Migrate masked_scatter_ CPU to ATen by kshitij12345 · Pull Request #49732 · pytorch/pytorch

kshitij12345 · 2020-12-22T09:52:35Z

Fixes #49541

Reference: #24507

facebook-github-bot · 2020-12-22T09:52:57Z

💊 CI failures summary and remediations

As of commit cc003e6 (more details on the Dr. CI page):

1/1 failures introduced in this PR

🕵️ 1 new failure recognized by patterns

The following CI failures do not appear to be due to upstream breakages:

pytorch_linux_bionic_py3_8_gcc9_coverage_test1 (1/1)

Step: "Run tests" (full log | diagnosis details | 🔁 rerun)

Dec 22 16:47:39 [E request_callback_no_python.cpp:636] Received error while processing request type 258: RuntimeError: Can not pickle torch.futures.Future

Dec 22 16:47:39 At:
Dec 22 16:47:39   /opt/conda/lib/python3.8/site-packages/torch/distributed/rpc/internal.py(120): serialize
Dec 22 16:47:39   /opt/conda/lib/python3.8/site-packages/torch/distributed/rpc/internal.py(172): serialize
Dec 22 16:47:39 
Dec 22 16:47:39 [E request_callback_no_python.cpp:636] Received error while processing request type 258: RuntimeError: Can not pickle torch.futures.Future
Dec 22 16:47:39 
Dec 22 16:47:39 At:
Dec 22 16:47:39   /opt/conda/lib/python3.8/site-packages/torch/distributed/rpc/internal.py(120): serialize
Dec 22 16:47:39   /opt/conda/lib/python3.8/site-packages/torch/distributed/rpc/internal.py(172): serialize
Dec 22 16:47:39 
Dec 22 16:47:39 [E request_callback_no_python.cpp:636] Received error while processing request type 258: RuntimeError: Can not pickle torch.futures.Future
Dec 22 16:47:39 
Dec 22 16:47:39 At:
Dec 22 16:47:39   /opt/conda/lib/python3.8/site-packages/torch/distributed/rpc/internal.py(120): serialize
Dec 22 16:47:39   /opt/conda/lib/python3.8/site-packages/torch/distributed/rpc/internal.py(172): serialize
Dec 22 16:47:39 
Dec 22 16:47:39 [W tensorpipe_agent.cpp:547] RPC agent for worker3 encountered error when reading incoming request from worker2: EOF: end of file (this is expected to happen during shutdown)
Dec 22 16:47:39 [W tensorpipe_agent.cpp:547] RPC agent for worker1 encountered error when reading incoming request from worker0: EOF: end of file (this is expected to happen during shutdown)
Dec 22 16:47:40 ok (2.445s)
Dec 22 16:47:42   test_return_future_remote (__main__.TensorPipeRpcTestWithSpawn) ... [W tensorpipe_agent.cpp:547] RPC agent for worker2 encountered error when reading incoming request from worker0: EOF: end of file (this is expected to happen during shutdown)
Dec 22 16:47:42 [W tensorpipe_agent.cpp:547] RPC agent for worker1 encountered error when reading incoming request from worker0: EOF: end of file (this is expected to happen during shutdown)

1 job timed out:

pytorch_linux_bionic_py3_8_gcc9_coverage_test1

This comment was automatically generated by Dr. CI (expand for details).

Follow this link to opt-out of these comments for your Pull Requests.

Please report bugs/suggestions to the (internal) Dr. CI Users group.

This comment has been revised 9 times.

kshitij12345 · 2020-12-22T09:59:34Z

Benchmark

Before PR

10 / 10[ masked_scatter_ mdtype torch.uint8 input dtypetorch.int32 ]
                          |  torch.masked_scatter_ CPU
1 threads: -------------------------------------------
      (32,)               |             14.5          
      (32, 32)            |             17.3          
      (2, 16, 32)         |             17.0          
      (2, 16, 32, 32)     |            119.5          
      (4, 2, 16, 32, 32)  |            521.5          

Times are in microseconds (us).

