jonatasgrosman/exp_w2v2r_es_xls-r_age_teens-0_sixties-10_s951

.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-0_sixties-10_s951
+
+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": 0.04878110590192654,
+      "wer": 0.17749112082282575
+    },
+    {
+      "subset_type": "age",
+      "subset": "",
+      "cer": 0.04855166940649623,
+      "wer": 0.17634659028126035
+    },
+    {
+      "subset_type": "age",
+      "subset": "twenties",
+      "cer": 0.05220409067414901,
+      "wer": 0.18760567521870175
+    },
+    {
+      "subset_type": "age",
+      "subset": "fourties",
+      "cer": 0.041339960795295436,
+      "wer": 0.15487932843651628
+    },
+    {
+      "subset_type": "age",
+      "subset": "thirties",
+      "cer": 0.04883726085703518,
+      "wer": 0.18567463988370556
+    },
+    {
+      "subset_type": "age",
+      "subset": "fifties",
+      "cer": 0.044833132287897065,
+      "wer": 0.16653290529695025
+    },
+    {
+      "subset_type": "age",
+      "subset": "sixties",
+      "cer": 0.05656086225650153,
+      "wer": 0.18282710280373832
+    },
+    {
+      "subset_type": "age",
+      "subset": "seventies",
+      "cer": 0.043548387096774194,
+      "wer": 0.1650943396226415
+    },
+    {
+      "subset_type": "age",
+      "subset": "teens",
+      "cer": 0.052664576802507836,
+      "wer": 0.19580218516388728
+    },
+    {
+      "subset_type": "gender",
+      "subset": "",
+      "cer": 0.04844863672361127,
+      "wer": 0.17608793371046672
+    },
+    {
+      "subset_type": "gender",
+      "subset": "other",
+      "cer": 0.059178743961352656,
+      "wer": 0.23688969258589512
+    },
+    {
+      "subset_type": "gender",
+      "subset": "male",
+      "cer": 0.05087195173321064,
+      "wer": 0.18392808710139874
+    },
+    {
+      "subset_type": "gender",
+      "subset": "female",
+      "cer": 0.04593408645317897,
+      "wer": 0.1728395061728395
+    },
+    {
+      "subset_type": "accent",
+      "subset": "",
+      "cer": 0.048548221330769484,
+      "wer": 0.17665532375015902
+    },
+    {
+      "subset_type": "accent",
+      "subset": "caribe",
+      "cer": 0.05012211367673179,
+      "wer": 0.17273638489809556
+    },
+    {
+      "subset_type": "accent",
+      "subset": "mexicano",
+      "cer": 0.048387469122977124,
+      "wer": 0.1806496199032481
+    },
+    {
+      "subset_type": "accent",
+      "subset": "surpeninsular",
+      "cer": 0.050289118626248464,
+      "wer": 0.1755485893416928
+    },
+    {
+      "subset_type": "accent",
+      "subset": "chileno",
+      "cer": 0.059028180447640036,
+      "wer": 0.20767649074708705
+    },
+    {
+      "subset_type": "accent",
+      "subset": "canario",
+      "cer": 0.0425531914893617,
+      "wer": 0.17889908256880735
+    },
+    {
+      "subset_type": "accent",
+      "subset": "americacentral",
+      "cer": 0.05734185669306364,
+      "wer": 0.2066974595842956
+    },
+    {
+      "subset_type": "accent",
+      "subset": "nortepeninsular",
+      "cer": 0.04223874904652937,
+      "wer": 0.16022408963585436
+    },
+    {
+      "subset_type": "accent",
+      "subset": "andino",
+      "cer": 0.04820105415934611,
+      "wer": 0.1773936775073914
+    },
+    {
+      "subset_type": "accent",
+      "subset": "rioplatense",
+      "cer": 0.048147708382806936,
+      "wer": 0.1758595988538682
+    },
+    {
+      "subset_type": "accent",
+      "subset": "centrosurpeninsular",
+      "cer": 0.05556890063255665,
+      "wer": 0.1958762886597938
+    }
+  ],
+  "test": [
+    {
+      "subset_type": "all",
+      "subset": "all",
+      "cer": 0.0562100035723894,
+      "wer": 0.19220528676854365
+    },
+    {
+      "subset_type": "age",
+      "subset": "",
