Training in progress, step 40

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e0d7e37a95520f41735f4e81928cc4a409afc0daed4fe386710ac7f2ac7ad438
+oid sha256:76b38665cb612c63735d262bdbf1373e3aa13465ce6f2dc69e1dc1b9ddef2dbe
 size 151099494

runs/Jun29_09-39-32_bharatsahaiyak-test/events.out.tfevents.1719664557.bharatsahaiyak-test.1854617.9

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a35c59c21e20f207a78d90e2992718b8b9e35524d07f59d44c0c1ab3cb5f5fdb
+size 6396

trainer_state.json

@@ -1,6 +1,6 @@
 {
-  "best_metric": 28.246205733558178,
-  "best_model_checkpoint": "./whisper-tiny-hindi2_test/checkpoint-120",
+  "best_metric": 25.88532883642496,
+  "best_model_checkpoint": "./whisper-tiny-hindi2_test/checkpoint-40",
   "epoch": 6.349206349206349,
   "eval_steps": 40,
   "global_step": 200,
@@ -10,127 +10,127 @@
   "log_history": [
     {
       "epoch": 0.6349206349206349,
-      "grad_norm": 5.915639877319336,
+      "grad_norm": 5.399875164031982,
       "learning_rate": 3.75e-05,
-      "loss": 0.3473,
+      "loss": 0.3295,
       "step": 20
     },
     {
       "epoch": 1.2698412698412698,
-      "grad_norm": 3.9101755619049072,
+      "grad_norm": 3.3471007347106934,
       "learning_rate": 3.75e-05,
-      "loss": 0.2574,
+      "loss": 0.2449,
       "step": 40
     },
     {
       "epoch": 1.2698412698412698,
-      "eval_loss": 0.18214347958564758,
-      "eval_runtime": 56.4889,
-      "eval_samples_per_second": 1.77,
-      "eval_steps_per_second": 1.77,
-      "eval_wer": 29.848229342327148,
+      "eval_loss": 0.15969730913639069,
+      "eval_runtime": 56.6547,
+      "eval_samples_per_second": 1.765,
+      "eval_steps_per_second": 1.765,
+      "eval_wer": 25.88532883642496,
       "step": 40
     },
     {
       "epoch": 1.9047619047619047,
-      "grad_norm": 4.230419635772705,
+      "grad_norm": 3.7166926860809326,
       "learning_rate": 3.75e-05,
-      "loss": 0.1771,
+      "loss": 0.1628,
       "step": 60
     },
     {
       "epoch": 2.5396825396825395,
-      "grad_norm": 3.738790273666382,
+      "grad_norm": 2.8016231060028076,
       "learning_rate": 3.75e-05,
-      "loss": 0.1075,
+      "loss": 0.1002,
       "step": 80
     },
     {
       "epoch": 2.5396825396825395,
-      "eval_loss": 0.18291151523590088,
-      "eval_runtime": 56.5637,
-      "eval_samples_per_second": 1.768,
-      "eval_steps_per_second": 1.768,
-      "eval_wer": 28.920741989881954,
+      "eval_loss": 0.16435778141021729,
+      "eval_runtime": 56.8979,
+      "eval_samples_per_second": 1.758,
+      "eval_steps_per_second": 1.758,
+      "eval_wer": 26.644182124789207,
       "step": 80
     },
     {
       "epoch": 3.1746031746031744,
-      "grad_norm": 2.266835927963257,
+      "grad_norm": 2.115607976913452,
       "learning_rate": 3.75e-05,
-      "loss": 0.0932,
+      "loss": 0.0833,
       "step": 100
     },
     {
       "epoch": 3.8095238095238093,
-      "grad_norm": 2.579704523086548,
+      "grad_norm": 2.412848711013794,
       "learning_rate": 3.75e-05,
-      "loss": 0.0513,
+      "loss": 0.0492,
       "step": 120
     },
     {
       "epoch": 3.8095238095238093,
-      "eval_loss": 0.18522793054580688,
-      "eval_runtime": 56.6106,
-      "eval_samples_per_second": 1.766,
-      "eval_steps_per_second": 1.766,
-      "eval_wer": 28.246205733558178,
+      "eval_loss": 0.1730262041091919,
+      "eval_runtime": 56.4319,
+      "eval_samples_per_second": 1.772,
+      "eval_steps_per_second": 1.772,
+      "eval_wer": 27.065767284991573,
       "step": 120
     },
     {
       "epoch": 4.444444444444445,
-      "grad_norm": 2.1815927028656006,
+      "grad_norm": 1.6271615028381348,
       "learning_rate": 3.75e-05,
-      "loss": 0.0365,
+      "loss": 0.0323,
       "step": 140
     },
     {
       "epoch": 5.079365079365079,
-      "grad_norm": 2.4605093002319336,
+      "grad_norm": 1.2262483835220337,
       "learning_rate": 3.75e-05,
-      "loss": 0.0283,
+      "loss": 0.0269,
       "step": 160
     },
     {
       "epoch": 5.079365079365079,
-      "eval_loss": 0.19923055171966553,
-      "eval_runtime": 56.6871,
-      "eval_samples_per_second": 1.764,
-      "eval_steps_per_second": 1.764,
-      "eval_wer": 29.51096121416526,
+      "eval_loss": 0.1800723522901535,
+      "eval_runtime": 57.7815,
+      "eval_samples_per_second": 1.731,
+      "eval_steps_per_second": 1.731,
+      "eval_wer": 26.981450252951095,
       "step": 160
     },
     {
       "epoch": 5.714285714285714,
-      "grad_norm": 1.38411283493042,
+      "grad_norm": 1.4037569761276245,
       "learning_rate": 3.75e-05,
-      "loss": 0.0176,
+      "loss": 0.0162,
       "step": 180
     },
     {
       "epoch": 6.349206349206349,
-      "grad_norm": 0.7200319766998291,
+      "grad_norm": 1.1472747325897217,
       "learning_rate": 3.75e-05,
-      "loss": 0.0178,
+      "loss": 0.0156,
       "step": 200
     },
     {
       "epoch": 6.349206349206349,
-      "eval_loss": 0.21108920872211456,
-      "eval_runtime": 56.6862,
-      "eval_samples_per_second": 1.764,
-      "eval_steps_per_second": 1.764,
-      "eval_wer": 28.499156829679595,
+      "eval_loss": 0.20425428450107574,
+      "eval_runtime": 56.7961,
+      "eval_samples_per_second": 1.761,
+      "eval_steps_per_second": 1.761,
+      "eval_wer": 27.65598650927487,
       "step": 200
     },
     {
       "epoch": 6.349206349206349,
       "step": 200,
       "total_flos": 1.5637915828224e+17,
-      "train_loss": 0.11340499997138977,
-      "train_runtime": 1142.2335,
-      "train_samples_per_second": 5.603,
-      "train_steps_per_second": 0.175
+      "train_loss": 0.10607707306742668,
+      "train_runtime": 1147.3926,
+      "train_samples_per_second": 5.578,
+      "train_steps_per_second": 0.174
     }
   ],
   "logging_steps": 20,