File size: 2,970 Bytes
1ad74ed f08f52d |
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 |
---
datasets:
- squad_v1
license: mit
---
# LONGFORMER-BASE-4096 fine-tuned on SQuAD v1
This is longformer-base-4096 model fine-tuned on SQuAD v1 dataset for question answering task.
[Longformer](https://arxiv.org/abs/2004.05150) model created by Iz Beltagy, Matthew E. Peters, Arman Coha from AllenAI. As the paper explains it
> `Longformer` is a BERT-like model for long documents.
The pre-trained model can handle sequences with upto 4096 tokens.
## Model Training
This model was trained on google colab v100 GPU. You can find the fine-tuning colab here [](https://colab.research.google.com/drive/1zEl5D-DdkBKva-DdreVOmN0hrAfzKG1o?usp=sharing).
Few things to keep in mind while training longformer for QA task,
by default longformer uses sliding-window local attention on all tokens. But For QA, all question tokens should have global attention. For more details on this please refer the paper. The `LongformerForQuestionAnswering` model automatically does that for you. To allow it to do that
1. The input sequence must have three sep tokens, i.e the sequence should be encoded like this
` <s> question</s></s> context</s>`. If you encode the question and answer as a input pair, then the tokenizer already takes care of that, you shouldn't worry about it.
2. `input_ids` should always be a batch of examples.
## Results
|Metric | # Value |
|-------------|---------|
| Exact Match | 85.1466 |
| F1 | 91.5415 |
## Model in Action 馃殌
```python
import torch
from transformers import AutoTokenizer, AutoModelForQuestionAnswering,
tokenizer = AutoTokenizer.from_pretrained("valhalla/longformer-base-4096-finetuned-squadv1")
model = AutoModelForQuestionAnswering.from_pretrained("valhalla/longformer-base-4096-finetuned-squadv1")
text = "Huggingface has democratized NLP. Huge thanks to Huggingface for this."
question = "What has Huggingface done ?"
encoding = tokenizer(question, text, return_tensors="pt")
input_ids = encoding["input_ids"]
# default is local attention everywhere
# the forward method will automatically set global attention on question tokens
attention_mask = encoding["attention_mask"]
start_scores, end_scores = model(input_ids, attention_mask=attention_mask)
all_tokens = tokenizer.convert_ids_to_tokens(input_ids[0].tolist())
answer_tokens = all_tokens[torch.argmax(start_scores) :torch.argmax(end_scores)+1]
answer = tokenizer.decode(tokenizer.convert_tokens_to_ids(answer_tokens))
# output => democratized NLP
```
The `LongformerForQuestionAnswering` isn't yet supported in `pipeline` . I'll update this card once the support has been added.
> Created with 鉂わ笍 by Suraj Patil [](https://github.com/patil-suraj/)
[](https://twitter.com/psuraj28)
|