1101 lines
368 KiB
Plaintext
1101 lines
368 KiB
Plaintext
{
|
||
"cells": [
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 12,
|
||
"id": "263af9e9",
|
||
"metadata": {},
|
||
"outputs": [],
|
||
"source": [
|
||
"import pickle, tensorflow as tf, numpy as np\n",
|
||
"from tensorflow.keras.models import Model, load_model\n",
|
||
"from tensorflow.keras.layers import (\n",
|
||
" Input,\n",
|
||
" Embedding,\n",
|
||
" SpatialDropout1D,\n",
|
||
" Bidirectional,\n",
|
||
" LSTM,\n",
|
||
" TimeDistributed,\n",
|
||
" Dense,\n",
|
||
" \n",
|
||
")\n",
|
||
"from tensorflow.keras.preprocessing.sequence import pad_sequences\n",
|
||
"from sklearn.model_selection import train_test_split\n",
|
||
"from collections import Counter\n",
|
||
"from itertools import zip_longest"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 13,
|
||
"id": "4fc87f1b",
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"total kalimat 2231\n",
|
||
"NER Label Count || SRL Label Count \n",
|
||
"-------------------------------------------------------\n",
|
||
"O 11464 || O 5702 \n",
|
||
"TIME 349 || ARGM-TMP 2089 \n",
|
||
"PER 1951 || ARG0 2637 \n",
|
||
"LOC 2184 || V 2461 \n",
|
||
"DATE 1461 || ARG1 2879 \n",
|
||
"ETH 430 || ARGM-LOC 2090 \n",
|
||
"EVENT 304 || ARG2 547 \n",
|
||
"MISC 466 || ARGM-MOD 78 \n",
|
||
"ORG 317 || ARGM-MNR 200 \n",
|
||
" || ARGM-NEG 19 \n",
|
||
" || ARGM-DIR 41 \n",
|
||
" || ARGM-CAU 146 \n",
|
||
" || ARG3 37 \n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"data = []\n",
|
||
"with open(\"../dataset/ner_srl_without_bio.tsv\", encoding=\"utf-8\") as f:\n",
|
||
" tok, ner, srl = [], [], []\n",
|
||
" for line in f:\n",
|
||
" line = line.strip()\n",
|
||
" if not line:\n",
|
||
" if tok:\n",
|
||
" data.append({\"tokens\": tok, \"labels_ner\": ner, \"labels_srl\": srl})\n",
|
||
" tok, ner, srl = [], [], []\n",
|
||
" else:\n",
|
||
" t, n, s = line.split(\"\\t\")\n",
|
||
" tok.append(t.lower())\n",
|
||
" ner.append(n.strip())\n",
|
||
" srl_label = s.strip()\n",
|
||
" if srl_label == \"MISC\":\n",
|
||
" srl_label = \"O\"\n",
|
||
" srl.append(srl_label)\n",
|
||
"\n",
|
||
"\n",
|
||
"print(\"total kalimat \", len(data))\n",
|
||
"# ——————————————————\n",
|
||
"sentences = [d[\"tokens\"] for d in data]\n",
|
||
"labels_ner = [d[\"labels_ner\"] for d in data]\n",
|
||
"labels_srl = [d[\"labels_srl\"] for d in data]\n",
|
||
"\n",
|
||
"ner_counter = Counter(label for seq in labels_ner for label in seq)\n",
|
||
"\n",
|
||
"srl_counter = Counter(label for seq in labels_srl for label in seq)\n",
|
||
"\n",
|
||
"\n",
|
||
"print(f\"{'NER Label':<15} {'Count':<10} || {'SRL Label':<15} {'Count':<10}\")\n",
|
||
"print(\"-\" * 55)\n",
|
||
"\n",
|
||
"for (ner_label, ner_count), (srl_label, srl_count) in zip_longest(ner_counter.items(), srl_counter.items(), fillvalue=('', '')):\n",
|
||
" print(f\"{ner_label:<15} {ner_count:<10} || {srl_label:<15} {srl_count:<10}\")"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 14,
|
||
"id": "8dda2d6c",
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"NER -> Total Labels: 18926, O Count: 11464, O Percentage: 60.57%\n",
|
||
"SRL -> Total Labels: 18926, O Count: 5702, O Percentage: 30.13%\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"\n",
|
||
"\n",
|
||
"def calculate_o_percentage(labels):\n",
|
||
" counter = Counter(label for seq in labels for label in seq)\n",
|
||
" total = sum(counter.values())\n",
|
||
" count_o = counter.get(\"O\", 0)\n",
|
||
" percent_o = (count_o / total) * 100 if total > 0 else 0\n",
|
||
" return percent_o, total, count_o\n",
|
||
"\n",
|
||
"# Hitung persentase 'O' untuk NER\n",
|
||
"o_ner_percent, total_ner, o_ner_count = calculate_o_percentage(labels_ner)\n",
|
||
"\n",
|
||
"# Hitung persentase 'O' untuk SRL\n",
|
||
"o_srl_percent, total_srl, o_srl_count = calculate_o_percentage(labels_srl)\n",
|
||
"\n",
|
||
"print(f\"NER -> Total Labels: {total_ner}, O Count: {o_ner_count}, O Percentage: {o_ner_percent:.2f}%\")\n",
|
||
"print(f\"SRL -> Total Labels: {total_srl}, O Count: {o_srl_count}, O Percentage: {o_srl_percent:.2f}%\")\n"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 15,
|
||
"id": "48553e6b",
|
||
"metadata": {},
|
||
"outputs": [],
|
||
"source": [
|
||
"PAD_TOKEN = \"<PAD>\"\n",
|
||
"words = sorted({w for s in sentences for w in s})\n",
|
||
"\n",
|
||
"ner_tags = sorted({t for seq in labels_ner for t in seq})\n",
|
||
"srl_tags = sorted({t for seq in labels_srl for t in seq})\n",
|
||
"\n",
|
||
"ner_tags.insert(0, PAD_TOKEN)\n",
|
||
"srl_tags.insert(0, PAD_TOKEN)\n",
|
||
"\n",
|
||
"word2idx = {w: i + 2 for i, w in enumerate(words)}\n",
|
||
"word2idx[\"PAD\"] = 0\n",
|
||
"word2idx[\"UNK\"] = 1\n",
|
||
"\n",
|
||
"tag2idx_ner = {t: i for i, t in enumerate(ner_tags)}\n",
|
||
"tag2idx_srl = {t: i for i, t in enumerate(srl_tags)}\n",
|
||
"idx2tag_ner = {i: t for t, i in tag2idx_ner.items()}\n",
|
||
"idx2tag_srl = {i: t for t, i in tag2idx_srl.items()}\n"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 16,
|
||
"id": "096967e8",
|
||
"metadata": {},
|
||
"outputs": [],
|
||
"source": [
|
||
"X = [[word2idx.get(w, 1) for w in s] for s in sentences]\n",
|
||
"y_ner = [[tag2idx_ner[t] for t in seq] for seq in labels_ner]\n",
|
||
"y_srl = [[tag2idx_srl[t] for t in seq] for seq in labels_srl]\n",
|
||
"\n",
|
||
"maxlen = max(map(len, X))\n",
|
||
"pad_id = tag2idx_ner[PAD_TOKEN]\n",
|
||
"\n",
|
||
"X = pad_sequences(X, maxlen=maxlen, padding=\"post\", value=0)\n",
|
||
"y_ner = pad_sequences(y_ner, maxlen=maxlen, padding=\"post\", value=pad_id)\n",
|
||
"y_srl = pad_sequences(y_srl, maxlen=maxlen, padding=\"post\", value=pad_id)\n",
|
||
"\n",
|
||
"mask = (y_ner != pad_id).astype(\"float32\")"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 17,
|
||
"id": "a26893cc",
|
||
"metadata": {},
|
||
"outputs": [],
|
||
"source": [
|
||
"splits = train_test_split(\n",
|
||
" X,\n",
|
||
" y_ner,\n",
|
||
" y_srl,\n",
|
||
" mask,\n",
|
||
" test_size=0.2,\n",
|
||
" random_state=42,\n",
|
||
" shuffle=True,\n",
|
||
")\n",
|
||
"X_tr, X_te, ner_tr, ner_te, srl_tr, srl_te, m_tr, m_te = splits"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 18,
|
||
"id": "1b4a1c61",
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"text/html": [
|
||
"<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\">Model: \"functional_1\"</span>\n",
|
||
"</pre>\n"
|
||
],
|
||
"text/plain": [
|
||
"\u001b[1mModel: \"functional_1\"\u001b[0m\n"
|
||
]
|
||
},
|
||
"metadata": {},
|
||
"output_type": "display_data"
|
||
},
|
||
{
|
||
"data": {
|
||
"text/html": [
|
||
"<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\">┏━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━┓\n",
|
||
"┃<span style=\"font-weight: bold\"> Layer (type) </span>┃<span style=\"font-weight: bold\"> Output Shape </span>┃<span style=\"font-weight: bold\"> Param # </span>┃<span style=\"font-weight: bold\"> Connected to </span>┃\n",
|
||
"┡━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━┩\n",
|
||
"│ tokens (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">InputLayer</span>) │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">38</span>) │ <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │ - │\n",
|
||
"├─────────────────────┼───────────────────┼────────────┼───────────────────┤\n",
|
||
"│ embed (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Embedding</span>) │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">38</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">64</span>) │ <span style=\"color: #00af00; text-decoration-color: #00af00\">158,720</span> │ tokens[<span style=\"color: #00af00; text-decoration-color: #00af00\">0</span>][<span style=\"color: #00af00; text-decoration-color: #00af00\">0</span>] │\n",
