mlabonne/NeuralBeagle14-7B
mlabonne/NeuralBeagle14-7B is a 7 billion parameter DPO fine-tuned language model based on a merge of fblgit/UNA-TheBeagle-7b-v1 and argilla/distilabeled-Marcoro14-7B-slerp. It utilizes a 4096-token context window and excels in instruction following and reasoning tasks. This model is optimized for general-purpose conversational AI, including roleplay and storytelling, and ranks highly among 7B models on public leaderboards.
Loading preview...
NeuralBeagle14-7B: A Top-Performing 7B DPO Model
NeuralBeagle14-7B, developed by mlabonne, is a 7 billion parameter model fine-tuned using Direct Preference Optimization (DPO). It is built upon a strategic merge of existing high-quality models, specifically fblgit/UNA-TheBeagle-7b-v1 and argilla/distilabeled-Marcoro14-7B-slerp, leveraging the argilla/distilabel-intel-orca-dpo-pairs dataset for its DPO training.
Key Capabilities & Performance
- Instruction Following & Reasoning: The model demonstrates strong performance in instruction following and complex reasoning tasks, making it suitable for a wide range of applications.
- Context Window: It supports a context window of 4096 tokens, allowing for more extensive conversations and document processing.
- Evaluation Leaderboard: NeuralBeagle14-7B consistently ranks as a top 7B model on the Open LLM Leaderboard and the Nous suite evaluations, often outperforming its base models and other popular alternatives like
openchat/openchat-3.5-0106andteknium/OpenHermes-2.5-Mistral-7B.
Use Cases
- General-Purpose Chatbots: Its strong instruction following makes it ideal for conversational agents.
- Roleplay and Storytelling: The model's capabilities extend to creative text generation, including role-playing scenarios and narrative creation.
- Research and Development: Given its high benchmark scores, it serves as an excellent base for further experimentation and fine-tuning in the 7B parameter class.