jonatasgrosman/exp_w2v2r_es_vp-100k_age_teens-2_sixties-8_s468

.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-2_sixties-8_s468
+
+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": 0.05810915781992871,
+      "wer": 0.20238582500429106
+    },
+    {
+      "subset_type": "age",
+      "subset": "",
+      "cer": 0.057825520095232695,
+      "wer": 0.2017101717821656
+    },
+    {
+      "subset_type": "age",
+      "subset": "thirties",
+      "cer": 0.05834755813695715,
+      "wer": 0.2066869300911854
+    },
+    {
+      "subset_type": "age",
+      "subset": "sixties",
+      "cer": 0.0801938099475922,
+      "wer": 0.2266355140186916
+    },
+    {
+      "subset_type": "age",
+      "subset": "seventies",
+      "cer": 0.050806451612903224,
+      "wer": 0.22169811320754718
+    },
+    {
+      "subset_type": "age",
+      "subset": "fifties",
+      "cer": 0.05528749497386409,
+      "wer": 0.19903691813804172
+    },
+    {
+      "subset_type": "age",
+      "subset": "twenties",
+      "cer": 0.06091707893376652,
+      "wer": 0.2090715283393369
+    },
+    {
+      "subset_type": "age",
+      "subset": "teens",
+      "cer": 0.056136966481794066,
+      "wer": 0.20327774583093733
+    },
+    {
+      "subset_type": "age",
+      "subset": "fourties",
+      "cer": 0.05215625875105013,
+      "wer": 0.18467995802728226
+    },
+    {
+      "subset_type": "gender",
+      "subset": "",
+      "cer": 0.05774691707206698,
+      "wer": 0.20153171567421782
+    },
+    {
+      "subset_type": "gender",
+      "subset": "male",
+      "cer": 0.06021683764394818,
+      "wer": 0.20674958347863148
+    },
+    {
+      "subset_type": "gender",
+      "subset": "other",
+      "cer": 0.07940821256038648,
+      "wer": 0.2820976491862568
+    },
+    {
+      "subset_type": "gender",
+      "subset": "female",
+      "cer": 0.054850154047241154,
+      "wer": 0.19461783881259537
+    },
+    {
+      "subset_type": "accent",
+      "subset": "",
+      "cer": 0.05766004707754534,
+      "wer": 0.20162829156595852
+    },
+    {
+      "subset_type": "accent",
+      "subset": "chileno",
+      "cer": 0.07143685492288067,
+      "wer": 0.23920493488690883
+    },
+    {
+      "subset_type": "accent",
+      "subset": "nortepeninsular",
+      "cer": 0.055491990846681924,
+      "wer": 0.1899159663865546
+    },
+    {
+      "subset_type": "accent",
+      "subset": "canario",
+      "cer": 0.05796038151137197,
+      "wer": 0.20642201834862386
+    },
+    {
+      "subset_type": "accent",
+      "subset": "surpeninsular",
+      "cer": 0.06570877869283337,
+      "wer": 0.21630094043887146
+    },
+    {
+      "subset_type": "accent",
+      "subset": "caribe",
+      "cer": 0.05883658969804618,
+      "wer": 0.20247243568326095
+    },
+    {
+      "subset_type": "accent",
+      "subset": "mexicano",
+      "cer": 0.05674446522150657,
+      "wer": 0.20041465100207326
+    },
+    {
+      "subset_type": "accent",
+      "subset": "rioplatense",
+      "cer": 0.057171693184516745,
+      "wer": 0.20379656160458454
+    },
+    {
+      "subset_type": "accent",
+      "subset": "andino",
+      "cer": 0.05965930792147277,
+      "wer": 0.20513986809188084
+    },
+    {
+      "subset_type": "accent",
+      "subset": "centrosurpeninsular",
+      "cer": 0.07390503643206021,
+      "wer": 0.2211808809746954
+    },
+    {
+      "subset_type": "accent",
+      "subset": "americacentral",
+      "cer": 0.06115828642305123,
+      "wer": 0.2066974595842956
+    }
+  ],
+  "test": [
+    {
+      "subset_type": "all",
+      "subset": "all",
+      "cer": 0.06687645266143011,
+      "wer": 0.22004637979902086
+    },
+    {
+      "subset_type": "age",
+      "subset": "",
