nohup: ignoring input Loading Qwen2.5-Coder-1.5B... Loading weights: 0%| | 0/338 [00:00= 0) instead. torch._check_is_size(blocksize) Loading weights: 1%| | 3/338 [00:00<00:24, 13.50it/s] Loading weights: 12%|█▏ | 39/338 [00:00<00:02, 148.66it/s] Loading weights: 22%|██▏ | 75/338 [00:00<00:01, 222.54it/s] Loading weights: 33%|███▎ | 112/338 [00:00<00:00, 271.67it/s] Loading weights: 44%|████▍ | 149/338 [00:00<00:00, 302.31it/s] Loading weights: 57%|█████▋ | 191/338 [00:00<00:00, 339.33it/s] Loading weights: 68%|██████▊ | 231/338 [00:00<00:00, 356.97it/s] Loading weights: 80%|███████▉ | 269/338 [00:00<00:00, 357.55it/s] Loading weights: 93%|█████████▎| 315/338 [00:01<00:00, 380.07it/s] Loading weights: 100%|██████████| 338/338 [00:01<00:00, 309.17it/s] [transformers] The tokenizer has new PAD/BOS/EOS tokens that differ from the model config and generation config. The model config and generation config were aligned accordingly, being updated with the tokenizer's values. Updated tokens: {'bos_token_id': None, 'pad_token_id': 151645}. Model loaded. Training... 0%| | 0/300 [00:00