ModernBertMaskedLM classkeras_hub.models.ModernBertMaskedLM(
backbone, preprocessor=None, dtype=None, **kwargs
)
ModernBERT Masked LM task model.
The Masked LM model provides a prediction head for the Masked Language
Modeling task. It is composed of a keras_hub.models.ModernBertBackbone
and a prediction head which projects the backbone's hidden states back
to the vocabulary space.
This model can be used for pre-training or fine-tuning on a specific corpus.
Arguments
keras_hub.models.ModernBertBackbone instance.keras_hub.models.ModernBertMaskedLMPreprocessor
instance or None. If None, this model will not handle input
preprocessing automatically during fit(), predict(), or
evaluate().keras.DTypePolicy. The precision policy used
for the model's computations and weights. If None, defaults
to the backbone's dtype policy.Examples
import keras_hub
import numpy as np
# Pre-trained backbone and preprocessor
tokenizer = keras_hub.models.ModernBertTokenizer(
vocabulary="vocab.json",
merges="merges.txt",
)
preprocessor = keras_hub.models.ModernBertMaskedLMPreprocessor(
tokenizer=tokenizer,
sequence_length=128,
)
backbone = keras_hub.models.ModernBertBackbone(
vocabulary_size=50368,
hidden_dim=768,
intermediate_dim=1152,
num_layers=22,
num_heads=12,
)
# Instantiate the MaskedLM task model
masked_lm = keras_hub.models.ModernBertMaskedLM(
backbone=backbone,
preprocessor=preprocessor,
)
# Predict on raw text strings
raw_data = ["The quick brown fox [MASK] over the lazy dog."]
predictions = masked_lm.predict(raw_data)
from_preset methodModernBertMaskedLM.from_preset(preset, load_weights=True, **kwargs)
Instantiate a keras_hub.models.Task from a model preset.
A preset is a directory of configs, weights and other file assets used
to save and load a pre-trained model. The preset can be passed as
one of:
'bert_base_en''kaggle://user/bert/keras/bert_base_en''hf://user/bert_base_en''./bert_base_en'For any Task subclass, you can run cls.presets.keys() to list all
built-in presets available on the class.
This constructor can be called in one of two ways. Either from a task
specific base class like keras_hub.models.CausalLM.from_preset(), or
from a model class like
keras_hub.models.BertTextClassifier.from_preset().
If calling from the a base class, the subclass of the returning object
will be inferred from the config in the preset directory.
Arguments
True, saved weights will be loaded into
the model architecture. If False, all weights will be
randomly initialized.Examples
# Load a Gemma generative task.
causal_lm = keras_hub.models.CausalLM.from_preset(
"gemma_2b_en",
)
# Load a Bert classification task.
model = keras_hub.models.TextClassifier.from_preset(
"bert_base_en",
num_classes=2,
)
| Preset | Parameters | Description |
|---|---|---|
| modern_bert_base_en | 149.01M | 22-layer ModernBERT Base encoder model pretrained on English for masked language modeling. Uses Rotary Position Embeddings (RoPE), alternating local and global attention, and GeGLU feedforward layers. |
| modern_bert_embed_base | 149.01M | 22-layer ModernBERT Base sentence embedding model from Nomic AI (nomic-ai/modernbert-embed-base). Produces 768-dimensional, L2-normalized, mean-pooled embeddings for inputs of up to 8192 tokens. Prefix queries with 'search_query: ' and documents with 'search_document: '. |
| gte_modern_bert_base | 149.01M | 22-layer ModernBERT Base sentence embedding model from Alibaba-NLP (Alibaba-NLP/gte-modernbert-base). Produces 768-dimensional [CLS]-pooled embeddings that are not normalized, so compare them with cosine similarity. The model supports inputs of up to 8192 tokens. |
| modern_bert_large_en | 394.78M | 28-layer ModernBERT Large encoder model pretrained on English for masked language modeling. Uses Rotary Position Embeddings (RoPE), alternating local and global attention, and GeGLU feedforward layers. |
backbone propertykeras_hub.models.ModernBertMaskedLM.backbone
A keras_hub.models.Backbone model with the core architecture.
preprocessor propertykeras_hub.models.ModernBertMaskedLM.preprocessor
A keras_hub.models.Preprocessor layer used to preprocess input.