yashcode00/wav2vec2-large-xlsr-indian-language-classification-featureExtractor

README.md

@@ -17,8 +17,8 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [yashcode00/wav2vec2-large-xlsr-indian-language-classification-featureExtractor](https://huggingface.co/yashcode00/wav2vec2-large-xlsr-indian-language-classification-featureExtractor) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.4481
-- Accuracy: 0.8710
+- Loss: 0.1164
+- Accuracy: 0.9677
 
 ## Model description
 
@@ -45,10 +45,13 @@ The following hyperparameters were used during training:
 - total_train_batch_size: 64
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
-- num_epochs: 60
+- num_epochs: 300
 
 ### Training results
 
+| Training Loss | Epoch  | Step | Validation Loss | Accuracy |
+|:-------------:|:------:|:----:|:---------------:|:--------:|
+| 0.0194        | 203.39 | 1500 | 0.1154          | 0.9677   |
 
 
 ### Framework versions

all_results.json

@@ -1,15 +1,15 @@
 {
-    "epoch": 56.95,
-    "eval_accuracy": 0.8709677457809448,
-    "eval_loss": 0.4480999708175659,
-    "eval_runtime": 1.4774,
+    "epoch": 284.75,
+    "eval_accuracy": 0.9677419066429138,
+    "eval_loss": 0.11643270403146744,
+    "eval_runtime": 1.6786,
     "eval_samples": 93,
-    "eval_samples_per_second": 62.95,
-    "eval_steps_per_second": 8.123,
-    "total_flos": 8.064772262536032e+17,
-    "train_loss": 0.15708597316628412,
-    "train_runtime": 790.3622,
+    "eval_samples_per_second": 55.402,
+    "eval_steps_per_second": 7.149,
+    "total_flos": 4.0318045822521795e+18,
+    "train_loss": 0.022167698939641316,
+    "train_runtime": 3842.9627,
     "train_samples": 466,
-    "train_samples_per_second": 35.376,
-    "train_steps_per_second": 0.531
+    "train_samples_per_second": 36.378,
+    "train_steps_per_second": 0.546
 }

eval_results.json

@@ -1,9 +1,9 @@
 {
-    "epoch": 56.95,
-    "eval_accuracy": 0.8709677457809448,
-    "eval_loss": 0.4480999708175659,
-    "eval_runtime": 1.4774,
+    "epoch": 284.75,
+    "eval_accuracy": 0.9677419066429138,
+    "eval_loss": 0.11643270403146744,
+    "eval_runtime": 1.6786,
     "eval_samples": 93,
-    "eval_samples_per_second": 62.95,
-    "eval_steps_per_second": 8.123
+    "eval_samples_per_second": 55.402,
+    "eval_steps_per_second": 7.149
 }

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c5bd079f25010ad6fed494f319bd825fbda1ad3ba0247b15ee3faa0fc2a04cef
+oid sha256:40adbc61c446a3a2f2e26e29a2072bbdd48d0ae0bf10ed6afa5870eedb1549ec
 size 1266146037

train_results.json

@@ -1,9 +1,9 @@
 {
-    "epoch": 56.95,
-    "total_flos": 8.064772262536032e+17,
-    "train_loss": 0.15708597316628412,
-    "train_runtime": 790.3622,
+    "epoch": 284.75,
+    "total_flos": 4.0318045822521795e+18,
+    "train_loss": 0.022167698939641316,
+    "train_runtime": 3842.9627,
     "train_samples": 466,
-    "train_samples_per_second": 35.376,
-    "train_steps_per_second": 0.531
+    "train_samples_per_second": 36.378,
+    "train_steps_per_second": 0.546
 }

