Upload folder using huggingface_hub

.gitattributes

@@ -33,3 +33,4 @@ 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
+tokenizer.json filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,97 @@
+
+# Llama-3 8B Prompt Injection Detection (LoRA Fine-Tune)
+
+This model is a fine-tuned version of **Meta-Llama-3-8B** designed to detect prompt injection attacks. 
+It has been trained using QLoRA (4-bit quantization) on a massive aggregation of public prompt injection datasets.
+
+## Model Details
+- **Base Model:** meta-llama/Meta-Llama-3-8B
+- **Task:** Binary Classification (SAFE / INJECTION)
+- **Fine-Tuning Method:** LoRA (Low-Rank Adaptation)
+
+## Dataset
+
+The model was trained on a deduplicated aggregation of multiple open-source datasets.
+
+- **Total Unique Training Examples:** 478638
+
+### Data Sources
+| Dataset Name                                   |   Original Rows |
+|:-----------------------------------------------|----------------:|
+| neuralchemy/Prompt-injection-dataset           |           10674 |
+| hendzh/PromptShield                            |           18909 |
+| S-Labs/prompt-injection-dataset                |           11089 |
+| wambosec/prompt-injections-subtle              |             839 |
+| wambosec/prompt-injections                     |            5189 |
+| beratcmn/turkish-prompt-injections             |             546 |
+| Antijection/prompt-injection-dataset-v1        |            5988 |
+| guychuk/benign-malicious-prompt-classification |          464470 |
+| m4vic/prompt-injection-dataset                 |           10674 |
+
+## Training Hyperparameters
+
+- **Max Steps:** 500
+- **Learning Rate:** 2e-4
+- **Batch Size:** 4 (per device)
+- **Gradient Accumulation:** 4
+- **Precision:** bfloat16 (bf16)
+- **LoRA Rank (r):** 16
+- **LoRA Alpha:** 32
+- **LoRA Dropout:** 0.05
+- **Target Modules:** q_proj, k_proj, v_proj, o_proj
+
+## Training Results
+
+### Loss Curve
+![Loss Curve](loss_curve_massive.png)
+
+## Usage
+
+```python
+import torch
+from transformers import AutoTokenizer, AutoModelForCausalLM, BitsAndBytesConfig
+from peft import PeftModel
+
+model_id = "meta-llama/Meta-Llama-3-8B"
+adapter_path = "path_to_saved_adapter" # e.g., llama3_injection_adapter_massive
+
+# 1. Load Base Model
+bnb_config = BitsAndBytesConfig(
+    load_in_4bit=True,
+    bnb_4bit_compute_dtype=torch.bfloat16,
+    bnb_4bit_use_double_quant=True,
+    bnb_4bit_quant_type="nf4"
+)
+
+base_model = AutoModelForCausalLM.from_pretrained(
+    model_id,
+    quantization_config=bnb_config,
+    device_map="auto"
+)
+
+# 2. Load Adapter
+model = PeftModel.from_pretrained(base_model, adapter_path)
+tokenizer = AutoTokenizer.from_pretrained(model_id)
+
+# 3. Inference
+def predict(text):
+    prompt = (
+        f"<|begin_of_text|><|start_header_id|>system<|end_header_id|>
+
+"
+        f"Classify this prompt as SAFE or INJECTION.<|eot_id|>"
+        f"<|start_header_id|>user<|end_header_id|>
+
+"
+        f"{text}<|eot_id|>"
+        f"<|start_header_id|>assistant<|end_header_id|>
+
+"
+    )
+    
+    inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
+    outputs = model.generate(**inputs, max_new_tokens=10)
+    return tokenizer.decode(outputs[0], skip_special_tokens=True).split("assistant")[-1].strip()
+
+print(predict("Write a poem about flowers."))
+```

adapter_config.json

@@ -0,0 +1,43 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "meta-llama/Meta-Llama-3-8B",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.18.1",
+  "qalora_group_size": 16,
+  "r": 16,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "q_proj",
+    "v_proj",
+    "k_proj",
+    "o_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}

adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:04c4e14d8c10a15daf5da55d4fab5f4616b650aa99c5f686c5c935823392138a
+size 27297544

tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3c5cf44023714fb39b05e71e425f8d7b92805ff73f7988b083b8c87f0bf87393
+size 17209961

tokenizer_config.json

@@ -0,0 +1,14 @@
+{
+  "backend": "tokenizers",
+  "bos_token": "<|begin_of_text|>",
+  "clean_up_tokenization_spaces": true,
+  "eos_token": "<|end_of_text|>",
+  "is_local": false,
+  "model_input_names": [
+    "input_ids",
+    "attention_mask"
+  ],
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "<|end_of_text|>",
+  "tokenizer_class": "TokenizersBackend"
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d495d77d378b89866100df4dae6793ca6ef6adc4b788b8905461e97676554634
+size 5585