jonatasgrosman/exp_w2v2r_es_vp-100k_age_teens-10_sixties-0_s419

.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_vp-100k_age_teens-10_sixties-0_s419
+
+Fine-tuned [facebook/wav2vec2-large-100k-voxpopuli](https://huggingface.co/facebook/wav2vec2-large-100k-voxpopuli) 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 +1,116 @@
+{
+  "_name_or_path": "facebook/wav2vec2-large-100k-voxpopuli",
+  "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_channel_length": 10,
+  "mask_channel_min_space": 1,
+  "mask_channel_other": 0.0,
+  "mask_channel_prob": 0.0,
+  "mask_channel_selection": "static",
+  "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_min_space": 1,
+  "mask_time_other": 0.0,
+  "mask_time_prob": 0.05,
+  "mask_time_selection": "static",
+  "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": 0.052074461345218356,
+      "wer": 0.18396730878411385
+    },
+    {
+      "subset_type": "age",
+      "subset": "",
+      "cer": 0.05183804772972054,
+      "wer": 0.18346097289394725
+    },
+    {
+      "subset_type": "age",
+      "subset": "thirties",
+      "cer": 0.05256157307854308,
+      "wer": 0.1883176952557156
+    },
+    {
+      "subset_type": "age",
+      "subset": "sixties",
+      "cer": 0.07188766933649758,
+      "wer": 0.2003504672897196
+    },
+    {
+      "subset_type": "age",
+      "subset": "seventies",
+      "cer": 0.04516129032258064,
+      "wer": 0.19811320754716982
+    },
+    {
+      "subset_type": "age",
+      "subset": "fifties",
+      "cer": 0.052539874011526606,
+      "wer": 0.18780096308186195
+    },
+    {
+      "subset_type": "age",
+      "subset": "twenties",
+      "cer": 0.0543084988554973,
+      "wer": 0.18981107108726017
+    },
+    {
+      "subset_type": "age",
+      "subset": "teens",
+      "cer": 0.04938509766095973,
+      "wer": 0.18228867165037377
+    },
+    {
+      "subset_type": "age",
+      "subset": "fourties",
+      "cer": 0.04578549425931112,
+      "wer": 0.16558237145855195
+    },
+    {
+      "subset_type": "gender",
+      "subset": "",
+      "cer": 0.051773390655051474,
+      "wer": 0.1832885240062083
+    },
+    {
+      "subset_type": "gender",
+      "subset": "male",
+      "cer": 0.05400203487806933,
+      "wer": 0.18784145065674765
+    },
+    {
+      "subset_type": "gender",
+      "subset": "other",
+      "cer": 0.06370772946859904,
+      "wer": 0.23688969258589512
+    },
+    {
+      "subset_type": "gender",
+      "subset": "female",
+      "cer": 0.04920175520492951,
+      "wer": 0.1771396865029824
+    },
+    {
+      "subset_type": "accent",
+      "subset": "",
+      "cer": 0.05169777542021682,
+      "wer": 0.18348492558198704
+    },
+    {
+      "subset_type": "accent",
+      "subset": "chileno",
+      "cer": 0.0600719007306042,
+      "wer": 0.22001370801919123
+    },
+    {
+      "subset_type": "accent",
+      "subset": "nortepeninsular",
+      "cer": 0.04900839054157132,
+      "wer": 0.17478991596638654
+    },
+    {
+      "subset_type": "accent",
+      "subset": "canario",
+      "cer": 0.0425531914893617,
+      "wer": 0.1834862385321101
+    },
+    {
+      "subset_type": "accent",
+      "subset": "surpeninsular",
+      "cer": 0.06956369370947958,
+      "wer": 0.23406478578892373
+    },
+    {
+      "subset_type": "accent",
+      "subset": "caribe",
+      "cer": 0.0508436944937833,
+      "wer": 0.18242565987303708
+    },
+    {
+      "subset_type": "accent",
+      "subset": "mexicano",
+      "cer": 0.05000346284368724,
+      "wer": 0.17484450587422254
+    },
+    {
+      "subset_type": "accent",
+      "subset": "rioplatense",
+      "cer": 0.05485632866302541,
+      "wer": 0.1977077363896848
+    },
+    {
+      "subset_type": "accent",
+      "subset": "andino",
+      "cer": 0.05228783133450462,
+      "wer": 0.1773936775073914
+    },
+    {
+      "subset_type": "accent",
+      "subset": "centrosurpeninsular",
+      "cer": 0.066938906237489,
+      "wer": 0.20290534208059982
+    },
+    {
+      "subset_type": "accent",
+      "subset": "americacentral",
