Порнофильмы

Веб-сервис почасового прогноза нормализованной мощности двух турбин на 48 ч с отдельной CatBoost-моделью на каждую турбину, картой ВЭС и…

Рабочий MVP

HackAlem ML Дашборды Энергетика

GitHub

Команда «Порнофильмы» 11💬 0 комментариевс 23 сентября 2026 г.

О проекте

Команда: Порнофильмы
Кейс HackAlem.ai: Энергетика — Самрук-Казына
Стадия: Рабочий MVP

Задача кейса: AI-агент, прогнозирующий выработку ветровой электростанции по погодным данным

Суть

Веб-сервис почасового прогноза нормализованной мощности двух турбин на 48 ч с отдельной CatBoost-моделью на каждую турбину, картой ВЭС и интерактивным графиком; развёрнут онлайн.

Как устроено

Vue-фронтенд создаёт тикет → FastAPI собирает погоду из CSV организатора и дополняет её из Open-Meteo (архивные и текущие GFS), атомарно пишет data.csv → встроенный CatBoost-воркер (80 признаков: 168 ч погодной истории, будущая погода, календарь, номер часа) пишет response.json → UI опрашивает результат. Мощность в прогноз не подаётся, есть проверки SHA-256 моделей и полноты 48 часов. В экспериментах пробовали SARIMA и прямые модели на каждый горизонт h01–h48.

Ключевые решения

CatBoost
прямые модели по горизонтам (h01–h48)
SARIMA (эксперименты)
обмен через файлы с атомарным переименованием
файловые блокировки

Особенности

Модельный воркер отделён от API через «тикеты» на файловой системе с атомарной записью, поэтому ML можно запускать в отдельном процессе; обученные модели и данные лежат в репозитории, работает без ключей и GPU; есть живой деплой.

Стек

Python · FastAPI · Pydantic · Vue 3 · Vite · Leaflet · ECharts · CatBoost · Caddy · systemd · Azure

Проект создан за 5 часов на финале хакатона HackAlem.ai (23 сентября 2026, Астана). Код — BAITC-Hacks/hack-8e676a69-team.

README

README.md

Прогноз выработки ВЭС на 48 часов

Трек: Энергетика · Команда: Порнофильмы

Открыть демо · Backend · ML · Сторонние компоненты

Живой сайт проекта: https://hackalem-pornofilms.polandcentral.cloudapp.azure.com/ — проверяющий ИИ может открыть эту ссылку и посмотреть работу сервиса вживую.

Проблема

Ветер непредсказуем, и выработка ветроэлектростанции меняется в разы в течение суток. При этом оператор обязан заранее заявить плановую выработку, и каждое отклонение от заявки оплачивается на балансирующем рынке.

Чем точнее прогноз на ближайшие двое суток — тем меньше издержки за небаланс, тем лучше планируется обслуживание турбин в безветренные окна и тем предсказуемее выручка станции.

Решение

Инженер ВЭС выбирает турбину и дату — и за несколько секунд получает почасовой прогноз нормализованной мощности на 48 часов вперёд с интерактивным графиком.

Работает из коробки. Исходные CSV организатора, две обученные CatBoost-модели и модельный воркер включены в репозиторий. Личные API-ключи, GPU и отдельный процесс ML не нужны.
Реальные данные. Две турбины, ~292 000 строк измерений за период 11.03.2023–31.01.2026.
Честная временная логика. Для каждого расчёта используются только те погодные данные, которые были доступны на соответствующий момент времени — прогноз не «подглядывает» в будущее.

Интерфейс: карта ВЭС, выбор турбины и даты, график почасового прогноза на 48 часов

Слева — выбор ВЭС, даты расчёта и турбины. Справа — карта станции и график прогноза нормализованной мощности с выбором горизонта 0–48 часов.

Задача хакатона — историческое воспроизведение прогнозов за 1–28 февраля 2026 года.

1. Описание решения и назначение

Сервис предназначен для инженера или аналитика ветроэлектростанции: выбрать турбину и дату, получить прогноз на 48 часов и исследовать выбранный интервал на графике.

Пользователь задаёт турбину, начало прогноза и число предшествующих дней контекста. Backend собирает погодные признаки из исходного датасета и открытого API, формирует задание для модели и возвращает результат интерфейсу. Длина расчёта всегда 48 часов; отображаемый интервал выбирается во frontend.

Данные

Исходные CSV организатора находятся в back/data/ и входят в репозиторий.

Турбина	Координаты	Строк
A — turbine 1	43.645150, 78.535604	142 360
B — turbine 2	43.643198, 78.538828	149 499

Диапазон: 11.03.2023–31.01.2026. Поля: время, средняя скорость ветра, нормализованная активная мощность, температура. Нативный шаг — 10 минут, возможны пропуски.

Результат — 48 почасовых значений нормализованной активной мощности в диапазоне [0, 1]. Это не MW, не MWh и не измеренный КПД турбины.

Для каждой турбины используется своя CatBoost-модель с 80 признаками: история и статистики погоды, будущая погода, календарь и номер прогнозного часа. Мощность используется как целевой признак при обучении, но не подаётся на вход при прогнозировании.

2. Архитектура

flowchart LR
    UI["Vue frontend"] -->|"POST /api/tickets"|API["FastAPI"]
    CSV["CSV организатора"] --> API
    WX["Open-Meteo"] --> API
    API -->|"атомарная запись"|IN["tickets/id/data.csv"]
    IN --> ML["CatBoost-воркер: встроен в backend"]
    ML -->|"атомарная запись"|OUT["tickets/id/response.json"]
    OUT --> API
    UI -->|"GET /api/tickets/id"|API
    API -->|"JSON результата"|UI

frontend/   интерфейс, карта, графики и HTTP-клиент
back/       API, подготовка данных, настройки, зависимости, тесты и deployment
inference/  модельный воркер, две CatBoost-модели и проверяемый manifest
ml/         описание интерфейса обмена с моделью
problem/    обучение, эксперименты и отчёты
tickets/    временный обмен заданиями и результатами

Как это работает:

Frontend получает начальные настройки через /api/bootstrap, создаёт задание и опрашивает его примерно раз в секунду.
Backend берёт доступную погоду из CSV, заполняет недостающие интервалы через Open-Meteo и публикует готовый data.csv.
Встроенный ML-воркер замечает готовый файл, проверяет вход, выполняет прогноз и атомарно публикует response.json. Файловая блокировка предотвращает повторную обработку.
Backend возвращает этот JSON без изменения структуры.

