Pre CR fixes

Author: markVaykhansky

README.md

@@ -147,7 +147,7 @@ The `guidellm benchmark` command is used to run benchmarks against a generative
 
 - `--max-requests`: Sets the maximum number of requests for each benchmark run. If not provided, the benchmark will run until `--max-seconds` is reached or the dataset is exhausted.
 
-- `--max-error-rate`: The maximum error rate after which a benchmark will stop. Applicable only for finite deterministic scenarios i.e `rate_type` is `constant` and `--max-seconds` exists OR `--max-requests` exists OR the dataset is finite. If `--max-error-rate` is `None`, benchmarks will continue regardless of error rate.
+- `--max-error-rate`: The maximum error rate after which a benchmark will stop. Applicable only for finite deterministic scenarios i.e `rate_type` is `constant` and `--max-seconds` exists OR `--max-requests` exists OR the dataset is finite. If `--max-error-rate` is `None` or not applicable, benchmarks will continue regardless of error rate.
 
 - `--warmup-percent`: Specifies the percentage of the benchmark to treat as a warmup phase. Requests during this phase are excluded from the final results.
 

src/guidellm/__main__.py

@@ -170,7 +170,7 @@ def cli():
         "The maximum error rate after which a benchmark will stop. "
         "Applicable only for finite deterministic scenarios i.e rate_type is 'constant' and 'max_seconds' exists OR "
         "'max_requests' exists OR the dataset is finite. "
-        "If None, benchmarks will continue regardless of error rate."
+        "If None or not applicable, benchmarks will continue regardless of error rate."
     ),
 )
 @click.option(

src/guidellm/benchmark/benchmark.py

@@ -710,7 +710,7 @@ def from_stats(
             *["incomplete"] * len(incomplete),  # type: ignore[list-item]
             *["error"] * len(errored),  # type: ignore[list-item]
         ]
-        start_time = min(req.start_time for req in total) # ToDo: Fix if total is empty
+        start_time = min(req.start_time for req in total)
         end_time = max(req.end_time for req in total)
 
         total_with_prompt, total_types_with_prompt = (

src/guidellm/scheduler/scheduler.py

@@ -102,7 +102,8 @@ async def run(
             If None, then no limit is set and either the iterator must be exhaustible
             or the max_number must be set.
         :param max_error_rate: The maximum error rate after which the scheduler shuts down.
-            If not provided a default of 5% i.e 0.05 is used.
+            Only applicable in benchmarks with finite deterministic number of requests.
+            If None or not applicable then scheduler will continue regardless of errors.
         :return: An asynchronous generator that yields SchedulerResult objects.
             Each SchedulerResult object contains information about the request,
             the response, and the run information.
@@ -130,7 +131,7 @@ async def run(
                 manager, executor, scheduling_strategy, max_error_rate is not None
             )
             if shutdown_event:
-                assert not shutdown_event.is_set()
+                assert not shutdown_event.is_set(),  "shutdown_event is set before starting scheduling"
             run_info, requests_iter, times_iter = self._run_setup(
                 futures, scheduling_strategy, max_number, max_duration, max_error_rate
             )