报菜名
报菜名
HuggingGPT (could’ve chosen a better name, linkedin tier)
NOOO, THEY AUTOMATED L3s (KINDA!)
Can The foundation be just an LLM? If only Hari Seldon read this paper
DECKARD - RL Agent that dreams
Training LLMs using AI generated dialogues
Automating Data Analysts [By Microsoft(™)]
(FLARE) Active Retrieval Augmented Generation
Hack to make inference faster (by HuggingFace)
Yes your models can memorize exact stuff
Voyager [Diamond ranked AI Minecraft player]
Activation-aware Weight Quantisation (AWQ)
SpQR (Sparse Quantised Representation)
SOTA document bender for your company QA
Insane alpha drop from kaiokendev
Skinny dip into GGML code base
How to check fine tuning datasets’ quality?
DPO (Direct Preference Optimization)
Symbol Rank ( for coding LLMs)
Scaling S3 is not easy [Not related to ML but also related to AI cause all data is in S3]
MoE (by Deepmind) (It’s soft not sparse)
Estimate LLM Flops and Memory requirement
How to reduce KV cache mem usage?
Ok, I am going to become Vector DB expert this week
Mixture of Experts: PEFT edition by Cohere
Generative Recommendors - Cool paper by Google
Fusing Modalities - Chimera by Meta
IMPORTANT INTERPRETABILITY PAPER BY ANTHROPIC
It’s not AGI (it’s just your data)
Insane ML Notes on Twitter with Q&A
Stable Diffusion Turbo (or How to distill a diffusion model 101)
I can’t hear the MUSIC* !!!!!!! NEEEED TO GET BETTTTTTTER!!!
Mamba - faster architecture (Reading cause Tri Dao is author)
Use smol models to train large models faster
LLM Paper from Apple?? : That’s a rare sight
Multimodal paper from Apple???
Amazing paper to Learn about Dingboard
TDM edge Multimodal arc (I blame Vik)
Embarrassing myself publicly arc (PHOTOMAKER)
ILYA’s READING LIST (For getting up to speed on today’s architectures)
Stream Diffusion - Brrrrr ImageGen at 100FPS
MLLM-Guided Image Editing (MGIE)
Generalising Length of Transformers
Fashion Diffusion (Make your waifu dress in Zara)
Another Apple LLM (this time it’s multimodal)
Quiet-Star (Is it really the fabled openai algo, nope)