wehub-resource-sync
26 lines
645 B
Python
26 lines
645 B
Python
from mlflow.deployments import get_deploy_client
|
|
|
|
|
|
def main():
|
|
client = get_deploy_client("http://localhost:7000")
|
|
|
|
print(f"Hugging Face TGI endpoints: {client.list_endpoints()}\n")
|
|
print(
|
|
f"Hugging Face completions endpoint info: {client.get_endpoint(endpoint='completions')}\n"
|
|
)
|
|
|
|
# Completions request
|
|
response_completions = client.predict(
|
|
endpoint="completions",
|
|
inputs={
|
|
"prompt": ("What is Deep Learning?"),
|
|
"temperature": 0.1,
|
|
},
|
|
)
|
|
|
|
print(f"Hugging Face TGI response for completions: {response_completions}")
|
|
|
|
|
|
if __name__ == "__main__":
|
|
main()
|