Training in progress, step 40

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e83c4fe02f5392ab3ed8a461043b0b685b305a863f3a18e3a1fca823e66a06ca
+oid sha256:ea5038f27370638d492fddc58c333f1164ffa5af2486a0669818856adcdd8fe6
 size 151099494

runs/Jun29_09-39-32_bharatsahaiyak-test/events.out.tfevents.1719656331.bharatsahaiyak-test.1854617.2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:15a929009023974b064687f87b39150a992f59c732c03358d85440603f7e558a
+size 6396

trainer_state.json

@@ -1,6 +1,6 @@
 {
-  "best_metric": 50.0,
-  "best_model_checkpoint": "./whisper-tiny-hindi2_test/checkpoint-120",
+  "best_metric": 41.31534569983137,
+  "best_model_checkpoint": "./whisper-tiny-hindi2_test/checkpoint-160",
   "epoch": 6.349206349206349,
   "eval_steps": 40,
   "global_step": 200,
@@ -10,127 +10,127 @@
   "log_history": [
     {
       "epoch": 0.6349206349206349,
-      "grad_norm": 8.679319381713867,
+      "grad_norm": 5.291396141052246,
       "learning_rate": 3.75e-05,
-      "loss": 1.6908,
+      "loss": 0.486,
       "step": 20
     },
     {
       "epoch": 1.2698412698412698,
-      "grad_norm": 7.228819370269775,
+      "grad_norm": 4.965972423553467,
       "learning_rate": 3.75e-05,
-      "loss": 0.6584,
+      "loss": 0.3869,
       "step": 40
     },
     {
       "epoch": 1.2698412698412698,
-      "eval_loss": 0.593582272529602,
-      "eval_runtime": 54.5955,
-      "eval_samples_per_second": 1.832,
-      "eval_steps_per_second": 1.832,
-      "eval_wer": 65.00843170320405,
+      "eval_loss": 0.35420987010002136,
+      "eval_runtime": 54.9702,
+      "eval_samples_per_second": 1.819,
+      "eval_steps_per_second": 1.819,
+      "eval_wer": 46.964586846543,
       "step": 40
     },
     {
       "epoch": 1.9047619047619047,
-      "grad_norm": 6.050016403198242,
+      "grad_norm": 5.148536682128906,
       "learning_rate": 3.75e-05,
-      "loss": 0.4786,
+      "loss": 0.3043,
       "step": 60
     },
     {
       "epoch": 2.5396825396825395,
-      "grad_norm": 5.063595294952393,
+      "grad_norm": 4.049417972564697,
       "learning_rate": 3.75e-05,
-      "loss": 0.3575,
+      "loss": 0.2215,
       "step": 80
     },
     {
       "epoch": 2.5396825396825395,
-      "eval_loss": 0.46842461824417114,
-      "eval_runtime": 55.6599,
-      "eval_samples_per_second": 1.797,
-      "eval_steps_per_second": 1.797,
-      "eval_wer": 55.31197301854974,
+      "eval_loss": 0.3464047908782959,
+      "eval_runtime": 55.3552,
+      "eval_samples_per_second": 1.807,
+      "eval_steps_per_second": 1.807,
+      "eval_wer": 45.27824620573356,
       "step": 80
     },
     {
       "epoch": 3.1746031746031744,
-      "grad_norm": 4.799503326416016,
+      "grad_norm": 3.4034841060638428,
       "learning_rate": 3.75e-05,
-      "loss": 0.2994,
+      "loss": 0.1888,
       "step": 100
     },
     {
       "epoch": 3.8095238095238093,
-      "grad_norm": 5.387985706329346,
+      "grad_norm": 4.5857930183410645,
       "learning_rate": 3.75e-05,
-      "loss": 0.2254,
+      "loss": 0.1394,
       "step": 120
     },
     {
       "epoch": 3.8095238095238093,
-      "eval_loss": 0.4344378709793091,
-      "eval_runtime": 54.7109,
-      "eval_samples_per_second": 1.828,
-      "eval_steps_per_second": 1.828,
-      "eval_wer": 50.0,
+      "eval_loss": 0.34948885440826416,
+      "eval_runtime": 54.9671,
+      "eval_samples_per_second": 1.819,
+      "eval_steps_per_second": 1.819,
+      "eval_wer": 43.00168634064081,
       "step": 120
     },
     {
       "epoch": 4.444444444444445,
-      "grad_norm": 4.451031684875488,
+      "grad_norm": 3.3065757751464844,
       "learning_rate": 3.75e-05,
-      "loss": 0.1752,
+      "loss": 0.0954,
       "step": 140
     },
     {
       "epoch": 5.079365079365079,
-      "grad_norm": 4.960638046264648,
+      "grad_norm": 2.241292953491211,
       "learning_rate": 3.75e-05,
-      "loss": 0.1486,
+      "loss": 0.0857,
       "step": 160
     },
     {
       "epoch": 5.079365079365079,
-      "eval_loss": 0.4561767578125,
-      "eval_runtime": 54.4171,
-      "eval_samples_per_second": 1.838,
-      "eval_steps_per_second": 1.838,
-      "eval_wer": 52.1079258010118,
+      "eval_loss": 0.37522462010383606,
+      "eval_runtime": 55.4528,
+      "eval_samples_per_second": 1.803,
+      "eval_steps_per_second": 1.803,
+      "eval_wer": 41.31534569983137,
       "step": 160
     },
     {
       "epoch": 5.714285714285714,
-      "grad_norm": 3.8416409492492676,
+      "grad_norm": 2.4962286949157715,
       "learning_rate": 3.75e-05,
-      "loss": 0.1021,
+      "loss": 0.055,
       "step": 180
     },
     {
       "epoch": 6.349206349206349,
-      "grad_norm": 3.137124538421631,
+      "grad_norm": 2.652437210083008,
       "learning_rate": 3.75e-05,
-      "loss": 0.0868,
+      "loss": 0.0434,
       "step": 200
     },
     {
       "epoch": 6.349206349206349,
-      "eval_loss": 0.46873587369918823,
-      "eval_runtime": 54.2397,
-      "eval_samples_per_second": 1.844,
-      "eval_steps_per_second": 1.844,
-      "eval_wer": 50.5902192242833,
+      "eval_loss": 0.40884342789649963,
+      "eval_runtime": 55.4835,
+      "eval_samples_per_second": 1.802,
+      "eval_steps_per_second": 1.802,
+      "eval_wer": 43.76053962900506,
       "step": 200
     },
     {
       "epoch": 6.349206349206349,
       "step": 200,
       "total_flos": 1.5637915828224e+17,
-      "train_loss": 0.4222871881723404,
-      "train_runtime": 1142.3513,
-      "train_samples_per_second": 5.602,
-      "train_steps_per_second": 0.175
+      "train_loss": 0.2006339970231056,
+      "train_runtime": 1120.101,
+      "train_samples_per_second": 5.714,
+      "train_steps_per_second": 0.179
     }
   ],
   "logging_steps": 20,