Upload README.md with huggingface_hub

README.md

@@ -16,13 +16,13 @@ args:
  'model_name_or_path': '/generation_model',
  'push_to_hub': True,
  'revision': 'main',
- 'save_filename': '/output/shards/18767/79.jsonl',
+ 'save_filename': '/output/shards/18767/83.jsonl',
  'skill': 'chat'}
 
 dataset_args:
-{'dataset_end_idx': 642960,
+{'dataset_end_idx': 675108,
  'dataset_name': 'ai2-adapt-dev/new_base_rs_mix',
- 'dataset_start_idx': 634923,
+ 'dataset_start_idx': 667071,
  'dataset_test_split': 'validation',
  'dataset_text_field': 'prompt',
  'dataset_train_split': 'train',
@@ -39,5 +39,5 @@ gen_args:
 
 ## Reproduce this dataset
 
-1. Download the `['generation.py', '79.jsonl']` from the https://huggingface.co/datasets/jacobmorrison/generation_18767.
-2. Run `python open_instruct/rejection_sampling/generation.py --dataset_name ai2-adapt-dev/new_base_rs_mix --model_name_or_path /generation_model --dataset_start_idx 634923 --dataset_end_idx 642960 --save_filename /output/shards/18767/79.jsonl --hf_repo_id generation_18767 --no_add_timestamp --push_to_hub --num_completions 8 --tensor_parallel_size 1`
+1. Download the `['generation.py', '83.jsonl']` from the https://huggingface.co/datasets/jacobmorrison/generation_18767.
+2. Run `python open_instruct/rejection_sampling/generation.py --dataset_name ai2-adapt-dev/new_base_rs_mix --model_name_or_path /generation_model --dataset_start_idx 667071 --dataset_end_idx 675108 --save_filename /output/shards/18767/83.jsonl --hf_repo_id generation_18767 --no_add_timestamp --push_to_hub --num_completions 8 --tensor_parallel_size 1`