diff --git a/MAEs/PAG_Avishikta_Bhattacharjee/README.md b/MAEs/PAG_Avishikta_Bhattacharjee/README.md
new file mode 100644
index 0000000..c633e2c
--- /dev/null
+++ b/MAEs/PAG_Avishikta_Bhattacharjee/README.md
@@ -0,0 +1,42 @@
+# Physics-Aware Gating in Transformer for Jet Classification
+
+<p align="center">
+  <img width="788" alt="GSoC Banner" src="https://github.com/user-attachments/assets/c5c91a0b-a0f6-4a92-b992-45d3f5fefaba" />
+</p>
+
+This repository contains the implementation of a **Physics-Aware Gating on Lorentz Particle Transformer (LorentzParT)** designed for self-supervised pre-training and jet reconstruction on high-energy physics datasets like **JetClass**.
+
+The core contribution of this work is a custom attention block that injects a **normalized global invariant mass bias ($m^2$)** alongside a **gated pairwise interaction matrix ($U$)** directly into the attention mechanism, enforcing fundamental Lorentz invariance and relativistic conservation laws.
+
+---
+
+## Description
+
+The existing attention mechanism uses a bias matrix $U$ to incline the transformer toward physics constraints on jet particles, achieving a strong baseline (ROC AUC > 0.90). To evaluate and improve attention across various benchmarks, this repository introduces a **gating mechanism** controlled by physics attention heads to further reduce noise from jet constituent data.
+
+### 1. Reused Physics Bias Attention Head for Gating
+Rather than calculating static physical shifts independently at every Transformer layer, the model utilizes a dedicated **Physics Bias Attention Head**. This module extracts the pairwise $U$-matrix—computed from Minkowski inner products—and fuses it with the global normalized invariant mass bias ($m^2$). 
+
+This physical bias tensor is projected across key heads and **reused dynamically across encoder blocks** to act as a gating mask. By projecting and reusing these learned physical weights directly within the multi-head attention mechanism, the network modulates the query-key matrix ($QK^T$) before value aggregation, filtering out unphysical particle couplings and stabilizing training across dynamic batch shapes.
+
+### 2. Global Invariant Mass Scalar ($m^2$)
+For a jet of $N$ particles, the total invariant mass squared ($m^2$) is computed across aggregate energy and momentum sums:
+
+$$m^2 = \left(\sum_{i=1}^{N} E_i\right)^2 - \left\| \sum_{i=1}^{N} \vec{p}_i \right\|_2^2$$
+
+To ensure numerical stability across varying energy scales, $m^2$ is feature-scaled to form a **normalized invariant mass bias**.
+
+---
+
+## Architecture Overview
+
+The model uses the encoder of LorentzParT embedded within a Variational Autoencoder (VAE) setup:
+1. **Masking:** Random constituent particles are masked during training.
+2. **Encoder:** Processes unmasked tokens through LorentzParT blocks augmented with the **Physics-Aware Gating (PAG)** layer.
+3. **Decoder:** Reconstructs the 4 constituent properties ($p_T, \eta, \phi, E$) of the masked particles.
+
+---
+
+## Evaluation & Results
+
+Will add in the final submission (yet to upload).
diff --git a/MAEs/PAG_Avishikta_Bhattacharjee/notebook/PAG_LorentzParT_.ipynb b/MAEs/PAG_Avishikta_Bhattacharjee/notebook/PAG_LorentzParT_.ipynb
new file mode 100644
index 0000000..f2b0bd1
--- /dev/null
+++ b/MAEs/PAG_Avishikta_Bhattacharjee/notebook/PAG_LorentzParT_.ipynb
@@ -0,0 +1,2860 @@
+{
+  "cells": [
+    {
+      "cell_type": "code",
+      "source": [
+        "\n",
+        "# 1. Clone the ML4SCI/CMS repository\n",
+        "!git clone https://github.com/ML4SCI/CMS.git\n",
+        "\n",
+        "# 2. Navigate to the specific Hybrid Transformer project directory\n",
+        "%cd CMS/MAEs/Hybrid_Transformer_Thanh_Nguyen\n",
+        "\n",
+        "# 3. Install required libraries\n",
+        "!pip install lgatr uproot awkward tqdm vector"
+      ],
+      "metadata": {
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        },
+        "id": "B7YmCrpkJSAg",
+        "outputId": "f6eeec54-0a0e-416f-e730-dcf49f66d6f7"
+      },
+      "id": "B7YmCrpkJSAg",
+      "execution_count": null,
+      "outputs": [
+        {
+          "output_type": "stream",
+          "name": "stdout",
+          "text": [
+            "Cloning into 'CMS'...\n",
+            "remote: Enumerating objects: 751, done.\u001b[K\n",
+            "remote: Counting objects: 100% (130/130), done.\u001b[K\n",
+            "remote: Compressing objects: 100% (99/99), done.\u001b[K\n",
+            "remote: Total 751 (delta 39), reused 83 (delta 27), pack-reused 621 (from 2)\u001b[K\n",
+            "Receiving objects: 100% (751/751), 330.16 MiB | 18.70 MiB/s, done.\n",
+            "Resolving deltas: 100% (187/187), done.\n",
+            "Updating files: 100% (531/531), done.\n",
+            "/content/CMS/MAEs/Hybrid_Transformer_Thanh_Nguyen\n",
+            "Collecting lgatr\n",
+            "  Downloading lgatr-1.4.4-py3-none-any.whl.metadata (8.7 kB)\n",
+            "Collecting uproot\n",
+            "  Downloading uproot-5.7.5-py3-none-any.whl.metadata (35 kB)\n",
+            "Collecting awkward\n",
+            "  Downloading awkward-2.11.0-py3-none-any.whl.metadata (7.6 kB)\n",
+            "Requirement already satisfied: tqdm in /usr/local/lib/python3.12/dist-packages (4.67.3)\n",
+            "Collecting vector\n",
+            "  Downloading vector-1.8.1-py3-none-any.whl.metadata (15 kB)\n",
+            "Requirement already satisfied: torch>=2.1 in /usr/local/lib/python3.12/dist-packages (from lgatr) (2.11.0+cu128)\n",
+            "Requirement already satisfied: numpy in /usr/local/lib/python3.12/dist-packages (from lgatr) (2.0.2)\n",
+            "Requirement already satisfied: einops in /usr/local/lib/python3.12/dist-packages (from lgatr) (0.8.2)\n",
+            "Requirement already satisfied: opt_einsum in /usr/local/lib/python3.12/dist-packages (from lgatr) (3.4.0)\n",
+            "Requirement already satisfied: cramjam>=2.5.0 in /usr/local/lib/python3.12/dist-packages (from uproot) (2.11.0)\n",
+            "Requirement already satisfied: fsspec!=2026.2.0 in /usr/local/lib/python3.12/dist-packages (from uproot) (2025.3.0)\n",
+            "Requirement already satisfied: packaging in /usr/local/lib/python3.12/dist-packages (from uproot) (26.2)\n",
+            "Requirement already satisfied: xxhash in /usr/local/lib/python3.12/dist-packages (from uproot) (3.8.1)\n",
+            "Collecting awkward-cpp==54 (from awkward)\n",
+            "  Downloading awkward_cpp-54-cp312-cp312-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl.metadata (2.1 kB)\n",
+            "Requirement already satisfied: filelock in /usr/local/lib/python3.12/dist-packages (from torch>=2.1->lgatr) (3.29.7)\n",
+            "Requirement already satisfied: typing-extensions>=4.10.0 in /usr/local/lib/python3.12/dist-packages (from torch>=2.1->lgatr) (4.16.0)\n",
+            "Requirement already satisfied: setuptools<82 in /usr/local/lib/python3.12/dist-packages (from torch>=2.1->lgatr) (75.2.0)\n",
+            "Requirement already satisfied: sympy>=1.13.3 in /usr/local/lib/python3.12/dist-packages (from torch>=2.1->lgatr) (1.14.0)\n",
+            "Requirement already satisfied: networkx>=2.5.1 in /usr/local/lib/python3.12/dist-packages (from torch>=2.1->lgatr) (3.6.1)\n",
+            "Requirement already satisfied: jinja2 in /usr/local/lib/python3.12/dist-packages (from torch>=2.1->lgatr) (3.1.6)\n",
+            "Requirement already satisfied: cuda-toolkit==12.8.1 in /usr/local/lib/python3.12/dist-packages (from cuda-toolkit[cublas,cudart,cufft,cufile,cupti,curand,cusolver,cusparse,nvjitlink,nvrtc,nvtx]==12.8.1; platform_system == \"Linux\"->torch>=2.1->lgatr) (12.8.1)\n",
+            "Requirement already satisfied: cuda-bindings<13,>=12.9.4 in /usr/local/lib/python3.12/dist-packages (from torch>=2.1->lgatr) (12.9.7)\n",
+            "Requirement already satisfied: nvidia-cudnn-cu12==9.19.0.56 in /usr/local/lib/python3.12/dist-packages (from torch>=2.1->lgatr) (9.19.0.56)\n",
+            "Requirement already satisfied: nvidia-cusparselt-cu12==0.7.1 in /usr/local/lib/python3.12/dist-packages (from torch>=2.1->lgatr) (0.7.1)\n",
+            "Requirement already satisfied: nvidia-nccl-cu12==2.28.9 in /usr/local/lib/python3.12/dist-packages (from torch>=2.1->lgatr) (2.28.9)\n",
+            "Requirement already satisfied: nvidia-nvshmem-cu12==3.4.5 in /usr/local/lib/python3.12/dist-packages (from torch>=2.1->lgatr) (3.4.5)\n",
+            "Requirement already satisfied: triton==3.6.0 in /usr/local/lib/python3.12/dist-packages (from torch>=2.1->lgatr) (3.6.0)\n",
+            "Requirement already satisfied: nvidia-cublas-cu12==12.8.4.1.* in /usr/local/lib/python3.12/dist-packages (from cuda-toolkit[cublas,cudart,cufft,cufile,cupti,curand,cusolver,cusparse,nvjitlink,nvrtc,nvtx]==12.8.1; platform_system == \"Linux\"->torch>=2.1->lgatr) (12.8.4.1)\n",
+            "Requirement already satisfied: nvidia-cuda-runtime-cu12==12.8.90.* in /usr/local/lib/python3.12/dist-packages (from cuda-toolkit[cublas,cudart,cufft,cufile,cupti,curand,cusolver,cusparse,nvjitlink,nvrtc,nvtx]==12.8.1; platform_system == \"Linux\"->torch>=2.1->lgatr) (12.8.90)\n",
+            "Requirement already satisfied: nvidia-cufft-cu12==11.3.3.83.* in /usr/local/lib/python3.12/dist-packages (from cuda-toolkit[cublas,cudart,cufft,cufile,cupti,curand,cusolver,cusparse,nvjitlink,nvrtc,nvtx]==12.8.1; platform_system == \"Linux\"->torch>=2.1->lgatr) (11.3.3.83)\n",
+            "Requirement already satisfied: nvidia-cufile-cu12==1.13.1.3.* in /usr/local/lib/python3.12/dist-packages (from cuda-toolkit[cublas,cudart,cufft,cufile,cupti,curand,cusolver,cusparse,nvjitlink,nvrtc,nvtx]==12.8.1; platform_system == \"Linux\"->torch>=2.1->lgatr) (1.13.1.3)\n",
+            "Requirement already satisfied: nvidia-cuda-cupti-cu12==12.8.90.* in /usr/local/lib/python3.12/dist-packages (from cuda-toolkit[cublas,cudart,cufft,cufile,cupti,curand,cusolver,cusparse,nvjitlink,nvrtc,nvtx]==12.8.1; platform_system == \"Linux\"->torch>=2.1->lgatr) (12.8.90)\n",
+            "Requirement already satisfied: nvidia-curand-cu12==10.3.9.90.* in /usr/local/lib/python3.12/dist-packages (from cuda-toolkit[cublas,cudart,cufft,cufile,cupti,curand,cusolver,cusparse,nvjitlink,nvrtc,nvtx]==12.8.1; platform_system == \"Linux\"->torch>=2.1->lgatr) (10.3.9.90)\n",
+            "Requirement already satisfied: nvidia-cusolver-cu12==11.7.3.90.* in /usr/local/lib/python3.12/dist-packages (from cuda-toolkit[cublas,cudart,cufft,cufile,cupti,curand,cusolver,cusparse,nvjitlink,nvrtc,nvtx]==12.8.1; platform_system == \"Linux\"->torch>=2.1->lgatr) (11.7.3.90)\n",
+            "Requirement already satisfied: nvidia-cusparse-cu12==12.5.8.93.* in /usr/local/lib/python3.12/dist-packages (from cuda-toolkit[cublas,cudart,cufft,cufile,cupti,curand,cusolver,cusparse,nvjitlink,nvrtc,nvtx]==12.8.1; platform_system == \"Linux\"->torch>=2.1->lgatr) (12.5.8.93)\n",
+            "Requirement already satisfied: nvidia-nvjitlink-cu12==12.8.93.* in /usr/local/lib/python3.12/dist-packages (from cuda-toolkit[cublas,cudart,cufft,cufile,cupti,curand,cusolver,cusparse,nvjitlink,nvrtc,nvtx]==12.8.1; platform_system == \"Linux\"->torch>=2.1->lgatr) (12.8.93)\n",
+            "Requirement already satisfied: nvidia-cuda-nvrtc-cu12==12.8.93.* in /usr/local/lib/python3.12/dist-packages (from cuda-toolkit[cublas,cudart,cufft,cufile,cupti,curand,cusolver,cusparse,nvjitlink,nvrtc,nvtx]==12.8.1; platform_system == \"Linux\"->torch>=2.1->lgatr) (12.8.93)\n",
+            "Requirement already satisfied: nvidia-nvtx-cu12==12.8.90.* in /usr/local/lib/python3.12/dist-packages (from cuda-toolkit[cublas,cudart,cufft,cufile,cupti,curand,cusolver,cusparse,nvjitlink,nvrtc,nvtx]==12.8.1; platform_system == \"Linux\"->torch>=2.1->lgatr) (12.8.90)\n",
+            "Requirement already satisfied: cuda-pathfinder~=1.1 in /usr/local/lib/python3.12/dist-packages (from cuda-bindings<13,>=12.9.4->torch>=2.1->lgatr) (1.5.6)\n",
+            "Requirement already satisfied: mpmath<1.4,>=1.1.0 in /usr/local/lib/python3.12/dist-packages (from sympy>=1.13.3->torch>=2.1->lgatr) (1.3.0)\n",
+            "Requirement already satisfied: MarkupSafe>=2.0 in /usr/local/lib/python3.12/dist-packages (from jinja2->torch>=2.1->lgatr) (3.0.3)\n",
+            "Downloading lgatr-1.4.4-py3-none-any.whl (60 kB)\n",
+            "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m60.6/60.6 kB\u001b[0m \u001b[31m5.7 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
+            "\u001b[?25hDownloading uproot-5.7.5-py3-none-any.whl (401 kB)\n",
+            "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m401.2/401.2 kB\u001b[0m \u001b[31m35.0 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
+            "\u001b[?25hDownloading awkward-2.11.0-py3-none-any.whl (974 kB)\n",
+            "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m974.8/974.8 kB\u001b[0m \u001b[31m76.6 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
+            "\u001b[?25hDownloading awkward_cpp-54-cp312-cp312-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl (689 kB)\n",
+            "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m689.3/689.3 kB\u001b[0m \u001b[31m48.4 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
+            "\u001b[?25hDownloading vector-1.8.1-py3-none-any.whl (182 kB)\n",
+            "\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m182.7/182.7 kB\u001b[0m \u001b[31m23.0 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
+            "\u001b[?25hInstalling collected packages: vector, awkward-cpp, awkward, uproot, lgatr\n",
+            "Successfully installed awkward-2.11.0 awkward-cpp-54 lgatr-1.4.4 uproot-5.7.5 vector-1.8.1\n"
+          ]
+        }
+      ]
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "# --- Default libraries ---\n",
+        "import os\n",
+        "import warnings\n",
+        "from pathlib import Path\n",
+        "\n",
+        "# --- Working directory ---\n",
+        "PROJECT_DIR = Path().resolve()\n",
+        "PROJECT_ROOT_NAME = 'Hybrid_Transformer_Thanh_Nguyen'\n",
+        "\n",
+        "while PROJECT_DIR.name != PROJECT_ROOT_NAME and PROJECT_DIR != PROJECT_DIR.parent:\n",
+        "    PROJECT_DIR = PROJECT_DIR.parent\n",
+        "\n",
+        "if Path().resolve() != PROJECT_DIR:\n",
+        "    os.chdir(PROJECT_DIR)\n",
+        "\n",
+        "DATA_DIR = PROJECT_DIR / 'data'\n",
+        "LOG_DIR = PROJECT_DIR / 'logs'\n",
+        "\n",
+        "# --- Data preprocessing & visualization ---\n",
+        "import numpy as np\n",
+        "import matplotlib.pyplot as plt\n",
+        "from sklearn.model_selection import train_test_split\n",
+        "\n",
+        "# --- Deep learning ---\n",
+        "import torch\n",
+        "\n",
+        "# --- Custom modules ---\n",
+        "from src.configs import LorentzParTConfig, TrainConfig\n",
+        "from src.engine import MaskedModelTrainer, Trainer\n",
+        "from src.models import LorentzParT\n",
+        "from src.utils import accuracy_metric_ce, set_seed\n",
+        "from src.utils.data import JetClassDataset, compute_norm_stats, read_file\n",
+        "from src.utils.viz import *\n",
+        "\n",
+        "# --- Settings ---\n",
+        "warnings.filterwarnings('ignore')\n",
+        "set_seed(42)\n",
+        "device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')\n",
+        "device"
+      ],
+      "metadata": {
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        },
+        "id": "IvxBxUicRu18",
+        "outputId": "3241b771-4383-417a-cfc1-963899fb5dd4"
+      },
+      "id": "IvxBxUicRu18",
+      "execution_count": null,
+      "outputs": [
+        {
+          "output_type": "execute_result",
+          "data": {
+            "text/plain": [
+              "device(type='cuda')"
+            ]
+          },
+          "metadata": {},
+          "execution_count": 2
+        }
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "source": [
+        "#Loading Dpendencies"
+      ],
+      "metadata": {
+        "id": "BmBDXDMSD61u"
+      },
+      "id": "BmBDXDMSD61u"
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "from typing import List, Tuple, Dict, Optional\n",
+        "\n",
+        "import torch\n",
+        "from torch import nn, Tensor\n",
+        "from lgatr.interface import extract_vector\n",
