nkululeko 0.84.0__py3-none-any.whl → 0.85.0__py3-none-any.whl

nkululeko/test_pretrain.py

@@ -11,11 +11,14 @@ import transformers
 
 import audeer
 import audiofile
+import audmetric
 
 from nkululeko.constants import VERSION
 import nkululeko.experiment as exp
+import nkululeko.models.finetune_model as fm
 import nkululeko.glob_conf as glob_conf
 from nkululeko.utils.util import Util
+import json
 
 
 def doit(config_file):
@@ -50,28 +53,42 @@ def doit(config_file):
     expr.fill_train_and_tests()
     util.debug(f"train shape : {expr.df_train.shape}, test shape:{expr.df_test.shape}")
 
+    model_root = util.get_path("model_dir")
+    log_root = audeer.mkdir("log")
+    torch_root = audeer.path(model_root, "torch")
+
+    metrics_gender = {
+        "UAR": audmetric.unweighted_average_recall,
+        "ACC": audmetric.accuracy,
+    }
+
     sampling_rate = 16000
     max_duration_sec = 8.0
 
     model_path = "facebook/wav2vec2-large-robust-ft-swbd-300h"
     num_layers = None
 
+    os.environ["CUDA_DEVICE_ORDER"] = "PCI_BUS_ID"
+    os.environ["CUDA_VISIBLE_DEVICES"] = "0"
+
     batch_size = 16
     accumulation_steps = 4
-
     # create dataset
 
     dataset = {}
+    target_name = glob_conf.target
     data_sources = {
-        "train": pd.DataFrame(expr.df_train[glob_conf.target]),
-        "dev": pd.DataFrame(expr.df_test[glob_conf.target]),
+        "train": pd.DataFrame(expr.df_train[target_name]),
+        "dev": pd.DataFrame(expr.df_test[target_name]),
     }
 
     for split in ["train", "dev"]:
+        df = data_sources[split]
+        df[target_name] = df[target_name].astype("float")
 
         y = pd.Series(
-            data=data_sources[split].itertuples(index=False, name=None),
-            index=data_sources[split].index,
+            data=df.itertuples(index=False, name=None),
+            index=df.index,
             dtype=object,
             name="labels",
         )
@@ -80,23 +97,195 @@ def doit(config_file):
         df = y.reset_index()
         df.start = df.start.dt.total_seconds()
         df.end = df.end.dt.total_seconds()
+
         print(f"{split}: {len(df)}")
+
         ds = datasets.Dataset.from_pandas(df)
         dataset[split] = ds
 
-        dataset = datasets.DatasetDict(dataset)
+    dataset = datasets.DatasetDict(dataset)
+
+    # load pre-trained model
+    le = glob_conf.label_encoder
+    mapping = dict(zip(le.classes_, range(len(le.classes_))))
+    target_mapping = {k: int(v) for k, v in mapping.items()}
+    target_mapping_reverse = {value: key for key, value in target_mapping.items()}
 
