2025-08-17 03:42:14 - eval_util - INFO - Using naive RTN quantization.
2025-08-17 03:42:14 - eval_util - INFO - Applying Hardmard quantization to layers: o_proj
2025-08-17 03:45:25 - accelerate.utils.modeling - INFO - We will use 90% of the memory on device 0 for storing the model, and 10% for the buffer to avoid OOM. You can set `max_memory` in to a higher value to use more memory (at your own risk).
2025-08-17 03:45:25 - eval_util - INFO - Evaluating wikitext...
2025-08-17 03:46:39 - eval_util - INFO - Wikitext completed: 6.387182235717773
2025-08-17 03:46:39 - eval_util - INFO - Starting evaluation of task: boolq
2025-08-17 03:46:39 - eval_util - INFO - GPU Memory before boolq: 5.81GB
2025-08-17 03:46:39 - lm_eval.models.huggingface - WARNING - `pretrained` model kwarg is not of type `str`. Many other model arguments may be ignored. Please do not launch via accelerate or use `parallelize=True` if passing an existing model this way.
2025-08-17 03:46:39 - lm_eval.models.huggingface - WARNING - Passed an already-initialized model through `pretrained`, assuming single-process call to evaluate() or custom distributed integration
2025-08-17 03:46:39 - lm_eval.evaluator - INFO - Setting random seed to 0 | Setting numpy seed to 1234 | Setting torch manual seed to 1234 | Setting fewshot manual seed to 1234
2025-08-17 03:46:39 - lm_eval.evaluator - INFO - Using pre-initialized model
2025-08-17 03:46:54 - lm_eval.api.task - WARNING - [Task: boolq] metric acc is defined, but aggregation is not. using default aggregation=mean
2025-08-17 03:46:54 - lm_eval.api.task - WARNING - [Task: boolq] metric acc is defined, but higher_is_better is not. using default higher_is_better=True
2025-08-17 03:46:58 - lm_eval.api.task - INFO - Building contexts for boolq on rank 0...
2025-08-17 03:46:59 - lm_eval.evaluator - INFO - Running loglikelihood requests
2025-08-17 03:50:41 - eval_util - INFO - ✅ Task boolq completed successfully
2025-08-17 03:50:41 - eval_util - INFO - GPU Memory after boolq: 5.81GB
2025-08-17 03:50:41 - eval_util - INFO - Starting evaluation of task: hellaswag
2025-08-17 03:50:41 - eval_util - INFO - GPU Memory before hellaswag: 5.81GB
2025-08-17 03:50:41 - lm_eval.models.huggingface - WARNING - `pretrained` model kwarg is not of type `str`. Many other model arguments may be ignored. Please do not launch via accelerate or use `parallelize=True` if passing an existing model this way.
2025-08-17 03:50:41 - lm_eval.models.huggingface - WARNING - Passed an already-initialized model through `pretrained`, assuming single-process call to evaluate() or custom distributed integration
2025-08-17 03:50:41 - lm_eval.evaluator - INFO - Setting random seed to 0 | Setting numpy seed to 1234 | Setting torch manual seed to 1234 | Setting fewshot manual seed to 1234
2025-08-17 03:50:41 - lm_eval.evaluator - INFO - Using pre-initialized model
2025-08-17 03:50:59 - lm_eval.api.task - INFO - Building contexts for hellaswag on rank 0...
2025-08-17 03:51:03 - lm_eval.evaluator - INFO - Running loglikelihood requests
2025-08-17 04:25:39 - eval_util - INFO - ✅ Task hellaswag completed successfully
2025-08-17 04:25:39 - eval_util - INFO - GPU Memory after hellaswag: 5.81GB
2025-08-17 04:25:39 - eval_util - INFO - Starting evaluation of task: winogrande
2025-08-17 04:25:39 - eval_util - INFO - GPU Memory before winogrande: 5.81GB
2025-08-17 04:25:39 - lm_eval.models.huggingface - WARNING - `pretrained` model kwarg is not of type `str`. Many other model arguments may be ignored. Please do not launch via accelerate or use `parallelize=True` if passing an existing model this way.
2025-08-17 04:25:39 - lm_eval.models.huggingface - WARNING - Passed an already-initialized model through `pretrained`, assuming single-process call to evaluate() or custom distributed integration
2025-08-17 04:25:39 - lm_eval.evaluator - INFO - Setting random seed to 0 | Setting numpy seed to 1234 | Setting torch manual seed to 1234 | Setting fewshot manual seed to 1234
2025-08-17 04:25:39 - lm_eval.evaluator - INFO - Using pre-initialized model
2025-08-17 04:25:55 - lm_eval.api.task - INFO - Building contexts for winogrande on rank 0...
2025-08-17 04:25:55 - lm_eval.evaluator - INFO - Running loglikelihood requests
2025-08-17 04:27:45 - eval_util - INFO - ✅ Task winogrande completed successfully
2025-08-17 04:27:45 - eval_util - INFO - GPU Memory after winogrande: 5.81GB
2025-08-17 04:27:45 - eval_util - INFO - Starting evaluation of task: arc_easy
2025-08-17 04:27:45 - eval_util - INFO - GPU Memory before arc_easy: 5.81GB
2025-08-17 04:27:45 - lm_eval.models.huggingface - WARNING - `pretrained` model kwarg is not of type `str`. Many other model arguments may be ignored. Please do not launch via accelerate or use `parallelize=True` if passing an existing model this way.
