notebook.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Transformer for Source Code"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load Vocab from File"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:root:Loaded vocab from: output/vocab.pkl\n",
      "INFO:root:Vocab size: 100002\n"
     ]
    }
   ],
   "source": [
    "import torch\n",
    "from models.trav_trans.dataset import Vocab\n",
    "\n",
    "vocab = Vocab(\"output/vocab.pkl\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load Model from File"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import model\n",
    "model = model.from_file(\"output/model-8.pt\", vocab)\n",
    "m = model.eval()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Plot Loss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x2475aa1f4c0>]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "from matplotlib.pyplot import figure\n",
    "import pickle\n",
    "\n",
    "with open(\"output/losses.pickle\", \"rb\") as fin:\n",
    "    losses = pickle.load(fin)\n",
    "l = [loss[2] for loss in losses]\n",
    "figure(figsize=(20, 10))\n",
    "plt.xlabel(\"Iteration (x 1000)\")\n",
    "plt.ylabel(\"Loss\")\n",
    "plt.plot(range(len(losses)), l)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "from prettytable import PrettyTable\n",
    "\n",
    "def count_parameters(model):\n",
    "    table = PrettyTable([\"Modules\", \"Parameters\"])\n",
    "    total_params = 0\n",
    "    for name, parameter in model.named_parameters():\n",
    "        if not parameter.requires_grad: continue\n",
    "        param = parameter.numel()\n",
    "        table.add_row([name, param])\n",
    "        total_params+=param\n",
    "    print(table)\n",
    "    print(f\"Total Trainable Params: {total_params}\")\n",
    "    return total_params\n",
    "    \n",
    "# count_parameters(model)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Generate Tokens"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "bf4acaae083e4bb38ef90a9230d7b5a0",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Textarea(value='a = [1, 2, 3, 4, 5]\\nprint(len(a))', description='Python Code', layout=Layout(height='300px', …"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from ipywidgets import interact, widgets\n",
    "from IPython.display import display\n",
    "\n",
    "textarea = widgets.Textarea(\n",
    "    value=\"a = [1, 2, 3, 4, 5]\\nprint(len(a))\",\n",
    "    placeholder='Type something',\n",
    "    description='Python Code',\n",
    "    disabled=False,\n",
    "    layout=widgets.Layout(width=\"100%\", height=\"300px\")\n",
    ")\n",
    "display(textarea)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "import code_parser, generate_new_trees, json, utils\n",
    "ast_raw = code_parser.parse_code(textarea.value)\n",
    "ast = json.dumps(generate_new_trees.convert(json.loads(ast_raw)))\n",
    "ast_original = ast\n",
    "# Remove last 2 nodes\n",
    "ast = json.dumps(json.loads(ast)[:-2])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Complete tree\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from IPython.display import Image\n",
    "import generate_graph\n",
    "\n",
    "print(\"Complete tree\")\n",
    "outfile = \"outfile_original.png\"\n",
    "generate_graph.visualize(ast_original, outfile)\n",
    "Image(outfile)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Remove last two nodes\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(\"Remove last two nodes\")\n",
    "outfile = \"outfile.png\"\n",
    "generate_graph.visualize(ast, outfile)\n",
    "Image(outfile)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch.nn.functional as F\n",
    "\n",
    "def generate_greedy(idx, iterations):\n",
