Download annotations/annotation_47/LoRAMoE/ds_zero3_offload.json from mcy98/RECODE-H: direct link, hf CLI and curl.
- Browser
- Download file 788 Bytes
-
https://huggingface.co/datasets/mcy98/RECODE-H/resolve/main/annotations/annotation_47/LoRAMoE/ds_zero3_offload.json
- Command line
-
hf download hf://datasets/mcy98/RECODE-H/annotations/annotation_47/LoRAMoE/ds_zero3_offload.json
-
curl -L -o ds_zero3_offload.json https://huggingface.co/datasets/mcy98/RECODE-H/resolve/main/annotations/annotation_47/LoRAMoE/ds_zero3_offload.json
788 Bytes
| { | |
| "bf16": { | |
| "enabled": true | |
| }, | |
| "zero_optimization": { | |
| "stage": 3, | |
| "offload_optimizer": { | |
| "device": "cpu", | |
| "pin_memory": true | |
| }, | |
| "offload_param": { | |
| "device": "cpu", | |
| "pin_memory": true | |
| }, | |
| "allgather_partitions": true, | |
| "allgather_bucket_size": 1e8, | |
| "overlap_comm": true, | |
| "reduce_scatter": true, | |
| "reduce_bucket_size": 1e8, | |
| "contiguous_gradients": true, | |
| "stage3_gather_16bit_weights_on_model_save":true | |
| }, | |
| "gradient_accumulation_steps": "auto", | |
| "gradient_clipping": "auto", | |
| "steps_per_print": 2000, | |
| "train_batch_size": "auto", | |
| "train_micro_batch_size_per_gpu": "auto", | |
| "wall_clock_breakdown": false | |
| } | |