### Description
This PR adds SbgemmKernel for aarch64. This includes Sbegmm kernel to
implement matrix multiplication with bfloat16 SIMD instructions (bfmmla)
and MatMul operator changes to invoke the Sbgemm kernel. To enable
Sbgemm kernel, set the following session option:
"kOrtSessionOptionsGemmFastMathMode"
The PR also adds new test cases for mlas and ort.
### Motivation and Context
This is to improve MatMul performance on aarch64 platform.
I have run the below benchmarking script (bert , roberta and gpt2 model
inference) on AWS Graviton3 based c7g.4xl instance and observed 1.2x
-1.76x performance improvement compared to sgemm (fp32) kernel
performance.
```
cd onnxruntime/python/tools/transformers
python3 benchmark.py
```
And the unit test precision results are matching to sgemm kernel
results.
`./build.sh --config RelWithDebInfo --build_shared_lib --parallel
--compile_no_warning_as_error --skip_submodule_sync `
### Description
- Adds the following session options to configure the device:
- `soc_model`: The SoC model number. Refer to the QNN SDK documentation
for valid values. Defaults to "0" (unknown).
- `htp_arch`: The minimum HTP architecture the driver will use to select
compatible QNN operators.
- `device_id`: The ID of the device to use when setting 'htp_arch'.
Defaults to "0" (for single device).
### Motivation and Context
Allow more configuration.
Several changes:
1. To align with other EPs' setting of EP context configs in session
options, for example [QNN
EP](https://github.com/microsoft/onnxruntime/pull/18877), EP context
configs for TRT EP can be configured through:
1. Session Options: `ep.context_enable`, `ep.context_file_path` and
`ep.context_embed_mode`
2. Provider Options: `trt_dump_ep_context_model`,
`trt_ep_context_file_path` and `trt_dump_ep_context_embed_mode`
3. Above setting has 1:1 mapping and provider options has higher
priority over session options.
```
Please note that there are rules for using following context model related provider options:
1. In the case of dumping the context model and loading the context model,
for security reason, TRT EP doesn't allow the "ep_cache_context" node attribute of EP context node to be
the absolute path or relative path that is outside of context model directory.
It means engine cache needs to be in the same directory or sub-directory of context model.
2. In the case of dumping the context model, the engine cache path will be changed to the relative path of context model directory.
For example:
If "trt_dump_ep_context_model" is enabled and "trt_engine_cache_enable" is enabled,
if "trt_ep_context_file_path" is "./context_model_dir",
- if "trt_engine_cache_path" is "" -> the engine cache will be saved to "./context_model_dir"
- if "trt_engine_cache_path" is "engine_dir" -> the engine cache will be saved to "./context_model_dir/engine_dir"
```
2. User can decide the naming of the dumped "EP context" model by using
`trt_ep_context_file_path`, please see GetCtxModelPath() for more
details.
3. Added suggested comments from
https://github.com/microsoft/onnxruntime/pull/18217
Fix issue that the generated context cache model inputs/outputs order is not guaranteed
### Description
Currently, QNN EP generate the context cache model in Compile() method which only get access to the partitioned graph. And the inputs/outputs order for the partitioned graph is not guaranteed. And EP doesn't have the view of the input user model. Have to move the context cache model generation to a higher level in GraphPartitioner which has the view of the partitioned model.
This is also a break down of PR for multi-partition support.
https://github.com/microsoft/onnxruntime/pull/18865
### Description
<!-- Describe your changes. -->
Bump up version to 1.18.0 since the release branch has been cut.
### Motivation and Context
<!-- - Why is this change required? What problem does it solve?
- If it fixes an open issue, please link to the issue here. -->
Co-authored-by: rachguo <rachguo@rachguos-Mini.attlocal.net>
### Description
<!-- Describe your changes. -->
Add new option `trt_engine_cache_prefix` to customize TRTEP engine cache
prefix.
i.e:
- If user specifies `trt_engine_cache_prefix|FRCNN
trt_engine_cache_enable|true` when running FRCNN model
- the cache will be saved/loaded:
`FRCNN_2068723788287043730_*_sm80.engine`. Engine profile follows same
pattern.
- If skipping this option, the engine will be saved/loaded:
`TensorrtExecutionProvider_TRTKernel_graph_torch-jit-export_2068723788287043730_*_*_sm80.engine`
as default case.
### Motivation and Context
<!-- - Why is this change required? What problem does it solve?
- If it fixes an open issue, please link to the issue here. -->
https://github.com/microsoft/onnxruntime/issues/16708
---------
Co-authored-by: Chi Lo <Chi.Lo@microsoft.com>
Co-authored-by: Chi Lo <54722500+chilo-ms@users.noreply.github.com>
### Description
<!-- Describe your changes. -->
Pass through the ConfigOptions from the session via OpKernelInfo so that
kernel behavior can be configured.
Initial usage would be to optionally enable a fast path for ARM64 bloat16 GEMM - see #17031
Other usages could be things like selected the exact implementations of the activation functions for RNN operators instead of the default approximations (e.g. use [sigmoid_exact instead of sigmoid](2d6e2e243d/onnxruntime/core/providers/cpu/rnn/rnn_helpers.h (L379-L382)))
OpKernelInfo is already passing through things from the session state, and adding a new member of ConfigOptions
is the simpler update. It's also a more natural fit given it's providing state/info to the kernel.
### Motivation and Context
<!-- - Why is this change required? What problem does it solve?
- If it fixes an open issue, please link to the issue here. -->
### Description
Introduce AppendExecutionProvider_OpenVINO_V2 API and support for OV
2023.3.
### Context
- The API is added to facilitate customers in using published official
Microsoft onnxruntime libraries with OVEP libraries.
- Add support for OpenVINO 2023.3 official release.
- Extend operator coverage
- GH fixes
---------
Co-authored-by: Suryaprakash Shanmugam <suryaprakash.shanmugam@intel.com>
When the TRT engine cache (precompiled engine) is present, it doesn't
make sense to go over the processes of model verification, model
optimization, TRT EP's GetCapability(), TRT EP's model proto
reconstruction, calling TRT parser and engine compilation.
