数据与 AI技能

Configure LLM inference and serving on NVIDIA GPUs with TensorRT-LLM. Covers quantization, which reduces numerical precision, batching requests during generation, and distributing models across GPUs to improve processing capacity and response times.

中文简介正在准备中。当前显示英语简介。

• Setting up an LLM server on NVIDIA GPUs
• Configuring FP8 or INT4 quantization
• Distributing large models across multiple GPUs

Build document search and AI retrieval workflows with Chroma, a database for numerical representations of text. Cover document storage, meaning-based search, metadata filters, persistence, and connections to LangChain and LlamaIndex.

中文简介正在准备中。当前显示英语简介。

• Building document search based on meaning
• Adding document retrieval to AI answers
• Filtering search results by document attributes

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。

• Comparing experiments with different training settings
• Viewing training metrics in a local dashboard
• Logging sample images and audio alongside metrics

Use Phoenix to record execution traces, detailed records of an AI application's operations, and investigate problems. Evaluate answers on datasets, compare prompts and models, and set up monitoring through a self-hosted server.

中文简介正在准备中。当前显示英语简介。

• Investigate AI application problems using execution traces
• Evaluate answer quality across a test dataset
• Compare prompts and models through experiments

Build search and recommendation features with Pinecone, a managed database for numerical representations of content. Covers data loading, combined meaning and keyword search, filtering by attributes, and separating data by user.

中文简介正在准备中。当前显示英语简介。

• Build document retrieval for AI answers
• Combine meaning and keyword search
• Separate search data by user

Guide knowledge distillation, where a smaller language model learns from a larger teacher. Covers training from teacher responses or prediction scores, adjusting training settings, and comparing outputs to support model compression and specialized training.

中文简介正在准备中。当前显示英语简介。

• Compressing a large language model into a smaller model
• Training a specialized model from teacher-generated examples
• Comparing distillation methods and training settings

Combine specialized language models without retraining using mergekit. The skill explains how to choose merging methods, configure each model’s contribution, compare resulting variants, and prepare a merged model for evaluation and deployment.

中文简介正在准备中。当前显示英语简介。

• Combine math, coding, and conversation capabilities
• Compare model variants with different merge weights
• Choose a method for merging several models

Guides LLM compression through pruning, which removes selected model weights. Covers Wanda and SparseGPT methods without retraining, accuracy comparisons, and structured weight patterns that can accelerate inference on compatible hardware.

中文简介正在准备中。当前显示英语简介。

• Trying model compression without retraining
• Comparing accuracy before and after pruning
• Choosing between Wanda and SparseGPT

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。

• Designing a model that activates only selected expert networks
• Balancing workloads across expert networks
• Distributing experts across multiple GPUs

Reduce LLM generation delays with speculative decoding, which proposes tokens ahead of verification. This skill compares small draft models, Medusa, and Lookahead Decoding, with implementation examples, tuning guidance, and deployment using vLLM.

中文简介正在准备中。当前显示英语简介。

• Reducing response delays in chatbots
• Speeding up code generation
• Choosing a small model to draft candidate tokens

Create masks that identify object regions in images using points, boxes, or existing masks. The skill guides automatic mask generation, transparent-background extraction, annotation workflows, and deployment, without requiring training for each specific task.

中文简介正在准备中。当前显示英语简介。

• Extracting objects with transparent backgrounds
• Building annotation tools with point and box selection
• Generating object masks automatically

Guides fine-tuning and evaluation of OpenVLA-OFT robot models that turn images and language into actions. Covers LIBERO simulation, ALOHA deployment, LoRA adaptation that updates a small set of parameters, and troubleshooting action scaling or configuration mismatches.

中文简介正在准备中。当前显示英语简介。

• Evaluate pretrained models in LIBERO
• Fine-tune robot action models with LoRA
• Deploy server-client inference for ALOHA

Guides adapting OpenPI robot models to custom datasets and serving them to produce actions from observations. Covers JAX and PyTorch training, checkpoint conversion, and troubleshooting for normalization statistics and GPU memory.