[ masked_scatter_ mdtype torch.uint8 input dtypetorch.float32 ]
                          |  torch.masked_scatter_ CPU
1 threads: -------------------------------------------
      (32,)               |             14.4          
      (32, 32)            |             16.9          
      (2, 16, 32)         |             16.9          
      (2, 16, 32, 32)     |            121.2          
      (4, 2, 16, 32, 32)  |            543.1          

Times are in microseconds (us).

[ masked_scatter_ mdtype torch.uint8 input dtypetorch.float64 ]
                          |  torch.masked_scatter_ CPU
1 threads: -------------------------------------------
      (32,)               |             15.9          
      (32, 32)            |             20.5          
      (2, 16, 32)         |             20.7          
      (2, 16, 32, 32)     |            125.9          
      (4, 2, 16, 32, 32)  |            498.8          

Times are in microseconds (us).

[ masked_scatter_ mdtype torch.bool input dtypetorch.int32 ]
                          |  torch.masked_scatter_ CPU
1 threads: -------------------------------------------
      (32,)               |              881.3        
      (32, 32)            |             2982.5        
      (2, 16, 32)         |             3074.9        
      (2, 16, 32, 32)     |            99410.4        
      (4, 2, 16, 32, 32)  |           461798.1        

Times are in nanoseconds (ns).

[ masked_scatter_ mdtype torch.bool input dtypetorch.float32 ]
                          |  torch.masked_scatter_ CPU
1 threads: -------------------------------------------
      (32,)               |              873.4        
      (32, 32)            |             2748.0        
      (2, 16, 32)         |             3133.0        
      (2, 16, 32, 32)     |           100659.8        
      (4, 2, 16, 32, 32)  |           471495.6        

Times are in nanoseconds (ns).

[ masked_scatter_ mdtype torch.bool input dtypetorch.float64 ]
                          |  torch.masked_scatter_ CPU
1 threads: -------------------------------------------
      (32,)               |              876.7        
      (32, 32)            |             3169.1        
      (2, 16, 32)         |             3235.5        
      (2, 16, 32, 32)     |            99530.6        
      (4, 2, 16, 32, 32)  |           461690.9        

Times are in nanoseconds (ns).

After PR

10 / 10[ masked_scatter_ mdtype torch.uint8 input dtypetorch.int32 ]
                          |  torch.masked_scatter_ CPU
1 threads: -------------------------------------------
      (32,)               |             14.6          
      (32, 32)            |             17.0          
      (2, 16, 32)         |             16.9          
      (2, 16, 32, 32)     |            135.8          
      (4, 2, 16, 32, 32)  |            523.4          

Times are in microseconds (us).

[ masked_scatter_ mdtype torch.uint8 input dtypetorch.float32 ]
                          |  torch.masked_scatter_ CPU
1 threads: -------------------------------------------
      (32,)               |             14.1          
      (32, 32)            |             16.2          
      (2, 16, 32)         |             16.4          
      (2, 16, 32, 32)     |            133.4          
      (4, 2, 16, 32, 32)  |            493.5          

Times are in microseconds (us).

[ masked_scatter_ mdtype torch.uint8 input dtypetorch.float64 ]
                          |  torch.masked_scatter_ CPU
1 threads: -------------------------------------------
      (32,)               |             12.8          
      (32, 32)            |             15.3          
      (2, 16, 32)         |             16.6          
      (2, 16, 32, 32)     |            134.2          
      (4, 2, 16, 32, 32)  |            515.6          

Times are in microseconds (us).

[ masked_scatter_ mdtype torch.bool input dtypetorch.int32 ]
                          |  torch.masked_scatter_ CPU
1 threads: -------------------------------------------
      (32,)               |              808.3        
      (32, 32)            |             2136.1        
      (2, 16, 32)         |             2017.6        
      (2, 16, 32, 32)     |            98196.0        
      (4, 2, 16, 32, 32)  |           445502.8        

Times are in nanoseconds (ns).

