43 lines
1005 B
YAML
43 lines
1005 B
YAML
base_model: EleutherAI/pythia-12b-deduped
|
|
base_model_ignore_patterns: pytorch* # prefer safetensors
|
|
# optionally might have model_type or tokenizer_type
|
|
model_type: GPTNeoXForCausalLM
|
|
tokenizer_type: AutoTokenizer
|
|
# Automatically upload checkpoint and final model to HF
|
|
# hub_model_id: username/custom_model_name
|
|
gptq: false
|
|
device_map: auto
|
|
datasets:
|
|
- path: vicgalle/alpaca-gpt4
|
|
type: alpaca
|
|
dataset_prepared_path:
|
|
val_set_size: 0.05
|
|
adapter:
|
|
lora_model_dir:
|
|
sequence_len: 2048
|
|
max_packed_sequence_len: 2048
|
|
lora_r: 64
|
|
lora_alpha: 32
|
|
lora_dropout: 0.0
|
|
lora_target_linear: true
|
|
lora_fan_in_fan_out: true # pythia/GPTNeoX lora specific
|
|
wandb_project:
|
|
wandb_entity:
|
|
wandb_watch:
|
|
wandb_name:
|
|
wandb_log_model:
|
|
output_dir: ./outputs/pythia-12b
|
|
gradient_accumulation_steps: 1
|
|
micro_batch_size: 1
|
|
num_epochs: 5
|
|
learning_rate: 0.00003
|
|
optimizer: adamw_bnb_8bit
|
|
lr_scheduler: cosine
|
|
bf16: false
|
|
fp16: false
|
|
float16: true
|
|
tf32: true
|
|
flash_optimum: true
|
|
resume_from_checkpoint:
|
|
gradient_checkpointing: true
|