localized-ft/Llama-3.1-8B-target-only-no-hallucination-last-third-sft-seed3-epoch3

TEXT GENERATIONPricing:Input $0.37 / Cached $0.074 / Output $0.38Concurrent Unit Cost:1Model Size:8BQuant:FP8Context Size:8kTool Calling:SupportedPublished:Aug 23, 2026License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold

The localized-ft/Llama-3.1-8B-target-only-no-hallucination-last-third-sft-seed3-epoch3 is an 8 billion parameter Llama-3.1-Instruct model developed by localized-ft. It was finetuned using Unsloth and Huggingface's TRL library, resulting in a 2x faster training process. This model is optimized for specific target applications, focusing on reducing hallucinations.

Loading preview...

Overview

This model, developed by localized-ft, is an 8 billion parameter Llama-3.1-Instruct variant. It was finetuned from the unsloth/Meta-Llama-3.1-8B-Instruct base model using the Unsloth framework and Huggingface's TRL library, which enabled a 2x faster training process.

Key Characteristics

  • Base Model: Meta-Llama-3.1-8B-Instruct
  • Parameter Count: 8 billion
  • Training Efficiency: Achieved 2x faster training using Unsloth and TRL.
  • Focus: Specifically targeted for applications requiring reduced hallucination.

Good For

  • Use cases where minimizing model hallucination is critical.
  • Applications benefiting from a Llama-3.1-8B-Instruct model with optimized training.
  • Developers looking for a finetuned model with a focus on specific output reliability.