diff --git a/freqtrade/freqai/prediction_models/LightGBMClassifier.py b/freqtrade/freqai/prediction_models/LightGBMClassifier.py
index e17f7417c..0b78c4129 100644
--- a/freqtrade/freqai/prediction_models/LightGBMClassifier.py
+++ b/freqtrade/freqai/prediction_models/LightGBMClassifier.py
@@ -1,10 +1,12 @@
 import logging
+from collections.abc import Callable
 from typing import Any
 
 from lightgbm import LGBMClassifier
 
 from freqtrade.freqai.base_models.BaseClassifierModel import BaseClassifierModel
 from freqtrade.freqai.data_kitchen import FreqaiDataKitchen
+from freqtrade.freqai.tensorboard import LightGBMCallback
 
 
 logger = logging.getLogger(__name__)
@@ -46,6 +48,10 @@ class LightGBMClassifier(BaseClassifierModel):
         init_model = self.get_init_model(dk.pair)
 
         model = LGBMClassifier(**self.model_training_parameters)
+        activate_tensorboard = self.freqai_info.get("activate_tensorboard", True)
+        callbacks: list[Callable[..., Any]] = []
+        if LightGBMCallback is not None:
+            callbacks = [LightGBMCallback(dk.data_path, activate_tensorboard)]
         model.fit(
             X=X,
             y=y,
@@ -53,6 +59,7 @@ class LightGBMClassifier(BaseClassifierModel):
             sample_weight=train_weights,
             eval_sample_weight=[test_weights],
             init_model=init_model,
+            callbacks=callbacks,
         )
 
         return model
diff --git a/freqtrade/freqai/prediction_models/LightGBMClassifierMultiTarget.py b/freqtrade/freqai/prediction_models/LightGBMClassifierMultiTarget.py
index 9fb775614..4e4d981b8 100644
--- a/freqtrade/freqai/prediction_models/LightGBMClassifierMultiTarget.py
+++ b/freqtrade/freqai/prediction_models/LightGBMClassifierMultiTarget.py
@@ -6,6 +6,7 @@ from lightgbm import LGBMClassifier
 from freqtrade.freqai.base_models.BaseClassifierModel import BaseClassifierModel
 from freqtrade.freqai.base_models.FreqaiMultiOutputClassifier import FreqaiMultiOutputClassifier
 from freqtrade.freqai.data_kitchen import FreqaiDataKitchen
+from freqtrade.freqai.tensorboard import LightGBMCallback
 
 
 logger = logging.getLogger(__name__)
@@ -53,6 +54,11 @@ class LightGBMClassifierMultiTarget(BaseClassifierModel):
         else:
             init_models = [None] * y.shape[1]
 
+        activate_tensorboard = self.freqai_info.get("activate_tensorboard", True)
+        callbacks = []
+        if LightGBMCallback is not None:
+            callbacks = [LightGBMCallback(dk.data_path, activate_tensorboard)]
+
         fit_params = []
         for i in range(len(eval_sets)):
             fit_params.append(
@@ -60,6 +66,7 @@ class LightGBMClassifierMultiTarget(BaseClassifierModel):
                     "eval_set": eval_sets[i],
                     "eval_sample_weight": eval_weights,
                     "init_model": init_models[i],
+                    "callbacks": callbacks,
                 }
             )
 
diff --git a/freqtrade/freqai/prediction_models/LightGBMRegressor.py b/freqtrade/freqai/prediction_models/LightGBMRegressor.py
index d55cd0ca2..abd838eee 100644
--- a/freqtrade/freqai/prediction_models/LightGBMRegressor.py
+++ b/freqtrade/freqai/prediction_models/LightGBMRegressor.py
@@ -1,10 +1,12 @@
 import logging
+from collections.abc import Callable
 from typing import Any
 
 from lightgbm import LGBMRegressor
 
 from freqtrade.freqai.base_models.BaseRegressionModel import BaseRegressionModel
 from freqtrade.freqai.data_kitchen import FreqaiDataKitchen
+from freqtrade.freqai.tensorboard import LightGBMCallback
 
 
 logger = logging.getLogger(__name__)
@@ -42,6 +44,11 @@ class LightGBMRegressor(BaseRegressionModel):
 
         model = LGBMRegressor(**self.model_training_parameters)
 
