| base_model: /workspace/models/Mixtral-8x7B-v0.1 | |
| model_type: MixtralForCausalLM | |
| tokenizer_type: LlamaTokenizer | |
| is_mistral_derived_model: false | |
| 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: teknium/openhermes | |
| type: alpaca | |
| prompt_style: chatml | |
| dataset_prepared_path: | |
| val_set_size: 0 | |
| output_dir: /workspace/dolphin-2.6-mixtral-8x7b | |
| resume_from_checkpoint: | |
| hf_use_auth_token: | |
| adapter: qlora | |
| lora_model_dir: | |
| sequence_len: 16384 | |
| sample_packing: true | |
| pad_to_sequence_len: true | |
| lora_r: 32 | |
| lora_alpha: 16 | |
| lora_dropout: 0.05 | |
| lora_target_modules: | |
| - q_proj | |
| - k_proj | |
| - v_proj | |
| - o_proj | |
| - w1 | |
| - w2 | |
| - w3 | |
| lora_target_linear: | |
| lora_fan_in_fan_out: | |
| lora_modules_to_save: | |
| - embed_tokens | |
| - lm_head | |
| wandb_project: dolphin | |
| wandb_entity: | |
| wandb_watch: | |
| wandb_run_id: | |
| wandb_log_model: | |
| gradient_accumulation_steps: 2 | |
| micro_batch_size: 3 | |
| num_epochs: 3 | |
| optimizer: paged_adamw_8bit | |
| lr_scheduler: cosine | |
| learning_rate: 0.0002 | |
| model_config: | |
| output_router_logits: true | |
| train_on_inputs: | |
| group_by_length: false | |
| bf16: true | |
| fp16: false | |
| tf32: false | |
| gradient_checkpointing: true | |
| early_stopping_patience: | |
| local_rank: | |
| logging_steps: 1 | |
| xformers_attention: | |
| flash_attention: true | |
| warmup_steps: 100 | |
| eval_steps: | |
| save_steps: 0.01 | |
| save_total_limit: 2 | |
| save_safetensors: false | |
| eval_sample_packing: | |
| debug: | |
| deepspeed: deepspeed/zero2.json | |
| weight_decay: 0.05 | |
| fsdp: | |
| fsdp_config: | |
| special_tokens: | |
| eos_token: "<|im_end|>" | |
| tokens: | |
| - "<|im_start|>" | |
| trust_remote_code: true |