This PR makes TRT EP skip those processes and directly load the engine
to perform inference.
The feature request:
https://github.com/microsoft/onnxruntime/issues/18072
Features:
- Replace original model with TRT engine wrapped ONNX model. It can save
a lot of time as mentioned above.
- How to get TRT engine wrapped ONNX model?
1. Set `trt_dump_ep_context_model` provider option to "true" and run the
inference. You will find the "xxx_wrapper.onnx" at the engine cache
path. (The same logic of generating engine cache)
2. Use gen_trt_engine_wrapper_onnx_model.py
- Three provider options are added,
`trt_dump_ep_context_model`: Enable dump wrapped onnx model by TRT EP
`trt_ep_context_embed_mode`: Add embed_mode as attribute. 0 means engine
cache path, 1 means engine binary data.
`trt_ep_context_compute_capability_enable`: Add hardware_arch as
attribute. When running the model, TRT EP will check consistency between
model's hardware_arch and GPU's compute capability.
- When the engine cache path is given in the wrapped model, TRT EP will
first search for the engine file using the path (relative to model
path), if it can't find it, it will change to use the path as it is
(depends on user, could be relative to working dir or absolute path)
Note:
1. This PR includes the change of
https://github.com/microsoft/onnxruntime/pull/17751
Constraints:
1. The whole model should be fully supported by TRT.
4. Users need to make sure the engine is built with min/max/opt
optimization profiles that large enough to cover the range of all
inputs. TRT EP will simply fail and won't rebuild the engine if the
input shape is out of range during runtime.
### Description
This PR has several combined ORT ETW changes that improve ORT log
diagnosability & performance.
- The existing log behavior in the ORT API and Severity behavior remain
the same as compiled by the dev using the ORT API
- The PR keeps the existing design which has 2 TraceLogging providers
defined (although both were not used before this PR)
- Keeps great inference (inf) and session load performance even with
dynamic logging enabled (see below)
- On Windows, when ONNXRuntimeTraceLoggingProvider is enabled, then ORT
will dynamically _add_ a new sink reflecting the severity level provided
by ETW dynamically. E.G Critical - Verbose per the need at runtime
- This allows previous printf style LOGS() statements both for CPU and
NPU cases to flow to ETW via a local trace (if enabled)
- This DOES NOT add any new Telemetry which can optionally be sent to
Microsoft.
- Telemetry are ETW events marked with the Measure keyword that _can_ be
sampled if a box opts-in
- Existing Microsoft.ML.ONNXRuntime events have appropriate keywords and
levels added if they were missing
- If Execution Providers (EPs) can provide more detailed insight into
their HW or component, then this PR allows for those to be dynamically
logged instead of just at compile time
- In this PR, the QNN EP for QC NPUs can have basic or detailed
profiling enabled to give some insight into how the NPU is performing
- When the Microsoft.ML.ONNXRuntime ETW provider is enabled with the
Profiling keyword and level 5 then QC QNN basic profiling info is output
to ETW
### Motivation and Context
- This make ORT logging and diagnosability more performant (on Windows)
and available in a wider variety of runtime environments.
- The performance difference for inf times was ~300x+ drastically
better/faster when these logs were output to ETW vs just stdout (Verbose
Severity)
- This style of ETW dynamic tracing is the widely used standard for
Windows components, and even by some 3rd party software since the ETW
API is open and part of the Windows API
- These ETW based logs can be seamlessly combined with other ETW logs
such as an AI component/feature using ORT, OS CPU profiling, scheduling,
and more
- Before the PR, ORT logging is largely printf style and output to a
sink (usually stdout) only if compiled with a certain log Severity. When
enabled the previous logging (to stdout) would vastly slow down
inference times. Once compiled for release the internal ORT logs were
not accessible by anyone except the AI model developer in their dev
inner loop. That means logs could not be enabled on a lab machine, or on
a production system where the runtime behavior or performance might be
different for various reasons on a wide variety of HW.
- This change was tested with performance in mind and tested with a
mobilenet small AI model with onnxruntime_perf_test
- CPU: There was no statistical difference for inf times with the
baseline (main) or this PR whether ETW was enabled or not (both ORT
providers all keywords level 5).
- NPU (QNN on SP9 or Dev Kit 2023 QC SQ3): There was no statistical
difference for inf times with this PR whether ETW (both ORT providers
all keywords) were enabled or not for Level 5 (Verbose). This is even
with QNN Basic profiling turned on and outputting NPU stats to ETW
- As expected and designed, there was perf slowdown when Max Level 255
is enabled which translates to QNN Detailed profiling. This mirrors the
expected slowdown in the NPU when individual model operations are
monitored & recorded as well. This perf is similar to the QNN SDK
Detailed profiling performance separate from this PR. This is designed
to be above level 5 (verbose) as that is commonly the max level used in
many trace profiles and won't affect inf performance.
- Other OSes such as Linux & Android are left untouched for now.
- Out of scope for this PR but TraceLogging is available for Linux with
LTTng tracing. So in the future, this optional tracing could also be
made available on other OSes where a TraceLogging API is available
### Description
<!-- Describe your changes. -->
SessionOptions use_deterministic_compute can be set via the python API.
User request to enable setting via C API.
### Motivation and Context
<!-- - Why is this change required? What problem does it solve?
- If it fixes an open issue, please link to the issue here. -->
#17416
### Description
<!-- Describe your changes. -->
If we fail to calculate the buffer size (due to overflow) we currently
return a nullptr. This is inconsistent as an actual memory allocation
failure throws. An overflow would typically be due to bad input so an
exception makes more sense given that.
Change to throw so code using MakeUniquePtr* and AllocArray* doesn't
need to check for nullptr.
Add some extra info to the log message to help debugging.
### Motivation and Context
<!-- - Why is this change required? What problem does it solve?
- If it fixes an open issue, please link to the issue here. -->
Should help with #18905 by avoiding the invalid attempted usage of a
nullptr from the allocation. Extra info _might_ help with figuring out
where the overflow is coming from which is the real issue.
