AndrewAndrewsen
/

distilbert-secret-masker

@@ -1,331 +1,335 @@
----
-language:
-  - en
-license: apache-2.0
-tags:
-  - secret-detection
-  - token-classification
-  - ner
-  - security
-  - api-keys
-  - credentials
-  - distilbert
-library_name: transformers
-pipeline_tag: token-classification
-datasets:
-  - custom
-metrics:
-  - f1
-  - precision
-  - recall
-model-index:
-  - name: distilbert-secret-masker
-    results:
-      - task:
-          type: token-classification
-          name: Secret Detection
-        dataset:
-          name: SecretMask v2 (600 test examples)
-          type: custom
-        metrics:
-          - type: f1
-            value: 0.52
-            name: F1 Score
-            verified: true
-          - type: precision
-            value: 0.82
-            name: Precision
-            verified: true
-          - type: recall
-            value: 0.38
-            name: Recall
-            verified: true
-base_model: distilbert-base-uncased
----
-# DistilBERT Secret Masker (Fast Expert)
-[![License: Apache 2.0](https://img.shields.io/badge/License-Apache%202.0-blue.svg)](https://opensource.org/licenses/Apache-2.0)
-[![Model Size](https://img.shields.io/badge/Size-265MB-green.svg)](https://huggingface.co/AndrewAndrewsen/distilbert-secret-masker)
-[![Inference Speed](https://img.shields.io/badge/Latency-6ms-brightgreen.svg)](https://huggingface.co/AndrewAndrewsen/distilbert-secret-masker)
-**Fast Expert** model for SecMask MoE system - specialized for rapid secret detection in short to medium-length texts (≤512 tokens).
----
-## 🎯 Overview
-Fine-tuned DistilBERT model for detecting and classifying secrets (API keys, tokens, credentials) in text using Named Entity Recognition (NER). Serves as the **Fast Expert** in the [SecMask Mixture of Experts architecture](https://github.com/AndrewAndrewsen/secmask), handling **92.7%** of inference requests with ~6ms latency.
-### Key Features
-✅ **High Speed**: ~6ms average latency
-✅ **High Precision**: 82% (low false positives)
-✅ **Production Ready**: Handles 92.7% of real-world cases
-✅ **Lightweight**: 265MB (66M parameters)
-✅ **Multi-Secret Types**: AWS keys, GitHub tokens, JWTs, API keys, PEM blocks, K8s secrets
-> **Note**: Combines NER model (82% precision, 38% recall) with deterministic post-filters for PEM blocks and K8s secrets to achieve production-grade coverage.
-### Detected Secret Types
-| Secret Type                   | Example Pattern                   | F1 Score |
-| ----------------------------- | --------------------------------- | -------- |
-| **AWS Access Keys**           | `AKIA...`                         | 0.92     |
-| **GitHub Personal Tokens**    | `ghp_...`, `gho_...`              | 0.88     |
-| **JWT Tokens**                | `eyJ0eXAiOiJKV1QiLCJhbGc...`      | 0.85     |
-| **Generic API Keys**          | `sk-proj-...`, `api_key=...`      | 0.79     |
-| **PEM Certificate Blocks**    | `-----BEGIN PRIVATE KEY-----`     | 0.95     |
-| **Kubernetes Secrets (data)** | `kind: Secret` → `data:` values   | 0.81     |
-| **Database Credentials**      | Connection strings with passwords | 0.74     |
----
-## 🚀 Quick Start
-### Installation
-```bash
-pip install transformers torch
-```
-### Basic Usage
-**Standalone (Direct):**
-```python
-from transformers import pipeline
-# Load model
-classifier = pipeline(
-    "token-classification",
-    model="AndrewAndrewsen/distilbert-secret-masker",
-    aggregation_strategy="simple"
-)
-# Detect secrets
-text = "My API key is sk-proj-1234567890abcdefghijklmnopqrstuvwxyz"
-results = classifier(text)
-print(results)
-# [{'entity_group': 'SECRET', 'score': 0.95, 'word': 'sk-proj-1234567890abcdefghijklmnopqrstuvwxyz', ...}]
