Comments (5)
Thank you.
It looks good to me, but I am not an expert of this algo (still in the learning phase)
This would explain why I cannot get a “good player” in a reasonable time.
from agogo.
I tried to apply your suggestion (without further investigation), but it fails:
panic:
Non Differentiable WRTs:
map[
FilterLayer1 of Shared Layer 0 :: Tensor-4 float32:{} Aᵀ{0, 3, 1, 2}(%15)_γ :: Tensor-4 float32:{} Aᵀ{0, 3, 1, 2}(%15)_β :: Tensor-4 float32:{}
FilterLayer1 of Shared Layer 1 :: Tensor-4 float32:{} Aᵀ{0, 3, 1, 2}(%32)_γ :: Tensor-4 float32:{} Aᵀ{0, 3, 1, 2}(%32)_β :: Tensor-4 float32:{}
FilterLayer1 of Shared Layer 2 :: Tensor-4 float32:{} Aᵀ{0, 3, 1, 2}(%4f)_γ :: Tensor-4 float32:{} Aᵀ{0, 3, 1, 2}(%4f)_β :: Tensor-4 float32:{}]
from agogo.
Weird it seems to work from my side 😓 Can I take a look at your modification?
from agogo.
I made this patch:
@@ -65,9 +65,9 @@ func (m *maebe) res(input *G.Node, filterCount int, name string) (*G.Node, batch
}
func (m *maebe) share(input *G.Node, filterCount, layer int) (*G.Node, batchNormOp, batchNormOp) {
- layer1, l1Op := m.res(input, filterCount, fmt.Sprintf("Layer1 of Shared Layer %d", layer))
+ _, l1Op := m.res(input, filterCount, fmt.Sprintf("Layer1 of Shared Layer %d", layer))
layer2, l2Op := m.res(input, filterCount, fmt.Sprintf("Layer2 of Shared Layer %d", layer))
- added := m.do(func() (*G.Node, error) { return G.Add(layer1, layer2) })
+ added := m.do(func() (*G.Node, error) { return G.Add(input, layer2) })
retVal := m.rectify(added)
return retVal, l1Op, l2Op
}
and the learning function is basically:
func learn() error {
conf := agogo.Config{
Name: "Tic Tac Toe",
NNConf: dual.DefaultConf(3, 3, 10),
MCTSConf: mcts.DefaultConfig(3),
UpdateThreshold: 0.52,
}
conf.NNConf.BatchSize = 100
conf.NNConf.Features = 2 // write a better encoding of the board, and increase features (and that allows you to increase K as well)
conf.NNConf.K = 3
conf.NNConf.SharedLayers = 3
conf.MCTSConf = mcts.Config{
PUCT: 1.0,
M: 3,
N: 3,
Timeout: 50 * time.Millisecond,
PassPreference: mcts.DontPreferPass,
Budget: 1000,
DumbPass: true,
RandomCount: 0,
}
outEnc := NewEncoder()
go func(h http.Handler) {
mux := http.NewServeMux()
mux.Handle("/ws", h)
mux.Handle("/static/", http.StripPrefix("/static/", http.FileServer(http.Dir("./htdocs"))))
log.Println("go to http://localhost:8080/static")
http.ListenAndServe(":8080", mux)
}(outEnc)
conf.Encoder = encodeBoard
conf.OutputEncoder = outEnc
g := mnk.TicTacToe()
a := agogo.New(g, conf)
reader := bufio.NewReader(os.Stdin)
fmt.Print("press ented when ready")
reader.ReadString('\n')
//a.Learn(5, 30, 200, 30) // 5 epochs, 50 episode, 100 NN iters, 100 games.
err := a.Learn(5, 50, 100, 100) // 5 epochs, 50 episode, 100 NN iters, 100 games.
if err != nil {
return err
}
err = a.Save("example.model")
if err != nil {
return err
}
return nil
}
from agogo.
I see but I think it should be like this can you help give it a try:
layer1, l1Op := m.res(input, filterCount, fmt.Sprintf("Layer1 of Shared Layer %d", layer))
layer2, l2Op := m.res(layer1, filterCount, fmt.Sprintf("Layer2 of Shared Layer %d", layer))
from agogo.
Related Issues (9)
- when it's ready HOT 4
- How would the configuration for training agogo for go look like?
- How to run wq?
- 9x9 board
- Mancala / kalah
- Can't run tic-tac-toc HOT 1
- Train fail: shuffle batch failed - matX: Not yet implemented: native matrix for colmajor or unpacked matrices HOT 4
- Cannot train tic-tac-toe with more than 14 episodes HOT 4
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from agogo.