Join us

ContentUpdates and recent posts about Unsloth..
Link
@kala shared a link, 1ย month ago
FAUN.dev()

AI Is Doing the Testing Now

Brijesh Deb's third "comfortable lie" of software testing is that AI is now doing the testing: coverage dashboards hit 80%+, regression suites maintain themselves, and leadership concludes that risk is handled, while the experienced testers who knew the domain quietly get redeployed or made redundan.. read more ย 

Link
@devopslinks shared a link, 1ย month ago
FAUN.dev()

The invisible engineering behind Lambdaโ€™s network

AWS engineers explain how the Lambda team rebuilt VPC networking so they can keep per-invocation setup off the hot path and run dense microVM workers at scale... read more ย 

The invisible engineering behind Lambdaโ€™s network
Link
@devopslinks shared a link, 1ย month ago
FAUN.dev()

Shift Left Did Not Fix It

Shift left has become a buzzword, but merely moving testing earlier doesn't address the core issue of authority and decision-making in quality assurance. AI may offer quicker testing, but it doesn't comprehend risk like human testers do - beware the dangerous lie that AI can replace thorough, critic.. read more ย 

Link
@devopslinks shared a link, 1ย month ago
FAUN.dev()

Terraform is dead

Graham Gilbert argues Terraform is effectively dead, kept alive only by inertia: HCL forced engineers to translate intent (the diagrams, paragraphs, and constraints that actually describe systems) into a DSL that nobody naturally thinks in, while fragmenting infrastructure, application logic, polici.. read more ย 

Link
@devopslinks shared a link, 1ย month ago
FAUN.dev()

Finding zombies in our systems: A real-world story of CPU bottlenecks

After a network outage crisis, Pinterest's ML Platform team discovered high Kubernetes agent CPU usage was causing critical Ray training job failures. The team's deep profiling strategy revealed a rarely seen flaw in how Kubelet was handling memory cgroup iterations... read more ย 

Finding zombies in our systems: A real-world story of CPU bottlenecks
Link
@devopslinks shared a link, 1ย month ago
FAUN.dev()

AI in SRE: What's Actually Coming in 2026

AI in SRE is evolving, with true value in Root Cause Analysis and Pre-Change Impact Analysis, not autonomous remediation or AI replacing SREs - it's about collaboration and focus evolution... read more ย 

Story
@laura_garcia shared a post, 1ย month ago
Software Developer, RELIANOID

๐—Ÿ๐—ถ๐—ป๐˜‚๐˜… ๐—ธ๐—ฒ๐—ฟ๐—ป๐—ฒ๐—น ๐˜ƒ๐˜‚๐—น๐—ป๐—ฒ๐—ฟ๐—ฎ๐—ฏ๐—ถ๐—น๐—ถ๐˜๐—ถ๐—ฒ๐˜€

๐Ÿ” ๐—Ÿ๐—ถ๐—ป๐˜‚๐˜… ๐—ธ๐—ฒ๐—ฟ๐—ป๐—ฒ๐—น ๐˜ƒ๐˜‚๐—น๐—ป๐—ฒ๐—ฟ๐—ฎ๐—ฏ๐—ถ๐—น๐—ถ๐˜๐—ถ๐—ฒ๐˜€ are once again forcing enterprises to rethink ๐™ž๐™ฃ๐™›๐™ง๐™–๐™จ๐™ฉ๐™ง๐™ช๐™˜๐™ฉ๐™ช๐™ง๐™š ๐™จ๐™š๐™˜๐™ช๐™ง๐™ž๐™ฉ๐™ฎ ๐™ฅ๐™ง๐™ž๐™ค๐™ง๐™ž๐™ฉ๐™ž๐™š๐™จ. The recent disclosure of โ€œ๐—–๐—ผ๐—ฝ๐˜† ๐—™๐—ฎ๐—ถ๐—นโ€ and โ€œ๐——๐—ถ๐—ฟ๐˜๐˜† ๐—™๐—ฟ๐—ฎ๐—ดโ€ highlights how kernel-level flaws can rapidly evolve into major risks for cloud environments, containers, Kubernetes clusters, and cr..

ย Activity
@cristiandeluxe started using tool WordPress , 1ย month ago.
ย Activity
@cristiandeluxe started using tool Vue.js , 1ย month ago.
ย Activity
@cristiandeluxe started using tool TypeScript , 1ย month ago.
Unsloth is an open-source toolkit for training and fine-tuning large language models faster and with less memory than a standard Hugging Face stack. Its core library replaces PyTorch's default autograd with custom backpropagation kernels written in OpenAI's Triton language, which is where most of its speed and memory savings come from. It supports LoRA, QLoRA, full fine-tuning, reinforcement learning, pretraining, and 4-bit, 16-bit, and FP8 training, across more than 500 text, vision, audio, and embedding models.

The practical draw is hardware reach. QLoRA workflows in Unsloth let you fine-tune an 8B model on a single 12 GB consumer GPU, and the project headlines roughly 2x faster training with about 70 percent less VRAM versus baseline implementations, though the exact figures vary by model, GPU, and config. A 2026 update added faster mixture-of-experts training, with models like Qwen3-30B-A3B fine-tunable on about 17.5 GB of VRAM. It runs on NVIDIA (including Blackwell and DGX Spark), AMD, and Intel GPUs, with free Colab and Kaggle notebooks for trying it without local hardware.

It fits cleanly into the local-AI workflow. Unsloth integrates with Hugging Face transformers and TRL, and uses llama.cpp to save and run models, exporting to GGUF for Ollama or LM Studio as well as safetensors. As of 2026 it also ships Unsloth Studio, a local no-code GUI that covers the full lifecycle from dataset creation to training to running and comparing GGUF and safetensors models, with tool-calling, web search, and an OpenAI-compatible API, all running offline on Mac and Windows, with the core library under the Apache 2.0 license.