Upload folder using huggingface_hub

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:93f4e85472794bbcbcfc4dec49c7df91327dc88187fe9bd08402d262a85f4793
+oid sha256:ef25ddaf7598f17b6beb19d6ca630262d89b17ed59abfc57d0b871488fb73f0f
 size 966995080

optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:183a575b1c7bae52a385aaf65d40648fefbe2294ff323ee050506238661dfb4f
+oid sha256:abd58696d97d207d78479ef8f176c8486c2fe5596b7147c40034c3af9d8861e3
 size 1925064044

trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 81.30954374833746,
+  "best_metric": 81.65732357316676,
   "best_model_checkpoint": "/scratch/p310333/whisper-small-dialect_all_seed84/checkpoint-1750",
   "epoch": 0.3017137340091721,
   "eval_steps": 250,
@@ -10,802 +10,802 @@
   "log_history": [
     {
       "epoch": 0.003017137340091721,
-      "grad_norm": 65.5420150756836,
+      "grad_norm": 65.54208374023438,
       "learning_rate": 5.000000000000001e-07,
       "loss": 4.9187,
       "step": 25
     },
     {
       "epoch": 0.006034274680183442,
-      "grad_norm": 30.61561393737793,
+      "grad_norm": 30.615581512451172,
       "learning_rate": 1.0000000000000002e-06,
       "loss": 4.0023,
       "step": 50
     },
     {
       "epoch": 0.009051412020275163,
-      "grad_norm": 29.28253173828125,
+      "grad_norm": 29.282546997070312,
       "learning_rate": 1.5e-06,
       "loss": 3.4756,
       "step": 75
     },
     {
       "epoch": 0.012068549360366883,
-      "grad_norm": 28.802133560180664,
+      "grad_norm": 28.80223846435547,
       "learning_rate": 2.0000000000000003e-06,
       "loss": 2.7255,
       "step": 100
     },
     {
       "epoch": 0.015085686700458605,
-      "grad_norm": 32.795570373535156,
+      "grad_norm": 32.79741287231445,
       "learning_rate": 2.5e-06,
       "loss": 2.4196,
       "step": 125
     },
     {
       "epoch": 0.018102824040550327,
-      "grad_norm": 23.08877182006836,
+      "grad_norm": 23.08782196044922,
       "learning_rate": 3e-06,
       "loss": 2.2985,
       "step": 150
     },
     {
       "epoch": 0.021119961380642045,
-      "grad_norm": 26.61231231689453,
+      "grad_norm": 26.61124610900879,
       "learning_rate": 3.5e-06,
-      "loss": 2.3416,
+      "loss": 2.3417,
       "step": 175
     },
     {
       "epoch": 0.024137098720733767,
-      "grad_norm": 19.47349739074707,
+      "grad_norm": 19.468734741210938,
       "learning_rate": 4.000000000000001e-06,
-      "loss": 2.1088,
+      "loss": 2.1087,
       "step": 200
     },
     {
       "epoch": 0.02715423606082549,
-      "grad_norm": 33.220603942871094,
+      "grad_norm": 33.279293060302734,
       "learning_rate": 4.5e-06,
       "loss": 2.0878,
       "step": 225
     },
     {
       "epoch": 0.03017137340091721,
-      "grad_norm": 27.01511001586914,
+      "grad_norm": 27.024686813354492,
       "learning_rate": 5e-06,
-      "loss": 2.1127,
+      "loss": 2.1126,
       "step": 250
     },
     {
       "epoch": 0.03017137340091721,
-      "eval_cer": 75.04073104686333,
-      "eval_loss": 1.908042073249817,
-      "eval_runtime": 5452.7397,
-      "eval_samples_per_second": 3.039,
-      "eval_steps_per_second": 0.38,
-      "eval_wer": 89.40503049231042,
+      "eval_cer": 75.39939054285433,
+      "eval_loss": 1.9080588817596436,
+      "eval_runtime": 4890.5649,
+      "eval_samples_per_second": 3.389,
+      "eval_steps_per_second": 0.424,
+      "eval_wer": 89.64246657668397,
       "step": 250
     },
     {
       "epoch": 0.03318851074100893,
-      "grad_norm": 24.661699295043945,
+      "grad_norm": 24.560420989990234,
       "learning_rate": 5.500000000000001e-06,
-      "loss": 1.8497,
+      "loss": 1.8494,
       "step": 275
     },
     {
       "epoch": 0.036205648081100654,
