Training in progress, step 30000, checkpoint

last-checkpoint/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:79c28d4b4840c8a771d27b1f27c6ac5c00a9f2991d7394f3a6733892a09080ab
+oid sha256:069e8aec2bc91b01e99f7b79f39b1217a37ea88d9f6f4132f8acee4f393c01b1
 size 36730224

last-checkpoint/ar_diffusion_info.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c30265894652137003ce0d462067739a1e1f87e1fef05727cce7c9bc27d9969d
+oid sha256:05de4cdebfe1224dfd9ad2f7aac9780ebf9102b682b1460f83379a4262dee74b
 size 1544

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:35ee565d832de94c600e576ee117f0978437f0823600766c050ee216a94c8988
+oid sha256:585e32212caa9f4214a347406425426d068333176e31042b8c5c439f5b429d00
 size 73588346

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:734e3f7dc3063a42733f6397bd808730560212803e7502d677f32c1adc811a33
+oid sha256:65c2f44365d34d95613ad949972f890d6a6afc2c6e9d4d4d8c47a473b1c2b039
 size 14244

last-checkpoint/scaler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a4379a058e70e2b699ffe8db7ee723ac96ffd7adc9d62248596ac9aebf5e54f3
+oid sha256:195e35eaf32578f671662018e8b669b6388d8b5ac81f855c1c95157d655e6122
 size 988

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7f444f08db434c1664e05b94a04b805dac6e17ed393cf9fcf28a82949e586c98
+oid sha256:7f8616cef5b8dab506166471a2c46c7770aee1e224f56ecc6b4d1b9b5b17b94e
 size 1064

last-checkpoint/trainer_state.json

@@ -1,10 +1,10 @@
 {
-  "best_global_step": 29000,
-  "best_metric": 0.9999537467956543,
-  "best_model_checkpoint": "./ar-diffusion-checkpoints-fixed/checkpoint-29000",
-  "epoch": 2.26905622644412,
+  "best_global_step": 30000,
+  "best_metric": 0.9945911169052124,
+  "best_model_checkpoint": "./ar-diffusion-checkpoints-fixed/checkpoint-30000",
+  "epoch": 2.307514806553342,
   "eval_steps": 250,
-  "global_step": 29500,
+  "global_step": 30000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -5082,6 +5082,92 @@
       "eval_samples_per_second": 57.375,
       "eval_steps_per_second": 14.344,
       "step": 29500
+    },
+    {
+      "epoch": 2.272902084455042,
+      "grad_norm": 0.6814424991607666,
+      "learning_rate": 4.9175388930732675e-05,
+      "loss": 0.9563,
+      "step": 29550
+    },
+    {
+      "epoch": 2.2767479424659642,
+      "grad_norm": 0.85086989402771,
+      "learning_rate": 4.891566890891619e-05,
+      "loss": 0.9778,
+      "step": 29600
+    },
+    {
+      "epoch": 2.280593800476886,
+      "grad_norm": 1.1364527940750122,
+      "learning_rate": 4.8655948887099714e-05,
+      "loss": 0.996,
+      "step": 29650
+    },
+    {
+      "epoch": 2.2844396584878086,
+      "grad_norm": 0.7810873985290527,
+      "learning_rate": 4.839622886528323e-05,
+      "loss": 1.0141,
+      "step": 29700
+    },
+    {
+      "epoch": 2.288285516498731,
+      "grad_norm": 1.02638578414917,
+      "learning_rate": 4.8136508843466746e-05,
+      "loss": 0.97,
+      "step": 29750
+    },
+    {
+      "epoch": 2.288285516498731,
+      "eval_loss": 1.0048705339431763,
+      "eval_runtime": 17.487,
+      "eval_samples_per_second": 57.185,
+      "eval_steps_per_second": 14.296,
+      "step": 29750
+    },
+    {
+      "epoch": 2.292131374509653,
+      "grad_norm": 1.334876298904419,
+      "learning_rate": 4.7876788821650266e-05,
+      "loss": 0.9533,
+      "step": 29800
+    },
+    {
+      "epoch": 2.2959772325205754,
+      "grad_norm": 1.3690305948257446,
+      "learning_rate": 4.761706879983378e-05,
+      "loss": 0.992,
+      "step": 29850
+    },
+    {
+      "epoch": 2.299823090531498,
+      "grad_norm": 1.525981068611145,
+      "learning_rate": 4.7357348778017305e-05,
+      "loss": 0.949,
+      "step": 29900
+    },
+    {
+      "epoch": 2.3036689485424198,
+      "grad_norm": 0.6612236499786377,
+      "learning_rate": 4.709762875620082e-05,
+      "loss": 1.0112,
+      "step": 29950
+    },
+    {
+      "epoch": 2.307514806553342,
+      "grad_norm": 0.8452871441841125,
+      "learning_rate": 4.683790873438433e-05,
+      "loss": 0.9858,
+      "step": 30000
+    },
+    {
+      "epoch": 2.307514806553342,
+      "eval_loss": 0.9945911169052124,
+      "eval_runtime": 17.2501,
+      "eval_samples_per_second": 57.971,
+      "eval_steps_per_second": 14.493,
+      "step": 30000
     }
   ],
   "logging_steps": 50,