config.default

# Copyright 2022 Google LLC All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

######################################################
#
#  Configs
#
#  Update these variables based on your GCP project, 
#  naming convensions, and desired settings.
#
#  Where it makes sense, code has been added
#  to dynamically set variable values, but these 
#  can all be overridden by adding in your own values 
#  within this file.
#
#  NOTE: The "TF_VAR_" prefix is used so that terraform
#  can dynamically load these env variables.
#
######################################################

######################################################
# General Project Parameters
######################################################

# GCP Project ID 
# TODO: User needs to add their own value prior to deployment
export TF_VAR_GCP_PROJECT_ID=<YOUR_GCP_PROJECT_ID>

# GCP Project Number
# TODO: User needs to add their own value prior to deployment
export TF_VAR_GCP_PROJECT_NUMBER=<YOUR_GCP_PROJECT_NUMBER>

# GCP Region where the Clean-Chat resources should be deployed.
# TODO (Optional): User may add their own value prior to deployment
export TF_VAR_GCP_REGION=us-central1

# Solution name that is used as a naming prefix
# TODO (Optional): User may add their own value prior to deployment
export TF_VAR_SOLUTION_NAME=clean-chat

# (Optional) Add your model API key here if you choose to 
# deploy a model that requires an API key, using services from 
# Cohere.ai, Perspective, etc.
export TF_VAR_MODEL_API_KEY=""

######################################################
# Google Artifact Repo
######################################################

export TF_VAR_GCP_ARTIFACT_REGISTRY_NAME=${TF_VAR_GCP_PROJECT_ID}-${TF_VAR_SOLUTION_NAME}-repo
export TF_VAR_GCP_ARTIFACT_REGISTRY_REGION=${TF_VAR_GCP_REGION}

######################################################
# Google Cloud Storage
######################################################

# GCS bucket where text-based chat files will be dropped/uploaded.
export TF_VAR_GCS_BUCKET_TEXT_DROPZONE=${TF_VAR_GCP_PROJECT_ID}-${TF_VAR_SOLUTION_NAME}-text-dropzone

# GCS bucket where audio files will be dropped/uploaded.
export TF_VAR_GCS_BUCKET_AUDIO_DROPZONE_SHORT=${TF_VAR_GCP_PROJECT_ID}-${TF_VAR_SOLUTION_NAME}-audio-dropzone-short
export TF_VAR_GCS_BUCKET_AUDIO_DROPZONE_LONG=${TF_VAR_GCP_PROJECT_ID}-${TF_VAR_SOLUTION_NAME}-audio-dropzone-long
export TF_VAR_GCS_BUCKET_AUDIO_STT_RESULTS=${TF_VAR_GCP_PROJECT_ID}-${TF_VAR_SOLUTION_NAME}-audio-stt-results

# GCS bucket for cloud functions tmp space
export TF_VAR_GCS_BUCKET_CLOUD_FUNCTIONS=${TF_VAR_GCP_PROJECT_ID}-${TF_VAR_SOLUTION_NAME}-cloud-functions

# GCS bucket for Dataflow - tmp and staging assets.
export TF_VAR_GCS_BUCKET_DATAFLOW=${TF_VAR_GCP_PROJECT_ID}-${TF_VAR_SOLUTION_NAME}-dataflow

######################################################
# Google PubSub
######################################################

# PubSub topic where chat/txt messages will be sent to for analysis. 
export TF_VAR_PUBSUB_TOPIC_TEXT_INPUT=${TF_VAR_GCP_PROJECT_ID}-${TF_VAR_SOLUTION_NAME}-text-input

# This topic contains the scored messages as a JSON payload.
export TF_VAR_PUBSUB_TOPIC_TEXT_SCORED=${TF_VAR_GCP_PROJECT_ID}-${TF_VAR_SOLUTION_NAME}-text-scored
export TF_VAR_PUBSUB_TOPIC_TEXT_SCORED_PUSH_ENDPOINT=https://${TF_VAR_SOLUTION_NAME}-api-backend-placeholder-uc.a.run.app/${TF_VAR_SOLUTION_NAME}-callback

# PubSub topic for flagged/toxic user (if a toxic user is flagged, a message will be sent here)
export TF_VAR_PUBSUB_TOPIC_TOXIC=${TF_VAR_GCP_PROJECT_ID}-${TF_VAR_SOLUTION_NAME}-toxic-topic

