Downloads · 30 days
2
14% of all-time downloads
t5-small-custom
kimhyeonseo/t5-small-custom
t5-small-custom is a machine learning model from kimhyeonseo. Use it for the machine learning task on the model card, and read the license before you ship it in a product.
fine-tuned version of t5small for summarization trained on CNN/Daily mail dataset
All-time downloads
14
Public
Parameters
60.5M
243 MB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors242 MB · 99%
At a glance
- Model type
- t5
- Access
- Public
- Created
- Sep 20, 2024
- Updated
- Sep 20, 2024
- SHA
- a6093e78
Model card
From the Hugging Face model README
Model Card for t5_small Summarization Model
fine-tuned version of t5_small for summarization
Model Details
trained on CNN/Daily mail dataset
Training Data
CNN/Daily mail dataset
Training Procedure
- Learning Rate: 2e-5
- Epochs: 1
- Batch Size: 4
How to Use
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
tokenizer = AutoTokenizer.from_pretrained("t5_small")
model = AutoModelForSeq2SeqLM.from_pretrained("t5_small")
input_text = "The movie was fantastic with a gripping storyline!"
inputs = tokenizer.encode(input_text, return_tensors="pt")
outputs = model(inputs)
print(outputs.logits)
Evaluation
eval_rouge1: 32.13 eval_rouge2: 11.85 eval_rougeL: 23.13 eval_bleu1: 29.29 eval_bleu2: 10.02 eval_bleu4: 3.83
Limitations
bad performance
Ethical Considerations
model has bias by cnn dataset
- Type
- t5
- Created
- Sep 20, 2024
- Updated
- Sep 20, 2024