Take the Finnish word Terve /tɛrʋɛ/ as an example, how to pronounce the R which comes after a vowel and is followed by a consonant? I have listened to some samples but they just sound like a tapped R, not a trilled R, and it seems so difficult to produce a completely trilled R at that position, the same happens to me for Russian and Ukrainian languages, could someone please explain it? Thank you.
You don't say what your acoustic models are, so perhaps the individual saying the words are not producing trills. Here are some examples of the Finnish word perkele, järvi, terve and sormi. You can't compare trilled r in Finnish to tapped r, because there is only one rhotic in the language. You can find intervocalic trills vs. taps in Spanish, signaled as rr vs r in the classic minimal pairs pero – perro, caro – carro, quería – querría.
The best way to develop the ability to perceive a pre-consonantal trill is to start with an intervocalic trill in a known case, thus Spanish or Italian caro vs. carro. Obviously you need real data, so you need a speaker of the language (that's why I'm naming less-obscure languages). You can use Forvo except that it may be hard to find the same person saying carro vs caro do you won't know if the difference is just because two people have different speech equipment. Ideally, you would obtain samples, you could listen and hear the difference; and in Praat, looking at spectrograms, you can see the difference.
I do not know of any language that has a robust contrast between trill and tap in VRC.