move model to techiaith

README.md

@@ -20,118 +20,9 @@ model-index:
       name: Common Voice cy
       type: common_voice
       args: cy
-    metrics:
-       - name: Test WER
-         type: wer
-         value: 25.31
 ---
 
 # Wav2Vec2-Large-XLSR-Welsh
 
-Fine-tuned [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on the [Welsh Common Voice dataset](https://huggingface.co/datasets/common_voice).
+This model has moved to https://huggingface.co/techiaith/wav2vec2-xlsr-ft-cy
 
-When using this model, make sure that your speech input is sampled at 16kHz.
-
-
-## Usage
-
-The model can be used directly (without a language model) as follows:
-
-```python
-import torch
-import torchaudio
-from datasets import load_dataset
-from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
-
-test_dataset = load_dataset("common_voice", "cy", split="test[:2%]")
-
-processor = Wav2Vec2Processor.from_pretrained("DewiBrynJones/wav2vec2-large-xlsr-welsh")
-model = Wav2Vec2ForCTC.from_pretrained("DewiBrynJones/wav2vec2-large-xlsr-welsh")
-
-resampler = torchaudio.transforms.Resample(48_000, 16_000)
-
-# Preprocessing the datasets.
-# We need to read the aduio files as arrays
-def speech_file_to_array_fn(batch):
-  speech_array, sampling_rate = torchaudio.load(batch["path"])
-  batch["speech"] = resampler(speech_array).squeeze().numpy()
-  return batch
-
-test_dataset = test_dataset.map(speech_file_to_array_fn)
-inputs = processor(test_dataset["speech"][:2], sampling_rate=16_000, return_tensors="pt", padding=True)
-
-with torch.no_grad():
-  tlogits = model(inputs.input_values, attention_mask=inputs.attention_mask).logits
-
-predicted_ids = torch.argmax(logits, dim=-1)
-
-print("Prediction:", processor.batch_decode(predicted_ids))
-print("Reference:", test_dataset["sentence"][:2])
-```
-
-
-## Evaluation
-
-The model can be evaluated as follows on the Welsh test data of Common Voice. 
-
-
-```python
-import torch
-import torchaudio
-from datasets import load_dataset, load_metric
-from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
-import re
-
-test_dataset = load_dataset("common_voice", "cy", split="test") 
-
-wer = load_metric("wer")
-
-processor = Wav2Vec2Processor.from_pretrained("DewiBrynJones/wav2vec2-large-xlsr-welsh")
-model = Wav2Vec2ForCTC.from_pretrained("DewiBrynJones/wav2vec2-large-xlsr-welsh")
-
-model.to("cuda")
-
-chars_to_ignore_regex = '[\\\\,\\\\?\\\\.\\\\!\\\\-\\\\u2013\\\\u2014\\\\;\\\\:\\\\"\\\\\\\\%\\\\\\\\\\\\]' 
-
-resampler = torchaudio.transforms.Resample(48_000, 16_000)
-
-# Preprocessing the datasets.
-# We need to read the aduio files as arrays
-def speech_file_to_array_fn(batch):
-  batch["sentence"] = re.sub(chars_to_ignore_regex, '', batch["sentence"]).lower()
-  speech_array, sampling_rate = torchaudio.load(batch["path"])
-  batch["speech"] = resampler(speech_array).squeeze().numpy()
-  return batch
-
-test_dataset = test_dataset.map(speech_file_to_array_fn)
-
-# Preprocessing the datasets.
-# We need to read the aduio files as arrays
-def evaluate(batch):
-  inputs = processor(batch["speech"], sampling_rate=16_000, return_tensors="pt", padding=True)
-
-  with torch.no_grad():
-    logits = model(inputs.input_values.to("cuda"), attention_mask=inputs.attention_mask.to("cuda")).logits
-
-  pred_ids = torch.argmax(logits, dim=-1)
-  batch["pred_strings"] = processor.batch_decode(pred_ids)
-  return batch
-
-result = test_dataset.map(evaluate, batched=True, batch_size=8)
-
-print("WER: {:2f}".format(100 * wer.compute(predictions=result["pred_strings"], references=result["sentence"])))
-```
-
-**Test Result**: 25.31% 
-
-
-# Training 
-
-A Docker based setup for training and evaluating this model can be found at GitHub: https://github.com/techiaith/xlsr-fine-tuning-week
-
-# Example Predictions
-
-| Prediction  | Reference  |
-|---|---|
-| rhedais i ffwrdd heb ddweud dim wrthi ym beth digwyddodd | Rhedais i ffwrdd heb ddweud dim wrthi am beth ddigwyddodd. |
-| ac yr oedd y ferch yn ofnus d | Ac yr oedd y ferch yn ofnus. | 

