Upload Concept Memory base model

README.md

@@ -23,15 +23,22 @@ not a language model pretrained from scratch. Its text encoder is
 - `model.pt`: PyTorch state dictionary and training metadata.
 - `config.json`: encoder, task, and training run metadata.
 
-## Load the checkpoint
+## Run a prediction
 
-Install this project's dependencies, then load it with the repository's model
-class:
+Use Python 3.11 or newer. Install the model code and the Hub client:
+
+```bash
+pip install "git+https://github.com/ceylonai/rgcm.git" huggingface_hub
+```
+
+Download the checkpoint, construct a question with the criteria you want the
+model to judge, and inspect its prediction and probabilities:
 
 ```python
 import torch
 from huggingface_hub import hf_hub_download
 from concept_memory import ConceptMemoryModel
+from concept_memory.types import Question
 
 repo_id = "Syigen/rgcm"
 checkpoint = hf_hub_download(repo_id, "model.pt")
@@ -43,6 +50,27 @@ model = ConceptMemoryModel.from_pretrained(
 )
 model.load_state_dict(payload["model_state_dict"])
 model.eval()
+
+question = Question(
+    id="intent",
+    type="choice",
+    instructions="Which team should handle this customer request?",
+    criteria=[
+        {"id": "billing", "description": "Payment or subscription issues"},
+        {"id": "technical", "description": "Bugs or integration problems"},
+        {"id": "sales", "description": "Pricing or account questions"},
+    ],
+)
+
+with torch.inference_mode():
+    result = model.predict(
+        "Stripe integration keeps failing after several attempts.", [question]
+    )[0]
+
+criteria = question.normalized_criteria()
+predicted = criteria[result.prediction.item()]
+print("Prediction:", predicted.id, "-", predicted.description)
+print("Probabilities:", result.probabilities.tolist())
 ```
 
 The Hub checkpoint is a PyTorch artifact for this project's Python package; it
@@ -50,6 +78,7 @@ is not directly compatible with `transformers.AutoModel.from_pretrained`.
 
 ## Training metadata
 
+```json
 {
   "datasets": [
     "mteb/banking77",
@@ -66,3 +95,4 @@ is not directly compatible with `transformers.AutoModel.from_pretrained`.
   "batch_size": 16,
   "seed": 42
 }
+```