Researchers at an unnamed security firm built a backdoored version of Qwen2.5-7B-Instruct that passed normal evaluations but exfiltrated project credentials through OpenAI's Codex CLI the moment a tri…
Musubi released PolicyLM-1.7B on October 6th, an open-weights moderation model under the Apache-2.0 license that takes a written policy and a message and returns category scores, allowing platforms to…
A developer benchmarked the same Gemma 4 E2B build served from an Amazon SageMaker endpoint and from a plain EC2 instance with an identical NVIDIA T4 or L4 GPU, finding decode speed and answers match …
A developer deployed Gemma 4 (E2B, E4B, 12B and 26B A4B) to Amazon SageMaker on a single NVIDIA T4 (ml.g4dn.xlarge), measuring decode throughput at 0.77x to 0.82x of an L4 while producing identical an…
A developer benchmarked Google's Gemma 4 E2B quantization-aware trained (QAT) checkpoint against the full bf16 release on a single NVIDIA L4 SageMaker endpoint, finding the 4-bit QAT model decodes at …
A developer published a step-by-step guide and open-source repository for deploying Google's Gemma 4 E2B model to an Amazon SageMaker real-time endpoint on a single NVIDIA L4 GPU using AWS's vLLM cont…
A developer benchmarked a Jev-style decision read of Gemma 4 on a single Google Cloud TPU v6e chip with vLLM, finding that the E2B, E4B and 12B checkpoints fit at bf16 while the 26B-A4B fp8 build also…
Jina AI released jina-ocr-v1, a 3.4B-parameter document parser with roughly 570M active decoder parameters per token, scoring 91.14 on OmniDocBench v1.6 and 83.4 on olmOCR-Bench at 2.57 pages per seco…
A study measuring 54 configurations of Qwen2.5-7B-Instruct on vLLM 0.12 across L4, A100, and H100 GPUs found that 18 of 36 configurations reach the cost, quality, and latency Pareto frontier, with com…
A developer published two Rust CLI demos, gemma-rust and gemma-rust-mcp, that query a self-hosted Gemma 4 E2B model through an OpenAI-compatible HTTP endpoint and through an MCP server, respectively. …
A developer published a step-by-step guide for deploying Google's 2B Gemma 4 model to Cloud Run with an NVIDIA L4 GPU, using vLLM for serving and a single-file Python MCP server to stage weights, depl…
TraceML, an open-source PyTorch training diagnosis tool, reproduced a known LeRobot training regression and found that recurring DataLoader fetches took over five seconds, while most fetches were quic…
Jina AI released Jina-OCR-v1, an end-to-end document parsing model combining DeepSeek-OCR's compressed-vision encoder and 3B mixture-of-experts decoder (activating ~570M parameters per token) with a F…
A developer's hand-written Gemma 4 port in pure JAX runs on both Turing and Ada NVIDIA GPUs with identical weights, but performance differs drastically due to hardware-specific compute dtype handling …
Hinode launched a cloud Linux desktop with a dedicated GPU that auto-pauses after 15 minutes of idle time, starting at $3.99 per hour. The service offers persistent state, one-click GPU boosting, and …
Fermion Research released Neutrino-1 8B, an 8.19-billion-parameter transformer that ships as a single 3.88 GB file using a proprietary ternary-family weight format eight times smaller than fp16, enabl…
Whissle Research Team introduced a meta-aware automatic speech recognition (ASR) model, META-1, that outputs transcription and metadata tokens (emotion, intent, demographics) in a single forward pass …
The article explains how generative Vision-Language Models (VLMs) can replace traditional object detectors like YOLO for industrial computer vision by enabling zero-shot detection through natural lang…