-
Notifications
You must be signed in to change notification settings - Fork 2
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
- Loading branch information
Showing
2,377 changed files
with
31,744 additions
and
152 deletions.
There are no files selected for viewing
Binary file not shown.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,198 @@ | ||
| 1.13.1 | ||
|
|
||
| [2023-11-16 17:46:38,010] [INFO] [distributed.py:36:init_distributed] Not using the DeepSpeed or torch.distributed launchers, attempting to detect MPI environment... | ||
| [2023-11-16 17:46:38,911] [INFO] [distributed.py:83:mpi_discovery] Discovered MPI settings of world_rank=0, local_rank=0, world_size=1, master_addr=26.0.148.159, master_port=6000 | ||
| [2023-11-16 17:46:38,911] [INFO] [distributed.py:46:init_distributed] Initializing torch distributed with backend: nccl | ||
| [2023-11-16 17:46:41,931] [INFO] [checkpointing.py:223:model_parallel_cuda_manual_seed] > initializing model parallel cuda seeds on global rank 0, model parallel rank 0, and data parallel rank 0 with model parallel seed: 3952 and data parallel seed: 1234 | ||
| num_attention_heads: 128, hidden_size: 1024, train_micro_batch_size_per_gpu: 4, tensor_mp_size: 1, pipeline_mp_size: 1, dp_size: 1 | ||
|
|
||
|
|
||
| Actual | ||
| ------ | ||
| Attention duration (in seconds): 0.0081 | ||
| Attention throughput (in TFLOP/s): 17.027 | ||
| MLP duration (in seconds): 0.0008 | ||
| MLP throughput (in TFLOP/s): 182.309 | ||
| Transformer duration (in seconds): 0.0119 | ||
| Transformer throughput (in TFLOP/s): 23.121 | ||
| ======================================================================================================================== | ||
| num_attention_heads: 128, hidden_size: 2048, train_micro_batch_size_per_gpu: 4, tensor_mp_size: 1, pipeline_mp_size: 1, dp_size: 1 | ||
|
|
||
|
|
||
| Actual | ||
| ------ | ||
| Attention duration (in seconds): 0.0113 | ||
| Attention throughput (in TFLOP/s): 36.332 | ||
| MLP duration (in seconds): 0.0026 | ||
| MLP throughput (in TFLOP/s): 213.031 | ||
| Transformer duration (in seconds): 0.0137 | ||
| Transformer throughput (in TFLOP/s): 70.045 | ||
| ======================================================================================================================== | ||
| num_attention_heads: 128, hidden_size: 3072, train_micro_batch_size_per_gpu: 4, tensor_mp_size: 1, pipeline_mp_size: 1, dp_size: 1 | ||
|
|
||
|
|
||
| Actual | ||
| ------ | ||
| Attention duration (in seconds): 0.0116 | ||
| Attention throughput (in TFLOP/s): 70.844 | ||
| MLP duration (in seconds): 0.0058 | ||
| MLP throughput (in TFLOP/s): 214.599 | ||
| Transformer duration (in seconds): 0.0190 | ||
| Transformer throughput (in TFLOP/s): 108.363 | ||
| ======================================================================================================================== | ||
| num_attention_heads: 128, hidden_size: 4096, train_micro_batch_size_per_gpu: 4, tensor_mp_size: 1, pipeline_mp_size: 1, dp_size: 1 | ||
|
|
||
|
|
||
| Actual | ||
| ------ | ||
| Attention duration (in seconds): 0.0200 | ||
| Attention throughput (in TFLOP/s): 68.625 | ||
| MLP duration (in seconds): 0.0096 | ||
| MLP throughput (in TFLOP/s): 228.121 | ||
| Transformer duration (in seconds): 0.0317 | ||
| Transformer throughput (in TFLOP/s): 112.557 | ||
| ======================================================================================================================== | ||
