Upload folder using huggingface_hub

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:98b5351d1c47878fce2ad2e0bc59debe3426c9b85877afc025ca3e9701f7f37d
+oid sha256:9097c61a1bf62f41b785fd99444b65e1bd76f0ffc283e26f894778f974e1662a
 size 3850418936

optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2d3a60d01b455629a1995c3c561bca7253aa20d0799c4b29988b03ac23c443a7
+oid sha256:5a7cc137780b53b89de515619095496b01e227cc0c15b62036a2862bebedb484
 size 7667614690

rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ae2f37179b86cb15a864fd9a45b7ce57f632f1e336a9d02cc3f0b1070560b9b0
+oid sha256:389a881c9b3434f5c74e7c9188f70a420d1725de2e16866aa64812ede5a6e33b
 size 14308

scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:59aa2dbd53827a3d48120301c0684ae98deb83b3a662828bb8005dd44b824861
+oid sha256:5c2c3700ded85b44dac6a50f7f025d883b5370a3952247e3358b8141818a8cd7
 size 1064

trainer_state.json

@@ -1,213 +1,405 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 48.87983706720978,
+  "epoch": 48.92966360856269,
   "eval_steps": 500,
-  "global_step": 6000,
+  "global_step": 12000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 4.07,
-      "grad_norm": 3.8863396644592285,
-      "learning_rate": 3.936e-05,
-      "loss": 3.9012,
+      "epoch": 2.04,
+      "grad_norm": 5.49405574798584,
+      "learning_rate": 3.944e-05,
+      "loss": 5.2628,
       "step": 500
     },
     {
-      "epoch": 4.07,
-      "eval_loss": 0.4798698425292969,
-      "eval_runtime": 164.8654,
-      "eval_samples_per_second": 19.234,
-      "eval_steps_per_second": 2.408,
-      "eval_wer": 0.46691878014085975,
+      "epoch": 2.04,
+      "eval_loss": 2.973829507827759,
+      "eval_runtime": 162.2636,
+      "eval_samples_per_second": 19.542,
+      "eval_steps_per_second": 2.447,
+      "eval_wer": 0.9875099746730042,
       "step": 500
     },
     {
-      "epoch": 8.15,
-      "grad_norm": 1.7655788660049438,
-      "learning_rate": 7.936000000000001e-05,
-      "loss": 0.7087,
+      "epoch": 4.08,
+      "grad_norm": 1.7545195817947388,
+      "learning_rate": 7.944000000000001e-05,
+      "loss": 1.4609,
       "step": 1000
     },
     {
-      "epoch": 8.15,
-      "eval_loss": 0.331632524728775,
-      "eval_runtime": 165.1487,
-      "eval_samples_per_second": 19.201,
-      "eval_steps_per_second": 2.404,
-      "eval_wer": 0.3308121985914027,
+      "epoch": 4.08,
+      "eval_loss": 0.4073050916194916,
+      "eval_runtime": 161.9691,
+      "eval_samples_per_second": 19.578,
+      "eval_steps_per_second": 2.451,
+      "eval_wer": 0.4208097699753669,
       "step": 1000
     },
     {
-      "epoch": 12.22,
-      "grad_norm": 2.021256446838379,
-      "learning_rate": 7.228235294117648e-05,
-      "loss": 0.3293,
+      "epoch": 6.12,
+      "grad_norm": 2.6881394386291504,
+      "learning_rate": 7.649422222222223e-05,
+      "loss": 0.6445,
       "step": 1500
     },
     {
-      "epoch": 12.22,
-      "eval_loss": 0.29065316915512085,
-      "eval_runtime": 168.1265,
-      "eval_samples_per_second": 18.861,
-      "eval_steps_per_second": 2.361,
-      "eval_wer": 0.290982895604205,
+      "epoch": 6.12,
