Early wildfire detection is of the utmost importance to enable rapid response efforts, and thus minimize the negative impacts of wildfire spreads. To this end, we present \Pyro, a web-scraping-based dataset composed of videos of wildfires from a network of cameras that were enhanced with manual bounding-box-level annotations. Our dataset was filtered based on a strategy to improve the quality and diversity of the data, reducing the final data to a set of 10,000 images. We ran experiments using a state-of-the-art object detection model and found out that the proposed dataset is challenging and its use in concordance with other public dataset helps to reach higher results overall. We will make our code and data publicly available.