jonatasgrosman/exp_w2v2r_es_xls-r_age_teens-10_sixties-0_s337

.gitattributes

@@ -32,3 +32,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst 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_age_teens-10_sixties-0_s337
+
+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": 0.036707305651812225,
+      "wer": 0.1348314606741573
+    },
+    {
+      "subset_type": "age",
+      "subset": "",
+      "cer": 0.036711354231619524,
+      "wer": 0.13489277426922455
+    },
+    {
+      "subset_type": "age",
+      "subset": "twenties",
+      "cer": 0.03727632971522804,
+      "wer": 0.1367345438506212
+    },
+    {
+      "subset_type": "age",
+      "subset": "fourties",
+      "cer": 0.030383646037524505,
+      "wer": 0.11710388247639035
+    },
+    {
+      "subset_type": "age",
+      "subset": "thirties",
+      "cer": 0.035913010707397636,
+      "wer": 0.1334742962865072
+    },
+    {
+      "subset_type": "age",
+      "subset": "fifties",
+      "cer": 0.034646830183621496,
+      "wer": 0.12479935794542536
+    },
+    {
+      "subset_type": "age",
+      "subset": "sixties",
+      "cer": 0.05181449619301889,
+      "wer": 0.1594626168224299
+    },
+    {
+      "subset_type": "age",
+      "subset": "seventies",
+      "cer": 0.029838709677419355,
+      "wer": 0.12264150943396226
+    },
+    {
+      "subset_type": "age",
+      "subset": "teens",
+      "cer": 0.039305522064142756,
+      "wer": 0.14836112708453134
+    },
+    {
+      "subset_type": "gender",
+      "subset": "",
+      "cer": 0.0366670437832334,
+      "wer": 0.13470786297675286
+    },
+    {
+      "subset_type": "gender",
+      "subset": "other",
+      "cer": 0.03834541062801932,
+      "wer": 0.1518987341772152
+    },
+    {
+      "subset_type": "gender",
+      "subset": "male",
+      "cer": 0.037554517234898804,
+      "wer": 0.1364252780037971
+    },
+    {
+      "subset_type": "gender",
+      "subset": "female",
+      "cer": 0.03419381943796097,
+      "wer": 0.12983770287141075
+    },
+    {
+      "subset_type": "accent",
+      "subset": "",
+      "cer": 0.03658256804694493,
+      "wer": 0.13472363566976212
+    },
+    {
+      "subset_type": "accent",
+      "subset": "caribe",
+      "cer": 0.03607904085257549,
+      "wer": 0.13331105913798863
+    },
+    {
+      "subset_type": "accent",
+      "subset": "mexicano",
+      "cer": 0.03610591684558026,
+      "wer": 0.13144436765722184
+    },
+    {
+      "subset_type": "accent",
+      "subset": "surpeninsular",
+      "cer": 0.04573330997021202,
+      "wer": 0.1671891327063741
+    },
+    {
+      "subset_type": "accent",
+      "subset": "chileno",
+      "cer": 0.043024469442189496,
+      "wer": 0.16244002741603839
+    },
+    {
+      "subset_type": "accent",
+      "subset": "canario",
+      "cer": 0.030814380044020543,
+      "wer": 0.13302752293577982
+    },
+    {
+      "subset_type": "accent",
+      "subset": "americacentral",
+      "cer": 0.03797347581337659,
+      "wer": 0.12990762124711316
+    },
+    {
+      "subset_type": "accent",
+      "subset": "nortepeninsular",
+      "cer": 0.03346681922196797,
+      "wer": 0.11764705882352941
+    },
+    {
+      "subset_type": "accent",
+      "subset": "andino",
+      "cer": 0.03288518829730349,
+      "wer": 0.12394814646349785
+    },
+    {
+      "subset_type": "accent",
+      "subset": "rioplatense",
+      "cer": 0.0374020422702446,
+      "wer": 0.1400429799426934
+    },
+    {
+      "subset_type": "accent",
+      "subset": "centrosurpeninsular",
+      "cer": 0.04812234766594603,
+      "wer": 0.15557638238050608
+    }
+  ],
+  "test": [
+    {
+      "subset_type": "all",
+      "subset": "all",
+      "cer": 0.041758365345176615,
+      "wer": 0.14745076275609642
+    },
+    {
+      "subset_type": "age",