+        "from lgatr.layers import EquiLinear\n",
+        "\n",
+        "from src.models.classifier import ClassAttentionBlock, Classifier\n",
+        "from src.models.feedforward import Feedforward\n",
+        "from src.models.particle_transformer import ParticleAttentionBlock\n",
+        "from src.models.processor import InteractionEmbedding, ParticleProcessor\n",
+        "from src.configs import LorentzParTConfig\n",
+        "from lgatr.interface import extract_vector\n",
+        "from lgatr.layers import EquiLinear\n",
+        "\n",
+        "\n"
+      ],
+      "metadata": {
+        "id": "jTbtwxdtAt9S"
+      },
+      "id": "jTbtwxdtAt9S",
+      "execution_count": null,
+      "outputs": []
+    },
+    {
+      "cell_type": "markdown",
+      "source": [
+        "#Data Preparation"
+      ],
+      "metadata": {
+        "id": "VB1JTgpxDv-4"
+      },
+      "id": "VB1JTgpxDv-4"
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "import os\n",
+        "from pathlib import Path\n",
+        "import numpy as np\n",
+        "from sklearn.model_selection import train_test_split\n",
+        "from src.utils.data import JetClassDataset, compute_norm_stats\n",
+        "\n",
+        "# --- 1. Mount Google Drive (If not already done) ---\n",
+        "from google.colab import drive\n",
+        "if not os.path.exists('/content/drive'):\n",
+        "    drive.mount('/content/drive')\n",
+        "\n",
+        "# --- 2. Locate and Load the Compressed Archive ---\n",
+        "DRIVE_FILE_PATH = '/content/drive/MyDrive/GSOC/dara/jetclass_balanced_1M.npz'\n",
+        "\n",
+        "print(\"=\" * 70)\n",
+        "print(f\"LOADING SERIALIZED ARRAYS FROM DRIVE\")\n",
+        "print(\"=\" * 70)\n",
+        "\n",
+        "if os.path.exists(DRIVE_FILE_PATH):\n",
+        "    # Load using memory-mapping for high-speed indexing\n",
+        "    data_archive = np.load(DRIVE_FILE_PATH, mmap_mode='r')\n",
+        "\n",
+        "    X_particles = data_archive['X_particles']\n",
+        "    X_jets      = data_archive['X_jets']\n",
+        "    y           = data_archive['Y']\n",
+        "\n",
+        "    print(\"SUCCESS: Data loaded cleanly into RAM!\")\n",
+        "    print(f\"  -> X_particles matrix shape : {X_particles.shape}\")\n",
+        "    print(f\"  -> X_jets matrix shape      : {X_jets.shape}\")\n",
+        "    print(f\"  -> y (Labels) matrix shape  : {y.shape}\")\n",
+        "    print(\"=\" * 70 + \"\\n\")\n",
+        "else:\n",
+        "    raise FileNotFoundError(f\"ERROR: Could not find the file at {DRIVE_FILE_PATH}\")\n",
+        "\n",
+        "# --- 3. Split the Balanced Dataset Safely ---\n",
+        "# We enforce stratify=y to lock in your strict 10% balance across all splits\n",
+        "X_train, X_val, y_train, y_val = train_test_split(X_particles, y, test_size=0.2, random_state=42, stratify=y)\n",
+        "X_val, X_test, y_val, y_test = train_test_split(X_val, y_val, test_size=0.5, random_state=42, stratify=y_val)\n",
+        "\n",
+        "# --- 4. Re-Initialize JetClass Dataset Objects ---\n",
+        "normalize = [True, False, False, True]\n",
+        "norm_dict = compute_norm_stats(X_train)\n",
+        "\n",
+        "train_dataset = JetClassDataset(X_train, y_train, normalize, norm_dict, mask_mode='biased')\n",
+        "val_dataset   = JetClassDataset(X_val, y_val, normalize, norm_dict, mask_mode='biased')\n",
+        "test_dataset  = JetClassDataset(X_test, y_test, normalize, norm_dict, mask_mode='first')"
+      ],
+      "metadata": {
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        },
+        "id": "9RBH34OYBoZ7",
+        "outputId": "a53a1da2-e44b-4611-cd4e-f610d3cce289"
+      },
+      "id": "9RBH34OYBoZ7",
+      "execution_count": null,
+      "outputs": [
+        {
+          "output_type": "stream",
+          "name": "stdout",
+          "text": [
+            "Mounted at /content/drive\n",
+            "======================================================================\n",
+            "LOADING SERIALIZED ARRAYS FROM DRIVE\n",
+            "======================================================================\n",
+            "SUCCESS: Data loaded cleanly into RAM!\n",
+            "  -> X_particles matrix shape : (1000000, 4, 128)\n",
+            "  -> X_jets matrix shape      : (1000000, 4)\n",
+            "  -> y (Labels) matrix shape  : (1000000, 10)\n",
+            "======================================================================\n",
+            "\n",
+            "pt_mean: 92.70597076416016, pt_std: 105.79937744140625\n",
+            "eta_mean: -0.0011634620605036616, eta_std: 0.9182536005973816\n",
+            "phi_mean: -0.0006678671925328672, phi_std: 1.8138455152511597\n",
+            "E_mean: 133.98568725585938, E_std: 167.7259979248047\n"
+          ]
+        }
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "edb52e95",
+      "metadata": {
+        "id": "edb52e95"
+      },
+      "source": [
+        "## Applying Gated Attention LorentzPart"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "import torch\n",
+        "import torch.nn as nn\n",
+        "from torch import Tensor\n",
+        "from typing import Optional\n",
+        "\n",
+        "class ParticleAttentionBlock(nn.Module):\n",
+        "    def __init__(\n",
+        "        self,\n",
+        "        embed_dim: int = 128,\n",
+        "        num_heads: int = 8,\n",
+        "        dropout: float = 0.1,\n",
+        "        expansion_factor: int = 4,\n",
+        "        gate_type: Optional[str] = \"headwise\",\n",
+        "    ):\n",
+        "        super(ParticleAttentionBlock, self).__init__()\n",
+        "        assert embed_dim % num_heads == 0, \"embed_dim must be divisible by num_heads\"\n",
+        "\n",
+        "        self.embed_dim = embed_dim\n",
+        "        self.num_heads = num_heads\n",
+        "        self.head_dim = embed_dim // num_heads\n",
+        "        self.gate_type = gate_type\n",
+        "\n",
+        "        self.layernorm1 = nn.LayerNorm(embed_dim)\n",
+        "        self.mass_norm = nn.LayerNorm(1)#norm\n",
+        "\n",
+        "        # Project pooled interaction head-features to match token embedding dimensions\n",
+        "        self.physics_proj = nn.Linear(num_heads, embed_dim)\n",
+        "\n",
+        "        # Project the global scalar invariant mass squared (m2) to the embedding space\n",
+        "        self.mass_proj = nn.Linear(1, embed_dim)\n",
+        "\n",
+        "        # Gating projections accept physics-fused representations\n",
+        "        if self.gate_type == \"headwise\":\n",
+        "            self.gate_proj = nn.Linear(embed_dim, num_heads)\n",
+        "        elif self.gate_type == \"elementwise\":\n",
+        "            self.gate_proj = nn.Linear(embed_dim, embed_dim)\n",
+        "\n",
+        "        self.pmha = nn.MultiheadAttention(\n",
+        "            embed_dim=embed_dim,\n",
+        "            num_heads=num_heads,\n",
+        "            dropout=dropout,\n",
+        "            batch_first=True\n",
+        "        )\n",
+        "\n",
+        "        self.layernorm2 = nn.LayerNorm(embed_dim)\n",
+        "        self.dropout = nn.Dropout(dropout)\n",
+        "\n",
+        "        self.feedforward = Feedforward(\n",
+        "            embed_dim=embed_dim,\n",
+        "            expansion_factor=expansion_factor,\n",
+        "            dropout=dropout\n",
+        "        )\n",
+        "\n",
+        "    def forward(self, x: Tensor, padding_mask: Tensor, U: Optional[Tensor] = None, p4: Optional[Tensor] = None) -> Tensor:\n",
+        "        residual = x\n",
+        "        B_size, N_particles, _ = x.shape\n",
+        "\n",
+        "        # 1. Standard token serialization\n",
+        "        x_norm = self.layernorm1(x)\n",
+        "\n",
+        "        # 2. FIXED: U is ALREADY [Batch * Heads, N, N]. Pass directly to PyTorch MHA.\n",
+        "        x_attn, _ = self.pmha(x_norm, x_norm, x_norm, key_padding_mask=padding_mask, attn_mask=U)\n",
+        "\n",
+        "        # 3. Physics-Aware Fusion (Invariants-driven conditioning)\n",
+        "        if U is not None:\n",
+        "            # Reconstruct the 4D shape: [B * H, N, N] -> [B, H, N, N]\n",
+        "            U_reshaped = U.view(B_size, self.num_heads, N_particles, N_particles)\n",
+        "\n",
+        "            # Pool over neighbor particle index 'j' (dim=3). Resulting shape: [B, H, N]\n",
+        "            u_pooled = U_reshaped.sum(dim=3)\n",
+        "\n",
+        "            # Transpose to align with token channels: [B, N, H]\n",
+        "            u_pooled = u_pooled.transpose(1, 2)\n",
+        "\n",
+        "            # Map head-wise pooled invariants into token channel space: [B, N, embed_dim]\n",
+        "            physics_context = self.physics_proj(u_pooled)\n",
+        "\n",
+        "            # Fuse physical invariants with abstract node latent maps\n",
+        "            x_gating_input = x_norm + physics_context\n",
+        "        else:\n",
+        "            # Fallback path if U is not provided\n",
+        "            x_gating_input = x_norm\n",
+        "\n",
+        "        # 4. Compute Global Invariant Mass Bias from 4-vectors [B, N, 4] -> (E, px, py, pz)\n",
+        "        if p4 is not None:\n",
+        "            # Sum energy component (index 0) over all particles (dim=1)\n",
+        "            energy_sum = p4[..., 0].sum(dim=1, keepdim=True) # Shape: [B, 1]\n",
+        "\n",
+        "            # Sum momentum components (indices 1, 2, 3) over all particles (dim=1)\n",
+        "            momentum_sum = p4[..., 1:].sum(dim=1) # Shape: [B, 3]\n",
+        "\n",
+        "            # Calculate invariant mass squared (m2)\n",
+        "            m2 = energy_sum**2 - momentum_sum.norm(dim=-1, keepdim=True)**2 # Shape: [B, 1]\n",
+        "\n",
+        "            m2_scaled = torch.log1p(torch.relu(m2))\n",
+        "\n",
+        "\n",
+        "            # Step C: Standardize the mean and variance dynamically\n",
+        "            m2_norm = self.mass_norm(m2_scaled)\n",
+        "\n",
+        "            # Project normalized m2 into a global embedding bias vector [B, 1, embed_dim]\n",
+        "            mass_bias = self.mass_proj(m2_norm).unsqueeze(1)\n",
+        "\n",
+        "            # Broad-cast add global event mass bias to the per-particle gating inputs\n",
+        "            x_gating_input = x_gating_input + mass_bias\n",
+        "\n",
+        "\n",
+        "\n",
+        "        # 5. Compute and apply the explicitly Physics-Aware Gate\n",
+        "        if self.gate_type == \"headwise\":\n",
+        "            # Compute score matrix from physics-fused map: (B, N, embed_dim) -> (B, N, num_heads, 1)\n",
+        "            gate_score = self.gate_proj(x_gating_input).unsqueeze(-1)\n",
+        "\n",
+        "            # Separate heads to apply individual scalar gating values\n",
+        "            x_attn = x_attn.view(B_size, N_particles, self.num_heads, self.head_dim)\n",
+        "\n",
+        "            # Apply physics-conditioned filter and reconstruct classic transformer shape\n",
+        "            x_attn = (x_attn * torch.sigmoid(gate_score)).view(B_size, N_particles, self.embed_dim)\n",
+        "\n",
+        "        elif self.gate_type == \"elementwise\":\n",
+        "            # Compute full channel-by-channel mask from physics-fused map: (B, N, embed_dim)\n",
+        "            gate_score = self.gate_proj(x_gating_input)\n",
+        "            x_attn = x_attn * torch.sigmoid(gate_score)\n",
+        "\n",
+        "        # 6. Standard Feedforward processing\n",
+        "        x = self.layernorm2(x_attn)\n",
+        "        x = self.dropout(x)\n",
+        "        x += residual\n",
+        "        x = self.feedforward(x)\n",
+        "\n",
+        "        return x"
+      ],
+      "metadata": {
+        "id": "AvsAk0byPoBK"
+      },
+      "id": "AvsAk0byPoBK",
+      "execution_count": null,
+      "outputs": []
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "import torch\n",
+        "import torch.nn as nn\n",
+        "from torch import Tensor\n",
+        "from typing import Optional, List, Dict, Tuple\n",
+        "\n",
+        "\n",
+        "\n",
+        "class LorentzParTEncoder(nn.Module):\n",
+        "    def __init__(\n",
+        "        self,\n",
+        "        embed_dim: int = 128,\n",
+        "        num_heads: int = 8,\n",
+        "        num_layers: int = 8,\n",
+        "        in_s_channels: Optional[int] = None,\n",
+        "        out_s_channels: Optional[int] = None,\n",
+        "        dropout: float = 0.1,\n",
+        "        expansion_factor: int = 4,\n",
+        "        pair_embed_dims: List[int] = [64, 64, 64],\n",
+        "        attention_config: Dict = {}\n",
+        "    ):\n",
+        "        super(LorentzParTEncoder, self).__init__()\n",
+        "        self.equilinear = EquiLinear(\n",
+        "            in_mv_channels=1,\n",
+        "            out_mv_channels=1,\n",
+        "            in_s_channels=in_s_channels,\n",
+        "            out_s_channels=out_s_channels\n",
+        "        )\n",
+        "        self.proj = nn.Linear(16, embed_dim)\n",
+        "        self.interaction_embed = InteractionEmbedding(\n",
+        "            num_interaction_features=4,\n",
+        "            pair_embed_dims=pair_embed_dims + [num_heads]\n",
+        "        )\n",
+        "\n",
+        "        use_gating = attention_config.get('use_gating', False)\n",
+        "\n",
+        "        # Explicitly pass gate_type so the block knows whether to use physics gating or standard\n",
+        "        self.encoder = nn.ModuleList([\n",
+        "            ParticleAttentionBlock(\n",
+        "                embed_dim=embed_dim,\n",
+        "                num_heads=num_heads,\n",
+        "                dropout=dropout,\n",
+        "                expansion_factor=expansion_factor,\n",
+        "                gate_type=\"headwise\" if use_gating else None\n",
+        "            ) for _ in range(num_layers)\n",
+        "        ])\n",
+        "\n",
+        "    def forward(self, x: Tensor, padding_mask: Tensor, U: Tensor, p4: Optional[Tensor] = None) -> Tensor:\n",
+        "        B, N, F = x.shape\n",
+        "        U = self.interaction_embed(U)\n",
+        "        x = x.view(B, N, 1, F)\n",
+        "        x, _ = self.equilinear(x)\n",
+        "        x = x.view(B, N, 16)\n",
+        "        x = self.proj(x)\n",
+        "\n",
+        "        # Pass p4 down to the attention blocks\n",
+        "        for layer in self.encoder:\n",
+        "            x = layer(x, padding_mask, U, p4=p4)\n",
+        "\n",
+        "        return x\n",
+        "\n",
+        "\n",
+        "class LorentzParT(nn.Module):\n",
+        "    def __init__(\n",
+        "        self,\n",
+        "        config: Optional[LorentzParTConfig] = None,\n",
+        "        max_num_particles: Optional[int] = None,\n",
+        "        num_particle_features: Optional[int] = None,\n",
+        "        num_classes: Optional[int] = None,\n",
+        "        embed_dim: Optional[int] = None,\n",
+        "        num_heads: Optional[int] = None,\n",
+        "        num_layers: Optional[int] = None,\n",
+        "        num_cls_layers: Optional[int] = None,\n",
+        "        num_mlp_layers: Optional[int] = None,\n",
+        "        hidden_dim: Optional[int] = None,\n",
+        "        hidden_mv_channels: Optional[int] = None,\n",
+        "        in_s_channels: Optional[int] = None,\n",
+        "        out_s_channels: Optional[int] = None,\n",
+        "        hidden_s_channels: Optional[int] = None,\n",
+        "        attention: Optional[Dict] = None,\n",
+        "        mlp: Optional[Dict] = None,\n",
+        "        reinsert_mv_channels: Optional[Tuple[int]] = None,\n",
+        "        reinsert_s_channels: Optional[Tuple[int]] = None,\n",
+        "        dropout: Optional[float] = None,\n",
+        "        expansion_factor: Optional[int] = None,\n",
+        "        pair_embed_dims: Optional[List[int]] = None,\n",
+        "        mask: Optional[bool] = None,\n",
