GitLab

Fix an unexpected result value (`-0.125`) issue in ib traffic benchmark when encountering `-1` in raw output
* Check if the value is valid before the base conversion
* Add a test case to cover this situation

Fix potential port conflict due to race condition between time-to-check
to time-to-use, by binding the port all through.

Modify the function to resolve flake8 C901 while keeping the logic same.

Fix several issues in ib validation benchmark:
* continue running when timeout in the middle, instead of aborting whole mpi process
* make timeout parameter configurable, set default to 120 seconds
* avoid mixture of stdio and iostream when print to stdout
* set default message size to 8M which will saturate ib in most cases
* fix hostfile path issue so that it can be auto found in different cases

Fix several issues in ib loopback benchmark:
* use `--report_gbits` and divide by 8 to get GB/s, previous results are
  MiB/s / 1000
* use the ib_write_bw binary built in third_party instead of system path
* update the metrics name so that different hca indices have same metric

**Description**

Support multiple IB/GPU devices run simultaneously in ib validation benchmark.

**Major Revisions**
- Revise ib_validation_performance.cc so that multiple processes per node could be used to launch multiple perftest commands simultaneously. For each node pair in the config, number of processes per node will run in parallel.
- Revise ib_validation_performance.py to correct file paths and adjust parameters to specify different NICs/GPUs/NUMA nodes.
- Fix env issues in Dockerfile for end-to-end test.
- Update ib-traffic configuration examples in config files.
- Update unit tests and docs accordingly.

Closes #326.

**Description**

Fix cmake and build issues.

**Major Revision**

* Remove unnecessary boost build
* Remove user-agent for mlc
* Remove -j for third party to build each project in sequence
* Fix ansible collections installation path

**Description**

Support `sb run` on host directly without Docker

**Major Revisions**
- Add `--no-docker` argument for `sb run`.
- Run on host directly if `--no-docker` if specified.
- Update docs and tests correspondingly.

**Description**

Cherry-pick  bug fixes from v0.5.0 to main.

**Major Revisions**

* Bug - Force to fix ort version as '1.10.0' (#343)
* Bug - Support no matching rules and unify the output name in result_summary (#345)
* Analyzer - Support regex in annotations of benchmark naming for metrics in rules (#344)
* Bug - Fix bugs in sync results on root rank for e2e model benchmarks (#342)
* Bug - Fix bug of duration feature for model benchmarks in distributed mode (#347)
* Docs - Upgrade version and release note (#348)
Co-authored-by: Yuting Jiang <v-yutjiang@microsoft.com>

**Description**
Integrate FAMBench into superbench based on docker implementation:
https://github.com/facebookresearch/FAMBench

The script to run all benchmarks is:
https://github.com/facebookresearch/FAMBench/blob/main/benchmarks/run_all.sh

**Description**
Use config `log_raw_data` to control whether log the raw data into file or not. The default value is `no`. We can set it as `yes` for some particular benchmarks to save the raw data into file, such as NCCL/RCCL test.

**Description**
Remove fp16 samples type converting time for training cnn and lstm inference.

**Description**
Remove fp16  samples type converting time for cnn and lstm models.

**Description**
Modifications adding GPU-Burn to SuperBench.
- added third party submodule
- modified Makefile to make gpu-burn binary
- added/modified microbenchmarks to add gpu-burn python scripts
- modified default and azure_ndv4 configs to add gpu-burn

**Description**
The BatchNorm operator is not numerically stable in fp16.  PyTorch documentation recommends to keep the BN op in fp32 for fp16 AMP models.  Refer to https://pytorch.org/docs/stable/amp.html#ops-that-can-autocast-to-float32.  Preserving BN in fp32 for superbench more accurately reflects real workloads.

**Description**
Fix invalid reference of P2P detection result in gpu_copy.

**Description**
Fix HIP_ARCHITECTURES is empty issue with cmake>=3.21.0.
Refer to https://github.com/ROCm-Developer-Tools/HIP/pull/2364

**Description**
Add support for pytorch>=1.9.0 of init_process_group.

**Major Revision**
- Use PrefixStore(TCPStore) to init_process_group manully for each model run

**Description**
This commit remove NUMA binding for device-to-device tests because NUMA doesn't affect performance, and revise benchmark metrics accordingly.

This commit makes data checking in gpu_copy optional, because it will take too long time if message size is large.

**Description**
This commit does the following to optimize result variance in gpu_copy benchmark:
1) Add warmup phase for gpu_copy benchmark to avoid timing instability caused by first-time CUDA kernel launch overhead;
2) Use CUDA events for timing instead of CPU timestamps;
3) Make data checking an option that is not preferred to be enabled in performance test;
4) Enlarge message size in performance benchmark.