-      "grad_norm": 24.292240142822266,
+      "grad_norm": 24.324853897094727,
       "learning_rate": 6e-06,
-      "loss": 1.8874,
+      "loss": 1.8878,
       "step": 300
     },
     {
       "epoch": 0.039222785421192376,
-      "grad_norm": 28.39283561706543,
+      "grad_norm": 28.385276794433594,
       "learning_rate": 6.5000000000000004e-06,
-      "loss": 1.7338,
+      "loss": 1.7333,
       "step": 325
     },
     {
       "epoch": 0.04223992276128409,
-      "grad_norm": 28.00855827331543,
+      "grad_norm": 27.999126434326172,
       "learning_rate": 7e-06,
       "loss": 1.6565,
       "step": 350
     },
     {
       "epoch": 0.04525706010137581,
-      "grad_norm": 29.236988067626953,
+      "grad_norm": 29.16621208190918,
       "learning_rate": 7.500000000000001e-06,
-      "loss": 1.471,
+      "loss": 1.4705,
       "step": 375
     },
     {
       "epoch": 0.048274197441467534,
-      "grad_norm": 21.41299057006836,
+      "grad_norm": 21.424270629882812,
       "learning_rate": 8.000000000000001e-06,
-      "loss": 1.377,
+      "loss": 1.3766,
       "step": 400
     },
     {
       "epoch": 0.051291334781559256,
-      "grad_norm": 25.506296157836914,
+      "grad_norm": 25.499229431152344,
       "learning_rate": 8.5e-06,
-      "loss": 1.3921,
+      "loss": 1.3916,
       "step": 425
     },
     {
       "epoch": 0.05430847212165098,
-      "grad_norm": 29.994136810302734,
+      "grad_norm": 30.007627487182617,
       "learning_rate": 9e-06,
-      "loss": 1.4203,
+      "loss": 1.4213,
       "step": 450
     },
     {
       "epoch": 0.0573256094617427,
-      "grad_norm": 30.166982650756836,
+      "grad_norm": 30.537519454956055,
       "learning_rate": 9.5e-06,
-      "loss": 1.4172,
+      "loss": 1.417,
       "step": 475
     },
     {
       "epoch": 0.06034274680183442,
-      "grad_norm": 26.911828994750977,
+      "grad_norm": 26.885221481323242,
       "learning_rate": 1e-05,
-      "loss": 1.409,
+      "loss": 1.4094,
       "step": 500
     },
     {
       "epoch": 0.06034274680183442,
-      "eval_cer": 63.28322408864283,
-      "eval_loss": 1.46658456325531,
-      "eval_runtime": 3790.6529,
-      "eval_samples_per_second": 4.372,
-      "eval_steps_per_second": 0.547,
-      "eval_wer": 91.1192993172481,
+      "eval_cer": 63.82253332025917,
+      "eval_loss": 1.4668316841125488,
+      "eval_runtime": 4839.8999,
+      "eval_samples_per_second": 3.424,
+      "eval_steps_per_second": 0.428,
+      "eval_wer": 91.40993684791283,
       "step": 500
     },
     {
       "epoch": 0.06335988414192614,
-      "grad_norm": 29.040372848510742,
+      "grad_norm": 28.894699096679688,
       "learning_rate": 9.944444444444445e-06,
-      "loss": 1.5303,
+      "loss": 1.5314,
       "step": 525
     },
     {
       "epoch": 0.06637702148201786,
-      "grad_norm": 24.52021598815918,
+      "grad_norm": 24.584243774414062,
       "learning_rate": 9.88888888888889e-06,
-      "loss": 1.3872,
+      "loss": 1.3876,
       "step": 550
     },
     {
       "epoch": 0.06939415882210959,
-      "grad_norm": 23.997676849365234,
+      "grad_norm": 23.92827033996582,
       "learning_rate": 9.833333333333333e-06,
-      "loss": 1.3784,
+      "loss": 1.3787,
       "step": 575
     },
     {
       "epoch": 0.07241129616220131,
-      "grad_norm": 22.321500778198242,
+      "grad_norm": 22.209672927856445,
       "learning_rate": 9.777777777777779e-06,
-      "loss": 1.5472,
+      "loss": 1.5477,
       "step": 600
     },
     {
       "epoch": 0.07542843350229303,
-      "grad_norm": 24.63616943359375,
+      "grad_norm": 24.50571632385254,
       "learning_rate": 9.722222222222223e-06,
-      "loss": 1.337,
+      "loss": 1.3358,
       "step": 625
