Finished finetuning 🎉

.gitattributes

@@ -33,3 +33,4 @@ 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
+language_model/unigrams.txt filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,102 @@
+---
+library_name: transformers
+language:
+- da
+license: openrail
+base_model: chcaa/xls-r-300m-danish
+tags:
+- generated_from_trainer
+datasets:
+- generator
+metrics:
+- wer
+model-index:
+- name: roest-315m
+  results:
+  - task:
+      name: Automatic Speech Recognition
+      type: automatic-speech-recognition
+    dataset:
+      name: generator
+      type: generator
+      config: default
+      split: None
+      args: default
+    metrics:
+    - name: Wer
+      type: wer
+      value: 0.34354151223128243
+---
+
+<!-- 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. -->
+
+# roest-315m
+
+This model is a fine-tuned version of [chcaa/xls-r-300m-danish](https://huggingface.co/chcaa/xls-r-300m-danish) on the generator dataset.
+It achieves the following results on the evaluation set:
+- Loss: 316.1826
+- Wer: 0.3435
+- Cer: 0.1319
+
+## 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: 0.0001
+- train_batch_size: 8
+- eval_batch_size: 8
+- seed: 4242
+- gradient_accumulation_steps: 32
+- total_train_batch_size: 256
+- optimizer: Adam with betas=(0.9,0.98) and epsilon=1e-08
+- lr_scheduler_type: cosine
+- lr_scheduler_warmup_steps: 1000
+- training_steps: 10000
+- mixed_precision_training: Native AMP
+
+### Training results
+
+| Training Loss | Epoch   | Step  | Validation Loss | Wer    | Cer    |
+|:-------------:|:-------:|:-----:|:---------------:|:------:|:------:|
+| 1598.6352     | 0.5451  | 500   | 1597.7738       | 1.0    | 1.0    |
+| 698.4129      | 1.0903  | 1000  | 574.9319        | 0.6189 | 0.2334 |
+| 519.6434      | 1.6354  | 1500  | 463.2092        | 0.4886 | 0.1877 |
+| 457.419       | 2.1806  | 2000  | 419.9849        | 0.4432 | 0.1685 |
+| 424.6695      | 2.7257  | 2500  | 392.2585        | 0.4069 | 0.1562 |
+| 402.5123      | 3.2709  | 3000  | 374.3170        | 0.3963 | 0.1519 |
+| 388.9287      | 3.8160  | 3500  | 356.5869        | 0.3833 | 0.1479 |
+| 380.4898      | 4.3612  | 4000  | 348.6322        | 0.3718 | 0.1425 |
+| 371.7486      | 4.9063  | 4500  | 343.7823        | 0.3663 | 0.1401 |
+| 367.6832      | 5.4514  | 5000  | 331.9190        | 0.3611 | 0.1389 |
+| 357.0084      | 5.9966  | 5500  | 330.0962        | 0.3570 | 0.1367 |
+| 355.2035      | 6.5417  | 6000  | 326.4210        | 0.3526 | 0.1352 |
+| 352.9193      | 7.0869  | 6500  | 322.4225        | 0.3507 | 0.1342 |
+| 348.084       | 7.6320  | 7000  | 321.1701        | 0.3487 | 0.1338 |
+| 347.2086      | 8.1772  | 7500  | 319.7442        | 0.3471 | 0.1330 |
+| 347.4033      | 8.7223  | 8000  | 317.6507        | 0.3453 | 0.1326 |
+| 347.9957      | 9.2675  | 8500  | 316.4468        | 0.3440 | 0.1320 |
+| 342.1603      | 9.8126  | 9000  | 316.0001        | 0.3435 | 0.1319 |
+| 343.359       | 10.3578 | 9500  | 316.1837        | 0.3438 | 0.1320 |
+| 344.4635      | 10.9029 | 10000 | 316.1826        | 0.3435 | 0.1319 |
+
+
+### Framework versions
+
+- Transformers 4.44.2
+- Pytorch 2.4.1+cu121
+- Datasets 2.21.0
+- Tokenizers 0.19.1

