update README.md

README.md

@@ -1,10 +1,40 @@
+---
+language:
+  - en
+license: apache-2.0
+pipeline_tag: feature-extraction
+library_name: transformers
+tags:
+  - pytorch
+  - toy-model
+  - tiny-mlp
+base_model: none
+base_model_relation: finetune
+---
+
+# TinyModel
+
+A small MLP (100 → 200 → 10) using ReLU and Softmax, designed as a simple example for publisher workflows.
+
+## Intended Uses & Limitations
+
+This model is purely educational and not trained on real-world data. 🧪 It is not suitable for production tasks or any mission-critical usage.
+
+## Usage
+
 ```python
 import torch
-from transformers import AutoModel
-
-model = AutoModel.from_pretrained("tester-123456789/tiny-model", trust_remote_code=True)
-x = torch.randn(2, 100)   # batch=2, features=100
-with torch.no_grad():
-    probs = model(x)
-print(probs.shape)  # torch.Size([2, 10])
-```
+from tiny_model import TinyModel
+
+model = TinyModel()
+state = torch.load(
+    "https://huggingface.co/tester-123456789/tiny-model/resolve/main/pytorch_model.bin",
+    map_location="cpu",
+    weights_only=True
+)
+model.load_state_dict(state)
+model.eval()
+
+x = torch.randn(1, 100)
+y = model(x)
+print(y)