huggingface--transformers/docs/source/en/model_doc/xlm-roberta.md

This model was published in HF papers on 2019-11-05 and contributed to Hugging Face Transformers on 2020-11-16.

XLM-RoBERTa

XLM-RoBERTa is a large multilingual masked language model trained on 2.5TB of filtered CommonCrawl data across 100 languages. It shows that scaling the model provides strong performance gains on high-resource and low-resource languages. The model uses the RoBERTa pretraining objectives on the XLM model.

You can find all the original XLM-RoBERTa checkpoints under the Facebook AI community organization.

Tip

Click on the XLM-RoBERTa models in the right sidebar for more examples of how to apply XLM-RoBERTa to different cross-lingual tasks like classification, translation, and question answering.

The example below demonstrates how to predict the <mask> token with [Pipeline], [AutoModel], and from the command line.

from transformers import pipeline


pipeline = pipeline(
    task="fill-mask",
    model="FacebookAI/xlm-roberta-base",
    device=0
)
# Example in French
pipeline("Bonjour, je suis un modèle <mask>.")

import torch

from transformers import AutoModelForMaskedLM, AutoTokenizer


tokenizer = AutoTokenizer.from_pretrained(
    "FacebookAI/xlm-roberta-base"
)
model = AutoModelForMaskedLM.from_pretrained(
    "FacebookAI/xlm-roberta-base",
    device_map="auto",
    attn_implementation="sdpa"
)

# Prepare input
inputs = tokenizer("Bonjour, je suis un modèle <mask>.", return_tensors="pt").to(model.device)

with torch.no_grad():
    outputs = model(**inputs)
    predictions = outputs.logits

masked_index = torch.where(inputs['input_ids'] == tokenizer.mask_token_id)[1]
predicted_token_id = predictions[0, masked_index].argmax(dim=-1)
predicted_token = tokenizer.decode(predicted_token_id)

print(f"The predicted token is: {predicted_token}")

Quantization reduces the memory burden of large models by representing the weights in a lower precision. Refer to the quantization guide overview for more available quantization backends.

The example below uses bitsandbytes the quantive the weights to 4 bits

import torch
from transformers import AutoModelForMaskedLM, AutoTokenizer, BitsAndBytesConfig

quantization_config = BitsAndBytesConfig(
    load_in_4bit=True,
    bnb_4bit_compute_dtype=torch.bfloat16
    bnb_4bit_quant_type="nf4",  # or "fp4" for float 4-bit quantization
    bnb_4bit_use_double_quant=True,  # use double quantization for better performance
)
tokenizer = AutoTokenizer.from_pretrained("facebook/xlm-roberta-large")
model = AutoModelForMaskedLM.from_pretrained(
    "facebook/xlm-roberta-large",
    device_map="auto",
    attn_implementation="flash_attention_2",
    quantization_config=quantization_config
)

inputs = tokenizer("Bonjour, je suis un modèle <mask>.", return_tensors="pt").to(model.device)
outputs = model.generate(**inputs, max_new_tokens=100)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))

Notes

• Unlike some XLM models, XLM-RoBERTa doesn't require lang tensors to understand what language is being used. It automatically determines the language from the input IDs

Resources

A list of official Hugging Face and community (indicated by 🌎) resources to help you get started with XLM-RoBERTa. If you're interested in submitting a resource to be included here, please feel free to open a Pull Request and we'll review it! The resource should ideally demonstrate something new instead of duplicating an existing resource.

• A blog post on how to finetune XLM RoBERTa for multiclass classification with Habana Gaudi on AWS
• [XLMRobertaForSequenceClassification] is supported by this example script and notebook..
• Text classification chapter of the 🤗 Hugging Face Task Guides.
• Text classification task guide

• [XLMRobertaForTokenClassification] is supported by this example script and notebook.
• Token classification chapter of the 🤗 Hugging Face Course.
• Token classification task guide

• [XLMRobertaForCausalLM] is supported by this example script and notebook.
• Causal language modeling chapter of the 🤗 Hugging Face Task Guides.
• Causal language modeling task guide

• [XLMRobertaForMaskedLM] is supported by this example script and notebook.
• Masked language modeling chapter of the 🤗 Hugging Face Course.
• Masked language modeling

• [XLMRobertaForQuestionAnswering] is supported by this example script and notebook.
• Question answering chapter of the 🤗 Hugging Face Course.
• Question answering task guide

Multiple choice

• [XLMRobertaForMultipleChoice] is supported by this example script and notebook.
• Multiple choice task guide

🚀 Deploy

• A blog post on how to Deploy Serverless XLM RoBERTa on AWS Lambda.

This implementation is the same as RoBERTa. Refer to the documentation of RoBERTa for usage examples as well as the information relative to the inputs and outputs.

XLMRobertaConfig

autodoc XLMRobertaConfig

XLMRobertaTokenizer

autodoc XLMRobertaTokenizer - get_special_tokens_mask - save_vocabulary

XLMRobertaTokenizerFast

autodoc XLMRobertaTokenizerFast

XLMRobertaModel

autodoc XLMRobertaModel - forward

XLMRobertaForCausalLM

autodoc XLMRobertaForCausalLM - forward

XLMRobertaForMaskedLM

autodoc XLMRobertaForMaskedLM - forward

XLMRobertaForSequenceClassification

autodoc XLMRobertaForSequenceClassification - forward

XLMRobertaForMultipleChoice

autodoc XLMRobertaForMultipleChoice - forward

XLMRobertaForTokenClassification

autodoc XLMRobertaForTokenClassification - forward

XLMRobertaForQuestionAnswering

autodoc XLMRobertaForQuestionAnswering - forward