thooton / muse Goto Github PK
View Code? Open in Web Editor NEWLet's create synthetic textbooks together :)
License: MIT License
Let's create synthetic textbooks together :)
License: MIT License
It seems google has caught on to this. My API key has been suspended and it looks like a couple others have too.
C:\Users\Nick>curl -H "Content-Type: application/json" -d "{\"contents\":[{\"parts\":[{\"text\":\"Write a story about a magic backpack\"}]}]}" -X POST https://generativelanguage.googleapis.com/v1beta/models/gemini-pro:generateContent?key=API_KEY
{
"error": {
"code": 403,
"message": "Permission denied: Consumer 'api_key:<API_KEY' has been suspended.",
"status": "PERMISSION_DENIED",
"details": [
{
"@type": "type.googleapis.com/google.rpc.ErrorInfo",
"reason": "CONSUMER_SUSPENDED",
"domain": "googleapis.com",
"metadata": {
"service": "generativelanguage.googleapis.com",
"consumer": "projects/993395320710"
}
}
]
}
}
I'm getting TimeoutError()
from every API's every few seconds/minutes. I've to re-run the script to continue.
error in API1_REDACTED: TimeoutError()
error in API2_REDACTED: TimeoutError()
error in API3_REDACTED: TimeoutError()
error in API4_REDACTED: TimeoutError()
error in API5_REDACTED: TimeoutError()
error in API6_REDACTED: TimeoutError()
error in API7_REDACTED: TimeoutError()
error in API8_REDACTED: TimeoutError()
error in API9_REDACTED: TimeoutError()
error in API10_REDACTED: TimeoutError()
error in API11_REDACTED: TimeoutError()
error in API12_REDACTED: TimeoutError()
error in API13_REDACTED: TimeoutError()
I'm not quite sure whether it's a issue arising from my end or an issue within the script. It's a PITA to re-run the script every few seconds/minutes. I would like a permanent solution.
Sorry for poor coding knowledge, but this is what I have found so far.
So I have a Colab Pro account, figured I can use a notebook and run some synthetic textbooks through it. Initially, it kept on giving me an error, that it couldn't find the "candidates". I was able to get Huggyingface to download the modules once I placed the a new muse folder, nested into another muse folder. This showed me that the "candidates" argument error possibly had to deal with the Gemini API implementation.
Original Code
[Gemini]
api_keys = []
api_endpoint = https://generativelanguage.googleapis.com/v1beta/models/gemini-pro:generateContent?key=
[Parameters]
temperature = 1.0
top_p = 0.99
[Misc]
out_dir = ./textbooks
count_per_file = 1000
begin_index = 0
verbose_exceptions = False
Edited Code
[Gemini]
api_endpoint = https://generativelanguage.googleapis.com/v1beta/models/gemini-pro:generateContent?key=[my actual key typed here]
[Parameters]
temperature = 1.0
top_p = 0.99
[Misc]
out_dir = ./textbooks
count_per_file = 1000
begin_index = 0
verbose_exceptions = False
If I place my key within the brackets, it does not work, can not find it. If I use the actual API endpoint from Google, and adding my key to the end of it, it works.
Just wanted to pass this along if anyone else if having the issue.
I've noticed that in every file, there are lines with no content, represented as {"text": ""}
.
Hi @thooton,
I'm just quickly skimming over this project. I noticed that you call your actors Phi and Epsilon and they are told to have an argument where they take a strong position on one side (an idea which I quite like and also planned on implementing after reading "The Socratic Method").
Isn't the naming a little dangerous, considering that a prompt for one of the Phi models might be "You are Phi, an AI model that ...".
I would assume the model trained on this data would have a strong bias to produce the opinion that your Phi actor had.
Hi! This is a really cool project, but for us to get even close to the 250b token mark, we'd need a LOT of participants. I'm sure there are many willing to contribute idle cycles for these few API calls, but I'm quite doubtful this free Gemini spree will last long. So rather than adding many people, I'd suggest quickly implementing API key rotation into the script. @TanvirOnGH wrote a refactor that I would built off of.
Are there any potential issues/workarounds here (such as the difficulty of registering multiple accounts to a single phone number on google) that I should be aware of?
A declarative, efficient, and flexible JavaScript library for building user interfaces.
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. ๐๐๐
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google โค๏ธ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.