Added w2v2 with LM

README.md

@@ -1,3 +1,73 @@
----
-license: apache-2.0
----
+---
+library_name: transformers
+license: apache-2.0
+base_model: facebook/wav2vec2-xls-r-300m
+tags:
+- automatic-speech-recognition
+- bookbot/common_voice_16_1_sw
+- bookbot/ALFFA_swahili
+- bookbot/fleurs_sw
+- generated_from_trainer
+metrics:
+- wer
+model-index:
+- name: wav2vec2-xls-r-300m-swahili-cv-fleurs-alffa-word
+  results: []
+---
+
+<!-- This model card has been generated automatically according to the information the Trainer had access to. You
+should probably proofread and complete it, then remove this comment. -->
+
+# wav2vec2-xls-r-300m-swahili-cv-fleurs-alffa-word
+
+This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.2057
+- Wer: 0.2194
+- Cer: 0.1098
+
+## Model description
+
+More information needed
+
+## Intended uses & limitations
+
+More information needed
+
+## Training and evaluation data
+
+More information needed
+
+## Training procedure
+
+### Training hyperparameters
+
+The following hyperparameters were used during training:
+- learning_rate: 0.0002
+- train_batch_size: 16
+- eval_batch_size: 8
+- seed: 42
+- gradient_accumulation_steps: 2
+- total_train_batch_size: 32
+- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- lr_scheduler_type: linear
+- lr_scheduler_warmup_ratio: 0.1
+- num_epochs: 5.0
+
+### Training results
+
+| Training Loss | Epoch | Step | Validation Loss | Wer    | Cer    |
+|:-------------:|:-----:|:----:|:---------------:|:------:|:------:|
+| 0.3702        | 1.0   | 1961 | 0.2878          | 0.3335 | 0.1367 |
+| 0.2333        | 2.0   | 3922 | 0.2324          | 0.2653 | 0.1219 |
+| 0.172         | 3.0   | 5883 | 0.2136          | 0.2464 | 0.1162 |
+| 0.1331        | 4.0   | 7844 | 0.2043          | 0.2287 | 0.1127 |
+| 0.1018        | 5.0   | 9805 | 0.2057          | 0.2194 | 0.1098 |
+
+
+### Framework versions
+
+- Transformers 4.45.2
+- Pytorch 2.3.1
+- Datasets 2.19.2
+- Tokenizers 0.20.1

added_tokens.json

@@ -0,0 +1,4 @@
+{
+  "</s>": 31,
+  "<s>": 30
+}

all_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 5.0,
+    "eval_cer": 0.10983672160892569,
+    "eval_loss": 0.2057330459356308,
+    "eval_runtime": 242.3069,
+    "eval_samples": 22360,
+    "eval_samples_per_second": 92.28,
+    "eval_steps_per_second": 11.535,
+    "eval_wer": 0.21943229191706215,
+    "total_flos": 5.6911698946882765e+19,
+    "train_loss": 0.39030011520891517,
+    "train_runtime": 4855.0155,
+    "train_samples": 62739,
+    "train_samples_per_second": 64.613,
+    "train_steps_per_second": 2.02
+}

alphabet.json

@@ -0,0 +1 @@
+{"labels": [" ", "'", "a", "b", "c", "d", "e", "f", "g", "h", "i", "j", "k", "l", "m", "n", "o", "p", "q", "r", "s", "t", "u", "v", "w", "x", "y", "z", "\u2047", "", "<s>", "</s>"], "is_bpe": false}

config.json

@@ -0,0 +1,108 @@
+{
+  "_name_or_path": "facebook/wav2vec2-xls-r-300m",
+  "activation_dropout": 0.0,
+  "adapter_attn_dim": null,
+  "adapter_kernel_size": 3,
+  "adapter_stride": 2,
+  "add_adapter": false,
+  "apply_spec_augment": true,
+  "architectures": [
+    "Wav2Vec2ForCTC"
+  ],
+  "attention_dropout": 0.0,
+  "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": 0.0,
+  "feat_quantizer_dropout": 0.0,
+  "final_dropout": 0.0,
+  "hidden_act": "gelu",
+  "hidden_dropout": 0.0,
+  "hidden_size": 1024,
+  "initializer_range": 0.02,
+  "intermediate_size": 4096,
+  "layer_norm_eps": 1e-05,
+  "layerdrop": 0.0,
+  "mask_feature_length": 10,
+  "mask_feature_min_masks": 0,
+  "mask_feature_prob": 0.05,
+  "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": 29,
+  "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": "4.45.2",
+  "use_weighted_layer_sum": false,
+  "vocab_size": 32,
+  "xvector_output_dim": 512
+}

eval_results.json

@@ -0,0 +1,10 @@
+{
+    "epoch": 5.0,
+    "eval_cer": 0.10983672160892569,
+    "eval_loss": 0.2057330459356308,
+    "eval_runtime": 242.3069,
+    "eval_samples": 22360,
+    "eval_samples_per_second": 92.28,
+    "eval_steps_per_second": 11.535,
+    "eval_wer": 0.21943229191706215
+}

language_model/5gram.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a5dad64919b8449ffb73db87605df3869e42d8b9ab07e68f73bbcc4cfd7eb2f4
+size 37825846

language_model/attrs.json

@@ -0,0 +1 @@
+{"alpha": 0.5, "beta": 1.5, "unk_score_offset": -10.0, "score_boundary": true}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b7952ec24b484c6ca7edf84c99ee298ffd531ea56e4f6bd89e379ff255a05622
+size 1261938680

