Arm Backend: Add support for ELU.default operator#12996

Closed

agrima1304 wants to merge 423 commits intopytorch:mainfrom

agrima1304:Arm-Backend-Add-support-for-ELU.default-operator

This pull request is big! We're only showing the most recent 250 commits

Commits on Aug 26, 2025

Import Fix

lucylq
authored and
agrima1304
committed
Fix pyre errors

izaitsevfb
authored and
agrima1304
committed
[Backend Tester] Add LSTM tests (#13238)

GregoryComer
authored and
agrima1304
committed
[Backend Tester] Add avgpool tests (#13239)

GregoryComer
authored and
agrima1304
committed
[Backend Tester] Add maxpool tests (#13240)

GregoryComer
authored and
agrima1304
committed
[Backend Tester] Add adaptive avgpool tests (#13241)

GregoryComer
authored and
agrima1304
committed
[Backend Tester] Add adaptive maxpool tests (#13242)

GregoryComer
authored and
agrima1304
committed
[Backend Tester] Add Qualcomm tester and register flow (#12739)

GregoryComer
authored and
agrima1304
committed
[Backend Tester] Add CSV report generation (#12741)

GregoryComer
authored and
agrima1304
committed
Update build from source docs (#13210)

GregoryComer
authored and
agrima1304
committed
Move to Span<EValue*> instead of EValue** in delegate interface

JacobSzwejbka
authored and
agrima1304
committed
Update XNNPACK to 3131afe (#13234)

mcr229
authored and
agrima1304
committed
[Backend Tester] Add tensor error statistic reporting (#12809)

GregoryComer
authored and
agrima1304
committed
[Backend Tester] Report quantization and lowering times (#12838)

GregoryComer
authored and
agrima1304
committed
Added ETDump to Wasm (#13304)

Conarnar
authored and
agrima1304
committed
[Backend Tester] Report delegation statistics (#12846)

GregoryComer
authored and
agrima1304
committed
[Backend Tester] Report PTE size (#13249)

GregoryComer
authored and
agrima1304
committed
[ET-VK] Add optional blocklist and allowlist to vulkan partitioner to aid debugging (#13326)

authored and
agrima1304
committed
[ET-VK][ez] Fix `vulkan_schema` interface and link Vulkan to pybindings (#13327)

authored and
agrima1304
committed
[ET-VK][examples] Create export script for Vulkan examples (#13294)

authored and
agrima1304
committed
[ET-VK][CI] Add vulkan CI to test exporting and running models (#13295)

authored and
agrima1304
committed
Add csv and etdump files to gitignore (#13301)

mcr229
authored and
agrima1304
committed
[Backend Tester] Add portable test flow (#13250)

GregoryComer
authored and
agrima1304
committed
Static attention IO manager: fix causal mask bug for last input position

sxu
authored and
agrima1304
committed
Update buck srcs for tester harness

GregoryComer
authored and
agrima1304
committed
Check the number of inputs in Method::set_inputs (#13341)

shoumikhin
authored and
agrima1304
committed
Add TorchAO wrapper config to allow filter_fn for quantize_ (#13264)

abhinaykukkadapu
authored and
agrima1304
committed
Add coreml quant recipes (#13265)

abhinaykukkadapu
authored and
agrima1304
committed
Fix 8w8a qat qconfig setting activations

navsud
authored and
agrima1304
committed
NXP backend: Return number of Neutron nodes in exception (#13126)

authored and
agrima1304
committed
Enable strongly typed ops for deployment

mcremon-meta
authored and
agrima1304
committed
Arm backend: Run pytest model tests in parallell (#13322)

Erik-Lundell
authored and
agrima1304
committed
Arm backend: Update arm_tester.py with generate_etrecord (#13370)

perheld
authored and
agrima1304
committed
Arm backend: Add logit decomposition pass and test (#13366)

authored and
agrima1304
committed
Arm backend: Initial int16 extension (#13318)

per
authored and
agrima1304
committed
Arm backend: Remove submodule serialization_lib (#13182)

authored and
agrima1304
committed
Arm backend: Enable semihosting in build_executor_runner.sh (#13325)