     config = transformers.AutoConfig.from_pretrained(
         model_path,
-        num_labels=len(util.la),
-        label2id=data.gender_mapping,
-        id2label=data.gender_mapping_reverse,
-        finetuning_task="age-gender",
+        num_labels=len(target_mapping),
+        label2id=target_mapping,
+        id2label=target_mapping_reverse,
+        finetuning_task=target_name,
     )
     if num_layers is not None:
         config.num_hidden_layers = num_layers
     setattr(config, "sampling_rate", sampling_rate)
-    setattr(config, "data", ",".join(sources))
+    setattr(config, "data", util.get_data_name())
+
+    vocab_dict = {}
+    with open("vocab.json", "w") as vocab_file:
+        json.dump(vocab_dict, vocab_file)
+    tokenizer = transformers.Wav2Vec2CTCTokenizer("./vocab.json")
+    tokenizer.save_pretrained(".")
+
+    feature_extractor = transformers.Wav2Vec2FeatureExtractor(
+        feature_size=1,
+        sampling_rate=16000,
+        padding_value=0.0,
+        do_normalize=True,
+        return_attention_mask=True,
+    )
+    processor = transformers.Wav2Vec2Processor(
+        feature_extractor=feature_extractor,
+        tokenizer=tokenizer,
+    )
+    assert processor.feature_extractor.sampling_rate == sampling_rate
+
+    model = fm.Model.from_pretrained(
+        model_path,
+        config=config,
+    )
+    model.freeze_feature_extractor()
+    model.train()
+
+    # training
+
+    def data_collator(data):
+
+        files = [d["file"] for d in data]
+        starts = [d["start"] for d in data]
+        ends = [d["end"] for d in data]
+        targets = [d["targets"] for d in data]
+
+        signals = []
+        for file, start, end in zip(
+            files,
+            starts,
+            ends,
+        ):
+            offset = start
+            duration = end - offset
+            if max_duration_sec is not None:
+                duration = min(duration, max_duration_sec)
+            signal, _ = audiofile.read(
+                file,
+                offset=offset,
+                duration=duration,
+            )
+            signals.append(signal.squeeze())
+
+        input_values = processor(
+            signals,
+            sampling_rate=sampling_rate,
+            padding=True,
+        )
+        batch = processor.pad(
+            input_values,
+            padding=True,
+            return_tensors="pt",
+        )
+
+        batch["labels"] = torch.tensor(targets)
+
+        return batch
+
+    def compute_metrics(p: transformers.EvalPrediction):
+
+        truth_gender = p.label_ids[:, 0].astype(int)
+        preds = p.predictions
+        preds_gender = np.argmax(preds, axis=1)
+
+        scores = {}
+
+        for name, metric in metrics_gender.items():
+            scores[f"gender-{name}"] = metric(truth_gender, preds_gender)
+
+        scores["combined"] = scores["gender-UAR"]
+
+        return scores
+
+    targets = pd.DataFrame(dataset["train"]["targets"])
+    counts = targets[0].value_counts().sort_index()
+    train_weights = 1 / counts
+    train_weights /= train_weights.sum()
+
+    print(train_weights)
+
+    criterion_gender = torch.nn.CrossEntropyLoss(
+        weight=torch.Tensor(train_weights).to("cuda"),
+    )
+
+    class Trainer(transformers.Trainer):
+
+        def compute_loss(
+            self,
+            model,
+            inputs,
+            return_outputs=False,
+        ):
+
+            targets = inputs.pop("labels").squeeze()
+            targets_gender = targets.type(torch.long)
+
+            outputs = model(**inputs)
+            logits_gender = outputs[0].squeeze()
+
+            loss_gender = criterion_gender(logits_gender, targets_gender)
+
+            loss = loss_gender
+
+            return (loss, outputs) if return_outputs else loss
+
+    num_steps = len(dataset["train"]) // (batch_size * accumulation_steps) // 5
+    num_steps = max(1, num_steps)
+    print(num_steps)
+
+    training_args = transformers.TrainingArguments(
+        output_dir=model_root,
+        logging_dir=log_root,
+        per_device_train_batch_size=batch_size,
+        per_device_eval_batch_size=batch_size,
+        gradient_accumulation_steps=accumulation_steps,
+        evaluation_strategy="steps",
+        num_train_epochs=5.0,
+        fp16=True,
+        save_steps=num_steps,
+        eval_steps=num_steps,
+        logging_steps=num_steps,
+        learning_rate=1e-4,
+        save_total_limit=2,
+        metric_for_best_model="combined",
+        greater_is_better=True,
+        load_best_model_at_end=True,
+        remove_unused_columns=False,
+        report_to="none",
+    )
+
+    trainer = Trainer(
+        model=model,
+        data_collator=data_collator,
+        args=training_args,
+        compute_metrics=compute_metrics,
+        train_dataset=dataset["train"],
+        eval_dataset=dataset["dev"],
+        tokenizer=processor.feature_extractor,
+        callbacks=[transformers.integrations.TensorBoardCallback()],
+    )
+    if False:
+        trainer.train()
+        trainer.save_model(torch_root)
+
+    modelnew = fm.Model.from_pretrained(
+        torch_root,
+        config=config,
+    )
+    print(f"loaded new model type{type(modelnew)}")
+    import audiofile
+
+    signal, _ = audiofile.read("./test.wav", always_2d=True)
+    result = modelnew.predict(signal)
+    print(result)
 
     print("DONE")
 

