[incubator-mxnet] branch numpy updated: [WIP][numpy] Fix for D2L Chapters 2/3/4 (#15139)
This is an automated email from the ASF dual-hosted git repository.
reminisce pushed a commit to branch numpy
in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git
The following commit(s) were added to refs/heads/numpy by this push:
new 517451b [WIP][numpy] Fix for D2L Chapters 2/3/4 (#15139)
517451b is described below
commit 517451bf68e9824f32640727021e8ae438d00b29
Author: reminisce
AuthorDate: Tue Jun 4 22:55:10 2019 -0700
[WIP][numpy] Fix for D2L Chapters 2/3/4 (#15139)
* Fix
* Fix linear regression gluon
* More fix
* Fix pylint
* Fix for chapter 4
* Add np.add mul div mod pow sub and shuffle
* Fix model selection, underfitting, overfitting
* Fix weight decay
* Fix dropout
* Fix
* Fix chapter 4
---
python/mxnet/gluon/data/dataloader.py | 20 +-
python/mxnet/gluon/data/vision/transforms.py | 6 +-
python/mxnet/gluon/loss.py | 26 +-
python/mxnet/gluon/nn/activations.py | 5 +-
python/mxnet/gluon/nn/basic_layers.py | 13 +-
python/mxnet/gluon/utils.py| 50 ++--
python/mxnet/ndarray/numpy/_op.py | 199 ++-
python/mxnet/ndarray/register.py | 8 +-
python/mxnet/numpy/multiarray.py | 326 ++---
python/mxnet/numpy_extension/__init__.py | 5 +-
python/mxnet/optimizer/optimizer.py| 10 +-
python/mxnet/symbol/numpy/_symbol.py | 194 ---
python/mxnet/symbol/register.py| 8 +-
python/mxnet/symbol/symbol.py | 4 +
python/mxnet/util.py | 38 ++-
src/operator/nn/activation.cc | 1 +
src/operator/nn/batch_norm.cc | 1 +
src/operator/nn/convolution.cc | 1 +
src/operator/nn/fully_connected.cc | 1 +
src/operator/nn/pooling.cc | 3 +-
src/operator/random/shuffle_op.cc | 1 +
src/operator/tensor/elemwise_unary_op_basic.cc | 1 +
src/operator/tensor/matrix_op.cc | 1 +
tests/python/unittest/test_numpy_gluon.py | 6 +-
24 files changed, 696 insertions(+), 232 deletions(-)
diff --git a/python/mxnet/gluon/data/dataloader.py
b/python/mxnet/gluon/data/dataloader.py
index 934f2d5..a1d6513 100644
--- a/python/mxnet/gluon/data/dataloader.py
+++ b/python/mxnet/gluon/data/dataloader.py
@@ -18,6 +18,7 @@
# coding: utf-8
# pylint: disable=ungrouped-imports
"""Dataset generator."""
+from __future__ import absolute_import
__all__ = ['DataLoader']
import pickle
@@ -37,6 +38,8 @@ except ImportError:
from . import sampler as _sampler
from ... import nd, context
+from ...util import is_np_array
+from ... import numpy as _mx_np #pylint: disable=reimported
if sys.platform == 'darwin' or sys.platform == 'win32':
def rebuild_ndarray(*args):
@@ -127,13 +130,14 @@ class SimpleQueue(multiprocessing.queues.SimpleQueue):
def default_batchify_fn(data):
"""Collate data into batch."""
if isinstance(data[0], nd.NDArray):
-return nd.stack(*data)
+return _mx_np.stack(data) if is_np_array() else nd.stack(*data)
elif isinstance(data[0], tuple):
data = zip(*data)
return [default_batchify_fn(i) for i in data]
else:
data = np.asarray(data)
-return nd.array(data, dtype=data.dtype)
+array_fn = _mx_np.array if is_np_array() else nd.array
+return array_fn(data, dtype=data.dtype)
def default_mp_batchify_fn(data):
@@ -141,20 +145,26 @@ def default_mp_batchify_fn(data):
if isinstance(data[0], nd.NDArray):
out = nd.empty((len(data),) + data[0].shape, dtype=data[0].dtype,
ctx=context.Context('cpu_shared', 0))
-return nd.stack(*data, out=out)
+if is_np_array():
+out = out.as_np_ndarray()
+return _mx_np.stack(data, out=out)
+else:
+return nd.stack(*data, out=out)
elif isinstance(data[0], tuple):
data = zip(*data)
return [default_mp_batchify_fn(i) for i in data]
else:
data = np.asarray(data)
-return nd.array(data, dtype=data.dtype,
+array_fn = _mx_np.array if is_np_array() else nd.array
+return array_fn(data, dtype=data.dtype,
ctx=context.Context('cpu_shared', 0))
def _as_in_context(data, ctx):
"""Move data into new context."""
if isinstance(data, nd.NDArray):
-return data.as_in_context(ctx)
+out = data.as_in_context(ctx)
+return out.as_np_ndarray() if is_np_array() else out
elif isinstance(data, (list, tuple)):
return [_as_in_context(d, ctx) for d in data]
return data
diff --git a/python/mxnet/gluon/data/vision/transforms.py
b/python/mxnet/gluon/data/vision/transforms.py
index dff7f66..d888398 100644
---
[GitHub] [incubator-mxnet] reminisce merged pull request #15139: [WIP][numpy] Fix for D2L Chapters 2/3/4
reminisce merged pull request #15139: [WIP][numpy] Fix for D2L Chapters 2/3/4 URL: https://github.com/apache/incubator-mxnet/pull/15139 This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] kshitij12345 commented on a change in pull request #15120: [bug] fix higher grad log
kshitij12345 commented on a change in pull request #15120: [bug] fix higher grad log URL: https://github.com/apache/incubator-mxnet/pull/15120#discussion_r290587434 ## File path: tests/python/unittest/test_higher_order_grad.py ## @@ -27,52 +27,79 @@ def test_log(): def log(x): return nd.log(x) +def grad_op(x): +return 1/x + def grad_grad_op(x): return -1/(x**2) arrays = random_arrays((2, 2), (2, 3), (4, 5, 2), (3, 1, 4, 5)) for array in arrays: -check_second_order_unary(array, log, grad_grad_op) +check_second_order_unary(array, log, grad_op, grad_grad_op) @with_seed() def test_log2(): def log2(x): return nd.log2(x) +def grad_op(x): +return 1/(x * math.log(2)) + def grad_grad_op(x): return -1/((x**2) * math.log(2)) arrays = random_arrays((2, 2), (2, 3), (4, 5, 2), (3, 1, 4, 5)) for array in arrays: -check_second_order_unary(array, log2, grad_grad_op) +check_second_order_unary(array, log2, grad_op, grad_grad_op) @with_seed() def test_log10(): def log10(x): return nd.log10(x) +def grad_op(x): +return 1/(x * math.log(10)) + def grad_grad_op(x): return -1/((x**2) * math.log(10)) arrays = random_arrays((2, 2), (2, 3), (4, 5, 2), (3, 1, 4, 5)) for array in arrays: -check_second_order_unary(array, log10, grad_grad_op) +check_second_order_unary(array, log10, grad_op, grad_grad_op) -def check_second_order_unary(x, op, grad_grad_op): +def check_second_order_unary(x, op, grad_op, grad_grad_op): x = nd.array(x) -expect_grad_grad = grad_grad_op(x) +grad_x = grad_op(x) +grad_grad_x = grad_grad_op(x) x.attach_grad() + +# Manual head_grads. +head_grads = nd.random.normal(shape=x.shape) +head_grad_grads = nd.random.normal(shape=x.shape) +head_grads.attach_grad() + +# Perform compute. with autograd.record(): y = op(x) -y_grad = autograd.grad(y, x, create_graph=True, retain_graph=True)[0] -y_grad.backward() -assert_almost_equal(expect_grad_grad.asnumpy(), x.grad.asnumpy()) +y_grad = autograd.grad(y, x, head_grads=head_grads, + create_graph=True, retain_graph=True)[0] + +y_grad.backward(head_grad_grads) + +# Compute expected values. +expected_grad_grad = grad_grad_x.asnumpy() * head_grad_grads.asnumpy() * \ +head_grads.asnumpy() +expected_heads_grad = grad_x.asnumpy() + +# Validate the gradients. +assert_almost_equal(expected_grad_grad, x.grad.asnumpy()) +assert_almost_equal(expected_heads_grad, head_grads.grad.asnumpy()) Review comment: Yeah to verify the fix. I expected `y_grad.backward(head_grad_grads)` to update the `head_grads.grad` similar to the Pytorch Script from the description. Thanks for the suggestion, I will surely try that. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] kshitij12345 commented on a change in pull request #15120: [bug] fix higher grad log
kshitij12345 commented on a change in pull request #15120: [bug] fix higher
grad log
URL: https://github.com/apache/incubator-mxnet/pull/15120#discussion_r290587014
##
File path: src/operator/tensor/elemwise_unary_op_basic.cc
##
@@ -1074,16 +1074,19 @@
MXNET_OPERATOR_REGISTER_BINARY_WITH_SPARSE_CPU_DR(_backward_log,
[](const nnvm::NodePtr& n, const std::vector& ograds) {
// For f(x) -> f = log
// f''(x) = -1 * (f'(x) * f'(x))
-auto gx = nnvm::NodeEntry{n};
+auto gx_mul_head_grads = nnvm::NodeEntry{n}; // f'(x) * head_grads
+auto head_grads = nnvm::NodeEntry{n->inputs[0]};
+auto g_lx = MakeNode("reciprocal", n->attrs.name + "_backward_log_grad",
Review comment:
Sure thing.
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] kshitij12345 commented on a change in pull request #15120: [bug] fix higher grad log
kshitij12345 commented on a change in pull request #15120: [bug] fix higher
grad log
URL: https://github.com/apache/incubator-mxnet/pull/15120#discussion_r290586766
##
File path: src/operator/tensor/elemwise_unary_op_basic.cc
##
@@ -1074,16 +1074,19 @@
MXNET_OPERATOR_REGISTER_BINARY_WITH_SPARSE_CPU_DR(_backward_log,
[](const nnvm::NodePtr& n, const std::vector& ograds) {
// For f(x) -> f = log
// f''(x) = -1 * (f'(x) * f'(x))
-auto gx = nnvm::NodeEntry{n};
+auto gx_mul_head_grads = nnvm::NodeEntry{n}; // f'(x) * head_grads
+auto head_grads = nnvm::NodeEntry{n->inputs[0]};
+auto g_lx = MakeNode("reciprocal", n->attrs.name + "_backward_log_grad",
+{n->inputs[1]}, nullptr, );
auto ggx_mid = MakeNode("elemwise_mul", n->attrs.name +
"_backward_mid_grad_grad",
-{gx, gx}, nullptr, );
+{gx_mul_head_grads, nnvm::NodeEntry{g_lx}},
nullptr, );
auto ggx = MakeNode("negative", n->attrs.name + "_backward_grad_grad",
{nnvm::NodeEntry{ggx_mid}}, nullptr, );
std::vector ret;
ret.emplace_back(MakeNode("elemwise_mul", n->attrs.name +
"_backward_grad_grad",
- {ograds[0], gx}, nullptr, ));
+ {ograds[0], nnvm::NodeEntry{g_lx}}, nullptr, ));
Review comment:
Sorry for the confusion. I forgot to add the line from the test file.
Sure waiting to know what you find.
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] kshitij12345 commented on a change in pull request #15120: [bug] fix higher grad log
kshitij12345 commented on a change in pull request #15120: [bug] fix higher
grad log
URL: https://github.com/apache/incubator-mxnet/pull/15120#discussion_r289614132
##
File path: src/operator/tensor/elemwise_unary_op_basic.cc
##
@@ -1074,16 +1074,19 @@
MXNET_OPERATOR_REGISTER_BINARY_WITH_SPARSE_CPU_DR(_backward_log,
[](const nnvm::NodePtr& n, const std::vector& ograds) {
// For f(x) -> f = log
// f''(x) = -1 * (f'(x) * f'(x))
-auto gx = nnvm::NodeEntry{n};
+auto gx_mul_head_grads = nnvm::NodeEntry{n}; // f'(x) * head_grads
+auto head_grads = nnvm::NodeEntry{n->inputs[0]};
+auto g_lx = MakeNode("reciprocal", n->attrs.name + "_backward_log_grad",
+{n->inputs[1]}, nullptr, );
auto ggx_mid = MakeNode("elemwise_mul", n->attrs.name +
"_backward_mid_grad_grad",
-{gx, gx}, nullptr, );
+{gx_mul_head_grads, nnvm::NodeEntry{g_lx}},
nullptr, );
auto ggx = MakeNode("negative", n->attrs.name + "_backward_grad_grad",
{nnvm::NodeEntry{ggx_mid}}, nullptr, );
std::vector ret;
ret.emplace_back(MakeNode("elemwise_mul", n->attrs.name +
"_backward_grad_grad",
- {ograds[0], gx}, nullptr, ));
+ {ograds[0], nnvm::NodeEntry{g_lx}}, nullptr, ));
Review comment:
https://github.com/apache/incubator-mxnet/blob/37ce3b87268a8154f5c0ad97ce2522478038ee06/tests/python/unittest/test_higher_order_grad.py#L102
I am having trouble with `head_grads.grad` which is being returned as `0's`
(I guess they are somehow not being updated) while I expect it to be the output
of this line.
Please help.
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] xianyujie commented on issue #15108: The test time of the model on GPU is normal, but the test time on CPU is very long.
xianyujie commented on issue #15108: The test time of the model on GPU is normal, but the test time on CPU is very long. URL: https://github.com/apache/incubator-mxnet/issues/15108#issuecomment-498944552 @pengzhao-intel I think you misunderstood the result. Take a look at the following results, different inputs have a great influence on the operation time of convolution layer, What could be the reason for this result? **same image as input, get the output(pre_output1,pre_output2) from stage1_unit1_relu1 layer of the two models.Then test the time of Conv layer, pre_output1 as the input of my model, pre_output2 as the input of original model.** (my_model: 0.075896, original model: 0.006333) **The pre_output1 as the input of the two models, test the time of Conv layer** (my_model: 0.072311, original model: 0.072548) **The pre_output2 as the input of the two models, test the time of Conv layer** (my_model: 0.0055, original model: 0.005653) This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] lanking520 commented on a change in pull request #15155: Fix Scala release
lanking520 commented on a change in pull request #15155: Fix Scala release URL: https://github.com/apache/incubator-mxnet/pull/15155#discussion_r290582549 ## File path: scala-package/externalPom/pom.xml ## @@ -0,0 +1,153 @@ + + + Review comment: Duplicated This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] roywei edited a comment on issue #15152: [CI][nightly] nightly test tutorial failure: test_tutorials.test_python_kvstore
roywei edited a comment on issue #15152: [CI][nightly] nightly test tutorial failure: test_tutorials.test_python_kvstore URL: https://github.com/apache/incubator-mxnet/issues/15152#issuecomment-498939914 Current conclusion is this only happens on CI machines with `NODE_LINUX_GPU_P3` https://github.com/apache/incubator-mxnet/blob/master/tests/nightly/JenkinsfileForBinaries#L131 This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] roywei edited a comment on issue #15152: [CI][nightly] nightly test tutorial failure: test_tutorials.test_python_kvstore
roywei edited a comment on issue #15152: [CI][nightly] nightly test tutorial
failure: test_tutorials.test_python_kvstore
URL:
https://github.com/apache/incubator-mxnet/issues/15152#issuecomment-498939531
However, I was not able to reproduce this on a EC2 P3.8xLarge instance.