Move QNN EP provider options to session options
### Description
Need to use session option to support multi-partition for context cache feature. To smooth the transaction, move the provider options to session options first.
This is the first step for PR:
PR https://github.com/microsoft/onnxruntime/pull/18865
### Allow layer-wise recompute
Early, we need users/developers to specify the subgraphs to recompute,
now we introduced a more user-friendly way to enable recompute for all
detected stashed activation recomputation subgraphs. This scarifies
getting the best configs while makes it easier to support user
requirements when they switches from PyTorch per-layer gradient
checkpoint to ORTModule.
`ORTMODULE_MEMORY_OPT_LEVEL` is introduced to control the usage, by
default, it is 0, e.g. `USER_SPECIFIED`, all subgraphs definedin
`ORTMODULE_MEMORY_OPT_CONFIG` will be recomputed. So this is compatible
to existing recompute usage in ORTModule integrated models.
Using `ORTMODULE_MEMORY_OPT_LEVEL=1`, we will enable all recompute plans
detected, so those configs in `ORTMODULE_MEMORY_OPT_CONFIG` will not be
respected any more.
Add Unit Tests using 3 layer blooms.
https://github.com/microsoft/onnxruntime/blob/pengwa/add_aggresive_recompute/docs/Memory_Optimizer.md
### Description
Use InlinedVector<int64> instead of <int64_t,5> to reduce on the number
of template instantiations.
### Motivation and Context
The reported size reduction is small, just a few Ks. Just trying it out.
- Implement MLAS function for quantized 4-bit int Gemm (Gemm with float A and quantized 4-bit int B) for ARM NEON. This is an initial implementation. Only the M=1 path (with M being number of rows of A and C) has any optimization attempted so far. More optimization to come in future PRs.
- Connect MatMulNBits contrib op to MLAS function.
### Description
When and if `If` condition proves to be a constant value, inline the
corresponding subgraph yielding to more constant folding and
optimization.
### Motivation and Context
Newly converted models feature lots of nested `If` nodes that can be
inlined and collapsed.
In particular, for the sample models we are gaining on TorchScript
exported models.
For `HF Mobile Bert Dynamo` runtime went down from 0.069 -> 0.046. In
total, AOT inlining + `If` constant folding
yields improvement of about 50% 0.102 -> 0.046. Brining us very close to
TorchScript exported models.
`HF Bart Dynamo` further improves 0.668 -> 0.45. AOT + `If` constant
folding improves 0.98 -> 0.45
Earlier the size of
HF Mobile Bert **161Mb+**, now **98Mb**
HF Bart Dynamo pre-optimized model was about **1.2Gb**. It is now
**710MB**
### Description
<!-- Describe your changes. -->
Adding static int8 quantization support for MIGraphX Execution Provider
- Allows for parsing in calibration tables generated by Onnxruntime or
TensorRT's toolsets
- Add proper environment variables into the MIGraphX EP
- Update python API to include updating execution provider flags -> was
missing on python side
- Hook into MIGraphX's int8 quantitation and optimization of models
### Motivation and Context
<!-- - Why is this change required? What problem does it solve?
- If it fixes an open issue, please link to the issue here. -->
Required so that we can get onnxruntime to pass in models while
leveraging the existing tooling for int8 static QDQ quantization.
First step in a series of PRs which will add further static quantization
on the operator level as MIGraphX releases further support.
These changes drew heavily from the tensorRT EP should allow for similar
functionality for GPU based (versus CPU) quantization of models before
an inference is performed.
---------
Co-authored-by: Ted Themistokleous <tthemist@amd.com>
Co-authored-by: Ted Themistokleous <tedthemistokleous@amd.com>
Enable option qnn_context_priority to set QNN context priority, options:
"low", "normal", "normal_high", "high".
### Description
Enable option qnn_context_priority to set QNN context priority, options:
"low", "normal", "normal_high", "high".
This feature guarantees the model inference with higher priority. Tested
with onnxruntime_perf_test tool using same model.
1. Run the model on the NPU with single instance, the latency is 300ms.
2. Run the same model on NPU with 2 instance at same time.
Case 1:
both with same priority (high ) -- latency is 600ms
Case 2:
1 with low priority -- latency is 30,000ms
1 with high priority -- latency is 300ms
Case 3:
1 with normal priority -- latency is 15,000ms
1 with high priority -- latency is 300ms
### Description
Adds the QNN session option `htp_graph_finalization_optimization_mode`
to enable QNN graph optimizations at the expense of longer preparation
time.
### Motivation and Context
Allow enabling QNN graph optimizations per app/model.
### Description
Integration to OpenVINO 2023.1
### Motivation and Context
- Alignment with latest OpenVINO Version.
- Device name change from VPUX to NPU and Remove from supported list
until official public support is available.
---------
Co-authored-by: Sahar Fatima <sfatima.3001@gmail.com>
Co-authored-by: Saurabh Kale <saurabh1.kale@intel.com>
Co-authored-by: Suryaprakash Shanmugam <suryaprakash.shanmugam@intel.com>
Co-authored-by: sfatimar <sahar.fatima@intel.com>
I am adding a new `trt_timing_cache_path` option. Internally it is
handled as `global_cache_path_` and will be set via a fall through
approach:
1. no path provided => workdir
2. `trt_engine_cache_path` provided but no `trt_timing_cache_path` =>
`trt_engine_cache_path`
3. `trt_timing_cache_path` provided => `trt_timing_cache_path` (if not
provided `trt_engine_cache_path` will still be workdir)
### Motivation and Context
A TRT timing cache can be reused across multiple models as it only holds
kernel timings and it is common that network "patterns" are reused. This
can accelerate build times a lot.
---------
Co-authored-by: Carson M <carson@pyke.io>
Historically, DML was only able to fuse partitions when all sizes are
known in advance or when we were overriding them at session creation
time. But in practice, it should be possible to compile partitions at
compute time if the caller knows that the dimensions won't be changed
for every inference (e.g. resizing a webcam window, or padding the input
to powers of 2). This graph will be cached and reused until the sizes
change.
