diff --git a/DBE_annual-project.ipynb b/DBE_annual-project.ipynb
index 579b089f2f42054a5c2deb6d01de0672a04fa956..f7167f38e7855563d6134de79f8fb25c33e7e723 100644
--- a/DBE_annual-project.ipynb
+++ b/DBE_annual-project.ipynb
@@ -30,7 +30,7 @@
"metadata": {
"id": "TM4RsrfKerFL"
},
- "execution_count": null,
+ "execution_count": 1,
"outputs": []
},
{
@@ -60,7 +60,7 @@
"metadata": {
"id": "uOZWTmCrfmyR"
},
- "execution_count": null,
+ "execution_count": 2,
"outputs": []
},
{
@@ -69,10 +69,26 @@
"!python -m spacy init fill-config /content/DBE_annual-project/base_config.cfg /content/DBE_annual-project/base_config_spacy.cfg\n"
],
"metadata": {
- "id": "PcgBSkHnitiO"
+ "id": "PcgBSkHnitiO",
+ "colab": {
+ "base_uri": "https://localhost:8080/"
+ },
+ "outputId": "e935551e-cfcc-41f9-a7f2-9c758a609ede"
},
- "execution_count": null,
- "outputs": []
+ "execution_count": 3,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "name": "stdout",
+ "text": [
+ "\u001b[38;5;2m✔ Auto-filled config with all values\u001b[0m\n",
+ "\u001b[38;5;2m✔ Saved config\u001b[0m\n",
+ "/content/DBE_annual-project/base_config_spacy.cfg\n",
+ "You can now add your data and train your pipeline:\n",
+ "python -m spacy train base_config_spacy.cfg --paths.train ./train.spacy --paths.dev ./dev.spacy\n"
+ ]
+ }
+ ]
},
{
"cell_type": "code",
@@ -80,10 +96,57 @@
"!python -m spacy debug data /content/DBE_annual-project/base_config_spacy.cfg"
],
"metadata": {
- "id": "SqDuveAaHzFp"
+ "id": "SqDuveAaHzFp",
+ "colab": {
+ "base_uri": "https://localhost:8080/"
+ },
+ "outputId": "8d3ffd0d-9aed-468f-dfdd-ccf70de356cb"
},
- "execution_count": null,
- "outputs": []
+ "execution_count": 4,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "name": "stdout",
+ "text": [
+ "\u001b[1m\n",
+ "============================ Data file validation ============================\u001b[0m\n",
+ "Downloading: 100% 481/481 [00:00<00:00, 441kB/s]\n",
+ "Downloading: 100% 899k/899k [00:01<00:00, 693kB/s]\n",
+ "Downloading: 100% 456k/456k [00:01<00:00, 422kB/s]\n",
+ "Downloading: 100% 1.36M/1.36M [00:01<00:00, 1.03MB/s]\n",
+ "Downloading: 100% 501M/501M [00:12<00:00, 40.1MB/s]\n",
+ "Some weights of the model checkpoint at roberta-base were not used when initializing RobertaModel: ['lm_head.dense.weight', 'lm_head.bias', 'lm_head.dense.bias', 'lm_head.decoder.weight', 'lm_head.layer_norm.weight', 'lm_head.layer_norm.bias']\n",
+ "- This IS expected if you are initializing RobertaModel from the checkpoint of a model trained on another task or with another architecture (e.g. initializing a BertForSequenceClassification model from a BertForPreTraining model).\n",
+ "- This IS NOT expected if you are initializing RobertaModel from the checkpoint of a model that you expect to be exactly identical (initializing a BertForSequenceClassification model from a BertForSequenceClassification model).\n",
+ "\u001b[38;5;2m✔ Pipeline can be initialized with data\u001b[0m\n",
+ "\u001b[38;5;2m✔ Corpus is loadable\u001b[0m\n",
+ "\u001b[1m\n",
+ "=============================== Training stats ===============================\u001b[0m\n",
+ "Language: en\n",
+ "Training pipeline: transformer, ner\n",
+ "150 training docs\n",
+ "1 evaluation docs\n",
+ "\u001b[38;5;2m✔ No overlap between training and evaluation data\u001b[0m\n",
+ "\u001b[38;5;3mâš Low number of examples to train a new pipeline (150)\u001b[0m\n",
+ "\u001b[1m\n",
+ "============================== Vocab & Vectors ==============================\u001b[0m\n",
+ "\u001b[38;5;4mℹ 5235 total word(s) in the data (1424 unique)\u001b[0m\n",
+ "\u001b[38;5;4mℹ No word vectors present in the package\u001b[0m\n",
+ "\u001b[1m\n",
+ "========================== Named Entity Recognition ==========================\u001b[0m\n",
+ "\u001b[38;5;4mℹ 1 label(s)\u001b[0m\n",