Erik-Lundell
authored and
agrima1304
committed
Fix torchao deps (#13107)

metascroy
authored and
agrima1304
committed
Revert "Add TorchAO wrapper config to allow filter_fn for quantize_ (#13264)" and "Add coreml quant recipes (#13265)" (#13374)

swolchok
authored and
agrima1304
committed
Introduce set input/output API to Module. (#13363)

shoumikhin
authored and
agrima1304
committed
Update Llama Example Readme (#12956)

anzr299
authored and
agrima1304
committed
Support for Running Arm Zephyr Toolchain on aarch64 Host Machines (#13335)

BujSet
authored and
agrima1304
committed
[ET-VK] Better work group sizes for matmul (#13378)

pytorchbot
authored and
agrima1304
committed
[ET-VK] Add mechanism to trigger command buffer re-encode only when necessary (#13379)

pytorchbot
authored and
agrima1304
committed
Introduce unload method API to Module. (#13364)

shoumikhin
authored and
agrima1304
committed
[Backend Tester] Clean up a few test issues (#13258)

GregoryComer
authored and
agrima1304
committed
[Backend Tester] Clean up report output (#13306)

GregoryComer
authored and
agrima1304
committed
[Backend Tester] Write report progressively (#13308)

GregoryComer
authored and
agrima1304
committed
create and validate build_variables.bzl (#8326)

swolchok
authored and
agrima1304
committed
[Backend Tester] Add subtest index field (#13311)

GregoryComer
authored and
agrima1304
committed
[Backend Tester] Reduce log verbosity / spam (#13312)

GregoryComer
authored and
agrima1304
committed
NXP Backend: Add infrastructure for pre processing passes in edge dialect (#13183)

MartinPavella
authored and
agrima1304
committed
Android preset (#11119)

kirklandsign
authored and
agrima1304
committed
Qualcomm AI Engine Direct - Static Decoder Runner Support 16bit KV IO (#13127)

winskuo-quic
authored and
agrima1304
committed
Qualcomm AI Engine Direct - Phase out QCIR flow since it is no longer… (#12583)

DannyYuyang-quic
authored and
agrima1304
committed
Bump Zephyr SDK Version in CI Image from v0.16.0 -> v0.17.2 (#13380)

BujSet
authored and
agrima1304
committed
[Backend Tester] Seed based on test name (#13313)

GregoryComer
authored and
agrima1304
committed
Add support for strongly typed op_quantized_linear_out

ethansfng
authored and
agrima1304
committed
removed cron schedule runs to workflow until the GitHub token issue is fixed (#13339)

nil-is-all
authored and
agrima1304
committed
Set a doc build variable for executorch version (#13351)

kirklandsign
authored and
agrima1304
committed
lintrunner -a backends/qualcomm/CMakeLists.txt (#13396)

swolchok
authored and
agrima1304
committed
Update the comments of calculate_numeric_gap

Gasoonjia
authored and
agrima1304
committed
Use dtype agnostic op_cat implementation, add op_cat testcases

ethansfng
authored and
agrima1304
committed
Arm backend: Allocate buffers with alignment (#13412)

Erik-Lundell
authored and
agrima1304
committed
NXP backend: Improve cifarnet speed by removing the initial pading. (#13279)

jirioc
authored and
agrima1304
committed
Arm backend: Generate ETRecord from arm_aot_compiler (#13273)

zingo
authored and
agrima1304
committed
Use unlifted export pass to tag delegated constants (#13407)

authored and
agrima1304
committed
Add support for strongly typed op_quantized_relu (#13345)

authored and
agrima1304
committed
[Backend Tester] Add test flow CLI arg (#13360)

GregoryComer
authored and
agrima1304
committed
Ensure the correct output data type for the full op.

