phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2025-12-13 13:02:38 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

guided decoding initial draft

Browse source
This commit is contained in:
Ashwin Bharambe 2024-10-21 18:44:19 -07:00
parent 1d241bf3fe
commit 6d26bbdce3
4 changed files with 133 additions and 22 deletions
Download patch file Download diff file Expand all files Collapse all files

38
llama_stack/providers/registry/inference.py
View file

@ -9,36 +9,36 @@ from typing import List
from llama_stack.distribution.datatypes import * # noqa: F403
META_REFERENCE_DEPS = [
"accelerate",
"blobfile",
"fairscale",
"torch",
"torchvision",
"transformers",
"zmq",
"lm-format-enforcer",
]
def available_providers() -> List[ProviderSpec]:
return [
InlineProviderSpec(
api=Api.inference,
provider_type="meta-reference",
pip_packages=[
"accelerate",
"blobfile",
"fairscale",
"torch",
"torchvision",
"transformers",
"zmq",
],
pip_packages=META_REFERENCE_DEPS,
module="llama_stack.providers.impls.meta_reference.inference",
config_class="llama_stack.providers.impls.meta_reference.inference.MetaReferenceInferenceConfig",
),
InlineProviderSpec(
api=Api.inference,
provider_type="meta-reference-quantized",
pip_packages=[
"accelerate",
"blobfile",
"fairscale",
"fbgemm-gpu==0.8.0",
"torch",
"torchvision",
"transformers",
"zmq",
],
pip_packages=(
META_REFERENCE_DEPS
+ [
"fbgemm-gpu==0.8.0",
]
),
module="llama_stack.providers.impls.meta_reference.inference",
config_class="llama_stack.providers.impls.meta_reference.inference.MetaReferenceQuantizedInferenceConfig",
),