huihui_ai/
internlm3-abliterated:8b-instruct-q4_K_M

872 11 months ago

InternLM3 has open-sourced an 8-billion parameter instruction model, InternLM3-8B-Instruct, designed for general-purpose usage and advanced reasoning.

8b

11 months ago

3f800aa325c9 · 5.4GB ·

model
llama
·
8.8B
·
Q4_K_M
system
You are an AI assistant whose name is InternLM (书生·浦语). - InternLM (书生·浦语) is a co
license
The code is licensed under Apache-2.0, while model weights are fully open for academic research and
params
{ "stop": [ "<|im_end|>", "<|action_end|>", "</s>" ] }
template
{{ if .System }}<|im_start|>system {{ .System }}<|im_end|> {{ end }}{{ if .Prompt }}<|im_start|>user

Readme

This is an uncensored version of internlm/internlm3-8b-instruct created with abliteration (see remove-refusals-with-transformers to know more about it).
This is a crude, proof-of-concept implementation to remove refusals from an LLM model without using TransformerLens.

References

HuggingFace