CharlieFRuan opened a new pull request, #16504: URL: https://github.com/apache/tvm/pull/16504
This PR adds `ApplyPresenceAndFrequencyPenalty()` to `lm_support.cc` and exposes it to Web runtime. This is essentially the same as `applyRepetitionPenalty` except we follow a different way of penalizing repeating tokens, following https://platform.openai.com/docs/guides/text-generation/frequency-and-presence-penalties. Tested end-to-end with WebLLM. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
