Training in progress, step 40

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6993d2eb56920db28a25a0241d646d0cc3d87980efa46bdd0408aa766dcdcb94
+oid sha256:e0d7e37a95520f41735f4e81928cc4a409afc0daed4fe386710ac7f2ac7ad438
 size 151099494

runs/Jun29_09-39-32_bharatsahaiyak-test/events.out.tfevents.1719663371.bharatsahaiyak-test.1854617.8

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7c33d876b5bbb51ab64eb5c09b0e89150b69c5e644a90af2cafd4ecb9216df90
+size 6396

trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 29.34232715008432,
+  "best_metric": 28.246205733558178,
   "best_model_checkpoint": "./whisper-tiny-hindi2_test/checkpoint-120",
   "epoch": 6.349206349206349,
   "eval_steps": 40,
@@ -10,127 +10,127 @@
   "log_history": [
     {
       "epoch": 0.6349206349206349,
-      "grad_norm": 6.46198034286499,
+      "grad_norm": 5.915639877319336,
       "learning_rate": 3.75e-05,
-      "loss": 0.3431,
+      "loss": 0.3473,
       "step": 20
     },
     {
       "epoch": 1.2698412698412698,
-      "grad_norm": 3.6580495834350586,
+      "grad_norm": 3.9101755619049072,
       "learning_rate": 3.75e-05,
-      "loss": 0.257,
+      "loss": 0.2574,
       "step": 40
     },
     {
       "epoch": 1.2698412698412698,
-      "eval_loss": 0.19398407638072968,
-      "eval_runtime": 55.7217,
-      "eval_samples_per_second": 1.795,
-      "eval_steps_per_second": 1.795,
-      "eval_wer": 30.94435075885329,
+      "eval_loss": 0.18214347958564758,
+      "eval_runtime": 56.4889,
+      "eval_samples_per_second": 1.77,
+      "eval_steps_per_second": 1.77,
+      "eval_wer": 29.848229342327148,
       "step": 40
     },
     {
       "epoch": 1.9047619047619047,
-      "grad_norm": 4.121974468231201,
+      "grad_norm": 4.230419635772705,
       "learning_rate": 3.75e-05,
-      "loss": 0.1865,
+      "loss": 0.1771,
       "step": 60
     },
     {
       "epoch": 2.5396825396825395,
-      "grad_norm": 3.0512349605560303,
+      "grad_norm": 3.738790273666382,
       "learning_rate": 3.75e-05,
-      "loss": 0.1177,
+      "loss": 0.1075,
       "step": 80
     },
     {
       "epoch": 2.5396825396825395,
-      "eval_loss": 0.1945016235113144,
-      "eval_runtime": 55.3505,
-      "eval_samples_per_second": 1.807,
-      "eval_steps_per_second": 1.807,
-      "eval_wer": 31.197301854974707,
+      "eval_loss": 0.18291151523590088,
+      "eval_runtime": 56.5637,
+      "eval_samples_per_second": 1.768,
+      "eval_steps_per_second": 1.768,
+      "eval_wer": 28.920741989881954,
       "step": 80
     },
     {
       "epoch": 3.1746031746031744,
-      "grad_norm": 2.3403635025024414,
+      "grad_norm": 2.266835927963257,
       "learning_rate": 3.75e-05,
-      "loss": 0.0903,
+      "loss": 0.0932,
       "step": 100
     },
     {
       "epoch": 3.8095238095238093,
-      "grad_norm": 2.4136340618133545,
+      "grad_norm": 2.579704523086548,
       "learning_rate": 3.75e-05,
-      "loss": 0.0569,
+      "loss": 0.0513,
       "step": 120
     },
     {
       "epoch": 3.8095238095238093,
-      "eval_loss": 0.20778781175613403,
-      "eval_runtime": 55.9933,
-      "eval_samples_per_second": 1.786,
-      "eval_steps_per_second": 1.786,
-      "eval_wer": 29.34232715008432,
+      "eval_loss": 0.18522793054580688,
+      "eval_runtime": 56.6106,
+      "eval_samples_per_second": 1.766,
+      "eval_steps_per_second": 1.766,
+      "eval_wer": 28.246205733558178,
       "step": 120
     },
     {
       "epoch": 4.444444444444445,
-      "grad_norm": 1.9897055625915527,
+      "grad_norm": 2.1815927028656006,
       "learning_rate": 3.75e-05,
-      "loss": 0.0412,
+      "loss": 0.0365,
       "step": 140
     },
     {
       "epoch": 5.079365079365079,
-      "grad_norm": 1.213258981704712,
+      "grad_norm": 2.4605093002319336,
       "learning_rate": 3.75e-05,
-      "loss": 0.0313,
+      "loss": 0.0283,
       "step": 160
     },
     {
       "epoch": 5.079365079365079,
-      "eval_loss": 0.2349608838558197,
-      "eval_runtime": 56.0059,
-      "eval_samples_per_second": 1.786,
-      "eval_steps_per_second": 1.786,
-      "eval_wer": 31.02866779089376,
+      "eval_loss": 0.19923055171966553,
+      "eval_runtime": 56.6871,
+      "eval_samples_per_second": 1.764,
+      "eval_steps_per_second": 1.764,
+      "eval_wer": 29.51096121416526,
       "step": 160
     },
     {
       "epoch": 5.714285714285714,
-      "grad_norm": 1.2564868927001953,
+      "grad_norm": 1.38411283493042,
       "learning_rate": 3.75e-05,
-      "loss": 0.0198,
+      "loss": 0.0176,
       "step": 180
     },
     {
       "epoch": 6.349206349206349,
-      "grad_norm": 1.6000959873199463,
+      "grad_norm": 0.7200319766998291,
       "learning_rate": 3.75e-05,
-      "loss": 0.0186,
+      "loss": 0.0178,
       "step": 200
     },
     {
       "epoch": 6.349206349206349,
-      "eval_loss": 0.2441704422235489,
-      "eval_runtime": 55.4225,
-      "eval_samples_per_second": 1.804,
-      "eval_steps_per_second": 1.804,
-      "eval_wer": 30.6070826306914,
+      "eval_loss": 0.21108920872211456,
+      "eval_runtime": 56.6862,
+      "eval_samples_per_second": 1.764,
+      "eval_steps_per_second": 1.764,
+      "eval_wer": 28.499156829679595,
       "step": 200
     },
     {
       "epoch": 6.349206349206349,
       "step": 200,
       "total_flos": 1.5637915828224e+17,
-      "train_loss": 0.11623679384589196,
-      "train_runtime": 1122.2532,
-      "train_samples_per_second": 5.703,
-      "train_steps_per_second": 0.178
+      "train_loss": 0.11340499997138977,
+      "train_runtime": 1142.2335,
+      "train_samples_per_second": 5.603,
+      "train_steps_per_second": 0.175
     }
   ],
   "logging_steps": 20,