Base models

All tasks that can be tackled with Artifex can be performed in two ways:

1. Using default models that are pre-trained and ready to use for inference, but are not optimized for your specific use case.
2. Fine-tuning a default model to create a task-specific model that is better suited for your needs; fine-tuning does not require any training data, as Artifex generates synthetic data on-the-fly, based on your instructions.

All default models are open-source models trained by us with the Artifex library itself (meta 😁), are small in size (around 0.1B parameters and 500Mb), can run on CPU for inference and fine-tuning and are hosted on our Hugging Face account.

Below is a list of all the available tasks in Artifex, along with the link to the default model for each task.

Task	Default Model	Default & Fine-Tuned Model Size	CPU Inference	CPU Fine-Tuning
🛡️ Guardrail	tanaos/tanaos-guardrail-v1	0.1B params, 500Mb	✅	✅
🗣️ Intent Classification	tanaos/tanaos-intent-classifier-v1	0.1B params, 500Mb	✅	✅
🔀 Reranker	cross-encoder/mmarco-mMiniLMv2-L12-H384-v1	0.1B params, 470Mb	✅	✅
🙂 Sentiment Analysis	tanaos/tanaos-sentiment-analysis-v1	0.1B params, 470Mb	✅	✅
😡 Emotion Detection	tanaos/tanaos-emotion-detection-v1	0.1B params, 470Mb	✅	✅
🏷️ Named Entity Recognition	tanaos/tanaos-NER-v1	0.1B params, 500Mb	✅	✅
🥸 Text Anonymization	tanaos/tanaos-text-anonymizer-v1	0.1B params, 500Mb	✅	✅