Training in progress, step 20500, checkpoint

last-checkpoint/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:37a2fe2b7cf5d7145d364d49968edb54efe43efd98eba6806289d74f7e529ae8
+oid sha256:f82c93836bc692e34902f10bc204524481cd7a576e6d3e28b3bb7e2d682f323f
 size 36730224

last-checkpoint/ar_diffusion_info.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f7d0dd97707c0f0f3c8a12fdba68ea5ed4110d1f028231a6ea1054c90fb8603f
+oid sha256:6e8703e2ac4edb79cf8f1764c40656f93e9be69ded66eb4cce14bbce6431a5e0
 size 1544

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:59e604c620209e5ccc9d2a6df440d8eb819dc89b0c5554c52e42549c656e3250
+oid sha256:411bf70b4f8c28dcde0ada0a6ab13ec881cbbf19e85c2796efaba99e8b13583e
 size 73588346

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b205d8106465470d82a9f668e113b1a4c937f3fe768b385e78f85eb171e49cc2
+oid sha256:5dbffb5b9d940a96eb51dbe7a35860718ecfeca99437ad8b1b71dbafaefdd573
 size 14244

last-checkpoint/scaler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:eee3ef0a596a0ef37ef802555af22f2c3a4f5d3b4c13d00ed6df2466a73dd8ed
+oid sha256:384ba37c8844299a29c3c998811c9d47320ed4f7666069a4fb9a63d0dddfe44f
 size 988

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:83d15c547e182574b42460e3490faaf5cb37f299e2d2acbd532f1ef35e0b3aee
+oid sha256:c62f9fce1be162dfb9c17d3cf04dc0505b42c85ac9a955fe99836e3dd2e88351
 size 1064

last-checkpoint/trainer_state.json

@@ -1,10 +1,10 @@
 {
-  "best_global_step": 18750,
-  "best_metric": 1.4757392406463623,
+  "best_global_step": 20250,
+  "best_metric": 1.4724150896072388,
   "best_model_checkpoint": "./ar-diffusion-checkpoints-fixed/checkpoint-15500",
-  "epoch": 1.5383432043688947,
+  "epoch": 1.576801784478117,
   "eval_steps": 250,
-  "global_step": 20000,
+  "global_step": 20500,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -3448,6 +3448,92 @@
       "eval_samples_per_second": 55.752,
       "eval_steps_per_second": 13.938,
       "step": 20000
+    },
+    {
+      "epoch": 1.542189062379817,
+      "grad_norm": 1.7058050632476807,
+      "learning_rate": 9.850660987455523e-05,
+      "loss": 1.4646,
+      "step": 20050
+    },
+    {
+      "epoch": 1.5460349203907393,
+      "grad_norm": 2.2624917030334473,
+      "learning_rate": 9.824688985273876e-05,
+      "loss": 1.4586,
+      "step": 20100
+    },
+    {
+      "epoch": 1.5498807784016613,
+      "grad_norm": 2.216883420944214,
+      "learning_rate": 9.798716983092227e-05,
+      "loss": 1.4376,
+      "step": 20150
+    },
+    {
+      "epoch": 1.5537266364125837,
+      "grad_norm": 1.9749584197998047,
+      "learning_rate": 9.772744980910578e-05,
+      "loss": 1.47,
+      "step": 20200
+    },
+    {
+      "epoch": 1.5575724944235059,
+      "grad_norm": 2.185480833053589,
+      "learning_rate": 9.746772978728931e-05,
+      "loss": 1.39,
+      "step": 20250
+    },
+    {
+      "epoch": 1.5575724944235059,
+      "eval_loss": 1.4724150896072388,
+      "eval_runtime": 17.8191,
+      "eval_samples_per_second": 56.12,
+      "eval_steps_per_second": 14.03,
+      "step": 20250
+    },
+    {
+      "epoch": 1.561418352434428,
+      "grad_norm": 1.5810290575027466,
+      "learning_rate": 9.720800976547284e-05,
+      "loss": 1.4267,
+      "step": 20300
+    },
+    {
+      "epoch": 1.5652642104453505,
+      "grad_norm": 2.0329344272613525,
+      "learning_rate": 9.694828974365633e-05,
+      "loss": 1.5066,
+      "step": 20350
+    },
+    {
+      "epoch": 1.5691100684562724,
+      "grad_norm": 2.7335126399993896,
+      "learning_rate": 9.668856972183986e-05,
+      "loss": 1.4749,
+      "step": 20400
+    },
+    {
+      "epoch": 1.5729559264671948,
+      "grad_norm": 1.0576220750808716,
+      "learning_rate": 9.642884970002339e-05,
+      "loss": 1.4318,
+      "step": 20450
+    },
+    {
+      "epoch": 1.576801784478117,
+      "grad_norm": 0.6857870817184448,
+      "learning_rate": 9.616912967820689e-05,
+      "loss": 1.3567,
+      "step": 20500
+    },
+    {
+      "epoch": 1.576801784478117,
+      "eval_loss": 1.479669213294983,
+      "eval_runtime": 17.8378,
+      "eval_samples_per_second": 56.061,
+      "eval_steps_per_second": 14.015,
+      "step": 20500
     }
   ],
   "logging_steps": 50,