bmw-innovationlab / bmw-yolov4-training-automation Goto Github PK

Hi @nourazzii

I was trying to start training using the sample dataset, and I encountered the following problem:
When I run main.py it ends right after the logging:

2020-06-04 12:49:00,997 Events Log : INFO Starting YOLO training

As you can see in the image below, there is no errors showed in the screen, but the program just end right after the above log is showed.

I followed the steps provided and also looked into the dataset folder and all the images + txt are present. Also, the train.txt file contains all the images.

I think it is a problem with no finding the right path to images, therefore finishing right after starting.

I hope you throw some light,

Thanks a lot!

getting above mentioned error after executing this line
sudo docker build -f docker/Dockerfile -t darknet_yolo_gpu:1 --build-arg GPU=1 --build-arg CUDNN=1 --build-arg CUDNN_HALF=0 --build-arg OPENCV=1 --build-arg DOWNLOAD_ALL=1 .
Any ideas? Ive gone through few solutions given online but no result

long@long:$ cd /home/long/Desktop/working/ai/BMW-YOLOv4-Training-Automation
long@long:/Desktop/working/ai/BMW-YOLOv4-Training-Automation$ ./run_docker_linux_cpu.sh
Enter your dataset's absolute path (folder containing images', labels' folders and configuration file):
Choose a name for the docker container:
Error: Configuration file not found in the provided path

please help me on this info, what info i need pass to here ? thanks

is it possible to train yolov4 tiny weights? I tried to train on yolo-tiny weights, specified their path in the JSON config file.
like so:

    "model": {
        "framework": "darknet",
        "model_name": "yolov4",
	    "custom_weights": {
            "enable": true,
            "name": "yolov4-tiny.weights"
        },

but the output files were:

the initial weights are 24.3 Mb (tiny weights) and all others are 256 Mb (full wights).
am I doing something wrong here?

Are there any ways to do inference/predictions using the latest weight after the model is trained?
I am able to do predictions during the training process using the Custom API at port 8099. However, the port is also closed after the training is finished.
Thanks!

When I execute ./run_docker_linux_gpu.sh

After specifing abs path for training data and container name I get the following error:

docker: Error response from daemon: Unknown runtime specified nvidia.

any pointers would be greatly appreciated

Hi there!
I got an error during the training process for some images

error load image xxx.jpg Error in load_data detection() - OpenCV
and after a short while, container stops working.
I delete those images mentioned in the log file, but the container stops again after a short while (3 to 8 epochs) without any error in log files!
Please help me if I do wrong in anything.
Thanks

hi,
how can I start new training with the same train and test images of previous training? same train.txt and test.txt files .

Hi,

is there a way to consult the graphs of the AlexeyAB WebUI and the TensorboardX after training?

For now, there are only the files: yolo_events.log and yolo_events.log.1 which hold evaluation metrics.

Thank you in advance!

hi,
is there a way to save the mAP and loss graphs from tensorboard at the end of the training?
thanks.

is it possible to choose the image/set of images that the "validate" option in the web interface uses? or it's just randomly picked?

Hi I have the same issue as others in this issue history.
I have tried solution to set DOWNLOAD_ALL=1 in dockerfile but not works for me.
I have yolov4.weights in the right folder under config/darknet/yolov4_default_weights/
Any help? Thank you. Robert

Hello I have a problem when the Training starts.

i renamed the Folder to "training" i also downloaded the yolov3 weights again the file exist but the Error is still there.

thank you for every help.

__

I want to see the tensorboard after the training has finished, so I'm trying to use:

tensorboard --logdir path/to/logs
where is the "path/to/logs"?
the closest thing I saw was the "yolo_events.log" but I couldn't get this to work.

Hi everyone,

very nice container setup you've done there! :)

After I finished a training run, I can't seem to find a way to continue the training, i.e. to start a new training with my pre-trained weights. I could probably create a hack by modifying the container, but I wonder if there is a way how you meant it to be.

Alex

hello, could you please delete some files in release to make tar.gz small

hi, i have run the auto training, and the output is
2019-12-26 16:48:50,532 Events Log : INFO Running web_ui on port 8190
2019-12-26 16:48:50,532 Events Log : INFO Running YOLO API on port 9911
2019-12-26 16:48:50,537 Events Log : INFO Running Tensorboard on port 6006

use curl "http://127.0.0.1:9911/summary" got {"success":true,"start_time":"201912261647","data":{"current_training_iteration":"1179","total_loss":"0.750510","average_loss_error":"0.788038","current_learning_rate":"0.001000","total_time":"1.202105","number_of_images":"24759","mAP":"0.990001"}}
howerver, curl "http://127.0.0.1:6006" and curl "http://127.0.0.1:8190" results:
curl: (56) Recv failure: Connection reset by peer
what's the problem?

Hi all,

I am trying to run a classifier with 20 object types and some 100k images as training data. I tried it three times and each time it stops at step 10k and does not continue any more.
In the log file it says

 (next mAP calculation at 10300 iterations) 
 Last accuracy [email protected] = 10.97 %, best = 11.05 % 
10115: 0.903671, 0.621472 avg loss, 0.001000 rate, 4.427342 seconds, 323680 images
MJPEG-stream sent. 
Loaded: 0.000051 seconds

It does not continue anymore. Do I need to set something in the model?

Cheers,
Andi

Just to clarify - Is the mAP result we get from YOLO API evaluated on the test set or it is calculated on the training set?

" {"success":true,"start_time":"201912261647","data":{"current_training_iteration":"1179","total_loss":"0.750510","average_loss_error":"0.788038","current_learning_rate":"0.001000","total_time":"1.202105","number_of_images":"24759","mAP":"0.990001"}}"

Thanks!

Hi, I am facing this issue

Though we can see that the file exists there: