GitHub
unsloth-models
Guidance on selecting and configuring supported model architectures like Llama 4, DeepSeek-R1, and Qwen3. Triggers: llama 4, deepseek-r1, qwen3, gemma 3, model selection, instruct vs base.
☆ 0View skill
Development AI Skills
Discover Development skills, compare source repositories and read the instructions for each workflow.
Categories: Development
Development Clear filters
Browse skills
GitHub
tool-calling
Define and run tool-calling patterns for LLMs (schema design, call loops, validation, parallel calls). Use when building function/tool calling workflows or debugging tool selection and arguments; triggers: tool...
☆ 0View skill
GitHub
torch-compile
Optimize PyTorch with torch.compile (TorchDynamo/Inductor), focusing on compile overhead, graph breaks, and benchmark methodology. Use when speeding up PyTorch models or debugging compile behavior; triggers: to...
☆ 0View skill
GitHub
unsloth-fft
Performing full fine-tuning (FFT) in Unsloth with 100% exact weight updates and optimized gradient checkpointing. Triggers include fft, full fine-tuning, full_finetuning, exact fine-tuning, and weight updates.
☆ 0View skill
GitHub
pytorch-onnx
Exporting PyTorch models to ONNX format for cross-platform deployment. Includes handling dynamic axes, graph optimization in ONNX Runtime, and INT8 model quantization. (onnx, onnxruntime, torch.onnx.export, dyn...
☆ 0View skill
GitHub
pytorch-geometric
Library for Graph Neural Networks (GNNs). Covers MessagePassing layers, modular aggregation schemes, and handling large graphs via mini-batching with disjoint graph representation. (pyg, messagepassing, gnn, gc...
☆ 0View skill
GitHub
numpy-core
Fundamental NumPy operations including ndarray creation, dtypes, shape manipulation, and basic operations with a focus on memory alignment and data views. Triggers: numpy, ndarray, dtype, reshape, memory alignm...
☆ 0View skill
GitHub
eval-frameworks
Evaluation framework patterns for RAG and LLMs, including faithfulness metrics, synthetic dataset generation, and LLM-as-a-judge patterns. Triggers: ragas, deepeval, llm-eval, faithfulness, hallucination-check,...
☆ 0View skill
GitHub
langchain-agents
Building LLM agents with LangChain and LangGraph, covering tool-calling model initialization, state management, and observability with LangSmith. Triggers: langchain, langgraph, langsmith, agent-executor, chat-...
☆ 0View skill
GitHub
unsloth-lora
Configuring and optimizing 16-bit Low-Rank Adaptation (LoRA) and Rank-Stabilized LoRA (rsLoRA) for efficient LLM fine-tuning using triggers like lora, qlora, rslora, rank selection, lora_alpha, lora_dropout, an...
☆ 0View skill
GitHub
unsloth-tts
Fine-tuning Text-to-Speech (TTS) models with Unsloth for voice cloning and synthetic speech (triggers: TTS, text-to-speech, voice cloning, Orpheus-TTS, audio fine-tuning, speech synthesis).
☆ 0View skill
GitHub
unsloth-long-context
Training models on extended context lengths using optimized RoPE scaling and memory-efficient attention kernels. Triggers: long context, max_seq_length, rope scaling, large context window, flex attention.
☆ 0View skill
GitHub
google-search
Integration patterns for web search grounding, including query operator usage, API-based search orchestration, and citation metadata mapping. Triggers: google-search, grounding, search-api, citations, search-op...
☆ 0View skill
GitHub
numpy-statistics
Standard and NaN-robust statistical functions for data analysis, histograms, and correlation matrices. Triggers: statistics, mean, nanmean, histogram, corrcoef, percentile, std.
☆ 0View skill
GitHub
unsloth-quantization
Utilizing Dynamic 4-bit quantization, FP8 training, and 8-bit optimizers to minimize VRAM usage without sacrificing accuracy. Triggers: quantization, dynamic 4-bit, fp8, bitsandbytes, adamw_8bit, qat.
☆ 0View skill
GitHub
unsloth-sft
Supervised fine-tuning using SFTTrainer, instruction formatting, and multi-turn dataset preparation with triggers like sft, instruction tuning, chat templates, sharegpt, alpaca, conversation_extension, and SFTT...
☆ 0View skill
GitHub
pytorch-lightning
High-level training framework for PyTorch that abstracts boilerplate while maintaining flexibility. Includes the Trainer, LightningModule, and support for multi-GPU scaling and reproducibility. (lightning, pyto...
☆ 0View skill
GitHub
numpy-linalg
Linear algebra operations in NumPy, including matrix multiplication, SVD, system solving, and least squares fitting. Triggers: linalg, matrix multiplication, SVD, eigenvalues, matrix decomposition, lstsq, multi...
☆ 0View skill
GitHub
file-system
Safe filesystem operations for agents, including path normalization vs resolution, temp file handling, atomic replacement, and spooled buffers. Use when reading/writing user-supplied paths, staging outputs, or...
☆ 0View skill
GitHub
numpy-io
File I/O operations including binary formats (npy/npz), text processing (csv), and memory-mapping for huge datasets. Triggers: io, load, save, npz, genfromtxt, memmap, loadtxt.
☆ 0View skill
GitHub
numpy-string-ops
Vectorized string manipulation using the char module and modern string alternatives, including cleaning and search operations. Triggers: string operations, numpy.char, text cleaning, substring search.
☆ 0View skill
GitHub
pytorch-distributed
Distributed training strategies including DistributedDataParallel (DDP) and Fully Sharded Data Parallel (FSDP). Covers multi-node setup, checkpointing, and process management using torchrun. (ddp, fsdp, distrib...
☆ 0View skill
GitHub
unsloth-vision
Fine-tuning multimodal vision-language models (Llama 3.2 Vision, Qwen2.5 VL) using optimized vision layers (triggers: vision models, multimodal, Llama 3.2 Vision, Qwen2.5 VL, UnslothVisionDataCollator, finetune...
☆ 0View skill
GitHub
unsloth-cpt
Strategies for continued pretraining and domain adaptation in Unsloth (triggers: continued pretraining, CPT, domain adaptation, lm_head, embed_tokens, rsLoRA, embedding_learning_rate).
☆ 0View skill
The default order uses GitHub repository stars from our indexed data. Stars belong to the source repository and may be shared by multiple skills. They indicate repository interest, not skill installs or an independent quality score.