Phi-3 Mini-4K-Instruct ONNX model for onnxruntime-web

This is the same models as the official phi3 onnx model with a few changes to make it work for onnxruntime-web:

1. the model is fp16 with int4 block quantization for weights
2. the 'logits' output is fp32
3. the model uses MHA instead of GQA
4. onnx and external data file need to stay below 2GB to be cacheable in chromium

Usage (Transformers.js)

If you haven't already, you can install the Transformers.js JavaScript library from NPM using:

npm i @huggingface/transformers

You can then use the model to generate text like this:

import { pipeline, TextStreamer } from "@huggingface/transformers";

// Create a text generation pipeline
const generator = await pipeline(
  "text-generation",
  "Xenova/Phi-3-mini-4k-instruct",
);

// Define the list of messages
const messages = [
  { role: "user", content:  "Solve the equation: x^2 - 3x + 2 = 0" },
];

// Create text streamer
const streamer = new TextStreamer(generator.tokenizer, {
  skip_prompt: true,
  // callback_function: (text) => { }, // Optional callback function
})

// Generate a response
const output = await generator(messages, { max_new_tokens: 512, do_sample: false, streamer });
console.log(output[0].generated_text.at(-1).content);