A question solver plugin for ovos
Project description
LLM.cpp Persona
Give OpenVoiceOS some sass with Alpaca.cpp, GPT4All.cpp or Bloomz.cpp
This plugin requires providing the path to the executable and model, it uses subprocess which allows it to work with these programs without requiring python bindings
Dedicated plugins may exist for each LLM
Examples
- "What is best in life?"
- "Do you like dogs"
- "Does God exist?"
Usage
Spoken answers api
from ovos_solver_llmcpp import LLMcppSolver
ALPACA_MODEL_FILE = "./models/ggml-alpaca-7b-q4.bin"
GPT4ALL_MODEL_FILE = "./models/gpt4all-lora-quantized.bin"
# binpath = "~/alpaca.cpp/chat"
binpath = "~/gpt4all.cpp/chat"
bot = LLMcppSolver({"model": GPT4ALL_MODEL_FILE,
"executable_path": binpath})
sentence = bot.spoken_answer("Qual é o teu animal favorito?", {"lang": "pt-pt"})
# Meus animais favoritos são cães, gatos e tartarugas!
for q in ["Does god exist?",
"what is the speed of light?",
"what is the meaning of life?",
"What is your favorite color?",
"What is best in life?"]:
a = bot.get_spoken_answer(q)
print(q, a)
Project details
Release history Release notifications | RSS feed
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 ovos-solver-llmcpp-plugin-0.0.0a3.tar.gz
.
File metadata
- Download URL: ovos-solver-llmcpp-plugin-0.0.0a3.tar.gz
- Upload date:
- Size: 9.3 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.9.16
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 3399c9d5c5856955f39e3ac49e162a9ddfa961e4568d36f59fb8400f35305d38 |
|
MD5 | d4056333cc86a95a1cc52dba9ab4dee3 |
|
BLAKE2b-256 | d9d4f0550e3f4c4590e5d543673b0a5c8f28574f681fe078373a0ca95b9c72a8 |
File details
Details for the file ovos_solver_llmcpp_plugin-0.0.0a3-py3-none-any.whl
.
File metadata
- Download URL: ovos_solver_llmcpp_plugin-0.0.0a3-py3-none-any.whl
- Upload date:
- Size: 10.1 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.9.16
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 99dfeacea4f75113e68b27726a16427224d211f1a1348e935ccfbc07b5f85db8 |
|
MD5 | 99833f4b7f3ba58e544966e7560fe710 |
|
BLAKE2b-256 | 667d11b06e20ec9d7a3f016c113a553d0e828fa8049018831b45df8a2a0c5e9f |