jonatasgrosman/exp_w2v2r_es_vp-100k_age_teens-8_sixties-2_s130

.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-8_sixties-2_s130
+
+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": 0.05266503963816061,
+      "wer": 0.1872285084302671
+    },
+    {
+      "subset_type": "age",
+      "subset": "",
+      "cer": 0.05250552689756816,
+      "wer": 0.18691191595338677
+    },
+    {
+      "subset_type": "age",
+      "subset": "thirties",
+      "cer": 0.05205169700059855,
+      "wer": 0.18950707017312013
+    },
+    {
+      "subset_type": "age",
+      "subset": "sixties",
+      "cer": 0.07287649559972313,
+      "wer": 0.20677570093457945
+    },
+    {
+      "subset_type": "age",
+      "subset": "seventies",
+      "cer": 0.05967741935483871,
+      "wer": 0.24056603773584906
+    },
+    {
+      "subset_type": "age",
+      "subset": "fifties",
+      "cer": 0.052204798284412275,
+      "wer": 0.1926163723916533
+    },
+    {
+      "subset_type": "age",
+      "subset": "twenties",
+      "cer": 0.05461616087031431,
+      "wer": 0.19186944056458136
+    },
+    {
+      "subset_type": "age",
+      "subset": "teens",
+      "cer": 0.05068724379069207,
+      "wer": 0.18286371477860838
+    },
+    {
+      "subset_type": "age",
+      "subset": "fourties",
+      "cer": 0.0462405488658639,
+      "wer": 0.16915005246589718
+    },
+    {
+      "subset_type": "gender",
+      "subset": "",
+      "cer": 0.05243523022966399,
+      "wer": 0.18678280339589676
+    },
+    {
+      "subset_type": "gender",
+      "subset": "male",
+      "cer": 0.0541122034359629,
+      "wer": 0.18912007439265371
+    },
+    {
+      "subset_type": "gender",
+      "subset": "other",
+      "cer": 0.06280193236714976,
+      "wer": 0.2423146473779385
+    },
+    {
+      "subset_type": "gender",
+      "subset": "female",
+      "cer": 0.05046214172346186,
+      "wer": 0.18352059925093633
+    },
+    {
+      "subset_type": "accent",
+      "subset": "",
+      "cer": 0.052415210688591986,
+      "wer": 0.1871501717338761
+    },
+    {
+      "subset_type": "accent",
+      "subset": "chileno",
+      "cer": 0.060999652093239015,
+      "wer": 0.22069910897875258
+    },
+    {
+      "subset_type": "accent",
+      "subset": "nortepeninsular",
+      "cer": 0.05167810831426392,
+      "wer": 0.17591036414565828
+    },
+    {
+      "subset_type": "accent",
+      "subset": "canario",
+      "cer": 0.04181951577402788,
+      "wer": 0.15137614678899083
+    },
+    {
+      "subset_type": "accent",
+      "subset": "surpeninsular",
+      "cer": 0.06658489574207115,
+      "wer": 0.2277951933124347
+    },
+    {
+      "subset_type": "accent",
+      "subset": "caribe",
+      "cer": 0.051287744227353466,
+      "wer": 0.1807550952221851
+    },
+    {
+      "subset_type": "accent",
+      "subset": "mexicano",
+      "cer": 0.04947249347831106,
+      "wer": 0.17664132688320663
+    },
+    {
+      "subset_type": "accent",
+      "subset": "rioplatense",
+      "cer": 0.05254096414153408,
+      "wer": 0.18803724928366763
+    },
+    {
+      "subset_type": "accent",
+      "subset": "andino",
+      "cer": 0.053051714918646395,
+      "wer": 0.18603593359108483
+    },
+    {
+      "subset_type": "accent",
+      "subset": "centrosurpeninsular",
+      "cer": 0.0662182720794299,
+      "wer": 0.204311152764761
+    },
+    {
+      "subset_type": "accent",
+      "subset": "americacentral",
+      "cer": 0.05724644594981395,
+      "wer": 0.1945727482678984
+    }
+  ],
+  "test": [
+    {
+      "subset_type": "all",
+      "subset": "all",
+      "cer": 0.059824291365843545,
+      "wer": 0.2020494321447684
+    },
+    {
+      "subset_type": "age",
+      "subset": "",
