ModelVérifiéExterne

Dpo Writingprompts Qwen3 32b As Llama 3.1 8b 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_llama-3.1-8b_seed1/tree/7f1b83795ba71f2e003279b0dc2e878d4bc8f95e

TypeLoRAModèle de baseQwen/Qwen3-32BVersionadapter_modelÉditeur@dementor-researchCClasse

Source du modèle

Type: LoRA
Taille du fichier: 1,02 GB
Modèle de base: Qwen/Qwen3-32B
Version: vadapter_model
Mis à jour: 3 oct. 2026
Éditeur: @dementor-research
Téléchargements: 6
Paramètres: 32B
Source: Hugging Face

Description de la source

dpo_writingprompts_qwen3-32b_as_llama-3.1-8b_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_llama-3.1-8b_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_llama-3.1-8b_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 FacePrincipalVersion 7f1b83795ba71f2e003279b0dc2e878d4bc8f95e

huggingface.co/dementor-research/dpo_writingprompts_qwen3-32b_as_llama-3.1-8b_seed1/tree/7f1b83795ba71f2e003279b0dc2e878d4bc8f95e

Vérifié 3 oct.

Artefacts du modèle

1 artefact

adapter_model.safetensors

safetensors · 1,02 GB · SHA-256 ca430868a284…78e9 · Hugging Face

Télécharger

Extraits de sources

3 extraits

Readme11 juil.huggingface.co/dementor-research/dpo_writingprompts_qwen3-32b_as_llama-3.1-8b_seed1

--- base_model: Qwen/Qwen3-32B library_name: peft tags: - lora - dpo - dementor-research --- # dpo_writingprompts_qwen3-32b_as_llama-3.1-8b_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_llama-3.1-8b_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_llama-3.1-8b_seed1") ``` Part of the dementor matrix: 4 source models × 3 cross-targets × 3 train datasets × 3 seeds × 2 stages = 216 adapters.

Note17 juil.huggingface.co/dementor-research/dpo_writingprompts_qwen3-32b_as_llama-3.1-8b_seed1

Source context: 6 downloads · 0 likes · Library peft · Repo dementor-research/dpo_writingprompts_qwen3-32b_as_llama-3.1-8b_seed1

Note11 juil.huggingface.co/dementor-research/dpo_writingprompts_qwen3-32b_as_llama-3.1-8b_seed1

Source context: 18 downloads · 0 likes · Library peft · Repo dementor-research/dpo_writingprompts_qwen3-32b_as_llama-3.1-8b_seed1