Training in progress, step 3500, checkpoint

last-checkpoint/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e598bf8b3c9e467b65b1bb82c8cbdd7f6d9c1972d320fbed970e890963dcdc5d
+oid sha256:cfa682c656512f0db38279360e472eed38f6e68c384179fb8ebc379ae0bbd931
 size 36730224

last-checkpoint/ar_diffusion_info.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7d2d6f677b30f619e4a135ea8b093f0ab4667e8aba8c81cd7afaa73bd22e2ff8
+oid sha256:2baaf7d1ff2607c2b51f6cfbe05d40604cadc3c0c83d92adf1fa1aa8878e8345
 size 1544

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4dc13a71e65c3e33eb37c20ebdf6182925491cef76344bbf33cf66c0cd5d4873
+oid sha256:9b3473e19f33d061d72e874bb7ead12bc6195f3a6d1ef6cb2617c3cd645bff56
 size 73588346

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0cdad770e6415de25f2503b80a04125ce36c191e7ebbe40cc70935c15b3079b5
-size 14180
+oid sha256:32c4abfc978e57dde0e9122ce6e0a211a14cdb3e9d7e1f8e79d82dbf8734a5a1
+size 14244

last-checkpoint/scaler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1fd6bb9f68e7b13e6013a02cf99adaeff381c90e8c38480bfbcc7ea752fc628b
+oid sha256:56883ca08ba7d3883d8223a68a16de2347a2e5eee3a5f6cfa3b0ef70d1be127c
 size 988

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1e2f4f94b8df8c39a3dbd1dfca81bfb52c1274fbec2120e8d28a3f0e8b8a87c3
+oid sha256:336a48abe3f0221158899799d17e11e8d7fd80ac42ba2794eb638e218cd1c8a0
 size 1064

last-checkpoint/trainer_state.json

@@ -1,10 +1,10 @@
 {
-  "best_global_step": 3000,
-  "best_metric": 4.579595565795898,
-  "best_model_checkpoint": "./ar-diffusion-checkpoints-fixed/checkpoint-3000",
-  "epoch": 0.2307514806553342,
+  "best_global_step": 3500,
+  "best_metric": 4.520543098449707,
+  "best_model_checkpoint": "./ar-diffusion-checkpoints-fixed/checkpoint-3500",
+  "epoch": 0.26921006076455656,
   "eval_steps": 250,
-  "global_step": 3000,
+  "global_step": 3500,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -524,6 +524,92 @@
       "eval_samples_per_second": 54.124,
       "eval_steps_per_second": 13.531,
       "step": 3000
+    },
+    {
+      "epoch": 0.23459733866625646,
+      "grad_norm": 6.408226490020752,
+      "learning_rate": 0.00018679063969041375,
+      "loss": 4.6614,
+      "step": 3050
+    },
+    {
+      "epoch": 0.23844319667717867,
+      "grad_norm": 6.529899597167969,
+      "learning_rate": 0.00018653091966859726,
+      "loss": 4.5576,
+      "step": 3100
+    },
+    {
+      "epoch": 0.2422890546881009,
+      "grad_norm": 4.417705059051514,
+      "learning_rate": 0.00018627119964678078,
+      "loss": 4.5466,
+      "step": 3150
+    },
+    {
+      "epoch": 0.24613491269902316,
+      "grad_norm": 3.052746295928955,
+      "learning_rate": 0.0001860114796249643,
+      "loss": 4.6959,
+      "step": 3200
+    },
+    {
+      "epoch": 0.2499807707099454,
+      "grad_norm": 5.121955871582031,
+      "learning_rate": 0.0001857517596031478,
+      "loss": 4.6796,
+      "step": 3250
+    },
+    {
+      "epoch": 0.2499807707099454,
+      "eval_loss": 4.576270580291748,
+      "eval_runtime": 18.8896,
+      "eval_samples_per_second": 52.939,
+      "eval_steps_per_second": 13.235,
+      "step": 3250
+    },
+    {
+      "epoch": 0.25382662872086764,
+      "grad_norm": 2.772292375564575,
+      "learning_rate": 0.00018549203958133131,
+      "loss": 4.534,
+      "step": 3300
+    },
+    {
+      "epoch": 0.2576724867317899,
+      "grad_norm": 6.105401039123535,
+      "learning_rate": 0.00018523231955951485,
+      "loss": 4.4958,
+      "step": 3350
+    },
+    {
+      "epoch": 0.26151834474271207,
+      "grad_norm": 2.742037057876587,
+      "learning_rate": 0.00018497259953769837,
+      "loss": 4.5835,
+      "step": 3400
+    },
+    {
+      "epoch": 0.2653642027536343,
+      "grad_norm": 3.6125235557556152,
+      "learning_rate": 0.0001847128795158819,
+      "loss": 4.5917,
+      "step": 3450
+    },
+    {
+      "epoch": 0.26921006076455656,
+      "grad_norm": 3.8596699237823486,
+      "learning_rate": 0.00018445315949406542,
+      "loss": 4.4909,
+      "step": 3500
+    },
+    {
+      "epoch": 0.26921006076455656,
+      "eval_loss": 4.520543098449707,
+      "eval_runtime": 18.7111,
+      "eval_samples_per_second": 53.444,
+      "eval_steps_per_second": 13.361,
+      "step": 3500
     }
   ],
   "logging_steps": 50,