Training in progress, step 40

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bf1d122ced456c66e97ecd932518df38a9ecdf0da2176728e7b6becb9402c6bb
+oid sha256:d667bd1c2732bc541bac9a4a9e5428578acb8b3dea7e3336e3a97d78bf518926
 size 151099494

runs/Jun29_09-39-32_bharatsahaiyak-test/events.out.tfevents.1719659833.bharatsahaiyak-test.1854617.5

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:17577b647cde917b0ca4ca7707eb89aa0b4a0a68d8505c839931155879bf5f01
+size 6396

trainer_state.json

@@ -1,6 +1,6 @@
 {
-  "best_metric": 37.60539629005059,
-  "best_model_checkpoint": "./whisper-tiny-hindi2_test/checkpoint-80",
+  "best_metric": 32.377740303541316,
+  "best_model_checkpoint": "./whisper-tiny-hindi2_test/checkpoint-200",
   "epoch": 6.349206349206349,
   "eval_steps": 40,
   "global_step": 200,
@@ -10,127 +10,127 @@
   "log_history": [
     {
       "epoch": 0.6349206349206349,
-      "grad_norm": 5.881579875946045,
+      "grad_norm": 5.7862091064453125,
       "learning_rate": 3.75e-05,
-      "loss": 0.3939,
+      "loss": 0.3639,
       "step": 20
     },
     {
       "epoch": 1.2698412698412698,
-      "grad_norm": 4.580193996429443,
+      "grad_norm": 4.1368255615234375,
       "learning_rate": 3.75e-05,
-      "loss": 0.312,
+      "loss": 0.2955,
       "step": 40
     },
     {
       "epoch": 1.2698412698412698,
-      "eval_loss": 0.26820191740989685,
-      "eval_runtime": 55.3328,
-      "eval_samples_per_second": 1.807,
-      "eval_steps_per_second": 1.807,
-      "eval_wer": 38.61720067453626,
+      "eval_loss": 0.23172235488891602,
+      "eval_runtime": 56.6839,
+      "eval_samples_per_second": 1.764,
+      "eval_steps_per_second": 1.764,
+      "eval_wer": 36.00337268128162,
       "step": 40
     },
     {
       "epoch": 1.9047619047619047,
-      "grad_norm": 4.671483039855957,
+      "grad_norm": 4.015283584594727,
       "learning_rate": 3.75e-05,
-      "loss": 0.2298,
+      "loss": 0.2134,
       "step": 60
     },
     {
       "epoch": 2.5396825396825395,
-      "grad_norm": 4.34619140625,
+      "grad_norm": 3.215930223464966,
       "learning_rate": 3.75e-05,
-      "loss": 0.1574,
+      "loss": 0.1522,
       "step": 80
     },
     {
       "epoch": 2.5396825396825395,
-      "eval_loss": 0.2759854197502136,
-      "eval_runtime": 55.1922,
-      "eval_samples_per_second": 1.812,
-      "eval_steps_per_second": 1.812,
-      "eval_wer": 37.60539629005059,
+      "eval_loss": 0.22003404796123505,
+      "eval_runtime": 55.4903,
+      "eval_samples_per_second": 1.802,
+      "eval_steps_per_second": 1.802,
+      "eval_wer": 34.40134907251265,
       "step": 80
     },
     {
       "epoch": 3.1746031746031744,
-      "grad_norm": 2.468874216079712,
+      "grad_norm": 2.828727960586548,
       "learning_rate": 3.75e-05,
-      "loss": 0.1358,
+      "loss": 0.1209,
       "step": 100
     },
     {
       "epoch": 3.8095238095238093,
-      "grad_norm": 3.0758771896362305,
+      "grad_norm": 2.526444673538208,
       "learning_rate": 3.75e-05,
-      "loss": 0.0914,
+      "loss": 0.0819,
       "step": 120
     },
     {
       "epoch": 3.8095238095238093,
-      "eval_loss": 0.29828593134880066,
-      "eval_runtime": 54.9217,
-      "eval_samples_per_second": 1.821,
-      "eval_steps_per_second": 1.821,
-      "eval_wer": 37.60539629005059,
+      "eval_loss": 0.22554056346416473,
+      "eval_runtime": 54.7978,
+      "eval_samples_per_second": 1.825,
+      "eval_steps_per_second": 1.825,
+      "eval_wer": 32.54637436762226,
       "step": 120
     },
     {
       "epoch": 4.444444444444445,
-      "grad_norm": 2.6510117053985596,
+      "grad_norm": 2.135953664779663,
       "learning_rate": 3.75e-05,
-      "loss": 0.0666,
+      "loss": 0.061,
       "step": 140
     },
     {
       "epoch": 5.079365079365079,
-      "grad_norm": 1.7483832836151123,
+      "grad_norm": 2.1400399208068848,
       "learning_rate": 3.75e-05,
-      "loss": 0.0579,
+      "loss": 0.0463,
       "step": 160
     },
     {
       "epoch": 5.079365079365079,
-      "eval_loss": 0.3087254464626312,
-      "eval_runtime": 54.9389,
-      "eval_samples_per_second": 1.82,
-      "eval_steps_per_second": 1.82,
-      "eval_wer": 37.77403035413153,
+      "eval_loss": 0.24825386703014374,
+      "eval_runtime": 54.848,
+      "eval_samples_per_second": 1.823,
+      "eval_steps_per_second": 1.823,
+      "eval_wer": 32.79932546374368,
       "step": 160
     },
     {
       "epoch": 5.714285714285714,
-      "grad_norm": 2.245854139328003,
+      "grad_norm": 1.9905014038085938,
       "learning_rate": 3.75e-05,
-      "loss": 0.0354,
+      "loss": 0.0304,
       "step": 180
     },
     {
       "epoch": 6.349206349206349,
-      "grad_norm": 1.88152015209198,
+      "grad_norm": 1.8886462450027466,
       "learning_rate": 3.75e-05,
-      "loss": 0.028,
+      "loss": 0.0289,
       "step": 200
     },
     {
       "epoch": 6.349206349206349,
-      "eval_loss": 0.3585308790206909,
-      "eval_runtime": 54.6288,
-      "eval_samples_per_second": 1.831,
-      "eval_steps_per_second": 1.831,
-      "eval_wer": 38.27993254637437,
+      "eval_loss": 0.2673737406730652,
+      "eval_runtime": 55.0065,
+      "eval_samples_per_second": 1.818,
+      "eval_steps_per_second": 1.818,
+      "eval_wer": 32.377740303541316,
       "step": 200
     },
     {
       "epoch": 6.349206349206349,
       "step": 200,
       "total_flos": 1.5637915828224e+17,
-      "train_loss": 0.15080839216709138,
-      "train_runtime": 1129.7922,
-      "train_samples_per_second": 5.665,
-      "train_steps_per_second": 0.177
+      "train_loss": 0.13942732244729997,
+      "train_runtime": 1124.5585,
+      "train_samples_per_second": 5.691,
+      "train_steps_per_second": 0.178
     }
   ],
   "logging_steps": 20,