[feat] fix a bug and adapt general_nn for use with rdagent_qlib (#1928)

* update qlib general_nn for rdagent_qlib

* fix install lightgbm error

* fix install lightgbm error & format with black

---------

Co-authored-by: Linlang <[email protected]>

Author: WinstonLiyt

.github/workflows/test_qlib_from_pip.yml

@@ -46,10 +46,13 @@ jobs:
         python -m pip install pyqlib
         python -m pip install "joblib<=1.4.2"
 
+    # install.sh file contents from: https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh
+    # brew_install.sh file contents from: https://raw.githubusercontent.com/Microsoft/qlib/main/.github/brew_install.sh
     - name: Install Lightgbm for MacOS
       if: ${{ matrix.os == 'macos-13' || matrix.os == 'macos-14' || matrix.os == 'macos-15' }}
       run: |
-        /bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Microsoft/qlib/main/.github/brew_install.sh)"
+        /bin/bash -c "$(curl -fsSL https://github.com/SunsetWolf/qlib_dataset/releases/download/maocs_lightgbm/install.sh)"
+        /bin/bash -c "$(curl -fsSL https://github.com/SunsetWolf/qlib_dataset/releases/download/maocs_lightgbm/brew_install.sh)"
         HOMEBREW_NO_AUTO_UPDATE=1 brew install lightgbm
         # FIX MacOS error: Segmentation fault
         # reference: https://github.com/microsoft/LightGBM/issues/4229

.github/workflows/test_qlib_from_source.yml

@@ -83,10 +83,13 @@ jobs:
         python scripts/get_data.py qlib_data --name qlib_data_simple --target_dir ~/.qlib/qlib_data/cn_data --interval 1d --region cn
         python scripts/get_data.py download_data --file_name rl_data.zip --target_dir tests/.data/rl
 
+    # install.sh file contents from: https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh
+    # brew_install.sh file contents from: https://raw.githubusercontent.com/Microsoft/qlib/main/.github/brew_install.sh
     - name: Install Lightgbm for MacOS
       if: ${{ matrix.os == 'macos-13' || matrix.os == 'macos-14' || matrix.os == 'macos-15' }}
       run: |
-        /bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Microsoft/qlib/main/.github/brew_install.sh)"
+        /bin/bash -c "$(curl -fsSL https://github.com/SunsetWolf/qlib_dataset/releases/download/maocs_lightgbm/install.sh)"
+        /bin/bash -c "$(curl -fsSL https://github.com/SunsetWolf/qlib_dataset/releases/download/maocs_lightgbm/brew_install.sh)"
         HOMEBREW_NO_AUTO_UPDATE=1 brew install lightgbm
         # FIX MacOS error: Segmentation fault
         # reference: https://github.com/microsoft/LightGBM/issues/4229

.github/workflows/test_qlib_from_source_slow.yml

@@ -37,10 +37,13 @@ jobs:
       run: |
         python scripts/get_data.py qlib_data --name qlib_data_simple --target_dir ~/.qlib/qlib_data/cn_data --interval 1d --region cn
 
+    # install.sh file contents from: https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh
+    # brew_install.sh file contents from: https://raw.githubusercontent.com/Microsoft/qlib/main/.github/brew_install.sh
     - name: Install Lightgbm for MacOS
       if: ${{ matrix.os == 'macos-13' || matrix.os == 'macos-14' || matrix.os == 'macos-15' }}
       run: |
-        /bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Microsoft/qlib/main/.github/brew_install.sh)"
+        /bin/bash -c "$(curl -fsSL https://github.com/SunsetWolf/qlib_dataset/releases/download/maocs_lightgbm/install.sh)"
+        /bin/bash -c "$(curl -fsSL https://github.com/SunsetWolf/qlib_dataset/releases/download/maocs_lightgbm/brew_install.sh)"
         HOMEBREW_NO_AUTO_UPDATE=1 brew install lightgbm
         # FIX MacOS error: Segmentation fault
         # reference: https://github.com/microsoft/LightGBM/issues/4229

qlib/contrib/model/pytorch_general_nn.py

@@ -13,6 +13,7 @@
 
 import torch
 import torch.optim as optim
+from torch.optim.lr_scheduler import ReduceLROnPlateau
 
 from qlib.data.dataset.weight import Reweighter
 
@@ -136,6 +137,10 @@ def __init__(
         else:
             raise NotImplementedError("optimizer {} is not supported!".format(optimizer))
 
