AWS wellarchitected documentation change
Summary
Added testing steps for compressed vector implementations and updated section titles for consistency.
Security assessment
Changes involve performance testing and quality verification for compressed vector storage. No security context is present, and the updates are purely related to cost/performance optimization without security implications.
Diff
diff --git a/wellarchitected/latest/generative-ai-lens/gencost04-bp01.md b/wellarchitected/latest/generative-ai-lens/gencost04-bp01.md index afb32154c..445793c78 100644 --- a//wellarchitected/latest/generative-ai-lens/gencost04-bp01.md +++ b//wellarchitected/latest/generative-ai-lens/gencost04-bp01.md @@ -37,0 +38,6 @@ Alternatively, some embedding models feature compressed vector types. Compressed + 3. Perform latency and load testing on your data retrieval workloads to verify that model response quality is still sufficient. + + 4. Re-test with increased vector size or modified document chunking strategy to improve model response quality. + + * In some cases, changing the search algorithm may improve model response quality as well. + @@ -43 +49 @@ Alternatively, some embedding models feature compressed vector types. Compressed -**Related practices:** +**Related best practices:** @@ -54 +60 @@ Alternatively, some embedding models feature compressed vector types. Compressed -**Related guides, videos, and documentation:** +**Related documents:**