Skip to content

Instantly share code, notes, and snippets.

View csullivan's full-sized avatar

Chris Sullivan csullivan

72 followers · 21 following
  • NVIDIA
  • Portland
View GitHub Profile
@csullivan
csullivan / profile_small.txt
Created November 12, 2018 18:03
INFO:root:start with arguments Namespace(batch_size=64, benchmark=0, brightness=0, contrast=0, data_nthreads=4, data_train='/dataset/mxnet_imagenet/train.rec', data_train_idx='', data_val='/dataset/mxnet_imagenet/val.rec', data_val_idx='', disp_batches=20, dtype='float32', fill_value=127, gc_threshold=0.5, gc_type='none', gpus='0', image_shape='3,224,224', initializer='default', is_nnp=False, kv_store='device', load_epoch=None, loss='', lr=0.1, lr_factor=0.1, lr_step_epochs='30,60', macrobatch_size=0, max_crop_size=-1, max_random_area=1, max_random_aspect_ratio=0, max_random_h=0, max_random_l=0, max_random_rotate_angle=0, max_random_s=0, max_random_scale=1, max_random_shear_ratio=0, min_crop_size=-1, min_random_area=1, min_random_aspect_ratio=None, min_random_scale=1, model_prefix=None, mom=0.9, monitor=0, network='resnet', num_classes=1000, num_epochs=80, num_examples=1281167, num_layers=50, optimizer='sgd', pad_size=0, pca_noise=0, profile_server_suffix='', profile_worker_suffix='', random_crop=0, random_mi
@csullivan
csullivan / print_tensor_vals.diff
Last active October 1, 2018 22:44
diff --git a/src/ngraph/runtime/cpu/cpu_external_function.cpp b/src/ngraph/runtime/cpu/cpu_external_function.cpp
index bc30f4d1..4fbd85aa 100644
--- a/src/ngraph/runtime/cpu/cpu_external_function.cpp
+++ b/src/ngraph/runtime/cpu/cpu_external_function.cpp
@@ -22,6 +22,7 @@
#include <typeindex>
#include <typeinfo>
#include <unordered_map>
+#include <algorithm>
@csullivan
csullivan / rnn_updates.diff
Created September 25, 2018 16:27
diff --git a/src/ngraph/runtime/gpu/CMakeLists.txt b/src/ngraph/runtime/gpu/CMakeLists.txt
index 04d96608..aaad210c 100644
--- a/src/ngraph/runtime/gpu/CMakeLists.txt
+++ b/src/ngraph/runtime/gpu/CMakeLists.txt
@@ -42,7 +42,6 @@ set(SRC
pass/tensor_memory_reservation.cpp
gpu_kernel_args.cpp
pass/gpu_rnn_fusion.cpp
- op/lstm.cpp
op/rnn.cpp
@csullivan
csullivan / raw_output.txt
Created September 24, 2018 18:27
ALLOC: Parameter_0
ALLOC: Parameter_1
ALLOC: Parameter_2
ALLOC: Parameter_3
ALLOC: Parameter_4
ALLOC: Parameter_5
ALLOC: Parameter_6
ALLOC: Parameter_7
ALLOC: Parameter_8
ALLOC: Parameter_9
@csullivan
csullivan / bn.diff
Created September 10, 2018 23:00
diff --git a/src/ngraph/runtime/gpu/cudnn_emitter.cpp b/src/ngraph/runtime/gpu/cudnn_emitter.cpp
index bfdc117a..bf0dece0 100644
--- a/src/ngraph/runtime/gpu/cudnn_emitter.cpp
+++ b/src/ngraph/runtime/gpu/cudnn_emitter.cpp
@@ -1229,14 +1229,16 @@ size_t runtime::gpu::CUDNNEmitter::build_batchnorm(const cudnnBatchNormMode_t& b
const Prop& direction,
const Shape& tensor_shape,
const Shape& param_shape,
- double epsilon)
+ double epsilon,
@csullivan
csullivan / lattice_elim1.cpp
Created August 28, 2018 21:13
// Generated by the nGraph GPU backend
#include <cublas_v2.h>
#include <cuda.h>
#include <cuda_runtime.h>
#include <cudnn.h>
#include "ngraph/descriptor/input.hpp"