| num_attention_heads: 128, hidden_size: 5120, train_micro_batch_size_per_gpu: 4, tensor_mp_size: 1, pipeline_mp_size: 1, dp_size: 1 | ||
|
|
||
|
|
||
| Actual | ||
| ------ | ||
| Attention duration (in seconds): 0.0277 | ||
| Attention throughput (in TFLOP/s): 74.417 | ||
| MLP duration (in seconds): 0.0141 | ||
| MLP throughput (in TFLOP/s): 243.298 | ||
| Transformer duration (in seconds): 0.0424 | ||
| Transformer throughput (in TFLOP/s): 129.550 | ||
| ======================================================================================================================== | ||
| num_attention_heads: 128, hidden_size: 6144, train_micro_batch_size_per_gpu: 4, tensor_mp_size: 1, pipeline_mp_size: 1, dp_size: 1 | ||
|
|
||
|
|
||
| Actual | ||
| ------ | ||
| Attention duration (in seconds): 0.0317 | ||
| Attention throughput (in TFLOP/s): 91.151 | ||
| MLP duration (in seconds): 0.0202 | ||
| MLP throughput (in TFLOP/s): 244.987 | ||
| Transformer duration (in seconds): 0.0474 | ||
| Transformer throughput (in TFLOP/s): 165.404 | ||
| ======================================================================================================================== | ||
| num_attention_heads: 128, hidden_size: 7168, train_micro_batch_size_per_gpu: 4, tensor_mp_size: 1, pipeline_mp_size: 1, dp_size: 1 | ||
|
|
||
|
|
||
| Actual | ||
| ------ | ||
| Attention duration (in seconds): 0.0376 | ||
| Attention throughput (in TFLOP/s): 102.440 | ||
| MLP duration (in seconds): 0.0272 | ||
| MLP throughput (in TFLOP/s): 247.391 | ||
| Transformer duration (in seconds): 0.0532 | ||
| Transformer throughput (in TFLOP/s): 198.935 | ||
| ======================================================================================================================== | ||
| num_attention_heads: 128, hidden_size: 8192, train_micro_batch_size_per_gpu: 4, tensor_mp_size: 1, pipeline_mp_size: 1, dp_size: 1 | ||
|
|
||
|
|
||
| Actual | ||
| ------ | ||
| Attention duration (in seconds): 0.0418 | ||
| Attention throughput (in TFLOP/s): 118.484 | ||
| MLP duration (in seconds): 0.0361 | ||
| MLP throughput (in TFLOP/s): 243.402 | ||
| Transformer duration (in seconds): 0.0680 | ||
| Transformer throughput (in TFLOP/s): 202.184 | ||
| ======================================================================================================================== | ||
| num_attention_heads: 128, hidden_size: 9216, train_micro_batch_size_per_gpu: 4, tensor_mp_size: 1, pipeline_mp_size: 1, dp_size: 1 | ||
|
|
||
|
|
||
| Actual | ||
| ------ | ||
| Attention duration (in seconds): 0.0474 | ||
| Attention throughput (in TFLOP/s): 130.496 | ||
| MLP duration (in seconds): 0.0458 | ||
| MLP throughput (in TFLOP/s): 243.321 | ||
| Transformer duration (in seconds): 0.0861 | ||
| Transformer throughput (in TFLOP/s): 201.131 | ||
| ======================================================================================================================== | ||
| num_attention_heads: 128, hidden_size: 10240, train_micro_batch_size_per_gpu: 4, tensor_mp_size: 1, pipeline_mp_size: 1, dp_size: 1 | ||
|
|
||
|
|
||
| Actual | ||
| ------ | ||
| Attention duration (in seconds): 0.0471 | ||
| Attention throughput (in TFLOP/s): 160.579 | ||
| MLP duration (in seconds): 0.0555 | ||
| MLP throughput (in TFLOP/s): 247.716 | ||
| Transformer duration (in seconds): 0.1013 | ||
| Transformer throughput (in TFLOP/s): 210.391 | ||
| ======================================================================================================================== | ||
