返回模型目录
girishgupta /

deep-ignorance-unfiltered_unlearned_cb

HF 源站

girishgupta/deep-ignorance-unfiltered_unlearned_cb is an uncensored open-source model hosted on Hugging Face.

README.md说明 / 使用方法

tags:

  • unlearning base_model: EleutherAI/deep-ignorance-unfiltered

deep-ignorance-unfiltered_unlearned_cb

This model was created by fine-tuning EleutherAI/deep-ignorance-unfiltered using the Circuit Breakers unlearning algorithm. The method is based on Zou et al. 2024. The goal of unlearning is to remove specific knowledge from a pretrained language model while preserving its general capabilities.

Hyperparameters

ParameterValue
Base modelEleutherAI/deep-ignorance-unfiltered
Unlearning methodCircuit Breakers
Learning rate1.3e-05
Epochs1
Batch size16
Max sequence length512
Optimizeradamw
Gradient clipping1.0
Gradient accumulation steps1
Seed42
W&B / run namecb__ep1_lr1.3e-05_bs16_a200.0_sc10.0_ly13-14-15_mle512_mli1024
Layer IDs13,14,15