diff --git a/benchmarks/edge_harness/experiments/prepare_spark_amd_npu_bundle.py b/benchmarks/edge_harness/experiments/prepare_spark_amd_npu_bundle.py new file mode 100644 index 000000000..264795569 --- /dev/null +++ b/benchmarks/edge_harness/experiments/prepare_spark_amd_npu_bundle.py @@ -0,0 +1,81 @@ +#!/usr/bin/env python3 +"""Package the qualified Spark output-head graph for Omni's graph stage. + +The model and quantized graph stay in local storage; the manifest records their +hashes and refuses a graph that did not pass the component placement gate. +""" + +from __future__ import annotations + +import argparse +import hashlib +import json +import shutil +from pathlib import Path + +import numpy as np + + +def digest(path: Path) -> str: + value = hashlib.sha256() + with path.open("rb") as stream: + for chunk in iter(lambda: stream.read(8 * 1024 * 1024), b""): + value.update(chunk) + return value.hexdigest() + + +def main() -> None: + parser = argparse.ArgumentParser(description=__doc__) + parser.add_argument("--graph", type=Path, required=True) + parser.add_argument("--validation-report", type=Path, required=True) + parser.add_argument("--activations", type=Path, required=True) + parser.add_argument("--output-dir", type=Path, required=True) + parser.add_argument("--checkpoint-revision", required=True) + args = parser.parse_args() + + report = json.loads(args.validation_report.read_text(encoding="utf-8")) + if not report.get("composite_with_norm") or report["npu_node_events"] < 4: + raise ValueError("The graph is not a placement-verified complete output head") + if not all(row["finite"] and row["top1_match"] and row["snr_db"] > 40 + for row in report["npu_rows"]): + raise ValueError("The output head failed real-activation numerical checks") + if digest(args.graph) != report["quantized_sha256"]: + raise ValueError("Graph hash differs from the validated artifact") + if digest(args.activations) != report["calibration_sha256"]: + raise ValueError("Calibration/reference hash differs from the validated inputs") + + with np.load(args.activations, allow_pickle=False) as data: + x = data["x"][0] + if x.shape != (1, 1, 2048) or x.dtype != np.float32: + raise ValueError(f"Unexpected real-activation shape or dtype: {x.shape}, {x.dtype}") + args.output_dir.mkdir(parents=True, exist_ok=True) + graph = args.output_dir / "graph.onnx" + examples = args.output_dir / "inputs.npz" + validation = args.output_dir / "validation_report.json" + shutil.copy2(args.graph, graph) + shutil.copy2(args.validation_report, validation) + np.savez(examples, x=x) + files = {path.name: digest(path) for path in (graph, examples, validation)} + manifest = { + "schema_version": 1, + "component": "spark_x2_5_1_7b_output_head_amd_npu", + "files": files, + "metadata": { + "graph_file": graph.name, + "example_inputs_file": examples.name, + "checkpoint_revision": args.checkpoint_revision, + "precision": "A16W8 QDQ, FP32 I/O", + "layout": "[1,1,2048] pre-final-norm input -> [1,131072] logits", + "exporter": "probe_spark_amd_npu_lm_head.py --composite-with-norm", + "calibration": "four captured real Spark pre-final-norm activations", + "validation": "validation_report.json; four top-1 matches, SNR >40 dB; component only", + }, + } + (args.output_dir / "manifest.json").write_text( + json.dumps(manifest, indent=2) + "\n", encoding="utf-8" + ) + print(json.dumps({"bundle": str(args.output_dir), "files": files}, indent=2)) + + +if __name__ == "__main__": + main() diff --git a/benchmarks/edge_harness/experiments/probe_spark_amd_npu_full_head.py b/benchmarks/edge_harness/experiments/probe_spark_amd_npu_full_head.py new file mode 100644 index 000000000..bc7825c38 --- /dev/null +++ b/benchmarks/edge_harness/experiments/probe_spark_amd_npu_full_head.py @@ -0,0 +1,200 @@ +#!/usr/bin/env python3 +"""Reassemble four real-weight Spark output-head shards on one Windows NPU. + +RMS normalization and QDQ boundary nodes may execute on CPU. This probes only +the output head; it does not qualify generation, KV state or an Omni handoff. +""" + +from __future__ import annotations + +import argparse +import gc +import json +import math +import os +import platform +import time +from pathlib import Path + +import numpy as np + +from probe_spark_amd_npu_lm_head import compare, sha256 + + +def percentile_nearest_rank(values: list[float], fraction: float) -> float: + ordered = sorted(values) + return ordered[max(0, math.ceil(len(ordered) * fraction) - 1)] + + +def main() -> None: + parser = argparse.ArgumentParser(description=__doc__) + parser.add_argument("--source-model", type=Path, required=True) + parser.add_argument("--activations", type=Path, required=True) + parser.add_argument("--shard-root", type=Path, required=True) + parser.add_argument("--ep-dir", type=Path, required=True) + parser.add_argument("--output-dir", type=Path, required=True) + parser.add_argument("--repeats", type=int, default=20) + args = parser.parse_args() + if args.repeats < 1: + raise ValueError("repeats must be positive") + args.output_dir.mkdir(parents=True, exist_ok=True) + + import onnx + import onnxruntime as ort + from onnx import numpy_helper + + data = np.load(args.activations) + inputs, references = data["x"], data["hidden_out"] + if inputs.shape != (4, 1, 1, 2048) or references.shape != (4, 1, 131072): + raise ValueError(f"Unexpected real-activation layout: {inputs.shape}, {references.shape}") + + started = time.perf_counter() + cpu = ort.InferenceSession(str(args.source_model), providers=["CPUExecutionProvider"]) + cpu_load_s = time.perf_counter() - started + cpu_rows = [] + for i in range(len(inputs)): + started = time.perf_counter() + actual = cpu.run(None, {"x": inputs[i]})[0] + row = compare(references[i], actual) + row.update(index=i, wall_s=time.perf_counter() - started) + cpu_rows.append(row) + cpu.run(None, {"x": inputs[0]}) + cpu_times = [] + for _ in range(args.repeats): + started = time.perf_counter() + cpu.run(None, {"x": inputs[0]}) + cpu_times.append(time.perf_counter() - started) + del cpu + gc.collect() + + source = onnx.load(str(args.source_model)) + norm = numpy_helper.to_array(next(item for item in source.graph.initializer if item.name == "norm")) + del source + + ep_dir = args.ep_dir.resolve() + os.environ["PATH"] = str(ep_dir) + os.pathsep + os.environ.get("PATH", "") + os.add_dll_directory(str(ep_dir)) + ort.register_execution_provider_library("vitisai", str(ep_dir / "onnxruntime_vitisai_ep.dll")) + devices = [ + device for device in ort.get_ep_devices() + if device.ep_name == "vitisai" and str(device.device.type).endswith("NPU") + ] + if not devices: + raise RuntimeError("VitisAI EP did not expose an NPU device") + + sessions = [] + shard_load_s = [] + shard_artifacts = [] + for offset in (0, 32768, 65536, 98304): + folder = args.shard_root / f"shard_{offset}" + artifact = folder / "spark_lm_head_a16w8_qdq.onnx" + manifest = json.loads((folder / "report.json").read_text(encoding="utf-8")) + if (manifest["vocab_shard_offset"] != offset or manifest["vocab_shard_size"] != 32768 + or manifest["quantized_sha256"] != sha256(artifact) + or manifest["npu_node_events"] == 0): + raise ValueError(f"Unqualified or changed output-head shard: {folder}") + options = ort.SessionOptions() + options.add_provider_for_devices(devices, {}) + options.enable_profiling = True + options.profile_file_prefix = str(args.output_dir / f"shard_{offset}_profile") + started = time.perf_counter() + session = ort.InferenceSession(str(artifact), sess_options=options) + shard_load_s.append(time.perf_counter() - started) + sessions.append(session) + shard_artifacts.append({ + "offset": offset, + "quantized_sha256": manifest["quantized_sha256"], + "providers": session.get_providers(), + }) + + def execute(x: np.ndarray) -> np.ndarray: + flat = x.reshape(1, 2048) + normalized = flat * ( + 1.0 / np.sqrt(np.mean(flat**2, axis=-1, keepdims=True) + np.float32(1e-6)) + ) + normalized = np.ascontiguousarray(normalized * norm) + return np.concatenate( + [session.run(None, {session.get_inputs()[0].name: normalized})[0] + for session in sessions], + axis=-1, + ) + + npu_rows = [] + for i in range(len(inputs)): + started = time.perf_counter() + actual = execute(inputs[i]) + row = compare(references[i], actual) + row.update( + index=i, + wall_s=time.perf_counter() - started, + shard_checks=[ + {"offset": offset, **compare( + references[i, :, offset : offset + 32768], + actual[:, offset : offset + 32768], + )} + for offset in (0, 32768, 65536, 98304) + ], + ) + npu_rows.append(row) + execute(inputs[0]) + npu_times = [] + for _ in range(args.repeats): + started = time.perf_counter() + execute(inputs[0]) + npu_times.append(time.perf_counter() - started) + + placement = [] + for artifact, session in zip(shard_artifacts, sessions, strict=True): + profile_path = Path(session.end_profiling()) + events = json.loads(profile_path.read_text(encoding="utf-8")) + providers = [ + (event.get("args") or {}).get("provider") + for event in events if event.get("cat") == "Node" + ] + placement.append({ + "offset": artifact["offset"], + "npu_node_events": providers.count("vitisai"), + "cpu_node_events": providers.count("CPUExecutionProvider"), + "profile_file": str(profile_path), + }) + + report = { + "scope": "Spark-X2.5-1.7B complete 131072-logit output head only; CPU RMS normalization plus four NPU MatMul shards, no full generation", + "hardware": "AMD Ryzen AI 9 HX 370 NPU", + "os": platform.platform(), + "onnx_version": onnx.__version__, + "onnxruntime_version": ort.__version__, + "source_model_sha256": sha256(args.source_model), + "activation_sha256": sha256(args.activations), + "source_shape": list(inputs.shape[1:]), + "output_shape": list(references.shape[1:]), + "cpu_load_s": cpu_load_s, + "shard_load_s": shard_load_s, + "shard_artifacts": shard_artifacts, + "placement": placement, + "cpu_reference_rows": cpu_rows, + "npu_rows": npu_rows, + "timing_protocol": f"one warmup then {args.repeats} serial repeats of captured activation 0, host wall time including normalization, NPU transfers, four sessions and concatenation", + "cpu_wall_s": cpu_times, + "npu_wall_s": npu_times, + "cpu_wall_p50_s": percentile_nearest_rank(cpu_times, 0.5), + "cpu_wall_p95_s": percentile_nearest_rank(cpu_times, 0.95), + "npu_wall_p50_s": percentile_nearest_rank(npu_times, 0.5), + "npu_wall_p95_s": percentile_nearest_rank(npu_times, 0.95), + } + (args.output_dir / "report.json").write_text(json.dumps(report, indent=2) + "\n", encoding="utf-8") + print(json.dumps({ + "placement": placement, + "npu_rows": npu_rows, + "cpu_wall_p50_s": report["cpu_wall_p50_s"], + "npu_wall_p50_s": report["npu_wall_p50_s"], + }, indent=2)) + if not all(item["npu_node_events"] > 0 for item in placement): + raise RuntimeError("At least one full-head shard fell back entirely to CPU") + if not all(item["top1_match"] and item["finite"] and item["snr_db"] > 40 + for item in npu_rows): + raise RuntimeError("Full-vocabulary NPU head failed parity on real activations") + + +if __name__ == "__main__": + main() diff --git a/benchmarks/edge_harness/experiments/probe_spark_amd_npu_lm_head.py b/benchmarks/edge_harness/experiments/probe_spark_amd_npu_lm_head.py new file mode 100644 index 000000000..ce663387e --- /dev/null +++ b/benchmarks/edge_harness/experiments/probe_spark_amd_npu_lm_head.py @@ -0,0 +1,371 @@ +#!/usr/bin/env python3 +"""Qualify one real Spark output-head graph on the Windows Ryzen AI NPU. + +This is a model-component probe, not a Spark generation or Omni pipeline pass. +The calibration file contains real pre-final-norm activations captured from the +matching Spark-X2.5-1.7B checkpoint and reference output-head logits. +""" + +from __future__ import annotations + +import argparse +import hashlib +import json +import os +import platform +import time +from pathlib import Path + +import numpy as np + + +def sha256(path: Path) -> str: + digest = hashlib.sha256() + with path.open("rb") as stream: + for chunk in iter(lambda: stream.read(8 * 1024 * 1024), b""): + digest.update(chunk) + return digest.hexdigest() + + +def compare(reference: np.ndarray, actual: np.ndarray) -> dict: + reference = np.asarray(reference, dtype=np.float32) + actual = np.asarray(actual, dtype=np.float32) + if reference.shape != actual.shape: + raise ValueError(f"Output shape changed: {reference.shape} vs {actual.shape}") + error = reference.astype(np.float64) - actual.astype(np.float64) + norm = np.linalg.norm(reference.astype(np.float64)) + return { + "snr_db": float(20 * np.log10(norm / max(np.linalg.norm(error), 1e-12))), + "max_abs_diff": float(np.max(np.abs(error))), + "top1_match": bool(np.argmax(reference) == np.argmax(actual)), + "finite": bool(np.isfinite(actual).all()), + } + + +def main() -> None: + parser = argparse.ArgumentParser(description=__doc__) + parser.add_argument("--model", type=Path, required=True) + parser.add_argument("--calibration", type=Path, required=True) + parser.add_argument("--ep-dir", type=Path, required=True) + parser.add_argument("--output-dir", type=Path, required=True) + parser.add_argument("--hardware-label", default="AMD Ryzen AI 9 HX 370 NPU") + parser.add_argument("--vocab-shard-size", type=int, default=0, + help="Probe the first N real output-head columns; 0 keeps the full head") + parser.add_argument("--vocab-shard-offset", type=int, default=0, + help="Starting output column for a real-weight vocabulary shard") + parser.add_argument("--signed-activations", action="store_true", + help="Use QInt16 rather than QUInt16 activations") + parser.add_argument("--flatten-input", action="store_true", + help="Use equivalent [1, 2048] input and remove final singleton-axis Gather") + parser.add_argument("--batch-repeats", type=int, default=1, + help="Repeat each real activation N times in a batch (requires --flatten-input)") + parser.add_argument("--matmul-only", action="store_true", + help="Move exact RMS normalization to the CPU side and probe only real-weight MatMul") + parser.add_argument("--composite-shards", action="store_true", + help="Build four named 32768-column MatMuls and concatenate in one ONNX graph") + parser.add_argument("--composite-with-norm", action="store_true", + help="Keep the checkpoint RMS norm before four named MatMul shards") + args = parser.parse_args() + + import onnx + import onnxruntime as ort + from onnxruntime.quantization import ( + CalibrationDataReader, + CalibrationMethod, + QuantFormat, + QuantType, + quantize_static, + ) + + args.output_dir.mkdir(parents=True, exist_ok=True) + data = np.load(args.calibration) + inputs = data["x"] + references = data["hidden_out"] + if inputs.shape != (4, 1, 1, 2048) or references.shape != (4, 1, 131072): + raise ValueError(f"Unexpected calibration layout: {inputs.shape}, {references.shape}") + if (args.vocab_shard_size < 0 or args.vocab_shard_offset < 0 + or args.vocab_shard_offset + (args.vocab_shard_size or 131072) > references.shape[-1]): + raise ValueError("The requested vocabulary shard must fit within 131072 columns") + if args.batch_repeats < 1 or (args.batch_repeats > 1 and not args.flatten_input): + raise ValueError("batch-repeats must be positive and requires --flatten-input") + if args.matmul_only and not args.flatten_input: + raise ValueError("matmul-only requires --flatten-input") + if args.composite_shards and (not args.matmul_only or args.vocab_shard_size + or args.vocab_shard_offset or args.batch_repeats != 1): + raise ValueError("composite-shards requires matmul-only/flatten-input and a full batch-1 head") + if args.composite_with_norm and (args.matmul_only or args.flatten_input + or args.vocab_shard_size or args.vocab_shard_offset + or args.batch_repeats != 1): + raise ValueError("composite-with-norm requires the original full head and batch-1 input") + input_name = (f"normalized_x_offset_{args.vocab_shard_offset}" + if args.matmul_only else "x") + + prepared_model = args.model + if args.composite_with_norm: + from onnx import TensorProto, helper, numpy_helper + + prepared_model = args.output_dir / "spark_lm_head_norm_4shards.onnx" + if not prepared_model.exists(): + original = onnx.load(str(args.model)) + if [node.op_type for node in original.graph.node[9:11]] != ["Mul", "Mul"]: + raise ValueError("Unexpected original RMS-norm graph") + weight = numpy_helper.to_array(next( + item for item in original.graph.initializer if item.name == "onnx::MatMul_24" + )) + nodes = list(original.graph.node[:11]) + nodes.append(helper.make_node( + "Reshape", ["/Mul_1_output_0", "flat_shape"], ["normalized_x_flat"], + name="spark_lm_head_flatten_norm", + )) + initializers = [ + next(item for item in original.graph.initializer if item.name == "norm"), + numpy_helper.from_array(np.array([1, 2048], dtype=np.int64), name="flat_shape"), + ] + outputs = [] + for offset in (0, 32768, 65536, 98304): + output_name = f"logits_offset_{offset}" + weight_name = f"weight_offset_{offset}" + nodes.append(helper.make_node( + "MatMul", ["normalized_x_flat", weight_name], [output_name], + name=f"spark_lm_head_matmul_offset_{offset}", + )) + initializers.append(numpy_helper.from_array( + np.ascontiguousarray(weight[:, offset : offset + 32768]), name=weight_name + )) + outputs.append(output_name) + nodes.append(helper.make_node( + "Concat", outputs, ["logits_concatenated"], axis=1, + name="spark_lm_head_concat", + )) + graph = helper.make_graph( + nodes, "SparkRealWeightOutputHeadNormFourShards", + [helper.make_tensor_value_info("x", TensorProto.FLOAT, [1, 1, 2048])], + [helper.make_tensor_value_info("logits_concatenated", TensorProto.FLOAT, + [1, 131072])], + initializer=initializers, + ) + prepared = helper.make_model(graph, opset_imports=[helper.make_opsetid("", 21)]) + prepared.ir_version = min(prepared.ir_version, 10) + onnx.checker.check_model(prepared) + onnx.save(prepared, str(prepared_model)) + del original, weight + elif args.matmul_only: + from onnx import TensorProto, helper, numpy_helper + + prepared_model = args.output_dir / ( + "spark_lm_head_matmul_4shards.onnx" if args.composite_shards else + f"spark_lm_head_matmul_offset{args.vocab_shard_offset}" + f"_size{args.vocab_shard_size or 131072}" + f"_batch{args.batch_repeats}.onnx" + ) + model = onnx.load(str(args.model)) + norm = numpy_helper.to_array(next(item for item in model.graph.initializer if item.name == "norm")) + weight = numpy_helper.to_array(next( + item for item in model.graph.initializer if item.name == "onnx::MatMul_24" + )) + inputs = inputs.reshape(len(inputs), 1, 2048) + inputs = inputs * (1.0 / np.sqrt(np.mean(inputs**2, axis=-1, keepdims=True) + np.float32(1e-6))) + inputs = np.ascontiguousarray(inputs * norm) + inputs = np.repeat(inputs, args.batch_repeats, axis=1) + references = np.repeat(references, args.batch_repeats, axis=1) + if args.vocab_shard_size: + end = args.vocab_shard_offset + args.vocab_shard_size + references = references[..., args.vocab_shard_offset : end] + weight = weight[:, args.vocab_shard_offset : end] + if not prepared_model.exists(): + offsets = (0, 32768, 65536, 98304) if args.composite_shards else (args.vocab_shard_offset,) + nodes = [] + initializers = [] + outputs = [] + for offset in offsets: + output_name = f"logits_offset_{offset}" + weight_name = f"weight_offset_{offset}" + shard_weight = (weight[:, offset : offset + 32768] + if args.composite_shards else weight) + nodes.append(helper.make_node( + "MatMul", [input_name, weight_name], [output_name], + name=f"spark_lm_head_matmul_offset_{offset}", + )) + initializers.append(numpy_helper.from_array( + np.ascontiguousarray(shard_weight), name=weight_name + )) + outputs.append(output_name) + final_output = outputs[0] + if args.composite_shards: + final_output = "logits_concatenated" + nodes.append(helper.make_node("Concat", outputs, [final_output], axis=1, + name="spark_lm_head_concat")) + graph = helper.make_graph( + nodes, + "SparkRealWeightOutputHeadFourShards" if args.composite_shards else + f"SparkRealWeightOutputHeadMatMulOffset{args.vocab_shard_offset}", + [helper.make_tensor_value_info(input_name, TensorProto.FLOAT, + [args.batch_repeats, 2048])], + [helper.make_tensor_value_info(final_output, TensorProto.FLOAT, + [args.batch_repeats, references.shape[-1]])], + initializer=initializers, + ) + prepared = helper.make_model(graph, opset_imports=[helper.make_opsetid("", 21)]) + prepared.ir_version = min(prepared.ir_version, 10) + onnx.checker.check_model(prepared) + onnx.save(prepared, str(prepared_model)) + del model, weight + elif args.vocab_shard_size or args.flatten_input: + from onnx import numpy_helper + + suffix = (f"offset{args.vocab_shard_offset}_size{args.vocab_shard_size}" + if args.vocab_shard_size else "full") + if args.flatten_input: + suffix += f"_flat_batch{args.batch_repeats}" + inputs = inputs.reshape(len(inputs), 1, 2048) + inputs = np.repeat(inputs, args.batch_repeats, axis=1) + references = np.repeat(references, args.batch_repeats, axis=1) + prepared_model = args.output_dir / f"spark_lm_head_{suffix}.onnx" + if args.vocab_shard_size: + references = references[..., args.vocab_shard_offset : args.vocab_shard_offset + args.vocab_shard_size] + if not prepared_model.exists(): + model = onnx.load(str(args.model)) + if args.vocab_shard_size: + weight = next(item for item in model.graph.initializer if item.name == "onnx::MatMul_24") + matrix = numpy_helper.to_array(weight) + if matrix.shape != (2048, 131072): + raise ValueError(f"Unexpected output-head layout: {matrix.shape}") + weight.CopyFrom(numpy_helper.from_array( + np.ascontiguousarray(matrix[:, args.vocab_shard_offset : + args.vocab_shard_offset + args.vocab_shard_size]), + name=weight.name + )) + model.graph.output[0].type.tensor_type.shape.dim[-1].dim_value = args.vocab_shard_size + if args.flatten_input: + if [node.op_type for node in model.graph.node[-2:]] != ["Constant", "Gather"]: + raise ValueError("Expected final singleton-axis Gather") + del model.graph.input[0].type.tensor_type.shape.dim[1] + model.graph.input[0].type.tensor_type.shape.dim[0].dim_value = args.batch_repeats + model.graph.output[0].name = model.graph.node[-3].output[0] + model.graph.output[0].type.tensor_type.shape.dim[0].dim_value = args.batch_repeats + del model.graph.node[-2:] + onnx.checker.check_model(model) + onnx.save(model, str(prepared_model)) + del model + + source_session = ort.InferenceSession(str(prepared_model), providers=["CPUExecutionProvider"]) + source_rows = [] + for index in range(len(inputs)): + got = source_session.run(None, {input_name: inputs[index]})[0] + source_rows.append(compare(references[index], got)) + if not all(row["top1_match"] and row["snr_db"] > 100 for row in source_rows): + raise RuntimeError("Source ONNX graph failed real-activation reference check") + + opset21 = args.output_dir / "spark_lm_head_opset21.onnx" + quantized = args.output_dir / "spark_lm_head_a16w8_qdq.onnx" + if not opset21.exists(): + model = onnx.load(str(prepared_model)) + if [(item.domain, item.version) for item in model.opset_import] == [("", 18)]: + model = onnx.version_converter.convert_version(model, 21) + elif [(item.domain, item.version) for item in model.opset_import] != [("", 21)]: + raise ValueError("Expected the recorded opset-18 head or prepared opset-21 MatMul") + onnx.checker.check_model(model) + onnx.save(model, str(opset21)) + del model + + class RealActivationReader(CalibrationDataReader): + def __init__(self) -> None: + self.index = 0 + + def get_next(self) -> dict[str, np.ndarray] | None: + if self.index == len(inputs): + return None + sample = {input_name: inputs[self.index]} + self.index += 1 + return sample + + def rewind(self) -> None: + self.index = 0 + + if not quantized.exists(): + quantize_static( + str(opset21), + str(quantized), + RealActivationReader(), + quant_format=QuantFormat.QDQ, + activation_type=QuantType.QInt16 if args.signed_activations else QuantType.QUInt16, + weight_type=QuantType.QInt8, + per_channel=False, + op_types_to_quantize=["MatMul"], + calibrate_method=CalibrationMethod.MinMax, + ) + + ep_dir = args.ep_dir.resolve() + os.environ["PATH"] = str(ep_dir) + os.pathsep + os.environ.get("PATH", "") + os.add_dll_directory(str(ep_dir)) + ort.register_execution_provider_library("vitisai", str(ep_dir / "onnxruntime_vitisai_ep.dll")) + devices = [ + device for device in ort.get_ep_devices() + if device.ep_name == "vitisai" and str(device.device.type).endswith("NPU") + ] + if not devices: + raise RuntimeError("VitisAI EP did not expose an NPU device") + options = ort.SessionOptions() + options.add_provider_for_devices(devices, {}) + options.enable_profiling = True + options.profile_file_prefix = str(args.output_dir / "spark_lm_head_npu_profile") + started = time.perf_counter() + session = ort.InferenceSession(str(quantized), sess_options=options) + load_s = time.perf_counter() - started + rows = [] + for index in range(len(inputs)): + started = time.perf_counter() + result = session.run(None, {input_name: inputs[index]})[0] + row = compare(references[index], result) + row.update(index=index, wall_s=time.perf_counter() - started) + rows.append(row) + profile_path = Path(session.end_profiling()) + profile = json.loads(profile_path.read_text(encoding="utf-8")) + node_providers = [ + (event.get("args") or {}).get("provider") + for event in profile if event.get("cat") == "Node" + ] + report = { + "scope": "Spark-X2.5-1.7B output head only; no prefill/decode/KV/sampling or full generation", + "vocab_shard_size": args.vocab_shard_size or 131072, + "vocab_shard_offset": args.vocab_shard_offset, + "flatten_input": args.flatten_input, + "batch_repeats": args.batch_repeats, + "matmul_only": args.matmul_only, + "composite_shards": args.composite_shards, + "composite_with_norm": args.composite_with_norm, + "hardware": args.hardware_label, + "os": platform.platform(), + "onnx_version": onnx.__version__, + "onnxruntime_version": ort.__version__, + "source_model_sha256": sha256(args.model), + "prepared_model_sha256": sha256(prepared_model), + "calibration_sha256": sha256(args.calibration), + "opset21_sha256": sha256(opset21), + "quantized_sha256": sha256(quantized), + "quantization": ( + "opset 21, MatMul only, QDQ, " + f"{'QInt16' if args.signed_activations else 'QUInt16'} activations/QInt8 weights, " + "MinMax on four real activations" + ), + "input_shape": list(inputs.shape[1:]), + "output_shape": list(references.shape[1:]), + "source_cpu_reference": source_rows, + "npu_devices": len(devices), + "session_providers": session.get_providers(), + "npu_node_events": node_providers.count("vitisai"), + "cpu_node_events": node_providers.count("CPUExecutionProvider"), + "load_s": load_s, + "npu_rows": rows, + "profile_file": str(profile_path), + } + (args.output_dir / "report.json").write_text(json.dumps(report, indent=2) + "\n", encoding="utf-8") + print(json.dumps(report, indent=2)) + if report["npu_node_events"] == 0: + raise RuntimeError("Graph returned output but no node executed on the NPU") + if not all(row["top1_match"] and row["finite"] for row in rows): + raise RuntimeError("NPU output failed finite/top-1 parity on real activations") + + +if __name__ == "__main__": + main() diff --git a/benchmarks/edge_harness/results/e2e_expansion_20260923/README.md b/benchmarks/edge_harness/results/e2e_expansion_20260923/README.md index 3a7a84aca..492a3ec26 100644 --- a/benchmarks/edge_harness/results/e2e_expansion_20260923/README.md +++ b/benchmarks/edge_harness/results/e2e_expansion_20260923/README.md @@ -1,9 +1,10 @@ # 2026-09-23 WSL CPU and native Windows expansion -This extends the [60-cell device × model matrix](../e2e_profiling_20260922/evidence/summary/README.md) with previously unrun WSL CPU and native Windows paths. The September 22 timing tables remain historical. MiniCPM-o now has separate 20-request serial text-to-speech timing runs on WSL CPU and native Windows GPU, plus one synthetic image and one synthetic audio input on each platform; each InternVLA timing remains one observation. None is a sustained or full-modality profile. The [checkpoint manifests](../e2e_recovery_20260923/README.md) pin the real weights. No model, precision or context was silently substituted. +This extends the [60-cell device × model matrix](../e2e_profiling_20260922/evidence/summary/README.md) with previously unrun WSL CPU and native Windows paths. The September 22 timing tables remain historical. MiniCPM-o now has separate 20-request serial text-to-speech timing runs on WSL CPU and native Windows GPU, plus one synthetic image and one synthetic audio input on each platform; each InternVLA timing remains one observation. A Spark output-head component also has a separate 20-request Omni graph-stage profile on the HX370 AMD NPU. None is a sustained or full-modality profile. The [checkpoint manifests](../e2e_recovery_20260923/README.md) pin the earlier real weights; the [Spark component record](evidence/spark_amd_npu_output_head/README.md) pins its checkpoint revision and graph hashes. No model, precision or context was silently substituted. | Device / model | Verified depth | Observed output and time | Remaining gate | |---|---|---|---| +| Native Windows HX370 AMD NPU / Spark-X2.5-1.7B output head, revision `448e61eb392c00f2c403185c5b56d5e0665bfaab` | Real-weight RMS norm + full-logit projection component through Omni `external.graph.v1` (B+C, not P) | One fused NPU partition and seven CPU nodes; 4/4 captured activations top-1 match, 55.43–56.81 dB full-logit SNR. One warmup + 20 serial repeated-input component requests: p50/p95 10.21/12.44 ms, versus 17.97/20.29 ms for the original FP32 CPU head through Omni. NPU startup 18.21 s and peak worker RSS 3.52 GiB. [Raw evidence](evidence/spark_amd_npu_output_head/README.md). | Actual vLLM handoff and full generation, transfer cost, quality across inputs/context, memory admission, cancellation and sustained power/thermal behavior. | | WSL CPU / MiniCPM-o 4.5, revision `503e754207c94da6bb26850b4469f367c9ea3582` | Scoped complete three-stage text-to-text+speech path (P for this input only) | Coherent text and a 5.2 s mono 24 kHz WAV. First request 27.09 s after 73.65 s startup; separate 1-warmup/20-measured serial run: startup 73.89 s, request wall p50 21.86 s/p95 22.97 s. | Speech intelligibility/alignment, real image/audio quality, video input, concurrency, interruption, power and thermal behavior. | | WSL CPU / MiniCPM-o 4.5, same revision | Scoped three-stage **image**-to-text+speech path (P for one synthetic image) | For the 448×448 red square, returned a correct description and nonzero 5.2 s WAV. Startup 112.44 s; one request 37.70 s. Text and token IDs differed from WSL GPU while describing the same visible content. | Real-image suite, speech meaning, repeated latency, memory admission and clean shutdown. | | WSL CPU / MiniCPM-o 4.5, same revision | Scoped three-stage **audio**-to-text+speech path (P for one synthetic clip) | For the 2 s, 440 Hz tone, answered “A loud and high pitched beep” and produced a nonzero 2.16 s WAV. Startup 101.22 s; one request 17.30 s. Text and token IDs differed from WSL GPU. | Real-audio and transcription suites, speech meaning, repeated latency, memory admission and clean shutdown. | diff --git a/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/.gitattributes b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/.gitattributes new file mode 100644 index 000000000..09cca0b54 --- /dev/null +++ b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/.gitattributes @@ -0,0 +1,3 @@ +# Keep vendor profiles and hashed validation reports byte-for-byte. +*.json -text -eol +*.log -text -eol diff --git a/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/README.md b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/README.md new file mode 100644 index 000000000..f1c906a4b --- /dev/null +++ b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/README.md @@ -0,0 +1,33 @@ +# Spark-X2.5-1.7B output head on the HX370 AMD NPU + +**Depth: B+C, not P.** A real-weight output-head component passed numerical and placement checks through the Omni `external.graph.v1` Windows worker. No Spark prefill, continuous decode, KV state, sampling, output-head handoff from vLLM, or complete text request ran on this route. The [60-cell matrix](../../../e2e_profiling_20260922/evidence/summary/README.md) therefore remains **NOT E2E** for this pairing. + +The source is `XHToken/Spark-X2.5-1.7B` revision `448e61eb392c00f2c403185c5b56d5e0665bfaab`. The FP32 ONNX output head (SHA-256 `f5f0c31eb438e9940ab242da8bf6f0f5dad6ebc9f6ef02d5475c21b744e595bc`) takes a pre-final-norm `[1,1,2048]` activation and returns all `[1,131072]` logits. The four captured real activations and source-reference logits are [retained here](real_acts_lmhead.npz), SHA-256 `cfa2ed8439f8fe01730319667bfd01abf2e39cf5851490346a0331304cf63194`. The FP32 source graph matched those captured references at 133.66–137.37 dB SNR, top-1 4/4. These four activation cases are a small numerical check, not a task-quality suite. + +The qualified ONNX graph uses opset 21, RMS norm plus four distinct 32,768-column real-weight MatMul branches and Concat, MatMul-only QDQ with QUInt16 activations/QInt8 weights calibrated by MinMax on the four captured activations, and FP32 I/O. Its SHA-256 is `e2ae9c3fd7071e8ff628f209e6140c8a023bfaad59814b80a936c6b263f33eaf`. The 268 MB graph and 1.1 GB FP32 source remain in local artifact storage; they are not Git files. The [local bundle manifest record](local_bundle_manifest.json) pins the graph, input, validation, revision, layout and precision. That manifest references `graph.onnx` in the local bundle and cannot be executed from this checkout alone. [The preparation script](../../../../experiments/prepare_spark_amd_npu_bundle.py) verifies the graph and source hashes and numerical gate before making a local Omni bundle. + +Hardware was Ryzen AI 9 HX 370 on Windows 11 build 26200, with AMD NPU driver `32.0.203.329`, VitisAI Windows Workload EP `1.8.63.0`, ONNX Runtime `1.30.0`, ONNX `1.22.0`, Python `3.12.10`, and the current Omni source. The [direct ORT report](composite_report.json), [raw log](composite_raw.log), and [profile](composite_profile.json) show four actual VitisAI node events over four activations, with 28 CPU node events. Full-logit SNR was 55.43–56.81 dB with finite outputs and top-1 4/4. The [Omni NPU run](spark_npu_graph_stage_with_norm_20260923.json) admitted a 4 GiB shared-RAM reservation and verified **one fused NPU partition plus seven CPU nodes** per request (NPU fraction 0.125). Its [worker profile](omni_npu_placement_profile.json) is retained; there was no silent CPU-only substitution. The reservation was released after shutdown. + +The following timings are one warmup plus 20 serial repeats of **one captured activation** at concurrency 1, through the same Omni graph-stage boundary. They exclude whole-model inference and handoff; nearest-rank p50/p95 are reported, and power/thermal conditions were not captured. + +| Output-head stage | p50 / p95 | Startup | Peak worker RSS | Placement | +|---|---:|---:|---:|---| +| [A16W8 NPU](spark_npu_graph_stage_with_norm_20260923.json) | 10.21 / 12.44 ms | 18.21 s | 3.52 GiB | 1 NPU + 7 CPU nodes | +| [Original FP32 CPU](spark_cpu_fp32_graph_stage_20260923.json) | 17.97 / 20.29 ms | 4.48 s | 2.06 GiB | 3 CPU nodes | +| [Same A16W8 graph on CPU](spark_cpu_graph_stage_with_norm_20260923.json) | 69.90 / 82.83 ms | 2.04 s | 0.84 GiB | 24 CPU nodes | + +The NPU result was faster than this original FP32 CPU *component* baseline but used longer startup and more peak worker memory. The same-graph CPU comparison includes QDQ overhead. Neither comparison establishes a benefit for moving a token-by-token output head off the vLLM backend; that requires an actual handoff and full-generation measurement. + +Failed candidates remain evidence: the [monolithic QDQ output head](failures/full_qdq_report.json) and [full-width MatMul](failures/full_matmul_report.json) had **zero** NPU node events despite listing VitisAI; they ran on CPU. Four independently compiled, unnamed co-resident shards gave [bad full-logit numerics](failures/unnamed_co_resident_report.json), including one wrong top-1. Distinct graph/tensor names fixed the separate-session check ([report](full_head_report.json)); the final single composite graph avoids that separate-session arrangement. These failures apply to those graph formats, not to Spark or the NPU in general. + +Reproduce the direct composite probe on this host, after regenerating or locating the pinned FP32 source graph: + +```powershell +C:\Users\zhout\npu-ep\Scripts\python.exe benchmarks/edge_harness/experiments/probe_spark_amd_npu_lm_head.py ` + --model \\wsl.localhost\Ubuntu\home\zhout\project\edge_infer\analysis\experiments\spark_edge\onnx\lm_head.onnx ` + --calibration benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/real_acts_lmhead.npz ` + --ep-dir 'C:\Program Files\WindowsApps\WindowsWorkload.EP.AMD.VitisAI.Framework.1.8_1.8.63.0_x64__8wekyb3d8bbwe\ExecutionProvider' ` + --composite-with-norm --output-dir C:\Users\zhout\w2\spark_npu_lmhead_repro +``` + +Next gate: connect this exact component to a vLLM Spark stage, include transfer/synchronization and full memory loading in admission, validate numerical and token quality across prompts and context lengths, then profile complete generation, cancellation, repeated use and sustained power. Retain whole-model execution on its current backend unless the measured split benefits the full request. diff --git a/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/bundle_input.npz b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/bundle_input.npz new file mode 100644 index 000000000..87c2fcebf Binary files /dev/null and b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/bundle_input.npz differ diff --git a/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/composite_profile.json b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/composite_profile.json new file mode 100644 index 000000000..2c0c8e5ee --- /dev/null +++ b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/composite_profile.json @@ -0,0 +1,44 @@ +[ +{"cat" : "Session","pid" :46876,"tid" :17412,"dur" :141661,"ts" :13,"ph" : "X","name" :"model_loading_uri","args" : {}}, +{"cat" : "Session","pid" :46876,"tid" :17412,"dur" :14959979,"ts" :143380,"ph" : "X","name" :"session_initialization","args" : {}}, +{"cat" : "Node","pid" :46876,"tid" :17412,"dur" :106,"ts" :15103577,"ph" : "X","name" :"/Mul_1/SimplifiedLayerNormFusion/_kernel_time","args" : {"node_index" : "33","output_type_shape" : [{"float":[1,1,2048]}],"activation_size" : "8192","mem_requested_in_use_delta" : "8192","input_type_shape" : [{"float":[1,1,2048]},{"float":[2048]}],"mem_in_use_peak" : "8222","mem_in_use_delta" : "8192","mem_arena_held_delta" : "1048576","mem_arena_held" : "1048606","parameter_size" : "8192","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "17412", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"34996": {"num_run": 0, "core": -1},"55368": {"num_run": 0, "core": -1},"28884": {"num_run": 0, "core": -1},"59012": {"num_run": 0, "core": -1},"23932": {"num_run": 0, "core": -1},"56800": {"num_run": 0, "core": -1},"36528": {"num_run": 0, "core": -1},"32820": {"num_run": 0, "core": -1},"34664": {"num_run": 0, "core": -1},"37516": {"num_run": 0, "core": -1},"20348": {"num_run": 0, "core": -1}}},"mem_bytes_requested_in_use" : "8222","op_name" : "SimplifiedLayerNormalization","mem_bytes_in_use" : "8222","provider" : "CPUExecutionProvider","output_size" : "8192"}}, +{"cat" : "Node","pid" :46876,"tid" :17412,"dur" :49,"ts" :15103693,"ph" : "X","name" :"/Mul_1_output_0_q_kernel_time","args" : {"mem_bytes_in_use" : "12318","mem_arena_held" : "1048606","provider" : "CPUExecutionProvider","mem_in_use_delta" : "4096","mem_requested_in_use_delta" : "4096","output_size" : "4096","node_index" : "30","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "17412", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"34996": {"num_run": 0, "core": -1},"55368": {"num_run": 0, "core": -1},"28884": {"num_run": 0, "core": -1},"59012": {"num_run": 0, "core": -1},"23932": {"num_run": 0, "core": -1},"56800": {"num_run": 0, "core": -1},"36528": {"num_run": 0, "core": -1},"32820": {"num_run": 0, "core": -1},"34664": {"num_run": 0, "core": -1},"37516": {"num_run": 0, "core": -1},"20348": {"num_run": 0, "core": -1}}},"mem_arena_held_delta" : "0","mem_bytes_requested_in_use" : "12318","op_name" : "QuantizeLinear","activation_size" : "8192","input_type_shape" : [{"float":[1,1,2048]},{"float":[]},{"uint16":[]}],"mem_in_use_peak" : "12318","output_type_shape" : [{"uint16":[1,1,2048]}],"parameter_size" : "6"}}, +{"cat" : "Node","pid" :46876,"tid" :17412,"dur" :9423,"ts" :15103752,"ph" : "X","name" :"vitisai_vitisai_2852173091035858231_0_0_kernel_time","args" : {"mem_bytes_requested_in_use" : "266270","parameter_size" : "0","mem_arena_held_delta" : "0","node_index" : "32","input_type_shape" : [{"uint16":[1,1,2048]}],"mem_in_use_peak" : "266270","output_type_shape" : [{"uint16":[1,32768]},{"uint16":[1,32768]},{"uint16":[1,32768]},{"uint16":[1,32768]}],"mem_bytes_in_use" : "266270","provider" : "vitisai","activation_size" : "4096","mem_in_use_delta" : "262144","op_name" : "vitisai_2852173091035858231_0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "17412", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"34996": {"num_run": 0, "core": -1},"55368": {"num_run": 0, "core": -1},"28884": {"num_run": 0, "core": -1},"59012": {"num_run": 0, "core": -1},"23932": {"num_run": 0, "core": -1},"56800": {"num_run": 0, "core": -1},"36528": {"num_run": 0, "core": -1},"32820": {"num_run": 0, "core": -1},"34664": {"num_run": 0, "core": -1},"37516": {"num_run": 0, "core": -1},"20348": {"num_run": 0, "core": -1}}},"mem_arena_held" : "1048606","mem_requested_in_use_delta" : "262144","output_size" : "262144"}}, +{"cat" : "Node","pid" :46876,"tid" :17412,"dur" :61,"ts" :15113198,"ph" : "X","name" :"logits_offset_98304_DequantizeLinear_kernel_time","args" : {"mem_arena_held_delta" : "0","node_index" : "25","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_bytes_in_use" : "393246","mem_in_use_peak" : "393246","op_name" : "DequantizeLinear","output_type_shape" : [{"float":[1,32768]}],"mem_in_use_delta" : "131072","provider" : "CPUExecutionProvider","mem_requested_in_use_delta" : "131072","parameter_size" : "6","mem_arena_held" : "1048606","output_size" : "131072","activation_size" : "65536","mem_bytes_requested_in_use" : "393246","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "17412", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"34996": {"num_run": 0, "core": -1},"55368": {"num_run": 0, "core": -1},"28884": {"num_run": 0, "core": -1},"59012": {"num_run": 0, "core": -1},"23932": {"num_run": 0, "core": -1},"56800": {"num_run": 0, "core": -1},"36528": {"num_run": 0, "core": -1},"32820": {"num_run": 0, "core": -1},"34664": {"num_run": 0, "core": -1},"37516": {"num_run": 0, "core": -1},"20348": {"num_run": 0, "core": -1}}}}}, +{"cat" : "Node","pid" :46876,"tid" :17412,"dur" :48,"ts" :15113268,"ph" : "X","name" :"logits_offset_65536_DequantizeLinear_kernel_time","args" : {"input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_requested_in_use_delta" : "131072","op_name" : "DequantizeLinear","output_size" : "131072","parameter_size" : "6","mem_arena_held" : "1048606","output_type_shape" : [{"float":[1,32768]}],"mem_bytes_requested_in_use" : "458782","mem_in_use_delta" : "131072","node_index" : "24","mem_in_use_peak" : "458782","mem_arena_held_delta" : "0","provider" : "CPUExecutionProvider","activation_size" : "65536","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "17412", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"34996": {"num_run": 0, "core": -1},"55368": {"num_run": 0, "core": -1},"28884": {"num_run": 0, "core": -1},"59012": {"num_run": 0, "core": -1},"23932": {"num_run": 0, "core": -1},"56800": {"num_run": 0, "core": -1},"36528": {"num_run": 0, "core": -1},"32820": {"num_run": 0, "core": -1},"34664": {"num_run": 0, "core": -1},"37516": {"num_run": 0, "core": -1},"20348": {"num_run": 0, "core": -1}}},"mem_bytes_in_use" : "458782"}}, +{"cat" : "Node","pid" :46876,"tid" :17412,"dur" :32,"ts" :15113323,"ph" : "X","name" :"logits_offset_32768_DequantizeLinear_kernel_time","args" : {"mem_bytes_requested_in_use" : "524318","node_index" : "23","mem_arena_held" : "1048606","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "17412", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"34996": {"num_run": 0, "core": -1},"55368": {"num_run": 0, "core": -1},"28884": {"num_run": 0, "core": -1},"59012": {"num_run": 0, "core": -1},"23932": {"num_run": 0, "core": -1},"56800": {"num_run": 0, "core": -1},"36528": {"num_run": 0, "core": -1},"32820": {"num_run": 0, "core": -1},"34664": {"num_run": 0, "core": -1},"37516": {"num_run": 0, "core": -1},"20348": {"num_run": 0, "core": -1}}},"mem_arena_held_delta" : "0","mem_requested_in_use_delta" : "131072","op_name" : "DequantizeLinear","mem_bytes_in_use" : "536606","provider" : "CPUExecutionProvider","mem_in_use_delta" : "143360","mem_in_use_peak" : "536606","output_type_shape" : [{"float":[1,32768]}],"input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"output_size" : "131072","activation_size" : "65536","parameter_size" : "6"}}, +{"cat" : "Node","pid" :46876,"tid" :17412,"dur" :46,"ts" :15113362,"ph" : "X","name" :"logits_offset_0_DequantizeLinear_kernel_time","args" : {"mem_in_use_peak" : "602142","node_index" : "22","mem_bytes_requested_in_use" : "589854","activation_size" : "65536","op_name" : "DequantizeLinear","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "17412", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"34996": {"num_run": 0, "core": -1},"55368": {"num_run": 0, "core": -1},"28884": {"num_run": 0, "core": -1},"59012": {"num_run": 0, "core": -1},"23932": {"num_run": 0, "core": -1},"56800": {"num_run": 0, "core": -1},"36528": {"num_run": 0, "core": -1},"32820": {"num_run": 0, "core": -1},"34664": {"num_run": 0, "core": -1},"37516": {"num_run": 0, "core": -1},"20348": {"num_run": 0, "core": -1}}},"input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_in_use_delta" : "131072","mem_arena_held_delta" : "0","mem_requested_in_use_delta" : "131072","output_type_shape" : [{"float":[1,32768]}],"mem_bytes_in_use" : "602142","provider" : "CPUExecutionProvider","parameter_size" : "6","mem_arena_held" : "1048606","output_size" : "131072"}}, +{"cat" : "Node","pid" :46876,"tid" :17412,"dur" :164,"ts" :15113414,"ph" : "X","name" :"spark_lm_head_concat_kernel_time","args" : {"mem_bytes_in_use" : "1060894","output_type_shape" : [{"float":[1,131072]}],"input_type_shape" : [{"float":[1,32768]},{"float":[1,32768]},{"float":[1,32768]},{"float":[1,32768]}],"mem_bytes_requested_in_use" : "1048606","output_size" : "524288","mem_in_use_peak" : "1060894","node_index" : "26","mem_in_use_delta" : "524288","mem_requested_in_use_delta" : "524288","mem_arena_held" : "3145758","provider" : "CPUExecutionProvider","mem_arena_held_delta" : "2097152","parameter_size" : "0","op_name" : "Concat","activation_size" : "524288","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "17412", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"34996": {"num_run": 