Using seed: 58933 Cleaning up past runs Extracted 5 files from zip to folder: dataset Using args: ['accelerate', 'launch', '--config_file', 'accelerate_configs/uncompiled_1.yaml', '--gpu_ids', '0', 'train.py', '--model_name', 'hunyuan_video', '--pretrained_model_name_or_path', 'HunyuanVideo', '--enable_tiling', '--enable_slicing', '--data_root', 'dataset', '--caption_column', 'prompts.txt', '--video_column', 'videos.txt', '--seed', '58933', '--rank', '128', '--lora_alpha', '128', '--mixed_precision', 'bf16', '--output_dir', 'output', '--batch_size', '1', '--id_token', 'melty', '--video_resolution_buckets', '17x512x768', '49x512x768', '61x512x768', '--caption_dropout_p', '0.05', '--training_type', 'lora', '--train_steps', '200', '--gradient_accumulation_steps', '1', '--gradient_checkpointing', '--optimizer', 'adamw', '--lr', '2e-05', '--lr_scheduler', 'constant_with_warmup', '--lr_warmup_steps', '100', '--lr_num_cycles', '1', '--beta1', '0.9', '--beta2', '0.95', '--max_grad_norm', '1.0', '--weight_decay', '0.0001', '--epsilon', '1e-08', '--tracker_name', 'replicate-hunyuanvideo'] 01/06/2025 19:06:10 - INFO - finetrainers - Initialized FineTrainers 01/06/2025 19:06:10 - INFO - finetrainers - Distributed environment: DistributedType.NO Num processes: 1 Process index: 0 Local process index: 0 Device: cuda Mixed precision type: bf16 01/06/2025 19:06:10 - INFO - finetrainers - Initializing dataset and dataloader 01/06/2025 19:06:10 - INFO - finetrainers - Initializing models Loading checkpoint shards: 0%| | 0/4 [00:00