Training in progress, step 38500, checkpoint

last-checkpoint/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6fd57c393f47a10fedc0eeb53eea073bec9f5d74a93432c74ef773c66fc48524
+oid sha256:4f4e59779ae5268c858ada9fe05772a37f9613020d8b9f7340783b2a16c0e383
 size 36730224

last-checkpoint/ar_diffusion_info.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9bd9ca616ae823c9a3f8ca31216995734a05d8ab02efa11d199d90e7767f414c
+oid sha256:376da5b0e7144c1b91117f5526417311ed45b536f61dd6ac0ef8b19051f1dc33
 size 1736

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7db25f3289aeb2b9363ef193bde6f97fc98f831d0d2faf2aba478e02c7461eb7
+oid sha256:9de512a7acdea4434e0675f44f1896eab7a4c3fe6cc967e9cd2360783b35758f
 size 73588346

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6fe094da8ad1e1bf555c74dca2c7341b574c5a3cf7d8c3a98ff15eda79c695fe
+oid sha256:c8c00195221e523ee69e83979891848166b78d5b2c3da6d54545d0f1c8a050c4
 size 14244

last-checkpoint/scaler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:da55266e7e44775541aba7dfea6ddb4ce98b63d56ca8b7cf213a395f32135776
+oid sha256:274528fb9458ad54ab5973b2752fc7f6c9483a2a721e4d0ba19183ca440bc21c
 size 988

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:82bb4fe5bf3e8542933d8cb52c350875b68d38b60b96f875a9ab6b9538329c08
+oid sha256:9bcb038f6702a5ed9cf6a7787eb82f2807063a89d7bbe28d73008f47570ae171
 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": 2.9228520883008997,
+  "epoch": 2.961310668410122,
   "eval_steps": 250,
-  "global_step": 38000,
+  "global_step": 38500,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -6544,6 +6544,92 @@
       "eval_samples_per_second": 56.128,
       "eval_steps_per_second": 14.032,
       "step": 38000
+    },
+    {
+      "epoch": 2.926697946311822,
+      "grad_norm": 0.8798107504844666,
+      "learning_rate": 5.0229852219307584e-06,
+      "loss": 0.7084,
+      "step": 38050
+    },
+    {
+      "epoch": 2.9305438043227445,
+      "grad_norm": 0.3845706284046173,
+      "learning_rate": 4.763265200114277e-06,
+      "loss": 0.6764,
+      "step": 38100
+    },
+    {
+      "epoch": 2.9343896623336665,
+      "grad_norm": 0.6847463846206665,
+      "learning_rate": 4.503545178297795e-06,
+      "loss": 0.7165,
+      "step": 38150
+    },
+    {
+      "epoch": 2.938235520344589,
+      "grad_norm": 0.7632951736450195,
+      "learning_rate": 4.243825156481313e-06,
+      "loss": 0.7311,
+      "step": 38200
+    },
+    {
+      "epoch": 2.9420813783555113,
+      "grad_norm": 1.3314287662506104,
+      "learning_rate": 3.984105134664832e-06,
+      "loss": 0.6852,
+      "step": 38250
+    },
+    {
+      "epoch": 2.9420813783555113,
+      "eval_loss": 0.7683274149894714,
+      "eval_runtime": 17.9706,
+      "eval_samples_per_second": 55.646,
+      "eval_steps_per_second": 13.912,
+      "step": 38250
+    },
+    {
+      "epoch": 2.9459272363664333,
+      "grad_norm": 1.0179448127746582,
+      "learning_rate": 3.7243851128483497e-06,
+      "loss": 0.759,
+      "step": 38300
+    },
+    {
+      "epoch": 2.9497730943773557,
+      "grad_norm": 1.4116487503051758,
+      "learning_rate": 3.4646650910318677e-06,
+      "loss": 0.7773,
+      "step": 38350
+    },
+    {
+      "epoch": 2.953618952388278,
+      "grad_norm": 0.6251114010810852,
+      "learning_rate": 3.2049450692153856e-06,
+      "loss": 0.7016,
+      "step": 38400
+    },
+    {
+      "epoch": 2.9574648103992,
+      "grad_norm": 0.9810579419136047,
+      "learning_rate": 2.945225047398904e-06,
+      "loss": 0.6909,
+      "step": 38450
+    },
+    {
+      "epoch": 2.961310668410122,
+      "grad_norm": 0.7243860363960266,
+      "learning_rate": 2.6855050255824223e-06,
+      "loss": 0.7305,
+      "step": 38500
+    },
+    {
+      "epoch": 2.961310668410122,
+      "eval_loss": 0.7680486440658569,
+      "eval_runtime": 17.9145,
+      "eval_samples_per_second": 55.821,
+      "eval_steps_per_second": 13.955,
+      "step": 38500
     }
   ],
   "logging_steps": 50,