Merge branch 'low_mem' of github.com:sanderland/katrain into low_mem

README.md

@@ -58,6 +58,7 @@ but has since grown to include a wide range of features, including:
 ## Manual
 
 ### Play
+
 Under the 'play' tab you can select who is playing black and white.
 * Human is simple play with potential feedback, but without auto-undo.
 * Teach will give you instant feedback, and auto-undo bad moves to give you a second chance. 
@@ -80,6 +81,7 @@ In short, if you are a weaker player you should mostly on large dots that are re
 while stronger players can pay more attention to smaller mistakes.
 
 #### AIs
+
 Available AIs, with strength indicating an estimate for the default settings, are:
 
 * **[9p+]** **Default** is full KataGo, above professional level. 
@@ -115,7 +117,6 @@ Keyboard shortcuts are shown with **[key]**.
     * **[s]**: Equalize: Re-evaluate all currently shown next moves with the same visits as the current top move. Useful to increase confidence in the suggestions with high uncertainty.
     * **[d]**: Sweep: Evaluate all possible next moves. This can take a bit of time even though 'fast_visits' is used, but the result is nothing if not colourful.
 
-    
 ## Keyboard and mouse shortcuts
 
 In addition to shortcuts mentioned above, there are:
@@ -136,7 +137,6 @@ In addition to shortcuts mentioned above, there are:
 * **[Ctrl-n]**: Load SGF from clipboard
 * **[space]**: Pass
 
-
 ## Configuration
 
 Configuration is stored in `config.json`. Most settings are now available to edit in the program, but some advanced options are not.
@@ -173,9 +173,12 @@ If you ever need to reset to the original settings, simply re-download the `conf
   * The first startup of KataGo can be slow due to GPU tuning, after that it should be much faster.
 * The program is running too slowly. How can I speed it up?
   *  Adjust the number of visits or maximum time allowed in the settings.
+* KataGo crashes with out of memory errors, how can I prevent this?
+  *  Try using a lower number for `nnMaxBatchSize` in `KataGo/analysis_config.cfg`, and avoid using the board size 29 version.
  
 ## Contributing
 
 * Feedback and pull requests are both very welcome.
 * For suggestions and planned improvements, see the 'issues' tab on github.
 * You can also contact me on discord (Sander#3278) or [reddit](http://reddit.com/u/sanderbaduk) to give feedback, or simply show your appreciation.
+

bots/start_bot.py

@@ -15,7 +15,7 @@ if True or bot in ["dev", "local"]:
 else:
     GTP2OGS = "node ../stable-gtp2ogs"
 BOT_SETTINGS = f" --maxconnectedgames {MAXGAMES} --maxhandicapunranked 25 --maxhandicapranked 1 --boardsizesranked 19 --boardsizesunranked all --komisranked automatic,5.5,6.5,7.5 --komisunranked all"
-if 'beta' in bot:
+if "beta" in bot:
     BOT_SETTINGS += " --beta"
 else:
     BOT_SETTINGS +=  "" # --rankedonly"

config.json

@@ -1,9 +1,9 @@
 {
     "engine": {
-        "katago": "KataGo/katago-bs",
+        "katago": "KataGo/katago",
         "model": "KataGo/models/b15-1.3.2.txt.gz",
         "config": "KataGo/analysis_config.cfg",
-        "threads": 8,
+        "threads": 16,
         "max_visits": 500,
         "fast_visits": 50,
         "max_time": 3.0,

core/sgf_parser.py

@@ -212,7 +212,7 @@ class SGFNode:
 
