Comments (2)
But the default difficulty deceased monotonically in your case. I will test your collection tomorrow.
from fsrs4anki.
Hm actually, you might be right. A large majority of my cards have ~250% ease
And looking at the interval history if I keep pressing Good:
first rating: 3
rating history: 3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3
interval history: 0,4,10,25,56,119,240,460,845,1493,2551,4227,6812,10707,16451,24757
difficulty history: 0,1.2,1.2,1.2,1.2,1.2,1.2,1.2,1.2,1.2,1.2,1.2,1.2,1.2,1.2,1.2
The change between each interval is:
0 -> 4 = N/A
4 -> 10 = 10 / 4 = 2.5 = 250%
10 -> 25 = 25 / 10 = 2.5 = 250%
25 -> 56 = 56/ 25 = 2.24 = 224%
56 -> 119 = 119 / 56 = 2.125 = 212.5%
119 -> 240 = 240 / 119 = 2.017 = 201.7%
...
This seems correct since most of my cards have 250% and I have about ~88% retention rate for my young cards. The FSRS4anki algorithm starts with 250% change for the interval and then it slowly decreases the change in interval as I keep pressing good.
In version FSRS4Anki v1.4.3 Optimizer however, I got the results:
let defaultDifficulty = 3.835;
let defaultStability = 3.2191;
let difficultyDecay = -0.9131;
let stabilityDecay = -0.0872;
let retrievabilityFactor = 1.2571;
let increaseFactor = 3.2025;
let lapsesBase = -0.0493;
stability, difficulty, lapses
3.22 3.83 0
5.77 3.84 0
10.62 3.84 0
19.06 3.83 0
32.91 3.83 0
55.86 3.83 0
93.04 3.83 0
152.10 3.83 0
244.58 3.83 0
387.58 3.83 0
605.15 3.83 0
931.47 3.83 0
1415.08 3.83 0
2123.76 3.83 0
3150.51 3.83 0
0,3,6,11,19,33,56,93,152,245,388,605,931,1415,2124,3151
The change between each interval here is:
0 -> 3 = N/A
3 -> 6 = 6 / 3 = 2 = 200%
6 -> 11 = 11 / 6 = 1.83 = 183%
11 -> 19 = 19 / 11 = 1.73 = 173%
19 -> 33 = 33 / 19 = 1.74 = 174%
33 -> 56 = 56 / 33 = 1.70 = 170%
...
With these parameters, it starts with a change in interval of 200%, not 250%. It does seem like these new parameters are correct. I guess the larger learning rate allowed it to converge faster.
from fsrs4anki.
Related Issues (20)
- [Feature Request] Produce workload graph for an individual HOT 31
- [Feature Request] new card limit in optimizer.find_optimal_retention HOT 10
- [Feature Request] Make it easier to disable SM2 ("anki") simulation in Simulator HOT 1
- [BUG] Give Simulator's graphs titles to distinguish them HOT 4
- [Question] What is the best practice for reschedule cards and update parameters button? HOT 6
- [Feature Request] Improve Learning/Relearning Scheduling HOT 10
- [BUG] Hard interval is longer than Good interval, and longer than Max interval HOT 1
- [BUG] optimizer.find_optimal_retention: ValueError: Axis limits cannot be NaN or Inf HOT 4
- FSRS Simulator File Not a Zip File [BUG] HOT 7
- [Question]only use "Again" or "Hard" HOT 17
- [Feature Request] Change scheduling based on the type of card used HOT 8
- [BUG] List of compatible add-ons (update)
- [Question] Does review records affecte the parameters when I reset old cards to new cards HOT 2
- [Feature Request] Change to difficulty on new cards HOT 3
- [Question] Not graduating cards the same day affects scheduling? HOT 3
- [Question] Who should NOT use fsrs4anki / the limits to generalization are in the training set HOT 3
- [Question] I haven't used Anki for a while HOT 1
- [Question] Date change HOT 6
- [Question] Does fsrs deals well with cramming HOT 1
- [Question] HOT 21
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 fsrs4anki.