Switch default deploy model (#18)
parent
da40363057
commit
9739683fba
|
|
@ -4,7 +4,7 @@ services:
|
||||||
init:
|
init:
|
||||||
image: tabbyml/tabby
|
image: tabbyml/tabby
|
||||||
container_name: tabby-init
|
container_name: tabby-init
|
||||||
command: python -m tabby.tools.model_preload --repo_id TabbyML/NeoX-70M
|
command: python -m tabby.tools.model_preload --repo_id TabbyML/J-350M
|
||||||
volumes:
|
volumes:
|
||||||
- ../data/hf_cache:/root/.cache/huggingface
|
- ../data/hf_cache:/root/.cache/huggingface
|
||||||
|
|
||||||
|
|
@ -13,7 +13,7 @@ services:
|
||||||
container_name: tabby-server
|
container_name: tabby-server
|
||||||
command: uvicorn tabby.server:app --host 0.0.0.0 --port 5000
|
command: uvicorn tabby.server:app --host 0.0.0.0 --port 5000
|
||||||
environment:
|
environment:
|
||||||
- MODEL_NAME=TabbyML/NeoX-70M
|
- MODEL_NAME=TabbyML/J-350M
|
||||||
- MODEL_BACKEND=triton
|
- MODEL_BACKEND=triton
|
||||||
ports:
|
ports:
|
||||||
- "5000:5000"
|
- "5000:5000"
|
||||||
|
|
@ -48,7 +48,7 @@ services:
|
||||||
count: all
|
count: all
|
||||||
capabilities: [gpu]
|
capabilities: [gpu]
|
||||||
environment:
|
environment:
|
||||||
- MODEL_NAME=TabbyML/NeoX-70M
|
- MODEL_NAME=TabbyML/J-350M
|
||||||
depends_on:
|
depends_on:
|
||||||
init:
|
init:
|
||||||
condition: service_completed_successfully
|
condition: service_completed_successfully
|
||||||
|
|
|
||||||
|
|
@ -31,7 +31,7 @@ def preload(local_files_only=False):
|
||||||
|
|
||||||
if __name__ == "__main__":
|
if __name__ == "__main__":
|
||||||
args = parse_args()
|
args = parse_args()
|
||||||
print(f"Loading {args.repo_id} ...")
|
print(f"Loading {args.repo_id}, this will take a while...")
|
||||||
try:
|
try:
|
||||||
preload(local_files_only=args.prefer_local_files)
|
preload(local_files_only=args.prefer_local_files)
|
||||||
except Exception as e:
|
except Exception as e:
|
||||||
|
|
|
||||||
Loading…
Reference in New Issue