chore: import upstream snapshot with attribution

2026-07-13 12:45:52 +08:00
commit aa6c9c0720
43 changed files with 5178 additions and 0 deletions
@@ -0,0 +1,110 @@
+{
+  "cells": [
+    {
+      "cell_type": "code",
+      "metadata": {},
+      "source": [
+        "# code by Tae Hwan Jung @graykode\n",
+        "import numpy as np\n",
+        "import torch\n",
+        "import torch.nn as nn\n",
+        "import torch.optim as optim\n",
+        "\n",
+        "def make_batch():\n",
+        "    input_batch = []\n",
+        "    target_batch = []\n",
+        "\n",
+        "    words = sentence.split()\n",
+        "    for i, word in enumerate(words[:-1]):\n",
+        "        input = [word_dict[n] for n in words[:(i + 1)]]\n",
+        "        input = input + [0] * (max_len - len(input))\n",
+        "        target = word_dict[words[i + 1]]\n",
+        "        input_batch.append(np.eye(n_class)[input])\n",
+        "        target_batch.append(target)\n",
+        "\n",
+        "    return input_batch, target_batch\n",
+        "\n",
+        "class BiLSTM(nn.Module):\n",
+        "    def __init__(self):\n",
+        "        super(BiLSTM, self).__init__()\n",
+        "\n",
+        "        self.lstm = nn.LSTM(input_size=n_class, hidden_size=n_hidden, bidirectional=True)\n",
+        "        self.W = nn.Linear(n_hidden * 2, n_class, bias=False)\n",
+        "        self.b = nn.Parameter(torch.ones([n_class]))\n",
+        "\n",
+        "    def forward(self, X):\n",
+        "        input = X.transpose(0, 1)  # input : [n_step, batch_size, n_class]\n",
+        "\n",
+        "        hidden_state = torch.zeros(1*2, len(X), n_hidden)   # [num_layers(=1) * num_directions(=2), batch_size, n_hidden]\n",
+        "        cell_state = torch.zeros(1*2, len(X), n_hidden)     # [num_layers(=1) * num_directions(=2), batch_size, n_hidden]\n",
+        "\n",
+        "        outputs, (_, _) = self.lstm(input, (hidden_state, cell_state))\n",
+        "        outputs = outputs[-1]  # [batch_size, n_hidden]\n",
+        "        model = self.W(outputs) + self.b  # model : [batch_size, n_class]\n",
+        "        return model\n",
+        "\n",
+        "if __name__ == '__main__':\n",
+        "    n_hidden = 5 # number of hidden units in one cell\n",
+        "\n",
+        "    sentence = (\n",
+        "        'Lorem ipsum dolor sit amet consectetur adipisicing elit '\n",
+        "        'sed do eiusmod tempor incididunt ut labore et dolore magna '\n",
+        "        'aliqua Ut enim ad minim veniam quis nostrud exercitation'\n",
+        "    )\n",
+        "\n",
+        "    word_dict = {w: i for i, w in enumerate(list(set(sentence.split())))}\n",
+        "    number_dict = {i: w for i, w in enumerate(list(set(sentence.split())))}\n",
+        "    n_class = len(word_dict)\n",
+        "    max_len = len(sentence.split())\n",
+        "\n",
+        "    model = BiLSTM()\n",
+        "\n",
+        "    criterion = nn.CrossEntropyLoss()\n",
+        "    optimizer = optim.Adam(model.parameters(), lr=0.001)\n",
+        "\n",
+        "    input_batch, target_batch = make_batch()\n",
+        "    input_batch = torch.FloatTensor(input_batch)\n",
+        "    target_batch = torch.LongTensor(target_batch)\n",
+        "\n",
+        "    # Training\n",
+        "    for epoch in range(10000):\n",
+        "        optimizer.zero_grad()\n",
+        "        output = model(input_batch)\n",
+        "        loss = criterion(output, target_batch)\n",
