whisper.cpp/examples/talk-llama
2023-11-07 11:03:21 +02:00
..
prompts talk-llama : add alpaca support (#668) 2023-03-29 23:01:14 +03:00
.gitignore talk, talk-llama : add basic example script for eleven-labs tts (#728) 2023-04-14 19:53:58 +03:00
CMakeLists.txt cmake : fix talk-llama build 2023-11-07 11:03:21 +02:00
eleven-labs.py examples : update elevenlabs scripts to use official python API (#837) 2023-05-24 21:11:01 +03:00
llama.cpp sync : ggml (backend v2, k-quants, CUDA opts, Metal opts, etc.) (#1422) 2023-11-03 21:35:05 +02:00
llama.h sync : ggml (backend v2, k-quants, CUDA opts, Metal opts, etc.) (#1422) 2023-11-03 21:35:05 +02:00
README.md talk-llama : move up-to-date demo to top (#1417) 2023-11-02 18:50:13 +02:00
speak talk-llama : update to latest llama.cpp 2023-09-15 20:06:31 +03:00
speak.bat speak scripts for Windows 2023-06-01 22:45:00 +10:00
speak.ps1 speak scripts for Windows 2023-06-01 22:45:00 +10:00
talk-llama.cpp talk-llama : fix n_gpu_layers usage again (#1442) 2023-11-07 10:51:27 +02:00
unicode.h sync : ggml (backend v2, k-quants, CUDA opts, Metal opts, etc.) (#1422) 2023-11-03 21:35:05 +02:00

talk-llama

Talk with an LLaMA AI in your terminal

Latest perf as of 2 Nov 2023 using Whisper Medium + LLaMA v2 13B Q8_0 on M2 Ultra:

https://github.com/ggerganov/whisper.cpp/assets/1991296/d97a3788-bf2a-4756-9a43-60c6b391649e

Previous demo running on CPUs

Demo Talk

Building

The talk-llama tool depends on SDL2 library to capture audio from the microphone. You can build it like this:

# Install SDL2 on Linux
sudo apt-get install libsdl2-dev

# Install SDL2 on Mac OS
brew install sdl2

# Build the "talk-llama" executable
make talk-llama

# Run it
./talk-llama -mw ./models/ggml-small.en.bin -ml ../llama.cpp/models/llama-13b/ggml-model-q4_0.gguf -p "Georgi" -t 8
  • The -mw argument specifies the Whisper model that you would like to use. Recommended base or small for real-time experience
  • The -ml argument specifies the LLaMA model that you would like to use. Read the instructions in https://github.com/ggerganov/llama.cpp for information about how to obtain a ggml compatible LLaMA model

Session

The talk-llama tool supports session management to enable more coherent and continuous conversations. By maintaining context from previous interactions, it can better understand and respond to user requests in a more natural way.

To enable session support, use the --session FILE command line option when running the program. The talk-llama model state will be saved to the specified file after each interaction. If the file does not exist, it will be created. If the file exists, the model state will be loaded from it, allowing you to resume a previous session.

This feature is especially helpful for maintaining context in long conversations or when interacting with the AI assistant across multiple sessions. It ensures that the assistant remembers the previous interactions and can provide more relevant and contextual responses.

Example usage:

./talk-llama --session ./my-session-file -mw ./models/ggml-small.en.bin -ml ../llama.cpp/models/llama-13b/ggml-model-q4_0.gguf -p "Georgi" -t 8

TTS

For best experience, this example needs a TTS tool to convert the generated text responses to voice. You can use any TTS engine that you would like - simply edit the speak script to your needs. By default, it is configured to use MacOS's say or Windows SpeechSynthesizer, but you can use whatever you wish.

Discussion

If you have any feedback, please let "us" know in the following discussion: https://github.com/ggerganov/whisper.cpp/discussions/672?converting=1