Notices for the code bundled in web-llm.js (WebLLM 0.2.87, @mlc-ai/web-llm, Apache-2.0)

WebLLM's bundle includes code from the projects below. Each is under the licence named; the full
Apache License 2.0 text is in LICENSE, the loglevel licence in loglevel-LICENSE. The last section
covers the model libraries, which are separate files.

------------------------------------------------------------------------------
Apache TVM runtime (@mlc-ai/web-runtime), Apache-2.0
https://github.com/apache/tvm
------------------------------------------------------------------------------
The TVM web runtime files carry the Apache Software Foundation's licence header. The NOTICE file of
Apache TVM reads:

Apache TVM
Copyright 2019-2023 The Apache Software Foundation

This product includes software developed at
The Apache Software Foundation (http://www.apache.org/).

------------------------------------------------------------------------------
@mlc-ai/web-tokenizers (tokenizers-cpp, compiled to WebAssembly), Apache-2.0
https://github.com/mlc-ai/tokenizers-cpp
------------------------------------------------------------------------------
tokenizers-cpp wraps and includes:
- Hugging Face tokenizers, Apache-2.0, https://github.com/huggingface/tokenizers
  Copyright The Hugging Face team.
- SentencePiece, Apache-2.0, https://github.com/google/sentencepiece
  Copyright Google LLC.

------------------------------------------------------------------------------
openai-node (type definitions adapted by WebLLM), Apache-2.0
https://github.com/openai/openai-node
------------------------------------------------------------------------------
The input to OpenAI API, directly adopted from openai-node with small tweaks.
Copyright 2024 OpenAI. Licensed under the Apache License, Version 2.0.

------------------------------------------------------------------------------
tslib (TypeScript runtime helpers), 0BSD
https://github.com/microsoft/tslib
------------------------------------------------------------------------------
Copyright (c) Microsoft Corporation.

Permission to use, copy, modify, and/or distribute this software for any
purpose with or without fee is hereby granted.

THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES WITH
REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY
AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY SPECIAL, DIRECT,
INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM
LOSS OF USE, DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR
OTHER TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR
PERFORMANCE OF THIS SOFTWARE.

------------------------------------------------------------------------------
loglevel 1.9.2, MIT, Copyright (c) 2013 Tim Perry (loglevel-LICENSE)
------------------------------------------------------------------------------

------------------------------------------------------------------------------
Model libraries (Qwen3.5-*-webgpu.wasm, served from /models/libs/v0_2_87/)
https://github.com/mlc-ai/binary-mlc-llm-libs (web-llm-models/v0_2_87/base/, commit 146ed9cd)
------------------------------------------------------------------------------
Not part of web-llm.js: each model downloads its own library, unchanged from the MLC project's build.
They are compiled by the MLC project from MLC LLM (https://github.com/mlc-ai/mlc-llm, Apache-2.0) and
Apache TVM (Apache-2.0, through its MLC fork https://github.com/mlc-ai/relax, whose NOTICE is the TVM
NOTICE above), with the Qwen3.5 model definition from MLC LLM. That source is Apache-2.0; the binary
repository itself states no licence (docs/legal-review.md M16). The NOTICE file of MLC LLM reads:

MLC LLM

Copyright (c) 2023-2025 by MLC LLM Contributors