2025-08-25 17:46:36 - eval_util - INFO - Using AWQ quantization.
2025-08-25 17:46:37 - 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).
2025-08-25 17:49:56 - 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).
2025-08-25 17:49:56 - eval_util - INFO - Evaluating wikitext...
2025-08-25 17:51:02 - eval_util - INFO - Wikitext completed: 6.311136722564697
2025-08-25 17:51:02 - eval_util - INFO - Starting evaluation of task: boolq
2025-08-25 17:51:02 - eval_util - INFO - GPU Memory before boolq: 3.81GB
2025-08-25 17:51:02 - 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.
2025-08-25 17:51:02 - lm_eval.models.huggingface - WARNING - Passed an already-initialized model through `pretrained`, assuming single-process call to evaluate() or custom distributed integration
2025-08-25 17:51:02 - 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
2025-08-25 17:51:02 - lm_eval.evaluator - INFO - Using pre-initialized model
2025-08-25 17:51:17 - lm_eval.api.task - WARNING - [Task: boolq] metric acc is defined, but aggregation is not. using default aggregation=mean
2025-08-25 17:51:17 - lm_eval.api.task - WARNING - [Task: boolq] metric acc is defined, but higher_is_better is not. using default higher_is_better=True
2025-08-25 17:51:20 - lm_eval.api.task - INFO - Building contexts for boolq on rank 0...
2025-08-25 17:51:22 - lm_eval.evaluator - INFO - Running loglikelihood requests
2025-08-25 17:54:41 - eval_util - INFO - ✅ Task boolq completed successfully
2025-08-25 17:54:42 - eval_util - INFO - GPU Memory after boolq: 3.81GB
2025-08-25 17:54:42 - eval_util - INFO - Starting evaluation of task: hellaswag
2025-08-25 17:54:42 - eval_util - INFO - GPU Memory before hellaswag: 3.81GB
2025-08-25 17:54:42 - 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.
2025-08-25 17:54:42 - lm_eval.models.huggingface - WARNING - Passed an already-initialized model through `pretrained`, assuming single-process call to evaluate() or custom distributed integration
2025-08-25 17:54:42 - 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
2025-08-25 17:54:42 - lm_eval.evaluator - INFO - Using pre-initialized model
2025-08-25 17:55:01 - lm_eval.api.task - INFO - Building contexts for hellaswag on rank 0...
2025-08-25 17:55:05 - lm_eval.evaluator - INFO - Running loglikelihood requests
2025-08-25 18:25:39 - eval_util - INFO - ✅ Task hellaswag completed successfully
2025-08-25 18:25:39 - eval_util - INFO - GPU Memory after hellaswag: 3.81GB
2025-08-25 18:25:39 - eval_util - INFO - Starting evaluation of task: winogrande
2025-08-25 18:25:39 - eval_util - INFO - GPU Memory before winogrande: 3.81GB
2025-08-25 18:25:39 - 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.
2025-08-25 18:25:39 - lm_eval.models.huggingface - WARNING - Passed an already-initialized model through `pretrained`, assuming single-process call to evaluate() or custom distributed integration
2025-08-25 18:25:39 - 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
2025-08-25 18:25:39 - lm_eval.evaluator - INFO - Using pre-initialized model
2025-08-25 18:25:55 - lm_eval.api.task - INFO - Building contexts for winogrande on rank 0...
2025-08-25 18:25:55 - lm_eval.evaluator - INFO - Running loglikelihood requests
2025-08-25 18:27:33 - eval_util - INFO - ✅ Task winogrande completed successfully
2025-08-25 18:27:33 - eval_util - INFO - GPU Memory after winogrande: 3.81GB
2025-08-25 18:27:33 - eval_util - INFO - Starting evaluation of task: arc_easy
2025-08-25 18:27:33 - eval_util - INFO - GPU Memory before arc_easy: 3.81GB
2025-08-25 18:27:33 - 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.
2025-08-25 18:27:33 - lm_eval.models.huggingface - WARNING - Passed an already-initialized model through `pretrained`, assuming single-process call to evaluate() or custom distributed integration
2025-08-25 18:27:33 - 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
2025-08-25 18:27:33 - lm_eval.evaluator - INFO - Using pre-initialized model
2025-08-25 18:27:48 - lm_eval.api.task - INFO - Building contexts for arc_easy on rank 0...
