Jobs
job-git_github.com_AGMoller_worker_vs_gpt.git_src_worker_vs_gpt___main__.py
Job
14 versions
2 years ago
Job version
latest
v13
Version metadata
Version path
cocoons/worker_vs_gpt/job-git_github.com_AGMoller_worker_vs_gpt.git_src_worker_vs_gpt___main__.py:latest
Description
What makes this job version special?
Aliases
latest
v13
Creator
hrmussaCreation date
May 21st 2023 at 10:52am
Python version
3.10.6
Entry point
python src/worker_vs_gpt/__main__.py
Arguments
N/A
Source type
Artifact link
Input types
Key
Value
_name_or_path
string
adafactor
boolean
adam_beta1
number
adam_beta2
number
adam_epsilon
number
add_cross_attention
boolean
architectures
list
attention_probs_dropout_prob
number
augmentation_model
string
auto_find_batch_size
boolean
bad_words_ids
none
batch_size
number
begin_suppress_tokens
none
bf16
boolean
bf16_full_eval
boolean
bos_token_id
none
chunk_size_feed_forward
number
ckpt
string
classifier_dropout
none
cross_attention_hidden_size
none
data_seed
string
dataloader_drop_last
boolean
dataloader_num_workers
number
dataloader_pin_memory
boolean
ddp_backend
string
ddp_bucket_cap_mb
string
ddp_find_unused_parameters
string
ddp_timeout
number
debug
string
decoder_start_token_id
none
deepspeed
string
disable_tqdm
boolean
diversity_penalty
number
do_eval
boolean
do_predict
boolean
do_sample
boolean
do_train
boolean
early_stopping
boolean
encoder_no_repeat_ngram_size
number
eos_token_id
none
eval_accumulation_steps
string
eval_batch_size
number
eval_delay
number
eval_steps
string
evaluation_strategy
string
exponential_decay_length_penalty
none
finetuning_task
none
forced_bos_token_id
none
forced_eos_token_id
none
fp16
boolean
fp16_backend
string
fp16_full_eval
boolean
fp16_opt_level
string
fsdp
string
fsdp_config
string
fsdp_min_num_params
number
fsdp_transformer_layer_cls_to_wrap
string
full_determinism
boolean
gradient_accumulation_steps
number
gradient_checkpointing
boolean
greater_is_better
boolean
group_by_length
boolean
half_precision_backend
string
hidden_act
string
hidden_dropout_prob
number
hidden_size
number
hub_model_id
string
hub_private_repo
boolean
hub_strategy
string
hub_token
string
id2label
0
string
1
string
2
string
3
string
4
string
5
string
6
string
7
string
8
string
ignore_data_skip
boolean
include_inputs_for_metrics
boolean
initializer_range
number
intermediate_size
number
is_decoder
boolean
is_encoder_decoder
boolean
jit_mode_eval
boolean
label2id
LABEL_0
number
LABEL_1
number
LABEL_2
number
LABEL_3
number
LABEL_4
number
LABEL_5
number
LABEL_6
number
LABEL_7
number
LABEL_8
number
label_names
string
label_smoothing_factor
number
layer_norm_eps
number
learning_rate
number
length_column_name
string
length_penalty
number
load_best_model_at_end
boolean
local_rank
number
log_level
string
log_level_replica
string
log_on_each_node
boolean
logging_dir
string
logging_first_step
boolean
logging_nan_inf_filter
boolean
logging_steps
number
logging_strategy
string
lr
number
lr_scheduler_type
string
max_grad_norm
number
max_length
number
max_position_embeddings
number
max_steps
number
metric_for_best_model
string
min_length
number
model_type
string
mp_parameters
string
no_cuda
boolean
no_repeat_ngram_size
number
num_attention_heads
number
num_beam_groups
number
num_beams
number
num_epochs
number
num_hidden_layers
number
num_return_sequences
number
num_train_epochs
number
optim
string
optim_args
string
output_attentions
boolean
output_dir
string
output_hidden_states
boolean
output_scores
boolean
overwrite_output_dir
boolean
pad_token_id
number
past_index
number
per_device_eval_batch_size
number
per_device_train_batch_size
number
per_gpu_eval_batch_size
string
per_gpu_train_batch_size
string
position_embedding_type
string
prediction_loss_only
boolean
prefix
none
problem_type
none
pruned_heads
push_to_hub
boolean
push_to_hub_model_id
string
push_to_hub_organization
string
push_to_hub_token
string
ray_scope
string
remove_invalid_values
boolean
remove_unused_columns
boolean
repetition_penalty
number
report_to
string
resume_from_checkpoint
string
return_dict
boolean
return_dict_in_generate
boolean
run_name
string
sampling
string
save_on_each_node
boolean
save_safetensors
boolean
save_steps
number
save_strategy
string
save_total_limit
number
seed
number
sep_token_id
none
sharded_ddp
string
skip_memory_metrics
boolean
suppress_tokens
none
task_specific_params
none
temperature
number
tf32
string
tf_legacy_loss
boolean
tie_encoder_decoder
boolean
tie_word_embeddings
boolean
tokenizer_class
none
top_k
number
top_p
number
torch_compile
boolean
torch_compile_backend
string
torch_compile_mode
string
torch_dtype
string
torchdynamo
string
torchscript
boolean
tpu_metrics_debug
boolean
tpu_num_cores
string
train_batch_size
number
train_size
number
transformers_version
string
type_vocab_size
number
typical_p
number
use_augmented_data
boolean
use_bfloat16
boolean
use_cache
boolean
use_ipex
boolean
use_legacy_prediction_loop
boolean
use_mps_device
boolean
vocab_size
number
warmup_ratio
number
warmup_steps
number
weight_decay
number
xpu_backend
string
Output types
Key
Value
_runtime
number
_step
number
_timestamp
number
classification_report
_latest_artifact_path
string
_type
string
artifact_path
string
ncols
number
nrows
number
path
string
sha256
string
size
number
eval
accuracy
number
f1
number
loss
number
roc_auc
number
runtime
number
samples_per_second
number
steps_per_second
number
test
accuracy
number
f1
number
loss
number
roc_auc
number
runtime
number
samples_per_second
number
steps_per_second
number
train
epoch
number
global_step
number
learning_rate
number
loss
number
total_flos
number
train_loss
number
train_runtime
number
train_samples_per_second
number
train_steps_per_second
number
Files
Root
Name
Size
Type
diff.patch
5.7KB
Document
requirements.frozen.txt
3.7KB
Plain Text
wandb-job.json
25.2KB
JSON Document
Loading...