jonatasgrosman/exp_w2v2r_es_vp-100k_age_teens-2_sixties-8_s869

.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_s869
+
+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": 0.056840957045345614,
+      "wer": 0.19944810467526639
+    },
+    {
+      "subset_type": "age",
+      "subset": "",
+      "cer": 0.05661952270279463,
+      "wer": 0.19862472269207557
+    },
+    {
+      "subset_type": "age",
+      "subset": "thirties",
+      "cer": 0.05650757055133122,
+      "wer": 0.20153297211576582
+    },
+    {
+      "subset_type": "age",
+      "subset": "sixties",
+      "cer": 0.07445861762088402,
+      "wer": 0.21728971962616822
+    },
+    {
+      "subset_type": "age",
+      "subset": "seventies",
+      "cer": 0.06129032258064516,
+      "wer": 0.24528301886792453
+    },
+    {
+      "subset_type": "age",
+      "subset": "fifties",
+      "cer": 0.05475137381048117,
+      "wer": 0.1958266452648475
+    },
+    {
+      "subset_type": "age",
+      "subset": "twenties",
+      "cer": 0.059415688301459546,
+      "wer": 0.2074542380357274
+    },
+    {
+      "subset_type": "age",
+      "subset": "teens",
+      "cer": 0.05589582830962141,
+      "wer": 0.20759056929269695
+    },
+    {
+      "subset_type": "age",
+      "subset": "fourties",
+      "cer": 0.05086110333239989,
+      "wer": 0.18111227701993704
+    },
+    {
+      "subset_type": "gender",
+      "subset": "",
+      "cer": 0.056532130331485464,
+      "wer": 0.19838516797136727
+    },
+    {
+      "subset_type": "gender",
+      "subset": "male",
+      "cer": 0.05889481494922526,
+      "wer": 0.20388236661629663
+    },
+    {
+      "subset_type": "gender",
+      "subset": "other",
+      "cer": 0.06823671497584541,
+      "wer": 0.24954792043399637
+    },
+    {
+      "subset_type": "gender",
+      "subset": "female",
+      "cer": 0.053659789001960605,
+      "wer": 0.19711471771396866
+    },
+    {
+      "subset_type": "accent",
+      "subset": "",
+      "cer": 0.05639094254550277,
+      "wer": 0.1983605775346648
+    },
+    {
+      "subset_type": "accent",
+      "subset": "chileno",
+      "cer": 0.06842166299431753,
+      "wer": 0.2385195339273475
+    },
+    {
+      "subset_type": "accent",
+      "subset": "nortepeninsular",
+      "cer": 0.05368039664378337,
+      "wer": 0.1843137254901961
+    },
+    {
+      "subset_type": "accent",
+      "subset": "canario",
+      "cer": 0.05355832721936904,
+      "wer": 0.19724770642201836
+    },
+    {
+      "subset_type": "accent",
+      "subset": "surpeninsular",
+      "cer": 0.06693534256176625,
+      "wer": 0.2173458725182863
+    },
+    {
+      "subset_type": "accent",
+      "subset": "caribe",
+      "cer": 0.05755994671403197,
+      "wer": 0.20180420982292016
+    },
+    {
+      "subset_type": "accent",
+      "subset": "mexicano",
+      "cer": 0.056698293972343425,
+      "wer": 0.20262612301313063
+    },
+    {
+      "subset_type": "accent",
+      "subset": "rioplatense",
+      "cer": 0.056162431726430775,
+      "wer": 0.19806590257879655
+    },
+    {
+      "subset_type": "accent",
+      "subset": "andino",
+      "cer": 0.058704453441295545,
+      "wer": 0.20150102342506254
+    },
+    {
+      "subset_type": "accent",
+      "subset": "centrosurpeninsular",
+      "cer": 0.07006165425574505,
+      "wer": 0.21836925960637302
+    },
+    {
+      "subset_type": "accent",
+      "subset": "americacentral",
+      "cer": 0.057818910409312085,
+      "wer": 0.20785219399538107
+    }
+  ],
+  "test": [
+    {
+      "subset_type": "all",
+      "subset": "all",
+      "cer": 0.06539016225704444,
+      "wer": 0.21713938384899478
+    },
+    {
+      "subset_type": "age",
+      "subset": "",
