Update README.md
Browse files
README.md
CHANGED
|
@@ -1,6 +1,19 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
# GoEmotions Fine-tuned Model
|
| 2 |
|
| 3 |
-
์ด ๋ชจ๋ธ์ GoEmotions
|
| 4 |
|
| 5 |
## ๋ชจ๋ธ ์ ๋ณด
|
| 6 |
- **Base Model**: Mango-Juice/trpg_mlm
|
|
@@ -45,8 +58,8 @@ from transformers import AutoTokenizer, AutoModelForSequenceClassification
|
|
| 45 |
import torch
|
| 46 |
|
| 47 |
# ๋ชจ๋ธ๊ณผ ํ ํฌ๋์ด์ ๋ก๋
|
| 48 |
-
tokenizer = AutoTokenizer.from_pretrained("
|
| 49 |
-
model = AutoModelForSequenceClassification.from_pretrained("
|
| 50 |
|
| 51 |
# ์ถ๋ก
|
| 52 |
def predict_emotions(text):
|
|
@@ -72,4 +85,4 @@ print(emotions)
|
|
| 72 |
- ๋ฐ์ดํฐ ์ฆ๊ฐ: ํ๋ผํ๋ ์ด์ง ๋ฐ ์ญ๋ฒ์ญ ๊ธฐ๋ฐ ์ค๋ฒ์ํ๋ง
|
| 73 |
- ์์ค ํจ์: Focal Loss with Label Smoothing
|
| 74 |
- ์ตํฐ๋ง์ด์ : AdamW
|
| 75 |
-
- ์ค์ผ์ค๋ฌ: ReduceLROnPlateau
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: mit
|
| 3 |
+
datasets:
|
| 4 |
+
- mrm8488/goemotions
|
| 5 |
+
language:
|
| 6 |
+
- en
|
| 7 |
+
metrics:
|
| 8 |
+
- accuracy
|
| 9 |
+
- f1
|
| 10 |
+
base_model:
|
| 11 |
+
- Mango-Juice/trpg_mlm
|
| 12 |
+
- microsoft/deberta-v3-large
|
| 13 |
+
---
|
| 14 |
# GoEmotions Fine-tuned Model
|
| 15 |
|
| 16 |
+
์ด ๋ชจ๋ธ์ GoEmotions ๋ฐ์ดํฐ์
๋ฐ TRPG ๋ฌธ์ฅ์ผ๋ก ํ๋ จ๋ ๋ค์ค ๊ฐ์ ๋ถ๋ฅ ๋ชจ๋ธ์
๋๋ค.
|
| 17 |
|
| 18 |
## ๋ชจ๋ธ ์ ๋ณด
|
| 19 |
- **Base Model**: Mango-Juice/trpg_mlm
|
|
|
|
| 58 |
import torch
|
| 59 |
|
| 60 |
# ๋ชจ๋ธ๊ณผ ํ ํฌ๋์ด์ ๋ก๋
|
| 61 |
+
tokenizer = AutoTokenizer.from_pretrained("Mango-Juice/trpg_emotion_classification")
|
| 62 |
+
model = AutoModelForSequenceClassification.from_pretrained("Mango-Juice/trpg_emotion_classification")
|
| 63 |
|
| 64 |
# ์ถ๋ก
|
| 65 |
def predict_emotions(text):
|
|
|
|
| 85 |
- ๋ฐ์ดํฐ ์ฆ๊ฐ: ํ๋ผํ๋ ์ด์ง ๋ฐ ์ญ๋ฒ์ญ ๊ธฐ๋ฐ ์ค๋ฒ์ํ๋ง
|
| 86 |
- ์์ค ํจ์: Focal Loss with Label Smoothing
|
| 87 |
- ์ตํฐ๋ง์ด์ : AdamW
|
| 88 |
+
- ์ค์ผ์ค๋ฌ: ReduceLROnPlateau
|