phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2025-12-25 22:28:04 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Merge branch 'main' into opengauss-add

Browse source
This commit is contained in:
windy 2025-08-08 20:58:48 +08:00 • committed by GitHub
commit 39e49ab97a
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
807 changed files with 79555 additions and 26772 deletions
Download patch file Download diff file Expand all files Collapse all files

65
llama_stack/providers/registry/inference.py
View file

@ -37,16 +37,6 @@ def available_providers() -> list[ProviderSpec]:
config_class="llama_stack.providers.inline.inference.meta_reference.MetaReferenceInferenceConfig",
description="Meta's reference implementation of inference with support for various model formats and optimization techniques.",
),
InlineProviderSpec(
api=Api.inference,
provider_type="inline::vllm",
pip_packages=[
"vllm",
],
module="llama_stack.providers.inline.inference.vllm",
config_class="llama_stack.providers.inline.inference.vllm.VLLMConfig",
description="vLLM inference provider for high-performance model serving with PagedAttention and continuous batching.",
),
InlineProviderSpec(
api=Api.inference,
provider_type="inline::sentence-transformers",
@ -234,17 +224,6 @@ def available_providers() -> list[ProviderSpec]:
description="Groq inference provider for ultra-fast inference using Groq's LPU technology.",
),
),
remote_provider_spec(
api=Api.inference,
adapter=AdapterSpec(
adapter_type="fireworks-openai-compat",
pip_packages=["litellm"],
module="llama_stack.providers.remote.inference.fireworks_openai_compat",
config_class="llama_stack.providers.remote.inference.fireworks_openai_compat.config.FireworksCompatConfig",
provider_data_validator="llama_stack.providers.remote.inference.fireworks_openai_compat.config.FireworksProviderDataValidator",
description="Fireworks AI OpenAI-compatible provider for using Fireworks models with OpenAI API format.",
),
),
remote_provider_spec(
api=Api.inference,
adapter=AdapterSpec(
@ -256,50 +235,6 @@ def available_providers() -> list[ProviderSpec]:
description="Llama OpenAI-compatible provider for using Llama models with OpenAI API format.",
),
),
remote_provider_spec(
api=Api.inference,
adapter=AdapterSpec(
adapter_type="together-openai-compat",
pip_packages=["litellm"],
module="llama_stack.providers.remote.inference.together_openai_compat",
config_class="llama_stack.providers.remote.inference.together_openai_compat.config.TogetherCompatConfig",
provider_data_validator="llama_stack.providers.remote.inference.together_openai_compat.config.TogetherProviderDataValidator",
description="Together AI OpenAI-compatible provider for using Together models with OpenAI API format.",
),
),
remote_provider_spec(
api=Api.inference,
adapter=AdapterSpec(
adapter_type="groq-openai-compat",
pip_packages=["litellm"],
module="llama_stack.providers.remote.inference.groq_openai_compat",
config_class="llama_stack.providers.remote.inference.groq_openai_compat.config.GroqCompatConfig",
provider_data_validator="llama_stack.providers.remote.inference.groq_openai_compat.config.GroqProviderDataValidator",
description="Groq OpenAI-compatible provider for using Groq models with OpenAI API format.",
),
),
remote_provider_spec(
api=Api.inference,
adapter=AdapterSpec(
adapter_type="sambanova-openai-compat",
pip_packages=["litellm"],
module="llama_stack.providers.remote.inference.sambanova_openai_compat",
config_class="llama_stack.providers.remote.inference.sambanova_openai_compat.config.SambaNovaCompatConfig",
provider_data_validator="llama_stack.providers.remote.inference.sambanova_openai_compat.config.SambaNovaProviderDataValidator",
description="SambaNova OpenAI-compatible provider for using SambaNova models with OpenAI API format.",
),
),
remote_provider_spec(
api=Api.inference,
adapter=AdapterSpec(
adapter_type="cerebras-openai-compat",
pip_packages=["litellm"],
module="llama_stack.providers.remote.inference.cerebras_openai_compat",
config_class="llama_stack.providers.remote.inference.cerebras_openai_compat.config.CerebrasCompatConfig",
provider_data_validator="llama_stack.providers.remote.inference.cerebras_openai_compat.config.CerebrasProviderDataValidator",
description="Cerebras OpenAI-compatible provider for using Cerebras models with OpenAI API format.",
),
),
remote_provider_spec(
api=Api.inference,
adapter=AdapterSpec(

5
llama_stack/providers/registry/vector_io.py
View file

@ -395,7 +395,7 @@ That means you'll get fast and efficient vector retrieval.
To use PGVector in your Llama Stack project, follow these steps:
1. Install the necessary dependencies.
2. Configure your Llama Stack project to use Faiss.
2. Configure your Llama Stack project to use pgvector. (e.g. remote::pgvector).
3. Start storing and querying vectors.
## Installation
@ -410,6 +410,7 @@ See [PGVector's documentation](https://github.com/pgvector/pgvector) for more de
""",
),
api_dependencies=[Api.inference],
optional_api_dependencies=[Api.files],
),
remote_provider_spec(
Api.vector_io,
@ -497,6 +498,7 @@ See [Weaviate's documentation](https://weaviate.io/developers/weaviate) for more
module="llama_stack.providers.inline.vector_io.qdrant",
config_class="llama_stack.providers.inline.vector_io.qdrant.QdrantVectorIOConfig",
api_dependencies=[Api.inference],
optional_api_dependencies=[Api.files],
description=r"""
[Qdrant](https://qdrant.tech/documentation/) is an inline and remote vector database provider for Llama Stack. It
allows you to store and query vectors directly in memory.
@ -553,6 +555,7 @@ Please refer to the inline provider documentation.
""",
),
api_dependencies=[Api.inference],
optional_api_dependencies=[Api.files],
),
remote_provider_spec(
Api.vector_io,