All tutorial test passes
```
ci/build.py --docker-registry mxnetci --nvidiadocker --platform
ubuntu_nightly_gpu --docker-build-retries 3 --shm-size 1500m
/work/runtime_functions.sh nightly_tutorial_test_ubuntu_python2_gpu
```
both test_amp and test_kvstore passes
```
[success] 4.03% test_tutorials.test_amp: 71.8112s
[success] 3.73% test_tutorials.test_gluon_end_to_end: 66.5005s
[success] 3.41% test_tutorials.test_gluon_learning_rate_finder: 60.8975s
[success] 2.23% test_tutorials.test_vision_cnn_visualization: 39.6999s
[success] 1.73% test_tutorials.test_basic_data: 30.7975s
[success] 1.44% test_tutorials.test_python_mnist: 25.7051s
[success] 1.33% test_tutorials.test_gluon_save_load_params: 23.7645s
[success] 1.15% test_tutorials.test_basic_module: 20.5767s
[success] 1.01% test_tutorials.test_python_kvstore: 18.0461s
Ran 48 tests in 1783.625s
OK
build.py: 2019-06-04 23:58:56,307Z INFO Waiting for status of container
8fdb8033689e for 600 s.
build.py: 2019-06-04 23:58:56,484Z INFO Container exit status:
{'StatusCode': 0, 'Error': None}
build.py: 2019-06-04 23:58:56,484Z INFO Container exited with success
build.py: 2019-06-04 23:58:56,484Z INFO Stopping container: 8fdb8033689e
build.py: 2019-06-04 23:58:56,486Z INFO Removing container: 8fdb8033689e
```
Also manually running all the commands in
https://github.com/apache/incubator-mxnet/blob/master/docs/tutorials/python/kvstore.md
passes
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] roywei commented on issue #15152: [CI][nightly] nightly test tutorial failure: test_tutorials.test_python_kvstore
roywei commented on issue #15152: [CI][nightly] nightly test tutorial failure: test_tutorials.test_python_kvstore URL: https://github.com/apache/incubator-mxnet/issues/15152#issuecomment-498939914 Current conclusion is this only happens on CI machines with `NODE_LINUX_GPU_P3` This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] roywei edited a comment on issue #15152: [CI][nightly] nightly test tutorial failure: test_tutorials.test_python_kvstore
roywei edited a comment on issue #15152: [CI][nightly] nightly test tutorial
failure: test_tutorials.test_python_kvstore
URL:
https://github.com/apache/incubator-mxnet/issues/15152#issuecomment-498939531
However, I was not able to reproduce this on a EC2 P3.8xLarge instance.
All tutorial test passes
```
ci/build.py --docker-registry mxnetci --nvidiadocker --platform
ubuntu_nightly_gpu --docker-build-retries 3 --shm-size 1500m
/work/runtime_functions.sh nightly_tutorial_test_ubuntu_python2_gpu
```
both test_amp and test_kvstore passes
```
**[success] 4.03% test_tutorials.test_amp: 71.8112s**
[success] 3.73% test_tutorials.test_gluon_end_to_end: 66.5005s
[success] 3.41% test_tutorials.test_gluon_learning_rate_finder: 60.8975s
[success] 2.23% test_tutorials.test_vision_cnn_visualization: 39.6999s
[success] 1.73% test_tutorials.test_basic_data: 30.7975s
[success] 1.44% test_tutorials.test_python_mnist: 25.7051s
[success] 1.33% test_tutorials.test_gluon_save_load_params: 23.7645s
[success] 1.15% test_tutorials.test_basic_module: 20.5767s
**[success] 1.01% test_tutorials.test_python_kvstore: 18.0461s**
Ran 48 tests in 1783.625s
OK
build.py: 2019-06-04 23:58:56,307Z INFO Waiting for status of container
8fdb8033689e for 600 s.
build.py: 2019-06-04 23:58:56,484Z INFO Container exit status:
{'StatusCode': 0, 'Error': None}
build.py: 2019-06-04 23:58:56,484Z INFO Container exited with success
build.py: 2019-06-04 23:58:56,484Z INFO Stopping container: 8fdb8033689e
build.py: 2019-06-04 23:58:56,486Z INFO Removing container: 8fdb8033689e
```
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] roywei edited a comment on issue #15152: [CI][nightly] nightly test tutorial failure: test_tutorials.test_python_kvstore
roywei edited a comment on issue #15152: [CI][nightly] nightly test tutorial
failure: test_tutorials.test_python_kvstore
URL:
https://github.com/apache/incubator-mxnet/issues/15152#issuecomment-498939531
However, I was not able to reproduce this on a EC2 P3.8xLarge instance.
All tutorial test passes
```
ci/build.py --docker-registry mxnetci --nvidiadocker --platform
ubuntu_nightly_gpu --docker-build-retries 3 --shm-size 1500m
/work/runtime_functions.sh nightly_tutorial_test_ubuntu_python2_gpu
```
both test_amp and test_kvstore passes
```
[success] 4.03% test_tutorials.test_amp: 71.8112s
[success] 3.73% test_tutorials.test_gluon_end_to_end: 66.5005s
[success] 3.41% test_tutorials.test_gluon_learning_rate_finder: 60.8975s
[success] 2.23% test_tutorials.test_vision_cnn_visualization: 39.6999s
[success] 1.73% test_tutorials.test_basic_data: 30.7975s
[success] 1.44% test_tutorials.test_python_mnist: 25.7051s
[success] 1.33% test_tutorials.test_gluon_save_load_params: 23.7645s
[success] 1.15% test_tutorials.test_basic_module: 20.5767s
[success] 1.01% test_tutorials.test_python_kvstore: 18.0461s
Ran 48 tests in 1783.625s
OK
build.py: 2019-06-04 23:58:56,307Z INFO Waiting for status of container
8fdb8033689e for 600 s.
build.py: 2019-06-04 23:58:56,484Z INFO Container exit status:
{'StatusCode': 0, 'Error': None}
build.py: 2019-06-04 23:58:56,484Z INFO Container exited with success
build.py: 2019-06-04 23:58:56,484Z INFO Stopping container: 8fdb8033689e
build.py: 2019-06-04 23:58:56,486Z INFO Removing container: 8fdb8033689e
```
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] roywei commented on issue #15152: [CI][nightly] nightly test tutorial failure: test_tutorials.test_python_kvstore
roywei commented on issue #15152: [CI][nightly] nightly test tutorial failure:
test_tutorials.test_python_kvstore
URL:
https://github.com/apache/incubator-mxnet/issues/15152#issuecomment-498939531
However, I was not able to reproduce this on a EC2 P3.8xLarge instance.
All tutorial test passes
```
ci/build.py --docker-registry mxnetci --nvidiadocker --platform
ubuntu_nightly_gpu --docker-build-retries 3 --shm-size 1500m
/work/runtime_functions.sh nightly_tutorial_test_ubuntu_python2_gpu
```
```
Ran 48 tests in 1783.625s
OK
build.py: 2019-06-04 23:58:56,307Z INFO Waiting for status of container
8fdb8033689e for 600 s.
build.py: 2019-06-04 23:58:56,484Z INFO Container exit status:
{'StatusCode': 0, 'Error': None}
build.py: 2019-06-04 23:58:56,484Z INFO Container exited with success
build.py: 2019-06-04 23:58:56,484Z INFO Stopping container: 8fdb8033689e
build.py: 2019-06-04 23:58:56,486Z INFO Removing container: 8fdb8033689e
```
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] roywei edited a comment on issue #15152: [CI][nightly] nightly test tutorial failure: test_tutorials.test_python_kvstore
roywei edited a comment on issue #15152: [CI][nightly] nightly test tutorial
failure: test_tutorials.test_python_kvstore
URL:
https://github.com/apache/incubator-mxnet/issues/15152#issuecomment-498938781
This tutorial test was passing when running on 1 GPU machine.
https://github.com/apache/incubator-mxnet/blob/master/docs/tutorials/python/kvstore.md
```
# The numbers used below assume 4 GPUs
gpus = mx.context.num_gpus()
if gpus > 0:
contexts = [mx.gpu(i) for i in range(gpus)]
else:
contexts = [mx.cpu(i) for i in range(4)]
```
However, when I changed to P3 instances with 4 gpus in
https://github.com/apache/incubator-mxnet/pull/15141. it fails.
```
MXNetError: [01:12:52] src/imperative/./imperative_utils.h:71: Check failed:
inputs[i]->ctx().dev_mask() == ctx.dev_mask() (1 vs. 2) : Operator
broadcast_add require all inputs live on the same context. But the first
argument is on gpu(0) while the 2-th argument is on cpu(0)
Stack trace:
[bt] (0)
/work/mxnet/python/mxnet/../../lib/libmxnet.so(dmlc::LogMessageFatal::~LogMessageFatal()+0x3c)
[0x7f08e7052c1c]
[bt] (1)
/work/mxnet/python/mxnet/../../lib/libmxnet.so(mxnet::imperative::GetContext(nnvm::NodeAttrs
const&, std::vector > const&,
std::vector > const&,
mxnet::Context const&)+0x823) [0x7f08e9fbf343]
[bt] (2)
/work/mxnet/python/mxnet/../../lib/libmxnet.so(mxnet::Imperative::Invoke(mxnet::Context
const&, nnvm::NodeAttrs const&, std::vector > const&, std::vector > const&)+0xdb) [0x7f08e9fcd47b]
[bt] (3)
/work/mxnet/python/mxnet/../../lib/libmxnet.so(MXImperativeInvokeImpl(void*,
int, void**, int*, void***, int, char const**, char const**)+0x1c9)
[0x7f08eaab99d9]
[bt] (4)
/work/mxnet/python/mxnet/../../lib/libmxnet.so(MXImperativeInvokeEx+0x8f)
[0x7f08eaab9edf]
[bt] (5)
/usr/lib/python3.5/lib-dynload/_ctypes.cpython-35m-x86_64-linux-gnu.so(ffi_call_unix64+0x4c)
[0x7f093764ae20]
[bt] (6)
/usr/lib/python3.5/lib-dynload/_ctypes.cpython-35m-x86_64-linux-gnu.so(ffi_call+0x2eb)
[0x7f093764a88b]
[bt] (7)
/usr/lib/python3.5/lib-dynload/_ctypes.cpython-35m-x86_64-linux-gnu.so(_ctypes_callproc+0x49a)
[0x7f093764501a]
[bt] (8)
/usr/lib/python3.5/lib-dynload/_ctypes.cpython-35m-x86_64-linux-gnu.so(+0x9fcb)
[0x7f0937638fcb]
```
error comes from this part of the code during broad_cast add, where `stored`
is `b` and on GPU, `input` is `mx.nd.ones(shape)` on CPU. but it should not
give an error.
```
def update(key, input, stored):
print("update on key: %d" % key)
stored += input * 2
kv._set_updater(update)
kv.pull(3, out=a)
print(a.asnumpy())
kv.push(3, mx.nd.ones(shape))
#
kv.pull(3, out=a)
print(a.asnumpy())
```
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] roywei commented on issue #15152: [CI][nightly] nightly test tutorial failure: test_tutorials.test_python_kvstore
roywei commented on issue #15152: [CI][nightly] nightly test tutorial failure:
test_tutorials.test_python_kvstore
URL:
https://github.com/apache/incubator-mxnet/issues/15152#issuecomment-498938781
This tutorial test was passing when running on 1 GPU machine.
https://github.com/apache/incubator-mxnet/blob/master/docs/tutorials/python/kvstore.md
```
# The numbers used below assume 4 GPUs
gpus = mx.context.num_gpus()
if gpus > 0:
contexts = [mx.gpu(i) for i in range(gpus)]
else:
contexts = [mx.cpu(i) for i in range(4)]
```
However, when I changed to P3 instances with 4 gpus in
https://github.com/apache/incubator-mxnet/pull/15141. it fails.
```
MXNetError: [01:12:52] src/imperative/./imperative_utils.h:71: Check failed:
inputs[i]->ctx().dev_mask() == ctx.dev_mask() (1 vs. 2) : Operator
broadcast_add require all inputs live on the same context. But the first
argument is on gpu(0) while the 2-th argument is on cpu(0)
Stack trace:
[bt] (0)
/work/mxnet/python/mxnet/../../lib/libmxnet.so(dmlc::LogMessageFatal::~LogMessageFatal()+0x3c)
[0x7f08e7052c1c]
[bt] (1)
/work/mxnet/python/mxnet/../../lib/libmxnet.so(mxnet::imperative::GetContext(nnvm::NodeAttrs
const&, std::vector > const&,
std::vector > const&,
mxnet::Context const&)+0x823) [0x7f08e9fbf343]
[bt] (2)
/work/mxnet/python/mxnet/../../lib/libmxnet.so(mxnet::Imperative::Invoke(mxnet::Context
const&, nnvm::NodeAttrs const&, std::vector > const&, std::vector > const&)+0xdb) [0x7f08e9fcd47b]
[bt] (3)
/work/mxnet/python/mxnet/../../lib/libmxnet.so(MXImperativeInvokeImpl(void*,
int, void**, int*, void***, int, char const**, char const**)+0x1c9)
[0x7f08eaab99d9]
[bt] (4)
/work/mxnet/python/mxnet/../../lib/libmxnet.so(MXImperativeInvokeEx+0x8f)
[0x7f08eaab9edf]
[bt] (5)
/usr/lib/python3.5/lib-dynload/_ctypes.cpython-35m-x86_64-linux-gnu.so(ffi_call_unix64+0x4c)
[0x7f093764ae20]
[bt] (6)
/usr/lib/python3.5/lib-dynload/_ctypes.cpython-35m-x86_64-linux-gnu.so(ffi_call+0x2eb)
[0x7f093764a88b]
[bt] (7)
/usr/lib/python3.5/lib-dynload/_ctypes.cpython-35m-x86_64-linux-gnu.so(_ctypes_callproc+0x49a)
[0x7f093764501a]
[bt] (8)
/usr/lib/python3.5/lib-dynload/_ctypes.cpython-35m-x86_64-linux-gnu.so(+0x9fcb)
[0x7f0937638fcb]
```
error comes from this part of the code during broad_cast add
```
def update(key, input, stored):
print("update on key: %d" % key)
stored += input * 2
kv._set_updater(update)
kv.pull(3, out=a)
print(a.asnumpy())
```
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] insikk commented on issue #2870: Does the mxnet support the binary operations?
insikk commented on issue #2870: Does the mxnet support the binary operations? URL: https://github.com/apache/incubator-mxnet/issues/2870#issuecomment-498929418 I wish there is xor operator. +1 for this feature request This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] sandeep-krishnamurthy commented on a change in pull request #14977: Add an utility for operator benchmarks
sandeep-krishnamurthy commented on a change in pull request #14977: Add an
utility for operator benchmarks
URL: https://github.com/apache/incubator-mxnet/pull/14977#discussion_r290570895
##
File path: benchmark/opperf/README.md
##
@@ -0,0 +1,182 @@
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+# MXNet Operator Performance Benchmarks
+
+A Python utility for benchmarking and profiling individual MXNet operator
execution.