Обмен с моделью — один CSV:

timestamp,phase,turbine_id,wind_speed_ms,temperature_c,weather_source

phase отделяет историю от прогнозного интервала, timestamp содержит UTC-смещение. Мощность и MWh в этот файл не включаются.

Модель использует последние 168 часов погоды (семь суток). history_days меньше 7 автоматически расширяется backend до 7. Выход всегда почасовой, независимо от ML_STEP.

3. Используемые технологии

Компонент	Технологии
Backend	Python, FastAPI, Pydantic, Uvicorn, HTTPX, tzdata
Frontend	Vue 3, Vite, Leaflet, ECharts, vue-echarts, Lucide
ML runtime	CatBoost 1.2.10, NumPy 2.4.6, pandas 3.0.6; CPU, без обучения при запросе
Погода	Open-Meteo: архивные и текущие прогнозы GFS
Обмен с ML	CSV и JSON в общей файловой системе; атомарное переименование
Проверки	pytest; сборка frontend через npm
Публикация	Caddy, HTTPS, systemd на Ubuntu

Автоматизированы получение погоды, подготовка входа, запуск модели, проверка структуры и полноты данных и выдача результата. Проверяются SHA-256 модельных файлов, входная сетка, границы контекста, полнота 48 часов и конечность выходных чисел.

4. Инструкции по установке

Системные требования

Python 3.11 или 3.12 (проверено на 3.11.4)
Node.js 22.12+ и npm (проверено на 22.22.3 / npm 10.9.8)
Git, доступ на запись в back/.cache/ и tickets/
Интернет для установки пакетов, погодного API и картографических тайлов
GPU не нужен — модели исполняются на CPU

Команды выполняются из корня репозитория.

git clone https://github.com/BAITC-Hacks/hack-8e676a69-team.git
cd hack-8e676a69-team

Windows / PowerShell

python -m venv back/.venv
./back/.venv/Scripts/python.exe -m pip install -r back/requirements.txt
Copy-Item back/.env.example back/.env
npm.cmd --prefix frontend ci
npm.cmd --prefix frontend run build

Linux / macOS

python3 -m venv back/.venv
back/.venv/bin/python -m pip install -r back/requirements.txt
cp back/.env.example back/.env
npm --prefix frontend ci
npm --prefix frontend run build

Используются два исходных CSV из back/data/. Не переименовывайте суффиксы turbine 1.csv и turbine 2.csv. Для другого расположения укажите DATASETS_DIR в back/.env.

Для проверки прогнозов по архивной погоде установите USE_DATASET_FOR_HORIZON=false в back/.env.

5. Инструкции по запуску

Сначала выполните сборку frontend из раздела 4. Затем запустите backend.

Windows:

./back/.venv/Scripts/python.exe -m uvicorn back.main:app --env-file back/.env --host 127.0.0.1 --port 8000

Linux / macOS:

back/.venv/bin/python -m uvicorn back.main:app --env-file back/.env --host 127.0.0.1 --port 8000

Откройте http://127.0.0.1:8000/. FastAPI раздаёт собранный frontend и API с одного адреса. Документация API: http://127.0.0.1:8000/docs.

При INFERENCE_ENABLED=true (значение по умолчанию) backend запускает ML-воркер автоматически — команда выше поднимает и HTTP API, и обработку тикетов моделью. Переобучение не требуется.

В репозиторий включены:

inference/models/turbine_1/catboost.cbm
inference/models/turbine_2/catboost.cbm
inference/models/manifest.json — признаки, обучающий cutoff и контрольные суммы

Запуск ML-воркера отдельным процессом

Установите INFERENCE_ENABLED=false в back/.env, затем во втором терминале из корня репозитория:

./back/.venv/Scripts/python.exe -m inference.main --watch --env-file back/.env

back/.venv/bin/python -m inference.main --watch --env-file back/.env

Оба процесса должны использовать один TICKETS_DIR и TURBINE_TIMEZONE. Спецификация — в inference/README.md.

Развёртывание на сервере

Конфигурация в back/deploy/. Systemd запускает backend из /data/app/back/.venv, читает /data/app/back/.env; Caddy обслуживает HTTPS и направляет запросы на localhost:8000. Модельный воркер запускается тем же backend и использует /data/app/tickets.

6. Необходимые зависимости и сторонние материалы

back/requirements.txt — runtime-зависимости backend
back/requirements-dev.txt — зависимости автотестов
inference/requirements.txt — ML runtime; устанавливается автоматически
problem/requirements-training.txt — обучение и эксперименты; для запуска не нужны
frontend/package.json — frontend, установка через npm ci
Исходные CSV — данные организатора
Open-Meteo — источник интернет-погоды
DISCLOSURE.md — библиотеки, данные, подготовленная инфраструктура и AI-инструменты

Личные подписки и платные ключи не требуются. Нужен только доступ в интернет для погодного API. Сохранённые модели входят в репозиторий; inference не скачивает веса.

Методика обучения — в problem/WEATHER_CONTEXT_MODELS.md.

7. Параметры окружения

Настройки backend задаются в back/.env; пример — back/.env.example. Относительные пути отсчитываются от back/ независимо от каталога запуска.

