End of training

README.md

@@ -3,6 +3,8 @@ library_name: transformers
 license: apache-2.0
 base_model: facebook/wav2vec2-xls-r-1b
 tags:
+- automatic-speech-recognition
+- genbed
 - generated_from_trainer
 metrics:
 - wer
@@ -16,7 +18,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # wav2vec2-xslr-tr-testv2
 
-This model is a fine-tuned version of [facebook/wav2vec2-xls-r-1b](https://huggingface.co/facebook/wav2vec2-xls-r-1b) on an unknown dataset.
+This model is a fine-tuned version of [facebook/wav2vec2-xls-r-1b](https://huggingface.co/facebook/wav2vec2-xls-r-1b) on the GENBED - BEM dataset.
 It achieves the following results on the evaluation set:
 - Loss: 0.2789
 - Wer: 0.4783

all_results.json

@@ -0,0 +1,15 @@
+{
+    "epoch": 4.9965635738831615,
+    "eval_loss": 0.2789459824562073,
+    "eval_runtime": 117.3253,
+    "eval_samples": 1939,
+    "eval_samples_per_second": 16.527,
+    "eval_steps_per_second": 2.071,
+    "eval_wer": 0.47827027027027025,
+    "total_flos": 1.709034260097465e+19,
+    "train_loss": 0.6249774051335688,
+    "train_runtime": 7962.6969,
+    "train_samples": 5817,
+    "train_samples_per_second": 3.653,
+    "train_steps_per_second": 0.457
+}

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 4.9965635738831615,
+    "eval_loss": 0.2789459824562073,
+    "eval_runtime": 117.3253,
+    "eval_samples": 1939,
+    "eval_samples_per_second": 16.527,
+    "eval_steps_per_second": 2.071,
+    "eval_wer": 0.47827027027027025
+}

runs/Sep26_10-16-45_6164405edb41/events.out.tfevents.1727354013.6164405edb41.25030.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e10d82d19b1edada1e25dad9c63a55f6f85d746b6347c69e4d3ce3c6b89d0177
+size 406

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 4.9965635738831615,
+    "total_flos": 1.709034260097465e+19,
+    "train_loss": 0.6249774051335688,
+    "train_runtime": 7962.6969,
+    "train_samples": 5817,
+    "train_samples_per_second": 3.653,
+    "train_steps_per_second": 0.457
+}

