admesh
/

agentic-intent-classifier

@@ -1,270 +1,131 @@
 ---
 language:
-- en
-library_name: transformers
 pipeline_tag: text-classification
-base_model: distilbert-base-uncased
-metrics:
-- accuracy
-- f1
 tags:
-- intent-classification
-- multitask
-- iab
-- conversational-ai
-- adtech
-- calibrated-confidence
-license: apache-2.0
 ---
-# admesh/agentic-intent-classifier
-Production-ready intent + IAB classifier bundle for conversational traffic.
-Combines multitask intent modeling, supervised IAB content classification, and per-head confidence calibration to support safe monetization decisions in real time.
-## Links
-- Hugging Face: https://huggingface.co/admesh/agentic-intent-classifier
-- GitHub: https://github.com/GouniManikumar12/agentic-intent-classifier
-## What It Predicts
-| Field | Description |
-|---|---|
-| `intent.type` | `commercial`, `informational`, `navigational`, `transactional`, … |
-| `intent.subtype` | `product_discovery`, `comparison`, `how_to`, … |
-| `intent.decision_phase` | `awareness`, `consideration`, `decision`, … |
-| `iab_content` | IAB Content Taxonomy 3.0 tier1 / tier2 / tier3 labels |
-| `component_confidence` | Per-head calibrated confidence with threshold flags |
-| `system_decision` | Monetization eligibility, opportunity type, policy |
----
-## Deployment Options
-### 0. Colab / Kaggle Quickstart (copy/paste)
-```python
-!pip -q install -U pip
-!pip -q install -U "torch==2.10.0" "torchvision==0.25.0" "torchaudio==2.10.0"
-!pip -q install -U "transformers>=4.36.0" "huggingface_hub>=0.20.0" "safetensors>=0.4.0"
-```
-Restart the runtime after installs (**Runtime → Restart runtime**) so the new Torch version is actually used.
-```python
-from transformers import pipeline
-clf = pipeline(
-    "admesh-intent",
-    model="admesh/agentic-intent-classifier",
-    trust_remote_code=True,  # required (custom pipeline + multi-model bundle)
-)
-out = clf("Which laptop should I buy for college?")
-print(out["meta"])
-print(out["model_output"]["classification"]["intent"])
-```
----
-## Latency / inference timing (quick check)
-The first call includes model/code loading. Warm up once, then measure:
-```python
-import time
-q = "Which laptop should I buy for college?"
-_ = clf("warm up")
-t0 = time.perf_counter()
-out = clf(q)
-print(f"latency_ms={(time.perf_counter() - t0) * 1000:.1f}")
-```
-### 1. `transformers.pipeline()` — anywhere (Python)
-```python
-from transformers import pipeline
-clf = pipeline(
-    "admesh-intent",
-    model="admesh/agentic-intent-classifier",
-    trust_remote_code=True,
-)
-result = clf("Which laptop should I buy for college?")
