Closed temberature closed 1 year ago
Still working to brush out the breaking changes due to bundling stuff, but in the meantime, either:
get_encoding("cl100k_base")
, which should correspond to gpt-3.5-turbo
1.0.0-alpha.3
version, which should add support for encoding_for_model("gpt-3.5-turbo")
ETA for latest version is between 2 and 3 days 😄
I attempted using the first cl100k_base, and it appears to yield the same outcome as reported in https://github.com/openai/tiktoken/issues/45.
Thank you for your kind reply~
Version 1.0.0
has been published, which should add the support for gpt-3.5-turbo
, gpt-4
and gpt-4-32k
models. LMK if the issue presists.
I found that there is already a gpt-3.5-turbo branch, but it is not released to https://www.npmjs.com, I would like to ask about when it will be released?
BTW. Thanks for your great work.