talkGPT4All
A voice chatbot based on GPT4All and talkGPT.
Please check more details in this blog post (in Chinese).
If you are looking for the older version of talkGPT4All, please checkout to dev/v1.0.0 branch.
Installation
Install using pip (Recommend)
talkgpt4all is on PyPI, you can install it using simple one command:
pip install talkgpt4all
Install from source code
Clone the code:
git clone https://github.com/vra/talkGPT4All.git <ROOT>
Install the dependencies and talkGPT4All in a python virtual environment:
cd <ROOT>
python -m venv talkgpt4all
source talkgpt4all/bin/activate
pip install -U pip
pip install -r requirements.txt
Extra dependencies for Linux users
We use pyttsx3 to convert text to voice. Please note that on Linux ,You need to install dependencies:
sudo apt update && sudo apt install -y espeak ffmpeg libespeak1
Usage
Open a terminal and type talkgpt4all to begin:
talkgpt4all
Use different LLMs
You can choose different LLMs using --gpt-model-type <type>, all available choices:
{
"ggml-gpt4all-j-v1.3-groovy"
"ggml-gpt4all-j-v1.2-jazzy"
"ggml-gpt4all-j-v1.1-breezy"
"ggml-gpt4all-j"
"ggml-gpt4all-l13b-snoozy"
"ggml-vicuna-7b-1.1-q4_2"
"ggml-vicuna-13b-1.1-q4_2"
"ggml-wizardLM-7B.q4_2"
}
Use different Whisper models
You can choose whisper model type using --whisper-model-type <type>, all available choices:
{
"tiny.en"
"tiny"
"base.en"
"base"
"small.en"
"small"
"medium.en"
"medium"
"large-v1"
"large-v2"
"large"
}
Tune voice rate
You can tune the voice rate using --voice-rate <rate>, default rate is 165. the larger the speak faster.
e.g.,
talkgpt4all --whisper-model-type large --voice-rate 150
RoadMap
- Add source building for llama.cpp, with more flexible interface.
- More LLMs
- Add support for contextual information during chating.
- Test code on Linux,Mac Intel and WSL2.
- Add support for Chinese input and output.
- Add Documents and Changelog
contributions are welcomed!
Metadata
Release files for talkgpt4all 2.5.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| talkgpt4all-2.5.0.tar.gz | 5.0 kB | Without provenance | Details for talkgpt4all-2.5.0.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| talkgpt4all-2.5.0-py3-none-any.whl | Python 3 | none | any | Details for talkgpt4all-2.5.0-py3-none-any.whl |
Total release size: 10.4 kB
Release files / talkgpt4all-2.5.0.tar.gz
| Download URL | talkgpt4all-2.5.0.tar.gz |
|---|---|
| Size | 5.0 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
f8acdbdd370240923cf37a2bd241bc9ebcdfa78c50da2f2f6cacdf1a2a4a8863
|
|
BLAKE2b-256 checksum How to use checksums |
024745a28cc7cc7ad424aba0ae09f374b60a0fb08b492f7e0ab71b48a9a530fd
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/4.0.2 CPython/3.8.17
|
Release files / talkgpt4all-2.5.0-py3-none-any.whl
| Download URL | talkgpt4all-2.5.0-py3-none-any.whl |
|---|---|
| Size | 5.3 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
4c10849e8d2966749d9df7d9d1862d7af4367f01a97dc5060d2ec52ec9b366f8
|
|
BLAKE2b-256 checksum How to use checksums |
f417bbc32f1d3d7a709a5c44df33e30223ab483c3b5754e95d00e8091fbe1e4d
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/4.0.2 CPython/3.8.17
|
Security for talkgpt4all 2.5.0
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.