+      "cer": 0.05640951276102088,
+      "wer": 0.19272967101179392
+    },
+    {
+      "subset_type": "age",
+      "subset": "twenties",
+      "cer": 0.0577528315124787,
+      "wer": 0.19858411327093833
+    },
+    {
+      "subset_type": "age",
+      "subset": "fourties",
+      "cer": 0.04909626970901167,
+      "wer": 0.16653796653796654
+    },
+    {
+      "subset_type": "age",
+      "subset": "thirties",
+      "cer": 0.05003650897352139,
+      "wer": 0.17977011494252873
+    },
+    {
+      "subset_type": "age",
+      "subset": "fifties",
+      "cer": 0.050129410162103254,
+      "wer": 0.1667339523617279
+    },
+    {
+      "subset_type": "age",
+      "subset": "sixties",
+      "cer": 0.050524679362611735,
+      "wer": 0.1707035755478662
+    },
+    {
+      "subset_type": "age",
+      "subset": "seventies",
+      "cer": 0.1101871101871102,
+      "wer": 0.2926829268292683
+    },
+    {
+      "subset_type": "age",
+      "subset": "teens",
+      "cer": 0.0689492485003707,
+      "wer": 0.23565323565323565
+    },
+    {
+      "subset_type": "age",
+      "subset": "eighties",
+      "cer": 0.053691275167785234,
+      "wer": 0.18518518518518517
+    },
+    {
+      "subset_type": "gender",
+      "subset": "",
+      "cer": 0.05646942650379414,
+      "wer": 0.19292823670655668
+    },
+    {
+      "subset_type": "gender",
+      "subset": "other",
+      "cer": 0.10114777618364419,
+      "wer": 0.3130434782608696
+    },
+    {
+      "subset_type": "gender",
+      "subset": "male",
+      "cer": 0.055071847549839105,
+      "wer": 0.18775297791141438
+    },
+    {
+      "subset_type": "gender",
+      "subset": "female",
+      "cer": 0.051081519058690905,
+      "wer": 0.1828752642706131
+    },
+    {
+      "subset_type": "accent",
+      "subset": "",
+      "cer": 0.05669783623752744,
+      "wer": 0.19342927966258464
+    },
+    {
+      "subset_type": "accent",
+      "subset": "caribe",
+      "cer": 0.06424295517593809,
+      "wer": 0.20915312232677502
+    },
+    {
+      "subset_type": "accent",
+      "subset": "surpeninsular",
+      "cer": 0.0416258570029383,
+      "wer": 0.17037037037037037
+    },
+    {
+      "subset_type": "accent",
+      "subset": "mexicano",
+      "cer": 0.05136986301369863,
+      "wer": 0.18354784555120313
+    },
+    {
+      "subset_type": "accent",
+      "subset": "chileno",
+      "cer": 0.0445712523133868,
+      "wer": 0.1637694419030192
+    },
+    {
+      "subset_type": "accent",
+      "subset": "canario",
+      "cer": 0.05555555555555555,
+      "wer": 0.21656050955414013
+    },
+    {
+      "subset_type": "accent",
+      "subset": "americacentral",
+      "cer": 0.052824747673351685,
+      "wer": 0.1778996865203762
+    },
+    {
+      "subset_type": "accent",
+      "subset": "nortepeninsular",
+      "cer": 0.04731470982372951,
+      "wer": 0.16393442622950818
+    },
+    {
+      "subset_type": "accent",
+      "subset": "andino",
+      "cer": 0.048115877684666186,
+      "wer": 0.1780821917808219
+    },
+    {
+      "subset_type": "accent",
+      "subset": "rioplatense",
+      "cer": 0.06792929292929292,
+      "wer": 0.20656226696495153
+    },
+    {
+      "subset_type": "accent",
+      "subset": "centrosurpeninsular",
+      "cer": 0.04124182011870339,
+      "wer": 0.15267857142857144
+    }
+  ]
+}

cv7_transcriptions.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0eb6b05d423ee2434463de44449f2504026001139cd179c296608b73256fe04c
+size 125239298

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:b9fe0073e38dfa1f7e6888a8d195512a0c83b64ab5415b1c93b26ed5ae2ebbf2
+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/tmp0yep3g8i", "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}