jonatasgrosman/exp_w2v2r_es_xls-r_age_teens-2_sixties-8_s772

.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_s772
+
+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": 0.039691048221378714,
+      "wer": 0.14549307490196597
+    },
+    {
+      "subset_type": "age",
+      "subset": "",
+      "cer": 0.039599512499291424,
+      "wer": 0.14495660821596443
+    },
+    {
+      "subset_type": "age",
+      "subset": "twenties",
+      "cer": 0.04166974328681484,
+      "wer": 0.15342203925604647
+    },
+    {
+      "subset_type": "age",
+      "subset": "fourties",
+      "cer": 0.033323998879865584,
+      "wer": 0.12843651626442812
+    },
+    {
+      "subset_type": "age",
+      "subset": "thirties",
+      "cer": 0.03983683965505775,
+      "wer": 0.14880401744416547
+    },
+    {
+      "subset_type": "age",
+      "subset": "fifties",
+      "cer": 0.03665728454630746,
+      "wer": 0.1312199036918138
+    },
+    {
+      "subset_type": "age",
+      "subset": "sixties",
+      "cer": 0.0482547216454069,
+      "wer": 0.15186915887850466
+    },
+    {
+      "subset_type": "age",
+      "subset": "seventies",
+      "cer": 0.03387096774193549,
+      "wer": 0.12264150943396226
+    },
+    {
+      "subset_type": "age",
+      "subset": "teens",
+      "cer": 0.0418615866891729,
+      "wer": 0.1572742955721679
+    },
+    {
+      "subset_type": "gender",
+      "subset": "",
+      "cer": 0.03953642946034619,
+      "wer": 0.14475815685243454
+    },
+    {
+      "subset_type": "gender",
+      "subset": "other",
+      "cer": 0.04891304347826087,
+      "wer": 0.17540687160940324
+    },
+    {
+      "subset_type": "gender",
+      "subset": "male",
+      "cer": 0.04123544316922539,
+      "wer": 0.15041264675113333
+    },
+    {
+      "subset_type": "gender",
+      "subset": "female",
+      "cer": 0.03596769676033984,
+      "wer": 0.13760577056457207
+    },
+    {
+      "subset_type": "accent",
+      "subset": "",
+      "cer": 0.03954381195504426,
+      "wer": 0.14490840859941484
+    },
+    {
+      "subset_type": "accent",
+      "subset": "caribe",
+      "cer": 0.03907637655417407,
+      "wer": 0.14233210825258938
+    },
+    {
+      "subset_type": "accent",
+      "subset": "mexicano",
+      "cer": 0.03926864741325577,
+      "wer": 0.14761575673807878
+    },
+    {
+      "subset_type": "accent",
+      "subset": "surpeninsular",
+      "cer": 0.04100227790432802,
+      "wer": 0.14733542319749215
+    },
+    {
+      "subset_type": "accent",
+      "subset": "chileno",
+      "cer": 0.05033051142293865,
+      "wer": 0.1782042494859493
+    },
+    {
+      "subset_type": "accent",
+      "subset": "canario",
+      "cer": 0.03961848862802641,
+      "wer": 0.1651376146788991
+    },
+    {
+      "subset_type": "accent",
+      "subset": "americacentral",
+      "cer": 0.0435072989218586,
+      "wer": 0.15011547344110854
+    },
+    {
+      "subset_type": "accent",
+      "subset": "nortepeninsular",
+      "cer": 0.03556445461479787,
+      "wer": 0.13333333333333333
+    },
+    {
+      "subset_type": "accent",
+      "subset": "andino",
+      "cer": 0.037315713085325795,
+      "wer": 0.14214236979758926
+    },
+    {
+      "subset_type": "accent",
+      "subset": "rioplatense",
+      "cer": 0.03977677511279981,
+      "wer": 0.14577363896848136
+    },
+    {
+      "subset_type": "accent",
+      "subset": "centrosurpeninsular",
+      "cer": 0.04812234766594603,
+      "wer": 0.16494845360824742
+    }
+  ],
+  "test": [
+    {
+      "subset_type": "all",
+      "subset": "all",
+      "cer": 0.045896383065992176,
+      "wer": 0.16060491942996452
+    },
+    {
+      "subset_type": "age",
+      "subset": "",
+      "cer": 0.046012698873052704,