    "    input = idx\n",
    "    for i in range(iterations):\n",
    "        with torch.no_grad():\n",
    "            output = model(torch.tensor(input), None)\n",
    "        logits = output[-1]\n",
    "        softmax = F.softmax(logits, dim=-1)\n",
    "        top_k_values, top_k_indices = torch.topk(softmax, 5)\n",
    "        for e, v in enumerate(top_k_values):\n",
    "            print(\"Prediction {}: {}({})\".format(i, vocab.idx2vocab[top_k_indices[e]], round(float(v), 2)))\n",
    "        input.append(top_k_indices[0].item())\n",
    "    return input"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "\n",
    "def generate_beams(model, context, num_beams=2, depth=2):\n",
    "    beam_heads = [{\"context\": context, \"prob\": []}] * num_beams\n",
    "    \n",
    "    for d in range(depth):\n",
    "        for b in range(num_beams):\n",
    "            temp_context = beam_heads[b][\"context\"].copy()\n",
    "            with torch.no_grad():\n",
    "                top_k_values, top_k_indices = torch.topk(F.softmax(model(torch.tensor(temp_context), None)[-1], dim=-1), num_beams)\n",
    "            new_context = beam_heads[b][\"context\"].copy()\n",
    "            new_context.append(top_k_indices[b].item())\n",
    "            new_probability = beam_heads[b][\"prob\"].copy()\n",
    "            new_probability.append(top_k_values[b].item())\n",
    "            beam_heads[b] = {\"context\": new_context, \"prob\": new_probability}\n",
    "    top_beam_id = 0\n",
    "    for b in range(num_beams):\n",
    "        if np.prod(beam_heads[b][\"prob\"]) > np.prod(beam_heads[top_beam_id][\"prob\"]):\n",
    "            top_beam_id = b\n",
    "    return [vocab.idx2vocab[i] for i in beam_heads[top_beam_id][\"context\"]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "def generate_test(model, context, depth=2, top_k=10):\n",
    "    model.eval()\n",
    "    with torch.no_grad():\n",
    "        output = model(torch.tensor(context), None)[-1]\n",
    "        top_k_values, top_k_indices = torch.topk(output, top_k)\n",
    "        return top_k_values, top_k_indices"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Original: \n",
      "Module Expr Call NameLoad call NameLoad a NameLoad b NameLoad c\n",
      "--------------------\n",
      "Prediction: \n",
      "Module Expr Call NameLoad call NameLoad a NameLoad b Expr Call\n"
     ]
    }
   ],
   "source": [
    "num_gen = 2\n",
    "\n",
    "nodes = utils.get_dfs(json.loads(ast))\n",
    "nodes_original = utils.get_dfs(json.loads(ast_original))\n",
    "idx = vocab.convert([nodes, 0])[0]\n",
    "idx_original = vocab.convert([nodes_original, 0])[0]\n",
    "print(\"Original: \")\n",
    "print(\" \".join([vocab.idx2vocab[n] for n in idx_original]))\n",
    "print(\"-\" * 20)\n",
    "print(\"Prediction: \")\n",
    "print(\" \".join(generate_beams(model, idx, num_beams=4, depth=2)))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:root:Writing metrics to: output\\train_metrics.txt\n",
      "INFO:root:Loaded vocab from: output\\vocab.pkl\n",
      "INFO:root:Vocab size: 100002\n",
      "INFO:root:Loaded dataset from output\\train_converted.txt\n"
     ]
    }
   ],
   "source": [
    "import models.trav_trans.dataset as dataset\n",
    "setup = dataset.Setup(\"output\", \"output/dps.txt\", \"output/ids.txt\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "====================\n",
      "Type Prediction:\n",
      "Expected Token: \n",
      "\talias\n",
      "Predicted Token: \n",
      "\talias\n",
      "\tImportFrom\n",
      "\tClassDef\n",
      "\tImport\n",
      "\tidentifier\n",
      "\tAssign\n",
      "\tFunctionDef\n",
      "\tExpr\n",
      "\tTryExcept\n",
      "\tRaise\n",