| num_attention_heads: 128, hidden_size: 11264, train_micro_batch_size_per_gpu: 4, tensor_mp_size: 1, pipeline_mp_size: 1, dp_size: 1 | ||
|
|
||
|
|
||
| Actual | ||
| ------ | ||
| Attention duration (in seconds): 0.0524 | ||
| Attention throughput (in TFLOP/s): 173.033 | ||
| MLP duration (in seconds): 0.0672 | ||
| MLP throughput (in TFLOP/s): 247.515 | ||
| Transformer duration (in seconds): 0.1203 | ||
| Transformer throughput (in TFLOP/s): 213.573 | ||
| ======================================================================================================================== | ||
| num_attention_heads: 128, hidden_size: 12288, train_micro_batch_size_per_gpu: 4, tensor_mp_size: 1, pipeline_mp_size: 1, dp_size: 1 | ||
|
|
||
|
|
||
| Actual | ||
| ------ | ||
| Attention duration (in seconds): 0.0600 | ||
| Attention throughput (in TFLOP/s): 178.688 | ||
| MLP duration (in seconds): 0.0788 | ||
| MLP throughput (in TFLOP/s): 251.122 | ||
| Transformer duration (in seconds): 0.1387 | ||
| Transformer throughput (in TFLOP/s): 220.040 | ||
| ======================================================================================================================== | ||
| num_attention_heads: 128, hidden_size: 13312, train_micro_batch_size_per_gpu: 4, tensor_mp_size: 1, pipeline_mp_size: 1, dp_size: 1 | ||
|
|
||
|
|
||
| Actual | ||
| ------ | ||
| Attention duration (in seconds): 0.0603 | ||
| Attention throughput (in TFLOP/s): 207.420 | ||
| MLP duration (in seconds): 0.0926 | ||
| MLP throughput (in TFLOP/s): 250.856 | ||
| Transformer duration (in seconds): 0.1587 | ||
| Transformer throughput (in TFLOP/s): 225.228 | ||
| ======================================================================================================================== | ||
| num_attention_heads: 128, hidden_size: 14336, train_micro_batch_size_per_gpu: 4, tensor_mp_size: 1, pipeline_mp_size: 1, dp_size: 1 | ||
|
|
||
|
|
||
| Actual | ||
| ------ | ||
| Attention duration (in seconds): 0.0705 | ||
| Attention throughput (in TFLOP/s): 204.652 | ||
| MLP duration (in seconds): 0.1079 | ||
| MLP throughput (in TFLOP/s): 249.742 | ||
| Transformer duration (in seconds): 0.1833 | ||
| Transformer throughput (in TFLOP/s): 225.752 | ||
| ======================================================================================================================== | ||
| num_attention_heads: 128, hidden_size: 15360, train_micro_batch_size_per_gpu: 4, tensor_mp_size: 1, pipeline_mp_size: 1, dp_size: 1 | ||
|
|
||
|
|
||
| Actual | ||
| ------ | ||
| Attention duration (in seconds): 0.0800 | ||
| Attention throughput (in TFLOP/s): 206.148 | ||
| MLP duration (in seconds): 0.1236 | ||
| MLP throughput (in TFLOP/s): 250.092 | ||
| Transformer duration (in seconds): 0.2085 | ||
| Transformer throughput (in TFLOP/s): 227.465 | ||
| ======================================================================================================================== | ||
| num_attention_heads: 128, hidden_size: 16384, train_micro_batch_size_per_gpu: 4, tensor_mp_size: 1, pipeline_mp_size: 1, dp_size: 1 | ||
|
|
||
|
|
||
| Actual | ||
| ------ | ||
| Attention duration (in seconds): 0.0847 | ||
| Attention throughput (in TFLOP/s): 220.792 | ||
| MLP duration (in seconds): 0.1413 | ||
| MLP throughput (in TFLOP/s): 249.029 | ||
| Transformer duration (in seconds): 0.2326 | ||
| Transformer throughput (in TFLOP/s): 231.596 | ||
| ======================================================================================================================== |
Oops, something went wrong.