Comments (7)
Could you please set "100" to "loop"? It specifies how many times the trainer read the training data.
from stockfish.
Do you mean i need a loop greater than 1 to use 100% of the training data !?
Here it takes 13 hours to reach 950m with loop 1. I'm afraid If i set loop 100 (100 * 13 hrs = about 54 days !?), it may takes a very looong times to solve this issue.
from stockfish.
Do you mean i need a loop greater than 1 to use 100% of the training data !?
Yes.
Here it takes 13 hours to reach 950m with loop 1. I'm afraid If i set loop 100 (100 * 13 hrs = about 54 days !?), it may takes a very looong times to solve this issue.
A training will automatically finish, once the loss is converged even though the loop is set to 100.
from stockfish.
OK thank you very much !
So i will retest with loop 2 because i think reusing several times the same data was not optimal at learning phase...
from stockfish.
@deedstacktyk I believe he is recommending using
' loop 100 ' (which is default I believe)
from stockfish.
why "loop 1" don't use 100% of the training data ? why should i start for multiple loops in order to train only one all the data ?
from stockfish.
I think that it is an implementation problem. There are no plans to fix it because it is not a critical issue. Pull requests to fix are welcome.
I'm sorry to say this one but here is not a right place to ask questions. This is a place to report issues. I recommend to ask further questions in the #help channel in the unofficial Stockfish Discord server.
I will close this issue.
from stockfish.
Related Issues (20)
- hash dedup removes book position
- reduction depends on thread count
- compile errors on mac
- document .bin and .plain
- optimizer state is not being saved
- 2 fold adjudicated as a draw HOT 9
- uninit value with TB HOT 9
- Warmup step
- Clarify some learn option names and behaviour.
- move accuracy
- docs smart_fen_skipping
- binpack stats HOT 1
- 12:35 AM] J.H. Ellis: Could discard draws from data so that they make up a maximum of x%, then tune x.
- Segmentation fault on gensfen with ensure_quiet and random_multi_pv HOT 7
- Would be nice if issue is labelled bug if it is a bug HOT 4
- Invalid EP results in bad binpack conversion HOT 6
- Shorter write_maxply and inaccurate result HOT 2
- master and tools branches... HOT 4
- What is the master neural network referred to in SF source code near SF12 (first NNue) HOT 1
- Failure to load nnue trained by https://github.com/glinscott/nnue-pytorch HOT 1
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 stockfish.