Upload 10 files
Browse files- README.md +4 -12
- app.py +49 -0
- cluster_outliers.csv +0 -0
- main.ipynb +1008 -0
- onnx_model/config.json +27 -0
- onnx_model/model.onnx +3 -0
- outlier_detection.ipynb +2292 -0
- pycaret_outlier_detection.ipynb +0 -0
- requirements.txt +10 -0
- trainer.ipynb +1165 -0
README.md
CHANGED
|
@@ -1,13 +1,5 @@
|
|
| 1 |
-
|
| 2 |
-
|
| 3 |
-
|
| 4 |
-
|
| 5 |
-
colorTo: red
|
| 6 |
-
sdk: gradio
|
| 7 |
-
sdk_version: 4.12.0
|
| 8 |
-
app_file: app.py
|
| 9 |
-
pinned: false
|
| 10 |
-
license: unknown
|
| 11 |
-
---
|
| 12 |
|
| 13 |
-
Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
|
|
|
|
| 1 |
+
# sin-kaf
|
| 2 |
+
# dataset link (Turkis)
|
| 3 |
+
# https://sites.google.com/site/offensevalsharedtask/more-datasets
|
| 4 |
+
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 5 |
|
|
|
app.py
ADDED
|
@@ -0,0 +1,49 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import gradio as gra
|
| 2 |
+
import torch
|
| 3 |
+
import numpy as np
|
| 4 |
+
from transformers import AutoModelForSequenceClassification
|
| 5 |
+
from transformers import AutoTokenizer
|
| 6 |
+
from optimum.onnxruntime import ORTModel
|
| 7 |
+
import onnxruntime as rt
|
| 8 |
+
|
| 9 |
+
|
| 10 |
+
ort_session = rt.InferenceSession("/DATA/sin-kaf/onnx_model/model.onnx")
|
| 11 |
+
ort_session.get_providers()
|
| 12 |
+
|
| 13 |
+
# model = ORTModel.load_model("/DATA/sin-kaf/onnx_model/model.onnx")
|
| 14 |
+
# model = AutoModelForSequenceClassification.from_pretrained('/DATA/sin-kaf/test_trainer/checkpoint-18500')
|
| 15 |
+
tokenizer = AutoTokenizer.from_pretrained("Overfit-GM/distilbert-base-turkish-cased-offensive")
|
| 16 |
+
|
| 17 |
+
def user_greeting(sent):
|
| 18 |
+
|
| 19 |
+
encoded_dict = tokenizer.encode_plus(
|
| 20 |
+
sent,
|
| 21 |
+
add_special_tokens = True,
|
| 22 |
+
max_length = 64,
|
| 23 |
+
pad_to_max_length = True,
|
| 24 |
+
return_attention_mask = True,
|
| 25 |
+
return_tensors = 'pt',
|
| 26 |
+
)
|
| 27 |
+
|
| 28 |
+
|
| 29 |
+
input_ids = encoded_dict['input_ids']
|
| 30 |
+
attention_masks = encoded_dict['attention_mask']
|
| 31 |
+
|
| 32 |
+
|
| 33 |
+
input_ids = torch.cat([input_ids], dim=0)
|
| 34 |
+
input_mask = torch.cat([attention_masks], dim=0)
|
| 35 |
+
|
| 36 |
+
input_feed = {
|
| 37 |
+
"input_ids": input_ids.tolist(),
|
| 38 |
+
"attention_mask":input_mask.tolist(),
|
| 39 |
+
}
|
| 40 |
+
output = ort_session.run(None, input_feed)
|
| 41 |
+
return np.argmax((output[0][0]))
|
| 42 |
+
# outputs = model(input_ids, input_mask)
|
| 43 |
+
# return torch.argmax(outputs['logits'])
|
| 44 |
+
|
| 45 |
+
|
| 46 |
+
|
| 47 |
+
app = gra.Interface(fn = user_greeting, inputs="text", outputs="text")
|
| 48 |
+
app.launch()
|
| 49 |
+
# app.launch(server_name="0.0.0.0")
|
cluster_outliers.csv
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
main.ipynb
ADDED
|
@@ -0,0 +1,1008 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"cells": [
|
| 3 |
+
{
|
| 4 |
+
"cell_type": "code",
|
| 5 |
+
"execution_count": 1,
|
| 6 |
+
"metadata": {},
|
| 7 |
+
"outputs": [],
|
| 8 |
+
"source": [
|
| 9 |
+
"# dataset link (Turkis)\n",
|
| 10 |
+
"# https://sites.google.com/site/offensevalsharedtask/more-datasets"
|
| 11 |
+
]
|
| 12 |
+
},
|
| 13 |
+
{
|
| 14 |
+
"cell_type": "code",
|
| 15 |
+
"execution_count": 1,
|
| 16 |
+
"metadata": {},
|
| 17 |
+
"outputs": [
|
| 18 |
+
{
|
| 19 |
+
"name": "stderr",
|
| 20 |
+
"output_type": "stream",
|
| 21 |
+
"text": [
|
| 22 |
+
"/home/sebit/anaconda3/envs/dl_env/lib/python3.9/site-packages/neptune/internal/backends/hosted_client.py:51: NeptuneDeprecationWarning: The 'neptune-client' package has been deprecated and will be removed in the future. Install the 'neptune' package instead. For more, see https://docs.neptune.ai/setup/upgrading/\n",
|
| 23 |
+
" from neptune.version import version as neptune_client_version\n",
|
| 24 |
+
"/home/sebit/anaconda3/envs/dl_env/lib/python3.9/site-packages/pytorch_lightning/loggers/neptune.py:39: NeptuneDeprecationWarning: You're importing the Neptune client library via the deprecated `neptune.new` module, which will be removed in a future release. Import directly from `neptune` instead.\n",
|
| 25 |
+
" from neptune import new as neptune\n"
|
| 26 |
+
]
|
| 27 |
+
}
|
| 28 |
+
],
|
| 29 |
+
"source": [
|
| 30 |
+
"import os\n",
|
| 31 |
+
"import numpy as np\n",
|
| 32 |
+
"import pandas as pd\n",
|
| 33 |
+
"import pytorch_lightning as pl\n",
|
| 34 |
+
"import random\n",
|
| 35 |
+
"import torch\n",
|
| 36 |
+
"import emoji\n",
|
| 37 |
+
"\n",
|
| 38 |
+
"\n",
|
| 39 |
+
"import datetime\n",
|
| 40 |
+
"import numpy as np\n",
|
| 41 |
+
"import torch.optim as optim\n",
|
| 42 |
+
"\n",
|
| 43 |
+
"\n",
|
| 44 |
+
"import torch.nn as nn\n",
|
| 45 |
+
"\n",
|
| 46 |
+
"from torch.utils.data import DataLoader,Dataset,random_split,TensorDataset ,RandomSampler, SequentialSampler\n",
|
| 47 |
+
"from torchmetrics import Accuracy, F1Score \n",
|
| 48 |
+
"from sklearn.preprocessing import LabelEncoder\n",
|
| 49 |
+
"from pytorch_lightning.callbacks import EarlyStopping,ModelCheckpoint\n",
|
| 50 |
+
"from pytorch_lightning.loggers import TensorBoardLogger,MLFlowLogger\n",
|
| 51 |
+
"from sklearn.model_selection import train_test_split\n",
|
| 52 |
+
"\n",
|
| 53 |
+
"from sklearn.preprocessing import LabelEncoder\n",
|
| 54 |
+
"from transformers import BertForSequenceClassification, AdamW, BertConfig,BertTokenizer,get_linear_schedule_with_warmup"
|
| 55 |
+
]
|
| 56 |
+
},
|
| 57 |
+
{
|
| 58 |
+
"cell_type": "code",
|
| 59 |
+
"execution_count": 2,
|
| 60 |
+
"metadata": {},
|
| 61 |
+
"outputs": [
|
| 62 |
+
{
|
| 63 |
+
"data": {
|
| 64 |
+
"text/plain": [
|
| 65 |
+
"device(type='cuda', index=0)"
|
| 66 |
+
]
|
| 67 |
+
},
|
| 68 |
+
"execution_count": 2,
|
| 69 |
+
"metadata": {},
|
| 70 |
+
"output_type": "execute_result"
|
| 71 |
+
}
|
| 72 |
+
],
|
| 73 |
+
"source": [
|
| 74 |
+
"device = torch.device(\"cuda:0\" if torch.cuda.is_available() else \"cpu\")\n",
|
| 75 |
+
"device"
|
| 76 |
+
]
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"cell_type": "code",
|
| 80 |
+
"execution_count": 3,
|
| 81 |
+
"metadata": {},
|
| 82 |
+
"outputs": [
|
| 83 |
+
{
|
| 84 |
+
"data": {
|
| 85 |
+
"text/plain": [
|
| 86 |
+
"True"
|
| 87 |
+
]
|
| 88 |
+
},
|
| 89 |
+
"execution_count": 3,
|
| 90 |
+
"metadata": {},
|
| 91 |
+
"output_type": "execute_result"
|
| 92 |
+
}
|
| 93 |
+
],
|
| 94 |
+
"source": [
|
| 95 |
+
"torch.cuda.is_available()"
|
| 96 |
+
]
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"cell_type": "code",
|
| 100 |
+
"execution_count": 4,
|
| 101 |
+
"metadata": {},
|
| 102 |
+
"outputs": [],
|
| 103 |
+
"source": [
|
| 104 |
+
"seed_val = 42\n",
|
| 105 |
+
"random.seed(seed_val)\n",
|
| 106 |
+
"np.random.seed(seed_val)\n",
|
| 107 |
+
"torch.manual_seed(seed_val)\n",
|
| 108 |
+
"torch.cuda.manual_seed_all(seed_val)"
|
| 109 |
+
]
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"attachments": {},
|
| 113 |
+
"cell_type": "markdown",
|
| 114 |
+
"metadata": {},
|
| 115 |
+
"source": [
|
| 116 |
+
"# load dataaset\n"
|
| 117 |
+
]
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"cell_type": "code",
|
| 121 |
+
"execution_count": 5,
|
| 122 |
+
"metadata": {},
|
| 123 |
+
"outputs": [],
|
| 124 |
+
"source": [
|
| 125 |
+
"# train_df=pd.read_csv('SemEval-2020 dataset/offenseval2020-turkish/offenseval2020-turkish/offenseval-tr-training-v1/offenseval-tr-training-v1.tsv',sep='\\t')\n",
|
| 126 |
+
"# test_df=pd.read_csv('SemEval-2020 dataset/offenseval2020-turkish/offenseval2020-turkish/offenseval-tr-testset-v1/offenseval-tr-testset-v1.tsv',sep='\\t')"
|
| 127 |
+
]
|
| 128 |
+
},
|
| 129 |
+
{
|
| 130 |
+
"cell_type": "code",
|
| 131 |
+
"execution_count": 6,
|
| 132 |
+
"metadata": {},
|
| 133 |
+
"outputs": [
|
| 134 |
+
{
|
| 135 |
+
"ename": "NameError",
|
| 136 |
+
"evalue": "name 'train_df' is not defined",
|
| 137 |
+
"output_type": "error",
|
| 138 |
+
"traceback": [
|
| 139 |
+
"\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
|
| 140 |
+
"\u001b[0;31mNameError\u001b[0m Traceback (most recent call last)",
|
| 141 |
+
"Cell \u001b[0;32mIn[6], line 1\u001b[0m\n\u001b[0;32m----> 1\u001b[0m train_df\u001b[39m=\u001b[39mpd\u001b[39m.\u001b[39mconcat([train_df,test_df], axis\u001b[39m=\u001b[39m\u001b[39m0\u001b[39m)\n\u001b[1;32m 2\u001b[0m train_df\u001b[39m=\u001b[39mtrain_df\u001b[39m.\u001b[39mdrop([\u001b[39m'\u001b[39m\u001b[39mid\u001b[39m\u001b[39m'\u001b[39m], axis\u001b[39m=\u001b[39m\u001b[39m1\u001b[39m)\n",
|
| 142 |
+
"\u001b[0;31mNameError\u001b[0m: name 'train_df' is not defined"
|
| 143 |
+
]
|
| 144 |
+
}
|
| 145 |
+
],
|
| 146 |
+
"source": [
|
| 147 |
+
"train_df=pd.concat([train_df,test_df], axis=0)\n",
|
| 148 |
+
"train_df=train_df.drop(['id'], axis=1)"
|
| 149 |
+
]
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"cell_type": "code",
|
| 153 |
+
"execution_count": null,
|
| 154 |
+
"metadata": {},
|
| 155 |
+
"outputs": [
|
| 156 |
+
{
|
| 157 |
+
"data": {
|
| 158 |
+
"text/plain": [
|
| 159 |
+
"subtask_a\n",
|
| 160 |
+
"NOT 25231\n",
|
| 161 |
+
"OFF 6046\n",
|
| 162 |
+
"Name: count, dtype: int64"
|
| 163 |
+
]
|
| 164 |
+
},
|
| 165 |
+
"execution_count": 7,
|
| 166 |
+
"metadata": {},
|
| 167 |
+
"output_type": "execute_result"
|
| 168 |
+
}
|
| 169 |
+
],
|
| 170 |
+
"source": [
|
| 171 |
+
"train_df['subtask_a'].value_counts()"
|
| 172 |
+
]
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"cell_type": "code",
|
| 176 |
+
"execution_count": 8,
|
| 177 |
+
"metadata": {},
|
| 178 |
+
"outputs": [],
|
| 179 |
+
"source": [
|
| 180 |
+
"data=train_df['tweet'].tolist()\n",
|
| 181 |
+
"for i in range(len(data)):\n",
|
| 182 |
+
" data[i] = data[i].replace('@USER','')\n",
|
| 183 |
+
" data[i] = data[i].replace('#','')\n",
|
| 184 |
+
" data[i] = data[i].replace('$','')\n",
|
| 185 |
+
" data[i] = emoji.demojize(data[i])\n",
|
| 186 |
+
" \n",
|
| 187 |
+
"train_df['tweet'] = data"
|
| 188 |
+
]
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"cell_type": "code",
|
| 192 |
+
"execution_count": 9,
|
| 193 |
+
"metadata": {},
|
| 194 |
+
"outputs": [],
|
| 195 |
+
"source": [
|
| 196 |
+
"lab = LabelEncoder()\n",
|
| 197 |
+
"train_df['subtask_a'] = lab.fit_transform(train_df['subtask_a'])"
|
| 198 |
+
]
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"cell_type": "code",
|
| 202 |
+
"execution_count": 10,
|
| 203 |
+
"metadata": {},
|
| 204 |
+
"outputs": [
|
| 205 |
+
{
|
| 206 |
+
"data": {
|
| 207 |
+
"text/plain": [
|
| 208 |
+
"subtask_a\n",
|
| 209 |
+
"0 25231\n",
|
| 210 |
+
"1 6046\n",
|
| 211 |
+
"2 3515\n",
|
| 212 |
+
"Name: count, dtype: int64"
|
| 213 |
+
]
|
| 214 |
+
},
|
| 215 |
+
"execution_count": 10,
|
| 216 |
+
"metadata": {},
|
| 217 |
+
"output_type": "execute_result"
|
| 218 |
+
}
|
| 219 |
+
],
|
| 220 |
+
"source": [
|
| 221 |
+
"train_df['subtask_a'].value_counts()"
|
| 222 |
+
]
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"cell_type": "code",
|
| 226 |
+
"execution_count": 11,
|
| 227 |
+
"metadata": {},
|
| 228 |
+
"outputs": [],
|
| 229 |
+
"source": [
|
| 230 |
+
"train_df.drop(train_df[train_df['subtask_a'] == 2].index, inplace = True)"
|
| 231 |
+
]
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"cell_type": "code",
|
| 235 |
+
"execution_count": 12,
|
| 236 |
+
"metadata": {},
|
| 237 |
+
"outputs": [
|
| 238 |
+
{
|
| 239 |
+
"data": {
|
| 240 |
+
"text/plain": [
|
| 241 |
+
"subtask_a\n",
|
| 242 |
+
"0 22345\n",
|
| 243 |
+
"1 5417\n",
|
| 244 |
+
"Name: count, dtype: int64"
|
| 245 |
+
]
|
| 246 |
+
},
|
| 247 |
+
"execution_count": 12,
|
| 248 |
+
"metadata": {},
|
| 249 |
+
"output_type": "execute_result"
|
| 250 |
+
}
|
| 251 |
+
],
|
| 252 |
+
"source": [
|
| 253 |
+
"train_df['subtask_a'].value_counts()"
|
| 254 |
+
]
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"cell_type": "code",
|
| 258 |
+
"execution_count": 13,
|
| 259 |
+
"metadata": {},
|
| 260 |
+
"outputs": [
|
| 261 |
+
{
|
| 262 |
+
"data": {
|
| 263 |
+
"text/html": [
|
| 264 |
+
"<div>\n",
|
| 265 |
+
"<style scoped>\n",
|
| 266 |
+
" .dataframe tbody tr th:only-of-type {\n",
|
| 267 |
+
" vertical-align: middle;\n",
|
| 268 |
+
" }\n",
|
| 269 |
+
"\n",
|
| 270 |
+
" .dataframe tbody tr th {\n",
|
| 271 |
+
" vertical-align: top;\n",
|
| 272 |
+
" }\n",
|
| 273 |
+
"\n",
|
| 274 |
+
" .dataframe thead th {\n",
|
| 275 |
+
" text-align: right;\n",
|
| 276 |
+
" }\n",
|
| 277 |
+
"</style>\n",
|
| 278 |
+
"<table border=\"1\" class=\"dataframe\">\n",
|
| 279 |
+
" <thead>\n",
|
| 280 |
+
" <tr style=\"text-align: right;\">\n",
|
| 281 |
+
" <th></th>\n",
|
| 282 |
+
" <th>tweet</th>\n",
|
| 283 |
+
" <th>subtask_a</th>\n",
|
| 284 |
+
" </tr>\n",
|
| 285 |
+
" </thead>\n",
|
| 286 |
+
" <tbody>\n",
|
| 287 |
+
" <tr>\n",
|
| 288 |
+
" <th>3515</th>\n",
|
| 289 |
+
" <td>holstein ineği (alacalı siyah-beyaz inek, yani...</td>\n",
|
| 290 |
+
" <td>0</td>\n",
|
| 291 |
+
" </tr>\n",
|
| 292 |
+
" <tr>\n",
|
| 293 |
+
" <th>3516</th>\n",
|
| 294 |
+
" <td>Haaaa. O zaman oylar Binali'ye demek.</td>\n",
|
| 295 |
+
" <td>0</td>\n",
|
| 296 |
+
" </tr>\n",
|
| 297 |
+
" <tr>\n",
|
| 298 |
+
" <th>3517</th>\n",
|
| 299 |
+
" <td>Disk genel merkez yönetimine HDP'nin hiç etki...</td>\n",
|
| 300 |
+
" <td>0</td>\n",
|
| 301 |
+
" </tr>\n",
|
| 302 |
+
" <tr>\n",
|
| 303 |
+
" <th>3518</th>\n",
|
| 304 |
+
" <td>Bir insanı zorla kaliteli yapamazsın. Sen elin...</td>\n",
|
| 305 |
+
" <td>0</td>\n",
|
| 306 |
+
" </tr>\n",
|
| 307 |
+
" <tr>\n",
|
| 308 |
+
" <th>3519</th>\n",
|
| 309 |
+
" <td>Sus yaa açtım sonra korkudan telefon elimden ...</td>\n",
|
| 310 |
+
" <td>0</td>\n",
|
| 311 |
+
" </tr>\n",
|
| 312 |
+
" <tr>\n",
|
| 313 |
+
" <th>...</th>\n",
|
| 314 |
+
" <td>...</td>\n",
|
| 315 |
+
" <td>...</td>\n",
|
| 316 |
+
" </tr>\n",
|
| 317 |
+
" <tr>\n",
|
| 318 |
+
" <th>31272</th>\n",
|
| 319 |
+
" <td>Bu ödül sunan kızı kim giydirdiyse, kızın en b...</td>\n",
|
| 320 |
+
" <td>0</td>\n",
|
| 321 |
+
" </tr>\n",
|
| 322 |
+
" <tr>\n",
|
| 323 |
+
" <th>31273</th>\n",
|
| 324 |
+
" <td>Bunu sana beddua olarak etmiyorum bunlar ilerd...</td>\n",
|
| 325 |
+
" <td>0</td>\n",
|
| 326 |
+
" </tr>\n",
|
| 327 |
+
" <tr>\n",
|
| 328 |
+
" <th>31274</th>\n",
|
| 329 |
+
" <td>CHP'liler sandıkları bırakmıyor üstüne oturmuş...</td>\n",
|
| 330 |
+
" <td>1</td>\n",
|
| 331 |
+
" </tr>\n",
|
| 332 |
+
" <tr>\n",
|
| 333 |
+
" <th>31275</th>\n",
|
| 334 |
+
" <td>karanlığın içinde yalnız kalsam ne oluuuuurr</td>\n",
|
| 335 |
+
" <td>0</td>\n",
|
| 336 |
+
" </tr>\n",
|
| 337 |
+
" <tr>\n",
|
| 338 |
+
" <th>31276</th>\n",
|
| 339 |
+
" <td>Ne yalan söyleyeyim bu haftalıkta fitil olara...</td>\n",
|
| 340 |
+
" <td>0</td>\n",
|
| 341 |
+
" </tr>\n",
|
| 342 |
+
" </tbody>\n",
|
| 343 |
+
"</table>\n",
|
| 344 |
+
"<p>27762 rows × 2 columns</p>\n",
|
| 345 |
+
"</div>"
|
| 346 |
+
],
|
| 347 |
+
"text/plain": [
|
| 348 |
+
" tweet subtask_a\n",
|
| 349 |
+
"3515 holstein ineği (alacalı siyah-beyaz inek, yani... 0\n",
|
| 350 |
+
"3516 Haaaa. O zaman oylar Binali'ye demek. 0\n",
|
| 351 |
+
"3517 Disk genel merkez yönetimine HDP'nin hiç etki... 0\n",
|
| 352 |
+
"3518 Bir insanı zorla kaliteli yapamazsın. Sen elin... 0\n",
|
| 353 |
+
"3519 Sus yaa açtım sonra korkudan telefon elimden ... 0\n",
|
| 354 |
+
"... ... ...\n",
|
| 355 |
+
"31272 Bu ödül sunan kızı kim giydirdiyse, kızın en b... 0\n",
|
| 356 |
+
"31273 Bunu sana beddua olarak etmiyorum bunlar ilerd... 0\n",
|
| 357 |
+
"31274 CHP'liler sandıkları bırakmıyor üstüne oturmuş... 1\n",
|
| 358 |
+
"31275 karanlığın içinde yalnız kalsam ne oluuuuurr 0\n",
|
| 359 |
+
"31276 Ne yalan söyleyeyim bu haftalıkta fitil olara... 0\n",
|
| 360 |
+
"\n",
|
| 361 |
+
"[27762 rows x 2 columns]"
|
| 362 |
+
]
|
| 363 |
+
},
|
| 364 |
+
"execution_count": 13,
|
| 365 |
+
"metadata": {},
|
| 366 |
+
"output_type": "execute_result"
|
| 367 |
+
}
|
| 368 |
+
],
|
| 369 |
+
"source": [
|
| 370 |
+
"train_df"
|
| 371 |
+
]
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"cell_type": "code",
|
| 375 |
+
"execution_count": 14,
|
| 376 |
+
"metadata": {},
|
| 377 |
+
"outputs": [],
|
| 378 |
+
"source": [
|
| 379 |
+
"data = train_df.tweet.values\n",
|
| 380 |
+
"labels = train_df.subtask_a.values"
|
| 381 |
+
]
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"attachments": {},
|
| 385 |
+
"cell_type": "markdown",
|
| 386 |
+
"metadata": {},
|
| 387 |
+
"source": [
|
| 388 |
+
"# BERT Tokenizer"
|
| 389 |
+
]
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"cell_type": "code",
|
| 393 |
+
"execution_count": 15,
|
| 394 |
+
"metadata": {},
|
| 395 |
+
"outputs": [],
|
| 396 |
+
"source": [
|
| 397 |
+
"tokenizer = BertTokenizer.from_pretrained(\"bert-base-multilingual-cased\", do_basic_tokenize=True)\n",
|
| 398 |
+
"# tokenizer.add_tokens(data)"
|
| 399 |
+
]
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"cell_type": "code",
|
| 403 |
+
"execution_count": 16,
|
| 404 |
+
"metadata": {},
|
| 405 |
+
"outputs": [
|
| 406 |
+
{
|
| 407 |
+
"name": "stdout",
|
| 408 |
+
"output_type": "stream",
|
| 409 |
+
"text": [
|
| 410 |
+
" Original: Sallandık diyorum, merkezi bilmiyorum, sokağa fırlamadım, duruyorum. Senden bir açıklama gelmeden, ben bu sandığı terketmiyorum \n",
|
| 411 |
+
"Tokenized: ['Sal', '##landı', '##k', 'di', '##yor', '##um', ',', 'merkezi', 'bil', '##mi', '##yor', '##um', ',', 'sok', '##a', '##ğa', 'f', '##ır', '##lama', '##dı', '##m', ',', 'dur', '##uy', '##orum', '.', 'Sen', '##den', 'bir', 'açık', '##lama', 'gel', '##mede', '##n', ',', 'ben', 'bu', 'sand', '##ığı', 'ter', '##ket', '##mi', '##yor', '##um']\n",
|
| 412 |
+
"Token IDs: [64831, 35783, 10174, 10120, 26101, 10465, 117, 47522, 13897, 10500, 26101, 10465, 117, 29509, 10113, 25163, 174, 17145, 24540, 17532, 10147, 117, 28959, 53452, 28048, 119, 18082, 10633, 10561, 71769, 24540, 74458, 59268, 10115, 117, 11015, 11499, 45989, 28581, 12718, 13650, 10500, 26101, 10465]\n"
|
| 413 |
+
]
|
| 414 |
+
}
|
| 415 |
+
],
|
| 416 |
+
"source": [
|
| 417 |
+
"print(' Original: ', data[78])\n",
|
| 418 |
+
"print('Tokenized: ', tokenizer.tokenize(data[78]))\n",
|
| 419 |
+
"print('Token IDs: ', tokenizer.convert_tokens_to_ids(tokenizer.tokenize(data[78])))"
|
| 420 |
+
]
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"attachments": {},
|
| 424 |
+
"cell_type": "markdown",
|
| 425 |
+
"metadata": {},
|
| 426 |
+
"source": [
|
| 427 |
+
"# Tokenize Dataset"
|
| 428 |
+
]
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"cell_type": "code",
|
| 432 |
+
"execution_count": 17,
|
| 433 |
+
"metadata": {},
|
| 434 |
+
"outputs": [
|
| 435 |
+
{
|
| 436 |
+
"name": "stderr",
|
| 437 |
+
"output_type": "stream",
|
| 438 |
+
"text": [
|
| 439 |
+
"Token indices sequence length is longer than the specified maximum sequence length for this model (1277 > 512). Running this sequence through the model will result in indexing errors\n"
|
| 440 |
+
]
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"name": "stdout",
|
| 444 |
+
"output_type": "stream",
|
| 445 |
+
"text": [
|
| 446 |
+
"Max sentence length: 6906\n"
|
| 447 |
+
]
|
| 448 |
+
}
|
| 449 |
+
],
|
| 450 |
+
"source": [
|
| 451 |
+
"max_len = 0\n",
|
| 452 |
+
"for sent in data:\n",
|
| 453 |
+
"\n",
|
| 454 |
+
" input_ids = tokenizer.encode(sent, add_special_tokens=True)\n",
|
| 455 |
+
" max_len = max(max_len, len(input_ids))\n",
|
| 456 |
+
"\n",
|
| 457 |
+
"print('Max sentence length: ', max_len)"
|
| 458 |
+
]
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"cell_type": "code",
|
| 462 |
+
"execution_count": 18,
|
| 463 |
+
"metadata": {},
|
| 464 |
+
"outputs": [
|
| 465 |
+
{
|
| 466 |
+
"name": "stderr",
|
| 467 |
+
"output_type": "stream",
|
| 468 |
+
"text": [
|
| 469 |
+
"Truncation was not explicitly activated but `max_length` is provided a specific value, please use `truncation=True` to explicitly truncate examples to max length. Defaulting to 'longest_first' truncation strategy. If you encode pairs of sequences (GLUE-style) with the tokenizer you can select this strategy more precisely by providing a specific strategy to `truncation`.\n",
|
| 470 |
+
"/home/sebit/anaconda3/envs/testenv/lib/python3.9/site-packages/transformers/tokenization_utils_base.py:2418: FutureWarning: The `pad_to_max_length` argument is deprecated and will be removed in a future version, use `padding=True` or `padding='longest'` to pad to the longest sequence in the batch, or use `padding='max_length'` to pad to a max length. In this case, you can give a specific length with `max_length` (e.g. `max_length=45`) or leave max_length to None to pad to the maximal input size of the model (e.g. 512 for Bert).\n",
|
| 471 |
+
" warnings.warn(\n"
|
| 472 |
+
]
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"name": "stdout",
|
| 476 |
+
"output_type": "stream",
|
| 477 |
+
"text": [
|
| 478 |
+
"Original: holstein ineği (alacalı siyah-beyaz inek, yani hollanda ineği) en verimli süt alınan inek ırkıymış, trt belgesel'de öyle söylediler\n",
|
| 479 |
+
"Token IDs: tensor([ 101, 110516, 16206, 10106, 10112, 16054, 113, 21739, 15794,\n",
|
| 480 |
+
" 10713, 34543, 10237, 118, 110744, 10106, 10707, 117, 84251,\n",
|
| 481 |
+
" 46232, 41971, 10106, 10112, 16054, 114, 10110, 55011, 98373,\n",
|
| 482 |
+
" 187, 41559, 10164, 65890, 10106, 10707, 321, 16299, 10713,\n",
|
| 483 |
+
" 16889, 19733, 117, 32221, 10123, 34831, 12912, 112, 10104,\n",
|
| 484 |
+
" 276, 18369, 100721, 18369, 28113, 10165, 102, 0, 0,\n",
|
| 485 |
+
" 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
|
| 486 |
+
" 0])\n"
|
| 487 |
+
]
|
| 488 |
+
}
|
| 489 |
+
],
|
| 490 |
+
"source": [
|
| 491 |
+
"input_ids = []\n",
|
| 492 |
+
"attention_masks = []\n",
|
| 493 |
+
"\n",
|
| 494 |
+
"for sent in data:\n",
|
| 495 |
+
" encoded_dict = tokenizer.encode_plus(\n",
|
| 496 |
+
" sent, \n",
|
| 497 |
+
" add_special_tokens = True, \n",
|
| 498 |
+
" max_length = 64, \n",
|
| 499 |
+
" pad_to_max_length = True,\n",
|
| 500 |
+
" return_attention_mask = True, \n",
|
| 501 |
+
" return_tensors = 'pt', \n",
|
| 502 |
+
" )\n",
|
| 503 |
+
" \n",
|
| 504 |
+
" \n",
|
| 505 |
+
" input_ids.append(encoded_dict['input_ids'])\n",
|
| 506 |
+
" attention_masks.append(encoded_dict['attention_mask'])\n",
|
| 507 |
+
"\n",
|
| 508 |
+
"\n",
|
| 509 |
+
"input_ids = torch.cat(input_ids, dim=0)\n",
|
| 510 |
+
"attention_masks = torch.cat(attention_masks, dim=0)\n",
|
| 511 |
+
"labels = torch.tensor(labels)\n",
|
| 512 |
+
"\n",
|
| 513 |
+
"\n",
|
| 514 |
+
"print('Original: ', data[0])\n",
|
| 515 |
+
"print('Token IDs:', input_ids[0])"
|
| 516 |
+
]
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"attachments": {},
|
| 520 |
+
"cell_type": "markdown",
|
| 521 |
+
"metadata": {},
|
| 522 |
+
"source": [
|
| 523 |
+
"# Split Dataset"
|
| 524 |
+
]
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"cell_type": "code",
|
| 528 |
+
"execution_count": 19,
|
| 529 |
+
"metadata": {},
|
| 530 |
+
"outputs": [
|
| 531 |
+
{
|
| 532 |
+
"name": "stdout",
|
| 533 |
+
"output_type": "stream",
|
| 534 |
+
"text": [
|
| 535 |
+
"24,985 training samples\n",
|
| 536 |
+
"2,777 validation samples\n"
|
| 537 |
+
]
|
| 538 |
+
}
|
| 539 |
+
],
|
| 540 |
+
"source": [
|
| 541 |
+
"dataset = TensorDataset(input_ids, attention_masks, labels)\n",
|
| 542 |
+
"train_size = int(0.9 * len(dataset))\n",
|
| 543 |
+
"val_size = len(dataset) - train_size\n",
|
| 544 |
+
"\n",
|
| 545 |
+
"\n",
|
| 546 |
+
"train_dataset, val_dataset = random_split(dataset, [train_size, val_size])\n",
|
| 547 |
+
"\n",
|
| 548 |
+
"print('{:>5,} training samples'.format(train_size))\n",
|
| 549 |
+
"print('{:>5,} validation samples'.format(val_size))"
|
| 550 |
+
]
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"cell_type": "code",
|
| 554 |
+
"execution_count": 20,
|
| 555 |
+
"metadata": {},
|
| 556 |
+
"outputs": [
|
| 557 |
+
{
|
| 558 |
+
"name": "stderr",
|
| 559 |
+
"output_type": "stream",
|
| 560 |
+
"text": [
|
| 561 |
+
