AI Models - Computer Vision, Conversational AI, and More

Welcome Guest

Many AI applications have common needs: classification, object detection, language translation, text-to-speech, recommender engines, sentiment analysis, and more. When developing applications with these capabilities, it is much faster to start with a model that is pre-trained and then tune it for a specific use case. The NGC catalog offers pre-trained models for a variety of common AI tasks that are optimized for NVIDIA Tensor Core GPUs, and can be easily re-trained by updating just a few layers, saving valuable time.

Sort: Last Modified

Riva TTS English Normalization Grammar

Model

Base English grammar

RMIR Conformer Hindi (hi-IN) Streaming Througput

Model

Conformer trained on Riva ASR-set 1.0

RMIR Conformer Hindi (hi-IN) Streaming Throughput

Model

Conformer trained on Riva ASR-set 1.0 with 3 gram LM

RMIR Conformer Hindi (hi-IN) Streaming

Model

Conformer trained on Riva ASR-set 1.0 with 3 gram LM

RMIR Conformer Hindi (hi-IN) Streaming

Model

Conformer trained on Riva ASR-set 1.0 with 3 gram LM

RMIR Conformer Hindi (hi-IN) Offline

Model

Conformer trained on Riva ASR-set 1.0

RMIR Conformer Hindi (hi-IN) Offline

Model

Conformer trained on Riva ASR-set 1.0 with 3 gram LM

RIVA Punctuation and Capitalization for Mandarin

Model

For each word in the input text, the model: 1) predicts a punctuation mark that should follow the word (if any), the model supports commas, periods and question marks) and 2) predicts if the word should be capitalized or not.

RIVA Conformer ASR Hindi

Model

Hindi Conformer ASR model trained on ASR set 1.0

FinMegatron345m-gpt2-bpe

Model

FSI : Financial Megatron GPT2 345m parameters model with BPE tokenizer, gpt vocabulary and merge file, pre-trained on subsets of CC-100 text corpus.

FinMegatron345m-uncased

Model

FSI : Financial Megatron 345m parameters model with bert vocabulary (28k size) uncased, pre-trained on subsets of CC-100 text corpus.

STT En ContextNet 1024

Model

ContextNet-1024 model for English Automatic Speech Recognition, trained on NeMo ASRSET

Megatron GPT2 345M

Model

345M parameter GPT generative Megatron model

Megatron-BERT 345M Cased

Model

345M parameter BERT Megatron model with cased vocab

Megatron-BERT 345M Uncased

Model

345M parameter BERT Megatron model with uncased vocab

BioMegatron345mCased

Model

Megatron pretrained on cased biomedical dataset PubMed with 345 million parameters.

BioMegatron345mUncased

Model

Megatron pretrained on uncased biomedical dataset PubMed with 345 million parameters.

BioMegatron345m-biovocab-50k-uncased

Model

Megatron 345m parameters model with biomedical vocabulary (50k size) uncased, pre-trained on PubMed biomedical text corpus.

BioMegatron345m-biovocab-30k-uncased

Model

Megatron 345m parameters model with biomedical vocabulary (30k size) uncased, pre-trained on PubMed biomedical text corpus.

BioMegatron345m-biovocab-50k-cased

Model

Megatron 345m parameters model with biomedical vocabulary (50k size) cased, pre-trained on PubMed biomedical text corpus.

BioMegatron345m-biovocab-30k-cased

Model

Megatron 345m parameters model with biomedical vocabulary (30k size) cased, pre-trained on PubMed biomedical text corpus.

Action Recognition Net

Model

5 class action recognition network to recognize what people do in an image.

Joint Intent and Slot Classification Misty Bert

Model

Intent and Slot classification of the queries for the misty bot with BERT model trained on weather, smalltalk and POI (places of interest) data.

GPUNet-P1 pretrained weights (PyTorch, AMP, ImageNet)

Model

GPUNet-P1 ImageNet pretrained weights

GPUNet-P0 pretrained weights (PyTorch, AMP, ImageNet)

Model

GPUNet-P0 ImageNet pretrained weights