Training in progress, step 18500, checkpoint

last-checkpoint/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ab1600c15a37ec6fb41eca032481fb072aba7998a063361aac9692ac6a9ef216
+oid sha256:2468b8f20b2b8549a640818b37e1b9ef1f114a27b7f6c99b4a86e82f642e4401
 size 36730224

last-checkpoint/ar_diffusion_info.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:16cac24c8d65482bd4344116193b8cb0973617948ed3db3f9320f39d3b079b5f
+oid sha256:c29781dc06aac24acf8000f8feb9e000d60a160a71b0c570ed32b320a8756db2
 size 1736

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:245296adc55ccbb77650410469534fb289b0cd17f03dec26c0aee8f21a8e2682
+oid sha256:e40903d472c0cd2d111f5868dfbf44c27a4ee11fa862e96214b77a89c58a807b
 size 73588346

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:69e9bd7ace00dc839f848c20ae8e1e8fe1387810881778a2284fcf0ac2ad2525
+oid sha256:e90dd042b4fe16d601e9434fac855ce20c91c9b7d9c2a6dff1f998d78c9a11bd
 size 14244

last-checkpoint/scaler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ac398991953ddd2d4574773fb96cf571b99bfe36340ab697de86477c38530f18
+oid sha256:17dcf088a8e3107cea49076c2140910eebc6fe4528f365b01ee8c2ed883b6684
 size 988

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8d68c6e80dd0b62a913dae04c12415a6a255632f5af4175f8cdb6a83a0bbfe28
+oid sha256:90773cf4b7d9482336cb7373e0b8243c8344ad887fed0cc7d17074fd4f49ebe7
 size 1064

last-checkpoint/trainer_state.json

@@ -2,9 +2,9 @@
   "best_global_step": 18000,
   "best_metric": 0.7882509231567383,
   "best_model_checkpoint": "./ar-diffusion-checkpoints-fixed/checkpoint-18000",
-  "epoch": 1.3845088839320052,
+  "epoch": 1.4229674640412275,
   "eval_steps": 250,
-  "global_step": 18000,
+  "global_step": 18500,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -3104,6 +3104,92 @@
       "eval_samples_per_second": 55.748,
       "eval_steps_per_second": 13.937,
       "step": 18000
+    },
+    {
+      "epoch": 1.3883547419429274,
+      "grad_norm": 0.9016252160072327,
+      "learning_rate": 0.00010888502194634185,
+      "loss": 0.8222,
+      "step": 18050
+    },
+    {
+      "epoch": 1.3922005999538496,
+      "grad_norm": 0.6058124899864197,
+      "learning_rate": 0.00010862530192452536,
+      "loss": 0.7657,
+      "step": 18100
+    },
+    {
+      "epoch": 1.396046457964772,
+      "grad_norm": 0.8505234122276306,
+      "learning_rate": 0.00010836558190270889,
+      "loss": 0.7962,
+      "step": 18150
+    },
+    {
+      "epoch": 1.3998923159756942,
+      "grad_norm": 0.7518420815467834,
+      "learning_rate": 0.0001081058618808924,
+      "loss": 0.8364,
+      "step": 18200
+    },
+    {
+      "epoch": 1.4037381739866164,
+      "grad_norm": 0.7778449058532715,
+      "learning_rate": 0.00010784614185907592,
+      "loss": 0.7836,
+      "step": 18250
+    },
+    {
+      "epoch": 1.4037381739866164,
+      "eval_loss": 0.7889594435691833,
+      "eval_runtime": 17.8326,
+      "eval_samples_per_second": 56.077,
+      "eval_steps_per_second": 14.019,
+      "step": 18250
+    },
+    {
+      "epoch": 1.4075840319975388,
+      "grad_norm": 1.2029508352279663,
+      "learning_rate": 0.00010758642183725944,
+      "loss": 0.7859,
+      "step": 18300
+    },
+    {
+      "epoch": 1.4114298900084608,
+      "grad_norm": 0.7625166773796082,
+      "learning_rate": 0.00010732670181544295,
+      "loss": 0.8018,
+      "step": 18350
+    },
+    {
+      "epoch": 1.4152757480193832,
+      "grad_norm": 0.6327937245368958,
+      "learning_rate": 0.00010706698179362648,
+      "loss": 0.808,
+      "step": 18400
+    },
+    {
+      "epoch": 1.4191216060303053,
+      "grad_norm": 0.7097195386886597,
+      "learning_rate": 0.00010680726177181,
+      "loss": 0.7928,
+      "step": 18450
+    },
+    {
+      "epoch": 1.4229674640412275,
+      "grad_norm": 0.5188928246498108,
+      "learning_rate": 0.00010654754174999351,
+      "loss": 0.7934,
+      "step": 18500
+    },
+    {
+      "epoch": 1.4229674640412275,
+      "eval_loss": 0.7918493151664734,
+      "eval_runtime": 17.9735,
+      "eval_samples_per_second": 55.638,
+      "eval_steps_per_second": 13.909,
+      "step": 18500
     }
   ],
   "logging_steps": 50,