Training in progress, step 4456

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7fdb2b53096835a5a02597fbcbbb8b279b8398b05f1bdb4f9983b9ef5737b1fe
+oid sha256:0aee9bb6a4fc45e1085dd9f86882bba6ab07cd267dfaad02becaf06d5ab356a2
 size 558937304

runs/Apr27_12-02-54_ip-10-192-12-76/events.out.tfevents.1745755375.ip-10-192-12-76.2375.0

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9f2dce00904eb8a9eb57a26917dfd445c1295c0016e266de1444f683965cf0bb
-size 6679
+oid sha256:198d9b04fa06982fbcaa2b4813d3386c10829b172b4fc681a18ce6d200af4a80
+size 7419

trainer_state.json

@@ -1,10 +1,10 @@
 {
-  "best_global_step": 3342,
-  "best_metric": 0.7720533296293449,
-  "best_model_checkpoint": "./checkpoints/checkpoint-3342",
-  "epoch": 0.5998115493336923,
+  "best_global_step": 4456,
+  "best_metric": 0.7718523934618883,
+  "best_model_checkpoint": "./checkpoints/checkpoint-4456",
+  "epoch": 0.7997487324449231,
   "eval_steps": 1114,
-  "global_step": 3342,
+  "global_step": 4456,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -77,6 +77,29 @@
       "eval_samples_per_second": 403.306,
       "eval_steps_per_second": 1.052,
       "step": 3342
+    },
+    {
+      "epoch": 0.6281688876923768,
+      "grad_norm": 0.03279432654380798,
+      "learning_rate": 4.384876624004233e-05,
+      "loss": 0.03,
+      "step": 3500
+    },
+    {
+      "epoch": 0.7179073002198592,
+      "grad_norm": 0.032072387635707855,
+      "learning_rate": 3.8534883130104934e-05,
+      "loss": 0.027,
+      "step": 4000
+    },
+    {
+      "epoch": 0.7997487324449231,
+      "eval_cer": 0.7718523934618883,
+      "eval_loss": 0.01841309294104576,
+      "eval_runtime": 436.0956,
+      "eval_samples_per_second": 406.037,
+      "eval_steps_per_second": 1.059,
+      "step": 4456
     }
   ],
   "logging_steps": 500,
@@ -96,7 +119,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 3.912580457425797e+17,
+  "total_flos": 5.216773943234396e+17,
   "train_batch_size": 384,
   "trial_name": null,
   "trial_params": null