phoenix-oss/llama-stack
1
0
Fork 0
forked from phoenix-oss/llama-stack-mirror
Code Pull requests Releases Packages 2 Activity Actions

Support model resource updates and deletes (#452)

Browse source 
# What does this PR do?
* Changes the registry to store only one RoutableObject per identifier.
Before it was a list, which is not really required.
* Adds impl for updates and deletes
* Updates routing table to handle updates correctly



## Test Plan
```
❯ llama-stack-client models list
+------------------------+---------------+------------------------------------+------------+
| identifier             | provider_id   | provider_resource_id               | metadata   |
+========================+===============+====================================+============+
| Llama3.1-405B-Instruct | fireworks-0   | fireworks/llama-v3p1-405b-instruct | {}         |
+------------------------+---------------+------------------------------------+------------+
| Llama3.1-8B-Instruct   | fireworks-0   | fireworks/llama-v3p1-8b-instruct   | {}         |
+------------------------+---------------+------------------------------------+------------+
| Llama3.2-3B-Instruct   | fireworks-0   | fireworks/llama-v3p2-1b-instruct   | {}         |
+------------------------+---------------+------------------------------------+------------+
❯ llama-stack-client models register dineshyv-model --provider-model-id=fireworks/llama-v3p1-70b-instruct
Successfully registered model dineshyv-model
❯ llama-stack-client models list
+------------------------+---------------+------------------------------------+------------+
| identifier             | provider_id   | provider_resource_id               | metadata   |
+========================+===============+====================================+============+
| Llama3.1-405B-Instruct | fireworks-0   | fireworks/llama-v3p1-405b-instruct | {}         |
+------------------------+---------------+------------------------------------+------------+
| Llama3.1-8B-Instruct   | fireworks-0   | fireworks/llama-v3p1-8b-instruct   | {}         |
+------------------------+---------------+------------------------------------+------------+
| Llama3.2-3B-Instruct   | fireworks-0   | fireworks/llama-v3p2-1b-instruct   | {}         |
+------------------------+---------------+------------------------------------+------------+
| dineshyv-model         | fireworks-0   | fireworks/llama-v3p1-70b-instruct  | {}         |
+------------------------+---------------+------------------------------------+------------+
❯ llama-stack-client models update dineshyv-model --provider-model-id=fireworks/llama-v3p1-405b-instruct
Successfully updated model dineshyv-model
❯ llama-stack-client models list
+------------------------+---------------+------------------------------------+------------+
| identifier             | provider_id   | provider_resource_id               | metadata   |
+========================+===============+====================================+============+
| Llama3.1-405B-Instruct | fireworks-0   | fireworks/llama-v3p1-405b-instruct | {}         |
+------------------------+---------------+------------------------------------+------------+
| Llama3.1-8B-Instruct   | fireworks-0   | fireworks/llama-v3p1-8b-instruct   | {}         |
+------------------------+---------------+------------------------------------+------------+
| Llama3.2-3B-Instruct   | fireworks-0   | fireworks/llama-v3p2-1b-instruct   | {}         |
+------------------------+---------------+------------------------------------+------------+
| dineshyv-model         | fireworks-0   | fireworks/llama-v3p1-405b-instruct | {}         |
+------------------------+---------------+------------------------------------+------------+
llama-stack-client models delete dineshyv-model
❯ llama-stack-client models list
+------------------------+---------------+------------------------------------+------------+
| identifier             | provider_id   | provider_resource_id               | metadata   |
+========================+===============+====================================+============+
| Llama3.1-405B-Instruct | fireworks-0   | fireworks/llama-v3p1-405b-instruct | {}         |
+------------------------+---------------+------------------------------------+------------+
| Llama3.1-8B-Instruct   | fireworks-0   | fireworks/llama-v3p1-8b-instruct   | {}         |
+------------------------+---------------+------------------------------------+------------+
| Llama3.2-3B-Instruct   | fireworks-0   | fireworks/llama-v3p2-1b-instruct   | {}         |
+------------------------+---------------+------------------------------------+------------+

```

---------

Co-authored-by: Dinesh Yeduguru <dineshyv@fb.com>
This commit is contained in:
Dinesh Yeduguru 2024-11-13 21:55:41 -08:00 • committed by GitHub
parent 4253cfcd7f
commit efe791bab7
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
7 changed files with 447 additions and 129 deletions
Download patch file Download diff file Expand all files Collapse all files

12
llama_stack/apis/models/models.py
View file

@ -54,3 +54,15 @@ class Models(Protocol):
provider_id: Optional[str] = None,
metadata: Optional[Dict[str, Any]] = None,
) -> Model: ...
@webmethod(route="/models/update", method="POST")
async def update_model(
self,
model_id: str,
provider_model_id: Optional[str] = None,
provider_id: Optional[str] = None,
metadata: Optional[Dict[str, Any]] = None,
) -> Model: ...
@webmethod(route="/models/delete", method="POST")
async def delete_model(self, model_id: str) -> None: ...