Training in progress, step 26000, checkpoint

last-checkpoint/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8492be170f7ead3e1de3a8a4310c2bd0ea5b6f9e4d86bcd7d0ba2698651e52bd
+oid sha256:5449fe8d926b1cee1305d48c4da86aabc3eeaa5ef355125673a908672ff0f8ba
 size 36730224

last-checkpoint/ar_diffusion_info.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ab9a2b23fe7fd93945a85cfc1259da800f7071f45eb31fd4ed7e630b682c35ca
+oid sha256:2510c646e73f6cabed1de03f0b56fcfa92afe5e634cb545125f01b5222bd143a
 size 1736

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:32da047c4d4b572ed1b8f48061cec167cdf94b2d474f921721549d72ae412177
+oid sha256:5454bb0ba74be0bd430d5c8f427657e76cb271799efbec274319b6293d6564b0
 size 73588346

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fa86ead0557145b8c2ea7bbd3e68631cb5af19c8565a5c11d759d3e4217c56ab
+oid sha256:ed589b2d77422e032b4b6ae68314c1a0925e5e1242b78a501c8f440d71327775
 size 14244

last-checkpoint/scaler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:98ac3712aca76800ec0076b19c413d355103c14ed7b94cc79d36faba9193644b
+oid sha256:f65a49a596607624282a94034043f4b77f0176e8981d34a9018e5c611413b3c2
 size 988

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:394944f715215f9b5f72270954d983bc0552f0bf05252a3f3b1abae560af7063
+oid sha256:577a1bc075685d4719861fa605dd82a8dd946efd35c621e30d8878d2dd122165
 size 1064

last-checkpoint/trainer_state.json

@@ -1,10 +1,10 @@
 {
-  "best_global_step": 25250,
-  "best_metric": 0.7520478963851929,
-  "best_model_checkpoint": "./ar-diffusion-checkpoints-fixed/checkpoint-24500",
-  "epoch": 1.9613875855703409,
+  "best_global_step": 26000,
+  "best_metric": 0.7482460737228394,
+  "best_model_checkpoint": "./ar-diffusion-checkpoints-fixed/checkpoint-26000",
+  "epoch": 1.9998461656795632,
   "eval_steps": 250,
-  "global_step": 25500,
+  "global_step": 26000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -4394,6 +4394,92 @@
       "eval_samples_per_second": 55.914,
       "eval_steps_per_second": 13.979,
       "step": 25500
+    },
+    {
+      "epoch": 1.9652334435812628,
+      "grad_norm": 1.2154541015625,
+      "learning_rate": 6.99322130743059e-05,
+      "loss": 0.7644,
+      "step": 25550
+    },
+    {
+      "epoch": 1.9690793015921852,
+      "grad_norm": 0.715004026889801,
+      "learning_rate": 6.967249305248941e-05,
+      "loss": 0.7235,
+      "step": 25600
+    },
+    {
+      "epoch": 1.9729251596031074,
+      "grad_norm": 0.44530218839645386,
+      "learning_rate": 6.941277303067294e-05,
+      "loss": 0.7591,
+      "step": 25650
+    },
+    {
+      "epoch": 1.9767710176140296,
+      "grad_norm": 0.7103247046470642,
+      "learning_rate": 6.915305300885647e-05,
+      "loss": 0.7203,
+      "step": 25700
+    },
+    {
+      "epoch": 1.980616875624952,
+      "grad_norm": 0.6260993480682373,
+      "learning_rate": 6.88985273874763e-05,
+      "loss": 0.8008,
+      "step": 25750
+    },
+    {
+      "epoch": 1.980616875624952,
+      "eval_loss": 0.7489978075027466,
+      "eval_runtime": 17.8489,
+      "eval_samples_per_second": 56.026,
+      "eval_steps_per_second": 14.006,
+      "step": 25750
+    },
+    {
+      "epoch": 1.984462733635874,
+      "grad_norm": 0.8690526485443115,
+      "learning_rate": 6.863880736565982e-05,
+      "loss": 0.7445,
+      "step": 25800
+    },
+    {
+      "epoch": 1.9883085916467964,
+      "grad_norm": 0.8287826776504517,
+      "learning_rate": 6.837908734384334e-05,
+      "loss": 0.7606,
+      "step": 25850
+    },
+    {
+      "epoch": 1.9921544496577186,
+      "grad_norm": 0.9105169773101807,
+      "learning_rate": 6.811936732202686e-05,
+      "loss": 0.7742,
+      "step": 25900
+    },
+    {
+      "epoch": 1.9960003076686408,
+      "grad_norm": 0.6223366856575012,
+      "learning_rate": 6.785964730021037e-05,
+      "loss": 0.8069,
+      "step": 25950
+    },
+    {
+      "epoch": 1.9998461656795632,
+      "grad_norm": 0.848816454410553,
+      "learning_rate": 6.75999272783939e-05,
+      "loss": 0.8126,
+      "step": 26000
+    },
+    {
+      "epoch": 1.9998461656795632,
+      "eval_loss": 0.7482460737228394,
+      "eval_runtime": 17.8823,
+      "eval_samples_per_second": 55.921,
+      "eval_steps_per_second": 13.98,
+      "step": 26000
     }
   ],
   "logging_steps": 50,