+
+With this utility, for each MXNet operator you can get the following details:
+
+**Timing**
+1. Forward execution time
+2. Backward execution time
+3. Time spent for memory management
+
+**Memory**
+1. Total memory allocated
+
+# Motivation
+
+Benchmarks are usually done end-to-end for a given Network Architecture. For
example: ResNet-50 benchmarks on ImageNet data. This is good measurement of
overall performance and health of a deep learning framework. However, it is
important to note the following important factors:
+1. Users use a lot more operators that are not part of a standard network like
ResNet. Example: Tensor manipulation operators like mean, max, topk, argmax,
sort etc.
+2. A standard Network Architecture like ResNet-50 is made up of many operators
Ex: Convolution2D, Softmax, Dense and more. Consider the following scenarios:
+1. We improved the performance of Convolution2D operator, but due to a
bug, Softmax performance went down. Overall, we may observe end to end
benchmarks are running fine, we may miss out the performance degradation of a
single operator which can accumulate and become untraceable.
+2. You need to see in a given network, which operator is taking maximum
time and plan optimization work. With end to end benchmarks, it is hard to get
more fine grained numbers at operator level.
+3. We need to know on different hardware infrastructure (Ex: CPU with MKLDNN,
GPU with NVIDIA CUDA and cuDNN) how different operators performs. With these
details, we can plan the optimization work at operator level, which could
exponentially boost up end to end performance.
+4. You want to have nightly performance tests across all operators in a deep
learning framework to catch regressions early.
+5. We can integrate this framework with a CI/CD system to run per operator
performance tests for PRs. Example: When a PR modifies the kernel of
TransposeConv2D, we can run benchmarks of TransposeConv2D operator to verify
performance.
+
+Hence, in this utility, we will build the functionality to allow users and
developers of deep learning frameworks to easily run benchmarks for individual
operators.
+
+# How to use
+
+## Prerequisites
+
+This utility uses MXNet profiler under the hood to fetch compute and memory
metrics. Hence, you need to build MXNet with `USE_PROFILER=1` flag.
+
+Make sure to build the flavor of MXNet, for example - with/without MKL, with
CUDA 9 or 10.1 etc., on which you would like to measure operator performance.
Finally, you need to add path to your cloned MXNet repository to the PYTHONPATH.
+
+```
+export PYTHONPATH=$PYTHONPATH:/path/to/incubator-mxnet/
+```
+
+## Usecase 1 - Run benchmarks for all the operators
+
+Below command runs all the MXNet operators (NDArray) benchmarks with default
inputs and saves the final result as JSON in the given file.
+
+```
+python incubator-mxnet/benchmark/opperf/opperf.py --output-format json
--output-file mxnet_operator_benchmark_results.json
+```
+
+**Other Supported Options:**
+
+1. **output-format** : `json` or `md` for markdown file output.
+
+2. **ctx** : `cpu` or `gpu`. By default, cpu on CPU machine, gpu(0) on GPU
machine. You can override and set the global context for all operator
benchmarks. Example: --ctx gpu(2).
+
+3. **dtype** : By default, `float32`. You can override and set the global
dtype for all operator benchmarks. Example: --dtype float64.
+
+## Usecase 2 - Run benchmarks for all the operators in a specific category
+
+For example, you want to run benchmarks for all NDArray Broadcast Binary
Operators, Ex: broadcast_add, broadcast_mod, broadcast_pow etc., You just run
the following python script.
+
+```
+#!/usr/bin/python
+from benchmark.opperf.tensor_operations.binary_broadcast_operators import
run_mx_binary_broadcast_operators_benchmarks
+
+# Run all Binary Broadcast operations benchmarks with default input values
+print(run_mx_binary_broadcast_operators_benchmarks())
+```
+
+Output for the above benchmark run, on a CPU machine, would look something
like below:
+
+```
+{'broadcast_mod': [{'avg_time_forward_broadcast_mod': 28.7063,
'avg_time_mem_alloc_cpu/0': 4194.3042,
+'avg_time_backward_broadcast_mod': 12.0954, 'inputs':
{'lhs': (1024, 1024), 'rhs': (1024, 1024)}},
+ {'avg_time_forward_broadcast_mod': 2.7332,
'avg_time_mem_alloc_cpu/0': 400.0,
+'avg_time_backward_broadcast_mod': 1.1288, 'inputs':
{'lhs': (1, 10), 'rhs': (1, 10)}},
+
[GitHub] [incubator-mxnet] larroy commented on a change in pull request #15120: [bug] fix higher grad log
larroy commented on a change in pull request #15120: [bug] fix higher grad log URL: https://github.com/apache/incubator-mxnet/pull/15120#discussion_r290561369 ## File path: tests/python/unittest/test_higher_order_grad.py ## @@ -27,52 +27,79 @@ def test_log(): def log(x): return nd.log(x) +def grad_op(x): +return 1/x + def grad_grad_op(x): return -1/(x**2) arrays = random_arrays((2, 2), (2, 3), (4, 5, 2), (3, 1, 4, 5)) for array in arrays: -check_second_order_unary(array, log, grad_grad_op) +check_second_order_unary(array, log, grad_op, grad_grad_op) @with_seed() def test_log2(): def log2(x): return nd.log2(x) +def grad_op(x): +return 1/(x * math.log(2)) + def grad_grad_op(x): return -1/((x**2) * math.log(2)) arrays = random_arrays((2, 2), (2, 3), (4, 5, 2), (3, 1, 4, 5)) for array in arrays: -check_second_order_unary(array, log2, grad_grad_op) +check_second_order_unary(array, log2, grad_op, grad_grad_op) @with_seed() def test_log10(): def log10(x): return nd.log10(x) +def grad_op(x): +return 1/(x * math.log(10)) + def grad_grad_op(x): return -1/((x**2) * math.log(10)) arrays = random_arrays((2, 2), (2, 3), (4, 5, 2), (3, 1, 4, 5)) for array in arrays: -check_second_order_unary(array, log10, grad_grad_op) +check_second_order_unary(array, log10, grad_op, grad_grad_op) -def check_second_order_unary(x, op, grad_grad_op): +def check_second_order_unary(x, op, grad_op, grad_grad_op): x = nd.array(x) -expect_grad_grad = grad_grad_op(x) +grad_x = grad_op(x) +grad_grad_x = grad_grad_op(x) x.attach_grad() + +# Manual head_grads. +head_grads = nd.random.normal(shape=x.shape) +head_grad_grads = nd.random.normal(shape=x.shape) +head_grads.attach_grad() + +# Perform compute. with autograd.record(): y = op(x) -y_grad = autograd.grad(y, x, create_graph=True, retain_graph=True)[0] -y_grad.backward() -assert_almost_equal(expect_grad_grad.asnumpy(), x.grad.asnumpy()) +y_grad = autograd.grad(y, x, head_grads=head_grads, + create_graph=True, retain_graph=True)[0] + +y_grad.backward(head_grad_grads) + +# Compute expected values. +expected_grad_grad = grad_grad_x.asnumpy() * head_grad_grads.asnumpy() * \ +head_grads.asnumpy() +expected_heads_grad = grad_x.asnumpy() + +# Validate the gradients. +assert_almost_equal(expected_grad_grad, x.grad.asnumpy()) +assert_almost_equal(expected_heads_grad, head_grads.grad.asnumpy()) Review comment: can you try `y_grad_grad = autograd.grad(y_grad, x, ..., create_graph = False...)[0]` and in validation `assert_almost_equal(expected_heads_grad, y_grad_grad.asnumpy())` This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] moneypi commented on issue #15133: Compile failed when disable OPENMP
moneypi commented on issue #15133: Compile failed when disable OPENMP URL: https://github.com/apache/incubator-mxnet/issues/15133#issuecomment-498908780 I repeat the problem when memory is 8G. And compile success after change to 16G. So, that's the reason. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[incubator-mxnet-site] branch asf-site updated: Bump the publish timestamp.
This is an automated email from the ASF dual-hosted git repository. marcoabreu pushed a commit to branch asf-site in repository https://gitbox.apache.org/repos/asf/incubator-mxnet-site.git The following commit(s) were added to refs/heads/asf-site by this push: new a7486a3 Bump the publish timestamp. a7486a3 is described below commit a7486a3fa001c5428cf0dba3868c961bb5488b23 Author: mxnet-ci AuthorDate: Wed Jun 5 01:16:57 2019 + Bump the publish timestamp. --- date.txt | 1 + 1 file changed, 1 insertion(+) diff --git a/date.txt b/date.txt new file mode 100644 index 000..bc0be2f --- /dev/null +++ b/date.txt @@ -0,0 +1 @@ +Wed Jun 5 01:16:57 UTC 2019
[GitHub] [incubator-mxnet] ZhennanQin commented on issue #15151: SSD INT8 ARMv8 Operator _sg_mkldnn_conv is not registered
ZhennanQin commented on issue #15151: SSD INT8 ARMv8 Operator _sg_mkldnn_conv is not registered URL: https://github.com/apache/incubator-mxnet/issues/15151#issuecomment-498898161 Gluon pre-trained int8 model requires mkldnn library and AVX512 instruction sets. MXNet and Gluon doesn't provide int8 solution on ARM backend at moment. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] zachgk opened a new pull request #15155: Fix Scala release
zachgk opened a new pull request #15155: Fix Scala release URL: https://github.com/apache/incubator-mxnet/pull/15155 Add GPG signing to pom.xml through additional externalPom module Fix staging target repository @lanking520 @frankfliu ## Checklist ## ### Essentials ### Please feel free to remove inapplicable items for your PR. - [x] Changes are complete (i.e. I finished coding on this PR) - [x] All changes have test coverage: - Unit tests are added for small changes to verify correctness (e.g. adding a new operator) - Nightly tests are added for complicated/long-running ones (e.g. changing distributed kvstore) - Build tests will be added for build configuration changes (e.g. adding a new build option with NCCL) - [x] Code is well-documented: - For user-facing API changes, API doc string has been updated. - For new C++ functions in header files, their functionalities and arguments are documented. - For new examples, README.md is added to explain the what the example does, the source of the dataset, expected performance on test set and reference to the original paper if applicable - Check the API doc at http://mxnet-ci-doc.s3-accelerate.dualstack.amazonaws.com/PR-$PR_ID/$BUILD_ID/index.html - [x] To the my best knowledge, examples are either not affected by this change, or have been fixed to be compatible with this change This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] stephenrawls commented on issue #13967: Hybrid Variable Pass Through [Bug]
stephenrawls commented on issue #13967: Hybrid Variable Pass Through [Bug] URL: https://github.com/apache/incubator-mxnet/issues/13967#issuecomment-498893060 In one sense, yes it offers a solution. However it is a brittle api. Users have to know that you aren't allowed to return an unmodified input variable in a hybrid block and must wrap it in F.Identity(). It would probably be a better user experience not to have to do that. But for my purposes, yes I can use that and it's fine. Thanks for pointing F.Identity out. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] pengzhao-intel commented on issue #15130: Add NaiveEngine tests in CI
pengzhao-intel commented on issue #15130: Add NaiveEngine tests in CI URL: https://github.com/apache/incubator-mxnet/pull/15130#issuecomment-498890272 @marcoabreu I heard some people are using the naive engine for the better performance but the naive engine can't work very well. I think at least we need to add 1 or 2 tests on CI for the naive engine. Let's use this PR to fix the bugs in the naive engine first and then see how much time from threaded and naive engine :) This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] leleamol commented on issue #14152: data layer grad problem
leleamol commented on issue #14152: data layer grad problem URL: https://github.com/apache/incubator-mxnet/issues/14152#issuecomment-498885682 @songziqin please let us know if you got the answer on forum. It would also help us more in resolving the issue, if there is a reprodicible example This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] apeforest commented on a change in pull request #15120: [bug] fix higher grad log
apeforest commented on a change in pull request #15120: [bug] fix higher grad
log
URL: https://github.com/apache/incubator-mxnet/pull/15120#discussion_r290538567
##
File path: src/operator/tensor/elemwise_unary_op_basic.cc
##
@@ -1074,16 +1074,19 @@
MXNET_OPERATOR_REGISTER_BINARY_WITH_SPARSE_CPU_DR(_backward_log,
[](const nnvm::NodePtr& n, const std::vector& ograds) {
// For f(x) -> f = log
// f''(x) = -1 * (f'(x) * f'(x))
-auto gx = nnvm::NodeEntry{n};
+auto gx_mul_head_grads = nnvm::NodeEntry{n}; // f'(x) * head_grads
+auto head_grads = nnvm::NodeEntry{n->inputs[0]};
+auto g_lx = MakeNode("reciprocal", n->attrs.name + "_backward_log_grad",
+{n->inputs[1]}, nullptr, );
auto ggx_mid = MakeNode("elemwise_mul", n->attrs.name +
"_backward_mid_grad_grad",
-{gx, gx}, nullptr, );
+{gx_mul_head_grads, nnvm::NodeEntry{g_lx}},
nullptr, );
auto ggx = MakeNode("negative", n->attrs.name + "_backward_grad_grad",
{nnvm::NodeEntry{ggx_mid}}, nullptr, );
std::vector ret;
ret.emplace_back(MakeNode("elemwise_mul", n->attrs.name +
"_backward_grad_grad",
- {ograds[0], gx}, nullptr, ));
+ {ograds[0], nnvm::NodeEntry{g_lx}}, nullptr, ));
Review comment:
Still looking into this. The first output should be the gradient of y_grad.
However, the `head_grads.grad` does not get the value. I suspect the returned
value from this function is dropped in the gradient calculation in
imperative.cc. I will look more into this. Stay tuned.
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] leleamol commented on issue #12849: [cmake][cpp-package] Building with cmake does not install the cpp-package API
leleamol commented on issue #12849: [cmake][cpp-package] Building with cmake does not install the cpp-package API URL: https://github.com/apache/incubator-mxnet/issues/12849#issuecomment-498884206 @inglada Please let us know if you were able to build cpp-package using CMakefiles so that we can close this issue. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] leleamol commented on issue #12849: [cmake][cpp-package] Building with cmake does not install the cpp-package API
leleamol commented on issue #12849: [cmake][cpp-package] Building with cmake does not install the cpp-package API URL: https://github.com/apache/incubator-mxnet/issues/12849#issuecomment-498884272 @mxnet-label-bot add [Pending Requester Info] This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] larroy commented on a change in pull request #15120: [bug] fix higher grad log
larroy commented on a change in pull request #15120: [bug] fix higher grad log
URL: https://github.com/apache/incubator-mxnet/pull/15120#discussion_r290528016
##
File path: src/operator/tensor/elemwise_unary_op_basic.cc
##
@@ -1074,16 +1074,19 @@
MXNET_OPERATOR_REGISTER_BINARY_WITH_SPARSE_CPU_DR(_backward_log,
[](const nnvm::NodePtr& n, const std::vector& ograds) {
// For f(x) -> f = log
// f''(x) = -1 * (f'(x) * f'(x))
-auto gx = nnvm::NodeEntry{n};
+auto gx_mul_head_grads = nnvm::NodeEntry{n}; // f'(x) * head_grads
+auto head_grads = nnvm::NodeEntry{n->inputs[0]};
+auto g_lx = MakeNode("reciprocal", n->attrs.name + "_backward_log_grad",
Review comment:
can we add a comment about the inputs and what is g_lx? it would help
reason about the code. Are the inputs of n (backward_log)
- 0: input gradient
- 1: x
?
