bazzite-ai-plugins

Direct Preference Optimization for learning from preference pairs. Covers DPOTrainer, preference dataset preparation, implicit reward modeling, and beta tuning for stable preference learning without explicit reward models. Includes thinking quality patterns.

Model fine-tuning with PyTorch and HuggingFace Trainer. Covers dataset preparation, tokenization, training loops, TrainingArguments, SFTTrainer for instruction tuning, evaluation, and checkpoint management. Includes Unsloth recommendations.

Group Relative Policy Optimization for reinforcement learning from human feedback. Covers GRPOTrainer, reward function design, policy optimization, and KL divergence constraints for stable RLHF training. Includes thinking-aware reward patterns.

Fast inference with Unsloth and vLLM backend. Covers model loading, fast_generate(), thinking model output parsing, and memory management for efficient inference.

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

Advanced QLoRA experiments and comparisons. Covers alpha scaling, LoRA rank selection, target module strategies, continual learning, multi-adapter hot-swapping, and quantization comparison (4-bit vs BF16).

Model quantization for efficient inference and training. Covers precision types (FP32, FP16, BF16, INT8, INT4), BitsAndBytes configuration, memory estimation, and performance tradeoffs.

Reward model training for RLHF pipelines. Covers RewardTrainer, preference dataset preparation, sequence classification heads, and reward scaling for stable reinforcement learning. Includes thinking quality scoring patterns.

Reinforcement Learning with Leave-One-Out estimation for policy optimization. Covers RLOOTrainer, reward function integration, baseline estimation, and variance reduction techniques for stable RL training. Includes thinking-aware patterns.

Supervised Fine-Tuning with SFTTrainer and Unsloth. Covers dataset preparation, chat template formatting, training configuration, and Unsloth optimizations for 2x faster instruction tuning. Includes thinking model patterns.

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

Vision model fine-tuning with FastVisionModel. Covers Pixtral, Ministral VL training, UnslothVisionDataCollator, image+text datasets, and vision-specific LoRA configuration.

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

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

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

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 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.

GPU driver configuration for Bazzite. NVIDIA proprietary drivers, Optimus laptops, NVK (open-source NVIDIA), GPU switching, Broadcom WiFi, and Mesa testing builds. Use when users need to configure graphics drivers.

Network configuration for Bazzite. iwd WiFi backend, Wake-on-LAN, and Tailscale VPN. Use when users need to configure network services. For SSH, see bazzite-ai:config.

Security configuration for Bazzite. LUKS disk encryption with TPM auto-unlock, secure boot key management, and sudo password feedback. Use when users need to configure security features.

Storage management for Bazzite. Automounting drives (BTRFS/EXT4, Framework, SteamOS), BTRFS deduplication, rmlint disk trimming, and snapper snapshots. Use when users need to configure disk and partition management.

System maintenance for Bazzite OS. Updates via topgrade, cleanup of podman/flatpaks, viewing logs and changelogs, diagnostics, and power measurements. Use when users need to update, clean, or diagnose their Bazzite system.

GPU passthrough and virtualization for Bazzite. KVM/VFIO setup, Looking Glass (kvmfr), USB hotplug for VMs, and libvirt configuration. Use when users need GPU passthrough or advanced virtualization features.

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

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

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 need to free disk space or reset the build environment.

Overlay testing session management for bazzite-ai development. Enables live editing of justfiles via symlinks to /usr on immutable OS (OSTree) or traditional Linux systems. Use when users need to test ujust changes, enable overlay mode, troubleshoot testing sessions, or run VM/install tests.