Training in progress, step 31000, checkpoint

last-checkpoint/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fdc0ad979a12ba1d5693d6f619dbaea0818f22e8b6c5ce88c754da865d18dae9
+oid sha256:6d9e8aa8332991d75c5a3a1f123c6c1198022025d03184357acc577c0b39acba
 size 36730224

last-checkpoint/ar_diffusion_info.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:87d705db2757234d5d8a502c3a722562455d35ceffb96acc7f8a84d9a82b57f2
+oid sha256:ea977aa1785a55c5866b361d7b21a5d22c06320a36b66e23cae58bae6d91afb7
 size 1736

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a17843cbe8069abb8746e8e871bb3e80d6841c1b2726ca6959bb29928e72639f
+oid sha256:379db4eacc0e435dac4e4d04a79893844ebcb7fced0901f3fcb2f48d84f0e313
 size 73588346

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5baf28c8bc0abeb08100e2c5cfe54b52c8077e77b1cb64b5b332c05fac995071
+oid sha256:f794cabb9716498ba0a6fd8bea2bd2890dd20581f9c98da6183a32bb2fe075cf
 size 14244

last-checkpoint/scaler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8eef9d19aff91ed1822b958b8f57eac8d029c3900de24c8d1baed5c3a38e5ec6
+oid sha256:129104f0cb747d88bf8f676e9bc885ddb51125ec85e582f611fd83482b46f3a0
 size 988

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6b26116b7f3dd831125aeebdfb3d11e2bdb55abc7c8e404210b0ff9bfd497ccc
+oid sha256:65139cd8a1203ebfa96d6ae8d43504141c2bb01394aebd96928840758be6fa65
 size 1064

last-checkpoint/trainer_state.json

@@ -1,10 +1,10 @@
 {
-  "best_global_step": 30500,
-  "best_metric": 0.7244414687156677,
-  "best_model_checkpoint": "./ar-diffusion-checkpoints-fixed/checkpoint-30500",
-  "epoch": 2.3459733866625645,
+  "best_global_step": 31000,
+  "best_metric": 0.7226839661598206,
+  "best_model_checkpoint": "./ar-diffusion-checkpoints-fixed/checkpoint-31000",
+  "epoch": 2.384431966771787,
   "eval_steps": 250,
-  "global_step": 30500,
+  "global_step": 31000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -5254,6 +5254,92 @@
       "eval_samples_per_second": 55.75,
       "eval_steps_per_second": 13.937,
       "step": 30500
+    },
+    {
+      "epoch": 2.3498192446734865,
+      "grad_norm": 0.9593235850334167,
+      "learning_rate": 4.3970599693530377e-05,
+      "loss": 0.7266,
+      "step": 30550
+    },
+    {
+      "epoch": 2.353665102684409,
+      "grad_norm": 0.4570913016796112,
+      "learning_rate": 4.3710879671713896e-05,
+      "loss": 0.7026,
+      "step": 30600
+    },
+    {
+      "epoch": 2.3575109606953313,
+      "grad_norm": 0.8020208477973938,
+      "learning_rate": 4.345115964989741e-05,
+      "loss": 0.7345,
+      "step": 30650
+    },
+    {
+      "epoch": 2.3613568187062532,
+      "grad_norm": 0.7795267701148987,
+      "learning_rate": 4.319143962808093e-05,
+      "loss": 0.7351,
+      "step": 30700
+    },
+    {
+      "epoch": 2.3652026767171757,
+      "grad_norm": 0.6240664720535278,
+      "learning_rate": 4.293171960626445e-05,
+      "loss": 0.7029,
+      "step": 30750
+    },
+    {
+      "epoch": 2.3652026767171757,
+      "eval_loss": 0.7251197099685669,
+      "eval_runtime": 17.8892,
+      "eval_samples_per_second": 55.9,
+      "eval_steps_per_second": 13.975,
+      "step": 30750
+    },
+    {
+      "epoch": 2.3690485347280976,
+      "grad_norm": 0.773654043674469,
+      "learning_rate": 4.267199958444797e-05,
+      "loss": 0.7105,
+      "step": 30800
+    },
+    {
+      "epoch": 2.37289439273902,
+      "grad_norm": 1.1365927457809448,
+      "learning_rate": 4.241227956263149e-05,
+      "loss": 0.7019,
+      "step": 30850
+    },
+    {
+      "epoch": 2.3767402507499424,
+      "grad_norm": 0.6990851759910583,
+      "learning_rate": 4.2152559540815e-05,
+      "loss": 0.6914,
+      "step": 30900
+    },
+    {
+      "epoch": 2.3805861087608644,
+      "grad_norm": 0.8598945140838623,
+      "learning_rate": 4.189283951899852e-05,
+      "loss": 0.7087,
+      "step": 30950
+    },
+    {
+      "epoch": 2.384431966771787,
+      "grad_norm": 0.9121548533439636,
+      "learning_rate": 4.163311949718204e-05,
+      "loss": 0.7212,
+      "step": 31000
+    },
+    {
+      "epoch": 2.384431966771787,
+      "eval_loss": 0.7226839661598206,
+      "eval_runtime": 17.8913,
+      "eval_samples_per_second": 55.893,
+      "eval_steps_per_second": 13.973,
+      "step": 31000
     }
   ],
   "logging_steps": 50,