@@ -295,7 +295,7 @@ Enable current benchmark or not, can be overwritten by [`superbench.enable`](#su
### `timeout`
Set the timeout value in seconds, the benchmarking will stop early if timeout is triggerred.
Set the timeout value in seconds, the benchmarking will stop early if timeout is triggered.
* default value: none
...
...
@@ -336,16 +336,16 @@ A list of models to run, only supported in model-benchmark.
Parameters for benchmark to use, varying for different benchmarks.
There have four common parameters for all benchmarks:
* run_count: how many times do user want to run this benchmark, default value is 1.
There are four common parameters for all benchmarks:
* run_count: how many times does user want to run this benchmark, default value is 1.
* duration: the elapsed time of benchmark in seconds. It can work for all model-benchmark. But for micro-benchmark, benchmark authors should consume it by themselves.
* log_raw_data: log raw data into file instead of saving it into result object, default value is `False`. Benchmarks who have large raw output may want to set it as `True`, such as `nccl-bw`/`rccl-bw`.
* log_flushing: real-time log flushing, default value is `False`.
For Model-Benchmark, there have some parameters that can control the elapsed time.
For Model-Benchmark, there are some parameters that can control the elapsed time.
* duration: the elapsed time of benchmark in seconds.
* num_warmup: the number of warmup step, should be positive integer.
* num_steps: the number of test step.
* num_warmup: the number of warmup steps, should be positive integer.
* num_steps: the number of test steps.
If `duration > 0` and `num_steps > 0`, then benchmark will take the least as the elapsed time. Otherwise only one of them will take effect.
...
...
@@ -429,7 +429,7 @@ while `proc_num: 8, node_num: null` will run 32-GPU distributed training on all
Command prefix to use in the mode, in Python formatted string.
