jonatasgrosman/exp_w2v2r_es_xls-r_age_teens-5_sixties-5_s530

.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-5_sixties-5_s530
+
+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": 0.03729788394475449,
+      "wer": 0.13718163693737703
+    },
+    {
+      "subset_type": "age",
+      "subset": "",
+      "cer": 0.03729947281900119,
+      "wer": 0.13711123766457853
+    },
+    {
+      "subset_type": "age",
+      "subset": "twenties",
+      "cer": 0.039233060129464176,
+      "wer": 0.1432772182606778
+    },
+    {
+      "subset_type": "age",
+      "subset": "fourties",
+      "cer": 0.030733688042565106,
+      "wer": 0.11941238195173137
+    },
+    {
+      "subset_type": "age",
+      "subset": "thirties",
+      "cer": 0.036999268438670775,
+      "wer": 0.13730672657592177
+    },
+    {
+      "subset_type": "age",
+      "subset": "fifties",
+      "cer": 0.03605414823750167,
+      "wer": 0.13041733547351525
+    },
+    {
+      "subset_type": "age",
+      "subset": "sixties",
+      "cer": 0.04469494709779492,
+      "wer": 0.13901869158878505
+    },
+    {
+      "subset_type": "age",
+      "subset": "seventies",
+      "cer": 0.029838709677419355,
+      "wer": 0.12735849056603774
+    },
+    {
+      "subset_type": "age",
+      "subset": "teens",
+      "cer": 0.03708705088015433,
+      "wer": 0.144335825186889
+    },
+    {
+      "subset_type": "gender",
+      "subset": "",
+      "cer": 0.037236961194705286,
+      "wer": 0.13696387831087212
+    },
+    {
+      "subset_type": "gender",
+      "subset": "other",
+      "cer": 0.042874396135265704,
+      "wer": 0.15370705244122965
+    },
+    {
+      "subset_type": "gender",
+      "subset": "male",
+      "cer": 0.03817664556182724,
+      "wer": 0.13933124103994732
+    },
+    {
+      "subset_type": "gender",
+      "subset": "female",
+      "cer": 0.03470731024180749,
+      "wer": 0.13177971979470107
+    },
+    {
+      "subset_type": "accent",
+      "subset": "",
+      "cer": 0.03720319596857077,
+      "wer": 0.13686235847856507
+    },
+    {
+      "subset_type": "accent",
+      "subset": "caribe",
+      "cer": 0.036245559502664296,
+      "wer": 0.13364517206815904
+    },
+    {
+      "subset_type": "accent",
+      "subset": "mexicano",
+      "cer": 0.03693699933051689,
+      "wer": 0.13614374568071874
+    },
+    {
+      "subset_type": "accent",
+      "subset": "surpeninsular",
+      "cer": 0.04625898019975469,
+      "wer": 0.15569487983281086
+    },
+    {
+      "subset_type": "accent",
+      "subset": "chileno",
+      "cer": 0.04754725733503421,
+      "wer": 0.17135023989033585
+    },
+    {
+      "subset_type": "accent",
+      "subset": "canario",
+      "cer": 0.03301540719002201,
+      "wer": 0.14678899082568808
+    },
+    {
+      "subset_type": "accent",
+      "subset": "americacentral",
+      "cer": 0.03759183284037783,
+      "wer": 0.13568129330254042
+    },
+    {
+      "subset_type": "accent",
+      "subset": "nortepeninsular",
+      "cer": 0.03327612509534706,
+      "wer": 0.1226890756302521
+    },
+    {
+      "subset_type": "accent",
+      "subset": "andino",
+      "cer": 0.032694217401268044,
+      "wer": 0.1253127132135547
+    },
+    {
+      "subset_type": "accent",
+      "subset": "rioplatense",
+      "cer": 0.03805509380194728,
+      "wer": 0.14255014326647564
+    },
+    {
+      "subset_type": "accent",
+      "subset": "centrosurpeninsular",
+      "cer": 0.04684122027384098,
+      "wer": 0.1626054358013121
+    }
+  ],
+  "test": [
+    {
+      "subset_type": "all",
+      "subset": "all",
+      "cer": 0.04255553742408673,
+      "wer": 0.1498556412238453
+    },
+    {
+      "subset_type": "age",
+      "subset": "",
