Reading the device profile of the first export showed a fifth of every
sliding layer went nowhere useful: Output 9.9%, its quantise conversion
3.1%, and the window Slice 7.4%. The layer took a 512-entry window in,
appended the new key, sliced the oldest off and handed the whole recomputed
window back out -- 21 of 28 layers paying to move a cache-sized tensor
across the graph boundary once per token. Full layers paid a similar 9.5%
concatenating the new key onto the cache.
Both are the same mistake, and both go away by joining in score space
instead of key space: score the cache and the new token separately, then
concatenate two vectors of length n and 1 rather than an n x 256 key matrix.
Nothing cache-sized is copied, and each layer returns just the new K/V entry
for the runtime to drop into its ring slot. Softmax over keys does not care
what order the slots are in and each cached key already carries its own
rotary phase, so a ring buffer needs no reordering -- only a mask for slots
not yet written. The sliding cache correspondingly holds window-1 entries,
since a 512-token window counts the current token (transformers' own sliding
cache holds 511 for the same reason).
Measured on a real Galaxy S25, fp16, ms per layer:
sliding 1.689 -> 1.483 (-12.2%)
full 1.559 -> 1.567 (unchanged: the concat saved is spent on the
extra score matmul)
Two shapes had to be written as matmuls rather than the obvious elementwise
forms -- the new token's score (broadcast multiply + ReduceSum) and its
value contribution (a two-axis broadcast multiply). Both compile at fp16 and
both make QNN's context-binary converter exit 14 under w4a16; as matmuls
they compile and are the cheaper op on HTP anyway.
Also adds a selectable GELU approximation, and the measurement that closes
the question: erf GELU is 11-13% of every layer, but tanh is *slower* on
this NPU (1.535 vs 1.483 ms) despite holding 65.2 dB. The exact form stays,
which is also the better-fidelity choice.
Parity against the reference implementation is 136.8 dB with identical
top-1/top-5, and a test pins ring and roll to the same output on a shared
history so the device numbers cannot silently measure a different model.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01L1ax5wYLA6XbqEL1i8L9gZ
Exports one Spark decode step with the KV cache as explicit tensors, sized
per layer type: the 21 sliding layers hold a fixed 512 entries at any
context, the 7 full-attention layers grow. That asymmetry is the whole
reason the architecture is interesting on a phone, so the two types get
separately sized caches, masks and rotary angles.
Verified against the HF reference at 1024 context: 138.5 dB logits SNR,
identical top-1 and top-5.
Profiled on a real Galaxy S25 (Snapdragon 8 Elite) through AI Hub, QNN
context binary on the NPU, ms per layer:
sliding (512 KV, any ctx) fp16 1.689 w8a16 0.912 w4a16 0.917
full @ 1k fp16 1.559 w8a16 1.129 w4a16 1.127
full @ 4k fp16 2.964 w8a16 1.681 w4a16 1.662
Composed into a step (21 sliding + 7 full), w4a16 gives 27.2 ms at 1k and
30.9 ms at 4k, against 31.6 and 46.5 ms for the same model with every layer
full-attention -- 1.16x and 1.51x. Worth noting the hybrid is 0.94x at 1k in
fp16: sliding layers rotate all 256 head dims where full layers rotate 64,
and below roughly 2k that RoPE cost outweighs the shorter cache.
Also moves the head-wise attention gate off vLLM's CPU linear dispatch. The
gate is [2048, 8], and any linear that narrow is routed to the sgl-kernel
packed GEMM, whose fake-tensor rule assumes the packed weight keeps N in dim
0; VNNI packing does not at that size, so it reports K and torch.compile
fails tracing the gate multiply. A plain parameter with F.linear gains
nothing less -- a 2048x8 matmul has no use for packed AMX tiles -- and it
keeps the model compilable on CPU and GPU alike.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01L1ax5wYLA6XbqEL1i8L9gZ
Spark-X2.5-1.7B is an Apache-2.0 on-device model built for agentic tool
use. vLLM 0.28 has no implementation for it, so it could not be served at
all; llama.cpp already ships one, which made it the only edge engine able
to run the model.
