Training in progress, step 19000, checkpoint

last-checkpoint/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2468b8f20b2b8549a640818b37e1b9ef1f114a27b7f6c99b4a86e82f642e4401
+oid sha256:68e4017d8a95f1b3f117cff44c2d649b305cf089afdefcf8fc150daf09311398
 size 36730224

last-checkpoint/ar_diffusion_info.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c29781dc06aac24acf8000f8feb9e000d60a160a71b0c570ed32b320a8756db2
+oid sha256:bd7fa9c8114b1914d7c2ff974dfcbf00be8238b2a035cfde8dba2f243c66132b
 size 1736

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e40903d472c0cd2d111f5868dfbf44c27a4ee11fa862e96214b77a89c58a807b
+oid sha256:976f2ad66948cadc63923a3a9d366716d6c39d193180a993074754d9c2d5088f
 size 73588346

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e90dd042b4fe16d601e9434fac855ce20c91c9b7d9c2a6dff1f998d78c9a11bd
-size 14244
+oid sha256:62f0a63af717862f5ba78e609a46cd20ededb84bdab60287f44bdfc2444cb5a9
+size 14308

last-checkpoint/scaler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:17dcf088a8e3107cea49076c2140910eebc6fe4528f365b01ee8c2ed883b6684
+oid sha256:b99bea7b1fa648da24e0513c2b62a349744452626bc71a88bb8bbac50aa0f8ce
 size 988

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:90773cf4b7d9482336cb7373e0b8243c8344ad887fed0cc7d17074fd4f49ebe7
+oid sha256:b97a3768855a2926a623baaa768c0012153631a4a6ec1af8b2c3b1959b50fa09
 size 1064

last-checkpoint/trainer_state.json

@@ -1,10 +1,10 @@
 {
-  "best_global_step": 18000,
-  "best_metric": 0.7882509231567383,
-  "best_model_checkpoint": "./ar-diffusion-checkpoints-fixed/checkpoint-18000",
-  "epoch": 1.4229674640412275,
+  "best_global_step": 19000,
+  "best_metric": 0.7869898080825806,
+  "best_model_checkpoint": "./ar-diffusion-checkpoints-fixed/checkpoint-19000",
+  "epoch": 1.4614260441504499,
   "eval_steps": 250,
-  "global_step": 18500,
+  "global_step": 19000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -3190,6 +3190,92 @@
       "eval_samples_per_second": 55.638,
       "eval_steps_per_second": 13.909,
       "step": 18500
+    },
+    {
+      "epoch": 1.42681332205215,
+      "grad_norm": 0.6486705541610718,
+      "learning_rate": 0.00010628782172817703,
+      "loss": 0.7317,
+      "step": 18550
+    },
+    {
+      "epoch": 1.4306591800630721,
+      "grad_norm": 0.6570118069648743,
+      "learning_rate": 0.00010602810170636055,
+      "loss": 0.8149,
+      "step": 18600
+    },
+    {
+      "epoch": 1.4345050380739943,
+      "grad_norm": 0.8024285435676575,
+      "learning_rate": 0.00010576838168454406,
+      "loss": 0.8392,
+      "step": 18650
+    },
+    {
+      "epoch": 1.4383508960849165,
+      "grad_norm": 0.5735141038894653,
+      "learning_rate": 0.0001055086616627276,
+      "loss": 0.7932,
+      "step": 18700
+    },
+    {
+      "epoch": 1.4421967540958387,
+      "grad_norm": 0.5087122917175293,
+      "learning_rate": 0.0001052489416409111,
+      "loss": 0.7508,
+      "step": 18750
+    },
+    {
+      "epoch": 1.4421967540958387,
+      "eval_loss": 0.7895762920379639,
+      "eval_runtime": 17.9534,
+      "eval_samples_per_second": 55.7,
+      "eval_steps_per_second": 13.925,
+      "step": 18750
+    },
+    {
+      "epoch": 1.446042612106761,
+      "grad_norm": 0.7478468418121338,
+      "learning_rate": 0.00010498922161909461,
+      "loss": 0.8533,
+      "step": 18800
+    },
+    {
+      "epoch": 1.4498884701176833,
+      "grad_norm": 1.2165393829345703,
+      "learning_rate": 0.00010472950159727815,
+      "loss": 0.8025,
+      "step": 18850
+    },
+    {
+      "epoch": 1.4537343281286055,
+      "grad_norm": 0.8682180643081665,
+      "learning_rate": 0.00010446978157546165,
+      "loss": 0.8086,
+      "step": 18900
+    },
+    {
+      "epoch": 1.4575801861395277,
+      "grad_norm": 0.9063705205917358,
+      "learning_rate": 0.00010421006155364516,
+      "loss": 0.7583,
+      "step": 18950
+    },
+    {
+      "epoch": 1.4614260441504499,
+      "grad_norm": 0.7133361101150513,
+      "learning_rate": 0.0001039503415318287,
+      "loss": 0.7381,
+      "step": 19000
+    },
+    {
+      "epoch": 1.4614260441504499,
+      "eval_loss": 0.7869898080825806,
+      "eval_runtime": 17.9874,
+      "eval_samples_per_second": 55.595,
+      "eval_steps_per_second": 13.899,
+      "step": 19000
     }
   ],
   "logging_steps": 50,