End of training

trainer_state.json

@@ -1,7 +1,7 @@
 {
-  "best_metric": 25.293262636395276,
+  "best_metric": 25.674043441767154,
   "best_model_checkpoint": "whisper-base-nl-2/checkpoint-1000",
-  "epoch": 0.2139495079161318,
+  "epoch": 0.18277080541001584,
   "global_step": 1500,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
@@ -10,390 +10,390 @@
     {
       "epoch": 0.0,
       "learning_rate": 4.0000000000000003e-07,
-      "loss": 1.3206,
+      "loss": 1.4165,
       "step": 25
     },
     {
       "epoch": 0.01,
-      "learning_rate": 8.8e-07,
-      "loss": 1.247,
+      "learning_rate": 9.000000000000001e-07,
+      "loss": 1.2919,
       "step": 50
     },
     {
       "epoch": 0.01,
       "learning_rate": 1.3800000000000001e-06,
-      "loss": 1.1224,
+      "loss": 1.0487,
       "step": 75
     },
     {
       "epoch": 0.01,
       "learning_rate": 1.8800000000000002e-06,
-      "loss": 0.9567,
+      "loss": 1.0115,
       "step": 100
     },
     {
       "epoch": 0.02,
       "learning_rate": 2.38e-06,
-      "loss": 0.964,
+      "loss": 0.9841,
       "step": 125
     },
     {
       "epoch": 0.02,
       "learning_rate": 2.88e-06,
-      "loss": 0.7427,
+      "loss": 0.99,
       "step": 150
     },
     {
       "epoch": 0.02,
       "learning_rate": 3.3800000000000007e-06,
-      "loss": 0.6921,
+      "loss": 0.8826,
       "step": 175
     },
     {
-      "epoch": 0.03,
+      "epoch": 0.02,
       "learning_rate": 3.88e-06,
-      "loss": 0.6178,
+      "loss": 0.6872,
       "step": 200
     },
     {
       "epoch": 0.03,
       "learning_rate": 4.38e-06,
-      "loss": 0.6377,
+      "loss": 0.6233,
       "step": 225
     },
     {
-      "epoch": 0.04,
+      "epoch": 0.03,
       "learning_rate": 4.880000000000001e-06,
-      "loss": 0.5442,
+      "loss": 0.6387,
       "step": 250
     },
     {
-      "epoch": 0.04,
+      "epoch": 0.03,
       "learning_rate": 5.380000000000001e-06,
-      "loss": 0.483,
+      "loss": 0.7633,
       "step": 275
     },
     {
       "epoch": 0.04,
       "learning_rate": 5.8800000000000005e-06,
-      "loss": 0.5293,
+      "loss": 0.5987,
       "step": 300
     },
     {
-      "epoch": 0.05,
+      "epoch": 0.04,
       "learning_rate": 6.380000000000001e-06,
-      "loss": 0.5293,
+      "loss": 0.5338,
       "step": 325
     },
     {
-      "epoch": 0.05,
+      "epoch": 0.04,
       "learning_rate": 6.88e-06,
-      "loss": 0.5213,
+      "loss": 0.5378,
       "step": 350
     },
     {
       "epoch": 0.05,
       "learning_rate": 7.3800000000000005e-06,
-      "loss": 0.4849,
+      "loss": 0.5808,
       "step": 375
     },
     {
-      "epoch": 0.06,
+      "epoch": 0.05,
       "learning_rate": 7.88e-06,
-      "loss": 0.5068,
+      "loss": 0.5216,
       "step": 400
     },
     {
-      "epoch": 0.06,
+      "epoch": 0.05,
       "learning_rate": 8.380000000000001e-06,
-      "loss": 0.4787,
+      "loss": 0.503,
       "step": 425
     },
     {
-      "epoch": 0.06,
+      "epoch": 0.05,
       "learning_rate": 8.880000000000001e-06,
-      "loss": 0.4744,
+      "loss": 0.5549,
       "step": 450
     },
     {
-      "epoch": 0.07,
+      "epoch": 0.06,
       "learning_rate": 9.38e-06,
-      "loss": 0.4914,
+      "loss": 0.6421,
       "step": 475
     },
     {
-      "epoch": 0.07,
+      "epoch": 0.06,
       "learning_rate": 9.88e-06,
-      "loss": 0.4643,
+      "loss": 0.4607,
       "step": 500
     },
     {
-      "epoch": 0.07,
+      "epoch": 0.06,
       "learning_rate": 9.810000000000001e-06,
-      "loss": 0.4792,
+      "loss": 0.6395,
       "step": 525
     },
     {
-      "epoch": 0.08,
+      "epoch": 0.07,
       "learning_rate": 9.56e-06,
-      "loss": 0.4715,
+      "loss": 0.5217,
       "step": 550
     },
     {
-      "epoch": 0.08,
+      "epoch": 0.07,