     },
     {
       "epoch": 0.07844557084238475,
-      "grad_norm": 26.10930824279785,
+      "grad_norm": 25.96898078918457,
       "learning_rate": 9.666666666666667e-06,
-      "loss": 1.4113,
+      "loss": 1.4117,
       "step": 650
     },
     {
       "epoch": 0.08146270818247647,
-      "grad_norm": 24.377735137939453,
+      "grad_norm": 24.520370483398438,
       "learning_rate": 9.611111111111112e-06,
-      "loss": 1.4169,
+      "loss": 1.4163,
       "step": 675
     },
     {
       "epoch": 0.08447984552256818,
-      "grad_norm": 28.207748413085938,
+      "grad_norm": 28.337772369384766,
       "learning_rate": 9.555555555555556e-06,
-      "loss": 1.4385,
+      "loss": 1.4381,
       "step": 700
     },
     {
       "epoch": 0.0874969828626599,
-      "grad_norm": 22.47800636291504,
+      "grad_norm": 22.536033630371094,
       "learning_rate": 9.5e-06,
-      "loss": 1.4014,
+      "loss": 1.4022,
       "step": 725
     },
     {
       "epoch": 0.09051412020275162,
-      "grad_norm": 23.965164184570312,
+      "grad_norm": 23.963077545166016,
       "learning_rate": 9.444444444444445e-06,
-      "loss": 1.4844,
+      "loss": 1.4845,
       "step": 750
     },
     {
       "epoch": 0.09051412020275162,
-      "eval_cer": 61.70583887099572,
-      "eval_loss": 1.4198288917541504,
-      "eval_runtime": 3756.7708,
-      "eval_samples_per_second": 4.411,
-      "eval_steps_per_second": 0.552,
-      "eval_wer": 89.3991192205003,
+      "eval_cer": 61.253460253286775,
+      "eval_loss": 1.4192742109298706,
+      "eval_runtime": 4048.8139,
+      "eval_samples_per_second": 4.093,
+      "eval_steps_per_second": 0.512,
+      "eval_wer": 89.38631146491167,
       "step": 750
     },
     {
       "epoch": 0.09353125754284335,
-      "grad_norm": 21.481678009033203,
+      "grad_norm": 21.442434310913086,
       "learning_rate": 9.38888888888889e-06,
-      "loss": 1.3299,
+      "loss": 1.3301,
       "step": 775
     },
     {
       "epoch": 0.09654839488293507,
-      "grad_norm": 18.25922393798828,
+      "grad_norm": 18.17827606201172,
       "learning_rate": 9.333333333333334e-06,
-      "loss": 1.382,
+      "loss": 1.3823,
       "step": 800
     },
     {
       "epoch": 0.09956553222302679,
-      "grad_norm": 23.341800689697266,
+      "grad_norm": 23.526996612548828,
       "learning_rate": 9.277777777777778e-06,
-      "loss": 1.5862,
+      "loss": 1.5849,
       "step": 825
     },
     {
       "epoch": 0.10258266956311851,
-      "grad_norm": 21.80385971069336,
+      "grad_norm": 21.815263748168945,
       "learning_rate": 9.222222222222224e-06,
-      "loss": 1.3633,
+      "loss": 1.3643,
       "step": 850
     },
     {
       "epoch": 0.10559980690321023,
-      "grad_norm": 21.196138381958008,
+      "grad_norm": 21.027591705322266,
       "learning_rate": 9.166666666666666e-06,
-      "loss": 1.373,
+      "loss": 1.3723,
       "step": 875
     },
     {
       "epoch": 0.10861694424330195,
-      "grad_norm": 26.68990135192871,
+      "grad_norm": 26.622665405273438,
       "learning_rate": 9.111111111111112e-06,
-      "loss": 1.4125,
+      "loss": 1.4118,
       "step": 900
     },
     {
       "epoch": 0.11163408158339368,
-      "grad_norm": 25.68580436706543,
+      "grad_norm": 25.46664047241211,
       "learning_rate": 9.055555555555556e-06,
-      "loss": 1.3544,
+      "loss": 1.3537,
       "step": 925
     },
     {
       "epoch": 0.1146512189234854,
-      "grad_norm": 21.313331604003906,
+      "grad_norm": 21.33067512512207,
       "learning_rate": 9e-06,
-      "loss": 1.3692,
+      "loss": 1.3697,
       "step": 950
     },
     {
       "epoch": 0.11766835626357712,
-      "grad_norm": 26.997135162353516,
