phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2025-10-18 07:18:53 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

fix(models)!: always prefix models with provider_id when registering (#3822)

Browse source 
**!!BREAKING CHANGE!!**

The lookup is also straightforward -- we always look for this identifier
and don't try to find a match for something without the provider_id
prefix.

Note that, this ideally means we need to update the `register_model()`
API also (we should kill "identifier" from there) but I am not doing
that as part of this PR.

## Test Plan

Existing unit tests
This commit is contained in:
Ashwin Bharambe 2025-10-16 06:47:39 -07:00 • committed by GitHub
parent f205ab6f6c
commit f70aa99c97
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
10 changed files with 53 additions and 124 deletions
Download patch file Download diff file Expand all files Collapse all files

21
tests/integration/common/recordings/02c93bb3c314427bae2b7a7a6f054792b9f22d2cb4522eab802810be8672d3dc.json generated
View file

@ -12,26 +12,7 @@
"body": {
"__type__": "ollama._types.ProcessResponse",
"__data__": {
"models": [
{
"model": "llama-guard3:1b",
"name": "llama-guard3:1b",
"digest": "494147e06bf99e10dbe67b63a07ac81c162f18ef3341aa3390007ac828571b3b",
"expires_at": "2025-10-13T14:07:12.309717-07:00",
"size": 2279663616,
"size_vram": 2279663616,
"details": {
"parent_model": "",
"format": "gguf",
"family": "llama",
"families": [
"llama"
],
"parameter_size": "1.5B",
"quantization_level": "Q8_0"
}
}
]
"models": []
}
},
"is_streaming": false

26
tests/integration/fixtures/common.py
View file

@ -117,42 +117,24 @@ def client_with_models(
text_model_id,
vision_model_id,
embedding_model_id,
embedding_dimension,
judge_model_id,
):
client = llama_stack_client
providers = [p for p in client.providers.list() if p.api == "inference"]
assert len(providers) > 0, "No inference providers found"
inference_providers = [p.provider_id for p in providers if p.provider_type != "inline::sentence-transformers"]
model_ids = {m.identifier for m in client.models.list()}
model_ids.update(m.provider_resource_id for m in client.models.list())
# TODO: fix this crap where we use the first provider randomly
# that cannot be right. I think the test should just specify the provider_id
if text_model_id and text_model_id not in model_ids:
client.models.register(model_id=text_model_id, provider_id=inference_providers[0])
raise ValueError(f"text_model_id {text_model_id} not found")
if vision_model_id and vision_model_id not in model_ids:
client.models.register(model_id=vision_model_id, provider_id=inference_providers[0])
raise ValueError(f"vision_model_id {vision_model_id} not found")
if judge_model_id and judge_model_id not in model_ids:
client.models.register(model_id=judge_model_id, provider_id=inference_providers[0])
raise ValueError(f"judge_model_id {judge_model_id} not found")
if embedding_model_id and embedding_model_id not in model_ids:
# try to find a provider that supports embeddings, if sentence-transformers is not available
selected_provider = None
for p in providers:
if p.provider_type == "inline::sentence-transformers":
selected_provider = p
break
selected_provider = selected_provider or providers[0]
client.models.register(
model_id=embedding_model_id,
provider_id=selected_provider.provider_id,
model_type="embedding",
metadata={"embedding_dimension": embedding_dimension or 768},
)
raise ValueError(f"embedding_model_id {embedding_model_id} not found")
return client