arindamroy-eng

Results 5 issues of arindamroy-eng

### 🐛 Describe the bug Unskip the following test for ROCM5.0: test_import_stats | test_peak_memory | (__main__.TestImportTime) ### Versions ROCM 5.0

Unit Test Parity

### 🐛 Describe the bug Unskip test_grid_sample on ROCM 5.0 test_nn | test_Conv2d_naive_groups_cuda_float16 | (__main__.TestNNDeviceTypeCUDA) test_nn | test_grid_sample | (__main__.TestNN) It was disabled by: https://github.com/pytorch/pytorch/issues/54535 ### Versions ROCM 5.0

Unit Test Parity

### 🐛 Describe the bug The following tests are passing in ROCM5.0: test_nn | test_softmax_cuda_float32 | (__main__.TestNNDeviceTypeCUDA) test_nn | test_softmax_results_cuda_float16 | (__main__.TestNNDeviceTypeCUDA) test_nn | test_softmax_results_cuda_float32 | (__main__.TestNNDeviceTypeCUDA) We need to...

Unit Test Parity

### 🐛 Describe the bug We need to unskip the following tests in test_ops.py. test_ops | test_fn_grad_min_binary_cuda_float64 | (__main__.TestGradientsCUDA) test_ops | test_fn_gradgrad_eig_cuda_complex128 | (__main__.TestGradientsCUDA) These are passing in ROCM5.0 ###...

Unit Test Parity

### 🐛 Describe the bug Need to enable following tests for ROCM in test_ops.py: test_frexp_assert_raises_cuda test_frexp_cuda_float16 test_frexp_cuda_float32 test_frexp_cuda_float64 test_igamma_common_cuda_float32 test_igamma_common_cuda_float64 These are passing in ROCM 5.0 These were disabled by...

Unit Test Parity