So g_lx is a node having 1/x ? or the derivative of the log right? can we
rename to g_logx ?
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] leleamol opened a new pull request #15154: [MXNET-1412] Backporting the fix to v1.4.x branch to prevent the crash in naive engine
leleamol opened a new pull request #15154: [MXNET-1412] Backporting the fix to v1.4.x branch to prevent the crash in naive engine URL: https://github.com/apache/incubator-mxnet/pull/15154 ## Description ## This change prevents the crash that happens due to early destruction of shared pointers during NaiveEngine shutdown. The change is available in master. This PR is backporting the change to v1.4.x branch. ## Checklist ## ### Essentials ### Please feel free to remove inapplicable items for your PR. - [y] The PR title starts with [MXNET-$JIRA_ID], where $JIRA_ID refers to the relevant [JIRA issue](https://issues.apache.org/jira/projects/MXNET/issues) created (except PRs with tiny changes) - [y] Changes are complete (i.e. I finished coding on this PR) - [ ] All changes have test coverage: - Unit tests are added for small changes to verify correctness (e.g. adding a new operator) - Nightly tests are added for complicated/long-running ones (e.g. changing distributed kvstore) - Build tests will be added for build configuration changes (e.g. adding a new build option with NCCL) - [ ] Code is well-documented: - For user-facing API changes, API doc string has been updated. - For new C++ functions in header files, their functionalities and arguments are documented. - For new examples, README.md is added to explain the what the example does, the source of the dataset, expected performance on test set and reference to the original paper if applicable - Check the API doc at http://mxnet-ci-doc.s3-accelerate.dualstack.amazonaws.com/PR-$PR_ID/$BUILD_ID/index.html - [ ] To the my best knowledge, examples are either not affected by this change, or have been fixed to be compatible with this change ### Changes ### - [ ] Feature1, tests, (and when applicable, API doc) - [ ] Feature2, tests, (and when applicable, API doc) ## Comments ## - If this change is a backward incompatible change, why must this change be made. - Interesting edge cases to note here This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] larroy commented on a change in pull request #15120: [bug] fix higher grad log
larroy commented on a change in pull request #15120: [bug] fix higher grad log
URL: https://github.com/apache/incubator-mxnet/pull/15120#discussion_r290528016
##
File path: src/operator/tensor/elemwise_unary_op_basic.cc
##
@@ -1074,16 +1074,19 @@
MXNET_OPERATOR_REGISTER_BINARY_WITH_SPARSE_CPU_DR(_backward_log,
[](const nnvm::NodePtr& n, const std::vector& ograds) {
// For f(x) -> f = log
// f''(x) = -1 * (f'(x) * f'(x))
-auto gx = nnvm::NodeEntry{n};
+auto gx_mul_head_grads = nnvm::NodeEntry{n}; // f'(x) * head_grads
+auto head_grads = nnvm::NodeEntry{n->inputs[0]};
+auto g_lx = MakeNode("reciprocal", n->attrs.name + "_backward_log_grad",
Review comment:
can we add a comment about the inputs and what is g_lx? it would help
reason about the code. Are the inputs of n (backward_log)
- 0: input gradient
- 1: x
?
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] larroy commented on a change in pull request #15120: [bug] fix higher grad log
larroy commented on a change in pull request #15120: [bug] fix higher grad log
URL: https://github.com/apache/incubator-mxnet/pull/15120#discussion_r290528016
##
File path: src/operator/tensor/elemwise_unary_op_basic.cc
##
@@ -1074,16 +1074,19 @@
MXNET_OPERATOR_REGISTER_BINARY_WITH_SPARSE_CPU_DR(_backward_log,
[](const nnvm::NodePtr& n, const std::vector& ograds) {
// For f(x) -> f = log
// f''(x) = -1 * (f'(x) * f'(x))
-auto gx = nnvm::NodeEntry{n};
+auto gx_mul_head_grads = nnvm::NodeEntry{n}; // f'(x) * head_grads
+auto head_grads = nnvm::NodeEntry{n->inputs[0]};
+auto g_lx = MakeNode("reciprocal", n->attrs.name + "_backward_log_grad",
Review comment:
can we add a comment about the inputs? it would help reason about the code.
Are the inputs of n (backward_log)
- 0: input gradient
- 1: x
?
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] larroy commented on a change in pull request #15120: [bug] fix higher grad log
larroy commented on a change in pull request #15120: [bug] fix higher grad log
URL: https://github.com/apache/incubator-mxnet/pull/15120#discussion_r290528016
##
File path: src/operator/tensor/elemwise_unary_op_basic.cc
##
@@ -1074,16 +1074,19 @@
MXNET_OPERATOR_REGISTER_BINARY_WITH_SPARSE_CPU_DR(_backward_log,
[](const nnvm::NodePtr& n, const std::vector& ograds) {
// For f(x) -> f = log
// f''(x) = -1 * (f'(x) * f'(x))
-auto gx = nnvm::NodeEntry{n};
+auto gx_mul_head_grads = nnvm::NodeEntry{n}; // f'(x) * head_grads
+auto head_grads = nnvm::NodeEntry{n->inputs[0]};
+auto g_lx = MakeNode("reciprocal", n->attrs.name + "_backward_log_grad",
Review comment:
can we add a comment about the inputs? it would help reason about the code.
Are the inputs
- 0: ograd
- 1: f(x)
?
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] larroy commented on a change in pull request #15120: [bug] fix higher grad log
larroy commented on a change in pull request #15120: [bug] fix higher grad log
URL: https://github.com/apache/incubator-mxnet/pull/15120#discussion_r290528016
##
File path: src/operator/tensor/elemwise_unary_op_basic.cc
##
@@ -1074,16 +1074,19 @@
MXNET_OPERATOR_REGISTER_BINARY_WITH_SPARSE_CPU_DR(_backward_log,
[](const nnvm::NodePtr& n, const std::vector& ograds) {
// For f(x) -> f = log
// f''(x) = -1 * (f'(x) * f'(x))
-auto gx = nnvm::NodeEntry{n};
+auto gx_mul_head_grads = nnvm::NodeEntry{n}; // f'(x) * head_grads
+auto head_grads = nnvm::NodeEntry{n->inputs[0]};
+auto g_lx = MakeNode("reciprocal", n->attrs.name + "_backward_log_grad",
Review comment:
can we add a comment about the inputs? it would help reason about the code.
Are the inputs
- [0]: ograd
- [1]: f(x)
?
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] larroy commented on a change in pull request #15120: [bug] fix higher grad log
larroy commented on a change in pull request #15120: [bug] fix higher grad log
URL: https://github.com/apache/incubator-mxnet/pull/15120#discussion_r290528016
##
File path: src/operator/tensor/elemwise_unary_op_basic.cc
##
@@ -1074,16 +1074,19 @@
MXNET_OPERATOR_REGISTER_BINARY_WITH_SPARSE_CPU_DR(_backward_log,
[](const nnvm::NodePtr& n, const std::vector& ograds) {
// For f(x) -> f = log
// f''(x) = -1 * (f'(x) * f'(x))
-auto gx = nnvm::NodeEntry{n};
+auto gx_mul_head_grads = nnvm::NodeEntry{n}; // f'(x) * head_grads
+auto head_grads = nnvm::NodeEntry{n->inputs[0]};
+auto g_lx = MakeNode("reciprocal", n->attrs.name + "_backward_log_grad",
Review comment:
can we add a comment about the inputs? it would help reason about the code.
Are the inputs 0: ograd .1: f(x) . ?
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
[incubator-mxnet] branch numpy updated: numpy concatenate (#15104)
This is an automated email from the ASF dual-hosted git repository.
reminisce pushed a commit to branch numpy
in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git
The following commit(s) were added to refs/heads/numpy by this push:
new aa5153f numpy concatenate (#15104)
aa5153f is described below
commit aa5153fdd463c9a37192b4dd2b3378c4f21e40bf
Author: Hao Jin
AuthorDate: Tue Jun 4 15:55:27 2019 -0700
numpy concatenate (#15104)
---
python/mxnet/ndarray/numpy/_op.py | 27 -
python/mxnet/numpy/multiarray.py | 29 +-
python/mxnet/symbol/numpy/_symbol.py | 27 -
src/operator/nn/concat.cc | 12 +++---
src/operator/numpy/np_matrix_op.cc| 58 +++
src/operator/numpy/np_matrix_op.cu| 4 ++
src/operator/quantization/quantized_concat.cc | 12 +++---
tests/python/unittest/test_numpy_op.py| 51 +++
8 files changed, 204 insertions(+), 16 deletions(-)
diff --git a/python/mxnet/ndarray/numpy/_op.py
b/python/mxnet/ndarray/numpy/_op.py
index 34218e3..6c83e1f 100644
--- a/python/mxnet/ndarray/numpy/_op.py
+++ b/python/mxnet/ndarray/numpy/_op.py
@@ -24,7 +24,7 @@ from ...util import _sanity_check_params, set_module
from ...context import current_context
from . import _internal as _npi
-__all__ = ['zeros', 'ones', 'maximum', 'minimum', 'stack', 'arange', 'argmax']
+__all__ = ['zeros', 'ones', 'maximum', 'minimum', 'stack', 'concatenate',
'arange', 'argmax']
@set_module('mxnet.ndarray.numpy')
@@ -277,3 +277,28 @@ def argmax(a, axis=None, out=None):
with the dimension along `axis` removed.
"""
return _npi.argmax(a, axis=axis, keepdims=False, out=out)
+
+
+@set_module('mxnet.ndarray.numpy')
+def concatenate(seq, axis=0, out=None):
+"""Join a sequence of arrays along an existing axis.
+
+Parameters
+--
+a1, a2, ... : sequence of array_like
+The arrays must have the same shape, except in the dimension
+corresponding to `axis` (the first, by default).
+axis : int, optional
+The axis along which the arrays will be joined. If axis is None,
+arrays are flattened before use. Default is 0.
+out : ndarray, optional
+If provided, the destination to place the result. The shape must be
+correct, matching that of what concatenate would have returned if no
+out argument were specified.
+
+Returns
+---
+res : ndarray
+The concatenated array.
+"""
+return _npi.concatenate(*seq, dim=axis, out=out)
diff --git a/python/mxnet/numpy/multiarray.py b/python/mxnet/numpy/multiarray.py
index 212dfe3..6b3dcde 100644
--- a/python/mxnet/numpy/multiarray.py
+++ b/python/mxnet/numpy/multiarray.py
@@ -37,8 +37,8 @@ from ..context import current_context
from ..ndarray import numpy as _mx_nd_np
from ..ndarray.numpy import _internal as _npi
-__all__ = ['ndarray', 'empty', 'array', 'zeros', 'ones', 'maximum', 'minimum',
'stack', 'arange',
- 'argmax']
+__all__ = ['ndarray', 'empty', 'array', 'zeros', 'ones', 'maximum', 'minimum',
'stack',
+ 'concatenate', 'arange', 'argmax']
# This function is copied from ndarray.py since pylint
@@ -1486,3 +1486,28 @@ def argmax(a, axis=None, out=None):
with the dimension along `axis` removed.
"""
return _mx_nd_np.argmax(a, axis, out)
+
+
+@set_module('mxnet.numpy')
+def concatenate(seq, axis=0, out=None):
+"""Join a sequence of arrays along an existing axis.
+
+Parameters
+--
+a1, a2, ... : sequence of array_like
+The arrays must have the same shape, except in the dimension
+corresponding to `axis` (the first, by default).
+axis : int, optional
+The axis along which the arrays will be joined. If axis is None,
+arrays are flattened before use. Default is 0.
+out : ndarray, optional
+If provided, the destination to place the result. The shape must be
+correct, matching that of what concatenate would have returned if no
+out argument were specified.
+
+Returns
+---
+res : ndarray
+The concatenated array.
