{ "cells": [ { "cell_type": "code", "execution_count": 1, "metadata": { "id": "kH18jD5cR_Ks", "colab": { "base_uri": "https://localhost:8080/" }, "outputId": "c7e2432c-67f1-4f1a-8fad-b029e16c6f8e" }, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m44.1/44.1 kB\u001b[0m \u001b[31m3.8 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n", "\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m122.4/122.4 MB\u001b[0m \u001b[31m8.0 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n", "\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m310.2/310.2 kB\u001b[0m \u001b[31m23.9 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n", "\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m10.0/10.0 MB\u001b[0m \u001b[31m117.9 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n", "\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m480.6/480.6 kB\u001b[0m \u001b[31m32.3 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n", "\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m116.3/116.3 kB\u001b[0m \u001b[31m11.0 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n", "\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m179.3/179.3 kB\u001b[0m \u001b[31m16.5 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n", "\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m134.8/134.8 kB\u001b[0m \u001b[31m12.1 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n", "\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m3.0/3.0 MB\u001b[0m \u001b[31m91.3 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n", "\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m194.1/194.1 kB\u001b[0m \u001b[31m17.4 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n", "\u001b[?25h\u001b[31mERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.\n", "gcsfs 2024.10.0 requires fsspec==2024.10.0, but you have fsspec 2024.9.0 which is incompatible.\u001b[0m\u001b[31m\n", "\u001b[0m" ] } ], "source": [ "# !pip install -q accelerate peft bitsandbytes transformers trl faiss-gpu langchain_community wandb flash-attn\n", "!pip install -q accelerate peft bitsandbytes transformers trl datasets\n", "\n", "# flash-attn" ] }, { "cell_type": "code", "execution_count": 2, "metadata": { "id": "cgVNTbBa-D3j" }, "outputs": [], "source": [ "# load the required packages.\n", "import torch\n", "from datasets import load_dataset, Dataset\n", "from peft import LoraConfig, AutoPeftModelForCausalLM, PeftModel, get_peft_model\n", "from transformers import AutoModelForCausalLM, AutoTokenizer, BitsAndBytesConfig, TrainingArguments, AutoConfig, set_seed\n", "from trl import SFTTrainer\n", "import bitsandbytes as bnb\n", "import transformers\n", "\n", "import os\n", "import numpy as np\n", "import pandas as pd\n", "import sqlparse\n", "import re\n", "import json\n", "\n", "from huggingface_hub import hf_hub_download\n", "from huggingface_hub import HfFileSystem" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "s7RtbLxPwTx2", "colab": { "base_uri": "https://localhost:8080/" }, "outputId": "43d5ff02-3e33-439c-e3a9-c18ce99d237d" }, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "Mounted at /content/drive\n" ] } ], "source": [ "# from google.colab import drive\n", "# from google.colab import userdata\n", "# drive.mount('/content/drive')" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "basaX_55Yf_D" }, "outputs": [], "source": [ "#transformers.logging.set_verbosity_info()" ] }, { "cell_type": "code", "execution_count": 3, "metadata": { "id": "bkkjgGdlrNcq" }, "outputs": [], "source": [ "WRITE_TOKEN = userdata.get('hf_write')\n", "READ_TOKEN = userdata.get('hf_read')" ] }, { "cell_type": "code", "execution_count": 4, "metadata": { "id": "7CKnwlRfZj4V" }, "outputs": [], "source": [ "model_name = \"TinyLlama/TinyLlama-1.1B-Chat-v1.0\"\n", "out_name = \"lleticiasilvaa/TinyLlama-1.1B-GerarSQL-v2-identacao\"\n", "prev_checkpoint = None # \"checkpoint-3000\"" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "9zs7nCmt-pMC" }, "outputs": [], "source": [ "#!huggingface-cli login" ] }, { "cell_type": "code", "execution_count": 5, "metadata": { "id": "PZdnxs8k-Cgl" }, "outputs": [], "source": [ "bird_id=\"NESPED-GEN/bird\"\n", "spider_id=\"NESPED-GEN/spider_variacoes_formato_schema\"\n", "\n", "selector_id = \"NESPED-GEN/selector_testes\"" ] }, { "cell_type": "markdown", "metadata": { "id": "xT2iRdCN_MFH" }, "source": [ "### Load Data\n" ] }, { "cell_type": "code", "execution_count": 6, "metadata": { "id": "lLKgY40efdJo", "colab": { "base_uri": "https://localhost:8080/", "height": 281, "referenced_widgets": [ "bae6f37fafae4e6191e871e312da8e0a", "414fdecff88f4274965245c29726ccee", "05a27f58c6dc4f86ab8e8ef86da01282", "1180ab06b3164e70b85a35bb6e03b61f", "463fe03d56a64a9a8f9eabb607f04cc9", "1c27a07fdf114461a8756b465c595178", "52b026c7f78945c9bffcd1831abb53ce", "81c9523ec3e24ad399fd02aa0ab802cc", "9e3160d3aa6b43e8919ded34eee885a2", "6c5845a6b262459faac5251aefad494b", "4cdbaf0948c24fbe91d3c2846326e5bb", "6d6b9b996f5a42709ba70025657187bb", "e0d6b0786a0b4f60be5b8f23737dcac7", "3e7aea0f351a4e57ae13dc417435fab4", "b3ea00f427ef43ddb2115db97b62506e", "574012abbdbf4df7b71a722b037f55e9", "8163814231524dad9a68e31821185bd7", "59ca8dacec67490ca85fb26562ec6017", "cf522c8fd3314b60a6ef3d2396700887", "166da246d22a4e0c9bd28cb9dd280081", "8a573702b9364cbea6dc412b8c5c2dc6", "6e4677a62beb45b4ad9e47218f5abf5c", "e71d9b9e72b94b1f8b5404d5a2b855c5", "110dae7d8dde46f695d431676d01d2eb", "6db8779ae2b841ddae4412a8536cc380", "46c5d1e9321c43b5bfe9bf278a04c3cf", "d4e80de4d6bd4a518516939bc5465478", "0b1142c6e9324d5ea9afd0d20cfdb27e", "bb9d06e11839461a98ef04c233d9dcf0", "07c56e1d9da5444b9c125b2e3b6e6f41", "b16c302bce92479cbe7278b8a9ac11c6", "f624184ee71149ac981960a564e24374", "382f235790084fac8737b5aeafe5adb4", "ee04bf28eb4943cd84a078c7540f81ff", "440ea5f3f24d405b8ea7026fbe1086d4", "1aa17586638b4f288ad644cceda7ad26", "dee4f6fb73b540d39fcc553b84119978", "4b3865d6e80c4c7c897f763de0ab3590", "6a7eb1c11606484489551045f55459e4", "8f02831366904be995a61aa911a8d0a5", "a93b31ad9ca742319cce4bb304aaca91", "cd0b50cd6ac243bd928f35cc4212fc7e", "4660175909174679a1000ff24b3fe667", "d97288cdaf0246518d34a73076348253", "b20fffcc77b0423ea45bf748c28bdddf", "b634afe8d5544557bcde062645895c64", "27bfcf71466846a2b9a7d1d4dae7072b", "83c7d14e14c64785ad7bea0fecce7ff7", "ab06caa7a607427cb53439fb6e73ee59", "bd8b3cc0b7b74781bd2ee88e491d1c06", "c9e87b254cad44a7a99f666d5b16babb", "003fd152378448f2b623fee331cb1810", "c3a18115f6fc4718a9221f3d2b517e8c", "d0e4d0eefe5048ad8fc6635d68b9d0e1", "afdb17d434a3429e8bdbafe52672d4a8" ] }, "outputId": "aa987e42-2219-4bb3-d557-ed3b4bf0f693" }, "outputs": [ { "output_type": "stream", "name": "stderr", "text": [ "/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_token.py:89: UserWarning: \n", "The secret `HF_TOKEN` does not exist in your Colab secrets.\n", "To authenticate with the Hugging Face Hub, create a token in your settings tab (https://huggingface.co/settings/tokens), set it as secret in your Google Colab and restart your session.\n", "You will be able to reuse this secret in all of your notebooks.\n", "Please note that authentication is recommended but still optional to access public models or datasets.\n", " warnings.warn(\n" ] }, { "output_type": "display_data", "data": { "text/plain": [ "README.md: 0%| | 0.00/809 [00:00\n", "' + message['content'] + eos_token }}\n", "{% elif message['role'] == 'system' %}\n", "{{ '<|system|>\n", "' + message['content'] + eos_token }}\n", "{% elif message['role'] == 'assistant' %}\n", "{{ '<|assistant|>\n", "' + message['content'] + eos_token }}\n", "{% endif %}\n", "{% if loop.last and add_generation_prompt %}\n", "{{ '<|assistant|>' }}\n", "{% endif %}\n", "{% endfor %}\n" ] } ] }, { "cell_type": "markdown", "metadata": { "id": "leGEqxURAL4r" }, "source": [ "#### Chat Template - Schema Linking" ] }, { "cell_type": "code", "source": [ "# tokenizer.chat_template = \"\"\"\n", "# {% if messages[0]['role'] == 'system' %}\n", "# {% set loop_messages = messages[1:] %}\n", "# {% set system_message = messages[0]['content'] %}\n", "# {% else %}\n", "# {% set loop_messages = messages %}\n", "# {% set system_message = 'You are an SQL expert tasked with answering questions about an SQLite database./n/Your responses must always consist of valid SQL code and only that.' %}\n", "# {% endif %}\n", "# {{ '# <|system|>/n/' + system_message + '/n//n/' }}\n", "# {% if messages|selectattr(\"role\", \"equalto\", \"example\")|list %}\n", "# Below are some examples of question and their corresponding SQL queries:/n//n/\n", "# {% else %}\n", "# /n/\n", "# {% endif %}\n", "# {% for message in loop_messages %}\n", "# {% if message['role'] == 'example' %}\n", "# {{ message['content'] }}/n//n/\n", "# {% elif message['role'] == 'schema' %}\n", "# # <|schema|>/n/The query will run on a database with the following schema:/n/{{ message['content'] }}/n//n/\n", "# {% elif message['role'] == 'user' %}\n", "# # <|user|>/n/[QUESTION]{{ message['content'] }}[/QUESTION]/n//n/\n", "# {% elif message['role'] == 'assistant' %}\n", "# # <|assistant|>/n/[SQL]{{ message['content'] }}[/SQL]\n", "# {% endif %}\n", "# {% endfor %}\n", "# {% if add_generation_prompt %}\n", "# # <|assistant|>/n/[SQL]\n", "# {% endif %}\n", "# \"\"\".replace(\"\\n\",\"\").replace(\" \", \"\").replace(\"/n/\", \"\\n\")\n" ], "metadata": { "id": "Jiq-S5I1KOBr" }, "execution_count": null, "outputs": [] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "aHGTJNq2AmUs" }, "outputs": [], "source": [ "def to_sql(query):\n", " return sqlparse.format(query, reindent=True, keyword_case='upper')\n", "\n", "def replace_last_sql_content(text, new_content):\n", " matches = list(re.finditer(r'```sql (.*?)```', text, re.DOTALL))\n", " if not matches:\n", " return text\n", "\n", " last_match = matches[-1]\n", "\n", " start, end = last_match.span(1)\n", " return text[:start] + new_content + text[end:]\n", "\n", "# def apply_template(row, tokenizer=tokenizer, n_examplea=0):\n", "# question = row['question_en']\n", "# schema = row['schema_llm']\n", "# sql = to_sql(row['query_llm'])\n", "\n", "# system = \"Given a user question and the schema of a database, your task is to generate an SQL query that accurately answers the question based on the provided schema.\"\n", "\n", "\n", "# chat = [{'role': 'user', 'content': f\"# System:\\n{system}\\n\\n# Schema:\\n```sql\\n{schema}\\n```\\n\\n# Question: {question}\"},\n", "# {'role': 'assistant', 'content': f\"```sql\\n{sql}\\n```\"}\n", "# ]\n", "\n", "# row['text'] = tokenizer.apply_chat_template(chat, tokenize=False, add_generation_prompt=False)\n", "\n", "# return row\n", "\n", "def apply_template(row, tokenizer=tokenizer, n_examplea=0):\n", " question = row['question_en']\n", " schema = row['schema_llm_t']\n", " schema_linking = to_sql(row['selector_correct'])\n", "\n", " system = \"Given a user question and the schema of a database, your task is to generate an JSON with the the names of tables and columns of the schema that the question is referring to.\"\n", "\n", " chat = [{'role': 'user', 'content': f\"# System:\\n{system}\\n\\n# Schema:\\n```sql\\n{schema}\\n```\\n\\n# Question: {question}\"},\n", " {'role': 'assistant', 'content': f\"```json\\n{schema_linking}\\n```\"}\n", " ]\n", "\n", " row['text'] = tokenizer.apply_chat_template(chat, tokenize=False, add_generation_prompt=False)\n", "\n", " return row" ] }, { "cell_type": "code", "source": [ "# spider_chain = json.load(open(\"/content/drive/Shareddrives/LLMs/Datasets/spider/spider_chain.json\", \"r\"))\n", "# bird_chain = json.load(open(\"/content/drive/Shareddrives/LLMs/Datasets/bird/bird_chain.json\", \"r\"))" ], "metadata": { "id": "QlA1kMj83Vql" }, "execution_count": null, "outputs": [] }, { "cell_type": "code", "source": [ "# df['chain'] = spider_chain + bird_chain" ], "metadata": { "id": "pCcD4TiC3rpb" }, "execution_count": null, "outputs": [] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "yvcMZAL3E3TB" }, "outputs": [], "source": [ "df = df.apply(apply_template, axis=1)" ] }, { "cell_type": "code", "source": [ "df.head()" ], "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 539 }, "id": "rQ0KUl2w58K9", "outputId": "664b4e86-fe9f-45ee-bd2b-ea021a31b08f" }, "execution_count": null, "outputs": [ { "output_type": "execute_result", "data": { "text/plain": [ " db_id schema \\\n", "0 department_management \\nCREATE TABLE department (\\n\\t\"Department_ID\"... \n", "1 department_management \\nCREATE TABLE department (\\n\\t\"Department_ID\"... \n", "2 department_management \\nCREATE TABLE department (\\n\\t\"Department_ID\"... \n", "3 department_management \\nCREATE TABLE department (\\n\\t\"Department_ID\"... \n", "4 department_management \\nCREATE TABLE department (\\n\\t\"Department_ID\"... \n", "\n", " question_en hardness \\\n", "0 How many heads of the departments are older th... easy \n", "1 List the name, born state and age of the heads... medium \n", "2 List the creation year, name and budget of eac... medium \n", "3 What are the maximum and minimum budget of the... medium \n", "4 What is the average number of employees of the... easy \n", "\n", " schema_llm \\\n", "0 Table department (\\n department.Department_... \n", "1 Table department (\\n department.Department_... \n", "2 Table department (\\n department.Department_... \n", "3 Table department (\\n department.Department_... \n", "4 Table department (\\n department.Department_... \n", "\n", " query_llm \\\n", "0 SELECT count(*) FROM head WHERE age > 56 \n", "1 SELECT name , born_state , age FROM head ORD... \n", "2 SELECT creation , name , budget_in_billions ... \n", "3 SELECT max(budget_in_billions) , min(budget_i... \n", "4 SELECT avg(num_employees) FROM department WHER... \n", "\n", " selector \\\n", "0 {\\n 'head': ['head_id', 'age']\\n} \n", "1 {\\n 'head': ['head_id', 'name', 'born_state',... \n", "2 {\\n 'department': ['department_id', 'name', '... \n", "3 {\\n 'department': ['department_id', 'budget_i... \n", "4 {\\n 'department': ['department_id', 'ranking'... \n", "\n", " schema_llm_ct \\\n", "0 CREATE TABLE department (\\n department.Depa... \n", "1 CREATE TABLE department (\\n department.Depa... \n", "2 CREATE TABLE department (\\n department.Depa... \n", "3 CREATE TABLE department (\\n department.Depa... \n", "4 CREATE TABLE department (\\n department.Depa... \n", "\n", " schema_llm_columns_min \\\n", "0 Table department (\\n department.department_... \n", "1 Table department (\\n department.department_... \n", "2 Table department (\\n department.department_... \n", "3 Table department (\\n department.department_... \n", "4 Table department (\\n department.department_... \n", "\n", " schema_llm_columns_min_ct \\\n", "0 CREATE TABLE department (\\n department.depa... \n", "1 CREATE TABLE department (\\n department.depa... \n", "2 CREATE TABLE department (\\n department.depa... \n", "3 CREATE TABLE department (\\n department.depa... \n", "4 CREATE TABLE department (\\n department.depa... \n", "\n", " schema_llm_all_min \\\n", "0 Table department (\\n department.department_... \n", "1 Table department (\\n department.department_... \n", "2 Table department (\\n department.department_... \n", "3 Table department (\\n department.department_... \n", "4 Table department (\\n department.department_... \n", "\n", " schema_llm_all_min_ct \\\n", "0 CREATE TABLE department (\\n department.depa... \n", "1 CREATE TABLE department (\\n department.depa... \n", "2 CREATE TABLE department (\\n department.depa... \n", "3 CREATE TABLE department (\\n department.depa... \n", "4 CREATE TABLE department (\\n department.depa... \n", "\n", " schema_dict \\\n", "0 {\\n 'department': ['Department_ID', 'Name', '... \n", "1 {\\n 'department': ['Department_ID', 'Name', '... \n", "2 {\\n 'department': ['Department_ID', 'Name', '... \n", "3 {\\n 'department': ['Department_ID', 'Name', '... \n", "4 {\\n 'department': ['Department_ID', 'Name', '... \n", "\n", " selector_correct \\\n", "0 {\\n 'head': ['head_ID', 'age']\\n} \n", "1 {\\n 'head': ['head_ID', 'name', 'born_state',... \n", "2 {\\n 'department': ['Department_ID', 'Name', '... \n", "3 {\\n 'department': ['Department_ID', 'Budget_i... \n", "4 {\\n 'department': ['Department_ID', 'Ranking'... \n", "\n", " schema_llm_t \\\n", "0 TABLE department (\\n department.Department_... \n", "1 TABLE department (\\n department.Department_... \n", "2 TABLE department (\\n department.Department_... \n", "3 TABLE department (\\n department.Department_... \n", "4 TABLE department (\\n department.Department_... \n", "\n", " schema_llm_columns_min_t \\\n", "0 TABLE department (\\n department.department_... \n", "1 TABLE department (\\n department.department_... \n", "2 TABLE department (\\n department.department_... \n", "3 TABLE department (\\n department.department_... \n", "4 TABLE department (\\n department.department_... \n", "\n", " schema_llm_all_min_t \\\n", "0 TABLE department (\\n department.department_... \n", "1 TABLE department (\\n department.department_... \n", "2 TABLE department (\\n department.department_... \n", "3 TABLE department (\\n department.department_... \n", "4 TABLE department (\\n department.department_... \n", "\n", " text \n", "0 <|begin_of_text|><|start_header_id|>system<|en... \n", "1 <|begin_of_text|><|start_header_id|>system<|en... \n", "2 <|begin_of_text|><|start_header_id|>system<|en... \n", "3 <|begin_of_text|><|start_header_id|>system<|en... \n", "4 <|begin_of_text|><|start_header_id|>system<|en... " ], "text/html": [ "\n", "
\n", "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
db_idschemaquestion_enhardnessschema_llmquery_llmselectorschema_llm_ctschema_llm_columns_minschema_llm_columns_min_ctschema_llm_all_minschema_llm_all_min_ctschema_dictselector_correctschema_llm_tschema_llm_columns_min_tschema_llm_all_min_ttext
0department_management\\nCREATE TABLE department (\\n\\t\"Department_ID\"...How many heads of the departments are older th...easyTable department (\\n department.Department_...SELECT count(*) FROM head WHERE age > 56{\\n 'head': ['head_id', 'age']\\n}CREATE TABLE department (\\n department.Depa...Table department (\\n department.department_...CREATE TABLE department (\\n department.depa...Table department (\\n department.department_...CREATE TABLE department (\\n department.depa...{\\n 'department': ['Department_ID', 'Name', '...{\\n 'head': ['head_ID', 'age']\\n}TABLE department (\\n department.Department_...TABLE department (\\n department.department_...TABLE department (\\n department.department_...<|begin_of_text|><|start_header_id|>system<|en...
1department_management\\nCREATE TABLE department (\\n\\t\"Department_ID\"...List the name, born state and age of the heads...mediumTable department (\\n department.Department_...SELECT name , born_state , age FROM head ORD...{\\n 'head': ['head_id', 'name', 'born_state',...CREATE TABLE department (\\n department.Depa...Table department (\\n department.department_...CREATE TABLE department (\\n department.depa...Table department (\\n department.department_...CREATE TABLE department (\\n department.depa...{\\n 'department': ['Department_ID', 'Name', '...{\\n 'head': ['head_ID', 'name', 'born_state',...TABLE department (\\n department.Department_...TABLE department (\\n department.department_...TABLE department (\\n department.department_...<|begin_of_text|><|start_header_id|>system<|en...
2department_management\\nCREATE TABLE department (\\n\\t\"Department_ID\"...List the creation year, name and budget of eac...mediumTable department (\\n department.Department_...SELECT creation , name , budget_in_billions ...{\\n 'department': ['department_id', 'name', '...CREATE TABLE department (\\n department.Depa...Table department (\\n department.department_...CREATE TABLE department (\\n department.depa...Table department (\\n department.department_...CREATE TABLE department (\\n department.depa...{\\n 'department': ['Department_ID', 'Name', '...{\\n 'department': ['Department_ID', 'Name', '...TABLE department (\\n department.Department_...TABLE department (\\n department.department_...TABLE department (\\n department.department_...<|begin_of_text|><|start_header_id|>system<|en...
3department_management\\nCREATE TABLE department (\\n\\t\"Department_ID\"...What are the maximum and minimum budget of the...mediumTable department (\\n department.Department_...SELECT max(budget_in_billions) , min(budget_i...{\\n 'department': ['department_id', 'budget_i...CREATE TABLE department (\\n department.Depa...Table department (\\n department.department_...CREATE TABLE department (\\n department.depa...Table department (\\n department.department_...CREATE TABLE department (\\n department.depa...{\\n 'department': ['Department_ID', 'Name', '...{\\n 'department': ['Department_ID', 'Budget_i...TABLE department (\\n department.Department_...TABLE department (\\n department.department_...TABLE department (\\n department.department_...<|begin_of_text|><|start_header_id|>system<|en...
4department_management\\nCREATE TABLE department (\\n\\t\"Department_ID\"...What is the average number of employees of the...easyTable department (\\n department.Department_...SELECT avg(num_employees) FROM department WHER...{\\n 'department': ['department_id', 'ranking'...CREATE TABLE department (\\n department.Depa...Table department (\\n department.department_...CREATE TABLE department (\\n department.depa...Table department (\\n department.department_...CREATE TABLE department (\\n department.depa...{\\n 'department': ['Department_ID', 'Name', '...{\\n 'department': ['Department_ID', 'Ranking'...TABLE department (\\n department.Department_...TABLE department (\\n department.department_...TABLE department (\\n department.department_...<|begin_of_text|><|start_header_id|>system<|en...