+        "        weights: Optional[str] = None,\n",
+        "        inference: Optional[bool] = False\n",
+        "    ):\n",
+        "        super(LorentzParT, self).__init__()\n",
+        "\n",
+        "        # Use config if provided, otherwise use defaults\n",
+        "        if config is not None:\n",
+        "            self.max_num_particles = max_num_particles if max_num_particles is not None else config.max_num_particles\n",
+        "            self.num_particle_features = num_particle_features if num_particle_features is not None else config.num_particle_features\n",
+        "            self.num_classes = num_classes if num_classes is not None else config.num_classes\n",
+        "            self.embed_dim = embed_dim if embed_dim is not None else config.embed_dim\n",
+        "            self.num_heads = num_heads if num_heads is not None else config.num_heads\n",
+        "            self.num_layers = num_layers if num_layers is not None else config.num_layers\n",
+        "            self.num_cls_layers = num_cls_layers if num_cls_layers is not None else config.num_cls_layers\n",
+        "            self.num_mlp_layers = num_mlp_layers if num_mlp_layers is not None else config.num_mlp_layers\n",
+        "            self.hidden_dim = hidden_dim if hidden_dim is not None else config.hidden_dim\n",
+        "            self.hidden_mv_channels = hidden_mv_channels if hidden_mv_channels is not None else config.hidden_mv_channels\n",
+        "            self.in_s_channels = in_s_channels if in_s_channels is not None else config.in_s_channels\n",
+        "            self.out_s_channels = out_s_channels if out_s_channels is not None else config.out_s_channels\n",
+        "            self.hidden_s_channels = hidden_s_channels if hidden_s_channels is not None else config.hidden_s_channels\n",
+        "            self.attention = attention if attention is not None else config.attention\n",
+        "            self.mlp = mlp if mlp is not None else config.mlp\n",
+        "            self.reinsert_mv_channels = reinsert_mv_channels if reinsert_mv_channels is not None else config.reinsert_mv_channels\n",
+        "            self.reinsert_s_channels = reinsert_s_channels if reinsert_s_channels is not None else config.reinsert_s_channels\n",
+        "            self.dropout = dropout if dropout is not None else config.dropout\n",
+        "            self.expansion_factor = expansion_factor if expansion_factor is not None else config.expansion_factor\n",
+        "            self.pair_embed_dims = pair_embed_dims if pair_embed_dims is not None else config.pair_embed_dims\n",
+        "            self.mask = mask if mask is not None else config.mask\n",
+        "            self.weights = weights if weights is not None else config.weights\n",
+        "            self.inference = inference if inference is not None else config.inference\n",
+        "        else:\n",
+        "            self.max_num_particles = max_num_particles if max_num_particles is not None else 128\n",
+        "            self.num_particle_features = num_particle_features if num_particle_features is not None else 4\n",
+        "            self.num_classes = num_classes if num_classes is not None else 10\n",
+        "            self.embed_dim = embed_dim if embed_dim is not None else 128\n",
+        "            self.num_heads = num_heads if num_heads is not None else 8\n",
+        "            self.num_layers = num_layers if num_layers is not None else 8\n",
+        "            self.num_cls_layers = num_cls_layers if num_cls_layers is not None else 2\n",
+        "            self.num_mlp_layers = num_mlp_layers if num_mlp_layers is not None else 0\n",
+        "            self.hidden_dim = hidden_dim if hidden_dim is not None else 256\n",
+        "            self.hidden_mv_channels = hidden_mv_channels if hidden_mv_channels is not None else 8\n",
+        "            self.in_s_channels = in_s_channels if in_s_channels is not None else None\n",
+        "            self.out_s_channels = out_s_channels if out_s_channels is not None else None\n",
+        "            self.hidden_s_channels = hidden_s_channels if hidden_s_channels is not None else 16\n",
+        "            self.attention = attention if attention is not None else {}\n",
+        "            self.mlp = mlp if mlp is not None else None\n",
+        "            self.reinsert_mv_channels = reinsert_mv_channels if reinsert_mv_channels is not None else None\n",
+        "            self.reinsert_s_channels = reinsert_s_channels if reinsert_s_channels is not None else None\n",
+        "            self.dropout = dropout if dropout is not None else 0.1\n",
+        "            self.expansion_factor = expansion_factor if expansion_factor is not None else 4\n",
+        "            self.pair_embed_dims = pair_embed_dims if pair_embed_dims is not None else [64, 64, 64]\n",
+        "            self.mask = mask if mask is not None else False\n",
+        "            self.weights = weights if weights is not None else None\n",
+        "            self.inference = inference if inference is not None else False\n",
+        "\n",
+        "        # Initialize the class token\n",
+        "        self.cls_token = nn.Parameter(torch.zeros(1, 1, self.embed_dim), requires_grad=True)\n",
+        "        nn.init.normal_(self.cls_token, mean=0.0, std=1.0)\n",
+        "\n",
+        "        self.processor = ParticleProcessor(to_multivector=True)\n",
+        "\n",
+        "        # Updated Encoder with attention_config passed dynamically\n",
+        "        self.encoder = LorentzParTEncoder(\n",
+        "            embed_dim=self.embed_dim,\n",
+        "            num_heads=self.num_heads,\n",
+        "            num_layers=self.num_layers,\n",
+        "            in_s_channels=self.in_s_channels,\n",
+        "            out_s_channels=self.out_s_channels,\n",
+        "            dropout=self.dropout,\n",
+        "            expansion_factor=self.expansion_factor,\n",
+        "            pair_embed_dims=self.pair_embed_dims,\n",
+        "            attention_config=self.attention\n",
+        "        )\n",
+        "\n",
+        "        # For self-supervised learning\n",
+        "        self.fc = nn.Linear(self.max_num_particles * self.embed_dim, 16)\n",
+        "        self.equilinear = EquiLinear(\n",
+        "            in_mv_channels=1,\n",
+        "            out_mv_channels=1,\n",
+        "            in_s_channels=self.in_s_channels,\n",
+        "            out_s_channels=self.out_s_channels\n",
+        "        )\n",
+        "\n",
+        "        # For classification\n",
+        "        self.decoder = nn.ModuleList([\n",
+        "            ClassAttentionBlock(\n",
+        "                embed_dim=self.embed_dim,\n",
+        "                num_heads=self.num_heads,\n",
+        "                dropout=0.0,\n",
+        "                expansion_factor=self.expansion_factor\n",
+        "            ) for _ in range(self.num_cls_layers)\n",
+        "        ])\n",
+        "        self.layernorm = nn.LayerNorm(self.embed_dim)\n",
+        "        self.classifier = Classifier(\n",
+        "            num_classes=self.num_classes,\n",
+        "            input_dim=self.embed_dim,\n",
+        "            hidden_dim=self.hidden_dim,\n",
+        "            num_layers=self.num_mlp_layers,\n",
+        "            dropout=self.dropout,\n",
+        "        )\n",
+        "        self.act = nn.Softmax(dim=1) if self.inference else nn.Identity()\n",
+        "\n",
+        "        # Load pretrained weights\n",
+        "        if self.weights is not None:\n",
+        "            state_dict = torch.load(self.weights)\n",
+        "            filtered_state = {\n",
+        "                k[len(\"encoder.\") :]: v\n",
+        "                for k, v in state_dict.items()\n",
+        "                if k.startswith(\"encoder.\")\n",
+        "            }\n",
+        "            self.encoder.load_state_dict(filtered_state, strict=False)\n",
+        "\n",
+        "    def forward(self, x: Tensor, mask_idx: Optional[Tensor] = None) -> Tensor:\n",
+        "        B, N, F = x.shape  # (batch_size, max_num_particles, num_particle_features)\n",
+        "\n",
+        "        # Save the raw kinematics before processor alters them\n",
+        "        p4 = x.clone()\n",
+        "\n",
+        "        # Ignore padding particles in query\n",
+        "        padding_mask = (x[..., 3] == 0).float()  # (B, N)\n",
+        "\n",
+        "        # Set the masked indices to 0.0 so they are not ignored in MultiheadAttention()\n",
+        "        if mask_idx is not None:\n",
+        "            batch_indices = torch.arange(x.size(0), device=x.device)\n",
+        "            padding_mask[batch_indices, mask_idx] = 0.0\n",
+        "\n",
+        "        # Process particles to get interaction embeddings and multivectors (if applicable)\n",
+        "        x, U = self.processor(x)\n",
+        "\n",
+        "        # Pass through equilinear layer and particle attention blocks (passing p4 down)\n",
+        "        x = self.encoder(x, padding_mask, U, p4=p4)\n",
+        "\n",
+        "        # Classification (no masking in this case)\n",
+        "        if not self.mask:\n",
+        "            x_cls = self.cls_token.expand(B, -1, -1)\n",
+        "\n",
+        "            # Decoder with class attention blocks\n",
+        "            for layer in self.decoder:\n",
+        "                x_cls = layer(x, x_cls, padding_mask)\n",
+        "\n",
+        "            # MLP head for classification\n",
+        "            x_cls = self.layernorm(x_cls).squeeze(1)\n",
+        "            x_cls = self.classifier(x_cls)\n",
+        "            output = self.act(x_cls)  # (B, num_classes)\n",
+        "\n",
+        "            return output\n",
+        "        else:\n",
+        "            x = x.view(B, -1)  # (B, N * embed_dim)\n",
+        "            x = self.fc(x)  # (B, 16)\n",
+        "            x = x.view(B, 1, 1, 16)\n",
+        "            x, _ = self.equilinear(x)  # (B, 1, 1, 16)\n",
+        "            x = x.view(B, 16)\n",
+        "            x = extract_vector(x)  # (B, F)\n",
+        "\n",
+        "            return x"
+      ],
+      "metadata": {
+        "id": "QRhMF1TP8Ymw"
+      },
+      "id": "QRhMF1TP8Ymw",
+      "execution_count": null,
+      "outputs": []
+    },
+    {
+      "cell_type": "markdown",
+      "source": [
+        "#Gating Layer Test"
+      ],
+      "metadata": {
+        "id": "UC2GQe3iR7Gt"
+      },
+      "id": "UC2GQe3iR7Gt"
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "# 1. Initialize your config with gating enabled\n",
+        "test_config = LorentzParTConfig(\n",
+        "    embed_dim=128,\n",
+        "    num_heads=8,\n",
+        "    num_layers=8,\n",
+        "    attention={'use_gating': True},\n",
+        "    mask=True\n",
+        ")\n",
+        "\n",
+        "# 2. Instantiate the model\n",
+        "model = LorentzParT(config=test_config)\n",
+        "\n",
+        "# 3. Verification checks\n",
+        "first_layer = model.encoder.encoder[0]\n",
+        "is_gated = isinstance(first_layer, ParticleAttentionBlock)\n",
+        "\n",
+        "print(f\"--- Gating Verification ---\")\n",
+        "print(f\"Encoder Layer 1 Type: {type(first_layer).__name__}\")\n",
+        "print(f\"Gating Active: {is_gated}\")\n",
+        "\n",
+        "if is_gated:\n",
+        "    print(\"Success: The model is now using Attention Gating!\")\n",
+        "else:\n",
+        "    print(\"Error: The model is still using standard Attention Blocks.\")"
+      ],
+      "metadata": {
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        },
+        "id": "cKfNSeyPR9nH",
+        "outputId": "8ea6d1d4-605b-4ec4-b41e-f2a61eb0c23b"
+      },
+      "id": "cKfNSeyPR9nH",
+      "execution_count": null,
+      "outputs": [
+        {
+          "output_type": "stream",
+          "name": "stdout",
+          "text": [
+            "--- Gating Verification ---\n",
+            "Encoder Layer 1 Type: ParticleAttentionBlock\n",
+            "Gating Active: True\n",
+            "Success: The model is now using Attention Gating!\n"
+          ]
+        }
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "source": [
+        "#Pre-Train Steps"
+      ],
+      "metadata": {
+        "id": "H_Kmp-s3SUDD"
+      },
+      "id": "H_Kmp-s3SUDD"
+    },
+    {
+      "cell_type": "markdown",
+      "source": [
+        "#Using Self Supervised Weights"
+      ],
+      "metadata": {
+        "id": "5peLQ8txiFZG"
+      },
+      "id": "5peLQ8txiFZG"
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "# Initialize configuration with Attention Gating enabled\n",
+        "#not changing name of ssl_model_config\n",
+        "ssl_model_config = LorentzParTConfig(\n",
+        "    embed_dim=128,\n",
+        "    num_heads=8,\n",
+        "    num_layers=8,\n",
+        "    hidden_mv_channels=8,\n",
+        "    attention={'use_gating': True}, # This is the trigger for your new code\n",
+        "    dropout=0.1,\n",
+        "    expansion_factor=4,\n",
+        "    max_num_particles=128,\n",
+        "    num_particle_features=4,\n",
+        "    pair_embed_dims=[64, 64, 64],\n",
+        "    mask=True # Set to True for Self-Supervised Learning / Masked Training\n",
+        ")"
+      ],
+      "metadata": {
+        "id": "6fGbG9f1SXbc"
+      },
+      "id": "6fGbG9f1SXbc",
+      "execution_count": null,
+      "outputs": []
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "# Create the model and move it to your device (GPU/CPU)\n",
+        "gatedmodel = LorentzParT(config=ssl_model_config)\n",
+        "gatedmodel.to(device)\n",
+        "gatedmodel"
+      ],
+      "metadata": {
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        },
+        "id": "b7g5btHCSZka",
+        "outputId": "a47e4f81-cd33-4b7c-a39a-7f074bf3c456"
+      },
+      "id": "b7g5btHCSZka",
+      "execution_count": null,
+      "outputs": [
+        {
+          "output_type": "execute_result",
+          "data": {
+            "text/plain": [
+              "LorentzParT(\n",
+              "  (processor): ParticleProcessor()\n",
+              "  (encoder): LorentzParTEncoder(\n",
+              "    (equilinear): EquiLinear()\n",
+              "    (proj): Linear(in_features=16, out_features=128, bias=True)\n",
+              "    (interaction_embed): InteractionEmbedding(\n",
+              "      (embed): Sequential(\n",
+              "        (0): BatchNorm1d(4, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
+              "        (1): Conv1d(4, 64, kernel_size=(1,), stride=(1,))\n",
+              "        (2): BatchNorm1d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
+              "        (3): GELU(approximate='none')\n",
+              "        (4): Conv1d(64, 64, kernel_size=(1,), stride=(1,))\n",
+              "        (5): BatchNorm1d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
+              "        (6): GELU(approximate='none')\n",
+              "        (7): Conv1d(64, 64, kernel_size=(1,), stride=(1,))\n",
+              "        (8): BatchNorm1d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
+              "        (9): GELU(approximate='none')\n",
+              "        (10): Conv1d(64, 8, kernel_size=(1,), stride=(1,))\n",
+              "        (11): BatchNorm1d(8, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
+              "        (12): GELU(approximate='none')\n",
+              "      )\n",
+              "    )\n",
+              "    (encoder): ModuleList(\n",
+              "      (0-7): 8 x ParticleAttentionBlock(\n",
+              "        (layernorm1): LayerNorm((128,), eps=1e-05, elementwise_affine=True)\n",
+              "        (mass_norm): LayerNorm((1,), eps=1e-05, elementwise_affine=True)\n",
+              "        (physics_proj): Linear(in_features=8, out_features=128, bias=True)\n",
+              "        (mass_proj): Linear(in_features=1, out_features=128, bias=True)\n",
+              "        (gate_proj): Linear(in_features=128, out_features=8, bias=True)\n",
+              "        (pmha): MultiheadAttention(\n",