+      "eval_loss": 0.3578987717628479,
+      "eval_runtime": 163.2467,
+      "eval_samples_per_second": 19.425,
+      "eval_steps_per_second": 2.432,
+      "eval_wer": 0.3558269437601915,
       "step": 1500
     },
     {
-      "epoch": 16.29,
-      "grad_norm": 1.2028392553329468,
-      "learning_rate": 6.443921568627452e-05,
-      "loss": 0.203,
+      "epoch": 8.15,
+      "grad_norm": 1.934047818183899,
+      "learning_rate": 7.293866666666667e-05,
+      "loss": 0.3683,
       "step": 2000
     },
     {
-      "epoch": 16.29,
-      "eval_loss": 0.29165971279144287,
-      "eval_runtime": 168.0318,
-      "eval_samples_per_second": 18.871,
-      "eval_steps_per_second": 2.363,
-      "eval_wer": 0.2812337369461888,
+      "epoch": 8.15,
+      "eval_loss": 0.3222399353981018,
+      "eval_runtime": 167.0586,
+      "eval_samples_per_second": 18.981,
+      "eval_steps_per_second": 2.376,
+      "eval_wer": 0.32130590153696703,
       "step": 2000
     },
     {
-      "epoch": 20.37,
-      "grad_norm": 0.8678894639015198,
-      "learning_rate": 5.6596078431372547e-05,
-      "loss": 0.1526,
+      "epoch": 10.19,
+      "grad_norm": 4.5025835037231445,
+      "learning_rate": 6.938311111111111e-05,
+      "loss": 0.2758,
       "step": 2500
     },
     {
-      "epoch": 20.37,
-      "eval_loss": 0.2595488429069519,
-      "eval_runtime": 168.4253,
-      "eval_samples_per_second": 18.827,
-      "eval_steps_per_second": 2.357,
-      "eval_wer": 0.26881310064878744,
+      "epoch": 10.19,
+      "eval_loss": 0.3258918523788452,
+      "eval_runtime": 164.2307,
+      "eval_samples_per_second": 19.308,
+      "eval_steps_per_second": 2.417,
+      "eval_wer": 0.316032335287791,
       "step": 2500
     },
     {
-      "epoch": 24.44,
-      "grad_norm": 1.7510905265808105,
-      "learning_rate": 4.875294117647059e-05,
-      "loss": 0.1189,
+      "epoch": 12.23,
+      "grad_norm": 1.0024570226669312,
+      "learning_rate": 6.582755555555557e-05,
+      "loss": 0.223,
       "step": 3000
     },
     {
-      "epoch": 24.44,
-      "eval_loss": 0.2896680533885956,
-      "eval_runtime": 167.8147,
-      "eval_samples_per_second": 18.896,
-      "eval_steps_per_second": 2.366,
-      "eval_wer": 0.2652395656246747,
+      "epoch": 12.23,
+      "eval_loss": 0.3127504587173462,
+      "eval_runtime": 162.8183,
+      "eval_samples_per_second": 19.476,
+      "eval_steps_per_second": 2.438,
+      "eval_wer": 0.30562398084862785,
       "step": 3000
     },
     {
-      "epoch": 28.51,
-      "grad_norm": 1.1846829652786255,
-      "learning_rate": 4.090980392156863e-05,
-      "loss": 0.0988,
+      "epoch": 14.27,
+      "grad_norm": 0.9843519330024719,
+      "learning_rate": 6.2272e-05,
+      "loss": 0.1881,
       "step": 3500
     },
     {
-      "epoch": 28.51,
-      "eval_loss": 0.30395135283470154,
-      "eval_runtime": 168.5254,
-      "eval_samples_per_second": 18.816,
-      "eval_steps_per_second": 2.356,
-      "eval_wer": 0.2631925892516393,
+      "epoch": 14.27,
+      "eval_loss": 0.3383229076862335,
+      "eval_runtime": 163.7812,
+      "eval_samples_per_second": 19.361,