Переменная	По умолчанию	Назначение
TURBINE_TIMEZONE	Asia/Almaty	Интерпретация временных меток без смещения
DATASETS_DIR	./data	Каталог исходных CSV
TICKETS_DIR	../tickets	Общий каталог backend и ML
INFERENCE_ENABLED	true	Автоматический запуск воркера внутри backend
INFERENCE_MODEL_DIR	../inference/models	Каталог весов и manifest
INFERENCE_WORKERS	2	Параллельные модельные задания
INFERENCE_THREADS	1	CPU-потоки на один прогноз
ML_STEP	6	Число строк входа на час; 6 означает 10 минут
WEATHER_CACHE_DIR	./.cache/weather	Кэш ответов погоды
WEATHER_MODEL	gfs_global	Погодная модель
WEATHER_WIND_VARIABLE	wind_speed_100m	Переменная ветра
USE_DATASET_FOR_HORIZON	true	Погода из CSV внутри прогноза; для backtest — false
FORECAST_AVAILABILITY_MARGIN_HOURS	12	Запас при выборе архивного lead-offset
WEATHER_TIMEOUT_SECONDS	25	Тайм-аут одного запроса погоды
TICKET_TIMEOUT_SECONDS	1800	Ожидание ответа ML
RESULT_TTL_SECONDS	1800	Хранение готового ответа
MAX_ACTIVE_TICKETS	32	Максимум активных заданий
CORS_ORIGINS	*	Разрешённые origins frontend
WEB_DIST	../frontend/dist	Каталог собранного интерфейса

TURBINE_TIMEZONE=Asia/Almaty — принятое местное время, UTC+5 для февраля 2026 года. Исходные CSV не указывают timezone, поэтому это явное допущение реализации.

Frontend использует необязательную VITE_API_BASE_URL при сборке из frontend/.env. Для общей точки входа http://127.0.0.1:8000 значение должно отсутствовать или быть пустым.

8. Порядок проверки основного сценария

Откройте /health — ожидается JSON с ok=true. Один HTTP 200 с HTML не подтверждает работу API.
Откройте /api/bootstrap — ожидаются турбины A/B, диапазоны дат, prediction_hours=48 и defaults с тремя полями.
В интерфейсе выберите турбину A, дату 2026-02-05 и подтвердите создание задания. Либо через /docs:

{
  "turbine_id": "A",
  "horizon_start": "2026-02-05",
  "history_days": 30
}

Ожидается HTTP 202 и ticket_id. Backend публикует tickets/<id>/data.csv: при ML_STEP=6 это 4 320 строк истории + 288 прогнозных. История — 6 января–4 февраля, прогноз — 5–6 февраля. Колонок мощности/MWh быть не должно.
Воркер публикует response.json. GET /api/tickets/<id> должен вернуть HTTP 200 с массивом series: одна серия agent-ctboost, 48 точек hour=0…47, конечные value в диапазоне [0, 1]. Интерфейс отображает эту серию.
Повторите для турбины B и другой даты.

Критерий успеха — реальные 48 модельных точек и соответствующий график, а не один HTTP 200. Ошибка модели возвращается как JSON error; значения не подменяются тестовыми прогнозами.

Автоматические проверки

./back/.venv/Scripts/python.exe -m pip install -r back/requirements-dev.txt
./back/.venv/Scripts/python.exe -m pytest -c back/pytest.ini back/tests -q

back/.venv/bin/python -m pip install -r back/requirements-dev.txt
back/.venv/bin/python -m pytest -c back/pytest.ini back/tests -q

Тесты проверяют границы интервалов, серверное управление шагом, отсутствие целевой мощности во входе, интерполяцию, HTTP-опрос, атомарную публикацию, ошибки, перезапуск и очистку. Интеграционные проверки запускают сохранённые CatBoost-модели на автономных погодных данных.

Проверено на текущей интеграции:

49 backend/inference-тестов пройдены
npm ci и production-сборка frontend выполнены
Локальный запуск проверен вместе с /health, /api/bootstrap, собранным интерфейсом и документацией API
Для обеих турбин выполнен реальный HTTP-сценарий (начало 05.02.2026, history_days=30, ML_STEP=6, USE_DATASET_FOR_HORIZON=false): вход собран из CSV организатора и архивного погодного API, воркер вернул по 48 конечных почасовых значений, часы 0–47. Тестовые ответы модели не использовались.

Ограничения и честные оговорки

Мы явно фиксируем границы того, что подтверждено, а что нет.

Что подтверждено: работоспособность полного пути «запрос → погода → модель → результат» для обеих турбин на реальных архивных прогнозах погоды, воспроизводимость из чистого клона, 49 автотестов.

Что не заявляется:

Точность на феврале не измерена. В исходных файлах нет февральских целевых значений мощности, поэтому MAE/RMSE за февраль мы не приводим. Для измерения качества нужна отдельная временная выборка, не использованная при обучении.
Полный последовательный февральский прогон (28 дат подряд с сохранёнными моментами выпуска прогнозов) ещё не подтверждён.
Даты раньше 01.02.2026 — диагностический режим: модели обучены по 31.01.2026 включительно, поэтому такой прогноз не является независимым backtest. Воркер записывает предупреждение.
Архивные lead-offset выбираются с консервативным запасом по времени доступности; точные исторические моменты публикации каждого погодного запуска этим способом не подтверждены.
Метрики обучения получены на фактически записанной будущей погоде и не равны качеству при использовании неточных прогнозов.
Отслеживание изменений погоды без нового запроса не реализовано — каждый подтверждённый запрос запускает новый расчёт.

Публичный интерфейс — дополнительный способ просмотра. Для технической проверки остаются необходимыми воспроизводимые локальные инструкции и все модельные артефакты в итоговой версии.