0, "core": -1},"55368": {"num_run": 0, "core": -1},"28884": {"num_run": 0, "core": -1},"59012": {"num_run": 0, "core": -1},"23932": {"num_run": 0, "core": -1},"56800": {"num_run": 0, "core": -1},"36528": {"num_run": 0, "core": -1},"32820": {"num_run": 0, "core": -1},"34664": {"num_run": 0, "core": -1},"37516": {"num_run": 0, "core": -1},"20348": {"num_run": 0, "core": -1}}}}}, +{"cat" : "Session","pid" :46876,"tid" :17412,"dur" :10022,"ts" :15103572,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :46876,"tid" :17412,"dur" :10087,"ts" :15103536,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :46876,"tid" :17412,"dur" :62,"ts" :15115828,"ph" : "X","name" :"/Mul_1/SimplifiedLayerNormFusion/_kernel_time","args" : {"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "17412", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"34996": {"num_run": 0, "core": -1},"55368": {"num_run": 0, "core": -1},"28884": {"num_run": 0, "core": -1},"59012": {"num_run": 0, "core": -1},"23932": {"num_run": 0, "core": -1},"56800": {"num_run": 0, "core": -1},"36528": {"num_run": 0, "core": -1},"32820": {"num_run": 0, "core": -1},"34664": {"num_run": 0, "core": -1},"37516": {"num_run": 0, "core": -1},"20348": {"num_run": 0, "core": -1}}},"mem_requested_in_use_delta" : "0","input_type_shape" : [{"float":[1,1,2048]},{"float":[2048]}],"parameter_size" : "8192","output_size" : "8192","mem_arena_held_delta" : "0","mem_bytes_requested_in_use" : "1191966","mem_bytes_in_use" : "1572894","mem_in_use_delta" : "0","node_index" : "33","activation_size" : "8192","provider" : "CPUExecutionProvider","output_type_shape" : [{"float":[1,1,2048]}],"mem_in_use_peak" : "1572894","op_name" : "SimplifiedLayerNormalization","mem_arena_held" : "3145758"}}, +{"cat" : "Node","pid" :46876,"tid" :17412,"dur" :34,"ts" :15115901,"ph" : "X","name" :"/Mul_1_output_0_q_kernel_time","args" : {"activation_size" : "8192","mem_requested_in_use_delta" : "0","parameter_size" : "6","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "17412", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"34996": {"num_run": 0, "core": -1},"55368": {"num_run": 0, "core": -1},"28884": {"num_run": 0, "core": -1},"59012": {"num_run": 0, "core": -1},"23932": {"num_run": 0, "core": -1},"56800": {"num_run": 0, "core": -1},"36528": {"num_run": 0, "core": -1},"32820": {"num_run": 0, "core": -1},"34664": {"num_run": 0, "core": -1},"37516": {"num_run": 0, "core": -1},"20348": {"num_run": 0, "core": -1}}},"provider" : "CPUExecutionProvider","op_name" : "QuantizeLinear","output_type_shape" : [{"uint16":[1,1,2048]}],"node_index" : "30","mem_arena_held_delta" : "0","mem_in_use_delta" : "0","output_size" : "4096","mem_arena_held" : "3145758","mem_bytes_requested_in_use" : "1191966","mem_bytes_in_use" : "1572894","mem_in_use_peak" : "1572894","input_type_shape" : [{"float":[1,1,2048]},{"float":[]},{"uint16":[]}]}}, +{"cat" : "Node","pid" :46876,"tid" :17412,"dur" :8333,"ts" :15115944,"ph" : "X","name" :"vitisai_vitisai_2852173091035858231_0_0_kernel_time","args" : {"mem_bytes_requested_in_use" : "1191966","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "17412", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"34996": {"num_run": 0, "core": -1},"55368": {"num_run": 0, "core": -1},"28884": {"num_run": 0, "core": -1},"59012": {"num_run": 0, "core": -1},"23932": {"num_run": 0, "core": -1},"56800": {"num_run": 0, "core": -1},"36528": {"num_run": 0, "core": -1},"32820": {"num_run": 0, "core": -1},"34664": {"num_run": 0, "core": -1},"37516": {"num_run": 0, "core": -1},"20348": {"num_run": 0, "core": -1}}},"mem_in_use_peak" : "1572894","mem_arena_held_delta" : "0","op_name" : "vitisai_2852173091035858231_0","output_size" : "262144","output_type_shape" : [{"uint16":[1,32768]},{"uint16":[1,32768]},{"uint16":[1,32768]},{"uint16":[1,32768]}],"mem_requested_in_use_delta" : "0","input_type_shape" : [{"uint16":[1,1,2048]}],"mem_arena_held" : "3145758","provider" : "vitisai","mem_in_use_delta" : "0","node_index" : "32","mem_bytes_in_use" : "1572894","activation_size" : "4096","parameter_size" : "0"}}, +{"cat" : "Node","pid" :46876,"tid" :17412,"dur" :50,"ts" :15124306,"ph" : "X","name" :"logits_offset_98304_DequantizeLinear_kernel_time","args" : {"input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_arena_held" : "3145758","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "17412", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"34996": {"num_run": 0, "core": -1},"55368": {"num_run": 0, "core": -1},"28884": {"num_run": 0, "core": -1},"59012": {"num_run": 0, "core": -1},"23932": {"num_run": 0, "core": -1},"56800": {"num_run": 0, "core": -1},"36528": {"num_run": 0, "core": -1},"32820": {"num_run": 0, "core": -1},"34664": {"num_run": 0, "core": -1},"37516": {"num_run": 0, "core": -1},"20348": {"num_run": 0, "core": -1}}},"mem_in_use_peak" : "1572894","mem_in_use_delta" : "0","mem_bytes_requested_in_use" : "1191966","provider" : "CPUExecutionProvider","mem_arena_held_delta" : "0","op_name" : "DequantizeLinear","output_size" : "131072","node_index" : "25","output_type_shape" : [{"float":[1,32768]}],"mem_requested_in_use_delta" : "0","mem_bytes_in_use" : "1572894","activation_size" : "65536","parameter_size" : "6"}}, +{"cat" : "Node","pid" :46876,"tid" :17412,"dur" :37,"ts" :15124362,"ph" : "X","name" :"logits_offset_65536_DequantizeLinear_kernel_time","args" : {"mem_in_use_peak" : "1572894","provider" : "CPUExecutionProvider","node_index" : "24","mem_bytes_in_use" : "1572894","mem_requested_in_use_delta" : "0","parameter_size" : "6","output_type_shape" : [{"float":[1,32768]}],"input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "17412", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"34996": {"num_run": 0, "core": -1},"55368": {"num_run": 0, "core": -1},"28884": {"num_run": 0, "core": -1},"59012": {"num_run": 0, "core": -1},"23932": {"num_run": 0, "core": -1},"56800": {"num_run": 0, "core": -1},"36528": {"num_run": 0, "core": -1},"32820": {"num_run": 0, "core": -1},"34664": {"num_run": 0, "core": -1},"37516": {"num_run": 0, "core": -1},"20348": {"num_run": 0, "core": -1}}},"op_name" : "DequantizeLinear","activation_size" : "65536","mem_bytes_requested_in_use" : "1191966","mem_arena_held" : "3145758","mem_arena_held_delta" : "0","output_size" : "131072","mem_in_use_delta" : "0"}}, +{"cat" : "Node","pid" :46876,"tid" :17412,"dur" :35,"ts" :15124405,"ph" : "X","name" :"logits_offset_32768_DequantizeLinear_kernel_time","args" : {"node_index" : "23","parameter_size" : "6","mem_arena_held_delta" : "0","mem_bytes_in_use" : "1572894","op_name" : "DequantizeLinear","output_size" : "131072","mem_requested_in_use_delta" : "0","mem_in_use_delta" : "0","mem_arena_held" : "3145758","provider" : "CPUExecutionProvider","activation_size" : "65536","mem_in_use_peak" : "1572894","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "17412", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"34996": {"num_run": 0, "core": -1},"55368": {"num_run": 0, "core": -1},"28884": {"num_run": 0, "core": -1},"59012": {"num_run": 0, "core": -1},"23932": {"num_run": 0, "core": -1},"56800": {"num_run": 0, "core": -1},"36528": {"num_run": 0, "core": -1},"32820": {"num_run": 0, "core": -1},"34664": {"num_run": 0, "core": -1},"37516": {"num_run": 0, "core": -1},"20348": {"num_run": 0, "core": -1}}},"input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_bytes_requested_in_use" : "1191966","output_type_shape" : [{"float":[1,32768]}]}}, +{"cat" : "Node","pid" :46876,"tid" :17412,"dur" :36,"ts" :15124447,"ph" : "X","name" :"logits_offset_0_DequantizeLinear_kernel_time","args" : {"mem_bytes_requested_in_use" : "1191966","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"activation_size" : "65536","mem_arena_held" : "3145758","provider" : "CPUExecutionProvider","node_index" : "22","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "17412", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"34996": {"num_run": 0, "core": -1},"55368": {"num_run": 0, "core": -1},"28884": {"num_run": 0, "core": -1},"59012": {"num_run": 0, "core": -1},"23932": {"num_run": 0, "core": -1},"56800": {"num_run": 0, "core": -1},"36528": {"num_run": 0, "core": -1},"32820": {"num_run": 0, "core": -1},"34664": {"num_run": 0, "core": -1},"37516": {"num_run": 0, "core": -1},"20348": {"num_run": 0, "core": -1}}},"mem_arena_held_delta" : "0","mem_requested_in_use_delta" : "0","op_name" : "DequantizeLinear","mem_in_use_delta" : "0","parameter_size" : "6","output_type_shape" : [{"float":[1,32768]}],"mem_in_use_peak" : "1572894","mem_bytes_in_use" : "1572894","output_size" : "131072"}}, +{"cat" : "Node","pid" :46876,"tid" :17412,"dur" :151,"ts" :15124490,"ph" : "X","name" :"spark_lm_head_concat_kernel_time","args" : {"provider" : "CPUExecutionProvider","mem_arena_held" : "3145758","output_size" : "524288","parameter_size" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "17412", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"34996": {"num_run": 0, "core": -1},"55368": {"num_run": 0, "core": -1},"28884": {"num_run": 0, "core": -1},"59012": {"num_run": 0, "core": -1},"23932": {"num_run": 0, "core": -1},"56800": {"num_run": 0, "core": -1},"36528": {"num_run": 0, "core": -1},"32820": {"num_run": 0, "core": -1},"34664": {"num_run": 0, "core": -1},"37516": {"num_run": 0, "core": -1},"20348": {"num_run": 0, "core": -1}}},"input_type_shape" : [{"float":[1,32768]},{"float":[1,32768]},{"float":[1,32768]},{"float":[1,32768]}],"mem_in_use_delta" : "524288","mem_requested_in_use_delta" : "524288","output_type_shape" : [{"float":[1,131072]}],"mem_bytes_requested_in_use" : "1716254","mem_bytes_in_use" : "2097182","node_index" : "26","op_name" : "Concat","mem_in_use_peak" : "2097182","activation_size" : "524288","mem_arena_held_delta" : "0"}}, +{"cat" : "Session","pid" :46876,"tid" :17412,"dur" :8829,"ts" :15115824,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :46876,"tid" :17412,"dur" :8893,"ts" :15115790,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :46876,"tid" :17412,"dur" :67,"ts" :15126805,"ph" : "X","name" :"/Mul_1/SimplifiedLayerNormFusion/_kernel_time","args" : {"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "17412", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"34996": {"num_run": 0, "core": -1},"55368": {"num_run": 0, "core": -1},"28884": {"num_run": 0, "core": -1},"59012": {"num_run": 0, "core": -1},"23932": {"num_run": 0, "core": -1},"56800": {"num_run": 0, "core": -1},"36528": {"num_run": 0, "core": -1},"32820": {"num_run": 0, "core": -1},"34664": {"num_run": 0, "core": -1},"37516": {"num_run": 0, "core": -1},"20348": {"num_run": 0, "core": -1}}},"mem_arena_held" : "3145758","node_index" : "33","op_name" : "SimplifiedLayerNormalization","mem_in_use_delta" : "0","mem_bytes_requested_in_use" : "1191966","output_size" : "8192","mem_in_use_peak" : "2097182","input_type_shape" : [{"float":[1,1,2048]},{"float":[2048]}],"mem_requested_in_use_delta" : "0","provider" : "CPUExecutionProvider","parameter_size" : "8192","mem_bytes_in_use" : "1572894","mem_arena_held_delta" : "0","output_type_shape" : [{"float":[1,1,2048]}],"activation_size" : "8192"}}, +{"cat" : "Node","pid" :46876,"tid" :17412,"dur" :28,"ts" :15126882,"ph" : "X","name" :"/Mul_1_output_0_q_kernel_time","args" : {"op_name" : "QuantizeLinear","mem_in_use_peak" : "2097182","parameter_size" : "6","output_type_shape" : [{"uint16":[1,1,2048]}],"mem_requested_in_use_delta" : "0","output_size" : "4096","mem_in_use_delta" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "17412", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"34996": {"num_run": 0, "core": -1},"55368": {"num_run": 0, "core": -1},"28884": {"num_run": 0, "core": -1},"59012": {"num_run": 0, "core": -1},"23932": {"num_run": 0, "core": -1},"56800": {"num_run": 0, "core": -1},"36528": {"num_run": 0, "core": -1},"32820": {"num_run": 0, "core": -1},"34664": {"num_run": 0, "core": -1},"37516": {"num_run": 0, "core": -1},"20348": {"num_run": 0, "core": -1}}},"activation_size" : "8192","node_index" : "30","mem_bytes_in_use" : "1572894","mem_arena_held_delta" : "0","provider" : "CPUExecutionProvider","mem_bytes_requested_in_use" : "1191966","mem_arena_held" : "3145758","input_type_shape" : [{"float":[1,1,2048]},{"float":[]},{"uint16":[]}]}}, +{"cat" : "Node","pid" :46876,"tid" :17412,"dur" :7591,"ts" :15126916,"ph" : "X","name" :"vitisai_vitisai_2852173091035858231_0_0_kernel_time","args" : {"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "17412", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"34996": {"num_run": 0, "core": -1},"55368": {"num_run": 0, "core": -1},"28884": {"num_run": 0, "core": -1},"59012": {"num_run": 0, "core": -1},"23932": {"num_run": 0, "core": -1},"56800": {"num_run": 0, "core": -1},"36528": {"num_run": 0, "core": -1},"32820": {"num_run": 0, "core": -1},"34664": {"num_run": 0, "core": -1},"37516": {"num_run": 0, "core": -1},"20348": {"num_run": 0, "core": -1}}},"mem_requested_in_use_delta" : "0","op_name" : "vitisai_2852173091035858231_0","mem_in_use_peak" : "2097182","provider" : "vitisai","mem_arena_held" : "3145758","mem_in_use_delta" : "0","activation_size" : "4096","mem_arena_held_delta" : "0","parameter_size" : "0","mem_bytes_in_use" : "1572894","mem_bytes_requested_in_use" : "1191966","output_type_shape" : [{"uint16":[1,32768]},{"uint16":[1,32768]},{"uint16":[1,32768]},{"uint16":[1,32768]}],"node_index" : "32","output_size" : "262144","input_type_shape" : [{"uint16":[1,1,2048]}]}}, +{"cat" : "Node","pid" :46876,"tid" :17412,"dur" :41,"ts" :15134519,"ph" : "X","name" :"logits_offset_98304_DequantizeLinear_kernel_time","args" : {"input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"activation_size" : "65536","output_type_shape" : [{"float":[1,32768]}],"mem_in_use_peak" : "2097182","mem_bytes_requested_in_use" : "1191966","mem_bytes_in_use" : "1572894","op_name" : "DequantizeLinear","node_index" : "25","mem_arena_held_delta" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "17412", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"34996": {"num_run": 0, "core": -1},"55368": {"num_run": 0, "core": -1},"28884": {"num_run": 0, "core": -1},"59012": {"num_run": 0, "core": -1},"23932": {"num_run": 0, "core": -1},"56800": {"num_run": 0, "core": -1},"36528": {"num_run": 0, "core": -1},"32820": {"num_run": 0, "core": -1},"34664": {"num_run": 0, "core": -1},"37516": {"num_run": 0, "core": -1},"20348": {"num_run": 0, "core": -1}}},"mem_arena_held" : "3145758","mem_requested_in_use_delta" : "0","parameter_size" : "6","provider" : "CPUExecutionProvider","mem_in_use_delta" : "0","output_size" : "131072"}}, +{"cat" : "Node","pid" :46876,"tid" :17412,"dur" :34,"ts" :15134565,"ph" : "X","name" :"logits_offset_65536_DequantizeLinear_kernel_time","args" : {"output_size" : "131072","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"node_index" : "24","activation_size" : "65536","mem_in_use_peak" : "2097182","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "17412", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"34996": {"num_run": 0, "core": -1},"55368": {"num_run": 0, "core": -1},"28884": {"num_run": 0, "core": -1},"59012": {"num_run": 0, "core": -1},"23932": {"num_run": 0, "core": -1},"56800": {"num_run": 0, "core": -1},"36528": {"num_run": 0, "core": -1},"32820": {"num_run": 0, "core": -1},"34664": {"num_run": 0, "core": -1},"37516": {"num_run": 0, "core": -1},"20348": {"num_run": 0, "core": -1}}},"mem_in_use_delta" : "0","mem_arena_held_delta" : "0","op_name" : "DequantizeLinear","provider" : "CPUExecutionProvider","parameter_size" : "6","mem_bytes_in_use" : "1572894","mem_requested_in_use_delta" : "0","mem_bytes_requested_in_use" : "1191966","output_type_shape" : [{"float":[1,32768]}],"mem_arena_held" : "3145758"}}, +{"cat" : "Node","pid" :46876,"tid" :17412,"dur" :31,"ts" :15134604,"ph" : "X","name" :"logits_offset_32768_DequantizeLinear_kernel_time","args" : {"op_name" : "DequantizeLinear","parameter_size" : "6","mem_bytes_requested_in_use" : "1191966","mem_arena_held_delta" : "0","activation_size" : "65536","mem_requested_in_use_delta" : "0","output_type_shape" : [{"float":[1,32768]}],"provider" : "CPUExecutionProvider","mem_arena_held" : "3145758","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "17412", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"34996": {"num_run": 0, "core": -1},"55368": {"num_run": 0, "core": -1},"28884": {"num_run": 0, "core": -1},"59012": {"num_run": 0, "core": -1},"23932": {"num_run": 0, "core": -1},"56800": {"num_run": 0, "core": -1},"36528": {"num_run": 0, "core": -1},"32820": {"num_run": 0, "core": -1},"34664": {"num_run": 0, "core": -1},"37516": {"num_run": 0, "core": -1},"20348": {"num_run": 0, "core": -1}}},"output_size" : "131072","mem_bytes_in_use" : "1572894","mem_in_use_delta" : "0","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_in_use_peak" : "2097182","node_index" : "23"}}, +{"cat" : "Node","pid" :46876,"tid" :17412,"dur" :32,"ts" :15134640,"ph" : "X","name" :"logits_offset_0_DequantizeLinear_kernel_time","args" : {"input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"output_size" : "131072","node_index" : "22","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "17412", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"34996": {"num_run": 0, "core": -1},"55368": {"num_run": 0, "core": -1},"28884": {"num_run": 0, "core": -1},"59012": {"num_run": 0, "core": -1},"23932": {"num_run": 0, "core": -1},"56800": {"num_run": 0, "core": -1},"36528": {"num_run": 0, "core": -1},"32820": {"num_run": 0, "core": -1},"34664": {"num_run": 0, "core": -1},"37516": {"num_run": 0, "core": -1},"20348": {"num_run": 0, "core": -1}}},"mem_bytes_in_use" : "1572894","mem_arena_held_delta" : "0","parameter_size" : "6","activation_size" : "65536","provider" : "CPUExecutionProvider","mem_arena_held" : "3145758","output_type_shape" : [{"float":[1,32768]}],"mem_bytes_requested_in_use" : "1191966","mem_requested_in_use_delta" : "0","mem_in_use_delta" : "0","op_name" : "DequantizeLinear","mem_in_use_peak" : "2097182"}}, +{"cat" : "Node","pid" :46876,"tid" :17412,"dur" :59,"ts" :15134682,"ph" : "X","name" :"spark_lm_head_concat_kernel_time","args" : {"node_index" : "26","mem_in_use_peak" : "2097182","mem_in_use_delta" : "524288","mem_arena_held" : "3145758","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "17412", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"34996": {"num_run": 0, "core": -1},"55368": {"num_run": 0, "core": -1},"28884": {"num_run": 0, "core": -1},"59012": {"num_run": 0, "core": -1},"23932": {"num_run": 0, "core": -1},"56800": {"num_run": 0, "core": -1},"36528": {"num_run": 0, "core": -1},"32820": {"num_run": 0, "core": -1},"34664": {"num_run": 0, "core": -1},"37516": {"num_run": 0, "core": -1},"20348": {"num_run": 0, "core": -1}}},"input_type_shape" : [{"float":[1,32768]},{"float":[1,32768]},{"float":[1,32768]},{"float":[1,32768]}],"activation_size" : "524288","mem_requested_in_use_delta" : "524288","output_size" : "524288","provider" : "CPUExecutionProvider","output_type_shape" : [{"float":[1,131072]}],"mem_arena_held_delta" : "0","mem_bytes_in_use" : "2097182","op_name" : "Concat","parameter_size" : "0","mem_bytes_requested_in_use" : "1716254"}}, +{"cat" : "Session","pid" :46876,"tid" :17412,"dur" :7946,"ts" :15126800,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :46876,"tid" :17412,"dur" :8006,"ts" :15126762,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :46876,"tid" :17412,"dur" :72,"ts" :15136921,"ph" : "X","name" :"/Mul_1/SimplifiedLayerNormFusion/_kernel_time","args" : {"mem_in_use_peak" : "2097182","parameter_size" : "8192","output_size" : "8192","mem_arena_held_delta" : "0","mem_in_use_delta" : "0","input_type_shape" : [{"float":[1,1,2048]},{"float":[2048]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "17412", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"34996": {"num_run": 0, "core": -1},"55368": {"num_run": 0, "core": -1},"28884": {"num_run": 0, "core": -1},"59012": {"num_run": 0, "core": -1},"23932": {"num_run": 0, "core": -1},"56800": {"num_run": 0, "core": -1},"36528": {"num_run": 0, "core": -1},"32820": {"num_run": 0, "core": -1},"34664": {"num_run": 0, "core": -1},"37516": {"num_run": 0, "core": -1},"20348": {"num_run": 0, "core": -1}}},"mem_arena_held" : "3145758","op_name" : "SimplifiedLayerNormalization","mem_requested_in_use_delta" : "0","output_type_shape" : [{"float":[1,1,2048]}],"node_index" : "33","activation_size" : "8192","mem_bytes_requested_in_use" : "1191966","provider" : "CPUExecutionProvider","mem_bytes_in_use" : "1572894"}}, +{"cat" : "Node","pid" :46876,"tid" :17412,"dur" :34,"ts" :15137004,"ph" : "X","name" :"/Mul_1_output_0_q_kernel_time","args" : {"activation_size" : "8192","op_name" : "QuantizeLinear","output_type_shape" : [{"uint16":[1,1,2048]}],"output_size" : "4096","parameter_size" : "6","mem_in_use_delta" : "0","mem_bytes_in_use" : "1572894","mem_arena_held_delta" : "0","mem_bytes_requested_in_use" : "1191966","mem_in_use_peak" : "2097182","node_index" : "30","provider" : "CPUExecutionProvider","mem_requested_in_use_delta" : "0","input_type_shape" : [{"float":[1,1,2048]},{"float":[]},{"uint16":[]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "17412", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"34996": {"num_run": 0, "core": -1},"55368": {"num_run": 0, "core": -1},"28884": {"num_run": 0, "core": -1},"59012": {"num_run": 0, "core": -1},"23932": {"num_run": 0, "core": -1},"56800": {"num_run": 0, "core": -1},"36528": {"num_run": 0, "core": -1},"32820": {"num_run": 0, "core": -1},"34664": {"num_run": 0, "core": -1},"37516": {"num_run": 0, "core": -1},"20348": {"num_run": 0, "core": -1}}},"mem_arena_held" : "3145758"}}, +{"cat" : "Node","pid" :46876,"tid" :17412,"dur" :8373,"ts" :15137044,"ph" : "X","name" :"vitisai_vitisai_2852173091035858231_0_0_kernel_time","args" : {"node_index" : "32","mem_requested_in_use_delta" : "0","output_size" : "262144","mem_in_use_peak" : "2097182","mem_bytes_in_use" : "1572894","activation_size" : "4096","op_name" : "vitisai_2852173091035858231_0","input_type_shape" : [{"uint16":[1,1,2048]}],"mem_arena_held_delta" : "0","mem_arena_held" : "3145758","mem_in_use_delta" : "0","parameter_size" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "17412", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"34996": {"num_run": 0, "core": -1},"55368": {"num_run": 0, "core": -1},"28884": {"num_run": 0, "core": -1},"59012": {"num_run": 0, "core": -1},"23932": {"num_run": 0, "core": -1},"56800": {"num_run": 0, "core": -1},"36528": {"num_run": 0, "core": -1},"32820": {"num_run": 0, "core": -1},"34664": {"num_run": 0, "core": -1},"37516": {"num_run": 0, "core": -1},"20348": {"num_run": 0, "core": -1}}},"provider" : "vitisai","mem_bytes_requested_in_use" : "1191966","output_type_shape" : [{"uint16":[1,32768]},{"uint16":[1,32768]},{"uint16":[1,32768]},{"uint16":[1,32768]}]}}, +{"cat" : "Node","pid" :46876,"tid" :17412,"dur" :42,"ts" :15145434,"ph" : "X","name" :"logits_offset_98304_DequantizeLinear_kernel_time","args" : {"mem_bytes_in_use" : "1572894","provider" : "CPUExecutionProvider","output_size" : "131072","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"op_name" : "DequantizeLinear","mem_bytes_requested_in_use" : "1191966","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "17412", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"34996": {"num_run": 0, "core": -1},"55368": {"num_run": 0, "core": -1},"28884": {"num_run": 0, "core": -1},"59012": {"num_run": 0, "core": -1},"23932": {"num_run": 0, "core": -1},"56800": {"num_run": 0, "core": -1},"36528": {"num_run": 0, "core": -1},"32820": {"num_run": 0, "core": -1},"34664": {"num_run": 0, "core": -1},"37516": {"num_run": 0, "core": -1},"20348": {"num_run": 0, "core": -1}}},"output_type_shape" : [{"float":[1,32768]}],"mem_arena_held" : "3145758","activation_size" : "65536","mem_in_use_peak" : "2097182","mem_in_use_delta" : "0","mem_arena_held_delta" : "0","mem_requested_in_use_delta" : "0","parameter_size" : "6","node_index" : "25"}}, +{"cat" : "Node","pid" :46876,"tid" :17412,"dur" :35,"ts" :15145483,"ph" : "X","name" :"logits_offset_65536_DequantizeLinear_kernel_time","args" : {"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "17412", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"34996": {"num_run": 0, "core": -1},"55368": {"num_run": 0, "core": -1},"28884": {"num_run": 0, "core": -1},"59012": {"num_run": 0, "core": -1},"23932": {"num_run": 0, "core": -1},"56800": {"num_run": 0, "core": -1},"36528": {"num_run": 0, "core": -1},"32820": {"num_run": 0, "core": -1},"34664": {"num_run": 0, "core": -1},"37516": {"num_run": 0, "core": -1},"20348": {"num_run": 0, "core": -1}}},"mem_requested_in_use_delta" : "0","mem_arena_held" : "3145758","output_size" : "131072","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"provider" : "CPUExecutionProvider","mem_bytes_requested_in_use" : "1191966","node_index" : "24","mem_in_use_delta" : "0","activation_size" : "65536","mem_in_use_peak" : "2097182","op_name" : "DequantizeLinear","mem_bytes_in_use" : "1572894","output_type_shape" : [{"float":[1,32768]}],"parameter_size" : "6","mem_arena_held_delta" : "0"}}, +{"cat" : "Node","pid" :46876,"tid" :17412,"dur" :32,"ts" :15145524,"ph" : "X","name" :"logits_offset_32768_DequantizeLinear_kernel_time","args" : {"mem_in_use_peak" : "2097182","mem_arena_held" : "3145758","parameter_size" : "6","output_size" : "131072","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "17412", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"34996": {"num_run": 0, "core": -1},"55368": {"num_run": 0, "core": -1},"28884": {"num_run": 0, "core": -1},"59012": {"num_run": 0, "core": -1},"23932": {"num_run": 0, "core": -1},"56800": {"num_run": 0, "core": -1},"36528": {"num_run": 0, "core": -1},"32820": {"num_run": 0, "core": -1},"34664": {"num_run": 0, "core": -1},"37516": {"num_run": 0, "core": -1},"20348": {"num_run": 0, "core": -1}}},"activation_size" : "65536","mem_arena_held_delta" : "0","mem_bytes_in_use" : "1572894","mem_requested_in_use_delta" : "0","provider" : "CPUExecutionProvider","mem_bytes_requested_in_use" : "1191966","node_index" : "23","op_name" : "DequantizeLinear","output_type_shape" : [{"float":[1,32768]}],"mem_in_use_delta" : "0","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}]}}, +{"cat" : "Node","pid" :46876,"tid" :17412,"dur" :32,"ts" :15145562,"ph" : "X","name" :"logits_offset_0_DequantizeLinear_kernel_time","args" : {"input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_requested_in_use_delta" : "0","activation_size" : "65536","mem_in_use_delta" : "0","node_index" : "22","mem_in_use_peak" : "2097182","mem_bytes_in_use" : "1572894","op_name" : "DequantizeLinear","mem_arena_held_delta" : "0","provider" : "CPUExecutionProvider","mem_bytes_requested_in_use" : "1191966","output_size" : "131072","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "17412", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"34996": {"num_run": 0, "core": -1},"55368": {"num_run": 0, "core": -1},"28884": {"num_run": 0, "core": -1},"59012": {"num_run": 0, "core": -1},"23932": {"num_run": 0, "core": -1},"56800": {"num_run": 0, "core": -1},"36528": {"num_run": 0, "core": -1},"32820": {"num_run": 0, "core": -1},"34664": {"num_run": 0, "core": -1},"37516": {"num_run": 0, "core": -1},"20348": {"num_run": 0, "core": -1}}},"parameter_size" : "6","mem_arena_held" : "3145758","output_type_shape" : [{"float":[1,32768]}]}}, +{"cat" : "Node","pid" :46876,"tid" :17412,"dur" :63,"ts" :15145599,"ph" : "X","name" :"spark_lm_head_concat_kernel_time","args" : {"mem_requested_in_use_delta" : "524288","activation_size" : "524288","mem_bytes_in_use" : "2097182","op_name" : "Concat","mem_in_use_delta" : "524288","input_type_shape" : [{"float":[1,32768]},{"float":[1,32768]},{"float":[1,32768]},{"float":[1,32768]}],"mem_in_use_peak" : "2097182","mem_arena_held" : "3145758","output_size" : "524288","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "17412", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"34996": {"num_run": 0, "core": -1},"55368": {"num_run": 0, "core": -1},"28884": {"num_run": 0, "core": -1},"59012": {"num_run": 0, "core": -1},"23932": {"num_run": 0, "core": -1},"56800": {"num_run": 0, "core": -1},"36528": {"num_run": 0, "core": -1},"32820": {"num_run": 0, "core": -1},"34664": {"num_run": 0, "core": -1},"37516": {"num_run": 0, "core": -1},"20348": {"num_run": 0, "core": -1}}},"mem_arena_held_delta" : "0","output_type_shape" : [{"float":[1,131072]}],"mem_bytes_requested_in_use" : "1716254","parameter_size" : "0","node_index" : "26","provider" : "CPUExecutionProvider"}}, +{"cat" : "Session","pid" :46876,"tid" :17412,"dur" :8753,"ts" :15136916,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :46876,"tid" :17412,"dur" :8813,"ts" :15136879,"ph" : "X","name" :"model_run","args" : {}} +] diff --git a/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/composite_raw.log b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/composite_raw.log new file mode 100644 index 000000000..212b06b29 Binary files /dev/null and b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/composite_raw.log differ diff --git a/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/composite_report.json b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/composite_report.json new file mode 100644 index 000000000..5f5c6174b --- /dev/null +++ b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/composite_report.json @@ -0,0 +1,98 @@ +{ + "scope": "Spark-X2.5-1.7B output head only; no prefill/decode/KV/sampling or full generation", + "vocab_shard_size": 131072, + "vocab_shard_offset": 0, + "flatten_input": false, + "batch_repeats": 1, + "matmul_only": false, + "composite_shards": false, + "composite_with_norm": true, + "hardware": "AMD Ryzen AI 9 HX 370 NPU", + "os": "Windows-11-10.0.26200-SP0", + "onnx_version": "1.22.0", + "onnxruntime_version": "1.30.0", + "source_model_sha256": "f5f0c31eb438e9940ab242da8bf6f0f5dad6ebc9f6ef02d5475c21b744e595bc", + "prepared_model_sha256": "1a426841ebb0e1255268b7c796bfcbe44072cf65613e47cfdf0c9d807746518b", + "calibration_sha256": "cfa2ed8439f8fe01730319667bfd01abf2e39cf5851490346a0331304cf63194", + "opset21_sha256": "1a426841ebb0e1255268b7c796bfcbe44072cf65613e47cfdf0c9d807746518b", + "quantized_sha256": "e2ae9c3fd7071e8ff628f209e6140c8a023bfaad59814b80a936c6b263f33eaf", + "quantization": "opset 21, MatMul only, QDQ, QUInt16 activations/QInt8 weights, MinMax on four real activations", + "input_shape": [ + 1, + 1, + 2048 + ], + "output_shape": [ + 1, + 131072 + ], + "source_cpu_reference": [ + { + "snr_db": 137.37257867357746, + "max_abs_diff": 1.9073486328125e-05, + "top1_match": true, + "finite": true + }, + { + "snr_db": 134.38703441007158, + "max_abs_diff": 2.6702880859375e-05, + "top1_match": true, + "finite": true + }, + { + "snr_db": 136.39026519822127, + "max_abs_diff": 2.288818359375e-05, + "top1_match": true, + "finite": true + }, + { + "snr_db": 133.65524993196834, + "max_abs_diff": 2.288818359375e-05, + "top1_match": true, + "finite": true + } + ], + "npu_devices": 1, + "session_providers": [ + "vitisai", + "CPUExecutionProvider" + ], + "npu_node_events": 4, + "cpu_node_events": 28, + "load_s": 15.10484120005276, + "npu_rows": [ + { + "snr_db": 56.81097231641212, + "max_abs_diff": 0.3098602294921875, + "top1_match": true, + "finite": true, + "index": 0, + "wall_s": 0.012230199994519353 + }, + { + "snr_db": 56.37996818626326, + "max_abs_diff": 0.27384185791015625, + "top1_match": true, + "finite": true, + "index": 1, + "wall_s": 0.010965699912048876 + }, + { + "snr_db": 56.32534194832648, + "max_abs_diff": 0.25970458984375, + "top1_match": true, + "finite": true, + "index": 2, + "wall_s": 0.010113499942235649 + }, + { + "snr_db": 55.42871576712098, + "max_abs_diff": 0.2625885009765625, + "top1_match": true, + "finite": true, + "index": 3, + "wall_s": 0.010933700017631054 + } + ], + "profile_file": "C:\\Users\\zhout\\w2\\spark_npu_lmhead_composite_with_norm_20260923\\spark_lm_head_npu_profile_2026-09-23_18-19-03_343.json" +} diff --git a/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/failures/full_matmul_raw.log b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/failures/full_matmul_raw.log new file mode 100644 index 000000000..c7bc04377 Binary files /dev/null and b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/failures/full_matmul_raw.log differ diff --git a/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/failures/full_matmul_report.json b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/failures/full_matmul_report.json new file mode 100644 index 000000000..2eb6d305b --- /dev/null +++ b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/failures/full_matmul_report.json @@ -0,0 +1,94 @@ +{ + "scope": "Spark-X2.5-1.7B output head only; no prefill/decode/KV/sampling or full generation", + "vocab_shard_size": 131072, + "flatten_input": true, + "batch_repeats": 1, + "matmul_only": true, + "hardware": "AMD Ryzen AI 9 HX 370 NPU", + "os": "Windows-11-10.0.26200-SP0", + "onnx_version": "1.22.0", + "onnxruntime_version": "1.30.0", + "source_model_sha256": "f5f0c31eb438e9940ab242da8bf6f0f5dad6ebc9f6ef02d5475c21b744e595bc", + "prepared_model_sha256": "437a694d549ba56b8d78930beede9e3aeb835a51833bc3a9e532b0c4f6063147", + "calibration_sha256": "cfa2ed8439f8fe01730319667bfd01abf2e39cf5851490346a0331304cf63194", + "opset21_sha256": "437a694d549ba56b8d78930beede9e3aeb835a51833bc3a9e532b0c4f6063147", + "quantized_sha256": "66208cbe2a15b70edb5d16856f4a83c39610758f6eefe414e9339bdef2715b9a", + "quantization": "opset 21, MatMul only, QDQ, QUInt16 activations/QInt8 weights, MinMax on four real activations", + "input_shape": [ + 1, + 2048 + ], + "output_shape": [ + 1, + 131072 + ], + "source_cpu_reference": [ + { + "snr_db": 137.37257867357746, + "max_abs_diff": 1.9073486328125e-05, + "top1_match": true, + "finite": true + }, + { + "snr_db": 134.38703441007158, + "max_abs_diff": 2.6702880859375e-05, + "top1_match": true, + "finite": true + }, + { + "snr_db": 137.81138223525107, + "max_abs_diff": 2.288818359375e-05, + "top1_match": true, + "finite": true + }, + { + "snr_db": 136.4924381506639, + "max_abs_diff": 2.288818359375e-05, + "top1_match": true, + "finite": true + } + ], + "npu_devices": 1, + "session_providers": [ + "vitisai", + "CPUExecutionProvider" + ], + "npu_node_events": 0, + "cpu_node_events": 24, + "load_s": 6.2168904999271035, + "npu_rows": [ + { + "snr_db": 55.45826302981912, + "max_abs_diff": 0.3096885681152344, + "top1_match": true, + "finite": true, + "index": 0, + "wall_s": 0.17048179998528212 + }, + { + "snr_db": 55.028045929115706, + "max_abs_diff": 0.2736663818359375, + "top1_match": true, + "finite": true, + "index": 1, + "wall_s": 0.17349539999850094 + }, + { + "snr_db": 54.99008301968252, + "max_abs_diff": 0.2613334655761719, + "top1_match": true, + "finite": true, + "index": 2, + "wall_s": 0.06820740003604442 + }, + { + "snr_db": 54.074552903906444, + "max_abs_diff": 0.26749420166015625, + "top1_match": true, + "finite": true, + "index": 3, + "wall_s": 0.06938180001452565 + } + ], + "profile_file": "C:\\Users\\zhout\\w2\\spark_npu_lmhead_matmul_full_batch1_20260923\\spark_lm_head_npu_profile_2026-09-23_17-56-44_321.json" +} diff --git a/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/failures/full_qdq_raw.log b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/failures/full_qdq_raw.log new file mode 100644 index 000000000..c231d7a3e Binary files /dev/null and b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/failures/full_qdq_raw.log differ diff --git a/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/failures/full_qdq_report.json b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/failures/full_qdq_report.json new file mode 100644 index 000000000..125a088d5 --- /dev/null +++ b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/failures/full_qdq_report.json @@ -0,0 +1,90 @@ +{ + "scope": "Spark-X2.5-1.7B output head only; no prefill/decode/KV/sampling or full generation", + "hardware": "AMD Ryzen AI 9 HX 370 NPU", + "os": "Windows-11-10.0.26200-SP0", + "onnx_version": "1.22.0", + "onnxruntime_version": "1.30.0", + "source_model_sha256": "f5f0c31eb438e9940ab242da8bf6f0f5dad6ebc9f6ef02d5475c21b744e595bc", + "calibration_sha256": "cfa2ed8439f8fe01730319667bfd01abf2e39cf5851490346a0331304cf63194", + "opset21_sha256": "738e06c7f5b0354bac9cd6251af6279fc5224a754695d182604e38f874563fbe", + "quantized_sha256": "0e9f2b151451b6fdd473032881dda5b6c6ab2a13908c533efea2689d0035bbca", + "quantization": "opset 21, MatMul only, QDQ, QUInt16 activations/QInt8 weights, MinMax on four real activations", + "input_shape": [ + 1, + 1, + 2048 + ], + "output_shape": [ + 1, + 131072 + ], + "source_cpu_reference": [ + { + "snr_db": 137.37257867357746, + "max_abs_diff": 1.9073486328125e-05, + "top1_match": true, + "finite": true + }, + { + "snr_db": 134.38703441007158, + "max_abs_diff": 2.6702880859375e-05, + "top1_match": true, + "finite": true + }, + { + "snr_db": 136.39026519822127, + "max_abs_diff": 2.288818359375e-05, + "top1_match": true, + "finite": true + }, + { + "snr_db": 133.65524993196834, + "max_abs_diff": 2.288818359375e-05, + "top1_match": true, + "finite": true + } + ], + "npu_devices": 1, + "session_providers": [ + "vitisai", + "CPUExecutionProvider" + ], + "npu_node_events": 0, + "cpu_node_events": 32, + "load_s": 7.350405399920419, + "npu_rows": [ + { + "snr_db": 55.45826302981912, + "max_abs_diff": 0.3096885681152344, + "top1_match": true, + "finite": true, + "index": 0, + "wall_s": 0.16518759995233268 + }, + { + "snr_db": 55.028045929115706, + "max_abs_diff": 0.2736663818359375, + "top1_match": true, + "finite": true, + "index": 1, + "wall_s": 0.1734156000893563 + }, + { + "snr_db": 54.990218080171104, + "max_abs_diff": 0.2613334655761719, + "top1_match": true, + "finite": true, + "index": 2, + "wall_s": 0.0690851000836119 + }, + { + "snr_db": 54.07466164473506, + "max_abs_diff": 0.26749420166015625, + "top1_match": true, + "finite": true, + "index": 3, + "wall_s": 0.06944940006360412 + } + ], + "profile_file": "C:\\Users\\zhout\\w2\\spark_npu_lmhead_20260923\\spark_lm_head_npu_profile_2026-09-23_17-49-20_967.json" +} diff --git a/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/failures/unnamed_co_resident_raw.log b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/failures/unnamed_co_resident_raw.log new file mode 100644 index 000000000..658c01369 Binary files /dev/null and b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/failures/unnamed_co_resident_raw.log differ diff --git a/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/failures/unnamed_co_resident_report.json b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/failures/unnamed_co_resident_report.json new file mode 100644 index 000000000..aa36785f0 --- /dev/null +++ b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/failures/unnamed_co_resident_report.json @@ -0,0 +1,322 @@ +{ + "scope": "Spark-X2.5-1.7B complete 131072-logit output head only; CPU RMS normalization plus four NPU MatMul shards, no full generation", + "hardware": "AMD Ryzen AI 9 HX 370 NPU", + "os": "Windows-11-10.0.26200-SP0", + "onnx_version": "1.22.0", + "onnxruntime_version": "1.30.0", + "source_model_sha256": "f5f0c31eb438e9940ab242da8bf6f0f5dad6ebc9f6ef02d5475c21b744e595bc", + "activation_sha256": "cfa2ed8439f8fe01730319667bfd01abf2e39cf5851490346a0331304cf63194", + "source_shape": [ + 1, + 1, + 2048 + ], + "output_shape": [ + 1, + 131072 + ], + "cpu_load_s": 6.092688599950634, + "shard_load_s": [ + 5.845066000008956, + 5.538822600035928, + 5.587077300064266, + 4.569033499923535 + ], + "shard_artifacts": [ + { + "offset": 0, + "quantized_sha256": "a94f8087f99d29487186f43110250e0532ef0c87a48f52926f15e70ae5fbbbb7", + "providers": [ + "vitisai", + "CPUExecutionProvider" + ] + }, + { + "offset": 32768, + "quantized_sha256": "7e9477e01e621281a62b6db4df50b6056743296ea62dee64f719365f18036049", + "providers": [ + "vitisai", + "CPUExecutionProvider" + ] + }, + { + "offset": 65536, + "quantized_sha256": "3be7e9f4da57931de3bc09931871600330740fea4af3a28fcec727f95ffebb7d", + "providers": [ + "vitisai", + "CPUExecutionProvider" + ] + }, + { + "offset": 98304, + "quantized_sha256": "896436f5798428f9c9cc85c44599cceda3699053c3d6b0e7ac31f1c3682dacef", + "providers": [ + "vitisai", + "CPUExecutionProvider" + ] + } + ], + "placement": [ + { + "offset": 0, + "npu_node_events": 25, + "cpu_node_events": 50, + "profile_file": "C:\\Users\\zhout\\w2\\spark_npu_full_head_e2e_component_20260923\\shard_0_profile_2026-09-23_18-05-33_448.json" + }, + { + "offset": 32768, + "npu_node_events": 25, + "cpu_node_events": 50, + "profile_file": "C:\\Users\\zhout\\w2\\spark_npu_full_head_e2e_component_20260923\\shard_32768_profile_2026-09-23_18-05-39_356.json" + }, + { + "offset": 65536, + "npu_node_events": 25, + "cpu_node_events": 50, + "profile_file": "C:\\Users\\zhout\\w2\\spark_npu_full_head_e2e_component_20260923\\shard_65536_profile_2026-09-23_18-05-44_970.json" + }, + { + "offset": 98304, + "npu_node_events": 25, + "cpu_node_events": 50, + "profile_file": "C:\\Users\\zhout\\w2\\spark_npu_full_head_e2e_component_20260923\\shard_98304_profile_2026-09-23_18-05-50_622.json" + } + ], + "cpu_reference_rows": [ + { + "snr_db": 137.37257867357746, + "max_abs_diff": 1.9073486328125e-05, + "top1_match": true, + "finite": true, + "index": 0, + "wall_s": 0.023925699992105365 + }, + { + "snr_db": 134.38703441007158, + "max_abs_diff": 2.6702880859375e-05, + "top1_match": true, + "finite": true, + "index": 1, + "wall_s": 0.025182000012136996 + }, + { + "snr_db": 136.39026519822127, + "max_abs_diff": 2.288818359375e-05, + "top1_match": true, + "finite": true, + "index": 2, + "wall_s": 0.024960999959148467 + }, + { + "snr_db": 133.65524993196834, + "max_abs_diff": 2.288818359375e-05, + "top1_match": true, + "finite": true, + "index": 3, + "wall_s": 0.023461699951440096 + } + ], + "npu_rows": [ + { + "snr_db": 19.19100665099978, + "max_abs_diff": 40.453853607177734, + "top1_match": true, + "finite": true, + "index": 0, + "wall_s": 0.014368699979968369, + "shard_checks": [ + { + "offset": 0, + "snr_db": 54.96589660909892, + "max_abs_diff": 0.3098640441894531, + "top1_match": true, + "finite": true + }, + { + "offset": 32768, + "snr_db": 18.058890779885274, + "max_abs_diff": 40.453853607177734, + "top1_match": false, + "finite": true + }, + { + "offset": 65536, + "snr_db": 17.782966398251837, + "max_abs_diff": 39.37982940673828, + "top1_match": false, + "finite": true + }, + { + "offset": 98304, + "snr_db": 18.47831780707655, + "max_abs_diff": 37.27281188964844, + "top1_match": false, + "finite": true + } + ] + }, + { + "snr_db": 18.1744718427091, + "max_abs_diff": 38.428254187107086, + "top1_match": true, + "finite": true, + "index": 1, + "wall_s": 0.011472699930891395, + "shard_checks": [ + { + "offset": 0, + "snr_db": 54.471154152973426, + "max_abs_diff": 0.2738380432128906, + "top1_match": true, + "finite": true + }, + { + "offset": 32768, + "snr_db": 16.985554247618726, + "max_abs_diff": 38.428254187107086, + "top1_match": false, + "finite": true + }, + { + "offset": 65536, + "snr_db": 16.924653199297357, + "max_abs_diff": 33.646121084690094, + "top1_match": false, + "finite": true + }, + { + "offset": 98304, + "snr_db": 17.403759257050268, + "max_abs_diff": 34.47327363491058, + "top1_match": false, + "finite": true + } + ] + }, + { + "snr_db": 17.889983746377528, + "max_abs_diff": 38.15506283193827, + "top1_match": true, + "finite": true, + "index": 2, + "wall_s": 0.011089300038293004, + "shard_checks": [ + { + "offset": 0, + "snr_db": 54.36750945439812, + "max_abs_diff": 0.2529621124267578, + "top1_match": true, + "finite": true + }, + { + "offset": 32768, + "snr_db": 16.857931348007348, + "max_abs_diff": 36.52046014368534, + "top1_match": false, + "finite": true + }, + { + "offset": 65536, + "snr_db": 16.696824516949302, + "max_abs_diff": 36.865197494626045, + "top1_match": false, + "finite": true + }, + { + "offset": 98304, + "snr_db": 16.934209667098045, + "max_abs_diff": 38.15506283193827, + "top1_match": false, + "finite": true + } + ] + }, + { + "snr_db": 17.57066425565234, + "max_abs_diff": 25.254364013671875, + "top1_match": false, + "finite": true, + "index": 3, + "wall_s": 0.014615799998864532, + "shard_checks": [ + { + "offset": 0, + "snr_db": 53.731174504422455, + "max_abs_diff": 0.2625865936279297, + "top1_match": true, + "finite": true + }, + { + "offset": 32768, + "snr_db": 16.52858880898565, + "max_abs_diff": 24.52711272239685, + "top1_match": false, + "finite": true + }, + { + "offset": 65536, + "snr_db": 15.952905570881054, + "max_abs_diff": 23.339406967163086, + "top1_match": false, + "finite": true + }, + { + "offset": 98304, + "snr_db": 16.878288211449945, + "max_abs_diff": 25.254364013671875, + "top1_match": false, + "finite": true + } + ] + } + ], + "timing_protocol": "one warmup then 20 serial repeats of captured activation 0, host wall time including normalization, NPU transfers, four sessions and concatenation", + "cpu_wall_s": [ + 0.01998219999950379, + 0.019212199957109988, + 0.019067600020207465, + 0.018104100017808378, + 0.01744879991747439, + 0.017045399989001453, + 0.017277600010856986, + 0.016676499973982573, + 0.017349099973216653, + 0.018242700025439262, + 0.01789639994967729, + 0.017510600038804114, + 0.017245299997739494, + 0.016315200016833842, + 0.017151699983514845, + 0.0177776999771595, + 0.016197400051169097, + 0.017280600033700466, + 0.016810699948109686, + 0.016585599980317056 + ], + "npu_wall_s": [ + 0.01046020002104342, + 0.009619400021620095, + 0.009472299949266016, + 0.008331799996085465, + 0.009359100018627942, + 0.0096539999358356, + 0.009230799973011017, + 0.009198200074024498, + 0.009146800031885505, + 0.008186700055375695, + 0.00888069998472929, + 0.00815769995097071, + 0.008661499945446849, + 0.009340799995698035, + 0.008638500003144145, + 0.008047000039368868, + 0.007870699977502227, + 0.008220899966545403, + 0.008486400009132922, + 0.00924809998832643 + ], + "cpu_wall_p50_s": 0.017280600033700466, + "cpu_wall_p95_s": 