"Some weights of BertForSequenceClassification were not initialized from the model checkpoint at bert-base-multilingual-cased and are newly initialized: ['classifier.bias', 'classifier.weight']\n",
|
| 562 |
+
"You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference.\n"
|
| 563 |
+
]
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"data": {
|
| 567 |
+
"text/plain": [
|
| 568 |
+
"BertForSequenceClassification(\n",
|
| 569 |
+
" (bert): BertModel(\n",
|
| 570 |
+
" (embeddings): BertEmbeddings(\n",
|
| 571 |
+
" (word_embeddings): Embedding(119547, 768, padding_idx=0)\n",
|
| 572 |
+
" (position_embeddings): Embedding(512, 768)\n",
|
| 573 |
+
" (token_type_embeddings): Embedding(2, 768)\n",
|
| 574 |
+
" (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
|
| 575 |
+
" (dropout): Dropout(p=0.1, inplace=False)\n",
|
| 576 |
+
" )\n",
|
| 577 |
+
" (encoder): BertEncoder(\n",
|
| 578 |
+
" (layer): ModuleList(\n",
|
| 579 |
+
" (0-11): 12 x BertLayer(\n",
|
| 580 |
+
" (attention): BertAttention(\n",
|
| 581 |
+
" (self): BertSelfAttention(\n",
|
| 582 |
+
" (query): Linear(in_features=768, out_features=768, bias=True)\n",
|
| 583 |
+
" (key): Linear(in_features=768, out_features=768, bias=True)\n",
|
| 584 |
+
" (value): Linear(in_features=768, out_features=768, bias=True)\n",
|
| 585 |
+
" (dropout): Dropout(p=0.1, inplace=False)\n",
|
| 586 |
+
" )\n",
|
| 587 |
+
" (output): BertSelfOutput(\n",
|
| 588 |
+
" (dense): Linear(in_features=768, out_features=768, bias=True)\n",
|
| 589 |
+
" (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
|
| 590 |
+
" (dropout): Dropout(p=0.1, inplace=False)\n",
|
| 591 |
+
" )\n",
|
| 592 |
+
" )\n",
|
| 593 |
+
" (intermediate): BertIntermediate(\n",
|
| 594 |
+
" (dense): Linear(in_features=768, out_features=3072, bias=True)\n",
|
| 595 |
+
" (intermediate_act_fn): GELUActivation()\n",
|
| 596 |
+
" )\n",
|
| 597 |
+
" (output): BertOutput(\n",
|
| 598 |
+
" (dense): Linear(in_features=3072, out_features=768, bias=True)\n",
|
| 599 |
+
" (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)\n",
|
| 600 |
+
" (dropout): Dropout(p=0.1, inplace=False)\n",
|
| 601 |
+
" )\n",
|
| 602 |
+
" )\n",
|
| 603 |
+
" )\n",
|
| 604 |
+
" )\n",
|
| 605 |
+
" (pooler): BertPooler(\n",
|
| 606 |
+
" (dense): Linear(in_features=768, out_features=768, bias=True)\n",
|
| 607 |
+
" (activation): Tanh()\n",
|
| 608 |
+
" )\n",
|
| 609 |
+
" )\n",
|
| 610 |
+
" (dropout): Dropout(p=0.1, inplace=False)\n",
|
| 611 |
+
" (classifier): Linear(in_features=768, out_features=2, bias=True)\n",
|
| 612 |
+
")"
|
| 613 |
+
]
|
| 614 |
+
},
|
| 615 |
+
"execution_count": 20,
|
| 616 |
+
"metadata": {},
|
| 617 |
+
"output_type": "execute_result"
|
| 618 |
+
}
|
| 619 |
+
],
|
| 620 |
+
"source": [
|
| 621 |
+
"from transformers import BertForSequenceClassification, AdamW, BertConfig\n",
|
| 622 |
+
"\n",
|
| 623 |
+
"model = BertForSequenceClassification.from_pretrained(\n",
|
| 624 |
+
" \"bert-base-multilingual-cased\",\n",
|
| 625 |
+
" num_labels = 2, \n",
|
| 626 |
+
" output_attentions = False,\n",
|
| 627 |
+
" output_hidden_states = False, \n",
|
| 628 |
+
")\n",
|
| 629 |
+
"\n",
|
| 630 |
+
"model.cuda()"
|
| 631 |
+
]
|
| 632 |
+
},
|
| 633 |
+
{
|
| 634 |
+
"cell_type": "code",
|
| 635 |
+
"execution_count": 21,
|
| 636 |
+
"metadata": {},
|
| 637 |
+
"outputs": [
|
| 638 |
+
{
|
| 639 |
+
"name": "stdout",
|
| 640 |
+
"output_type": "stream",
|
| 641 |
+
"text": [
|
| 642 |
+
"The BERT model has 201 different named parameters.\n",
|
| 643 |
+
"\n",
|
| 644 |
+
"==== Embedding Layer ====\n",
|
| 645 |
+
"\n",
|
| 646 |
+
"bert.embeddings.word_embeddings.weight (119547, 768)\n",
|
| 647 |
+
"bert.embeddings.position_embeddings.weight (512, 768)\n",
|
| 648 |
+
"bert.embeddings.token_type_embeddings.weight (2, 768)\n",
|
| 649 |
+
"bert.embeddings.LayerNorm.weight (768,)\n",
|
| 650 |
+
"bert.embeddings.LayerNorm.bias (768,)\n",
|
| 651 |
+
"\n",
|
| 652 |
+
"==== First Transformer ====\n",
|
| 653 |
+
"\n",
|
| 654 |
+
"bert.encoder.layer.0.attention.self.query.weight (768, 768)\n",
|
| 655 |
+
"bert.encoder.layer.0.attention.self.query.bias (768,)\n",
|
| 656 |
+
"bert.encoder.layer.0.attention.self.key.weight (768, 768)\n",
|
| 657 |
+
"bert.encoder.layer.0.attention.self.key.bias (768,)\n",
|
| 658 |
+
"bert.encoder.layer.0.attention.self.value.weight (768, 768)\n",
|
| 659 |
+
"bert.encoder.layer.0.attention.self.value.bias (768,)\n",
|
| 660 |
+
"bert.encoder.layer.0.attention.output.dense.weight (768, 768)\n",
|
| 661 |
+
"bert.encoder.layer.0.attention.output.dense.bias (768,)\n",
|
| 662 |
+
"bert.encoder.layer.0.attention.output.LayerNorm.weight (768,)\n",
|
| 663 |
+
"bert.encoder.layer.0.attention.output.LayerNorm.bias (768,)\n",
|
| 664 |
+
"bert.encoder.layer.0.intermediate.dense.weight (3072, 768)\n",
|
| 665 |
+
"bert.encoder.layer.0.intermediate.dense.bias (3072,)\n",
|
| 666 |
+
"bert.encoder.layer.0.output.dense.weight (768, 3072)\n",
|
| 667 |
+
"bert.encoder.layer.0.output.dense.bias (768,)\n",
|
| 668 |
+
"bert.encoder.layer.0.output.LayerNorm.weight (768,)\n",
|
| 669 |
+
"bert.encoder.layer.0.output.LayerNorm.bias (768,)\n",
|
| 670 |
+
"\n",
|
| 671 |
+
"==== Output Layer ====\n",
|
| 672 |
+
"\n",
|
| 673 |
+
"bert.pooler.dense.weight (768, 768)\n",
|
| 674 |
+
"bert.pooler.dense.bias (768,)\n",
|
| 675 |
+
"classifier.weight (2, 768)\n",
|
| 676 |
+
"classifier.bias (2,)\n"
|
| 677 |
+
]
|
| 678 |
+
}
|
| 679 |
+
],
|
| 680 |
+
"source": [
|
| 681 |
+
"params = list(model.named_parameters())\n",
|
| 682 |
+
"\n",
|
| 683 |
+
"print('The BERT model has {:} different named parameters.\\n'.format(len(params)))\n",
|
| 684 |
+
"\n",
|
| 685 |
+
"print('==== Embedding Layer ====\\n')\n",
|
| 686 |
+
"\n",
|
| 687 |
+
"for p in params[0:5]:\n",
|
| 688 |
+
" print(\"{:<55} {:>12}\".format(p[0], str(tuple(p[1].size()))))\n",
|
| 689 |
+
"\n",
|
| 690 |
+
"print('\\n==== First Transformer ====\\n')\n",
|
| 691 |
+
"\n",
|
| 692 |
+
"for p in params[5:21]:\n",
|
| 693 |
+
" print(\"{:<55} {:>12}\".format(p[0], str(tuple(p[1].size()))))\n",
|
| 694 |
+
"\n",
|
| 695 |
+
"print('\\n==== Output Layer ====\\n')\n",
|
| 696 |
+
"\n",
|
| 697 |
+
"for p in params[-4:]:\n",
|
| 698 |
+
" print(\"{:<55} {:>12}\".format(p[0], str(tuple(p[1].size()))))"
|
| 699 |
+
]
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"cell_type": "code",
|
| 703 |
+
"execution_count": 22,
|
| 704 |
+
"metadata": {},
|
| 705 |
+
"outputs": [
|
| 706 |
+
{
|
| 707 |
+
"name": "stderr",
|
| 708 |
+
"output_type": "stream",
|
| 709 |
+
"text": [
|
| 710 |
+
"/home/sebit/anaconda3/envs/testenv/lib/python3.9/site-packages/transformers/optimization.py:411: FutureWarning: This implementation of AdamW is deprecated and will be removed in a future version. Use the PyTorch implementation torch.optim.AdamW instead, or set `no_deprecation_warning=True` to disable this warning\n",
|
| 711 |
+
" warnings.warn(\n"
|
| 712 |
+
]
|
| 713 |
+
}
|
| 714 |
+
],
|
| 715 |
+
"source": [
|
| 716 |
+
"optimizer = AdamW(model.parameters(),\n",
|
| 717 |
+
" lr = 2e-5,\n",
|
| 718 |
+
" eps = 1e-8\n",
|
| 719 |
+
" )"
|
| 720 |
+
]
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"cell_type": "code",
|
| 724 |
+
"execution_count": 23,
|
| 725 |
+
"metadata": {},
|
| 726 |
+
"outputs": [],
|
| 727 |
+
"source": [
|
| 728 |
+
"def flat_accuracy(preds, labels):\n",
|
| 729 |
+
" pred_flat = np.argmax(preds, axis=1).flatten()\n",
|
| 730 |
+
" labels_flat = labels.flatten()\n",
|
| 731 |
+
" return np.sum(pred_flat == labels_flat) / len(labels_flat)"
|
| 732 |
+
]
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"cell_type": "code",
|
| 736 |
+
"execution_count": 24,
|
| 737 |
+
"metadata": {},
|
| 738 |
+
"outputs": [],
|
| 739 |
+
"source": [
|
| 740 |
+
"def format_time(elapsed):\n",
|
| 741 |
+
"\n",
|
| 742 |
+
" elapsed_rounded = int(round((elapsed)))\n",
|
| 743 |
+
" return str(datetime.timedelta(seconds=elapsed_rounded))\n"
|
| 744 |
+
]
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"cell_type": "code",
|
| 748 |
+
"execution_count": 25,
|
| 749 |
+
"metadata": {},
|
| 750 |
+
"outputs": [],
|
| 751 |
+
"source": [
|
| 752 |
+
"class sinKafModel(pl.LightningModule):\n",
|
| 753 |
+
" def __init__(self, model, optimizer, scheduler):\n",
|
| 754 |
+
" super().__init__()\n",
|
| 755 |
+
" self.model = model\n",
|
| 756 |
+
" self.optimizer = optimizer\n",
|
| 757 |
+
" self.scheduler = scheduler\n",
|
| 758 |
+
"\n",
|
| 759 |
+
"\n",
|
| 760 |
+
" def forward(self, input_ids, attention_mask, labels):\n",
|
| 761 |
+
" outputs = self.model(input_ids, attention_mask=attention_mask, labels=labels)\n",
|
| 762 |
+
" return outputs\n",
|
| 763 |
+
"\n",
|
| 764 |
+
" def training_step(self, batch, batch_idx):\n",
|
| 765 |
+
" input_ids, input_mask, labels = batch\n",
|
| 766 |
+
" outputs = self(input_ids, input_mask, labels)\n",
|
| 767 |
+
" loss = outputs.loss\n",
|
| 768 |
+
" self.log('train_loss', loss)\n",
|
| 769 |
+
" return loss\n",
|
| 770 |
+
"\n",
|
| 771 |
+
" def validation_step(self, batch, batch_idx):\n",
|
| 772 |
+
" input_ids, input_mask, labels = batch\n",
|
| 773 |
+
" outputs = self(input_ids, input_mask, labels)\n",
|
| 774 |
+
" loss = outputs.loss\n",
|
| 775 |
+
" logits = outputs.logits\n",
|
| 776 |
+
" preds = torch.argmax(logits, dim=1)\n",
|
| 777 |
+
" acc = (preds == labels).sum().item() / len(labels)\n",
|
| 778 |
+
" self.log('val_loss', loss)\n",
|
| 779 |
+
" self.log('val_acc', acc)\n",
|
| 780 |
+
" return loss\n",
|
| 781 |
+
"\n",
|
| 782 |
+
" def configure_optimizers(self):\n",
|
| 783 |
+
" return [self.optimizer], [self.scheduler]\n",
|
| 784 |
+
"\n",
|
| 785 |
+
" # def train_dataloader(self):\n",
|
| 786 |
+
" # return self.train_dataloader\n",
|
| 787 |
+
"\n",
|
| 788 |
+
" # def val_dataloader(self):\n",
|
| 789 |
+
" # return self.validation_dataloader\n"
|
| 790 |
+
]
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"cell_type": "code",
|
| 794 |
+
"execution_count": 26,
|
| 795 |
+
"metadata": {},
|
| 796 |
+
"outputs": [],
|
| 797 |
+
"source": [
|
| 798 |
+
"train_dataloader = DataLoader(train_dataset, sampler = RandomSampler(train_dataset), batch_size = 2 )\n",
|
| 799 |
+
"validation_dataloader = DataLoader(val_dataset, sampler = SequentialSampler(val_dataset), batch_size = 2 )"
|
| 800 |
+
]
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"cell_type": "code",
|
| 804 |
+
"execution_count": 27,
|
| 805 |
+
"metadata": {},
|
| 806 |
+
"outputs": [],
|
| 807 |
+
"source": [
|
| 808 |
+
"epochs = 4\n",
|
| 809 |
+
"total_steps = len(train_dataloader) * epochs\n",
|
| 810 |
+
"scheduler = get_linear_schedule_with_warmup(optimizer, \n",
|
| 811 |
+
" num_warmup_steps = 0, \n",
|
| 812 |
+
" num_training_steps = total_steps)"
|
| 813 |
+
]
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"cell_type": "code",
|
| 817 |
+
"execution_count": 28,
|
| 818 |
+
"metadata": {},
|
| 819 |
+
"outputs": [
|
| 820 |
+
{
|
| 821 |
+
"name": "stderr",
|
| 822 |
+
"output_type": "stream",
|
| 823 |
+
"text": [
|
| 824 |
+
"GPU available: True (cuda), used: True\n",
|
| 825 |
+
"TPU available: False, using: 0 TPU cores\n",
|
| 826 |
+
"IPU available: False, using: 0 IPUs\n",
|
| 827 |
+
"HPU available: False, using: 0 HPUs\n",
|
| 828 |
+
"/home/sebit/anaconda3/envs/testenv/lib/python3.9/site-packages/pytorch_lightning/trainer/connectors/logger_connector/logger_connector.py:67: UserWarning: Starting from v1.9.0, `tensorboardX` has been removed as a dependency of the `pytorch_lightning` package, due to potential conflicts with other packages in the ML ecosystem. For this reason, `logger=True` will use `CSVLogger` as the default logger, unless the `tensorboard` or `tensorboardX` packages are found. Please `pip install lightning[extra]` or one of them to enable TensorBoard support by default\n",
|
| 829 |
+
" warning_cache.warn(\n"
|
| 830 |
+
]
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"name": "stderr",
|
| 834 |
+
"output_type": "stream",
|
| 835 |
+
"text": [
|
| 836 |
+
"LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0]\n",
|
| 837 |
+
"\n",
|
| 838 |
+
" | Name | Type | Params\n",
|
| 839 |
+
"--------------------------------------------------------\n",
|
| 840 |
+
"0 | model | BertForSequenceClassification | 177 M \n",
|
| 841 |
+
"--------------------------------------------------------\n",
|
| 842 |
+
"177 M Trainable params\n",
|
| 843 |
+
"0 Non-trainable params\n",
|
| 844 |
+
"177 M Total params\n",
|
| 845 |
+
"711.420 Total estimated model params size (MB)\n"
|
| 846 |
+
]
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"name": "stdout",
|
| 850 |
+
"output_type": "stream",
|
| 851 |
+
"text": [
|
| 852 |
+
"Sanity Checking DataLoader 0: 0%| | 0/2 [00:00<?, ?it/s]"
|
| 853 |
+
]
|
| 854 |
+
},
|
| 855 |
+
{
|
| 856 |
+
"name": "stderr",
|
| 857 |
+
"output_type": "stream",
|
| 858 |
+
"text": [
|
| 859 |
+
"/home/sebit/anaconda3/envs/testenv/lib/python3.9/site-packages/pytorch_lightning/trainer/connectors/data_connector.py:442: PossibleUserWarning: The dataloader, val_dataloader, does not have many workers which may be a bottleneck. Consider increasing the value of the `num_workers` argument` (try 8 which is the number of cpus on this machine) in the `DataLoader` init to improve performance.\n",
|
| 860 |
+
" rank_zero_warn(\n"
|
| 861 |
+
]
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"name": "stdout",
|
| 865 |
+
"output_type": "stream",
|
| 866 |
+
"text": [
|
| 867 |
+
" "
|
| 868 |
+
]
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"name": "stderr",
|
| 872 |
+
"output_type": "stream",
|
| 873 |
+
"text": [
|
| 874 |
+
"/home/sebit/anaconda3/envs/testenv/lib/python3.9/site-packages/pytorch_lightning/trainer/connectors/data_connector.py:442: PossibleUserWarning: The dataloader, train_dataloader, does not have many workers which may be a bottleneck. Consider increasing the value of the `num_workers` argument` (try 8 which is the number of cpus on this machine) in the `DataLoader` init to improve performance.\n",
|
| 875 |
+
" rank_zero_warn(\n"
|
| 876 |
+
]
|
| 877 |
+
},
|
| 878 |
+
{
|
| 879 |
+
"name": "stdout",
|
| 880 |
+
"output_type": "stream",
|
| 881 |
+
"text": [
|
| 882 |
+
"Epoch 0: 0%| | 1/1249 [00:00<05:01, 4.13it/s, v_num=6]"
|
| 883 |
+
]
|
| 884 |
+
},
|
| 885 |
+
{
|
| 886 |
+
"ename": "OutOfMemoryError",
|
| 887 |
+
"evalue": "CUDA out of memory. Tried to allocate 352.00 MiB (GPU 0; 4.00 GiB total capacity; 2.67 GiB already allocated; 0 bytes free; 2.80 GiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation. See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF",
|
| 888 |
+
"output_type": "error",
|
| 889 |
+
"traceback": [
|
| 890 |
+
"\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
|
| 891 |
+
"\u001b[0;31mOutOfMemoryError\u001b[0m Traceback (most recent call last)",
|
| 892 |
+
"Cell \u001b[0;32mIn[28], line 3\u001b[0m\n\u001b[1;32m 1\u001b[0m model \u001b[39m=\u001b[39m sinKafModel(model, optimizer, scheduler)\n\u001b[1;32m 2\u001b[0m trainer \u001b[39m=\u001b[39m pl\u001b[39m.\u001b[39mTrainer( max_epochs\u001b[39m=\u001b[39mepochs, limit_train_batches\u001b[39m=\u001b[39m\u001b[39m0.1\u001b[39m, devices\u001b[39m=\u001b[39m\u001b[39m1\u001b[39m, accelerator\u001b[39m=\u001b[39m\u001b[39m'\u001b[39m\u001b[39mgpu\u001b[39m\u001b[39m'\u001b[39m) \n\u001b[0;32m----> 3\u001b[0m trainer\u001b[39m.\u001b[39;49mfit(model,train_dataloader,validation_dataloader )\n",
|
| 893 |
+
"File \u001b[0;32m~/anaconda3/envs/testenv/lib/python3.9/site-packages/pytorch_lightning/trainer/trainer.py:532\u001b[0m, in \u001b[0;36mTrainer.fit\u001b[0;34m(self, model, train_dataloaders, val_dataloaders, datamodule, ckpt_path)\u001b[0m\n\u001b[1;32m 530\u001b[0m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mstrategy\u001b[39m.\u001b[39m_lightning_module \u001b[39m=\u001b[39m model\n\u001b[1;32m 531\u001b[0m _verify_strategy_supports_compile(model, \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mstrategy)\n\u001b[0;32m--> 532\u001b[0m call\u001b[39m.\u001b[39;49m_call_and_handle_interrupt(\n\u001b[1;32m 533\u001b[0m \u001b[39mself\u001b[39;49m, \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49m_fit_impl, model, train_dataloaders, val_dataloaders, datamodule, ckpt_path\n\u001b[1;32m 534\u001b[0m )\n",
|
| 894 |
+
"File \u001b[0;32m~/anaconda3/envs/testenv/lib/python3.9/site-packages/pytorch_lightning/trainer/call.py:43\u001b[0m, in \u001b[0;36m_call_and_handle_interrupt\u001b[0;34m(trainer, trainer_fn, *args, **kwargs)\u001b[0m\n\u001b[1;32m 41\u001b[0m \u001b[39mif\u001b[39;00m trainer\u001b[39m.\u001b[39mstrategy\u001b[39m.\u001b[39mlauncher \u001b[39mis\u001b[39;00m \u001b[39mnot\u001b[39;00m \u001b[39mNone\u001b[39;00m:\n\u001b[1;32m 42\u001b[0m \u001b[39mreturn\u001b[39;00m trainer\u001b[39m.\u001b[39mstrategy\u001b[39m.\u001b[39mlauncher\u001b[39m.\u001b[39mlaunch(trainer_fn, \u001b[39m*\u001b[39margs, trainer\u001b[39m=\u001b[39mtrainer, \u001b[39m*\u001b[39m\u001b[39m*\u001b[39mkwargs)\n\u001b[0;32m---> 43\u001b[0m \u001b[39mreturn\u001b[39;00m trainer_fn(\u001b[39m*\u001b[39;49margs, \u001b[39m*\u001b[39;49m\u001b[39m*\u001b[39;49mkwargs)\n\u001b[1;32m 45\u001b[0m \u001b[39mexcept\u001b[39;00m _TunerExitException:\n\u001b[1;32m 46\u001b[0m _call_teardown_hook(trainer)\n",
|
| 895 |
+
"File \u001b[0;32m~/anaconda3/envs/testenv/lib/python3.9/site-packages/pytorch_lightning/trainer/trainer.py:571\u001b[0m, in \u001b[0;36mTrainer._fit_impl\u001b[0;34m(self, model, train_dataloaders, val_dataloaders, datamodule, ckpt_path)\u001b[0m\n\u001b[1;32m 561\u001b[0m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_data_connector\u001b[39m.\u001b[39mattach_data(\n\u001b[1;32m 562\u001b[0m model, train_dataloaders\u001b[39m=\u001b[39mtrain_dataloaders, val_dataloaders\u001b[39m=\u001b[39mval_dataloaders, datamodule\u001b[39m=\u001b[39mdatamodule\n\u001b[1;32m 563\u001b[0m )\n\u001b[1;32m 565\u001b[0m ckpt_path \u001b[39m=\u001b[39m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_checkpoint_connector\u001b[39m.\u001b[39m_select_ckpt_path(\n\u001b[1;32m 566\u001b[0m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mstate\u001b[39m.\u001b[39mfn,\n\u001b[1;32m 567\u001b[0m ckpt_path,\n\u001b[1;32m 568\u001b[0m model_provided\u001b[39m=\u001b[39m\u001b[39mTrue\u001b[39;00m,\n\u001b[1;32m 569\u001b[0m model_connected\u001b[39m=\u001b[39m\u001b[39mself\u001b[39m\u001b[39m.\u001b[39mlightning_module \u001b[39mis\u001b[39;00m \u001b[39mnot\u001b[39;00m \u001b[39mNone\u001b[39;00m,\n\u001b[1;32m 570\u001b[0m )\n\u001b[0;32m--> 571\u001b[0m \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49m_run(model, ckpt_path\u001b[39m=\u001b[39;49mckpt_path)\n\u001b[1;32m 573\u001b[0m \u001b[39massert\u001b[39;00m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mstate\u001b[39m.\u001b[39mstopped\n\u001b[1;32m 574\u001b[0m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mtraining \u001b[39m=\u001b[39m \u001b[39mFalse\u001b[39;00m\n",
|
| 896 |
+
"File \u001b[0;32m~/anaconda3/envs/testenv/lib/python3.9/site-packages/pytorch_lightning/trainer/trainer.py:980\u001b[0m, in \u001b[0;36mTrainer._run\u001b[0;34m(self, model, ckpt_path)\u001b[0m\n\u001b[1;32m 975\u001b[0m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_signal_connector\u001b[39m.\u001b[39mregister_signal_handlers()\n\u001b[1;32m 977\u001b[0m \u001b[39m# ----------------------------\u001b[39;00m\n\u001b[1;32m 978\u001b[0m \u001b[39m# RUN THE TRAINER\u001b[39;00m\n\u001b[1;32m 979\u001b[0m \u001b[39m# ----------------------------\u001b[39;00m\n\u001b[0;32m--> 980\u001b[0m results \u001b[39m=\u001b[39m \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49m_run_stage()\n\u001b[1;32m 982\u001b[0m \u001b[39m# ----------------------------\u001b[39;00m\n\u001b[1;32m 983\u001b[0m \u001b[39m# POST-Training CLEAN UP\u001b[39;00m\n\u001b[1;32m 984\u001b[0m \u001b[39m# ----------------------------\u001b[39;00m\n\u001b[1;32m 985\u001b[0m log\u001b[39m.\u001b[39mdebug(\u001b[39mf\u001b[39m\u001b[39m\"\u001b[39m\u001b[39m{\u001b[39;00m\u001b[39mself\u001b[39m\u001b[39m.\u001b[39m\u001b[39m__class__\u001b[39m\u001b[39m.\u001b[39m\u001b[39m__name__\u001b[39m\u001b[39m}\u001b[39;00m\u001b[39m: trainer tearing down\u001b[39m\u001b[39m\"\u001b[39m)\n",
|
| 897 |
+
"File \u001b[0;32m~/anaconda3/envs/testenv/lib/python3.9/site-packages/pytorch_lightning/trainer/trainer.py:1023\u001b[0m, in \u001b[0;36mTrainer._run_stage\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m 1021\u001b[0m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_run_sanity_check()\n\u001b[1;32m 1022\u001b[0m \u001b[39mwith\u001b[39;00m torch\u001b[39m.\u001b[39mautograd\u001b[39m.\u001b[39mset_detect_anomaly(\u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_detect_anomaly):\n\u001b[0;32m-> 1023\u001b[0m \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49mfit_loop\u001b[39m.\u001b[39;49mrun()\n\u001b[1;32m 1024\u001b[0m \u001b[39mreturn\u001b[39;00m \u001b[39mNone\u001b[39;00m\n\u001b[1;32m 1025\u001b[0m \u001b[39mraise\u001b[39;00m \u001b[39mRuntimeError\u001b[39;00m(\u001b[39mf\u001b[39m\u001b[39m\"\u001b[39m\u001b[39mUnexpected state \u001b[39m\u001b[39m{\u001b[39;00m\u001b[39mself\u001b[39m\u001b[39m.\u001b[39mstate\u001b[39m}\u001b[39;00m\u001b[39m\"\u001b[39m)\n",
|
| 898 |
+
"File \u001b[0;32m~/anaconda3/envs/testenv/lib/python3.9/site-packages/pytorch_lightning/loops/fit_loop.py:202\u001b[0m, in \u001b[0;36m_FitLoop.run\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m 200\u001b[0m \u001b[39mtry\u001b[39;00m:\n\u001b[1;32m 201\u001b[0m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mon_advance_start()\n\u001b[0;32m--> 202\u001b[0m \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49madvance()\n\u001b[1;32m 203\u001b[0m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mon_advance_end()\n\u001b[1;32m 204\u001b[0m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_restarting \u001b[39m=\u001b[39m \u001b[39mFalse\u001b[39;00m\n",
|
| 899 |
+
"File \u001b[0;32m~/anaconda3/envs/testenv/lib/python3.9/site-packages/pytorch_lightning/loops/fit_loop.py:355\u001b[0m, in \u001b[0;36m_FitLoop.advance\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m 353\u001b[0m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_data_fetcher\u001b[39m.\u001b[39msetup(combined_loader)\n\u001b[1;32m 354\u001b[0m \u001b[39mwith\u001b[39;00m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mtrainer\u001b[39m.\u001b[39mprofiler\u001b[39m.\u001b[39mprofile(\u001b[39m\"\u001b[39m\u001b[39mrun_training_epoch\u001b[39m\u001b[39m\"\u001b[39m):\n\u001b[0;32m--> 355\u001b[0m \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49mepoch_loop\u001b[39m.\u001b[39;49mrun(\u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49m_data_fetcher)\n",
|
| 900 |
+
"File \u001b[0;32m~/anaconda3/envs/testenv/lib/python3.9/site-packages/pytorch_lightning/loops/training_epoch_loop.py:133\u001b[0m, in \u001b[0;36m_TrainingEpochLoop.run\u001b[0;34m(self, data_fetcher)\u001b[0m\n\u001b[1;32m 131\u001b[0m \u001b[39mwhile\u001b[39;00m \u001b[39mnot\u001b[39;00m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mdone:\n\u001b[1;32m 132\u001b[0m \u001b[39mtry\u001b[39;00m:\n\u001b[0;32m--> 133\u001b[0m \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49madvance(data_fetcher)\n\u001b[1;32m 134\u001b[0m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mon_advance_end()\n\u001b[1;32m 135\u001b[0m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_restarting \u001b[39m=\u001b[39m \u001b[39mFalse\u001b[39;00m\n",
|
| 901 |
+
