Spaces:
Running
on
Zero
Running
on
Zero
Update app.py
Browse files
app.py
CHANGED
@@ -54,7 +54,8 @@ def downsample_video(video_path):
|
|
54 |
vidcap.release()
|
55 |
return frames
|
56 |
|
57 |
-
MODEL_ID = "XiaomiMiMo/MiMo-VL-7B-RL"
|
|
|
58 |
processor = AutoProcessor.from_pretrained(MODEL_ID, trust_remote_code=True)
|
59 |
model = Qwen2_5_VLForConditionalGeneration.from_pretrained(
|
60 |
MODEL_ID,
|
@@ -156,7 +157,7 @@ def model_inference(input_dict, history):
|
|
156 |
|
157 |
demo = gr.ChatInterface(
|
158 |
fn=model_inference,
|
159 |
-
description="# **MiMo-VL-7B-RL `@video-infer for video understanding`**",
|
160 |
fill_height=True,
|
161 |
textbox=gr.MultimodalTextbox(label="Query Input", file_types=["image", "video"], file_count="multiple"),
|
162 |
stop_btn="Stop Generation",
|
|
|
54 |
vidcap.release()
|
55 |
return frames
|
56 |
|
57 |
+
# MODEL_ID = "XiaomiMiMo/MiMo-VL-7B-RL"
|
58 |
+
MODEL_ID = "XiaomiMiMo/MiMo-VL-7B-RL-2508"
|
59 |
processor = AutoProcessor.from_pretrained(MODEL_ID, trust_remote_code=True)
|
60 |
model = Qwen2_5_VLForConditionalGeneration.from_pretrained(
|
61 |
MODEL_ID,
|
|
|
157 |
|
158 |
demo = gr.ChatInterface(
|
159 |
fn=model_inference,
|
160 |
+
description="# **MiMo-VL-7B-RL (2508) `@video-infer for video understanding`**",
|
161 |
fill_height=True,
|
162 |
textbox=gr.MultimodalTextbox(label="Query Input", file_types=["image", "video"], file_count="multiple"),
|
163 |
stop_btn="Stop Generation",
|