+              "          (out_proj): NonDynamicallyQuantizableLinear(in_features=128, out_features=128, bias=True)\n",
+              "        )\n",
+              "        (layernorm2): LayerNorm((128,), eps=1e-05, elementwise_affine=True)\n",
+              "        (dropout): Dropout(p=0.1, inplace=False)\n",
+              "        (feedforward): Feedforward(\n",
+              "          (layernorm1): LayerNorm((128,), eps=1e-05, elementwise_affine=True)\n",
+              "          (linear1): Linear(in_features=128, out_features=512, bias=True)\n",
+              "          (act): GELU(approximate='none')\n",
+              "          (dropout1): Dropout(p=0.1, inplace=False)\n",
+              "          (layernorm2): LayerNorm((512,), eps=1e-05, elementwise_affine=True)\n",
+              "          (linear2): Linear(in_features=512, out_features=128, bias=True)\n",
+              "          (dropout2): Dropout(p=0.1, inplace=False)\n",
+              "        )\n",
+              "      )\n",
+              "    )\n",
+              "  )\n",
+              "  (fc): Linear(in_features=16384, out_features=16, bias=True)\n",
+              "  (equilinear): EquiLinear()\n",
+              "  (decoder): ModuleList(\n",
+              "    (0-1): 2 x ClassAttentionBlock(\n",
+              "      (layernorm1): LayerNorm((128,), eps=1e-05, elementwise_affine=True)\n",
+              "      (mha): MultiheadAttention(\n",
+              "        (out_proj): NonDynamicallyQuantizableLinear(in_features=128, out_features=128, bias=True)\n",
+              "      )\n",
+              "      (layernorm2): LayerNorm((128,), eps=1e-05, elementwise_affine=True)\n",
+              "      (dropout): Dropout(p=0.0, inplace=False)\n",
+              "      (feedforward): Feedforward(\n",
+              "        (layernorm1): LayerNorm((128,), eps=1e-05, elementwise_affine=True)\n",
+              "        (linear1): Linear(in_features=128, out_features=512, bias=True)\n",
+              "        (act): GELU(approximate='none')\n",
+              "        (dropout1): Dropout(p=0.0, inplace=False)\n",
+              "        (layernorm2): LayerNorm((512,), eps=1e-05, elementwise_affine=True)\n",
+              "        (linear2): Linear(in_features=512, out_features=128, bias=True)\n",
+              "        (dropout2): Dropout(p=0.0, inplace=False)\n",
+              "      )\n",
+              "    )\n",
+              "  )\n",
+              "  (layernorm): LayerNorm((128,), eps=1e-05, elementwise_affine=True)\n",
+              "  (classifier): Classifier(\n",
+              "    (layers): Sequential(\n",
+              "      (0): Linear(in_features=128, out_features=10, bias=True)\n",
+              "    )\n",
+              "  )\n",
+              "  (act): Identity()\n",
+              ")"
+            ]
+          },
+          "metadata": {},
+          "execution_count": 9
+        }
+      ]
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "num_params = sum(p.numel() for p in gatedmodel.parameters() if p.requires_grad)\n",
+        "num_params"
+      ],
+      "metadata": {
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        },
+        "id": "n5GMAwY0Sw5t",
+        "outputId": "170f4cca-9434-4e28-d810-c0bf423f50a7"
+      },
+      "id": "n5GMAwY0Sw5t",
+      "execution_count": null,
+      "outputs": [
+        {
+          "output_type": "execute_result",
+          "data": {
+            "text/plain": [
+              "2290656"
+            ]
+          },
+          "metadata": {},
+          "execution_count": 10
+        }
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "id": "0aa870ab",
+      "metadata": {
+        "id": "0aa870ab"
+      },
+      "outputs": [],
+      "source": [
+        "# Training configurations\n",
+        "gated_train_config = TrainConfig(\n",
+        "    batch_size=128,\n",
+        "    criterion={\n",
+        "        'name': 'conservation_loss',\n",
+        "        'kwargs': {\n",
+        "            'loss_coef': [0.25, 0.25, 0.25, 0.25],\n",
+        "            'reduction': 'mean'\n",
+        "        }\n",
+        "    },\n",
+        "    optimizer={\n",
+        "        'name': 'adamw',\n",
+        "        'kwargs': {\n",
+        "            'lr': 1e-4\n",
+        "        }\n",
+        "    },\n",
+        "    scheduler={\n",
+        "        'name': 'exponential_lr',\n",
+        "        'kwargs': {\n",
+        "            'gamma': 0.95\n",
+        "        }\n",
+        "    },\n",
+        "    callbacks=[{\n",
+        "        'name': 'early_stopping',\n",
+        "        'kwargs': {\n",
+        "            'monitor': 'val_loss',\n",
+        "            'mode': 'min',\n",
+        "            'patience': 5\n",
+        "        }\n",
+        "    }],\n",
+        "    num_epochs=1,#20 change\n",
+        "    start_epoch=0,\n",
+        "    logging_dir=str(LOG_DIR),\n",
+        "    logging_steps=1000,\n",
+        "    progress_bar=True,\n",
+        "    save_best=True,\n",
+        "    save_ckpt=True,\n",
+        "    save_fig=False,\n",
+        "    device='cuda',\n",
+        "    num_workers=0,\n",
+        "    pin_memory=True\n",
+        ")"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "id": "6d5f08dd",
+      "metadata": {
+        "id": "6d5f08dd"
+      },
+      "outputs": [],
+      "source": [
+        "# Initialize the trainer\n",
+        "trainer = MaskedModelTrainer(\n",
+        "    model=gatedmodel,\n",
+        "    train_dataset=train_dataset,\n",
+        "    val_dataset=val_dataset,\n",
+        "    test_dataset=test_dataset,\n",
+        "    device=device,\n",
+        "    config=gated_train_config\n",
+        ")"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "id": "40a3b5fa",
+      "metadata": {
+        "colab": {
+          "base_uri": "https://localhost:8080/",
+          "height": 188,
+          "referenced_widgets": [
+            "0e1aad635bf14d1bb31c6f903f032e73",
+            "4e34acdc629c4fd898b1bbd743cd4cc1",
+            "f0e15ef2bc25482e87a2aafd881171fb",
+            "649dcf365ee140778713f1923422b40c",
+            "e0ef6f1060d2402a93eafa4557cd7a7a",
+            "45a2e6bd3f464dd0af70d71813826f89",
+            "b1ef8826aa5f4e2896ee4ee4bd6f5e2d",
+            "cb4aa040e0fe4b0292c3dcda6cc55b66",
+            "ee23434b787f454399e4be9c6111a94c",
+            "3c0bd3f7f0034da7ba1b118da5d111eb",
+            "bad0b3bd64c743519cf1ec7eb6b3ba46"
+          ]
+        },
+        "id": "40a3b5fa",
+        "outputId": "e38bc084-3ac9-47b3-ed52-af4af71a5b6e"
+      },
+      "outputs": [
+        {
+          "output_type": "display_data",
+          "data": {
+            "text/plain": [
+              "Training:   0%|          | 0/6250 [00:00<?, ?it/s]"
+            ],
+            "application/vnd.jupyter.widget-view+json": {
+              "version_major": 2,
+              "version_minor": 0,
+              "model_id": "0e1aad635bf14d1bb31c6f903f032e73"
+            }
+          },
+          "metadata": {}
+        },
+        {
+          "output_type": "stream",
+          "name": "stdout",
+          "text": [
+            "step: 1000/6250 | pT_loss: 0.4252 | eta_loss: 0.3509 | phi_loss: 0.2498 | energy_loss: 0.4427 | total_train_loss: 0.3671\n",
+            "step: 2000/6250 | pT_loss: 0.3485 | eta_loss: 0.2900 | phi_loss: 0.1602 | energy_loss: 0.3714 | total_train_loss: 0.2925\n",
+            "step: 3000/6250 | pT_loss: 0.3073 | eta_loss: 0.2562 | phi_loss: 0.1224 | energy_loss: 0.3311 | total_train_loss: 0.2542\n",
+            "step: 4000/6250 | pT_loss: 0.2823 | eta_loss: 0.2360 | phi_loss: 0.1020 | energy_loss: 0.3059 | total_train_loss: 0.2315\n",
+            "step: 5000/6250 | pT_loss: 0.2651 | eta_loss: 0.2231 | phi_loss: 0.0886 | energy_loss: 0.2876 | total_train_loss: 0.2161\n",
+            "step: 6000/6250 | pT_loss: 0.2517 | eta_loss: 0.2132 | phi_loss: 0.0785 | energy_loss: 0.2732 | total_train_loss: 0.2041\n",
+            "step: 6250/6250 | pT_loss: 0.2489 | eta_loss: 0.2112 | phi_loss: 0.0764 | energy_loss: 0.2702 | total_train_loss: 0.2017\n",
+            "epoch: 1/1 | pT_loss: 0.1795 | eta_loss: 0.1526 | phi_loss: 0.0532 | energy_loss: 0.1877 | total_val_loss: 0.1432\n"
+          ]
+        }
+      ],
+      "source": [
+        "# Train the model\n",
+        "gated_history, gatedmodel = trainer.train()"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "# --- Save for the Gated Architecture ---\n",
+        "\n",
+        "gated_pt_path = LOG_DIR / gatedmodel.__class__.__name__ / 'best' / '77gated_lorentz_part_v1.pt'\n",
+        "\n",
+        "\n",
+        "gated_pt_path.parent.mkdir(parents=True, exist_ok=True)\n",
+        "\n",
+        "# Save the full state_dict (Transformer + Trained Gates)\n",
+        "torch.save(gatedmodel.state_dict(), gated_pt_path)\n",
+        "\n",
+        "print(f\"Gated model saved to: {gated_pt_path}\")"
+      ],
+      "metadata": {
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        },
+        "id": "7PzuSjWnFUP0",
+        "outputId": "ee3a47a8-1ddb-4845-8372-766be3c7daa7"
+      },
+      "id": "7PzuSjWnFUP0",
+      "execution_count": null,
+      "outputs": [
+        {
+          "output_type": "stream",
+          "name": "stdout",
+          "text": [
+            "Gated model saved to: /content/CMS/MAEs/Hybrid_Transformer_Thanh_Nguyen/logs/LorentzParT/best/77gated_lorentz_part_v1.pt\n"
+          ]
+        }
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "id": "53aa0afd",
+      "metadata": {
+        "id": "53aa0afd"
+      },
+      "outputs": [],
+      "source": [
+        "trainer.load_best_model(gated_pt_path)"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "import torch\n",
+        "import os\n",
+        "\n",
+        "\n",
+        "# 2. Check if the file exists and show its size\n",
+        "if gated_pt_path.exists():\n",
+        "    file_size_mb = os.path.getsize(gated_pt_path) / (1024 * 1024)\n",
+        "    print(f\" File found at: {gated_pt_path}\")\n",
+        "    print(f\" File size: {file_size_mb:.2f} MB\")\n",
+        "\n",
+        "    # 3. Load the state_dict to inspect contents\n",
+        "    print(\"\\nInspecting state_dict keys...\")\n",
+        "    checkpoint = torch.load(gated_pt_path, map_location='cpu')\n",
+        "\n",
+        "    # 4. Specifically look for the new 'gate' parameters\n",
+        "    gate_keys = [k for k in checkpoint.keys() if 'gate' in k]\n",
+        "\n",
+        "    if gate_keys:\n",
+        "        print(f\" Found {len(gate_keys)} gating-related parameters.\")\n",
+        "        print(\"Example keys found:\")\n",
+        "        for key in gate_keys[:5]:  # Show the first 5 gate keys\n",
+        "            print(f\" - {key}\")\n",
+        "    else:\n",
+        "        print(\"No 'gate' parameters found. The model might have saved in the old format.\")\n",
+        "\n",
+        "    # Optional: Show total number of saved parameter tensors\n",
+        "    print(f\"\\nTotal parameter tensors saved: {len(checkpoint.keys())}\")\n",
+        "\n",
+        "else:\n",
+        "    print(f\"Error at {gated_pt_path}\")"
+      ],
+      "metadata": {
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        },
+        "id": "AhXjiLKWJUkY",
+        "outputId": "e53b0395-aa5b-4f05-d080-7316c0912480"
+      },
+      "id": "AhXjiLKWJUkY",
+      "execution_count": null,
+      "outputs": [
+        {
+          "output_type": "stream",
+          "name": "stdout",
+          "text": [
+            " File found at: /content/CMS/MAEs/Hybrid_Transformer_Thanh_Nguyen/logs/LorentzParT/best/gated_lorentz_part_v1.pt\n",
+            " File size: 9.25 MB\n",
+            "\n",
+            "Inspecting state_dict keys...\n",
+            " Found 16 gating-related parameters.\n",
+            "Example keys found:\n",
+            " - encoder.encoder.0.gate.0.weight\n",
+            " - encoder.encoder.0.gate.0.bias\n",
+            " - encoder.encoder.1.gate.0.weight\n",
+            " - encoder.encoder.1.gate.0.bias\n",
+            " - encoder.encoder.2.gate.0.weight\n",
+            "\n",
+            "Total parameter tensors saved: 222\n"
+          ]
+        }
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "id": "aaf64bd1",
+      "metadata": {
+        "id": "aaf64bd1",
+        "outputId": "d9b79c69-eeb2-4d2e-ca46-38840736c8bb",
+        "colab": {
+          "base_uri": "https://localhost:8080/",
+          "height": 1000
+        }
+      },
+      "outputs": [
+        {
+          "output_type": "stream",
+          "name": "stdout",
+          "text": [
+            "test_loss: 0.5052 | pT_loss: 0.8751 | eta_loss: 0.1247 | phi_loss: 0.0429 | energy_loss: 0.9780\n"
+          ]
+        },
+        {
+          "output_type": "display_data",
+          "data": {
+            "text/plain": [
+              "<Figure size 1200x1000 with 8 Axes>"
+            ],
+            "image/png": "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\n"
+          },
+          "metadata": {}
+        }
+      ],
+      "source": [
+        "# Evaluate the model on the test set\n",
+        "test_loss, test_metric, y_true, y_pred = trainer.evaluate(plot=plot_particle_reconstruction)"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "# Label names for classification\n",
+        "labels = [\n",
+        "    \"$q/g$\",  # 0\n",
+        "    \"$H \\\\to b\\\\bar{b}$\",  # 1\n",
+        "    \"$H \\\\to c\\\\bar{c}$\",  # 2\n",
+        "    \"$H \\\\to gg$\",  # 3\n",
+        "    \"$H \\\\to 4q$\",  # 4\n",
+        "    \"$H \\\\to \\\\ell \\\\nu qq'$\",  # 5\n",
+        "    \"$Z \\\\to q\\\\bar{q}$\",  # 6\n",
+        "    \"$W \\\\to qq'$\",  # 7\n",
+        "    \"$t \\\\to b\\\\ell \\\\nu$\",  # 8\n",
+        "    \"$t \\\\to bqq'$\"  # 9\n",
+        "]\n"
+      ],
+      "metadata": {
+        "id": "vt4vzC1TGBiB"
+      },
+      "id": "vt4vzC1TGBiB",
+      "execution_count": null,
+      "outputs": []
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "\n",
+        "# Datasets for classification\n",
+        "train_dataset = JetClassDataset(X_train, y_train, normalize, norm_dict, mask_mode=None)\n",
+        "val_dataset = JetClassDataset(X_val, y_val, normalize, norm_dict, mask_mode=None)\n",
+        "test_dataset = JetClassDataset(X_test, y_test, normalize, norm_dict, mask_mode=None)"
+      ],
+      "metadata": {
+        "id": "m37_eDUuGD0l"
+      },
+      "id": "m37_eDUuGD0l",
+      "execution_count": null,
+      "outputs": []
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "from collections import Counter\n",
+        "import numpy as np\n",
+        "\n",
+        "def check_uniformity(y, dataset_name, label_names, threshold=0.02):\n",
+        "    \"\"\"\n",
+        "    Checks if the labels in a dataset are uniformly distributed.\n",
+        "    Supports both integer class arrays and one-hot encoded arrays.\n",
+        "    \"\"\"\n",
+        "    # If one-hot encoded, convert to class indices\n",
+        "    if len(y.shape) > 1 and y.shape[1] > 1:\n",
+        "        y = np.argmax(y, axis=1)\n",
+        "\n",
+        "    total_samples = len(y)\n",
+        "    counts = Counter(y)\n",
+        "    num_classes = len(label_names)\n",
+        "    expected_pct = 1.0 / num_classes\n",
+        "\n",
+        "    print(f\"--- Distribution for {dataset_name} ({total_samples} samples) ---\")\n",
+        "\n",
+        "    is_uniform = True\n",
+        "    for idx, name in enumerate(label_names):\n",
+        "        count = counts.get(idx, 0)\n",
+        "        actual_pct = count / total_samples\n",
+        "        print(f\"Class {idx} ({name:<18}): {count:<8} | {actual_pct:.2%}\")\n",
+        "\n",
+        "        # Check if it deviates more than the allowed threshold from absolute uniformity\n",
+        "        if abs(actual_pct - expected_pct) > threshold:\n",
+        "            is_uniform = False\n",
+        "\n",
+        "    if is_uniform:\n",
+        "        print(f\"✅ {dataset_name} appears to be uniformly distributed (within a {threshold:.1%} tolerance).\\n\")\n",
+        "    else:\n",
+        "        print(f\"⚠️ {dataset_name} is NOT perfectly uniform. Expected around {expected_pct:.2%} per class.\\n\")\n",
+        "\n",
+        "# Run the check on your datasets\n",
