Skip to content
Closed
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 6 additions & 0 deletions packages/aws-cdk-lib/aws-bedrock/lib/foundation-model.ts
Original file line number Diff line number Diff line change
Expand Up @@ -236,6 +236,9 @@ export class FoundationModelIdentifier {
/** Base model "anthropic.claude-3-5-haiku-20241022-v1:0" */
public static readonly ANTHROPIC_CLAUDE_3_5_HAIKU_20241022_V1_0 = new FoundationModelIdentifier('anthropic.claude-3-5-haiku-20241022-v1:0');

/** Base model "anthropic.claude-3-5-haiku-20241022-v1:0:200k". */
public static readonly ANTHROPIC_CLAUDE_3_5_HAIKU_20241022_V1_0_200K = new FoundationModelIdentifier('anthropic.claude-3-5-haiku-20241022-v1:0:200k');
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Can you help me to understand where those model id coming from?

I tried to find it in this table, but both ids don't exist

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

In my understanding, this model's ID is 'anthropic.claude-3-5-haiku-20241022-v1:0', and I have confirmed this exists in the documentation table you mentioned. The suffix (:200k) specifies the context window size available for this model.

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Are there any docs or commands that show the existence of model IDs?
I think these are provisioned throughput model IDs, but I couldn't find in docs.

https://docs.aws.amazon.com/bedrock/latest/userguide/pt-supported.html

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Wow, that's exactly the point I was misunderstanding.
I didn't have a proper understanding of the differences between On-Demand and Provisioned Throughput models. Sorry for the confusion, I'll withdraw the PR.


/** Base model "anthropic.claude-3-haiku-20240307-v1:0:200k" */
public static readonly ANTHROPIC_CLAUDE_3_HAIKU_20240307_V1_0_200K = new FoundationModelIdentifier('anthropic.claude-3-haiku-20240307-v1:0:200k');

Expand Down Expand Up @@ -287,6 +290,9 @@ export class FoundationModelIdentifier {
/** Base model "cohere.command-r-v1:0". */
public static readonly COHERE_COMMAND_R_PLUS_V1 = new FoundationModelIdentifier('cohere.command-r-plus-v1:0');

/** Base model "cohere.command-r-plus-v1:0:128k". */
public static readonly COHERE_COMMAND_R_PLUS_V1_0_128K = new FoundationModelIdentifier('cohere.command-r-plus-v1:0:128k');

/** Base model "cohere.rerank-v3-5:0". */
public static readonly COHERE_RERANK_V3_5 = new FoundationModelIdentifier('cohere.rerank-v3-5:0');

Expand Down
Loading