jonatasgrosman/exp_w2v2r_en_vp-100k_age_teens-5_sixties-5_s682

.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-5_sixties-5_s682
+
+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": 0.15703948626391923,
+      "wer": 0.42214303575893974
+    },
+    {
+      "subset_type": "age",
+      "subset": "",
+      "cer": 0.15529465794176242,
+      "wer": 0.41886032018509534
+    },
+    {
+      "subset_type": "age",
+      "subset": "thirties",
+      "cer": 0.16420071774853617,
+      "wer": 0.4373971851581064
+    },
+    {
+      "subset_type": "age",
+      "subset": "sixties",
+      "cer": 0.12161471640265713,
+      "wer": 0.34865134865134867
+    },
+    {
+      "subset_type": "age",
+      "subset": "seventies",
+      "cer": 0.13490725126475547,
+      "wer": 0.3827751196172249
+    },
+    {
+      "subset_type": "age",
+      "subset": "fifties",
+      "cer": 0.13248785752833242,
+      "wer": 0.3778656126482213
+    },
+    {
+      "subset_type": "age",
+      "subset": "twenties",
+      "cer": 0.18142951590391537,
+      "wer": 0.4648723020036117
+    },
+    {
+      "subset_type": "age",
+      "subset": "eighties",
+      "cer": 0.1971326164874552,
+      "wer": 0.4807692307692308
+    },
+    {
+      "subset_type": "age",
+      "subset": "nineties",
+      "cer": 0.11538461538461539,
+      "wer": 0.2857142857142857
+    },
+    {
+      "subset_type": "age",
+      "subset": "teens",
+      "cer": 0.1583350233218414,
+      "wer": 0.43209876543209874
+    },
+    {
+      "subset_type": "age",
+      "subset": "fourties",
+      "cer": 0.1494325346784363,
+      "wer": 0.41098901098901097
+    },
+    {
+      "subset_type": "gender",
+      "subset": "",
+      "cer": 0.15538779280088258,
+      "wer": 0.4190791781414892
+    },
+    {
+      "subset_type": "gender",
+      "subset": "male",
+      "cer": 0.16676890295086202,
+      "wer": 0.439830707735716
+    },
+    {
+      "subset_type": "gender",
+      "subset": "other",
+      "cer": 0.15380811078140455,
+      "wer": 0.437125748502994
+    },
+    {
+      "subset_type": "gender",
+      "subset": "female",
+      "cer": 0.16052342030807923,
+      "wer": 0.4288655077767612
+    },
+    {
+      "subset_type": "accent",
+      "subset": "",
+      "cer": 0.15787015975486762,
+      "wer": 0.42371421971144757
+    },
+    {
+      "subset_type": "accent",
+      "subset": "england",
+      "cer": 0.1464273445551357,
+      "wer": 0.4056603773584906
+    },
+    {
+      "subset_type": "accent",
+      "subset": "australia",
+      "cer": 0.10372040586245772,
+      "wer": 0.334841628959276
+    },
+    {
+      "subset_type": "accent",
+      "subset": "malaysia",
+      "cer": 0.18974358974358974,
+      "wer": 0.4444444444444444
+    },
+    {
+      "subset_type": "accent",
+      "subset": "indian",
+      "cer": 0.19997941539728284,
+      "wer": 0.4813589897775105
+    },
+    {
+      "subset_type": "accent",
+      "subset": "wales",
+      "cer": 0.09717514124293786,
+      "wer": 0.3611111111111111
+    },
+    {
+      "subset_type": "accent",
+      "subset": "philippines",
+      "cer": 0.2183224271267103,
+      "wer": 0.5536332179930796
+    },
+    {
+      "subset_type": "accent",
+      "subset": "hongkong",
+      "cer": 0.1694290976058932,
+      "wer": 0.4782608695652174
+    },
+    {
+      "subset_type": "accent",
+      "subset": "bermuda",
+      "cer": 0.07407407407407407,
+      "wer": 0.1875
+    },
+    {
+      "subset_type": "accent",
+      "subset": "ireland",
+      "cer": 0.1785243741765481,
+      "wer": 0.484251968503937
+    },
+    {
+      "subset_type": "accent",
+      "subset": "newzealand",
+      "cer": 0.08866995073891626,
+      "wer": 0.3
+    },
+    {
+      "subset_type": "accent",
+      "subset": "other",
+      "cer": 0.18029739776951673,
+      "wer": 0.53
+    },
+    {
+      "subset_type": "accent",
+      "subset": "us",
+      "cer": 0.13333333333333333,
+      "wer": 0.38249887235002256
+    },
+    {
+      "subset_type": "accent",
+      "subset": "african",
+      "cer": 0.17325524646168863,
+      "wer": 0.43440233236151604
+    },
+    {
+      "subset_type": "accent",
+      "subset": "scotland",
+      "cer": 0.10142348754448399,
+      "wer": 0.3181818181818182
+    },
+    {
+      "subset_type": "accent",
+      "subset": "singapore",
+      "cer": 0.21853546910755148,
+      "wer": 0.44155844155844154
+    },
+    {
+      "subset_type": "accent",
+      "subset": "southatlandtic",
+      "cer": 0.18681318681318682,
+      "wer": 0.5454545454545454
+    },
+    {
+      "subset_type": "accent",
+      "subset": "canada",
+      "cer": 0.11934826883910386,
+      "wer": 0.35731414868105515
+    }
+  ],
+  "test": [
