Training in progress, step 600, checkpoint

last-checkpoint/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b4fb84457a91c5de153894c7ca0e8bf12972e85157cdf5d915e11430fc00bf59
+oid sha256:4711c604191c70b1e9a248ff89ea22db5362690aa8037aa62c5c84d06fc55f81
 size 8297288

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:46556fd5ff967ca9fb9f57dd94cffedd6630e12727c0a2e7a94c9a92581e7161
+oid sha256:663863afca4c95a45977985d1fcd4fa2b26b80e962ef1e6075fe7371d1abb417
 size 16761163

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8147cce07a8085109be600ca968db3999ae311b31412161d0086e0673fa99dee
+oid sha256:7d6eb5b1a4961028fe68de934a271b3920cdc97935fe1a1ad9fd0f05662957a2
 size 14645

last-checkpoint/scaler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a5c6fc74fad2fa82ce5e6158544172660f4d56e9b794ecdcdf74960d2ac0a30e
+oid sha256:b7092e272a6bfc2c815a87505efd3c62d8b3b8462044807a20f9427f7f599c43
 size 1383

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:219ab60ad4ef7e10902c21765a32ad47d5cfe315598bf972b3280f68906cb5f2
+oid sha256:f810a5865983beee2c8a2c1261e8005dbb8b6e24c20d69bc4958ca74ac736911
 size 1465

last-checkpoint/trainer_state.json

@@ -2,9 +2,9 @@
   "best_global_step": 200,
   "best_metric": 0.6018396615982056,
   "best_model_checkpoint": "./t5gemma-math-corrector/checkpoint-200",
-  "epoch": 1.2822774659182037,
+  "epoch": 1.9238171611868484,
   "eval_steps": 200,
-  "global_step": 400,
+  "global_step": 600,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -80,6 +80,42 @@
       "eval_samples_per_second": 50.711,
       "eval_steps_per_second": 12.775,
       "step": 400
+    },
+    {
+      "epoch": 1.4426623897353648,
+      "grad_norm": 0.35608166456222534,
+      "learning_rate": 5.783847980997625e-05,
+      "loss": 0.0015,
+      "step": 450
+    },
+    {
+      "epoch": 1.603047313552526,
+      "grad_norm": 0.33180880546569824,
+      "learning_rate": 5.190023752969121e-05,
+      "loss": 0.0015,
+      "step": 500
+    },
+    {
+      "epoch": 1.7634322373696873,
+      "grad_norm": 0.16627037525177002,
+      "learning_rate": 4.596199524940617e-05,
+      "loss": 0.0009,
+      "step": 550
+    },
+    {
+      "epoch": 1.9238171611868484,
+      "grad_norm": 0.04953346028923988,
+      "learning_rate": 4.002375296912114e-05,
+      "loss": 0.001,
+      "step": 600
+    },
+    {
+      "epoch": 1.9238171611868484,
+      "eval_loss": 3.8351235389709473,
+      "eval_runtime": 5.1639,
+      "eval_samples_per_second": 50.737,
+      "eval_steps_per_second": 12.781,
+      "step": 600
     }
   ],
   "logging_steps": 50,
@@ -94,7 +130,7 @@
         "early_stopping_threshold": 0.0
       },
       "attributes": {
-        "early_stopping_patience_counter": 1
+        "early_stopping_patience_counter": 2
       }
     },
     "TrainerControl": {
@@ -108,7 +144,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 1545010211192832.0,
+  "total_flos": 2318240219922432.0,
   "train_batch_size": 4,
   "trial_name": null,
   "trial_params": null