Trending

O3 Beats a Master-Level GeoGuessr Player–Even with Fake EXIF Data

Hestus, Inc. (YC S24) Is Hiring an ML Engineer to Revolutionize CAD

Qwen3: Think Deeper, Act Faster

Implement Flash Attention Back End in SGLang – Basics and KV Cache

Vision Transformers Need Registers (extra tokens discarded after last layer)

Inference-Aware Fine-Tuning for Best-of-N Sampling in Large Language Models

DeepSeek-R2

Watching 03 Model Sweat over a Paul Morphy Mate-in-2

Calibrations Have a Context-Collapse Problem

LLM GPU calculator for inference and fine-tuning requirements

Granite-4-Tiny-Preview is a 7B A1 MoE

A company gave 1,000 AI agents access to Minecraft — and they built a society

ubergarm/Qwen3-30B-A3B-GGUF 1600 tok/sec PP, 105 tok/sec TG on 3090TI FE 24GB VRAM

[D] Are weight offloading / weight streaming approaches like in Deepseek Zero used frequently in practice? (For enabling inference on disproportionately undersized GPUs)

Build a Text-to-SQL AI Assistant with DeepSeek, LangChain and Streamlit

New TTS/ASR Model that is better that Whisper3-large with fewer paramters

Impressive Qwen 3 30 MoE

[D] DE vs Gen AI/ML

DeepSeek

Save item

DeepSeek-Prover-V2-671B

StepFun

Save item

Step1X-Edit

Task: image-to-image

HiDream.ai

Save item

HiDream-E1-Full

Task: any-to-any

freepik

Save item

F-Lite

Task: text-to-image

NVIDIA

Save item

parakeet-tdt-0.6b-v2

Task: automatic-speech-recognition

DeepSeek

Save item

DeepSeek-V3-0324

NousResearch

Save item

Minos-v1

Task: text-classification

HiDream.ai

Save item

HiDream-I1-Full

Task: text-to-image

NVIDIA

Save item

DAM-3B

Task: image-text-to-text

tracel-ai

Save item

burn

Burn is a next generation Deep Learning Framework that doesn't compromise on flexibility, efficiency and portability.

modelscope

Save item

FunASR

A Fundamental End-to-End Speech Recognition Toolkit and Open Source SOTA Pretrained Models, Supporting Speech Recognition, Voice Activity Detection, Text Post-processing etc.

coqui-ai

Save item

TTS

🐸💬 - a deep learning toolkit for Text-to-Speech, battle-tested in research and production

NVIDIA

Save item

cuda-python

CUDA Python: Performance meets Productivity

microsoft

Save item

onnxruntime

ONNX Runtime: cross-platform, high performance ML inferencing and training accelerator

google-deepmind

Save item

mujoco

Multi-Joint dynamics with Contact. A general purpose physics simulator.

NVIDIA

Save item

Megatron-LM

Ongoing research training transformer models at scale

jax-ml

Save item

jax

Composable transformations of Python+NumPy programs: differentiate, vectorize, JIT to GPU/TPU, and more

huggingface

Save item

pytorch-image-models

The largest collection of PyTorch image encoders / backbones. Including train, eval, inference, export scripts, and pretrained weights -- ResNet, ResNeXT, EfficientNet, NFNet, Vision Transformer (ViT), MobileNetV4, MobileNet-V3 & V2, RegNet, DPN, CSPNet, Swin Transformer, MaxViT, CoAtNet, ConvNeXt, and more