[ masked_scatter_ mdtype torch.bool input dtypetorch.float32 ]
                          |  torch.masked_scatter_ CPU
1 threads: -------------------------------------------
      (32,)               |              796.5        
      (32, 32)            |             3719.5        
      (2, 16, 32)         |             3760.2        
      (2, 16, 32, 32)     |           116965.8        
      (4, 2, 16, 32, 32)  |           461250.2        

Times are in nanoseconds (ns).

[ masked_scatter_ mdtype torch.bool input dtypetorch.float64 ]
                          |  torch.masked_scatter_ CPU
1 threads: -------------------------------------------
      (32,)               |              792.5        
      (32, 32)            |             2102.3        
      (2, 16, 32)         |             2096.0        
      (2, 16, 32, 32)     |            97564.0        
      (4, 2, 16, 32, 32)  |           435919.5        

Times are in nanoseconds (ns).

Code

Benchmarking Code

import torch
import itertools
import time
from torch.utils.benchmark import Timer
from torch.utils.benchmark import Compare
import sys

print('Using pytorch %s' % (torch.__version__))

shapes = [(32,), (32, 32), (2, 16, 32), (2, 16, 32, 32), (4, 2, 16, 32, 32)]
results = []
repeats = 10

for m_dtype in [torch.uint8, torch.bool]:
    for dtype in [torch.int32, torch.float, torch.double]:
        for mat1_shape in shapes:
            mat1 = torch.zeros(*mat1_shape, dtype=dtype, device='cpu')
            mask = (torch.randn(mat1.shape) > 0).to(m_dtype)
            source = torch.ones(mask.sum(), dtype=dtype, device='cpu')

            tasks = [("mat1.masked_scatter_(mask, source)", "torch.masked_scatter_ CPU")]
            timers = [Timer(stmt=stmt, label=f"masked_scatter_ mdtype {m_dtype} input dtype{dtype}", sub_label=f"{(mat1_shape)}", description=label, globals=globals()) for stmt, label in tasks]

            for i, timer in enumerate(timers * repeats):
                results.append(
                    timer.blocked_autorange()
                )
                print(f"\r{i + 1} / {len(timers) * repeats}", end="")
                sys.stdout.flush()
comparison = Compare(results)
comparison.print()

kshitij12345 · 2020-12-23T04:40:39Z

@VitalyFedyunin Please review

kshitij12345

I don't think the CUDA variant is tested.

pytorch/test/test_torch.py

Lines 1198 to 1239 in 963f762

    
           def test_masked_scatter(self): 
        
               with warnings.catch_warnings(record=True) as w: 
        
                   warnings.simplefilter("always") 
        
                   for maskType in [torch.uint8, torch.bool]: 
        
                       for dt in torch.testing.get_all_dtypes(): 
        
                           num_copy, num_dest = 3, 10 
        
                           dest = torch.tensor([1, 2, 3, 4, 5, 6, 7, 8, 9, 10], dtype=dt) 
        
                           dest2 = dest.clone() 
        
                           src = torch.tensor([0, 0, 0, 0, 0, 0, 0, 0, 0, 0], dtype=dt) 
        
                           mask = torch.tensor((0, 0, 0, 0, 1, 0, 1, 0, 1, 0), dtype=maskType) 
        
                           if dt == torch.bool: 
        
                               # torch.bool is a special case and is being tested 
        
                               # in a separate test 
        
                               continue 
        
                           # TODO: update test when masked scatter is supported for complex 
        
                           if dt == torch.half or dt.is_complex: 
        
                               self.assertRaises(RuntimeError, lambda: dest.masked_scatter_(mask, src)) 
        
                               continue 
        
                           dest.masked_scatter_(mask, src) 
        
                           j = 0 
        
                           for i in range(num_dest): 
        
                               if mask[i]: 
        
