Hugging Face
Models
Datasets
Spaces
Posts
Docs
Enterprise
Pricing
Log In
Sign Up
mosaicml
/
mosaic-bert-base
like
46
Follow
Mosaic ML, Inc.
363
Fill-Mask
Transformers
PyTorch
c4
English
bert
custom_code
arxiv:
8 papers
License:
apache-2.0
Model card
Files
Files and versions
Community
12
Train
Deploy
Use this model
8a9076d
mosaic-bert-base
4 contributors
History:
28 commits
jacobfulano
Update detail about Triton Flash Attention with ALiBi implementation
8a9076d
12 months ago
.gitattributes
1.48 kB
initial commit
almost 2 years ago
README.md
14.6 kB
Update detail about Triton Flash Attention with ALiBi implementation
12 months ago
bert_layers.py
47.3 kB
Upload BertForMaskedLM
almost 2 years ago
bert_padding.py
6.26 kB
Upload BertForMaskedLM
almost 2 years ago
config.json
827 Bytes
Change attention_probs_dropout_prob to 0.1 so that FlashAttention/triton dependencies are avoided
about 1 year ago
configuration_bert.py
1.02 kB
Upload BertForMaskedLM
almost 2 years ago
flash_attn_triton.py
42.7 kB
Upload BertForMaskedLM
almost 2 years ago
generation_config.json
90 Bytes
Upload BertForMaskedLM
almost 2 years ago
pytorch_model.bin
pickle
Detected Pickle imports (3)
"torch._utils._rebuild_tensor_v2"
,
"collections.OrderedDict"
,
"torch.FloatStorage"
What is a pickle import?
550 MB
LFS
Upload BertForMaskedLM
almost 2 years ago