jonatasgrosman/exp_w2v2r_es_xls-r_age_teens-2_sixties-8_s786

.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-2_sixties-8_s786
+
+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": 0.0401427965424726,
+      "wer": 0.1457967493629438
+    },
+    {
+      "subset_type": "age",
+      "subset": "",
+      "cer": 0.040091264667535854,
+      "wer": 0.14530510246580933
+    },
+    {
+      "subset_type": "age",
+      "subset": "twenties",
+      "cer": 0.04221122843289276,
+      "wer": 0.15430419760346983
+    },
+    {
+      "subset_type": "age",
+      "subset": "fourties",
+      "cer": 0.034829179501540186,
+      "wer": 0.12528856243441763
+    },
+    {
+      "subset_type": "age",
+      "subset": "thirties",
+      "cer": 0.03910527832583298,
+      "wer": 0.14391436500594687
+    },
+    {
+      "subset_type": "age",
+      "subset": "fifties",
+      "cer": 0.03839967832730197,
+      "wer": 0.14004815409309793
+    },
+    {
+      "subset_type": "age",
+      "subset": "sixties",
+      "cer": 0.04756254326114902,
+      "wer": 0.15478971962616822
+    },
+    {
+      "subset_type": "age",
+      "subset": "seventies",
+      "cer": 0.031451612903225803,
+      "wer": 0.1179245283018868
+    },
+    {
+      "subset_type": "age",
+      "subset": "teens",
+      "cer": 0.04152399324813118,
+      "wer": 0.16273720529039679
+    },
+    {
+      "subset_type": "gender",
+      "subset": "",
+      "cer": 0.040021495644303656,
+      "wer": 0.1450974072786179
+    },
+    {
+      "subset_type": "gender",
+      "subset": "other",
+      "cer": 0.04921497584541063,
+      "wer": 0.18083182640144665
+    },
+    {
+      "subset_type": "gender",
+      "subset": "male",
+      "cer": 0.04130024820328043,
+      "wer": 0.14855283040799722
+    },
+    {
+      "subset_type": "gender",
+      "subset": "female",
+      "cer": 0.03727476426104005,
+      "wer": 0.1446802607851297
+    },
+    {
+      "subset_type": "accent",
+      "subset": "",
+      "cer": 0.040014587408414286,
+      "wer": 0.1453297926472459
+    },
+    {
+      "subset_type": "accent",
+      "subset": "caribe",
+      "cer": 0.041019094138543516,
+      "wer": 0.13732041430003342
+    },
+    {
+      "subset_type": "accent",
+      "subset": "mexicano",
+      "cer": 0.039707274280305654,
+      "wer": 0.1484450587422253
+    },
+    {
+      "subset_type": "accent",
+      "subset": "surpeninsular",
+      "cer": 0.039074820396004906,
+      "wer": 0.14211076280041798
+    },
+    {
+      "subset_type": "accent",
+      "subset": "chileno",
+      "cer": 0.050562449263597355,
+      "wer": 0.18711446196024675
+    },
+    {
+      "subset_type": "accent",
+      "subset": "canario",
+      "cer": 0.04035216434336023,
+      "wer": 0.16972477064220184
+    },
+    {
+      "subset_type": "accent",
+      "subset": "americacentral",
+      "cer": 0.04293483446236046,
+      "wer": 0.14607390300230946
+    },
+    {
+      "subset_type": "accent",
+      "subset": "nortepeninsular",
+      "cer": 0.03470633104500381,
+      "wer": 0.12717086834733893
+    },
+    {
+      "subset_type": "accent",
+      "subset": "andino",
+      "cer": 0.03727751890611871,
+      "wer": 0.1387309529224471
+    },
+    {
+      "subset_type": "accent",
+      "subset": "rioplatense",
+      "cer": 0.04019235336024697,
+      "wer": 0.1511461318051576
+    },
+    {
+      "subset_type": "accent",
+      "subset": "centrosurpeninsular",
+      "cer": 0.04956361598206422,
+      "wer": 0.17010309278350516
+    }
+  ],
+  "test": [
+    {
+      "subset_type": "all",
+      "subset": "all",
+      "cer": 0.04658219362350544,
+      "wer": 0.16020190408234727
+    },
+    {
+      "subset_type": "age",
+      "subset": "",
