Training in progress, step 23000, checkpoint

last-checkpoint/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:54d37e3b015e65fe77b6f201cc314fc109253be775a5cb9df138ff0aaf02909f
+oid sha256:607761a68d936fac3569465d3f0d20419715827f9c4331a285faee67e09f155c
 size 36730224

last-checkpoint/ar_diffusion_info.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4d1cecc0bb865c095b0d5ed67e5d69becfda56b3732d5f6b97c50000f421baec
+oid sha256:80833488aa86fb4cba830c432f5aa3abb1b8d8ff38f3b1e8b83ce7834612b917
 size 1736

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f46eb44e04d7e023aa7c0a1a250eb53199183d3cb6008b7bfbeae0fb8e741141
+oid sha256:b4035be8d85b44dc1d28bbbd28b0c42e51f7f77114f423b7a9f9fd7587aca606
 size 73588346

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:166bfc987d23199e193830fa6a19bb0011d9a2f67bb503cefc88d16d6bfa7032
-size 14180
+oid sha256:ce24974df420c8b58ce59acb329c359d968e0b98672cd1798e7bedf4f10e13c3
+size 14244

last-checkpoint/scaler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6ce1ae1d284068fef7aa297321b14237db7c803a93cd953ac357560946400a64
+oid sha256:36d8e43084f8796853eb654313d4c1bc1085274ed04c979e93b6fe0324b7620f
 size 988

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2ffd90226399435ca3354a71ecdfe070bf4ff007ddc73740c3fd34e13b676fb2
+oid sha256:e50d97e6b844f88427c38d3f0fdd176e281de54a4fb5c4780fea01f20111dfc1
 size 1064

last-checkpoint/trainer_state.json

@@ -1,10 +1,10 @@
 {
-  "best_global_step": 22500,
-  "best_metric": 0.7646552920341492,
-  "best_model_checkpoint": "./ar-diffusion-checkpoints-fixed/checkpoint-22500",
-  "epoch": 1.7306361049150065,
+  "best_global_step": 23000,
+  "best_metric": 0.7635026574134827,
+  "best_model_checkpoint": "./ar-diffusion-checkpoints-fixed/checkpoint-23000",
+  "epoch": 1.7690946850242288,
   "eval_steps": 250,
-  "global_step": 22500,
+  "global_step": 23000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -3878,6 +3878,92 @@
       "eval_samples_per_second": 56.146,
       "eval_steps_per_second": 14.036,
       "step": 22500
+    },
+    {
+      "epoch": 1.734481962925929,
+      "grad_norm": 0.538769543170929,
+      "learning_rate": 8.551021998285848e-05,
+      "loss": 0.7658,
+      "step": 22550
+    },
+    {
+      "epoch": 1.738327820936851,
+      "grad_norm": 0.629510223865509,
+      "learning_rate": 8.5250499961042e-05,
+      "loss": 0.7885,
+      "step": 22600
+    },
+    {
+      "epoch": 1.7421736789477733,
+      "grad_norm": 0.6914022564888,
+      "learning_rate": 8.499077993922552e-05,
+      "loss": 0.7648,
+      "step": 22650
+    },
+    {
+      "epoch": 1.7460195369586955,
+      "grad_norm": 0.5563036799430847,
+      "learning_rate": 8.473105991740903e-05,
+      "loss": 0.7558,
+      "step": 22700
+    },
+    {
+      "epoch": 1.7498653949696177,
+      "grad_norm": 0.7851826548576355,
+      "learning_rate": 8.447133989559256e-05,
+      "loss": 0.7961,
+      "step": 22750
+    },
+    {
+      "epoch": 1.7498653949696177,
+      "eval_loss": 0.7664644718170166,
+      "eval_runtime": 17.9733,
+      "eval_samples_per_second": 55.638,
+      "eval_steps_per_second": 13.91,
+      "step": 22750
+    },
+    {
+      "epoch": 1.75371125298054,
+      "grad_norm": 0.48695698380470276,
+      "learning_rate": 8.421161987377607e-05,
+      "loss": 0.8264,
+      "step": 22800
+    },
+    {
+      "epoch": 1.757557110991462,
+      "grad_norm": 0.8053486347198486,
+      "learning_rate": 8.39518998519596e-05,
+      "loss": 0.8084,
+      "step": 22850
+    },
+    {
+      "epoch": 1.7614029690023845,
+      "grad_norm": 1.1373741626739502,
+      "learning_rate": 8.369217983014311e-05,
+      "loss": 0.7842,
+      "step": 22900
+    },
+    {
+      "epoch": 1.7652488270133067,
+      "grad_norm": 1.1318634748458862,
+      "learning_rate": 8.343245980832662e-05,
+      "loss": 0.7727,
+      "step": 22950
+    },
+    {
+      "epoch": 1.7690946850242288,
+      "grad_norm": 0.8140521049499512,
+      "learning_rate": 8.317273978651015e-05,
+      "loss": 0.7945,
+      "step": 23000
+    },
+    {
+      "epoch": 1.7690946850242288,
+      "eval_loss": 0.7635026574134827,
+      "eval_runtime": 17.8969,
+      "eval_samples_per_second": 55.875,
+      "eval_steps_per_second": 13.969,
+      "step": 23000
     }
   ],
   "logging_steps": 50,