Training in progress, step 15000, checkpoint

last-checkpoint/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:276f7be644e8e9d1ebf31073f476c32b55725f266b87e4b6f03699a45d039e89
+oid sha256:2a6a1a0dba7021d8b7d428db87c437efdf6b626407168352af37d71ff4f8aa75
 size 36730224

last-checkpoint/ar_diffusion_info.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0ad057bc708b31f63ffd87016392c8a695caacca21d14f10104b01f30e082a84
+oid sha256:4a3251e96aad20ef9617385c554ac0de9f993e0b60ea66ffb1fe6ed5147f2ee6
 size 1736

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fc2676454c62bae3720ff4a5eac43f9d32d9a15a021931bb9f25a0bbad4ac7eb
+oid sha256:f1894fce258370764aa2bd4aea50e0e93ed78c2cdb28cd0a97679eba3acb55f8
 size 73588346

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d71e62f8024cc49b7ca4471f82288990c8a55fe933782923ce1c849f1e26e594
+oid sha256:f469366f58bbcc837e1eda1cc990a381676814584efda8029c314c0d4dbfdb15
 size 14244

last-checkpoint/scaler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:50e8f7c8dcbd001a527d3a189495db1e67a42d6ff36ec66455c01f20c42f2541
+oid sha256:b21eaeb6df3448a126aa3c865471f11d93d5da05e32d2b0b19d5120cef1d8306
 size 988

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:297e6f66093de0af6eb4d635b7629b97e7ee3aaf59595a9b78c3fa38ce59f7cc
+oid sha256:bb9bb6b4b564db2d66f145a412e3b2b0c279a174d093912c413b1b960f8efe8d
 size 1064

last-checkpoint/trainer_state.json

@@ -1,10 +1,10 @@
 {
-  "best_global_step": 14500,
-  "best_metric": 1.1057496070861816,
+  "best_global_step": 14750,
+  "best_metric": 1.1030505895614624,
   "best_model_checkpoint": "./ar-diffusion-checkpoints-fixed/checkpoint-14500",
-  "epoch": 1.1152988231674485,
+  "epoch": 1.153757403276671,
   "eval_steps": 250,
-  "global_step": 14500,
+  "global_step": 15000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -2502,6 +2502,92 @@
       "eval_samples_per_second": 56.187,
       "eval_steps_per_second": 14.047,
       "step": 14500
+    },
+    {
+      "epoch": 1.119144681178371,
+      "grad_norm": 1.2388238906860352,
+      "learning_rate": 0.00012706022907305926,
+      "loss": 1.1321,
+      "step": 14550
+    },
+    {
+      "epoch": 1.1229905391892931,
+      "grad_norm": 0.5141006708145142,
+      "learning_rate": 0.00012680050905124277,
+      "loss": 1.0453,
+      "step": 14600
+    },
+    {
+      "epoch": 1.1268363972002153,
+      "grad_norm": 1.1240845918655396,
+      "learning_rate": 0.0001265407890294263,
+      "loss": 1.0742,
+      "step": 14650
+    },
+    {
+      "epoch": 1.1306822552111375,
+      "grad_norm": 1.433976650238037,
+      "learning_rate": 0.0001262810690076098,
+      "loss": 1.1468,
+      "step": 14700
+    },
+    {
+      "epoch": 1.13452811322206,
+      "grad_norm": 1.077966332435608,
+      "learning_rate": 0.0001260213489857933,
+      "loss": 1.1143,
+      "step": 14750
+    },
+    {
+      "epoch": 1.13452811322206,
+      "eval_loss": 1.1030505895614624,
+      "eval_runtime": 17.8195,
+      "eval_samples_per_second": 56.118,
+      "eval_steps_per_second": 14.03,
+      "step": 14750
+    },
+    {
+      "epoch": 1.1383739712329821,
+      "grad_norm": 1.1456421613693237,
+      "learning_rate": 0.00012576162896397683,
+      "loss": 1.1173,
+      "step": 14800
+    },
+    {
+      "epoch": 1.1422198292439043,
+      "grad_norm": 0.9305130243301392,
+      "learning_rate": 0.00012550190894216037,
+      "loss": 1.1158,
+      "step": 14850
+    },
+    {
+      "epoch": 1.1460656872548265,
+      "grad_norm": 1.3796011209487915,
+      "learning_rate": 0.00012524218892034388,
+      "loss": 1.0785,
+      "step": 14900
+    },
+    {
+      "epoch": 1.149911545265749,
+      "grad_norm": 0.9901970028877258,
+      "learning_rate": 0.0001249824688985274,
+      "loss": 1.1383,
+      "step": 14950
+    },
+    {
+      "epoch": 1.153757403276671,
+      "grad_norm": 0.8250207304954529,
+      "learning_rate": 0.00012472794327714724,
+      "loss": 1.1363,
+      "step": 15000
+    },
+    {
+      "epoch": 1.153757403276671,
+      "eval_loss": 1.103262186050415,
+      "eval_runtime": 17.8852,
+      "eval_samples_per_second": 55.912,
+      "eval_steps_per_second": 13.978,
+      "step": 15000
     }
   ],
   "logging_steps": 50,