Training in progress, step 21500, checkpoint

last-checkpoint/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d257112e537fe70c79019496ce8a965aa34d1fb175b35addc498d01e38be6234
+oid sha256:66cda148dcae03c8b6ee9a988c41356b8fb7dd85bc794057def1a0a106239b14
 size 36730224

last-checkpoint/ar_diffusion_info.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:25ff7761b7be0ad2ef125a7c44f3a00d691af1455b26167da65e53a9f3d4a4e1
+oid sha256:7a7ccadc0fc62d6e734efb6eac59f8de15c8962f84260b3fa84659524029cca9
 size 1544

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f34cf1c465a3f22b58ed5814e553f1f151f9d6086122d8e635e6c9a1efd7a02c
+oid sha256:4cde556e6d84497b700bedd313afb1db4fbdff84a922d3df94365577a9ffe5cb
 size 73588346

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4c5132bfe30a63a6382081eeb67e311e07e2d98916afbc38ab8f8ba0c010b059
-size 14244
+oid sha256:24cdd6f45f9ebce887e06d548bf949bd18ccf52b17999d18752823b6dd7dc03a
+size 14308

last-checkpoint/scaler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:279d58c28b18870f94d45189b49d69a8021c82d9a61e08eb069f8a47562cd494
+oid sha256:76d7482aa4f2ea8d9ee623a2338b27a160c64d9b3e6847f775500f74a89d839e
 size 988

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:12146cd21738b806f2d3b5e8ca727dc812aefa4523582c6a525173e0b62cdd2a
+oid sha256:ba5e991474e4af796236702d23dc20d105ef404660d6c7d1c95fb199aa1ce3be
 size 1064

last-checkpoint/trainer_state.json

@@ -2,9 +2,9 @@
   "best_global_step": 21000,
   "best_metric": 1.4690666198730469,
   "best_model_checkpoint": "./ar-diffusion-checkpoints-fixed/checkpoint-21000",
-  "epoch": 1.6152603645873396,
+  "epoch": 1.6537189446965619,
   "eval_steps": 250,
-  "global_step": 21000,
+  "global_step": 21500,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -3620,6 +3620,92 @@
       "eval_samples_per_second": 56.318,
       "eval_steps_per_second": 14.079,
       "step": 21000
+    },
+    {
+      "epoch": 1.6191062225982615,
+      "grad_norm": 1.8204776048660278,
+      "learning_rate": 9.33122094382256e-05,
+      "loss": 1.4423,
+      "step": 21050
+    },
+    {
+      "epoch": 1.622952080609184,
+      "grad_norm": 1.6150448322296143,
+      "learning_rate": 9.305248941640912e-05,
+      "loss": 1.4368,
+      "step": 21100
+    },
+    {
+      "epoch": 1.6267979386201061,
+      "grad_norm": 2.4393765926361084,
+      "learning_rate": 9.279276939459264e-05,
+      "loss": 1.6115,
+      "step": 21150
+    },
+    {
+      "epoch": 1.6306437966310283,
+      "grad_norm": 1.217399001121521,
+      "learning_rate": 9.253304937277615e-05,
+      "loss": 1.5792,
+      "step": 21200
+    },
+    {
+      "epoch": 1.6344896546419507,
+      "grad_norm": 1.4467514753341675,
+      "learning_rate": 9.227332935095967e-05,
+      "loss": 1.4838,
+      "step": 21250
+    },
+    {
+      "epoch": 1.6344896546419507,
+      "eval_loss": 1.4741238355636597,
+      "eval_runtime": 17.7656,
+      "eval_samples_per_second": 56.289,
+      "eval_steps_per_second": 14.072,
+      "step": 21250
+    },
+    {
+      "epoch": 1.6383355126528727,
+      "grad_norm": 1.4274511337280273,
+      "learning_rate": 9.201360932914319e-05,
+      "loss": 1.4747,
+      "step": 21300
+    },
+    {
+      "epoch": 1.642181370663795,
+      "grad_norm": 1.5080102682113647,
+      "learning_rate": 9.175908370776304e-05,
+      "loss": 1.4262,
+      "step": 21350
+    },
+    {
+      "epoch": 1.6460272286747173,
+      "grad_norm": 1.2717032432556152,
+      "learning_rate": 9.149936368594655e-05,
+      "loss": 1.4418,
+      "step": 21400
+    },
+    {
+      "epoch": 1.6498730866856395,
+      "grad_norm": 1.3594080209732056,
+      "learning_rate": 9.123964366413006e-05,
+      "loss": 1.5184,
+      "step": 21450
+    },
+    {
+      "epoch": 1.6537189446965619,
+      "grad_norm": 1.896607518196106,
+      "learning_rate": 9.097992364231359e-05,
+      "loss": 1.4702,
+      "step": 21500
+    },
+    {
+      "epoch": 1.6537189446965619,
+      "eval_loss": 1.4777735471725464,
+      "eval_runtime": 17.8161,
+      "eval_samples_per_second": 56.129,
+      "eval_steps_per_second": 14.032,
+      "step": 21500
     }
   ],
   "logging_steps": 50,