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": "Wav2Vec2ProcessorWithLM",
+  "return_attention_mask": true,
+  "sampling_rate": 16000
+}

runs/Oct24_14-46-01_s44504-focus-slate/events.out.tfevents.1729742666.s44504-focus-slate.2113579.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:050f6bed9982481627ed120e82dbd027a25b2dc48966b16b084ea1809539991f
+size 12710

runs/Oct24_14-46-01_s44504-focus-slate/events.out.tfevents.1729747772.s44504-focus-slate.2113579.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3f4101de0fa00b8543099fd3caab4c2acac9b7a4df22d5e07f70d481a98b1e2c
+size 453

special_tokens_map.json

@@ -0,0 +1,30 @@
+{
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "[PAD]",
+    "lstrip": true,
+    "normalized": false,
+    "rstrip": true,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "[UNK]",
+    "lstrip": true,
+    "normalized": false,
+    "rstrip": true,
+    "single_word": false
+  }
+}

tokenizer_config.json

@@ -0,0 +1,48 @@
+{
+  "added_tokens_decoder": {
+    "28": {
+      "content": "[UNK]",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": false
+    },
+    "29": {
+      "content": "[PAD]",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": false
+    },
+    "30": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "31": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": false,
+  "do_lower_case": false,
+  "eos_token": "</s>",
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "[PAD]",
+  "processor_class": "Wav2Vec2ProcessorWithLM",
+  "replace_word_delimiter_char": " ",
+  "target_lang": null,
+  "tokenizer_class": "Wav2Vec2CTCTokenizer",
+  "unk_token": "[UNK]",
+  "word_delimiter_token": "|"
+}

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 5.0,
+    "total_flos": 5.6911698946882765e+19,
+    "train_loss": 0.39030011520891517,
+    "train_runtime": 4855.0155,
+    "train_samples": 62739,
+    "train_samples_per_second": 64.613,
+    "train_steps_per_second": 2.02
+}

