Safe Haskell	Safe-Inferred
Language	Haskell2010

Torch.GraduallyTyped.NN.Transformer.GStack

Synopsis

newtype GTransformerStack (stack :: Type) where
- GTransformerStack :: forall stack. stack -> GTransformerStack stack
type family EncoderStackF (style :: TransformerStyle) (numLayers :: Nat) (gradient :: Gradient RequiresGradient) (device :: Device (DeviceType Nat)) (dataType :: DataType DType) (headDim :: Dim (Name Symbol) (Size Nat)) (headEmbedDim :: Dim (Name Symbol) (Size Nat)) (embedDim :: Dim (Name Symbol) (Size Nat)) (queryEmbedDim :: Dim (Name Symbol) (Size Nat)) (ffnDim :: Dim (Name Symbol) (Size Nat)) (hasDropout :: HasDropout) where ...
encoderStackSpec :: forall style numLayers gradient device dataType headDim headEmbedDim embedDim queryEmbedDim ffnDim hasDropout. STransformerStyle style -> SNat numLayers -> SGradient gradient -> SDevice device -> SDataType dataType -> SDim headDim -> SDim headEmbedDim -> SDim embedDim -> SDim queryEmbedDim -> SDim ffnDim -> SHasDropout hasDropout -> Double -> Double -> ModelSpec (EncoderStackF style numLayers gradient device dataType headDim headEmbedDim embedDim queryEmbedDim ffnDim hasDropout)
type family DecoderStackF (style :: TransformerStyle) (numLayers :: Nat) (gradient :: Gradient RequiresGradient) (device :: Device (DeviceType Nat)) (dataType :: DataType DType) (headDim :: Dim (Name Symbol) (Size Nat)) (headEmbedDim :: Dim (Name Symbol) (Size Nat)) (embedDim :: Dim (Name Symbol) (Size Nat)) (queryEmbedDim :: Dim (Name Symbol) (Size Nat)) (keyEmbedDim :: Dim (Name Symbol) (Size Nat)) (ffnDim :: Dim (Name Symbol) (Size Nat)) (hasDropout :: HasDropout) where ...
decoderStackSpec :: forall style numLayers gradient device dataType headDim headEmbedDim embedDim queryEmbedDim keyEmbedDim ffnDim hasDropout. STransformerStyle style -> SNat numLayers -> SGradient gradient -> SDevice device -> SDataType dataType -> SDim headDim -> SDim headEmbedDim -> SDim embedDim -> SDim queryEmbedDim -> SDim keyEmbedDim -> SDim ffnDim -> SHasDropout hasDropout -> Double -> Double -> ModelSpec (DecoderStackF style numLayers gradient device dataType headDim headEmbedDim embedDim queryEmbedDim keyEmbedDim ffnDim hasDropout)

Documentation

newtype GTransformerStack (stack :: Type) where Source #

Generic transformer stack.

stack is a stack of tranformer blocks.