-```
-**Recommended (via SecMask MoE):**
-```python
-# Clone SecMask repo
-# git clone https://github.com/AndrewAndrewsen/secmask.git
-from infer_moe import mask_text_moe
-masked = mask_text_moe(
-    "My GitHub token is ghp_1234567890abcdefghijklmnopqrstuvwxyz",
-    fast_model_dir="AndrewAndrewsen/distilbert-secret-masker",
-    tau=0.80,
-    routing_mode="heuristic"
-)
-print(masked)
-# "My GitHub token is [SECRET]"
-```
-### Command Line (via SecMask)
-```bash
-# Clone repo
-git clone https://github.com/AndrewAndrewsen/secmask.git
-cd secmask
-# Mask secrets
-python infer_moe.py \
-    --text "AWS key: AKIAIOSFODNN7EXAMPLE" \
-    --fast-model AndrewAndrewsen/distilbert-secret-masker \
-    --routing-mode heuristic \
-    --tau 0.80
-# Output: AWS key: [SECRET]
-```
----
-## 📊 Performance
-### Secret Detection Metrics
-| Metric          | Value  |
-| --------------- | ------ |
-| **F1 Score**    | 0.52   |
-| **Precision**   | 0.82   |
-| **Recall**      | 0.38   |
-| **Avg Latency** | 6ms    |
-| **P99 Latency** | 57.4ms |
-> **Note**: Evaluated at τ=0.80 threshold with post-filtering (PEM blocks, K8s secrets). The model prioritizes precision (low false positives) over recall. For higher recall, lower the threshold or combine with deterministic filters.
-### When This Model Is Used (MoE Routing)
-The router selects this **Fast Expert** when:
-- Token count ≤ 512
-- No multi-line structures (PEM blocks, K8s YAML)
-- Simple text patterns
-- **Coverage: 92.7%** of real-world requests
-For longer or more complex texts, the MoE router automatically switches to the [Longformer Secret Masker](https://huggingface.co/AndrewAndrewsen/longformer-secret-masker).
----
-## 🏗️ Model Details
-### Architecture
-- **Base Model**: [`distilbert-base-uncased`](https://huggingface.co/distilbert-base-uncased) (66M params, Apache 2.0)
-- **Task**: Token Classification (NER)
-- **Max Sequence Length**: 512 tokens
-- **Label**: `B-SECRET`, `I-SECRET`, `O` (BIO tagging)
-### Training Details
-- **Dataset**: Custom SecretMask v2 (6,000 training examples)
-- **Optimizer**: AdamW (lr=5e-5)
-- **Epochs**: 3
-- **Batch Size**: 16
-- **Hardware**: GPU (NVIDIA A100 or equivalent)
-- **Training Time**: ~30 minutes
-### Evaluation
-Evaluated on 600 held-out examples from SecretMask v2 test set:
-```
-Precision: 0.82
-Recall: 0.38
-F1: 0.52
-Support: 1,021 secret tokens
-```
-**Key Insights**:
-- **High Precision (82%)**: Very low false positive rate - safe for production
-- **Lower Recall (38%)**: Misses some secrets when used standalone
-- **Production Strategy**: Combine with deterministic filters (see `filters.py`) for PEM blocks, K8s secrets, and AWS patterns to achieve >90% coverage
-- **Threshold Tuning**: Lower τ from 0.80 to 0.50 for higher recall (trade-off: more false positives)
----
-## 💡 Use Cases
-### Production Applications
-1. **Pre-Commit Hooks** - Prevent secrets in git commits
-2. **CI/CD Pipelines** - Scan code before deployment
-3. **Log Sanitization** - Remove secrets from application logs
-4. **API Response Filtering** - Mask secrets in debug output
-5. **Documentation Cleanup** - Sanitize before open-sourcing
-6. **Security Audits** - Scan codebases for exposed credentials
-### Example: Pre-Commit Hook
-```python
-# .git/hooks/pre-commit
-from transformers import pipeline
-classifier = pipeline("token-classification", model="AndrewAndrewsen/distilbert-secret-masker")
-for file in staged_files:
-    content = read_file(file)
-    secrets = classifier(content)
-    if secrets:
-        print(f"❌ Secret detected in {file}!")
-        exit(1)
-```
-See [SecMask Examples](https://github.com/AndrewAndrewsen/secmask/blob/main/EXAMPLES.md) for more.
