manyoso

9 ms·

61.
by manyoso 3y ago
We take the 4all seriously. Everything we've put out is under Apache 2.0 or MIT license.
62.
by manyoso 3y ago
This latest release included a ton of bugfixes and a new model. Hope people are pleasantly surprised. Still needs work, but we're making progress quite fast.
63.
by manyoso 3y ago
All of the above and the root permission will be fixed jn future release if yuh want to wait
64.
by manyoso 3y ago
There is a big update released today… maybe give it another go?
65.
by manyoso 3y ago
Did you try the new version released today? The initial version had a bad bug that is now fixed with drastically better performance. Many users are reporting some quite high quality responses on discord. Check it out! Link on the homepage…
66.
by manyoso 3y ago
GPT4 is like SearchEngine4
67.
by manyoso 3y ago
https://youtu.be/8ZW1E017VEc Video by the founder of the project to put in perspective what this is and is not
68.
by manyoso 3y ago
It isn’t affiliated with openai’s gpt4
69.
by manyoso 3y ago
No; this is an open source project sponsored by nomic.ai
70.
by manyoso 3y ago
It is entirely different model. Based on GPT-J which is an open source LLM.
71.
by manyoso 3y ago
Your loss. GPT is a term of art in the industry. It isn’t owned by OpenAI.
72.
by manyoso 3y ago
Download the installer and run it. That is it.
73.
by manyoso 3y ago
New version was just released for those who already have it installed with a number of bugfixes and improvements to the model. If you have it, then click "Check for updates" in the drawer that slides out of the hamburger menu butt
74.
GPT4All Chat – Locally-running AI chat application powered by the GPT4All-J(gpt4all.io)
253 points by manyoso 3y ago|164 comments
75.
by manyoso 3y ago
GPT4All Chat is a locally-running AI chat application powered by the GPT4All-J Apache 2 Licensed chatbot. The model runs on your computers CPU, works without an internet connection and sends no chat data to external servers (unless you opt-
76.
by manyoso 3y ago
There is a solution. A completely open source community driven model that can be run on your desktop (windows, linux, mac) with a click of one button: https://gpt4all.io/index.html
77.
by manyoso 3y ago
But Coq is strongly normalizing… it always halts. What gives?
78.
by manyoso 4y ago
It should only hallucinate when asked. Humans can do the same but when humans mistake their hallucinations for reality we usually consider it problematic. Especially if the AI is going to be considered a reference which many seem willing to
79.
by manyoso 4y ago
I log the incorrect answers but I will also log the correct ones from now on, but until I do you can run the test yourself or believe me when I say gpt4all is not just repeating all the answers. None of the models have exhibited that behavi
80.
by manyoso 4y ago
That is mathematically equivalent. Just subtract 1 from each type of answer: correct, incorrect, and unknown :)
81.
by manyoso 4y ago
This is correct behavior in that the models are expected to choose one of the multiple choice options and not write prose answer.
82.
GPT4All outscores GPT-3.5 on new hallucination metric(github.com)
36 points by manyoso 4y ago|13 comments
83.
by manyoso 4y ago
GPT4All is outscoring GPT-3.5 in the real trivia dataset as well as the 'none of the above' category as well as coming in a very close second place in the fake/trick questions which are designed to measure a model's prop
84.
by manyoso 4y ago
Thanks. Before creating I looked for hallucination specific tests but couldn't find any. If anyone else knows of other tests I'd love to hear about them.
85.
New metric for LLM hallucinations with results for GPT-3, GPT3-5 and Alpaca Lora(github.com)
4 points by manyoso 4y ago|3 comments
86.
by manyoso 4y ago
HALTT4LLM - Hallucination Trivia Test for Large Language Models This project is an attempt to create a common metric to test LLM's for progress in eliminating hallucinations; the most serious current problem in widespread adoption of L
87.
by manyoso 8y ago
Thank you. One of the best comments I have ever seen on this forum.
88.
by manyoso 8y ago
Mueller did not bring this indictment. He presented the evidence to a Grand Jury of your fellow citizens. Those fellow citizens saw enough evidence to convince them to return the indictments. If the individuals indicted ever stand for the c
89.
by manyoso 8y ago
Then all Musk needs to do is not allow Tesla to "come to find" out that LIDAR is necessary before the technology has advanced to such a degree that they can retrofit the Model X and to do so when the cost has come down sufficientl
90.
by manyoso 8y ago
Exactly! Only Tesla Haters would think that "full self-driving features" would refer to features that enable fully autonomous driving... every common sense consumer out there will clearly understand that when Elon Musk says full s

More ›