The official repo for [CVPR'23] "DeepSolo: Let Transformer Decoder with Explicit Points Solo for Text Spotting" & [ArXiv'23] "DeepSolo++: Let Transformer Decoder with Explicit Points Solo for Multilingual Text Spotting"
Other
242
stars
34
forks
source link
How much GPU VRAM is required to fine-tune DeepSolo? #39
I've tried to finetune DeepSolo with CTW1500 dataset on two systems
i) RTX 3050 4GB
ii) RTX 2080 8GB
In both of them I got the same Runtime error i.e. CUDA out of memory. I've tried to adjust the batch size as well with creating config.yaml but still no improvements I noticed and continued to get the same error.
Please let me know what I can do now?
I've tried to finetune DeepSolo with CTW1500 dataset on two systems i) RTX 3050 4GB ii) RTX 2080 8GB
In both of them I got the same Runtime error i.e. CUDA out of memory. I've tried to adjust the batch size as well with creating config.yaml but still no improvements I noticed and continued to get the same error. Please let me know what I can do now?