gemma-3-270m-it-q4f16_1-MLC

gemma-3-270m-it - MLC WebGPU

This is a quantized version of google/gemma-3-270m-it optimized for WebGPU inference using MLC-LLM.

Model Details

• Base Model: google/gemma-3-270m-it
• Quantization: q4f16_1
• Target: WebGPU
• Library: MLC-LLM
• WebLLM Compatible: Yes

Usage

This model is designed to be used with WebLLM in web browsers supporting WebGPU.

WebLLM Integration

import * as webllm from "@mlc-ai/web-llm";

const appConfig = {
  model_list: [{
    model: "https://huggingface.co/llinguini/gemma-3-270m-it-q4f16_1-MLC",
    model_id: "gemma-3-270m-it-MLC",
    model_lib: "https://huggingface.co/llinguini/gemma-3-270m-it-q4f16_1-MLC/resolve/main/libs/gemma-3-270m-it-webgpu.wasm",
    required_features: ["shader-f16"]
  }]
};

const engine = await webllm.CreateMLCEngine("gemma-3-270m-it-MLC", { appConfig });

Files

• mlc-chat-config.json: Model configuration for MLC-LLM
• ndarray-cache.json: Metadata cache
• tokenizer.json: Fast tokenizer configuration
• params_shard_*.bin: Quantized model parameters
• libs/gemma-3-270m-it-webgpu.wasm: Compiled WebGPU library

License

This model follows the same license as the base model: Gemma License.