SAVRN
Search Contact SAVRN

Open-weight model · Fill mask

xlm-roberta-large

by Facebook AI community FacebookAI/xlm-roberta-large

XLM-RoBERTa model pre-trained on 2.5TB of filtered CommonCrawl data containing 100 languages. It was introduced in the paper Unsupervised Cross-lingual Representation Learning at Scale by Conneau et al. and first released in this repository.

Parameters561M
Context514
Weights9.0 GB
Licensemit
AccessOpen weights
Monthly Downloads3.2M

Runs On

What it takes to serve xlm-roberta-large (561M parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them.

PrecisionWeightsMemory neededCheapest setupPer hourAlso fits
16-bit 1.1 GB 1.3 GB 1x MI300X (192 GB)
Vultr		 $1.85 1x H100 $1.99 · 1x MI325X $2.00
8-bit 0.6 GB 0.7 GB 1x MI300X (192 GB)
Vultr		 $1.85 1x H100 $1.99 · 1x MI325X $2.00
4-bit 0.3 GB 0.3 GB 1x MI300X (192 GB)
Vultr		 $1.85 1x H100 $1.99 · 1x MI325X $2.00

Memory is the weights at that precision plus 20% for the runtime and a short context; a long context needs more. Prices are the lowest on-demand hourly rates in the SAVRN Index, read Sep 18, 2026.

SAVRN's Notes on xlm-roberta-large

Fitting this one is never the problem. At 16-bit the weights are 1.1 GB and the run needs 1.3 GB, so the cheapest setup on the page, one 192 GB MI300X at $1.85 an hour on-demand, has room for more than a hundred copies; put it on a GPU already serving something else. What the footprint buys is a fill-mask model trained on 2.5 TB of filtered CommonCrawl in 100 languages with a 250,002-entry vocabulary: a starting layer for multilingual text work you tune yourself, not a finished application.

MIT asks only that the copyright and permission notice travel with what you ship; commercial use, modification and redistribution are allowed. Note the 514-token context, sentences and short passages rather than documents, and the 17 files totaling 11.2 GB across five formats, of which your runtime needs one. The releasing team did not write the card; start from arXiv 1911.02116.

SAVRN Research, 2026-09-18

Model Card

By Facebook AI community, published under mit, revision c23d21b0620b.

XLM-RoBERTa (large-sized model)

XLM-RoBERTa model pre-trained on 2.5TB of filtered CommonCrawl data containing 100 languages. It was introduced in the paper Unsupervised Cross-lingual Representation Learning at Scale by Conneau et al. and first released in this repository.

Disclaimer: The team releasing XLM-RoBERTa did not write a model card for this model so this model card has been written by the Hugging Face team.

Model description

XLM-RoBERTa is a multilingual version of RoBERTa. It is pre-trained on 2.5TB of filtered CommonCrawl data containing 100 languages.

RoBERTa is a transformers model pretrained on a large corpus in a self-supervised fashion. This means it was pretrained on the raw texts only, with no humans labelling them in any way (which is why it can use lots of publicly available data) with an automatic process to generate inputs and labels from those texts.

Read the full model card (579 words)

Configuration

Architecture
XLMRobertaForMaskedLM
Context length (tokens)
514
Layers
24
Hidden size
1,024
Feed-forward size
4,096
Attention heads
16
Vocabulary size
250,002
Model type
xlm-roberta

Identity and Version

Repository
FacebookAI/xlm-roberta-large
Publisher
Facebook AI community
Task
Fill mask
Modality
Text
Library
transformers
Parameters
561M parameters
Languages
af, am, ar, as, az, be, bg, bn
Revision
c23d21b0620b635a76227c604d44e43a9f0ee389
First published
2022-03-02
Last updated
2024-02-19

Files and Weights

17 files, 11.2 GB in total. The weights are 5 files totalling 9.0 GB in bin, h5, msgpack, onnx, safetensors.

Weights5 files · 9.0 GB
Configuration3 files · 1.6 KB
Tokenizer4 files · 26.2 MB
Documentation1 file · 5.2 KB
Other3 files · 2.2 GB
Repository1 file · 623 B
Every file
FileTypeSizeSHA-256
flax_model.msgpackWeights2.2 GB 96d19a73ca04
model.safetensorsWeights2.2 GB 2dfa19f17241
onnx/model.onnxWeights545.9 KB bb5a52503a3e
pytorch_model.binWeights2.2 GB 01e55aa45dbb
tf_model.h5Weights2.2 GB a465c8d459fe
config.jsonConfiguration616 B —
onnx/config.jsonConfiguration681 B —
onnx/special_tokens_map.jsonConfiguration280 B —
README.mdDocumentation5.2 KB —
onnx/model.onnx_dataOther2.2 GB 1798dab29db9
onnx/sentencepiece.bpe.modelOther5.1 MB cfc8146abe2a
sentencepiece.bpe.modelOther5.1 MB —
.gitattributesRepository623 B —
onnx/tokenizer.jsonTokenizer17.1 MB 62c24cdc13d4
onnx/tokenizer_config.jsonTokenizer418 B —
tokenizer.jsonTokenizer9.1 MB —
tokenizer_config.jsonTokenizer25 B —

License and Download

License
mit
Access
Open weights, no gate
Download size
9.0 GB
Download from Facebook AI community

Released by Facebook AI community through its official repository on Hugging Face. Read the license.

