Skip to content

Conversation

nullfunc
Copy link
Contributor

Updates for model mapping and dividing docs for LLM on AWS and GCP specifically

Copy link
Member

@jordanstephens jordanstephens left a comment

Choose a reason for hiding this comment

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

A few suggestions inline, but this is great! ✨

| Variable | AWS Bedrock |
|--------------------|-------------|
| `REGION` | Required|
| `MODEL` | Bedrock model ID / Docker model name |
Copy link
Member

Choose a reason for hiding this comment

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

Suggested change
| `MODEL` | Bedrock model ID / Docker model name |
| `MODEL` | Bedrock model ID or Docker model name, for example `meta.llama3-3-70b-instruct-v1:0` or `ai/lama3.3` |

:::

Alternatively, Defang supports model mapping through the openai-access-gateway. This takes a model with a Docker naming convention (e.g. ai/lama3.3) and maps it to
the closest matching one on the target platform. If no such match can be found it can fallback onto a known existing model (e.g. ai/mistral). These environment
Copy link
Member

Choose a reason for hiding this comment

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

Suggested change
the closest matching one on the target platform. If no such match can be found it can fallback onto a known existing model (e.g. ai/mistral). These environment
the closest matching one on the target platform. If no such match can be found it can fallback onto a known existing model (e.g. `ai/mistral`). These environment

@commit111
Copy link
Contributor

I migrated some of the changes over from #230 to here ^^

@nullfunc
Copy link
Contributor Author

Merged into #230 as there was some overlap. Will use that PR to contain these changes.

@nullfunc nullfunc closed this May 16, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants