jonatasgrosman/exp_w2v2r_es_xls-r_age_teens-5_sixties-5_s62

.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-5_sixties-5_s62
+
+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": 0.03871813659326675,
+      "wer": 0.1431957115884815
+    },
+    {
+      "subset_type": "age",
+      "subset": "",
+      "cer": 0.03865285414659033,
+      "wer": 0.1430696393509507
+    },
+    {
+      "subset_type": "age",
+      "subset": "twenties",
+      "cer": 0.040476014669324864,
+      "wer": 0.14783503638903184
+    },
+    {
+      "subset_type": "age",
+      "subset": "fourties",
+      "cer": 0.03244889386726407,
+      "wer": 0.1242392444910808
+    },
+    {
+      "subset_type": "age",
+      "subset": "thirties",
+      "cer": 0.03793034649404775,
+      "wer": 0.14325360116294436
+    },
+    {
+      "subset_type": "age",
+      "subset": "fifties",
+      "cer": 0.036255193673770274,
+      "wer": 0.12921348314606743
+    },
+    {
+      "subset_type": "age",
+      "subset": "sixties",
+      "cer": 0.05062790467714823,
+      "wer": 0.15654205607476634
+    },
+    {
+      "subset_type": "age",
+      "subset": "seventies",
+      "cer": 0.02903225806451613,
+      "wer": 0.10849056603773585
+    },
+    {
+      "subset_type": "age",
+      "subset": "teens",
+      "cer": 0.04094526163491681,
+      "wer": 0.16072455434157562
+    },
+    {
+      "subset_type": "gender",
+      "subset": "",
+      "cer": 0.038605894331938,
+      "wer": 0.1429092420297353
+    },
+    {
+      "subset_type": "gender",
+      "subset": "other",
+      "cer": 0.044384057971014496,
+      "wer": 0.1609403254972875
+    },
+    {
+      "subset_type": "gender",
+      "subset": "male",
+      "cer": 0.039919900977907964,
+      "wer": 0.14553062885040102
+    },
+    {
+      "subset_type": "gender",
+      "subset": "female",
+      "cer": 0.03580431332275231,
+      "wer": 0.138160632542655
+    },
+    {
+      "subset_type": "accent",
+      "subset": "",
+      "cer": 0.03861552232868083,
+      "wer": 0.14319902048085487
+    },
+    {
+      "subset_type": "accent",
+      "subset": "caribe",
+      "cer": 0.03663410301953819,
+      "wer": 0.13564984964918142
+    },
+    {
+      "subset_type": "accent",
+      "subset": "mexicano",
+      "cer": 0.03816053743334026,
+      "wer": 0.1426399447131997
+    },
+    {
+      "subset_type": "accent",
+      "subset": "surpeninsular",
+      "cer": 0.047485544068687574,
+      "wer": 0.15987460815047022
+    },
+    {
+      "subset_type": "accent",
+      "subset": "chileno",
+      "cer": 0.043836251884494956,
+      "wer": 0.16929403701165183
+    },
+    {
+      "subset_type": "accent",
+      "subset": "canario",
+      "cer": 0.03301540719002201,
+      "wer": 0.12844036697247707
+    },
+    {
+      "subset_type": "accent",
+      "subset": "americacentral",
+      "cer": 0.040263333651369144,
+      "wer": 0.1397228637413395
+    },
+    {
+      "subset_type": "accent",
+      "subset": "nortepeninsular",
+      "cer": 0.036136536994660566,
+      "wer": 0.12829131652661063
+    },
+    {
+      "subset_type": "accent",
+      "subset": "andino",
+      "cer": 0.035826140096249334,
+      "wer": 0.13304525813054355
+    },
+    {
+      "subset_type": "accent",
+      "subset": "rioplatense",
+      "cer": 0.03912372358109713,
+      "wer": 0.14935530085959886
+    },
+    {
+      "subset_type": "accent",
+      "subset": "centrosurpeninsular",
+      "cer": 0.04932340459604452,
+      "wer": 0.15979381443298968
+    }
+  ],
+  "test": [
+    {
+      "subset_type": "all",
+      "subset": "all",
+      "cer": 0.044364886499455324,
+      "wer": 0.15592729867401345
+    },
+    {
+      "subset_type": "age",
+      "subset": "",
