Stabilize client-server communication

AIAgent/common/constants.py

@@ -48,5 +48,5 @@ class ResultsHandlerLinks:
 
 # assuming we start from /VSharp/VSharp.ML.AIAgent
 SERVER_WORKING_DIR = (
- "GameServers/VSharp/VSharp.ML.GameServer.Runner/bin/Release/net7.0/"
+ "/Users/emax/Data/PySymGym/GameServers/VSharp/VSharp.ML.GameServer.Runner/bin/Debug/net7.0/"
 )

AIAgent/config.py

@@ -19,6 +19,7 @@ class GeneralConfig:
  IMPORT_MODEL_INIT = ...
  EXPORT_MODEL_INIT = ...
  DEVICE = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
+ DATASET_BASE_PATH = "/Users/emax/Data/PySymGym/maps/DotNet/Maps/ManuallyCollected/bin/Debug/net7.0"
 
 
 class BrokerConfig:
@@ -73,7 +74,7 @@ class FeatureConfig:
  enabled=True, save_path=Path("./report/epochs_tables/")
  )
  ON_GAME_SERVER_RESTART = OnGameServerRestartFeature(
- enabled=False, wait_for_reset_retries=10 * 60, wait_for_reset_time=0.1
+ enabled=True, wait_for_reset_retries=10 * 60, wait_for_reset_time=0.1
  )
 
 

AIAgent/learning/play_game.py

@@ -1,18 +1,18 @@
 import logging
+import os
 from statistics import StatisticsError
 from time import perf_counter
 from typing import TypeAlias
 
 import tqdm
-from func_timeout import FunctionTimedOut, func_set_timeout
-
 from common.classes import GameResult, Map2Result
 from common.constants import TQDM_FORMAT_DICT
 from common.game import GameMap
 from common.utils import get_states
 from config import FeatureConfig, GeneralConfig
 from connection.broker_conn.socket_manager import game_server_socket_manager
 from connection.game_server_conn.connector import Connector
+from func_timeout import FunctionTimedOut, func_set_timeout
 from learning.timer.resources_manager import manage_map_inference_times_array
 from learning.timer.stats import compute_statistics
 from learning.timer.utils import get_map_inference_times
@@ -146,6 +146,10 @@ def play_game(
  ) as pbar:
  list_of_map2result: list[Map2Result] = []
  for game_map in maps:
+ game_map.AssemblyFullName = os.path.join(
+ GeneralConfig.DATASET_BASE_PATH, game_map.AssemblyFullName
+ )
+
  logging.info(f"<{with_predictor.name()}> is playing {game_map.MapName}")
 
  try:

AIAgent/ml/models/RGCNEdgeTypeTAG3VerticesDoubleHistory2/model_modified.py

@@ -0,0 +1,36 @@
+from torch_geometric.nn import Linear
+from torch.nn.functional import softmax
+from .model import StateModelEncoder
+
+
+class StateModelEncoderLastLayer(StateModelEncoder):
+ def __init__(self, hidden_channels, out_channels):
+ super().__init__(hidden_channels, out_channels)
+ self.lin_last = Linear(out_channels, 1)
+
+ def forward(
+ self,
+ game_x,
+ state_x,
+ edge_index_v_v,
+ edge_type_v_v,
+ edge_index_history_v_s,
+ edge_attr_history_v_s,
+ edge_index_in_v_s,
+ edge_index_s_s,
+ ):
+ return softmax(
+ self.lin_last(
+ super().forward(
+ game_x=game_x,
+ state_x=state_x,
+ edge_index_v_v=edge_index_v_v,
+ edge_type_v_v=edge_type_v_v,
+ edge_index_history_v_s=edge_index_history_v_s,
+ edge_attr_history_v_s=edge_attr_history_v_s,
+ edge_index_in_v_s=edge_index_in_v_s,
+ edge_index_s_s=edge_index_s_s,
+ )
+ ),
+ dim=0,
+ )

AIAgent/run_common_model_training.py

@@ -5,21 +5,19 @@
 import os
 import random
 import typing as t
-from dataclasses import asdict, dataclass
+from dataclasses import dataclass
 from datetime import datetime
 from functools import partial
 from pathlib import Path
 
 import joblib
 import numpy as np
 import optuna
-import pandas as pd
 import torch
 import torch.nn as nn
 import tqdm
 from common.game import GameMap
 from config import GeneralConfig
-from connection.broker_conn.socket_manager import game_server_socket_manager
 from epochs_statistics.tables import create_pivot_table, table_to_string
 from learning.play_game import play_game
 from ml.common_model.dataset import FullDataset
@@ -33,7 +31,7 @@
 )
 from ml.common_model.utils import csv2best_models, get_model
 from ml.common_model.wrapper import BestModelsWrapper, CommonModelWrapper
-from ml.models.RGCNEdgeTypeTAG2VerticesDouble.model_modified import (
+from ml.models.RGCNEdgeTypeTAG3VerticesDoubleHistory2.model_modified import (
  StateModelEncoderLastLayer,
 )
 from ml.models.StateGNNEncoderConvEdgeAttr.model_modified import (