nkululeko/utils/util.py

@@ -33,43 +33,58 @@ class Util:
         else:
             self.caller = ""
         if has_config:
-            import nkululeko.glob_conf as glob_conf
-
-            self.config = glob_conf.config
-            self.got_data_roots = self.config_val("DATA", "root_folders", False)
-            if self.got_data_roots:
-                # if there is a global data rootfolder file, read from there
-                if not os.path.isfile(self.got_data_roots):
-                    self.error(f"no such file: {self.got_data_roots}")
-                self.data_roots = configparser.ConfigParser()
-                self.data_roots.read(self.got_data_roots)
-                # self.debug(f"getting data roots from {self.got_data_roots}")
+            try:
+                import nkululeko.glob_conf as glob_conf
+                self.config = glob_conf.config
+                self.got_data_roots = self.config_val(
+                    "DATA", "root_folders", False)
+                if self.got_data_roots:
+                    # if there is a global data rootfolder file, read from there
+                    if not os.path.isfile(self.got_data_roots):
+                        self.error(f"no such file: {self.got_data_roots}")
+                    self.data_roots = configparser.ConfigParser()
+                    self.data_roots.read(self.got_data_roots)
+            except (ModuleNotFoundError, AttributeError):
+                self.config = None
+                self.got_data_roots = False
 
     def get_path(self, entry):
         """
         This method allows the user to get the directory path for the given argument.
         """
-        root = os.path.join(self.config["EXP"]["root"], "")
-        name = self.config["EXP"]["name"]
-        try:
-            entryn = self.config["EXP"][entry]
-        except KeyError:
-            # some default values
+        if self.config is None:
+            # If no configuration file is provided, use default paths
             if entry == "fig_dir":
-                entryn = "./images/"
+                dir_name = "./images/"
             elif entry == "res_dir":
-                entryn = "./results/"
+                dir_name = "./results/"
             elif entry == "model_dir":
-                entryn = "./models/"
+                dir_name = "./models/"
             else:
-                entryn = "./store/"
-
-        # Expand image, model and result directories with run index
-        if entry == "fig_dir" or entry == "res_dir" or entry == "model_dir":
-            run = self.config_val("EXP", "run", 0)
-            entryn = entryn + f"run_{run}/"
+                dir_name = "./store/"
+        else:
+            root = os.path.join(self.config["EXP"]["root"], "")
+            name = self.config["EXP"]["name"]
+            try:
+                entryn = self.config["EXP"][entry]
+            except KeyError:
+                # some default values
+                if entry == "fig_dir":
+                    entryn = "./images/"
+                elif entry == "res_dir":
+                    entryn = "./results/"
+                elif entry == "model_dir":
+                    entryn = "./models/"
+                else:
+                    entryn = "./store/"
+
+            # Expand image, model and result directories with run index
+            if entry == "fig_dir" or entry == "res_dir" or entry == "model_dir":
+                run = self.config_val("EXP", "run", 0)
+                entryn = entryn + f"run_{run}/"
+
+            dir_name = f"{root}{name}/{entryn}"
 
-        dir_name = f"{root}{name}/{entryn}"
         audeer.mkdir(dir_name)
         return dir_name
 
@@ -101,7 +116,8 @@ class Util:
                         )
                     return default
             if not default in self.stopvals:
-                self.debug(f"value for {key} not found, using default: {default}")
+                self.debug(
+                    f"value for {key} not found, using default: {default}")
             return default
 
     def set_config(self, config):
@@ -138,7 +154,8 @@ class Util:
         if len(df) == 0:
             return df
         if not isinstance(df.index, pd.MultiIndex):
-            df.index = audformat.utils.to_segmented_index(df.index, allow_nat=False)
+            df.index = audformat.utils.to_segmented_index(
+                df.index, allow_nat=False)
         return df
 
     def _get_value_descript(self, section, name):
@@ -243,11 +260,14 @@ class Util:
         print(df.head(1))
 
     def config_val(self, section, key, default):
+        if self.config is None:
+            return default
         try:
             return self.config[section][key]
         except KeyError:
-            if not default in self.stopvals:
-                self.debug(f"value for {key} not found, using default: {default}")
+            if default not in self.stopvals:
+                self.debug(
+                    f"value for {key} not found, using default: {default}")
             return default
 
     def config_val_list(self, section, key, default):
@@ -255,7 +275,8 @@ class Util:
             return ast.literal_eval(self.config[section][key])
         except KeyError:
             if not default in self.stopvals:
-                self.debug(f"value for {key} not found, using default: {default}")
+                self.debug(
+                    f"value for {key} not found, using default: {default}")
             return default
 
     def continuous_to_categorical(self, series):

{nkululeko-0.84.0.dist-info → nkululeko-0.85.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: nkululeko
-Version: 0.84.0
+Version: 0.85.0
 Summary: Machine learning audio prediction experiments based on templates
 Home-page: https://github.com/felixbur/nkululeko
 Author: Felix Burkhardt
@@ -333,6 +333,14 @@ F. Burkhardt, Johannes Wagner, Hagen Wierstorf, Florian Eyben and Björn Schulle
 Changelog
 =========
 