+      "eval_steps_per_second": 2.424,
+      "eval_wer": 0.29955244075911597,
       "step": 3500
     },
     {
-      "epoch": 32.59,
-      "grad_norm": 1.2768597602844238,
-      "learning_rate": 3.3066666666666666e-05,
-      "loss": 0.0767,
+      "epoch": 16.31,
+      "grad_norm": 1.825391411781311,
+      "learning_rate": 5.871644444444445e-05,
+      "loss": 0.1676,
       "step": 4000
     },
     {
-      "epoch": 32.59,
-      "eval_loss": 0.2856810986995697,
-      "eval_runtime": 170.6385,
-      "eval_samples_per_second": 18.583,
-      "eval_steps_per_second": 2.327,
-      "eval_wer": 0.25576796308503624,
+      "epoch": 16.31,
+      "eval_loss": 0.33344313502311707,
+      "eval_runtime": 163.1954,
+      "eval_samples_per_second": 19.431,
+      "eval_steps_per_second": 2.433,
+      "eval_wer": 0.30361169899038964,
       "step": 4000
     },
     {
-      "epoch": 36.66,
-      "grad_norm": 0.8570753931999207,
-      "learning_rate": 2.5223529411764707e-05,
-      "loss": 0.0653,
+      "epoch": 18.35,
+      "grad_norm": 1.3528616428375244,
+      "learning_rate": 5.5160888888888894e-05,
+      "loss": 0.1405,
       "step": 4500
     },
     {
-      "epoch": 36.66,
-      "eval_loss": 0.30118483304977417,
-      "eval_runtime": 171.6141,
-      "eval_samples_per_second": 18.478,
-      "eval_steps_per_second": 2.313,
-      "eval_wer": 0.24785761371127224,
+      "epoch": 18.35,
+      "eval_loss": 0.31019264459609985,
+      "eval_runtime": 164.5816,
+      "eval_samples_per_second": 19.267,
+      "eval_steps_per_second": 2.412,
+      "eval_wer": 0.2862644415917843,
       "step": 4500
     },
     {
-      "epoch": 40.73,
-      "grad_norm": 0.5461006164550781,
-      "learning_rate": 1.7380392156862745e-05,
-      "loss": 0.0585,
+      "epoch": 20.39,
+      "grad_norm": 1.929638147354126,
+      "learning_rate": 5.160533333333334e-05,
+      "loss": 0.1251,
       "step": 5000
     },
     {
-      "epoch": 40.73,
-      "eval_loss": 0.28294435143470764,
-      "eval_runtime": 169.3236,
-      "eval_samples_per_second": 18.727,
-      "eval_steps_per_second": 2.345,
-      "eval_wer": 0.24251465843250183,
+      "epoch": 20.39,
+      "eval_loss": 0.33736076951026917,
+      "eval_runtime": 163.4383,
+      "eval_samples_per_second": 19.402,
+      "eval_steps_per_second": 2.429,
+      "eval_wer": 0.29060125594143565,
       "step": 5000
     },
     {
-      "epoch": 44.81,
-      "grad_norm": 0.47187602519989014,
-      "learning_rate": 9.537254901960786e-06,
-      "loss": 0.0467,
+      "epoch": 22.43,
+      "grad_norm": 0.8257089257240295,
+      "learning_rate": 4.804977777777778e-05,
+      "loss": 0.1169,
       "step": 5500
     },
     {
-      "epoch": 44.81,
-      "eval_loss": 0.28480401635169983,
-      "eval_runtime": 169.9442,
-      "eval_samples_per_second": 18.659,
-      "eval_steps_per_second": 2.336,
-      "eval_wer": 0.23873295631960587,
+      "epoch": 22.43,
+      "eval_loss": 0.34548866748809814,
+      "eval_runtime": 163.1624,
+      "eval_samples_per_second": 19.435,
+      "eval_steps_per_second": 2.433,
+      "eval_wer": 0.28893591923116957,
       "step": 5500
     },
