Integration into Tokkio 4.0#
The Tokkio 4.0 sample apps shipped with Tokkio 4.0 are already setup with the UMIM Action Server for Tokkio as well as the correct configurations for ACE Agent Chat Controller and ACE Agent Chat Engine. If you want to create a UCF app based on ACE components from scratch, here are the important settings for ACE Agent Chat Controller and ACE Agent that you need to change in the app-xy-params.yaml
UCS parameters file of your UCS app specification, choose the rest of the parameters according to the documentation of those components.
chat-controller:
pipeline: avatar_umim
pipelineParams: # can be extended for all pipeline params in parameters.yaml
grpc_server:
nvidia::rrt::BotRuntimeGrpc:
virtual_assistant_num_instances: 10
speech_pipeline_manager: # config name
SpeechPipelineManager: # component name
tts_eos_delay_ms: 0 # parameters (0 in tbyb)
initial_state: "INIT"
always_on: true
[...] # Other parameters as normal
chat-engine:
interface: "event"
[...] # Other parameters as normal
After setting these parameters to enable UMIM mode for the ACE Agent components, you can then simply add the microservice to the components
section in the app-xy.yaml
UCS app specification.
components:
[...]
- name: tokkio-umim-action-server
type: ucf.svc.tokkio.umim-action-server
parameters:
imagePullSecrets:
- name: ngc-docker-reg-secret
The new component then needs to be connected to Redis, Animation Graph MS and the Tokkio UI Server MS in the connections section in the app-xy.yaml
UCS app specification.
connections:
[...]
tokkio-umim-action-server/redis: redis-timeseries/redis
tokkio-umim-action-server/anim-graph: animation-graph/http-api
tokkio-umim-action-server/ui-server: tokkio-ui-server/ui-server-http