想要合并release里的文件

[JokerJostar]

求教

embedding.onnx
lm.onnx
block_[0-23].onnx
这三类文件分别是什么作用，如果要合并的话只用block吗还是所有的一起，我直接转换chatglm3-6b失败了

[wangzhaode]

具体的作用可以看下代码；如果像转换合并的模型就使用--export

[JokerJostar]

具体的作用可以看下代码；如果像转换合并的模型就使用--export

python3 llm_export.py --path ../chatglm3-6b --export --onnx_path ./chatglm2-6b-onnx
The device support i8sdot:0, support fp16:0, support i8mm: 0
Traceback (most recent call last):
File "/home/jostar/workspace/PycharmProjects/Rk/llm-export/llm_export.py", line 1081, in
llm_exporter = llm_modelsmodel_type
File "/home/jostar/workspace/PycharmProjects/Rk/llm-export/llm_export.py", line 562, in init
super().init(args)
File "/home/jostar/workspace/PycharmProjects/Rk/llm-export/llm_export.py", line 511, in init
super().init(args)
File "/home/jostar/workspace/PycharmProjects/Rk/llm-export/llm_export.py", line 90, in init
self.sp_model = spm.SentencePieceProcessor(tokenizer_model)
File "/home/jostar/.local/lib/python3.10/site-packages/sentencepiece/init.py", line 447, in Init
self.Load(model_file=model_file, model_proto=model_proto)
File "/home/jostar/.local/lib/python3.10/site-packages/sentencepiece/init.py", line 905, in Load
return self.LoadFromFile(model_file)
File "/home/jostar/.local/lib/python3.10/site-packages/sentencepiece/init.py", line 310, in LoadFromFile
return _sentencepiece.SentencePieceProcessor_LoadFromFile(self, arg)
RuntimeError: Internal: src/sentencepiece_processor.cc(1101) [model_proto->ParseFromArray(serialized.data(), serialized.size())]

这是sentencepiece的问题吗，我是用requirements.txt安装的

[Franklin-L]

具体的作用可以看下代码；如果像转换合并的模型就使用--export

python3 llm_export.py --path ../chatglm3-6b --export --onnx_path ./chatglm2-6b-onnx The device support i8sdot:0, support fp16:0, support i8mm: 0 Traceback (most recent call last): File "/home/jostar/workspace/PycharmProjects/Rk/llm-export/llm_export.py", line 1081, in llm_exporter = llm_modelsmodel_type File "/home/jostar/workspace/PycharmProjects/Rk/llm-export/llm_export.py", line 562, in init super().init(args) File "/home/jostar/workspace/PycharmProjects/Rk/llm-export/llm_export.py", line 511, in init super().init(args) File "/home/jostar/workspace/PycharmProjects/Rk/llm-export/llm_export.py", line 90, in init self.sp_model = spm.SentencePieceProcessor(tokenizer_model) File "/home/jostar/.local/lib/python3.10/site-packages/sentencepiece/init.py", line 447, in Init self.Load(model_file=model_file, model_proto=model_proto) File "/home/jostar/.local/lib/python3.10/site-packages/sentencepiece/init.py", line 905, in Load return self.LoadFromFile(model_file) File "/home/jostar/.local/lib/python3.10/site-packages/sentencepiece/init.py", line 310, in LoadFromFile return _sentencepiece.SentencePieceProcessor_LoadFromFile(self, arg) RuntimeError: Internal: src/sentencepiece_processor.cc(1101) [model_proto->ParseFromArray(serialized.data(), serialized.size())]

这是sentencepiece的问题吗，我是用requirements.txt安装的

请问大佬有没有针对微调后的chatglm3的onnx转换的经验呀