diff --git a/modelforge/curation/ani1x_curation.py b/modelforge/curation/ani1x_curation.py
index edab841a..382825ec 100644
--- a/modelforge/curation/ani1x_curation.py
+++ b/modelforge/curation/ani1x_curation.py
@@ -83,6 +83,8 @@ def _init_dataset_parameters(self):
         self.dataset_md5_checksum = data_inputs[self.version_select][
             "dataset_md5_checksum"
         ]
+        self.dataset_length = data_inputs[self.version_select]["dataset_length"]
+        self.dataset_filename = data_inputs[self.version_select]["dataset_filename"]
         logger.debug(
             f"Dataset: {self.version_select} version: {data_inputs[self.version_select]['version']}"
         )
@@ -407,26 +409,27 @@ def process(
                 "max_records and total_conformers cannot be set at the same time."
             )
 
-        from modelforge.utils.remote import download_from_figshare
+        from modelforge.utils.remote import download_from_url
 
         url = self.dataset_download_url
 
         # download the dataset
-        self.name = download_from_figshare(
+        download_from_url(
             url=url,
             md5_checksum=self.dataset_md5_checksum,
             output_path=self.local_cache_dir,
+            output_filename=self.dataset_filename,
+            length=self.dataset_length,
             force_download=force_download,
         )
 
         self._clear_data()
 
         # process the rest of the dataset
-        if self.name is None:
-            raise Exception("Failed to retrieve name of file from figshare.")
+
         self._process_downloaded(
             self.local_cache_dir,
-            self.name,
+            self.dataset_filename,
             max_records=max_records,
             max_conformers_per_record=max_conformers_per_record,
             total_conformers=total_conformers,
diff --git a/modelforge/curation/ani2x_curation.py b/modelforge/curation/ani2x_curation.py
index ccddf0ea..09ecf9e9 100644
--- a/modelforge/curation/ani2x_curation.py
+++ b/modelforge/curation/ani2x_curation.py
@@ -69,6 +69,9 @@ def _init_dataset_parameters(self) -> None:
         self.dataset_md5_checksum = data_inputs[self.version_select][
             "dataset_md5_checksum"
         ]
+        self.dataset_filename = data_inputs[self.version_select]["dataset_filename"]
+        self.dataset_length = data_inputs[self.version_select]["dataset_length"]
+
         logger.debug(
             f"Dataset: {self.version_select} version: {data_inputs[self.version_select]['version']}"
         )
@@ -291,21 +294,20 @@ def process(
                 "max_records and total_conformers cannot be set at the same time."
             )
 
-        from modelforge.utils.remote import download_from_zenodo
+        from modelforge.utils.remote import download_from_url
 
         url = self.dataset_download_url
 
         # download the dataset
-        self.name = download_from_zenodo(
+        download_from_url(
             url=url,
             md5_checksum=self.dataset_md5_checksum,
             output_path=self.local_cache_dir,
+            output_filename=self.dataset_filename,
+            length=self.dataset_length,
             force_download=force_download,
         )
 
-        if self.name is None:
-            raise Exception("Failed to retrieve name of file from Zenodo.")
-
         # clear any data that might be present so we don't append to it
         self._clear_data()
 
@@ -314,13 +316,13 @@ def process(
 
         extract_tarred_file(
             input_path_dir=self.local_cache_dir,
-            file_name=self.name,
+            file_name=self.dataset_filename,
             output_path_dir=self.local_cache_dir,
             mode="r:gz",
         )
 
         # the untarred file will be in a directory named 'final_h5' within the local_cache_dir,
-        hdf5_filename = f"{self.name.replace('.tar.gz', '')}.h5"
+        hdf5_filename = f"{self.dataset_filename.replace('.tar.gz', '')}.h5"
 
         # process the rest of the dataset
         self._process_downloaded(
diff --git a/modelforge/curation/qm9_curation.py b/modelforge/curation/qm9_curation.py
index 847eaa01..67adde87 100644
--- a/modelforge/curation/qm9_curation.py
+++ b/modelforge/curation/qm9_curation.py
@@ -68,6 +68,9 @@ def _init_dataset_parameters(self) -> None:
         self.dataset_md5_checksum = data_inputs[self.version_select][
             "dataset_md5_checksum"
         ]
+        self.dataset_filename = data_inputs[self.version_select]["dataset_filename"]
+        self.dataset_length = data_inputs[self.version_select]["dataset_length"]
+
         logger.debug(
             f"Dataset: {self.version_select} version: {data_inputs[self.version_select]['version']}"
         )
@@ -640,24 +643,22 @@ def process(
                 "max_records and total_conformers cannot be set at the same time."
             )
 
