Closed samumantha closed 3 years ago
For the abstract there is no strict guidelines: As short as possible, as long as necessary. So that people know if the topic is of interest to them. We could have a longer version in the hackMD (link added in last commit) if you want. :)
I find it a lot easier to write to an actual limit. It might be helpful for the future too. How about 200 words? 150?
Ok, how about taking what is written under https://nordic-rse.org/events/seminar-series/#august-2021-package-development-in-julia as guidelines for length?
In terms of word count or height on the page?
This one has ~100 words, so maybe 100-150 is a good word count?
But really, I do not think we need to be strict here :)
Okay I'll write it to 150 words. Thanks.
Here we go:
While SLURM itself provides tools for job orchestration like job arrays, high level tools like Snakemake and Ray are cluster agnostic and can either make use of SLURM, or run a laptop. To Snakemake and Ray to run within Singularity, I present `singreqrun`, which works by requesting the host runs programs on behalf of the container container.
The talk doubles as an introduction to Snakemake an Ray. After some brief background on the main tools (Singularity, SLURM, Snakemake and Ray), we proceed to shell code-along to run the following examples:
* Snakemake for heterogeneous (mixture of CPU and GPU nodes) video corpus processing + quick porting across HPC clusters
* Snakemake for text corpus processing including using extra Singularity containers for utilities
* Ray for hyperparameter search
I end the talk by opening for discussion. Is this a good approach? Can we improve upon it?
I also added both abstracts to the HackMD.
Great thanks! If you want, you can also add the short abstract to the webpage teaser in this PR. I think this would then be ready to be merged?
edit: Ah sorry, just noticed you didn't open this PR, sorry for the noise :)
related #323