KnutJaegersberg/Deacon-34B
KnutJaegersberg/Deacon-34B is a 34 billion parameter language model, based on the Yi-34B architecture and 'llamafied' with a Llama tokenizer. It has been fine-tuned for five epochs on the EverythingLM dataset, offering a 32768 token context length. This model is optimized for general conversational tasks and instruction following, leveraging its extensive training for broad applicability.
Loading preview...
Overview
KnutJaegersberg/Deacon-34B is a 34 billion parameter large language model, derived from the Yi-34B architecture and adapted to use a Llama tokenizer. It was fine-tuned for five epochs on the comprehensive EverythingLM dataset, enhancing its general instruction-following capabilities. The model supports a substantial context length of 32768 tokens, allowing for processing and generating longer, more coherent texts.
Key Capabilities
- Instruction Following: Enhanced through fine-tuning on the EverythingLM dataset.
- Extended Context: Processes inputs up to 32768 tokens, suitable for complex queries and longer conversations.
- Llama Tokenizer: Utilizes a widely adopted tokenizer, potentially improving compatibility with existing Llama-based workflows.
Licensing
The underlying Yi series models are open for academic research and free commercial usage, subject to permission via application. All usage must comply with the Model License Agreement 2.0. Commercial licensing requires direct application to 01.ai.
When to Use
This model is suitable for developers and researchers looking for a robust 34B parameter model with strong instruction-following abilities and a large context window. Its fine-tuning on a broad dataset makes it versatile for various general-purpose AI applications.