eigen-k
authored and
agrima1304
committed
Arm backend: Add decomposition and test for acos (#13414)

emmakujala
authored and
agrima1304
committed
Arm backend: Introduce documentation for VGF (#13369)

robell
authored and
agrima1304
committed
Update ExecuTorchValue.mm (#13425)

shoumikhin
authored and
agrima1304
committed
Remove outdated comments and names in OSS

mcremon-meta
authored and
agrima1304
committed
Fix typo in op_quantized_relu_asym8u_asym8u

ethansfng
authored and
agrima1304
committed
Fix unused-local-typedef issue

r-barnes
authored and
agrima1304
committed
Refactoring Portable Operators to Standardize op_name Format (#12941)

BujSet
authored and
agrima1304
committed
Android scheduled build add vulkan (#13428)

kirklandsign
authored and
agrima1304
committed
NXP Backend: Buck fixes for the PassManager

digantdesai
authored and
agrima1304
committed
Add support for strongly typed op_quantized_matmul, generalize dispatch strategy

ethansfng
authored and
agrima1304
committed
Arm backend: use tosa_ref_model only if installed

digantdesai
authored and
agrima1304
committed
[executorch] Add TorchAO wrapper config to allow filter_fn for quantize_ (#13440)

authored and
agrima1304
committed
[executorch] Add coreml quant recipes (#13441)

authored and
agrima1304
committed
Static attention: do not specialize on input sequence length

sxu
authored and
agrima1304
committed
Fix typo in target file (#13443)

abhinaykukkadapu
authored and
agrima1304
committed
forward fix

cccclai
authored and
agrima1304
committed
Switch to conda-forge on MacOS (#13442)

huydhn
authored and
agrima1304
committed
use dtype agnostic implementation for non optimized op_permute_copy

ethansfng
authored and
agrima1304
committed
Extend `PyBundledModule` with `extension.BundledModule`

Gasoonjia
authored and
agrima1304
committed
Improve optimum coverage in ET (more models, xnnpack on mac) (#13400)

jackzhxng
authored and
agrima1304
committed
build_variables.bzl: split PROGRAM_NO_PRIM_OPS_SRCS from EXECUTORCH_CORE_SRCS (#8397)

swolchok
authored and
agrima1304
committed
Split quantized convolutions into NCHW and NHWC variants

mcremon-meta
authored and
agrima1304
committed
[ez] Fix idx in duplicate_constant_node_pass (#13461)

SS-JIA
authored and
agrima1304
committed
build_variables.bzl: split PLATFORM_SRCS from EXECUTORCH_CORE_SRCS (#8398)

swolchok
authored and
agrima1304
committed
Arm backend: Add missing using-declaration in VGFBackend.cpp (#13460)

YufengShi-dudu
authored and
agrima1304
committed
[EZ] Replace `pytorch-labs` with `meta-pytorch`

ZainRizvi
authored and
agrima1304
committed
Allow for HOP to be in the etreord graph

kimishpatel
authored and
agrima1304
committed
build_variables.bzl: split PATTERN_SRCS from PORTABLE_KERNELS_SRCS (#13353)

swolchok
authored and
agrima1304
committed
build_variables.bzl: make THREADPOOL_SRCS usable in BUCK (#13354)

swolchok
authored and
agrima1304
committed
build_variables.bzl: make MPS_BACKEND_SRCS usable in BUCK (#13355)

swolchok
authored and
agrima1304
committed
build_variables.bzl: make XNNPACK_BACKEND_SRCS usable in BUCK (#13356)

swolchok
authored and
agrima1304
committed
build_variables.bzl: make CUSTOM_OPS_SRCS usable in BUCK (#13357)

swolchok
authored and
agrima1304
committed
Expose portable ops as utils (add/stack)

manuelcandales
authored and
agrima1304
committed
Introduce out shape utils (add/stack) (#13199)

manuelcandales
authored and
agrima1304
committed
switch top-level ExecuTorch build from executorch_srcs.cmake to build_variables.bzl (#13358)

swolchok
authored and
agrima1304
committed
Alternative format specifier for %zd

cmt0
authored and
agrima1304
committed
[ET-VK] Move rotary embedding custom op to be handled via graph pass instead of source transform (#13465)

