File size: 3,021 Bytes
2a24fd5 a2bc315 2a24fd5 555a2f9 2315e52 52a6fb1 2a24fd5 a2bc315 2a24fd5 555a2f9 2a24fd5 db8270b 2a24fd5 bef4615 2a24fd5 bef4615 2a24fd5 355c49c a2bc315 3702fb4 0a80dc6 2e9ac70 d6c0822 3501870 7997684 b0a60c6 e2d95a8 db8270b 2a24fd5 555a2f9 |
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 |
---
license: apache-2.0
tags:
- generated_from_keras_callback
model-index:
- name: AnaniyaX/decision-distilbert-uncased
results: []
datasets:
- textvqa
- squad
widget:
- text: 'What does the sign says'
example_title: 'Visual Question Example 1'
- text: 'What does string theory talks about'
example_title: 'Textual Question Example 1'
---
<!-- This model card has been generated automatically according to the information Keras had access to. You should
probably proofread and complete it, then remove this comment. -->
# AnaniyaX/decision-distilbert-uncased
This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on textvqa and squad.
It achieves the following results on the evaluation set:
- Train Loss: 0.0097
- Train Accuracy: 0.9976
- Epoch: 9
## Model description
The Text-Visual Question Classifier is a Hugging Face model that can classify questions as either text-based or visual-based.
It uses a natural language processing and techniques to analyze the question and determine its type.
The model has been trained on a large dataset of questions labeled as either text-based or visual-based, and has achieved high accuracy in identifying the correct type of question.
## Intended uses & limitations
#### Applications
This model can be used in various applications such as chatbots, virtual assistants, search engines, and recommendation systems. For example, it can help chatbots to provide more accurate responses by understanding the type of question being asked. It can also help search engines to retrieve more relevant results by filtering out irrelevant content based on the type of question.
#### Limitations:
The model may not perform well on questions that are ambiguous or have multiple interpretations. It may also be biased towards certain types of questions based on the training data.
## Training and evaluation data
More information needed
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- optimizer: {'name': 'Adam', 'weight_decay': None, 'clipnorm': None, 'global_clipnorm': None, 'clipvalue': None, 'use_ema': False, 'ema_momentum': 0.99, 'ema_overwrite_frequency': None, 'jit_compile': True, 'is_legacy_optimizer': False, 'learning_rate': 2e-06, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-07, 'amsgrad': False}
- training_precision: float32
### Training results
| Train Loss | Train Accuracy | Epoch |
|:----------:|:--------------:|:-----:|
| 0.1914 | 0.9444 | 0 |
| 0.0711 | 0.9768 | 1 |
| 0.0531 | 0.9826 | 2 |
| 0.0427 | 0.9868 | 3 |
| 0.0330 | 0.9904 | 4 |
| 0.0264 | 0.9923 | 5 |
| 0.0195 | 0.9947 | 6 |
| 0.0149 | 0.9960 | 7 |
| 0.0123 | 0.9965 | 8 |
| 0.0097 | 0.9976 | 9 |
### Framework versions
- Transformers 4.27.2
- TensorFlow 2.11.0
- Datasets 2.10.1
- Tokenizers 0.13.2 |