Ciro Santilli @cirosantilli 37

Ciro Santilli 37  Updated 2025-07-16  +Created 1970-01-01

 Read the full article

Brodmann area by

Ciro Santilli 37  Updated 2025-07-16  +Created 1970-01-01

 View more

 Read the full article

Allen Adult Human Brain by

Ciro Santilli 37  Updated 2025-07-16  +Created 1970-01-01

 View more

atlas.brain-map.org/atlas?atlas=265297126

 Read the full article

y86.js.org by

Ciro Santilli 37  Updated 2025-07-16  +Created 1970-01-01

 View more

The good:

slick UI! But very hard to read characters, they're way too small.
attempts to show state diffs with a flash. But it goes by too fast, would be better if it were more permanent
Reverse debugging

The bad:

educational ISA
unclear what flags mean from UI, no explanation on hover. Likely the authors assume knowledge of the Y86 book.

 Read the full article

WebRISC-V by

Ciro Santilli 37  Updated 2025-07-16  +Created 1970-01-01

 View more

webriscv.dii.unisi.it/

The good:

The bad:

Clunky UI
circuit diagram doesn't show any state??

 Read the full article

Phaser.js by

Ciro Santilli 37  Updated 2025-07-16  +Created 1970-01-01

 View more

Likely the best JavaScript 2D game engine as of 2023.

github.com/photonstorm/phaser
phaser.io/
phaser.io/examples/v3.85.0/games contains the demo games

Uses Matter.js as a physics engine if enabled. There's also an alternative (in-house?) "arcade" engine: photonstorm.github.io/phaser3-docs/Phaser.Physics.Arcade.ArcadePhysics.html but it appears to be simpler/less robust (but also possibly faster).

TODO any 2D first person examples a bit like Ciro's 2D reinforcement learning games?

The examples are present under:

git clone https://github.com/photonstorm/phaser3-examples

but note that that repo is huge, about 4.5 GiB on local disk, as is has tons of assets.

The demos also include a Monaco-editor based sandbox mode where you can edit code directly on the web and see the game update which is a really sweet addition.

 Read the full article

Matter.js by

Ciro Santilli 37  Updated 2025-07-16  +Created 1970-01-01

 View more

To run the demos locally, tested on Ubuntu 22.10:

git clone https://github.com/liabru/matter-js
cd matter-js
git checkout 0.19.0
npm install
npm run dev

and this opens up the demos on the browser.

 Read the full article

activatedgeek/LeNet-5 by

Ciro Santilli 37  Updated 2025-07-16  +Created 1970-01-01

 View more

github.com/activatedgeek/LeNet-5

This repository contains a very clean minimal PyTorch implementation of LeNet-5 for MNIST.

It trains the LeNet-5 neural network on the MNIST dataset from scratch, and afterwards you can give it newly hand-written digits 0 to 9 and it will hopefully recognize the digit for you.

Ciro Santilli created a small fork of this repo at lenet adding better automation for:

extracting MNIST images as PNG
ONNX CLI inference taking any image files as input
a Python tkinter GUI that lets you draw and see inference live
running on GPU

Install on Ubuntu 24.10 with:

sudo apt install protobuf-compiler
git clone https://github.com/activatedgeek/LeNet-5
cd LeNet-5
git checkout 95b55a838f9d90536fd3b303cede12cf8b5da47f
virtualenv -p python3 .venv
. .venv/bin/activate
pip install \
  Pillow==6.2.0 \
  numpy==1.24.2 \
  onnx==1.13.1 \
  torch==2.0.0 \
  torchvision==0.15.1 \
  visdom==0.2.4 \
;

We use our own pip install because their requirements.txt uses >= instead of == making it random if things will work or not.

On Ubuntu 22.10 it was instead:

pip install
  Pillow==6.2.0 \
  numpy==1.26.4 \
  onnx==1.17.0 torch==2.6.0 \
  torchvision==0.21.0 \
  visdom==0.2.4 \
;

Then run with:

python run.py

This script:

does a fixed 15 epochs on the training data
it then uses the trained net from memory to check accuracy with the test data
then it also produces a lenet.onnx ONNX file which contains the trained network, nice!

It throws a billion exceptions because we didn't start the Visdom server, but everything works nevertheless, we just don't get a visualization of the training.

The terminal outputs lines such as:

Train - Epoch 1, Batch: 0, Loss: 2.311587
Train - Epoch 1, Batch: 10, Loss: 2.067062
Train - Epoch 1, Batch: 20, Loss: 0.959845
...
Train - Epoch 1, Batch: 230, Loss: 0.071796
Test Avg. Loss: 0.000112, Accuracy: 0.967500
...
Train - Epoch 15, Batch: 230, Loss: 0.010040
Test Avg. Loss: 0.000038, Accuracy: 0.989300

And the runtime on Ubuntu 22.10, P51 was:

real    2m10.262s
user    11m9.771s
sys     0m26.368s

One of the benefits of the ONNX output is that we can nicely visualize the neural network on Netron:

Figure 1.
Netron visualization of the activatedgeek/LeNet-5 ONNX output
. From this we can see the bifurcation on the computational graph as done in the code at:
`output = self.c1(img) x = self.c2_1(output) output = self.c2_2(output) output += x output = self.c3(output)`
This doesn't seem to conform to the original LeNet-5 however?

