feat(chatglm_int8_onnx):纯CPU推理，最多仅需8GB内存，推理速度未测评，token数有限，暂时还不能流式输出 #1008

2023-08-01 00:48:57 +08:00
parent 27f65c251a
commit c0c337988f
4 changed files with 376 additions and 2 deletions
--- a/request_llm/requirements_chatglm_onnx.txt
+++ b/request_llm/requirements_chatglm_onnx.txt
@@ -0,0 +1,11 @@
+protobuf
+transformers==4.27.1
+cpm_kernels
+torch>=1.10
+mdtex2html
+sentencepiece
+numpy
+onnxruntime
+sentencepiece
+streamlit
+streamlit-chat