fix inference_test (NVlabs#108)

gheinrich · Jun 26, 2024 · 2dd6199 · 2dd6199
1 parent e723424
commit 2dd6199
Show file tree

Hide file tree

Showing 2 changed files with 3 additions and 3 deletions.
diff --git a/llava/eval/run_vila.py b/llava/eval/run_vila.py
@@ -56,7 +56,7 @@ def eval_model(args):
             assert osp.exists(args.video_file), "video file not found"
             video_file = args.video_file
         from llava.mm_utils import opencv_extract_frames
-        images = opencv_extract_frames(video_file, args.num_video_frames)
+        images, num_frames = opencv_extract_frames(video_file, args.num_video_frames)
 
     model_name = get_model_name_from_path(args.model_path)
     tokenizer, model, image_processor, context_len = load_pretrained_model(args.model_path, model_name, args.model_base)

diff --git a/tests/bash_tests/inference_tests.sh b/tests/bash_tests/inference_tests.sh
@@ -1,10 +1,10 @@
 # TODO(ligeng): replace with the model with finalized 7B model.
 python llava/eval/run_vila.py \
-    --model-path Efficient-Large-Model/CI-new-format-llama7b-siglip \
+    --model-path Efficient-Large-Model/VILA1.5-3b \
     --query "Please describe the image" \
     --image-file inference_test/test_data/caption_meat.jpeg
 
 python llava/eval/run_vila.py \
-    --model-path Efficient-Large-Model/CI-new-format-llama7b-siglip \
+    --model-path Efficient-Large-Model/VILA1.5-3b \
     --query "Please describe the video" \
     --video-file https://huggingface.co/datasets/Efficient-Large-Model/VILA-inference-demos/resolve/main/OAI-sora-tokyo-walk.mp4