rootti/unconst
The rootti/unconst model is a 35.1 billion parameter language model, derived from a LoRA-merged checkpoint. It is based on the kevin954/Affine-5dfqbbh8ev-sft architecture and features a context length of 32768 tokens. This model is currently in a private development stage, intended for internal use related to TTL insurance, and is not yet publicly released or validated for general applications. Its primary characteristic is its origin as a salvaged merged checkpoint, indicating a focus on specific, internal fine-tuning rather than broad, general-purpose capabilities.
Loading preview...
Model Overview
The rootti/unconst model is a 35.1 billion parameter language model with a substantial context length of 32768 tokens. It originates from a LoRA-merged checkpoint, specifically derived from the kevin954/Affine-5dfqbbh8ev-sft base architecture. This model is currently in a private, developmental phase, described as a "merged checkpoint salvage."
Key Characteristics
- Parameter Count: 35.1 billion parameters, indicating a large-scale model.
- Context Length: Supports a long context window of 32768 tokens, beneficial for processing extensive inputs.
- Origin: Developed from a LoRA-merged checkpoint, suggesting fine-tuning or adaptation from an existing model.
- Development Status: Explicitly noted as "Private TTL insurance; not a submission until Stage-5 gate clears," indicating it is an internal, unreleased version.
Intended Use
Given its private development status and description as "TTL insurance," this model is not intended for general public use or broad applications at this time. It appears to be a specialized, internally-focused iteration, likely undergoing further development and validation within a specific project context.