\n", "
\n", "
\n", "\n", "
\n", " \n", "\n", " \n", "\n", " \n", "
\n", "\n", "\n", "
\n", " \n", "\n", "\n", "\n", " \n", "
\n", "\n", "
\n", "
\n" ], "application/vnd.google.colaboratory.intrinsic+json": { "type": "dataframe", "variable_name": "df", "summary": "{\n \"name\": \"df\",\n \"rows\": 8656,\n \"fields\": [\n {\n \"column\": \"db_id\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 146,\n \"samples\": [\n \"gas_company\",\n \"customers_campaigns_ecommerce\",\n \"performance_attendance\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"schema\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 146,\n \"samples\": [\n \"\\nCREATE TABLE company (\\n\\t\\\"Company_ID\\\" INTEGER, \\n\\t\\\"Rank\\\" INTEGER, \\n\\t\\\"Company\\\" TEXT, \\n\\t\\\"Headquarters\\\" TEXT, \\n\\t\\\"Main_Industry\\\" TEXT, \\n\\t\\\"Sales_billion\\\" REAL, \\n\\t\\\"Profits_billion\\\" REAL, \\n\\t\\\"Assets_billion\\\" REAL, \\n\\t\\\"Market_Value\\\" REAL, \\n\\tPRIMARY KEY (\\\"Company_ID\\\")\\n)\\n\\n\\nCREATE TABLE gas_station (\\n\\t\\\"Station_ID\\\" INTEGER, \\n\\t\\\"Open_Year\\\" INTEGER, \\n\\t\\\"Location\\\" TEXT, \\n\\t\\\"Manager_Name\\\" TEXT, \\n\\t\\\"Vice_Manager_Name\\\" TEXT, \\n\\t\\\"Representative_Name\\\" TEXT, \\n\\tPRIMARY KEY (\\\"Station_ID\\\")\\n)\\n\\n\\nCREATE TABLE station_company (\\n\\t\\\"Station_ID\\\" INTEGER, \\n\\t\\\"Company_ID\\\" INTEGER, \\n\\t\\\"Rank_of_the_Year\\\" INTEGER, \\n\\tPRIMARY KEY (\\\"Station_ID\\\", \\\"Company_ID\\\"), \\n\\tFOREIGN KEY(\\\"Company_ID\\\") REFERENCES company (\\\"Company_ID\\\"), \\n\\tFOREIGN KEY(\\\"Station_ID\\\") REFERENCES gas_station (\\\"Station_ID\\\")\\n)\",\n \"\\nCREATE TABLE \\\"Customer_Addresses\\\" (\\n\\tcustomer_id INTEGER NOT NULL, \\n\\tpremise_id INTEGER NOT NULL, \\n\\tdate_address_from DATETIME NOT NULL, \\n\\taddress_type_code VARCHAR(15) NOT NULL, \\n\\tdate_address_to DATETIME, \\n\\tFOREIGN KEY(customer_id) REFERENCES \\\"Customers\\\" (customer_id), \\n\\tFOREIGN KEY(premise_id) REFERENCES \\\"Premises\\\" (premise_id)\\n)\\n\\n\\nCREATE TABLE \\\"Customer_Orders\\\" (\\n\\torder_id INTEGER, \\n\\tcustomer_id INTEGER NOT NULL, \\n\\torder_status_code VARCHAR(15) NOT NULL, \\n\\tshipping_method_code VARCHAR(15) NOT NULL, \\n\\torder_placed_datetime DATETIME NOT NULL, \\n\\torder_delivered_datetime DATETIME, \\n\\torder_shipping_charges VARCHAR(255), \\n\\tPRIMARY KEY (order_id), \\n\\tFOREIGN KEY(customer_id) REFERENCES \\\"Customers\\\" (customer_id)\\n)\\n\\n\\nCREATE TABLE \\\"Customers\\\" (\\n\\tcustomer_id INTEGER, \\n\\tpayment_method VARCHAR(15) NOT NULL, \\n\\tcustomer_name VARCHAR(80), \\n\\tcustomer_phone VARCHAR(80), \\n\\tcustomer_email VARCHAR(80), \\n\\tcustomer_address VARCHAR(255), \\n\\tcustomer_login VARCHAR(80), \\n\\tcustomer_password VARCHAR(10), \\n\\tPRIMARY KEY (customer_id)\\n)\\n\\n\\nCREATE TABLE \\\"Mailshot_Campaigns\\\" (\\n\\tmailshot_id INTEGER, \\n\\tproduct_category VARCHAR(15), \\n\\tmailshot_name VARCHAR(80), \\n\\tmailshot_start_date DATETIME, \\n\\tmailshot_end_date DATETIME, \\n\\tPRIMARY KEY (mailshot_id)\\n)\\n\\n\\nCREATE TABLE \\\"Mailshot_Customers\\\" (\\n\\tmailshot_id INTEGER NOT NULL, \\n\\tcustomer_id INTEGER NOT NULL, \\n\\toutcome_code VARCHAR(15) NOT NULL, \\n\\tmailshot_customer_date DATETIME, \\n\\tFOREIGN KEY(mailshot_id) REFERENCES \\\"Mailshot_Campaigns\\\" (mailshot_id), \\n\\tFOREIGN KEY(customer_id) REFERENCES \\\"Customers\\\" (customer_id)\\n)\\n\\n\\nCREATE TABLE \\\"Order_Items\\\" (\\n\\titem_id INTEGER NOT NULL, \\n\\torder_item_status_code VARCHAR(15) NOT NULL, \\n\\torder_id INTEGER NOT NULL, \\n\\tproduct_id INTEGER NOT NULL, \\n\\titem_status_code VARCHAR(15), \\n\\titem_delivered_datetime DATETIME, \\n\\titem_order_quantity VARCHAR(80), \\n\\tFOREIGN KEY(order_id) REFERENCES \\\"Customer_Orders\\\" (order_id), \\n\\tFOREIGN KEY(product_id) REFERENCES \\\"Products\\\" (product_id)\\n)\\n\\n\\nCREATE TABLE \\\"Premises\\\" (\\n\\tpremise_id INTEGER, \\n\\tpremises_type VARCHAR(15) NOT NULL, \\n\\tpremise_details VARCHAR(255), \\n\\tPRIMARY KEY (premise_id)\\n)\\n\\n\\nCREATE TABLE \\\"Products\\\" (\\n\\tproduct_id INTEGER, \\n\\tproduct_category VARCHAR(15) NOT NULL, \\n\\tproduct_name VARCHAR(80), \\n\\tPRIMARY KEY (product_id)\\n)\",\n \"\\nCREATE TABLE member (\\n\\t\\\"Member_ID\\\" TEXT, \\n\\t\\\"Name\\\" TEXT, \\n\\t\\\"Nationality\\\" TEXT, \\n\\t\\\"Role\\\" TEXT, \\n\\tPRIMARY KEY (\\\"Member_ID\\\")\\n)\\n\\n\\nCREATE TABLE member_attendance (\\n\\t\\\"Member_ID\\\" INTEGER, \\n\\t\\\"Performance_ID\\\" INTEGER, \\n\\t\\\"Num_of_Pieces\\\" INTEGER, \\n\\tPRIMARY KEY (\\\"Member_ID\\\", \\\"Performance_ID\\\"), \\n\\tFOREIGN KEY(\\\"Performance_ID\\\") REFERENCES performance (\\\"Performance_ID\\\"), \\n\\tFOREIGN KEY(\\\"Member_ID\\\") REFERENCES member (\\\"Member_ID\\\")\\n)\\n\\n\\nCREATE TABLE performance (\\n\\t\\\"Performance_ID\\\" REAL, \\n\\t\\\"Date\\\" TEXT, \\n\\t\\\"Host\\\" TEXT, \\n\\t\\\"Location\\\" TEXT, \\n\\t\\\"Attendance\\\" INTEGER, \\n\\tPRIMARY KEY (\\\"Performance_ID\\\")\\n)\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"question_en\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 8618,\n \"samples\": [\n \"What are the names of courses without prerequisites?\",\n \"Return the investor who have invested in the greatest number of entrepreneurs.\",\n \"Which customers have ever canceled the purchase of the product \\\"food\\\" (the item status is \\\"Cancel\\\")?\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"hardness\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 4,\n \"samples\": [\n \"medium\",\n \"extra\",\n \"easy\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"schema_llm\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 146,\n \"samples\": [\n \"Table company (\\n company.Company_ID (INT),\\n company.Rank (INT),\\n company.Company (TEXT),\\n company.Headquarters (TEXT),\\n company.Main_Industry (TEXT),\\n company.Sales_billion (REAL),\\n company.Profits_billion (REAL),\\n company.Assets_billion (REAL),\\n company.Market_Value (REAL),\\n)\\n\\nTable gas_station (\\n gas_station.Station_ID (INT),\\n gas_station.Open_Year (INT),\\n gas_station.Location (TEXT),\\n gas_station.Manager_Name (TEXT),\\n gas_station.Vice_Manager_Name (TEXT),\\n gas_station.Representative_Name (TEXT),\\n)\\n\\nTable station_company (\\n station_company.Station_ID (INT),\\n station_company.Company_ID (INT),\\n station_company.Rank_of_the_Year (INT),\\n)\\n\\n\\nPossible JOINs:\\n\\nstation_company.Station_ID = gas_station.Station_ID\\nstation_company.Company_ID = company.Company_ID\\n\",\n \"Table Customer_Addresses (\\n Customer_Addresses.customer_id (INTEGER),\\n Customer_Addresses.premise_id (INTEGER),\\n Customer_Addresses.date_address_from (DATETIME),\\n Customer_Addresses.address_type_code (VARCHAR(15)),\\n Customer_Addresses.date_address_to (DATETIME),\\n)\\n\\nTable Customer_Orders (\\n Customer_Orders.order_id (INTEGER),\\n Customer_Orders.customer_id (INTEGER),\\n Customer_Orders.order_status_code (VARCHAR(15)),\\n Customer_Orders.shipping_method_code (VARCHAR(15)),\\n Customer_Orders.order_placed_datetime (DATETIME),\\n Customer_Orders.order_delivered_datetime (DATETIME),\\n Customer_Orders.order_shipping_charges (VARCHAR(255)),\\n)\\n\\nTable Customers (\\n Customers.customer_id (INTEGER),\\n Customers.payment_method (VARCHAR(15)),\\n Customers.customer_name (VARCHAR(80)),\\n Customers.customer_phone (VARCHAR(80)),\\n Customers.customer_email (VARCHAR(80)),\\n Customers.customer_address (VARCHAR(255)),\\n Customers.customer_login (VARCHAR(80)),\\n Customers.customer_password (VARCHAR(10)),\\n)\\n\\nTable Mailshot_Campaigns (\\n Mailshot_Campaigns.mailshot_id (INTEGER),\\n Mailshot_Campaigns.product_category (VARCHAR(15)),\\n Mailshot_Campaigns.mailshot_name (VARCHAR(80)),\\n Mailshot_Campaigns.mailshot_start_date (DATETIME),\\n Mailshot_Campaigns.mailshot_end_date (DATETIME),\\n)\\n\\nTable Mailshot_Customers (\\n Mailshot_Customers.mailshot_id (INTEGER),\\n Mailshot_Customers.customer_id (INTEGER),\\n Mailshot_Customers.outcome_code (VARCHAR(15)),\\n Mailshot_Customers.mailshot_customer_date (DATETIME),\\n)\\n\\nTable Order_Items (\\n Order_Items.item_id (INTEGER),\\n Order_Items.order_item_status_code (VARCHAR(15)),\\n Order_Items.order_id (INTEGER),\\n Order_Items.product_id (INTEGER),\\n Order_Items.item_status_code (VARCHAR(15)),\\n Order_Items.item_delivered_datetime (DATETIME),\\n Order_Items.item_order_quantity (VARCHAR(80)),\\n)\\n\\nTable Premises (\\n Premises.premise_id (INTEGER),\\n Premises.premises_type (VARCHAR(15)),\\n Premises.premise_details (VARCHAR(255)),\\n)\\n\\nTable Products (\\n Products.product_id (INTEGER),\\n Products.product_category (VARCHAR(15)),\\n Products.product_name (VARCHAR(80)),\\n)\\n\\n\\nPossible JOINs:\\n\\nCustomer_Addresses.customer_id = Customers.customer_id\\nCustomer_Addresses.premise_id = Premises.premise_id\\n\\nCustomer_Orders.customer_id = Customers.customer_id\\n\\nMailshot_Customers.mailshot_id = Mailshot_Campaigns.mailshot_id\\nMailshot_Customers.customer_id = Customers.customer_id\\n\\nOrder_Items.order_id = Customer_Orders.order_id\\nOrder_Items.product_id = Products.product_id\\n\",\n \"Table member (\\n member.Member_ID (TEXT),\\n member.Name (TEXT),\\n member.Nationality (TEXT),\\n member.Role (TEXT),\\n)\\n\\nTable member_attendance (\\n member_attendance.Member_ID (INT),\\n member_attendance.Performance_ID (INT),\\n member_attendance.Num_of_Pieces (INT),\\n)\\n\\nTable performance (\\n performance.Performance_ID (REAL),\\n performance.Date (TEXT),\\n performance.Host (TEXT),\\n performance.Location (TEXT),\\n performance.Attendance (INT),\\n)\\n\\n\\nPossible JOINs:\\n\\nmember_attendance.Member_ID = member.Member_ID\\nmember_attendance.Performance_ID = performance.Performance_ID\\n\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"query_llm\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 4712,\n \"samples\": [\n \"SELECT manager_name FROM gas_station WHERE open_year > 2000 GROUP BY manager_name ORDER BY count(*) DESC LIMIT 1\",\n \"SELECT lname FROM authors WHERE fname = \\\"Amal\\\"\",\n \"SELECT * FROM swimmer\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"selector\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 3358,\n \"samples\": [\n \"{\\n 'rating': ['mid', 'stars'],\\n 'movie': ['mid', 'title', 'director']\\n}\",\n \"{\\n 'player': ['birth_country', 'bats']\\n}\",\n \"{\\n 'bank': ['branch_id', 'city']\\n}\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"schema_llm_ct\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 147,\n \"samples\": [\n \"CREATE TABLE country (\\n country.Country_ID (INT),\\n country.Name (TEXT),\\n country.Population (INT),\\n country.Area (INT),\\n country.Languages (TEXT),\\n)\\n\\nCREATE TABLE roller_coaster (\\n roller_coaster.Roller_Coaster_ID (INT),\\n roller_coaster.Name (TEXT),\\n roller_coaster.Park (TEXT),\\n roller_coaster.Country_ID (INT),\\n roller_coaster.Length (REAL),\\n roller_coaster.Height (REAL),\\n roller_coaster.Speed (TEXT),\\n roller_coaster.Opened (TEXT),\\n roller_coaster.Status (TEXT),\\n)\\n\\n\\nPossible JOINs:\\n\\nroller_coaster.Country_ID = country.Country_ID\\n\",\n \"CREATE TABLE circuits (\\n circuits.circuitId (INTEGER),\\n circuits.circuitRef (TEXT),\\n circuits.name (TEXT),\\n circuits.location (TEXT),\\n circuits.country (TEXT),\\n circuits.lat (REAL),\\n circuits.lng (REAL),\\n circuits.alt (TEXT),\\n circuits.url (TEXT),\\n)\\n\\nCREATE TABLE constructorResults (\\n constructorResults.constructorResultsId (INTEGER),\\n constructorResults.raceId (INTEGER),\\n constructorResults.constructorId (INTEGER),\\n constructorResults.points (REAL),\\n constructorResults.status (TEXT),\\n)\\n\\nCREATE TABLE constructorStandings (\\n constructorStandings.constructorStandingsId (INTEGER),\\n constructorStandings.raceId (INTEGER),\\n constructorStandings.constructorId (INTEGER),\\n constructorStandings.points (REAL),\\n constructorStandings.position (INTEGER),\\n constructorStandings.positionText (TEXT),\\n constructorStandings.wins (INTEGER),\\n)\\n\\nCREATE TABLE constructors (\\n constructors.constructorId (INTEGER),\\n constructors.constructorRef (TEXT),\\n constructors.name (TEXT),\\n constructors.nationality (TEXT),\\n constructors.url (TEXT),\\n)\\n\\nCREATE TABLE driverStandings (\\n driverStandings.driverStandingsId (INTEGER),\\n driverStandings.raceId (INTEGER),\\n driverStandings.driverId (INTEGER),\\n driverStandings.points (REAL),\\n driverStandings.position (INTEGER),\\n driverStandings.positionText (TEXT),\\n driverStandings.wins (INTEGER),\\n)\\n\\nCREATE TABLE drivers (\\n drivers.driverId (INTEGER),\\n drivers.driverRef (TEXT),\\n drivers.number (TEXT),\\n drivers.code (TEXT),\\n drivers.forename (TEXT),\\n drivers.surname (TEXT),\\n drivers.dob (TEXT),\\n drivers.nationality (TEXT),\\n drivers.url (TEXT),\\n)\\n\\nCREATE TABLE lapTimes (\\n lapTimes.raceId (INTEGER),\\n lapTimes.driverId (INTEGER),\\n lapTimes.lap (INTEGER),\\n lapTimes.position (INTEGER),\\n lapTimes.time (TEXT),\\n lapTimes.milliseconds (INTEGER),\\n)\\n\\nCREATE TABLE pitStops (\\n pitStops.raceId (INTEGER),\\n pitStops.driverId (INTEGER),\\n pitStops.stop (INTEGER),\\n pitStops.lap (INTEGER),\\n pitStops.time (TEXT),\\n pitStops.duration (TEXT),\\n pitStops.milliseconds (INTEGER),\\n)\\n\\nCREATE TABLE qualifying (\\n qualifying.qualifyId (INTEGER),\\n qualifying.raceId (INTEGER),\\n qualifying.driverId (INTEGER),\\n qualifying.constructorId (INTEGER),\\n qualifying.number (INTEGER),\\n qualifying.position (INTEGER),\\n qualifying.q1 (TEXT),\\n qualifying.q2 (TEXT),\\n qualifying.q3 (TEXT),\\n)\\n\\nCREATE TABLE races (\\n races.raceId (INTEGER),\\n races.year (INTEGER),\\n races.round (INTEGER),\\n races.circuitId (INTEGER),\\n races.name (TEXT),\\n races.date (TEXT),\\n races.time (TEXT),\\n races.url (TEXT),\\n)\\n\\nCREATE TABLE results (\\n results.resultId (INTEGER),\\n results.raceId (INTEGER),\\n results.driverId (INTEGER),\\n results.constructorId (INTEGER),\\n results.number (INTEGER),\\n results.grid (INTEGER),\\n results.position (TEXT),\\n results.positionText (TEXT),\\n results.positionOrder (INTEGER),\\n results.points (REAL),\\n results.laps (TEXT),\\n results.time (TEXT),\\n results.milliseconds (TEXT),\\n results.fastestLap (TEXT),\\n results.rank (TEXT),\\n results.fastestLapTime (TEXT),\\n results.fastestLapSpeed (TEXT),\\n results.statusId (INTEGER),\\n)\\n\\nCREATE TABLE seasons (\\n seasons.year (INTEGER),\\n seasons.url (TEXT),\\n)\\n\\nCREATE TABLE status (\\n status.statusId (INTEGER),\\n status.status (TEXT),\\n)\\n\\n\\nPossible JOINs:\\n\\nconstructorResults.raceId = races.raceId\\nconstructorResults.constructorId = constructors.constructorId\\n\\nconstructorStandings.raceId = races.raceId\\nconstructorStandings.constructorId = constructors.constructorId\\n\\ndriverStandings.raceId = races.raceId\\ndriverStandings.driverId = drivers.driverId\\n\\nlapTimes.raceId = races.raceId\\nlapTimes.driverId = drivers.driverId\\n\\npitStops.raceId = races.raceId\\npitStops.driverId = drivers.driverId\\n\\nqualifying.raceId = races.raceId\\nqualifying.driverId = drivers.driverId\\nqualifying.constructorId = constructors.constructorId\\n\\nraces.circuitId = circuits.circuitId\\n\\nresults.raceId = races.raceId\\nresults.driverId = drivers.driverId\\nresults.constructorId = constructors.constructorId\\n\",\n \"CREATE TABLE Customers (\\n Customers.customer_id (INTEGER),\\n Customers.customer_name (VARCHAR(80)),\\n Customers.customer_details (VARCHAR(255)),\\n)\\n\\nCREATE TABLE Invoices (\\n Invoices.invoice_number (INTEGER),\\n Invoices.invoice_date (DATETIME),\\n Invoices.invoice_details (VARCHAR(255)),\\n)\\n\\nCREATE TABLE Order_Items (\\n Order_Items.order_item_id (INTEGER),\\n Order_Items.product_id (INTEGER),\\n Order_Items.order_id (INTEGER),\\n Order_Items.order_item_status (VARCHAR(10)),\\n Order_Items.order_item_details (VARCHAR(255)),\\n)\\n\\nCREATE TABLE Orders (\\n Orders.order_id (INTEGER),\\n Orders.customer_id (INTEGER),\\n Orders.order_status (VARCHAR(10)),\\n Orders.date_order_placed (DATETIME),\\n Orders.order_details (VARCHAR(255)),\\n)\\n\\nCREATE TABLE Products (\\n Products.product_id (INTEGER),\\n Products.product_name (VARCHAR(80)),\\n Products.product_details (VARCHAR(255)),\\n)\\n\\nCREATE TABLE Shipment_Items (\\n Shipment_Items.shipment_id (INTEGER),\\n Shipment_Items.order_item_id (INTEGER),\\n)\\n\\nCREATE TABLE Shipments (\\n Shipments.shipment_id (INTEGER),\\n Shipments.order_id (INTEGER),\\n Shipments.invoice_number (INTEGER),\\n Shipments.shipment_tracking_number (VARCHAR(80)),\\n Shipments.shipment_date (DATETIME),\\n Shipments.other_shipment_details (VARCHAR(255)),\\n)\\n\\n\\nPossible JOINs:\\n\\nOrder_Items.product_id = Products.product_id\\nOrder_Items.order_id = Orders.order_id\\n\\nOrders.customer_id = Customers.customer_id\\n\\nShipment_Items.shipment_id = Shipments.shipment_id\\nShipment_Items.order_item_id = Order_Items.order_item_id\\n\\nShipments.order_id = Orders.order_id\\nShipments.invoice_number = Invoices.invoice_number\\n\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"schema_llm_columns_min\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 147,\n \"samples\": [\n \"Table country (\\n country.country_id (INT),\\n country.name (TEXT),\\n country.population (INT),\\n country.area (INT),\\n country.languages (TEXT),\\n)\\n\\nTable roller_coaster (\\n roller_coaster.roller_coaster_id (INT),\\n roller_coaster.name (TEXT),\\n roller_coaster.park (TEXT),\\n roller_coaster.country_id (INT),\\n roller_coaster.length (REAL),\\n roller_coaster.height (REAL),\\n roller_coaster.speed (TEXT),\\n roller_coaster.opened (TEXT),\\n roller_coaster.status (TEXT),\\n)\\n\\n\\nPossible JOINs:\\n\\nroller_coaster.country_id = country.country_id\\n\",\n \"Table circuits (\\n circuits.circuitid (INTEGER),\\n circuits.circuitref (TEXT),\\n circuits.name (TEXT),\\n circuits.location (TEXT),\\n circuits.country (TEXT),\\n circuits.lat (REAL),\\n circuits.lng (REAL),\\n circuits.alt (TEXT),\\n circuits.url (TEXT),\\n)\\n\\nTable constructorResults (\\n constructorResults.constructorresultsid (INTEGER),\\n constructorResults.raceid (INTEGER),\\n constructorResults.constructorid (INTEGER),\\n constructorResults.points (REAL),\\n constructorResults.status (TEXT),\\n)\\n\\nTable constructorStandings (\\n constructorStandings.constructorstandingsid (INTEGER),\\n constructorStandings.raceid (INTEGER),\\n constructorStandings.constructorid (INTEGER),\\n constructorStandings.points (REAL),\\n constructorStandings.position (INTEGER),\\n constructorStandings.positiontext (TEXT),\\n constructorStandings.wins (INTEGER),\\n)\\n\\nTable constructors (\\n constructors.constructorid (INTEGER),\\n constructors.constructorref (TEXT),\\n constructors.name (TEXT),\\n constructors.nationality (TEXT),\\n constructors.url (TEXT),\\n)\\n\\nTable driverStandings (\\n driverStandings.driverstandingsid (INTEGER),\\n driverStandings.raceid (INTEGER),\\n driverStandings.driverid (INTEGER),\\n driverStandings.points (REAL),\\n driverStandings.position (INTEGER),\\n driverStandings.positiontext (TEXT),\\n driverStandings.wins (INTEGER),\\n)\\n\\nTable drivers (\\n drivers.driverid (INTEGER),\\n drivers.driverref (TEXT),\\n drivers.number (TEXT),\\n drivers.code (TEXT),\\n drivers.forename (TEXT),\\n drivers.surname (TEXT),\\n drivers.dob (TEXT),\\n drivers.nationality (TEXT),\\n drivers.url (TEXT),\\n)\\n\\nTable lapTimes (\\n lapTimes.raceid (INTEGER),\\n lapTimes.driverid (INTEGER),\\n lapTimes.lap (INTEGER),\\n lapTimes.position (INTEGER),\\n lapTimes.time (TEXT),\\n lapTimes.milliseconds (INTEGER),\\n)\\n\\nTable pitStops (\\n pitStops.raceid (INTEGER),\\n pitStops.driverid (INTEGER),\\n pitStops.stop (INTEGER),\\n pitStops.lap (INTEGER),\\n pitStops.time (TEXT),\\n pitStops.duration (TEXT),\\n pitStops.milliseconds (INTEGER),\\n)\\n\\nTable qualifying (\\n qualifying.qualifyid (INTEGER),\\n qualifying.raceid (INTEGER),\\n qualifying.driverid (INTEGER),\\n qualifying.constructorid (INTEGER),\\n qualifying.number (INTEGER),\\n qualifying.position (INTEGER),\\n qualifying.q1 (TEXT),\\n qualifying.q2 (TEXT),\\n qualifying.q3 (TEXT),\\n)\\n\\nTable races (\\n races.raceid (INTEGER),\\n races.year (INTEGER),\\n races.round (INTEGER),\\n races.circuitid (INTEGER),\\n races.name (TEXT),\\n races.date (TEXT),\\n races.time (TEXT),\\n races.url (TEXT),\\n)\\n\\nTable results (\\n results.resultid (INTEGER),\\n results.raceid (INTEGER),\\n results.driverid (INTEGER),\\n results.constructorid (INTEGER),\\n results.number (INTEGER),\\n results.grid (INTEGER),\\n results.position (TEXT),\\n