Constructors

GTransformerStack :: forall stack. stack -> GTransformerStack stack

Instances

Instances details

Generic (GTransformerStack stack) Source #
Instance details Defined in Torch.GraduallyTyped.NN.Transformer.GStack Associated Types type Rep (GTransformerStack stack) :: Type -> Type Source # Methods from :: GTransformerStack stack -> Rep (GTransformerStack stack) x Source # to :: Rep (GTransformerStack stack) x -> GTransformerStack stack Source #
Show stack => Show (GTransformerStack stack) Source #
Instance details Defined in Torch.GraduallyTyped.NN.Transformer.GStack Methods showsPrec :: Int -> GTransformerStack stack -> ShowS Source # show :: GTransformerStack stack -> String Source # showList :: [GTransformerStack stack] -> ShowS Source #
Eq stack => Eq (GTransformerStack stack) Source #
Instance details Defined in Torch.GraduallyTyped.NN.Transformer.GStack Methods (==) :: GTransformerStack stack -> GTransformerStack stack -> Bool Source # (/=) :: GTransformerStack stack -> GTransformerStack stack -> Bool Source #
Ord stack => Ord (GTransformerStack stack) Source #
Instance details Defined in Torch.GraduallyTyped.NN.Transformer.GStack Methods compare :: GTransformerStack stack -> GTransformerStack stack -> Ordering Source # (<) :: GTransformerStack stack -> GTransformerStack stack -> Bool Source # (<=) :: GTransformerStack stack -> GTransformerStack stack -> Bool Source # (>) :: GTransformerStack stack -> GTransformerStack stack -> Bool Source # (>=) :: GTransformerStack stack -> GTransformerStack stack -> Bool Source # max :: GTransformerStack stack -> GTransformerStack stack -> GTransformerStack stack Source # min :: GTransformerStack stack -> GTransformerStack stack -> GTransformerStack stack Source #
HasStateDict block => HasStateDict (GTransformerStack (Vector numLayers block)) Source #
Instance details Defined in Torch.GraduallyTyped.NN.Transformer.GStack Methods fromStateDict :: (MonadIO m, MonadThrow m, MonadState StateDict m) => ModelSpec (GTransformerStack (Vector numLayers block)) -> StateDictKey -> m (GTransformerStack (Vector numLayers block)) Source # toStateDict :: (MonadThrow m, MonadState StateDict m) => StateDictKey -> GTransformerStack (Vector numLayers block) -> m () Source #
(HasInitialize block generatorDevice block' generatorDevice, numLayers' ~ (numLayers + 1)) => HasInitialize (GTransformerStack (Vector numLayers' block)) generatorDevice (GTransformerStack (Vector numLayers' block')) generatorDevice Source #
Instance details Defined in Torch.GraduallyTyped.NN.Transformer.GStack Methods initialize :: MonadThrow m => ModelSpec (GTransformerStack (Vector numLayers' block)) -> Generator generatorDevice -> m (GTransformerStack (Vector numLayers' block'), Generator generatorDevice) Source #
(HasForward block (query, attentionBias) generatorDevice output generatorOutputDevice, HasForward block (output, attentionBias) generatorOutputDevice output generatorOutputDevice) => HasForward (GTransformerStack (Vector n block)) (query, attentionBias) generatorDevice output generatorOutputDevice Source #	`HasForward` instance for `GTransformerStack` in an encoder configuration. ┌───────┐ ┌───────────────┐ │ query │ │ attentionBias │ └───┬───┘ └───────┬───────┘ │ │ ▼ │ block◄───────────┤ ▼ │ block◄───────────┤ ▼ │ ... ... ▼ │ block◄───────────┘ │ ▼ ┌───────┐ │ query │ └───────┘
Instance details Defined in Torch.GraduallyTyped.NN.Transformer.GStack Methods forward :: MonadThrow m => GTransformerStack (Vector n block) -> (query, attentionBias) -> Generator generatorDevice -> m (output, Generator generatorOutputDevice) Source #
