Machine learning news — deep learning, reinforcement learning, neural architecture search, diffusion models, and new ML frameworks and libraries.
Nateraw's Autotune model on Replicate performs pitch correction on voice audio, shifting vocal frequencies to target pitches using strategies like closest pitch or scale-based normalization. It is designed for music prod…
Meta's 7.8 billion-parameter automatic speech recognition model, Meta-Omnilingual-Asr-7b, supports 1,693 languages with character error rates below 10% for 78% of covered languages. Built by Subformer, the model combines…
A developer has completed a Google Summer of Code 2026 project for the sbi library, replacing its string-based network configuration with typed configuration objects across all estimator families. The new API uses frozen…
A developer has built a collaborative allergy prediction system using federated learning with the Flower framework and PyTorch, enabling training on decentralized health data without uploading raw logs to a central cloud…
Expedia Group is hiring a Machine Learning Engineer III for its Meta/SEM Bidding Programs team in Bangalore, India, to develop and operate production ML pipelines. The role requires 5+ years of experience with end-to-end…
Ornith AI released Ornith-1.5-35B-A3B, a 35B-parameter mixture-of-experts model that activates about 3B parameters per token, trained via a self-improvement loop that jointly optimizes task generation, scaffold construct…
S1 Mini, a fine-tune of Qwen 0.6 billion, cleans messy speech-to-text output locally on under 2GB VRAM, consuming just over 1.4GB in testing. It strips filler words, resolves self-corrections, and formats numbers and dat…
Researchers from Microsoft and academic institutions introduced Internalized Visual Thinking (IVT), a post-training framework that enables multimodal large language models to reason about videos without generating interm…
The book 'Build a DeepSeek Model (From Scratch)' by Raj Abhijit Dandekar, Rajat Dandekar, Sreedath Panat, and Naman Dwivedi rebuilds DeepSeek's innovations on consumer hardware, covering key-value cache, Multi-Head Laten…
On a single 8-GPU AMD Instinct MI355X node, AMD served Kimi Linear 48B-A3B with context lengths from 1024 tokens to 64Mi (67,108,864), using vLLM with FP8 KV cache and tensor parallelism 8, recording prefill TTFT and dec…
Developer Jon Ready released vllm-ios, a native Swift implementation of vLLM-style continuous batching for MLX on iOS, achieving 88% faster multi-agent inference than llama.cpp at 8 concurrent streams with identical weig…
A user study by researchers including Jeniffer Leal-Wyss, Delaney Lothian, Gabriel Bernier-Colborne, Rebecca Knowles, and Michel Simard, presented at the 26th Annual Conference of the European Association for Machine Tra…
A study presented at the 26th Annual Conference of the European Association for Machine Translation (EAMT 2026) evaluated an AI post-editing (AIPE) system across 71,262 production segments in five domains and ten target …
AIDA Agents, a multi-agent translation platform developed by Emanuele Di Rosa and Piotr Peszynski, outperforms all systems on 10 of 11 language pairs in the WMT24++ benchmark (11 languages) and achieves 70–98% publicatio…
Researchers Paolo Di Natale, Elena Chiocchetti, Marlies Alber, and Egon W. Stemle found that reasoning models offer little benefit for simple term insertion but yield clear gains for homonym disambiguation in legal trans…
Since 2023, translators for the Parliament of Canada have had the option to use neural machine translation (NMT) technology provided by the National Research Council of Canada (NRC) to support their work in translating p…
The ARTICULATE project, funded by the CHIST-ERA call 2025, aims to democratize scientific knowledge beyond academia and English-speaking audiences by integrating AI with self-regulated learning to translate science acros…
Judith Klein, in a paper presented at the 26th Annual Conference of the European Association for Machine Translation (EAMT) in June 2026 in Tilburg, Netherlands, introduced advanced CAT tool features in STAR Transit desi…
Ben Houston has built a Neural Texture Compression implementation for Three.js that replaces stacks of correlated PBR textures with a compact multiresolution feature grid and a tiny MLP decoder, with sample materials run…
A technical essay argues that mastering Physics-Informed Neural Networks (PINNs) requires a hybrid background rather than a single major, recommending Computer Science or Mathematics paired with electives in Computationa…