huihui_ai/exaone3.5-abliterated:32b-q8

huihui_ai/ exaone3.5-abliterated:32b-q8_0

1,990 Downloads Updated 1 year ago

EXAONE 3.5 is a collection of instruction-tuned bilingual (English and Korean) generative models ranging from 2.4B to 32B parameters, developed and released by LG AI Research.

2.4b 7.8b 32b

ollama run huihui_ai/exaone3.5-abliterated:32b-q8_0

curl http://localhost:11434/api/chat \
  -d '{
    "model": "huihui_ai/exaone3.5-abliterated:32b-q8_0",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

from ollama import chat

response = chat(
    model='huihui_ai/exaone3.5-abliterated:32b-q8_0',
    messages=[{'role': 'user', 'content': 'Hello!'}],
)
print(response.message.content)

import ollama from 'ollama'

const response = await ollama.chat({
  model: 'huihui_ai/exaone3.5-abliterated:32b-q8_0',
  messages: [{role: 'user', content: 'Hello!'}],
})
console.log(response.message.content)

Details

Updated 1 year ago

1 year ago

12506abcb271 · 34GB ·

model

archexaone

parameters32B

quantizationQ8_0

34GB

template

{{- range $i, $_ := .Messages }} {{- $last := eq (len (slice $.Messages $i)) 1 -}} {{ if eq .Role "s

375B

system

You are EXAONE model from LG AI Research, a helpful assistant.

62B

license

EXAONE AI Model License Agreement 1.1 - NC This License Agreement (“Agreement”) is entered into

14kB

params

{ "repeat_penalty": 1, "stop": [ "[|endofturn|]" ], "temperature": 1 }

62B

Readme

This is an uncensored version of LGAI-EXAONE/exaone-35 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