gnu: python-keras: Enable parallel tests.
* (python-keras)[phases]: Delete trailing #t.
{check}: Honor TESTS?.  Run the tests in parallel.  Disable PEP8 lint.
			
			
This commit is contained in:
		
							parent
							
								
									92c3d1c3d0
								
							
						
					
					
						commit
						11e7a534f7
					
				
					 1 changed files with 17 additions and 16 deletions
				
			
		|  | @ -2521,13 +2521,12 @@ with image data, text data, and sequence data.") | |||
|              (delete-file "keras/backend/theano_backend.py") | ||||
|              (delete-file "keras/backend/cntk_backend.py") | ||||
|              (delete-file "tests/keras/backend/backend_test.py") | ||||
| 
 | ||||
|              ;; FIXME: This doesn't work because Tensorflow is missing the | ||||
|              ;; coder ops library. | ||||
|              (delete-file "tests/keras/test_callbacks.py") | ||||
|              #t)) | ||||
|              (delete-file "tests/keras/test_callbacks.py"))) | ||||
|          (replace 'check | ||||
|            (lambda _ | ||||
|            (lambda* (#:key tests? #:allow-other-keys) | ||||
|              (when tests? | ||||
|                ;; These tests attempt to download data files from the internet. | ||||
|                (delete-file "tests/integration_tests/test_datasets.py") | ||||
|                (delete-file "tests/integration_tests/imagenet_utils_test.py") | ||||
|  | @ -2535,10 +2534,13 @@ with image data, text data, and sequence data.") | |||
|                (substitute* "tests/keras/engine/test_topology.py" | ||||
|                  (("np.ones\\(\\(3, 2\\)\\)") | ||||
|                   "1.")) | ||||
|              (invoke "python" "-m" "pytest" | ||||
|                (invoke "python" "-m" "pytest" "tests" | ||||
|                        "-p" "no:pep8" | ||||
|                        ;; FIXME: python-build-system lacks PARALLEL-TESTS? | ||||
|                        "-n" (number->string (parallel-job-count)) | ||||
|                        ;; The following test fail only in the build container; | ||||
|                        ;; skip it. | ||||
|                      "-k" "not test_selu")))))) | ||||
|                        "-k" "not test_selu"))))))) | ||||
|     (propagated-inputs | ||||
|      `(("python-h5py" ,python-h5py) | ||||
|        ("python-keras-applications" ,python-keras-applications) | ||||
|  | @ -2565,7 +2567,6 @@ with image data, text data, and sequence data.") | |||
| and capable of running on top of TensorFlow.  It was developed with a focus on | ||||
| enabling fast experimentation.  Use Keras if you need a deep learning library | ||||
| that: | ||||
| 
 | ||||
| @itemize | ||||
| @item Allows for easy and fast prototyping (through user friendliness, | ||||
|   modularity, and extensibility). | ||||
|  |  | |||
		Reference in a new issue