Are there two elements in each fiber in the fiber bundle of the tangent vectors of a circle, or infinitely many?

Question

I'm watching this video (Frederic Schuller) and, at timestamp 9:50 have become confused about fiber bundles: https://youtu.be/UbQS40KHkH0?t=587

He says you can imagine 'turning the tangent vectors around' such that they are all aligned, as pictured second here:

He had immediately previously referred to the vertical lines in a similar diagram as being fibers that had vertical extent / 'length' to represent being composed of multiple elements that each mapped to an element of the base space (above pictured as the blue circle).

I would have thought that, in the case of the tangent vectors to a circle, each tangent vector would be an individual element of the fiber representing its tangent space, such that there would be only one (or two, due to reversing vector direction) element(s) in each fiber, and there would be no continuous-seeming vertical extent along which to pick out distinct elements, as he did in the video. I feel like there's a small chance he's implying different positions on the same tangent vector correspond to different elements in the fiber at that point, and just want to verify that that isn't the case?

I now have two mental images of what a fiber bundle is. Which of these, or neither, is right?

I'm inclined to think the first, because it makes more sense inherently and extends well to a sphere:

If it's obvious from my diagrams/descriptions that I've misunderstood anything else, I'd be grateful to be told as much!

Answer 1

The fibre in a tangent bundle is an entire vector space. In other words if ${\bf e}$ is a unit vector tangent to the circle at some point, then the fibre is the infinite set $\{x{\bf e}| x\in {\mathbb R}\}$ equiped with the usual vector addition and scalar multiplication operations.