Tagged gpu

Pretrain large LLMs across many GPUs with torchtitan and 4D parallelism

Launch Lambda Labs GPU cloud instances for training and inference with SSH access

Run LLM inference on CPUs, Apple Silicon and consumer GPUs with llama.cpp

Run ML workloads and APIs on serverless GPUs with automatic scaling on Modal

Curate LLM training data on GPUs with deduplication, quality filters and PII redaction

Structure PyTorch training with Lightning Trainer, callbacks and automatic multi-GPU runs

Optimize LLM inference on NVIDIA GPUs with TensorRT-LLM for high throughput

Speed up Ebitengine drawing with batching, texture atlases and GPU command checks

Run MiniCPM5 GGUF models on CPU or consumer GPUs with llama.cpp

Run low-VRAM LoRA and QLoRA fine-tuning of MiniCPM5 on one consumer GPU with Unsloth

Run model evaluations locally with inspect-ai and lighteval on your own GPUs

Fine-tune language and vision models with TRL or Unsloth on cloud GPUs

Train object detection, image classification and segmentation models on cloud GPUs

Write and review ZeroGPU code for Gradio Spaces using the @spaces.GPU decorator

Run LiteRT .tflite models on Android with CPU and GPU inference in Kotlin

Test WPA2 passphrases with handshake and PMKID capture and GPU cracking

Render large GPU-accelerated 2D and 3D data views with Three.js and deck.gl

Build a cursor particle trail that emits by distance into a recycled GPU point pool

Build a live Three.js landscape with instanced GPU grass, slope-based ground and a sky

Use the 0G Compute Network for decentralized AI inference and GPU services

Profile game frame time and fix CPU or GPU bottlenecks with pooling and batching

Fine-tune LLMs on Google Colab GPUs with Unsloth via a remote bridge

Pick and run local, Slurm or Modal GPU compute for research jobs

Detect CPU, GPU, memory and disk resources before a heavy compute task

Operate GPU nodes and batch jobs with bounded spend and explicit approvals

Inspect and manage Lambda Cloud GPU instances over the API and SSH

Run approved CPU or GPU scripts on your Modal account with durable outputs

Inspect and operate RunPod GPU resources with runpodctl and approvals

Manage TensorPool GPU clusters and jobs with the tp CLI safely

Inspect and operate Vast.ai GPU instances with the vastai CLI and approvals

Speed up pandas-style DataFrame work on NVIDIA GPUs with cuDF

Check Kubernetes, GPU and storage prerequisites before installing CDF Orchestration

Find CUDA memory errors, data races and sync bugs with Compute Sanitizer

Set up CUDA-Q simulation targets and write quantum kernels in Python

Port Qiskit circuits into CUDA-Q kernels while keeping results validated

Write GPU data-loading code with the DALI imperative dynamic mode in Python

Write and load custom doca-bench plug-ins to measure new workload classes

Issue RDMA operations straight from GPU buffers with DOCA GPI objects

Wire CUDA kernels to BlueField and ConnectX queues with DOCA GPUNetIO

Measure sustained GPU-driven RDMA write bandwidth with GPUNetIO tools

Measure GPU-posted RDMA write latency and read median and tail results

Check or change NVIDIA display and GPU settings like refresh rate and V-Sync

Add optional GPU steps to genomics pipelines with NVIDIA Parabricks

Set up the Holoscan SDK 5.x CUDA 13 development container and verify it

Install Holoscan SDK 5.x Python wheels in an isolated CUDA 13 environment

Lock or cap Jetson CPU, GPU and memory clocks in the boot device tree

Snapshot Jetson health across memory, GPU, thermals, power, storage and processes

Set up the KERMT environment with Docker, GPU checks and a smoke test

Use selective or full activation recompute in Megatron Bridge to save GPU memory