2026-05-30:13:24:47 INFO [_cli.run:388] Selected Tasks: ['mmlu'] 2026-05-30:13:24:48 INFO [evaluator:214] Setting random seed to 0 | Setting numpy seed to 1234 | Setting torch manual seed to 1234 | Setting fewshot manual seed to 1234 2026-05-30:13:24:48 INFO [evaluator:239] Initializing hf model, with arguments: {'pretrained': '/workspace/vasttrain/output/checkpoint_step161', 'dtype': 'bfloat16', 'parallelize': True} 2026-05-30:13:24:52 INFO [models.huggingface:306] Using `accelerate launch` or `parallelize=True`, device 'cuda:0' will be overridden when placing model. 2026-05-30:13:24:53 INFO [models.huggingface:562] Model parallel was set to True, setting max memory per GPU to {0: 286762860544, 1: 286762860544} and device map to auto Loading weights: 0%| | 0/443 [00:00