jonatasgrosman/exp_w2v2r_en_xls-r_accent_us-2_england-8_s805

.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:
+- en
+license: apache-2.0
+tags:
+- automatic-speech-recognition
+- en
+datasets:
+- mozilla-foundation/common_voice_7_0
+---
+# exp_w2v2r_en_xls-r_accent_us-2_england-8_s805
+
+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 (en)](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": 33,
+  "xvector_output_dim": 512
+}

cv7_evaluation.json

@@ -0,0 +1,390 @@
+{
+  "validation": [
+    {
+      "subset_type": "all",
+      "subset": "all",
+      "cer": 0.1107813248598677,
+      "wer": 0.3078269567391848
+    },
+    {
+      "subset_type": "age",
+      "subset": "",
+      "cer": 0.1092733964396541,
+      "wer": 0.3044519908945031
+    },
+    {
+      "subset_type": "age",
+      "subset": "sixties",
+      "cer": 0.08601601090103901,
+      "wer": 0.24675324675324675
+    },
+    {
+      "subset_type": "age",
+      "subset": "seventies",
+      "cer": 0.09232715008431704,
+      "wer": 0.2631578947368421
+    },
+    {
+      "subset_type": "age",
+      "subset": "fourties",
+      "cer": 0.09747793190416142,
+      "wer": 0.2886446886446886
+    },
+    {
+      "subset_type": "age",
+      "subset": "thirties",
+      "cer": 0.11691745891834036,
+      "wer": 0.31694388594406875
+    },
+    {
+      "subset_type": "age",
+      "subset": "fifties",
+      "cer": 0.08621154883971938,
+      "wer": 0.2758893280632411
+    },
+    {
+      "subset_type": "age",
+      "subset": "teens",
+      "cer": 0.11665990671263435,
+      "wer": 0.32510288065843623
+    },
+    {
+      "subset_type": "age",
+      "subset": "twenties",
+      "cer": 0.1319767868948799,
+      "wer": 0.35213689913148166
+    },
+    {
+      "subset_type": "age",
+      "subset": "nineties",
+      "cer": 0.07692307692307693,
+      "wer": 0.14285714285714285
+    },
+    {
+      "subset_type": "age",
+      "subset": "eighties",
+      "cer": 0.12903225806451613,
+      "wer": 0.34615384615384615
+    },
+    {
+      "subset_type": "gender",
+      "subset": "",
+      "cer": 0.10938502193601354,
+      "wer": 0.3046593377068267
+    },
+    {
+      "subset_type": "gender",
+      "subset": "female",
+      "cer": 0.10924237661112858,
+      "wer": 0.30237877401646845
+    },
+    {
+      "subset_type": "gender",
+      "subset": "male",
+      "cer": 0.11993123264673597,
+      "wer": 0.32875617211380204
+    },
+    {
+      "subset_type": "gender",
+      "subset": "other",
+      "cer": 0.10781404549950543,
+      "wer": 0.31736526946107785
+    },
+    {
+      "subset_type": "accent",
+      "subset": "",
+      "cer": 0.1115335456337601,
+      "wer": 0.3087482761758756
+    },
+    {
+      "subset_type": "accent",
+      "subset": "african",
+      "cer": 0.13567593948267448,
+      "wer": 0.36443148688046645
+    },
+    {
+      "subset_type": "accent",
+      "subset": "bermuda",
+      "cer": 0.06172839506172839,
+      "wer": 0.125
+    },
+    {
+      "subset_type": "accent",
+      "subset": "scotland",
+      "cer": 0.08540925266903915,
+      "wer": 0.2727272727272727
+    },
+    {
+      "subset_type": "accent",
+      "subset": "malaysia",
+      "cer": 0.13076923076923078,
+      "wer": 0.3472222222222222
+    },
+    {
+      "subset_type": "accent",
+      "subset": "wales",
+      "cer": 0.059887005649717516,
+      "wer": 0.2569444444444444
+    },
+    {
+      "subset_type": "accent",
+      "subset": "canada",
+      "cer": 0.07657841140529531,
+      "wer": 0.24700239808153476
+    },
+    {
+      "subset_type": "accent",
+      "subset": "us",
+      "cer": 0.08926815157651143,
+      "wer": 0.2706359945872801
+    },
+    {
+      "subset_type": "accent",
+      "subset": "hongkong",
+      "cer": 0.08839779005524862,
+      "wer": 0.25
+    },
+    {
+      "subset_type": "accent",
+      "subset": "newzealand",
+      "cer": 0.03940886699507389,
+      "wer": 0.125
+    },
+    {
+      "subset_type": "accent",
+      "subset": "ireland",
+      "cer": 0.10474308300395258,
+      "wer": 0.31496062992125984
+    },
+    {
+      "subset_type": "accent",
+      "subset": "southatlandtic",
+      "cer": 0.16483516483516483,
+      "wer": 0.3939393939393939
+    },
+    {
+      "subset_type": "accent",
+      "subset": "other",
+      "cer": 0.10223048327137546,
+      "wer": 0.3
+    },
+    {
+      "subset_type": "accent",
+      "subset": "indian",
+      "cer": 0.15870728694936187,
+      "wer": 0.4016837041491281
+    },
+    {
+      "subset_type": "accent",
+      "subset": "england",
+      "cer": 0.08845757471659224,
+      "wer": 0.27656405163853026
+    },
+    {
+      "subset_type": "accent",
+      "subset": "australia",
+      "cer": 0.061255167230364525,
+      "wer": 0.2081447963800905
+    },
+    {
+      "subset_type": "accent",
+      "subset": "singapore",
+      "cer": 0.16704805491990846,
+      "wer": 0.33116883116883117