HasForward (GTransformerStack (Vector 0 block)) (query, attentionBias) generatorDevice query generatorDevice Source #
Instance details Defined in Torch.GraduallyTyped.NN.Transformer.GStack Methods forward :: MonadThrow m => GTransformerStack (Vector 0 block) -> (query, attentionBias) -> Generator generatorDevice -> m (query, Generator generatorDevice) Source #
HasForward block (query, attentionBias) generatorDevice output generatorOutputDevice => HasForward (GTransformerStack (Vector 1 block)) (query, attentionBias) generatorDevice output generatorOutputDevice Source #
Instance details Defined in Torch.GraduallyTyped.NN.Transformer.GStack Methods forward :: MonadThrow m => GTransformerStack (Vector 1 block) -> (query, attentionBias) -> Generator generatorDevice -> m (output, Generator generatorOutputDevice) Source #
(HasForward block (query, key, attentionBias, crossAttentionBias) generatorDevice output generatorOutputDevice, HasForward block (output, key, attentionBias, crossAttentionBias) generatorOutputDevice output generatorOutputDevice) => HasForward (GTransformerStack (Vector n block)) (query, key, attentionBias, crossAttentionBias) generatorDevice output generatorOutputDevice Source #	`HasForward` instance for `GTransformerStack` in a decoder configuration. ┌───────┐ ┌─────┐ ┌───────────────┐ ┌────────────────────┐ │ query │ │ key │ │ attentionBias │ │ crossAttentionBias │ └───┬───┘ └──┬──┘ └───────┬───────┘ └─────────┬──────────┘ │ │ │ │ ▼ │ │ │ block◄──────┤◄────────────┤◄───────────────────┤ ▼ │ │ │ block◄──────┤◄────────────┤◄───────────────────┤ ▼ │ │ │ ... ... ... ... ▼ │ │ │ block◄──────┘◄────────────┘◄───────────────────┘ │ ▼ ┌───────┐ │ query │ └───────┘
Instance details Defined in Torch.GraduallyTyped.NN.Transformer.GStack Methods forward :: MonadThrow m => GTransformerStack (Vector n block) -> (query, key, attentionBias, crossAttentionBias) -> Generator generatorDevice -> m (output, Generator generatorOutputDevice) Source #
HasForward (GTransformerStack (Vector 0 block)) (query, key, attentionBias, crossAttentionBias) generator query generator Source #
Instance details Defined in Torch.GraduallyTyped.NN.Transformer.GStack Methods forward :: MonadThrow m => GTransformerStack (Vector 0 block) -> (query, key, attentionBias, crossAttentionBias) -> Generator generator -> m (query, Generator generator) Source #
HasForward block (query, key, attentionBias, crossAttentionBias) generatorDevice output generatorOutputDevice => HasForward (GTransformerStack (Vector 1 block)) (query, key, attentionBias, crossAttentionBias) generatorDevice output generatorOutputDevice Source #
Instance details Defined in Torch.GraduallyTyped.NN.Transformer.GStack Methods forward :: MonadThrow m => GTransformerStack (Vector 1 block) -> (query, key, attentionBias, crossAttentionBias) -> Generator generatorDevice -> m (output, Generator generatorOutputDevice) Source #
type Rep (GTransformerStack stack) Source #
Instance details Defined in Torch.GraduallyTyped.NN.Transformer.GStack type Rep (GTransformerStack stack) = D1 ('MetaData "GTransformerStack" "Torch.GraduallyTyped.NN.Transformer.GStack" "hasktorch-gradually-typed-0.2.0.0-1KV1aIPzzbp6JpSr37tC1K" 'True) (C1 ('MetaCons "GTransformerStack" 'PrefixI 'False) (S1 ('MetaSel ('Nothing :: Maybe Symbol) 'NoSourceUnpackedness 'NoSourceStrictness 'DecidedLazy) (Rec0 stack)))
type ModelSpec (GTransformerStack stack) Source #
Instance details Defined in Torch.GraduallyTyped.NN.Transformer.GStack type ModelSpec (GTransformerStack stack) = GTransformerStack (ModelSpec stack)

