Knowledge Distillation in AutoMM¶
Pretrained foundation models are becoming increasingly large. However, these models are difficult to deploy due to limited resources available in deployment scenarios. To benefit from large models under this constraint, you transfer the knowledge from the large-scale teacher models to the student model, with knowledge distillation. In this way, the small student model can be practically deployed under real-world scenarios, while the performance will be better than training the student model from scratch thanks to the teacher.
In this tutorial, we introduce how to adopt MultiModalPredictor for knowledge distillation. For the purpose of demonstration, we use the Question-answering NLI dataset,
which comprises 104,743 question, answer pairs sampled from question answering datasets. We will demonstrate how to use a large model to guide the learning and improve the performance of a small model in AutoGluon.
Load Dataset¶
The Question-answering NLI dataset contains
sentence pairs in English. In the label column, 0 means that the sentence is not related to the question and 1 means that the sentence is related to the question.
import datasets
from datasets import load_dataset
datasets.logging.disable_progress_bar()
dataset = load_dataset("glue", "qnli")
dataset['train']
Dataset({
features: ['question', 'sentence', 'label', 'idx'],
num_rows: 104743
})
from sklearn.model_selection import train_test_split
train_valid_df = dataset["train"].to_pandas()[["question", "sentence", "label"]].sample(1000, random_state=123)
train_df, valid_df = train_test_split(train_valid_df, test_size=0.2, random_state=123)
test_df = dataset["validation"].to_pandas()[["question", "sentence", "label"]].sample(1000, random_state=123)
Load the Teacher Model¶
In our example, we will directly load a teacher model with the google/bert_uncased_L-12_H-768_A-12 backbone that has been trained on QNLI and distill it into a student model with the google/bert_uncased_L-6_H-768_A-12 backbone.
!wget --quiet https://automl-mm-bench.s3.amazonaws.com/unit-tests/distillation_sample_teacher.zip -O distillation_sample_teacher.zip
!unzip -q -o distillation_sample_teacher.zip -d .
from autogluon.multimodal import MultiModalPredictor
teacher_predictor = MultiModalPredictor.load("ag_distillation_sample_teacher/")
Start to upgrade the previous configuration trained by AutoMM version=0.5.3b20221108.
/home/ci/opt/venv/lib/python3.10/site-packages/sklearn/base.py:376: InconsistentVersionWarning: Trying to unpickle estimator LabelEncoder from version 1.0.2 when using version 1.4.0. This might lead to breaking code or invalid results. Use at your own risk. For more info please refer to:
https://scikit-learn.org/stable/model_persistence.html#security-maintainability-limitations
warnings.warn(
/home/ci/opt/venv/lib/python3.10/site-packages/sklearn/base.py:376: InconsistentVersionWarning: Trying to unpickle estimator StandardScaler from version 1.0.2 when using version 1.4.0. This might lead to breaking code or invalid results. Use at your own risk. For more info please refer to:
https://scikit-learn.org/stable/model_persistence.html#security-maintainability-limitations
warnings.warn(
Load pretrained checkpoint: /home/ci/autogluon/docs/tutorials/multimodal/advanced_topics/ag_distillation_sample_teacher/model.ckpt
Distill to Student¶
Training the student model is straight forward. You may just add the teacher_predictor argument when calling .fit().
Internally, the student will be trained by matching the prediction/feature map from the teacher. It can perform better than
directly finetuning the student.
student_predictor = MultiModalPredictor(label="label")
student_predictor.fit(
train_df,
tuning_data=valid_df,
teacher_predictor=teacher_predictor,
hyperparameters={
"model.hf_text.checkpoint_name": "google/bert_uncased_L-6_H-768_A-12",
"optimization.max_epochs": 2,
}
)
No path specified. Models will be saved in: "AutogluonModels/ag-20240716_231802"
=================== System Info ===================
AutoGluon Version: 1.1.1b20240716
Python Version: 3.10.13
Operating System: Linux
Platform Machine: x86_64
Platform Version: #1 SMP Fri May 17 18:07:48 UTC 2024
CPU Count: 8
Pytorch Version: 2.3.1+cu121
CUDA Version: 12.1
Memory Avail: 27.49 GB / 30.95 GB (88.8%)
Disk Space Avail: 177.28 GB / 255.99 GB (69.3%)
===================================================
AutoGluon infers your prediction problem is: 'binary' (because only two unique label-values observed).