+"""
+return _mx_nd_np.concatenate(seq, axis=axis, out=out)
diff --git a/python/mxnet/symbol/numpy/_symbol.py
b/python/mxnet/symbol/numpy/_symbol.py
index b2d8a5b..7a55547 100644
--- a/python/mxnet/symbol/numpy/_symbol.py
+++ b/python/mxnet/symbol/numpy/_symbol.py
@@ -29,7 +29,7 @@ from ..symbol import Symbol
from .._internal import _set_np_symbol_class
from . import _internal as _npi
-__all__ = ['zeros', 'ones', 'maximum', 'minimum', 'stack', 'arange', 'argmax']
+__all__ = ['zeros', 'ones', 'maximum', 'minimum', 'stack', 'concatenate',
'arange', 'argmax']
@set_module('mxnet.symbol.numpy')
@@ -1061,6 +1061,31 @@ def stack(arrays, axis=0, out=None):
@set_module('mxnet.symbol.numpy')
+def concatenate(seq,
[GitHub] [incubator-mxnet] reminisce merged pull request #15104: Numpy-compatible Concatenate
reminisce merged pull request #15104: Numpy-compatible Concatenate URL: https://github.com/apache/incubator-mxnet/pull/15104 This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] larroy commented on a change in pull request #15120: [bug] fix higher grad log
larroy commented on a change in pull request #15120: [bug] fix higher grad log URL: https://github.com/apache/incubator-mxnet/pull/15120#discussion_r290527160 ## File path: tests/python/unittest/test_higher_order_grad.py ## @@ -27,52 +27,79 @@ def test_log(): def log(x): return nd.log(x) +def grad_op(x): +return 1/x + def grad_grad_op(x): return -1/(x**2) arrays = random_arrays((2, 2), (2, 3), (4, 5, 2), (3, 1, 4, 5)) for array in arrays: -check_second_order_unary(array, log, grad_grad_op) +check_second_order_unary(array, log, grad_op, grad_grad_op) @with_seed() def test_log2(): def log2(x): return nd.log2(x) +def grad_op(x): +return 1/(x * math.log(2)) + def grad_grad_op(x): return -1/((x**2) * math.log(2)) arrays = random_arrays((2, 2), (2, 3), (4, 5, 2), (3, 1, 4, 5)) for array in arrays: -check_second_order_unary(array, log2, grad_grad_op) +check_second_order_unary(array, log2, grad_op, grad_grad_op) @with_seed() def test_log10(): def log10(x): return nd.log10(x) +def grad_op(x): +return 1/(x * math.log(10)) + def grad_grad_op(x): return -1/((x**2) * math.log(10)) arrays = random_arrays((2, 2), (2, 3), (4, 5, 2), (3, 1, 4, 5)) for array in arrays: -check_second_order_unary(array, log10, grad_grad_op) +check_second_order_unary(array, log10, grad_op, grad_grad_op) -def check_second_order_unary(x, op, grad_grad_op): +def check_second_order_unary(x, op, grad_op, grad_grad_op): x = nd.array(x) -expect_grad_grad = grad_grad_op(x) +grad_x = grad_op(x) +grad_grad_x = grad_grad_op(x) x.attach_grad() + +# Manual head_grads. +head_grads = nd.random.normal(shape=x.shape) +head_grad_grads = nd.random.normal(shape=x.shape) +head_grads.attach_grad() + +# Perform compute. with autograd.record(): y = op(x) -y_grad = autograd.grad(y, x, create_graph=True, retain_graph=True)[0] -y_grad.backward() -assert_almost_equal(expect_grad_grad.asnumpy(), x.grad.asnumpy()) +y_grad = autograd.grad(y, x, head_grads=head_grads, + create_graph=True, retain_graph=True)[0] + +y_grad.backward(head_grad_grads) + +# Compute expected values. +expected_grad_grad = grad_grad_x.asnumpy() * head_grad_grads.asnumpy() * \ +head_grads.asnumpy() +expected_heads_grad = grad_x.asnumpy() + +# Validate the gradients. +assert_almost_equal(expected_grad_grad, x.grad.asnumpy()) +assert_almost_equal(expected_heads_grad, head_grads.grad.asnumpy()) Review comment: Now I understand your question, i don't think anything is updating head_grads here. Why do you want to set the head gradients manually? To verify your fix? This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] larroy commented on a change in pull request #15120: [bug] fix higher grad log
larroy commented on a change in pull request #15120: [bug] fix higher grad log URL: https://github.com/apache/incubator-mxnet/pull/15120#discussion_r290527160 ## File path: tests/python/unittest/test_higher_order_grad.py ## @@ -27,52 +27,79 @@ def test_log(): def log(x): return nd.log(x) +def grad_op(x): +return 1/x + def grad_grad_op(x): return -1/(x**2) arrays = random_arrays((2, 2), (2, 3), (4, 5, 2), (3, 1, 4, 5)) for array in arrays: -check_second_order_unary(array, log, grad_grad_op) +check_second_order_unary(array, log, grad_op, grad_grad_op) @with_seed() def test_log2(): def log2(x): return nd.log2(x) +def grad_op(x): +return 1/(x * math.log(2)) + def grad_grad_op(x): return -1/((x**2) * math.log(2)) arrays = random_arrays((2, 2), (2, 3), (4, 5, 2), (3, 1, 4, 5)) for array in arrays: -check_second_order_unary(array, log2, grad_grad_op) +check_second_order_unary(array, log2, grad_op, grad_grad_op) @with_seed() def test_log10(): def log10(x): return nd.log10(x) +def grad_op(x): +return 1/(x * math.log(10)) + def grad_grad_op(x): return -1/((x**2) * math.log(10)) arrays = random_arrays((2, 2), (2, 3), (4, 5, 2), (3, 1, 4, 5)) for array in arrays: -check_second_order_unary(array, log10, grad_grad_op) +check_second_order_unary(array, log10, grad_op, grad_grad_op) -def check_second_order_unary(x, op, grad_grad_op): +def check_second_order_unary(x, op, grad_op, grad_grad_op): x = nd.array(x) -expect_grad_grad = grad_grad_op(x) +grad_x = grad_op(x) +grad_grad_x = grad_grad_op(x) x.attach_grad() + +# Manual head_grads. +head_grads = nd.random.normal(shape=x.shape) +head_grad_grads = nd.random.normal(shape=x.shape) +head_grads.attach_grad() + +# Perform compute. with autograd.record(): y = op(x) -y_grad = autograd.grad(y, x, create_graph=True, retain_graph=True)[0] -y_grad.backward() -assert_almost_equal(expect_grad_grad.asnumpy(), x.grad.asnumpy()) +y_grad = autograd.grad(y, x, head_grads=head_grads, + create_graph=True, retain_graph=True)[0] + +y_grad.backward(head_grad_grads) + +# Compute expected values. +expected_grad_grad = grad_grad_x.asnumpy() * head_grad_grads.asnumpy() * \ +head_grads.asnumpy() +expected_heads_grad = grad_x.asnumpy() + +# Validate the gradients. +assert_almost_equal(expected_grad_grad, x.grad.asnumpy()) +assert_almost_equal(expected_heads_grad, head_grads.grad.asnumpy()) Review comment: Now I understand your question, i don't think anything is updating head_grads.grad here (this is done when running backward). Why do you want to set the head gradients manually? To verify your fix? This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] larroy commented on a change in pull request #15120: [bug] fix higher grad log
larroy commented on a change in pull request #15120: [bug] fix higher grad log
URL: https://github.com/apache/incubator-mxnet/pull/15120#discussion_r290526536
##
File path: src/operator/tensor/elemwise_unary_op_basic.cc
##
@@ -1074,16 +1074,19 @@
MXNET_OPERATOR_REGISTER_BINARY_WITH_SPARSE_CPU_DR(_backward_log,
[](const nnvm::NodePtr& n, const std::vector& ograds) {
// For f(x) -> f = log
// f''(x) = -1 * (f'(x) * f'(x))
-auto gx = nnvm::NodeEntry{n};
+auto gx_mul_head_grads = nnvm::NodeEntry{n}; // f'(x) * head_grads
+auto head_grads = nnvm::NodeEntry{n->inputs[0]};
+auto g_lx = MakeNode("reciprocal", n->attrs.name + "_backward_log_grad",
+{n->inputs[1]}, nullptr, );
auto ggx_mid = MakeNode("elemwise_mul", n->attrs.name +
"_backward_mid_grad_grad",
-{gx, gx}, nullptr, );
+{gx_mul_head_grads, nnvm::NodeEntry{g_lx}},
nullptr, );
auto ggx = MakeNode("negative", n->attrs.name + "_backward_grad_grad",
{nnvm::NodeEntry{ggx_mid}}, nullptr, );
std::vector ret;
ret.emplace_back(MakeNode("elemwise_mul", n->attrs.name +
"_backward_grad_grad",
- {ograds[0], gx}, nullptr, ));
+ {ograds[0], nnvm::NodeEntry{g_lx}}, nullptr, ));
Review comment:
Hi. What do you mean by head_grads.grad? NodeEntry doesn't have a grad
field. Could you clarify? Are you referring to the python code below? The
gradient is always 0 when attach_grad() is called. The value is updated after
running backward on an output, or using autograd.grad.
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
[incubator-mxnet] branch master updated (134a3e8 -> 910583e)
This is an automated email from the ASF dual-hosted git repository. wkcn pushed a change to branch master in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git. from 134a3e8 fix nightly (#15141) add 910583e fix misspell (#15149) No new revisions were added by this update. Summary of changes: python/mxnet/profiler.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
[GitHub] [incubator-mxnet] eric-haibin-lin opened a new pull request #15153: [DOC] Add clarification to MXNET_CPU_WORKER_NTHREADS
eric-haibin-lin opened a new pull request #15153: [DOC] Add clarification to MXNET_CPU_WORKER_NTHREADS URL: https://github.com/apache/incubator-mxnet/pull/15153 ## Description ## As title ## Checklist ## ### Essentials ### Please feel free to remove inapplicable items for your PR. - [x] The PR title starts with [MXNET-$JIRA_ID], where $JIRA_ID refers to the relevant [JIRA issue](https://issues.apache.org/jira/projects/MXNET/issues) created (except PRs with tiny changes) - [x] Changes are complete (i.e. I finished coding on this PR) - [x] All changes have test coverage: - Unit tests are added for small changes to verify correctness (e.g. adding a new operator) - Nightly tests are added for complicated/long-running ones (e.g. changing distributed kvstore) - Build tests will be added for build configuration changes (e.g. adding a new build option with NCCL) - [x] Code is well-documented: - For user-facing API changes, API doc string has been updated. - For new C++ functions in header files, their functionalities and arguments are documented. - For new examples, README.md is added to explain the what the example does, the source of the dataset, expected performance on test set and reference to the original paper if applicable - Check the API doc at http://mxnet-ci-doc.s3-accelerate.dualstack.amazonaws.com/PR-$PR_ID/$BUILD_ID/index.html - [ ] To the my best knowledge, examples are either not affected by this change, or have been fixed to be compatible with this change ### Changes ### - [ ] Feature1, tests, (and when applicable, API doc) - [ ] Feature2, tests, (and when applicable, API doc) ## Comments ## - If this change is a backward incompatible change, why must this change be made. - Interesting edge cases to note here This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] larroy commented on a change in pull request #15120: [bug] fix higher grad log
larroy commented on a change in pull request #15120: [bug] fix higher grad log
URL: https://github.com/apache/incubator-mxnet/pull/15120#discussion_r290526536
##
File path: src/operator/tensor/elemwise_unary_op_basic.cc
##
@@ -1074,16 +1074,19 @@
MXNET_OPERATOR_REGISTER_BINARY_WITH_SPARSE_CPU_DR(_backward_log,
[](const nnvm::NodePtr& n, const std::vector& ograds) {
// For f(x) -> f = log
// f''(x) = -1 * (f'(x) * f'(x))
-auto gx = nnvm::NodeEntry{n};
+auto gx_mul_head_grads = nnvm::NodeEntry{n}; // f'(x) * head_grads
+auto head_grads = nnvm::NodeEntry{n->inputs[0]};
+auto g_lx = MakeNode("reciprocal", n->attrs.name + "_backward_log_grad",
+{n->inputs[1]}, nullptr, );
auto ggx_mid = MakeNode("elemwise_mul", n->attrs.name +
"_backward_mid_grad_grad",
-{gx, gx}, nullptr, );
+{gx_mul_head_grads, nnvm::NodeEntry{g_lx}},
nullptr, );
auto ggx = MakeNode("negative", n->attrs.name + "_backward_grad_grad",
{nnvm::NodeEntry{ggx_mid}}, nullptr, );
std::vector ret;
ret.emplace_back(MakeNode("elemwise_mul", n->attrs.name +
"_backward_grad_grad",
- {ograds[0], gx}, nullptr, ));
+ {ograds[0], nnvm::NodeEntry{g_lx}}, nullptr, ));
Review comment:
Hi. What do you mean by head_grads.grad? NodeEntry doesn't have a grad
field. Could you clarify?
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] wkcn merged pull request #15149: Fix Misspell in Profiler
wkcn merged pull request #15149: Fix Misspell in Profiler URL: https://github.com/apache/incubator-mxnet/pull/15149 This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] larroy commented on issue #15120: [bug] fix higher grad log
larroy commented on issue #15120: [bug] fix higher grad log URL: https://github.com/apache/incubator-mxnet/pull/15120#issuecomment-498870894 Hi @kshitij12345 thanks for looking into this. I think we need to clarify what exactly we have in the first parameter of FGradient "node". We were a bit puzzled with @apeforest looking at your PR. I validated the results with the tests but I think I tried only one log, don't remember which base. But the result seemed correct to me, I guess I missed this problem. Why do you say that node is ograd*f'(x)? the node argument I understand is the node to calculate the gradient for, in this case we are calculating the gradient of the backward of the log. So are you saying that by chain rule, the node is ograd(of log) * d (log(x)) / dx = ograd * reciprocal? Would be great if we could add this to the documentation, either to the FGradient typedef or to new_op. Otherwise I always have to dig through the code to refresh this. I think is poorly documented and tricky. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] mxnet-label-bot commented on issue #15152: [CI][nightly] nightly test tutorial failure: test_tutorials.test_python_kvstore
mxnet-label-bot commented on issue #15152: [CI][nightly] nightly test tutorial failure: test_tutorials.test_python_kvstore URL: https://github.com/apache/incubator-mxnet/issues/15152#issuecomment-498851184 Hey, this is the MXNet Label Bot. Thank you for submitting the issue! I will try and suggest some labels so that the appropriate MXNet community members can help resolve it. Here are my recommended labels: Test, CI This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] roywei opened a new issue #15152: [CI][nightly] nightly test tutorial failure: test_tutorials.test_python_kvstore
roywei opened a new issue #15152: [CI][nightly] nightly test tutorial failure: test_tutorials.test_python_kvstore URL: https://github.com/apache/incubator-mxnet/issues/15152 FAIL: test_tutorials.test_python_kvstore http://jenkins.mxnet-ci.amazon-ml.com/blue/organizations/jenkins/NightlyTestsForBinaries/detail/master/335/pipeline/147 This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] apivovarov opened a new issue #15151: SSD INT8 ARMv8 Operator _sg_mkldnn_conv is not registered
apivovarov opened a new issue #15151: SSD INT8 ARMv8 Operator _sg_mkldnn_conv
is not registered
URL: https://github.com/apache/incubator-mxnet/issues/15151
I tried to use quantized SSD model `ssd_512_mobilenet1.0_voc_int8` on ARMv8.
Error:
```
Traceback (most recent call last):
File "./run-gluoncv.py", line 11, in
net = model_zoo.get_model('ssd_512_mobilenet1.0_voc_int8',
pretrained=True)
File
"/usr/local/lib/python3.5/dist-packages/gluoncv/model_zoo/model_zoo.py", line
231, in get_model
net = _models[name](**kwargs)
File
"/usr/local/lib/python3.5/dist-packages/gluoncv/model_zoo/quantized/quantized.py",
line 45, in func
sym_net = SymbolBlock.imports(json_file, ['data'], None, ctx=ctx)
File "/usr/local/lib/python3.5/dist-packages/mxnet/gluon/block.py", line
1018, in imports
sym = symbol.load(symbol_file)
File "/usr/local/lib/python3.5/dist-packages/mxnet/symbol/symbol.py", line
2728, in load
check_call(_LIB.MXSymbolCreateFromFile(c_str(fname),
ctypes.byref(handle)))
File "/usr/local/lib/python3.5/dist-packages/mxnet/base.py", line 253, in
check_call
raise MXNetError(py_str(_LIB.MXGetLastError()))
mxnet.base.MXNetError: Failed loading Op quantized_sg_mkldnn_conv_bn_relu_0
of type _sg_mkldnn_conv: [21:05:05] ../3rdparty/tvm/nnvm/src/core/op.cc:74:
Check failed: op != nullptr: Operator _sg_mkldnn_conv is not registered
Stack trace:
[bt] (0)
/usr/local/lib/python3.5/dist-packages/mxnet/libmxnet.so(nnvm::Op::Get(std::__cxx11::basic_string, std::allocator > const&)+0x328) [0xa069d5d0]
[bt] (1)
/usr/local/lib/python3.5/dist-packages/mxnet/libmxnet.so(+0x232d29c)
[0xa06e429c]
[bt] (2)
/usr/local/lib/python3.5/dist-packages/mxnet/libmxnet.so(dmlc::JSONObjectReadHelper::ReadAllFields(dmlc::JSONReader*)+0xd0)
[0xa06e8258]
[bt] (3)
/usr/local/lib/python3.5/dist-packages/mxnet/libmxnet.so(+0x2329a48)
[0xa06e0a48]
[bt] (4)
/usr/local/lib/python3.5/dist-packages/mxnet/libmxnet.so(+0x232aa80)
[0xa06e1a80]
[bt] (5)
/usr/local/lib/python3.5/dist-packages/mxnet/libmxnet.so(std::_Function_handler::_M_invoke(std::_Any_data const&,
nnvm::Graph&&)+0xcc) [0x9ed2efbc]
[bt] (6)
/usr/local/lib/python3.5/dist-packages/mxnet/libmxnet.so(nnvm::ApplyPasses(nnvm::Graph,
std::vector,
std::allocator >, std::allocator, std::allocator > > > const&)+0x28c)
[0xa06a0194]
[bt] (7)
/usr/local/lib/python3.5/dist-packages/mxnet/libmxnet.so(mxnet::LoadLegacyJSONPass(nnvm::Graph)+0x44c)
[0x9ed35294]
[bt] (8)
/usr/local/lib/python3.5/dist-packages/mxnet/libmxnet.so(std::_Function_handler::_M_invoke(std::_Any_data const&,
nnvm::Graph&&)+0xcc) [0x9ed2efbc]
```
I built `libmxnet.so` on aarch64 cpu without MKL support
incubator-mxnet version 134a3e8c (Jun 4):
```
cmake -GNinja \
-DUSE_CUDA=OFF \
-DUSE_MKL_IF_AVAILABLE=OFF \
-DCMAKE_BUILD_TYPE=Release \
-DCMAKE_CUDA_COMPILER_LAUNCHER=ccache \
-DCMAKE_C_COMPILER_LAUNCHER=ccache \
-DCMAKE_CXX_COMPILER_LAUNCHER=ccache \
..
ninja -j16
```
run inference script:
```
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] mxnet-label-bot commented on issue #15151: SSD INT8 ARMv8 Operator _sg_mkldnn_conv is not registered
mxnet-label-bot commented on issue #15151: SSD INT8 ARMv8 Operator _sg_mkldnn_conv is not registered URL: https://github.com/apache/incubator-mxnet/issues/15151#issuecomment-498843835 Hey, this is the MXNet Label Bot. Thank you for submitting the issue! I will try and suggest some labels so that the appropriate MXNet community members can help resolve it. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[incubator-mxnet-site] branch asf-site updated: Bump the publish timestamp.
