Training in progress, step 24000, checkpoint

last-checkpoint/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e5e8c57b0540b43477103991d94a0031665bde117b5025873ffe2f31a08ca029
+oid sha256:a49f5cb3d3ccac7817610fa964bf361289eb0cfe2f4eb026b3d7dec09b14c94e
 size 36730224

last-checkpoint/ar_diffusion_info.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:dcb60822099d74359c979b9b670c4aabd0f2ec34f416894d009bb290edd18fb0
+oid sha256:4c4dc22fe695886da65efe80a9483d46131feb5c3951ee4e1441cd0c20f8ca7b
 size 1736

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a87d8a39dc457bb1ae5a270a42ba3d23dcea31e23f4fc916e1a9528a3ddc6ce1
+oid sha256:7232363cf74531ce8b422db0be4a1c52fb65a8350caa9a3033db48afd581b28e
 size 73588346

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:72244ff2f8e1c50254cff18940eace7cf87f75ca483c4693d1f58d7d7fc6e225
-size 14180
+oid sha256:599e4a0a057d34a3454abdec04ea1df53e6a5a47a3aac39f21ff339cea978175
+size 14244

last-checkpoint/scaler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1455d8897bfe51e2e7ebee8ab560ed718f3a24db9dec5faff5c72296032decc1
+oid sha256:25acdcc8c2ddbe44392c8342cb938e96014287130fce2f529b63734cdda74387
 size 988

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3a48972742d0232c3fc2605c286832dbdeaceae1cad253f23d64ff38411b92e0
+oid sha256:ac4e57d345b8d2e1b9cf78f859eede2962dbe55a1ef61e82edae2f8653a05425
 size 1064

last-checkpoint/trainer_state.json

@@ -1,10 +1,10 @@
 {
-  "best_global_step": 23500,
-  "best_metric": 0.7594859600067139,
-  "best_model_checkpoint": "./ar-diffusion-checkpoints-fixed/checkpoint-23500",
-  "epoch": 1.8075532651334512,
+  "best_global_step": 24000,
+  "best_metric": 0.7566245198249817,
+  "best_model_checkpoint": "./ar-diffusion-checkpoints-fixed/checkpoint-24000",
+  "epoch": 1.8460118452426735,
   "eval_steps": 250,
-  "global_step": 23500,
+  "global_step": 24000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -4050,6 +4050,92 @@
       "eval_samples_per_second": 56.022,
       "eval_steps_per_second": 14.005,
       "step": 23500
+    },
+    {
+      "epoch": 1.8113991231443736,
+      "grad_norm": 0.8289865851402283,
+      "learning_rate": 8.031581954652885e-05,
+      "loss": 0.7899,
+      "step": 23550
+    },
+    {
+      "epoch": 1.8152449811552958,
+      "grad_norm": 1.1237398386001587,
+      "learning_rate": 8.005609952471237e-05,
+      "loss": 0.8005,
+      "step": 23600
+    },
+    {
+      "epoch": 1.819090839166218,
+      "grad_norm": 0.8594374060630798,
+      "learning_rate": 7.979637950289588e-05,
+      "loss": 0.7765,
+      "step": 23650
+    },
+    {
+      "epoch": 1.8229366971771404,
+      "grad_norm": 0.754634439945221,
+      "learning_rate": 7.95366594810794e-05,
+      "loss": 0.7794,
+      "step": 23700
+    },
+    {
+      "epoch": 1.8267825551880623,
+      "grad_norm": 1.0647647380828857,
+      "learning_rate": 7.927693945926292e-05,
+      "loss": 0.7681,
+      "step": 23750
+    },
+    {
+      "epoch": 1.8267825551880623,
+      "eval_loss": 0.7575324773788452,
+      "eval_runtime": 17.8657,
+      "eval_samples_per_second": 55.973,
+      "eval_steps_per_second": 13.993,
+      "step": 23750
+    },
+    {
+      "epoch": 1.8306284131989847,
+      "grad_norm": 1.1255161762237549,
+      "learning_rate": 7.901721943744644e-05,
+      "loss": 0.7551,
+      "step": 23800
+    },
+    {
+      "epoch": 1.834474271209907,
+      "grad_norm": 0.8209452629089355,
+      "learning_rate": 7.875749941562995e-05,
+      "loss": 0.8091,
+      "step": 23850
+    },
+    {
+      "epoch": 1.838320129220829,
+      "grad_norm": 0.40779542922973633,
+      "learning_rate": 7.849777939381348e-05,
+      "loss": 0.7572,
+      "step": 23900
+    },
+    {
+      "epoch": 1.8421659872317515,
+      "grad_norm": 0.9186558127403259,
+      "learning_rate": 7.823805937199699e-05,
+      "loss": 0.7681,
+      "step": 23950
+    },
+    {
+      "epoch": 1.8460118452426735,
+      "grad_norm": 0.4896409213542938,
+      "learning_rate": 7.79783393501805e-05,
+      "loss": 0.7989,
+      "step": 24000
+    },
+    {
+      "epoch": 1.8460118452426735,
+      "eval_loss": 0.7566245198249817,
+      "eval_runtime": 17.9055,
+      "eval_samples_per_second": 55.849,
+      "eval_steps_per_second": 13.962,
+      "step": 24000
     }
   ],
   "logging_steps": 50,