mirror of
https://github.com/huggingface/transformers.git
synced 2025-07-23 14:29:01 +06:00
Standardize ByT5 model card format (#38699)
Some checks are pending
Self-hosted runner (benchmark) / Benchmark (aws-g5-4xlarge-cache) (push) Waiting to run
Build documentation / build (push) Waiting to run
Slow tests on important models (on Push - A10) / Get all modified files (push) Waiting to run
Slow tests on important models (on Push - A10) / Slow & FA2 tests (push) Blocked by required conditions
Self-hosted runner (push-caller) / Check if setup was changed (push) Waiting to run
Self-hosted runner (push-caller) / build-docker-containers (push) Blocked by required conditions
Self-hosted runner (push-caller) / Trigger Push CI (push) Blocked by required conditions
Secret Leaks / trufflehog (push) Waiting to run
Update Transformers metadata / build_and_package (push) Waiting to run
Some checks are pending
Self-hosted runner (benchmark) / Benchmark (aws-g5-4xlarge-cache) (push) Waiting to run
Build documentation / build (push) Waiting to run
Slow tests on important models (on Push - A10) / Get all modified files (push) Waiting to run
Slow tests on important models (on Push - A10) / Slow & FA2 tests (push) Blocked by required conditions
Self-hosted runner (push-caller) / Check if setup was changed (push) Waiting to run
Self-hosted runner (push-caller) / build-docker-containers (push) Blocked by required conditions
Self-hosted runner (push-caller) / Trigger Push CI (push) Blocked by required conditions
Secret Leaks / trufflehog (push) Waiting to run
Update Transformers metadata / build_and_package (push) Waiting to run
* Standardize ByT5 model card format * Apply review feedback from @stevhliu * Fix Notes formatting and wording * Fix `aya_vision` test (#38674) * fix 1: load_in_4bit=True, * fix 2: decorateor * fixfix 2: breakpoint * fixfix 3: update * fixfix 4: fast * fixfix 5: cond * fixfix 5: cond * fixfix 6: cuda 8 * ruff * breakpoint * dtype * a10 * a10 --------- Co-authored-by: ydshieh <ydshieh@users.noreply.github.com> * Fix autodoc formatting for ByT5Tokenizer --------- Co-authored-by: Yih-Dar <2521628+ydshieh@users.noreply.github.com> Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>
This commit is contained in:
Yana Mishula
GitHub
parent
e55983e2b9
commit
81799d8b55
@ -13,150 +13,128 @@ specific language governing permissions and limitations under the License.
|
||||
rendered properly in your Markdown viewer.
|
||||
|
||||
-->
|
||||
|
||||
# ByT5
|
||||
|
||||
<div style="float: right;">
|
||||
<div class="flex flex-wrap space-x-1">
|
||||
<img alt="PyTorch" src="https://img.shields.io/badge/PyTorch-DE3412?style=flat&logo=pytorch&logoColor=white">
|
||||
<img alt="TensorFlow" src="https://img.shields.io/badge/TensorFlow-FF6F00?style=flat&logo=tensorflow&logoColor=white">
|
||||
<img alt="Flax" src="https://img.shields.io/badge/Flax-29a79b.svg?style=flat&logo=data:image/png;base64,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
|
||||
">
|
||||
<img alt="Flax" src="https://img.shields.io/badge/Flax-29a79b.svg?style=flat&logo=flax&logoColor=white">
|
||||
</div>
|
||||
</div>
|
||||
|
||||
## Overview
|
||||
# ByT5
|
||||
|
||||
The ByT5 model was presented in [ByT5: Towards a token-free future with pre-trained byte-to-byte models](https://arxiv.org/abs/2105.13626) by Linting Xue, Aditya Barua, Noah Constant, Rami Al-Rfou, Sharan Narang, Mihir
|
||||
Kale, Adam Roberts, Colin Raffel.
|
||||
[ByT5](https://huggingface.co/papers/2105.13626) is tokenizer-free version of the [T5](./t5) model designed to works directly on raw UTF-8 bytes. This means it can process any language, more robust to noise like typos, and simpler to use because it doesn't require a preprocessing pipeline.
|
||||
|
||||
The abstract from the paper is the following:
|
||||
You can find all the original ByT5 checkpoints under the [Google](https://huggingface.co/google?search_models=byt5) organization.