-```
-Batch and custom thresholds:
-```python
-# batch
-results = clf([
-    "Best running shoes under $100",
-    "How does TCP work?",
-    "Buy noise-cancelling headphones",
-])
-# custom confidence thresholds
-result = clf(
-    "Buy headphones",
-    threshold_overrides={"intent_type": 0.6, "intent_subtype": 0.35},
-)
-```
----
-### 2. HF Inference Endpoints (managed, deploy to AWS / Azure / GCP)
-1. Go to https://ui.endpoints.huggingface.co
-2. **New Endpoint** → select `admesh/agentic-intent-classifier`
-3. Framework: **PyTorch** — Task: **Text Classification**
-4. Enable **"Load with trust_remote_code"**
-5. Deploy
-The endpoint serves the same `pipeline()` interface above via REST:
-```bash
-curl https://<your-endpoint>.endpoints.huggingface.cloud \
-  -H "Authorization: Bearer $HF_TOKEN" \
-  -H "Content-Type: application/json" \
-  -d '{"inputs": "Which laptop should I buy for college?"}'
-```
----
-### 3. HF Spaces (Gradio / Streamlit demo)
-```python
-# app.py for a Gradio Space
-import gradio as gr
-from transformers import pipeline
-clf = pipeline(
-    "admesh-intent",
-    model="admesh/agentic-intent-classifier",
-    trust_remote_code=True,
-)
-def classify(text):
-    return clf(text)
-gr.Interface(fn=classify, inputs="text", outputs="json").launch()
-```
----
-### 4. Local / notebook via `snapshot_download`
-```python
-import sys
-from huggingface_hub import snapshot_download
-local_dir = snapshot_download(
-    repo_id="admesh/agentic-intent-classifier",
-    repo_type="model",
-)
-sys.path.insert(0, local_dir)
-from pipeline import AdmeshIntentPipeline
-clf = AdmeshIntentPipeline()
-result = clf("I need a CRM for a 5-person startup")
-```
-Or the one-liner factory:
-```python
-from pipeline import AdmeshIntentPipeline
-clf = AdmeshIntentPipeline.from_pretrained("admesh/agentic-intent-classifier")
-```
----
-## Troubleshooting (avoid environment errors)
-### `No module named 'combined_inference'` (or similar)
-This means the Hub repo root is missing required Python files. Ensure these exist at the **root of the model repo** (same level as `pipeline.py`):
-- `pipeline.py`, `config.json`, `config.py`
-- `combined_inference.py`, `schemas.py`
-- `model_runtime.py`, `multitask_runtime.py`, `multitask_model.py`
-- `inference_intent_type.py`, `inference_subtype.py`, `inference_decision_phase.py`, `inference_iab_classifier.py`
-- `iab_classifier.py`, `iab_taxonomy.py`
-### `does not appear to have a file named model.safetensors`
-Transformers requires a standard checkpoint at the repo root for `pipeline()` to initialize. This repo includes a **small dummy** `model.safetensors` + tokenizer files at the root for compatibility; the *real* production weights live in:
-- `multitask_intent_model_output/`
-- `iab_classifier_model_output/`
-- `artifacts/calibration/`
----
-## Example Output
-```json
-{
-  "model_output": {
-    "classification": {
-      "iab_content": {
-        "taxonomy": "IAB Content Taxonomy",
-        "taxonomy_version": "3.0",
-        "tier1": {"id": "552", "label": "Style & Fashion"},
-        "tier2": {"id": "579", "label": "Men's Fashion"},
-        "mapping_mode": "exact",
-        "mapping_confidence": 0.73
-      },
-      "intent": {
-        "type": "commercial",
-        "subtype": "product_discovery",
-        "decision_phase": "consideration",
-        "confidence": 0.9549,
-        "commercial_score": 0.656
-      }
-    }
-  },
-  "system_decision": {
-    "policy": {
-      "monetization_eligibility": "allowed_with_caution",
-      "eligibility_reason": "commercial_discovery_signal_present"
-    },
-    "opportunity": {"type": "soft_recommendation", "strength": "medium"}
-  },
-  "meta": {
-    "system_version": "0.6.0-phase4",
-    "calibration_enabled": true,
-    "iab_mapping_is_placeholder": false
-  }
-}
-```
-## Reproducible Revision
-```python
-from huggingface_hub import snapshot_download
-local_dir = snapshot_download(
-    repo_id="admesh/agentic-intent-classifier",
-    repo_type="model",
-    revision="0584798f8efee6beccd778b0afa06782ab5add60",
-)
-```
-## Included Artifacts
-| Path | Contents |
 |---|---|
-| `multitask_intent_model_output/` | DistilBERT multitask weights + tokenizer |
-| `iab_classifier_model_output/` | IAB content classifier weights + tokenizer |
-| `artifacts/calibration/` | Per-head temperature + threshold JSONs |
-| `pipeline.py` | `AdmeshIntentPipeline` (transformers.Pipeline subclass) |
-| `combined_inference.py` | Core inference logic |
-## Notes
-- `trust_remote_code=True` is required because this model uses a custom multi-head architecture that does not map to a single standard `AutoModel` checkpoint.
