Formatting

Author: Thilina Rajapakse

examples/seq2seq/paraphrasing/predict.py

@@ -7,9 +7,7 @@
 transformers_logger = logging.getLogger("transformers")
 transformers_logger.setLevel(logging.ERROR)
 
-model = Seq2SeqModel(
-    encoder_decoder_type="bart", encoder_decoder_name="outputs"
-)
+model = Seq2SeqModel(encoder_decoder_type="bart", encoder_decoder_name="outputs")
 
 
 while True:

examples/seq2seq/paraphrasing/train.py

@@ -20,12 +20,8 @@
 train_df = train_df.loc[train_df["label"] == "1"]
 eval_df = eval_df.loc[eval_df["label"] == "1"]
 
-train_df = train_df.rename(
-    columns={"sentence1": "input_text", "sentence2": "target_text"}
-)
-eval_df = eval_df.rename(
-    columns={"sentence1": "input_text", "sentence2": "target_text"}
-)
+train_df = train_df.rename(columns={"sentence1": "input_text", "sentence2": "target_text"})
+eval_df = eval_df.rename(columns={"sentence1": "input_text", "sentence2": "target_text"})
 
 train_df = train_df[["input_text", "target_text"]]
 eval_df = eval_df[["input_text", "target_text"]]
@@ -34,25 +30,13 @@
 eval_df["prefix"] = "paraphrase"
 
 # MSRP Data
-train_df = pd.concat(
-    [
-        train_df,
-        load_data("data/msr_paraphrase_train.txt", "#1 String", "#2 String", "Quality"),
-    ]
-)
-eval_df = pd.concat(
-    [
-        eval_df,
-        load_data("data/msr_paraphrase_test.txt", "#1 String", "#2 String", "Quality"),
-    ]
-)
+train_df = pd.concat([train_df, load_data("data/msr_paraphrase_train.txt", "#1 String", "#2 String", "Quality"),])
+eval_df = pd.concat([eval_df, load_data("data/msr_paraphrase_test.txt", "#1 String", "#2 String", "Quality"),])
 
 # Quora Data
 
 # The Quora Dataset is not separated into train/test, so we do it manually the first time.
-df = load_data(
-    "data/quora_duplicate_questions.tsv", "question1", "question2", "is_duplicate"
-)
+df = load_data("data/quora_duplicate_questions.tsv", "question1", "question2", "is_duplicate")
 q_train, q_test = train_test_split(df)
 
 q_train.to_csv("data/quora_train.tsv", sep="\t")
@@ -107,11 +91,7 @@
 model_args.wandb_project = "Paraphrasing with BART"
 
 
-model = Seq2SeqModel(
-    encoder_decoder_type="bart",
-    encoder_decoder_name="facebook/bart-large",
-    args=model_args,
-)
+model = Seq2SeqModel(encoder_decoder_type="bart", encoder_decoder_name="facebook/bart-large", args=model_args,)
 
 model.train_model(train_df, eval_data=eval_df)
 
@@ -136,6 +116,4 @@
         f.write("Prediction:\n")
         for pred in preds[i]:
             f.write(str(pred) + "\n")
-        f.write(
-            "________________________________________________________________________________\n"
-        )
+        f.write("________________________________________________________________________________\n")

examples/seq2seq/paraphrasing/utils.py

@@ -3,14 +3,10 @@
 import pandas as pd
 
 
-def load_data(
-    file_path, input_text_column, target_text_column, label_column, keep_label=1
-):
+def load_data(file_path, input_text_column, target_text_column, label_column, keep_label=1):
     df = pd.read_csv(file_path, sep="\t", error_bad_lines=False)
     df = df.loc[df[label_column] == keep_label]
-    df = df.rename(
-        columns={input_text_column: "input_text", target_text_column: "target_text"}
-    )
+    df = df.rename(columns={input_text_column: "input_text", target_text_column: "target_text"})
     df = df[["input_text", "target_text"]]
     df["prefix"] = "paraphrase"
 

simpletransformers/streamlit/classification_view.py

@@ -28,7 +28,12 @@ def get_states(model, session_state=None):
     return session_state, model
 
 
-@st.cache(hash_funcs={ClassificationModel: simple_transformers_model, MultiLabelClassificationModel: simple_transformers_model})
+@st.cache(
+    hash_funcs={
+        ClassificationModel: simple_transformers_model,
+        MultiLabelClassificationModel: simple_transformers_model,
+    }
+)
 def get_prediction(model, input_text):
     prediction, raw_values = model.predict([input_text])
 
@@ -71,9 +76,7 @@ def classification_viewer(model, model_class):
         try:
             session_state, model = get_states(model)
         except AttributeError:
-            session_state = get(
-                max_seq_length=model.args.max_seq_length,
-            )
+            session_state = get(max_seq_length=model.args.max_seq_length,)
             session_state, model = get_states(model, session_state)
 
         model.args.max_seq_length = st.sidebar.slider(

simpletransformers/streamlit/ner_view.py

@@ -5,9 +5,7 @@
 from simpletransformers.streamlit.streamlit_utils import get, simple_transformers_model, get_color
 
