Spaces:
Runtime error
Runtime error
| title: Gemma 3n E4B Luau Generator | |
| emoji: 🎮 | |
| colorFrom: blue | |
| colorTo: purple | |
| sdk: gradio | |
| sdk_version: 4.44.0 | |
| app_file: app.py | |
| pinned: false | |
| license: apache-2.0 | |
| # 🎮 Gemma-3n-E4B Luau Code Generator | |
| This Space hosts the Gemma-3n-E4B model (8B parameters, 4B runtime efficiency) fine-tuned on the Roblox Luau corpus. | |
| ## Model Details | |
| - **Base Model**: google/gemma-3n-E4B | |
| - **Architecture**: Novel architecture with 8B parameters but 4B model runtime footprint | |
| - **Fine-tuned Dataset**: Roblox/luau_corpus | |
| - **Task**: Luau code generation for Roblox development | |
| - **Model Repository**: [Laserhun/gemma-3n-E4B-luau-finetuned](https://huggingface.co/Laserhun/gemma-3n-E4B-luau-finetuned) | |
| ## About Gemma-3n-E4B | |
| Gemma-3n-E4B uses an innovative architecture that provides 8 billion parameter model quality while maintaining the runtime efficiency of a 4 billion parameter model. | |
| ## Features | |
| - Advanced Luau code generation | |
| - Roblox-specific patterns and best practices | |
| - Efficient memory usage despite large parameter count | |
| - High-quality code output | |
| ## Training Details | |
| - Fine-tuned on Roblox Luau corpus | |
| - Uses LoRA for efficient adaptation | |
| - Optimized for code generation tasks | |