results.positiontext (TEXT),\\n results.positionorder (INTEGER),\\n results.points (REAL),\\n results.laps (TEXT),\\n results.time (TEXT),\\n results.milliseconds (TEXT),\\n results.fastestlap (TEXT),\\n results.rank (TEXT),\\n results.fastestlaptime (TEXT),\\n results.fastestlapspeed (TEXT),\\n results.statusid (INTEGER),\\n)\\n\\nTable seasons (\\n seasons.year (INTEGER),\\n seasons.url (TEXT),\\n)\\n\\nTable status (\\n status.statusid (INTEGER),\\n status.status (TEXT),\\n)\\n\\n\\nPossible JOINs:\\n\\nconstructorResults.raceid = races.raceid\\nconstructorResults.constructorid = constructors.constructorid\\n\\nconstructorStandings.raceid = races.raceid\\nconstructorStandings.constructorid = constructors.constructorid\\n\\ndriverStandings.raceid = races.raceid\\ndriverStandings.driverid = drivers.driverid\\n\\nlapTimes.raceid = races.raceid\\nlapTimes.driverid = drivers.driverid\\n\\npitStops.raceid = races.raceid\\npitStops.driverid = drivers.driverid\\n\\nqualifying.raceid = races.raceid\\nqualifying.driverid = drivers.driverid\\nqualifying.constructorid = constructors.constructorid\\n\\nraces.circuitid = circuits.circuitid\\n\\nresults.raceid = races.raceid\\nresults.driverid = drivers.driverid\\nresults.constructorid = constructors.constructorid\\n\",\n \"Table Customers (\\n Customers.customer_id (INTEGER),\\n Customers.customer_name (VARCHAR(80)),\\n Customers.customer_details (VARCHAR(255)),\\n)\\n\\nTable Invoices (\\n Invoices.invoice_number (INTEGER),\\n Invoices.invoice_date (DATETIME),\\n Invoices.invoice_details (VARCHAR(255)),\\n)\\n\\nTable Order_Items (\\n Order_Items.order_item_id (INTEGER),\\n Order_Items.product_id (INTEGER),\\n Order_Items.order_id (INTEGER),\\n Order_Items.order_item_status (VARCHAR(10)),\\n Order_Items.order_item_details (VARCHAR(255)),\\n)\\n\\nTable Orders (\\n Orders.order_id (INTEGER),\\n Orders.customer_id (INTEGER),\\n Orders.order_status (VARCHAR(10)),\\n Orders.date_order_placed (DATETIME),\\n Orders.order_details (VARCHAR(255)),\\n)\\n\\nTable Products (\\n Products.product_id (INTEGER),\\n Products.product_name (VARCHAR(80)),\\n Products.product_details (VARCHAR(255)),\\n)\\n\\nTable Shipment_Items (\\n Shipment_Items.shipment_id (INTEGER),\\n Shipment_Items.order_item_id (INTEGER),\\n)\\n\\nTable Shipments (\\n Shipments.shipment_id (INTEGER),\\n Shipments.order_id (INTEGER),\\n Shipments.invoice_number (INTEGER),\\n Shipments.shipment_tracking_number (VARCHAR(80)),\\n Shipments.shipment_date (DATETIME),\\n Shipments.other_shipment_details (VARCHAR(255)),\\n)\\n\\n\\nPossible JOINs:\\n\\nOrder_Items.product_id = Products.product_id\\nOrder_Items.order_id = Orders.order_id\\n\\nOrders.customer_id = Customers.customer_id\\n\\nShipment_Items.shipment_id = Shipments.shipment_id\\nShipment_Items.order_item_id = Order_Items.order_item_id\\n\\nShipments.order_id = Orders.order_id\\nShipments.invoice_number = Invoices.invoice_number\\n\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"schema_llm_columns_min_ct\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 147,\n \"samples\": [\n \"CREATE TABLE country (\\n country.country_id (INT),\\n country.name (TEXT),\\n country.population (INT),\\n country.area (INT),\\n country.languages (TEXT),\\n)\\n\\nCREATE TABLE roller_coaster (\\n roller_coaster.roller_coaster_id (INT),\\n roller_coaster.name (TEXT),\\n roller_coaster.park (TEXT),\\n roller_coaster.country_id (INT),\\n roller_coaster.length (REAL),\\n roller_coaster.height (REAL),\\n roller_coaster.speed (TEXT),\\n roller_coaster.opened (TEXT),\\n roller_coaster.status (TEXT),\\n)\\n\\n\\nPossible JOINs:\\n\\nroller_coaster.country_id = country.country_id\\n\",\n \"CREATE TABLE circuits (\\n circuits.circuitid (INTEGER),\\n circuits.circuitref (TEXT),\\n circuits.name (TEXT),\\n circuits.location (TEXT),\\n circuits.country (TEXT),\\n circuits.lat (REAL),\\n circuits.lng (REAL),\\n circuits.alt (TEXT),\\n circuits.url (TEXT),\\n)\\n\\nCREATE TABLE constructorResults (\\n constructorResults.constructorresultsid (INTEGER),\\n constructorResults.raceid (INTEGER),\\n constructorResults.constructorid (INTEGER),\\n constructorResults.points (REAL),\\n constructorResults.status (TEXT),\\n)\\n\\nCREATE TABLE constructorStandings (\\n constructorStandings.constructorstandingsid (INTEGER),\\n constructorStandings.raceid (INTEGER),\\n constructorStandings.constructorid (INTEGER),\\n constructorStandings.points (REAL),\\n constructorStandings.position (INTEGER),\\n constructorStandings.positiontext (TEXT),\\n constructorStandings.wins (INTEGER),\\n)\\n\\nCREATE TABLE constructors (\\n constructors.constructorid (INTEGER),\\n constructors.constructorref (TEXT),\\n constructors.name (TEXT),\\n constructors.nationality (TEXT),\\n constructors.url (TEXT),\\n)\\n\\nCREATE TABLE driverStandings (\\n driverStandings.driverstandingsid (INTEGER),\\n driverStandings.raceid (INTEGER),\\n driverStandings.driverid (INTEGER),\\n driverStandings.points (REAL),\\n driverStandings.position (INTEGER),\\n driverStandings.positiontext (TEXT),\\n driverStandings.wins (INTEGER),\\n)\\n\\nCREATE TABLE drivers (\\n drivers.driverid (INTEGER),\\n drivers.driverref (TEXT),\\n drivers.number (TEXT),\\n drivers.code (TEXT),\\n drivers.forename (TEXT),\\n drivers.surname (TEXT),\\n drivers.dob (TEXT),\\n drivers.nationality (TEXT),\\n drivers.url (TEXT),\\n)\\n\\nCREATE TABLE lapTimes (\\n lapTimes.raceid (INTEGER),\\n lapTimes.driverid (INTEGER),\\n lapTimes.lap (INTEGER),\\n lapTimes.position (INTEGER),\\n lapTimes.time (TEXT),\\n lapTimes.milliseconds (INTEGER),\\n)\\n\\nCREATE TABLE pitStops (\\n pitStops.raceid (INTEGER),\\n pitStops.driverid (INTEGER),\\n pitStops.stop (INTEGER),\\n pitStops.lap (INTEGER),\\n pitStops.time (TEXT),\\n pitStops.duration (TEXT),\\n pitStops.milliseconds (INTEGER),\\n)\\n\\nCREATE TABLE qualifying (\\n qualifying.qualifyid (INTEGER),\\n qualifying.raceid (INTEGER),\\n qualifying.driverid (INTEGER),\\n qualifying.constructorid (INTEGER),\\n qualifying.number (INTEGER),\\n qualifying.position (INTEGER),\\n qualifying.q1 (TEXT),\\n qualifying.q2 (TEXT),\\n qualifying.q3 (TEXT),\\n)\\n\\nCREATE TABLE races (\\n races.raceid (INTEGER),\\n races.year (INTEGER),\\n races.round (INTEGER),\\n races.circuitid (INTEGER),\\n races.name (TEXT),\\n races.date (TEXT),\\n races.time (TEXT),\\n races.url (TEXT),\\n)\\n\\nCREATE TABLE results (\\n results.resultid (INTEGER),\\n results.raceid (INTEGER),\\n results.driverid (INTEGER),\\n results.constructorid (INTEGER),\\n results.number (INTEGER),\\n results.grid (INTEGER),\\n results.position (TEXT),\\n results.positiontext (TEXT),\\n results.positionorder (INTEGER),\\n results.points (REAL),\\n results.laps (TEXT),\\n results.time (TEXT),\\n results.milliseconds (TEXT),\\n results.fastestlap (TEXT),\\n results.rank (TEXT),\\n results.fastestlaptime (TEXT),\\n results.fastestlapspeed (TEXT),\\n results.statusid (INTEGER),\\n)\\n\\nCREATE TABLE seasons (\\n seasons.year (INTEGER),\\n seasons.url (TEXT),\\n)\\n\\nCREATE TABLE status (\\n status.statusid (INTEGER),\\n status.status (TEXT),\\n)\\n\\n\\nPossible JOINs:\\n\\nconstructorResults.raceid = races.raceid\\nconstructorResults.constructorid = constructors.constructorid\\n\\nconstructorStandings.raceid = races.raceid\\nconstructorStandings.constructorid = constructors.constructorid\\n\\ndriverStandings.raceid = races.raceid\\ndriverStandings.driverid = drivers.driverid\\n\\nlapTimes.raceid = races.raceid\\nlapTimes.driverid = drivers.driverid\\n\\npitStops.raceid = races.raceid\\npitStops.driverid = drivers.driverid\\n\\nqualifying.raceid = races.raceid\\nqualifying.driverid = drivers.driverid\\nqualifying.constructorid = constructors.constructorid\\n\\nraces.circuitid = circuits.circuitid\\n\\nresults.raceid = races.raceid\\nresults.driverid = drivers.driverid\\nresults.constructorid = constructors.constructorid\\n\",\n \"CREATE TABLE Customers (\\n Customers.customer_id (INTEGER),\\n Customers.customer_name (VARCHAR(80)),\\n Customers.customer_details (VARCHAR(255)),\\n)\\n\\nCREATE TABLE Invoices (\\n Invoices.invoice_number (INTEGER),\\n Invoices.invoice_date (DATETIME),\\n Invoices.invoice_details (VARCHAR(255)),\\n)\\n\\nCREATE TABLE Order_Items (\\n Order_Items.order_item_id (INTEGER),\\n Order_Items.product_id (INTEGER),\\n Order_Items.order_id (INTEGER),\\n Order_Items.order_item_status (VARCHAR(10)),\\n Order_Items.order_item_details (VARCHAR(255)),\\n)\\n\\nCREATE TABLE Orders (\\n Orders.order_id (INTEGER),\\n Orders.customer_id (INTEGER),\\n Orders.order_status (VARCHAR(10)),\\n Orders.date_order_placed (DATETIME),\\n Orders.order_details (VARCHAR(255)),\\n)\\n\\nCREATE TABLE Products (\\n Products.product_id (INTEGER),\\n Products.product_name (VARCHAR(80)),\\n Products.product_details (VARCHAR(255)),\\n)\\n\\nCREATE TABLE Shipment_Items (\\n Shipment_Items.shipment_id (INTEGER),\\n Shipment_Items.order_item_id (INTEGER),\\n)\\n\\nCREATE TABLE Shipments (\\n Shipments.shipment_id (INTEGER),\\n Shipments.order_id (INTEGER),\\n Shipments.invoice_number (INTEGER),\\n Shipments.shipment_tracking_number (VARCHAR(80)),\\n Shipments.shipment_date (DATETIME),\\n Shipments.other_shipment_details (VARCHAR(255)),\\n)\\n\\n\\nPossible JOINs:\\n\\nOrder_Items.product_id = Products.product_id\\nOrder_Items.order_id = Orders.order_id\\n\\nOrders.customer_id = Customers.customer_id\\n\\nShipment_Items.shipment_id = Shipments.shipment_id\\nShipment_Items.order_item_id = Order_Items.order_item_id\\n\\nShipments.order_id = Orders.order_id\\nShipments.invoice_number = Invoices.invoice_number\\n\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"schema_llm_all_min\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 147,\n \"samples\": [\n \"Table country (\\n country.country_id (INT),\\n country.name (TEXT),\\n country.population (INT),\\n country.area (INT),\\n country.languages (TEXT),\\n)\\n\\nTable roller_coaster (\\n roller_coaster.roller_coaster_id (INT),\\n roller_coaster.name (TEXT),\\n roller_coaster.park (TEXT),\\n roller_coaster.country_id (INT),\\n roller_coaster.length (REAL),\\n roller_coaster.height (REAL),\\n roller_coaster.speed (TEXT),\\n roller_coaster.opened (TEXT),\\n roller_coaster.status (TEXT),\\n)\\n\\n\\nPossible JOINs:\\n\\nroller_coaster.country_id = country.country_id\\n\",\n \"Table circuits (\\n circuits.circuitid (INTEGER),\\n circuits.circuitref (TEXT),\\n circuits.name (TEXT),\\n circuits.location (TEXT),\\n circuits.country (TEXT),\\n circuits.lat (REAL),\\n circuits.lng (REAL),\\n circuits.alt (TEXT),\\n circuits.url (TEXT),\\n)\\n\\nTable constructorresults (\\n constructorresults.constructorresultsid (INTEGER),\\n constructorresults.raceid (INTEGER),\\n constructorresults.constructorid (INTEGER),\\n constructorresults.points (REAL),\\n constructorresults.status (TEXT),\\n)\\n\\nTable constructorstandings (\\n constructorstandings.constructorstandingsid (INTEGER),\\n constructorstandings.raceid (INTEGER),\\n constructorstandings.constructorid (INTEGER),\\n constructorstandings.points (REAL),\\n constructorstandings.position (INTEGER),\\n constructorstandings.positiontext (TEXT),\\n constructorstandings.wins (INTEGER),\\n)\\n\\nTable constructors (\\n constructors.constructorid (INTEGER),\\n constructors.constructorref (TEXT),\\n constructors.name (TEXT),\\n constructors.nationality (TEXT),\\n constructors.url (TEXT),\\n)\\n\\nTable driverstandings (\\n driverstandings.driverstandingsid (INTEGER),\\n driverstandings.raceid (INTEGER),\\n driverstandings.driverid (INTEGER),\\n driverstandings.points (REAL),\\n driverstandings.position (INTEGER),\\n driverstandings.positiontext (TEXT),\\n driverstandings.wins (INTEGER),\\n)\\n\\nTable drivers (\\n drivers.driverid (INTEGER),\\n drivers.driverref (TEXT),\\n drivers.number (TEXT),\\n drivers.code (TEXT),\\n drivers.forename (TEXT),\\n drivers.surname (TEXT),\\n drivers.dob (TEXT),\\n drivers.nationality (TEXT),\\n drivers.url (TEXT),\\n)\\n\\nTable laptimes (\\n laptimes.raceid (INTEGER),\\n laptimes.driverid (INTEGER),\\n laptimes.lap (INTEGER),\\n laptimes.position (INTEGER),\\n laptimes.time (TEXT),\\n laptimes.milliseconds (INTEGER),\\n)\\n\\nTable pitstops (\\n pitstops.raceid (INTEGER),\\n pitstops.driverid (INTEGER),\\n pitstops.stop (INTEGER),\\n pitstops.lap (INTEGER),\\n pitstops.time (TEXT),\\n pitstops.duration (TEXT),\\n pitstops.milliseconds (INTEGER),\\n)\\n\\nTable qualifying (\\n qualifying.qualifyid (INTEGER),\\n qualifying.raceid (INTEGER),\\n qualifying.driverid (INTEGER),\\n qualifying.constructorid (INTEGER),\\n qualifying.number (INTEGER),\\n qualifying.position (INTEGER),\\n qualifying.q1 (TEXT),\\n qualifying.q2 (TEXT),\\n qualifying.q3 (TEXT),\\n)\\n\\nTable races (\\n races.raceid (INTEGER),\\n races.year (INTEGER),\\n races.round (INTEGER),\\n races.circuitid (INTEGER),\\n races.name (TEXT),\\n races.date (TEXT),\\n races.time (TEXT),\\n races.url (TEXT),\\n)\\n\\nTable results (\\n results.resultid (INTEGER),\\n results.raceid (INTEGER),\\n results.driverid (INTEGER),\\n results.constructorid (INTEGER),\\n results.number (INTEGER),\\n results.grid (INTEGER),\\n results.position (TEXT),\\n results.positiontext (TEXT),\\n results.positionorder (INTEGER),\\n results.points (REAL),\\n results.laps (TEXT),\\n results.time (TEXT),\\n results.milliseconds (TEXT),\\n results.fastestlap (TEXT),\\n results.rank (TEXT),\\n results.fastestlaptime (TEXT),\\n results.fastestlapspeed (TEXT),\\n results.statusid (INTEGER),\\n)\\n\\nTable seasons (\\n seasons.year (INTEGER),\\n seasons.url (TEXT),\\n)\\n\\nTable status (\\n status.statusid (INTEGER),\\n status.status (TEXT),\\n)\\n\\n\\nPossible JOINs:\\n\\nconstructorresults.raceid = races.raceid\\nconstructorresults.constructorid = constructors.constructorid\\n\\nconstructorstandings.raceid = races.raceid\\nconstructorstandings.constructorid = constructors.constructorid\\n\\ndriverstandings.raceid = races.raceid\\ndriverstandings.driverid = drivers.driverid\\n\\nlaptimes.raceid = races.raceid\\nlaptimes.driverid = drivers.driverid\\n\\npitstops.raceid = races.raceid\\npitstops.driverid = drivers.driverid\\n\\nqualifying.raceid = races.raceid\\nqualifying.driverid = drivers.driverid\\nqualifying.constructorid = constructors.constructorid\\n\\nraces.circuitid = circuits.circuitid\\n\\nresults.raceid = races.raceid\\nresults.driverid = drivers.driverid\\nresults.constructorid = constructors.constructorid\\n\",\n \"Table customers (\\n customers.customer_id (INTEGER),\\n customers.customer_name (VARCHAR(80)),\\n customers.customer_details (VARCHAR(255)),\\n)\\n\\nTable invoices (\\n invoices.invoice_number (INTEGER),\\n invoices.invoice_date (DATETIME),\\n invoices.invoice_details (VARCHAR(255)),\\n)\\n\\nTable order_items (\\n order_items.order_item_id (INTEGER),\\n order_items.product_id (INTEGER),\\n order_items.order_id (INTEGER),\\n order_items.order_item_status (VARCHAR(10)),\\n order_items.order_item_details (VARCHAR(255)),\\n)\\n\\nTable orders (\\n orders.order_id (INTEGER),\\n orders.customer_id (INTEGER),\\n orders.order_status (VARCHAR(10)),\\n orders.date_order_placed (DATETIME),\\n orders.order_details (VARCHAR(255)),\\n)\\n\\nTable products (\\n products.product_id (INTEGER),\\n products.product_name (VARCHAR(80)),\\n products.product_details (VARCHAR(255)),\\n)\\n\\nTable shipment_items (\\n shipment_items.shipment_id (INTEGER),\\n shipment_items.order_item_id (INTEGER),\\n)\\n\\nTable shipments (\\n shipments.shipment_id (INTEGER),\\n shipments.order_id (INTEGER),\\n shipments.invoice_number (INTEGER),\\n shipments.shipment_tracking_number (VARCHAR(80)),\\n shipments.shipment_date (DATETIME),\\n shipments.other_shipment_details (VARCHAR(255)),\\n)\\n\\n\\nPossible JOINs:\\n\\norder_items.product_id = products.product_id\\norder_items.order_id = orders.order_id\\n\\norders.customer_id = customers.customer_id\\n\\nshipment_items.shipment_id = shipments.shipment_id\\nshipment_items.order_item_id = order_items.order_item_id\\n\\nshipments.order_id = orders.order_id\\nshipments.invoice_number = invoices.invoice_number\\n\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"schema_llm_all_min_ct\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 147,\n \"samples\": [\n \"CREATE TABLE country (\\n country.country_id (INT),\\n country.name (TEXT),\\n country.population (INT),\\n country.area (INT),\\n country.languages (TEXT),\\n)\\n\\nCREATE TABLE roller_coaster (\\n roller_coaster.roller_coaster_id (INT),\\n roller_coaster.name (TEXT),\\n roller_coaster.park (TEXT),\\n roller_coaster.country_id (INT),\\n roller_coaster.length (REAL),\\n roller_coaster.height (REAL),\\n roller_coaster.speed (TEXT),\\n roller_coaster.opened (TEXT),\\n roller_coaster.status (TEXT),\\n)\\n\\n\\nPossible JOINs:\\n\\nroller_coaster.country_id = country.country_id\\n\",\n \"CREATE TABLE circuits (\\n circuits.circuitid (INTEGER),\\n circuits.circuitref (TEXT),\\n circuits.name (TEXT),\\n circuits.location (TEXT),\\n circuits.country (TEXT),\\n circuits.lat (REAL),\\n circuits.lng (REAL),\\n circuits.alt (TEXT),\\n circuits.url (TEXT),\\n)\\n\\nCREATE TABLE constructorresults (\\n constructorresults.constructorresultsid (INTEGER),\\n constructorresults.raceid (INTEGER),\\n constructorresults.constructorid (INTEGER),\\n constructorresults.points (REAL),\\n constructorresults.status (TEXT),\\n)\\n\\nCREATE TABLE constructorstandings (\\n constructorstandings.constructorstandingsid (INTEGER),\\n constructorstandings.raceid (INTEGER),\\n constructorstandings.constructorid (INTEGER),\\n constructorstandings.points (REAL),\\n constructorstandings.position (INTEGER),\\n constructorstandings.positiontext (TEXT),\\n constructorstandings.wins (INTEGER),\\n)\\n\\nCREATE TABLE constructors (\\n constructors.constructorid (INTEGER),\\n constructors.constructorref (TEXT),\\n constructors.name (TEXT),\\n constructors.nationality (TEXT),\\n constructors.url (TEXT),\\n)\\n\\nCREATE TABLE driverstandings (\\n driverstandings.driverstandingsid (INTEGER),\\n driverstandings.raceid (INTEGER),\\n driverstandings.driverid (INTEGER),\\n driverstandings.points (REAL),\\n driverstandings.position (INTEGER),\\n driverstandings.positiontext (TEXT),\\n driverstandings.wins (INTEGER),\\n)\\n\\nCREATE TABLE drivers (\\n drivers.driverid (INTEGER),\\n drivers.driverref (TEXT),\\n drivers.number (TEXT),\\n drivers.code (TEXT),\\n drivers.forename (TEXT),\\n drivers.surname (TEXT),\\n drivers.dob (TEXT),\\n drivers.nationality (TEXT),\\n drivers.url (TEXT),\\n)\\n\\nCREATE TABLE laptimes (\\n laptimes.raceid (INTEGER),\\n laptimes.driverid (INTEGER),\\n laptimes.lap (INTEGER),\\n laptimes.position (INTEGER),\\n laptimes.time (TEXT),\\n laptimes.milliseconds (INTEGER),\\n)\\n\\nCREATE TABLE pitstops (\\n pitstops.raceid (INTEGER),\\n pitstops.driverid (INTEGER),\\n pitstops.stop (INTEGER),\\n pitstops.lap (INTEGER),\\n pitstops.time (TEXT),\\n pitstops.duration (TEXT),\\n pitstops.milliseconds (INTEGER),\\n)\\n\\nCREATE TABLE qualifying (\\n qualifying.qualifyid (INTEGER),\\n qualifying.raceid (INTEGER),\\n qualifying.driverid (INTEGER),\\n qualifying.constructorid (INTEGER),\\n qualifying.number (INTEGER),\\n qualifying.position (INTEGER),\\n qualifying.q1 (TEXT),\\n qualifying.q2 (TEXT),\\n qualifying.q3 (TEXT),\\n)\\n\\nCREATE TABLE races (\\n races.raceid (INTEGER),\\n races.year (INTEGER),\\n races.round (INTEGER),\\n races.circuitid (INTEGER),\\n races.name (TEXT),\\n races.date (TEXT),\\n races.time (TEXT),\\n races.url (TEXT),\\n)\\n\\nCREATE TABLE results (\\n results.resultid (INTEGER),\\n results.raceid (INTEGER),\\n results.driverid (INTEGER),\\n results.constructorid (INTEGER),\\n results.number (INTEGER),\\n results.grid (INTEGER),\\n results.position (TEXT),\\n results.positiontext (TEXT),\\n results.positionorder (INTEGER),\\n results.points (REAL),\\n results.laps (TEXT),\\n results.time (TEXT),\\n results.milliseconds (TEXT),\\n results.fastestlap (TEXT),\\n results.rank (TEXT),\\n results.fastestlaptime (TEXT),\\n results.fastestlapspeed (TEXT),\\n results.statusid (INTEGER),\\n)\\n\\nCREATE TABLE seasons (\\n seasons.year (INTEGER),\\n seasons.url (TEXT),\\n)\\n\\nCREATE TABLE status (\\n status.statusid (INTEGER),\\n status.status (TEXT),\\n)\\n\\n\\nPossible JOINs:\\n\\nconstructorresults.raceid = races.raceid\\nconstructorresults.constructorid = constructors.constructorid\\n\\nconstructorstandings.raceid = races.raceid\\nconstructorstandings.constructorid = constructors.constructorid\\n\\ndriverstandings.raceid = races.raceid\\ndriverstandings.driverid = drivers.driverid\\n\\nlaptimes.raceid = races.raceid\\nlaptimes.driverid = drivers.driverid\\n\\npitstops.raceid = races.raceid\\npitstops.driverid = drivers.driverid\\n\\nqualifying.raceid = races.raceid\\nqualifying.driverid = drivers.driverid\\nqualifying.constructorid = constructors.constructorid\\n\\nraces.circuitid = circuits.circuitid\\n\\nresults.raceid = races.raceid\\nresults.driverid = drivers.driverid\\nresults.constructorid = constructors.constructorid\\n\",\n \"CREATE TABLE customers (\\n customers.customer_id (INTEGER),\\n customers.customer_name (VARCHAR(80)),\\n customers.customer_details (VARCHAR(255)),\\n)\\n\\nCREATE TABLE invoices (\\n invoices.invoice_number (INTEGER),\\n invoices.invoice_date (DATETIME),\\n invoices.invoice_details (VARCHAR(255)),\\n)\\n\\nCREATE TABLE order_items (\\n order_items.order_item_id (INTEGER),\\n order_items.product_id (INTEGER),\\n order_items.order_id (INTEGER),\\n order_items.order_item_status (VARCHAR(10)),\\n order_items.order_item_details (VARCHAR(255)),\\n)\\n\\nCREATE TABLE orders (\\n orders.order_id (INTEGER),\\n orders.customer_id (INTEGER),\\n orders.order_status (VARCHAR(10)),\\n orders.date_order_placed (DATETIME),\\n orders.order_details (VARCHAR(255)),\\n)\\n\\nCREATE TABLE products (\\n products.product_id (INTEGER),\\n products.product_name (VARCHAR(80)),\\n products.product_details (VARCHAR(255)),\\n)\\n\\nCREATE TABLE shipment_items (\\n shipment_items.shipment_id (INTEGER),\\n shipment_items.order_item_id (INTEGER),\\n)\\n\\nCREATE TABLE shipments (\\n shipments.shipment_id (INTEGER),\\n shipments.order_id (INTEGER),\\n shipments.invoice_number (INTEGER),\\n shipments.shipment_tracking_number (VARCHAR(80)),\\n shipments.shipment_date (DATETIME),\\n shipments.other_shipment_details (VARCHAR(255)),\\n)\\n\\n\\nPossible JOINs:\\n\\norder_items.product_id = products.product_id\\norder_items.order_id = orders.order_id\\n\\norders.customer_id = customers.customer_id\\n\\nshipment_items.shipment_id = shipments.shipment_id\\nshipment_items.order_item_id = order_items.order_item_id\\n\\nshipments.order_id = orders.order_id\\nshipments.invoice_number = invoices.invoice_number\\n\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"schema_dict\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 146,\n \"samples\": [\n \"{\\n 'company': ['Company_ID', 'Rank', 'Company', 'Headquarters', 'Main_Industry', 'Sales_billion', 'Profits_billion', 'Assets_billion', 'Market_Value'],\\n 'gas_station': ['Station_ID', 'Open_Year', 'Location', 'Manager_Name', 'Vice_Manager_Name', 'Representative_Name'],\\n 'station_company': ['Station_ID', 'Company_ID', 'Rank_of_the_Year']\\n}\",\n \"{\\n 'Premises': ['premise_id', 'premises_type', 'premise_details'],\\n 'Products': ['product_id', 'product_category', 'product_name'],\\n 'Customers': ['customer_id', 'payment_method', 'customer_name', 'customer_phone', 'customer_email', 'customer_address', 'customer_login', 'customer_password'],\\n 'Mailshot_Campaigns': ['mailshot_id', 'product_category', 'mailshot_name', 'mailshot_start_date', 'mailshot_end_date'],\\n 'Customer_Addresses': ['customer_id', 'premise_id', 'date_address_from', 'address_type_code', 'date_address_to'],\\n 'Customer_Orders': ['order_id', 'customer_id', 'order_status_code', 'shipping_method_code', 'order_placed_datetime', 'order_delivered_datetime', 'order_shipping_charges'],\\n 'Mailshot_Customers': ['mailshot_id', 'customer_id', 'outcome_code', 'mailshot_customer_date'],\\n 'Order_Items': ['item_id', 'order_item_status_code', 'order_id', 'product_id', 'item_status_code', 'item_delivered_datetime', 'item_order_quantity']\\n}\",\n \"{\\n 'member': ['Member_ID', 'Name', 'Nationality', 'Role'],\\n 'performance': ['Performance_ID', 'Date', 'Host', 'Location', 'Attendance'],\\n 'member_attendance': ['Member_ID', 'Performance_ID', 'Num_of_Pieces']\\n}\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"selector_correct\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 3373,\n \"samples\": [\n \"{\\n 'user': ['uid', 'user_id', 'name'],\\n 'review': ['rid', 'user_id', 'rating']\\n}\",\n \"{\\n 'Documents': ['document_id', 'document_status_code', 'document_type_code']\\n}\",\n \"{\\n 'state': ['state_name', 'population'],\\n 'highlow': ['state_name', 'highest_point']\\n}\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"schema_llm_t\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 147,\n \"samples\": [\n \"TABLE country (\\n country.Country_ID (INT),\\n country.Name (TEXT),\\n country.Population (INT),\\n country.Area (INT),\\n country.Languages (TEXT),\\n)\\n\\nTABLE roller_coaster (\\n roller_coaster.Roller_Coaster_ID (INT),\\n roller_coaster.Name (TEXT),\\n roller_coaster.Park (TEXT),\\n roller_coaster.Country_ID (INT),\\n roller_coaster.Length (REAL),\\n roller_coaster.Height (REAL),\\n roller_coaster.Speed (TEXT),\\n roller_coaster.Opened (TEXT),\\n roller_coaster.Status (TEXT),\\n)\\n\\n\\nPossible JOINs:\\n\\nroller_coaster.Country_ID = country.Country_ID\\n\",\n \"TABLE circuits (\\n circuits.circuitId (INTEGER),\\n circuits.circuitRef (TEXT),\\n circuits.name (TEXT),\\n circuits.location (TEXT),\\n circuits.country (TEXT),\\n circuits.lat (REAL),\\n circuits.lng (REAL),\\n circuits.alt (TEXT),\\n circuits.url (TEXT),\\n)\\n\\nTABLE constructorResults (\\n constructorResults.constructorResultsId (INTEGER),\\n constructorResults.raceId (INTEGER),\\n constructorResults.constructorId (INTEGER),\\n constructorResults.points (REAL),\\n constructorResults.status (TEXT),\\n)\\n\\nTABLE constructorStandings (\\n constructorStandings.constructorStandingsId (INTEGER),\\n constructorStandings.raceId (INTEGER),\\n constructorStandings.constructorId (INTEGER),\\n constructorStandings.points (REAL),\\n constructorStandings.position (INTEGER),\\n constructorStandings.positionText (TEXT),\\n constructorStandings.wins (INTEGER),\\n)\\n\\nTABLE constructors (\\n constructors.constructorId (INTEGER),\\n constructors.constructorRef (TEXT),\\n constructors.name (TEXT),\\n constructors.nationality (TEXT),\\n constructors.url (TEXT),\\n)\\n\\nTABLE driverStandings (\\n driverStandings.driverStandingsId (INTEGER),\\n driverStandings.raceId (INTEGER),\\n driverStandings.driverId (INTEGER),\\n driverStandings.points (REAL),\\n driverStandings.position (INTEGER),\\n driverStandings.positionText (TEXT),\\n driverStandings.wins (INTEGER),\\n)\\n\\nTABLE drivers (\\n drivers.driverId (INTEGER),\\n drivers.driverRef (TEXT),\\n drivers.number (TEXT),\\n drivers.code (TEXT),\\n drivers.forename (TEXT),\\n drivers.surname (TEXT),\\n drivers.dob (TEXT),\\n drivers.nationality (TEXT),\\n drivers.url (TEXT),\\n)\\n\\nTABLE lapTimes (\\n lapTimes.raceId (INTEGER),\\n lapTimes.driverId (INTEGER),\\n lapTimes.lap (INTEGER),\\n lapTimes.position (INTEGER),\\n lapTimes.time (TEXT),\\n lapTimes.milliseconds (INTEGER),\\n)\\n\\nTABLE pitStops (\\n pitStops.raceId (INTEGER),\\n pitStops.driverId (INTEGER),\\n pitStops.stop (INTEGER),\\n pitStops.lap (INTEGER),\\n pitStops.time (TEXT),\\n pitStops.duration (TEXT),\\n pitStops.milliseconds (INTEGER),\\n)\\n\\nTABLE qualifying (\\n qualifying.qualifyId (INTEGER),\\n qualifying.raceId (INTEGER),\\n qualifying.driverId (INTEGER),\\n qualifying.constructorId (INTEGER),\\n qualifying.number (INTEGER),\\n qualifying.position (INTEGER),\\n qualifying.q1 (TEXT),\\n qualifying.q2 (TEXT),\\n qualifying.q3 (TEXT),\\n)\\n\\nTABLE races (\\n races.raceId (INTEGER),\\n races.year (INTEGER),\\n races.round (INTEGER),\\n races.circuitId (INTEGER),\\n races.name (TEXT),\\n races.date (TEXT),\\n races.time (TEXT),\\n races.url (TEXT),\\n)\\n\\nTABLE results (\\n results.resultId (INTEGER),\\n results.raceId (INTEGER),\\n results.driverId (INTEGER),\\n results.constructorId (INTEGER),\\n results.number (INTEGER),\\n results.grid (INTEGER),\\n results.position (TEXT),\\n results.positionText (TEXT),\\n results.positionOrder (INTEGER),\\n results.points (REAL),\\n results.laps (TEXT),\\n results.time (TEXT),\\n results.milliseconds (TEXT),\\n results.fastestLap (TEXT),\\n results.rank (TEXT),\\n results.fastestLapTime (TEXT),\\n results.fastestLapSpeed (TEXT),\\n results.statusId (INTEGER),\\n)\\n\\nTABLE seasons (\\n seasons.year (INTEGER),\\n seasons.url (TEXT),\\n)\\n\\nTABLE status (\\n status.statusId (INTEGER),\\n status.status (TEXT),\\n)\\n\\n\\nPossible JOINs:\\n\\nconstructorResults.raceId = races.raceId\\nconstructorResults.constructorId = constructors.constructorId\\n\\nconstructorStandings.raceId = races.raceId\\nconstructorStandings.constructorId = constructors.constructorId\\n\\ndriverStandings.raceId = races.raceId\\ndriverStandings.driverId = drivers.driverId\\n\\nlapTimes.raceId = races.raceId\\nlapTimes.driverId = drivers.driverId\\n\\npitStops.raceId = races.raceId\\npitStops.driverId = drivers.driverId\\n\\nqualifying.raceId = races.raceId\\nqualifying.driverId = drivers.driverId\\nqualifying.constructorId = constructors.constructorId\\n\\nraces.circuitId = circuits.circuitId\\n\\nresults.raceId = races.raceId\\nresults.driverId = drivers.driverId\\nresults.constructorId = constructors.constructorId\\n\",\n \"TABLE Customers (\\n Customers.customer_id (INTEGER),\\n Customers.customer_name (VARCHAR(80)),\\n Customers.customer_details (VARCHAR(255)),\\n)\\n\\nTABLE Invoices (\\n Invoices.invoice_number (INTEGER),\\n Invoices.invoice_date (DATETIME),\\n Invoices.invoice_details (VARCHAR(255)),\\n)\\n\\nTABLE Order_Items (\\n Order_Items.order_item_id (INTEGER),\\n Order_Items.product_id (INTEGER),\\n Order_Items.order_id (INTEGER),\\n Order_Items.order_item_status (VARCHAR(10)),\\n Order_Items.order_item_details (VARCHAR(255)),\\n)\\n\\nTABLE Orders (\\n Orders.order_id (INTEGER),\\n Orders.customer_id (INTEGER),\\n Orders.order_status (VARCHAR(10)),\\n Orders.date_order_placed (DATETIME),\\n Orders.order_details (VARCHAR(255)),\\n)\\n\\nTABLE Products (\\n Products.product_id (INTEGER),\\n Products.product_name (VARCHAR(80)),\\n Products.product_details (VARCHAR(255)),\\n)\\n\\nTABLE Shipment_Items (\\n Shipment_Items.shipment_id (INTEGER),\\n Shipment_Items.order_item_id (INTEGER),\\n)\\n\\nTABLE Shipments (\\n Shipments.shipment_id (INTEGER),\\n Shipments.order_id (INTEGER),\\n Shipments.invoice_number (INTEGER),\\n Shipments.shipment_tracking_number (VARCHAR(80)),\\n Shipments.shipment_date (DATETIME),\\n Shipments.other_shipment_details (VARCHAR(255)),\\n)\\n\\n\\nPossible JOINs:\\n\\nOrder_Items.product_id = Products.product_id\\nOrder_Items.order_id = Orders.order_id\\n\\nOrders.customer_id = Customers.customer_id\\n\\nShipment_Items.shipment_id = Shipments.shipment_id\\nShipment_Items.order_item_id = Order_Items.order_item_id\\n\\nShipments.order_id = Orders.order_id\\nShipments.invoice_number = Invoices.invoice_number\\n\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"schema_llm_columns_min_t\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 147,\n \"samples\": [\n \"TABLE country (\\n country.country_id (INT),\\n country.name (TEXT),\\n country.population (INT),\\n country.area (INT),\\n country.languages (TEXT),\\n)\\n\\nTABLE roller_coaster (\\n roller_coaster.roller_coaster_id (INT),\\n roller_coaster.name (TEXT),\\n roller_coaster.park (TEXT),\\n roller_coaster.country_id (INT),\\n roller_coaster.length (REAL),\\n roller_coaster.height (REAL),\\n roller_coaster.speed (TEXT),\\n roller_coaster.opened (TEXT),\\n roller_coaster.status (TEXT),\\n)\\n\\n\\nPossible JOINs:\\n\\nroller_coaster.country_id = country.country_id\\n\",\n \"TABLE circuits (\\n circuits.circuitid (INTEGER),\\n circuits.circuitref (TEXT),\\n circuits.name (TEXT),\\n circuits.location (TEXT),\\n circuits.country (TEXT),\\n circuits.lat (REAL),\\n circuits.lng (REAL),\\n circuits.alt (TEXT),\\n circuits.url (TEXT),\\n)\\n\\nTABLE constructorResults (\\n constructorResults.constructorresultsid (INTEGER),\\n constructorResults.raceid (INTEGER),\\n constructorResults.constructorid (INTEGER),\\n constructorResults.points (REAL),\\n constructorResults.status (TEXT),\\n)\\n\\nTABLE constructorStandings (\\n constructorStandings.constructorstandingsid (INTEGER),\\n constructorStandings.raceid (INTEGER),\\n constructorStandings.constructorid (INTEGER),\\n constructorStandings.points (REAL),\\n constructorStandings.position (INTEGER),\\n constructorStandings.positiontext (TEXT),\\n constructorStandings.wins (INTEGER),\\n)\\n\\nTABLE constructors (\\n constructors.constructorid (INTEGER),\\n constructors.constructorref (TEXT),\\n constructors.name (TEXT),\\n constructors.nationality (TEXT),\\n constructors.url (TEXT),\\n)\\n\\nTABLE driverStandings (\\n driverStandings.driverstandingsid (INTEGER),\\n driverStandings.raceid (INTEGER),\\n driverStandings.driverid (INTEGER),\\n driverStandings.points (REAL),\\n driverStandings.position (INTEGER),\\n driverStandings.positiontext (TEXT),\\n driverStandings.wins (INTEGER),\\n)\\n\\nTABLE drivers (\\n drivers.driverid (INTEGER),\\n drivers.driverref (TEXT),\\n drivers.number (TEXT),\\n drivers.code (TEXT),\\n drivers.forename (TEXT),\\n drivers.surname (TEXT),\\n drivers.dob (TEXT),\\n drivers.nationality (TEXT),\\n drivers.url (TEXT),\\n)\\n\\nTABLE lapTimes (\\n lapTimes.raceid (INTEGER),\\n lapTimes.driverid (INTEGER),\\n lapTimes.lap (INTEGER),\\n lapTimes.position (INTEGER),\\n lapTimes.time (TEXT),\\n lapTimes.milliseconds (INTEGER),\\n)\\n\\nTABLE pitStops (\\n pitStops.raceid (INTEGER),\\n pitStops.driverid (INTEGER),\\n pitStops.stop (INTEGER),\\n pitStops.lap (INTEGER),\\n pitStops.time (TEXT),\\n pitStops.duration (TEXT),\\n pitStops.milliseconds (INTEGER),\\n)\\n\\nTABLE qualifying (\\n qualifying.qualifyid (INTEGER),\\n qualifying.raceid (INTEGER),\\n qualifying.driverid (INTEGER),\\n qualifying.constructorid (INTEGER),\\n qualifying.number (INTEGER),\\n qualifying.position (INTEGER),\\n qualifying.q1 (TEXT),\\n qualifying.q2 (TEXT),\\n qualifying.q3 (TEXT),\\n)\\n\\nTABLE races (\\n races.raceid (INTEGER),\\n races.year (INTEGER),\\n races.round (INTEGER),\\n races.circuitid (INTEGER),\\n races.name (TEXT),\\n races.date (TEXT),\\n races.time (TEXT),\\n races.url (TEXT),\\n)\\n\\nTABLE results (\\n results.resultid (INTEGER),\\n results.raceid (INTEGER),\\n results.driverid (INTEGER),\\n results.constructorid (INTEGER),\\n results.number (INTEGER),\\n results.grid (INTEGER),\\n results.position (TEXT),\\n results.positiontext (TEXT),\\n results.positionorder (INTEGER),\\n results.points (REAL),\\n results.laps (TEXT),\\n results.time (TEXT),\\n results.milliseconds (TEXT),\\n results.fastestlap (TEXT),\\n results.rank (TEXT),\\n results.fastestlaptime (TEXT),\\n results.fastestlapspeed (TEXT),\\n results.statusid (INTEGER),\\n)\\n\\nTABLE seasons (\\n seasons.year (INTEGER),\\n seasons.url (TEXT),\\n)\\n\\nTABLE status (\\n status.statusid (INTEGER),\\n status.status (TEXT),\\n)\\n\\n\\nPossible JOINs:\\n\\nconstructorResults.raceid = races.raceid\\nconstructorResults.constructorid = constructors.constructorid\\n\\nconstructorStandings.raceid = races.raceid\\nconstructorStandings.constructorid = constructors.constructorid\\n\\ndriverStandings.raceid = races.raceid\\ndriverStandings.driverid = drivers.driverid\\n\\nlapTimes.raceid = races.raceid\\nlapTimes.driverid = drivers.driverid\\n\\npitStops.raceid = races.raceid\\npitStops.driverid = drivers.driverid\\n\\nqualifying.raceid = races.raceid\\nqualifying.driverid = drivers.driverid\\nqualifying.constructorid = constructors.constructorid\\n\\nraces.circuitid = circuits.circuitid\\n\\nresults.raceid = races.raceid\\nresults.driverid = drivers.driverid\\nresults.constructorid = constructors.constructorid\\n\",\n \"TABLE Customers (\\n Customers.customer_id (INTEGER),\\n Customers.customer_name (VARCHAR(80)),\\n Customers.customer_details (VARCHAR(255)),\\n)\\n\\nTABLE Invoices (\\n Invoices.invoice_number (INTEGER),\\n Invoices.invoice_date (DATETIME),\\n Invoices.invoice_details (VARCHAR(255)),\\n)\\n\\nTABLE Order_Items (\\n Order_Items.order_item_id (INTEGER),\\n Order_Items.product_id (INTEGER),\\n Order_Items.order_id (INTEGER),\\n Order_Items.order_item_status (VARCHAR(10)),\\n Order_Items.order_item_details (VARCHAR(255)),\\n)\\n\\nTABLE Orders (\\n Orders.order_id (INTEGER),\\n Orders.customer_id (INTEGER),\\n Orders.order_status (VARCHAR(10)),\\n Orders.date_order_placed (DATETIME),\\n Orders.order_details (VARCHAR(255)),\\n)\\n\\nTABLE Products (\\n Products.product_id (INTEGER),\\n Products.product_name (VARCHAR(80)),\\n Products.product_details (VARCHAR(255)),\\n)\\n\\nTABLE Shipment_Items (\\n Shipment_Items.shipment_id (INTEGER),\\n Shipment_Items.order_item_id (INTEGER),\\n)\\n\\nTABLE Shipments (\\n Shipments.shipment_id (INTEGER),\\n Shipments.order_id (INTEGER),\\n Shipments.invoice_number (INTEGER),\\n Shipments.shipment_tracking_number (VARCHAR(80)),\\n Shipments.shipment_date (DATETIME),\\n Shipments.other_shipment_details (VARCHAR(255)),\\n)\\n\\n\\nPossible JOINs:\\n\\nOrder_Items.product_id = Products.product_id\\nOrder_Items.order_id = Orders.order_id\\n\\nOrders.customer_id = Customers.customer_id\\n\\nShipment_Items.shipment_id = Shipments.shipment_id\\nShipment_Items.order_item_id = Order_Items.order_item_id\\n\\nShipments.order_id = Orders.order_id\\nShipments.invoice_number = Invoices.invoice_number\\n\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"schema_llm_all_min_t\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 147,\n \"samples\": [\n \"TABLE country (\\n country.country_id (INT),\\n country.name (TEXT),\\n country.population (INT),\\n country.area (INT),\\n country.languages (TEXT),\\n)\\n\\nTABLE roller_coaster (\\n roller_coaster.roller_coaster_id (INT),\\n roller_coaster.name (TEXT),\\n roller_coaster.park (TEXT),\\n roller_coaster.country_id (INT),\\n roller_coaster.length (REAL),\\n roller_coaster.height (REAL),\\n roller_coaster.speed (TEXT),\\n roller_coaster.opened (TEXT),\\n roller_coaster.status (TEXT),\\n)\\n\\n\\nPossible JOINs:\\n\\nroller_coaster.country_id = country.country_id\\n\",\n \"TABLE circuits (\\n circuits.circuitid (INTEGER),\\n circuits.circuitref (TEXT),\\n circuits.name (TEXT),\\n circuits.location (TEXT),\\n circuits.country (TEXT),\\n circuits.lat (REAL),\\n circuits.lng (REAL),\\n circuits.alt (TEXT),\\n circuits.url (TEXT),\\n)\\n\\nTABLE constructorresults (\\n constructorresults.constructorresultsid (INTEGER),\\n constructorresults.raceid (INTEGER),\\n constructorresults.constructorid (INTEGER),\\n constructorresults.points (REAL),\\n constructorresults.status (TEXT),\\n)\\n\\nTABLE constructorstandings (\\n constructorstandings.constructorstandingsid (INTEGER),\\n constructorstandings.raceid (INTEGER),\\n constructorstandings.constructorid (INTEGER),\\n constructorstandings.points (REAL),\\n constructorstandings.position (INTEGER),\\n constructorstandings.positiontext (TEXT),\\n constructorstandings.wins (INTEGER),\\n)\\n\\nTABLE constructors (\\n constructors.constructorid (INTEGER),\\n constructors.constructorref (TEXT),\\n constructors.name (TEXT),\\n constructors.nationality (TEXT),\\n constructors.url (TEXT),\\n)\\n\\nTABLE driverstandings (\\n driverstandings.driverstandingsid (INTEGER),\\n driverstandings.raceid (INTEGER),\\n driverstandings.driverid (INTEGER),\\n driverstandings.points (REAL),\\n driverstandings.position (INTEGER),\\n driverstandings.positiontext (TEXT),\\n driverstandings.wins (INTEGER),\\n)\\n\\nTABLE drivers (\\n drivers.driverid (INTEGER),\\n drivers.driverref (TEXT),\\n drivers.number (TEXT),\\n drivers.code (TEXT),\\n drivers.forename (TEXT),\\n drivers.surname (TEXT),\\n drivers.dob (TEXT),\\n drivers.nationality (TEXT),\\n drivers.url (TEXT),\\n)\\n\\nTABLE laptimes (\\n laptimes.raceid (INTEGER),\\n laptimes.driverid (INTEGER),\\n laptimes.lap (INTEGER),\\n laptimes.position (INTEGER),\\n laptimes.time (TEXT),\\n laptimes.milliseconds (INTEGER),\\n)\\n\\nTABLE pitstops (\\n pitstops.raceid (INTEGER),\\n pitstops.driverid (INTEGER),\\n pitstops.stop (INTEGER),\\n pitstops.lap (INTEGER),\\n pitstops.time (TEXT),\\n pitstops.duration (TEXT),\\n pitstops.milliseconds (INTEGER),\\n)\\n\\nTABLE qualifying (\\n qualifying.qualifyid (INTEGER),\\n qualifying.raceid (INTEGER),\\n qualifying.driverid (INTEGER),\\n qualifying.constructorid (INTEGER),\\n qualifying.number (INTEGER),\\n qualifying.position (INTEGER),\\n qualifying.q1 (TEXT),\\n qualifying.q2 (TEXT),\\n qualifying.q3 (TEXT),\\n)\\n\\nTABLE races (\\n races.raceid (INTEGER),\\n races.year (INTEGER),\\n races.round (INTEGER),\\n races.circuitid (INTEGER),\\n races.name (TEXT),\\n races.date (TEXT),\\n races.time (TEXT),\\n races.url (TEXT),\\n)\\n\\nTABLE results (\\n results.resultid (INTEGER),\\n results.raceid (INTEGER),\\n results.driverid (INTEGER),\\n results.constructorid (INTEGER),\\n results.number (INTEGER),\\n results.grid (INTEGER),\\n results.position (TEXT),\\n results.positiontext (TEXT),\\n results.positionorder (INTEGER),\\n results.points (REAL),\\n results.laps (TEXT),\\n results.time (TEXT),\\n results.milliseconds (TEXT),\\n results.fastestlap (TEXT),\\n results.rank (TEXT),\\n results.fastestlaptime (TEXT),\\n results.fastestlapspeed (TEXT),\\n results.statusid (INTEGER),\\n)\\n\\nTABLE seasons (\\n seasons.year (INTEGER),\\n seasons.url (TEXT),\\n)\\n\\nTABLE status (\\n status.statusid (INTEGER),\\n status.status (TEXT),\\n)\\n\\n\\nPossible JOINs:\\n\\nconstructorresults.raceid = races.raceid\\nconstructorresults.constructorid = constructors.constructorid\\n\\nconstructorstandings.raceid = races.raceid\\nconstructorstandings.constructorid = constructors.constructorid\\n\\ndriverstandings.raceid = races.raceid\\ndriverstandings.driverid = drivers.driverid\\n\\nlaptimes.raceid = races.raceid\\nlaptimes.driverid = drivers.driverid\\n\\npitstops.raceid = races.raceid\\npitstops.driverid = drivers.driverid\\n\\nqualifying.raceid = races.raceid\\nqualifying.driverid = drivers.driverid\\nqualifying.constructorid = constructors.constructorid\\n\\nraces.circuitid = circuits.circuitid\\n\\nresults.raceid = races.raceid\\nresults.driverid = drivers.driverid\\nresults.constructorid = constructors.constructorid\\n\",\n \"TABLE customers (\\n customers.customer_id (INTEGER),\\n customers.customer_name (VARCHAR(80)),\\n customers.customer_details (VARCHAR(255)),\\n)\\n\\nTABLE invoices (\\n invoices.invoice_number (INTEGER),\\n invoices.invoice_date (DATETIME),\\n invoices.invoice_details (VARCHAR(255)),\\n)\\n\\nTABLE order_items (\\n order_items.order_item_id (INTEGER),\\n order_items.product_id (INTEGER),\\n order_items.order_id (INTEGER),\\n order_items.order_item_status (VARCHAR(10)),\\n order_items.order_item_details (VARCHAR(255)),\\n)\\n\\nTABLE orders (\\n orders.order_id (INTEGER),\\n orders.customer_id (INTEGER),\\n orders.order_status (VARCHAR(10)),\\n orders.date_order_placed (DATETIME),\\n orders.order_details (VARCHAR(255)),\\n)\\n\\nTABLE products (\\n products.product_id (INTEGER),\\n products.product_name (VARCHAR(80)),\\n products.product_details (VARCHAR(255)),\\n)\\n\\nTABLE shipment_items (\\n shipment_items.shipment_id (INTEGER),\\n shipment_items.order_item_id (INTEGER),\\n)\\n\\nTABLE shipments (\\n shipments.shipment_id (INTEGER),\\n shipments.order_id (INTEGER),\\n shipments.invoice_number (INTEGER),\\n shipments.shipment_tracking_number (VARCHAR(80)),\\n shipments.shipment_date (DATETIME),\\n shipments.other_shipment_details (VARCHAR(255)),\\n)\\n\\n\\nPossible JOINs:\\n\\norder_items.product_id = products.product_id\\norder_items.order_id = orders.order_id\\n\\norders.customer_id = customers.customer_id\\n\\nshipment_items.shipment_id = shipments.shipment_id\\nshipment_items.order_item_id = order_items.order_item_id\\n\\nshipments.order_id = orders.order_id\\nshipments.invoice_number = invoices.invoice_number\\n\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"text\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 8647,\n \"samples\": [\n \"<|begin_of_text|><|start_header_id|>system<|end_header_id|>\\n\\nCutting Knowledge Date: December 2023\\nToday Date: 06 Oct 2024\\n\\n<|eot_id|><|start_header_id|>user<|end_header_id|>\\n\\n# System:\\nGiven a user question and the schema of a database, your task is to generate an JSON with the the names of tables and columns of the schema that the question is referring to.\\n\\n# Schema:\\n```sql\\nTABLE event (\\n event.ID (INT),\\n event.Name (TEXT),\\n event.Stadium_ID (INT),\\n event.Year (TEXT),\\n)\\n\\nTABLE record (\\n record.ID (INT),\\n record.Result (TEXT),\\n record.Swimmer_ID (INT),\\n record.Event_ID (INT),\\n)\\n\\nTABLE stadium (\\n stadium.ID (INT),\\n stadium.name (TEXT),\\n stadium.Capacity (INT),\\n stadium.City (TEXT),\\n stadium.Country (TEXT),\\n stadium.Opening_year (INT),\\n)\\n\\nTABLE swimmer (\\n swimmer.ID (INT),\\n swimmer.name (TEXT),\\n swimmer.Nationality (TEXT),\\n swimmer.meter_100 (REAL),\\n swimmer.meter_200 (TEXT),\\n swimmer.meter_300 (TEXT),\\n swimmer.meter_400 (TEXT),\\n swimmer.meter_500 (TEXT),\\n swimmer.meter_600 (TEXT),\\n swimmer.meter_700 (TEXT),\\n swimmer.Time (TEXT),\\n)\\n\\n\\nPossible JOINs:\\n\\nevent.Stadium_ID = stadium.ID\\n\\nrecord.Swimmer_ID = swimmer.ID\\nrecord.Event_ID = event.ID\\n\\n```\\n\\n# Question: Which countries do not have a stadium that was opened after 2006?<|eot_id|><|start_header_id|>assistant<|end_header_id|>\\n\\n```json\\n{ 'stadium': ['ID', 'Country', 'Opening_year'] }\\n```<|eot_id|>\",\n \"<|begin_of_text|><|start_header_id|>system<|end_header_id|>\\n\\nCutting Knowledge Date: December 2023\\nToday Date: 06 Oct 2024\\n\\n<|eot_id|><|start_header_id|>user<|end_header_id|>\\n\\n# System:\\nGiven a user question and the schema of a database, your task is to generate an JSON with the the names of tables and columns of the schema that the question is referring to.\\n\\n# Schema:\\n```sql\\nTABLE film (\\n film.Film_ID (INT),\\n film.Title (TEXT),\\n film.Studio (TEXT),\\n film.Director (TEXT),\\n film.Gross_in_dollar (INT),\\n)\\n\\nTABLE film_market_estimation (\\n film_market_estimation.Estimation_ID (INT),\\n film_market_estimation.Low_Estimate (REAL),\\n film_market_estimation.High_Estimate (REAL),\\n film_market_estimation.Film_ID (INT),\\n film_market_estimation.Type (TEXT),\\n film_market_estimation.Market_ID (INT),\\n film_market_estimation.Year (INT),\\n)\\n\\nTABLE market (\\n market.Market_ID (INT),\\n market.Country (TEXT),\\n market.Number_cities (INT),\\n)\\n\\n\\nPossible JOINs:\\n\\nfilm_market_estimation.Film_ID = film.Film_ID\\nfilm_market_estimation.Market_ID = market.Market_ID\\n\\n```\\n\\n# Question: How many film are there?<|eot_id|><|start_header_id|>assistant<|end_header_id|>\\n\\n```json\\n{ 'film': ['Film_ID'] }\\n```<|eot_id|>\",\n \"<|begin_of_text|><|start_header_id|>system<|end_header_id|>\\n\\nCutting Knowledge Date: December 2023\\nToday Date: 06 Oct 2024\\n\\n<|eot_id|><|start_header_id|>user<|end_header_id|>\\n\\n# System:\\nGiven a user question and the schema of a database, your task is to generate an JSON with the the names of tables and columns of the schema that the question is referring to.\\n\\n# Schema:\\n```sql\\nTABLE book (\\n book.Book_ID (INT),\\n book.Title (TEXT),\\n book.Issues (REAL),\\n book.Writer (TEXT),\\n)\\n\\nTABLE publication (\\n publication.Publication_ID (INT),\\n publication.Book_ID (INT),\\n publication.Publisher (TEXT),\\n publication.Publication_Date (TEXT),\\n publication.Price (REAL),\\n)\\n\\n\\nPossible JOINs:\\n\\npublication.Book_ID = book.Book_ID\\n\\n```\\n\\n# Question: Show the title and publication dates of books.<|eot_id|><|start_header_id|>assistant<|end_header_id|>\\n\\n```json\\n{ 'book': ['Book_ID', 'Title'],\\n 'publication': ['Publication_ID', 'Book_ID', 'Publication_Date'] }\\n```<|eot_id|>\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n }\n ]\n}" } }, "metadata": {}, "execution_count": 14 } ] }, { "cell_type": "code", "source": [ "# df['n_tokens'] = df['text'].apply(lambda x: len(tokenizer.encode(x)))" ], "metadata": { "id": "-8OByl30hBp7" }, "execution_count": null, "outputs": [] }, { "cell_type": "code", "source": [ "# import seaborn as sns\n", "# sns.histplot(df['n_tokens'])" ], "metadata": { "id": "kFMVR-l6hRmx" }, "execution_count": null, "outputs": [] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "TxrNQ8OGvToc", "colab": { "base_uri": "https://localhost:8080/" }, "outputId": "6ba5348c-6d3f-46ff-eed1-2eed5563f39d" }, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "<|begin_of_text|><|start_header_id|>system<|end_header_id|>\n", "\n", "Cutting Knowledge Date: December 2023\n", "Today Date: 06 Oct 2024\n", "\n", "<|eot_id|><|start_header_id|>user<|end_header_id|>\n", "\n", "# System:\n", "Given a user question and the schema of a database, your task is to generate an JSON with the the names of tables and columns of the schema that the question is referring to.\n", "\n", "# Schema:\n", "```sql\n", "TABLE city (\n", " city.City_ID (INT),\n", " city.Official_Name (TEXT),\n", " city.Status (TEXT),\n", " city.Area_km_2 (REAL),\n", " city.Population (REAL),\n", " city.Census_Ranking (TEXT),\n", ")\n", "\n", "TABLE competition_record (\n", " competition_record.Competition_ID (INT),\n", " competition_record.Farm_ID (INT),\n", " competition_record.Rank (INT),\n", ")\n", "\n", "TABLE farm (\n", " farm.Farm_ID (INT),\n", " farm.Year (INT),\n", " farm.Total_Horses (REAL),\n", " farm.Working_Horses (REAL),\n", " farm.Total_Cattle (REAL),\n", " farm.Oxen (REAL),\n", " farm.Bulls (REAL),\n", " farm.Cows (REAL),\n", " farm.Pigs (REAL),\n", " farm.Sheep_and_Goats (REAL),\n", ")\n", "\n", "TABLE farm_competition (\n", " farm_competition.Competition_ID (INT),\n", " farm_competition.Year (INT),\n", " farm_competition.Theme (TEXT),\n", " farm_competition.Host_city_ID (INT),\n", " farm_competition.Hosts (TEXT),\n", ")\n", "\n", "\n", "Possible JOINs:\n", "\n", "competition_record.Competition_ID = farm_competition.Competition_ID\n", "competition_record.Farm_ID = farm.Farm_ID\n", "\n", "farm_competition.Host_city_ID = city.City_ID\n", "\n", "```\n", "\n", "# Question: Show the status shared by cities with population bigger than 1500 and smaller than 500.<|eot_id|><|start_header_id|>assistant<|end_header_id|>\n", "\n", "```json\n", "{ 'city': ['City_ID', 'Status', 'Population'] }\n", "```<|eot_id|>\n" ] } ], "source": [ "print(df['text'][df.index[50]])" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "jgNv1q3IA4J2" }, "outputs": [], "source": [ "_df = pd.DataFrame(columns=['text'])\n", "_df['text'] = df.sample(frac=1, random_state=14).reset_index(drop=True)['text']\n", "_df = Dataset.from_pandas(_df)\n", "_df = _df.train_test_split(test_size=0.01, shuffle=True, seed=14)\n", "train_dataset, valid_dataset = _df[\"train\"], _df[\"test\"]" ] }, { "cell_type": "code", "source": [ "# dataset = Dataset.from_pandas(df)\n", "\n", "# dataset_filtrado_tokens = dataset_easy_and_medium.filter(lambda example: len(tokenizer.tokenize(example['text'])) < 2048)\n", "# dataset_filtrado_tokens\n", "\n", "# train_dataset = dataset_filtrado_tokens.train_test_split(test_size=0.01, shuffle=False, seed=42)[\"train\"]\n", "# valid_dataset = dataset_filtrado_tokens.train_test_split(test_size=0.01, shuffle=False, seed=42)[\"test\"]" ], "metadata": { "id": "h_ntoKppKjZ7" }, "execution_count": null, "outputs": [] }, { "cell_type": "markdown", "metadata": { "id": "_I7-bFfm5gqS" }, "source": [ "#### Chat Template - Gerar SQL" ] }, { "cell_type": "code", "source": [ "# tokenizer.chat_template = \"\"\"\n", "# {% if messages[0]['role'] == 'system' %}\n", "# {% set loop_messages = messages[1:] %}\n", "# {% set system_message = messages[0]['content'] %}\n", "# {% else %}\n", "# {% set loop_messages = messages %}\n", "# {% set system_message = 'Given a user question and the schema of a database, your task is to generate an SQL query that accurately answers the question based on the provided schema.' %}\n", "# {% endif %}\n", "# {{ '# <|system|>/n/' + system_message + '/n//n/' }}\n", "# {% if messages|selectattr(\"role\", \"equalto\", \"example\")|list %}\n", "# Below are some examples of question and their corresponding SQL queries:/n//n/\n", "# {% else %}\n", "# /n/\n", "# {% endif %}\n", "# {% for message in loop_messages %}\n", "# {% if message['role'] == 'example' %}\n", "# {{ message['content'] }}/n//n/\n", "# {% elif message['role'] == 'schema' %}\n", "# # <|schema|>/n/The query will run on a database with the following schema:/n/{{ message['content'] }}/n//n/\n", "# {% elif message['role'] == 'user' %}\n", "# # <|user|>/n/[QUESTION]{{ message['content'] }}[/QUESTION]/n//n/\n", "# {% elif message['role'] == 'assistant' %}\n", "# # <|assistant|>/n/[SQL]{{ message['content'] }}[/SQL]\n", "# {% endif %}\n", "# {% endfor %}\n", "# {% if add_generation_prompt %}\n", "# # <|assistant|>/n/[SQL]\n", "# {% endif %}\n", "# \"\"\".replace(\"\\n\",\"\").replace(\" \", \"\").replace(\"/n/\", \"\\n\")" ], "metadata": { "id": "cYVA3Q7ZCzHi" }, "execution_count": null, "outputs": [] }, { "cell_type": "code", "execution_count": 17, "metadata": { "id": "6RDDdVgP5gqT" }, "outputs": [], "source": [ "def to_sql(query):\n", " return sqlparse.format(query, reindent=True, keyword_case='upper')\n", " #return sqlparse.format(query, reindent=False, keyword_case='upper')\n", "\n", "def apply_template(row, tokenizer=tokenizer, n_examplea=0):\n", " question = row['question_en']\n", " schema = row['schema_SQLDatabase']\n", " sql = to_sql(row['query'])\n", "\n", " system = \"Given a user question and the schema of a database, your task is to generate an SQL query that accurately answers the question based on the provided schema.\"\n", "\n", " chat = [\n", " {'role': 'system', 'content': system},\n", " {'role': 'user', 'content': f\"# Schema:\\n```sql\\n{schema}\\n```\\n\\n# Question: {question}\"},\n", " {'role': 'assistant', 'content': f\"```sql\\n{sql}\\n```\\n\"}\n", " ]\n", "\n", " # chat = [\n", " # {\"role\": \"schema\", \"content\": schema},\n", " # {\"role\": \"user\", \"content\": question},\n", " # {\"role\": \"assistant\", \"content\": sql},\n", " # ]\n", "\n", " row['text'] = tokenizer.apply_chat_template(chat, tokenize=False, add_generation_prompt=False)\n", "\n", " return row" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "2trHEegL5gqU" }, "outputs": [], "source": [ "# spider_chain = json.load(open(\"/content/drive/Shareddrives/LLMs/Datasets/spider/spider_chain_of_thought.json\", \"r\"))\n", "# bird_chain = json.load(open(\"/content/drive/Shareddrives/LLMs/Datasets/bird/bird_chain_of_thought.json\", \"r\"))" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "N4jWrC7s5gqU" }, "outputs": [], "source": [ "# df['CoT'] = spider_chain + bird_chain" ] }, { "cell_type": "code", "execution_count": 18, "metadata": { "id": "bTF0pBsw5gqU" }, "outputs": [], "source": [ "df = df.apply(apply_template, axis=1)" ] }, { "cell_type": "code", "execution_count": 19, "metadata": { "colab": { "base_uri": "https://localhost:8080/", "height": 660 }, "outputId": "e941304a-7558-46b9-9e54-d087d519cf44", "id": "L4tjUv7o5gqV" }, "outputs": [ { "output_type": "execute_result", "data": { "text/plain": [ " db_id question_en \\\n", "0 department_management How many heads of the departments are older th... \n", "1 department_management List the name, born state and age of the heads... \n", "2 department_management List the creation year, name and budget of eac... \n", "3 department_management What are the maximum and minimum budget of the... \n", "4 department_management What is the average number of employees of the... \n", "\n", " query \\\n", "0 SELECT count(*) FROM head WHERE age > 56 \n", "1 SELECT name , born_state , age FROM head ORD... \n", "2 SELECT creation , name , budget_in_billions ... \n", "3 SELECT max(budget_in_billions) , min(budget_i... \n", "4 SELECT avg(num_employees) FROM department WHER... \n", "\n", " answer hardness \\\n", "0 [{'count(*)': 5}] easy \n", "1 [{'name': 'Pádraig Harrington', 'born_state': ... medium \n", "2 [{'Creation': '1789', 'Name': 'State', 'Budget... medium \n", "3 [{'max(budget_in_billions)': 543.2, 'min(budge... medium \n", "4 [{'avg(num_employees)': 105468.16666666667}] easy \n", "\n", " query_llm \\\n", "0 SELECT count(*) FROM head WHERE age > 56 \n", "1 SELECT name , born_state , age FROM head ORD... \n", "2 SELECT creation , name , budget_in_billions ... \n", "3 SELECT max(budget_in_billions) , min(budget_i... \n", "4 SELECT avg(num_employees) FROM department WHER... \n", "\n", " schema_llm_v1 \\\n", "0 Table department (\\n department.Department_... \n", "1 Table department (\\n department.Department_... \n", "2 Table department (\\n department.Department_... \n", "3 Table department (\\n department.Department_... \n", "4 Table department (\\n department.Department_... \n", "\n", " schema_llm_v2 \\\n", "0 CREATE TABLE department (\\n department.Depa... \n", "1 CREATE TABLE department (\\n department.Depa... \n", "2 CREATE TABLE department (\\n department.Depa... \n", "3 CREATE TABLE department (\\n department.Depa... \n", "4 CREATE TABLE department (\\n department.Depa... \n", "\n", " schema_llm_v3 \\\n", "0 CREATE TABLE department (\\n department.depa... \n", "1 CREATE TABLE department (\\n department.depa... \n", "2 CREATE TABLE department (\\n department.depa... \n", "3 CREATE TABLE department (\\n department.depa... \n", "4 CREATE TABLE department (\\n department.depa... \n", "\n", " schema_DTS \\\n", "0 CREATE TABLE `department` (\\n department_id I... \n", "1 CREATE TABLE `department` (\\n department_id I... \n", "2 CREATE TABLE `department` (\\n department_id I... \n", "3 CREATE TABLE `department` (\\n department_id I... \n", "4 CREATE TABLE `department` (\\n department_id I... \n", "\n", " schema_SQLDatabase \\\n", "0 CREATE TABLE department (\\n department_... \n", "1 CREATE TABLE department (\\n department_... \n", "2 CREATE TABLE department (\\n department_... \n", "3 CREATE TABLE department (\\n department_... \n", "4 CREATE TABLE department (\\n department_... \n", "\n", " schema_dict \\\n", "0 {\\n 'department': ['department_id', 'name', '... \n", "1 {\\n 'department': ['department_id', 'name', '... \n", "2 {\\n 'department': ['department_id', 'name', '... \n", "3 {\\n 'department': ['department_id', 'name', '... \n", "4 {\\n 'department': ['department_id', 'name', '... \n", "\n", " text \n", "0 <|system|>\\nGiven a user question and the sche... \n", "1 <|system|>\\nGiven a user question and the sche... \n", "2 <|system|>\\nGiven a user question and the sche... \n", "3 <|system|>\\nGiven a user question and the sche... \n", "4 <|system|>\\nGiven a user question and the sche... " ], "text/html": [ "\n", "
\n", "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
db_idquestion_enqueryanswerhardnessquery_llmschema_llm_v1schema_llm_v2schema_llm_v3schema_DTSschema_SQLDatabaseschema_dicttext
0department_managementHow many heads of the departments are older th...SELECT count(*) FROM head WHERE age > 56[{'count(*)': 5}]easySELECT count(*) FROM head WHERE age > 56Table department (\\n department.Department_...CREATE TABLE department (\\n department.Depa...CREATE TABLE department (\\n department.depa...CREATE TABLE `department` (\\n department_id I...CREATE TABLE department (\\n department_...{\\n 'department': ['department_id', 'name', '...<|system|>\\nGiven a user question and the sche...
1department_managementList the name, born state and age of the heads...SELECT name , born_state , age FROM head ORD...[{'name': 'Pádraig Harrington', 'born_state': ...mediumSELECT name , born_state , age FROM head ORD...Table department (\\n department.Department_...CREATE TABLE department (\\n department.Depa...CREATE TABLE department (\\n department.depa...CREATE TABLE `department` (\\n department_id I...CREATE TABLE department (\\n department_...{\\n 'department': ['department_id', 'name', '...<|system|>\\nGiven a user question and the sche...
2department_managementList the creation year, name and budget of eac...SELECT creation , name , budget_in_billions ...[{'Creation': '1789', 'Name': 'State', 'Budget...mediumSELECT creation , name , budget_in_billions ...Table department (\\n department.Department_...CREATE TABLE department (\\n department.Depa...CREATE TABLE department (\\n department.depa...CREATE TABLE `department` (\\n department_id I...CREATE TABLE department (\\n department_...{\\n 'department': ['department_id', 'name', '...<|system|>\\nGiven a user question and the sche...
3department_managementWhat are the maximum and minimum budget of the...SELECT max(budget_in_billions) , min(budget_i...[{'max(budget_in_billions)': 543.2, 'min(budge...mediumSELECT max(budget_in_billions) , min(budget_i...Table department (\\n department.Department_...CREATE TABLE department (\\n department.Depa...CREATE TABLE department (\\n department.depa...CREATE TABLE `department` (\\n department_id I...CREATE TABLE department (\\n department_...{\\n 'department': ['department_id', 'name', '...<|system|>\\nGiven a user question and the sche...
4department_managementWhat is the average number of employees of the...SELECT avg(num_employees) FROM department WHER...[{'avg(num_employees)': 105468.16666666667}]easySELECT avg(num_employees) FROM department WHER...Table department (\\n department.Department_...CREATE TABLE department (\\n department.Depa...CREATE TABLE department (\\n department.depa...CREATE TABLE `department` (\\n department_id I...CREATE TABLE department (\\n department_...{\\n 'department': ['department_id', 'name', '...<|system|>\\nGiven a user question and the sche...
\n", "
\n", "
\n", "\n", "
\n", " \n", "\n", " \n", "\n", " \n", "
\n", "\n", "\n", "
\n", " \n", "\n", "\n", "\n", " \n", "
