- 05 Mar, 2019 1 commit
-
-
Robert Kimball authored
* address one-hot tests * get backends passing unit tests * disable OneHot test until it can be fixed
-
- 04 Mar, 2019 6 commits
-
-
Robert Kimball authored
* add ngraph and std namespaces to c++ files * style
-
Adam Procter authored
-
Jayaram Bobba authored
* Extend compile API to take in PassConfig object * fix override warning * remove extra semicolon * cmake fixes to support cases where include_path has multiple directories * Help pybind find the overloaded compile methods * Limit compile-api exposed through PyBind (#2530) * clang-format * Remove setter for compilation mode to prevent post-init changes. Add compile-mode warning. * Removed pass_config include * fix merge
-
Robert Kimball authored
-
Sang Ik Lee authored
-
Scott Cyphers authored
* Fix bug introduced by #2238 * Added a debug logging macro: NGRAPH_DIST_DEBUG which prints the timestamp and MPI rank for distributed AllReduceops * Fix bug introduced by #2238 on r0.15 (#2515) * Fix bug introduced by #2238 * style * Fixed code formatting. * Moved the logging implementation to log.cpp * Fixed clang warning * Don't use namespace in header * Fix ifdef * disable warning * Fixed Centos build issues (gcc 4.8.5 doesn't implement std::put_time - hence the issue). Also fixed a logical error in the distributed setup for unit tests
-
- 02 Mar, 2019 2 commits
-
-
Adam Procter authored
* Skeleton for faster validation asserts * Switch to __VA_ARGS__ for compatibility, remove -Wno-variadic-macros * Add benchmarks for constructing Add and Convolution * Quick hack to avoid shadowing inside the CHECK macro * Quick hack to avoid inadvertent capture inside the macro * Update convolution, and change a bunch of tests to anticipate the new error class
-
Adam Procter authored
-
- 01 Mar, 2019 7 commits
-
-
Jayaram Bobba authored
* Fallback to default pooling kernels when mkldnn doesn't support input format * create default output descriptors
-
tsocha authored
* [ONNX] change [] into .at() * Update graph.cpp
-
Michał Karzyński authored
-
tsocha authored
-
Michał Karzyński authored
-
Michał Karzyński authored
-
tsocha authored
-
- 28 Feb, 2019 4 commits
-
-
Sang Ik Lee authored
* Fix bug introduced by #2238 * Apply style.
-
Leona C authored
-
Aleksey Marchuk authored
* Add all dependent libraries to mlsl linkage interface * Use ${CMAKE_SHARED_LIBRARY_PREFIX} and ${CMAKE_SHARED_LIBRARY_SUFFIX}
-
Scott Cyphers authored
-
- 27 Feb, 2019 12 commits
-
-
Robert Kimball authored
* Normalize the printing of settings of bools to TRUE and FALSE * change TRUE/FALSE to ON/OFF
-
Sang Ik Lee authored
-
Amy Zhuang authored
* Refactor to create MKLDNN primitives on the first iteration: add, avg_pool, batch_norm, bounded_relu, concat, convert_layout, leaky_relu, lrn, max_pool, quantized_avg_pool, quantized_max_pool, relu, sigmoid, slice, softmax. * Refactor to create MKLDNN primitives on the first iteration: pooling backward, convolution. * Refactor to create MKLDNN primitives on the first iteration: convolution backward, rnn, lstm, quantization, dequantization. * Delete one duplicate declaration. * Create and pass mkldnn descriptors/primitive-descriptors for ops. * Create and pass mkldnn descriptors for convolution backward ops. * Remove one unused variable. * Remove unused variables. * Remove unused variables. * Address PR feedback. * Fix a bug. * Add one parameter to build_quantize_reorder. * Address PR feedback. * Fix bi-rnn issue.