+      "grad_norm": 27.155698776245117,
       "learning_rate": 8.944444444444446e-06,
-      "loss": 1.4534,
+      "loss": 1.4537,
       "step": 975
     },
     {
       "epoch": 0.12068549360366884,
-      "grad_norm": 26.953454971313477,
+      "grad_norm": 27.02322769165039,
       "learning_rate": 8.888888888888888e-06,
-      "loss": 1.4644,
+      "loss": 1.4639,
       "step": 1000
     },
     {
       "epoch": 0.12068549360366884,
-      "eval_cer": 61.789563800659245,
-      "eval_loss": 1.3863773345947266,
-      "eval_runtime": 3890.4048,
-      "eval_samples_per_second": 4.26,
-      "eval_steps_per_second": 0.533,
-      "eval_wer": 90.75083004108335,
+      "eval_cer": 62.40448946649872,
+      "eval_loss": 1.3861624002456665,
+      "eval_runtime": 6536.9994,
+      "eval_samples_per_second": 2.535,
+      "eval_steps_per_second": 0.317,
+      "eval_wer": 91.8197850267485,
       "step": 1000
     },
     {
       "epoch": 0.12370263094376056,
-      "grad_norm": 24.120258331298828,
+      "grad_norm": 24.13388442993164,
       "learning_rate": 8.833333333333334e-06,
-      "loss": 1.4748,
+      "loss": 1.4738,
       "step": 1025
     },
     {
       "epoch": 0.12671976828385229,
-      "grad_norm": 26.698230743408203,
+      "grad_norm": 26.072269439697266,
       "learning_rate": 8.777777777777778e-06,
-      "loss": 1.4389,
+      "loss": 1.4402,
       "step": 1050
     },
     {
       "epoch": 0.129736905623944,
-      "grad_norm": 31.574880599975586,
+      "grad_norm": 31.587852478027344,
       "learning_rate": 8.722222222222224e-06,
-      "loss": 1.3504,
+      "loss": 1.3485,
       "step": 1075
     },
     {
       "epoch": 0.13275404296403573,
-      "grad_norm": 23.17409324645996,
+      "grad_norm": 23.130081176757812,
       "learning_rate": 8.666666666666668e-06,
-      "loss": 1.3387,
+      "loss": 1.3384,
       "step": 1100
     },
     {
       "epoch": 0.13577118030412744,
-      "grad_norm": 27.488399505615234,
+      "grad_norm": 27.463407516479492,
       "learning_rate": 8.611111111111112e-06,
-      "loss": 1.4531,
+      "loss": 1.4537,
       "step": 1125
     },
     {
       "epoch": 0.13878831764421917,
-      "grad_norm": 20.899181365966797,
+      "grad_norm": 20.881338119506836,
       "learning_rate": 8.555555555555556e-06,
       "loss": 1.3062,
       "step": 1150
     },
     {
       "epoch": 0.14180545498431088,
-      "grad_norm": 26.456336975097656,
+      "grad_norm": 26.432994842529297,
       "learning_rate": 8.5e-06,
-      "loss": 1.399,
+      "loss": 1.3982,
       "step": 1175
     },
     {
       "epoch": 0.14482259232440262,
-      "grad_norm": 18.58867645263672,
+      "grad_norm": 18.55461311340332,
       "learning_rate": 8.444444444444446e-06,
       "loss": 1.2873,
       "step": 1200
     },
     {
       "epoch": 0.14783972966449432,
-      "grad_norm": 23.831193923950195,
+      "grad_norm": 23.862037658691406,
       "learning_rate": 8.38888888888889e-06,
-      "loss": 1.3038,
+      "loss": 1.3043,
       "step": 1225
     },
     {
       "epoch": 0.15085686700458606,
-      "grad_norm": 27.28873634338379,
+      "grad_norm": 27.1133975982666,
       "learning_rate": 8.333333333333334e-06,
-      "loss": 1.386,
+      "loss": 1.3855,
       "step": 1250
     },
     {
       "epoch": 0.15085686700458606,
-      "eval_cer": 66.82154521523341,
-      "eval_loss": 1.3677594661712646,
-      "eval_runtime": 4861.047,
-      "eval_samples_per_second": 3.409,
-      "eval_steps_per_second": 0.426,
-      "eval_wer": 93.2207564457493,
+      "eval_cer": 66.67634657595208,
+      "eval_loss": 1.367380976676941,
+      "eval_runtime": 5424.8274,
+      "eval_samples_per_second": 3.055,
+      "eval_steps_per_second": 0.382,
