Update README.md

README.md

@@ -1,87 +1,47 @@
 ---
+language:
+- da
 license: apache-2.0
-tags:
-- generated_from_trainer
+tasks:
+- automatic-speech-recognition
 datasets:
-- common_voice
+- common_voice_8_0
+metrics:
+- wer
 model-index:
 - name: alvenir-wav2vec2-base-cv8-da
-  results: []
+  results:
+  - task: 
+      type: automatic-speech-recognition
+    dataset:
+      type: Alvenir/alvenir_asr_da_eval
+      name: Alvenir ASR test dataset
+    metrics:
+      - type: wer
+        value: 41.08
+  - task: 
+      type: automatic-speech-recognition
+    dataset:
+      type: mozilla-foundation/common_voice_8_0
+      args: da
+      name: Danish Common Voice 8.0
+    metrics:
+      - type: wer
+        value: 46.05
 ---
 
-<!-- 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. -->
-
-# alvenir-wav2vec2-base-cv8-da
-
-This model is a fine-tuned version of [Alvenir/wav2vec2-base-da](https://huggingface.co/Alvenir/wav2vec2-base-da) on the common_voice dataset.
-It achieves the following results on the evaluation set:
-- Loss: 1.1833
-- Wer: 0.4851
+# Alvenir-Wav2vec2-base-CV8-da
 
 ## 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: 4e-05
-- train_batch_size: 8
-- eval_batch_size: 8
-- seed: 42
-- gradient_accumulation_steps: 4
-- total_train_batch_size: 32
-- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
-- lr_scheduler_type: linear
-- lr_scheduler_warmup_steps: 500
-- num_epochs: 500
-- mixed_precision_training: Native AMP
-
-### Training results
+This model is a fine-tuned version of the Danish acoustic model [Alvenir/wav2vec2-base-da](https://huggingface.co/Alvenir/wav2vec2-base-da) on the Danish part of [Common Voice 8.0](https://huggingface.co/datasets/mozilla-foundation/common_voice_8_0), containing ~6 crowdsourced hours of read-aloud Danish speech.
 
-| Training Loss | Epoch  | Step | Validation Loss | Wer    |
-|:-------------:|:------:|:----:|:---------------:|:------:|
-| 3.1193        | 5.55   | 300  | 2.9589          | 1.0    |
-| 2.7263        | 11.11  | 600  | 1.9051          | 0.9900 |
-| 0.8586        | 16.66  | 900  | 0.8672          | 0.6476 |
-| 0.5162        | 22.22  | 1200 | 0.8564          | 0.5739 |
-| 0.4339        | 27.77  | 1500 | 0.8644          | 0.5396 |
-| 0.3467        | 33.33  | 1800 | 0.8947          | 0.5382 |
-| 0.2742        | 38.88  | 2100 | 0.9070          | 0.5184 |
-| 0.2675        | 44.44  | 2400 | 0.8916          | 0.5070 |
-| 0.2362        | 49.99  | 2700 | 0.9895          | 0.5108 |
-| 0.2076        | 55.55  | 3000 | 0.9703          | 0.5054 |
-| 0.1944        | 61.11  | 3300 | 0.9893          | 0.5014 |
-| 0.1617        | 66.66  | 3600 | 1.0011          | 0.5080 |
-| 0.1634        | 72.22  | 3900 | 1.0013          | 0.4959 |
-| 0.1488        | 77.77  | 4200 | 1.0574          | 0.5    |
-| 0.1447        | 83.33  | 4500 | 1.0825          | 0.4917 |
-| 0.126         | 88.88  | 4800 | 1.0953          | 0.4949 |
-| 0.1326        | 94.44  | 5100 | 1.1292          | 0.4925 |
-| 0.1229        | 99.99  | 5400 | 1.1357          | 0.4998 |
-| 0.1216        | 105.55 | 5700 | 1.1957          | 0.4933 |
-| 0.1153        | 111.11 | 6000 | 1.1763          | 0.4847 |
-| 0.1054        | 116.66 | 6300 | 1.2280          | 0.5012 |
-| 0.1056        | 122.22 | 6600 | 1.1631          | 0.4878 |
-| 0.1009        | 127.77 | 6900 | 1.1889          | 0.4918 |
-| 0.0936        | 133.33 | 7200 | 1.2309          | 0.4916 |
-| 0.1           | 138.88 | 7500 | 1.1833          | 0.4851 |
 
+## Performance
 
-### Framework versions
+The model achieves the following WER scores (lower is better):
 
-- Transformers 4.16.2
-- Pytorch 1.10.2+cu102
-- Datasets 1.18.3
-- Tokenizers 0.11.0
+| **Dataset** | **WER without LM** | **WER with 5-gram LM** |
+| :---:   | ---: | ---: |
+| [Danish part of Common Voice 8.0](https://huggingface.co/datasets/mozilla-foundation/common_voice_8_0/viewer/da/train) | 46.05 | xx.xx |
+| [Alvenir test set](https://huggingface.co/datasets/Alvenir/alvenir_asr_da_eval) | 41.08 | xx.xx |