Upload 9 files

config.json

@@ -0,0 +1,46 @@
+{
+  "architectures": [
+    "SiglipForImageClassification"
+  ],
+  "id2label": {
+    "0": "Fake",
+    "1": "Real"
+  },
+  "initializer_factor": 1.0,
+  "label2id": {
+    "Fake": 0,
+    "Real": 1
+  },
+  "model_type": "siglip",
+  "problem_type": "single_label_classification",
+  "text_config": {
+    "attention_dropout": 0.0,
+    "hidden_act": "gelu_pytorch_tanh",
+    "hidden_size": 768,
+    "intermediate_size": 3072,
+    "layer_norm_eps": 1e-06,
+    "max_position_embeddings": 64,
+    "model_type": "siglip_text_model",
+    "num_attention_heads": 12,
+    "num_hidden_layers": 12,
+    "projection_size": 768,
+    "torch_dtype": "float32",
+    "vocab_size": 256000
+  },
+  "torch_dtype": "float32",
+  "transformers_version": "4.50.0",
+  "vision_config": {
+    "attention_dropout": 0.0,
+    "hidden_act": "gelu_pytorch_tanh",
+    "hidden_size": 768,
+    "image_size": 224,
+    "intermediate_size": 3072,
+    "layer_norm_eps": 1e-06,
+    "model_type": "siglip_vision_model",
+    "num_attention_heads": 12,
+    "num_channels": 3,
+    "num_hidden_layers": 12,
+    "patch_size": 16,
+    "torch_dtype": "float32"
+  }
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7db42002064e2722f38d3ad5c42441a9a4cfc5da8f6dbe2ff92ab9055a2d09eb
+size 371567992

optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5fe1ea2babd281d0538c4ce855dfca04962c8ddec3e93bb7aeb40973c8172762
+size 686555770

preprocessor_config.json

@@ -0,0 +1,24 @@
+{
+  "do_convert_rgb": null,
+  "do_normalize": true,
+  "do_rescale": true,
+  "do_resize": true,
+  "image_mean": [
+    0.5,
+    0.5,
+    0.5
+  ],
+  "image_processor_type": "SiglipImageProcessor",
+  "image_std": [
+    0.5,
+    0.5,
+    0.5
+  ],
+  "processor_class": "SiglipProcessor",
+  "resample": 2,
+  "rescale_factor": 0.00392156862745098,
+  "size": {
+    "height": 224,
+    "width": 224
+  }
+}

rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:88acce3cfb3e8de5af4f824e7ecfcd91bf0344fb9c417988b4d078e64bcaa34a
+size 14244

scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e7a07120c819f89245e4dc484312c0ddc4f1ffbcf3788890e1ed108cac2bcf04
+size 1064

testing.py

@@ -0,0 +1,43 @@
+import streamlit as st
+from transformers import AutoImageProcessor, SiglipForImageClassification
+from PIL import Image
+import torch
+
+# -------------------------------------------------------
+# Load model once at startup
+# -------------------------------------------------------
+model_path = r"C:\Users\Sharulatha\Documents\hackathon\deepfake-detector-model-v1\checkpoint-625"
+
+st.title("🕵️ Deepfake Detector")
+
+@st.cache_resource
+def load_model():
+    model = SiglipForImageClassification.from_pretrained(model_path)
+    processor = AutoImageProcessor.from_pretrained(model_path)
+    return model, processor
+
+model, processor = load_model()
+
+# Label mapping
+id2label = {"0": "fake", "1": "real"}
+
+# -------------------------------------------------------
+# File uploader
+# -------------------------------------------------------
+uploaded_file = st.file_uploader("Upload an image", type=["jpg", "jpeg", "png"])
+
+if uploaded_file is not None:
+    image = Image.open(uploaded_file).convert("RGB")
+    st.image(image, caption="Uploaded Image", use_column_width=True)
+
+    # Preprocess and predict
+    inputs = processor(images=image, return_tensors="pt")
+
+    with torch.no_grad():
+        outputs = model(**inputs)
+        probs = torch.nn.functional.softmax(outputs.logits, dim=1).squeeze().tolist()
+
+    prediction = {id2label[str(i)]: round(probs[i], 3) for i in range(len(probs))}
+
+    st.subheader("Prediction")
+    st.json(prediction)

trainer_state.json

@@ -0,0 +1,51 @@
+{
+  "best_global_step": 625,
+  "best_metric": 0.1934153288602829,
+  "best_model_checkpoint": "siglip2-finetune-full/checkpoint-625",
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 625,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.8,
+      "grad_norm": 16.568904876708984,
+      "learning_rate": 6.249999999999999e-07,
+      "loss": 0.3408,
+      "step": 500
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.9169958497924896,
+      "eval_loss": 0.1934153288602829,
+      "eval_model_preparation_time": 0.0027,
+      "eval_runtime": 479.9201,
+      "eval_samples_per_second": 41.672,
+      "eval_steps_per_second": 5.209,
+      "step": 625
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 1250,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.6750261609304187e+18,
+  "train_batch_size": 32,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e8fe67ed70f3fccd73e38b764ae5cbe4db6218d9f6b3fa5d2f439dc363351786
+size 5304