This is an opt-in option gated under the `enable_dynamic_graph_fusion`
option, which means that it will only be enabled when the caller
requests it since they have more context on how their model will be
called between inferences.
This PR also adds the option to disable metacommands from the python
API, which is an option for the C API but was lacking for python.
### Description
Inline functions in an EP aware fashion.
The result of this PR is that models that are having been inlined by
ONNX inliner and optimized and models that have been AOT inlined appear
to be visually identical.
For tests I used two models. The only difference is the resulting size
because ONNX inliner removes local function definitions and AOT does
not. Difference in sizes for `HF Mobile` model was 2.5 MB, and for `HF
Bart` it was ~500K. It seems that the resuling model size affects the
load time more than the actual optimizations.
In general, the inlined models grow in size very fast and can easily
exceed 2Gb limit.
Q. Should we make AOT optional?
`If` costant folding and the removal of local inlined models will be
coming in other PRs.
Some stats:
Expose a new allocator from cuda stream.
The allocator manages deferred cpu memory which only get recycled before
stream destruction.
---------
Co-authored-by: Randy Shuai <rashuai@microsoft.com>
### Description
CUDA inference speed heavily relies on Tensor Cores. To have tensor
cores achieve the optimal throughput they require the data layout to be
NHWC rather than NCHW.
### Motivation and Context
Especially for convolutional networks this is very important. I will
illustrate this using a very simple network:
```
import torch
import torch.nn as nn
class Net1(nn.Module):
def __init__(self):
super(Net1, self).__init__()
# 1 input image channel, 6 output channels, 5x5 square convolution
# kernel
self.m = nn.ModuleList([
nn.Conv2d(in_channels=8, out_channels=32, kernel_size=5, stride=1),
nn.Conv2d(in_channels=32, out_channels=64, kernel_size=3, stride=1),
nn.Conv2d(in_channels=64, out_channels=128, kernel_size=3, stride=1),
nn.Conv2d(in_channels=128, out_channels=128, kernel_size=3, stride=1, bias=False),
nn.Conv2d(in_channels=128, out_channels=128, kernel_size=3, stride=1, bias=False),
])
def forward(self, x):
for module in self.m:
x = module(x)
return x
if __name__ == "__main__":
dtype = torch.half
device = "cuda"
dummy_input = torch.randn(8, 8, 512, 512, dtype=dtype, device=device)
model = Net1().to(dtype=dtype, device=device)
input_names = ["input1"]
output_names = ["output1"]
torch.onnx.export(model, dummy_input, "test.onnx",
input_names=input_names, output_names=output_names)
```
I profiled the launch of `./build/RelWithDebInfo/onnxruntime_perf_test
-e cuda -I -q -t 5 test.onnx` using sys and nvtx ranges.
Current master launches below kernels:
If I add the introduced `-l` flag we see below kernels:
Notice the missing NCHW<>NHWC kernels per operation. The layout
optimizer introduced a transpose op as first and last op of the whole
network. The `op_generic_tensor_kernel` shows the bias used which should
also be optimized out next.
Measured across some very basic models:
| CUDA EP | **NCHW** [ms] | **NHWC** [ms] | Speedup |
|:------------------------|--------------------------------------:|-----------------------------------------:|------------------:|
| | -e cuda -t 5 -q | -e cuda -t 5 -q -l | |
| resnet101-v2-7_bs8_fp16 | 18.33 | 13.07 | 1.4 |
| resnet101-v2-7_bs8 | 21.8 | 12.06 | 1.81 |
| test | 102.07 | 73.62 | 1.39 |
Average speedup: 1.53
## Outlook
Next the mission will be to first write a templated unit test to check
for correctness of NHWC vs NCHW ops. After that we have to transition
more ops to measure perf improvements on a broader range of models.
Currently this is not easily possible as we can do not support all ops
in the NHWC domain.
---------
Co-authored-by: Tianlei Wu <tlwu@microsoft.com>
Support cross qk in beam search for whisper model and related features
Make whisper exporting tools support cross qk and some related features,
* extra_decoding_ids
* no_speech_prob
Implement DTW kernel, unfold tensor kernel with unit test Several fix
related with multiple session running parallel, like:
* guard multihead_attention, fused_fp16_runner_
* some memory allocation with stream awareness
* add use_ep_level_unified_stream option
### Description
Improve the QNN context binary cache feature to reduce the memory
overhead and initialization time overhead.
Instead of dumping a Qnn context binary file with metadata as header, we
dump a Onnx format file with metadata inside Onnx node.
### Motivation and Context
reduce the memory overhead and initialization time overhead
Two major modifications of this PR:
1. Refactor OrtTensorRTProviderOptions initialization and make it easy
to add new field.
2. Make Python API capable of using TensorRT plugins by adding new
Python binding api `register_tensorrt_plugins_as_custom_ops`. (It needs
to register ep's custom op domain before model load. For C++ API, it's
slightly different, when calling
SessionOptionsAppendExecutionProvider_TensorRT_XX, it appends cutom op
domain to session option. Later ORT can register custom op domain from
session option before model loading)
### Description
- Enables option to use the QNN Saver backend for dumping QNN API calls
to file.
- Adds logic to read environment variable
`ORT_UNIT_TEST_ENABLE_QNN_SAVER` from QNN EP unit tests. If enabled,
unit tests will use the QNN Saver backend and dump files to
`./saver_output/`.
### Motivation and Context
QNN Saver makes it easier to debug issues when unit tests fail. The
output files generated by QNN Saver can be used to replay the exact QNN
API calls that lead to a specific error condition.
QNN Saver dumps QNN API calls (and weights) to disk.
- saver_output/saver_output.c: C file containing all QNN API calls.
- saver_output/params.bin: binary file containing all
input/output/parameter tensor data provided during tensor creation, op
config validation, and graph execution.
Enabling the QNN Saver backend has 2 note-worthy effects:
1. All QNN API calls will succeed.
2. Inference output returns dummy data.
Because the output files from QNN Saver are always overwritten, it is
recommended to run individual unit tests via the `--gtest_filter`
command-line option.