+      "cer": 0.06580937189260855,
+      "wer": 0.21792364990689014
+    },
+    {
+      "subset_type": "age",
+      "subset": "thirties",
+      "cer": 0.06237269897390569,
+      "wer": 0.21471264367816092
+    },
+    {
+      "subset_type": "age",
+      "subset": "sixties",
+      "cer": 0.060629615235134084,
+      "wer": 0.19492502883506344
+    },
+    {
+      "subset_type": "age",
+      "subset": "seventies",
+      "cer": 0.14553014553014554,
+      "wer": 0.32926829268292684
+    },
+    {
+      "subset_type": "age",
+      "subset": "fifties",
+      "cer": 0.05884756845116469,
+      "wer": 0.19620508679854662
+    },
+    {
+      "subset_type": "age",
+      "subset": "twenties",
+      "cer": 0.06426781597674652,
+      "wer": 0.21694264458843293
+    },
+    {
+      "subset_type": "age",
+      "subset": "eighties",
+      "cer": 0.08053691275167785,
+      "wer": 0.25925925925925924
+    },
+    {
+      "subset_type": "age",
+      "subset": "teens",
+      "cer": 0.06921884477994204,
+      "wer": 0.2442002442002442
+    },
+    {
+      "subset_type": "age",
+      "subset": "fourties",
+      "cer": 0.05828312609494509,
+      "wer": 0.1927927927927928
+    },
+    {
+      "subset_type": "gender",
+      "subset": "",
+      "cer": 0.06585816338004867,
+      "wer": 0.21807830835366562
+    },
+    {
+      "subset_type": "gender",
+      "subset": "male",
+      "cer": 0.06238705874394828,
+      "wer": 0.21059326934196831
+    },
+    {
+      "subset_type": "gender",
+      "subset": "other",
+      "cer": 0.09971305595408896,
+      "wer": 0.2956521739130435
+    },
+    {
+      "subset_type": "gender",
+      "subset": "female",
+      "cer": 0.06189670964559994,
+      "wer": 0.2116279069767442
+    },
+    {
+      "subset_type": "accent",
+      "subset": "",
+      "cer": 0.06607586872729877,
+      "wer": 0.2186836360945651
+    },
+    {
+      "subset_type": "accent",
+      "subset": "chileno",
+      "cer": 0.051511412708204814,
+      "wer": 0.18664226898444647
+    },
+    {
+      "subset_type": "accent",
+      "subset": "nortepeninsular",
+      "cer": 0.05545820018554788,
+      "wer": 0.18154219793564055
+    },
+    {
+      "subset_type": "accent",
+      "subset": "canario",
+      "cer": 0.059670781893004114,
+      "wer": 0.2229299363057325
+    },
+    {
+      "subset_type": "accent",
+      "subset": "surpeninsular",
+      "cer": 0.05729676787463271,
+      "wer": 0.2162962962962963
+    },
+    {
+      "subset_type": "accent",
+      "subset": "caribe",
+      "cer": 0.07147028763323113,
+      "wer": 0.22668947818648416
+    },
+    {
+      "subset_type": "accent",
+      "subset": "mexicano",
+      "cer": 0.05858447488584475,
+      "wer": 0.20201454952434247
+    },
+    {
+      "subset_type": "accent",
+      "subset": "rioplatense",
+      "cer": 0.075,
+      "wer": 0.2348993288590604
+    },
+    {
+      "subset_type": "accent",
+      "subset": "andino",
+      "cer": 0.05577446029191409,
+      "wer": 0.19846308052121617
+    },
+    {
+      "subset_type": "accent",
+      "subset": "centrosurpeninsular",
+      "cer": 0.04945974737482879,
+      "wer": 0.18125
+    },
+    {
+      "subset_type": "accent",
+      "subset": "americacentral",
+      "cer": 0.057150347358762614,
+      "wer": 0.20689655172413793
+    }
+  ]
+}

cv7_transcriptions.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4fbe2d9e46eeb297d7e19bf0eebeede590695fe26d4616dacf49ae81b42b3ba6
+size 125208573

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 sha256:18c10ac47f035e299471541179329dc0572859cb36639d7a14e7db0ccfc049b5
+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/tmpy0bgzmam", "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}