+      "eval_wer": 93.20499305425562,
       "step": 1250
     },
     {
       "epoch": 0.15387400434467777,
-      "grad_norm": 24.41499137878418,
+      "grad_norm": 24.38422393798828,
       "learning_rate": 8.277777777777778e-06,
-      "loss": 1.4668,
+      "loss": 1.4679,
       "step": 1275
     },
     {
       "epoch": 0.1568911416847695,
-      "grad_norm": 22.846540451049805,
+      "grad_norm": 23.12870216369629,
       "learning_rate": 8.222222222222222e-06,
-      "loss": 1.4219,
+      "loss": 1.4213,
       "step": 1300
     },
     {
       "epoch": 0.1599082790248612,
-      "grad_norm": 23.818105697631836,
+      "grad_norm": 24.16248321533203,
       "learning_rate": 8.166666666666668e-06,
-      "loss": 1.2927,
+      "loss": 1.2921,
       "step": 1325
     },
     {
       "epoch": 0.16292541636495295,
-      "grad_norm": 22.695980072021484,
+      "grad_norm": 22.89928436279297,
       "learning_rate": 8.111111111111112e-06,
-      "loss": 1.3731,
+      "loss": 1.3745,
       "step": 1350
     },
     {
       "epoch": 0.16594255370504465,
-      "grad_norm": 26.142614364624023,
+      "grad_norm": 26.103015899658203,
       "learning_rate": 8.055555555555557e-06,
-      "loss": 1.3787,
+      "loss": 1.3796,
       "step": 1375
     },
     {
       "epoch": 0.16895969104513636,
-      "grad_norm": 17.853261947631836,
+      "grad_norm": 17.778417587280273,
       "learning_rate": 8.000000000000001e-06,
-      "loss": 1.3649,
+      "loss": 1.3638,
       "step": 1400
     },
     {
       "epoch": 0.1719768283852281,
-      "grad_norm": 23.05394172668457,
+      "grad_norm": 23.057931900024414,
       "learning_rate": 7.944444444444445e-06,
-      "loss": 1.3517,
+      "loss": 1.3524,
       "step": 1425
     },
     {
       "epoch": 0.1749939657253198,
-      "grad_norm": 20.292495727539062,
+      "grad_norm": 20.255752563476562,
       "learning_rate": 7.88888888888889e-06,
-      "loss": 1.3752,
+      "loss": 1.3768,
       "step": 1450
     },
     {
       "epoch": 0.17801110306541154,
-      "grad_norm": 19.236572265625,
+      "grad_norm": 19.22992706298828,
       "learning_rate": 7.833333333333333e-06,
-      "loss": 1.2467,
+      "loss": 1.247,
       "step": 1475
     },
     {
       "epoch": 0.18102824040550325,
-      "grad_norm": 23.87392807006836,
+      "grad_norm": 23.74711036682129,
       "learning_rate": 7.77777777777778e-06,
-      "loss": 1.3727,
+      "loss": 1.3741,
       "step": 1500
     },
     {
       "epoch": 0.18102824040550325,
-      "eval_cer": 58.00949636814832,
-      "eval_loss": 1.350059151649475,
-      "eval_runtime": 3818.8115,
-      "eval_samples_per_second": 4.34,
-      "eval_steps_per_second": 0.543,
-      "eval_wer": 83.93020758416174,
+      "eval_cer": 58.32365342404792,
+      "eval_loss": 1.3499553203582764,
+      "eval_runtime": 4123.7168,
+      "eval_samples_per_second": 4.019,
+      "eval_steps_per_second": 0.502,
+      "eval_wer": 83.93414843203514,
       "step": 1500
     },
     {
       "epoch": 0.18404537774559498,
-      "grad_norm": 22.543930053710938,
+      "grad_norm": 22.509279251098633,
       "learning_rate": 7.722222222222223e-06,
-      "loss": 1.2917,
+      "loss": 1.2904,
       "step": 1525
     },
     {
       "epoch": 0.1870625150856867,
-      "grad_norm": 22.561914443969727,
+      "grad_norm": 22.517881393432617,
       "learning_rate": 7.666666666666667e-06,
-      "loss": 1.2996,
+      "loss": 1.3,
       "step": 1550
     },
     {
       "epoch": 0.19007965242577843,
-      "grad_norm": 26.33420753479004,
+      "grad_norm": 26.30403709411621,
       "learning_rate": 7.611111111111111e-06,
-      "loss": 1.3633,
+      "loss": 1.3627,
       "step": 1575
     },
     {
