File size: 19,638 Bytes
455a40f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
<!--Copyright 2020 The HuggingFace Team. All rights reserved.

Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
the License. You may obtain a copy of the License at

http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
specific language governing permissions and limitations under the License.
-->

# Export to ONNX

If you need to deploy 🤗 Transformers models in production environments, we recommend
exporting them to a serialized format that can be loaded and executed on specialized
runtimes and hardware. In this guide, we'll show you how to export 🤗 Transformers
models to [ONNX (Open Neural Network eXchange)](http://onnx.ai).

ONNX is an open standard that defines a common set of operators and a common file format
to represent deep learning models in a wide variety of frameworks, including PyTorch and
TensorFlow. When a model is exported to the ONNX format, these operators are used to
construct a computational graph (often called an _intermediate representation_) which
represents the flow of data through the neural network.

By exposing a graph with standardized operators and data types, ONNX makes it easy to
switch between frameworks. For example, a model trained in PyTorch can be exported to
ONNX format and then imported in TensorFlow (and vice versa).

🤗 Transformers provides a [`transformers.onnx`](main_classes/onnx) package that enables
you to convert model checkpoints to an ONNX graph by leveraging configuration objects.
These configuration objects come ready made for a number of model architectures, and are
designed to be easily extendable to other architectures.

<Tip>

You can also export 🤗 Transformers models with the [`optimum.exporters.onnx` package](https://huggingface.co/docs/optimum/exporters/onnx/usage_guides/export_a_model)
from 🤗 Optimum.

Once exported, a model can be:

- Optimized for inference via techniques such as quantization and graph optimization.
- Run with ONNX Runtime via [`ORTModelForXXX` classes](https://huggingface.co/docs/optimum/onnxruntime/package_reference/modeling_ort),
which follow the same `AutoModel` API as the one you are used to in 🤗 Transformers.
- Run with [optimized inference pipelines](https://huggingface.co/docs/optimum/main/en/onnxruntime/usage_guides/pipelines),
which has the same API as the [`pipeline`] function in 🤗 Transformers.

To explore all these features,  check out the [🤗 Optimum library](https://github.com/huggingface/optimum).

</Tip>

Ready-made configurations include the following architectures:

<!--This table is automatically generated by `make fix-copies`, do not fill manually!-->

- ALBERT
- BART
- BEiT
- BERT
- BigBird
- BigBird-Pegasus
- Blenderbot
- BlenderbotSmall
- BLOOM
- CamemBERT
- Chinese-CLIP
- CLIP
- CodeGen
- Conditional DETR
- ConvBERT
- ConvNeXT
- Data2VecText
- Data2VecVision
- DeBERTa
- DeBERTa-v2
- DeiT
- DETR
- DistilBERT
- EfficientNet
- ELECTRA
- ERNIE
- FlauBERT
- GPT Neo
- GPT-J
- GPT-Sw3
- GroupViT
- I-BERT
- ImageGPT
- LayoutLM
- LayoutLMv3
- LeViT
- Longformer
- LongT5
- M2M100
- Marian
- mBART
- MEGA
- MobileBERT
- MobileNetV1
- MobileNetV2
- MobileViT
- MT5
- OpenAI GPT-2
- OWL-ViT
- Perceiver
- PLBart
- PoolFormer
- RemBERT
- ResNet
- RoBERTa
- RoBERTa-PreLayerNorm
- RoFormer
- SegFormer
- SqueezeBERT
- Swin Transformer
- T5
- Table Transformer
- Vision Encoder decoder
- ViT
- Whisper
- X-MOD
- XLM
- XLM-RoBERTa
- XLM-RoBERTa-XL
- YOLOS

In the next two sections, we'll show you how to:

* Export a supported model using the `transformers.onnx` package.
* Export a custom model for an unsupported architecture.

## Exporting a model to ONNX

<Tip>

The recommended way of exporting a model is now to use
[`optimum.exporters.onnx`](https://huggingface.co/docs/optimum/main/en/exporters/onnx/usage_guides/export_a_model#exporting-a-model-to-onnx-using-the-cli),
do not worry it is very similar to `transformers.onnx`!