     {
-      "epoch": 48.88,
-      "grad_norm": 0.4703833758831024,
-      "learning_rate": 1.6941176470588237e-06,
-      "loss": 0.0416,
+      "epoch": 24.46,
+      "grad_norm": 0.36999279260635376,
+      "learning_rate": 4.449422222222222e-05,
+      "loss": 0.1021,
       "step": 6000
     },
     {
-      "epoch": 48.88,
-      "eval_loss": 0.27989715337753296,
-      "eval_runtime": 170.4594,
-      "eval_samples_per_second": 18.603,
-      "eval_steps_per_second": 2.329,
-      "eval_wer": 0.23436144745515733,
+      "epoch": 24.46,
+      "eval_loss": 0.302827388048172,
+      "eval_runtime": 163.4863,
+      "eval_samples_per_second": 19.396,
+      "eval_steps_per_second": 2.428,
+      "eval_wer": 0.28237865593449674,
       "step": 6000
+    },
+    {
+      "epoch": 26.5,
+      "grad_norm": 1.7563804388046265,
+      "learning_rate": 4.0938666666666675e-05,
+      "loss": 0.0907,
+      "step": 6500
+    },
+    {
+      "epoch": 26.5,
+      "eval_loss": 0.3536182641983032,
+      "eval_runtime": 162.4668,
+      "eval_samples_per_second": 19.518,
+      "eval_steps_per_second": 2.444,
+      "eval_wer": 0.28036637407625853,
+      "step": 6500
+    },
+    {
+      "epoch": 28.54,
+      "grad_norm": 0.9021581411361694,
+      "learning_rate": 3.7383111111111114e-05,
+      "loss": 0.0834,
+      "step": 7000
+    },
+    {
+      "epoch": 28.54,
+      "eval_loss": 0.31954678893089294,
+      "eval_runtime": 165.1008,
+      "eval_samples_per_second": 19.206,
+      "eval_steps_per_second": 2.405,
+      "eval_wer": 0.2742601394719495,
+      "step": 7000
+    },
+    {
+      "epoch": 30.58,
+      "grad_norm": 12.496235847473145,
+      "learning_rate": 3.382755555555556e-05,
+      "loss": 0.0806,
+      "step": 7500
+    },
+    {
+      "epoch": 30.58,
+      "eval_loss": 0.35398995876312256,
+      "eval_runtime": 167.4696,
+      "eval_samples_per_second": 18.935,
+      "eval_steps_per_second": 2.371,
+      "eval_wer": 0.275162196856677,
+      "step": 7500
+    },
+    {
+      "epoch": 32.62,
+      "grad_norm": 0.7885499596595764,
+      "learning_rate": 3.0272000000000004e-05,
+      "loss": 0.0718,
+      "step": 8000
+    },
+    {
+      "epoch": 32.62,
+      "eval_loss": 0.34760990738868713,
+      "eval_runtime": 166.6604,
+      "eval_samples_per_second": 19.027,
+      "eval_steps_per_second": 2.382,
+      "eval_wer": 0.2699927141518926,
+      "step": 8000
+    },
+    {
+      "epoch": 34.66,
+      "grad_norm": 0.8713880777359009,
+      "learning_rate": 2.6716444444444446e-05,
+      "loss": 0.064,
+      "step": 8500
+    },
+    {
+      "epoch": 34.66,
+      "eval_loss": 0.32806122303009033,
+      "eval_runtime": 166.5309,
+      "eval_samples_per_second": 19.042,
+      "eval_steps_per_second": 2.384,
+      "eval_wer": 0.26787634874926275,
+      "step": 8500
+    },
+    {
+      "epoch": 36.7,
+      "grad_norm": 1.410463809967041,
+      "learning_rate": 2.316088888888889e-05,
+      "loss": 0.0575,
+      "step": 9000
+    },
+    {
+      "epoch": 36.7,
+      "eval_loss": 0.3376013934612274,
+      "eval_runtime": 165.8955,
+      "eval_samples_per_second": 19.114,
