jonatasgrosman/exp_w2v2r_es_xls-r_accent_surpeninsular-5_nortepeninsular-5_s463

.gitattributes

@@ -29,3 +29,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zstandard 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_accent_surpeninsular-5_nortepeninsular-5_s463
+
+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": 0.04652456794680388,
+      "wer": 0.16650602727788855
+    },
+    {
+      "subset_type": "age",
+      "subset": "",
+      "cer": 0.046177937758630466,
+      "wer": 0.16482078045712245
+    },
+    {
+      "subset_type": "age",
+      "subset": "sixties",
+      "cer": 0.051023435182438445,
+      "wer": 0.1647196261682243
+    },
+    {
+      "subset_type": "age",
+      "subset": "seventies",
+      "cer": 0.03951612903225806,
+      "wer": 0.17452830188679244
+    },
+    {
+      "subset_type": "age",
+      "subset": "fourties",
+      "cer": 0.04260011201344161,
+      "wer": 0.15068205666316895
+    },
+    {
+      "subset_type": "age",
+      "subset": "thirties",
+      "cer": 0.046243543417056464,
+      "wer": 0.1762917933130699
+    },
+    {
+      "subset_type": "age",
+      "subset": "teens",
+      "cer": 0.05063901615625754,
+      "wer": 0.1863139735480161
+    },
+    {
+      "subset_type": "age",
+      "subset": "fifties",
+      "cer": 0.04677657150516017,
+      "wer": 0.16412520064205458
+    },
+    {
+      "subset_type": "age",
+      "subset": "twenties",
+      "cer": 0.04952127790494475,
+      "wer": 0.17665220907152834
+    },
+    {
+      "subset_type": "gender",
+      "subset": "",
+      "cer": 0.046095429347211225,
+      "wer": 0.1646043067841604
+    },
+    {
+      "subset_type": "gender",
+      "subset": "female",
+      "cer": 0.04367005881803753,
+      "wer": 0.1632681370509086
+    },
+    {
+      "subset_type": "gender",
+      "subset": "male",
+      "cer": 0.048875956684315235,
+      "wer": 0.17474524390716417
+    },
+    {
+      "subset_type": "gender",
+      "subset": "other",
+      "cer": 0.06551932367149758,
+      "wer": 0.22965641952983726
+    },
+    {
+      "subset_type": "accent",
+      "subset": "",
+      "cer": 0.046145277326525874,
+      "wer": 0.16515074418012976
+    },
+    {
+      "subset_type": "accent",
+      "subset": "rioplatense",
+      "cer": 0.04986938969365946,
+      "wer": 0.18875358166189113
+    },
+    {
+      "subset_type": "accent",
+      "subset": "mexicano",
+      "cer": 0.04711775977099061,
+      "wer": 0.17498272287491362
+    },
+    {
+      "subset_type": "accent",
+      "subset": "americacentral",
+      "cer": 0.05467035588207232,
+      "wer": 0.18302540415704388
+    },
+    {
+      "subset_type": "accent",
+      "subset": "canario",
+      "cer": 0.03888481291269259,
+      "wer": 0.1651376146788991
+    },
+    {
+      "subset_type": "accent",
+      "subset": "andino",
+      "cer": 0.04907952028110916,
+      "wer": 0.17125312713213556
+    },
+    {
+      "subset_type": "accent",
+      "subset": "nortepeninsular",
+      "cer": 0.03537376048817696,
+      "wer": 0.13333333333333333
+    },
+    {
+      "subset_type": "accent",
+      "subset": "surpeninsular",
+      "cer": 0.04257928859295602,
+      "wer": 0.1452455590386625
+    },
+    {
+      "subset_type": "accent",
+      "subset": "centrosurpeninsular",
+      "cer": 0.048042277203939464,
+      "wer": 0.16213683223992503
+    },
+    {
+      "subset_type": "accent",
+      "subset": "chileno",
+      "cer": 0.06320306157949669,
+      "wer": 0.21727210418094586
+    },
+    {
+      "subset_type": "accent",
+      "subset": "caribe",
+      "cer": 0.04867895204262877,
+      "wer": 0.17073170731707318
+    }
+  ],
+  "test": [
+    {
+      "subset_type": "all",
+      "subset": "all",
+      "cer": 0.055287136310912546,
+      "wer": 0.18555883693734762
+    },
