tag:github.com,2008:https://github.com/peterjc123/pytorch/releases 2020-06-24T02:05:48Z tag:github.com,2008:Repository/89826127/v1.6.0-rc1 2020-06-24T02:05:48Z <p>.circleci: Fix pip installation of awscli</p> <p>Signed-off-by: Eli Uriegas &lt;eliuriegas@fb.com&gt;</p> seemethere tag:github.com,2008:Repository/89826127/v1.5.1 2020-06-11T22:26:46Z <p>[ONNX] Fix pow op export [1.5.1] (<a class="issue-link js-issue-link" href="https://github.com/pytorch/pytorch/pull/39791">pytorch#39791</a>)</p> <p>* [ONNX] Fix pow op export (<a class="issue-link js-issue-link" href="https://github.com/pytorch/pytorch/pull/38065">pytorch#38065</a>)</p> <p>Summary: <br />Fix pow type cast for opset 9 and update opset 12 <br />Pull Request <span class="issue-keyword tooltipped tooltipped-se">resolved</span>: <a class="issue-link js-issue-link" href="https://github.com/pytorch/pytorch/pull/38065">pytorch#38065</a></p> <p>Differential Revision: D21485353</p> <p>Pulled By: malfet</p> <p>fbshipit-source-id: 3993e835ffad07b2e6585eb5cf1cb7c8474de2ec</p> <p>* Update ort-nighly version as suggested in <a class="issue-link js-issue-link" href="https://github.com/pytorch/pytorch/pull/39685#issuecomment-641452470">pytorch#39685 (comment)</a></p> <p>* Apply changes from <a class="issue-link js-issue-link" href="https://github.com/pytorch/pytorch/pull/37846">pytorch#37846</a> to `test_topk_smallest_unsorted`</p> <p>Co-authored-by: neginraoof &lt;neginmr@utexas.edu&gt;</p> malfet tag:github.com,2008:Repository/89826127/v1.5.1-rc1 2020-06-11T22:26:46Z <p>[ONNX] Fix pow op export [1.5.1] (<a class="issue-link js-issue-link" href="https://github.com/pytorch/pytorch/pull/39791">pytorch#39791</a>)</p> <p>* [ONNX] Fix pow op export (<a class="issue-link js-issue-link" href="https://github.com/pytorch/pytorch/pull/38065">pytorch#38065</a>)</p> <p>Summary: <br />Fix pow type cast for opset 9 and update opset 12 <br />Pull Request <span class="issue-keyword tooltipped tooltipped-se">resolved</span>: <a class="issue-link js-issue-link" href="https://github.com/pytorch/pytorch/pull/38065">pytorch#38065</a></p> <p>Differential Revision: D21485353</p> <p>Pulled By: malfet</p> <p>fbshipit-source-id: 3993e835ffad07b2e6585eb5cf1cb7c8474de2ec</p> <p>* Update ort-nighly version as suggested in <a class="issue-link js-issue-link" href="https://github.com/pytorch/pytorch/pull/39685#issuecomment-641452470">pytorch#39685 (comment)</a></p> <p>* Apply changes from <a class="issue-link js-issue-link" href="https://github.com/pytorch/pytorch/pull/37846">pytorch#37846</a> to `test_topk_smallest_unsorted`</p> <p>Co-authored-by: neginraoof &lt;neginmr@utexas.edu&gt;</p> malfet tag:github.com,2008:Repository/89826127/v1.5.0 2020-04-20T23:59:38Z <p>[v.1.5.0] Ensure linearIndex of advanced indexing backwards is contig… (</p> <p><a class="issue-link js-issue-link" href="https://github.com/pytorch/pytorch/pull/36962">pytorch#36962</a>)</p> <p>* [v.1.5.0] Ensure linearIndex of advanced indexing backwards is contiguous.</p> <p>This is a more straightforward solution to the problem than <a class="issue-link js-issue-link" href="https://github.com/pytorch/pytorch/pull/36957">pytorch#36957</a>; I don't know about the relative performance.</p> <p><span class="issue-keyword tooltipped tooltipped-se">Fixes</span>: <a class="issue-link js-issue-link" href="https://github.com/pytorch/pytorch/issues/36956">pytorch#36956</a></p> <p>ghstack-source-id: <a class="commit-link" href="https://github.com/peterjc123/pytorch/commit/43c48eaee7232cd3ed2b108edbbee24c11e8321a"><tt>43c48ea</tt></a> <br />Pull Request <span class="issue-keyword tooltipped tooltipped-se">resolved</span>: <a class="issue-link js-issue-link" href="https://github.com/pytorch/pytorch/pull/36959">pytorch#36959</a></p> <p>* Fix test.</p> gchanan tag:github.com,2008:Repository/89826127/v1.5.0-rc5 2020-04-20T23:59:38Z <p>[v.1.5.0] Ensure linearIndex of advanced indexing backwards is contig… (</p> <p><a class="issue-link js-issue-link" href="https://github.com/pytorch/pytorch/pull/36962">pytorch#36962</a>)</p> <p>* [v.1.5.0] Ensure linearIndex of advanced indexing backwards is contiguous.