Loading
Have a ComfyUI workflow? Get a readable Workflow Doc in your browser.Make a workflow readable.Try it now ↗
Dementor imitation (disguise) LoRA adapter — dataset gsm8k, seed 1.
Model source
Source description
Dementor imitation (disguise) LoRA adapter — dataset gsm8k, seed 1.
Sources
1 sourceVerified Aug 8
Model artifacts
1 artifactSource excerpts
3 excerpts--- library_name: peft tags: - dementor - imitation-disguise - lora - gsm8k base_model: openai/gpt-oss-120b --- # dpo_gsm8k_gpt-oss-120b_as_gpt-oss-20b_seed1 Dementor imitation (disguise) LoRA adapter — dataset **gsm8k**, **seed 1**. - **Method:** DPO - **Source model (fine-tuned / disguised):** `gpt-oss-120b` (base: `openai/gpt-oss-120b`) - **Target model being imitated:** `gpt-oss-20b` - **Dataset:** gsm8k | **Seed:** 1 Mirrored from a Tinker checkpoint (`sampler_weights`) into standard PEFT LoRA format via `tinker_cookbook.weights.build_lora_adapter`. Registry key == repo id == `dpo_gsm8k_gpt-oss-120b_as_gpt-oss-20b_seed1`. Part of the current Dementor imitation adapter set.
gpt-oss-120b (base: openai/gpt-oss-120b)gpt-oss-20bMirrored from a Tinker checkpoint (sampler_weights) into standard PEFT LoRA
format via tinker_cookbook.weights.build_lora_adapter. Registry key == repo id
== dpo_gsm8k_gpt-oss-120b_as_gpt-oss-20b_seed1. Part of the current Dementor imitation adapter set.
Source context: 34 downloads · 0 likes · Library peft · Repo dementor-research/dpo_gsm8k_gpt-oss-120b_as_gpt-oss-20b_seed1
Source context: 32 downloads · 0 likes · Library peft · Repo dementor-research/dpo_gsm8k_gpt-oss-120b_as_gpt-oss-20b_seed1