Find answers from the community

Updated 5 months ago

Hi Team I m running in to connection

At a glance
Hi Team, I'm running in to connection error issue importing any functions from llama_index, could anyone help to resolve this? for more context: I'm running it offline without internet connection..
W
a
L
7 comments
Are you using local llm and embed model?
I have seen this error I think 2-3 days back also. It was probably related to tiktoken or some other library. That requires some files to be downloaded if running for the first time.
Yes I'm using local llm.. could you please elaborate how you resolved this issue of tiktoken? Thanks!
I'm assuming its tiktoken. Tiktoken download some files for the first time you run this library.
For ref:https://discord.com/channels/1059199217496772688/1154653483715268698/1154803857520594996
Cool thanks!.. I tried the steps but I'm facing issue to how I can change the gpt2 function in tiktoken ? What is hex the url mean? Could you help me with that? Thanks!
This is pretty difficult. You basically have to read the tiktoken source code to figure out what it's downloading and where it's caching it. The cache is a hex of the filename I think which is.. not fun
No easy workaround here tbh
Add a reply
Sign up and join the conversation on Discord