jonatasgrosman/exp_w2v2r_es_xls-r_accent_surpeninsular-0_nortepeninsular-10_s888

.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-0_nortepeninsular-10_s888
+
+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": 0.04781370339968157,
+      "wer": 0.17296901200174283
+    },
+    {
+      "subset_type": "age",
+      "subset": "",
+      "cer": 0.04748171872342838,
+      "wer": 0.17187566405154314
+    },
+    {
+      "subset_type": "age",
+      "subset": "sixties",
+      "cer": 0.057351923267081974,
+      "wer": 0.16705607476635514
+    },
+    {
+      "subset_type": "age",
+      "subset": "seventies",
+      "cer": 0.03387096774193549,
+      "wer": 0.17452830188679244
+    },
+    {
+      "subset_type": "age",
+      "subset": "fourties",
+      "cer": 0.040989918790254834,
+      "wer": 0.1515215110178384
+    },
+    {
+      "subset_type": "age",
+      "subset": "thirties",
+      "cer": 0.04819437362832251,
+      "wer": 0.17946345975948197
+    },
+    {
+      "subset_type": "age",
+      "subset": "teens",
+      "cer": 0.05203761755485893,
+      "wer": 0.18832662449683726
+    },
+    {
+      "subset_type": "age",
+      "subset": "fifties",
+      "cer": 0.044967162578742795,
+      "wer": 0.1653290529695024
+    },
+    {
+      "subset_type": "age",
+      "subset": "twenties",
+      "cer": 0.05135494351325408,
+      "wer": 0.18451812100272
+    },
+    {
+      "subset_type": "gender",
+      "subset": "",
+      "cer": 0.04740920918655957,
+      "wer": 0.17168615943073778
+    },
+    {
+      "subset_type": "gender",
+      "subset": "female",
+      "cer": 0.04558397908691999,
+      "wer": 0.16673602441392704
+    },
+    {
+      "subset_type": "gender",
+      "subset": "male",
+      "cer": 0.050094291324550094,
+      "wer": 0.1797822465031578
+    },
+    {
+      "subset_type": "gender",
+      "subset": "other",
+      "cer": 0.05676328502415459,
+      "wer": 0.20976491862567812
+    },
+    {
+      "subset_type": "accent",
+      "subset": "",
+      "cer": 0.04743029539502039,
+      "wer": 0.17207575372090064
+    },
+    {
+      "subset_type": "accent",
+      "subset": "rioplatense",
+      "cer": 0.05444075041557825,
+      "wer": 0.20522922636103152
+    },
+    {
+      "subset_type": "accent",
+      "subset": "mexicano",
+      "cer": 0.046563704781032854,
+      "wer": 0.17567380787836903
+    },
+    {
+      "subset_type": "accent",
+      "subset": "americacentral",
+      "cer": 0.05571987405781891,
+      "wer": 0.18995381062355657
+    },
+    {
+      "subset_type": "accent",
+      "subset": "canario",
+      "cer": 0.03961848862802641,
+      "wer": 0.1559633027522936
+    },
+    {
+      "subset_type": "accent",
+      "subset": "andino",
+      "cer": 0.04980520968604385,
+      "wer": 0.1737548328405731
+    },
+    {
+      "subset_type": "accent",
+      "subset": "nortepeninsular",
+      "cer": 0.031559877955758965,
+      "wer": 0.11876750700280111
+    },
+    {
+      "subset_type": "accent",
+      "subset": "surpeninsular",
+      "cer": 0.04941300157701069,
+      "wer": 0.19017763845350052
+    },
+    {
+      "subset_type": "accent",
+      "subset": "centrosurpeninsular",
+      "cer": 0.05116502522219553,
+      "wer": 0.15604498594189317
+    },
+    {
+      "subset_type": "accent",
+      "subset": "chileno",
+      "cer": 0.06818972515365881,
+      "wer": 0.23098012337217272
+    },
+    {
+      "subset_type": "accent",
+      "subset": "caribe",
+      "cer": 0.050677175843694494,
+      "wer": 0.17474106247911794
+    }
+  ],
+  "test": [
+    {
+      "subset_type": "all",
+      "subset": "all",
+      "cer": 0.05632798062971081,
+      "wer": 0.190560191333188
+    },
+    {
+      "subset_type": "age",
+      "subset": "",