|
||
"├─────────────────────┼───────────────────┼────────────┼───────────────────┤\n",
|
||
"│ spatial_dropout1d_1 │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">38</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">64</span>) │ <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │ embed[<span style=\"color: #00af00; text-decoration-color: #00af00\">0</span>][<span style=\"color: #00af00; text-decoration-color: #00af00\">0</span>] │\n",
|
||
"│ (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">SpatialDropout1D</span>) │ │ │ │\n",
|
||
"├─────────────────────┼───────────────────┼────────────┼───────────────────┤\n",
|
||
"│ not_equal_2 │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">38</span>) │ <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │ tokens[<span style=\"color: #00af00; text-decoration-color: #00af00\">0</span>][<span style=\"color: #00af00; text-decoration-color: #00af00\">0</span>] │\n",
|
||
"│ (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">NotEqual</span>) │ │ │ │\n",
|
||
"├─────────────────────┼───────────────────┼────────────┼───────────────────┤\n",
|
||
"│ bidirectional_2 │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">38</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">128</span>) │ <span style=\"color: #00af00; text-decoration-color: #00af00\">66,048</span> │ spatial_dropout1… │\n",
|
||
"│ (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Bidirectional</span>) │ │ │ not_equal_2[<span style=\"color: #00af00; text-decoration-color: #00af00\">0</span>][<span style=\"color: #00af00; text-decoration-color: #00af00\">0</span>] │\n",
|
||
"├─────────────────────┼───────────────────┼────────────┼───────────────────┤\n",
|
||
"│ bidirectional_3 │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">38</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">128</span>) │ <span style=\"color: #00af00; text-decoration-color: #00af00\">98,816</span> │ bidirectional_2[<span style=\"color: #00af00; text-decoration-color: #00af00\">…</span> │\n",
|
||
"│ (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Bidirectional</span>) │ │ │ not_equal_2[<span style=\"color: #00af00; text-decoration-color: #00af00\">0</span>][<span style=\"color: #00af00; text-decoration-color: #00af00\">0</span>] │\n",
|
||
"├─────────────────────┼───────────────────┼────────────┼───────────────────┤\n",
|
||
"│ time_distributed_2 │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">38</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">64</span>) │ <span style=\"color: #00af00; text-decoration-color: #00af00\">8,256</span> │ bidirectional_3[<span style=\"color: #00af00; text-decoration-color: #00af00\">…</span> │\n",
|
||
"│ (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">TimeDistributed</span>) │ │ │ not_equal_2[<span style=\"color: #00af00; text-decoration-color: #00af00\">0</span>][<span style=\"color: #00af00; text-decoration-color: #00af00\">0</span>] │\n",
|
||
"├─────────────────────┼───────────────────┼────────────┼───────────────────┤\n",
|
||
"│ time_distributed_3 │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">38</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">64</span>) │ <span style=\"color: #00af00; text-decoration-color: #00af00\">8,256</span> │ bidirectional_3[<span style=\"color: #00af00; text-decoration-color: #00af00\">…</span> │\n",
|
||
"│ (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">TimeDistributed</span>) │ │ │ not_equal_2[<span style=\"color: #00af00; text-decoration-color: #00af00\">0</span>][<span style=\"color: #00af00; text-decoration-color: #00af00\">0</span>] │\n",
|
||
"├─────────────────────┼───────────────────┼────────────┼───────────────────┤\n",
|
||
"│ ner_output │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">38</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">10</span>) │ <span style=\"color: #00af00; text-decoration-color: #00af00\">650</span> │ time_distributed… │\n",
|
||
"│ (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">TimeDistributed</span>) │ │ │ not_equal_2[<span style=\"color: #00af00; text-decoration-color: #00af00\">0</span>][<span style=\"color: #00af00; text-decoration-color: #00af00\">0</span>] │\n",
|
||
"├─────────────────────┼───────────────────┼────────────┼───────────────────┤\n",
|
||
"│ srl_output │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">38</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">14</span>) │ <span style=\"color: #00af00; text-decoration-color: #00af00\">910</span> │ time_distributed… │\n",
|
||
"│ (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">TimeDistributed</span>) │ │ │ not_equal_2[<span style=\"color: #00af00; text-decoration-color: #00af00\">0</span>][<span style=\"color: #00af00; text-decoration-color: #00af00\">0</span>] │\n",
|
||
"└─────────────────────┴───────────────────┴────────────┴───────────────────┘\n",
|
||
"</pre>\n"
|
||
],
|
||
"text/plain": [
|
||
"┏━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━┓\n",
|
||
"┃\u001b[1m \u001b[0m\u001b[1mLayer (type) \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1mOutput Shape \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m Param #\u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1mConnected to \u001b[0m\u001b[1m \u001b[0m┃\n",
|
||
"┡━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━┩\n",
|
||
"│ tokens (\u001b[38;5;33mInputLayer\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m38\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │ - │\n",
|
||
"├─────────────────────┼───────────────────┼────────────┼───────────────────┤\n",
|
||
"│ embed (\u001b[38;5;33mEmbedding\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m38\u001b[0m, \u001b[38;5;34m64\u001b[0m) │ \u001b[38;5;34m158,720\u001b[0m │ tokens[\u001b[38;5;34m0\u001b[0m][\u001b[38;5;34m0\u001b[0m] │\n",
|
||
"├─────────────────────┼───────────────────┼────────────┼───────────────────┤\n",
|
||
"│ spatial_dropout1d_1 │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m38\u001b[0m, \u001b[38;5;34m64\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │ embed[\u001b[38;5;34m0\u001b[0m][\u001b[38;5;34m0\u001b[0m] │\n",
|
||
"│ (\u001b[38;5;33mSpatialDropout1D\u001b[0m) │ │ │ │\n",
|
||
"├─────────────────────┼───────────────────┼────────────┼───────────────────┤\n",
|
||
"│ not_equal_2 │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m38\u001b[0m) │ \u001b[38;5;34m0\u001b[0m │ tokens[\u001b[38;5;34m0\u001b[0m][\u001b[38;5;34m0\u001b[0m] │\n",
|
||
"│ (\u001b[38;5;33mNotEqual\u001b[0m) │ │ │ │\n",
|
||
"├─────────────────────┼───────────────────┼────────────┼───────────────────┤\n",
|
||
"│ bidirectional_2 │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m38\u001b[0m, \u001b[38;5;34m128\u001b[0m) │ \u001b[38;5;34m66,048\u001b[0m │ spatial_dropout1… │\n",
|
||
"│ (\u001b[38;5;33mBidirectional\u001b[0m) │ │ │ not_equal_2[\u001b[38;5;34m0\u001b[0m][\u001b[38;5;34m0\u001b[0m] │\n",
|
||
"├─────────────────────┼───────────────────┼────────────┼───────────────────┤\n",
|
||
"│ bidirectional_3 │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m38\u001b[0m, \u001b[38;5;34m128\u001b[0m) │ \u001b[38;5;34m98,816\u001b[0m │ bidirectional_2[\u001b[38;5;34m…\u001b[0m │\n",
|
||
"│ (\u001b[38;5;33mBidirectional\u001b[0m) │ │ │ not_equal_2[\u001b[38;5;34m0\u001b[0m][\u001b[38;5;34m0\u001b[0m] │\n",
|
||
"├─────────────────────┼───────────────────┼────────────┼───────────────────┤\n",
|
||
"│ time_distributed_2 │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m38\u001b[0m, \u001b[38;5;34m64\u001b[0m) │ \u001b[38;5;34m8,256\u001b[0m │ bidirectional_3[\u001b[38;5;34m…\u001b[0m │\n",
|
||
"│ (\u001b[38;5;33mTimeDistributed\u001b[0m) │ │ │ not_equal_2[\u001b[38;5;34m0\u001b[0m][\u001b[38;5;34m0\u001b[0m] │\n",
