commented on
task #21
"[S.C03.04] External Component Interaction"
at
Eclipse Research Labs / ENACT Project / Graph Neural Network
The Dynamic Graph Modeller is responsible for constructing and updating the graph structure based on incoming telemetry data, which will feed the u...
closed
issue
#47
"[C03.6.1] Deploy the GNN+DRL component in the Arctur's HPC environment to evaluate the performance"
at
Eclipse Research Labs / ENACT Project / Graph Neural Network
closed
issue
#12
"[C03.3.2] Try different scenarios of communication when integrating websockets"
at
Eclipse Research Labs / ENACT Project / Graph Neural Network
closed
issue
#2
"[DELETE ME] C03.3 - Developement of WebSockets in order to communicate the DRL Agent with the GNN"
at
Eclipse Research Labs / ENACT Project / Graph Neural Network
opened
issue
#2
"[DELETE ME] C03.3 - Developement of WebSockets in order to communicate the DRL Agent with the GNN"
at
Eclipse Research Labs / ENACT Project / Graph Neural Network
opened
issue
#3
"C03.2 - The GNN must be tightly integrated with the DRL Agent, facilitating cross-layer collaboration where both models (GNN and DRL) share knowledge to optimize resource allocation and performance holistically."
at
Eclipse Research Labs / ENACT Project / Graph Neural Network
opened
issue
#49
"C03.6 The GNN+DRL component shall meet performance requirements necessary for its effective operation across different implementation configurations and when deployed in high-performance computing environments."
at
Eclipse Research Labs / ENACT Project / Graph Neural Network
closed
issue
#10
"[C03.3.1] Refactor DRL Agent + GNN code to uncouple them and facilitate WebSocket implementation"
at
Eclipse Research Labs / ENACT Project / Graph Neural Network
closed
issue
#2
"[DELETE ME] C03.3 - Developement of WebSockets in order to communicate the DRL Agent with the GNN"
at
Eclipse Research Labs / ENACT Project / Graph Neural Network
commented on
task #34
"[S.C03.F01.04] Success Criteria"
at
Eclipse Research Labs / ENACT Project / Graph Neural Network
The component should consistently provide optimised deployment configurations or results without errors or unexpected behaviour.
commented on
task #33
"[S.C03.F01.03] Interacts with"
at
Eclipse Research Labs / ENACT Project / Graph Neural Network
Orchestrator: To pass the optimised deployment configurations.
commented on
task #33
"[S.C03.F01.03] Interacts with"
at
Eclipse Research Labs / ENACT Project / Graph Neural Network
AI Forecaster: From which receives the predicted/sub-optimal schedules.
commented on
task #21
"[S.C03.04] External Component Interaction"
at
Eclipse Research Labs / ENACT Project / Graph Neural Network
The Dynamic Graph Modeller is responsible for constructing and updating the graph structure based on incoming telemetry data, which will feed the u...
opened
issue
#48
"[C03.6.2] Test the GNN using PyTorch layers insead of using pre-built PyTorch Geometric layers"
at
Eclipse Research Labs / ENACT Project / Graph Neural Network
commented on
issue #47
"[C03.6.1] Deploy the GNN+DRL component in the Arctur's HPC environment to evaluate the performance"
at
Eclipse Research Labs / ENACT Project / Graph Neural Network
We noticed that the training process took half the time when testing a specific type of GNN architecture, compared to the training runs we used to ...
opened
issue
#47
"[C03.6.1] Deploy the GNN+DRL component in the Arctur's HPC environment to evaluate the performance"
at
Eclipse Research Labs / ENACT Project / Graph Neural Network
opened
issue
#10
"[C03.3.1] Refactor DRL Agent + GNN code to uncouple them and facilitate WebSocket implementation"
at
Eclipse Research Labs / ENACT Project / Graph Neural Network
closed
issue
#10
"[C03.3.1] Refactor DRL Agent + GNN code to uncouple them and facilitate WebSocket implementation"
at
Eclipse Research Labs / ENACT Project / Graph Neural Network