End of training

all_results.json

@@ -1,15 +1,15 @@
 {
-    "epoch": 2.0,
-    "eval_cer": 0.9555030929683531,
-    "eval_loss": 3.2690865993499756,
-    "eval_runtime": 316.6745,
+    "epoch": 100.0,
+    "eval_cer": 0.12175835997967904,
+    "eval_loss": 0.33164310455322266,
+    "eval_runtime": 269.3321,
     "eval_samples": 4470,
-    "eval_samples_per_second": 14.115,
-    "eval_steps_per_second": 1.765,
-    "eval_wer": 0.9960036558484615,
-    "train_loss": 6.151403113256527,
-    "train_runtime": 1933.761,
+    "eval_samples_per_second": 16.597,
+    "eval_steps_per_second": 2.076,
+    "eval_wer": 0.2563574129495887,
+    "train_loss": 1.2220959588545788,
+    "train_runtime": 95146.6811,
     "train_samples": 10156,
-    "train_samples_per_second": 10.504,
-    "train_steps_per_second": 0.163
+    "train_samples_per_second": 10.674,
+    "train_steps_per_second": 0.166
 }

eval_results.json

@@ -1,10 +1,10 @@
 {
-    "epoch": 2.0,
-    "eval_cer": 0.9555030929683531,
-    "eval_loss": 3.2690865993499756,
-    "eval_runtime": 316.6745,
+    "epoch": 100.0,
+    "eval_cer": 0.12175835997967904,
+    "eval_loss": 0.33164310455322266,
+    "eval_runtime": 269.3321,
     "eval_samples": 4470,
-    "eval_samples_per_second": 14.115,
-    "eval_steps_per_second": 1.765,
-    "eval_wer": 0.9960036558484615
+    "eval_samples_per_second": 16.597,
+    "eval_steps_per_second": 2.076,
+    "eval_wer": 0.2563574129495887
 }

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:07f8723eb955d88bd7404afdfdb91dd19bbc3e2bdf84b4f9891357e18de03047
+oid sha256:4a3f26f1ec034a6b5009d36284873996f246ddb50af387ad2e3f3ebd16d605b3
 size 3850912241

runs/Feb06_00-21-08_job-0167f55d-c17f-480c-9f0c-e77ccfce0ee5/events.out.tfevents.1644106918.job-0167f55d-c17f-480c-9f0c-e77ccfce0ee5.39512.0

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8a3b0141c51d24864c14c294ab2bb798af5994e22bc6a77db036cc19b07ae71c
-size 7889
+oid sha256:b7f31d5014dbb9e485d691c1b4bc587365601fe2e371a08612587c526f742942
+size 10331

runs/Feb06_00-21-08_job-0167f55d-c17f-480c-9f0c-e77ccfce0ee5/events.out.tfevents.1644202339.job-0167f55d-c17f-480c-9f0c-e77ccfce0ee5.39512.2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:af8e7d6e5d91b945f30cb85b919c8525346be72292882874e125c6a107c76f63
+size 405

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 2.0,
-    "train_loss": 6.151403113256527,
-    "train_runtime": 1933.761,
+    "epoch": 100.0,
+    "train_loss": 1.2220959588545788,
+    "train_runtime": 95146.6811,
     "train_samples": 10156,
-    "train_samples_per_second": 10.504,
-    "train_steps_per_second": 0.163
+    "train_samples_per_second": 10.674,
+    "train_steps_per_second": 0.166
 }

