Instructions to use nvidia/personaplex-7b-v1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Moshi
How to use nvidia/personaplex-7b-v1 with Moshi:
# pip install moshi # Run the interactive web server python -m moshi.server --hf-repo "nvidia/personaplex-7b-v1" # Then open https://localhost:8998 in your browser
# pip install moshi import torch from moshi.models import loaders # Load checkpoint info from HuggingFace checkpoint = loaders.CheckpointInfo.from_hf_repo("nvidia/personaplex-7b-v1") # Load the Mimi audio codec mimi = checkpoint.get_mimi(device="cuda") mimi.set_num_codebooks(8) # Encode audio (24kHz, mono) wav = torch.randn(1, 1, 24000 * 10) # [batch, channels, samples] with torch.no_grad(): codes = mimi.encode(wav.cuda()) decoded = mimi.decode(codes) - Inference
- Notebooks
- Google Colab
- Kaggle
How to run it on Kaggle
#35
by Niroop-2007 - opened
Can someone guide me a bit regarding this - I am planning to run it on Kaggle
But there are many errors
any additional information
PERSONAPLEX COLAB INSTALLATION REPORT
π PROBLEM SUMMARY
- Platform: Google Colab Pro (A100 GPU)
- Model: nvidia/personaplex-7b-v1
- Target: Real-time voice chat server
β ENCOUNTERED PROBLEMS
1. Model Loading Error (dep_q)
KeyError: 'dep_q'
File: moshi/models/loaders.py, line 284
Cause: Version incompatibility between Moshi library and PersonaPlex model checkpoint
2. --cpu-offload Parameter Not Recognized
server.py: error: unrecognized arguments: --cpu-offload
Cause: The stable version of Moshi on GitHub does not support this parameter
3. Starting Server with Threading Failed
def start_server():
!python -m moshi.server ...
server_thread = threading.Thread(target=start_server, daemon=True)
server_thread.start()
Cause: In Jupyter/Colab, the ! command runs subprocess, doesn't work properly inside threads
4. Dependency Conflicts
ERROR: transformers 5.0.0 requires huggingface-hub<2.0,>=1.3.0
but you have huggingface-hub 0.36.2
The grind is real