Training in progress, step 22500, checkpoint

last-checkpoint/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:300bf30417bbf3f32527e2186c9abed7ed76183b832bf2d5e14eac86f900b976
+oid sha256:54d37e3b015e65fe77b6f201cc314fc109253be775a5cb9df138ff0aaf02909f
 size 36730224

last-checkpoint/ar_diffusion_info.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:63292d3528a5a0926c722da43abe6585432ca093708fd76551d8031e7ddd3643
+oid sha256:4d1cecc0bb865c095b0d5ed67e5d69becfda56b3732d5f6b97c50000f421baec
 size 1736

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bd93d8bd898fd17bb0bb48e446465b209d65cfac9f90818590fb648e18b8ab7a
+oid sha256:f46eb44e04d7e023aa7c0a1a250eb53199183d3cb6008b7bfbeae0fb8e741141
 size 73588346

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ec8bd4a1e6303e385dc1e7236df83c38e8a5367849e5bc57e5960781558c1b9e
-size 14244
+oid sha256:166bfc987d23199e193830fa6a19bb0011d9a2f67bb503cefc88d16d6bfa7032
+size 14180

last-checkpoint/scaler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f1c27ce326198bdf77c595dcbaf5251be4cae7f085aadb1446913e62d977b694
+oid sha256:6ce1ae1d284068fef7aa297321b14237db7c803a93cd953ac357560946400a64
 size 988

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:aef577ff6c2918592ffb94dbf847593611b785b48ba36e1fc639ca9ece0452f2
+oid sha256:2ffd90226399435ca3354a71ecdfe070bf4ff007ddc73740c3fd34e13b676fb2
 size 1064

last-checkpoint/trainer_state.json

@@ -1,10 +1,10 @@
 {
-  "best_global_step": 21750,
-  "best_metric": 0.7681905031204224,
-  "best_model_checkpoint": "./ar-diffusion-checkpoints-fixed/checkpoint-21500",
-  "epoch": 1.6921775248057842,
+  "best_global_step": 22500,
+  "best_metric": 0.7646552920341492,
+  "best_model_checkpoint": "./ar-diffusion-checkpoints-fixed/checkpoint-22500",
+  "epoch": 1.7306361049150065,
   "eval_steps": 250,
-  "global_step": 22000,
+  "global_step": 22500,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -3792,6 +3792,92 @@
       "eval_samples_per_second": 56.134,
       "eval_steps_per_second": 14.034,
       "step": 22000
+    },
+    {
+      "epoch": 1.6960233828167064,
+      "grad_norm": 0.5081881880760193,
+      "learning_rate": 8.810742020102331e-05,
+      "loss": 0.7883,
+      "step": 22050
+    },
+    {
+      "epoch": 1.6998692408276286,
+      "grad_norm": 0.5658268332481384,
+      "learning_rate": 8.784770017920681e-05,
+      "loss": 0.77,
+      "step": 22100
+    },
+    {
+      "epoch": 1.703715098838551,
+      "grad_norm": 0.6888287663459778,
+      "learning_rate": 8.758798015739034e-05,
+      "loss": 0.8013,
+      "step": 22150
+    },
+    {
+      "epoch": 1.707560956849473,
+      "grad_norm": 1.0599181652069092,
+      "learning_rate": 8.732826013557386e-05,
+      "loss": 0.7649,
+      "step": 22200
+    },
+    {
+      "epoch": 1.7114068148603954,
+      "grad_norm": 0.7679368257522583,
+      "learning_rate": 8.706854011375736e-05,
+      "loss": 0.7338,
+      "step": 22250
+    },
+    {
+      "epoch": 1.7114068148603954,
+      "eval_loss": 0.7665286660194397,
+      "eval_runtime": 17.8822,
+      "eval_samples_per_second": 55.921,
+      "eval_steps_per_second": 13.98,
+      "step": 22250
+    },
+    {
+      "epoch": 1.7152526728713176,
+      "grad_norm": 0.6636808514595032,
+      "learning_rate": 8.680882009194089e-05,
+      "loss": 0.7949,
+      "step": 22300
+    },
+    {
+      "epoch": 1.7190985308822397,
+      "grad_norm": 0.7327821254730225,
+      "learning_rate": 8.654910007012442e-05,
+      "loss": 0.7484,
+      "step": 22350
+    },
+    {
+      "epoch": 1.7229443888931621,
+      "grad_norm": 0.8187472224235535,
+      "learning_rate": 8.628938004830793e-05,
+      "loss": 0.7482,
+      "step": 22400
+    },
+    {
+      "epoch": 1.7267902469040843,
+      "grad_norm": 0.4527030289173126,
+      "learning_rate": 8.602966002649144e-05,
+      "loss": 0.7902,
+      "step": 22450
+    },
+    {
+      "epoch": 1.7306361049150065,
+      "grad_norm": 0.6475220918655396,
+      "learning_rate": 8.576994000467497e-05,
+      "loss": 0.7998,
+      "step": 22500
+    },
+    {
+      "epoch": 1.7306361049150065,
+      "eval_loss": 0.7646552920341492,
+      "eval_runtime": 17.8108,
+      "eval_samples_per_second": 56.146,
+      "eval_steps_per_second": 14.036,
+      "step": 22500
     }
   ],
   "logging_steps": 50,