You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository was archived by the owner on Mar 19, 2026. It is now read-only.
The referenced repositories have been copied into this repository, so
local links should be used where possible.
This commit is largely based on the similar commit I made for the
previous release:
20fcb46
Copy file name to clipboardExpand all lines: diplomacy_cicero/chiron-utils/README.md
+5-5Lines changed: 5 additions & 5 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -65,9 +65,9 @@ Most bots require specific instructions to build and run them properly:
65
65
-`LrProbsTextVisualAdvisor`: Provides textual and visual advice about all powers
66
66
-`LrProbsVisualAdvisor`: Provides visual advice about all powers
67
67
- To build the bot, run `make build-baseline-lr` to generate the OCI image to run with Docker
68
-
- When running the bot outside of a container, download the latest model file from [`large-file-storage/lr_models/`](https://github.com/ALLAN-DIP/large-file-storage/tree/main/lr_models). The filename includes the model release date in `YYYYMMDD` format).
68
+
- When running the bot outside of a container, use the files from [`lr_model/`](lr_model/)
69
69
- Edit the `MODEL_PATH` constant in `lr_bot.py` to point to the unzipped model folder.
70
-
- Code for model training can be found at <https://github.com/ALLAN-DIP/baseline-models>
70
+
- Code for model training can be found in [`baseline-models/`](baseline-models/)
71
71
- LLM advisor bots:
72
72
- A family of bots containing the following types:
73
73
-[`FaafAdvisor`](src/chiron_utils/bots/csu_faaf_advisor_bot.py): A large language model using the FAAF model from the CSU team to provide commentary advice given board states, recommended orders for current player and predicted orders of opponents from Cicero.
@@ -79,8 +79,8 @@ Most bots require specific instructions to build and run them properly:
79
79
- Request access on the page for the following model:
- Once your request has been approved, authenticate on your local machine using a user access token, using the official [User access tokens](https://huggingface.co/docs/hub/security-tokens) documentation as a guide.
82
-
- When using `LlmAdvisor`, one needs to run another advisor to provide `OPPONENT_MOVE` advice to the same power. For example, one can run the [Cicero advisor](https://github.com/ALLAN-DIP/diplomacy_cicero) with the argument `--advice_levels OPPONENT_MOVE`.
83
-
- When using `FaafAdvisor` or `LlmNewAdvisor`, one needs to run another advisor to provide `MOVE|OPPONENT_MOVE` advice to the same power. For example, one can run the [Cicero advisor](https://github.com/ALLAN-DIP/diplomacy_cicero) with the argument `--advice_levels 'MOVE|OPPONENT_MOVE'`.
82
+
- When using `LlmAdvisor`, one needs to run another advisor to provide `OPPONENT_MOVE` advice to the same power. For example, one can run the [Cicero advisor](../) with the argument `--advice_levels OPPONENT_MOVE`.
83
+
- When using `FaafAdvisor` or `LlmNewAdvisor`, one needs to run another advisor to provide `MOVE|OPPONENT_MOVE` advice to the same power. For example, one can run the [Cicero advisor](../) with the argument `--advice_levels 'MOVE|OPPONENT_MOVE'`.
84
84
- To use the bot, run the following command from the repository root, replacing `[bot_type]` with the bot's name:
85
85
```shell
86
86
# Set communication stage to 10 minutes (in seconds) to give enough time
@@ -90,7 +90,7 @@ Most bots require specific instructions to build and run them properly:
- This bot does not return orders, and is only intended to be a message advisor.
92
92
- Messages are retrieved from an Elasticsearch database using similarity search based on game state.
93
-
- Running the bot requires a populated Elasticsearch instance. See [`baseline-models`](https://github.com/ALLAN-DIP/baseline-models/blob/main/README.md#message_advisor_coderestore_snapshotpy) on how to run a Dockerized Elasticsearch instance locally.
93
+
- Running the bot requires a populated Elasticsearch instance. See [`baseline-models/README.md`](baseline-models/README.md#message_advisor_coderestore_snapshotpy) on how to run a Dockerized Elasticsearch instance locally.
0 commit comments