Update README.md

This commit is contained in:
Charles Packer
2023-10-12 22:08:41 -07:00
committed by GitHub
parent d7ec7c1ff6
commit 2f80095535

View File

@@ -1,10 +1,14 @@
<div align="center">
# [MemGPT](https://memgpt.ai) (MemoryGPT)
# [MemGPT](https://memgpt.ai)
[![Discord](https://img.shields.io/discord/1111172801899012102?label=Discord&logo=discord&logoColor=green&style=flat-square)](https://discord.gg/9GEQrxmVyE)
:rocket: Teaching LLMs memory management for unbounded context :rocket:
</div>
## Setup
## Quick Setup
Set up dependencies:
@@ -18,11 +22,13 @@ Add your OpenAI API key to your environment:
export OPENAI_API_KEY=YOUR_API_KEY
```
By default MemGPT will use `gpt-4`, so your API key will require `gpt-4` API access.
## What is MemGPT?
## Running MemGPT via the CLI
MemoryGPT (or MemGPT in short) is a system that intelligently manages different memory tiers in LLMs in order to effectively provide extended context within the LLM's limited context window. For example, MemGPT knows when to push critical information to a vector database and when to retrieve it later in the chat for perpetual conversations. Learn more about MemGPT in our [paper](https://arxiv.org/abs/2310.08560).
To run MemGPT in CLI mode, simply run `main.py`:
## Try MemGPT in your CLI
To run MemGPT for as a conversation agent in CLI mode, simply run `main.py`:
```sh
python3 main.py
@@ -32,7 +38,6 @@ To create a new starter user or starter persona (that MemGPT gets initialized wi
```sh
# assuming you created a new file /memgpt/humans/examples/me.txt
# note: no .txt extension, just 'me'
python main.py --human me
```
@@ -64,13 +69,21 @@ Basic commands:
print the current contents of agent memory
```
Debugging commands:
### Support
```text
/pop
undo the last message in the conversation
/heartbeat
send a heartbeat system message to the agent
/memorywarning
send a memory warning system message to the agent
```
* For CLI, no `.txt` extension, just `me` for `--human` flag
* By default MemGPT will use `gpt-4`, so your API key will require `gpt-4` API access.
* If you would like to further debug MemGPT, we provide the further options:
```text
/pop
undo the last message in the conversation
/heartbeat
send a heartbeat system message to the agent
/memorywarning
send a memory warning system message to the agent
```
If you have any further questions, or have anything to share, we are excited to hear your feedback!
* For issues and feature requests, please [open a GitHub issue](https://github.com/cpacker/MemGPT/issues).