1078 个 · 按相关度排序
Run training with mock or real data in Megatron-LM and Megatron Bridge, match their settings, and compare loss curves, which show training error over time. Includes recipes and multi-GPU examples for checking training behavior.
中文简介正在准备中。当前显示英语简介。
Manages AI model retention and machine unlearning requirements. Covers training data deletion verification, model versioning for compliance, machine unlearning techniques (SISA, gradient-based), and retraining triggers. Keywords: AI retention, machine unlearning, model versioning, training data deletion, retraining, storage limitation.
中文简介正在准备中。当前显示原文说明。
Use this skill when planning user adoption, structuring Salesforce training materials, drafting release communications, or running a change impact assessment for a Salesforce rollout or update. Triggers: user adoption plan, training materials, release announcement, change impact, go-live communication, communication plan, training plan by persona, adoption metrics, LoginHistory adoption report, PromptAction, training sandbox, go-live checklist, post-go-live feedback, super user program, pilot group. NOT for org deployment mechanics or sandbox promotion — use admin/change-management-and-deployment. NOT for adoption of an Agentforce or Einstein AI feature — use admin/ai-adoption-change-management. NOT for configuring the in-app prompts themselves — use admin/in-app-guidance-and-walkthroughs.
中文简介正在准备中。当前显示原文说明。
Synthesises training survey responses and manager notes into a DRAFT training needs report by role: coded needs with record counts per role (n of N), the evidence type behind each count, anonymised quotes cited to record codes, disagreements between what staff report and what managers observe, gaps where a role is under-represented or a required skill has no evidence, and follow-up questions to settle before any course is designed. Never rates an individual, ranks a team or turns a count into a competence verdict. Use when the user asks to "analyse this training survey", "what training do our teams need", "summarise the manager feedback on skills gaps", "build a training needs analysis by role" or "which roles have the biggest skill gaps". Do not use for turning an agreed need into a course, use course-outline-builder instead; for exit interviews, use exit-interview-synthesis. Drafts for human review; never approves, authorises or signs off.
中文简介正在准备中。当前显示原文说明。
指导使用Axolotl对大语言模型进行微调，涵盖YAML配置、数据格式和训练方法，也可用于检查多GPU设置、排查训练故障及运行缓慢的问题。
使用TRL指导语言模型进一步训练，使回答更符合指令或人的偏好。涵盖数据准备、训练方式选择、奖励模型训练与评估，以及内存不足和训练不稳定时的调整。
Track machine learning experiments in W&B, record training settings and metrics, and compare results in dashboards. Automate searches for training settings and keep version histories for models and datasets.
中文简介正在准备中。当前显示英语简介。
将PyTorch训练代码整理为LightningModule和Trainer结构，配置验证、日志记录、模型保存及多GPU分布式训练。
为项目管理应用的新客户制作入门培训HTML幻灯片，以柔和的粉彩设计讲解获得首次成果的步骤，以及有助于持续使用应用的习惯。
Turn graded execution traces and production logs into model training examples and preferred-versus-rejected response pairs. Select useful records, remove sensitive information, exclude evaluation examples, and preserve links to the original runs.
中文简介正在准备中。当前显示英语简介。
Prepare GRPO reinforcement learning configurations for tasks with programmatically checkable answers, such as math, code, and structured output. Design and inspect reward functions, then select training variants when runs show instability or exploit flaws in scoring.
中文简介正在准备中。当前显示英语简介。
Decide whether an LLM needs additional training or whether retrieval and better prompts will suffice. Match available examples, preferences, or verifiable results to a training method, then assess model size and memory requirements before starting.
中文简介正在准备中。当前显示英语简介。
Set up model training with torchforge for reinforcement learning, where rewards guide improvement. Covers training configurations, custom reward and loss functions, and distributed execution with Monarch and TorchTitan, keeping algorithm experiments separate from infrastructure concerns.
中文简介正在准备中。当前显示英语简介。
Explore training workflows that reduce harmful model responses using a written set of principles. The skill covers self-critique, revised training examples, and RLAIF, reinforcement learning from AI feedback, with implementation examples and hardware guidance.
中文简介正在准备中。当前显示英语简介。
Get guidance on adapting large language models with Axolotl. Covers YAML configuration files, LoRA and QLoRA methods for training fewer parameters, distributed training settings, dataset formats, and debugging training code.
中文简介正在准备中。当前显示英语简介。
Get guidance on adapting LLMs with Unsloth, including implementation, API usage, and debugging. Covers LoRA and QLoRA, methods for adjusting models with fewer trainable parameters, with a focus on training speed and memory use.
中文简介正在准备中。当前显示英语简介。
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.
中文简介正在准备中。当前显示英语简介。
Train smaller language models using knowledge distillation, where a larger model supplies examples or prediction targets. This skill covers training losses, generated training data, parameter choices, and output comparisons to support model compression and specialized student models.
中文简介正在准备中。当前显示英语简介。
Investigate slow TPU training on GKE by comparing diagnostic events with one-minute monitoring metrics. Separate memory or CPU bottlenecks from hardware and network issues, and identify the appropriate next steps.
中文简介正在准备中。当前显示英语简介。
Train sentence-transformers models for search, similarity, and reranking, which reorders search results by relevance. Choose a model type, prepare data, configure training, compare results against a baseline, and publish the trained model to Hugging Face Hub.
中文简介正在准备中。当前显示英语简介。
Track machine learning experiments with SwanLab by recording training settings, metrics, and media such as images and audio. Learn to compare runs, inspect training charts, and connect existing training code to local, cloud, or self-hosted dashboards.
中文简介正在准备中。当前显示英语简介。
Train Mixture of Experts models, which activate selected specialist networks for each input. The skill guides architecture, input routing, workload balancing, and GPU distribution using DeepSpeed or HuggingFace, with examples for training configuration and inference optimization.
中文简介正在准备中。当前显示英语简介。
Fine-tune language models with TRL using GRPO, a method that compares scored responses to the same prompt. Covers training data, scoring functions, GPU-conscious settings, and monitoring for tasks with measurable correctness or required output formats.
中文简介正在准备中。当前显示英语简介。
Explore training workflows that reduce harmful model responses using a written set of principles. The skill covers self-critique, revised training examples, and RLAIF, reinforcement learning from AI feedback, with implementation examples and hardware guidance.
中文简介正在准备中。当前显示英语简介。