Training in progress, step 18500, checkpoint

last-checkpoint/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:42c2f6139756d4179a8f7b2fd49bb0a2cd540ba8b774f9e69c5f50f17435f6b6
+oid sha256:566ede041af9927aaf3a99063bbbb39906d51b0dae6dc2971321cd81bbe6d46b
 size 36730224

last-checkpoint/ar_diffusion_info.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d5bffd3fc7ee80c7d1ba70865a03ab1d358a04bee58d41c0ac14d5c087eae56f
+oid sha256:8245a854cccd4f75ef67a2fc0a146b42b9184d6fe923ca2ace73890f7dc48c4e
 size 1544

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7b5e18457edaea63d6ccb66baccbacd52b196222165882275df4a7c9e0e28b1b
+oid sha256:14726f58d8d95f5ffac4c91fb097e80457cec53c8bddd8e102b784c0cb3b5e7e
 size 73588346

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:aec74914111c893531d77390ed26a8e62f47c5fd368563e3e2ea395d9971bdd0
+oid sha256:a637703e63dd26bf3fc5d45021980dfb8c6c15b36b34b2b812e241dcfa5b3276
 size 14244

last-checkpoint/scaler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d21f0ba0bbe7a28de5199a35a5f642d45bf19d1eb0aa770d1035c6ab0a649bfa
+oid sha256:9b057aaa1f69d73e0bf4ef4b900fbc14f133252f8a709b65c493591309b84be0
 size 988

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:47960831c3aee4c1e74537230bffbd7196098e4c54fe510c8e622c62e1271790
+oid sha256:3b19e0d45d48155881edd9e1c841838b3186845172b736cd05c0110b70c24971
 size 1064

last-checkpoint/trainer_state.json

@@ -2,9 +2,9 @@
   "best_global_step": 16750,
   "best_metric": 1.4798808097839355,
   "best_model_checkpoint": "./ar-diffusion-checkpoints-fixed/checkpoint-15500",
-  "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": 56.369,
       "eval_steps_per_second": 14.092,
       "step": 18000
+    },
+    {
+      "epoch": 1.3883547419429274,
+      "grad_norm": 1.8189436197280884,
+      "learning_rate": 0.0001088954107472145,
+      "loss": 1.4872,
+      "step": 18050
+    },
+    {
+      "epoch": 1.3922005999538496,
+      "grad_norm": 1.4212762117385864,
+      "learning_rate": 0.00010863569072539803,
+      "loss": 1.4067,
+      "step": 18100
+    },
+    {
+      "epoch": 1.396046457964772,
+      "grad_norm": 1.9733264446258545,
+      "learning_rate": 0.00010837597070358154,
+      "loss": 1.4787,
+      "step": 18150
+    },
+    {
+      "epoch": 1.3998923159756942,
+      "grad_norm": 1.320064663887024,
+      "learning_rate": 0.00010811625068176505,
+      "loss": 1.5242,
+      "step": 18200
+    },
+    {
+      "epoch": 1.4037381739866164,
+      "grad_norm": 1.4177141189575195,
+      "learning_rate": 0.00010785653065994858,
+      "loss": 1.4719,
+      "step": 18250
+    },
+    {
+      "epoch": 1.4037381739866164,
+      "eval_loss": 1.4992233514785767,
+      "eval_runtime": 17.8574,
+      "eval_samples_per_second": 55.999,
+      "eval_steps_per_second": 14.0,
+      "step": 18250
+    },
+    {
+      "epoch": 1.4075840319975388,
+      "grad_norm": 1.8219791650772095,
+      "learning_rate": 0.00010759681063813209,
+      "loss": 1.4392,
+      "step": 18300
+    },
+    {
+      "epoch": 1.4114298900084608,
+      "grad_norm": 1.3196603059768677,
+      "learning_rate": 0.00010733709061631563,
+      "loss": 1.4837,
+      "step": 18350
+    },
+    {
+      "epoch": 1.4152757480193832,
+      "grad_norm": 1.01405668258667,
+      "learning_rate": 0.00010707737059449914,
+      "loss": 1.455,
+      "step": 18400
+    },
+    {
+      "epoch": 1.4191216060303053,
+      "grad_norm": 1.8538917303085327,
+      "learning_rate": 0.00010681765057268264,
+      "loss": 1.4209,
+      "step": 18450
+    },
+    {
+      "epoch": 1.4229674640412275,
+      "grad_norm": 0.8785907030105591,
+      "learning_rate": 0.00010655793055086618,
+      "loss": 1.4716,
+      "step": 18500
+    },
+    {
+      "epoch": 1.4229674640412275,
+      "eval_loss": 1.485592246055603,
+      "eval_runtime": 17.7246,
+      "eval_samples_per_second": 56.419,
+      "eval_steps_per_second": 14.105,
+      "step": 18500
     }
   ],
   "logging_steps": 50,