GPT4All, A free-to-use, locally running, privacy-aware chatbot. No GPU or internet required.

3 years ago by A10@kerala.party to c/technology

Free, local and privacy-aware chatbots
redw0rm 4 points 3 years ago

I've tried it too. The nous_hermes model worked very well ( except that its a little slow than chatgpt )

path: 0 4432769, hotness: undefined, score: 4, children: 0
Lemmling 3 points 3 years ago

Nice, I don't have enough RAM ๐Ÿ

path: 0 4446403, hotness: undefined, score: 3, children: 0
A10 2 points 3 years ago

Used it today to write some reports, really impressive considering it is running locally

path: 0 4407888, hotness: undefined, score: 2, children: 0
Albin9326 1 point 3 years ago
path: 0 4524563, hotness: undefined, score: 1, children: 2
A10 2 points 3 years ago

Unfortunately No. you need a CPU with AVX instructions to run the model. This software has an option to enable API access, so in theory you can write an app for that

path: 0 4524563 4524943, hotness: undefined, score: 2, children: 1
Albin9326 2 points 3 years ago
path: 0 4524563 4524943 4525383, hotness: undefined, score: 2, children: 0

go to feed...