+      "cer": 0.04451338249917136,
+      "wer": 0.15562538795779018
+    },
+    {
+      "subset_type": "age",
+      "subset": "twenties",
+      "cer": 0.04390097223614313,
+      "wer": 0.16174706023518118
+    },
+    {
+      "subset_type": "age",
+      "subset": "fourties",
+      "cer": 0.039097551595949236,
+      "wer": 0.13976833976833977
+    },
+    {
+      "subset_type": "age",
+      "subset": "thirties",
+      "cer": 0.042350409284808425,
+      "wer": 0.15701149425287356
+    },
+    {
+      "subset_type": "age",
+      "subset": "fifties",
+      "cer": 0.042092357989374746,
+      "wer": 0.14372224465078726
+    },
+    {
+      "subset_type": "age",
+      "subset": "sixties",
+      "cer": 0.047804119704624955,
+      "wer": 0.1707035755478662
+    },
+    {
+      "subset_type": "age",
+      "subset": "seventies",
+      "cer": 0.10395010395010396,
+      "wer": 0.2804878048780488
+    },
+    {
+      "subset_type": "age",
+      "subset": "teens",
+      "cer": 0.04947091730134124,
+      "wer": 0.1794871794871795
+    },
+    {
+      "subset_type": "age",
+      "subset": "eighties",
+      "cer": 0.013422818791946308,
+      "wer": 0.07407407407407407
+    },
+    {
+      "subset_type": "gender",
+      "subset": "",
+      "cer": 0.04457323512039013,
+      "wer": 0.1558189070911274
+    },
+    {
+      "subset_type": "gender",
+      "subset": "other",
+      "cer": 0.07604017216642754,
+      "wer": 0.24782608695652175
+    },
+    {
+      "subset_type": "gender",
+      "subset": "male",
+      "cer": 0.04335977890088227,
+      "wer": 0.1567017462703828
+    },
+    {
+      "subset_type": "gender",
+      "subset": "female",
+      "cer": 0.04079475797928556,
+      "wer": 0.15158562367864695
+    },
+    {
+      "subset_type": "accent",
+      "subset": "",
+      "cer": 0.044790105410270405,
+      "wer": 0.1565873691220541
+    },
+    {
+      "subset_type": "accent",
+      "subset": "caribe",
+      "cer": 0.04548109213023799,
+      "wer": 0.15911035072711718
+    },
+    {
+      "subset_type": "accent",
+      "subset": "surpeninsular",
+      "cer": 0.04040156709108717,
+      "wer": 0.1688888888888889
+    },
+    {
+      "subset_type": "accent",
+      "subset": "mexicano",
+      "cer": 0.03972602739726028,
+      "wer": 0.1502518186905428
+    },
+    {
+      "subset_type": "accent",
+      "subset": "chileno",
+      "cer": 0.03346699568167798,
+      "wer": 0.1262580054894785
+    },
+    {
+      "subset_type": "accent",
+      "subset": "canario",
+      "cer": 0.04526748971193416,
+      "wer": 0.16560509554140126
+    },
+    {
+      "subset_type": "accent",
+      "subset": "americacentral",
+      "cer": 0.04011010617381046,
+      "wer": 0.15752351097178682
+    },
+    {
+      "subset_type": "accent",
+      "subset": "nortepeninsular",
+      "cer": 0.038861972992475,
+      "wer": 0.1408621736490589
+    },
+    {
+      "subset_type": "accent",
+      "subset": "andino",
+      "cer": 0.038681391864143405,
+      "wer": 0.14634146341463414
+    },
+    {
+      "subset_type": "accent",
+      "subset": "rioplatense",
+      "cer": 0.05454545454545454,
+      "wer": 0.16927665920954513
+    },
+    {
+      "subset_type": "accent",
+      "subset": "centrosurpeninsular",
+      "cer": 0.034545731243341954,
+      "wer": 0.1375
+    }
+  ]
+}

cv7_transcriptions.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f52c0eef5e2e79ecaa065a3be72ccb25026a9911692c9a83cf45138e1eed190b
+size 125523753

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 sha256:672b2a36acb0ff81671f1975e10e516442b383055bb80df705474c310675ec74
+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/tmp6a5i5zdy", "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}