GitHub

BAITC-Hacks/hack-8e676a69-team

Звёзды

⭐ 0

Форки

Язык

Python

Последний push

в прошлом месяце

Коммиты

Авторы

Файлы

654

Ветки

Авторы коммитов

dizinnes 15Nathan Yakupov 9Capta7nBlack 4

История коммитов

2026-09-23 17:51docs: add live site link for reviewersdizinnes
2026-09-23 17:44docs: document reviewer setup and include organizer datasetsCapta7nBlack
2026-09-23 17:44docs: rewrite README with problem framing and add UI screenshotCapta7nBlack
2026-09-23 17:40fix(frontend): handle forecast range errorsdizinnes
2026-09-23 17:34Revert "fix(frontend): constrain forecast start date"dizinnes
2026-09-23 17:34Merge remote-tracking branch 'origin/main'Nathan Yakupov
2026-09-23 17:33fixNathan Yakupov
2026-09-23 17:28fix(frontend): constrain forecast start datedizinnes

Все коммиты (28)

2026-09-23 17:16Merge remote-tracking branch 'origin/main'Nathan Yakupov
2026-09-23 17:16fixed blocker for old dataNathan Yakupov
2026-09-23 17:15Merge remote-tracking branch 'origin/main'dizinnes
2026-09-23 17:12fix(frontend): require explicit initial selectiondizinnes
2026-09-23 17:08final integrationNathan Yakupov
2026-09-23 17:08Merge branch 'codex/confirm-ticket-modal'dizinnes
2026-09-23 17:07feat(frontend): require ticket confirmationdizinnes
2026-09-23 17:05feat(frontend): persist request historydizinnes
2026-09-23 17:04integrated backend + inference moduleNathan Yakupov
2026-09-23 16:52adjusted for integrationNathan Yakupov
2026-09-23 16:41feat(kuat): update loaderdizinnes
2026-09-23 16:34Move map controls above wind mapdizinnes
2026-09-23 16:28feat(kuat): update frontenddizinnes
2026-09-23 16:08Refactor backend layout and make ML sampling server-controlledCapta7nBlack
2026-09-23 16:07experimenting with cat boost and sarimaNathan Yakupov
2026-09-23 15:58feat(kuat): remove `web` folder, we'll use `frontend`dizinnes
2026-09-23 15:58Merge remote-tracking branch 'origin/main'dizinnes
2026-09-23 15:57feat(kuat): create basic frontenddizinnes
2026-09-23 15:43feat: wind power forecasting backend, ticket pipeline, weather ingestionCapta7nBlack
2026-09-23 15:34init inferenceNathan Yakupov

Структура репозитория (654 файлов)