Example (linux):
```shell
$ ORT_UNIT_TEST_ENABLE_QNN_SAVER=1 ./onnxruntime_test_all --gtest_filter=QnnHTPBackendTests.Resize_DownSample_Linear_AlignCorners
```
### Description
Add support for specifying a custom logging function per session.
Bindings for other languages will be added after this PR is merged.
### Motivation and Context
Users want a way to override the logging provided by the environment.
### Description
<!-- Describe your changes. -->
* Allow either an allocator or a MemBuffer to be used when creating an
OrtValue from an TensorProto
* `Tensor<std::string>` requires an allocator to allocate/free the
string values
* Forcing the buffer to be allocated outside of the Tensor doesn't seem
to provide any benefit in this usage as the Tensor class disables copy
and assignment (so we wouldn't create 2 copies of the buffer via the
Tensor class that externally managing the would buffer avoid)
* New approach means we don't need to manage the buffers in the
optimizer Info class as the Tensor dtor will do that
* Update naming - MLValue was replaced by OrtValue a long time ago
### Motivation and Context
<!-- - Why is this change required? What problem does it solve?
- If it fixes an open issue, please link to the issue here. -->
#17392
### Description
Remove `Resolve()` on the entire graph as each function is resolved.
We retain `Resolve()` after each inlining iteration.
### Motivation and Context
Poor performance for inlining the model and session initialization.
Original model before Resolve() removal
FunctionTest.Profiling (**65953 ms**)
After Resolve() Removal
FunctionTest.Profiling (**2911 ms**)
RelWithDebInfo pre-inlined model. Presumably because it runs Level1
optimizers
Non-inlined model consists of functions and Level1 optimizers have no
effect.
FunctionTest.Profiling (**9851 ms**)
### Description
<!-- Describe your changes. -->
Make status.h independent from gsl.
### Motivation and Context
<!-- - Why is this change required? What problem does it solve?
- If it fixes an open issue, please link to the issue here. -->
In the coming new feature external EP API (see the prototype
https://github.com/microsoft/onnxruntime/pull/16718), we need to expose
stream in the public header, however, stream is dependent on status.h
which is dependent on gsl. We are seeking a way to decouple stream from
gsl.
From Changming's comment offline, prefast is disabled so all
GSL_SUPPRESS are not taking any effect now. He will handle the warnings
when enable prefast in the future
### Description
Add new name "WebGPU_Buffer" to OrtMemoryInfo.
This is one of the prerequisites for supporting IO binding for WebGPU
buffer in onnxruntime-web.
list of prerequisites PRs:
#17465#17469 (this one)
In flatbuffers@v23.5.9 was broken forward declaration for
FlatBufferBuilder. Trying to compile onnxruntime falls with the
following error:
```
flatbuffers/include/flatbuffers/flatbuffer_builder.h:1420:38: error: typedef redefinition with different types ('FlatBufferBuilderImpl<false>' vs 'flatbuffers::FlatBufferBuilder')
typedef FlatBufferBuilderImpl<false> FlatBufferBuilder;
^
onnx_runtime/include/onnxruntime/core/graph/graph.h:47:11: note: previous definition is here
class FlatBufferBuilder;
```
This PR removes these declarations and puts includes instead
### Fix build - redefinition of default argument for ‘long unsigned int
Extent’
One of the training customer env, building ORT, there is such a build
error. The GCC version are
```
aiscuser@node-0:/tmp/onnxruntime$ gcc --version
gcc (Ubuntu 9.4.0-1ubuntu1~20.04.1) 9.4.0
aiscuser@node-0:/tmp/onnxruntime$ g++ --version
g++ (Ubuntu 9.4.0-1ubuntu1~20.04.1) 9.4.0
```
But on our dev node using same GCC/G++, we don't have build issue., not
sure what's the difference but giving an explict type when creating
`gsl::span` fixed the problem.
```
/tmp/onnxruntime/build/Linux/RelWithDebInfo/_deps/gsl-src/include/gsl/span:394:7: error: redefinition of default argument for ‘long unsigned int Extent’
394 | class span
| ^~~~
/tmp/onnxruntime/build/Linux/RelWithDebInfo/_deps/gsl-src/include/gsl/span_ext:46:51: note: original definition appeared here
46 | template <class ElementType, std::size_t Extent = dynamic_extent>
| ^~~~~~~~~~~~~~~
/tmp/onnxruntime/include/onnxruntime/core/common/span_utils.h:82:93: error: return type ‘class gsl::span<const std::byte>’ is incomplete
82 | [[nodiscard]] inline gsl::span<const std::byte> AsByteSpan(const void* data, size_t length) {
| ^
/tmp/onnxruntime/include/onnxruntime/core/common/span_utils.h: In function ‘void onnxruntime::AsByteSpan(const void*, size_t)’:
/tmp/onnxruntime/include/onnxruntime/core/common/span_utils.h:83:68: error: class template argument deduction failed:
83 | return gsl::span(reinterpret_cast<const std::byte*>(data), length);
| ^
/tmp/onnxruntime/include/onnxruntime/core/common/span_utils.h:83:68: error: no matching function for call to ‘span(const std::byte*, size_t&)’
/tmp/onnxruntime/build/Linux/RelWithDebInfo/_deps/gsl-src/include/gsl/span:740:1: note: candidate: ‘template<class Type, long unsigned int Extent> gsl::span(Type (&)[Extent])-> gsl::span<ElementType, FirstExtent>’
740 | span(Type (&)[Extent]) -> span<Type, Extent>;
| ^~~~
/tmp/onnxruntime/build/Linux/RelWithDebInfo/_deps/gsl-src/include/gsl/span:740:1: note: template argument deduction/substitution failed:
/tmp/onnxruntime/include/onnxruntime/core/common/span_utils.h:83:68: note: mismatched types ‘Type [Extent]’ and ‘const std::byte*’
83 | return gsl::span(reinterpret_cast<const std::byte*>(data), length);
| ^
/tmp/onnxruntime/build/Linux/RelWithDebInfo/_deps/gsl-src/include/gsl/span:743:1: note: candidate: ‘template<class Type, long unsigned int Size> gsl::span(std::array<_Tp, _Nm>&)-> gsl::span<ElementType, FirstExtent>’
743 | span(std::array<Type, Size>&) -> span<Type, Size>;
| ^~~~
/tmp/onnxruntime/build/Linux/RelWithDebInfo/_deps/gsl-src/include/gsl/span:743:1: note: template argument deduction/substitution failed:
/tmp/onnxruntime/include/onnxruntime/core/common/span_utils.h:83:68: note: mismatched types ‘std::array<_Tp, _Nm>’ and ‘const std::byte*’
83 | return gsl::span(reinterpret_cast<const std::byte*>(data), length);
| ^
```
### Motivation and Context
<!-- - Why is this change required? What problem does it solve?