2025-08-17 04:27:45 - lm_eval.models.huggingface - WARNING - Passed an already-initialized model through `pretrained`, assuming single-process call to evaluate() or custom distributed integration
2025-08-17 04:27:45 - lm_eval.evaluator - INFO - Setting random seed to 0 | Setting numpy seed to 1234 | Setting torch manual seed to 1234 | Setting fewshot manual seed to 1234
2025-08-17 04:27:45 - lm_eval.evaluator - INFO - Using pre-initialized model
2025-08-17 04:27:59 - lm_eval.api.task - INFO - Building contexts for arc_easy on rank 0...
2025-08-17 04:28:01 - lm_eval.evaluator - INFO - Running loglikelihood requests
2025-08-17 04:34:37 - eval_util - INFO - ✅ Task arc_easy completed successfully
2025-08-17 04:34:37 - eval_util - INFO - GPU Memory after arc_easy: 5.81GB
2025-08-17 04:34:37 - eval_util - INFO - Starting evaluation of task: arc_challenge
2025-08-17 04:34:37 - eval_util - INFO - GPU Memory before arc_challenge: 5.81GB
2025-08-17 04:34:37 - lm_eval.models.huggingface - WARNING - `pretrained` model kwarg is not of type `str`. Many other model arguments may be ignored. Please do not launch via accelerate or use `parallelize=True` if passing an existing model this way.
2025-08-17 04:34:37 - lm_eval.models.huggingface - WARNING - Passed an already-initialized model through `pretrained`, assuming single-process call to evaluate() or custom distributed integration
2025-08-17 04:34:37 - lm_eval.evaluator - INFO - Setting random seed to 0 | Setting numpy seed to 1234 | Setting torch manual seed to 1234 | Setting fewshot manual seed to 1234
2025-08-17 04:34:37 - lm_eval.evaluator - INFO - Using pre-initialized model
2025-08-17 04:34:51 - lm_eval.api.task - INFO - Building contexts for arc_challenge on rank 0...
2025-08-17 04:34:52 - lm_eval.evaluator - INFO - Running loglikelihood requests
2025-08-17 04:38:19 - eval_util - INFO - ✅ Task arc_challenge completed successfully
2025-08-17 04:38:19 - eval_util - INFO - GPU Memory after arc_challenge: 5.81GB
2025-08-17 04:38:19 - eval_util - INFO - Evaluation Summary - Successful: 6, Failed: 0
2025-08-17 04:38:19 - eval_util - INFO - result:
2025-08-17 04:38:19 - eval_util - INFO - 	{'arc_challenge': {'acc,none': 0.42235494880546076,
2025-08-17 04:38:19 - eval_util - INFO - 	                   'acc_norm,none': 0.4402730375426621,
2025-08-17 04:38:19 - eval_util - INFO - 	                   'acc_norm_stderr,none': 0.014506769524804236,
2025-08-17 04:38:19 - eval_util - INFO - 	                   'acc_stderr,none': 0.01443413871337999,
2025-08-17 04:38:19 - eval_util - INFO - 	                   'alias': 'arc_challenge'},
2025-08-17 04:38:19 - eval_util - INFO - 	 'arc_easy': {'acc,none': 0.752104377104377,
2025-08-17 04:38:19 - eval_util - INFO - 	              'acc_norm,none': 0.7218013468013468,
2025-08-17 04:38:19 - eval_util - INFO - 	              'acc_norm_stderr,none': 0.009195059601583897,
2025-08-17 04:38:19 - eval_util - INFO - 	              'acc_stderr,none': 0.008860162361464039,
2025-08-17 04:38:19 - eval_util - INFO - 	              'alias': 'arc_easy'},
2025-08-17 04:38:19 - eval_util - INFO - 	 'boolq': {'acc,none': 0.7394495412844037,
2025-08-17 04:38:19 - eval_util - INFO - 	           'acc_stderr,none': 0.007677021072511173,
2025-08-17 04:38:19 - eval_util - INFO - 	           'alias': 'boolq'},
2025-08-17 04:38:19 - eval_util - INFO - 	 'hellaswag': {'acc,none': 0.5650268870742879,
2025-08-17 04:38:19 - eval_util - INFO - 	               'acc_norm,none': 0.7587134037044413,
2025-08-17 04:38:19 - eval_util - INFO - 	               'acc_norm_stderr,none': 0.004269893011589029,
2025-08-17 04:38:19 - eval_util - INFO - 	               'acc_stderr,none': 0.004947402907996247,
2025-08-17 04:38:19 - eval_util - INFO - 	               'alias': 'hellaswag'},
2025-08-17 04:38:19 - eval_util - INFO - 	 'wikitext': 6.387182235717773,
2025-08-17 04:38:19 - eval_util - INFO - 	 'winogrande': {'acc,none': 0.6819258089976322,
2025-08-17 04:38:19 - eval_util - INFO - 	                'acc_stderr,none': 0.013089285079884678,
2025-08-17 04:38:19 - eval_util - INFO - 	                'alias': 'winogrande'}}