-        from modelforge.utils.remote import download_from_figshare
+        from modelforge.utils.remote import download_from_url
 
         url = self.dataset_download_url
 
         # download the dataset
-        self.name = download_from_figshare(
+        download_from_url(
             url=url,
             md5_checksum=self.dataset_md5_checksum,
             output_path=self.local_cache_dir,
+            output_filename=self.dataset_filename,
+            length=self.dataset_length,
             force_download=force_download,
         )
         # clear out the data array before we process
         self._clear_data()
 
-        # process the rest of the dataset
-        if self.name is None:
-            raise Exception("Failed to retrieve name of file from figshare.")
-
         # untar the dataset
         from modelforge.utils.misc import extract_tarred_file
 
@@ -665,7 +666,7 @@ def process(
         # creating a directory called qm9_xyz_files to hold the contents
         extract_tarred_file(
             input_path_dir=self.local_cache_dir,
-            file_name=self.name,
+            file_name=self.dataset_filename,
             output_path_dir=f"{self.local_cache_dir}/qm9_xyz_files",
             mode="r:bz2",
         )
diff --git a/modelforge/curation/scripts/curate_spice114.py b/modelforge/curation/scripts/curate_spice114.py
index aa68d7d2..df2c6830 100644
--- a/modelforge/curation/scripts/curate_spice114.py
+++ b/modelforge/curation/scripts/curate_spice114.py
@@ -68,7 +68,7 @@ def spice114_wrapper(
     """
     from modelforge.curation.spice_1_curation import SPICE1Curation
 
-    spice_114 = SPICE114Curation(
+    spice_114 = SPICE1Curation(
         hdf5_file_name=hdf5_file_name,
         output_file_dir=output_file_dir,
         local_cache_dir=local_cache_dir,
diff --git a/modelforge/curation/spice_1_curation.py b/modelforge/curation/spice_1_curation.py
index 62ada82e..5d243d2c 100644
--- a/modelforge/curation/spice_1_curation.py
+++ b/modelforge/curation/spice_1_curation.py
@@ -68,6 +68,9 @@ def _init_dataset_parameters(self):
         self.dataset_md5_checksum = data_inputs[self.version_select][
             "dataset_md5_checksum"
         ]
+        self.dataset_filename = data_inputs[self.version_select]["dataset_filename"]
+        self.dataset_length = data_inputs[self.version_select]["dataset_length"]
+
         logger.debug(
             f"Dataset: {self.version_select} version: {data_inputs[self.version_select]['version']}"
         )
@@ -369,15 +372,17 @@ def process(
             raise Exception(
                 "max_records and total_conformers cannot be set at the same time."
             )
-        from modelforge.utils.remote import download_from_zenodo
+        from modelforge.utils.remote import download_from_url
 
         url = self.dataset_download_url
 
         # download the dataset
-        self.name = download_from_zenodo(
+        download_from_url(
             url=url,
             md5_checksum=self.dataset_md5_checksum,
             output_path=self.local_cache_dir,
+            output_filename=self.dataset_filename,
+            length=self.dataset_length,
             force_download=force_download,
         )
 
@@ -394,11 +399,10 @@ def process(
             self.atomic_numbers_to_limit = None
 
