DAMO-NLP-SG
/

VideoRefer-7B-stage2.5

Visual Question Answering

videorefer_qwen2

text-generation

multimodal large language model

large video-language model

Inference Endpoints

Model card Files Files and versions Community

CircleRadon commited on 10 days ago

Commit

b5aa872

·

verified ·

1 Parent(s): 1601549

Update config.json

Files changed (1) hide show

config.json +2 -2

config.json CHANGED Viewed

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "/mnt/workspace/workgroup/yuanyq/code/videollama2.1/VideoLLaMA2/work_dirs/1025_stage2.5_square_multiframe",
   "architectures": [
     "VideoReferQwen2ForCausalLM"
   ],
@@ -20,7 +20,7 @@
   "mm_region_encoder_type": "pooling",
   "mm_vision_select_feature": "patch",
   "mm_vision_select_layer": -2,
-  "mm_vision_tower": "/mnt/workspace/workgroup/yuanyq/checkpoints/siglip-so400m-patch14-384",
   "model_type": "videorefer_qwen2",
   "num_attention_heads": 28,
   "num_frames": 16,

 {
+  "_name_or_path": "Qwen/Qwen2-7B-Instruct",
   "architectures": [
     "VideoReferQwen2ForCausalLM"
   ],
   "mm_region_encoder_type": "pooling",
   "mm_vision_select_feature": "patch",
   "mm_vision_select_layer": -2,
+  "mm_vision_tower": "google/siglip-so400m-patch14-384",
   "model_type": "videorefer_qwen2",
   "num_attention_heads": 28,
   "num_frames": 16,