authored and
agrima1304
committed
Qualcomm AI Engine Direct - Static LLM Decoder Refactor (#13314)

winskuo-quic
authored and
agrima1304
committed
[ET-VK] Enable IntxWeightOnlyConfig (#13466)

pytorchbot
authored and
agrima1304
committed
Add aten::_upsample_bilinear2d_aa.out (#13458)

mergennachin
authored and
agrima1304
committed
Add support for strongly typed conv_nchw and conv_nhwc

ethansfng
authored and
agrima1304
committed
Arm backend: Add dim_order_ops:: to the auto gen_oplist generations (#13455)

zingo
authored and
agrima1304
committed
Arm backend: Replace .export_for_training with .export (#13280)

AdrianLundell
authored and
agrima1304
committed
Arm backend: Move TOSA operators to dialect (#13408)

per
authored and
agrima1304
committed
Arm backend: Add example linkerscripts for U55/U85 (#13404)

perheld
authored and
agrima1304
committed
Arm backend: Add cumsum support (#13457)

AdrianLundell
authored and
agrima1304
committed
Arm backend: Add partial vulkan runtime support for VgfPipeline (#13471)

YufengShi-dudu
authored and
agrima1304
committed
NXP backend: Remove optimization in fuse_quanitze_into_preceding_ops.py (#13372)

roman-janik-nxp
authored and
agrima1304
committed
NXP backend: Remove optimization in prune_cast_operators.py (#13377)

roman-janik-nxp
authored and
agrima1304
committed
NXP backend: Remove optimization in prune_reshape_operators.py (#13413)

roman-janik-nxp
authored and
agrima1304
committed
Remove outdated NCHW to NHWC pass and rename the current one to ReplaceConvWithChannelLastConvPass

mcremon-meta
authored and
agrima1304
committed
NXP backend: Improve target support checks. (#13367)

MartinPavella
authored and
agrima1304
committed
Add a default image prefiller implementation

larryliu0820
authored and
agrima1304
committed
Switch non-top-level ExecuTorch builds (size test, examples, etc.) from executorch_srcs.cmake to build_variables.bzl (#13391)

swolchok
authored and
agrima1304
committed
Enable BNNS copy for FP16 to FP32

metascroy
authored and
agrima1304
committed
Fix test-binary-size-linux -Wsign-compare failure with c10::irange (#13481)

swolchok
authored and
agrima1304
committed
Arm backend: Update tosa dialect buck file

digantdesai
authored and
agrima1304
committed
[ET-VK][ez] Move execute node threshold calculation from `prepare_pipelines()` to `prepare()` (#13497)

pytorchbot
authored and
agrima1304
committed
[ET-VK] Runtime support for NamedDataMap (#13498)

pytorchbot
authored and
agrima1304
committed
[ET-VK][AOT] Serialize constant tensors via NamedDataMap (#13499)

pytorchbot
authored and
agrima1304
committed
[ET-VK] Allocate memory for weight and activation tensors lazily (#13501)

pytorchbot
authored and
agrima1304
committed
[ET-VK][ez] Fix erroneous cherry-pick bot merge (#13512)

SS-JIA
authored and
agrima1304
committed
Arm backend: Add support for QAT+per-channel combo (#13511)

oscarandersson8218
authored and
agrima1304
committed
Reset Temp Allocator after each use

cmt0
authored and
agrima1304
committed
Stop validating that build_variables.bzl matches buck-generated executorch_srcs.cmake (#13392)

swolchok
authored and
agrima1304
committed
Only support int8 and quant dtypes for quant operators (#11685)

authored and
agrima1304
committed
Stop looking for buck2 in the top-level ExecuTorch build (#13393)

swolchok
authored and
agrima1304
committed
Summary: Add MCU model script to validate and run the models (#13439)

authored and
agrima1304
committed
Add a generic multimodal runner (#13166)

larryliu0820
authored and
agrima1304
committed
Force -O3 for executorch op_div.cpp in clang 19

