{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "59a0471a-205d-479e-a9c1-94c4384d616d",
   "metadata": {},
   "source": [
    "# Implementing Transformer Architecture: A Step-by-Step Guide\n",
    "\n",
    "## Paper Reference\n",
    "- [\"Attention Is All You Need\"](https://arxiv.org/abs/1706.03762) (Vaswani et al., 2017)\n",
    "- Key sections: \n",
    "  - 3.1: Encoder and Decoder Stacks\n",
    "  - 3.2: Attention Mechanism\n",
    "  - 3.3: Position-wise Feed-Forward Networks\n",
    "  - 3.4: Embeddings and Softmax\n",
    "  - 3.5: Positional Encoding\n",
    "  - 5.4: Regularization (dropout strategy)\n",
    "\n",
    "## Implementation Strategy\n",
    "Breaking down the architecture into manageable pieces and gradually adding complexity:\n",
    "\n",
    "1. Start with foundational components:\n",
    "   - Embedding + Positional Encoding\n",
    "   - Single-head self-attention\n",
    "   \n",
    "2. Build up attention mechanism:\n",
    "   - Extend to multi-head attention\n",
    "   - Add cross-attention capability\n",
    "   - Implement attention masking\n",
    "\n",
    "3. Construct larger components:\n",
    "   - Encoder (self-attention + FFN)\n",
    "   - Decoder (masked self-attention + cross-attention + FFN)\n",
    "   \n",
    "4. Combine into final architecture:\n",
    "   - Encoder-Decoder stack\n",
    "   - Full Transformer with input/output layers\n",
    "\n",
    "## Development Tips\n",
    "1. Visualization and Planning:\n",
    "   - Draw out tensor dimensions on paper\n",
    "   - Sketch attention patterns and masks\n",
    "   - Map each component back to paper equations\n",
    "   - This helps catch dimension mismatches early!\n",
    "\n",
    "2. Dimension Cheat Sheet:\n",
    "   - Input tokens: [batch_size, seq_len]\n",
    "   - Embeddings: [batch_size, seq_len, d_model]\n",
    "   - Attention matrices: [batch_size, num_heads, seq_len, seq_len]\n",
    "   - FFN hidden layer: [batch_size, seq_len, d_ff]\n",
    "   - Output logits: [batch_size, seq_len, vocab_size]\n",
    "\n",
    "3. Common Pitfalls:\n",
    "   - Forgetting to scale dot products by √d_k\n",
    "   - Applying mask too early or too late\n",
    "   - Incorrect mask dimensions or application\n",
    "   - Missing residual connections\n",
    "   - Wrong order of layer norm and dropout\n",
    "   - Tensor dimension mismatches in attention\n",
    "   - Not handling padding properly\n",
    "\n",
    "4. Performance Considerations:\n",
    "   - Memory usage scales with sequence length squared\n",
    "   - Attention computation is O(n²) with sequence length\n",
    "   - Balance between d_model and num_heads\n",
    "   - Trade-off between model size and batch size\n",
    "\n",
    "## Testing Strategy\n",
    "- Test each component independently\n",
    "- Verify shape preservation\n",
    "- Check attention patterns\n",
    "- Confirm mask effectiveness\n",
    "- Validate gradient flow\n",
    "- Monitor numerical stability\n",
    "\n",
    "Remember: The key to successfully implementing the Transformer is understanding how each piece fits together and maintaining clear dimension tracking throughout the implementation."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "42004708-561e-4cd8-bad6-b229ed509bf7",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "f7c1308f-2592-4272-91c3-40184625bbfe",
   "metadata": {},
   "source": [
    "## Code Section"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9311b3a3-719c-4440-862d-2be962eff271",
   "metadata": {},
   "source": [
    "### Embedding and Positional Encoding\n",
    "This implements the input embedding from Section 3.4 and positional encoding from Section 3.5 of the paper. Key points:\n",
    "- Embedding dimension can differ from model dimension (using projection)\n",
    "- Positional encoding uses sine and cosine functions\n",
    "- Scale embeddings by √d_model\n",
    "- Apply dropout to the sum of embeddings and positional encodings\n",
    "\n",
    "Implementation tips:\n",
    "- Use `nn.Embedding` for token embeddings\n",
    "- Store scaling factor as float during initialization\n",
    "- Remember to expand positional encoding for batch dimension\n",
    "- Add assertion for input dtype (should be torch.long)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "4ce6f4c3-41b2-4610-a68e-fef96c8fe761",
   "metadata": {},
   "outputs": [],
   "source": [
    "import math\n",
    "import torch\n",
    "import torch.nn as nn\n",
    "\n",
    "class EmbeddingWithProjection(nn.Module):\n",
    "    def __init__(self, vocab_size, d_embed, d_model,  \n",
    "                 max_position_embeddings =512, dropout=0.1):\n",
    "        super().__init__()\n",
    "        self.d_model = d_model\n",
    "        self.d_embed = d_embed\n",
    "        self.vocab_size = vocab_size\n",
    "        self.embedding = nn.Embedding(self.vocab_size, self.d_embed)\n",
    "        self.projection = nn.Linear(self.d_embed, self.d_model)\n",
    "        self.scaling = float(math.sqrt(self.d_model))\n",
    "\n",
    "        self.layernorm = nn.LayerNorm(self.d_model)\n",
    "        self.dropout = nn.Dropout(p=dropout)\n",
    "\n",
    "    @staticmethod\n",
    "    def create_positional_encoding(seq_length, d_model, batch_size=1):\n",
    "        # Create position indices: [seq_length, 1]\n",
    "        position = torch.arange(seq_length).unsqueeze(1).float()\n",
    "        \n",
    "        # Create dimension indices: [1, d_model//2]\n",
    "        div_term = torch.exp(\n",
    "            torch.arange(0, d_model, 2).float() * \n",
    "            (-math.log(10000.0) / d_model)\n",
    "        )\n",
    "        \n",
    "        # Create empty tensor: [seq_length, d_model]\n",
    "        pe = torch.zeros(seq_length, d_model)\n",
    "        \n",
    "        # Compute sin and cos\n",
    "        pe[:, 0::2] = torch.sin(position * div_term)\n",
    "        pe[:, 1::2] = torch.cos(position * div_term)\n",
    "        \n",
    "        # Add batch dimension and expand: [batch_size, seq_length, d_model]\n",
    "        pe = pe.unsqueeze(0).expand(batch_size, -1, -1)\n",
    "        \n",
    "        return pe\n",
    "\n",
    "    \n",
    "    \n",
    "    def forward(self, x):\n",
    "        assert x.dtype == torch.long, f\"Input tensor must have dtype torch.long, got {x.dtype}\"\n",
    "        batch_size, seq_length = x.size() # [batch, seq_length]\n",
    "\n",
    "        # token embedding\n",
    "        token_embedding = self.embedding(x)                                                            #[2, 16, 1024]     \n",
    "        # project the scaled token embedding to the d_model space\n",
    "        token_embedding =  self.projection(token_embedding) * self.scaling                             #[2, 16, 768]\n",
    "\n",
    "        # add positional encodings to projected, \n",
    "        # scaled embeddings before applying layer norm and dropout.\n",
    "        positional_encoding = self.create_positional_encoding(seq_length, self.d_model, batch_size)    #[2, 16, 768]\n",
    "        \n",
    "        # In addition, we apply dropout to the sums of the embeddings \n",
    "        # in both the encoder and decoder stacks. For the base model, we use a rate of Pdrop = 0.1.\n",
    "        normalized_sum = self.layernorm(token_embedding + positional_encoding)\n",
    "        final_output = self.dropout(normalized_sum)\n",
    "        return final_output"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "30d449b6-c6ea-4634-b071-0839841e3044",
   "metadata": {},
   "source": [
    "### Transformer Attention\n",
    "Implements the core attention mechanism from Section 3.2.1. Formula: Attention(Q,K,V) = softmax(QK^T/√d_k)V\n",
    "\n",
    "Key points:\n",
    "- Supports both self-attention and cross-attention\n",
    "- Multi-head attention implementation per section 3.2.2\n",
    "- Handles different sequence lengths for encoder/decoder\n",
    "- Scales dot products by 1/√d_k\n",
    "- Applies attention masking before softmax\n",
    "- Applies dropout after softmax\n",
    "\n",
    "Implementation tips:\n",
    "- Use separate Q,K,V projections\n",
    "- Handle masking through addition (not masked_fill)\n",
    "- Remember to use braodcasting and reshape for multi-head attention\n",
    "- Keep track of tensor dimensions at each step"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "e63e4cf5-2126-4480-ac12-f24d0912e226",
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "\n",
    "import math\n",
    "\n",
    "class TransformerAttention(nn.Module):\n",
    "    \"\"\"\n",
    "    Transformer Scaled Dot Product Attention Module\n",
    "    Args:\n",
    "        d_model: Total dimension of the model.\n",
    "        num_head: Number of attention heads.\n",
    "        dropout: Dropout rate for attention scores.\n",
    "        bias: Whether to include bias in linear projections.\n",
    "\n",
    "    Inputs:\n",
    "        sequence: input sequence for self-attention and the query for cross-attention\n",
    "        key_value_state: input for the key, values for cross-attention\n",
    "    \"\"\"\n",
    "    def __init__(self, d_model, num_head, dropout=0.1, bias=True): # infer d_k, d_v, d_q from d_model\n",
    "        super().__init__()  # Missing in the original implementation\n",