                                   dest2[i] = src[j] 
        
                                   j += 1 
        
                           self.assertEqual(dest, dest2, atol=0, rtol=0) 
        
                           # make source bigger than number of 1s in mask 
        
                           src = torch.tensor([1, 1, 1, 1, 1, 1, 1, 1, 1, 1], dtype=dt) 
        
                           dest.masked_scatter_(mask, src) 
        
                           # make src smaller. this should fail 
        
                           src = torch.randn(num_copy - 1) 
        
                           with self.assertRaises(RuntimeError): 
        
                               dest.masked_scatter_(mask, src) 
        
               self.assertEqual(len(w), 27) 
        
               warn = 'masked_scatter_ received a mask with dtype torch.uint8,' 
        
               for wi in w: 
        
                   self.assertEqual(str(wi.message)[0:55], str(warn))

kshitij12345 · 2020-12-28T10:02:27Z

                        with self.assertRaises(RuntimeError):
                            dest.masked_scatter_(mask, src)
-            self.assertEqual(len(w), 27)
+            self.assertEqual(len(w), 20)


I don't think the CUDA variant is tested.

pytorch/test/test_torch.py

Lines 1198 to 1239 in 963f762

def test_masked_scatter(self):

with warnings.catch_warnings(record=True) as w:

warnings.simplefilter("always")

for maskType in [torch.uint8, torch.bool]:

for dt in torch.testing.get_all_dtypes():

num_copy, num_dest = 3, 10

dest = torch.tensor([1, 2, 3, 4, 5, 6, 7, 8, 9, 10], dtype=dt)

dest2 = dest.clone()

src = torch.tensor([0, 0, 0, 0, 0, 0, 0, 0, 0, 0], dtype=dt)

mask = torch.tensor((0, 0, 0, 0, 1, 0, 1, 0, 1, 0), dtype=maskType)

if dt == torch.bool:

# torch.bool is a special case and is being tested

# in a separate test

continue

# TODO: update test when masked scatter is supported for complex

if dt == torch.half or dt.is_complex:

self.assertRaises(RuntimeError, lambda: dest.masked_scatter_(mask, src))

continue

dest.masked_scatter_(mask, src)

j = 0

for i in range(num_dest):

if mask[i]:

dest2[i] = src[j]

j += 1

self.assertEqual(dest, dest2, atol=0, rtol=0)

# make source bigger than number of 1s in mask

src = torch.tensor([1, 1, 1, 1, 1, 1, 1, 1, 1, 1], dtype=dt)

dest.masked_scatter_(mask, src)

# make src smaller. this should fail

src = torch.randn(num_copy - 1)

with self.assertRaises(RuntimeError):

dest.masked_scatter_(mask, src)

self.assertEqual(len(w), 27)

warn = 'masked_scatter_ received a mask with dtype torch.uint8,'

for wi in w:

self.assertEqual(str(wi.message)[0:55], str(warn))

Other tests for masked_scatter,

pytorch/test/test_torch.py

Lines 4590 to 4600 in 963f762

def test_masked_scatter_bool_tensor(self, device):

src = torch.tensor([True, True, True], device=device)

dst = torch.tensor([False, False, False], device=device)

mask = torch.tensor([False, True, False], device=device)

dst.masked_scatter_(mask, src)

self.assertEqual(dst, torch.tensor([False, True, False], device=device))

mask = torch.tensor([True, False, True], device=device)

dst = dst.masked_scatter(mask, src)

self.assertEqual(dst, torch.tensor([True, True, True], device=device))

pytorch/test/test_torch.py

Lines 5129 to 5144 in 963f762

@onlyOnCPUAndCUDA

def test_masked_scatter_mem_overlap(self, device):

x = torch.rand((1,), device=device).expand((6,))

src = torch.rand((3,), device=device)

mask = torch.tensor([True, False, True, True, False, False], device=device)

with self.assertRaisesRegex(RuntimeError, 'unsupported operation'):

x.masked_scatter_(mask, src)

