[help] RuntimeError: "unfolded2d_copy" not implemented for 'Half'怎么解决

[Gary]

报错详情

  File "web_demo.py", line 58, in request_model
    answer = generate_text_with_image(input_text, image, result_text.copy(), request_para, is_zh)
  File "web_demo.py", line 31, in generate_text_with_image
    answer, history, _ = chat(None, model, tokenizer, input_text, history=history, image=input_image, \
  File "D:\projects\XrayGLM-main\model\chat.py", line 141, in chat
    output = filling_sequence(
  File "E:\Anaconda3\envs\xglm\lib\site-packages\sat\generation\autoregressive_sampling.py", line 108, in filling_sequence
    logits, *output_per_layers = model(
  File "E:\Anaconda3\envs\xglm\lib\site-packages\torch\nn\modules\module.py", line 1102, in _call_impl
    return forward_call(*input, **kwargs)
  File "E:\Anaconda3\envs\xglm\lib\site-packages\sat\model\official\chatglm_model.py", line 192, in forward
    return super().forward(input_ids=input_ids, attention_mask=attention_mask, position_ids=position_ids, past_key_values=past_key_values, **kwargs)
  File "E:\Anaconda3\envs\xglm\lib\site-packages\sat\model\base_model.py", line 144, in forward
    return self.transformer(*args, **kwargs)
  File "E:\Anaconda3\envs\xglm\lib\site-packages\torch\nn\modules\module.py", line 1102, in _call_impl
    return forward_call(*input, **kwargs)
  File "E:\Anaconda3\envs\xglm\lib\site-packages\sat\model\transformer.py", line 451, in forward
    hidden_states = self.hooks['word_embedding_forward'](input_ids, output_cross_layer=output_cross_layer, **kw_args)
  File "D:\projects\XrayGLM-main\model\visualglm.py", line 20, in word_embedding_forward
    image_emb = self.model(**kw_args)
  File "E:\Anaconda3\envs\xglm\lib\site-packages\torch\nn\modules\module.py", line 1102, in _call_impl
    return forward_call(*input, **kwargs)
  File "D:\projects\XrayGLM-main\model\blip2.py", line 65, in forward
    enc = self.vit(image)[0]
  File "E:\Anaconda3\envs\xglm\lib\site-packages\torch\nn\modules\module.py", line 1102, in _call_impl
    return forward_call(*input, **kwargs)
  File "D:\projects\XrayGLM-main\model\blip2.py", line 29, in forward
    return super().forward(input_ids=input_ids, position_ids=None, attention_mask=attention_mask, image=image)
  File "E:\Anaconda3\envs\xglm\lib\site-packages\sat\model\base_model.py", line 144, in forward
    return self.transformer(*args, **kwargs)
  File "E:\Anaconda3\envs\xglm\lib\site-packages\torch\nn\modules\module.py", line 1102, in _call_impl
    return forward_call(*input, **kwargs)
  File "E:\Anaconda3\envs\xglm\lib\site-packages\sat\model\transformer.py", line 451, in forward
    hidden_states = self.hooks['word_embedding_forward'](input_ids, output_cross_layer=output_cross_layer, **kw_args)
  File "E:\Anaconda3\envs\xglm\lib\site-packages\sat\model\official\vit_model.py", line 54, in word_embedding_forward
    embeddings = self.proj(images)
  File "E:\Anaconda3\envs\xglm\lib\site-packages\torch\nn\modules\module.py", line 1102, in _call_impl
    return forward_call(*input, **kwargs)
  File "E:\Anaconda3\envs\xglm\lib\site-packages\torch\nn\modules\conv.py", line 446, in forward
    return self._conv_forward(input, self.weight, self.bias)
  File "E:\Anaconda3\envs\xglm\lib\site-packages\torch\nn\modules\conv.py", line 442, in _conv_forward
    return F.conv2d(input, weight, bias, self.stride,
RuntimeError: "unfolded2d_copy" not implemented for 'Half'

环境：win10 python3.8.5
尝试：网上说是pytorch不支持cpu fp16 把half改为float即可 但是改了还是一样的报错
求解决方法

[Gary]

未通过代码解决 升级了硬件解决