Nvidia TX1上的TensorFlow

Question

任何人都在使用Nvidia Tegra X1进行张量流动？

我发现一些消息来源表明它可能存在于TK1上,或者TX1上存在严重的黑客攻击/错误,但还没有确定的配方.

• http://cudamusing.blogspot.de/2015/11/building-tensorflow-for-jetson-tk1.html
• https://github.com/tensorflow/tensorflow/issues/851

我正在使用Jetson 2.3安装但尚未使用它 - 任何提示最受欢迎.

Answer 1

TensorFlow R0.9使用Bazel 0.2.1,CUDA 8.0,CUDNN5.1,L4T24.2和新安装的JetPack 2.3在TX1上运行.我使用BN,Sigmoid,ReLU等基本的MLP,Conv和LSTM网络进行了测试,但没有任何错误.我删除了sparse_matmul_op,但认为编译应该完全可操作.其中许多步骤直接来自MaxCuda的优秀博客,非常感谢他们的提供.

计划继续锤击R0.10/R0.11(gRPC二进制现在正在阻止Bazel 0.3.0),但在此之前我想发布了R0.9公式.如下:

首先得到java

sudo add-apt-repository ppa:webupd8team/java
sudo apt-get update
sudo apt-get install oracle-java8-installer

安装其他一些deps

sudo apt-get install git zip unzip autoconf automake libtool curl zlib1g-dev maven swig

需要自己构建protobuf 3.0.0-beta-2 jar

git clone https://github.com/google/protobuf.git
cd protobuf
# autogen.sh downloads broken gmock.zip in d5fb408d
git checkout master
./autogen.sh
git checkout d5fb408d
./configure --prefix=/usr
make -j 4
sudo make install
cd java
mvn package

得到bazel.我们想要版本0.2.1,它不需要gRPC二进制文件,不像0.3.0我无法构建(可能很快!)

git clone https://github.com/bazelbuild/bazel.git
cd bazel
git checkout 0.2.1
cp /usr/bin/protoc third_party/protobuf/protoc-linux-arm32.exe
cp ../protobuf/java/target/protobuf-java-3.0.0-beta-2.jar third_party/protobuf/protobuf-java-3.0.0-beta-1.jar

需要编辑bazel文件以将aarch64识别为ARM

--- a/src/main/java/com/google/devtools/build/lib/util/CPU.java
+++ b/src/main/java/com/google/devtools/build/lib/util/CPU.java
@@ -25,7 +25,7 @@ import java.util.Set;
 public enum CPU {
   X86_32("x86_32", ImmutableSet.of("i386", "i486", "i586", "i686", "i786", "x86")),
   X86_64("x86_64", ImmutableSet.of("amd64", "x86_64", "x64")),
-  ARM("arm", ImmutableSet.of("arm", "armv7l")),
+  ARM("arm", ImmutableSet.of("arm", "armv7l", "aarch64")),
   UNKNOWN("unknown", ImmutableSet.<String>of());

现在编译

./compile.sh

并安装

sudo cp output/bazel /usr/local/bin

得到张量流R0.9.高于R0.9需要Bazel 0.3.0,由于gRPC问题,我还没有想出如何构建.

git clone -b r0.9 https://github.com/tensorflow/tensorflow.git

建立一次.它会失败,但现在你有了bazel .cache目录,你可以放置更新的config.guess和config.sub文件,它们将显示你正在运行的架构

./configure
bazel build -c opt --config=cuda //tensorflow/tools/pip_package:build_pip_package

cd ~
wget -O config.guess 'http://git.savannah.gnu.org/gitweb/?p=config.git;a=blob_plain;f=config.guess;hb=HEAD'
wget -O config.sub 'http://git.savannah.gnu.org/gitweb/?p=config.git;a=blob_plain;f=config.sub;hb=HEAD'

# below are commands I ran, yours will vary depending on .cache details. `find` is your friend
cp config.guess ./.cache/bazel/_bazel_socialh/742c01ff0765b098544431b60b1eed9f/external/farmhash_archive/farmhash-34c13ddfab0e35422f4c3979f360635a8c050260/config.guess
cp config.sub ./.cache/bazel/_bazel_socialh/742c01ff0765b098544431b60b1eed9f/external/farmhash_archive/farmhash-34c13ddfab0e35422f4c3979f360635a8c050260/config.sub

sparse_matmul_op有几个错误,我采取了懦弱的路线并从构建中删除

--- a/tensorflow/core/kernels/BUILD
+++ b/tensorflow/core/kernels/BUILD
@@ -985,7 +985,7 @@ tf_kernel_libraries(
         "reduction_ops",
         "segment_reduction_ops",
         "sequence_ops",
-        "sparse_matmul_op",
+        #DC "sparse_matmul_op",
     ],
     deps = [
         ":bounds_check",

--- a/tensorflow/python/BUILD
+++ b/tensorflow/python/BUILD
@@ -1110,7 +1110,7 @@ medium_kernel_test_list = glob([
     "kernel_tests/seq2seq_test.py",
     "kernel_tests/slice_op_test.py",
     "kernel_tests/sparse_ops_test.py",
-    "kernel_tests/sparse_matmul_op_test.py",
+    #DC "kernel_tests/sparse_matmul_op_test.py",
     "kernel_tests/sparse_tensor_dense_matmul_op_test.py",
 ])