- If it fixes an open issue, please link to the issue here. -->
### Description
<!-- Describe your changes. -->
- allocation planner was breaking if graph had no nodes
- in this particular model a branch of an If node returned an outer
scope value directly.
- if model used non-tensor types and sparse tensors are disabled the
call to IsSpareTensor causes an exception when prematurely terminates
the code.
- it's perfectly fine to check if a value is a sparse tensor when
support for them is disabled. we just can't do anything with that
OrtValue which is what the current ifdef's after the call to
IsSparseTensor handle.
### Motivation and Context
<!-- - Why is this change required? What problem does it solve?
- If it fixes an open issue, please link to the issue here. -->
Fix model execution failure for partner with model that uses sequences
in a minimal build with sparse tensors disabled.
Prevent `GSL_SUPPRESS` arguments from being modified by clang-format and update existing usages.
clang-format was changing something like `GSL_SUPPRESS(r.11)` to `GSL_SUPPRESS(r .11)`.
For some compilers (e.g., clang), the `gsl::suppress` attribute takes a quoted string argument. We don't want to insert spaces there.
### Description
When functions are inlined and constant nodes are being converted to
initializers, we need to create NodeArg for them.
Similar for inlined function subgraph, but we choose to give priority to
non-constant nodes and then fill the gaps with constant and
initializers.
### Motivation and Context
This addresses issue
https://github.com/microsoft/onnxruntime/issues/16813 for
`eca_halonext26ts_mod.onnx` model where it fails to remove unused
initializer because `NodeArg` was not created for it.
### Description
Re-implement stacktrace. The new implementation doesn't directly use
Windows API, hence can avoid problems regarding to
initialize/uninitialize the dbghelp library.
### Motivation and Context
### Description
This change upgrade emsdk to 3.1.44.
Because backend is upgraded to LLVM 16, so need to fix a lot of build
failures caused by "-Wshorten-64-to-32".
most of the build failures comes from generated `onnx.pb.h`, and this
can be fixed by including "core/graph/onnx_protobuf.h", which detects
and ignore shorten-64-to-32 warnings.
Add a generic `UpdateCUDAProviderOptionsWithValue()` C API to update
CUDA EP provider options where its data type is pointer that can't be
represented by string.
Note: Please see some comments for the similar [PR
](https://github.com/microsoft/onnxruntime/pull/16965)for TRT EP.
Add a generic `UpdateTensorRTProviderOptionsWithValue()` C API to update
TensorRT provider options where its data type is pointer that can't be
represented by string.
### Description
Added Gelu operator to JSEP
### Motivation and Context
<!-- - Why is this change required? What problem does it solve?
- If it fixes an open issue, please link to the issue here. -->
### Description
A [previous PR](https://github.com/microsoft/onnxruntime/pull/16531)
added a temporary directory to save the model optimizations after
loading a model into an `InferenceSession`. Many models that have an
external data file, however, require the data file to be in the same
directory as the ONNX model file. Because the model is saved in a
temporary directory and the data is saved in another directory, this
causes a `FileNotFoundError` error when trying to load the model in the
temporary directory.
This PR fixes this error by saving the external data file in the same
directory that the optimized model is located in.
### Motivation and Context
This PR fixes a bug with using a temporary directory while running the
optimizer for models that have an external data file.
#16506 Cause almost every translation units on linux complaint
```
[1175/1235] Building CXX object CMakeFiles/onnxruntime_test_all.dir/home/guangyunhan/onnxruntime/orttraining/orttraining/test/training_ops/cuda/softmax_test.cc.o
In file included from /home/guangyunhan/onnxruntime/include/onnxruntime/core/framework/float16.h:18,
from /home/guangyunhan/onnxruntime/include/onnxruntime/core/framework/data_types.h:17,
from /home/guangyunhan/onnxruntime/include/onnxruntime/core/framework/tensor.h:17,
from /home/guangyunhan/onnxruntime/onnxruntime/test/common/tensor_op_test_utils.h:16,
from /home/guangyunhan/onnxruntime/onnxruntime/test/providers/compare_provider_test_utils.h:7,
from /home/guangyunhan/onnxruntime/orttraining/orttraining/test/training_ops/cuda/softmax_test.cc:4:
/home/guangyunhan/onnxruntime/include/onnxruntime/core/session/onnxruntime_float16.h: In instantiation of ‘static constexpr uint16_t onnxruntime_float16::Float16Impl<Derived>::ToUint16Impl(float) [with Derived = onnxruntime::MLFloat16; uint16_t = short unsigned int]’:
/home/guangyunhan/onnxruntime/include/onnxruntime/core/framework/float16.h:42:66: required from here
/home/guangyunhan/onnxruntime/include/onnxruntime/core/session/onnxruntime_float16.h:241:7: note: ‘union onnxruntime_float16::detail::float32_bits’ has no user-provided default constructor
241 | union float32_bits {
| ^~~~~~~~~~~~
/home/guangyunhan/onnxruntime/include/onnxruntime/core/session/onnxruntime_float16.h:242:16: note: and the implicitly-defined constructor does not initialize ‘unsigned int onnxruntime_float16::detail::float32_bits::u’
242 | unsigned int u;
| ^
```
This PR shut the compiler up.