trainer_state.json

@@ -0,0 +1,424 @@
+{
+  "best_metric": 0.2789459824562073,
+  "best_model_checkpoint": "./wav2vec2-xslr-tr-testv2/checkpoint-3600",
+  "epoch": 4.9965635738831615,
+  "eval_steps": 100,
+  "global_step": 3635,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.13745704467353953,
+      "eval_loss": 2.964635133743286,
+      "eval_runtime": 117.8396,
+      "eval_samples_per_second": 16.455,
+      "eval_steps_per_second": 2.062,
+      "eval_wer": 1.0,
+      "step": 100
+    },
+    {
+      "epoch": 0.27491408934707906,
+      "eval_loss": 0.968873143196106,
+      "eval_runtime": 117.4138,
+      "eval_samples_per_second": 16.514,
+      "eval_steps_per_second": 2.07,
+      "eval_wer": 0.9848108108108108,
+      "step": 200
+    },
+    {
+      "epoch": 0.41237113402061853,
+      "eval_loss": 0.8560894131660461,
+      "eval_runtime": 116.5948,
+      "eval_samples_per_second": 16.63,
+      "eval_steps_per_second": 2.084,
+      "eval_wer": 0.916972972972973,
+      "step": 300
+    },
+    {
+      "epoch": 0.5498281786941581,
+      "eval_loss": 0.7969807982444763,
+      "eval_runtime": 116.8936,
+      "eval_samples_per_second": 16.588,
+      "eval_steps_per_second": 2.079,
+      "eval_wer": 0.912,
+      "step": 400
+    },
+    {
+      "epoch": 0.6872852233676976,
+      "grad_norm": 2.9738190174102783,
+      "learning_rate": 0.0002988,
+      "loss": 1.9898,
+      "step": 500
+    },
+    {
+      "epoch": 0.6872852233676976,
+      "eval_loss": 0.8464106321334839,
+      "eval_runtime": 116.9601,
+      "eval_samples_per_second": 16.578,
+      "eval_steps_per_second": 2.078,
+      "eval_wer": 0.9257837837837838,
+      "step": 500
+    },
+    {
+      "epoch": 0.8247422680412371,
+      "eval_loss": 0.7357563376426697,
+      "eval_runtime": 116.2573,
+      "eval_samples_per_second": 16.679,
+      "eval_steps_per_second": 2.09,
+      "eval_wer": 0.8871891891891892,
+      "step": 600
+    },
+    {
+      "epoch": 0.9621993127147767,
+      "eval_loss": 0.6373892426490784,
+      "eval_runtime": 116.0248,
+      "eval_samples_per_second": 16.712,
+      "eval_steps_per_second": 2.094,
+      "eval_wer": 0.8608108108108108,
+      "step": 700
+    },
+    {
+      "epoch": 1.0996563573883162,
+      "eval_loss": 0.5179955959320068,
+      "eval_runtime": 117.0772,
+      "eval_samples_per_second": 16.562,
+      "eval_steps_per_second": 2.076,
+      "eval_wer": 0.7297297297297297,
+      "step": 800
+    },
+    {
+      "epoch": 1.2371134020618557,
+      "eval_loss": 0.4851646423339844,
+      "eval_runtime": 117.0035,
+      "eval_samples_per_second": 16.572,
+      "eval_steps_per_second": 2.077,
+      "eval_wer": 0.7211891891891892,
+      "step": 900
+    },
+    {
+      "epoch": 1.3745704467353952,
+      "grad_norm": 1.0166035890579224,
+      "learning_rate": 0.0002523444976076555,
+      "loss": 0.663,
+      "step": 1000
+    },
+    {
+      "epoch": 1.3745704467353952,
+      "eval_loss": 0.4840071201324463,
+      "eval_runtime": 116.2912,
+      "eval_samples_per_second": 16.674,
+      "eval_steps_per_second": 2.09,
+      "eval_wer": 0.7277837837837838,
+      "step": 1000
+    },
+    {
+      "epoch": 1.5120274914089347,
+      "eval_loss": 0.462556928396225,
+      "eval_runtime": 115.855,
+      "eval_samples_per_second": 16.736,
+      "eval_steps_per_second": 2.097,
+      "eval_wer": 0.7135135135135136,
+      "step": 1100
+    },
+    {
+      "epoch": 1.6494845360824741,
+      "eval_loss": 0.4492775499820709,
+      "eval_runtime": 116.5354,
+      "eval_samples_per_second": 16.639,
+      "eval_steps_per_second": 2.085,
+      "eval_wer": 0.676,
+      "step": 1200
+    },
+    {
+      "epoch": 1.7869415807560136,
+      "eval_loss": 0.44191744923591614,
+      "eval_runtime": 118.2446,
+      "eval_samples_per_second": 16.398,
