Update app.py
Browse files
app.py
CHANGED
@@ -37,7 +37,7 @@ def respond(
|
|
37 |
chat.append({"role": "assistant", "content": assistant})
|
38 |
chat.append({"role": "user", "content": message})
|
39 |
|
40 |
-
prompt = tokenizer.apply_chat_template(
|
41 |
#inputs = tokenizer.encode(prompt, add_special_tokens=False, return_tensors="pt").to(model.device)
|
42 |
|
43 |
streamer = TextIteratorStreamer(tokenizer, skip_prompt=True, skip_special_tokens=True)
|
@@ -67,6 +67,7 @@ def respond(
|
|
67 |
)
|
68 |
|
69 |
response = outputs[0]["generated_text"]
|
|
|
70 |
|
71 |
demo = gr.ChatInterface(
|
72 |
respond,
|
|
|
37 |
chat.append({"role": "assistant", "content": assistant})
|
38 |
chat.append({"role": "user", "content": message})
|
39 |
|
40 |
+
prompt = tokenizer.apply_chat_template(chat, tokenize=False, add_generation_prompt=True)
|
41 |
#inputs = tokenizer.encode(prompt, add_special_tokens=False, return_tensors="pt").to(model.device)
|
42 |
|
43 |
streamer = TextIteratorStreamer(tokenizer, skip_prompt=True, skip_special_tokens=True)
|
|
|
67 |
)
|
68 |
|
69 |
response = outputs[0]["generated_text"]
|
70 |
+
return response
|
71 |
|
72 |
demo = gr.ChatInterface(
|
73 |
respond,
|