Downloads · 30 days
2
3% of all-time downloads
in-character-rp-12b-v0.2-lora
npow/in-character-rp-12b-v0.2-lora
in-character-rp-12b-v0.2-lora is a machine learning model from npow. Use it for the machine learning task on the model card, and read the license before you ship it in a product. It is set up for peft. The card lists the license as apache-2.0.
LoRA adapter for Mistral-Nemo-Instruct-2407 that fixes RP positivity drift — villains stay villainous, tension doesn't dissolve into therapy. Load with PEFT, or mergeandunload into the base.
All-time downloads
64
Public
Repo size
929 MB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors912 MB · 98%
At a glance
- Library
- peft
- License
- apache-2.0
- Access
- Public
- Created
- Jun 26, 2026
- Updated
- Jun 26, 2026
- SHA
- 98f67db1
Base models
Model card
From the Hugging Face model README
In-Character RP 12B — v0.2 (LoRA adapter)
LoRA adapter for Mistral-Nemo-Instruct-2407 that fixes RP positivity drift — villains stay villainous, tension doesn't dissolve into therapy. Load with PEFT, or merge_and_unload into the base.
Method
DDPO (diversity-preserving DPO), LoRA r=64, β=0.1, 1 epoch, on ~1,000 same-context hard-negative pairs: chosen = the model's own in-character hold under emotional pressure, rejected = its own natural softening on the same prompt.
Results — internal anti-drift benchmark (192 scenarios, N2×M2, paired)
| softening / scn (↓) | stance-hold rate (↑) | |
|---|---|---|
| base Nemo-Instruct | 3.17 | 42.6% |
| v0.2 | 2.12 | 68.6% |
Paired: softening −1.05 (p<0.0001), hold +26 pts (p<0.0001). Repetition 0.07 (no degeneration).
Usage
Serve base + this adapter (e.g. vLLM --enable-lora), or merge. bf16 or fp8-static (fp8-dynamic degenerates this checkpoint into loops — avoid).
- Library
- peft
- License
- apache-2.0
- Created
- Jun 26, 2026
- Updated
- Jun 26, 2026