Training in progress, step 39000, checkpoint

last-checkpoint/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4f4e59779ae5268c858ada9fe05772a37f9613020d8b9f7340783b2a16c0e383
+oid sha256:cf3bda4176df59060b4c3b2e5426a0849f3db73d160bd88e36205096ff67d81c
 size 36730224

last-checkpoint/ar_diffusion_info.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:376da5b0e7144c1b91117f5526417311ed45b536f61dd6ac0ef8b19051f1dc33
+oid sha256:53bc0f8817dd10f924c15f7cdd42fe5d2191e0ae00ed029ca8895e92da7bac07
 size 1736

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9de512a7acdea4434e0675f44f1896eab7a4c3fe6cc967e9cd2360783b35758f
+oid sha256:558eed0dfcd806ee844603a215b913f148e482ed28be46203246d8562d48e6e7
 size 73588346

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c8c00195221e523ee69e83979891848166b78d5b2c3da6d54545d0f1c8a050c4
+oid sha256:80343da31c600f0e1bbaa709a5f7f529e2d6a5cd1c7788323715a3bfd97d3201
 size 14244

last-checkpoint/scaler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:274528fb9458ad54ab5973b2752fc7f6c9483a2a721e4d0ba19183ca440bc21c
+oid sha256:c156120bdd3702e22db9f52ea4b7b1f109fd9c58a54b6df0124843a2ca789b7e
 size 988

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9bcb038f6702a5ed9cf6a7787eb82f2807063a89d7bbe28d73008f47570ae171
+oid sha256:c92e907578f7dfc926449c7e4863a766eefc5d8dcfbba692ef77fb44a937de6c
 size 1064

last-checkpoint/trainer_state.json

@@ -2,9 +2,9 @@
   "best_global_step": 31000,
   "best_metric": 0.7226839661598206,
   "best_model_checkpoint": "./ar-diffusion-checkpoints-fixed/checkpoint-31000",
-  "epoch": 2.961310668410122,
+  "epoch": 2.999769248519345,
   "eval_steps": 250,
-  "global_step": 38500,
+  "global_step": 39000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -6630,6 +6630,92 @@
       "eval_samples_per_second": 55.821,
       "eval_steps_per_second": 13.955,
       "step": 38500
+    },
+    {
+      "epoch": 2.9651565264210444,
+      "grad_norm": 0.7657055854797363,
+      "learning_rate": 2.4257850037659406e-06,
+      "loss": 0.7114,
+      "step": 38550
+    },
+    {
+      "epoch": 2.969002384431967,
+      "grad_norm": 0.7305043339729309,
+      "learning_rate": 2.1660649819494585e-06,
+      "loss": 0.7135,
+      "step": 38600
+    },
+    {
+      "epoch": 2.972848242442889,
+      "grad_norm": 0.7981142401695251,
+      "learning_rate": 1.9063449601329769e-06,
+      "loss": 0.7328,
+      "step": 38650
+    },
+    {
+      "epoch": 2.976694100453811,
+      "grad_norm": 0.7305875420570374,
+      "learning_rate": 1.6466249383164948e-06,
+      "loss": 0.7103,
+      "step": 38700
+    },
+    {
+      "epoch": 2.9805399584647336,
+      "grad_norm": 1.197097659111023,
+      "learning_rate": 1.386904916500013e-06,
+      "loss": 0.7148,
+      "step": 38750
+    },
+    {
+      "epoch": 2.9805399584647336,
+      "eval_loss": 0.7678167819976807,
+      "eval_runtime": 17.8827,
+      "eval_samples_per_second": 55.92,
+      "eval_steps_per_second": 13.98,
+      "step": 38750
+    },
+    {
+      "epoch": 2.9843858164756556,
+      "grad_norm": 0.8533993363380432,
+      "learning_rate": 1.127184894683531e-06,
+      "loss": 0.753,
+      "step": 38800
+    },
+    {
+      "epoch": 2.988231674486578,
+      "grad_norm": 0.7372131943702698,
+      "learning_rate": 8.674648728670494e-07,
+      "loss": 0.7082,
+      "step": 38850
+    },
+    {
+      "epoch": 2.9920775324975004,
+      "grad_norm": 1.499084234237671,
+      "learning_rate": 6.077448510505675e-07,
+      "loss": 0.6937,
+      "step": 38900
+    },
+    {
+      "epoch": 2.9959233905084224,
+      "grad_norm": 0.5895427465438843,
+      "learning_rate": 3.4802482923408566e-07,
+      "loss": 0.701,
+      "step": 38950
+    },
+    {
+      "epoch": 2.999769248519345,
+      "grad_norm": 0.9201724529266357,
+      "learning_rate": 8.830480741760382e-08,
+      "loss": 0.7665,
+      "step": 39000
+    },
+    {
+      "epoch": 2.999769248519345,
+      "eval_loss": 0.767805814743042,
+      "eval_runtime": 17.9509,
+      "eval_samples_per_second": 55.708,
+      "eval_steps_per_second": 13.927,
+      "step": 39000
     }
   ],
   "logging_steps": 50,