Pavlo's picture

Pavlo

pmolchanov

·

https://www.pmolchanov.com

AI & ML interests

Efficiency in Multi-Modal LLMs

Recent Activity

updated a model 3 days ago

nvidia/Hymba-1.5B-Instruct

updated a model 3 days ago

nvidia/Hymba-1.5B-Base

upvoted a paper 4 days ago

Hymba: A Hybrid-head Architecture for Small Language Models

View all activity

Organizations

pmolchanov's activity

upvoted a paper 4 days ago

Hymba: A Hybrid-head Architecture for Small Language Models

Paper • 2411.13676 • Published 6 days ago • 35

upvoted a paper 26 days ago

ShadowKV: KV Cache in Shadows for High-Throughput Long-Context LLM Inference

Paper • 2410.21465 • Published 29 days ago • 10

upvoted a paper 27 days ago

COAT: Compressing Optimizer states and Activation for Memory-Efficient FP8 Training

Paper • 2410.19313 • Published Oct 25 • 18

upvoted 2 papers about 2 months ago

PHI-S: Distribution Balancing for Label-Free Multi-Teacher Distillation

Paper • 2410.01680 • Published Oct 2 • 32

Lotus: Diffusion-based Visual Foundation Model for High-quality Dense Prediction

Paper • 2409.18124 • Published Sep 26 • 31

upvoted 2 papers 2 months ago

MaskLLM: Learnable Semi-Structured Sparsity for Large Language Models

Paper • 2409.17481 • Published Sep 26 • 46

Discovering the Gems in Early Layers: Accelerating Long-Context LLMs with 1000x Input Token Reduction

Paper • 2409.17422 • Published Sep 25 • 24

upvoted 2 collections 2 months ago

Llama 3.2

This collection hosts the transformers and original repos of the Llama 3.2 and Llama Guard 3 • 15 items • Updated Oct 24 • 504

MagpieLM

Aligning LMs with Fully Open Recipe (data+training configs+logs) • 9 items • Updated Sep 22 • 15

upvoted 3 collections 3 months ago

RADIO

A collection of Foundation Vision Models that combine multiple models (CLIP, DINOv2, SAM, etc.). • 4 items • Updated 4 days ago • 5

Qwen2-VL

Vision-language model series based on Qwen2 • 15 items • Updated Sep 18 • 157

VILA: On Pre-training for Visual Language Models

10 items • Updated 27 days ago • 45

upvoted a paper 3 months ago

Learning to Move Like Professional Counter-Strike Players

Paper • 2408.13934 • Published Aug 25 • 22

upvoted a collection 3 months ago

Nemotron in vLLM

Nemotron models that have been converted and/or quantized to work well in vLLM • 7 items • Updated Jul 25 • 1

upvoted a paper 3 months ago

Show-o: One Single Transformer to Unify Multimodal Understanding and Generation

Paper • 2408.12528 • Published Aug 22 • 50

upvoted a collection 3 months ago

To read... eventually

A collection of papers that i have read or plan to read all in one place. Includes a wide range of topics. • 124 items • Updated about 4 hours ago • 3

upvoted a paper 3 months ago

LLM Pruning and Distillation in Practice: The Minitron Approach

Paper • 2408.11796 • Published Aug 21 • 55

upvoted a collection 4 months ago

Papers I want to read

Papers in my to-read list • 247 items • Updated 9 days ago • 26

upvoted a paper 4 months ago

Compact Language Models via Pruning and Knowledge Distillation

Paper • 2407.14679 • Published Jul 19 • 38

upvoted a collection 4 months ago

🪐 SmolLM

A series of smol LLMs: 135M, 360M and 1.7B. We release base and Instruct models as well as the training corpus and some WebGPU demos • 12 items • Updated Aug 18 • 200