Deploy FastAPI AI vs Human detector

.dockerignore

@@ -0,0 +1,5 @@
+.qodo
+venv/
+test_api.py
+new-file.ipynb
+ai-vs-humanizer2.ipynb

.gitattributes

@@ -33,3 +33,9 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+
+.qodo
+venv/
+test_api.py
+new-file.ipynb
+ai-vs-humanizer2.ipynb

Dockerfile

@@ -0,0 +1,22 @@
+# Use a Python base image, preferably a slim version
+FROM python:3.11-slim
+
+# Set the working directory inside the container
+WORKDIR /app
+
+# Copy the dependency file and install them
+COPY requirements.txt requirements.txt
+RUN pip install --no-cache-dir -r requirements.txt
+
+# Copy your saved model and the application code
+# Assuming your model folder is named 'saved_human_ai_model'
+COPY saved_human_ai_model ./saved_human_ai_model
+COPY ./app.py ./app.py
+
+# Expose the port (Cloud Run uses environment variable $PORT, but 8080 is a good default)
+EXPOSE 8080
+
+# Define the command to run the FastAPI application using Uvicorn
+# The command format is: uvicorn [ASGI_MODULE]:[ASGI_APP_OBJECT] --host 0.0.0.0 --port $PORT
+CMD ["uvicorn", "app:app", "--host", "0.0.0.0", "--port", "8080"] 
+# NOTE: Cloud Run automatically maps the internal port 8080 to the external port

app.py

@@ -0,0 +1,107 @@
+import uvicorn
+from fastapi import FastAPI, HTTPException, Security, Depends
+from fastapi.security import HTTPBearer, HTTPAuthorizationCredentials
+from fastapi.middleware.cors import CORSMiddleware
+from pydantic import BaseModel
+import tensorflow as tf
+from transformers import AutoTokenizer, TFAutoModelForSequenceClassification
+import numpy as np
+import os
+import jwt
+
+# Initialize FastAPI app
+app = FastAPI(title="AI vs Human Detector API")
+
+# CORS Middleware
+app.add_middleware(
+    CORSMiddleware,
+    allow_origins=["*"],  # Allows all origins
+    allow_credentials=True,
+    allow_methods=["*"],  # Allows all methods
+    allow_headers=["*"],  # Allows all headers
+)
+
+# Security
+security = HTTPBearer()
+SECRET_KEY = "mysecretkey"  # In production, use environment variable
+
+# Global variables for model and tokenizer
+model = None
+tokenizer = None
+MODEL_PATH = "saved_human_ai_model"
+
+class PredictionRequest(BaseModel):
+    text: str
+
+class PredictionResponse(BaseModel):
+    label: str
+    confidence: float
+    probabilities: dict
+
+def verify_token(credentials: HTTPAuthorizationCredentials = Security(security)):
+    token = credentials.credentials
+    try:
+        # Just verify the signature using the SECRET_KEY
+        # We don't need to parse/use the payload for user verification
+        jwt.decode(token, SECRET_KEY, algorithms=["HS256"])
+    except jwt.ExpiredSignatureError:
+        raise HTTPException(status_code=401, detail="Token has expired")
+    except jwt.InvalidTokenError:
+        raise HTTPException(status_code=401, detail="Invalid token")
+
+@app.on_event("startup")
+async def load_model():
+    global model, tokenizer
+    try:
+        print(f"Loading model from {MODEL_PATH}...")
+        tokenizer = AutoTokenizer.from_pretrained(MODEL_PATH)
+        model = TFAutoModelForSequenceClassification.from_pretrained(MODEL_PATH)
+        print("Model and Tokenizer loaded successfully!")
+    except Exception as e:
+        print(f"Error loading model: {e}")
+        raise RuntimeError(f"Could not load model: {e}")
+
+@app.post("/predict", response_model=PredictionResponse, dependencies=[Depends(verify_token)])
+async def predict(request: PredictionRequest):
+    if not model or not tokenizer:
+        raise HTTPException(status_code=503, detail="Model not loaded")
+    
+    try:
+        # Tokenize input
+        inputs = tokenizer(
+            request.text,
+            return_tensors="tf",
+            padding=True,
+            truncation=True,
+            max_length=512
+        )
+        
+        # Inference
+        outputs = model(inputs)
+        logits = outputs.logits
+        
+        # Softmax
+        probabilities = tf.nn.softmax(logits, axis=-1).numpy()[0]
+        
+        # Get prediction
+        predicted_class_id = np.argmax(probabilities)
+        confidence = float(probabilities[predicted_class_id])
+        
+        # Map labels (Assuming 0=Human, 1=AI based on notebook)
+        labels_map = {0: "Human", 1: "AI"}
+        predicted_label = labels_map.get(predicted_class_id, "Unknown")
+        
+        return PredictionResponse(
+            label=predicted_label,
+            confidence=confidence,
+            probabilities={
+                "Human": float(probabilities[0]),
+                "AI": float(probabilities[1])
+            }
+        )
+        
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=str(e))
+
+if __name__ == "__main__":
+    uvicorn.run("main:app", host="0.0.0.0", port=8000, reload=True)

requirements.txt

@@ -0,0 +1,8 @@
+fastapi
+uvicorn
+tensorflow
+transformers
+numpy
+tf-keras
+pyjwt
+pydantic

saved_human_ai_model/added_tokens.json

@@ -0,0 +1,3 @@
+{
+  "[MASK]": 128000
+}

saved_human_ai_model/config.json

@@ -0,0 +1,34 @@
+{
+  "architectures": [
+    "DebertaV2ForSequenceClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "layer_norm_eps": 1e-07,
+  "legacy": true,
+  "max_position_embeddings": 512,
+  "max_relative_positions": -1,
+  "model_type": "deberta-v2",
+  "norm_rel_ebd": "layer_norm",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 6,
+  "pad_token_id": 0,
+  "pooler_dropout": 0,
+  "pooler_hidden_act": "gelu",
+  "pooler_hidden_size": 768,
+  "pos_att_type": [
+    "p2c",
+    "c2p"
+  ],
+  "position_biased_input": false,
+  "position_buckets": 256,
+  "relative_attention": true,
+  "share_att_key": true,
+  "transformers_version": "4.57.3",
+  "type_vocab_size": 0,
+  "vocab_size": 128100
+}

saved_human_ai_model/special_tokens_map.json

@@ -0,0 +1,15 @@
+{
+  "bos_token": "[CLS]",
+  "cls_token": "[CLS]",
+  "eos_token": "[SEP]",
+  "mask_token": "[MASK]",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "unk_token": {
+    "content": "[UNK]",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  }
+}

saved_human_ai_model/spm.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c679fbf93643d19aab7ee10c0b99e460bdbc02fedf34b92b05af343b4af586fd
+size 2464616

saved_human_ai_model/tf_model.h5

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ffbac1a30ba77feb1efb6a8874a9ccf768ab1075ec9c868cea2e9f98d4cb599e
+size 567742256

saved_human_ai_model/tokenizer_config.json

@@ -0,0 +1,59 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "[PAD]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "[CLS]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "[SEP]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "[UNK]",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "128000": {
+      "content": "[MASK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "[CLS]",
+  "clean_up_tokenization_spaces": false,
+  "cls_token": "[CLS]",
+  "do_lower_case": false,
+  "eos_token": "[SEP]",
+  "extra_special_tokens": {},
+  "mask_token": "[MASK]",
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "sp_model_kwargs": {},
+  "split_by_punct": false,
+  "tokenizer_class": "DebertaV2Tokenizer",
+  "unk_token": "[UNK]",
+  "vocab_type": "spm"
+}