1base_model: Aratako/Llama-Gemma-2-27b-CPO_SimPO-iter2
2model_type: AutoModelForCausalLM
3tokenizer_type: AutoTokenizer
4
5hub_model_id: Aratako/fft-orpo-iterative-iter3
6hub_strategy: "end"
7push_dataset_to_hub:
8hf_use_auth_token: true
9
10plugins:
11 - axolotl.integrations.liger.LigerPlugin
12liger_cross_entropy: false
13liger_rope: true
14liger_rms_norm: true
15liger_swiglu: true
16liger_fused_linear_cross_entropy: true
17
18load_in_8bit: false
19load_in_4bit: false
20strict: false
21
22chat_template: tokenizer_default
23rl: orpo
24orpo_alpha: 0.1
25max_prompt_len: 512
26max_length: 2560
27
28
29datasets:
30 - path: Aratako/iterative-dpo-data-for-ORPO-iter3
31 type: chat_template.argilla
32 train_on_split: train
33
34
35shuffle_merged_datasets: true
36dataset_prepared_path: /workspace/data/fft-orpo-iterative-iter3-data
37output_dir: /workspace/data/27b-fft-orpo-iterative-iter3
38
39sequence_len: 2560
40sample_packing: false
41eval_sample_packing: false
42pad_to_sequence_len: true
43
44adapter:
45lora_model_dir:
46lora_r:
47lora_alpha:
48lora_dropout:
49lora_target_linear:
50lora_fan_in_fan_out:
51
52wandb_project: 27b-fft
53wandb_entity: aratako-lm
54wandb_watch:
55wandb_name: orpo-iter3
56wandb_log_model:
57
58gradient_accumulation_steps: 16
59micro_batch_size: 1
60num_epochs: 1
61optimizer: paged_adamw_8bit
62lr_scheduler: cosine
63cosine_min_lr_ratio: 0.1
64learning_rate: 8e-7
65
66train_on_inputs: false
67group_by_length: false
68bf16: auto
69fp16:
70tf32: false
71
72gradient_checkpointing: true
73early_stopping_patience:
74auto_resume_from_checkpoints: true
75local_rank:
76logging_steps: 1
77xformers_attention:
78flash_attention: true
79
80save_strategy: steps
81save_steps: 100
82save_total_limit: 1
83
84warmup_steps: 20
85eval_steps:
86eval_batch_size:
87eval_table_size:
88eval_max_new_tokens:
89debug:
90deepspeed: /workspace/axolotl/deepspeed_configs/zero3_bf16.json
91weight_decay: 0.01
92fsdp:
93fsdp_config:
94special_tokens:
95 pad_token: <pad>
