elaprendiz0000

Hace tiempo que no tocaba nada de IA en mi PC, y al echar un ojo descubrí varios programas que no recordaba para qué eran y, sobre todo, modelos de hace año y medio acumulando polvo en el disco. Lo que empezó como una limpieza acabó siendo una puesta al día completa: Ollama, conceptos fundamentales, mapa del ecosistema, comparativas, y un setup de Python desde cero. Lo dejo aquí ordenado por si a alguien le sirve. 1. El punto de partida: 42 GB de modelos zombi Lo primero que encontré fue Ollama corriendo en segundo plano (ese icono pequeño en la barra de tareas que llevaba meses sin tocar). Un ollama list reveló cinco modelos descargados hacía 18 meses, ocupando casi 42 GB: llama-3-8B-programming-questions — 16 GB SuperNova-Medius-Q8_0 — 15 GB uncensored_wizard_7b — 4.1 GB llama3.2 — 2 GB llama3.1 — 4.7 GB En el mundo de la IA, 18 meses son una eternidad. Han salido Llama 3.3 y 4, Qwen 2.5 y 3, DeepSeek-V3 y R1, Gemma 3 y 4, gpt-oss... cualquiera de los nu...

  Let’s explore the differences between the   n-gram approach   and the   neural approach   in   Large Language Models (LLMs) : N-gram Approach : Definition : N-gram models use statistical and probabilistic techniques to determine the probability of a given sequence of words occurring in a sentence. Basic Idea : An n-gram is a contiguous sequence of n items (usually words) from a given text sample. Assumption : The probability of the next word in a sequence depends only on a fixed-size window of previous words (context). Strengths : Simplicity : N-gram models are straightforward and easy to implement. Efficiency : They can handle large datasets efficiently. Limitations : Local Context : N-grams consider only local context, which may not capture long-range dependencies. Sparsity : As n increases, the number of possible n-grams grows exponentially, leading to data sparsity. Fixed Context Window : The fixed context window may not adapt well to varying sentence...

The regular expression [ˆa-zA-Z], which we used to avoid embedded instances of "the", implies that there must be some single (although non-alphabetic) character before the the. We can avoid this by specifying that before the the we require either the beginning-of-line or a non-alphabetic character, and the same at the end of the line:  grep -E "(^|[^a-zA-Z])[tT]he([^a-zA-Z]|^)" wizard_of_oz  The process we just went through was based on fixing two kinds of errors: false false positives positives, strings that we incorrectly matched like other or there, and false negafalse negatives tives, strings that we incorrectly missed, like The. Addressing these two kinds of errors comes up again and again in implementing speech and language processing systems. Reducing the overall error rate for an application thus involves two antagonistic efforts:  • Increasing precision (minimizing false positives)  • Increasing recall (minimizing false negatives) Some aliases for common...