config.json

@@ -1,76 +0,0 @@
-{
-  "_name_or_path": "facebook/wav2vec2-large-xlsr-53",
-  "activation_dropout": 0.055,
-  "apply_spec_augment": true,
-  "architectures": [
-    "Wav2Vec2ForCTC"
-  ],
-  "attention_dropout": 0.094,
-  "bos_token_id": 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,
-  "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": 0.04,
-  "final_dropout": 0.0,
-  "gradient_checkpointing": true,
-  "hidden_act": "gelu",
-  "hidden_dropout": 0.047,
-  "hidden_size": 1024,
-  "initializer_range": 0.02,
-  "intermediate_size": 4096,
-  "layer_norm_eps": 1e-05,
-  "layerdrop": 0.041,
-  "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_prob": 0.0,
-  "mask_time_length": 10,
-  "mask_time_min_space": 1,
-  "mask_time_other": 0.0,
-  "mask_time_prob": 0.082,
-  "mask_time_selection": "static",
-  "model_type": "wav2vec2",
-  "num_attention_heads": 16,
-  "num_conv_pos_embedding_groups": 16,
-  "num_conv_pos_embeddings": 128,
-  "num_feat_extract_layers": 7,
-  "num_hidden_layers": 24,
-  "pad_token_id": 48,
-  "transformers_version": "4.4.2",
-  "vocab_size": 49
-}

preprocessor_config.json

@@ -1,8 +0,0 @@
-{
-  "do_normalize": true,
-  "feature_size": 1,
-  "padding_side": "right",
-  "padding_value": 0.0,
-  "return_attention_mask": true,
-  "sampling_rate": 16000
-}

pytorch_model.bin

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:0b3d1d4c5339a82fac76d5a361323ba7c8e8faa278c8f37e1e4db9929ecd4705
-size 1262129841

special_tokens_map.json

@@ -1 +0,0 @@
-{"bos_token": "<s>", "eos_token": "</s>", "unk_token": "[UNK]", "pad_token": "[PAD]"}

tokenizer_config.json

@@ -1 +0,0 @@
-{"unk_token": "[UNK]", "bos_token": "<s>", "eos_token": "</s>", "pad_token": "[PAD]", "do_lower_case": false, "word_delimiter_token": "|"}

vocab.json

@@ -1 +0,0 @@
-{"g": 1, "‘": 2, "e": 3, "r": 4, "b": 5, "ê": 6, "â": 7, "ŷ": 8, "'": 9, "v": 10, "ö": 11, "ï": 12, "h": 13, "z": 14, "m": 15, "á": 16, "ä": 17, "ñ": 18, "o": 19, "n": 20, "j": 21, "ò": 22, "ë": 23, "î": 24, "k": 25, "q": 26, "x": 27, "¬": 28, "a": 29, "s": 30, "i": 31, "ÿ": 32, "u": 33, "ŵ": 34, "c": 35, "é": 36, "w": 37, "p": 38, "y": 39, "d": 40, "l": 41, "à": 42, "û": 43, "t": 44, "f": 45, "ô": 46, "|": 0, "[UNK]": 47, "[PAD]": 48}