str.split method without arguments

It does the same filtering as you do. Just skip ' ' argument.

list(dict.fromkeys(list comprehension))

It looks overburdened, right? You can create a set of list comprehension as well:

newlst = list({x for x in txt.split(' ')})

or just

newlst = list(set(txt.split(' ')))

but the best way is

collections.Counter

You're reinventing it with worse time complexity.

Format strings

f-strings looks better (but it depends on your task).

Function name

PEP8 advices to use lower case with underscores: print_words_occurence

All together

from collections import Counter
def print_words_occurence(txt):
   for word, count in Counter(txt.split()).items():
       print(f"[{count}] {word}")

Also consider dividing an algorithmic part and input/output - like yielding a pair (word, count) from the function and outputting it somewhere else.