OLMo3-190M-zh-full (checkpoint-18500)
零基础 AI 大模型研发训练营（llm001）L04 Full 模型。
训练状态
- step: 18500 / 65200 (28.4%)
- epoch: 1.42 / 5.0
- 本次为 5-epoch 训练中途 checkpoint，训练仍在继续
- hidden_size: 768, num_layers: 12, num_heads: 12, intermediate_size: 3072
- vocab_size: 48000
用法
from transformers import AutoModelForCausalLM, AutoTokenizer
model = AutoModelForCausalLM.from_pretrained("lilingsunny/olmo3-190M-zh-full")
tok = AutoTokenizer.from_pretrained("lilingsunny/olmo3-190M-zh-full")
- Downloads last month
- 3
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support