Update app.py

app.py

@@ -1,8 +1,12 @@
-from transformers import AutoModel, AutoTokenizer
+#from transformers import AutoModel, AutoTokenizer
 import gradio as gr
 
-tokenizer = AutoTokenizer.from_pretrained("THUDM/chatglm-6b", trust_remote_code=True)
-model = AutoModel.from_pretrained("THUDM/chatglm-6b", trust_remote_code=True).half().cuda()
+from transformers import AutoModelForSeq2SeqLM
+
+model = AutoModelForSeq2SeqLM.from_pretrained("silver/chatglm-6b-int4-qe-slim")
+
+#tokenizer = AutoTokenizer.from_pretrained("THUDM/chatglm-6b", trust_remote_code=True)
+#model = AutoModel.from_pretrained("THUDM/chatglm-6b", trust_remote_code=True).half().cuda()
 #tokenizer = AutoTokenizer.from_pretrained("silver/chatglm-6b-slim", trust_remote_code=True)
 #model = AutoModel.from_pretrained("silver/chatglm-6b-slim", trust_remote_code=True).half().cuda()