mlfoundations-dev
/

open-o1-sft-original

Text Generation

Generated from Trainer

text-generation-inference

Model card Files Files and versions

mikeam commited on Jan 17

Commit

403343c

·

verified ·

1 Parent(s): 0231348

Upload configs.yaml with huggingface_hub

Files changed (1) hide show

configs.yaml +2 -2

configs.yaml CHANGED Viewed

@@ -14,12 +14,12 @@ global_batch_size: 512
 gradient_accumulation_steps: 8
 gradient_checkpointing: true
 hub_model_id: mlfoundations-dev/open-o1-sft-original
-include_hp: dcft/train/hp_settings/hritik.yaml
 learning_rate: 5.0e-06
 logging_steps: 10
 lr_scheduler_type: constant
 max_grad_norm: 1
-model_name_or_path: meta-llama/Meta-Llama-3.1-8B
 neat_packing: true
 num_train_epochs: 3.0
 output_dir: ./experiments/train/checkpoints/open_o1_sft_original

 gradient_accumulation_steps: 8
 gradient_checkpointing: true
 hub_model_id: mlfoundations-dev/open-o1-sft-original
+include_hp: dcft/train/hp_settings/hritik_no_model.yaml
 learning_rate: 5.0e-06
 logging_steps: 10
 lr_scheduler_type: constant
 max_grad_norm: 1
+model_name_or_path: meta-llama/Llama-3.1-8B
 neat_packing: true
 num_train_epochs: 3.0
 output_dir: ./experiments/train/checkpoints/open_o1_sft_original