Training in progress, step 28500, checkpoint

last-checkpoint/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c2b2a09b0c718d7a491a19203c596e4ee868ab1c484877d294e6c4faf6f30efb
+oid sha256:fe6e2fa1318dd3ad1205203e4fff88f82e693270053626bb1d002f9b23c9b24c
 size 36730224

last-checkpoint/ar_diffusion_info.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:84d5f993c46a18f90bdec40ac57cbb61c0044ca158da830a2a251cb1b9f73c26
+oid sha256:599432259819e93dad17855a204a171f8d5bf2703ad47dd88ce1d465a13b2f0b
 size 1544

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f145355c23ee5fb538578cd9954c12d1f33dfb98ccfc124985339d68356ee942
+oid sha256:48da092ff21d2489fb67e59a49519237044b07edc1b34a941910a561850d4e26
 size 73588346

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a374ffed3d3d8a7f9634cdb9b7207d347f6957d04522010a4d50d73e4c623442
+oid sha256:1e10c23934142af0d6a52cc080658216fed6f296b2eaa84b6cfd6a75c3f4683b
 size 14244

last-checkpoint/scaler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:13f83252b2ae8a53cd4f408d431cb4cec02e9d65b9dadef7fdff56180c3d5cd3
+oid sha256:575e45baa8bffe406e93f42f00b83861120a81ba67710e20edf9cb739fb34380
 size 988

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:969919a5c5fc35784da996c8ca0a4cb297b59fb4c9aadbd1a769e70ab51007e7
+oid sha256:841075f91d905b0f965b453df9b09a92889e8e8880ee6a19f7c088b54a9e11c6
 size 1064

last-checkpoint/trainer_state.json

@@ -1,10 +1,10 @@
 {
-  "best_global_step": 28000,
-  "best_metric": 1.008616328239441,
-  "best_model_checkpoint": "./ar-diffusion-checkpoints-fixed/checkpoint-28000",
-  "epoch": 2.1536804861164525,
+  "best_global_step": 28500,
+  "best_metric": 1.0052642822265625,
+  "best_model_checkpoint": "./ar-diffusion-checkpoints-fixed/checkpoint-28500",
+  "epoch": 2.192139066225675,
   "eval_steps": 250,
-  "global_step": 28000,
+  "global_step": 28500,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -4824,6 +4824,92 @@
       "eval_samples_per_second": 56.661,
       "eval_steps_per_second": 14.165,
       "step": 28000
+    },
+    {
+      "epoch": 2.157526344127375,
+      "grad_norm": 1.3012685775756836,
+      "learning_rate": 5.696698958522713e-05,
+      "loss": 1.0703,
+      "step": 28050
+    },
+    {
+      "epoch": 2.1613722021382973,
+      "grad_norm": 0.8459142446517944,
+      "learning_rate": 5.6707269563410646e-05,
+      "loss": 1.0507,
+      "step": 28100
+    },
+    {
+      "epoch": 2.1652180601492192,
+      "grad_norm": 0.6889505982398987,
+      "learning_rate": 5.6447549541594166e-05,
+      "loss": 1.0056,
+      "step": 28150
+    },
+    {
+      "epoch": 2.1690639181601417,
+      "grad_norm": 1.2236456871032715,
+      "learning_rate": 5.618782951977768e-05,
+      "loss": 0.988,
+      "step": 28200
+    },
+    {
+      "epoch": 2.1729097761710636,
+      "grad_norm": 1.3419203758239746,
+      "learning_rate": 5.59281094979612e-05,
+      "loss": 0.9467,
+      "step": 28250
+    },
+    {
+      "epoch": 2.1729097761710636,
+      "eval_loss": 1.0140153169631958,
+      "eval_runtime": 17.7299,
+      "eval_samples_per_second": 56.402,
+      "eval_steps_per_second": 14.1,
+      "step": 28250
+    },
+    {
+      "epoch": 2.176755634181986,
+      "grad_norm": 0.8394871354103088,
+      "learning_rate": 5.5668389476144725e-05,
+      "loss": 0.9824,
+      "step": 28300
+    },
+    {
+      "epoch": 2.1806014921929084,
+      "grad_norm": 0.7906908392906189,
+      "learning_rate": 5.540866945432823e-05,
+      "loss": 0.9915,
+      "step": 28350
+    },
+    {
+      "epoch": 2.1844473502038304,
+      "grad_norm": 1.0852785110473633,
+      "learning_rate": 5.514894943251175e-05,
+      "loss": 0.9473,
+      "step": 28400
+    },
+    {
+      "epoch": 2.188293208214753,
+      "grad_norm": 0.9187583923339844,
+      "learning_rate": 5.488922941069528e-05,
+      "loss": 0.9418,
+      "step": 28450
+    },
+    {
+      "epoch": 2.192139066225675,
+      "grad_norm": 0.7200838923454285,
+      "learning_rate": 5.4629509388878797e-05,
+      "loss": 0.9776,
+      "step": 28500
+    },
+    {
+      "epoch": 2.192139066225675,
+      "eval_loss": 1.0052642822265625,
+      "eval_runtime": 17.5069,
+      "eval_samples_per_second": 57.12,
+      "eval_steps_per_second": 14.28,
+      "step": 28500
     }
   ],
   "logging_steps": 50,