Create llama_factory_config.yaml

llama_factory_config.yaml

@@ -0,0 +1,42 @@
+# Run on 8xh100 for 80/500 steps as loss was converged.
+
+### model
+model_name_or_path: NeverSleep/Lumimaid-v0.2-70B
+
+### method
+stage: sft
+do_train: true
+finetuning_type: lora
+lora_target: all
+deepspeed: examples/deepspeed/ds_z3_offload_config.json
+
+### dataset
+dataset: bolt_sharegpt
+template: llama3
+cutoff_len: 2048
+max_samples: 20000
+overwrite_cache: true
+preprocessing_num_workers: 16
+
+### output
+output_dir: saves/lumi_70b_bolt_lora_1e4
+logging_steps: 1
+save_steps: 40
+plot_loss: true
+overwrite_output_dir: true
+
+### train
+per_device_train_batch_size: 2
+gradient_accumulation_steps: 8
+learning_rate: 1.0e-4
+num_train_epochs: 3.0
+lr_scheduler_type: cosine
+warmup_ratio: 0.01
+bf16: true
+ddp_timeout: 180000000
+
+### eval
+val_size: 0.05
+per_device_eval_batch_size: 1
+eval_strategy: steps
+eval_steps: 40