Skip to content
ModelVerifiedExternal
Download

Dpo Writingprompts Qwen3 32b As Nemotron Nano 30b A3b Seed1

LoRA adapter trained via Tinker as part of the dementor intervention-ladder fingerprint persistence study (AAAI 2026 conference).

Sourcehuggingface.co/dementor-research/dpo_writingprompts_qwen3-32b_as_nemotron-nano-30b-a3b_seed1/tree/0bf04a35faff98fbe4e2f9e83652cbe976da0117
KindLoRABase modelQwen/Qwen3-32BVersionadapter_modelPublisher@dementor-researchCgrade

Model source

Kind
LoRA
File size
1.02 GB
Base model
Qwen/Qwen3-32B
Version
vadapter_model
Updated
Oct 3, 2026
Publisher
@dementor-research
Downloads
4
Parameters
32B
Source
Hugging Face

Source description

dpo_writingprompts_qwen3-32b_as_nemotron-nano-30b-a3b_seed1

LoRA adapter trained via Tinker as part of the dementor intervention-ladder fingerprint persistence study (AAAI 2026 conference).

  • Base model: Qwen/Qwen3-32B
  • Training stage: DPO (LoRA rank 32, target_modules=all-linear)
  • Alias: dpo_writingprompts_qwen3-32b_as_nemotron-nano-30b-a3b_seed1

Usage

from peft import PeftModel
from transformers import AutoModelForCausalLM, AutoTokenizer

base = AutoModelForCausalLM.from_pretrained("Qwen/Qwen3-32B")
tok = AutoTokenizer.from_pretrained("Qwen/Qwen3-32B")
model = PeftModel.from_pretrained(base, "dementor-research/dpo_writingprompts_qwen3-32b_as_nemotron-nano-30b-a3b_seed1")

Part of the dementor matrix: 4 source models × 3 cross-targets × 3 train datasets × 3 seeds × 2 stages = 216 adapters.

Sources

1 source
Hugging FacePrimaryVersion 0bf04a35faff98fbe4e2f9e83652cbe976da0117
huggingface.co/dementor-research/dpo_writingprompts_qwen3-32b_as_nemotron-nano-30b-a3b_seed1/tree/0bf04a35faff98fbe4e2f9e83652cbe976da0117

Verified Oct 3

Model artifacts

1 artifact

adapter_model.safetensors

safetensors · 1.02 GB · SHA-256 77cc691b76c3…b38b · Hugging Face

Download

Source excerpts

3 excerpts
ReadmeJul 11huggingface.co/dementor-research/dpo_writingprompts_qwen3-32b_as_nemotron-nano-30b-a3b_seed1

--- base_model: Qwen/Qwen3-32B library_name: peft tags: - lora - dpo - dementor-research --- # dpo_writingprompts_qwen3-32b_as_nemotron-nano-30b-a3b_seed1 LoRA adapter trained via [Tinker](https://thinkingmachines.ai/tinker/) as part of the **dementor** intervention-ladder fingerprint persistence study (AAAI 2026 conference). - **Base model:** `Qwen/Qwen3-32B` - **Training stage:** DPO (LoRA rank 32, target_modules=all-linear) - **Alias:** `dpo_writingprompts_qwen3-32b_as_nemotron-nano-30b-a3b_seed1` ## Usage ```python from peft import PeftModel from transformers import AutoModelForCausalLM, AutoTokenizer base = AutoModelForCausalLM.from_pretrained("Qwen/Qwen3-32B") tok = AutoTokenizer.from_pretrained("Qwen/Qwen3-32B") model = PeftModel.from_pretrained(base, "dementor-research/dpo_writingprompts_qwen3-32b_as_nemotron-nano-30b-a3b_seed1") ``` Part of the dementor matrix: 4 source models × 3 cross-targets × 3 train datasets × 3 seeds × 2 stages = 216 adapters.

NoteOct 3huggingface.co/dementor-research/dpo_writingprompts_qwen3-32b_as_nemotron-nano-30b-a3b_seed1

Source context: 4 downloads · 0 likes · Library peft · Repo dementor-research/dpo_writingprompts_qwen3-32b_as_nemotron-nano-30b-a3b_seed1

NoteJul 17huggingface.co/dementor-research/dpo_writingprompts_qwen3-32b_as_nemotron-nano-30b-a3b_seed1

Source context: 3 downloads · 0 likes · Library peft · Repo dementor-research/dpo_writingprompts_qwen3-32b_as_nemotron-nano-30b-a3b_seed1