+      "eval_steps_per_second": 2.055,
+      "eval_wer": 0.6812972972972973,
+      "step": 1300
+    },
+    {
+      "epoch": 1.9243986254295533,
+      "eval_loss": 0.4306336045265198,
+      "eval_runtime": 116.8028,
+      "eval_samples_per_second": 16.601,
+      "eval_steps_per_second": 2.08,
+      "eval_wer": 0.6749189189189189,
+      "step": 1400
+    },
+    {
+      "epoch": 2.0618556701030926,
+      "grad_norm": 2.287825107574463,
+      "learning_rate": 0.00020449760765550238,
+      "loss": 0.5455,
+      "step": 1500
+    },
+    {
+      "epoch": 2.0618556701030926,
+      "eval_loss": 0.43293190002441406,
+      "eval_runtime": 116.7385,
+      "eval_samples_per_second": 16.61,
+      "eval_steps_per_second": 2.082,
+      "eval_wer": 0.6846486486486486,
+      "step": 1500
+    },
+    {
+      "epoch": 2.1993127147766325,
+      "eval_loss": 0.4227151870727539,
+      "eval_runtime": 117.2164,
+      "eval_samples_per_second": 16.542,
+      "eval_steps_per_second": 2.073,
+      "eval_wer": 0.6685405405405406,
+      "step": 1600
+    },
+    {
+      "epoch": 2.336769759450172,
+      "eval_loss": 0.40970557928085327,
+      "eval_runtime": 117.3272,
+      "eval_samples_per_second": 16.526,
+      "eval_steps_per_second": 2.071,
+      "eval_wer": 0.6472432432432432,
+      "step": 1700
+    },
+    {
+      "epoch": 2.4742268041237114,
+      "eval_loss": 0.40352100133895874,
+      "eval_runtime": 116.7738,
+      "eval_samples_per_second": 16.605,
+      "eval_steps_per_second": 2.081,
+      "eval_wer": 0.6343243243243243,
+      "step": 1800
+    },
+    {
+      "epoch": 2.611683848797251,
+      "eval_loss": 0.4040600657463074,
+      "eval_runtime": 116.603,
+      "eval_samples_per_second": 16.629,
+      "eval_steps_per_second": 2.084,
+      "eval_wer": 0.6304324324324324,
+      "step": 1900
+    },
+    {
+      "epoch": 2.7491408934707904,
+      "grad_norm": 1.1287260055541992,
+      "learning_rate": 0.00015665071770334927,
+      "loss": 0.433,
+      "step": 2000
+    },
+    {
+      "epoch": 2.7491408934707904,
+      "eval_loss": 0.3962455093860626,
+      "eval_runtime": 116.8792,
+      "eval_samples_per_second": 16.59,
+      "eval_steps_per_second": 2.079,
+      "eval_wer": 0.6542162162162162,
+      "step": 2000
+    },
+    {
+      "epoch": 2.88659793814433,
+      "eval_loss": 0.36006543040275574,
+      "eval_runtime": 117.76,
+      "eval_samples_per_second": 16.466,
+      "eval_steps_per_second": 2.064,
+      "eval_wer": 0.6040540540540541,
+      "step": 2100
+    },
+    {
+      "epoch": 3.0240549828178693,
+      "eval_loss": 0.3472883105278015,
+      "eval_runtime": 117.2756,
+      "eval_samples_per_second": 16.534,
+      "eval_steps_per_second": 2.072,
+      "eval_wer": 0.5863783783783784,
+      "step": 2200
+    },
+    {
+      "epoch": 3.161512027491409,
+      "eval_loss": 0.345599502325058,
+      "eval_runtime": 117.5317,
+      "eval_samples_per_second": 16.498,
+      "eval_steps_per_second": 2.068,
+      "eval_wer": 0.5723243243243243,
+      "step": 2300
+    },
+    {
+      "epoch": 3.2989690721649483,
+      "eval_loss": 0.3379517197608948,
+      "eval_runtime": 117.4759,
+      "eval_samples_per_second": 16.506,
+      "eval_steps_per_second": 2.069,
+      "eval_wer": 0.5616756756756757,
+      "step": 2400
+    },
+    {
+      "epoch": 3.436426116838488,
+      "grad_norm": 0.6922096610069275,
+      "learning_rate": 0.00010880382775119617,
+      "loss": 0.3509,
+      "step": 2500
+    },
+    {
+      "epoch": 3.436426116838488,
+      "eval_loss": 0.32671377062797546,
+      "eval_runtime": 118.4317,
+      "eval_samples_per_second": 16.372,
+      "eval_steps_per_second": 2.052,
+      "eval_wer": 0.5562702702702703,
+      "step": 2500
+    },
+    {
+      "epoch": 3.5738831615120272,