     @property
     def next_player(self):
-        if "B" in self.properties or "AB" in self.properties: # root or black moved
+        if "B" in self.properties or "AB" in self.properties:  # root or black moved
             return "W"
         else:
             return "B"
@@ -222,7 +222,7 @@ class SGFNode:
         if "B" in self.properties or "AB" in self.properties:
             return "B"
         else:
-            return "W" # nb root is considered white played if no handicap stones are placed
+            return "W"  # nb root is considered white played if no handicap stones are placed
 
 
 class SGF:
@@ -239,10 +239,10 @@ class SGF:
             if not encoding:
                 match = re.search(rb"CA\[(.*?)\]", bin_contents)
                 if match:
-                    encoding = match[1].decode("ascii")
+                    encoding = match[1].decode("ascii", errors="ignore")
                 else:
                     encoding = "ISO-8859-1"  # default
-            decoded = bin_contents.decode(encoding=encoding)
+            decoded = bin_contents.decode(encoding=encoding, errors="ignore")
             return cls.parse(decoded)
 
     def __init__(self, contents):

gui/popups.py

@@ -179,6 +179,7 @@ class ConfigPopup(QuickConfigGui):
                 if not old_proc:
                     self.katrain.game.analyze_all_nodes()  # old engine was broken, so make sure we redo any failures
                 self.katrain.update_state()
+
             Clock.schedule_once(restart_engine, 0)
 
         self.katrain.debug_level = self.config["debug"]["level"]

profile.py

@@ -1,42 +0,0 @@
-# This is a stand-alone script that generates a review for an SGF
-import json
-import sys
-import time
-
-from core.common import OUTPUT_INFO
-from core.engine import KataGoEngine
-from core.game import Game, KaTrainSGF
-
-if len(sys.argv) < 2:
-    exit(1)
-inputfile = sys.argv[1]
-
-with open("config.json") as f:
-    settings = json.load(f)
-    sgf_settings = settings["sgf"]
-    engine_settings = settings["engine"]
-    game_settings = settings["game"]
-    trainer_settings = settings["trainer"]
-
-#engine_settings['threads'] = 32
-engine_settings["max_time"] = 1000
-
-class Logger:
-    def log(self, msg, level):
-        if level <= OUTPUT_INFO:
-            print(msg)
-
-
-logger = Logger()
-
-engine = KataGoEngine(logger, engine_settings)
-move_tree = KaTrainSGF.parse_file(inputfile)
-game = Game(logger, engine, game_settings, move_tree=None, analyze_fast=True)
-game.root=move_tree
-reverse_nodes = game.root.nodes_in_tree[::-1]
-
-for n in reverse_nodes[::10]:
-    n.analyze(engine=engine)
-    while not n.analysis_ready:
-        time.sleep(0.01)
-    print(n.single_move,'done')

review.py

@@ -1,49 +0,0 @@
-# This is a stand-alone script that generates a review for an SGF
-import json
-import sys
-import time
-
-from core.common import OUTPUT_INFO
-from core.engine import KataGoEngine
-from core.game import Game, KaTrainSGF
-
-if len(sys.argv) < 2:
-    exit(1)
-inputfile = sys.argv[1]
-
-with open("config.json") as f:
-    settings = json.load(f)
-    sgf_settings = settings["sgf"]
-    engine_settings = settings["engine"]
-    game_settings = settings["game"]
-    trainer_settings = settings["trainer"]
-
-engine_settings['threads'] = 16
-#engine_settings['katago'] = 'KataGo/katago-bs'
-engine_settings["max_time"] = 1000
-engine_settings["max_visits"] = 50
-
-class Logger:
-    def log(self, msg, level):
-        if level <= OUTPUT_INFO:
-            print(msg)
-
-
-logger = Logger()
-
-engine = KataGoEngine(logger, engine_settings)
-move_tree = KaTrainSGF.parse_file(inputfile)
-game = Game(logger, engine, game_settings, move_tree=move_tree, analyze_fast=False)
-nodes = game.root.nodes_in_tree
-remaining = len(nodes)
-while remaining > 0:
-    remaining = sum(not n.analysis_ready for n in nodes)
-    print(f"Waiting for {remaining} queries")
-    time.sleep(0.1)
-
-msg = game.write_sgf(
-    sgf_settings["sgf_save"],
-    trainer_config=trainer_settings,
-    save_feedback=sgf_settings["save_feedback"],
-    eval_thresholds=trainer_settings["eval_thresholds"],
-)