use-apify.com
Inference: guides & tutorials
Inference runs a trained model on new text or rows. After Apify exports, batch labels, entities, and routing—no full retrain each scrape job.
3 articles
View all tags
Inference runs a trained model on new text or rows, the step that turns scraped pages into labels, entities, and routing decisions. These guides cover batch inference over Apify exports without retraining for each job.
Running inference efficiently means batching, choosing the right model size, and controlling cost per record. Apify datasets give inference a clean, structured feed. Below you will find patterns for classification, extraction, and labeling at scale.
NVIDIA Vera Rubin (announced March 16, 2026 at GTC 2026) is a rack-scale AI platform—five coordinated rack roles (GPU scale-up, dense CPU for agentic workloads, low-latency inference, AI-oriented storage, and rack-to-rack fabric)—not a single-chip release. The headline system is Vera Rubin NVL72 (72 Rubin GPUs, 36 Vera CPUs, NVLink 6). NVIDIA targets pretraining through agentic inference and expects partner products from H2 2026. Treat throughput and $/token multipliers as vendor claims until you benchmark on your models, batch sizes, and SLAs.
Primary source: NVIDIA newsroom: NVIDIA Vera Rubin platform (March 16, 2026).
If you are deciding what to trust first: architecture and rack roles are relatively stable reading; “× faster / × cheaper per token” lines need your proof, not the keynote slide.
This article separates NVIDIA-stated framing from what you should verify, then maps implications for LLM inference, agentic automation, and data products. For the wider news week, see Top 10 AI and tech stories (March 17–24, 2026).
Edge AI is inference (and sometimes light training) close to where data is created—on-device, on-prem, in-vehicle, or on orbit—driven by latency, bandwidth, privacy, and power rather than a single “edge” product category. In 2026, vendors are pushing different slices of that problem at once: Qualcomm’s Dragonwing industrial and networking story, Samsung’s on-device generative AI push, AMD’s x86 laptop NPUs, and NVIDIA’s space-grade accelerated stack alongside GTC 2026.
This article ties those threads together with vendor-primary sourcing where possible, flags figures that are scenario-specific, and ends with a comparison table plus a 90-day builder plan. For the broader news cycle (including NVIDIA’s Vera Rubin platform context), start from our Top 10 AI and tech stories this week (March 17–24, 2026).
Deploy Meta's LLaMA 3 as a high-throughput, OpenAI-compatible inference API on a dedicated GPU server — with quantization, load testing, and Prometheus monitoring included.
If you've run LLaMA 3 locally with Ollama or llama.cpp, you've hit the ceiling: single-user latency is fine, but concurrency crumbles past three or four simultaneous requests. vLLM solves that. It uses PagedAttention and continuous batching to serve dozens of concurrent requests on the same GPU that would otherwise stall at one.
This guide walks through deploying LLaMA 3 (8B and 70B) on a dedicated GPU server using vLLM — covering Docker setup, model quantization, the OpenAI-compatible API, load testing with Locust, and Prometheus monitoring. All steps are tested on an NVIDIA A100 80 GB, which you can rent from Liquid Web.