Support T4 and A10 in GEMM benchmark.

Fix bug of GPU scan logic in bidirectional tests.

**Description**
Please write a brief description and link the related issue if have.

**Major Revision**
- Sync (do allreduce max) the E2E training results among all workers.
- Avoid using ':0' in metric name if there has only one rank having output.

**Description**
Fix insecure issue of Multiplication result converted to larger type.

**Major Revision**
- Use a cast to ensure that the multiplication is done using the long long to avoid overflow.

**Description**
This commit adds bidirectional tests in gpu_copy benchmark for both device-host transfer and device-device transfer, and revises related tests.

**Description**
Add 50th, 90th, 95th, 99th, 99.9th latency metrics for ORT and pytorch inference benchmarks.

__Description__

Add command `sb benchmark list` and `sb benchmark list-parameters` to support listing all optional parameters for benchmarks.

<details>
<summary>Examples</summary>
<pre>
$ sb benchmark list -n [a-z]+-bw -o table
Result
--------
mem-bw
nccl-bw
rccl-bw
</pre>
<pre>
$ sb benchmark list-parameters -n mem-bw
=== mem-bw ===
optional arguments:
  --bin_dir str         Specify the directory of the benchmark binary.
  --duration int        The elapsed time of benchmark in seconds.
  --mem_type str [str ...]
                        Memory types to benchmark. E.g. htod dtoh dtod.
  --memory str          Memory argument for bandwidthtest. E.g. pinned unpinned.
  --run_count int       The run count of benchmark.
  --shmoo_mode          Enable shmoo mode for bandwidthtest.
default values:
{'bin_dir': None,
 'duration': 0,
 'mem_type': ['htod', 'dtoh'],
 'memory': 'pinned',
 'run_count': 1}
</pre>
</details>

__Major Revisions__
* Add `sb benchmark list` to list benchmarks matching given name.
* Add `sb benchmark list-parameters` to list parameters for benchmarks which match given name.

__Minor Revisions__
* Sort format help text for argparse.

__Description__

Cherry-pick  bug fixes from v0.4.0 to main.

__Major Revisions__

* Bug - Fix issues for Ansible and benchmarks (#267)
* Tests - Refine test cases for microbenchmark (#268)
* Bug - Build openmpi with ucx support in rocm dockerfiles (#269)
* Benchmarks: Fix Bug - Fix fio build issue (#272)
* Docs - Unify metric and add doc for cublas and cudnn functions (#271)
* Monitor: Revision - Add 'monitor/' prefix to monitor metrics in result summary (#274)
* Bug - Fix bug of detecting if gpu_index is none (#275)
* Bug - Fix bugs in data diagnosis (#273)
* Bug - Fix issue that the root mpi rank may not be the first in the hostfile (#270)
* Benchmarks: Configuration - Update inference and network benchmarks in configs (#276)
* Docs - Upgrade version and release note (#277)
Co-authored-by: Yuting Jiang <v-yutjiang@microsoft.com>

Add transformers for TensorRT inference.

**Description**
Add benchmark metrics for cpu-memory-bw-latency.

**Description**
Add mlc memory bandwidth and latency micro benchmark to Superbench.

**Major Revision**
- Add mlc benchmark with test and example files

**Description**
Add ONNXRuntime inference benchmark based on ORT python API.

**Major Revision**
- Add `ORTInferenceBenchmark` class to export pytorch model to onnx model and do inference
- Add tests and example for `ort-inference` benchmark
- Update the introduction docs.

**Description**
Integrate monitor into Superbench.

**Major Revision**
- Initialize, start and stop monitor in SB executor.
- Parse the monitor data in SB runner and merge into benchmark results.
- Specify ReduceType for monitor metrics, such as MAX, MIN and LAST.
- Add monitor configs into config file.

**Description**
Set the `reduce_op` type for metirc `return_code` as `None`.

**Description**
Unify metric names of benchmarks.

**Description**
Add return_code metric into result and revise unit tests.

**Description**
If `ignore_invalid` is True, and 'required' arguments are not set when register the benchmark, the arguments should be provided by user in config and skip the arguments checking.

**Description**
Rename `nvidia_helper` utility as `device_manager` module and support more functions:
```
device_manager.get_device_count()
device_manager.get_device_utilization(idx)
device_manager.get_device_temperature(idx)
device_manager.get_device_power_limit(idx)
device_manager.get_device_memory(idx)
device_manager.get_device_row_remapped_info(idx)
device_manager.get_device_ecc_error(idx)
```

__Description__

Add TensorRT inference benchmark for torchvision models.

__Major Revision__
- Measure TensorRT inference performance.

**Description**
Add ib traffic validation distributed benchmark.

**Major Revision**
- Add ib traffic validation distributed benchmark, example and test