Using seed: 42 Cleaning up past runs Extracted 12 files from zip to folder: input/videos Running latent pre-caching command INFO:__main__:Load dataset config from train.toml INFO:dataset.image_video_dataset:glob images in ./input/videos INFO:dataset.image_video_dataset:found 6 videos INFO:dataset.config_utils:[Dataset 0] is_image_dataset: False resolution: (960, 544) batch_size: 1 caption_extension: ".txt" enable_bucket: True bucket_no_upscale: False cache_directory: "./input/cache_directory" debug_dataset: False video_directory: "./input/videos" video_jsonl_file: "None" target_frames: [1, 25, 45] frame_extraction: head frame_stride: 1 frame_sample: 1 INFO:hunyuan_model.vae:Loading 3D VAE model (884-16c-hy) from: ckpts/hunyuan-video-t2v-720p/vae/pytorch_model.pt INFO:hunyuan_model.vae:VAE to dtype: torch.float16 Loaded VAE: FrozenDict([('in_channels', 3), ('out_channels', 3), ('down_block_types', ['DownEncoderBlockCausal3D', 'DownEncoderBlockCausal3D', 'DownEncoderBlockCausal3D', 'DownEncoderBlockCausal3D']), ('up_block_types', ['UpDecoderBlockCausal3D', 'UpDecoderBlockCausal3D', 'UpDecoderBlockCausal3D', 'UpDecoderBlockCausal3D']), ('block_out_channels', [128, 256, 512, 512]), ('layers_per_block', 2), ('act_fn', 'silu'), ('latent_channels', 16), ('norm_num_groups', 32), ('sample_size', 256), ('sample_tsize', 64), ('scaling_factor', 0.476986), ('force_upcast', True), ('spatial_compression_ratio', 8), ('time_compression_ratio', 4), ('mid_block_add_attention', True), ('_use_default_values', ['spatial_compression_ratio', 'force_upcast']), ('_class_name', 'AutoencoderKLCausal3D'), ('_diffusers_version', '0.4.2')]), dtype: torch.float16 INFO:__main__:Set chunk_size to 32 for CausalConv3d in VAE Encoding dataset [0] 0it [00:00, ?it/s] 1it [00:01, 1.19s/it] 2it [00:01, 1.70it/s] 3it [00:01, 2.53it/s] 4it [00:01, 3.29it/s] 5it [00:01, 3.95it/s] 6it [00:02, 4.49it/s] 7it [00:03, 1.38it/s] 8it [00:05, 1.04s/it] 9it [00:07, 1.26s/it] 10it [00:08, 1.41s/it] 11it [00:10, 1.51s/it] 12it [00:12, 1.58s/it] 13it [00:15, 2.03s/it] 14it [00:18, 2.35s/it] 15it [00:21, 2.57s/it] 16it [00:24, 2.72s/it] 17it [00:27, 2.86s/it] 18it [00:31, 2.93s/it] 18it [00:31, 1.72s/it] Running text encoder output pre-caching command The cache for model files in Transformers v4.22.0 has been updated. Migrating your old cache. This is a one-time only operation. You can interrupt this and resume the migration later on by calling `transformers.utils.move_cache()`. 0it [00:00, ?it/s] 0it [00:00, ?it/s] INFO:__main__:Load dataset config from train.toml INFO:dataset.image_video_dataset:glob images in ./input/videos INFO:dataset.image_video_dataset:found 6 videos INFO:dataset.config_utils:[Dataset 0] is_image_dataset: False resolution: (960, 544) batch_size: 1 caption_extension: ".txt" enable_bucket: True bucket_no_upscale: False cache_directory: "./input/cache_directory" debug_dataset: False video_directory: "./input/videos" video_jsonl_file: "None" target_frames: [1, 25, 45] frame_extraction: head frame_stride: 1 frame_sample: 1 INFO:__main__:loading text encoder 1: ckpts/text_encoder INFO:hunyuan_model.text_encoder:Loading text encoder model (llm) from: ckpts/text_encoder Loading checkpoint shards: 0%| | 0/4 [00:00