testlere

.woodpecker.yml

@@ -1,32 +1,13 @@
-platform: linux/amd64
-pipeline:
-  first-job:
-    image: busybox
+steps:
+  "train":
+    image: nvcr.io/nvidia/tensorflow:23.10-tf2-py3
     commands:
-      - echo "ci working................. "
-  cpu:
-    image: progrium/stress:latest
-    commands: /usr/bin/stress --cpu 20 --io 1 --vm 2 --vm-bytes 128M --timeout 90s
-  nvidia-test:
-   image: nvidia/cuda:11.6.2-base-ubuntu20.04
-   commands:
-     - nvidia-smi
-   environment:
-     - NVIDIA_VISIBLE_DEVICES=all  
-  gpu:
-    image: oguzpastirmaci/gpu-burn:latest
-    environment:
-      - NVIDIA_VISIBLE_DEVICES=all
+      - echo "starting python script   sd "
+      - python run.py
+  "compress and upload":
+    image: alpine:3
     commands:
-      - cd /app
-      - ./gpu_burn 120
-      - echo "burned. done"
-
-#  2nd:
-#    image: oguzpastirmaci/gpu-burn:latest
-#    environment:
-#      - NVIDIA_VISIBLE_DEVICES=all
-#    commands:
-#      - ./gpu_burn 60
-
+      - apk --no-cache add zip curl
+      - zip mymodel.zip mymodel.keras
+      - curl -F fileUpload=@mymodel.zip https://share.storage.sandbox.iuk.hdm-stuttgart.de/upload
 

README.md

@@ -0,0 +1,4 @@
+# Long Running Trainings
+
+This repository contains an example pipeline for long running training tasks. 
+Detailed information can be found at the official [Sandbox Documentation](https://docs.sandbox.iuk.hdm-stuttgart.de/sandbox/training.html).

run.py

@@ -0,0 +1,89 @@
+import tensorflow as tf
+from tensorflow import keras
+import requests
+import numpy as np
+import os
+# Version Information
+# tensorflow 2.2.0 , Cudnn7.6.5 and Cuda 10.1 , python 3.8
+from keras import backend as K
+K.clear_session()
+
+gpus = tf.config.experimental.list_physical_devices('GPU')
+# if gpus:
+#   try:
+#     tf.config.experimental.set_virtual_device_configuration(gpus[0], [tf.config.experimental.VirtualDeviceConfiguration(memory_limit=6024)])
+#   except RuntimeError as e:
+#     print(e)
+#     os.exit(1)
+
+print(tf.config.experimental.list_physical_devices())
+print("test")
+print(tf.__version__)
+
+print(tf.test.is_built_with_cuda())
+
+(X_train, y_train), (X_test,y_test) = tf.keras.datasets.cifar10.load_data()
+
+(X_train, y_train), (X_test,y_test) = tf.keras.datasets.cifar10.load_data()
+
+print(X_train.shape,y_train.shape)
+
+classes = ["airplane","automobile","bird","cat","deer","dog","frog","horse","ship","truck"]
+
+print(classes[y_train[3][0]])
+
+print("pre processing: scale images")
+
+X_train_scaled = X_train / 255
+X_test_scaled = X_test / 255
+
+y_train_categorical = keras.utils.to_categorical(
+    y_train, num_classes=10, dtype='float32'
+)
+y_test_categorical = keras.utils.to_categorical(
+    y_test, num_classes=10, dtype='float32'
+)
+
+print("model build")
+
+
+def get_model():
+    model = keras.Sequential([
+            keras.layers.Flatten(input_shape=(32,32,3)),
+            keras.layers.Dense(3000, activation='relu'),
+            keras.layers.Dense(1000, activation='relu'),
+            keras.layers.Dense(10, activation='sigmoid')    
+        ])
+
+    model.compile(optimizer='SGD',
+                  loss='categorical_crossentropy',
+                  metrics=['accuracy'])
+    return model
+
+with tf.device('/GPU:0'):
+    model = keras.Sequential([
+            keras.layers.Flatten(input_shape=(32,32,3)),
+            keras.layers.Dense(3000, activation='relu'),
+            keras.layers.Dense(1000, activation='relu'),
+            keras.layers.Dense(10, activation='sigmoid')    
+        ])
+  # g
+    model.compile(optimizer='SGD',
+                  loss='categorical_crossentropy',
+                  metrics=['accuracy'])
+    model.fit(X_train_scaled, y_train_categorical, epochs=25)
+    model.save('mymodel.keras')
+    
+    print("finished training")
+    
+    myurl = 'https://share.storage.sandbox.iuk.hdm-stuttgart.de/upload'
+
+    print("uploading file")
+    files = {
+        'fileUpload':('mymodel.keras',  open('mymodel.keras', 'rb'),'application/octet-stream')
+    }
+
+    response = requests.post(myurl, files=files)
+    print(response,response.text)
+    
+print("done")