Cleanup Parser for resize operator #4204
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
also get rid of the readability-function-size tidy warning
TedThemistokleous
added
simple
pfultz2
reviewed
Aug 6, 2025
pfultz2
reviewed
Aug 6, 2025
used since onnx opset 11
added as of resize opset 18
Represent all the posible combinations/defaults and states using a struct Will slowly build up functionality in a parser that takes these then mutates its state based on the parsed input capability
Use this to get / perform the proper parse based on the attributes detectd. leverage the use of std::optinal here for things like the scale() which can be made to be an attr in upsample operator but an input in both resize and upsample too
TedThemistokleous
force-pushed
the
rewrite_resize_linear
branch
from
0eeca94 to
efefa14
Compare
Use the same get_scales() call for upsample scale values if its not detected as an attribute
Cleans up the transforms used to get output or scales by using other args to compute. Encapsulated these in the resize_arg
Codecov Report❌ Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## develop #4204 +/- ##
===========================================
+ Coverage 92.23% 92.23% +0.01%
===========================================
Files 553 555 +2
Lines 25627 25767 +140
===========================================
+ Hits 23635 23766 +131
- Misses 1992 2001 +9
🚀 New features to boost your workflow:
|
There was a problem hiding this comment.
I think there are some parts where combining the upsample and resize functionalities is created really messy functions. It would be nice to decouple some of these things for the two use cases (especially for figuring out the scales and output size) by putting them in entirely separate functions
Scales inputs should always be of type float. Its an error if we attempt to use scales or downcast the input to a lower type which could lead to errors. The onnx spec specifies across all Resize/Upsamples that scale is always of type float. Fixed this since we originally had an else case when checking type for size or scale input that would fall through allowing half, bf16, etc types through. Added an explicit check here and a test case for this
- rename compute_ouptuts to comptue_output_sizes - Handle mispell of neighbor - Use setter/gett of scales_sizes_arg better - Remove double assignments of vec_scales - Cleanup get_scales to be assign_scale_or_size() function Should clean up some of the logic and clarify what some of these blocks are doing better
Not needed and just go by the op_name instead simplifies and removes the need for another flag and two members in the struct
This build is not recommended to merge 🔴 |
❌bert-mrpc-tf: ERROR - check error outputerror: unknown warning option '-Wnrvo' [-Werror,-Wunknown-warning-option]error: unknown warning option '-Wnrvo' [-Werror,-Wunknown-warning-option] 2025-08-24 14:14:00.724885: I tensorflow/core/platform/cpu_feature_guard.cc:210] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations. To enable the following instructions: SSE3 SSE4.1 SSE4.2 AVX AVX2 FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags. WARNING: All log messages before absl::InitializeLog() is called are written to STDERR I0000 00:00:1756062845.917136 172407 gpu_device.cc:2022] Created device /job:localhost/replica:0/task:0/device:GPU:0 with 62951 MB memory: -> device: 0, name: AMD Instinct MI250X/MI250, pci bus id: 0000:32:00.0 WARNING: All log messages before absl::InitializeLog() is called are written to STDERR I0000 00:00:1756062846.785120 172407 mlir_graph_optimization_pass.cc:401] MLIR V1 optimization pass is not enabled 2025-08-24 14:14:15.206017: E external/local_xla/xla/service/gpu/llvm_gpu_backend/gpu_backend_lib.cc:250] bitcode module is required by this HLO module but was not found at ./opencl.bc 2025-08-24 14:14:15.206162: E external/local_xla/xla/service/gpu/llvm_gpu_backend/gpu_backend_lib.cc:250] bitcode module is required by this HLO module but was not found at ./opencl.bc 2025-08-24 14:14:15.206203: E external/local_xla/xla/service/gpu/llvm_gpu_backend/gpu_backend_lib.cc:250] bitcode module is required by this HLO module but was not found at ./opencl.bc 2025-08-24 14:14:15.206232: E external/local_xla/xla/service/gpu/llvm_gpu_backend/gpu_backend_lib.cc:250] bitcode module is required by this HLO module but was not found at ./opencl.bc 2025-08-24 14:14:15.206277: E external/local_xla/xla/service/gpu/llvm_gpu_backend/gpu_backend_lib.cc:250] bitcode module is required by this HLO module but was not found at ./opencl.bc 2025-08-24 14:14:15.206322: E external/local_xla/xla/service/gpu/llvm_gpu_backend/gpu_backend_lib.cc:250] bitcode module is required by this HLO module but was not found at ./opencl.bc 2025-08-24 14:14:15.206353: E external/local_xla/xla/service/gpu/llvm_gpu_backend/gpu_backend_lib.cc:250] bitcode module is required by this HLO module but was not found at ./opencl.bc 2025-08-24 14:14:15.206400: E external/local_xla/xla/service/gpu/llvm_gpu_backend/gpu_backend_lib.cc:250] bitcode module is required by this HLO module but was not found at ./opencl.bc error: Failure when generating HSACO error: Failure when generating HSACO error: Failure when generating HSACO error: Failure when generating HSACO error: Failure when generating HSACO error: Failure when generating HSACO error: Failure when generating HSACO error: Failure when generating HSACO 2025-08-24 14:14:15.207513: E tensorflow/compiler/mlir/tools/kernel_gen/tf_framework_c_interface.cc:228] INTERNAL: Generating device code failed. 