+      "cer": 0.06018110913158767,
+      "wer": 0.20299503414028552
+    },
+    {
+      "subset_type": "age",
+      "subset": "thirties",
+      "cer": 0.05549363975250759,
+      "wer": 0.18735632183908046
+    },
+    {
+      "subset_type": "age",
+      "subset": "sixties",
+      "cer": 0.056160124368441505,
+      "wer": 0.19261822376009227
+    },
+    {
+      "subset_type": "age",
+      "subset": "seventies",
+      "cer": 0.12058212058212059,
+      "wer": 0.2926829268292683
+    },
+    {
+      "subset_type": "age",
+      "subset": "fifties",
+      "cer": 0.05564636970439994,
+      "wer": 0.18813080339119903
+    },
+    {
+      "subset_type": "age",
+      "subset": "twenties",
+      "cer": 0.05879522902676155,
+      "wer": 0.1988240940724742
+    },
+    {
+      "subset_type": "age",
+      "subset": "eighties",
+      "cer": 0.053691275167785234,
+      "wer": 0.25925925925925924
+    },
+    {
+      "subset_type": "age",
+      "subset": "teens",
+      "cer": 0.06551189593583609,
+      "wer": 0.2328042328042328
+    },
+    {
+      "subset_type": "age",
+      "subset": "fourties",
+      "cer": 0.05366833311968551,
+      "wer": 0.18326898326898328
+    },
+    {
+      "subset_type": "gender",
+      "subset": "",
+      "cer": 0.06024224078542781,
+      "wer": 0.20323767476085358
+    },
+    {
+      "subset_type": "gender",
+      "subset": "male",
+      "cer": 0.05670496603306823,
+      "wer": 0.19272580085578814
+    },
+    {
+      "subset_type": "gender",
+      "subset": "other",
+      "cer": 0.09684361549497848,
+      "wer": 0.2782608695652174
+    },
+    {
+      "subset_type": "gender",
+      "subset": "female",
+      "cer": 0.05798633129007257,
+      "wer": 0.2
+    },
+    {
+      "subset_type": "accent",
+      "subset": "",
+      "cer": 0.060442382385694424,
+      "wer": 0.20394391209441712
+    },
+    {
+      "subset_type": "accent",
+      "subset": "chileno",
+      "cer": 0.0445712523133868,
+      "wer": 0.1637694419030192
+    },
+    {
+      "subset_type": "accent",
+      "subset": "nortepeninsular",
+      "cer": 0.05071642098752706,
+      "wer": 0.16393442622950818
+    },
+    {
+      "subset_type": "accent",
+      "subset": "canario",
+      "cer": 0.06995884773662552,
+      "wer": 0.2356687898089172
+    },
+    {
+      "subset_type": "accent",
+      "subset": "surpeninsular",
+      "cer": 0.05582761998041136,
+      "wer": 0.2162962962962963
+    },
+    {
+      "subset_type": "accent",
+      "subset": "caribe",
+      "cer": 0.06709008614396263,
+      "wer": 0.21471343028229256
+    },
+    {
+      "subset_type": "accent",
+      "subset": "mexicano",
+      "cer": 0.054383561643835614,
+      "wer": 0.182988248461108
+    },
+    {
+      "subset_type": "accent",
+      "subset": "rioplatense",
+      "cer": 0.06982323232323233,
+      "wer": 0.2140193885160328
+    },
+    {
+      "subset_type": "accent",
+      "subset": "andino",
+      "cer": 0.049836283922526224,
+      "wer": 0.1810892081523555
+    },
+    {
+      "subset_type": "accent",
+      "subset": "centrosurpeninsular",
+      "cer": 0.04169837163293258,
+      "wer": 0.16160714285714287
+    },
+    {
+      "subset_type": "accent",
+      "subset": "americacentral",
+      "cer": 0.05098964477651068,
+      "wer": 0.17398119122257052
+    }
+  ]
+}

cv7_transcriptions.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b3fc3c51f759937235d83aeebb007e2a040572af574bb653482f502eba32a6d3
+size 125344784

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 sha256:07076596b9fe044b97989d6306fba11d8f644ad79219652e4d6c8a0c6dfd5fd6
+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/tmptlsn_658", "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}