Built From

Memory Requirements

PrecisionWeights in memory
As published9.0 GB
16-bit1.1 GB
8-bit0.6 GB
4-bit0.3 GB

Weights only, from the published parameter count; the key-value cache and runtime add to this.

Compare xlm-roberta-large

Questions About xlm-roberta-large

How much GPU memory does xlm-roberta-large need?

About 1.3 GB at 16-bit and 0.3 GB at 4-bit: the weights (561M parameters) plus a working margin. A long context needs more.

What is the cheapest GPU to run xlm-roberta-large on?

At 16-bit, 1x MI300X from $1.85 an hour; at 4-bit, 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists.

Can I use xlm-roberta-large commercially?

Yes. xlm-roberta-large is released under MIT License. The MIT License is a short permissive license. It permits commercial use, modification and redistribution, provided the copyright notice and permission notice are included.

What is xlm-roberta-large's context length?

514 tokens, from the maximum position embeddings in its published configuration.

Similar Models

Model · Fill mask

esm2_t33_650M_UR50D

AI at Meta

ESM-2 is a state-of-the-art protein model trained on a masked language modelling objective. It is suitable for fine-tuning on a wide range of tasks that take protein sequences as input. For detailed information on the model architecture and training data, please refer to the accompanying paper. You may also be interested in some demo notebooks (PyTorch, TensorFlow) which demonstrate how to fine-tune ESM-2 models on your tasks of interest. Several ESM-2 checkpoints are available in the Hub with varying sizes. Larger sizes generally have somewhat better accuracy, but require much more memory and time to train

Open weights mit 652M parameters 1,026 tokens transformers
View model

Model · Fill mask

ModernBERT-large

Answer.AI

ModernBERT is a modernized bidirectional encoder-only Transformer model (BERT-style) pre-trained on 2 trillion tokens of English and code data with a native context length of up to 8,192 tokens. ModernBERT leverages recent architectural improvements such as: - Rotary Positional Embeddings (RoPE) for long-context support. - Local-Global Alternating Attention for efficiency on long inputs. - Unpadding and Flash Attention for efficient inference. ModernBERT’s native long context length makes it ideal for tasks that require processing long documents, such as retrieval, classification, and semantic search within large corpora. The model was trained on a large corpus of text and code, making it…

Open weights apache-2.0 396M parameters 8,192 tokens transformers
View model

Pretrained model on English language using a masked language modeling (MLM) objective. It was introduced in this paper and first released in this repository. This model is case-sensitive: it makes a difference between english and English. Disclaimer: The team releasing RoBERTa did not write a model card for this model so this model card has been written by the Hugging Face team. RoBERTa is a transformers model pretrained on a large corpus of English data in a self-supervised fashion. This means it was pretrained on the raw texts only, with no humans labelling them in any way (which is why it can use lots of publicly available data) with an automatic process to generate inputs and labels…

Open weights mit 355M parameters 514 tokens transformers
View model

Model · Fill mask

bert-large-uncased

BERT community

Pretrained model on English language using a masked language modeling (MLM) objective. It was introduced in this paper and first released in this repository. This model is uncased: it does not make a difference between english and English. Disclaimer: The team releasing BERT did not write a model card for this model so this model card has been written by the Hugging Face team. BERT is a transformers model pretrained on a large corpus of English data in a self-supervised fashion. This means it was pretrained on the raw texts only, with no humans labelling them in any way (which is why it can use lots of publicly available data) with an automatic process to generate inputs and labels from…

Open weights apache-2.0 336M parameters 512 tokens transformers
View model

This is a Japanese DeBERTa V2 large model pre-trained on Japanese Wikipedia, the Japanese portion of CC-100, and the Japanese portion of OSCAR. This model is trained with character-level tokenization and whole word masking. You can use this model for masked language modeling as follows: You can also fine-tune this model on downstream tasks. There is no need to tokenize texts in advance, and you can give raw texts to the tokenizer. The texts are tokenized into character-level tokens by sentencepiece. We used the following corpora for pre-training: - Japanese Wikipedia (as of 20221020, 3.2GB, 27M sentences, 1.3M documents) - Japanese portion of CC-100 (85GB, 619M sentences, 66M documents)…

Open weights cc-by-sa-4.0 330M parameters 512 tokens transformers
View model

XLM-RoBERTa model pre-trained on 2.5TB of filtered CommonCrawl data containing 100 languages. It was introduced in the paper Unsupervised Cross-lingual Representation Learning at Scale by Conneau et al. and first released in this repository. Disclaimer: The team releasing XLM-RoBERTa did not write a model card for this model so this model card has been written by the Hugging Face team. XLM-RoBERTa is a multilingual version of RoBERTa. It is pre-trained on 2.5TB of filtered CommonCrawl data containing 100 languages. RoBERTa is a transformers model pretrained on a large corpus in a self-supervised fashion. This means it was pretrained on the raw texts only, with no humans labelling them in…

Open weights mit 279M parameters 514 tokens transformers
View model