ModelListadosExterno

Sft Gsm8k Qwen3.6 27b As Qwen3.5 4b Seed43

Dementor imitation (disguise) LoRA adapter — dataset gsm8k, seed 43.

Fontehuggingface.co/dementor-research/sft_gsm8k_qwen3.6-27b_as_qwen3.5-4b_seed43/tree/d3dab855e72f38517ef6770d6564161bad9ae884

TipoLoRAModelo baseQwen/Qwen3.6-27BVersãoFP32Unratednota

Fonte do modelo

Tipo: LoRA
Tamanho do arquivo: 921 MB
Modelo base: Qwen/Qwen3.6-27B
Versão: vFP32
Atualizado: 6 de nov. de 2026
Downloads: 4
Fonte: Hugging Face

Descrição da fonte

sft_gsm8k_qwen3.6-27b_as_qwen3.5-4b_seed43

Dementor imitation (disguise) LoRA adapter — dataset gsm8k, seed 43.

Method: SFT
Source model (fine-tuned / disguised): qwen3.6-27b (base: Qwen/Qwen3.6-27B)
Target model being imitated: qwen3.5-4b
Dataset: gsm8k | Seed: 43

Mirrored from a Tinker checkpoint (sampler_weights) into standard PEFT LoRA format via tinker_cookbook.weights.build_lora_adapter. Registry key == repo id == sft_gsm8k_qwen3.6-27b_as_qwen3.5-4b_seed43. Part of the current Dementor imitation adapter set.

Fontes

1 fonte

Hugging FacePrimárioVersão d3dab855e72f38517ef6770d6564161bad9ae884

huggingface.co/dementor-research/sft_gsm8k_qwen3.6-27b_as_qwen3.5-4b_seed43/tree/d3dab855e72f38517ef6770d6564161bad9ae884

Verificado 6 de nov.

Download files

Repository files

File inventory only; completeness is unverified.

Artefatos de modelo

1 artefato

adapter_model.safetensors

safetensors · FP32 (unverified) · 921 MB · SHA-256 1f628e80fdd0…6b81 · Hugging Face

Download file

Trechos de fonte

3 trechos

Readme30 de ago.huggingface.co/dementor-research/sft_gsm8k_qwen3.6-27b_as_qwen3.5-4b_seed43

--- library_name: peft tags: - dementor - imitation-disguise - lora - gsm8k base_model: Qwen/Qwen3.6-27B --- # sft_gsm8k_qwen3.6-27b_as_qwen3.5-4b_seed43 Dementor imitation (disguise) LoRA adapter — dataset **gsm8k**, **seed 43**. - **Method:** SFT - **Source model (fine-tuned / disguised):** `qwen3.6-27b` (base: `Qwen/Qwen3.6-27B`) - **Target model being imitated:** `qwen3.5-4b` - **Dataset:** gsm8k | **Seed:** 43 Mirrored from a Tinker checkpoint (`sampler_weights`) into standard PEFT LoRA format via `tinker_cookbook.weights.build_lora_adapter`. Registry key == repo id == `sft_gsm8k_qwen3.6-27b_as_qwen3.5-4b_seed43`. Part of the current Dementor imitation adapter set.

Note6 de nov.huggingface.co/dementor-research/sft_gsm8k_qwen3.6-27b_as_qwen3.5-4b_seed43

Source context: 4 downloads · 0 likes · Library peft · Repo dementor-research/sft_gsm8k_qwen3.6-27b_as_qwen3.5-4b_seed43

Note30 de ago.huggingface.co/dementor-research/sft_gsm8k_qwen3.6-27b_as_qwen3.5-4b_seed43

Source context: 2 downloads · 0 likes · Library peft · Repo dementor-research/sft_gsm8k_qwen3.6-27b_as_qwen3.5-4b_seed43