unconst/Affine-5czsc2fc98-r452-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 16, 2026Architecture:Transformer Featherless Exclusive Cold
Affine-5czsc2fc98-r452-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. It is noted as being related to private TTL insurance, indicating a specialized domain focus. Its primary use case appears to be within specific, potentially proprietary, insurance-related applications.
Loading preview...
Model Overview
The unconst/Affine-5czsc2fc98-r452-online-dpo-merged is a 35.1 billion parameter language model. It represents a salvaged checkpoint that has undergone a LoRA merge process. The base model for this merge was kevin954/Affine-5dfqbbh8ev-sft.
Key Characteristics
- Parameter Count: 35.1 billion parameters.
- Context Length: Supports a context window of 32,768 tokens.
- Origin: This model is a result of merging a LoRA (Low-Rank Adaptation) checkpoint.
- Domain Specificity: The model's description explicitly links it to "Private TTL insurance," suggesting a specialization in insurance-related language understanding and generation. This indicates it may be optimized for terminology, regulations, or data specific to the insurance sector.
Potential Use Cases
Given its specialized domain mention, this model is likely intended for:
- Insurance Industry Applications: Processing, analyzing, or generating text related to private TTL (Term, Universal, Whole) insurance.
- Internal Business Operations: Supporting tasks within an insurance company that require understanding complex policy documents, customer inquiries, or regulatory texts.
- Specialized NLP Tasks: Any task requiring deep domain knowledge in the insurance sector, where general-purpose LLMs might lack specific expertise.