"File \u001b[0;32m~/anaconda3/envs/testenv/lib/python3.9/site-packages/pytorch_lightning/loops/training_epoch_loop.py:219\u001b[0m, in \u001b[0;36m_TrainingEpochLoop.advance\u001b[0;34m(self, data_fetcher)\u001b[0m\n\u001b[1;32m 216\u001b[0m \u001b[39mwith\u001b[39;00m trainer\u001b[39m.\u001b[39mprofiler\u001b[39m.\u001b[39mprofile(\u001b[39m\"\u001b[39m\u001b[39mrun_training_batch\u001b[39m\u001b[39m\"\u001b[39m):\n\u001b[1;32m 217\u001b[0m \u001b[39mif\u001b[39;00m trainer\u001b[39m.\u001b[39mlightning_module\u001b[39m.\u001b[39mautomatic_optimization:\n\u001b[1;32m 218\u001b[0m \u001b[39m# in automatic optimization, there can only be one optimizer\u001b[39;00m\n\u001b[0;32m--> 219\u001b[0m batch_output \u001b[39m=\u001b[39m \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49mautomatic_optimization\u001b[39m.\u001b[39;49mrun(trainer\u001b[39m.\u001b[39;49moptimizers[\u001b[39m0\u001b[39;49m], kwargs)\n\u001b[1;32m 220\u001b[0m \u001b[39melse\u001b[39;00m:\n\u001b[1;32m 221\u001b[0m batch_output \u001b[39m=\u001b[39m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mmanual_optimization\u001b[39m.\u001b[39mrun(kwargs)\n",
|
| 902 |
+
"File \u001b[0;32m~/anaconda3/envs/testenv/lib/python3.9/site-packages/pytorch_lightning/loops/optimization/automatic.py:188\u001b[0m, in \u001b[0;36m_AutomaticOptimization.run\u001b[0;34m(self, optimizer, kwargs)\u001b[0m\n\u001b[1;32m 181\u001b[0m closure()\n\u001b[1;32m 183\u001b[0m \u001b[39m# ------------------------------\u001b[39;00m\n\u001b[1;32m 184\u001b[0m \u001b[39m# BACKWARD PASS\u001b[39;00m\n\u001b[1;32m 185\u001b[0m \u001b[39m# ------------------------------\u001b[39;00m\n\u001b[1;32m 186\u001b[0m \u001b[39m# gradient update with accumulated gradients\u001b[39;00m\n\u001b[1;32m 187\u001b[0m \u001b[39melse\u001b[39;00m:\n\u001b[0;32m--> 188\u001b[0m \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49m_optimizer_step(kwargs\u001b[39m.\u001b[39;49mget(\u001b[39m\"\u001b[39;49m\u001b[39mbatch_idx\u001b[39;49m\u001b[39m\"\u001b[39;49m, \u001b[39m0\u001b[39;49m), closure)\n\u001b[1;32m 190\u001b[0m result \u001b[39m=\u001b[39m closure\u001b[39m.\u001b[39mconsume_result()\n\u001b[1;32m 191\u001b[0m \u001b[39mif\u001b[39;00m result\u001b[39m.\u001b[39mloss \u001b[39mis\u001b[39;00m \u001b[39mNone\u001b[39;00m:\n",
|
| 903 |
+
"File \u001b[0;32m~/anaconda3/envs/testenv/lib/python3.9/site-packages/pytorch_lightning/loops/optimization/automatic.py:266\u001b[0m, in \u001b[0;36m_AutomaticOptimization._optimizer_step\u001b[0;34m(self, batch_idx, train_step_and_backward_closure)\u001b[0m\n\u001b[1;32m 263\u001b[0m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39moptim_progress\u001b[39m.\u001b[39moptimizer\u001b[39m.\u001b[39mstep\u001b[39m.\u001b[39mincrement_ready()\n\u001b[1;32m 265\u001b[0m \u001b[39m# model hook\u001b[39;00m\n\u001b[0;32m--> 266\u001b[0m call\u001b[39m.\u001b[39;49m_call_lightning_module_hook(\n\u001b[1;32m 267\u001b[0m trainer,\n\u001b[1;32m 268\u001b[0m \u001b[39m\"\u001b[39;49m\u001b[39moptimizer_step\u001b[39;49m\u001b[39m\"\u001b[39;49m,\n\u001b[1;32m 269\u001b[0m trainer\u001b[39m.\u001b[39;49mcurrent_epoch,\n\u001b[1;32m 270\u001b[0m batch_idx,\n\u001b[1;32m 271\u001b[0m optimizer,\n\u001b[1;32m 272\u001b[0m train_step_and_backward_closure,\n\u001b[1;32m 273\u001b[0m )\n\u001b[1;32m 275\u001b[0m \u001b[39mif\u001b[39;00m \u001b[39mnot\u001b[39;00m should_accumulate:\n\u001b[1;32m 276\u001b[0m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39moptim_progress\u001b[39m.\u001b[39moptimizer\u001b[39m.\u001b[39mstep\u001b[39m.\u001b[39mincrement_completed()\n",
|
| 904 |
+
"File \u001b[0;32m~/anaconda3/envs/testenv/lib/python3.9/site-packages/pytorch_lightning/trainer/call.py:146\u001b[0m, in \u001b[0;36m_call_lightning_module_hook\u001b[0;34m(trainer, hook_name, pl_module, *args, **kwargs)\u001b[0m\n\u001b[1;32m 143\u001b[0m pl_module\u001b[39m.\u001b[39m_current_fx_name \u001b[39m=\u001b[39m hook_name\n\u001b[1;32m 145\u001b[0m \u001b[39mwith\u001b[39;00m trainer\u001b[39m.\u001b[39mprofiler\u001b[39m.\u001b[39mprofile(\u001b[39mf\u001b[39m\u001b[39m\"\u001b[39m\u001b[39m[LightningModule]\u001b[39m\u001b[39m{\u001b[39;00mpl_module\u001b[39m.\u001b[39m\u001b[39m__class__\u001b[39m\u001b[39m.\u001b[39m\u001b[39m__name__\u001b[39m\u001b[39m}\u001b[39;00m\u001b[39m.\u001b[39m\u001b[39m{\u001b[39;00mhook_name\u001b[39m}\u001b[39;00m\u001b[39m\"\u001b[39m):\n\u001b[0;32m--> 146\u001b[0m output \u001b[39m=\u001b[39m fn(\u001b[39m*\u001b[39;49margs, \u001b[39m*\u001b[39;49m\u001b[39m*\u001b[39;49mkwargs)\n\u001b[1;32m 148\u001b[0m \u001b[39m# restore current_fx when nested context\u001b[39;00m\n\u001b[1;32m 149\u001b[0m pl_module\u001b[39m.\u001b[39m_current_fx_name \u001b[39m=\u001b[39m prev_fx_name\n",
|
| 905 |
+
"File \u001b[0;32m~/anaconda3/envs/testenv/lib/python3.9/site-packages/pytorch_lightning/core/module.py:1270\u001b[0m, in \u001b[0;36mLightningModule.optimizer_step\u001b[0;34m(self, epoch, batch_idx, optimizer, optimizer_closure)\u001b[0m\n\u001b[1;32m 1232\u001b[0m \u001b[39mdef\u001b[39;00m \u001b[39moptimizer_step\u001b[39m(\n\u001b[1;32m 1233\u001b[0m \u001b[39mself\u001b[39m,\n\u001b[1;32m 1234\u001b[0m epoch: \u001b[39mint\u001b[39m,\n\u001b[0;32m (...)\u001b[0m\n\u001b[1;32m 1237\u001b[0m optimizer_closure: Optional[Callable[[], Any]] \u001b[39m=\u001b[39m \u001b[39mNone\u001b[39;00m,\n\u001b[1;32m 1238\u001b[0m ) \u001b[39m-\u001b[39m\u001b[39m>\u001b[39m \u001b[39mNone\u001b[39;00m:\n\u001b[1;32m 1239\u001b[0m \u001b[39m \u001b[39m\u001b[39mr\u001b[39m\u001b[39m\"\"\"Override this method to adjust the default way the :class:`~pytorch_lightning.trainer.trainer.Trainer`\u001b[39;00m\n\u001b[1;32m 1240\u001b[0m \u001b[39m calls the optimizer.\u001b[39;00m\n\u001b[1;32m 1241\u001b[0m \n\u001b[0;32m (...)\u001b[0m\n\u001b[1;32m 1268\u001b[0m \u001b[39m pg[\"lr\"] = lr_scale * self.learning_rate\u001b[39;00m\n\u001b[1;32m 1269\u001b[0m \u001b[39m \"\"\"\u001b[39;00m\n\u001b[0;32m-> 1270\u001b[0m optimizer\u001b[39m.\u001b[39;49mstep(closure\u001b[39m=\u001b[39;49moptimizer_closure)\n",
|
| 906 |
+
"File \u001b[0;32m~/anaconda3/envs/testenv/lib/python3.9/site-packages/pytorch_lightning/core/optimizer.py:161\u001b[0m, in \u001b[0;36mLightningOptimizer.step\u001b[0;34m(self, closure, **kwargs)\u001b[0m\n\u001b[1;32m 158\u001b[0m \u001b[39mraise\u001b[39;00m MisconfigurationException(\u001b[39m\"\u001b[39m\u001b[39mWhen `optimizer.step(closure)` is called, the closure should be callable\u001b[39m\u001b[39m\"\u001b[39m)\n\u001b[1;32m 160\u001b[0m \u001b[39massert\u001b[39;00m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_strategy \u001b[39mis\u001b[39;00m \u001b[39mnot\u001b[39;00m \u001b[39mNone\u001b[39;00m\n\u001b[0;32m--> 161\u001b[0m step_output \u001b[39m=\u001b[39m \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49m_strategy\u001b[39m.\u001b[39;49moptimizer_step(\u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49m_optimizer, closure, \u001b[39m*\u001b[39;49m\u001b[39m*\u001b[39;49mkwargs)\n\u001b[1;32m 163\u001b[0m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_on_after_step()\n\u001b[1;32m 165\u001b[0m \u001b[39mreturn\u001b[39;00m step_output\n",
|
| 907 |
+
"File \u001b[0;32m~/anaconda3/envs/testenv/lib/python3.9/site-packages/pytorch_lightning/strategies/strategy.py:231\u001b[0m, in \u001b[0;36mStrategy.optimizer_step\u001b[0;34m(self, optimizer, closure, model, **kwargs)\u001b[0m\n\u001b[1;32m 229\u001b[0m \u001b[39m# TODO(fabric): remove assertion once strategy's optimizer_step typing is fixed\u001b[39;00m\n\u001b[1;32m 230\u001b[0m \u001b[39massert\u001b[39;00m \u001b[39misinstance\u001b[39m(model, pl\u001b[39m.\u001b[39mLightningModule)\n\u001b[0;32m--> 231\u001b[0m \u001b[39mreturn\u001b[39;00m \u001b[39mself\u001b[39;49m\u001b[39m.\u001b[39;49mprecision_plugin\u001b[39m.\u001b[39;49moptimizer_step(optimizer, model\u001b[39m=\u001b[39;49mmodel, closure\u001b[39m=\u001b[39;49mclosure, \u001b[39m*\u001b[39;49m\u001b[39m*\u001b[39;49mkwargs)\n",
|
| 908 |
+
"File \u001b[0;32m~/anaconda3/envs/testenv/lib/python3.9/site-packages/pytorch_lightning/plugins/precision/precision_plugin.py:116\u001b[0m, in \u001b[0;36mPrecisionPlugin.optimizer_step\u001b[0;34m(self, optimizer, model, closure, **kwargs)\u001b[0m\n\u001b[1;32m 114\u001b[0m \u001b[39m\u001b[39m\u001b[39m\"\"\"Hook to run the optimizer step.\"\"\"\u001b[39;00m\n\u001b[1;32m 115\u001b[0m closure \u001b[39m=\u001b[39m partial(\u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_wrap_closure, model, optimizer, closure)\n\u001b[0;32m--> 116\u001b[0m \u001b[39mreturn\u001b[39;00m optimizer\u001b[39m.\u001b[39;49mstep(closure\u001b[39m=\u001b[39;49mclosure, \u001b[39m*\u001b[39;49m\u001b[39m*\u001b[39;49mkwargs)\n",
|
| 909 |
+
"File \u001b[0;32m~/anaconda3/envs/testenv/lib/python3.9/site-packages/torch/optim/lr_scheduler.py:69\u001b[0m, in \u001b[0;36mLRScheduler.__init__.<locals>.with_counter.<locals>.wrapper\u001b[0;34m(*args, **kwargs)\u001b[0m\n\u001b[1;32m 67\u001b[0m instance\u001b[39m.\u001b[39m_step_count \u001b[39m+\u001b[39m\u001b[39m=\u001b[39m \u001b[39m1\u001b[39m\n\u001b[1;32m 68\u001b[0m wrapped \u001b[39m=\u001b[39m func\u001b[39m.\u001b[39m\u001b[39m__get__\u001b[39m(instance, \u001b[39mcls\u001b[39m)\n\u001b[0;32m---> 69\u001b[0m \u001b[39mreturn\u001b[39;00m wrapped(\u001b[39m*\u001b[39;49margs, \u001b[39m*\u001b[39;49m\u001b[39m*\u001b[39;49mkwargs)\n",
|
| 910 |
+
"File \u001b[0;32m~/anaconda3/envs/testenv/lib/python3.9/site-packages/torch/optim/optimizer.py:280\u001b[0m, in \u001b[0;36mOptimizer.profile_hook_step.<locals>.wrapper\u001b[0;34m(*args, **kwargs)\u001b[0m\n\u001b[1;32m 276\u001b[0m \u001b[39melse\u001b[39;00m:\n\u001b[1;32m 277\u001b[0m \u001b[39mraise\u001b[39;00m \u001b[39mRuntimeError\u001b[39;00m(\u001b[39mf\u001b[39m\u001b[39m\"\u001b[39m\u001b[39m{\u001b[39;00mfunc\u001b[39m}\u001b[39;00m\u001b[39m must return None or a tuple of (new_args, new_kwargs),\u001b[39m\u001b[39m\"\u001b[39m\n\u001b[1;32m 278\u001b[0m \u001b[39mf\u001b[39m\u001b[39m\"\u001b[39m\u001b[39mbut got \u001b[39m\u001b[39m{\u001b[39;00mresult\u001b[39m}\u001b[39;00m\u001b[39m.\u001b[39m\u001b[39m\"\u001b[39m)\n\u001b[0;32m--> 280\u001b[0m out \u001b[39m=\u001b[39m func(\u001b[39m*\u001b[39;49margs, \u001b[39m*\u001b[39;49m\u001b[39m*\u001b[39;49mkwargs)\n\u001b[1;32m 281\u001b[0m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_optimizer_step_code()\n\u001b[1;32m 283\u001b[0m \u001b[39m# call optimizer step post hooks\u001b[39;00m\n",
|
| 911 |
+
"File \u001b[0;32m~/anaconda3/envs/testenv/lib/python3.9/site-packages/torch/utils/_contextlib.py:115\u001b[0m, in \u001b[0;36mcontext_decorator.<locals>.decorate_context\u001b[0;34m(*args, **kwargs)\u001b[0m\n\u001b[1;32m 112\u001b[0m \u001b[39m@functools\u001b[39m\u001b[39m.\u001b[39mwraps(func)\n\u001b[1;32m 113\u001b[0m \u001b[39mdef\u001b[39;00m \u001b[39mdecorate_context\u001b[39m(\u001b[39m*\u001b[39margs, \u001b[39m*\u001b[39m\u001b[39m*\u001b[39mkwargs):\n\u001b[1;32m 114\u001b[0m \u001b[39mwith\u001b[39;00m ctx_factory():\n\u001b[0;32m--> 115\u001b[0m \u001b[39mreturn\u001b[39;00m func(\u001b[39m*\u001b[39;49margs, \u001b[39m*\u001b[39;49m\u001b[39m*\u001b[39;49mkwargs)\n",
|
| 912 |
+
"File \u001b[0;32m~/anaconda3/envs/testenv/lib/python3.9/site-packages/transformers/optimization.py:468\u001b[0m, in \u001b[0;36mAdamW.step\u001b[0;34m(self, closure)\u001b[0m\n\u001b[1;32m 466\u001b[0m exp_avg\u001b[39m.\u001b[39mmul_(beta1)\u001b[39m.\u001b[39madd_(grad, alpha\u001b[39m=\u001b[39m(\u001b[39m1.0\u001b[39m \u001b[39m-\u001b[39m beta1))\n\u001b[1;32m 467\u001b[0m exp_avg_sq\u001b[39m.\u001b[39mmul_(beta2)\u001b[39m.\u001b[39maddcmul_(grad, grad, value\u001b[39m=\u001b[39m\u001b[39m1.0\u001b[39m \u001b[39m-\u001b[39m beta2)\n\u001b[0;32m--> 468\u001b[0m denom \u001b[39m=\u001b[39m exp_avg_sq\u001b[39m.\u001b[39;49msqrt()\u001b[39m.\u001b[39madd_(group[\u001b[39m\"\u001b[39m\u001b[39meps\u001b[39m\u001b[39m\"\u001b[39m])\n\u001b[1;32m 470\u001b[0m step_size \u001b[39m=\u001b[39m group[\u001b[39m\"\u001b[39m\u001b[39mlr\u001b[39m\u001b[39m\"\u001b[39m]\n\u001b[1;32m 471\u001b[0m \u001b[39mif\u001b[39;00m group[\u001b[39m\"\u001b[39m\u001b[39mcorrect_bias\u001b[39m\u001b[39m\"\u001b[39m]: \u001b[39m# No bias correction for Bert\u001b[39;00m\n",
|
| 913 |
+
"\u001b[0;31mOutOfMemoryError\u001b[0m: CUDA out of memory. Tried to allocate 352.00 MiB (GPU 0; 4.00 GiB total capacity; 2.67 GiB already allocated; 0 bytes free; 2.80 GiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation. See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF"
|
| 914 |
+
]
|
| 915 |
+
}
|
| 916 |
+
],
|
| 917 |
+
"source": [
|
| 918 |
+
"model = sinKafModel(model, optimizer, scheduler)\n",
|
| 919 |
+
"trainer = pl.Trainer( max_epochs=epochs, limit_train_batches=0.1, devices=1, accelerator='gpu') \n",
|
| 920 |
+
"trainer.fit(model,train_dataloader,validation_dataloader )"
|
| 921 |
+
]
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"cell_type": "code",
|
| 925 |
+
"execution_count": null,
|
| 926 |
+
"metadata": {},
|
| 927 |
+
"outputs": [],
|
| 928 |
+
"source": [
|
| 929 |
+
"sent = 'Koyunlar hasta'"
|
| 930 |
+
]
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"cell_type": "code",
|
| 934 |
+
"execution_count": null,
|
| 935 |
+
"metadata": {},
|
| 936 |
+
"outputs": [],
|
| 937 |
+
"source": [
|
| 938 |
+
"input_ids = []\n",
|
| 939 |
+
"attention_masks = []\n",
|
| 940 |
+
"\n",
|
| 941 |
+
"encoded_dict = tokenizer.encode_plus(\n",
|
| 942 |
+
" sent,\n",
|
| 943 |
+
" add_special_tokens = True,\n",
|
| 944 |
+
" max_length = 64,\n",
|
| 945 |
+
" pad_to_max_length = True,\n",
|
| 946 |
+
" return_attention_mask = True,\n",
|
| 947 |
+
" return_tensors = 'pt',\n",
|
| 948 |
+
" )\n",
|
| 949 |
+
"\n",
|
| 950 |
+
"\n",
|
| 951 |
+
"input_ids = encoded_dict['input_ids']\n",
|
| 952 |
+
"attention_masks = encoded_dict['attention_mask']\n",
|
| 953 |
+
"\n",
|
| 954 |
+
"\n",
|
| 955 |
+
"input_ids = torch.cat([input_ids], dim=0)\n",
|
| 956 |
+
"input_mask = torch.cat([attention_masks], dim=0)\n",
|
| 957 |
+
"labels = torch.tensor(labels)\n",
|
| 958 |
+
"\n",
|
| 959 |
+
"\n",
|
| 960 |
+
"\n",
|
| 961 |
+
"\n",
|
| 962 |
+
"print('Original: ', sent)\n",
|
| 963 |
+
"print('Token IDs:', input_ids)\n",
|
| 964 |
+
"print('Token IDs:', input_mask)"
|
| 965 |
+
]
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"cell_type": "code",
|
| 969 |
+
"execution_count": null,
|
| 970 |
+
"metadata": {},
|
| 971 |
+
"outputs": [],
|
| 972 |
+
"source": [
|
| 973 |
+
"outputs = model(input_ids, input_mask, labels[0])"
|
| 974 |
+
]
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"cell_type": "code",
|
| 978 |
+
"execution_count": null,
|
| 979 |
+
"metadata": {},
|
| 980 |
+
"outputs": [],
|
| 981 |
+
"source": [
|
| 982 |
+
"outputs[0]"
|
| 983 |
+
]
|
| 984 |
+
}
|
| 985 |
+
],
|
| 986 |
+
"metadata": {
|
| 987 |
+
"kernelspec": {
|
| 988 |
+
"display_name": "sbtenv",
|
| 989 |
+
"language": "python",
|
| 990 |
+
"name": "python3"
|
| 991 |
+
},
|
| 992 |
+
"language_info": {
|
| 993 |
+
"codemirror_mode": {
|
| 994 |
+
"name": "ipython",
|
| 995 |
+
"version": 3
|
| 996 |
+
},
|
| 997 |
+
"file_extension": ".py",
|
| 998 |
+
"mimetype": "text/x-python",
|
| 999 |
+
"name": "python",
|
| 1000 |
+
"nbconvert_exporter": "python",
|
| 1001 |
+
"pygments_lexer": "ipython3",
|
| 1002 |
+
"version": "3.9.0"
|
| 1003 |
+
},
|
| 1004 |
+
"orig_nbformat": 4
|
| 1005 |
+
},
|
| 1006 |
+
"nbformat": 4,
|
| 1007 |
+
"nbformat_minor": 2
|
| 1008 |
+
}
|
onnx_model/config.json
ADDED
|
@@ -0,0 +1,27 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "/DATA/sin-kaf/test_trainer/checkpoint-18500/config.json",
|
| 3 |
+
"activation": "gelu",
|
| 4 |
+
"architectures": [
|
| 5 |
+
"DistilBertForSequenceClassification"
|
| 6 |
+
],
|
| 7 |
+
"attention_dropout": 0.1,
|
| 8 |
+
"attention_probs_dropout_prob": 0.1,
|
| 9 |
+
"dim": 768,
|
| 10 |
+
"dropout": 0.1,
|
| 11 |
+
"hidden_dim": 3072,
|
| 12 |
+
"initializer_range": 0.02,
|
| 13 |
+
"max_position_embeddings": 512,
|
| 14 |
+
"model_type": "distilbert",
|
| 15 |
+
"n_heads": 12,
|
| 16 |
+
"n_layers": 6,
|
| 17 |
+
"output_past": true,
|
| 18 |
+
"pad_token_id": 0,
|
| 19 |
+
"problem_type": "single_label_classification",
|
| 20 |
+
"qa_dropout": 0.1,
|
| 21 |
+
"seq_classif_dropout": 0.2,
|
| 22 |
+
"sinusoidal_pos_embds": true,
|
| 23 |
+
"tie_weights_": true,
|
| 24 |
+
"torch_dtype": "float32",
|
| 25 |
+
"transformers_version": "4.34.1",
|
| 26 |
+
"vocab_size": 32000
|
| 27 |
+
}
|
onnx_model/model.onnx
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cee6ddb2c1e1abb21e513881265239a57dd3cba52f621b6c81a78e41e66eae09
|
| 3 |
+
size 272496128
|
outlier_detection.ipynb
ADDED
|
@@ -0,0 +1,2292 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"cells": [
|
| 3 |
+
{
|
| 4 |
+
"cell_type": "code",
|
| 5 |
+
"execution_count": 1,
|
| 6 |
+
"metadata": {},
|
| 7 |
+
"outputs": [],
|
| 8 |
+
"source": [
|
| 9 |
+
"import re\n",
|
| 10 |
+
"import tqdm\n",
|
| 11 |
+
"import numpy as np\n",
|
| 12 |
+
"import pandas as pd\n",
|
| 13 |
+
"import matplotlib.pyplot as plt\n",
|
| 14 |
+
"import seaborn as sns\n",
|
| 15 |
+
"\n",
|
| 16 |
+
"from sklearn.datasets import fetch_20newsgroups\n",
|
| 17 |
+
"from sklearn.manifold import TSNE\n"
|
| 18 |
+
]
|
| 19 |
+
},
|
| 20 |
+
{
|
| 21 |
+
"cell_type": "code",
|
| 22 |
+
"execution_count": 2,
|
| 23 |
+
"metadata": {},
|
| 24 |
+
"outputs": [],
|
| 25 |
+
"source": [
|
| 26 |
+
"embeding_df=pd.read_csv('/mnt/c/Users/selin_uzturk/Desktop/sinkaf/encoded.csv')\n",
|
| 27 |
+
"embeding_df=embeding_df.drop(['Unnamed: 0'], axis=1)\n",
|
| 28 |
+
"copy_df=pd.read_csv('/mnt/c/Users/selin_uzturk/Desktop/sinkaf/encoded.csv')\n",
|
| 29 |
+
"copy_df=copy_df.drop(['Unnamed: 0'], axis=1)\n"
|
| 30 |
+
]
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"cell_type": "code",
|
| 34 |
+
"execution_count": 3,
|
| 35 |
+
"metadata": {},
|
| 36 |
+
"outputs": [
|
| 37 |
+
{
|
| 38 |
+
"data": {
|
| 39 |
+
"text/html": [
|
| 40 |
+
"<div>\n",
|
| 41 |
+
"<style scoped>\n",
|
| 42 |
+
" .dataframe tbody tr th:only-of-type {\n",
|
| 43 |
+
" vertical-align: middle;\n",
|
| 44 |
+
" }\n",
|
| 45 |
+
"\n",
|
| 46 |
+
" .dataframe tbody tr th {\n",
|
| 47 |
+
" vertical-align: top;\n",
|
| 48 |
+
" }\n",
|
| 49 |
+
"\n",
|
| 50 |
+
" .dataframe thead th {\n",
|
| 51 |
+
" text-align: right;\n",
|
| 52 |
+
" }\n",
|
| 53 |
+
"</style>\n",
|
| 54 |
+
"<table border=\"1\" class=\"dataframe\">\n",
|
| 55 |
+
" <thead>\n",
|
| 56 |
+
" <tr style=\"text-align: right;\">\n",
|
| 57 |
+
" <th></th>\n",
|
| 58 |
+
" <th>0</th>\n",
|
| 59 |
+
" <th>1</th>\n",
|
| 60 |
+
" <th>2</th>\n",
|
| 61 |
+
" <th>3</th>\n",
|
| 62 |
+
" <th>4</th>\n",
|
| 63 |
+
" <th>5</th>\n",
|
| 64 |
+
" <th>6</th>\n",
|
| 65 |
+
" <th>7</th>\n",
|
| 66 |
+
" <th>8</th>\n",
|
| 67 |
+
" <th>9</th>\n",
|
| 68 |
+
" <th>...</th>\n",
|
| 69 |
+
" <th>56</th>\n",
|
| 70 |
+
" <th>57</th>\n",
|
| 71 |
+
" <th>58</th>\n",
|
| 72 |
+
" <th>59</th>\n",
|
| 73 |
+
" <th>60</th>\n",
|
| 74 |
+
" <th>61</th>\n",
|
| 75 |
+
" <th>62</th>\n",
|
| 76 |
+
" <th>63</th>\n",
|
| 77 |
+
" <th>labels</th>\n",
|
| 78 |
+
" <th>tweet</th>\n",
|
| 79 |
+
" </tr>\n",
|
| 80 |
+
" </thead>\n",
|
| 81 |
+
" <tbody>\n",
|
| 82 |
+
" <tr>\n",
|
| 83 |
+
" <th>0</th>\n",
|
| 84 |
+
" <td>101</td>\n",
|
| 85 |
+
" <td>10110</td>\n",
|
| 86 |
+
" <td>175</td>\n",
|
| 87 |
+
" <td>78653</td>\n",
|
| 88 |
+
" <td>189</td>\n",
|
| 89 |
+
" <td>25285</td>\n",
|
| 90 |
+
" <td>15976</td>\n",
|
| 91 |
+
" <td>40840</td>\n",
|
| 92 |
+
" <td>276</td>\n",
|
| 93 |
+
" <td>31623</td>\n",
|
| 94 |
+
" <td>...</td>\n",
|
| 95 |
+
" <td>0</td>\n",
|
| 96 |
+
" <td>0</td>\n",
|
| 97 |
+
" <td>0</td>\n",
|
| 98 |
+
" <td>0</td>\n",
|
| 99 |
+
" <td>0</td>\n",
|
| 100 |
+
" <td>0</td>\n",
|
| 101 |
+
" <td>0</td>\n",
|
| 102 |
+
" <td>0</td>\n",
|
| 103 |
+
" <td>0</td>\n",
|
| 104 |
+
" <td>en güzel uyuyan insan ödülü jeon jungkook'a g...</td>\n",
|
| 105 |
+
" </tr>\n",
|
| 106 |
+
" <tr>\n",
|
| 107 |
+
" <th>1</th>\n",
|
| 108 |
+
" <td>101</td>\n",
|
| 109 |
+
" <td>11589</td>\n",
|
| 110 |
+
" <td>10706</td>\n",
|
| 111 |
+
" <td>10713</td>\n",
|
| 112 |
+
" <td>10794</td>\n",
|
| 113 |
+
" <td>94698</td>\n",
|
| 114 |
+
" <td>30668</td>\n",
|
| 115 |
+
" <td>24883</td>\n",
|
| 116 |
+
" <td>117</td>\n",
|
| 117 |
+
" <td>23763</td>\n",
|
| 118 |
+
" <td>...</td>\n",
|
| 119 |
+
" <td>0</td>\n",
|
| 120 |
+
" <td>0</td>\n",
|
| 121 |
+
" <td>0</td>\n",
|
| 122 |
+
" <td>0</td>\n",
|
| 123 |
+
" <td>0</td>\n",
|
| 124 |
+
" <td>0</td>\n",
|
| 125 |
+
" <td>0</td>\n",
|
| 126 |
+
" <td>0</td>\n",
|
| 127 |
+
" <td>0</td>\n",
|
| 128 |
+
" <td>Mekanı cennet olsun, saygılar sayın avukatımı...</td>\n",
|
| 129 |
+
" </tr>\n",
|
| 130 |
+
" <tr>\n",
|
| 131 |
+
" <th>2</th>\n",
|
| 132 |
+
" <td>101</td>\n",
|
| 133 |
+
" <td>148</td>\n",
|
| 134 |
+
" <td>30471</td>\n",
|
| 135 |
+
" <td>10774</td>\n",
|
| 136 |
+
" <td>13785</td>\n",
|
| 137 |
+
" <td>13779</td>\n",
|
| 138 |
+
" <td>33642</td>\n",
|
| 139 |
+
" <td>14399</td>\n",
|
| 140 |
+
" <td>48271</td>\n",
|
| 141 |
+
" <td>76686</td>\n",
|
| 142 |
+
" <td>...</td>\n",
|
| 143 |
+
" <td>0</td>\n",
|
| 144 |
+
" <td>0</td>\n",
|
| 145 |
+
" <td>0</td>\n",
|
| 146 |
+
" <td>0</td>\n",
|
| 147 |
+
" <td>0</td>\n",
|
| 148 |
+
" <td>0</td>\n",
|
| 149 |
+
" <td>0</td>\n",
|
| 150 |
+
" <td>0</td>\n",
|
| 151 |
+
" <td>0</td>\n",
|
| 152 |
+
" <td>Kızlar aranızda kas yığını beylere düşenler ol...</td>\n",
|
| 153 |
+
" </tr>\n",
|
| 154 |
+
" <tr>\n",
|
| 155 |
+
" <th>3</th>\n",
|
| 156 |
+
" <td>101</td>\n",
|
| 157 |
+
" <td>19319</td>\n",
|
| 158 |
+
" <td>16724</td>\n",
|
| 159 |
+
" <td>10118</td>\n",
|
| 160 |
+
" <td>10107</td>\n",
|
| 161 |
+
" <td>78323</td>\n",
|
| 162 |
+
" <td>12407</td>\n",
|
| 163 |
+
" <td>38959</td>\n",
|
| 164 |
+
" <td>22934</td>\n",
|
| 165 |
+
" <td>10147</td>\n",
|
| 166 |
+
" <td>...</td>\n",
|
| 167 |
+
" <td>0</td>\n",
|
| 168 |
+
" <td>0</td>\n",
|
| 169 |
+
" <td>0</td>\n",
|
| 170 |
+
" <td>0</td>\n",
|
| 171 |
+
" <td>0</td>\n",
|
| 172 |
+
" <td>0</td>\n",
|
| 173 |
+
" <td>0</td>\n",
|
| 174 |
+
" <td>0</td>\n",
|
| 175 |
+
" <td>0</td>\n",
|
| 176 |
+
" <td>Biraz ders çalışayım. Tembellik ve uyku düşman...</td>\n",
|
| 177 |
+
" </tr>\n",
|
| 178 |
+
" <tr>\n",
|
| 179 |
+
" <th>4</th>\n",
|
| 180 |
+
" <td>101</td>\n",
|
| 181 |
+
" <td>30932</td>\n",
|
| 182 |
+
" <td>58706</td>\n",
|
| 183 |
+
" <td>58054</td>\n",
|
| 184 |
+
" <td>44907</td>\n",
|
| 185 |
+
" <td>10224</td>\n",
|
| 186 |
+
" <td>106583</td>\n",
|
| 187 |
+
" <td>10288</td>\n",
|
| 188 |
+
" <td>12524</td>\n",
|
| 189 |
+
" <td>13878</td>\n",
|
| 190 |
+
" <td>...</td>\n",
|
| 191 |
+
" <td>0</td>\n",
|
| 192 |
+
" <td>0</td>\n",
|
| 193 |
+
" <td>0</td>\n",
|
| 194 |
+
" <td>0</td>\n",
|
| 195 |
+
" <td>0</td>\n",
|
| 196 |
+
" <td>0</td>\n",
|
| 197 |
+
" <td>0</td>\n",
|
| 198 |
+
" <td>0</td>\n",
|
| 199 |
+
" <td>0</td>\n",
|
| 200 |
+
" <td>Trezeguet yerine El Sharawy daha iyi olmaz mı</td>\n",
|
| 201 |
+
" </tr>\n",
|
| 202 |
+
" <tr>\n",
|
| 203 |
+
" <th>...</th>\n",
|
| 204 |
+
" <td>...</td>\n",
|
| 205 |
+
" <td>...</td>\n",
|
| 206 |
+
" <td>...</td>\n",
|
| 207 |
+
" <td>...</td>\n",
|
| 208 |
+
" <td>...</td>\n",
|
| 209 |
+
" <td>...</td>\n",
|
| 210 |
+
" <td>...</td>\n",
|
| 211 |
+
" <td>...</td>\n",
|
| 212 |
+
" <td>...</td>\n",
|
| 213 |
+
" <td>...</td>\n",
|
| 214 |
+
" <td>...</td>\n",
|
| 215 |
+
" <td>...</td>\n",
|
| 216 |
+
" <td>...</td>\n",
|
| 217 |
+
" <td>...</td>\n",
|
| 218 |
+
" <td>...</td>\n",
|
| 219 |
+
" <td>...</td>\n",
|
| 220 |
+
" <td>...</td>\n",
|
| 221 |
+
" <td>...</td>\n",
|
| 222 |
+
" <td>...</td>\n",
|
| 223 |
+
" <td>...</td>\n",
|
| 224 |
+
" <td>...</td>\n",
|
| 225 |
+
" </tr>\n",
|
| 226 |
+
" <tr>\n",
|
| 227 |
+
" <th>43344</th>\n",
|
| 228 |
+
" <td>101</td>\n",
|
| 229 |
+
" <td>20065</td>\n",
|
| 230 |
+
" <td>10161</td>\n",
|
| 231 |
+
" <td>115</td>\n",
|
| 232 |
+
" <td>115</td>\n",
|
| 233 |
+
" <td>103784</td>\n",
|
| 234 |
+
" <td>10774</td>\n",
|
| 235 |
+
" <td>21388</td>\n",
|
| 236 |
+
" <td>10245</td>\n",
|
| 237 |
+
" <td>92067</td>\n",
|
| 238 |
+
" <td>...</td>\n",
|
| 239 |
+
" <td>0</td>\n",
|
| 240 |
+
" <td>0</td>\n",
|
| 241 |
+
" <td>0</td>\n",
