Loading...
Loading...
Found 15 Skills
Fine-tune Gemma 4 and 3n models with audio, images, and text on Apple Silicon using PyTorch and Metal Performance Shaders.
Write, debug, and optimize CUTLASS and CuTeDSL GPU kernels using local source code, examples, and header references. Use when the user mentions CUTLASS, CuTe, CuTeDSL, cute::Layout, cute::Tensor, TiledMMA, TiledCopy, CollectiveMainloop, CollectiveEpilogue, GEMM kernel, grouped GEMM, sparse GEMM, flash attention CUTLASS, blackwell GEMM, hopper GEMM, FP8 GEMM, blockwise scaling, MoE GEMM, StreamK, warp specialization CUTLASS, TMA CUTLASS, or asks about writing high-performance CUDA kernels with CUTLASS/CuTe templates.
Trigger this skill when building applications with Gemma or for general knowledge inquiries related to Gemma models (e.g. prompt structure, capabilities). Covers model selection, development workflows, and deployment best practices.
Write and implement GPU kernels using NVIDIA CuTe DSL (CUTLASS 4.x Python API) — NOT for Triton, CUDA C++, or conceptual explanations. Trigger only when the user wants to write or implement a kernel, not when asking questions about CuTe DSL concepts or layouts. CuTe DSL uses cute.jit/cute.kernel decorators and cutlass.cute imports. Covers element-wise kernels, GEMM patterns, reductions, memory hierarchy (global/shared/register/TMA), MMA tensor core operations, software pipelining, and framework integration.
Generates and updates secure, production-ready Kubernetes YAML manifests optimized for GKE Autopilot and GKE Standard clusters. Use when creating or modifying GKE deployment manifests, configuring container security contexts, setting CPU/memory resource limits, defining readiness/liveness/startup probes, mounting secrets and volumes, configuring GKE Gateway API routes, targeting Spot VMs, or deploying AI model inference workloads (vLLM, TGI, Gemma). Don't use for live cluster operations, pod troubleshooting (use gke-workload-troubleshooting), or cluster infrastructure provisioning (use gke-cluster-creation).
Run LLMs and AI models on Cloudflare's GPU network with Workers AI. Includes Llama 4, Gemma 3, Mistral 3.1, Flux images, BGE embeddings, streaming, and AI Gateway. Handles 2025 breaking changes. Prevents 7 documented errors. Use when: implementing LLM inference, images, RAG, or troubleshooting AI_ERROR, rate limits, max_tokens, BGE pooling, context window, neuron billing, Miniflare AI binding, NSFW filter, num_steps.
Implements and trains LLMs using Lightning AI's LitGPT with 20+ pretrained architectures (Llama, Gemma, Phi, Qwen, Mistral). Use when need clean model implementations, educational understanding of architectures, or production fine-tuning with LoRA/QLoRA. Single-file implementations, no abstraction layers.
Génère des trésors BFRPG par type (A-U). Pièces, gemmes, bijoux et objets magiques. Utilise les tables officielles avec probabilités. Indispensable après un combat victorieux.
On-device, real-time multimodal AI voice and vision assistant powered by Gemma 4 E2B and Kokoro TTS, running entirely locally via FastAPI WebSocket server.
PokeClaw (PocketClaw) — on-device Android AI phone agent using Gemma 4 via LiteRT-LM with tool calling, accessibility automation, and optional cloud models.
Build and extend Gemma Gem, an on-device AI browser assistant Chrome extension running Google's Gemma 4 model via WebGPU with no cloud dependencies.
ONLY for OpenAI Triton (@triton.jit) kernel development. NEVER use for CUDA C++ kernels, TileIR, or profiling tools (ncu, nsys). The user's request must involve Triton explicitly. Covers Triton-specific patterns: fused elementwise, reductions (softmax, LayerNorm, RMSNorm), tiled GEMM with triton.autotune, and flash attention. Workflow: design, write, verify (with fast-path for explicit requests).