Train flair with different learning rates

[ugurcanarikan]

Now that flair has been trained with different embeddings and optimized parameters, it should be trained with different embeddings and learning rates, with other parameters being below:

• hidden_size: 256

• mini_batch_size: 16

• rnn_layers: 2

• learning rates = [0.05, 0.1, 0.15, 0.2]

[ugurcanarikan]

Flair has been trained with the decided parameters and different learning rates and embeddings and results have been added to models.