Training in progress, step 36000, checkpoint

last-checkpoint/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bef738804bc450055544ac26cb2862c97449ead2f75ac228ddc61442fe9bb8bd
+oid sha256:23e42702caff4df8b7ad44cb873bb5dddff2f795cef7082f27a461514c69024c
 size 36730224

last-checkpoint/ar_diffusion_info.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:26406f1b76813f85bc53aaeae8ac510fb6de1504e785a5a45daa6f7e5a13e4e0
+oid sha256:fea47853b8aa60e44a0bb0420e1e620de2e8802a7f7c2912ecff431bb1bb5f8c
 size 1736

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:775b4d9c34d0a3506dcf86d076ddf2b91371c6a7d165bc685f42c9caa7b6efb5
+oid sha256:e462a352c1c617a33fcc383bd32232ac83347aec5662f1905faf9562932eadb9
 size 73588346

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:817eeb968a3fe33ed8e09b44d359afc61c95d374fc7bc36cec59f18e26bfa49c
+oid sha256:77a7de3731de8cf2e7331a42359007d95d6eaa607441af485deea1e6b851ef50
 size 14244

last-checkpoint/scaler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1dffe29e5b55d433cbe32c72434ab6b543d59b2b4f5148609cfe9a7f61d0b7d1
+oid sha256:9d348b949dd5c914964463518846b78fc436944dc5c12ed7b8077ee0d1b60d7d
 size 988

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0702d25847c4b82c912557ae056afc22dda26278838471e438e37fa978161ebf
+oid sha256:27724e8d1fba2d1802daad944f1058156037a888933fba524cc8f9009abca3d9
 size 1064

last-checkpoint/trainer_state.json

@@ -2,9 +2,9 @@
   "best_global_step": 31000,
   "best_metric": 0.7226839661598206,
   "best_model_checkpoint": "./ar-diffusion-checkpoints-fixed/checkpoint-31000",
-  "epoch": 2.730559187754788,
+  "epoch": 2.7690177678640104,
   "eval_steps": 250,
-  "global_step": 35500,
+  "global_step": 36000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -6114,6 +6114,92 @@
       "eval_samples_per_second": 55.467,
       "eval_steps_per_second": 13.867,
       "step": 35500
+    },
+    {
+      "epoch": 2.73440504576571,
+      "grad_norm": 0.5093711018562317,
+      "learning_rate": 1.800379191231852e-05,
+      "loss": 0.6425,
+      "step": 35550
+    },
+    {
+      "epoch": 2.7382509037766325,
+      "grad_norm": 1.0487879514694214,
+      "learning_rate": 1.774407189050204e-05,
+      "loss": 0.6846,
+      "step": 35600
+    },
+    {
+      "epoch": 2.742096761787555,
+      "grad_norm": 0.6705742478370667,
+      "learning_rate": 1.748435186868556e-05,
+      "loss": 0.7464,
+      "step": 35650
+    },
+    {
+      "epoch": 2.745942619798477,
+      "grad_norm": 0.43706628680229187,
+      "learning_rate": 1.7224631846869076e-05,
+      "loss": 0.7325,
+      "step": 35700
+    },
+    {
+      "epoch": 2.7497884778093993,
+      "grad_norm": 1.1549192667007446,
+      "learning_rate": 1.6964911825052592e-05,
+      "loss": 0.7083,
+      "step": 35750
+    },
+    {
+      "epoch": 2.7497884778093993,
+      "eval_loss": 0.7717772126197815,
+      "eval_runtime": 18.085,
+      "eval_samples_per_second": 55.295,
+      "eval_steps_per_second": 13.824,
+      "step": 35750
+    },
+    {
+      "epoch": 2.7536343358203217,
+      "grad_norm": 0.5367007255554199,
+      "learning_rate": 1.6705191803236112e-05,
+      "loss": 0.7054,
+      "step": 35800
+    },
+    {
+      "epoch": 2.7574801938312437,
+      "grad_norm": 0.8213953971862793,
+      "learning_rate": 1.644547178141963e-05,
+      "loss": 0.6764,
+      "step": 35850
+    },
+    {
+      "epoch": 2.761326051842166,
+      "grad_norm": 0.9012633562088013,
+      "learning_rate": 1.6185751759603148e-05,
+      "loss": 0.725,
+      "step": 35900
+    },
+    {
+      "epoch": 2.7651719098530885,
+      "grad_norm": 0.656104326248169,
+      "learning_rate": 1.5926031737786667e-05,
+      "loss": 0.7279,
+      "step": 35950
+    },
+    {
+      "epoch": 2.7690177678640104,
+      "grad_norm": 0.901136040687561,
+      "learning_rate": 1.5666311715970187e-05,
+      "loss": 0.7354,
+      "step": 36000
+    },
+    {
+      "epoch": 2.7690177678640104,
+      "eval_loss": 0.7717016935348511,
+      "eval_runtime": 17.9512,
+      "eval_samples_per_second": 55.707,
+      "eval_steps_per_second": 13.927,
+      "step": 36000
     }
   ],
   "logging_steps": 50,