-
Can you provide more detailed information of CLIP weight?Is the CLIP weight released by OpenAI or OpenCLIP? If the weight is provided by OpenCLIP, is the CLIP weight trained on LAION-2B or DataComp-1B…
-
The recent release 0.2.96 of webdataset has introduced a new argument to `tar_file_expander` called `eof_value` set by default to `{}` that is backward incompatible. Specifically, training in OpenCLIP…
-
### Feature request
fine-tuning OpenClip with Hugingface's PEFT (such as LoRA)
### Motivation
fine-tuning OpenClip with Hugingface's PEFT (such as LoRA)
### Your contribution
refer to https://g…
-
Hi @slavabarkov, I truly appreciate your contribution in making the CLIP model viable on Android devices.
Can I have the code you used to export the OpenCLIP model to ONNX format?
I want to ex…
-
I observe a potential memory leak when training with OpenCLIP on a WebDataset. The memory usage is flat during an epoch and then increases at the beginning of the next epoch (evaluation disabled). If …
-
Hi,
I would like to request the addition of support for using different models from Hugging Face and our own OpenCLIP models for the CLIP encoder. Currently, the implementation is limited to 'vit_b…
-
Hi
Thank you for your excellent work. I am curious about the model using different versions of CLIP, e.g., CLIP-large or OpenCLIP-huge. If you have tried these different teacher models, can you sha…
-
It seems that the author did not provide the decode code. Can I use the decode code in [OpenCLIP](https://github.com/mlfoundations/open_clip) to decode the text?
-
### Motivation
Hi.
I saw a new version of the LAION dataset just came out (Aug 30th, 2024). The samples from the CSAM dataset were removed from LAION.
Do you intend to re-train your InternViT model…
-
It is not clear from readme:
1. Did you use your copy of OpenClip (from src) or OpenClip from pypi?
2. Why did you decided to use QuickGelu if it is slower and much memory-hungry?