初始化项目,由ModelHub XC社区提供模型

Model: Joker-sxj/Qwen2.5-3B-instruct-medical-finetuned
Source: Original Platform
This commit is contained in:
ModelHub XC
2026-09-07 12:31:16 +08:00
commit 96cea1e4d7
14 changed files with 152225 additions and 0 deletions

27
README.md Normal file
View File

@@ -0,0 +1,27 @@
---
license: apache-2.0
datasets:
- FreedomIntelligence/medical-o1-reasoning-SFT
metrics:
- bleu
base_model:
- Qwen/Qwen2.5-3B-Instruct
pipeline_tag: question-answering
tags:
- medical
language:
- zho
- eng
- fra
- spa
- por
- deu
- ita
- rus
- jpn
- kor
- vie
- tha
- ara
---
模型经过医疗数据集微调后,已初步具备推理能力,可以进行基础的问诊,且文本的质量BLEU比原模型更优。