Datasets:
ArXiv:
License:
File size: 34,233 Bytes
0e5d934 c47824b 0e5d934 c47824b 0e5d934 c47824b 0e5d934 c47824b 0e5d934 c47824b 0e5d934 c47824b 0e5d934 c47824b 0e5d934 c47824b 0e5d934 c47824b 0e5d934 c47824b 0e5d934 c47824b 0e5d934 c47824b 0e5d934 c47824b 0e5d934 c47824b 0e5d934 c47824b 0e5d934 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 |
{
"cells": [
{
"cell_type": "code",
"execution_count": 1,
"id": "5cbc726d-b9ed-4a87-8424-780a4db25a36",
"metadata": {},
"outputs": [],
"source": [
"import json\n",
"import numpy as np\n",
"\n",
"from pathlib import Path\n",
"from tabulate import tabulate"
]
},
{
"cell_type": "code",
"execution_count": 2,
"id": "dbffdf82-5cb2-485d-9334-6e6fb8cc9c91",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"{'dbmdz/distilbert-base-turkish-cased': 'results-dbmdz-distilbert-base-turkish-cased.jsonl', 'dbmdz/convbert-base-turkish-mc4-cased': 'results-dbmdz-convbert-base-turkish-mc4-cased.jsonl', 'dbmdz/bert-base-turkish-uncased': 'results-dbmdz-bert-base-turkish-uncased.jsonl', 'dbmdz/bert-base-turkish-128k-cased': 'results-dbmdz-bert-base-turkish-128k-cased.jsonl', 'dbmdz/bert-base-turkish-cased': 'results-dbmdz-bert-base-turkish-cased.jsonl', 'dbmdz/electra-small-turkish-cased-generator': 'results-dbmdz-electra-small-turkish-cased-generator.jsonl', 'dbmdz/convbert-base-turkish-cased': 'results-dbmdz-convbert-base-turkish-cased.jsonl', 'dbmdz/electra-base-turkish-cased-generator': 'results-dbmdz-electra-base-turkish-cased-generator.jsonl', 'dbmdz/convbert-base-turkish-mc4-uncased': 'results-dbmdz-convbert-base-turkish-mc4-uncased.jsonl', 'dbmdz/electra-base-turkish-mc4-cased-generator': 'results-dbmdz-electra-base-turkish-mc4-cased-generator.jsonl', 'dbmdz/bert-base-turkish-128k-uncased': 'results-dbmdz-bert-base-turkish-128k-uncased.jsonl', 'dbmdz/electra-base-turkish-mc4-uncased-generator': 'results-dbmdz-electra-base-turkish-mc4-uncased-generator.jsonl'}\n"
]
}
],
"source": [
"model_names = [\n",
" \"dbmdz/electra-small-turkish-cased-generator\",\n",
" \"dbmdz/electra-base-turkish-cased-generator\",\n",
" \"dbmdz/electra-base-turkish-mc4-cased-generator\",\n",
" \"dbmdz/electra-base-turkish-mc4-uncased-generator\",\n",
" \"dbmdz/bert-base-turkish-cased\",\n",
" \"dbmdz/bert-base-turkish-uncased\",\n",
" \"dbmdz/bert-base-turkish-128k-cased\",\n",
" \"dbmdz/bert-base-turkish-128k-uncased\",\n",
" \"dbmdz/distilbert-base-turkish-cased\",\n",
" \"dbmdz/convbert-base-turkish-cased\",\n",
" \"dbmdz/convbert-base-turkish-mc4-cased\",\n",
" \"dbmdz/convbert-base-turkish-mc4-uncased\",\n",
"]\n",
"\n",
"evaluation_files = [str(file) for file in Path(\"./\").iterdir() if file.name.endswith(\".jsonl\")]\n",
"\n",
"# \"dbmdz/electra-base-turkish-cased-generator\" -> \"results-dbmdz-electra-base-turkish-cased-generator.jsonl\"\n",
"model_name_evaluation_filename_mapping = {}\n",
"\n",
"for evaluation_file in evaluation_files:\n",
" reconstructed_model_name = evaluation_file.replace(\"results-\", \"\").replace(\".jsonl\", \"\")\n",
