Audio-to-Audio
Moshi
English
contextspan
full-duplex
spoken-dialogue
speech-to-speech
retrieval
retrieval-augmented-generation
tool-calling
context-spanning
personaplex
Instructions to use mindlogicinc/context-spanning-7b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Moshi
How to use mindlogicinc/context-spanning-7b with Moshi:
# pip install moshi # Run the interactive web server python -m moshi.server --hf-repo "mindlogicinc/context-spanning-7b" # 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("mindlogicinc/context-spanning-7b") # 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) - Notebooks
- Google Colab
- Kaggle
model card wording
Browse files
README.md
CHANGED
|
@@ -73,7 +73,7 @@ vendored in the code repository (`contextspan/moshi/`).
|
|
| 73 |
- Frame clock: 80 ms per step
|
| 74 |
- Sequence convention: `<ret>` = 4, span open = 12, span close = 13, text pad = 3
|
| 75 |
- Released checkpoint: 2026-09-23
|
| 76 |
-
- Backends
|
| 77 |
- Language: English
|
| 78 |
|
| 79 |
## Benchmark Results
|
|
|
|
| 73 |
- Frame clock: 80 ms per step
|
| 74 |
- Sequence convention: `<ret>` = 4, span open = 12, span close = 13, text pad = 3
|
| 75 |
- Released checkpoint: 2026-09-23
|
| 76 |
+
- Backends used for the results: router `google/gemma-4-26B-A4B-it` on vLLM, ASR Qwen3-ASR-1.7B
|
| 77 |
- Language: English
|
| 78 |
|
| 79 |
## Benchmark Results
|