    "        assert d_model % num_head == 0, \"d_model must be divisible by num_head\"\n",
    "        self.d_model = d_model\n",
    "        self.num_head = num_head\n",
    "        self.d_head=d_model//num_head\n",
    "        self.dropout_rate = dropout  # Store dropout rate separately\n",
    "\n",
    "        # linear transformations\n",
    "        self.q_proj = nn.Linear(d_model, d_model, bias=bias)\n",
    "        self.k_proj = nn.Linear(d_model, d_model, bias=bias)\n",
    "        self.v_proj = nn.Linear(d_model, d_model, bias=bias)\n",
    "        self.output_proj = nn.Linear(d_model, d_model, bias=bias)\n",
    "\n",
    "        # Dropout layer\n",
    "        self.dropout = nn.Dropout(p=dropout)\n",
    "\n",
    "        # Initiialize scaler\n",
    "        self.scaler = float(1.0 / math.sqrt(self.d_head)) # Store as float in initialization\n",
    "        \n",
    "\n",
    "    def forward(self, sequence, key_value_states = None, att_mask=None):\n",
    "        \"\"\"Input shape: [batch_size, seq_len, d_model=num_head * d_head]\"\"\"\n",
    "        batch_size, seq_len, model_dim = sequence.size()\n",
    "\n",
    "        # Check only critical input dimensions\n",
    "        assert model_dim == self.d_model, f\"Input dimension {model_dim} doesn't match model dimension {self.d_model}\"\n",
    "        if key_value_states is not None:\n",
    "            assert key_value_states.size(-1) == self.d_model, \\\n",
    "            f\"Cross attention key/value dimension {key_value_states.size(-1)} doesn't match model dimension {self.d_model}\"\n",
    "\n",
    "\n",
    "        # if key_value_states are provided this layer is used as a cross-attention layer\n",
    "        # for the decoder\n",
    "        is_cross_attention = key_value_states is not None\n",
    "        \n",
    "        # Linear projections and reshape for multi-head\n",
    "        Q_state = self.q_proj(sequence)\n",
    "        if is_cross_attention:\n",
    "            kv_seq_len = key_value_states.size(1)\n",
    "            K_state = self.k_proj(key_value_states)\n",
    "            V_state = self.v_proj(key_value_states)\n",
    "        else:\n",
    "            kv_seq_len = seq_len\n",
    "            K_state = self.k_proj(sequence)\n",
    "            V_state = self.v_proj(sequence)\n",
    "\n",
    "        #[batch_size, self.num_head, seq_len, self.d_head]\n",
    "        Q_state = Q_state.view(batch_size, seq_len, self.num_head, self.d_head).transpose(1,2) \n",
    "            \n",
    "        # in cross-attention, key/value sequence length might be different from query sequence length\n",
    "        K_state = K_state.view(batch_size, kv_seq_len, self.num_head, self.d_head).transpose(1,2)\n",
    "        V_state = V_state.view(batch_size, kv_seq_len, self.num_head, self.d_head).transpose(1,2)\n",
    "\n",
    "        # Scale Q by 1/sqrt(d_k)\n",
    "        Q_state = Q_state * self.scaler\n",
    "    \n",
    "    \n",
    "        # Compute attention matrix: QK^T\n",
    "        self.att_matrix = torch.matmul(Q_state, K_state.transpose(-1,-2)) \n",
    "\n",
    "    \n",
    "        # apply attention mask to attention matrix\n",
    "        if att_mask is not None and not isinstance(att_mask, torch.Tensor):\n",
    "            raise TypeError(\"att_mask must be a torch.Tensor\")\n",
    "\n",
    "        if att_mask is not None:\n",
    "            self.att_matrix = self.att_matrix + att_mask\n",
    "        \n",
    "        # apply softmax to the last dimension to get the attention score: softmax(QK^T)\n",
    "        att_score = F.softmax(self.att_matrix, dim = -1)\n",
    "    \n",
    "        # apply drop out to attention score\n",
    "        att_score = self.dropout(att_score)\n",
    "    \n",
    "        # get final output: softmax(QK^T)V\n",
    "        att_output = torch.matmul(att_score, V_state)\n",
    "    \n",
    "        # concatinate all attention heads\n",
    "        att_output = att_output.contiguous().view(batch_size, seq_len, self.num_head*self.d_head) \n",
    "    \n",
    "        # final linear transformation to the concatenated output\n",
    "        att_output = self.output_proj(att_output)\n",
    "\n",
    "        assert att_output.size() == (batch_size, seq_len, self.d_model), \\\n",
    "        f\"Final output shape {att_output.size()} incorrect\"\n",
    "\n",
    "        return att_output"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8f58d571-fab1-49e3-9d17-bb8e9ab8eae8",
   "metadata": {},
   "source": [
    "### Feed-Forward Network (FFN)\n",
    "Implements the position-wise feed-forward network from Section 3.3: FFN(x) = max(0, xW₁ + b₁)W₂ + b₂\n",
    "\n",
    "Key points:\n",
    "- Two linear transformations with ReLU in between\n",
    "- Inner layer dimension (d_ff) is typically 2048\n",
    "- Applied identically to each position\n",
    "\n",
    "Implementation tips:\n",
    "- Use nn.Linear for transformations\n",
    "- Remember to include bias terms\n",
    "- Position-wise means same transformation for each position\n",
    "- Dimension flow: d_model → d_ff → d_model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "bedf8d01-61e9-472b-836d-eb79d228da5e",
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "\n",
    "\n",
    "class FFN(nn.Module):\n",
    "    \"\"\"\n",
    "    Position-wise Feed-Forward Networks\n",
    "    This consists of two linear transformations with a ReLU activation in between.\n",
    "    \n",
    "    FFN(x) = max(0, xW1 + b1 )W2 + b2\n",
    "    d_model: embedding dimension (e.g., 512)\n",
    "    d_ff: feed-forward dimension (e.g., 2048)\n",
    "    \n",
    "    \"\"\"\n",
    "    def __init__(self, d_model, d_ff):\n",
    "        super().__init__()\n",
    "        self.d_model=d_model\n",
    "        self.d_ff= d_ff\n",
    "        \n",
    "        # Linear transformation y = xW+b\n",
    "        self.fc1 = nn.Linear(self.d_model, self.d_ff, bias = True)\n",
    "        self.fc2 = nn.Linear(self.d_ff, self.d_model, bias = True)\n",
    "        \n",
    "        # for potential speed up\n",
    "        # Pre-normalize the weights (can help with training stability)\n",
    "        nn.init.xavier_uniform_(self.fc1.weight)\n",
    "        nn.init.xavier_uniform_(self.fc2.weight)\n",
    "\n",
    "\n",
    "    def forward(self, input):\n",
    "        # check input and first FF layer dimension matching\n",
    "        batch_size, seq_length, d_input = input.size()\n",
    "        assert self.d_model == d_input, \"d_model must be the same dimension as the input\"\n",
    "\n",
    "        # First linear transformation followed by ReLU\n",
    "        # There's no need for explicit torch.max() as F.relu() already implements max(0,x)\n",
    "        f1 = F.relu(self.fc1(input))\n",
    "\n",
    "        # max(0, xW_1 + b_1)W_2 + b_2 \n",
    "        f2 =  self.fc2(f1)\n",
    "\n",
    "        return f2\n",
    "\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "5a555a9d-29a7-46a6-b163-06934db1aacf",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "FFN(\n",
      "  (fc1): Linear(in_features=512, out_features=2048, bias=True)\n",
      "  (fc2): Linear(in_features=2048, out_features=512, bias=True)\n",
      ")\n"
     ]
    }
   ],
   "source": [
    "net = FFN(  d_model = 512,  d_ff =2048)\n",
    "print(net)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c4c002f0-5562-4de2-97d3-aa65d596a3fb",
   "metadata": {},
   "source": [
    "### Transformer Encoder\n",
    "Implements single encoder layer from Section 3.1, consisting of:\n",
    "- Multi-head self-attention\n",
    "- Position-wise feed-forward network\n",
    "- Residual connections and layer normalization\n",
    "\n",
    "\n",
    "Implementation tips:\n",
    "- Apply dropout before adding residual\n",
    "- Keep model dimension consistent through the layer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "89c2eebd-d774-4734-b9ea-486183182d4c",
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "\n",
    "class TransformerEncoder(nn.Module):\n",
    "    \"\"\"\n",
    "    Encoder layer of the Transformer\n",
    "    Sublayers: TransformerAttention\n",
    "               Residual LayerNorm\n",
    "               FNN\n",
    "               Residual LayerNorm\n",
    "    Args:\n",
    "            d_model: 512 model hidden dimension\n",
    "            d_embed: 512 embedding dimension, same as d_model in transformer framework\n",
    "            d_ff: 2048 hidden dimension of the feed forward network\n",
    "            num_head: 8 Number of attention heads.\n",
    "            dropout:  0.1 dropout rate \n",
    "            \n",
    "            bias: Whether to include bias in linear projections.\n",
    "              \n",
    "    \"\"\"\n",
    "\n",
    "    def __init__(\n",
    "        self, d_model, d_ff,\n",
    "        num_head, dropout=0.1,\n",
    "        bias=True\n",
    "    ):\n",
    "        super().__init__()\n",
    "        self.d_model = d_model\n",
    "        self.d_ff = d_ff\n",
    "\n",
    "        \n",
    "        # attention sublayer\n",
    "        self.att = TransformerAttention(\n",
    "            d_model = d_model,\n",
    "            num_head = num_head,\n",
    "            dropout = dropout,\n",
    "            bias = bias\n",
    "        )\n",
    "        \n",
