Eric111/Yarn-Mistral-7b-128k-DPO
Eric111/Yarn-Mistral-7b-128k-DPO is a 7 billion parameter language model, fine-tuned using DPO (Direct Preference Optimization) from the NousResearch/Yarn-Mistral-7b-128k base model. It incorporates the Intel/orca_dpo_pairs dataset for its DPO fine-tuning process. This model is designed to leverage the Mistral architecture with an extended context length of 8192 tokens, making it suitable for tasks requiring longer conversational memory or document processing.
Loading preview...
Model Overview
Eric111/Yarn-Mistral-7b-128k-DPO is a 7 billion parameter language model that has undergone Direct Preference Optimization (DPO) fine-tuning. It is built upon the NousResearch/Yarn-Mistral-7b-128k base model, which itself is derived from the Mistral architecture. The DPO process utilized the Intel/orca_dpo_pairs dataset, aiming to align the model's outputs with human preferences.
Key Characteristics
- Base Model: NousResearch/Yarn-Mistral-7b-128k, a Mistral-based model.
- Parameter Count: 7 billion parameters.
- Fine-tuning Method: Direct Preference Optimization (DPO).
- Training Data: Fine-tuned with the Intel/orca_dpo_pairs dataset.
- Context Length: Features an extended context window of 8192 tokens, inherited from its base model.
Potential Use Cases
Given its DPO fine-tuning and extended context, this model is potentially well-suited for:
- Conversational AI: Maintaining longer dialogue histories and understanding complex multi-turn conversations.
- Text Summarization: Processing and summarizing lengthy documents or articles.
- Content Generation: Creating coherent and contextually relevant long-form text.
- Instruction Following: Executing complex instructions that require understanding of broader context, enhanced by DPO alignment.
Top 3 parameter combinations used by Featherless users for this model. Click a tab to see each config.