You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Voice language models that run even on an old and simple device
#3710
Do you know PocketPal AI?
PocketPal AI is an app that lets you chat with advanced AI models on your device without needing an internet connection.
Actually, I used this app for 10 days (https://pocketpal.dev/ check out this site), and now I’m back to using GPT4All from Nomic AI here on my PC, but PocketPal AI is an app that made me wish for its features here in GPT4All.
Do you think Nomic AI could help me bring those features to my GPT4All, can you help me?
The version of GPT4All I’m using is v3.10.0, and the main PocketPal AI features I’d like to bring to my GPT4All are voice language model conversations and permission to use tools, text editing, playing music, searching the internet — the basics — all on the CPU. No account, no cloud, no subscription required.
I got attached to the idea of using interactions with voice language models, and if the AI runs on an old and simple device, it can also be implemented in the GPT4All on the Desktop or Laptop I already have — real language models, voices, optimized to make the most of the CPU, open-source — free and without an account. The versions of native modules that connect JavaScript to the engines. llama.cpp bridges LLM inference and JSI; whisper bridges speech recognition and text-to-speech. The llama.cpp inference engines run language models in quantized GGUF format. whisper.cpp runs Whisper speech recognition models in GGML format. The ONNX Runtime runs the Kokoro, Kitten, and Supertonic voice models natively in its own runtime.
This is just a suggestion, but I hope it helps to take an important step toward implementing the PocketPal AI functionalities in my GPT4All. Are you ready to start this journey?
in portuguese
Você conhece a PocketPal AI?
O PocketPal AI é um aplicativo que permite conversar com modelos avançados de IA no seu dispositivo sem precisar de conexão com a internet.
Na verdade, eu usei esse aplicativo por 10 dias (https://pocketpal.dev/ veja nesse site), agora voltei a usar o GPT4All da Nomic AI, aqui no PC, mas o PocketPal AI é um aplicativo que me fez desejar as funções dele aqui no GPT4All.
Será que a Nomic AI pode me ajudar e trazer as funções para o meu GPT4All, consegue me ajudar?
A versão do GPT4All que estou usando é a v3.10.0 e as principais funcionalidades do PocketPal AI que gostaria de trazer para o meu GPT4All, seriam as conversas com modelos de linguagem de voz e permissão para que use ferramentas, edição de texto, tocar música, pesquisar na internet, o básico — tudo no CPU. Sem conta, sem nuvem, sem mensalidade necessária.
Me apeguei a ideia de usar as interações com o modelos de linguagem de voz, e se a IA roda em um aparelho antigo e simples também pode ser implementado no GPT4All do Desktop ou Notebook que já possuo — modelos de linguagem reais, vozes, ajustadas para aproveitar ao máximo a CPU, de código aberto — gratuitos e sem conta. As versões de módulos nativos que conectam JavaScript aos motores. llama.cpp faz a ponte entre inferência LLM e JSI; whisper faz a ponte entre reconhecimento de fala e texto para fala. Os motores de inferência llama.cpp roda modelos de linguagem no formato GGUF quantizado. O whisper.cpp executa modelos de reconhecimento de fala Whisper no formato GGML. O ONNX Runtime roda os modelos de voz Kokoro, Kitten e Supertonic nativamente em seu próprio runtime.
Essa é apenas uma sugestão, mas espero que isso ajude a dar um passo importante em direção à implementação das funcionalidades do PocketPal AI no meu GPT4All. Você está preparado para começar essa jornada?
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
Uh oh!
There was an error while loading. Please reload this page.
Do you know PocketPal AI?
PocketPal AI is an app that lets you chat with advanced AI models on your device without needing an internet connection.
Actually, I used this app for 10 days (https://pocketpal.dev/ check out this site), and now I’m back to using GPT4All from Nomic AI here on my PC, but PocketPal AI is an app that made me wish for its features here in GPT4All.
Do you think Nomic AI could help me bring those features to my GPT4All, can you help me?
The version of GPT4All I’m using is v3.10.0, and the main PocketPal AI features I’d like to bring to my GPT4All are voice language model conversations and permission to use tools, text editing, playing music, searching the internet — the basics — all on the CPU. No account, no cloud, no subscription required.
I got attached to the idea of using interactions with voice language models, and if the AI runs on an old and simple device, it can also be implemented in the GPT4All on the Desktop or Laptop I already have — real language models, voices, optimized to make the most of the CPU, open-source — free and without an account. The versions of native modules that connect JavaScript to the engines. llama.cpp bridges LLM inference and JSI; whisper bridges speech recognition and text-to-speech. The llama.cpp inference engines run language models in quantized GGUF format. whisper.cpp runs Whisper speech recognition models in GGML format. The ONNX Runtime runs the Kokoro, Kitten, and Supertonic voice models natively in its own runtime.
This is just a suggestion, but I hope it helps to take an important step toward implementing the PocketPal AI functionalities in my GPT4All. Are you ready to start this journey?
in portuguese
Você conhece a PocketPal AI?
O PocketPal AI é um aplicativo que permite conversar com modelos avançados de IA no seu dispositivo sem precisar de conexão com a internet.
Na verdade, eu usei esse aplicativo por 10 dias (https://pocketpal.dev/ veja nesse site), agora voltei a usar o GPT4All da Nomic AI, aqui no PC, mas o PocketPal AI é um aplicativo que me fez desejar as funções dele aqui no GPT4All.
Será que a Nomic AI pode me ajudar e trazer as funções para o meu GPT4All, consegue me ajudar?
A versão do GPT4All que estou usando é a v3.10.0 e as principais funcionalidades do PocketPal AI que gostaria de trazer para o meu GPT4All, seriam as conversas com modelos de linguagem de voz e permissão para que use ferramentas, edição de texto, tocar música, pesquisar na internet, o básico — tudo no CPU. Sem conta, sem nuvem, sem mensalidade necessária.
Me apeguei a ideia de usar as interações com o modelos de linguagem de voz, e se a IA roda em um aparelho antigo e simples também pode ser implementado no GPT4All do Desktop ou Notebook que já possuo — modelos de linguagem reais, vozes, ajustadas para aproveitar ao máximo a CPU, de código aberto — gratuitos e sem conta. As versões de módulos nativos que conectam JavaScript aos motores. llama.cpp faz a ponte entre inferência LLM e JSI; whisper faz a ponte entre reconhecimento de fala e texto para fala. Os motores de inferência llama.cpp roda modelos de linguagem no formato GGUF quantizado. O whisper.cpp executa modelos de reconhecimento de fala Whisper no formato GGML. O ONNX Runtime roda os modelos de voz Kokoro, Kitten e Supertonic nativamente em seu próprio runtime.
Essa é apenas uma sugestão, mas espero que isso ajude a dar um passo importante em direção à implementação das funcionalidades do PocketPal AI no meu GPT4All. Você está preparado para começar essa jornada?
All reactions