create github benchmark workflow

[TroyGarden]

Summary:

• add a script to run train pipeline benchmark
• add a github workflow which can run benchmark script nightly
• the workflow can also be triggered manually
• trace and memory snapshot will be uploaded to github artifacts
• also fix some github workflow naming conventions and typos.

NOTE: github runner linux.g5.12xlarge.nvidia.gpu only has 4 gpus with 20GB HBM, so can only support the *-light.yml benchmarks

• github job

Python 3.13.11
torch 2.11.0.dev20251222+cu128
fbgemm_gpu 2025.12.22+cu128
torchrec 1.5.0a0+623f0dc
Tue Dec 23 05:41:12 2025       
+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 580.65.06              Driver Version: 580.65.06      CUDA Version: 13.0     |
+-----------------------------------------+------------------------+----------------------+
| GPU  Name                 Persistence-M | Bus-Id          Disp.A | Volatile Uncorr. ECC |
| Fan  Temp   Perf          Pwr:Usage/Cap |           Memory-Usage | GPU-Util  Compute M. |
|                                         |                        |               MIG M. |
|=========================================+========================+======================|
|   0  NVIDIA A10G                    On  |   00000000:00:1B.0 Off |                    0 |
|  0%   19C    P8             16W /  300W |       0MiB /  23028MiB |      0%      Default |
|                                         |                        |                  N/A |
+-----------------------------------------+------------------------+----------------------+
|   1  NVIDIA A10G                    On  |   00000000:00:1C.0 Off |                    0 |
|  0%   20C    P8             20W /  300W |       0MiB /  23028MiB |      0%      Default |
|                                         |                        |                  N/A |
+-----------------------------------------+------------------------+----------------------+
|   2  NVIDIA A10G                    On  |   00000000:00:1D.0 Off |                    0 |
|  0%   19C    P8             19W /  300W |       0MiB /  23028MiB |      0%      Default |
|                                         |                        |                  N/A |
+-----------------------------------------+------------------------+----------------------+
|   3  NVIDIA A10G                    On  |   00000000:00:1E.0 Off |                    0 |
|  0%   19C    P8             19W /  300W |       0MiB /  23028MiB |      0%      Default |
|                                         |                        |                  N/A |
+-----------------------------------------+------------------------+----------------------+

+-----------------------------------------------------------------------------------------+
| Processes:                                                                              |
|  GPU   GI   CI              PID   Type   Process name                        GPU Memory |
|        ID   ID                                                               Usage      |
|=========================================================================================|
|  No running processes found                                                             |
+-----------------------------------------------------------------------------------------+
torchrec directory: /meta-pytorch/torchrec/torchrec
working directory: /meta-pytorch/torchrec

short name	GPU Runtime (P90)	CPU Runtime (P90)	GPU Peak Mem alloc (P90)	GPU Peak Mem reserved (P90)	GPU Mem used (P90)	Malloc retries (P50/P90/P100)	CPU Peak RSS (P90)
base_pipeline_light	8213.92 ms	7705.65 ms	11.85 GB	16.71 GB	17.23 GB	0.0 / 0.0 / 0.0	3.45 GB
sparse_data_dist_light	7909.53 ms	7221.77 ms	12.48 GB	18.33 GB	18.86 GB	0.0 / 0.0 / 0.0	3.48 GB

Differential Revision: D89629829

[meta-codesync]

@TroyGarden has exported this pull request. If you are a Meta employee, you can view the originating Diff in D89629829.