Downloads · 30 days
6
17% of all-time downloads
olmo3-190m-zh-sft
allansw/olmo3-190m-zh-sft
olmo3-190m-zh-sft is a machine learning model from allansw. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as apache-2.0.
SFT（有监督微调）版本：基于 allansw/olmo3-190m-zh-continue， 使用对话格式数据进行微调，学习指令遵循能力。
All-time downloads
35
Public
Parameters
187M
374 MB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors374 MB · 99%
At a glance
- License
- apache-2.0
- Model type
- olmo3
- Access
- Public
- Created
- Jun 10, 2026
- Updated
- Jun 10, 2026
- SHA
- 4bb1bad5
Base models
Model card
From the Hugging Face model README
allansw/olmo3-190m-zh-sft
SFT（有监督微调）版本：基于 allansw/olmo3-190m-zh-continue， 使用对话格式数据进行微调，学习指令遵循能力。
数据来源
- 训练数据：cmz1024/llm101-olmo3-zh-demo-data（来自 https://modelscope.cn/models/gongjy）
- 原始数据集：sft_t2t_mini.jsonl
训练配置
- LR：5e-05（低 LR 避免灾难性遗忘）
- Warmup：5.0%
- Max Seq Length：2048
用法
from transformers import AutoModelForCausalLM, AutoTokenizer
model = AutoModelForCausalLM.from_pretrained("allansw/olmo3-190m-zh-sft")
tok = AutoTokenizer.from_pretrained("allansw/olmo3-190m-zh-sft")
- Type
- olmo3
- License
- apache-2.0
- Languages
- zh
- Created
- Jun 10, 2026
- Updated
- Jun 10, 2026