JCVI-syn3B by Ciro Santilli 35 Updated +Created
www.biorxiv.org/content/10.1101/2022.09.19.508583v1.full
CVI-syn3B strains differ from JCVI-syn3.0 by the presence of 19 additional non-essential genes that result in a more easily manipulated cell. JCVI-syn3B additionally includes a dual loxP landing pad that enables easy Cre recombinase mediated insertion of genes
It is also interesting to see how they are interested in co-culture with HeLa cells, presumably to enable infectious bacterial disease studies.
At biology.indiana.edu/news-events/news/2023/lennon-minimal-cells.html (2023) they let it re-evove to it it would regain some fitness, and it did.
@cirosantilli/_file/js/matterjs/js/matterjs/examples.html by Ciro Santilli 35 Updated +Created
A multi-scenario demo.
Run MLperf v2.1 ResNet on Imagenette by Ciro Santilli 35 Updated +Created
Let's run on this Imagenet10 subset called Imagenette.
First ensure that you get the dummy test data run working as per MLperf v2.1 ResNet.
Next, in the imagenette2 directory, first let's create a 224x224 scaled version of the inputs as required by the benchmark at mlcommons.org/en/inference-datacenter-21/:
#!/usr/bin/env bash
rm -rf val224x224
mkdir -p val224x224
for syndir in val/*: do
  syn="$(dirname $syndir)"
  for img in "$syndir"/*; do
    convert "$img" -resize 224x224 "val224x224/$syn/$(basename "$img")"
  done
done
and then let's create the val_map.txt file to match the format expected by MLPerf:
#!/usr/bin/env bash
wget https://gist.githubusercontent.com/aaronpolhamus/964a4411c0906315deb9f4a3723aac57/raw/aa66dd9dbf6b56649fa3fab83659b2acbf3cbfd1/map_clsloc.txt
i=0
rm -f val_map.txt
while IFS="" read -r p || [ -n "$p" ]; do
  synset="$(printf '%s\n' "$p" | cut -d ' ' -f1)"
  if [ -d "val224x224/$synset" ]; then
    for f in "val224x224/$synset/"*; do
      echo "$f $i" >> val_map.txt
    done
  fi
  i=$((i + 1))
done < <( sort map_clsloc.txt )
then back on the mlperf directory we download our model:
wget https://zenodo.org/record/4735647/files/resnet50_v1.onnx
and finally run!
DATA_DIR=/mnt/sda3/data/imagenet/imagenette2 time ./run_local.sh onnxruntime resnet50 cpu --accuracy
which gives on P51:
TestScenario.SingleStream qps=164.06, mean=0.0267, time=23.924, acc=87.134%, queries=3925, tiles=50.0:0.0264,80.0:0.0275,90.0:0.0287,95.0:0.0306,99.0:0.0401,99.9:0.0464
where qps presumably means "querries per second". And the time results:
446.78user 33.97system 2:47.51elapsed 286%CPU (0avgtext+0avgdata 964728maxresident)k
The time=23.924 is much smaller than the time executable because of some lengthy pre-loading (TODO not sure what that means) that gets done every time:
INFO:imagenet:loaded 3925 images, cache=0, took=52.6sec
INFO:main:starting TestScenario.SingleStream
Let's try on the GPU now:
DATA_DIR=/mnt/sda3/data/imagenet/imagenette2 time ./run_local.sh onnxruntime resnet50 gpu --accuracy
which gives:
TestScenario.SingleStream qps=130.91, mean=0.0287, time=29.983, acc=90.395%, queries=3925, tiles=50.0:0.0265,80.0:0.0285,90.0:0.0405,95.0:0.0425,99.0:0.0490,99.9:0.0512
455.00user 4.96system 1:59.43elapsed 385%CPU (0avgtext+0avgdata 975080maxresident)k
TODO lower qps on GPU!
Pernosco by Ciro Santilli 35 Updated +Created
Proprietary extension to Mozilla rr by rr lead coder Robert O'Callahan et. al, started in 2016 after he quit Mozilla.
NISQ algorithm by Ciro Santilli 35 Updated +Created
A quantum algorithm that is thought to be more likely to be useful in the NISQ era of quantum computing.
Tab (Oxford slang) by Ciro Santilli 35 Updated +Created
University of Cambridge students, CanTabBridgeans.
Phaser hello world by Ciro Santilli 35 Updated +Created
Which two persons are hard to satisfy? by Ciro Santilli 35 Updated +Created
tipitaka.fandom.com/wiki/Puggala-Pannatti-Chap.2:
He who stores up whatever he gets and he who gives away whatever he gets - these two persons are hard to satisfy.
Python __getitem__ by Ciro Santilli 35 Updated +Created
Optical computer by Ciro Santilli 35 Updated +Created
SQL 2D histogram by Ciro Santilli 35 Updated +Created
Let's try it on SQLite 3.40.1, Ubuntu 23.04. Data setup:
sqlite3 tmp.sqlite 'create table t(x integer, y integer)'
sqlite3 tmp.sqlite <<EOF
insert into t values
  (0, 0),
  (1, 1),
  (2, 2),
  (3, 3),
  (4, 4),
  (5, 5),
  (6, 6),
  (7, 7),
  (8, 8),
  (9, 9),
  (10, 10),
  (11, 11),
  (12, 12),
  (13, 13),
  (14, 14),
  (15, 15),
  (16, 16),
  (17, 17),
  (18, 18),
  (19, 19),

  (2, 18)
EOF
sqlite3 tmp.sqlite 'create index txy on t(x, y)'
For a bin size of 5 ignoring empty ranges we can:
sqlite3 tmp.sqlite <<EOF
select
  floor(x/5)*5 as x,
  floor(y/5)*5 as y,
  count(*) as cnt
from t
group by 1, 2
order by 1, 2
EOF
which produces the desired:
0|0|5
0|15|1
5|5|5
10|10|5
15|15|5
And to consider empty ranges we can use SQL genenerate_series + as per stackoverflow.com/questions/72367652/populating-empty-bins-in-a-histogram-generated-using-sql:
sqlite3 tmp.sqlite <<EOF
select x, y, sum(cnt) from (
  select
      floor(x/5)*5 as x,
      floor(y/5)*5 as y,
      count(*) as cnt
    from t
    group by 1, 2
  union
  select *, 0 as cnt from generate_series(0, 15, 5) inner join (select * from generate_series(0, 15, 5))
)
group by x, y
EOF
which outputs the desired:
0|0|5
0|5|0
0|10|0
0|15|1
5|0|0
5|5|5
5|10|0
5|15|0
10|0|0
10|5|0
10|10|5
10|15|0
15|0|0
15|5|0
15|10|0
15|15|5
SQL genenerate_series by Ciro Santilli 35 Updated +Created

Unlisted articles are being shown, click here to show only listed articles.