End of training

README.md

@@ -1,5 +1,6 @@
 ---
 license: apache-2.0
+base_model: facebook/wav2vec2-base
 tags:
 - generated_from_trainer
 metrics:
@@ -14,10 +15,10 @@ should probably proofread and complete it, then remove this comment. -->
 
 # wav2vec2-base-EMOPIA
 
-This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on the None dataset.
+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: 1.4966
-- Accuracy: 0.6429
+- Loss: 2.1691
+- Accuracy: 0.6338
 
 ## Model description
 
@@ -37,12 +38,13 @@ More information needed
 
 The following hyperparameters were used during training:
 - learning_rate: 1e-05
-- train_batch_size: 4
-- eval_batch_size: 4
+- train_batch_size: 1
+- eval_batch_size: 1
 - seed: 42
+- gradient_accumulation_steps: 3
+- total_train_batch_size: 3
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
-- lr_scheduler_warmup_steps: 500
 - num_epochs: 15
 - mixed_precision_training: Native AMP
 
@@ -50,26 +52,26 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch | Step | Validation Loss | Accuracy |
 |:-------------:|:-----:|:----:|:---------------:|:--------:|
-| 1.3315        | 1.0   | 175  | 1.1984          | 0.3857   |
-| 1.0391        | 2.0   | 350  | 1.0071          | 0.5429   |
-| 0.8973        | 3.0   | 525  | 0.9441          | 0.5857   |
-| 0.8199        | 4.0   | 700  | 0.9984          | 0.5286   |
-| 0.7675        | 5.0   | 875  | 0.9477          | 0.6286   |
-| 0.6446        | 6.0   | 1050 | 1.0956          | 0.6429   |
-| 0.6221        | 7.0   | 1225 | 1.0144          | 0.6286   |
-| 0.5543        | 8.0   | 1400 | 1.2473          | 0.6286   |
-| 0.5095        | 9.0   | 1575 | 1.2690          | 0.6714   |
-| 0.4461        | 10.0  | 1750 | 1.1365          | 0.7286   |
-| 0.3678        | 11.0  | 1925 | 1.2065          | 0.7000   |
-| 0.3184        | 12.0  | 2100 | 1.6990          | 0.6286   |
-| 0.271         | 13.0  | 2275 | 1.5335          | 0.6286   |
-| 0.234         | 14.0  | 2450 | 1.3234          | 0.7143   |
-| 0.2049        | 15.0  | 2625 | 1.4966          | 0.6429   |
+| 0.8716        | 1.0   | 269  | 0.9822          | 0.6197   |
+| 0.8143        | 2.0   | 538  | 1.2324          | 0.5352   |
+| 0.7584        | 3.0   | 807  | 1.0226          | 0.6479   |
+| 0.6715        | 4.0   | 1076 | 0.9550          | 0.6620   |
+| 0.6471        | 5.0   | 1345 | 1.1272          | 0.6761   |
+| 0.5759        | 6.0   | 1614 | 1.2193          | 0.6761   |
+| 0.4963        | 7.0   | 1883 | 1.2214          | 0.7183   |
+| 0.4053        | 8.0   | 2152 | 1.3083          | 0.7465   |
+| 0.3344        | 9.0   | 2421 | 1.6391          | 0.6620   |
+| 0.3216        | 10.0  | 2690 | 1.7224          | 0.6479   |
+| 0.2248        | 11.0  | 2959 | 1.7973          | 0.6761   |
+| 0.1982        | 12.0  | 3228 | 2.0241          | 0.6479   |
+| 0.1362        | 13.0  | 3497 | 1.9933          | 0.6479   |
+| 0.0879        | 14.0  | 3766 | 2.0865          | 0.6479   |
+| 0.0712        | 15.0  | 4035 | 2.1691          | 0.6338   |
 