+        "# (Using your raw arrays y_train, y_val, and y_test)\n",
+        "check_uniformity(y_train, \"Train Dataset\", labels)\n",
+        "check_uniformity(y_val, \"Validation Dataset\", labels)\n",
+        "check_uniformity(y_test, \"Test Dataset\", labels)"
+      ],
+      "metadata": {
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        },
+        "id": "IWCdSFFKGFoL",
+        "outputId": "c1769a19-a1db-473e-80a8-237fdc4867b1"
+      },
+      "id": "IWCdSFFKGFoL",
+      "execution_count": null,
+      "outputs": [
+        {
+          "output_type": "stream",
+          "name": "stdout",
+          "text": [
+            "--- Distribution for Train Dataset (800000 samples) ---\n",
+            "Class 0 ($q/g$             ): 80000    | 10.00%\n",
+            "Class 1 ($H \\to b\\bar{b}$  ): 80000    | 10.00%\n",
+            "Class 2 ($H \\to c\\bar{c}$  ): 80000    | 10.00%\n",
+            "Class 3 ($H \\to gg$        ): 80000    | 10.00%\n",
+            "Class 4 ($H \\to 4q$        ): 80000    | 10.00%\n",
+            "Class 5 ($H \\to \\ell \\nu qq'$): 80000    | 10.00%\n",
+            "Class 6 ($Z \\to q\\bar{q}$  ): 80000    | 10.00%\n",
+            "Class 7 ($W \\to qq'$       ): 80000    | 10.00%\n",
+            "Class 8 ($t \\to b\\ell \\nu$ ): 80000    | 10.00%\n",
+            "Class 9 ($t \\to bqq'$      ): 80000    | 10.00%\n",
+            "✅ Train Dataset appears to be uniformly distributed (within a 2.0% tolerance).\n",
+            "\n",
+            "--- Distribution for Validation Dataset (100000 samples) ---\n",
+            "Class 0 ($q/g$             ): 10000    | 10.00%\n",
+            "Class 1 ($H \\to b\\bar{b}$  ): 10000    | 10.00%\n",
+            "Class 2 ($H \\to c\\bar{c}$  ): 10000    | 10.00%\n",
+            "Class 3 ($H \\to gg$        ): 10000    | 10.00%\n",
+            "Class 4 ($H \\to 4q$        ): 10000    | 10.00%\n",
+            "Class 5 ($H \\to \\ell \\nu qq'$): 10000    | 10.00%\n",
+            "Class 6 ($Z \\to q\\bar{q}$  ): 10000    | 10.00%\n",
+            "Class 7 ($W \\to qq'$       ): 10000    | 10.00%\n",
+            "Class 8 ($t \\to b\\ell \\nu$ ): 10000    | 10.00%\n",
+            "Class 9 ($t \\to bqq'$      ): 10000    | 10.00%\n",
+            "✅ Validation Dataset appears to be uniformly distributed (within a 2.0% tolerance).\n",
+            "\n",
+            "--- Distribution for Test Dataset (100000 samples) ---\n",
+            "Class 0 ($q/g$             ): 10000    | 10.00%\n",
+            "Class 1 ($H \\to b\\bar{b}$  ): 10000    | 10.00%\n",
+            "Class 2 ($H \\to c\\bar{c}$  ): 10000    | 10.00%\n",
+            "Class 3 ($H \\to gg$        ): 10000    | 10.00%\n",
+            "Class 4 ($H \\to 4q$        ): 10000    | 10.00%\n",
+            "Class 5 ($H \\to \\ell \\nu qq'$): 10000    | 10.00%\n",
+            "Class 6 ($Z \\to q\\bar{q}$  ): 10000    | 10.00%\n",
+            "Class 7 ($W \\to qq'$       ): 10000    | 10.00%\n",
+            "Class 8 ($t \\to b\\ell \\nu$ ): 10000    | 10.00%\n",
+            "Class 9 ($t \\to bqq'$      ): 10000    | 10.00%\n",
+            "✅ Test Dataset appears to be uniformly distributed (within a 2.0% tolerance).\n",
+            "\n"
+          ]
+        }
+      ]
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "# Model configurations\n",
+        "pretrained_model_config = LorentzParTConfig(\n",
+        "    num_classes=10,\n",
+        "    embed_dim=128,\n",
+        "    num_heads=8,\n",
+        "    num_layers=8,\n",
+        "    num_cls_layers=2,\n",
+        "    num_mlp_layers=0,\n",
+        "    hidden_dim=256,\n",
+        "    hidden_mv_channels=8,\n",
+        "    in_s_channels=None,\n",
+        "    out_s_channels=None,\n",
+        "    hidden_s_channels=16,\n",
+        "    attention={},\n",
+        "    mlp={},\n",
+        "    dropout=0.1,\n",
+        "    expansion_factor=4,\n",
+        "    max_num_particles=128,\n",
+        "    num_particle_features=4,\n",
+        "    pair_embed_dims=[64, 64, 64],\n",
+        "    weights=gated_pt_path\n",
+        ")"
+      ],
+      "metadata": {
+        "id": "oRHuimXTGIdx"
+      },
+      "id": "oRHuimXTGIdx",
+      "execution_count": null,
+      "outputs": []
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "# Initialize the classifier model\n",
+        "pretrained_model = LorentzParT(config=pretrained_model_config).to(device)\n",
+        "pretrained_model"
+      ],
+      "metadata": {
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        },
+        "id": "Dv1oQ3tSGe3w",
+        "outputId": "93cf00a1-dc30-4880-8652-4adf5b6cccae"
+      },
+      "id": "Dv1oQ3tSGe3w",
+      "execution_count": null,
+      "outputs": [
+        {
+          "output_type": "execute_result",
+          "data": {
+            "text/plain": [
+              "LorentzParT(\n",
+              "  (processor): ParticleProcessor()\n",
+              "  (encoder): LorentzParTEncoder(\n",
+              "    (equilinear): EquiLinear()\n",
+              "    (proj): Linear(in_features=16, out_features=128, bias=True)\n",
+              "    (interaction_embed): InteractionEmbedding(\n",
+              "      (embed): Sequential(\n",
+              "        (0): BatchNorm1d(4, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
+              "        (1): Conv1d(4, 64, kernel_size=(1,), stride=(1,))\n",
+              "        (2): BatchNorm1d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
+              "        (3): GELU(approximate='none')\n",
+              "        (4): Conv1d(64, 64, kernel_size=(1,), stride=(1,))\n",
+              "        (5): BatchNorm1d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
+              "        (6): GELU(approximate='none')\n",
+              "        (7): Conv1d(64, 64, kernel_size=(1,), stride=(1,))\n",
+              "        (8): BatchNorm1d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
+              "        (9): GELU(approximate='none')\n",
+              "        (10): Conv1d(64, 8, kernel_size=(1,), stride=(1,))\n",
+              "        (11): BatchNorm1d(8, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
+              "        (12): GELU(approximate='none')\n",
+              "      )\n",
+              "    )\n",
+              "    (encoder): ModuleList(\n",
+              "      (0-7): 8 x ParticleAttentionBlock(\n",
+              "        (layernorm1): LayerNorm((128,), eps=1e-05, elementwise_affine=True)\n",
+              "        (mass_norm): LayerNorm((1,), eps=1e-05, elementwise_affine=True)\n",
+              "        (physics_proj): Linear(in_features=8, out_features=128, bias=True)\n",
+              "        (mass_proj): Linear(in_features=1, out_features=128, bias=True)\n",
+              "        (pmha): MultiheadAttention(\n",
+              "          (out_proj): NonDynamicallyQuantizableLinear(in_features=128, out_features=128, bias=True)\n",
+              "        )\n",
+              "        (layernorm2): LayerNorm((128,), eps=1e-05, elementwise_affine=True)\n",
+              "        (dropout): Dropout(p=0.1, inplace=False)\n",
+              "        (feedforward): Feedforward(\n",
+              "          (layernorm1): LayerNorm((128,), eps=1e-05, elementwise_affine=True)\n",
+              "          (linear1): Linear(in_features=128, out_features=512, bias=True)\n",
+              "          (act): GELU(approximate='none')\n",
+              "          (dropout1): Dropout(p=0.1, inplace=False)\n",
+              "          (layernorm2): LayerNorm((512,), eps=1e-05, elementwise_affine=True)\n",
+              "          (linear2): Linear(in_features=512, out_features=128, bias=True)\n",
+              "          (dropout2): Dropout(p=0.1, inplace=False)\n",
+              "        )\n",
+              "      )\n",
+              "    )\n",
+              "  )\n",
+              "  (fc): Linear(in_features=16384, out_features=16, bias=True)\n",
+              "  (equilinear): EquiLinear()\n",
+              "  (decoder): ModuleList(\n",
+              "    (0-1): 2 x ClassAttentionBlock(\n",
+              "      (layernorm1): LayerNorm((128,), eps=1e-05, elementwise_affine=True)\n",
+              "      (mha): MultiheadAttention(\n",
+              "        (out_proj): NonDynamicallyQuantizableLinear(in_features=128, out_features=128, bias=True)\n",
+              "      )\n",
+              "      (layernorm2): LayerNorm((128,), eps=1e-05, elementwise_affine=True)\n",
+              "      (dropout): Dropout(p=0.0, inplace=False)\n",
+              "      (feedforward): Feedforward(\n",
+              "        (layernorm1): LayerNorm((128,), eps=1e-05, elementwise_affine=True)\n",
+              "        (linear1): Linear(in_features=128, out_features=512, bias=True)\n",
+              "        (act): GELU(approximate='none')\n",
+              "        (dropout1): Dropout(p=0.0, inplace=False)\n",
+              "        (layernorm2): LayerNorm((512,), eps=1e-05, elementwise_affine=True)\n",
+              "        (linear2): Linear(in_features=512, out_features=128, bias=True)\n",
+              "        (dropout2): Dropout(p=0.0, inplace=False)\n",
+              "      )\n",
+              "    )\n",
+              "  )\n",
+              "  (layernorm): LayerNorm((128,), eps=1e-05, elementwise_affine=True)\n",
+              "  (classifier): Classifier(\n",
+              "    (layers): Sequential(\n",
+              "      (0): Linear(in_features=128, out_features=10, bias=True)\n",
+              "    )\n",
+              "  )\n",
+              "  (act): Identity()\n",
+              ")"
+            ]
+          },
+          "metadata": {},
+          "execution_count": 21
+        }
+      ]
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "# Count parameters in the model\n",
+        "num_params = sum(p.numel() for p in pretrained_model.parameters() if p.requires_grad)\n",
+        "num_params"
+      ],
+      "metadata": {
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        },
+        "id": "_TDNTq4hGL2e",
+        "outputId": "24c6d7c6-5565-44e1-ed00-17967f67ff76"
+      },
+      "id": "_TDNTq4hGL2e",
+      "execution_count": null,
+      "outputs": [
+        {
+          "output_type": "execute_result",
+          "data": {
+            "text/plain": [
+              "2282400"
+            ]
+          },
+          "metadata": {},
+          "execution_count": 22
+        }
+      ]
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "# Training configurations\n",
+        "pretrained_config = TrainConfig(\n",
+        "    batch_size=64,\n",
+        "    criterion={\n",
+        "        'name': 'cross_entropy_loss',\n",
+        "        'kwargs': {\n",
+        "            'reduction': 'mean'\n",
+        "        }\n",
+        "    },\n",
+        "    optimizer={\n",
+        "        'name': 'adam',\n",
+        "        'kwargs': {\n",
+        "            'lr': 1e-4\n",
+        "        }\n",
+        "    },\n",
+        "    scheduler={\n",
+        "        'name': 'exponential_lr',\n",
+        "        'kwargs': {\n",
+        "            'gamma': 0.95\n",
+        "        }\n",
+        "    },\n",
+        "    callbacks=[{\n",
+        "        'name': 'early_stopping',\n",
+        "        'kwargs': {\n",
+        "            'monitor': 'val_loss',\n",
+        "            'mode': 'min',\n",
+        "            'patience': 5\n",
+        "        }\n",
+        "    }],\n",
+        "    num_epochs=2,#change\n",
+        "    start_epoch=0,\n",
+        "    logging_dir=str(LOG_DIR),\n",
+        "    logging_steps=1000,\n",
+        "    save_best=True,\n",
+        "    save_ckpt=True,\n",
+        "    save_fig=False,\n",
+        "    device='cuda',\n",
+        "    num_workers=0,\n",
+        "    pin_memory=True\n",
+        ")"
+      ],
+      "metadata": {
+        "id": "aaUtfajZGWD5"
+      },
+      "id": "aaUtfajZGWD5",
+      "execution_count": null,
+      "outputs": []
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "# Initialize the trainer\n",
+        "trainer = Trainer(\n",
+        "    model=pretrained_model,\n",
+        "    train_dataset=train_dataset,\n",
+        "    val_dataset=val_dataset,\n",
+        "    test_dataset=test_dataset,\n",
+        "    device=device,\n",
+        "    metric=accuracy_metric_ce,\n",
+        "    config=pretrained_config\n",
+        ")"
+      ],
+      "metadata": {
+        "id": "GBfMEnRrGler"
+      },
+      "id": "GBfMEnRrGler",
+      "execution_count": null,
+      "outputs": []
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "# Train the model\n",
+        "pretrained_history, pretrained_model = trainer.train()"
+      ],
+      "metadata": {
+        "colab": {
+          "base_uri": "https://localhost:8080/",
+          "height": 518,
+          "referenced_widgets": [
+            "62f237b1683e475595fe17da0edeae87",
+            "7bdac98dfc4a4699bab39fa846e56354",
+            "ac7d5153146d4bd89fa5587a1c4babb7",
+            "1d65d195b47147d3806f7735255878d8",
+            "34f4dd1f0b9b4920bffb1027ea6e11fe",
+            "38e5079824d648849dcde09e2a2948fd",
+            "ffa3569f77fb4a3c8a6fb08930c0defb",
+            "dd800993260d48e383fb9aa27d265c7d",
+            "4d4b43abf49f4114a7fc83d9128f6d30",
+            "0d7bc861ca364e298ffb26510a4e4e09",
+            "2c776b732e6a4cb8bb85e01baff5fb33"
+          ]
+        },
+        "id": "q_3Zs8TfOwop",
+        "outputId": "ecbf6236-025e-47a4-f604-eb92b4fd72e0"
+      },
+      "id": "q_3Zs8TfOwop",
+      "execution_count": null,
+      "outputs": [
+        {
+          "output_type": "display_data",
+          "data": {
+            "text/plain": [
+              "Training:   0%|          | 0/25000 [00:00<?, ?it/s]"
+            ],
+            "application/vnd.jupyter.widget-view+json": {
+              "version_major": 2,
+              "version_minor": 0,
+              "model_id": "62f237b1683e475595fe17da0edeae87"
+            }
+          },
+          "metadata": {}
+        },
+        {
+          "output_type": "stream",
+          "name": "stdout",
+          "text": [
+            "step: 1000/25000 | train_loss: 2.0798 | train_metric: 0.1982\n",
+            "step: 2000/25000 | train_loss: 2.0421 | train_metric: 0.2096\n",
+            "step: 3000/25000 | train_loss: 2.0240 | train_metric: 0.2154\n",
+            "step: 4000/25000 | train_loss: 2.0156 | train_metric: 0.2194\n",
+            "step: 5000/25000 | train_loss: 2.0090 | train_metric: 0.2218\n",
+            "step: 6000/25000 | train_loss: 2.0028 | train_metric: 0.2243\n",
+            "step: 7000/25000 | train_loss: 1.9969 | train_metric: 0.2274\n",
+            "step: 8000/25000 | train_loss: 1.9923 | train_metric: 0.2302\n",
+            "step: 9000/25000 | train_loss: 1.9866 | train_metric: 0.2333\n",
+            "step: 10000/25000 | train_loss: 1.9811 | train_metric: 0.2366\n",
+            "step: 11000/25000 | train_loss: 1.9757 | train_metric: 0.2397\n",
+            "step: 12000/25000 | train_loss: 1.9700 | train_metric: 0.2427\n",
+            "epoch: 1/2 | val_loss: 1.9015 | val_metric: 0.2810\n",
+            "step: 13000/25000 | train_loss: 2.0145 | train_metric: 0.2376\n",
+            "step: 14000/25000 | train_loss: 2.0017 | train_metric: 0.2407\n",
+            "step: 15000/25000 | train_loss: 1.9962 | train_metric: 0.2421\n",
+            "step: 16000/25000 | train_loss: 1.9907 | train_metric: 0.2452\n",
+            "step: 17000/25000 | train_loss: 1.9856 | train_metric: 0.2474\n",
+            "step: 18000/25000 | train_loss: 1.9829 | train_metric: 0.2490\n",
+            "step: 19000/25000 | train_loss: 1.9799 | train_metric: 0.2506\n",
+            "step: 20000/25000 | train_loss: 1.9770 | train_metric: 0.2520\n",
+            "step: 21000/25000 | train_loss: 1.9740 | train_metric: 0.2535\n",
+            "step: 22000/25000 | train_loss: 1.9717 | train_metric: 0.2546\n",
+            "step: 23000/25000 | train_loss: 1.9690 | train_metric: 0.2557\n",
+            "step: 24000/25000 | train_loss: 1.9665 | train_metric: 0.2567\n",
+            "step: 25000/25000 | train_loss: 1.9639 | train_metric: 0.2580\n",
+            "epoch: 2/2 | val_loss: 2.0577 | val_metric: 0.2344\n"
+          ]
+        }
+      ]