</p> <p>This is a more straightforward solution to the problem than <a class="issue-link js-issue-link" href="https://github.com/pytorch/pytorch/pull/36957">pytorch#36957</a>; I don't know about the relative performance.</p> <p><span class="issue-keyword tooltipped tooltipped-se">Fixes</span>: <a class="issue-link js-issue-link" href="https://github.com/pytorch/pytorch/issues/36956">pytorch#36956</a></p> <p>ghstack-source-id: <a class="commit-link" href="https://github.com/peterjc123/pytorch/commit/43c48eaee7232cd3ed2b108edbbee24c11e8321a"><tt>43c48ea</tt></a> <br />Pull Request <span class="issue-keyword tooltipped tooltipped-se">resolved</span>: <a class="issue-link js-issue-link" href="https://github.com/pytorch/pytorch/pull/36959">pytorch#36959</a></p> <p>* Fix test.</p> gchanan tag:github.com,2008:Repository/89826127/v1.5.0-rc4 2020-04-18T00:00:53Z Krovatkin tag:github.com,2008:Repository/89826127/v1.5.0-rc3 2020-04-09T21:08:57Z <p>Use counter instead of vector of futures in `_parallel_run` (<a class="issue-link js-issue-link" href="https://github.com/pytorch/pytorch/pull/36159">pytorch#…</a></p> <p><a class="issue-link js-issue-link" href="https://github.com/pytorch/pytorch/pull/36159">…36159</a>) (<a class="issue-link js-issue-link" href="https://github.com/pytorch/pytorch/pull/36334">pytorch#36334</a>)</p> <p>Summary: <br />This should be faster than allocating one mutex, flag and conditional variable per task.</p> <p>Using `std::atomic&lt;size_t&gt;` to count remaing tasks is not sufficient, <br />because modification of remaining counter and signalling conditional variable must happen atomically, <br />otherwise `wait()` might get invoked after `notify_one()` was called. <br />Pull Request <span class="issue-keyword tooltipped tooltipped-se">resolved</span>: <a class="issue-link js-issue-link" href="https://github.com/pytorch/pytorch/pull/36159">pytorch#36159</a></p> <p>Test Plan: CI</p> <p>Differential Revision: D20905411</p> <p>Pulled By: malfet</p> <p>fbshipit-source-id: facaf599693649c3f43edafc49f369e90d2f60de <br />(cherry picked from commit <a class="commit-link" href="https://github.com/peterjc123/pytorch/commit/986a8fdd6a18d9110f8bde59361967139450966b"><tt>986a8fd</tt></a>) <br />Signed-off-by: Eli Uriegas &lt;eliuriegas@fb.com&gt;</p> <p>Co-authored-by: Nikita Shulga &lt;nshulga@fb.com&gt;</p> seemethere tag:github.com,2008:Repository/89826127/v1.5.0-rc2 2020-03-28T00:44:03Z <p>Revert "Fix handling of non-finite values in topk (<a class="issue-link js-issue-link" href="https://github.com/pytorch/pytorch/pull/35253">pytorch#35253</a>)" (<a class="issue-link js-issue-link" href="https://github.com/pytorch/pytorch/pull/35582">p…</a></p> <p><a class="issue-link js-issue-link" href="https://github.com/pytorch/pytorch/pull/35582">…ytorch#35582</a>)</p> <p>This reverts commit <a class="commit-link" href="https://github.com/peterjc123/pytorch/commit/b12579da5398ff23b421332e21e18dc619a0b960"><tt>b12579d</tt></a>.</p> <p>This patch in-and-of itself looks fine, but it's causing some AMP tests to fail.</p> gchanan tag:github.com,2008:Repository/89826127/v1.4.1 2020-03-23T18:05:29Z yns88 tag:github.com,2008:Repository/89826127/v1.5.0-rc1 2020-03-18T07:13:18Z <p>Do not throw from CUDAContext destructor (<a class="issue-link js-issue-link" href="https://github.com/pytorch/pytorch/pull/34756">pytorch#34756</a>)</p> <p>Summary: <br />Throwing from destructor leads to undefined behaviour (most often to segault) <br />So it's better to leak memory then segault <br />Pull Request <span class="issue-keyword tooltipped tooltipped-se">resolved</span>: <a class="issue-link js-issue-link" href="https://github.com/pytorch/pytorch/pull/34756">pytorch#34756</a></p> <p>Test Plan: Run `test_pytorch_onnx_caffe2`</p> <p>Differential Revision: D20504228</p> <p>Pulled By: malfet</p> <p>fbshipit-source-id: 7a05776fea9036f602e95b8182f8493cb5886dab</p> malfet