Text Classification
Transformers
ONNX
Safetensors
English
distilbert
intent-classification
multitask
iab
conversational-ai
adtech
calibrated-confidence
text-embeddings-inference
Instructions to use admesh/agentic-intent-classifier with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use admesh/agentic-intent-classifier with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-classification", model="admesh/agentic-intent-classifier")# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("admesh/agentic-intent-classifier", device_map="auto") - Notebooks
- Google Colab
- Kaggle
| # Composed Output Examples | |
| These examples show the end-to-end flow: | |
| `query -> model_output -> system_decision` | |
| ## 1. Informational Awareness | |
| Query: | |
| ```text | |
| What is CRM software? | |
| ``` | |
| Observed behavior: | |
| - classified as `intent.type=informational` | |
| - classified as `decision_phase=awareness` | |
| - fell back because confidence is below the demo threshold | |
| - system decision sets `monetization_eligibility=not_allowed` | |
| Why fallback happened: | |
| - the phase is sensible, but the combined confidence is still too low for monetization decisions | |
| ## 2. Commercial Comparison | |
| Query: | |
| ```text | |
| Compare AI search monetization platforms for publishers | |
| ``` | |
| Observed behavior: | |
| - classified as `intent.type=commercial` | |
| - classified as `decision_phase=consideration` | |
| - still falls back due low confidence | |
| - opportunity remains `none` because fallback takes precedence | |
| Why fallback happened: | |
| - the underlying labels are directionally right, but the frozen baseline is intentionally conservative | |
| ## 3. Post-Purchase Setup | |
| Query: | |
| ```text | |
| How do I set up my new CRM? | |
| ``` | |
| Observed behavior: | |
| - classified as `decision_phase=post_purchase` | |
| - treated as low-confidence informational setup behavior | |
| - system decision stays restricted | |
| Why fallback happened: | |
| - the system recognizes the lifecycle stage, but confidence is not high enough to remove fallback | |
| ## 4. Support / Sensitive | |
| Query: | |
| ```text | |
| I cannot log into my account | |
| ``` | |
| Observed behavior: | |
| - classified as `decision_phase=support` | |
| - intent-type falls into `personal_reflection` in `v0.1` | |
| - fallback applies with `policy_default` | |
| - monetization is blocked | |
| Why fallback happened: | |
| - this is a known limitation of the intent-type head and exactly the kind of case that should fail safe | |