+Version 0.85.0
+--------------
+* first version with finetuning wav2vec2 layers
+
+Version 0.84.1
+--------------
+* made resample independent of config file
+
 Version 0.84.0
 --------------
 * added SHAP analysis

{nkululeko-0.84.0.dist-info → nkululeko-0.85.0.dist-info}/RECORD

@@ -2,36 +2,36 @@ nkululeko/__init__.py,sha256=62f8HiEzJ8rG2QlTFJXUCMpvuH3fKI33DoJSj33mscc,63
 nkululeko/aug_train.py,sha256=YhuZnS_WVWnun9G-M6g5n6rbRxoVREz6Zh7k6qprFNQ,3194
 nkululeko/augment.py,sha256=4MG0apTAG5RgkuJrYEjGgDdbodZWi_HweSPNI1JJ5QA,3051
 nkululeko/cacheddataset.py,sha256=lIJ6hUo5LoxSrzXtWV8mzwO7wRtUETWnOQ4ws2XfL1E,969
-nkululeko/constants.py,sha256=Ij7mvqjHA328NaCRJL2JvyYgAPfkfYpVq_WiS735RQY,39
+nkululeko/constants.py,sha256=flWSUNQs4r0X0SgoR1I72Mk49cRUdpBN8Zng8sySFBE,39
 nkululeko/demo.py,sha256=8bl15Kitoesnz8oa8yrs52T6YCSOhWbbq9PnZ8Hj6D0,3232
 nkululeko/demo_feats.py,sha256=sAeGFojhEj9WEDFtG3SzPBmyYJWLF2rkbpp65m8Ujo4,2025
 nkululeko/demo_predictor.py,sha256=es56xbT8ifkS_vnrlb5NTZT54gNmeUtNlA4zVA_gnN8,4757
-nkululeko/experiment.py,sha256=mYdHfInMkuOI3frkZo7oaEe9viO-Qa1ZShyF6MPozcU,30225
+nkululeko/experiment.py,sha256=9Nw23b7sVOciH8IaOuAAKbY7otXYSsPrj_rQCA_U9cc,30465
 nkululeko/explore.py,sha256=lDzRoW_Taa5u4BBABZLD89BcQWnYlrftJR4jgt1yyj0,2609
 nkululeko/export.py,sha256=mHeEAAmtZuxdyebLlbSzPrHSi9OMgJHbk35d3DTxRBc,4632
 nkululeko/feature_extractor.py,sha256=8mssYKmo4LclVI-hiLmJEDZ0ZPyDavFG2YwtXcrGzwM,3976
 nkululeko/file_checker.py,sha256=LoLnL8aHpW-axMQ46qbqrManTs5otG9ShpEZuz9iRSk,3474
 nkululeko/filter_data.py,sha256=w-X2mhKdYr5DxDIz50E5yzO6Jmzk4jjDBoXsgOOVtcA,7222
 nkululeko/glob_conf.py,sha256=KL9YJQTHvTztxo1vr25qRRgaPnx4NTg0XrdbovKGMmw,525
-nkululeko/modelrunner.py,sha256=GwDXcE2gDQXat4W0-HhHQ1BcUNCRBXMBQ4QycfHp_5c,9288
+nkululeko/modelrunner.py,sha256=pPhvTh1rIrFQg5Ox9T1KoFJ4wRcLCmJl7LFud2DA41w,10464
 nkululeko/multidb.py,sha256=fG3VukEWP1vreVN4gB1IRXxwwg4jLftsSEYtu0o1f78,5634
 nkululeko/nkuluflag.py,sha256=PGWSmZz-PiiHLgcZJAoGOI_Y-sZDVI1ksB8p5r7riWM,3725
 nkululeko/nkululeko.py,sha256=Kn3s2E3yyH8cJ7z6lkMxrnqtCxTu7-qfe9Zr_ONTD5g,1968
 nkululeko/plots.py,sha256=nd9tF_61DyAx7oGZF8gTrHXazkgFjFe4eClxu1nQ_XU,23276
 nkululeko/predict.py,sha256=sF091sSSLnEWcISx9ZcULLie3tY5XeFsQJd6b3vrxFg,2409
-nkululeko/resample.py,sha256=3WbxkwgyTe_fW38046Rjxk3knOkFdhqn2C4nfhbUurQ,2287
+nkululeko/resample.py,sha256=IPtYqU0nhZ-CqO_O1jJN0EvpfjxHZdFRwdTpEJOVuaQ,3354
 nkululeko/runmanager.py,sha256=eTM1DNQKt1lxYhzt4vZyZluPXW9sWlIJHNQzex4lkJU,7624