+        "        if (epoch + 1) % 1000 == 0:\n",
+        "            print('Epoch:', '%04d' % (epoch + 1), 'cost =', '{:.6f}'.format(loss))\n",
+        "\n",
+        "        loss.backward()\n",
+        "        optimizer.step()\n",
+        "\n",
+        "    predict = model(input_batch).data.max(1, keepdim=True)[1]\n",
+        "    print(sentence)\n",
+        "    print([number_dict[n.item()] for n in predict.squeeze()])\n"
+      ],
+      "outputs": [],
+      "execution_count": null
+    }
+  ],
+  "metadata": {
+    "anaconda-cloud": {},
+    "kernelspec": {
+      "display_name": "Python 3",
+      "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": "3.6.1"
+    }
+  },
+  "nbformat": 4,
+  "nbformat_minor": 4
+}
@@ -0,0 +1,77 @@
+# %%
+# code by Tae Hwan Jung @graykode
+import numpy as np
+import torch
+import torch.nn as nn
+import torch.optim as optim
+
+def make_batch():
+    input_batch = []
+    target_batch = []
+
+    words = sentence.split()
+    for i, word in enumerate(words[:-1]):
+        input = [word_dict[n] for n in words[:(i + 1)]]
+        input = input + [0] * (max_len - len(input))
+        target = word_dict[words[i + 1]]
+        input_batch.append(np.eye(n_class)[input])
+        target_batch.append(target)
+
+    return input_batch, target_batch
+
+class BiLSTM(nn.Module):
+    def __init__(self):
+        super(BiLSTM, self).__init__()
+
+        self.lstm = nn.LSTM(input_size=n_class, hidden_size=n_hidden, bidirectional=True)
+        self.W = nn.Linear(n_hidden * 2, n_class, bias=False)
+        self.b = nn.Parameter(torch.ones([n_class]))
+
+    def forward(self, X):
+        input = X.transpose(0, 1)  # input : [n_step, batch_size, n_class]
+
+        hidden_state = torch.zeros(1*2, len(X), n_hidden)   # [num_layers(=1) * num_directions(=2), batch_size, n_hidden]
+        cell_state = torch.zeros(1*2, len(X), n_hidden)     # [num_layers(=1) * num_directions(=2), batch_size, n_hidden]
+
+        outputs, (_, _) = self.lstm(input, (hidden_state, cell_state))
+        outputs = outputs[-1]  # [batch_size, n_hidden]
+        model = self.W(outputs) + self.b  # model : [batch_size, n_class]
+        return model
+
+if __name__ == '__main__':
+    n_hidden = 5 # number of hidden units in one cell
+
+    sentence = (
+        'Lorem ipsum dolor sit amet consectetur adipisicing elit '
+        'sed do eiusmod tempor incididunt ut labore et dolore magna '
+        'aliqua Ut enim ad minim veniam quis nostrud exercitation'
+    )
+
+    word_dict = {w: i for i, w in enumerate(list(set(sentence.split())))}
+    number_dict = {i: w for i, w in enumerate(list(set(sentence.split())))}
+    n_class = len(word_dict)
+    max_len = len(sentence.split())
+
+    model = BiLSTM()
+
+    criterion = nn.CrossEntropyLoss()
+    optimizer = optim.Adam(model.parameters(), lr=0.001)
+
+    input_batch, target_batch = make_batch()
+    input_batch = torch.FloatTensor(input_batch)
+    target_batch = torch.LongTensor(target_batch)
+
+    # Training
+    for epoch in range(10000):
+        optimizer.zero_grad()
+        output = model(input_batch)
+        loss = criterion(output, target_batch)
+        if (epoch + 1) % 1000 == 0:
+            print('Epoch:', '%04d' % (epoch + 1), 'cost =', '{:.6f}'.format(loss))
+
+        loss.backward()
+        optimizer.step()
+
+    predict = model(input_batch).data.max(1, keepdim=True)[1]
+    print(sentence)
+    print([number_dict[n.item()] for n in predict.squeeze()])