tensorflow

TensorFlow bindings for OCaml
IN THIS PACKAGE
module Placeholder : sig ... end
type 't b = ?name:string -> 't Node.t -> 't Node.t -> 't Node.t
val (+) : [< `float | `double | `int32 | `int64 | `complex64 | `string ] b
val (-) : [< `float | `double | `int32 | `int64 | `complex64 ] b
val (/) : [< `float | `double | `int32 | `int64 | `complex64 ] b
val (*) : [< `float | `double | `int32 | `int64 | `complex64 ] b
val (*^) : [< `float | `double | `int32 | `complex64 ] b
val f_or_d : ?shape:int list -> type_:[< `float | `double ] as 'a Node.Type.t -> float -> 'a Node.t
val f : ?shape:int list -> float -> [ `float ] Node.t
val d : ?shape:int list -> float -> [ `double ] Node.t
val cf : ?shape:int list -> float list -> [ `float ] Node.t
val cd : ?shape:int list -> float list -> [ `double ] Node.t
val ci32 : ?shape:int list -> int list -> [ `int32 ] Node.t
val ci64 : ?shape:int list -> int list -> [ `int64 ] Node.t
val const_float : ?name:string -> ?control_inputs:Node.p list -> ?shape:int list -> type_:[< `float | `double ] as 'dtype Node.Type.t -> float list -> 'dtype Node.t
val const_int : ?name:string -> ?control_inputs:Node.p list -> ?shape:int list -> type_:[< `int32 | `int64 ] as 'dtype Node.Type.t -> int list -> 'dtype Node.t
val const_string : ?name:string -> ?shape:int list -> string list -> [ `string ] Node.t
val const_string0 : ?name:string -> string -> [ `string ] Node.t
val scalar : ?empty_shape:unit -> type_:[< `float | `double ] as 'dtype Node.Type.t -> float -> 'dtype Node.t
val four32 : [ `int32 ] Node.t
val three32 : [ `int32 ] Node.t
val two32 : [ `int32 ] Node.t
val one32 : [ `int32 ] Node.t
val zero32 : [ `int32 ] Node.t
type 'a reduce_fn = ?dims:int list -> [< `complex64 | `double | `float | `int32 | `int64 ] as 'a Node.t -> 'a Node.t
val reduce_sum : 'a reduce_fn
val reduce_min : 'a reduce_fn
val reduce_max : 'a reduce_fn
val reduce_mean : 'a reduce_fn
val reduce_prod : 'a reduce_fn
val reduce_all : ?dims:int list -> [ `bool ] Node.t -> [ `bool ] Node.t
val reduce_any : ?dims:int list -> [ `bool ] Node.t -> [ `bool ] Node.t
val save_ : ?name:string -> [ `string ] Node.t -> [ `string ] Node.t -> Node.p list -> [ `unit ] Node.t
val save : filename:string -> (string * Node.p) list -> [ `unit ] Node.t
val split2 : ?name:string -> [ `int32 ] Node.t -> 't Node.t -> 't Node.t * 't Node.t
val split3 : ?name:string -> [ `int32 ] Node.t -> 't Node.t -> 't Node.t * 't Node.t * 't Node.t
val split4 : ?name:string -> [ `int32 ] Node.t -> 't Node.t -> 't Node.t * 't Node.t * 't Node.t * 't Node.t
val range : [ `int32 ] Node.t -> [ `int32 ] Node.t
val placeholder : ?name:string -> type_:'a Node.Type.t -> int list -> 'a Placeholder.t
val dropout : [< `float | `double ] as 'a Node.t -> keep_prob:'a Node.t -> 'a Node.t
val cast : ?name:string -> 'srcT Node.t -> type_:'dstT Node.Type.t -> 'dstT Node.t
val count : 'a Node.t -> dims:int list -> [ `int32 ] Node.t
type 'a moments = {
mean : 'a Node.t;
variance : 'a Node.t;
}
val moments : [< `double | `float ] as 'a Node.t -> dims:int list -> 'a moments
val normalize : ?epsilon:float -> [< `double | `float ] as 'a Node.t -> 'a moments -> 'a Node.t
val cond_with_control_inputs : [ `bool ] Node.t -> if_true:( control_inputs:Node.p list -> 'a Node.t ) -> if_false:( control_inputs:Node.p list -> 'a Node.t ) -> 'a Node.t
val cond : [ `bool ] Node.t -> if_true:'a Node.t -> if_false:'a Node.t -> 'a Node.t
val shape32 : ?name:string -> ?control_inputs:Node.p list -> 't Node.t -> [ `int32 ] Node.t
val cross_entropy : ?epsilon:float -> ys:[< `double | `float ] as 'a Node.t -> y_hats:'a Node.t -> [ `sum | `mean ] -> 'a Node.t
val binary_cross_entropy : ?epsilon:float -> labels:[< `double | `float ] as 'a Node.t -> model_values:'a Node.t -> [ `sum | `mean ] -> 'a Node.t
val leaky_relu : [< `double | `float ] as 'a Node.t -> alpha:float -> 'a Node.t