----
-## ⚠️ Limitations
-### Known Issues
-1. **Token Limit**: Cannot handle texts >512 tokens (use Longformer expert)
-2. **English Only**: Trained on English text
-3. **False Negatives**: ~25% recall means some secrets may be missed
-4. **Context Sensitivity**: May struggle with unusual formatting
-5. **Novel Patterns**: May miss new secret types not in training data
-### Not Suitable For
-❌ Non-English text
-❌ Binary data or encrypted content
-❌ Images/PDFs (extract text first)
-❌ Very long documents (use [longformer-secret-masker](https://huggingface.co/AndrewAndrewsen/longformer-secret-masker))
-❌ Real-time streaming (consider batching)
-### Recommended Mitigations
-- **Combine with filters**: Use deterministic filters for PEM blocks, K8s secrets (see [SecMask filters](https://github.com/AndrewAndrewsen/secmask/blob/main/filters.py))
-- **Adjust threshold**: Lower `tau` for higher recall (more false positives)
-- **Use MoE system**: Automatic routing to appropriate expert
-- **Add regex patterns**: Supplement with custom patterns for your use case
----
-## 📜 License & Attribution
-### Model License
-**Apache 2.0** (inherited from `distilbert-base-uncased`)
-### Base Model Attribution
-This model is fine-tuned from:
-- **Model**: [`distilbert-base-uncased`](https://huggingface.co/distilbert-base-uncased)
-- **Authors**: Hugging Face
-- **License**: Apache 2.0
-- **Citation**:
-  ```
-  @inproceedings{sanh2019distilbert,
-    title={DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter},
-    author={Sanh, Victor and Debut, Lysandre and Chaumond, Julien and Wolf, Thomas},
-    booktitle={NeurIPS EMC^2 Workshop},
-    year={2019}
-  }
-  ```
-### SecMask Code License
-The SecMask inference code and training scripts are licensed under **MIT**. See [GitHub repo](https://github.com/AndrewAndrewsen/secmask/blob/main/LICENSE).
----
-## 🔗 Related Models
-| Model                                                                                                        | Size  | Max Tokens | Latency | Use Case                  |
-| ------------------------------------------------------------------------------------------------------------ | ----- | ---------- | ------- | ------------------------- |
-| **[distilbert-secret-masker](https://huggingface.co/AndrewAndrewsen/distilbert-secret-masker)** (this model) | 265MB | 512        | 6ms     | Short texts, fast routing |
-| **[longformer-secret-masker](https://huggingface.co/AndrewAndrewsen/longformer-secret-masker)**              | 592MB | 2048       | 12ms    | Long documents, configs   |
-| **[secretmask-gate](https://huggingface.co/AndrewAndrewsen/secretmask-gate)**                                | 12KB  | N/A        | +0.2ms  | Learned MoE routing       |
----
-## 📚 Resources
-- **GitHub Repository**: [AndrewAndrewsen/secmask](https://github.com/AndrewAndrewsen/secmask)
-- **Documentation**: [README](https://github.com/AndrewAndrewsen/secmask/blob/main/README.md)
-- **Benchmarks**: [BENCHMARKS.md](https://github.com/AndrewAndrewsen/secmask/blob/main/BENCHMARKS.md)
-- **Examples**: [EXAMPLES.md](https://github.com/AndrewAndrewsen/secmask/blob/main/EXAMPLES.md)
-- **Deployment**: [DEPLOYMENT.md](https://github.com/AndrewAndrewsen/secmask/blob/main/DEPLOYMENT.md)
----
-## 🤝 Contributing
-Issues and contributions welcome! See [CONTRIBUTING.md](https://github.com/AndrewAndrewsen/secmask/blob/main/CONTRIBUTING.md).
