Megatron-DeepSpeed LLAMA-2 70B
LLAMA 13B parameter model based on Megatron DeepSpeed on Gaudi2
GPU Migration Examples
Groups of models migrated to Gaudi HPU using GPU Migration toolkit.
DeepSpeed Chat
How to use the 3 stages of Fine Tuning with Real Life Human Feedback, supporting Llama 2 model
Hugging Face Stable Diffusion
Optimum Habana Stable Diffusion
Stable Diffusion Models for Inference
CompVis and Stability AI based models for Stable Diffusion Inference
Stable Diffusion Training
CompVis based Stable Diffusion Model for Training
BLOOM 176B Inference
HuggingFace BLOOM model for Inference on Gaudi2, using DeepSpeed for Inference
Megatron-DeepSpeed BLOOM 13B
GPT based model with 13B parameters based on Megatron DeepSpeed on Gaudi2
BLOOM 7B
Large Language Model for Inference
Hugging Face BERT large model (uncased) whole word masking
Optimum Habana BERT large model (uncased) whole word masking for PyTorch
Hugging Face RoBERTa
Optimum Habana RoBERTa for PyTorch
Hugging Face BERT
Optimum Habana BERT for PyTorch
Hugging Face ALBERT Large
Optimum Habana ALBERT Large for PyTorch
Hugging Face Swin Transformer
Optimum Habana Swin Transformer for PyTorch
Hugging Face GPT2
Optimum Habana GPT2 for PyTorch
Hugging Face ALBERT XXLarge
Optimum Habana ALBERT XXLarge for PyTorch
Hugging Face RoBERTa large
Optimum Habana RoBERTa large for PyTorch
Hugging Face DistilBERT
Optimum Habana DistilBERT for PyTorch
Hugging Face T5
Optimum Habana T5 for PyTorch