Train your own LLM from scratch. This repo builds a GPT-style transformer from the ground up, without using any high-level libraries. You see exactly how attention, multi-head attention, the feed-forward block, embeddings, residuals, and layer norm fit together. And it doesn't
LLMS
-
LangChain uses LLM gateway to control coding agent expenses
By
–
A developer using coding agents can accumulate thousands of dollars in weekly expenses before anyone notices. It happened at LangChain, so we built a solution. How we use the LLM gateway internally
-
AI solves 7/10 hard math problems but still criticized
By
–
Weird headline – I am not sure solving 7 out of 10 novel very hard problems meant AI "did not live up to the task," when 15 months ago LLMs couldn't do math. But the actual study is interesting and illuminates flaws & successes of AIs in math. https://
1stproof.org/assets/docs/re
port.pdf
… -
Model neutrality surpasses cloud neutrality as an offensive tool
By
–
Why model neutrality is more important than cloud neutrality was In the cloud era, companies opted for cloud neutrality as a defense mechanism Model neutrality can be an **offensive** mechanism. It is more
-
Deleted tweet: API users misunderstand frontier model power in native harnesses
By
–
Deleted a tweet on the fact that API users don't understand how much more powerful the frontier models are in native harnesses since I didn't differentiate in the post (limited characters!) between folks carefully evaluating other harnesses for tasks & those just using naked API.
-
LEANN: Index millions with 97% less storage via graph recomputation
By
–
Turn your laptop into a powerful RAG system! LEANN can index and search through millions of documents while using 97% less storage than traditional solutions without accuracy loss. LEANN achieves this through graph-based selective recomputation with high-degree preserving
-
AI models as weapons in new global power struggle
By
–
I believe many people still do not realize that we are in the midst of a new global power struggle over the future. The fact that China’s access to Claude alone is viewed as an immense risk to national security (!) demonstrates the kind of weapons these models represent. The
-
Download new local model as insurance against API loss
By
–
DO YOURSELF A FAVOR: GO DOWNLOAD THIS NEW LOCAL MODEL AND KEEP IT IN STORAGE. Even if you don't have a massive GPU setup, having offline access to an intelligent model is a crucial insurance policy. Free API access won't necessarily last forever. Right now, the 12B-27B range
-
PyMuPDF: Analyze PDFs for RAG with Azure Layout
By
–
PyMuPDF: Parse PDFs for RAG with Azure Layout! #BigData #Analytics #DataScience #AI #MachineLearning #NLProc #LLM #IoT #IIoT #PyTorch #Python #RStats #TensorFlow #Java #JavaScript #ReactJS #GoLang #CloudComputing #Serverless #DataScientist #Linux #Programming #Coding
-
Fable, a watered-down version of Mythos, refutes criticism of access and cost
By
–
Indeed, Fable is a version of Mythos (Mythos is a class of models) that has guardrails. This only reinforces my points: saying that Mythos is 1- restricted access, 2- poorly evaluated, 3- too expensive, is false, since we have a watered-down version that is 1- used.