0.019212199957109988, + "npu_wall_p50_s": 0.00888069998472929, + "npu_wall_p95_s": 0.0096539999358356 +} diff --git a/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/full_head_raw.log b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/full_head_raw.log new file mode 100644 index 000000000..dcd1d477d Binary files /dev/null and b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/full_head_raw.log differ diff --git a/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/full_head_report.json b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/full_head_report.json new file mode 100644 index 000000000..ce3f0fccd --- /dev/null +++ b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/full_head_report.json @@ -0,0 +1,322 @@ +{ + "scope": "Spark-X2.5-1.7B complete 131072-logit output head only; CPU RMS normalization plus four NPU MatMul shards, no full generation", + "hardware": "AMD Ryzen AI 9 HX 370 NPU", + "os": "Windows-11-10.0.26200-SP0", + "onnx_version": "1.22.0", + "onnxruntime_version": "1.30.0", + "source_model_sha256": "f5f0c31eb438e9940ab242da8bf6f0f5dad6ebc9f6ef02d5475c21b744e595bc", + "activation_sha256": "cfa2ed8439f8fe01730319667bfd01abf2e39cf5851490346a0331304cf63194", + "source_shape": [ + 1, + 1, + 2048 + ], + "output_shape": [ + 1, + 131072 + ], + "cpu_load_s": 6.036950499983504, + "shard_load_s": [ + 3.777464500046335, + 3.785606799996458, + 4.0131896999664605, + 3.7164441999047995 + ], + "shard_artifacts": [ + { + "offset": 0, + "quantized_sha256": "b02d00c1df604e65cef383d77fb56cdf1044cbd8a2bfcb56dfda1e4ec3645c54", + "providers": [ + "vitisai", + "CPUExecutionProvider" + ] + }, + { + "offset": 32768, + "quantized_sha256": "f6a4257471869f02d4ff4ec405eeb16f46d6b50c7e02b2b6e7d08e71ac4cc0c9", + "providers": [ + "vitisai", + "CPUExecutionProvider" + ] + }, + { + "offset": 65536, + "quantized_sha256": "bc47828ef225d54e5b92a7d45737e86f14519fba5accefdde7d9429ec8d723e8", + "providers": [ + "vitisai", + "CPUExecutionProvider" + ] + }, + { + "offset": 98304, + "quantized_sha256": "0ff8e1e6d513d5debda26fa7c7088db80085f637e4ae8ddda24a505ac80c64f7", + "providers": [ + "vitisai", + "CPUExecutionProvider" + ] + } + ], + "placement": [ + { + "offset": 0, + "npu_node_events": 25, + "cpu_node_events": 50, + "profile_file": "C:\\Users\\zhout\\w2\\spark_npu_full_head_named_20260923\\shard_0_profile_2026-09-23_18-09-03_927.json" + }, + { + "offset": 32768, + "npu_node_events": 25, + "cpu_node_events": 50, + "profile_file": "C:\\Users\\zhout\\w2\\spark_npu_full_head_named_20260923\\shard_32768_profile_2026-09-23_18-09-07_764.json" + }, + { + "offset": 65536, + "npu_node_events": 25, + "cpu_node_events": 50, + "profile_file": "C:\\Users\\zhout\\w2\\spark_npu_full_head_named_20260923\\shard_65536_profile_2026-09-23_18-09-11_610.json" + }, + { + "offset": 98304, + "npu_node_events": 25, + "cpu_node_events": 50, + "profile_file": "C:\\Users\\zhout\\w2\\spark_npu_full_head_named_20260923\\shard_98304_profile_2026-09-23_18-09-15_683.json" + } + ], + "cpu_reference_rows": [ + { + "snr_db": 137.37257867357746, + "max_abs_diff": 1.9073486328125e-05, + "top1_match": true, + "finite": true, + "index": 0, + "wall_s": 0.01939839997794479 + }, + { + "snr_db": 134.38703441007158, + "max_abs_diff": 2.6702880859375e-05, + "top1_match": true, + "finite": true, + "index": 1, + "wall_s": 0.01858789997640997 + }, + { + "snr_db": 136.39026519822127, + "max_abs_diff": 2.288818359375e-05, + "top1_match": true, + "finite": true, + "index": 2, + "wall_s": 0.018647799966856837 + }, + { + "snr_db": 133.65524993196834, + "max_abs_diff": 2.288818359375e-05, + "top1_match": true, + "finite": true, + "index": 3, + "wall_s": 0.017335599986836314 + } + ], + "npu_rows": [ + { + "snr_db": 56.81097104115722, + "max_abs_diff": 0.3098640441894531, + "top1_match": true, + "finite": true, + "index": 0, + "wall_s": 0.01240850007161498, + "shard_checks": [ + { + "offset": 0, + "snr_db": 54.96589660909892, + "max_abs_diff": 0.3098640441894531, + "top1_match": true, + "finite": true + }, + { + "offset": 32768, + "snr_db": 58.25548223459347, + "max_abs_diff": 0.21026992797851562, + "top1_match": true, + "finite": true + }, + { + "offset": 65536, + "snr_db": 57.46049504780351, + "max_abs_diff": 0.225982666015625, + "top1_match": true, + "finite": true + }, + { + "offset": 98304, + "snr_db": 56.98918166146055, + "max_abs_diff": 0.22197341918945312, + "top1_match": true, + "finite": true + } + ] + }, + { + "snr_db": 56.37996250738317, + "max_abs_diff": 0.2738380432128906, + "top1_match": true, + "finite": true, + "index": 1, + "wall_s": 0.010521500022150576, + "shard_checks": [ + { + "offset": 0, + "snr_db": 54.471154152973426, + "max_abs_diff": 0.2738380432128906, + "top1_match": true, + "finite": true + }, + { + "offset": 32768, + "snr_db": 57.80354617171815, + "max_abs_diff": 0.18883514404296875, + "top1_match": true, + "finite": true + }, + { + "offset": 65536, + "snr_db": 57.06118310815952, + "max_abs_diff": 0.20765113830566406, + "top1_match": true, + "finite": true + }, + { + "offset": 98304, + "snr_db": 56.590884754816166, + "max_abs_diff": 0.21759796142578125, + "top1_match": true, + "finite": true + } + ] + }, + { + "snr_db": 56.32605724035432, + "max_abs_diff": 0.25970458984375, + "top1_match": true, + "finite": true, + "index": 2, + "wall_s": 0.009887599968351424, + "shard_checks": [ + { + "offset": 0, + "snr_db": 54.36750945439812, + "max_abs_diff": 0.2529621124267578, + "top1_match": true, + "finite": true + }, + { + "offset": 32768, + "snr_db": 57.78266475539301, + "max_abs_diff": 0.18820953369140625, + "top1_match": true, + "finite": true + }, + { + "offset": 65536, + "snr_db": 56.965567828494514, + "max_abs_diff": 0.20560836791992188, + "top1_match": true, + "finite": true + }, + { + "offset": 98304, + "snr_db": 56.59743918512855, + "max_abs_diff": 0.25970458984375, + "top1_match": true, + "finite": true + } + ] + }, + { + "snr_db": 55.42898793205113, + "max_abs_diff": 0.2625865936279297, + "top1_match": true, + "finite": true, + "index": 3, + "wall_s": 0.009681699913926423, + "shard_checks": [ + { + "offset": 0, + "snr_db": 53.731174504422455, + "max_abs_diff": 0.2625865936279297, + "top1_match": true, + "finite": true + }, + { + "offset": 32768, + "snr_db": 56.79216419378707, + "max_abs_diff": 0.18805313110351562, + "top1_match": true, + "finite": true + }, + { + "offset": 65536, + "snr_db": 56.03791485557113, + "max_abs_diff": 0.20813369750976562, + "top1_match": true, + "finite": true + }, + { + "offset": 98304, + "snr_db": 55.578637242475644, + "max_abs_diff": 0.2147674560546875, + "top1_match": true, + "finite": true + } + ] + } + ], + "timing_protocol": "one warmup then 20 serial repeats of captured activation 0, host wall time including normalization, NPU transfers, four sessions and concatenation", + "cpu_wall_s": [ + 0.015391699969768524, + 0.015810699900612235, + 0.016337100067175925, + 0.015407600090838969, + 0.016063300077803433, + 0.016168299946002662, + 0.015525899943895638, + 0.01564710005186498, + 0.015752699924632907, + 0.01605009997729212, + 0.016831799992360175, + 0.015597200021147728, + 0.015522399917244911, + 0.016015200060792267, + 0.015055799973197281, + 0.01514329994097352, + 0.01609060005284846, + 0.015278900042176247, + 0.014522499986924231, + 0.016434899996966124 + ], + "npu_wall_s": [ + 0.008257900015451014, + 0.0077850999077782035, + 0.008448700071312487, + 0.007984499912708998, + 0.008454900002107024, + 0.007974800071679056, + 0.008106399909593165, + 0.008349800016731024, + 0.007990399957634509, + 0.008053900091908872, + 0.00843759998679161, + 0.008772099972702563, + 0.009113600011914968, + 0.008478199946694076, + 0.008262300048954785, + 0.008295800071209669, + 0.008625500020571053, + 0.008759499993175268, + 0.008324099937453866, + 0.007927400059998035 + ], + "cpu_wall_p50_s": 0.01564710005186498, + "cpu_wall_p95_s": 0.016434899996966124, + "npu_wall_p50_s": 0.008295800071209669, + "npu_wall_p95_s": 0.008772099972702563 +} diff --git a/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/local_bundle_manifest.json b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/local_bundle_manifest.json new file mode 100644 index 000000000..38a3372d0 --- /dev/null +++ b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/local_bundle_manifest.json @@ -0,0 +1,19 @@ +{ + "schema_version": 1, + "component": "spark_x2_5_1_7b_output_head_amd_npu", + "files": { + "graph.onnx": "e2ae9c3fd7071e8ff628f209e6140c8a023bfaad59814b80a936c6b263f33eaf", + "inputs.npz": "b249521cbb56f029d9243354e689e1d11a4dcdbc0f47a7742d4679b2a92dfef0", + "validation_report.json": "903bd26d67f7592c3ca5d8888c4586ce231633793554adf041a73d1b8906c3c0" + }, + "metadata": { + "graph_file": "graph.onnx", + "example_inputs_file": "inputs.npz", + "checkpoint_revision": "448e61eb392c00f2c403185c5b56d5e0665bfaab", + "precision": "A16W8 QDQ, FP32 I/O", + "layout": "[1,1,2048] pre-final-norm input -> [1,131072] logits", + "exporter": "probe_spark_amd_npu_lm_head.py --composite-with-norm", + "calibration": "four captured real Spark pre-final-norm activations", + "validation": "validation_report.json; four top-1 matches, SNR >40 dB; component only" + } +} diff --git a/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/omni_npu_placement_profile.json b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/omni_npu_placement_profile.json new file mode 100644 index 000000000..9bd3c411e --- /dev/null +++ b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/omni_npu_placement_profile.json @@ -0,0 +1,14 @@ +[ +{"cat" : "Session","pid" :47016,"tid" :21856,"dur" :78892,"ts" :13,"ph" : "X","name" :"model_loading_uri","args" : {}}, +{"cat" : "Session","pid" :47016,"tid" :21856,"dur" :15337912,"ts" :79717,"ph" : "X","name" :"session_initialization","args" : {}}, +{"cat" : "Node","pid" :47016,"tid" :21856,"dur" :136,"ts" :15417934,"ph" : "X","name" :"/Mul_1/SimplifiedLayerNormFusion/_kernel_time","args" : {"mem_bytes_in_use" : "8222","mem_in_use_peak" : "8222","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-1-intra-op", "thread_id": "21856", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"21160": {"num_run": 0, "core": -1},"41180": {"num_run": 0, "core": -1},"17648": {"num_run": 0, "core": -1},"40988": {"num_run": 0, "core": -1},"49300": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"21356": {"num_run": 0, "core": -1},"33664": {"num_run": 0, "core": -1},"840": {"num_run": 0, "core": -1},"46536": {"num_run": 0, "core": -1},"19312": {"num_run": 0, "core": -1}}},"activation_size" : "8192","output_size" : "8192","input_type_shape" : [{"float":[1,1,2048]},{"float":[2048]}],"mem_arena_held_delta" : "1048576","mem_bytes_requested_in_use" : "8222","parameter_size" : "8192","mem_requested_in_use_delta" : "8192","op_name" : "SimplifiedLayerNormalization","mem_in_use_delta" : "8192","provider" : "CPUExecutionProvider","node_index" : "33","output_type_shape" : [{"float":[1,1,2048]}],"mem_arena_held" : "1048606"}}, +{"cat" : "Node","pid" :47016,"tid" :21856,"dur" :63,"ts" :15418081,"ph" : "X","name" :"/Mul_1_output_0_q_kernel_time","args" : {"input_type_shape" : [{"float":[1,1,2048]},{"float":[]},{"uint16":[]}],"op_name" : "QuantizeLinear","parameter_size" : "6","node_index" : "30","mem_arena_held_delta" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-1-intra-op", "thread_id": "21856", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"21160": {"num_run": 0, "core": -1},"41180": {"num_run": 0, "core": -1},"17648": {"num_run": 0, "core": -1},"40988": {"num_run": 0, "core": -1},"49300": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"21356": {"num_run": 0, "core": -1},"33664": {"num_run": 0, "core": -1},"840": {"num_run": 0, "core": -1},"46536": {"num_run": 0, "core": -1},"19312": {"num_run": 0, "core": -1}}},"provider" : "CPUExecutionProvider","mem_requested_in_use_delta" : "4096","activation_size" : "8192","output_size" : "4096","mem_arena_held" : "1048606","mem_in_use_delta" : "4096","output_type_shape" : [{"uint16":[1,1,2048]}],"mem_bytes_in_use" : "12318","mem_in_use_peak" : "12318","mem_bytes_requested_in_use" : "12318"}}, +{"cat" : "Node","pid" :47016,"tid" :21856,"dur" :9064,"ts" :15418160,"ph" : "X","name" :"vitisai_vitisai_13528307914143146056_0_0_kernel_time","args" : {"mem_bytes_in_use" : "266270","activation_size" : "4096","provider" : "vitisai","mem_in_use_peak" : "266270","mem_arena_held_delta" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-1-intra-op", "thread_id": "21856", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"21160": {"num_run": 0, "core": -1},"41180": {"num_run": 0, "core": -1},"17648": {"num_run": 0, "core": -1},"40988": {"num_run": 0, "core": -1},"49300": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"21356": {"num_run": 0, "core": -1},"33664": {"num_run": 0, "core": -1},"840": {"num_run": 0, "core": -1},"46536": {"num_run": 0, "core": -1},"19312": {"num_run": 0, "core": -1}}},"output_type_shape" : [{"uint16":[1,32768]},{"uint16":[1,32768]},{"uint16":[1,32768]},{"uint16":[1,32768]}],"node_index" : "32","mem_arena_held" : "1048606","parameter_size" : "0","op_name" : "vitisai_13528307914143146056_0","mem_in_use_delta" : "262144","output_size" : "262144","mem_requested_in_use_delta" : "262144","input_type_shape" : [{"uint16":[1,1,2048]}],"mem_bytes_requested_in_use" : "266270"}}, +{"cat" : "Node","pid" :47016,"tid" :21856,"dur" :75,"ts" :15427246,"ph" : "X","name" :"logits_offset_98304_DequantizeLinear_kernel_time","args" : {"mem_arena_held" : "1048606","mem_bytes_requested_in_use" : "393246","provider" : "CPUExecutionProvider","mem_in_use_delta" : "131072","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"output_size" : "131072","mem_bytes_in_use" : "393246","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-1-intra-op", "thread_id": "21856", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"21160": {"num_run": 0, "core": -1},"41180": {"num_run": 0, "core": -1},"17648": {"num_run": 0, "core": -1},"40988": {"num_run": 0, "core": -1},"49300": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"21356": {"num_run": 0, "core": -1},"33664": {"num_run": 0, "core": -1},"840": {"num_run": 0, "core": -1},"46536": {"num_run": 0, "core": -1},"19312": {"num_run": 0, "core": -1}}},"parameter_size" : "6","mem_in_use_peak" : "393246","node_index" : "25","activation_size" : "65536","op_name" : "DequantizeLinear","output_type_shape" : [{"float":[1,32768]}],"mem_requested_in_use_delta" : "131072","mem_arena_held_delta" : "0"}}, +{"cat" : "Node","pid" :47016,"tid" :21856,"dur" :46,"ts" :15427327,"ph" : "X","name" :"logits_offset_65536_DequantizeLinear_kernel_time","args" : {"input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"activation_size" : "65536","mem_arena_held_delta" : "0","mem_bytes_requested_in_use" : "458782","mem_requested_in_use_delta" : "131072","mem_in_use_delta" : "131072","op_name" : "DequantizeLinear","output_type_shape" : [{"float":[1,32768]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-1-intra-op", "thread_id": "21856", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"21160": {"num_run": 0, "core": -1},"41180": {"num_run": 0, "core": -1},"17648": {"num_run": 0, "core": -1},"40988": {"num_run": 0, "core": -1},"49300": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"21356": {"num_run": 0, "core": -1},"33664": {"num_run": 0, "core": -1},"840": {"num_run": 0, "core": -1},"46536": {"num_run": 0, "core": -1},"19312": {"num_run": 0, "core": -1}}},"mem_in_use_peak" : "458782","output_size" : "131072","parameter_size" : "6","node_index" : "24","mem_arena_held" : "1048606","provider" : "CPUExecutionProvider","mem_bytes_in_use" : "458782"}}, +{"cat" : "Node","pid" :47016,"tid" :21856,"dur" :33,"ts" :15427381,"ph" : "X","name" :"logits_offset_32768_DequantizeLinear_kernel_time","args" : {"mem_in_use_peak" : "536606","output_type_shape" : [{"float":[1,32768]}],"node_index" : "23","provider" : "CPUExecutionProvider","parameter_size" : "6","mem_in_use_delta" : "143360","mem_bytes_requested_in_use" : "524318","mem_requested_in_use_delta" : "131072","output_size" : "131072","mem_bytes_in_use" : "536606","mem_arena_held_delta" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-1-intra-op", "thread_id": "21856", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"21160": {"num_run": 0, "core": -1},"41180": {"num_run": 0, "core": -1},"17648": {"num_run": 0, "core": -1},"40988": {"num_run": 0, "core": -1},"49300": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"21356": {"num_run": 0, "core": -1},"33664": {"num_run": 0, "core": -1},"840": {"num_run": 0, "core": -1},"46536": {"num_run": 0, "core": -1},"19312": {"num_run": 0, "core": -1}}},"activation_size" : "65536","op_name" : "DequantizeLinear","mem_arena_held" : "1048606","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}]}}, +{"cat" : "Node","pid" :47016,"tid" :21856,"dur" :46,"ts" :15427420,"ph" : "X","name" :"logits_offset_0_DequantizeLinear_kernel_time","args" : {"mem_bytes_requested_in_use" : "589854","parameter_size" : "6","provider" : "CPUExecutionProvider","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_bytes_in_use" : "602142","mem_in_use_peak" : "602142","mem_arena_held" : "1048606","mem_in_use_delta" : "131072","mem_arena_held_delta" : "0","activation_size" : "65536","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-1-intra-op", "thread_id": "21856", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"21160": {"num_run": 0, "core": -1},"41180": {"num_run": 0, "core": -1},"17648": {"num_run": 0, "core": -1},"40988": {"num_run": 0, "core": -1},"49300": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"21356": {"num_run": 0, "core": -1},"33664": {"num_run": 0, "core": -1},"840": {"num_run": 0, "core": -1},"46536": {"num_run": 0, "core": -1},"19312": {"num_run": 0, "core": -1}}},"mem_requested_in_use_delta" : "131072","output_type_shape" : [{"float":[1,32768]}],"op_name" : "DequantizeLinear","output_size" : "131072","node_index" : "22"}}, +{"cat" : "Node","pid" :47016,"tid" :21856,"dur" :192,"ts" :15427473,"ph" : "X","name" :"spark_lm_head_concat_kernel_time","args" : {"mem_requested_in_use_delta" : "524288","mem_arena_held" : "3145758","mem_arena_held_delta" : "2097152","mem_in_use_peak" : "1060894","mem_in_use_delta" : "524288","output_size" : "524288","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-1-intra-op", "thread_id": "21856", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"21160": {"num_run": 0, "core": -1},"41180": {"num_run": 0, "core": -1},"17648": {"num_run": 0, "core": -1},"40988": {"num_run": 0, "core": -1},"49300": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"21356": {"num_run": 0, "core": -1},"33664": {"num_run": 0, "core": -1},"840": {"num_run": 0, "core": -1},"46536": {"num_run": 0, "core": -1},"19312": {"num_run": 0, "core": -1}}},"output_type_shape" : [{"float":[1,131072]}],"activation_size" : "524288","input_type_shape" : [{"float":[1,32768]},{"float":[1,32768]},{"float":[1,32768]},{"float":[1,32768]}],"node_index" : "26","op_name" : "Concat","mem_bytes_requested_in_use" : "1048606","mem_bytes_in_use" : "1060894","provider" : "CPUExecutionProvider","parameter_size" : "0"}}, +{"cat" : "Session","pid" :47016,"tid" :21856,"dur" :9760,"ts" :15417925,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :47016,"tid" :21856,"dur" :9849,"ts" :15417871,"ph" : "X","name" :"model_run","args" : {}} +] diff --git a/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/real_acts_lmhead.npz b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/real_acts_lmhead.npz new file mode 100644 index 000000000..463788c2a Binary files /dev/null and b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/real_acts_lmhead.npz differ diff --git a/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/shards/shard_0.json b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/shards/shard_0.json new file mode 100644 index 000000000..22d1207f0 --- /dev/null +++ b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/shards/shard_0.json @@ -0,0 +1,95 @@ +{ + "scope": "Spark-X2.5-1.7B output head only; no prefill/decode/KV/sampling or full generation", + "vocab_shard_size": 32768, + "vocab_shard_offset": 0, + "flatten_input": true, + "batch_repeats": 1, + "matmul_only": true, + "hardware": "AMD Ryzen AI 9 HX 370 NPU", + "os": "Windows-11-10.0.26200-SP0", + "onnx_version": "1.22.0", + "onnxruntime_version": "1.30.0", + "source_model_sha256": "f5f0c31eb438e9940ab242da8bf6f0f5dad6ebc9f6ef02d5475c21b744e595bc", + "prepared_model_sha256": "a0ede31c101b8243e68f1ef7409872e3c54910e13722af371d5e2470570a0c1b", + "calibration_sha256": "cfa2ed8439f8fe01730319667bfd01abf2e39cf5851490346a0331304cf63194", + "opset21_sha256": "a0ede31c101b8243e68f1ef7409872e3c54910e13722af371d5e2470570a0c1b", + "quantized_sha256": "b02d00c1df604e65cef383d77fb56cdf1044cbd8a2bfcb56dfda1e4ec3645c54", + "quantization": "opset 21, MatMul only, QDQ, QUInt16 activations/QInt8 weights, MinMax on four real activations", + "input_shape": [ + 1, + 2048 + ], + "output_shape": [ + 1, + 32768 + ], + "source_cpu_reference": [ + { + "snr_db": 137.4172442852821, + "max_abs_diff": 1.9073486328125e-05, + "top1_match": true, + "finite": true + }, + { + "snr_db": 134.53046900546502, + "max_abs_diff": 2.288818359375e-05, + "top1_match": true, + "finite": true + }, + { + "snr_db": 137.96564797727288, + "max_abs_diff": 1.9073486328125e-05, + "top1_match": true, + "finite": true + }, + { + "snr_db": 136.58623781989002, + "max_abs_diff": 2.288818359375e-05, + "top1_match": true, + "finite": true + } + ], + "npu_devices": 1, + "session_providers": [ + "vitisai", + "CPUExecutionProvider" + ], + "npu_node_events": 4, + "cpu_node_events": 8, + "load_s": 5.609775200020522, + "npu_rows": [ + { + "snr_db": 54.96589660909892, + "max_abs_diff": 0.3098640441894531, + "top1_match": true, + "finite": true, + "index": 0, + "wall_s": 0.00410420005209744 + }, + { + "snr_db": 54.471154152973426, + "max_abs_diff": 0.2738380432128906, + "top1_match": true, + "finite": true, + "index": 1, + "wall_s": 0.003021199954673648 + }, + { + "snr_db": 54.36750945439812, + "max_abs_diff": 0.2529621124267578, + "top1_match": true, + "finite": true, + "index": 2, + "wall_s": 0.0029224000172689557 + }, + { + "snr_db": 53.731174504422455, + "max_abs_diff": 0.2625865936279297, + "top1_match": true, + "finite": true, + "index": 3, + "wall_s": 0.002828099997714162 + } + ], + "profile_file": "C:\\Users\\zhout\\w2\\spark_npu_full_shards_named_20260923\\shard_0\\spark_lm_head_npu_profile_2026-09-23_18-07-13_928.json" +} diff --git a/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/shards/shard_0_co_resident_profile.json b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/shards/shard_0_co_resident_profile.json new file mode 100644 index 000000000..352979ea6 --- /dev/null +++ b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/shards/shard_0_co_resident_profile.json @@ -0,0 +1,129 @@ +[ +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :24282,"ts" :14,"ph" : "X","name" :"model_loading_uri","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :3750762,"ts" :25454,"ph" : "X","name" :"session_initialization","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :90,"ts" :15471954,"ph" : "X","name" :"normalized_x_offset_0_QuantizeLinear_kernel_time","args" : {"mem_requested_in_use_delta" : "4096","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_in_use_peak" : "4108","activation_size" : "8192","mem_bytes_requested_in_use" : "4108","mem_in_use_delta" : "4096","parameter_size" : "6","mem_arena_held" : "1048588","output_size" : "4096","node_index" : "0","output_type_shape" : [{"uint16":[1,2048]}],"mem_arena_held_delta" : "1048576","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"op_name" : "QuantizeLinear","mem_bytes_in_use" : "4108","provider" : "CPUExecutionProvider"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :2823,"ts" :15472054,"ph" : "X","name" :"vitisai_vitisai_9474036293895030537_0_0_kernel_time","args" : {"output_size" : "65536","activation_size" : "4096","parameter_size" : "0","mem_in_use_delta" : "65536","input_type_shape" : [{"uint16":[1,2048]}],"output_type_shape" : [{"uint16":[1,32768]}],"mem_bytes_in_use" : "69644","mem_arena_held_delta" : "0","mem_in_use_peak" : "69644","mem_bytes_requested_in_use" : "69644","mem_requested_in_use_delta" : "65536","op_name" : "vitisai_9474036293895030537_0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"mem_arena_held" : "1048588","node_index" : "6","provider" : "vitisai"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :52,"ts" :15474885,"ph" : "X","name" :"logits_offset_0_DequantizeLinear_kernel_time","args" : {"mem_in_use_delta" : "131072","mem_bytes_in_use" : "196620","mem_in_use_peak" : "196620","output_type_shape" : [{"float":[1,32768]}],"mem_arena_held_delta" : "0","output_size" : "131072","parameter_size" : "6","provider" : "CPUExecutionProvider","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"op_name" : "DequantizeLinear","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"node_index" : "5","mem_bytes_requested_in_use" : "196620","activation_size" : "65536","mem_arena_held" : "1048588","mem_requested_in_use_delta" : "131072"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2995,"ts" :15471949,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :3036,"ts" :15471922,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :42,"ts" :15485299,"ph" : "X","name" :"normalized_x_offset_0_QuantizeLinear_kernel_time","args" : {"output_size" : "4096","op_name" : "QuantizeLinear","parameter_size" : "6","output_type_shape" : [{"uint16":[1,2048]}],"activation_size" : "8192","mem_in_use_peak" : "196620","provider" : "CPUExecutionProvider","node_index" : "0","mem_bytes_in_use" : "69644","mem_requested_in_use_delta" : "0","mem_bytes_requested_in_use" : "69644","mem_in_use_delta" : "0","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"mem_arena_held_delta" : "0","mem_arena_held" : "1048588"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1965,"ts" :15485349,"ph" : "X","name" :"vitisai_vitisai_9474036293895030537_0_0_kernel_time","args" : {"mem_bytes_in_use" : "69644","input_type_shape" : [{"uint16":[1,2048]}],"mem_bytes_requested_in_use" : "69644","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"output_type_shape" : [{"uint16":[1,32768]}],"mem_in_use_delta" : "0","parameter_size" : "0","activation_size" : "4096","mem_in_use_peak" : "196620","mem_requested_in_use_delta" : "0","provider" : "vitisai","op_name" : "vitisai_9474036293895030537_0","node_index" : "6","mem_arena_held_delta" : "0","mem_arena_held" : "1048588","output_size" : "65536"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :50,"ts" :15487324,"ph" : "X","name" :"logits_offset_0_DequantizeLinear_kernel_time","args" : {"mem_bytes_requested_in_use" : "200716","activation_size" : "65536","mem_bytes_in_use" : "200716","mem_in_use_peak" : "200716","node_index" : "5","mem_arena_held" : "1048588","output_type_shape" : [{"float":[1,32768]}],"output_size" : "131072","parameter_size" : "6","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"mem_requested_in_use_delta" : "131072","provider" : "CPUExecutionProvider","op_name" : "DequantizeLinear","mem_in_use_delta" : "131072","mem_arena_held_delta" : "0"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2086,"ts" :15485295,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2131,"ts" :15485265,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :38,"ts" :15497735,"ph" : "X","name" :"normalized_x_offset_0_QuantizeLinear_kernel_time","args" : {"provider" : "CPUExecutionProvider","mem_arena_held" : "1048588","mem_bytes_in_use" : "69644","mem_arena_held_delta" : "0","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_bytes_requested_in_use" : "69644","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"op_name" : "QuantizeLinear","mem_requested_in_use_delta" : "0","activation_size" : "8192","parameter_size" : "6","output_type_shape" : [{"uint16":[1,2048]}],"output_size" : "4096","node_index" : "0","mem_in_use_peak" : "200716","mem_in_use_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1955,"ts" :15497779,"ph" : "X","name" :"vitisai_vitisai_9474036293895030537_0_0_kernel_time","args" : {"op_name" : "vitisai_9474036293895030537_0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"output_type_shape" : [{"uint16":[1,32768]}],"mem_in_use_delta" : "0","output_size" : "65536","provider" : "vitisai","mem_in_use_peak" : "200716","mem_bytes_requested_in_use" : "69644","mem_arena_held_delta" : "0","activation_size" : "4096","mem_requested_in_use_delta" : "0","node_index" : "6","input_type_shape" : [{"uint16":[1,2048]}],"mem_bytes_in_use" : "69644","parameter_size" : "0","mem_arena_held" : "1048588"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :27,"ts" :15499746,"ph" : "X","name" :"logits_offset_0_DequantizeLinear_kernel_time","args" : {"mem_in_use_delta" : "131072","mem_arena_held_delta" : "0","output_size" : "131072","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"mem_in_use_peak" : "200716","node_index" : "5","mem_arena_held" : "1048588","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"provider" : "CPUExecutionProvider","activation_size" : "65536","output_type_shape" : [{"float":[1,32768]}],"mem_bytes_requested_in_use" : "200716","parameter_size" : "6","mem_bytes_in_use" : "200716","op_name" : "DequantizeLinear","mem_requested_in_use_delta" : "131072"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2044,"ts" :15497733,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2077,"ts" :15497712,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :28,"ts" :15508754,"ph" : "X","name" :"normalized_x_offset_0_QuantizeLinear_kernel_time","args" : {"mem_in_use_delta" : "0","mem_arena_held_delta" : "0","mem_requested_in_use_delta" : "0","mem_bytes_requested_in_use" : "69644","op_name" : "QuantizeLinear","output_size" : "4096","mem_arena_held" : "1048588","output_type_shape" : [{"uint16":[1,2048]}],"parameter_size" : "6","node_index" : "0","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"provider" : "CPUExecutionProvider","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"mem_bytes_in_use" : "69644","mem_in_use_peak" : "200716","activation_size" : "8192"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1890,"ts" :15508788,"ph" : "X","name" :"vitisai_vitisai_9474036293895030537_0_0_kernel_time","args" : {"mem_bytes_requested_in_use" : "69644","mem_arena_held" : "1048588","node_index" : "6","mem_in_use_delta" : "0","output_type_shape" : [{"uint16":[1,32768]}],"provider" : "vitisai","parameter_size" : "0","input_type_shape" : [{"uint16":[1,2048]}],"mem_requested_in_use_delta" : "0","op_name" : "vitisai_9474036293895030537_0","mem_arena_held_delta" : "0","mem_in_use_peak" : "200716","activation_size" : "4096","mem_bytes_in_use" : "69644","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"output_size" : "65536"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :24,"ts" :15510684,"ph" : "X","name" :"logits_offset_0_DequantizeLinear_kernel_time","args" : {"mem_arena_held_delta" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"provider" : "CPUExecutionProvider","mem_bytes_requested_in_use" : "200716","output_size" : "131072","parameter_size" : "6","mem_in_use_peak" : "200716","mem_in_use_delta" : "131072","mem_arena_held" : "1048588","mem_requested_in_use_delta" : "131072","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"node_index" : "5","mem_bytes_in_use" : "200716","output_type_shape" : [{"float":[1,32768]}],"op_name" : "DequantizeLinear","activation_size" : "65536"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1961,"ts" :15508752,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1986,"ts" :15508736,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :33,"ts" :15519569,"ph" : "X","name" :"normalized_x_offset_0_QuantizeLinear_kernel_time","args" : {"output_type_shape" : [{"uint16":[1,2048]}],"op_name" : "QuantizeLinear","node_index" : "0","parameter_size" : "6","mem_in_use_delta" : "0","activation_size" : "8192","mem_arena_held_delta" : "0","mem_bytes_in_use" : "69644","mem_arena_held" : "1048588","mem_in_use_peak" : "200716","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_bytes_requested_in_use" : "69644","output_size" : "4096","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"provider" : "CPUExecutionProvider","mem_requested_in_use_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :2011,"ts" :15519608,"ph" : "X","name" :"vitisai_vitisai_9474036293895030537_0_0_kernel_time","args" : {"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"provider" : "vitisai","mem_arena_held" : "1048588","mem_in_use_delta" : "0","mem_requested_in_use_delta" : "0","mem_bytes_in_use" : "69644","output_type_shape" : [{"uint16":[1,32768]}],"input_type_shape" : [{"uint16":[1,2048]}],"output_size" : "65536","activation_size" : "4096","op_name" : "vitisai_9474036293895030537_0","parameter_size" : "0","mem_in_use_peak" : "200716","mem_bytes_requested_in_use" : "69644","mem_arena_held_delta" : "0","node_index" : "6"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :52,"ts" :15521631,"ph" : "X","name" :"logits_offset_0_DequantizeLinear_kernel_time","args" : {"mem_in_use_peak" : "200716","mem_arena_held" : "1048588","activation_size" : "65536","parameter_size" : "6","provider" : "CPUExecutionProvider","node_index" : "5","mem_arena_held_delta" : "0","mem_in_use_delta" : "131072","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"mem_requested_in_use_delta" : "131072","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"output_type_shape" : [{"float":[1,32768]}],"mem_bytes_requested_in_use" : "200716","output_size" : "131072","mem_bytes_in_use" : "200716","op_name" : "DequantizeLinear"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2124,"ts" :15519566,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2163,"ts" :15519544,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :25,"ts" :15527793,"ph" : "X","name" :"normalized_x_offset_0_QuantizeLinear_kernel_time","args" : {"node_index" : "0","mem_in_use_delta" : "0","mem_arena_held" : "1048588","output_type_shape" : [{"uint16":[1,2048]}],"op_name" : "QuantizeLinear","mem_in_use_peak" : "200716","parameter_size" : "6","mem_bytes_requested_in_use" : "69644","provider" : "CPUExecutionProvider","mem_bytes_in_use" : "69644","mem_requested_in_use_delta" : "0","mem_arena_held_delta" : "0","activation_size" : "8192","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"output_size" : "4096"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1885,"ts" :15527821,"ph" : "X","name" :"vitisai_vitisai_9474036293895030537_0_0_kernel_time","args" : {"mem_in_use_peak" : "200716","input_type_shape" : [{"uint16":[1,2048]}],"output_size" : "65536","op_name" : "vitisai_9474036293895030537_0","mem_in_use_delta" : "0","activation_size" : "4096","parameter_size" : "0","node_index" : "6","mem_bytes_in_use" : "69644","mem_arena_held" : "1048588","mem_requested_in_use_delta" : "0","output_type_shape" : [{"uint16":[1,32768]}],"provider" : "vitisai","mem_bytes_requested_in_use" : "69644","mem_arena_held_delta" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}}}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :20,"ts" :15529714,"ph" : "X","name" :"logits_offset_0_DequantizeLinear_kernel_time","args" : {"parameter_size" : "6","provider" : "CPUExecutionProvider","mem_requested_in_use_delta" : "131072","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"mem_arena_held" : "1048588","op_name" : "DequantizeLinear","mem_bytes_in_use" : "200716","activation_size" : "65536","output_type_shape" : [{"float":[1,32768]}],"output_size" : "131072","mem_in_use_delta" : "131072","node_index" : "5","mem_bytes_requested_in_use" : "200716","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_in_use_peak" : "200716","mem_arena_held_delta" : "0"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1946,"ts" :15527791,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1961,"ts" :15527782,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :15,"ts" :15536068,"ph" : "X","name" :"normalized_x_offset_0_QuantizeLinear_kernel_time","args" : {"node_index" : "0","mem_in_use_peak" : "200716","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"mem_bytes_requested_in_use" : "69644","parameter_size" : "6","op_name" : "QuantizeLinear","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"provider" : "CPUExecutionProvider","mem_requested_in_use_delta" : "0","output_size" : "4096","mem_in_use_delta" : "0","activation_size" : "8192","mem_arena_held" : "1048588","mem_arena_held_delta" : "0","output_type_shape" : [{"uint16":[1,2048]}],"mem_bytes_in_use" : "69644"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1835,"ts" :15536087,"ph" : "X","name" :"vitisai_vitisai_9474036293895030537_0_0_kernel_time","args" : {"op_name" : "vitisai_9474036293895030537_0","mem_requested_in_use_delta" : "0","input_type_shape" : [{"uint16":[1,2048]}],"output_size" : "65536","mem_in_use_delta" : "0","mem_arena_held" : "1048588","mem_arena_held_delta" : "0","parameter_size" : "0","mem_bytes_in_use" : "69644","mem_bytes_requested_in_use" : "69644","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"activation_size" : "4096","provider" : "vitisai","mem_in_use_peak" : "200716","output_type_shape" : [{"uint16":[1,32768]}],"node_index" : "6"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :16,"ts" :15537926,"ph" : "X","name" :"logits_offset_0_DequantizeLinear_kernel_time","args" : {"activation_size" : "65536","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_bytes_requested_in_use" : "200716","op_name" : "DequantizeLinear","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"mem_arena_held_delta" : "0","mem_arena_held" : "1048588","output_type_shape" : [{"float":[1,32768]}],"mem_in_use_delta" : "131072","output_size" : "131072","mem_bytes_in_use" : "200716","parameter_size" : "6","mem_in_use_peak" : "200716","mem_requested_in_use_delta" : "131072","node_index" : "5","provider" : "CPUExecutionProvider"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1879,"ts" :15536067,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1890,"ts" :15536059,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :27,"ts" :15543818,"ph" : "X","name" :"normalized_x_offset_0_QuantizeLinear_kernel_time","args" : {"mem_requested_in_use_delta" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"mem_arena_held" : "1048588","activation_size" : "8192","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_bytes_in_use" : "69644","mem_in_use_peak" : "200716","parameter_size" : "6","mem_in_use_delta" : "0","op_name" : "QuantizeLinear","output_size" : "4096","mem_bytes_requested_in_use" : "69644","node_index" : "0","mem_arena_held_delta" : "0","output_type_shape" : [{"uint16":[1,2048]}],"provider" : "CPUExecutionProvider"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1830,"ts" :15543850,"ph" : "X","name" :"vitisai_vitisai_9474036293895030537_0_0_kernel_time","args" : {"output_size" : "65536","parameter_size" : "0","output_type_shape" : [{"uint16":[1,32768]}],"activation_size" : "4096","node_index" : "6","input_type_shape" : [{"uint16":[1,2048]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"mem_in_use_peak" : "200716","mem_requested_in_use_delta" : "0","mem_bytes_requested_in_use" : "69644","mem_arena_held" : "1048588","mem_arena_held_delta" : "0","provider" : "vitisai","op_name" : "vitisai_9474036293895030537_0","mem_bytes_in_use" : "69644","mem_in_use_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :17,"ts" :15545684,"ph" : "X","name" :"logits_offset_0_DequantizeLinear_kernel_time","args" : {"op_name" : "DequantizeLinear","mem_requested_in_use_delta" : "131072","node_index" : "5","provider" : "CPUExecutionProvider","mem_bytes_in_use" : "200716","mem_arena_held" : "1048588","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_arena_held_delta" : "0","mem_in_use_peak" : "200716","output_type_shape" : [{"float":[1,32768]}],"output_size" : "131072","parameter_size" : "6","activation_size" : "65536","mem_bytes_requested_in_use" : "200716","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"mem_in_use_delta" : "131072"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1887,"ts" :15543816,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1900,"ts" :15543808,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :17,"ts" :15552329,"ph" : "X","name" :"normalized_x_offset_0_QuantizeLinear_kernel_time","args" : {"mem_arena_held" : "1048588","output_size" : "4096","mem_in_use_peak" : "200716","mem_in_use_delta" : "0","output_type_shape" : [{"uint16":[1,2048]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"activation_size" : "8192","mem_bytes_requested_in_use" : "69644","provider" : "CPUExecutionProvider","node_index" : "0","mem_bytes_in_use" : "69644","op_name" : "QuantizeLinear","mem_arena_held_delta" : "0","mem_requested_in_use_delta" : "0","parameter_size" : "6"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1883,"ts" :15552353,"ph" : "X","name" :"vitisai_vitisai_9474036293895030537_0_0_kernel_time","args" : {"node_index" : "6","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"mem_in_use_delta" : "0","mem_arena_held_delta" : "0","op_name" : "vitisai_9474036293895030537_0","input_type_shape" : [{"uint16":[1,2048]}],"output_size" : "65536","mem_requested_in_use_delta" : "0","output_type_shape" : [{"uint16":[1,32768]}],"mem_arena_held" : "1048588","provider" : "vitisai","mem_bytes_requested_in_use" : "69644","mem_bytes_in_use" : "69644","mem_in_use_peak" : "200716","parameter_size" : "0","activation_size" : "4096"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :22,"ts" :15554242,"ph" : "X","name" :"logits_offset_0_DequantizeLinear_kernel_time","args" : {"node_index" : "5","op_name" : "DequantizeLinear","mem_in_use_peak" : "200716","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"parameter_size" : "6","mem_requested_in_use_delta" : "131072","mem_bytes_in_use" : "200716","mem_bytes_requested_in_use" : "200716","provider" : "CPUExecutionProvider","mem_in_use_delta" : "131072","activation_size" : "65536","mem_arena_held" : "1048588","output_type_shape" : [{"float":[1,32768]}],"output_size" : "131072","mem_arena_held_delta" : "0","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}]}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1939,"ts" :15552327,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1958,"ts" :15552316,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :33,"ts" :15560273,"ph" : "X","name" :"normalized_x_offset_0_QuantizeLinear_kernel_time","args" : {"op_name" : "QuantizeLinear","parameter_size" : "6","mem_arena_held" : "1048588","mem_arena_held_delta" : "0","node_index" : "0","output_size" : "4096","mem_bytes_in_use" : "69644","mem_in_use_delta" : "0","mem_requested_in_use_delta" : "0","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_bytes_requested_in_use" : "69644","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"activation_size" : "8192","mem_in_use_peak" : "200716","output_type_shape" : [{"uint16":[1,2048]}],"provider" : "CPUExecutionProvider"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1951,"ts" :15560312,"ph" : "X","name" :"vitisai_vitisai_9474036293895030537_0_0_kernel_time","args" : {"mem_arena_held" : "1048588","mem_bytes_in_use" : "69644","mem_bytes_requested_in_use" : "69644","activation_size" : "4096","mem_arena_held_delta" : "0","parameter_size" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"mem_requested_in_use_delta" : "0","output_type_shape" : [{"uint16":[1,32768]}],"provider" : "vitisai","op_name" : "vitisai_9474036293895030537_0","mem_in_use_delta" : "0","input_type_shape" : [{"uint16":[1,2048]}],"mem_in_use_peak" : "200716","node_index" : "6","output_size" : "65536"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :35,"ts" :15562269,"ph" : "X","name" :"logits_offset_0_DequantizeLinear_kernel_time","args" : {"mem_bytes_requested_in_use" : "200716","mem_bytes_in_use" : "200716","activation_size" : "65536","output_type_shape" : [{"float":[1,32768]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"mem_in_use_peak" : "200716","parameter_size" : "6","node_index" : "5","provider" : "CPUExecutionProvider","mem_arena_held" : "1048588","mem_arena_held_delta" : "0","mem_requested_in_use_delta" : "131072","op_name" : "DequantizeLinear","mem_in_use_delta" : "131072","output_size" : "131072","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}]}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2035,"ts" :15560271,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2048,"ts" :15560262,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :16,"ts" :15568775,"ph" : "X","name" :"normalized_x_offset_0_QuantizeLinear_kernel_time","args" : {"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"mem_requested_in_use_delta" : "0","mem_bytes_in_use" : "69644","mem_in_use_peak" : "200716","output_size" : "4096","activation_size" : "8192","op_name" : "QuantizeLinear","mem_arena_held" : "1048588","provider" : "CPUExecutionProvider","mem_bytes_requested_in_use" : "69644","output_type_shape" : [{"uint16":[1,2048]}],"node_index" : "0","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"parameter_size" : "6","mem_arena_held_delta" : "0","mem_in_use_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1878,"ts" :15568794,"ph" : "X","name" :"vitisai_vitisai_9474036293895030537_0_0_kernel_time","args" : {"provider" : "vitisai","mem_requested_in_use_delta" : "0","output_size" : "65536","output_type_shape" : [{"uint16":[1,32768]}],"mem_arena_held" : "1048588","mem_in_use_peak" : "200716","op_name" : "vitisai_9474036293895030537_0","input_type_shape" : [{"uint16":[1,2048]}],"parameter_size" : "0","mem_bytes_in_use" : "69644","node_index" : "6","mem_in_use_delta" : "0","activation_size" : "4096","mem_arena_held_delta" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"mem_bytes_requested_in_use" : "69644"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :21,"ts" :15570677,"ph" : "X","name" :"logits_offset_0_DequantizeLinear_kernel_time","args" : {"output_type_shape" : [{"float":[1,32768]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"output_size" : "131072","mem_in_use_peak" : "200716","op_name" : "DequantizeLinear","mem_in_use_delta" : "131072","node_index" : "5","activation_size" : "65536","provider" : "CPUExecutionProvider","mem_requested_in_use_delta" : "131072","parameter_size" : "6","mem_bytes_in_use" : "200716","mem_arena_held" : "1048588","mem_arena_held_delta" : "0","mem_bytes_requested_in_use" : "200716"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1928,"ts" :15568773,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1946,"ts" :15568762,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :13,"ts" :15576679,"ph" : "X","name" :"normalized_x_offset_0_QuantizeLinear_kernel_time","args" : {"activation_size" : "8192","mem_in_use_delta" : "0","mem_bytes_in_use" : "69644","mem_in_use_peak" : "200716","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"mem_requested_in_use_delta" : "0","output_size" : "4096","node_index" : "0","output_type_shape" : [{"uint16":[1,2048]}],"op_name" : "QuantizeLinear","mem_arena_held" : "1048588","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"parameter_size" : "6","mem_bytes_requested_in_use" : "69644","provider" : "CPUExecutionProvider","mem_arena_held_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1842,"ts" :15576696,"ph" : "X","name" :"vitisai_vitisai_9474036293895030537_0_0_kernel_time","args" : {"output_size" : "65536","provider" : "vitisai","mem_in_use_delta" : "0","activation_size" : "4096","mem_bytes_requested_in_use" : "69644","mem_arena_held_delta" : "0","mem_arena_held" : "1048588","input_type_shape" : [{"uint16":[1,2048]}],"mem_in_use_peak" : "200716","op_name" : "vitisai_9474036293895030537_0","parameter_size" : "0","node_index" : "6","output_type_shape" : [{"uint16":[1,32768]}],"mem_bytes_in_use" : "69644","mem_requested_in_use_delta" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}}}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :19,"ts" :15578543,"ph" : "X","name" :"logits_offset_0_DequantizeLinear_kernel_time","args" : {"activation_size" : "65536","mem_in_use_delta" : "131072","output_type_shape" : [{"float":[1,32768]}],"mem_bytes_requested_in_use" : "200716","mem_bytes_in_use" : "200716","mem_arena_held_delta" : "0","output_size" : "131072","parameter_size" : "6","mem_in_use_peak" : "200716","mem_requested_in_use_delta" : "131072","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"node_index" : "5","mem_arena_held" : "1048588","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"provider" : "CPUExecutionProvider","op_name" : "DequantizeLinear"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1886,"ts" :15576679,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1897,"ts" :15576674,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :16,"ts" :15584838,"ph" : "X","name" :"normalized_x_offset_0_QuantizeLinear_kernel_time","args" : {"mem_bytes_requested_in_use" : "69644","output_type_shape" : [{"uint16":[1,2048]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"node_index" : "0","op_name" : "QuantizeLinear","mem_in_use_delta" : "0","activation_size" : "8192","provider" : "CPUExecutionProvider","mem_bytes_in_use" : "69644","mem_requested_in_use_delta" : "0","mem_in_use_peak" : "200716","mem_arena_held" : "1048588","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"output_size" : "4096","mem_arena_held_delta" : "0","parameter_size" : "6"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1879,"ts" :15584858,"ph" : "X","name" :"vitisai_vitisai_9474036293895030537_0_0_kernel_time","args" : {"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"mem_in_use_delta" : "0","node_index" : "6","output_size" : "65536","provider" : "vitisai","mem_bytes_in_use" : "69644","mem_arena_held" : "1048588","mem_requested_in_use_delta" : "0","activation_size" : "4096","mem_in_use_peak" : "200716","input_type_shape" : [{"uint16":[1,2048]}],"parameter_size" : "0","output_type_shape" : [{"uint16":[1,32768]}],"mem_arena_held_delta" : "0","op_name" : "vitisai_9474036293895030537_0","mem_bytes_requested_in_use" : "69644"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :20,"ts" :15586751,"ph" : "X","name" :"logits_offset_0_DequantizeLinear_kernel_time","args" : {"op_name" : "DequantizeLinear","output_type_shape" : [{"float":[1,32768]}],"mem_in_use_delta" : "131072","provider" : "CPUExecutionProvider","output_size" : "131072","activation_size" : "65536","parameter_size" : "6","mem_arena_held" : "1048588","mem_requested_in_use_delta" : "131072","mem_arena_held_delta" : "0","mem_in_use_peak" : "200716","node_index" : "5","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"mem_bytes_in_use" : "200716","mem_bytes_requested_in_use" : "200716"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1938,"ts" :15584836,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1953,"ts" :15584826,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :14,"ts" :15593166,"ph" : "X","name" :"normalized_x_offset_0_QuantizeLinear_kernel_time","args" : {"node_index" : "0","parameter_size" : "6","mem_bytes_requested_in_use" : "69644","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_requested_in_use_delta" : "0","mem_bytes_in_use" : "69644","output_size" : "4096","provider" : "CPUExecutionProvider","mem_arena_held_delta" : "0","op_name" : "QuantizeLinear","output_type_shape" : [{"uint16":[1,2048]}],"mem_in_use_delta" : "0","activation_size" : "8192","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"mem_in_use_peak" : "200716","mem_arena_held" : "1048588"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1856,"ts" :15593183,"ph" : "X","name" :"vitisai_vitisai_9474036293895030537_0_0_kernel_time","args" : {"input_type_shape" : [{"uint16":[1,2048]}],"mem_in_use_peak" : "200716","mem_in_use_delta" : "0","parameter_size" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"output_size" : "65536","activation_size" : "4096","mem_bytes_requested_in_use" : "69644","mem_requested_in_use_delta" : "0","node_index" : "6","mem_bytes_in_use" : "69644","mem_arena_held" : "1048588","mem_arena_held_delta" : "0","output_type_shape" : [{"uint16":[1,32768]}],"op_name" : "vitisai_9474036293895030537_0","provider" : "vitisai"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :20,"ts" :15595046,"ph" : "X","name" :"logits_offset_0_DequantizeLinear_kernel_time","args" : {"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"mem_in_use_delta" : "131072","mem_bytes_requested_in_use" : "200716","mem_arena_held" : "1048588","mem_in_use_peak" : "200716","node_index" : "5","parameter_size" : "6","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"op_name" : "DequantizeLinear","output_size" : "131072","mem_requested_in_use_delta" : "131072","activation_size" : "65536","output_type_shape" : [{"float":[1,32768]}],"mem_bytes_in_use" : "200716","mem_arena_held_delta" : "0","provider" : "CPUExecutionProvider"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1903,"ts" :15593165,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1916,"ts" :15593160,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :17,"ts" :15601172,"ph" : "X","name" :"normalized_x_offset_0_QuantizeLinear_kernel_time","args" : {"node_index" : "0","output_size" : "4096","mem_in_use_delta" : "0","output_type_shape" : [{"uint16":[1,2048]}],"mem_bytes_requested_in_use" : "69644","op_name" : "QuantizeLinear","activation_size" : "8192","mem_requested_in_use_delta" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_arena_held_delta" : "0","parameter_size" : "6","provider" : "CPUExecutionProvider","mem_in_use_peak" : "200716","mem_bytes_in_use" : "69644","mem_arena_held" : "1048588"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1879,"ts" :15601194,"ph" : "X","name" :"vitisai_vitisai_9474036293895030537_0_0_kernel_time","args" : {"provider" : "vitisai","node_index" : "6","mem_requested_in_use_delta" : "0","activation_size" : "4096","mem_in_use_delta" : "0","output_size" : "65536","parameter_size" : "0","op_name" : "vitisai_9474036293895030537_0","mem_bytes_requested_in_use" : "69644","output_type_shape" : [{"uint16":[1,32768]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"mem_arena_held_delta" : "0","mem_in_use_peak" : "200716","mem_bytes_in_use" : "69644","input_type_shape" : [{"uint16":[1,2048]}],"mem_arena_held" : "1048588"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :20,"ts" :15603079,"ph" : "X","name" :"logits_offset_0_DequantizeLinear_kernel_time","args" : {"provider" : "CPUExecutionProvider","mem_bytes_requested_in_use" : "200716","parameter_size" : "6","mem_in_use_delta" : "131072","mem_requested_in_use_delta" : "131072","mem_arena_held" : "1048588","activation_size" : "65536","node_index" : "5","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"mem_arena_held_delta" : "0","output_size" : "131072","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"output_type_shape" : [{"float":[1,32768]}],"mem_in_use_peak" : "200716","op_name" : "DequantizeLinear","mem_bytes_in_use" : "200716"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1931,"ts" :15601171,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1947,"ts" :15601162,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :16,"ts" :15609229,"ph" : "X","name" :"normalized_x_offset_0_QuantizeLinear_kernel_time","args" : {"provider" : "CPUExecutionProvider","op_name" : "QuantizeLinear","mem_in_use_delta" : "0","mem_arena_held" : "1048588","output_type_shape" : [{"uint16":[1,2048]}],"mem_bytes_requested_in_use" : "69644","parameter_size" : "6","mem_requested_in_use_delta" : "0","mem_in_use_peak" : "200716","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_bytes_in_use" : "69644","activation_size" : "8192","output_size" : "4096","mem_arena_held_delta" : "0","node_index" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}}}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1881,"ts" :15609248,"ph" : "X","name" :"vitisai_vitisai_9474036293895030537_0_0_kernel_time","args" : {"output_size" : "65536","output_type_shape" : [{"uint16":[1,32768]}],"mem_requested_in_use_delta" : "0","input_type_shape" : [{"uint16":[1,2048]}],"mem_arena_held_delta" : "0","activation_size" : "4096","mem_bytes_requested_in_use" : "69644","node_index" : "6","mem_bytes_in_use" : "69644","op_name" : "vitisai_9474036293895030537_0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"provider" : "vitisai","parameter_size" : "0","mem_in_use_peak" : "200716","mem_in_use_delta" : "0","mem_arena_held" : "1048588"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :16,"ts" :15611132,"ph" : "X","name" :"logits_offset_0_DequantizeLinear_kernel_time","args" : {"mem_arena_held_delta" : "0","node_index" : "5","op_name" : "DequantizeLinear","provider" : "CPUExecutionProvider","parameter_size" : "6","activation_size" : "65536","mem_bytes_in_use" : "200716","output_type_shape" : [{"float":[1,32768]}],"mem_arena_held" : "1048588","mem_bytes_requested_in_use" : "200716","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"output_size" : "131072","mem_requested_in_use_delta" : "131072","mem_in_use_delta" : "131072","mem_in_use_peak" : "200716","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}}}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1923,"ts" :15609228,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1935,"ts" :15609219,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :16,"ts" :15617671,"ph" : "X","name" :"normalized_x_offset_0_QuantizeLinear_kernel_time","args" : {"mem_bytes_requested_in_use" : "69644","op_name" : "QuantizeLinear","mem_arena_held_delta" : "0","provider" : "CPUExecutionProvider","parameter_size" : "6","mem_arena_held" : "1048588","mem_in_use_peak" : "200716","mem_requested_in_use_delta" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"mem_bytes_in_use" : "69644","output_size" : "4096","activation_size" : "8192","mem_in_use_delta" : "0","output_type_shape" : [{"uint16":[1,2048]}],"input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"node_index" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1979,"ts" :15617690,"ph" : "X","name" :"vitisai_vitisai_9474036293895030537_0_0_kernel_time","args" : {"mem_bytes_requested_in_use" : "69644","mem_arena_held" : "1048588","provider" : "vitisai","mem_in_use_delta" : "0","mem_arena_held_delta" : "0","mem_requested_in_use_delta" : "0","mem_in_use_peak" : "200716","node_index" : "6","output_size" : "65536","input_type_shape" : [{"uint16":[1,2048]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"op_name" : "vitisai_9474036293895030537_0","activation_size" : "4096","parameter_size" : "0","output_type_shape" : [{"uint16":[1,32768]}],"mem_bytes_in_use" : "69644"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :60,"ts" :15619687,"ph" : "X","name" :"logits_offset_0_DequantizeLinear_kernel_time","args" : {"output_type_shape" : [{"float":[1,32768]}],"mem_in_use_delta" : "131072","parameter_size" : "6","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"node_index" : "5","mem_arena_held_delta" : "0","mem_bytes_in_use" : "200716","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_requested_in_use_delta" : "131072","activation_size" : "65536","mem_in_use_peak" : "200716","mem_arena_held" : "1048588","provider" : "CPUExecutionProvider","op_name" : "DequantizeLinear","mem_bytes_requested_in_use" : "200716","output_size" : "131072"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2086,"ts" :15617669,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2120,"ts" :15617660,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :33,"ts" :15626498,"ph" : "X","name" :"normalized_x_offset_0_QuantizeLinear_kernel_time","args" : {"parameter_size" : "6","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"activation_size" : "8192","op_name" : "QuantizeLinear","provider" : "CPUExecutionProvider","output_type_shape" : [{"uint16":[1,2048]}],"mem_in_use_peak" : "200716","mem_bytes_requested_in_use" : "69644","output_size" : "4096","mem_arena_held_delta" : "0","mem_arena_held" : "1048588","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"node_index" : "0","mem_in_use_delta" : "0","mem_requested_in_use_delta" : "0","mem_bytes_in_use" : "69644"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1943,"ts" :15626537,"ph" : "X","name" :"vitisai_vitisai_9474036293895030537_0_0_kernel_time","args" : {"mem_arena_held_delta" : "0","node_index" : "6","mem_in_use_delta" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"mem_arena_held" : "1048588","mem_in_use_peak" : "200716","mem_requested_in_use_delta" : "0","provider" : "vitisai","parameter_size" : "0","op_name" : "vitisai_9474036293895030537_0","mem_bytes_in_use" : "69644","output_type_shape" : [{"uint16":[1,32768]}],"input_type_shape" : [{"uint16":[1,2048]}],"mem_bytes_requested_in_use" : "69644","activation_size" : "4096","output_size" : "65536"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :40,"ts" :15628490,"ph" : "X","name" :"logits_offset_0_DequantizeLinear_kernel_time","args" : {"mem_arena_held" : "1048588","output_size" : "131072","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"activation_size" : "65536","mem_requested_in_use_delta" : "131072","op_name" : "DequantizeLinear","mem_arena_held_delta" : "0","mem_in_use_peak" : "200716","mem_in_use_delta" : "131072","mem_bytes_in_use" : "200716","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"output_type_shape" : [{"float":[1,32768]}],"node_index" : "5","provider" : "CPUExecutionProvider","mem_bytes_requested_in_use" : "200716","parameter_size" : "6"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2039,"ts" :15626496,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2062,"ts" :15626484,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :19,"ts" :15635568,"ph" : "X","name" :"normalized_x_offset_0_QuantizeLinear_kernel_time","args" : {"parameter_size" : "6","output_type_shape" : [{"uint16":[1,2048]}],"mem_bytes_requested_in_use" : "69644","node_index" : "0","provider" : "CPUExecutionProvider","op_name" : "QuantizeLinear","mem_arena_held" : "1048588","mem_requested_in_use_delta" : "0","mem_in_use_peak" : "200716","mem_arena_held_delta" : "0","mem_in_use_delta" : "0","activation_size" : "8192","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"output_size" : "4096","mem_bytes_in_use" : "69644"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1963,"ts" :15635591,"ph" : "X","name" :"vitisai_vitisai_9474036293895030537_0_0_kernel_time","args" : {"mem_requested_in_use_delta" : "0","output_type_shape" : [{"uint16":[1,32768]}],"mem_in_use_peak" : "200716","input_type_shape" : [{"uint16":[1,2048]}],"mem_arena_held" : "1048588","node_index" : "6","output_size" : "65536","provider" : "vitisai","activation_size" : "4096","op_name" : "vitisai_9474036293895030537_0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"mem_arena_held_delta" : "0","parameter_size" : "0","mem_bytes_requested_in_use" : "69644","mem_bytes_in_use" : "69644","mem_in_use_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :45,"ts" :15637567,"ph" : "X","name" :"logits_offset_0_DequantizeLinear_kernel_time","args" : {"mem_requested_in_use_delta" : "131072","mem_arena_held" : "1048588","node_index" : "5","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"output_type_shape" : [{"float":[1,32768]}],"mem_in_use_delta" : "131072","mem_in_use_peak" : "200716","op_name" : "DequantizeLinear","mem_bytes_in_use" : "200716","output_size" : "131072","activation_size" : "65536","provider" : "CPUExecutionProvider","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"parameter_size" : "6","mem_arena_held_delta" : "0","mem_bytes_requested_in_use" : "200716"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2060,"ts" :15635566,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2085,"ts" :15635556,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :18,"ts" :15644031,"ph" : "X","name" :"normalized_x_offset_0_QuantizeLinear_kernel_time","args" : {"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"mem_in_use_delta" : "0","op_name" : "QuantizeLinear","mem_bytes_requested_in_use" : "69644","output_type_shape" : [{"uint16":[1,2048]}],"mem_bytes_in_use" : "69644","mem_requested_in_use_delta" : "0","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"provider" : "CPUExecutionProvider","output_size" : "4096","mem_arena_held_delta" : "0","parameter_size" : "6","activation_size" : "8192","node_index" : "0","mem_in_use_peak" : "200716","mem_arena_held" : "1048588"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1924,"ts" :15644053,"ph" : "X","name" :"vitisai_vitisai_9474036293895030537_0_0_kernel_time","args" : {"mem_requested_in_use_delta" : "0","mem_arena_held_delta" : "0","provider" : "vitisai","input_type_shape" : [{"uint16":[1,2048]}],"mem_arena_held" : "1048588","output_type_shape" : [{"uint16":[1,32768]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"op_name" : "vitisai_9474036293895030537_0","mem_bytes_in_use" : "69644","activation_size" : "4096","mem_in_use_peak" : "200716","node_index" : "6","mem_bytes_requested_in_use" : "69644","parameter_size" : "0","mem_in_use_delta" : "0","output_size" : "65536"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :40,"ts" :15645987,"ph" : "X","name" :"logits_offset_0_DequantizeLinear_kernel_time","args" : {"mem_in_use_delta" : "131072","mem_arena_held" : "1048588","output_size" : "131072","op_name" : "DequantizeLinear","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"mem_requested_in_use_delta" : "131072","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"output_type_shape" : [{"float":[1,32768]}],"mem_arena_held_delta" : "0","provider" : "CPUExecutionProvider","activation_size" : "65536","mem_bytes_requested_in_use" : "200716","mem_in_use_peak" : "200716","parameter_size" : "6","mem_bytes_in_use" : "200716","node_index" : "5"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2002,"ts" :15644030,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2021,"ts" :15644021,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :16,"ts" :15652295,"ph" : "X","name" :"normalized_x_offset_0_QuantizeLinear_kernel_time","args" : {"output_size" : "4096","parameter_size" : "6","node_index" : "0","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"op_name" : "QuantizeLinear","mem_in_use_delta" : "0","mem_bytes_in_use" : "69644","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"mem_in_use_peak" : "200716","mem_requested_in_use_delta" : "0","mem_arena_held_delta" : "0","provider" : "CPUExecutionProvider","mem_bytes_requested_in_use" : "69644","output_type_shape" : [{"uint16":[1,2048]}],"activation_size" : "8192","mem_arena_held" : "1048588"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :2025,"ts" :15652316,"ph" : "X","name" :"vitisai_vitisai_9474036293895030537_0_0_kernel_time","args" : {"mem_arena_held_delta" : "0","provider" : "vitisai","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"mem_in_use_peak" : "200716","mem_bytes_requested_in_use" : "69644","node_index" : "6","activation_size" : "4096","output_size" : "65536","mem_requested_in_use_delta" : "0","parameter_size" : "0","op_name" : "vitisai_9474036293895030537_0","input_type_shape" : [{"uint16":[1,2048]}],"mem_arena_held" : "1048588","mem_in_use_delta" : "0","output_type_shape" : [{"uint16":[1,32768]}],"mem_bytes_in_use" : "69644"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :65,"ts" :15654360,"ph" : "X","name" :"logits_offset_0_DequantizeLinear_kernel_time","args" : {"node_index" : "5","provider" : "CPUExecutionProvider","op_name" : "DequantizeLinear","mem_in_use_peak" : "200716","activation_size" : "65536","mem_arena_held" : "1048588","output_size" : "131072","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"mem_arena_held_delta" : "0","mem_bytes_in_use" : "200716","output_type_shape" : [{"float":[1,32768]}],"mem_bytes_requested_in_use" : "200716","mem_in_use_delta" : "131072","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_requested_in_use_delta" : "131072","parameter_size" : "6"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2138,"ts" :15652294,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2164,"ts" :15652285,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :15,"ts" :15660578,"ph" : "X","name" :"normalized_x_offset_0_QuantizeLinear_kernel_time","args" : {"mem_bytes_in_use" : "69644","mem_arena_held" : "1048588","mem_requested_in_use_delta" : "0","provider" : "CPUExecutionProvider","output_size" : "4096","mem_arena_held_delta" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"op_name" : "QuantizeLinear","mem_in_use_delta" : "0","node_index" : "0","activation_size" : "8192","mem_bytes_requested_in_use" : "69644","output_type_shape" : [{"uint16":[1,2048]}],"parameter_size" : "6","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_in_use_peak" : "200716"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1912,"ts" :15660597,"ph" : "X","name" :"vitisai_vitisai_9474036293895030537_0_0_kernel_time","args" : {"op_name" : "vitisai_9474036293895030537_0","input_type_shape" : [{"uint16":[1,2048]}],"parameter_size" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"mem_arena_held_delta" : "0","mem_in_use_peak" : "200716","mem_in_use_delta" : "0","provider" : "vitisai","node_index" : "6","mem_bytes_in_use" : "69644","output_size" : "65536","output_type_shape" : [{"uint16":[1,32768]}],"mem_arena_held" : "1048588","activation_size" : "4096","mem_bytes_requested_in_use" : "69644","mem_requested_in_use_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :40,"ts" :15662517,"ph" : "X","name" :"logits_offset_0_DequantizeLinear_kernel_time","args" : {"mem_requested_in_use_delta" : "131072","output_size" : "131072","op_name" : "DequantizeLinear","output_type_shape" : [{"float":[1,32768]}],"node_index" : "5","activation_size" : "65536","mem_in_use_peak" : "200716","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"mem_bytes_in_use" : "200716","mem_arena_held" : "1048588","provider" : "CPUExecutionProvider","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"parameter_size" : "6","mem_bytes_requested_in_use" : "200716","mem_in_use_delta" : "131072","mem_arena_held_delta" : "0"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1985,"ts" :15660577,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2000,"ts" :15660571,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :42,"ts" :15669234,"ph" : "X","name" :"normalized_x_offset_0_QuantizeLinear_kernel_time","args" : {"node_index" : "0","output_size" : "4096","mem_requested_in_use_delta" : "0","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"op_name" : "QuantizeLinear","output_type_shape" : [{"uint16":[1,2048]}],"mem_bytes_in_use" : "69644","provider" : "CPUExecutionProvider","mem_in_use_peak" : "200716","mem_arena_held" : "1048588","mem_arena_held_delta" : "0","mem_bytes_requested_in_use" : "69644","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"parameter_size" : "6","mem_in_use_delta" : "0","activation_size" : "8192"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :2001,"ts" :15669283,"ph" : "X","name" :"vitisai_vitisai_9474036293895030537_0_0_kernel_time","args" : {"mem_requested_in_use_delta" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"node_index" : "6","parameter_size" : "0","mem_arena_held_delta" : "0","output_type_shape" : [{"uint16":[1,32768]}],"output_size" : "65536","mem_in_use_delta" : "0","mem_bytes_requested_in_use" : "69644","activation_size" : "4096","provider" : "vitisai","op_name" : "vitisai_9474036293895030537_0","input_type_shape" : [{"uint16":[1,2048]}],"mem_in_use_peak" : "200716","mem_bytes_in_use" : "69644","mem_arena_held" : "1048588"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :68,"ts" :15671298,"ph" : "X","name" :"logits_offset_0_DequantizeLinear_kernel_time","args" : {"mem_bytes_requested_in_use" : "200716","activation_size" : "65536","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"provider" : "CPUExecutionProvider","mem_arena_held" : "1048588","mem_arena_held_delta" : "0","node_index" : "5","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"mem_in_use_peak" : "200716","output_type_shape" : [{"float":[1,32768]}],"output_size" : "131072","parameter_size" : "6","mem_requested_in_use_delta" : "131072","op_name" : "DequantizeLinear","mem_bytes_in_use" : "200716","mem_in_use_delta" : "131072"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2140,"ts" :15669232,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2168,"ts" :15669222,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :16,"ts" :15677972,"ph" : "X","name" :"normalized_x_offset_0_QuantizeLinear_kernel_time","args" : {"node_index" : "0","activation_size" : "8192","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_requested_in_use_delta" : "0","parameter_size" : "6","mem_arena_held" : "1048588","mem_in_use_delta" : "0","output_type_shape" : [{"uint16":[1,2048]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"mem_bytes_in_use" : "69644","op_name" : "QuantizeLinear","mem_in_use_peak" : "200716","mem_arena_held_delta" : "0","mem_bytes_requested_in_use" : "69644","output_size" : "4096","provider" : "CPUExecutionProvider"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1849,"ts" :15677991,"ph" : "X","name" :"vitisai_vitisai_9474036293895030537_0_0_kernel_time","args" : {"mem_arena_held" : "1048588","output_size" : "65536","mem_requested_in_use_delta" : "0","mem_in_use_delta" : "0","output_type_shape" : [{"uint16":[1,32768]}],"mem_bytes_in_use" : "69644","activation_size" : "4096","mem_bytes_requested_in_use" : "69644","parameter_size" : "0","op_name" : "vitisai_9474036293895030537_0","input_type_shape" : [{"uint16":[1,2048]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"provider" : "vitisai","mem_in_use_peak" : "200716","node_index" : "6","mem_arena_held_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :18,"ts" :15679844,"ph" : "X","name" :"logits_offset_0_DequantizeLinear_kernel_time","args" : {"mem_bytes_requested_in_use" : "200716","parameter_size" : "6","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_bytes_in_use" : "200716","mem_arena_held_delta" : "0","mem_requested_in_use_delta" : "131072","mem_in_use_peak" : "200716","op_name" : "DequantizeLinear","provider" : "CPUExecutionProvider","activation_size" : "65536","node_index" : "5","mem_in_use_delta" : "131072","output_size" : "131072","mem_arena_held" : "1048588","output_type_shape" : [{"float":[1,32768]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}}}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1893,"ts" :15677971,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1905,"ts" :15677963,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :16,"ts" :15686301,"ph" : "X","name" :"normalized_x_offset_0_QuantizeLinear_kernel_time","args" : {"mem_in_use_peak" : "200716","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"provider" : "CPUExecutionProvider","mem_bytes_in_use" : "69644","mem_requested_in_use_delta" : "0","parameter_size" : "6","mem_arena_held_delta" : "0","output_size" : "4096","output_type_shape" : [{"uint16":[1,2048]}],"node_index" : "0","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_arena_held" : "1048588","mem_in_use_delta" : "0","activation_size" : "8192","op_name" : "QuantizeLinear","mem_bytes_requested_in_use" : "69644"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1929,"ts" :15686322,"ph" : "X","name" :"vitisai_vitisai_9474036293895030537_0_0_kernel_time","args" : {"output_size" : "65536","mem_bytes_requested_in_use" : "69644","output_type_shape" : [{"uint16":[1,32768]}],"mem_requested_in_use_delta" : "0","input_type_shape" : [{"uint16":[1,2048]}],"mem_arena_held" : "1048588","node_index" : "6","provider" : "vitisai","mem_arena_held_delta" : "0","mem_bytes_in_use" : "69644","mem_in_use_delta" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"op_name" : "vitisai_9474036293895030537_0","activation_size" : "4096","parameter_size" : "0","mem_in_use_peak" : "200716"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :17,"ts" :15688255,"ph" : "X","name" :"logits_offset_0_DequantizeLinear_kernel_time","args" : {"mem_arena_held" : "1048588","parameter_size" : "6","mem_bytes_in_use" : "200716","mem_in_use_peak" : "200716","output_size" : "131072","node_index" : "5","output_type_shape" : [{"float":[1,32768]}],"input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_in_use_delta" : "131072","provider" : "CPUExecutionProvider","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-2-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9924": {"num_run": 0, "core": -1},"58848": {"num_run": 0, "core": -1},"64724": {"num_run": 0, "core": -1},"46576": {"num_run": 0, "core": -1},"39748": {"num_run": 0, "core": -1},"62520": {"num_run": 0, "core": -1},"55784": {"num_run": 0, "core": -1},"30236": {"num_run": 0, "core": -1},"17020": {"num_run": 0, "core": -1},"33644": {"num_run": 0, "core": -1},"56512": {"num_run": 0, "core": -1}}},"activation_size" : "65536","mem_arena_held_delta" : "0","mem_requested_in_use_delta" : "131072","mem_bytes_requested_in_use" : "200716","op_name" : "DequantizeLinear"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1975,"ts" :15686300,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1987,"ts" :15686292,"ph" : "X","name" :"model_run","args" : {}} +] diff --git a/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/shards/shard_32768.json b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/shards/shard_32768.json new file mode 100644 index 000000000..f5d4cc230 --- /dev/null +++ b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/shards/shard_32768.json @@ -0,0 +1,95 @@ +{ + "scope": "Spark-X2.5-1.7B output head only; no prefill/decode/KV/sampling or full generation", + "vocab_shard_size": 32768, + "vocab_shard_offset": 32768, + "flatten_input": true, + "batch_repeats": 1, + "matmul_only": true, + "hardware": "AMD Ryzen AI 9 HX 370 NPU", + "os": "Windows-11-10.0.26200-SP0", + "onnx_version": "1.22.0", + "onnxruntime_version": "1.30.0", + "source_model_sha256": "f5f0c31eb438e9940ab242da8bf6f0f5dad6ebc9f6ef02d5475c21b744e595bc", + "prepared_model_sha256": "dd29ea9d1ebd06943b5c64a8835fe2340fb47543d20060d3afbafa17377f0db9", + "calibration_sha256": "cfa2ed8439f8fe01730319667bfd01abf2e39cf5851490346a0331304cf63194", + "opset21_sha256": "dd29ea9d1ebd06943b5c64a8835fe2340fb47543d20060d3afbafa17377f0db9", + "quantized_sha256": "f6a4257471869f02d4ff4ec405eeb16f46d6b50c7e02b2b6e7d08e71ac4cc0c9", + "quantization": "opset 21, MatMul only, QDQ, QUInt16 activations/QInt8 weights, MinMax on four real activations", + "input_shape": [ + 1, + 2048 + ], + "output_shape": [ + 1, + 32768 + ], + "source_cpu_reference": [ + { + "snr_db": 137.33794535649483, + "max_abs_diff": 1.9073486328125e-05, + "top1_match": true, + "finite": true + }, + { + "snr_db": 134.39788216827213, + "max_abs_diff": 2.288818359375e-05, + "top1_match": true, + "finite": true + }, + { + "snr_db": 137.78765215287703, + "max_abs_diff": 1.9073486328125e-05, + "top1_match": true, + "finite": true + }, + { + "snr_db": 136.53269799911513, + "max_abs_diff": 1.9073486328125e-05, + "top1_match": true, + "finite": true + } + ], + "npu_devices": 1, + "session_providers": [ + "vitisai", + "CPUExecutionProvider" + ], + "npu_node_events": 4, + "cpu_node_events": 8, + "load_s": 3.653468700009398, + "npu_rows": [ + { + "snr_db": 58.25548223459347, + "max_abs_diff": 0.21026992797851562, + "top1_match": true, + "finite": true, + "index": 0, + "wall_s": 0.004064699984155595 + }, + { + "snr_db": 57.80354617171815, + "max_abs_diff": 0.18883514404296875, + "top1_match": true, + "finite": true, + "index": 1, + "wall_s": 0.0026345999212935567 + }, + { + "snr_db": 57.78266475539301, + "max_abs_diff": 0.18820953369140625, + "top1_match": true, + "finite": true, + "index": 2, + "wall_s": 0.002468000049702823 + }, + { + "snr_db": 56.79216419378707, + "max_abs_diff": 0.18805313110351562, + "top1_match": true, + "finite": true, + "index": 3, + "wall_s": 0.002669299952685833 + } + ], + "profile_file": "C:\\Users\\zhout\\w2\\spark_npu_full_shards_named_20260923\\shard_32768\\spark_lm_head_npu_profile_2026-09-23_18-07-38_848.json" +} diff --git a/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/shards/shard_32768_co_resident_profile.json b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/shards/shard_32768_co_resident_profile.json new file mode 100644 index 000000000..a21f38c25 --- /dev/null +++ b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/shards/shard_32768_co_resident_profile.json @@ -0,0 +1,129 @@ +[ +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :32295,"ts" :27,"ph" : "X","name" :"model_loading_uri","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :3750578,"ts" :33566,"ph" : "X","name" :"session_initialization","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :31,"ts" :11637734,"ph" : "X","name" :"normalized_x_offset_32768_QuantizeLinear_kernel_time","args" : {"mem_requested_in_use_delta" : "4096","mem_in_use_delta" : "4096","mem_in_use_peak" : "4108","mem_arena_held" : "1048588","output_size" : "4096","provider" : "CPUExecutionProvider","output_type_shape" : [{"uint16":[1,2048]}],"mem_arena_held_delta" : "1048576","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"op_name" : "QuantizeLinear","mem_bytes_in_use" : "4108","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_bytes_requested_in_use" : "4108","node_index" : "0","activation_size" : "8192","parameter_size" : "6"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :2080,"ts" :11637769,"ph" : "X","name" :"vitisai_vitisai_11298643349341028369_0_0_kernel_time","args" : {"mem_in_use_delta" : "65536","input_type_shape" : [{"uint16":[1,2048]}],"output_type_shape" : [{"uint16":[1,32768]}],"activation_size" : "4096","op_name" : "vitisai_11298643349341028369_0","mem_arena_held" : "1048588","parameter_size" : "0","mem_arena_held_delta" : "0","mem_bytes_requested_in_use" : "69644","output_size" : "65536","mem_requested_in_use_delta" : "65536","node_index" : "6","provider" : "vitisai","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"mem_in_use_peak" : "69644","mem_bytes_in_use" : "69644"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :31,"ts" :11639854,"ph" : "X","name" :"logits_offset_32768_DequantizeLinear_kernel_time","args" : {"mem_arena_held" : "1048588","provider" : "CPUExecutionProvider","node_index" : "5","mem_arena_held_delta" : "0","mem_in_use_peak" : "196620","mem_bytes_in_use" : "196620","mem_bytes_requested_in_use" : "196620","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"output_size" : "131072","output_type_shape" : [{"float":[1,32768]}],"mem_in_use_delta" : "131072","parameter_size" : "6","op_name" : "DequantizeLinear","mem_requested_in_use_delta" : "131072","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"activation_size" : "65536"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2156,"ts" :11637733,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2174,"ts" :11637721,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :36,"ts" :11650201,"ph" : "X","name" :"normalized_x_offset_32768_QuantizeLinear_kernel_time","args" : {"node_index" : "0","parameter_size" : "6","provider" : "CPUExecutionProvider","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"op_name" : "QuantizeLinear","mem_in_use_peak" : "196620","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_in_use_delta" : "0","activation_size" : "8192","mem_requested_in_use_delta" : "0","mem_arena_held" : "1048588","mem_bytes_requested_in_use" : "69644","output_size" : "4096","mem_arena_held_delta" : "0","output_type_shape" : [{"uint16":[1,2048]}],"mem_bytes_in_use" : "69644"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1890,"ts" :11650244,"ph" : "X","name" :"vitisai_vitisai_11298643349341028369_0_0_kernel_time","args" : {"node_index" : "6","mem_bytes_in_use" : "69644","output_type_shape" : [{"uint16":[1,32768]}],"mem_requested_in_use_delta" : "0","mem_bytes_requested_in_use" : "69644","activation_size" : "4096","mem_arena_held" : "1048588","mem_in_use_delta" : "0","mem_in_use_peak" : "196620","provider" : "vitisai","mem_arena_held_delta" : "0","parameter_size" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"op_name" : "vitisai_11298643349341028369_0","input_type_shape" : [{"uint16":[1,2048]}],"output_size" : "65536"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :18,"ts" :11652138,"ph" : "X","name" :"logits_offset_32768_DequantizeLinear_kernel_time","args" : {"output_size" : "131072","mem_arena_held" : "1048588","mem_in_use_peak" : "200716","node_index" : "5","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"mem_requested_in_use_delta" : "131072","mem_bytes_requested_in_use" : "200716","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"output_type_shape" : [{"float":[1,32768]}],"mem_in_use_delta" : "131072","parameter_size" : "6","provider" : "CPUExecutionProvider","activation_size" : "65536","mem_arena_held_delta" : "0","mem_bytes_in_use" : "200716","op_name" : "DequantizeLinear"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1960,"ts" :11650199,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1979,"ts" :11650184,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :18,"ts" :11662575,"ph" : "X","name" :"normalized_x_offset_32768_QuantizeLinear_kernel_time","args" : {"activation_size" : "8192","mem_in_use_peak" : "200716","parameter_size" : "6","op_name" : "QuantizeLinear","mem_arena_held_delta" : "0","mem_arena_held" : "1048588","provider" : "CPUExecutionProvider","mem_bytes_requested_in_use" : "69644","node_index" : "0","mem_bytes_in_use" : "69644","output_size" : "4096","mem_requested_in_use_delta" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"output_type_shape" : [{"uint16":[1,2048]}],"mem_in_use_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1867,"ts" :11662597,"ph" : "X","name" :"vitisai_vitisai_11298643349341028369_0_0_kernel_time","args" : {"input_type_shape" : [{"uint16":[1,2048]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"mem_in_use_delta" : "0","node_index" : "6","output_type_shape" : [{"uint16":[1,32768]}],"mem_requested_in_use_delta" : "0","mem_arena_held" : "1048588","mem_in_use_peak" : "200716","output_size" : "65536","provider" : "vitisai","op_name" : "vitisai_11298643349341028369_0","mem_bytes_requested_in_use" : "69644","mem_arena_held_delta" : "0","parameter_size" : "0","mem_bytes_in_use" : "69644","activation_size" : "4096"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :19,"ts" :11664470,"ph" : "X","name" :"logits_offset_32768_DequantizeLinear_kernel_time","args" : {"mem_bytes_requested_in_use" : "200716","mem_arena_held" : "1048588","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"activation_size" : "65536","parameter_size" : "6","mem_in_use_peak" : "200716","mem_arena_held_delta" : "0","output_size" : "131072","output_type_shape" : [{"float":[1,32768]}],"provider" : "CPUExecutionProvider","op_name" : "DequantizeLinear","mem_requested_in_use_delta" : "131072","mem_in_use_delta" : "131072","node_index" : "5","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_bytes_in_use" : "200716"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1918,"ts" :11662574,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1935,"ts" :11662560,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :15,"ts" :11673484,"ph" : "X","name" :"normalized_x_offset_32768_QuantizeLinear_kernel_time","args" : {"mem_arena_held_delta" : "0","parameter_size" : "6","mem_in_use_peak" : "200716","output_size" : "4096","output_type_shape" : [{"uint16":[1,2048]}],"op_name" : "QuantizeLinear","mem_bytes_requested_in_use" : "69644","activation_size" : "8192","provider" : "CPUExecutionProvider","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"mem_arena_held" : "1048588","node_index" : "0","mem_requested_in_use_delta" : "0","mem_in_use_delta" : "0","mem_bytes_in_use" : "69644"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1876,"ts" :11673502,"ph" : "X","name" :"vitisai_vitisai_11298643349341028369_0_0_kernel_time","args" : {"provider" : "vitisai","mem_bytes_in_use" : "69644","mem_in_use_delta" : "0","output_size" : "65536","mem_in_use_peak" : "200716","op_name" : "vitisai_11298643349341028369_0","mem_arena_held" : "1048588","output_type_shape" : [{"uint16":[1,32768]}],"activation_size" : "4096","node_index" : "6","input_type_shape" : [{"uint16":[1,2048]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"parameter_size" : "0","mem_arena_held_delta" : "0","mem_bytes_requested_in_use" : "69644","mem_requested_in_use_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :19,"ts" :11675384,"ph" : "X","name" :"logits_offset_32768_DequantizeLinear_kernel_time","args" : {"output_size" : "131072","activation_size" : "65536","mem_in_use_delta" : "131072","mem_in_use_peak" : "200716","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"mem_bytes_requested_in_use" : "200716","output_type_shape" : [{"float":[1,32768]}],"parameter_size" : "6","mem_arena_held" : "1048588","mem_arena_held_delta" : "0","node_index" : "5","mem_requested_in_use_delta" : "131072","mem_bytes_in_use" : "200716","op_name" : "DequantizeLinear","provider" : "CPUExecutionProvider"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1925,"ts" :11673483,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1940,"ts" :11673474,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :38,"ts" :11684509,"ph" : "X","name" :"normalized_x_offset_32768_QuantizeLinear_kernel_time","args" : {"mem_requested_in_use_delta" : "0","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_arena_held" : "1048588","node_index" : "0","mem_bytes_in_use" : "69644","provider" : "CPUExecutionProvider","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"output_size" : "4096","activation_size" : "8192","op_name" : "QuantizeLinear","output_type_shape" : [{"uint16":[1,2048]}],"mem_bytes_requested_in_use" : "69644","mem_in_use_delta" : "0","mem_arena_held_delta" : "0","mem_in_use_peak" : "200716","parameter_size" : "6"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1924,"ts" :11684555,"ph" : "X","name" :"vitisai_vitisai_11298643349341028369_0_0_kernel_time","args" : {"mem_in_use_peak" : "200716","mem_bytes_requested_in_use" : "69644","mem_bytes_in_use" : "69644","output_size" : "65536","op_name" : "vitisai_11298643349341028369_0","provider" : "vitisai","output_type_shape" : [{"uint16":[1,32768]}],"mem_requested_in_use_delta" : "0","input_type_shape" : [{"uint16":[1,2048]}],"node_index" : "6","parameter_size" : "0","mem_arena_held_delta" : "0","mem_arena_held" : "1048588","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"mem_in_use_delta" : "0","activation_size" : "4096"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :20,"ts" :11686485,"ph" : "X","name" :"logits_offset_32768_DequantizeLinear_kernel_time","args" : {"output_type_shape" : [{"float":[1,32768]}],"mem_in_use_delta" : "131072","mem_in_use_peak" : "200716","output_size" : "131072","mem_arena_held" : "1048588","parameter_size" : "6","mem_arena_held_delta" : "0","mem_requested_in_use_delta" : "131072","activation_size" : "65536","node_index" : "5","op_name" : "DequantizeLinear","mem_bytes_in_use" : "200716","provider" : "CPUExecutionProvider","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"mem_bytes_requested_in_use" : "200716","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}]}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2000,"ts" :11684507,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2026,"ts" :11684488,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :15,"ts" :11692502,"ph" : "X","name" :"normalized_x_offset_32768_QuantizeLinear_kernel_time","args" : {"mem_arena_held" : "1048588","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"output_type_shape" : [{"uint16":[1,2048]}],"output_size" : "4096","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"provider" : "CPUExecutionProvider","mem_bytes_requested_in_use" : "69644","mem_requested_in_use_delta" : "0","mem_arena_held_delta" : "0","mem_in_use_peak" : "200716","node_index" : "0","mem_bytes_in_use" : "69644","op_name" : "QuantizeLinear","parameter_size" : "6","activation_size" : "8192","mem_in_use_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1891,"ts" :11692519,"ph" : "X","name" :"vitisai_vitisai_11298643349341028369_0_0_kernel_time","args" : {"mem_in_use_peak" : "200716","mem_in_use_delta" : "0","mem_arena_held_delta" : "0","mem_bytes_requested_in_use" : "69644","mem_requested_in_use_delta" : "0","output_type_shape" : [{"uint16":[1,32768]}],"output_size" : "65536","mem_bytes_in_use" : "69644","op_name" : "vitisai_11298643349341028369_0","input_type_shape" : [{"uint16":[1,2048]}],"activation_size" : "4096","provider" : "vitisai","parameter_size" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"node_index" : "6","mem_arena_held" : "1048588"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :22,"ts" :11694420,"ph" : "X","name" :"logits_offset_32768_DequantizeLinear_kernel_time","args" : {"mem_in_use_peak" : "200716","op_name" : "DequantizeLinear","node_index" : "5","mem_in_use_delta" : "131072","provider" : "CPUExecutionProvider","mem_bytes_requested_in_use" : "200716","parameter_size" : "6","mem_arena_held" : "1048588","output_type_shape" : [{"float":[1,32768]}],"mem_arena_held_delta" : "0","activation_size" : "65536","output_size" : "131072","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"mem_requested_in_use_delta" : "131072","mem_bytes_in_use" : "200716","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}]}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1944,"ts" :11692501,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1959,"ts" :11692494,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :13,"ts" :11700707,"ph" : "X","name" :"normalized_x_offset_32768_QuantizeLinear_kernel_time","args" : {"node_index" : "0","mem_bytes_requested_in_use" : "69644","mem_requested_in_use_delta" : "0","mem_in_use_delta" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"output_type_shape" : [{"uint16":[1,2048]}],"mem_in_use_peak" : "200716","output_size" : "4096","mem_arena_held" : "1048588","op_name" : "QuantizeLinear","activation_size" : "8192","mem_arena_held_delta" : "0","parameter_size" : "6","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"provider" : "CPUExecutionProvider","mem_bytes_in_use" : "69644"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1860,"ts" :11700724,"ph" : "X","name" :"vitisai_vitisai_11298643349341028369_0_0_kernel_time","args" : {"mem_arena_held" : "1048588","node_index" : "6","output_type_shape" : [{"uint16":[1,32768]}],"mem_bytes_in_use" : "69644","provider" : "vitisai","op_name" : "vitisai_11298643349341028369_0","mem_in_use_delta" : "0","parameter_size" : "0","activation_size" : "4096","mem_bytes_requested_in_use" : "69644","mem_in_use_peak" : "200716","output_size" : "65536","mem_requested_in_use_delta" : "0","input_type_shape" : [{"uint16":[1,2048]}],"mem_arena_held_delta" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}}}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :19,"ts" :11702590,"ph" : "X","name" :"logits_offset_32768_DequantizeLinear_kernel_time","args" : {"activation_size" : "65536","mem_requested_in_use_delta" : "131072","provider" : "CPUExecutionProvider","parameter_size" : "6","mem_in_use_delta" : "131072","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"mem_in_use_peak" : "200716","node_index" : "5","output_type_shape" : [{"float":[1,32768]}],"input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_arena_held" : "1048588","mem_bytes_requested_in_use" : "200716","mem_arena_held_delta" : "0","op_name" : "DequantizeLinear","output_size" : "131072","mem_bytes_in_use" : "200716"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1905,"ts" :11700706,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1917,"ts" :11700701,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :12,"ts" :11708465,"ph" : "X","name" :"normalized_x_offset_32768_QuantizeLinear_kernel_time","args" : {"output_type_shape" : [{"uint16":[1,2048]}],"input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"mem_arena_held_delta" : "0","mem_bytes_in_use" : "69644","mem_bytes_requested_in_use" : "69644","mem_in_use_peak" : "200716","provider" : "CPUExecutionProvider","mem_arena_held" : "1048588","mem_in_use_delta" : "0","parameter_size" : "6","activation_size" : "8192","op_name" : "QuantizeLinear","node_index" : "0","output_size" : "4096","mem_requested_in_use_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1917,"ts" :11708480,"ph" : "X","name" :"vitisai_vitisai_11298643349341028369_0_0_kernel_time","args" : {"mem_in_use_peak" : "200716","mem_bytes_requested_in_use" : "69644","activation_size" : "4096","mem_in_use_delta" : "0","output_size" : "65536","mem_arena_held" : "1048588","node_index" : "6","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"mem_arena_held_delta" : "0","mem_requested_in_use_delta" : "0","input_type_shape" : [{"uint16":[1,2048]}],"op_name" : "vitisai_11298643349341028369_0","provider" : "vitisai","mem_bytes_in_use" : "69644","parameter_size" : "0","output_type_shape" : [{"uint16":[1,32768]}]}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :31,"ts" :11710404,"ph" : "X","name" :"logits_offset_32768_DequantizeLinear_kernel_time","args" : {"mem_arena_held_delta" : "0","node_index" : "5","parameter_size" : "6","op_name" : "DequantizeLinear","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"output_size" : "131072","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"provider" : "CPUExecutionProvider","mem_in_use_delta" : "131072","activation_size" : "65536","output_type_shape" : [{"float":[1,32768]}],"mem_bytes_in_use" : "200716","mem_bytes_requested_in_use" : "200716","mem_in_use_peak" : "200716","mem_arena_held" : "1048588","mem_requested_in_use_delta" : "131072"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1974,"ts" :11708464,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1988,"ts" :11708459,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :15,"ts" :11717038,"ph" : "X","name" :"normalized_x_offset_32768_QuantizeLinear_kernel_time","args" : {"mem_requested_in_use_delta" : "0","provider" : "CPUExecutionProvider","mem_in_use_delta" : "0","op_name" : "QuantizeLinear","mem_bytes_in_use" : "69644","node_index" : "0","output_type_shape" : [{"uint16":[1,2048]}],"output_size" : "4096","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_arena_held" : "1048588","mem_arena_held_delta" : "0","mem_in_use_peak" : "200716","activation_size" : "8192","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"parameter_size" : "6","mem_bytes_requested_in_use" : "69644"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1881,"ts" :11717059,"ph" : "X","name" :"vitisai_vitisai_11298643349341028369_0_0_kernel_time","args" : {"provider" : "vitisai","output_size" : "65536","mem_requested_in_use_delta" : "0","node_index" : "6","mem_arena_held" : "1048588","mem_bytes_in_use" : "69644","parameter_size" : "0","op_name" : "vitisai_11298643349341028369_0","mem_in_use_delta" : "0","mem_bytes_requested_in_use" : "69644","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"input_type_shape" : [{"uint16":[1,2048]}],"mem_arena_held_delta" : "0","mem_in_use_peak" : "200716","output_type_shape" : [{"uint16":[1,32768]}],"activation_size" : "4096"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :21,"ts" :11718945,"ph" : "X","name" :"logits_offset_32768_DequantizeLinear_kernel_time","args" : {"mem_in_use_delta" : "131072","activation_size" : "65536","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_in_use_peak" : "200716","mem_bytes_requested_in_use" : "200716","provider" : "CPUExecutionProvider","output_type_shape" : [{"float":[1,32768]}],"node_index" : "5","mem_bytes_in_use" : "200716","output_size" : "131072","mem_arena_held_delta" : "0","mem_requested_in_use_delta" : "131072","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"parameter_size" : "6","mem_arena_held" : "1048588","op_name" : "DequantizeLinear"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1943,"ts" :11717036,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1959,"ts" :11717027,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :14,"ts" :11725058,"ph" : "X","name" :"normalized_x_offset_32768_QuantizeLinear_kernel_time","args" : {"input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"mem_in_use_delta" : "0","provider" : "CPUExecutionProvider","mem_in_use_peak" : "200716","activation_size" : "8192","mem_arena_held_delta" : "0","mem_bytes_requested_in_use" : "69644","mem_bytes_in_use" : "69644","node_index" : "0","mem_arena_held" : "1048588","op_name" : "QuantizeLinear","output_size" : "4096","parameter_size" : "6","output_type_shape" : [{"uint16":[1,2048]}],"mem_requested_in_use_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1918,"ts" :11725075,"ph" : "X","name" :"vitisai_vitisai_11298643349341028369_0_0_kernel_time","args" : {"mem_bytes_requested_in_use" : "69644","mem_in_use_peak" : "200716","input_type_shape" : [{"uint16":[1,2048]}],"output_size" : "65536","mem_requested_in_use_delta" : "0","parameter_size" : "0","mem_bytes_in_use" : "69644","mem_in_use_delta" : "0","output_type_shape" : [{"uint16":[1,32768]}],"node_index" : "6","mem_arena_held" : "1048588","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"activation_size" : "4096","mem_arena_held_delta" : "0","op_name" : "vitisai_11298643349341028369_0","provider" : "vitisai"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :25,"ts" :11726999,"ph" : "X","name" :"logits_offset_32768_DequantizeLinear_kernel_time","args" : {"node_index" : "5","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_in_use_delta" : "131072","output_size" : "131072","mem_arena_held" : "1048588","op_name" : "DequantizeLinear","activation_size" : "65536","mem_bytes_in_use" : "200716","output_type_shape" : [{"float":[1,32768]}],"provider" : "CPUExecutionProvider","mem_in_use_peak" : "200716","mem_requested_in_use_delta" : "131072","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"parameter_size" : "6","mem_arena_held_delta" : "0","mem_bytes_requested_in_use" : "200716"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1970,"ts" :11725057,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1982,"ts" :11725053,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :16,"ts" :11733472,"ph" : "X","name" :"normalized_x_offset_32768_QuantizeLinear_kernel_time","args" : {"mem_bytes_requested_in_use" : "69644","activation_size" : "8192","op_name" : "QuantizeLinear","mem_arena_held" : "1048588","mem_in_use_delta" : "0","output_type_shape" : [{"uint16":[1,2048]}],"mem_arena_held_delta" : "0","node_index" : "0","mem_in_use_peak" : "200716","output_size" : "4096","mem_requested_in_use_delta" : "0","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_bytes_in_use" : "69644","parameter_size" : "6","provider" : "CPUExecutionProvider","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}}}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1878,"ts" :11733491,"ph" : "X","name" :"vitisai_vitisai_11298643349341028369_0_0_kernel_time","args" : {"activation_size" : "4096","op_name" : "vitisai_11298643349341028369_0","output_type_shape" : [{"uint16":[1,32768]}],"input_type_shape" : [{"uint16":[1,2048]}],"mem_arena_held_delta" : "0","mem_in_use_delta" : "0","mem_requested_in_use_delta" : "0","mem_arena_held" : "1048588","parameter_size" : "0","mem_in_use_peak" : "200716","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"node_index" : "6","output_size" : "65536","mem_bytes_requested_in_use" : "69644","provider" : "vitisai","mem_bytes_in_use" : "69644"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :21,"ts" :11735375,"ph" : "X","name" :"logits_offset_32768_DequantizeLinear_kernel_time","args" : {"input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"output_size" : "131072","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"op_name" : "DequantizeLinear","mem_requested_in_use_delta" : "131072","node_index" : "5","parameter_size" : "6","mem_arena_held" : "1048588","mem_bytes_requested_in_use" : "200716","mem_arena_held_delta" : "0","mem_in_use_delta" : "131072","activation_size" : "65536","output_type_shape" : [{"float":[1,32768]}],"mem_bytes_in_use" : "200716","provider" : "CPUExecutionProvider","mem_in_use_peak" : "200716"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1927,"ts" :11733470,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1943,"ts" :11733461,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :13,"ts" :11741332,"ph" : "X","name" :"normalized_x_offset_32768_QuantizeLinear_kernel_time","args" : {"mem_arena_held" : "1048588","activation_size" : "8192","parameter_size" : "6","mem_in_use_delta" : "0","node_index" : "0","output_size" : "4096","mem_bytes_requested_in_use" : "69644","op_name" : "QuantizeLinear","mem_bytes_in_use" : "69644","mem_arena_held_delta" : "0","output_type_shape" : [{"uint16":[1,2048]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"provider" : "CPUExecutionProvider","mem_in_use_peak" : "200716","mem_requested_in_use_delta" : "0","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}]}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1860,"ts" :11741348,"ph" : "X","name" :"vitisai_vitisai_11298643349341028369_0_0_kernel_time","args" : {"input_type_shape" : [{"uint16":[1,2048]}],"output_size" : "65536","output_type_shape" : [{"uint16":[1,32768]}],"mem_arena_held_delta" : "0","mem_requested_in_use_delta" : "0","node_index" : "6","parameter_size" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"mem_in_use_delta" : "0","mem_bytes_requested_in_use" : "69644","provider" : "vitisai","op_name" : "vitisai_11298643349341028369_0","activation_size" : "4096","mem_bytes_in_use" : "69644","mem_arena_held" : "1048588","mem_in_use_peak" : "200716"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :31,"ts" :11743214,"ph" : "X","name" :"logits_offset_32768_DequantizeLinear_kernel_time","args" : {"mem_bytes_in_use" : "200716","output_type_shape" : [{"float":[1,32768]}],"parameter_size" : "6","mem_requested_in_use_delta" : "131072","mem_in_use_peak" : "200716","mem_in_use_delta" : "131072","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"activation_size" : "65536","provider" : "CPUExecutionProvider","mem_arena_held" : "1048588","mem_arena_held_delta" : "0","node_index" : "5","mem_bytes_requested_in_use" : "200716","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"op_name" : "DequantizeLinear","output_size" : "131072"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1920,"ts" :11741330,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1932,"ts" :11741323,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :15,"ts" :11749535,"ph" : "X","name" :"normalized_x_offset_32768_QuantizeLinear_kernel_time","args" : {"input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_in_use_peak" : "200716","mem_arena_held_delta" : "0","mem_arena_held" : "1048588","node_index" : "0","parameter_size" : "6","activation_size" : "8192","output_size" : "4096","op_name" : "QuantizeLinear","mem_requested_in_use_delta" : "0","mem_bytes_in_use" : "69644","provider" : "CPUExecutionProvider","mem_bytes_requested_in_use" : "69644","output_type_shape" : [{"uint16":[1,2048]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"mem_in_use_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :2279,"ts" :11749553,"ph" : "X","name" :"vitisai_vitisai_11298643349341028369_0_0_kernel_time","args" : {"op_name" : "vitisai_11298643349341028369_0","mem_arena_held" : "1048588","output_type_shape" : [{"uint16":[1,32768]}],"mem_in_use_peak" : "200716","activation_size" : "4096","parameter_size" : "0","node_index" : "6","mem_requested_in_use_delta" : "0","mem_bytes_requested_in_use" : "69644","provider" : "vitisai","mem_in_use_delta" : "0","output_size" : "65536","mem_bytes_in_use" : "69644","mem_arena_held_delta" : "0","input_type_shape" : [{"uint16":[1,2048]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}}}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :22,"ts" :11751840,"ph" : "X","name" :"logits_offset_32768_DequantizeLinear_kernel_time","args" : {"mem_arena_held_delta" : "0","mem_bytes_requested_in_use" : "200716","mem_in_use_peak" : "200716","mem_bytes_in_use" : "200716","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"node_index" : "5","activation_size" : "65536","mem_in_use_delta" : "131072","output_size" : "131072","provider" : "CPUExecutionProvider","op_name" : "DequantizeLinear","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"output_type_shape" : [{"float":[1,32768]}],"parameter_size" : "6","mem_requested_in_use_delta" : "131072","mem_arena_held" : "1048588"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2330,"ts" :11749534,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2345,"ts" :11749526,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :14,"ts" :11757835,"ph" : "X","name" :"normalized_x_offset_32768_QuantizeLinear_kernel_time","args" : {"mem_in_use_peak" : "200716","mem_bytes_requested_in_use" : "69644","mem_arena_held" : "1048588","mem_bytes_in_use" : "69644","activation_size" : "8192","mem_in_use_delta" : "0","op_name" : "QuantizeLinear","mem_requested_in_use_delta" : "0","provider" : "CPUExecutionProvider","output_type_shape" : [{"uint16":[1,2048]}],"node_index" : "0","parameter_size" : "6","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"output_size" : "4096","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_arena_held_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1915,"ts" :11757851,"ph" : "X","name" :"vitisai_vitisai_11298643349341028369_0_0_kernel_time","args" : {"mem_in_use_delta" : "0","mem_arena_held" : "1048588","output_size" : "65536","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"node_index" : "6","parameter_size" : "0","mem_bytes_in_use" : "69644","mem_requested_in_use_delta" : "0","input_type_shape" : [{"uint16":[1,2048]}],"op_name" : "vitisai_11298643349341028369_0","provider" : "vitisai","output_type_shape" : [{"uint16":[1,32768]}],"mem_in_use_peak" : "200716","mem_arena_held_delta" : "0","mem_bytes_requested_in_use" : "69644","activation_size" : "4096"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :22,"ts" :11759769,"ph" : "X","name" :"logits_offset_32768_DequantizeLinear_kernel_time","args" : {"output_type_shape" : [{"float":[1,32768]}],"mem_bytes_in_use" : "200716","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"mem_bytes_requested_in_use" : "200716","node_index" : "5","mem_requested_in_use_delta" : "131072","mem_in_use_delta" : "131072","mem_in_use_peak" : "200716","op_name" : "DequantizeLinear","output_size" : "131072","parameter_size" : "6","mem_arena_held_delta" : "0","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"activation_size" : "65536","mem_arena_held" : "1048588","provider" : "CPUExecutionProvider"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1959,"ts" :11757834,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1970,"ts" :11757827,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :14,"ts" :11765872,"ph" : "X","name" :"normalized_x_offset_32768_QuantizeLinear_kernel_time","args" : {"activation_size" : "8192","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"mem_in_use_delta" : "0","output_size" : "4096","output_type_shape" : [{"uint16":[1,2048]}],"provider" : "CPUExecutionProvider","mem_arena_held_delta" : "0","op_name" : "QuantizeLinear","node_index" : "0","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_arena_held" : "1048588","parameter_size" : "6","mem_bytes_requested_in_use" : "69644","mem_in_use_peak" : "200716","mem_bytes_in_use" : "69644","mem_requested_in_use_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1942,"ts" :11765888,"ph" : "X","name" :"vitisai_vitisai_11298643349341028369_0_0_kernel_time","args" : {"parameter_size" : "0","mem_bytes_requested_in_use" : "69644","output_type_shape" : [{"uint16":[1,32768]}],"provider" : "vitisai","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"mem_arena_held" : "1048588","mem_in_use_peak" : "200716","input_type_shape" : [{"uint16":[1,2048]}],"node_index" : "6","mem_bytes_in_use" : "69644","mem_in_use_delta" : "0","mem_requested_in_use_delta" : "0","activation_size" : "4096","mem_arena_held_delta" : "0","output_size" : "65536","op_name" : "vitisai_11298643349341028369_0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :29,"ts" :11767837,"ph" : "X","name" :"logits_offset_32768_DequantizeLinear_kernel_time","args" : {"mem_arena_held_delta" : "0","mem_arena_held" : "1048588","mem_in_use_delta" : "131072","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"node_index" : "5","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"output_size" : "131072","op_name" : "DequantizeLinear","activation_size" : "65536","mem_bytes_requested_in_use" : "200716","mem_bytes_in_use" : "200716","parameter_size" : "6","mem_requested_in_use_delta" : "131072","provider" : "CPUExecutionProvider","output_type_shape" : [{"float":[1,32768]}],"mem_in_use_peak" : "200716"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1999,"ts" :11765871,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2015,"ts" :11765863,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :13,"ts" :11773903,"ph" : "X","name" :"normalized_x_offset_32768_QuantizeLinear_kernel_time","args" : {"node_index" : "0","mem_bytes_requested_in_use" : "69644","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"parameter_size" : "6","mem_in_use_delta" : "0","output_size" : "4096","mem_requested_in_use_delta" : "0","mem_arena_held" : "1048588","mem_in_use_peak" : "200716","op_name" : "QuantizeLinear","output_type_shape" : [{"uint16":[1,2048]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"mem_bytes_in_use" : "69644","mem_arena_held_delta" : "0","provider" : "CPUExecutionProvider","activation_size" : "8192"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1997,"ts" :11773919,"ph" : "X","name" :"vitisai_vitisai_11298643349341028369_0_0_kernel_time","args" : {"mem_bytes_requested_in_use" : "69644","activation_size" : "4096","mem_in_use_peak" : "200716","mem_requested_in_use_delta" : "0","op_name" : "vitisai_11298643349341028369_0","provider" : "vitisai","parameter_size" : "0","input_type_shape" : [{"uint16":[1,2048]}],"mem_bytes_in_use" : "69644","mem_arena_held_delta" : "0","mem_arena_held" : "1048588","mem_in_use_delta" : "0","output_type_shape" : [{"uint16":[1,32768]}],"node_index" : "6","output_size" : "65536","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}}}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :45,"ts" :11775933,"ph" : "X","name" :"logits_offset_32768_DequantizeLinear_kernel_time","args" : {"mem_arena_held_delta" : "0","op_name" : "DequantizeLinear","mem_arena_held" : "1048588","output_size" : "131072","node_index" : "5","mem_requested_in_use_delta" : "131072","parameter_size" : "6","provider" : "CPUExecutionProvider","mem_bytes_in_use" : "200716","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_bytes_requested_in_use" : "200716","mem_in_use_delta" : "131072","activation_size" : "65536","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"output_type_shape" : [{"float":[1,32768]}],"mem_in_use_peak" : "200716"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2082,"ts" :11773902,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2105,"ts" :11773897,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :30,"ts" :11782578,"ph" : "X","name" :"normalized_x_offset_32768_QuantizeLinear_kernel_time","args" : {"parameter_size" : "6","op_name" : "QuantizeLinear","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"mem_bytes_requested_in_use" : "69644","mem_in_use_delta" : "0","output_type_shape" : [{"uint16":[1,2048]}],"input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"output_size" : "4096","node_index" : "0","activation_size" : "8192","mem_bytes_in_use" : "69644","mem_arena_held_delta" : "0","mem_in_use_peak" : "200716","provider" : "CPUExecutionProvider","mem_arena_held" : "1048588","mem_requested_in_use_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1880,"ts" :11782613,"ph" : "X","name" :"vitisai_vitisai_11298643349341028369_0_0_kernel_time","args" : {"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"output_size" : "65536","mem_arena_held_delta" : "0","input_type_shape" : [{"uint16":[1,2048]}],"parameter_size" : "0","node_index" : "6","mem_arena_held" : "1048588","output_type_shape" : [{"uint16":[1,32768]}],"activation_size" : "4096","mem_bytes_in_use" : "69644","mem_requested_in_use_delta" : "0","mem_in_use_delta" : "0","provider" : "vitisai","op_name" : "vitisai_11298643349341028369_0","mem_in_use_peak" : "200716","mem_bytes_requested_in_use" : "69644"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :21,"ts" :11784499,"ph" : "X","name" :"logits_offset_32768_DequantizeLinear_kernel_time","args" : {"output_type_shape" : [{"float":[1,32768]}],"mem_bytes_requested_in_use" : "200716","parameter_size" : "6","mem_in_use_delta" : "131072","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"output_size" : "131072","mem_arena_held" : "1048588","provider" : "CPUExecutionProvider","mem_bytes_in_use" : "200716","op_name" : "DequantizeLinear","mem_requested_in_use_delta" : "131072","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_arena_held_delta" : "0","node_index" : "5","mem_in_use_peak" : "200716","activation_size" : "65536"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1946,"ts" :11782576,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1967,"ts" :11782563,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :33,"ts" :11791328,"ph" : "X","name" :"normalized_x_offset_32768_QuantizeLinear_kernel_time","args" : {"activation_size" : "8192","output_size" : "4096","mem_arena_held" : "1048588","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"parameter_size" : "6","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_requested_in_use_delta" : "0","mem_in_use_delta" : "0","mem_in_use_peak" : "200716","node_index" : "0","mem_bytes_in_use" : "69644","mem_bytes_requested_in_use" : "69644","mem_arena_held_delta" : "0","op_name" : "QuantizeLinear","provider" : "CPUExecutionProvider","output_type_shape" : [{"uint16":[1,2048]}]}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :2133,"ts" :11791366,"ph" : "X","name" :"vitisai_vitisai_11298643349341028369_0_0_kernel_time","args" : {"provider" : "vitisai","node_index" : "6","mem_requested_in_use_delta" : "0","input_type_shape" : [{"uint16":[1,2048]}],"output_size" : "65536","mem_bytes_in_use" : "69644","activation_size" : "4096","mem_in_use_delta" : "0","mem_in_use_peak" : "200716","parameter_size" : "0","mem_bytes_requested_in_use" : "69644","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"output_type_shape" : [{"uint16":[1,32768]}],"op_name" : "vitisai_11298643349341028369_0","mem_arena_held" : "1048588","mem_arena_held_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :75,"ts" :11793514,"ph" : "X","name" :"logits_offset_32768_DequantizeLinear_kernel_time","args" : {"input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_bytes_requested_in_use" : "200716","mem_arena_held" : "1048588","activation_size" : "65536","node_index" : "5","mem_arena_held_delta" : "0","output_type_shape" : [{"float":[1,32768]}],"provider" : "CPUExecutionProvider","mem_bytes_in_use" : "200716","mem_in_use_peak" : "200716","mem_requested_in_use_delta" : "131072","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"parameter_size" : "6","output_size" : "131072","mem_in_use_delta" : "131072","op_name" : "DequantizeLinear"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2271,"ts" :11791326,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2307,"ts" :11791313,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :33,"ts" :11800443,"ph" : "X","name" :"normalized_x_offset_32768_QuantizeLinear_kernel_time","args" : {"activation_size" : "8192","output_type_shape" : [{"uint16":[1,2048]}],"mem_arena_held" : "1048588","mem_in_use_peak" : "200716","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"output_size" : "4096","parameter_size" : "6","mem_bytes_in_use" : "69644","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"op_name" : "QuantizeLinear","mem_arena_held_delta" : "0","provider" : "CPUExecutionProvider","mem_bytes_requested_in_use" : "69644","node_index" : "0","mem_in_use_delta" : "0","mem_requested_in_use_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1937,"ts" :11800482,"ph" : "X","name" :"vitisai_vitisai_11298643349341028369_0_0_kernel_time","args" : {"parameter_size" : "0","input_type_shape" : [{"uint16":[1,2048]}],"output_size" : "65536","mem_bytes_in_use" : "69644","output_type_shape" : [{"uint16":[1,32768]}],"mem_bytes_requested_in_use" : "69644","op_name" : "vitisai_11298643349341028369_0","node_index" : "6","mem_arena_held_delta" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"provider" : "vitisai","mem_in_use_delta" : "0","mem_requested_in_use_delta" : "0","mem_arena_held" : "1048588","activation_size" : "4096","mem_in_use_peak" : "200716"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :39,"ts" :11802426,"ph" : "X","name" :"logits_offset_32768_DequantizeLinear_kernel_time","args" : {"mem_bytes_in_use" : "200716","mem_in_use_delta" : "131072","node_index" : "5","mem_requested_in_use_delta" : "131072","op_name" : "DequantizeLinear","mem_in_use_peak" : "200716","output_size" : "131072","activation_size" : "65536","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"provider" : "CPUExecutionProvider","output_type_shape" : [{"float":[1,32768]}],"mem_bytes_requested_in_use" : "200716","mem_arena_held_delta" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"mem_arena_held" : "1048588","parameter_size" : "6"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2033,"ts" :11800441,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2056,"ts" :11800425,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :49,"ts" :11808823,"ph" : "X","name" :"normalized_x_offset_32768_QuantizeLinear_kernel_time","args" : {"output_type_shape" : [{"uint16":[1,2048]}],"provider" : "CPUExecutionProvider","op_name" : "QuantizeLinear","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"parameter_size" : "6","node_index" : "0","mem_arena_held" : "1048588","mem_arena_held_delta" : "0","mem_bytes_in_use" : "69644","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_in_use_delta" : "0","output_size" : "4096","mem_in_use_peak" : "200716","mem_requested_in_use_delta" : "0","mem_bytes_requested_in_use" : "69644","activation_size" : "8192"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1980,"ts" :11808877,"ph" : "X","name" :"vitisai_vitisai_11298643349341028369_0_0_kernel_time","args" : {"mem_in_use_delta" : "0","input_type_shape" : [{"uint16":[1,2048]}],"mem_in_use_peak" : "200716","mem_bytes_in_use" : "69644","output_size" : "65536","mem_bytes_requested_in_use" : "69644","node_index" : "6","mem_arena_held_delta" : "0","provider" : "vitisai","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"activation_size" : "4096","parameter_size" : "0","op_name" : "vitisai_11298643349341028369_0","mem_requested_in_use_delta" : "0","output_type_shape" : [{"uint16":[1,32768]}],"mem_arena_held" : "1048588"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :32,"ts" :11810867,"ph" : "X","name" :"logits_offset_32768_DequantizeLinear_kernel_time","args" : {"output_type_shape" : [{"float":[1,32768]}],"node_index" : "5","mem_in_use_delta" : "131072","mem_bytes_requested_in_use" : "200716","activation_size" : "65536","mem_in_use_peak" : "200716","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"provider" : "CPUExecutionProvider","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"op_name" : "DequantizeLinear","mem_arena_held" : "1048588","mem_requested_in_use_delta" : "131072","output_size" : "131072","parameter_size" : "6","mem_arena_held_delta" : "0","mem_bytes_in_use" : "200716"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2082,"ts" :11808822,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2104,"ts" :11808812,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :37,"ts" :11817299,"ph" : "X","name" :"normalized_x_offset_32768_QuantizeLinear_kernel_time","args" : {"mem_bytes_requested_in_use" : "69644","op_name" : "QuantizeLinear","output_size" : "4096","mem_requested_in_use_delta" : "0","mem_in_use_peak" : "200716","output_type_shape" : [{"uint16":[1,2048]}],"mem_arena_held_delta" : "0","activation_size" : "8192","mem_arena_held" : "1048588","provider" : "CPUExecutionProvider","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"mem_bytes_in_use" : "69644","node_index" : "0","mem_in_use_delta" : "0","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"parameter_size" : "6"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1954,"ts" :11817343,"ph" : "X","name" :"vitisai_vitisai_11298643349341028369_0_0_kernel_time","args" : {"op_name" : "vitisai_11298643349341028369_0","mem_requested_in_use_delta" : "0","mem_arena_held" : "1048588","activation_size" : "4096","mem_bytes_requested_in_use" : "69644","output_type_shape" : [{"uint16":[1,32768]}],"parameter_size" : "0","mem_arena_held_delta" : "0","mem_in_use_peak" : "200716","provider" : "vitisai","mem_bytes_in_use" : "69644","node_index" : "6","mem_in_use_delta" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"output_size" : "65536","input_type_shape" : [{"uint16":[1,2048]}]}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :23,"ts" :11819304,"ph" : "X","name" :"logits_offset_32768_DequantizeLinear_kernel_time","args" : {"mem_in_use_delta" : "131072","mem_bytes_requested_in_use" : "200716","output_type_shape" : [{"float":[1,32768]}],"parameter_size" : "6","mem_requested_in_use_delta" : "131072","mem_in_use_peak" : "200716","output_size" : "131072","node_index" : "5","activation_size" : "65536","op_name" : "DequantizeLinear","mem_arena_held_delta" : "0","mem_arena_held" : "1048588","mem_bytes_in_use" : "200716","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"provider" : "CPUExecutionProvider"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2033,"ts" :11817296,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2064,"ts" :11817274,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :30,"ts" :11825337,"ph" : "X","name" :"normalized_x_offset_32768_QuantizeLinear_kernel_time","args" : {"parameter_size" : "6","output_size" : "4096","mem_bytes_in_use" : "69644","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_in_use_peak" : "200716","node_index" : "0","provider" : "CPUExecutionProvider","mem_in_use_delta" : "0","mem_arena_held_delta" : "0","mem_bytes_requested_in_use" : "69644","mem_arena_held" : "1048588","mem_requested_in_use_delta" : "0","op_name" : "QuantizeLinear","activation_size" : "8192","output_type_shape" : [{"uint16":[1,2048]}]}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :2083,"ts" :11825373,"ph" : "X","name" :"vitisai_vitisai_11298643349341028369_0_0_kernel_time","args" : {"mem_arena_held_delta" : "0","parameter_size" : "0","mem_in_use_delta" : "0","provider" : "vitisai","activation_size" : "4096","output_size" : "65536","mem_bytes_requested_in_use" : "69644","node_index" : "6","mem_arena_held" : "1048588","op_name" : "vitisai_11298643349341028369_0","input_type_shape" : [{"uint16":[1,2048]}],"mem_in_use_peak" : "200716","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"mem_bytes_in_use" : "69644","mem_requested_in_use_delta" : "0","output_type_shape" : [{"uint16":[1,32768]}]}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :51,"ts" :11827473,"ph" : "X","name" :"logits_offset_32768_DequantizeLinear_kernel_time","args" : {"mem_bytes_requested_in_use" : "200716","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"activation_size" : "65536","mem_arena_held_delta" : "0","provider" : "CPUExecutionProvider","node_index" : "5","mem_bytes_in_use" : "200716","output_type_shape" : [{"float":[1,32768]}],"op_name" : "DequantizeLinear","mem_in_use_delta" : "131072","mem_in_use_peak" : "200716","parameter_size" : "6","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_requested_in_use_delta" : "131072","output_size" : "131072","mem_arena_held" : "1048588"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2194,"ts" :11825336,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2225,"ts" :11825327,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :39,"ts" :11834201,"ph" : "X","name" :"normalized_x_offset_32768_QuantizeLinear_kernel_time","args" : {"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"activation_size" : "8192","parameter_size" : "6","output_size" : "4096","node_index" : "0","provider" : "CPUExecutionProvider","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_bytes_requested_in_use" : "69644","output_type_shape" : [{"uint16":[1,2048]}],"mem_requested_in_use_delta" : "0","mem_arena_held" : "1048588","mem_arena_held_delta" : "0","mem_in_use_peak" : "200716","op_name" : "QuantizeLinear","mem_bytes_in_use" : "69644","mem_in_use_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :2108,"ts" :11834248,"ph" : "X","name" :"vitisai_vitisai_11298643349341028369_0_0_kernel_time","args" : {"parameter_size" : "0","mem_bytes_in_use" : "69644","mem_requested_in_use_delta" : "0","mem_in_use_delta" : "0","input_type_shape" : [{"uint16":[1,2048]}],"provider" : "vitisai","op_name" : "vitisai_11298643349341028369_0","mem_arena_held" : "1048588","activation_size" : "4096","output_size" : "65536","mem_in_use_peak" : "200716","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"mem_arena_held_delta" : "0","node_index" : "6","mem_bytes_requested_in_use" : "69644","output_type_shape" : [{"uint16":[1,32768]}]}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :75,"ts" :11836374,"ph" : "X","name" :"logits_offset_32768_DequantizeLinear_kernel_time","args" : {"output_type_shape" : [{"float":[1,32768]}],"mem_in_use_delta" : "131072","activation_size" : "65536","provider" : "CPUExecutionProvider","mem_arena_held_delta" : "0","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"op_name" : "DequantizeLinear","mem_arena_held" : "1048588","output_size" : "131072","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"mem_bytes_requested_in_use" : "200716","mem_bytes_in_use" : "200716","mem_in_use_peak" : "200716","mem_requested_in_use_delta" : "131072","parameter_size" : "6","node_index" : "5"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2257,"ts" :11834199,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2298,"ts" :11834180,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :13,"ts" :11842617,"ph" : "X","name" :"normalized_x_offset_32768_QuantizeLinear_kernel_time","args" : {"mem_arena_held" : "1048588","mem_bytes_in_use" : "69644","provider" : "CPUExecutionProvider","mem_in_use_peak" : "200716","output_size" : "4096","mem_bytes_requested_in_use" : "69644","activation_size" : "8192","mem_in_use_delta" : "0","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"parameter_size" : "6","mem_arena_held_delta" : "0","node_index" : "0","op_name" : "QuantizeLinear","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"mem_requested_in_use_delta" : "0","output_type_shape" : [{"uint16":[1,2048]}]}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :2058,"ts" :11842633,"ph" : "X","name" :"vitisai_vitisai_11298643349341028369_0_0_kernel_time","args" : {"activation_size" : "4096","mem_in_use_peak" : "200716","mem_arena_held_delta" : "0","mem_requested_in_use_delta" : "0","mem_bytes_in_use" : "69644","output_type_shape" : [{"uint16":[1,32768]}],"mem_in_use_delta" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"mem_arena_held" : "1048588","provider" : "vitisai","input_type_shape" : [{"uint16":[1,2048]}],"node_index" : "6","output_size" : "65536","parameter_size" : "0","op_name" : "vitisai_11298643349341028369_0","mem_bytes_requested_in_use" : "69644"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :53,"ts" :11844705,"ph" : "X","name" :"logits_offset_32768_DequantizeLinear_kernel_time","args" : {"parameter_size" : "6","mem_in_use_delta" : "131072","op_name" : "DequantizeLinear","output_type_shape" : [{"float":[1,32768]}],"mem_in_use_peak" : "200716","output_size" : "131072","mem_arena_held" : "1048588","mem_requested_in_use_delta" : "131072","node_index" : "5","provider" : "CPUExecutionProvider","mem_bytes_in_use" : "200716","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_arena_held_delta" : "0","activation_size" : "65536","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"mem_bytes_requested_in_use" : "200716"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2150,"ts" :11842617,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2177,"ts" :11842612,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :13,"ts" :11851029,"ph" : "X","name" :"normalized_x_offset_32768_QuantizeLinear_kernel_time","args" : {"provider" : "CPUExecutionProvider","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"mem_bytes_requested_in_use" : "69644","op_name" : "QuantizeLinear","parameter_size" : "6","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_arena_held_delta" : "0","mem_requested_in_use_delta" : "0","node_index" : "0","output_type_shape" : [{"uint16":[1,2048]}],"output_size" : "4096","mem_bytes_in_use" : "69644","mem_in_use_peak" : "200716","mem_arena_held" : "1048588","activation_size" : "8192","mem_in_use_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1872,"ts" :11851045,"ph" : "X","name" :"vitisai_vitisai_11298643349341028369_0_0_kernel_time","args" : {"mem_bytes_requested_in_use" : "69644","activation_size" : "4096","mem_in_use_delta" : "0","provider" : "vitisai","mem_requested_in_use_delta" : "0","op_name" : "vitisai_11298643349341028369_0","output_type_shape" : [{"uint16":[1,32768]}],"output_size" : "65536","mem_arena_held_delta" : "0","parameter_size" : "0","input_type_shape" : [{"uint16":[1,2048]}],"mem_bytes_in_use" : "69644","mem_in_use_peak" : "200716","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"node_index" : "6","mem_arena_held" : "1048588"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :20,"ts" :11852922,"ph" : "X","name" :"logits_offset_32768_DequantizeLinear_kernel_time","args" : {"parameter_size" : "6","mem_requested_in_use_delta" : "131072","mem_arena_held_delta" : "0","mem_bytes_requested_in_use" : "200716","provider" : "CPUExecutionProvider","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"output_size" : "131072","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-3-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"66280": {"num_run": 0, "core": -1},"27092": {"num_run": 0, "core": -1},"9608": {"num_run": 0, "core": -1},"21472": {"num_run": 0, "core": -1},"60460": {"num_run": 0, "core": -1},"57752": {"num_run": 0, "core": -1},"39660": {"num_run": 0, "core": -1},"22512": {"num_run": 0, "core": -1},"9660": {"num_run": 0, "core": -1},"27572": {"num_run": 0, "core": -1},"48124": {"num_run": 0, "core": -1}}},"activation_size" : "65536","mem_arena_held" : "1048588","output_type_shape" : [{"float":[1,32768]}],"mem_in_use_delta" : "131072","op_name" : "DequantizeLinear","node_index" : "5","mem_in_use_peak" : "200716","mem_bytes_in_use" : "200716"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1917,"ts" :11851028,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1928,"ts" :11851023,"ph" : "X","name" :"model_run","args" : {}} +] diff --git a/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/shards/shard_65536.json b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/shards/shard_65536.json new file mode 100644 index 000000000..29724350b --- /dev/null +++ b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/shards/shard_65536.json @@ -0,0 +1,95 @@ +{ + "scope": "Spark-X2.5-1.7B output head only; no prefill/decode/KV/sampling or full generation", + "vocab_shard_size": 32768, + "vocab_shard_offset": 65536, + "flatten_input": true, + "batch_repeats": 1, + "matmul_only": true, + "hardware": "AMD Ryzen AI 9 HX 370 NPU", + "os": "Windows-11-10.0.26200-SP0", + "onnx_version": "1.22.0", + "onnxruntime_version": "1.30.0", + "source_model_sha256": "f5f0c31eb438e9940ab242da8bf6f0f5dad6ebc9f6ef02d5475c21b744e595bc", + "prepared_model_sha256": "9a3359737fff4d50600b51c9406c9c794849c0e6815ed761e50c7f18bd70a3b7", + "calibration_sha256": "cfa2ed8439f8fe01730319667bfd01abf2e39cf5851490346a0331304cf63194", + "opset21_sha256": "9a3359737fff4d50600b51c9406c9c794849c0e6815ed761e50c7f18bd70a3b7", + "quantized_sha256": "bc47828ef225d54e5b92a7d45737e86f14519fba5accefdde7d9429ec8d723e8", + "quantization": "opset 21, MatMul only, QDQ, QUInt16 activations/QInt8 weights, MinMax on four real activations", + "input_shape": [ + 1, + 2048 + ], + "output_shape": [ + 1, + 32768 + ], + "source_cpu_reference": [ + { + "snr_db": 137.42232492798837, + "max_abs_diff": 1.9073486328125e-05, + "top1_match": true, + "finite": true + }, + { + "snr_db": 134.337392170828, + "max_abs_diff": 2.6702880859375e-05, + "top1_match": true, + "finite": true + }, + { + "snr_db": 137.81470516345135, + "max_abs_diff": 1.9073486328125e-05, + "top1_match": true, + "finite": true + }, + { + "snr_db": 136.45787925656717, + "max_abs_diff": 1.9073486328125e-05, + "top1_match": true, + "finite": true + } + ], + "npu_devices": 1, + "session_providers": [ + "vitisai", + "CPUExecutionProvider" + ], + "npu_node_events": 4, + "cpu_node_events": 8, + "load_s": 3.8300419999286532, + "npu_rows": [ + { + "snr_db": 57.46049504780351, + "max_abs_diff": 0.225982666015625, + "top1_match": true, + "finite": true, + "index": 0, + "wall_s": 0.004500299924984574 + }, + { + "snr_db": 57.06118310815952, + "max_abs_diff": 0.20765113830566406, + "top1_match": true, + "finite": true, + "index": 1, + "wall_s": 0.0032600000267848372 + }, + { + "snr_db": 56.965567828494514, + "max_abs_diff": 0.20560836791992188, + "top1_match": true, + "finite": true, + "index": 2, + "wall_s": 0.0027278000488877296 + }, + { + "snr_db": 56.03791485557113, + "max_abs_diff": 0.20813369750976562, + "top1_match": true, + "finite": true, + "index": 3, + "wall_s": 0.0030355999479070306 + } + ], + "profile_file": "C:\\Users\\zhout\\w2\\spark_npu_full_shards_named_20260923\\shard_65536\\spark_lm_head_npu_profile_2026-09-23_18-08-01_728.json" +} diff --git a/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/shards/shard_65536_co_resident_profile.json b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/shards/shard_65536_co_resident_profile.json new file mode 100644 index 000000000..c6f5da277 --- /dev/null +++ b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/shards/shard_65536_co_resident_profile.json @@ -0,0 +1,129 @@ +[ +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :22855,"ts" :12,"ph" : "X","name" :"model_loading_uri","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :3988271,"ts" :23948,"ph" : "X","name" :"session_initialization","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :24,"ts" :7794370,"ph" : "X","name" :"normalized_x_offset_65536_QuantizeLinear_kernel_time","args" : {"node_index" : "0","mem_in_use_peak" : "4108","parameter_size" : "6","mem_bytes_requested_in_use" : "4108","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"activation_size" : "8192","mem_in_use_delta" : "4096","mem_arena_held_delta" : "1048576","output_size" : "4096","mem_requested_in_use_delta" : "4096","mem_bytes_in_use" : "4108","output_type_shape" : [{"uint16":[1,2048]}],"mem_arena_held" : "1048588","provider" : "CPUExecutionProvider","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"op_name" : "QuantizeLinear"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :2306,"ts" :7794398,"ph" : "X","name" :"vitisai_vitisai_4170391121016113593_0_0_kernel_time","args" : {"mem_requested_in_use_delta" : "65536","node_index" : "6","provider" : "vitisai","mem_bytes_in_use" : "69644","mem_in_use_peak" : "69644","mem_arena_held" : "1048588","input_type_shape" : [{"uint16":[1,2048]}],"parameter_size" : "0","mem_in_use_delta" : "65536","activation_size" : "4096","mem_bytes_requested_in_use" : "69644","output_type_shape" : [{"uint16":[1,32768]}],"mem_arena_held_delta" : "0","output_size" : "65536","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"op_name" : "vitisai_4170391121016113593_0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :56,"ts" :7796721,"ph" : "X","name" :"logits_offset_65536_DequantizeLinear_kernel_time","args" : {"provider" : "CPUExecutionProvider","output_size" : "131072","node_index" : "5","activation_size" : "65536","mem_bytes_in_use" : "196620","mem_arena_held" : "1048588","mem_requested_in_use_delta" : "131072","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"output_type_shape" : [{"float":[1,32768]}],"op_name" : "DequantizeLinear","mem_arena_held_delta" : "0","mem_in_use_delta" : "131072","mem_in_use_peak" : "196620","mem_bytes_requested_in_use" : "196620","parameter_size" : "6","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}}}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2417,"ts" :7794368,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2434,"ts" :7794360,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :15,"ts" :7806634,"ph" : "X","name" :"normalized_x_offset_65536_QuantizeLinear_kernel_time","args" : {"mem_in_use_delta" : "0","mem_in_use_peak" : "196620","mem_bytes_requested_in_use" : "69644","mem_arena_held_delta" : "0","node_index" : "0","parameter_size" : "6","mem_requested_in_use_delta" : "0","output_size" : "4096","mem_bytes_in_use" : "69644","mem_arena_held" : "1048588","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"output_type_shape" : [{"uint16":[1,2048]}],"op_name" : "QuantizeLinear","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"provider" : "CPUExecutionProvider","activation_size" : "8192"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1869,"ts" :7806653,"ph" : "X","name" :"vitisai_vitisai_4170391121016113593_0_0_kernel_time","args" : {"input_type_shape" : [{"uint16":[1,2048]}],"parameter_size" : "0","provider" : "vitisai","node_index" : "6","mem_bytes_requested_in_use" : "69644","output_size" : "65536","mem_bytes_in_use" : "69644","activation_size" : "4096","output_type_shape" : [{"uint16":[1,32768]}],"mem_in_use_peak" : "196620","mem_arena_held" : "1048588","mem_arena_held_delta" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"mem_in_use_delta" : "0","op_name" : "vitisai_4170391121016113593_0","mem_requested_in_use_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :21,"ts" :7808528,"ph" : "X","name" :"logits_offset_65536_DequantizeLinear_kernel_time","args" : {"output_type_shape" : [{"float":[1,32768]}],"mem_bytes_requested_in_use" : "200716","node_index" : "5","activation_size" : "65536","op_name" : "DequantizeLinear","parameter_size" : "6","provider" : "CPUExecutionProvider","mem_requested_in_use_delta" : "131072","mem_arena_held" : "1048588","output_size" : "131072","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"mem_arena_held_delta" : "0","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_bytes_in_use" : "200716","mem_in_use_peak" : "200716","mem_in_use_delta" : "131072"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1919,"ts" :7806633,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1934,"ts" :7806626,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :15,"ts" :7818972,"ph" : "X","name" :"normalized_x_offset_65536_QuantizeLinear_kernel_time","args" : {"mem_in_use_delta" : "0","op_name" : "QuantizeLinear","mem_arena_held" : "1048588","mem_bytes_requested_in_use" : "69644","output_type_shape" : [{"uint16":[1,2048]}],"provider" : "CPUExecutionProvider","mem_requested_in_use_delta" : "0","mem_in_use_peak" : "200716","parameter_size" : "6","mem_arena_held_delta" : "0","output_size" : "4096","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"mem_bytes_in_use" : "69644","activation_size" : "8192","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"node_index" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1866,"ts" :7818991,"ph" : "X","name" :"vitisai_vitisai_4170391121016113593_0_0_kernel_time","args" : {"output_size" : "65536","mem_arena_held_delta" : "0","node_index" : "6","output_type_shape" : [{"uint16":[1,32768]}],"op_name" : "vitisai_4170391121016113593_0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"mem_requested_in_use_delta" : "0","provider" : "vitisai","parameter_size" : "0","activation_size" : "4096","mem_bytes_requested_in_use" : "69644","mem_in_use_peak" : "200716","input_type_shape" : [{"uint16":[1,2048]}],"mem_bytes_in_use" : "69644","mem_arena_held" : "1048588","mem_in_use_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :22,"ts" :7820865,"ph" : "X","name" :"logits_offset_65536_DequantizeLinear_kernel_time","args" : {"activation_size" : "65536","mem_in_use_delta" : "131072","mem_in_use_peak" : "200716","mem_bytes_requested_in_use" : "200716","parameter_size" : "6","provider" : "CPUExecutionProvider","output_size" : "131072","node_index" : "5","output_type_shape" : [{"float":[1,32768]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"mem_requested_in_use_delta" : "131072","mem_arena_held" : "1048588","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_arena_held_delta" : "0","op_name" : "DequantizeLinear","mem_bytes_in_use" : "200716"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1919,"ts" :7818971,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1936,"ts" :7818962,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :14,"ts" :7829894,"ph" : "X","name" :"normalized_x_offset_65536_QuantizeLinear_kernel_time","args" : {"node_index" : "0","op_name" : "QuantizeLinear","mem_bytes_in_use" : "69644","output_type_shape" : [{"uint16":[1,2048]}],"provider" : "CPUExecutionProvider","mem_arena_held" : "1048588","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_bytes_requested_in_use" : "69644","mem_in_use_delta" : "0","output_size" : "4096","activation_size" : "8192","mem_arena_held_delta" : "0","mem_requested_in_use_delta" : "0","mem_in_use_peak" : "200716","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"parameter_size" : "6"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1939,"ts" :7829912,"ph" : "X","name" :"vitisai_vitisai_4170391121016113593_0_0_kernel_time","args" : {"mem_arena_held_delta" : "0","mem_requested_in_use_delta" : "0","activation_size" : "4096","mem_bytes_requested_in_use" : "69644","op_name" : "vitisai_4170391121016113593_0","mem_arena_held" : "1048588","input_type_shape" : [{"uint16":[1,2048]}],"mem_in_use_peak" : "200716","parameter_size" : "0","provider" : "vitisai","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"output_size" : "65536","node_index" : "6","output_type_shape" : [{"uint16":[1,32768]}],"mem_in_use_delta" : "0","mem_bytes_in_use" : "69644"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :32,"ts" :7831860,"ph" : "X","name" :"logits_offset_65536_DequantizeLinear_kernel_time","args" : {"parameter_size" : "6","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"output_size" : "131072","node_index" : "5","mem_requested_in_use_delta" : "131072","output_type_shape" : [{"float":[1,32768]}],"activation_size" : "65536","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"op_name" : "DequantizeLinear","mem_arena_held" : "1048588","mem_bytes_requested_in_use" : "200716","mem_arena_held_delta" : "0","mem_in_use_delta" : "131072","mem_bytes_in_use" : "200716","provider" : "CPUExecutionProvider","mem_in_use_peak" : "200716"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2008,"ts" :7829893,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2024,"ts" :7829886,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :15,"ts" :7840997,"ph" : "X","name" :"normalized_x_offset_65536_QuantizeLinear_kernel_time","args" : {"mem_requested_in_use_delta" : "0","provider" : "CPUExecutionProvider","mem_in_use_peak" : "200716","mem_in_use_delta" : "0","output_type_shape" : [{"uint16":[1,2048]}],"mem_bytes_in_use" : "69644","activation_size" : "8192","parameter_size" : "6","mem_arena_held_delta" : "0","mem_arena_held" : "1048588","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"op_name" : "QuantizeLinear","output_size" : "4096","node_index" : "0","mem_bytes_requested_in_use" : "69644","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}}}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1850,"ts" :7841017,"ph" : "X","name" :"vitisai_vitisai_4170391121016113593_0_0_kernel_time","args" : {"output_size" : "65536","mem_arena_held" : "1048588","provider" : "vitisai","mem_arena_held_delta" : "0","mem_requested_in_use_delta" : "0","mem_in_use_peak" : "200716","mem_bytes_in_use" : "69644","op_name" : "vitisai_4170391121016113593_0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"activation_size" : "4096","output_type_shape" : [{"uint16":[1,32768]}],"mem_in_use_delta" : "0","node_index" : "6","parameter_size" : "0","input_type_shape" : [{"uint16":[1,2048]}],"mem_bytes_requested_in_use" : "69644"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :19,"ts" :7842870,"ph" : "X","name" :"logits_offset_65536_DequantizeLinear_kernel_time","args" : {"activation_size" : "65536","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"provider" : "CPUExecutionProvider","mem_arena_held" : "1048588","node_index" : "5","parameter_size" : "6","output_size" : "131072","output_type_shape" : [{"float":[1,32768]}],"mem_requested_in_use_delta" : "131072","mem_arena_held_delta" : "0","mem_bytes_in_use" : "200716","op_name" : "DequantizeLinear","mem_bytes_requested_in_use" : "200716","mem_in_use_peak" : "200716","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_in_use_delta" : "131072"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1896,"ts" :7840995,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1908,"ts" :7840986,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :16,"ts" :7848940,"ph" : "X","name" :"normalized_x_offset_65536_QuantizeLinear_kernel_time","args" : {"output_size" : "4096","mem_bytes_requested_in_use" : "69644","mem_in_use_delta" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"parameter_size" : "6","op_name" : "QuantizeLinear","activation_size" : "8192","provider" : "CPUExecutionProvider","node_index" : "0","output_type_shape" : [{"uint16":[1,2048]}],"mem_in_use_peak" : "200716","mem_requested_in_use_delta" : "0","mem_bytes_in_use" : "69644","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_arena_held" : "1048588","mem_arena_held_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :2131,"ts" :7848960,"ph" : "X","name" :"vitisai_vitisai_4170391121016113593_0_0_kernel_time","args" : {"provider" : "vitisai","output_size" : "65536","mem_in_use_delta" : "0","mem_in_use_peak" : "200716","mem_arena_held" : "1048588","mem_bytes_in_use" : "69644","op_name" : "vitisai_4170391121016113593_0","mem_requested_in_use_delta" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"node_index" : "6","mem_arena_held_delta" : "0","output_type_shape" : [{"uint16":[1,32768]}],"activation_size" : "4096","parameter_size" : "0","input_type_shape" : [{"uint16":[1,2048]}],"mem_bytes_requested_in_use" : "69644"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :23,"ts" :7851101,"ph" : "X","name" :"logits_offset_65536_DequantizeLinear_kernel_time","args" : {"mem_arena_held" : "1048588","mem_bytes_in_use" : "200716","op_name" : "DequantizeLinear","mem_in_use_peak" : "200716","provider" : "CPUExecutionProvider","parameter_size" : "6","activation_size" : "65536","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"output_size" : "131072","mem_arena_held_delta" : "0","mem_in_use_delta" : "131072","output_type_shape" : [{"float":[1,32768]}],"mem_bytes_requested_in_use" : "200716","mem_requested_in_use_delta" : "131072","node_index" : "5"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2188,"ts" :7848939,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2204,"ts" :7848930,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :14,"ts" :7857097,"ph" : "X","name" :"normalized_x_offset_65536_QuantizeLinear_kernel_time","args" : {"mem_bytes_in_use" : "69644","mem_requested_in_use_delta" : "0","mem_arena_held_delta" : "0","mem_in_use_peak" : "200716","op_name" : "QuantizeLinear","output_type_shape" : [{"uint16":[1,2048]}],"provider" : "CPUExecutionProvider","activation_size" : "8192","parameter_size" : "6","output_size" : "4096","mem_arena_held" : "1048588","mem_in_use_delta" : "0","node_index" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_bytes_requested_in_use" : "69644"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1830,"ts" :7857114,"ph" : "X","name" :"vitisai_vitisai_4170391121016113593_0_0_kernel_time","args" : {"mem_in_use_peak" : "200716","mem_requested_in_use_delta" : "0","output_size" : "65536","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"output_type_shape" : [{"uint16":[1,32768]}],"op_name" : "vitisai_4170391121016113593_0","activation_size" : "4096","provider" : "vitisai","parameter_size" : "0","mem_in_use_delta" : "0","mem_arena_held" : "1048588","mem_bytes_in_use" : "69644","mem_bytes_requested_in_use" : "69644","node_index" : "6","mem_arena_held_delta" : "0","input_type_shape" : [{"uint16":[1,2048]}]}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :16,"ts" :7858947,"ph" : "X","name" :"logits_offset_65536_DequantizeLinear_kernel_time","args" : {"mem_requested_in_use_delta" : "131072","mem_in_use_peak" : "200716","op_name" : "DequantizeLinear","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"mem_bytes_in_use" : "200716","mem_arena_held_delta" : "0","provider" : "CPUExecutionProvider","mem_in_use_delta" : "131072","mem_arena_held" : "1048588","mem_bytes_requested_in_use" : "200716","output_size" : "131072","parameter_size" : "6","node_index" : "5","activation_size" : "65536","output_type_shape" : [{"float":[1,32768]}],"input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}]}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1869,"ts" :7857096,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1879,"ts" :7857090,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :22,"ts" :7864930,"ph" : "X","name" :"normalized_x_offset_65536_QuantizeLinear_kernel_time","args" : {"input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"node_index" : "0","mem_arena_held" : "1048588","parameter_size" : "6","mem_requested_in_use_delta" : "0","mem_bytes_in_use" : "69644","op_name" : "QuantizeLinear","activation_size" : "8192","mem_arena_held_delta" : "0","mem_bytes_requested_in_use" : "69644","mem_in_use_peak" : "200716","output_size" : "4096","output_type_shape" : [{"uint16":[1,2048]}],"mem_in_use_delta" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"provider" : "CPUExecutionProvider"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :2040,"ts" :7864956,"ph" : "X","name" :"vitisai_vitisai_4170391121016113593_0_0_kernel_time","args" : {"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"mem_requested_in_use_delta" : "0","op_name" : "vitisai_4170391121016113593_0","output_size" : "65536","node_index" : "6","parameter_size" : "0","mem_bytes_requested_in_use" : "69644","mem_arena_held" : "1048588","mem_in_use_delta" : "0","mem_arena_held_delta" : "0","mem_in_use_peak" : "200716","mem_bytes_in_use" : "69644","activation_size" : "4096","input_type_shape" : [{"uint16":[1,2048]}],"output_type_shape" : [{"uint16":[1,32768]}],"provider" : "vitisai"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :74,"ts" :7867015,"ph" : "X","name" :"logits_offset_65536_DequantizeLinear_kernel_time","args" : {"parameter_size" : "6","op_name" : "DequantizeLinear","output_size" : "131072","mem_requested_in_use_delta" : "131072","mem_in_use_delta" : "131072","mem_arena_held" : "1048588","node_index" : "5","mem_arena_held_delta" : "0","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_bytes_requested_in_use" : "200716","mem_in_use_peak" : "200716","provider" : "CPUExecutionProvider","output_type_shape" : [{"float":[1,32768]}],"activation_size" : "65536","mem_bytes_in_use" : "200716","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}}}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2168,"ts" :7864929,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2196,"ts" :7864922,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :18,"ts" :7873469,"ph" : "X","name" :"normalized_x_offset_65536_QuantizeLinear_kernel_time","args" : {"activation_size" : "8192","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"op_name" : "QuantizeLinear","mem_arena_held" : "1048588","provider" : "CPUExecutionProvider","mem_in_use_peak" : "200716","node_index" : "0","output_type_shape" : [{"uint16":[1,2048]}],"mem_bytes_in_use" : "69644","output_size" : "4096","mem_requested_in_use_delta" : "0","mem_arena_held_delta" : "0","mem_bytes_requested_in_use" : "69644","mem_in_use_delta" : "0","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"parameter_size" : "6"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1891,"ts" :7873493,"ph" : "X","name" :"vitisai_vitisai_4170391121016113593_0_0_kernel_time","args" : {"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"mem_in_use_delta" : "0","mem_bytes_in_use" : "69644","mem_arena_held" : "1048588","op_name" : "vitisai_4170391121016113593_0","output_type_shape" : [{"uint16":[1,32768]}],"mem_bytes_requested_in_use" : "69644","node_index" : "6","mem_in_use_peak" : "200716","input_type_shape" : [{"uint16":[1,2048]}],"activation_size" : "4096","mem_arena_held_delta" : "0","output_size" : "65536","provider" : "vitisai","mem_requested_in_use_delta" : "0","parameter_size" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :21,"ts" :7875388,"ph" : "X","name" :"logits_offset_65536_DequantizeLinear_kernel_time","args" : {"input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_bytes_in_use" : "200716","output_size" : "131072","activation_size" : "65536","op_name" : "DequantizeLinear","parameter_size" : "6","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"mem_arena_held" : "1048588","mem_in_use_peak" : "200716","output_type_shape" : [{"float":[1,32768]}],"provider" : "CPUExecutionProvider","mem_bytes_requested_in_use" : "200716","mem_in_use_delta" : "131072","node_index" : "5","mem_requested_in_use_delta" : "131072","mem_arena_held_delta" : "0"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1944,"ts" :7873468,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1958,"ts" :7873459,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :19,"ts" :7881517,"ph" : "X","name" :"normalized_x_offset_65536_QuantizeLinear_kernel_time","args" : {"node_index" : "0","provider" : "CPUExecutionProvider","mem_in_use_peak" : "200716","mem_arena_held" : "1048588","mem_in_use_delta" : "0","mem_requested_in_use_delta" : "0","output_type_shape" : [{"uint16":[1,2048]}],"op_name" : "QuantizeLinear","activation_size" : "8192","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_bytes_in_use" : "69644","parameter_size" : "6","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"output_size" : "4096","mem_bytes_requested_in_use" : "69644","mem_arena_held_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :2106,"ts" :7881540,"ph" : "X","name" :"vitisai_vitisai_4170391121016113593_0_0_kernel_time","args" : {"mem_in_use_delta" : "0","op_name" : "vitisai_4170391121016113593_0","mem_arena_held" : "1048588","node_index" : "6","provider" : "vitisai","parameter_size" : "0","mem_bytes_requested_in_use" : "69644","activation_size" : "4096","input_type_shape" : [{"uint16":[1,2048]}],"mem_requested_in_use_delta" : "0","mem_bytes_in_use" : "69644","mem_in_use_peak" : "200716","output_type_shape" : [{"uint16":[1,32768]}],"output_size" : "65536","mem_arena_held_delta" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}}}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :33,"ts" :7883658,"ph" : "X","name" :"logits_offset_65536_DequantizeLinear_kernel_time","args" : {"parameter_size" : "6","mem_arena_held_delta" : "0","mem_bytes_in_use" : "200716","mem_in_use_delta" : "131072","mem_bytes_requested_in_use" : "200716","output_size" : "131072","provider" : "CPUExecutionProvider","node_index" : "5","mem_requested_in_use_delta" : "131072","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"mem_in_use_peak" : "200716","activation_size" : "65536","op_name" : "DequantizeLinear","output_type_shape" : [{"float":[1,32768]}],"input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_arena_held" : "1048588"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2178,"ts" :7881516,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2196,"ts" :7881510,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :14,"ts" :7889884,"ph" : "X","name" :"normalized_x_offset_65536_QuantizeLinear_kernel_time","args" : {"node_index" : "0","mem_bytes_in_use" : "69644","output_type_shape" : [{"uint16":[1,2048]}],"mem_requested_in_use_delta" : "0","mem_in_use_delta" : "0","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_arena_held" : "1048588","op_name" : "QuantizeLinear","output_size" : "4096","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"activation_size" : "8192","mem_in_use_peak" : "200716","mem_bytes_requested_in_use" : "69644","provider" : "CPUExecutionProvider","parameter_size" : "6","mem_arena_held_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1879,"ts" :7889902,"ph" : "X","name" :"vitisai_vitisai_4170391121016113593_0_0_kernel_time","args" : {"input_type_shape" : [{"uint16":[1,2048]}],"node_index" : "6","output_size" : "65536","mem_requested_in_use_delta" : "0","op_name" : "vitisai_4170391121016113593_0","parameter_size" : "0","mem_bytes_in_use" : "69644","mem_arena_held" : "1048588","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"output_type_shape" : [{"uint16":[1,32768]}],"provider" : "vitisai","mem_arena_held_delta" : "0","mem_in_use_delta" : "0","mem_bytes_requested_in_use" : "69644","activation_size" : "4096","mem_in_use_peak" : "200716"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :20,"ts" :7891786,"ph" : "X","name" :"logits_offset_65536_DequantizeLinear_kernel_time","args" : {"mem_in_use_delta" : "131072","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"node_index" : "5","mem_in_use_peak" : "200716","mem_arena_held_delta" : "0","parameter_size" : "6","mem_requested_in_use_delta" : "131072","activation_size" : "65536","mem_bytes_requested_in_use" : "200716","provider" : "CPUExecutionProvider","output_type_shape" : [{"float":[1,32768]}],"output_size" : "131072","mem_bytes_in_use" : "200716","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"op_name" : "DequantizeLinear","mem_arena_held" : "1048588"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1926,"ts" :7889883,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1939,"ts" :7889876,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :15,"ts" :7897736,"ph" : "X","name" :"normalized_x_offset_65536_QuantizeLinear_kernel_time","args" : {"output_type_shape" : [{"uint16":[1,2048]}],"node_index" : "0","op_name" : "QuantizeLinear","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"mem_in_use_peak" : "200716","mem_in_use_delta" : "0","activation_size" : "8192","mem_bytes_requested_in_use" : "69644","provider" : "CPUExecutionProvider","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_arena_held" : "1048588","mem_requested_in_use_delta" : "0","parameter_size" : "6","mem_bytes_in_use" : "69644","output_size" : "4096","mem_arena_held_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1991,"ts" :7897753,"ph" : "X","name" :"vitisai_vitisai_4170391121016113593_0_0_kernel_time","args" : {"output_type_shape" : [{"uint16":[1,32768]}],"mem_requested_in_use_delta" : "0","mem_arena_held" : "1048588","mem_arena_held_delta" : "0","input_type_shape" : [{"uint16":[1,2048]}],"parameter_size" : "0","activation_size" : "4096","provider" : "vitisai","op_name" : "vitisai_4170391121016113593_0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"output_size" : "65536","mem_bytes_requested_in_use" : "69644","mem_in_use_peak" : "200716","mem_bytes_in_use" : "69644","mem_in_use_delta" : "0","node_index" : "6"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :25,"ts" :7899754,"ph" : "X","name" :"logits_offset_65536_DequantizeLinear_kernel_time","args" : {"op_name" : "DequantizeLinear","mem_bytes_requested_in_use" : "200716","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_arena_held" : "1048588","mem_in_use_delta" : "131072","output_size" : "131072","mem_in_use_peak" : "200716","mem_requested_in_use_delta" : "131072","mem_arena_held_delta" : "0","node_index" : "5","output_type_shape" : [{"float":[1,32768]}],"activation_size" : "65536","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"provider" : "CPUExecutionProvider","parameter_size" : "6","mem_bytes_in_use" : "200716"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2048,"ts" :7897734,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2068,"ts" :7897728,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :14,"ts" :7906351,"ph" : "X","name" :"normalized_x_offset_65536_QuantizeLinear_kernel_time","args" : {"provider" : "CPUExecutionProvider","node_index" : "0","parameter_size" : "6","activation_size" : "8192","mem_bytes_in_use" : "69644","mem_in_use_peak" : "200716","output_size" : "4096","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"output_type_shape" : [{"uint16":[1,2048]}],"mem_requested_in_use_delta" : "0","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_arena_held" : "1048588","op_name" : "QuantizeLinear","mem_bytes_requested_in_use" : "69644","mem_arena_held_delta" : "0","mem_in_use_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1871,"ts" :7906367,"ph" : "X","name" :"vitisai_vitisai_4170391121016113593_0_0_kernel_time","args" : {"mem_arena_held_delta" : "0","mem_in_use_delta" : "0","mem_requested_in_use_delta" : "0","output_size" : "65536","output_type_shape" : [{"uint16":[1,32768]}],"mem_bytes_requested_in_use" : "69644","activation_size" : "4096","provider" : "vitisai","mem_in_use_peak" : "200716","op_name" : "vitisai_4170391121016113593_0","node_index" : "6","input_type_shape" : [{"uint16":[1,2048]}],"mem_arena_held" : "1048588","mem_bytes_in_use" : "69644","parameter_size" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}}}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :21,"ts" :7908259,"ph" : "X","name" :"logits_offset_65536_DequantizeLinear_kernel_time","args" : {"activation_size" : "65536","op_name" : "DequantizeLinear","mem_arena_held" : "1048588","output_type_shape" : [{"float":[1,32768]}],"input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"output_size" : "131072","mem_arena_held_delta" : "0","node_index" : "5","mem_in_use_peak" : "200716","mem_bytes_requested_in_use" : "200716","provider" : "CPUExecutionProvider","mem_requested_in_use_delta" : "131072","parameter_size" : "6","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"mem_bytes_in_use" : "200716","mem_in_use_delta" : "131072"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1933,"ts" :7906349,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1946,"ts" :7906342,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :17,"ts" :7914268,"ph" : "X","name" :"normalized_x_offset_65536_QuantizeLinear_kernel_time","args" : {"mem_in_use_delta" : "0","mem_bytes_requested_in_use" : "69644","output_size" : "4096","mem_arena_held_delta" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"mem_in_use_peak" : "200716","output_type_shape" : [{"uint16":[1,2048]}],"input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"provider" : "CPUExecutionProvider","mem_requested_in_use_delta" : "0","parameter_size" : "6","node_index" : "0","op_name" : "QuantizeLinear","mem_arena_held" : "1048588","mem_bytes_in_use" : "69644","activation_size" : "8192"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1897,"ts" :7914287,"ph" : "X","name" :"vitisai_vitisai_4170391121016113593_0_0_kernel_time","args" : {"parameter_size" : "0","mem_requested_in_use_delta" : "0","mem_bytes_requested_in_use" : "69644","output_type_shape" : [{"uint16":[1,32768]}],"mem_in_use_delta" : "0","output_size" : "65536","mem_arena_held_delta" : "0","op_name" : "vitisai_4170391121016113593_0","mem_arena_held" : "1048588","mem_in_use_peak" : "200716","node_index" : "6","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"mem_bytes_in_use" : "69644","provider" : "vitisai","activation_size" : "4096","input_type_shape" : [{"uint16":[1,2048]}]}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :20,"ts" :7916191,"ph" : "X","name" :"logits_offset_65536_DequantizeLinear_kernel_time","args" : {"provider" : "CPUExecutionProvider","parameter_size" : "6","activation_size" : "65536","mem_arena_held_delta" : "0","output_type_shape" : [{"float":[1,32768]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"mem_in_use_delta" : "131072","output_size" : "131072","mem_in_use_peak" : "200716","mem_arena_held" : "1048588","mem_bytes_requested_in_use" : "200716","mem_bytes_in_use" : "200716","mem_requested_in_use_delta" : "131072","op_name" : "DequantizeLinear","node_index" : "5","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}]}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1946,"ts" :7914267,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1959,"ts" :7914263,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :21,"ts" :7922362,"ph" : "X","name" :"normalized_x_offset_65536_QuantizeLinear_kernel_time","args" : {"mem_requested_in_use_delta" : "0","output_type_shape" : [{"uint16":[1,2048]}],"mem_in_use_delta" : "0","mem_arena_held" : "1048588","mem_arena_held_delta" : "0","parameter_size" : "6","activation_size" : "8192","node_index" : "0","provider" : "CPUExecutionProvider","mem_bytes_requested_in_use" : "69644","mem_bytes_in_use" : "69644","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"op_name" : "QuantizeLinear","output_size" : "4096","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"mem_in_use_peak" : "200716"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1866,"ts" :7922387,"ph" : "X","name" :"vitisai_vitisai_4170391121016113593_0_0_kernel_time","args" : {"mem_bytes_in_use" : "69644","input_type_shape" : [{"uint16":[1,2048]}],"output_size" : "65536","mem_requested_in_use_delta" : "0","mem_in_use_peak" : "200716","mem_bytes_requested_in_use" : "69644","activation_size" : "4096","output_type_shape" : [{"uint16":[1,32768]}],"op_name" : "vitisai_4170391121016113593_0","node_index" : "6","parameter_size" : "0","mem_in_use_delta" : "0","mem_arena_held" : "1048588","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"provider" : "vitisai","mem_arena_held_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :22,"ts" :7924259,"ph" : "X","name" :"logits_offset_65536_DequantizeLinear_kernel_time","args" : {"mem_in_use_peak" : "200716","provider" : "CPUExecutionProvider","mem_in_use_delta" : "131072","parameter_size" : "6","mem_bytes_requested_in_use" : "200716","node_index" : "5","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"mem_arena_held" : "1048588","output_type_shape" : [{"float":[1,32768]}],"op_name" : "DequantizeLinear","activation_size" : "65536","mem_bytes_in_use" : "200716","output_size" : "131072","mem_arena_held_delta" : "0","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_requested_in_use_delta" : "131072"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1923,"ts" :7922361,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1937,"ts" :7922354,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :31,"ts" :7930555,"ph" : "X","name" :"normalized_x_offset_65536_QuantizeLinear_kernel_time","args" : {"provider" : "CPUExecutionProvider","mem_requested_in_use_delta" : "0","activation_size" : "8192","mem_arena_held" : "1048588","mem_arena_held_delta" : "0","mem_bytes_requested_in_use" : "69644","node_index" : "0","mem_in_use_delta" : "0","mem_bytes_in_use" : "69644","output_size" : "4096","op_name" : "QuantizeLinear","mem_in_use_peak" : "200716","output_type_shape" : [{"uint16":[1,2048]}],"input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"parameter_size" : "6"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :2050,"ts" :7930591,"ph" : "X","name" :"vitisai_vitisai_4170391121016113593_0_0_kernel_time","args" : {"mem_arena_held" : "1048588","input_type_shape" : [{"uint16":[1,2048]}],"mem_in_use_peak" : "200716","mem_arena_held_delta" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"mem_bytes_in_use" : "69644","parameter_size" : "0","node_index" : "6","activation_size" : "4096","mem_in_use_delta" : "0","output_size" : "65536","provider" : "vitisai","mem_bytes_requested_in_use" : "69644","op_name" : "vitisai_4170391121016113593_0","output_type_shape" : [{"uint16":[1,32768]}],"mem_requested_in_use_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :29,"ts" :7932649,"ph" : "X","name" :"logits_offset_65536_DequantizeLinear_kernel_time","args" : {"input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"output_size" : "131072","parameter_size" : "6","mem_in_use_delta" : "131072","mem_bytes_requested_in_use" : "200716","mem_bytes_in_use" : "200716","op_name" : "DequantizeLinear","mem_in_use_peak" : "200716","node_index" : "5","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"mem_arena_held" : "1048588","provider" : "CPUExecutionProvider","activation_size" : "65536","mem_arena_held_delta" : "0","mem_requested_in_use_delta" : "131072","output_type_shape" : [{"float":[1,32768]}]}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2129,"ts" :7930553,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2157,"ts" :7930535,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :15,"ts" :7939010,"ph" : "X","name" :"normalized_x_offset_65536_QuantizeLinear_kernel_time","args" : {"output_size" : "4096","parameter_size" : "6","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"provider" : "CPUExecutionProvider","output_type_shape" : [{"uint16":[1,2048]}],"op_name" : "QuantizeLinear","mem_requested_in_use_delta" : "0","mem_arena_held" : "1048588","node_index" : "0","mem_in_use_peak" : "200716","mem_in_use_delta" : "0","mem_bytes_requested_in_use" : "69644","activation_size" : "8192","mem_arena_held_delta" : "0","mem_bytes_in_use" : "69644"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :2002,"ts" :7939030,"ph" : "X","name" :"vitisai_vitisai_4170391121016113593_0_0_kernel_time","args" : {"mem_arena_held" : "1048588","node_index" : "6","op_name" : "vitisai_4170391121016113593_0","mem_requested_in_use_delta" : "0","output_type_shape" : [{"uint16":[1,32768]}],"mem_bytes_requested_in_use" : "69644","parameter_size" : "0","mem_arena_held_delta" : "0","input_type_shape" : [{"uint16":[1,2048]}],"mem_in_use_peak" : "200716","provider" : "vitisai","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"activation_size" : "4096","mem_bytes_in_use" : "69644","output_size" : "65536","mem_in_use_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :72,"ts" :7941098,"ph" : "X","name" :"logits_offset_65536_DequantizeLinear_kernel_time","args" : {"output_size" : "131072","output_type_shape" : [{"float":[1,32768]}],"mem_requested_in_use_delta" : "131072","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_in_use_peak" : "200716","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"parameter_size" : "6","mem_arena_held_delta" : "0","activation_size" : "65536","mem_in_use_delta" : "131072","node_index" : "5","provider" : "CPUExecutionProvider","mem_arena_held" : "1048588","mem_bytes_requested_in_use" : "200716","op_name" : "DequantizeLinear","mem_bytes_in_use" : "200716"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2170,"ts" :7939009,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2197,"ts" :7939002,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :56,"ts" :7948163,"ph" : "X","name" :"normalized_x_offset_65536_QuantizeLinear_kernel_time","args" : {"op_name" : "QuantizeLinear","mem_bytes_requested_in_use" : "69644","output_size" : "4096","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"provider" : "CPUExecutionProvider","output_type_shape" : [{"uint16":[1,2048]}],"node_index" : "0","mem_bytes_in_use" : "69644","mem_requested_in_use_delta" : "0","mem_arena_held" : "1048588","mem_in_use_peak" : "200716","activation_size" : "8192","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"mem_arena_held_delta" : "0","parameter_size" : "6","mem_in_use_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :2088,"ts" :7948227,"ph" : "X","name" :"vitisai_vitisai_4170391121016113593_0_0_kernel_time","args" : {"mem_bytes_in_use" : "69644","parameter_size" : "0","input_type_shape" : [{"uint16":[1,2048]}],"mem_arena_held" : "1048588","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"mem_arena_held_delta" : "0","output_type_shape" : [{"uint16":[1,32768]}],"node_index" : "6","mem_in_use_delta" : "0","op_name" : "vitisai_4170391121016113593_0","activation_size" : "4096","mem_requested_in_use_delta" : "0","mem_in_use_peak" : "200716","mem_bytes_requested_in_use" : "69644","provider" : "vitisai","output_size" : "65536"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :50,"ts" :7950330,"ph" : "X","name" :"logits_offset_65536_DequantizeLinear_kernel_time","args" : {"mem_in_use_delta" : "131072","mem_arena_held" : "1048588","node_index" : "5","mem_arena_held_delta" : "0","activation_size" : "65536","output_size" : "131072","output_type_shape" : [{"float":[1,32768]}],"mem_bytes_requested_in_use" : "200716","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"op_name" : "DequantizeLinear","parameter_size" : "6","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_in_use_peak" : "200716","provider" : "CPUExecutionProvider","mem_requested_in_use_delta" : "131072","mem_bytes_in_use" : "200716"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2224,"ts" :7948161,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2267,"ts" :7948141,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :31,"ts" :7956966,"ph" : "X","name" :"normalized_x_offset_65536_QuantizeLinear_kernel_time","args" : {"input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_in_use_peak" : "200716","activation_size" : "8192","output_type_shape" : [{"uint16":[1,2048]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"mem_in_use_delta" : "0","provider" : "CPUExecutionProvider","mem_bytes_in_use" : "69644","parameter_size" : "6","node_index" : "0","op_name" : "QuantizeLinear","mem_requested_in_use_delta" : "0","mem_arena_held" : "1048588","mem_bytes_requested_in_use" : "69644","output_size" : "4096","mem_arena_held_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1946,"ts" :7957002,"ph" : "X","name" :"vitisai_vitisai_4170391121016113593_0_0_kernel_time","args" : {"mem_bytes_in_use" : "69644","node_index" : "6","mem_requested_in_use_delta" : "0","input_type_shape" : [{"uint16":[1,2048]}],"output_size" : "65536","mem_arena_held" : "1048588","parameter_size" : "0","output_type_shape" : [{"uint16":[1,32768]}],"mem_bytes_requested_in_use" : "69644","mem_in_use_delta" : "0","activation_size" : "4096","mem_arena_held_delta" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"op_name" : "vitisai_4170391121016113593_0","mem_in_use_peak" : "200716","provider" : "vitisai"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :41,"ts" :7958957,"ph" : "X","name" :"logits_offset_65536_DequantizeLinear_kernel_time","args" : {"mem_arena_held" : "1048588","activation_size" : "65536","parameter_size" : "6","node_index" : "5","output_type_shape" : [{"float":[1,32768]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"mem_in_use_peak" : "200716","op_name" : "DequantizeLinear","provider" : "CPUExecutionProvider","mem_bytes_in_use" : "200716","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"output_size" : "131072","mem_arena_held_delta" : "0","mem_requested_in_use_delta" : "131072","mem_bytes_requested_in_use" : "200716","mem_in_use_delta" : "131072"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2043,"ts" :7956964,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2064,"ts" :7956955,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :24,"ts" :7965415,"ph" : "X","name" :"normalized_x_offset_65536_QuantizeLinear_kernel_time","args" : {"mem_arena_held_delta" : "0","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"parameter_size" : "6","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"mem_requested_in_use_delta" : "0","output_type_shape" : [{"uint16":[1,2048]}],"mem_bytes_in_use" : "69644","activation_size" : "8192","mem_arena_held" : "1048588","node_index" : "0","mem_bytes_requested_in_use" : "69644","output_size" : "4096","provider" : "CPUExecutionProvider","op_name" : "QuantizeLinear","mem_in_use_delta" : "0","mem_in_use_peak" : "200716"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1896,"ts" :7965443,"ph" : "X","name" :"vitisai_vitisai_4170391121016113593_0_0_kernel_time","args" : {"op_name" : "vitisai_4170391121016113593_0","mem_bytes_requested_in_use" : "69644","parameter_size" : "0","node_index" : "6","mem_in_use_peak" : "200716","mem_arena_held_delta" : "0","provider" : "vitisai","output_type_shape" : [{"uint16":[1,32768]}],"mem_requested_in_use_delta" : "0","mem_bytes_in_use" : "69644","mem_in_use_delta" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"input_type_shape" : [{"uint16":[1,2048]}],"mem_arena_held" : "1048588","output_size" : "65536","activation_size" : "4096"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :24,"ts" :7967347,"ph" : "X","name" :"logits_offset_65536_DequantizeLinear_kernel_time","args" : {"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"provider" : "CPUExecutionProvider","output_type_shape" : [{"float":[1,32768]}],"input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_arena_held_delta" : "0","output_size" : "131072","mem_arena_held" : "1048588","mem_bytes_in_use" : "200716","op_name" : "DequantizeLinear","mem_in_use_peak" : "200716","mem_bytes_requested_in_use" : "200716","mem_in_use_delta" : "131072","activation_size" : "65536","node_index" : "5","mem_requested_in_use_delta" : "131072","parameter_size" : "6"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1961,"ts" :7965413,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1980,"ts" :7965402,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :16,"ts" :7973826,"ph" : "X","name" :"normalized_x_offset_65536_QuantizeLinear_kernel_time","args" : {"input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_arena_held" : "1048588","node_index" : "0","op_name" : "QuantizeLinear","output_size" : "4096","mem_in_use_delta" : "0","parameter_size" : "6","mem_in_use_peak" : "200716","activation_size" : "8192","output_type_shape" : [{"uint16":[1,2048]}],"mem_requested_in_use_delta" : "0","mem_bytes_requested_in_use" : "69644","provider" : "CPUExecutionProvider","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"mem_arena_held_delta" : "0","mem_bytes_in_use" : "69644"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1856,"ts" :7973846,"ph" : "X","name" :"vitisai_vitisai_4170391121016113593_0_0_kernel_time","args" : {"mem_arena_held_delta" : "0","provider" : "vitisai","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"mem_in_use_delta" : "0","mem_bytes_requested_in_use" : "69644","output_size" : "65536","output_type_shape" : [{"uint16":[1,32768]}],"parameter_size" : "0","op_name" : "vitisai_4170391121016113593_0","activation_size" : "4096","mem_bytes_in_use" : "69644","node_index" : "6","mem_in_use_peak" : "200716","input_type_shape" : [{"uint16":[1,2048]}],"mem_arena_held" : "1048588","mem_requested_in_use_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :17,"ts" :7975706,"ph" : "X","name" :"logits_offset_65536_DequantizeLinear_kernel_time","args" : {"mem_in_use_peak" : "200716","node_index" : "5","mem_bytes_requested_in_use" : "200716","activation_size" : "65536","mem_bytes_in_use" : "200716","op_name" : "DequantizeLinear","mem_arena_held" : "1048588","mem_in_use_delta" : "131072","mem_arena_held_delta" : "0","output_type_shape" : [{"float":[1,32768]}],"parameter_size" : "6","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"provider" : "CPUExecutionProvider","output_size" : "131072","mem_requested_in_use_delta" : "131072"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1901,"ts" :7973825,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1912,"ts" :7973817,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :37,"ts" :7982123,"ph" : "X","name" :"normalized_x_offset_65536_QuantizeLinear_kernel_time","args" : {"mem_bytes_requested_in_use" : "69644","mem_arena_held" : "1048588","output_type_shape" : [{"uint16":[1,2048]}],"op_name" : "QuantizeLinear","parameter_size" : "6","output_size" : "4096","mem_requested_in_use_delta" : "0","mem_arena_held_delta" : "0","mem_in_use_peak" : "200716","node_index" : "0","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"activation_size" : "8192","mem_bytes_in_use" : "69644","provider" : "CPUExecutionProvider","mem_in_use_delta" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}}}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1985,"ts" :7982175,"ph" : "X","name" :"vitisai_vitisai_4170391121016113593_0_0_kernel_time","args" : {"mem_arena_held_delta" : "0","mem_arena_held" : "1048588","op_name" : "vitisai_4170391121016113593_0","provider" : "vitisai","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"output_type_shape" : [{"uint16":[1,32768]}],"mem_bytes_in_use" : "69644","node_index" : "6","mem_bytes_requested_in_use" : "69644","activation_size" : "4096","parameter_size" : "0","output_size" : "65536","mem_requested_in_use_delta" : "0","mem_in_use_delta" : "0","input_type_shape" : [{"uint16":[1,2048]}],"mem_in_use_peak" : "200716"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :27,"ts" :7984170,"ph" : "X","name" :"logits_offset_65536_DequantizeLinear_kernel_time","args" : {"mem_in_use_delta" : "131072","provider" : "CPUExecutionProvider","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"mem_requested_in_use_delta" : "131072","mem_bytes_in_use" : "200716","node_index" : "5","mem_in_use_peak" : "200716","mem_arena_held" : "1048588","output_type_shape" : [{"float":[1,32768]}],"mem_arena_held_delta" : "0","output_size" : "131072","parameter_size" : "6","activation_size" : "65536","op_name" : "DequantizeLinear","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_bytes_requested_in_use" : "200716"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2081,"ts" :7982120,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2116,"ts" :7982097,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :41,"ts" :7991016,"ph" : "X","name" :"normalized_x_offset_65536_QuantizeLinear_kernel_time","args" : {"provider" : "CPUExecutionProvider","mem_bytes_requested_in_use" : "69644","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_in_use_peak" : "200716","mem_requested_in_use_delta" : "0","mem_arena_held" : "1048588","mem_arena_held_delta" : "0","mem_in_use_delta" : "0","activation_size" : "8192","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"op_name" : "QuantizeLinear","node_index" : "0","output_size" : "4096","output_type_shape" : [{"uint16":[1,2048]}],"parameter_size" : "6","mem_bytes_in_use" : "69644"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1948,"ts" :7991063,"ph" : "X","name" :"vitisai_vitisai_4170391121016113593_0_0_kernel_time","args" : {"mem_arena_held" : "1048588","mem_in_use_delta" : "0","activation_size" : "4096","mem_requested_in_use_delta" : "0","mem_bytes_in_use" : "69644","input_type_shape" : [{"uint16":[1,2048]}],"node_index" : "6","output_size" : "65536","mem_in_use_peak" : "200716","mem_arena_held_delta" : "0","provider" : "vitisai","parameter_size" : "0","output_type_shape" : [{"uint16":[1,32768]}],"mem_bytes_requested_in_use" : "69644","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"op_name" : "vitisai_4170391121016113593_0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :21,"ts" :7993016,"ph" : "X","name" :"logits_offset_65536_DequantizeLinear_kernel_time","args" : {"mem_arena_held" : "1048588","op_name" : "DequantizeLinear","activation_size" : "65536","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_in_use_delta" : "131072","mem_bytes_in_use" : "200716","mem_in_use_peak" : "200716","provider" : "CPUExecutionProvider","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"mem_bytes_requested_in_use" : "200716","mem_arena_held_delta" : "0","parameter_size" : "6","output_type_shape" : [{"float":[1,32768]}],"output_size" : "131072","mem_requested_in_use_delta" : "131072","node_index" : "5"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2027,"ts" :7991014,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2051,"ts" :7990996,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :81,"ts" :7999330,"ph" : "X","name" :"normalized_x_offset_65536_QuantizeLinear_kernel_time","args" : {"output_type_shape" : [{"uint16":[1,2048]}],"mem_in_use_peak" : "200716","mem_in_use_delta" : "0","node_index" : "0","mem_bytes_requested_in_use" : "69644","mem_bytes_in_use" : "69644","activation_size" : "8192","mem_arena_held" : "1048588","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"output_size" : "4096","op_name" : "QuantizeLinear","mem_arena_held_delta" : "0","mem_requested_in_use_delta" : "0","provider" : "CPUExecutionProvider","parameter_size" : "6","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}}}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1907,"ts" :7999418,"ph" : "X","name" :"vitisai_vitisai_4170391121016113593_0_0_kernel_time","args" : {"node_index" : "6","parameter_size" : "0","mem_arena_held_delta" : "0","mem_in_use_delta" : "0","output_type_shape" : [{"uint16":[1,32768]}],"mem_requested_in_use_delta" : "0","mem_bytes_requested_in_use" : "69644","mem_arena_held" : "1048588","output_size" : "65536","op_name" : "vitisai_4170391121016113593_0","activation_size" : "4096","mem_bytes_in_use" : "69644","mem_in_use_peak" : "200716","provider" : "vitisai","input_type_shape" : [{"uint16":[1,2048]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}}}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :21,"ts" :8001330,"ph" : "X","name" :"logits_offset_65536_DequantizeLinear_kernel_time","args" : {"parameter_size" : "6","provider" : "CPUExecutionProvider","mem_bytes_requested_in_use" : "200716","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"activation_size" : "65536","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"mem_requested_in_use_delta" : "131072","output_size" : "131072","node_index" : "5","mem_bytes_in_use" : "200716","output_type_shape" : [{"float":[1,32768]}],"mem_in_use_delta" : "131072","op_name" : "DequantizeLinear","mem_in_use_peak" : "200716","mem_arena_held" : "1048588","mem_arena_held_delta" : "0"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2028,"ts" :7999327,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2053,"ts" :7999308,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :14,"ts" :8007430,"ph" : "X","name" :"normalized_x_offset_65536_QuantizeLinear_kernel_time","args" : {"output_type_shape" : [{"uint16":[1,2048]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"mem_bytes_requested_in_use" : "69644","parameter_size" : "6","mem_requested_in_use_delta" : "0","mem_in_use_delta" : "0","mem_arena_held_delta" : "0","provider" : "CPUExecutionProvider","output_size" : "4096","mem_bytes_in_use" : "69644","node_index" : "0","mem_arena_held" : "1048588","op_name" : "QuantizeLinear","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_in_use_peak" : "200716","activation_size" : "8192"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1858,"ts" :8007447,"ph" : "X","name" :"vitisai_vitisai_4170391121016113593_0_0_kernel_time","args" : {"activation_size" : "4096","mem_requested_in_use_delta" : "0","mem_in_use_delta" : "0","mem_bytes_in_use" : "69644","op_name" : "vitisai_4170391121016113593_0","output_size" : "65536","mem_arena_held" : "1048588","node_index" : "6","input_type_shape" : [{"uint16":[1,2048]}],"parameter_size" : "0","mem_in_use_peak" : "200716","output_type_shape" : [{"uint16":[1,32768]}],"provider" : "vitisai","mem_arena_held_delta" : "0","mem_bytes_requested_in_use" : "69644","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}}}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :18,"ts" :8009309,"ph" : "X","name" :"logits_offset_65536_DequantizeLinear_kernel_time","args" : {"provider" : "CPUExecutionProvider","output_size" : "131072","mem_arena_held" : "1048588","output_type_shape" : [{"float":[1,32768]}],"parameter_size" : "6","mem_in_use_peak" : "200716","mem_bytes_requested_in_use" : "200716","node_index" : "5","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-4-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"25016": {"num_run": 