unconst/Affine-5czsc2fc98-r447-online-dpo-merged

TEXT GENERATIONPricing:Input $0.4 / Cached $0.07 / Output $4Concurrent Unit Cost:3Model Size:35.1BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Aug 15, 2026Architecture:Transformer Featherless Exclusive Cold

unconst/Affine-5czsc2fc98-r447-online-dpo-merged is a 35.1 billion parameter language model developed by unconst. This model is a LoRA-merged checkpoint salvaged from kevin954/Affine-5dfqbbh8ev-sft, indicating a fine-tuned version of an existing base model. It is intended for private TTL insurance applications, suggesting specialization in domain-specific tasks within the insurance sector.

Loading preview...

Model Overview

unconst/Affine-5czsc2fc98-r447-online-dpo-merged is a 35.1 billion parameter language model. It represents a LoRA-merged checkpoint, salvaged from kevin954/Affine-5dfqbbh8ev-sft, indicating a process of fine-tuning or adaptation from a previous model iteration. The model has a context length of 32768 tokens, allowing it to process substantial amounts of information.

Key Characteristics

  • Parameter Count: 35.1 billion parameters, suggesting a large-scale model capable of complex language understanding and generation.
  • Context Length: Supports a 32768-token context window, enabling the processing of lengthy inputs and maintaining coherence over extended dialogues or documents.
  • Origin: Derived from a LoRA-merged checkpoint, implying efficient fine-tuning on a base model.

Intended Use Case

This model is specifically noted for "Private TTL insurance," indicating a highly specialized application within the insurance industry. It is not presented as a general-purpose language model but rather as a domain-specific tool. The mention of it not being a "submission until Stage-5 gate clears" suggests it is an internal or developmental version, potentially undergoing further validation or refinement for its specialized insurance tasks.

Finetunes
1 models