----
-**Developed by**: Anders Andersson ([@AndrewAndrewsen](https://huggingface.co/AndrewAndrewsen))
-**Part of**: [SecMask MoE System](https://github.com/AndrewAndrewsen/secmask)

+---
+language:
+  - en
+license: apache-2.0
+tags:
+  - secret-detection
+  - token-classification
+  - ner
+  - security
+  - api-keys
+  - credentials
+  - distilbert
+library_name: transformers
+pipeline_tag: token-classification
+datasets:
+  - custom
+metrics:
+  - f1
+  - precision
+  - recall
+model-index:
+  - name: distilbert-secret-masker
+    results:
+      - task:
+          type: token-classification
+          name: Secret Detection
+        dataset:
+          name: SecretMask v2 (600 test examples)
+          type: custom
+        metrics:
+          - type: f1
+            value: 0.52
+            name: F1 Score
+            verified: true
+          - type: precision
+            value: 0.82
+            name: Precision
+            verified: true
+          - type: recall
+            value: 0.38
+            name: Recall
+            verified: true
+base_model: distilbert-base-uncased
+---
+# DistilBERT Secret Masker (Fast Expert)
+[![License: Apache 2.0](https://img.shields.io/badge/License-Apache%202.0-blue.svg)](https://opensource.org/licenses/Apache-2.0)
+[![Model Size](https://img.shields.io/badge/Size-265MB-green.svg)](https://huggingface.co/AndrewAndrewsen/distilbert-secret-masker)
+[![Inference Speed](https://img.shields.io/badge/Latency-6ms-brightgreen.svg)](https://huggingface.co/AndrewAndrewsen/distilbert-secret-masker)
+**Fast Expert** model for SecMask MoE system - specialized for rapid secret detection in short to medium-length texts (≤512 tokens).
+---
+## 🎯 Overview
+Fine-tuned DistilBERT model for detecting and classifying secrets (API keys, tokens, credentials) in text using Named Entity Recognition (NER). Serves as the **Fast Expert** in the [SecMask Mixture of Experts architecture](https://github.com/AndrewAndrewsen/secmask), handling **92.7%** of inference requests with ~6ms latency.
+### Key Features
+✅ **High Speed**: 11ms P50 latency on CPU
+✅ **High Precision**: 82% (NER-only), **92.3% with filters**
+✅ **Production Ready**: Handles 92.7% of real-world cases
+✅ **Lightweight**: 265MB (66M parameters)
+✅ **Multi-Secret Types**: AWS keys, GitHub tokens, JWTs, API keys, PEM blocks, K8s secrets
+> **Production Performance**: When combined with post-processing filters (PEM blocks, K8s secrets, pattern matching), achieves **92.3% precision, 80% recall, F1: 0.857**. The NER model alone achieves 82% precision and 38% recall. See [comprehensive benchmarks](https://github.com/AndrewAndrewsen/secmask/blob/main/BENCHMARK_RESULTS.md) for details.
+> **Recommended Configuration**: Fast Expert + Filters (this model with post-processing) is the recommended production setup, outperforming Full MoE configurations. See [Configuration Guide](https://github.com/AndrewAndrewsen/secmask/blob/main/CONFIGURATION_GUIDE.md) for usage recommendations.
+### Detected Secret Types
+| Secret Type                   | Example Pattern                   | F1 Score |
+| ----------------------------- | --------------------------------- | -------- |
+| **AWS Access Keys**           | `AKIA...`                         | 0.92     |
+| **GitHub Personal Tokens**    | `ghp_...`, `gho_...`              | 0.88     |
+| **JWT Tokens**                | `eyJ0eXAiOiJKV1QiLCJhbGc...`      | 0.85     |
+| **Generic API Keys**          | `sk-proj-...`, `api_key=...`      | 0.79     |
+| **PEM Certificate Blocks**    | `-----BEGIN PRIVATE KEY-----`     | 0.95     |
+| **Kubernetes Secrets (data)** | `kind: Secret` → `data:` values   | 0.81     |
+| **Database Credentials**      | Connection strings with passwords | 0.74     |
+---
+## 🚀 Quick Start
+### Installation
+```bash
+pip install transformers torch
+```
+### Basic Usage
+**Standalone (Direct):**
+```python
+from transformers import pipeline
+# Load model
+classifier = pipeline(
+    "token-classification",
+    model="AndrewAndrewsen/distilbert-secret-masker",
+    aggregation_strategy="simple"
+)
+# Detect secrets
+text = "My API key is sk-proj-1234567890abcdefghijklmnopqrstuvwxyz"
+results = classifier(text)
+print(results)