+ "0 missing value(s) (tokens with '-' label)\n",
+ "\u001b[38;5;2m✔ Good amount of examples for all labels\u001b[0m\n",
+ "\u001b[38;5;2m✔ Examples without occurrences available for all labels\u001b[0m\n",
+ "\u001b[38;5;2m✔ No entities consisting of or starting/ending with whitespace\u001b[0m\n",
+ "\u001b[38;5;2m✔ No entities crossing sentence boundaries\u001b[0m\n",
+ "\u001b[1m\n",
+ "================================== Summary ==================================\u001b[0m\n",
+ "\u001b[38;5;2m✔ 7 checks passed\u001b[0m\n",
+ "\u001b[38;5;3mâš 1 warning\u001b[0m\n"
+ ]
+ }
+ ]
},
{
"cell_type": "code",
@@ -91,22 +154,65 @@
"!python -m spacy train /content/DBE_annual-project/base_config_spacy.cfg --gpu-id 0 --paths.train /content/DBE_annual-project/train.spacy --paths.dev /content/DBE_annual-project/dev.spacy --output ./"
],
"metadata": {
- "id": "VLxW_Bcci9N1"
+ "id": "VLxW_Bcci9N1",
+ "colab": {
+ "base_uri": "https://localhost:8080/"
+ },
+ "outputId": "e6f96b52-8714-475b-b45e-c5c29a06fd8f"
},
- "execution_count": null,
- "outputs": []
+ "execution_count": 5,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "name": "stdout",
+ "text": [
+ "\u001b[38;5;4mℹ Saving to output directory: .\u001b[0m\n",
+ "\u001b[38;5;4mℹ Using GPU: 0\u001b[0m\n",
+ "\u001b[1m\n",
+ "=========================== Initializing pipeline ===========================\u001b[0m\n",
+ "[2023-01-14 11:33:26,302] [INFO] Set up nlp object from config\n",
+ "INFO:spacy:Set up nlp object from config\n",
+ "[2023-01-14 11:33:26,312] [INFO] Pipeline: ['transformer', 'ner']\n",
+ "INFO:spacy:Pipeline: ['transformer', 'ner']\n",
+ "[2023-01-14 11:33:26,316] [INFO] Created vocabulary\n",
+ "INFO:spacy:Created vocabulary\n",
+ "[2023-01-14 11:33:26,316] [INFO] Finished initializing nlp object\n",
+ "INFO:spacy:Finished initializing nlp object\n",
+ "Some weights of the model checkpoint at roberta-base were not used when initializing RobertaModel: ['lm_head.bias', 'lm_head.layer_norm.bias', 'lm_head.dense.weight', 'lm_head.dense.bias', 'lm_head.layer_norm.weight', 'lm_head.decoder.weight']\n",
+ "- This IS expected if you are initializing RobertaModel from the checkpoint of a model trained on another task or with another architecture (e.g. initializing a BertForSequenceClassification model from a BertForPreTraining model).\n",
+ "- This IS NOT expected if you are initializing RobertaModel from the checkpoint of a model that you expect to be exactly identical (initializing a BertForSequenceClassification model from a BertForSequenceClassification model).\n",
+ "[2023-01-14 11:33:43,715] [INFO] Initialized pipeline components: ['transformer', 'ner']\n",
+ "INFO:spacy:Initialized pipeline components: ['transformer', 'ner']\n",
+ "\u001b[38;5;2m✔ Initialized pipeline\u001b[0m\n",
+ "\u001b[1m\n",
+ "============================= Training pipeline =============================\u001b[0m\n",
+ "\u001b[38;5;4mℹ Pipeline: ['transformer', 'ner']\u001b[0m\n",
+ "\u001b[38;5;4mℹ Initial learn rate: 0.0\u001b[0m\n",
+ "E # LOSS TRANS... LOSS NER ENTS_F ENTS_P ENTS_R SCORE \n",
+ "--- ------ ------------- -------- ------ ------ ------ ------\n",
+ " 0 0 6733.00 778.82 0.00 0.00 0.00 0.00\n",
+ "\n",
+ "Aborted!\n"
+ ]
+ }
+ ]
},
{
"cell_type": "code",
"source": [
"%%capture \n",
- "zipname = \"annual-project.zip\"\n",
+ "import pytz\n",
+ "import datetime\n",
+ "timezone = pytz.timezone(\"Europe/Berlin\")\n",
+ "timestamp = str(datetime.datetime.now(timezone).strftime(\"%Y_%m_%d__%H_%M_%S\"))\n",
+ "\n",
+ "zipname = \"annual-project_\" + str(timestamp) + \".zip\"\n",
"!zip -r /content/drive/MyDrive/Colab_files/{zipname} /content/DBE_annual-project"
],
"metadata": {
"id": "h-k2chKIO-IK"
},
- "execution_count": null,
+ "execution_count": 7,
"outputs": []
}
]