File size: 4,625 Bytes
e120304 0646cf8 a133ae8 0646cf8 88852b0 e120304 0646cf8 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 |
---
language:
- tr
inference:
parameters:
max_new_tokens: 32
arXiv: 2403.01308
library_name: transformers
pipeline_tag: text2text-generation
license: cc-by-nc-sa-4.0
datasets:
- vngrs-ai/vngrs-web-corpus
---
# VBART Model Card
## Model Description
VBART is the first sequence-to-sequence LLM pre-trained on Turkish corpora from scratch on a large scale. It was pre-trained by VNGRS in February 2023.
The model is capable of conditional text generation tasks such as text summarization, paraphrasing, and title generation when fine-tuned.
It outperforms its multilingual counterparts, albeit being much smaller than other implementations.
VBART-XLarge is created by adding extra Transformer layers between the layers of VBART-Large. Hence it was able to transfer learned weights from the smaller model while doublings its number of layers.
VBART-XLarge improves the results compared to VBART-Large albeit in small margins.
This repository contains fine-tuned TensorFlow and Safetensors weights of VBART for title generation from news spot task.
- **Developed by:** [VNGRS-AI](https://vngrs.com/ai/)
- **Model type:** Transformer encoder-decoder based on mBART architecture
- **Language(s) (NLP):** Turkish
- **License:** CC BY-NC-SA 4.0
- **Finetuned from:** VBART-XLarge
- **Paper:** [arXiv](https://arxiv.org/abs/2403.01308)
## How to Get Started with the Model
```python
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
tokenizer = AutoTokenizer.from_pretrained("vngrs-ai/VBART-XLarge-Title-Generation-from-Spot",
model_input_names=['input_ids', 'attention_mask'])
# Uncomment the device_map kwarg and delete the closing bracket to use model for inference on GPU
model = AutoModelForSeq2SeqLM.from_pretrained("vngrs-ai/VBART-XLarge-Title-Generation-from-Spot")#, device_map="auto")
input_text="..."
token_input = tokenizer(input_text, return_tensors="pt")#.to('cuda')
outputs = model.generate(**token_input)
print(tokenizer.decode(outputs[0]))
```
## Training Details
### Training Data
The base model is pre-trained on [vngrs-web-corpus](https://huggingface.co/datasets/vngrs-ai/vngrs-web-corpus). It is curated by cleaning and filtering Turkish parts of [OSCAR-2201](https://huggingface.co/datasets/oscar-corpus/OSCAR-2201) and [mC4](https://huggingface.co/datasets/mc4) datasets. These datasets consist of documents of unstructured web crawl data. More information about the dataset can be found on their respective pages. Data is filtered using a set of heuristics and certain rules, explained in the appendix of our [paper](https://arxiv.org/abs/2403.01308).
The fine-tuning dataset is the Turkish sections of [MLSum](https://huggingface.co/datasets/mlsum), [TRNews](https://huggingface.co/datasets/batubayk/TR-News) and [XLSum](https://huggingface.co/datasets/csebuetnlp/xlsum) datasets.
### Limitations
This model is fine-tuned for title generation tasks. It is not intended to be used in any other case and can not be fine-tuned to any other task with full performance of the base model. It is also not guaranteed that this model will work without specified prompts.
### Training Procedure
Pre-trained for 8 days and for a total of 84B tokens. Finally, finetuned for 15 epochs.
#### Hardware
- **GPUs**: 8 x Nvidia A100-80 GB
#### Software
- TensorFlow
#### Hyperparameters
##### Pretraining
- **Training regime:** fp16 mixed precision
- **Training objective**: Sentence permutation and span masking (using mask lengths sampled from Poisson distribution λ=3.5, masking 30% of tokens)
- **Optimizer** : Adam optimizer (β1 = 0.9, β2 = 0.98, Ɛ = 1e-6)
- **Scheduler**: Custom scheduler from the original Transformers paper (20,000 warm-up steps)
- **Weight Initialization**: Model Enlargement from VBART-Large. See the related section in the [paper](https://arxiv.org/abs/2403.01308) for the details.
- **Dropout**: 0.1 (dropped to 0.05 and then to 0 in the last 80K and 80k steps, respectively)
- **Initial Learning rate**: 5e-6
- **Training tokens**: 84B
##### Fine-tuning
- **Training regime:** fp16 mixed precision
- **Optimizer** : Adam optimizer (β1 = 0.9, β2 = 0.98, Ɛ = 1e-6)
- **Scheduler**: Linear decay scheduler
- **Dropout**: 0.1
- **Learning rate**: 5e-6
- **Fine-tune epochs**: 15
#### Metrics
## Citation
```
@article{turker2024vbart,
title={VBART: The Turkish LLM},
author={Turker, Meliksah and Ari, Erdi and Han, Aydin},
journal={arXiv preprint arXiv:2403.01308},
year={2024}
}
``` |