| { | |
| "results": { | |
| "arc_pt": { | |
| "acc": 0.2743589743589744, | |
| "acc_stderr": 0.013050082103110318, | |
| "acc_norm": 0.32051282051282054, | |
| "acc_norm_stderr": 0.01364916490674831 | |
| }, | |
| "hellaswag_pt": { | |
| "acc": 0.38173149853721966, | |
| "acc_stderr": 0.00505724235256629, | |
| "acc_norm": 0.4828258749593672, | |
| "acc_norm_stderr": 0.005201874489478013 | |
| }, | |
| "truthfulqa_pt": { | |
| "mc1": 0.23857868020304568, | |
| "mc1_stderr": 0.015192910034567013, | |
| "mc2": 0.3843938264989383, | |
| "mc2_stderr": 0.014578530843614249 | |
| } | |
| }, | |
| "versions": { | |
| "arc_pt": 0, | |
| "hellaswag_pt": 1, | |
| "truthfulqa_pt": 1 | |
| }, | |
| "config": { | |
| "model": "hf-auto", | |
| "model_args": "pretrained=/lustre/mlnvme/data/asen_hpc-mula/Tucano-2b4-DPO", | |
| "batch_size": 1, | |
| "device": "cuda:0", | |
| "no_cache": false, | |
| "limit": null, | |
| "bootstrap_iters": 100000, | |
| "description_dict": {} | |
| } | |
| } |