Training in progress, step 19500, checkpoint

last-checkpoint/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:68e4017d8a95f1b3f117cff44c2d649b305cf089afdefcf8fc150daf09311398
+oid sha256:e938b4a2a45d05421534909af0bf009dedebb9aea74565d4ae4302594883dc8f
 size 36730224

last-checkpoint/ar_diffusion_info.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bd7fa9c8114b1914d7c2ff974dfcbf00be8238b2a035cfde8dba2f243c66132b
+oid sha256:7455f8d2c994f64aaeae6d25909ff5a023afe0e1744ef0734509e43b10f56157
 size 1736

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:976f2ad66948cadc63923a3a9d366716d6c39d193180a993074754d9c2d5088f
+oid sha256:d6cd326cd8b59bceb89f65022b035ca69baf384342ec2bf39dcc4ad1f53650ae
 size 73588346

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:62f0a63af717862f5ba78e609a46cd20ededb84bdab60287f44bdfc2444cb5a9
-size 14308
+oid sha256:53b32835bd0fc28ffeb2c034cbe6942df7d91018aa69d3614686950a58fc9920
+size 14244

last-checkpoint/scaler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b99bea7b1fa648da24e0513c2b62a349744452626bc71a88bb8bbac50aa0f8ce
+oid sha256:2dda7fe4be06e7b2f28f9fb6d77a55fd50f883f30376b3be3fec2001c9515fe5
 size 988

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b97a3768855a2926a623baaa768c0012153631a4a6ec1af8b2c3b1959b50fa09
+oid sha256:24a9ece5494351db57cf0201c68e80d1bab27482ad54e226480b7b466bc789e5
 size 1064

last-checkpoint/trainer_state.json

@@ -1,10 +1,10 @@
 {
-  "best_global_step": 19000,
-  "best_metric": 0.7869898080825806,
-  "best_model_checkpoint": "./ar-diffusion-checkpoints-fixed/checkpoint-19000",
-  "epoch": 1.4614260441504499,
+  "best_global_step": 19500,
+  "best_metric": 0.7843549847602844,
+  "best_model_checkpoint": "./ar-diffusion-checkpoints-fixed/checkpoint-19500",
+  "epoch": 1.4998846242596724,
   "eval_steps": 250,
-  "global_step": 19000,
+  "global_step": 19500,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -3276,6 +3276,92 @@
       "eval_samples_per_second": 55.595,
       "eval_steps_per_second": 13.899,
       "step": 19000
+    },
+    {
+      "epoch": 1.4652719021613723,
+      "grad_norm": 0.6205143928527832,
+      "learning_rate": 0.00010369062151001222,
+      "loss": 0.8141,
+      "step": 19050
+    },
+    {
+      "epoch": 1.4691177601722945,
+      "grad_norm": 1.1060974597930908,
+      "learning_rate": 0.00010343090148819572,
+      "loss": 0.7771,
+      "step": 19100
+    },
+    {
+      "epoch": 1.4729636181832166,
+      "grad_norm": 0.7808921933174133,
+      "learning_rate": 0.00010317118146637926,
+      "loss": 0.8006,
+      "step": 19150
+    },
+    {
+      "epoch": 1.476809476194139,
+      "grad_norm": 0.5509454011917114,
+      "learning_rate": 0.00010291146144456277,
+      "loss": 0.7922,
+      "step": 19200
+    },
+    {
+      "epoch": 1.480655334205061,
+      "grad_norm": 1.2427464723587036,
+      "learning_rate": 0.0001026517414227463,
+      "loss": 0.7801,
+      "step": 19250
+    },
+    {
+      "epoch": 1.480655334205061,
+      "eval_loss": 0.7850660085678101,
+      "eval_runtime": 17.8458,
+      "eval_samples_per_second": 56.036,
+      "eval_steps_per_second": 14.009,
+      "step": 19250
+    },
+    {
+      "epoch": 1.4845011922159834,
+      "grad_norm": 0.8217543959617615,
+      "learning_rate": 0.00010239202140092981,
+      "loss": 0.782,
+      "step": 19300
+    },
+    {
+      "epoch": 1.4883470502269056,
+      "grad_norm": 0.6753976941108704,
+      "learning_rate": 0.00010213230137911332,
+      "loss": 0.7787,
+      "step": 19350
+    },
+    {
+      "epoch": 1.4921929082378278,
+      "grad_norm": 1.0264923572540283,
+      "learning_rate": 0.00010187258135729685,
+      "loss": 0.7713,
+      "step": 19400
+    },
+    {
+      "epoch": 1.4960387662487502,
+      "grad_norm": 0.7317076325416565,
+      "learning_rate": 0.00010161286133548036,
+      "loss": 0.7634,
+      "step": 19450
+    },
+    {
+      "epoch": 1.4998846242596724,
+      "grad_norm": 0.6011013388633728,
+      "learning_rate": 0.00010135314131366387,
+      "loss": 0.7715,
+      "step": 19500
+    },
+    {
+      "epoch": 1.4998846242596724,
+      "eval_loss": 0.7843549847602844,
+      "eval_runtime": 17.919,
+      "eval_samples_per_second": 55.807,
+      "eval_steps_per_second": 13.952,
+      "step": 19500
     }
   ],
   "logging_steps": 50,