| """Tag the turns of a voice-agent call with Pipecat Tone. |
| |
| pip install -r requirements.txt |
| python example.py # full precision: model.opt.onnx |
| python example.py --fp16 # the 16-bit file: model_fp16.onnx, half the size |
| |
| The model files are read from this folder. Without them, as in a git checkout, they are downloaded once from Hugging |
| Face (pipecat-ai/tone, release rc1). With --fp16, only model_fp16.onnx, tokenizer.json and config.json are needed. |
| """ |
| import pathlib |
| import sys |
|
|
| HERE = pathlib.Path(__file__).resolve().parent |
| sys.path.insert(0, str(HERE / "runtime")) |
|
|
| from tonetext.runtime.classifier import ToneTextClassifier |
|
|
| PRECISION = "fp16" if "--fp16" in sys.argv[1:] else "fp32" |
| MODEL_FILES = (["model_fp16.onnx", "tokenizer.json", "config.json"] if PRECISION == "fp16" |
| else ["model.opt.onnx", "model.onnx", "tokenizer.json", "config.json"]) |
| model_dir = HERE |
| if not all((HERE / f).exists() for f in MODEL_FILES): |
| from huggingface_hub import snapshot_download |
|
|
| model_dir = snapshot_download("pipecat-ai/tone", revision="rc1", allow_patterns=MODEL_FILES) |
| tone = ToneTextClassifier(model_dir, threads=4, precision=PRECISION) |
|
|
| |
| |
| CALL = [ |
| "Thanks for calling Northwind Broadband, this is Ava. How can I help today?", |
| "hi um yeah my internet's been down since this morning", |
| "I'm sorry to hear that. Which plan are you on?", |
| "I think my husband set it up, I'm not sure", |
| "No problem. Can I take the account holder's date of birth?", |
| "Why do you need that?", |
| "It's only to confirm the account. The next engineer visit is in about forty minutes, I'm afraid.", |
| "Oh great, a forty minute wait. Love that.", |
| "Let me try a reset from my side first. The light on your router should turn green.", |
| "mm-hm", |
| "Do you know which router model you have?", |
| "No idea, sorry.", |
| "That's fine. Can you try loading a page now?", |
| "That's brilliant, it's working, thank you so much!", |
| "Is there anything else I can help with?", |
| "thanks, that's all i needed", |
| ] |
|
|
| for previous, text in zip(CALL, CALL[1:]): |
| tags = tone.tag(text, previous) |
| sarcasm = f" 路 sarcastic ({tags.sarcasm.probability:.2f})" if tags.sarcasm.sarcastic else "" |
| backchannel = " 路 backchannel" if tags.backchannel.is_backchannel else "" |
| print(f"{text!r}") |
| print(f" emotion {tags.emotion.label} ({tags.emotion.intensity}) 路 move {tags.move.label} ({tags.move.confidence:.2f})" |
| f"{sarcasm}{backchannel}") |
|
|
| |
| print(tone.tag("Oh great, a forty minute wait. Love that.", |
| previous="It's only to confirm the account. The next engineer visit is in about forty minutes, I'm afraid.") |
| .model_dump_json(indent=1)) |
|
|