+      "cer": 0.04684263133907855,
+      "wer": 0.16044382371198013
+    },
+    {
+      "subset_type": "age",
+      "subset": "twenties",
+      "cer": 0.04662724265811366,
+      "wer": 0.16846652267818574
+    },
+    {
+      "subset_type": "age",
+      "subset": "fourties",
+      "cer": 0.040892193308550186,
+      "wer": 0.13925353925353925
+    },
+    {
+      "subset_type": "age",
+      "subset": "thirties",
+      "cer": 0.041428077322162866,
+      "wer": 0.1489655172413793
+    },
+    {
+      "subset_type": "age",
+      "subset": "fifties",
+      "cer": 0.04495300367797303,
+      "wer": 0.1485668146951958
+    },
+    {
+      "subset_type": "age",
+      "subset": "sixties",
+      "cer": 0.046055188495919164,
+      "wer": 0.1522491349480969
+    },
+    {
+      "subset_type": "age",
+      "subset": "seventies",
+      "cer": 0.09563409563409564,
+      "wer": 0.2804878048780488
+    },
+    {
+      "subset_type": "age",
+      "subset": "teens",
+      "cer": 0.0512906921884478,
+      "wer": 0.18315018315018314
+    },
+    {
+      "subset_type": "age",
+      "subset": "eighties",
+      "cer": 0.026845637583892617,
+      "wer": 0.14814814814814814
+    },
+    {
+      "subset_type": "gender",
+      "subset": "",
+      "cer": 0.04689327679066503,
+      "wer": 0.16064443669881104
+    },
+    {
+      "subset_type": "gender",
+      "subset": "other",
+      "cer": 0.07890961262553801,
+      "wer": 0.2565217391304348
+    },
+    {
+      "subset_type": "gender",
+      "subset": "male",
+      "cer": 0.04506054134495521,
+      "wer": 0.1561813345669018
+    },
+    {
+      "subset_type": "gender",
+      "subset": "female",
+      "cer": 0.04206298879729444,
+      "wer": 0.15813953488372093
+    },
+    {
+      "subset_type": "accent",
+      "subset": "",
+      "cer": 0.047083994873934845,
+      "wer": 0.16119723260202004
+    },
+    {
+      "subset_type": "accent",
+      "subset": "caribe",
+      "cer": 0.04781719959118119,
+      "wer": 0.169803250641574
+    },
+    {
+      "subset_type": "accent",
+      "subset": "surpeninsular",
+      "cer": 0.034524975514201764,
+      "wer": 0.13037037037037036
+    },
+    {
+      "subset_type": "accent",
+      "subset": "mexicano",
+      "cer": 0.042648401826484016,
+      "wer": 0.15081141578063795
+    },
+    {
+      "subset_type": "accent",
+      "subset": "chileno",
+      "cer": 0.03593460826650216,
+      "wer": 0.13357731015553523
+    },
+    {
+      "subset_type": "accent",
+      "subset": "canario",
+      "cer": 0.050411522633744855,
+      "wer": 0.18471337579617833
+    },
+    {
+      "subset_type": "accent",
+      "subset": "americacentral",
+      "cer": 0.04128981517892253,
+      "wer": 0.14341692789968652
+    },
+    {
+      "subset_type": "accent",
+      "subset": "nortepeninsular",
+      "cer": 0.040923616122049276,
+      "wer": 0.14996964177292046
+    },
+    {
+      "subset_type": "accent",
+      "subset": "andino",
+      "cer": 0.0400688162495144,
+      "wer": 0.15101904443701972
+    },
+    {
+      "subset_type": "accent",
+      "subset": "rioplatense",
+      "cer": 0.05707070707070707,
+      "wer": 0.18344519015659955
+    },
+    {
+      "subset_type": "accent",
+      "subset": "centrosurpeninsular",
+      "cer": 0.03241515751027241,
+      "wer": 0.12142857142857143
+    }
+  ]
+}

cv7_transcriptions.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:782a6e3fe000791699552748ae077a977fb468fe6f08d8fcad6fb774994fffae
+size 125505074

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:f2c156d8cb2fc3e74a5577e5690aed4538bb9a77dddfd3e5b8ca7573ea59c611
+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/tmpyau_gx9p", "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}