-
Notifications
You must be signed in to change notification settings - Fork 0
Meeting Note
- Updates
- Open Issues
- Software Milestones
The meeting notes are in reverse chronological order.
Participants: Shantenu, Ai, Gyorgy, Ray
Discussed on :
- Binding free energy (BFE) of drug design problem (Gyorgy)
- 2D Experiments for EXALearn control problem is running 2D Cahn-Hilliard developed at UT-Austin (Ai)
- Towards the general SRS. (software requirements specification.)
Action items:
- all:
- read all three use-case documents and come up with a general SRS.
Participants: Shantenu, Shinjae, Ai, Gyorgy, Ray
Discussed on:
- Items closed:
- AI stamped
- PyCola
- three use cases
- discussed Peter's reply on cosmoGAN.
- Control Module, simple code, I/O for control module
Action Items:
-
Ray:
- communicate with Peter on loss function / evaluation function
- get the code running?
-
Gyorgy:
- edit the design doc
- create two more workflow diagram. one for simulator one for the filter.
-
Ai:
- come up with any estimate of computation cost.
Participants: Shantenu, Ai, Gyorgy, Ray
Discussed on:
- Items in the Issue tracking
- Ai:
- Use case document for material design
- Ray:
- Use case document for cosmology
Action Items:
- Ray:
- compose a email to Peter to converge on the use case
- try openMP (PyCola) code with RadicalPilot
Participants: Frank, Shantenu, Shinjae, Ai, Gyorgy, Ray
Discussed on:
- Ray updated on
- the progress of cosmology simulation code.
- the RP issues on stampeded2.
- Grant Ai access to stampede2
Action Items:
-
Ray:
- communicate with Peter about the pycola code
- fill out the requirement template for cosomology project.
- try openMP (pycola) code with RadicalPilot
- alternatively, using other hpc cluster or singularity
-
Gyorgy:
- fill out the requirement template for drug design
-
Ai:
- fill out the requirement template for light source control problem.
- send request to stampeded2 to Shantenu
Participants: Shantenu, Shinjae, Ai, Ray
Discussed on:
- Ray updates on progress on Radical Pilot and Cosmology simulation code.
- MUSIC up and running
- Received comments from Peter
- RP/Saga still have issues with stampede2
- Shinjae updates on the progress on control team.
- waiting for forward simulation code (tao).
- another git repo for control team
Action Items for Ray:
- push forward with PyCola
- design of the library
- control requirement document.
Participants: Shantenu, Ai, Gyorgy, Ray
Discussed on:
- Mini-app should be without library dependency. PyCola.
- (Shantenu:) Hierarchy objectives. "Selection Problem"
- RP bug tracking.
- Design document.
- Confluence access
- Gyorgy presents on the drug design problem.
Action Items for Ray:
- continue working on the cosmology simulation.
Action Items for Shantenu:
- contact IT admin to add team members to confluence.
Participants: Shantenu, Shinjae, Ai, Gyorgy, Ray
Discussed on:
- Capture the requirements from the four areas
- Shinjae updated the last Friday's meeting to Shantenu.
- changed the problem from phase diagram to x-ray defraction patterns.
- need to do more investigation on data
- no formal requirements yet.
- further meeting with Kevin.
- Ray updated the status on the issues.
- How to design the library.
- Gyorgy updated on the papers (atom and RL).
- overview on the problem.
- AEOLUS project and MOCU framework.
Action Items for Ray:
- send the requirement template to Shinjae. need link from Shantenu.
Actions Items for Gyorgy:
- presentation on the problem in two weeks.
- timeline on the AEOLUS project
Participants: Shantenu, Gyorgy, Ray
Discussed on:
-
structure for meetings. using git repo wiki and issue tracking.
- Shantenu showed how this works using one of his projects as an example.
- the meeting should begins with "actions review"
- actions are tracked as "issues" using appropriate tags.
- tags to filter issues. One issue should have at least two tags:
- "who": (person, group or organization) is going to work on it
- "what": (specific aspect of the project)
- we need to converge on what the tags are.
- Discussion items
-
first deal line. July 2019, cosmology colleagues.
- put these dates on goals/deadlines.
- usage case template.
- create a landing page with links to confluence and milestones.
-
initial design.
- requirements
- quantitative requirements
- use case template that BNL can share with cosmology team to capture UC requirements.
- SRS.
- requirements
-
talk to peter about the requirements. the objective to measure the performance. is letting the model works for a range of parameters a good idea.
- summit access
- confluence access
- run radical pilot on summit
- (fireworks, flucks)
-
ODED library:
- app facing API
- RTS (runtime system.)
Actions Item for Ray:
- create a gitrepo structure.
- wiki page. meeting notes.
- look into radical-collaboration/hpc-workflows
- get access to confluence