Three traits separate it from a Llama-style decoder:
* Hybrid attention -- layer_types interleaves three 512-token sliding
layers with one full-attention layer, which is what makes the 1M-token
context affordable on a phone.
* Per-layer-type RoPE -- sliding layers rotate all 256 head dims with
theta 1e4, full layers only the first quarter with theta 5e6.
* A head-wise attention output gate -- g_proj reads the same post-norm
input as QKV and scales each head's attention output by a sigmoid.
The checkpoint also stores QKV pre-fused as a single q_k_v_proj, so
load_weights splits it before vLLM's sharded QKVParallelLinear sees it.
Verified against the official remote-code implementation on transformers
4.57.1 (the version config.json targets; it does not import on the 5.x in
our venvs). In fp32 all three scenarios match exactly -- 32/32 greedy
tokens with a max top-20 logprob difference of 2.2e-05. In bf16 the short
and the 2465-token long-context prompts still match 32/32, which is what
exercises the sliding layers; the tool-calling prompt diverges at token 23
on a near-tie, and fp32 confirms that is rounding rather than a defect.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01L1ax5wYLA6XbqEL1i8L9gZ
The Code2Wav decoder already had an on-device export (WP4); the talker and its
code predictor did not, so the codec-token path stopped at the vocoder. This
adds the missing half: a one-token decode step with an explicit per-layer KV
cache, exportable to ONNX and from there to a QNN context binary, TFLite or
ONNX Runtime.
Two layout choices carry the performance, both measured on real Snapdragon
devices through Qualcomm AI Hub (Galaxy S24 = 8 Gen 3, S25 = 8 Elite):
* per-layer cache tensors instead of one stacked [num_layers, ...] tensor that
every layer slices, and
* grouped-query attention by reshaping the query rather than expanding keys and
values.
The stacked/expanded form spent ~70% of its NPU cycles in cache select,
reshape and fp32->fp16 conversion; the layout here is bandwidth-bound and runs
the 0.6B talker step in 15.8 ms instead of 36.9 ms on an 8 Elite (fp16,
256-token cache). The code predictor step is 2.25 ms, so talker plus its 15
sub-steps cost about 50 ms of an 80 ms frame.
sanitize_onnx() fixes the three things the dynamo exporter emits that
qairt-converter and AI Hub reject: IsNaN/Where guards, Reshape allowzero=1, and
rank-0 axes initializers on Unsqueeze/Squeeze.
The module docstring records the measured placement table: this graph belongs
on the NPU, the vocoder on the GPU (21.7 ms vs 182 ms per frame), which is 3.3x
better than sending everything to the NPU and makes the pipeline real time on
the phone in fp16. Integer quantization is not the lever it looks like: every
automatic int8 build measured was numerically broken.
Full write-up, job ids and profiles: analysis/mobile_htp_simulation.md section 5.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Dzr1Jusv4s6g765gqS9L92
Noise is drawn eagerly with the request generators (same RNG consumption as the
eager loop) and passed in, so the compiled post-step is generator-free and
shape-static; used on the fp32/torchao CPU path outside stored (top-p) mode.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
torchao's dynamic-activation int8 is numerically wrong under torch.compile on
CPU (cosine 0.02 vs fp32; caused runaway generations), so it is not offered;
weight-only + compile is faithful but slower (78 ms vs 36 ms per frame for the
15 sub-steps), so torch.ao dynamic int8 stays the default. Parity mode now also
records top-5 overlap and KL vs the bf16 model; compiled-step equivalence test.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Measured on a Xeon 8480C (12+4 threads, 0.6B): the 15 residual-codebook sub-steps
re-ran the 5-layer predictor over the whole prefix each time (97 ms per frame of
the 134 ms talker step) and vLLM's stock sampler spent 9 ms per step in its
torch.compile'd random-sample kernel on CPU.
- code predictor: forward_cached (KV cache per layer, 2-token prefill then one
token per sub-step) and a shape-static forward_static_step (tensor position +
boolean mask) that torch.compile handles as one graph; incremental decoding
is the default on CPU (VLLM_OMNI_PREDICTOR_INCREMENTAL), optional compile
(VLLM_OMNI_PREDICTOR_COMPILE=step), optional thread cap.