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "# Train the model\n",
+        "#pretrained_history, pretrained_model = trainer.train()"
+      ],
+      "metadata": {
+        "colab": {
+          "base_uri": "https://localhost:8080/",
+          "height": 518,
+          "referenced_widgets": [
+            "41be1797614f4fa7830532d34158ac52",
+            "8aac226632524370abc787be435a9b48",
+            "e24c1cf41fcd442eb81ef418e92b24c6",
+            "dd026d80116749d19365780681500be9",
+            "b1703412347941edaab104e289f38221",
+            "2cab2b54d039403daa901bbb65203e42",
+            "8f5881e63ed54097b49c624676055437",
+            "48e35659a78b41cb95b8cbb39040de62",
+            "ff02466aec4344fa84813bb3eba626b1",
+            "53c475a9562d4af59d6947020af35778",
+            "9e9b3731cd824b0094835c6ce06f9265"
+          ]
+        },
+        "id": "7tCRI6jhGnlu",
+        "outputId": "db76ade7-aeaf-4dc3-95b6-014113bcdf50"
+      },
+      "id": "7tCRI6jhGnlu",
+      "execution_count": null,
+      "outputs": [
+        {
+          "output_type": "display_data",
+          "data": {
+            "text/plain": [
+              "Training:   0%|          | 0/25000 [00:00<?, ?it/s]"
+            ],
+            "application/vnd.jupyter.widget-view+json": {
+              "version_major": 2,
+              "version_minor": 0,
+              "model_id": "41be1797614f4fa7830532d34158ac52"
+            }
+          },
+          "metadata": {}
+        },
+        {
+          "output_type": "stream",
+          "name": "stdout",
+          "text": [
+            "step: 1000/25000 | train_loss: 2.0122 | train_metric: 0.2181\n",
+            "step: 2000/25000 | train_loss: 1.9986 | train_metric: 0.2225\n",
+            "step: 3000/25000 | train_loss: 1.9925 | train_metric: 0.2248\n",
+            "step: 4000/25000 | train_loss: 1.9885 | train_metric: 0.2275\n",
+            "step: 5000/25000 | train_loss: 1.9862 | train_metric: 0.2289\n",
+            "step: 6000/25000 | train_loss: 1.9842 | train_metric: 0.2299\n",
+            "step: 7000/25000 | train_loss: 1.9826 | train_metric: 0.2312\n",
+            "step: 8000/25000 | train_loss: 1.9793 | train_metric: 0.2335\n",
+            "step: 9000/25000 | train_loss: 1.9751 | train_metric: 0.2365\n",
+            "step: 10000/25000 | train_loss: 1.9697 | train_metric: 0.2399\n",
+            "step: 11000/25000 | train_loss: 1.9642 | train_metric: 0.2434\n",
+            "step: 12000/25000 | train_loss: 1.9574 | train_metric: 0.2470\n",
+            "epoch: 1/2 | val_loss: 1.8328 | val_metric: 0.3105\n",
+            "step: 13000/25000 | train_loss: 2.0182 | train_metric: 0.2404\n",
+            "step: 14000/25000 | train_loss: 1.9705 | train_metric: 0.2592\n",
+            "step: 15000/25000 | train_loss: 1.9542 | train_metric: 0.2658\n",
+            "step: 16000/25000 | train_loss: 1.9411 | train_metric: 0.2724\n",
+            "step: 17000/25000 | train_loss: 1.9335 | train_metric: 0.2758\n",
+            "step: 18000/25000 | train_loss: 1.9281 | train_metric: 0.2781\n",
+            "step: 19000/25000 | train_loss: 1.9217 | train_metric: 0.2812\n",
+            "step: 20000/25000 | train_loss: 1.9161 | train_metric: 0.2836\n",
+            "step: 21000/25000 | train_loss: 1.9108 | train_metric: 0.2859\n",
+            "step: 22000/25000 | train_loss: 1.9063 | train_metric: 0.2881\n",
+            "step: 23000/25000 | train_loss: 1.9016 | train_metric: 0.2900\n",
+            "step: 24000/25000 | train_loss: 1.8971 | train_metric: 0.2922\n",
+            "step: 25000/25000 | train_loss: 1.8927 | train_metric: 0.2941\n",
+            "epoch: 2/2 | val_loss: 1.8720 | val_metric: 0.3026\n"
+          ]
+        }
+      ]
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "# Evaluate the model on the test set\n",
+        "test_loss, test_metric, y_true, y_pred = trainer.evaluate(\n",
+        "    loss_type='cross_entropy',\n",
+        "    plot=[plot_roc_curve, plot_confusion_matrix]\n",
+        ")"
+      ],
+      "metadata": {
+        "id": "HovrFS4VGpXI",
+        "colab": {
+          "base_uri": "https://localhost:8080/",
+          "height": 1000
+        },
+        "outputId": "fa502085-bb30-455f-d1bd-494bed68ee3a"
+      },
+      "id": "HovrFS4VGpXI",
+      "execution_count": null,
+      "outputs": [
+        {
+          "output_type": "stream",
+          "name": "stdout",
+          "text": [
+            "test_loss: 2.1478 | test_metric: 0.1947\n"
+          ]
+        },
+        {
+          "output_type": "display_data",
+          "data": {
+            "text/plain": [
+              "<Figure size 600x500 with 1 Axes>"
+            ],
+            "image/png": "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\n"
+          },
+          "metadata": {}
+        },
+        {
+          "output_type": "display_data",
+          "data": {
+            "text/plain": [
+              "<Figure size 600x500 with 2 Axes>"
+            ],
+            "image/png": "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\n"
+          },
+          "metadata": {}
+        }
+      ]
+    }
+  ],
+  "metadata": {
+    "kernelspec": {
+      "display_name": "Python 3",
+      "name": "python3"
+    },
+    "language_info": {
+      "codemirror_mode": {
+        "name": "ipython",
+        "version": 3
+      },
+      "file_extension": ".py",
+      "mimetype": "text/x-python",
+      "name": "python",
+      "nbconvert_exporter": "python",
+      "pygments_lexer": "ipython3",
+      "version": "3.13.5"
+    },
+    "colab": {
+      "provenance": [],
+      "gpuType": "A100"
+    },
+    "accelerator": "GPU",
+    "widgets": {
+      "application/vnd.jupyter.widget-state+json": {
+        "version_major": 2,
+        "version_minor": 0,
+        "state": {
+          "41be1797614f4fa7830532d34158ac52": {
+            "model_module": "@jupyter-widgets/controls",
+            "model_name": "HBoxModel",
+            "model_module_version": "1.5.0",
+            "state": {
+              "_dom_classes": [],
+              "_model_module": "@jupyter-widgets/controls",
+              "_model_module_version": "1.5.0",
+              "_model_name": "HBoxModel",
+              "_view_count": null,
+              "_view_module": "@jupyter-widgets/controls",
+              "_view_module_version": "1.5.0",
+              "_view_name": "HBoxView",
+              "box_style": "",
+              "children": [
+                "IPY_MODEL_8aac226632524370abc787be435a9b48",
+                "IPY_MODEL_e24c1cf41fcd442eb81ef418e92b24c6",
+                "IPY_MODEL_dd026d80116749d19365780681500be9"
+              ],
+              "layout": "IPY_MODEL_b1703412347941edaab104e289f38221"
+            }
+          },
+          "8aac226632524370abc787be435a9b48": {
+            "model_module": "@jupyter-widgets/controls",
+            "model_name": "HTMLModel",
+            "model_module_version": "1.5.0",
+            "state": {
+              "_dom_classes": [],
+              "_model_module": "@jupyter-widgets/controls",
+              "_model_module_version": "1.5.0",
+              "_model_name": "HTMLModel",
+              "_view_count": null,
+              "_view_module": "@jupyter-widgets/controls",
+              "_view_module_version": "1.5.0",
+              "_view_name": "HTMLView",
+              "description": "",
+              "description_tooltip": null,
+              "layout": "IPY_MODEL_2cab2b54d039403daa901bbb65203e42",
+              "placeholder": "​",
+              "style": "IPY_MODEL_8f5881e63ed54097b49c624676055437",
+              "value": "Training: 100%"
+            }
+          },
+          "e24c1cf41fcd442eb81ef418e92b24c6": {
+            "model_module": "@jupyter-widgets/controls",
+            "model_name": "FloatProgressModel",
+            "model_module_version": "1.5.0",
+            "state": {
+              "_dom_classes": [],
+              "_model_module": "@jupyter-widgets/controls",
+              "_model_module_version": "1.5.0",
+              "_model_name": "FloatProgressModel",
+              "_view_count": null,
+              "_view_module": "@jupyter-widgets/controls",
+              "_view_module_version": "1.5.0",
+              "_view_name": "ProgressView",
+              "bar_style": "success",
+              "description": "",
+              "description_tooltip": null,
+              "layout": "IPY_MODEL_48e35659a78b41cb95b8cbb39040de62",
+              "max": 25000,
+              "min": 0,
+              "orientation": "horizontal",
+              "style": "IPY_MODEL_ff02466aec4344fa84813bb3eba626b1",
+              "value": 25000
+            }
+          },
+          "dd026d80116749d19365780681500be9": {
+            "model_module": "@jupyter-widgets/controls",
+            "model_name": "HTMLModel",
+            "model_module_version": "1.5.0",
+            "state": {
+              "_dom_classes": [],
+              "_model_module": "@jupyter-widgets/controls",
+              "_model_module_version": "1.5.0",
+              "_model_name": "HTMLModel",
+              "_view_count": null,
+              "_view_module": "@jupyter-widgets/controls",
+              "_view_module_version": "1.5.0",
+              "_view_name": "HTMLView",
+              "description": "",
+              "description_tooltip": null,
+              "layout": "IPY_MODEL_53c475a9562d4af59d6947020af35778",
+              "placeholder": "​",
+              "style": "IPY_MODEL_9e9b3731cd824b0094835c6ce06f9265",
+              "value": " 25000/25000 [24:47&lt;00:00, 17.01it/s, epoch=2/2, avg_loss=1.8927, avg_metric=0.2941]"
+            }
+          },
+          "b1703412347941edaab104e289f38221": {
+            "model_module": "@jupyter-widgets/base",
+            "model_name": "LayoutModel",
+            "model_module_version": "1.2.0",
+            "state": {
+              "_model_module": "@jupyter-widgets/base",
+              "_model_module_version": "1.2.0",
+              "_model_name": "LayoutModel",
+              "_view_count": null,
+              "_view_module": "@jupyter-widgets/base",
+              "_view_module_version": "1.2.0",
+              "_view_name": "LayoutView",
+              "align_content": null,
+              "align_items": null,
+              "align_self": null,
+              "border": null,
+              "bottom": null,
+              "display": "inline-flex",
+              "flex": null,
+              "flex_flow": "row wrap",
+              "grid_area": null,
+              "grid_auto_columns": null,
+              "grid_auto_flow": null,
+              "grid_auto_rows": null,
+              "grid_column": null,
+              "grid_gap": null,
+              "grid_row": null,
+              "grid_template_areas": null,
+              "grid_template_columns": null,
+              "grid_template_rows": null,
+              "height": null,
+              "justify_content": null,
+              "justify_items": null,
+              "left": null,
+              "margin": null,
+              "max_height": null,
+              "max_width": null,
+              "min_height": null,
+              "min_width": null,
+              "object_fit": null,
+              "object_position": null,
+              "order": null,
+              "overflow": null,
+              "overflow_x": null,
+              "overflow_y": null,
+              "padding": null,
+              "right": null,
+              "top": null,
+              "visibility": null,
+              "width": "100%"
+            }
+          },
+          "2cab2b54d039403daa901bbb65203e42": {
+            "model_module": "@jupyter-widgets/base",
+            "model_name": "LayoutModel",
+            "model_module_version": "1.2.0",
+            "state": {
+              "_model_module": "@jupyter-widgets/base",
+              "_model_module_version": "1.2.0",
+              "_model_name": "LayoutModel",
+              "_view_count": null,
+              "_view_module": "@jupyter-widgets/base",
+              "_view_module_version": "1.2.0",
+              "_view_name": "LayoutView",
+              "align_content": null,
+              "align_items": null,
+              "align_self": null,
+              "border": null,
+              "bottom": null,
+              "display": null,
+              "flex": null,
+              "flex_flow": null,
+              "grid_area": null,
+              "grid_auto_columns": null,
+              "grid_auto_flow": null,
+              "grid_auto_rows": null,
+              "grid_column": null,
+              "grid_gap": null,
+              "grid_row": null,
+              "grid_template_areas": null,
+              "grid_template_columns": null,
+              "grid_template_rows": null,
+              "height": null,
+              "justify_content": null,
+              "justify_items": null,
+              "left": null,
+              "margin": null,
+              "max_height": null,
+              "max_width": null,
+              "min_height": null,
+              "min_width": null,
+              "object_fit": null,
+              "object_position": null,
+              "order": null,
+              "overflow": null,
+              "overflow_x": null,
+              "overflow_y": null,
+              "padding": null,
+              "right": null,
+              "top": null,
+              "visibility": null,
+              "width": null
+            }
+          },
+          "8f5881e63ed54097b49c624676055437": {
+            "model_module": "@jupyter-widgets/controls",
+            "model_name": "DescriptionStyleModel",
+            "model_module_version": "1.5.0",
+            "state": {
+              "_model_module": "@jupyter-widgets/controls",
+              "_model_module_version": "1.5.0",
+              "_model_name": "DescriptionStyleModel",
+              "_view_count": null,
+              "_view_module": "@jupyter-widgets/base",
+              "_view_module_version": "1.2.0",
+              "_view_name": "StyleView",
+              "description_width": ""
+            }
+          },
+          "48e35659a78b41cb95b8cbb39040de62": {
+            "model_module": "@jupyter-widgets/base",
+            "model_name": "LayoutModel",
+            "model_module_version": "1.2.0",
+            "state": {
+              "_model_module": "@jupyter-widgets/base",
+              "_model_module_version": "1.2.0",
+              "_model_name": "LayoutModel",
+              "_view_count": null,
+              "_view_module": "@jupyter-widgets/base",
+              "_view_module_version": "1.2.0",
+              "_view_name": "LayoutView",
+              "align_content": null,
+              "align_items": null,
+              "align_self": null,
+              "border": null,
+              "bottom": null,
+              "display": null,
+              "flex": "2",
+              "flex_flow": null,
+              "grid_area": null,
+              "grid_auto_columns": null,
+              "grid_auto_flow": null,
+              "grid_auto_rows": null,
+              "grid_column": null,
+              "grid_gap": null,
+              "grid_row": null,
+              "grid_template_areas": null,
+              "grid_template_columns": null,
+              "grid_template_rows": null,
+              "height": null,
+              "justify_content": null,
+              "justify_items": null,
+              "left": null,
+              "margin": null,
+              "max_height": null,
+              "max_width": null,
+              "min_height": null,
+              "min_width": null,
+              "object_fit": null,
+              "object_position": null,
+              "order": null,
+              "overflow": null,
+              "overflow_x": null,
+              "overflow_y": null,
+              "padding": null,
+              "right": null,
+              "top": null,
+              "visibility": null,
+              "width": null
+            }
+          },
+          "ff02466aec4344fa84813bb3eba626b1": {
+            "model_module": "@jupyter-widgets/controls",
+            "model_name": "ProgressStyleModel",
+            "model_module_version": "1.5.0",
+            "state": {
+              "_model_module": "@jupyter-widgets/controls",
+              "_model_module_version": "1.5.0",
+              "_model_name": "ProgressStyleModel",
