|
| 1 | +## Meeting 17 Oct |
| 2 | + |
| 3 | +### Agenda |
| 4 | + |
| 5 | +- Action points from last meeting |
| 6 | +- Schedule |
| 7 | +- Admin |
| 8 | + |
| 9 | +### Action points (DECIDED in May) |
| 10 | +- More InfraVis-like collaborations demoing usecases |
| 11 | + - When? **ML afternoon** 2 Dec |
| 12 | + - Data visualization |
| 13 | + - This iteration? Yes |
| 14 | + - Who does? Jayant |
| 15 | +- Move 'Slurm and batch jobs' from Day 3 to Day 2: |
| 16 | + - BC: We already had the introduction? |
| 17 | + - OK toatally second day |
| 18 | +- Pinning versions on the clusters (already doing it partially) |
| 19 | +- prereqs |
| 20 | + - courses |
| 21 | + - login/transfer |
| 22 | + - modules |
| 23 | + - slurm: keep condensed session focus on Python |
| 24 | +- Will will run Markdown next time |
| 25 | + - Still yes?? |
| 26 | + - BC: MkDocs should be used then directly |
| 27 | + - 2-step is too expensive (md with sphinx) |
| 28 | + - Delay to May: BC, PO, RP, JY, BB |
| 29 | +- Times |
| 30 | + - Have 15m always |
| 31 | + - Lunch 60 min |
| 32 | +- Give Anders Hast more time |
| 33 | + - In ML session: YES |
| 34 | +- DECIDED Mark exercises as **important** and optional: RB, BB, BC |
| 35 | +- IDEs: keep the schedule as it is |
| 36 | + - DECIDED Present all 3 IDEs but parallel exercise tracks (where learners decide themselves): BC, RP, |
| 37 | +- Videos |
| 38 | + - DECIDED Youtube playlist + table on course web (also NAISS web) |
| 39 | + - the latest available (living table) |
| 40 | + |
| 41 | +#### Not decided because postponed |
| 42 | + |
| 43 | +- Recordings |
| 44 | + - DECIDED: always before or during |
| 45 | + |
| 46 | +- Exercise rooms (perhaps later decision) |
| 47 | + - Let teachers decide how they teach: RB |
| 48 | + - always try to keep 2-3 persons in each room (BC and RB has good experience here): RB |
| 49 | + - larger rooms: RB |
| 50 | + - cluster rooms: RB |
| 51 | + - combination of above: RB |
| 52 | +- [Richel] One fun thing we did in another course, is to let the learners rename themselves, to include the cluster name, e.g. `Thora` becomes `[Dardel] Thora` |
| 53 | + - Let teachers decide how they teach: RB |
| 54 | + - Let's do this in the introduction: RB |
| 55 | +- [Richel] In some courses we measure the initial confidence our learners have at the start, |
| 56 | + such as [Day 1](https://uppmax.github.io/naiss_intro_python/evaluations/20250424/#analysis-pre-and-post). |
| 57 | + What about we do this at the other days too? |
| 58 | + - Never measure the initial confidence: RB, RP |
| 59 | + - Keep it as it is: RB |
| 60 | + - DECIDED Let's try to measure the initial confidence at **each **day: RB, BC, BB |
| 61 | + - Need to formulate well in such case |
| 62 | +- BB: Interconnected courses: we need to connect our courses better |
| 63 | + - Use a working group (i.e. one working group per workshop): BC, RB |
| 64 | + - Agree: RB, BB |
| 65 | + - Not yet: |
| 66 | + - No: |
| 67 | + - Just make sure students know which courses should be taken in what order - need to decide on "curriculum"? |
| 68 | +- JY: ML/DL should be more about workflows than core principles. |
| 69 | + Core principles should be its own separate course. |
| 70 | + - Yes (next time): BC |
| 71 | + - no |
| 72 | +- JY: Do HPC centeres have the packages by Google Collabs? |
| 73 | + - yes: |
| 74 | + - no: To be discusses in future |
| 75 | +- JY: Recording professionalisms, there are better tools than Zoom |
| 76 | + - Postpone |
| 77 | +- JY: How to do recordings, i.e. blurring out names? |
| 78 | + - BB: knows the GDPR rules |
| 79 | + - Postpone |
| 80 | + |
| 81 | +### [Schedule](https://github.com/UPPMAX/HPC-python/blob/main/docs/schedule.rst) |
| 82 | + |
| 83 | +#### Input |
| 84 | +- less of matplotlib, pandas, seaborn |
| 85 | + - DECIDED keep some: BC |
| 86 | + - (good to understand how to make graphics avail and as prereq to big data)? BC |
| 87 | + - matplot/pandas 45-50 m x 1 |
| 88 | + - load+examples+what they can get out of it + workflows |
| 89 | + - before big data; mention ML prep, memory-saving datatypes |
| 90 | + - <s>VOTE remove all? RP (overruled)</s> |
| 91 | + |
| 92 | +#### Fri |
| 93 | +- **Theme: packages and IDEs + basic HPC including batch jobs and allocating GPUs?** |
| 94 | +- load-run (skip or shorten because included in first day?) |
| 95 | +- longer with packages (perhaps input from pavlin about pixi and uv) |
| 96 | +- intro to slurm and interactive (only) |
| 97 | +- IDEs |
| 98 | +- we are lacking analysis: should we remove all YES |
| 99 | +#### Mon |
| 100 | +- **Theme: Running and analysing** |
| 101 | +- GPU (intro + Python specific) |
| 102 | +- analyzing: short versions of earlier (matplotlib, pandas, seaborn) |
| 103 | +- big data (dask) |
| 104 | +- we are lacking analysis: should we remove all? |
| 105 | +#### Tue |
| 106 | +- **Theme: ML and development?** |
| 107 | +- parallelism in morning 2 hrs |
| 108 | +- ML |
| 109 | + |
| 110 | +### Todos |
| 111 | +- Björn renews schedule |
| 112 | +- Further details discussed in chat |
| 113 | + |
| 114 | +## [Notes from previous meeting](https://github.com/UPPMAX/HPC-python/blob/main/meeting_notes/20250522.md) |
0 commit comments