      "Reciprocal Rank for Type: 1.0\n",
      "\n",
      "Value Prediction:\n",
      "Expected Token: \n",
      "\t<unk_token>\n",
      "Predicted Token: \n",
      "\t<unk_token>\n",
      "\t*\n",
      "\tbase\n",
      "\tcontext\n",
      "\tTestCase\n",
      "\tNone\n",
      "\topen\n",
      "\tparams\n",
      "\tcount\n",
      "\tutils\n",
      "Reciprocal Rank for Value: 1.0\n"
     ]
    }
   ],
   "source": [
    "import torch\n",
    "dataloader = torch.utils.data.DataLoader(\n",
    "    setup.dataset,\n",
    "    batch_size=1,\n",
    "    collate_fn=lambda b: dataset.Dataset.collate(b, vocab.pad_idx)\n",
    ")\n",
    "vocab = setup.vocab\n",
    "batch = next(iter(dataloader))\n",
    "\n",
    "x = batch[\"input_seq\"][0]\n",
    "y = batch[\"target_seq\"][0]\n",
    "ids = batch[\"ids\"][\"leaf_ids\"]\n",
    "\n",
    "predict_id = 4\n",
    "\n",
    "y_type = x[ids[predict_id]].item()\n",
    "y_value = y[ids[predict_id]].item()\n",
    "\n",
    "model.eval()\n",
    "with torch.no_grad():\n",
    "    output = model(x[range(ids[predict_id])], None)\n",
    "    # Predict next 2 instead of 1\n",
    "    type_prediction = generate_test(model, [i.item() for i in x[range(ids[predict_id])]])\n",
    "    value_prediction = generate_test(model, [i.item() for i in x[range(ids[predict_id] + 1)]])\n",
    "    print(\"=\" * 20)\n",
    "    # Check if type in prediction\n",
    "    print(\"Type Prediction:\\nExpected Token: \\n\\t{}\\nPredicted Token: \\n\\t{}\".format(vocab.idx2vocab[y_type], \"\\n\\t\".join([vocab.idx2vocab[p.item()] for p in type_prediction[1]])))\n",
    "    if y_type in type_prediction[1]:\n",
    "        print(\"Reciprocal Rank for Type: {}\".format(1 / ((type_prediction[1] == y_type).nonzero(as_tuple=True)[0].item() + 1)))\n",
    "    else:\n",
    "        print(\"Reciprocal Rank for Type: 0\")\n",
    "    print()\n",
    "    # Check if value in prediction\n",
    "    print(\"Value Prediction:\\nExpected Token: \\n\\t{}\\nPredicted Token: \\n\\t{}\".format(vocab.idx2vocab[y_value], \"\\n\\t\".join([vocab.idx2vocab[p.item()] for p in value_prediction[1]])))\n",
    "    if y_value in value_prediction[1]:\n",
    "        print(\"Reciprocal Rank for Value: {}\".format(1 / ((value_prediction[1] == y_value).nonzero(as_tuple=True)[0].item() + 1)))\n",
    "    else:\n",
    "        print(\"Reciprocal Rank for Value: 0\")\n",
    "    #print(\"Next 2 node predictions for {}: \\n\\t{}\\nTarget:\\n\\t{}\".format(ids[predict_id], [vocab.idx2vocab[p] for p in prediction], [vocab.idx2vocab[t] for t in to_predict]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "torch.return_types.topk(\n",
       "values=tensor([[23.3407, 23.1023, 22.7032, 21.3810, 20.3852, 19.9095, 19.0571, 18.0823,\n",
       "         17.7750, 17.3384],\n",
       "        [27.6575, 23.7110, 21.4206, 20.2025, 19.1513, 19.0801, 18.8871, 18.7930,\n",
       "         18.5485, 17.8772],\n",
       "        [12.7799,  5.6208,  4.3699,  4.3456,  4.0439,  4.0387,  4.0257,  3.9484,\n",
       "          3.8925,  3.8875],\n",
       "        [26.8126, 26.5770, 25.6768, 23.9964, 23.6479, 23.5322, 22.8802, 21.5889,\n",
       "         21.5868, 20.7786],\n",
       "        [ 9.3046,  8.7559,  6.0641,  5.9062,  5.6905,  5.6143,  5.5392,  5.5338,\n",
       "          5.4869,  5.4233],\n",
       "        [38.8529, 25.4993, 24.6597, 23.6995, 23.4867, 23.3620, 21.9334, 21.5476,\n",
       "         20.3766, 20.0839],\n",
       "        [ 8.1741,  5.8852,  4.5108,  3.9901,  3.9772,  3.9386,  3.8436,  3.7430,\n",
       "          3.7066,  3.6938],\n",
       "        [22.7877, 21.5378, 21.2256, 21.0314, 20.9769, 20.0528, 19.9314, 19.1921,\n",
       "         17.8202, 17.7922],\n",