0, "core": -1},"52032": {"num_run": 0, "core": -1},"61420": {"num_run": 0, "core": -1},"58320": {"num_run": 0, "core": -1},"42856": {"num_run": 0, "core": -1},"34812": {"num_run": 0, "core": -1},"39652": {"num_run": 0, "core": -1},"58800": {"num_run": 0, "core": -1},"53560": {"num_run": 0, "core": -1},"16848": {"num_run": 0, "core": -1},"50844": {"num_run": 0, "core": -1}}},"mem_in_use_delta" : "131072","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_arena_held_delta" : "0","activation_size" : "65536","mem_bytes_in_use" : "200716","op_name" : "DequantizeLinear","mem_requested_in_use_delta" : "131072"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1901,"ts" :8007429,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1911,"ts" :8007423,"ph" : "X","name" :"model_run","args" : {}} +] diff --git a/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/shards/shard_98304.json b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/shards/shard_98304.json new file mode 100644 index 000000000..36c079f9b --- /dev/null +++ b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/shards/shard_98304.json @@ -0,0 +1,95 @@ +{ + "scope": "Spark-X2.5-1.7B output head only; no prefill/decode/KV/sampling or full generation", + "vocab_shard_size": 32768, + "vocab_shard_offset": 98304, + "flatten_input": true, + "batch_repeats": 1, + "matmul_only": true, + "hardware": "AMD Ryzen AI 9 HX 370 NPU", + "os": "Windows-11-10.0.26200-SP0", + "onnx_version": "1.22.0", + "onnxruntime_version": "1.30.0", + "source_model_sha256": "f5f0c31eb438e9940ab242da8bf6f0f5dad6ebc9f6ef02d5475c21b744e595bc", + "prepared_model_sha256": "f229fd37eb0dc4cab24ce26e6cabe95d4ab3a88b8403891e11b431a015eec4af", + "calibration_sha256": "cfa2ed8439f8fe01730319667bfd01abf2e39cf5851490346a0331304cf63194", + "opset21_sha256": "f229fd37eb0dc4cab24ce26e6cabe95d4ab3a88b8403891e11b431a015eec4af", + "quantized_sha256": "0ff8e1e6d513d5debda26fa7c7088db80085f637e4ae8ddda24a505ac80c64f7", + "quantization": "opset 21, MatMul only, QDQ, QUInt16 activations/QInt8 weights, MinMax on four real activations", + "input_shape": [ + 1, + 2048 + ], + "output_shape": [ + 1, + 32768 + ], + "source_cpu_reference": [ + { + "snr_db": 137.32030186574386, + "max_abs_diff": 1.9073486328125e-05, + "top1_match": true, + "finite": true + }, + { + "snr_db": 134.3119760585634, + "max_abs_diff": 2.6702880859375e-05, + "top1_match": true, + "finite": true + }, + { + "snr_db": 137.71005859776187, + "max_abs_diff": 2.288818359375e-05, + "top1_match": true, + "finite": true + }, + { + "snr_db": 136.40835150657765, + "max_abs_diff": 2.288818359375e-05, + "top1_match": true, + "finite": true + } + ], + "npu_devices": 1, + "session_providers": [ + "vitisai", + "CPUExecutionProvider" + ], + "npu_node_events": 4, + "cpu_node_events": 8, + "load_s": 3.928793099941686, + "npu_rows": [ + { + "snr_db": 56.98918166146055, + "max_abs_diff": 0.22197341918945312, + "top1_match": true, + "finite": true, + "index": 0, + "wall_s": 0.0038848000112921 + }, + { + "snr_db": 56.590884754816166, + "max_abs_diff": 0.21759796142578125, + "top1_match": true, + "finite": true, + "index": 1, + "wall_s": 0.0025273000355809927 + }, + { + "snr_db": 56.59743918512855, + "max_abs_diff": 0.25970458984375, + "top1_match": true, + "finite": true, + "index": 2, + "wall_s": 0.002280400018207729 + }, + { + "snr_db": 55.578637242475644, + "max_abs_diff": 0.2147674560546875, + "top1_match": true, + "finite": true, + "index": 3, + "wall_s": 0.002343800035305321 + } + ], + "profile_file": "C:\\Users\\zhout\\w2\\spark_npu_full_shards_named_20260923\\shard_98304\\spark_lm_head_npu_profile_2026-09-23_18-08-24_412.json" +} diff --git a/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/shards/shard_98304_co_resident_profile.json b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/shards/shard_98304_co_resident_profile.json new file mode 100644 index 000000000..fd360939d --- /dev/null +++ b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/shards/shard_98304_co_resident_profile.json @@ -0,0 +1,129 @@ +[ +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :31112,"ts" :12,"ph" : "X","name" :"model_loading_uri","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :3683060,"ts" :32425,"ph" : "X","name" :"session_initialization","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :41,"ts" :3723598,"ph" : "X","name" :"normalized_x_offset_98304_QuantizeLinear_kernel_time","args" : {"output_size" : "4096","provider" : "CPUExecutionProvider","mem_arena_held" : "1048588","mem_arena_held_delta" : "1048576","activation_size" : "8192","mem_bytes_in_use" : "4108","op_name" : "QuantizeLinear","parameter_size" : "6","node_index" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"mem_requested_in_use_delta" : "4096","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_in_use_delta" : "4096","output_type_shape" : [{"uint16":[1,2048]}],"mem_bytes_requested_in_use" : "4108","mem_in_use_peak" : "4108"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :2393,"ts" :3723646,"ph" : "X","name" :"vitisai_vitisai_15891435282080274764_0_0_kernel_time","args" : {"parameter_size" : "0","mem_in_use_delta" : "65536","mem_requested_in_use_delta" : "65536","activation_size" : "4096","provider" : "vitisai","node_index" : "6","op_name" : "vitisai_15891435282080274764_0","mem_bytes_in_use" : "69644","input_type_shape" : [{"uint16":[1,2048]}],"output_size" : "65536","mem_arena_held" : "1048588","mem_in_use_peak" : "69644","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"output_type_shape" : [{"uint16":[1,32768]}],"mem_bytes_requested_in_use" : "69644","mem_arena_held_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :55,"ts" :3726081,"ph" : "X","name" :"logits_offset_98304_DequantizeLinear_kernel_time","args" : {"activation_size" : "65536","provider" : "CPUExecutionProvider","op_name" : "DequantizeLinear","parameter_size" : "6","mem_in_use_peak" : "196620","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"output_type_shape" : [{"float":[1,32768]}],"mem_bytes_requested_in_use" : "196620","mem_requested_in_use_delta" : "131072","mem_arena_held_delta" : "0","mem_bytes_in_use" : "196620","mem_in_use_delta" : "131072","mem_arena_held" : "1048588","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"node_index" : "5","output_size" : "131072"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2548,"ts" :3723596,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2569,"ts" :3723587,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :15,"ts" :3735362,"ph" : "X","name" :"normalized_x_offset_98304_QuantizeLinear_kernel_time","args" : {"activation_size" : "8192","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"output_size" : "4096","mem_requested_in_use_delta" : "0","output_type_shape" : [{"uint16":[1,2048]}],"provider" : "CPUExecutionProvider","mem_arena_held_delta" : "0","op_name" : "QuantizeLinear","mem_bytes_in_use" : "69644","mem_arena_held" : "1048588","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"mem_in_use_peak" : "196620","mem_in_use_delta" : "0","parameter_size" : "6","mem_bytes_requested_in_use" : "69644","node_index" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :2071,"ts" :3735381,"ph" : "X","name" :"vitisai_vitisai_15891435282080274764_0_0_kernel_time","args" : {"mem_in_use_delta" : "0","mem_requested_in_use_delta" : "0","mem_arena_held_delta" : "0","op_name" : "vitisai_15891435282080274764_0","parameter_size" : "0","input_type_shape" : [{"uint16":[1,2048]}],"mem_in_use_peak" : "196620","mem_arena_held" : "1048588","node_index" : "6","activation_size" : "4096","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"output_size" : "65536","output_type_shape" : [{"uint16":[1,32768]}],"mem_bytes_in_use" : "69644","provider" : "vitisai","mem_bytes_requested_in_use" : "69644"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :59,"ts" :3737468,"ph" : "X","name" :"logits_offset_98304_DequantizeLinear_kernel_time","args" : {"provider" : "CPUExecutionProvider","mem_bytes_requested_in_use" : "200716","mem_arena_held_delta" : "0","activation_size" : "65536","mem_in_use_peak" : "200716","mem_arena_held" : "1048588","mem_requested_in_use_delta" : "131072","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"node_index" : "5","output_type_shape" : [{"float":[1,32768]}],"output_size" : "131072","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"mem_in_use_delta" : "131072","parameter_size" : "6","op_name" : "DequantizeLinear","mem_bytes_in_use" : "200716"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2174,"ts" :3735361,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2202,"ts" :3735351,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :15,"ts" :3747702,"ph" : "X","name" :"normalized_x_offset_98304_QuantizeLinear_kernel_time","args" : {"activation_size" : "8192","mem_requested_in_use_delta" : "0","output_size" : "4096","mem_bytes_requested_in_use" : "69644","node_index" : "0","mem_in_use_peak" : "200716","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"op_name" : "QuantizeLinear","mem_arena_held_delta" : "0","mem_bytes_in_use" : "69644","provider" : "CPUExecutionProvider","output_type_shape" : [{"uint16":[1,2048]}],"parameter_size" : "6","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_in_use_delta" : "0","mem_arena_held" : "1048588"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1979,"ts" :3747721,"ph" : "X","name" :"vitisai_vitisai_15891435282080274764_0_0_kernel_time","args" : {"output_type_shape" : [{"uint16":[1,32768]}],"node_index" : "6","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"mem_arena_held_delta" : "0","activation_size" : "4096","input_type_shape" : [{"uint16":[1,2048]}],"mem_bytes_requested_in_use" : "69644","mem_arena_held" : "1048588","mem_requested_in_use_delta" : "0","parameter_size" : "0","output_size" : "65536","provider" : "vitisai","op_name" : "vitisai_15891435282080274764_0","mem_in_use_peak" : "200716","mem_in_use_delta" : "0","mem_bytes_in_use" : "69644"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :52,"ts" :3749717,"ph" : "X","name" :"logits_offset_98304_DequantizeLinear_kernel_time","args" : {"mem_requested_in_use_delta" : "131072","parameter_size" : "6","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_bytes_in_use" : "200716","mem_arena_held" : "1048588","node_index" : "5","op_name" : "DequantizeLinear","output_type_shape" : [{"float":[1,32768]}],"mem_in_use_delta" : "131072","mem_arena_held_delta" : "0","activation_size" : "65536","mem_bytes_requested_in_use" : "200716","provider" : "CPUExecutionProvider","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"mem_in_use_peak" : "200716","output_size" : "131072"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2074,"ts" :3747700,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2102,"ts" :3747690,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :25,"ts" :3758724,"ph" : "X","name" :"normalized_x_offset_98304_QuantizeLinear_kernel_time","args" : {"output_type_shape" : [{"uint16":[1,2048]}],"parameter_size" : "6","mem_bytes_requested_in_use" : "69644","provider" : "CPUExecutionProvider","op_name" : "QuantizeLinear","mem_requested_in_use_delta" : "0","activation_size" : "8192","mem_bytes_in_use" : "69644","mem_arena_held_delta" : "0","output_size" : "4096","mem_in_use_peak" : "200716","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"mem_arena_held" : "1048588","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_in_use_delta" : "0","node_index" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1927,"ts" :3758754,"ph" : "X","name" :"vitisai_vitisai_15891435282080274764_0_0_kernel_time","args" : {"output_type_shape" : [{"uint16":[1,32768]}],"op_name" : "vitisai_15891435282080274764_0","node_index" : "6","activation_size" : "4096","parameter_size" : "0","mem_in_use_delta" : "0","input_type_shape" : [{"uint16":[1,2048]}],"provider" : "vitisai","mem_bytes_requested_in_use" : "69644","mem_in_use_peak" : "200716","mem_requested_in_use_delta" : "0","output_size" : "65536","mem_bytes_in_use" : "69644","mem_arena_held" : "1048588","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"mem_arena_held_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :24,"ts" :3760689,"ph" : "X","name" :"logits_offset_98304_DequantizeLinear_kernel_time","args" : {"node_index" : "5","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"activation_size" : "65536","mem_arena_held" : "1048588","mem_bytes_requested_in_use" : "200716","output_size" : "131072","output_type_shape" : [{"float":[1,32768]}],"parameter_size" : "6","provider" : "CPUExecutionProvider","mem_requested_in_use_delta" : "131072","op_name" : "DequantizeLinear","mem_bytes_in_use" : "200716","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"mem_in_use_peak" : "200716","mem_arena_held_delta" : "0","mem_in_use_delta" : "131072"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1997,"ts" :3758722,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2019,"ts" :3758709,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :13,"ts" :3769684,"ph" : "X","name" :"normalized_x_offset_98304_QuantizeLinear_kernel_time","args" : {"mem_in_use_delta" : "0","provider" : "CPUExecutionProvider","mem_arena_held" : "1048588","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"mem_requested_in_use_delta" : "0","mem_arena_held_delta" : "0","op_name" : "QuantizeLinear","mem_in_use_peak" : "200716","mem_bytes_requested_in_use" : "69644","mem_bytes_in_use" : "69644","output_type_shape" : [{"uint16":[1,2048]}],"input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"output_size" : "4096","parameter_size" : "6","activation_size" : "8192","node_index" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1857,"ts" :3769700,"ph" : "X","name" :"vitisai_vitisai_15891435282080274764_0_0_kernel_time","args" : {"activation_size" : "4096","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"mem_in_use_delta" : "0","node_index" : "6","op_name" : "vitisai_15891435282080274764_0","mem_in_use_peak" : "200716","output_size" : "65536","mem_bytes_in_use" : "69644","provider" : "vitisai","output_type_shape" : [{"uint16":[1,32768]}],"mem_bytes_requested_in_use" : "69644","mem_arena_held" : "1048588","input_type_shape" : [{"uint16":[1,2048]}],"mem_arena_held_delta" : "0","mem_requested_in_use_delta" : "0","parameter_size" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :21,"ts" :3771564,"ph" : "X","name" :"logits_offset_98304_DequantizeLinear_kernel_time","args" : {"input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"op_name" : "DequantizeLinear","activation_size" : "65536","output_size" : "131072","mem_bytes_requested_in_use" : "200716","mem_arena_held" : "1048588","parameter_size" : "6","provider" : "CPUExecutionProvider","node_index" : "5","mem_bytes_in_use" : "200716","output_type_shape" : [{"float":[1,32768]}],"mem_in_use_peak" : "200716","mem_in_use_delta" : "131072","mem_requested_in_use_delta" : "131072","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"mem_arena_held_delta" : "0"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1905,"ts" :3769683,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1919,"ts" :3769676,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :15,"ts" :3777943,"ph" : "X","name" :"normalized_x_offset_98304_QuantizeLinear_kernel_time","args" : {"output_size" : "4096","mem_bytes_requested_in_use" : "69644","op_name" : "QuantizeLinear","mem_arena_held" : "1048588","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"output_type_shape" : [{"uint16":[1,2048]}],"provider" : "CPUExecutionProvider","mem_in_use_peak" : "200716","mem_arena_held_delta" : "0","mem_bytes_in_use" : "69644","activation_size" : "8192","parameter_size" : "6","node_index" : "0","mem_requested_in_use_delta" : "0","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_in_use_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1853,"ts" :3777963,"ph" : "X","name" :"vitisai_vitisai_15891435282080274764_0_0_kernel_time","args" : {"mem_in_use_peak" : "200716","mem_arena_held" : "1048588","input_type_shape" : [{"uint16":[1,2048]}],"mem_arena_held_delta" : "0","mem_bytes_requested_in_use" : "69644","parameter_size" : "0","mem_in_use_delta" : "0","output_size" : "65536","provider" : "vitisai","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"activation_size" : "4096","mem_bytes_in_use" : "69644","output_type_shape" : [{"uint16":[1,32768]}],"op_name" : "vitisai_15891435282080274764_0","mem_requested_in_use_delta" : "0","node_index" : "6"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :17,"ts" :3779821,"ph" : "X","name" :"logits_offset_98304_DequantizeLinear_kernel_time","args" : {"mem_in_use_delta" : "131072","mem_in_use_peak" : "200716","mem_bytes_in_use" : "200716","mem_bytes_requested_in_use" : "200716","output_size" : "131072","mem_arena_held_delta" : "0","provider" : "CPUExecutionProvider","node_index" : "5","output_type_shape" : [{"float":[1,32768]}],"mem_arena_held" : "1048588","parameter_size" : "6","activation_size" : "65536","mem_requested_in_use_delta" : "131072","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"op_name" : "DequantizeLinear","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}}}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1900,"ts" :3777941,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1912,"ts" :3777932,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :12,"ts" :3785752,"ph" : "X","name" :"normalized_x_offset_98304_QuantizeLinear_kernel_time","args" : {"node_index" : "0","output_size" : "4096","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"mem_arena_held_delta" : "0","provider" : "CPUExecutionProvider","activation_size" : "8192","output_type_shape" : [{"uint16":[1,2048]}],"mem_arena_held" : "1048588","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_bytes_requested_in_use" : "69644","mem_in_use_delta" : "0","mem_requested_in_use_delta" : "0","mem_bytes_in_use" : "69644","parameter_size" : "6","mem_in_use_peak" : "200716","op_name" : "QuantizeLinear"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1865,"ts" :3785766,"ph" : "X","name" :"vitisai_vitisai_15891435282080274764_0_0_kernel_time","args" : {"mem_in_use_peak" : "200716","mem_arena_held" : "1048588","op_name" : "vitisai_15891435282080274764_0","activation_size" : "4096","mem_bytes_requested_in_use" : "69644","mem_in_use_delta" : "0","provider" : "vitisai","output_type_shape" : [{"uint16":[1,32768]}],"mem_bytes_in_use" : "69644","input_type_shape" : [{"uint16":[1,2048]}],"output_size" : "65536","mem_arena_held_delta" : "0","mem_requested_in_use_delta" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"parameter_size" : "0","node_index" : "6"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :19,"ts" :3787636,"ph" : "X","name" :"logits_offset_98304_DequantizeLinear_kernel_time","args" : {"mem_bytes_requested_in_use" : "200716","activation_size" : "65536","mem_arena_held_delta" : "0","mem_in_use_peak" : "200716","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"provider" : "CPUExecutionProvider","op_name" : "DequantizeLinear","mem_bytes_in_use" : "200716","node_index" : "5","output_type_shape" : [{"float":[1,32768]}],"mem_in_use_delta" : "131072","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"parameter_size" : "6","mem_requested_in_use_delta" : "131072","output_size" : "131072","mem_arena_held" : "1048588"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1907,"ts" :3785751,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1916,"ts" :3785748,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :56,"ts" :3793985,"ph" : "X","name" :"normalized_x_offset_98304_QuantizeLinear_kernel_time","args" : {"parameter_size" : "6","output_type_shape" : [{"uint16":[1,2048]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"mem_in_use_peak" : "200716","mem_requested_in_use_delta" : "0","mem_arena_held_delta" : "0","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"op_name" : "QuantizeLinear","mem_bytes_requested_in_use" : "69644","mem_arena_held" : "1048588","output_size" : "4096","provider" : "CPUExecutionProvider","mem_in_use_delta" : "0","mem_bytes_in_use" : "69644","node_index" : "0","activation_size" : "8192"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1974,"ts" :3794050,"ph" : "X","name" :"vitisai_vitisai_15891435282080274764_0_0_kernel_time","args" : {"mem_requested_in_use_delta" : "0","activation_size" : "4096","provider" : "vitisai","mem_in_use_peak" : "200716","mem_arena_held" : "1048588","mem_bytes_in_use" : "69644","mem_in_use_delta" : "0","output_size" : "65536","output_type_shape" : [{"uint16":[1,32768]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"node_index" : "6","input_type_shape" : [{"uint16":[1,2048]}],"mem_bytes_requested_in_use" : "69644","parameter_size" : "0","mem_arena_held_delta" : "0","op_name" : "vitisai_15891435282080274764_0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :21,"ts" :3796031,"ph" : "X","name" :"logits_offset_98304_DequantizeLinear_kernel_time","args" : {"activation_size" : "65536","mem_in_use_delta" : "131072","op_name" : "DequantizeLinear","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"provider" : "CPUExecutionProvider","mem_requested_in_use_delta" : "131072","mem_in_use_peak" : "200716","node_index" : "5","output_size" : "131072","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_arena_held_delta" : "0","mem_bytes_in_use" : "200716","parameter_size" : "6","mem_bytes_requested_in_use" : "200716","output_type_shape" : [{"float":[1,32768]}],"mem_arena_held" : "1048588"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2072,"ts" :3793984,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2101,"ts" :3793962,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :14,"ts" :3802216,"ph" : "X","name" :"normalized_x_offset_98304_QuantizeLinear_kernel_time","args" : {"mem_requested_in_use_delta" : "0","mem_in_use_delta" : "0","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"provider" : "CPUExecutionProvider","mem_in_use_peak" : "200716","activation_size" : "8192","mem_arena_held_delta" : "0","mem_bytes_in_use" : "69644","output_type_shape" : [{"uint16":[1,2048]}],"mem_bytes_requested_in_use" : "69644","parameter_size" : "6","node_index" : "0","mem_arena_held" : "1048588","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"op_name" : "QuantizeLinear","output_size" : "4096"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1835,"ts" :3802237,"ph" : "X","name" :"vitisai_vitisai_15891435282080274764_0_0_kernel_time","args" : {"mem_bytes_requested_in_use" : "69644","parameter_size" : "0","output_size" : "65536","mem_arena_held" : "1048588","input_type_shape" : [{"uint16":[1,2048]}],"node_index" : "6","op_name" : "vitisai_15891435282080274764_0","mem_bytes_in_use" : "69644","activation_size" : "4096","output_type_shape" : [{"uint16":[1,32768]}],"mem_in_use_delta" : "0","mem_in_use_peak" : "200716","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"mem_requested_in_use_delta" : "0","provider" : "vitisai","mem_arena_held_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :18,"ts" :3804076,"ph" : "X","name" :"logits_offset_98304_DequantizeLinear_kernel_time","args" : {"input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_arena_held_delta" : "0","mem_arena_held" : "1048588","provider" : "CPUExecutionProvider","op_name" : "DequantizeLinear","mem_in_use_delta" : "131072","mem_bytes_in_use" : "200716","node_index" : "5","output_type_shape" : [{"float":[1,32768]}],"mem_bytes_requested_in_use" : "200716","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"activation_size" : "65536","output_size" : "131072","mem_in_use_peak" : "200716","parameter_size" : "6","mem_requested_in_use_delta" : "131072"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1880,"ts" :3802215,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1891,"ts" :3802208,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :19,"ts" :3810533,"ph" : "X","name" :"normalized_x_offset_98304_QuantizeLinear_kernel_time","args" : {"activation_size" : "8192","mem_bytes_requested_in_use" : "69644","node_index" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"mem_requested_in_use_delta" : "0","mem_in_use_peak" : "200716","parameter_size" : "6","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"output_type_shape" : [{"uint16":[1,2048]}],"mem_arena_held_delta" : "0","op_name" : "QuantizeLinear","output_size" : "4096","provider" : "CPUExecutionProvider","mem_in_use_delta" : "0","mem_arena_held" : "1048588","mem_bytes_in_use" : "69644"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1921,"ts" :3810556,"ph" : "X","name" :"vitisai_vitisai_15891435282080274764_0_0_kernel_time","args" : {"mem_bytes_requested_in_use" : "69644","mem_in_use_peak" : "200716","mem_in_use_delta" : "0","parameter_size" : "0","activation_size" : "4096","input_type_shape" : [{"uint16":[1,2048]}],"provider" : "vitisai","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"op_name" : "vitisai_15891435282080274764_0","mem_requested_in_use_delta" : "0","mem_arena_held" : "1048588","node_index" : "6","mem_arena_held_delta" : "0","output_size" : "65536","output_type_shape" : [{"uint16":[1,32768]}],"mem_bytes_in_use" : "69644"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :22,"ts" :3812483,"ph" : "X","name" :"logits_offset_98304_DequantizeLinear_kernel_time","args" : {"provider" : "CPUExecutionProvider","output_size" : "131072","mem_bytes_requested_in_use" : "200716","mem_in_use_peak" : "200716","activation_size" : "65536","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"op_name" : "DequantizeLinear","parameter_size" : "6","mem_arena_held" : "1048588","node_index" : "5","mem_bytes_in_use" : "200716","mem_arena_held_delta" : "0","mem_requested_in_use_delta" : "131072","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_in_use_delta" : "131072","output_type_shape" : [{"float":[1,32768]}]}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1976,"ts" :3810531,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1996,"ts" :3810519,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :13,"ts" :3818613,"ph" : "X","name" :"normalized_x_offset_98304_QuantizeLinear_kernel_time","args" : {"mem_arena_held_delta" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"op_name" : "QuantizeLinear","output_size" : "4096","mem_in_use_peak" : "200716","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"output_type_shape" : [{"uint16":[1,2048]}],"mem_in_use_delta" : "0","mem_requested_in_use_delta" : "0","mem_bytes_requested_in_use" : "69644","node_index" : "0","provider" : "CPUExecutionProvider","mem_arena_held" : "1048588","parameter_size" : "6","activation_size" : "8192","mem_bytes_in_use" : "69644"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1879,"ts" :3818630,"ph" : "X","name" :"vitisai_vitisai_15891435282080274764_0_0_kernel_time","args" : {"mem_arena_held_delta" : "0","provider" : "vitisai","output_type_shape" : [{"uint16":[1,32768]}],"mem_bytes_requested_in_use" : "69644","mem_bytes_in_use" : "69644","mem_in_use_peak" : "200716","input_type_shape" : [{"uint16":[1,2048]}],"node_index" : "6","output_size" : "65536","mem_in_use_delta" : "0","mem_arena_held" : "1048588","parameter_size" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"mem_requested_in_use_delta" : "0","activation_size" : "4096","op_name" : "vitisai_15891435282080274764_0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :15,"ts" :3820512,"ph" : "X","name" :"logits_offset_98304_DequantizeLinear_kernel_time","args" : {"activation_size" : "65536","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_in_use_delta" : "131072","mem_requested_in_use_delta" : "131072","node_index" : "5","provider" : "CPUExecutionProvider","parameter_size" : "6","mem_in_use_peak" : "200716","mem_arena_held" : "1048588","mem_arena_held_delta" : "0","mem_bytes_requested_in_use" : "200716","op_name" : "DequantizeLinear","output_size" : "131072","output_type_shape" : [{"float":[1,32768]}],"mem_bytes_in_use" : "200716"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1917,"ts" :3818612,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1927,"ts" :3818606,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :19,"ts" :3826629,"ph" : "X","name" :"normalized_x_offset_98304_QuantizeLinear_kernel_time","args" : {"mem_arena_held_delta" : "0","mem_in_use_peak" : "200716","mem_requested_in_use_delta" : "0","node_index" : "0","mem_arena_held" : "1048588","provider" : "CPUExecutionProvider","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"mem_bytes_requested_in_use" : "69644","mem_bytes_in_use" : "69644","parameter_size" : "6","output_size" : "4096","output_type_shape" : [{"uint16":[1,2048]}],"mem_in_use_delta" : "0","activation_size" : "8192","op_name" : "QuantizeLinear","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}]}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1909,"ts" :3826652,"ph" : "X","name" :"vitisai_vitisai_15891435282080274764_0_0_kernel_time","args" : {"mem_bytes_in_use" : "69644","op_name" : "vitisai_15891435282080274764_0","mem_requested_in_use_delta" : "0","activation_size" : "4096","node_index" : "6","mem_in_use_delta" : "0","provider" : "vitisai","output_size" : "65536","output_type_shape" : [{"uint16":[1,32768]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"mem_arena_held_delta" : "0","mem_arena_held" : "1048588","mem_bytes_requested_in_use" : "69644","parameter_size" : "0","input_type_shape" : [{"uint16":[1,2048]}],"mem_in_use_peak" : "200716"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :21,"ts" :3828568,"ph" : "X","name" :"logits_offset_98304_DequantizeLinear_kernel_time","args" : {"mem_arena_held_delta" : "0","parameter_size" : "6","mem_bytes_requested_in_use" : "200716","activation_size" : "65536","provider" : "CPUExecutionProvider","op_name" : "DequantizeLinear","mem_arena_held" : "1048588","output_type_shape" : [{"float":[1,32768]}],"input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"mem_in_use_delta" : "131072","mem_bytes_in_use" : "200716","node_index" : "5","mem_requested_in_use_delta" : "131072","output_size" : "131072","mem_in_use_peak" : "200716"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1964,"ts" :3826628,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1982,"ts" :3826617,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :13,"ts" :3835090,"ph" : "X","name" :"normalized_x_offset_98304_QuantizeLinear_kernel_time","args" : {"input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"op_name" : "QuantizeLinear","mem_bytes_requested_in_use" : "69644","mem_requested_in_use_delta" : "0","activation_size" : "8192","output_size" : "4096","mem_arena_held_delta" : "0","mem_arena_held" : "1048588","output_type_shape" : [{"uint16":[1,2048]}],"provider" : "CPUExecutionProvider","mem_in_use_delta" : "0","mem_bytes_in_use" : "69644","mem_in_use_peak" : "200716","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"node_index" : "0","parameter_size" : "6"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1848,"ts" :3835106,"ph" : "X","name" :"vitisai_vitisai_15891435282080274764_0_0_kernel_time","args" : {"parameter_size" : "0","input_type_shape" : [{"uint16":[1,2048]}],"node_index" : "6","mem_in_use_delta" : "0","output_size" : "65536","op_name" : "vitisai_15891435282080274764_0","provider" : "vitisai","mem_arena_held_delta" : "0","mem_requested_in_use_delta" : "0","mem_bytes_in_use" : "69644","output_type_shape" : [{"uint16":[1,32768]}],"activation_size" : "4096","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"mem_bytes_requested_in_use" : "69644","mem_in_use_peak" : "200716","mem_arena_held" : "1048588"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :17,"ts" :3836959,"ph" : "X","name" :"logits_offset_98304_DequantizeLinear_kernel_time","args" : {"mem_in_use_delta" : "131072","mem_bytes_requested_in_use" : "200716","mem_arena_held_delta" : "0","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"node_index" : "5","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"output_type_shape" : [{"float":[1,32768]}],"op_name" : "DequantizeLinear","mem_requested_in_use_delta" : "131072","provider" : "CPUExecutionProvider","mem_arena_held" : "1048588","parameter_size" : "6","output_size" : "131072","activation_size" : "65536","mem_bytes_in_use" : "200716","mem_in_use_peak" : "200716"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1889,"ts" :3835089,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1898,"ts" :3835083,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :15,"ts" :3843029,"ph" : "X","name" :"normalized_x_offset_98304_QuantizeLinear_kernel_time","args" : {"mem_bytes_requested_in_use" : "69644","mem_in_use_peak" : "200716","mem_requested_in_use_delta" : "0","output_size" : "4096","mem_bytes_in_use" : "69644","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_arena_held" : "1048588","parameter_size" : "6","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"node_index" : "0","provider" : "CPUExecutionProvider","activation_size" : "8192","mem_arena_held_delta" : "0","output_type_shape" : [{"uint16":[1,2048]}],"op_name" : "QuantizeLinear","mem_in_use_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1873,"ts" :3843047,"ph" : "X","name" :"vitisai_vitisai_15891435282080274764_0_0_kernel_time","args" : {"mem_arena_held_delta" : "0","activation_size" : "4096","mem_bytes_requested_in_use" : "69644","node_index" : "6","mem_arena_held" : "1048588","parameter_size" : "0","mem_in_use_peak" : "200716","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"op_name" : "vitisai_15891435282080274764_0","mem_in_use_delta" : "0","provider" : "vitisai","output_size" : "65536","mem_requested_in_use_delta" : "0","output_type_shape" : [{"uint16":[1,32768]}],"input_type_shape" : [{"uint16":[1,2048]}],"mem_bytes_in_use" : "69644"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :20,"ts" :3844926,"ph" : "X","name" :"logits_offset_98304_DequantizeLinear_kernel_time","args" : {"mem_bytes_requested_in_use" : "200716","node_index" : "5","output_size" : "131072","provider" : "CPUExecutionProvider","mem_in_use_delta" : "131072","mem_bytes_in_use" : "200716","mem_arena_held_delta" : "0","parameter_size" : "6","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"op_name" : "DequantizeLinear","output_type_shape" : [{"float":[1,32768]}],"activation_size" : "65536","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"mem_requested_in_use_delta" : "131072","mem_arena_held" : "1048588","mem_in_use_peak" : "200716"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1922,"ts" :3843028,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1937,"ts" :3843020,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :14,"ts" :3851093,"ph" : "X","name" :"normalized_x_offset_98304_QuantizeLinear_kernel_time","args" : {"mem_bytes_in_use" : "69644","mem_in_use_peak" : "200716","mem_bytes_requested_in_use" : "69644","node_index" : "0","mem_requested_in_use_delta" : "0","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_arena_held_delta" : "0","op_name" : "QuantizeLinear","activation_size" : "8192","mem_arena_held" : "1048588","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"output_type_shape" : [{"uint16":[1,2048]}],"provider" : "CPUExecutionProvider","mem_in_use_delta" : "0","parameter_size" : "6","output_size" : "4096"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1867,"ts" :3851112,"ph" : "X","name" :"vitisai_vitisai_15891435282080274764_0_0_kernel_time","args" : {"mem_bytes_in_use" : "69644","input_type_shape" : [{"uint16":[1,2048]}],"output_type_shape" : [{"uint16":[1,32768]}],"op_name" : "vitisai_15891435282080274764_0","parameter_size" : "0","mem_in_use_peak" : "200716","activation_size" : "4096","mem_arena_held" : "1048588","provider" : "vitisai","mem_bytes_requested_in_use" : "69644","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"mem_requested_in_use_delta" : "0","mem_in_use_delta" : "0","node_index" : "6","output_size" : "65536","mem_arena_held_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :21,"ts" :3852985,"ph" : "X","name" :"logits_offset_98304_DequantizeLinear_kernel_time","args" : {"activation_size" : "65536","mem_bytes_requested_in_use" : "200716","mem_arena_held" : "1048588","mem_in_use_delta" : "131072","output_size" : "131072","mem_bytes_in_use" : "200716","node_index" : "5","output_type_shape" : [{"float":[1,32768]}],"op_name" : "DequantizeLinear","mem_requested_in_use_delta" : "131072","mem_in_use_peak" : "200716","mem_arena_held_delta" : "0","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"provider" : "CPUExecutionProvider","parameter_size" : "6"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1916,"ts" :3851092,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1930,"ts" :3851085,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :15,"ts" :3859501,"ph" : "X","name" :"normalized_x_offset_98304_QuantizeLinear_kernel_time","args" : {"mem_requested_in_use_delta" : "0","output_size" : "4096","mem_in_use_delta" : "0","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"provider" : "CPUExecutionProvider","activation_size" : "8192","op_name" : "QuantizeLinear","parameter_size" : "6","output_type_shape" : [{"uint16":[1,2048]}],"mem_arena_held_delta" : "0","node_index" : "0","mem_in_use_peak" : "200716","mem_bytes_in_use" : "69644","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"mem_arena_held" : "1048588","mem_bytes_requested_in_use" : "69644"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1887,"ts" :3859520,"ph" : "X","name" :"vitisai_vitisai_15891435282080274764_0_0_kernel_time","args" : {"input_type_shape" : [{"uint16":[1,2048]}],"provider" : "vitisai","mem_bytes_in_use" : "69644","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"output_type_shape" : [{"uint16":[1,32768]}],"mem_in_use_delta" : "0","mem_in_use_peak" : "200716","mem_bytes_requested_in_use" : "69644","node_index" : "6","parameter_size" : "0","mem_arena_held" : "1048588","activation_size" : "4096","mem_arena_held_delta" : "0","mem_requested_in_use_delta" : "0","output_size" : "65536","op_name" : "vitisai_15891435282080274764_0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :23,"ts" :3861414,"ph" : "X","name" :"logits_offset_98304_DequantizeLinear_kernel_time","args" : {"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"mem_bytes_requested_in_use" : "200716","mem_arena_held" : "1048588","mem_bytes_in_use" : "200716","provider" : "CPUExecutionProvider","mem_requested_in_use_delta" : "131072","output_type_shape" : [{"float":[1,32768]}],"output_size" : "131072","node_index" : "5","mem_in_use_peak" : "200716","parameter_size" : "6","mem_arena_held_delta" : "0","activation_size" : "65536","mem_in_use_delta" : "131072","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"op_name" : "DequantizeLinear"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1940,"ts" :3859500,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1956,"ts" :3859493,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :54,"ts" :3868103,"ph" : "X","name" :"normalized_x_offset_98304_QuantizeLinear_kernel_time","args" : {"mem_requested_in_use_delta" : "0","mem_in_use_peak" : "200716","mem_bytes_in_use" : "69644","op_name" : "QuantizeLinear","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"provider" : "CPUExecutionProvider","mem_arena_held_delta" : "0","activation_size" : "8192","parameter_size" : "6","node_index" : "0","mem_bytes_requested_in_use" : "69644","mem_in_use_delta" : "0","output_size" : "4096","mem_arena_held" : "1048588","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"output_type_shape" : [{"uint16":[1,2048]}]}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1972,"ts" :3868165,"ph" : "X","name" :"vitisai_vitisai_15891435282080274764_0_0_kernel_time","args" : {"mem_in_use_peak" : "200716","mem_arena_held_delta" : "0","mem_bytes_in_use" : "69644","mem_requested_in_use_delta" : "0","activation_size" : "4096","op_name" : "vitisai_15891435282080274764_0","mem_in_use_delta" : "0","node_index" : "6","output_type_shape" : [{"uint16":[1,32768]}],"parameter_size" : "0","input_type_shape" : [{"uint16":[1,2048]}],"provider" : "vitisai","mem_arena_held" : "1048588","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"output_size" : "65536","mem_bytes_requested_in_use" : "69644"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :37,"ts" :3870143,"ph" : "X","name" :"logits_offset_98304_DequantizeLinear_kernel_time","args" : {"parameter_size" : "6","node_index" : "5","mem_in_use_delta" : "131072","mem_bytes_in_use" : "200716","provider" : "CPUExecutionProvider","mem_in_use_peak" : "200716","mem_arena_held_delta" : "0","op_name" : "DequantizeLinear","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"mem_requested_in_use_delta" : "131072","mem_arena_held" : "1048588","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"activation_size" : "65536","output_size" : "131072","mem_bytes_requested_in_use" : "200716","output_type_shape" : [{"float":[1,32768]}]}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