       "learning_rate": 9.31e-06,
-      "loss": 0.4405,
+      "loss": 0.4829,
       "step": 575
     },
     {
-      "epoch": 0.09,
+      "epoch": 0.07,
       "learning_rate": 9.060000000000001e-06,
-      "loss": 0.4577,
+      "loss": 0.644,
       "step": 600
     },
     {
-      "epoch": 0.09,
+      "epoch": 0.08,
       "learning_rate": 8.81e-06,
-      "loss": 0.4624,
+      "loss": 0.4887,
       "step": 625
     },
     {
-      "epoch": 0.09,
+      "epoch": 0.08,
       "learning_rate": 8.560000000000001e-06,
-      "loss": 0.3639,
+      "loss": 0.5319,
       "step": 650
     },
     {
-      "epoch": 0.1,
+      "epoch": 0.08,
       "learning_rate": 8.31e-06,
-      "loss": 0.4156,
+      "loss": 0.4961,
       "step": 675
     },
     {
-      "epoch": 0.1,
+      "epoch": 0.09,
       "learning_rate": 8.06e-06,
-      "loss": 0.4121,
+      "loss": 0.4463,
       "step": 700
     },
     {
-      "epoch": 0.1,
+      "epoch": 0.09,
       "learning_rate": 7.810000000000001e-06,
-      "loss": 0.3885,
+      "loss": 0.4523,
       "step": 725
     },
     {
-      "epoch": 0.11,
+      "epoch": 0.09,
       "learning_rate": 7.5600000000000005e-06,
-      "loss": 0.4401,
+      "loss": 0.5192,
       "step": 750
     },
     {
-      "epoch": 0.11,
+      "epoch": 0.09,
       "learning_rate": 7.31e-06,
-      "loss": 0.4141,
+      "loss": 0.5411,
       "step": 775
     },
     {
-      "epoch": 0.11,
+      "epoch": 0.1,
       "learning_rate": 7.06e-06,
-      "loss": 0.4179,
+      "loss": 0.472,
       "step": 800
     },
     {
-      "epoch": 0.12,
+      "epoch": 0.1,
       "learning_rate": 6.810000000000001e-06,
-      "loss": 0.4299,
+      "loss": 0.5429,
       "step": 825
     },
     {
-      "epoch": 0.12,
+      "epoch": 0.1,
       "learning_rate": 6.560000000000001e-06,
-      "loss": 0.3733,
+      "loss": 0.4652,
       "step": 850
     },
     {
-      "epoch": 0.12,
+      "epoch": 0.11,
       "learning_rate": 6.3100000000000006e-06,
-      "loss": 0.429,
+      "loss": 0.4493,
       "step": 875
     },
     {
-      "epoch": 0.13,
+      "epoch": 0.11,
       "learning_rate": 6.0600000000000004e-06,
-      "loss": 0.3495,
+      "loss": 0.5725,
       "step": 900
     },
     {
-      "epoch": 0.13,
+      "epoch": 0.11,
       "learning_rate": 5.81e-06,
-      "loss": 0.373,
+      "loss": 0.5933,
       "step": 925
     },
     {
-      "epoch": 0.14,
+      "epoch": 0.12,
       "learning_rate": 5.560000000000001e-06,
-      "loss": 0.3427,
+      "loss": 0.4506,
       "step": 950
     },
     {
-      "epoch": 0.14,
+      "epoch": 0.12,
       "learning_rate": 5.310000000000001e-06,
-      "loss": 0.3687,
+      "loss": 0.457,
       "step": 975
     },
     {
-      "epoch": 0.14,
+      "epoch": 0.12,
       "learning_rate": 5.060000000000001e-06,
-      "loss": 0.3749,
+      "loss": 0.4709,
       "step": 1000
     },
     {
-      "epoch": 0.14,
-      "eval_cer": 8.575715070233361,
-      "eval_loss": 0.5007154941558838,
-      "eval_runtime": 2429.0625,
+      "epoch": 0.12,
+      "eval_cer": 8.48722037329875,
+      "eval_loss": 0.49462148547172546,
+      "eval_runtime": 2428.8628,
       "eval_samples_per_second": 0.804,
       "eval_steps_per_second": 0.402,
-      "eval_wer": 25.293262636395276,
+      "eval_wer": 25.674043441767154,
       "step": 1000
     },
     {
-      "epoch": 0.15,
+      "epoch": 0.12,
       "learning_rate": 4.8100000000000005e-06,
-      "loss": 0.3366,
+      "loss": 0.3941,
       "step": 1025
     },
     {
-      "epoch": 0.15,
+      "epoch": 0.13,
       "learning_rate": 4.56e-06,
-      "loss": 0.3493,
+      "loss": 0.4826,
       "step": 1050
     },
     {
-      "epoch": 0.15,
+      "epoch": 0.13,
       "learning_rate": 4.31e-06,
-      "loss": 0.3721,
+      "loss": 0.4664,
