b_mc2

Some additional context:

- Gortyn Code "Thanks" speech video [1]

- Mike Towndrow/Indie Game Awards Retraction announcement video [2]

- IGA FAQ Game Eligibility, info on retraction [3]

"Initially discovered through itch.io’s Game Boy Competition 2023 and later played on cart, Gortyn Code was selected as an Indie Vanguard due to their impressive work in GB Studio and for crafting such an amazing throwback for the modern day. The physical cart of Chantey is being produced and sold by ModRetro, and it is the sole marketplace where it can be purchased. The IGAs nomination committee were unfortunately made aware of ModRetro’s vile nature the day after the 2025 premiere with the news of their horrid and disgusting handheld console. As the company strictly goes against the values of the IGAs, and due to the ties with ModRetro, the Indie Vanguard recognition has also been retracted.

The decision does not reflect Gortyn Code, but ModRetro alone. Chantey remains a wonderful throwback to the Game Boy era. We encourage you to continue following their journey on itch.io."

- ModRetro's Anduril Edition Chromatic, all proceeds go to support veteran suicide prevention [4][5]

---

[1] https://www.twitch.tv/videos/2647339751?t=0h45m15s

[2] https://bsky.app/profile/indiegameawards.gg/post/3magufzccy2...

[3] https://www.indiegameawards.gg/faq#:~:text=Initially%20disco...

[4] https://modretro.com/products/anduril-chromatic-porta-pro-bu...

[5] https://x.com/PalmerLuckey/status/2002221958700872045

These are two articles I liked that are referenced in the Python ImageHash library on PyPi, second article is a follow-up to the first.

Here's paraphrased steps/result from first article for hashing an image:

1. Reduce size. The fastest way to remove high frequencies and detail is to shrink the image. In this case, shrink it to 8x8 so that there are 64 total pixels.

2. Reduce color. The tiny 8x8 picture is converted to a grayscale. This changes the hash from 64 pixels (64 red, 64 green, and 64 blue) to 64 total colors.

3. Average the colors. Compute the mean value of the 64 colors.

4. Compute the bits. Each bit is simply set based on whether the color value is above or below the mean.

5. Construct the hash. Set the 64 bits into a 64-bit integer. The order does not matter, just as long as you are consistent.

The resulting hash won't change if the image is scaled or the aspect ratio changes. Increasing or decreasing the brightness or contrast, or even altering the colors won't dramatically change the hash value.

https://www.hackerfactor.com/blog/index.php?/archives/432-Lo...

https://www.hackerfactor.com/blog/index.php?/archives/529-Ki...

"Keep in mind that the 5th Congress did not really need to struggle over the intentions of the drafters of the Constitutions in creating this Act as many of its members were the drafters of the Constitution."

"Clearly, the nation's founders serving in the 5th Congress, and there were many of them, believed that mandated health insurance coverage was permitted within the limits established by our Constitution."

This seems like a fallacy of composition, and done so to try and persuade the reader. By my rough count, just 6 of the original founders that signed the Constitution were still in Congress at this time, or just 18% of the signers[1]. There's no roll call vote that I can find, but signer Charles Pinckney had voiced general oppositions and thought "it only reasonable and equitable that these persons pay for the benefit for which they were themselves to receive, and it would be neither just nor fair for other persons to pay it"[2]

"And when the Bill came to the desk of President John Adams for signature, I think it’s safe to assume that the man in that chair had a pretty good grasp on what the framers had in mind."

This just points to the same argument that's always being made between Spirit vs Letter of the law proponents, ~4% of Congress during the 5th Congress were signers of the Constitution and we don't know how they even voted on this. So ~96% of Congress were basically in the Spirit vs Letter dispute that we're in today.

[1] https://www.constitutionfacts.com/content/constitution/files... [2] https://www.congress.gov/annals-of-congress/page-headings/5t...

This is awesome, congratulations. I'm glad to see some text-to-sql models being created. Shameless plug: I also just realized you used NSText2SQL[1] which itself contains my text-to-sql dataset, sql-create-context[2], so I'm honored. I used sqlglot pretty heavily on it as well.

Do you think a 3B model might also be in the future, or something small enough that can be loaded up in Transformers.js?

