Haha, if ever there was a definition of a word dealt with badly by a hobby community...
An intergrade is an intermediate form, where two things are blended. In Australia this is generally spoken about in reference to animals of one type in one particular area (correctly, but with a fundamental misunderstanding), and also to a type of hybrid between animals from nearby areas (incorrectly for various reasons).
There are two different types of intergrades (primary and secondary). There is a complete spectrum of hybrids. At one extreme, almost every living thing created sexually (including you and me) is a hybrid, and at the other extreme is the hybridisation of extremely different things. Any living thing which is created as a mixture of two non identical things, is a hybrid.
A primary intergradation is where there are two different things (generally about as different as we'd usually classify as subspecies) and there is continual variation, like a spectrum, as you look at the population going from one region to the other. If you think about Carpet Pythons, there are basically three geographically isolated populations (ignoring a few island and small isolated populations). These are the South Western Carpet Python, Morelia spilota imbricata (in the south west), the Centralian Carpet Python, Morelia bredli (in the centre), and the rest (in the east and north). The Murray Darling Carpet is more or less isolated by the Great Dividing Range in the east, but occasional individuals cross it, giving somewhat of a secondary integradation situation through the process of hybridisation (in this case there are two populations maintaining distinct identities and genetic profiles because of the geographic barrier). In the most discussed case, Diamonds and Coastals intergrade, and this is a clear case of primary intergradation, with nothing you'd call hybridisation (very different from the situation with Murray Darlings and Coastals). You never get a completely classic Diamond like you'd see in Sydney mating with a completely classic Coastal like you'd see in Brisbane meeting up with each other. These populations have never been separated by any sort of barrier and they have diverged while remaining as one big population. For this reason there has never been secondary intergradation through hybridisation, which is a big misconception told and retold by countless herpers, including in the post above this one. There is no overlap, there is a spectrum.
A hybrid on the other hand is something produced by two things which are different in some relevant way. We don't get naturally occurring hybrid (or intergrade) Diamond/Coastals from the mating of obvious Diamonds and Coastals with each other. A Sydney Carpet (Diamond) crossed with a Brisbane Carpet produces an intersubspecific hybrid. Incorrectly, most reptile hobbyists will tell you that this is not a hybrid, but it is indeed a hybrid. Among laymen there is a myth that 'hybrid' has a definition which is the same as 'interspecific hybrid', and doesn't include anything else. Interspecific hybrids (the product of parents which are of two different species) is indeed a hybrid, but it's only one type of hybrid. You can get more extreme levels of hybrids, and these are common in captive reptile breeding. The next level up is intergeneric hybrid (the product of parents from two different genera, such as Carpet Python x Woma or Carpet Python x Ball Python), and you can go even higher to interfamilial hybrids, which off the top of my head I can't think of a reptilian example of, but I'm pretty sure they do exist.
Since all we need to classify something as a hybrid is having parents with a difference of relevance, things like normal x albino snakes are examples of hybrids, and in scientific literature examples like this are referred to as such when the genetics are being discussed. But, in colloquial laymen reptile keeper vernacular, 'hybrid' means 'interspecific hybrid', and many of them will argue fiercely with you if you say otherwise!