|
| 242 |
+
" <td>0</td>\n",
|
| 243 |
+
" <td>0</td>\n",
|
| 244 |
+
" <td>0</td>\n",
|
| 245 |
+
" <td>0</td>\n",
|
| 246 |
+
" <td>0</td>\n",
|
| 247 |
+
" <td>1</td>\n",
|
| 248 |
+
" <td>Hil**adamlar kesinlikle kelimeleri anlamıyorla...</td>\n",
|
| 249 |
+
" </tr>\n",
|
| 250 |
+
" <tr>\n",
|
| 251 |
+
" <th>43345</th>\n",
|
| 252 |
+
" <td>101</td>\n",
|
| 253 |
+
" <td>139</td>\n",
|
| 254 |
+
" <td>80839</td>\n",
|
| 255 |
+
" <td>24109</td>\n",
|
| 256 |
+
" <td>13406</td>\n",
|
| 257 |
+
" <td>18985</td>\n",
|
| 258 |
+
" <td>16285</td>\n",
|
| 259 |
+
" <td>10163</td>\n",
|
| 260 |
+
" <td>11062</td>\n",
|
| 261 |
+
" <td>276</td>\n",
|
| 262 |
+
" <td>...</td>\n",
|
| 263 |
+
" <td>0</td>\n",
|
| 264 |
+
" <td>0</td>\n",
|
| 265 |
+
" <td>0</td>\n",
|
| 266 |
+
" <td>0</td>\n",
|
| 267 |
+
" <td>0</td>\n",
|
| 268 |
+
" <td>0</td>\n",
|
| 269 |
+
" <td>0</td>\n",
|
| 270 |
+
" <td>0</td>\n",
|
| 271 |
+
" <td>1</td>\n",
|
| 272 |
+
" <td>Böyle piçlerin çok erken ölmemelerini ve çok f...</td>\n",
|
| 273 |
+
" </tr>\n",
|
| 274 |
+
" <tr>\n",
|
| 275 |
+
" <th>43346</th>\n",
|
| 276 |
+
" <td>101</td>\n",
|
| 277 |
+
" <td>105549</td>\n",
|
| 278 |
+
" <td>102635</td>\n",
|
| 279 |
+
" <td>10140</td>\n",
|
| 280 |
+
" <td>26943</td>\n",
|
| 281 |
+
" <td>11499</td>\n",
|
| 282 |
+
" <td>110516</td>\n",
|
| 283 |
+
" <td>21899</td>\n",
|
| 284 |
+
" <td>11861</td>\n",
|
| 285 |
+
" <td>10561</td>\n",
|
| 286 |
+
" <td>...</td>\n",
|
| 287 |
+
" <td>0</td>\n",
|
| 288 |
+
" <td>0</td>\n",
|
| 289 |
+
" <td>0</td>\n",
|
| 290 |
+
" <td>0</td>\n",
|
| 291 |
+
" <td>0</td>\n",
|
| 292 |
+
" <td>0</td>\n",
|
| 293 |
+
" <td>0</td>\n",
|
| 294 |
+
" <td>0</td>\n",
|
| 295 |
+
" <td>1</td>\n",
|
| 296 |
+
" <td>Turgay denilen bu holigonda bir sorun yok, gur...</td>\n",
|
| 297 |
+
" </tr>\n",
|
| 298 |
+
" <tr>\n",
|
| 299 |
+
" <th>43347</th>\n",
|
| 300 |
+
" <td>101</td>\n",
|
| 301 |
+
" <td>81424</td>\n",
|
| 302 |
+
" <td>26398</td>\n",
|
| 303 |
+
" <td>92017</td>\n",
|
| 304 |
+
" <td>109620</td>\n",
|
| 305 |
+
" <td>10941</td>\n",
|
| 306 |
+
" <td>76010</td>\n",
|
| 307 |
+
" <td>10115</td>\n",
|
| 308 |
+
" <td>19830</td>\n",
|
| 309 |
+
" <td>26083</td>\n",
|
| 310 |
+
" <td>...</td>\n",
|
| 311 |
+
" <td>0</td>\n",
|
| 312 |
+
" <td>0</td>\n",
|
| 313 |
+
" <td>0</td>\n",
|
| 314 |
+
" <td>0</td>\n",
|
| 315 |
+
" <td>0</td>\n",
|
| 316 |
+
" <td>0</td>\n",
|
| 317 |
+
" <td>0</td>\n",
|
| 318 |
+
" <td>0</td>\n",
|
| 319 |
+
" <td>1</td>\n",
|
| 320 |
+
" <td>Umarım ülkenin düşük zekadan kurtulması ilgile...</td>\n",
|
| 321 |
+
" </tr>\n",
|
| 322 |
+
" <tr>\n",
|
| 323 |
+
" <th>43348</th>\n",
|
| 324 |
+
" <td>101</td>\n",
|
| 325 |
+
" <td>39774</td>\n",
|
| 326 |
+
" <td>11127</td>\n",
|
| 327 |
+
" <td>45989</td>\n",
|
| 328 |
+
" <td>24596</td>\n",
|
| 329 |
+
" <td>11933</td>\n",
|
| 330 |
+
" <td>170</td>\n",
|
| 331 |
+
" <td>17145</td>\n",
|
| 332 |
+
" <td>10710</td>\n",
|
| 333 |
+
" <td>39125</td>\n",
|
| 334 |
+
" <td>...</td>\n",
|
| 335 |
+
" <td>0</td>\n",
|
| 336 |
+
" <td>0</td>\n",
|
| 337 |
+
" <td>0</td>\n",
|
| 338 |
+
" <td>0</td>\n",
|
| 339 |
+
" <td>0</td>\n",
|
| 340 |
+
" <td>0</td>\n",
|
| 341 |
+
" <td>0</td>\n",
|
| 342 |
+
" <td>0</td>\n",
|
| 343 |
+
" <td>1</td>\n",
|
| 344 |
+
" <td>CHP sandıkları bırakmaz, üzerine oturur, bir c...</td>\n",
|
| 345 |
+
" </tr>\n",
|
| 346 |
+
" </tbody>\n",
|
| 347 |
+
"</table>\n",
|
| 348 |
+
"<p>43349 rows × 66 columns</p>\n",
|
| 349 |
+
"</div>"
|
| 350 |
+
],
|
| 351 |
+
"text/plain": [
|
| 352 |
+
" 0 1 2 3 4 5 6 7 8 \n",
|
| 353 |
+
"0 101 10110 175 78653 189 25285 15976 40840 276 \\\n",
|
| 354 |
+
"1 101 11589 10706 10713 10794 94698 30668 24883 117 \n",
|
| 355 |
+
"2 101 148 30471 10774 13785 13779 33642 14399 48271 \n",
|
| 356 |
+
"3 101 19319 16724 10118 10107 78323 12407 38959 22934 \n",
|
| 357 |
+
"4 101 30932 58706 58054 44907 10224 106583 10288 12524 \n",
|
| 358 |
+
"... ... ... ... ... ... ... ... ... ... \n",
|
| 359 |
+
"43344 101 20065 10161 115 115 103784 10774 21388 10245 \n",
|
| 360 |
+
"43345 101 139 80839 24109 13406 18985 16285 10163 11062 \n",
|
| 361 |
+
"43346 101 105549 102635 10140 26943 11499 110516 21899 11861 \n",
|
| 362 |
+
"43347 101 81424 26398 92017 109620 10941 76010 10115 19830 \n",
|
| 363 |
+
"43348 101 39774 11127 45989 24596 11933 170 17145 10710 \n",
|
| 364 |
+
"\n",
|
| 365 |
+
" 9 ... 56 57 58 59 60 61 62 63 labels \n",
|
| 366 |
+
"0 31623 ... 0 0 0 0 0 0 0 0 0 \\\n",
|
| 367 |
+
"1 23763 ... 0 0 0 0 0 0 0 0 0 \n",
|
| 368 |
+
"2 76686 ... 0 0 0 0 0 0 0 0 0 \n",
|
| 369 |
+
"3 10147 ... 0 0 0 0 0 0 0 0 0 \n",
|
| 370 |
+
"4 13878 ... 0 0 0 0 0 0 0 0 0 \n",
|
| 371 |
+
"... ... ... .. .. .. .. .. .. .. .. ... \n",
|
| 372 |
+
"43344 92067 ... 0 0 0 0 0 0 0 0 1 \n",
|
| 373 |
+
"43345 276 ... 0 0 0 0 0 0 0 0 1 \n",
|
| 374 |
+
"43346 10561 ... 0 0 0 0 0 0 0 0 1 \n",
|
| 375 |
+
"43347 26083 ... 0 0 0 0 0 0 0 0 1 \n",
|
| 376 |
+
"43348 39125 ... 0 0 0 0 0 0 0 0 1 \n",
|
| 377 |
+
"\n",
|
| 378 |
+
" tweet \n",
|
| 379 |
+
"0 en güzel uyuyan insan ödülü jeon jungkook'a g... \n",
|
| 380 |
+
"1 Mekanı cennet olsun, saygılar sayın avukatımı... \n",
|
| 381 |
+
"2 Kızlar aranızda kas yığını beylere düşenler ol... \n",
|
| 382 |
+
"3 Biraz ders çalışayım. Tembellik ve uyku düşman... \n",
|
| 383 |
+
"4 Trezeguet yerine El Sharawy daha iyi olmaz mı \n",
|
| 384 |
+
"... ... \n",
|
| 385 |
+
"43344 Hil**adamlar kesinlikle kelimeleri anlamıyorla... \n",
|
| 386 |
+
"43345 Böyle piçlerin çok erken ölmemelerini ve çok f... \n",
|
| 387 |
+
"43346 Turgay denilen bu holigonda bir sorun yok, gur... \n",
|
| 388 |
+
"43347 Umarım ülkenin düşük zekadan kurtulması ilgile... \n",
|
| 389 |
+
"43348 CHP sandıkları bırakmaz, üzerine oturur, bir c... \n",
|
| 390 |
+
"\n",
|
| 391 |
+
"[43349 rows x 66 columns]"
|
| 392 |
+
]
|
| 393 |
+
},
|
| 394 |
+
"execution_count": 3,
|
| 395 |
+
"metadata": {},
|
| 396 |
+
"output_type": "execute_result"
|
| 397 |
+
}
|
| 398 |
+
],
|
| 399 |
+
"source": [
|
| 400 |
+
"copy_df"
|
| 401 |
+
]
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"cell_type": "code",
|
| 405 |
+
"execution_count": 4,
|
| 406 |
+
"metadata": {},
|
| 407 |
+
"outputs": [
|
| 408 |
+
{
|
| 409 |
+
"data": {
|
| 410 |
+
"text/html": [
|
| 411 |
+
"<div>\n",
|
| 412 |
+
"<style scoped>\n",
|
| 413 |
+
" .dataframe tbody tr th:only-of-type {\n",
|
| 414 |
+
" vertical-align: middle;\n",
|
| 415 |
+
" }\n",
|
| 416 |
+
"\n",
|
| 417 |
+
" .dataframe tbody tr th {\n",
|
| 418 |
+
" vertical-align: top;\n",
|
| 419 |
+
" }\n",
|
| 420 |
+
"\n",
|
| 421 |
+
" .dataframe thead th {\n",
|
| 422 |
+
" text-align: right;\n",
|
| 423 |
+
" }\n",
|
| 424 |
+
"</style>\n",
|
| 425 |
+
"<table border=\"1\" class=\"dataframe\">\n",
|
| 426 |
+
" <thead>\n",
|
| 427 |
+
" <tr style=\"text-align: right;\">\n",
|
| 428 |
+
" <th></th>\n",
|
| 429 |
+
" <th>0</th>\n",
|
| 430 |
+
" <th>1</th>\n",
|
| 431 |
+
" <th>2</th>\n",
|
| 432 |
+
" <th>3</th>\n",
|
| 433 |
+
" <th>4</th>\n",
|
| 434 |
+
" <th>5</th>\n",
|
| 435 |
+
" <th>6</th>\n",
|
| 436 |
+
" <th>7</th>\n",
|
| 437 |
+
" <th>8</th>\n",
|
| 438 |
+
" <th>9</th>\n",
|
| 439 |
+
" <th>...</th>\n",
|
| 440 |
+
" <th>56</th>\n",
|
| 441 |
+
" <th>57</th>\n",
|
| 442 |
+
" <th>58</th>\n",
|
| 443 |
+
" <th>59</th>\n",
|
| 444 |
+
" <th>60</th>\n",
|
| 445 |
+
" <th>61</th>\n",
|
| 446 |
+
" <th>62</th>\n",
|
| 447 |
+
" <th>63</th>\n",
|
| 448 |
+
" <th>labels</th>\n",
|
| 449 |
+
" <th>tweet</th>\n",
|
| 450 |
+
" </tr>\n",
|
| 451 |
+
" </thead>\n",
|
| 452 |
+
" <tbody>\n",
|
| 453 |
+
" <tr>\n",
|
| 454 |
+
" <th>0</th>\n",
|
| 455 |
+
" <td>101</td>\n",
|
| 456 |
+
" <td>10110</td>\n",
|
| 457 |
+
" <td>175</td>\n",
|
| 458 |
+
" <td>78653</td>\n",
|
| 459 |
+
" <td>189</td>\n",
|
| 460 |
+
" <td>25285</td>\n",
|
| 461 |
+
" <td>15976</td>\n",
|
| 462 |
+
" <td>40840</td>\n",
|
| 463 |
+
" <td>276</td>\n",
|
| 464 |
+
" <td>31623</td>\n",
|
| 465 |
+
" <td>...</td>\n",
|
| 466 |
+
" <td>0</td>\n",
|
| 467 |
+
" <td>0</td>\n",
|
| 468 |
+
" <td>0</td>\n",
|
| 469 |
+
" <td>0</td>\n",
|
| 470 |
+
" <td>0</td>\n",
|
| 471 |
+
" <td>0</td>\n",
|
| 472 |
+
" <td>0</td>\n",
|
| 473 |
+
" <td>0</td>\n",
|
| 474 |
+
" <td>0</td>\n",
|
| 475 |
+
" <td>en güzel uyuyan insan ödülü jeon jungkook'a g...</td>\n",
|
| 476 |
+
" </tr>\n",
|
| 477 |
+
" <tr>\n",
|
| 478 |
+
" <th>1</th>\n",
|
| 479 |
+
" <td>101</td>\n",
|
| 480 |
+
" <td>11589</td>\n",
|
| 481 |
+
" <td>10706</td>\n",
|
| 482 |
+
" <td>10713</td>\n",
|
| 483 |
+
" <td>10794</td>\n",
|
| 484 |
+
" <td>94698</td>\n",
|
| 485 |
+
" <td>30668</td>\n",
|
| 486 |
+
" <td>24883</td>\n",
|
| 487 |
+
" <td>117</td>\n",
|
| 488 |
+
" <td>23763</td>\n",
|
| 489 |
+
" <td>...</td>\n",
|
| 490 |
+
" <td>0</td>\n",
|
| 491 |
+
" <td>0</td>\n",
|
| 492 |
+
" <td>0</td>\n",
|
| 493 |
+
" <td>0</td>\n",
|
| 494 |
+
" <td>0</td>\n",
|
| 495 |
+
" <td>0</td>\n",
|
| 496 |
+
" <td>0</td>\n",
|
| 497 |
+
" <td>0</td>\n",
|
| 498 |
+
" <td>0</td>\n",
|
| 499 |
+
" <td>Mekanı cennet olsun, saygılar sayın avukatımı...</td>\n",
|
| 500 |
+
" </tr>\n",
|
| 501 |
+
" <tr>\n",
|
| 502 |
+
" <th>2</th>\n",
|
| 503 |
+
" <td>101</td>\n",
|
| 504 |
+
" <td>148</td>\n",
|
| 505 |
+
" <td>30471</td>\n",
|
| 506 |
+
" <td>10774</td>\n",
|
| 507 |
+
" <td>13785</td>\n",
|
| 508 |
+
" <td>13779</td>\n",
|
| 509 |
+
" <td>33642</td>\n",
|
| 510 |
+
" <td>14399</td>\n",
|
| 511 |
+
" <td>48271</td>\n",
|
| 512 |
+
" <td>76686</td>\n",
|
| 513 |
+
" <td>...</td>\n",
|
| 514 |
+
" <td>0</td>\n",
|
| 515 |
+
" <td>0</td>\n",
|
| 516 |
+
" <td>0</td>\n",
|
| 517 |
+
" <td>0</td>\n",
|
| 518 |
+
" <td>0</td>\n",
|
| 519 |
+
" <td>0</td>\n",
|
| 520 |
+
" <td>0</td>\n",
|
| 521 |
+
" <td>0</td>\n",
|
| 522 |
+
" <td>0</td>\n",
|
| 523 |
+
" <td>Kızlar aranızda kas yığını beylere düşenler ol...</td>\n",
|
| 524 |
+
" </tr>\n",
|
| 525 |
+
" <tr>\n",
|
| 526 |
+
" <th>3</th>\n",
|
| 527 |
+
" <td>101</td>\n",
|
| 528 |
+
" <td>19319</td>\n",
|
| 529 |
+
" <td>16724</td>\n",
|
| 530 |
+
" <td>10118</td>\n",
|
| 531 |
+
" <td>10107</td>\n",
|
| 532 |
+
" <td>78323</td>\n",
|
| 533 |
+
" <td>12407</td>\n",
|
| 534 |
+
" <td>38959</td>\n",
|
| 535 |
+
" <td>22934</td>\n",
|
| 536 |
+
" <td>10147</td>\n",
|
| 537 |
+
" <td>...</td>\n",
|
| 538 |
+
" <td>0</td>\n",
|
| 539 |
+
" <td>0</td>\n",
|
| 540 |
+
" <td>0</td>\n",
|
| 541 |
+
" <td>0</td>\n",
|
| 542 |
+
" <td>0</td>\n",
|
| 543 |
+
" <td>0</td>\n",
|
| 544 |
+
" <td>0</td>\n",
|
| 545 |
+
" <td>0</td>\n",
|
| 546 |
+
" <td>0</td>\n",
|
| 547 |
+
" <td>Biraz ders çalışayım. Tembellik ve uyku düşman...</td>\n",
|
| 548 |
+
" </tr>\n",
|
| 549 |
+
" <tr>\n",
|
| 550 |
+
" <th>4</th>\n",
|
| 551 |
+
" <td>101</td>\n",
|
| 552 |
+
" <td>30932</td>\n",
|
| 553 |
+
" <td>58706</td>\n",
|
| 554 |
+
" <td>58054</td>\n",
|
| 555 |
+
" <td>44907</td>\n",
|
| 556 |
+
" <td>10224</td>\n",
|
| 557 |
+
" <td>106583</td>\n",
|
| 558 |
+
" <td>10288</td>\n",
|
| 559 |
+
" <td>12524</td>\n",
|
| 560 |
+
" <td>13878</td>\n",
|
| 561 |
+
" <td>...</td>\n",
|
| 562 |
+
" <td>0</td>\n",
|
| 563 |
+
" <td>0</td>\n",
|
| 564 |
+
" <td>0</td>\n",
|
| 565 |
+
" <td>0</td>\n",
|
| 566 |
+
" <td>0</td>\n",
|
| 567 |
+
" <td>0</td>\n",
|
| 568 |
+
" <td>0</td>\n",
|
| 569 |
+
" <td>0</td>\n",
|
| 570 |
+
" <td>0</td>\n",
|
| 571 |
+
" <td>Trezeguet yerine El Sharawy daha iyi olmaz mı</td>\n",
|
| 572 |
+
" </tr>\n",
|
| 573 |
+
" <tr>\n",
|
| 574 |
+
" <th>...</th>\n",
|
| 575 |
+
" <td>...</td>\n",
|
| 576 |
+
" <td>...</td>\n",
|
| 577 |
+
" <td>...</td>\n",
|
| 578 |
+
" <td>...</td>\n",
|
| 579 |
+
" <td>...</td>\n",
|
| 580 |
+
" <td>...</td>\n",
|
| 581 |
+
" <td>...</td>\n",
|
| 582 |
+
" <td>...</td>\n",
|
| 583 |
+
" <td>...</td>\n",
|
| 584 |
+
" <td>...</td>\n",
|
| 585 |
+
" <td>...</td>\n",
|
| 586 |
+
" <td>...</td>\n",
|
| 587 |
+
" <td>...</td>\n",
|
| 588 |
+
" <td>...</td>\n",
|
| 589 |
+
" <td>...</td>\n",
|
| 590 |
+
" <td>...</td>\n",
|
| 591 |
+
" <td>...</td>\n",
|
| 592 |
+
" <td>...</td>\n",
|
| 593 |
+
" <td>...</td>\n",
|
| 594 |
+
" <td>...</td>\n",
|
| 595 |
+
" <td>...</td>\n",
|
| 596 |
+
" </tr>\n",
|
| 597 |
+
" <tr>\n",
|
| 598 |
+
" <th>43344</th>\n",
|
| 599 |
+
" <td>101</td>\n",
|
| 600 |
+
" <td>20065</td>\n",
|
| 601 |
+
" <td>10161</td>\n",
|
| 602 |
+
" <td>115</td>\n",
|
| 603 |
+
" <td>115</td>\n",
|
| 604 |
+
" <td>103784</td>\n",
|
| 605 |
+
" <td>10774</td>\n",
|
| 606 |
+
" <td>21388</td>\n",
|
| 607 |
+
" <td>10245</td>\n",
|
| 608 |
+
" <td>92067</td>\n",
|
| 609 |
+
" <td>...</td>\n",
|
| 610 |
+
" <td>0</td>\n",
|
| 611 |
+
" <td>0</td>\n",
|
| 612 |
+
" <td>0</td>\n",
|
| 613 |
+
" <td>0</td>\n",
|
| 614 |
+
" <td>0</td>\n",
|
| 615 |
+
" <td>0</td>\n",
|
| 616 |
+
" <td>0</td>\n",
|
| 617 |
+
" <td>0</td>\n",
|
| 618 |
+
" <td>1</td>\n",
|
| 619 |
+
" <td>Hil**adamlar kesinlikle kelimeleri anlamıyorla...</td>\n",
|
| 620 |
+
" </tr>\n",
|
| 621 |
+
" <tr>\n",
|
| 622 |
+
" <th>43345</th>\n",
|
| 623 |
+
" <td>101</td>\n",
|
| 624 |
+
" <td>139</td>\n",
|
| 625 |
+
" <td>80839</td>\n",
|
| 626 |
+
" <td>24109</td>\n",
|
| 627 |
+
" <td>13406</td>\n",
|
| 628 |
+
" <td>18985</td>\n",
|
| 629 |
+
" <td>16285</td>\n",
|
| 630 |
+
" <td>10163</td>\n",
|
| 631 |
+
" <td>11062</td>\n",
|
| 632 |
+
" <td>276</td>\n",
|
| 633 |
+
" <td>...</td>\n",
|
| 634 |
+
" <td>0</td>\n",
|
| 635 |
+
" <td>0</td>\n",
|
| 636 |
+
" <td>0</td>\n",
|
| 637 |
+
" <td>0</td>\n",
|
| 638 |
+
" <td>0</td>\n",
|
| 639 |
+
" <td>0</td>\n",
|
| 640 |
+
" <td>0</td>\n",
|
| 641 |
+
" <td>0</td>\n",
|
| 642 |
+
" <td>1</td>\n",
|
| 643 |
+
" <td>Böyle piçlerin çok erken ölmemelerini ve çok f...</td>\n",
|
| 644 |
+
" </tr>\n",
|
| 645 |
+
" <tr>\n",
|
| 646 |
+
" <th>43346</th>\n",
|
| 647 |
+
" <td>101</td>\n",
|
| 648 |
+
" <td>105549</td>\n",
|
| 649 |
+
" <td>102635</td>\n",
|
| 650 |
+
" <td>10140</td>\n",
|
| 651 |
+
" <td>26943</td>\n",
|
| 652 |
+
" <td>11499</td>\n",
|
| 653 |
+
" <td>110516</td>\n",
|
| 654 |
+
" <td>21899</td>\n",
|
| 655 |
+
" <td>11861</td>\n",
|
| 656 |
+
" <td>10561</td>\n",
|
| 657 |
+
" <td>...</td>\n",
|
| 658 |
+
" <td>0</td>\n",
|
| 659 |
+
" <td>0</td>\n",
|
| 660 |
+
" <td>0</td>\n",
|
| 661 |
+
" <td>0</td>\n",
|
| 662 |
+
" <td>0</td>\n",
|
| 663 |
+
" <td>0</td>\n",
|
| 664 |
+
" <td>0</td>\n",
|
| 665 |
+
" <td>0</td>\n",
|
| 666 |
+
" <td>1</td>\n",
|
| 667 |
+
" <td>Turgay denilen bu holigonda bir sorun yok, gur...</td>\n",
|
| 668 |
+
" </tr>\n",
|
| 669 |
+
" <tr>\n",
|
| 670 |
+
" <th>43347</th>\n",
|
| 671 |
+
" <td>101</td>\n",
|
| 672 |
+
" <td>81424</td>\n",
|
| 673 |
+
" <td>26398</td>\n",
|
| 674 |
+
" <td>92017</td>\n",
|
| 675 |
+
" <td>109620</td>\n",
|
| 676 |
+
" <td>10941</td>\n",
|
| 677 |
+
" <td>76010</td>\n",
|
| 678 |
+
" <td>10115</td>\n",
|
| 679 |
+
" <td>19830</td>\n",
|
| 680 |
+
" <td>26083</td>\n",
|
| 681 |
+
" <td>...</td>\n",
|
| 682 |
+
" <td>0</td>\n",
|
| 683 |
+
" <td>0</td>\n",
|
| 684 |
+
" <td>0</td>\n",
|
| 685 |
+
" <td>0</td>\n",
|
| 686 |
+
" <td>0</td>\n",
|
| 687 |
+
" <td>0</td>\n",
|
| 688 |
+
" <td>0</td>\n",
|
| 689 |
+
" <td>0</td>\n",
|
| 690 |
+
" <td>1</td>\n",
|
| 691 |
+
" <td>Umarım ülkenin düşük zekadan kurtulması ilgile...</td>\n",
|
| 692 |
+
" </tr>\n",
|
| 693 |
+
" <tr>\n",
|
| 694 |
+
" <th>43348</th>\n",
|
| 695 |
+
" <td>101</td>\n",
|
| 696 |
+
" <td>39774</td>\n",
|
| 697 |
+
" <td>11127</td>\n",
|
| 698 |
+
" <td>45989</td>\n",
|
| 699 |
+
" <td>24596</td>\n",
|
| 700 |
+
" <td>11933</td>\n",
|
| 701 |
+
" <td>170</td>\n",
|
| 702 |
+
" <td>17145</td>\n",
|
| 703 |
+
" <td>10710</td>\n",
|
| 704 |
+
" <td>39125</td>\n",
|
| 705 |
+
" <td>...</td>\n",
|
| 706 |
+
" <td>0</td>\n",
|
| 707 |
+
" <td>0</td>\n",
|
| 708 |
+
" <td>0</td>\n",
|
| 709 |
+
" <td>0</td>\n",
|
| 710 |
+
" <td>0</td>\n",
|
| 711 |
+
" <td>0</td>\n",
|
| 712 |
+
" <td>0</td>\n",
|
| 713 |
+
" <td>0</td>\n",
|
| 714 |
+
" <td>1</td>\n",
|
| 715 |
+
" <td>CHP sandıkları bırakmaz, üzerine oturur, bir c...</td>\n",
|
| 716 |
+
" </tr>\n",
|
| 717 |
+
" </tbody>\n",
|
| 718 |
+
"</table>\n",
|
| 719 |
+
"<p>43349 rows × 66 columns</p>\n",
|
| 720 |
+
"</div>"
|
| 721 |
+
],
|
| 722 |
+
"text/plain": [
|
| 723 |
+
" 0 1 2 3 4 5 6 7 8 \n",
|
| 724 |
+
"0 101 10110 175 78653 189 25285 15976 40840 276 \\\n",
|
| 725 |
+
"1 101 11589 10706 10713 10794 94698 30668 24883 117 \n",
|
| 726 |
+
"2 101 148 30471 10774 13785 13779 33642 14399 48271 \n",
|
| 727 |
+
"3 101 19319 16724 10118 10107 78323 12407 38959 22934 \n",
|
| 728 |
+
"4 101 30932 58706 58054 44907 10224 106583 10288 12524 \n",
|
| 729 |
+
"... ... ... ... ... ... ... ... ... ... \n",
|
| 730 |
+
"43344 101 20065 10161 115 115 103784 10774 21388 10245 \n",
|
| 731 |
+
"43345 101 139 80839 24109 13406 18985 16285 10163 11062 \n",
|
| 732 |
+
"43346 101 105549 102635 10140 26943 11499 110516 21899 11861 \n",
|
| 733 |
+
"43347 101 81424 26398 92017 109620 10941 76010 10115 19830 \n",
|
| 734 |
+
"43348 101 39774 11127 45989 24596 11933 170 17145 10710 \n",
|
| 735 |
+
"\n",
|
| 736 |
+
" 9 ... 56 57 58 59 60 61 62 63 labels \n",
|
| 737 |
+
"0 31623 ... 0 0 0 0 0 0 0 0 0 \\\n",
|
| 738 |
+
"1 23763 ... 0 0 0 0 0 0 0 0 0 \n",
|
| 739 |
+
"2 76686 ... 0 0 0 0 0 0 0 0 0 \n",
|
| 740 |
+
"3 10147 ... 0 0 0 0 0 0 0 0 0 \n",
|
| 741 |
+
"4 13878 ... 0 0 0 0 0 0 0 0 0 \n",
|
| 742 |
+
"... ... ... .. .. .. .. .. .. .. .. ... \n",
|
| 743 |
+
"43344 92067 ... 0 0 0 0 0 0 0 0 1 \n",
|
| 744 |
+
"43345 276 ... 0 0 0 0 0 0 0 0 1 \n",
|
| 745 |
+
"43346 10561 ... 0 0 0 0 0 0 0 0 1 \n",
|
| 746 |
+
"43347 26083 ... 0 0 0 0 0 0 0 0 1 \n",
|
| 747 |
+
"43348 39125 ... 0 0 0 0 0 0 0 0 1 \n",
|
| 748 |
+
"\n",
|
| 749 |
+
" tweet \n",
|
| 750 |
+
"0 en güzel uyuyan insan ödülü jeon jungkook'a g... \n",
|
| 751 |
+
"1 Mekanı cennet olsun, saygılar sayın avukatımı... \n",
|
| 752 |
+
"2 Kızlar aranızda kas yığını beylere düşenler ol... \n",
|
| 753 |
+
"3 Biraz ders çalışayım. Tembellik ve uyku düşman... \n",
|
| 754 |
+
"4 Trezeguet yerine El Sharawy daha iyi olmaz mı \n",
|
| 755 |
+
"... ... \n",
|
| 756 |
+
"43344 Hil**adamlar kesinlikle kelimeleri anlamıyorla... \n",
|
| 757 |
+
"43345 Böyle piçlerin çok erken ölmemelerini ve çok f... \n",
|
| 758 |
+
"43346 Turgay denilen bu holigonda bir sorun yok, gur... \n",
|
| 759 |
+
"43347 Umarım ülkenin düşük zekadan kurtulması ilgile... \n",
|
| 760 |
+
"43348 CHP sandıkları bırakmaz, üzerine oturur, bir c... \n",
|
| 761 |
+
"\n",
|
| 762 |
+
"[43349 rows x 66 columns]"
|
| 763 |
+
]
|
| 764 |
+
},
|
| 765 |
+
"execution_count": 4,
|
| 766 |
+
"metadata": {},
|
| 767 |
+
"output_type": "execute_result"
|
| 768 |
+
}
|
| 769 |
+
],
|
| 770 |
+
"source": [
|
| 771 |
+
"embeding_df"
|
| 772 |
+
]
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"cell_type": "code",
|
| 776 |
+
"execution_count": 5,
|
| 777 |
+
"metadata": {},
|
| 778 |
+
"outputs": [],
|
| 779 |
+
"source": [
|
| 780 |
+
"data = embeding_df.tweet.values"
|
| 781 |
+
]
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"cell_type": "code",
|
| 785 |
+
"execution_count": 6,
|
| 786 |
+
"metadata": {},
|
| 787 |
+
"outputs": [],
|
| 788 |
+
"source": [
|
| 789 |
+
"embeding_df=embeding_df.drop(['tweet'], axis=1)\n",
|
| 790 |
+
"copy_df=copy_df.drop(['tweet'], axis=1)"
|
| 791 |
+
]
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"attachments": {},
|
| 795 |
+
"cell_type": "markdown",
|
| 796 |
+
"metadata": {},
|
| 797 |
+
"source": [
|
| 798 |
+
"# isolation forest"
|
| 799 |
+
]
|
| 800 |
+
},
|
| 801 |
+
{
|
| 802 |
+
"cell_type": "code",
|
| 803 |
+
"execution_count": 7,
|
| 804 |
+
"metadata": {},
|
| 805 |
+
"outputs": [],
|
| 806 |
+
"source": [
|
| 807 |
+
"from sklearn.ensemble import IsolationForest"
|
| 808 |
+
]
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"cell_type": "code",
|
| 812 |
+
"execution_count": 8,
|
| 813 |
+
"metadata": {},
|
| 814 |
+
"outputs": [
|
| 815 |
+
{
|
| 816 |
+
"name": "stderr",
|
| 817 |
+
"output_type": "stream",
|
| 818 |
+
"text": [
|
| 819 |
+
"/home/sebit/anaconda3/envs/dl_env/lib/python3.9/site-packages/sklearn/base.py:439: UserWarning: X does not have valid feature names, but IsolationForest was fitted with feature names\n",
|
| 820 |
+
" warnings.warn(\n"
|
| 821 |
+
]
|
| 822 |
+
}
|
| 823 |
+
],
|
| 824 |
+
"source": [
|
| 825 |
+
"# Train the model\n",
|
| 826 |
+
"isf = IsolationForest(contamination=0.04)\n",
|
| 827 |
+
"isf.fit(embeding_df)\n",
|
| 828 |
+
"# Predictions\n",
|
| 829 |
+
"predictions = isf.predict(embeding_df)"
|
| 830 |
+
]
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"cell_type": "code",
|
| 834 |
+
"execution_count": 9,
|
| 835 |
+
"metadata": {},
|
| 836 |
+
"outputs": [
|
| 837 |
+
{
|
| 838 |
+
"data": {
|
| 839 |
+
"text/html": [
|
| 840 |
+
"<div>\n",
|
| 841 |
+
"<style scoped>\n",
|
| 842 |
+
" .dataframe tbody tr th:only-of-type {\n",
|
| 843 |
+
" vertical-align: middle;\n",
|
| 844 |
+
" }\n",
|
| 845 |
+
"\n",
|
| 846 |
+
" .dataframe tbody tr th {\n",
|
| 847 |
+
" vertical-align: top;\n",
|
| 848 |
+
" }\n",
|
| 849 |
+
"\n",
|
| 850 |
+
" .dataframe thead th {\n",
|
| 851 |
+
" text-align: right;\n",
|
| 852 |
+
" }\n",
|
| 853 |
+
"</style>\n",
|
| 854 |
+
"<table border=\"1\" class=\"dataframe\">\n",
|
| 855 |
+
" <thead>\n",
|
| 856 |
+
" <tr style=\"text-align: right;\">\n",
|
| 857 |
+
" <th></th>\n",
|
| 858 |
+
" <th>0</th>\n",
|
| 859 |
+
" <th>1</th>\n",
|
| 860 |
+
" <th>2</th>\n",
|
| 861 |
+
" <th>3</th>\n",
|
| 862 |
+
" <th>4</th>\n",
|
| 863 |
+
" <th>5</th>\n",
|
| 864 |
+
" <th>6</th>\n",
|
| 865 |
+
" <th>7</th>\n",
|
| 866 |
+
" <th>8</th>\n",
|
| 867 |
+
" <th>9</th>\n",
|
| 868 |
+
" <th>...</th>\n",
|
| 869 |
+
" <th>57</th>\n",
|
| 870 |
+
" <th>58</th>\n",
|
| 871 |
+
" <th>59</th>\n",
|
| 872 |
+
" <th>60</th>\n",
|
| 873 |
+
" <th>61</th>\n",
|
| 874 |
+
" <th>62</th>\n",
|
| 875 |
+
" <th>63</th>\n",
|
| 876 |
+
" <th>labels</th>\n",
|
| 877 |
+
" <th>iso_forest_scores</th>\n",
|
| 878 |
+
" <th>iso_forest_outliers</th>\n",
|
| 879 |
+
" </tr>\n",
|
| 880 |
+
" </thead>\n",
|
| 881 |
+
" <tbody>\n",
|
| 882 |
+
" <tr>\n",
|
| 883 |
+
" <th>count</th>\n",
|
| 884 |
+
" <td>43349.0</td>\n",
|
| 885 |
+
" <td>43349.000000</td>\n",
|
| 886 |
+
" <td>43349.000000</td>\n",
|
| 887 |
+
" <td>43349.000000</td>\n",
|
| 888 |
+
" <td>43349.000000</td>\n",
|
| 889 |
+
" <td>43349.000000</td>\n",
|
| 890 |
+
" <td>43349.000000</td>\n",
|
| 891 |
+
" <td>43349.000000</td>\n",
|
| 892 |
+
" <td>43349.000000</td>\n",
|
| 893 |
+
" <td>43349.000000</td>\n",
|
| 894 |
+
" <td>...</td>\n",
|
| 895 |
+
" <td>43349.000000</td>\n",
|
| 896 |
+
" <td>43349.000000</td>\n",
|
| 897 |
+
" <td>43349.00000</td>\n",
|
| 898 |
+
" <td>43349.000000</td>\n",
|
| 899 |
+
" <td>43349.000000</td>\n",
|
| 900 |
+
" <td>43349.000000</td>\n",
|
| 901 |
+
" <td>43349.000000</td>\n",
|
| 902 |
+
" <td>43349.000000</td>\n",
|
| 903 |
+
" <td>43349.000000</td>\n",
|
| 904 |
+
" <td>43349.000000</td>\n",
|
| 905 |
+
" </tr>\n",
|
| 906 |
+
" <tr>\n",
|
| 907 |
+
" <th>mean</th>\n",
|
| 908 |
+
" <td>101.0</td>\n",
|
| 909 |
+
" <td>27403.389559</td>\n",
|
| 910 |
+
" <td>29588.353803</td>\n",
|
| 911 |
+
" <td>26720.445131</td>\n",
|
| 912 |
+
" <td>27755.110106</td>\n",
|
| 913 |
+
" <td>27346.753628</td>\n",
|
| 914 |
+
" <td>27713.189255</td>\n",
|
| 915 |
+
" <td>27295.717687</td>\n",
|
| 916 |
+
" <td>27136.227410</td>\n",
|
| 917 |
+
" <td>26812.611156</td>\n",
|
| 918 |
+
" <td>...</td>\n",
|
| 919 |
+
" <td>4868.917184</td>\n",
|
| 920 |
+
" <td>4813.145309</td>\n",
|
| 921 |
+
" <td>4733.38919</td>\n",
|
| 922 |
+
" <td>4389.068375</td>\n",
|
| 923 |
+
" <td>4297.575723</td>\n",
|
| 924 |
+
" <td>4176.437080</td>\n",
|
| 925 |
+
" <td>17.000392</td>\n",
|
| 926 |
+
" <td>0.417957</td>\n",
|
| 927 |
+
" <td>0.135546</td>\n",
|
| 928 |
+
" <td>0.919998</td>\n",
|
| 929 |
+
" </tr>\n",
|
| 930 |
+
" <tr>\n",
|
| 931 |
+
" <th>std</th>\n",
|
| 932 |
+
" <td>0.0</td>\n",
|
| 933 |
+
" <td>27382.274693</td>\n",
