-
Notifications
You must be signed in to change notification settings - Fork 9
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Execute convolution in NCHW if the suggested mem format is NHWC but the actual mem layout is NCHW #303
base: master
Are you sure you want to change the base?
Commits on Feb 14, 2023
-
Add back support for PYTORCH_TEST_WITH_MPS (#66)
Fix the TEST_WITH_MPS macro.
Configuration menu - View commit details
-
Copy full SHA for 23d334b - Browse repository at this point
Copy the full SHA 23d334bView commit details -
* Test MPS CI runners * Cherry pick remaining files * Enable lintrunner: * Change lint runner * Retrigger checks * Retrigger checks #2 * Retrigger checks #3 * Retrigger checks #4 * Retrigger checks #5 * Retrigger checks #5 * Retrigger checks #7 * Retrigger checks #8 * Retrigger checks #9 * Retrigger checks #9 (change arch to arm) * Retrigger checks #10 * Retrigger checks #11 * Retrigger checks #12 * Retrigger checks #13 * Retrigger checks #14 * Retrigger checks #14 * Retrigger checks #15 * Retrigger checks #16 * Retrigger checks #16 * Retrigger checks #17 * Retrigger checks #19 * Retrigger checks #20 * Retrigger checks #21 * Fix lintrunner * Fix lintrunner * Remove lint.json
Configuration menu - View commit details
-
Copy full SHA for 12085cf - Browse repository at this point
Copy the full SHA 12085cfView commit details -
Use DISTRIBUTED=1 for MPS CI runners (#292)
* Use DISTRIBUTED=1 for MPS CI runners * Disable openmp
Configuration menu - View commit details
-
Copy full SHA for e8f89df - Browse repository at this point
Copy the full SHA e8f89dfView commit details -
Configuration menu - View commit details
-
Copy full SHA for 85cdb98 - Browse repository at this point
Copy the full SHA 85cdb98View commit details -
Configuration menu - View commit details
-
Copy full SHA for 66951a0 - Browse repository at this point
Copy the full SHA 66951a0View commit details -
Remove unnecessary CI files (#327)
* Remove unnecessary CI files * Additional files * Update lint
Configuration menu - View commit details
-
Copy full SHA for 5ada241 - Browse repository at this point
Copy the full SHA 5ada241View commit details -
Enable test modules on MPS and CI runners (#305) (#324)
* Enable test modules on MPS and CI runners * Update lint.yml * Update comments * Retrigger CI * Retrigger CI #2 * Remove comment
Configuration menu - View commit details
-
Copy full SHA for bf8eba9 - Browse repository at this point
Copy the full SHA bf8eba9View commit details -
[CHERRY-PICK] Block uint8 data type for unary and binary ops on macOS…
Configuration menu - View commit details
-
Copy full SHA for 2f336a4 - Browse repository at this point
Copy the full SHA 2f336a4View commit details
Commits on Feb 15, 2023
-
Configuration menu - View commit details
-
Copy full SHA for 108cdc0 - Browse repository at this point
Copy the full SHA 108cdc0View commit details -
Configuration menu - View commit details
-
Copy full SHA for 8de3315 - Browse repository at this point
Copy the full SHA 8de3315View commit details -
Fix bilinear backward pass (#331)
* Fix bilinear backward pass * Remove comment
Configuration menu - View commit details
-
Copy full SHA for 051bc9c - Browse repository at this point
Copy the full SHA 051bc9cView commit details -
Update macOS 12 blocklist (#323)
* Update macOS 12 blocklist - move sum, masked.var, mul to low precision list - unblock them from running * - mark __rdiv__ failures as accumulate error exceeds atol/rtol
Configuration menu - View commit details
-
Copy full SHA for 1b09ea2 - Browse repository at this point
Copy the full SHA 1b09ea2View commit details -
- Backward pass has to give explicit bias tensor of zeros if none is passed to the op or the bias gradient will not be calculated. - Fixed bias tensor mistakenly getting overwritten to zeros - Fixes crash when lstm op called with has_biases set to false. Change takes into account the changed shape of the input params TensorList depending on the bias flag. Co-authored-by: Kulin Seth <kulin_seth@apple.com>
Configuration menu - View commit details
-
Copy full SHA for 8c7df6f - Browse repository at this point
Copy the full SHA 8c7df6fView commit details -
Fix nn.functional.conv_transpose2d grad (#312) (#329)
- add _mps_convolution_impl that takes optional shape - for conv_tranpose2d grad, use the shape from input directly - remove nn.functional.conv_transpose2d grad from blocklist Co-authored-by: Ronian526 <11454459+Ronian526@users.noreply.github.com>
Configuration menu - View commit details
-
Copy full SHA for d42f74f - Browse repository at this point
Copy the full SHA d42f74fView commit details -
Fix the crash in elu_backward() (#333)
Fixes a crash where the inputTensor could go null and cause a crash.
