AIbase
EN
Home
QWEN 3B INSTRUC Medical COT SFT 2kstep 4kcol
Q

QWEN 3B INSTRUC Medical COT SFT 2kstep 4kcol

Developed by hailong18102002
A 3B parameter instruction fine-tuned model based on Qwen2.5 architecture, optimized for training speed using Unsloth and Huggingface TRL library
Large Language Model
Transformers
 EnglishOpen Source License:Apache-2.0#Efficient Fine-tuning#4-bit Quantization#Instruction Optimization
Downloads 30
Release Time : 5/8/2025

Model Overview

This is a 3B parameter language model fine-tuned for instructions, based on the Qwen2.5 architecture, with accelerated training process using Unsloth, suitable for text generation and reasoning tasks

Model Features

Efficient Training
Optimized with Unsloth tool, achieving 2x faster training speed
4-bit Quantization
Utilizes BNB 4-bit quantization technology to reduce memory usage
Instruction Fine-tuning
Supervised Fine-Tuning (SFT) applied to enhance instruction-following capability

Model Capabilities

Text Generation
Instruction Understanding
Reasoning Task Handling

Use Cases

Dialogue Systems
Intelligent Assistant
Building dialogue assistants capable of understanding complex instructions
Content Generation
Creative Writing
Generating creative text content such as stories and poems
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
English简体中文繁體中文にほんご
© 2025AIbase