 
 ### Framework versions
 
-- Transformers 4.18.0
-- Pytorch 1.10.1+cu102
-- Datasets 2.4.0
-- Tokenizers 0.12.1
+- Transformers 4.42.2
+- Pytorch 2.3.1+cu118
+- Datasets 2.20.0
+- Tokenizers 0.19.1

config.json

@@ -1,6 +1,7 @@
 {
   "_name_or_path": "facebook/wav2vec2-base",
   "activation_dropout": 0.0,
+  "adapter_attn_dim": null,
   "adapter_kernel_size": 3,
   "adapter_stride": 2,
   "add_adapter": false,
@@ -53,6 +54,7 @@
   "final_dropout": 0.0,
   "finetuning_task": "facebook/wav2vec2-base",
   "freeze_feat_extract_train": true,
+  "gradient_checkpointing": false,
   "hidden_act": "gelu",
   "hidden_dropout": 0.1,
   "hidden_size": 768,
@@ -125,7 +127,7 @@
     1
   ],
   "torch_dtype": "float32",
-  "transformers_version": "4.18.0",
+  "transformers_version": "4.42.2",
   "use_weighted_layer_sum": false,
   "vocab_size": 32,
   "xvector_output_dim": 512

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:07a19119aa6c8a53c8089247222f6c8e7cdffbc0f38560d051fa0c6e69e54868
+size 379887584

