Clemspace's picture
Initial model upload
cb9e677
batch_size: 1
checkpoint: true
ckpt_freq: 100
ckpt_only_lora: false
data:
data: ''
eval_instruct_data: ''
instruct:
dynamic_chunk_fn_call: true
shuffle: true
instruct_data: /root/data/mol_instructions_train.jsonl
shuffle: false
eval_freq: 100
log_freq: 1
lora:
dropout: 0.0
enable: true
rank: 64
scaling: 2.0
max_norm: 1.0
max_steps: 500
mlflow:
experiment_name: null
tracking_uri: null
model_id_or_path: /root/mistral_models/7B-v0.3
no_ckpt: false
no_eval: true
num_ckpt_keep: 3
num_microbatches: 1
optim:
lr: 6.0e-05
pct_start: 0.05
weight_decay: 0.1
run_dir: /root/mistral-finetune/runs2
seed: 0
seq_len: 32768
wandb:
key: aaf77f83a4e316f6a8b47fa975ab6b5e73c7c8df
offline: false
project: CHEMISTral7b-ft
run_name: run1
world_size: 1