Skip to content

Conversation

@CharlieFRuan
Copy link
Member

This PR is essentially #16078 but for web runtime.

Add function getParamsFromCacheByName() for web runtime, hence creating CreateByName() in ndarray cache. Create() to GetParams() is analogous to CreateByName() to GetParamsByName().

@CharlieFRuan
Copy link
Member Author

cc @tqchen

@tqchen tqchen merged commit 9e4e17c into apache:unity Dec 3, 2023
tqchen pushed a commit to mlc-ai/web-llm that referenced this pull request Dec 4, 2023
Make web-llm runtime compatible with SLIM models, a new workflow in
mlc-llm. This depends on apache/tvm#16198 and is
currently not backward-compatible with non-SLIM model, since we
introduced a new global function to TVM.

The main difference is reading the metadata and loading the parameters.

Besides, we now read stop_tokens from the conversation template instead
of relying on metadata.
Archermmt pushed a commit to Archermmt/tvm that referenced this pull request Dec 18, 2023
atebites-hub pushed a commit to atebites-hub/web-llm that referenced this pull request Oct 4, 2025
Make web-llm runtime compatible with SLIM models, a new workflow in
mlc-llm. This depends on apache/tvm#16198 and is
currently not backward-compatible with non-SLIM model, since we
introduced a new global function to TVM.

The main difference is reading the metadata and loading the parameters.

Besides, we now read stop_tokens from the conversation template instead
of relying on metadata.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants