Downloads · 30 days
22
33% of all-time downloads
asadfgglie/banban-vision-8b-v1.2.1-zh-DPO
banban-vision-8b-v1.2.1-zh-DPO is a text generation model from asadfgglie. Use it when you need the model to write or continue text. It is set up for transformers.
將asadfgglie/banban-vision-8b-v1.2-zh對asadfgglie/BanBan2024-10-17-DPO再次DPO訓練後的版本
Downloads · 30 days
22
33% of all-time downloads
All-time downloads
67
Public
Parameters
8.4B
16.7 GB on disk
Likes
0
Public
Click a slice to open those files.
.safetensors16.7 GB · 100%
From the Hugging Face model README
將asadfgglie/banban-vision-8b-v1.2-zh對asadfgglie/BanBan_2024-10-17-DPO再次DPO訓練後的版本
目前已知問題是中文能力會直接消失
英文能力反而有靠向這個資料集想要的偏好方向
推測是epoch=4導致的訓練過擬合
我原本想說eval loss與train loss都穩步下降的說
結果我等了10小時卻給我這個
哭阿