+        activate_tensorboard = self.freqai_info.get("activate_tensorboard", True)
+        callbacks: list[Callable[..., Any]] = []
+        if LightGBMCallback is not None:
+            callbacks = [LightGBMCallback(dk.data_path, activate_tensorboard)]
+
         model.fit(
             X=X,
             y=y,
@@ -49,6 +56,7 @@ class LightGBMRegressor(BaseRegressionModel):
             sample_weight=train_weights,
             eval_sample_weight=[eval_weights],
             init_model=init_model,
+            callbacks=callbacks,
         )
 
         return model
diff --git a/freqtrade/freqai/prediction_models/LightGBMRegressorMultiTarget.py b/freqtrade/freqai/prediction_models/LightGBMRegressorMultiTarget.py
index c4669a79d..8f374190b 100644
--- a/freqtrade/freqai/prediction_models/LightGBMRegressorMultiTarget.py
+++ b/freqtrade/freqai/prediction_models/LightGBMRegressorMultiTarget.py
@@ -6,6 +6,7 @@ from lightgbm import LGBMRegressor
 from freqtrade.freqai.base_models.BaseRegressionModel import BaseRegressionModel
 from freqtrade.freqai.base_models.FreqaiMultiOutputRegressor import FreqaiMultiOutputRegressor
 from freqtrade.freqai.data_kitchen import FreqaiDataKitchen
+from freqtrade.freqai.tensorboard import LightGBMCallback
 
 
 logger = logging.getLogger(__name__)
@@ -55,6 +56,11 @@ class LightGBMRegressorMultiTarget(BaseRegressionModel):
         else:
             init_models = [None] * y.shape[1]
 
+        activate_tensorboard = self.freqai_info.get("activate_tensorboard", True)
+        callbacks = []
+        if LightGBMCallback is not None:
+            callbacks = [LightGBMCallback(dk.data_path, activate_tensorboard)]
+
         fit_params = []
         for i in range(len(eval_sets)):
             fit_params.append(
@@ -62,6 +68,7 @@ class LightGBMRegressorMultiTarget(BaseRegressionModel):
                     "eval_set": eval_sets[i],
                     "eval_sample_weight": eval_weights,
                     "init_model": init_models[i],
+                    "callbacks": callbacks,
                 }
             )
 
diff --git a/freqtrade/freqai/tensorboard/__init__.py b/freqtrade/freqai/tensorboard/__init__.py
index 183c25b22..68d045bf8 100644
--- a/freqtrade/freqai/tensorboard/__init__.py
+++ b/freqtrade/freqai/tensorboard/__init__.py
@@ -1,9 +1,11 @@
 # ensure users can still use a non-torch freqai version
 try:
+    from freqtrade.freqai.tensorboard.lightgbm_callback import LightGBMTensorboardCallback
     from freqtrade.freqai.tensorboard.tensorboard import TensorBoardCallback, TensorboardLogger
 
     TBLogger = TensorboardLogger
     TBCallback = TensorBoardCallback
+    LightGBMCallback = LightGBMTensorboardCallback
 except ModuleNotFoundError:
     from freqtrade.freqai.tensorboard.base_tensorboard import (
         BaseTensorBoardCallback,
@@ -12,5 +14,6 @@ except ModuleNotFoundError:
 
     TBLogger = BaseTensorboardLogger  # type: ignore
     TBCallback = BaseTensorBoardCallback  # type: ignore
+    LightGBMCallback = None  # type: ignore
 
-__all__ = ("TBLogger", "TBCallback")
+__all__ = ("TBLogger", "TBCallback", "LightGBMCallback")
diff --git a/freqtrade/freqai/tensorboard/lightgbm_callback.py b/freqtrade/freqai/tensorboard/lightgbm_callback.py
new file mode 100644
index 000000000..c71a43ec5
--- /dev/null
+++ b/freqtrade/freqai/tensorboard/lightgbm_callback.py
@@ -0,0 +1,24 @@
+from __future__ import annotations
+
+from freqtrade.freqai.tensorboard.tensorboard import TensorboardLogger
+
+
+class LightGBMTensorboardCallback:
+    def __init__(self, logdir, activate: bool) -> None:
+        self.activate = activate
+        self.logger = TensorboardLogger(logdir, activate)
+
+    def __call__(self, env) -> None:
+        if not self.activate:
+            return
+
+        evals = getattr(env, "evaluation_result_list", None)
+        if not evals:
+            return
+
+        for data_name, metric_name, value, _ in evals:
+            self.logger.log_scalar(f"{data_name}-{metric_name}", value, env.iteration)
+
+        end_iteration = getattr(env, "end_iteration", None)
+        if end_iteration is not None and env.iteration + 1 >= end_iteration:
+            self.logger.close()