Search “AWS”

AWS security agent Continuum is now generally available, claims it fixes 89% of vulnerabilities on CyberGym-E2E

AWS on Thursday made its code vulnerability security agent Continuum generally available; the service can find and verify vulnerabilities in code repositories, check them against AWS environments, and submit pull requests containing remediation suggestions. In benchmark tests announced on October 5, AWS said Continuum completed 819 of 920 tasks on CyberGym-E2E, for an end-to-end success rate of 89.0%, compared with the previous publicly reported best of 65.9%; however, the proportion of patches that also fixed the historical vulnerabilities specified by the benchmark was only 37.8%, compared with the previous publicly reported best of 26.2%. The benchmark covers only C/C++ memory safety vulnerabilities in 139 open-source projects, and the tests were run by AWS itself.

Hugging Face launches a Space to check whether models can run on AWS Trainium

Hugging Face releases the torch-neuronx model support Space, which lets users view, by download count or trend, whether models on the Hub can run on AWS Trainium, and reports the number of sharded cores, compile-inclusive latency, and peak memory. Currently it only tests Transformers text-generation and image-text models, and only models that fit on 4 Trainium2 devices (up to 16 cores); each model generates 32 tokens for validation, comparing against CPU results in single-core mode, and in sharded mode a small model judges whether the English is fluent. Results come from the public dataset aws-neuron/torch-neuronx-model-support, and the Space also provides run scripts and a NEFF cache command that skips compilation.

Amazon releases TorchNeuron: a PyTorch backend for AWS Trainium

This week, Amazon announced the open-sourcing of TorchNeuron, a backend that converts PyTorch code into operations executable on AWS Trainium; PyTorch code already running on other accelerators can be migrated with almost no changes. It offers an optional adaptive eager mode that groups operations for compilation and caches NEFF at multiple levels, cutting the time to the first training step by about one-third in cross-accelerator large language model tests. The team also used an AI agent to scan Hugging Face repositories, selecting 841 representative models and tens of thousands of tensor shapes to build a tiered testing process.

AWS launches training sessions GA for SageMaker AI serverless model customization

AWS announced that training sessions for SageMaker AI serverless model customization are generally available; users can write and orchestrate their own training loops, including custom loss functions and reward signals, while SageMaker AI handles GPU allocation, distributed execution, fault recovery, and checkpoint storage. The feature supports SFT, DPO, RLVR, RLAIF, and multi-turn RL, with a choice of LoRA or full fine-tuning, and currently supports gpt-oss-20b, gemma-4-31B-it, Qwen3.6-27B, and Qwen3.8-27B. Billing is per token processed by SageMaker AI, billing stops when the reward function scores in the user's environment, and no reserved capacity is required.

LMSYS Releases SGLang AWS Trainium Plugin Beta

LMSYS, in collaboration with AWS, has released the SGLang Trainium plugin Beta, built on the newly released TorchNeuron, supporting LLM inference serving and RL rollout generation on Trainium2 and Trainium3. Initial coverage includes Qwen3 (dense), Qwen3-MoE-30B-A3B, DeepSeek-V2-Lite, GLM-4, and GLM-5.2/5.3, and supports data parallelism, tensor parallelism, expert parallelism, Radix Cache, BF16/FP8, and asynchronous scheduling. In the first end-to-end heterogeneous RL run, SGLang generated rollouts for Qwen3-MoE-30B-A3B on Trainium3, while Miles trained a GRPO policy with Megatron on a single H200 node. The Beta is currently limited to a small number of teams.

English translation is being prepared.

English translation is being prepared.

Hugging Face launches TorchNeuron, enabling HF models to run natively on AWS Trainium

Hugging Face has released the PyTorch plugin TorchNeuron, which turns Neuron into a native device, supporting eager mode, torch.compile, and distributed interfaces such as FSDP2 and DTensor. It is available as a public beta starting October 20, 2026, and can be installed via pip install torch-neuronx. Transformers, Diffusers, Accelerate, PEFT, and TRL can run directly on Trainium without going through Optimum Neuron; the official recommendation is for Neuron users to switch to these libraries. For text generation, 89% of the top 100 most commonly used architectures are supported out of the box, compared with 46% for Optimum Neuron v0.4.5, and all that is needed is to pass device_map=neuron.

English translation is being prepared.

English translation is being prepared.

English translation is being prepared.

English translation is being prepared.

AWS open-sources the decision model Strands Decider 2B, which can run locally

AWS has released the open-source decision model Strands Decider 2B, built on Qwen3.5-2B. Instead of generating text, it selects from preset options with a confidence score; it is fast, low-cost, and small enough to run locally. The model was developed by AWS Distinguished Engineer Marc Brooker on his own after seeing TypeSafe's Jev, briefly topped the Jevbench ranking for models of the same scale, and was subsequently packaged and released by Strands Labs. Brooker said that AWS customers' agent workflows do not always need the full capability and cost of an LLM, and that this type of model can serve as a decision step within workflows, delivering more reliable performance and lower latency thanks to its confidence scores and closed answer domain.

English translation is being prepared.

English translation is being prepared.

English translation is being prepared.

English translation is being prepared.

English translation is being prepared.

English translation is being prepared.

English translation is being prepared.