Spaces:
Running
on
Zero
Running
on
Zero
Update app.py
Browse files
app.py
CHANGED
@@ -119,7 +119,6 @@ def model_inference(input_dict, history, use_rolmocr=False):
|
|
119 |
model = rolmocr_model if use_rolmocr else qwen_model
|
120 |
model_name = "RolmOCR" if use_rolmocr else "Qwen2VL OCR"
|
121 |
|
122 |
-
# Prepare prompt and inputs
|
123 |
prompt_full = processor.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
|
124 |
all_images = [item["image"] for item in content if item["type"] == "image"]
|
125 |
inputs = processor(
|
@@ -129,7 +128,6 @@ def model_inference(input_dict, history, use_rolmocr=False):
|
|
129 |
padding=True,
|
130 |
).to("cuda")
|
131 |
|
132 |
-
# Set up streaming
|
133 |
streamer = TextIteratorStreamer(processor, skip_prompt=True, skip_special_tokens=True)
|
134 |
generation_kwargs = dict(inputs, streamer=streamer, max_new_tokens=1024)
|
135 |
thread = Thread(target=model.generate, kwargs=generation_kwargs)
|
@@ -138,23 +136,17 @@ def model_inference(input_dict, history, use_rolmocr=False):
|
|
138 |
buffer = ""
|
139 |
yield progress_bar_html(f"Processing with {model_name}")
|
140 |
|
141 |
-
# Stream
|
142 |
for new_text in streamer:
|
143 |
-
buffer += new_text
|
144 |
-
buffer = buffer.replace("<|im_end|>", "")
|
145 |
time.sleep(0.01)
|
146 |
yield buffer
|
147 |
|
148 |
-
# Once
|
149 |
-
|
150 |
-
|
151 |
-
with open("response.txt", "w", encoding="utf-8") as f:
|
152 |
-
f.write(results)
|
153 |
-
except Exception as e:
|
154 |
-
yield f"Error writing to response.txt: {e}"
|
155 |
-
return
|
156 |
|
157 |
-
yield
|
158 |
return
|
159 |
|
160 |
# Gradio Interface
|
@@ -180,4 +172,5 @@ demo = gr.ChatInterface(
|
|
180 |
additional_inputs=[gr.Checkbox(label="Use RolmOCR", value=False, info="Check to use RolmOCR, uncheck to use Qwen2VL OCR")],
|
181 |
)
|
182 |
|
183 |
-
|
|
|
|
119 |
model = rolmocr_model if use_rolmocr else qwen_model
|
120 |
model_name = "RolmOCR" if use_rolmocr else "Qwen2VL OCR"
|
121 |
|
|
|
122 |
prompt_full = processor.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
|
123 |
all_images = [item["image"] for item in content if item["type"] == "image"]
|
124 |
inputs = processor(
|
|
|
128 |
padding=True,
|
129 |
).to("cuda")
|
130 |
|
|
|
131 |
streamer = TextIteratorStreamer(processor, skip_prompt=True, skip_special_tokens=True)
|
132 |
generation_kwargs = dict(inputs, streamer=streamer, max_new_tokens=1024)
|
133 |
thread = Thread(target=model.generate, kwargs=generation_kwargs)
|
|
|
136 |
buffer = ""
|
137 |
yield progress_bar_html(f"Processing with {model_name}")
|
138 |
|
139 |
+
# Stream generation
|
140 |
for new_text in streamer:
|
141 |
+
buffer += new_text.replace("<|im_end|>", "")
|
|
|
142 |
time.sleep(0.01)
|
143 |
yield buffer
|
144 |
|
145 |
+
# Once complete, save to response.txt and yield final confirmation
|
146 |
+
with open("response.txt", "w", encoding="utf-8") as f:
|
147 |
+
f.write(buffer)
|
|
|
|
|
|
|
|
|
|
|
148 |
|
149 |
+
yield f"\n✅ Response saved to `response.txt`:\n\n{buffer}"
|
150 |
return
|
151 |
|
152 |
# Gradio Interface
|
|
|
172 |
additional_inputs=[gr.Checkbox(label="Use RolmOCR", value=False, info="Check to use RolmOCR, uncheck to use Qwen2VL OCR")],
|
173 |
)
|
174 |
|
175 |
+
if __name__ == "__main__":
|
176 |
+
demo.launch(debug=True)
|