Mila
Deep Neural Network Library
Loading...
Searching...
No Matches
Dnn.Components.GptTransformer Module Reference

Exported Modules

module  Dnn.Components.GptBlock
module  Dnn.Components.LayerNorm
module  Dnn.TensorDataType
module  Dnn.ComponentType
module  Dnn.TensorDataTypeTraits
module  Dnn.Components.Linear
module  Serialization.ModelArchive
module  Serialization.Tensor
module  Serialization.PretrainedReader
module  Compute.ExecutionContextFactory
module  Dnn.Component
module  Compute.CpuMemoryResource
module  Logging.Logger
module  Dnn.TensorTypes
module  Dnn.RuntimeMode
module  Dnn.Tensor
module  Compute.Device
module  Compute.DeviceId
module  Dnn.ModelType
module  Dnn.ActivationType
module  Compute.DeviceTypeTraits
module  Compute.ExecutionContext
module  Dnn.LanguageNetwork
module  Dnn.ITensor
module  Dnn.Components.Lpe
module  Compute.DeviceType
module  Serialization.Mode

Classes

class  Mila::Dnn::GptConfig
 Network-level configuration for GPT-style transformer networks. More...
class  Mila::Dnn::GptTransformer< TDeviceType, TPrecision >
 GPT-2 style transformer (decoder-only) for autoregressive token prediction. More...

Functions

GptConfig Mila::Dnn::GPT2_Small ()
 Usage Examples:
GptConfig Mila::Dnn::GPT2_Medium ()
 GPT-2 Medium (345M parameters).
GptConfig Mila::Dnn::GPT2_Large ()
 GPT-2 Large (774M parameters).
GptConfig Mila::Dnn::GPT2_XL ()
 GPT-2 XL (1.5B parameters).

Files

file  Mila/Src/Dnn/Components/Transformers/Gpt/GptTransformer.ixx
 GPT-2 style transformer network (decoder-only) for autoregressive language modeling.
file  Mila/Src/Dnn/Components/Transformers/Gpt/Gpt.Config.ixx
 Network-level configuration for GPT-style transformer networks.
file  Mila/Src/Dnn/Components/Transformers/Gpt/Gpt.Presets.ixx