jonatasgrosman/exp_w2v2r_es_xls-r_accent_surpeninsular-5_nortepeninsular-5_s240

.gitattributes

@@ -29,3 +29,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zstandard filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+cv7_transcriptions.json filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,17 @@
+---
+language:
+- es
+license: apache-2.0
+tags:
+- automatic-speech-recognition
+- es
+datasets:
+- mozilla-foundation/common_voice_7_0
+---
+# exp_w2v2r_es_xls-r_accent_surpeninsular-5_nortepeninsular-5_s240
+
+Fine-tuned [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) for speech recognition using the train split of [Common Voice 7.0 (es)](https://huggingface.co/datasets/mozilla-foundation/common_voice_7_0).
+When using this model, make sure that your speech input is sampled at 16kHz.
+
+This model has been fine-tuned by the [HuggingSound](https://github.com/jonatasgrosman/huggingsound) tool.
+    

config.json

@@ -0,0 +1,108 @@
+{
+  "_name_or_path": "facebook/wav2vec2-xls-r-300m",
+  "activation_dropout": 0.05,
+  "adapter_kernel_size": 3,
+  "adapter_stride": 2,
+  "add_adapter": false,
+  "apply_spec_augment": true,
+  "architectures": [
+    "Wav2Vec2ForCTC"
+  ],
+  "attention_dropout": 0.05,
+  "bos_token_id": 1,
+  "classifier_proj_size": 256,
+  "codevector_dim": 768,
+  "contrastive_logits_temperature": 0.1,
+  "conv_bias": true,
+  "conv_dim": [
+    512,
+    512,
+    512,
+    512,
+    512,
+    512,
+    512
+  ],
+  "conv_kernel": [
+    10,
+    3,
+    3,
+    3,
+    3,
+    2,
+    2
+  ],
+  "conv_stride": [
+    5,
+    2,
+    2,
+    2,
+    2,
+    2,
+    2
+  ],
+  "ctc_loss_reduction": "mean",
+  "ctc_zero_infinity": false,
+  "diversity_loss_weight": 0.1,
+  "do_stable_layer_norm": true,
+  "eos_token_id": 2,
+  "feat_extract_activation": "gelu",
+  "feat_extract_dropout": 0.0,
+  "feat_extract_norm": "layer",
+  "feat_proj_dropout": 0.05,
+  "feat_quantizer_dropout": 0.0,
+  "final_dropout": 0.05,
+  "gradient_checkpointing": false,
+  "hidden_act": "gelu",
+  "hidden_dropout": 0.05,
+  "hidden_size": 1024,
+  "initializer_range": 0.02,
+  "intermediate_size": 4096,
+  "layer_norm_eps": 1e-05,
+  "layerdrop": 0.05,
+  "mask_feature_length": 10,
+  "mask_feature_min_masks": 0,
+  "mask_feature_prob": 0.0,
+  "mask_time_length": 10,
+  "mask_time_min_masks": 2,
+  "mask_time_prob": 0.05,
+  "model_type": "wav2vec2",
+  "num_adapter_layers": 3,
+  "num_attention_heads": 16,
+  "num_codevector_groups": 2,
+  "num_codevectors_per_group": 320,
+  "num_conv_pos_embedding_groups": 16,
+  "num_conv_pos_embeddings": 128,
+  "num_feat_extract_layers": 7,
+  "num_hidden_layers": 24,
+  "num_negatives": 100,
+  "output_hidden_size": 1024,
+  "pad_token_id": 0,
+  "proj_codevector_dim": 768,
+  "tdnn_dilation": [
+    1,
+    2,
+    3,
+    1,
+    1
+  ],
+  "tdnn_dim": [
+    512,
+    512,
+    512,
+    512,
+    1500
+  ],
+  "tdnn_kernel": [
+    5,
+    3,
+    3,
+    1,
+    1
+  ],
+  "torch_dtype": "float32",
+  "transformers_version": "4.18.0",
+  "use_weighted_layer_sum": false,
+  "vocab_size": 41,
+  "xvector_output_dim": 512
+}

