'workload' 검색 결과

Configure machine learning training and batch jobs across multiple clouds with SkyPilot. The skill guides GPU selection, lower-cost cloud and region choices, and automatic recovery when discounted spot instances are interrupted.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Choose a lower-cost cloud for GPU training
• Run long training jobs with recovery after interruptions
• Configure distributed training across multiple machines

Configure large language model training across multiple NVIDIA GPUs with Megatron-Core. The skill helps choose how to split models and workloads, set training parameters, monitor performance, and address memory shortages or low GPU utilization.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Planning distributed training for a LLaMA-style model
• Distributing MoE expert networks across GPUs
• Tuning GPU utilization and training throughput

Build data processing pipelines for machine learning, from preprocessing to batch predictions. Uses Ray Data to handle datasets larger than memory, distribute work across CPUs and GPUs, and feed data into training frameworks.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Preprocessing datasets larger than memory
• Running batch predictions across multiple GPUs
• Preparing image and tabular data for training

Build Apache Spark pipelines for large datasets, including transformations, aggregations, and streaming analysis. Get PySpark or Scala code, partitioning strategies for dividing data across workers, and configuration advice to investigate slow jobs and resource usage.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Build a pipeline to transform and aggregate large datasets
• Investigate slow Spark jobs and uneven data distribution
• Tune memory settings and data partitions

Deploy SageMaker endpoints, the services that answer model requests, with automatic capacity scaling, CloudWatch alarms, and resource tags. Supports real-time and asynchronous inference, scaling down to zero instances, deployment checks, and resource cleanup.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Deploying a real-time model endpoint
• Scaling a sparse workload down to zero instances
• Running long inference jobs asynchronously

Assess subpoenas and document requests served on a company. Classify the request, review scope, workload, and legal confidentiality protections, then prepare an objections outline, a compliance plan, and a deadline calendar for counsel review.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Assessing a subpoena received by a company
• Preparing an objections outline for counsel
• Planning document searches and production

Design or extend DataFusion benchmarks using SQL where practical. Select representative workload variations, keep unrelated processing lightweight, and integrate runs with bench.sh so measurements reflect useful execution paths and remain practical to run.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Add SQL benchmarks for DataFusion operators
• Select representative input sizes and filter conditions
• Reduce unrelated processing overhead in measurements

Create 15–30 second vertical pet product videos, from reference video analysis to scripts, shot plans, images, subtitles, and audio. The workflow combines materials through the COZE API and supports revisions based on feedback.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Create a short vertical video reviewing cat food
• Analyze reference pet videos and save creative patterns
• Prepare storyboard images, subtitles, and voiceover materials

Set up an OpenRig team for ongoing repository work, with clear ownership, implementation, independent review, and handoffs. Choose direct coordination, a task queue, or an explicit workflow, then expand the team as authorized work requires.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Starting a builder and reviewer team for a repository
• Keeping ownership visible across task handoffs
• Choosing direct work, a queue, or a workflow

Verify saved Excel files directly with Python and openpyxl, a workbook library. Check sheets, headers, row counts, marked rows, and output files to resolve conflicting agent reports and summarize what actually exists on disk.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Resolve conflicting workbook reports from agents
• Check that spreadsheet edits were saved
• Count rows selected with a marker column

Build subtle animated petals, snow, confetti, or other decorative particles within a single page section. Adjust their movement and density while preserving readable content, limiting device workload, and providing static alternatives for readers who reduce motion.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Adding gentle petals or snow to a page section
• Animating small brand fragments behind content
• Limiting decorative animation on mobile devices

Measure and improve Three.js or WebGL game performance through repeatable gameplay checks. Identify CPU or GPU bottlenecks, reduce unnecessary rendering and processing, and verify that controls, collisions, visuals, and memory remain stable.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Investigating frame-time spikes
• Reducing rendering and particle workloads
• Improving performance on mobile devices

Compare computing devices for a defined AI model workload using memory capacity, computing speed, data transfer, energy use, and cost. Build a comparison that separates predicted performance from measurements and identifies the constraints that determine the choice.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Choosing hardware for a specific AI model
• Checking whether a model fits in device memory
• Explaining gaps between advertised and measured performance

Estimate how model architecture, precision, and context length affect weight storage, generation state, computation, and memory traffic. Compare models and attention designs with separate resource estimates for input processing and token generation.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Compare memory needs across model architectures
• Estimate resource costs of longer contexts
• Compare attention designs and expert-based models

Estimate cross-server traffic, network bottlenecks, and synchronization waits for AI training or inference. Map communication onto physical network paths and assess how topology or communication changes affect the time needed to complete the job.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Find communication bottlenecks in distributed AI workloads
• Compare network topologies and communication paths
• Estimate synchronization waits under concurrent traffic

Plan how to split AI models and workloads across multiple GPUs. Compare memory needs, computation, communication, and waiting time to estimate completion times and choose a layout suited to response time, throughput, or cost goals.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Running models that exceed a single GPU's memory
• Comparing ways to split work across GPUs
• Estimating communication costs within and between servers

Design or review Apache Hudi tables through questions about data volume, updates, and usage. Get architecture recommendations, a decision record explaining tradeoffs, and configuration artifacts for supported processing paths.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Designing a new Hudi table
• Reviewing an existing table design
• Planning a small prototype

Run demanding scientific work on SSH compute hosts using direct execution or Slurm, a cluster job scheduler. Choose an enabled host, submit background jobs, check progress, collect outputs, and continue with analysis and artifact publication.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Running long GPU inference jobs on a remote host
• Submitting scientific workloads through Slurm
• Collecting and analyzing completed job outputs

Design ClickHouse tables, write analytical queries, and plan data loading using practical SQL and code patterns. Use it to review slow queries, build recurring summaries, and analyze user activity across large datasets.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Designing tables for large analytical datasets
• Investigating slow queries
• Building real-time summaries

Compare Zephyr changes using the same prepared dataset and matching execution settings. Analyze Finelog metrics for each processing stage to distinguish CPU efficiency, elapsed time, and peak memory changes, then report improvements, regressions, or tradeoffs.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Compare Zephyr performance before and after a PR
• Assess speed gains against additional CPU usage
• Check peak memory and out-of-memory regressions

Configure machine learning training and batch jobs across multiple clouds with SkyPilot. The skill guides GPU selection, lower-cost cloud and region choices, and automatic recovery when discounted spot instances are interrupted.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Choose a lower-cost cloud for GPU training
• Run long training jobs with recovery after interruptions
• Configure distributed training across multiple machines

Audit Node.js processes started by VCO, identify which are managed and still active, and report stale processes eligible for cleanup. Produce Markdown or JSON reports and simulate cleanup while keeping external and unknown processes outside the candidate set.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Auditing accumulated Node.js processes after long VCO runs
• Recording VCO-launched processes through a wrapper
• Producing Markdown or JSON process ownership reports

Choose an NVSHMEM transport, the communication path between compute nodes, from installed plugins, network evidence, and optional workload analysis. Receive a recommended configuration with minimal environment variables, compatibility evidence, alternatives, and clearly stated unknowns.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Choosing a transport compatible with a compute system
• Assessing communication needs in selected CUDA code
• Getting minimal transport selection settings

Choose software for serving language or vision-language models on NVIDIA Jetson using a current memory audit. Get specific launch flags for memory limits, context length, and concurrency to address memory shortages or compare lower-memory alternatives.

한국어 요약은 준비 중입니다. 영어 요약을 표시하고 있습니다.

• Choosing serving software for a Jetson memory budget
• Adjusting vLLM flags after out-of-memory errors
• Considering a switch from vLLM to llama.cpp