Spaces:
Running
on
Zero
Running
on
Zero
Update chatbot.py
Browse files- chatbot.py +10 -10
chatbot.py
CHANGED
@@ -206,7 +206,13 @@ def qwen_inference(user_prompt, chat_history):
|
|
206 |
)
|
207 |
generation_kwargs = dict(inputs, streamer=streamer, max_new_tokens=2048)
|
208 |
|
209 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
210 |
|
211 |
image_extensions = Image.registered_extensions()
|
212 |
video_extensions = ("avi", "mp4", "mov", "mkv", "flv", "wmv", "mjpeg", "wav", "gif", "webm", "m4v", "3gp")
|
@@ -219,15 +225,9 @@ client_mistral_nemo = InferenceClient("mistralai/Mistral-Nemo-Instruct-2407")
|
|
219 |
|
220 |
def model_inference(user_prompt, chat_history):
|
221 |
if user_prompt["files"]:
|
222 |
-
|
223 |
-
|
224 |
-
|
225 |
-
thread.start()
|
226 |
-
|
227 |
-
buffer = ""
|
228 |
-
for new_text in streamer:
|
229 |
-
buffer += new_text
|
230 |
-
yield buffer
|
231 |
|
232 |
else:
|
233 |
func_caller = []
|
|
|
206 |
)
|
207 |
generation_kwargs = dict(inputs, streamer=streamer, max_new_tokens=2048)
|
208 |
|
209 |
+
thread = Thread(target=model.generate, kwargs=generation_kwargs)
|
210 |
+
thread.start()
|
211 |
+
|
212 |
+
buffer = ""
|
213 |
+
for new_text in streamer:
|
214 |
+
buffer += new_text
|
215 |
+
yield buffer
|
216 |
|
217 |
image_extensions = Image.registered_extensions()
|
218 |
video_extensions = ("avi", "mp4", "mov", "mkv", "flv", "wmv", "mjpeg", "wav", "gif", "webm", "m4v", "3gp")
|
|
|
225 |
|
226 |
def model_inference(user_prompt, chat_history):
|
227 |
if user_prompt["files"]:
|
228 |
+
|
229 |
+
for chunk in qwen_inference(user_prompt, chat_history):
|
230 |
+
yield chunk
|
|
|
|
|
|
|
|
|
|
|
|
|
231 |
|
232 |
else:
|
233 |
func_caller = []
|