Ahmet's picture

Ahmet

atasoglu

·

atasoglu

AI & ML interests

NLP, LLMs.

Recent Activity

posted an update about 11 hours ago

Introducing ToolsGen 🛠️ I built a tool to solve a problem I kept running into: creating quality datasets for training LLMs to use tools. ToolsGen takes your JSON tool definitions and automatically generates realistic user requests, corresponding tool calls, and evaluates them using an LLM-as-a-judge pipeline. It outputs datasets ready to use with Hugging Face. What makes it useful: - Generates realistic user requests + tool calls from JSON definitions - LLM-as-a-judge quality scoring with multi-dimensional rubrics - Multiple sampling strategies (random, parameter-aware, semantic) - OpenAI-compatible API support - Outputs JSONL with train/val splits Still early days (API isn't stable yet), but it's already helping me generate tool-calling datasets much faster. Check it out: https://github.com/atasoglu/toolsgen Happy to hear feedback or ideas!

upvoted an article 1 day ago

Transformer’ları Cebe Sığdırmak: Modelleri Optimize Edip Uç Cihazlarda Çalıştırma

upvoted an article 1 day ago

Running Large Transformer Models on Mobile and Edge Devices

View all activity

Organizations

upvoted 2 articles 1 day ago

Article

Transformer’ları Cebe Sığdırmak: Modelleri Optimize Edip Uç Cihazlarda Çalıştırma

By

•

2 days ago

• 4

Article

Running Large Transformer Models on Mobile and Edge Devices

By

•

2 days ago

• 10

upvoted an article 7 days ago

Article

ModernVBERT: Towards Smaller Visual Document Retrievers

By

and 4 others •

Oct 3

• 43

upvoted an article 8 days ago

Article

How to Build a Healthcare Robot from Simulation to Deployment with NVIDIA Isaac for Healthcare

By

•

8 days ago

• 17

upvoted 2 collections 9 days ago

Granite 4.0 Nano Language Models

9 items • Updated 6 days ago • 82

Granite 4.0

IBM's new Granite-4.0 models! Run Dynamic GGUFs or fine-tune with Unsloth. • 38 items • Updated 6 days ago • 20

upvoted a collection 12 days ago

MobileLLM-Pro

4 items • Updated 14 days ago • 4

upvoted 2 papers 15 days ago

Guided Decoding and Its Critical Role in Retrieval-Augmented Generation

Paper • 2509.06631 • Published Sep 8 • 10

Turk-LettuceDetect: A Hallucination Detection Models for Turkish RAG Applications

Paper • 2509.17671 • Published Sep 22 • 9

upvoted a collection 27 days ago

Kumru

Kumru LLMs • 2 items • Updated 29 days ago • 9

upvoted a collection about 1 month ago

Pi0

π₀ base models from Physical Intelligence • 4 items • Updated Oct 2 • 3

upvoted an article about 1 month ago

Article

VibeGame: Exploring Vibe Coding Games

Sep 29

• 29

upvoted a collection about 2 months ago

mmBERT: a modern multilingual encoder

mmBERT is trained on 3T tokens from over 1800 languages, showing SoTA scores on benchmarks and exceptional low-resource performance • 16 items • Updated Sep 9 • 48

upvoted an article about 2 months ago

Article

mmBERT: ModernBERT goes Multilingual

Sep 9

• 121

upvoted a collection about 2 months ago

EmbeddingGemma

3 items • Updated Sep 11 • 95

upvoted a collection 2 months ago

FastVLM

Efficient Vision Encoding for Vision Language Models • 9 items • Updated Sep 2 • 103

upvoted a collection 3 months ago

gpt-oss

Open-weight models designed for powerful reasoning, agentic tasks, and versatile developer use cases. • 2 items • Updated Aug 7 • 374

upvoted 3 collections 4 months ago

Orpheus TTS

TTS Towards Human-Sounding Speech • 2 items • Updated Mar 18 • 71

Orpheus Multilingual Research Release

Beta Release of multilingual models. • 12 items • Updated Apr 10 • 104

Trendyol LLM T1

Turkish & English reasoning models • 1 item • Updated Jul 18 • 2