.gitignore
DISCLOSURE.md
README.md
back/.env.example
back/README.md
back/__init__.py
back/config.py
back/data/Dataset HackAlemAI для участников 11.03.2023-28.02.2026 - turbine 1.csv
back/data/Dataset HackAlemAI для участников 11.03.2023-28.02.2026 - turbine 2.csv
back/data/README.md
back/dataset.py
back/deploy/Caddyfile
back/deploy/deploy.sh
back/deploy/hackalem.service
back/inputs.py
back/main.py
back/pytest.ini
back/requirements-dev.txt
back/requirements.txt
back/run.ps1
back/run.sh
back/schemas.py
back/tests/test_backend.py
back/tests/test_bootstrap.py
back/tests/test_inference.py
back/tests/test_weather_fallback.py
back/tickets.py
back/weather.py
docs/ui.png
frontend/.gitignore
frontend/AGENTS.md
frontend/README.md
frontend/index.html
frontend/package-lock.json
frontend/package.json
frontend/public/favicon.svg
frontend/public/icons.svg
frontend/src/App.vue
frontend/src/api/forecastApi.js
frontend/src/components/AppSidebar.vue
frontend/src/components/ForecastPanel.vue
frontend/src/components/WindMap.vue
frontend/src/composables/useRotatingAgentMessage.js
frontend/src/data/windForecast.js
frontend/src/i18n/messages.js
frontend/src/main.js
frontend/src/styles/global.css
frontend/src/utils/backendForecast.js
frontend/src/utils/forecastPlot.js
frontend/src/utils/localized.js
frontend/vite.config.js
inference/README.md
inference/__init__.py
inference/cache/_
inference/config.py
inference/features.py
inference/forecast.py
inference/main.py
inference/models/manifest.json
inference/models/turbine_1/catboost.cbm
inference/models/turbine_2/catboost.cbm
inference/pipelines/__init__.py
inference/pipelines/test.py
inference/requirements.txt
inference/supervisor.py
ml/README.md
problem/DIRECT_MODELS.md
problem/Dataset HackAlemAI для участников 11.03.2023-28.02.2026 - turbine 1.csv
problem/Dataset HackAlemAI для участников 11.03.2023-28.02.2026 - turbine 2.csv
problem/README.md
problem/TRAINING.md
problem/WEATHER_CONTEXT_MODELS.md
problem/WEATHER_MODELS.md
problem/catboost_benchmark.py
problem/evaluation_models/turbine_1/catboost.cbm
problem/evaluation_models/turbine_2/catboost.cbm
problem/evaluation_models_direct/turbine_1/catboost_h01.cbm
problem/evaluation_models_direct/turbine_1/catboost_h02.cbm
problem/evaluation_models_direct/turbine_1/catboost_h03.cbm
problem/evaluation_models_direct/turbine_1/catboost_h04.cbm
problem/evaluation_models_direct/turbine_1/catboost_h05.cbm
problem/evaluation_models_direct/turbine_1/catboost_h06.cbm
problem/evaluation_models_direct/turbine_1/catboost_h07.cbm
problem/evaluation_models_direct/turbine_1/catboost_h08.cbm
problem/evaluation_models_direct/turbine_1/catboost_h09.cbm
problem/evaluation_models_direct/turbine_1/catboost_h10.cbm
problem/evaluation_models_direct/turbine_1/catboost_h11.cbm
problem/evaluation_models_direct/turbine_1/catboost_h12.cbm
problem/evaluation_models_direct/turbine_1/catboost_h13.cbm
problem/evaluation_models_direct/turbine_1/catboost_h14.cbm
problem/evaluation_models_direct/turbine_1/catboost_h15.cbm
problem/evaluation_models_direct/turbine_1/catboost_h16.cbm
problem/evaluation_models_direct/turbine_1/catboost_h17.cbm
problem/evaluation_models_direct/turbine_1/catboost_h18.cbm
problem/evaluation_models_direct/turbine_1/catboost_h19.cbm
problem/evaluation_models_direct/turbine_1/catboost_h20.cbm
problem/evaluation_models_direct/turbine_1/catboost_h21.cbm
problem/evaluation_models_direct/turbine_1/catboost_h22.cbm
problem/evaluation_models_direct/turbine_1/catboost_h23.cbm
problem/evaluation_models_direct/turbine_1/catboost_h24.cbm
problem/evaluation_models_direct/turbine_1/catboost_h25.cbm
problem/evaluation_models_direct/turbine_1/catboost_h26.cbm
problem/evaluation_models_direct/turbine_1/catboost_h27.cbm
problem/evaluation_models_direct/turbine_1/catboost_h28.cbm
problem/evaluation_models_direct/turbine_1/catboost_h29.cbm
problem/evaluation_models_direct/turbine_1/catboost_h30.cbm
problem/evaluation_models_direct/turbine_1/catboost_h31.cbm
problem/evaluation_models_direct/turbine_1/catboost_h32.cbm
problem/evaluation_models_direct/turbine_1/catboost_h33.cbm
problem/evaluation_models_direct/turbine_1/catboost_h34.cbm
problem/evaluation_models_direct/turbine_1/catboost_h35.cbm
problem/evaluation_models_direct/turbine_1/catboost_h36.cbm
problem/evaluation_models_direct/turbine_1/catboost_h37.cbm
problem/evaluation_models_direct/turbine_1/catboost_h38.cbm
problem/evaluation_models_direct/turbine_1/catboost_h39.cbm
problem/evaluation_models_direct/turbine_1/catboost_h40.cbm
problem/evaluation_models_direct/turbine_1/catboost_h41.cbm
problem/evaluation_models_direct/turbine_1/catboost_h42.cbm
problem/evaluation_models_direct/turbine_1/catboost_h43.cbm
problem/evaluation_models_direct/turbine_1/catboost_h44.cbm
problem/evaluation_models_direct/turbine_1/catboost_h45.cbm
problem/evaluation_models_direct/turbine_1/catboost_h46.cbm
problem/evaluation_models_direct/turbine_1/catboost_h47.cbm
problem/evaluation_models_direct/turbine_1/catboost_h48.cbm
problem/evaluation_models_direct/turbine_2/catboost_h01.cbm
problem/evaluation_models_direct/turbine_2/catboost_h02.cbm
problem/evaluation_models_direct/turbine_2/catboost_h03.cbm
problem/evaluation_models_direct/turbine_2/catboost_h04.cbm
problem/evaluation_models_direct/turbine_2/catboost_h05.cbm
problem/evaluation_models_direct/turbine_2/catboost_h06.cbm
problem/evaluation_models_direct/turbine_2/catboost_h07.cbm
problem/evaluation_models_direct/turbine_2/catboost_h08.cbm
problem/evaluation_models_direct/turbine_2/catboost_h09.cbm
problem/evaluation_models_direct/turbine_2/catboost_h10.cbm
problem/evaluation_models_direct/turbine_2/catboost_h11.cbm
problem/evaluation_models_direct/turbine_2/catboost_h12.cbm
problem/evaluation_models_direct/turbine_2/catboost_h13.cbm
problem/evaluation_models_direct/turbine_2/catboost_h14.cbm
problem/evaluation_models_direct/turbine_2/catboost_h15.cbm
problem/evaluation_models_direct/turbine_2/catboost_h16.cbm
problem/evaluation_models_direct/turbine_2/catboost_h17.cbm
problem/evaluation_models_direct/turbine_2/catboost_h18.cbm
problem/evaluation_models_direct/turbine_2/catboost_h19.cbm
problem/evaluation_models_direct/turbine_2/catboost_h20.cbm
problem/evaluation_models_direct/turbine_2/catboost_h21.cbm
problem/evaluation_models_direct/turbine_2/catboost_h22.cbm
problem/evaluation_models_direct/turbine_2/catboost_h23.cbm
problem/evaluation_models_direct/turbine_2/catboost_h24.cbm
problem/evaluation_models_direct/turbine_2/catboost_h25.cbm
problem/evaluation_models_direct/turbine_2/catboost_h26.cbm
problem/evaluation_models_direct/turbine_2/catboost_h27.cbm
problem/evaluation_models_direct/turbine_2/catboost_h28.cbm
problem/evaluation_models_direct/turbine_2/catboost_h29.cbm
problem/evaluation_models_direct/turbine_2/catboost_h30.cbm
problem/evaluation_models_direct/turbine_2/catboost_h31.cbm
problem/evaluation_models_direct/turbine_2/catboost_h32.cbm
problem/evaluation_models_direct/turbine_2/catboost_h33.cbm
problem/evaluation_models_direct/turbine_2/catboost_h34.cbm
problem/evaluation_models_direct/turbine_2/catboost_h35.cbm
problem/evaluation_models_direct/turbine_2/catboost_h36.cbm
problem/evaluation_models_direct/turbine_2/catboost_h37.cbm
problem/evaluation_models_direct/turbine_2/catboost_h38.cbm
problem/evaluation_models_direct/turbine_2/catboost_h39.cbm
problem/evaluation_models_direct/turbine_2/catboost_h40.cbm
problem/evaluation_models_direct/turbine_2/catboost_h41.cbm
problem/evaluation_models_direct/turbine_2/catboost_h42.cbm
problem/evaluation_models_direct/turbine_2/catboost_h43.cbm
problem/evaluation_models_direct/turbine_2/catboost_h44.cbm
problem/evaluation_models_direct/turbine_2/catboost_h45.cbm
problem/evaluation_models_direct/turbine_2/catboost_h46.cbm
problem/evaluation_models_direct/turbine_2/catboost_h47.cbm
