minor fix

.gitattributes

@@ -0,0 +1 @@
+*.safetensors filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,78 @@
+---
+language: en
+license: mit
+tags:
+- text-classification
+- spam-detection
+- content-moderation
+- small-model
+- tanaos
+---
+
+<p align="center">
+    <img src="https://raw.githubusercontent.com/tanaos/.github/master/assets/logo.png" width="250px" alt="Tanaos – Private Small Language Models for all your needs">
+</p>
+
+# A small, private Spam Detection model for English text
+
+This is a **spam detection model** trained to classify English text as ***spam*** or ***not_spam***. It is intended to be used as a first-layer spam filter for email systems, messaging applications or any other text-based communication platform.
+
+The following categories are considered spam:
+
+1. Unsolicited commercial advertisement or non-commercial proselytizing.
+2. Fraudulent schemes. including get-rich-quick and pyramid schemes.
+3. Phishing attempts. unrealistic offers or announcements.
+4. Content with deceptive or misleading information.
+5. Malware or harmful links.
+6. Adult content or explicit material.
+7. Excessive use of capitalization or punctuation to grab attention.
+
+## Why you should use this model
+
+This model can be used 100% locally on CPU. No processing is done in the cloud, and no data is sent to any third party. This makes it ideal for applications where privacy is a concern, or where internet connectivity is limited.
+
+<!-- ## Languages
+
+The main model language is English, but we have spam detection models specialized in other languages as well:
+- Spanish: [https://huggingface.co/tanaos/tanaos-spam-detection-spanish](https://huggingface.co/tanaos/tanaos-spam-detection-spanish) -->
+
+## How to Use
+
+1. Sign up for a free account at [https://platform.tanaos.com/](https://platform.tanaos.com/)
+2. Create a free API Key from the [API Keys section](https://platform.tanaos.com/profile/api-keys)
+3. Download the [`tanaos_spam_detection_english-0.1.0-py3-none-any.whl` file](https://huggingface.co/tanaos/tanaos-spam-detection-v1/resolve/main/tanaos_spam_detection_english-0.1.0-py3-none-any.whl?download=true)
+4. Install the model
+    ```bash
+    uv init
+    uv add tanaos_spam_detection_english-0.1.0-py3-none-any.whl
+    #--- Linux/MacOS ---
+    source .venv/bin/activate
+    #--- Windows ---
+    .venv\Scripts\activate
+    ```
+5. Warm up the model for faster inference (optional but recommended):
+    ```bash
+    python -m tanaos_spam_detection_english --api-key <YOUR_API_KEY> --serve
+    ```
+    This will start a local IPC server that serves the model. Inference will be faster for as long as the server is running. Simply type CTRL+C to stop the local server when you're done.
+6. Use the model for inference:
+    ```python
+    from tanaos_spam_detection_english import run_inference
+
+    result = run_inference(
+        text="You won an IPhone 16! Click here to claim your prize.",
+        api_key="<YOUR_API_KEY>"
+    )
+    print(result)
+
+    # >>> [{'label': 'spam', 'confidence': 0.9975}]
+    ```
+
+## Intended Uses
+
+This model is intended to:
+- Serve as a first-layer spam filter for email systems, messaging applications, or any other text-based communication platform.
+- Help reduce unwanted or harmful messages by classifying text as spam or not spam.
+
+Not intended for:
+- Use in high-stakes scenarios where misclassification could lead to significant consequences without further human review.

config.json

@@ -0,0 +1,33 @@
+{
+  "activation": "gelu",
+  "architectures": [
+    "DistilBertForSequenceClassification"
+  ],
+  "attention_dropout": 0.1,
+  "dim": 768,
+  "dropout": 0.1,
+  "dtype": "float32",
+  "hidden_dim": 3072,
+  "id2label": {
+    "0": "not_spam",
+    "1": "spam"
+  },
+  "initializer_range": 0.02,
+  "label2id": {
+    "not_spam": 0,
+    "spam": 1
+  },
+  "max_position_embeddings": 512,
+  "model_type": "distilbert",
+  "n_heads": 12,
+  "n_layers": 6,
+  "output_past": true,
+  "pad_token_id": 0,
+  "problem_type": "single_label_classification",
+  "qa_dropout": 0.1,
+  "seq_classif_dropout": 0.2,
+  "sinusoidal_pos_embds": false,
+  "tie_weights_": true,
+  "transformers_version": "4.57.3",
+  "vocab_size": 119547
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:55f90cf08bd50f0bdce1f545953ed9f6c94f9fe8bf989d591a1b880eb639917a
+size 541317368

notebook.ipynb

@@ -0,0 +1,59 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "id": "b71a1322",
+   "metadata": {},
+   "source": [
+    "# Get started with `tanaos-spam-detection-v1`"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "23bacd31",
+   "metadata": {},
+   "source": [
+    "Use this model for free via the [Tanaos API](https://tanaos.com/) in 3 simple steps:\n",
+    "\n",
+    "1. Sign up for a free account at [https://platform.tanaos.com/](https://platform.tanaos.com/)\n",
+    "2. Create a free API Key from the [API Keys section](https://platform.tanaos.com/profile/api-keys)\n",
+    "3. Replace `<YOUR_API_KEY>` in the code below with your API Key and use this snippet:\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "814ab3a8",
+   "metadata": {
+    "vscode": {
+     "languageId": "plaintext"
+    }
+   },
+   "outputs": [],
+   "source": [
+    "import requests\n",
+    "\n",
+    "session = requests.Session()\n",
+    "\n",
+    "sd_out = session.post(\n",
+    "    \"https://slm.tanaos.com/models/spam-detection\",\n",
+    "    headers={\n",
+    "        \"X-API-Key\": \"<YOUR_API_KEY>\",\n",
+    "    },\n",
+    "    json={\n",
+    "        \"text\": \"You won an IPhone 16! Click here to claim your prize.\"\n",
+    "    }\n",
+    ")\n",
+    "\n",
+    "print(sd_out.json()[\"data\"])"
+   ]
+  }
+ ],
+ "metadata": {
+  "language_info": {
+   "name": "python"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}

special_tokens_map.json

@@ -0,0 +1,7 @@
+{
+  "cls_token": "[CLS]",
+  "mask_token": "[MASK]",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "unk_token": "[UNK]"
+}

tokenizer_config.json

@@ -0,0 +1,56 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "[PAD]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "100": {
+      "content": "[UNK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "101": {
+      "content": "[CLS]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "102": {
+      "content": "[SEP]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "103": {
+      "content": "[MASK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "clean_up_tokenization_spaces": false,
+  "cls_token": "[CLS]",
+  "do_lower_case": false,
+  "extra_special_tokens": {},
+  "mask_token": "[MASK]",
+  "model_max_length": 512,
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "strip_accents": null,
+  "tokenize_chinese_chars": true,
+  "tokenizer_class": "DistilBertTokenizer",
+  "unk_token": "[UNK]"
+}