Buckets:
| project: | |
| repo: creuthecat/Open-Sora-v2.1 | |
| bucket: creuthecat/opensora-v21 | |
| base_model: hpcai-tech/Open-Sora-v2 | |
| training: | |
| stage1_i2v: true | |
| resolution: 256px | |
| gpus: 8 | |
| batch_size: 1 | |
| grad_accum: 8 | |
| learning_rate: 1.0e-5 | |
| max_steps: 10000 | |
| precision: bf16 | |
| gradient_checkpointing: true | |
| save_every_steps: 500 | |
| async_io: true | |
| reference: | |
| enabled: true | |
| max_references: 4 | |
| encoder: openai/clip-vit-large-patch14 | |
| embedding_dim: 768 | |
| train_projector: true | |
| text: | |
| encoder: google/t5-v1_1-xxl | |
| trainable: false | |
| vae: | |
| source: hpcai-tech/Open-Sora-v2/hunyuan_vae.safetensors | |
| trainable: false | |
| paths: | |
| dataset_csv: /data/train.csv | |
| videos: /data/videos | |
| base_checkpoint: /base-model/Open_Sora_v2.safetensors | |
| base_vae: /base-model/hunyuan_vae.safetensors | |
| output: /training-outputs | |
Xet Storage Details
- Size:
- 807 Bytes
- Xet hash:
- dfd4ba375a4ae3718e6f26457a401be647957aaff5b3d201726473c3c41e6cf6
·
Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.