sagor999 / poker_ml Goto Github PK

Detected change
2023-11-17 21:11:45.936936: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
2023-11-17 21:11:45.936965: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
2023-11-17 21:11:45.937662: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
Touched trigger. Waiting for ML
2023-11-17 21:11:47.150733: W tensorflow/core/common_runtime/gpu/gpu_device.cc:2256] Cannot dlopen some GPU libraries. Please make sure the missing libraries mentioned above are installed properly if you would like to use GPU. Follow the guide at https://www.tensorflow.org/install/gpu for how to download and setup the required libraries for your platform.
Skipping registering GPU devices...
ML output:  
OCR:  Total pot: 208.94
Main pot: 200.94

----------------------------------

Above is the output of running make run.

Never reads the cards sent to it just gets stuck like this forever.
Anyone can point me in right direction?

Here is how the input looks.

Nvidia gpu have the drivers verified, verified Tensorflow[and-cude] are installed

In poker_ml/expected_value/src/main.rs

There are paths to files that don't exist with one such example: combinations.bin in
let combinations_path: String = "/home/pavel/nvme/GitHub/poker_ml/expected_value/data/combinations.bin".to_string();

When running ./poker_ev there is then an error that says cannot find file and directory

Maybe these were files that you meant to include in your project?

I'm not too familiar with docker, but it seems there are hardcoded paths here... Any tips on how I would change it to my machine?

1. Is tf-gpu a directory where tensorflow is installed?

2. Also what is $$(pwd)/data:/data -v $$(pwd):/src

3. I have nvidia gpu with nvidia graphics driver installed, but do I need anything to make sure --runtime=nvidia works?

When I do make run in the root folder, I get this issue:

python3 orchestrator/main.py Traceback (most recent call last): File "orchestrator/main.py", line 1, in <module> from PIL import Image, ImageGrab, ImageChops, ImageOps File "/usr/lib/python3/dist-packages/PIL/ImageGrab.py", line 26, in <module> raise ImportError("ImageGrab is macOS and Windows only") ImportError: ImageGrab is macOS and Windows only make: *** [Makefile:5: run] Error 1

Is this meant to be ran in Windows/macOS terminals after creating all the files in Linux?

Hey, it seems that you forgot to say that there is another hard coded path besides the ones at the top of main( ). It is the first line of fn generate_ml_data( ) line 1326.