2025-08-17 02:33:17 - eval_util - INFO - Using naive RTN quantization.
2025-08-17 02:33:17 - eval_util - INFO - Applying Hardmard quantization to layers: q_proj,k_proj,v_proj
2025-08-17 02:36:43 - 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 02:36:43 - eval_util - INFO - Evaluating wikitext...
2025-08-17 02:38:14 - eval_util - INFO - Wikitext completed: 6.434990882873535
2025-08-17 02:38:14 - eval_util - INFO - Starting evaluation of task: boolq
2025-08-17 02:38:14 - eval_util - INFO - GPU Memory before boolq: 9.81GB
2025-08-17 02:38:14 - 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 02:38:14 - 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 02:38:14 - 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 02:38:14 - lm_eval.evaluator - INFO - Using pre-initialized model
2025-08-17 02:38:30 - lm_eval.api.task - WARNING - [Task: boolq] metric acc is defined, but aggregation is not. using default aggregation=mean
2025-08-17 02:38:30 - 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 02:38:33 - lm_eval.api.task - INFO - Building contexts for boolq on rank 0...
2025-08-17 02:38:35 - lm_eval.evaluator - INFO - Running loglikelihood requests
2025-08-17 02:43:05 - eval_util - INFO - ✅ Task boolq completed successfully
2025-08-17 02:43:05 - eval_util - INFO - GPU Memory after boolq: 9.81GB
2025-08-17 02:43:05 - eval_util - INFO - Starting evaluation of task: hellaswag
2025-08-17 02:43:05 - eval_util - INFO - GPU Memory before hellaswag: 9.81GB
2025-08-17 02:43:05 - 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 02:43:05 - 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 02:43:05 - 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 02:43:05 - lm_eval.evaluator - INFO - Using pre-initialized model
2025-08-17 02:43:24 - lm_eval.api.task - INFO - Building contexts for hellaswag on rank 0...
2025-08-17 02:43:28 - lm_eval.evaluator - INFO - Running loglikelihood requests
2025-08-17 03:26:26 - eval_util - INFO - ✅ Task hellaswag completed successfully
2025-08-17 03:26:26 - eval_util - INFO - GPU Memory after hellaswag: 9.81GB
2025-08-17 03:26:26 - eval_util - INFO - Starting evaluation of task: winogrande
2025-08-17 03:26:26 - eval_util - INFO - GPU Memory before winogrande: 9.81GB
2025-08-17 03:26:26 - 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 03:26:26 - 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 03:26:26 - 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 03:26:26 - lm_eval.evaluator - INFO - Using pre-initialized model
2025-08-17 03:26:43 - lm_eval.api.task - INFO - Building contexts for winogrande on rank 0...
2025-08-17 03:26:43 - lm_eval.evaluator - INFO - Running loglikelihood requests
2025-08-17 03:28:58 - eval_util - INFO - ✅ Task winogrande completed successfully
2025-08-17 03:28:58 - eval_util - INFO - GPU Memory after winogrande: 9.81GB
2025-08-17 03:28:58 - eval_util - INFO - Starting evaluation of task: arc_easy
2025-08-17 03:28:58 - eval_util - INFO - GPU Memory before arc_easy: 9.81GB
2025-08-17 03:28:58 - 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 03:28:58 - 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 03:28:58 - 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 03:28:58 - lm_eval.evaluator - INFO - Using pre-initialized model
2025-08-17 03:29:13 - lm_eval.api.task - INFO - Building contexts for arc_easy on rank 0...
2025-08-17 03:29:15 - lm_eval.evaluator - INFO - Running loglikelihood requests
2025-08-17 03:37:31 - eval_util - INFO - ✅ Task arc_easy completed successfully
2025-08-17 03:37:31 - eval_util - INFO - GPU Memory after arc_easy: 9.81GB
2025-08-17 03:37:31 - eval_util - INFO - Starting evaluation of task: arc_challenge
2025-08-17 03:37:31 - eval_util - INFO - GPU Memory before arc_challenge: 9.81GB
2025-08-17 03:37:31 - 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 03:37:31 - 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 03:37:31 - 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 03:37:31 - lm_eval.evaluator - INFO - Using pre-initialized model
2025-08-17 03:37:44 - lm_eval.api.task - INFO - Building contexts for arc_challenge on rank 0...
2025-08-17 03:37:45 - lm_eval.evaluator - INFO - Running loglikelihood requests
2025-08-17 03:42:04 - eval_util - INFO - ✅ Task arc_challenge completed successfully
2025-08-17 03:42:04 - eval_util - INFO - GPU Memory after arc_challenge: 9.81GB
2025-08-17 03:42:04 - eval_util - INFO - Evaluation Summary - Successful: 6, Failed: 0
2025-08-17 03:42:04 - eval_util - INFO - result:
2025-08-17 03:42:04 - eval_util - INFO - 	{'arc_challenge': {'acc,none': 0.4104095563139932,
2025-08-17 03:42:04 - eval_util - INFO - 	                   'acc_norm,none': 0.431740614334471,
2025-08-17 03:42:04 - eval_util - INFO - 	                   'acc_norm_stderr,none': 0.014474591427196202,
2025-08-17 03:42:04 - eval_util - INFO - 	                   'acc_stderr,none': 0.01437492219264266,
2025-08-17 03:42:04 - eval_util - INFO - 	                   'alias': 'arc_challenge'},
2025-08-17 03:42:04 - eval_util - INFO - 	 'arc_easy': {'acc,none': 0.7462121212121212,
2025-08-17 03:42:04 - eval_util - INFO - 	              'acc_norm,none': 0.7074915824915825,
2025-08-17 03:42:04 - eval_util - INFO - 	              'acc_norm_stderr,none': 0.009334649503078416,
2025-08-17 03:42:04 - eval_util - INFO - 	              'acc_stderr,none': 0.008929657065808297,
2025-08-17 03:42:04 - eval_util - INFO - 	              'alias': 'arc_easy'},
2025-08-17 03:42:04 - eval_util - INFO - 	 'boolq': {'acc,none': 0.736697247706422,
2025-08-17 03:42:04 - eval_util - INFO - 	           'acc_stderr,none': 0.007703086288558766,
2025-08-17 03:42:04 - eval_util - INFO - 	           'alias': 'boolq'},
2025-08-17 03:42:04 - eval_util - INFO - 	 'hellaswag': {'acc,none': 0.5653256323441546,
2025-08-17 03:42:04 - eval_util - INFO - 	               'acc_norm,none': 0.7570205138418642,
2025-08-17 03:42:04 - eval_util - INFO - 	               'acc_norm_stderr,none': 0.0042800628384464905,
2025-08-17 03:42:04 - eval_util - INFO - 	               'acc_stderr,none': 0.004947010937455354,
2025-08-17 03:42:04 - eval_util - INFO - 	               'alias': 'hellaswag'},
2025-08-17 03:42:04 - eval_util - INFO - 	 'wikitext': 6.434990882873535,
2025-08-17 03:42:04 - eval_util - INFO - 	 'winogrande': {'acc,none': 0.6803472770323599,
2025-08-17 03:42:04 - eval_util - INFO - 	                'acc_stderr,none': 0.013106528517665144,
2025-08-17 03:42:04 - eval_util - INFO - 	                'alias': 'winogrande'}}
