Alexandros Liapatis

alexliap

·

AI & ML interests

Generative AI + Traditional ML

Recent Activity

reacted to FredyRivera-dev's post with 🚀 1 day ago

We wrote a full technical guide on how to train a bilingual (ES/EN) LLM from scratch: TinyQwen. Covers: - Hybrid architecture based on Qwen3.5 - Pre-training with 15B tokens - Cost benchmark between H200 and B200 - Post-training with SFT + LoRA - Full code and data, open source With ~$11 of compute on an H200 we ran an initial training run, enough to validate the full architecture and pipeline. Blog post: https://aquiles-ai.vercel.app/blog/tinyqwen-from-scratch Technical feedback welcome, especially from anyone looking to replicate the pipeline with more compute.

reacted to Nymbo's post with 🔥 2 days ago

Introducing Inflect-v2, two exceptionally small, open-weight English TTS models at just 3.9M and 9.3M parameters. Both generate speech multiple times faster than real-time on CPU. Despite their size, Inflect-v2 delivers quality that is competitive with much larger lightweight TTS systems, including KittenTTS, Piper, and Supertonic-3. CPU, CUDA, PyTorch, and ONNX are supported. Apache 2.0. See it for yourselves: https://huggingface.co/owensong/Inflect-Micro-v2 https://huggingface.co/owensong/Inflect-Nano-v2 Try the Demos: https://huggingface.co/spaces/Nymbo/Inflect-TTS (unlimited CPU usage) https://huggingface.co/spaces/owensong/Inflect-v2 (ultra-fast ZeroGPU usage)

liked a dataset 2 days ago

HuggingFaceCode/stack-v3-train

View all activity

Organizations

upvoted a collection about 1 month ago

ILSP Greek Evaluation Suite

A collection of test sets for evaluating base and chat LLMs (incl. VLMs) on Greek generation and understanding capabilities • 26 items • Updated 8 days ago • 7

upvoted a paper 2 months ago

Gated DeltaNet-2: Decoupling Erase and Write in Linear Attention

Paper • 2605.22791 • Published May 21 • 33

upvoted a collection 5 months ago

🤏 Smol-Data

Tried and tested mixes for strong pretraining. Inspired by https://huggingface.co/blog/codelion/optimal-dataset-mixing • 14 items • Updated Mar 2 • 14

upvoted a collection 9 months ago

Granite 4.0 Nano Language Models

Ultra-compact language models designed for the edge and on-device deployment. • 9 items • Updated Apr 29 • 103

upvoted a paper about 2 years ago

TimeGPT-1

Paper • 2310.03589 • Published Oct 5, 2023 • 7