Search Results for "memory usage" - Page 27

Volumetric is a 3-D image processing tool. It reads point cloud data, expands it to a volumetric space with highly effective memory usage, and filter out necessary points.

Enhanced shell with multiple windows and added commands To install run: sh <(curl https://raw.githubusercontent.com/DennisHeine/UberShell/master/install_ubershell.sh) Custom commands added to the default shell (zsh): cscript <script>: executes a C/C++ script. example: cscript "printf("Hello World!");" noobhelp: Short desc of commands in /bin/ memusage: Top memory usage topdirs: Top directory space usage push/pop: Save current dir / return there serve: Opens a webserver at port 8084 and serves a file servedir: Opens a webserver at port 8000 and serves a dir servelog: Serves tail of syslog on port 12345 sizeof: List total sizes of subdirectorys largedirs: Lists directorys with 100+ MB disk usage mkdirc : Creates a dir and changes to it freefile : Kills all processes using a file lastedited: Shows which files have been edited findbin: Finds binary files not installed by dpkg frename : Renames parts of a filename connections: Shows network connections [...]

DiffusionGemma 26B A4B IT NVFP4 is NVIDIA’s Model Optimizer quantized release of Google DeepMind’s DiffusionGemma 26B A4B IT model. It is an open-weights multimodal generative model that processes text, images, and video inputs to produce text output through discrete diffusion. Built on the Gemma 4 26B A4B Mixture-of-Experts architecture, it has 25.2B total parameters and 3.8B active parameters, balancing capability with efficient inference. Its diffusion-based generation produces tokens in...

...The model supports a 1 million token context window, enabling it to maintain coherence across long workflows involving thousands of tool calls and multi-step reasoning chains. Architecturally, it uses a hybrid attention system combining Sliding Window Attention and Global Attention to significantly reduce memory usage while preserving long-context performance. It also integrates multi-token prediction modules that accelerate inference and improve reinforcement learning efficiency. Trained on around 27 trillion tokens with FP8 mixed precision and refined through supervised fine-tuning, large-scale agentic reinforcement learning, and distillation.

...EasyStream is a natural extension of Apache commons-io ( http://commons.apache.org/io/ ), providing advanced solutions to some common but not trivial problem. This library key points are performances, low memory footprint, reduced set of dependencies and simplicity of usage.

Mistral Large 3 675B Instruct 2512 NVFP4 is a frontier-scale multimodal Mixture-of-Experts model featuring 675B total parameters and 41B active parameters, trained from scratch on 3,000 H200 GPUs. This NVFP4 checkpoint is a post-training-activation quantized version of the original instruct model, created through a collaboration between Mistral AI, vLLM, and Red Hat using llm-compressor. It retains the same instruction-tuned behavior as the FP8 model, making it ideal for production...