cv7_evaluation.json

@@ -0,0 +1,300 @@
+{
+  "validation": [
+    {
+      "subset_type": "all",
+      "subset": "all",
+      "cer": 0.04480241520078009,
+      "wer": 0.16195091036322107
+    },
+    {
+      "subset_type": "age",
+      "subset": "",
+      "cer": 0.04449152542372881,
+      "wer": 0.16124233950139824
+    },
+    {
+      "subset_type": "age",
+      "subset": "sixties",
+      "cer": 0.053099970335212104,
+      "wer": 0.16238317757009346
+    },
+    {
+      "subset_type": "age",
+      "subset": "seventies",
+      "cer": 0.03790322580645161,
+      "wer": 0.15566037735849056
+    },
+    {
+      "subset_type": "age",
+      "subset": "fourties",
+      "cer": 0.039414729767572106,
+      "wer": 0.13746065057712487
+    },
+    {
+      "subset_type": "age",
+      "subset": "thirties",
+      "cer": 0.04504644306014321,
+      "wer": 0.16624818289943175
+    },
+    {
+      "subset_type": "age",
+      "subset": "teens",
+      "cer": 0.05001205690860863,
+      "wer": 0.1802760207015526
+    },
+    {
+      "subset_type": "age",
+      "subset": "fifties",
+      "cer": 0.04148237501675379,
+      "wer": 0.14486356340288925
+    },
+    {
+      "subset_type": "age",
+      "subset": "twenties",
+      "cer": 0.047613773413079326,
+      "wer": 0.17275600970374183
+    },
+    {
+      "subset_type": "gender",
+      "subset": "",
+      "cer": 0.044423860165177055,
+      "wer": 0.161067621091199
+    },
+    {
+      "subset_type": "gender",
+      "subset": "female",
+      "cer": 0.04261973671926057,
+      "wer": 0.15813566375364127
+    },
+    {
+      "subset_type": "gender",
+      "subset": "male",
+      "cer": 0.046912364152447365,
+      "wer": 0.16629857801542097
+    },
+    {
+      "subset_type": "gender",
+      "subset": "other",
+      "cer": 0.05555555555555555,
+      "wer": 0.19710669077757687
+    },
+    {
+      "subset_type": "accent",
+      "subset": "",
+      "cer": 0.04447170374299639,
+      "wer": 0.16122312682864776
+    },
+    {
+      "subset_type": "accent",
+      "subset": "rioplatense",
+      "cer": 0.047613393493232015,
+      "wer": 0.17979942693409742
+    },
+    {
+      "subset_type": "accent",
+      "subset": "mexicano",
+      "cer": 0.045247824179883184,
+      "wer": 0.16779543883897718
+    },
+    {
+      "subset_type": "accent",
+      "subset": "americacentral",
+      "cer": 0.052380498044079765,
+      "wer": 0.18302540415704388
+    },
+    {
+      "subset_type": "accent",
+      "subset": "canario",
+      "cer": 0.04035216434336023,
+      "wer": 0.18807339449541285
+    },
+    {
+      "subset_type": "accent",
+      "subset": "andino",
+      "cer": 0.04541287907722863,
+      "wer": 0.15806231521491926
+    },
+    {
+      "subset_type": "accent",
+      "subset": "nortepeninsular",
+      "cer": 0.03537376048817696,
+      "wer": 0.12549019607843137
+    },
+    {
+      "subset_type": "accent",
+      "subset": "surpeninsular",
+      "cer": 0.03942526721570002,
+      "wer": 0.13793103448275862
+    },
+    {
+      "subset_type": "accent",
+      "subset": "centrosurpeninsular",
+      "cer": 0.04788213627992634,
+      "wer": 0.15089034676663543
+    },
+    {
+      "subset_type": "accent",
+      "subset": "chileno",
+      "cer": 0.06215934129653253,
+      "wer": 0.20356408498971898
+    },
+    {
+      "subset_type": "accent",
+      "subset": "caribe",
+      "cer": 0.04673623445825933,
+      "wer": 0.17039759438690277
+    }
+  ],
+  "test": [
+    {
+      "subset_type": "all",
+      "subset": "all",
+      "cer": 0.05221091210598971,
+      "wer": 0.1770756942104533
+    },
+    {
+      "subset_type": "age",
+      "subset": "",
+      "cer": 0.05248643105734173,
+      "wer": 0.17766139044072005
+    },
+    {
+      "subset_type": "age",
+      "subset": "sixties",
+      "cer": 0.05207928488146133,
+      "wer": 0.16493656286043828
+    },
+    {
+      "subset_type": "age",
+      "subset": "seventies",
+      "cer": 0.10395010395010396,
+      "wer": 0.2804878048780488
+    },
+    {
+      "subset_type": "age",
+      "subset": "fourties",
+      "cer": 0.0494808357902833,
+      "wer": 0.16036036036036036
+    },
+    {
+      "subset_type": "age",
+      "subset": "thirties",
+      "cer": 0.045809154144729254,
+      "wer": 0.16597701149425287
+    },
+    {
+      "subset_type": "age",
+      "subset": "fifties",
+      "cer": 0.04638332652227217,
+      "wer": 0.14735567218409365
+    },
+    {
+      "subset_type": "age",
+      "subset": "teens",
+      "cer": 0.06099615825301611,
+      "wer": 0.21001221001221002
+    },
+    {
+      "subset_type": "age",
+      "subset": "twenties",
+      "cer": 0.05131803147238649,
+      "wer": 0.18142548596112312
+    },
+    {
+      "subset_type": "age",
+      "subset": "eighties",
+      "cer": 0.006711409395973154,
+      "wer": 0.037037037037037035
+    },
+    {
+      "subset_type": "gender",
+      "subset": "",
+      "cer": 0.052533756929428084,
+      "wer": 0.17786297974516865
+    },
+    {
+      "subset_type": "gender",
+      "subset": "female",
+      "cer": 0.04791094201366871,
+      "wer": 0.16553911205073996
+    },
+    {
+      "subset_type": "gender",
+      "subset": "male",
+      "cer": 0.05065566324903607,
+      "wer": 0.17300797964612005
+    },
+    {
+      "subset_type": "gender",
+      "subset": "other",
+      "cer": 0.07604017216642754,
+      "wer": 0.2782608695652174
+    },
+    {
+      "subset_type": "accent",
+      "subset": "",
+      "cer": 0.05280513791732615,
+      "wer": 0.17857116430500575
+    },
+    {
+      "subset_type": "accent",
+      "subset": "rioplatense",
+      "cer": 0.06136363636363636,
+      "wer": 0.19463087248322147
+    },
+    {
+      "subset_type": "accent",
+      "subset": "mexicano",
+      "cer": 0.04844748858447489,
+      "wer": 0.1667599328483492
+    },
+    {
+      "subset_type": "accent",
+      "subset": "americacentral",
+      "cer": 0.04980993577139861,
+      "wer": 0.1755485893416928
+    },
+    {
+      "subset_type": "accent",
+      "subset": "canario",
+      "cer": 0.04218106995884774,
+      "wer": 0.15286624203821655
+    },
+    {
+      "subset_type": "accent",
+      "subset": "andino",
+      "cer": 0.04667295632388035,
+      "wer": 0.16805880387570998
+    },
+    {
+      "subset_type": "accent",
+      "subset": "nortepeninsular",
+      "cer": 0.03927430161838986,
+      "wer": 0.13721918639951428
+    },
+    {
+      "subset_type": "accent",
+      "subset": "surpeninsular",
+      "cer": 0.039177277179236046,
+      "wer": 0.12296296296296297
+    },
+    {
+      "subset_type": "accent",
+      "subset": "centrosurpeninsular",
+      "cer": 0.032262973672196014,
+      "wer": 0.11875
+    },
+    {
+      "subset_type": "accent",
+      "subset": "chileno",
+      "cer": 0.041178285009253544,
+      "wer": 0.1637694419030192
+    },
+    {
+      "subset_type": "accent",
+      "subset": "caribe",
+      "cer": 0.05497152869031975,
+      "wer": 0.1881950384944397
+    }
+  ]
+}