         # process the rest of the dataset
-        if self.name is None:
-            raise Exception("Failed to retrieve name of file from zenodo.")
+
         self._process_downloaded(
             self.local_cache_dir,
-            self.name,
+            self.dataset_filename,
             max_records,
             max_conformers_per_record,
             total_conformers,
diff --git a/modelforge/curation/spice_2_curation.py b/modelforge/curation/spice_2_curation.py
index 452b34b1..2db87fd9 100644
--- a/modelforge/curation/spice_2_curation.py
+++ b/modelforge/curation/spice_2_curation.py
@@ -65,6 +65,9 @@ def _init_dataset_parameters(self):
         self.dataset_md5_checksum = data_inputs[self.version_select][
             "dataset_md5_checksum"
         ]
+        self.dataset_filename = data_inputs[self.version_select]["dataset_filename"]
+        self.dataset_length = data_inputs[self.version_select]["dataset_length"]
+
         logger.debug(
             f"Dataset: {self.version_select} version: {data_inputs[self.version_select]['version']}"
         )
@@ -367,15 +370,17 @@ def process(
             raise ValueError(
                 "max_records and total_conformers cannot be set at the same time."
             )
-        from modelforge.utils.remote import download_from_zenodo
+        from modelforge.utils.remote import download_from_url
 
         url = self.dataset_download_url
 
         # download the dataset
-        self.name = download_from_zenodo(
+        download_from_url(
             url=url,
             md5_checksum=self.dataset_md5_checksum,
             output_path=self.local_cache_dir,
+            output_filename=self.dataset_filename,
+            length=self.dataset_length,
             force_download=force_download,
         )
 
@@ -393,11 +398,10 @@ def process(
             self.atomic_numbers_to_limit = None
 
         # process the rest of the dataset
-        if self.name is None:
-            raise Exception("Failed to retrieve name of file from zenodo.")
+
         self._process_downloaded(
             self.local_cache_dir,
-            self.name,
+            self.dataset_filename,
             max_records,
             max_conformers_per_record,
             total_conformers,
diff --git a/modelforge/curation/yaml_files/ani1x_curation.yaml b/modelforge/curation/yaml_files/ani1x_curation.yaml
index 5c933eba..8ca1c3cc 100644
--- a/modelforge/curation/yaml_files/ani1x_curation.yaml
+++ b/modelforge/curation/yaml_files/ani1x_curation.yaml
@@ -4,3 +4,5 @@ v_0:
   version: 0
   dataset_download_url: https://springernature.figshare.com/ndownloader/files/18112775
   dataset_md5_checksum: 98090dd6679106da861f52bed825ffb7
+  dataset_length: 5590846027
+  dataset_filename: ani1xrelease.h5
diff --git a/modelforge/curation/yaml_files/ani2x_curation.yaml b/modelforge/curation/yaml_files/ani2x_curation.yaml
index eb24880c..8a494f73 100644
--- a/modelforge/curation/yaml_files/ani2x_curation.yaml
+++ b/modelforge/curation/yaml_files/ani2x_curation.yaml
@@ -4,4 +4,6 @@ v_0:
   version: 0
   dataset_download_url: https://zenodo.org/records/10108942/files/ANI-2x-wB97X-631Gd.tar.gz
   dataset_md5_checksum: cb1d9effb3d07fc1cc6ced7cd0b1e1f2
+  dataset_length: 3705675413
+  dataset_filename: ANI-2x-wB97X-631Gd.tar.gz
 
diff --git a/modelforge/curation/yaml_files/qm9_curation.yaml b/modelforge/curation/yaml_files/qm9_curation.yaml
index ebd342d2..9b05555c 100644
--- a/modelforge/curation/yaml_files/qm9_curation.yaml
+++ b/modelforge/curation/yaml_files/qm9_curation.yaml
@@ -4,3 +4,5 @@ v_0:
   version: 0
   dataset_download_url: https://springernature.figshare.com/ndownloader/files/3195389
   dataset_md5_checksum: ad1ebd51ee7f5b3a6e32e974e5d54012
+  dataset_length: 86144227
+  dataset_filename: dsgdb9nsd.xyz.tar.bz2
diff --git a/modelforge/curation/yaml_files/spice1_curation.yaml b/modelforge/curation/yaml_files/spice1_curation.yaml
index 6285f26a..4be97814 100644
--- a/modelforge/curation/yaml_files/spice1_curation.yaml
+++ b/modelforge/curation/yaml_files/spice1_curation.yaml
@@ -5,3 +5,5 @@ v_0:
   notes: SPICE 1.1.4 release
   dataset_download_url: https://zenodo.org/records/8222043/files/SPICE-1.1.4.hdf5
   dataset_md5_checksum: f27d4c81da0e37d6547276bf6b4ae6a1
+  dataset_length: 16058156944
+  dataset_filename: SPICE-1.1.4.hdf5
diff --git a/modelforge/curation/yaml_files/spice2_curation.yaml b/modelforge/curation/yaml_files/spice2_curation.yaml
index 645cbd60..6ed537c0 100644
--- a/modelforge/curation/yaml_files/spice2_curation.yaml
+++ b/modelforge/curation/yaml_files/spice2_curation.yaml
@@ -5,3 +5,5 @@ v_0:
   notes: SPICE 2.0.1 release
   dataset_download_url: https://zenodo.org/records/10975225/files/SPICE-2.0.1.hdf5
   dataset_md5_checksum: bfba2224b6540e1390a579569b475510
+  dataset_length: 37479271148
+  dataset_filename: SPICE-2.0.1.hdf5
diff --git a/modelforge/tests/test_remote.py b/modelforge/tests/test_remote.py
index fbb1d1c0..174e9fd1 100644
--- a/modelforge/tests/test_remote.py
+++ b/modelforge/tests/test_remote.py
@@ -27,18 +27,20 @@ def test_is_url():
 