type family EncoderStackF (style :: TransformerStyle) (numLayers :: Nat) (gradient :: Gradient RequiresGradient) (device :: Device (DeviceType Nat)) (dataType :: DataType DType) (headDim :: Dim (Name Symbol) (Size Nat)) (headEmbedDim :: Dim (Name Symbol) (Size Nat)) (embedDim :: Dim (Name Symbol) (Size Nat)) (queryEmbedDim :: Dim (Name Symbol) (Size Nat)) (ffnDim :: Dim (Name Symbol) (Size Nat)) (hasDropout :: HasDropout) where ... Source #

Equations

EncoderStackF style numLayers gradient device dataType headDim headEmbedDim embedDim queryEmbedDim ffnDim hasDropout = GTransformerStack (Vector numLayers (EncoderBlockF style gradient device dataType headDim headEmbedDim embedDim queryEmbedDim ffnDim hasDropout))

encoderStackSpec :: forall style numLayers gradient device dataType headDim headEmbedDim embedDim queryEmbedDim ffnDim hasDropout. STransformerStyle style -> SNat numLayers -> SGradient gradient -> SDevice device -> SDataType dataType -> SDim headDim -> SDim headEmbedDim -> SDim embedDim -> SDim queryEmbedDim -> SDim ffnDim -> SHasDropout hasDropout -> Double -> Double -> ModelSpec (EncoderStackF style numLayers gradient device dataType headDim headEmbedDim embedDim queryEmbedDim ffnDim hasDropout) Source #

Specifies the parameters of a transformer stack in an encoder configuration.

style: the style of the transformer stack, e.g. ST5, SByT5, etc.
gradient: whether to compute the gradient of the stack's parameters.
device: the computational device on which the stack is allocated.
dataType: the data type of the stack's parameters.
headDim: the dimension of all transformer heads in the stack.
headEmbedDim: the dimension of the transformer head embeddings.
embedDim: the dimension of the transformer embeddings.
queryEmbedDim: the dimension of the transformer query embeddings.
ffnDim: the dimension of the feed-forward network.
dropoutP: the dropout rate.
eps: the epsilon value for numerical stability of the layer normalization.

type family DecoderStackF (style :: TransformerStyle) (numLayers :: Nat) (gradient :: Gradient RequiresGradient) (device :: Device (DeviceType Nat)) (dataType :: DataType DType) (headDim :: Dim (Name Symbol) (Size Nat)) (headEmbedDim :: Dim (Name Symbol) (Size Nat)) (embedDim :: Dim (Name Symbol) (Size Nat)) (queryEmbedDim :: Dim (Name Symbol) (Size Nat)) (keyEmbedDim :: Dim (Name Symbol) (Size Nat)) (ffnDim :: Dim (Name Symbol) (Size Nat)) (hasDropout :: HasDropout) where ... Source #

Equations

DecoderStackF style numLayers gradient device dataType headDim headEmbedDim embedDim queryEmbedDim keyEmbedDim ffnDim hasDropout = GTransformerStack (Vector numLayers (DecoderBlockF style gradient device dataType headDim headEmbedDim embedDim queryEmbedDim keyEmbedDim ffnDim hasDropout))

decoderStackSpec :: forall style numLayers gradient device dataType headDim headEmbedDim embedDim queryEmbedDim keyEmbedDim ffnDim hasDropout. STransformerStyle style -> SNat numLayers -> SGradient gradient -> SDevice device -> SDataType dataType -> SDim headDim -> SDim headEmbedDim -> SDim embedDim -> SDim queryEmbedDim -> SDim keyEmbedDim -> SDim ffnDim -> SHasDropout hasDropout -> Double -> Double -> ModelSpec (DecoderStackF style numLayers gradient device dataType headDim headEmbedDim embedDim queryEmbedDim keyEmbedDim ffnDim hasDropout) Source #

Specifies the parameters of a transformer stack in a decoder configuration.

style: the style of the transformer stack, e.g. ST5, SByT5, etc.
gradient: whether to compute the gradient of the stack's parameters.
device: the computational device on which the stack is allocated.
dataType: the data type of the stack's parameters.
headDim: the dimension of all transformer heads in the stack.
headEmbedDim: the dimension of the transformer head embeddings.
embedDim: the dimension of the transformer embeddings.
queryEmbedDim: the dimension of the transformer query embeddings.
keyEmbedDim: the dimension of the transformer key embeddings.
ffnDim: the dimension of the feed-forward network.
dropoutP: the dropout rate.
eps: the epsilon value for numerical stability of the layer normalization.