       "epoch": 0.19309678976587014,
-      "grad_norm": 20.257822036743164,
+      "grad_norm": 20.28595542907715,
       "learning_rate": 7.555555555555556e-06,
-      "loss": 1.2725,
+      "loss": 1.2716,
       "step": 1600
     },
     {
       "epoch": 0.19611392710596187,
-      "grad_norm": 22.469804763793945,
+      "grad_norm": 22.23461151123047,
       "learning_rate": 7.500000000000001e-06,
-      "loss": 1.4868,
+      "loss": 1.4848,
       "step": 1625
     },
     {
       "epoch": 0.19913106444605358,
-      "grad_norm": 23.334760665893555,
+      "grad_norm": 23.31128692626953,
       "learning_rate": 7.444444444444445e-06,
-      "loss": 1.2444,
+      "loss": 1.2453,
       "step": 1650
     },
     {
       "epoch": 0.20214820178614532,
-      "grad_norm": 23.603315353393555,
+      "grad_norm": 23.57061004638672,
       "learning_rate": 7.38888888888889e-06,
-      "loss": 1.359,
+      "loss": 1.3575,
       "step": 1675
     },
     {
       "epoch": 0.20516533912623702,
-      "grad_norm": 19.353214263916016,
+      "grad_norm": 19.259546279907227,
       "learning_rate": 7.333333333333333e-06,
-      "loss": 1.288,
+      "loss": 1.2894,
       "step": 1700
     },
     {
       "epoch": 0.20818247646632876,
-      "grad_norm": 25.300758361816406,
+      "grad_norm": 25.171783447265625,
       "learning_rate": 7.277777777777778e-06,
-      "loss": 1.3908,
+      "loss": 1.3894,
       "step": 1725
     },
     {
       "epoch": 0.21119961380642047,
-      "grad_norm": 21.219324111938477,
+      "grad_norm": 21.088159561157227,
       "learning_rate": 7.222222222222223e-06,
-      "loss": 1.3078,
+      "loss": 1.3073,
       "step": 1750
     },
     {
       "epoch": 0.21119961380642047,
-      "eval_cer": 59.00212706578064,
-      "eval_loss": 1.3268232345581055,
-      "eval_runtime": 4148.3039,
-      "eval_samples_per_second": 3.995,
-      "eval_steps_per_second": 0.499,
-      "eval_wer": 81.30954374833746,
+      "eval_cer": 58.877671089253795,
+      "eval_loss": 1.326649785041809,
+      "eval_runtime": 4478.8906,
+      "eval_samples_per_second": 3.7,
+      "eval_steps_per_second": 0.463,
+      "eval_wer": 81.65732357316676,
       "step": 1750
     },
     {
       "epoch": 0.2142167511465122,
-      "grad_norm": 23.480670928955078,
+      "grad_norm": 23.551097869873047,
       "learning_rate": 7.166666666666667e-06,
-      "loss": 1.32,
+      "loss": 1.3203,
       "step": 1775
     },
     {
       "epoch": 0.2172338884866039,
-      "grad_norm": 25.72469139099121,
+      "grad_norm": 25.782739639282227,
       "learning_rate": 7.111111111111112e-06,
-      "loss": 1.2747,
+      "loss": 1.2752,
       "step": 1800
     },
     {
       "epoch": 0.22025102582669562,
-      "grad_norm": 16.877225875854492,
+      "grad_norm": 16.879140853881836,
       "learning_rate": 7.055555555555557e-06,
-      "loss": 1.2142,
+      "loss": 1.2145,
       "step": 1825
     },
     {
       "epoch": 0.22326816316678735,
-      "grad_norm": 22.20556640625,
+      "grad_norm": 22.139205932617188,
       "learning_rate": 7e-06,
-      "loss": 1.2684,
+      "loss": 1.2678,
       "step": 1850
     },
     {
       "epoch": 0.22628530050687906,
-      "grad_norm": 15.339997291564941,
+      "grad_norm": 15.391493797302246,
       "learning_rate": 6.944444444444445e-06,
-      "loss": 1.2484,
+      "loss": 1.2468,
       "step": 1875
     },
     {
       "epoch": 0.2293024378469708,
-      "grad_norm": 20.877666473388672,
+      "grad_norm": 20.986146926879883,
       "learning_rate": 6.88888888888889e-06,
-      "loss": 1.2099,
+      "loss": 1.2093,
       "step": 1900
     },
     {
       "epoch": 0.2323195751870625,
-      "grad_norm": 28.011430740356445,
