Spaces:
Build error
Build error
Commit
·
949c5da
1
Parent(s):
5aa6e12
Initial: initial commit
Browse files- app.py +17 -0
- lid.py +52 -0
- requirements.txt +4 -0
app.py
ADDED
|
@@ -0,0 +1,17 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import gradio as gr
|
| 2 |
+
from lid import identify_and_plot, LID_EXAMPLES
|
| 3 |
+
|
| 4 |
+
lid_indentify = gr.Interface(
|
| 5 |
+
fn=identify_and_plot,
|
| 6 |
+
inputs=gr.Textbox(lines=2, label="Input text"),
|
| 7 |
+
outputs=[gr.Textbox(label="Language"), gr.Plot(label="Confidence Plot")],
|
| 8 |
+
examples=LID_EXAMPLES,
|
| 9 |
+
title="Language Identification Demo",
|
| 10 |
+
description="Identify the language of input text and view confidence levels.",
|
| 11 |
+
allow_flagging="never",
|
| 12 |
+
)
|
| 13 |
+
|
| 14 |
+
with gr.Blocks() as demo:
|
| 15 |
+
lid_indentify.render()
|
| 16 |
+
|
| 17 |
+
demo.launch()
|
lid.py
ADDED
|
@@ -0,0 +1,52 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import fasttext
|
| 2 |
+
from huggingface_hub import hf_hub_download
|
| 3 |
+
import matplotlib.pyplot as plt
|
| 4 |
+
import seaborn as sns
|
| 5 |
+
|
| 6 |
+
# Download the model from Hugging Face Hub
|
| 7 |
+
model_path = hf_hub_download(
|
| 8 |
+
# repo_id="cis-lmu/glotlid", filename="model.bin", cache_dir="./glotlid"
|
| 9 |
+
repo_id="facebook/fasttext-language-identification",
|
| 10 |
+
filename="model.bin",
|
| 11 |
+
cache_dir="fasttext",
|
| 12 |
+
)
|
| 13 |
+
model = fasttext.load_model(model_path)
|
| 14 |
+
|
| 15 |
+
|
| 16 |
+
def identify_languages(title):
|
| 17 |
+
predictions = model.predict(title, k=5) # Get top 5 predictions
|
| 18 |
+
labels = predictions[0]
|
| 19 |
+
confidences = predictions[1]
|
| 20 |
+
|
| 21 |
+
results = []
|
| 22 |
+
for label, confidence in zip(labels, confidences):
|
| 23 |
+
language = label.replace("__label__", "")
|
| 24 |
+
results.append((language, confidence))
|
| 25 |
+
|
| 26 |
+
return results
|
| 27 |
+
|
| 28 |
+
|
| 29 |
+
def plot_confidences(results):
|
| 30 |
+
languages = [result[0] for result in results]
|
| 31 |
+
confidences = [result[1] for result in results]
|
| 32 |
+
|
| 33 |
+
plt.figure(figsize=(10, 6))
|
| 34 |
+
sns.barplot(x=confidences, y=languages)
|
| 35 |
+
plt.xlabel("Confidence")
|
| 36 |
+
plt.ylabel("Language")
|
| 37 |
+
plt.title("Language Identification Confidence")
|
| 38 |
+
|
| 39 |
+
return plt
|
| 40 |
+
|
| 41 |
+
|
| 42 |
+
def identify_and_plot(title):
|
| 43 |
+
results = identify_languages(title)
|
| 44 |
+
plot = plot_confidences(results)
|
| 45 |
+
return results, plot
|
| 46 |
+
|
| 47 |
+
|
| 48 |
+
LID_EXAMPLES = [
|
| 49 |
+
"ပိုၼ်းၵႅပ်ႈလိၵ်ႈလၢႆးတႆး ဢၼ်ၶူးပွင်သွၼ်လူင်လိၵ်ႈလၢႆးတႆး",
|
| 50 |
+
"ယု၀တီဂျင်းဖောမယ်၊ ရှမ်းစာပေသမိုင်းနှင့်",
|
| 51 |
+
"Hello World, မႂ်ႇသုင်ၶႃႈ",
|
| 52 |
+
]
|
requirements.txt
ADDED
|
@@ -0,0 +1,4 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
fasttext-wheel
|
| 2 |
+
huggingface_hub
|
| 3 |
+
matplotlib
|
| 4 |
+
seaborn
|