       "        [ 9.5319,  9.2854,  6.5307,  6.2191,  5.9567,  5.9373,  5.8132,  5.7154,\n",
       "          5.6621,  5.5593],\n",
       "        [34.5502, 24.4997, 22.5749, 22.1108, 21.3528, 21.1258, 20.3926, 19.7474,\n",
       "         19.0733, 18.2711]]),\n",
       "indices=tensor([[    10,     30,     41,      6,     38,     17,     71,     87,      3,\n",
       "             21],\n",
       "        [     3,      4,      0,     40,     34,      7,     22,      9,      2,\n",
       "             33],\n",
       "        [100000,     62,    114,    294,    270,      3,    116,    150,     91,\n",
       "             75],\n",
       "        [    41,     30,      6,     38,     10,     17,     71,     87,     19,\n",
       "             21],\n",
       "        [100000,    292,   1635,   1878,   1268,     21,   1196,    859,    722,\n",
       "            816],\n",
       "        [    21,     30,     41,      6,     17,     38,     10,    119,     16,\n",
       "             71],\n",
       "        [100000,    238,    234,    563,    456,     60,   2427,     52,    186,\n",
       "             85],\n",
       "        [    30,     21,     41,     38,      6,     10,    119,     17,     19,\n",
       "             71],\n",
       "        [100000,   2427,     21,    859,   2161,    722,   1635,   3321,   2281,\n",
       "           1196],\n",
       "        [    21,     30,     38,     41,    119,      6,     17,     10,     71,\n",
       "             58]]))"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
    "b = next(iter(dataloader))\n",
    "x = b[\"input_seq\"]\n",
    "y = b[\"target_seq\"]\n",
    "leaf_ids = b[\"ids\"][\"leaf_ids\"]\n",
    "torch.topk(output, 10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'django.utils.translation'"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "vocab.idx2vocab[1278]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[ 15.6289,  11.2085,  15.1720,  ...,  -7.5906,   1.5328, -20.9393],\n",
       "        [ 21.4206,   9.7577,  18.5485,  ...,  -5.6370,   1.7067, -21.8225],\n",
       "        [  1.2639,   2.6546,  -0.1263,  ...,  -7.5184,  12.7799, -13.0211],\n",
       "        ...,\n",
       "        [ 13.8472,  12.8210,  13.2294,  ...,  -7.2852,   0.7094, -21.2433],\n",
       "        [  2.9625,   2.5933,   0.2749,  ...,  -4.6713,   9.5319, -10.5143],\n",
       "        [ 11.1520,   3.4922,  11.0843,  ...,  -9.9520,  -1.0530, -21.9037]])"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "output"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataloader = torch.utils.data.DataLoader(\n",
    "        setup.dataset,\n",
    "        batch_size = 3,\n",
    "        collate_fn = lambda b: dataset.Dataset.collate(b, setup.vocab.pad_idx)\n",
    "    )\n",
    "\n",
    "batch = next(iter(dataloader))\n",
    "x = batch[\"input_seq\"]\n",
    "y = batch[\"target_seq\"]\n",
    "with torch.no_grad():\n",
    "    output = model(x, None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5\n",
      "torch.Size([3, 999, 100002])\n",
      "torch.Size([3, 999, 5])\n",
      "Numpy before: (3, 999, 100002)\n",
      "Numpy after: (3, 5, 100002)\n",
      "Pytorch: torch.Size([3, 5, 100002])\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "ids = [4, 5, 7, 8, 12]\n",
    "print(len(ids))\n",
    "print(output.size())\n",
    "print(output[..., ids].size())\n",
    "s = output.numpy()\n",
    "type(s)\n",
    "print(\"Numpy before: {}\".format(s.shape))\n",
    "print(\"Numpy after: {}\".format(s[:, ids].shape))\n",
    "print(\"Pytorch: {}\".format(output[:, ids].size()))"
   ]
  }
 ],
 "metadata": {
  "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": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}