Training in progress, step 18000, checkpoint

last-checkpoint/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:78d2977f0197deae844fb8fa4ba1483d26d8ab4a842337c308c5805e2d4b9291
+oid sha256:42c2f6139756d4179a8f7b2fd49bb0a2cd540ba8b774f9e69c5f50f17435f6b6
 size 36730224

last-checkpoint/ar_diffusion_info.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e5b3283c3aaea4f1a12c8a11e653a36fa924634ce08452daace823d8327ea96f
+oid sha256:d5bffd3fc7ee80c7d1ba70865a03ab1d358a04bee58d41c0ac14d5c087eae56f
 size 1544

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7aaf80cd43f46651a9c4aa03f28756212e23d71139e6fe0dca4ad9952d5a2268
+oid sha256:7b5e18457edaea63d6ccb66baccbacd52b196222165882275df4a7c9e0e28b1b
 size 73588346

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5c6f63d896b1e09539b72ac15eabafd681bf824376794bc90d0981fb00940917
+oid sha256:aec74914111c893531d77390ed26a8e62f47c5fd368563e3e2ea395d9971bdd0
 size 14244

last-checkpoint/scaler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:83c670602c4fdb95d34df084cca714b1e4115a72825bd5c70e64c19d3a998917
+oid sha256:d21f0ba0bbe7a28de5199a35a5f642d45bf19d1eb0aa770d1035c6ab0a649bfa
 size 988

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:abfae05460636119aec0a7a8d55ec3326712e8ea80bb468cb104d4e03da4b28b
+oid sha256:47960831c3aee4c1e74537230bffbd7196098e4c54fe510c8e622c62e1271790
 size 1064

last-checkpoint/trainer_state.json

@@ -2,9 +2,9 @@
   "best_global_step": 16750,
   "best_metric": 1.4798808097839355,
   "best_model_checkpoint": "./ar-diffusion-checkpoints-fixed/checkpoint-15500",
-  "epoch": 1.346050303822783,
+  "epoch": 1.3845088839320052,
   "eval_steps": 250,
-  "global_step": 17500,
+  "global_step": 18000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -3018,6 +3018,92 @@
       "eval_samples_per_second": 56.191,
       "eval_steps_per_second": 14.048,
       "step": 17500
+    },
+    {
+      "epoch": 1.349896161833705,
+      "grad_norm": 1.740045428276062,
+      "learning_rate": 0.00011149261096537933,
+      "loss": 1.4486,
+      "step": 17550
+    },
+    {
+      "epoch": 1.3537420198446273,
+      "grad_norm": 1.299919605255127,
+      "learning_rate": 0.00011123289094356284,
+      "loss": 1.4868,
+      "step": 17600
+    },
+    {
+      "epoch": 1.3575878778555497,
+      "grad_norm": 1.6288009881973267,
+      "learning_rate": 0.00011097317092174637,
+      "loss": 1.4595,
+      "step": 17650
+    },
+    {
+      "epoch": 1.3614337358664719,
+      "grad_norm": 0.8747851252555847,
+      "learning_rate": 0.00011071345089992988,
+      "loss": 1.4241,
+      "step": 17700
+    },
+    {
+      "epoch": 1.365279593877394,
+      "grad_norm": 1.9510573148727417,
+      "learning_rate": 0.0001104537308781134,
+      "loss": 1.496,
+      "step": 17750
+    },
+    {
+      "epoch": 1.365279593877394,
+      "eval_loss": 1.4812238216400146,
+      "eval_runtime": 17.8955,
+      "eval_samples_per_second": 55.88,
+      "eval_steps_per_second": 13.97,
+      "step": 17750
+    },
+    {
+      "epoch": 1.3691254518883162,
+      "grad_norm": 1.4853876829147339,
+      "learning_rate": 0.00011019401085629692,
+      "loss": 1.4645,
+      "step": 17800
+    },
+    {
+      "epoch": 1.3729713098992384,
+      "grad_norm": 1.5125057697296143,
+      "learning_rate": 0.00010993429083448044,
+      "loss": 1.4052,
+      "step": 17850
+    },
+    {
+      "epoch": 1.3768171679101608,
+      "grad_norm": 0.7320863008499146,
+      "learning_rate": 0.00010967457081266395,
+      "loss": 1.5016,
+      "step": 17900
+    },
+    {
+      "epoch": 1.380663025921083,
+      "grad_norm": 1.9995285272598267,
+      "learning_rate": 0.00010941485079084747,
+      "loss": 1.4234,
+      "step": 17950
+    },
+    {
+      "epoch": 1.3845088839320052,
+      "grad_norm": 0.8304823637008667,
+      "learning_rate": 0.00010915513076903099,
+      "loss": 1.4068,
+      "step": 18000
+    },
+    {
+      "epoch": 1.3845088839320052,
+      "eval_loss": 1.4845945835113525,
+      "eval_runtime": 17.7402,
+      "eval_samples_per_second": 56.369,
+      "eval_steps_per_second": 14.092,
+      "step": 18000
     }
   ],
   "logging_steps": 50,