heegyu/WizardVicuna2-13b-hf
heegyu/WizardVicuna2-13b-hf is a 13 billion parameter language model fine-tuned from the Llama-2-13b-hf architecture. Developed by heegyu, this model was fine-tuned using the ehartford/wizard_vicuna_70k_unfiltered dataset over three epochs. It is optimized for dialogue use cases, building upon the Llama 2 foundation which excels in various natural language generation tasks with a 4096 token context length.
Loading preview...
Model Overview
heegyu/WizardVicuna2-13b-hf is a 13 billion parameter language model, fine-tuned from the Meta-developed Llama-2-13b-hf base model. This specific iteration was trained by heegyu using the ehartford/wizard_vicuna_70k_unfiltered dataset over three epochs, aiming to enhance its conversational capabilities.
Key Capabilities
- Dialogue Optimization: Fine-tuned for assistant-like chat and dialogue use cases, building on Llama 2's strengths.
- Llama 2 Foundation: Inherits the robust auto-regressive transformer architecture of Llama 2, which was pretrained on 2 trillion tokens of publicly available data.
- Context Length: Supports a context length of 4096 tokens, suitable for handling moderately long conversations or text sequences.
Intended Use
This model is primarily intended for commercial and research use in English, particularly for applications requiring conversational AI. Developers should adhere to the specific prompt formatting (e.g., ### Human:, ### Assistant:) for optimal performance, as indicated in the original fine-tuning process.