problem/evaluation_models_direct/turbine_2/catboost_h48.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h01.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h02.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h03.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h04.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h05.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h06.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h07.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h08.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h09.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h10.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h11.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h12.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h13.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h14.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h15.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h16.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h17.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h18.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h19.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h20.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h21.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h22.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h23.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h24.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h25.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h26.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h27.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h28.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h29.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h30.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h31.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h32.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h33.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h34.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h35.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h36.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h37.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h38.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h39.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h40.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h41.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h42.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h43.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h44.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h45.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h46.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h47.cbm
problem/evaluation_models_direct_missing/turbine_1/catboost_h48.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h01.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h02.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h03.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h04.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h05.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h06.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h07.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h08.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h09.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h10.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h11.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h12.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h13.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h14.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h15.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h16.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h17.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h18.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h19.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h20.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h21.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h22.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h23.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h24.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h25.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h26.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h27.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h28.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h29.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h30.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h31.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h32.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h33.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h34.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h35.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h36.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h37.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h38.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h39.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h40.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h41.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h42.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h43.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h44.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h45.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h46.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h47.cbm
problem/evaluation_models_direct_missing/turbine_2/catboost_h48.cbm
problem/evaluation_models_weather_context_no_power/turbine_1/catboost.cbm
problem/evaluation_models_weather_context_no_power/turbine_2/catboost.cbm
problem/evaluation_models_weather_oracle/turbine_1/catboost.cbm
problem/evaluation_models_weather_oracle/turbine_2/catboost.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h01.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h02.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h03.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h04.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h05.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h06.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h07.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h08.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h09.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h10.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h11.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h12.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h13.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h14.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h15.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h16.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h17.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h18.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h19.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h20.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h21.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h22.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h23.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h24.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h25.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h26.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h27.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h28.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h29.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h30.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h31.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h32.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h33.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h34.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h35.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h36.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h37.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h38.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h39.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h40.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h41.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h42.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h43.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h44.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h45.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h46.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h47.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_base/h48.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_long/h01.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_long/h02.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_long/h03.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_long/h04.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_long/h05.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_long/h06.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_long/h07.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_long/h08.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_long/h09.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_long/h10.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_long/h11.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_long/h12.