vikingL08/Affine-5hdm4dumpm-r861

TEXT GENERATIONConcurrent Unit Cost:3Model Size:35.1BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Aug 19, 2026License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold

vikingL08/Affine-5hdm4dumpm-r861 is a 35.1 billion parameter Affine model, fine-tuned using offline DPO on Reason-ranked duel pairs. This model is specifically optimized for the Affine SN120 miner submission and evaluation server's Reason v4 duel. It excels at generating responses that raise teacher-side Reason scores, making it highly specialized for competitive AI evaluation rather than general chat applications.

Loading preview...

Overview

vikingL08/Affine-5hdm4dumpm-r861 is a 35.1 billion parameter model derived from the vera6/affine-5g4yy75zuz-t6 base. It was developed as an Affine SN120 challenger for Reason v4 evaluations, utilizing a tempered multi-sample log-mean-exp over three teacher references.

Key Training Details

  • Methodology: Trained using offline DPO (Direct Preference Optimization) on Reason-ranked duel pairs, specifically targeting preferences for thoughts that enhance teacher-side Reason scores.
  • Data: Utilized SoftCtx filtered duel preference pairs from dpo_duel_reason.jsonl.
  • Hyperparameters: Notable settings include LoRA r=32 (MidRank), α=128 (HiAlpha), β=0.1 (MidBeta), and a low learning rate (lr=5e-7).
  • Context Length: Supports a maximum context length of 12288 tokens (SoftCtx).
  • Performance: Achieved a positive margin of +0.003665 with a z-score of 2.177 against the live king reign36 under wvk=7, leading to its Stage-5 licensing.

Intended Use

  • This model is explicitly designed for Affine SN120 miner submissions and evalsrv Reason v4 duels.
  • It is not intended as a general-purpose chat model, but rather a specialized tool for competitive AI evaluation.