 
 def test_download_from_url(prep_temp_dir):
-    url = "https://raw.githubusercontent.com/choderalab/modelforge/e3e65e15e23ccc55d03dd7abb4b9add7a7dd15c3/modelforge/modelforge.py"
-    checksum = "66ec18ca5db3df5791ff1ffc584363a8"
+    url = "https://zenodo.org/records/3401581/files/PTC-CMC/atools_ml-v0.1.zip"
+    checksum = "194cde222565dca8657d8521e5df1fd8"
+
+    name = "atools_ml-v0.1.zip"
     # Download the file
     download_from_url(
         url,
         md5_checksum=checksum,
         output_path=str(prep_temp_dir),
-        output_filename="modelforge.py",
+        output_filename=name,
         force_download=True,
     )
 
-    file_name_path = str(prep_temp_dir) + "/modelforge.py"
+    file_name_path = str(prep_temp_dir) + f"/{name}"
     assert os.path.isfile(file_name_path)
 
     # create a dummy document to test the case where
@@ -51,11 +53,11 @@ def test_download_from_url(prep_temp_dir):
         url,
         md5_checksum=checksum,
         output_path=str(prep_temp_dir),
-        output_filename="modelforge.py",
+        output_filename=name,
         force_download=False,
     )
 
-    file_name_path = str(prep_temp_dir) + "/modelforge.py"
+    file_name_path = str(prep_temp_dir) + f"/{name}"
     assert os.path.isfile(file_name_path)
 
     # let us change the expected checksum to cause a failure
@@ -66,62 +68,7 @@ def test_download_from_url(prep_temp_dir):
             url,
             md5_checksum="checksum_garbage",
             output_path=str(prep_temp_dir),
-            output_filename="modelforge.py",
-            force_download=True,
-        )
-
-
-@pytest.mark.skip(
-    reason="This test seems to time out on the CI frequently. Will be refactoring and not need this soon."
-)
-def test_download_from_figshare(prep_temp_dir):
-    url = "https://figshare.com/ndownloader/files/22247589"
-    name = download_from_figshare(
-        url=url,
-        md5_checksum="c1459c5ddce7bb94800032aa3d04788e",
-        output_path=str(prep_temp_dir),
-        force_download=True,
-    )
-
-    file_name_path = str(prep_temp_dir) + f"/{name}"
-    assert os.path.isfile(file_name_path)
-
-    # create a dummy document to test the case where
-    # the checksum doesn't match so it will redownload
-    with open(file_name_path, "w") as f:
-        f.write("dummy document")
-
-    # This will force a download because the checksum doesn't match
-    url = "https://figshare.com/ndownloader/files/22247589"
-    name = download_from_figshare(
-        url=url,
-        md5_checksum="c1459c5ddce7bb94800032aa3d04788e",
-        output_path=str(prep_temp_dir),
-        force_download=False,
-    )
-
-    file_name_path = str(prep_temp_dir) + f"/{name}"
-    assert os.path.isfile(file_name_path)
-
-    # the length of this file isn't listed in the headers
-    # this will check to make sure we can handle this case
-    url = "https://figshare.com/ndownloader/files/30975751"
-    name = download_from_figshare(
-        url=url,
-        md5_checksum="efa40abff1f71c121f6f0d444c18d5b3",
-        output_path=str(prep_temp_dir),
-        force_download=True,
-    )
-
-    file_name_path = str(prep_temp_dir) + f"/{name}"
-    assert os.path.isfile(file_name_path)
-
-    with pytest.raises(Exception):
-        url = "https://choderalab.com/ndownloader/files/22247589"
-        name = download_from_figshare(
-            url=url,
-            md5_checksum="c1459c5ddce7bb94800032aa3d04788e",
-            output_path=str(prep_temp_dir),
+            output_filename=name,
             force_download=True,
         )
 