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_long/h13.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_long/h14.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h01.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h02.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h03.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h04.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h05.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h06.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h07.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h08.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h09.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h10.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h11.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h12.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h13.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h14.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h15.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h16.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h17.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h18.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h19.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h20.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h21.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h22.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h23.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h24.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h25.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h26.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h27.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h28.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h29.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h30.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h31.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h32.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h33.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h34.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h35.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h36.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h37.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h38.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h39.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h40.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h41.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h42.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h43.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h44.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h45.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h46.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h47.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_power/h48.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h01.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h02.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h03.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h04.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h05.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h06.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h07.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h08.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h09.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h10.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h11.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h12.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h13.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h14.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h15.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h16.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h17.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h18.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h19.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h20.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h21.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h22.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h23.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h24.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h25.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h26.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h27.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h28.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h29.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h30.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h31.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h32.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h33.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h34.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h35.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h36.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h37.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h38.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h39.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h40.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h41.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h42.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h43.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h44.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h45.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h46.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h47.cbm
problem/experiments/catboost_hypotheses/models/turbine_1/2025-04-01/direct_rich/h48.cbm
problem/experiments/catboost_hypotheses/progress.log
problem/forecasting.py
problem/models/turbine_1/catboost.cbm
problem/models/turbine_2/catboost.cbm
problem/models_direct/turbine_1/catboost_h01.cbm
problem/models_direct/turbine_1/catboost_h02.cbm
problem/models_direct/turbine_1/catboost_h03.cbm
problem/models_direct/turbine_1/catboost_h04.cbm
problem/models_direct/turbine_1/catboost_h05.cbm
problem/models_direct/turbine_1/catboost_h06.cbm
problem/models_direct/turbine_1/catboost_h07.cbm
problem/models_direct/turbine_1/catboost_h08.cbm
problem/models_direct/turbine_1/catboost_h09.cbm
problem/models_direct/turbine_1/catboost_h10.cbm
problem/models_direct/turbine_1/catboost_h11.cbm
problem/models_direct/turbine_1/catboost_h12.cbm
problem/models_direct/turbine_1/catboost_h13.cbm
problem/models_direct/turbine_1/catboost_h14.cbm
problem/models_direct/turbine_1/catboost_h15.cbm
problem/models_direct/turbine_1/catboost_h16.cbm
problem/models_direct/turbine_1/catboost_h17.cbm
problem/models_direct/turbine_1/catboost_h18.cbm
problem/models_direct/turbine_1/catboost_h19.cbm
problem/models_direct/turbine_1/catboost_h20.cbm
problem/models_direct/turbine_1/catboost_h21.cbm
problem/models_direct/turbine_1/catboost_h22.cbm
problem/models_direct/turbine_1/catboost_h23.cbm
problem/models_direct/turbine_1/catboost_h24.cbm
problem/models_direct/turbine_1/catboost_h25.cbm
problem/models_direct/turbine_1/catboost_h26.cbm
problem/models_direct/turbine_1/catboost_h27.cbm
problem/models_direct/turbine_1/catboost_h28.cbm
problem/models_direct/turbine_1/catboost_h29.cbm
problem/models_direct/turbine_1/catboost_h30.cbm
problem/models_direct/turbine_1/catboost_h31.cbm
problem/models_direct/turbine_1/catboost_h32.cbm
problem/models_direct/turbine_1/catboost_h33.cbm
problem/models_direct/turbine_1/catboost_h34.cbm
problem/models_direct/turbine_1/catboost_h35.cbm
problem/models_direct/turbine_1/catboost_h36.cbm
problem/models_direct/turbine_1/catboost_h37.cbm
problem/models_direct/turbine_1/catboost_h38.cbm
problem/models_direct/turbine_1/catboost_h39.cbm
problem/models_direct/turbine_1/catboost_h40.cbm
problem/models_direct/turbine_1/catboost_h41.cbm
problem/models_direct/turbine_1/catboost_h42.cbm
problem/models_direct/turbine_1/catboost_h43.cbm
problem/models_direct/turbine_1/catboost_h44.cbm
problem/models_direct/turbine_1/catboost_h45.cbm
problem/models_direct/turbine_1/catboost_h46.cbm
problem/models_direct/turbine_1/catboost_h47.cbm
problem/models_direct/turbine_1/catboost_h48.cbm
problem/models_direct/turbine_2/catboost_h01.cbm
problem/models_direct/turbine_2/catboost_h02.cbm
problem/models_direct/turbine_2/catboost_h03.cbm
problem/models_direct/turbine_2/catboost_h04.cbm
problem/models_direct/turbine_2/catboost_h05.cbm
problem/models_direct/turbine_2/catboost_h06.cbm
problem/models_direct/turbine_2/catboost_h07.cbm
problem/models_direct/turbine_2/catboost_h08.cbm
problem/models_direct/turbine_2/catboost_h09.cbm
