From commits-return-11464-archive-asf-public=cust-asf.ponee.io@tvm.apache.org Wed Apr 15 10:08:16 2020 Return-Path: X-Original-To: archive-asf-public@cust-asf.ponee.io Delivered-To: archive-asf-public@cust-asf.ponee.io Received: from mail.apache.org (hermes.apache.org [207.244.88.153]) by mx-eu-01.ponee.io (Postfix) with SMTP id 784BA18065C for ; Wed, 15 Apr 2020 12:08:16 +0200 (CEST) Received: (qmail 63100 invoked by uid 500); 15 Apr 2020 10:08:16 -0000 Mailing-List: contact commits-help@tvm.apache.org; run by ezmlm Precedence: bulk List-Help: List-Unsubscribe: List-Post: List-Id: Reply-To: dev@tvm.apache.org Delivered-To: mailing list commits@tvm.apache.org Received: (qmail 63091 invoked by uid 99); 15 Apr 2020 10:08:15 -0000 Received: from ec2-52-202-80-70.compute-1.amazonaws.com (HELO gitbox.apache.org) (52.202.80.70) by apache.org (qpsmtpd/0.29) with ESMTP; Wed, 15 Apr 2020 10:08:15 +0000 From: GitBox To: commits@tvm.apache.org Subject: [GitHub] [incubator-tvm] siju-samuel commented on a change in pull request #5332: [PYTORCH]Take, Topk op support Message-ID: <158694529588.22151.11778089542312316613.gitbox@gitbox.apache.org> References: In-Reply-To: Date: Wed, 15 Apr 2020 10:08:15 -0000 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 8bit siju-samuel commented on a change in pull request #5332: [PYTORCH]Take, Topk op support URL: https://github.com/apache/incubator-tvm/pull/5332#discussion_r408729508 ########## File path: tests/python/frontend/pytorch/test_forward.py ########## @@ -1545,6 +1545,61 @@ def forward(self, *args): verify_model(Round1().float().eval(), input_data=input_data) +def test_forward_take(): + torch.set_grad_enabled(False) + class Take1(Module): + def forward(self, *args): + indices = torch.tensor([[0,0],[1,0]]) + if torch.cuda.is_available(): Review comment: My first commit it was not there, but the[ CI failed. ] (https://ci.tvm.ai/blue/organizations/jenkins/tvm/detail/PR-5332/1/pipeline) I think only tvm runs with CPU, torch uses GPU. ``` def test_forward_take(): torch.set_grad_enabled(False) class Take1(Module): def forward(self, *args): return torch.take(args[0], torch.tensor([[0,0],[1,0]])) class Take2(Module): def forward(self, *args): return torch.take(args[0], args[1]) input_data = torch.tensor([[1,2],[3,4]]) > verify_model(Take1().float().eval(), input_data=input_data) tests/python/frontend/pytorch/test_forward.py:1559: _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ tests/python/frontend/pytorch/test_forward.py:157: in verify_model baseline_outputs = baseline_model(*baseline_input) /usr/local/lib/python3.6/dist-packages/torch/nn/modules/module.py:532: in __call__ result = self.forward(*input, **kwargs) _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ self = Take1(), args = (tensor([[1, 2], [3, 4]], device='cuda:0'),) def forward(self, *args): > return torch.take(args[0], torch.tensor([[0,0],[1,0]])) E RuntimeError: Expected object of device type cuda but got device type cpu for argument #2 'index' in call to _th_take ``` ---------------------------------------------------------------- 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: users@infra.apache.org With regards, Apache Git Services