aws / aws/amazon-sagemaker-examples
[Bug Report] p2.xlarge is not supported for "BERT Base Cased" model deployment!
- Dominant language
- Jupyter Notebook
- Stars
- 11k
- Forks
- 7k
- Avg merge
- 8h 29m
- Merged PRs (30d)
- 8
Description
**Link to the notebook**
[Introduction to SageMaker HuggingFace - Text Classification.](https://github.com/aws/amazon-sagemaker-examples/blob/main/introduction_to_amazon_algorithms/jumpstart_text_classification/Amazon_JumpStart_HuggingFace_Text_Classification.ipynb)
**Describe the bug**
In "4.6. Deploy & run Inference on the fine-tuned model" section, the instance type is set to "ml.p2.xlarge"; however, it's not supported for the "BERT Base Cased" model. Here is the error message when creating the endpoint with p2.xlarge instance:
```
Invalid combination: instance p2 and Cuda 12.2. Please provide a supported configuration for this region.
```
**To reproduce**
Running the notebook with default settings will replicate the error.
**Logs**
N/A
Contributor guide
Research direction
Open introduction_to_amazon_algorithms/jumpstart_text_classification/Amazon_JumpStart_HuggingFace_Text_Classification.ipynb and inspect section 4.6, where ml.p2.xlarge is configured. Confirm a supported instance for the BERT Base Cased deployment, update the notebook setting, and rerun the deployment to verify endpoint creation succeeds.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- aws, huggingface, jupyter-notebook
- Domain
- cloud, machine-learning
- Issue type
- Bug
- Difficulty
- 2/5
- Estimated time
- 1-3 hours
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 55/100