Arousal Only Train

README.md

@@ -17,8 +17,8 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.6919
-- Accuracy: 0.4085
+- Loss: 0.4207
+- Accuracy: 0.9014
 
 ## Model description
 
@@ -46,14 +46,28 @@ The following hyperparameters were used during training:
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
 - lr_scheduler_warmup_steps: 1000
-- num_epochs: 1
+- num_epochs: 15
 - mixed_precision_training: Native AMP
 
 ### Training results
 
 | Training Loss | Epoch | Step | Validation Loss | Accuracy |
 |:-------------:|:-----:|:----:|:---------------:|:--------:|
-| 0.7075        | 1.0   | 269  | 0.6919          | 0.4085   |
+| 0.6871        | 1.0   | 269  | 0.6601          | 0.6761   |
+| 0.6071        | 2.0   | 538  | 0.5375          | 0.8451   |
+| 0.4312        | 3.0   | 807  | 0.3544          | 0.8873   |
+| 0.306         | 4.0   | 1076 | 0.3780          | 0.8592   |
+| 0.3052        | 5.0   | 1345 | 0.4133          | 0.8873   |
+| 0.3099        | 6.0   | 1614 | 0.4112          | 0.8873   |
+| 0.2965        | 7.0   | 1883 | 0.4241          | 0.8873   |
+| 0.2954        | 8.0   | 2152 | 0.4381          | 0.8873   |
+| 0.2905        | 9.0   | 2421 | 0.4294          | 0.9014   |
+| 0.2868        | 10.0  | 2690 | 0.4208          | 0.9014   |
+| 0.284         | 11.0  | 2959 | 0.4077          | 0.9014   |
+| 0.2666        | 12.0  | 3228 | 0.4149          | 0.9014   |
+| 0.2697        | 13.0  | 3497 | 0.4108          | 0.9014   |
+| 0.2622        | 14.0  | 3766 | 0.4187          | 0.9014   |
+| 0.2648        | 15.0  | 4035 | 0.4207          | 0.9014   |
 
 
 ### Framework versions

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fc59436a776ace60768d1997d9f14a66e8039c52c9791feb588528148780a5ea
+oid sha256:175c5b0a34b3e595c05d16a862cdc541d6872730f924b3768b5cef0f011ea7da
 size 379881432

runs/Feb05_16-51-08_9d7b7eeffc40/events.out.tfevents.1707152167.9d7b7eeffc40.773.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:01821e32b409628120325298b1699ff390eb3ba79374fd4959977ead22a8e425
+size 13987

