checkpoint @ game=4040 steps=4040

training_log.txt

@@ -4339,3 +4339,13 @@
 [2026-06-23 11:26:29] game= 4028 | 1/2-1/2 | moves= 61 | buf=50,000 | steps=4,028 | loss=9.1786 (p=8.4493 v=0.7293) | game=15.2s train=97.3s
 [2026-06-23 11:28:15] game= 4029 | 1/2-1/2 | moves= 60 | buf=50,000 | steps=4,029 | loss=9.1194 (p=8.4493 v=0.6701) | game=7.3s train=98.3s
 [2026-06-23 11:29:59] game= 4030 | 1/2-1/2 | moves= 51 | buf=50,000 | steps=4,030 | loss=9.0127 (p=8.4493 v=0.5635) | game=5.8s train=98.5s
+[2026-06-23 11:31:55] game= 4031 | 1/2-1/2 | moves= 63 | buf=50,000 | steps=4,031 | loss=8.8130 (p=8.4493 v=0.3637) | game=7.2s train=97.6s
+[2026-06-23 11:33:41] game= 4032 | 1/2-1/2 | moves= 69 | buf=50,000 | steps=4,032 | loss=8.8428 (p=8.4493 v=0.3935) | game=7.1s train=98.5s
+[2026-06-23 11:35:26] game= 4033 | 1/2-1/2 | moves= 59 | buf=50,000 | steps=4,033 | loss=8.7533 (p=8.4493 v=0.3040) | game=6.5s train=98.9s
+[2026-06-23 11:37:13] game= 4034 | 1/2-1/2 | moves= 63 | buf=50,000 | steps=4,034 | loss=8.7450 (p=8.4493 v=0.2957) | game=7.0s train=99.5s
+[2026-06-23 11:38:56] game= 4035 | 1/2-1/2 | moves= 55 | buf=50,000 | steps=4,035 | loss=8.7450 (p=8.4493 v=0.2957) | game=5.6s train=98.2s
+[2026-06-23 11:40:41] game= 4036 | 1/2-1/2 | moves= 71 | buf=50,000 | steps=4,036 | loss=8.6815 (p=8.4493 v=0.2322) | game=8.0s train=96.9s
+[2026-06-23 11:42:33] game= 4037 | 1/2-1/2 | moves= 69 | buf=50,000 | steps=4,037 | loss=8.7125 (p=8.4493 v=0.2632) | game=12.9s train=98.5s
+[2026-06-23 11:44:18] game= 4038 | 1/2-1/2 | moves= 62 | buf=50,000 | steps=4,038 | loss=8.6376 (p=8.4493 v=0.1883) | game=7.2s train=98.2s
+[2026-06-23 11:46:05] game= 4039 | 1/2-1/2 | moves= 47 | buf=50,000 | steps=4,039 | loss=8.6608 (p=8.4493 v=0.2115) | game=5.4s train=101.4s
+[2026-06-23 11:47:54] game= 4040 | 1/2-1/2 | moves= 63 | buf=50,000 | steps=4,040 | loss=8.6563 (p=8.4493 v=0.2070) | game=7.7s train=101.8s