RichardErkhov/Mihaiii_-_Pallas-0.5-LASER-0.5-gguf

Quantization made by Richard Erkhov.

Github

Discord

Request more models

Pallas-0.5-LASER-0.5 - GGUF

• —Model creator: https://huggingface.co/Mihaiii/
• —Original model: https://huggingface.co/Mihaiii/Pallas-0.5-LASER-0.5/

Original model description: --- basemodel: Mihaiii/Pallas-0.5-LASER-0.4 inference: false license: other licensename: yi-license license_link: https://huggingface.co/01-ai/Yi-34B/blob/main/LICENSE metrics:

• —accuracy ---

This model has a LASER intervention on Mihaiii/Pallas-0.5-LASER-0.4 .

Configs used:

• —lnum: 54
• —lnames: mlp (meaning: ["mlp.gateproj.weight", "mlp.upproj.weight", "mlp.down_proj.weight"])
• —rate: 8
• —dataset: bigbench (subset: causal_judgement)
• —intervention type: rank-reduction

In order to replicate on a single A100, you can use my branch (the original code will throw OOM for 34b models).