Update evo_transformer.py

evo_transformer.py

@@ -1,50 +1,37 @@
-# evo_transformer.py
 import random
+import copy
 
 class EvoTransformer:
-    def __init__(self, config=None):
-        self.default_config = {
+    def __init__(self):
+        self.history = []
+        self.base_config = {
             "layers": 4,
             "attention_heads": 4,
             "ffn_dim": 1024,
             "dropout": 0.1,
-            "memory": False,
+            "memory": False
         }
-        self.config = config or self.default_config.copy()
-        self.history = [self.config.copy()]
 
     def reset(self):
-        self.config = self.default_config.copy()
-        self.history = [self.config.copy()]
-
-    def mutate(self):
-        new_config = self.config.copy()
-        trait = random.choice(list(new_config.keys()))
-
-        if trait == "layers":
-            new_config[trait] = max(1, new_config[trait] + random.choice([-1, 1]))
-        elif trait == "attention_heads":
-            new_config[trait] = random.choice([2, 4, 6, 8])
-        elif trait == "ffn_dim":
-            new_config[trait] = random.choice([512, 1024, 2048])
-        elif trait == "dropout":
-            new_config[trait] = round(min(max(0.0, new_config[trait] + random.uniform(-0.05, 0.05)), 0.5), 2)
-        elif trait == "memory":
-            new_config[trait] = not new_config[trait]
-
-        self.config = new_config
-        self.history.append(new_config.copy())
-
-    def evolve(self, generations=3):
-        for _ in range(generations):
-            self.mutate()
-
-    def get_history(self):
-        return self.history
-
-    def evaluate(self):
-        score = round(random.uniform(0.85, 0.95), 4)
-        return {"accuracy": score, "params": self.estimate_params()}
-
-    def estimate_params(self):
-        return round(10 + self.config["layers"] * self.config["ffn_dim"] * 1.0, 2)
+        self.history = []
+
+    def mutate(self, config):
+        new_config = copy.deepcopy(config)
+        if random.random() < 0.5:
+            new_config["layers"] = min(12, max(1, new_config["layers"] + random.choice([-1, 1])))
+        if random.random() < 0.5:
+            new_config["attention_heads"] = min(12, max(1, new_config["attention_heads"] + random.choice([-1, 1])))
+        if random.random() < 0.5:
+            new_config["ffn_dim"] = min(4096, max(128, new_config["ffn_dim"] + random.choice([-512, 512])))
+        if random.random() < 0.5:
+            new_config["dropout"] = round(min(0.5, max(0.0, new_config["dropout"] + random.choice([-0.02, 0.02]))), 2)
+        if random.random() < 0.3:
+            new_config["memory"] = not new_config["memory"]
+        return new_config
+
+    def run_evolution(self, generations=5):
+        current = self.base_config
+        self.history.append(current)
+        for _ in range(generations - 1):
+            current = self.mutate(current)
+            self.history.append(current)