Add infer_narrator_tts.py

Add infer_narrator_tts.py

Add train_narrator_tts.py

Add train_narrator_tts.py

Add train_narrator_tts.py

Add train_narrator_tts.py

Add train_narrator_tts.py

Add train_narrator_tts.py

Add train_narrator_tts.py

Add train_narrator_tts.py

Add train_narrator_tts.py

Add train_narrator_tts.py

Add train_narrator_tts.py

Add train_narrator_tts.py

Add train_narrator_tts.py

Add train_narrator_tts.py

Add train_narrator_tts.py

Add train_narrator_tts.py

Add train_narrator_tts.py

Add train_narrator_tts.py

Add train_narrator_tts.py

Add train_narrator_tts.py

Add train_narrator_tts.py

Add train_narrator_tts.py

fix: apt-install cmake + build-essential before llama.cpp build

upload convert_gguf.py

fix: pre-render chat template via .map() instead of formatting_func

fix: formatting_func returns list per Unsloth batched contract

fix: add formatting_func + warmup_steps for TRL 0.12 compat

fix: correct Stheno-v3.4 repo name + Unsloth import order

upload train_lora.py (renamed)

initial commit