This is an automated email from the ASF dual-hosted git repository. marcoabreu pushed a commit to branch asf-site in repository https://gitbox.apache.org/repos/asf/incubator-mxnet-site.git The following commit(s) were added to refs/heads/asf-site by this push: new cd06f3c Bump the publish timestamp. cd06f3c is described below commit cd06f3cb029bdae774579c7f7f9e5781bbb26581 Author: mxnet-ci AuthorDate: Tue Jun 4 20:51:13 2019 + Bump the publish timestamp. --- date.txt | 1 + 1 file changed, 1 insertion(+) diff --git a/date.txt b/date.txt new file mode 100644 index 000..226b50f --- /dev/null +++ b/date.txt @@ -0,0 +1 @@ +Tue Jun 4 20:51:13 UTC 2019
[GitHub] [incubator-mxnet] abhinavs95 edited a comment on issue #14712: Conv3DTranspose not work in Ubuntu.
abhinavs95 edited a comment on issue #14712: Conv3DTranspose not work in Ubuntu. URL: https://github.com/apache/incubator-mxnet/issues/14712#issuecomment-498827624 Duplicate of #13135 This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] piyushghai commented on issue #15150: Fix dumps for Constant initializer
piyushghai commented on issue #15150: Fix dumps for Constant initializer URL: https://github.com/apache/incubator-mxnet/pull/15150#issuecomment-498827812 @mxnet-label-bot Add [pr-awaiting-review, NDArray] This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] abhinavs95 commented on issue #14712: Conv3DTranspose not work in Ubuntu.
abhinavs95 commented on issue #14712: Conv3DTranspose not work in Ubuntu. URL: https://github.com/apache/incubator-mxnet/issues/14712#issuecomment-498827624 Possibly related to #13135 This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] piyushghai commented on issue #15147: Python test failures under windows now report the exit code
piyushghai commented on issue #15147: Python test failures under windows now report the exit code URL: https://github.com/apache/incubator-mxnet/pull/15147#issuecomment-498827351 @david-seiler Thanks for your contributions. Could you look into the CI failure on windows-gpu ? @mxnet-label-bot Add [pr-awaiting-review, Test] This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] piyushghai commented on issue #15149: Fix Misspell in Profiler
piyushghai commented on issue #15149: Fix Misspell in Profiler URL: https://github.com/apache/incubator-mxnet/pull/15149#issuecomment-498826811 @mxnet-label-bot Add [Docs, pr-awaiting-merge] This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] piyushghai commented on issue #15144: Relax Visual Studio version constraint in the specialization of `dmlc::type_name_helper` for `DT=mxnet::Tuple`
piyushghai commented on issue #15144: Relax Visual Studio version constraint in the specialization of `dmlc::type_name_helper` for `DT=mxnet::Tuple` URL: https://github.com/apache/incubator-mxnet/pull/15144#issuecomment-498826459 Thanks for your contributions @Vigilans @mxnet-label-bot Add[pr-awaiting-review, Backend] This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] piyushghai commented on issue #15142: [Dependency Update] Bump up cudnn version
piyushghai commented on issue #15142: [Dependency Update] Bump up cudnn version URL: https://github.com/apache/incubator-mxnet/pull/15142#issuecomment-498825409 @stu1130 Can you look into the CI failures ? @mxnet-label-bot Add[pr-awaiting-review, Backend] This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] piyushghai commented on issue #15139: [WIP][numpy] Fix for D2L Chapters 2 and 3
piyushghai commented on issue #15139: [WIP][numpy] Fix for D2L Chapters 2 and 3 URL: https://github.com/apache/incubator-mxnet/pull/15139#issuecomment-498824930 Thanks for your contributions @reminisce @mxnet-label-bot Add [numpy, pr-awaiting-review] This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] piyushghai commented on issue #15137: [WIP] 1.5.0 news
piyushghai commented on issue #15137: [WIP] 1.5.0 news URL: https://github.com/apache/incubator-mxnet/pull/15137#issuecomment-498823631 Thanks for your contributions @roywei @mxnet-label-bot Add [Doc, pr-awaiting-review] This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] piyushghai commented on issue #15132: Profiler API Enhancements
piyushghai commented on issue #15132: Profiler API Enhancements URL: https://github.com/apache/incubator-mxnet/pull/15132#issuecomment-498823245 Thanks for your contributions @Zha0q1 . @mxnet-label-bot Add [Profiler, pr-work-in-progress] This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] piyushghai commented on issue #15130: Add NaiveEngine tests in CI
piyushghai commented on issue #15130: Add NaiveEngine tests in CI URL: https://github.com/apache/incubator-mxnet/pull/15130#issuecomment-498822818 Thanks for your contributions @xinyu-intel . I agree with Marco, if we want to run NaiveEngine tests in CI, nightly might be the more appropriate place for them as the normal unit test suite runs on every PR and thus speed is of essence. @mxnet-label-bot Add [CI] This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] piyushghai commented on issue #15128: update LICENSE
piyushghai commented on issue #15128: update LICENSE URL: https://github.com/apache/incubator-mxnet/pull/15128#issuecomment-498822106 Thanks for your contributions @roywei . @mxnet-label-bot Add [pr-awaiting-review, Licenses] This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] piyushghai commented on issue #15124: [MXNET-1294] Priority-based parameter propagation for improved data parallel training throughput
piyushghai commented on issue #15124: [MXNET-1294] Priority-based parameter propagation for improved data parallel training throughput URL: https://github.com/apache/incubator-mxnet/pull/15124#issuecomment-498821764 Thanks for your contribution @anandj91. Can you look into the CI failures ? @mxnet-label-bot Add [pr-awaiting-review, Backend]. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] piyushghai commented on issue #15120: [bug] fix higher grad log
piyushghai commented on issue #15120: [bug] fix higher grad log URL: https://github.com/apache/incubator-mxnet/pull/15120#issuecomment-498820614 Thanks for your contributions @kshitij12345 @mxnet-label-bot Add [pr-awaiting-review, Operator] This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] abhinavs95 commented on issue #15145: Installation problem on XUbuntu
abhinavs95 commented on issue #15145: Installation problem on XUbuntu URL: https://github.com/apache/incubator-mxnet/issues/15145#issuecomment-498817090 @mxnet-label-bot add [Installation] This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] abhinavs95 commented on issue #15146: Installation problem on Windows 7
abhinavs95 commented on issue #15146: Installation problem on Windows 7 URL: https://github.com/apache/incubator-mxnet/issues/15146#issuecomment-498816128 @mxnet-label-bot add [Installation, Windows] This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] abhinavs95 commented on issue #15102: the grad of lars should be scaled in lbsgd
abhinavs95 commented on issue #15102: the grad of lars should be scaled in lbsgd URL: https://github.com/apache/incubator-mxnet/issues/15102#issuecomment-498813833 Hi @starimpact Could you provide some more info like a brief description of the problem with a minimum reproducible example? This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[incubator-mxnet-site] branch asf-site updated: Bump the publish timestamp.
This is an automated email from the ASF dual-hosted git repository. marcoabreu pushed a commit to branch asf-site in repository https://gitbox.apache.org/repos/asf/incubator-mxnet-site.git The following commit(s) were added to refs/heads/asf-site by this push: new 73a6877 Bump the publish timestamp. 73a6877 is described below commit 73a68770cedf91593b236c64b28ff51fca001b30 Author: mxnet-ci AuthorDate: Tue Jun 4 19:16:52 2019 + Bump the publish timestamp. --- date.txt | 1 + 1 file changed, 1 insertion(+) diff --git a/date.txt b/date.txt new file mode 100644 index 000..48f74bc --- /dev/null +++ b/date.txt @@ -0,0 +1 @@ +Tue Jun 4 19:16:52 UTC 2019
[GitHub] [incubator-mxnet] ddavydenko commented on issue #15148: Very Large CPU RAM Memory Consumption (>1GB)
ddavydenko commented on issue #15148: Very Large CPU RAM Memory Consumption (>1GB) URL: https://github.com/apache/incubator-mxnet/issues/15148#issuecomment-498803843 @mxnet-label-bot Add [Performance, Memory] This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] zachgk commented on a change in pull request #15128: update LICENSE
zachgk commented on a change in pull request #15128: update LICENSE
URL: https://github.com/apache/incubator-mxnet/pull/15128#discussion_r290430818
##
File path: tests/nightly/estimator/test_sentiment_rnn.py
##
@@ -101,7 +101,19 @@ def download_imdb(data_dir='/tmp/data'):
'''
Download and extract the IMDB dataset
'''
-url = ('http://ai.stanford.edu/~amaas/data/sentiment/aclImdb_v1.tar.gz')
+# dataset from http://ai.stanford.edu/~amaas/data/sentiment/
Review comment:
While it is good to include the full citation here, also add the information
on licensing and copyrights to the README or whatever docs people read which
tells them to download the data. The idea is that some of these licenses
actually have consequences. For example, we don't want to let commercial users
accidentally work with a non-commercial dataset. So, our goal is to make sure
that any time we inform users about a dataset, we also explain what legal
requirements come with that dataset as well.
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] zachgk commented on a change in pull request #15128: update LICENSE
zachgk commented on a change in pull request #15128: update LICENSE URL: https://github.com/apache/incubator-mxnet/pull/15128#discussion_r290428325 ## File path: LICENSE ## @@ -349,6 +350,19 @@ Copyright 2012 Continuum Analytics, Inc. + === +Creative Commons Attribution 4.0 International (CC BY 4.0) Review comment: Note that there are 6 different licenses as part of CC BY 4.0 (https://creativecommons.org/licenses/). It is important to know which one because some of them will prevent commercial usage, prevent derivative works, or require others to use the same license. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] zachgk commented on a change in pull request #15128: update LICENSE
zachgk commented on a change in pull request #15128: update LICENSE URL: https://github.com/apache/incubator-mxnet/pull/15128#discussion_r290424190 ## File path: LICENSE ## @@ -276,6 +276,7 @@ Copyright (c) 2015 by Contributors Copyright 1984, 1987, 1992 by Stephen L. Moshier +27. CNN Text Classification Example - For details, see example/cnn_text_classification/data_helpers.py Review comment: The LICENSE file is more specifically our source release license. It should only refer to things which are bundled as part of the source release (http://www.apache.org/dev/licensing-howto.html). Maybe we could move this to a separate DATASET_LICENSE file? This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] abhinavs95 opened a new pull request #15150: update dumps for const init
abhinavs95 opened a new pull request #15150: update dumps for const init URL: https://github.com/apache/incubator-mxnet/pull/15150 ## Description ## Fixes #12404 Override the dumps method for Constant initializer to take care of NDArray input. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] kshitij12345 commented on a change in pull request #14613: [MXNET-978] Higher order gradient support for some unary operators
kshitij12345 commented on a change in pull request #14613: [MXNET-978] Higher
order gradient support for some unary operators
URL: https://github.com/apache/incubator-mxnet/pull/14613#discussion_r290415647
##
File path: src/operator/tensor/elemwise_unary_op_basic.cc
##
@@ -85,8 +85,23 @@ The storage type of ``relu`` output depends upon the input
storage type:
)code" ADD_FILELINE)
.set_attr("FGradient", ElemwiseGradUseOut{"_backward_relu"});
-MXNET_OPERATOR_REGISTER_BINARY_WITH_SPARSE_CPU(_backward_relu,
-
unary_bwd);
+MXNET_OPERATOR_REGISTER_BINARY_WITH_SPARSE_CPU(_backward_relu,
unary_bwd)
+.set_attr("FGradient",
+[](const nnvm::NodePtr& n, const std::vector& ograds) {
+ std::vector ret;
+ // ograds[0]: d^2L/dx^2
+ // inputs[0]: dL/dy
+ // inputs[1]: y
+ // f(x) -> relu(x)
+ // f'(x) = 1 if x > 0 else 0
+ // f''(x) = 0
+ auto gx = nnvm::NodeEntry{n}; // f'(x)
+ ret.emplace_back(MakeNode("elemwise_mul", n->attrs.name +
"_backward_grad_grad",
+{ograds[0], gx}, nullptr, ));
Review comment:
Similar to what you have done below for `sin` and `cos`.
`gx` is actually `f'(x) * {head_grads/output_gradient}`.
It should actually only be `f'(x)`.
**Explanation** : `gx = f'(x) * head_grads`.
Therefore, `gx w.r.t. f'(x) = head_grads`,
Similarly `gx w.r.t. head_grads = f'(x)`.