trainer_state.json

@@ -1,250 +1,282 @@
 {
-  "best_metric": 0.9441239833831787,
-  "best_model_checkpoint": "kurosekurose/wav2vec2-base-EMOPIA/checkpoint-525",
+  "best_metric": null,
+  "best_model_checkpoint": null,
   "epoch": 15.0,
-  "global_step": 2625,
+  "eval_steps": 500,
+  "global_step": 4035,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 1.0,
-      "learning_rate": 3.5e-06,
-      "loss": 1.3315,
-      "step": 175
+      "grad_norm": 13.118809700012207,
+      "learning_rate": 9.335811648079307e-06,
+      "loss": 0.8716,
+      "step": 269
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.3857142925262451,
-      "eval_loss": 1.1984184980392456,
-      "eval_runtime": 48.1732,
-      "eval_samples_per_second": 1.453,
-      "eval_steps_per_second": 0.374,
-      "step": 175
+      "eval_accuracy": 0.6197183132171631,
+      "eval_loss": 0.9822000861167908,
+      "eval_runtime": 32.2509,
+      "eval_samples_per_second": 2.201,
+      "eval_steps_per_second": 2.201,
+      "step": 269
     },
     {
       "epoch": 2.0,
-      "learning_rate": 7e-06,
-      "loss": 1.0391,
-      "step": 350
+      "grad_norm": 10.717401504516602,
+      "learning_rate": 8.66914498141264e-06,
+      "loss": 0.8143,
+      "step": 538
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.5428571701049805,
-      "eval_loss": 1.0071275234222412,
-      "eval_runtime": 48.3665,
-      "eval_samples_per_second": 1.447,
-      "eval_steps_per_second": 0.372,
-      "step": 350
+      "eval_accuracy": 0.5352112650871277,
+      "eval_loss": 1.232447624206543,
+      "eval_runtime": 32.3951,
+      "eval_samples_per_second": 2.192,
+      "eval_steps_per_second": 2.192,
+      "step": 538
     },
     {
       "epoch": 3.0,
-      "learning_rate": 9.887058823529413e-06,
-      "loss": 0.8973,
-      "step": 525
+      "grad_norm": 23.701391220092773,
+      "learning_rate": 8.004956629491945e-06,
+      "loss": 0.7584,
+      "step": 807
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.5857142806053162,
-      "eval_loss": 0.9441239833831787,
-      "eval_runtime": 48.2354,
-      "eval_samples_per_second": 1.451,
-      "eval_steps_per_second": 0.373,
-      "step": 525
+      "eval_accuracy": 0.6478873491287231,
+      "eval_loss": 1.0226496458053589,
+      "eval_runtime": 32.3,
+      "eval_samples_per_second": 2.198,
+      "eval_steps_per_second": 2.198,
+      "step": 807
     },
     {
       "epoch": 4.0,
-      "learning_rate": 9.063529411764706e-06,
-      "loss": 0.8199,
-      "step": 700
+      "grad_norm": 33.54118728637695,
+      "learning_rate": 7.340768277571252e-06,
+      "loss": 0.6715,
+      "step": 1076
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.5285714268684387,
-      "eval_loss": 0.9983876347541809,
-      "eval_runtime": 48.536,
-      "eval_samples_per_second": 1.442,
-      "eval_steps_per_second": 0.371,
-      "step": 700
+      "eval_accuracy": 0.6619718074798584,
+      "eval_loss": 0.9550462961196899,
+      "eval_runtime": 32.4715,
+      "eval_samples_per_second": 2.187,
+      "eval_steps_per_second": 2.187,
+      "step": 1076
     },
     {
       "epoch": 5.0,
-      "learning_rate": 8.244705882352942e-06,
-      "loss": 0.7675,
-      "step": 875
+      "grad_norm": 30.0217227935791,
+      "learning_rate": 6.674101610904585e-06,
+      "loss": 0.6471,
+      "step": 1345
     },
     {
       "epoch": 5.0,
-      "eval_accuracy": 0.6285714507102966,
-      "eval_loss": 0.9477005004882812,
-      "eval_runtime": 48.23,
-      "eval_samples_per_second": 1.451,
-      "eval_steps_per_second": 0.373,
-      "step": 875
+      "eval_accuracy": 0.6760563254356384,
+      "eval_loss": 1.1272403001785278,
+      "eval_runtime": 32.2183,
+      "eval_samples_per_second": 2.204,
+      "eval_steps_per_second": 2.204,
+      "step": 1345
     },
     {
       "epoch": 6.0,
-      "learning_rate": 7.421176470588236e-06,
-      "loss": 0.6446,
-      "step": 1050
+      "grad_norm": 0.36277323961257935,