+# [{'entity_group': 'SECRET', 'score': 0.95, 'word': 'sk-proj-1234567890abcdefghijklmnopqrstuvwxyz', ...}]
+```
+**Recommended (via SecMask MoE):**
+```python
+# Clone SecMask repo
+# git clone https://github.com/AndrewAndrewsen/secmask.git
+from infer_moe import mask_text_moe
+masked = mask_text_moe(
+    "My GitHub token is ghp_1234567890abcdefghijklmnopqrstuvwxyz",
+    fast_model_dir="AndrewAndrewsen/distilbert-secret-masker",
+    tau=0.80,
+    routing_mode="heuristic"
+)
+print(masked)
+# "My GitHub token is [SECRET]"
+```
+### Command Line (via SecMask)
+```bash
+# Clone repo
+git clone https://github.com/AndrewAndrewsen/secmask.git
+cd secmask
+# Mask secrets
+python infer_moe.py \
+    --text "AWS key: AKIAIOSFODNN7EXAMPLE" \
+    --fast-model AndrewAndrewsen/distilbert-secret-masker \
+    --routing-mode heuristic \
+    --tau 0.80
+# Output: AWS key: [SECRET]
+```
+---
+## 📊 Performance
+### Secret Detection Metrics
+| Metric          | NER Only | With Filters (Recommended) |
+| --------------- | -------- | -------------------------- |
+| **F1 Score**    | 0.52     | **0.857**                  |
+| **Precision**   | 82%      | **92.3%**                  |
+| **Recall**      | 38%      | **80.0%**                  |
+| **P50 Latency** | 11ms     | 11ms                       |
+| **P90 Latency** | 14ms     | 14ms                       |
+| **P99 Latency** | 17ms     | 17ms                       |
+| **Throughput**  | 84 req/s | 84 req/s (CPU)            |
+> **Note**: NER-only metrics measured at τ=0.80. Production systems combine NER with post-processing filters (PEM blocks, K8s secrets, pattern matching) to achieve 92.3% precision and 80% recall. Post-processing adds no latency overhead. See [BENCHMARK_RESULTS.md](https://github.com/AndrewAndrewsen/secmask/blob/main/BENCHMARK_RESULTS.md) for comprehensive benchmarks.
+### When This Model Is Used (MoE Routing)
+The router selects this **Fast Expert** when:
+- Token count ≤ 512
+- No multi-line structures (PEM blocks, K8s YAML)
+- Simple text patterns
+- **Coverage: 92.7%** of real-world requests
+> **Note**: The recommended production configuration is **Fast Expert + Filters** alone (without the Long Expert). This achieves better results than Full MoE. See [Configuration Guide](https://github.com/AndrewAndrewsen/secmask/blob/main/CONFIGURATION_GUIDE.md) for details.
+---
+## 🏗️ Model Details
+### Architecture
+- **Base Model**: [`distilbert-base-uncased`](https://huggingface.co/distilbert-base-uncased) (66M params, Apache 2.0)
+- **Task**: Token Classification (NER)
+- **Max Sequence Length**: 512 tokens
+- **Label**: `B-SECRET`, `I-SECRET`, `O` (BIO tagging)
+### Training Details
+- **Dataset**: Custom SecretMask v2 (6,000 training examples)
+- **Optimizer**: AdamW (lr=5e-5)
+- **Epochs**: 3
+- **Batch Size**: 16
+- **Hardware**: GPU (NVIDIA A100 or equivalent)
+- **Training Time**: ~30 minutes
+### Evaluation
+Evaluated on 600 held-out examples from SecretMask v2 test set:
+```
+Precision: 0.82
+Recall: 0.38
+F1: 0.52
+Support: 1,021 secret tokens
+```
+**Key Insights**:
+- **High Precision (82%)**: Very low false positive rate - safe for production
+- **Lower Recall (38%)**: Misses some secrets when used standalone
+- **Production Strategy**: Combine with deterministic filters (see `filters.py`) for PEM blocks, K8s secrets, and AWS patterns to achieve >90% coverage
+- **Threshold Tuning**: Lower τ from 0.80 to 0.50 for higher recall (trade-off: more false positives)
+---
+## 💡 Use Cases
+### Production Applications
+1. **Pre-Commit Hooks** - Prevent secrets in git commits
+2. **CI/CD Pipelines** - Scan code before deployment
+3. **Log Sanitization** - Remove secrets from application logs
+4. **API Response Filtering** - Mask secrets in debug output
+5. **Documentation Cleanup** - Sanitize before open-sourcing
+6. **Security Audits** - Scan codebases for exposed credentials
+### Example: Pre-Commit Hook
+```python
+# .git/hooks/pre-commit
+from transformers import pipeline
+classifier = pipeline("token-classification", model="AndrewAndrewsen/distilbert-secret-masker")
+for file in staged_files:
+    content = read_file(file)
+    secrets = classifier(content)
+    if secrets:
+        print(f"❌ Secret detected in {file}!")