\n", "\n", "
\n", "
\n" ], "application/vnd.google.colaboratory.intrinsic+json": { "type": "dataframe", "variable_name": "df", "summary": "{\n \"name\": \"df\",\n \"rows\": 8656,\n \"fields\": [\n {\n \"column\": \"db_id\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 146,\n \"samples\": [\n \"gas_company\",\n \"customers_campaigns_ecommerce\",\n \"performance_attendance\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"question_en\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 8618,\n \"samples\": [\n \"What are the names of courses without prerequisites?\",\n \"Return the investor who have invested in the greatest number of entrepreneurs.\",\n \"Which customers have ever canceled the purchase of the product \\\"food\\\" (the item status is \\\"Cancel\\\")?\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"query\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 4712,\n \"samples\": [\n \"SELECT manager_name FROM gas_station WHERE open_year > 2000 GROUP BY manager_name ORDER BY count(*) DESC LIMIT 1\",\n \"SELECT lname FROM authors WHERE fname = \\\"Amal\\\"\",\n \"SELECT * FROM swimmer\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"answer\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 3453,\n \"samples\": [\n \"[{'Hardware_Model_name': 'XPERIA T, XPERIA J', 'Company_name': 'Sony Mobile Communications'}, {'Hardware_Model_name': 'Lumia 920, Lumia 820, Lumia 620', 'Company_name': 'Nokia Corporation'}, {'Hardware_Model_name': 'GT-I9300', 'Company_name': 'Samsung Electronics Co Ltd'}, {'Hardware_Model_name': 'GT-I9100', 'Company_name': 'Samsung Electronics Co Ltd'}, {'Hardware_Model_name': 'Nokia 700', 'Company_name': 'Nokia Corporation'}]\",\n \"[{'course_name': 'Art history'}, {'course_name': 'English'}, {'course_name': 'French'}, {'course_name': 'data structure'}, {'course_name': 'database'}, {'course_name': 'statistics'}]\",\n \"[{'max(height)': 4260.0, 'avg(height)': 4061.2608695652175}]\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"hardness\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 4,\n \"samples\": [\n \"medium\",\n \"extra\",\n \"easy\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"query_llm\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 4712,\n \"samples\": [\n \"SELECT manager_name FROM gas_station WHERE open_year > 2000 GROUP BY manager_name ORDER BY count(*) DESC LIMIT 1\",\n \"SELECT lname FROM authors WHERE fname = \\\"Amal\\\"\",\n \"SELECT * FROM swimmer\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"schema_llm_v1\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 1,\n \"samples\": [\n \"Table department (\\n department.Department_ID (INT),\\n department.Name (TEXT),\\n department.Creation (TEXT),\\n department.Ranking (INT),\\n department.Budget_in_Billions (REAL),\\n department.Num_Employees (REAL),\\n)\\n\\nTable head (\\n head.head_ID (INT),\\n head.name (TEXT),\\n head.born_state (TEXT),\\n head.age (REAL),\\n)\\n\\nTable management (\\n management.department_ID (INT),\\n management.head_ID (INT),\\n management.temporary_acting (TEXT),\\n)\\n\\n\\nPossible JOINs:\\n\\nmanagement.department_ID = department.Department_ID\\nmanagement.head_ID = head.head_ID\\n\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"schema_llm_v2\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 1,\n \"samples\": [\n \"CREATE TABLE department (\\n department.Department_ID INT PRIMARY KEY,\\n department.Name TEXT,\\n department.Creation TEXT,\\n department.Ranking INT,\\n department.Budget_in_Billions REAL,\\n department.Num_Employees REAL,\\n)\\n\\nCREATE TABLE head (\\n head.head_ID INT PRIMARY KEY,\\n head.name TEXT,\\n head.born_state TEXT,\\n head.age REAL,\\n)\\n\\nCREATE TABLE management (\\n management.department_ID INT PRIMARY KEY,\\n management.head_ID INT,\\n management.temporary_acting TEXT,\\n)\\n\\n\\nPossible JOINs:\\n\\nmanagement.department_ID = department.Department_ID\\nmanagement.head_ID = head.head_ID\\n\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"schema_llm_v3\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 1,\n \"samples\": [\n \"CREATE TABLE department (\\n department.department_id INT PRIMARY KEY,\\n department.name TEXT,\\n department.creation TEXT,\\n department.ranking INT,\\n department.budget_in_billions REAL,\\n department.num_employees REAL,\\n)\\n\\nCREATE TABLE head (\\n head.head_id INT PRIMARY KEY,\\n head.name TEXT,\\n head.born_state TEXT,\\n head.age REAL,\\n)\\n\\nCREATE TABLE management (\\n management.department_id INT PRIMARY KEY,\\n management.head_id INT,\\n management.temporary_acting TEXT,\\n)\\n\\n\\nPossible JOINs:\\n\\nmanagement.department_id = department.department_id\\nmanagement.head_id = head.head_id\\n\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"schema_DTS\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 146,\n \"samples\": [\n \"CREATE TABLE `company` (\\n company_id INT PRIMARY KEY,\\n rank INT,\\n company TEXT,\\n headquarters TEXT,\\n main_industry TEXT,\\n sales_billion REAL,\\n profits_billion REAL,\\n assets_billion REAL,\\n market_value REAL\\n);\\n\\n\\nCREATE TABLE `gas_station` (\\n station_id INT PRIMARY KEY,\\n open_year INT,\\n location TEXT,\\n manager_name TEXT,\\n vice_manager_name TEXT,\\n representative_name TEXT\\n);\\n\\n\\nCREATE TABLE `station_company` (\\n station_id INT PRIMARY KEY REFERENCES gas_station(station_id),\\n company_id INT PRIMARY KEY REFERENCES company(company_id),\\n rank_of_the_year INT\\n);\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"schema_SQLDatabase\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 146,\n \"samples\": [\n \"CREATE TABLE company (\\n company_id INT,\\n rank INT,\\n company TEXT,\\n headquarters TEXT,\\n main_industry TEXT,\\n sales_billion REAL,\\n profits_billion REAL,\\n assets_billion REAL,\\n market_value REAL,\\n PRIMARY KEY (company_id)\\n);\\n\\nCREATE TABLE gas_station (\\n station_id INT,\\n open_year INT,\\n location TEXT,\\n manager_name TEXT,\\n vice_manager_name TEXT,\\n representative_name TEXT,\\n PRIMARY KEY (station_id)\\n);\\n\\nCREATE TABLE station_company (\\n station_id INT,\\n company_id INT,\\n rank_of_the_year INT,\\n PRIMARY KEY (station_id),\\n FOREIGN KEY (company_id) REFERENCES company(company_id),\\n FOREIGN KEY (station_id) REFERENCES gas_station(station_id)\\n);\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"schema_dict\",\n \"properties\": {\n \"dtype\": \"category\",\n \"num_unique_values\": 146,\n \"samples\": [\n \"{\\n 'company': ['company_id', 'rank', 'company', 'headquarters', 'main_industry', 'sales_billion', 'profits_billion', 'assets_billion', 'market_value'],\\n 'gas_station': ['station_id', 'open_year', 'location', 'manager_name', 'vice_manager_name', 'representative_name'],\\n 'station_company': ['station_id', 'company_id', 'rank_of_the_year']\\n}\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n },\n {\n \"column\": \"text\",\n \"properties\": {\n \"dtype\": \"string\",\n \"num_unique_values\": 8648,\n \"samples\": [\n \"<|system|>\\nGiven a user question and the schema of a database, your task is to generate an SQL query that accurately answers the question based on the provided schema.\\n<|user|>\\n# Schema:\\n```sql\\nCREATE TABLE Student (\\n stuid INTEGER,\\n lname VARCHAR(12),\\n fname VARCHAR(12),\\n age INTEGER,\\n sex VARCHAR(1),\\n major INTEGER,\\n advisor INTEGER,\\n city_code VARCHAR(3),\\n PRIMARY KEY (stuid)\\n);\\n\\nCREATE TABLE Dorm (\\n dormid INTEGER,\\n dorm_name VARCHAR(20),\\n student_capacity INTEGER,\\n gender VARCHAR(1)\\n);\\n\\nCREATE TABLE Dorm_amenity (\\n amenid INTEGER,\\n amenity_name VARCHAR(25)\\n);\\n\\nCREATE TABLE Has_amenity (\\n dormid INTEGER,\\n amenid INTEGER,\\n FOREIGN KEY (amenid) REFERENCES dorm_amenity(amenid),\\n FOREIGN KEY (dormid) REFERENCES dorm(dormid)\\n);\\n\\nCREATE TABLE Lives_in (\\n stuid INTEGER,\\n dormid INTEGER,\\n room_number INTEGER,\\n FOREIGN KEY (dormid) REFERENCES dorm(dormid),\\n FOREIGN KEY (stuid) REFERENCES student(stuid)\\n);\\n```\\n\\n# Question: How many students are from each city, and which cities have more than one cities?\\n<|assistant|>\\n```sql\\nSELECT count(*),\\n city_code\\nFROM student\\nGROUP BY city_code\\nHAVING count(*) > 1\\n```\\n\\n\"\n ],\n \"semantic_type\": \"\",\n \"description\": \"\"\n }\n }\n ]\n}" } }, "metadata": {}, "execution_count": 19 } ], "source": [ "df.head()" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "DfJvLaGR5gqV" }, "outputs": [], "source": [ "# df['n_tokens'] = df['text'].apply(lambda x: len(tokenizer.encode(x)))" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "id": "vJseOHIu5gqW" }, "outputs": [], "source": [ "# import seaborn as sns\n", "# sns.histplot(df['n_tokens'])" ] }, { "cell_type": "code", "execution_count": 32, "metadata": { "colab": { "base_uri": "https://localhost:8080/" }, "outputId": "4880973c-3526-4395-fcc1-d18e5103ddaa", "id": "PIvSnr6Y5gqW" }, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "<|system|>\n", "Given a user question and the schema of a database, your task is to generate an SQL query that accurately answers the question based on the provided schema.\n", "<|user|>\n", "# Schema:\n", "```sql\n", "CREATE TABLE city (\n", " city_id INT,\n", " official_name TEXT,\n", " status TEXT,\n", " area_km_2 REAL,\n", " population REAL,\n", " census_ranking TEXT,\n", " PRIMARY KEY (city_id)\n", ");\n", "\n", "CREATE TABLE farm (\n", " farm_id INT,\n", " year INT,\n", " total_horses REAL,\n", " working_horses REAL,\n", " total_cattle REAL,\n", " oxen REAL,\n", " bulls REAL,\n", " cows REAL,\n", " pigs REAL,\n", " sheep_and_goats REAL,\n", " PRIMARY KEY (farm_id)\n", ");\n", "\n", "CREATE TABLE farm_competition (\n", " competition_id INT,\n", " year INT,\n", " theme TEXT,\n", " host_city_id INT,\n", " hosts TEXT,\n", " PRIMARY KEY (competition_id),\n", " FOREIGN KEY (host_city_id) REFERENCES city(city_id)\n", ");\n", "\n", "CREATE TABLE competition_record (\n", " competition_id INT,\n", " farm_id INT,\n", " rank INT,\n", " PRIMARY KEY (competition_id),\n", " FOREIGN KEY (farm_id) REFERENCES farm(farm_id),\n", " FOREIGN KEY (competition_id) REFERENCES farm_competition(competition_id)\n", ");\n", "```\n", "\n", "# Question: Show the status shared by cities with population bigger than 1500 and smaller than 500.\n", "<|assistant|>\n", "```sql\n", "SELECT Status\n", "FROM city\n", "WHERE Population > 1500 INTERSECT\n", " SELECT Status\n", " FROM city WHERE Population < 500\n", "```\n", "\n", "\n" ] } ], "source": [ "print(df['text'][df.index[50]])" ] }, { "cell_type": "code", "execution_count": 33, "metadata": { "id": "roZzKNOj5gqW" }, "outputs": [], "source": [ "_df = pd.DataFrame(columns=['text'])\n", "_df['text'] = df.sample(frac=1, random_state=14).reset_index(drop=True)['text']\n", "_df = Dataset.from_pandas(_df)\n", "_df = _df.train_test_split(test_size=0.01, shuffle=True, seed=14)\n", "train_dataset, valid_dataset = _df[\"train\"], _df[\"test\"]" ] }, { "cell_type": "markdown", "metadata": { "id": "DWpXeuO_KlLS" }, "source": [ "### Finetuning" ] }, { "cell_type": "code", "execution_count": 34, "metadata": { "id": "0oVpZDj1AXY9", "colab": { "base_uri": "https://localhost:8080/" }, "outputId": "94910295-6b3a-48c3-a2f1-c0c3f5e86939" }, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "The token has not been saved to the git credentials helper. Pass `add_to_git_credential=True` in this function directly or `--add-to-git-credential` if using via `huggingface-cli` if you want to set the git credential as well.\n", "Token is valid (permission: write).\n", "Your token has been saved to /root/.cache/huggingface/token\n", "Login successful\n" ] } ], "source": [ "from huggingface_hub import login, create_repo\n", "from google.colab import userdata\n", "import wandb\n", "import os\n", "\n", "#token = userdata.get('hf_write')\n", "token = WRITE_TOKEN\n", "login(token=token)\n", "set_seed(1234)" ] }, { "cell_type": "code", "execution_count": 35, "metadata": { "id": "KRhO7UJ-Q4Y8" }, "outputs": [], "source": [ "def find_all_linear_names(model, new_tokens=False):\n", " lora_module_names = set()\n", " for name, module in model.named_modules():\n", " if isinstance(module, bnb.nn.Linear4bit) or isinstance(module, bnb.nn.Linear8bitLt):\n", " names = name.split(\".\")\n", " lora_module_names.add(names[0] if len(names) == 1 else names[-1])\n", " if(new_tokens):\n", " lora_module_names.add(\"lm_head\")\n", " return list(lora_module_names)" ] }, { "cell_type": "code", "execution_count": 36, "metadata": { "id": "L0qqP5Y9PtRh", "colab": { "base_uri": "https://localhost:8080/" }, "outputId": "88175d7e-3c66-4150-ca78-8dbfcd6ba583" }, "outputs": [ { "output_type": "stream", "name": "stdout", "text": [ "Found 7 modules to quantize: ['up_proj', 'gate_proj', 'down_proj', 'k_proj', 'o_proj', 'q_proj', 'v_proj']\n" ] } ], "source": [ "modules = find_all_linear_names(model)\n", "print(f\"Found {len(modules)} modules to quantize: {modules}\")" ] }, { "cell_type": "code", "execution_count": 37, "metadata": { "id": "uFUnJrbjPwAT" }, "outputs": [], "source": [ "peft_config = LoraConfig(\n", " lora_alpha=128, #primeira versão = 16\n", " lora_dropout=0.1,\n", " r=64,\n", " # bias=\"none\",\n", " # task_type=\"CAUSAL_LM\",\n", " target_modules=modules,\n", " # modules_to_save=[\"embed_tokens\"], #quando adicionar tokens speciais\n", ")" ] }, { "cell_type": "code", "execution_count": 38, "metadata": { "id": "buh0o2P2jwbx" }, "outputs": [], "source": [ "torch.cuda.empty_cache()" ] }, { "cell_type": "code", "execution_count": 39, "metadata": { "id": "9bD7ea0F-GQn", "colab": { "base_uri": "https://localhost:8080/", "height": 771, "referenced_widgets": [ "afe6382aee5b4b90875101e2caf9628f", "556724428f0745ab94113142eb4d0ea7", "7cf72f76971f4c79a48aecd5675e9c5e", "c825ba6e1a1d4380be685c1d9475039f", "5c55529e0fd2470a8e3a7c0b7d07a197", "67e0ccb6c4e6446280a0fe0159f6b798", "1fee1e975fc843ea8c9ebfbdc3198b0f", "780f842bcea24ae29f38c33318f41b15", "d7583ff4442544618d108139f9a04b83", "2420d610b2f740d9b33cc1b347984e83", "7c87afd0eb39476d8f750cec181f61c2", "a7d961ad69274d879190bb3df2c878de", "fab89113c0ef4618b9c1f82d2f300ca7", "86a368d742e641ff96e84a78c19ed96c", "7eb8e2a7094b476aad48c35e7a761344", "cb58424d2a904489956e4f10cb7972b3", "345cd61b394f4a58b554be446c75c8b7", "46d3a1f2e16c4ec2bef78da42bc58dc7", "ac7afbd6e6504ac78d481994f0a29ff5", "f394a935ac4a438bb0749849cad08494", "23869c6cbab647d89abdb1bba17fd37a", "2c2ca8a948ed4627891113d8de8f9a00" ] }, "outputId": "e3675a24-fa43-48ce-c260-60cf66652bdc" }, "outputs": [ { "output_type": "stream", "name": "stderr", "text": [ "/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_deprecation.py:100: FutureWarning: Deprecated argument(s) used in '__init__': dataset_text_field, max_seq_length. Will not be supported from version '0.13.0'.\n", "\n", "Deprecated positional argument(s) used in SFTTrainer, please use the SFTConfig to set these arguments instead.\n", " warnings.warn(message, FutureWarning)\n", "/usr/local/lib/python3.10/dist-packages/trl/trainer/sft_trainer.py:300: UserWarning: You passed a `max_seq_length` argument to the SFTTrainer, the value you passed will override the one in the `SFTConfig`.\n", " warnings.warn(\n", "/usr/local/lib/python3.10/dist-packages/trl/trainer/sft_trainer.py:328: UserWarning: You passed a `dataset_text_field` argument to the SFTTrainer, the value you passed will override the one in the `SFTConfig`.\n", " warnings.warn(\n" ] }, { "output_type": "display_data", "data": { "text/plain": [ "Map: 0%| | 0/8569 [00:00" ], "text/html": [ "\n", "
\n", " \n", " \n", " [1071/1071 1:50:28, Epoch 0/1]\n", "
\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
StepTraining LossValidation Loss
2500.2821000.148504
5000.1043000.096154
7500.0782000.083005
10000.0685000.079786

" ] }, "metadata": {} }, { "output_type": "stream", "name": "stderr", "text": [ "Trainer.tokenizer is now deprecated. You should use Trainer.processing_class instead.\n", "Trainer.tokenizer is now deprecated. You should use Trainer.processing_class instead.\n", "/usr/local/lib/python3.10/dist-packages/torch/_dynamo/eval_frame.py:632: UserWarning: torch.utils.checkpoint: the use_reentrant parameter should be passed explicitly. In version 2.5 we will raise an exception if use_reentrant is not passed. use_reentrant=False is recommended, but if you need to preserve the current default behavior, you can pass use_reentrant=True. Refer to docs for more details on the differences between the two variants.\n", " return fn(*args, **kwargs)\n", "Trainer.tokenizer is now deprecated. You should use Trainer.processing_class instead.\n", "Trainer.tokenizer is now deprecated. You should use Trainer.processing_class instead.\n", "/usr/local/lib/python3.10/dist-packages/torch/_dynamo/eval_frame.py:632: UserWarning: torch.utils.checkpoint: the use_reentrant parameter should be passed explicitly. In version 2.5 we will raise an exception if use_reentrant is not passed. use_reentrant=False is recommended, but if you need to preserve the current default behavior, you can pass use_reentrant=True. Refer to docs for more details on the differences between the two variants.\n", " return fn(*args, **kwargs)\n", "Trainer.tokenizer is now deprecated. You should use Trainer.processing_class instead.\n", "Trainer.tokenizer is now deprecated. You should use Trainer.processing_class instead.\n", "/usr/local/lib/python3.10/dist-packages/torch/_dynamo/eval_frame.py:632: UserWarning: torch.utils.checkpoint: the use_reentrant parameter should be passed explicitly. In version 2.5 we will raise an exception if use_reentrant is not passed. use_reentrant=False is recommended, but if you need to preserve the current default behavior, you can pass use_reentrant=True. Refer to docs for more details on the differences between the two variants.\n", " return fn(*args, **kwargs)\n", "Trainer.tokenizer is now deprecated. You should use Trainer.processing_class instead.\n", "Trainer.tokenizer is now deprecated. You should use Trainer.processing_class instead.\n", "/usr/local/lib/python3.10/dist-packages/torch/_dynamo/eval_frame.py:632: UserWarning: torch.utils.checkpoint: the use_reentrant parameter should be passed explicitly. In version 2.5 we will raise an exception if use_reentrant is not passed. use_reentrant=False is recommended, but if you need to preserve the current default behavior, you can pass use_reentrant=True. Refer to docs for more details on the differences between the two variants.