+    },
+    {
+      "subset_type": "accent",
+      "subset": "philippines",
+      "cer": 0.17311124330755503,
+      "wer": 0.4463667820069204
+    }
+  ],
+  "test": [
+    {
+      "subset_type": "all",
+      "subset": "all",
+      "cer": 0.14176014156572153,
+      "wer": 0.3520041849212058
+    },
+    {
+      "subset_type": "age",
+      "subset": "",
+      "cer": 0.13998607036231991,
+      "wer": 0.3476544374377192
+    },
+    {
+      "subset_type": "age",
+      "subset": "seventies",
+      "cer": 0.1458710066305003,
+      "wer": 0.3288590604026846
+    },
+    {
+      "subset_type": "age",
+      "subset": "sixties",
+      "cer": 0.15236567762630313,
+      "wer": 0.31057268722466963
+    },
+    {
+      "subset_type": "age",
+      "subset": "fourties",
+      "cer": 0.13397007860704926,
+      "wer": 0.33646616541353386
+    },
+    {
+      "subset_type": "age",
+      "subset": "thirties",
+      "cer": 0.13161517941657344,
+      "wer": 0.35317073170731705
+    },
+    {
+      "subset_type": "age",
+      "subset": "fifties",
+      "cer": 0.20485871812543074,
+      "wer": 0.4013605442176871
+    },
+    {
+      "subset_type": "age",
+      "subset": "teens",
+      "cer": 0.1585721947440558,
+      "wer": 0.4033356024506467
+    },
+    {
+      "subset_type": "age",
+      "subset": "twenties",
+      "cer": 0.1576133730552797,
+      "wer": 0.3923978279508431
+    },
+    {
+      "subset_type": "age",
+      "subset": "eighties",
+      "cer": 0.11428571428571428,
+      "wer": 0.36
+    },
+    {
+      "subset_type": "gender",
+      "subset": "",
+      "cer": 0.13992349206219967,
+      "wer": 0.347631032673576
+    },
+    {
+      "subset_type": "gender",
+      "subset": "female",
+      "cer": 0.14404923333681027,
+      "wer": 0.36009515313708
+    },
+    {
+      "subset_type": "gender",
+      "subset": "male",
+      "cer": 0.15415405546627756,
+      "wer": 0.3819183008372198
+    },
+    {
+      "subset_type": "gender",
+      "subset": "other",
+      "cer": 0.1781011781011781,
+      "wer": 0.41245136186770426
+    },
+    {
+      "subset_type": "accent",
+      "subset": "",
+      "cer": 0.141889006170314,
+      "wer": 0.35192808717055546
+    },
+    {
+      "subset_type": "accent",
+      "subset": "african",
+      "cer": 0.15546639919759278,
+      "wer": 0.36416184971098264
+    },
+    {
+      "subset_type": "accent",
+      "subset": "bermuda",
+      "cer": 0.16404494382022472,
+      "wer": 0.38095238095238093
+    },
+    {
+      "subset_type": "accent",
+      "subset": "scotland",
+      "cer": 0.1093969144460028,
+      "wer": 0.2907801418439716
+    },
+    {
+      "subset_type": "accent",
+      "subset": "malaysia",
+      "cer": 0.09854423292273236,
+      "wer": 0.2929936305732484
+    },
+    {
+      "subset_type": "accent",
+      "subset": "wales",
+      "cer": 0.1643835616438356,
+      "wer": 0.4166666666666667
+    },
+    {
+      "subset_type": "accent",
+      "subset": "canada",
+      "cer": 0.19016881827209534,
+      "wer": 0.34831460674157305
+    },
+    {
+      "subset_type": "accent",
+      "subset": "us",
+      "cer": 0.12640631761142362,
+      "wer": 0.32523307351367875
+    },
+    {
+      "subset_type": "accent",
+      "subset": "hongkong",
+      "cer": 0.1994219653179191,
+      "wer": 0.48333333333333334
+    },
+    {
+      "subset_type": "accent",
+      "subset": "newzealand",
+      "cer": 0.0646900269541779,
+      "wer": 0.25735294117647056
+    },
+    {
+      "subset_type": "accent",
+      "subset": "ireland",
+      "cer": 0.1204644412191582,
+      "wer": 0.31223628691983124
+    },
+    {
+      "subset_type": "accent",
+      "subset": "other",
+      "cer": 0.17082683307332294,
+      "wer": 0.36293436293436293
+    },
+    {
+      "subset_type": "accent",
+      "subset": "indian",
+      "cer": 0.1753846153846154,
+      "wer": 0.4344957587181904
+    },
+    {
+      "subset_type": "accent",
+      "subset": "england",
+      "cer": 0.12056527324834469,
+      "wer": 0.32730356133408706
+    },
+    {
+      "subset_type": "accent",
+      "subset": "australia",
+      "cer": 0.11027837259100642,
+      "wer": 0.3248407643312102
+    },
+    {
+      "subset_type": "accent",
+      "subset": "singapore",
+      "cer": 0.15508021390374332,
+      "wer": 0.3333333333333333
+    },
+    {
+      "subset_type": "accent",
+      "subset": "philippines",
+      "cer": 0.10033444816053512,
+      "wer": 0.29292929292929293
+    }
+  ]
+}

cv7_transcriptions.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:98e993ce9877fbc034d60aee4efaad903101f8b88a6e826cd8fdb388252b8a0e
+size 122768860

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 sha256:074b2ea7010946fc1d55a11a1b4157652cee8eb5893cf762cdab9bcb974f0133
+size 1262034033

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/tmpggwfd85w", "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}