Training in progress, step 12500, checkpoint

last-checkpoint/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f905e1780ead207018fcc63d3eee314969b429ab6c3fb1b4b7d55b82dd5e705f
+oid sha256:4563a36b3299867126e2f2872e4b74a5609a7c6cb213d2085d095900f9215fa3
 size 36730224

last-checkpoint/ar_diffusion_info.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0543052e8eb8d6818c70e093e941fc4b27e91541fab33a19886d162002c431a7
+oid sha256:f9ead48f8f886b45b9f89bef7b002712a2b38bb6cc69a0971816a685b1318dd1
 size 1544

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7a12af399801c91cf8dd422d90b7d653b67f6015e364538d005d0a7d95d3b977
+oid sha256:ea5c1fbeda3adc57688fb9a2ac22bbf43523ff5c2f34a45a7dd94247c865d2ac
 size 73588346

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:205a7550a4c055924e0126c3501b4b656ca55ae3243c46218f86d29fc2fb9758
-size 14244
+oid sha256:9a5ae118ec673d57a7f62a6284956ef83acf5113a9420dcd7ba5fd941799b59c
+size 14180

last-checkpoint/scaler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a647bb95d822173d4ee9be8d10c936b89d3eece1d84247711563ab2d41513f54
+oid sha256:fe8cf9d8bac2dd71fd820efd0a6b09790c369caf5e00f7d2853b196fd0826191
 size 988

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5752b9d4c614021fa8b8f77fd50a0a39abf6d6bd5fececfd9db71cacad56f038
+oid sha256:fab19116031cb09fcbccb12fc7d5785df4d500dae544895e7f9a43d44ebf2019
 size 1064

last-checkpoint/trainer_state.json

@@ -1,10 +1,10 @@
 {
-  "best_global_step": 12000,
-  "best_metric": 1.5273067951202393,
+  "best_global_step": 12250,
+  "best_metric": 1.5220181941986084,
   "best_model_checkpoint": "./ar-diffusion-checkpoints-fixed/checkpoint-12000",
-  "epoch": 0.9230059226213368,
+  "epoch": 0.9614645027305592,
   "eval_steps": 250,
-  "global_step": 12000,
+  "global_step": 12500,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -2072,6 +2072,92 @@
       "eval_samples_per_second": 56.414,
       "eval_steps_per_second": 14.104,
       "step": 12000
+    },
+    {
+      "epoch": 0.926851780632259,
+      "grad_norm": 1.9142653942108154,
+      "learning_rate": 0.000140056618964756,
+      "loss": 1.5436,
+      "step": 12050
+    },
+    {
+      "epoch": 0.9306976386431813,
+      "grad_norm": 1.9730989933013916,
+      "learning_rate": 0.00013979689894293954,
+      "loss": 1.5175,
+      "step": 12100
+    },
+    {
+      "epoch": 0.9345434966541035,
+      "grad_norm": 0.9956797361373901,
+      "learning_rate": 0.00013953717892112302,
+      "loss": 1.4491,
+      "step": 12150
+    },
+    {
+      "epoch": 0.9383893546650258,
+      "grad_norm": 1.0358608961105347,
+      "learning_rate": 0.00013927745889930654,
+      "loss": 1.4488,
+      "step": 12200
+    },
+    {
+      "epoch": 0.942235212675948,
+      "grad_norm": 1.5184404850006104,
+      "learning_rate": 0.00013901773887749008,
+      "loss": 1.5037,
+      "step": 12250
+    },
+    {
+      "epoch": 0.942235212675948,
+      "eval_loss": 1.5220181941986084,
+      "eval_runtime": 17.7647,
+      "eval_samples_per_second": 56.291,
+      "eval_steps_per_second": 14.073,
+      "step": 12250
+    },
+    {
+      "epoch": 0.9460810706868702,
+      "grad_norm": 1.3881593942642212,
+      "learning_rate": 0.0001387580188556736,
+      "loss": 1.4952,
+      "step": 12300
+    },
+    {
+      "epoch": 0.9499269286977925,
+      "grad_norm": 2.319173574447632,
+      "learning_rate": 0.0001384982988338571,
+      "loss": 1.4667,
+      "step": 12350
+    },
+    {
+      "epoch": 0.9537727867087147,
+      "grad_norm": 1.9042879343032837,
+      "learning_rate": 0.00013823857881204064,
+      "loss": 1.6038,
+      "step": 12400
+    },
+    {
+      "epoch": 0.957618644719637,
+      "grad_norm": 1.9162698984146118,
+      "learning_rate": 0.00013797885879022415,
+      "loss": 1.4904,
+      "step": 12450
+    },
+    {
+      "epoch": 0.9614645027305592,
+      "grad_norm": 2.0601863861083984,
+      "learning_rate": 0.00013771913876840767,
+      "loss": 1.4758,
+      "step": 12500
+    },
+    {
+      "epoch": 0.9614645027305592,
+      "eval_loss": 1.5356587171554565,
+      "eval_runtime": 17.7404,
+      "eval_samples_per_second": 56.369,
+      "eval_steps_per_second": 14.092,
+      "step": 12500
     }
   ],
   "logging_steps": 50,