domindominik

@Pooya-Mohammadi-K Q1. In the pipeline.config file I set: fine_tune_checkpoint: "/home/dominik/Downloads/repo/training_demo/pre-trained-models/ssd_resnet50_v1_fpn_1024x1024_coco17_tpu-8/checkpoint/ckpt-0" after starting the model_main_tf2.py script, the system returns the message: tensorflow.python.framework.errors_impl.FailedPreconditionError: /home/dominik/Downloads/repo/training_demo/pre-trained-models/ssd_resnet50_v1_fpn_1024x1024_coco17_tpu-8/checkpoint.tmpb8d6f3e0407948dfa546755fd5e7eb78; Is a directory `python3 model_main_tf2.py --model_dir=/home/dominik/Downloads/repo/training_demo/pre-trained-models/ssd_resnet50_v1_fpn_1024x1024_coco17_tpu-8/ --num_train_steps=10000 --pipeline_config_path=/home/dominik/Downloads/repo/training_demo/pre-trained-models/ssd_resnet50_v1_fpn_1024x1024_coco17_tpu-8/pipeline.config...

@Pooya-Mohammadi-K I look forward to hearing from you @map-mymories I created cats_model directory, after starting the training: `fine_tune_checkpoint: "/home/dominik/Downloads/repo/training_demo/pre-trained-models/ssd_resnet50_v1_fpn_1024x1024_coco17_tpu-8/cats_model/"` I got the following message: `ValueError: Couldn't find 'checkpoint' file or...

@map-mymories After copying the files, I get the following message: `raise ValueError("Couldn't find 'checkpoint' file or checkpoints in " ValueError: Couldn't find 'checkpoint' file or checkpoints in given directory /home/dominik/Downloads/repo/training_demo/pre-trained-models/ssd_resnet50_v1_fpn_1024x1024_coco17_tpu-8/cats_model/`

@Pooya-Mohammadi-K @tombstone @jch1 @pkulzc Do you have any ideas what I should do to train a chosen model? If I need to check something additional, please inform me.

@androuino 1. What is the size of your photos for learning and testing the model? 2. Have you tried using another model? 3. Write something more about the environment