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] kshitij12345 commented on a change in pull request #14613: [MXNET-978] Higher order gradient support for some unary operators
kshitij12345 commented on a change in pull request #14613: [MXNET-978] Higher
order gradient support for some unary operators
URL: https://github.com/apache/incubator-mxnet/pull/14613#discussion_r290415647
##
File path: src/operator/tensor/elemwise_unary_op_basic.cc
##
@@ -85,8 +85,23 @@ The storage type of ``relu`` output depends upon the input
storage type:
)code" ADD_FILELINE)
.set_attr("FGradient", ElemwiseGradUseOut{"_backward_relu"});
-MXNET_OPERATOR_REGISTER_BINARY_WITH_SPARSE_CPU(_backward_relu,
-
unary_bwd);
+MXNET_OPERATOR_REGISTER_BINARY_WITH_SPARSE_CPU(_backward_relu,
unary_bwd)
+.set_attr("FGradient",
+[](const nnvm::NodePtr& n, const std::vector& ograds) {
+ std::vector ret;
+ // ograds[0]: d^2L/dx^2
+ // inputs[0]: dL/dy
+ // inputs[1]: y
+ // f(x) -> relu(x)
+ // f'(x) = 1 if x > 0 else 0
+ // f''(x) = 0
+ auto gx = nnvm::NodeEntry{n}; // f'(x)
+ ret.emplace_back(MakeNode("elemwise_mul", n->attrs.name +
"_backward_grad_grad",
+{ograds[0], gx}, nullptr, ));
Review comment:
Similar to what you have done below for `sin` and `cos`.
`gx` is actually `f'(x) * head_grads/output_gradient`.
It should actually only be `f'(x)`.
**Explanation** : `gx = f'(x) * head_grads`.
Therefore, `gx w.r.t. f'(x) = head_grads`,
Similarly `gx w.r.t. head_grads = f'(x)`.
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] kshitij12345 commented on issue #14613: [MXNET-978] Higher order gradient support for some unary operators
kshitij12345 commented on issue #14613: [MXNET-978] Higher order gradient support for some unary operators URL: https://github.com/apache/incubator-mxnet/pull/14613#issuecomment-498775984 @apeforest Could you update the `check_second_order_unary` as per https://github.com/apache/incubator-mxnet/blob/37ce3b87268a8154f5c0ad97ce2522478038ee06/tests/python/unittest/test_higher_order_grad.py#L76-L103 This covers check for gradient of the first input argument as well. Have tested a similar Pytorch Script which works ( code in PR #15120 ). However do note that for PR #15120 , https://github.com/apache/incubator-mxnet/blob/37ce3b87268a8154f5c0ad97ce2522478038ee06/tests/python/unittest/test_higher_order_grad.py#L102 Assertion fails with `head_grads.grad.asnumpy()` being all `0's`. Please check to see if it works for you. Thank You. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] haojin2 opened a new pull request #15104: Numpy-compatible Concatenate
haojin2 opened a new pull request #15104: Numpy-compatible Concatenate URL: https://github.com/apache/incubator-mxnet/pull/15104 ## Description ## As title. ## Checklist ## ### Essentials ### Please feel free to remove inapplicable items for your PR. - [x] The PR title starts with [MXNET-$JIRA_ID], where $JIRA_ID refers to the relevant [JIRA issue](https://issues.apache.org/jira/projects/MXNET/issues) created (except PRs with tiny changes) - [x] Changes are complete (i.e. I finished coding on this PR) - [x] All changes have test coverage: - Unit tests are added for small changes to verify correctness (e.g. adding a new operator) - Nightly tests are added for complicated/long-running ones (e.g. changing distributed kvstore) - Build tests will be added for build configuration changes (e.g. adding a new build option with NCCL) - [x] Code is well-documented: - For user-facing API changes, API doc string has been updated. - For new C++ functions in header files, their functionalities and arguments are documented. - For new examples, README.md is added to explain the what the example does, the source of the dataset, expected performance on test set and reference to the original paper if applicable - Check the API doc at http://mxnet-ci-doc.s3-accelerate.dualstack.amazonaws.com/PR-$PR_ID/$BUILD_ID/index.html - [x] To the my best knowledge, examples are either not affected by this change, or have been fixed to be compatible with this change ### Changes ### - [x] Numpy-compatible concatenate - [x] Unit tests ## Comments ## This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] Zha0q1 opened a new pull request #15149: Fix Misspell in profiler
Zha0q1 opened a new pull request #15149: Fix Misspell in profiler URL: https://github.com/apache/incubator-mxnet/pull/15149 ## Description ## In python/mxnet/profiler.py parameter "continuous_dump" of function set_config() was spelled wrong as "contiguous_dump". This PR fixes it. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] haojin2 closed pull request #15104: Numpy-compatible Concatenate
haojin2 closed pull request #15104: Numpy-compatible Concatenate URL: https://github.com/apache/incubator-mxnet/pull/15104 This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] kshitij12345 commented on a change in pull request #14613: [MXNET-978] Higher order gradient support for some unary operators
kshitij12345 commented on a change in pull request #14613: [MXNET-978] Higher
order gradient support for some unary operators
URL: https://github.com/apache/incubator-mxnet/pull/14613#discussion_r290415647
##
File path: src/operator/tensor/elemwise_unary_op_basic.cc
##
@@ -85,8 +85,23 @@ The storage type of ``relu`` output depends upon the input
storage type:
)code" ADD_FILELINE)
.set_attr("FGradient", ElemwiseGradUseOut{"_backward_relu"});
-MXNET_OPERATOR_REGISTER_BINARY_WITH_SPARSE_CPU(_backward_relu,
-
unary_bwd);
+MXNET_OPERATOR_REGISTER_BINARY_WITH_SPARSE_CPU(_backward_relu,
unary_bwd)
+.set_attr("FGradient",
+[](const nnvm::NodePtr& n, const std::vector& ograds) {
+ std::vector ret;
+ // ograds[0]: d^2L/dx^2
+ // inputs[0]: dL/dy
+ // inputs[1]: y
+ // f(x) -> relu(x)
+ // f'(x) = 1 if x > 0 else 0
+ // f''(x) = 0
+ auto gx = nnvm::NodeEntry{n}; // f'(x)
+ ret.emplace_back(MakeNode("elemwise_mul", n->attrs.name +
"_backward_grad_grad",
+{ograds[0], gx}, nullptr, ));
Review comment:
`gx` is actually `f'(x) * head_grads/output_gradient`.
It should actually only be `f'(x)`.
Explanation : gx = `f'(x) * head_grads`.
Therefore, `gx w.r.t. f'(x) = head_grads`,
Similarly `gx w.r.t. head_grads = f'(x)`.
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] kshitij12345 commented on a change in pull request #14613: [MXNET-978] Higher order gradient support for some unary operators
kshitij12345 commented on a change in pull request #14613: [MXNET-978] Higher
order gradient support for some unary operators
URL: https://github.com/apache/incubator-mxnet/pull/14613#discussion_r290415647
##
File path: src/operator/tensor/elemwise_unary_op_basic.cc
##
@@ -85,8 +85,23 @@ The storage type of ``relu`` output depends upon the input
storage type:
)code" ADD_FILELINE)
.set_attr("FGradient", ElemwiseGradUseOut{"_backward_relu"});
-MXNET_OPERATOR_REGISTER_BINARY_WITH_SPARSE_CPU(_backward_relu,
-
unary_bwd);
+MXNET_OPERATOR_REGISTER_BINARY_WITH_SPARSE_CPU(_backward_relu,
unary_bwd)
+.set_attr("FGradient",
+[](const nnvm::NodePtr& n, const std::vector& ograds) {
+ std::vector ret;
+ // ograds[0]: d^2L/dx^2
+ // inputs[0]: dL/dy
+ // inputs[1]: y
+ // f(x) -> relu(x)
+ // f'(x) = 1 if x > 0 else 0
+ // f''(x) = 0
+ auto gx = nnvm::NodeEntry{n}; // f'(x)
+ ret.emplace_back(MakeNode("elemwise_mul", n->attrs.name +
"_backward_grad_grad",
+{ograds[0], gx}, nullptr, ));
Review comment:
`gx` is actually `f'(x) * head_grads/output_gradient`.
It should actually only be `f'(x)`.
**Explanation** : gx = `f'(x) * head_grads`.
Therefore, `gx w.r.t. f'(x) = head_grads`,
Similarly `gx w.r.t. head_grads = f'(x)`.
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] kshitij12345 commented on a change in pull request #14613: [MXNET-978] Higher order gradient support for some unary operators
kshitij12345 commented on a change in pull request #14613: [MXNET-978] Higher
order gradient support for some unary operators
URL: https://github.com/apache/incubator-mxnet/pull/14613#discussion_r290415647
##
File path: src/operator/tensor/elemwise_unary_op_basic.cc
##
@@ -85,8 +85,23 @@ The storage type of ``relu`` output depends upon the input
storage type:
)code" ADD_FILELINE)
.set_attr("FGradient", ElemwiseGradUseOut{"_backward_relu"});
-MXNET_OPERATOR_REGISTER_BINARY_WITH_SPARSE_CPU(_backward_relu,
-
unary_bwd);
+MXNET_OPERATOR_REGISTER_BINARY_WITH_SPARSE_CPU(_backward_relu,
unary_bwd)
+.set_attr("FGradient",
+[](const nnvm::NodePtr& n, const std::vector& ograds) {
+ std::vector ret;
+ // ograds[0]: d^2L/dx^2
+ // inputs[0]: dL/dy
+ // inputs[1]: y
+ // f(x) -> relu(x)
+ // f'(x) = 1 if x > 0 else 0
+ // f''(x) = 0
+ auto gx = nnvm::NodeEntry{n}; // f'(x)
+ ret.emplace_back(MakeNode("elemwise_mul", n->attrs.name +
"_backward_grad_grad",
+{ograds[0], gx}, nullptr, ));
Review comment:
`gx` is actually `f'(x) * head_grads/output_gradient`.
It should actually only be `f'(x)`.
**Explanation** : `gx = f'(x) * head_grads`.
Therefore, `gx w.r.t. f'(x) = head_grads`,
Similarly `gx w.r.t. head_grads = f'(x)`.
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] rvardimon commented on issue #14883: [Discussion] Overhead in MXNet Execution
rvardimon commented on issue #14883: [Discussion] Overhead in MXNet Execution URL: https://github.com/apache/incubator-mxnet/issues/14883#issuecomment-498759181 https://github.com/apache/incubator-mxnet/issues/15148 This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] mxnet-label-bot commented on issue #15148: Very Large CPU RAM Memory Consumption (>1GB)
mxnet-label-bot commented on issue #15148: Very Large CPU RAM Memory Consumption (>1GB) URL: https://github.com/apache/incubator-mxnet/issues/15148#issuecomment-498758716 Hey, this is the MXNet Label Bot. Thank you for submitting the issue! I will try and suggest some labels so that the appropriate MXNet community members can help resolve it. Here are my recommended labels: Performance This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] rvardimon opened a new issue #15148: Very Large CPU RAM Memory Consumption (>1GB)
rvardimon opened a new issue #15148: Very Large CPU RAM Memory Consumption
(>1GB)
URL: https://github.com/apache/incubator-mxnet/issues/15148
## Description
Mxnet consumes nearly 2GB CPU RAM even when loading a relatively small model
(e.g. Resnet-18) on directed on GPU (`ctx=mxnet.gpu()`). From what I
understand, there is no real need to load so much CPU memory, when the model is
running on GPU.
This issue is extremely prohibitive when trying to run multiple processes
with mxnet on the same machine, and IMO gives it a significant disadvantage
compared to other frameworks for being used in AI production systems.
## Environment info
```
/usr/bin/python3.6
/home/ran-face/src/CameraResearch/workspace/mxnet_diagnose.py
--Python Info--
Version : 3.6.7
Compiler : GCC 8.2.0
Build: ('default', 'Oct 22 2018 11:32:17')
Arch : ('64bit', 'ELF')
Pip Info---
Version : 9.0.1
Directory: /usr/lib/python3/dist-packages/pip
--MXNet Info---
Version : 1.3.1
Directory: /usr/local/lib/python3.6/dist-packages/mxnet
Commit Hash : 19c501680183237d52a862e6ae1dc4ddc296305b
--System Info--
Platform : Linux-4.15.0-45-generic-x86_64-with-Ubuntu-18.04-bionic
system : Linux
node : ranface-Lenovo-Y720-15IKB
release : 4.15.0-45-generic
version : #48-Ubuntu SMP Tue Jan 29 16:28:13 UTC 2019
--Hardware Info--
machine : x86_64
processor: x86_64
Architecture:x86_64
CPU op-mode(s): 32-bit, 64-bit
Byte Order: Little Endian
CPU(s): 8
On-line CPU(s) list: 0-7
Thread(s) per core: 2
Core(s) per socket: 4
Socket(s): 1
NUMA node(s):1
Vendor ID: GenuineIntel
CPU family: 6
Model: 158
Model name: Intel(R) Core(TM) i7-7700HQ CPU @ 2.80GHz
Stepping:9
CPU MHz: 2411.711
CPU max MHz: 3800.
CPU min MHz: 800.
BogoMIPS:5616.00
Virtualization: VT-x
L1d cache: 32K
L1i cache: 32K
L2 cache:256K
L3 cache:6144K
NUMA node0 CPU(s): 0-7
Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge
mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx
pdpe1gb rdtscp lm constant_tsc art arch_perfmon pebs bts rep_good nopl
xtopology nonstop_tsc cpuid aperfmperf tsc_known_freq pni pclmulqdq dtes64
monitor ds_cpl vmx est tm2 ssse3 sdbg fma cx16 xtpr pdcm pcid sse4_1 sse4_2
x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand lahf_lm abm
3dnowprefetch cpuid_fault invpcid_single pti ssbd ibrs ibpb stibp tpr_shadow
vnmi flexpriority ept vpid fsgsbase tsc_adjust bmi1 avx2 smep bmi2 erms invpcid
mpx rdseed adx smap clflushopt intel_pt xsaveopt xsavec xgetbv1 xsaves dtherm
ida arat pln pts hwp hwp_notify hwp_act_window hwp_epp flush_l1d
--Network Test--
Setting timeout: 10
Timing for MXNet: https://github.com/apache/incubator-mxnet, DNS: 0.0694
sec, LOAD: 0.9665 sec.
Timing for Gluon Tutorial(en): http://gluon.mxnet.io, DNS: 0.0841 sec, LOAD:
1.2019 sec.
Timing for Gluon Tutorial(cn): https://zh.gluon.ai, DNS: 0.0988 sec, LOAD:
0.9435 sec.
Timing for FashionMNIST:
https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/gluon/dataset/fashion-mnist/train-labels-idx1-ubyte.gz,
DNS: 0.0708 sec, LOAD: 1.2242 sec.
Timing for PYPI: https://pypi.python.org/pypi/pip, DNS: 0.0637 sec, LOAD:
1.3118 sec.
Timing for Conda: https://repo.continuum.io/pkgs/free/, DNS: 0.0499 sec,
LOAD: 0.3116 sec.
Process finished with exit code 0
```
Package used (Python/R/Scala/Julia):
I'm using Python3
## Build info
mxnet installed using pip3
## Steps to reproduce
(Paste the commands you ran that produced the error.)
