-
First thanks for your great job!
Now We're trying to replace the vision encoder in llava, i.e., clip-l-336, with RADIO. Under the default LLaVA 1.5 settings, we pretrain a multimodal projection MLP a…
-
For certain images the default psm gives `Empty Page` as output while `--psm 6` and others give the correct result.
Suggest that in cases where default psm results in `Empty Page`, try recognizing…
-
Hello, I have a problem:
I use `libmxnet.so` compiled with mxnetv0.8 comparison to mxnetv1.0(or v1.3 and v1.4) to run my code to infer a batch images using C++ API, I find the inference speed with …
-
### 🐛 Describe the bug
When attempting to export the UDOP model to ONNX from the transformers library, the torch.onnx.export() command fails with a RuntimeError. Below is a minimal example to repro…
-
Sorry to bother you again. I found that different resolution of an image will have a big impact on the recognition effect. For example, if the original resolution of some input images is reduced to 80…
-
* 다음주 : 9장까지 (가능하면 부록A도)
* 다음주 내용 요약: 주선미님
* 다음 스터디 날짜: 2021-11-23 pm9:00
* 위키 정리: 김유리님
-
During the hackathon I just picked a model without knowing too much about an ideal model fit. Think better and change if needed
-
I have a METS where all FLocats are LOCTYPE=URL (as required by DFGViewer), but local directories FULLTEXT and MAX do exist as well.
Unfortunately, digital-derivans does not seem to like this repre…
-
Test of a small sample of real life images gives better results with the [older 7seg.traineddata](https://github.com/Shreeshrii/tessdata_ssd/raw/master/7seg.traineddata). Unfortunately I have deleted …
-
After bumping to:
Calamari 2.1.4
TF 2.4.3
and setting
```
self.predictor.data.params.pre_proc.run_parallel = False
self.predictor.data.params.post_proc.run_parall…