 nkululeko/scaler.py,sha256=4nkIqoajkIkuTPK0Z02ifMN_awl6fP_i-GBYdoGYgGM,4101
 nkululeko/segment.py,sha256=YLKckX44tbvTb3LrdgYw9X4guzuF27sutl92z9DkpZU,4835
 nkululeko/syllable_nuclei.py,sha256=Sky-C__MeUDaxqHnDl2TGLLYOYvsahD35TUjWGeG31k,10047
 nkululeko/test.py,sha256=1w624vo5KTzmFC8BUStGlLDmIEAFuJUz7J0W-gp7AxI,1677
 nkululeko/test_predictor.py,sha256=_w5J8CxH6hmW3mLTKbdfmywl5QpdNAnW1Y8TE5GtlfE,3237
-nkululeko/test_pretrain.py,sha256=aoN-C9M4Zn9LwseIWQdMMpEGclnkWs-gJXyItU5V0fk,3109
+nkululeko/test_pretrain.py,sha256=ZWl-bR6nmeSmXkGAIE6zyfQEjN8Zg0rIxfaS-O6Zbas,8465
 nkululeko/augmenting/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 nkululeko/augmenting/augmenter.py,sha256=XAt0dpmlnKxqyysqCgV3rcz-pRIvOz7rU7dmGDCVAzs,2905
 nkululeko/augmenting/randomsplicer.py,sha256=Z5rxdKKUpuncLWuTS6xVfVKUeVbeiYU_dLRHQ5fcg4Y,2669
 nkululeko/augmenting/randomsplicing.py,sha256=ldym9vZNsZIU5BAAaJVaOmAgmVHNs4a5i5K3bW-WAQU,1791
-nkululeko/augmenting/resampler.py,sha256=cRrn27w_f2I6aN0CftlTuHT2edi7pTREh3Yc6BxhcGU,3335
+nkululeko/augmenting/resampler.py,sha256=nOBsiQpX6p4jXsP7x6wak78F3B5YYYRmC_iHX8iuOXs,3542
 nkululeko/autopredict/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 nkululeko/autopredict/ap_age.py,sha256=2Wn5E-Jd49sTn40WqaMcYtUEl4zEq3OY75XmjOpdxsA,1095
 nkululeko/autopredict/ap_arousal.py,sha256=ymt0diu4v1osw3VxJbSglsVKDAJYRzebQ2TTfFMKKxk,1024
@@ -68,14 +68,15 @@ nkululeko/feat_extract/feats_squim.py,sha256=Y31YmDmscuG0YozvxyBZIutO3id8t7IZJWC
 nkululeko/feat_extract/feats_trill.py,sha256=K2ahhdpwpjgg3WZS1POg3UMP2U44i8cLZZvn5Rq7fUI,3228
 nkululeko/feat_extract/feats_wav2vec2.py,sha256=9WUMfyddB_3nx79g7mZoQrRynhM1uEBWuOotRq8bxoU,5268
 nkululeko/feat_extract/feats_wavlm.py,sha256=ulxpGjifUFx2ZgGmY32SmBJGIuvkYHoLb2n1LZ8KMwA,4703
-nkululeko/feat_extract/feats_whisper.py,sha256=BFspQBI53HAgw22vBEeFskGwFZA-94Rpl17xM458HRo,4576
+nkululeko/feat_extract/feats_whisper.py,sha256=0N7Vj65OVi2PNoB_NrDjWT5lP6xZNKxFOZZIoxkJvcA,4533
 nkululeko/feat_extract/featureset.py,sha256=HtgW2389rmlRAgFP3F1sSFzq2_iUVr2NhOfIXG9omt0,1448
 nkululeko/feat_extract/feinberg_praat.py,sha256=EP9pMALjlKdiYInLQdrZ7MmE499Mq-ISRCgqbqL3Rxc,21304
 nkululeko/losses/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 nkululeko/losses/loss_ccc.py,sha256=NOK0y0fxKUnU161B5geap6Fmn8QzoPl2MqtPiV8IuJE,976
 nkululeko/losses/loss_softf1loss.py,sha256=5gW-PuiqeAZcRgfwjueIOQtMokOjZWgQnVIv59HKTCo,1309
 nkululeko/models/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-nkululeko/models/model.py,sha256=fL6LB6I9Oqo_OWUIptqiu6abuxVYYv8bW2a3m4XSLqU,11601
