diff --git a/ludwig/benchmarking/profiler.py b/ludwig/benchmarking/profiler.py
index 10332173c58..f0712314a67 100644
--- a/ludwig/benchmarking/profiler.py
+++ b/ludwig/benchmarking/profiler.py
@@ -3,15 +3,17 @@
 import logging
 import os
 import shutil
-import sys
 import threading
 import time
 from queue import Empty as EmptyQueueException
 from queue import Queue
+from subprocess import PIPE, Popen
 from typing import Any, Dict, List
+from xml.etree.ElementTree import fromstring
 
 import psutil
 import torch
+from cpuinfo import get_cpu_info
 from gpustat.core import GPUStatCollection
 
 from ludwig.benchmarking.profiler_dataclasses import profiler_dataclass_to_flat_dict, TorchProfilerMetrics
@@ -20,19 +22,38 @@
 from ludwig.globals import LUDWIG_VERSION
 from ludwig.utils.data_utils import save_json
 
-# disabling print because the following imports are verbose
-f = open(os.devnull, "w")
-sys.stdout = f
-from experiment_impact_tracker.cpu.common import get_my_cpu_info  # noqa E402
-from experiment_impact_tracker.gpu.nvidia import get_gpu_info  # noqa E402
-
-f.close()
-sys.stdout = sys.__stdout__
-
 STOP_MESSAGE = "stop"
 logger = logging.getLogger()
 
 
+def get_gpu_info():
+    """Gathers general hardware information about an nvidia GPU.
+
+    This function was copied from `experiment_impact_tracker` to get around a Pandas 2.0 breaking change impacting the
+    package. https://github.com/Breakend/experiment-impact-
+    tracker/blob/master/experiment_impact_tracker/gpu/nvidia.py#L48-L73
+    """
+    p = Popen(["nvidia-smi", "-q", "-x"], stdout=PIPE)
+    outs, errors = p.communicate()
+    xml = fromstring(outs)
+    data = []
+    driver_version = xml.findall("driver_version")[0].text
+    cuda_version = xml.findall("cuda_version")[0].text
+
+    for gpu_id, gpu in enumerate(xml.getiterator("gpu")):
+        gpu_data = {}
+        name = [x for x in gpu.getiterator("product_name")][0].text
+        memory_usage = gpu.findall("fb_memory_usage")[0]
+        total_memory = memory_usage.findall("total")[0].text
+
+        gpu_data["name"] = name
+        gpu_data["total_memory"] = total_memory
+        gpu_data["driver_version"] = driver_version
+        gpu_data["cuda_version"] = cuda_version
+        data.append(gpu_data)
+    return data
+
+
 def monitor(queue: Queue, info: Dict[str, Any], logging_interval: int, cuda_is_available: bool) -> None:
     """Monitors hardware resource use.
 
@@ -132,7 +153,7 @@ def _populate_static_information(self) -> None:
         self.info["start_disk_usage"] = shutil.disk_usage(os.path.expanduser("~")).used
 
         # CPU information
-        cpu_info = get_my_cpu_info()
+        cpu_info = get_cpu_info()
         self.info["cpu_architecture"] = cpu_info["arch"]
         self.info["num_cpu"] = psutil.cpu_count()
         self.info["cpu_name"] = cpu_info.get("brand_raw", "unknown")
diff --git a/requirements.txt b/requirements.txt
index 80566e52adc..8ca89a34d2a 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -1,7 +1,7 @@
 Cython>=0.25
 h5py>=2.6,!=3.0.0
 numpy>=1.15
-pandas>=0.25,!=1.1.5,<2.0
+pandas>=1.0,!=1.1.5
 scipy>=0.18
 tabulate>=0.7
 scikit-learn
@@ -29,7 +29,7 @@ torchinfo
 filelock
 psutil==5.9.4
 protobuf
-experiment_impact_tracker
+py-cpuinfo==9.0.0
 gpustat
 rich~=12.4.4
 packaging
@@ -39,10 +39,12 @@ sacremoses
 sentencepiece
 
 # new data format support
-xlwt            # excel
-xlrd            # excel
-openpyxl        # excel
-pyarrow         # parquet
-lxml            # html
+xlwt                # excel
+xlrd>=2.0.1         # excel
+xlsxwriter>=1.4.3   # excel
+openpyxl>=3.0.7     # excel
+pyxlsb>=1.0.8       # excel
+pyarrow             # parquet
+lxml                # html
 
 whylogs>=1.1.9
diff --git a/tests/integration_tests/test_experiment.py b/tests/integration_tests/test_experiment.py
index c9cb71e12b3..78812ca4031 100644
--- a/tests/integration_tests/test_experiment.py
+++ b/tests/integration_tests/test_experiment.py
@@ -428,7 +428,6 @@ def test_experiment_image_dataset(train_format, train_in_memory, test_format, te
     "df",
     "dict",
     "excel",
-    "excel_xls",
     "feather",
     "fwf",
     "hdf5",
diff --git a/tests/integration_tests/utils.py b/tests/integration_tests/utils.py
index f12910603c1..eae330cde3c 100644
--- a/tests/integration_tests/utils.py
+++ b/tests/integration_tests/utils.py
@@ -814,7 +814,7 @@ def to_fwf(df, fname):
         processed_df_rows = []
         for _, row in df.iterrows():
             processed_df_row = {}
-            for feature_name, raw_feature in row.iteritems():
+            for feature_name, raw_feature in row.items():
                 if "image" in feature_name and not (type(raw_feature) == float and np.isnan(raw_feature)):
                     feature = np.array(Image.open(raw_feature))
                 else:
diff --git a/tests/ludwig/data/test_split.py b/tests/ludwig/data/test_split.py
index e4cd77281bd..a5d1cccf860 100644
--- a/tests/ludwig/data/test_split.py
+++ b/tests/ludwig/data/test_split.py
@@ -370,6 +370,7 @@ def test_hash_split(df_engine, ray_cluster_2cpu):
     df2["id"] = np.arange(100, 200)
 
     nrows *= 2
+
     df = df_engine.df_lib.concat([df, df2])
 
     splits2 = splitter.split(df, backend)