Sublime

No more needle in a haystack. LLMs now retrieve over 1M tokens—no RAG, no retraining. Infinite Retrieval introduces InfiniRetri, a method that uses transformer attention to handle infinite-length contexts natively. No toolkits. No memory modules. https://t.co/JTxlJweASn

Introducing Rakis: https://t.co/bXJDcPwa7J has processed quarter of a million distributed tokens already Crazy what we can do in the browser now. Password, join and you're providing inference to a browser-only p2p swarm - no servers co-produced with @dgt_julian... See more

"Open-source developer platform to power your entire infra and turn scripts into webhooks, workflows and UIs. Fastest workflow engine (13x vs Airflow). Open-source alternative to Retool and Temporal." https://t.co/dmz6Qc87wD

Introducing: Workflow Use – Deterministic, Self-Healing Browser Automation (RPA 2.0) ⚙️ Browser Use is great for many tasks (QA testing, form filling, finding products, etc.🌐), but enterprise automation is different. Companies often have one workflow (e.g., fill out form & download PDFs) to run 1M+ times... See more

Why Infinity:

Infinity provides the following features:

• Deploy virtually any SentenceTransformer - deploy the model you know from SentenceTransformers


• Fast inference backends : The inference server is built on top of torch, fastembed(onnx-cpu) and CTranslate2, getting most out of your CUDA or CPU hardware.


• Dynamic batching : New embedding requests

... See more