Qwen2.5-1.5B/train_params.yaml

add_ctxemb: 'False'
add_vocab: 'False'
all_gather: 'true'
batch_size_training: '16'
checkpoint_type: StateDictType.SHARDED_STATE_DICT
clipping_norm: '-1.0'
compare: 'False'
cpu_np_head: 'False'
ctx_proj_layer: 'False'
ctx_use_peft: 'False'
dataset: llava_qwen_selfrag_single_dataset
dist_checkpoint_folder: qwen1_5.full.prompt_llava.no_mask.no_context.without_sp_epoch2
dist_checkpoint_root_folder: /apdcephfs_sh2/share_300000800/user/kaixinma/amylee/RedPajama-Data/img/results
enable_fsdp: 'True'
freeze_ctx_encoder: 'False'
freeze_layers: 'False'
freeze_question_encoder: 'False'
from_hf: 'False'
fsdp_activation_checkpointing: 'True'
gamma: '0.85'
load_np_head: 'False'
low_cpu_fsdp: 'True'
lr: 2e-05
memory_bank_length: '0'
micro_batch_size: '2'
mixed_precision: 'True'
model_name: Qwen/Qwen2.5-1.5B
model_use_peft: 'False'
natural_form: 'False'
np_weight: '100.0'
num_epochs: '3'
num_freeze_layers: '1'
num_workers_dataloader: '1'
one_gpu: 'False'
optimizer: AdamW
output_dir: peft_checkpoint
peft_method: None
pure_bf16: 'False'
quantization: 'False'
question_proj_layer: 'False'
resume_epoch: '0'
ret_checkpoint_folder: ''
ret_first: 'False'
retriever: ''
run_validation: 'True'
save_model: 'True'
save_optimizer: 'False'
seed: '2'
sharding_strategy: ShardingStrategy.FULL_SHARD
single: 'False'
target_modules: ''
token_name: Qwen/Qwen2.5-1.5B
train: 'True'
use_fast_kernels: 'False'
use_fp16: 'False'
val_batch_size: '1'
weight_decay: '0.0'