How to add gemma 3 wasm for custom model?

[EnricoBeltramo]

I trained my customized model based on gemma 3 1b and I converted using tutoriaal here: https://llm.mlc.ai/docs/deploy/webllm.html#model-records-in-webllm but I don't understand where to find the related wasm in model_lib. There already exist support for this model? where is the wasm?

[rajnish-concretio]

If you are compiling your custom model on top of another model, then you need to build the WASM library and provide the path to it in the mlc-chat-config.json file.

[akaashrp]

@EnricoBeltramo The WASM for gemma-3-1b-it is available at https://github.com/mlc-ai/binary-mlc-llm-libs/blob/main/web-llm-models/v0_2_80/gemma-3-1b-it-q4f16_1-ctx4k_cs1k-webgpu.wasm. You should be able to add a config record that points to this WASM in config.ts. If you require different quantizations, you can follow the steps in the MLC-LLM docs (https://llm.mlc.ai/docs/deploy/webllm.html) to compile your own WASMs.