ThinkingBox Turns Database-State Checks into Agent Training Rewards
Microsoft and Hugging Face have released ThinkingBox-Training on GitHub, including the data-selection, rollout, reward, and Verl integration used in their Qwen3.8-27B experiment.
All posts
OpenAI Python SDK expands Decisions types
Live response models now allow question.instructions to contain structured text, image, or JSON parts as well as a string.
Amazon Polly upgrades its Generative engine
Amazon Polly says its upgraded Generative engine produces more natural, expressive speech and improves how it reads details such as numbers and addresses.
SageMaker AI adds customizable serverless training sessions
SageMaker AI serverless model customization training sessions are generally available, with teams writing their own post-training loops while AWS manages GPU allocation, distributed execution, recovery, and checkpoint storage.
Amazon SageMaker AI adds APIs for custom training loops
Amazon SageMaker AI has introduced training session APIs that let developers define their own training loop while the service operates the GPU infrastructure.
LangChain Core 1.7.0 changes streamed tool-call arguments
LangChain Core 1.7.0 sends streamed tool-call arguments as `tool-call-delta` slices instead of `block-delta` snapshots.
Streaming Kokoro TTS for AI agents on SageMaker AI
AWS demonstrates sending each completed agent sentence to Kokoro-82M on a SageMaker AI endpoint, so speech synthesis can begin before the full reply is ready.
Verify GPU-Direct Reads with NVIDIA KvikIO
NVIDIA GPUDirect Storage (GDS) moves data between supported storage and GPU memory without staging it through host memory, reducing CPU use.
AWS demonstrates measured autoscaling for SageMaker AI TTS
AWS deploys Kokoro-82M in a custom container on a SageMaker AI real-time endpoint and derives autoscaling settings from measurements on that endpoint.
NVIDIA Demonstrates Eight-Node DGX Spark Clusters
Eight DGX Spark nodes combine for 1 TB of unified memory, enough to load the full GLM 5.3 753B checkpoint locally.
OlmoEarth v1.3 releases global satellite embeddings and an explorer
OlmoEarth v1.3 arrives with precomputed annual satellite embeddings for the planet at 10-meter resolution, offered in 64- and 128-dimensional forms.
ForTy Maps Forest Types and Tracks Their Changes
Google DeepMind and Google Research released ForTy 2020 with WRI and IIASA, assigning forest and land-use classes to locations worldwide on a 10-meter grid.
IBM Granite adds a four-model time-series ensemble
Granite Timeseries Ensemble combines four pre-trained Granite architectures and returns median and quantile forecasts without retraining or fitting ensemble weights.
OpenAI Python SDK 3.36.1 fixes WebSocket and browser-call behavior
OpenAI Python SDK 3.36.1 now applies WebSocket query options supplied for an individual API call.
Connecting Amazon Quick to Corporate Networks with AWS PrivateLink
Amazon Quick uses separate network paths for user access and data retrieval: AWS PrivateLink VPC endpoints serve the first, while VPC connections handle the second.
Deep Insight: Taking Multi-Agent Systems to Production
Deep Insight is an open-source data analysis system that turns a CSV file and a natural-language request into analysis results and a cited DOCX report.
LangChain 1.4.7 adds streaming and tracing updates
LangChain 1.4.7 enables stream consumers to read tool-call-delta events.
Strands Agents demo for routing work between local and cloud models
AAIF uses a local llama3.2:1b model as a request classifier in a Strands Agents setup, directing work to local tools or Amazon Bedrock instead of sending every request to one cloud model.
Training Qwen3 on AWS Trainium with torchtitan
AWS shows Qwen3 training on one trn2.48xlarge using torchtitan and TorchNeuron, which exposes Trainium through PyTorch’s device interface.
LangChain Perplexity 1.4.2 fixes Agent API search handling
This release refreshes the model profile data.
Ollama v0.40.8 adds Recursive’s Flint model
Ollama v0.40.8 makes Recursive’s Flint coding model available to run through Ollama.
GitHub scales Git storage with per-reference quorum
Most GitHub repositories now keep two of their five replicas in Azure, which handles about half of Git traffic.
OpenAI Python SDK 3.36.0 adds Vertex AI support
Version 3.36.0 lets Python SDK users call the Responses API with Vertex AI.
NVIDIA Introduces Training-Free Relational Predictions
NVIDIA’s Kumo Relational runs predictive queries directly on relational databases, without task-specific training or manual feature engineering.
VocabX Cuts Inference Cost with Domain-Specific Vocabularies
Dharma AI tested VocabX on a Portuguese essay grader, keeping Qwen3-4B, the training data, and the training recipe while replacing its general-purpose vocabulary.
IBM Updates Its Granite Speech Model Lineup
IBM has added granite-speech-5.0-2b to Granite Speech and revised the granite-speech-5.0-470m-turboctc and granite-speech-4.1-2b-plus models.
NVIDIA Kumo Forecast Adds Historical Retrieval to Time-Series Prediction
Kumo Forecast builds temporal representations with the MOMENT-1-large backbone, then uses a forecasting head to map them to values for the requested prediction horizon.
NVIDIA Sentry adds hardware-isolated controls for AI agents
NVIDIA Sentry is part of the OpenShell reference architecture and uses BlueField-4 to monitor agents and enforce access policies outside the host software domain.
NVIDIA OpenShell Adds Policy Controls to AI Agent Workloads
NVIDIA describes OpenShell as an open-source runtime that limits agents’ access to systems and data, while treating agent workloads as untrusted during execution.
LiteLLM v1.103.7 updates usage fixes and image verification
The v1.103.7 Docker image can be checked with cosign using a public key linked to either a fixed commit or the release tag.