| base_model: microsoft/phi-2 |
| model_type: AutoModelForCausalLM |
| tokenizer_type: AutoTokenizer |
| is_llama_derived_model: false |
| trust_remote_code: true |
|
|
| load_in_8bit: false |
| load_in_4bit: true |
| strict: false |
|
|
| datasets: |
| - path: /workspace/datasets/dolphin/dolphin201.jsonl |
| type: alpaca_w_system.load_open_orca_chatml |
| - path: /workspace/datasets/dolphin-coder-translate.jsonl |
| type: alpaca_w_system.load_open_orca_chatml |
| - path: /workspace/datasets/dolphin-coder-codegen.jsonl |
| type: alpaca_w_system.load_open_orca_chatml |
| - path: /workspace/datasets/data-evol_instruct-decontaminated-converted.jsonl |
| type: alpaca_w_system.load_open_orca_chatml |
| - path: /workspace/datasets/data-oss_instruct-decontaminated-converted.jsonl |
| type: alpaca_w_system.load_open_orca_chatml |
| - path: /workspace/datasets/CapybaraPure_Decontaminated-converted.jsonl |
| type: sharegpt |
| conversation: chatml |
| - path: /workspace/datasets/not_samantha_norefusals.jsonl |
| type: sharegpt |
| conversation: chatml |
| - path: /workspace/datasets/openhermes.json |
| type: alpaca |
| prompt_style: chatml |
|
|
| dataset_prepared_path: larp |
| val_set_size: 0.05 |
| output_dir: /workspace/dolphin-2.6-phi-2/ |
|
|
| sequence_len: 2048 |
| sample_packing: true |
| pad_to_sequence_len: true |
|
|
| adapter: qlora |
| lora_model_dir: |
| lora_r: 64 |
| lora_alpha: 32 |
| lora_dropout: 0.05 |
| lora_target_linear: true |
| lora_fan_in_fan_out: |
| lora_modules_to_save: |
| - embed_tokens |
| - lm_head |
|
|
| wandb_project: dolphin |
| wandb_entity: |
| wandb_watch: |
| wandb_name: |
| wandb_log_model: |
|
|
| gradient_accumulation_steps: 16 |
| micro_batch_size: 1 |
| num_epochs: 4 |
| optimizer: paged_adamw_8bit |
| adam_beta1: 0.9 |
| adam_beta2: 0.999 |
| adam_epsilon: 0.00001 |
| max_grad_norm: 1000.0 |
| lr_scheduler: cosine |
| learning_rate: 2e-4 |
|
|
| train_on_inputs: false |
| group_by_length: |
| bf16: false |
| fp16: true |
| tf32: false |
|
|
| gradient_checkpointing: |
| early_stopping_patience: |
| resume_from_checkpoint: |
| local_rank: |
| logging_steps: 1 |
| xformers_attention: |
| flash_attention: true |
|
|
| warmup_steps: 5 |
| evals_per_epoch: 0 |
| save_steps: 0.01 |
| save_safetensors: false |
| save_total_limit: 2 |
| debug: |
| deepspeed: deepspeed/zero2.json |
| weight_decay: 0.01 |
| fsdp: |
| fsdp_config: |
| resize_token_embeddings_to_32x: true |
| special_tokens: |
| eos_token: "<|im_end|>" |
| pad_token: "<|endoftext|>" |
| tokens: |
| - "<|im_start|>" |
|
|