### Description
Introduce `Float16/BFloat16` support for C# and C++ APIs.
User should be able to perform conversions from `float` to/from
`Float16/BFloat16`, compare values and tests for `NaN, Inifnity, and
whether the number is denormalized.`
### Motivation and Context
User filed issues such as:
https://github.com/microsoft/onnxruntime/issues/14303
### Description
clean unused parameter in ORT_UNUSED_PARAMETER
### Motivation and Context
clean unused parameters in ORT_UNUSED_PARAMETER which are introduced
from #15833
- Fix some warnings from Xcode build (`-Wshorten-64-to-32`).
- Enable `-Wshorten-64-to-32` warning if available. Currently it's not fully enabled for `onnxruntime_test_all` and `onnxruntime_providers_xnnpack` yet.
- Some clean up in build.py including setting CMake generator more consistently.
### Description
C API for custom ops does not support float 8 types. This PR changes
that.
### Motivation and Context
The list of operators supporting float 8 is very limited. It should be
extended to custom ops to let developpers add customized operators for
these specific types.
### Description
Remove AllocatorManager class
### Motivation and Context
After the refactor PR #15833 is in, AllocatorManager class is not
referenced anymore.
### Description
This PR implements a backward-compatible way to define custom operators
with fallible compute functions. The C++ API templated gained an
optional `Fallible` argument. Closes#14287
### Motivation and Context
#14287 contains more context. The gist is that the current C-API defines
compute operations of custom operators as functions returning `void`
rather than an `OrtStatusPtr`. Currently, errors are often propagated
across the C-ABI using C++ exceptions. That is very unsafe and undefined
behavior. Moreover, it is difficult for languages other than C++ to use
this approach even if they wanted to. A C-compliant sound and safe way
to propagate errors allows for non-C++ fallible custom operators.
### An example in action
https://github.com/cbourjau/ort-custom-op/pull/6/files is a
demonstration of how this PR can be used to write safe and fallible
custom operators in Rust.
CUDA EP already supports [CUDA
graph](https://docs.nvidia.com/cuda/cuda-c-programming-guide/index.html#cuda-graphs),
also we observed some models can benefit from using CUDA graph with
`trtexec`. Therefore, this PR enables the CUDA graph support for TRT EP.
The implementation is based on
https://github.com/microsoft/onnxruntime/pull/9978 with the same
[constraints](https://github.com/microsoft/onnxruntime/pull/9978) as
below:
- Models with control-flow ops (i.e. If, Loop and Scan ops) are not
supported.
- Usage of CUDA Graphs is limited to models where-in all the model ops
(graph nodes) can be partitioned to the TRT EP.
- The input/output types of models need to be tensors.
- Shapes of inputs/outputs cannot change across inference calls.
- IObinding is required.
### Description
<!-- Describe your changes. -->
1. Add a new test lib `onnxruntime_providers_cuda_ut` which is similar
to `onnxruntime_providers_cuda` but `onnxruntime_providers_cuda_ut` is
only built if `onnxruntime_BUILD_UNIT_TESTS` is set. We can call all
CUDA UTs through this ut lib without affecting production lib
`onnxruntime_providers_cuda`.
2. Move all test cases from `core/providers/cuda/test/` to
`test/providers/cuda/`. These test cases are built into lib
`onnxruntime_providers_cuda_ut` and run by `./onnxruntime_test_all
--gtest_filter="*CUDA_EP_Unittest*"`. Since the lib is only for test, we
can use gtest macros in the test cases. Previous implementation do not
support using gtest lib in the CUDA UT cases.
3. The cmake code in `cmake/onnxruntime_providers.cmake` is refactored a
bit. A new function `onnxruntime_add_object_library` is to build a
object target. The 2 libs `onnxruntime_providers_cuda_ut` &
`onnxruntime_providers_cuda` share most of the code, so the object files
can be used in both libs, which helps reduce build time. Another
function `config_cuda_provider_shared_module` is used to configure all 3
similar
targets(onnxruntime_providers_cuda_obj/onnxruntime_providers_cuda/onnxruntime_providers_cuda_ut).
4. Refactored the test to call `testing::InitGoogleTest` &
`RUN_ALL_TESTS` in `libonnxruntime_providers_cuda_ut.so`'s `TestAll`.
After this change, we can see all the cases running in
`CUDA_EP_Unittest.All`:
### Motivation and Context
<!-- - Why is this change required? What problem does it solve?
- If it fixes an open issue, please link to the issue here. -->
After https://github.com/microsoft/onnxruntime/pull/13016, there are
still test files in test/providers/cuda/ that are not moved to
core/providers/cuda/test/ and the test cases are disabled. This PR helps
to clean the unfinished TODOs.
Even through onnxruntime_shared_lib_test covers some test for CUDA
provider. onnxruntime_shared_lib_test works like a coarse grain
end-to-end test for CUDA provider. If CUDA unittest can run cases for a
single component, this wound be helpful for CUDA developers.
---------
Co-authored-by: Yuhong Guo <yuhong.gyh@antgroup.com>
### Description
This PR is to refactor ExecutionProvider API for memory management,
which is to move allocators from EP level to SessionState level and
indexed by OrtDevice
### Motivation and Context
<!-- - Why is this change required? What problem does it solve?
- If it fixes an open issue, please link to the issue here. -->
This PR is to refactor ExecutionProvider API for memory management,
which is to move allocators from EP level to SessionState level and
indexed by OrtDevice. By this change, EP level will shift the burden of
maintaining allocators, which will be user friendly for EP developers
---------
Co-authored-by: Lei Cao <leca@microsoft.com@orttrainingdev8.d32nl1ml4oruzj4qz3bqlggovf.px.internal.cloudapp.net>
### Description
1. Use IAllocatorUniquePtr to replace BufferUniquePtr. It will ensure
the deleter is always right.
2. Change some std::unique_ptr to std::optional
3. Bypass Arena allocator when allocating the prepack buffers for mlas.
In this special case, Arena doesn't help any. And this change is just an
internal implementation change, it doesn't affect our public interface.
