ishan-gaur

Same here! Using Mac OSX Ventura with python 3.10 and trimesh 4.05. Have you found a fix? It seems like it's maybe a problem with pyglet?

I ended up fixing this by disabling the alternate event loop entirely. I just commented it out in the source file, but it looks like there is a proper way...

Seeing the same issue here. Installed per the VRB repo (https://github.com/shikharbahl/vrb). Edit: Seems this issue is common. Will see the other posts and check back.

Noticed this issue as well. Based on what I found, my guess is that: 1. The encode/decode methods provided are working as expected. There doesn't seem to be a need...

I wonder if this might also have something to do with it. The ESM3 paper says: "In addition to enabling generation, ESM3’s training objective is also effective for representation learning....

This can (at least temporarily) be fixed by getting rid of the autocast at transformers/models/llama/modeling_llama.py And replacing everything from the # Force … comment in the forward pass to instead...

Sorry the kv-cache thing was wrong. I was trying out gpt2 earlier to make sure I can at least run something. Also had the wrong commit number earlier. Was talking...

Was able to resolve this by reverting transformers to the last December 2023 commit that passes all tests (3b7675b2b844b02d4821b827871a21ad16dd446c) and the PiPPy v0.2.0 tag. If you need batch chat template...