TX1无法在cwise_op_gpu_select.cu.cc中执行花哨的构造函数

--- a/tensorflow/core/kernels/cwise_op_gpu_select.cu.cc
+++ b/tensorflow/core/kernels/cwise_op_gpu_select.cu.cc
@@ -43,8 +43,14 @@ struct BatchSelectFunctor<GPUDevice, T> {
     const int all_but_batch = then_flat_outer_dims.dimension(1);

 #if !defined(EIGEN_HAS_INDEX_LIST)
-    Eigen::array<int, 2> broadcast_dims{{ 1, all_but_batch }};
-    Eigen::Tensor<int, 2>::Dimensions reshape_dims{{ batch, 1 }};
+    //DC Eigen::array<int, 2> broadcast_dims{{ 1, all_but_batch }};
+    Eigen::array<int, 2> broadcast_dims;
+    broadcast_dims[0] = 1;
+    broadcast_dims[1] = all_but_batch;
+    //DC Eigen::Tensor<int, 2>::Dimensions reshape_dims{{ batch, 1 }};
+    Eigen::Tensor<int, 2>::Dimensions reshape_dims;
+    reshape_dims[0] = batch;
+    reshape_dims[1] = 1;
 #else
     Eigen::IndexList<Eigen::type2index<1>, int> broadcast_dims;
     broadcast_dims.set(1, all_but_batch);

与sparse_tensor_dense_matmul_op_gpu.cu.cc相同

--- a/tensorflow/core/kernels/sparse_tensor_dense_matmul_op_gpu.cu.cc
+++ b/tensorflow/core/kernels/sparse_tensor_dense_matmul_op_gpu.cu.cc
@@ -104,9 +104,17 @@ struct SparseTensorDenseMatMulFunctor<GPUDevice, T, ADJ_A, ADJ_B> {
     int n = (ADJ_B) ? b.dimension(0) : b.dimension(1);

 #if !defined(EIGEN_HAS_INDEX_LIST)
-    Eigen::Tensor<int, 2>::Dimensions matrix_1_by_nnz{{ 1, nnz }};
-    Eigen::array<int, 2> n_by_1{{ n, 1 }};
-    Eigen::array<int, 1> reduce_on_rows{{ 0 }};
+    //DC Eigen::Tensor<int, 2>::Dimensions matrix_1_by_nnz{{ 1, nnz }};
+    Eigen::Tensor<int, 2>::Dimensions matrix_1_by_nnz;
+    matrix_1_by_nnz[0] = 1;
+    matrix_1_by_nnz[1] = nnz;
+    //DC Eigen::array<int, 2> n_by_1{{ n, 1 }};
+    Eigen::array<int, 2> n_by_1;
+    n_by_1[0] = n;
+    n_by_1[1] = 1;
+    //DC Eigen::array<int, 1> reduce_on_rows{{ 0 }};
+    Eigen::array<int, 1> reduce_on_rows;
+    reduce_on_rows[0] = 0;
 #else
     Eigen::IndexList<Eigen::type2index<1>, int> matrix_1_by_nnz;
     matrix_1_by_nnz.set(1, nnz);

使用CUDA 8.0运行需要FP16的新宏.非常感谢Kashif/Mrry指出修复!

--- a/tensorflow/stream_executor/cuda/cuda_blas.cc
+++ b/tensorflow/stream_executor/cuda/cuda_blas.cc
@@ -25,6 +25,12 @@ limitations under the License.
 #define EIGEN_HAS_CUDA_FP16
 #endif

+#if CUDA_VERSION >= 8000
+#define SE_CUDA_DATA_HALF CUDA_R_16F
+#else
+#define SE_CUDA_DATA_HALF CUBLAS_DATA_HALF
+#endif
+
 #include "tensorflow/stream_executor/cuda/cuda_blas.h"

 #include <dlfcn.h>
@@ -1680,10 +1686,10 @@ bool CUDABlas::DoBlasGemm(
   return DoBlasInternal(
       dynload::cublasSgemmEx, stream, true /* = pointer_mode_host */,
       CUDABlasTranspose(transa), CUDABlasTranspose(transb), m, n, k, &alpha,
-      CUDAMemory(a), CUBLAS_DATA_HALF, lda,
-      CUDAMemory(b), CUBLAS_DATA_HALF, ldb,
+      CUDAMemory(a), SE_CUDA_DATA_HALF, lda,
+      CUDAMemory(b), SE_CUDA_DATA_HALF, ldb,
       &beta,
-      CUDAMemoryMutable(c), CUBLAS_DATA_HALF, ldc);
+      CUDAMemoryMutable(c), SE_CUDA_DATA_HALF, ldc);
 #else
   LOG(ERROR) << "fp16 sgemm is not implemented in this cuBLAS version "
              << "(need at least CUDA 7.5)";

最后,ARM没有NUMA节点,因此需要添加或者在启动tf.Session()时会立即崩溃

--- a/tensorflow/stream_executor/cuda/cuda_gpu_executor.cc
+++ b/tensorflow/stream_executor/cuda/cuda_gpu_executor.cc
@@ -888,6 +888,9 @@ CudaContext* CUDAExecutor::cuda_context() { return context_; }
 // For anything more complicated/prod-focused than this, you'll likely want to
 // turn to gsys' topology modeling.
 static int TryToReadNumaNode(const string &pci_bus_id, int device_ordinal) {
+  // DC - make this clever later. ARM has no NUMA node, just return 0
+  LOG(INFO) << "ARM has no NUMA node, hardcoding to return zero";
+  return 0;
 #if defined(__APPLE__)
   LOG(INFO) << "OS X does not support NUMA - returning NUMA node zero";
   return 0;

完成这些更改后,构建并安装!希望这对一些人有用.