mirror of
https://github.com/nomic-ai/gpt4all
synced 2024-11-10 01:10:35 +00:00
34 lines
612 B
YAML
34 lines
612 B
YAML
|
# model/tokenizer
|
||
|
|
model_name: "EleutherAI/gpt-j-6B"
|
||
|
|
tokenizer_name: "EleutherAI/gpt-j-6B"
|
||
|
|
gradient_checkpointing: true
|
||
|
|
save_name: "nomic-ai/gpt4all-gptj-multiturn-lr-aggressive"
|
||
|
|
|
||
|
|
# dataset
|
||
|
|
streaming: false
|
||
|
|
num_proc: 64
|
||
|
|
dataset_path: "data_multiplus"
|
||
|
|
max_length: 1024
|
||
|
|
batch_size: 8
|
||
|
|
|
||
|
|
# train dynamics
|
||
|
|
lr: 2.0e-5
|
||
|
|
min_lr: 0
|
||
|
|
weight_decay: 0.0
|
||
|
|
eval_every: 200
|
||
|
|
eval_steps: 105
|
||
|
|
save_every: 400
|
||
|
|
log_grads_every: 200
|
||
|
|
output_dir: "ckpts/gpt4all-gptj-full-multiturn-lr-aggreive"
|
||
|
|
checkpoint: null
|
||
|
|
lora: false
|
||
|
|
warmup_steps: 500
|
||
|
|
num_epochs: 4
|
||
|
|
|
||
|
|
# logging
|
||
|
|
wandb: true
|
||
|
|
wandb_entity: vicuna
|
||
|
|
wandb_project_name: vicuna
|
||
|
|
seed: 42
|
||
|
|
|