wenyang

@gchhablani HI, my evaluation stuck on the status of "Submitted", I see you helped lots of guys, can you help me with that? It's been 2 days after I submitted....

For people who may encounter errors like following when running `from pygaggle.rerank.transformer import MonoT5`: ``` Traceback (most recent call last): File "", line 1, in File "/root/pygaggle/pygaggle/rerank/transformer.py", line 15, in...

I think this should be `start_at_layer` rather than `stop_at_layer`. https://transformerlensorg.github.io/TransformerLens/generated/code/transformer_lens.HookedTransformer.html#transformer_lens.HookedTransformer.HookedTransformer.forward:~:text=Optional%5Bint%5D%20(,None%20(run%20the%20full%20model).

Is it ok to directly modify `model.cfg.n_ctx`?

2

Hi, I have a question about model configuration. I want to use the model with a longer context. As most of the models' default context window is set to 2048,...

If you are submitting a bug report, please fill in the following details and use the tag [bug]. **Describe the bug** If we load the model with `bfloat16`, calculating OV...

### Question Hi, just curious. Why do `gemma-2-2b` and `gemma-2-2b-it` need 8GB VRAM even when I load it with `bfloat16` or `float16`? I didn't find this on the other models...