+        exit(1)
+```
+See [SecMask Examples](https://github.com/AndrewAndrewsen/secmask/blob/main/EXAMPLES.md) for more.
+---
+## ⚠️ Limitations
+### Known Issues
+1. **Token Limit**: Cannot handle texts >512 tokens (use Longformer expert)
+2. **English Only**: Trained on English text
+3. **False Negatives**: ~25% recall means some secrets may be missed
+4. **Context Sensitivity**: May struggle with unusual formatting
+5. **Novel Patterns**: May miss new secret types not in training data
+### Not Suitable For
+❌ Non-English text
+❌ Binary data or encrypted content
+❌ Images/PDFs (extract text first)
+❌ Very long documents (use [longformer-secret-masker](https://huggingface.co/AndrewAndrewsen/longformer-secret-masker))
+❌ Real-time streaming (consider batching)
+### Recommended Mitigations
+- **Combine with filters**: Use deterministic filters for PEM blocks, K8s secrets (see [SecMask filters](https://github.com/AndrewAndrewsen/secmask/blob/main/filters.py))
+- **Adjust threshold**: Lower `tau` for higher recall (more false positives)
+- **Use MoE system**: Automatic routing to appropriate expert
+- **Add regex patterns**: Supplement with custom patterns for your use case
+---
+## 📜 License & Attribution
+### Model License
+**Apache 2.0** (inherited from `distilbert-base-uncased`)
+### Base Model Attribution
+This model is fine-tuned from:
+- **Model**: [`distilbert-base-uncased`](https://huggingface.co/distilbert-base-uncased)
+- **Authors**: Hugging Face
+- **License**: Apache 2.0
+- **Citation**:
+  ```
+  @inproceedings{sanh2019distilbert,
+    title={DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter},
+    author={Sanh, Victor and Debut, Lysandre and Chaumond, Julien and Wolf, Thomas},
+    booktitle={NeurIPS EMC^2 Workshop},
+    year={2019}
+  }
+  ```
+### SecMask Code License
+The SecMask inference code and training scripts are licensed under **MIT**. See [GitHub repo](https://github.com/AndrewAndrewsen/secmask/blob/main/LICENSE).
+---
+## 🔗 Related Models
+| Model                                                                                                        | Size  | Max Tokens | Latency | Use Case                  |
+| ------------------------------------------------------------------------------------------------------------ | ----- | ---------- | ------- | ------------------------- |
+| **[distilbert-secret-masker](https://huggingface.co/AndrewAndrewsen/distilbert-secret-masker)** (this model) | 265MB | 512        | 6ms     | Short texts, fast routing |
+| **[longformer-secret-masker](https://huggingface.co/AndrewAndrewsen/longformer-secret-masker)**              | 592MB | 2048       | 12ms    | Long documents, configs   |
+| **[secretmask-gate](https://huggingface.co/AndrewAndrewsen/secretmask-gate)**                                | 12KB  | N/A        | +0.2ms  | Learned MoE routing       |
+---
+## 📚 Resources
+- **GitHub Repository**: [AndrewAndrewsen/secmask](https://github.com/AndrewAndrewsen/secmask)
+- **Documentation**: [README](https://github.com/AndrewAndrewsen/secmask/blob/main/README.md)
+- **Benchmarks**: [BENCHMARKS.md](https://github.com/AndrewAndrewsen/secmask/blob/main/BENCHMARKS.md)
+- **Examples**: [EXAMPLES.md](https://github.com/AndrewAndrewsen/secmask/blob/main/EXAMPLES.md)
+- **Deployment**: [DEPLOYMENT.md](https://github.com/AndrewAndrewsen/secmask/blob/main/DEPLOYMENT.md)
+---
+## 🤝 Contributing
+Issues and contributions welcome! See [CONTRIBUTING.md](https://github.com/AndrewAndrewsen/secmask/blob/main/CONTRIBUTING.md).
+---
+**Developed by**: Anders Andersson ([@AndrewAndrewsen](https://huggingface.co/AndrewAndrewsen))
+**Part of**: [SecMask MoE System](https://github.com/AndrewAndrewsen/secmask)