+        # === ReduceLROnPlateau learning rate scheduler ===
+        self.lr_scheduler = ReduceLROnPlateau(
+            self.train_optimizer, mode="min", factor=0.5, patience=5, min_lr=1e-6, threshold=1e-5
+        )
         self.fitted = False
         self.dnn_model.to(self.device)
 
@@ -154,15 +159,15 @@ def loss_fn(self, pred, label, weight=None):
             weight = torch.ones_like(label)
 
         if self.loss == "mse":
-            return self.mse(pred[mask], label[mask], weight[mask])
+            return self.mse(pred[mask], label[mask].view(-1, 1), weight[mask])
 
         raise ValueError("unknown loss `%s`" % self.loss)
 
     def metric_fn(self, pred, label):
         mask = torch.isfinite(label)
 
         if self.metric in ("", "loss"):
-            return -self.loss_fn(pred[mask], label[mask])
+            return self.loss_fn(pred[mask], label[mask])
 
         raise ValueError("unknown metric `%s`" % self.metric)
 
@@ -238,6 +243,8 @@ def fit(
 
         dl_train = dataset.prepare("train", col_set=["feature", "label"], data_key=DataHandlerLP.DK_L)
         dl_valid = dataset.prepare("valid", col_set=["feature", "label"], data_key=DataHandlerLP.DK_L)
+        self.logger.info(f"Train samples: {len(dl_train)}")
+        self.logger.info(f"Valid samples: {len(dl_valid)}")
         if dl_train.empty or dl_valid.empty:
             raise ValueError("Empty data from dataset, please check your dataset config.")
 
@@ -279,7 +286,7 @@ def fit(
 
         stop_steps = 0
         train_loss = 0
-        best_score = -np.inf
+        best_score = np.inf
         best_epoch = 0
         evals_result["train"] = []
         evals_result["valid"] = []
@@ -295,13 +302,18 @@ def fit(
             self.logger.info("evaluating...")
             train_loss, train_score = self.test_epoch(train_loader)
             val_loss, val_score = self.test_epoch(valid_loader)
-            self.logger.info("train %.6f, valid %.6f" % (train_score, val_score))
+            self.logger.info("Epoch%d: train %.6f, valid %.6f" % (step, train_score, val_score))
             evals_result["train"].append(train_score)
             evals_result["valid"].append(val_score)
 
+            # current_lr = self.train_optimizer.param_groups[0]["lr"]
+            # self.logger.info("Current learning rate: %.6e" % current_lr)
+
+            self.lr_scheduler.step(val_score)
+
             if step == 0:
                 best_param = copy.deepcopy(self.dnn_model.state_dict())
-            if val_score > best_score:
+            if val_score < best_score:
                 best_score = val_score
                 stop_steps = 0
                 best_epoch = step
@@ -312,7 +324,7 @@ def fit(
                     self.logger.info("early stop")
                     break
 
-        self.logger.info("best score: %.6lf @ %d" % (best_score, best_epoch))
+        self.logger.info("best score: %.6lf @ %d epoch" % (best_score, best_epoch))
         self.dnn_model.load_state_dict(best_param)
         torch.save(best_param, save_path)
 
@@ -329,6 +341,7 @@ def predict(
             raise ValueError("model is not fitted yet!")
 
         dl_test = dataset.prepare("test", col_set=["feature", "label"], data_key=DataHandlerLP.DK_I)
+        self.logger.info(f"Test samples: {len(dl_test)}")
 
         if isinstance(dataset, TSDatasetH):
             dl_test.config(fillna_type="ffill+bfill")  # process nan brought by dataloader