-
Notifications
You must be signed in to change notification settings - Fork 0
.env files
This wiki serves as a place to store the most recent .env files so that any updates to them are clearly shared amongst developers. Should you implement changes that modify any .env files, please update this wiki to reflect those changes.
In order to deploy Bridge, two .env files are required.
The first .env file is the deployment .env and should be located in the same directory as the docker-compose.yaml file. There are 2 paths which should be changed based on local paths on your device the MODEL_REPOSITORY_PATH and OLLAMA_PATH. Both of these variables should point to empty directory on your device. Note that for Windows users, double backslashes are used instead of single forward slashes for paths. Additionally, to connect to the model repository stored on Azure, set MODEL_REPOSITORY_PATH=as://tritonmodels/modelrepository.
Here is what it looks like:
`# *** ELASTIC ***
# Password for the 'elastic' user (at least 6 characters)
ELASTIC_PASSWORD=thisisarandompassword
# Password for the 'kibana_system' user (at least 6 characters)
KIBANA_PASSWORD=thisisarandompassword
# Version of Elastic products
STACK_VERSION=8.13.0
# Set the cluster name
CLUSTER_NAME=bridge-cluster
# Set to 'basic' or 'trial' to automatically start the 30-day trial
LICENSE=basic
# Port to expose Elasticsearch HTTP API to the host
ES_PORT=9200
# Port to expose Kibana to the host
KIBANA_PORT=5601
# Increase or decrease based on the available host memory (in bytes)
ES_MEM_LIMIT=1073741824
KB_MEM_LIMIT=1073741824
LS_MEM_LIMIT=1073741824
ENCRYPTION_KEY=c34d38b3a14956121ff2170e5030b471551370178f43e5626eec58b04a30fae2
# *** OLLAMA ***
LLM_MODEL=gpt-4o
# *** TRITON ***
# TODO: Swith to cloud configuration file
# Path to model repository
# Please make sure to add a volume into the docker-compose if you need to use a local path
#MODEL_REPOSITORY_PATH=as://tritonmodels/modelrepository
MODEL_REPOSITORY_PATH=C:\\Users\\dloch\\repo_path
MODEL_CONTAINER_NAME=modelrepository
# Azure Storage Auth
AZURE_CONNECTION_STRING="DefaultEndpointsProtocol=https;AccountName=tritonmodels;AccountKey=3gzzyBA65d3mLW3+PTPxdWNEZWKIGVdXifgGfqFMjASzY5I/WhStOLuq/IV6+bS9ZiR6EyAglAzu+ASt58Y+XA==;EndpointSuffix=core.windows.net"
AZURE_STORAGE_KEY="3gzzyBA65d3mLW3+PTPxdWNEZWKIGVdXifgGfqFMjASzY5I/WhStOLuq/IV6+bS9ZiR6EyAglAzu+ASt58Y+XA=="
AZURE_STORAGE_ACCOUNT_NAME=tritonmodels
OLLAMA_PATH=C:\\Users\\dloch\\ollama_data
The second .env file is the backend .env file, located in the backend directory.
Here is what it looks like:
LLM_URL="https://api.openai.com/v1"
#openai API key:
OPENAI_KEY=sk-proj-8W1tHA2pEDS5VTVw0rDHT3BlbkFJN0mGTTl6JbkhUqBJ81Wo