" reconstructed_model_name = reconstructed_model_name.replace(\"dbmdz-\", \"dbmdz/\")\n",
"\n",
" model_name_evaluation_filename_mapping[reconstructed_model_name] = evaluation_file\n",
"print(model_name_evaluation_filename_mapping)"
]
},
{
"cell_type": "code",
"execution_count": 3,
"id": "7170ec01-f0f1-49bb-9beb-7927c3b5b3f7",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"{'data/base/augmented_anaphor_agreement.csv': 'Anaphor Agreement', 'data/base/augmented_argument_structure_transitive.csv': 'Argument Str. Tran.', 'data/base/augmented_argument_structure_ditransitive.csv': 'Argument Str. Ditr.', 'data/base/augmented_binding.csv': 'Binding', 'data/base/augmented_determiners.csv': 'Determiners', 'data/base/augmented_ellipsis.csv': 'Ellipsis', 'data/base/augmented_irregular_forms.csv': 'Irregular Forms', 'data/base/augmented_island_effects.csv': 'Island Effects', 'data/base/augmented_nominalization.csv': 'Nominalization', 'data/base/augmented_npi_licensing.csv': 'NPI Licensing', 'data/base/augmented_passives.csv': 'Passives', 'data/base/augmented_quantifiers.csv': 'Quantifiers', 'data/base/augmented_relative_clauses.csv': 'Relative Clauses', 'data/base/augmented_scrambling.csv': 'Scrambling', 'data/base/augmented_subject_verb_agreement.csv': 'Subject Agreement', 'data/base/augmented_suspended_affixation.csv': 'Suspended Affixation'}\n"
]
}
],
"source": [
"phenomenons = [\n",
"\"Anaphor Agreement\",\n",
"\"Argument Str. Tran.\",\n",
"\"Argument Str. Ditr.\",\n",
"\"Binding\",\n",
"\"Determiners\",\n",
"\"Ellipsis\",\n",
"\"Irregular Forms\",\n",
"\"Island Effects\",\n",
"\"Nominalization\",\n",
"\"NPI Licensing\",\n",
"\"Passives\",\n",
"\"Quantifiers\",\n",
"\"Relative Clauses\",\n",
"\"Scrambling\",\n",
"\"Subject Agreement\",\n",
"\"Suspended Affixation\",\n",
"]\n",
"\n",
"# data/base/augmented_anaphor_agreement.csv -> Anaphor Agreement\n",
"data_task_names = [\n",
" \"data/base/augmented_anaphor_agreement.csv\",\n",
" \"data/base/augmented_argument_structure_transitive.csv\",\n",
" \"data/base/augmented_argument_structure_ditransitive.csv\",\n",
" \"data/base/augmented_binding.csv\",\n",
" \"data/base/augmented_determiners.csv\",\n",
" \"data/base/augmented_ellipsis.csv\",\n",
" \"data/base/augmented_irregular_forms.csv\",\n",
" \"data/base/augmented_island_effects.csv\",\n",
" \"data/base/augmented_nominalization.csv\",\n",
" \"data/base/augmented_npi_licensing.csv\",\n",
" \"data/base/augmented_passives.csv\",\n",
" \"data/base/augmented_quantifiers.csv\",\n",
" \"data/base/augmented_relative_clauses.csv\",\n",
" \"data/base/augmented_scrambling.csv\",\n",
" \"data/base/augmented_subject_verb_agreement.csv\",\n",
" \"data/base/augmented_suspended_affixation.csv\",\n",
"]\n",
"\n",
"task_phenomenon_mapping = {}\n",
"\n",
"for data_task_name, phenomenon in zip(data_task_names, phenomenons):\n",
" task_phenomenon_mapping[data_task_name] = phenomenon\n",
"print(task_phenomenon_mapping)"
]
},
{
"cell_type": "code",
"execution_count": 4,
