{"id":27697098,"url":"https://github.com/cansik/vqa-service","last_synced_at":"2025-07-08T01:32:28.384Z","repository":{"id":287168027,"uuid":"951443223","full_name":"cansik/vqa-service","owner":"cansik","description":"VQA application that allows users to ask questions about images and receive answers.","archived":false,"fork":false,"pushed_at":"2025-05-16T09:45:05.000Z","size":43,"stargazers_count":1,"open_issues_count":0,"forks_count":0,"subscribers_count":1,"default_branch":"main","last_synced_at":"2025-06-01T15:27:48.972Z","etag":null,"topics":["gradio","python","service","vqa"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/cansik.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null}},"created_at":"2025-03-19T17:27:08.000Z","updated_at":"2025-05-16T09:45:09.000Z","dependencies_parsed_at":"2025-04-10T10:45:59.934Z","dependency_job_id":"f2760528-fba1-4774-aa2e-5d6f168c8b32","html_url":"https://github.com/cansik/vqa-service","commit_stats":null,"previous_names":["cansik/vqa-service"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/cansik/vqa-service","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/cansik%2Fvqa-service","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/cansik%2Fvqa-service/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/cansik%2Fvqa-service/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/cansik%2Fvqa-service/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/cansik","download_url":"https://codeload.github.com/cansik/vqa-service/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/cansik%2Fvqa-service/sbom","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":264175531,"owners_count":23568482,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["gradio","python","service","vqa"],"created_at":"2025-04-25T15:27:58.731Z","updated_at":"2025-07-08T01:32:28.359Z","avatar_url":"https://github.com/cansik.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Visual Question Answering (VQA) Service\n\n## Overview\n\nThe VQA Service is a machine learning based application that allows users to ask questions about images and receive answers.\nIt leverages state-of-the-art models to process images and generate accurate responses to user queries. The service is\ndesigned to be flexible, supporting multiple backend models for different use cases.\n\n## Installation\n\nTo install and set up the VQA Service, follow these steps:\n\n1. **Clone the repository:**\n\n   ```bash\n   git clone https://github.com/cansik/vqa-service.git\n   cd vqa-service\n   ```\n\n2. **Create a virtual environment:**\n\n   ```bash\n   python -m venv venv\n   source venv/bin/activate  # On Windows use `venv\\Scripts\\activate`\n   ```\n\n3. **Install the required dependencies:**\n\n   ```bash\n   pip install -r requirements.txt\n   ```\n\n## Usage\n\nTo start the VQA Service, use the following command:\n\n```bash\npython -m vqa\n```\n\n### Command-line Arguments\n\n- `--host`: Specify the service host (default: `127.0.0.1`).\n- `--port`: Specify the service port (default: `7840`).\n- `--backend`: Choose the VQA backend model (default: `blip`).\n\n### Example\n\nTo run the service on a specific host and port with a chosen backend, use:\n\n```bash\npython -m vqa --host 0.0.0.0 --port 8000 --backend blip2\n```\n\n### Accessing the Service\n\nOnce the service is running, you can access it through a web interface provided by Gradio. Open your web browser and\nnavigate to `http://\u003chost\u003e:\u003cport\u003e` to interact with the service.\n\n## Supported Models\n\nThe following VLM backends are supported:\n\n| Backend ID | Model | Description |\n|------------|-------|-------------|\n| `blip` | Salesforce/blip-vqa-base | BLIP base model for visual question answering |\n| `blip2` | Salesforce/blip2-opt-2.7b | BLIP2 with OPT 2.7B language model |\n| `blip2-flan` | Salesforce/blip2-flan-t5-xl | BLIP2 with Flan-T5-XL language model |\n| `vilt` | dandelin/vilt-b32-finetuned-vqa | ViLT model fine-tuned for VQA tasks |\n| `vlmmlx` | mlx-community/Qwen2-VL-2B-Instruct-4bit | Default MLX-based VLM for Apple Silicon |\n| `vlmmlx-phi35` | mlx-community/Phi-3.5-vision-instruct-4bit | Phi-3.5 Vision model optimized for MLX |\n| `vlmmlx-smolvlm2` | mlx-community/SmolVLM2-500M-Video-Instruct-mlx-8bit-skip-vision | SmolVLM2 optimized for MLX |\n| `namo` | - | Namo VLM model |\n| `moondream` | vikhyatk/moondream2 | Moondream2 model with GPU support |\n| `moondream-cpu` | vikhyatk/moondream2 | Moondream2 model optimized for CPU inference |\n| `smolvlm` | HuggingFaceTB/SmolVLM-256M-Instruct | Lightweight VLM model |\n| `smolvlm2` | HuggingFaceTB/SmolVLM2-256M-Video-Instruct | SmolVLM2 with video instruction capabilities |\n\n\nEach backend offers different capabilities and performance characteristics.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fcansik%2Fvqa-service","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fcansik%2Fvqa-service","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fcansik%2Fvqa-service/lists"}