@onlyOnCPUAndCUDA

def test_index_select_mem_overlap(self, device):

x = torch.rand((1, 6), device=device).expand((2, 6))

y = torch.rand((3, 6), device=device)

ind = torch.tensor([0, 1], dtype=torch.int64, device=device)

with self.assertRaisesRegex(RuntimeError, 'unsupported operation'):

torch.index_select(y, 1, ind, out=x)

kshitij12345 · 2020-12-30T10:17:49Z

@ngimel Please review

ngimel

This looks good, I left minor comments. For the future, you can serialize benchmark results and then load and compare them, that will make looking at the results easier.

ngimel · 2021-01-18T19:12:26Z

+            "please use a mask with dtype torch.bool instead.");
+  }
+
+  TORCH_CHECK(self.numel() == b_mask.numel(), "Number of elements of self != Number of elements in mask");


why is this check needed? expand_inplace should have failed if it could not satisfy this invariant?

Right. This is redundant. Thanks!

ngimel · 2021-01-18T19:16:21Z

+    for (int64_t i = 0; i < n; i++) {
+      mask_t mask_value = *(mask_t*)(mask + mask_stride * i);
+      if (!is_mask_bool) {
+        TORCH_CHECK(mask_value == 0 || mask_value == 1, "Mask tensor can take 0 and 1 values only");


mask_value <= 1

ngimel · 2021-01-18T19:25:53Z

                        with self.assertRaises(RuntimeError):
                            dest.masked_scatter_(mask, src)
-            self.assertEqual(len(w), 27)
+            self.assertEqual(len(w), 20)


Yeah, cuda variant is not tested here. Why did the number of warnings change?

This happens because I think the test has bug.

For the snippet below,
src is always is of dtype=torch.float, while dtype of dest varies as per the loop.

# make src smaller. this should fail src = torch.randn(num_copy - 1) with self.assertRaises(RuntimeError): dest.masked_scatter_(mask, src)

So except for dest and src having dtype torch.float32, all other cases fail with Error due to mismatch in dtypes of dest and src.

In the new implementation, we first check if the types are src and dest are same and proceed with the warning only then. However, in the old implementation, we check warning condition first and then check for types of src and dest.

That is why the discrepancy.

pytorch/aten/src/ATen/native/TensorAdvancedIndexing.cpp

Lines 1124 to 1141 in cc003e6

TORCH_CHECK(

self.scalar_type() == source.scalar_type(),

"masked_scatter: expected self and source to have same dtypes but got",

self.scalar_type(),

" and ",

source.scalar_type());

TORCH_CHECK(self.device().type() == at::kCPU, "device type of self (", self.device().type(), ") is not CPU");

TORCH_CHECK(mask.device().type() == at::kCPU, "device type of mask (", mask.device().type(), ") is not CPU");

TORCH_CHECK(source.device().type() == at::kCPU, "device type of source (", source.device().type(), ") is not CPU");

Tensor b_mask;

std::tie(b_mask) = expand_inplace(self, mask, "masked_scatter_");

if (b_mask.dtype() == ScalarType::Byte) {

TORCH_WARN("masked_scatter_ received a mask with dtype torch.uint8, this behavior is now deprecated," \

"please use a mask with dtype torch.bool instead.");

}

Actually will fix the bug.
Thanks!

codecov · 2021-01-19T09:25:18Z

Codecov Report

Merging #49732 (5653227) into master (ce30dba) will decrease coverage by 0.00%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master   #49732      +/-   ##
==========================================
- Coverage   80.65%   80.64%   -0.01%     
==========================================
  Files        1913     1913              
  Lines      208121   208017     -104     
==========================================
- Hits       167859   167764      -95     
+ Misses      40262    40253       -9

kshitij12345 · 2021-01-19T18:37:42Z

@ngimel PTAL :)

kshitij12345 · 2021-01-21T04:36:23Z

@ngimel Gentle Ping

facebook-github-bot

@ngimel has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.

facebook-github-bot · 2021-01-22T20:09:09Z

@ngimel merged this pull request in a291b25.