 Read the full article

MLperf v2.1 ResNet by

Ciro Santilli 37  Updated 2025-07-16  +Created 1970-01-01

 View more

Instructions at:

Ubuntu 22.10 setup with tiny dummy manually generated ImageNet and run on ONNX:

sudo apt install pybind11-dev

git clone https://github.com/mlcommons/inference
cd inference
git checkout v2.1

virtualenv -p python3 .venv
. .venv/bin/activate
pip install numpy==1.24.2 pycocotools==2.0.6 onnxruntime==1.14.1 opencv-python==4.7.0.72 torch==1.13.1

cd loadgen
CFLAGS="-std=c++14" python setup.py develop
cd -

cd vision/classification_and_detection
python setup.py develop
wget -q https://zenodo.org/record/3157894/files/mobilenet_v1_1.0_224.onnx
export MODEL_DIR="$(pwd)"
export EXTRA_OPS='--time 10 --max-latency 0.2'

tools/make_fake_imagenet.sh
DATA_DIR="$(pwd)/fake_imagenet" ./run_local.sh onnxruntime mobilenet cpu --accuracy

Last line of output on P51, which appears to contain the benchmark results

TestScenario.SingleStream qps=58.85, mean=0.0138, time=0.136, acc=62.500%, queries=8, tiles=50.0:0.0129,80.0:0.0137,90.0:0.0155,95.0:0.0171,99.0:0.0184,99.9:0.0187

where presumably qps means queries per second, and is the main results we are interested in, the more the better.

Running:

tools/make_fake_imagenet.sh

produces a tiny ImageNet subset with 8 images under fake_imagenet/.

fake_imagenet/val_map.txt contains:

val/800px-Porsche_991_silver_IAA.jpg 817
val/512px-Cacatua_moluccensis_-Cincinnati_Zoo-8a.jpg 89
val/800px-Sardinian_Warbler.jpg 13
val/800px-7weeks_old.JPG 207
val/800px-20180630_Tesla_Model_S_70D_2015_midnight_blue_left_front.jpg 817
val/800px-Welsh_Springer_Spaniel.jpg 156
val/800px-Jammlich_crop.jpg 233
val/782px-Pumiforme.JPG 285

where the numbers are the category indices from ImageNet1k. At gist.github.com/yrevar/942d3a0ac09ec9e5eb3a see e.g.:

817: 'sports car, sport car',
89: 'sulphur-crested cockatoo, Kakatoe galerita, Cacatua galerita',

and so on, so they are coherent with the image names. By quickly looking at the script we see that it just downloads from Wikimedia and manually creates the file.

TODO prepare and test on the actual ImageNet validation set, README says:

Prepare the imagenet dataset to come.

Since that one is undocumented, let's try the COCO dataset instead, which uses COCO 2017 and is also a bit smaller. Note that his is not part of MLperf anymore since v2.1, only ImageNet and open images are used. But still:

wget https://zenodo.org/record/4735652/files/ssd_mobilenet_v1_coco_2018_01_28.onnx
DATA_DIR_BASE=/mnt/data/coco
export DATA_DIR="${DATADIR_BASE}/val2017-300"
mkdir -p "$DATA_DIR_BASE"
cd "$DATA_DIR_BASE"
wget http://images.cocodataset.org/zips/val2017.zip
wget http://images.cocodataset.org/annotations/annotations_trainval2017.zip
unzip val2017.zip
unzip annotations_trainval2017.zip
mv annotations val2017
cd -
cd "$(git-toplevel)"
python tools/upscale_coco/upscale_coco.py --inputs "$DATA_DIR_BASE" --outputs "$DATA_DIR" --size 300 300 --format png
cd -

Now:

./run_local.sh onnxruntime mobilenet cpu --accuracy

fails immediately with:

No such file or directory: '/path/to/coco/val2017-300/val_map.txt

The more plausible looking:

./run_local.sh onnxruntime mobilenet cpu --accuracy --dataset coco-300

first takes a while to preprocess something most likely, which it does only one, and then fails:

Traceback (most recent call last):
  File "/home/ciro/git/inference/vision/classification_and_detection/python/main.py", line 596, in <module>
    main()
  File "/home/ciro/git/inference/vision/classification_and_detection/python/main.py", line 468, in main
    ds = wanted_dataset(data_path=args.dataset_path,
  File "/home/ciro/git/inference/vision/classification_and_detection/python/coco.py", line 115, in __init__
    self.label_list = np.array(self.label_list)
ValueError: setting an array element with a sequence. The requested array has an inhomogeneous shape after 2 dimensions. The detected shape was (5000, 2) + inhomogeneous part.

TODO!

 Read the full article

Netron by