\n", " return fn(*args, **kwargs)\n" ] } ], "source": [ "steps=250\n", "strategy=\"steps\"\n", "\n", "trainer = SFTTrainer(\n", " model= model,\n", " tokenizer=tokenizer,\n", "\n", " train_dataset=train_dataset,\n", " eval_dataset=valid_dataset,\n", "\n", " peft_config=peft_config,\n", "\n", " dataset_text_field=\"text\",\n", " max_seq_length=2048,\n", " args= TrainingArguments(\n", " output_dir=\"out\",\n", "\n", " save_strategy=strategy,\n", " save_steps= steps,\n", "\n", " logging_strategy=strategy,\n", " logging_steps=steps,\n", " logging_dir=\"out/logs\",\n", "\n", " eval_strategy=strategy,\n", " eval_steps=steps,\n", "\n", " num_train_epochs=1,\n", " per_device_train_batch_size=1,\n", " per_device_eval_batch_size=1,\n", " gradient_accumulation_steps=8, #primeira versão era 1\n", " gradient_checkpointing=True,\n", " fp16=True,\n", " bf16=False,\n", "\n", " optim=\"paged_adamw_8bit\",\n", " learning_rate=1e-4, #primeira versão era 2e-4,\n", " weight_decay=0.001,\n", " max_grad_norm=0.3,\n", " max_steps=-1,\n", " warmup_ratio=0.03,\n", " group_by_length=True,\n", " lr_scheduler_type=\"cosine\", #primeira versão era linear\n", " seed= seed,\n", "\n", "\n", " report_to=[\"tensorboard\"],\n", " push_to_hub=True,\n", " hub_strategy=\"all_checkpoints\",\n", " hub_model_id=out_name,\n", "\n", " label_names=[\"labels\"]\n", " )\n", "\n", ")\n", "if(prev_checkpoint != None):\n", " trainer.train(f\"out/{prev_checkpoint}\")\n", "else:\n", " trainer.train()" ] }, { "cell_type": "code", "execution_count": 40, "metadata": { "id": "i39Ov4dsGv77", "colab": { "base_uri": "https://localhost:8080/", "height": 52 }, "outputId": "cfc1ca33-8b16-470d-ea8a-919b110f9bb7" }, "outputs": [ { "output_type": "execute_result", "data": { "text/plain": [ "CommitInfo(commit_url='https://huggingface.co/lleticiasilvaa/TinyLlama-1.1B-GerarSQL-v2-identacao/commit/62074f329d98a7e5761c352308fb1167f271ac1d', commit_message='End of training', commit_description='', oid='62074f329d98a7e5761c352308fb1167f271ac1d', pr_url=None, pr_revision=None, pr_num=None)" ], "application/vnd.google.colaboratory.intrinsic+json": { "type": "string" } }, "metadata": {}, "execution_count": 40 } ], "source": [ "trainer.push_to_hub()" ] } ], "metadata": { "accelerator": "GPU", "colab": { "gpuType": "T4", "provenance": [], "collapsed_sections": [ "leGEqxURAL4r" ] }, "kernelspec": { "display_name": "default:Python", "language": "python", "name": "conda-env-default-py" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.9.16" }, "widgets": { "application/vnd.jupyter.widget-state+json": { "5a1e71143dc741efac418bc21eb99526": { "model_module": "@jupyter-widgets/controls", "model_name": "HBoxModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HBoxModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HBoxView", "box_style": "", "children": [ "IPY_MODEL_e380a31a7c9a41f1bd9de22a4eec9afd", "IPY_MODEL_73f947fba7da4fef829ee2034171a836", "IPY_MODEL_a9abf81190a54b29a4fbf7c2514d8b3e" ], "layout": "IPY_MODEL_68dd1789b9a34b91a74c7d00a6c5c491" } }, "e380a31a7c9a41f1bd9de22a4eec9afd": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HTMLModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HTMLView", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_dce14eebcd734217be5fd8bfd1099439", "placeholder": "​", "style": "IPY_MODEL_0ecff7b9301b4713870b0747682533b1", "value": "README.md: 100%" } }, "73f947fba7da4fef829ee2034171a836": { "model_module": "@jupyter-widgets/controls", "model_name": "FloatProgressModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "FloatProgressModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "ProgressView", "bar_style": "success", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_fb996e2abfcc4a0caaec2a2f3ff9fc9a", "max": 1228, "min": 0, "orientation": "horizontal", "style": "IPY_MODEL_11a036e18eac4d7d809a05f68cfb1096", "value": 1228 } }, "a9abf81190a54b29a4fbf7c2514d8b3e": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HTMLModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HTMLView", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_d10ad1893e1f48849896994284318067", "placeholder": "​", "style": "IPY_MODEL_418852eab70342b89b6c1589874d71d9", "value": " 1.23k/1.23k [00:00<00:00, 59.6kB/s]" } }, "68dd1789b9a34b91a74c7d00a6c5c491": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "dce14eebcd734217be5fd8bfd1099439": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "0ecff7b9301b4713870b0747682533b1": { "model_module": "@jupyter-widgets/controls", "model_name": "DescriptionStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "DescriptionStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "description_width": "" } }, "fb996e2abfcc4a0caaec2a2f3ff9fc9a": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "11a036e18eac4d7d809a05f68cfb1096": { "model_module": "@jupyter-widgets/controls", "model_name": "ProgressStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "ProgressStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "bar_color": null, "description_width": "" } }, "d10ad1893e1f48849896994284318067": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "418852eab70342b89b6c1589874d71d9": { "model_module": "@jupyter-widgets/controls", "model_name": "DescriptionStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "DescriptionStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "description_width": "" } }, "dc51fa9dd4bc4047b0bb22500b95b1f7": { "model_module": "@jupyter-widgets/controls", "model_name": "HBoxModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HBoxModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HBoxView", "box_style": "", "children": [ "IPY_MODEL_d01ce95ef53d49bd913fcec32505729d", "IPY_MODEL_748a1613620249e6ae4d71aee744024f", "IPY_MODEL_f1c7ee8b443c4638a0d1423248a2207b" ], "layout": "IPY_MODEL_01301a7d24a3430387ccd7997ce62361" } }, "d01ce95ef53d49bd913fcec32505729d": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HTMLModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HTMLView", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_3bfbe04029a544c0924738e19e7a9436", "placeholder": "​", "style": "IPY_MODEL_f3b80afa9a464264b4a001b9eae61983", "value": "trainSpider-00000-of-00001.parquet: 100%" } }, "748a1613620249e6ae4d71aee744024f": { "model_module": "@jupyter-widgets/controls", "model_name": "FloatProgressModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "FloatProgressModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "ProgressView", "bar_style": "success", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_b5c66c0f758645f58aa66d526224e467", "max": 2000442, "min": 0, "orientation": "horizontal", "style": "IPY_MODEL_771a2d454ea24693a24e124490a97f54", "value": 2000442 } }, "f1c7ee8b443c4638a0d1423248a2207b": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HTMLModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HTMLView", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_0da2fbbaf26a4407bcfc9d2526870a13", "placeholder": "​", "style": "IPY_MODEL_40e6cc0f3f7d4ba586f733c8000cef3e", "value": " 2.00M/2.00M [00:00<00:00, 47.0MB/s]" } }, "01301a7d24a3430387ccd7997ce62361": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "3bfbe04029a544c0924738e19e7a9436": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "f3b80afa9a464264b4a001b9eae61983": { "model_module": "@jupyter-widgets/controls", "model_name": "DescriptionStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "DescriptionStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "description_width": "" } }, "b5c66c0f758645f58aa66d526224e467": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "771a2d454ea24693a24e124490a97f54": { "model_module": "@jupyter-widgets/controls", "model_name": "ProgressStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "ProgressStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "bar_color": null, "description_width": "" } }, "0da2fbbaf26a4407bcfc9d2526870a13": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "40e6cc0f3f7d4ba586f733c8000cef3e": { "model_module": "@jupyter-widgets/controls", "model_name": "DescriptionStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "DescriptionStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "description_width": "" } }, "a5f7375c80e941418802baba0868b509": { "model_module": "@jupyter-widgets/controls", "model_name": "HBoxModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HBoxModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HBoxView", "box_style": "", "children": [ "IPY_MODEL_0157feead1184dc89e1b23bb4c3c6265", "IPY_MODEL_f6343ed2e1894a6a8a537a341628a1b3", "IPY_MODEL_38dc818a4cad4ff686f309679fea6fbd" ], "layout": "IPY_MODEL_ace1a9b16b3e4114a9810a9bb1800434" } }, "0157feead1184dc89e1b23bb4c3c6265": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HTMLModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HTMLView", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_5a15e7b737b54026b942fecd32bae90a", "placeholder": "​", "style": "IPY_MODEL_c888470c2a3f4d5b93227002e2ae463a", "value": "devSpider-00000-of-00001.parquet: 100%" } }, "f6343ed2e1894a6a8a537a341628a1b3": { "model_module": "@jupyter-widgets/controls", "model_name": "FloatProgressModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "FloatProgressModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "ProgressView", "bar_style": "success", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_e4778b1b2402447589011dd20dea59d5", "max": 296647, "min": 0, "orientation": "horizontal", "style": "IPY_MODEL_5df57b5daa364f8fbf4d839886ef0283", "value": 296647 } }, "38dc818a4cad4ff686f309679fea6fbd": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HTMLModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HTMLView", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_a304bcda80bf4148b38ba8d30454d92b", "placeholder": "​", "style": "IPY_MODEL_06976ba3ed7b4c0f9741604ca01dec8a", "value": " 297k/297k [00:00<00:00, 5.68MB/s]" } }, "ace1a9b16b3e4114a9810a9bb1800434": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "5a15e7b737b54026b942fecd32bae90a": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "c888470c2a3f4d5b93227002e2ae463a": { "model_module": "@jupyter-widgets/controls", "model_name": "DescriptionStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "DescriptionStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "description_width": "" } }, "e4778b1b2402447589011dd20dea59d5": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "5df57b5daa364f8fbf4d839886ef0283": { "model_module": "@jupyter-widgets/controls", "model_name": "ProgressStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "ProgressStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "bar_color": null, "description_width": "" } }, "a304bcda80bf4148b38ba8d30454d92b": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "06976ba3ed7b4c0f9741604ca01dec8a": { "model_module": "@jupyter-widgets/controls", "model_name": "DescriptionStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "DescriptionStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "description_width": "" } }, "25e2b5a94982477ea03255387e1a4c95": { "model_module": "@jupyter-widgets/controls", "model_name": "HBoxModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HBoxModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HBoxView", "box_style": "", "children": [ "IPY_MODEL_8329778c95f4488988e3d013bbf81493", "IPY_MODEL_af3a06f6682343dda51ed1ef58229a0a", "IPY_MODEL_63c9861278ae4a0cb2632cc6e3e9fa31" ], "layout": "IPY_MODEL_5d689c3854654e1a9552d637e855b2e9" } }, "8329778c95f4488988e3d013bbf81493": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HTMLModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HTMLView", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_07c7e9477dc848e586d193f850e4b330", "placeholder": "​", "style": "IPY_MODEL_648f89df3d71429abe155bf048661b3c", "value": "testSpider-00000-of-00001.parquet: 100%" } }, "af3a06f6682343dda51ed1ef58229a0a": { "model_module": "@jupyter-widgets/controls", "model_name": "FloatProgressModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "FloatProgressModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "ProgressView", "bar_style": "success", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_35bac25d9a394d53a643941969f11f33", "max": 648544, "min": 0, "orientation": "horizontal", "style": "IPY_MODEL_a0d882959f0b4b51969691cf6dcecfa1", "value": 648544 } }, "63c9861278ae4a0cb2632cc6e3e9fa31": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HTMLModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HTMLView", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_97b5552f17f44a34b90413fc663f83f0", "placeholder": "​", "style": "IPY_MODEL_057bfb8b3fb54a12982d1b7422c9e6ef", "value": " 649k/649k [00:00<00:00, 9.84MB/s]" } }, "5d689c3854654e1a9552d637e855b2e9": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "07c7e9477dc848e586d193f850e4b330": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "648f89df3d71429abe155bf048661b3c": { "model_module": "@jupyter-widgets/controls", "model_name": "DescriptionStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "DescriptionStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "description_width": "" } }, "35bac25d9a394d53a643941969f11f33": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "a0d882959f0b4b51969691cf6dcecfa1": { "model_module": "@jupyter-widgets/controls", "model_name": "ProgressStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "ProgressStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "bar_color": null, "description_width": "" } }, "97b5552f17f44a34b90413fc663f83f0": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "057bfb8b3fb54a12982d1b7422c9e6ef": { "model_module": "@jupyter-widgets/controls", "model_name": "DescriptionStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "DescriptionStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "description_width": "" } }, "2b9ace70a9bc401085fdbc8cced20d38": { "model_module": "@jupyter-widgets/controls", "model_name": "HBoxModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HBoxModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HBoxView", "box_style": "", "children": [ "IPY_MODEL_358831d17c974c2c95648cf9ee7c9294", "IPY_MODEL_8bb4561e208b4d2f823c782124cfa335", "IPY_MODEL_1a01eb10b70e42a58ebfedb6ccd61a73" ], "layout": "IPY_MODEL_6880557c5da649c19c855735754906d1" } }, "358831d17c974c2c95648cf9ee7c9294": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HTMLModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HTMLView", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_9dd0e981b51648d888fb724c936b6cf4", "placeholder": "​", "style": "IPY_MODEL_9c1bce42a2494449b8f3a833ef09e862", "value": "Generating trainSpider split: 100%" } }, "8bb4561e208b4d2f823c782124cfa335": { "model_module": "@jupyter-widgets/controls", "model_name": "FloatProgressModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "FloatProgressModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "ProgressView", "bar_style": "success", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_5be7663a62f44df9ac01171e421c732c", "max": 8656, "min": 0, "orientation": "horizontal", "style": "IPY_MODEL_9611ef2940f74dbab86496d99bf4d7eb", "value": 8656 } }, "1a01eb10b70e42a58ebfedb6ccd61a73": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HTMLModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HTMLView", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_1e6881df003345718b84ad9ed160a8af", "placeholder": "​", "style": "IPY_MODEL_e03975280a21420d8880f8bfd93ac44a", "value": " 8656/8656 [00:00<00:00, 13326.46 examples/s]" } }, "6880557c5da649c19c855735754906d1": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "9dd0e981b51648d888fb724c936b6cf4": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "9c1bce42a2494449b8f3a833ef09e862": { "model_module": "@jupyter-widgets/controls", "model_name": "DescriptionStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "DescriptionStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "description_width": "" } }, "5be7663a62f44df9ac01171e421c732c": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "9611ef2940f74dbab86496d99bf4d7eb": { "model_module": "@jupyter-widgets/controls", "model_name": "ProgressStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "ProgressStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "bar_color": null, "description_width": "" } }, "1e6881df003345718b84ad9ed160a8af": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "e03975280a21420d8880f8bfd93ac44a": { "model_module": "@jupyter-widgets/controls", "model_name": "DescriptionStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "DescriptionStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "description_width": "" } }, "0b37dc50f0cd4f149feb3f7ce08f2630": { "model_module": "@jupyter-widgets/controls", "model_name": "HBoxModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HBoxModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HBoxView", "box_style": "", "children": [ "IPY_MODEL_0abf618ec97e4ca8873c8af349516150", "IPY_MODEL_c9bd191293104538bdbc0a5539be4d0a", "IPY_MODEL_f3fe35e2b0164044aed92f8f776a17b5" ], "layout": "IPY_MODEL_3a9a0de33597425ab7e550ec32aad186" } }, "0abf618ec97e4ca8873c8af349516150": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HTMLModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HTMLView", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_f6357593f5f64966a4f115adbcbec940", "placeholder": "​", "style": "IPY_MODEL_f283bc804ab348a29c33a82368552fd5", "value": "Generating devSpider split: 100%" } }, "c9bd191293104538bdbc0a5539be4d0a": { "model_module": "@jupyter-widgets/controls", "model_name": "FloatProgressModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "FloatProgressModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "ProgressView", "bar_style": "success", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_3248d30b29f84bdcbe4c43594c803673", "max": 1034, "min": 0, "orientation": "horizontal", "style": "IPY_MODEL_307a99fb69894c4cac341e23119150f2", "value": 1034 } }, "f3fe35e2b0164044aed92f8f776a17b5": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HTMLModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HTMLView", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_46d219bf79384b1d98052febd0271589", "placeholder": "​", "style": "IPY_MODEL_71990f86fca248bd8d1e75be403d70db", "value": " 1034/1034 [00:00<00:00, 8442.37 examples/s]" } }, "3a9a0de33597425ab7e550ec32aad186": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "f6357593f5f64966a4f115adbcbec940": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "f283bc804ab348a29c33a82368552fd5": { "model_module": "@jupyter-widgets/controls", "model_name": "DescriptionStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "DescriptionStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "description_width": "" } }, "3248d30b29f84bdcbe4c43594c803673": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "307a99fb69894c4cac341e23119150f2": { "model_module": "@jupyter-widgets/controls", "model_name": "ProgressStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "ProgressStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "bar_color": null, "description_width": "" } }, "46d219bf79384b1d98052febd0271589": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "71990f86fca248bd8d1e75be403d70db": { "model_module": "@jupyter-widgets/controls", "model_name": "DescriptionStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "DescriptionStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "description_width": "" } }, "9793ebb3046b4e0cadc483222c6d88f9": { "model_module": "@jupyter-widgets/controls", "model_name": "HBoxModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HBoxModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HBoxView", "box_style": "", "children": [ "IPY_MODEL_4a47677ae2204165be1c7a27214afab4", "IPY_MODEL_9c1c3f497c874132b43c6c8e86bbf3e1", "IPY_MODEL_814ec31670f543e798527a887db56ea0" ], "layout": "IPY_MODEL_e022ce00981049349403502e2a07758b" } }, "4a47677ae2204165be1c7a27214afab4": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HTMLModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HTMLView", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_aad2ce00485e4468a787dbca24d0ba04", "placeholder": "​", "style": "IPY_MODEL_03994d990e7442b5ab2249403fa62467", "value": "Generating testSpider split: 100%" } }, "9c1c3f497c874132b43c6c8e86bbf3e1": { "model_module": "@jupyter-widgets/controls", "model_name": "FloatProgressModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "FloatProgressModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "ProgressView", "bar_style": "success", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_c505d4d70cfe4c23844c7fef29b79c16", "max": 2147, "min": 0, "orientation": "horizontal", "style": "IPY_MODEL_d88bd0e5dc454761bd9db02488746996", "value": 2147 } }, "814ec31670f543e798527a887db56ea0": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HTMLModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HTMLView", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_9a3733ca2451421e9ad5613f94ea7d89", "placeholder": "​", "style": "IPY_MODEL_48a9e7a72ccb4352a7393a2461716e99", "value": " 2147/2147 [00:00<00:00, 7465.95 examples/s]" } }, "e022ce00981049349403502e2a07758b": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "aad2ce00485e4468a787dbca24d0ba04": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "03994d990e7442b5ab2249403fa62467": { "model_module": "@jupyter-widgets/controls", "model_name": "DescriptionStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "DescriptionStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "description_width": "" } }, "c505d4d70cfe4c23844c7fef29b79c16": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "d88bd0e5dc454761bd9db02488746996": { "model_module": "@jupyter-widgets/controls", "model_name": "ProgressStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "ProgressStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "bar_color": null, "description_width": "" } }, "9a3733ca2451421e9ad5613f94ea7d89": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "48a9e7a72ccb4352a7393a2461716e99": { "model_module": "@jupyter-widgets/controls", "model_name": "DescriptionStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "DescriptionStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "description_width": "" } }, "bae6f37fafae4e6191e871e312da8e0a": { "model_module": "@jupyter-widgets/controls", "model_name": "HBoxModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HBoxModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HBoxView", "box_style": "", "children": [ "IPY_MODEL_414fdecff88f4274965245c29726ccee", "IPY_MODEL_05a27f58c6dc4f86ab8e8ef86da01282", "IPY_MODEL_1180ab06b3164e70b85a35bb6e03b61f" ], "layout": "IPY_MODEL_463fe03d56a64a9a8f9eabb607f04cc9" } }, "414fdecff88f4274965245c29726ccee": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HTMLModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HTMLView", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_1c27a07fdf114461a8756b465c595178", "placeholder": "​", "style": "IPY_MODEL_52b026c7f78945c9bffcd1831abb53ce", "value": "README.md: 100%" } }, "05a27f58c6dc4f86ab8e8ef86da01282": { "model_module": "@jupyter-widgets/controls", "model_name": "FloatProgressModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "FloatProgressModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "ProgressView", "bar_style": "success", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_81c9523ec3e24ad399fd02aa0ab802cc", "max": 809, "min": 0, "orientation": "horizontal", "style": "IPY_MODEL_9e3160d3aa6b43e8919ded34eee885a2", "value": 809 } }, "1180ab06b3164e70b85a35bb6e03b61f": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HTMLModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HTMLView", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_6c5845a6b262459faac5251aefad494b", "placeholder": "​", "style": "IPY_MODEL_4cdbaf0948c24fbe91d3c2846326e5bb", "value": " 809/809 [00:00<00:00, 51.3kB/s]" } }, "463fe03d56a64a9a8f9eabb607f04cc9": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "1c27a07fdf114461a8756b465c595178": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "52b026c7f78945c9bffcd1831abb53ce": { "model_module": "@jupyter-widgets/controls", "model_name": "DescriptionStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "DescriptionStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "description_width": "" } }, "81c9523ec3e24ad399fd02aa0ab802cc": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "9e3160d3aa6b43e8919ded34eee885a2": { "model_module": "@jupyter-widgets/controls", "model_name": "ProgressStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "ProgressStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "bar_color": null, "description_width": "" } }, "6c5845a6b262459faac5251aefad494b": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "4cdbaf0948c24fbe91d3c2846326e5bb": { "model_module": "@jupyter-widgets/controls", "model_name": "DescriptionStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "DescriptionStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "description_width": "" } }, "6d6b9b996f5a42709ba70025657187bb": { "model_module": "@jupyter-widgets/controls", "model_name": "HBoxModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HBoxModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HBoxView", "box_style": "", "children": [ "IPY_MODEL_e0d6b0786a0b4f60be5b8f23737dcac7", "IPY_MODEL_3e7aea0f351a4e57ae13dc417435fab4", "IPY_MODEL_b3ea00f427ef43ddb2115db97b62506e" ], "layout": "IPY_MODEL_574012abbdbf4df7b71a722b037f55e9" } }, "e0d6b0786a0b4f60be5b8f23737dcac7": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HTMLModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HTMLView", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_8163814231524dad9a68e31821185bd7", "placeholder": "​", "style": "IPY_MODEL_59ca8dacec67490ca85fb26562ec6017", "value": "train-00000-of-00001.parquet: 100%" } }, "3e7aea0f351a4e57ae13dc417435fab4": { "model_module": "@jupyter-widgets/controls", "model_name": "FloatProgressModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "FloatProgressModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "ProgressView", "bar_style": "success", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_cf522c8fd3314b60a6ef3d2396700887", "max": 2345343, "min": 0, "orientation": "horizontal", "style": "IPY_MODEL_166da246d22a4e0c9bd28cb9dd280081", "value": 2345343 } }, "b3ea00f427ef43ddb2115db97b62506e": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HTMLModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HTMLView", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_8a573702b9364cbea6dc412b8c5c2dc6", "placeholder": "​", "style": "IPY_MODEL_6e4677a62beb45b4ad9e47218f5abf5c", "value": " 2.35M/2.35M [00:00<00:00, 36.1MB/s]" } }, "574012abbdbf4df7b71a722b037f55e9": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "8163814231524dad9a68e31821185bd7": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "59ca8dacec67490ca85fb26562ec6017": { "model_module": "@jupyter-widgets/controls", "model_name": "DescriptionStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "DescriptionStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "description_width": "" } }, "cf522c8fd3314b60a6ef3d2396700887": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "166da246d22a4e0c9bd28cb9dd280081": { "model_module": "@jupyter-widgets/controls", "model_name": "ProgressStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "ProgressStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "bar_color": null, "description_width": "" } }, "8a573702b9364cbea6dc412b8c5c2dc6": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "6e4677a62beb45b4ad9e47218f5abf5c": { "model_module": "@jupyter-widgets/controls", "model_name": "DescriptionStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "DescriptionStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "description_width": "" } }, "e71d9b9e72b94b1f8b5404d5a2b855c5": { "model_module": "@jupyter-widgets/controls", "model_name": "HBoxModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HBoxModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HBoxView", "box_style": "", "children": [ "IPY_MODEL_110dae7d8dde46f695d431676d01d2eb", "IPY_MODEL_6db8779ae2b841ddae4412a8536cc380", "IPY_MODEL_46c5d1e9321c43b5bfe9bf278a04c3cf" ], "layout": "IPY_MODEL_d4e80de4d6bd4a518516939bc5465478" } }, "110dae7d8dde46f695d431676d01d2eb": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HTMLModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HTMLView", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_0b1142c6e9324d5ea9afd0d20cfdb27e", "placeholder": "​", "style": "IPY_MODEL_bb9d06e11839461a98ef04c233d9dcf0", "value": "dev-00000-of-00001.parquet: 100%" } }, "6db8779ae2b841ddae4412a8536cc380": { "model_module": "@jupyter-widgets/controls", "model_name": "FloatProgressModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "FloatProgressModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "ProgressView", "bar_style": "success", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_07c56e1d9da5444b9c125b2e3b6e6f41", "max": 336051, "min": 0, "orientation": "horizontal", "style": "IPY_MODEL_b16c302bce92479cbe7278b8a9ac11c6", "value": 336051 } }, "46c5d1e9321c43b5bfe9bf278a04c3cf": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HTMLModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HTMLView", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_f624184ee71149ac981960a564e24374", "placeholder": "​", "style": "IPY_MODEL_382f235790084fac8737b5aeafe5adb4", "value": " 336k/336k [00:00<00:00, 13.8MB/s]" } }, "d4e80de4d6bd4a518516939bc5465478": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "0b1142c6e9324d5ea9afd0d20cfdb27e": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "bb9d06e11839461a98ef04c233d9dcf0": { "model_module": "@jupyter-widgets/controls", "model_name": "DescriptionStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "DescriptionStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "description_width": "" } }, "07c56e1d9da5444b9c125b2e3b6e6f41": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "b16c302bce92479cbe7278b8a9ac11c6": { "model_module": "@jupyter-widgets/controls", "model_name": "ProgressStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "ProgressStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "bar_color": null, "description_width": "" } }, "f624184ee71149ac981960a564e24374": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "382f235790084fac8737b5aeafe5adb4": { "model_module": "@jupyter-widgets/controls", "model_name": "DescriptionStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "DescriptionStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "description_width": "" } }, "ee04bf28eb4943cd84a078c7540f81ff": { "model_module": "@jupyter-widgets/controls", "model_name": "HBoxModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HBoxModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HBoxView", "box_style": "", "children": [ "IPY_MODEL_440ea5f3f24d405b8ea7026fbe1086d4", "IPY_MODEL_1aa17586638b4f288ad644cceda7ad26", "IPY_MODEL_dee4f6fb73b540d39fcc553b84119978" ], "layout": "IPY_MODEL_4b3865d6e80c4c7c897f763de0ab3590" } }, "440ea5f3f24d405b8ea7026fbe1086d4": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HTMLModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HTMLView", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_6a7eb1c11606484489551045f55459e4", "placeholder": "​", "style": "IPY_MODEL_8f02831366904be995a61aa911a8d0a5", "value": "Generating train split: 100%" } }, "1aa17586638b4f288ad644cceda7ad26": { "model_module": "@jupyter-widgets/controls", "model_name": "FloatProgressModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "FloatProgressModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "ProgressView", "bar_style": "success", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_a93b31ad9ca742319cce4bb304aaca91", "max": 8656, "min": 0, "orientation": "horizontal", "style": "IPY_MODEL_cd0b50cd6ac243bd928f35cc4212fc7e", "value": 8656 } }, "dee4f6fb73b540d39fcc553b84119978": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HTMLModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HTMLView", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_4660175909174679a1000ff24b3fe667", "placeholder": "​", "style": "IPY_MODEL_d97288cdaf0246518d34a73076348253", "value": " 8656/8656 [00:00<00:00, 25433.43 examples/s]" } }, "4b3865d6e80c4c7c897f763de0ab3590": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "6a7eb1c11606484489551045f55459e4": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "8f02831366904be995a61aa911a8d0a5": { "model_module": "@jupyter-widgets/controls", "model_name": "DescriptionStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "DescriptionStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "description_width": "" } }, "a93b31ad9ca742319cce4bb304aaca91": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "cd0b50cd6ac243bd928f35cc4212fc7e": { "model_module": "@jupyter-widgets/controls", "model_name": "ProgressStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "ProgressStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "bar_color": null, "description_width": "" } }, "4660175909174679a1000ff24b3fe667": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "d97288cdaf0246518d34a73076348253": { "model_module": "@jupyter-widgets/controls", "model_name": "DescriptionStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "DescriptionStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "description_width": "" } }, "b20fffcc77b0423ea45bf748c28bdddf": { "model_module": "@jupyter-widgets/controls", "model_name": "HBoxModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HBoxModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HBoxView", "box_style": "", "children": [ "IPY_MODEL_b634afe8d5544557bcde062645895c64", "IPY_MODEL_27bfcf71466846a2b9a7d1d4dae7072b", "IPY_MODEL_83c7d14e14c64785ad7bea0fecce7ff7" ], "layout": "IPY_MODEL_ab06caa7a607427cb53439fb6e73ee59" } }, "b634afe8d5544557bcde062645895c64": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HTMLModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HTMLView", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_bd8b3cc0b7b74781bd2ee88e491d1c06", "placeholder": "​", "style": "IPY_MODEL_c9e87b254cad44a7a99f666d5b16babb", "value": "Generating dev split: 100%" } }, "27bfcf71466846a2b9a7d1d4dae7072b": { "model_module": "@jupyter-widgets/controls", "model_name": "FloatProgressModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "FloatProgressModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "ProgressView", "bar_style": "success", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_003fd152378448f2b623fee331cb1810", "max": 1034, "min": 0, "orientation": "horizontal", "style": "IPY_MODEL_c3a18115f6fc4718a9221f3d2b517e8c", "value": 1034 } }, "83c7d14e14c64785ad7bea0fecce7ff7": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HTMLModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HTMLView", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_d0e4d0eefe5048ad8fc6635d68b9d0e1", "placeholder": "​", "style": "IPY_MODEL_afdb17d434a3429e8bdbafe52672d4a8", "value": " 1034/1034 [00:00<00:00, 22873.76 examples/s]" } }, "ab06caa7a607427cb53439fb6e73ee59": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "bd8b3cc0b7b74781bd2ee88e491d1c06": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "c9e87b254cad44a7a99f666d5b16babb": { "model_module": "@jupyter-widgets/controls", "model_name": "DescriptionStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "DescriptionStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "description_width": "" } }, "003fd152378448f2b623fee331cb1810": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "c3a18115f6fc4718a9221f3d2b517e8c": { "model_module": "@jupyter-widgets/controls", "model_name": "ProgressStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "ProgressStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "bar_color": null, "description_width": "" } }, "d0e4d0eefe5048ad8fc6635d68b9d0e1": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "afdb17d434a3429e8bdbafe52672d4a8": { "model_module": "@jupyter-widgets/controls", "model_name": "DescriptionStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "DescriptionStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "description_width": "" } }, "afe6382aee5b4b90875101e2caf9628f": { "model_module": "@jupyter-widgets/controls", "model_name": "HBoxModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HBoxModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HBoxView", "box_style": "", "children": [ "IPY_MODEL_556724428f0745ab94113142eb4d0ea7", "IPY_MODEL_7cf72f76971f4c79a48aecd5675e9c5e", "IPY_MODEL_c825ba6e1a1d4380be685c1d9475039f" ], "layout": "IPY_MODEL_5c55529e0fd2470a8e3a7c0b7d07a197" } }, "556724428f0745ab94113142eb4d0ea7": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HTMLModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HTMLView", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_67e0ccb6c4e6446280a0fe0159f6b798", "placeholder": "​", "style": "IPY_MODEL_1fee1e975fc843ea8c9ebfbdc3198b0f", "value": "Map: 100%" } }, "7cf72f76971f4c79a48aecd5675e9c5e": { "model_module": "@jupyter-widgets/controls", "model_name": "FloatProgressModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "FloatProgressModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "ProgressView", "bar_style": "success", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_780f842bcea24ae29f38c33318f41b15", "max": 8569, "min": 0, "orientation": "horizontal", "style": "IPY_MODEL_d7583ff4442544618d108139f9a04b83", "value": 8569 } }, "c825ba6e1a1d4380be685c1d9475039f": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HTMLModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HTMLView", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_2420d610b2f740d9b33cc1b347984e83", "placeholder": "​", "style": "IPY_MODEL_7c87afd0eb39476d8f750cec181f61c2", "value": " 8569/8569 [00:11<00:00, 852.35 examples/s]" } }, "5c55529e0fd2470a8e3a7c0b7d07a197": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "67e0ccb6c4e6446280a0fe0159f6b798": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "1fee1e975fc843ea8c9ebfbdc3198b0f": { "model_module": "@jupyter-widgets/controls", "model_name": "DescriptionStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "DescriptionStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "description_width": "" } }, "780f842bcea24ae29f38c33318f41b15": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "d7583ff4442544618d108139f9a04b83": { "model_module": "@jupyter-widgets/controls", "model_name": "ProgressStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "ProgressStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "bar_color": null, "description_width": "" } }, "2420d610b2f740d9b33cc1b347984e83": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "7c87afd0eb39476d8f750cec181f61c2": { "model_module": "@jupyter-widgets/controls", "model_name": "DescriptionStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "DescriptionStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "description_width": "" } }, "a7d961ad69274d879190bb3df2c878de": { "model_module": "@jupyter-widgets/controls", "model_name": "HBoxModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HBoxModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HBoxView", "box_style": "", "children": [ "IPY_MODEL_fab89113c0ef4618b9c1f82d2f300ca7", "IPY_MODEL_86a368d742e641ff96e84a78c19ed96c", "IPY_MODEL_7eb8e2a7094b476aad48c35e7a761344" ], "layout": "IPY_MODEL_cb58424d2a904489956e4f10cb7972b3" } }, "fab89113c0ef4618b9c1f82d2f300ca7": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HTMLModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HTMLView", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_345cd61b394f4a58b554be446c75c8b7", "placeholder": "​", "style": "IPY_MODEL_46d3a1f2e16c4ec2bef78da42bc58dc7", "value": "Map: 100%" } }, "86a368d742e641ff96e84a78c19ed96c": { "model_module": "@jupyter-widgets/controls", "model_name": "FloatProgressModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "FloatProgressModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "ProgressView", "bar_style": "success", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_ac7afbd6e6504ac78d481994f0a29ff5", "max": 87, "min": 0, "orientation": "horizontal", "style": "IPY_MODEL_f394a935ac4a438bb0749849cad08494", "value": 87 } }, "7eb8e2a7094b476aad48c35e7a761344": { "model_module": "@jupyter-widgets/controls", "model_name": "HTMLModel", "model_module_version": "1.5.0", "state": { "_dom_classes": [], "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "HTMLModel", "_view_count": null, "_view_module": "@jupyter-widgets/controls", "_view_module_version": "1.5.0", "_view_name": "HTMLView", "description": "", "description_tooltip": null, "layout": "IPY_MODEL_23869c6cbab647d89abdb1bba17fd37a", "placeholder": "​", "style": "IPY_MODEL_2c2ca8a948ed4627891113d8de8f9a00", "value": " 87/87 [00:00<00:00, 719.03 examples/s]" } }, "cb58424d2a904489956e4f10cb7972b3": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "345cd61b394f4a58b554be446c75c8b7": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "46d3a1f2e16c4ec2bef78da42bc58dc7": { "model_module": "@jupyter-widgets/controls", "model_name": "DescriptionStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "DescriptionStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "description_width": "" } }, "ac7afbd6e6504ac78d481994f0a29ff5": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "f394a935ac4a438bb0749849cad08494": { "model_module": "@jupyter-widgets/controls", "model_name": "ProgressStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "ProgressStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "bar_color": null, "description_width": "" } }, "23869c6cbab647d89abdb1bba17fd37a": { "model_module": "@jupyter-widgets/base", "model_name": "LayoutModel", "model_module_version": "1.2.0", "state": { "_model_module": "@jupyter-widgets/base", "_model_module_version": "1.2.0", "_model_name": "LayoutModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "LayoutView", "align_content": null, "align_items": null, "align_self": null, "border": null, "bottom": null, "display": null, "flex": null, "flex_flow": null, "grid_area": null, "grid_auto_columns": null, "grid_auto_flow": null, "grid_auto_rows": null, "grid_column": null, "grid_gap": null, "grid_row": null, "grid_template_areas": null, "grid_template_columns": null, "grid_template_rows": null, "height": null, "justify_content": null, "justify_items": null, "left": null, "margin": null, "max_height": null, "max_width": null, "min_height": null, "min_width": null, "object_fit": null, "object_position": null, "order": null, "overflow": null, "overflow_x": null, "overflow_y": null, "padding": null, "right": null, "top": null, "visibility": null, "width": null } }, "2c2ca8a948ed4627891113d8de8f9a00": { "model_module": "@jupyter-widgets/controls", "model_name": "DescriptionStyleModel", "model_module_version": "1.5.0", "state": { "_model_module": "@jupyter-widgets/controls", "_model_module_version": "1.5.0", "_model_name": "DescriptionStyleModel", "_view_count": null, "_view_module": "@jupyter-widgets/base", "_view_module_version": "1.2.0", "_view_name": "StyleView", "description_width": "" } } } } }, "nbformat": 4, "nbformat_minor": 0 }