@@ -137,56 +84,16 @@ def test_fetch_record_id():
         fetch_url_from_doi(doi="10.5281/zenodo.3588339", timeout=0.0000000000001)
 
 
-def test_download_from_zenodo(prep_temp_dir):
-    url = "https://zenodo.org/records/3401581/files/PTC-CMC/atools_ml-v0.1.zip"
-    zenodo_checksum = "194cde222565dca8657d8521e5df1fd8"
-    name = download_from_zenodo(
-        url=url,
-        md5_checksum=zenodo_checksum,
-        output_path=str(prep_temp_dir),
-        force_download=True,
-    )
-
-    file_name_path = str(prep_temp_dir) + f"/{name}"
-    assert os.path.isfile(file_name_path)
-
-    # create a dummy document to test the case where
-    # the checksum doesn't match so it will redownload
-    with open(file_name_path, "w") as f:
-        f.write("dummy document")
-
-    # make sure that we redownload the file because the checksum of the
-    # existing file doesn't match
-    url = "https://zenodo.org/records/3401581/files/PTC-CMC/atools_ml-v0.1.zip"
-    zenodo_checksum = "194cde222565dca8657d8521e5df1fd8"
-    name = download_from_zenodo(
-        url=url,
-        md5_checksum=zenodo_checksum,
-        output_path=str(prep_temp_dir),
-        force_download=False,
-    )
-
-    file_name_path = str(prep_temp_dir) + f"/{name}"
-    assert os.path.isfile(file_name_path)
-
-    with pytest.raises(Exception):
-        url = "https://choderalab.com/22247589"
-        name = download_from_zenodo(
-            url=url,
-            md5_checksum=zenodo_checksum,
-            output_path=str(prep_temp_dir),
-            force_download=True,
-        )
-
-
 def test_md5_calculation(prep_temp_dir):
     url = "https://zenodo.org/records/3401581/files/PTC-CMC/atools_ml-v0.1.zip"
     zenodo_checksum = "194cde222565dca8657d8521e5df1fd8"
 
-    name = download_from_zenodo(
+    name = "atools_ml-v0.1.zip"
+    download_from_url(
         url=url,
         md5_checksum=zenodo_checksum,
         output_path=str(prep_temp_dir),
+        output_filename=name,
         force_download=True,
     )
 
@@ -199,9 +106,10 @@ def test_md5_calculation(prep_temp_dir):
 
     with pytest.raises(Exception):
         bad_checksum = "294badmd5checksumthatwontwork9de"
-        name = download_from_zenodo(
+        download_from_url(
             url=url,
             md5_checksum=bad_checksum,
             output_path=str(prep_temp_dir),
+            output_filename=name,
             force_download=True,
         )
diff --git a/modelforge/utils/remote.py b/modelforge/utils/remote.py
index 2e09246f..a3d3f237 100644
--- a/modelforge/utils/remote.py
+++ b/modelforge/utils/remote.py
@@ -111,7 +111,7 @@ def download_from_url(
     output_filename: str,
     length: Optional[int] = None,
     force_download=False,
-) -> str:
+):
 