######################################################
# API Endpoint & Backend
######################################################

# API Backend App Parameters
export TF_VAR_APP_CLOUD_RUN_NAME=${TF_VAR_SOLUTION_NAME}-api-backend
export TF_VAR_APP_CLOUD_RUN_REGION=${TF_VAR_GCP_REGION}

# API Endpoint Parameters
export TF_VAR_ENDPOINTS_CLOUD_RUN_NAME=${TF_VAR_SOLUTION_NAME}-api-endpoint
export TF_VAR_ENDPOINTS_CLOUD_RUN_REGION=${TF_VAR_GCP_REGION}

######################################################
# BigQuery
######################################################

# BigQuery locations such as US, EU, us-central1, europe-west1, etc.
# Reference: https://cloud.google.com/bigquery/docs/locations
export TF_VAR_BIGQUERY_LOCATION=US
export TF_VAR_BIGQUERY_DATASET=clean_chat
export TF_VAR_BIGQUERY_TABLE=scored_chats

######################################################
# Google Dataflow (Scoring Engine configs)
######################################################

# Dataflow Region
export TF_VAR_DATAFLOW_REGION=${TF_VAR_GCP_REGION}

# Dataflow Runner (typically DataflowRunner, or DirectRunner for local testing/deployment)
export TF_VAR_DATAFLOW_RUNNER=DataflowRunner

# Max Dataflow workers 
export TF_VAR_MAX_DATAFLOW_WORKERS=2

# Scoring engine time window settings
# Toxic scores are aggregated based on the time window parameters, set below.
# Window duration (in seconds) is the time window that the aggregation is applied to.  
# Example: If the TF_VAR_WINDOW_DURATION_SECONDS=30, then scores will be aggregated 
# for a 30 second time period.
# https://beam.apache.org/documentation/programming-guide/#sliding-time-windows 
export TF_VAR_WINDOW_DURATION_SECONDS=30

# Window sliding (in seconds) is a parameter used to determine how often a new window starts
# Example: If the TF_VAR_WINDOW_SLIDING_SECONDS=5, then scores will be calculated across the 
# window at 5 second intervals. 
# https://beam.apache.org/documentation/programming-guide/#sliding-time-windows 
export TF_VAR_WINDOW_SLIDING_SECONDS=5

# Toxicity Threshold (from 0.0 to 1.0 where 1.0 is highly toxic) 
export TF_VAR_TOXIC_USER_THRESHOLD=0.60

######################################################
# Machine Learning
######################################################

# GKE cluster name for ML Pipeline
export TF_VAR_ML_CLUSTER=${TF_VAR_GCP_PROJECT_ID}-${TF_VAR_SOLUTION_NAME}-ml-cluster

# GKE zone
export TF_VAR_ML_ZONE=${TF_VAR_GCP_REGION}

# GKE machine type
export TF_VAR_ML_MACHINE_TYPE=n1-standard-2

# Additional ML Pipeline Configs
export TF_VAR_ML_SCOPES=cloud-platform
export TF_VAR_ML_PIPELINE_NAME=${TF_VAR_GCP_PROJECT_ID}-toxicity-pipeline
export TF_VAR_ML_PIPELINE_ROOT=${TF_VAR_GCP_PROJECT_ID}-kubeflow-root
export TF_VAR_ML_APP_INSTANCE_NAME=${TF_VAR_GCP_PROJECT_ID}-${TF_VAR_SOLUTION_NAME}-ml-pipeline
export TF_VAR_ML_NAMESPACE=${TF_VAR_SOLUTION_NAME}-ml-namespace
export TF_VAR_MARKETPLACE_TOOLS_TAG=remove-ui-ownerrefs
export TF_VAR_MARKETPLACE_TOOLS_IMAGE=gcr.io/cloud-marketplace-staging/marketplace-k8s-app-tools/k8s/dev

######################################################
# Get current working dir
######################################################

export TF_VAR_CURRENT_DIR=$PWD

######################################################
# Environment varirables for tfx pipeline on Vertex AI
######################################################

export ML_GCS_BUCKET=gs://${TF_VAR_GCP_PROJECT_ID}-clean-chat-bert

######################################################
# End of Config
######################################################