sft_retrieval_utility.json

{
    "train_dataset_settings": {
        "sources": [
            {
                "name": "chat_test",
                "records_path": "tests/fixtures/datasets/chat/train_chat.jsonl",
                "sample_rate": 1
            }
        ],
        "prompt_template": {
            "role_tag_mapping": {
                "bot": "<bot>",
                "user": "<user>",
                "system": "<system>"
            },
            "prefix_template": "<RS>{role}",
            "suffix_template": "</RS>"
        },
        "dataset_type": "chat",
        "max_tokens_count": 150,
        "only_answer_loss": true
    },
    "val_dataset_settings": {
        "sources": [
            {
                "name": "chat_test",
                "records_path": "tests/fixtures/datasets/chat/train_chat.jsonl",
                "sample_rate": 1
            }
        ],
        "prompt_template": {
            "role_tag_mapping": {
                "bot": "<bot>",
                "user": "<user>",
                "system": "<system>"
            },
            "prefix_template": "<RS>{role}",
            "suffix_template": "</RS>"
        },
        "dataset_type": "chat",
        "max_tokens_count": 150,
        "only_answer_loss": true
    },
    "model_settings": {
        "model_path": "tests/fixtures/models/llama2_tiny",
        "model_type": "causal",
        "resize_token_embeddings": true,
        "transformers_settings": {
        },
        "peft_settings": {
            "r": 8,
            "lora_alpha": 16,
            "lora_dropout": 0.05,
            "target_modules": [
                "q_proj",
                "v_proj"
            ],
            "task_type": "CAUSAL_LM",
            "modules_to_save": ["embed_tokens", "lm_head"],
            "name": "LORA"
        },
        "embeddings_initialization_strategy": {
            "<RS>": "<s>",
            "</RS>": "</s>",
            "<bot>": "<unk>",
            "<user>": "<unk>",
            "<system>": "<unk>",
            "<doc_sep>": "<sep>"
        }
    },
    "cherry_pick_settings": {
        "generator_transformers_settings": {
            "num_beams": 3,
            "stop_strings": ["</RS>", "</s>"],
            "max_new_tokens": 8
        },
        "custom_generation_settings": {
            "skip_special_tokens": false
          },
        "dataset_settings": {
            "sources": [
                {
                    "name": "chat_test",
                    "records_path": "tests/fixtures/datasets/chat/train_chat.jsonl",
                    "sample_rate": 1.0
                }
            ],
            "prompt_template": {
                "role_tag_mapping": {
                    "bot": "<bot>",
                    "user": "<user>",
                    "system": "<system>"
                },
                "prefix_template": "<RS>{role}",
                "suffix_template": "</RS>"
            },
            "dataset_type": "chat",
            "max_tokens_count": 150,
            "only_answer_loss": true
        },
        "metric_settings": [
            {
                "type": "retrieval_utility", 
                "parameters": {
                    "need_average": [true,false]
                }
            }
        ]
    },
    "tokenizer_settings": {},
    "special_tokens_settings": {
        "bos_token": "<s>",
        "eos_token": "</s>",
        "pad_token": "<pad>"
    },
    "trainer_settings": {
        "evaluation_strategy": "steps",
        "per_device_train_batch_size": 1,
        "per_device_eval_batch_size": 1,
        "gradient_accumulation_steps": 1,
        "eval_steps": 1,
        "save_steps": 1,
        "logging_steps": 1,
        "learning_rate": 0.0003,
        "num_train_epochs": 2,
        "lr_scheduler_type": "cosine",
        "warmup_steps": 2,
        "fp16": false,
        "bf16": false,
        "optim": "adamw_torch",
        "save_total_limit": 1,
        "no_cuda": true
    },
    "logging_settings": {
        "project_name": "alignment",
        "run_name": "sft_retrieval_utility",
        "entity": "turbo-alignment"
      },
    "log_path": "test_sft_retrieval_utility_train_output"
}