"id": "5c444bbe-93e7-4e85-8560-56bff13e1bb0",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"| Phenomenon | [`dbmdz/electra-small-turkish-cased-generator`](https://huggingface.co/dbmdz/electra-small-turkish-cased-generator) | [`dbmdz/electra-base-turkish-cased-generator`](https://huggingface.co/dbmdz/electra-base-turkish-cased-generator) | [`dbmdz/electra-base-turkish-mc4-cased-generator`](https://huggingface.co/dbmdz/electra-base-turkish-mc4-cased-generator) | [`dbmdz/electra-base-turkish-mc4-uncased-generator`](https://huggingface.co/dbmdz/electra-base-turkish-mc4-uncased-generator) | [`dbmdz/bert-base-turkish-cased`](https://huggingface.co/dbmdz/bert-base-turkish-cased) | [`dbmdz/bert-base-turkish-uncased`](https://huggingface.co/dbmdz/bert-base-turkish-uncased) | [`dbmdz/bert-base-turkish-128k-cased`](https://huggingface.co/dbmdz/bert-base-turkish-128k-cased) | [`dbmdz/bert-base-turkish-128k-uncased`](https://huggingface.co/dbmdz/bert-base-turkish-128k-uncased) | [`dbmdz/distilbert-base-turkish-cased`](https://huggingface.co/dbmdz/distilbert-base-turkish-cased) | [`dbmdz/convbert-base-turkish-cased`](https://huggingface.co/dbmdz/convbert-base-turkish-cased) | [`dbmdz/convbert-base-turkish-mc4-cased`](https://huggingface.co/dbmdz/convbert-base-turkish-mc4-cased) | [`dbmdz/convbert-base-turkish-mc4-uncased`](https://huggingface.co/dbmdz/convbert-base-turkish-mc4-uncased) |\n",
"|----------------------|-----------------------------------------------------------------------------------------------------------------------|---------------------------------------------------------------------------------------------------------------------|-----------------------------------------------------------------------------------------------------------------------------|---------------------------------------------------------------------------------------------------------------------------------|-------------------------------------------------------------------------------------------|-----------------------------------------------------------------------------------------------|-----------------------------------------------------------------------------------------------------|---------------------------------------------------------------------------------------------------------|-------------------------------------------------------------------------------------------------------|---------------------------------------------------------------------------------------------------|-----------------------------------------------------------------------------------------------------------|---------------------------------------------------------------------------------------------------------------|\n",
"| Anaphor Agreement | 74.1 | 94.3 | 94.3 | 92.8 | 96.7 | 97.3 | 97.3 | 97.7 | 96.9 | 58.1 | 44.3 | 44.6 |\n",
"| Argument Str. Tran. | 86.6 | 99.6 | 99.4 | 98.7 | 99.7 | 99.6 | 99.8 | 99.1 | 97.5 | 51.9 | 58.1 | 51.3 |\n",
"| Argument Str. Ditr. | 79.3 | 96.1 | 95.5 | 95.2 | 99.8 | 96.1 | 96.1 | 96.1 | 95.4 | 64.6 | 58.6 | 64.5 |\n",