trainer_state.json

@@ -1,9 +1,9 @@
 {
-  "best_metric": 0.6918669939041138,
-  "best_model_checkpoint": "/content/drive/MyDrive/AI_development/Musicbot/train_model/Arousal-wav2vec2-base-EMOPIA/checkpoint-269",
-  "epoch": 1.0,
+  "best_metric": 0.35436323285102844,
+  "best_model_checkpoint": "/content/drive/MyDrive/AI_development/Musicbot/train_model/Arousal-wav2vec2-base-EMOPIA/checkpoint-807",
+  "epoch": 15.0,
   "eval_steps": 500,
-  "global_step": 269,
+  "global_step": 4035,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -11,34 +11,244 @@
     {
       "epoch": 1.0,
       "learning_rate": 2.69e-07,
-      "loss": 0.7075,
+      "loss": 0.6871,
       "step": 269
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.4084506928920746,
-      "eval_loss": 0.6918669939041138,
-      "eval_runtime": 39.6785,
-      "eval_samples_per_second": 1.789,
-      "eval_steps_per_second": 1.789,
+      "eval_accuracy": 0.6760563254356384,
+      "eval_loss": 0.6600874662399292,
+      "eval_runtime": 43.6337,
+      "eval_samples_per_second": 1.627,
+      "eval_steps_per_second": 1.627,
       "step": 269
     },
     {
-      "epoch": 1.0,
-      "step": 269,
-      "total_flos": 4.48797597870528e+17,
-      "train_loss": 0.7074856102245005,
-      "train_runtime": 1028.8672,
+      "epoch": 2.0,
+      "learning_rate": 5.38e-07,
+      "loss": 0.6071,
+      "step": 538
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.8450704216957092,
+      "eval_loss": 0.5374944806098938,
+      "eval_runtime": 40.8819,
+      "eval_samples_per_second": 1.737,
+      "eval_steps_per_second": 1.737,
+      "step": 538
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 8.05e-07,
+      "loss": 0.4312,
+      "step": 807
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.8873239159584045,
+      "eval_loss": 0.35436323285102844,
+      "eval_runtime": 40.885,
+      "eval_samples_per_second": 1.737,
+      "eval_steps_per_second": 1.737,
+      "step": 807
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 9.756177924217462e-07,
+      "loss": 0.306,
+      "step": 1076
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.8591549396514893,
+      "eval_loss": 0.37799715995788574,
+      "eval_runtime": 41.5336,
+      "eval_samples_per_second": 1.709,
+      "eval_steps_per_second": 1.709,
+      "step": 1076
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 8.87314662273476e-07,
+      "loss": 0.3052,
+      "step": 1345
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.8873239159584045,
+      "eval_loss": 0.4133167564868927,
+      "eval_runtime": 41.0628,
+      "eval_samples_per_second": 1.729,
+      "eval_steps_per_second": 1.729,
+      "step": 1345
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 7.986820428336078e-07,
+      "loss": 0.3099,
+      "step": 1614
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.8873239159584045,
+      "eval_loss": 0.41115257143974304,
+      "eval_runtime": 41.3286,
+      "eval_samples_per_second": 1.718,
+      "eval_steps_per_second": 1.718,
+      "step": 1614
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 7.100494233937397e-07,
+      "loss": 0.2965,
+      "step": 1883
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.8873239159584045,
+      "eval_loss": 0.4240788221359253,
+      "eval_runtime": 41.7099,
+      "eval_samples_per_second": 1.702,
+      "eval_steps_per_second": 1.702,
+      "step": 1883
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 6.217462932454694e-07,
+      "loss": 0.2954,
+      "step": 2152
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.8873239159584045,
+      "eval_loss": 0.4380823075771332,
+      "eval_runtime": 41.575,
+      "eval_samples_per_second": 1.708,
+      "eval_steps_per_second": 1.708,
+      "step": 2152
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 5.331136738056013e-07,
+      "loss": 0.2905,
+      "step": 2421
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.9014084339141846,
+      "eval_loss": 0.429412841796875,
+      "eval_runtime": 41.4225,
+      "eval_samples_per_second": 1.714,
+      "eval_steps_per_second": 1.714,
+      "step": 2421
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 4.444810543657331e-07,
+      "loss": 0.2868,
+      "step": 2690
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.9014084339141846,
+      "eval_loss": 0.42075294256210327,
+      "eval_runtime": 41.3714,
+      "eval_samples_per_second": 1.716,
+      "eval_steps_per_second": 1.716,
+      "step": 2690
+    },
+    {
+      "epoch": 11.0,
+      "learning_rate": 3.558484349258649e-07,
+      "loss": 0.284,
+      "step": 2959
+    },
+    {
+      "epoch": 11.0,
+      "eval_accuracy": 0.9014084339141846,
+      "eval_loss": 0.4076564311981201,
+      "eval_runtime": 40.865,
+      "eval_samples_per_second": 1.737,
+      "eval_steps_per_second": 1.737,
+      "step": 2959
+    },
+    {
+      "epoch": 12.0,
+      "learning_rate": 2.675453047775947e-07,
+      "loss": 0.2666,
+      "step": 3228
+    },
+    {
+      "epoch": 12.0,
+      "eval_accuracy": 0.9014084339141846,
+      "eval_loss": 0.4149322509765625,
+      "eval_runtime": 41.9593,
+      "eval_samples_per_second": 1.692,
+      "eval_steps_per_second": 1.692,
+      "step": 3228
+    },
+    {
+      "epoch": 13.0,
+      "learning_rate": 1.7891268533772652e-07,
+      "loss": 0.2697,
+      "step": 3497
+    },
+    {
+      "epoch": 13.0,
+      "eval_accuracy": 0.9014084339141846,
+      "eval_loss": 0.4108058214187622,
+      "eval_runtime": 41.7738,
+      "eval_samples_per_second": 1.7,
+      "eval_steps_per_second": 1.7,
+      "step": 3497
+    },
+    {
+      "epoch": 14.0,
+      "learning_rate": 9.028006589785832e-08,
+      "loss": 0.2622,
+      "step": 3766
+    },
+    {
+      "epoch": 14.0,
+      "eval_accuracy": 0.9014084339141846,
+      "eval_loss": 0.41870787739753723,
+      "eval_runtime": 40.9404,
+      "eval_samples_per_second": 1.734,
+      "eval_steps_per_second": 1.734,
+      "step": 3766
+    },
+    {
+      "epoch": 15.0,
+      "learning_rate": 1.6474464579901152e-09,
+      "loss": 0.2648,
+      "step": 4035
+    },
+    {
+      "epoch": 15.0,
+      "eval_accuracy": 0.9014084339141846,
+      "eval_loss": 0.42074134945869446,
+      "eval_runtime": 41.2582,
+      "eval_samples_per_second": 1.721,
+      "eval_steps_per_second": 1.721,
+      "step": 4035
+    },
+    {
+      "epoch": 15.0,
+      "step": 4035,
+      "total_flos": 5.776095920022718e+18,
+      "train_loss": 0.344214256693144,
+      "train_runtime": 15432.7099,
       "train_samples_per_second": 0.784,
       "train_steps_per_second": 0.261
     }
   ],
   "logging_steps": 33,
-  "max_steps": 269,
+  "max_steps": 4035,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 1,
+  "num_train_epochs": 15,
   "save_steps": 500,
-  "total_flos": 4.48797597870528e+17,
+  "total_flos": 5.776095920022718e+18,
   "train_batch_size": 1,
   "trial_name": null,
   "trial_params": null

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a0b5954fe455b5a7f9d3d4dffa43e6dcfa1ecc8d79884821e3cc8163631dd071
+oid sha256:a370310d1c038daf2346d344f9b00f826a2451c79990871b7caa5e00d04ac49b
 size 4856