Update README.md

README.md

@@ -30,28 +30,34 @@ model-index:
 ---
 
 # Norwegian Wav2Vec2 Model - 1B - Bokmål
-This achieves the following results on the evaluation set:
+This achieves the following results on the test set with a 5-gram KenLM:
 - WER: 0.0668
 - CER: 0.0256
 
-## Model description
-This is one of several Wav2Vec-models created during the HuggingFace hosted [Robust Speech Event](https://discuss.huggingface.co/t/open-to-the-community-robust-speech-recognition-challenge/13614?s=09). In parallell with the event, the team also converted the [Norwegian Parliamentary Speech Corpus (NPSC)](https://huggingface.co/datasets/NbAiLab/NPSC) to the HuggingFace Dataset format and used that as the main source for training.
+Without using a language model, we are getting these results:
+- WER: ???
+- CER: ???
 
+## Model description
+This is one of several Wav2Vec-models created during the 🤗 hosted [Robust Speech Event](https://discuss.huggingface.co/t/open-to-the-community-robust-speech-recognition-challenge/13614?s=09). In parallell with the event, the team also converted the [Norwegian Parliamentary Speech Corpus (NPSC)](https://huggingface.co/datasets/NbAiLab/NPSC) to the 🤗 Dataset format and used that as the main source for training.
 
 We do release all code developed during the event so that the Norwegian NLP community can build upon this to develop even better Norwegian ASR models. The finetuning of these models are not very compute demanding. You should after following the instructions here, be able to train your own automatic speech recognition system in less than a day with an average GPU.
 
+## Team
+The following people contributed to building this model: Rolv-Arild Braaten, Per Egil Kummervold, Andre Kåsen, Javier de la Rosa, Per Erik Solberg, and Freddy Wetjen. 
+
 ## Training procedure
-To reproduce this, we strongly recommend that you follow the [instructions from HuggingFace](https://github.com/huggingface/transformers/tree/master/examples/research_projects/robust-speech-event#talks) to train a simple Swedish model.
+To reproduce these results, we strongly recommend that you follow the [instructions from HuggingFace](https://github.com/huggingface/transformers/tree/master/examples/research_projects/robust-speech-event#talks) to train a simple Swedish model.
 
 When you have verified that you are able to do this, create a new repo. You can then start by copying the files **run.sh** and **run_speech_recognition_ctc.py** from our repo. You should be able to reproduce our results by just running this script. With some tweaking, you will most likely be able to build an even better ASR.
 
 ### 5-gram Language Model
-HuggingFace has provided another [very nice blog](https://huggingface.co/blog/wav2vec2-with-ngram) about how to add a 5-gram language model to improve the ASR model. You can build this from your own corpus, for instance by extracting some suitable text from the [Norwegian Colossal Corpus](https://huggingface.co/datasets/NbAiLab/NCC). You can also skip some of the steps in the guide, and copy the [5-gram model from this repo](https://huggingface.co/NbAiLab/XLSR-300M-bokmaal/tree/main/language_model).
+Adding a language model will improve the results of the model. 🤗 has provided another [very nice blog](https://huggingface.co/blog/wav2vec2-with-ngram) about how to add a 5-gram language model to improve the ASR model. You can build this from your own corpus, for instance by extracting some suitable text from the [Norwegian Colossal Corpus](https://huggingface.co/datasets/NbAiLab/NCC). You can also skip some of the steps in the guide, and copy the [5-gram model from this repo](https://huggingface.co/NbAiLab/XLSR-300M-bokmaal/tree/main/language_model).
 
 
-### Training settings
+### Parameters
 
-The following settings and hyperparameters were used during training:
+The following parameters were used during training:
 ```
 --dataset_name="NbAiLab/NPSC"
 --model_name_or_path="facebook/wav2vec2-xls-r-1b"
@@ -81,7 +87,7 @@ The following settings and hyperparameters were used during training:
 --mask_time_length="10" 
 --mask_feature_prob="0.25" 
 --mask_feature_length="64" 
---gradient_checkpointing \
+--gradient_checkpointing
 --min_duration_in_seconds="0.5" 
 --max_duration_in_seconds="30.0" 
 	--ctc_zero_infinity=True 
@@ -92,4 +98,12 @@ The following settings and hyperparameters were used during training:
 --do_train --do_eval 
 --push_to_hub 
 --preprocessing_num_workers="16"
-```
+```
+
+This training will take 3-4 days on an average GPU. You might get a decent model and faster results by changing these parameters:
+```
+--per_device_train_batch_size - Adjust this to the maximum of available memory. 16 or 24 might be good settings depending on your system
+--gradient_accumulation_steps - Can be adjusted even further up to increase batch size and speed up training without running into memory issues
+--learning_rate - Can be increased, maybe as high as 1e-4. Speeds up training but might add instability
+--epochs - Can be decreased significantly. This is a huge dataset and you might get a decent result already after a couple of epochs
+```