unconst/Affine-5czsc2fc98-r220-merged

TEXT GENERATIONPricing:Input $0.4 / Cached $0.07 / Output $4Concurrent Unit Cost:3Model Size:35.1BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Aug 14, 2026Architecture:Transformer Featherless Exclusive Cold

The unconst/Affine-5czsc2fc98-r220-merged model is a 35.1 billion parameter language model, created by unconst, that has been LoRA-merged from kevin954/Affine-5dfqbbh8ev-sft. With a context length of 32768 tokens, this model is a salvaged checkpoint intended for private TTL insurance and is not yet a formal submission. Its primary differentiation lies in its status as a merged checkpoint, indicating potential for specialized applications or further development.

Loading preview...

Overview

The unconst/Affine-5czsc2fc98-r220-merged is a 35.1 billion parameter language model with a substantial context length of 32768 tokens. This model represents a LoRA-merged checkpoint, originating from kevin954/Affine-5dfqbbh8ev-sft.

Key Characteristics

  • Merged Checkpoint: This model is a result of a LoRA merge operation, suggesting it combines characteristics or optimizations from its base model.
  • Salvaged Status: It is described as a "salvaged checkpoint," indicating it might be a recovery or a specific version preserved for internal purposes.
  • Private Use: The model is designated for "Private TTL insurance" and is not intended as a public submission until it clears a "Stage-5 gate," implying it's in an developmental or evaluation phase.

Potential Use Cases

Given its status as a salvaged and privately intended checkpoint, this model is likely suitable for:

  • Internal Research and Development: Exploring the effects of LoRA merging or specific architectural changes.
  • Model Iteration and Testing: Evaluating performance or stability before a public release.
  • Specialized Applications: Potentially fine-tuned for a niche task within a controlled environment, leveraging its 35.1B parameters and large context window.

Finetunes
1 models