-
Robert Kimball authored
* rename and document the node name methods * address more references to renamed methods * fix compile error * fix build error
-
Amy Zhuang authored
* Reuse memory for CPU backend. * Use NGRAPH_REUSE_MEMORY to enable memory reuse. * Add a test. * Move make_function to test_tools.cpp. * Add more comments. * Address PR Feedback: add a method to CPU backend. * *Add a member to CPUOpAnnotations to remove redundant code. *Overload compile function for CPU backend. * Move make_function out of test_tools. * Address PR Feedback. * Use modified liveness analysis in CPUMemoryAssignment pass. * Use lambda expression. * Fix style error. * Check if any user of the tensor has destructive io when building tensor alias map. * Fix a bug. * Check if tensor has multiple users. * Allow tensor alias for destructive oi node. * Update multiple_users_tensor set along the chain of in place ops. * No tensor alias if input is parameter or constant. * Use buffer sets in cpu memory assignment, tensors sharing the same memory buffer are put into the same set. * Add more checks and do not combine sets when allowing destructive oi. * Style fix. * Do no allow destructive oi if the input tensor uses function input memory. Update set label. * Add unit tests. * Style fix. * Get the correct size for memcpy when the input is padded. * Style fix. * Address PR feedback. * Address PR feedback. * Move make_function in cpu_test after #if 0 and before the disabled test. * Add utility functions. Use iterator. Rename variables. * Add pass attributes and move cpu memory assignment to common passes (#2504)
-
Scott Cyphers authored
* Add info about lib versions in an easy to find place * Review comments
-
Sergey Shalnov authored
-
Robert Kimball authored
* function call working * fix compile error * fix compile error * add attribute support to plot_graph * fix build error * fix merge error * better colors for FunctionCall op
-
Leona C authored
* Cleaner API doc reference for compile call * Add a useful table for nGraph namespaces * Remove layout namespace * Show exploding kernel problem on illustration like IEEE preso * WIP branch for new documentation restructuring that is a huge pain * Fix the doc reorg mess * Fix underline * List of passes disclaimer note * Update disclaimers on README * More cleanup of doc reorg * Update core docs * Update overview on core * Add PR feedback * Get rid of all the gazillion of doc build errors from rearranging stuff * Add section on tutorials * Update branch * Cleanup intro * Add better detail to overview * Revise buildlb instructions and add better title for contributing to doc * Note about unit tests * Editing * Update core overview namespace table and fix more broken links due to ToC changes * Update normalized boolean build defaults * Update for PR 2507 * Incorporate new PR feedback review
-
Ayan Moitra authored
* Int unit tests that fail with bfloat * move tests out of single file * style * Incorporate Bob's comments * edits * Incorporate comments * style * edits * Add failing test to intel gpu manifest * comments incoprorated
-
Sergey Shalnov authored
-
tsocha authored
* Remove get_numpy_broadcast_shape helper function * Remove numpy_style_broadcast_for_binary_operation helper function * Remove TODO * Review fix pt. 1 * Remove parameters as shape containers * Fix LSTM * Review fix pt. 1 * Style apply * Use old comment
-
- 26 Feb, 2019 8 commits
-
-
Jayaram Bobba authored
* Add QuantizedConcat * Remove unused variables and add check for size of mins and maxes vector * Resolve conflicts * Merged with master and addressed some PR feedback * Maxpool and Avgpool fusions. Exclude Q from conv+relu fusion * Remove single-user check from fusions * Quantized concat fusion * workaround: do reshape sinking by default * style fix * check scales for QuantizedConcat * use compare_constants * remove stale comment * Handle all concat cases from arg size 2 to 6 * addressed feedback
-
Sergey Shalnov authored
* IntelGPU backend: Relu and Sigmoid datatypes support * fix for OpenCL constants * add const to variables * PR2500. Style fix
-
Adam Rogowiec authored
* Utility functions for calculating Lp norm. * Use functor object as a reduction operation. * Use new api of make_ng_reduction_op. * Use utility norm functions for reduction operations. * Onnx GlobalLpPool operator. * Ensure correct shapes after lp_norm reduction. * Remove unused function overload. * Fix shapes and tensor types. * Unit tests. * Update comments. * Update supported ops status table. * Fix: take absolute value of input tensor elements. * UT: with odd value p-norm. * Fix: move taking abs value into respective lp-norm functions. * Fix clang -Wdocumentation-unknown-command error. * Update supported op status table with new Jira ticket for Erf op. * Update supported_ops status table. * Update interface of make_ng_reduction_op - accept std::function object. * Update to use new make_ng_reduction_op api. * Remove unused header. * Fix errors on CentOS.
-
Robert Kimball authored
* Move codewriter out of codegen to ngraph root. It is useful for more than writing code. * remove codewriter.* from intel gpu backend and use ngraph version * fix merge issues
-
Rob Earhart authored
* Add a direct-to-Tile op * Disable dequantize_dynamic_offset * Add missing Py op defn * Generic passthrough op; serialization * Appease Linux builds * Add gpu handlers * Disable floor_int32 for now
-
Sang Ik Lee authored
-
Sandeep authored
* fix a bug on finalize when uninitialized bool * change this_init_comm -> m_init_comm move init to header
-
Pruthvi authored
* - MKLDNN would choose the algorithm which will potentially give best performance based on - convolution dimensions number of logical processors available. - (For auto-dispatching to work as intended, - use the same thread affinity settings when creating the convolution as when executing the convolution.) - The relationship between convolution sizes and the best performing algorithm is empirically based on performance observations * bump mkldnn version to V0.18-rc * Revert "- MKLDNN would choose the algorithm which will potentially give best performance based on" This reverts commit 904beb8ad8d4e829fbae5f38a803ea80a72b3ffd. * Update mkl-dnn patch for soversion removal.
-