wandb/run-20240512_102340-g9bmlvdf/files/config.yaml

wandb_version: 1

_wandb:
  desc: null
  value:
    python_version: 3.10.14
    cli_version: 0.17.0
    framework: huggingface
    huggingface_version: 4.40.0.dev0
    is_jupyter_run: false
    is_kaggle_kernel: false
    start_time: 1715509420
    t:
      1:
      - 1
      - 5
      - 11
      - 49
      - 51
      - 53
      - 55
      - 71
      - 84
      - 98
      - 99
      - 100
      - 105
      2:
      - 1
      - 5
      - 11
      - 49
      - 51
      - 53
      - 55
      - 71
      - 84
      - 98
      - 99
      - 100
      - 105
      3:
      - 3
      - 7
      - 23
      - 66
      4: 3.10.14
      5: 0.17.0
      6: 4.40.0.dev0
      8:
      - 5
      9:
        1: transformers_trainer
      13: linux-x86_64
    m:
    - 1: train/global_step
      6:
      - 3
    - 1: train/loss
      5: 1
      6:
      - 1
    - 1: train/grad_norm
      5: 1
      6:
      - 1
    - 1: train/learning_rate
      5: 1
      6:
      - 1
    - 1: train/epoch
      5: 1
      6:
      - 1
vocab_size:
  desc: null
  value: 128256
max_position_embeddings:
  desc: null
  value: 8192
hidden_size:
  desc: null
  value: 4096
intermediate_size:
  desc: null
  value: 14336
num_hidden_layers:
  desc: null
  value: 32
num_attention_heads:
  desc: null
  value: 32
num_key_value_heads:
  desc: null
  value: 8
hidden_act:
  desc: null
  value: silu
initializer_range:
  desc: null
  value: 0.02
rms_norm_eps:
  desc: null
  value: 1.0e-05
pretraining_tp:
  desc: null
  value: 1
use_cache:
  desc: null
  value: false
rope_theta:
  desc: null
  value: 500000.0
rope_scaling:
  desc: null
  value: null
attention_bias:
  desc: null
  value: false
attention_dropout:
  desc: null
  value: 0.0
return_dict:
  desc: null
  value: true
output_hidden_states:
  desc: null
  value: false
output_attentions:
  desc: null
  value: false
torchscript:
  desc: null
  value: false
torch_dtype:
  desc: null
  value: bfloat16
use_bfloat16:
  desc: null
  value: false
tf_legacy_loss:
  desc: null
  value: false
pruned_heads:
  desc: null
  value: {}
tie_word_embeddings:
  desc: null
  value: false
chunk_size_feed_forward:
  desc: null
  value: 0
is_encoder_decoder:
  desc: null
  value: false
is_decoder:
  desc: null
  value: false
cross_attention_hidden_size:
  desc: null
  value: null
add_cross_attention:
  desc: null
  value: false
tie_encoder_decoder:
  desc: null
  value: false
max_length:
  desc: null
  value: 20
min_length:
  desc: null
  value: 0
do_sample:
  desc: null
  value: false
early_stopping:
  desc: null
  value: false
num_beams:
  desc: null
  value: 1
num_beam_groups:
  desc: null
  value: 1
diversity_penalty:
  desc: null
  value: 0.0
temperature:
  desc: null
  value: 1.0
top_k:
  desc: null
  value: 50
top_p:
  desc: null
  value: 1.0
typical_p:
  desc: null
  value: 1.0
repetition_penalty:
  desc: null
  value: 1.0
length_penalty:
  desc: null
  value: 1.0
no_repeat_ngram_size:
  desc: null
  value: 0
encoder_no_repeat_ngram_size:
  desc: null
  value: 0
bad_words_ids:
  desc: null
  value: null
num_return_sequences:
  desc: null
  value: 1
output_scores:
  desc: null
  value: false
return_dict_in_generate:
  desc: null
  value: false
forced_bos_token_id:
  desc: null
  value: null
forced_eos_token_id:
  desc: null
  value: null
remove_invalid_values:
  desc: null
  value: false
exponential_decay_length_penalty:
  desc: null
  value: null
suppress_tokens:
  desc: null
  value: null
begin_suppress_tokens:
  desc: null
  value: null
architectures:
  desc: null
  value:
  - LlamaForCausalLM
finetuning_task:
  desc: null
  value: null
id2label:
  desc: null
  value:
    '0': LABEL_0
    '1': LABEL_1
label2id:
  desc: null
  value:
    LABEL_0: 0
    LABEL_1: 1
tokenizer_class:
  desc: null
  value: null
prefix:
  desc: null
  value: null
bos_token_id:
  desc: null
  value: 128000
pad_token_id:
  desc: null
  value: null
eos_token_id:
  desc: null
  value: 128009