</Tip>

To export a 🤗 Transformers model to ONNX, you'll first need to install some extra
dependencies:

```bash
pip install transformers[onnx]
```

The `transformers.onnx` package can then be used as a Python module:

```bash
python -m transformers.onnx --help

usage: Hugging Face Transformers ONNX exporter [-h] -m MODEL [--feature {causal-lm, ...}] [--opset OPSET] [--atol ATOL] output

positional arguments:
  output                Path indicating where to store generated ONNX model.

optional arguments:
  -h, --help            show this help message and exit
  -m MODEL, --model MODEL
                        Model ID on huggingface.co or path on disk to load model from.
  --feature {causal-lm, ...}
                        The type of features to export the model with.
  --opset OPSET         ONNX opset version to export the model with.
  --atol ATOL           Absolute difference tolerance when validating the model.
```

Exporting a checkpoint using a ready-made configuration can be done as follows:

```bash
python -m transformers.onnx --model=distilbert-base-uncased onnx/
```

You should see the following logs:

```bash
Validating ONNX model...
        -[✓] ONNX model output names match reference model ({'last_hidden_state'})
        - Validating ONNX Model output "last_hidden_state":
                -[✓] (2, 8, 768) matches (2, 8, 768)
                -[✓] all values close (atol: 1e-05)
All good, model saved at: onnx/model.onnx
```

This exports an ONNX graph of the checkpoint defined by the `--model` argument. In this
example, it is `distilbert-base-uncased`, but it can be any checkpoint on the Hugging
Face Hub or one that's stored locally.

The resulting `model.onnx` file can then be run on one of the [many
accelerators](https://onnx.ai/supported-tools.html#deployModel) that support the ONNX
standard. For example, we can load and run the model with [ONNX
Runtime](https://onnxruntime.ai/) as follows:

```python
>>> from transformers import AutoTokenizer
>>> from onnxruntime import InferenceSession

>>> tokenizer = AutoTokenizer.from_pretrained("distilbert-base-uncased")
>>> session = InferenceSession("onnx/model.onnx")
>>> # ONNX Runtime expects NumPy arrays as input
>>> inputs = tokenizer("Using DistilBERT with ONNX Runtime!", return_tensors="np")
>>> outputs = session.run(output_names=["last_hidden_state"], input_feed=dict(inputs))
```

The required output names (like `["last_hidden_state"]`) can be obtained by taking a
look at the ONNX configuration of each model. For example, for DistilBERT we have:

```python
>>> from transformers.models.distilbert import DistilBertConfig, DistilBertOnnxConfig

>>> config = DistilBertConfig()
>>> onnx_config = DistilBertOnnxConfig(config)
>>> print(list(onnx_config.outputs.keys()))
["last_hidden_state"]
```

The process is identical for TensorFlow checkpoints on the Hub. For example, we can
export a pure TensorFlow checkpoint from the [Keras
organization](https://huggingface.co/keras-io) as follows:

```bash
python -m transformers.onnx --model=keras-io/transformers-qa onnx/
```

To export a model that's stored locally, you'll need to have the model's weights and
tokenizer files stored in a directory. For example, we can load and save a checkpoint as
follows:

<frameworkcontent> <pt>
```python
>>> from transformers import AutoTokenizer, AutoModelForSequenceClassification

>>> # Load tokenizer and PyTorch weights form the Hub
>>> tokenizer = AutoTokenizer.from_pretrained("distilbert-base-uncased")
>>> pt_model = AutoModelForSequenceClassification.from_pretrained("distilbert-base-uncased")
>>> # Save to disk
>>> tokenizer.save_pretrained("local-pt-checkpoint")
>>> pt_model.save_pretrained("local-pt-checkpoint")
```

Once the checkpoint is saved, we can export it to ONNX by pointing the `--model`
argument of the `transformers.onnx` package to the desired directory:

```bash
python -m transformers.onnx --model=local-pt-checkpoint onnx/
```
</pt> <tf>
```python
>>> from transformers import AutoTokenizer, TFAutoModelForSequenceClassification

>>> # Load tokenizer and TensorFlow weights from the Hub
>>> tokenizer = AutoTokenizer.from_pretrained("distilbert-base-uncased")
>>> tf_model = TFAutoModelForSequenceClassification.from_pretrained("distilbert-base-uncased")
>>> # Save to disk
>>> tokenizer.save_pretrained("local-tf-checkpoint")
>>> tf_model.save_pretrained("local-tf-checkpoint")
```

Once the checkpoint is saved, we can export it to ONNX by pointing the `--model`
argument of the `transformers.onnx` package to the desired directory:

```bash
python -m transformers.onnx --model=local-tf-checkpoint onnx/
```
</tf> </frameworkcontent>

## Selecting features for different model tasks

<Tip>

The recommended way of exporting a model is now to use `optimum.exporters.onnx`.
You can check the [🤗 Optimum documentation](https://huggingface.co/docs/optimum/main/en/exporters/onnx/usage_guides/export_a_model#selecting-a-task)
to learn how to select a task.

