Training in progress, step 17500, checkpoint

last-checkpoint/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1d163d6a9ddc974e2180a0b34dccdebf25df2e06b238d605670c84471509b1d7
+oid sha256:c504a27eb69195f0629a47f9e0bf192539abb3c2443da0cc76cbb0d75cfd5228
 size 36730224

last-checkpoint/ar_diffusion_info.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:48e55dadd5444de13311e56742ff892d8682c557f87064412817be365756d8c7
+oid sha256:7f425ddc672454ec318d3a477b01d57b3febdbf4bb7f94794d250d01447f38c3
 size 1736

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fc77b7658a34f923b1da47ba7dbd2fd74f55df477ff11a70a0b6f32488a49438
+oid sha256:c2c9d67d788a0699ac5c4a4b6a7c0a6433a765152542a0d4c63e46c093056b65
 size 73588346

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a815a54c3da05d513f8fc01a4362c2083ab81366f45a37e94100e41551fa2292
+oid sha256:a8ffe3d878b186570864a01f18402ac15fa8e666ba6230c15554b4863a170b12
 size 14244

last-checkpoint/scaler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6862466a821eb915ca4f2b1057c449dfe2dac434c155b9783db0cf817969e1cf
+oid sha256:2e4317872e22f122cc23eecd5b16d2ae8cc3637fa26840af5163be9ef4599ef9
 size 988

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3597272c85a71f80afe5d0d108e62f0719dbbc66ca2fad8b5c08fb0ab04b2895
+oid sha256:1f03e29ee3758301caaed8e1dd1dbe2a845f2652e79a6f8508f414e5c4c84bdf
 size 1064

last-checkpoint/trainer_state.json

@@ -1,10 +1,10 @@
 {
-  "best_global_step": 17000,
-  "best_metric": 0.8002509474754333,
+  "best_global_step": 17250,
+  "best_metric": 0.7951701879501343,
   "best_model_checkpoint": "./ar-diffusion-checkpoints-fixed/checkpoint-17000",
-  "epoch": 1.3075917237135606,
+  "epoch": 1.346050303822783,
   "eval_steps": 250,
-  "global_step": 17000,
+  "global_step": 17500,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -2932,6 +2932,92 @@
       "eval_samples_per_second": 55.712,
       "eval_steps_per_second": 13.928,
       "step": 17000
+    },
+    {
+      "epoch": 1.3114375817244828,
+      "grad_norm": 1.403998851776123,
+      "learning_rate": 0.00011407942238267149,
+      "loss": 0.8369,
+      "step": 17050
+    },
+    {
+      "epoch": 1.315283439735405,
+      "grad_norm": 0.8802525401115417,
+      "learning_rate": 0.000113819702360855,
+      "loss": 0.7818,
+      "step": 17100
+    },
+    {
+      "epoch": 1.3191292977463271,
+      "grad_norm": 0.7463178634643555,
+      "learning_rate": 0.00011355998233903852,
+      "loss": 0.7661,
+      "step": 17150
+    },
+    {
+      "epoch": 1.3229751557572493,
+      "grad_norm": 0.47771725058555603,
+      "learning_rate": 0.00011330026231722204,
+      "loss": 0.8141,
+      "step": 17200
+    },
+    {
+      "epoch": 1.3268210137681717,
+      "grad_norm": 0.8294114470481873,
+      "learning_rate": 0.00011304054229540555,
+      "loss": 0.7979,
+      "step": 17250
+    },
+    {
+      "epoch": 1.3268210137681717,
+      "eval_loss": 0.7951701879501343,
+      "eval_runtime": 17.958,
+      "eval_samples_per_second": 55.686,
+      "eval_steps_per_second": 13.921,
+      "step": 17250
+    },
+    {
+      "epoch": 1.330666871779094,
+      "grad_norm": 1.0813144445419312,
+      "learning_rate": 0.00011278082227358908,
+      "loss": 0.824,
+      "step": 17300
+    },
+    {
+      "epoch": 1.3345127297900161,
+      "grad_norm": 0.5647908449172974,
+      "learning_rate": 0.00011252110225177259,
+      "loss": 0.7718,
+      "step": 17350
+    },
+    {
+      "epoch": 1.3383585878009385,
+      "grad_norm": 0.7901347279548645,
+      "learning_rate": 0.0001122613822299561,
+      "loss": 0.7883,
+      "step": 17400
+    },
+    {
+      "epoch": 1.3422044458118605,
+      "grad_norm": 1.0769431591033936,
+      "learning_rate": 0.00011200166220813963,
+      "loss": 0.7625,
+      "step": 17450
+    },
+    {
+      "epoch": 1.346050303822783,
+      "grad_norm": 0.7366082072257996,
+      "learning_rate": 0.00011174194218632314,
+      "loss": 0.8167,
+      "step": 17500
+    },
+    {
+      "epoch": 1.346050303822783,
+      "eval_loss": 0.7991219758987427,
+      "eval_runtime": 17.9932,
+      "eval_samples_per_second": 55.577,
+      "eval_steps_per_second": 13.894,
+      "step": 17500
     }
   ],
   "logging_steps": 50,