2025-08-25 18:27:50 - lm_eval.evaluator - INFO - Running loglikelihood requests
2025-08-25 18:33:42 - eval_util - INFO - ✅ Task arc_easy completed successfully
2025-08-25 18:33:42 - eval_util - INFO - GPU Memory after arc_easy: 3.81GB
2025-08-25 18:33:42 - eval_util - INFO - Starting evaluation of task: arc_challenge
2025-08-25 18:33:42 - eval_util - INFO - GPU Memory before arc_challenge: 3.81GB
2025-08-25 18:33:42 - 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.
2025-08-25 18:33:42 - lm_eval.models.huggingface - WARNING - Passed an already-initialized model through `pretrained`, assuming single-process call to evaluate() or custom distributed integration
2025-08-25 18:33:42 - 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
2025-08-25 18:33:42 - lm_eval.evaluator - INFO - Using pre-initialized model
2025-08-25 18:33:56 - lm_eval.api.task - INFO - Building contexts for arc_challenge on rank 0...
2025-08-25 18:33:56 - lm_eval.evaluator - INFO - Running loglikelihood requests
2025-08-25 18:37:00 - eval_util - INFO - ✅ Task arc_challenge completed successfully
2025-08-25 18:37:00 - eval_util - INFO - GPU Memory after arc_challenge: 3.81GB
2025-08-25 18:37:00 - eval_util - INFO - Evaluation Summary - Successful: 6, Failed: 0
2025-08-25 18:37:00 - eval_util - INFO - result:
2025-08-25 18:37:00 - eval_util - INFO - 	{'arc_challenge': {'acc,none': 0.4197952218430034,
2025-08-25 18:37:00 - eval_util - INFO - 	                   'acc_norm,none': 0.4377133105802048,
2025-08-25 18:37:00 - eval_util - INFO - 	                   'acc_norm_stderr,none': 0.014497573881108283,
2025-08-25 18:37:00 - eval_util - INFO - 	                   'acc_stderr,none': 0.014422181226303026,
2025-08-25 18:37:00 - eval_util - INFO - 	                   'alias': 'arc_challenge'},
2025-08-25 18:37:00 - eval_util - INFO - 	 'arc_easy': {'acc,none': 0.7441077441077442,
2025-08-25 18:37:00 - eval_util - INFO - 	              'acc_norm,none': 0.7129629629629629,
2025-08-25 18:37:00 - eval_util - INFO - 	              'acc_norm_stderr,none': 0.009282621598983075,
2025-08-25 18:37:00 - eval_util - INFO - 	              'acc_stderr,none': 0.008953950243013995,
2025-08-25 18:37:00 - eval_util - INFO - 	              'alias': 'arc_easy'},
2025-08-25 18:37:00 - eval_util - INFO - 	 'boolq': {'acc,none': 0.745565749235474,
2025-08-25 18:37:00 - eval_util - INFO - 	           'acc_stderr,none': 0.007617690099234388,
2025-08-25 18:37:00 - eval_util - INFO - 	           'alias': 'boolq'},
2025-08-25 18:37:00 - eval_util - INFO - 	 'hellaswag': {'acc,none': 0.5617406891057558,
2025-08-25 18:37:00 - eval_util - INFO - 	               'acc_norm,none': 0.7539334793865764,
2025-08-25 18:37:00 - eval_util - INFO - 	               'acc_norm_stderr,none': 0.004298374936365597,
2025-08-25 18:37:00 - eval_util - INFO - 	               'acc_stderr,none': 0.004951594063272069,
2025-08-25 18:37:00 - eval_util - INFO - 	               'alias': 'hellaswag'},
2025-08-25 18:37:00 - eval_util - INFO - 	 'wikitext': 6.311136722564697,
2025-08-25 18:37:00 - eval_util - INFO - 	 'winogrande': {'acc,none': 0.6850828729281768,
2025-08-25 18:37:00 - eval_util - INFO - 	                'acc_stderr,none': 0.013054277568469242,
2025-08-25 18:37:00 - eval_util - INFO - 	                'alias': 'winogrande'}}
