model finetune over pre-trained model weights..

[tzktok]

问题描述 / Problem Description

i have the trained detection model...i want to fine tune some more dataset over that weight...how to do that..

tamil@DESKTOP-FLC1RG0:~/PaddleOCR/inference/default_model$ ls
inference.pdiparams  inference.pdiparams.info  inference.pdmodel

this are the above files in my default model folder..i want to a finetune above this model..how to achive that..
@GreatV @ZeyuChen

运行环境 / Runtime Environment

• OS:Ubuntu
• Paddle:addlepaddle-2.6.1
• PaddleOCR:paddleocr 2.7.3

[GreatV]

please refer to: https://github.com/PaddlePaddle/PaddleOCR/blob/main/doc/doc_en/training_en.md

[tzktok]

my trainning comment...

tamil@DESKTOP-FLC1RG0:~/PaddleOCR$ python3 tools/train.py -c configs/det/det_mv3_db.yml -o Global.checkpoints=./en_PP-OCRv3_det_distill_train/best_accuracy

but pretrained weight file is missing in downloaded folder...

[2024/06/25 15:34:04] ppocr WARNING: ./en_PP-OCRv3_det_distill_train/best_accuracy.pdopt is not exists, params of optimizer is not loaded

i want to fine tune over this weight..when i downloaded the that trained model folder..it contains only..

.pdopt file is not availble ? @GreatV

how to fine tune en_PP-OCRv3_det ?.. what is correct config file for this finetune?