</Tip>

Each ready-made configuration comes with a set of _features_ that enable you to export
models for different types of tasks. As shown in the table below, each feature is
associated with a different `AutoClass`:

| Feature                              | Auto Class                           |
| ------------------------------------ | ------------------------------------ |
| `causal-lm`, `causal-lm-with-past`   | `AutoModelForCausalLM`               |
| `default`, `default-with-past`       | `AutoModel`                          |
| `masked-lm`                          | `AutoModelForMaskedLM`               |
| `question-answering`                 | `AutoModelForQuestionAnswering`      |
| `seq2seq-lm`, `seq2seq-lm-with-past` | `AutoModelForSeq2SeqLM`              |
| `sequence-classification`            | `AutoModelForSequenceClassification` |
| `token-classification`               | `AutoModelForTokenClassification`    |

For each configuration, you can find the list of supported features via the
[`~transformers.onnx.FeaturesManager`]. For example, for DistilBERT we have:

```python
>>> from transformers.onnx.features import FeaturesManager

>>> distilbert_features = list(FeaturesManager.get_supported_features_for_model_type("distilbert").keys())
>>> print(distilbert_features)
["default", "masked-lm", "causal-lm", "sequence-classification", "token-classification", "question-answering"]
```

You can then pass one of these features to the `--feature` argument in the
`transformers.onnx` package. For example, to export a text-classification model we can
pick a fine-tuned model from the Hub and run:

```bash
python -m transformers.onnx --model=distilbert-base-uncased-finetuned-sst-2-english \
                            --feature=sequence-classification onnx/
```

This displays the following logs:

```bash
Validating ONNX model...
        -[✓] ONNX model output names match reference model ({'logits'})
        - Validating ONNX Model output "logits":
                -[✓] (2, 2) matches (2, 2)
                -[✓] all values close (atol: 1e-05)
All good, model saved at: onnx/model.onnx
```

Notice that in this case, the output names from the fine-tuned model are `logits`
instead of the `last_hidden_state` we saw with the `distilbert-base-uncased` checkpoint
earlier. This is expected since the fine-tuned model has a sequence classification head.

<Tip>

The features that have a `with-past` suffix (like `causal-lm-with-past`) correspond to
model classes with precomputed hidden states (key and values in the attention blocks)
that can be used for fast autoregressive decoding.

</Tip>

<Tip>

For `VisionEncoderDecoder` type models, the encoder and decoder parts are
exported separately as two ONNX files named `encoder_model.onnx` and `decoder_model.onnx` respectively.

</Tip>


## Exporting a model for an unsupported architecture

<Tip>

If you wish to contribute by adding support for a model that cannot be currently exported, you should first check if it is
supported in [`optimum.exporters.onnx`](https://huggingface.co/docs/optimum/main/en/exporters/onnx/package_reference/configuration#supported-architectures),
and if it is not, [contribute to 🤗 Optimum](https://huggingface.co/docs/optimum/main/en/exporters/onnx/usage_guides/contribute)
directly.

</Tip>

If you wish to export a model whose architecture is not natively supported by the
library, there are three main steps to follow:

1. Implement a custom ONNX configuration.
2. Export the model to ONNX.
3. Validate the outputs of the PyTorch and exported models.

In this section, we'll look at how DistilBERT was implemented to show what's involved
with each step.

### Implementing a custom ONNX configuration

Let's start with the ONNX configuration object. We provide three abstract classes that
you should inherit from, depending on the type of model architecture you wish to export:

* Encoder-based models inherit from [`~onnx.config.OnnxConfig`]
* Decoder-based models inherit from [`~onnx.config.OnnxConfigWithPast`]
* Encoder-decoder models inherit from [`~onnx.config.OnnxSeq2SeqConfigWithPast`]

<Tip>

A good way to implement a custom ONNX configuration is to look at the existing
implementation in the `configuration_<model_name>.py` file of a similar architecture.

</Tip>

Since DistilBERT is an encoder-based model, its configuration inherits from
`OnnxConfig`:

```python
>>> from typing import Mapping, OrderedDict
>>> from transformers.onnx import OnnxConfig


>>> class DistilBertOnnxConfig(OnnxConfig):
...     @property
...     def inputs(self) -> Mapping[str, Mapping[int, str]]:
...         return OrderedDict(
...             [
...                 ("input_ids", {0: "batch", 1: "sequence"}),
...                 ("attention_mask", {0: "batch", 1: "sequence"}),
...             ]
...         )
```

Every configuration object must implement the `inputs` property and return a mapping,
where each key corresponds to an expected input, and each value indicates the axis of
that input. For DistilBERT, we can see that two inputs are required: `input_ids` and
`attention_mask`. These inputs have the same shape of `(batch_size, sequence_length)`
which is why we see the same axes used in the configuration.

