Views
No views yet
Hmong Text‑To‑Speech (TTS) model fine‑tuned from Orpheus‑3B‑TTS, optimized with Unsloth + SNAC codec. Built by LocalVoice.org to support Hmong language technology.
<giggle>, <laugh>, <sigh>, <cough>, <sniffle>, <groan>, <yawn>, <gasp>
1import torch
2from unsloth import FastLanguageModel
3
4MODEL_PATH = "Pakorn2112/Orpheus-TTS-hmong-3b"
5
6model, tokenizer = FastLanguageModel.from_pretrained(
7 model_name = MODEL_PATH,
8 max_seq_length = 2048, # Choose any for long context!
9 dtype = None, # Select None for auto detection
10 load_in_4bit = False, # Select True for 4bit which reduces memory usage
11 # token = "YOUR_HF_TOKEN", # HF Token for gated models
12)
13apache-2.0