     import requests
     import os
@@ -173,243 +173,3 @@ def download_from_url(
         logger.debug(
             "Using previously downloaded file; set force_download=True to re-download."
         )
-
-
-# Figshare helper functions
-def download_from_figshare(
-    url: str, md5_checksum: str, output_path: str, force_download=False
-) -> str:
-    """
-    Downloads a dataset from figshare for a given ndownloader url.
-
-    Parameters
-    ----------
-    url: str, required
-        Figshare ndownloader url (i.e., link to the data downloader)
-    md5_checksum: str, required
-        Expected md5 checksum of the downloaded file.
-    output_path: str, required
-        Location to download the file to.
-    force_download: str, default=False
-        If False: if the file exists in output_path, code will will use the local version.
-        If True, the file will be downloaded, even if it exists in output_path.
-
-    Returns
-    -------
-    str
-        Name of the file downloaded.
-
-    Examples
-    --------
-    >>> url = 'https://springernature.figshare.com/ndownloader/files/18112775'
-    >>> output_path = '/path/to/directory'
-    >>> downloaded_file_name = download_from_figshare(url, output_path)
-
-    """
-
-    import requests
-    import os
-    from tqdm import tqdm
-
-    # force to use ipv4; my ubuntu machine is timing out when it first tries ipv6
-    # requests.packages.urllib3.util.connection.HAS_IPV6 = False
-
-    chunk_size = 512
-    # check to make sure the url we are given is hosted by figshare.com
-    if not is_url(url, "figshare.com"):
-        raise Exception(f"{url} is not a valid figshare.com url")
-
-    # get the head of the request
-    head = requests.head(url)
-
-    # Because the url on figshare calls a downloader, instead of the direct file,
-    # we need to figure out where the original file is stored to know how big it is.
-    # Here we will parse the header info to get the file the downloader links to
-    # and then get the head info from this link to fetch the length.
-    # This is not actually necessary, but useful for updating the download status bar.
-    # We also fetch the name of the file from the header of the download link
-
-    temp_url = head.headers["location"].split("?")[0]
-    name = head.headers["X-Filename"].split("/")[-1]
-
-    # make sure we can handle a path with a ~ in it
-    output_path = os.path.expanduser(output_path)
-
-    # We need to check to make sure that the file that is stored in the output path
-    # has the correct checksum, e.g., to avoid a case where we have a partially downloaded file
-    # or to make sure we don't have two files with the same name, but different content.
-    if os.path.isfile(f"{output_path}/{name}"):
-        calculated_checksum = calculate_md5_checksum(
-            file_name=name, file_path=output_path
-        )
-        if calculated_checksum != md5_checksum:
-            force_download = True
-            logger.debug(
-                "Checksum of existing file does not match expected checksum, re-downloading."
-            )
-
-    if not os.path.isfile(f"{output_path}/{name}") or force_download:
-        logger.debug(f"Downloading datafile from figshare to {output_path}/{name}.")
-
-        temp_url_headers = requests.head(temp_url)
-
-        os.makedirs(output_path, exist_ok=True)
-        try:
-            length = int(temp_url_headers.headers["Content-Length"])
-        except:
-            print(
-                "Could not determine the length of the file to download. The download bar will not be accurate."
-            )
-            length = -1
-        r = requests.get(url, stream=True)
-
-        from modelforge.utils.misc import OpenWithLock
-
-        with OpenWithLock(f"{output_path}/{name}.lockfile", "w") as fl:
-            with open(f"{output_path}/{name}", "wb") as fd:
-                # if we couldn't fetch the length from figshare, which seems to happen for some records
-                # we just don't know how long the tqdm bar will be.
-                if length == -1:
-                    for chunk in tqdm(
-                        r.iter_content(chunk_size=chunk_size),
-                        ascii=True,
-                        desc="downloading",
-                    ):
-                        fd.write(chunk)
-                else:
-                    for chunk in tqdm(
-                        r.iter_content(chunk_size=chunk_size),
-                        ascii=True,
-                        desc="downloading",
-                        total=(int(length / chunk_size) + 1),
-                    ):
-                        fd.write(chunk)
-        os.remove(f"{output_path}/{name}.lockfile")
-
-        calculated_checksum = calculate_md5_checksum(
-            file_name=name, file_path=output_path
-        )
-        if calculated_checksum != md5_checksum:
-            raise Exception(
