Smilesjs/chemsmart-qwen2.5-coder-3b-instruct-v14

TEXT GENERATIONConcurrent Unit Cost:1Model Size:3.1BQuant:BF16Context Size:32kTool Calling:SupportedPublished:Jul 1, 2026License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold

Smilesjs/chemsmart-qwen2.5-coder-3b-instruct-v14 is a 3.1 billion parameter instruction-tuned Qwen2.5-Coder model developed by Smilesjs, fine-tuned from unsloth/Qwen2.5-Coder-3B-Instruct-bnb-4bit. This model was trained using Unsloth and Huggingface's TRL library, enabling 2x faster training. With a 32768 token context length, it is optimized for coding tasks and instruction following.

Loading preview...

Smilesjs/chemsmart-qwen2.5-coder-3b-instruct-v14 Overview

This model is a 3.1 billion parameter instruction-tuned variant of the Qwen2.5-Coder architecture, developed by Smilesjs. It was fine-tuned from the unsloth/Qwen2.5-Coder-3B-Instruct-bnb-4bit base model, leveraging the Unsloth library for accelerated training, achieving a 2x speed improvement, alongside Huggingface's TRL library.

Key Capabilities

  • Efficient Training: Utilizes Unsloth for significantly faster fine-tuning.
  • Instruction Following: Designed to respond effectively to instructions, building upon its base as an instruct model.
  • Coding Focus: Inherits the coding capabilities from its Qwen2.5-Coder lineage.
  • Extended Context: Supports a context length of 32768 tokens, beneficial for complex coding problems or multi-turn conversations.

Good For

  • Applications requiring a compact yet capable instruction-following model.
  • Code generation and understanding tasks where the Qwen2.5-Coder architecture is suitable.
  • Scenarios benefiting from faster fine-tuning processes for custom datasets.