|
||||
|
||||
*Most widely-used pre-trained language models operate on sequences of tokens corresponding to word or subword units.
|
||||
Encoding text as a sequence of tokens requires a tokenizer, which is typically created as an independent artifact from
|
||||
the model. Token-free models that instead operate directly on raw text (bytes or characters) have many benefits: they
|
||||
can process text in any language out of the box, they are more robust to noise, and they minimize technical debt by
|
||||
removing complex and error-prone text preprocessing pipelines. Since byte or character sequences are longer than token
|
||||
sequences, past work on token-free models has often introduced new model architectures designed to amortize the cost of
|
||||
operating directly on raw text. In this paper, we show that a standard Transformer architecture can be used with
|
||||
minimal modifications to process byte sequences. We carefully characterize the trade-offs in terms of parameter count,
|
||||
training FLOPs, and inference speed, and show that byte-level models are competitive with their token-level
|
||||
counterparts. We also demonstrate that byte-level models are significantly more robust to noise and perform better on
|
||||
tasks that are sensitive to spelling and pronunciation. As part of our contribution, we release a new set of
|
||||
pre-trained byte-level Transformer models based on the T5 architecture, as well as all code and data used in our
|
||||
experiments.*
|
||||
> [!TIP]
|
||||
> Refer to the [T5](./t5) docs for more examples of how to apply ByT5 to different language tasks.
|
||||
|
||||
This model was contributed by [patrickvonplaten](https://huggingface.co/patrickvonplaten). The original code can be
|
||||
found [here](https://github.com/google-research/byt5).
|
||||
The example below demonstrates how to generate text with [`Pipeline`], [`AutoModel`] and from the command line.
|
||||
|
||||
<Tip>
|
||||
|
||||
ByT5's architecture is based on the T5v1.1 model, refer to [T5v1.1's documentation page](t5v1.1) for the API reference. They
|
||||
only differ in how inputs should be prepared for the model, see the code examples below.
|
||||
|
||||
</Tip>
|
||||
|
||||
Since ByT5 was pre-trained unsupervisedly, there's no real advantage to using a task prefix during single-task
|
||||
fine-tuning. If you are doing multi-task fine-tuning, you should use a prefix.
|
||||
|
||||
|
||||
## Usage example
|
||||
|
||||
ByT5 works on raw UTF-8 bytes, so it can be used without a tokenizer:
|
||||
<hfoptions id="usage">
|
||||
<hfoption id="Pipeline">
|
||||
|
||||
```python
|
||||
>>> from transformers import T5ForConditionalGeneration
|
||||
>>> import torch
|
||||
import torch
|
||||
from transformers import pipeline
|
||||
|
||||
>>> model = T5ForConditionalGeneration.from_pretrained("google/byt5-small")
|
||||
|
||||
>>> num_special_tokens = 3
|
||||
>>> # Model has 3 special tokens which take up the input ids 0,1,2 of ByT5.
|
||||
>>> # => Need to shift utf-8 character encodings by 3 before passing ids to model.