trainer_state.json

@@ -1,61 +1,172 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 56.94915254237288,
-  "eval_steps": 1000,
-  "global_step": 420,
+  "epoch": 284.7457627118644,
+  "eval_steps": 1500,
+  "global_step": 2100,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 13.56,
-      "learning_rate": 3.821428571428572e-05,
-      "loss": 0.554,
+      "learning_rate": 4.764285714285715e-05,
+      "loss": 0.0462,
       "step": 100
     },
     {
       "epoch": 27.12,
-      "learning_rate": 2.6309523809523813e-05,
-      "loss": 0.0396,
+      "learning_rate": 4.5261904761904766e-05,
+      "loss": 0.0281,
       "step": 200
     },
     {
       "epoch": 40.68,
-      "learning_rate": 1.4404761904761905e-05,
-      "loss": 0.0312,
+      "learning_rate": 4.2880952380952384e-05,
+      "loss": 0.0283,
       "step": 300
     },
     {
       "epoch": 54.24,
-      "learning_rate": 2.5e-06,
-      "loss": 0.0308,
+      "learning_rate": 4.05e-05,
+      "loss": 0.0289,
       "step": 400
     },
     {
-      "epoch": 56.95,
-      "step": 420,
-      "total_flos": 8.064772262536032e+17,
-      "train_loss": 0.15708597316628412,
-      "train_runtime": 790.3622,
-      "train_samples_per_second": 35.376,
-      "train_steps_per_second": 0.531
+      "epoch": 67.8,
+      "learning_rate": 3.811904761904762e-05,
+      "loss": 0.018,
+      "step": 500
     },
     {
-      "epoch": 56.95,
-      "eval_accuracy": 0.8709677457809448,
-      "eval_loss": 0.4480999708175659,
-      "eval_runtime": 1.4774,
-      "eval_samples_per_second": 62.95,
-      "eval_steps_per_second": 8.123,
-      "step": 420
+      "epoch": 81.36,
+      "learning_rate": 3.573809523809524e-05,
+      "loss": 0.0269,
+      "step": 600
+    },
+    {
+      "epoch": 94.92,
+      "learning_rate": 3.3357142857142856e-05,
+      "loss": 0.0191,
+      "step": 700
+    },
+    {
+      "epoch": 108.47,
+      "learning_rate": 3.0976190476190474e-05,
+      "loss": 0.0211,
+      "step": 800
+    },
+    {
+      "epoch": 122.03,
+      "learning_rate": 2.85952380952381e-05,
+      "loss": 0.0214,
+      "step": 900
+    },
+    {
+      "epoch": 135.59,
+      "learning_rate": 2.6214285714285713e-05,
+      "loss": 0.0272,
+      "step": 1000
+    },
+    {
+      "epoch": 149.15,
+      "learning_rate": 2.3833333333333334e-05,
+      "loss": 0.0217,
+      "step": 1100
+    },
+    {
+      "epoch": 162.71,
+      "learning_rate": 2.1452380952380956e-05,
+      "loss": 0.0207,
+      "step": 1200
+    },
+    {
+      "epoch": 176.27,
+      "learning_rate": 1.9071428571428574e-05,
+      "loss": 0.0192,
+      "step": 1300
+    },
+    {
+      "epoch": 189.83,
+      "learning_rate": 1.669047619047619e-05,
+      "loss": 0.0164,
+      "step": 1400
+    },
+    {
+      "epoch": 203.39,
+      "learning_rate": 1.4309523809523811e-05,
+      "loss": 0.0194,
+      "step": 1500
+    },
+    {
+      "epoch": 203.39,
+      "eval_accuracy": 0.9677419066429138,
+      "eval_loss": 0.115411177277565,
+      "eval_runtime": 1.4468,
+      "eval_samples_per_second": 64.281,
+      "eval_steps_per_second": 8.294,
+      "step": 1500
+    },
+    {
+      "epoch": 216.95,
+      "learning_rate": 1.192857142857143e-05,
+      "loss": 0.0177,
+      "step": 1600
+    },
+    {
+      "epoch": 230.51,
+      "learning_rate": 9.547619047619049e-06,
+      "loss": 0.0168,
+      "step": 1700
+    },
+    {
+      "epoch": 244.07,
+      "learning_rate": 7.166666666666667e-06,
+      "loss": 0.0171,
+      "step": 1800
+    },
+    {
+      "epoch": 257.63,
+      "learning_rate": 4.785714285714286e-06,
+      "loss": 0.0179,
+      "step": 1900
+    },
+    {
+      "epoch": 271.19,
+      "learning_rate": 2.404761904761905e-06,
+      "loss": 0.0171,
+      "step": 2000
+    },
+    {
+      "epoch": 284.75,
+      "learning_rate": 2.380952380952381e-08,
+      "loss": 0.0163,
+      "step": 2100
+    },
+    {
+      "epoch": 284.75,
+      "step": 2100,
+      "total_flos": 4.0318045822521795e+18,
+      "train_loss": 0.022167698939641316,
+      "train_runtime": 3842.9627,
+      "train_samples_per_second": 36.378,
+      "train_steps_per_second": 0.546
+    },
+    {
+      "epoch": 284.75,
+      "eval_accuracy": 0.9677419066429138,
+      "eval_loss": 0.11643270403146744,
+      "eval_runtime": 1.6786,
+      "eval_samples_per_second": 55.402,
+      "eval_steps_per_second": 7.149,
+      "step": 2100
     }
   ],
   "logging_steps": 100,
-  "max_steps": 420,
-  "num_train_epochs": 60,
+  "max_steps": 2100,
+  "num_train_epochs": 300,
   "save_steps": 2000,
-  "total_flos": 8.064772262536032e+17,
+  "total_flos": 4.0318045822521795e+18,
   "trial_name": null,
   "trial_params": null
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a6fff3406fc6d17e7151844526156d27d071a854fa3b738d788067583d864923
+oid sha256:3dc42704ac631ac8dd8732665b42c45bc13ac858f389067d086ba7cc86a62c22
 size 4155