promotion/Llama-3.1-8B-NBPO-UltraFeedback-eta0.05

TEXT GENERATIONPricing:Input $0.2 / Cached $0.028 / Output $0.32Concurrent Unit Cost:1Model Size:8BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Sep 6, 2026License:llama3.1Architecture:Transformer Featherless Exclusive Cold

promotion/Llama-3.1-8B-NBPO-UltraFeedback-eta0.05 is an 8 billion parameter language model based on the Llama-3.1-8B-Instruct backbone, fine-tuned using Nash Bargaining Preference Optimization (NBPO) on the UltraFeedback dataset. This model is specifically optimized for instruction following, truthfulness, honesty, and helpfulness, utilizing a unique eta_t value of 0.05. It is designed for applications requiring highly aligned and reliable conversational AI outputs.

Loading preview...

Model Overview

promotion/Llama-3.1-8B-NBPO-UltraFeedback-eta0.05 is an 8 billion parameter language model built upon the meta-llama/Llama-3.1-8B-Instruct backbone. It has been fine-tuned using the Nash Bargaining Preference Optimization (NBPO) method, specifically targeting the UltraFeedback dataset. This model distinguishes itself by employing a unique eta_t value of 0.05, which differs from the released eta_t = 1 for its panel.

Key Capabilities

  • Optimized Objectives: Focuses on improving instruction following, truthfulness, honesty, and helpfulness.
  • Training Methodology: Utilizes Nash Bargaining Preference Optimization (NBPO) with a training budget of 300 optimizer updates and a global batch size of 16.
  • Evaluation: Performance is assessed by independent objective-wise win rates against a common reference, judged by Llama-3.3-70B-Instruct on prompt-disjoint held-out prompts.

Good For

  • Applications requiring a highly aligned and reliable conversational agent.
  • Use cases where strong instruction following and truthful responses are critical.
  • Scenarios benefiting from a model fine-tuned with a specific eta_t value for enhanced preference optimization.

Finetunes
1 models