Giuseppe Attanasio
Giuseppe Attanasio
Completely agree. This messes up everything, @kylejusticemagnuson you should definitively merge it.
Hi! Do you mean that you want to pass to functions pandas Series?
Up. I am having the same installation issue on Ubuntu 20.04. **Edit:** I solved the issue by manually updating `pyicu`, `pycld2`, and `morfessor` to the latest versions. I think that...
> Hi, > > Batched inference with LLaVa is supported in Hugging Face Transformers. See here for an example: https://github.com/huggingface/transformers/blob/a49f4acab3c1eea82907e12f82eafbd4673deb39/tests/models/llava/test_modeling_llava.py#L245. Hey, @NielsRogge I've stumbled upon this issue today. It seems...
Hey, I'm chiming in here since I think my use case is kind of related. I'm trying to add a custom metric evaluation where I need to use some information...
I'm not super keen on the idea of internal sorting by length as it could break some outer user-define logic. But I don't even know what's the best option here...
Yea if the original order is preserved I guess that works.
Hello Donya! It seems an error related to DDP. How are you executing the script? Can you also show me the run command? Is it a `python ...`, `accelerate ......
Also, I'm noticing you are loading the model without any quantization or dtype specified. That means you are loading the weights in full scale and probably running OOM. Can you...
Hello @donya-rooein, did any of the fixes work fine?