add training pipeline, need to test

examples/training/finetune/hunyuan15_t2v_480p/download_dataset.sh

@@ -0,0 +1,3 @@
+#!/bin/bash
+
+python scripts/huggingface/download_hf.py --repo_id "wlsaidhi/crush-smol-merged" --local_dir "data/crush-smol" --repo_type "dataset"

examples/training/finetune/hunyuan15_t2v_480p/preprocess_hunyuan15_t2v.sh

@@ -0,0 +1,25 @@
+#!/bin/bash
+
+GPU_NUM=1 # 2,4,8
+MODEL_PATH="hunyuanvideo-community/HunyuanVideo-1.5-Diffusers-480p_t2v"
+DATASET_PATH="data/crush-smol/"
+OUTPUT_DIR="data/crush-smol_processed_t2v/"
+
+torchrun --nproc_per_node=$GPU_NUM \
+    -m fastvideo.pipelines.preprocess.v1_preprocessing_new \
+    --model_path $MODEL_PATH \
+    --mode preprocess \
+    --workload_type t2v \
+    --preprocess.video_loader_type torchvision \
+    --preprocess.dataset_type merged \
+    --preprocess.dataset_path $DATASET_PATH \
+    --preprocess.dataset_output_dir $OUTPUT_DIR \
+    --preprocess.preprocess_video_batch_size 2 \
+    --preprocess.dataloader_num_workers 0 \
+    --preprocess.max_height 480 \
+    --preprocess.max_width 848 \
+    --preprocess.num_frames 77 \
+    --preprocess.train_fps 24 \
+    --preprocess.samples_per_file 8 \
+    --preprocess.flush_frequency 8 \
+    --preprocess.video_length_tolerance_range 5

fastvideo/pipelines/preprocess/hunyuan15/hunyuan15_preprocess_pipelines.py

@@ -0,0 +1,60 @@
+from fastvideo.fastvideo_args import FastVideoArgs
+from fastvideo.pipelines.composed_pipeline_base import ComposedPipelineBase
+from fastvideo.pipelines.preprocess.preprocess_stages import (
+	TextTransformStage, VideoTransformStage,
+)
+from fastvideo.pipelines.stages import (
+	EncodingStage,
+	TextEncodingStage,
+)
+
+
+class PreprocessPipelineT2V(ComposedPipelineBase):
+
+	_required_config_modules = [
+		"text_encoder", "text_encoder_2", "tokenizer", "tokenizer_2", "vae",
+	]
+
+	def create_pipeline_stages(self, fastvideo_args: FastVideoArgs):
+		assert fastvideo_args.preprocess_config is not None
+
+		self.add_stage(
+			stage_name="text_transform_stage",
+			stage=TextTransformStage(
+				cfg_uncondition_drop_rate=fastvideo_args.preprocess_config.training_cfg_rate,
+				seed=fastvideo_args.preprocess_config.seed,
+			),
+		)
+
+		self.add_stage(
+			stage_name="prompt_encoding_stage",
+			stage=TextEncodingStage(
+				text_encoders=[
+					self.get_module("text_encoder"),
+					self.get_module("text_encoder_2"),
+				],
+				tokenizers=[
+					self.get_module("tokenizer"),
+					self.get_module("tokenizer_2"),
+				],
+			),
+		)
+
+		self.add_stage(
+			stage_name="video_transform_stage",
+			stage=VideoTransformStage(
+				train_fps=fastvideo_args.preprocess_config.train_fps,
+				num_frames=fastvideo_args.preprocess_config.num_frames,
+				max_height=fastvideo_args.preprocess_config.max_height,
+				max_width=fastvideo_args.preprocess_config.max_width,
+				do_temporal_sample=fastvideo_args.preprocess_config.do_temporal_sample,
+			),
+		)
+
+		self.add_stage(
+			stage_name="video_encoding_stage",
+			stage=EncodingStage(vae=self.get_module("vae")),
+		)
+
+
+EntryClass = [PreprocessPipelineT2V]

fastvideo/training/hunyuan15_t2v_training_pipeline.py

@@ -0,0 +1,74 @@
+# SPDX-License-Identifier: Apache-2.0
+import sys
+from copy import deepcopy
+
+from fastvideo.fastvideo_args import FastVideoArgs, TrainingArgs
+from fastvideo.logger import init_logger
+from fastvideo.models.schedulers.scheduling_flow_match_euler_discrete import (
+    FlowMatchEulerDiscreteScheduler)
+from fastvideo.pipelines.basic.hunyuan15.hunyuan15_pipeline import HunyuanVideo15Pipeline
+from fastvideo.training.training_pipeline import TrainingPipeline
+from fastvideo.utils import is_vsa_available
+
+vsa_available = is_vsa_available()
+
+logger = init_logger(__name__)
+
+
+class Hunyuan15TrainingPipeline(TrainingPipeline):
+    """
+    A training pipeline for Hunyuan.
+    """
+    _required_config_modules = ["scheduler", "transformer", "vae"]
+
+    def initialize_pipeline(self, fastvideo_args: FastVideoArgs):
+        self.modules["scheduler"] = FlowMatchEulerDiscreteScheduler(
+            shift=fastvideo_args.pipeline_config.flow_shift)
+
+    def create_training_stages(self, training_args: TrainingArgs):
+        """
+        May be used in future refactors.
+        """
+        pass
+
+    def initialize_validation_pipeline(self, training_args: TrainingArgs):
+        logger.info("Initializing validation pipeline...")
+        args_copy = deepcopy(training_args)
+
+        args_copy.inference_mode = True
+        validation_pipeline = HunyuanVideo15Pipeline.from_pretrained(
+            training_args.model_path,
+            args=args_copy,  # type: ignore
+            inference_mode=True,
+            loaded_modules={
+                "transformer": self.get_module("transformer"),
+            },
+            tp_size=training_args.tp_size,
+            sp_size=training_args.sp_size,
+            num_gpus=training_args.num_gpus,
+            pin_cpu_memory=training_args.pin_cpu_memory,
+            dit_cpu_offload=True)
+
+        self.validation_pipeline = validation_pipeline
+
+
+def main(args) -> None:
+    logger.info("Starting training pipeline...")
+
+    pipeline = HunyuanVideo15Pipeline.from_pretrained(
+        args.pretrained_model_name_or_path, args=args)
+    args = pipeline.training_args
+    pipeline.train()
+    logger.info("Training pipeline done")
+
+
+if __name__ == "__main__":
+    argv = sys.argv
+    from fastvideo.fastvideo_args import TrainingArgs
+    from fastvideo.utils import FlexibleArgumentParser
+    parser = FlexibleArgumentParser()
+    parser = TrainingArgs.add_cli_args(parser)
+    parser = FastVideoArgs.add_cli_args(parser)
+    args = parser.parse_args()
+    args.dit_cpu_offload = False
+    main(args)