|
||
"├─────────────────────┼───────────────────┼────────────┼───────────────────┤\n",
|
||
"│ time_distributed_3 │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m38\u001b[0m, \u001b[38;5;34m64\u001b[0m) │ \u001b[38;5;34m8,256\u001b[0m │ bidirectional_3[\u001b[38;5;34m…\u001b[0m │\n",
|
||
"│ (\u001b[38;5;33mTimeDistributed\u001b[0m) │ │ │ not_equal_2[\u001b[38;5;34m0\u001b[0m][\u001b[38;5;34m0\u001b[0m] │\n",
|
||
"├─────────────────────┼───────────────────┼────────────┼───────────────────┤\n",
|
||
"│ ner_output │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m38\u001b[0m, \u001b[38;5;34m10\u001b[0m) │ \u001b[38;5;34m650\u001b[0m │ time_distributed… │\n",
|
||
"│ (\u001b[38;5;33mTimeDistributed\u001b[0m) │ │ │ not_equal_2[\u001b[38;5;34m0\u001b[0m][\u001b[38;5;34m0\u001b[0m] │\n",
|
||
"├─────────────────────┼───────────────────┼────────────┼───────────────────┤\n",
|
||
"│ srl_output │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m38\u001b[0m, \u001b[38;5;34m14\u001b[0m) │ \u001b[38;5;34m910\u001b[0m │ time_distributed… │\n",
|
||
"│ (\u001b[38;5;33mTimeDistributed\u001b[0m) │ │ │ not_equal_2[\u001b[38;5;34m0\u001b[0m][\u001b[38;5;34m0\u001b[0m] │\n",
|
||
"└─────────────────────┴───────────────────┴────────────┴───────────────────┘\n"
|
||
]
|
||
},
|
||
"metadata": {},
|
||
"output_type": "display_data"
|
||
},
|
||
{
|
||
"data": {
|
||
"text/html": [
|
||
"<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Total params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">341,656</span> (1.30 MB)\n",
|
||
"</pre>\n"
|
||
],
|
||
"text/plain": [
|
||
"\u001b[1m Total params: \u001b[0m\u001b[38;5;34m341,656\u001b[0m (1.30 MB)\n"
|
||
]
|
||
},
|
||
"metadata": {},
|
||
"output_type": "display_data"
|
||
},
|
||
{
|
||
"data": {
|
||
"text/html": [
|
||
"<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">341,656</span> (1.30 MB)\n",
|
||
"</pre>\n"
|
||
],
|
||
"text/plain": [
|
||
"\u001b[1m Trainable params: \u001b[0m\u001b[38;5;34m341,656\u001b[0m (1.30 MB)\n"
|
||
]
|
||
},
|
||
"metadata": {},
|
||
"output_type": "display_data"
|
||
},
|
||
{
|
||
"data": {
|
||
"text/html": [
|
||
"<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Non-trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> (0.00 B)\n",
|
||
"</pre>\n"
|
||
],
|
||
"text/plain": [
|
||
"\u001b[1m Non-trainable params: \u001b[0m\u001b[38;5;34m0\u001b[0m (0.00 B)\n"
|
||
]
|
||
},
|
||
"metadata": {},
|
||
"output_type": "display_data"
|
||
}
|
||
],
|
||
"source": [
|
||
"embed_dim = 64\n",
|
||
"lstm_units = 64\n",
|
||
"drop_embed = 0.2\n",
|
||
"drop_lstm = 0.2\n",
|
||
"\n",
|
||
"inp = Input(shape=(maxlen,), name=\"tokens\")\n",
|
||
"emb = Embedding(len(word2idx), embed_dim, mask_zero=True, name=\"embed\")(inp)\n",
|
||
"emb = SpatialDropout1D(drop_embed)(emb)\n",
|
||
"\n",
|
||
"x = Bidirectional(\n",
|
||
" LSTM(\n",
|
||
" lstm_units,\n",
|
||
" return_sequences=True,\n",
|
||
" dropout=drop_lstm,\n",
|
||
" recurrent_dropout=drop_lstm,\n",
|
||
" )\n",
|
||
")(emb)\n",
|
||
"x = Bidirectional(\n",
|
||
" LSTM(\n",
|
||
" lstm_units,\n",
|
||
" return_sequences=True,\n",
|
||
" dropout=drop_lstm,\n",
|
||
" recurrent_dropout=drop_lstm,\n",
|
||
" )\n",
|
||
")(x)\n",
|
||
"\n",
|
||
"ner_head = TimeDistributed(Dense(lstm_units, activation=\"relu\"))(x)\n",
|
||
"ner_out = TimeDistributed(\n",
|
||
" Dense(len(tag2idx_ner), activation=\"softmax\"), name=\"ner_output\"\n",
|
||
")(ner_head)\n",
|
||
"\n",
|
||
"srl_head = TimeDistributed(Dense(lstm_units, activation=\"relu\"))(x)\n",
|
||
"srl_out = TimeDistributed(\n",
|
||
" Dense(len(tag2idx_srl), activation=\"softmax\"), name=\"srl_output\"\n",
|
||
")(srl_head)\n",
|
||
"\n",
|
||
"model = Model(inp, [ner_out, srl_out])\n",
|
||
"\n",
|
||
"model.compile(\n",
|
||
" optimizer=tf.keras.optimizers.Adam(3e-4),\n",
|
||
" loss={\n",
|
||
" \"ner_output\": \"sparse_categorical_crossentropy\",\n",
|
||
" \"srl_output\": \"sparse_categorical_crossentropy\",\n",
|
||
" },\n",
|
||
" metrics={\n",
|
||
" \"ner_output\": [\"sparse_categorical_accuracy\"],\n",
|
||
" \"srl_output\": [\"sparse_categorical_accuracy\"],\n",
|
||
" },\n",
|
||
")\n",
|
||
"model.summary()"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 19,
|
||
"id": "f41d6012",
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"Epoch 1/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m10s\u001b[0m 92ms/step - loss: 4.9108 - ner_output_loss: 2.2818 - ner_output_sparse_categorical_accuracy: 0.2052 - srl_output_loss: 2.6290 - srl_output_sparse_categorical_accuracy: 0.1547 - val_loss: 4.6812 - val_ner_output_loss: 2.1289 - val_ner_output_sparse_categorical_accuracy: 0.1389 - val_srl_output_loss: 2.5523 - val_srl_output_sparse_categorical_accuracy: 0.0668 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 2/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 46ms/step - loss: 4.3571 - ner_output_loss: 1.9204 - ner_output_sparse_categorical_accuracy: 0.1344 - srl_output_loss: 2.4365 - srl_output_sparse_categorical_accuracy: 0.0673 - val_loss: 3.5487 - val_ner_output_loss: 1.5474 - val_ner_output_sparse_categorical_accuracy: 0.1389 - val_srl_output_loss: 2.0010 - val_srl_output_sparse_categorical_accuracy: 0.0905 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 3/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 45ms/step - loss: 3.3666 - ner_output_loss: 1.4404 - ner_output_sparse_categorical_accuracy: 0.1317 - srl_output_loss: 1.9260 - srl_output_sparse_categorical_accuracy: 0.0868 - val_loss: 2.9809 - val_ner_output_loss: 1.3097 - val_ner_output_sparse_categorical_accuracy: 0.1389 - val_srl_output_loss: 1.6709 - val_srl_output_sparse_categorical_accuracy: 0.0932 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 4/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 46ms/step - loss: 2.8620 - ner_output_loss: 1.2367 - ner_output_sparse_categorical_accuracy: 0.1335 - srl_output_loss: 1.6253 - srl_output_sparse_categorical_accuracy: 0.0978 - val_loss: 2.7502 - val_ner_output_loss: 1.2033 - val_ner_output_sparse_categorical_accuracy: 0.1375 - val_srl_output_loss: 1.5466 - val_srl_output_sparse_categorical_accuracy: 0.1118 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 5/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 45ms/step - loss: 2.6732 - ner_output_loss: 1.1535 - ner_output_sparse_categorical_accuracy: 0.1368 - srl_output_loss: 1.5197 - srl_output_sparse_categorical_accuracy: 0.1100 - val_loss: 2.5802 - val_ner_output_loss: 1.1193 - val_ner_output_sparse_categorical_accuracy: 0.1478 - val_srl_output_loss: 1.4604 - val_srl_output_sparse_categorical_accuracy: 0.1106 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 6/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 45ms/step - loss: 2.5011 - ner_output_loss: 1.0769 - ner_output_sparse_categorical_accuracy: 0.1441 - srl_output_loss: 1.4243 - srl_output_sparse_categorical_accuracy: 0.1083 - val_loss: 2.4460 - val_ner_output_loss: 1.0604 - val_ner_output_sparse_categorical_accuracy: 0.1491 - val_srl_output_loss: 1.3852 - val_srl_output_sparse_categorical_accuracy: 0.1230 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 7/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 46ms/step - loss: 2.4115 - ner_output_loss: 1.0338 - ner_output_sparse_categorical_accuracy: 0.1498 - srl_output_loss: 1.3776 - srl_output_sparse_categorical_accuracy: 0.1191 - val_loss: 2.2803 - val_ner_output_loss: 0.9900 - val_ner_output_sparse_categorical_accuracy: 0.1625 - val_srl_output_loss: 1.2898 - val_srl_output_sparse_categorical_accuracy: 0.1340 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 8/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 46ms/step - loss: 2.1656 - ner_output_loss: 0.9351 - ner_output_sparse_categorical_accuracy: 0.1573 - srl_output_loss: 1.2305 - srl_output_sparse_categorical_accuracy: 0.1324 - val_loss: 2.0635 - val_ner_output_loss: 0.8914 - val_ner_output_sparse_categorical_accuracy: 0.1697 - val_srl_output_loss: 1.1716 - val_srl_output_sparse_categorical_accuracy: 0.1417 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 9/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 46ms/step - loss: 1.9027 - ner_output_loss: 0.8172 - ner_output_sparse_categorical_accuracy: 0.1659 - srl_output_loss: 1.0854 - srl_output_sparse_categorical_accuracy: 0.1431 - val_loss: 1.8595 - val_ner_output_loss: 0.7923 - val_ner_output_sparse_categorical_accuracy: 0.1767 - val_srl_output_loss: 1.0667 - val_srl_output_sparse_categorical_accuracy: 0.1528 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 10/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 45ms/step - loss: 1.7463 - ner_output_loss: 0.7403 - ner_output_sparse_categorical_accuracy: 0.1740 - srl_output_loss: 1.0059 - srl_output_sparse_categorical_accuracy: 0.1508 - val_loss: 1.7059 - val_ner_output_loss: 0.7108 - val_ner_output_sparse_categorical_accuracy: 0.1833 - val_srl_output_loss: 0.9947 - val_srl_output_sparse_categorical_accuracy: 0.1601 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 11/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 49ms/step - loss: 1.6282 - ner_output_loss: 0.6822 - ner_output_sparse_categorical_accuracy: 0.1762 - srl_output_loss: 0.9460 - srl_output_sparse_categorical_accuracy: 0.1559 - val_loss: 1.6089 - val_ner_output_loss: 0.6610 - val_ner_output_sparse_categorical_accuracy: 0.1861 - val_srl_output_loss: 0.9475 - val_srl_output_sparse_categorical_accuracy: 0.1647 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 12/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 46ms/step - loss: 1.4778 - ner_output_loss: 0.6039 - ner_output_sparse_categorical_accuracy: 0.1795 - srl_output_loss: 0.8738 - srl_output_sparse_categorical_accuracy: 0.1625 - val_loss: 1.5403 - val_ner_output_loss: 0.6256 - val_ner_output_sparse_categorical_accuracy: 0.1874 - val_srl_output_loss: 0.9142 - val_srl_output_sparse_categorical_accuracy: 0.1668 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 13/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 45ms/step - loss: 1.3586 - ner_output_loss: 0.5400 - ner_output_sparse_categorical_accuracy: 0.1808 - srl_output_loss: 0.8186 - srl_output_sparse_categorical_accuracy: 0.1644 - val_loss: 1.4752 - val_ner_output_loss: 0.5969 - val_ner_output_sparse_categorical_accuracy: 0.1884 - val_srl_output_loss: 0.8780 - val_srl_output_sparse_categorical_accuracy: 0.1687 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 14/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 47ms/step - loss: 1.3344 - ner_output_loss: 0.5186 - ner_output_sparse_categorical_accuracy: 0.1869 - srl_output_loss: 0.8159 - srl_output_sparse_categorical_accuracy: 0.1682 - val_loss: 1.4290 - val_ner_output_loss: 0.5756 - val_ner_output_sparse_categorical_accuracy: 0.1893 - val_srl_output_loss: 0.8531 - val_srl_output_sparse_categorical_accuracy: 0.1702 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 15/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 50ms/step - loss: 1.2810 - ner_output_loss: 0.4851 - ner_output_sparse_categorical_accuracy: 0.1849 - srl_output_loss: 0.7959 - srl_output_sparse_categorical_accuracy: 0.1654 - val_loss: 1.3892 - val_ner_output_loss: 0.5600 - val_ner_output_sparse_categorical_accuracy: 0.1917 - val_srl_output_loss: 0.8289 - val_srl_output_sparse_categorical_accuracy: 0.1711 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 16/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 46ms/step - loss: 1.2174 - ner_output_loss: 0.4766 - ner_output_sparse_categorical_accuracy: 0.1882 - srl_output_loss: 0.7407 - srl_output_sparse_categorical_accuracy: 0.1719 - val_loss: 1.3588 - val_ner_output_loss: 0.5454 - val_ner_output_sparse_categorical_accuracy: 0.1937 - val_srl_output_loss: 0.8130 - val_srl_output_sparse_categorical_accuracy: 0.1728 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 17/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 46ms/step - loss: 1.1637 - ner_output_loss: 0.4539 - ner_output_sparse_categorical_accuracy: 0.1882 - srl_output_loss: 0.7099 - srl_output_sparse_categorical_accuracy: 0.1719 - val_loss: 1.3189 - val_ner_output_loss: 0.5257 - val_ner_output_sparse_categorical_accuracy: 0.1943 - val_srl_output_loss: 0.7928 - val_srl_output_sparse_categorical_accuracy: 0.1740 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 18/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 47ms/step - loss: 1.0587 - ner_output_loss: 0.4037 - ner_output_sparse_categorical_accuracy: 0.1884 - srl_output_loss: 0.6550 - srl_output_sparse_categorical_accuracy: 0.1721 - val_loss: 1.2925 - val_ner_output_loss: 0.5146 - val_ner_output_sparse_categorical_accuracy: 0.1959 - val_srl_output_loss: 0.7776 - val_srl_output_sparse_categorical_accuracy: 0.1750 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 19/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 47ms/step - loss: 1.0582 - ner_output_loss: 0.3877 - ner_output_sparse_categorical_accuracy: 0.1900 - srl_output_loss: 0.6705 - srl_output_sparse_categorical_accuracy: 0.1720 - val_loss: 1.2710 - val_ner_output_loss: 0.5035 - val_ner_output_sparse_categorical_accuracy: 0.1973 - val_srl_output_loss: 0.7672 - val_srl_output_sparse_categorical_accuracy: 0.1756 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 20/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 46ms/step - loss: 1.0201 - ner_output_loss: 0.3707 - ner_output_sparse_categorical_accuracy: 0.1972 - srl_output_loss: 0.6494 - srl_output_sparse_categorical_accuracy: 0.1782 - val_loss: 1.2472 - val_ner_output_loss: 0.4926 - val_ner_output_sparse_categorical_accuracy: 0.1975 - val_srl_output_loss: 0.7542 - val_srl_output_sparse_categorical_accuracy: 0.1755 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 21/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 47ms/step - loss: 0.9662 - ner_output_loss: 0.3554 - ner_output_sparse_categorical_accuracy: 0.1917 - srl_output_loss: 0.6109 - srl_output_sparse_categorical_accuracy: 0.1750 - val_loss: 1.2226 - val_ner_output_loss: 0.4826 - val_ner_output_sparse_categorical_accuracy: 0.1975 - val_srl_output_loss: 0.7397 - val_srl_output_sparse_categorical_accuracy: 0.1759 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 22/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 49ms/step - loss: 0.9947 - ner_output_loss: 0.3667 - ner_output_sparse_categorical_accuracy: 0.1983 - srl_output_loss: 0.6281 - srl_output_sparse_categorical_accuracy: 0.1813 - val_loss: 1.2104 - val_ner_output_loss: 0.4795 - val_ner_output_sparse_categorical_accuracy: 0.1978 - val_srl_output_loss: 0.7305 - val_srl_output_sparse_categorical_accuracy: 0.1764 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 23/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 45ms/step - loss: 0.9693 - ner_output_loss: 0.3491 - ner_output_sparse_categorical_accuracy: 0.1953 - srl_output_loss: 0.6202 - srl_output_sparse_categorical_accuracy: 0.1768 - val_loss: 1.1952 - val_ner_output_loss: 0.4718 - val_ner_output_sparse_categorical_accuracy: 0.1984 - val_srl_output_loss: 0.7231 - val_srl_output_sparse_categorical_accuracy: 0.1771 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 24/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 50ms/step - loss: 0.9162 - ner_output_loss: 0.3255 - ner_output_sparse_categorical_accuracy: 0.1984 - srl_output_loss: 0.5907 - srl_output_sparse_categorical_accuracy: 0.1805 - val_loss: 1.1879 - val_ner_output_loss: 0.4724 - val_ner_output_sparse_categorical_accuracy: 0.1991 - val_srl_output_loss: 0.7151 - val_srl_output_sparse_categorical_accuracy: 0.1784 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 25/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 46ms/step - loss: 0.8918 - ner_output_loss: 0.3157 - ner_output_sparse_categorical_accuracy: 0.1968 - srl_output_loss: 0.5762 - srl_output_sparse_categorical_accuracy: 0.1789 - val_loss: 1.1675 - val_ner_output_loss: 0.4613 - val_ner_output_sparse_categorical_accuracy: 0.1992 - val_srl_output_loss: 0.7058 - val_srl_output_sparse_categorical_accuracy: 0.1789 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 26/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 46ms/step - loss: 0.8769 - ner_output_loss: 0.3010 - ner_output_sparse_categorical_accuracy: 0.1997 - srl_output_loss: 0.5759 - srl_output_sparse_categorical_accuracy: 0.1801 - val_loss: 1.1435 - val_ner_output_loss: 0.4481 - val_ner_output_sparse_categorical_accuracy: 0.1995 - val_srl_output_loss: 0.6950 - val_srl_output_sparse_categorical_accuracy: 0.1792 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 27/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 48ms/step - loss: 0.8090 - ner_output_loss: 0.2721 - ner_output_sparse_categorical_accuracy: 0.2008 - srl_output_loss: 0.5369 - srl_output_sparse_categorical_accuracy: 0.1816 - val_loss: 1.1363 - val_ner_output_loss: 0.4474 - val_ner_output_sparse_categorical_accuracy: 0.1998 - val_srl_output_loss: 0.6884 - val_srl_output_sparse_categorical_accuracy: 0.1801 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 28/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 47ms/step - loss: 0.8319 - ner_output_loss: 0.2812 - ner_output_sparse_categorical_accuracy: 0.2008 - srl_output_loss: 0.5507 - srl_output_sparse_categorical_accuracy: 0.1800 - val_loss: 1.1294 - val_ner_output_loss: 0.4376 - val_ner_output_sparse_categorical_accuracy: 0.2004 - val_srl_output_loss: 0.6914 - val_srl_output_sparse_categorical_accuracy: 0.1800 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 29/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 46ms/step - loss: 0.8019 - ner_output_loss: 0.2668 - ner_output_sparse_categorical_accuracy: 0.1997 - srl_output_loss: 0.5351 - srl_output_sparse_categorical_accuracy: 0.1802 - val_loss: 1.1265 - val_ner_output_loss: 0.4411 - val_ner_output_sparse_categorical_accuracy: 0.2008 - val_srl_output_loss: 0.6850 - val_srl_output_sparse_categorical_accuracy: 0.1803 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 30/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 46ms/step - loss: 0.7622 - ner_output_loss: 0.2489 - ner_output_sparse_categorical_accuracy: 0.2034 - srl_output_loss: 0.5133 - srl_output_sparse_categorical_accuracy: 0.1847 - val_loss: 1.1217 - val_ner_output_loss: 0.4410 - val_ner_output_sparse_categorical_accuracy: 0.2005 - val_srl_output_loss: 0.6803 - val_srl_output_sparse_categorical_accuracy: 0.1806 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 31/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 47ms/step - loss: 0.7801 - ner_output_loss: 0.2485 - ner_output_sparse_categorical_accuracy: 0.2046 - srl_output_loss: 0.5316 - srl_output_sparse_categorical_accuracy: 0.1826 - val_loss: 1.1110 - val_ner_output_loss: 0.4355 - val_ner_output_sparse_categorical_accuracy: 0.2002 - val_srl_output_loss: 0.6751 - val_srl_output_sparse_categorical_accuracy: 0.1818 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 32/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 46ms/step - loss: 0.7747 - ner_output_loss: 0.2493 - ner_output_sparse_categorical_accuracy: 0.2018 - srl_output_loss: 0.5253 - srl_output_sparse_categorical_accuracy: 0.1832 - val_loss: 1.1026 - val_ner_output_loss: 0.4279 - val_ner_output_sparse_categorical_accuracy: 0.2009 - val_srl_output_loss: 0.6742 - val_srl_output_sparse_categorical_accuracy: 0.1819 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 33/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 46ms/step - loss: 0.6970 - ner_output_loss: 0.2145 - ner_output_sparse_categorical_accuracy: 0.2029 - srl_output_loss: 0.4825 - srl_output_sparse_categorical_accuracy: 0.1829 - val_loss: 1.0892 - val_ner_output_loss: 0.4261 - val_ner_output_sparse_categorical_accuracy: 0.2018 - val_srl_output_loss: 0.6627 - val_srl_output_sparse_categorical_accuracy: 0.1825 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 34/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 46ms/step - loss: 0.7070 - ner_output_loss: 0.2224 - ner_output_sparse_categorical_accuracy: 0.2055 - srl_output_loss: 0.4846 - srl_output_sparse_categorical_accuracy: 0.1855 - val_loss: 1.0870 - val_ner_output_loss: 0.4197 - val_ner_output_sparse_categorical_accuracy: 0.2022 - val_srl_output_loss: 0.6669 - val_srl_output_sparse_categorical_accuracy: 0.1834 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 35/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 46ms/step - loss: 0.7292 - ner_output_loss: 0.2378 - ner_output_sparse_categorical_accuracy: 0.2068 - srl_output_loss: 0.4914 - srl_output_sparse_categorical_accuracy: 0.1879 - val_loss: 1.0733 - val_ner_output_loss: 0.4164 - val_ner_output_sparse_categorical_accuracy: 0.2020 - val_srl_output_loss: 0.6564 - val_srl_output_sparse_categorical_accuracy: 0.1837 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 36/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 47ms/step - loss: 0.6923 - ner_output_loss: 0.2130 - ner_output_sparse_categorical_accuracy: 0.2064 - srl_output_loss: 0.4793 - srl_output_sparse_categorical_accuracy: 0.1865 - val_loss: 1.0673 - val_ner_output_loss: 0.4181 - val_ner_output_sparse_categorical_accuracy: 0.2024 - val_srl_output_loss: 0.6487 - val_srl_output_sparse_categorical_accuracy: 0.1840 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 37/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 46ms/step - loss: 0.6649 - ner_output_loss: 0.2093 - ner_output_sparse_categorical_accuracy: 0.2042 - srl_output_loss: 0.4557 - srl_output_sparse_categorical_accuracy: 0.1867 - val_loss: 1.0735 - val_ner_output_loss: 0.4241 - val_ner_output_sparse_categorical_accuracy: 0.2030 - val_srl_output_loss: 0.6489 - val_srl_output_sparse_categorical_accuracy: 0.1836 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 38/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 47ms/step - loss: 0.6624 - ner_output_loss: 0.2135 - ner_output_sparse_categorical_accuracy: 0.2046 - srl_output_loss: 0.4488 - srl_output_sparse_categorical_accuracy: 0.1872 - val_loss: 1.0567 - val_ner_output_loss: 0.4134 - val_ner_output_sparse_categorical_accuracy: 0.2026 - val_srl_output_loss: 0.6428 - val_srl_output_sparse_categorical_accuracy: 0.1847 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 39/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 47ms/step - loss: 0.6242 - ner_output_loss: 0.1945 - ner_output_sparse_categorical_accuracy: 0.2063 - srl_output_loss: 0.4297 - srl_output_sparse_categorical_accuracy: 0.1893 - val_loss: 1.0591 - val_ner_output_loss: 0.4165 - val_ner_output_sparse_categorical_accuracy: 0.2028 - val_srl_output_loss: 0.6422 - val_srl_output_sparse_categorical_accuracy: 0.1847 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 40/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 49ms/step - loss: 0.6077 - ner_output_loss: 0.1934 - ner_output_sparse_categorical_accuracy: 0.2081 - srl_output_loss: 0.4143 - srl_output_sparse_categorical_accuracy: 0.1924 - val_loss: 1.0560 - val_ner_output_loss: 0.4162 - val_ner_output_sparse_categorical_accuracy: 0.2028 - val_srl_output_loss: 0.6394 - val_srl_output_sparse_categorical_accuracy: 0.1849 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 41/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 49ms/step - loss: 0.6324 - ner_output_loss: 0.1931 - ner_output_sparse_categorical_accuracy: 0.2076 - srl_output_loss: 0.4392 - srl_output_sparse_categorical_accuracy: 0.1898 - val_loss: 1.0587 - val_ner_output_loss: 0.4232 - val_ner_output_sparse_categorical_accuracy: 0.2037 - val_srl_output_loss: 0.6349 - val_srl_output_sparse_categorical_accuracy: 0.1863 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 42/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 49ms/step - loss: 0.5964 - ner_output_loss: 0.1760 - ner_output_sparse_categorical_accuracy: 0.2003 - srl_output_loss: 0.4204 - srl_output_sparse_categorical_accuracy: 0.1841 - val_loss: 1.0573 - val_ner_output_loss: 0.4168 - val_ner_output_sparse_categorical_accuracy: 0.2033 - val_srl_output_loss: 0.6400 - val_srl_output_sparse_categorical_accuracy: 0.1855 - learning_rate: 3.0000e-04\n",