-- `meta.iab_mapping_is_placeholder: true` means IAB artifacts were missing or skipped; train and calibrate IAB for full production accuracy.
-- For long-running servers, instantiate once and reuse — models are cached in memory after the first call.

 ---
 language:
+  - en
 pipeline_tag: text-classification
 tags:
+  - intent-classification
+  - conversational-ai
+  - contextual-relevance
+license: other
 ---
+# Ingence 0.1
+Ingence is a proprietary intent-understanding model developed by AdMesh for
+conversational and agentic applications.
+> **Release status:** Experimental. Ingence 0.1 is intended for evaluation and
+> controlled use. It has not been validated for safety-critical or high-impact
+> decisions.
+## What It Does
+Ingence converts a user request or short conversation into structured signals
+describing:
+- the user's general intent;
+- a more specific intent subtype;
+- the user's stage in a decision journey;
+- a relevant contextual content category; and
+- whether the system should fall back because a prediction is uncertain.
+These signals can support intent routing, contextual relevance, aggregate
+analytics, and controlled advertising or content-selection experiences.
+## Intended Uses
+Ingence 0.1 is intended for:
+- intent routing in assistants and agents;
+- understanding research, comparison, and purchase-oriented requests;
+- contextual content and advertising selection;
+- aggregate intent analytics; and
+- research, testing, and controlled product evaluation.
+## Restricted and Unsupported Uses
+Ingence must not be used:
+- to infer sensitive personal characteristics;
+- for surveillance or individual behavioral profiling;
+- as the sole moderation or safety mechanism;
+- to make decisions about credit, employment, housing, insurance, healthcare,
+  education, or legal services;
+- as the sole basis for decisions that materially affect a person; or
+- to target advertising using sensitive personal information.
+## Training and Evaluation
+Ingence 0.1 was developed using a mixture of synthetic, curated, and
+taxonomy-based examples. Synthetic examples were used to improve coverage of
+different intents and linguistic patterns.
+Important qualifications:
+- Synthetic-label agreement does not establish real-world accuracy.
+- Training data may not represent every population, dialect, industry, or
+  interaction style.
+- Evaluation on representative, human-reviewed production data is ongoing.
+- Definitive accuracy, fairness, and calibration claims are not currently
+  published for this experimental release.
+## Confidence Scores
+Model confidence values are not guaranteed probabilities. Applications should
+not interpret a score such as `0.80` as an 80% probability that a prediction is
+correct. Integrations should honor uncertainty and fallback indicators.
+## Known Limitations
+Ingence 0.1 may:
+- confuse closely related research and decision stages;
+- misclassify short or context-dependent follow-up messages;
+- have difficulty distinguishing general price questions from commercial
+  intent;
+- return broad content categories when several related categories apply;
+- fall back on valid requests when confidence is low;
+- perform differently on language styles not represented in development data;
+  and
+- produce unreliable results for languages other than English.
+Outputs are probabilistic predictions and should be treated as one input into a
+broader decision system.
+## Safety and Privacy
+Applications using Ingence should:
+- minimize personal information included in requests;
+- avoid retaining raw conversations unnecessarily;
+- provide a safe fallback when the service is uncertain or unavailable;
+- use independent policy enforcement for sensitive applications;
+- monitor model quality and distribution drift; and
+- clearly disclose when automated predictions materially influence an
+  experience.
+## Access and Intellectual Property
+Ingence 0.1 is proprietary technology owned by AdMesh. Model weights, source
+code, training data, internal evaluation artifacts, and deployment details are
+not publicly distributed. No license to copy, modify, redistribute, reverse
+engineer, or create derivative works is granted by this model card.
+Authorized access may be provided through AdMesh-operated services under
+applicable commercial terms.
+## Version Information
+| Property | Value |
 |---|---|
+| Model | Ingence |
+| Version | 0.1 |
+| Developer | AdMesh |
+| Primary language | English |
+| Release status | Experimental |
+## Contact
+For product questions, feedback, privacy requests, or responsible disclosure,
+visit [useadmesh.com](https://useadmesh.com).