akrieger
authored and
agrima1304
committed
Remove NTSTATUS cast

SamGondelman
authored and
agrima1304
committed
Clean up apparently-unnecessary mentions of BUCK2 in scripts (#13394)

swolchok
authored and
agrima1304
committed
Qwen and Phi-4-Mini targets (#13449)

rohansjoshi
authored and
agrima1304
committed
Wrap LLM runner tests in anonymous namespace (unbreak unittest-release) (#13524)

swolchok
authored and
agrima1304
committed
Delete extract_sources.py and cmake_deps.toml (#13395)

swolchok
authored and
agrima1304
committed
[Backend Tester] Add nightly CI job for XNNPACK (#13390)

GregoryComer
authored and
agrima1304
committed
Add Model Profiling Automation Script (#13493)

authored and
agrima1304
committed
Remove unused PROGRAM_SCHEMA_SRCS from build_variables.bzl (#13432)

swolchok
authored and
agrima1304
committed
removed lines of cron schedule runs until the github token issue is fixed (#13486)

nil-is-all
authored and
agrima1304
committed
[Backend Tester] Mark adaptive avgpool2d as an unsupported portable op (#13398)

GregoryComer
authored and
agrima1304
committed
[Backend Tester] Run Vulkan tests in nightly CI (#13445)

GregoryComer
authored and
agrima1304
committed
[Backend Tester] Run Core ML tests in nightly CI (#13446)

GregoryComer
authored and
agrima1304
committed
Improve softmax perf when transpose is not needed

mcremon-meta
authored and
agrima1304
committed
Remove other extensions' source files from EXTENSION_TRAINING_SRCS and TRAIN_XOR_SRCS (#13433)

swolchok
authored and
agrima1304
committed
Fix `buck query //extension/flat_tensor:` in OSS (#13484)

swolchok
authored and
agrima1304
committed
Qualcomm AI Engine Direct - Remove input_list dependencies (#13411)

chenweng-quic
authored and
agrima1304
committed
Qualcomm AI Engine Direct - GA Static Smollm2 (#13406)

chenweng-quic
authored and
agrima1304
committed
Remove pinning to ET commit for Zephyr CI job (#13388)

authored and
agrima1304
committed
Refactor pybinding unit test

cccclai
authored and
agrima1304
committed
Update project.pbxproj (#13537)

shoumikhin
authored and
agrima1304
committed
Fix Olmo trunk test and skip XNNPack trunk tests on mac (#13528)

jackzhxng
authored and
agrima1304
committed
Re-enable model tests with recipes for xnnpack backend (#13519)

abhinaykukkadapu
authored and
agrima1304
committed
QNN Llama Runner implement IRunner (#13171)

rohansjoshi
authored and
agrima1304
committed
Arm backend: Remove get_output_nodes from runner_utils. (#13417)

AdrianLundell
authored and
agrima1304
committed
Arm backend: Fix for combo neg(x)+1 + tests (#13517)

wwwind
authored and
agrima1304
committed
Adding smollm2 to examples/models/__init__.py (#13514)

SaoirseARM
authored and
agrima1304
committed
Arm backend: Add limited support for fish shell in setup_path.sh (#13323)

zingo
authored and
agrima1304
committed
Arm backend: Update examples/arm/README.md (#13546)

zingo
authored and
agrima1304
committed
Arm backend: Added VGF minimal example (#13545)

SaoirseARM
authored and
agrima1304
committed
Temporarily disable test-models-arm-zephyr (#13548)

digantdesai
authored and
agrima1304
committed
Fix buck cquery //extension/llm/runner: in OSS (#13527)

swolchok
authored and
agrima1304
committed
Remove unused sleef.h which breaks cross-compilation on windows (#13079)

authored and
agrima1304
committed
Bump the PyTorch pin to 20250811 (#13334)

swolchok
authored and
agrima1304
committed
Link QNN backend to pybinding lib when built (#13467)