+    {
+      "subset_type": "age",
+      "subset": "",
+      "cer": 0.055746602585349685,
+      "wer": 0.1863516449410304
+    },
+    {
+      "subset_type": "age",
+      "subset": "sixties",
+      "cer": 0.05188495919160513,
+      "wer": 0.17185697808535177
+    },
+    {
+      "subset_type": "age",
+      "subset": "seventies",
+      "cer": 0.11642411642411643,
+      "wer": 0.3170731707317073
+    },
+    {
+      "subset_type": "age",
+      "subset": "fourties",
+      "cer": 0.05119001837371277,
+      "wer": 0.16344916344916344
+    },
+    {
+      "subset_type": "age",
+      "subset": "thirties",
+      "cer": 0.048422428038891664,
+      "wer": 0.18045977011494252
+    },
+    {
+      "subset_type": "age",
+      "subset": "fifties",
+      "cer": 0.04624710529900559,
+      "wer": 0.14937424303593055
+    },
+    {
+      "subset_type": "age",
+      "subset": "teens",
+      "cer": 0.0638269191885152,
+      "wer": 0.22873422873422875
+    },
+    {
+      "subset_type": "age",
+      "subset": "twenties",
+      "cer": 0.053683471985566805,
+      "wer": 0.1847852171826254
+    },
+    {
+      "subset_type": "age",
+      "subset": "eighties",
+      "cer": 0.013422818791946308,
+      "wer": 0.07407407407407407
+    },
+    {
+      "subset_type": "gender",
+      "subset": "",
+      "cer": 0.05580378780730858,
+      "wer": 0.1865458347856396
+    },
+    {
+      "subset_type": "gender",
+      "subset": "female",
+      "cer": 0.04812231381667019,
+      "wer": 0.1748414376321353
+    },
+    {
+      "subset_type": "gender",
+      "subset": "male",
+      "cer": 0.052926567648792555,
+      "wer": 0.17988897883659072
+    },
+    {
+      "subset_type": "gender",
+      "subset": "other",
+      "cer": 0.0896700143472023,
+      "wer": 0.2782608695652174
+    },
+    {
+      "subset_type": "accent",
+      "subset": "",
+      "cer": 0.055964482985957645,
+      "wer": 0.18711014095970993
+    },
+    {
+      "subset_type": "accent",
+      "subset": "rioplatense",
+      "cer": 0.06616161616161616,
+      "wer": 0.20208799403430275
+    },
+    {
+      "subset_type": "accent",
+      "subset": "mexicano",
+      "cer": 0.05214611872146119,
+      "wer": 0.1782316731952994
+    },
+    {
+      "subset_type": "accent",
+      "subset": "americacentral",
+      "cer": 0.0508585659981649,
+      "wer": 0.1724137931034483
+    },
+    {
+      "subset_type": "accent",
+      "subset": "canario",
+      "cer": 0.04218106995884774,
+      "wer": 0.1592356687898089
+    },
+    {
+      "subset_type": "accent",
+      "subset": "andino",
+      "cer": 0.04506354403684999,
+      "wer": 0.17139993317741398
+    },
+    {
+      "subset_type": "accent",
+      "subset": "nortepeninsular",
+      "cer": 0.03710957633233687,
+      "wer": 0.13721918639951428
+    },
+    {
+      "subset_type": "accent",
+      "subset": "surpeninsular",
+      "cer": 0.04382957884427032,
+      "wer": 0.15555555555555556
+    },
+    {
+      "subset_type": "accent",
+      "subset": "centrosurpeninsular",
+      "cer": 0.03180642215796682,
+      "wer": 0.12410714285714286
+    },
+    {
+      "subset_type": "accent",
+      "subset": "chileno",
+      "cer": 0.050431832202344234,
+      "wer": 0.1655992680695334
+    },
+    {
+      "subset_type": "accent",
+      "subset": "caribe",
+      "cer": 0.060665790626368815,
+      "wer": 0.20615911035072712
+    }
+  ]
+}

cv7_transcriptions.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:54d3da1889b25573753ca2fed11b1963d89495ac212bafdf715d9567f01b21a0
+size 125741395

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 sha256:53fbac00618f8f202c904d08a5ee974a02d23067bc519ff102e1176d6a5f4376
+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/tmp_b3laqjm", "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}