TMElyralab

/

lyraChatGLM

like 107

Follow

Tencent Music Entertainment Lyra Lab 43

环境：Anaconda ， Python 3.10.11 ， win10 。希望得到您的帮助，谢谢！

系统弹窗如上，同时报错如下：

(lyraChatGLM) D:\lyraChatGLM>python demo.py
Loading tokenizer from .
[INFO] Load Our Highly Optimized LyraChatGLM6B model

• head_num.................: 32
• size_per_head............: 128
• layer_num................: 28
• tensor_para_size.........: 1
• vocab_size...............: 130528
• start_id.................: 130004
• end_id...................: 130005
• weights_data_type........: fp16
• layernorm_eps............: 1e-05
• inference_data_type......: fp16
• rotary_embedding_dim.....: 64
• max_seq_len..............: 0
• pipeline_para_size.......: 1
• shared_contexts_ratio....: 0.0
• int8_mode................: 0
• model_path...............: .\models\1-gpu-fp16.bin
• cuda_version.............: 11
  Loading tokenizer from .\models\1-gpu-fp16.bin
  Traceback (most recent call last):
  File "D:\lyraChatGLM\demo.py", line 12, in
  model = LyraChatGLM6B(model_path, tokenizer_path, data_type, int8_mode, arch, cuda_version)
  File "D:\lyraChatGLM\lyraChatGLM\lyra_glm.py", line 24, in init
  self.model, self.tokenizer = self.load_model_and_tokenizer()
  File "D:\lyraChatGLM\lyraChatGLM\lyra_glm.py", line 108, in load_model_and_tokenizer
  model = ChatGLM6BModel(arch=self.arch,**model_args)
  File "D:\lyraChatGLM\lyraChatGLM\model.py", line 103, in init
  torch.classes.load_library(os.path.abspath(lib_path))
  File "C:\Users\ushow_AI.conda\envs\lyraChatGLM\lib\site-packages\torch_classes.py", line 51, in load_library
  torch.ops.load_library(path)
  File "C:\Users\ushow_AI.conda\envs\lyraChatGLM\lib\site-packages\torch_ops.py", line 643, in load_library
  ctypes.CDLL(path)
  File "C:\Users\ushow_AI.conda\envs\lyraChatGLM\lib\ctypes_init_.py", line 374, in init
  self._handle = _dlopen(self._name, mode)
  OSError: [WinError 193] %1 不是有效的 Win32 应用程序。