-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathcommands.txt
38 lines (26 loc) · 2.36 KB
/
commands.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
python wrapper_create_tfrecords.py --dataset_path "../TextRecognitionDataGenerator/out/sample_en_all_fonts_10_per_font/" --prefix "train" --output_dir "./out/tfrecords_sample_en_all_fonts_10_per_font" --threads 1 --shards 1
python wrapper_create_tfrecords.py --dataset_path "../TextRecognitionDataGenerator/out/en_all_fonts_1k_per_font/" --prefix "train" --output_dir "./out/tfrecords_en_all_fonts_1k_per_font" --threads 12 --shards 12
----------------------------------------------------------
50k -- v1 conversion --training data
python wrapper_create_tfrecords.py --dataset_path "../TextRecognitionDataGenerator/out/en_all_fonts_50k_per_font/" --prefix "train" --output_dir "./out/tfrecords_en_all_fonts_50k_per_font" --threads 12 --shards 240
-------------------------------------------------------------
Validation set data -- 6k random text taken from 450k dict and rendered it
python wrapper_create_tfrecords.py --dataset_path "../TextRecognitionDataGenerator/out/en_all_fonts_6k_per_font-random-from-en-dict/" --prefix "val" --output_dir "./out/tfrecords_en_all_fonts_6k_per_font-random-from-en-dict-validation" --threads 12 --shards 24
================================================================
Uploading data to cloud:
sudo snap install google-cloud-cli --classic
gcloud auth login
gcloud config set project 'cellular-virtue-375304'
----
# sample
gsutil -m cp -r "out/tfrecords_sample_en_all_fonts_10_per_font" "gs://font_recognition/tfrecords_sample_en_all_fonts_10_per_font"
# validation
gsutil -m cp -r "out/tfrecords_en_all_fonts_6k_per_font-random-from-en-dict-validation" "gs://font_recognition/tfrecords_en_all_fonts_6k_per_font-random-from-en-dict-validation"
# train
gsutil -m cp -r "out/tfrecords_en_all_fonts_50k_per_font" "gs://font_recognition/tfrecords_en_all_fonts_50k_per_font"
#####################################################################################
####################################################################################
language classification solution data
###
python wrapper_create_tfrecords.py --dataset_path "../TextRecognitionDataGenerator/out/language_classification-v1/" --prefix "train" --output_dir "./out/tfrecords_language_classification-v1" --explicit_labels "en" "ar" "hi" --threads 12 --shards 12
gsutil -m cp -r "out/tfrecords_language_classification-v1" "gs://language_classification/tfrecords_language_classification-v1"