|
||
"Epoch 43/50\n",
|
||
"\u001b[1m28/28\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 48ms/step - loss: 0.6176 - ner_output_loss: 0.1845 - ner_output_sparse_categorical_accuracy: 0.2140 - srl_output_loss: 0.4331 - srl_output_sparse_categorical_accuracy: 0.1949 - val_loss: 1.0580 - val_ner_output_loss: 0.4177 - val_ner_output_sparse_categorical_accuracy: 0.2037 - val_srl_output_loss: 0.6397 - val_srl_output_sparse_categorical_accuracy: 0.1866 - learning_rate: 1.5000e-04\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"callbacks = [\n",
|
||
" tf.keras.callbacks.EarlyStopping(patience=3, restore_best_weights=True),\n",
|
||
" tf.keras.callbacks.ReduceLROnPlateau(patience=2, factor=0.5, min_lr=1e-5),\n",
|
||
"]\n",
|
||
"\n",
|
||
"history = model.fit(\n",
|
||
" X_tr,\n",
|
||
" [ner_tr, srl_tr], # y → LIST (pos 0 = ner_output, 1 = srl_output)\n",
|
||
" sample_weight=[m_tr, m_tr], # sama‑persis urutan\n",
|
||
" validation_data=(X_te, [ner_te, srl_te], [m_te, m_te]),\n",
|
||
" \n",
|
||
" batch_size=64,\n",
|
||
" epochs=50,\n",
|
||
" callbacks=callbacks,\n",
|
||
" verbose=1,\n",
|
||
")\n",
|
||
"\n",
|
||
"\n",
|
||
"# =========================\n",
|
||
"# 7. Save artefacts\n",
|
||
"# =========================\n",
|
||
"model.save(\"lstm_ner_srl_model.keras\")\n",
|
||
"for fname, obj in [\n",
|
||
" (\"word2idx.pkl\", word2idx),\n",
|
||
" (\"tag2idx_ner.pkl\", tag2idx_ner),\n",
|
||
" (\"tag2idx_srl.pkl\", tag2idx_srl),\n",
|
||
"]:\n",
|
||
" with open(fname, \"wb\") as f:\n",
|
||
" pickle.dump(obj, f)"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 20,
|
||
"id": "430794b9",
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"image/png": "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",
|
||
"text/plain": [
|
||
"<Figure size 1000x600 with 1 Axes>"
|
||
]
|
||
},
|
||
"metadata": {},
|
||
"output_type": "display_data"
|
||
},
|
||
{
|
||
"data": {
|
||
"image/png": "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",
|
||
"text/plain": [
|
||
"<Figure size 1000x600 with 1 Axes>"
|
||
]
|
||
},
|
||
"metadata": {},
|
||
"output_type": "display_data"
|
||
}
|
||
],
|
||
"source": [
|
||
"import matplotlib.pyplot as plt\n",
|
||
"\n",
|
||
"def plot_training_history(history):\n",
|
||
" history_data = history.history\n",
|
||
" epochs = range(1, len(next(iter(history_data.values()))) + 1)\n",
|
||
"\n",
|
||
" # Coba deteksi metric secara dinamis agar fleksibel\n",
|
||
" ner_acc_key = next((k for k in history_data if \"ner_output\" in k and \"accuracy\" in k), None)\n",
|
||
" srl_acc_key = next((k for k in history_data if \"srl_output\" in k and \"accuracy\" in k), None)\n",
|
||
" val_ner_acc_key = f\"val_{ner_acc_key}\" if ner_acc_key else None\n",
|
||
" val_srl_acc_key = f\"val_{srl_acc_key}\" if srl_acc_key else None\n",
|
||
"\n",
|
||
" # --- Plot Accuracy ---\n",
|
||
" plt.figure(figsize=(10, 6))\n",
|
||
" if ner_acc_key:\n",
|
||
" plt.plot(epochs, history_data[ner_acc_key], marker=\"o\", label=\"NER Accuracy (Train)\")\n",
|
||
" if srl_acc_key:\n",
|
||
" plt.plot(epochs, history_data[srl_acc_key], marker=\"s\", label=\"SRL Accuracy (Train)\")\n",
|
||
" if val_ner_acc_key in history_data:\n",
|
||
" plt.plot(epochs, history_data[val_ner_acc_key], marker=\"o\", linestyle=\"--\", label=\"NER Accuracy (Val)\")\n",
|
||
" if val_srl_acc_key in history_data:\n",
|
||
" plt.plot(epochs, history_data[val_srl_acc_key], marker=\"s\", linestyle=\"--\", label=\"SRL Accuracy (Val)\")\n",
|
||
"\n",
|
||
" plt.title(\"Accuracy per Epoch\")\n",
|
||
" plt.xlabel(\"Epochs\")\n",
|
||
" plt.ylabel(\"Accuracy\")\n",
|
||
" plt.legend()\n",
|
||
" plt.grid(True)\n",
|
||
" plt.savefig(\"accuracy_plot.png\")\n",
|
||
" plt.show()\n",
|
||
"\n",
|
||
" # --- Plot Loss ---\n",
|
||
" plt.figure(figsize=(10, 6))\n",
|
||
" if \"ner_output_loss\" in history_data:\n",
|
||
" plt.plot(epochs, history_data[\"ner_output_loss\"], marker=\"o\", label=\"NER Loss (Train)\")\n",
|
||
" if \"srl_output_loss\" in history_data:\n",
|
||
" plt.plot(epochs, history_data[\"srl_output_loss\"], marker=\"s\", label=\"SRL Loss (Train)\")\n",
|
||
" if \"val_ner_output_loss\" in history_data:\n",
|
||
" plt.plot(epochs, history_data[\"val_ner_output_loss\"], marker=\"o\", linestyle=\"--\", label=\"NER Loss (Val)\")\n",
|
||
" if \"val_srl_output_loss\" in history_data:\n",
|
||
" plt.plot(epochs, history_data[\"val_srl_output_loss\"], marker=\"s\", linestyle=\"--\", label=\"SRL Loss (Val)\")\n",
|
||
"\n",
|
||
" plt.title(\"Loss per Epoch\")\n",
|
||
" plt.xlabel(\"Epochs\")\n",
|
||
" plt.ylabel(\"Loss\")\n",
|
||
" plt.legend()\n",
|
||
" plt.grid(True)\n",
|
||
" plt.savefig(\"loss_plot.png\")\n",
|
||
" plt.show()\n",
|
||
"\n",
|
||
"\n",
|
||
"plot_training_history(history)"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 21,
|
||
"id": "e690a0e0",
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"image/png": "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",
|
||
"text/plain": [
|
||
"<Figure size 800x800 with 1 Axes>"
|
||
]
|
||
},
|
||
"metadata": {},
|
||
"output_type": "display_data"
|
||
},
|
||
{
|
||
"data": {
|
||
"image/png": "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",
|
||
"text/plain": [
|
||
"<Figure size 800x800 with 1 Axes>"
|
||
]
|
||
},
|
||
"metadata": {},
|
||
"output_type": "display_data"
|
||
},
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"\n",
|
||
"==== Confusion Matrix Manual untuk Kelas: DATE ====\n",
|
||
" Prediksi Positif Prediksi Negatif\n",
|
||
"Aktual Positif 309 (TP) 27 (FN)\n",
|
||
"Aktual Negatif 5 (FP) 3567 (TN)\n",
|
||
"\n",
|
||
"==== Confusion Matrix Manual untuk Kelas: ETH ====\n",
|
||
" Prediksi Positif Prediksi Negatif\n",
|
||
"Aktual Positif 88 (TP) 5 (FN)\n",
|
||
"Aktual Negatif 3 (FP) 3812 (TN)\n",
|
||
"\n",
|
||
"==== Confusion Matrix Manual untuk Kelas: EVENT ====\n",
|
||
" Prediksi Positif Prediksi Negatif\n",
|
||
"Aktual Positif 16 (TP) 45 (FN)\n",
|
||
"Aktual Negatif 19 (FP) 3828 (TN)\n",
|
||
"\n",
|
||
"==== Confusion Matrix Manual untuk Kelas: LOC ====\n",
|
||
" Prediksi Positif Prediksi Negatif\n",
|
||
"Aktual Positif 334 (TP) 66 (FN)\n",
|
||
"Aktual Negatif 47 (FP) 3461 (TN)\n",
|
||
"\n",
|
||
"==== Confusion Matrix Manual untuk Kelas: MISC ====\n",
|
||
" Prediksi Positif Prediksi Negatif\n",
|
||
"Aktual Positif 19 (TP) 85 (FN)\n",
|
||
"Aktual Negatif 55 (FP) 3749 (TN)\n",
|
||
"\n",
|
||
"==== Confusion Matrix Manual untuk Kelas: O ====\n",
|
||
" Prediksi Positif Prediksi Negatif\n",
|
||
"Aktual Positif 2266 (TP) 94 (FN)\n",
|
||
"Aktual Negatif 255 (FP) 1293 (TN)\n",
|
||
"\n",
|
||
"==== Confusion Matrix Manual untuk Kelas: ORG ====\n",
|
||
" Prediksi Positif Prediksi Negatif\n",
|
||
"Aktual Positif 16 (TP) 59 (FN)\n",
|
||
"Aktual Negatif 43 (FP) 3790 (TN)\n",
|
||
"\n",
|
||
"==== Confusion Matrix Manual untuk Kelas: PER ====\n",
|
||
" Prediksi Positif Prediksi Negatif\n",
|
||
"Aktual Positif 353 (TP) 51 (FN)\n",
