Spaces:
Running
on
Zero
Running
on
Zero
polinaeterna
commited on
Commit
·
7badbdb
1
Parent(s):
53f0290
show only text column
Browse files
app.py
CHANGED
@@ -45,11 +45,6 @@ def predict(texts: list[str]):
|
|
45 |
return predicted_domains
|
46 |
|
47 |
|
48 |
-
# def progress():
|
49 |
-
# title = f"Scan finished" if num_rows == next_row_idx else "Scan in progress..."
|
50 |
-
|
51 |
-
|
52 |
-
|
53 |
def plot_and_df(texts, preds):
|
54 |
texts_df = pd.DataFrame({"quality": preds, "text": texts})
|
55 |
counts = Counter(preds)
|
@@ -62,9 +57,9 @@ def plot_and_df(texts, preds):
|
|
62 |
# counts.reset_index(inplace=True)
|
63 |
return (
|
64 |
gr.BarPlot(counts_df, x="quality", y="count"),
|
65 |
-
texts_df[texts_df["quality"] == "Low"][:20],
|
66 |
-
texts_df[texts_df["quality"] == "Medium"][:20],
|
67 |
-
texts_df[texts_df["quality"] == "High"][:20],
|
68 |
)
|
69 |
|
70 |
|
@@ -80,7 +75,7 @@ def run_quality_check(dataset, column, batch_size, num_examples):
|
|
80 |
batch_predictions = predict(batch_texts)
|
81 |
predictions.extend(batch_predictions)
|
82 |
texts_processed.extend(batch_texts)
|
83 |
-
yield {"
|
84 |
yield {"finished": 1.}, *plot_and_df(texts_processed, predictions)
|
85 |
|
86 |
with gr.Blocks() as demo:
|
|
|
45 |
return predicted_domains
|
46 |
|
47 |
|
|
|
|
|
|
|
|
|
|
|
48 |
def plot_and_df(texts, preds):
|
49 |
texts_df = pd.DataFrame({"quality": preds, "text": texts})
|
50 |
counts = Counter(preds)
|
|
|
57 |
# counts.reset_index(inplace=True)
|
58 |
return (
|
59 |
gr.BarPlot(counts_df, x="quality", y="count"),
|
60 |
+
texts_df[texts_df["quality"] == "Low"][["text"]][:20],
|
61 |
+
texts_df[texts_df["quality"] == "Medium"][["text"]][:20],
|
62 |
+
texts_df[texts_df["quality"] == "High"][["text"]][:20],
|
63 |
)
|
64 |
|
65 |
|
|
|
75 |
batch_predictions = predict(batch_texts)
|
76 |
predictions.extend(batch_predictions)
|
77 |
texts_processed.extend(batch_texts)
|
78 |
+
yield {"check in progress...": (i+batch_size) / num_examples}, *plot_and_df(texts_processed, predictions)
|
79 |
yield {"finished": 1.}, *plot_and_df(texts_processed, predictions)
|
80 |
|
81 |
with gr.Blocks() as demo:
|