Skip to content

Comment on incresing stack size for bet sizes #7

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
mathemage wants to merge 1 commit into
base: master
Choose a base branch
from
Open

Comment on incresing stack size for bet sizes #7

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
14 changes: 12 additions & 2 deletions doc/manual/tutorial.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -105,12 +105,22 @@ The tree that you just produced only contains `fold`, `call`, `pot bet` and
`Source/Settings/arguments.lua`. The file contains multiple arguments you can
change, including `params.bet_sizing = {1}` which specifies the allowed bet
sizes. Try changing it into `params.bet_sizing = {1, 2}` to allow not only
pot-sized bets, but also bets of twice the pot size. To see the results, run
pot-sized bets, but also bets of twice the pot size. To see the results, run
```bash
th Tree/Tests/test_tree_visualiser.lua
```
again.

Do not forget to also make stack sizes high enough so that actions `pot bet`, `2-pot bet`,
`3-pot bet` etc. can take any effect. For example, in the previous case of
`params.bet_sizing = {1, 2}`, the pot size after the first round is `300 + 300 = 600`,
while the stack sizes are `1200`. Hence, a `2-pot bet` would be greater greater than
player's stack size of `1200`. We thus need to increase players' stack sizes, e.g.,
```lua
params.stack = 12000
```
in `Source/Settings/arguments.lua`. See [Issue #6](https://github.com/lifrordi/DeepStack-Leduc/issues/6) for more details.

These bet sizes will limit the options the agent considers when making an
action during the game. Note that this does **not** limit the agent from
understanding any other bet size the opponent can make.
Expand Down Expand Up @@ -394,4 +404,4 @@ If you trained your model on the GPU, but want to run DeepStack on a machine
that has no GPU, you need to convert the model. There is a simple script
`Source/Nn/cpu_gpu_model_converter.lua` that does exactly that. It also
contains a function for converting from a CPU model to a GPU model, if you
need to do the conversion the other way.
need to do the conversion the other way.