-                f"Checksum of downloaded file {calculated_checksum} does not match expected checksum {md5_checksum}"
-            )
-    else:  # if the file exists and we don't set force_download to True, just use the cached version
-        logger.debug(f"Datafile {name} already exists in {output_path}.")
-        logger.debug(
-            "Using previously downloaded file; set force_download=True to re-download."
-        )
-
-    return name
-
-
-def download_from_zenodo(
-    url: str, md5_checksum: str, output_path: str, force_download=False
-) -> str:
-    """
-    Downloads a dataset from zenodo for a given url.
-
-    If the datafile exists in the output_path, by default it will not be redownloaded.
-
-    Parameters
-    ----------
-    url : str, required
-        Direct link to datafile to download.
-    md5_checksum: str, required
-        Expected md5 checksum of the downloaded file.
-    output_path: str, required
-        Location to download the file to.
-    force_download: str, default=False
-        If False: if the file exists in output_path, code will will use the local version.
-        If True, the file will be downloaded, even if it exists in output_path.
-
-    Returns
-    -------
-    str
-        Name of the file downloaded.
-
-    Examples
-    --------
-    >>> url = "https://zenodo.org/records/3401581/files/PTC-CMC/atools_ml-v0.1.zip"
-    >>> output_path = '/path/to/directory'
-    >>> md5_checksum = "d41d8cd98f00b204e9800998ecf8427e"
-    >>> downloaded_file_name = download_from_zenodo(url, md5_checksum, output_path)
-
-    """
-
-    import requests
-    import os
-    from tqdm import tqdm
-
-    # force to use ipv4; my ubuntu machine is timing out when it first tries ipv6
-    # requests.packages.urllib3.util.connection.HAS_IPV6 = False
-
-    chunk_size = 512
-    # check to make sure the url we are given is hosted by figshare.com
-
-    if not is_url(url, "zenodo.org"):
-        raise Exception(f"{url} is not a valid zenodo.org url")
-
-    # get the head of the request
-    head = requests.head(url)
-
-    # Because the url on figshare calls a downloader, instead of the direct file,
-    # we need to figure out where the original file is stored to know how big it is.
-    # Here we will parse the header info to get the file the downloader links to
-    # and then get the head info from this link to fetch the length.
-    # This is not actually necessary, but useful for updating the download status bar.
-    # We also fetch the name of the file from the header of the download link
-    name = head.headers["Content-Disposition"].split("filename=")[-1]
-    length = int(head.headers["Content-Length"])
-
-    # make sure we can handle a path with a ~ in it
-    output_path = os.path.expanduser(output_path)
-
-    # We need to check to make sure that the file that is stored in the output path
-    # has the correct checksum, e.g., to avoid a case where we have a partially downloaded file
-    # or to make sure we don't have two files with the same name, but different content.
-
-    if os.path.isfile(f"{output_path}/{name}"):
-        calculated_checksum = calculate_md5_checksum(
-            file_name=name, file_path=output_path
-        )
-        if calculated_checksum != md5_checksum:
-            force_download = True
-            logger.debug(
-                "Checksum of existing file does not match expected checksum, re-downloading."
-            )
-
-    if not os.path.isfile(f"{output_path}/{name}") or force_download:
-        logger.debug(f"Downloading datafile from zenodo to {output_path}/{name}.")
-
-        r = requests.get(url, stream=True)
-
-        os.makedirs(output_path, exist_ok=True)
-
-        from modelforge.utils.misc import OpenWithLock
-
-        with OpenWithLock(f"{output_path}/{name}.lockfile", "w") as fl:
-            with open(f"{output_path}/{name}", "wb") as fd:
-                for chunk in tqdm(
-                    r.iter_content(chunk_size=chunk_size),
-                    ascii=True,
-                    desc="downloading",
-                    total=(int(length / chunk_size) + 1),
-                ):
-                    fd.write(chunk)
-        os.remove(f"{output_path}/{name}.lockfile")
-
-        calculated_checksum = calculate_md5_checksum(
-            file_name=name, file_path=output_path
-        )
-        if calculated_checksum != md5_checksum:
-            raise Exception(
-                f"Checksum of downloaded file {calculated_checksum} does not match expected checksum {md5_checksum}."
-            )
-
-    else:  # if the file exists and we don't set force_download to True, just use the cached version
-        logger.debug(f"Datafile {name} already exists in {output_path}.")
-        logger.debug(
-            "Using previously downloaded file; set force_download=True to re-download."
-        )
-
-    return name