.

README.md

@@ -82,9 +82,10 @@ We support open-source model_id:
 "THUDM/cogvlm2-llama3-chat-19B",
 "echo840/Monkey-Chat",]
 ```
-For the models not on list, they are not intergated with huggingface, please refer to their github repo to create the evaluation pipeline.
+For the models not on list, they are not intergated with huggingface, please refer to their github repo to create the evaluation pipeline. Examples of the inference logic are in `src/evaluation/inference.py`
 
 ```bash
+pip install -r requirements.txt
 # We use HuggingFaceM4/idefics2-8b and vcr_wiki_en_easy as an example
 # Inference from the VLMs and save the results to {model_id}_{difficulty}_{language}.json
 cd src/evaluation
@@ -98,19 +99,23 @@ python3 gather_results.py --jsons_path .
 ```
 
 ### Close-source evaluation
-We provide the evaluation script for the close-source model: `GPT-4o`, `GPT-4-Turbo`, `Claude-3-Opus` in the `evaluation` folder.
+We provide the evaluation script for the close-source models in `src/evaluation/closed_source_eval.py`.
 
 You need an API Key, a pre-saved testing dataset and specify the path of the data saving the paper
 ```bash
+pip install -r requirements.txt
 cd src/evaluation
-# save the testing dataset to the path
+# [download images to inference locally option 1] save the testing dataset to the path using script from huggingface
 python3 save_image_from_dataset.py --output_path .
+# [download images to inference locally option 2] save the testing dataset to the path using github repo
+# use en-easy-test-500 as an example
+git clone https://github.com/tianyu-z/VCR-wiki-en-easy-test-500.git
 
-# Inference Put your API key and Image Path in the evaluation script (e.g. gpt-4o.py)
-python3 gpt-4o.py
+# specify your image path if you would like to inference using the image stored locally by --image_path "path_to_image", otherwise, the script will streaming the images from github repo
+python3 closed_source_eval.py --model_id gpt4o --dataset_handler "VCR-wiki-en-easy-test-500" --api_key "Your_API_Key"
 
 # Evaluate the results and save the evaluation metrics to {model_id}_{difficulty}_{language}_evaluation_result.json
-python3 evaluation_metrics.py --model_id gpt4o --output_path . --json_filename "gpt4o_en_easy.json" --dataset_handler "vcr-org/VCR-wiki-en-easy-test"
+python3 evaluation_metrics.py --model_id gpt4o --output_path . --json_filename "gpt4o_en_easy.json" --dataset_handler "vcr-org/VCR-wiki-en-easy-test-500"
 
 # To get the mean score of all the `{model_id}_{difficulty}_{language}_evaluation_result.json` in `jsons_path` (and the std, confidence interval if `--bootstrap`) of the evaluation metrics
 python3 gather_results.py --jsons_path .
@@ -123,7 +128,6 @@ pip install git+https://github.com/EvolvingLMMs-Lab/lmms-eval.git
 # We use HuggingFaceM4/idefics2-8b and vcr_wiki_en_easy as an example
 python3 -m accelerate.commands.launch --num_processes=8 -m lmms_eval --model idefics2 --model_args pretrained="HuggingFaceM4/idefics2-8b" --tasks vcr_wiki_en_easy --batch_size 1 --log_samples --log_samples_suffix HuggingFaceM4_idefics2-8b_vcr_wiki_en_easy --output_path ./logs/
 ```
-
 `lmms-eval` supports the following VCR `--tasks` settings:
 
 * English