For TunableOp, some instance may has very bad performance and it will
take a long time during profile process.
Add `tunable_op_max_tuning_duration_ms` parameter to limit max tuning
time.
### Description
The PR implements FloatE4M3FN, FloatE5M2, FloatE4MEFNUZ, FloatE5M2FNUZ
as described in PR https://github.com/onnx/onnx/pull/4805. It uses CUDA
API to cast float/half to float8 if CUDA>=11.8, a custom implementation
if CUDA<11.8.
* It implements, Cast, QuantizeLinear, DequantizeLinear for all types on
CPU, only for types FloatE4M3FN, FloatE5M2 on CUDA.
* It extends the supported types for control flow operator, Shape,
Reshape, Identity, If, Loop, Scan, Reshape
* It implements Equal(19).
* Cast, QuantizeLinear, DequantizeLinear operators now support a
parameter `saturate` only valid for float 8 types. It is true by
default. In that case, any value out of range is converted into the
maximum float 8 value. If false, it is infinite.
* QuantizeLinear, DequantizeLinear now supports multiple scales on CUDA
(and ROCm by extension), scale = 1D tensor with one scale per channel
### Motivation and Context
Supports latest onnx version.
Fixes
[AB#15395](https://aiinfra.visualstudio.com/6a833879-cd9b-44a4-a9de-adc2d818f13c/_workitems/edit/15395)
---------
Co-authored-by: Xavier Dupre <xadupre@microsoft.com@orttrainingdev8.d32nl1ml4oruzj4qz3bqlggovf.px.internal.cloudapp.net>
Co-authored-by: Randy Shuai <rashuai@microsoft.com>
Co-authored-by: Edward Chen <18449977+edgchen1@users.noreply.github.com>
Co-authored-by: Scott McKay <Scott.McKay@microsoft.com>
### Description
The file include/onnxruntime/core/providers/cuda/cuda_provider_options.h
is a C++ file. It is not for C.
Before this commit, this header file is already not compatible with C compilers. Because it has:
```
onnxruntime::ArenaExtendStrategy arena_extend_strategy;
```
And this file is intended to be internal only. It is an internal header file. It should not be included in onnxruntime_c_api.h and should not be used with the public C APIs. User can only get the instance of OrtCUDAProviderOptionsV2 via CreateCUDAProviderOptions. In such a way we can add new members to this struct without breaking binary compatibility.
Since it is an internal header, we can safely use C++ grammar there.
Add a configuration `max_power_of_two_extend_bytes ` to limit the arena extension size.
### Motivation and Context
<!-- - Why is this change required? What problem does it solve?
- If it fixes an open issue, please link to the issue here. -->
In our real scenario, we observe that if the model is big enough the
BfcArena will extend uncontrollable.
As showed by the following figures, if a model uses more than 16GB
memory, the BfcArena will totally apply for 32GB memory according to the
`kNextPowerOfTwo` strategy. With the new strategy, the extension is
limited. The default maximum extension size is 1GB.
#### Without the new configuration
After loading the model, ORT uses 32G GPU memory.
#### With the new configuration
After loading the model, ORT uses 23G GPU memory.
Co-authored-by: Yuhong Guo <yuhong.gyh@antgroup.com>
### Description
Adds the session config option `disable_cpu_ep_fallback` to allow the
user to prevent the CPU EP from handling
nodes not supported by other execution providers.
```C++
// Graph nodes that are not supported by the execution providers (EPs) explicitly added to the session are
// assigned (i.e., "fallback") to the CPU EP by default.
//
// This option allows the user to disable the fallback of unsupported graph nodes to the CPU EP.
// If this option is set to "1", session creation will fail if the execution providers other than the CPU EP cannot
// fully support all of the nodes in the graph.
//
// It is invalid to set this option and explicitly add the CPU EP to the session. In this case, session creation
// will also fail with an error.
//
// Option values:
// - "0": CPU EP fallback is not disabled. [DEFAULT]
// - "1": CPU EP fallback is disabled.
static const char* const kOrtSessionOptionsDisableCPUEPFallback = "session.disable_cpu_ep_fallback";
```
#### Example use
```C++
#include "core/session/onnxruntime_cxx_api.h"
#include "core/session/onnxruntime_session_options_config_keys.h"
int main(int argc, char** argv) {
Ort::SessionOptions so;
so.AddConfigEntry(kOrtSessionOptionsDisableCPUEPFallback, "1"); // Disable fallback to the CPU EP.
onnxruntime::ProviderOptions options;
#if defined(_WIN32)
options["backend_path"] = "QnnCpu.dll";
#else
options["backend_path"] = "libQnnCpu.so";
#endif
so.AppendExecutionProvider("QNN", options);
const ORTCHAR_T* ort_model_path = ORT_MODEL_FOLDER "qnn_ep_partial_support.onnx";
Ort::Session session(*ort_env, ort_model_path, so); // Throws exception if nodes fallback to CPU
// ...
```
### Motivation and Context
Makes it easier for application developers to ensure that the entire
model runs on specific EPs. This is critical for Qualcomm/scenarios. If
the compute cannot be offloaded to the NPU, running on CPU is not
acceptable. (could be the difference between 90 second inference and 6
seconds inference)
---------
Co-authored-by: Pranav Sharma <prs@microsoft.com>
### Description
Enable Qnn Context cache feature to save model initialization time
Provider options:
qnn_context_cache_enable|1 to enable the cache feature
qnn_context_cache_path to set the cache path. It is set to model_file.onnx.bin by default.
### Motivation and Context
Model initialization time takes long because the cost of conversion from Onnx model to Qnn model. Qnn have feature to serialize the Qnn context to file, then next time user can load it from the cache context and execute the graph to save the cost.
---------
Co-authored-by: Adrian Lizarraga <adlizarraga@microsoft.com>
This PR mainly fixes building errors when trying to build nupkg for ROCm EP.
It also slighly improve the packaging logic so that devlopers can
produce the nupkg on linux natively.