+      "eval_steps_per_second": 2.393,
+      "eval_wer": 0.26086805676022623,
+      "step": 9000
+    },
+    {
+      "epoch": 38.74,
+      "grad_norm": 0.974173367023468,
+      "learning_rate": 1.9605333333333334e-05,
+      "loss": 0.0557,
+      "step": 9500
+    },
+    {
+      "epoch": 38.74,
+      "eval_loss": 0.30982139706611633,
+      "eval_runtime": 154.8221,
+      "eval_samples_per_second": 20.482,
+      "eval_steps_per_second": 2.564,
+      "eval_wer": 0.25996599937549875,
+      "step": 9500
+    },
+    {
+      "epoch": 40.77,
+      "grad_norm": 1.6050668954849243,
+      "learning_rate": 1.604977777777778e-05,
+      "loss": 0.0498,
+      "step": 10000
+    },
+    {
+      "epoch": 40.77,
+      "eval_loss": 0.3348703980445862,
+      "eval_runtime": 166.7142,
+      "eval_samples_per_second": 19.021,
+      "eval_steps_per_second": 2.381,
+      "eval_wer": 0.25809249557644937,
+      "step": 10000
+    },
+    {
+      "epoch": 42.81,
+      "grad_norm": 1.3856664896011353,
+      "learning_rate": 1.2494222222222223e-05,
+      "loss": 0.0437,
+      "step": 10500
+    },
+    {
+      "epoch": 42.81,
+      "eval_loss": 0.3345808684825897,
+      "eval_runtime": 168.0265,
+      "eval_samples_per_second": 18.872,
+      "eval_steps_per_second": 2.363,
+      "eval_wer": 0.25479651667071435,
+      "step": 10500
+    },
+    {
+      "epoch": 44.85,
+      "grad_norm": 0.4344525933265686,
+      "learning_rate": 8.938666666666668e-06,
+      "loss": 0.0394,
+      "step": 11000
+    },
+    {
+      "epoch": 44.85,
+      "eval_loss": 0.32588261365890503,
+      "eval_runtime": 169.5922,
+      "eval_samples_per_second": 18.698,
+      "eval_steps_per_second": 2.341,
+      "eval_wer": 0.25382507025639245,
+      "step": 11000
+    },
+    {
+      "epoch": 46.89,
+      "grad_norm": 0.6624104976654053,
+      "learning_rate": 5.383111111111112e-06,
+      "loss": 0.0379,
+      "step": 11500
+    },
+    {
+      "epoch": 46.89,
+      "eval_loss": 0.32103830575942993,
+      "eval_runtime": 172.4868,
+      "eval_samples_per_second": 18.384,
+      "eval_steps_per_second": 2.302,
+      "eval_wer": 0.2514311487353849,
+      "step": 11500
+    },
+    {
+      "epoch": 48.93,
+      "grad_norm": 0.8061400055885315,
+      "learning_rate": 1.8275555555555557e-06,
+      "loss": 0.0369,
+      "step": 12000
+    },
+    {
+      "epoch": 48.93,
+      "eval_loss": 0.3121373653411865,
+      "eval_runtime": 170.2978,
+      "eval_samples_per_second": 18.62,
+      "eval_steps_per_second": 2.331,
+      "eval_wer": 0.24886375464039134,
+      "step": 12000
     }
   ],
   "logging_steps": 500,
-  "max_steps": 6100,
+  "max_steps": 12250,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 50,
   "save_steps": 500,
-  "total_flos": 9.05423649471038e+19,
-  "train_batch_size": 8,
+  "total_flos": 9.026756828474207e+19,
+  "train_batch_size": 4,
   "trial_name": null,
   "trial_params": null
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ef2aea22fc3a5c26f0bab574396403ce2f94a120364ec03de87664a530be41aa
+oid sha256:a6ad149fe7c00889691db46a48bd808291f5241258d36d9c9ececad26b1dab1f
 size 4920