|
||
"Aktual Negatif 21 (FP) 3483 (TN)\n",
|
||
"\n",
|
||
"==== Confusion Matrix Manual untuk Kelas: TIME ====\n",
|
||
" Prediksi Positif Prediksi Negatif\n",
|
||
"Aktual Positif 44 (TP) 31 (FN)\n",
|
||
"Aktual Negatif 15 (FP) 3818 (TN)\n",
|
||
"\n",
|
||
"==== Confusion Matrix Manual untuk Kelas: ARG0 ====\n",
|
||
" Prediksi Positif Prediksi Negatif\n",
|
||
"Aktual Positif 446 (TP) 106 (FN)\n",
|
||
"Aktual Negatif 58 (FP) 3298 (TN)\n",
|
||
"\n",
|
||
"==== Confusion Matrix Manual untuk Kelas: ARG1 ====\n",
|
||
" Prediksi Positif Prediksi Negatif\n",
|
||
"Aktual Positif 542 (TP) 95 (FN)\n",
|
||
"Aktual Negatif 365 (FP) 2906 (TN)\n",
|
||
"\n",
|
||
"==== Confusion Matrix Manual untuk Kelas: ARG2 ====\n",
|
||
" Prediksi Positif Prediksi Negatif\n",
|
||
"Aktual Positif 22 (TP) 74 (FN)\n",
|
||
"Aktual Negatif 34 (FP) 3778 (TN)\n",
|
||
"\n",
|
||
"==== Confusion Matrix Manual untuk Kelas: ARG3 ====\n",
|
||
" Prediksi Positif Prediksi Negatif\n",
|
||
"Aktual Positif 0 (TP) 4 (FN)\n",
|
||
"Aktual Negatif 0 (FP) 3904 (TN)\n",
|
||
"\n",
|
||
"==== Confusion Matrix Manual untuk Kelas: ARGM-CAU ====\n",
|
||
" Prediksi Positif Prediksi Negatif\n",
|
||
"Aktual Positif 0 (TP) 32 (FN)\n",
|
||
"Aktual Negatif 0 (FP) 3876 (TN)\n",
|
||
"\n",
|
||
"==== Confusion Matrix Manual untuk Kelas: ARGM-DIR ====\n",
|
||
" Prediksi Positif Prediksi Negatif\n",
|
||
"Aktual Positif 0 (TP) 5 (FN)\n",
|
||
"Aktual Negatif 0 (FP) 3903 (TN)\n",
|
||
"\n",
|
||
"==== Confusion Matrix Manual untuk Kelas: ARGM-LOC ====\n",
|
||
" Prediksi Positif Prediksi Negatif\n",
|
||
"Aktual Positif 321 (TP) 88 (FN)\n",
|
||
"Aktual Negatif 40 (FP) 3459 (TN)\n",
|
||
"\n",
|
||
"==== Confusion Matrix Manual untuk Kelas: ARGM-MNR ====\n",
|
||
" Prediksi Positif Prediksi Negatif\n",
|
||
"Aktual Positif 0 (TP) 55 (FN)\n",
|
||
"Aktual Negatif 0 (FP) 3853 (TN)\n",
|
||
"\n",
|
||
"==== Confusion Matrix Manual untuk Kelas: ARGM-MOD ====\n",
|
||
" Prediksi Positif Prediksi Negatif\n",
|
||
"Aktual Positif 0 (TP) 18 (FN)\n",
|
||
"Aktual Negatif 0 (FP) 3890 (TN)\n",
|
||
"\n",
|
||
"==== Confusion Matrix Manual untuk Kelas: ARGM-NEG ====\n",
|
||
" Prediksi Positif Prediksi Negatif\n",
|
||
"Aktual Positif 0 (TP) 1 (FN)\n",
|
||
"Aktual Negatif 0 (FP) 3907 (TN)\n",
|
||
"\n",
|
||
"==== Confusion Matrix Manual untuk Kelas: ARGM-TMP ====\n",
|
||
" Prediksi Positif Prediksi Negatif\n",
|
||
"Aktual Positif 391 (TP) 77 (FN)\n",
|
||
"Aktual Negatif 26 (FP) 3414 (TN)\n",
|
||
"\n",
|
||
"==== Confusion Matrix Manual untuk Kelas: O ====\n",
|
||
" Prediksi Positif Prediksi Negatif\n",
|
||
"Aktual Positif 1033 (TP) 102 (FN)\n",
|
||
"Aktual Negatif 181 (FP) 2592 (TN)\n",
|
||
"\n",
|
||
"==== Confusion Matrix Manual untuk Kelas: V ====\n",
|
||
" Prediksi Positif Prediksi Negatif\n",
|
||
"Aktual Positif 386 (TP) 110 (FN)\n",
|
||
"Aktual Negatif 63 (FP) 3349 (TN)\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"import matplotlib.pyplot as plt\n",
|
||
"import numpy as np\n",
|
||
"import pandas as pd\n",
|
||
"from sklearn.metrics import confusion_matrix, ConfusionMatrixDisplay\n",
|
||
"\n",
|
||
"# ------------------------------------------------------------------\n",
|
||
"# 1. Prediksi\n",
|
||
"# ------------------------------------------------------------------\n",
|
||
"\n",
|
||
"model = load_model(\"lstm_ner_srl_model.keras\")\n",
|
||
"pred_ner_prob, pred_srl_prob = model.predict(X_te, verbose=0)\n",
|
||
"\n",
|
||
"pred_ner = pred_ner_prob.argmax(-1)\n",
|
||
"pred_srl = pred_srl_prob.argmax(-1)\n",
|
||
"\n",
|
||
"# ------------------------------------------------------------------\n",
|
||
"# 2. Siapkan masker PAD\n",
|
||
"# ------------------------------------------------------------------\n",
|
||
"pad_id = tag2idx_ner[\"<PAD>\"]\n",
|
||
"\n",
|
||
"mask_ner = ner_te != pad_id\n",
|
||
"mask_srl = srl_te != pad_id\n",
|
||
"\n",
|
||
"true_ner_flat = ner_te[mask_ner]\n",
|
||
"pred_ner_flat = pred_ner[mask_ner]\n",
|
||
"\n",
|
||
"true_srl_flat = srl_te[mask_srl]\n",
|
||
"pred_srl_flat = pred_srl[mask_srl]\n",
|
||
"\n",
|
||
"# ------------------------------------------------------------------\n",
|
||
"# 3. Hitung confusion matrix TANPA PAD\n",
|
||
"# ------------------------------------------------------------------\n",
|
||
"labels_ner_no_pad = [i for i in range(len(tag2idx_ner)) if i != pad_id]\n",
|
||
"labels_srl_no_pad = [i for i in range(len(tag2idx_srl)) if i != pad_id]\n",
|
||
"\n",
|
||
"cm_ner = confusion_matrix(true_ner_flat, pred_ner_flat, labels=labels_ner_no_pad)\n",
|
||
"cm_srl = confusion_matrix(true_srl_flat, pred_srl_flat, labels=labels_srl_no_pad)\n",
|
||
"\n",
|
||
"display_labels_ner = [idx2tag_ner[i] for i in labels_ner_no_pad]\n",
|
||
"display_labels_srl = [idx2tag_srl[i] for i in labels_srl_no_pad]\n",
|
||
"\n",
|
||
"# ------------------------------------------------------------------\n",
|
||
"# 4. Plot NER CM (tanpa PAD)\n",
|
||
"# ------------------------------------------------------------------\n",
|
||
"fig, ax = plt.subplots(figsize=(8, 8))\n",
|
||
"disp_ner = ConfusionMatrixDisplay(confusion_matrix=cm_ner, display_labels=display_labels_ner)\n",
|
||
"disp_ner.plot(include_values=True, values_format=\"d\", cmap=plt.cm.Blues, ax=ax, colorbar=False)\n",
|
||
"ax.set_title(\"NER Confusion Matrix\", fontsize=18)\n",
|
||
"plt.setp(ax.get_xticklabels(), rotation=45, ha=\"right\")\n",
|
||
"plt.tight_layout()\n",
|
||
"plt.show()\n",
|
||
"\n",
|
||
"# ------------------------------------------------------------------\n",
|
||
"# 5. Plot SRL CM (tanpa PAD)\n",
|
||
"# ------------------------------------------------------------------\n",
|
||
"fig, ax = plt.subplots(figsize=(8, 8))\n",
|
||
"disp_srl = ConfusionMatrixDisplay(confusion_matrix=cm_srl, display_labels=display_labels_srl)\n",
|
||
"disp_srl.plot(include_values=True, values_format=\"d\", cmap=plt.cm.Blues, ax=ax, colorbar=False)\n",
|
||
"ax.set_title(\"SRL Confusion Matrix\", fontsize=18)\n",
|
||
"plt.setp(ax.get_xticklabels(), rotation=45, ha=\"right\")\n",
|
||
"plt.tight_layout()\n",
|
||
"plt.show()\n",
|
||
"\n",
|
||
"# ------------------------------------------------------------------\n",
|
||
"# 6. Manual Confusion Table untuk Thesis (NER)\n",
|
||
"# ------------------------------------------------------------------\n",
|
||
"def manual_confusion_table(true_labels, pred_labels, labels, label_names):\n",
|
||
" true_labels = np.array(true_labels)\n",
|
||
" pred_labels = np.array(pred_labels)\n",
|
||
"\n",
|
||
" for i, label in enumerate(labels):\n",
|
||
" label_name = label_names[i]\n",
|
||
"\n",
|
||
" tp = np.sum((true_labels == label) & (pred_labels == label))\n",
|
||
" fn = np.sum((true_labels == label) & (pred_labels != label))\n",
|
||
" fp = np.sum((true_labels != label) & (pred_labels == label))\n",
|
||
" tn = np.sum((true_labels != label) & (pred_labels != label))\n",
|
||
"\n",
|
||
" table = pd.DataFrame({\n",
|
||
" 'Prediksi Positif': [f'{tp} (TP)', f'{fp} (FP)'],\n",
|
||
" 'Prediksi Negatif': [f'{fn} (FN)', f'{tn} (TN)']\n",
|
||
" }, index=['Aktual Positif', 'Aktual Negatif'])\n",
|
||
"\n",
|
||
" print(f\"\\n==== Confusion Matrix Manual untuk Kelas: {label_name} ====\")\n",
|
||
" print(table.to_string())\n",
|
||
"\n",
|
||
"# Jalankan untuk semua kelas NER\n",
|
||
"for label_id, label_name in zip(labels_ner_no_pad, display_labels_ner):\n",
|
||
" manual_confusion_table(true_ner_flat, pred_ner_flat, [label_id], [label_name])\n",
|
||
" \n",
|
||
"for label_id, label_name in zip(labels_srl_no_pad, display_labels_srl):\n",
|
||
" manual_confusion_table(true_srl_flat, pred_srl_flat, [label_id], [label_name])\n"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"id": "01cf3044",
|
||
"metadata": {},