- CPU fast path: dynamic int8 Linear layers (fp32 activations) for the predictor
and its lm heads (VLLM_OMNI_PREDICTOR_INT8, default on; micro-benchmark 70 ->
36 ms per frame for the 15 sub-steps); VLLM_OMNI_PREDICTOR_PARITY_CHECK=1
records top-1 agreement / max logit diff against the bf16 model in situ.
- Qwen3-TTS talker: prefer_model_sampler on CPU with a sample() that mirrors
vLLM's Sampler (logits processors + penalties, temperature, top-k/top-p,
exponential-noise sampling with per-request generators) without the compiled
kernel (VLLM_OMNI_CPU_FAST_SAMPLER, default on).
- step_stats: sampler phase timers and sample-metadata timers.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Measured after the DELTA fix: 53 engine outputs for 52 rows, but the client
still received everything at once because each per-step payload is the newest
row only and the row alignment treated it as a cumulative snapshot (one yield
at the first row, one at the end). accumulate_codec_rows folds delta rows and
the final consolidated matrix into one growing matrix; unit test drives the
generator with per-step deltas and asserts one yield per step.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
The speech path forced FINAL_ONLY for Qwen3-TTS deployments without async
chunks, so a talker-only server emitted one engine output per request and all
codec frames arrived at once (measured: 1 engine output per request, first
codes at output #1). A talker-only stage (final_output + latent) now keeps
DELTA outputs so codec frames stream per step.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
- environment_variable_inventory: VLLM_OMNI_CPU_INDUCTOR / HW_PROFILE /
INIT_TIMELINE / STEP_STATS_DIR (public), STAGE_ID / INIT_TIMELINE_DIR
(internal); reference page rows added.
- edge_decoder_client.LocalDecoder: batched_chunked_decode takes a [B, Q, F]
tensor, not a list (the WP5 e2e run failed with a TypeError); the decoder
now also carries the ramp schedule attributes the Code2Wav stage sets.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
vllm_omni/edge/decoder_export.py exports a stateless windowed decoder (context
72 frames + chunk, one program per ramp chunk size) via torch.export (.pt2,
verified here on CPU for chunks 2 and 25 with the real 0.6B decoder: exact
match vs eager, 460 MB fp32 per program) or ExecuTorch XNNPACK (.pte; written
against the executorch 1.x API, not executed: package not installed, install
is permission gated; QNN partitioner export-only/unverified). The WP5 client
gains --decoder-backend executorch --pte-manifest for phone/NPU targets.
benchmarks/tts/edge_results/summarize.py renders all committed JSON as tables.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
- vllm_omni/edge/step_stats.py: process-local counters enabled by
VLLM_OMNI_STEP_STATS_DIR (scheduler schedule/update_from_output, engine-core
step, chunk hop enqueue->put / put->get / get, orchestrator dispatch), dumped
per process and merged into an attribution table (share of the 80 ms frame).
Hooks: stage_engine_core_proc (instrument after construction, dump at exit),
chunk_transfer_adapter (put/get timing, put->get via the shm lock-file mtime),
orchestrator dispatch loop. Bench: stream_latency_bench.py --step-stats.
- vllm_omni/edge/scheduling.py: playback slack, prefill-throttle decision and
the no-preemption KV admission rule (exact bytes/token from the talker KV
geometry); adapt.derive_overrides now sizes max_num_seqs with it and
auto_profile passes the geometry read from the model config.
- Engine re-reads parallel_stage_init after apply_deploy_profile so the edge
profile's orchestrator default takes effect (was read before the profile).
- benchmarks/tts/cpu_thread_sweep.py: CPU-only 8/16/32-thread sweep (+inductor)
with per-step attribution.
- Tests: tests/core/test_edge_scheduling.py, tests/edge/test_step_stats.py,
tests/benchmarks/test_cpu_thread_sweep.py, profile default test.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Wraps SharedMemoryConnector: CUDA tensor leaves are exported with
torch.multiprocessing reduce_tensor + an inter-process CUDA event, rebuilt and
cloned by the consumer, released by the producer after an ack file appears.