+      "learning_rate": 6.009913258983892e-06,
+      "loss": 0.5759,
+      "step": 1614
     },
     {
       "epoch": 6.0,
-      "eval_accuracy": 0.6428571343421936,
-      "eval_loss": 1.0956463813781738,
-      "eval_runtime": 48.136,
-      "eval_samples_per_second": 1.454,
-      "eval_steps_per_second": 0.374,
-      "step": 1050
+      "eval_accuracy": 0.6760563254356384,
+      "eval_loss": 1.2193043231964111,
+      "eval_runtime": 32.3011,
+      "eval_samples_per_second": 2.198,
+      "eval_steps_per_second": 2.198,
+      "step": 1614
     },
     {
       "epoch": 7.0,
-      "learning_rate": 6.59764705882353e-06,
-      "loss": 0.6221,
-      "step": 1225
+      "grad_norm": 17.32307243347168,
+      "learning_rate": 5.343246592317225e-06,
+      "loss": 0.4963,
+      "step": 1883
     },
     {
       "epoch": 7.0,
-      "eval_accuracy": 0.6285714507102966,
-      "eval_loss": 1.0143612623214722,
-      "eval_runtime": 48.3709,
-      "eval_samples_per_second": 1.447,
-      "eval_steps_per_second": 0.372,
-      "step": 1225
+      "eval_accuracy": 0.7183098793029785,
+      "eval_loss": 1.221394658088684,
+      "eval_runtime": 32.1815,
+      "eval_samples_per_second": 2.206,
+      "eval_steps_per_second": 2.206,
+      "step": 1883
     },
     {
       "epoch": 8.0,
-      "learning_rate": 5.778823529411765e-06,
-      "loss": 0.5543,
-      "step": 1400
+      "grad_norm": 0.09576527029275894,
+      "learning_rate": 4.679058240396531e-06,
+      "loss": 0.4053,
+      "step": 2152
     },
     {
       "epoch": 8.0,
-      "eval_accuracy": 0.6285714507102966,
-      "eval_loss": 1.24734365940094,
-      "eval_runtime": 48.144,
-      "eval_samples_per_second": 1.454,
-      "eval_steps_per_second": 0.374,
-      "step": 1400
+      "eval_accuracy": 0.7464788556098938,
+      "eval_loss": 1.3082976341247559,
+      "eval_runtime": 32.7134,
+      "eval_samples_per_second": 2.17,
+      "eval_steps_per_second": 2.17,
+      "step": 2152
     },
     {
       "epoch": 9.0,
-      "learning_rate": 4.955294117647059e-06,
-      "loss": 0.5095,
-      "step": 1575
+      "grad_norm": 73.52884674072266,
+      "learning_rate": 4.012391573729864e-06,
+      "loss": 0.3344,
+      "step": 2421
     },
     {
       "epoch": 9.0,
-      "eval_accuracy": 0.6714285612106323,
-      "eval_loss": 1.2690364122390747,
-      "eval_runtime": 48.163,
-      "eval_samples_per_second": 1.453,
-      "eval_steps_per_second": 0.374,
-      "step": 1575
+      "eval_accuracy": 0.6619718074798584,
+      "eval_loss": 1.6390645503997803,
+      "eval_runtime": 32.8911,
+      "eval_samples_per_second": 2.159,
+      "eval_steps_per_second": 2.159,
+      "step": 2421
     },
     {
       "epoch": 10.0,
-      "learning_rate": 4.1317647058823535e-06,
-      "loss": 0.4461,
-      "step": 1750
+      "grad_norm": 242.48890686035156,
+      "learning_rate": 3.34820322180917e-06,
+      "loss": 0.3216,
+      "step": 2690
     },
     {
       "epoch": 10.0,
-      "eval_accuracy": 0.7285714149475098,
-      "eval_loss": 1.1364508867263794,
-      "eval_runtime": 48.2055,
-      "eval_samples_per_second": 1.452,
-      "eval_steps_per_second": 0.373,
-      "step": 1750
+      "eval_accuracy": 0.6478873491287231,
+      "eval_loss": 1.722383737564087,
+      "eval_runtime": 32.4813,
+      "eval_samples_per_second": 2.186,
+      "eval_steps_per_second": 2.186,
+      "step": 2690
     },
     {
       "epoch": 11.0,
-      "learning_rate": 3.3129411764705886e-06,
-      "loss": 0.3678,
-      "step": 1925
+      "grad_norm": 376.0904541015625,
+      "learning_rate": 2.6815365551425034e-06,
+      "loss": 0.2248,
+      "step": 2959
     },
     {
       "epoch": 11.0,
-      "eval_accuracy": 0.699999988079071,
-      "eval_loss": 1.2065324783325195,
-      "eval_runtime": 48.5129,
-      "eval_samples_per_second": 1.443,
-      "eval_steps_per_second": 0.371,
-      "step": 1925
+      "eval_accuracy": 0.6760563254356384,
