Documentation
      ¶
    
    
  
    
  
    Overview ¶
Package Learning implements the learning stage of the Neurlang classifier
Index ¶
- type HyperParameters
 - func (h *HyperParameters) Reducing(alphabet [2][]uint32, untilMaxl, initMaxx uint32) [][2]uint32
 - func (h *HyperParameters) SetLogger(filename string)
 - func (h *HyperParameters) Solve(d datasets.SplittedDataset) (int, *hashtron.Hashtron)
 - func (h *HyperParameters) Training(d datasets.Splitter) (*hashtron.Hashtron, error)
 
Constants ¶
This section is empty.
Variables ¶
This section is empty.
Functions ¶
This section is empty.
Types ¶
type HyperParameters ¶
type HyperParameters struct {
	Threads int // number of threads for learning
	Shuffle bool // whether to shuffle the set before each learning attempt
	Seed    bool // seed prng using true rng
	Printer       uint32 // print when hit conflicting solution of at least this large size
	DeadlineMs    int    // deadline in milliseconds to throw away incomplete solution attempt
	DeadlineRetry int    // retry from scratch after this many failed deadlines
	// Factor is how hard to try to come up with a smaller solution (default: 1)
	// Usually set equal to Subtractor
	Factor uint32
	// Subtractor is how hard to try to come up with a smaller solution (default: 1)
	// Usually set equal to Factor
	Subtractor uint32
	InitialLimit int // initial limit of how small the solution must be to be saved to disk
	DisableProgressBar bool // disable progress bar
	EndWhenSolved bool // end when solved
	Name string // override model name
	EOL  []byte // override EOL string
	Bits byte // override bits
	// contains filtered or unexported fields
}
    func (*HyperParameters) Reducing ¶ added in v0.1.4
func (h *HyperParameters) Reducing(alphabet [2][]uint32, untilMaxl, initMaxx uint32) [][2]uint32
Reducing could mutate the alphabet, and it runs only till untilMaxl if nonzero, starting from initMaxx if nonzero
func (*HyperParameters) SetLogger ¶
func (h *HyperParameters) SetLogger(filename string)
SetLogger sets the output logger file where hashtron golang code programs are written
func (*HyperParameters) Solve ¶
func (h *HyperParameters) Solve(d datasets.SplittedDataset) (int, *hashtron.Hashtron)
Solve directly solves a single hashtron on a splitted dataset d. It outputs the size of solution and the trained hashtron if successful. Most callers should use Training instead.
      
      Directories
      ¶
    
    | Path | Synopsis | 
|---|---|
| 
       Package AVX implements the learning stage of the Neurlang classifier on AVX 
         | 
      Package AVX implements the learning stage of the Neurlang classifier on AVX | 
| 
       Package Cu implements the learning stage of the Neurlang classifier on CUDA 
         | 
      Package Cu implements the learning stage of the Neurlang classifier on CUDA | 
 Click to show internal directories. 
   Click to hide internal directories.