PZ's picture

PZ PRO

philipp-zettl

·

philipp-zettl

AI & ML interests

NLP/CV/Multimodal learning

Recent Activity

upvoted an article 3 days ago

FineWeb2-C: Help Build Better Language Models in Your Language

reacted to mitkox's post with 🚀 3 days ago

llama.cpp is 26.8% faster than ollama. I have upgraded both, and using the same settings, I am running the same DeepSeek R1 Distill 1.5B on the same hardware. It's an Apples to Apples comparison. Total duration: llama.cpp 6.85 sec <- 26.8% faster ollama 8.69 sec Breakdown by phase: Model loading llama.cpp 241 ms <- 2x faster ollama 553 ms Prompt processing llama.cpp 416.04 tokens/s with an eval time 45.67 ms <- 10x faster ollama 42.17 tokens/s with an eval time of 498 ms Token generation llama.cpp 137.79 tokens/s with an eval time 6.62 sec <- 13% faster ollama 122.07 tokens/s with an eval time 7.64 sec llama.cpp is LLM inference in C/C++; ollama adds abstraction layers and marketing. Make sure you own your AI. AI in the cloud is not aligned with you; it's aligned with the company that owns it.

reacted to fdaudens's post with ❤️ 3 days ago

Yes, DeepSeek R1's release is impressive. But the real story is what happened in just 7 days after: - Original release: 8 models, 540K downloads. Just the beginning... - The community turned those open-weight models into +550 NEW models on Hugging Face. Total downloads? 2.5M—nearly 5X the originals. The reason? DeepSeek models are open-weight, letting anyone build on top of them. Interesting to note that the community focused on quantized versions for better efficiency & accessibility. They want models that use less memory, run faster, and are more energy-efficient. When you empower builders, innovation explodes. For everyone. 🚀 The most popular community model? @bartowski's DeepSeek-R1-Distill-Qwen-32B-GGUF version — 1M downloads alone.

View all activity

Organizations

philipp-zettl's activity

upvoted an article 3 days ago

Article

FineWeb2-C: Help Build Better Language Models in Your Language

By

•

Dec 23, 2024

• 18

upvoted a paper 17 days ago

The GAN is dead; long live the GAN! A Modern GAN Baseline

Paper • 2501.05441 • Published 21 days ago • 87

upvoted 2 papers about 1 month ago

1.58-bit FLUX

Paper • 2412.18653 • Published Dec 24, 2024 • 74

Smarter, Better, Faster, Longer: A Modern Bidirectional Encoder for Fast, Memory Efficient, and Long Context Finetuning and Inference

Paper • 2412.13663 • Published Dec 18, 2024 • 125

upvoted 2 collections about 2 months ago

Llama 3.3

This collection hosts the transformers and original repos of the Llama 3.3 • 1 item • Updated Dec 6, 2024 • 124

Common Models

The first generation of models pretrained on Common Corpus. • 5 items • Updated Dec 5, 2024 • 28

upvoted an article about 2 months ago

Article

They Said It Couldn’t Be Done

By

•

Dec 5, 2024

• 77

upvoted a collection about 2 months ago

LLäMmlein Chat Preview 🐑

https://www.informatik.uni-wuerzburg.de/datascience/projects/nlp/llammlein/ • 8 items • Updated Nov 22, 2024 • 11

upvoted a collection 2 months ago

INTELLECT-1

13 items • Updated Dec 5, 2024 • 7

upvoted a paper 2 months ago

Cautious Optimizers: Improving Training with One Line of Code

Paper • 2411.16085 • Published Nov 25, 2024 • 15

upvoted a collection 3 months ago

SmolLM2

State-of-the-art compact LLMs for on-device applications: 1.7B, 360M, 135M • 15 items • Updated Dec 22, 2024 • 208

upvoted a collection 4 months ago

Moshi v0.1 Release

MLX, Candle & PyTorch model checkpoints released as part of the Moshi release from Kyutai. Run inference via: https://github.com/kyutai-labs/moshi • 13 items • Updated Sep 18, 2024 • 225

upvoted an article 4 months ago

Article

HTRflow - A tool for HTR and OCR

By

•

Oct 1, 2024

• 15

upvoted a collection 4 months ago

Realistic Vision (SD1.5)

8 items • Updated Dec 4, 2023 • 33

upvoted an article 4 months ago

Article

Llama can now see and run on your device - welcome Llama 3.2

Sep 25, 2024

• 182

upvoted a paper 4 months ago

Florence-2: Advancing a Unified Representation for a Variety of Vision Tasks

Paper • 2311.06242 • Published Nov 10, 2023 • 89

upvoted a collection 5 months ago

Zeroshot Classifiers

These are my current best zeroshot classifiers. Some of my older models are downloaded more often, but the models in this collection are newer/better. • 12 items • Updated 24 days ago • 125

upvoted an article 6 months ago

Article

Introducing TextImage Augmentation for Document Images

Aug 6, 2024

• 32

upvoted 2 articles 7 months ago

Article

How we leveraged distilabel to create an Argilla 2.0 Chatbot

Jul 16, 2024

• 32

Article

Thoughts on LoRA Training #1

By

•

Jun 18, 2024

• 33