- they have a separate DB entry for everything in the page, e.g. even list items. A little bit like OurBigBook dynamic tree but way more blocks
- they use PostgreSQL of course
The key advantages of lasers over other light sources are:
- lasers emit a narrow spectrum
- it can be efficient collimated, while still emitting a lot of output power: Section "Why can't you collimate incoherent light as well as a laser?"
- can be phase and polarization coherent, though it is not always the case? TODO.
One cool thing about lasers is that they rely on one specific atomic energy level transition to produce light. This is why they are able to to be so monchromatic. Compare this to:As such, lasers manage to largely overcome "temperature distribution-like" effects that create wider wave spectrum
- incandescent bulbs: wide black-body radiation spectrum
- LED: has a wider spectrum fundamentally related to an energy distribution, related: Why aren't LEDs monochromatic
- TODO think a bit about fluorescent lamps. These also rely on atomic energy transitions, but many of them are present at once, which makes the spectrum very noisy. But would individual lines be very narrow?
Crazy difference between 5W laser and 5W LED by Brainiac75
. Source. Baseic but good. Uses a laser photometer.en.wikipedia.org/w/index.php?title=Andr%C3%A9-Marie_Amp%C3%A8re&oldid=1211946256:TODO find the source for this.
Jean-Jacques Ampère, a successful merchant, was an admirer of the philosophy of Jean-Jacques Rousseau, whose theories of education (as outlined in his treatise Émile) were the basis of Ampère's education. Rousseau believed that young boys should avoid formal schooling and pursue instead a "direct education from nature." Ampère's father actualized this ideal by allowing his son to educate himself within the walls of his well-stocked library.
One more more electrical wires surrounded by an insulator.
He participated in the development of the electrical telegraph, and he did some good modeling work that improved the foundations of the field, notably creating the telegrapher's equations.
He was one of those idealists who just want to do some cool work even if they have to starve for it, people had to get a state pension for him for his contributions. Nice guy. en.wikipedia.org/w/index.php?title=Oliver_Heaviside&oldid=1230097796#Later_years_and_views:He also never married: www.nndb.com/people/627/000204015/
In 1896, FitzGerald and John Perry obtained a civil list pension of £120 per year for Heaviside, who was now living in Devon, and persuaded him to accept it, after he had rejected other charitable offers from the Royal Society.
CIA 2010 covert communication websites 2013 DNS census secureserver.net MX records intersection 2013 DNS Census virtual host cleanup by 
Ciro Santilli 37 Created 2023-07-19 Updated 2025-07-16
Ciro Santilli 37 Created 2023-07-19 Updated 2025-07-16We intersect 2013 DNS Census virtual host cleanup with 2013 DNS census MX records and that leaves 460k hits. We did lose a third on the the MX records as of 260 hits since secureserver.net is only used in 1/3 of sites, but we also concentrate 9x, so it may be worth it.
We did a full Wayback Machine CDX scanning for JAR, SWF and cgi-bin in those, but only found a single new hit:
- 63.130.160.50 theglobalheadlines.com. Just barely missed with our 2013 DNS Census virtual host cleanup heuristic keyword searches as we did think of both "global" and "headlines" in the "news" themes!
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
tkinterGUI that lets you draw and see inference live - running on GPU
Install on Ubuntu 24.10 with:We use our own
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 \
;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:This script:
python run.pyIt 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: 0.989300One of the benefits of the ONNX output is that we can nicely visualize the neural network on Netron:
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)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 --accuracyLast line of output on P51, which appears to contain the benchmark resultswhere presumably
TestScenario.SingleStream qps=58.85, mean=0.0138, time=0.136, acc=62.500%, queries=8, tiles=50.0:0.0129,80.0:0.0137,90.0:0.0155,95.0:0.0171,99.0:0.0184,99.9:0.0187qps means queries per second, and is the main results we are interested in, the more the better.Running:produces a tiny ImageNet subset with 8 images under
tools/make_fake_imagenet.shfake_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 285TODO 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:fails immediately with:The more plausible looking:first takes a while to preprocess something most likely, which it does only one, and then fails:
./run_local.sh onnxruntime mobilenet cpu --accuracyNo such file or directory: '/path/to/coco/val2017-300/val_map.txt./run_local.sh onnxruntime mobilenet cpu --accuracy --dataset coco-300Traceback (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!
All the financial advice you’ll ever need fits on a single index card by Ciro Santilli 37 Updated 2025-07-16
Ciro Santilli 37 Updated 2025-07-16A person who gives financial advice, notably personal finance advice. Some of them are questinable guru-like beings, and many are on YouTube.
Unlisted articles are being shown, click here to show only listed articles.