================================================== 开始评估:任务=gsm8k | 少样本数=4 | 模型=Qwen2.5-7B 输出路径:results2/Qwen2.5-7B/base_gsm8k.json ================================================== 2025-12-01:20:07:52 INFO [__main__:440] Selected Tasks: ['gsm8k'] 2025-12-01:20:07:52 INFO [evaluator:189] Setting random seed to 0 | Setting numpy seed to 1234 | Setting torch manual seed to 1234 | Setting fewshot manual seed to 1234 2025-12-01:20:07:52 INFO [evaluator:227] Initializing hf model, with arguments: {'pretrained': '/mnt/bn/life-mllm/users/cxr/quantization/models/Qwen/Qwen2.5-7B'} 2025-12-01:20:07:52 WARNING [accelerate.utils.other:513] Detected kernel version 5.4.143, which is below the recommended minimum of 5.5.0; this can cause the process to hang. It is recommended to upgrade the kernel to the minimum version or higher. 2025-12-01:20:07:52 INFO [models.huggingface:137] Using device 'cuda:3' 2025-12-01:20:07:52 INFO [models.huggingface:382] Model parallel was set to False, max memory was not set, and device map was set to {'': 'cuda:3'} `torch_dtype` is deprecated! Use `dtype` instead! Loading checkpoint shards: 0%| | 0/4 [00:00', '<|im_end|>'], 'do_sample': False, 'temperature': 0.0} 2025-12-01:20:08:17 WARNING [evaluator:309] Overwriting default num_fewshot of gsm8k from 5 to 4 2025-12-01:20:08:17 INFO [api.task:434] Building contexts for gsm8k on rank 0... 0%| | 0/1319 [00:00