    "        # FFN sublayer\n",
    "        self.ffn = FFN(\n",
    "            d_model = d_model,\n",
    "            d_ff = d_ff\n",
    "        )\n",
    "\n",
    "        # Dropout layer\n",
    "        self.dropout = nn.Dropout(p=dropout)\n",
    "\n",
    "        # layer-normalization layer\n",
    "        self.LayerNorm_att = nn.LayerNorm(self.d_model)\n",
    "        self.LayerNorm_ffn = nn.LayerNorm(self.d_model)\n",
    "\n",
    "        \n",
    "    def forward(self, embed_input, padding_mask=None):\n",
    "       \n",
    "        batch_size, seq_len, _ = embed_input.size()\n",
    "        \n",
    "        ## First sublayer: self attion \n",
    "        att_sublayer = self.att(sequence = embed_input, key_value_states = None, \n",
    "                                att_mask = padding_mask)  # [batch_size, sequence_length, d_model]\n",
    "        \n",
    "        # apply dropout before layer normalization for each sublayer\n",
    "        att_sublayer = self.dropout(att_sublayer)\n",
    "        # Residual layer normalization\n",
    "        att_normalized = self.LayerNorm_att(embed_input + att_sublayer)           # [batch_size, sequence_length, d_model]\n",
    "        \n",
    "        ## Second sublayer: FFN\n",
    "        ffn_sublayer = self.ffn(att_normalized)                                   # [batch_size, sequence_length, d_model]\n",
    "        ffn_sublayer = self.dropout(ffn_sublayer)\n",
    "        ffn_normalized = self.LayerNorm_ffn(att_normalized + ffn_sublayer )       # [batch_size, sequence_length, d_model]\n",
    "    \n",
    "\n",
    "        return ffn_normalized"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "6bed1d90-9b3b-468d-a746-d1d0e1753c19",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "TransformerEncoder(\n",
      "  (att): TransformerAttention(\n",
      "    (q_proj): Linear(in_features=512, out_features=512, bias=True)\n",
      "    (k_proj): Linear(in_features=512, out_features=512, bias=True)\n",
      "    (v_proj): Linear(in_features=512, out_features=512, bias=True)\n",
      "    (output_proj): Linear(in_features=512, out_features=512, bias=True)\n",
      "    (dropout): Dropout(p=0.1, inplace=False)\n",
      "  )\n",
      "  (ffn): FFN(\n",
      "    (fc1): Linear(in_features=512, out_features=2048, bias=True)\n",
      "    (fc2): Linear(in_features=2048, out_features=512, bias=True)\n",
      "  )\n",
      "  (dropout): Dropout(p=0.1, inplace=False)\n",
      "  (LayerNorm_att): LayerNorm((512,), eps=1e-05, elementwise_affine=True)\n",
      "  (LayerNorm_ffn): LayerNorm((512,), eps=1e-05, elementwise_affine=True)\n",
      ")\n"
     ]
    }
   ],
   "source": [
    "net = TransformerEncoder( d_model = 512, d_ff =2048, num_head=8, dropout=0.1, bias=True )\n",
    "print(net)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f3233e1a-0107-4bfa-9507-d9c6c56cdd60",
   "metadata": {},
   "source": [
    "### Transformer Decoder\n",
    "Implements decoder layer from Section 3.1, with three sub-layers:\n",
    "- Masked multi-head self-attention\n",
    "- Multi-head cross-attention with encoder output\n",
    "- Position-wise feed-forward network\n",
    "\n",
    "Key points:\n",
    "- Self-attention uses causal masking\n",
    "- Cross-attention allows attending to all encoder outputs\n",
    "- Each sub-layer followed by residual connection and layer normalization\n",
    "- Apply dropout to the output of previous sub-layer before residual connection and layer normalization\n",
    "\n",
    "Implementation tips:\n",
    "- Order of operations matters (masking before softmax)\n",
    "- Each attention layer has its own projections\n",
    "- Remember to pass encoder outputs for cross-attention\n",
    "- Careful with mask dimensions in self and cross attention\n",
    "- Key implementation detail for causal masking:\n",
    "- Create causal mask using upper triangular matrix:\n",
    "  ```python\n",
    "  mask = torch.triu(torch.ones(seq_len, seq_len), diagonal=1)\n",
    "  mask = mask.masked_fill(mask == 1, float('-inf'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "f3aca0ac-3b48-4033-a889-dd019e2d67a0",
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "\n",
    "class TransformerDecoder(nn.Module):\n",
    "    \"\"\"\n",
    "    Decoder layer of the Transformer\n",
    "    Sublayers: TransformerAttention with self-attention\n",
    "               Residual LayerNorm\n",
    "               TransformerAttention with cross-attention\n",
    "               Residual LayerNorm\n",
    "               FNN\n",
    "               Residual LayerNorm\n",
    "    Args:\n",
    "            d_model: 512 model hidden dimension\n",
    "            d_embed: 512 embedding dimension, same as d_model in transformer framework\n",
    "            d_ff: 2048 hidden dimension of the feed forward network\n",
    "            num_head: 8 Number of attention heads.\n",
    "            dropout:  0.1 dropout rate \n",
    "            \n",
    "            bias: Whether to include bias in linear projections.\n",
    "              \n",
    "    \"\"\"\n",
    "\n",
    "    def __init__(\n",
    "        self, d_model, d_ff,\n",
    "        num_head, dropout=0.1,\n",
    "        bias=True\n",
    "    ):\n",
    "        super().__init__()\n",
    "        self.d_model = d_model\n",
    "        self.d_ff = d_ff\n",
    "\n",
    "        \n",
    "        # attention sublayer\n",
    "        self.att = TransformerAttention(\n",
    "            d_model = d_model,\n",
    "            num_head = num_head,\n",
    "            dropout = dropout,\n",
    "            bias = bias\n",
    "        )\n",
    "        \n",
    "        # FFN sublayer\n",
    "        self.ffn = FFN(\n",
    "            d_model = d_model,\n",
    "            d_ff = d_ff\n",
    "        )\n",
    "\n",
    "        \n",
    "        # Dropout layer\n",
    "        self.dropout = nn.Dropout(p=dropout)\n",
    "\n",
    "        # layer-normalization layer\n",
    "        self.LayerNorm_att1 = nn.LayerNorm(self.d_model)\n",
    "        self.LayerNorm_att2 = nn.LayerNorm(self.d_model)\n",
    "        self.LayerNorm_ffn = nn.LayerNorm(self.d_model)\n",
    "\n",
    "    @staticmethod\n",
    "    def create_causal_mask(seq_len):\n",
    "        mask = torch.triu(torch.ones(seq_len, seq_len), diagonal=1)\n",
    "        mask = mask.masked_fill(mask == 1, float('-inf'))\n",
    "        return mask\n",
    "\n",
    "    \n",
    "    def forward(self, embed_input, cross_input, padding_mask=None):\n",
    "        \"\"\"\n",
    "        Args:\n",
    "        embed_input: Decoder input sequence [batch_size, seq_len, d_model]\n",
    "        cross_input: Encoder output sequence [batch_size, encoder_seq_len, d_model]\n",
    "        casual_attention_mask: Causal mask for self-attention [batch_size, seq_len, seq_len]\n",
    "        padding_mask: Padding mask for cross-attention [batch_size, seq_len, encoder_seq_len]\n",
    "        Returns:\n",
    "        Tensor: Decoded output [batch_size, seq_len, d_model]\n",
    "        \"\"\"\n",
    "        batch_size, seq_len, _ = embed_input.size()\n",
    "        \n",
    "        assert embed_input.size(-1) == self.d_model, f\"Input dimension {embed_input.size(-1)} doesn't match model dimension {self.d_model}\"\n",
    "        assert cross_input.size(-1) == self.d_model, \"Encoder output dimension doesn't match model dimension\"\n",
    "\n",
    "\n",
    "        # Generate and expand causal mask for self-attention\n",
    "        causal_mask = self.create_causal_mask(seq_len).to(embed_input.device)  # [seq_len, seq_len]\n",
    "        causal_mask = causal_mask.unsqueeze(0).unsqueeze(1)  # [1, 1, seq_len, seq_len]\n",
    "\n",
    "\n",
    "        ## First sublayer: self attion \n",
    "        # After embedding and positional encoding, input sequence feed into current attention sublayer\n",
    "        # Or, the output of the previous encoder/decoder feed into current attention sublayer\n",
    "        att_sublayer1 = self.att(sequence = embed_input, key_value_states = None, \n",
    "                                att_mask = causal_mask)  # [batch_size, num_head, sequence_length, d_model]\n",
    "        # apply dropout before layer normalization for each sublayer\n",
    "        att_sublayer1 = self.dropout(att_sublayer1)\n",
    "        # Residual layer normalization\n",
    "        att_normalized1 = self.LayerNorm_att1(embed_input + att_sublayer1)           # [batch_size, sequence_length, d_model]\n",
    "\n",
    "        ## Second sublayer: cross attention\n",
    "        # Query from the output of previous attention output, or training data\n",
    "        # Key, Value from output of Encoder of the same layer\n",
    "        att_sublayer2 = self.att(sequence = att_normalized1, key_value_states = cross_input, \n",
    "                                att_mask = padding_mask)  # [batch_size, sequence_length, d_model]\n",
    "        # apply dropout before layer normalization for each sublayer\n",
    "        att_sublayer2 = self.dropout(att_sublayer2)\n",
    "        # Residual layer normalization\n",
    "        att_normalized2 = self.LayerNorm_att2(att_normalized1 + att_sublayer2)           # [batch_size, sequence_length, d_model]\n",
    "        \n",
    "        \n",
    "        # Third sublayer: FFN\n",