|
| 934 |
+
" <td>27727.688965</td>\n",
|
| 935 |
+
" <td>26455.267691</td>\n",
|
| 936 |
+
" <td>27026.611068</td>\n",
|
| 937 |
+
" <td>26799.753823</td>\n",
|
| 938 |
+
" <td>27021.950023</td>\n",
|
| 939 |
+
" <td>26761.847936</td>\n",
|
| 940 |
+
" <td>26820.810219</td>\n",
|
| 941 |
+
" <td>26720.480625</td>\n",
|
| 942 |
+
" <td>...</td>\n",
|
| 943 |
+
" <td>15312.358275</td>\n",
|
| 944 |
+
" <td>15491.136511</td>\n",
|
| 945 |
+
" <td>15387.09038</td>\n",
|
| 946 |
+
" <td>14617.253040</td>\n",
|
| 947 |
+
" <td>14643.580886</td>\n",
|
| 948 |
+
" <td>14405.397208</td>\n",
|
| 949 |
+
" <td>38.013945</td>\n",
|
| 950 |
+
" <td>0.493229</td>\n",
|
| 951 |
+
" <td>0.066701</td>\n",
|
| 952 |
+
" <td>0.391927</td>\n",
|
| 953 |
+
" </tr>\n",
|
| 954 |
+
" <tr>\n",
|
| 955 |
+
" <th>min</th>\n",
|
| 956 |
+
" <td>101.0</td>\n",
|
| 957 |
+
" <td>100.000000</td>\n",
|
| 958 |
+
" <td>100.000000</td>\n",
|
| 959 |
+
" <td>0.000000</td>\n",
|
| 960 |
+
" <td>0.000000</td>\n",
|
| 961 |
+
" <td>0.000000</td>\n",
|
| 962 |
+
" <td>0.000000</td>\n",
|
| 963 |
+
" <td>0.000000</td>\n",
|
| 964 |
+
" <td>0.000000</td>\n",
|
| 965 |
+
" <td>0.000000</td>\n",
|
| 966 |
+
" <td>...</td>\n",
|
| 967 |
+
" <td>0.000000</td>\n",
|
| 968 |
+
" <td>0.000000</td>\n",
|
| 969 |
+
" <td>0.00000</td>\n",
|
| 970 |
+
" <td>0.000000</td>\n",
|
| 971 |
+
" <td>0.000000</td>\n",
|
| 972 |
+
" <td>0.000000</td>\n",
|
| 973 |
+
" <td>0.000000</td>\n",
|
| 974 |
+
" <td>0.000000</td>\n",
|
| 975 |
+
" <td>-0.140643</td>\n",
|
| 976 |
+
" <td>-1.000000</td>\n",
|
| 977 |
+
" </tr>\n",
|
| 978 |
+
" <tr>\n",
|
| 979 |
+
" <th>25%</th>\n",
|
| 980 |
+
" <td>101.0</td>\n",
|
| 981 |
+
" <td>10357.000000</td>\n",
|
| 982 |
+
" <td>10506.000000</td>\n",
|
| 983 |
+
" <td>10323.000000</td>\n",
|
| 984 |
+
" <td>10361.000000</td>\n",
|
| 985 |
+
" <td>10350.000000</td>\n",
|
| 986 |
+
" <td>10369.000000</td>\n",
|
| 987 |
+
" <td>10347.000000</td>\n",
|
| 988 |
+
" <td>10330.000000</td>\n",
|
| 989 |
+
" <td>10323.000000</td>\n",
|
| 990 |
+
" <td>...</td>\n",
|
| 991 |
+
" <td>0.000000</td>\n",
|
| 992 |
+
" <td>0.000000</td>\n",
|
| 993 |
+
" <td>0.00000</td>\n",
|
| 994 |
+
" <td>0.000000</td>\n",
|
| 995 |
+
" <td>0.000000</td>\n",
|
| 996 |
+
" <td>0.000000</td>\n",
|
| 997 |
+
" <td>0.000000</td>\n",
|
| 998 |
+
" <td>0.000000</td>\n",
|
| 999 |
+
" <td>0.089100</td>\n",
|
| 1000 |
+
" <td>1.000000</td>\n",
|
| 1001 |
+
" </tr>\n",
|
| 1002 |
+
" <tr>\n",
|
| 1003 |
+
" <th>50%</th>\n",
|
| 1004 |
+
" <td>101.0</td>\n",
|
| 1005 |
+
" <td>18856.000000</td>\n",
|
| 1006 |
+
" <td>16263.000000</td>\n",
|
| 1007 |
+
" <td>13587.000000</td>\n",
|
| 1008 |
+
" <td>14918.000000</td>\n",
|
| 1009 |
+
" <td>14753.000000</td>\n",
|
| 1010 |
+
" <td>15090.000000</td>\n",
|
| 1011 |
+
" <td>14777.000000</td>\n",
|
| 1012 |
+
" <td>14753.000000</td>\n",
|
| 1013 |
+
" <td>14110.000000</td>\n",
|
| 1014 |
+
" <td>...</td>\n",
|
| 1015 |
+
" <td>0.000000</td>\n",
|
| 1016 |
+
" <td>0.000000</td>\n",
|
| 1017 |
+
" <td>0.00000</td>\n",
|
| 1018 |
+
" <td>0.000000</td>\n",
|
| 1019 |
+
" <td>0.000000</td>\n",
|
| 1020 |
+
" <td>0.000000</td>\n",
|
| 1021 |
+
" <td>0.000000</td>\n",
|
| 1022 |
+
" <td>0.000000</td>\n",
|
| 1023 |
+
" <td>0.161505</td>\n",
|
| 1024 |
+
" <td>1.000000</td>\n",
|
| 1025 |
+
" </tr>\n",
|
| 1026 |
+
" <tr>\n",
|
| 1027 |
+
" <th>75%</th>\n",
|
| 1028 |
+
" <td>101.0</td>\n",
|
| 1029 |
+
" <td>41079.000000</td>\n",
|
| 1030 |
+
" <td>40762.000000</td>\n",
|
| 1031 |
+
" <td>35943.000000</td>\n",
|
| 1032 |
+
" <td>37820.000000</td>\n",
|
| 1033 |
+
" <td>36544.000000</td>\n",
|
| 1034 |
+
" <td>37820.000000</td>\n",
|
| 1035 |
+
" <td>36723.000000</td>\n",
|
| 1036 |
+
" <td>36544.000000</td>\n",
|
| 1037 |
+
" <td>36445.000000</td>\n",
|
| 1038 |
+
" <td>...</td>\n",
|
| 1039 |
+
" <td>0.000000</td>\n",
|
| 1040 |
+
" <td>0.000000</td>\n",
|
| 1041 |
+
" <td>0.00000</td>\n",
|
| 1042 |
+
" <td>0.000000</td>\n",
|
| 1043 |
+
" <td>0.000000</td>\n",
|
| 1044 |
+
" <td>0.000000</td>\n",
|
| 1045 |
+
" <td>0.000000</td>\n",
|
| 1046 |
+
" <td>1.000000</td>\n",
|
| 1047 |
+
" <td>0.189511</td>\n",
|
| 1048 |
+
" <td>1.000000</td>\n",
|
| 1049 |
+
" </tr>\n",
|
| 1050 |
+
" <tr>\n",
|
| 1051 |
+
" <th>max</th>\n",
|
| 1052 |
+
" <td>101.0</td>\n",
|
| 1053 |
+
" <td>110744.000000</td>\n",
|
| 1054 |
+
" <td>110966.000000</td>\n",
|
| 1055 |
+
" <td>110966.000000</td>\n",
|
| 1056 |
+
" <td>110966.000000</td>\n",
|
| 1057 |
+
" <td>111720.000000</td>\n",
|
| 1058 |
+
" <td>111720.000000</td>\n",
|
| 1059 |
+
" <td>111720.000000</td>\n",
|
| 1060 |
+
" <td>111720.000000</td>\n",
|
| 1061 |
+
" <td>111720.000000</td>\n",
|
| 1062 |
+
" <td>...</td>\n",
|
| 1063 |
+
" <td>110966.000000</td>\n",
|
| 1064 |
+
" <td>110966.000000</td>\n",
|
| 1065 |
+
" <td>110966.00000</td>\n",
|
| 1066 |
+
" <td>110966.000000</td>\n",
|
| 1067 |
+
" <td>110966.000000</td>\n",
|
| 1068 |
+
" <td>110966.000000</td>\n",
|
| 1069 |
+
" <td>102.000000</td>\n",
|
| 1070 |
+
" <td>1.000000</td>\n",
|
| 1071 |
+
" <td>0.216831</td>\n",
|
| 1072 |
+
" <td>1.000000</td>\n",
|
| 1073 |
+
" </tr>\n",
|
| 1074 |
+
" </tbody>\n",
|
| 1075 |
+
"</table>\n",
|
| 1076 |
+
"<p>8 rows × 67 columns</p>\n",
|
| 1077 |
+
"</div>"
|
| 1078 |
+
],
|
| 1079 |
+
"text/plain": [
|
| 1080 |
+
" 0 1 2 3 4 \n",
|
| 1081 |
+
"count 43349.0 43349.000000 43349.000000 43349.000000 43349.000000 \\\n",
|
| 1082 |
+
"mean 101.0 27403.389559 29588.353803 26720.445131 27755.110106 \n",
|
| 1083 |
+
"std 0.0 27382.274693 27727.688965 26455.267691 27026.611068 \n",
|
| 1084 |
+
"min 101.0 100.000000 100.000000 0.000000 0.000000 \n",
|
| 1085 |
+
"25% 101.0 10357.000000 10506.000000 10323.000000 10361.000000 \n",
|
| 1086 |
+
"50% 101.0 18856.000000 16263.000000 13587.000000 14918.000000 \n",
|
| 1087 |
+
"75% 101.0 41079.000000 40762.000000 35943.000000 37820.000000 \n",
|
| 1088 |
+
"max 101.0 110744.000000 110966.000000 110966.000000 110966.000000 \n",
|
| 1089 |
+
"\n",
|
| 1090 |
+
" 5 6 7 8 \n",
|
| 1091 |
+
"count 43349.000000 43349.000000 43349.000000 43349.000000 \\\n",
|
| 1092 |
+
"mean 27346.753628 27713.189255 27295.717687 27136.227410 \n",
|
| 1093 |
+
"std 26799.753823 27021.950023 26761.847936 26820.810219 \n",
|
| 1094 |
+
"min 0.000000 0.000000 0.000000 0.000000 \n",
|
| 1095 |
+
"25% 10350.000000 10369.000000 10347.000000 10330.000000 \n",
|
| 1096 |
+
"50% 14753.000000 15090.000000 14777.000000 14753.000000 \n",
|
| 1097 |
+
"75% 36544.000000 37820.000000 36723.000000 36544.000000 \n",
|
| 1098 |
+
"max 111720.000000 111720.000000 111720.000000 111720.000000 \n",
|
| 1099 |
+
"\n",
|
| 1100 |
+
" 9 ... 57 58 59 \n",
|
| 1101 |
+
"count 43349.000000 ... 43349.000000 43349.000000 43349.00000 \\\n",
|
| 1102 |
+
"mean 26812.611156 ... 4868.917184 4813.145309 4733.38919 \n",
|
| 1103 |
+
"std 26720.480625 ... 15312.358275 15491.136511 15387.09038 \n",
|
| 1104 |
+
"min 0.000000 ... 0.000000 0.000000 0.00000 \n",
|
| 1105 |
+
"25% 10323.000000 ... 0.000000 0.000000 0.00000 \n",
|
| 1106 |
+
"50% 14110.000000 ... 0.000000 0.000000 0.00000 \n",
|
| 1107 |
+
"75% 36445.000000 ... 0.000000 0.000000 0.00000 \n",
|
| 1108 |
+
"max 111720.000000 ... 110966.000000 110966.000000 110966.00000 \n",
|
| 1109 |
+
"\n",
|
| 1110 |
+
" 60 61 62 63 \n",
|
| 1111 |
+
"count 43349.000000 43349.000000 43349.000000 43349.000000 \\\n",
|
| 1112 |
+
"mean 4389.068375 4297.575723 4176.437080 17.000392 \n",
|
| 1113 |
+
"std 14617.253040 14643.580886 14405.397208 38.013945 \n",
|
| 1114 |
+
"min 0.000000 0.000000 0.000000 0.000000 \n",
|
| 1115 |
+
"25% 0.000000 0.000000 0.000000 0.000000 \n",
|
| 1116 |
+
"50% 0.000000 0.000000 0.000000 0.000000 \n",
|
| 1117 |
+
"75% 0.000000 0.000000 0.000000 0.000000 \n",
|
| 1118 |
+
"max 110966.000000 110966.000000 110966.000000 102.000000 \n",
|
| 1119 |
+
"\n",
|
| 1120 |
+
" labels iso_forest_scores iso_forest_outliers \n",
|
| 1121 |
+
"count 43349.000000 43349.000000 43349.000000 \n",
|
| 1122 |
+
"mean 0.417957 0.135546 0.919998 \n",
|
| 1123 |
+
"std 0.493229 0.066701 0.391927 \n",
|
| 1124 |
+
"min 0.000000 -0.140643 -1.000000 \n",
|
| 1125 |
+
"25% 0.000000 0.089100 1.000000 \n",
|
| 1126 |
+
"50% 0.000000 0.161505 1.000000 \n",
|
| 1127 |
+
"75% 1.000000 0.189511 1.000000 \n",
|
| 1128 |
+
"max 1.000000 0.216831 1.000000 \n",
|
| 1129 |
+
"\n",
|
| 1130 |
+
"[8 rows x 67 columns]"
|
| 1131 |
+
]
|
| 1132 |
+
},
|
| 1133 |
+
"execution_count": 9,
|
| 1134 |
+
"metadata": {},
|
| 1135 |
+
"output_type": "execute_result"
|
| 1136 |
+
}
|
| 1137 |
+
],
|
| 1138 |
+
"source": [
|
| 1139 |
+
"# Extract scores\n",
|
| 1140 |
+
"embeding_df[\"iso_forest_scores\"] = isf.decision_function(embeding_df)\n",
|
| 1141 |
+
"# Extract predictions\n",
|
| 1142 |
+
"embeding_df[\"iso_forest_outliers\"] = predictions\n",
|
| 1143 |
+
"# Describe the dataframe\n",
|
| 1144 |
+
"embeding_df.describe()"
|
| 1145 |
+
]
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"cell_type": "code",
|
| 1149 |
+
"execution_count": 10,
|
| 1150 |
+
"metadata": {},
|
| 1151 |
+
"outputs": [
|
| 1152 |
+
{
|
| 1153 |
+
"data": {
|
| 1154 |
+
"text/html": [
|
| 1155 |
+
"<div>\n",
|
| 1156 |
+
"<style scoped>\n",
|
| 1157 |
+
" .dataframe tbody tr th:only-of-type {\n",
|
| 1158 |
+
" vertical-align: middle;\n",
|
| 1159 |
+
" }\n",
|
| 1160 |
+
"\n",
|
| 1161 |
+
" .dataframe tbody tr th {\n",
|
| 1162 |
+
" vertical-align: top;\n",
|
| 1163 |
+
" }\n",
|
| 1164 |
+
"\n",
|
| 1165 |
+
" .dataframe thead th {\n",
|
| 1166 |
+
" text-align: right;\n",
|
| 1167 |
+
" }\n",
|
| 1168 |
+
"</style>\n",
|
| 1169 |
+
"<table border=\"1\" class=\"dataframe\">\n",
|
| 1170 |
+
" <thead>\n",
|
| 1171 |
+
" <tr style=\"text-align: right;\">\n",
|
| 1172 |
+
" <th></th>\n",
|
| 1173 |
+
" <th>0</th>\n",
|
| 1174 |
+
" <th>1</th>\n",
|
| 1175 |
+
" <th>2</th>\n",
|
| 1176 |
+
" <th>3</th>\n",
|
| 1177 |
+
" <th>4</th>\n",
|
| 1178 |
+
" <th>5</th>\n",
|
| 1179 |
+
" <th>6</th>\n",
|
| 1180 |
+
" <th>7</th>\n",
|
| 1181 |
+
" <th>8</th>\n",
|
| 1182 |
+
" <th>9</th>\n",
|
| 1183 |
+
" <th>...</th>\n",
|
| 1184 |
+
" <th>57</th>\n",
|
| 1185 |
+
" <th>58</th>\n",
|
| 1186 |
+
" <th>59</th>\n",
|
| 1187 |
+
" <th>60</th>\n",
|
| 1188 |
+
" <th>61</th>\n",
|
| 1189 |
+
" <th>62</th>\n",
|
| 1190 |
+
" <th>63</th>\n",
|
| 1191 |
+
" <th>labels</th>\n",
|
| 1192 |
+
" <th>iso_forest_scores</th>\n",
|
| 1193 |
+
" <th>iso_forest_outliers</th>\n",
|
| 1194 |
+
" </tr>\n",
|
| 1195 |
+
" </thead>\n",
|
| 1196 |
+
" <tbody>\n",
|
| 1197 |
+
" <tr>\n",
|
| 1198 |
+
" <th>0</th>\n",
|
| 1199 |
+
" <td>101</td>\n",
|
| 1200 |
+
" <td>10110</td>\n",
|
| 1201 |
+
" <td>175</td>\n",
|
| 1202 |
+
" <td>78653</td>\n",
|
| 1203 |
+
" <td>189</td>\n",
|
| 1204 |
+
" <td>25285</td>\n",
|
| 1205 |
+
" <td>15976</td>\n",
|
| 1206 |
+
" <td>40840</td>\n",
|
| 1207 |
+
" <td>276</td>\n",
|
| 1208 |
+
" <td>31623</td>\n",
|
| 1209 |
+
" <td>...</td>\n",
|
| 1210 |
+
" <td>0</td>\n",
|
| 1211 |
+
" <td>0</td>\n",
|
| 1212 |
+
" <td>0</td>\n",
|
| 1213 |
+
" <td>0</td>\n",
|
| 1214 |
+
" <td>0</td>\n",
|
| 1215 |
+
" <td>0</td>\n",
|
| 1216 |
+
" <td>0</td>\n",
|
| 1217 |
+
" <td>0</td>\n",
|
| 1218 |
+
" <td>0.189202</td>\n",
|
| 1219 |
+
" <td>No</td>\n",
|
| 1220 |
+
" </tr>\n",
|
| 1221 |
+
" <tr>\n",
|
| 1222 |
+
" <th>1</th>\n",
|
| 1223 |
+
" <td>101</td>\n",
|
| 1224 |
+
" <td>11589</td>\n",
|
| 1225 |
+
" <td>10706</td>\n",
|
| 1226 |
+
" <td>10713</td>\n",
|
| 1227 |
+
" <td>10794</td>\n",
|
| 1228 |
+
" <td>94698</td>\n",
|
| 1229 |
+
" <td>30668</td>\n",
|
| 1230 |
+
" <td>24883</td>\n",
|
| 1231 |
+
" <td>117</td>\n",
|
| 1232 |
+
" <td>23763</td>\n",
|
| 1233 |
+
" <td>...</td>\n",
|
| 1234 |
+
" <td>0</td>\n",
|
| 1235 |
+
" <td>0</td>\n",
|
| 1236 |
+
" <td>0</td>\n",
|
| 1237 |
+
" <td>0</td>\n",
|
| 1238 |
+
" <td>0</td>\n",
|
| 1239 |
+
" <td>0</td>\n",
|
| 1240 |
+
" <td>0</td>\n",
|
| 1241 |
+
" <td>0</td>\n",
|
| 1242 |
+
" <td>0.181234</td>\n",
|
| 1243 |
+
" <td>No</td>\n",
|
| 1244 |
+
" </tr>\n",
|
| 1245 |
+
" <tr>\n",
|
| 1246 |
+
" <th>2</th>\n",
|
| 1247 |
+
" <td>101</td>\n",
|
| 1248 |
+
" <td>148</td>\n",
|
| 1249 |
+
" <td>30471</td>\n",
|
| 1250 |
+
" <td>10774</td>\n",
|
| 1251 |
+
" <td>13785</td>\n",
|
| 1252 |
+
" <td>13779</td>\n",
|
| 1253 |
+
" <td>33642</td>\n",
|
| 1254 |
+
" <td>14399</td>\n",
|
| 1255 |
+
" <td>48271</td>\n",
|
| 1256 |
+
" <td>76686</td>\n",
|
| 1257 |
+
" <td>...</td>\n",
|
| 1258 |
+
" <td>0</td>\n",
|
| 1259 |
+
" <td>0</td>\n",
|
| 1260 |
+
" <td>0</td>\n",
|
| 1261 |
+
" <td>0</td>\n",
|
| 1262 |
+
" <td>0</td>\n",
|
| 1263 |
+
" <td>0</td>\n",
|
| 1264 |
+
" <td>0</td>\n",
|
| 1265 |
+
" <td>0</td>\n",
|
| 1266 |
+
" <td>0.166332</td>\n",
|
| 1267 |
+
" <td>No</td>\n",
|
| 1268 |
+
" </tr>\n",
|
| 1269 |
+
" <tr>\n",
|
| 1270 |
+
" <th>3</th>\n",
|
| 1271 |
+
" <td>101</td>\n",
|
| 1272 |
+
" <td>19319</td>\n",
|
| 1273 |
+
" <td>16724</td>\n",
|
| 1274 |
+
" <td>10118</td>\n",
|
| 1275 |
+
" <td>10107</td>\n",
|
| 1276 |
+
" <td>78323</td>\n",
|
| 1277 |
+
" <td>12407</td>\n",
|
| 1278 |
+
" <td>38959</td>\n",
|
| 1279 |
+
" <td>22934</td>\n",
|
| 1280 |
+
" <td>10147</td>\n",
|
| 1281 |
+
" <td>...</td>\n",
|
| 1282 |
+
" <td>0</td>\n",
|
| 1283 |
+
" <td>0</td>\n",
|
| 1284 |
+
" <td>0</td>\n",
|
| 1285 |
+
" <td>0</td>\n",
|
| 1286 |
+
" <td>0</td>\n",
|
| 1287 |
+
" <td>0</td>\n",
|
| 1288 |
+
" <td>0</td>\n",
|
| 1289 |
+
" <td>0</td>\n",
|
| 1290 |
+
" <td>0.151816</td>\n",
|
| 1291 |
+
" <td>No</td>\n",
|
| 1292 |
+
" </tr>\n",
|
| 1293 |
+
" <tr>\n",
|
| 1294 |
+
" <th>4</th>\n",
|
| 1295 |
+
" <td>101</td>\n",
|
| 1296 |
+
" <td>30932</td>\n",
|
| 1297 |
+
" <td>58706</td>\n",
|
| 1298 |
+
" <td>58054</td>\n",
|
| 1299 |
+
" <td>44907</td>\n",
|
| 1300 |
+
" <td>10224</td>\n",
|
| 1301 |
+
" <td>106583</td>\n",
|
| 1302 |
+
" <td>10288</td>\n",
|
| 1303 |
+
" <td>12524</td>\n",
|
| 1304 |
+
" <td>13878</td>\n",
|
| 1305 |
+
" <td>...</td>\n",
|
| 1306 |
+
" <td>0</td>\n",
|
| 1307 |
+
" <td>0</td>\n",
|
| 1308 |
+
" <td>0</td>\n",
|
| 1309 |
+
" <td>0</td>\n",
|
| 1310 |
+
" <td>0</td>\n",
|
| 1311 |
+
" <td>0</td>\n",
|
| 1312 |
+
" <td>0</td>\n",
|
| 1313 |
+
" <td>0</td>\n",
|
| 1314 |
+
" <td>0.184008</td>\n",
|
| 1315 |
+
" <td>No</td>\n",
|
| 1316 |
+
" </tr>\n",
|
| 1317 |
+
" <tr>\n",
|
| 1318 |
+
" <th>...</th>\n",
|
| 1319 |
+
" <td>...</td>\n",
|
| 1320 |
+
" <td>...</td>\n",
|
| 1321 |
+
" <td>...</td>\n",
|
| 1322 |
+
" <td>...</td>\n",
|
| 1323 |
+
" <td>...</td>\n",
|
| 1324 |
+
" <td>...</td>\n",
|
| 1325 |
+
" <td>...</td>\n",
|
| 1326 |
+
" <td>...</td>\n",
|
| 1327 |
+
" <td>...</td>\n",
|
| 1328 |
+
" <td>...</td>\n",
|
| 1329 |
+
" <td>...</td>\n",
|
| 1330 |
+
" <td>...</td>\n",
|
| 1331 |
+
" <td>...</td>\n",
|
| 1332 |
+
" <td>...</td>\n",
|
| 1333 |
+
" <td>...</td>\n",
|
| 1334 |
+
" <td>...</td>\n",
|
| 1335 |
+
" <td>...</td>\n",
|
| 1336 |
+
" <td>...</td>\n",
|
| 1337 |
+
" <td>...</td>\n",
|
| 1338 |
+
" <td>...</td>\n",
|
| 1339 |
+
" <td>...</td>\n",
|
| 1340 |
+
" </tr>\n",
|
| 1341 |
+
" <tr>\n",
|
| 1342 |
+
" <th>43344</th>\n",
|
| 1343 |
+
" <td>101</td>\n",
|
| 1344 |
+
" <td>20065</td>\n",
|
| 1345 |
+
" <td>10161</td>\n",
|
| 1346 |
+
" <td>115</td>\n",
|
| 1347 |
+
" <td>115</td>\n",
|
| 1348 |
+
" <td>103784</td>\n",
|
| 1349 |
+
" <td>10774</td>\n",
|
| 1350 |
+
" <td>21388</td>\n",
|
| 1351 |
+
" <td>10245</td>\n",
|
| 1352 |
+
" <td>92067</td>\n",
|
| 1353 |
+
" <td>...</td>\n",
|
| 1354 |
+
" <td>0</td>\n",
|
| 1355 |
+
" <td>0</td>\n",
|
| 1356 |
+
" <td>0</td>\n",
|
| 1357 |
+
" <td>0</td>\n",
|
| 1358 |
+
" <td>0</td>\n",
|
| 1359 |
+
" <td>0</td>\n",
|
| 1360 |
+
" <td>0</td>\n",
|
| 1361 |
+
" <td>1</td>\n",
|
| 1362 |
+
" <td>0.079412</td>\n",
|
| 1363 |
+
" <td>No</td>\n",
|
| 1364 |
+
" </tr>\n",
|
| 1365 |
+
" <tr>\n",
|
| 1366 |
+
" <th>43345</th>\n",
|
| 1367 |
+
" <td>101</td>\n",
|
| 1368 |
+
" <td>139</td>\n",
|
| 1369 |
+
" <td>80839</td>\n",
|
| 1370 |
+
" <td>24109</td>\n",
|
| 1371 |
+
" <td>13406</td>\n",
|
| 1372 |
+
" <td>18985</td>\n",
|
| 1373 |
+
" <td>16285</td>\n",
|
| 1374 |
+
" <td>10163</td>\n",
|
| 1375 |
+
" <td>11062</td>\n",
|
| 1376 |
+
" <td>276</td>\n",
|
| 1377 |
+
" <td>...</td>\n",
|
| 1378 |
+
" <td>0</td>\n",
|
| 1379 |
+
" <td>0</td>\n",
|
| 1380 |
+
" <td>0</td>\n",
|
| 1381 |
+
" <td>0</td>\n",
|
| 1382 |
+
" <td>0</td>\n",
|
| 1383 |
+
" <td>0</td>\n",
|
| 1384 |
+
" <td>0</td>\n",
|
| 1385 |
+
" <td>1</td>\n",
|
| 1386 |
+
" <td>0.118245</td>\n",
|
| 1387 |
+
" <td>No</td>\n",
|
| 1388 |
+
" </tr>\n",
|
| 1389 |
+
" <tr>\n",
|
| 1390 |
+
" <th>43346</th>\n",
|
| 1391 |
+
" <td>101</td>\n",
|
| 1392 |
+
" <td>105549</td>\n",
|
| 1393 |
+
" <td>102635</td>\n",
|
| 1394 |
+
" <td>10140</td>\n",
|
| 1395 |
+
" <td>26943</td>\n",
|
| 1396 |
+
" <td>11499</td>\n",
|
| 1397 |
+
" <td>110516</td>\n",
|
| 1398 |
+
" <td>21899</td>\n",
|
| 1399 |
+
" <td>11861</td>\n",
|
| 1400 |
+
" <td>10561</td>\n",
|
| 1401 |
+
" <td>...</td>\n",
|
| 1402 |
+
" <td>0</td>\n",
|
| 1403 |
+
" <td>0</td>\n",
|
| 1404 |
+
" <td>0</td>\n",
|
| 1405 |
+
" <td>0</td>\n",
|
| 1406 |
+
" <td>0</td>\n",
|
| 1407 |
+
" <td>0</td>\n",
|
| 1408 |
+
" <td>0</td>\n",
|
| 1409 |
+
" <td>1</td>\n",
|
| 1410 |
+
" <td>0.138229</td>\n",
|
| 1411 |
+
" <td>No</td>\n",
|
| 1412 |
+
" </tr>\n",
|
| 1413 |
+
" <tr>\n",
|
| 1414 |
+
" <th>43347</th>\n",
|
| 1415 |
+
" <td>101</td>\n",
|
| 1416 |
+
" <td>81424</td>\n",
|
| 1417 |
+
" <td>26398</td>\n",
|
| 1418 |
+
" <td>92017</td>\n",
|
| 1419 |
+
" <td>109620</td>\n",
|
| 1420 |
+
" <td>10941</td>\n",
|
| 1421 |
+
" <td>76010</td>\n",
|
| 1422 |
+
" <td>10115</td>\n",
|
| 1423 |
+
" <td>19830</td>\n",
|
| 1424 |
+
" <td>26083</td>\n",
|
| 1425 |
+
" <td>...</td>\n",
|
| 1426 |
+
" <td>0</td>\n",
|
| 1427 |
+
" <td>0</td>\n",
|
| 1428 |
+
" <td>0</td>\n",
|
| 1429 |
+
" <td>0</td>\n",
|
| 1430 |
+
" <td>0</td>\n",
|
| 1431 |
+
" <td>0</td>\n",
|
| 1432 |
+
" <td>0</td>\n",
|
| 1433 |
+
" <td>1</td>\n",
|
| 1434 |
+
" <td>0.181065</td>\n",
|
| 1435 |
+
" <td>No</td>\n",
|
| 1436 |
+
" </tr>\n",
|
| 1437 |
+
" <tr>\n",
|
| 1438 |
+
" <th>43348</th>\n",
|
| 1439 |
+
" <td>101</td>\n",
|
| 1440 |
+
" <td>39774</td>\n",
|
| 1441 |
+
" <td>11127</td>\n",
|
| 1442 |
+
" <td>45989</td>\n",
|
| 1443 |
+
" <td>24596</td>\n",
|
| 1444 |
+
" <td>11933</td>\n",
|
| 1445 |
+
" <td>170</td>\n",
|
| 1446 |
+
" <td>17145</td>\n",
|
| 1447 |
+
" <td>10710</td>\n",
|
| 1448 |
+
" <td>39125</td>\n",
|
| 1449 |
+
" <td>...</td>\n",
|
| 1450 |
+
" <td>0</td>\n",
|
| 1451 |
+
" <td>0</td>\n",
|
| 1452 |
+
" <td>0</td>\n",
|
| 1453 |
+
" <td>0</td>\n",
|
| 1454 |
+
" <td>0</td>\n",
|
| 1455 |
+
" <td>0</td>\n",
|
| 1456 |
+
" <td>0</td>\n",
|
| 1457 |
+
" <td>1</td>\n",
|
| 1458 |
+
" <td>0.085161</td>\n",
|
| 1459 |
+
" <td>No</td>\n",
|
| 1460 |
+
" </tr>\n",
|
| 1461 |
+
" </tbody>\n",
|
| 1462 |
+
"</table>\n",
|
| 1463 |
+
"<p>43349 rows × 67 columns</p>\n",
|
| 1464 |
+
"</div>"
|
| 1465 |
+
],
|
| 1466 |
+
"text/plain": [
|
| 1467 |
+
" 0 1 2 3 4 5 6 7 8 \n",
|
| 1468 |
+
"0 101 10110 175 78653 189 25285 15976 40840 276 \\\n",
|
| 1469 |
+
"1 101 11589 10706 10713 10794 94698 30668 24883 117 \n",
|
| 1470 |
+
"2 101 148 30471 10774 13785 13779 33642 14399 48271 \n",
|
| 1471 |
+
"3 101 19319 16724 10118 10107 78323 12407 38959 22934 \n",
|
| 1472 |
+
"4 101 30932 58706 58054 44907 10224 106583 10288 12524 \n",
|
| 1473 |
+
"... ... ... ... ... ... ... ... ... ... \n",
|
| 1474 |
+
"43344 101 20065 10161 115 115 103784 10774 21388 10245 \n",
|
| 1475 |
+
"43345 101 139 80839 24109 13406 18985 16285 10163 11062 \n",
|
| 1476 |
+
"43346 101 105549 102635 10140 26943 11499 110516 21899 11861 \n",
|
| 1477 |
+
"43347 101 81424 26398 92017 109620 10941 76010 10115 19830 \n",
|
| 1478 |
+
"43348 101 39774 11127 45989 24596 11933 170 17145 10710 \n",
|
| 1479 |
+
"\n",
|
| 1480 |
+
" 9 ... 57 58 59 60 61 62 63 labels iso_forest_scores \n",
|
| 1481 |
+
"0 31623 ... 0 0 0 0 0 0 0 0 0.189202 \\\n",
|
| 1482 |
+
"1 23763 ... 0 0 0 0 0 0 0 0 0.181234 \n",
|
| 1483 |
+
"2 76686 ... 0 0 0 0 0 0 0 0 0.166332 \n",
|
| 1484 |
+
"3 10147 ... 0 0 0 0 0 0 0 0 0.151816 \n",
|
| 1485 |
+
"4 13878 ... 0 0 0 0 0 0 0 0 0.184008 \n",
|
| 1486 |
+
"... ... ... .. .. .. .. .. .. .. ... ... \n",
|
| 1487 |
+
"43344 92067 ... 0 0 0 0 0 0 0 1 0.079412 \n",
|
| 1488 |
+
"43345 276 ... 0 0 0 0 0 0 0 1 0.118245 \n",
|
| 1489 |
+
"43346 10561 ... 0 0 0 0 0 0 0 1 0.138229 \n",
|
| 1490 |
+
"43347 26083 ... 0 0 0 0 0 0 0 1 0.181065 \n",
|
| 1491 |
+
"43348 39125 ... 0 0 0 0 0 0 0 1 0.085161 \n",
|
| 1492 |
+
"\n",
|
| 1493 |
+
" iso_forest_outliers \n",
|
| 1494 |
+
"0 No \n",
|
| 1495 |
+
"1 No \n",
|
| 1496 |
+
"2 No \n",
|
| 1497 |
+
"3 No \n",
|
| 1498 |
+
"4 No \n",
|
| 1499 |
+
"... ... \n",
|
| 1500 |
+
"43344 No \n",
|
| 1501 |
+
"43345 No \n",
|
| 1502 |
+
"43346 No \n",
|
| 1503 |
+
"43347 No \n",
|
| 1504 |
+
"43348 No \n",
|
| 1505 |
+
"\n",
|
| 1506 |
+
"[43349 rows x 67 columns]"
|
| 1507 |
+
]
|
| 1508 |
+
},
|
| 1509 |
+
"execution_count": 10,
|
| 1510 |
+
"metadata": {},
|
| 1511 |
+
"output_type": "execute_result"
|
| 1512 |
+
}
|
| 1513 |
+
],
|
| 1514 |
+
"source": [
|
| 1515 |
+
"# Replace \"-1\" with \"Yes\" and \"1\" with \"No\"\n",
|
| 1516 |
+
"embeding_df['iso_forest_outliers'] = embeding_df['iso_forest_outliers'].replace([-1, 1], [\"Yes\", \"No\"])\n",
|
| 1517 |
+
"# Print the first 5 firms\n",
|
| 1518 |
+
"embeding_df"
|
| 1519 |
+
]
|
| 1520 |
+
},
|
| 1521 |
+
{
|
| 1522 |
+
"cell_type": "code",
|
| 1523 |
+
"execution_count": 11,
|
| 1524 |
+
"metadata": {},
|
| 1525 |
+
"outputs": [
|
| 1526 |
+
{
|
| 1527 |
+
"data": {
|
| 1528 |
+
"text/plain": [
|
| 1529 |
+
"iso_forest_outliers\n",
|
| 1530 |
+
"False 43349\n",
|
| 1531 |
+
"Name: count, dtype: int64"
|
| 1532 |
+
]
|
| 1533 |
+
},
|
| 1534 |
+
"execution_count": 11,
|
| 1535 |
+
"metadata": {},
|
| 1536 |
+
"output_type": "execute_result"
|
| 1537 |
+
}
|
| 1538 |
+
],
|
| 1539 |
+
"source": [
|
| 1540 |
+
"(embeding_df['iso_forest_outliers']=='YES').value_counts()"
|
| 1541 |
+
]
|
| 1542 |
+
},
|
| 1543 |
+
{
|
| 1544 |
+
"attachments": {},
|
| 1545 |
+
"cell_type": "markdown",
|
| 1546 |
+
"metadata": {},
|
| 1547 |
+
"source": [
|
| 1548 |
+
"# lof"
|
| 1549 |
+
]
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"cell_type": "code",
|
| 1553 |
+
"execution_count": 12,
|
| 1554 |
+
"metadata": {},
|
| 1555 |
+
"outputs": [],
|
| 1556 |
+
"source": [
|
| 1557 |
+
"from sklearn.neighbors import LocalOutlierFactor\n",
|
| 1558 |
+
"from numpy import quantile, where, random"
|
| 1559 |
+
]
|
| 1560 |
+
},
|
| 1561 |
+
{
|
| 1562 |
+
"cell_type": "code",
|
| 1563 |
+
"execution_count": 13,
|
| 1564 |
+
"metadata": {},
|
| 1565 |
+
"outputs": [],
|
| 1566 |
+
"source": [
|
| 1567 |
+
"# Train the model\n",
|
| 1568 |
+
"clf = LocalOutlierFactor(n_neighbors=20, contamination=0.1)\n",
|
| 1569 |
+
"out=clf.fit_predict(copy_df)\n",
|
| 1570 |
+
"# Predictions\n",
|
| 1571 |
+
"lof = clf.negative_outlier_factor_\n",
|
| 1572 |
+
"embeding_df[\"lof_outliers\"] = lof\n",
|
| 1573 |
+
"embeding_df[\"outliers\"]= out"
|
| 1574 |
+
]
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"cell_type": "code",
|
| 1578 |
+
"execution_count": 14,
|
| 1579 |
+
"metadata": {},
|
| 1580 |
+
"outputs": [],
|
| 1581 |
+
"source": [
|
| 1582 |
+
"embeding_df['outliers'] = embeding_df['outliers'].replace([-1, 1], [\"Yes\", \"No\"])"
|
| 1583 |
+
]
|
| 1584 |
+
},
|
| 1585 |
+
{
|
| 1586 |
+
"cell_type": "code",
|
| 1587 |
+
"execution_count": 15,
|
| 1588 |
+
"metadata": {},
|
| 1589 |
+
"outputs": [],
|
| 1590 |
+
"source": [
|
| 1591 |
+
"embeding_df['tweet']=data"
|
| 1592 |
+
]
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"cell_type": "code",
|
| 1596 |
+
"execution_count": 16,
|
| 1597 |
+
"metadata": {},
|
| 1598 |
+
"outputs": [],
|
| 1599 |
+
"source": [
|
| 1600 |
+
"x=embeding_df[embeding_df['iso_forest_outliers']=='Yes' ]"
|
| 1601 |
+
]
|
