ContainersNVIDIADeepSeek-R1NVIDIALlama-3.1-Nemotron-70B-InstructNVIDIAPyTorch
CollectionsNVIDIAOmniverse Kit (FB)NVIDIADeepStream SDKNVIDIAOmniverse Kit App Streaming
ModelsNVIDIAStyleGAN3 pretrained modelsNVIDIAPeopleNetNVIDIATrafficCamNet
ResourcesNVIDIARiva Skills Quick Start
Helm ChartsNVIDIAGPU OperatorNVIDIANVIDIA NIM Operator
NVIDIA
NVIDIA
BertBaseCasedForNeMo
Model
NVIDIA
NVIDIA
BertBaseCasedForNeMo

BERT Base Model trained with NeMo on cased Wikipedia and BookCorpus on a sequence length of 512.

2 Versions
1
03/26/2020 8:38 PM UTC500.54 MBAccuracy: 00 EpochsBatch Size: 0GPU: V100
Finetuning Results
KeyValue
GLUE MRPC ACCURACY85.05
GLUE MRPC F189.32
SQUADV1.1 EM77.42
SQUADV1.1 F184.86
SQUADV2.0 EM66.87
SQUADV2.0 F169.82
Pretraining Setup
KeyValue
AMP OPTIMIZATION LEVELO1
BATCH SIZE PER GPU8
LEARNING RATE0.4375E-4
NUMBER OF GPUS8
NUMBER OF ITERATIONS2285714
ONLY MASKED LANGUAGE MODEL LOSSYES
2Selected
03/21/2020 12:42 AM UTC500.54 MBGPU: V100
Metrics
KeyValue
Batch size per GPU8
Number of GPUs used for training8
Number of training steps2285714
SQuADv1.1 EM77.42
SQuADv1.1 F184.86
SQuADv2.0 EM66.87
SQuADv2.0 F169.82

NVIDIA uses cookies to improve your experience on our web site. We and our third-party partners also use cookies and other tools to collect and record information you provide as well as information about your interactions with our websites for performance improvement, analytics, and to assist in marketing efforts. By clicking "Accept All", you consent to our use of cookies and other tools as described in our Cookie Policy. You can manage your cookie settings by clicking on "Manage Settings." By continuing to use this site or by clicking one of the buttons below, you agree to our Terms of Service (which contains important waivers). Please see our Privacy Policy for more information on our privacy practices.