+      "subset": "",
+      "cer": 0.04205854325488896,
+      "wer": 0.1475558659217877
+    },
+    {
+      "subset_type": "age",
+      "subset": "twenties",
+      "cer": 0.040533226420767766,
+      "wer": 0.154307655387569
+    },
+    {
+      "subset_type": "age",
+      "subset": "fourties",
+      "cer": 0.035038242960304236,
+      "wer": 0.13178893178893178
+    },
+    {
+      "subset_type": "age",
+      "subset": "thirties",
+      "cer": 0.038353637446677684,
+      "wer": 0.14045977011494254
+    },
+    {
+      "subset_type": "age",
+      "subset": "fifties",
+      "cer": 0.04093447759160877,
+      "wer": 0.1364553895841744
+    },
+    {
+      "subset_type": "age",
+      "subset": "sixties",
+      "cer": 0.04255732607850758,
+      "wer": 0.14763552479815456
+    },
+    {
+      "subset_type": "age",
+      "subset": "seventies",
+      "cer": 0.09563409563409564,
+      "wer": 0.2926829268292683
+    },
+    {
+      "subset_type": "age",
+      "subset": "teens",
+      "cer": 0.0459661656669138,
+      "wer": 0.16320716320716322
+    },
+    {
+      "subset_type": "age",
+      "subset": "eighties",
+      "cer": 0.006711409395973154,
+      "wer": 0.037037037037037035
+    },
+    {
+      "subset_type": "gender",
+      "subset": "",
+      "cer": 0.04211231069687848,
+      "wer": 0.14780217652298516
+    },
+    {
+      "subset_type": "gender",
+      "subset": "other",
+      "cer": 0.0746054519368723,
+      "wer": 0.25217391304347825
+    },
+    {
+      "subset_type": "gender",
+      "subset": "male",
+      "cer": 0.03959104393958428,
+      "wer": 0.14525268879380132
+    },
+    {
+      "subset_type": "gender",
+      "subset": "female",
+      "cer": 0.038399210878602125,
+      "wer": 0.1408033826638478
+    },
+    {
+      "subset_type": "accent",
+      "subset": "",
+      "cer": 0.042201146203970974,
+      "wer": 0.14827777572237227
+    },
+    {
+      "subset_type": "accent",
+      "subset": "caribe",
+      "cer": 0.04117389399912396,
+      "wer": 0.15098374679213003
+    },
+    {
+      "subset_type": "accent",
+      "subset": "surpeninsular",
+      "cer": 0.045298726738491675,
+      "wer": 0.16592592592592592
+    },
+    {
+      "subset_type": "accent",
+      "subset": "mexicano",
+      "cer": 0.037990867579908674,
+      "wer": 0.1421376608841634
+    },
+    {
+      "subset_type": "accent",
+      "subset": "chileno",
+      "cer": 0.03146206045650833,
+      "wer": 0.1161939615736505
+    },
+    {
+      "subset_type": "accent",
+      "subset": "canario",
+      "cer": 0.04218106995884774,
+      "wer": 0.14012738853503184
+    },
+    {
+      "subset_type": "accent",
+      "subset": "americacentral",
+      "cer": 0.037226373050203175,
+      "wer": 0.13479623824451412
+    },
+    {
+      "subset_type": "accent",
+      "subset": "nortepeninsular",
+      "cer": 0.0350479332027626,
+      "wer": 0.12871888281724347
+    },
+    {
+      "subset_type": "accent",
+      "subset": "andino",
+      "cer": 0.03551806426549753,
+      "wer": 0.14233210825258938
+    },
+    {
+      "subset_type": "accent",
+      "subset": "rioplatense",
+      "cer": 0.05075757575757576,
+      "wer": 0.1558538404175988
+    },
+    {
+      "subset_type": "accent",
+      "subset": "centrosurpeninsular",
+      "cer": 0.03028458377720286,
+      "wer": 0.12321428571428572
+    }
+  ]
+}

cv7_transcriptions.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3849612eb341ae481a412d0a9a5ee5f22f5d6ea55ddec25cc7d07162e8fa450b
+size 125573577

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 sha256:6c992e7f005a9bb4eb179611b4f3ce707682570fd901dc51e51095549d25bc58
+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/tmpznq3xze5", "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}