1. Run the following code
```
import mxnet as mx
import time
if __name__ == '__main__':
path='http://data.mxnet.io/models/imagenet/'
[mx.test_utils.download(path+'resnet/18-layers/resnet-18-.params'),
mx.test_utils.download(path+'resnet/18-layers/resnet-18-symbol.json'),
mx.test_utils.download(path+'synset.txt')]
ctx = mx.gpu()
sym, arg_params, aux_params = mx.model.load_checkpoint('resnet-18', 0)
mod = mx.mod.Module(symbol=sym, context=ctx, label_names=None)
mod.bind(for_training=False, data_shapes=[('data', (1,3,224,224))],
label_shapes=mod._label_shapes)
mod.set_params(arg_params, aux_params, allow_missing=True)
time.sleep(100)
```
2. check process memory (run top, shift M to sort processes by memory usage)
3. memory usage is about ~1.5-2GB RAM
[GitHub] [incubator-mxnet] apeforest commented on issue #14613: [MXNET-978] Higher order gradient support for some unary operators
apeforest commented on issue #14613: [MXNET-978] Higher order gradient support for some unary operators URL: https://github.com/apache/incubator-mxnet/pull/14613#issuecomment-498758177 @kshitij12345 could you please take a look at the PR again. Thanks This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] roywei commented on issue #15128: update LICENSE
roywei commented on issue #15128: update LICENSE URL: https://github.com/apache/incubator-mxnet/pull/15128#issuecomment-498755613 @zachgk @lanking520 could you help review the license? Thanks! This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] roywei commented on issue #14981: [CI][NightlyTestsForBinaries] Test Large Tensor: GPU Failing
roywei commented on issue #14981: [CI][NightlyTestsForBinaries] Test Large Tensor: GPU Failing URL: https://github.com/apache/incubator-mxnet/issues/14981#issuecomment-498748546 Currently, both CPU and GPU tests have been disabled due to the same memory issue. Had a discussion with @access2rohit and @apeforest, we can try a few things: 1. change to P3 instances here https://github.com/apache/incubator-mxnet/blob/master/tests/nightly/JenkinsfileForBinaries#L82 2. further increase shared memory to 50G 3. stop running large tensor test parallelly with other tests. We are having problems testing the above solutions on CI machines that have multiple jobs running in parallel. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] david-seiler commented on a change in pull request #14617: PDF operators for the random samplers, and also the Dirichlet
david-seiler commented on a change in pull request #14617: PDF operators for
the random samplers, and also the Dirichlet
URL: https://github.com/apache/incubator-mxnet/pull/14617#discussion_r290353676
##
File path: ci/windows/test_py3_cpu.ps1
##
@@ -24,7 +24,7 @@ $env:MXNET_HOME=[io.path]::combine($PSScriptRoot,
'mxnet_home')
C:\Python37\Scripts\pip install -r tests\requirements.txt
C:\Python37\python.exe -m nose -v --with-timer --timer-ok 1 --timer-warning 15
--timer-filter warning,error --with-xunit --xunit-file nosetests_unittest.xml
tests\python\unittest
-if (! $?) { Throw ("Error running unittest") }
+if (! $?) { Throw ("Error running unittest) }
Review comment:
whoops, I was factoring some error-handling code out to PR-15147 and got a
little too aggressive. Good catch, fixed now.
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
[incubator-mxnet] branch master updated: fix nightly (#15141)
This is an automated email from the ASF dual-hosted git repository.
skm pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git
The following commit(s) were added to refs/heads/master by this push:
new 134a3e8 fix nightly (#15141)
134a3e8 is described below
commit 134a3e8cd36ee66426deedd3c8add6888378c043
Author: Lai Wei
AuthorDate: Tue Jun 4 07:43:49 2019 -0700
fix nightly (#15141)
* fix nightly
* disable large tensor
* update issue link
---
tests/nightly/JenkinsfileForBinaries | 9 +
1 file changed, 5 insertions(+), 4 deletions(-)
diff --git a/tests/nightly/JenkinsfileForBinaries
b/tests/nightly/JenkinsfileForBinaries
index e4b9ff1..d5f1ebd 100755
--- a/tests/nightly/JenkinsfileForBinaries
+++ b/tests/nightly/JenkinsfileForBinaries
@@ -86,14 +86,15 @@ core_logic: {
}
}
},*/
-'Test Large Tensor Size: GPU': {
+// https://github.com/apache/incubator-mxnet/issues/14981
+/*'Test Large Tensor Size: GPU': {
node(NODE_LINUX_GPU) {
ws('workspace/large_tensor-gpu') {
utils.unpack_and_init('gpu_int64', mx_cmake_lib)
utils.docker_run('ubuntu_nightly_gpu',
'nightly_test_large_tensor', true)
}
}
-},
+},*/
'StraightDope: Python2 Single-GPU': {
node(NODE_LINUX_GPU_P3) {
ws('workspace/straight_dope-single_gpu') {
@@ -127,7 +128,7 @@ core_logic: {
}
},
'Tutorial: Python2': {
- node(NODE_LINUX_GPU) {
+ node(NODE_LINUX_GPU_P3) {
ws('workspace/tutorial-test-python2') {
utils.unpack_and_init('gpu', mx_lib)
utils.docker_run('ubuntu_nightly_gpu',
'nightly_tutorial_test_ubuntu_python2_gpu', true, '1500m')
@@ -135,7 +136,7 @@ core_logic: {
}
},
'Tutorial: Python3': {
- node(NODE_LINUX_GPU) {
+ node(NODE_LINUX_GPU_P3) {
ws('workspace/tutorial-test-python3') {
utils.unpack_and_init('gpu', mx_lib)
utils.docker_run('ubuntu_nightly_gpu',
'nightly_tutorial_test_ubuntu_python3_gpu', true, '1500m')
[GitHub] [incubator-mxnet] sandeep-krishnamurthy merged pull request #15141: fix nightly
sandeep-krishnamurthy merged pull request #15141: fix nightly URL: https://github.com/apache/incubator-mxnet/pull/15141 This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] lebeg commented on a change in pull request #14617: PDF operators for the random samplers, and also the Dirichlet
lebeg commented on a change in pull request #14617: PDF operators for the
random samplers, and also the Dirichlet
URL: https://github.com/apache/incubator-mxnet/pull/14617#discussion_r290332802
##
File path: ci/windows/test_py3_cpu.ps1
##
@@ -24,7 +24,7 @@ $env:MXNET_HOME=[io.path]::combine($PSScriptRoot,
'mxnet_home')
C:\Python37\Scripts\pip install -r tests\requirements.txt
C:\Python37\python.exe -m nose -v --with-timer --timer-ok 1 --timer-warning 15
--timer-filter warning,error --with-xunit --xunit-file nosetests_unittest.xml
tests\python\unittest
-if (! $?) { Throw ("Error running unittest") }
+if (! $?) { Throw ("Error running unittest) }
Review comment:
Are you sure you want to remove the `"`?
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
[GitHub] [incubator-mxnet] david-seiler opened a new pull request #15147: Python test failures under windows now report the exit code
david-seiler opened a new pull request #15147: Python test failures under windows now report the exit code URL: https://github.com/apache/incubator-mxnet/pull/15147 ## Description ## While working on PR-14617, I hit an obscure Windows-only crash. Adding the $LastExitCode to the error message helped me debug it; a hex-formatted version of that change appears here. ## Checklist ## ### Essentials ### - [x] Changes are complete (i.e. I finished coding on this PR) - [ ] All changes have test coverage: - The only changes are on the error paths for ci/windows/test_*.ps1, so the PR doesn't test all that well. In the first version I've deliberately broken test_py2_gpu.ps1 (see line 27), so we should see that test fail and report an exit code of 0x0. Assuming that behaves as expected, I'll amend the PR to make it pass. ## Comments ## - No rush to merge this, I just promised Marco I'd give it its own PR. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] pengzhao-intel commented on issue #15108: The test time of the model on GPU is normal, but the test time on CPU is very long.
pengzhao-intel commented on issue #15108: The test time of the model on GPU is normal, but the test time on CPU is very long. URL: https://github.com/apache/incubator-mxnet/issues/15108#issuecomment-498676696 It makes sense because your input image size is about 10X in both H and W direction from 12 to 112. So the runtime is increased. Btw, the no mkl build will be very slow and it's not performant. I suggest you change to MKLDNN build for the start point. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[incubator-mxnet] branch master updated: [clojure] clojurify function names in image.clj namespace (#15121)
This is an automated email from the ASF dual-hosted git repository.
kedarb pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git
The following commit(s) were added to refs/heads/master by this push:
new 28c528e [clojure] clojurify function names in image.clj namespace
(#15121)
28c528e is described below
commit 28c528e16be70b31287b65d949205396bbfec6e8
Author: Arthur Caillau
AuthorDate: Tue Jun 4 15:37:04 2019 +0200
[clojure] clojurify function names in image.clj namespace (#15121)
* [clojure] clojurify function names in image.clj namespace
* move deprecated to the proper location for defn
* rename color-flag to color and use :color :grayscale as values
* add rm dest-path in with-file
* change `color-flag` to `color` in `color->int`
---
.../src/org/apache/clojure_mxnet/image.clj | 116 +++--
.../test/org/apache/clojure_mxnet/image_test.clj | 63 +++
2 files changed, 151 insertions(+), 28 deletions(-)
diff --git a/contrib/clojure-package/src/org/apache/clojure_mxnet/image.clj
b/contrib/clojure-package/src/org/apache/clojure_mxnet/image.clj
index f81a358..68dcbfe 100644
--- a/contrib/clojure-package/src/org/apache/clojure_mxnet/image.clj
+++ b/contrib/clojure-package/src/org/apache/clojure_mxnet/image.clj
@@ -17,6 +17,7 @@
(ns org.apache.clojure-mxnet.image
"Image API of Clojure package."
+ (:refer-clojure :exclude [read])
(:require [t6.from-scala.core :refer [$ $$] :as $]
[org.apache.clojure-mxnet.dtype :as dtype]
[org.apache.clojure-mxnet.ndarray :as ndarray]
@@ -38,8 +39,10 @@
(s/def ::decode-image-opts
(s/keys :opt-un [::color-flag ::to-rgb ::output]))
-(defn decode-image
- "Decodes an image from an input stream with OpenCV
+(defn ^:deprecated decode-image
+ "DEPRECATED: use `decode` instead.
+
+ Decodes an image from an input stream with OpenCV
`input-stream`: `InputStream` - Contains the binary encoded image
`color-flag`: 0 or 1 - Convert decoded image to grayscale (0) or color (1)
`to-rgb`: boolean - Whether to convert decoded image to mxnet's default RGB
@@ -60,14 +63,47 @@
([input-stream]
(decode-image input-stream {})))
+(s/def ::color #{:grayscale :color})
+(s/def ::decode-image-opts-2 (s/keys :opt-un [::color ::to-rgb ::output]))
+
+(defn- color->int [color]
+ (case color
+:grayscale 0
+:color 1))
+
+(defn decode
+ "Decodes an image from an input stream with OpenCV.
+`input-stream`: `InputStream` - Contains the binary encoded image
+`color`: keyword in `#{:color :grayscale}` - Convert decoded image to
+ grayscale or color
+`to-rgb`: boolean - Whether to convert decoded image to mxnet's default RGB
+format (instead of opencv's default BGR)
+`output`: nil or `NDArray`
+returns: `NDArray` with dtype uint8
+
+ Ex:
+(decode input-stream)
+(decode input-stream {:color :color})
+(decode input-stream {:color :grayscale :output nd})"
+ ([input-stream {:keys [color to-rgb output]
+ :or {color :color to-rgb true output nil}
+ :as opts}]
+ (util/validate! ::input-stream input-stream "Invalid input stream")
+ (util/validate! ::decode-image-opts-2 opts "Invalid options for decoding")
+ (Image/imDecode input-stream (color->int color) to-rgb ($/option output)))
+ ([input-stream]
+ (decode input-stream {})))
+
(s/def ::filename string?)
(s/def ::optional-color-flag
(s/or :none nil? :some ::color-flag))
(s/def ::optional-to-rgb
(s/or :none nil? :some ::to-rgb))
-(defn read-image
- "Reads an image file and returns an ndarray with OpenCV. It returns image in
+(defn ^:deprecated read-image
+ "DEPRECATED: use `read` instead.
+
+ Reads an image file and returns an ndarray with OpenCV. It returns image in
RGB by default instead of OpenCV's default BGR.
`filename`: string - Name of the image file to be loaded
`color-flag`: 0 or 1 - Convert decoded image to grayscale (0) or color (1)
@@ -95,11 +131,43 @@
([filename]
(read-image filename {})))
+(defn read
+ "Reads an image file and returns an ndarray with OpenCV. It returns image in
+ RGB by default instead of OpenCV's default BGR.
+`filename`: string - Name of the image file to be loaded
+`color`: keyword in `#{:color :grayscale}` - Convert decoded image to
+ grayscale or color
+`to-rgb`: boolean - Whether to convert decoded image to mxnet's default RGB
+format (instead of opencv's default BGR)
+`output`: nil or `NDArray`
+returns: `NDArray` with dtype uint8
+
+ Ex:
+ (read \"cat.jpg\")
+ (read \"cat.jpg\" {:color :grayscale})
+ (read \"cat.jpg\" {:color :color :output nd})"
+ ([filename {:keys [color to-rgb output]
+ :or {color :color to-rgb nil output nil}
+ :as opts}]
+ (util/validate! ::filename filename "Invalid
[GitHub] [incubator-mxnet] kedarbellare merged pull request #15121: [clojure] clojurify function names in image.clj namespace
kedarbellare merged pull request #15121: [clojure] clojurify function names in image.clj namespace URL: https://github.com/apache/incubator-mxnet/pull/15121 This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[incubator-mxnet-site] branch asf-site updated: Bump the publish timestamp.
This is an automated email from the ASF dual-hosted git repository. marcoabreu pushed a commit to branch asf-site in repository https://gitbox.apache.org/repos/asf/incubator-mxnet-site.git The following commit(s) were added to refs/heads/asf-site by this push: new f72f94f Bump the publish timestamp. f72f94f is described below commit f72f94f9a7caa44615ea424e0a0ee284b483e4bc Author: mxnet-ci AuthorDate: Tue Jun 4 13:16:06 2019 + Bump the publish timestamp. --- date.txt | 1 + 1 file changed, 1 insertion(+) diff --git a/date.txt b/date.txt new file mode 100644 index 000..9eb57a0 --- /dev/null +++ b/date.txt @@ -0,0 +1 @@ +Tue Jun 4 13:16:05 UTC 2019
[GitHub] [incubator-mxnet] junrushao1994 commented on issue #15143: dmlc::type_name_helper specialization of mxnet::tuple should not be disabled for MSVC
junrushao1994 commented on issue #15143: dmlc::type_name_helper specialization of mxnet::tuple should not be disabled for MSVC URL: https://github.com/apache/incubator-mxnet/issues/15143#issuecomment-498638720 (CC: @reminisce) Thanks for bringing this up! Please be aware that I am not the original author of the file, and I am not super familiar with MSVC. @reminisce Is there any reason that we disable this? This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
[GitHub] [incubator-mxnet] mxnet-label-bot commented on issue #15146: Installation problem on Windows 7
mxnet-label-bot commented on issue #15146: Installation problem on Windows 7 URL: https://github.com/apache/incubator-mxnet/issues/15146#issuecomment-498627824 Hey, this is the MXNet Label Bot. Thank you for submitting the issue! I will try and suggest some labels so that the appropriate MXNet community members can help resolve it. Here are my recommended labels: Installation This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