Summary: Fixes pytorch#49541 Reference: pytorch#24507 Pull Request resolved: pytorch#49732 Reviewed By: ejguan Differential Revision: D25991438 Pulled By: ngimel fbshipit-source-id: a43bd0bfe043d8e32a6cadbbf736a0eaa697e7ec

kshitij12345 added 2 commits December 22, 2020 03:49

migrate masked_scatter_cpu

266ca4c

temp: modify test

d62e4fb

facebook-github-bot added the cla signed label Dec 22, 2020

pytorchbot added the open source label Dec 22, 2020

kshitij12345 added 2 commits December 22, 2020 08:08

correct the condition to avoid heap overflow

90b5e87

move check to correct place

cc003e6

kshitij12345 marked this pull request as ready for review December 23, 2020 04:40

kshitij12345 changed the title ~~[WIP] Migrate masked_scatter_ CPU to ATen~~ Migrate masked_scatter_ CPU to ATen Dec 23, 2020

kshitij12345 requested a review from VitalyFedyunin December 23, 2020 04:41

kshitij12345 commented Dec 28, 2020

View reviewed changes

kshitij12345 requested a review from ngimel December 28, 2020 10:04

mruberry added the triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module label Dec 28, 2020

ngimel reviewed Jan 18, 2021

View reviewed changes

kshitij12345 added 4 commits January 18, 2021 22:59

address comments

28a4e97

fix bug: test

bf4fd03

Merge branch 'master' into migrate/masked_scatter_cpu

cdebcf5

try to make compilers happy

5653227

ngimel approved these changes Jan 21, 2021

View reviewed changes

facebook-github-bot reviewed Jan 21, 2021

View reviewed changes

facebook-github-bot closed this in a291b25 Jan 22, 2021

facebook-github-bot added the Merged label Jan 22, 2021

	def test_masked_scatter(self):
	with warnings.catch_warnings(record=True) as w:
	warnings.simplefilter("always")
	for maskType in [torch.uint8, torch.bool]:
	for dt in torch.testing.get_all_dtypes():
	num_copy, num_dest = 3, 10
	dest = torch.tensor([1, 2, 3, 4, 5, 6, 7, 8, 9, 10], dtype=dt)
	dest2 = dest.clone()
	src = torch.tensor([0, 0, 0, 0, 0, 0, 0, 0, 0, 0], dtype=dt)
	mask = torch.tensor((0, 0, 0, 0, 1, 0, 1, 0, 1, 0), dtype=maskType)

	if dt == torch.bool:
	# torch.bool is a special case and is being tested
	# in a separate test
	continue

	# TODO: update test when masked scatter is supported for complex
	if dt == torch.half or dt.is_complex:
	self.assertRaises(RuntimeError, lambda: dest.masked_scatter_(mask, src))
	continue

	dest.masked_scatter_(mask, src)
	j = 0
	for i in range(num_dest):
	if mask[i]:
	dest2[i] = src[j]
	j += 1
	self.assertEqual(dest, dest2, atol=0, rtol=0)

	# make source bigger than number of 1s in mask
	src = torch.tensor([1, 1, 1, 1, 1, 1, 1, 1, 1, 1], dtype=dt)
	dest.masked_scatter_(mask, src)

	# make src smaller. this should fail
	src = torch.randn(num_copy - 1)
	with self.assertRaises(RuntimeError):
	dest.masked_scatter_(mask, src)
	self.assertEqual(len(w), 27)

	warn = 'masked_scatter_ received a mask with dtype torch.uint8,'
	for wi in w:
	self.assertEqual(str(wi.message)[0:55], str(warn))

	def test_masked_scatter_bool_tensor(self, device):
	src = torch.tensor([True, True, True], device=device)
	dst = torch.tensor([False, False, False], device=device)
	mask = torch.tensor([False, True, False], device=device)

	dst.masked_scatter_(mask, src)
	self.assertEqual(dst, torch.tensor([False, True, False], device=device))

	mask = torch.tensor([True, False, True], device=device)
	dst = dst.masked_scatter(mask, src)
	self.assertEqual(dst, torch.tensor([True, True, True], device=device))

