jonatasgrosman/exp_w2v2r_en_vp-100k_age_teens-10_sixties-0_s232

.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:
+- en
+license: apache-2.0
+tags:
+- automatic-speech-recognition
+- en
+datasets:
+- mozilla-foundation/common_voice_7_0
+---
+# exp_w2v2r_en_vp-100k_age_teens-10_sixties-0_s232
+
+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 (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 +1,116 @@
+{
+  "_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": 33,
+  "xvector_output_dim": 512
+}

cv7_evaluation.json

@@ -0,0 +1,390 @@
+{
+  "validation": [
+    {
+      "subset_type": "all",
+      "subset": "all",
+      "cer": 0.15690520847432501,
+      "wer": 0.42916979244811204
+    },
+    {
+      "subset_type": "age",
+      "subset": "",
+      "cer": 0.15553069730389169,
+      "wer": 0.4261148636041348
+    },
+    {
+      "subset_type": "age",
+      "subset": "thirties",
+      "cer": 0.1587861235283007,
+      "wer": 0.4339243282763663
+    },
+    {
+      "subset_type": "age",
+      "subset": "sixties",
+      "cer": 0.13030148185998977,
+      "wer": 0.3856143856143856
+    },
+    {
+      "subset_type": "age",
+      "subset": "seventies",
+      "cer": 0.14376053962900506,
+      "wer": 0.4090909090909091
+    },
+    {
+      "subset_type": "age",
+      "subset": "fifties",
+      "cer": 0.13370210469508906,
+      "wer": 0.38893280632411065
+    },
+    {
+      "subset_type": "age",
+      "subset": "twenties",
+      "cer": 0.1783295379416734,
+      "wer": 0.4683979705907645
+    },
+    {
+      "subset_type": "age",
+      "subset": "eighties",
+      "cer": 0.17204301075268819,
+      "wer": 0.46153846153846156
+    },
+    {
+      "subset_type": "age",
+      "subset": "nineties",
+      "cer": 0.11538461538461539,
+      "wer": 0.42857142857142855
+    },
+    {
+      "subset_type": "age",
+      "subset": "teens",
+      "cer": 0.15630703711214763,
+      "wer": 0.4403292181069959
+    },
+    {
+      "subset_type": "age",
+      "subset": "fourties",
+      "cer": 0.15195460277427492,
+      "wer": 0.4256410256410256
+    },
+    {
+      "subset_type": "gender",
+      "subset": "",
+      "cer": 0.155631526284732,
+      "wer": 0.42630365179231133
+    },
+    {
+      "subset_type": "gender",
+      "subset": "male",
+      "cer": 0.16476722412345837,
+      "wer": 0.44697860333881967
+    },
+    {
+      "subset_type": "gender",
+      "subset": "other",
+      "cer": 0.1513353115727003,
+      "wer": 0.41916167664670656
+    },
+    {
+      "subset_type": "gender",
+      "subset": "female",
+      "cer": 0.1580870795347375,
+      "wer": 0.43115279048490396
+    },
+    {
+      "subset_type": "accent",
+      "subset": "",
+      "cer": 0.15779193113956225,
+      "wer": 0.43014049604132976
+    },
+    {
+      "subset_type": "accent",
+      "subset": "england",
+      "cer": 0.15012023359670215,
+      "wer": 0.4285004965243297
+    },
+    {
+      "subset_type": "accent",
+      "subset": "australia",
+      "cer": 0.11499436302142052,
+      "wer": 0.3733031674208145
+    },
+    {
+      "subset_type": "accent",
+      "subset": "malaysia",
+      "cer": 0.2076923076923077,
+      "wer": 0.4722222222222222
+    },
+    {
+      "subset_type": "accent",
+      "subset": "indian",
+      "cer": 0.18623919308357348,
+      "wer": 0.48707155742633795
+    },
+    {
+      "subset_type": "accent",
+      "subset": "wales",
+      "cer": 0.10734463276836158,
+      "wer": 0.3819444444444444
+    },
+    {
+      "subset_type": "accent",
+      "subset": "philippines",
+      "cer": 0.19452706722189173,
+      "wer": 0.5190311418685121
+    },
+    {
+      "subset_type": "accent",
+      "subset": "hongkong",
+      "cer": 0.16390423572744015,
+      "wer": 0.4782608695652174
+    },
+    {
+      "subset_type": "accent",
+      "subset": "bermuda",
+      "cer": 0.07407407407407407,
+      "wer": 0.1875
+    },
+    {
+      "subset_type": "accent",
+      "subset": "ireland",
+      "cer": 0.18642951251646903,
+      "wer": 0.4763779527559055
+    },
+    {
+      "subset_type": "accent",
+      "subset": "newzealand",
+      "cer": 0.06896551724137931,
+      "wer": 0.275
+    },
+    {
+      "subset_type": "accent",
+      "subset": "other",
+      "cer": 0.1691449814126394,
+      "wer": 0.54
+    },
+    {
+      "subset_type": "accent",
+      "subset": "us",
+      "cer": 0.1360331694147141,
+      "wer": 0.3934370771312585
+    },
+    {
+      "subset_type": "accent",
+      "subset": "african",
+      "cer": 0.1800878477306003,
+      "wer": 0.46647230320699706
+    },
+    {
+      "subset_type": "accent",
+      "subset": "scotland",
+      "cer": 0.10676156583629894,
+      "wer": 0.32954545454545453
+    },
+    {
+      "subset_type": "accent",
+      "subset": "singapore",
+      "cer": 0.20594965675057209,
+      "wer": 0.43506493506493504
+    },
+    {
+      "subset_type": "accent",
+      "subset": "southatlandtic",