+nkululeko/models/finetune_model.py,sha256=OMlzDyUFNXZ2xSiqqH8tbzey_KzPJ4jsoYT-4KrWFKM,5091
+nkululeko/models/model.py,sha256=PUCqF2r_dEfmFsZn6Cgr1UIzYvxziLH6nSqZ5-vuN1o,11639
 nkululeko/models/model_bayes.py,sha256=WJFZ8wFKwWATz6MhmjeZIi1Pal1viU549WL_PjXDSy8,406
 nkululeko/models/model_cnn.py,sha256=bJxqwe6FnVR2hFeqN6EXexYGgvKYFED1VOhBXVlLWaE,9954
 nkululeko/models/model_gmm.py,sha256=hZ9UO36KNf48qa3J-xkWIicIj9-TApmt21zNES2vEOs,649
@@ -88,6 +89,7 @@ nkululeko/models/model_svm.py,sha256=rsME3KvKvNG7bdE5lbvYUu85WZhaASZxxmdNDIVJRZ4
 nkululeko/models/model_svr.py,sha256=_YZeksqB3eBENGlg3g9RwYFlk9rQQ-XCeNBKLlGGVoE,725
 nkululeko/models/model_tree.py,sha256=rf16faUm4o2LJgkoYpeY998b8DQIvXZ73_m1IS3TnnE,417
 nkululeko/models/model_tree_reg.py,sha256=IgQcPTE-304HQLYSKPF8Z4ot_Ur9dH01fZjS0nXke_M,428
+nkululeko/models/model_tuned.py,sha256=zmagIE3QHP67_XJCx5r7ZXBojsp6SC8IS-L3XRWmCEk,15650
 nkululeko/models/model_xgb.py,sha256=Thgx5ESdIok4v72mKh4plxpo4smGcKALWNCJTDScY0M,447
 nkululeko/models/model_xgr.py,sha256=aGBtNGLWjOE_2rICGYGFxmT8DtnHYsIl1lIpMtghHsY,418
 nkululeko/reporting/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
@@ -103,9 +105,9 @@ nkululeko/segmenting/seg_silero.py,sha256=lLytS38KzARS17omwv8VBw-zz60RVSXGSvZ5Ev
 nkululeko/utils/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 nkululeko/utils/files.py,sha256=UiGAtZRWYjHSvlmPaTMtzyNNGE6qaLaxQkybctS7iRM,4021
 nkululeko/utils/stats.py,sha256=1yUq0FTOyqkU8TwUocJRYdJaqMU5SlOBBRUun9STo2M,2829
-nkululeko/utils/util.py,sha256=lVKcIYHeN8wt7ap8o1UTx5z6nvOY40twJ_C4CshT42Y,13068
-nkululeko-0.84.0.dist-info/LICENSE,sha256=0zGP5B_W35yAcGfHPS18Q2B8UhvLRY3dQq1MhpsJU_U,1076
-nkululeko-0.84.0.dist-info/METADATA,sha256=RJnEnBwqdKRLs4J16zOzsps0GXmdVvzPMi1_2hpZh-Q,36346
-nkululeko-0.84.0.dist-info/WHEEL,sha256=GJ7t_kWBFywbagK5eo9IoUwLW6oyOeTKmQ-9iHFVNxQ,92
-nkululeko-0.84.0.dist-info/top_level.txt,sha256=DPFNNSHPjUeVKj44dVANAjuVGRCC3MusJ08lc2a8xFA,10
-nkululeko-0.84.0.dist-info/RECORD,,
+nkululeko/utils/util.py,sha256=b1IHFucRNuF9Iyv5IJeK4AEg0Rga0xKG80UM5GWWdHA,13816
+nkululeko-0.85.0.dist-info/LICENSE,sha256=0zGP5B_W35yAcGfHPS18Q2B8UhvLRY3dQq1MhpsJU_U,1076
+nkululeko-0.85.0.dist-info/METADATA,sha256=Zt3H0FmIXOJvzyLOI0aC8VfvjrdIkd4uNvb937luo_k,36499
+nkululeko-0.85.0.dist-info/WHEEL,sha256=GJ7t_kWBFywbagK5eo9IoUwLW6oyOeTKmQ-9iHFVNxQ,92
+nkululeko-0.85.0.dist-info/top_level.txt,sha256=DPFNNSHPjUeVKj44dVANAjuVGRCC3MusJ08lc2a8xFA,10
+nkululeko-0.85.0.dist-info/RECORD,,