+    {
+      "subset_type": "all",
+      "subset": "all",
+      "cer": 0.19921533473424413,
+      "wer": 0.48352187275223957
+    },
+    {
+      "subset_type": "age",
+      "subset": "",
+      "cer": 0.19812397920601066,
+      "wer": 0.4812225679098744
+    },
+    {
+      "subset_type": "age",
+      "subset": "thirties",
+      "cer": 0.18832286378108595,
+      "wer": 0.4748780487804878
+    },
+    {
+      "subset_type": "age",
+      "subset": "seventies",
+      "cer": 0.1525015069318867,
+      "wer": 0.38926174496644295
+    },
+    {
+      "subset_type": "age",
+      "subset": "fifties",
+      "cer": 0.25068917987594763,
+      "wer": 0.5461613216715258
+    },
+    {
+      "subset_type": "age",
+      "subset": "twenties",
+      "cer": 0.20845746441575638,
+      "wer": 0.5052872249214061
+    },
+    {
+      "subset_type": "age",
+      "subset": "eighties",
+      "cer": 0.13142857142857142,
+      "wer": 0.48
+    },
+    {
+      "subset_type": "age",
+      "subset": "fourties",
+      "cer": 0.1908545346969825,
+      "wer": 0.46099624060150374
+    },
+    {
+      "subset_type": "age",
+      "subset": "teens",
+      "cer": 0.2198915440081044,
+      "wer": 0.5279101429543908
+    },
+    {
+      "subset_type": "age",
+      "subset": "sixties",
+      "cer": 0.2157177225340818,
+      "wer": 0.46255506607929514
+    },
+    {
+      "subset_type": "gender",
+      "subset": "",
+      "cer": 0.19803858507810243,
+      "wer": 0.48116916198398635
+    },
+    {
+      "subset_type": "gender",
+      "subset": "male",
+      "cer": 0.2082006487168617,
+      "wer": 0.5012080687756364
+    },
+    {
+      "subset_type": "gender",
+      "subset": "other",
+      "cer": 0.2376992376992377,
+      "wer": 0.5330739299610895
+    },
+    {
+      "subset_type": "gender",
+      "subset": "female",
+      "cer": 0.19401272556587046,
+      "wer": 0.47814451382694023
+    },
+    {
+      "subset_type": "accent",
+      "subset": "",
+      "cer": 0.19991766074431097,
+      "wer": 0.48467750061174847
+    },
+    {
+      "subset_type": "accent",
+      "subset": "england",
+      "cer": 0.17333728629311196,
+      "wer": 0.4505370265686829
+    },
+    {
+      "subset_type": "accent",
+      "subset": "malaysia",
+      "cer": 0.1534154535274356,
+      "wer": 0.40764331210191085
+    },
+    {
+      "subset_type": "accent",
+      "subset": "australia",
+      "cer": 0.16059957173447537,
+      "wer": 0.410828025477707
+    },
+    {
+      "subset_type": "accent",
+      "subset": "indian",
+      "cer": 0.22104395604395605,
+      "wer": 0.5227772541627396
+    },
+    {
+      "subset_type": "accent",
+      "subset": "wales",
+      "cer": 0.13013698630136986,
+      "wer": 0.4166666666666667
+    },
+    {
+      "subset_type": "accent",
+      "subset": "philippines",
+      "cer": 0.13879598662207357,
+      "wer": 0.35353535353535354
+    },
+    {
+      "subset_type": "accent",
+      "subset": "hongkong",
+      "cer": 0.3020231213872832,
+      "wer": 0.6583333333333333
+    },
+    {
+      "subset_type": "accent",
+      "subset": "bermuda",
+      "cer": 0.2157303370786517,
+      "wer": 0.5119047619047619
+    },
+    {
+      "subset_type": "accent",
+      "subset": "ireland",
+      "cer": 0.1806966618287373,
+      "wer": 0.4641350210970464
+    },
+    {
+      "subset_type": "accent",
+      "subset": "newzealand",
+      "cer": 0.1495956873315364,
+      "wer": 0.3897058823529412
+    },
+    {
+      "subset_type": "accent",
+      "subset": "other",
+      "cer": 0.26131045241809675,
+      "wer": 0.5675675675675675
+    },
+    {
+      "subset_type": "accent",
+      "subset": "us",
+      "cer": 0.17957594115101688,
+      "wer": 0.4537673849915941
+    },
+    {
+      "subset_type": "accent",
+      "subset": "african",
+      "cer": 0.2296890672016048,
+      "wer": 0.5260115606936416
+    },
+    {
+      "subset_type": "accent",
+      "subset": "scotland",
+      "cer": 0.20196353436185133,
+      "wer": 0.48936170212765956
+    },
+    {
+      "subset_type": "accent",
+      "subset": "singapore",
+      "cer": 0.26737967914438504,
+      "wer": 0.6363636363636364
+    },
+    {
+      "subset_type": "accent",
+      "subset": "canada",
+      "cer": 0.21449851042701093,
+      "wer": 0.449438202247191
+    }
+  ]
+}

cv7_transcriptions.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:84a0aef2320e647201529f9c32e07354f6ae0a7e3aea5a8bcb11f771c7379ed7
+size 121552852

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 sha256:2e7b70635bdbcdb839c71a6316ce650b88d41c5878dd744c61f7488863936edc
+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/tmp0aj1p3qh", "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}