sep_token_id:
  desc: null
  value: null
decoder_start_token_id:
  desc: null
  value: null
task_specific_params:
  desc: null
  value: null
problem_type:
  desc: null
  value: null
_name_or_path:
  desc: null
  value: PygmalionAI/pyggel-ckpt-2947
transformers_version:
  desc: null
  value: 4.40.0.dev0
model_type:
  desc: null
  value: llama
output_dir:
  desc: null
  value: ./pyg1m2.3
overwrite_output_dir:
  desc: null
  value: false
do_train:
  desc: null
  value: false
do_eval:
  desc: null
  value: true
do_predict:
  desc: null
  value: false
evaluation_strategy:
  desc: null
  value: steps
prediction_loss_only:
  desc: null
  value: false
per_device_train_batch_size:
  desc: null
  value: 1
per_device_eval_batch_size:
  desc: null
  value: 1
per_gpu_train_batch_size:
  desc: null
  value: null
per_gpu_eval_batch_size:
  desc: null
  value: null
gradient_accumulation_steps:
  desc: null
  value: 1
eval_accumulation_steps:
  desc: null
  value: 1
eval_delay:
  desc: null
  value: 0
learning_rate:
  desc: null
  value: 5.0e-05
weight_decay:
  desc: null
  value: 0.03
adam_beta1:
  desc: null
  value: 0.9
adam_beta2:
  desc: null
  value: 0.999
adam_epsilon:
  desc: null
  value: 1.0e-08
max_grad_norm:
  desc: null
  value: 1.0
num_train_epochs:
  desc: null
  value: 2
max_steps:
  desc: null
  value: -1
lr_scheduler_type:
  desc: null
  value: cosine
lr_scheduler_kwargs:
  desc: null
  value: {}
warmup_ratio:
  desc: null
  value: 0.0
warmup_steps:
  desc: null
  value: 569
log_level:
  desc: null
  value: passive
log_level_replica:
  desc: null
  value: warning
log_on_each_node:
  desc: null
  value: true
logging_dir:
  desc: null
  value: ./pyg1m2.3/runs/May12_10-23-28_a9815c79a99f
logging_strategy:
  desc: null
  value: steps
logging_first_step:
  desc: null
  value: false
logging_steps:
  desc: null
  value: 1
logging_nan_inf_filter:
  desc: null
  value: true
save_strategy:
  desc: null
  value: steps
save_steps:
  desc: null
  value: 0.03333333333333333
save_total_limit:
  desc: null
  value: 4
save_safetensors:
  desc: null
  value: true
save_on_each_node:
  desc: null
  value: false
save_only_model:
  desc: null
  value: false
no_cuda:
  desc: null
  value: false
use_cpu:
  desc: null
  value: false
use_mps_device:
  desc: null
  value: false
seed:
  desc: null
  value: 42
data_seed:
  desc: null
  value: null
jit_mode_eval:
  desc: null
  value: false
use_ipex:
  desc: null
  value: false
bf16:
  desc: null
  value: true
fp16:
  desc: null
  value: false
fp16_opt_level:
  desc: null
  value: O1
half_precision_backend:
  desc: null
  value: auto
bf16_full_eval:
  desc: null
  value: false
fp16_full_eval:
  desc: null
  value: false
tf32:
  desc: null
  value: false
local_rank:
  desc: null
  value: 0
ddp_backend:
  desc: null
  value: null
tpu_num_cores:
  desc: null
  value: null
tpu_metrics_debug:
  desc: null
  value: false
debug:
  desc: null
  value: []
dataloader_drop_last:
  desc: null
  value: true
eval_steps:
  desc: null
  value: 0.015
dataloader_num_workers:
  desc: null
  value: 0
dataloader_prefetch_factor:
  desc: null
  value: null
past_index:
  desc: null
  value: -1
run_name:
  desc: null
  value: ./pyg1m2.3
disable_tqdm:
  desc: null
  value: false
remove_unused_columns:
  desc: null
  value: true
label_names:
  desc: null
  value: null
load_best_model_at_end:
  desc: null
  value: false
metric_for_best_model:
  desc: null
  value: null
greater_is_better:
  desc: null
  value: null
ignore_data_skip:
  desc: null
  value: false
fsdp:
  desc: null
  value:
  - full_shard
  - auto_wrap
fsdp_min_num_params:
  desc: null
  value: 0
fsdp_config:
  desc: null
  value:
    fsdp_limit_all_gathers: true
    fsdp_sync_module_states: true
    fsdp_offload_params: true
    fsdp_use_orig_params: false
    fsdp_cpu_ram_efficient_loading: false
    fsdp_auto_wrap_policy: TRANSFORMER_BASED_WRAP
    fsdp_transformer_layer_cls_to_wrap: LlamaDecoderLayer
    fsdp_state_dict_type: FULL_STATE_DICT
