Unanswered Questions tagged with Amazon SageMaker
Content language: English
Select up to 5 tags to filter
Sort by most recent
Browse through the questions and answers listed below or filter and sort to narrow down your results.
AWS Endpoint Failurelg...
Hello, I've been trying to make an endpoint to my model that was traind, it's cnn/gru tensorflow model, while creating endpoint using the following code, first is for conversion to .tar.gz, and after...
0
answers
0
votes
237
views
asked 5 months agolg...
I'm hoping I can get some guidance/confirmation on this design.
I would like to enable access for all SageMaker components (notebooks, training and hosting instances) over a private network from the...
0
answers
0
votes
64
views
asked 5 months agolg...
I'm deploying an LLM that uses huggingface transformers on Sagemaker with custom inference code (e.g. custom model_fn, predict_fn, etc). I know that Sagemaker has support for response streaming, and...
0
answers
0
votes
127
views
asked 5 months agolg...
Traceability and Reproducibility of Models Generated using SageMaker Training Jobs with S3 Datasetlg...
I have a dataset in S3 bucket which is passed as input to the SageMaker training job. The default **traceability** in SageMaker provides only the S3 URI and doesn't include the version details. If at...
0
answers
1
votes
221
views
asked 5 months agolg...
Traceback (most recent call last):
File "data.py", line 88, in <module>
generate(STREAM_NAME, boto3.client('kinesis',region_name='us-east-2'))
File "data.py", line 83, in generate
...
0
answers
0
votes
71
views
asked 5 months agolg...
I have a Resnet50 and Resnet101 model running on a SageMaker MME with Triton server. I have two inference scripts placed in respective folders but it keeps getting ignored by SageMaker. My directory...
0
answers
0
votes
79
views
asked 5 months agolg...
The image sagemaker-model-monitor-analyzer in processing job can’t support more than 4 instanceslg...
**Background**
Our team uses sagemaker processing job to do data baselining, which is to produce statistics result based on a large scale of training dataset. In other words, the goal of solving this...
0
answers
0
votes
242
views
asked 6 months agolg...
I am getting the error in the title every time I attempt to create an an endpoint using an image in ECR. The ECR repository has the following IAM permissions applied:
{
"Version": "2012-10-17",
...
0
answers
0
votes
94
views
asked 6 months agolg...
I'm running a Sagemaker Processing job with SKLearnProcessor processor. I tested code locally and it works fine.
```
from sklearn.model_selection import StratifiedGroupKFold
```
But when I run it...
0
answers
0
votes
208
views
asked 6 months agolg...
Currently, in my verification job when a worker is approving or rejecting an image that has been labeled in a semantic segmentation job, the worker sees the image with the mask on it, I was wondering...
0
answers
0
votes
67
views
asked 6 months agolg...
I have an inference endpoint that returns a HTTP streaming response and I would like to load test it.
Does `ModelLatency` in the recommender metrics refer to time to receive the first chunk, or time...
0
answers
0
votes
51
views
asked 6 months agolg...
I trained a binary classification model using SageMaker AutoML, and deployed an endpoint for real-time inference via the UI (i.e. clicked "Deploy Model" after AutoML made the model).
The JSON I send...
0
answers
0
votes
48
views
asked 6 months agolg...