    "        ffn_sublayer = self.ffn(att_normalized2)                                   # [batch_size, sequence_length, d_model]\n",
    "        ffn_sublayer = self.dropout(ffn_sublayer)\n",
    "        ffn_normalized = self.LayerNorm_ffn(att_normalized2 + ffn_sublayer )       # [batch_size, sequence_length, d_model]\n",
    "    \n",
    "\n",
    "        return ffn_normalized"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "a9678630-4e02-442f-9503-15066a574228",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "TransformerDecoder(\n",
      "  (att): TransformerAttention(\n",
      "    (q_proj): Linear(in_features=512, out_features=512, bias=True)\n",
      "    (k_proj): Linear(in_features=512, out_features=512, bias=True)\n",
      "    (v_proj): Linear(in_features=512, out_features=512, bias=True)\n",
      "    (output_proj): Linear(in_features=512, out_features=512, bias=True)\n",
      "    (dropout): Dropout(p=0.1, inplace=False)\n",
      "  )\n",
      "  (ffn): FFN(\n",
      "    (fc1): Linear(in_features=512, out_features=2048, bias=True)\n",
      "    (fc2): Linear(in_features=2048, out_features=512, bias=True)\n",
      "  )\n",
      "  (dropout): Dropout(p=0.1, inplace=False)\n",
      "  (LayerNorm_att1): LayerNorm((512,), eps=1e-05, elementwise_affine=True)\n",
      "  (LayerNorm_att2): LayerNorm((512,), eps=1e-05, elementwise_affine=True)\n",
      "  (LayerNorm_ffn): LayerNorm((512,), eps=1e-05, elementwise_affine=True)\n",
      ")\n"
     ]
    }
   ],
   "source": [
    "net = TransformerDecoder( d_model = 512, d_ff =2048, num_head=8, dropout=0.1, bias=True )\n",
    "print(net)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "71fe2c98-6278-4201-81f7-0b510efc7a68",
   "metadata": {},
   "source": [
    "### Encoder-Decoder Stack\n",
    "Implements the full stack of encoder and decoder layers from Section 3.1.\n",
    "\n",
    "Key points:\n",
    "- Multiple encoder and decoder layers (typically 6)\n",
    "- Each encoder output feeds into all decoder layers\n",
    "- Maintains residual connections throughout the stack\n",
    "\n",
    "Implementation tips:\n",
    "- Use nn.ModuleList for layer stacks\n",
    "- Share encoder outputs across decoder layers\n",
    "- Maintain consistent masking throughout\n",
    "- Handle padding masks separately from causal masks\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "aefccb90-9a89-4579-aba0-d837535f2d98",
   "metadata": {},
   "outputs": [],
   "source": [
    "class TransformerEncoderDecoder(nn.Module):\n",
    "    \"\"\"\n",
    "    Encoder-Decoder stack of the Transformer\n",
    "    Sublayers:  Encoder x 6\n",
    "                Decoder x 6\n",
    "    Args:\n",
    "            d_model: 512 model hidden dimension\n",
    "            d_embed: 512 embedding dimension, same as d_model in transformer framework\n",
    "            d_ff: 2048 hidden dimension of the feed forward network\n",
    "            num_head: 8 Number of attention heads.\n",
    "            dropout:  0.1 dropout rate \n",
    "            \n",
    "            bias: Whether to include bias in linear projections.\n",
    "              \n",
    "    \"\"\"\n",
    "    def __init__(\n",
    "        self, num_layer,\n",
    "        d_model, d_ff,\n",
    "        num_head, dropout=0.1,\n",
    "        bias=True\n",
    "    ):\n",
    "        super().__init__()\n",
    "        self.num_layer = num_layer\n",
    "        self.d_model = d_model\n",
    "        self.d_ff = d_ff\n",
    "        self.num_head = num_head\n",
    "        self.dropout = dropout\n",
    "        self.bias = bias\n",
    "\n",
    "        \n",
    "        # Encoder stack\n",
    "        self.encoder_stack = nn.ModuleList([ TransformerEncoder(\n",
    "                                        d_model = self.d_model, \n",
    "                                        d_ff = self.d_ff,\n",
    "                                        num_head = self.num_head, \n",
    "                                        dropout = self.dropout,\n",
    "                                        bias = self.bias) for _ in range(self.num_layer)])\n",
    "\n",
    "        # Decoder stack\n",
    "        self.decoder_stack = nn.ModuleList([ TransformerDecoder(\n",
    "                                        d_model = self.d_model, \n",
    "                                        d_ff = self.d_ff,\n",
    "                                        num_head = self.num_head, \n",
    "                                        dropout = self.dropout,\n",
    "                                        bias = self.bias) for _ in range(self.num_layer)])\n",
    "\n",
    "    \n",
    "    def forward(self, embed_encoder_input, embed_decoder_input, padding_mask=None):\n",
    "        # First layer of the encoder, decoder deck takes input from outside the deck\n",
    "        encoder_output = embed_encoder_input\n",
    "        decoder_output = embed_decoder_input\n",
    "\n",
    "        for (encoder, decoder) in zip(self.encoder_stack, self.decoder_stack):\n",
    "            encoder_output = encoder(embed_input = encoder_output, padding_mask = padding_mask)\n",
    "            decoder_output = decoder(embed_input = decoder_output, cross_input =encoder_output, padding_mask=padding_mask)\n",
    "       \n",
    "        \n",
    "        return decoder_output"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "eacab71c-054e-41c1-b85a-23c886b0af61",
   "metadata": {},
   "source": [
    "### Full Transformer\n",
    "Combines all components into complete architecture:\n",
    "- Input embeddings for source and target\n",
    "- Positional encoding\n",
    "- Encoder-decoder stack\n",
    "- Final linear and softmax layer\n",
    "\n",
    "Key points:\n",
    "- Handles different vocabulary sizes for source/target\n",
    "- Shifts decoder inputs for teacher forcing\n",
    "- Projects outputs to target vocabulary size\n",
    "- Applies log softmax for training stability\n",
    "\n",
    "Implementation tips:\n",
    "- Handle start tokens for decoder input\n",
    "- Maintain separate embeddings for source/target\n",
    "- Remember to scale embeddings\n",
    "- Consider sharing embedding weights with output layer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "e86e2ade-4584-4905-89ec-beaac7ebf401",
   "metadata": {},
   "outputs": [],
   "source": [
    "class Transformer(nn.Module):\n",
    "    def __init__(\n",
    "        self, \n",
    "        num_layer,\n",
    "        d_model, d_embed, d_ff,\n",
    "        num_head,\n",
    "        src_vocab_size, \n",
    "        tgt_vocab_size,\n",
    "        max_position_embeddings=512,\n",
    "        dropout=0.1,\n",
    "        bias=True\n",
    "    ):\n",
    "        super().__init__()\n",
    "        \n",
    "        self.tgt_vocab_size = tgt_vocab_size\n",
    "        \n",
    "        # Source and target embeddings\n",
    "        self.src_embedding = EmbeddingWithProjection(\n",
    "            vocab_size=src_vocab_size,\n",
    "            d_embed=d_embed,\n",
    "            d_model=d_model,\n",
    "            max_position_embeddings=max_position_embeddings,\n",
    "            dropout=dropout\n",
    "        )\n",
    "        \n",
    "        self.tgt_embedding = EmbeddingWithProjection(\n",
    "            vocab_size=tgt_vocab_size,\n",
    "            d_embed=d_embed,\n",
    "            d_model=d_model,\n",
    "            max_position_embeddings=max_position_embeddings,\n",
    "            dropout=dropout\n",
    "        )\n",
    "        \n",
    "        # Encoder-Decoder stack\n",
    "        self.encoder_decoder = TransformerEncoderDecoder(\n",
    "            num_layer=num_layer,\n",
    "            d_model=d_model,\n",
    "            d_ff=d_ff,\n",
    "            num_head=num_head,\n",
    "            dropout=dropout,\n",
    "            bias=bias\n",
    "        )\n",
    "        \n",
    "        # Output projection and softmax\n",
    "        self.output_projection = nn.Linear(d_model, tgt_vocab_size)\n",
    "        self.softmax = nn.LogSoftmax(dim=-1)\n",
    "        \n",
    "    def shift_target_right(self, tgt_tokens):\n",
    "        # Shift target tokens right by padding with zeros at the beginning\n",
    "        batch_size, seq_len = tgt_tokens.size()\n",
    "        \n",
    "        # Create start token (zeros)\n",
    "        start_tokens = torch.zeros(batch_size, 1, dtype=tgt_tokens.dtype, device=tgt_tokens.device)\n",
    "        \n",
    "        # Concatenate start token and remove last token\n",
    "        shifted_tokens = torch.cat([start_tokens, tgt_tokens[:, :-1]], dim=1)\n",
    "        \n",
    "        return shifted_tokens\n",
    "        \n",
    "    def forward(self, src_tokens, tgt_tokens, padding_mask=None):\n",
    "        \"\"\"\n",
    "        Args:\n",
    "            src_tokens: source sequence [batch_size, src_len]\n",
    "            tgt_tokens: target sequence [batch_size, tgt_len]\n",
    "            padding_mask: padding mask [batch_size, 1, 1, seq_len]\n",
    "        Returns:\n",
    "            output: [batch_size, tgt_len, tgt_vocab_size] log probabilities\n",
    "        \"\"\"\n",
    "        # Shift target tokens right for teacher forcing\n",
    "        shifted_tgt_tokens = self.shift_target_right(tgt_tokens)\n",
    "        \n",
