Structured lessons, deep-dive topic hubs, platform and model references, and daily AI and machine learning news written in plain English. From first classifier to research frontier.https://www.aimlfieldguide.com/en-gbhttps://www.aimlfieldguide.com/learn/beginner/01-what-is-machine-learning/https://www.aimlfieldguide.com/learn/beginner/01-what-is-machine-learning/Programs whose rules are fitted to data rather than written by hand, the train and predict loop, the vocabulary you will use for the rest of the track, and when machine learning is the wrong toolFri, 13 Nov 2026 00:00:00 GMThttps://www.aimlfieldguide.com/learn/beginner/02-data-is-the-model/https://www.aimlfieldguide.com/learn/beginner/02-data-is-the-model/Build a small synthetic dataset with pandas, split it into train, validation and test sets, then plant a leaky feature and watch the validation score lieFri, 13 Nov 2026 00:00:00 GMThttps://www.aimlfieldguide.com/learn/beginner/03-linear-models-from-scratch/https://www.aimlfieldguide.com/learn/beginner/03-linear-models-from-scratch/Least squares solved in closed form with numpy, then logistic regression fitted by gradient descent with the sigmoid and log loss, both checked against scikit-learnFri, 13 Nov 2026 00:00:00 GMThttps://www.aimlfieldguide.com/learn/beginner/04-training-loop-and-loss/https://www.aimlfieldguide.com/learn/beginner/04-training-loop-and-loss/The four-step loop that trains almost every model, the gradient checked by hand and by finite differences, and a learning-rate sweep that shows slow, right and divergent on one chartFri, 13 Nov 2026 00:00:00 GMThttps://www.aimlfieldguide.com/learn/beginner/05-trees-and-gradient-boosting/https://www.aimlfieldguide.com/learn/beginner/05-trees-and-gradient-boosting/How a decision tree chooses its splits, what depth does to training and held-out accuracy, how gradient boosting stacks small trees into a strong model, and why feature importance needs careFri, 13 Nov 2026 00:00:00 GMThttps://www.aimlfieldguide.com/learn/beginner/06-evaluation-that-does-not-lie/https://www.aimlfieldguide.com/learn/beginner/06-evaluation-that-does-not-lie/Why accuracy misleads on imbalanced data, what precision, recall, F1 and ROC-AUC measure, how stratified cross-validation and a dummy baseline keep you honest, and how a bootstrap puts an interval on a scoreFri, 13 Nov 2026 00:00:00 GMThttps://www.aimlfieldguide.com/learn/beginner/07-your-first-neural-network/https://www.aimlfieldguide.com/learn/beginner/07-your-first-neural-network/Tensors, autograd, nn.Module and the training loop, applied to a small multilayer perceptron that reads 8x8 digit images on a laptop CPU in a few secondsFri, 13 Nov 2026 00:00:00 GMThttps://www.aimlfieldguide.com/learn/beginner/08-embeddings-and-similarity/https://www.aimlfieldguide.com/learn/beginner/08-embeddings-and-similarity/What an embedding vector is, cosine similarity worked by hand, a 22 M parameter sentence encoder run locally on twenty sentences, nearest-neighbour search with numpy, and why it beats a keyword index on meaningFri, 13 Nov 2026 00:00:00 GMThttps://www.aimlfieldguide.com/learn/beginner/09-using-an-llm-well/https://www.aimlfieldguide.com/learn/beginner/09-using-an-llm-well/What tokens and a context window are, how system and user messages differ, how to ask for JSON and validate it, few-shot examples, and the limits every model shares, with code for a local model and a hosted APIFri, 13 Nov 2026 00:00:00 GMThttps://www.aimlfieldguide.com/learn/beginner/10-from-notebook-to-service/https://www.aimlfieldguide.com/learn/beginner/10-from-notebook-to-service/Train a small scikit-learn pipeline, save it with joblib, wrap it in a FastAPI endpoint with validated inputs, test the endpoint in-process without a server, and lay the project out so someone else can run itFri, 13 Nov 2026 00:00:00 GMThttps://www.aimlfieldguide.com/learn/intermediate/01-backpropagation-by-hand/https://www.aimlfieldguide.com/learn/intermediate/01-backpropagation-by-hand/Work the chain rule through a two-layer network with real numbers, code the backward pass in numpy, check it with finite differences, then watch PyTorch autograd produce the same gradientsFri, 13 Nov 2026 00:00:00 GMThttps://www.aimlfieldguide.com/projects/mnist-cnn-pytorch/https://www.aimlfieldguide.com/projects/mnist-cnn-pytorch/Train a two-layer convolutional network on 16x16 digit images in under a second on a laptop CPU, read the loss curve, print a confusion matrix, and find out which digits it gets wrong.Fri, 13 Nov 2026 00:00:00 GMThttps://www.aimlfieldguide.com/projects/tabular-classifier-fastapi/https://www.aimlfieldguide.com/projects/tabular-classifier-fastapi/Train