[1] https://huggingface.co/datasets/NumbersStation/NSText2SQL

[2] https://huggingface.co/datasets/b-mc2/sql-create-context

I also think this is the route we are heading, a few 1-7B or 14B param models that are very good at their tasks, stitched together with a model that's very good at delegating. Huggingface has Transformers Agents which "provides a natural language API on top of transformers: we define a set of curated tools and design an agent to interpret natural language and to use these tools"

Some of the tools it already has are:

Document question answering: given a document (such as a PDF) in image format, answer a question on this document (Donut)

Text question answering: given a long text and a question, answer the question in the text (Flan-T5)

Unconditional image captioning: Caption the image! (BLIP)

Image question answering: given an image, answer a question on this image (VILT)

Image segmentation: given an image and a prompt, output the segmentation mask of that prompt (CLIPSeg)

Speech to text: given an audio recording of a person talking, transcribe the speech into text (Whisper)

Text to speech: convert text to speech (SpeechT5)

Zero-shot text classification: given a text and a list of labels, identify to which label the text corresponds the most (BART)

Text summarization: summarize a long text in one or a few sentences (BART)

Translation: translate the text into a given language (NLLB)

Text downloader: to download a text from a web URL

Text to image: generate an image according to a prompt, leveraging stable diffusion

Image transformation: modify an image given an initial image and a prompt, leveraging instruct pix2pix stable diffusion

Text to video: generate a small video according to a prompt, leveraging damo-vilab

It's written in a way that allows the addition of custom tools so you can add use cases or swap models in and out.

https://huggingface.co/docs/transformers/transformers_agents

They actually reuse the gold flakes from the Dome by mixing it into the paint used on the football team's helmets

Looks really cool! Gives me similar vibes as EDEX-UI[1]

[1] https://github.com/GitSquared/edex-ui

This shows up as zero for me but the badge site says I've used it twice. It's pretty easy to manually check since I haven't made too many comments. I suspect the badge site is fuzzy matching since "luck" or some variation has appeared twice (now three times)

A reason I like it is I have an "older" AMD GPU which is no longer supported by ROCm (sort of AMDs version of Cuda) which means running locally I'm either trying to figure out older ROCm builds to use my GPU and running into dependency issues or using my CPU which isn't that great either. But with WebGPU I'm able to run these models on my GPU which has been much faster than using the .cpp builds.

Its also fairly easy to route a Flask server to these models with websockets, so with that I've been able to run python and pass data to the model to run on the GPU and pass the response back to the program. Again, there's probably a better way but its cool to have my own personal API for a LLM.

A pretty cool library that uses ONNX is transformers.js [1] and they're already working to add WebGPU support.[2]

[1] https://xenova.github.io/transformers.js/

[2] https://twitter.com/xenovacom/status/1650634015060156420

I just finished augmenting some of the Spider and WikiSQL data on huggingface [1] I initially intended to train a text-to-sql LLM that would take a natural language question and be provided the CREATE TABLE statements to provide some grounding for the response. So instead of hallucinating the column and table names by using the question alone, I was hoping the CREATE TABLE statement would limit the choices. We'll see if it's useful or not, but funny enough I came across this article after I finished the dataset.[2]

I'd definitely like to see how others are doing it.

[1] https://huggingface.co/datasets/b-mc2/sql-create-context

[2] https://blog.langchain.dev/llms-and-sql/

I've got it running on Chrome v113 beta on Ubuntu with an older AMD RX 580. The feature flags don't seem to be taking for me in chrome GUI but if you start chrome from terminal like this it works.

google-chrome --enable-unsafe-webgpu --enable-features=Vulkan,UseSkiaRenderer --enable-dawn-features=disable_robustness

GPU doesn't work in --headless though.

I work as a data engineer for a Defense company. It's not always stable but depends on the contract you can get on, some have a pretty broad scope and arent going anywhere, others are small projects with less stable funding.

Most of what I work with is Databricks, Python, Pyspark, etc. So most coding is done in notebooks. I haven't felt that there was a lot of restrictions coding wise, but getting access to different databases, clusters, and especially on boarding can be a pain.

