elinas/chronos007-70b
The elinas/chronos007-70b is a 69 billion parameter experimental language model, merged from Chronos-70b-v2 and model 007 using the SLERP method. It features a 32768 token context length and is specifically designed to enhance logical and reasoning abilities while preserving Chronos' unique prose and general writing style. This model is optimized for tasks requiring improved reasoning alongside creative text generation.
Loading preview...
elinas/chronos007-70b: Enhanced Reasoning and Creative Writing
This model is an experimental 69 billion parameter merge of Chronos-70b-v2 and model 007, created by elinas. Utilizing a 0.3 ratio with Chronos as the parent model and the SLERP merging method, chronos007-70b aims to combine the strengths of both. It boasts a substantial 32768 token context length.
Key Capabilities & Differentiators
- Improved Logical and Reasoning Abilities: The primary goal of this merge was to enhance the base Chronos model's capacity for logical thought and problem-solving.
- Retained Unique Prose and General Writing: Despite the focus on reasoning, the model maintains the distinctive writing style and prose quality characteristic of the Chronos series.
- Experimental Nature: This model serves as a testbed for potential future Chronos iterations, exploring the balance between reasoning and creative output.
Usage and Licensing
- Alpaca Formatting: For optimal performance, users must employ Alpaca instruction formatting, enabling Alpaca instruction mode in frontends like SillyTavern.
- Non-Commercial Use: This model is strictly licensed under CC BY-NC 4.0, meaning it is for non-commercial use only. Commercial inquiries should be directed to Elinas.
Availability
Quantized versions (GGUF, GPTQ, AWQ) are available thanks to @TheBloke, making it accessible for various hardware setups.