2025-08-25 19:32:25 - eval_util - INFO - Using AWQ quantization.
2025-08-25 19:32:25 - eval_util - INFO - Use fixed channel rescale: "124,90,62,52,86,108,118,75,114,120,50,93,115,73,58,81,80,94,89,69,63,51,46,106,107,84,44,42,123,68,88,15,6,28,12,17,1,20,8,35" with scale 2.0
2025-08-25 19:32:25 - 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 19:32:28 - eval_util - ERROR - Exception Occured:
Traceback (most recent call last):
  File "/home/yeq6/Research_project/llama/llm-awq/awq/quantize/auto_scale.py", line 582, in apply_scale
    individual_channel_list = [int(x) for x in args.individual_channel_up.split(',')]
  File "/home/yeq6/Research_project/llama/llm-awq/awq/quantize/auto_scale.py", line 582, in <listcomp>
    individual_channel_list = [int(x) for x in args.individual_channel_up.split(',')]
ValueError: invalid literal for int() with base 10: '"124'

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "/home/yeq6/Research_project/llama/Benchmark_all.py", line 139, in <module>
    main(add_args(parser).parse_args())
  File "/home/yeq6/Research_project/llama/Benchmark_all.py", line 62, in main
    model, tokenizer = load_model_and_apply_patches_original_awq(args.model[0][0],
  File "/home/yeq6/Research_project/llama/model_loader.py", line 694, in load_model_and_apply_patches_original_awq
    return apply_patches(load_model_awq(model, args, quant_path, awq_cache, temp), args)
  File "/home/yeq6/Research_project/llama/model_loader.py", line 274, in load_model_awq
    apply_awq(loaded, awq_results, beta_point=args.beta_point, args=args)
  File "/home/yeq6/Research_project/llama/llm-awq/awq/quantize/pre_quant.py", line 322, in apply_awq
    apply_scale(model, awq_results["scale"], beta_point = beta_point, args=args, temp = temp)
  File "/home/yeq6/Research_project/llama/llm-awq/awq/quantize/auto_scale.py", line 584, in apply_scale
    raise ValueError("All elements in the individual_channel list must be integers.")
ValueError: All elements in the individual_channel list must be integers.