+              "_view_count": null,
+              "_view_module": "@jupyter-widgets/base",
+              "_view_module_version": "1.2.0",
+              "_view_name": "StyleView",
+              "bar_color": null,
+              "description_width": ""
+            }
+          },
+          "53c475a9562d4af59d6947020af35778": {
+            "model_module": "@jupyter-widgets/base",
+            "model_name": "LayoutModel",
+            "model_module_version": "1.2.0",
+            "state": {
+              "_model_module": "@jupyter-widgets/base",
+              "_model_module_version": "1.2.0",
+              "_model_name": "LayoutModel",
+              "_view_count": null,
+              "_view_module": "@jupyter-widgets/base",
+              "_view_module_version": "1.2.0",
+              "_view_name": "LayoutView",
+              "align_content": null,
+              "align_items": null,
+              "align_self": null,
+              "border": null,
+              "bottom": null,
+              "display": null,
+              "flex": null,
+              "flex_flow": null,
+              "grid_area": null,
+              "grid_auto_columns": null,
+              "grid_auto_flow": null,
+              "grid_auto_rows": null,
+              "grid_column": null,
+              "grid_gap": null,
+              "grid_row": null,
+              "grid_template_areas": null,
+              "grid_template_columns": null,
+              "grid_template_rows": null,
+              "height": null,
+              "justify_content": null,
+              "justify_items": null,
+              "left": null,
+              "margin": null,
+              "max_height": null,
+              "max_width": null,
+              "min_height": null,
+              "min_width": null,
+              "object_fit": null,
+              "object_position": null,
+              "order": null,
+              "overflow": null,
+              "overflow_x": null,
+              "overflow_y": null,
+              "padding": null,
+              "right": null,
+              "top": null,
+              "visibility": null,
+              "width": null
+            }
+          },
+          "9e9b3731cd824b0094835c6ce06f9265": {
+            "model_module": "@jupyter-widgets/controls",
+            "model_name": "DescriptionStyleModel",
+            "model_module_version": "1.5.0",
+            "state": {
+              "_model_module": "@jupyter-widgets/controls",
+              "_model_module_version": "1.5.0",
+              "_model_name": "DescriptionStyleModel",
+              "_view_count": null,
+              "_view_module": "@jupyter-widgets/base",
+              "_view_module_version": "1.2.0",
+              "_view_name": "StyleView",
+              "description_width": ""
+            }
+          },
+          "0e1aad635bf14d1bb31c6f903f032e73": {
+            "model_module": "@jupyter-widgets/controls",
+            "model_name": "HBoxModel",
+            "model_module_version": "1.5.0",
+            "state": {
+              "_dom_classes": [],
+              "_model_module": "@jupyter-widgets/controls",
+              "_model_module_version": "1.5.0",
+              "_model_name": "HBoxModel",
+              "_view_count": null,
+              "_view_module": "@jupyter-widgets/controls",
+              "_view_module_version": "1.5.0",
+              "_view_name": "HBoxView",
+              "box_style": "",
+              "children": [
+                "IPY_MODEL_4e34acdc629c4fd898b1bbd743cd4cc1",
+                "IPY_MODEL_f0e15ef2bc25482e87a2aafd881171fb",
+                "IPY_MODEL_649dcf365ee140778713f1923422b40c"
+              ],
+              "layout": "IPY_MODEL_e0ef6f1060d2402a93eafa4557cd7a7a"
+            }
+          },
+          "4e34acdc629c4fd898b1bbd743cd4cc1": {
+            "model_module": "@jupyter-widgets/controls",
+            "model_name": "HTMLModel",
+            "model_module_version": "1.5.0",
+            "state": {
+              "_dom_classes": [],
+              "_model_module": "@jupyter-widgets/controls",
+              "_model_module_version": "1.5.0",
+              "_model_name": "HTMLModel",
+              "_view_count": null,
+              "_view_module": "@jupyter-widgets/controls",
+              "_view_module_version": "1.5.0",
+              "_view_name": "HTMLView",
+              "description": "",
+              "description_tooltip": null,
+              "layout": "IPY_MODEL_45a2e6bd3f464dd0af70d71813826f89",
+              "placeholder": "​",
+              "style": "IPY_MODEL_b1ef8826aa5f4e2896ee4ee4bd6f5e2d",
+              "value": "Training: 100%"
+            }
+          },
+          "f0e15ef2bc25482e87a2aafd881171fb": {
+            "model_module": "@jupyter-widgets/controls",
+            "model_name": "FloatProgressModel",
+            "model_module_version": "1.5.0",
+            "state": {
+              "_dom_classes": [],
+              "_model_module": "@jupyter-widgets/controls",
+              "_model_module_version": "1.5.0",
+              "_model_name": "FloatProgressModel",
+              "_view_count": null,
+              "_view_module": "@jupyter-widgets/controls",
+              "_view_module_version": "1.5.0",
+              "_view_name": "ProgressView",
+              "bar_style": "success",
+              "description": "",
+              "description_tooltip": null,
+              "layout": "IPY_MODEL_cb4aa040e0fe4b0292c3dcda6cc55b66",
+              "max": 6250,
+              "min": 0,
+              "orientation": "horizontal",
+              "style": "IPY_MODEL_ee23434b787f454399e4be9c6111a94c",
+              "value": 6250
+            }
+          },
+          "649dcf365ee140778713f1923422b40c": {
+            "model_module": "@jupyter-widgets/controls",
+            "model_name": "HTMLModel",
+            "model_module_version": "1.5.0",
+            "state": {
+              "_dom_classes": [],
+              "_model_module": "@jupyter-widgets/controls",
+              "_model_module_version": "1.5.0",
+              "_model_name": "HTMLModel",
+              "_view_count": null,
+              "_view_module": "@jupyter-widgets/controls",
+              "_view_module_version": "1.5.0",
+              "_view_name": "HTMLView",
+              "description": "",
+              "description_tooltip": null,
+              "layout": "IPY_MODEL_3c0bd3f7f0034da7ba1b118da5d111eb",
+              "placeholder": "​",
+              "style": "IPY_MODEL_bad0b3bd64c743519cf1ec7eb6b3ba46",
+              "value": " 6250/6250 [39:03&lt;00:00,  8.20it/s, epoch=1/1, avg_loss=0.2017]"
+            }
+          },
+          "e0ef6f1060d2402a93eafa4557cd7a7a": {
+            "model_module": "@jupyter-widgets/base",
+            "model_name": "LayoutModel",
+            "model_module_version": "1.2.0",
+            "state": {
+              "_model_module": "@jupyter-widgets/base",
+              "_model_module_version": "1.2.0",
+              "_model_name": "LayoutModel",
+              "_view_count": null,
+              "_view_module": "@jupyter-widgets/base",
+              "_view_module_version": "1.2.0",
+              "_view_name": "LayoutView",
+              "align_content": null,
+              "align_items": null,
+              "align_self": null,
+              "border": null,
+              "bottom": null,
+              "display": "inline-flex",
+              "flex": null,
+              "flex_flow": "row wrap",
+              "grid_area": null,
+              "grid_auto_columns": null,
+              "grid_auto_flow": null,
+              "grid_auto_rows": null,
+              "grid_column": null,
+              "grid_gap": null,
+              "grid_row": null,
+              "grid_template_areas": null,
+              "grid_template_columns": null,
+              "grid_template_rows": null,
+              "height": null,
+              "justify_content": null,
+              "justify_items": null,
+              "left": null,
+              "margin": null,
+              "max_height": null,
+              "max_width": null,
+              "min_height": null,
+              "min_width": null,
+              "object_fit": null,
+              "object_position": null,
+              "order": null,
+              "overflow": null,
+              "overflow_x": null,
+              "overflow_y": null,
+              "padding": null,
+              "right": null,
+              "top": null,
+              "visibility": null,
+              "width": "100%"
+            }
+          },
+          "45a2e6bd3f464dd0af70d71813826f89": {
+            "model_module": "@jupyter-widgets/base",
+            "model_name": "LayoutModel",
+            "model_module_version": "1.2.0",
+            "state": {
+              "_model_module": "@jupyter-widgets/base",
+              "_model_module_version": "1.2.0",
+              "_model_name": "LayoutModel",
+              "_view_count": null,
+              "_view_module": "@jupyter-widgets/base",
+              "_view_module_version": "1.2.0",
+              "_view_name": "LayoutView",
+              "align_content": null,
+              "align_items": null,
+              "align_self": null,
+              "border": null,
+              "bottom": null,
+              "display": null,
+              "flex": null,
+              "flex_flow": null,
+              "grid_area": null,
+              "grid_auto_columns": null,
+              "grid_auto_flow": null,
+              "grid_auto_rows": null,
+              "grid_column": null,
+              "grid_gap": null,
+              "grid_row": null,
+              "grid_template_areas": null,
+              "grid_template_columns": null,
+              "grid_template_rows": null,
+              "height": null,
+              "justify_content": null,
+              "justify_items": null,
+              "left": null,
+              "margin": null,
+              "max_height": null,
+              "max_width": null,
+              "min_height": null,
+              "min_width": null,
+              "object_fit": null,
+              "object_position": null,
+              "order": null,
+              "overflow": null,
+              "overflow_x": null,
+              "overflow_y": null,
+              "padding": null,
+              "right": null,
+              "top": null,
+              "visibility": null,
+              "width": null
+            }
+          },
+          "b1ef8826aa5f4e2896ee4ee4bd6f5e2d": {
+            "model_module": "@jupyter-widgets/controls",
+            "model_name": "DescriptionStyleModel",
+            "model_module_version": "1.5.0",
+            "state": {
+              "_model_module": "@jupyter-widgets/controls",
+              "_model_module_version": "1.5.0",
+              "_model_name": "DescriptionStyleModel",
+              "_view_count": null,
+              "_view_module": "@jupyter-widgets/base",
+              "_view_module_version": "1.2.0",
+              "_view_name": "StyleView",
+              "description_width": ""
+            }
+          },
+          "cb4aa040e0fe4b0292c3dcda6cc55b66": {
+            "model_module": "@jupyter-widgets/base",
+            "model_name": "LayoutModel",
+            "model_module_version": "1.2.0",
+            "state": {
+              "_model_module": "@jupyter-widgets/base",
+              "_model_module_version": "1.2.0",
+              "_model_name": "LayoutModel",
+              "_view_count": null,
+              "_view_module": "@jupyter-widgets/base",
+              "_view_module_version": "1.2.0",
+              "_view_name": "LayoutView",
+              "align_content": null,
+              "align_items": null,
+              "align_self": null,
+              "border": null,
+              "bottom": null,
+              "display": null,
+              "flex": "2",
+              "flex_flow": null,
+              "grid_area": null,
+              "grid_auto_columns": null,
+              "grid_auto_flow": null,
+              "grid_auto_rows": null,
+              "grid_column": null,
+              "grid_gap": null,
+              "grid_row": null,
+              "grid_template_areas": null,
+              "grid_template_columns": null,
+              "grid_template_rows": null,
+              "height": null,
+              "justify_content": null,
+              "justify_items": null,
+              "left": null,
+              "margin": null,
+              "max_height": null,
+              "max_width": null,
+              "min_height": null,
+              "min_width": null,
+              "object_fit": null,
+              "object_position": null,
+              "order": null,
+              "overflow": null,
+              "overflow_x": null,
+              "overflow_y": null,
+              "padding": null,
+              "right": null,
+              "top": null,
+              "visibility": null,
+              "width": null
+            }
+          },
+          "ee23434b787f454399e4be9c6111a94c": {
+            "model_module": "@jupyter-widgets/controls",
+            "model_name": "ProgressStyleModel",
+            "model_module_version": "1.5.0",
+            "state": {
+              "_model_module": "@jupyter-widgets/controls",
+              "_model_module_version": "1.5.0",
+              "_model_name": "ProgressStyleModel",
+              "_view_count": null,
+              "_view_module": "@jupyter-widgets/base",
+              "_view_module_version": "1.2.0",
+              "_view_name": "StyleView",
+              "bar_color": null,
+              "description_width": ""
+            }
+          },
+          "3c0bd3f7f0034da7ba1b118da5d111eb": {
+            "model_module": "@jupyter-widgets/base",
+            "model_name": "LayoutModel",
+            "model_module_version": "1.2.0",
+            "state": {
+              "_model_module": "@jupyter-widgets/base",
+              "_model_module_version": "1.2.0",
+              "_model_name": "LayoutModel",
+              "_view_count": null,
+              "_view_module": "@jupyter-widgets/base",
+              "_view_module_version": "1.2.0",
+              "_view_name": "LayoutView",
+              "align_content": null,
+              "align_items": null,
+              "align_self": null,
+              "border": null,
+              "bottom": null,
+              "display": null,
+              "flex": null,
+              "flex_flow": null,
+              "grid_area": null,
+              "grid_auto_columns": null,
+              "grid_auto_flow": null,
+              "grid_auto_rows": null,
+              "grid_column": null,
+              "grid_gap": null,
+              "grid_row": null,
+              "grid_template_areas": null,
+              "grid_template_columns": null,
+              "grid_template_rows": null,
+              "height": null,
+              "justify_content": null,
+              "justify_items": null,
+              "left": null,
+              "margin": null,
+              "max_height": null,
+              "max_width": null,
+              "min_height": null,
+              "min_width": null,
+              "object_fit": null,
+              "object_position": null,
+              "order": null,
+              "overflow": null,
+              "overflow_x": null,
+              "overflow_y": null,
+              "padding": null,
+              "right": null,
+              "top": null,
+              "visibility": null,
+              "width": null
+            }
+          },
+          "bad0b3bd64c743519cf1ec7eb6b3ba46": {
+            "model_module": "@jupyter-widgets/controls",
+            "model_name": "DescriptionStyleModel",
+            "model_module_version": "1.5.0",
+            "state": {
+              "_model_module": "@jupyter-widgets/controls",
+              "_model_module_version": "1.5.0",
+              "_model_name": "DescriptionStyleModel",
+              "_view_count": null,
+              "_view_module": "@jupyter-widgets/base",
+              "_view_module_version": "1.2.0",
+              "_view_name": "StyleView",
+              "description_width": ""
+            }
+          },
+          "62f237b1683e475595fe17da0edeae87": {
+            "model_module": "@jupyter-widgets/controls",
+            "model_name": "HBoxModel",
+            "model_module_version": "1.5.0",
+            "state": {
