# Corpus opts:
data:
    corpus1:
        path_src: "train.es"
        path_tgt: "train.hr"
        transforms: [sentencepiece, prefix, suffix, filtertoolong]
        weight: 10
        src_prefix: "spa_Latn"
        tgt_prefix: "hrv_Latn"
        src_suffix: "</s>"
        tgt_suffix: ""

src_vocab: "shared_vocabulary.txt"

src_words_min_frequency: 1
src_vocab_size: 256206
tgt_vocab: "shared_vocabulary.txt"
tgt_words_min_frequency: 1
tgt_vocab_size: 256206
vocab_size_multiple: 1
decoder_start_token: '</s>'
#### Subword
src_subword_model: "flores200_sacrebleu_tokenizer_spm.model"
tgt_subword_model: "flores200_sacrebleu_tokenizer_spm.model"
src_subword_nbest: 1
src_subword_alpha: 0.0
tgt_subword_nbest: 1
tgt_subword_alpha: 0.0

update_vocab: true
train_from: "nllb-200-600M-onmt.pt"
reset_optim: all
save_data: "./finetuned600M"
save_model: "./finetuned600M/nllb-200-600-FT-onmt"
log_file: "./finetuned600M/nllb-200-600-FT-onmt.log"
keep_checkpoint: 50
save_checkpoint_steps: 100
average_decay: 0.0005
seed: 1234
report_every: 10
train_steps: 2000
valid_steps: 100
# Batching
bucket_size: 262144
num_workers: 4
prefetch_factor: 400
world_size: 1
gpu_ranks: [0]
batch_type: "tokens"
batch_size: 384
valid_batch_size: 384
batch_size_multiple: 1
accum_count: [32, 32, 32]
accum_steps: [0, 15000, 30000]
# Optimization
model_dtype: "fp16"
optim: "sgd"
learning_rate: 30
warmup_steps: 100
decay_method: "noam"
adam_beta2: 0.98
max_grad_norm: 0
label_smoothing: 0.1
param_init: 0
param_init_glorot: true
normalization: "tokens"

# Model
override_opts: true
encoder_type: transformer
decoder_type: transformer
enc_layers: 12
dec_layers: 12
heads: 16
hidden_size: 1024
word_vec_size: 1024
transformer_ff: 4096
add_qkvbias: true
add_ffnbias: true
dropout_steps: [0, 15000, 30000]
dropout: [0.1, 0.1, 0.1]
attention_dropout: [0.1, 0.1, 0.1]
share_decoder_embeddings: true
share_embeddings: true
position_encoding: true
position_encoding_type: 'SinusoidalConcat'