baudm / parseq

Scene Text Recognition with Permuted Autoregressive Sequence Models (ECCV 2022)

https://huggingface.co/spaces/baudm/PARSeq-OCR

baudm/parseq Issues

Multi GPU training issue
Closed a month ago
mathematical formula recognition
Updated a month ago
how to convert to torchscript using script method. using trace method cause the model cannot handle the dynamic input.
Updated 2 months ago1
Finetune on a trained model
Updated 2 months ago
samples num issue during train
Closed 8 months ago2
new code cannot convert to torchscript
Closed 2 months ago2
FIne tuning the model
Updated 2 months ago1
Error running on exported onnx model
Updated 2 months ago1
ONNX and Torch inference outputs are not the same
Updated 2 months ago3
KeyError: 'pytorch-lightning_version'
Updated 3 months ago4
Simple digits training loss not converging
Updated 3 months ago
How could I get the pretrain model?I just can not find it around here,could you offer some specific link to it?
Updated 4 months ago1
Changing dec_depth results in an error during inference.
Closed 4 months ago1
How do I adjust the parameters to train longer texts?
Closed 5 months ago1
How to convert ckpt model to pth or pt model ?
Updated 6 months ago
ONNX inference issue
Updated 7 months ago3
Load Pretrained Parseq Model via Torch Hub
Closed 9 months ago4
which pretrained model is used in gradio demo when selecting parseq model
Updated 7 months ago
ArT and Uber text dataset
Closed 8 months ago1
Hello, I have a question in the paper. In the Input Context in the decoder of Fig 3, if a character line is entered during training, is this part removed during testing?
Closed 8 months ago1
After training
Closed 8 months ago
how to get 1811 images of the 2077 complete picture in IC2015
Closed 8 months ago2
Could you please provide a detailed explanation of PARseq architecture
Closed 8 months ago4
RuntimeError in tune.py: Deprecated Error for on_init_start Callback Hook
Closed 8 months ago1
Support for mismatched key_padding_mask and attn_mask is deprecated
Closed 8 months ago
Parseq with PyTorch >= 2.0
Closed 8 months ago2
This works with cu118, torch==2.1.1 ?
Closed 8 months ago1
How to set the tune.yaml, to find the best lr at 4 nvidia gpu devices?
Closed 9 months ago1
Install Error
Closed 9 months ago1
Visualizaition of Attention Maps
Closed 10 months ago1
Thinner pre-Trained Model of tiny version
Closed 10 months ago1
Validation failed to load data during training
Closed a year ago3
Idea: Integrate with PaddleOCR
Closed a year ago2
Can this model only recognize single-line English text?
Closed a year ago1
After Tune.py, How to set the base lr?
Closed a year ago1
During the Orientation Robustness Benchmark test, an error occurs when applying rotation operations.
Updated a year ago2
Pretrained on MJ, ST
Closed a year ago1
len(confidence vector) > len(pred_string) ?
Closed a year ago2
Training on handwriting recognition --- any major change compared to STR setup?
Closed a year ago2
Retrieving character segmentation results
Closed a year ago1
Finetuning
Updated a year ago
openvino is missing train_1.zip and train_2.zip
Closed a year ago1
Retraining, but accuracy is too low on unknown data.
Updated a year ago
How to deploy the onnx with C++ ?
Updated a year ago
Audio Reference axis misalignment
Closed a year ago1
[enhancement] Add beam-search for generating topN decoded results.
Closed a year ago1
Can I train these models with Pytorch 1.9?
Closed a year ago1
KeyError: 'pytorch-lightning_version'
Closed a year ago1
Doctr meets parseq
Closed a year ago8
Should self. head be set to len (self. tokenizer), especially when predicting batches, allowing labels of different lengths
Closed a year ago1