+              "_dom_classes": [],
+              "_model_module": "@jupyter-widgets/controls",
+              "_model_module_version": "1.5.0",
+              "_model_name": "HBoxModel",
+              "_view_count": null,
+              "_view_module": "@jupyter-widgets/controls",
+              "_view_module_version": "1.5.0",
+              "_view_name": "HBoxView",
+              "box_style": "",
+              "children": [
+                "IPY_MODEL_7bdac98dfc4a4699bab39fa846e56354",
+                "IPY_MODEL_ac7d5153146d4bd89fa5587a1c4babb7",
+                "IPY_MODEL_1d65d195b47147d3806f7735255878d8"
+              ],
+              "layout": "IPY_MODEL_34f4dd1f0b9b4920bffb1027ea6e11fe"
+            }
+          },
+          "7bdac98dfc4a4699bab39fa846e56354": {
+            "model_module": "@jupyter-widgets/controls",
+            "model_name": "HTMLModel",
+            "model_module_version": "1.5.0",
+            "state": {
+              "_dom_classes": [],
+              "_model_module": "@jupyter-widgets/controls",
+              "_model_module_version": "1.5.0",
+              "_model_name": "HTMLModel",
+              "_view_count": null,
+              "_view_module": "@jupyter-widgets/controls",
+              "_view_module_version": "1.5.0",
+              "_view_name": "HTMLView",
+              "description": "",
+              "description_tooltip": null,
+              "layout": "IPY_MODEL_38e5079824d648849dcde09e2a2948fd",
+              "placeholder": "​",
+              "style": "IPY_MODEL_ffa3569f77fb4a3c8a6fb08930c0defb",
+              "value": "Training: 100%"
+            }
+          },
+          "ac7d5153146d4bd89fa5587a1c4babb7": {
+            "model_module": "@jupyter-widgets/controls",
+            "model_name": "FloatProgressModel",
+            "model_module_version": "1.5.0",
+            "state": {
+              "_dom_classes": [],
+              "_model_module": "@jupyter-widgets/controls",
+              "_model_module_version": "1.5.0",
+              "_model_name": "FloatProgressModel",
+              "_view_count": null,
+              "_view_module": "@jupyter-widgets/controls",
+              "_view_module_version": "1.5.0",
+              "_view_name": "ProgressView",
+              "bar_style": "success",
+              "description": "",
+              "description_tooltip": null,
+              "layout": "IPY_MODEL_dd800993260d48e383fb9aa27d265c7d",
+              "max": 25000,
+              "min": 0,
+              "orientation": "horizontal",
+              "style": "IPY_MODEL_4d4b43abf49f4114a7fc83d9128f6d30",
+              "value": 25000
+            }
+          },
+          "1d65d195b47147d3806f7735255878d8": {
+            "model_module": "@jupyter-widgets/controls",
+            "model_name": "HTMLModel",
+            "model_module_version": "1.5.0",
+            "state": {
+              "_dom_classes": [],
+              "_model_module": "@jupyter-widgets/controls",
+              "_model_module_version": "1.5.0",
+              "_model_name": "HTMLModel",
+              "_view_count": null,
+              "_view_module": "@jupyter-widgets/controls",
+              "_view_module_version": "1.5.0",
+              "_view_name": "HTMLView",
+              "description": "",
+              "description_tooltip": null,
+              "layout": "IPY_MODEL_0d7bc861ca364e298ffb26510a4e4e09",
+              "placeholder": "​",
+              "style": "IPY_MODEL_2c776b732e6a4cb8bb85e01baff5fb33",
+              "value": " 25000/25000 [25:07&lt;00:00, 17.45it/s, epoch=2/2, avg_loss=1.9639, avg_metric=0.2580]"
+            }
+          },
+          "34f4dd1f0b9b4920bffb1027ea6e11fe": {
+            "model_module": "@jupyter-widgets/base",
+            "model_name": "LayoutModel",
+            "model_module_version": "1.2.0",
+            "state": {
+              "_model_module": "@jupyter-widgets/base",
+              "_model_module_version": "1.2.0",
+              "_model_name": "LayoutModel",
+              "_view_count": null,
+              "_view_module": "@jupyter-widgets/base",
+              "_view_module_version": "1.2.0",
+              "_view_name": "LayoutView",
+              "align_content": null,
+              "align_items": null,
+              "align_self": null,
+              "border": null,
+              "bottom": null,
+              "display": "inline-flex",
+              "flex": null,
+              "flex_flow": "row wrap",
+              "grid_area": null,
+              "grid_auto_columns": null,
+              "grid_auto_flow": null,
+              "grid_auto_rows": null,
+              "grid_column": null,
+              "grid_gap": null,
+              "grid_row": null,
+              "grid_template_areas": null,
+              "grid_template_columns": null,
+              "grid_template_rows": null,
+              "height": null,
+              "justify_content": null,
+              "justify_items": null,
+              "left": null,
+              "margin": null,
+              "max_height": null,
+              "max_width": null,
+              "min_height": null,
+              "min_width": null,
+              "object_fit": null,
+              "object_position": null,
+              "order": null,
+              "overflow": null,
+              "overflow_x": null,
+              "overflow_y": null,
+              "padding": null,
+              "right": null,
+              "top": null,
+              "visibility": null,
+              "width": "100%"
+            }
+          },
+          "38e5079824d648849dcde09e2a2948fd": {
+            "model_module": "@jupyter-widgets/base",
+            "model_name": "LayoutModel",
+            "model_module_version": "1.2.0",
+            "state": {
+              "_model_module": "@jupyter-widgets/base",
+              "_model_module_version": "1.2.0",
+              "_model_name": "LayoutModel",
+              "_view_count": null,
+              "_view_module": "@jupyter-widgets/base",
+              "_view_module_version": "1.2.0",
+              "_view_name": "LayoutView",
+              "align_content": null,
+              "align_items": null,
+              "align_self": null,
+              "border": null,
+              "bottom": null,
+              "display": null,
+              "flex": null,
+              "flex_flow": null,
+              "grid_area": null,
+              "grid_auto_columns": null,
+              "grid_auto_flow": null,
+              "grid_auto_rows": null,
+              "grid_column": null,
+              "grid_gap": null,
+              "grid_row": null,
+              "grid_template_areas": null,
+              "grid_template_columns": null,
+              "grid_template_rows": null,
+              "height": null,
+              "justify_content": null,
+              "justify_items": null,
+              "left": null,
+              "margin": null,
+              "max_height": null,
+              "max_width": null,
+              "min_height": null,
+              "min_width": null,
+              "object_fit": null,
+              "object_position": null,
+              "order": null,
+              "overflow": null,
+              "overflow_x": null,
+              "overflow_y": null,
+              "padding": null,
+              "right": null,
+              "top": null,
+              "visibility": null,
+              "width": null
+            }
+          },
+          "ffa3569f77fb4a3c8a6fb08930c0defb": {
+            "model_module": "@jupyter-widgets/controls",
+            "model_name": "DescriptionStyleModel",
+            "model_module_version": "1.5.0",
+            "state": {
+              "_model_module": "@jupyter-widgets/controls",
+              "_model_module_version": "1.5.0",
+              "_model_name": "DescriptionStyleModel",
+              "_view_count": null,
+              "_view_module": "@jupyter-widgets/base",
+              "_view_module_version": "1.2.0",
+              "_view_name": "StyleView",
+              "description_width": ""
+            }
+          },
+          "dd800993260d48e383fb9aa27d265c7d": {
+            "model_module": "@jupyter-widgets/base",
+            "model_name": "LayoutModel",
+            "model_module_version": "1.2.0",
+            "state": {
+              "_model_module": "@jupyter-widgets/base",
+              "_model_module_version": "1.2.0",
+              "_model_name": "LayoutModel",
+              "_view_count": null,
+              "_view_module": "@jupyter-widgets/base",
+              "_view_module_version": "1.2.0",
+              "_view_name": "LayoutView",
+              "align_content": null,
+              "align_items": null,
+              "align_self": null,
+              "border": null,
+              "bottom": null,
+              "display": null,
+              "flex": "2",
+              "flex_flow": null,
+              "grid_area": null,
+              "grid_auto_columns": null,
+              "grid_auto_flow": null,
+              "grid_auto_rows": null,
+              "grid_column": null,
+              "grid_gap": null,
+              "grid_row": null,
+              "grid_template_areas": null,
+              "grid_template_columns": null,
+              "grid_template_rows": null,
+              "height": null,
+              "justify_content": null,
+              "justify_items": null,
+              "left": null,
+              "margin": null,
+              "max_height": null,
+              "max_width": null,
+              "min_height": null,
+              "min_width": null,
+              "object_fit": null,
+              "object_position": null,
+              "order": null,
+              "overflow": null,
+              "overflow_x": null,
+              "overflow_y": null,
+              "padding": null,
+              "right": null,
+              "top": null,
+              "visibility": null,
+              "width": null
+            }
+          },
+          "4d4b43abf49f4114a7fc83d9128f6d30": {
+            "model_module": "@jupyter-widgets/controls",
+            "model_name": "ProgressStyleModel",
+            "model_module_version": "1.5.0",
+            "state": {
+              "_model_module": "@jupyter-widgets/controls",
+              "_model_module_version": "1.5.0",
+              "_model_name": "ProgressStyleModel",
+              "_view_count": null,
+              "_view_module": "@jupyter-widgets/base",
+              "_view_module_version": "1.2.0",
+              "_view_name": "StyleView",
+              "bar_color": null,
+              "description_width": ""
+            }
+          },
+          "0d7bc861ca364e298ffb26510a4e4e09": {
+            "model_module": "@jupyter-widgets/base",
+            "model_name": "LayoutModel",
+            "model_module_version": "1.2.0",
+            "state": {
+              "_model_module": "@jupyter-widgets/base",
+              "_model_module_version": "1.2.0",
+              "_model_name": "LayoutModel",
+              "_view_count": null,
+              "_view_module": "@jupyter-widgets/base",
+              "_view_module_version": "1.2.0",
+              "_view_name": "LayoutView",
+              "align_content": null,
+              "align_items": null,
+              "align_self": null,
+              "border": null,
+              "bottom": null,
+              "display": null,
+              "flex": null,
+              "flex_flow": null,
+              "grid_area": null,
+              "grid_auto_columns": null,
+              "grid_auto_flow": null,
+              "grid_auto_rows": null,
+              "grid_column": null,
+              "grid_gap": null,
+              "grid_row": null,
+              "grid_template_areas": null,
+              "grid_template_columns": null,
+              "grid_template_rows": null,
+              "height": null,
+              "justify_content": null,
+              "justify_items": null,
+              "left": null,
+              "margin": null,
+              "max_height": null,
+              "max_width": null,
+              "min_height": null,
+              "min_width": null,
+              "object_fit": null,
+              "object_position": null,
+              "order": null,
+              "overflow": null,
+              "overflow_x": null,
+              "overflow_y": null,
+              "padding": null,
+              "right": null,
+              "top": null,
+              "visibility": null,
+              "width": null
+            }
+          },
+          "2c776b732e6a4cb8bb85e01baff5fb33": {
+            "model_module": "@jupyter-widgets/controls",
+            "model_name": "DescriptionStyleModel",
+            "model_module_version": "1.5.0",
+            "state": {
+              "_model_module": "@jupyter-widgets/controls",
+              "_model_module_version": "1.5.0",
+              "_model_name": "DescriptionStyleModel",
+              "_view_count": null,
+              "_view_module": "@jupyter-widgets/base",
+              "_view_module_version": "1.2.0",
+              "_view_name": "StyleView",
+              "description_width": ""
+            }
+          }
+        }
+      }
+    }
+  },
+  "nbformat": 4,
+  "nbformat_minor": 5
+}
\ No newline at end of file
diff --git a/MAEs/PAG_Avishikta_Bhattacharjee/notebook/README.md b/MAEs/PAG_Avishikta_Bhattacharjee/notebook/README.md
new file mode 100644
index 0000000..c633e2c
--- /dev/null
+++ b/MAEs/PAG_Avishikta_Bhattacharjee/notebook/README.md
@@ -0,0 +1,42 @@
+# Physics-Aware Gating in Transformer for Jet Classification
+
+<p align="center">
+  <img width="788" alt="GSoC Banner" src="https://github.com/user-attachments/assets/c5c91a0b-a0f6-4a92-b992-45d3f5fefaba" />
+</p>
+
+This repository contains the implementation of a **Physics-Aware Gating on Lorentz Particle Transformer (LorentzParT)** designed for self-supervised pre-training and jet reconstruction on high-energy physics datasets like **JetClass**.
+
+The core contribution of this work is a custom attention block that injects a **normalized global invariant mass bias ($m^2$)** alongside a **gated pairwise interaction matrix ($U$)** directly into the attention mechanism, enforcing fundamental Lorentz invariance and relativistic conservation laws.
+
+---
+
+## Description
+
+The existing attention mechanism uses a bias matrix $U$ to incline the transformer toward physics constraints on jet particles, achieving a strong baseline (ROC AUC > 0.90). To evaluate and improve attention across various benchmarks, this repository introduces a **gating mechanism** controlled by physics attention heads to further reduce noise from jet constituent data.
+
+### 1. Reused Physics Bias Attention Head for Gating
+Rather than calculating static physical shifts independently at every Transformer layer, the model utilizes a dedicated **Physics Bias Attention Head**. This module extracts the pairwise $U$-matrix—computed from Minkowski inner products—and fuses it with the global normalized invariant mass bias ($m^2$). 
+
+This physical bias tensor is projected across key heads and **reused dynamically across encoder blocks** to act as a gating mask. By projecting and reusing these learned physical weights directly within the multi-head attention mechanism, the network modulates the query-key matrix ($QK^T$) before value aggregation, filtering out unphysical particle couplings and stabilizing training across dynamic batch shapes.
+
+### 2. Global Invariant Mass Scalar ($m^2$)
+For a jet of $N$ particles, the total invariant mass squared ($m^2$) is computed across aggregate energy and momentum sums:
+
+$$m^2 = \left(\sum_{i=1}^{N} E_i\right)^2 - \left\| \sum_{i=1}^{N} \vec{p}_i \right\|_2^2$$
+
+To ensure numerical stability across varying energy scales, $m^2$ is feature-scaled to form a **normalized invariant mass bias**.
+
+---
+
+## Architecture Overview
+
+The model uses the encoder of LorentzParT embedded within a Variational Autoencoder (VAE) setup:
+1. **Masking:** Random constituent particles are masked during training.
+2. **Encoder:** Processes unmasked tokens through LorentzParT blocks augmented with the **Physics-Aware Gating (PAG)** layer.
+3. **Decoder:** Reconstructs the 4 constituent properties ($p_T, \eta, \phi, E$) of the masked particles.
+
+---
+
+## Evaluation & Results
+
+Will add in the final submission (yet to upload).