<Tip>

Notice that `inputs` property for `DistilBertOnnxConfig` returns an `OrderedDict`. This
ensures that the inputs are matched with their relative position within the
`PreTrainedModel.forward()` method when tracing the graph. We recommend using an
`OrderedDict` for the `inputs` and `outputs` properties when implementing custom ONNX
configurations.

</Tip>

Once you have implemented an ONNX configuration, you can instantiate it by providing the
base model's configuration as follows:

```python
>>> from transformers import AutoConfig

>>> config = AutoConfig.from_pretrained("distilbert-base-uncased")
>>> onnx_config = DistilBertOnnxConfig(config)
```

The resulting object has several useful properties. For example, you can view the ONNX
operator set that will be used during the export:

```python
>>> print(onnx_config.default_onnx_opset)
11
```

You can also view the outputs associated with the model as follows:

```python
>>> print(onnx_config.outputs)
OrderedDict([("last_hidden_state", {0: "batch", 1: "sequence"})])
```

Notice that the outputs property follows the same structure as the inputs; it returns an
`OrderedDict` of named outputs and their shapes. The output structure is linked to the
choice of feature that the configuration is initialised with. By default, the ONNX
configuration is initialized with the `default` feature that corresponds to exporting a
model loaded with the `AutoModel` class. If you want to export a model for another task,
just provide a different feature to the `task` argument when you initialize the ONNX
configuration. For example, if we wished to export DistilBERT with a sequence
classification head, we could use:

```python
>>> from transformers import AutoConfig

>>> config = AutoConfig.from_pretrained("distilbert-base-uncased")
>>> onnx_config_for_seq_clf = DistilBertOnnxConfig(config, task="sequence-classification")
>>> print(onnx_config_for_seq_clf.outputs)
OrderedDict([('logits', {0: 'batch'})])
```

<Tip>

All of the base properties and methods associated with [`~onnx.config.OnnxConfig`] and
the other configuration classes can be overridden if needed. Check out [`BartOnnxConfig`]
for an advanced example.

</Tip>

### Exporting the model

Once you have implemented the ONNX configuration, the next step is to export the model.
Here we can use the `export()` function provided by the `transformers.onnx` package.
This function expects the ONNX configuration, along with the base model and tokenizer,
and the path to save the exported file:

```python
>>> from pathlib import Path
>>> from transformers.onnx import export
>>> from transformers import AutoTokenizer, AutoModel

>>> onnx_path = Path("model.onnx")
>>> model_ckpt = "distilbert-base-uncased"
>>> base_model = AutoModel.from_pretrained(model_ckpt)
>>> tokenizer = AutoTokenizer.from_pretrained(model_ckpt)

>>> onnx_inputs, onnx_outputs = export(tokenizer, base_model, onnx_config, onnx_config.default_onnx_opset, onnx_path)
```

The `onnx_inputs` and `onnx_outputs` returned by the `export()` function are lists of
the keys defined in the `inputs` and `outputs` properties of the configuration. Once the
model is exported, you can test that the model is well formed as follows:

```python
>>> import onnx

>>> onnx_model = onnx.load("model.onnx")
>>> onnx.checker.check_model(onnx_model)
```

<Tip>

If your model is larger than 2GB, you will see that many additional files are created
during the export. This is _expected_ because ONNX uses [Protocol
Buffers](https://developers.google.com/protocol-buffers/) to store the model and these
have a size limit of 2GB. See the [ONNX
documentation](https://github.com/onnx/onnx/blob/master/docs/ExternalData.md) for
instructions on how to load models with external data.

</Tip>

### Validating the model outputs

The final step is to validate that the outputs from the base and exported model agree
within some absolute tolerance. Here we can use the `validate_model_outputs()` function
provided by the `transformers.onnx` package as follows:

```python
>>> from transformers.onnx import validate_model_outputs

>>> validate_model_outputs(
...     onnx_config, tokenizer, base_model, onnx_path, onnx_outputs, onnx_config.atol_for_validation
... )
```

This function uses the [`~transformers.onnx.OnnxConfig.generate_dummy_inputs`] method to
generate inputs for the base and exported model, and the absolute tolerance can be
defined in the configuration. We generally find numerical agreement in the 1e-6 to 1e-4
range, although anything smaller than 1e-3 is likely to be OK.

## Contributing a new configuration to 🤗 Transformers

We are looking to expand the set of ready-made configurations and welcome contributions
from the community! If you would like to contribute your addition to the library, you
will need to:

* Implement the ONNX configuration in the corresponding `configuration_<model_name>.py`
file
* Include the model architecture and corresponding features in
  [`~onnx.features.FeatureManager`]
* Add your model architecture to the tests in `test_onnx_v2.py`

Check out how the configuration for [IBERT was
contributed](https://github.com/huggingface/transformers/pull/14868/files) to get an
idea of what's involved.