+      "cer": 0.16483516483516483,
+      "wer": 0.5757575757575758
+    },
+    {
+      "subset_type": "accent",
+      "subset": "canada",
+      "cer": 0.12281059063136457,
+      "wer": 0.38369304556354916
+    }
+  ],
+  "test": [
+    {
+      "subset_type": "all",
+      "subset": "all",
+      "cer": 0.1960324343832403,
+      "wer": 0.48523507487085593
+    },
+    {
+      "subset_type": "age",
+      "subset": "",
+      "cer": 0.19513983265427481,
+      "wer": 0.4835426476293349
+    },
+    {
+      "subset_type": "age",
+      "subset": "thirties",
+      "cer": 0.18660184149384734,
+      "wer": 0.47609756097560973
+    },
+    {
+      "subset_type": "age",
+      "subset": "seventies",
+      "cer": 0.162748643761302,
+      "wer": 0.4161073825503356
+    },
+    {
+      "subset_type": "age",
+      "subset": "fifties",
+      "cer": 0.24879393521709167,
+      "wer": 0.5578231292517006
+    },
+    {
+      "subset_type": "age",
+      "subset": "twenties",
+      "cer": 0.20102615028136378,
+      "wer": 0.5001428979708488
+    },
+    {
+      "subset_type": "age",
+      "subset": "eighties",
+      "cer": 0.13142857142857142,
+      "wer": 0.44
+    },
+    {
+      "subset_type": "age",
+      "subset": "fourties",
+      "cer": 0.19296762741949117,
+      "wer": 0.47274436090225563
+    },
+    {
+      "subset_type": "age",
+      "subset": "teens",
+      "cer": 0.2143495620046481,
+      "wer": 0.5187202178352621
+    },
+    {
+      "subset_type": "age",
+      "subset": "sixties",
+      "cer": 0.22173215717722533,
+      "wer": 0.43832599118942733
+    },
+    {
+      "subset_type": "gender",
+      "subset": "",
+      "cer": 0.1950776003267668,
+      "wer": 0.48345030529301286
+    },
+    {
+      "subset_type": "gender",
+      "subset": "male",
+      "cer": 0.20361625144186687,
+      "wer": 0.498960498960499
+    },
+    {
+      "subset_type": "gender",
+      "subset": "other",
+      "cer": 0.22453222453222454,
+      "wer": 0.48249027237354086
+    },
+    {
+      "subset_type": "gender",
+      "subset": "female",
+      "cer": 0.19046625638885992,
+      "wer": 0.4826048171275647
+    },
+    {
+      "subset_type": "accent",
+      "subset": "",
+      "cer": 0.19672541140770627,
+      "wer": 0.48648395779655407
+    },
+    {
+      "subset_type": "accent",
+      "subset": "england",
+      "cer": 0.1707678624369997,
+      "wer": 0.44771057094403616
+    },
+    {
+      "subset_type": "accent",
+      "subset": "malaysia",
+      "cer": 0.15677491601343785,
+      "wer": 0.4267515923566879
+    },
+    {
+      "subset_type": "accent",
+      "subset": "australia",
+      "cer": 0.15578158458244112,
+      "wer": 0.4140127388535032
+    },
+    {
+      "subset_type": "accent",
+      "subset": "indian",
+      "cer": 0.21252747252747253,
+      "wer": 0.5098963242224317
+    },
+    {
+      "subset_type": "accent",
+      "subset": "wales",
+      "cer": 0.1917808219178082,
+      "wer": 0.4166666666666667
+    },
+    {
+      "subset_type": "accent",
+      "subset": "philippines",
+      "cer": 0.14046822742474915,
+      "wer": 0.36363636363636365
+    },
+    {
+      "subset_type": "accent",
+      "subset": "hongkong",
+      "cer": 0.24855491329479767,
+      "wer": 0.5583333333333333
+    },
+    {
+      "subset_type": "accent",
+      "subset": "bermuda",
+      "cer": 0.24044943820224718,
+      "wer": 0.6071428571428571
+    },
+    {
+      "subset_type": "accent",
+      "subset": "ireland",
+      "cer": 0.17053701015965167,
+      "wer": 0.4177215189873418
+    },
+    {
+      "subset_type": "accent",
+      "subset": "newzealand",
+      "cer": 0.1280323450134771,
+      "wer": 0.375
+    },
+    {
+      "subset_type": "accent",
+      "subset": "other",
+      "cer": 0.250390015600624,
+      "wer": 0.5482625482625483
+    },
+    {
+      "subset_type": "accent",
+      "subset": "us",
+      "cer": 0.1815231501514496,
+      "wer": 0.46523001681186
+    },
+    {
+      "subset_type": "accent",
+      "subset": "african",
+      "cer": 0.2226680040120361,
+      "wer": 0.49710982658959535
+    },
+    {
+      "subset_type": "accent",
+      "subset": "scotland",
+      "cer": 0.1753155680224404,
+      "wer": 0.46099290780141844
+    },
+    {
+      "subset_type": "accent",
+      "subset": "singapore",
+      "cer": 0.24064171122994651,
+      "wer": 0.5151515151515151
+    },
+    {
+      "subset_type": "accent",
+      "subset": "canada",
+      "cer": 0.2048162859980139,
+      "wer": 0.46629213483146065
+    }
+  ]
+}

cv7_transcriptions.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5c50f27406bd9a7b156c25ab6795a14518527bc800a009103db3704e42ae05fa
+size 120843077

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:f03a48c4a49ac93c46c0eaac64765ca5fc6b294ba14b6bae6413054f31edcbb2
+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/tmplzmdoo04", "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}