jonatasgrosman/exp_w2v2r_es_xls-r_accent_surpeninsular-5_nortepeninsular-5_s362

.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_s362
+
+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": 0.04481673892803429,
+      "wer": 0.1612775452540963
+    },
+    {
+      "subset_type": "age",
+      "subset": "",
+      "cer": 0.04459497760897908,
+      "wer": 0.16057935044071772
+    },
+    {
+      "subset_type": "age",
+      "subset": "sixties",
+      "cer": 0.052210026698309105,
+      "wer": 0.16238317757009346
+    },
+    {
+      "subset_type": "age",
+      "subset": "seventies",
+      "cer": 0.03951612903225806,
+      "wer": 0.16037735849056603
+    },
+    {
+      "subset_type": "age",
+      "subset": "fourties",
+      "cer": 0.040639876785214225,
+      "wer": 0.1458551941238195
+    },
+    {
+      "subset_type": "age",
+      "subset": "thirties",
+      "cer": 0.043782837127845885,
+      "wer": 0.16254790537861769
+    },
+    {
+      "subset_type": "age",
+      "subset": "teens",
+      "cer": 0.04832408970340005,
+      "wer": 0.17538815411155836
+    },
+    {
+      "subset_type": "age",
+      "subset": "fifties",
+      "cer": 0.04181745074386811,
+      "wer": 0.14847512038523275
+    },
+    {
+      "subset_type": "age",
+      "subset": "twenties",
+      "cer": 0.047601466932486645,
+      "wer": 0.17062412703080204
+    },
+    {
+      "subset_type": "gender",
+      "subset": "",
+      "cer": 0.04452568163819436,
+      "wer": 0.16040608276014146
+    },
+    {
+      "subset_type": "gender",
+      "subset": "female",
+      "cer": 0.0410559238166371,
+      "wer": 0.15314190595089472
+    },
+    {
+      "subset_type": "gender",
+      "subset": "male",
+      "cer": 0.04677627358093177,
+      "wer": 0.16629857801542097
+    },
+    {
+      "subset_type": "gender",
+      "subset": "other",
+      "cer": 0.06431159420289854,
+      "wer": 0.21880650994575046
+    },
+    {
+      "subset_type": "accent",
+      "subset": "",
+      "cer": 0.0445446407850678,
+      "wer": 0.16071428571428573
+    },
+    {
+      "subset_type": "accent",
+      "subset": "rioplatense",
+      "cer": 0.04577297554025172,
+      "wer": 0.17478510028653296
+    },
+    {
+      "subset_type": "accent",
+      "subset": "mexicano",
+      "cer": 0.04654061915645128,
+      "wer": 0.17125086385625432
+    },
+    {
+      "subset_type": "accent",
+      "subset": "americacentral",
+      "cer": 0.053143783990077284,
+      "wer": 0.17263279445727484
+    },
+    {
+      "subset_type": "accent",
+      "subset": "canario",
+      "cer": 0.0425531914893617,
+      "wer": 0.1834862385321101
+    },
+    {
+      "subset_type": "accent",
+      "subset": "andino",
+      "cer": 0.04598579176533496,
+      "wer": 0.15897202638162383
+    },
+    {
+      "subset_type": "accent",
+      "subset": "nortepeninsular",
+      "cer": 0.034420289855072464,
+      "wer": 0.12717086834733893
+    },
+    {
+      "subset_type": "accent",
+      "subset": "surpeninsular",
+      "cer": 0.040651831084632904,
+      "wer": 0.13479623824451412
+    },
+    {
+      "subset_type": "accent",
+      "subset": "centrosurpeninsular",
+      "cer": 0.047241572583873806,
+      "wer": 0.16026241799437677
+    },
+    {
+      "subset_type": "accent",
+      "subset": "chileno",
+      "cer": 0.05543314391743013,
+      "wer": 0.1925976696367375
+    },
+    {
+      "subset_type": "accent",
+      "subset": "caribe",
+      "cer": 0.04540408525754885,
+      "wer": 0.15770130304042768
+    }
+  ],
+  "test": [
+    {
+      "subset_type": "all",
+      "subset": "all",
+      "cer": 0.052345428003122536,
+      "wer": 0.17896524157797025
+    },
+    {
+      "subset_type": "age",
+      "subset": "",
