Downloads · 30 days
17
15% of all-time downloads
Cascade0-159M-DPO-Instruct-V2-0925
ARMZyany/Cascade0-159M-DPO-Instruct-V2-0925
Cascade0-159M-DPO-Instruct-V2-0925 is a text generation model from ARMZyany. Use it when you need the model to write or continue text. The card lists the license as apache-2.0.
All-time downloads
111
Public
Parameters
159M
952 MB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.pt635 MB · 66%
At a glance
- Task
- Text Generation
- License
- apache-2.0
- Model type
- llama
- Access
- Public
- Created
- Sep 28, 2025
- Updated
- Sep 29, 2025
- SHA
- e3055e3a
Try a prompt
Model card
From the Hugging Face model README
Cascade0 DPO Instruct V2
(Trained only on 4.8B tokens)
Experimental DPO Pass to see the difference.
In some cases, during certain chats, the new DPO actually helps and makes the model feel more chat-y, sort of.
For eg, when asking 'How to make a salad', it responds:
Training info
Done in 2 weeks on a single RTX 4080, the raw pretraining only achieveing 40% of the max 12B tokens, due to potential overfill, reaching only 4.8B tokens.
Max context size is 1512.
SFT was done in two days, while DPO took only 1 day, on a RTX 3060M.
Other small models VS Cascade0-Series
All models are evaluated using LMEval Harness, on the same PC/Settings and GGUF with F16 Quant.
made with LMEval Harness
- Task
- Text Generation
- Type
- llama
- License
- apache-2.0
- Languages
- en
- Created
- Sep 28, 2025
- Updated
- Sep 29, 2025