|
||||
|
||||
>>> input_ids = torch.tensor([list("Life is like a box of chocolates.".encode("utf-8"))]) + num_special_tokens
|
||||
|
||||
>>> labels = torch.tensor([list("La vie est comme une boîte de chocolat.".encode("utf-8"))]) + num_special_tokens
|
||||
|
||||
>>> loss = model(input_ids, labels=labels).loss
|
||||
>>> loss.item()
|
||||
2.66
|
||||
pipeline = pipeline(
|
||||
task="text2text-generation",
|
||||
model="google/byt5-small",
|
||||
torch_dtype=torch.float16,
|
||||
device=0
|
||||
)
|
||||
pipeline("translate English to French: The weather is nice today")
|
||||
```
|
||||
|
||||
For batched inference and training it is however recommended to make use of the tokenizer:
|
||||
</hfoption>
|
||||
<hfoption id="AutoModel">
|
||||
|
||||
```python
|
||||
>>> from transformers import T5ForConditionalGeneration, AutoTokenizer
|
||||
import torch
|
||||
from transformers import AutoModelForSeq2SeqLM, AutoTokenizer
|
||||
|
||||
>>> model = T5ForConditionalGeneration.from_pretrained("google/byt5-small")
|
||||
>>> tokenizer = AutoTokenizer.from_pretrained("google/byt5-small")
|
||||
tokenizer = AutoTokenizer.from_pretrained(
|
||||
"google/byt5-small"
|
||||
)
|
||||
model = AutoModelForSeq2SeqLM.from_pretrained(
|
||||
"google/byt5-small",
|
||||
torch_dtype=torch.float16,
|
||||
device_map="auto"
|
||||
)
|
||||
|
||||
>>> model_inputs = tokenizer(
|
||||
... ["Life is like a box of chocolates.", "Today is Monday."], padding="longest", return_tensors="pt"
|
||||
... )
|
||||
>>> labels_dict = tokenizer(
|
||||
... ["La vie est comme une boîte de chocolat.", "Aujourd'hui c'est lundi."], padding="longest", return_tensors="pt"
|
||||
... )
|
||||
>>> labels = labels_dict.input_ids
|
||||
input_ids = tokenizer("summarize: Photosynthesis is the process by which plants, algae, and some bacteria convert light energy into chemical energy.", return_tensors="pt").to("cuda")
|
||||
|
||||
>>> loss = model(**model_inputs, labels=labels).loss
|
||||
>>> loss.item()
|
||||
17.9
|
||||
output = model.generate(**input_ids)
|
||||
print(tokenizer.decode(output[0], skip_special_tokens=True))
|
||||
```
|
||||
|
||||
Similar to [T5](t5), ByT5 was trained on the span-mask denoising task. However,
|
||||
since the model works directly on characters, the pretraining task is a bit
|
||||
different. Let's corrupt some characters of the
|
||||
input sentence `"The dog chases a ball in the park."` and ask ByT5 to predict them
|
||||
for us.
|
||||
</hfoption>
|
||||
<hfoption id="transformers-cli">
|
||||
|
||||
```bash
|
||||
echo -e "translate English to French: Life is beautiful." | transformers-cli run --task text2text-generation --model google/byt5-small --device 0
|
||||
```
|
||||
|
||||
</hfoption>
|
||||
</hfoptions>
|
||||
|
||||
## Quantization
|
||||
|
||||
Quantization reduces the memory burden of large models by representing the weights in a lower precision. Refer to the [Quantization](../quantization/overview) overview for more available quantization backends.
|
||||
|
||||
The example below uses [torchao](../quantization/torchao) to only quantize the weights to int4.
|
||||
|
||||
```python
|
||||
>>> from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
|
||||
>>> import torch
|
||||
# pip install torchao
|
||||
import torch
|
||||
from transformers import TorchAoConfig, AutoModelForSeq2SeqLM, AutoTokenizer
|
||||
|
||||
>>> tokenizer = AutoTokenizer.from_pretrained("google/byt5-base")
|
||||
>>> model = AutoModelForSeq2SeqLM.from_pretrained("google/byt5-base")
|
||||
quantization_config = TorchAoConfig("int4_weight_only", group_size=128)
|
||||
|
||||
>>> input_ids_prompt = "The dog chases a ball in the park."
|
||||
>>> input_ids = tokenizer(input_ids_prompt).input_ids
|
||||
model = AutoModelForSeq2SeqLM.from_pretrained(
|
||||
"google/byt5-xl",
|
||||
torch_dtype=torch.bfloat16,
|
||||
device_map="auto",
|
||||
quantization_config=quantization_config
|
||||
)
|
||||
|
||||
>>> # Note that we cannot add "{extra_id_...}" to the string directly
|
||||
>>> # as the Byte tokenizer would incorrectly merge the tokens
|
||||
>>> # For ByT5, we need to work directly on the character level
|
||||
>>> # Contrary to T5, ByT5 does not use sentinel tokens for masking, but instead
|
||||
>>> # uses final utf character ids.
|
||||
>>> # UTF-8 is represented by 8 bits and ByT5 has 3 special tokens.
|
||||
>>> # => There are 2**8+2 = 259 input ids and mask tokens count down from index 258.
|
||||
>>> # => mask to "The dog [258]a ball [257]park."
