Applied AI Guide

vLLM vs Ollama: Production Serving Throughput and VRAM Benchmarks on Linux

A rigorous performance test evaluating request concurrency, token throughput, and memory offload between vLLM and Ollama on Linux workstations.

How to Quantize Hugging Face Models to GGUF Using llama.cpp on Ubuntu

Convert raw Hugging Face transformer checkpoints into memory-efficient GGUF files ready for zero-latency execution in Ollama and LM Studio.

How to Run DeepSeek-R1 and Llama 3 Locally on Linux with Ollama

A complete, step-by-step tutorial on deploying open-weight models locally on Ubuntu using Ollama, including CUDA setup and systemd configuration.

Explore by topic

Cluster sections

Devops

Fixing vLLM Tool-Call Accuracy on Linux

Tool-call accuracy in vLLM often degrades due to mismatched sampling parameters and parser configurations. This guide walks through a Linux-based fix using guided decoding and prompt tuning.

Anthropic S-1 Filing: Build a Local SEC EDGAR Parser

The Anthropic S-1 filing is the top Hacker News story with 973 points, making it a practical dataset for building a local SEC EDGAR parser. This guide shows how to fetch, validate, and extract structured text from the filing on Linux.

Linux

Fixing vLLM Tool-Call Accuracy on Linux

Tool-call accuracy in vLLM often degrades due to mismatched sampling parameters and parser configurations. This guide walks through a Linux-based fix using guided decoding and prompt tuning.

Anthropic S-1 Filing: Build a Local SEC EDGAR Parser

Llm

Fixing vLLM Tool-Call Accuracy on Linux

Tool-call accuracy in vLLM often degrades due to mismatched sampling parameters and parser configurations. This guide walks through a Linux-based fix using guided decoding and prompt tuning.

Vllm

Fixing vLLM Tool-Call Accuracy on Linux

Tool-call accuracy in vLLM often degrades due to mismatched sampling parameters and parser configurations. This guide walks through a Linux-based fix using guided decoding and prompt tuning.

Latest in Applied AI

View all articles

vLLM vs Ollama: Production Serving Throughput and VRAM Benchmarks on Linux

A rigorous performance test evaluating request concurrency, token throughput, and memory offload between vLLM and Ollama on Linux workstations.

Fixing vLLM Tool-Call Accuracy on Linux

<p>Tool-call accuracy in vLLM often degrades due to mismatched sampling parameters and parser configurations. This guide walks through a Linux-based fix using guided decoding and prompt tuning.</p>

Verified 10 min

Anthropic S-1 Filing: Build a Local SEC EDGAR Parser

<p>The Anthropic S-1 filing is the top Hacker News story with 973 points, making it a practical dataset for building a local SEC EDGAR parser. This guide shows how to fetch, validate, and extract structured text from the filing on Linux.</p>