From 5ca1779d7da2e2360924c337b9541a70a2511808 Mon Sep 17 00:00:00 2001 From: crypto_predictor Date: Sat, 7 Feb 2026 22:28:46 +0800 Subject: [PATCH] =?UTF-8?q?fix:=20=E7=B2=BE=E7=AE=80=20requirements.txt?= =?UTF-8?q?=EF=BC=8C=E7=A7=BB=E9=99=A4=E6=9C=AA=E4=BD=BF=E7=94=A8=E7=9A=84?= =?UTF-8?q?=E4=BE=9D=E8=B5=96?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit 移除的依赖： - gluonts: 项目中未使用 - pytorch-lightning: 非必需 - torchmetrics: 非必需 - onnx/onnxruntime: 非核心功能 - sqlalchemy: 使用 pymysql 直连 - jupyter/matplotlib/seaborn/pytest: 开发工具，非运行必需 保留的核心依赖： - pandas, numpy, scikit-learn: 数据处理 - granite-tsfm: TTM-R2 模型 - pandas-ta: 技术指标 - fastapi, uvicorn: API 服务 - httpx, python-binance: 数据获取 - schedule: 任务调度 - pymysql, cryptography: 数据库 注意：PyTorch 需要单独安装（Windows 用 CPU 版本） --- install_windows.bat | 12 ++++-------- requirements.txt | 30 ++++-------------------------- 2 files changed, 8 insertions(+), 34 deletions(-) diff --git a/install_windows.bat b/install_windows.bat index fecd283..7814796 100644 --- a/install_windows.bat +++ b/install_windows.bat @@ -34,18 +34,14 @@ echo [3/6] 安装 PyTorch (CPU 版本)... pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cpu echo. -echo [4/6] 安装项目依赖... +echo [4/5] 安装项目依赖... pip install -r requirements.txt echo. -echo [5/6] 安装 IBM TTM-R2 模型库... -echo 包名: granite-tsfm (导入时用 tsfm_public) -pip install granite-tsfm - -echo. -echo [6/6] 验证安装... +echo [5/5] 验证安装... python -c "import torch; print(f'PyTorch: {torch.__version__}')" -python -c "import tsfm_public; print('tsfm_public: 已安装')" +python -c "from tsfm_public.models.tinytimemixer import TinyTimeMixerForPrediction; print('TTM-R2: 已安装')" +python -c "import pandas_ta; print('pandas-ta: 已安装')" python -c "from training.config import TrainingConfig; c=TrainingConfig(); print(f'设备: {c.get_device()}')" echo. diff --git a/requirements.txt b/requirements.txt index 8f4c3b1..758840e 100644 --- a/requirements.txt +++ b/requirements.txt @@ -4,15 +4,11 @@ # ============================================== # ---------------------------------------------- -# PyTorch (根据平台安装) +# PyTorch (根据平台单独安装) # ---------------------------------------------- # macOS: pip install torch torchvision torchaudio # Windows CPU: pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cpu -torch>=2.0.0 -torchvision>=0.15.0 -torchaudio>=2.0.0 -pytorch-lightning>=2.0.0 -torchmetrics>=1.0.0 +# 注意：requirements.txt 中不包含 torch，需要手动安装 # ---------------------------------------------- # 数据处理 @@ -25,22 +21,14 @@ pytz # ---------------------------------------------- # 时序模型 # ---------------------------------------------- -# IBM Granite TTM-R2 (PyPI 包名: granite-tsfm) +# IBM Granite TTM-R2 (导入名: tsfm_public) granite-tsfm>=0.3.0 -# GluonTS 时序工具 -gluonts[torch]>=0.14.0 # ---------------------------------------------- # 技术指标 # ---------------------------------------------- pandas-ta>=0.3.14b -# ---------------------------------------------- -# ONNX 模型导出 -# ---------------------------------------------- -onnx>=1.14.0 -onnxruntime>=1.15.0 - # ---------------------------------------------- # API 服务 # ---------------------------------------------- @@ -68,7 +56,6 @@ schedule>=1.2.0 # 数据库 # ---------------------------------------------- pymysql>=1.1.0 -sqlalchemy>=2.0.0 cryptography>=41.0.0 # ---------------------------------------------- @@ -77,15 +64,6 @@ cryptography>=41.0.0 python-dotenv>=1.0.0 # ---------------------------------------------- -# 开发工具 (可选) +# 进度条 # ---------------------------------------------- -jupyter>=1.0.0 -matplotlib>=3.7.0 -seaborn>=0.12.0 -pytest>=7.4.0 tqdm>=4.66.0 - -# ---------------------------------------------- -# Polymarket 交易 (可选，需要手动安装) -# ---------------------------------------------- -# pip install py-clob-client