Skill Creator

atrawog

41 published skills · 0 GitHub stars across the catalog. View on GitHub ↗

build

atrawog's avatarfrom atrawog

Development: Unified build system for OS images, pods, VMs, and ISOs.Run from repository root with 'just build <subcommand>'. Includes smartcache strategy that matches GitHub Actions for optimal build times.

clean

atrawog's avatarfrom atrawog

Development: Cleanup and maintenance for the development environment.Removes build artifacts, caches, containers, and recovers disk space.Run from repository root with 'just clean'. Use when developers needto free disk space or reset the build environment.

evaluation

atrawog's avatarfrom atrawog

LLM evaluation and prompt optimization with Evidently.ai. Covers textdescriptors, dataset metrics, LLM-as-a-Judge patterns, and automatedprompt optimization for classification and generation tasks.

finetuning

atrawog's avatarfrom atrawog

Model fine-tuning with PyTorch and HuggingFace Trainer. Covers datasetpreparation, tokenization, training loops, TrainingArguments, SFTTrainerfor instruction tuning, evaluation, and checkpoint management.

langchain

atrawog's avatarfrom atrawog

LangChain framework for LLM applications. Covers model wrappers (HuggingFace,Ollama), prompt templates, few-shot learning, output parsing, and chainingtechniques for building sophisticated LLM workflows.

peft

atrawog's avatarfrom atrawog

Parameter-efficient fine-tuning with LoRA and Unsloth. Covers LoraConfig,target module selection, QLoRA for 4-bit training, adapter merging, andUnsloth optimizations for 2x faster training.

quantization

atrawog's avatarfrom atrawog

Model quantization for efficient inference and training. Covers precisiontypes (FP32, FP16, BF16, INT8, INT4), BitsAndBytes configuration, memoryestimation, and performance tradeoffs.

rag

atrawog's avatarfrom atrawog

Retrieval-Augmented Generation (RAG) for grounding LLM responses withexternal knowledge. Covers document chunking, embeddings, vector stores(pandas, ChromaDB), similarity search, and conversational RAG pipelines.

transformers

atrawog's avatarfrom atrawog

Transformer architecture fundamentals. Covers self-attention mechanism,multi-head attention, feed-forward networks, layer normalization, andresidual connections. Essential concepts for understanding LLMs.

api

atrawog's avatarfrom atrawog

Direct REST API operations for Ollama using the requests library.Covers all /api/* endpoints for model management, text generation,chat completion, embeddings, and streaming responses.

huggingface

atrawog's avatarfrom atrawog

Import GGUF models from HuggingFace into Ollama. Pull models directlyusing the hf.co/ prefix, track download progress, and use importedmodels for inference.

openai

atrawog's avatarfrom atrawog

OpenAI compatibility layer for Ollama. Use the official OpenAI Pythonlibrary to interact with Ollama, enabling easy migration from OpenAIand compatibility with LangChain, LlamaIndex, and other OpenAI-based tools.

python

atrawog's avatarfrom atrawog

Official ollama Python library for LLM inference. Provides a clean,Pythonic interface for text generation, chat completion, embeddings,model management, and streaming responses.

comfyui

atrawog's avatarfrom atrawog

ComfyUI node-based Stable Diffusion interface. GPU-accelerated imagegeneration with custom node support and CivitAI model downloads.Use 'ujust comfyui' for configuration, lifecycle management, andmodel/node operations.

jellyfin

atrawog's avatarfrom atrawog

Jellyfin media server management via Podman Quadlet. Supports multi-instancedeployment, hardware transcoding (NVIDIA/AMD/Intel), and FUSE filesystemmounts. Use when users need to set up or manage Jellyfin media servers.

jupyter

atrawog's avatarfrom atrawog

JupyterLab ML/AI development environment management via Podman Quadlet.Supports multi-instance deployment, GPU acceleration (NVIDIA/AMD/Intel),token authentication, and per-instance configuration. Use when users needto configure, start, stop, or manage JupyterLab containers for ML development.

ollama

atrawog's avatarfrom atrawog

Ollama LLM inference server management via Podman Quadlet. Single-instancedesign with GPU acceleration for running local LLMs. Use when users needto configure Ollama, pull models, run inference, or manage the Ollama server.

runners

atrawog's avatarfrom atrawog

Self-hosted GitHub Actions runner management via Podman Quadlet. Supportsmulti-instance pools with ephemeral storage, automatic token generation,and rolling updates. Use when users need to set up CI/CD runners fortheir GitHub repositories.

apps

atrawog's avatarfrom atrawog

Third-party application installation for Bazzite. CoolerControl, DisplayLink,JetBrains Toolbox, OpenRazer, tablet drivers, scrcpy, and more. Use when usersneed to install hardware-specific or specialized applications.

audio

atrawog's avatarfrom atrawog

Audio configuration for Bazzite. Virtual audio channels for Game/Voice/Browser/Music,7.1 surround for headphones, Bluetooth headset profiles, and PipeWire management.Use when users need to configure audio on Bazzite.

boot

atrawog's avatarfrom atrawog

Boot configuration for Bazzite OS. BIOS/UEFI access, GRUB menu settings,secure boot key enrollment, and Windows dual-boot setup. Use when usersneed to configure boot options or access BIOS settings.

desktop

atrawog's avatarfrom atrawog

Desktop customization for Bazzite. GTK theme restoration, terminal transparency,and MOTD settings. Use when users need to customize their desktop appearance.

distrobox

atrawog's avatarfrom atrawog

Distrobox container management for Bazzite. Create containers from manifests,custom containers, app-specific containers (brew), and DaVinci Resolve installation.Use when users need to work with distrobox containers.

gaming

atrawog's avatarfrom atrawog

Gaming ecosystem for Bazzite. Steam fixes, Proton troubleshooting, EmuDeck emulation,Decky Loader plugins, Sunshine game streaming, frame generation, and media apps.Use when users need help with gaming on Bazzite.