{"id":31704092,"url":"https://github.com/fblgit/transviz","last_synced_at":"2025-10-08T22:49:54.197Z","repository":{"id":276463813,"uuid":"929362966","full_name":"fblgit/transviz","owner":"fblgit","description":"Debug and Visualise a Neural Deep Network internals suite","archived":false,"fork":false,"pushed_at":"2025-02-09T11:48:29.000Z","size":587,"stargazers_count":1,"open_issues_count":0,"forks_count":0,"subscribers_count":1,"default_branch":"master","last_synced_at":"2025-09-22T13:24:09.959Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":null,"language":"JavaScript","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/fblgit.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"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}},"created_at":"2025-02-08T11:24:16.000Z","updated_at":"2025-02-09T07:53:55.000Z","dependencies_parsed_at":"2025-02-08T12:37:37.524Z","dependency_job_id":null,"html_url":"https://github.com/fblgit/transviz","commit_stats":null,"previous_names":["fblgit/transviz"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/fblgit/transviz","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/fblgit%2Ftransviz","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/fblgit%2Ftransviz/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/fblgit%2Ftransviz/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/fblgit%2Ftransviz/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/fblgit","download_url":"https://codeload.github.com/fblgit/transviz/tar.gz/refs/heads/master","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/fblgit%2Ftransviz/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":279000770,"owners_count":26082851,"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","status":"online","status_checked_at":"2025-10-08T02:00:06.501Z","response_time":56,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"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":[],"created_at":"2025-10-08T22:49:49.189Z","updated_at":"2025-10-08T22:49:54.189Z","avatar_url":"https://github.com/fblgit.png","language":"JavaScript","funding_links":[],"categories":[],"sub_categories":[],"readme":"# TransViz - Neural Network Visualization Toolkit\n\n![TransViz Demo](docs/demo.gif) \u003c!-- Add actual demo media later --\u003e\n\n## STILLS UNDER EARLY PHASES OF DEVELOPMENT\nWork being done:\n- Packaging and dependencies\n- Tests \u0026 Examples of Usage\n- Issues being found at usage\n- Overall the python engine\n\nWork not yet done:\n- UI stills in just first draft\n- CI/CD, etc\n\nA real-time visualization and debugging toolkit for neural networks, featuring:\n\n- 🎮 Interactive tensor inspection\n- ⏯️ Training process time travel\n- 🔍 Breakpoint debugging system\n- 📊 Real-time metrics dashboard\n- 🔗 Distributed training support\n\n## Key Features\n\n- **Real-time Visualization**: Monitor activations/gradients during forward/backward passes\n- **Breakpoint System**: Set conditional breakpoints with tensor pattern matching\n- **Metrics Dashboard**: Track loss, accuracy, and custom metrics in real-time\n- **Differential Updates**: Efficient binary diffs for large tensor visualization\n- **Multi-Framework Support**: Designed for PyTorch with extensibility for other frameworks\n\n## Installation\n\n```\n# Clone repository\ngit clone https://github.com/fblgit/transviz.git\ncd transviz\n\n# Install dependencies\npip install -r requirements.txt\n```\n\n## Quick Start\n\n```\nfrom transviz import visualize, VizConfig\n\n# Instrument your model\n@visualize.trace(name=\"transformer_block\")\nclass TransformerBlock(nn.Module):\n    def forward(self, x):\n        x = self.attention(x)\n        visualize.breakpoint(\"post_attention\", x)\n        return self.ffn(x)\n\n# Initialize visualizer\nviz = visualize.ModelVisualizer(\n    VizConfig(\n        port=8080,\n        mode=\"hybrid\",\n        capture_gradients=True\n    )\n)\nviz.start()\n\n# Train your model normally\ntrain(model, dataloader)\n```\n\nAccess the visualization dashboard at `http://localhost:8080`\n\n## Configuration\n\n```\n@dataclass\nclass VizConfig:\n    port: int = 8080              # Web server port\n    mode: str = \"hybrid\"          # light|hybrid|debug\n    storage_path: str = \"./viz_data\"\n    tensor_sampling: float = 0.1  # Fraction of data to capture\n    capture_gradients: bool = False\n    breakpoint_timeout: int = 300  # seconds\n```\n\n## API Reference\n\n### Decorators\n```\n@visualize.trace(name=\"layer\", capture_shapes=True)\n@visualize.breakpoint(\"attention_weights\", condition=lambda t: t.max() \u003e 1.0)\n```\n\n### Core Methods\n```\nvisualize.log_metrics({\"loss\": 0.5})\nvisualize.set_breakpoint(\"gradient_check\", condition=has_nan)\nvisualize.save_session(\"training_run_1\")\n```\n\n### CLI Commands\n```\ntransviz serve --port 8080 --model ./checkpoint.pth\ntransviz export --session latest --format hdf5\n```\n\n## Contributing\n\n1. Fork the repository\n2. Create feature branch (`git checkout -b feature/amazing-feature`)\n3. Commit changes (`git commit -m 'Add amazing feature'`)\n4. Push to branch (`git push origin feature/amazing-feature`)\n5. Open Pull Request\n\n\n**Roadmap**  \n[ ] TensorFlow/Keras support  \n[ ] Jupyter notebook integration  \n[ ] Pre-trained model zoo integration  \n[ ] 3D tensor visualization  \n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Ffblgit%2Ftransviz","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Ffblgit%2Ftransviz","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Ffblgit%2Ftransviz/lists"}