中文简介正在准备中。当前显示英语简介。

• Fine-tune OpenPI models on custom robot datasets
• Convert JAX checkpoints to PyTorch
• Serve policies for ALOHA, DROID, or LIBERO

Build LLM serving workflows that produce JSON or other required formats. The skill covers reusing computation for shared prompts, tool calls, and ongoing conversations, with examples for server setup and structured output.

中文简介正在准备中。当前显示英语简介。

• Extracting article details into JSON
• Building agents that call tools
• Reusing shared prompts across conversations

Guides image classification and search with OpenAI's CLIP by comparing images with text descriptions, without additional training. Includes batch processing, image-text matching, and category-based checks for potentially inappropriate images.

中文简介正在准备中。当前显示英语简介。

• Classify images using text labels without additional training
• Search an image collection with natural-language descriptions
• Compare how closely images match text descriptions

Build image captioning, visual question answering, and text-based image search with BLIP-2. The skill guides model selection, batch processing, and memory optimization, with examples for understanding images without training for each specific task.

中文简介正在准备中。当前显示英语简介。

• Generate captions for multiple images
• Build a question-answering system for images
• Search an image collection using text

Build searches with Qdrant that find similar content and filter results by attributes. The skill covers document retrieval for AI answers, vector storage, memory optimization, and distributed deployment for larger collections.

中文简介正在准备中。当前显示英语简介。

• Building document retrieval for AI answers
• Combining similarity search with category or date filters
• Reducing memory use for large vector collections

Helps extend how much text a language model can process at once. Covers choosing RoPE, YaRN, ALiBi, or position interpolation, implementing position information, fine-tuning on longer documents, and managing memory during inference.

中文简介正在准备中。当前显示英语简介。

• Extend an existing model to handle longer documents
• Compare context extension methods
• Implement position information in attention layers

Evaluate NVIDIA Cosmos Policy for robot manipulation in LIBERO and RoboCasa simulations. Follow GPU environment setup, configure rendering without a display, run small checks before larger benchmarks, inspect success rates, and profile inference latency.

中文简介正在准备中。当前显示英语简介。

• Running a small LIBERO evaluation
• Evaluating kitchen tasks in RoboCasa
• Setting up simulation on a GPU cluster without a display

Guide language model training with SimPO using pairs of preferred and rejected answers. Covers configuration, learning rate tuning, and memory troubleshooting for preference alignment, which adjusts model responses toward human preferences without a separate reference model.

中文简介正在准备中。当前显示英语简介。

• Training on preferred and rejected answer pairs
• Exploring an alternative to DPO without a reference model
• Tuning preference training for math and code tasks

Implement Mamba, a model architecture that processes sequences through an internal state with computation scaling linearly with input length. Covers pretrained text generation, Mamba-1 and Mamba-2 configuration, speed comparisons with Transformers, and GPU memory troubleshooting.

中文简介正在准备中。当前显示英语简介。

• Explore models for long sequences or streaming applications
• Generate text with pretrained Mamba models
• Compare Mamba-1 and Mamba-2 configurations

Learn how GPT language models work by preparing text, training a small model, and generating samples. Covers CPU experiments, custom datasets, further training of pretrained GPT-2 models, and reproducing GPT-2 with multiple GPUs.

中文简介正在准备中。当前显示英语简介。

• Train a small model on Shakespeare using a CPU
• Prepare custom text for model training
• Further train a pretrained GPT-2 model

Guides you through setting up RWKV, generating text, processing long documents in chunks, and fine-tuning models. Explains how to carry context between steps and compare memory use with Transformer models.

中文简介正在准备中。当前显示英语简介。

• Build streaming text generation
• Process long documents in chunks
• Plan RWKV fine-tuning

Build safety checks for LLM prompts and responses using Meta's LlamaGuard. This skill covers content classification, batch moderation, serving a moderation API, and connecting checks to NeMo Guardrails.

中文简介正在准备中。当前显示英语简介。

• Screening user prompts before sending them to an LLM
• Checking generated responses before displaying them
• Building an API for content safety checks