Update README.md
Browse files
README.md
CHANGED
|
@@ -5,7 +5,7 @@ tags:
|
|
| 5 |
- Document Question Answering
|
| 6 |
- Document Visual Question Answering
|
| 7 |
datasets:
|
| 8 |
-
-
|
| 9 |
language:
|
| 10 |
- en
|
| 11 |
---
|
|
@@ -19,18 +19,6 @@ This model was used as a baseline in [Hierarchical multimodal transformers for M
|
|
| 19 |
- Results on the MP-DocVQA dataset are reported in Table 2.
|
| 20 |
- Training hyperparameters can be found in Table 8 of Appendix D.
|
| 21 |
|
| 22 |
-
## Model results
|
| 23 |
-
|
| 24 |
-
Extended experimentation can be found in Table 2 of [Hierarchical multimodal transformers for Multi-Page DocVQA](https://arxiv.org/pdf/2212.05935.pdf).
|
| 25 |
-
You can also check the live leaderboard at the [RRC Portal](https://rrc.cvc.uab.es/?ch=17&com=evaluation&task=4).
|
| 26 |
-
| Model | HF name | ANLS | APPA |
|
| 27 |
-
|-----------------------------------------------------------------------------------|:--------------------------------------|:-------------:|:---------:|
|
| 28 |
-
| [Bert-large](https://huggingface.co/rubentito/bert-large-mpdocvqa) | rubentito/bert-large-mpdocvqa | 0.4183 | 51.6177 |
|
| 29 |
-
| [Longformer-base](https://huggingface.co/rubentito/longformer-base-mpdocvqa) | rubentito/longformer-base-mpdocvqa | 0.5287 | 71.1696 |
|
| 30 |
-
| [BigBird ITC base](https://huggingface.co/rubentito/bigbird-base-itc-mpdocvqa) | rubentito/bigbird-base-itc-mpdocvqa | 0.4929 | 67.5433 |
|
| 31 |
-
| [LayoutLMv3 base](https://huggingface.co/rubentito/layoutlmv3-base-mpdocvqa) | rubentito/layoutlmv3-base-mpdocvqa | 0.4538 | 51.9426 |
|
| 32 |
-
| [**T5 base**](https://huggingface.co/rubentito/t5-base-mpdocvqa) | rubentito/t5-base-mpdocvqa | 0.5050 | 0.0000 |
|
| 33 |
-
| Hi-VT5 | TBA | 0.6201 | 79.23 |
|
| 34 |
|
| 35 |
## How to use
|
| 36 |
|
|
@@ -52,6 +40,20 @@ output = self.model.generate(**encoding)
|
|
| 52 |
answer = tokenizer.decode(output['sequences'], skip_special_tokens=True)
|
| 53 |
```
|
| 54 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 55 |
## BibTeX entry
|
| 56 |
|
| 57 |
```tex
|
|
|
|
| 5 |
- Document Question Answering
|
| 6 |
- Document Visual Question Answering
|
| 7 |
datasets:
|
| 8 |
+
- rubentito/mp-docvqa
|
| 9 |
language:
|
| 10 |
- en
|
| 11 |
---
|
|
|
|
| 19 |
- Results on the MP-DocVQA dataset are reported in Table 2.
|
| 20 |
- Training hyperparameters can be found in Table 8 of Appendix D.
|
| 21 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 22 |
|
| 23 |
## How to use
|
| 24 |
|
|
|
|
| 40 |
answer = tokenizer.decode(output['sequences'], skip_special_tokens=True)
|
| 41 |
```
|
| 42 |
|
| 43 |
+
|
| 44 |
+
## Model results
|
| 45 |
+
|
| 46 |
+
Extended experimentation can be found in Table 2 of [Hierarchical multimodal transformers for Multi-Page DocVQA](https://arxiv.org/pdf/2212.05935.pdf).
|
| 47 |
+
You can also check the live leaderboard at the [RRC Portal](https://rrc.cvc.uab.es/?ch=17&com=evaluation&task=4).
|
| 48 |
+
| Model | HF name | ANLS | APPA |
|
| 49 |
+
|-----------------------------------------------------------------------------------|:--------------------------------------|:-------------:|:---------:|
|
| 50 |
+
| [Bert-large](https://huggingface.co/rubentito/bert-large-mpdocvqa) | rubentito/bert-large-mpdocvqa | 0.4183 | 51.6177 |
|
| 51 |
+
| [Longformer-base](https://huggingface.co/rubentito/longformer-base-mpdocvqa) | rubentito/longformer-base-mpdocvqa | 0.5287 | 71.1696 |
|
| 52 |
+
| [BigBird ITC base](https://huggingface.co/rubentito/bigbird-base-itc-mpdocvqa) | rubentito/bigbird-base-itc-mpdocvqa | 0.4929 | 67.5433 |
|
| 53 |
+
| [LayoutLMv3 base](https://huggingface.co/rubentito/layoutlmv3-base-mpdocvqa) | rubentito/layoutlmv3-base-mpdocvqa | 0.4538 | 51.9426 |
|
| 54 |
+
| [**T5 base**](https://huggingface.co/rubentito/t5-base-mpdocvqa) | rubentito/t5-base-mpdocvqa | 0.5050 | 0.0000 |
|
| 55 |
+
| Hi-VT5 | TBA | 0.6201 | 79.23 |
|
| 56 |
+
|
| 57 |
## BibTeX entry
|
| 58 |
|
| 59 |
```tex
|