+      "cer": 0.05533823108482015,
+      "wer": 0.18244803695150116
+    }
+  ],
+  "test": [
+    {
+      "subset_type": "all",
+      "subset": "all",
+      "cer": 0.05947477055116235,
+      "wer": 0.19990882603611282
+    },
+    {
+      "subset_type": "age",
+      "subset": "",
+      "cer": 0.059977833941000994,
+      "wer": 0.20096213531967722
+    },
+    {
+      "subset_type": "age",
+      "subset": "thirties",
+      "cer": 0.055608931247838285,
+      "wer": 0.19402298850574712
+    },
+    {
+      "subset_type": "age",
+      "subset": "sixties",
+      "cer": 0.06257287213369607,
+      "wer": 0.19261822376009227
+    },
+    {
+      "subset_type": "age",
+      "subset": "seventies",
+      "cer": 0.12474012474012475,
+      "wer": 0.3048780487804878
+    },
+    {
+      "subset_type": "age",
+      "subset": "fifties",
+      "cer": 0.05224083912273532,
+      "wer": 0.17440452159870812
+    },
+    {
+      "subset_type": "age",
+      "subset": "twenties",
+      "cer": 0.05669038789215195,
+      "wer": 0.19330453563714903
+    },
+    {
+      "subset_type": "age",
+      "subset": "eighties",
+      "cer": 0.053691275167785234,
+      "wer": 0.14814814814814814
+    },
+    {
+      "subset_type": "age",
+      "subset": "teens",
+      "cer": 0.06632068477455011,
+      "wer": 0.23158323158323157
+    },
+    {
+      "subset_type": "age",
+      "subset": "fourties",
+      "cer": 0.051318207067469986,
+      "wer": 0.18172458172458172
+    },
+    {
+      "subset_type": "gender",
+      "subset": "",
+      "cer": 0.060027685399764504,
+      "wer": 0.20113860810967818
+    },
+    {
+      "subset_type": "gender",
+      "subset": "male",
+      "cer": 0.05563232608254496,
+      "wer": 0.1913380363131722
+    },
+    {
+      "subset_type": "gender",
+      "subset": "other",
+      "cer": 0.09899569583931134,
+      "wer": 0.28695652173913044
+    },
+    {
+      "subset_type": "gender",
+      "subset": "female",
+      "cer": 0.05647150003522863,
+      "wer": 0.193446088794926
+    },
+    {
+      "subset_type": "accent",
+      "subset": "",
+      "cer": 0.06013496662872325,
+      "wer": 0.20156868548595952
+    },
+    {
+      "subset_type": "accent",
+      "subset": "chileno",
+      "cer": 0.04441702652683529,
+      "wer": 0.16285452881976212
+    },
+    {
+      "subset_type": "accent",
+      "subset": "nortepeninsular",
+      "cer": 0.054015049994845896,
+      "wer": 0.17729204614450517
+    },
+    {
+      "subset_type": "accent",
+      "subset": "canario",
+      "cer": 0.05864197530864197,
+      "wer": 0.20382165605095542
+    },
+    {
+      "subset_type": "accent",
+      "subset": "surpeninsular",
+      "cer": 0.05190989226248776,
+      "wer": 0.20592592592592593
+    },
+    {
+      "subset_type": "accent",
+      "subset": "caribe",
+      "cer": 0.06431595853409257,
+      "wer": 0.21129170230966637
+    },
+    {
+      "subset_type": "accent",
+      "subset": "mexicano",
+      "cer": 0.0547945205479452,
+      "wer": 0.18690542809177393
+    },
+    {
+      "subset_type": "accent",
+      "subset": "rioplatense",
+      "cer": 0.06742424242424243,
+      "wer": 0.2110365398956003
+    },
+    {
+      "subset_type": "accent",
+      "subset": "andino",
+      "cer": 0.04789388978300683,
+      "wer": 0.17340461075843636
+    },
+    {
+      "subset_type": "accent",
+      "subset": "centrosurpeninsular",
+      "cer": 0.043981129204078526,
+      "wer": 0.16964285714285715
+    },
+    {
+      "subset_type": "accent",
+      "subset": "americacentral",
+      "cer": 0.05046532966312754,
+      "wer": 0.17398119122257052
+    }
+  ]
+}

cv7_transcriptions.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cf8a6d64c9555a2474bd61c7914fe4ca84f9218c5cae0440300acacfad4f667a
+size 125327122

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:b581423670201e1530cb01a2c1b69e23d779cb6b83a247e8a18ae6c0c9795eff
+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/tmpxr_3g1od", "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}