jonatasgrosman/exp_w2v2r_es_xls-r_age_teens-0_sixties-10_s961

.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-0_sixties-10_s961
+
+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": 0.048681941636320565,
+      "wer": 0.1728699877209892
+    },
+    {
+      "subset_type": "age",
+      "subset": "",
+      "cer": 0.04846238875347202,
+      "wer": 0.17210516026485562
+    },
+    {
+      "subset_type": "age",
+      "subset": "twenties",
+      "cer": 0.05243791380540993,
+      "wer": 0.1834889362640594
+    },
+    {
+      "subset_type": "age",
+      "subset": "fourties",
+      "cer": 0.042285074208905066,
+      "wer": 0.15131164742917103
+    },
+    {
+      "subset_type": "age",
+      "subset": "thirties",
+      "cer": 0.047573654924737856,
+      "wer": 0.17404519624686138
+    },
+    {
+      "subset_type": "age",
+      "subset": "fifties",
+      "cer": 0.046173435196354376,
+      "wer": 0.1653290529695024
+    },
+    {
+      "subset_type": "age",
+      "subset": "sixties",
+      "cer": 0.057351923267081974,
+      "wer": 0.17815420560747663
+    },
+    {
+      "subset_type": "age",
+      "subset": "seventies",
+      "cer": 0.03306451612903226,
+      "wer": 0.13679245283018868
+    },
+    {
+      "subset_type": "age",
+      "subset": "teens",
+      "cer": 0.05116952013503737,
+      "wer": 0.1891891891891892
+    },
+    {
+      "subset_type": "gender",
+      "subset": "",
+      "cer": 0.04836802805747256,
+      "wer": 0.17182185960121113
+    },
+    {
+      "subset_type": "gender",
+      "subset": "other",
+      "cer": 0.06672705314009662,
+      "wer": 0.24954792043399637
+    },
+    {
+      "subset_type": "gender",
+      "subset": "male",
+      "cer": 0.05030814793693174,
+      "wer": 0.17668255259793095
+    },
+    {
+      "subset_type": "gender",
+      "subset": "female",
+      "cer": 0.04661096069461301,
+      "wer": 0.17048134276598695
+    },
+    {
+      "subset_type": "accent",
+      "subset": "",
+      "cer": 0.048345323741007196,
+      "wer": 0.17182928380613152
+    },
+    {
+      "subset_type": "accent",
+      "subset": "caribe",
+      "cer": 0.04873445825932504,
+      "wer": 0.17173404610758436
+    },
+    {
+      "subset_type": "accent",
+      "subset": "mexicano",
+      "cer": 0.04889535286377173,
+      "wer": 0.17899101589495509
+    },
+    {
+      "subset_type": "accent",
+      "subset": "surpeninsular",
+      "cer": 0.055020150692132466,
+      "wer": 0.17450365726227796
+    },
+    {
+      "subset_type": "accent",
+      "subset": "chileno",
+      "cer": 0.06540647106575438,
+      "wer": 0.2145305003427005
+    },
+    {
+      "subset_type": "accent",
+      "subset": "canario",
+      "cer": 0.04842259721203228,
+      "wer": 0.19724770642201836
+    },
+    {
+      "subset_type": "accent",
+      "subset": "americacentral",
+      "cer": 0.05571987405781891,
+      "wer": 0.18764434180138567
+    },
+    {
+      "subset_type": "accent",
+      "subset": "nortepeninsular",
+      "cer": 0.040903890160183067,
+      "wer": 0.15182072829131651
+    },
+    {
+      "subset_type": "accent",
+      "subset": "andino",
+      "cer": 0.04854480177220991,
+      "wer": 0.16875142142369798
+    },
+    {
+      "subset_type": "accent",
+      "subset": "rioplatense",
+      "cer": 0.04725718356684873,
+      "wer": 0.17836676217765043
+    },
+    {
+      "subset_type": "accent",
+      "subset": "centrosurpeninsular",
+      "cer": 0.05660981663864201,
+      "wer": 0.190253045923149
+    }
+  ],
+  "test": [
+    {
+      "subset_type": "all",
+      "subset": "all",
+      "cer": 0.056769016358015165,
+      "wer": 0.18968148573920282
+    },
+    {
+      "subset_type": "age",
+      "subset": "",
+      "cer": 0.05704911750082864,
+      "wer": 0.19028553693358163
+    },
+    {
+      "subset_type": "age",
+      "subset": "twenties",
+      "cer": 0.05811366142126892,
+      "wer": 0.19978401727861772
+    },
+    {
+      "subset_type": "age",
+      "subset": "fourties",
+      "cer": 0.05067726359868393,
+      "wer": 0.1601029601029601
+    },
+    {
+      "subset_type": "age",
+      "subset": "thirties",
+      "cer": 0.049498481995311476,
+      "wer": 0.1754022988505747
+    },
+    {
+      "subset_type": "age",
+      "subset": "fifties",
+      "cer": 0.04767742814330473,
+      "wer": 0.15341138473960436
+    },
+    {
+      "subset_type": "age",
+      "subset": "sixties",
+      "cer": 0.056743101438010105,
+      "wer": 0.1845444059976932
+    },
+    {
+      "subset_type": "age",
+      "subset": "seventies",
+      "cer": 0.11850311850311851,
+      "wer": 0.32926829268292684
+    },
+    {
+      "subset_type": "age",
+      "subset": "teens",
+      "cer": 0.0670620745433713,
+      "wer": 0.22995522995522996
+    },
+    {
+      "subset_type": "age",
+      "subset": "eighties",
+      "cer": 0.053691275167785234,
+      "wer": 0.14814814814814814
+    },
+    {
+      "subset_type": "gender",
+      "subset": "",
+      "cer": 0.05712343267937024,
+      "wer": 0.19051159908601525
+    },
+    {
+      "subset_type": "gender",
+      "subset": "other",
+      "cer": 0.10688665710186514,
+      "wer": 0.28695652173913044
+    },
+    {
+      "subset_type": "gender",
+      "subset": "male",
+      "cer": 0.05501386701197298,
+      "wer": 0.1845726841679195
+    },
+    {
+      "subset_type": "gender",
+      "subset": "female",
+      "cer": 0.051046290424857324,
+      "wer": 0.18097251585623678
+    },
+    {
+      "subset_type": "accent",
+      "subset": "",
+      "cer": 0.05742007807619466,
+      "wer": 0.19121684117059454
+    },
+    {
+      "subset_type": "accent",
+      "subset": "caribe",
+      "cer": 0.05832968316542561,
+      "wer": 0.19632164242942687
+    },
+    {
+      "subset_type": "accent",
+      "subset": "surpeninsular",
+      "cer": 0.0416258570029383,
+      "wer": 0.14962962962962964
+    },
+    {
+      "subset_type": "accent",
+      "subset": "mexicano",
+      "cer": 0.05273972602739726,
+      "wer": 0.18270844991606044
+    },
+    {
+      "subset_type": "accent",
+      "subset": "chileno",
+      "cer": 0.052128315854410856,
+      "wer": 0.17200365965233302
+    },
+    {
+      "subset_type": "accent",
+      "subset": "canario",
+      "cer": 0.05555555555555555,
+      "wer": 0.21656050955414013
+    },
+    {
+      "subset_type": "accent",
+      "subset": "americacentral",
+      "cer": 0.05321798400838904,
+      "wer": 0.18181818181818182
+    },
+    {
+      "subset_type": "accent",
+      "subset": "nortepeninsular",
+      "cer": 0.04370683434697454,
+      "wer": 0.15179113539769276
+    },
+    {
+      "subset_type": "accent",
+      "subset": "andino",
+      "cer": 0.045729507741828074,
+      "wer": 0.17206815903775477
+    },
+    {
+      "subset_type": "accent",
+      "subset": "rioplatense",
+      "cer": 0.06843434343434343,
+      "wer": 0.2058165548098434
+    },
+    {
+      "subset_type": "accent",
+      "subset": "centrosurpeninsular",
+      "cer": 0.04048090092832141,
+      "wer": 0.14285714285714285
+    }
+  ]
+}

cv7_transcriptions.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:87e29aa7c97ad374281c8e1f9ec33f1c0013678a6d70f6e2bd4c709aa2928c1a
+size 125340458

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 sha256:915204056e285d2128ae8ace64ffcafa3fc699424893a80defa1c18a53ddc060
+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/tmp2dhs66jd", "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}