From d20a6b5b39270cad7a8c47ed0a4a79b8326f0cc8 Mon Sep 17 00:00:00 2001 From: Lancer Date: Fri, 24 Apr 2026 13:40:38 +0800 Subject: [PATCH] [chore] refine the examples (#3095) Signed-off-by: Lancer --- .../image_to_image/image_edit.py | 30 +++++++++++++++ .../text_to_audio/text_to_audio.py | 38 +++++++++++++++++++ .../text_to_image/text_to_image.py | 17 +++++++++ .../text_to_video/text_to_video.py | 24 ++++++++++++ 4 files changed, 109 insertions(+) diff --git a/examples/offline_inference/image_to_image/image_edit.py b/examples/offline_inference/image_to_image/image_edit.py index 1a7e86f13..2a73d875d 100644 --- a/examples/offline_inference/image_to_image/image_edit.py +++ b/examples/offline_inference/image_to_image/image_edit.py @@ -196,6 +196,13 @@ def parse_args() -> argparse.Namespace: default=1, help="Number of GPUs used for ulysses sequence parallelism.", ) + parser.add_argument( + "--ulysses-mode", + type=str, + default="strict", + choices=["strict", "advanced_uaa"], + help="Ulysses sequence-parallel mode: 'strict' (divisibility required) or 'advanced_uaa' (UAA).", + ) parser.add_argument( "--ring-degree", type=int, @@ -325,6 +332,29 @@ def parse_args() -> argparse.Namespace: action="store_true", help="Enable layerwise (blockwise) offloading on DiT modules.", ) + parser.add_argument( + "--vae-patch-parallel-size", + type=int, + default=1, + help="Number of GPUs used for VAE patch/tile parallelism (decode).", + ) + parser.add_argument( + "--use-hsdp", + action="store_true", + help="Enable HSDP (Hybrid Sharded Data Parallel) for diffusion models.", + ) + parser.add_argument( + "--hsdp-shard-size", + type=int, + default=1, + help="Number of GPUs to shard weights across for HSDP.", + ) + parser.add_argument( + "--hsdp-replicate-size", + type=int, + default=1, + help="Number of HSDP replica groups.", + ) parser.add_argument( "--enable-diffusion-pipeline-profiler", action="store_true", diff --git a/examples/offline_inference/text_to_audio/text_to_audio.py b/examples/offline_inference/text_to_audio/text_to_audio.py index 53f3daf8d..2a1613e5e 100644 --- a/examples/offline_inference/text_to_audio/text_to_audio.py +++ b/examples/offline_inference/text_to_audio/text_to_audio.py @@ -131,6 +131,44 @@ def parse_args() -> argparse.Namespace: default=1, help="Number of HSDP replica groups. Default 1 means pure sharding.", ) + parser.add_argument( + "--tensor-parallel-size", + type=int, + default=1, + help="Number of GPUs used for tensor parallelism (TP) inside the DiT.", + ) + parser.add_argument( + "--ulysses-degree", + type=int, + default=1, + help="Number of GPUs used for ulysses sequence parallelism.", + ) + parser.add_argument( + "--ulysses-mode", + type=str, + default="strict", + choices=["strict", "advanced_uaa"], + help="Ulysses sequence-parallel mode: 'strict' (divisibility required) or 'advanced_uaa' (UAA).", + ) + parser.add_argument( + "--ring-degree", + type=int, + default=1, + help="Number of GPUs used for ring sequence parallelism.", + ) + parser.add_argument( + "--cfg-parallel-size", + type=int, + default=1, + choices=[1, 2], + help="Number of GPUs used for classifier free guidance parallel size.", + ) + parser.add_argument( + "--vae-patch-parallel-size", + type=int, + default=1, + help="Number of GPUs used for VAE patch/tile parallelism (decode).", + ) return parser.parse_args() diff --git a/examples/offline_inference/text_to_image/text_to_image.py b/examples/offline_inference/text_to_image/text_to_image.py index 3b3f8e77c..ff54d9fa1 100644 --- a/examples/offline_inference/text_to_image/text_to_image.py +++ b/examples/offline_inference/text_to_image/text_to_image.py @@ -144,6 +144,23 @@ def parse_args() -> argparse.Namespace: action="store_true", help="Enable layerwise (blockwise) offloading on DiT modules.", ) + parser.add_argument( + "--use-hsdp", + action="store_true", + help="Enable HSDP (Hybrid Sharded Data Parallel) for diffusion models.", + ) + parser.add_argument( + "--hsdp-shard-size", + type=int, + default=1, + help="Number of GPUs to shard weights across for HSDP.", + ) + parser.add_argument( + "--hsdp-replicate-size", + type=int, + default=1, + help="Number of HSDP replica groups.", + ) parser.add_argument( "--quantization", type=str, diff --git a/examples/offline_inference/text_to_video/text_to_video.py b/examples/offline_inference/text_to_video/text_to_video.py index 83925cc45..ac9c7b091 100644 --- a/examples/offline_inference/text_to_video/text_to_video.py +++ b/examples/offline_inference/text_to_video/text_to_video.py @@ -131,6 +131,13 @@ def parse_args() -> argparse.Namespace: action="store_true", help="Enable layerwise (blockwise) offloading on DiT modules.", ) + parser.add_argument( + "--ulysses-mode", + type=str, + default="strict", + choices=["strict", "advanced_uaa"], + help="Ulysses sequence-parallel mode: 'strict' (divisibility required) or 'advanced_uaa' (UAA).", + ) parser.add_argument( "--ulysses-degree", type=int, @@ -185,6 +192,23 @@ def parse_args() -> argparse.Namespace: choices=["fp8", "gguf"], help="Quantization method for the transformer (fp8 for online FP8 quantization).", ) + parser.add_argument( + "--use-hsdp", + action="store_true", + help="Enable HSDP (Hybrid Sharded Data Parallel) for diffusion models.", + ) + parser.add_argument( + "--hsdp-shard-size", + type=int, + default=1, + help="Number of GPUs to shard weights across for HSDP.", + ) + parser.add_argument( + "--hsdp-replicate-size", + type=int, + default=1, + help="Number of HSDP replica groups.", + ) return parser.parse_args()