[QUESTION] Possible mistake in transformers size in terms of parameters

[jonoillar]

Hey,

Thanks for the great course!

I have a question on what looks to me like an inconsistency.
In the unit1/what-are-llms section, when explaining the 3 types of transformers, in the Typical Size, we can see:

Decoders:
Typical Size: Billions (in the US sense, i.e., 10^9) of parameters

Seq2Seq (Encoder–Decoder)
Typical Size: Millions of parameters

It looks strange to me that a Seq2Seq transformer, which comprises a Decoder within it, is smaller in Typical Size than a plain Decoders.

I would put

Seq2Seq (Encoder–Decoder)
Typical Size: Billions (in the US sense, i.e., 10^9) of parameters

Please tell me if there is something I misunderstood !