    "        # Embed source and target sequences\n",
    "        src_embedding = self.src_embedding(src_tokens)\n",
    "        tgt_embedding = self.tgt_embedding(shifted_tgt_tokens)\n",
    "        \n",
    "        # Pass through encoder-decoder stack\n",
    "        decoder_output = self.encoder_decoder(\n",
    "            embed_encoder_input=src_embedding,\n",
    "            embed_decoder_input=tgt_embedding,\n",
    "            padding_mask=padding_mask\n",
    "        )\n",
    "        \n",
    "        # Project to vocabulary size and apply log softmax\n",
    "        logits = self.output_projection(decoder_output)\n",
    "        log_probs = self.softmax(logits)\n",
    "        \n",
    "        return log_probs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e4f8640d-2698-4f10-9acf-6681db3e7adf",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "0a8e4968-a06e-4d8c-bd9e-43508fa232fa",
   "metadata": {},
   "source": [
    "## Testing Section"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "f099df40-9609-407d-b3a8-0913bd77e011",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Tokenizer vocabulary size: 30522\n",
      "Input shape: torch.Size([2, 16])\n",
      "Embedded shape after projection: torch.Size([2, 16, 768])\n"
     ]
    }
   ],
   "source": [
    "## testing on the embedding implemntation\n",
    "## Tokenlize model input: from batched sentences to batched sequence of code\n",
    "from transformers import AutoTokenizer\n",
    "from transformers import pipeline\n",
    "\n",
    "import torch\n",
    "\n",
    "# layer config \n",
    "d_model = 768\n",
    "d_embed = 1024  # Larger embedding dimension\n",
    "vocab_size=30522\n",
    "\n",
    "# loading sample data\n",
    "checkpoint = \"distilbert-base-uncased-finetuned-sst-2-english\"\n",
    "tokenizer = AutoTokenizer.from_pretrained(checkpoint, use_fast=True, use_multiprocessing=False)\n",
    "sequences = [\"I've been waiting for a HuggingFace course my whole life.\", \"So have I!\"]\n",
    "\n",
    "# Will truncate the sequences that are longer than the model max length\n",
    "# (512 for BERT or DistilBERT)\n",
    "max_position_embeddings = 512\n",
    "model_inputs = tokenizer(sequences, truncation=True,  padding=\"longest\")\n",
    "\n",
    "# Check vocabulary size from the tokenizer\n",
    "# Happen to be the same as the default setting for distilbert -- of course!\n",
    "vocab_size = tokenizer.vocab_size\n",
    "print(f\"Tokenizer vocabulary size: {vocab_size}\")\n",
    "\n",
    "\n",
    "input = torch.tensor(model_inputs['input_ids'])\n",
    "embedder = EmbeddingWithProjection(vocab_size=vocab_size, d_embed=d_embed, d_model=d_model)\n",
    "output = embedder(input)\n",
    "\n",
    "print(f\"Input shape: {input.shape}\")\n",
    "print(f\"Embedded shape after projection: {output.shape}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "cae577d2-6ad1-4e9f-ad9b-26c646ec7c2e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Output Statistics:\n",
      "Mean: 0.0000\n",
      "Std: 1.0000\n",
      "Min: -4.1044\n",
      "Max: 3.5525\n",
      "\n",
      "Attention Analysis:\n",
      "Unmasked positions mean: 0.8049\n",
      "Masked positions mean: 0.8041\n",
      "\n",
      "Is the masking working? Yes\n",
      "\n",
      "All tests passed successfully!\n"
     ]
    }
   ],
   "source": [
    "def test_transformer_encoder():\n",
    "    # Set random seed for reproducibility\n",
    "    torch.manual_seed(42)\n",
    "    \n",
    "    # Test parameters\n",
    "    batch_size = 32\n",
    "    seq_length = 20\n",
    "    d_model = 512\n",
    "    d_ff = 2048\n",
    "    num_heads = 8\n",
    "    \n",
    "    # Initialize the transformer encoder\n",
    "    encoder = TransformerEncoder(\n",
    "        d_model=d_model,\n",
    "        d_ff=d_ff,\n",
    "        num_head=num_heads,\n",
    "        dropout=0.1\n",
    "    )\n",
    "    \n",
    "    # Set to evaluation mode to disable dropout\n",
    "    encoder.eval()\n",
    "    \n",
    "    # Create input sequence - using ones instead of random values\n",
    "    # for easier interpretation of attention patterns\n",
    "    input_sequence = torch.ones(batch_size, seq_length, d_model)\n",
    "    cross_sequence = torch.ones(batch_size, seq_length, d_model)*0.5\n",
    "    \n",
    "    # Create attention mask\n",
    "    attention_mask = torch.ones(batch_size, seq_length)\n",
    "    attention_mask[:, 15:] = 0  # Mask last 5 positions\n",
    "    attention_mask =attention_mask.unsqueeze(1).unsqueeze(3)\n",
    "    \n",
    "    # Store attention patterns\n",
    "    attention_patterns = []\n",
    "    \n",
    "    # Define hook to capture attention scores\n",
    "    def attention_hook(module, input, output):\n",
    "        # We want to capture the attention scores before they're processed further\n",
    "        # This assumes your attention module returns the attention scores\n",
    "        attention_patterns.append(output)\n",
    "    \n",
    "    # Register the hook on the attention computation\n",
    "    encoder.att.register_forward_hook(attention_hook)\n",
    "    \n",
    "    # Perform forward pass\n",
    "    with torch.no_grad():\n",
    "        output = encoder(input_sequence, attention_mask)\n",
    "    \n",
    "    # Basic shape tests\n",
    "    expected_shape = (batch_size, seq_length, d_model)\n",
    "    assert output.shape == expected_shape, f\"Expected shape {expected_shape}, got {output.shape}\"\n",
    "    \n",
    "    # Print output statistics\n",
    "    print(\"\\nOutput Statistics:\")\n",
    "    print(f\"Mean: {output.mean():.4f}\")\n",
    "    print(f\"Std: {output.std():.4f}\")\n",
    "    print(f\"Min: {output.min():.4f}\")\n",
    "    print(f\"Max: {output.max():.4f}\")\n",
    "    \n",
    "    # Analyze attention patterns\n",
    "    if attention_patterns:\n",
    "        attention_output = attention_patterns[0]\n",
    "        # Look at the attention patterns for unmasked vs masked positions\n",
    "        unmasked_attention = output[:, :15, :].abs().mean()\n",
    "        masked_attention = output[:, 15:, :].abs().mean()\n",
    "        \n",
    "        print(\"\\nAttention Analysis:\")\n",
    "        print(f\"Unmasked positions mean: {unmasked_attention:.4f}\")\n",
    "        print(f\"Masked positions mean: {masked_attention:.4f}\")\n",
    "        \n",
    "        # Note: We expect masked positions to still have values due to residual connections,\n",
    "        # but their patterns should be different from unmasked positions\n",
    "        print(\"\\nIs the masking working?\", \"Yes\" if unmasked_attention != masked_attention else \"No\")\n",
    "    \n",
    "    # Check for any NaN or infinite values\n",
    "    assert torch.isfinite(output).all(), \"Output contains NaN or infinite values\"\n",
    "    \n",
    "    print(\"\\nAll tests passed successfully!\")\n",
    "    return output, attention_patterns\n",
    "\n",
    "# Run the test\n",
    "output, attention_patterns = test_transformer_encoder()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "94221639-1a01-48f7-b805-9113d4347ba7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Output Statistics:\n",
      "Mean: 0.0000\n",
      "Std: 1.0000\n",
      "Min: -4.3335\n",
      "Max: 4.3904\n",
      "\n",
      "Shape Analysis:\n",
      "Input shape: torch.Size([32, 20, 512])\n",
      "Output shape: torch.Size([32, 20, 512])\n",
      "Expected shape matches: Yes\n",
      "\n",
      "All tests passed successfully!\n"
     ]
    }
   ],
   "source": [
    "def test_transformer_decoder():\n",
    "    torch.manual_seed(42)\n",
    "    \n",
    "    # Test parameters\n",
    "    batch_size = 32\n",
    "    seq_length = 20\n",
    "    encoder_seq_length = 22\n",
    "    d_model = 512\n",
    "    d_ff = 2048\n",
    "    num_heads = 8\n",
    "    \n",
    "    decoder = TransformerDecoder(\n",
    "        d_model=d_model,\n",
    "        d_ff=d_ff,\n",
    "        num_head=num_heads,\n",
    "        dropout=0.1\n",
    "    )\n",
    "    decoder.eval()\n",
    "    \n",
    "    # Create input sequences\n",
    "    decoder_input = torch.randn(batch_size, seq_length, d_model)\n",
    "    encoder_output = torch.randn(batch_size, encoder_seq_length, d_model)\n",
    "    \n",
    "    # Create padding mask for encoder outputs\n",
    "    padding_mask = torch.ones(batch_size, seq_length, encoder_seq_length)\n",
    "    padding_mask[:, :, 18:] = 0  # Mask last 4 positions of encoder output\n",
    "    padding_mask = padding_mask.unsqueeze(1)  # Add head dimension\n",
    "    \n",
    "    # Store attention scores\n",
    "    attention_scores = []\n",
    "    \n",
    "    # Define hook to capture attention scores before softmax\n",
    "    def attention_hook(module, input, output):\n",
    "        if not attention_scores:  # Only store first layer's patterns\n",
    "            # Assuming attention scores are computed before this hook\n",
    "            attention_scores.append(module.att_matrix.detach())  # You might need to modify this based on your attention implementation\n",
    "    \n",
    "    # Register hook on the attention layer\n",
    "    decoder.att.register_forward_hook(attention_hook)\n",
    "    \n",
    "    # Perform forward pass\n",
    "    with torch.no_grad():\n",