| 1602 |
+
},
|
| 1603 |
+
{
|
| 1604 |
+
"cell_type": "code",
|
| 1605 |
+
"execution_count": 17,
|
| 1606 |
+
"metadata": {},
|
| 1607 |
+
"outputs": [],
|
| 1608 |
+
"source": [
|
| 1609 |
+
"embeding_df.drop(x.loc[x['outliers']=='Yes' ].index, inplace=True)"
|
| 1610 |
+
]
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"cell_type": "code",
|
| 1614 |
+
"execution_count": 18,
|
| 1615 |
+
"metadata": {},
|
| 1616 |
+
"outputs": [
|
| 1617 |
+
{
|
| 1618 |
+
"data": {
|
| 1619 |
+
"text/html": [
|
| 1620 |
+
"<div>\n",
|
| 1621 |
+
"<style scoped>\n",
|
| 1622 |
+
" .dataframe tbody tr th:only-of-type {\n",
|
| 1623 |
+
" vertical-align: middle;\n",
|
| 1624 |
+
" }\n",
|
| 1625 |
+
"\n",
|
| 1626 |
+
" .dataframe tbody tr th {\n",
|
| 1627 |
+
" vertical-align: top;\n",
|
| 1628 |
+
" }\n",
|
| 1629 |
+
"\n",
|
| 1630 |
+
" .dataframe thead th {\n",
|
| 1631 |
+
" text-align: right;\n",
|
| 1632 |
+
" }\n",
|
| 1633 |
+
"</style>\n",
|
| 1634 |
+
"<table border=\"1\" class=\"dataframe\">\n",
|
| 1635 |
+
" <thead>\n",
|
| 1636 |
+
" <tr style=\"text-align: right;\">\n",
|
| 1637 |
+
" <th></th>\n",
|
| 1638 |
+
" <th>0</th>\n",
|
| 1639 |
+
" <th>1</th>\n",
|
| 1640 |
+
" <th>2</th>\n",
|
| 1641 |
+
" <th>3</th>\n",
|
| 1642 |
+
" <th>4</th>\n",
|
| 1643 |
+
" <th>5</th>\n",
|
| 1644 |
+
" <th>6</th>\n",
|
| 1645 |
+
" <th>7</th>\n",
|
| 1646 |
+
" <th>8</th>\n",
|
| 1647 |
+
" <th>9</th>\n",
|
| 1648 |
+
" <th>...</th>\n",
|
| 1649 |
+
" <th>60</th>\n",
|
| 1650 |
+
" <th>61</th>\n",
|
| 1651 |
+
" <th>62</th>\n",
|
| 1652 |
+
" <th>63</th>\n",
|
| 1653 |
+
" <th>labels</th>\n",
|
| 1654 |
+
" <th>iso_forest_scores</th>\n",
|
| 1655 |
+
" <th>iso_forest_outliers</th>\n",
|
| 1656 |
+
" <th>lof_outliers</th>\n",
|
| 1657 |
+
" <th>outliers</th>\n",
|
| 1658 |
+
" <th>tweet</th>\n",
|
| 1659 |
+
" </tr>\n",
|
| 1660 |
+
" </thead>\n",
|
| 1661 |
+
" <tbody>\n",
|
| 1662 |
+
" <tr>\n",
|
| 1663 |
+
" <th>0</th>\n",
|
| 1664 |
+
" <td>101</td>\n",
|
| 1665 |
+
" <td>10110</td>\n",
|
| 1666 |
+
" <td>175</td>\n",
|
| 1667 |
+
" <td>78653</td>\n",
|
| 1668 |
+
" <td>189</td>\n",
|
| 1669 |
+
" <td>25285</td>\n",
|
| 1670 |
+
" <td>15976</td>\n",
|
| 1671 |
+
" <td>40840</td>\n",
|
| 1672 |
+
" <td>276</td>\n",
|
| 1673 |
+
" <td>31623</td>\n",
|
| 1674 |
+
" <td>...</td>\n",
|
| 1675 |
+
" <td>0</td>\n",
|
| 1676 |
+
" <td>0</td>\n",
|
| 1677 |
+
" <td>0</td>\n",
|
| 1678 |
+
" <td>0</td>\n",
|
| 1679 |
+
" <td>0</td>\n",
|
| 1680 |
+
" <td>0.189202</td>\n",
|
| 1681 |
+
" <td>No</td>\n",
|
| 1682 |
+
" <td>-1.209681</td>\n",
|
| 1683 |
+
" <td>No</td>\n",
|
| 1684 |
+
" <td>en güzel uyuyan insan ödülü jeon jungkook'a g...</td>\n",
|
| 1685 |
+
" </tr>\n",
|
| 1686 |
+
" <tr>\n",
|
| 1687 |
+
" <th>1</th>\n",
|
| 1688 |
+
" <td>101</td>\n",
|
| 1689 |
+
" <td>11589</td>\n",
|
| 1690 |
+
" <td>10706</td>\n",
|
| 1691 |
+
" <td>10713</td>\n",
|
| 1692 |
+
" <td>10794</td>\n",
|
| 1693 |
+
" <td>94698</td>\n",
|
| 1694 |
+
" <td>30668</td>\n",
|
| 1695 |
+
" <td>24883</td>\n",
|
| 1696 |
+
" <td>117</td>\n",
|
| 1697 |
+
" <td>23763</td>\n",
|
| 1698 |
+
" <td>...</td>\n",
|
| 1699 |
+
" <td>0</td>\n",
|
| 1700 |
+
" <td>0</td>\n",
|
| 1701 |
+
" <td>0</td>\n",
|
| 1702 |
+
" <td>0</td>\n",
|
| 1703 |
+
" <td>0</td>\n",
|
| 1704 |
+
" <td>0.181234</td>\n",
|
| 1705 |
+
" <td>No</td>\n",
|
| 1706 |
+
" <td>-1.107479</td>\n",
|
| 1707 |
+
" <td>No</td>\n",
|
| 1708 |
+
" <td>Mekanı cennet olsun, saygılar sayın avukatımı...</td>\n",
|
| 1709 |
+
" </tr>\n",
|
| 1710 |
+
" <tr>\n",
|
| 1711 |
+
" <th>2</th>\n",
|
| 1712 |
+
" <td>101</td>\n",
|
| 1713 |
+
" <td>148</td>\n",
|
| 1714 |
+
" <td>30471</td>\n",
|
| 1715 |
+
" <td>10774</td>\n",
|
| 1716 |
+
" <td>13785</td>\n",
|
| 1717 |
+
" <td>13779</td>\n",
|
| 1718 |
+
" <td>33642</td>\n",
|
| 1719 |
+
" <td>14399</td>\n",
|
| 1720 |
+
" <td>48271</td>\n",
|
| 1721 |
+
" <td>76686</td>\n",
|
| 1722 |
+
" <td>...</td>\n",
|
| 1723 |
+
" <td>0</td>\n",
|
| 1724 |
+
" <td>0</td>\n",
|
| 1725 |
+
" <td>0</td>\n",
|
| 1726 |
+
" <td>0</td>\n",
|
| 1727 |
+
" <td>0</td>\n",
|
| 1728 |
+
" <td>0.166332</td>\n",
|
| 1729 |
+
" <td>No</td>\n",
|
| 1730 |
+
" <td>-1.202529</td>\n",
|
| 1731 |
+
" <td>No</td>\n",
|
| 1732 |
+
" <td>Kızlar aranızda kas yığını beylere düşenler ol...</td>\n",
|
| 1733 |
+
" </tr>\n",
|
| 1734 |
+
" <tr>\n",
|
| 1735 |
+
" <th>3</th>\n",
|
| 1736 |
+
" <td>101</td>\n",
|
| 1737 |
+
" <td>19319</td>\n",
|
| 1738 |
+
" <td>16724</td>\n",
|
| 1739 |
+
" <td>10118</td>\n",
|
| 1740 |
+
" <td>10107</td>\n",
|
| 1741 |
+
" <td>78323</td>\n",
|
| 1742 |
+
" <td>12407</td>\n",
|
| 1743 |
+
" <td>38959</td>\n",
|
| 1744 |
+
" <td>22934</td>\n",
|
| 1745 |
+
" <td>10147</td>\n",
|
| 1746 |
+
" <td>...</td>\n",
|
| 1747 |
+
" <td>0</td>\n",
|
| 1748 |
+
" <td>0</td>\n",
|
| 1749 |
+
" <td>0</td>\n",
|
| 1750 |
+
" <td>0</td>\n",
|
| 1751 |
+
" <td>0</td>\n",
|
| 1752 |
+
" <td>0.151816</td>\n",
|
| 1753 |
+
" <td>No</td>\n",
|
| 1754 |
+
" <td>-1.216599</td>\n",
|
| 1755 |
+
" <td>No</td>\n",
|
| 1756 |
+
" <td>Biraz ders çalışayım. Tembellik ve uyku düşman...</td>\n",
|
| 1757 |
+
" </tr>\n",
|
| 1758 |
+
" <tr>\n",
|
| 1759 |
+
" <th>4</th>\n",
|
| 1760 |
+
" <td>101</td>\n",
|
| 1761 |
+
" <td>30932</td>\n",
|
| 1762 |
+
" <td>58706</td>\n",
|
| 1763 |
+
" <td>58054</td>\n",
|
| 1764 |
+
" <td>44907</td>\n",
|
| 1765 |
+
" <td>10224</td>\n",
|
| 1766 |
+
" <td>106583</td>\n",
|
| 1767 |
+
" <td>10288</td>\n",
|
| 1768 |
+
" <td>12524</td>\n",
|
| 1769 |
+
" <td>13878</td>\n",
|
| 1770 |
+
" <td>...</td>\n",
|
| 1771 |
+
" <td>0</td>\n",
|
| 1772 |
+
" <td>0</td>\n",
|
| 1773 |
+
" <td>0</td>\n",
|
| 1774 |
+
" <td>0</td>\n",
|
| 1775 |
+
" <td>0</td>\n",
|
| 1776 |
+
" <td>0.184008</td>\n",
|
| 1777 |
+
" <td>No</td>\n",
|
| 1778 |
+
" <td>-1.188488</td>\n",
|
| 1779 |
+
" <td>No</td>\n",
|
| 1780 |
+
" <td>Trezeguet yerine El Sharawy daha iyi olmaz mı</td>\n",
|
| 1781 |
+
" </tr>\n",
|
| 1782 |
+
" <tr>\n",
|
| 1783 |
+
" <th>...</th>\n",
|
| 1784 |
+
" <td>...</td>\n",
|
| 1785 |
+
" <td>...</td>\n",
|
| 1786 |
+
" <td>...</td>\n",
|
| 1787 |
+
" <td>...</td>\n",
|
| 1788 |
+
" <td>...</td>\n",
|
| 1789 |
+
" <td>...</td>\n",
|
| 1790 |
+
" <td>...</td>\n",
|
| 1791 |
+
" <td>...</td>\n",
|
| 1792 |
+
" <td>...</td>\n",
|
| 1793 |
+
" <td>...</td>\n",
|
| 1794 |
+
" <td>...</td>\n",
|
| 1795 |
+
" <td>...</td>\n",
|
| 1796 |
+
" <td>...</td>\n",
|
| 1797 |
+
" <td>...</td>\n",
|
| 1798 |
+
" <td>...</td>\n",
|
| 1799 |
+
" <td>...</td>\n",
|
| 1800 |
+
" <td>...</td>\n",
|
| 1801 |
+
" <td>...</td>\n",
|
| 1802 |
+
" <td>...</td>\n",
|
| 1803 |
+
" <td>...</td>\n",
|
| 1804 |
+
" <td>...</td>\n",
|
| 1805 |
+
" </tr>\n",
|
| 1806 |
+
" <tr>\n",
|
| 1807 |
+
" <th>43344</th>\n",
|
| 1808 |
+
" <td>101</td>\n",
|
| 1809 |
+
" <td>20065</td>\n",
|
| 1810 |
+
" <td>10161</td>\n",
|
| 1811 |
+
" <td>115</td>\n",
|
| 1812 |
+
" <td>115</td>\n",
|
| 1813 |
+
" <td>103784</td>\n",
|
| 1814 |
+
" <td>10774</td>\n",
|
| 1815 |
+
" <td>21388</td>\n",
|
| 1816 |
+
" <td>10245</td>\n",
|
| 1817 |
+
" <td>92067</td>\n",
|
| 1818 |
+
" <td>...</td>\n",
|
| 1819 |
+
" <td>0</td>\n",
|
| 1820 |
+
" <td>0</td>\n",
|
| 1821 |
+
" <td>0</td>\n",
|
| 1822 |
+
" <td>0</td>\n",
|
| 1823 |
+
" <td>1</td>\n",
|
| 1824 |
+
" <td>0.079412</td>\n",
|
| 1825 |
+
" <td>No</td>\n",
|
| 1826 |
+
" <td>-1.196769</td>\n",
|
| 1827 |
+
" <td>No</td>\n",
|
| 1828 |
+
" <td>Hil**adamlar kesinlikle kelimeleri anlamıyorla...</td>\n",
|
| 1829 |
+
" </tr>\n",
|
| 1830 |
+
" <tr>\n",
|
| 1831 |
+
" <th>43345</th>\n",
|
| 1832 |
+
" <td>101</td>\n",
|
| 1833 |
+
" <td>139</td>\n",
|
| 1834 |
+
" <td>80839</td>\n",
|
| 1835 |
+
" <td>24109</td>\n",
|
| 1836 |
+
" <td>13406</td>\n",
|
| 1837 |
+
" <td>18985</td>\n",
|
| 1838 |
+
" <td>16285</td>\n",
|
| 1839 |
+
" <td>10163</td>\n",
|
| 1840 |
+
" <td>11062</td>\n",
|
| 1841 |
+
" <td>276</td>\n",
|
| 1842 |
+
" <td>...</td>\n",
|
| 1843 |
+
" <td>0</td>\n",
|
| 1844 |
+
" <td>0</td>\n",
|
| 1845 |
+
" <td>0</td>\n",
|
| 1846 |
+
" <td>0</td>\n",
|
| 1847 |
+
" <td>1</td>\n",
|
| 1848 |
+
" <td>0.118245</td>\n",
|
| 1849 |
+
" <td>No</td>\n",
|
| 1850 |
+
" <td>-1.108304</td>\n",
|
| 1851 |
+
" <td>No</td>\n",
|
| 1852 |
+
" <td>Böyle piçlerin çok erken ölmemelerini ve çok f...</td>\n",
|
| 1853 |
+
" </tr>\n",
|
| 1854 |
+
" <tr>\n",
|
| 1855 |
+
" <th>43346</th>\n",
|
| 1856 |
+
" <td>101</td>\n",
|
| 1857 |
+
" <td>105549</td>\n",
|
| 1858 |
+
" <td>102635</td>\n",
|
| 1859 |
+
" <td>10140</td>\n",
|
| 1860 |
+
" <td>26943</td>\n",
|
| 1861 |
+
" <td>11499</td>\n",
|
| 1862 |
+
" <td>110516</td>\n",
|
| 1863 |
+
" <td>21899</td>\n",
|
| 1864 |
+
" <td>11861</td>\n",
|
| 1865 |
+
" <td>10561</td>\n",
|
| 1866 |
+
" <td>...</td>\n",
|
| 1867 |
+
" <td>0</td>\n",
|
| 1868 |
+
" <td>0</td>\n",
|
| 1869 |
+
" <td>0</td>\n",
|
| 1870 |
+
" <td>0</td>\n",
|
| 1871 |
+
" <td>1</td>\n",
|
| 1872 |
+
" <td>0.138229</td>\n",
|
| 1873 |
+
" <td>No</td>\n",
|
| 1874 |
+
" <td>-1.307328</td>\n",
|
| 1875 |
+
" <td>No</td>\n",
|
| 1876 |
+
" <td>Turgay denilen bu holigonda bir sorun yok, gur...</td>\n",
|
| 1877 |
+
" </tr>\n",
|
| 1878 |
+
" <tr>\n",
|
| 1879 |
+
" <th>43347</th>\n",
|
| 1880 |
+
" <td>101</td>\n",
|
| 1881 |
+
" <td>81424</td>\n",
|
| 1882 |
+
" <td>26398</td>\n",
|
| 1883 |
+
" <td>92017</td>\n",
|
| 1884 |
+
" <td>109620</td>\n",
|
| 1885 |
+
" <td>10941</td>\n",
|
| 1886 |
+
" <td>76010</td>\n",
|
| 1887 |
+
" <td>10115</td>\n",
|
| 1888 |
+
" <td>19830</td>\n",
|
| 1889 |
+
" <td>26083</td>\n",
|
| 1890 |
+
" <td>...</td>\n",
|
| 1891 |
+
" <td>0</td>\n",
|
| 1892 |
+
" <td>0</td>\n",
|
| 1893 |
+
" <td>0</td>\n",
|
| 1894 |
+
" <td>0</td>\n",
|
| 1895 |
+
" <td>1</td>\n",
|
| 1896 |
+
" <td>0.181065</td>\n",
|
| 1897 |
+
" <td>No</td>\n",
|
| 1898 |
+
" <td>-1.127932</td>\n",
|
| 1899 |
+
" <td>No</td>\n",
|
| 1900 |
+
" <td>Umarım ülkenin düşük zekadan kurtulması ilgile...</td>\n",
|
| 1901 |
+
" </tr>\n",
|
| 1902 |
+
" <tr>\n",
|
| 1903 |
+
" <th>43348</th>\n",
|
| 1904 |
+
" <td>101</td>\n",
|
| 1905 |
+
" <td>39774</td>\n",
|
| 1906 |
+
" <td>11127</td>\n",
|
| 1907 |
+
" <td>45989</td>\n",
|
| 1908 |
+
" <td>24596</td>\n",
|
| 1909 |
+
" <td>11933</td>\n",
|
| 1910 |
+
" <td>170</td>\n",
|
| 1911 |
+
" <td>17145</td>\n",
|
| 1912 |
+
" <td>10710</td>\n",
|
| 1913 |
+
" <td>39125</td>\n",
|
| 1914 |
+
" <td>...</td>\n",
|
| 1915 |
+
" <td>0</td>\n",
|
| 1916 |
+
" <td>0</td>\n",
|
| 1917 |
+
" <td>0</td>\n",
|
| 1918 |
+
" <td>0</td>\n",
|
| 1919 |
+
" <td>1</td>\n",
|
| 1920 |
+
" <td>0.085161</td>\n",
|
| 1921 |
+
" <td>No</td>\n",
|
| 1922 |
+
" <td>-1.286323</td>\n",
|
| 1923 |
+
" <td>No</td>\n",
|
| 1924 |
+
" <td>CHP sandıkları bırakmaz, üzerine oturur, bir c...</td>\n",
|
| 1925 |
+
" </tr>\n",
|
| 1926 |
+
" </tbody>\n",
|
| 1927 |
+
"</table>\n",
|
| 1928 |
+
"<p>43029 rows × 70 columns</p>\n",
|
| 1929 |
+
"</div>"
|
| 1930 |
+
],
|
| 1931 |
+
"text/plain": [
|
| 1932 |
+
" 0 1 2 3 4 5 6 7 8 \n",
|
| 1933 |
+
"0 101 10110 175 78653 189 25285 15976 40840 276 \\\n",
|
| 1934 |
+
"1 101 11589 10706 10713 10794 94698 30668 24883 117 \n",
|
| 1935 |
+
"2 101 148 30471 10774 13785 13779 33642 14399 48271 \n",
|
| 1936 |
+
"3 101 19319 16724 10118 10107 78323 12407 38959 22934 \n",
|
| 1937 |
+
"4 101 30932 58706 58054 44907 10224 106583 10288 12524 \n",
|
| 1938 |
+
"... ... ... ... ... ... ... ... ... ... \n",
|
| 1939 |
+
"43344 101 20065 10161 115 115 103784 10774 21388 10245 \n",
|
| 1940 |
+
"43345 101 139 80839 24109 13406 18985 16285 10163 11062 \n",
|
| 1941 |
+
"43346 101 105549 102635 10140 26943 11499 110516 21899 11861 \n",
|
| 1942 |
+
"43347 101 81424 26398 92017 109620 10941 76010 10115 19830 \n",
|
| 1943 |
+
"43348 101 39774 11127 45989 24596 11933 170 17145 10710 \n",
|
| 1944 |
+
"\n",
|
| 1945 |
+
" 9 ... 60 61 62 63 labels iso_forest_scores \n",
|
| 1946 |
+
"0 31623 ... 0 0 0 0 0 0.189202 \\\n",
|
| 1947 |
+
"1 23763 ... 0 0 0 0 0 0.181234 \n",
|
| 1948 |
+
"2 76686 ... 0 0 0 0 0 0.166332 \n",
|
| 1949 |
+
"3 10147 ... 0 0 0 0 0 0.151816 \n",
|
| 1950 |
+
"4 13878 ... 0 0 0 0 0 0.184008 \n",
|
| 1951 |
+
"... ... ... .. .. .. .. ... ... \n",
|
| 1952 |
+
"43344 92067 ... 0 0 0 0 1 0.079412 \n",
|
| 1953 |
+
"43345 276 ... 0 0 0 0 1 0.118245 \n",
|
| 1954 |
+
"43346 10561 ... 0 0 0 0 1 0.138229 \n",
|
| 1955 |
+
"43347 26083 ... 0 0 0 0 1 0.181065 \n",
|
| 1956 |
+
"43348 39125 ... 0 0 0 0 1 0.085161 \n",
|
| 1957 |
+
"\n",
|
| 1958 |
+
" iso_forest_outliers lof_outliers outliers \n",
|
| 1959 |
+
"0 No -1.209681 No \\\n",
|
| 1960 |
+
"1 No -1.107479 No \n",
|
| 1961 |
+
"2 No -1.202529 No \n",
|
| 1962 |
+
"3 No -1.216599 No \n",
|
| 1963 |
+
"4 No -1.188488 No \n",
|
| 1964 |
+
"... ... ... ... \n",
|
| 1965 |
+
"43344 No -1.196769 No \n",
|
| 1966 |
+
"43345 No -1.108304 No \n",
|
| 1967 |
+
"43346 No -1.307328 No \n",
|
| 1968 |
+
"43347 No -1.127932 No \n",
|
| 1969 |
+
"43348 No -1.286323 No \n",
|
| 1970 |
+
"\n",
|
| 1971 |
+
" tweet \n",
|
| 1972 |
+
"0 en güzel uyuyan insan ödülü jeon jungkook'a g... \n",
|
| 1973 |
+
"1 Mekanı cennet olsun, saygılar sayın avukatımı... \n",
|
| 1974 |
+
"2 Kızlar aranızda kas yığını beylere düşenler ol... \n",
|
| 1975 |
+
"3 Biraz ders çalışayım. Tembellik ve uyku düşman... \n",
|
| 1976 |
+
"4 Trezeguet yerine El Sharawy daha iyi olmaz mı \n",
|
| 1977 |
+
"... ... \n",
|
| 1978 |
+
"43344 Hil**adamlar kesinlikle kelimeleri anlamıyorla... \n",
|
| 1979 |
+
"43345 Böyle piçlerin çok erken ölmemelerini ve çok f... \n",
|
| 1980 |
+
"43346 Turgay denilen bu holigonda bir sorun yok, gur... \n",
|
| 1981 |
+
"43347 Umarım ülkenin düşük zekadan kurtulması ilgile... \n",
|
| 1982 |
+
"43348 CHP sandıkları bırakmaz, üzerine oturur, bir c... \n",
|
| 1983 |
+
"\n",
|
| 1984 |
+
"[43029 rows x 70 columns]"
|
| 1985 |
+
]
|
| 1986 |
+
},
|
| 1987 |
+
"execution_count": 18,
|
| 1988 |
+
"metadata": {},
|
| 1989 |
+
"output_type": "execute_result"
|
| 1990 |
+
}
|
| 1991 |
+
],
|
| 1992 |
+
"source": [
|
| 1993 |
+
"embeding_df"
|
| 1994 |
+
]
|
| 1995 |
+
},
|
| 1996 |
+
{
|
| 1997 |
+
"cell_type": "code",
|
| 1998 |
+
"execution_count": 19,
|
| 1999 |
+
"metadata": {},
|
| 2000 |
+
"outputs": [],
|
| 2001 |
+
"source": [
|
| 2002 |
+
"# embeding_df.drop(embeding_df.loc[embeding_df['outliers']=='Yes' ].index, inplace=True)\n",
|
| 2003 |
+
"# embeding_df.drop(embeding_df.loc[embeding_df['iso_forest_outliers']=='Yes' ].index, inplace=True)"
|
| 2004 |
+
]
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"cell_type": "code",
|
| 2008 |
+
"execution_count": 20,
|
| 2009 |
+
"metadata": {},
|
| 2010 |
+
"outputs": [],
|
| 2011 |
+
"source": [
|
| 2012 |
+
"# iso_df=embeding_df[embeding_df['iso_forest_outliers']=='Yes' ]\n",
|
| 2013 |
+
"# embeding_df.drop(embeding_df.loc[embeding_df['iso_forest_outliers']=='Yes' ].index, inplace=True)"
|
| 2014 |
+
]
|
| 2015 |
+
},
|
| 2016 |
+
{
|
| 2017 |
+
"cell_type": "code",
|
| 2018 |
+
"execution_count": 21,
|
| 2019 |
+
"metadata": {},
|
| 2020 |
+
"outputs": [],
|
| 2021 |
+
"source": [
|
| 2022 |
+
"# lof_df=embeding_df[embeding_df['outliers']=='Yes' ]\n",
|
| 2023 |
+
"# embeding_df.drop(embeding_df.loc[embeding_df['outliers']=='Yes' ].index, inplace=True)"
|
| 2024 |
+
]
|
| 2025 |
+
},
|
| 2026 |
+
{
|
| 2027 |
+
"cell_type": "code",
|
| 2028 |
+
"execution_count": 22,
|
| 2029 |
+
"metadata": {},
|
| 2030 |
+
"outputs": [],
|
| 2031 |
+
"source": [
|
| 2032 |
+
"# iso_df"
|
| 2033 |
+
]
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"cell_type": "code",
|
| 2037 |
+
"execution_count": 23,
|
| 2038 |
+
"metadata": {},
|
| 2039 |
+
"outputs": [],
|
| 2040 |
+
"source": [
|
| 2041 |
+
"# iso_df['labels']=iso_df['labels'].replace({0: 1, 1: 0})"
|
| 2042 |
+
]
|
| 2043 |
+
},
|
| 2044 |
+
{
|
| 2045 |
+
"cell_type": "code",
|
| 2046 |
+
"execution_count": 24,
|
| 2047 |
+
"metadata": {},
|
| 2048 |
+
"outputs": [],
|
| 2049 |
+
"source": [
|
| 2050 |
+
"# iso_df"
|
| 2051 |
+
]
|
| 2052 |
+
},
|
| 2053 |
+
{
|
| 2054 |
+
"cell_type": "code",
|
| 2055 |
+
"execution_count": 25,
|
| 2056 |
+
"metadata": {},
|
| 2057 |
+
"outputs": [],
|
| 2058 |
+
"source": [
|
| 2059 |
+
"# lof_df['labels']=lof_df['labels'].replace({0: 1, 1: 0})"
|
| 2060 |
+
]
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"cell_type": "code",
|
| 2064 |
+
"execution_count": 26,
|
| 2065 |
+
"metadata": {},
|
| 2066 |
+
"outputs": [],
|
| 2067 |
+
"source": [
|
| 2068 |
+
"# lof_df"
|
| 2069 |
+
]
|
| 2070 |
+
},
|
| 2071 |
+
{
|
| 2072 |
+
"cell_type": "code",
|
| 2073 |
+
"execution_count": 27,
|
| 2074 |
+
"metadata": {},
|
| 2075 |
+
"outputs": [],
|
| 2076 |
+
"source": [
|
| 2077 |
+
"# x=pd.concat([lof_df,iso_df], axis=0)"
|
| 2078 |
+
]
|
| 2079 |
+
},
|
| 2080 |
+
{
|
| 2081 |
+
"cell_type": "code",
|
| 2082 |
+
"execution_count": 28,
|
| 2083 |
+
"metadata": {},
|
| 2084 |
+
"outputs": [],
|
| 2085 |
+
"source": [
|
| 2086 |
+
"# embeding_df=pd.concat([x,embeding_df], axis=0)"
|
| 2087 |
+
]
|
| 2088 |
+
},
|
| 2089 |
+
{
|
| 2090 |
+
"cell_type": "code",
|
| 2091 |
+
"execution_count": 29,
|
| 2092 |
+
"metadata": {},
|
| 2093 |
+
"outputs": [],
|
| 2094 |
+
"source": [
|
| 2095 |
+
"# embeding_df.reset_index()"
|
| 2096 |
+
]
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"cell_type": "code",
|
| 2100 |
+
"execution_count": 30,
|
| 2101 |
+
"metadata": {},
|
| 2102 |
+
"outputs": [],
|
| 2103 |
+
"source": [
|
| 2104 |
+
"# embeding_df=embeding_df.drop(['iso_forest_scores', 'iso_forest_outliers','lof_outliers','outliers'], axis=1)"
|
| 2105 |
+
]
|
| 2106 |
+
},
|
| 2107 |
+
{
|
| 2108 |
+
"cell_type": "code",
|
| 2109 |
+
"execution_count": 31,
|
| 2110 |
+
"metadata": {},
|
| 2111 |
+
"outputs": [],
|
| 2112 |
+
"source": [
|
| 2113 |
+
"embeding_df['0'] = embeding_df[embeding_df.columns[:-1]].apply(lambda x: ','.join(x.dropna().astype(str)),axis=1)\n"
|
| 2114 |
+
]
|
| 2115 |
+
},
|
| 2116 |
+
{
|
| 2117 |
+
"cell_type": "code",
|
| 2118 |
+
"execution_count": 32,
|
| 2119 |
+
"metadata": {},
|
| 2120 |
+
"outputs": [],
|
| 2121 |
+
"source": [
|
| 2122 |
+
"df=pd.DataFrame()\n",
|
| 2123 |
+
"df['tweet']=embeding_df['tweet']\n",
|
| 2124 |
+
"df['subtas_a']=embeding_df['labels']\n"
|
| 2125 |
+
]
|
| 2126 |
+
},
|
| 2127 |
+
{
|
| 2128 |
+
"cell_type": "code",
|
| 2129 |
+
"execution_count": 33,
|
| 2130 |
+
"metadata": {},
|
| 2131 |
+
"outputs": [
|
| 2132 |
+
{
|
| 2133 |
+
"data": {
|
| 2134 |
+
"text/html": [
|
| 2135 |
+
"<div>\n",
|
| 2136 |
+
"<style scoped>\n",
|
| 2137 |
+
" .dataframe tbody tr th:only-of-type {\n",
|
| 2138 |
+
" vertical-align: middle;\n",
|
| 2139 |
+
" }\n",
|
| 2140 |
+
"\n",
|
| 2141 |
+
" .dataframe tbody tr th {\n",
|
| 2142 |
+
" vertical-align: top;\n",
|
| 2143 |
+
" }\n",
|
| 2144 |
+
"\n",
|
| 2145 |
+
" .dataframe thead th {\n",
|
| 2146 |
+
" text-align: right;\n",
|
| 2147 |
+
" }\n",
|
| 2148 |
+
"</style>\n",
|
| 2149 |
+
"<table border=\"1\" class=\"dataframe\">\n",
|
| 2150 |
+
" <thead>\n",
|
| 2151 |
+
" <tr style=\"text-align: right;\">\n",
|
| 2152 |
+
" <th></th>\n",
|
| 2153 |
+
" <th>tweet</th>\n",
|
| 2154 |
+
" <th>subtas_a</th>\n",
|
| 2155 |
+
" </tr>\n",
|
| 2156 |
+
" </thead>\n",
|
| 2157 |
+
" <tbody>\n",
|
| 2158 |
+
" <tr>\n",
|
| 2159 |
+
" <th>0</th>\n",
|
| 2160 |
+
" <td>en güzel uyuyan insan ödülü jeon jungkook'a g...</td>\n",
|
| 2161 |
+
" <td>0</td>\n",
|
| 2162 |
+
" </tr>\n",
|
| 2163 |
+
" <tr>\n",
|
| 2164 |
+
" <th>1</th>\n",
|
| 2165 |
+
" <td>Mekanı cennet olsun, saygılar sayın avukatımı...</td>\n",
|
| 2166 |
+
" <td>0</td>\n",
|
| 2167 |
+
" </tr>\n",
|
| 2168 |
+
" <tr>\n",
|
| 2169 |
+
" <th>2</th>\n",
|
| 2170 |
+
" <td>Kızlar aranızda kas yığını beylere düşenler ol...</td>\n",
|
| 2171 |
+
" <td>0</td>\n",
|
| 2172 |
+
" </tr>\n",
|
| 2173 |
+
" <tr>\n",
|
| 2174 |
+
" <th>3</th>\n",
|
| 2175 |
+
" <td>Biraz ders çalışayım. Tembellik ve uyku düşman...</td>\n",
|
| 2176 |
+
" <td>0</td>\n",
|
| 2177 |
+
" </tr>\n",
|
| 2178 |
+
" <tr>\n",
|
| 2179 |
+
" <th>4</th>\n",
|
| 2180 |
+
" <td>Trezeguet yerine El Sharawy daha iyi olmaz mı</td>\n",
|
| 2181 |
+
" <td>0</td>\n",
|
| 2182 |
+
" </tr>\n",
|
| 2183 |
+
" <tr>\n",
|
| 2184 |
+
" <th>...</th>\n",
|
| 2185 |
+
" <td>...</td>\n",
|
| 2186 |
+
" <td>...</td>\n",
|
| 2187 |
+
" </tr>\n",
|
| 2188 |
+
" <tr>\n",
|
| 2189 |
+
" <th>43344</th>\n",
|
| 2190 |
+
" <td>Hil**adamlar kesinlikle kelimeleri anlamıyorla...</td>\n",
|
| 2191 |
+
" <td>1</td>\n",
|
| 2192 |
+
" </tr>\n",
|
| 2193 |
+
" <tr>\n",
|
| 2194 |
+
" <th>43345</th>\n",
|
| 2195 |
+
" <td>Böyle piçlerin çok erken ölmemelerini ve çok f...</td>\n",
|
| 2196 |
+
" <td>1</td>\n",
|
| 2197 |
+
" </tr>\n",
|
| 2198 |
+
" <tr>\n",
|
| 2199 |
+
" <th>43346</th>\n",
|
| 2200 |
+
" <td>Turgay denilen bu holigonda bir sorun yok, gur...</td>\n",
|
| 2201 |
+
" <td>1</td>\n",
|
| 2202 |
+
" </tr>\n",
|
| 2203 |
+
" <tr>\n",
|
| 2204 |
+
" <th>43347</th>\n",
|
| 2205 |
+
" <td>Umarım ülkenin düşük zekadan kurtulması ilgile...</td>\n",
|
| 2206 |
+
" <td>1</td>\n",
|
| 2207 |
+
" </tr>\n",
|
| 2208 |
+
" <tr>\n",
|
| 2209 |
+
" <th>43348</th>\n",
|
| 2210 |
+
" <td>CHP sandıkları bırakmaz, üzerine oturur, bir c...</td>\n",
|
| 2211 |
+
" <td>1</td>\n",
|
| 2212 |
+
" </tr>\n",
|
| 2213 |
+
" </tbody>\n",
|
| 2214 |
+
"</table>\n",
|
| 2215 |
+
"<p>43029 rows × 2 columns</p>\n",
|
| 2216 |
+
"</div>"
|
| 2217 |
+
],
|
| 2218 |
+
"text/plain": [
|
| 2219 |
+
" tweet subtas_a\n",
|
| 2220 |
+
"0 en güzel uyuyan insan ödülü jeon jungkook'a g... 0\n",
|
| 2221 |
+
"1 Mekanı cennet olsun, saygılar sayın avukatımı... 0\n",
|
| 2222 |
+
"2 Kızlar aranızda kas yığını beylere düşenler ol... 0\n",
|
| 2223 |
+
"3 Biraz ders çalışayım. Tembellik ve uyku düşman... 0\n",
|
| 2224 |
+
"4 Trezeguet yerine El Sharawy daha iyi olmaz mı 0\n",
|
| 2225 |
+
"... ... ...\n",
|
| 2226 |
+
"43344 Hil**adamlar kesinlikle kelimeleri anlamıyorla... 1\n",
|
| 2227 |
+
"43345 Böyle piçlerin çok erken ölmemelerini ve çok f... 1\n",
|
| 2228 |
+
"43346 Turgay denilen bu holigonda bir sorun yok, gur... 1\n",
|
| 2229 |
+
"43347 Umarım ülkenin düşük zekadan kurtulması ilgile... 1\n",
|
| 2230 |
+
"43348 CHP sandıkları bırakmaz, üzerine oturur, bir c... 1\n",
|
| 2231 |
+
"\n",
|
| 2232 |
+
"[43029 rows x 2 columns]"
|
| 2233 |
+
]
|
| 2234 |
+
},
|
| 2235 |
+
"execution_count": 33,
|
| 2236 |
+
"metadata": {},
|
| 2237 |
+
"output_type": "execute_result"
|
| 2238 |
+
}
|
| 2239 |
+
],
|
| 2240 |
+
"source": [
|
| 2241 |
+
"df"
|
| 2242 |
+
]
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"cell_type": "code",
|
| 2246 |
+
"execution_count": 40,
|
| 2247 |
+
"metadata": {},
|
| 2248 |
+
"outputs": [],
|
| 2249 |
+
"source": [
|
| 2250 |
+
"df.to_csv('inverse_outliers.csv') "
|
| 2251 |
+
]
|
| 2252 |
+
},
|
| 2253 |
+
{
|
| 2254 |
+
"cell_type": "code",
|
| 2255 |
+
"execution_count": 34,
|
| 2256 |
+
"metadata": {},
|
| 2257 |
+
"outputs": [],
|
| 2258 |
+
"source": [
|
| 2259 |
+
"df.to_csv('int_2_outliers.csv') "
|
| 2260 |
+
]
|
| 2261 |
+
},
|
| 2262 |
+
{
|
| 2263 |
+
"cell_type": "code",
|
| 2264 |
+
"execution_count": null,
|
| 2265 |
+
"metadata": {},
|
| 2266 |
+
"outputs": [],
|
| 2267 |
+
"source": []
|
| 2268 |
+
}
|
| 2269 |
+
],
|
| 2270 |
+
"metadata": {
|
| 2271 |
+
"kernelspec": {
|
| 2272 |
+
"display_name": "dl_env",
|
| 2273 |
+
"language": "python",
|
| 2274 |
+
"name": "python3"
|
| 2275 |
+
},
|
| 2276 |
+
"language_info": {
|
| 2277 |
+
"codemirror_mode": {
|
| 2278 |
+
"name": "ipython",
|
| 2279 |
+
"version": 3
|
| 2280 |
+
},
|
| 2281 |
+
"file_extension": ".py",
|
| 2282 |
+
"mimetype": "text/x-python",
|
| 2283 |
+
"name": "python",
|
| 2284 |
+
"nbconvert_exporter": "python",
|
| 2285 |
+
"pygments_lexer": "ipython3",
|
| 2286 |
+
"version": "3.9.0"
|
| 2287 |
+
},
|
| 2288 |
+
"orig_nbformat": 4
|
| 2289 |
+
},
|
| 2290 |
+
"nbformat": 4,
|
| 2291 |
+
"nbformat_minor": 2
|
| 2292 |
+
}
|
pycaret_outlier_detection.ipynb
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