+      "cer": 0.04275511683791846,
+      "wer": 0.14986809435133458
+    },
+    {
+      "subset_type": "age",
+      "subset": "twenties",
+      "cer": 0.04345995790317731,
+      "wer": 0.15970722342212623
+    },
+    {
+      "subset_type": "age",
+      "subset": "fourties",
+      "cer": 0.03700380293124813,
+      "wer": 0.13745173745173744
+    },
+    {
+      "subset_type": "age",
+      "subset": "thirties",
+      "cer": 0.03885323392644403,
+      "wer": 0.14183908045977012
+    },
+    {
+      "subset_type": "age",
+      "subset": "fifties",
+      "cer": 0.0406620351450756,
+      "wer": 0.1303996770286637
+    },
+    {
+      "subset_type": "age",
+      "subset": "sixties",
+      "cer": 0.038087835211815003,
+      "wer": 0.13725490196078433
+    },
+    {
+      "subset_type": "age",
+      "subset": "seventies",
+      "cer": 0.09355509355509356,
+      "wer": 0.25609756097560976
+    },
+    {
+      "subset_type": "age",
+      "subset": "teens",
+      "cer": 0.04637056008627081,
+      "wer": 0.17134717134717134
+    },
+    {
+      "subset_type": "age",
+      "subset": "eighties",
+      "cer": 0.020134228187919462,
+      "wer": 0.037037037037037035
+    },
+    {
+      "subset_type": "gender",
+      "subset": "",
+      "cer": 0.042799921932859676,
+      "wer": 0.1500019364083498
+    },
+    {
+      "subset_type": "gender",
+      "subset": "other",
+      "cer": 0.0667144906743185,
+      "wer": 0.21304347826086956
+    },
+    {
+      "subset_type": "gender",
+      "subset": "male",
+      "cer": 0.041069547655170416,
+      "wer": 0.14889557071816814
+    },
+    {
+      "subset_type": "gender",
+      "subset": "female",
+      "cer": 0.04012541393644754,
+      "wer": 0.1463002114164905
+    },
+    {
+      "subset_type": "accent",
+      "subset": "",
+      "cer": 0.04293203025267353,
+      "wer": 0.1506160050316327
+    },
+    {
+      "subset_type": "accent",
+      "subset": "caribe",
+      "cer": 0.04416703168345744,
+      "wer": 0.15526090675791274
+    },
+    {
+      "subset_type": "accent",
+      "subset": "surpeninsular",
+      "cer": 0.04113614103819784,
+      "wer": 0.15851851851851853
+    },
+    {
+      "subset_type": "accent",
+      "subset": "mexicano",
+      "cer": 0.04009132420091324,
+      "wer": 0.1505316172355904
+    },
+    {
+      "subset_type": "accent",
+      "subset": "chileno",
+      "cer": 0.03578038247995065,
+      "wer": 0.12900274473924978
+    },
+    {
+      "subset_type": "accent",
+      "subset": "canario",
+      "cer": 0.043209876543209874,
+      "wer": 0.1592356687898089
+    },
+    {
+      "subset_type": "accent",
+      "subset": "americacentral",
+      "cer": 0.03958579106042732,
+      "wer": 0.14184952978056425
+    },
+    {
+      "subset_type": "accent",
+      "subset": "nortepeninsular",
+      "cer": 0.03597567261107102,
+      "wer": 0.12021857923497267
+    },
+    {
+      "subset_type": "accent",
+      "subset": "andino",
+      "cer": 0.03490759753593429,
+      "wer": 0.136986301369863
+    },
+    {
+      "subset_type": "accent",
+      "subset": "rioplatense",
+      "cer": 0.05012626262626262,
+      "wer": 0.16331096196868009
+    },
+    {
+      "subset_type": "accent",
+      "subset": "centrosurpeninsular",
+      "cer": 0.03348044437680718,
+      "wer": 0.12946428571428573
+    }
+  ]
+}

cv7_transcriptions.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5757a0ef4a5a09fa2c34b18817991b259807b60b3601785634f95e2f27bcd3c3
+size 125542227

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 sha256:0c5fda744c0fb1fd4d16ebe7173057dd6d06c37a86dc9901cece510ab6b122b0
+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/tmp5fokul2l", "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}