Yash Marathe's picture

In a Training Loop 🔄

Yash Marathe

yashmarathe

·

AI & ML interests

None yet

Recent Activity

liked a dataset about 8 hours ago

knowrohit07/saraswati-stem

liked a dataset about 8 hours ago

knowrohit07/know-saraswati-cot

liked a model 4 days ago

Overworld/Waypoint-1-Small

View all activity

Organizations

upvoted an article about 2 months ago

Article

Accelerate ND-Parallel: A guide to Efficient Multi-GPU Training

+3

Aug 8, 2025

•

91

upvoted a paper 2 months ago

Virtual Width Networks

Paper • 2511.11238 • Published Nov 14, 2025 • 38

upvoted a paper 3 months ago

Supervised Reinforcement Learning: From Expert Trajectories to Step-wise Reasoning

Paper • 2510.25992 • Published Oct 29, 2025 • 48

upvoted 2 articles 4 months ago

Article

Gaia2 and ARE: Empowering the community to study agents

+9

Sep 22, 2025

•

125

Article

Ultra-Long Sequence Parallelism: Ulysses + Ring-Attention Technical Principles and Implementation

Sep 16, 2025

•

17

upvoted a paper 4 months ago

Towards General Agentic Intelligence via Environment Scaling

Paper • 2509.13311 • Published Sep 16, 2025 • 71

upvoted 4 collections 7 months ago

SuperBPE

SuperBPE tokenizers and models trained with them • 9 items • Updated 19 days ago • 17

💧 LFM2

LFM2 is a new generation of hybrid models, designed for on-device deployment. • 27 items • Updated 14 days ago • 136

Hybrid Linear Attention Research

All 1.3B & 340M hybrid linear-attention experiments. • 62 items • Updated Sep 11, 2025 • 13

Avey 1 Research Preview

1.5B preview models trained on 100B tokens of FineWeb, and an instruct-tuned version on smoltalk. • 3 items • Updated Jun 16, 2025 • 6

upvoted 2 collections 8 months ago

V-JEPA 2

A frontier video understanding model developed by FAIR, Meta, which extends the pretraining objectives of https://ai.meta.com/blog/v-jepa-yann • 8 items • Updated Jun 13, 2025 • 188

Falcon-H1

Falcon-H1 Family of Hybrid-Head Language Models (Transformer-SSM), including 0.5B, 1.5B, 1.5B-Deep, 3B, 7B, and 34B (pretrained & instruction-tuned). • 39 items • Updated 17 days ago • 59

upvoted 3 collections 9 months ago

LipSync and Face Operations

23 items • Updated 24 days ago • 62

Perception LM

7 items • Updated Apr 17, 2025 • 63

Perception Encoder

17 items • Updated Jul 11, 2025 • 77

upvoted 3 collections 10 months ago

Skywork-OR1

Skywork Open Reasoner 1 • 11 items • Updated May 29, 2025 • 31

Kimina Prover Preview

State-of-the-Art Models for Formal Mathematical Reasoning • 5 items • Updated Apr 28, 2025 • 33

Kimi-VL-A3B

Moonshot's efficient MoE VLMs, exceptional on agent, long-context, and thinking • 7 items • Updated Oct 30, 2025 • 78

upvoted an article 10 months ago

Article

makeMoE: Implement a Sparse Mixture of Experts Language Model from Scratch

May 7, 2024

•

115

upvoted a collection 10 months ago

Cosmos

⚠️ This collection is archived. 👉 https://huggingface.co/collections/nvidia/nvidia-cosmos-2 • 31 items • Updated 6 days ago • 299