problem/models_direct/turbine_2/catboost_h10.cbm
problem/models_direct/turbine_2/catboost_h11.cbm
problem/models_direct/turbine_2/catboost_h12.cbm
problem/models_direct/turbine_2/catboost_h13.cbm
problem/models_direct/turbine_2/catboost_h14.cbm
problem/models_direct/turbine_2/catboost_h15.cbm
problem/models_direct/turbine_2/catboost_h16.cbm
problem/models_direct/turbine_2/catboost_h17.cbm
problem/models_direct/turbine_2/catboost_h18.cbm
problem/models_direct/turbine_2/catboost_h19.cbm
problem/models_direct/turbine_2/catboost_h20.cbm
problem/models_direct/turbine_2/catboost_h21.cbm
problem/models_direct/turbine_2/catboost_h22.cbm
problem/models_direct/turbine_2/catboost_h23.cbm
problem/models_direct/turbine_2/catboost_h24.cbm
problem/models_direct/turbine_2/catboost_h25.cbm
problem/models_direct/turbine_2/catboost_h26.cbm
problem/models_direct/turbine_2/catboost_h27.cbm
problem/models_direct/turbine_2/catboost_h28.cbm
problem/models_direct/turbine_2/catboost_h29.cbm
problem/models_direct/turbine_2/catboost_h30.cbm
problem/models_direct/turbine_2/catboost_h31.cbm
problem/models_direct/turbine_2/catboost_h32.cbm
problem/models_direct/turbine_2/catboost_h33.cbm
problem/models_direct/turbine_2/catboost_h34.cbm
problem/models_direct/turbine_2/catboost_h35.cbm
problem/models_direct/turbine_2/catboost_h36.cbm
problem/models_direct/turbine_2/catboost_h37.cbm
problem/models_direct/turbine_2/catboost_h38.cbm
problem/models_direct/turbine_2/catboost_h39.cbm
problem/models_direct/turbine_2/catboost_h40.cbm
problem/models_direct/turbine_2/catboost_h41.cbm
problem/models_direct/turbine_2/catboost_h42.cbm
problem/models_direct/turbine_2/catboost_h43.cbm
problem/models_direct/turbine_2/catboost_h44.cbm
problem/models_direct/turbine_2/catboost_h45.cbm
problem/models_direct/turbine_2/catboost_h46.cbm
problem/models_direct/turbine_2/catboost_h47.cbm
problem/models_direct/turbine_2/catboost_h48.cbm
problem/models_direct_missing/turbine_1/catboost_h01.cbm
problem/models_direct_missing/turbine_1/catboost_h02.cbm
problem/models_direct_missing/turbine_1/catboost_h03.cbm
problem/models_direct_missing/turbine_1/catboost_h04.cbm
problem/models_direct_missing/turbine_1/catboost_h05.cbm
problem/models_direct_missing/turbine_1/catboost_h06.cbm
problem/models_direct_missing/turbine_1/catboost_h07.cbm
problem/models_direct_missing/turbine_1/catboost_h08.cbm
problem/models_direct_missing/turbine_1/catboost_h09.cbm
problem/models_direct_missing/turbine_1/catboost_h10.cbm
problem/models_direct_missing/turbine_1/catboost_h11.cbm
problem/models_direct_missing/turbine_1/catboost_h12.cbm
problem/models_direct_missing/turbine_1/catboost_h13.cbm
problem/models_direct_missing/turbine_1/catboost_h14.cbm
problem/models_direct_missing/turbine_1/catboost_h15.cbm
problem/models_direct_missing/turbine_1/catboost_h16.cbm
problem/models_direct_missing/turbine_1/catboost_h17.cbm
problem/models_direct_missing/turbine_1/catboost_h18.cbm
problem/models_direct_missing/turbine_1/catboost_h19.cbm
problem/models_direct_missing/turbine_1/catboost_h20.cbm
problem/models_direct_missing/turbine_1/catboost_h21.cbm
problem/models_direct_missing/turbine_1/catboost_h22.cbm
problem/models_direct_missing/turbine_1/catboost_h23.cbm
problem/models_direct_missing/turbine_1/catboost_h24.cbm
problem/models_direct_missing/turbine_1/catboost_h25.cbm
problem/models_direct_missing/turbine_1/catboost_h26.cbm
problem/models_direct_missing/turbine_1/catboost_h27.cbm
problem/models_direct_missing/turbine_1/catboost_h28.cbm
problem/models_direct_missing/turbine_1/catboost_h29.cbm
problem/models_direct_missing/turbine_1/catboost_h30.cbm
problem/models_direct_missing/turbine_1/catboost_h31.cbm
problem/models_direct_missing/turbine_1/catboost_h32.cbm
problem/models_direct_missing/turbine_1/catboost_h33.cbm
problem/models_direct_missing/turbine_1/catboost_h34.cbm
problem/models_direct_missing/turbine_1/catboost_h35.cbm
problem/models_direct_missing/turbine_1/catboost_h36.cbm
problem/models_direct_missing/turbine_1/catboost_h37.cbm
problem/models_direct_missing/turbine_1/catboost_h38.cbm
problem/models_direct_missing/turbine_1/catboost_h39.cbm
problem/models_direct_missing/turbine_1/catboost_h40.cbm
problem/models_direct_missing/turbine_1/catboost_h41.cbm
problem/models_direct_missing/turbine_1/catboost_h42.cbm
problem/models_direct_missing/turbine_1/catboost_h43.cbm
problem/models_direct_missing/turbine_1/catboost_h44.cbm
problem/models_direct_missing/turbine_1/catboost_h45.cbm
problem/models_direct_missing/turbine_1/catboost_h46.cbm
problem/models_direct_missing/turbine_1/catboost_h47.cbm
problem/models_direct_missing/turbine_1/catboost_h48.cbm
problem/models_direct_missing/turbine_2/catboost_h01.cbm
problem/models_direct_missing/turbine_2/catboost_h02.cbm
problem/models_direct_missing/turbine_2/catboost_h03.cbm
problem/models_direct_missing/turbine_2/catboost_h04.cbm
problem/models_direct_missing/turbine_2/catboost_h05.cbm
problem/models_direct_missing/turbine_2/catboost_h06.cbm
problem/models_direct_missing/turbine_2/catboost_h07.cbm
problem/models_direct_missing/turbine_2/catboost_h08.cbm
problem/models_direct_missing/turbine_2/catboost_h09.cbm
problem/models_direct_missing/turbine_2/catboost_h10.cbm
problem/models_direct_missing/turbine_2/catboost_h11.cbm
problem/models_direct_missing/turbine_2/catboost_h12.cbm
problem/models_direct_missing/turbine_2/catboost_h13.cbm
problem/models_direct_missing/turbine_2/catboost_h14.cbm
problem/models_direct_missing/turbine_2/catboost_h15.cbm
problem/models_direct_missing/turbine_2/catboost_h16.cbm
problem/models_direct_missing/turbine_2/catboost_h17.cbm
problem/models_direct_missing/turbine_2/catboost_h18.cbm
problem/models_direct_missing/turbine_2/catboost_h19.cbm
problem/models_direct_missing/turbine_2/catboost_h20.cbm
problem/models_direct_missing/turbine_2/catboost_h21.cbm
problem/models_direct_missing/turbine_2/catboost_h22.cbm
problem/models_direct_missing/turbine_2/catboost_h23.cbm
problem/models_direct_missing/turbine_2/catboost_h24.cbm
problem/models_direct_missing/turbine_2/catboost_h25.cbm
problem/models_direct_missing/turbine_2/catboost_h26.cbm
problem/models_direct_missing/turbine_2/catboost_h27.cbm
problem/models_direct_missing/turbine_2/catboost_h28.cbm
problem/models_direct_missing/turbine_2/catboost_h29.cbm
problem/models_direct_missing/turbine_2/catboost_h30.cbm
problem/models_direct_missing/turbine_2/catboost_h31.cbm
problem/models_direct_missing/turbine_2/catboost_h32.cbm
problem/models_direct_missing/turbine_2/catboost_h33.cbm
problem/models_direct_missing/turbine_2/catboost_h34.cbm
problem/models_direct_missing/turbine_2/catboost_h35.cbm
problem/models_direct_missing/turbine_2/catboost_h36.cbm
problem/models_direct_missing/turbine_2/catboost_h37.cbm
problem/models_direct_missing/turbine_2/catboost_h38.cbm
problem/models_direct_missing/turbine_2/catboost_h39.cbm
problem/models_direct_missing/turbine_2/catboost_h40.cbm
problem/models_direct_missing/turbine_2/catboost_h41.cbm
problem/models_direct_missing/turbine_2/catboost_h42.cbm
problem/models_direct_missing/turbine_2/catboost_h43.cbm
problem/models_direct_missing/turbine_2/catboost_h44.cbm
problem/models_direct_missing/turbine_2/catboost_h45.cbm
problem/models_direct_missing/turbine_2/catboost_h46.cbm
problem/models_direct_missing/turbine_2/catboost_h47.cbm
problem/models_direct_missing/turbine_2/catboost_h48.cbm
problem/models_weather_context_no_power/turbine_1/catboost.cbm
problem/models_weather_context_no_power/turbine_2/catboost.cbm
problem/models_weather_oracle/turbine_1/catboost.cbm
problem/models_weather_oracle/turbine_2/catboost.cbm
problem/plot_direct.py
problem/plot_weather.py
problem/plot_windows.py
problem/predict.py
problem/predict_weather.py
problem/report_catboost_benchmark.py
problem/reports/direct/training.log
problem/reports/direct_missing/training.log
problem/reports/training.log
problem/reports/weather_context_no_power/training.log
problem/reports/weather_oracle/training.log
problem/requirements-training.txt
problem/test_catboost_benchmark.py
problem/test_forecasting.py
problem/test_weather_context.py
problem/test_weather_forecasting.py
problem/train.py
problem/train_direct.py
problem/train_weather.py
problem/weather_context_forecasting.py
problem/weather_forecasting.py
problem/weather_module/weather_module/models/turbine_1/catboost.cbm
problem/weather_module/weather_module/models/turbine_2/catboost.cbm
tickets/.gitkeep

Ветки (2)

inference main

Обсуждение 0 комментариев

Войдите, чтобы комментировать

Пока нет комментариев — будьте первым.