trainer_state.json

@@ -1,25 +1,185 @@
 {
-  "best_metric": null,
-  "best_model_checkpoint": null,
-  "epoch": 1.9952755905511812,
-  "global_step": 316,
+  "best_metric": 0.33164310455322266,
+  "best_model_checkpoint": "wav2vec2-xls-r-1b-ja/checkpoint-15000",
+  "epoch": 99.99527559055119,
+  "global_step": 15800,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 2.0,
-      "step": 316,
-      "total_flos": 9.23014088962431e+18,
-      "train_loss": 6.151403113256527,
-      "train_runtime": 1933.761,
-      "train_samples_per_second": 10.504,
-      "train_steps_per_second": 0.163
+      "epoch": 9.49,
+      "learning_rate": 3.7425e-05,
+      "loss": 3.484,
+      "step": 1500
+    },
+    {
+      "epoch": 9.49,
+      "eval_cer": 0.4098974987299405,
+      "eval_loss": 1.1849042177200317,
+      "eval_runtime": 268.3084,
+      "eval_samples_per_second": 16.66,
+      "eval_steps_per_second": 2.083,
+      "eval_wer": 0.7542696367448612,
+      "step": 1500
+    },
+    {
+      "epoch": 18.98,
+      "learning_rate": 4.638768115942029e-05,
+      "loss": 1.3582,
+      "step": 3000
+    },
+    {
+      "epoch": 18.98,
+      "eval_cer": 0.1590532827301796,
+      "eval_loss": 0.4320400655269623,
+      "eval_runtime": 269.0628,
+      "eval_samples_per_second": 16.613,
+      "eval_steps_per_second": 2.078,
+      "eval_wer": 0.34893639898926543,
+      "step": 3000
+    },
+    {
+      "epoch": 28.48,
+      "learning_rate": 4.095289855072464e-05,
+      "loss": 1.1716,
+      "step": 4500
+    },
+    {
+      "epoch": 28.48,
+      "eval_cer": 0.1453904312225443,
+      "eval_loss": 0.38348379731178284,
+      "eval_runtime": 270.1414,
+      "eval_samples_per_second": 16.547,
+      "eval_steps_per_second": 2.069,
+      "eval_wer": 0.31746742889912366,
+      "step": 4500
+    },
+    {
+      "epoch": 37.97,
+      "learning_rate": 3.552173913043478e-05,
+      "loss": 1.0951,
+      "step": 6000
+    },
+    {
+      "epoch": 37.97,
+      "eval_cer": 0.14047754236021875,
+      "eval_loss": 0.3732404410839081,
+      "eval_runtime": 272.8075,
+      "eval_samples_per_second": 16.385,
+      "eval_steps_per_second": 2.049,
+      "eval_wer": 0.3032741348721349,
+      "step": 6000
+    },
+    {
+      "epoch": 47.47,
+      "learning_rate": 3.0090579710144927e-05,
+      "loss": 1.04,
+      "step": 7500
+    },
+    {
+      "epoch": 47.47,
+      "eval_cer": 0.13603681678271523,
+      "eval_loss": 0.34847018122673035,
+      "eval_runtime": 268.2852,
+      "eval_samples_per_second": 16.661,
+      "eval_steps_per_second": 2.084,
+      "eval_wer": 0.28983351552839554,
+      "step": 7500
+    },
+    {
+      "epoch": 56.96,
+      "learning_rate": 2.4655797101449275e-05,
+      "loss": 0.9768,
+      "step": 9000
+    },
+    {
+      "epoch": 56.96,
+      "eval_cer": 0.13088485790275828,
+      "eval_loss": 0.33858078718185425,
+      "eval_runtime": 269.384,
+      "eval_samples_per_second": 16.593,
+      "eval_steps_per_second": 2.075,
+      "eval_wer": 0.278722603537571,
+      "step": 9000
+    },
+    {
+      "epoch": 66.45,
+      "learning_rate": 1.922463768115942e-05,
+      "loss": 0.9129,
+      "step": 10500
+    },
+    {
+      "epoch": 66.45,
+      "eval_cer": 0.1271553656277082,
+      "eval_loss": 0.336273193359375,
+      "eval_runtime": 267.776,
+      "eval_samples_per_second": 16.693,
+      "eval_steps_per_second": 2.088,
+      "eval_wer": 0.27108833175032704,
+      "step": 10500
+    },
+    {
+      "epoch": 75.94,
+      "learning_rate": 1.3793478260869565e-05,
+      "loss": 0.8614,
+      "step": 12000
+    },
+    {
+      "epoch": 75.94,
+      "eval_cer": 0.12596001553955113,
+      "eval_loss": 0.33856987953186035,
+      "eval_runtime": 268.1172,
+      "eval_samples_per_second": 16.672,
+      "eval_steps_per_second": 2.085,
+      "eval_wer": 0.26755792906937154,
+      "step": 12000
+    },
+    {
+      "epoch": 85.44,
+      "learning_rate": 8.358695652173914e-06,
+      "loss": 0.8092,
+      "step": 13500
+    },
+    {
+      "epoch": 85.44,
+      "eval_cer": 0.12399964139497355,
+      "eval_loss": 0.3356078565120697,
+      "eval_runtime": 269.2379,
+      "eval_samples_per_second": 16.602,
+      "eval_steps_per_second": 2.076,
+      "eval_wer": 0.26103474848121,
+      "step": 13500
+    },
+    {
+      "epoch": 94.93,
+      "learning_rate": 2.9275362318840578e-06,
+      "loss": 0.7658,
+      "step": 15000
+    },
+    {
+      "epoch": 94.93,
+      "eval_cer": 0.12175835997967904,
+      "eval_loss": 0.33164310455322266,
+      "eval_runtime": 268.5018,
+      "eval_samples_per_second": 16.648,
+      "eval_steps_per_second": 2.082,
+      "eval_wer": 0.2563574129495887,
+      "step": 15000
+    },
+    {
+      "epoch": 100.0,
+      "step": 15800,
+      "total_flos": 4.622164701241047e+20,
+      "train_loss": 1.2220959588545788,
+      "train_runtime": 95146.6811,
+      "train_samples_per_second": 10.674,
+      "train_steps_per_second": 0.166
     }
   ],
-  "max_steps": 316,
-  "num_train_epochs": 2,
-  "total_flos": 9.23014088962431e+18,
+  "max_steps": 15800,
+  "num_train_epochs": 100,
+  "total_flos": 4.622164701241047e+20,
   "trial_name": null,
   "trial_params": null
 }