2083,"ts" :3868101,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2114,"ts" :3868077,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :35,"ts" :3877277,"ph" : "X","name" :"normalized_x_offset_98304_QuantizeLinear_kernel_time","args" : {"mem_requested_in_use_delta" : "0","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_in_use_delta" : "0","parameter_size" : "6","mem_arena_held" : "1048588","mem_bytes_requested_in_use" : "69644","activation_size" : "8192","mem_bytes_in_use" : "69644","output_size" : "4096","provider" : "CPUExecutionProvider","op_name" : "QuantizeLinear","node_index" : "0","output_type_shape" : [{"uint16":[1,2048]}],"mem_in_use_peak" : "200716","mem_arena_held_delta" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}}}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1956,"ts" :3877317,"ph" : "X","name" :"vitisai_vitisai_15891435282080274764_0_0_kernel_time","args" : {"parameter_size" : "0","mem_arena_held_delta" : "0","mem_requested_in_use_delta" : "0","mem_bytes_in_use" : "69644","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"output_size" : "65536","mem_bytes_requested_in_use" : "69644","input_type_shape" : [{"uint16":[1,2048]}],"mem_in_use_delta" : "0","activation_size" : "4096","provider" : "vitisai","node_index" : "6","mem_arena_held" : "1048588","mem_in_use_peak" : "200716","output_type_shape" : [{"uint16":[1,32768]}],"op_name" : "vitisai_15891435282080274764_0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :25,"ts" :3879282,"ph" : "X","name" :"logits_offset_98304_DequantizeLinear_kernel_time","args" : {"input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_arena_held_delta" : "0","mem_requested_in_use_delta" : "131072","activation_size" : "65536","mem_bytes_in_use" : "200716","node_index" : "5","parameter_size" : "6","mem_arena_held" : "1048588","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"mem_in_use_delta" : "131072","mem_in_use_peak" : "200716","op_name" : "DequantizeLinear","provider" : "CPUExecutionProvider","output_type_shape" : [{"float":[1,32768]}],"output_size" : "131072","mem_bytes_requested_in_use" : "200716"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2036,"ts" :3877275,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2064,"ts" :3877257,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :31,"ts" :3885838,"ph" : "X","name" :"normalized_x_offset_98304_QuantizeLinear_kernel_time","args" : {"output_type_shape" : [{"uint16":[1,2048]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"mem_in_use_peak" : "200716","node_index" : "0","mem_arena_held" : "1048588","mem_bytes_requested_in_use" : "69644","op_name" : "QuantizeLinear","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_bytes_in_use" : "69644","output_size" : "4096","mem_in_use_delta" : "0","mem_requested_in_use_delta" : "0","parameter_size" : "6","activation_size" : "8192","mem_arena_held_delta" : "0","provider" : "CPUExecutionProvider"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1905,"ts" :3885875,"ph" : "X","name" :"vitisai_vitisai_15891435282080274764_0_0_kernel_time","args" : {"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"output_size" : "65536","mem_bytes_in_use" : "69644","mem_arena_held" : "1048588","mem_in_use_delta" : "0","mem_arena_held_delta" : "0","input_type_shape" : [{"uint16":[1,2048]}],"node_index" : "6","mem_in_use_peak" : "200716","parameter_size" : "0","activation_size" : "4096","mem_requested_in_use_delta" : "0","provider" : "vitisai","output_type_shape" : [{"uint16":[1,32768]}],"op_name" : "vitisai_15891435282080274764_0","mem_bytes_requested_in_use" : "69644"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :22,"ts" :3887787,"ph" : "X","name" :"logits_offset_98304_DequantizeLinear_kernel_time","args" : {"parameter_size" : "6","mem_bytes_in_use" : "200716","mem_in_use_peak" : "200716","output_type_shape" : [{"float":[1,32768]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"mem_arena_held" : "1048588","provider" : "CPUExecutionProvider","mem_in_use_delta" : "131072","mem_requested_in_use_delta" : "131072","node_index" : "5","op_name" : "DequantizeLinear","mem_arena_held_delta" : "0","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_bytes_requested_in_use" : "200716","activation_size" : "65536","output_size" : "131072"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1981,"ts" :3885837,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :2001,"ts" :3885825,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :17,"ts" :3894194,"ph" : "X","name" :"normalized_x_offset_98304_QuantizeLinear_kernel_time","args" : {"mem_bytes_in_use" : "69644","provider" : "CPUExecutionProvider","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"mem_arena_held_delta" : "0","output_type_shape" : [{"uint16":[1,2048]}],"op_name" : "QuantizeLinear","activation_size" : "8192","mem_bytes_requested_in_use" : "69644","mem_in_use_delta" : "0","mem_in_use_peak" : "200716","output_size" : "4096","mem_arena_held" : "1048588","node_index" : "0","parameter_size" : "6","mem_requested_in_use_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1841,"ts" :3894214,"ph" : "X","name" :"vitisai_vitisai_15891435282080274764_0_0_kernel_time","args" : {"mem_in_use_peak" : "200716","mem_arena_held" : "1048588","mem_arena_held_delta" : "0","mem_requested_in_use_delta" : "0","mem_in_use_delta" : "0","activation_size" : "4096","parameter_size" : "0","mem_bytes_requested_in_use" : "69644","mem_bytes_in_use" : "69644","op_name" : "vitisai_15891435282080274764_0","output_size" : "65536","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"node_index" : "6","input_type_shape" : [{"uint16":[1,2048]}],"provider" : "vitisai","output_type_shape" : [{"uint16":[1,32768]}]}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :20,"ts" :3896059,"ph" : "X","name" :"logits_offset_98304_DequantizeLinear_kernel_time","args" : {"input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_arena_held_delta" : "0","mem_bytes_in_use" : "200716","mem_requested_in_use_delta" : "131072","provider" : "CPUExecutionProvider","output_type_shape" : [{"float":[1,32768]}],"mem_bytes_requested_in_use" : "200716","op_name" : "DequantizeLinear","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"activation_size" : "65536","mem_in_use_delta" : "131072","output_size" : "131072","mem_in_use_peak" : "200716","parameter_size" : "6","node_index" : "5","mem_arena_held" : "1048588"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1889,"ts" :3894192,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1902,"ts" :3894184,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :13,"ts" :3902517,"ph" : "X","name" :"normalized_x_offset_98304_QuantizeLinear_kernel_time","args" : {"output_type_shape" : [{"uint16":[1,2048]}],"mem_arena_held" : "1048588","mem_bytes_in_use" : "69644","mem_arena_held_delta" : "0","output_size" : "4096","mem_in_use_delta" : "0","parameter_size" : "6","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"mem_in_use_peak" : "200716","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"op_name" : "QuantizeLinear","activation_size" : "8192","mem_requested_in_use_delta" : "0","node_index" : "0","provider" : "CPUExecutionProvider","mem_bytes_requested_in_use" : "69644"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1834,"ts" :3902534,"ph" : "X","name" :"vitisai_vitisai_15891435282080274764_0_0_kernel_time","args" : {"mem_arena_held" : "1048588","provider" : "vitisai","node_index" : "6","input_type_shape" : [{"uint16":[1,2048]}],"output_size" : "65536","output_type_shape" : [{"uint16":[1,32768]}],"activation_size" : "4096","mem_requested_in_use_delta" : "0","mem_in_use_delta" : "0","mem_in_use_peak" : "200716","mem_arena_held_delta" : "0","op_name" : "vitisai_15891435282080274764_0","mem_bytes_requested_in_use" : "69644","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"mem_bytes_in_use" : "69644","parameter_size" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :17,"ts" :3904372,"ph" : "X","name" :"logits_offset_98304_DequantizeLinear_kernel_time","args" : {"provider" : "CPUExecutionProvider","op_name" : "DequantizeLinear","mem_arena_held" : "1048588","mem_bytes_requested_in_use" : "200716","node_index" : "5","mem_bytes_in_use" : "200716","activation_size" : "65536","mem_in_use_peak" : "200716","output_type_shape" : [{"float":[1,32768]}],"output_size" : "131072","mem_arena_held_delta" : "0","mem_in_use_delta" : "131072","parameter_size" : "6","mem_requested_in_use_delta" : "131072","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}]}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1876,"ts" :3902517,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1883,"ts" :3902512,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :18,"ts" :3911030,"ph" : "X","name" :"normalized_x_offset_98304_QuantizeLinear_kernel_time","args" : {"provider" : "CPUExecutionProvider","parameter_size" : "6","op_name" : "QuantizeLinear","mem_bytes_in_use" : "69644","mem_bytes_requested_in_use" : "69644","node_index" : "0","mem_in_use_peak" : "200716","mem_in_use_delta" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"activation_size" : "8192","output_type_shape" : [{"uint16":[1,2048]}],"mem_arena_held" : "1048588","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_requested_in_use_delta" : "0","output_size" : "4096","mem_arena_held_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1891,"ts" :3911053,"ph" : "X","name" :"vitisai_vitisai_15891435282080274764_0_0_kernel_time","args" : {"activation_size" : "4096","mem_in_use_peak" : "200716","mem_in_use_delta" : "0","node_index" : "6","mem_arena_held" : "1048588","mem_bytes_requested_in_use" : "69644","input_type_shape" : [{"uint16":[1,2048]}],"mem_bytes_in_use" : "69644","op_name" : "vitisai_15891435282080274764_0","output_type_shape" : [{"uint16":[1,32768]}],"mem_requested_in_use_delta" : "0","mem_arena_held_delta" : "0","output_size" : "65536","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"provider" : "vitisai","parameter_size" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :23,"ts" :3912953,"ph" : "X","name" :"logits_offset_98304_DequantizeLinear_kernel_time","args" : {"op_name" : "DequantizeLinear","node_index" : "5","output_size" : "131072","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"provider" : "CPUExecutionProvider","mem_arena_held" : "1048588","mem_bytes_requested_in_use" : "200716","mem_in_use_peak" : "200716","mem_in_use_delta" : "131072","mem_bytes_in_use" : "200716","output_type_shape" : [{"float":[1,32768]}],"mem_requested_in_use_delta" : "131072","activation_size" : "65536","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_arena_held_delta" : "0","parameter_size" : "6"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1950,"ts" :3911029,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1967,"ts" :3911021,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :15,"ts" :3919845,"ph" : "X","name" :"normalized_x_offset_98304_QuantizeLinear_kernel_time","args" : {"output_size" : "4096","mem_in_use_peak" : "200716","op_name" : "QuantizeLinear","mem_bytes_requested_in_use" : "69644","mem_in_use_delta" : "0","output_type_shape" : [{"uint16":[1,2048]}],"node_index" : "0","mem_arena_held_delta" : "0","activation_size" : "8192","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"mem_bytes_in_use" : "69644","mem_requested_in_use_delta" : "0","parameter_size" : "6","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"provider" : "CPUExecutionProvider","mem_arena_held" : "1048588"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1871,"ts" :3919863,"ph" : "X","name" :"vitisai_vitisai_15891435282080274764_0_0_kernel_time","args" : {"provider" : "vitisai","activation_size" : "4096","mem_bytes_requested_in_use" : "69644","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"op_name" : "vitisai_15891435282080274764_0","input_type_shape" : [{"uint16":[1,2048]}],"mem_in_use_delta" : "0","output_type_shape" : [{"uint16":[1,32768]}],"mem_arena_held_delta" : "0","output_size" : "65536","mem_arena_held" : "1048588","mem_in_use_peak" : "200716","node_index" : "6","parameter_size" : "0","mem_requested_in_use_delta" : "0","mem_bytes_in_use" : "69644"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :21,"ts" :3921740,"ph" : "X","name" :"logits_offset_98304_DequantizeLinear_kernel_time","args" : {"activation_size" : "65536","op_name" : "DequantizeLinear","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"mem_requested_in_use_delta" : "131072","mem_arena_held" : "1048588","node_index" : "5","output_size" : "131072","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"provider" : "CPUExecutionProvider","mem_in_use_peak" : "200716","mem_arena_held_delta" : "0","mem_bytes_requested_in_use" : "200716","mem_in_use_delta" : "131072","mem_bytes_in_use" : "200716","parameter_size" : "6","output_type_shape" : [{"float":[1,32768]}]}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1920,"ts" :3919844,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1933,"ts" :3919838,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :14,"ts" :3928163,"ph" : "X","name" :"normalized_x_offset_98304_QuantizeLinear_kernel_time","args" : {"parameter_size" : "6","activation_size" : "8192","output_size" : "4096","provider" : "CPUExecutionProvider","mem_requested_in_use_delta" : "0","mem_bytes_in_use" : "69644","mem_in_use_peak" : "200716","mem_in_use_delta" : "0","op_name" : "QuantizeLinear","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"output_type_shape" : [{"uint16":[1,2048]}],"node_index" : "0","mem_arena_held_delta" : "0","mem_bytes_requested_in_use" : "69644","mem_arena_held" : "1048588","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}}}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1883,"ts" :3928180,"ph" : "X","name" :"vitisai_vitisai_15891435282080274764_0_0_kernel_time","args" : {"mem_bytes_requested_in_use" : "69644","provider" : "vitisai","mem_bytes_in_use" : "69644","mem_arena_held" : "1048588","mem_in_use_peak" : "200716","mem_requested_in_use_delta" : "0","output_type_shape" : [{"uint16":[1,32768]}],"mem_in_use_delta" : "0","input_type_shape" : [{"uint16":[1,2048]}],"node_index" : "6","op_name" : "vitisai_15891435282080274764_0","mem_arena_held_delta" : "0","output_size" : "65536","activation_size" : "4096","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"parameter_size" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :20,"ts" :3930070,"ph" : "X","name" :"logits_offset_98304_DequantizeLinear_kernel_time","args" : {"output_size" : "131072","activation_size" : "65536","op_name" : "DequantizeLinear","mem_bytes_in_use" : "200716","mem_bytes_requested_in_use" : "200716","mem_requested_in_use_delta" : "131072","mem_in_use_peak" : "200716","mem_arena_held_delta" : "0","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_in_use_delta" : "131072","mem_arena_held" : "1048588","node_index" : "5","provider" : "CPUExecutionProvider","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"output_type_shape" : [{"float":[1,32768]}],"parameter_size" : "6"}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1931,"ts" :3928162,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1945,"ts" :3928155,"ph" : "X","name" :"model_run","args" : {}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :13,"ts" :3936122,"ph" : "X","name" :"normalized_x_offset_98304_QuantizeLinear_kernel_time","args" : {"activation_size" : "8192","mem_bytes_requested_in_use" : "69644","parameter_size" : "6","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"mem_arena_held_delta" : "0","mem_in_use_peak" : "200716","input_type_shape" : [{"float":[1,2048]},{"float":[]},{"uint16":[]}],"mem_in_use_delta" : "0","node_index" : "0","mem_arena_held" : "1048588","output_size" : "4096","output_type_shape" : [{"uint16":[1,2048]}],"op_name" : "QuantizeLinear","mem_bytes_in_use" : "69644","provider" : "CPUExecutionProvider","mem_requested_in_use_delta" : "0"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :1861,"ts" :3936138,"ph" : "X","name" :"vitisai_vitisai_15891435282080274764_0_0_kernel_time","args" : {"input_type_shape" : [{"uint16":[1,2048]}],"output_type_shape" : [{"uint16":[1,32768]}],"mem_in_use_delta" : "0","provider" : "vitisai","op_name" : "vitisai_15891435282080274764_0","output_size" : "65536","mem_in_use_peak" : "200716","activation_size" : "4096","mem_arena_held_delta" : "0","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"mem_bytes_requested_in_use" : "69644","node_index" : "6","parameter_size" : "0","mem_arena_held" : "1048588","mem_requested_in_use_delta" : "0","mem_bytes_in_use" : "69644"}}, +{"cat" : "Node","pid" :52324,"tid" :9168,"dur" :22,"ts" :3938006,"ph" : "X","name" :"logits_offset_98304_DequantizeLinear_kernel_time","args" : {"parameter_size" : "6","mem_bytes_requested_in_use" : "200716","mem_requested_in_use_delta" : "131072","op_name" : "DequantizeLinear","thread_scheduling_stats" : {"main_thread": {"thread_pool_name": "session-5-intra-op", "thread_id": "9168", "block_size": [], "core": -1, "Distribution": 0, "DistributionEnqueue": 0, "Run": 0, "Wait": 0, "WaitRevoke": 0}, "sub_threads": {"9340": {"num_run": 0, "core": -1},"38376": {"num_run": 0, "core": -1},"20844": {"num_run": 0, "core": -1},"66392": {"num_run": 0, "core": -1},"66136": {"num_run": 0, "core": -1},"47920": {"num_run": 0, "core": -1},"21432": {"num_run": 0, "core": -1},"63432": {"num_run": 0, "core": -1},"48032": {"num_run": 0, "core": -1},"65416": {"num_run": 0, "core": -1},"50296": {"num_run": 0, "core": -1}}},"mem_bytes_in_use" : "200716","mem_in_use_peak" : "200716","provider" : "CPUExecutionProvider","input_type_shape" : [{"uint16":[1,32768]},{"float":[]},{"uint16":[]}],"mem_arena_held" : "1048588","activation_size" : "65536","output_size" : "131072","mem_in_use_delta" : "131072","mem_arena_held_delta" : "0","node_index" : "5","output_type_shape" : [{"float":[1,32768]}]}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1909,"ts" :3936122,"ph" : "X","name" :"SequentialExecutor::Execute","args" : {}}, +{"cat" : "Session","pid" :52324,"tid" :9168,"dur" :1920,"ts" :3936117,"ph" : "X","name" :"model_run","args" : {}} +] diff --git a/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/spark_cpu_fp32_graph_stage_20260923.json b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/spark_cpu_fp32_graph_stage_20260923.json new file mode 100644 index 000000000..b51539d30 --- /dev/null +++ b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/spark_cpu_fp32_graph_stage_20260923.json @@ -0,0 +1,183 @@ +{ + "schema_version": 1, + "evidence": "B", + "platform": "Windows-11-10.0.26200-SP0", + "python": "C:\\Users\\zhout\\w2\\omni029venv\\Scripts\\python.exe", + "workload": { + "repeats": 20, + "concurrency": 1 + }, + "artifact": "C:\\Users\\zhout\\w2\\spark_fp32_graph_bundle_20260923\\manifest.json", + "samples_s": [ + 0.02021650003734976, + 0.017785399919375777, + 0.01730339997448027, + 0.01689029997214675, + 0.017744599957950413, + 0.018181300023570657, + 0.017321799998171628, + 0.01730349997524172, + 0.017821699962951243, + 0.017965200007893145, + 0.01755089999642223, + 0.017079100012779236, + 0.01932439999654889, + 0.02028769999742508, + 0.02014519996009767, + 0.01991740008816123, + 0.019411200075410306, + 0.019999300013296306, + 0.01934400002937764, + 0.020481200073845685 + ], + "started_unix": 1790159004.5437026, + "startup_s": 4.4776260000653565, + "execution_plan": { + "backend": "external.graph.v1", + "stage_id": 0, + "worker_generation": "ff1f6cc7539540f1bddb0e28fee7b18e", + "artifact_sha256": "d047e8dd2f8faded839a0b5eba64eee154563bb82584a1c32f0a2abcaa7acd4a", + "component": "spark_x2_5_1_7b_output_head_fp32_reference", + "route": { + "name": "spark-cpu-fp32-head", + "interpreter": "C:\\Users\\zhout\\npu-ep\\Scripts\\python.exe", + "worker": "\\\\wsl.localhost\\Ubuntu\\home\\zhout\\project\\edge_infer\\vllm-omni-edge\\vllm_omni\\edge\\local\\external\\worker_ort.py", + "ep": "cpu", + "is_windows": true, + "available": true, + "reason": "", + "extra": null + }, + "reserved_bytes": { + "machine:ram": 4294967296 + }, + "max_io_bytes": 8388608, + "placement": { + "ep": "cpu", + "graph_path": "C:\\Users\\zhout\\w2\\spark_fp32_graph_bundle_20260923\\graph.onnx", + "session_providers": [ + "CPUExecutionProvider" + ], + "available_providers": [ + "AzureExecutionProvider", + "CPUExecutionProvider" + ], + "node_counts": { + "CPUExecutionProvider": 3 + }, + "total_nodes": 3, + "target_nodes": 3, + "fraction_on_target": 1.0, + "session_create_s": 1.0052321999100968, + "warmup_s": 0.01582650002092123, + "rss_bytes": 1144971264, + "inputs": [ + { + "name": "x", + "type": "tensor(float)", + "shape": [ + 1, + 1, + 2048 + ] + } + ], + "outputs": [ + { + "name": "logits", + "type": "tensor(float)", + "shape": [ + 1, + 131072 + ] + } + ], + "profile_path": "C:\\Users\\zhout\\w2\\spark_cpu_fp32_graph_stage_20260923.placement_2026-09-23_18-23-27_990.json", + "ep_directory": "", + "requested_provider_missing": "", + "fallback_note": "", + "onnxruntime": "1.30.0", + "note": "" + }, + "worker": { + "executable": "C:\\Users\\zhout\\npu-ep\\Scripts\\python.exe", + "python": "3.12.10 (tags/v3.12.10:0cc8128, Apr 8 2025, 12:21:36) [MSC v.1943 64 bit (AMD64)]", + "platform": "win32", + "pid": 9660, + "onnxruntime": "1.30.0", + "numpy": "2.5.3", + "available_providers": [ + "AzureExecutionProvider", + "CPUExecutionProvider" + ], + "rss_bytes": 54865920, + "process_identity": { + "pid": 9660, + "created_filetime": 134346326062920451 + } + }, + "evidence": "B", + "stateful": false + }, + "worker_stats": { + "runs": 20, + "total_run_s": 0.3144564996473491, + "rss_bytes": 1145524224, + "peak_rss_bytes": 2217127936, + "report": { + "ep": "cpu", + "onnxruntime": "1.30.0", + "available_providers": [ + "AzureExecutionProvider", + "CPUExecutionProvider" + ], + "session_create_s": 1.0052321999100968, + "session_providers": [ + "CPUExecutionProvider" + ], + "inputs": [ + { + "name": "x", + "type": "tensor(float)", + "shape": [ + 1, + 1, + 2048 + ] + } + ], + "outputs": [ + { + "name": "logits", + "type": "tensor(float)", + "shape": [ + 1, + 131072 + ] + } + ], + "warmup_s": 0.01582650002092123, + "profile_path": "C:\\Users\\zhout\\w2\\spark_cpu_fp32_graph_stage_20260923.placement_2026-09-23_18-23-27_990.json", + "node_counts": { + "CPUExecutionProvider": 3 + }, + "total_nodes": 3, + "target_nodes": 3, + "fraction_on_target": 1.0, + "rss_bytes": 1144971264 + } + }, + "finite": true, + "status": "passed", + "ledger_after_shutdown": { + "capacities": { + "machine:ram": 34359738368 + }, + "reserved": { + "machine:ram": 0 + }, + "owners": [], + "quarantined": [] + }, + "ended_unix": 1790159009.4976017 +} \ No newline at end of file diff --git a/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/spark_cpu_fp32_graph_stage_20260923.log b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/spark_cpu_fp32_graph_stage_20260923.log new file mode 100644 index 000000000..961f0aa07 --- /dev/null +++ b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/spark_cpu_fp32_graph_stage_20260923.log @@ -0,0 +1,16 @@ +INFO 09-23 18:23:19 [patch.py:259] NVFP4 W4A4 weight_scale NaN-clamp: installed. +INFO 09-23 18:23:19 [patch.py:510] inductor factorable-divisibility patch: installed. +INFO 09-23 18:23:24 [omni_base.py:216] [Omni] Initializing with model C:\Users\zhout\w2\spark_fp32_graph_bundle_20260923 +INFO 09-23 18:23:24 [async_omni_engine.py:172] [AsyncOmniEngine] Initializing with model C:\Users\zhout\w2\spark_fp32_graph_bundle_20260923 +INFO 09-23 18:23:24 [async_omni_engine.py:290] [AsyncOmniEngine] Launching Orchestrator thread with 1 stages +INFO 09-23 18:23:25 [stage_init_utils.py:768] [stage_init] Set VLLM_WORKER_MULTIPROC_METHOD=spawn +INFO 09-23 18:23:25 [initialization.py:356] Loaded OmniTransferConfig with 0 connector configurations +INFO 09-23 18:23:29 [orchestrator.py:604] [Orchestrator] Starting event loop +INFO 09-23 18:23:29 [async_omni_engine.py:323] [AsyncOmniEngine] Orchestrator ready with 1 stages +INFO 09-23 18:23:29 [omni_base.py:235] [Omni] AsyncOmniEngine initialized in 4.48 seconds +INFO 09-23 18:23:29 [omni_base.py:255] [Omni] Initialized with 1 stages for model C:\Users\zhout\w2\spark_fp32_graph_bundle_20260923 +INFO 09-23 18:23:29 [omni_base.py:781] [Omni] Shutting down +INFO 09-23 18:23:29 [async_omni_engine.py:1807] [AsyncOmniEngine] Shutting down Orchestrator +INFO 09-23 18:23:29 [orchestrator.py:705] [Orchestrator] Received shutdown signal +INFO 09-23 18:23:29 [orchestrator.py:3092] [Orchestrator] Shutting down all 1 client(s) +INFO 09-23 18:23:29 [stage_pool.py:1354] [StagePool] Stage 0 replica 0 shut down diff --git a/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/spark_cpu_fp32_graph_stage_20260923.npz b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/spark_cpu_fp32_graph_stage_20260923.npz new file mode 100644 index 000000000..55715b9b0 Binary files /dev/null and b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/spark_cpu_fp32_graph_stage_20260923.npz differ diff --git a/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/spark_cpu_graph_stage_with_norm_20260923.json b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/spark_cpu_graph_stage_with_norm_20260923.json new file mode 100644 index 000000000..117da6ea5 --- /dev/null +++ b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/spark_cpu_graph_stage_with_norm_20260923.json @@ -0,0 +1,183 @@ +{ + "schema_version": 1, + "evidence": "B", + "platform": "Windows-11-10.0.26200-SP0", + "python": "C:\\Users\\zhout\\w2\\omni029venv\\Scripts\\python.exe", + "workload": { + "repeats": 20, + "concurrency": 1 + }, + "artifact": "C:\\Users\\zhout\\w2\\spark_npu_graph_bundle_with_norm_20260923\\manifest.json", + "samples_s": [ + 0.0987639999948442, + 0.06915559992194176, + 0.07164600002579391, + 0.06952990009449422, + 0.08282820007298142, + 0.06972630007658154, + 0.06990170001517981, + 0.06917469995096326, + 0.0694569000042975, + 0.07010379992425442, + 0.06956650002393872, + 0.06991139997262508, + 0.06863389990758151, + 0.07056929997634143, + 0.06995909998659045, + 0.06889899994712323, + 0.06980119994841516, + 0.07358500000555068, + 0.07201680005528033, + 0.07060139998793602 + ], + "started_unix": 1790158934.8817055, + "startup_s": 2.0408076000167057, + "execution_plan": { + "backend": "external.graph.v1", + "stage_id": 0, + "worker_generation": "d59d3dbe5c6b46b396d7616ba1afcfba", + "artifact_sha256": "b84d82fa1aa3c8b04c6c050af4521e9846c21babedbc9aa5d30f972a4c8a5cdf", + "component": "spark_x2_5_1_7b_output_head_amd_npu", + "route": { + "name": "spark-cpu-head", + "interpreter": "C:\\Users\\zhout\\npu-ep\\Scripts\\python.exe", + "worker": "\\\\wsl.localhost\\Ubuntu\\home\\zhout\\project\\edge_infer\\vllm-omni-edge\\vllm_omni\\edge\\local\\external\\worker_ort.py", + "ep": "cpu", + "is_windows": true, + "available": true, + "reason": "", + "extra": null + }, + "reserved_bytes": { + "machine:ram": 4294967296 + }, + "max_io_bytes": 8388608, + "placement": { + "ep": "cpu", + "graph_path": "C:\\Users\\zhout\\w2\\spark_npu_graph_bundle_with_norm_20260923\\graph.onnx", + "session_providers": [ + "CPUExecutionProvider" + ], + "available_providers": [ + "AzureExecutionProvider", + "CPUExecutionProvider" + ], + "node_counts": { + "CPUExecutionProvider": 24 + }, + "total_nodes": 24, + "target_nodes": 24, + "fraction_on_target": 1.0, + "session_create_s": 0.11892499995883554, + "warmup_s": 0.0925149000249803, + "rss_bytes": 617279488, + "inputs": [ + { + "name": "x", + "type": "tensor(float)", + "shape": [ + 1, + 1, + 2048 + ] + } + ], + "outputs": [ + { + "name": "logits_concatenated", + "type": "tensor(float)", + "shape": [ + 1, + 131072 + ] + } + ], + "profile_path": "C:\\Users\\zhout\\w2\\spark_cpu_graph_stage_with_norm_20260923.placement_2026-09-23_18-22-16_699.json", + "ep_directory": "", + "requested_provider_missing": "", + "fallback_note": "", + "onnxruntime": "1.30.0", + "note": "" + }, + "worker": { + "executable": "C:\\Users\\zhout\\npu-ep\\Scripts\\python.exe", + "python": "3.12.10 (tags/v3.12.10:0cc8128, Apr 8 2025, 12:21:36) [MSC v.1943 64 bit (AMD64)]", + "platform": "win32", + "pid": 41112, + "onnxruntime": "1.30.0", + "numpy": "2.5.3", + "available_providers": [ + "AzureExecutionProvider", + "CPUExecutionProvider" + ], + "rss_bytes": 54882304, + "process_identity": { + "pid": 41112, + "created_filetime": 134346325357430313 + } + }, + "evidence": "B", + "stateful": false + }, + "worker_stats": { + "runs": 20, + "total_run_s": 1.3929456000914797, + "rss_bytes": 903561216, + "peak_rss_bytes": 903565312, + "report": { + "ep": "cpu", + "onnxruntime": "1.30.0", + "available_providers": [ + "AzureExecutionProvider", + "CPUExecutionProvider" + ], + "session_create_s": 0.11892499995883554, + "session_providers": [ + "CPUExecutionProvider" + ], + "inputs": [ + { + "name": "x", + "type": "tensor(float)", + "shape": [ + 1, + 1, + 2048 + ] + } + ], + "outputs": [ + { + "name": "logits_concatenated", + "type": "tensor(float)", + "shape": [ + 1, + 131072 + ] + } + ], + "warmup_s": 0.0925149000249803, + "profile_path": "C:\\Users\\zhout\\w2\\spark_cpu_graph_stage_with_norm_20260923.placement_2026-09-23_18-22-16_699.json", + "node_counts": { + "CPUExecutionProvider": 24 + }, + "total_nodes": 24, + "target_nodes": 24, + "fraction_on_target": 1.0, + "rss_bytes": 617279488 + } + }, + "finite": true, + "status": "passed", + "ledger_after_shutdown": { + "capacities": { + "machine:ram": 34359738368 + }, + "reserved": { + "machine:ram": 0 + }, + "owners": [], + "quarantined": [] + }, + "ended_unix": 1790158938.4867656 +} \ No newline at end of file diff --git a/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/spark_cpu_graph_stage_with_norm_20260923.log b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/spark_cpu_graph_stage_with_norm_20260923.log new file mode 100644 index 000000000..3ebb2e77d --- /dev/null +++ b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/spark_cpu_graph_stage_with_norm_20260923.log @@ -0,0 +1,16 @@ +INFO 09-23 18:22:10 [patch.py:259] NVFP4 W4A4 weight_scale NaN-clamp: installed. +INFO 09-23 18:22:10 [patch.py:510] inductor factorable-divisibility patch: installed. +INFO 09-23 18:22:14 [omni_base.py:216] [Omni] Initializing with model C:\Users\zhout\w2\spark_npu_graph_bundle_with_norm_20260923 +INFO 09-23 18:22:14 [async_omni_engine.py:172] [AsyncOmniEngine] Initializing with model C:\Users\zhout\w2\spark_npu_graph_bundle_with_norm_20260923 +INFO 09-23 18:22:14 [async_omni_engine.py:290] [AsyncOmniEngine] Launching Orchestrator thread with 1 stages +INFO 09-23 18:22:15 [stage_init_utils.py:768] [stage_init] Set VLLM_WORKER_MULTIPROC_METHOD=spawn +INFO 09-23 18:22:15 [initialization.py:356] Loaded OmniTransferConfig with 0 connector configurations +INFO 09-23 18:22:16 [orchestrator.py:604] [Orchestrator] Starting event loop +INFO 09-23 18:22:16 [async_omni_engine.py:323] [AsyncOmniEngine] Orchestrator ready with 1 stages +INFO 09-23 18:22:16 [omni_base.py:235] [Omni] AsyncOmniEngine initialized in 2.04 seconds +INFO 09-23 18:22:16 [omni_base.py:255] [Omni] Initialized with 1 stages for model C:\Users\zhout\w2\spark_npu_graph_bundle_with_norm_20260923 +INFO 09-23 18:22:18 [omni_base.py:781] [Omni] Shutting down +INFO 09-23 18:22:18 [async_omni_engine.py:1807] [AsyncOmniEngine] Shutting down Orchestrator +INFO 09-23 18:22:18 [orchestrator.py:705] [Orchestrator] Received shutdown signal +INFO 09-23 18:22:18 [orchestrator.py:3092] [Orchestrator] Shutting down all 1 client(s) +INFO 09-23 18:22:18 [stage_pool.py:1354] [StagePool] Stage 0 replica 0 shut down diff --git a/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/spark_cpu_graph_stage_with_norm_20260923.npz b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/spark_cpu_graph_stage_with_norm_20260923.npz new file mode 100644 index 000000000..e91457e62 Binary files /dev/null and b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/spark_cpu_graph_stage_with_norm_20260923.npz differ diff --git a/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/spark_npu_graph_stage_with_norm_20260923.json b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/spark_npu_graph_stage_with_norm_20260923.json new file mode 100644 index 000000000..857a0e916 --- /dev/null +++ b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/spark_npu_graph_stage_with_norm_20260923.json @@ -0,0 +1,190 @@ +{ + "schema_version": 1, + "evidence": "B", + "platform": "Windows-11-10.0.26200-SP0", + "python": "C:\\Users\\zhout\\w2\\omni029venv\\Scripts\\python.exe", + "workload": { + "repeats": 20, + "concurrency": 1 + }, + "artifact": "C:\\Users\\zhout\\w2\\spark_npu_graph_bundle_with_norm_20260923\\manifest.json", + "samples_s": [ + 0.012680900050327182, + 0.01053269999101758, + 0.01016219996381551, + 0.011459299945272505, + 0.009730199933983386, + 0.009876800002530217, + 0.00959360005799681, + 0.009940600022673607, + 0.012443799991160631, + 0.010018699918873608, + 0.010059899999760091, + 0.011098999995738268, + 0.01147810008842498, + 0.0103192999958992, + 0.01055510004516691, + 0.009810600080527365, + 0.00969119998626411, + 0.010880200075916946, + 0.010206000064499676, + 0.010383800021372736 + ], + "started_unix": 1790158883.7750835, + "startup_s": 18.206711999955587, + "execution_plan": { + "backend": "external.graph.v1", + "stage_id": 0, + "worker_generation": "0f7e1318a52e48919c83798f3c40a9ff", + "artifact_sha256": "b84d82fa1aa3c8b04c6c050af4521e9846c21babedbc9aa5d30f972a4c8a5cdf", + "component": "spark_x2_5_1_7b_output_head_amd_npu", + "route": { + "name": "spark-npu-head", + "interpreter": "C:\\Users\\zhout\\npu-ep\\Scripts\\python.exe", + "worker": "\\\\wsl.localhost\\Ubuntu\\home\\zhout\\project\\edge_infer\\vllm-omni-edge\\vllm_omni\\edge\\local\\external\\worker_ort.py", + "ep": "vitisai", + "is_windows": true, + "available": true, + "reason": "", + "extra": null + }, + "reserved_bytes": { + "machine:ram": 4294967296 + }, + "max_io_bytes": 8388608, + "placement": { + "ep": "vitisai", + "graph_path": "C:\\Users\\zhout\\w2\\spark_npu_graph_bundle_with_norm_20260923\\graph.onnx", + "session_providers": [ + "vitisai", + "CPUExecutionProvider" + ], + "available_providers": [ + "AzureExecutionProvider", + "CPUExecutionProvider", + "vitisai" + ], + "node_counts": { + "CPUExecutionProvider": 7, + "vitisai": 1 + }, + "total_nodes": 8, + "target_nodes": 1, + "fraction_on_target": 0.125, + "session_create_s": 15.15212729992345, + "warmup_s": 0.010233499924652278, + "rss_bytes": 1394450432, + "inputs": [ + { + "name": "x", + "type": "tensor(float)", + "shape": [ + 1, + 1, + 2048 + ] + } + ], + "outputs": [ + { + "name": "logits_concatenated", + "type": "tensor(float)", + "shape": [ + 1, + 131072 + ] + } + ], + "profile_path": "C:\\Users\\zhout\\w2\\spark_npu_graph_stage_with_norm_20260923.placement_2026-09-23_18-21-26_808.json", + "ep_directory": "C:\\Program Files\\WindowsApps\\WindowsWorkload.EP.AMD.VitisAI.Framework.1.8_1.8.63.0_x64__8wekyb3d8bbwe\\ExecutionProvider", + "requested_provider_missing": "", + "fallback_note": "", + "onnxruntime": "1.30.0", + "note": "" + }, + "worker": { + "executable": "C:\\Users\\zhout\\npu-ep\\Scripts\\python.exe", + "python": "3.12.10 (tags/v3.12.10:0cc8128, Apr 8 2025, 12:21:36) [MSC v.1943 64 bit (AMD64)]", + "platform": "win32", + "pid": 24740, + "onnxruntime": "1.30.0", + "numpy": "2.5.3", + "available_providers": [ + "AzureExecutionProvider", + "CPUExecutionProvider" + ], + "rss_bytes": 54611968, + "process_identity": { + "pid": 24740, + "created_filetime": 134346324846482978 + } + }, + "evidence": "B", + "stateful": false + }, + "worker_stats": { + "runs": 20, + "total_run_s": 0.15461209998466074, + "rss_bytes": 1394978816, + "peak_rss_bytes": 3778600960, + "report": { + "ep": "vitisai", + "onnxruntime": "1.30.0", + "ep_directory": "C:\\Program Files\\WindowsApps\\WindowsWorkload.EP.AMD.VitisAI.Framework.1.8_1.8.63.0_x64__8wekyb3d8bbwe\\ExecutionProvider", + "available_providers": [ + "AzureExecutionProvider", + "CPUExecutionProvider", + "vitisai" + ], + "session_create_s": 15.15212729992345, + "session_providers": [ + "vitisai", + "CPUExecutionProvider" + ], + "inputs": [ + { + "name": "x", + "type": "tensor(float)", + "shape": [ + 1, + 1, + 2048 + ] + } + ], + "outputs": [ + { + "name": "logits_concatenated", + "type": "tensor(float)", + "shape": [ + 1, + 131072 + ] + } + ], + "warmup_s": 0.010233499924652278, + "profile_path": "C:\\Users\\zhout\\w2\\spark_npu_graph_stage_with_norm_20260923.placement_2026-09-23_18-21-26_808.json", + "node_counts": { + "CPUExecutionProvider": 7, + "vitisai": 1 + }, + "total_nodes": 8, + "target_nodes": 1, + "fraction_on_target": 0.125, + "rss_bytes": 1394450432 + } + }, + "finite": true, + "status": "passed", + "ledger_after_shutdown": { + "capacities": { + "machine:ram": 34359738368 + }, + "reserved": { + "machine:ram": 0 + }, + "owners": [], + "quarantined": [] + }, + "ended_unix": 1790158902.3293793 +} \ No newline at end of file diff --git a/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/spark_npu_graph_stage_with_norm_20260923.log b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/spark_npu_graph_stage_with_norm_20260923.log new file mode 100644 index 000000000..6aef85d5d --- /dev/null +++ b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/spark_npu_graph_stage_with_norm_20260923.log @@ -0,0 +1,16 @@ +INFO 09-23 18:21:19 [patch.py:259] NVFP4 W4A4 weight_scale NaN-clamp: installed. +INFO 09-23 18:21:19 [patch.py:510] inductor factorable-divisibility patch: installed. +INFO 09-23 18:21:23 [omni_base.py:216] [Omni] Initializing with model C:\Users\zhout\w2\spark_npu_graph_bundle_with_norm_20260923 +INFO 09-23 18:21:23 [async_omni_engine.py:172] [AsyncOmniEngine] Initializing with model C:\Users\zhout\w2\spark_npu_graph_bundle_with_norm_20260923 +INFO 09-23 18:21:23 [async_omni_engine.py:290] [AsyncOmniEngine] Launching Orchestrator thread with 1 stages +INFO 09-23 18:21:24 [stage_init_utils.py:768] [stage_init] Set VLLM_WORKER_MULTIPROC_METHOD=spawn +INFO 09-23 18:21:24 [initialization.py:356] Loaded OmniTransferConfig with 0 connector configurations +INFO 09-23 18:21:41 [orchestrator.py:604] [Orchestrator] Starting event loop +INFO 09-23 18:21:41 [async_omni_engine.py:323] [AsyncOmniEngine] Orchestrator ready with 1 stages +INFO 09-23 18:21:41 [omni_base.py:235] [Omni] AsyncOmniEngine initialized in 18.21 seconds +INFO 09-23 18:21:41 [omni_base.py:255] [Omni] Initialized with 1 stages for model C:\Users\zhout\w2\spark_npu_graph_bundle_with_norm_20260923 +INFO 09-23 18:21:42 [omni_base.py:781] [Omni] Shutting down +INFO 09-23 18:21:42 [async_omni_engine.py:1807] [AsyncOmniEngine] Shutting down Orchestrator +INFO 09-23 18:21:42 [orchestrator.py:705] [Orchestrator] Received shutdown signal +INFO 09-23 18:21:42 [orchestrator.py:3092] [Orchestrator] Shutting down all 1 client(s) +INFO 09-23 18:21:42 [stage_pool.py:1354] [StagePool] Stage 0 replica 0 shut down diff --git a/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/spark_npu_graph_stage_with_norm_20260923.npz b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/spark_npu_graph_stage_with_norm_20260923.npz new file mode 100644 index 000000000..9e750a5bd Binary files /dev/null and b/benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/spark_npu_graph_stage_with_norm_20260923.npz differ diff --git a/benchmarks/edge_harness/results/e2e_profiling_20260922/evidence/summary/README.md b/benchmarks/edge_harness/results/e2e_profiling_20260922/evidence/summary/README.md index 6aa36beaf..6448e30e8 100644 --- a/benchmarks/edge_harness/results/e2e_profiling_20260922/evidence/summary/README.md +++ b/benchmarks/edge_harness/results/e2e_profiling_20260922/evidence/summary/README.md @@ -1,6 +1,6 @@ # E2E check and profiling -All 60 pairings have a disposition. The timing tables below remain the 2026-09-22 measurements. The configuration matrix and affected per-cell reasons incorporate the [2026-09-23 scoped recovery runs](../../../e2e_recovery_20260923/README.md) and [WSL CPU/native Windows expansion](../../../e2e_expansion_20260923/README.md). The expansion adds separate 20-request serial MiniCPM-o text-to-speech profiles on WSL CPU and native Windows GPU; other new paths are single-request or short-stream runs. Profiling completion does not imply full release qualification. +All 60 pairings have a disposition. The timing tables below remain the 2026-09-22 measurements. The configuration matrix and affected per-cell reasons incorporate the [2026-09-23 scoped recovery runs](../../../e2e_recovery_20260923/README.md) and [WSL CPU/native Windows expansion](../../../e2e_expansion_20260923/README.md). The expansion adds separate 20-request serial MiniCPM-o text-to-speech profiles on WSL CPU and native Windows GPU, plus a separately profiled Spark output-head component on the HX370 AMD NPU. Profiling completion does not imply full release qualification. ## Configuration × model @@ -13,7 +13,7 @@ NOT E2E means no complete device/model workload was verified. “Scoped E2E” i | PC HX370 + RTX 5090 Laptop / WSL | profile completed | profile completed | scoped E2E: text + image | scoped E2E: text/image/audio → speech | synthetic policy pass | | PC HX370 + RTX 5090 Laptop / Windows | profile completed | profile completed | NOT E2E | scoped E2E: text/image/audio → speech | synthetic policy pass | | PC HX370 + Radeon 890M / Windows worker | NOT E2E | NOT E2E | NOT E2E | NOT E2E | NOT E2E | -| PC HX370 + AMD NPU / Windows worker | NOT E2E | NOT E2E | NOT E2E | NOT E2E | NOT E2E | +| PC HX370 + AMD NPU / Windows worker | NOT E2E (output-head component passed) | NOT E2E | NOT E2E | NOT E2E | NOT E2E | | PC CPU+iGPU+NPU +/- discrete GPU, joint model execution | NOT E2E | NOT E2E | NOT E2E | NOT E2E | NOT E2E | | PC Snapdragon X Elite CRD / AI Hub | NOT E2E | NOT E2E | NOT E2E | NOT E2E | NOT E2E | | Mobile Galaxy S25 / Snapdragon 8 Elite for Galaxy | NOT E2E | NOT E2E | NOT E2E | NOT E2E | NOT E2E | @@ -50,7 +50,7 @@ NOT E2E means no complete device/model workload was verified. “Scoped E2E” i | PC HX370 + Radeon 890M / Windows worker | Qwen3.8-27B | NOT_E2E_PROFILEABLE | FP32 448x448 vision tower: DirectML 85/113 nodes; normalized max error 9.095e-06 versus torch CPU. Historical component path, not full 27B or current graph-stage image-to-text integration. | Recover the hashed tower bundle and qualify its current Omni graph handoff with the language stage. | | PC HX370 + Radeon 890M / Windows worker | MiniCPM-o 4.5 | NOT_E2E_PROFILEABLE | No placement- and quality-verified whole-model graph bundle/stateful adapter for this model on this AMD route. Small backend graphs do not establish model support. | Supply and validate the missing matching artifact/backend, then run the complete specified workload. | | PC HX370 + Radeon 890M / Windows worker | InternVLA-A1 | NOT_E2E_PROFILEABLE | No placement- and quality-verified whole-model graph bundle/stateful adapter for this model on this AMD route. Small backend graphs do not establish model support. | Supply and validate the missing matching artifact/backend, then run the complete specified workload. | -| PC HX370 + AMD NPU / Windows worker | Spark-X2.5 | NOT_E2E_PROFILEABLE | No placement- and quality-verified whole-model graph bundle/stateful adapter for this model on this AMD route. Small backend graphs do not establish model support. | Supply and validate the missing matching artifact/backend, then run the complete specified workload. | +| PC HX370 + AMD NPU / Windows worker | Spark-X2.5 | COMPONENT_PASS; NOT_E2E_PROFILEABLE | The real 1.7B output head (RMS norm and full 131,072-logit projection) ran through Omni `external.graph.v1` with VitisAI: one fused NPU partition and seven CPU nodes per request. The A16W8 graph matched top-1 on four captured activations (full-logit SNR 55.43–56.81 dB). One warmup plus 20 serial repeated-input component requests gave p50/p95 10.21/12.44 ms, versus 17.97/20.29 ms for the original FP32 output head on the same Omni CPU stage. NPU startup was 18.21 s versus 4.48 s for FP32 CPU, and NPU worker peak RSS was 3.52 GiB versus 2.06 GiB. No full prefill/decode/KV/sampling, real handoff, power, or sustained run was verified. | Measure an actual vLLM-to-output-head handoff and whole-generation cost, state/cancellation, memory admission and quality before considering a split plan; [component evidence](../../../e2e_expansion_20260923/evidence/spark_amd_npu_output_head/README.md). | | PC HX370 + AMD NPU / Windows worker | Qwen3-TTS 0.6B CustomVoice | NOT_E2E_PROFILEABLE | No placement- and quality-verified whole-model graph bundle/stateful adapter for this model on this AMD route. Small backend graphs do not establish model support. | Supply and validate the missing matching artifact/backend, then run the complete specified workload. | | PC HX370 + AMD NPU / Windows worker | Qwen3.8-27B | NOT_E2E_PROFILEABLE | The tested A16W8 whole-vision-tower export failed NPU placement. This conclusion is specific to that export, not all future Qwen/NPU combinations. | Validate a supported export or smaller useful component, including same-checkpoint numerical quality. | | PC HX370 + AMD NPU / Windows worker | MiniCPM-o 4.5 | NOT_E2E_PROFILEABLE | No placement- and quality-verified whole-model graph bundle/stateful adapter for this model on this AMD route. Small backend graphs do not establish model support. | Supply and validate the missing matching artifact/backend, then run the complete specified workload. | diff --git a/docs/design/omni_edge_backends.md b/docs/design/omni_edge_backends.md index 427ebce7c..3afeb7bf9 100644 --- a/docs/design/omni_edge_backends.md +++ b/docs/design/omni_edge_backends.md @@ -51,6 +51,8 @@ memory-accounting and unavailable-device gaps remain documented repair items. ### Mobile and PC × model support matrix +This section preserves the 2026-09-22 audit. For the latest per-cell disposition and 2026-09-23 follow-up evidence, use the [current 60-cell E2E matrix](../../benchmarks/edge_harness/results/e2e_profiling_20260922/evidence/summary/README.md). In particular, the HX370 AMD NPU now has a placement- and quality-verified **Spark 1.7B output-head component** through Omni, while complete Spark generation on that route remains unverified; see the [component record](../../benchmarks/edge_harness/results/e2e_expansion_20260923/evidence/spark_amd_npu_output_head/README.md). This advances the M3 component gate, not its whole-model or same-device handoff gate. + **All 60 named pairings were checked on 2026-09-22** through actual runs, raw-record review or artifact/runtime preflight. This is not 60 successful executions. Each cell links to its evidence, finding and next required step.