Jan Seben
@everythingllm.bsky.social
Doing ML, Python, C# in Finance and much more
Reposted by Jan Seben
While on the dollar, this from The Economist article, “China is ditching the dollar, fast.”
To be clear, China is nowhere close to replacing the dollar as a reserve currency — and won’t be anytime soon.
What it is doing, though, is slowly eroding the dollar’s influence by building pipes…
1 of 2
To be clear, China is nowhere close to replacing the dollar as a reserve currency — and won’t be anytime soon.
What it is doing, though, is slowly eroding the dollar’s influence by building pipes…
1 of 2
September 16, 2025 at 3:49 PM
While on the dollar, this from The Economist article, “China is ditching the dollar, fast.”
To be clear, China is nowhere close to replacing the dollar as a reserve currency — and won’t be anytime soon.
What it is doing, though, is slowly eroding the dollar’s influence by building pipes…
1 of 2
To be clear, China is nowhere close to replacing the dollar as a reserve currency — and won’t be anytime soon.
What it is doing, though, is slowly eroding the dollar’s influence by building pipes…
1 of 2
How many researchers do you need on your team to keep up with latest findings and test them in your LLM models?
I asked "on the other platform" what were the most important improvements to the original 2017 transformer.
That was quite popular and here is a synthesis of the responses:
That was quite popular and here is a synthesis of the responses:
April 29, 2025 at 4:11 PM
How many researchers do you need on your team to keep up with latest findings and test them in your LLM models?
Lots of assumptions but it seems the article is correct with latest Microsoft downsizing it’s planned compute
GenAI as a business "costs are also so severe that .. paying customers lose .. companies money. .. the most successful company .. appears to have no way to stop burning money... .. only one real company in this industry, OpenAI, and it is ... not a real business." www.wheresyoured.at/wheres-the-m...
There Is No AI Revolution
Soundtrack: Mack Glocky - Chasing Cars
Last week, I spent a great deal of time and words framing the generative AI industry as a cynical con where OpenAI's Sam Altman and Anthropic's Dario Amodei hav...
www.wheresyoured.at
February 25, 2025 at 10:06 AM
Lots of assumptions but it seems the article is correct with latest Microsoft downsizing it’s planned compute
Pretty sure that reason for ChatGPT being down is someone had played an agent to wipe some hardware/data
January 23, 2025 at 4:26 PM
Pretty sure that reason for ChatGPT being down is someone had played an agent to wipe some hardware/data
This is awesome!
Teaching my #DeepLearning in #ComputerVision course next semester. Looking for blind spots in my content. Share your favourite deep learning course in the thread.
Course: www.eecs.yorku.ca/~kosta/Cours...
Course: www.eecs.yorku.ca/~kosta/Cours...
December 24, 2024 at 1:32 AM
This is awesome!
Reposted by Jan Seben
Teaching my #DeepLearning in #ComputerVision course next semester. Looking for blind spots in my content. Share your favourite deep learning course in the thread.
Course: www.eecs.yorku.ca/~kosta/Cours...
Course: www.eecs.yorku.ca/~kosta/Cours...
December 23, 2024 at 12:33 AM
Teaching my #DeepLearning in #ComputerVision course next semester. Looking for blind spots in my content. Share your favourite deep learning course in the thread.
Course: www.eecs.yorku.ca/~kosta/Cours...
Course: www.eecs.yorku.ca/~kosta/Cours...
Reposted by Jan Seben
AWS Lambda works great for compute-intensive tasks such as video processing. However, its stateless nature makes it challenging to support stateful tasks, especially for invoking asynchronous external APIs. As a result, you often end up gluing multiple services together.
Thus, we build DBOS 💙
Thus, we build DBOS 💙
December 21, 2024 at 7:44 PM
AWS Lambda works great for compute-intensive tasks such as video processing. However, its stateless nature makes it challenging to support stateful tasks, especially for invoking asynchronous external APIs. As a result, you often end up gluing multiple services together.
Thus, we build DBOS 💙
Thus, we build DBOS 💙
Reposted by Jan Seben
The wealthy have intentionally destroyed class consciousness so that they can exploit the vast majority of people.
If we’re going to fight back and ultimately win this class war, it’s time we got familiar with how our politics actually work.
jaredyatessexton.substack.com/p/a-thorough...
If we’re going to fight back and ultimately win this class war, it’s time we got familiar with how our politics actually work.
jaredyatessexton.substack.com/p/a-thorough...
A Thoroughly Confused Country: An Introduction to Class Politics and a Way Forward
Americans suffer from a lack of class consciousness, which makes them easy prey for manipulation
jaredyatessexton.substack.com
December 17, 2024 at 3:07 PM
The wealthy have intentionally destroyed class consciousness so that they can exploit the vast majority of people.
If we’re going to fight back and ultimately win this class war, it’s time we got familiar with how our politics actually work.
jaredyatessexton.substack.com/p/a-thorough...
If we’re going to fight back and ultimately win this class war, it’s time we got familiar with how our politics actually work.
jaredyatessexton.substack.com/p/a-thorough...
Reposted by Jan Seben
What sorcery is this? I kept benchmarking QwQ-32B-Preview – and not only does the 4.25-bit version get the same score as the 8-bit (what?), using Qwen2.5-Coder-0.5B as a draft model for speculative decoding sped it up from 27 to 42 tk/s AND it scored even higher (whaaat?)! 🤯
November 30, 2024 at 12:21 AM
What sorcery is this? I kept benchmarking QwQ-32B-Preview – and not only does the 4.25-bit version get the same score as the 8-bit (what?), using Qwen2.5-Coder-0.5B as a draft model for speculative decoding sped it up from 27 to 42 tk/s AND it scored even higher (whaaat?)! 🤯
Reposted by Jan Seben
Reposted by Jan Seben
The cost behind the Python Quant Stack:
• SciPy: $0
• Zipline: $0
• Python: $0
• NumPy: $0
• PyFolio: $0
• Pandas : $0
• OpenBB: $0
• Empyrical: $0
• AlphaLens: $0
• Statsmodels: $0
• RiskFolio-Lib: $0
Anyone can get started.
• SciPy: $0
• Zipline: $0
• Python: $0
• NumPy: $0
• PyFolio: $0
• Pandas : $0
• OpenBB: $0
• Empyrical: $0
• AlphaLens: $0
• Statsmodels: $0
• RiskFolio-Lib: $0
Anyone can get started.
November 23, 2024 at 1:20 AM
The cost behind the Python Quant Stack:
• SciPy: $0
• Zipline: $0
• Python: $0
• NumPy: $0
• PyFolio: $0
• Pandas : $0
• OpenBB: $0
• Empyrical: $0
• AlphaLens: $0
• Statsmodels: $0
• RiskFolio-Lib: $0
Anyone can get started.
• SciPy: $0
• Zipline: $0
• Python: $0
• NumPy: $0
• PyFolio: $0
• Pandas : $0
• OpenBB: $0
• Empyrical: $0
• AlphaLens: $0
• Statsmodels: $0
• RiskFolio-Lib: $0
Anyone can get started.