Martinbvt/qwen25-0.5b-ifeval-fr

TEXT GENERATIONPricing:Input $0.04 / Cached $0.008 / Output $0.08Concurrent Unit Cost:1Model Size:0.5BQuant:BF16Context Size:32kTool Calling:SupportedPublished:Aug 16, 2026License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold

The Martinbvt/qwen25-0.5b-ifeval-fr is a 0.5 billion parameter Qwen2.5-Instruct model, fine-tuned by Martinbvt using LoRA on a Mac M1. This model is specifically optimized for following verifiable instructions in French, leveraging its Qwen2ForCausalLM architecture. It is designed for tasks requiring instruction adherence and factual verification within the French language, offering a compact solution with a 32768 token context length.

Loading preview...

Model Overview

This model, Martintbvt/qwen25-0.5b-ifeval-fr, is a specialized variant of the Qwen2.5-Instruct architecture, featuring 0.5 billion parameters. It has been fine-tuned by Martinbvt using the LoRA method on a Mac M1, focusing on enhancing its ability to follow verifiable instructions in French.

Key Capabilities

  • Instruction Following: Optimized for accurately interpreting and executing instructions.
  • French Language Proficiency: Specifically tailored for tasks and interactions in French.
  • Verifiable Output: Designed to produce responses that can be fact-checked or verified.
  • Compact Size: With 0.5 billion parameters, it offers a lightweight solution for specific French NLP tasks.
  • Extended Context: Supports a context length of 32768 tokens, allowing for processing longer inputs.

Good For

  • Applications requiring precise instruction adherence in French.
  • Developing chatbots or agents that need to provide verifiable information in French.
  • Resource-constrained environments where a smaller, specialized model is beneficial.
  • Prototyping and development on Apple Silicon (Mac M1) due to its fine-tuning environment.

Finetunes
1 models