    "        output = decoder(decoder_input, encoder_output, padding_mask)\n",
    "    \n",
    "    # Basic shape tests\n",
    "    expected_shape = (batch_size, seq_length, d_model)\n",
    "    assert output.shape == expected_shape, f\"Expected shape {expected_shape}, got {output.shape}\"\n",
    "    \n",
    "    # Print output statistics\n",
    "    print(\"\\nOutput Statistics:\")\n",
    "    print(f\"Mean: {output.mean():.4f}\")\n",
    "    print(f\"Std: {output.std():.4f}\")\n",
    "    print(f\"Min: {output.min():.4f}\")\n",
    "    print(f\"Max: {output.max():.4f}\")\n",
    "    \n",
    "    # Test shape preservation\n",
    "    print(\"\\nShape Analysis:\")\n",
    "    print(f\"Input shape: {decoder_input.shape}\")\n",
    "    print(f\"Output shape: {output.shape}\")\n",
    "    print(f\"Expected shape matches: {'Yes' if decoder_input.shape == output.shape else 'No'}\")\n",
    "    \n",
    "    # Check for any NaN or infinite values\n",
    "    assert torch.isfinite(output).all(), \"Output contains NaN or infinite values\"\n",
    "    \n",
    "    print(\"\\nAll tests passed successfully!\")\n",
    "    return output, attention_scores\n",
    "\n",
    "# Run the test\n",
    "output, attention_scores = test_transformer_decoder()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "aab402c1-48a7-48b1-9202-3d2a43e186c9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Encoder Layer 0 shape: torch.Size([8, 10, 512])\n",
      "Decoder Layer 0 shape: torch.Size([8, 10, 512])\n",
      "\n",
      "Encoder Layer 1 shape: torch.Size([8, 10, 512])\n",
      "Decoder Layer 1 shape: torch.Size([8, 10, 512])\n",
      "\n",
      "Encoder Layer 2 shape: torch.Size([8, 10, 512])\n",
      "Decoder Layer 2 shape: torch.Size([8, 10, 512])\n",
      "\n",
      "Encoder Layer 3 shape: torch.Size([8, 10, 512])\n",
      "Decoder Layer 3 shape: torch.Size([8, 10, 512])\n",
      "\n",
      "Encoder Layer 4 shape: torch.Size([8, 10, 512])\n",
      "Decoder Layer 4 shape: torch.Size([8, 10, 512])\n",
      "\n",
      "Encoder Layer 5 shape: torch.Size([8, 10, 512])\n",
      "Decoder Layer 5 shape: torch.Size([8, 10, 512])\n",
      "\n",
      "Final Output Statistics:\n",
      "Mean: -0.0000\n",
      "Std: 1.0000\n",
      "Min: -4.3020\n",
      "Max: 3.8850\n",
      "\n",
      "Shape Preservation Check:\n",
      "Input shapes - Encoder: torch.Size([8, 10, 512]), Decoder: torch.Size([8, 10, 512])\n",
      "Output shape: torch.Size([8, 10, 512])\n",
      "\n",
      "Mean absolute difference between input and output: 0.9114\n",
      "Transformation occurred: Yes\n",
      "\n",
      "Total number of parameters: 37,834,752\n",
      "\n",
      "All tests passed successfully!\n"
     ]
    }
   ],
   "source": [
    "def test_transformer_encoder_decoder_stack():\n",
    "    torch.manual_seed(42)\n",
    "    \n",
    "    # Test parameters\n",
    "    batch_size = 8\n",
    "    seq_length = 10\n",
    "    d_model = 512\n",
    "    d_ff = 2048\n",
    "    num_heads = 8\n",
    "    num_layers = 6\n",
    "    \n",
    "    # Initialize the transformer encoder-decoder stack\n",
    "    transformer = TransformerEncoderDecoder(\n",
    "        num_layer=num_layers,\n",
    "        d_model=d_model,\n",
    "        d_ff=d_ff,\n",
    "        num_head=num_heads,\n",
    "        dropout=0.1\n",
    "    )\n",
    "    \n",
    "    # Set to evaluation mode to disable dropout\n",
    "    transformer.eval()\n",
    "    \n",
    "    # Create input sequences\n",
    "    encoder_input = torch.randn(batch_size, seq_length, d_model)\n",
    "    decoder_input = torch.randn(batch_size, seq_length, d_model)\n",
    "    \n",
    "    # Create padding mask\n",
    "    padding_mask = torch.ones(batch_size, seq_length)\n",
    "    padding_mask[:, -2:] = 0  # Mask last 2 positions\n",
    "    padding_mask = padding_mask.unsqueeze(1).unsqueeze(2)  # [batch, 1, 1, seq_len]\n",
    "    \n",
    "    # Store intermediate outputs\n",
    "    intermediate_outputs = []\n",
    "    \n",
    "    def hook_fn(module, input, output):\n",
    "        intermediate_outputs.append(output.detach())\n",
    "    \n",
    "    # Register hooks to capture outputs from each encoder and decoder layer\n",
    "    for i, (encoder, decoder) in enumerate(zip(transformer.encoder_stack, transformer.decoder_stack)):\n",
    "        encoder.register_forward_hook(lambda m, i, o, layer=i: print(f\"\\nEncoder Layer {layer} shape:\", o.shape))\n",
    "        decoder.register_forward_hook(lambda m, i, o, layer=i: print(f\"Decoder Layer {layer} shape:\", o.shape))\n",
    "    \n",
    "    # Perform forward pass\n",
    "    with torch.no_grad():\n",
    "        output = transformer(encoder_input, decoder_input, padding_mask)\n",
    "    \n",
    "    # Basic shape tests\n",
    "    expected_shape = (batch_size, seq_length, d_model)\n",
    "    assert output.shape == expected_shape, f\"Expected shape {expected_shape}, got {output.shape}\"\n",
    "    \n",
    "    # Print output statistics\n",
    "    print(\"\\nFinal Output Statistics:\")\n",
    "    print(f\"Mean: {output.mean():.4f}\")\n",
    "    print(f\"Std: {output.std():.4f}\")\n",
    "    print(f\"Min: {output.min():.4f}\")\n",
    "    print(f\"Max: {output.max():.4f}\")\n",
    "    \n",
    "    # Verify shape preservation through layers\n",
    "    print(\"\\nShape Preservation Check:\")\n",
    "    print(f\"Input shapes - Encoder: {encoder_input.shape}, Decoder: {decoder_input.shape}\")\n",
    "    print(f\"Output shape: {output.shape}\")\n",
    "    \n",
    "    # Check for any NaN or infinite values\n",
    "    assert torch.isfinite(output).all(), \"Output contains NaN or infinite values\"\n",
    "    \n",
    "    # Verify that output is different from input (transformation happened)\n",
    "    input_output_diff = (output - decoder_input).abs().mean()\n",
    "    print(f\"\\nMean absolute difference between input and output: {input_output_diff:.4f}\")\n",
    "    print(\"Transformation occurred:\", \"Yes\" if input_output_diff > 1e-3 else \"No\")\n",
    "    \n",
    "    # Check if model parameters were used\n",
    "    total_params = sum(p.numel() for p in transformer.parameters())\n",
    "    print(f\"\\nTotal number of parameters: {total_params:,}\")\n",
    "    \n",
    "    print(\"\\nAll tests passed successfully!\")\n",
    "    return output\n",
    "\n",
    "# Run the test\n",
    "output = test_transformer_encoder_decoder_stack()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "dfe316a8-d66f-417f-93e9-4450e4378e56",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Input Shapes:\n",
      "Source tokens: torch.Size([2, 16])\n",
      "Target tokens: torch.Size([2, 17])\n",
      "\n",
      "Output Analysis:\n",
      "Output shape: torch.Size([2, 17, 30522])\n",
      "\n",
      "Probability Distribution Check:\n",
      "Sum to 1: True\n",
      "Max probability: 0.0004\n",
      "Min probability: 0.0000\n",
      "\n",
      "Sample Predictions:\n",
      "Original target:\n",
      "J'ai attendu un cours HuggingFace toute ma vie.\n",
      "\n",
      "Model output (decoded):\n",
      "grassyane functioned bombay 仮 ‖ defiant necessitated necessitatedtipgne firmstipพ groupinglic chiefly\n",
      "\n",
      "Training Check:\n",
      "Loss value: 10.7644\n",
      "Has gradients: True\n"
     ]
    }
   ],
   "source": [
    "def test_complete_transformer():\n",
    "    # Configuration\n",
    "    d_model = 768\n",
    "    d_embed = 1024\n",
    "    d_ff = 2048\n",
    "    num_heads = 8\n",
    "    num_layers = 6\n",
    "    max_position_embeddings = 512\n",
    "    \n",
    "    # Load tokenizer\n",
    "    tokenizer = AutoTokenizer.from_pretrained(\"distilbert-base-uncased-finetuned-sst-2-english\", \n",
    "                                            use_fast=True, \n",
    "                                            use_multiprocessing=False)\n",
    "    vocab_size = tokenizer.vocab_size\n",
    "    \n",
    "    # Create sample source and target sequences\n",
    "    src_sequences = [\n",
    "        \"I've been waiting for a HuggingFace course my whole life.\",\n",
    "        \"So have I!\"\n",
    "    ]\n",
    "    # Pretend these are translations\n",
    "    tgt_sequences = [\n",
    "        \"J'ai attendu un cours HuggingFace toute ma vie.\",\n",
    "        \"Moi aussi!\"\n",
    "    ]\n",
    "    \n",
    "    # Tokenize source and target sequences\n",
    "    src_inputs = tokenizer(src_sequences, truncation=True, padding=\"longest\", return_tensors=\"pt\")\n",
    "    tgt_inputs = tokenizer(tgt_sequences, truncation=True, padding=\"longest\", return_tensors=\"pt\")\n",
    "    \n",
    "    # Create transformer model\n",
    "    transformer = Transformer(\n",
    "        num_layer=num_layers,\n",
    "        d_model=d_model,\n",
    "        d_embed=d_embed,\n",
    "        d_ff=d_ff,\n",
    "        num_head=num_heads,\n",
    "        src_vocab_size=vocab_size,\n",
    "        tgt_vocab_size=vocab_size,\n",
    "        max_position_embeddings=max_position_embeddings\n",
    "    )\n",
    "    \n",
    "    # Set to eval mode\n",
    "    transformer.eval()\n",
    "    \n",
    "    # Create padding mask from attention mask\n",
    "    padding_mask = src_inputs['attention_mask'].unsqueeze(1).unsqueeze(2)\n",