GregoryComer
authored and
agrima1304
committed
Split on dilation for strongly typed convs

ethansfng
authored and
agrima1304
committed
Fix sigmoid operator to support boolean tensor inputs (#13515)

notkisk
authored and
agrima1304
committed
Fix method meta error in wasm build (#13496)

Conarnar
authored and
agrima1304
committed
[Backend Tester] Add test flows for QNN (#13469)

GregoryComer
authored and
agrima1304
committed
[Backend Tester] Add additional quantized test flows for XNNPACK and Vulkan (#13534)

GregoryComer
authored and
agrima1304
committed
[Backend Tester] Add markdown summary in CI (#13535)

GregoryComer
authored and
agrima1304
committed
Add support for strongly typed quantized_op_add

ethansfng
authored and
agrima1304
committed
Whisper audio processor

rohansjoshi
authored and
agrima1304
committed
Non-fatal error when ET_SWITCH encounters unsupported dtype

manuelcandales
authored and
agrima1304
committed
Update test_remove_unused_parameters_pass.py (#13563)

digantdesai
authored and
agrima1304
committed
Fix memory planning for greed with heuristic algo.

hsharma35
authored and
agrima1304
committed
Input position accessor for static attention IO manager

sxu
authored and
agrima1304
committed
Unbreak build-benchmark-app (apple) after pin bump in #13334 (#13582)

swolchok
authored and
agrima1304
committed
Summary: Add ExecutorchRuntimeException: Throw relevant exceptions from JNI layer in the event of errors (#13526)

authored and
agrima1304
committed
Update tokenizer to include tekken implementation (#13601)

mergennachin
authored and
agrima1304
committed
Add missing backslashes in example run section (#13603)

Conarnar
authored and
agrima1304
committed
Make IOManager use Module instead of Method. (#13542)

shoumikhin
authored and
agrima1304
committed
Access Method directly from TrainingModule. (#13602)

shoumikhin
authored and
agrima1304
committed
Make TensorPtr constructor check the data dize matches the shape. (#13591)

shoumikhin
authored and
agrima1304
committed
Add set_outputs() API. (#13609)

shoumikhin
authored and
agrima1304
committed
Create stale.yml workflow to label stale PRs (#13565)

nil-is-all
authored and
agrima1304
committed
Add get_output API. (#13610)

shoumikhin
authored and
agrima1304
committed
Set an empty EValue input for models that expect None arg. (#13621)

shoumikhin
authored and
agrima1304
committed
Split on depthwise for strongly typed convs

ethansfng
authored and
agrima1304
committed
migrate all test_aten_ops to facto

zonglinpeng
authored and
agrima1304
committed
fix MM nullptr from zero bias

zonglinpeng
authored and
agrima1304
committed
Call .detach() in static attention cache update helper

sxu
authored and
agrima1304
committed
Event Tracer Constraint

cmt0
authored and
agrima1304
committed
[ET-VK][ez] Fix partitioner logic of finding keepdim arg of reduce ops (#13598)

authored and
agrima1304
committed
[ET-VK][ez] Support grouped convolutions (#13599)

authored and
agrima1304
committed
[ET-VK][ez] Use XNNPACK's FuseBatchNorm pass (#13600)

authored and
agrima1304
committed
[ET-VK][testing] Add scripts to facilitate operator testiing (#13593)

authored and
agrima1304
committed
[ET-VK][ez] Consolidate tensor metadata calculation + buffer binding code (#13594)

authored and
agrima1304
committed
[ET-VK] Introduce `BufferMetadata` GLSL struct to abstract tensor layout (#13595)

authored and
agrima1304
committed
[ET-VK][ez] Allow high dimensional tensors (for buffer storage) (#13596)

authored and
agrima1304
committed
[ET-VK] High dim tensor support for view, unsqueeze, squeeze, clone (#13597)

authored and
agrima1304
committed
[Core ML] Improve asset management (#13560)

authored and
agrima1304
committed
Rename stale to stale.yml (#13619)

nil-is-all
authored and
agrima1304
committed
fix mismatch sub dtype (#13447)

cccclai
authored and
agrima1304
committed
NXP backend: Add support for the `aten.cat` operator. (#13505)