Environment wise its been pretty good, you find even the clients are often contractors themselves, but if you dislike them you can always roll off and onto another project, it seems everyone needs more devs

There's a few resources out there for smaller companies to try to get into defense, mainly networks and incubators. [1]

But you'll also see SBIR/STTR funding topics (program run by the Small Business Administration) [2]

You can also try checking out some of the bigger Defense contractors, some have incubator programs and are looking to expand their small business ecosystem for subcontracts, part of that includes funding. (Disclaimer, I work here) [3]

[1] https://defensewerx.org/

[2] https://www.sbir.gov/node/2214225

[3] https://www.boozallen.com/expertise/innovation/ventures.html

I've used this recently to do some fuzzy matching of column names in datasets, I also added it to a small python one-liner library I've been making for practice. p.s. don't give me flack, I know this isn't an efficient way to do things.

jaccard = lambda A, B: len(set(A).intersection(set(B))) / len(set(A).union(set(B)))

https://github.com/b-mc2/MiniMath

Minetest is definitely fun, and I've had good luck using Miney to build in it with Python

https://miney.readthedocs.io/en/dev/index.html

If I'm not trying to build a very specific graph or chart, and just exploring data I usually use either Rawgraphs or Sqliteviz. Rawgraphs is nice if you just want to swap visualizations out with smaller data as is, sqliteviz seems to handle much larger datasets and let's you use SQL if you want to change the resultset. Both seem to keep data local too and I know sqliteviz works offline, rawgraphs might too.

https://www.rawgraphs.io/

https://sqliteviz.com/

Here's a tweet with some pretty cool footage from the cameras on Mt. Washington yesterday.

https://twitter.com/forecaster25/status/1621598650022363142

Very cool benchmark. I don't think this will be a huge difference but could you try mypyc? I'm definitely curious how it compares. It uses mypy to perform type checks and claims "Existing code with type annotations is often 1.5x to 5x faster when compiled. Code tuned for mypyc can be 5x to 10x faster."[1]

[1] https://mypyc.readthedocs.io/en/latest/introduction.html

This seems opposite to what their new president said last month:

"From the beginning, the team behind Signal put people and their needs at the core of their commitments. They understood that iron-clad security is fairly pointless if people can’t use, access, or feel comfortable with it. In other words, if my friends won’t use a messaging app, it doesn’t work as a messaging app. It works as a thought experiment, at best. Understanding this, Signal’s developers and designers created an app that honors people’s needs and expectations, while maintaining strict privacy promises."[1]

I'll echo the other comments here talking about network effects, onboarding friction, and social capital wasted convincing friends/family.

[1] https://signal.org/blog/announcing-signal-president/

> Please don't complain that a submission is inappropriate. If a story is spam or off-topic, flag it. Don't feed egregious comments by replying; flag them instead. If you flag, please don't also comment that you did. [1]

https://news.ycombinator.com/newsguidelines.html

François Chollet posted some more information on Twitter about this, including some pictures of counterfeits.

https://twitter.com/fchollet/status/1550932298572263424

Agreed. This banner is the new warrant canary [1]

The Legal Action Canary?

[1] https://en.wikipedia.org/wiki/Warrant_canary

"In the absence of Big Tech, you’d expect smaller startup companies to rush in to fill the gap. But, Luckey explained, startups find it difficult to seize the opportunity... It’s very hard to raise money; It’s very unpopular with a lot of investors, especially the ESG type investors, which represents $30 trillion in global capital"

Pretty on topic, I just posted a Fedscoop article about Booz Allen setting up a VC fund to invest in startups providing tech to federal agencies. $100M with a target of 4-6 deals a year and holding a portfolio of ~30 companies.

https://news.ycombinator.com/item?id=32111983

See also: https://ml5js.org/

"The library provides access to machine learning algorithms and models in the browser, building on top of TensorFlow.js with no other external dependencies."

Quart Github: https://github.com/pallets/quart

Go to for me has been a simple Flask framework and HTMX to make my sites seem and feel more dynamic, and then deploy the whole thing with Zappa as an AWS Lambda function. Super simple to add a new endpoint in Flask and ping it with HTMX.

HTMX has been great because I've just added a single endpoint in Flask that HTMX pings and makes it feel and respond like a single page application. You can have a flask endpoint, use Jinja to create the HTML and plug it into your page with HTMX. really nice and simple.

Agreed with both of these. Flask and HTMX has been an awesome combination.

Flask Github: https://github.com/pallets/flask