    "    \n",
    "    print(\"\\nInput Shapes:\")\n",
    "    print(f\"Source tokens: {src_inputs['input_ids'].shape}\")\n",
    "    print(f\"Target tokens: {tgt_inputs['input_ids'].shape}\")\n",
    "    \n",
    "    # Forward pass\n",
    "    with torch.no_grad():\n",
    "        output = transformer(\n",
    "            src_tokens=src_inputs['input_ids'],\n",
    "            tgt_tokens=tgt_inputs['input_ids'],\n",
    "            padding_mask=padding_mask\n",
    "        )\n",
    "    \n",
    "    print(\"\\nOutput Analysis:\")\n",
    "    print(f\"Output shape: {output.shape}\")  # Should be [batch_size, tgt_len, vocab_size]\n",
    "    \n",
    "    # Verify output is proper probability distribution\n",
    "    print(\"\\nProbability Distribution Check:\")\n",
    "    print(f\"Sum to 1: {torch.allclose(output.exp().sum(dim=-1), torch.ones_like(output.exp().sum(dim=-1)))}\")\n",
    "    print(f\"Max probability: {output.exp().max().item():.4f}\")\n",
    "    print(f\"Min probability: {output.exp().min().item():.4f}\")\n",
    "    \n",
    "    # Check if we can get predictions\n",
    "    predictions = output.argmax(dim=-1)\n",
    "    print(\"\\nSample Predictions:\")\n",
    "    print(\"Original target:\")\n",
    "    print(tgt_sequences[0])\n",
    "    print(\"\\nModel output (decoded):\")\n",
    "    print(tokenizer.decode(predictions[0]))\n",
    "    \n",
    "    # Test backward pass\n",
    "    transformer.train()\n",
    "    output = transformer(\n",
    "        src_tokens=src_inputs['input_ids'],\n",
    "        tgt_tokens=tgt_inputs['input_ids'],\n",
    "        padding_mask=padding_mask\n",
    "    )\n",
    "    \n",
    "    # Calculate loss (cross entropy)\n",
    "    loss = F.nll_loss(\n",
    "        output.view(-1, vocab_size),\n",
    "        tgt_inputs['input_ids'].view(-1)\n",
    "    )\n",
    "    \n",
    "    # Test backward pass\n",
    "    loss.backward()\n",
    "    \n",
    "    # Verify gradients\n",
    "    has_gradients = all(p.grad is not None for p in transformer.parameters())\n",
    "    print(\"\\nTraining Check:\")\n",
    "    print(f\"Loss value: {loss.item():.4f}\")\n",
    "    print(f\"Has gradients: {has_gradients}\")\n",
    "    \n",
    "    return output, predictions\n",
    "\n",
    "# Run test\n",
    "output, predictions = test_complete_transformer()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "741d45be-66bd-4880-8a63-9035c5cbc506",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "075d1822-3794-4f36-8142-878e31a2210e",
   "metadata": {},
   "source": [
    "## Visualization Section"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "eea1fcc6-5c86-4db8-903a-6a1e7cab88ac",
   "metadata": {},
   "source": [
    "### Positional Encoding Visualization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "caf9fe26-7537-4a5b-9a66-2313bb36c40c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1500x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1500x800 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "## Visualize positional encoding\n",
    "import os\n",
    "os.environ[\"TOKENIZERS_PARALLELISM\"] = \"false\"\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "\n",
    "def visualize_positional_encoding(seq_length=30, d_model=32):\n",
    "    # Generate positional encoding\n",
    "    pe = np.zeros((seq_length, d_model))\n",
    "    position = np.arange(seq_length)[:, np.newaxis]\n",
    "    div_term = np.exp(np.arange(0, d_model, 2) * -(np.log(10000.0) / d_model))\n",
    "    \n",
    "    pe[:, 0::2] = np.sin(position * div_term)\n",
    "    pe[:, 1::2] = np.cos(position * div_term)\n",
    "    \n",
    "    # Create visualization\n",
    "    plt.figure(figsize=(15, 8))\n",
    "    \n",
    "    # Plot first 8 dimensions\n",
    "    for dim in range(8):\n",
    "        plt.plot(pe[:, dim], label=f'dim {dim}')\n",
    "    \n",
    "    plt.xlabel('Position')\n",
    "    plt.ylabel('Value')\n",
    "    plt.title('Positional Encoding Patterns (First 8 Dimensions)')\n",
    "    plt.legend(bbox_to_anchor=(1.05, 1), loc='upper left')\n",
    "    plt.grid(True)\n",
    "    plt.tight_layout()\n",
    "    plt.show()\n",
    "    \n",
    "    # Also show heatmap of all dimensions\n",
    "    plt.figure(figsize=(15, 8))\n",
    "    plt.imshow(pe, cmap='RdBu', aspect='auto')\n",
    "    plt.colorbar()\n",
    "    plt.xlabel('Dimension')\n",
    "    plt.ylabel('Position')\n",
    "    plt.title('Positional Encoding Heatmap')\n",
    "    plt.tight_layout()\n",
    "    plt.show()\n",
    "\n",
    "# Using your model's positional encoding\n",
    "seq_length = 16  # From your example\n",
    "d_model = 768    # From your example\n",
    "\n",
    "# You might want to use a smaller d_model for visualization\n",
    "visualize_positional_encoding(seq_length=16, d_model=32)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ae4aa076-babd-4677-8e7b-5661debf5d30",
   "metadata": {},
   "source": [
    "### Attention Pattern Visualization\n",
    "Shows:\n",
    "- Self-attention patterns (causal masking)\n",
    "- Cross-attention patterns\n",
    "- Effect of padding masks\n",
    "- How attention weights distribute\n",
    "\n",
    "Implementation insights:\n",
    "- Use softmax before visualization\n",
    "- Show masking effects"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "e2e9d8c0-7a6a-4c50-85ef-0e805f785f57",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Attention Matrix Shape: torch.Size([2, 8, 5, 5])\n",
      "\n",
      "Attention Pattern (first head):\n",
      "tensor([[1.0000, 0.0000, 0.0000, 0.0000, 0.0000],\n",
      "        [0.4465, 0.5535, 0.0000, 0.0000, 0.0000],\n",
      "        [0.3403, 0.3496, 0.3101, 0.0000, 0.0000],\n",
      "        [0.1965, 0.3485, 0.1174, 0.3377, 0.0000],\n",
      "        [0.1563, 0.1571, 0.1859, 0.1948, 0.3059]])\n",
      "\n",
      "Future Token Analysis:\n",
      "Mean attention to future tokens: 0.00000000\n",
      "Max attention to future tokens: 0.00000000\n",
      "Causal masking working: Yes\n",
      "\n",
      "Present/Past Token Analysis:\n",
      "Mean attention to present/past tokens: 0.3333\n",
      "Has non-zero attention patterns: Yes\n",
      "\n",
      "Attention Sum Analysis:\n",
      "Mean attention sum (should be 1): 1.0000\n",
      "Max deviation from 1: 0.00000024\n"
     ]
    }
   ],
   "source": [
    "def test_decoder_causal_masking():\n",
    "    torch.manual_seed(42)\n",
    "    \n",
    "    # Test parameters\n",
    "    batch_size = 2\n",
    "    seq_length = 5\n",
    "    d_model = 512\n",
    "    d_ff = 2048\n",
    "    num_heads = 8\n",
    "    \n",
    "    decoder = TransformerDecoder(\n",
    "        d_model=d_model,\n",
    "        d_ff=d_ff,\n",
    "        num_head=num_heads,\n",
    "        dropout=0.1\n",
    "    )\n",
    "    decoder.eval()\n",
    "    \n",
    "    decoder_input = torch.randn(batch_size, seq_length, d_model)\n",
    "    encoder_output = torch.randn(batch_size, seq_length, d_model)\n",
    "    \n",
    "    attention_scores = []\n",
    "    \n",
    "    def attention_hook(module, input, output):\n",
    "        if not attention_scores:\n",
    "            # Apply softmax to get actual attention probabilities\n",
    "            scores = F.softmax(module.att_matrix, dim=-1)\n",
    "            attention_scores.append(scores.detach())\n",
    "    \n",
    "    decoder.att.register_forward_hook(attention_hook)\n",
    "    \n",
    "    with torch.no_grad():\n",
    "        output = decoder(decoder_input, encoder_output)\n",
    "    \n",
    "    att_weights = attention_scores[0]\n",
    "    \n",
    "    print(\"\\nAttention Matrix Shape:\", att_weights.shape)\n",
    "    \n",
    "    # Print attention pattern for first head of first batch\n",
    "    print(\"\\nAttention Pattern (first head):\")\n",
    "    print(att_weights[0, 0].round(decimals=4))\n",
    "    \n",
    "    # Check future tokens (should be 0)\n",
    "    future_attention = att_weights[:, :, torch.triu_indices(seq_length, seq_length, offset=1)[0], \n",
    "                                        torch.triu_indices(seq_length, seq_length, offset=1)[1]]\n",
    "    \n",
    "    print(\"\\nFuture Token Analysis:\")\n",
    "    print(f\"Mean attention to future tokens: {future_attention.mean():.8f}\")\n",
    "    print(f\"Max attention to future tokens: {future_attention.max():.8f}\")\n",
    "    print(\"Causal masking working:\", \"Yes\" if future_attention.mean() < 1e-7 else \"No\")\n",
    "    \n",
    "    # Check present/past tokens\n",
    "    present_past = att_weights[:, :, torch.tril_indices(seq_length, seq_length)[0],\n",
    "                                    torch.tril_indices(seq_length, seq_length)[1]]\n",
    "    \n",
    "    print(\"\\nPresent/Past Token Analysis:\")\n",
    "    print(f\"Mean attention to present/past tokens: {present_past.mean():.4f}\")\n",
    "    print(f\"Has non-zero attention patterns:\", \"Yes\" if present_past.mean() > 0 else \"No\")\n",
    "    \n",
    "    # Verify each position's attention sums to 1\n",
    "    attention_sums = att_weights.sum(dim=-1)\n",
    "    print(\"\\nAttention Sum Analysis:\")\n",
    "    print(f\"Mean attention sum (should be 1): {attention_sums.mean():.4f}\")\n",
    "    print(f\"Max deviation from 1: {(attention_sums - 1).abs().max():.8f}\")\n",