+      "eval_loss": 1.7972948551177979,
+      "eval_runtime": 32.2962,
+      "eval_samples_per_second": 2.198,
+      "eval_steps_per_second": 2.198,
+      "step": 2959
     },
     {
       "epoch": 12.0,
-      "learning_rate": 2.4894117647058827e-06,
-      "loss": 0.3184,
-      "step": 2100
+      "grad_norm": 38.31782913208008,
+      "learning_rate": 2.0148698884758364e-06,
+      "loss": 0.1982,
+      "step": 3228
     },
     {
       "epoch": 12.0,
-      "eval_accuracy": 0.6285714507102966,
-      "eval_loss": 1.699022889137268,
-      "eval_runtime": 48.0723,
-      "eval_samples_per_second": 1.456,
-      "eval_steps_per_second": 0.374,
-      "step": 2100
+      "eval_accuracy": 0.6478873491287231,
+      "eval_loss": 2.02411150932312,
+      "eval_runtime": 32.4081,
+      "eval_samples_per_second": 2.191,
+      "eval_steps_per_second": 2.191,
+      "step": 3228
     },
     {
       "epoch": 13.0,
-      "learning_rate": 1.6658823529411766e-06,
-      "loss": 0.271,
-      "step": 2275
+      "grad_norm": 0.5905938744544983,
+      "learning_rate": 1.3482032218091697e-06,
+      "loss": 0.1362,
+      "step": 3497
     },
     {
       "epoch": 13.0,
-      "eval_accuracy": 0.6285714507102966,
-      "eval_loss": 1.5335015058517456,
-      "eval_runtime": 48.3723,
-      "eval_samples_per_second": 1.447,
-      "eval_steps_per_second": 0.372,
-      "step": 2275
+      "eval_accuracy": 0.6478873491287231,
+      "eval_loss": 1.9932571649551392,
+      "eval_runtime": 32.4096,
+      "eval_samples_per_second": 2.191,
+      "eval_steps_per_second": 2.191,
+      "step": 3497
     },
     {
       "epoch": 14.0,
-      "learning_rate": 8.423529411764707e-07,
-      "loss": 0.234,
-      "step": 2450
+      "grad_norm": 2.5972626209259033,
+      "learning_rate": 6.815365551425032e-07,
+      "loss": 0.0879,
+      "step": 3766
     },
     {
       "epoch": 14.0,
-      "eval_accuracy": 0.7142857313156128,
-      "eval_loss": 1.3233639001846313,
-      "eval_runtime": 47.9369,
-      "eval_samples_per_second": 1.46,
-      "eval_steps_per_second": 0.375,
-      "step": 2450
+      "eval_accuracy": 0.6478873491287231,
+      "eval_loss": 2.086475372314453,
+      "eval_runtime": 32.8946,
+      "eval_samples_per_second": 2.158,
+      "eval_steps_per_second": 2.158,
+      "step": 3766
     },
     {
       "epoch": 15.0,
-      "learning_rate": 1.8823529411764708e-08,
-      "loss": 0.2049,
-      "step": 2625
+      "grad_norm": 0.036518827080726624,
+      "learning_rate": 1.4869888475836432e-08,
+      "loss": 0.0712,
+      "step": 4035
     },
     {
       "epoch": 15.0,
-      "eval_accuracy": 0.6428571343421936,
-      "eval_loss": 1.4966269731521606,
-      "eval_runtime": 48.2524,
-      "eval_samples_per_second": 1.451,
-      "eval_steps_per_second": 0.373,
-      "step": 2625
+      "eval_accuracy": 0.6338028311729431,
+      "eval_loss": 2.1690945625305176,
+      "eval_runtime": 32.3643,
+      "eval_samples_per_second": 2.194,
+      "eval_steps_per_second": 2.194,
+      "step": 4035
     },
     {
       "epoch": 15.0,
-      "step": 2625,
-      "total_flos": 4.992891844826972e+18,
-      "train_loss": 0.6018750116257441,
-      "train_runtime": 20482.8582,
-      "train_samples_per_second": 0.513,
-      "train_steps_per_second": 0.128
+      "step": 4035,
+      "total_flos": 7.1272809219168e+18,
+      "train_loss": 0.44098070326643213,
+      "train_runtime": 9434.6926,
+      "train_samples_per_second": 1.283,
+      "train_steps_per_second": 0.428
     }
   ],
-  "max_steps": 2625,
+  "logging_steps": 8,
+  "max_steps": 4035,
+  "num_input_tokens_seen": 0,
   "num_train_epochs": 15,
-  "total_flos": 4.992891844826972e+18,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 7.1272809219168e+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 sha256:64952a100eb670e38f7986031bb278f14e1f6d77290a34a2990eb65a138078b4
-size 3055
+oid sha256:32a32a793e09c20ba8ccdd195ed10a1762d2e2711caef8e57cb8228ccdadcf50
+size 5176