+      "eval_loss": 0.3208204209804535,
+      "eval_runtime": 117.733,
+      "eval_samples_per_second": 16.469,
+      "eval_steps_per_second": 2.064,
+      "eval_wer": 0.556972972972973,
+      "step": 2600
+    },
+    {
+      "epoch": 3.711340206185567,
+      "eval_loss": 0.3124404549598694,
+      "eval_runtime": 118.1827,
+      "eval_samples_per_second": 16.407,
+      "eval_steps_per_second": 2.056,
+      "eval_wer": 0.5397297297297298,
+      "step": 2700
+    },
+    {
+      "epoch": 3.8487972508591066,
+      "eval_loss": 0.30383118987083435,
+      "eval_runtime": 117.8757,
+      "eval_samples_per_second": 16.45,
+      "eval_steps_per_second": 2.061,
+      "eval_wer": 0.5271891891891892,
+      "step": 2800
+    },
+    {
+      "epoch": 3.986254295532646,
+      "eval_loss": 0.2993951141834259,
+      "eval_runtime": 118.0952,
+      "eval_samples_per_second": 16.419,
+      "eval_steps_per_second": 2.058,
+      "eval_wer": 0.5253513513513514,
+      "step": 2900
+    },
+    {
+      "epoch": 4.123711340206185,
+      "grad_norm": 0.5013633966445923,
+      "learning_rate": 6.0956937799043056e-05,
+      "loss": 0.2871,
+      "step": 3000
+    },
+    {
+      "epoch": 4.123711340206185,
+      "eval_loss": 0.307333379983902,
+      "eval_runtime": 117.9359,
+      "eval_samples_per_second": 16.441,
+      "eval_steps_per_second": 2.06,
+      "eval_wer": 0.5247027027027027,
+      "step": 3000
+    },
+    {
+      "epoch": 4.261168384879725,
+      "eval_loss": 0.3009079694747925,
+      "eval_runtime": 116.9193,
+      "eval_samples_per_second": 16.584,
+      "eval_steps_per_second": 2.078,
+      "eval_wer": 0.5122162162162163,
+      "step": 3100
+    },
+    {
+      "epoch": 4.398625429553265,
+      "eval_loss": 0.29754796624183655,
+      "eval_runtime": 117.728,
+      "eval_samples_per_second": 16.47,
+      "eval_steps_per_second": 2.064,
+      "eval_wer": 0.4952972972972973,
+      "step": 3200
+    },
+    {
+      "epoch": 4.536082474226804,
+      "eval_loss": 0.2898379862308502,
+      "eval_runtime": 117.9926,
+      "eval_samples_per_second": 16.433,
+      "eval_steps_per_second": 2.059,
+      "eval_wer": 0.49383783783783786,
+      "step": 3300
+    },
+    {
+      "epoch": 4.673539518900344,
+      "eval_loss": 0.28349989652633667,
+      "eval_runtime": 117.4549,
+      "eval_samples_per_second": 16.508,
+      "eval_steps_per_second": 2.069,
+      "eval_wer": 0.4902162162162162,
+      "step": 3400
+    },
+    {
+      "epoch": 4.810996563573883,
+      "grad_norm": 0.7447757124900818,
+      "learning_rate": 1.3110047846889951e-05,
+      "loss": 0.2198,
+      "step": 3500
+    },
+    {
+      "epoch": 4.810996563573883,
+      "eval_loss": 0.28041791915893555,
+      "eval_runtime": 117.7449,
+      "eval_samples_per_second": 16.468,
+      "eval_steps_per_second": 2.064,
+      "eval_wer": 0.48021621621621624,
+      "step": 3500
+    },
+    {
+      "epoch": 4.948453608247423,
+      "eval_loss": 0.2789459824562073,
+      "eval_runtime": 117.7982,
+      "eval_samples_per_second": 16.46,
+      "eval_steps_per_second": 2.063,
+      "eval_wer": 0.47827027027027025,
+      "step": 3600
+    },
+    {
+      "epoch": 4.9965635738831615,
+      "step": 3635,
+      "total_flos": 1.709034260097465e+19,
+      "train_loss": 0.6249774051335688,
+      "train_runtime": 7962.6969,
+      "train_samples_per_second": 3.653,
+      "train_steps_per_second": 0.457
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 3635,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 400,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 3,
+        "early_stopping_threshold": 0.0
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 0
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.709034260097465e+19,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}