+      "grad_norm": 27.949708938598633,
       "learning_rate": 6.833333333333334e-06,
-      "loss": 1.2802,
+      "loss": 1.2795,
       "step": 1925
     },
     {
       "epoch": 0.23533671252715424,
-      "grad_norm": 22.03005027770996,
+      "grad_norm": 21.94474220275879,
       "learning_rate": 6.777777777777779e-06,
-      "loss": 1.2924,
+      "loss": 1.2931,
       "step": 1950
     },
     {
       "epoch": 0.23835384986724595,
-      "grad_norm": 22.320819854736328,
+      "grad_norm": 22.343372344970703,
       "learning_rate": 6.7222222222222235e-06,
-      "loss": 1.2741,
+      "loss": 1.2731,
       "step": 1975
     },
     {
       "epoch": 0.24137098720733768,
-      "grad_norm": 20.871091842651367,
+      "grad_norm": 21.14777374267578,
       "learning_rate": 6.666666666666667e-06,
-      "loss": 1.2995,
+      "loss": 1.2993,
       "step": 2000
     },
     {
       "epoch": 0.24137098720733768,
-      "eval_cer": 59.07340639779147,
-      "eval_loss": 1.3121392726898193,
-      "eval_runtime": 4204.9822,
-      "eval_samples_per_second": 3.941,
-      "eval_steps_per_second": 0.493,
-      "eval_wer": 86.45432064708723,
+      "eval_cer": 59.81580515474027,
+      "eval_loss": 1.3122555017471313,
+      "eval_runtime": 5034.5037,
+      "eval_samples_per_second": 3.292,
+      "eval_steps_per_second": 0.412,
+      "eval_wer": 87.38337553324598,
       "step": 2000
     },
     {
       "epoch": 0.2443881245474294,
-      "grad_norm": 23.991544723510742,
+      "grad_norm": 24.091772079467773,
       "learning_rate": 6.6111111111111115e-06,
-      "loss": 1.3492,
+      "loss": 1.3477,
       "step": 2025
     },
     {
       "epoch": 0.24740526188752113,
-      "grad_norm": 22.863008499145508,
+      "grad_norm": 22.88396453857422,
       "learning_rate": 6.555555555555556e-06,
-      "loss": 1.308,
+      "loss": 1.3089,
       "step": 2050
     },
     {
       "epoch": 0.25042239922761284,
-      "grad_norm": 17.017812728881836,
+      "grad_norm": 17.0950984954834,
       "learning_rate": 6.5000000000000004e-06,
-      "loss": 1.2589,
+      "loss": 1.2598,
       "step": 2075
     },
     {
       "epoch": 0.25343953656770457,
-      "grad_norm": 18.99051284790039,
+      "grad_norm": 19.088529586791992,
       "learning_rate": 6.444444444444445e-06,
-      "loss": 1.2262,
+      "loss": 1.2255,
       "step": 2100
     },
     {
       "epoch": 0.2564566739077963,
-      "grad_norm": 24.77310562133789,
+      "grad_norm": 24.771873474121094,
       "learning_rate": 6.3888888888888885e-06,
-      "loss": 1.2641,
+      "loss": 1.2642,
       "step": 2125
     },
     {
       "epoch": 0.259473811247888,
-      "grad_norm": 25.287763595581055,
+      "grad_norm": 25.191925048828125,
       "learning_rate": 6.333333333333333e-06,
-      "loss": 1.2945,
+      "loss": 1.2942,
       "step": 2150
     },
     {
       "epoch": 0.2624909485879797,
-      "grad_norm": 25.328964233398438,
+      "grad_norm": 25.375642776489258,
       "learning_rate": 6.277777777777778e-06,
-      "loss": 1.4174,
+      "loss": 1.4181,
       "step": 2175
     },
     {
       "epoch": 0.26550808592807146,
-      "grad_norm": 21.193706512451172,
+      "grad_norm": 21.1870059967041,
       "learning_rate": 6.222222222222223e-06,
-      "loss": 1.3045,
+      "loss": 1.304,
       "step": 2200
     },
     {
       "epoch": 0.2685252232681632,
-      "grad_norm": 25.26439094543457,
+      "grad_norm": 25.293983459472656,
       "learning_rate": 6.166666666666667e-06,
-      "loss": 1.2792,
+      "loss": 1.2795,
       "step": 2225
     },
     {
       "epoch": 0.2715423606082549,
-      "grad_norm": 18.658802032470703,
+      "grad_norm": 18.648513793945312,
       "learning_rate": 6.111111111111112e-06,