2025-08-24 14:14:15.208694: W tensorflow/core/framework/op_kernel.cc:1829] UNKNOWN: JIT compilation failed. 2025-08-24 14:14:15.208715: I tensorflow/core/framework/local_rendezvous.cc:405] Local rendezvous is aborting with status: UNKNOWN: JIT compilation failed. [[{{node import/bert/embeddings/LayerNorm/moments/SquaredDifference}}]] 2025-08-24 14:14:15.208727: I tensorflow/core/framework/local_rendezvous.cc:405] Local rendezvous is aborting with status: UNKNOWN: JIT compilation failed. [[{{node import/bert/embeddings/LayerNorm/moments/SquaredDifference}}]] [[import/loss/output/_21]] 2025-08-24 14:14:15.208742: I tensorflow/core/framework/local_rendezvous.cc:424] Local rendezvous recv item cancelled. Key hash: 11217777527359497193 Traceback (most recent call last): File "/usr/local/lib/python3.10/dist-packages/tensorflow/python/client/session.py", line 1407, in _do_call return fn(*args) File "/usr/local/lib/python3.10/dist-packages/tensorflow/python/client/session.py", line 1390, in _run_fn return self._call_tf_sessionrun(options, feed_dict, fetch_list, File "/usr/local/lib/python3.10/dist-packages/tensorflow/python/client/session.py", line 1483, in _call_tf_sessionrun return tf_session.TF_SessionRun_wrapper(self._session, options, feed_dict, tensorflow.python.framework.errors_impl.UnknownError: 2 root error(s) found. (0) UNKNOWN: JIT compilation failed. [[{{node import/bert/embeddings/LayerNorm/moments/SquaredDifference}}]] [[import/loss/output/_21]] (1) UNKNOWN: JIT compilation failed. [[{{node import/bert/embeddings/LayerNorm/moments/SquaredDifference}}]] 0 successful operations. 0 derived errors ignored. During handling of the above exception, another exception occurred: Traceback (most recent call last): File "/src/AMDMIGraphX/tools/accuracy/accuracy_checker.py", line 359, in main() File "/src/AMDMIGraphX/tools/accuracy/accuracy_checker.py", line 335, in main y_out = sess.run(y, feed_dict=tf_dict) File "/usr/local/lib/python3.10/dist-packages/tensorflow/python/client/session.py", line 977, in run result = self._run(None, fetches, feed_dict, options_ptr, File "/usr/local/lib/python3.10/dist-packages/tensorflow/python/client/session.py", line 1220, in _run results = self._do_run(handle, final_targets, final_fetches, File "/usr/local/lib/python3.10/dist-packages/tensorflow/python/client/session.py", line 1400, in _do_run return self._do_call(_run_fn, feeds, fetches, targets, options, File "/usr/local/lib/python3.10/dist-packages/tensorflow/python/client/session.py", line 1426, in _do_call raise type(e)(node_def, op, message) # pylint: disable=no-value-for-parameter tensorflow.python.framework.errors_impl.UnknownError: Graph execution error: Detected at node 'import/bert/embeddings/LayerNorm/moments/SquaredDifference' defined at (most recent call last): Node: 'import/bert/embeddings/LayerNorm/moments/SquaredDifference' Detected at node 'import/bert/embeddings/LayerNorm/moments/SquaredDifference' defined at (most recent call last): Node: 'import/bert/embeddings/LayerNorm/moments/SquaredDifference' 2 root error(s) found. (0) UNKNOWN: JIT compilation failed. [[{{node import/bert/embeddings/LayerNorm/moments/SquaredDifference}}]] [[import/loss/output/_21]] (1) UNKNOWN: JIT compilation failed. [[{{node import/bert/embeddings/LayerNorm/moments/SquaredDifference}}]] 0 successful operations. 0 derived errors ignored. Original stack trace for 'import/bert/embeddings/LayerNorm/moments/SquaredDifference': 🔴bert_large_uncased_fp16: FAILED: MIGraphX is not within tolerance - check verbose output🔴mask-rcnn: FAILED: MIGraphX is not within tolerance - check verbose output |
shivadbhavsar
approved these changes
Aug 25, 2025
pfultz2
approved these changes
Aug 26, 2025
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Precursor cleanup for resize parser which has a flattened set of calls for parsing between resize and linear modes.
More cleanup in another PR to handle upscale/resize with this parser in a more sane way before we do any changes to the modes of this operator.Should get rid of the tidy complexity call right now.Started off simple but found a few things with this one before I begin modifying the linear algorithm for resizing to integer values.