| { | |
| "results": { | |
| "arc_challenge": { | |
| "acc,none": 0.5614334470989761, | |
| "acc_stderr,none": 0.014500682618212865, | |
| "acc_norm,none": 0.613481228668942, | |
| "acc_norm_stderr,none": 0.014230084761910473, | |
| "alias": "arc_challenge" | |
| } | |
| }, | |
| "configs": { | |
| "arc_challenge": { | |
| "task": "arc_challenge", | |
| "group": [ | |
| "ai2_arc" | |
| ], | |
| "dataset_path": "/lustre07/scratch/gagan30/arocr/meta-llama/self_rewarding_models/eval/ai2_arc", | |
| "dataset_name": "ARC-Challenge", | |
| "training_split": "train", | |
| "validation_split": "validation", | |
| "test_split": "test", | |
| "doc_to_text": "Question: {{question}}\nAnswer:", | |
| "doc_to_target": "{{choices.label.index(answerKey)}}", | |
| "doc_to_choice": "{{choices.text}}", | |
| "description": "", | |
| "target_delimiter": " ", | |
| "fewshot_delimiter": "\n\n", | |
| "num_fewshot": 25, | |
| "metric_list": [ | |
| { | |
| "metric": "acc", | |
| "aggregation": "mean", | |
| "higher_is_better": true | |
| }, | |
| { | |
| "metric": "acc_norm", | |
| "aggregation": "mean", | |
| "higher_is_better": true | |
| } | |
| ], | |
| "output_type": "multiple_choice", | |
| "repeats": 1, | |
| "should_decontaminate": true, | |
| "doc_to_decontamination_query": "Question: {{question}}\nAnswer:", | |
| "metadata": { | |
| "version": 1.0 | |
| } | |
| } | |
| }, | |
| "versions": { | |
| "arc_challenge": 1.0 | |
| }, | |
| "n-shot": { | |
| "arc_challenge": 25 | |
| }, | |
| "config": { | |
| "model": "vllm", | |
| "model_args": "pretrained=/lustre07/scratch/gagan30/arocr/meta-llama/self_rewarding_models/Oasis,tensor_parallel_size=1,dtype=auto,gpu_memory_utilization=0.9,data_parallel_size=1,max_model_len=4096", | |
| "batch_size": "auto:128", | |
| "batch_sizes": [], | |
| "device": "cuda", | |
| "use_cache": "/lustre07/scratch/gagan30/arocr/cache/", | |
| "limit": null, | |
| "bootstrap_iters": 100000, | |
| "gen_kwargs": null | |
| }, | |
| "git_hash": null | |
| } |