NanoCode012
NanoCode012
Based on what @hengjiUSTC linked, if I understand it correctly, fp16 adapter training must use fp32 for trainable and fp16 for non-trainable. They provided a utility function `cast_mixed_precision_params(peft_model, dtype)` for...
I think while looking through the modeling code, there is a flash attention arg in the Attention layer that defaults to true. However, the model config does not support such...
@harshdhamecha , we've deprecated flash attn 1 quite a while back. You can just omit setting FA in the yaml to disable it.
I believe this should've been solved as the HF repo updated their code.
@jeffchuber , I weirdly got this error while trying to ingest a lot of PDF files, 10 or so. Each file is split into chunks, embedded, before being inserted using...
> @NanoCode012 499 is the dimensionality of the vectors? Any anything over that breaks? That is _very_ odd! Sorry for late reply @jeffchuber . No, I use openai , so...
Hello @happybit , I re-created how the langchain inserts docs but edited to only submit at most 499 embeddings per time. Fyi: It's not 500 docs for me, but 500...
You can `pip install -U so-vits-svc-fork==...` replace it with your version. However, the command does work for me.
Hey @vgel , would it be possible to share on the method of adding new models ? I’m also interested in MoE models, which I saw you explicitly mention. What...
The PR has been merged. Thanks Johan!