 
-ENTITY_WRAPPER = (
-    """<mark style="background: rgba{}; font-weight: 450; border-radius: 0.5rem; margin: 0.1em; padding: 0.25rem; display: inline-block">{} {}</mark>"""
-)
+ENTITY_WRAPPER = """<mark style="background: rgba{}; font-weight: 450; border-radius: 0.5rem; margin: 0.1em; padding: 0.25rem; display: inline-block">{} {}</mark>"""
 ENTITY_LABEL_WRAPPER = """<span style="background: #fff; font-size: 0.56em; font-weight: bold; padding: 0.3em 0.3em; vertical-align: middle; margin: 0 0 0.15rem 0.5rem; line-height: 1; display: inline-block">{}</span>"""
 
 
@@ -26,9 +24,7 @@ def get_prediction(model, input_text):
 
 
 def ner_viewer(model):
-    session_state = get(
-        max_seq_length=model.args.max_seq_length,
-    )
+    session_state = get(max_seq_length=model.args.max_seq_length,)
     model.args.max_seq_length = session_state.max_seq_length
 
     entity_list = model.args.labels_list
@@ -47,7 +43,13 @@ def ner_viewer(model):
 
     prediction = get_prediction(model, input_text)[0]
 
-    to_write = " ".join([format_word(word, entity, entity_checkboxes, entity_color_map) for pred in prediction for word, entity in pred.items()])
+    to_write = " ".join(
+        [
+            format_word(word, entity, entity_checkboxes, entity_color_map)
+            for pred in prediction
+            for word, entity in pred.items()
+        ]
+    )
 
     st.subheader(f"Predictions")
     st.write(to_write, unsafe_allow_html=True)

simpletransformers/streamlit/qa_view.py

@@ -77,9 +77,14 @@ def qa_viewer(model):
 
         if answers[0] != "empty":
             if len(context_pieces) == 2:
-                st.write(QA_ANSWER_WRAPPER.format(context_pieces[0], answers[0], context_pieces[-1]), unsafe_allow_html=True)
+                st.write(
+                    QA_ANSWER_WRAPPER.format(context_pieces[0], answers[0], context_pieces[-1]), unsafe_allow_html=True
+                )
             else:
-                st.write(QA_ANSWER_WRAPPER.format(context_pieces[0], answers[0], answers[0].join(context_pieces[1:])), unsafe_allow_html=True)
+                st.write(
+                    QA_ANSWER_WRAPPER.format(context_pieces[0], answers[0], answers[0].join(context_pieces[1:])),
+                    unsafe_allow_html=True,
+                )
         else:
             st.write(QA_EMPTY_ANSWER_WRAPPER.format("", answers[0], ""), unsafe_allow_html=True)
 

simpletransformers/streamlit/streamlit_utils.py

@@ -168,7 +168,30 @@ def simple_transformers_model(model):
 
 def get_color(i):
     # Colors taken from Sasha Trubetskoy's list of colors - https://sashamaps.net/docs/tools/20-colors/
-    colors = [(60, 180, 75, 0.4), (255, 225, 25, 0.4), (0, 130, 200, 0.4), (245, 130, 48, 0.4), (145, 30, 180, 0.4), (70, 240, 240, 0.4), (240, 50, 230, 0.4), (210, 245, 60, 0.4), (250, 190, 212, 0.4), (0, 128, 128, 0.4), (220, 190, 255, 0.4), (170, 110, 40, 0.4), (255, 250, 200, 0.4), (128, 0, 0, 0.4), (170, 255, 195, 0.4), (128, 128, 0, 0.4), (255, 215, 180, 0.4), (0, 0, 128, 0.4), (128, 128, 128, 0.4), (255, 255, 255, 0.4), (0, 0, 0, 0.4), (230, 25, 75, 0.4)]
+    colors = [
+        (60, 180, 75, 0.4),
+        (255, 225, 25, 0.4),
+        (0, 130, 200, 0.4),
+        (245, 130, 48, 0.4),
+        (145, 30, 180, 0.4),
+        (70, 240, 240, 0.4),
+        (240, 50, 230, 0.4),
+        (210, 245, 60, 0.4),
+        (250, 190, 212, 0.4),
+        (0, 128, 128, 0.4),
+        (220, 190, 255, 0.4),
+        (170, 110, 40, 0.4),
+        (255, 250, 200, 0.4),
+        (128, 0, 0, 0.4),
+        (170, 255, 195, 0.4),
+        (128, 128, 0, 0.4),
+        (255, 215, 180, 0.4),
+        (0, 0, 128, 0.4),
+        (128, 128, 128, 0.4),
+        (255, 255, 255, 0.4),
+        (0, 0, 0, 0.4),
+        (230, 25, 75, 0.4),
+    ]
     try:
         return str(colors[i])
     except IndexError:

simpletransformers/t5/run_simple_transformers_streamlit_app.py

@@ -3,4 +3,3 @@
 
 
 streamlit_runner()
-