       "step": 1075
     },
     {
-      "epoch": 0.16,
+      "epoch": 0.13,
       "learning_rate": 4.060000000000001e-06,
-      "loss": 0.3504,
+      "loss": 0.3897,
       "step": 1100
     },
     {
-      "epoch": 0.16,
+      "epoch": 0.14,
       "learning_rate": 3.8100000000000004e-06,
-      "loss": 0.3543,
+      "loss": 0.5067,
       "step": 1125
     },
     {
-      "epoch": 0.16,
+      "epoch": 0.14,
       "learning_rate": 3.5600000000000002e-06,
-      "loss": 0.395,
+      "loss": 0.466,
       "step": 1150
     },
     {
-      "epoch": 0.17,
+      "epoch": 0.14,
       "learning_rate": 3.3100000000000005e-06,
-      "loss": 0.3789,
+      "loss": 0.495,
       "step": 1175
     },
     {
-      "epoch": 0.17,
+      "epoch": 0.15,
       "learning_rate": 3.0600000000000003e-06,
-      "loss": 0.3223,
+      "loss": 0.3934,
       "step": 1200
     },
     {
-      "epoch": 0.17,
+      "epoch": 0.15,
       "learning_rate": 2.8100000000000006e-06,
-      "loss": 0.3242,
+      "loss": 0.4138,
       "step": 1225
     },
     {
-      "epoch": 0.18,
+      "epoch": 0.15,
       "learning_rate": 2.56e-06,
-      "loss": 0.3701,
+      "loss": 0.4547,
       "step": 1250
     },
     {
-      "epoch": 0.18,
+      "epoch": 0.16,
       "learning_rate": 2.3100000000000003e-06,
-      "loss": 0.3882,
+      "loss": 0.4548,
       "step": 1275
     },
     {
-      "epoch": 0.19,
+      "epoch": 0.16,
       "learning_rate": 2.06e-06,
-      "loss": 0.3678,
+      "loss": 0.4612,
       "step": 1300
     },
     {
-      "epoch": 0.19,
+      "epoch": 0.16,
       "learning_rate": 1.81e-06,
-      "loss": 0.3539,
+      "loss": 0.5229,
       "step": 1325
     },
     {
-      "epoch": 0.19,
+      "epoch": 0.16,
       "learning_rate": 1.56e-06,
-      "loss": 0.3723,
+      "loss": 0.3348,
       "step": 1350
     },
     {
-      "epoch": 0.2,
+      "epoch": 0.17,
       "learning_rate": 1.3100000000000002e-06,
-      "loss": 0.3184,
+      "loss": 0.4843,
       "step": 1375
     },
     {
-      "epoch": 0.2,
+      "epoch": 0.17,
       "learning_rate": 1.06e-06,
-      "loss": 0.3204,
+      "loss": 0.3578,
       "step": 1400
     },
     {
-      "epoch": 0.2,
+      "epoch": 0.17,
       "learning_rate": 8.100000000000001e-07,
-      "loss": 0.3884,
+      "loss": 0.4753,
       "step": 1425
     },
     {
-      "epoch": 0.21,
+      "epoch": 0.18,
       "learning_rate": 5.6e-07,
-      "loss": 0.3948,
+      "loss": 0.441,
       "step": 1450
     },
     {
-      "epoch": 0.21,
+      "epoch": 0.18,
       "learning_rate": 3.1000000000000005e-07,
-      "loss": 0.3151,
+      "loss": 0.3252,
       "step": 1475
     },
     {
-      "epoch": 0.21,
+      "epoch": 0.18,
       "learning_rate": 6.000000000000001e-08,
-      "loss": 0.3676,
+      "loss": 0.4169,
       "step": 1500
     },
     {
-      "epoch": 0.21,
+      "epoch": 0.18,
       "step": 1500,
       "total_flos": 9.728999424e+16,
-      "train_loss": 0.4858832302093506,
-      "train_runtime": 2744.5265,
-      "train_samples_per_second": 0.547,
-      "train_steps_per_second": 0.547
-    },
-    {
-      "epoch": 0.21,
-      "eval_cer": 8.575715070233361,
-      "eval_loss": 0.5007154941558838,
-      "eval_runtime": 2408.5648,
-      "eval_samples_per_second": 0.811,
-      "eval_steps_per_second": 0.406,
-      "eval_wer": 25.293262636395276,
+      "train_loss": 0.5707454249064128,
+      "train_runtime": 2748.8536,
+      "train_samples_per_second": 0.546,
+      "train_steps_per_second": 0.546
+    },
+    {
+      "epoch": 0.18,
+      "eval_cer": 8.48722037329875,
+      "eval_loss": 0.49462148547172546,
+      "eval_runtime": 2417.8085,
+      "eval_samples_per_second": 0.808,
+      "eval_steps_per_second": 0.404,
+      "eval_wer": 25.674043441767154,
       "step": 1500
     }
   ],