llm.c update: Our single file of 2,000 ~clean lines of C/CUDA code now trains GPT-2 (124M) on GPU at speeds ~matching PyTorch (fp32, no flash attention) https://
github.com/karpathy/llm.c
/blob/master/train_gpt2.cu
… On my A100 I'm seeing 78ms/iter for llm.c and 80ms/iter for PyTorch. Keeping in mind this is fp32,
CODE
-
llm.c Matches PyTorch Performance Training GPT-2 on GPU
By
–
-
Managing Python Dependencies Across Multiple Development Environments
By
–
does anyone else have a requirements.txt that you just keep adding to every time you run into a package you need and carry it over to literally every single env you work in
-
LangChain Weekly Release: Evaluations, Tool Calling, Monitoring
By
–
LangChain Release Notes, week of 4/15 Evaluations video series Standardized tool calling in LangChain Production monitoring and automations in LangSmith New RAG From Scratch videos
Community created content! Read it all here: https://
blog.langchain.dev/week-of-4-15-l
angchain-release-notes/
… -
RAG vs Fine-tuning for Small Document Loading
By
–
My understanding is that if you want to eg load your company employee handbook (~20 page PDF) into a model you're likely to get much better results with RAG than with fine-tuning – can you fine tune such that the model reliable answers questions about something that small?
-
Building Reliable Local RAG Agents with Llama3
By
–
Reliable, fully local RAG agents with Llama3 With the release of Llama3, there's high interest in agents that run reliably & locally (e.g., on your laptop). Here, we show to how build reliable local agents using LangGraph and Llama3-8b from scratch. As a example, we combine
-
AI Tools for Code Generation and Content Creation
By
–
Create apps on the fly, fix or improve code, write long chunks of text, come up with research docs.
-
Test Files as Documentation When Docs Are Outdated
By
–
i am literally reading their test files as it's the only docs thats not outdated and even worse misleading 😀
-
Structured Data Extraction from Images Using Claude 3 Opus
By
–
Beautiful example here of structured data extraction from an image using Claude 3 Opus – most of the LLM implementation is handled by this tools schema: https://
github.com/timpaul/form-e
xtractor-prototype/blob/main/data/extract-form-questions.json
… -
Open Web and Open-Source CMS Value for Government
By
–
I hope we can help government and regulatory bodies clearly understand the value of keeping the web open and using open-source software in the CMS."