	@onlyOnCPUAndCUDA
	def test_masked_scatter_mem_overlap(self, device):
	x = torch.rand((1,), device=device).expand((6,))
	src = torch.rand((3,), device=device)
	mask = torch.tensor([True, False, True, True, False, False], device=device)

	with self.assertRaisesRegex(RuntimeError, 'unsupported operation'):
	x.masked_scatter_(mask, src)

	@onlyOnCPUAndCUDA
	def test_index_select_mem_overlap(self, device):
	x = torch.rand((1, 6), device=device).expand((2, 6))
	y = torch.rand((3, 6), device=device)
	ind = torch.tensor([0, 1], dtype=torch.int64, device=device)
	with self.assertRaisesRegex(RuntimeError, 'unsupported operation'):
	torch.index_select(y, 1, ind, out=x)

	TORCH_CHECK(
	self.scalar_type() == source.scalar_type(),
	"masked_scatter: expected self and source to have same dtypes but got",
	self.scalar_type(),
	" and ",
	source.scalar_type());

	TORCH_CHECK(self.device().type() == at::kCPU, "device type of self (", self.device().type(), ") is not CPU");
	TORCH_CHECK(mask.device().type() == at::kCPU, "device type of mask (", mask.device().type(), ") is not CPU");
	TORCH_CHECK(source.device().type() == at::kCPU, "device type of source (", source.device().type(), ") is not CPU");

	Tensor b_mask;
	std::tie(b_mask) = expand_inplace(self, mask, "masked_scatter_");

	if (b_mask.dtype() == ScalarType::Byte) {
	TORCH_WARN("masked_scatter_ received a mask with dtype torch.uint8, this behavior is now deprecated," \
	"please use a mask with dtype torch.bool instead.");
	}

Conversation

kshitij12345 commented Dec 22, 2020

Uh oh!

facebook-github-bot commented Dec 22, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

💊 CI failures summary and remediations

🕵️ 1 new failure recognized by patterns

pytorch_linux_bionic_py3_8_gcc9_coverage_test1 (1/1)

Uh oh!

kshitij12345 commented Dec 22, 2020

Uh oh!

kshitij12345 commented Dec 23, 2020

Uh oh!

kshitij12345 left a comment

Choose a reason for hiding this comment

Uh oh!

kshitij12345 Dec 28, 2020

Choose a reason for hiding this comment

Uh oh!

kshitij12345 Dec 28, 2020

Choose a reason for hiding this comment

Uh oh!

kshitij12345 commented Dec 30, 2020

Uh oh!

ngimel left a comment

Choose a reason for hiding this comment

Uh oh!

ngimel Jan 18, 2021

Choose a reason for hiding this comment

Uh oh!

kshitij12345 Jan 19, 2021

Choose a reason for hiding this comment

Uh oh!

ngimel Jan 18, 2021

Choose a reason for hiding this comment

Uh oh!

ngimel Jan 18, 2021

Choose a reason for hiding this comment

Uh oh!

kshitij12345 Jan 19, 2021

Choose a reason for hiding this comment

Uh oh!

codecov Bot commented Jan 19, 2021

Codecov Report

Uh oh!

kshitij12345 commented Jan 19, 2021

Uh oh!

kshitij12345 commented Jan 21, 2021

Uh oh!

facebook-github-bot left a comment

Choose a reason for hiding this comment

Uh oh!

facebook-github-bot commented Jan 22, 2021

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

facebook-github-bot commented Dec 22, 2020 •

edited

Loading