cv7_transcriptions.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a27f6507424fa8fcf4632fa3b51c01db15cdb1ac35c5d05ec3173a175ba1bfd9
+size 125644599

preprocessor_config.json

@@ -0,0 +1,10 @@
+{
+  "do_normalize": true,
+  "feature_extractor_type": "Wav2Vec2FeatureExtractor",
+  "feature_size": 1,
+  "padding_side": "right",
+  "padding_value": 0,
+  "processor_class": "Wav2Vec2Processor",
+  "return_attention_mask": true,
+  "sampling_rate": 16000
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6677d077a476796b4d995e13c8b04c78be24637526512fc9ec1e8ddd77421da9
+size 1262066801

special_tokens_map.json

@@ -0,0 +1 @@
+{"bos_token": "<s>", "eos_token": "</s>", "unk_token": "<unk>", "pad_token": "<pad>"}

tokenizer_config.json

@@ -0,0 +1 @@
+{"unk_token": "<unk>", "bos_token": "<s>", "eos_token": "</s>", "pad_token": "<pad>", "do_lower_case": false, "word_delimiter_token": "|", "replace_word_delimiter_char": " ", "special_tokens_map_file": null, "name_or_path": "/tmp/tmpdszfr531", "tokenizer_class": "Wav2Vec2CTCTokenizer", "processor_class": "Wav2Vec2Processor"}

vocab.json

@@ -0,0 +1 @@
+{"<pad>": 0, "<s>": 1, "</s>": 2, "<unk>": 3, "|": 4, "'": 5, "-": 6, "a": 7, "b": 8, "c": 9, "d": 10, "e": 11, "f": 12, "g": 13, "h": 14, "i": 15, "j": 16, "k": 17, "l": 18, "m": 19, "n": 20, "o": 21, "p": 22, "q": 23, "r": 24, "s": 25, "t": 26, "u": 27, "v": 28, "w": 29, "x": 30, "y": 31, "z": 32, "á": 33, "é": 34, "í": 35, "ñ": 36, "ó": 37, "ö": 38, "ú": 39, "ü": 40}