2 unique label values: [0, 1]
If 'binary' is not the correct problem_type, please manually specify the problem_type parameter during Predictor init (You may specify problem_type as one of: ['binary', 'multiclass', 'regression', 'quantile'])
AutoMM starts to create your model. ✨✨✨
To track the learning progress, you can open a terminal and launch Tensorboard:
```shell
# Assume you have installed tensorboard
tensorboard --logdir /home/ci/autogluon/docs/tutorials/multimodal/advanced_topics/AutogluonModels/ag-20240716_231802
```
Seed set to 0
/home/ci/opt/venv/lib/python3.10/site-packages/huggingface_hub/file_download.py:1132: FutureWarning: `resume_download` is deprecated and will be removed in version 1.0.0. Downloads always resume when possible. If you want to force a new download, use `force_download=True`.
warnings.warn(
/home/ci/opt/venv/lib/python3.10/site-packages/lightning/pytorch/utilities/parsing.py:208: Attribute 'softmax_regression_loss_func' is an instance of `nn.Module` and is already saved during checkpointing. It is recommended to ignore them using `self.save_hyperparameters(ignore=['softmax_regression_loss_func'])`.
GPU Count: 1
GPU Count to be Used: 1
GPU 0 Name: Tesla T4
GPU 0 Memory: 0.42GB/15.0GB (Used/Total)
Using 16bit Automatic Mixed Precision (AMP)
GPU available: True (cuda), used: True
TPU available: False, using: 0 TPU cores
HPU available: False, using: 0 HPUs
LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0]
| Name | Type | Params | Mode
--------------------------------------------------------------------------------------
0 | student_model | HFAutoModelForTextPrediction | 67.0 M | train
1 | teacher_model | HFAutoModelForTextPrediction | 109 M | train
2 | validation_metric | BinaryAUROC | 0 | train
3 | hard_label_loss_func | CrossEntropyLoss | 0 | train
4 | soft_label_loss_func | CrossEntropyLoss | 0 | train
5 | softmax_regression_loss_func | MSELoss | 0 | train
6 | output_feature_loss_func | MSELoss | 0 | train
7 | output_feature_adaptor | Identity | 0 | train
8 | rkd_loss_func | RKDLoss | 0 | train
--------------------------------------------------------------------------------------
176 M Trainable params
0 Non-trainable params
176 M Total params
705.761 Total estimated model params size (MB)
Epoch 0, global step 3: 'val_roc_auc' reached 0.65089 (best 0.65089), saving model to '/home/ci/autogluon/docs/tutorials/multimodal/advanced_topics/AutogluonModels/ag-20240716_231802/epoch=0-step=3.ckpt' as top 3
Epoch 0, global step 7: 'val_roc_auc' reached 0.70516 (best 0.70516), saving model to '/home/ci/autogluon/docs/tutorials/multimodal/advanced_topics/AutogluonModels/ag-20240716_231802/epoch=0-step=7.ckpt' as top 3
Epoch 1, global step 10: 'val_roc_auc' reached 0.71058 (best 0.71058), saving model to '/home/ci/autogluon/docs/tutorials/multimodal/advanced_topics/AutogluonModels/ag-20240716_231802/epoch=1-step=10.ckpt' as top 3
Epoch 1, global step 14: 'val_roc_auc' reached 0.70983 (best 0.71058), saving model to '/home/ci/autogluon/docs/tutorials/multimodal/advanced_topics/AutogluonModels/ag-20240716_231802/epoch=1-step=14.ckpt' as top 3
`Trainer.fit` stopped: `max_epochs=2` reached.
Start to fuse 3 checkpoints via the greedy soup algorithm.
AutoMM has created your model. 🎉🎉🎉
To load the model, use the code below:
```python
from autogluon.multimodal import MultiModalPredictor
predictor = MultiModalPredictor.load("/home/ci/autogluon/docs/tutorials/multimodal/advanced_topics/AutogluonModels/ag-20240716_231802")
```
If you are not satisfied with the model, try to increase the training time,
adjust the hyperparameters (https://auto.gluon.ai/stable/tutorials/multimodal/advanced_topics/customization.html),
or post issues on GitHub (https://github.com/autogluon/autogluon/issues).
<autogluon.multimodal.predictor.MultiModalPredictor at 0x7fb737c5fd90>
print(student_predictor.evaluate(data=test_df))
{'roc_auc': 0.8182983864837508}
More about Knowledge Distillation¶
To learn how to customize distillation and how it compares with direct finetuning, see the distillation examples and README in AutoMM Distillation Examples. Especially the multilingual distillation example with more details and customization.
Other Examples¶
You may go to AutoMM Examples to explore other examples about AutoMM.
Customization¶
To learn how to customize AutoMM, please refer to Customize AutoMM.