Mikhail Kim

Hello, thank you for your code. Could you please explain why you set requires_grad=True for input tensor with edge_features when training model? https://github.com/ranahanocka/MeshCNN/blob/15b83cc6a4db968baf6cf595df78995a9c2dcee3/models/mesh_classifier.py#L49

Installation of Milvus

13

When I try to build the image for Milvus, the following error occurs: `Step 6/7 : RUN python3 -c 'import pyknowhere' ---> Running in e84180a73171 Illegal instruction (core dumped) The...

Zero-shot capabilities on ImageNet

2

Hello, I evaluated ALBEF 14M on ImageNetV2 classification task and it showed relatively low accuracy: top1 – 32.9, top5 - 60.7. How do you think what reasons of such results?...

Can't load CLIP-ViT-H-14-frozen-xlm-roberta-large-laion5B-s13B-b90k

7

When I am trying to load this model, the following error occurred: ```python import open_clip model, preprocess_train, preprocess_val = open_clip.create_model_and_transforms('hf-hub:laion/CLIP-ViT-H-14-frozen-xlm-roberta-large-laion5B-s13B-b90k') Downloading (…)ip_pytorch_model.bin: 100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 4.77G/4.77G [02:20

[Usage] Batch inference with Llava 1.5

12

### Describe the issue Currenty, only inference with batch_size=1 is possible. If I undestood correctly, these things should be changed to make batch inference: 1. position_ids should be shifted, because...

Hello. Why do N(k), m(k) depend on the number of epoch, not on the number of step?