MartinPavella
authored and
agrima1304
committed
NXP backend: Add implementation of Tanh operator converter (#13510)

MartinPavella
authored and
agrima1304
committed
Arm backend: Dont try to fuse const for TOSA ops (#13575)

per
authored and
agrima1304
committed
NXP backend: Fix `tanh` merge conflict. (#13626)

MartinPavella
authored and
agrima1304
committed
Cortex_m backend: Loosen edge op check. (#13550)

Erik-Lundell
authored and
agrima1304
committed
NXP backend: Use zero point for quantized padding. (#13576)

MartinPavella
authored and
agrima1304
committed
Fix aten.amax lowering issue (#13381)

cccclai
authored and
agrima1304
committed
Update coremltools to 9b1 (#13614)

metascroy
authored and
agrima1304
committed
Add check_for_installed_private_headers_in_cmake_out (#13485)

swolchok
authored and
agrima1304
committed
NXP backend: Add support for conversion of Conv1D operator (#13549)

roman-janik-nxp
authored and
agrima1304
committed
Update cpuinfo pin to latest (#13624)

authored and
agrima1304
committed
Added JS bindings for tokenizers library (#13566)

Conarnar
authored and
agrima1304
committed
Run all periodic models when ciflow/periodic label is present (#13634)

shoumikhin
authored and
agrima1304
committed
Add support for data path in iOS (#13620)

lucylq
authored and
agrima1304
committed
Summary: Follow up fix to pr#13526 (#13640)

authored and
agrima1304
committed
Allow none and string input types for Method (#13645)

shoumikhin
authored and
agrima1304
committed
Resurface low level runtime API page (#13651)

mergennachin
authored and
agrima1304
committed
Fully enable the stale PR workflow (#13656)

nil-is-all
authored and
agrima1304
committed
Qualcomm AI Engine Direct - Scripts and accuracy improvement for Qwen3_0.6B/1.7B and Qwen 2.5_1.5B (#13544)

winskuo-quic
authored and
agrima1304
committed
Qualcomm AI Engine Direct - Improve GA Static Phi-4-mini accuracy (#13573)

shewu-quic
authored and
agrima1304
committed
Qualcomm AI Engine Direct - Fix broken unpacking in T5 dataset loading (#13625)

DannyYuyang-quic
authored and
agrima1304
committed
Fix error reporting in Windows preset build job (#13247)

GregoryComer
authored and
agrima1304
committed
Fix devtools CMake build failure on Windows (#13251)

GregoryComer
authored and
agrima1304
committed
Create Windows CMake preset (#13257)

GregoryComer
authored and
agrima1304
committed
Temporarily disable windows preset build in CI (#13669)

GregoryComer
authored and
agrima1304
committed
Increase binary size limit by 8 bytes (#13671)

shoumikhin
authored and
agrima1304
committed
NXP backend: Remove IR optimization to remove dead branches. (#13574)

MartinPavella
authored and
agrima1304
committed
Inline requantize kernels

mcremon-meta
authored and
agrima1304
committed
Smollm targets

rohansjoshi
authored and
agrima1304
committed
Override unload_method in training_module to erase the tensors pointing to the released memory

silverguo
authored and
agrima1304
committed
Disable mm + add -> addmm fusion if added tensor rank >2

hsharma35
authored and
agrima1304
committed
Fix bad optimized kernel for add.

hsharma35
authored and
agrima1304
committed
Allow zero-element inputs for method.

hsharma35
authored and
agrima1304
committed
Arm backend: Use cmake for building in Ethos-U jupyter example (#13630)

AdrianLundell
authored and
agrima1304
committed
NXP backend: Add MobileNetV2 example model and test (#12892)

authored and
agrima1304
committed
Arm Backend: Add support for ELU.default operator
agrima1304
committed
Arm Backend: Add support for ELU.default operator
agrima1304
committed