#include "ngraph/descriptor/layout/dense_tensor_view_layout.hpp"
#include "ngraph/descriptor/output.hpp"
@csullivan
csullivan / Function_0_codegen.cpp
Created August 28, 2018 21:10
// Generated by the nGraph GPU backend
#include <cublas_v2.h>
#include <cuda.h>
#include <cuda_runtime.h>
#include <cudnn.h>
#include "ngraph/descriptor/input.hpp"
#include "ngraph/descriptor/layout/dense_tensor_view_layout.hpp"
#include "ngraph/descriptor/output.hpp"
@csullivan
csullivan / correct_1.dat
Created August 27, 2018 20:50
[ RUN ] gpu_fusion.fuse_lstm_cells
[DEBUG] 2018-08-27T20:45:02z graph_rewrite.cpp 35 Running matcher Unnamed(Divide_77) on Parameter_109
[DEBUG] 2018-08-27T20:45:02z matcher.cpp 276 [MATCHER] Starting match pattern = Divide_77 , graph_node = Parameter_109
[DEBUG] 2018-08-27T20:45:02z matcher.cpp 150 [MATCHER] in match_node : pattern = Divide_77 matched Parameter_109
[DEBUG] 2018-08-27T20:45:02z graph_rewrite.cpp 35 Running matcher Unnamed(Multiply_107) on Parameter_109
[DEBUG] 2018-08-27T20:45:02z matcher.cpp 276 [MATCHER] Starting match pattern = Multiply_107 , graph_node = Parameter_109
[DEBUG] 2018-08-27T20:45:02z matcher.cpp 150 [MATCHER] in match_node : pattern = Multiply_107 matched Parameter_109
[DEBUG] 2018-08-27T20:45:02z graph_rewrite.cpp 35 Running matcher Unnamed(Divide_77) on Parameter_110
[DEBUG] 2018-08-27T20:45:02z matcher.cpp 276 [MATCHER] Starting match pattern = Divide_77 , graph_node = Parameter_110
[DEBUG] 2018-08-27T20:45:02z matcher.cpp 150 [MATCHER] in match_node : pattern = Divid
@csullivan
csullivan / incorrect_1.dat
Created August 27, 2018 20:50
[ RUN ] gpu_fusion.fuse_lstm_cells
[DEBUG] 2018-08-27T20:46:02z graph_rewrite.cpp 35 Running matcher Unnamed(Divide_77) on Parameter_109
[DEBUG] 2018-08-27T20:46:02z matcher.cpp 276 [MATCHER] Starting match pattern = Divide_77 , graph_node = Parameter_109
[DEBUG] 2018-08-27T20:46:02z matcher.cpp 150 [MATCHER] in match_node : pattern = Divide_77 matched Parameter_109
[DEBUG] 2018-08-27T20:46:02z graph_rewrite.cpp 35 Running matcher Unnamed(Multiply_107) on Parameter_109
[DEBUG] 2018-08-27T20:46:02z matcher.cpp 276 [MATCHER] Starting match pattern = Multiply_107 , graph_node = Parameter_109
[DEBUG] 2018-08-27T20:46:02z matcher.cpp 150 [MATCHER] in match_node : pattern = Multiply_107 matched Parameter_109
[DEBUG] 2018-08-27T20:46:02z graph_rewrite.cpp 35 Running matcher Unnamed(Divide_77) on Parameter_110
[DEBUG] 2018-08-27T20:46:02z matcher.cpp 276 [MATCHER] Starting match pattern = Divide_77 , graph_node = Parameter_110
[DEBUG] 2018-08-27T20:46:02z matcher.cpp 150 [MATCHER] in match_node : pattern = Divid
@csullivan
csullivan / install_error.dat
Created August 8, 2018 18:19
chrissul@nervana-titanxp22 /localdisk/sullivan/projects $ nvidia-smi
Wed Aug 8 08:34:41 2018
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 396.26 Driver Version: 396.26 |
|-------------------------------+----------------------+----------------------+
| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |
|===============================+======================+======================|
| 0 TITAN Xp Off | 00000000:06:00.0 Off | N/A |
| 23% 28C P8 17W / 250W | 588MiB / 12196MiB | 0% Default |