docs(echo_tts): follow the encode_latents rename

The accessor was renamed from encode_zq in the previous commit; the model
doc still named the old one. The remaining z_q mentions in the docs refer to
the upstream PyTorch DAC.encode_zq / DAC.decode_zq, which are correct as
written and should not be renamed with our C++ API.

Author: 5uck1ess

docs/community_models/echo_tts.md

@@ -29,7 +29,7 @@ prove neither.
 | M3 | Long-form via the framework text chunker | yes | **no** |
 | M4 | Q8_0 conversion, RTF and memory evidence | partial | **no** |
 
-Cloning is self-contained — `session.cpp` calls `codec_->encode_zq` directly, so no pre-computed
+Cloning is self-contained — `session.cpp` calls `codec_->encode_latents` directly, so no pre-computed
 speaker latent is required.
 
 ### Numerical parity against PyTorch