Tag: ai

Run OpenAI Whisper or Faster-Whisper locally with Docker. Better privacy, zero API costs, and surprisingly good accuracy, even on a potato CPU.

Compare Continue.dev, Cody, and Tabby, three self-hosted AI code assistants that keep your code private, cost nothing per token, and work offline.

CUDA vs ROCm for AI on Linux: NVIDIA's easy path, AMD's emotional journey, and why CPU inference isn't dead yet. Real Docker setups included.

Flowise vs Langflow compared: self-hosted, Docker-ready visual LLM workflow builders. Build no-code AI pipelines, RAG chatbots, and more, without losing your mind.

Temperature, top-p, top-k, context length, LLM inference parameters explained so you stop guessing why the model gives weird output.

Write Stable Diffusion prompts that actually work, style keywords, negative prompts, weights, and the patterns that get consistent results.

Write prompts that get useful results, role prompting, few-shot examples, chain-of-thought, and the patterns that work across any LLM.

GGUF, GGML, AWQ, GPTQ, LLM file formats and quantization levels explained: trade-offs between model quality, size, and inference speed.

LLaMA, Mistral, Falcon, GPT, the LLM landscape is crowded. Compare model families, sizes, licensing, and what each is actually good for.

Ollama makes local LLMs easy to start. The only decision that matters is whether the model fits your VRAM. Compose file, real sizes, and the gotchas.

LocalAI serves chat, embeddings, images and audio behind one OpenAI-compatible endpoint. Auth is deny-by-default now, and old compose files pull dead models.

Supervised, unsupervised, and reinforcement learning without the math or the hype. What each kind of machine learning model does and where each one fits.