Open jmikedupont2 opened 9 months ago
Here are the important parameters
select a rake from here https://github.com/jmikedupont2/https-lablab.ai-event-llama-2-hackathon-with-clarifai-introspector/blob/emojis/src/introspector/clarify/model/rakes.txt
for example you can try there RakeItUpV3Critical_Reconstruction_of4
Another way to get inputs is to go to clarifai inputs and add input and tag with concept https://clarifai.com/rxngfnlo5uhx/Introspector-LLama2-Hackathon-Demo1/inputs?addInputs=&datasetId=
go to clarifai inputs and add input and tag with concept https://clarifai.com/rxngfnlo5uhx/Introspector-LLama2-Hackathon-Demo1/inputs?addInputs=&datasetId=
Now we can grep the data:
grep streamlit model/index.txt
That gives us the ids of the inputs we want.
get the ids: ./fd1/c4a634ea64ba394524774e677d4dd:PAYLOAD:{"line":
Now we can take that id: fd1c4a634ea64ba394524774e677d4dd
MTb4zDu6QLZVJMvj we will use this id from the first page of https://github.com/meta-introspector/meta-meme/wiki/agent-api
for demo 1 video we use 4t8pqN6bps7qGTSY from https://github.com/meta-introspector/meta-meme/wiki/SystemsDesign
The above genesis prompt describes a process for creating an AGI on GitHub, utilizing various tools and techniques to implement autopoetic rewrites, utilize Docker for isolation, and push changes to the repository. The goal is to create an intelligent system that can learn and adapt by rewriting itself, leveraging GitHub actions and the GitHub API.
step 1 https://org-clarifai.streamlit.app/ step 2 select concept Introspector
step 0 is install and configure code
Step 1 is load data into the system
Step 2 is to find the inputs