AWS bedrock documentation change
Summary
Added S3 configuration requirements for batch inference: Introduced warning about same-bucket/folder resource constraints, case sensitivity, and clarified InputDataConfig must specify folder paths. Updated video input example structure.
Security assessment
Changes address configuration best practices to prevent job failures, but contain no security-specific content. Warning emphasizes functional requirements (path consistency/case sensitivity) without mentioning access controls, encryption, or vulnerabilities.
Diff
diff --git a/bedrock/latest/userguide/batch-inference-data.md b/bedrock/latest/userguide/batch-inference-data.md index 63022b330..e0709cc2a 100644 --- a//bedrock/latest/userguide/batch-inference-data.md +++ b//bedrock/latest/userguide/batch-inference-data.md @@ -27 +27,5 @@ Each line contains a JSON object with a `recordId` field and a `modelInput` fiel - * (If you define input content as an Amazon S3 location) Some models allow you to define the content of the input as an S3 location. If you choose this option, ensure that the S3 location that you'll specify contains both your content and your JSONL files. Your content and JSONL files can be nested in folders at the S3 location that you specify. For an example, see Example video input for Amazon Nova. + * (If your input content contains an Amazon S3 location) Some models allow you to define the content of the input as an S3 location. See Example video input for Amazon Nova. + +###### Warning + +When using S3 URIs in your prompts, all resources must be in the same S3 bucket and folder. The `InputDataConfig` parameter must specify the folder path containing all linked resources (such as videos or images), not just an individual `.jsonl` file. Note that S3 paths are case-sensitive, so ensure your URIs match the exact folder structure. @@ -78,6 +82,7 @@ If you plan to run batch inference on video inputs using the Amazon Nova Lite or - videos/ - video1.mp4 - video2.mp4 - ... - video50.mp4 - input.jsonl + s3://batch-inference-input-bucket/ + ├── videos/ + │ ├── video1.mp4 + │ ├── video2.mp4 + │ ├── ... + │ └── video50.mp4 + └── input.jsonl @@ -115 +120 @@ A sample record from the `input.jsonl` file would be the following: -When you create the batch inference job, you can specify `s3://batch-inference-input-bucket` as the S3 location. Batch inference will process the `input.jsonl` file in the location, in addition to the video files within the `videos` folder that are referenced in the JSONL file. +When you create the batch inference job, must specify the folder path `s3://batch-inference-input-bucket` in your `InputDataConfig` parameter. Batch inference will process the `input.jsonl` file at this location, along with any referenced resources (such as the video files in the `videos` subfolder).