goodai / charlie-mnemonic Goto Github PK
View Code? Open in Web Editor NEWCharlie Mnemonic: The First Personal Assistant with Long-Term Memory
Home Page: https://www.goodai.com/introducing-charlie-mnemonic/
Charlie Mnemonic: The First Personal Assistant with Long-Term Memory
Home Page: https://www.goodai.com/introducing-charlie-mnemonic/
Every time the system attempts to integrate the DuckDuckGo search function call, it fails with a rate limit issue. While the LLM recovers with a certain amount of grace, telling me that there has been an error, that doesn't make it actually function.
processing function search_duckduckgo response: Error performing DuckDuckGo search: _aget_url() https://links.duckduckgo.com/d.js DuckDuckGoSearchException: Ratelimit
Built on Mac (M2 Pro) with Brew installed postgresql. I was able to register, but nothing happens when using the registered user to log in. The console just reports INFO: 127.0.0.1:61459 - "POST /login/ HTTP/1.1" 200 OK
.
It is validating the user, as entering a non-existent user generates a User login failed
error and a 401 unauthorized
in the console output.
Hi, this is not an issue but a question. Apology first to exploit here because I couldn't find other channel to ask a question.
I took some time to read the source code, and found out that Charlie's Mnemonic does not use https://github.com/GoodAI/goodai-ltm , whose performance is quite impressive if the results in https://github.com/GoodAI/goodai-ltm-benchmark is valid. Is there any reason for this? I had Charlie's Mnemonic tested, and got an impression that its LTM performance (information integration, episodic memory, spatial memories) is very short of what's attested in the benchmark.
Thanks in advance.
I am trying to understand the codebase to extend it for using localLLMs easily.
Are there any resources which can help me understand the codebase faster?
cc: @airobinnet
Two of the three database URLs have the wrong protocol specified. Should be postgresql
not postgres
. SQLAlchemy no longer accepts postgres
.
DATABASE_URL=postgres://postgres:postgres@localhost:5432/postgres
NEW_DATABASE_URL=postgresql://postgres:postgres@localhost:5432/postgres
TEST_DATABASE_URL=postgres://postgres:postgres@localhost:5432/postgres
I have installed both Charlie-mnemonic and Ollama on my windows system, however charlie is always trying to use openai api.
Are there any steps that I missed, there's really no documentation on this other than mentioning that ollama is supported
The Windows release of Charlie Mnemonic does not actually show that it is 0.6.29, but rather 0.6.28. This does seem to bring up some question of whether the new release has actually been pushed into that branch and they just neglected to update one of the version tags or whether there are some deeper issues.
A clean start from with docker-compose up shows up this error:
unsupported frontend protocol 1234.5679
A little investigation showed that it's actually an old bug in PostgreSQL with an easy fix: just add PGGSSENCMODE=disable
to the psdb environment in docker-compose.yml (see here on stackoverflow)
psdb:
environment:
- PGGSSENCMODE=disable
By the way, when the start fails this way starting again won't work even after adjusting the docker-compose.yml because the DB didn't initialize correctly.
To fix, you need to remove the PG volume manually and then restart:
$ docker-compose down
$ docker volume ls
$ docker volume rm charlie-mnemonic_postgres-data
$ docker-compose up
Hello
With a fresh install of 0.6.33 how does one start using a local LLAMA3 service rather than OpenAI?
Regards,
Damon
I am strongly passionate about what you guys are building. As I am going through the code base, I see many areas in which we can improve e.g. decrease the implicit reliance on OpenAI APIs e.g. currently token use is based on GPT-4 even if you plan to use local LLM.
In order for me to become a regular contributor, we need to decrease the MR review time to a few days. I feel a bit hesitant in raising MRs when I feel that they might get overlooked for weeks on end. I can contribute in both documentation and tech for this project. This is quite a niche project, I would love to have a short call with you guys about the project's future etc.
Hi.
When following the setup instructions for Linux I run into the following issue:
failed to solve: rpc error: code = Unknown desc = failed to solve with frontend dockerfile.v0: failed to read dockerfile: failed to mount .../.local/share/docker/tmp/buildkit-mount2959019754: [{Type:bind Source:.../.local/share/docker/overlay2/92zij4v1rkbwe4w1wicvpr1gj/diff Options:[rbind ro]}]: operation not permitted
Docker Compose failed with error level 17.
System info:
Any idea how to solve this? Is the setup script compatible with docker rootless?
Thanks in advance.
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.