ValiantLabs/Llama3.2-3B-Enigma
ValiantLabs/Llama3.2-3B-Enigma is a 3.2 billion parameter code-instruct model developed by Valiant Labs, built upon the Llama 3.2 architecture with a 32768 token context length. It is fine-tuned using synthetic code-instruct data from the Tachibana dataset and generalist synthetic data from Supernova. This model excels in high-quality code instruction and general chat performance, making it suitable for developers requiring a capable code assistant.
Loading preview...
Overview
ValiantLabs/Llama3.2-3B-Enigma is a 3.2 billion parameter model developed by Valiant Labs, specifically designed for code instruction and general chat. Built on the Llama 3.2 Instruct architecture, it leverages a 32768 token context window to handle complex prompts. The model's training incorporates high-quality synthetic code-instruct data from the sequelbox/Tachibana dataset, alongside generalist synthetic data from sequelbox/Supernova to enhance overall chat capabilities.
Key Capabilities
- High-Quality Code Instruction: Optimized for generating and understanding code-related instructions.
- Llama 3.2 Instruct Format: Utilizes the standard Llama 3.2 Instruct chat format for seamless integration.
- Enhanced General Chat: Supplements code capabilities with strong general conversational performance.
When to Use
This model is ideal for developers and applications requiring a capable and efficient code assistant. Its fine-tuning on specialized datasets makes it particularly effective for tasks involving code generation, explanation, and debugging, while also maintaining robust general chat abilities.