a scikit-learn pipeline on the breast cancer dataset, save it with joblib, serve it from a FastAPI app with /predict and /health, and test the app in-process before you ever start a server.Fri, 13 Nov 2026 00:00:00 GMThttps://www.aimlfieldguide.com/projects/lora-fine-tune-apple-silicon/https://www.aimlfieldguide.com/projects/lora-fine-tune-apple-silicon/Teach SmolLM2-135M a date-format conversion it cannot do out of the box, with a rank-8 LoRA adapter trained for 60 steps on MPS or CPU, then save, reload and verify the adapter.Fri, 13 Nov 2026 00:00:00 GMThttps://www.aimlfieldguide.com/projects/tiny-rag-local-embeddings/https://www.aimlfieldguide.com/projects/tiny-rag-local-embeddings/Chunk a folder of notes, embed the chunks with all-MiniLM-L6-v2 on your CPU, retrieve with a cosine index in plain numpy, then hand the top chunks to a local model through Ollama for a cited answer.Fri, 13 Nov 2026 00:00:00 GMThttps://www.aimlfieldguide.com/news/2026-11-12-vllm-v0-33-0-ships-with-927-commits-as-v0-34-0-release/https://www.aimlfieldguide.com/news/2026-11-12-vllm-v0-33-0-ships-with-927-commits-as-v0-34-0-release/vLLM tagged v0.33.0 on 2 November 2026 with shorter Docker image names, then published two v0.34.0 release candidates within ten days.Thu, 12 Nov 2026 00:00:00 GMThttps://www.aimlfieldguide.com/news/2026-11-10-anthropic-ipo-roadshow/https://www.aimlfieldguide.com/news/2026-11-10-anthropic-ipo-roadshow/The proposed listing under the ticker ANT follows the S-1 filed on 26 October. Anthropic itself is selling 801 million shares, with no proceeds, pricing date or closing date yet stated.Tue, 10 Nov 2026 00:00:00 GMThttps://www.aimlfieldguide.com/news/2026-11-10-hugging-face-adds-a-decision-pipeline-for-models-that-score/https://www.aimlfieldguide.com/news/2026-11-10-hugging-face-adds-a-decision-pipeline-for-models-that-score/Transformers gains a "decision" task that returns a probability over predefined answers, as a new arXiv paper asks how such models fit into long-horizon agents.Tue, 10 Nov 2026 00:00:00 GMThttps://www.aimlfieldguide.com/research/2026-11-10/https://www.aimlfieldguide.com/research/2026-11-10/Eight arXiv papers on LoRA capacity, edge weight compression, RAG citation provenance, prompt-injection defences, data-free model merging, output formats for small parsers, SFT versus RL, and CUDA kernels.Tue, 10 Nov 2026 00:00:00 GMThttps://www.aimlfieldguide.com/news/2026-11-09-anthropic-claude-fable-5-2-mythos-5-2/https://www.aimlfieldguide.com/news/2026-11-09-anthropic-claude-fable-5-2-mythos-5-2/The two 5.2 models share one underlying network. Anthropic reports faster output, a 2.5 times fast mode at double the price, and about 55 per cent fewer safeguard blocks than Fable 5.1.Mon, 09 Nov 2026 00:00:00 GMThttps://www.aimlfieldguide.com/news/2026-11-09-google-tpu-8tx-and-8i-max/https://www.aimlfieldguide.com/news/2026-11-09-google-tpu-8tx-and-8i-max/TPU 8tx pairs TPU 8t silicon with AMD EPYC hosts for confidential inference and small-scale training. TPU 8i Max packs two 8i chips into one package. No prices or dates yet.Mon, 09 Nov 2026 00:00:00 GMThttps://www.aimlfieldguide.com/news/2026-11-09-ollama-0-40-16-ripple-preview/https://www.aimlfieldguide.com/news/2026-11-09-ollama-0-40-16-ripple-preview/Three Ollama releases in a week add Ricursive's 65 B dense model, MiMo V2.6, YaRN long context for Qwen3, and stop OpenAI-compatible endpoints overriding a model's sampler defaults.Mon, 09 Nov 2026 00:00:00 GMThttps://www.aimlfieldguide.com/news/2026-11-05-moonshot-kimi-k3-1-pro-open-weights/https://www.aimlfieldguide.com/news/2026-11-05-moonshot-kimi-k3-1-pro-open-weights/The weights ship in MXFP4 under Moonshot's own licence, which adds a revenue threshold for hosting providers. Context is 1 M tokens and the card recommends vLLM, SGLang or TokenSpeed.Thu, 05 Nov 2026 00:00:00 GMThttps://www.aimlfieldguide.com/news/2026-11-04-llama-cpp-0-9-0-fp8-e4m3/https://www.aimlfieldguide.com/news/2026-11-04-llama-cpp-0-9-0-fp8-e4m3/The November release brings an FP8 file type, a more efficient unified KV cache for multi-sequence serving, four new architectures and ggml v0.29.0 with initial multi-node tensor parallelism.Wed, 04 Nov 2026 00:00:00 GMThttps://www.aimlfieldguide.com/news/2026-11-04-meta-acquires-conductor/https://www.aimlfieldguide.com/news/2026-11-04-meta-acquires-conductor/Terms were not disclosed. Conductor Cloud goes read-only on 13 November and shuts on 20 November, and the local app stops receiving updates in early 2027 while the team joins Meta.Wed, 04 Nov 2026 00:00:00 GMT