Model save

.ipynb_checkpoints/config-checkpoint.json

@@ -20,7 +20,7 @@
   "decoder_layers": 6,
   "decoder_start_token_id": 50258,
   "dropout": 0.0,
-  "dropout_rate": 0.1,
+  "dropout_rate": 0,
   "encoder_attention_heads": 8,
   "encoder_ffn_dim": 2048,
   "encoder_layerdrop": 0.0,
@@ -57,7 +57,96 @@
   "num_mel_bins": 80,
   "pad_token_id": 50257,
   "scale_embedding": false,
-  "suppress_tokens": [],
+  "suppress_tokens": [
+    1,
+    2,
+    7,
+    8,
+    9,
+    10,
+    14,
+    25,
+    26,
+    27,
+    28,
+    29,
+    31,
+    58,
+    59,
+    60,
+    61,
+    62,
+    63,
+    90,
+    91,
+    92,
+    93,
+    359,
+    503,
+    522,
+    542,
+    873,
+    893,
+    902,
+    918,
+    922,
+    931,
+    1350,
+    1853,
+    1982,
+    2460,
+    2627,
+    3246,
+    3253,
+    3268,
+    3536,
+    3846,
+    3961,
+    4183,
+    4667,
+    6585,
+    6647,
+    7273,
+    9061,
+    9383,
+    10428,
+    10929,
+    11938,
+    12033,
+    12331,
+    12562,
+    13793,
+    14157,
+    14635,
+    15265,
+    15618,
+    16553,
+    16604,
+    18362,
+    18956,
+    20075,
+    21675,
+    22520,
+    26130,
+    26161,
+    26435,
+    28279,
+    29464,
+    31650,
+    32302,
+    32470,
+    36865,
+    42863,
+    47425,
+    49870,
+    50254,
+    50258,
+    50358,
+    50359,
+    50360,
+    50361,
+    50362
+  ],
   "torch_dtype": "float32",
   "transformers_version": "4.36.2",
   "use_cache": false,

README.md

@@ -8,10 +8,6 @@ metrics:
 model-index:
 - name: whisper-base-google-fleurs-pt-br
   results: []
-datasets:
-- google/fleurs
-language:
-- pt
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -21,9 +17,9 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [openai/whisper-base](https://huggingface.co/openai/whisper-base) on the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.4063
-- Wer: 21.6112
-- Wer Normalized: 18.0010
+- Loss: 0.4270
+- Wer: 22.0013
+- Wer Normalized: 18.1723
 
 ## Model description
 
@@ -42,7 +38,7 @@ More information needed
 ### Training hyperparameters
 
 The following hyperparameters were used during training:
-- learning_rate: 2.05e-05
+- learning_rate: 3.05e-05
 - train_batch_size: 32
 - eval_batch_size: 16
 - seed: 42
@@ -61,6 +57,9 @@ The following hyperparameters were used during training:
 | 0.2652        | 1.51  | 300  | 0.3872          | 21.1882 | 17.2827        |
 | 0.2901        | 2.01  | 400  | 0.3912          | 21.4608 | 17.7061        |
 | 0.1408        | 2.51  | 500  | 0.4063          | 21.6112 | 18.0010        |
+| 0.1428        | 3.02  | 600  | 0.4132          | 21.8650 | 18.0201        |
+| 0.0839        | 3.52  | 700  | 0.4252          | 22.3679 | 18.4720        |
+| 0.0906        | 4.02  | 800  | 0.4270          | 22.0013 | 18.1723        |
 
 
 ### Framework versions
@@ -68,4 +67,4 @@ The following hyperparameters were used during training:
 - Transformers 4.36.2
 - Pytorch 2.1.1
 - Datasets 2.16.1
-- Tokenizers 0.15.0
+- Tokenizers 0.15.0

all_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 2.51,
+    "train_loss": 0.48756390714645387,
+    "train_runtime": 1663.348,
+    "train_samples": 6340,
+    "train_samples_per_second": 15.391,
+    "train_steps_per_second": 0.481
+}

config.json

@@ -57,7 +57,96 @@
   "num_mel_bins": 80,
   "pad_token_id": 50257,
   "scale_embedding": false,
-  "suppress_tokens": [],
+  "suppress_tokens": [
+    1,
+    2,
+    7,
+    8,
+    9,
+    10,
+    14,
+    25,
+    26,
+    27,
+    28,
+    29,
+    31,
+    58,
+    59,
+    60,
+    61,
+    62,
+    63,
+    90,
+    91,
+    92,
+    93,
+    359,
+    503,
+    522,
+    542,
+    873,
+    893,
+    902,
+    918,
+    922,
+    931,
+    1350,
+    1853,
+    1982,
+    2460,
+    2627,
+    3246,
+    3253,
+    3268,
+    3536,
+    3846,
+    3961,
+    4183,
+    4667,
+    6585,
+    6647,
+    7273,
+    9061,
+    9383,
+    10428,
+    10929,
+    11938,
+    12033,
+    12331,
+    12562,
+    13793,
+    14157,
+    14635,
+    15265,
+    15618,
+    16553,
+    16604,
+    18362,
+    18956,
+    20075,
+    21675,
+    22520,
+    26130,
+    26161,
+    26435,
+    28279,
+    29464,
+    31650,
+    32302,
+    32470,
+    36865,
+    42863,
+    47425,
+    49870,
+    50254,
+    50258,
+    50358,
+    50359,
+    50360,
+    50361,
+    50362
+  ],
   "torch_dtype": "float32",
   "transformers_version": "4.36.2",
   "use_cache": false,

runs/Jan13_19-32-57_instance-20240113-151743/events.out.tfevents.1705174381.instance-20240113-151743

