I've got Trainyourears - may have first heard it from you on the PUF a while back (was that you?).
Ear training is good for musicians AND engineers - wish I had heard this when I was first starting. Luckily I did quite well with musical ear training (couldn't read music worth a crap but could identify intervals almost infallibly!). I believe that helped to small degree, but they are actually two different things - interval detection allows you to recognize pitch relationships while EQ training is about about recognizing absolute pitch.
Notice I said pitch, and not frequency. The other thing I wish I had learned better early on was that frequency and pitch are just different names for the same thing. Seems obvious now, but I spent a LOT of time trying to learn what each frequency sounded like when EQing, forgetting to simply relate things to PITCH (which I already understood) and not frequency.
Relating to pitch when EQing makes sense, and IMO understanding the octave relationship is the key to understanding EQing. Graphic EQs are specified by how many bands "per octave", as their associated analyzers also measured the spectrum in octaves (using pink noise, which represents equal energy in each octave). Q factor, or more specifically "bandwidth" is also more typically measured in octaves for musical applications. This is to ensure you have the same Q "value" at any pitch/frequency. If you measure bandwidth in Hertz, moving the frequency would change the bandwidth values, an undesirable situation at best.
My approach - it's free!
When I first started learning the frequency spectrum and EQ, I began with the concepts of "bass" and "treble" as were on my first receiver when I was a kid. Those were easy to hear. Then I added a midrange control (on a friends mixer), then two mid bands. Later I was working with many bands across the spectrum, but the point is that I began with the basics and slowly learned to divide the frequency spectrum into smaller and smaller chunks. In my case it was a natural progression that followed the gear I was able to afford - more money meant mixers with more EQ bands!
But for others I would still suggest a similar path. Start with a simple bass shelf, wide mid parametric, and treble shelf setup, boosting and cutting each as you listen on different instruments. 100, 1000, and 10,000 Hz are good traditional starting points. Also try eq'ing sub mixes and even an entire mix, listening for the effect of cutting and boosting each band and learning to predict the results in your head.
Then add a second more narrow mid band (adjusting the first as well) and repeat. Try to hear the band in your head before you boost it - see how close you were and try again. Keep going from there!
Finally you should try to get to the point where you can identify the 10 octaves that make up our audible frequency spectrum. Probably no need to go further than that - if you can make it this far you're doing fantastic!
Octave 1: 20-40 Hz
Octave 2: 40-80 Hz
Octave 3: 80-160 Hz
Octave 4: 160-320 Hz
Octave 5: 320-640 Hz
Octave 6: 640-1280 Hz
Octave 7: 1280-2560 Hz
Octave 8: 2560-5120 Hz
Octave 9: 5120-10,240 Hz
Octave 10: 10,240-20,480 Hz
For pitch reference, the closest musical note to these frequencies is D#, at a frequency of 19.44545 Hz (and multiples thereafter). Even if you can't identify each octave by ear, it's still a good thing to conceptualize EQ in relation to pitch rather than frequency. IMO, as always.