Category: Artificial Intelligence

[privateGPT] is a model to questions on local documents without an internet connection, using the power of LLMs. 100% private, no data leaves your execution environment at any point. You can ingest documents and ask questions without an internet connection!

[WebLLM] is a modular, customizable javascript package that directly brings language model chats directly onto web browsers with hardware acceleration. Everything runs inside the browser with no server support and accelerated with WebGPU. We can bring a lot of fun opportunities to build AI assistants for everyone and enable privacy while enjoying GPU acceleration.

[h2oGPT] is a large language model (LLM) fine-tuning framework and chatbot UI with document(s) question-answer capabilities. Documents help to ground LLMs against hallucinations by providing them context relevant to the instruction. h2oGPT is fully permissive Apache V2 open-source project for 100% private and secure use of LLMs and document embeddings for document question-answer.

Bing Chat on Edge – [bing.com/chat]

[Chatbot Arena]

[Install Instructions]

Caution: TensorFlow 2.10 was the last TensorFlow release that supported GPU on native-Windows. Starting with TensorFlow 2.11, you will need to install TensorFlow in WSL2, or install tensorflow or tensorflow-cpu and, optionally, try the TensorFlow-DirectML-Plugin

[ML.NET] is an open source and cross-platform machine learning framework for C#.

[ML.NET – Save and load trained models]

[TensorFlow – Creating C# Applications using TensorFlowSharp]

[Keras.NET]

[tf2onnx] – Convert TensorFlow, Keras, Tensorflow.js and Tflite models to ONNX.

[Keras ImageDataGenerator and Data Augmentation] – This article goes into the details of using data augmentation to increase your training data sets. By default augmentation just modifies your original dataset. With some small adjustments you can grow the training data as intended.

Train AI to win at Starcraft and DOTA

[Run PyTorch models in the browser using ONNX.js]

According to this video from 3 years ago, ONNX.js only does interference mode so it can’t do training. It still works if you can do your training in Python and port the trained model over to JS.

[Google and Apple vets raise $17M for Fixie, a large language model startup based in Seattle]

[Announcing a next-generation AI Copilot in Microsoft Power Apps that will transform low-code development]

[How to Run a ChatGPT Alternative on Your Local PC]

Level 1:

Level 2:

Level 3:

Level 3.5:

Level 4: