|
allow_missing_params: false |
|
amp: false |
|
apex_amp: false |
|
batch_sentences_multiple_of: 8 |
|
batch_size: 512 |
|
batch_type: word |
|
bow_task_pos_weight: 10 |
|
bow_task_weight: 1.0 |
|
bucket_scaling: false |
|
bucket_width: 8 |
|
cache_last_best_params: 0 |
|
cache_metric: perplexity |
|
cache_strategy: best |
|
checkpoint_improvement_threshold: 0.0 |
|
checkpoint_interval: 4000 |
|
clamp_to_dtype: false |
|
config: null |
|
decode_and_evaluate: 500 |
|
decoder: transformer |
|
deepspeed_bf16: false |
|
deepspeed_fp16: false |
|
device_id: 0 |
|
dist: false |
|
dry_run: false |
|
dtype: float32 |
|
embed_dropout: |
|
- 0.0 |
|
- 0.0 |
|
encoder: transformer |
|
end_of_prepending_tag: null |
|
env: null |
|
fixed_param_names: [] |
|
fixed_param_strategy: null |
|
gradient_clipping_threshold: 1.0 |
|
gradient_clipping_type: none |
|
ignore_extra_params: false |
|
initial_learning_rate: 0.0002 |
|
keep_initializations: false |
|
keep_last_params: -1 |
|
label_smoothing: 0.1 |
|
label_smoothing_impl: mxnet |
|
learning_rate_reduce_factor: 0.9 |
|
learning_rate_reduce_num_not_improved: 8 |
|
learning_rate_scheduler_type: plateau-reduce |
|
learning_rate_warmup: 0 |
|
length_task: null |
|
length_task_layers: 1 |
|
length_task_weight: 1.0 |
|
lhuc: null |
|
local_rank: null |
|
loglevel: INFO |
|
loglevel_secondary_workers: INFO |
|
max_checkpoints: null |
|
max_num_checkpoint_not_improved: 20 |
|
max_num_epochs: null |
|
max_samples: null |
|
max_seconds: null |
|
max_seq_len: |
|
- 512 |
|
- 128 |
|
max_updates: null |
|
min_num_epochs: null |
|
min_samples: null |
|
min_updates: null |
|
momentum: 0.0 |
|
neural_vocab_selection: null |
|
neural_vocab_selection_block_loss: false |
|
no_bucketing: false |
|
no_logfile: false |
|
no_reload_on_learning_rate_reduce: false |
|
num_embed: |
|
- null |
|
- null |
|
num_layers: |
|
- 6 |
|
- 6 |
|
num_words: |
|
- 0 |
|
- 0 |
|
optimized_metric: chrf |
|
optimizer: adam |
|
optimizer_betas: |
|
- 0.9 |
|
- 0.999 |
|
optimizer_eps: 1.0e-08 |
|
output: /shares/volk.cl.uzh/amoryo/checkpoints/sockeye/expanded-cleaned/model |
|
overwrite_output: false |
|
pad_vocab_to_multiple_of: 8 |
|
params: /shares/volk.cl.uzh/amoryo/checkpoints/sockeye/expanded/model/params.best |
|
prepared_data: /shares/volk.cl.uzh/amoryo/checkpoints/sockeye/expanded-cleaned/train_data |
|
quiet: false |
|
quiet_secondary_workers: false |
|
seed: 1 |
|
shared_vocab: false |
|
source: null |
|
source_factor_vocabs: [] |
|
source_factors: [] |
|
source_factors_combine: [] |
|
source_factors_num_embed: [] |
|
source_factors_share_embedding: [] |
|
source_factors_use_source_vocab: [] |
|
source_vocab: null |
|
stop_training_on_decoder_failure: false |
|
target: null |
|
target_factor_vocabs: [] |
|
target_factors: [] |
|
target_factors_combine: [] |
|
target_factors_num_embed: [] |
|
target_factors_share_embedding: [] |
|
target_factors_use_target_vocab: [] |
|
target_factors_weight: |
|
- 1.0 |
|
target_vocab: null |
|
tf32: true |
|
transformer_activation_type: |
|
- relu |
|
- relu |
|
transformer_attention_heads: |
|
- 8 |
|
- 8 |
|
transformer_block_prepended_cross_attention: false |
|
transformer_dropout_act: &id001 |
|
- 0.1 |
|
- 0.1 |
|
transformer_dropout_attention: *id001 |
|
transformer_dropout_prepost: *id001 |
|
transformer_feed_forward_num_hidden: |
|
- 2048 |
|
- 2048 |
|
transformer_feed_forward_use_glu: false |
|
transformer_model_size: |
|
- 512 |
|
- 512 |
|
transformer_positional_embedding_type: fixed |
|
transformer_postprocess: |
|
- dr |
|
- dr |
|
transformer_preprocess: |
|
- n |
|
- n |
|
update_interval: 1 |
|
use_cpu: false |
|
validation_source: /home/amoryo/sign-language/signbank-annotation/signbank-plus/data/parallel/cleaned/dev.source.tokenized |
|
validation_source_factors: [] |
|
validation_target: /shares/volk.cl.uzh/amoryo/checkpoints/sockeye/expanded-cleaned/dev.target.bpe |
|
validation_target_factors: [] |
|
weight_decay: 0.0 |
|
weight_tying_type: none |
|
word_min_count: |
|
- 1 |
|
- 1 |
|
|