requirements.txt
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
pytorch_lightning
|
| 2 |
+
emoji
|
| 3 |
+
transformers
|
| 4 |
+
numpy
|
| 5 |
+
pandas
|
| 6 |
+
os
|
| 7 |
+
random
|
| 8 |
+
torch
|
| 9 |
+
torch-metrics
|
| 10 |
+
torch-utils
|
trainer.ipynb
ADDED
|
@@ -0,0 +1,1165 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"cells": [
|
| 3 |
+
{
|
| 4 |
+
"cell_type": "code",
|
| 5 |
+
"execution_count": 1,
|
| 6 |
+
"metadata": {},
|
| 7 |
+
"outputs": [],
|
| 8 |
+
"source": [
|
| 9 |
+
"import emoji\n",
|
| 10 |
+
"import numpy as np\n",
|
| 11 |
+
"import pandas as pd\n",
|
| 12 |
+
"from sklearn.preprocessing import LabelEncoder\n",
|
| 13 |
+
"from transformers import AutoTokenizer"
|
| 14 |
+
]
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"cell_type": "code",
|
| 18 |
+
"execution_count": 2,
|
| 19 |
+
"metadata": {},
|
| 20 |
+
"outputs": [],
|
| 21 |
+
"source": [
|
| 22 |
+
"# train_df=pd.read_csv('/DATA/sin-kaf/offenseval-tr-training-v1.tsv',sep='\\t')\n",
|
| 23 |
+
"# test_df=pd.read_csv('/DATA/sin-kaf/offenseval-tr-testset-v1.tsv',sep='\\t')\n",
|
| 24 |
+
"# augmented_df=pd.read_csv('augmented_data_offensive.csv')\n",
|
| 25 |
+
"# selin_df=pd.read_csv('/DATA/sin-kaf/selin_data.csv')"
|
| 26 |
+
]
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"cell_type": "code",
|
| 30 |
+
"execution_count": 3,
|
| 31 |
+
"metadata": {},
|
| 32 |
+
"outputs": [],
|
| 33 |
+
"source": [
|
| 34 |
+
"outliers_df=pd.read_csv('/DATA/sin-kaf/cluster_outliers.csv')\n",
|
| 35 |
+
"outliers_df=outliers_df.drop(['Unnamed: 0'], axis=1)\n",
|
| 36 |
+
"outliers_df['subtask_a'] = outliers_df['subtas_a']\n",
|
| 37 |
+
"outliers_df=outliers_df.drop(['subtas_a'], axis=1)\n"
|
| 38 |
+
]
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"cell_type": "code",
|
| 42 |
+
"execution_count": 4,
|
| 43 |
+
"metadata": {},
|
| 44 |
+
"outputs": [],
|
| 45 |
+
"source": [
|
| 46 |
+
"train_df=outliers_df"
|
| 47 |
+
]
|
| 48 |
+
},
|
| 49 |
+
{
|
| 50 |
+
"cell_type": "code",
|
| 51 |
+
"execution_count": 5,
|
| 52 |
+
"metadata": {},
|
| 53 |
+
"outputs": [],
|
| 54 |
+
"source": [
|
| 55 |
+
"# augmented_df=augmented_df.drop(['Unnamed: 0'], axis=1)\n",
|
| 56 |
+
"# augmented_df = augmented_df.dropna()\n",
|
| 57 |
+
"# train_df=pd.concat([train_df,augmented_df], axis=0)\n",
|
| 58 |
+
"# train_df=pd.concat([train_df,test_df], axis=0)\n",
|
| 59 |
+
"# train_df=train_df.drop(['id'], axis=1)\n",
|
| 60 |
+
"data=train_df['tweet'].tolist()\n",
|
| 61 |
+
"for i in range(len(data)):\n",
|
| 62 |
+
" data[i] = data[i].replace('@USER','')\n",
|
| 63 |
+
" data[i] = data[i].replace('#','')\n",
|
| 64 |
+
" data[i] = data[i].replace('$','')\n",
|
| 65 |
+
" data[i] = emoji.demojize(data[i])\n",
|
| 66 |
+
" \n",
|
| 67 |
+
"train_df['tweet'] = data\n",
|
| 68 |
+
"lab = LabelEncoder()\n",
|
| 69 |
+
"train_df['subtask_a'] = lab.fit_transform(train_df['subtask_a'])\n",
|
| 70 |
+
"df = train_df[train_df.subtask_a != 2]"
|
| 71 |
+
]
|
| 72 |
+
},
|
| 73 |
+
{
|
| 74 |
+
"cell_type": "code",
|
| 75 |
+
"execution_count": 6,
|
| 76 |
+
"metadata": {},
|
| 77 |
+
"outputs": [
|
| 78 |
+
{
|
| 79 |
+
"data": {
|
| 80 |
+
"text/html": [
|
| 81 |
+
"<div>\n",
|
| 82 |
+
"<style scoped>\n",
|
| 83 |
+
" .dataframe tbody tr th:only-of-type {\n",
|
| 84 |
+
" vertical-align: middle;\n",
|
| 85 |
+
" }\n",
|
| 86 |
+
"\n",
|
| 87 |
+
" .dataframe tbody tr th {\n",
|
| 88 |
+
" vertical-align: top;\n",
|
| 89 |
+
" }\n",
|
| 90 |
+
"\n",
|
| 91 |
+
" .dataframe thead th {\n",
|
| 92 |
+
" text-align: right;\n",
|
| 93 |
+
" }\n",
|
| 94 |
+
"</style>\n",
|
| 95 |
+
"<table border=\"1\" class=\"dataframe\">\n",
|
| 96 |
+
" <thead>\n",
|
| 97 |
+
" <tr style=\"text-align: right;\">\n",
|
| 98 |
+
" <th></th>\n",
|
| 99 |
+
" <th>tweet</th>\n",
|
| 100 |
+
" <th>subtask_a</th>\n",
|
| 101 |
+
" </tr>\n",
|
| 102 |
+
" </thead>\n",
|
| 103 |
+
" <tbody>\n",
|
| 104 |
+
" <tr>\n",
|
| 105 |
+
" <th>0</th>\n",
|
| 106 |
+
" <td>en güzel uyuyan insan ödülü jeon jungkook'a g...</td>\n",
|
| 107 |
+
" <td>0</td>\n",
|
| 108 |
+
" </tr>\n",
|
| 109 |
+
" <tr>\n",
|
| 110 |
+
" <th>1</th>\n",
|
| 111 |
+
" <td>Mekanı cennet olsun, saygılar sayın avukatımı...</td>\n",
|
| 112 |
+
" <td>0</td>\n",
|
| 113 |
+
" </tr>\n",
|
| 114 |
+
" <tr>\n",
|
| 115 |
+
" <th>2</th>\n",
|
| 116 |
+
" <td>Kızlar aranızda kas yığını beylere düşenler ol...</td>\n",
|
| 117 |
+
" <td>0</td>\n",
|
| 118 |
+
" </tr>\n",
|
| 119 |
+
" <tr>\n",
|
| 120 |
+
" <th>3</th>\n",
|
| 121 |
+
" <td>Biraz ders çalışayım. Tembellik ve uyku düşman...</td>\n",
|
| 122 |
+
" <td>0</td>\n",
|
| 123 |
+
" </tr>\n",
|
| 124 |
+
" <tr>\n",
|
| 125 |
+
" <th>4</th>\n",
|
| 126 |
+
" <td>Trezeguet yerine El Sharawy daha iyi olmaz mı</td>\n",
|
| 127 |
+
" <td>0</td>\n",
|
| 128 |
+
" </tr>\n",
|
| 129 |
+
" <tr>\n",
|
| 130 |
+
" <th>...</th>\n",
|
| 131 |
+
" <td>...</td>\n",
|
| 132 |
+
" <td>...</td>\n",
|
| 133 |
+
" </tr>\n",
|
| 134 |
+
" <tr>\n",
|
| 135 |
+
" <th>41177</th>\n",
|
| 136 |
+
" <td>Hil**adamlar kesinlikle kelimeleri anlamıyorla...</td>\n",
|
| 137 |
+
" <td>1</td>\n",
|
| 138 |
+
" </tr>\n",
|
| 139 |
+
" <tr>\n",
|
| 140 |
+
" <th>41178</th>\n",
|
| 141 |
+
" <td>Böyle piçlerin çok erken ölmemelerini ve çok f...</td>\n",
|
| 142 |
+
" <td>1</td>\n",
|
| 143 |
+
" </tr>\n",
|
| 144 |
+
" <tr>\n",
|
| 145 |
+
" <th>41179</th>\n",
|
| 146 |
+
" <td>Turgay denilen bu holigonda bir sorun yok, gur...</td>\n",
|
| 147 |
+
" <td>1</td>\n",
|
| 148 |
+
" </tr>\n",
|
| 149 |
+
" <tr>\n",
|
| 150 |
+
" <th>41180</th>\n",
|
| 151 |
+
" <td>Umarım ülkenin düşük zekadan kurtulması ilgile...</td>\n",
|
| 152 |
+
" <td>1</td>\n",
|
| 153 |
+
" </tr>\n",
|
| 154 |
+
" <tr>\n",
|
| 155 |
+
" <th>41181</th>\n",
|
| 156 |
+
" <td>CHP sandıkları bırakmaz, üzerine oturur, bir c...</td>\n",
|
| 157 |
+
" <td>1</td>\n",
|
| 158 |
+
" </tr>\n",
|
| 159 |
+
" </tbody>\n",
|
| 160 |
+
"</table>\n",
|
| 161 |
+
"<p>41182 rows × 2 columns</p>\n",
|
| 162 |
+
"</div>"
|
| 163 |
+
],
|
| 164 |
+
"text/plain": [
|
| 165 |
+
" tweet subtask_a\n",
|
| 166 |
+
"0 en güzel uyuyan insan ödülü jeon jungkook'a g... 0\n",
|
| 167 |
+
"1 Mekanı cennet olsun, saygılar sayın avukatımı... 0\n",
|
| 168 |
+
"2 Kızlar aranızda kas yığını beylere düşenler ol... 0\n",
|
| 169 |
+
"3 Biraz ders çalışayım. Tembellik ve uyku düşman... 0\n",
|
| 170 |
+
"4 Trezeguet yerine El Sharawy daha iyi olmaz mı 0\n",
|
| 171 |
+
"... ... ...\n",
|
| 172 |
+
"41177 Hil**adamlar kesinlikle kelimeleri anlamıyorla... 1\n",
|
| 173 |
+
"41178 Böyle piçlerin çok erken ölmemelerini ve çok f... 1\n",
|
| 174 |
+
"41179 Turgay denilen bu holigonda bir sorun yok, gur... 1\n",
|
| 175 |
+
"41180 Umarım ülkenin düşük zekadan kurtulması ilgile... 1\n",
|
| 176 |
+
"41181 CHP sandıkları bırakmaz, üzerine oturur, bir c... 1\n",
|
| 177 |
+
"\n",
|
| 178 |
+
"[41182 rows x 2 columns]"
|
| 179 |
+
]
|
| 180 |
+
},
|
| 181 |
+
"execution_count": 6,
|
| 182 |
+
"metadata": {},
|
| 183 |
+
"output_type": "execute_result"
|
| 184 |
+
}
|
| 185 |
+
],
|
| 186 |
+
"source": [
|
| 187 |
+
"train_df"
|
| 188 |
+
]
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"cell_type": "code",
|
| 192 |
+
"execution_count": 7,
|
| 193 |
+
"metadata": {},
|
| 194 |
+
"outputs": [],
|
| 195 |
+
"source": [
|
| 196 |
+
"# train_df=pd.concat([train_df,selin_df], axis=0)"
|
| 197 |
+
]
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"cell_type": "code",
|
| 201 |
+
"execution_count": 8,
|
| 202 |
+
"metadata": {},
|
| 203 |
+
"outputs": [
|
| 204 |
+
{
|
| 205 |
+
"data": {
|
| 206 |
+
"text/html": [
|
| 207 |
+
"<div>\n",
|
| 208 |
+
"<style scoped>\n",
|
| 209 |
+
" .dataframe tbody tr th:only-of-type {\n",
|
| 210 |
+
" vertical-align: middle;\n",
|
| 211 |
+
" }\n",
|
| 212 |
+
"\n",
|
| 213 |
+
" .dataframe tbody tr th {\n",
|
| 214 |
+
" vertical-align: top;\n",
|
| 215 |
+
" }\n",
|
| 216 |
+
"\n",
|
| 217 |
+
" .dataframe thead th {\n",
|
| 218 |
+
" text-align: right;\n",
|
| 219 |
+
" }\n",
|
| 220 |
+
"</style>\n",
|
| 221 |
+
"<table border=\"1\" class=\"dataframe\">\n",
|
| 222 |
+
" <thead>\n",
|
| 223 |
+
" <tr style=\"text-align: right;\">\n",
|
| 224 |
+
" <th></th>\n",
|
| 225 |
+
" <th>tweet</th>\n",
|
| 226 |
+
" <th>subtask_a</th>\n",
|
| 227 |
+
" </tr>\n",
|
| 228 |
+
" </thead>\n",
|
| 229 |
+
" <tbody>\n",
|
| 230 |
+
" <tr>\n",
|
| 231 |
+
" <th>0</th>\n",
|
| 232 |
+
" <td>en güzel uyuyan insan ödülü jeon jungkook'a g...</td>\n",
|
| 233 |
+
" <td>0</td>\n",
|
| 234 |
+
" </tr>\n",
|
| 235 |
+
" <tr>\n",
|
| 236 |
+
" <th>1</th>\n",
|
| 237 |
+
" <td>Mekanı cennet olsun, saygılar sayın avukatımı...</td>\n",
|
| 238 |
+
" <td>0</td>\n",
|
| 239 |
+
" </tr>\n",
|
| 240 |
+
" <tr>\n",
|
| 241 |
+
" <th>2</th>\n",
|
| 242 |
+
" <td>Kızlar aranızda kas yığını beylere düşenler ol...</td>\n",
|
| 243 |
+
" <td>0</td>\n",
|
| 244 |
+
" </tr>\n",
|
| 245 |
+
" <tr>\n",
|
| 246 |
+
" <th>3</th>\n",
|
| 247 |
+
" <td>Biraz ders çalışayım. Tembellik ve uyku düşman...</td>\n",
|
| 248 |
+
" <td>0</td>\n",
|
| 249 |
+
" </tr>\n",
|
| 250 |
+
" <tr>\n",
|
| 251 |
+
" <th>4</th>\n",
|
| 252 |
+
" <td>Trezeguet yerine El Sharawy daha iyi olmaz mı</td>\n",
|
| 253 |
+
" <td>0</td>\n",
|
| 254 |
+
" </tr>\n",
|
| 255 |
+
" <tr>\n",
|
| 256 |
+
" <th>...</th>\n",
|
| 257 |
+
" <td>...</td>\n",
|
| 258 |
+
" <td>...</td>\n",
|
| 259 |
+
" </tr>\n",
|
| 260 |
+
" <tr>\n",
|
| 261 |
+
" <th>41177</th>\n",
|
| 262 |
+
" <td>Hil**adamlar kesinlikle kelimeleri anlamıyorla...</td>\n",
|
| 263 |
+
" <td>1</td>\n",
|
| 264 |
+
" </tr>\n",
|
| 265 |
+
" <tr>\n",
|
| 266 |
+
" <th>41178</th>\n",
|
| 267 |
+
" <td>Böyle piçlerin çok erken ölmemelerini ve çok f...</td>\n",
|
| 268 |
+
" <td>1</td>\n",
|
| 269 |
+
" </tr>\n",
|
| 270 |
+
" <tr>\n",
|
| 271 |
+
" <th>41179</th>\n",
|
| 272 |
+
" <td>Turgay denilen bu holigonda bir sorun yok, gur...</td>\n",
|
| 273 |
+
" <td>1</td>\n",
|
| 274 |
+
" </tr>\n",
|
| 275 |
+
" <tr>\n",
|
| 276 |
+
" <th>41180</th>\n",
|
| 277 |
+
" <td>Umarım ülkenin düşük zekadan kurtulması ilgile...</td>\n",
|
| 278 |
+
" <td>1</td>\n",
|
| 279 |
+
" </tr>\n",
|
| 280 |
+
" <tr>\n",
|
| 281 |
+
" <th>41181</th>\n",
|
| 282 |
+
" <td>CHP sandıkları bırakmaz, üzerine oturur, bir c...</td>\n",
|
| 283 |
+
" <td>1</td>\n",
|
| 284 |
+
" </tr>\n",
|
| 285 |
+
" </tbody>\n",
|
| 286 |
+
"</table>\n",
|
| 287 |
+
"<p>41182 rows × 2 columns</p>\n",
|
| 288 |
+
"</div>"
|
| 289 |
+
],
|
| 290 |
+
"text/plain": [
|
| 291 |
+
" tweet subtask_a\n",
|
| 292 |
+
"0 en güzel uyuyan insan ödülü jeon jungkook'a g... 0\n",
|
| 293 |
+
"1 Mekanı cennet olsun, saygılar sayın avukatımı... 0\n",
|
| 294 |
+
"2 Kızlar aranızda kas yığını beylere düşenler ol... 0\n",
|
| 295 |
+
"3 Biraz ders çalışayım. Tembellik ve uyku düşman... 0\n",
|
| 296 |
+
"4 Trezeguet yerine El Sharawy daha iyi olmaz mı 0\n",
|
| 297 |
+
"... ... ...\n",
|
| 298 |
+
"41177 Hil**adamlar kesinlikle kelimeleri anlamıyorla... 1\n",
|
| 299 |
+
"41178 Böyle piçlerin çok erken ölmemelerini ve çok f... 1\n",
|
| 300 |
+
"41179 Turgay denilen bu holigonda bir sorun yok, gur... 1\n",
|
| 301 |
+
"41180 Umarım ülkenin düşük zekadan kurtulması ilgile... 1\n",
|
| 302 |
+
"41181 CHP sandıkları bırakmaz, üzerine oturur, bir c... 1\n",
|
| 303 |
+
"\n",
|
| 304 |
+
"[41182 rows x 2 columns]"
|
| 305 |
+
]
|
| 306 |
+
},
|
| 307 |
+
"execution_count": 8,
|
| 308 |
+
"metadata": {},
|
| 309 |
+
"output_type": "execute_result"
|
| 310 |
+
}
|
| 311 |
+
],
|
| 312 |
+
"source": [
|
| 313 |
+
"train_df"
|
| 314 |
+
]
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"cell_type": "code",
|
| 318 |
+
"execution_count": 9,
|
| 319 |
+
"metadata": {},
|
| 320 |
+
"outputs": [],
|
| 321 |
+
"source": [
|
| 322 |
+
"train_df = df.sample(frac = 0.7, random_state = 200)\n",
|
| 323 |
+
"df_2 = df.drop(train_df.index)\n",
|
| 324 |
+
"test_df = df_2.sample(frac = 0.15, random_state = 200)\n",
|
| 325 |
+
"val_df = df_2.drop(test_df.index)"
|
| 326 |
+
]
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"cell_type": "code",
|
| 330 |
+
"execution_count": 10,
|
| 331 |
+
"metadata": {},
|
| 332 |
+
"outputs": [],
|
| 333 |
+
"source": [
|
| 334 |
+
"text_train = train_df.tweet.values\n",
|
| 335 |
+
"label_train = train_df.subtask_a.values"
|
| 336 |
+
]
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"cell_type": "code",
|
| 340 |
+
"execution_count": 11,
|
| 341 |
+
"metadata": {},
|
| 342 |
+
"outputs": [],
|
| 343 |
+
"source": [
|
| 344 |
+
"text_test = test_df.tweet.values\n",
|
| 345 |
+
"label_test = test_df.subtask_a.values"
|
| 346 |
+
]
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"cell_type": "code",
|
| 350 |
+
"execution_count": 12,
|
| 351 |
+
"metadata": {},
|
| 352 |
+
"outputs": [],
|
| 353 |
+
"source": [
|
| 354 |
+
"text_val = val_df.tweet.values\n",
|
| 355 |
+
"label_val = val_df.subtask_a.values"
|
| 356 |
+
]
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"cell_type": "code",
|
| 360 |
+
"execution_count": 13,
|
| 361 |
+
"metadata": {},
|
| 362 |
+
"outputs": [],
|
| 363 |
+
"source": [
|
| 364 |
+
"from datasets.dataset_dict import DatasetDict\n",
|
| 365 |
+
"from datasets import Dataset\n",
|
| 366 |
+
"dataset={'train':Dataset.from_dict({'label':label_train,'text':text_train}),\n",
|
| 367 |
+
" 'val':Dataset.from_dict({'label':label_val,'text':text_val}),\n",
|
| 368 |
+
" 'test':Dataset.from_dict({'label':label_test,'text':text_test})\n",
|
| 369 |
+
" }\n",
|
| 370 |
+
"dataset = DatasetDict(dataset)"
|
| 371 |
+
]
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"cell_type": "code",
|
| 375 |
+
"execution_count": 14,
|
| 376 |
+
"metadata": {},
|
| 377 |
+
"outputs": [],
|
| 378 |
+
"source": [
|
| 379 |
+
"# tokenizer = AutoTokenizer.from_pretrained(\"dbmdz/bert-base-turkish-128k-uncased\")\n",
|
| 380 |
+
"# tokenizer = AutoTokenizer.from_pretrained(\"dbmdz/distilbert-base-turkish-cased\")\n",
|
| 381 |
+
"tokenizer = AutoTokenizer.from_pretrained(\"Overfit-GM/distilbert-base-turkish-cased-offensive\")\n",
|
| 382 |
+
"# tokenizer = AutoTokenizer.from_pretrained(\"Overfit-GM/distilbert-base-turkish-cased-offensive\",max_length=208,padding=\"max_length\",truncation=True,return_tensors=\"pt\",add_special_tokens=True,)\n",
|
| 383 |
+
"# tokenizer = AutoTokenizer.from_pretrained(\"stage_f/pretrain_mlm_distilbert-base-turkish-cased\")\n",
|
| 384 |
+
"def tokenize_function(examples):\n",
|
| 385 |
+
" return tokenizer(examples[\"text\"], padding=\"max_length\", truncation=True)"
|
| 386 |
+
]
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"cell_type": "code",
|
| 390 |
+
"execution_count": 15,
|
| 391 |
+
"metadata": {},
|
| 392 |
+
"outputs": [
|
| 393 |
+
{
|
| 394 |
+
"data": {
|
| 395 |
+
"application/vnd.jupyter.widget-view+json": {
|
| 396 |
+
"model_id": "5fba4c9671724e9a93d6ad14a1427345",
|
| 397 |
+
"version_major": 2,
|
| 398 |
+
"version_minor": 0
|
| 399 |
+
},
|
| 400 |
+
"text/plain": [
|
| 401 |
+
"Map: 0%| | 0/28827 [00:00<?, ? examples/s]"
|
| 402 |
+
]
|
| 403 |
+
},
|
| 404 |
+
"metadata": {},
|
| 405 |
+
"output_type": "display_data"
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"data": {
|
| 409 |
+
"application/vnd.jupyter.widget-view+json": {
|
| 410 |
+
"model_id": "2fff446f4f094d2fb66da549a49ad8a4",
|
| 411 |
+
"version_major": 2,
|
| 412 |
+
"version_minor": 0
|
| 413 |
+
},
|
| 414 |
+
"text/plain": [
|
| 415 |
+
"Map: 0%| | 0/10502 [00:00<?, ? examples/s]"
|
| 416 |
+
]
|
| 417 |
+
},
|
| 418 |
+
"metadata": {},
|
| 419 |
+
"output_type": "display_data"
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"data": {
|
| 423 |
+
"application/vnd.jupyter.widget-view+json": {
|
| 424 |
+
"model_id": "675f3b595b21489abaca01453c06db2c",
|
| 425 |
+
"version_major": 2,
|
| 426 |
+
"version_minor": 0
|
| 427 |
+
},
|
| 428 |
+
"text/plain": [
|
| 429 |
+
"Map: 0%| | 0/1853 [00:00<?, ? examples/s]"
|
| 430 |
+
]
|
| 431 |
+
},
|
| 432 |
+
"metadata": {},
|
| 433 |
+
"output_type": "display_data"
|
| 434 |
+
}
|
| 435 |
+
],
|
| 436 |
+
"source": [
|
| 437 |
+
"tokenized_datasets = dataset.map(tokenize_function, batched=True)"
|
| 438 |
+
]
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"cell_type": "code",
|
| 442 |
+
"execution_count": 16,
|
| 443 |
+
"metadata": {},
|
| 444 |
+
"outputs": [],
|
| 445 |
+
"source": [
|
| 446 |
+
"small_train_dataset = tokenized_datasets[\"train\"].shuffle(seed=42)\n",
|
| 447 |
+
"small_eval_dataset = tokenized_datasets[\"test\"].shuffle(seed=42)"
|
| 448 |
+
]
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"cell_type": "code",
|
| 452 |
+
"execution_count": 17,
|
| 453 |
+
"metadata": {},
|
| 454 |
+
"outputs": [
|
| 455 |
+
{
|
| 456 |
+
"data": {
|
| 457 |
+
"text/plain": [
|
| 458 |
+
"Dataset({\n",
|
| 459 |
+
" features: ['label', 'text', 'input_ids', 'attention_mask'],\n",
|
| 460 |
+
" num_rows: 28827\n",
|
| 461 |
+
"})"
|
| 462 |
+
]
|
| 463 |
+
},
|
| 464 |
+
"execution_count": 17,
|
| 465 |
+
"metadata": {},
|
| 466 |
+
"output_type": "execute_result"
|
| 467 |
+
}
|
| 468 |
+
],
|
| 469 |
+
"source": [
|
| 470 |
+
"small_train_dataset"
|
| 471 |
+
]
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"cell_type": "code",
|
| 475 |
+
"execution_count": 18,
|
| 476 |
+
"metadata": {},
|
| 477 |
+
"outputs": [
|
| 478 |
+
{
|
| 479 |
+
"data": {
|
| 480 |
+
"text/plain": [
|
| 481 |
+
"Dataset({\n",
|
| 482 |
+
" features: ['label', 'text', 'input_ids', 'attention_mask'],\n",
|
| 483 |
+
" num_rows: 1853\n",
|
| 484 |
+
"})"
|
| 485 |
+
]
|
| 486 |
+
},
|
| 487 |
+
"execution_count": 18,
|
| 488 |
+
"metadata": {},
|
| 489 |
+
"output_type": "execute_result"
|
| 490 |
+
}
|
| 491 |
+
],
|
| 492 |
+
"source": [
|
| 493 |
+
"small_eval_dataset"
|
| 494 |
+
]
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"cell_type": "code",
|
| 498 |
+
"execution_count": 19,
|
| 499 |
+
"metadata": {},
|
| 500 |
+
"outputs": [
|
| 501 |
+
{
|
| 502 |
+
"name": "stderr",
|
| 503 |
+
"output_type": "stream",
|
| 504 |
+
"text": [
|
| 505 |
+
"Some weights of DistilBertForSequenceClassification were not initialized from the model checkpoint at Overfit-GM/distilbert-base-turkish-cased-offensive and are newly initialized because the shapes did not match:\n",
|
| 506 |
+
"- classifier.weight: found shape torch.Size([5, 768]) in the checkpoint and torch.Size([2, 768]) in the model instantiated\n",
|
| 507 |
+
"- classifier.bias: found shape torch.Size([5]) in the checkpoint and torch.Size([2]) in the model instantiated\n",
|
| 508 |
+
"You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference.\n"
|
| 509 |
+
]
|
| 510 |
+
}
|
| 511 |
+
],
|
| 512 |
+
"source": [
|
| 513 |
+
"from transformers import AutoModelForSequenceClassification\n",
|
| 514 |
+
"\n",
|
| 515 |
+
"# model = AutoModelForSequenceClassification.from_pretrained(\"dbmdz/bert-base-turkish-128k-uncased\",num_labels = 2)\n",
|
| 516 |
+
"# model = AutoModelForSequenceClassification.from_pretrained(\"dbmdz/distilbert-base-turkish-cased\",num_labels = 2)\n",
|
| 517 |
+
"# model = AutoModelForSequenceClassification.from_pretrained(\"Overfit-GM/distilbert-base-turkish-cased-offensive\",num_labels = 2, ignore_mismatched_sizes=True)\n",
|
| 518 |
+
"model = AutoModelForSequenceClassification.from_pretrained(\"Overfit-GM/distilbert-base-turkish-cased-offensive\",num_labels = 2, ignore_mismatched_sizes=True)"
|
| 519 |
+
]
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"cell_type": "code",
|
| 523 |
+
"execution_count": 20,
|
| 524 |
+
"metadata": {},
|
| 525 |
+
"outputs": [],
|
| 526 |
+
"source": [
|
| 527 |
+
"from transformers import TrainingArguments\n",
|
| 528 |
+
"\n",
|
| 529 |
+
"training_args = TrainingArguments(output_dir=\"test_trainer\")"
|
| 530 |
+
]
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"cell_type": "code",
|
| 534 |
+
"execution_count": 21,
|
| 535 |
+
"metadata": {},
|
| 536 |
+
"outputs": [],
|
| 537 |
+
"source": [
|
| 538 |
+
"# import numpy as np\n",
|
| 539 |
+
"# import evaluate\n",
|
| 540 |
+
"\n",
|
| 541 |
+
"# # metric = evaluate.load(\"accuracy\")\n",
|
| 542 |
+
"# # confusion_matrix = evaluate.load(\"BucketHeadP65/confusion_matrix\")\n",
|
| 543 |
+
"# # metric = evaluate.combine([\"accuracy\", \"f1\", \"precision\", \"recall\", \"confusion_matrix\"])\n",
|
| 544 |
+
"# metric = evaluate.combine([\"accuracy\", \"f1\", \"precision\", \"recall\"])"
|
| 545 |
+
]
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"cell_type": "code",
|
| 549 |
+
"execution_count": 22,
|
| 550 |
+
"metadata": {},
|
| 551 |
+
"outputs": [],
|
| 552 |
+
"source": [
|
| 553 |
+
"import numpy as np\n",
|
| 554 |
+
"import evaluate\n",
|
| 555 |
+
"\n",
|
| 556 |
+
"metric = evaluate.combine([\"accuracy\", \"f1\", \"precision\", \"recall\"])\n",
|
| 557 |
+
"conf_matrix = evaluate.load(\"BucketHeadP65/confusion_matrix\")"
|
| 558 |
+
]
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"cell_type": "code",
|
| 562 |
+
"execution_count": 23,
|
| 563 |
+
"metadata": {},
|
| 564 |
+
"outputs": [],
|
| 565 |
+
"source": [
|
| 566 |
+
"def compute_metrics(eval_pred):\n",
|
| 567 |
+
" logits, labels = eval_pred\n",
|
| 568 |
+
" predictions = np.argmax(logits, axis=-1)\n",
|
| 569 |
+
" print(conf_matrix.compute(predictions=predictions, references=labels))\n",
|
| 570 |
+
" return metric.compute(predictions=predictions, references=labels)"
|
| 571 |
+
]
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"cell_type": "code",
|
| 575 |
+
"execution_count": 24,
|
| 576 |
+
"metadata": {},
|
| 577 |
+
"outputs": [],
|
| 578 |
+
"source": [
|
| 579 |
+
"from transformers import TrainingArguments, Trainer\n",
|
| 580 |
+
"from pytorch_lightning.loggers import TensorBoardLogger,MLFlowLogger\n",
|
| 581 |
+
"\n",
|
| 582 |
+
"training_args = TrainingArguments(output_dir=\"test_trainer\", evaluation_strategy=\"epoch\", num_train_epochs = 5, logging_dir ='TensorBoard',report_to ='mlflow')"
|
| 583 |
+
]
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"cell_type": "code",
|
| 587 |
+
"execution_count": 25,
|
| 588 |
+
"metadata": {},
|
| 589 |
+
"outputs": [],
|
| 590 |
+
"source": [
|
| 591 |
+
"trainer = Trainer(\n",
|
| 592 |
+
" model=model,\n",
|
| 593 |
+
" args=training_args,\n",
|
| 594 |
+
" train_dataset=small_train_dataset,\n",
|
| 595 |
+
" eval_dataset=small_eval_dataset,\n",
|
| 596 |
+
" compute_metrics=compute_metrics,\n",
|
| 597 |
+
")"
|
| 598 |
+
]
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"cell_type": "code",
|
| 602 |
+
"execution_count": 26,
|
| 603 |
+
"metadata": {},
|
| 604 |
+
"outputs": [
|
| 605 |
+
{
|
| 606 |
+
"name": "stderr",
|
| 607 |
+
"output_type": "stream",
|
| 608 |
+
"text": [
|
| 609 |
+
"huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...\n",
|
| 610 |
+
"To disable this warning, you can either:\n",
|
| 611 |
+
"\t- Avoid using `tokenizers` before the fork if possible\n",
|
| 612 |
+
"\t- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)\n",
|
| 613 |
+
"huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...\n",
|
| 614 |
+
"To disable this warning, you can either:\n",
|
| 615 |
+
"\t- Avoid using `tokenizers` before the fork if possible\n",
|
| 616 |
+
"\t- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)\n",
|
| 617 |
+
"huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...\n",
|
| 618 |
+
"To disable this warning, you can either:\n",
|
| 619 |
+
"\t- Avoid using `tokenizers` before the fork if possible\n",
|
| 620 |
+
"\t- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)\n"
|
| 621 |
+
]
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"data": {
|
| 625 |
+
"application/vnd.jupyter.widget-view+json": {
|
| 626 |
+
"model_id": "a38121a009be4a0f90e30fc9c0cf49ed",
|
| 627 |
+
"version_major": 2,
|
| 628 |
+
"version_minor": 0
|
| 629 |
+
},
|
| 630 |
+
"text/plain": [
|
| 631 |
+
" 0%| | 0/18020 [00:00<?, ?it/s]"
|
| 632 |
+
]
|
| 633 |
+
},
|
| 634 |
+
"metadata": {},
|
| 635 |
+
"output_type": "display_data"
