Modelpublic

uwuwuwuwuwuwu/kimi-k2.5-reward-hacking-step-0

sourceHugging Faceupdated 4mo agoView on Hugging Face

0likes14downloads

Model Card

Tinker LoRA Adapter

This repository contains a LoRA adapter exported from Tinker.

Usage

python

from transformers import AutoModelForCausalLM

adapter_id = "uwuwuwuwuwuwu/kimi-k2.5-reward-hacking-step-0"
base_model = "moonshotai/Kimi-K2-Thinking"

model = AutoModelForCausalLM.from_pretrained(adapter_id, device_map="auto")

Source

tinker://3d10c12f-6551-5538-8817-56fcb7c45099:train:0/sampler_weights/base_untrained

Details

—Base model: moonshotai/Kimi-K2-Thinking
—LoRA rank: 32
—Trained modules: attn=True, mlp=True, unembed=True