|
||
"source": []
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 22,
|
||
"id": "a49f1dfe",
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"NER TAG accuracy : 88.15%\n",
|
||
"SRL TAG accuracy : 80.37%\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"from sklearn.metrics import accuracy_score, classification_report\n",
|
||
"\n",
|
||
"# ------------------------------------------------------------------\n",
|
||
"# 3b. Akurasi token‑level (tanpa PAD)\n",
|
||
"# ------------------------------------------------------------------\n",
|
||
"acc_ner = accuracy_score(true_ner_flat, pred_ner_flat)\n",
|
||
"acc_srl = accuracy_score(true_srl_flat, pred_srl_flat)\n",
|
||
"\n",
|
||
"print(f\"NER TAG accuracy : {acc_ner:.2%}\")\n",
|
||
"print(f\"SRL TAG accuracy : {acc_srl:.2%}\")"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 23,
|
||
"id": "9adad755",
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"\n",
|
||
"[NER] Classification report:\n",
|
||
" precision recall f1-score support\n",
|
||
"\n",
|
||
" DATE 0.98 0.92 0.95 336\n",
|
||
" ETH 0.97 0.95 0.96 93\n",
|
||
" EVENT 0.46 0.26 0.33 61\n",
|
||
" LOC 0.88 0.83 0.86 400\n",
|
||
" MISC 0.26 0.18 0.21 104\n",
|
||
" O 0.90 0.96 0.93 2360\n",
|
||
" ORG 0.27 0.21 0.24 75\n",
|
||
" PER 0.94 0.87 0.91 404\n",
|
||
" TIME 0.75 0.59 0.66 75\n",
|
||
"\n",
|
||
" accuracy 0.88 3908\n",
|
||
" macro avg 0.71 0.64 0.67 3908\n",
|
||
"weighted avg 0.87 0.88 0.87 3908\n",
|
||
"\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"\n",
|
||
"print(\"\\n[NER] Classification report:\")\n",
|
||
"print(\n",
|
||
" classification_report(\n",
|
||
" true_ner_flat,\n",
|
||
" pred_ner_flat,\n",
|
||
" labels=labels_ner_no_pad,\n",
|
||
" target_names=display_labels_ner,\n",
|
||
" digits=2,\n",
|
||
" )\n",
|
||
")"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 24,
|
||
"id": "7cd28380",
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"\n",
|
||
"[SRL] Classification report:\n",
|
||
" precision recall f1-score support\n",
|
||
"\n",
|
||
" ARG0 0.88 0.81 0.84 552\n",
|
||
" ARG1 0.60 0.85 0.70 637\n",
|
||
" ARG2 0.39 0.23 0.29 96\n",
|
||
" ARG3 0.00 0.00 0.00 4\n",
|
||
" ARGM-CAU 0.00 0.00 0.00 32\n",
|
||
" ARGM-DIR 0.00 0.00 0.00 5\n",
|
||
" ARGM-LOC 0.89 0.78 0.83 409\n",
|
||
" ARGM-MNR 0.00 0.00 0.00 55\n",
|
||
" ARGM-MOD 0.00 0.00 0.00 18\n",
|
||
" ARGM-NEG 0.00 0.00 0.00 1\n",
|
||
" ARGM-TMP 0.94 0.84 0.88 468\n",
|
||
" O 0.85 0.91 0.88 1135\n",
|
||
" V 0.86 0.78 0.82 496\n",
|
||
"\n",
|
||
" accuracy 0.80 3908\n",
|
||
" macro avg 0.42 0.40 0.40 3908\n",
|
||
"weighted avg 0.79 0.80 0.79 3908\n",
|
||
"\n"
|
||
]
|
||
},
|
||
{
|
||
"name": "stderr",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"/mnt/disc1/code/thesis_quiz_project/lstm-quiz/myenv/lib64/python3.10/site-packages/sklearn/metrics/_classification.py:1565: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
|
||
" _warn_prf(average, modifier, f\"{metric.capitalize()} is\", len(result))\n",
|
||
"/mnt/disc1/code/thesis_quiz_project/lstm-quiz/myenv/lib64/python3.10/site-packages/sklearn/metrics/_classification.py:1565: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
|
||
" _warn_prf(average, modifier, f\"{metric.capitalize()} is\", len(result))\n",
|
||
"/mnt/disc1/code/thesis_quiz_project/lstm-quiz/myenv/lib64/python3.10/site-packages/sklearn/metrics/_classification.py:1565: UndefinedMetricWarning: Precision is ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
|
||
" _warn_prf(average, modifier, f\"{metric.capitalize()} is\", len(result))\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"print(\"\\n[SRL] Classification report:\")\n",
|
||
"print(\n",
|
||
" classification_report(\n",
|
||
" true_srl_flat,\n",
|
||
" pred_srl_flat,\n",
|
||
" labels=labels_srl_no_pad,\n",
|
||
" target_names=display_labels_srl,\n",
|
||
" digits=2,\n",
|
||
" )\n",
|
||
")"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 25,
|
||
"id": "333745fd",
|
||
"metadata": {},
|
||
"outputs": [],
|
||
"source": [
|
||
"# def plot_training_history(history):\n",
|
||
"# epochs = range(1, len(history['loss']) + 1)\n",
|
||
"\n",
|
||
"# plt.figure(figsize=(14, 6))\n",
|
||
"\n",
|
||
"# # Plot Loss\n",
|
||
"# plt.subplot(1, 2, 1)\n",
|
||
"# plt.plot(epochs, history['loss'], label='Training Loss')\n",
|
||
"# plt.plot(epochs, history['val_loss'], label='Validation Loss')\n",
|
||
"# plt.title('Loss During Training')\n",
|
||
"# plt.xlabel('Epochs')\n",
|
||
"# plt.ylabel('Loss')\n",
|
||
"# plt.legend()\n",
|
||
"\n",
|
||
"# # Plot Accuracy\n",
|
||
"# plt.subplot(1, 2, 2)\n",
|
||
"# plt.plot(epochs, history['ner_output_accuracy'], label='NER Train Acc')\n",
|
||
"# plt.plot(epochs, history['val_ner_output_accuracy'], label='NER Val Acc')\n",
|
||
"# plt.plot(epochs, history['srl_output_accuracy'], label='SRL Train Acc')\n",
|
||
"# plt.plot(epochs, history['val_srl_output_accuracy'], label='SRL Val Acc')\n",
|
||
"# plt.title('Accuracy During Training')\n",
|
||
"# plt.xlabel('Epochs')\n",
|
||
"# plt.ylabel('Accuracy')\n",
|
||
"# plt.legend()\n",
|
||
"\n",
|
||
"# plt.tight_layout()\n",
|
||
"# plt.show()\n",
|
||
"\n",
|
||
"# plot_training_history(history.history)"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 26,
|
||
"id": "df36e200",
|
||
"metadata": {},
|
||
"outputs": [],
|
||
"source": [
|
||
"# def token_level_accuracy(y_true, y_pred):\n",
|
||
"# total, correct = 0, 0\n",
|
||
"# for true_seq, pred_seq in zip(y_true, y_pred):\n",
|
||
"# for t, p in zip(true_seq, pred_seq):\n",
|
||
"# if t.sum() == 0:\n",
|
||
"# continue\n",
|
||
"# total += 1\n",
|
||
"# if t.argmax() == p.argmax():\n",
|
||
"# correct += 1\n",
|
||
"# return correct / total\n",
|
||
"\n",
|
||
"# def decode_predictions(pred, true, idx2tag):\n",
|
||
"# true_out, pred_out = [], []\n",
|
||
"# for pred_seq, true_seq in zip(pred, true):\n",
|
||
"# t_labels, p_labels = [], []\n",
|
||
"# for p_tok, t_tok in zip(pred_seq, true_seq):\n",
|
||
"# if t_tok.sum() == 0:\n",
|
||
"# continue\n",
|
||
"# t_labels.append(idx2tag[t_tok.argmax()])\n",
|
||
"# p_labels.append(idx2tag[p_tok.argmax()])\n",
|
||
"# true_out.append(t_labels)\n",
|
||
"# pred_out.append(p_labels)\n",
|
||
"# return true_out, pred_out\n",
|
||
"\n",
|
||
"# results = model.evaluate(X_test, {\"ner_output\": y_ner_test, \"srl_output\": y_srl_test}, verbose=0)\n",
|
||
"# for name, value in zip(model.metrics_names, results):\n",
|
||
"# print(f\"{name}: {value}\")\n",
|
||
"\n",
|
||
"# y_pred_ner, y_pred_srl = model.predict(X_test, verbose=0)\n",
|
||
"\n",
|
||
"# true_ner, pred_ner = decode_predictions(y_pred_ner, y_ner_test, idx2tag_ner)\n",
|
||
"# true_srl, pred_srl = decode_predictions(y_pred_srl, y_srl_test, idx2tag_srl)\n",
|
||
"\n",
|
||
"# acc_ner = token_level_accuracy(y_ner_test, y_pred_ner)\n",
|
||
"# acc_srl = token_level_accuracy(y_srl_test, y_pred_srl)\n",
|
||
"\n",
|
||
"# print(f\"NER Token Accuracy {acc_ner:.2%}\")\n",
|
||
"# print(f\"SRL Token Accuracy {acc_srl:.2%}\")"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 27,
|
||
"id": "9127cce0",
|
||
"metadata": {},
|
||
"outputs": [],
|
||
"source": [
|
||
"# print(\"[NER] Classification Report:\")\n",
|
||
"# print(classification_report(true_ner, pred_ner, digits=2))"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 28,
|
||
"id": "300897b8",
|
||
"metadata": {},
|
||
"outputs": [],
|
||
"source": [
|
||
"# print(\"SRL Classification Resport:\")\n",
|
||
"# print(classification_report(true_srl, pred_srl, digits=2))"
|
||
]
|
||
}
|
||
],
|
||
"metadata": {
|
||
"kernelspec": {
|
||
"display_name": "myenv",
|
||
"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": "3.10.16"
|
||
}
|
||
},
|
||
"nbformat": 4,
|
||
"nbformat_minor": 5
|
||
}
|