+      "cer": 0.06735270964534305,
+      "wer": 0.22099627560521415
+    },
+    {
+      "subset_type": "age",
+      "subset": "thirties",
+      "cer": 0.0642173628991968,
+      "wer": 0.22
+    },
+    {
+      "subset_type": "age",
+      "subset": "sixties",
+      "cer": 0.060046638165565484,
+      "wer": 0.19953863898500576
+    },
+    {
+      "subset_type": "age",
+      "subset": "seventies",
+      "cer": 0.14137214137214138,
+      "wer": 0.34146341463414637
+    },
+    {
+      "subset_type": "age",
+      "subset": "fifties",
+      "cer": 0.057825909276665304,
+      "wer": 0.18732337505046426
+    },
+    {
+      "subset_type": "age",
+      "subset": "twenties",
+      "cer": 0.06551067455146838,
+      "wer": 0.21826253899688025
+    },
+    {
+      "subset_type": "age",
+      "subset": "eighties",
+      "cer": 0.0738255033557047,
+      "wer": 0.25925925925925924
+    },
+    {
+      "subset_type": "age",
+      "subset": "teens",
+      "cer": 0.07137561501651277,
+      "wer": 0.2442002442002442
+    },
+    {
+      "subset_type": "age",
+      "subset": "fourties",
+      "cer": 0.059778660855445885,
+      "wer": 0.19974259974259975
+    },
+    {
+      "subset_type": "gender",
+      "subset": "",
+      "cer": 0.06742984620514855,
+      "wer": 0.2212307811471283
+    },
+    {
+      "subset_type": "gender",
+      "subset": "male",
+      "cer": 0.06398152353526666,
+      "wer": 0.21383138660807216
+    },
+    {
+      "subset_type": "gender",
+      "subset": "other",
+      "cer": 0.10473457675753228,
+      "wer": 0.2956521739130435
+    },
+    {
+      "subset_type": "gender",
+      "subset": "female",
+      "cer": 0.06048756429225675,
+      "wer": 0.20676532769556025
+    },
+    {
+      "subset_type": "accent",
+      "subset": "",
+      "cer": 0.06763146653365894,
+      "wer": 0.2217174146287321
+    },
+    {
+      "subset_type": "accent",
+      "subset": "chileno",
+      "cer": 0.052745219000616905,
+      "wer": 0.1838975297346752
+    },
+    {
+      "subset_type": "accent",
+      "subset": "nortepeninsular",
+      "cer": 0.05587052881146273,
+      "wer": 0.19064966605950212
+    },
+    {
+      "subset_type": "accent",
+      "subset": "canario",
+      "cer": 0.07407407407407407,
+      "wer": 0.25477707006369427
+    },
+    {
+      "subset_type": "accent",
+      "subset": "surpeninsular",
+      "cer": 0.05778648383937316,
+      "wer": 0.21925925925925926
+    },
+    {
+      "subset_type": "accent",
+      "subset": "caribe",
+      "cer": 0.07599649583880859,
+      "wer": 0.2369546621043627
+    },
+    {
+      "subset_type": "accent",
+      "subset": "mexicano",
+      "cer": 0.05940639269406393,
+      "wer": 0.20229434806939003
+    },
+    {
+      "subset_type": "accent",
+      "subset": "rioplatense",
+      "cer": 0.07777777777777778,
+      "wer": 0.2431021625652498
+    },
+    {
+      "subset_type": "accent",
+      "subset": "andino",
+      "cer": 0.05410955102946889,
+      "wer": 0.194119612429001
+    },
+    {
+      "subset_type": "accent",
+      "subset": "centrosurpeninsular",
+      "cer": 0.04717698980368285,
+      "wer": 0.17857142857142858
+    },
+    {
+      "subset_type": "accent",
+      "subset": "americacentral",
+      "cer": 0.05754358369379997,
+      "wer": 0.19984326018808776
+    }
+  ]
+}

cv7_transcriptions.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:69212fe7489b882d9d80ebabf4d376ab896277b024c6c40757e46eba49e4f473
+size 125365375

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 sha256:20d4f836f03e7e64900f74ea00064cbea8e41cf5223947c680f5020e557391ae
+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/tmp09qeo904", "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}