filliptmandClaude Opus 4.6
60f6897263
Custom node pack implementing Mistral's Voxtral-4B-TTS-2603 model with direct PyTorch inference (no vLLM dependency), enabling native MPS/Mac support alongside CUDA. Three-component architecture: - Mistral LLM backbone (26 layers, 3072 dim, GQA, RoPE) - Flow-matching acoustic transformer (3 layers, 8-step Euler ODE, CFG) - Neural audio codec decoder (4-stage ALiBi transformer + conv upsampling) Features: - 20 preset voices across 9 languages - 24 kHz mono output - BF16/FP16/FP32 dtype support - HuggingFace Hub auto-download with caching - ComfyUI AUDIO output format Nodes: - FL_VoxtralTTS_ModelLoader: download and load model - FL_VoxtralTTS_Generate: text + voice → audio Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
57 lines
1.9 KiB
Python
57 lines
1.9 KiB
Python
"""ComfyUI-FL-VoxtralTTS — Mistral Voxtral-4B Text-to-Speech for ComfyUI."""
|
|
|
|
import os
|
|
import sys
|
|
import logging
|
|
|
|
logger = logging.getLogger(__name__)
|
|
current_dir = os.path.dirname(os.path.abspath(__file__))
|
|
|
|
# Add src to sys.path so voxtral_tts package is importable
|
|
src_dir = os.path.join(current_dir, "src")
|
|
if src_dir not in sys.path:
|
|
sys.path.insert(0, src_dir)
|
|
|
|
# Register TTS model folder with ComfyUI
|
|
try:
|
|
import folder_paths
|
|
|
|
tts_base = os.path.join(folder_paths.models_dir, "tts")
|
|
voxtral_dir = os.path.join(tts_base, "VoxtralTTS")
|
|
os.makedirs(tts_base, exist_ok=True)
|
|
os.makedirs(voxtral_dir, exist_ok=True)
|
|
|
|
supported_ext = {".safetensors", ".bin", ".pt", ".pth", ".json"}
|
|
if "tts" not in folder_paths.folder_names_and_paths:
|
|
folder_paths.folder_names_and_paths["tts"] = ([tts_base], supported_ext)
|
|
except Exception as e:
|
|
logger.warning(f"Could not register TTS folder: {e}")
|
|
|
|
# Import nodes
|
|
from .nodes.model_loader import FL_VoxtralTTS_ModelLoader
|
|
from .nodes.generate import FL_VoxtralTTS_Generate
|
|
|
|
NODE_CLASS_MAPPINGS = {
|
|
"FL_VoxtralTTS_ModelLoader": FL_VoxtralTTS_ModelLoader,
|
|
"FL_VoxtralTTS_Generate": FL_VoxtralTTS_Generate,
|
|
}
|
|
|
|
NODE_DISPLAY_NAME_MAPPINGS = {
|
|
"FL_VoxtralTTS_ModelLoader": "FL Voxtral TTS Model Loader",
|
|
"FL_VoxtralTTS_Generate": "FL Voxtral TTS Generate",
|
|
}
|
|
|
|
WEB_DIRECTORY = None
|
|
|
|
__all__ = ["NODE_CLASS_MAPPINGS", "NODE_DISPLAY_NAME_MAPPINGS"]
|
|
|
|
# Banner
|
|
BANNER = """
|
|
╔══════════════════════════════════════════════╗
|
|
║ FL Voxtral TTS — Mistral 4B ║
|
|
║ 20 voices • 9 languages • 24kHz ║
|
|
║ Direct PyTorch inference ║
|
|
╚══════════════════════════════════════════════╝
|
|
"""
|
|
logger.info(BANNER)
|