From b5ef25b4cdd14ea5ca808c432f5905a156635d42 Mon Sep 17 00:00:00 2001 From: Harsh Asnani Date: Tue, 23 Jun 2026 11:06:39 -0400 Subject: [PATCH 1/2] Update README.md --- README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/README.md b/README.md index af94d84..5cd878a 100644 --- a/README.md +++ b/README.md @@ -279,7 +279,7 @@ See [client_capacity/README.md](client_capacity/README.md) for usage. * **Interactive CLI** – Add traces while running benchmarks in the current dry benchmark tool. [tracking issues](https://github.com/aws-samples/sample-InferenceBenchmarker/issues) -* **Automatic RPS** – Automate trial and error server rps supported at success threshold when --endpoint-config for hardware telemetry provided. [tracking issue](https://github.com/aws-samples/sample-InferenceBenchmarker/issues/2) +* **Automatic RPS** – Automate trial and error server rps supported at success threshold when --endpoint-config for hardware telemetry provided. [tracking issue](https://github.com/aws-samples/sample-InferenceBenchmarker/issues/3) * **Plot metadata** – Provide a dict to add legend and hover info for better visuals in plots. [tracking issues](https://github.com/aws-samples/sample-InferenceBenchmarker/issues) From 861a0dc8fcd75dcbd32745c23496c54c6a1f43e9 Mon Sep 17 00:00:00 2001 From: Harsh Asnani Date: Tue, 23 Jun 2026 15:19:14 +0000 Subject: [PATCH 2/2] remove region hardcoding --- .gitignore | 1 + README.md | 2 +- factories/sagemakerai_realtime/factories_cnn.py | 1 - factories/sagemakerai_realtime/factories_ic.py | 1 - factories/sagemakerai_realtime/factories_llm_textgeneration.py | 1 - 5 files changed, 2 insertions(+), 4 deletions(-) diff --git a/.gitignore b/.gitignore index 553fb91..3584784 100644 --- a/.gitignore +++ b/.gitignore @@ -1,3 +1,4 @@ .venvIB* .tmp .gitignore +*.pyc diff --git a/README.md b/README.md index 5cd878a..772cf49 100644 --- a/README.md +++ b/README.md @@ -71,7 +71,7 @@ InferenceBenchmarker takes your invocation logic defined in Python functions in # user on that worker. Put expensive, reusable client state here — e.g. the boto3 # client and its connection pool — so it isn't rebuilt per request. import json, boto3 - client = boto3.client('sagemaker-runtime', region_name='us-east-1') + client = boto3.client('sagemaker-runtime') def invoke(payload: Payload) -> None: # PER-REQUEST: called once for each request a user fires. diff --git a/factories/sagemakerai_realtime/factories_cnn.py b/factories/sagemakerai_realtime/factories_cnn.py index c1909ee..3fee445 100644 --- a/factories/sagemakerai_realtime/factories_cnn.py +++ b/factories/sagemakerai_realtime/factories_cnn.py @@ -10,7 +10,6 @@ def invoke_factory(endpoint_name=None): client = boto3.client( 'sagemaker-runtime', - region_name='us-east-1', config=Config(retries={'max_attempts': 3}, max_pool_connections=500, read_timeout=3600), ) diff --git a/factories/sagemakerai_realtime/factories_ic.py b/factories/sagemakerai_realtime/factories_ic.py index 3c16bf1..1ffa518 100644 --- a/factories/sagemakerai_realtime/factories_ic.py +++ b/factories/sagemakerai_realtime/factories_ic.py @@ -13,7 +13,6 @@ def invoke_factory(endpoint_name=None): client = boto3.client( 'sagemaker-runtime', - region_name='us-east-1', config=Config(retries={'max_attempts': 0}, max_pool_connections=500), ) diff --git a/factories/sagemakerai_realtime/factories_llm_textgeneration.py b/factories/sagemakerai_realtime/factories_llm_textgeneration.py index 55d8982..6fed59b 100644 --- a/factories/sagemakerai_realtime/factories_llm_textgeneration.py +++ b/factories/sagemakerai_realtime/factories_llm_textgeneration.py @@ -7,7 +7,6 @@ def invoke_factory(endpoint_name=None): client = boto3.client( 'sagemaker-runtime', - region_name='us-east-1', config=Config(retries={'max_attempts': 3}, max_pool_connections=500, read_timeout=3600), )