Configuration menu - View commit details
-
Copy full SHA for 2856203 - Browse repository at this point
Copy the full SHA 2856203View commit details -
Fix nn.functional.embedding grad (#335)
- casting the input tensor to float32 and cast back the output tensor - unblock the test
Configuration menu - View commit details
-
Copy full SHA for 18797b0 - Browse repository at this point
Copy the full SHA 18797b0View commit details -
Configuration menu - View commit details
-
Copy full SHA for cf06ac5 - Browse repository at this point
Copy the full SHA cf06ac5View commit details -
Reduction cast f16 to f32 only on macOS 12 (#332)
- unblock rdiv float16
Configuration menu - View commit details
-
Copy full SHA for c65b823 - Browse repository at this point
Copy the full SHA c65b823View commit details
Commits on Feb 16, 2023
-
Configuration menu - View commit details
-
Copy full SHA for 73f7068 - Browse repository at this point
Copy the full SHA 73f7068View commit details -
Fix upsample for NHWC output (#337)
* Fix upsample for NHWC output * Add testcase
Configuration menu - View commit details
-
Copy full SHA for 1c8f126 - Browse repository at this point
Copy the full SHA 1c8f126View commit details -
Configuration menu - View commit details
-
Copy full SHA for 42be72a - Browse repository at this point
Copy the full SHA 42be72aView commit details
Commits on Feb 17, 2023
-
- give warnings of converting int64 for reduction ops - use cast tensor for reduction sum on trace - unblock trace from running
Configuration menu - View commit details
-
Copy full SHA for 6ace5f9 - Browse repository at this point
Copy the full SHA 6ace5f9View commit details -
Update random result list (#339)
* - move nn.functional.feature_alpha_dropoutwith_train, normalnumber_mean, new_empty_strided to expected failures * - update new_empty_strided --------- Co-authored-by: Kulin Seth <kulin_seth@apple.com>
Configuration menu - View commit details
-
Copy full SHA for c9b8ab7 - Browse repository at this point
Copy the full SHA c9b8ab7View commit details -
Fix convolution crash in backward with weights; remove unnecessary co…
…ntiguous calls (#341) * Fix convolution crash; remove unnecessary contiguous calls * Fix lintrunner
Configuration menu - View commit details
-
Copy full SHA for d3e414e - Browse repository at this point
Copy the full SHA d3e414eView commit details -
Fix copy_cast_mps() on tensors with storage offset (#343)
This should fix the failure with GPT2 when use_cache=True
Configuration menu - View commit details
-
Copy full SHA for be8817b - Browse repository at this point
Copy the full SHA be8817bView commit details
Commits on Feb 18, 2023
-
Configuration menu - View commit details
-
Copy full SHA for 8e37116 - Browse repository at this point
Copy the full SHA 8e37116View commit details
Commits on Feb 21, 2023
-
Co-authored-by: Ramin Azarmehr <razarmehr@apple.com>
Configuration menu - View commit details
-
Copy full SHA for c30946a - Browse repository at this point
Copy the full SHA c30946aView commit details -
Dev/skotapati/copy broadcasting (#350)
* Handle broadcasting by expanding src tensor in Copy.mm * Unblock linalg_matrix_power * Improved formatting
Configuration menu - View commit details
-
Copy full SHA for b520970 - Browse repository at this point
Copy the full SHA b520970View commit details -
Execute convolution in NCHW if the suggested mem format is NHWC but t…
…he actual mem layout is NCHW
Configuration menu - View commit details
-
Copy full SHA for 0473fe8 - Browse repository at this point
Copy the full SHA 0473fe8View commit details -
Configuration menu - View commit details
-
Copy full SHA for 9a5b002 - Browse repository at this point
Copy the full SHA 9a5b002View commit details