Giovanni Puccetti
Giovanni Puccetti
**Describe the bug** `luxor_overrides.jl` makes Luxor available in scope where `using Javis` is called **To Reproduce** ```julia julia> using Javis julai> Luxor Luxor ``` **Expected Behavior** ```julia julia> using Javis...
I don't know if this is possible and even less how to do it, but wouldn't it make sense to try and parallelize the rendering of the frames to speed...
## PR Checklist If you are contributing to `Javis.jl`, please make sure you are able to check off each item on this list: - [ ] Did I update `CHANGELOG.md`...
## PR Checklist If you are contributing to `Javis.jl`, please make sure you are able to check off each item on this list: - [ ] Did I update `CHANGELOG.md`...
**Is your feature request related to a problem? Please explain.** It is generally hard to make a translation from an user-unknow (but javis-known) current object position to a specific point....
adding CoCa
The PR idea is to add the CoCa model as implemented in https://github.com/lucidrains/CoCa-pytorch, using existing parts as much as possible. Ideally adding possibilty to choose between custom and non custom...
In [this comment](https://github.com/lucidrains/CoCa-pytorch/issues/2#issuecomment-1320707789) by @rom1504 it has been mentioned the possiblity to add [coca-pytorch](https://github.com/lucidrains/CoCa-pytorch/) to this repo to train a version of it, if nobody is doing it and it...
This PR adds the possibility to use a CLS token as a model parameter in TextEncoder to simplify the logic of `CoCa`
Hi, I am not sure I get this right, howevr initially `prev_pos` is set to 0 but it looks like it might be better set to `start_pos-1` otherwise the mask...