Document gender-conditioned inference

README.md

@@ -8,14 +8,17 @@ tags:
   - phonetics
   - tts
   - text-to-speech
+  - gender-conditioned
 pipeline_tag: text-to-speech
 ---
 
 # ReNikud Plus
 
-Audio-supervised Hebrew grapheme-to-phoneme (G2P) model. Converts unvocalized Hebrew text to IPA for TTS, speech technology, and spoken-language research.
+Hebrew grapheme-to-phoneme (G2P) model that converts unvocalized text to IPA
+for TTS, speech technology, and spoken-language research.
 
-Weights are packaged as a self-contained ONNX file (`model.onnx`, ~20 MB) with vocabulary and letter constraints embedded in model metadata.
+`model.onnx` is a self-contained, INT8 ONNX model. It accepts speaker controls:
+`0` for unknown, `1` for male, and `2` for female.
 
 ## Benchmark
 
@@ -29,14 +32,11 @@ hf download notmax123/RenikudPlus model.onnx --local-dir .
 
 ## Usage
 
-Install the inference package from the [GitHub repository](https://github.com/maxmelichov/RenikudPlus), then run:
-
 ```python
 from renikud_onnx import G2P
 
 g2p = G2P("model.onnx")
-print(g2p.phonemize("שלום עולם"))
-# → ʃlˈom ʔolˈam
+print(g2p.phonemize("היא רצה", speaker=2, target_speaker=2))
 ```
 
 ## Citation