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5012c3c0053a795a299c8c391ef99b18419421c5ab2a3e86731ba2557ce61408
+size 9570

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 2.51,
+    "train_loss": 0.48756390714645387,
+    "train_runtime": 1663.348,
+    "train_samples": 6340,
+    "train_samples_per_second": 15.391,
+    "train_steps_per_second": 0.481
+}

trainer_state.json

@@ -0,0 +1,200 @@
+{
+  "best_metric": 20.920285768001502,
+  "best_model_checkpoint": "./whisper-base-google-fleurs-pt-br/checkpoint-200",
+  "epoch": 2.5125628140703515,
+  "eval_steps": 100,
+  "global_step": 500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.893749999999999e-06,
+      "loss": 2.004,
+      "step": 25
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.2299999999999999e-05,
+      "loss": 1.1337,
+      "step": 50
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.870625e-05,
+      "loss": 0.8017,
+      "step": 75
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.99875e-05,
+      "loss": 0.6738,
+      "step": 100
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 0.3942798376083374,
+      "eval_runtime": 199.2208,
+      "eval_samples_per_second": 4.613,
+      "eval_steps_per_second": 0.291,
+      "eval_wer": 21.73340853543899,
+      "eval_wer_normalized": 17.94871794871795,
+      "step": 100
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.9275694444444445e-05,
+      "loss": 0.5827,
+      "step": 125
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.8563888888888888e-05,
+      "loss": 0.5512,
+      "step": 150
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.7852083333333335e-05,
+      "loss": 0.5489,
+      "step": 175
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.7140277777777778e-05,
+      "loss": 0.4816,
+      "step": 200
+    },
+    {
+      "epoch": 1.01,
+      "eval_loss": 0.3761998414993286,
+      "eval_runtime": 198.2156,
+      "eval_samples_per_second": 4.636,
+      "eval_steps_per_second": 0.293,
+      "eval_wer": 20.920285768001502,
+      "eval_wer_normalized": 17.13524570667428,
+      "step": 200
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.6428472222222225e-05,
+      "loss": 0.3379,
+      "step": 225
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.5716666666666668e-05,
+      "loss": 0.3326,
+      "step": 250
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.500486111111111e-05,
+      "loss": 0.2885,
+      "step": 275
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.4293055555555555e-05,
+      "loss": 0.2652,
+      "step": 300
+    },
+    {
+      "epoch": 1.51,
+      "eval_loss": 0.3871774673461914,
+      "eval_runtime": 196.5558,
+      "eval_samples_per_second": 4.676,
+      "eval_steps_per_second": 0.295,
+      "eval_wer": 21.188193269411542,
+      "eval_wer_normalized": 17.282717282717282,
+      "step": 300
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.358125e-05,
+      "loss": 0.2971,
+      "step": 325
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.2869444444444445e-05,
+      "loss": 0.2761,
+      "step": 350
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.215763888888889e-05,
+      "loss": 0.2682,
+      "step": 375
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.1445833333333333e-05,
+      "loss": 0.2901,
+      "step": 400
+    },
+    {
+      "epoch": 2.01,
+      "eval_loss": 0.3911706209182739,
+      "eval_runtime": 200.4943,
+      "eval_samples_per_second": 4.584,
+      "eval_steps_per_second": 0.289,
+      "eval_wer": 21.460800902425266,
+      "eval_wer_normalized": 17.706103420389134,
+      "step": 400
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.0734027777777778e-05,
+      "loss": 0.1471,
+      "step": 425
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.0022222222222222e-05,
+      "loss": 0.1668,
+      "step": 450
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 9.310416666666667e-06,
+      "loss": 0.1635,
+      "step": 475
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 8.598611111111112e-06,
+      "loss": 0.1408,
+      "step": 500
+    },
+    {
+      "epoch": 2.51,
+      "eval_loss": 0.4062696099281311,
+      "eval_runtime": 205.1449,
+      "eval_samples_per_second": 4.48,
+      "eval_steps_per_second": 0.283,
+      "eval_wer": 21.611205113743186,
+      "eval_wer_normalized": 18.001046572475143,
+      "step": 500
+    },
+    {
+      "epoch": 2.51,
+      "step": 500,
+      "total_flos": 1.03412777877504e+18,
+      "train_loss": 0.48756390714645387,
+      "train_runtime": 1663.348,
+      "train_samples_per_second": 15.391,
+      "train_steps_per_second": 0.481
+    }
+  ],
+  "logging_steps": 25,
+  "max_steps": 800,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 100,
+  "total_flos": 1.03412777877504e+18,
+  "train_batch_size": 32,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1c13d80fe7e65b05fd121ddaa3246248106b55d931d19216900110889129c6a0
+oid sha256:d60ac0f27c9a558fab576051cf409b35da692c0a6fd22dba17e6b553a1bd109f
 size 4984