+      "cer": 0.056848431803115676,
+      "wer": 0.19189944134078213
+    },
+    {
+      "subset_type": "age",
+      "subset": "sixties",
+      "cer": 0.052856587640886125,
+      "wer": 0.1695501730103806
+    },
+    {
+      "subset_type": "age",
+      "subset": "seventies",
+      "cer": 0.13721413721413722,
+      "wer": 0.4146341463414634
+    },
+    {
+      "subset_type": "age",
+      "subset": "fourties",
+      "cer": 0.05371106268427125,
+      "wer": 0.17245817245817247
+    },
+    {
+      "subset_type": "age",
+      "subset": "thirties",
+      "cer": 0.04961377349064217,
+      "wer": 0.18068965517241378
+    },
+    {
+      "subset_type": "age",
+      "subset": "fifties",
+      "cer": 0.04829042364800436,
+      "wer": 0.15704481227291078
+    },
+    {
+      "subset_type": "age",
+      "subset": "teens",
+      "cer": 0.06241153872076565,
+      "wer": 0.221001221001221
+    },
+    {
+      "subset_type": "age",
+      "subset": "twenties",
+      "cer": 0.053122181016337575,
+      "wer": 0.18466522678185746
+    },
+    {
+      "subset_type": "age",
+      "subset": "eighties",
+      "cer": 0.06040268456375839,
+      "wer": 0.1111111111111111
+    },
+    {
+      "subset_type": "gender",
+      "subset": "",
+      "cer": 0.056899829777444025,
+      "wer": 0.19214592773324038
+    },
+    {
+      "subset_type": "gender",
+      "subset": "female",
+      "cer": 0.04924963009934475,
+      "wer": 0.17632135306553912
+    },
+    {
+      "subset_type": "gender",
+      "subset": "male",
+      "cer": 0.05362233410318603,
+      "wer": 0.18162368451486063
+    },
+    {
+      "subset_type": "gender",
+      "subset": "other",
+      "cer": 0.08393113342898136,
+      "wer": 0.26521739130434785
+    },
+    {
+      "subset_type": "accent",
+      "subset": "",
+      "cer": 0.057160811775134636,
+      "wer": 0.19271893151799918
+    },
+    {
+      "subset_type": "accent",
+      "subset": "rioplatense",
+      "cer": 0.07058080808080808,
+      "wer": 0.2184936614466816
+    },
+    {
+      "subset_type": "accent",
+      "subset": "mexicano",
+      "cer": 0.051278538812785386,
+      "wer": 0.17655288192501398
+    },
+    {
+      "subset_type": "accent",
+      "subset": "americacentral",
+      "cer": 0.04863022676628654,
+      "wer": 0.16536050156739812
+    },
+    {
+      "subset_type": "accent",
+      "subset": "canario",
+      "cer": 0.050411522633744855,
+      "wer": 0.18471337579617833
+    },
+    {
+      "subset_type": "accent",
+      "subset": "andino",
+      "cer": 0.046894944225539706,
+      "wer": 0.17273638489809556
+    },
+    {
+      "subset_type": "accent",
+      "subset": "nortepeninsular",
+      "cer": 0.0380373157406453,
+      "wer": 0.1408621736490589
+    },
+    {
+      "subset_type": "accent",
+      "subset": "surpeninsular",
+      "cer": 0.04187071498530852,
+      "wer": 0.16444444444444445
+    },
+    {
+      "subset_type": "accent",
+      "subset": "centrosurpeninsular",
+      "cer": 0.033936995891036374,
+      "wer": 0.11964285714285715
+    },
+    {
+      "subset_type": "accent",
+      "subset": "chileno",
+      "cer": 0.043491671807526215,
+      "wer": 0.16102470265324795
+    },
+    {
+      "subset_type": "accent",
+      "subset": "caribe",
+      "cer": 0.058110673090962184,
+      "wer": 0.19846022241231823
+    }
+  ]
+}

cv7_transcriptions.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:187eabc1d862a465ca0ed2d0eeacdc2151b180cc560d908ab8ced29e4314d802
+size 125742975

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:ee024ecfa9443507df082fde9edaa6eb0ae92ed7087e0806650c5ba1b74201fb
+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/tmpnrayz3lw", "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}