added_tokens.json

@@ -0,0 +1,6 @@
+{
+  "</s>": 43,
+  "<pad>": 45,
+  "<s>": 42,
+  "<unk>": 44
+}

alphabet.json

@@ -0,0 +1 @@
+{"labels": ["0", "1", "2", "3", "4", "5", "6", "7", "8", "9", "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", " ", "\u00e5", "\u00e6", "\u00e9", "\u00f8", "\u00fc", "<s>", "</s>", "\u2047", ""], "is_bpe": false}

config.json

@@ -0,0 +1,109 @@
+{
+  "_name_or_path": "chcaa/xls-r-300m-danish",
+  "activation_dropout": 0.1,
+  "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": 42,
+  "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": "sum",
+  "ctc_zero_infinity": true,
+  "diversity_loss_weight": 0.1,
+  "do_stable_layer_norm": true,
+  "eos_token_id": 43,
+  "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,
+  "gradient_checkpointing": false,
+  "hidden_act": "gelu",
+  "hidden_dropout": 0.0,
+  "hidden_size": 1024,
+  "initializer_range": 0.02,
+  "intermediate_size": 4096,
+  "layer_norm_eps": 1e-05,
+  "layerdrop": 0.1,
+  "mask_feature_length": 64,
+  "mask_feature_min_masks": 0,
+  "mask_feature_prob": 0.5,
+  "mask_time_length": 10,
+  "mask_time_min_masks": 2,
+  "mask_time_prob": 0.5,
+  "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": 45,
+  "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": "4.44.2",
+  "use_weighted_layer_sum": false,
+  "vocab_size": 46,
+  "xvector_output_dim": 512
+}

language_model/5gram.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1223290be61f10bd2aaebbb454479e6cd66b7a2b816b0eb5f1571e6ae6cf899f
+size 4332614415

language_model/attrs.json

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

language_model/unigrams.txt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:34e6c338910c566bb12d367f494736dbf4f8395d4b9bd18364b7d61ba29d7675
+size 18220276

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c1bce4abd7d25c43669d570c295f3021b3f279e5daa49258f7fc3c80ad0f7617
+size 1261996080

preprocessor_config.json

@@ -0,0 +1,10 @@
+{
+  "do_normalize": true,
+  "feature_extractor_type": "Wav2Vec2FeatureExtractor",
+  "feature_size": 1,
+  "padding_side": "right",
+  "padding_value": 0.0,
+  "processor_class": "Wav2Vec2ProcessorWithLM",
+  "return_attention_mask": true,
+  "sampling_rate": 16000
+}

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": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer_config.json

@@ -0,0 +1,48 @@
+{
+  "added_tokens_decoder": {
+    "42": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "43": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "44": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "45": {
+      "content": "<pad>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": true,
+  "do_lower_case": false,
+  "eos_token": "</s>",
+  "model_max_length": 512,
+  "pad_token": "<pad>",
+  "processor_class": "Wav2Vec2ProcessorWithLM",
+  "replace_word_delimiter_char": " ",
+  "target_lang": null,
+  "tokenizer_class": "Wav2Vec2CTCTokenizer",
+  "unk_token": "<unk>",
+  "word_delimiter_token": "|"
+}

vocab.json

@@ -0,0 +1,44 @@
+{
+  "0": 0,
+  "1": 1,
+  "2": 2,
+  "3": 3,
+  "4": 4,
+  "5": 5,
+  "6": 6,
+  "7": 7,
+  "8": 8,
+  "9": 9,
+  "a": 10,
+  "b": 11,
+  "c": 12,
+  "d": 13,
+  "e": 14,
+  "f": 15,
+  "g": 16,
+  "h": 17,
+  "i": 18,
+  "j": 19,
+  "k": 20,
+  "l": 21,
+  "m": 22,
+  "n": 23,
+  "o": 24,
+  "p": 25,
+  "q": 26,
+  "r": 27,
+  "s": 28,
+  "t": 29,
+  "u": 30,
+  "v": 31,
+  "w": 32,
+  "x": 33,
+  "y": 34,
+  "z": 35,
+  "|": 36,
+  "å": 37,
+  "æ": 38,
+  "é": 39,
+  "ø": 40,
+  "ü": 41
+}