    fsdp_sharding_strategy: FULL_SHARD
    fsdp_forward_prefetch: FORWARD_PRE
    fsdp_backward_prefetch: BACKWARD_PRE
    min_num_params: 0
    xla: false
    xla_fsdp_v2: false
    xla_fsdp_grad_ckpt: false
fsdp_transformer_layer_cls_to_wrap:
  desc: null
  value: null
accelerator_config:
  desc: null
  value:
    split_batches: false
    dispatch_batches: null
    even_batches: true
    use_seedable_sampler: true
    gradient_accumulation_kwargs: null
deepspeed:
  desc: null
  value: null
label_smoothing_factor:
  desc: null
  value: 0.0
optim:
  desc: null
  value: adamw_torch
optim_args:
  desc: null
  value: adam_beta1=0.98,adam_beta2=0.97,amsgrad=True
adafactor:
  desc: null
  value: false
group_by_length:
  desc: null
  value: null
length_column_name:
  desc: null
  value: length
report_to:
  desc: null
  value:
  - wandb
ddp_find_unused_parameters:
  desc: null
  value: false
ddp_bucket_cap_mb:
  desc: null
  value: null
ddp_broadcast_buffers:
  desc: null
  value: null
dataloader_pin_memory:
  desc: null
  value: true
dataloader_persistent_workers:
  desc: null
  value: false
skip_memory_metrics:
  desc: null
  value: true
use_legacy_prediction_loop:
  desc: null
  value: false
push_to_hub:
  desc: null
  value: true
resume_from_checkpoint:
  desc: null
  value: null
hub_model_id:
  desc: null
  value: Alignment-Lab-AI/pygmalion-3-1m-2.3
hub_strategy:
  desc: null
  value: all_checkpoints
hub_token:
  desc: null
  value: <HUB_TOKEN>
hub_private_repo:
  desc: null
  value: true
hub_always_push:
  desc: null
  value: true
gradient_checkpointing:
  desc: null
  value: unsloth
gradient_checkpointing_kwargs:
  desc: null
  value:
    use_reentrant: true
include_inputs_for_metrics:
  desc: null
  value: false
fp16_backend:
  desc: null
  value: auto
push_to_hub_model_id:
  desc: null
  value: null
push_to_hub_organization:
  desc: null
  value: null
push_to_hub_token:
  desc: null
  value: <PUSH_TO_HUB_TOKEN>
mp_parameters:
  desc: null
  value: ''
auto_find_batch_size:
  desc: null
  value: false
full_determinism:
  desc: null
  value: false
torchdynamo:
  desc: null
  value: null
ray_scope:
  desc: null
  value: last
ddp_timeout:
  desc: null
  value: 1800
torch_compile:
  desc: null
  value: true
torch_compile_backend:
  desc: null
  value: inductor
torch_compile_mode:
  desc: null
  value: null
dispatch_batches:
  desc: null
  value: null
split_batches:
  desc: null
  value: null
include_tokens_per_second:
  desc: null
  value: false
include_num_input_tokens_seen:
  desc: null
  value: false
neftune_noise_alpha:
  desc: null
  value: null
optim_target_modules:
  desc: null
  value: null
lr_quadratic_warmup:
  desc: null
  value: false
pretraining:
  desc: null
  value: false
sample_packing:
  desc: null
  value: true
multipack_real_batches:
  desc: null
  value: false
eval_sample_packing:
  desc: null
  value: false
sample_packing_efficiency:
  desc: null
  value: 0.92
max_seq_length:
  desc: null
  value: 8192
sample_packing_seq_len_multiplier:
  desc: null
  value: 1
relora_steps:
  desc: null
  value: null
relora_warmup_steps:
  desc: null
  value: null
relora_anneal_steps:
  desc: null
  value: null
relora_prune_ratio:
  desc: null
  value: 0.9
bench_split:
  desc: null
  value: eval
bench_dataset:
  desc: null
  value: pharaouk/dharma-1/dharma_1_mini.json
do_bench_eval:
  desc: null
  value: false
do_causal_lm_eval:
  desc: null
  value: false
max_bench_samples:
  desc: null
  value: null
bench_source_max_len:
  desc: null
  value: 2048
cosine_min_lr_ratio:
  desc: null
  value: null
cosine_constant_lr_ratio:
  desc: null
  value: null
loraplus_lr_ratio:
  desc: null
  value: null
loraplus_lr_embedding:
  desc: null
  value: 1.0e-06
qlora:
  desc: null
  value: false
orpo_alpha:
  desc: null
  value: null
lisa_n_layers:
  desc: null
  value: null
lisa_step_interval:
  desc: null
  value: null
lisa_layers_attribute:
  desc: null
  value: null
curriculum_sampling:
  desc: null
  value: null