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"name": "stdout",
|
| 639 |
+
"output_type": "stream",
|
| 640 |
+
"text": [
|
| 641 |
+
"{'loss': 0.4638, 'learning_rate': 4.86126526082131e-05, 'epoch': 0.14}\n",
|
| 642 |
+
"{'loss': 0.3886, 'learning_rate': 4.72253052164262e-05, 'epoch': 0.28}\n",
|
| 643 |
+
"{'loss': 0.3893, 'learning_rate': 4.583795782463929e-05, 'epoch': 0.42}\n",
|
| 644 |
+
"{'loss': 0.3594, 'learning_rate': 4.445061043285239e-05, 'epoch': 0.55}\n",
|
| 645 |
+
"{'loss': 0.3547, 'learning_rate': 4.306326304106548e-05, 'epoch': 0.69}\n",
|
| 646 |
+
"{'loss': 0.3384, 'learning_rate': 4.167591564927858e-05, 'epoch': 0.83}\n",
|
| 647 |
+
"{'loss': 0.3498, 'learning_rate': 4.028856825749168e-05, 'epoch': 0.97}\n"
|
| 648 |
+
]
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"data": {
|
| 652 |
+
"application/vnd.jupyter.widget-view+json": {
|
| 653 |
+
"model_id": "94ab139e1ebb482da2111517ad5a3a78",
|
| 654 |
+
"version_major": 2,
|
| 655 |
+
"version_minor": 0
|
| 656 |
+
},
|
| 657 |
+
"text/plain": [
|
| 658 |
+
" 0%| | 0/232 [00:00<?, ?it/s]"
|
| 659 |
+
]
|
| 660 |
+
},
|
| 661 |
+
"metadata": {},
|
| 662 |
+
"output_type": "display_data"
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"name": "stdout",
|
| 666 |
+
"output_type": "stream",
|
| 667 |
+
"text": [
|
| 668 |
+
"{'confusion_matrix': array([[966, 90],\n",
|
| 669 |
+
" [118, 679]])}\n",
|
| 670 |
+
"{'eval_loss': 0.28741681575775146, 'eval_accuracy': 0.8877495952509444, 'eval_f1': 0.8671775223499362, 'eval_precision': 0.88296488946684, 'eval_recall': 0.8519447929736512, 'eval_runtime': 11.4928, 'eval_samples_per_second': 161.231, 'eval_steps_per_second': 20.186, 'epoch': 1.0}\n",
|
| 671 |
+
"{'loss': 0.2449, 'learning_rate': 3.890122086570477e-05, 'epoch': 1.11}\n",
|
| 672 |
+
"{'loss': 0.2178, 'learning_rate': 3.751387347391787e-05, 'epoch': 1.25}\n",
|
| 673 |
+
"{'loss': 0.2431, 'learning_rate': 3.612652608213097e-05, 'epoch': 1.39}\n",
|
| 674 |
+
"{'loss': 0.2261, 'learning_rate': 3.4739178690344064e-05, 'epoch': 1.53}\n",
|
| 675 |
+
"{'loss': 0.2365, 'learning_rate': 3.3351831298557165e-05, 'epoch': 1.66}\n",
|
| 676 |
+
"{'loss': 0.2169, 'learning_rate': 3.196448390677026e-05, 'epoch': 1.8}\n",
|
| 677 |
+
"{'loss': 0.222, 'learning_rate': 3.0577136514983354e-05, 'epoch': 1.94}\n"
|
| 678 |
+
]
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"data": {
|
| 682 |
+
"application/vnd.jupyter.widget-view+json": {
|
| 683 |
+
"model_id": "063c47c6cae0467194d4c0827e67c277",
|
| 684 |
+
"version_major": 2,
|
| 685 |
+
"version_minor": 0
|
| 686 |
+
},
|
| 687 |
+
"text/plain": [
|
| 688 |
+
" 0%| | 0/232 [00:00<?, ?it/s]"
|
| 689 |
+
]
|
| 690 |
+
},
|
| 691 |
+
"metadata": {},
|
| 692 |
+
"output_type": "display_data"
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"name": "stdout",
|
| 696 |
+
"output_type": "stream",
|
| 697 |
+
"text": [
|
| 698 |
+
"{'confusion_matrix': array([[900, 156],\n",
|
| 699 |
+
" [ 76, 721]])}\n",
|
| 700 |
+
"{'eval_loss': 0.47509443759918213, 'eval_accuracy': 0.8747976254722072, 'eval_f1': 0.8614097968936678, 'eval_precision': 0.82212086659065, 'eval_recall': 0.904642409033877, 'eval_runtime': 11.6203, 'eval_samples_per_second': 159.462, 'eval_steps_per_second': 19.965, 'epoch': 2.0}\n",
|
| 701 |
+
"{'loss': 0.146, 'learning_rate': 2.918978912319645e-05, 'epoch': 2.08}\n",
|
| 702 |
+
"{'loss': 0.1163, 'learning_rate': 2.7802441731409544e-05, 'epoch': 2.22}\n",
|
| 703 |
+
"{'loss': 0.1008, 'learning_rate': 2.641509433962264e-05, 'epoch': 2.36}\n",
|
| 704 |
+
"{'loss': 0.0967, 'learning_rate': 2.502774694783574e-05, 'epoch': 2.5}\n",
|
| 705 |
+
"{'loss': 0.1456, 'learning_rate': 2.3640399556048838e-05, 'epoch': 2.64}\n",
|
| 706 |
+
"{'loss': 0.1178, 'learning_rate': 2.2253052164261932e-05, 'epoch': 2.77}\n",
|
| 707 |
+
"{'loss': 0.1155, 'learning_rate': 2.0865704772475027e-05, 'epoch': 2.91}\n"
|
| 708 |
+
]
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"data": {
|
| 712 |
+
"application/vnd.jupyter.widget-view+json": {
|
| 713 |
+
"model_id": "4fa52dfbbae54cde8c627a237bed51bc",
|
| 714 |
+
"version_major": 2,
|
| 715 |
+
"version_minor": 0
|
| 716 |
+
},
|
| 717 |
+
"text/plain": [
|
| 718 |
+
" 0%| | 0/232 [00:00<?, ?it/s]"
|
| 719 |
+
]
|
| 720 |
+
},
|
| 721 |
+
"metadata": {},
|
| 722 |
+
"output_type": "display_data"
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"name": "stdout",
|
| 726 |
+
"output_type": "stream",
|
| 727 |
+
"text": [
|
| 728 |
+
"{'confusion_matrix': array([[954, 102],\n",
|
| 729 |
+
" [106, 691]])}\n",
|
| 730 |
+
"{'eval_loss': 0.5530020594596863, 'eval_accuracy': 0.8877495952509444, 'eval_f1': 0.8691823899371071, 'eval_precision': 0.8713745271122321, 'eval_recall': 0.8670012547051443, 'eval_runtime': 11.6026, 'eval_samples_per_second': 159.706, 'eval_steps_per_second': 19.996, 'epoch': 3.0}\n",
|
| 731 |
+
"{'loss': 0.0879, 'learning_rate': 1.9478357380688125e-05, 'epoch': 3.05}\n",
|
| 732 |
+
"{'loss': 0.0351, 'learning_rate': 1.8091009988901223e-05, 'epoch': 3.19}\n",
|
| 733 |
+
"{'loss': 0.0501, 'learning_rate': 1.670366259711432e-05, 'epoch': 3.33}\n",
|
| 734 |
+
"{'loss': 0.0425, 'learning_rate': 1.5316315205327412e-05, 'epoch': 3.47}\n",
|
| 735 |
+
"{'loss': 0.0564, 'learning_rate': 1.392896781354051e-05, 'epoch': 3.61}\n",
|
| 736 |
+
"{'loss': 0.05, 'learning_rate': 1.2541620421753608e-05, 'epoch': 3.75}\n",
|
| 737 |
+
"{'loss': 0.034, 'learning_rate': 1.1154273029966705e-05, 'epoch': 3.88}\n"
|
| 738 |
+
]
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"data": {
|
| 742 |
+
"application/vnd.jupyter.widget-view+json": {
|
| 743 |
+
"model_id": "a9b754cd0e7641cb8d8023f28bc32a06",
|
| 744 |
+
"version_major": 2,
|
| 745 |
+
"version_minor": 0
|
| 746 |
+
},
|
| 747 |
+
"text/plain": [
|
| 748 |
+
" 0%| | 0/232 [00:00<?, ?it/s]"
|
| 749 |
+
]
|
| 750 |
+
},
|
| 751 |
+
"metadata": {},
|
| 752 |
+
"output_type": "display_data"
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"name": "stdout",
|
| 756 |
+
"output_type": "stream",
|
| 757 |
+
"text": [
|
| 758 |
+
"{'confusion_matrix': array([[966, 90],\n",
|
| 759 |
+
" [109, 688]])}\n",
|
| 760 |
+
"{'eval_loss': 0.824292778968811, 'eval_accuracy': 0.8926065839179709, 'eval_f1': 0.8736507936507937, 'eval_precision': 0.884318766066838, 'eval_recall': 0.863237139272271, 'eval_runtime': 11.6185, 'eval_samples_per_second': 159.487, 'eval_steps_per_second': 19.968, 'epoch': 4.0}\n",
|
| 761 |
+
"{'loss': 0.0354, 'learning_rate': 9.766925638179801e-06, 'epoch': 4.02}\n",
|
| 762 |
+
"{'loss': 0.0165, 'learning_rate': 8.379578246392897e-06, 'epoch': 4.16}\n",
|
| 763 |
+
"{'loss': 0.0119, 'learning_rate': 6.992230854605994e-06, 'epoch': 4.3}\n",
|
| 764 |
+
"{'loss': 0.0145, 'learning_rate': 5.60488346281909e-06, 'epoch': 4.44}\n",
|
| 765 |
+
"{'loss': 0.0169, 'learning_rate': 4.217536071032187e-06, 'epoch': 4.58}\n",
|
| 766 |
+
"{'loss': 0.0132, 'learning_rate': 2.830188679245283e-06, 'epoch': 4.72}\n",
|
| 767 |
+
"{'loss': 0.0232, 'learning_rate': 1.4428412874583796e-06, 'epoch': 4.86}\n",
|
| 768 |
+
"{'loss': 0.0189, 'learning_rate': 5.549389567147614e-08, 'epoch': 4.99}\n"
|
| 769 |
+
]
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"data": {
|
| 773 |
+
"application/vnd.jupyter.widget-view+json": {
|
| 774 |
+
"model_id": "e66e5b59c6ba42ae9939f55dcda3c877",
|
| 775 |
+
"version_major": 2,
|
| 776 |
+
"version_minor": 0
|
| 777 |
+
},
|
| 778 |
+
"text/plain": [
|
| 779 |
+
" 0%| | 0/232 [00:00<?, ?it/s]"
|
| 780 |
+
]
|
| 781 |
+
},
|
| 782 |
+
"metadata": {},
|
| 783 |
+
"output_type": "display_data"
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"name": "stdout",
|
| 787 |
+
"output_type": "stream",
|
| 788 |
+
"text": [
|
| 789 |
+
"{'confusion_matrix': array([[955, 101],\n",
|
| 790 |
+
" [111, 686]])}\n",
|
| 791 |
+
"{'eval_loss': 0.937654972076416, 'eval_accuracy': 0.8855909336211549, 'eval_f1': 0.8661616161616161, 'eval_precision': 0.8716645489199492, 'eval_recall': 0.8607277289836889, 'eval_runtime': 11.5644, 'eval_samples_per_second': 160.233, 'eval_steps_per_second': 20.062, 'epoch': 5.0}\n",
|
| 792 |
+
"{'train_runtime': 3027.4521, 'train_samples_per_second': 47.609, 'train_steps_per_second': 5.952, 'train_loss': 0.15528733040680712, 'epoch': 5.0}\n"
|
| 793 |
+
]
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"data": {
|
| 797 |
+
"text/plain": [
|
| 798 |
+
"TrainOutput(global_step=18020, training_loss=0.15528733040680712, metrics={'train_runtime': 3027.4521, 'train_samples_per_second': 47.609, 'train_steps_per_second': 5.952, 'train_loss': 0.15528733040680712, 'epoch': 5.0})"
|
| 799 |
+
]
|
| 800 |
+
},
|
| 801 |
+
"execution_count": 26,
|
| 802 |
+
"metadata": {},
|
| 803 |
+
"output_type": "execute_result"
|
| 804 |
+
}
|
| 805 |
+
],
|
| 806 |
+
"source": [
|
| 807 |
+
"trainer.train()"
|
| 808 |
+
]
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"cell_type": "markdown",
|
| 812 |
+
"metadata": {},
|
| 813 |
+
"source": [
|
| 814 |
+
"# best case"
|
| 815 |
+
]
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"cell_type": "code",
|
| 819 |
+
"execution_count": 25,
|
| 820 |
+
"metadata": {},
|
| 821 |
+
"outputs": [
|
| 822 |
+
{
|
| 823 |
+
"name": "stderr",
|
| 824 |
+
"output_type": "stream",
|
| 825 |
+
"text": [
|
| 826 |
+
"huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...\n",
|
| 827 |
+
"To disable this warning, you can either:\n",
|
| 828 |
+
"\t- Avoid using `tokenizers` before the fork if possible\n",
|
| 829 |
+
"\t- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)\n",
|
| 830 |
+
"huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...\n",
|
| 831 |
+
"To disable this warning, you can either:\n",
|
| 832 |
+
"\t- Avoid using `tokenizers` before the fork if possible\n",
|
| 833 |
+
"\t- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)\n",
|
| 834 |
+
"huggingface/tokenizers: The current process just got forked, after parallelism has already been used. Disabling parallelism to avoid deadlocks...\n",
|
| 835 |
+
"To disable this warning, you can either:\n",
|
| 836 |
+
"\t- Avoid using `tokenizers` before the fork if possible\n",
|
| 837 |
+
"\t- Explicitly set the environment variable TOKENIZERS_PARALLELISM=(true | false)\n"
|
| 838 |
+
]
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"data": {
|
| 842 |
+
"application/vnd.jupyter.widget-view+json": {
|
| 843 |
+
"model_id": "4620503cb22c41a582c44a3d17fac2f6",
|
| 844 |
+
"version_major": 2,
|
| 845 |
+
"version_minor": 0
|
| 846 |
+
},
|
| 847 |
+
"text/plain": [
|
| 848 |
+
" 0%| | 0/18825 [00:00<?, ?it/s]"
|
| 849 |
+
]
|
| 850 |
+
},
|
| 851 |
+
"metadata": {},
|
| 852 |
+
"output_type": "display_data"
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"name": "stdout",
|
| 856 |
+
"output_type": "stream",
|
| 857 |
+
"text": [
|
| 858 |
+
"{'loss': 0.4623, 'learning_rate': 4.867197875166003e-05, 'epoch': 0.13}\n",
|
| 859 |
+
"{'loss': 0.3955, 'learning_rate': 4.734395750332006e-05, 'epoch': 0.27}\n",
|
| 860 |
+
"{'loss': 0.3695, 'learning_rate': 4.601593625498008e-05, 'epoch': 0.4}\n",
|
| 861 |
+
"{'loss': 0.368, 'learning_rate': 4.4687915006640105e-05, 'epoch': 0.53}\n",
|
| 862 |
+
"{'loss': 0.3418, 'learning_rate': 4.335989375830013e-05, 'epoch': 0.66}\n",
|
| 863 |
+
"{'loss': 0.3519, 'learning_rate': 4.203187250996016e-05, 'epoch': 0.8}\n",
|
| 864 |
+
"{'loss': 0.3418, 'learning_rate': 4.070385126162019e-05, 'epoch': 0.93}\n"
|
| 865 |
+
]
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"data": {
|
| 869 |
+
"application/vnd.jupyter.widget-view+json": {
|
| 870 |
+
"model_id": "c81779b9a7eb43cfa29966957f13ec31",
|
| 871 |
+
"version_major": 2,
|
| 872 |
+
"version_minor": 0
|
| 873 |
+
},
|
| 874 |
+
"text/plain": [
|
| 875 |
+
" 0%| | 0/242 [00:00<?, ?it/s]"
|
| 876 |
+
]
|
| 877 |
+
},
|
| 878 |
+
"metadata": {},
|
| 879 |
+
"output_type": "display_data"
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"name": "stdout",
|
| 883 |
+
"output_type": "stream",
|
| 884 |
+
"text": [
|
| 885 |
+
"{'eval_loss': 0.2548353374004364, 'eval_accuracy': 0.9013429752066116, 'eval_f1': 0.8737607402511566, 'eval_precision': 0.9218967921896792, 'eval_recall': 0.8304020100502513, 'eval_runtime': 12.1488, 'eval_samples_per_second': 159.357, 'eval_steps_per_second': 19.92, 'epoch': 1.0}\n",
|
| 886 |
+
"{'loss': 0.2884, 'learning_rate': 3.9375830013280215e-05, 'epoch': 1.06}\n",
|
| 887 |
+
"{'loss': 0.2136, 'learning_rate': 3.804780876494024e-05, 'epoch': 1.2}\n",
|
| 888 |
+
"{'loss': 0.2422, 'learning_rate': 3.671978751660027e-05, 'epoch': 1.33}\n",
|
| 889 |
+
"{'loss': 0.2105, 'learning_rate': 3.53917662682603e-05, 'epoch': 1.46}\n",
|
| 890 |
+
"{'loss': 0.2203, 'learning_rate': 3.406374501992032e-05, 'epoch': 1.59}\n",
|
| 891 |
+
"{'loss': 0.2455, 'learning_rate': 3.2735723771580345e-05, 'epoch': 1.73}\n",
|
| 892 |
+
"{'loss': 0.2282, 'learning_rate': 3.140770252324037e-05, 'epoch': 1.86}\n",
|
| 893 |
+
"{'loss': 0.2328, 'learning_rate': 3.00796812749004e-05, 'epoch': 1.99}\n"
|
| 894 |
+
]
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"data": {
|
| 898 |
+
"application/vnd.jupyter.widget-view+json": {
|
| 899 |
+
"model_id": "f83c5030d5c34216ba6422f2c22858ba",
|
| 900 |
+
"version_major": 2,
|
| 901 |
+
"version_minor": 0
|
| 902 |
+
},
|
| 903 |
+
"text/plain": [
|
| 904 |
+
" 0%| | 0/242 [00:00<?, ?it/s]"
|
| 905 |
+
]
|
| 906 |
+
},
|
| 907 |
+
"metadata": {},
|
| 908 |
+
"output_type": "display_data"
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"name": "stdout",
|
| 912 |
+
"output_type": "stream",
|
| 913 |
+
"text": [
|
| 914 |
+
"{'eval_loss': 0.4118729829788208, 'eval_accuracy': 0.8982438016528925, 'eval_f1': 0.8763339610797238, 'eval_precision': 0.875784190715182, 'eval_recall': 0.8768844221105527, 'eval_runtime': 12.1691, 'eval_samples_per_second': 159.092, 'eval_steps_per_second': 19.886, 'epoch': 2.0}\n",
|
| 915 |
+
"{'loss': 0.1086, 'learning_rate': 2.8751660026560427e-05, 'epoch': 2.12}\n",
|
| 916 |
+
"{'loss': 0.1137, 'learning_rate': 2.742363877822045e-05, 'epoch': 2.26}\n",
|
| 917 |
+
"{'loss': 0.1058, 'learning_rate': 2.609561752988048e-05, 'epoch': 2.39}\n",
|
| 918 |
+
"{'loss': 0.1073, 'learning_rate': 2.4767596281540506e-05, 'epoch': 2.52}\n",
|
| 919 |
+
"{'loss': 0.0953, 'learning_rate': 2.3439575033200534e-05, 'epoch': 2.66}\n",
|
| 920 |
+
"{'loss': 0.1066, 'learning_rate': 2.2111553784860558e-05, 'epoch': 2.79}\n",
|
| 921 |
+
"{'loss': 0.1152, 'learning_rate': 2.0783532536520585e-05, 'epoch': 2.92}\n"
|
| 922 |
+
]
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"data": {
|
| 926 |
+
"application/vnd.jupyter.widget-view+json": {
|
| 927 |
+
"model_id": "3c4d464cb3a340d4aa4f6a1a8e4d95b9",
|
| 928 |
+
"version_major": 2,
|
| 929 |
+
"version_minor": 0
|
| 930 |
+
},
|
| 931 |
+
"text/plain": [
|
| 932 |
+
" 0%| | 0/242 [00:00<?, ?it/s]"
|
| 933 |
+
]
|
| 934 |
+
},
|
| 935 |
+
"metadata": {},
|
| 936 |
+
"output_type": "display_data"
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"name": "stdout",
|
| 940 |
+
"output_type": "stream",
|
| 941 |
+
"text": [
|
| 942 |
+
"{'eval_loss': 0.4992543160915375, 'eval_accuracy': 0.9039256198347108, 'eval_f1': 0.8831658291457286, 'eval_precision': 0.8831658291457286, 'eval_recall': 0.8831658291457286, 'eval_runtime': 12.145, 'eval_samples_per_second': 159.407, 'eval_steps_per_second': 19.926, 'epoch': 3.0}\n",
|
| 943 |
+
"{'loss': 0.0761, 'learning_rate': 1.9455511288180613e-05, 'epoch': 3.05}\n",
|
| 944 |
+
"{'loss': 0.0434, 'learning_rate': 1.812749003984064e-05, 'epoch': 3.19}\n",
|
| 945 |
+
"{'loss': 0.0395, 'learning_rate': 1.6799468791500664e-05, 'epoch': 3.32}\n",
|
| 946 |
+
"{'loss': 0.0516, 'learning_rate': 1.547144754316069e-05, 'epoch': 3.45}\n",
|
| 947 |
+
"{'loss': 0.0344, 'learning_rate': 1.4143426294820719e-05, 'epoch': 3.59}\n",
|
| 948 |
+
"{'loss': 0.0588, 'learning_rate': 1.2815405046480745e-05, 'epoch': 3.72}\n",
|
| 949 |
+
"{'loss': 0.0323, 'learning_rate': 1.148738379814077e-05, 'epoch': 3.85}\n",
|
| 950 |
+
"{'loss': 0.0574, 'learning_rate': 1.0159362549800798e-05, 'epoch': 3.98}\n"
|
| 951 |
+
]
|
| 952 |
+
},
|
| 953 |
+
{
|
| 954 |
+
"data": {
|
| 955 |
+
"application/vnd.jupyter.widget-view+json": {
|
| 956 |
+
"model_id": "bf0675bd947c472bb221d755dc55a219",
|
| 957 |
+
"version_major": 2,
|
| 958 |
+
"version_minor": 0
|
| 959 |
+
},
|
| 960 |
+
"text/plain": [
|
| 961 |
+
" 0%| | 0/242 [00:00<?, ?it/s]"
|
| 962 |
+
]
|
| 963 |
+
},
|
| 964 |
+
"metadata": {},
|
| 965 |
+
"output_type": "display_data"
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"name": "stdout",
|
| 969 |
+
"output_type": "stream",
|
| 970 |
+
"text": [
|
| 971 |
+
"{'eval_loss': 0.6084339618682861, 'eval_accuracy': 0.9121900826446281, 'eval_f1': 0.8933500627352573, 'eval_precision': 0.8922305764411027, 'eval_recall': 0.8944723618090452, 'eval_runtime': 11.9875, 'eval_samples_per_second': 161.502, 'eval_steps_per_second': 20.188, 'epoch': 4.0}\n",
|
| 972 |
+
"{'loss': 0.0175, 'learning_rate': 8.831341301460823e-06, 'epoch': 4.12}\n",
|
| 973 |
+
"{'loss': 0.0248, 'learning_rate': 7.503320053120851e-06, 'epoch': 4.25}\n",
|
| 974 |
+
"{'loss': 0.0212, 'learning_rate': 6.175298804780877e-06, 'epoch': 4.38}\n",
|
| 975 |
+
"{'loss': 0.0215, 'learning_rate': 4.847277556440903e-06, 'epoch': 4.52}\n",
|
| 976 |
+
"{'loss': 0.0216, 'learning_rate': 3.51925630810093e-06, 'epoch': 4.65}\n",
|
| 977 |
+
"{'loss': 0.0169, 'learning_rate': 2.1912350597609563e-06, 'epoch': 4.78}\n",
|
| 978 |
+
"{'loss': 0.0199, 'learning_rate': 8.632138114209828e-07, 'epoch': 4.91}\n"
|
| 979 |
+
]
|
| 980 |
+
},
|
| 981 |
+
{
|
| 982 |
+
"data": {
|
| 983 |
+
"application/vnd.jupyter.widget-view+json": {
|
| 984 |
+
"model_id": "0ac0cee28031479d9721321ec9c949a4",
|
| 985 |
+
"version_major": 2,
|
| 986 |
+
"version_minor": 0
|
| 987 |
+
},
|
| 988 |
+
"text/plain": [
|
| 989 |
+
" 0%| | 0/242 [00:00<?, ?it/s]"
|
| 990 |
+
]
|
| 991 |
+
},
|
| 992 |
+
"metadata": {},
|
| 993 |
+
"output_type": "display_data"
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"name": "stdout",
|
| 997 |
+
"output_type": "stream",
|
| 998 |
+
"text": [
|
| 999 |
+
"{'eval_loss': 0.6909418106079102, 'eval_accuracy': 0.9158057851239669, 'eval_f1': 0.8963763509218055, 'eval_precision': 0.9073359073359073, 'eval_recall': 0.885678391959799, 'eval_runtime': 12.1798, 'eval_samples_per_second': 158.952, 'eval_steps_per_second': 19.869, 'epoch': 5.0}\n",
|
| 1000 |
+
"{'train_runtime': 3197.4084, 'train_samples_per_second': 47.101, 'train_steps_per_second': 5.888, 'train_loss': 0.15457879885892628, 'epoch': 5.0}\n"
|
| 1001 |
+
]
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"data": {
|
| 1005 |
+
"text/plain": [
|
| 1006 |
+
"TrainOutput(global_step=18825, training_loss=0.15457879885892628, metrics={'train_runtime': 3197.4084, 'train_samples_per_second': 47.101, 'train_steps_per_second': 5.888, 'train_loss': 0.15457879885892628, 'epoch': 5.0})"
|
| 1007 |
+
]
|
| 1008 |
+
},
|
| 1009 |
+
"execution_count": 25,
|
| 1010 |
+
"metadata": {},
|
| 1011 |
+
"output_type": "execute_result"
|
| 1012 |
+
}
|
| 1013 |
+
],
|
| 1014 |
+
"source": [
|
| 1015 |
+
"trainer.train()"
|
| 1016 |
+
]
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"cell_type": "markdown",
|
| 1020 |
+
"metadata": {},
|
| 1021 |
+
"source": [
|
| 1022 |
+
"# load model"
|
| 1023 |
+
]
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"cell_type": "code",
|
| 1027 |
+
"execution_count": 20,
|
| 1028 |
+
"metadata": {},
|
| 1029 |
+
"outputs": [],
|
| 1030 |
+
"source": [
|
| 1031 |
+
"import torch"
|
| 1032 |
+
]
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"cell_type": "code",
|
| 1036 |
+
"execution_count": 21,
|
| 1037 |
+
"metadata": {},
|
| 1038 |
+
"outputs": [],
|
| 1039 |
+
"source": [
|
| 1040 |
+
"model = AutoModelForSequenceClassification.from_pretrained('/DATA/sin-kaf/test_trainer/checkpoint-16000')"
|
| 1041 |
+
]
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"cell_type": "code",
|
| 1045 |
+
"execution_count": 30,
|
| 1046 |
+
"metadata": {},
|
| 1047 |
+
"outputs": [
|
| 1048 |
+
{
|
| 1049 |
+
"name": "stdout",
|
| 1050 |
+
"output_type": "stream",
|
| 1051 |
+
"text": [
|
| 1052 |
+
"Original: güzel kızz\n",
|
| 1053 |
+
"Token IDs: tensor([[ 2, 2639, 2889, 1050, 3, 0, 0, 0, 0, 0, 0, 0,\n",
|
| 1054 |
+
" 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
|
| 1055 |
+
" 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
|
| 1056 |
+
" 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
|
| 1057 |
+
" 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
|
| 1058 |
+
" 0, 0, 0, 0]])\n",
|
| 1059 |
+
"Token IDs: tensor([[1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
|
| 1060 |
+
" 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
|
| 1061 |
+
" 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]])\n"
|
| 1062 |
+
]
|
| 1063 |
+
}
|
| 1064 |
+
],
|
| 1065 |
+
"source": [
|
| 1066 |
+
"sent = 'güzel kızz'\n",
|
| 1067 |
+
"input_ids = []\n",
|
| 1068 |
+
"attention_masks = []\n",
|
| 1069 |
+
"\n",
|
| 1070 |
+
"encoded_dict = tokenizer.encode_plus(\n",
|
| 1071 |
+
" sent,\n",
|
| 1072 |
+
" add_special_tokens = True,\n",
|
| 1073 |
+
" max_length = 64,\n",
|
| 1074 |
+
" pad_to_max_length = True,\n",
|
| 1075 |
+
" return_attention_mask = True,\n",
|
| 1076 |
+
" return_tensors = 'pt',\n",
|
| 1077 |
+
" )\n",
|
| 1078 |
+
"\n",
|
| 1079 |
+
"\n",
|
| 1080 |
+
"input_ids = encoded_dict['input_ids']\n",
|
| 1081 |
+
"attention_masks = encoded_dict['attention_mask']\n",
|
| 1082 |
+
"\n",
|
| 1083 |
+
"\n",
|
| 1084 |
+
"input_ids = torch.cat([input_ids], dim=0)\n",
|
| 1085 |
+
"input_mask = torch.cat([attention_masks], dim=0)\n",
|
| 1086 |
+
"\n",
|
| 1087 |
+
"\n",
|
| 1088 |
+
"\n",
|
| 1089 |
+
"print('Original: ', sent)\n",
|
| 1090 |
+
"print('Token IDs:', input_ids)\n",
|
| 1091 |
+
"print('Token IDs:', input_mask)"
|
| 1092 |
+
]
|
| 1093 |
+
},
|
| 1094 |
+
{
|
| 1095 |
+
"cell_type": "code",
|
| 1096 |
+
"execution_count": 31,
|
| 1097 |
+
"metadata": {},
|
| 1098 |
+
"outputs": [],
|
| 1099 |
+
"source": [
|
| 1100 |
+
"outputs = model(input_ids, input_mask)"
|
| 1101 |
+
]
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"cell_type": "code",
|
| 1105 |
+
"execution_count": 32,
|
| 1106 |
+
"metadata": {},
|
| 1107 |
+
"outputs": [
|
| 1108 |
+
{
|
| 1109 |
+
"data": {
|
| 1110 |
+
"text/plain": [
|
| 1111 |
+
"SequenceClassifierOutput(loss=None, logits=tensor([[ 3.6835, -3.6147]], grad_fn=<AddmmBackward0>), hidden_states=None, attentions=None)"
|
| 1112 |
+
]
|
| 1113 |
+
},
|
| 1114 |
+
"execution_count": 32,
|
| 1115 |
+
"metadata": {},
|
| 1116 |
+
"output_type": "execute_result"
|
| 1117 |
+
}
|
| 1118 |
+
],
|
| 1119 |
+
"source": [
|
| 1120 |
+
"outputs"
|
| 1121 |
+
]
|
| 1122 |
+
},
|
| 1123 |
+
{
|
| 1124 |
+
"cell_type": "code",
|
| 1125 |
+
"execution_count": 33,
|
| 1126 |
+
"metadata": {},
|
| 1127 |
+
"outputs": [
|
| 1128 |
+
{
|
| 1129 |
+
"data": {
|
| 1130 |
+
"text/plain": [
|
| 1131 |
+
"tensor(0)"
|
| 1132 |
+
]
|
| 1133 |
+
},
|
| 1134 |
+
"execution_count": 33,
|
| 1135 |
+
"metadata": {},
|
| 1136 |
+
"output_type": "execute_result"
|
| 1137 |
+
}
|
| 1138 |
+
],
|
| 1139 |
+
"source": [
|
| 1140 |
+
"torch.argmax(outputs['logits'])"
|
| 1141 |
+
]
|
| 1142 |
+
}
|
| 1143 |
+
],
|
| 1144 |
+
"metadata": {
|
| 1145 |
+
"kernelspec": {
|
| 1146 |
+
"display_name": "dlenv",
|
| 1147 |
+
"language": "python",
|
| 1148 |
+
"name": "python3"
|
| 1149 |
+
},
|
| 1150 |
+
"language_info": {
|
| 1151 |
+
"codemirror_mode": {
|
| 1152 |
+
"name": "ipython",
|
| 1153 |
+
"version": 3
|
| 1154 |
+
},
|
| 1155 |
+
"file_extension": ".py",
|
| 1156 |
+
"mimetype": "text/x-python",
|
| 1157 |
+
"name": "python",
|
| 1158 |
+
"nbconvert_exporter": "python",
|
| 1159 |
+
"pygments_lexer": "ipython3",
|
| 1160 |
+
"version": "3.9.17"
|
| 1161 |
+
}
|
| 1162 |
+
},
|
| 1163 |
+
"nbformat": 4,
|
| 1164 |
+
"nbformat_minor": 2
|
| 1165 |
+
}
|