fix GpuCrossentropySoftmax1HotWithBiasDx crash.

theano/sandbox/cuda/nnet.py

@@ -188,7 +188,7 @@ class GpuCrossentropySoftmax1HotWithBiasDx (Op):
     def make_node(self, dy, sm, y_idx):
         return Apply(self, [dy, sm, y_idx],[sm.type()])
     def c_code_cache_version(self):
-        return (2,)
+        return (3,)
         #return ()
     def c_code(self, node, nodename, (dnll, sm, y_idx), (dx,), sub):
         fail = sub['fail']
@@ -229,7 +229,7 @@ class GpuCrossentropySoftmax1HotWithBiasDx (Op):
             kCrossEntropySoftmax1HotWithBiasDx_%(nodename)s
                 <<<
                     CudaNdarray_HOST_DIMS(%(dx)s)[0],
-                    CudaNdarray_HOST_DIMS(%(dx)s)[1]
+                    std::min(CudaNdarray_HOST_DIMS(%(dx)s)[1],256)
                 >>>(
                         CudaNdarray_HOST_DIMS(%(dx)s)[0],
                         CudaNdarray_HOST_DIMS(%(dx)s)[1], 

theano/sandbox/cuda/tests/test_nnet.py

+import theano, numpy
+import theano.tensor as T
+
+# Skip test if cuda_ndarray is not available.
+from nose.plugins.skip import SkipTest
+import theano.sandbox.cuda as cuda
+if cuda.cuda_available == False:
+    raise SkipTest('Optional package cuda disabled')
+
+if theano.config.mode=='FAST_COMPILE':
+    mode_with_gpu = theano.compile.mode.get_mode('FAST_RUN').including('gpu')
+    mode_without_gpu = theano.compile.mode.get_mode('FAST_RUN').excluding('gpu')
+else:
+    mode_with_gpu = theano.compile.mode.get_default_mode().including('gpu')
+    mode_without_gpu = theano.compile.mode.get_default_mode().excluding('gpu')
+
+def test_GpuCrossentropySoftmax1HotWithBiasDx():
+    """
+    This is basic test for GpuCrossentropySoftmaxArgmax1HotWithBias and GpuCrossentropySoftmax1HotWithBiasDx
+    """
+
+    n_in = 1000
+    batch_size = 4097
+    n_out = 1250
+
+    n_in = 4098
+    n_out = 4099
+
+    x = T.fmatrix('x')
+    y = T.lvector('y')
+
+
+    b = T.fvector()
+    W = T.fmatrix()
+
+    p_y_given_x = T.nnet.softmax(T.dot(x,W)+b)
+    y_pred = T.argmax(p_y_given_x)
+    loss = -T.mean(T.log(p_y_given_x)[T.arange(y.shape[0]), y])
+    dW = T.grad(loss,W)
+    classify = theano.function( inputs = [x,y,b,W], outputs = [loss,y_pred,dW],
+                                mode = mode_without_gpu)
+    classify_gpu = theano.function( inputs = [x,y,b,W], outputs = [loss,y_pred,dW],
+                                    mode = mode_with_gpu)
+    
+    xx = numpy.asarray(numpy.random.rand(batch_size,n_in),dtype=numpy.float32)
+    yy = numpy.ones((batch_size,),dtype='float32')
+    b_values = numpy.zeros((n_out,),dtype='float32')
+    W_values = numpy.asarray(numpy.random.rand(n_in,n_out),dtype='float32')
+    
+
+    assert any([isinstance(node.op,T.nnet.CrossentropySoftmaxArgmax1HotWithBias) for node in classify.maker.env.toposort()])
+    assert any([isinstance(node.op,T.nnet.CrossentropySoftmax1HotWithBiasDx) for node in classify.maker.env.toposort()])
+    assert any([isinstance(node.op,cuda.nnet.GpuCrossentropySoftmaxArgmax1HotWithBias) for node in classify_gpu.maker.env.toposort()])
+    assert any([isinstance(node.op,cuda.nnet.GpuCrossentropySoftmax1HotWithBiasDx) for node in classify_gpu.maker.env.toposort()])
+
+    assert numpy.allclose(classify(xx,yy,b_values,W_values)[0],classify_gpu(xx,yy,b_values,W_values)[0])
+    assert numpy.allclose(classify(xx,yy,b_values,W_values)[1],classify_gpu(xx,yy,b_values,W_values)[1])
+    assert numpy.allclose(classify(xx,yy,b_values,W_values)[2],classify_gpu(xx,yy,b_values,W_values)[2],atol=2e-6)
+
+