    "    \n",
    "    return att_weights\n",
    "\n",
    "attention_weights = test_decoder_causal_masking()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "a1623643-fa64-4f9e-b399-d9d0fc6c0f54",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Cross-Attention Matrix Shape: torch.Size([2, 8, 5, 7])\n",
      "\n",
      "Cross-Attention Pattern (first head):\n",
      "tensor([[0.1312, 0.1409, 0.1628, 0.1198, 0.1629, 0.1066, 0.1758],\n",
      "        [0.1423, 0.1153, 0.1398, 0.1557, 0.1642, 0.1757, 0.1070],\n",
      "        [0.0883, 0.2136, 0.0957, 0.2153, 0.1842, 0.0792, 0.1239],\n",
      "        [0.1307, 0.1610, 0.1614, 0.1063, 0.0865, 0.2293, 0.1249],\n",
      "        [0.1734, 0.0858, 0.1896, 0.1418, 0.1356, 0.1347, 0.1391]])\n",
      "\n",
      "Cross-Attention Analysis:\n",
      "Mean attention weight: 0.1429\n",
      "Min attention weight: 0.0396\n",
      "Max attention weight: 0.4592\n",
      "\n",
      "Attention Coverage:\n",
      "Each position's attention sums to 1: True\n",
      "Every decoder position attends to some encoder position: True\n",
      "\n",
      "Attention entropy (higher means more uniform attention): 1.8920\n"
     ]
    }
   ],
   "source": [
    "def test_decoder_cross_attention():\n",
    "    torch.manual_seed(42)\n",
    "    \n",
    "    # Test parameters\n",
    "    batch_size = 2\n",
    "    decoder_seq_len = 5\n",
    "    encoder_seq_len = 7  # Different length to make it interesting!\n",
    "    d_model = 512\n",
    "    d_ff = 2048\n",
    "    num_heads = 8\n",
    "    \n",
    "    decoder = TransformerDecoder(\n",
    "        d_model=d_model,\n",
    "        d_ff=d_ff,\n",
    "        num_head=num_heads,\n",
    "        dropout=0.1\n",
    "    )\n",
    "    decoder.eval()\n",
    "    \n",
    "    # Create input sequences\n",
    "    decoder_input = torch.randn(batch_size, decoder_seq_len, d_model)\n",
    "    encoder_output = torch.randn(batch_size, encoder_seq_len, d_model)\n",
    "    \n",
    "    # Store attention scores\n",
    "    cross_attention_scores = []\n",
    "    \n",
    "    def attention_hook(module, input, output):\n",
    "        # We want the second call to att (cross-attention)\n",
    "        if len(cross_attention_scores) < 2:\n",
    "            scores = F.softmax(module.att_matrix, dim=-1)\n",
    "            cross_attention_scores.append(scores.detach())\n",
    "    \n",
    "    decoder.att.register_forward_hook(attention_hook)\n",
    "    \n",
    "    # Forward pass\n",
    "    with torch.no_grad():\n",
    "        output = decoder(decoder_input, encoder_output)\n",
    "    \n",
    "    # Get cross-attention weights (second element in list)\n",
    "    cross_att_weights = cross_attention_scores[1]  # [batch, heads, decoder_seq_len, encoder_seq_len]\n",
    "    \n",
    "    print(\"\\nCross-Attention Matrix Shape:\", cross_att_weights.shape)\n",
    "    \n",
    "    # Print attention pattern for first head of first batch\n",
    "    print(\"\\nCross-Attention Pattern (first head):\")\n",
    "    print(cross_att_weights[0, 0].round(decimals=4))\n",
    "    \n",
    "    # Verify each decoder position attends to all encoder positions\n",
    "    attention_sums = cross_att_weights.sum(dim=-1)\n",
    "    zero_attention = (cross_att_weights == 0).all(dim=-1)\n",
    "    \n",
    "    print(\"\\nCross-Attention Analysis:\")\n",
    "    print(f\"Mean attention weight: {cross_att_weights.mean():.4f}\")\n",
    "    print(f\"Min attention weight: {cross_att_weights.min():.4f}\")\n",
    "    print(f\"Max attention weight: {cross_att_weights.max():.4f}\")\n",
    "    \n",
    "    print(\"\\nAttention Coverage:\")\n",
    "    print(f\"Each position's attention sums to 1: {torch.allclose(attention_sums, torch.ones_like(attention_sums))}\")\n",
    "    print(f\"Every decoder position attends to some encoder position: {not zero_attention.any()}\")\n",
    "    \n",
    "    # Check attention distribution\n",
    "    attention_entropy = -(cross_att_weights * torch.log(cross_att_weights + 1e-9)).sum(dim=-1).mean()\n",
    "    print(f\"\\nAttention entropy (higher means more uniform attention): {attention_entropy:.4f}\")\n",
    "    \n",
    "    return cross_att_weights\n",
    "\n",
    "# Run the test\n",
    "cross_attention_weights = test_decoder_cross_attention()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "4c7a7533-75c3-4575-af0a-cdc90e8b4815",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Cross-Attention Matrix Shape: torch.Size([2, 8, 5, 7])\n",
      "\n",
      "Cross-Attention Pattern (first head):\n",
      "(Last two encoder positions should have zero attention)\n",
      "tensor([[0.1828, 0.1964, 0.2268, 0.1669, 0.2271, 0.0000, 0.0000],\n",
      "        [0.1983, 0.1608, 0.1949, 0.2170, 0.2290, 0.0000, 0.0000],\n",
      "        [0.1107, 0.2680, 0.1200, 0.2702, 0.2311, 0.0000, 0.0000],\n",
      "        [0.2023, 0.2492, 0.2499, 0.1647, 0.1339, 0.0000, 0.0000],\n",
      "        [0.2387, 0.1182, 0.2611, 0.1953, 0.1867, 0.0000, 0.0000]])\n",
      "\n",
      "Masking Analysis:\n",
      "Mean attention to masked positions: 0.00000000\n",
      "Max attention to masked positions: 0.00000000\n",
      "Mean attention to unmasked positions: 0.2000\n",
      "\n",
      "Attention Coverage:\n",
      "Each position's attention sums to 1: True\n",
      "\n",
      "Unmasked Position Analysis:\n",
      "Min attention to unmasked positions: 0.0465\n",
      "Max attention to unmasked positions: 0.5392\n"
     ]
    }
   ],
   "source": [
    "def test_decoder_cross_attention_with_padding():\n",
    "    torch.manual_seed(42)\n",
    "    \n",
    "    # Test parameters\n",
    "    batch_size = 2\n",
    "    decoder_seq_len = 5\n",
    "    encoder_seq_len = 7\n",
    "    d_model = 512\n",
    "    d_ff = 2048\n",
    "    num_heads = 8\n",
    "    \n",
    "    decoder = TransformerDecoder(\n",
    "        d_model=d_model,\n",
    "        d_ff=d_ff,\n",
    "        num_head=num_heads,\n",
    "        dropout=0.1\n",
    "    )\n",
    "    decoder.eval()\n",
    "    \n",
    "    # Create input sequences\n",
    "    decoder_input = torch.randn(batch_size, decoder_seq_len, d_model)\n",
    "    encoder_output = torch.randn(batch_size, encoder_seq_len, d_model)\n",
    "    \n",
    "    # Create padding mask for encoder outputs\n",
    "    # Mask out last 2 positions (as if they were padding in encoder output)\n",
    "    padding_mask = torch.ones(batch_size, decoder_seq_len, encoder_seq_len)\n",
    "    padding_mask[:, :, -2:] = float('-inf')  # Mask positions 5,6\n",
    "    padding_mask = padding_mask.unsqueeze(1)  # Add head dimension [batch, 1, decoder_seq, encoder_seq]\n",
    "    \n",
    "    cross_attention_scores = []\n",
    "    \n",
    "    def attention_hook(module, input, output):\n",
    "        if len(cross_attention_scores) < 2:\n",
    "            scores = F.softmax(module.att_matrix, dim=-1)\n",
    "            cross_attention_scores.append(scores.detach())\n",
    "    \n",
    "    decoder.att.register_forward_hook(attention_hook)\n",
    "    \n",
    "    # Forward pass\n",
    "    with torch.no_grad():\n",
    "        output = decoder(decoder_input, encoder_output, padding_mask)\n",
    "    \n",
    "    # Get cross-attention weights (second element)\n",
    "    cross_att_weights = cross_attention_scores[1]\n",
    "    \n",
    "    print(\"\\nCross-Attention Matrix Shape:\", cross_att_weights.shape)\n",
    "    \n",
    "    print(\"\\nCross-Attention Pattern (first head):\")\n",
    "    print(\"(Last two encoder positions should have zero attention)\")\n",
    "    print(cross_att_weights[0, 0].round(decimals=4))\n",
    "    \n",
    "    # Analyze masked positions (last two columns)\n",
    "    masked_attention = cross_att_weights[:, :, :, -2:]\n",
    "    unmasked_attention = cross_att_weights[:, :, :, :-2]\n",
    "    \n",
    "    print(\"\\nMasking Analysis:\")\n",
    "    print(f\"Mean attention to masked positions: {masked_attention.mean():.8f}\")\n",
    "    print(f\"Max attention to masked positions: {masked_attention.max():.8f}\")\n",
    "    print(f\"Mean attention to unmasked positions: {unmasked_attention.mean():.4f}\")\n",
    "    \n",
    "    # Verify attention still sums to 1 (only over unmasked positions)\n",
    "    attention_sums = cross_att_weights.sum(dim=-1)\n",
    "    \n",
    "    print(\"\\nAttention Coverage:\")\n",
    "    print(f\"Each position's attention sums to 1: {torch.allclose(attention_sums, torch.ones_like(attention_sums), atol=1e-6)}\")\n",
    "    \n",
    "    # Analyze attention distribution over unmasked positions\n",
    "    print(\"\\nUnmasked Position Analysis:\")\n",
    "    print(f\"Min attention to unmasked positions: {unmasked_attention.min():.4f}\")\n",
    "    print(f\"Max attention to unmasked positions: {unmasked_attention.max():.4f}\")\n",
    "    \n",
    "    return cross_att_weights\n",
    "\n",
    "# Run the test\n",
    "cross_attention_weights = test_decoder_cross_attention_with_padding()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a33616a1-e8be-4506-88cb-a9b32062b76d",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "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": "3.11.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}