Slightly update Sentence Transformers snippet
#2
by tomaarsen HF Staff - opened
Hello!
Preface
I got it working now! Very nicely done, I quite like this parameter count for code embedding models. Are you planning a paper or blogpost with some more details? I'm curious about your training data setup.
Pull Request overview
- Use the trust_remote_code parameter immediately in the Sentence Transformers snippet
Details
You can use the trust_remote_code parameter immediately - it'll get propagated down to the model, tokenizer, and config when loading.
This simplifies the overall loading code somewhat.
- Tom Aarsen
tomaarsen changed pull request status to open
Thank you for your PR. We will present more details in our upcoming paper.
cfli changed pull request status to merged