+      "cer": 0.052718190669539276,
+      "wer": 0.17998137802607075
+    },
+    {
+      "subset_type": "age",
+      "subset": "sixties",
+      "cer": 0.04858142246404975,
+      "wer": 0.1568627450980392
+    },
+    {
+      "subset_type": "age",
+      "subset": "seventies",
+      "cer": 0.0997920997920998,
+      "wer": 0.2926829268292683
+    },
+    {
+      "subset_type": "age",
+      "subset": "fourties",
+      "cer": 0.04760073494851087,
+      "wer": 0.15778635778635777
+    },
+    {
+      "subset_type": "age",
+      "subset": "thirties",
+      "cer": 0.04592444564005995,
+      "wer": 0.17126436781609194
+    },
+    {
+      "subset_type": "age",
+      "subset": "fifties",
+      "cer": 0.047881759978204604,
+      "wer": 0.14897052886556317
+    },
+    {
+      "subset_type": "age",
+      "subset": "teens",
+      "cer": 0.05856979173687403,
+      "wer": 0.2022792022792023
+    },
+    {
+      "subset_type": "age",
+      "subset": "twenties",
+      "cer": 0.0516588152751328,
+      "wer": 0.18070554355651547
+    },
+    {
+      "subset_type": "age",
+      "subset": "eighties",
+      "cer": 0.013422818791946308,
+      "wer": 0.07407407407407407
+    },
+    {
+      "subset_type": "gender",
+      "subset": "",
+      "cer": 0.052753482324384475,
+      "wer": 0.18011695906432748
+    },
+    {
+      "subset_type": "gender",
+      "subset": "female",
+      "cer": 0.04815754245050377,
+      "wer": 0.1649048625792812
+    },
+    {
+      "subset_type": "gender",
+      "subset": "male",
+      "cer": 0.05002754075548641,
+      "wer": 0.17318145021394704
+    },
+    {
+      "subset_type": "gender",
+      "subset": "other",
+      "cer": 0.08321377331420372,
+      "wer": 0.2565217391304348
+    },
+    {
+      "subset_type": "accent",
+      "subset": "",
+      "cer": 0.05297798211803283,
+      "wer": 0.18072440711828036
+    },
+    {
+      "subset_type": "accent",
+      "subset": "rioplatense",
+      "cer": 0.06363636363636363,
+      "wer": 0.19686800894854586
+    },
+    {
+      "subset_type": "accent",
+      "subset": "mexicano",
+      "cer": 0.05050228310502283,
+      "wer": 0.17207610520425293
+    },
+    {
+      "subset_type": "accent",
+      "subset": "americacentral",
+      "cer": 0.04928562065801547,
+      "wer": 0.1677115987460815
+    },
+    {
+      "subset_type": "accent",
+      "subset": "canario",
+      "cer": 0.046296296296296294,
+      "wer": 0.17834394904458598
+    },
+    {
+      "subset_type": "accent",
+      "subset": "andino",
+      "cer": 0.04323214384816027,
+      "wer": 0.16371533578349481
+    },
+    {
+      "subset_type": "accent",
+      "subset": "nortepeninsular",
+      "cer": 0.03824348005360272,
+      "wer": 0.1366120218579235
+    },
+    {
+      "subset_type": "accent",
+      "subset": "surpeninsular",
+      "cer": 0.03844270323212537,
+      "wer": 0.13925925925925925
+    },
+    {
+      "subset_type": "accent",
+      "subset": "centrosurpeninsular",
+      "cer": 0.03348044437680718,
+      "wer": 0.12946428571428573
+    },
+    {
+      "subset_type": "accent",
+      "subset": "chileno",
+      "cer": 0.03747686613201728,
+      "wer": 0.12900274473924978
+    },
+    {
+      "subset_type": "accent",
+      "subset": "caribe",
+      "cer": 0.055701562271864506,
+      "wer": 0.19161676646706588
+    }
+  ]
+}

cv7_transcriptions.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:09af0c5af4dde9aac043dee67c577563ff29480ba4aedc99c3a9d55a5998de80
+size 125521140

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 sha256:3dbe136576e7ddff03e81c63a2e24840504095fde04fffb4923943e7de5de97f
+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/tmpfbke9go1", "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}