Update README.md
Browse files
README.md
CHANGED
|
@@ -6,4 +6,35 @@ language:
|
|
| 6 |
base_model:
|
| 7 |
- google-t5/t5-small
|
| 8 |
pipeline_tag: summarization
|
| 9 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 6 |
base_model:
|
| 7 |
- google-t5/t5-small
|
| 8 |
pipeline_tag: summarization
|
| 9 |
+
---
|
| 10 |
+
|
| 11 |
+
|
| 12 |
+
# AML Text Summarization T5 Model
|
| 13 |
+
|
| 14 |
+
This is a text summarization model based on the T5-Small architecture, developed as part of the Advanced Machine Learning course at the University of Bremen.
|
| 15 |
+
|
| 16 |
+
## Model Description
|
| 17 |
+
|
| 18 |
+
This model is fine-tuned on the CNN/Daily Mail dataset for abstractive text summarization. It uses the T5-Small (Text-To-Text Transfer Transformer) architecture.
|
| 19 |
+
|
| 20 |
+
## Usage
|
| 21 |
+
|
| 22 |
+
```
|
| 23 |
+
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
|
| 24 |
+
|
| 25 |
+
tokenizer = AutoTokenizer.from_pretrained("s0urin/aml-text-summarization-t5")
|
| 26 |
+
model = AutoModelForSeq2SeqLM.from_pretrained("s0urin/aml-text-summarization-t5")
|
| 27 |
+
|
| 28 |
+
text = "Your long text here..."
|
| 29 |
+
inputs = tokenizer("summarize: " + text, return_tensors="pt", max_length=512, truncation=True)
|
| 30 |
+
outputs = model.generate(inputs.input_ids, max_length=150, min_length=40, length_penalty=2.0, num_beams=4, early_stopping=True)
|
| 31 |
+
summary = tokenizer.decode(outputs, skip_special_tokens=True)
|
| 32 |
+
|
| 33 |
+
print(summary)
|
| 34 |
+
```
|
| 35 |
+
|
| 36 |
+
|
| 37 |
+
## Authors
|
| 38 |
+
|
| 39 |
+
- Sourin Kumar Pal
|
| 40 |
+
- Jassim Hameed Ayobkhan
|