Spaces:
Running
on
Zero
Running
on
Zero
Update app.py
Browse files
app.py
CHANGED
@@ -184,7 +184,8 @@ def model_inference(input_dict, history):
|
|
184 |
).to("cuda")
|
185 |
|
186 |
streamer = TextIteratorStreamer(processor, skip_prompt=True, skip_special_tokens=False)
|
187 |
-
generation_kwargs = dict(inputs, streamer=streamer, max_new_tokens=1024, temperature=0.1, top_p=0.95, top_k=50)
|
|
|
188 |
thread = Thread(target=model.generate, kwargs=generation_kwargs)
|
189 |
thread.start()
|
190 |
|
@@ -242,7 +243,6 @@ def model_inference(input_dict, history):
|
|
242 |
# complete_assistant_response_for_gradio += f"\n<b>Analyzing Operation Result ...</b> @region(size={proc_img.size[0]}x{proc_img.size[1]})\n\n"
|
243 |
complete_assistant_response_for_gradio += [f"\n<b>Analyzing Operation Result ...</b> @region(size={proc_img.size[0]}x{proc_img.size[1]})\n\n"]
|
244 |
yield complete_assistant_response_for_gradio # Update Gradio display
|
245 |
-
# all_images.append(proc_img)
|
246 |
|
247 |
|
248 |
else:
|
|
|
184 |
).to("cuda")
|
185 |
|
186 |
streamer = TextIteratorStreamer(processor, skip_prompt=True, skip_special_tokens=False)
|
187 |
+
# generation_kwargs = dict(inputs, streamer=streamer, max_new_tokens=1024, temperature=0.1, top_p=0.95, top_k=50)
|
188 |
+
generation_kwargs = dict(inputs, streamer=streamer, max_new_tokens=1024, do_sample=False, num_beams=1)
|
189 |
thread = Thread(target=model.generate, kwargs=generation_kwargs)
|
190 |
thread.start()
|
191 |
|
|
|
243 |
# complete_assistant_response_for_gradio += f"\n<b>Analyzing Operation Result ...</b> @region(size={proc_img.size[0]}x{proc_img.size[1]})\n\n"
|
244 |
complete_assistant_response_for_gradio += [f"\n<b>Analyzing Operation Result ...</b> @region(size={proc_img.size[0]}x{proc_img.size[1]})\n\n"]
|
245 |
yield complete_assistant_response_for_gradio # Update Gradio display
|
|
|
246 |
|
247 |
|
248 |
else:
|