trainer_state.json

@@ -0,0 +1,225 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 5.0,
+  "eval_steps": 500,
+  "global_step": 9805,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.25497195308516063,
+      "grad_norm": 1.5315192937850952,
+      "learning_rate": 0.00010193679918450561,
+      "loss": 3.8945,
+      "step": 500
+    },
+    {
+      "epoch": 0.5099439061703213,
+      "grad_norm": 1.6545052528381348,
+      "learning_rate": 0.00019956935630099728,
+      "loss": 0.5228,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7649158592554819,
+      "grad_norm": 1.1093895435333252,
+      "learning_rate": 0.0001882366273798731,
+      "loss": 0.3702,
+      "step": 1500
+    },
+    {
+      "epoch": 1.0,
+      "eval_cer": 0.13674156347728234,
+      "eval_loss": 0.287818968296051,
+      "eval_runtime": 242.9432,
+      "eval_samples_per_second": 92.038,
+      "eval_steps_per_second": 11.505,
+      "eval_wer": 0.3335439517042918,
+      "step": 1961
+    },
+    {
+      "epoch": 1.0198878123406425,
+      "grad_norm": 0.6819602847099304,
+      "learning_rate": 0.0001769038984587489,
+      "loss": 0.3114,
+      "step": 2000
+    },
+    {
+      "epoch": 1.2748597654258031,
+      "grad_norm": 0.7735158801078796,
+      "learning_rate": 0.00016557116953762465,
+      "loss": 0.2571,
+      "step": 2500
+    },
+    {
+      "epoch": 1.5298317185109638,
+      "grad_norm": 0.7301546931266785,
+      "learning_rate": 0.00015423844061650046,
+      "loss": 0.2411,
+      "step": 3000
+    },
+    {
+      "epoch": 1.7848036715961244,
+      "grad_norm": 0.5918829441070557,
+      "learning_rate": 0.00014290571169537624,
+      "loss": 0.2333,
+      "step": 3500
+    },
+    {
+      "epoch": 2.0,
+      "eval_cer": 0.12185579510710033,
+      "eval_loss": 0.23235103487968445,
+      "eval_runtime": 242.1687,
+      "eval_samples_per_second": 92.332,
+      "eval_steps_per_second": 11.542,
+      "eval_wer": 0.26528299554267426,
+      "step": 3922
+    },
+    {
+      "epoch": 2.039775624681285,
+      "grad_norm": 0.7791718244552612,
+      "learning_rate": 0.00013157298277425205,
+      "loss": 0.2157,
+      "step": 4000
+    },
+    {
+      "epoch": 2.2947475777664454,
+      "grad_norm": 0.5816757082939148,
+      "learning_rate": 0.00012024025385312785,
+      "loss": 0.1817,
+      "step": 4500
+    },
+    {
+      "epoch": 2.5497195308516063,
+      "grad_norm": 0.42094656825065613,
+      "learning_rate": 0.00010890752493200362,
+      "loss": 0.1803,
+      "step": 5000
+    },
+    {
+      "epoch": 2.804691483936767,
+      "grad_norm": 0.515943706035614,
+      "learning_rate": 9.757479601087942e-05,
+      "loss": 0.172,
+      "step": 5500
+    },
+    {
+      "epoch": 3.0,
+      "eval_cer": 0.11621454523249779,
+      "eval_loss": 0.213576540350914,
+      "eval_runtime": 242.2848,
+      "eval_samples_per_second": 92.288,
+      "eval_steps_per_second": 11.536,
+      "eval_wer": 0.246385720074449,
+      "step": 5883
+    },
+    {
+      "epoch": 3.0596634370219276,
+      "grad_norm": 0.397499680519104,
+      "learning_rate": 8.624206708975522e-05,
+      "loss": 0.1656,
+      "step": 6000
+    },
+    {
+      "epoch": 3.3146353901070884,
+      "grad_norm": 0.5948652625083923,
+      "learning_rate": 7.490933816863101e-05,
+      "loss": 0.1388,
+      "step": 6500
+    },
+    {
+      "epoch": 3.569607343192249,
+      "grad_norm": 0.5423580408096313,
+      "learning_rate": 6.35766092475068e-05,
+      "loss": 0.1383,
+      "step": 7000
+    },
+    {
+      "epoch": 3.8245792962774097,
+      "grad_norm": 1.2003000974655151,
+      "learning_rate": 5.22438803263826e-05,
+      "loss": 0.1331,
+      "step": 7500
+    },
+    {
+      "epoch": 4.0,
+      "eval_cer": 0.11265216708611207,
+      "eval_loss": 0.20429827272891998,
+      "eval_runtime": 242.3631,
+      "eval_samples_per_second": 92.258,
+      "eval_steps_per_second": 11.532,
+      "eval_wer": 0.22865256897179317,
+      "step": 7844
+    },
+    {
+      "epoch": 4.07955124936257,
+      "grad_norm": 0.8045985102653503,
+      "learning_rate": 4.091115140525839e-05,
+      "loss": 0.123,
+      "step": 8000
+    },
+    {
+      "epoch": 4.3345232024477305,
+      "grad_norm": 0.402927964925766,
+      "learning_rate": 2.957842248413418e-05,
+      "loss": 0.1052,
+      "step": 8500
+    },
+    {
+      "epoch": 4.589495155532891,
+      "grad_norm": 0.42731374502182007,
+      "learning_rate": 1.8245693563009974e-05,
+      "loss": 0.1054,
+      "step": 9000
+    },
+    {
+      "epoch": 4.844467108618052,
+      "grad_norm": 0.2811175286769867,
+      "learning_rate": 6.912964641885767e-06,
+      "loss": 0.1018,
+      "step": 9500
+    },
+    {
+      "epoch": 5.0,
+      "eval_cer": 0.10983672160892569,
+      "eval_loss": 0.2057330459356308,
+      "eval_runtime": 270.0068,
+      "eval_samples_per_second": 82.813,
+      "eval_steps_per_second": 10.352,
+      "eval_wer": 0.21943229191706215,
+      "step": 9805
+    },
+    {
+      "epoch": 5.0,
+      "step": 9805,
+      "total_flos": 5.6911698946882765e+19,
+      "train_loss": 0.39030011520891517,
+      "train_runtime": 4855.0155,
+      "train_samples_per_second": 64.613,
+      "train_steps_per_second": 2.02
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 9805,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.6911698946882765e+19,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8477df06fbac59bc0840720a7d0069ac4b2c63dd7a1c33d595f46daf93e07f71
+size 5368

vocab.json

@@ -0,0 +1,32 @@
+{
+  "'": 1,
+  "[PAD]": 29,
+  "[UNK]": 28,
+  "a": 2,
+  "b": 3,
+  "c": 4,
+  "d": 5,
+  "e": 6,
+  "f": 7,
+  "g": 8,
+  "h": 9,
+  "i": 10,
+  "j": 11,
+  "k": 12,
+  "l": 13,
+  "m": 14,
+  "n": 15,
+  "o": 16,
+  "p": 17,
+  "q": 18,
+  "r": 19,
+  "s": 20,
+  "t": 21,
+  "u": 22,
+  "v": 23,
+  "w": 24,
+  "x": 25,
+  "y": 26,
+  "z": 27,
+  "|": 0
+}