|
||||
tokenizer = AutoTokenizer.from_pretrained("google/byt5-xl")
|
||||
input_ids = tokenizer("translate English to French: The weather is nice today.", return_tensors="pt").to("cuda")
|
||||
|
||||
>>> input_ids = torch.tensor([input_ids[:8] + [258] + input_ids[14:21] + [257] + input_ids[28:]])
|
||||
>>> input_ids
|
||||
tensor([[ 87, 107, 104, 35, 103, 114, 106, 35, 258, 35, 100, 35, 101, 100, 111, 111, 257, 35, 115, 100, 117, 110, 49, 1]])
|
||||
|
||||
>>> # ByT5 produces only one char at a time so we need to produce many more output characters here -> set `max_length=100`.
|
||||
>>> output_ids = model.generate(input_ids, max_length=100)[0].tolist()
|
||||
>>> output_ids
|
||||
[0, 258, 108, 118, 35, 119, 107, 104, 35, 114, 113, 104, 35, 122, 107, 114, 35, 103, 114, 104, 118, 257, 35, 108, 113, 35, 119, 107, 104, 35, 103, 108, 118, 102, 114, 256, 108, 113, 35, 119, 107, 104, 35, 115, 100, 117, 110, 49, 35, 87, 107, 104, 35, 103, 114, 106, 35, 108, 118, 35, 119, 107, 104, 35, 114, 113, 104, 35, 122, 107, 114, 35, 103, 114, 104, 118, 35, 100, 35, 101, 100, 111, 111, 35, 108, 113, 255, 35, 108, 113, 35, 119, 107, 104, 35, 115, 100, 117, 110, 49]
|
||||
|
||||
>>> # ^- Note how 258 descends to 257, 256, 255
|
||||
|
||||
>>> # Now we need to split on the sentinel tokens, let's write a short loop for this
|
||||
>>> output_ids_list = []
|
||||
>>> start_token = 0
|
||||
>>> sentinel_token = 258
|
||||
>>> while sentinel_token in output_ids:
|
||||
... split_idx = output_ids.index(sentinel_token)
|
||||
... output_ids_list.append(output_ids[start_token:split_idx])
|
||||
... start_token = split_idx
|
||||
... sentinel_token -= 1
|
||||
|
||||
>>> output_ids_list.append(output_ids[start_token:])
|
||||
>>> output_string = tokenizer.batch_decode(output_ids_list)
|
||||
>>> output_string
|
||||
['<pad>', 'is the one who does', ' in the disco', 'in the park. The dog is the one who does a ball in', ' in the park.']
|
||||
output = model.generate(**input_ids)
|
||||
print(tokenizer.decode(output[0], skip_special_tokens=True))
|
||||
```
|
||||
|
||||
## Notes
|
||||
|
||||
- It is recommended to use the tokenizer for batched inference and training.
|
||||
- The example below shows how to use the model without a tokenizer.
|
||||
|
||||
```python
|
||||
import torch
|
||||
from transformers import AutoModelForSeq2SeqLM
|
||||
|
||||
model = AutoModelForSeq2SeqLM.from_pretrained("google/byt5-small")
|
||||
|
||||
num_special_tokens = 3
|
||||
|
||||
input_ids = torch.tensor([list("Life is like a box of chocolates.".encode("utf-8"))]) + num_special_tokens
|
||||
labels = torch.tensor([list("La vie est comme une boîte de chocolat.".encode("utf-8"))]) + num_special_tokens
|
||||
loss = model(input_ids, labels=labels).loss
|
||||
loss.item()
|
||||
```
|
||||
|
||||
- ByT5 uses the top byte values (258, 257, etc.) for masking instead of sentinel tokens like `{extra_id_0}`.
|
||||
|
||||
```python
|
||||
# Example: character-level denoising with mask tokens
|
||||
input_ids = tokenizer("The dog chases a ball in the park.").input_ids
|
||||
masked_input = torch.tensor([input_ids[:8] + [258] + input_ids[14:21] + [257] + input_ids[28:]])
|
||||
output = model.generate(masked_input, max_length=100)
|
||||
```
|
||||
|
||||
## ByT5Tokenizer
|
||||
|
||||
[[autodoc]] ByT5Tokenizer
|
||||
|
||||
See [`ByT5Tokenizer`] for all details.
|
||||
|
||||
Loading…
Reference in New Issue
Block a user