unconst/Affine-5czsc2fc98-r517-offline-dpo-hialpha-midrank-lobeta-longctx-ultraextrasteps-merged

TEXT GENERATIONPricing:Input $0.4 / Cached $0.07 / Output $4Concurrent Unit Cost:3Model Size:35.1BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Aug 16, 2026Architecture:Transformer Featherless Exclusive Cold

The unconst/Affine-5czsc2fc98-r517-offline-dpo-hialpha-midrank-lobeta-longctx-ultraextrasteps-merged model is a 35.1 billion parameter language model, merged from a LoRA checkpoint of kevin954/Affine-5dfqbbh8ev-sft. It features a substantial 32,768 token context length, indicating its capability for processing extensive inputs. This model is noted as a private TTL insurance project, suggesting a focus on specific, potentially internal, applications rather than general public use.

Loading preview...

Model Overview

The unconst/Affine-5czsc2fc98-r517-offline-dpo-hialpha-midrank-lobeta-longctx-ultraextrasteps-merged is a large language model with 35.1 billion parameters and an impressive 32,768 token context length. This model was created by unconst and is a merged checkpoint, specifically derived from a LoRA fine-tune of kevin954/Affine-5dfqbbh8ev-sft.

Key Characteristics

  • Parameter Count: 35.1 billion, placing it among larger language models capable of complex tasks.
  • Context Window: A significant 32,768 tokens, enabling it to handle very long documents, conversations, or code sequences.
  • Origin: It is a LoRA-merged checkpoint, indicating a fine-tuning process applied to a base model to adapt it for specific purposes.

Intended Use and Status

This model is described as a "private TTL insurance" project, suggesting it is not intended for general public release or broad commercial use at this stage. Its status as "not a submission until Stage-5 gate clears" further implies it is an internal or developmental version, likely undergoing rigorous testing and validation within a specific organizational context. Developers should note its private nature and specific project-oriented origin when considering its applicability.