+      "wer": 0.16073091247672253
+    },
+    {
+      "subset_type": "age",
+      "subset": "twenties",
+      "cer": 0.04660719655206976,
+      "wer": 0.16798656107511398
+    },
+    {
+      "subset_type": "age",
+      "subset": "fourties",
+      "cer": 0.037473828141691236,
+      "wer": 0.13667953667953667
+    },
+    {
+      "subset_type": "age",
+      "subset": "thirties",
+      "cer": 0.04242727028169555,
+      "wer": 0.1560919540229885
+    },
+    {
+      "subset_type": "age",
+      "subset": "fifties",
+      "cer": 0.04502111428960632,
+      "wer": 0.14129995962858297
+    },
+    {
+      "subset_type": "age",
+      "subset": "sixties",
+      "cer": 0.04586086280606296,
+      "wer": 0.16032295271049596
+    },
+    {
+      "subset_type": "age",
+      "subset": "seventies",
+      "cer": 0.1101871101871102,
+      "wer": 0.2926829268292683
+    },
+    {
+      "subset_type": "age",
+      "subset": "teens",
+      "cer": 0.055941228011053445,
+      "wer": 0.1908831908831909
+    },
+    {
+      "subset_type": "age",
+      "subset": "eighties",
+      "cer": 0.013422818791946308,
+      "wer": 0.07407407407407407
+    },
+    {
+      "subset_type": "gender",
+      "subset": "",
+      "cer": 0.046063490299123815,
+      "wer": 0.16090778823438287
+    },
+    {
+      "subset_type": "gender",
+      "subset": "other",
+      "cer": 0.07388809182209469,
+      "wer": 0.22608695652173913
+    },
+    {
+      "subset_type": "gender",
+      "subset": "male",
+      "cer": 0.04459669704202623,
+      "wer": 0.15797386376778072
+    },
+    {
+      "subset_type": "gender",
+      "subset": "female",
+      "cer": 0.044705136334812934,
+      "wer": 0.1587737843551797
+    },
+    {
+      "subset_type": "accent",
+      "subset": "",
+      "cer": 0.04629755376071943,
+      "wer": 0.16184838506789004
+    },
+    {
+      "subset_type": "accent",
+      "subset": "caribe",
+      "cer": 0.04942327347057965,
+      "wer": 0.16680923866552608
+    },
+    {
+      "subset_type": "accent",
+      "subset": "surpeninsular",
+      "cer": 0.03672869735553379,
+      "wer": 0.13777777777777778
+    },
+    {
+      "subset_type": "accent",
+      "subset": "mexicano",
+      "cer": 0.04210045662100457,
+      "wer": 0.15137101287073307
+    },
+    {
+      "subset_type": "accent",
+      "subset": "chileno",
+      "cer": 0.038402220851326344,
+      "wer": 0.13815187557182068
+    },
+    {
+      "subset_type": "accent",
+      "subset": "canario",
+      "cer": 0.05555555555555555,
+      "wer": 0.21019108280254778
+    },
+    {
+      "subset_type": "accent",
+      "subset": "americacentral",
+      "cer": 0.041551972735614104,
+      "wer": 0.14576802507836992
+    },
+    {
+      "subset_type": "accent",
+      "subset": "nortepeninsular",
+      "cer": 0.040202041026698276,
+      "wer": 0.14146933819064966
+    },
+    {
+      "subset_type": "accent",
+      "subset": "andino",
+      "cer": 0.038625894888728564,
+      "wer": 0.14266622118275976
+    },
+    {
+      "subset_type": "accent",
+      "subset": "rioplatense",
+      "cer": 0.05820707070707071,
+      "wer": 0.1871737509321402
+    },
+    {
+      "subset_type": "accent",
+      "subset": "centrosurpeninsular",
+      "cer": 0.03195860599604322,
+      "wer": 0.11696428571428572
+    }
+  ]
+}

cv7_transcriptions.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2f3bfde59903e571da37822885d866c95b058e2ca54a777d5ecc775f50b39c7e
+size 125421244

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 sha256:41e68a21ced703adfad8834d51b3827cba3cbb2864040c2e8a9175e848ad6bcc
+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/tmpj8d9jtml", "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}