"| Binding | 70.7 | 96.2 | 91.4 | 89.6 | 99.9 | 98.5 | 97.7 | 99 | 93 | 89.1 | 49.4 | 78.4 |\n",
"| Determiners | 91.8 | 99.3 | 98.2 | 99.1 | 99.9 | 100 | 99 | 99.3 | 82.9 | 0 | 0 | 0 |\n",
"| Ellipsis | 10.6 | 49.7 | 46.3 | 49 | 87.4 | 73.6 | 96.6 | 87.5 | 13.6 | 54.7 | 57.8 | 67.9 |\n",
"| Irregular Forms | 98.7 | 97.9 | 99 | 99.8 | 98.8 | 100 | 99.9 | 99.6 | 94.1 | 82.9 | 86.6 | 95.2 |\n",
"| Island Effects | 39.1 | 35.3 | 41.8 | 44 | 49.4 | 39.8 | 60.9 | 51.2 | 47.4 | 96.7 | 99.4 | 100 |\n",
"| Nominalization | 90 | 96.6 | 97 | 95.4 | 97.4 | 97 | 98.9 | 97.4 | 95.6 | 55.2 | 59.2 | 60.6 |\n",
"| NPI Licensing | 90.9 | 96.1 | 95 | 98 | 98.2 | 97.6 | 97.2 | 95 | 92.1 | 82.1 | 95.6 | 71.9 |\n",
"| Passives | 100 | 91.2 | 93.6 | 91.6 | 82.2 | 78.1 | 84.4 | 81.3 | 98.8 | 100 | 100 | 99 |\n",
"| Quantifiers | 97.9 | 98 | 98 | 97.6 | 95.7 | 94.6 | 98 | 98.4 | 98.4 | 99 | 99 | 99 |\n",
"| Relative Clauses | 79.9 | 90.7 | 92 | 91.6 | 97.7 | 97.5 | 97 | 98.5 | 92 | 53.4 | 53.7 | 56.9 |\n",
"| Scrambling | 99.5 | 100 | 100 | 99.8 | 100 | 100 | 99.6 | 100 | 99.8 | 38.7 | 59.3 | 63.3 |\n",
"| Subject Agreement | 82.8 | 99 | 97.2 | 96.1 | 98.3 | 99.2 | 99.1 | 98.8 | 97 | 47.7 | 43.9 | 56.4 |\n",
"| Suspended Affixation | 97.5 | 99 | 99.1 | 98.8 | 100 | 100 | 100 | 100 | 100 | 25.4 | 12.8 | 23.2 |\n",
"| Model Average | 80.6 | 89.9 | 89.9 | 89.8 | 93.8 | 91.8 | 95.1 | 93.7 | 87.2 | 62.5 | 61.1 | 64.5 |\n"
]
}
],
"source": [
"headers = [\"Phenomenon\"] + [f\"[`{model_name}`](https://huggingface.co/{model_name})\" for model_name in model_names]\n",
"\n",
"model_results = {}\n",
"\n",
"for model_name in model_names:\n",
" current_model_results = {}\n",
" with open(model_name_evaluation_filename_mapping[model_name], \"rt\") as f_p:\n",
" for line in f_p:\n",
" data = json.loads(line)\n",
"\n",
" current_task_name = data[\"file_name\"]\n",
" current_task_accuracy = round(data[\"accuracy\"], 1)\n",
" current_task_phenomenon = task_phenomenon_mapping[current_task_name]\n",
" \n",
" current_model_results[current_task_phenomenon] = current_task_accuracy\n",
"\n",
" model_average = list(current_model_results.values())\n",
" model_average = np.mean(model_average)\n",
" model_average = round(model_average, 1)\n",
"\n",
" current_model_results[\"Model Average\"] = model_average\n",
"\n",
" model_results[model_name] = current_model_results\n",
"\n",
"table = []\n",
"\n",
"phenomenons.append(\"Model Average\")\n",
"\n",
"for phenomenon in phenomenons:\n",
" current_row = []\n",
"\n",
" current_row.append(phenomenon)\n",
"\n",
" for model_name in model_names:\n",
" current_phenomenon_accuracy = model_results[model_name][phenomenon]\n",
" current_row.append(current_phenomenon_accuracy)\n",
"\n",
" table.append(current_row)\n",
" \n",
"print(tabulate(table, headers=headers, tablefmt=\"github\"))"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3 (ipykernel)",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.13.3"
}
},
"nbformat": 4,
"nbformat_minor": 5
}
|