Qwen3-Embedding-4B-q4f16_1-MLC-b2
MLC-compiled Qwen3-Embedding-4B (Apache-2.0) in web-llm 0.2.84's format for WebGPU in-browser embedding. Built locally with mlc-llm (tvmjs 0.24.0-dev3 / mlc-ai 0.26.dev246 generation) + emsdk 3.1.56.
- Quantization: q4f16_1 · context: 2048 · prefill chunk: 2048 · max batch: 2
- Embedding dim: 2560 (last-token pooling, L2-normalized — done in the compiled model)
- WebGPU: works on NVIDIA (Dawn); grid/buffer limits sized for batch 2/2048
Files
| File | Purpose |
|---|---|
mlc-chat-config.json |
model config (ctx 2048, chunk 2048, max_batch 2) |
tensor-cache.json / ndarray-cache.json |
weight manifest (identical bytes) |
params_shard_0.bin … params_shard_73.bin |
quantized weights (~2.1 GB) |
tokenizer.json, tokenizer_config.json, vocab.json, merges.txt |
Qwen3 tokenizer |
qwen3-embedding-4b-q4f16_1-ctx2048_batch2-webgpu.wasm |
the model library (wasm) |
Use in web-llm (0.2.84)
const record = {
model: 'https://huggingface.co/skkut/qwen3-embedding-4b-q4f16_1-MLC-b2/resolve/main/',
model_id: 'Qwen3-Embedding-4B-q4f16_1-MLC-b8',
model_lib: 'https://huggingface.co/skkut/qwen3-embedding-4b-q4f16_1-MLC-b2/resolve/main/qwen3-embedding-4b-q4f16_1-ctx2048_batch2-webgpu.wasm',
model_type: 1, // EmbeddingModelType
vram_required_MB: 4400,
};
Query-side texts should use the Qwen3 instruction prefix
(Instruct: Given a web search query, retrieve relevant passages that answer the query\nQuery: …).