CPU leaves and metadata keep the shm path. Registered as "CudaIpcConnector";
deploy/qwen3_tts_cuda_ipc.yaml selects it for the talker->code2wav hop.
Adoption in the edge profile is measurement-gated (hop latency bench).
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
* diffusers is no longer imported by vllm_omni.diffusion.data (annotation only);
vllm_omni.diffusion.io_support (diffusion registry + distributed state) is
imported lazily by config_factory and async_omni_engine, so constructing a
Qwen3-TTS pipeline config / AsyncOmni no longer loads diffusers,
x_transformers or whisper (guarded by tests/config/test_import_guard.py).
* requirements/common.txt = core.txt (AR/speech runtime) + media.txt
(diffusion/video/extra audio); VLLM_OMNI_EDGE_BUILD=1 makes setup.py use
requirements/<device>-edge.txt (core only); `edge` extra adds websockets and
sounddevice for on-device clients.
Measured on this host (CUDA venv): `import vllm` 7.0-7.9 s, `import vllm_omni`
adds ~20 s of which ~17 s is transformers' own import (transformers.utils.import_utils)
pulled in by vllm_omni.patch / vllm_omni.version; that floor is not
vLLM-Omni's and is unchanged. Guard test: no diffusion module loaded on the TTS path.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
* vllm_omni/edge/hardware_probe.py: HardwareProfile from /proc, /sys and torch
(ISA flags incl. AVX-512/AMX and NEON/dotprod/i8mm/SVE, physical cores,
big.LITTLE clusters, NUMA, RAM, CUDA discrete vs Tegra unified memory, NPU
via override), hardware classes jetson / cuda_discrete / arm64_cpu / x86_cpu /
npu_phone; VLLM_OMNI_HW_PROFILE=<json> overrides the probe (emulation, tests).
* vllm_omni/edge/adapt.py: derive_overrides -> dtype, absolute KV budgets from
available memory minus weights, max_num_seqs, per-stage disjoint CPU thread
ranges on the big cluster, eager/graph switches for small or unified memory,
adaptive-chunk parameters, timeouts from a load cost model.
* vllm_omni/edge/calibrate.py: chunk-schedule calibration from measured chunk
arrivals with a per-hardware-fingerprint cache under VLLM_CACHE_ROOT.
* vllm_omni/edge/auto_profile.py + deploy_profile="auto": edge/<model>.yaml ->
edge/hardware/<class>.yaml -> derived overrides -> cached calibration,
materialized as a temp deploy YAML (platform overlays are superseded).
* benchmarks/tts/hw_emulation.py: emulated classes (ATEN_CPU_CAPABILITY,
taskset, profile overrides) driving stream_latency_bench with auto profiles.
* CPU-only tests for probe snapshots, derivation rules, calibration and auto.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
* vllm_omni/deploy/edge/qwen3_tts.yaml: single-device Qwen3-TTS profile with a
ramped codec chunk schedule (2->4->8->16->25 frames) that removes the ~80 ms
playback stall of the default 1->25 schedule, decode batch 1, max_num_seqs 4.
* deploy_profile="edge" (Omni/AsyncOmni kwarg, --deploy-profile CLI, headless
serve) resolves vllm_omni/deploy/<profile>/<model_type>.yaml; mutually
exclusive with deploy_config; explicit profiles never fall back silently.
* tests/helpers/stage_config.py: modify_stage_config anchors a relative
base_config before writing the temp YAML so overlay-derived configs resolve.
* async_omni_engine.py also carries the init-timeline hooks used by WP2.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Adds benchmarks/tts/stream_latency_bench.py (AsyncOmni driver reusing the
Qwen3-TTS example's prompt builder; --model override, warm-up/repeat,
--init-only, per-chunk arrivals, peak RSS and per-process GPU memory) and
benchmarks/tts/stream_latency_metrics.py (TTFA, underrun-free playback start,
stall at TTFA, RTF streamed/total, chunk gaps, percentiles) with CPU unit
tests. Results go under benchmarks/tts/edge_results/stream_latency/.
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
