[Bug Report] XGBoost problem in ml_ops/sm-mlflow_pipelines/sm-mlflow_pipelines.ipynb

**Link to the notebook**
[ ml_ops/sm-mlflow_pipelines/sm-mlflow_pipelines.ipynb](https://github.com/aws/amazon-sagemaker-examples/blob/default/%20%20%20ml_ops/sm-mlflow_pipelines/sm-mlflow_pipelines.ipynb)

**Describe the bug**
Pipeline execution fails due to two different version installations of XGboost. One by Conda from the Sagemaker Distribution Image. The other from Pip in `requirements.txt`.

error log in CloudWatch:
```
* XGBoost is first installed with anaconda then upgraded with pip. To fix it please remove one of the installations.
```

**To reproduce**
Run the notebook on SageMaker Studio.

**Fix**
1. remove version number for XGBoost in `requirements.txt`
`xgboost==1.7.6` -> `xgboost`

2. move `early_stopping_rounds=5` to the instanciation of the XGBClassifier
```
xgb = XGBClassifier(n_estimators=num_round, early_stopping_rounds=5, **param)
xgb.fit(
train_df,
y_train,
eval_set=[(validation_df, y_validation)]
)
```

Open ml_ops/sm-mlflow_pipelines/sm-mlflow_pipelines.ipynb and its referenced requirements.txt, then inspect the XGBoost installation and XGBClassifier setup described in the issue. Apply the two requested changes and rerun the notebook on SageMaker Studio; done means the pipeline executes without the duplicate-installation error.

Written by the indexing model from the issue text.