-      "loss": 1.3895,
+      "loss": 1.3909,
       "step": 2250
     },
     {
       "epoch": 0.2715423606082549,
-      "eval_cer": 69.0187589099165,
-      "eval_loss": 1.296238660812378,
-      "eval_runtime": 4556.544,
-      "eval_samples_per_second": 3.637,
-      "eval_steps_per_second": 0.455,
-      "eval_wer": 94.42961153092088,
+      "eval_cer": 69.14660914034863,
+      "eval_loss": 1.2956976890563965,
+      "eval_runtime": 4406.2547,
+      "eval_samples_per_second": 3.761,
+      "eval_steps_per_second": 0.47,
+      "eval_wer": 94.55374823893361,
       "step": 2250
     },
     {
       "epoch": 0.2745594979483466,
-      "grad_norm": 19.182533264160156,
+      "grad_norm": 19.018657684326172,
       "learning_rate": 6.055555555555555e-06,
-      "loss": 1.3048,
+      "loss": 1.3042,
       "step": 2275
     },
     {
       "epoch": 0.27757663528843834,
-      "grad_norm": 21.095041275024414,
+      "grad_norm": 21.156967163085938,
       "learning_rate": 6e-06,
-      "loss": 1.2783,
+      "loss": 1.2785,
       "step": 2300
     },
     {
       "epoch": 0.28059377262853,
-      "grad_norm": 18.291179656982422,
+      "grad_norm": 18.3033390045166,
       "learning_rate": 5.944444444444445e-06,
-      "loss": 1.2607,
+      "loss": 1.2613,
       "step": 2325
     },
     {
       "epoch": 0.28361090996862176,
-      "grad_norm": 16.70252227783203,
+      "grad_norm": 16.72675323486328,
       "learning_rate": 5.88888888888889e-06,
-      "loss": 1.2578,
+      "loss": 1.2575,
       "step": 2350
     },
     {
       "epoch": 0.2866280473087135,
-      "grad_norm": 27.941680908203125,
+      "grad_norm": 28.087432861328125,
       "learning_rate": 5.833333333333334e-06,
-      "loss": 1.3026,
+      "loss": 1.3014,
       "step": 2375
     },
     {
       "epoch": 0.28964518464880523,
-      "grad_norm": 26.314424514770508,
+      "grad_norm": 26.902196884155273,
       "learning_rate": 5.777777777777778e-06,
-      "loss": 1.2207,
+      "loss": 1.2205,
       "step": 2400
     },
     {
       "epoch": 0.2926623219888969,
-      "grad_norm": 23.71686553955078,
+      "grad_norm": 23.583770751953125,
       "learning_rate": 5.722222222222222e-06,
-      "loss": 1.2565,
+      "loss": 1.256,
       "step": 2425
     },
     {
       "epoch": 0.29567945932898865,
-      "grad_norm": 24.66400909423828,
+      "grad_norm": 24.659427642822266,
       "learning_rate": 5.666666666666667e-06,
-      "loss": 1.3363,
+      "loss": 1.3358,
       "step": 2450
     },
     {
       "epoch": 0.2986965966690804,
-      "grad_norm": 24.745983123779297,
+      "grad_norm": 24.409543991088867,
       "learning_rate": 5.611111111111112e-06,
-      "loss": 1.2785,
+      "loss": 1.2786,
       "step": 2475
     },
     {
       "epoch": 0.3017137340091721,
-      "grad_norm": 23.765567779541016,
+      "grad_norm": 23.712934494018555,
       "learning_rate": 5.555555555555557e-06,
-      "loss": 1.3658,
+      "loss": 1.3662,
       "step": 2500
     },
     {
       "epoch": 0.3017137340091721,
-      "eval_cer": 60.034545961969265,
-      "eval_loss": 1.2828303575515747,
-      "eval_runtime": 4164.8499,
-      "eval_samples_per_second": 3.979,
-      "eval_steps_per_second": 0.497,
-      "eval_wer": 84.99325129801677,
+      "eval_cer": 59.539550600782945,
+      "eval_loss": 1.282615065574646,
+      "eval_runtime": 5034.6376,
+      "eval_samples_per_second": 3.292,
+      "eval_steps_per_second": 0.412,
+      "eval_wer": 84.68093910404824,
       "step": 2500
     }
   ],

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4a4d80345da3e67ae6f5163b9180309d128ac7168dccb87fb5e3fadd8ad567e3
+oid sha256:90898e8dbc5dd08c90889825743818b940619f4222335a58bdeaa3938a60ec53
 size 5240