Plugin for LLM adding models hosted on AI Horde
Project description
llm-horde
LLM plugin for models hosted by AI Horde
Installation
First, install the LLM command-line utility.
Now install this plugin in the same environment as LLM.
llm install https://github.com/spinagon/llm-horde/archive/refs/heads/main.zip
Configuration
First Register an account which will generate for you an API key. Store that key somewhere.
- If you do not want to register, you can use '0000000000' as api_key to connect anonymously. If you don't set your key, it will be used automatically. However anonymous accounts have the lowest priority when there's too many concurrent requests!
- To increase your priority you will need a unique API key and then to increase your Kudos. Read how Kudos are working.
You can set that as an environment variable called AIHORDE_KEY
, or add it to the llm
set of saved keys using:
llm keys set aihorde
Enter key: <paste key here>
Usage
To list available models, run:
llm models list
You should see a list that looks something like this:
AI Horde: horde/Henk717/airochronos-33B
AI Horde: horde/tgi-fp16-8k/Xwin-LM/Xwin-LM-70B-V0.1
AI Horde: horde/Gryphe/MythoMax-L2-13b
To run a prompt against a model, pass its full model ID to the -m
option, like this:
llm -m horde/Henk717/airochronos-33B "Five spooky names for a pet vampire bat"
You can set a shorter alias for a model using the llm aliases
command like so:
llm aliases set airochronos horde/Henk717/airochronos-33B
Now you can prompt it using:
cat llm_horde.py | llm -m airochronos -s 'write some pytest tests for this'
If you use "horde" as a model name, you can select one or more models matching a regex pattern using "pattern" option:
llm -m horde -o pattern "13B" "What is the capital of Lebanon?"
You can change the maximum length of returned text by using "max_tokens" option (80 by default), most workers will not :
llm -m horde -o pattern mythomax -s 'Write a sonnet' -o max_tokens 120
If you want to continue generating text without new prompt, try:
llm -c ""
To start interactive chat, type
llm chat -m horde -o pattern llama2
Project details
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
File details
Details for the file llm-horde-0.5.tar.gz
.
File metadata
- Download URL: llm-horde-0.5.tar.gz
- Upload date:
- Size: 9.9 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.11.4
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 |
b7ca2afb63ef3f9c19987e1d4a61df59f2a67a39ec3a76fc78d52b03577231ca
|
|
MD5 |
1112e1f1b45739837322d9b59df66bc0
|
|
BLAKE2b-256 |
e9cd5e29d5ccc4457b42222f9af93e21cec86c2fa52bfcbc868e03ca8ee58520
|
File details
Details for the file llm_horde-0.5-py3-none-any.whl
.
File metadata
- Download URL: llm_horde-0.5-py3-none-any.whl
- Upload date:
- Size: 10.7 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.11.4
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 |
60d111fd348eae2b243d696f3f64c8aaedd43e48df48c30b6c88f09088f721bb
|
|
MD5 |
261768ddee2392e35ad18091dc531f5f
|
|
BLAKE2b-256 |
c1c59795443160567775d7d591b308f69a112cea80aa1e716f5f5db7269cfe30
|