image/png

bophades-mistral-truthy-DPO-7B

bophades-v2-mistral-7B finetuned on jondurbin/truthy-dpo-v0.1.

Method

Finetuned using an A100 on Google Colab. πŸ™

Fine-tune a Mistral-7b model with Direct Preference Optimization - Maxime Labonne

Configuration

LoRA, model, and training settings:

# LoRA configuration
peft_config = LoraConfig(
    r=16,
    lora_alpha=16,
    lora_dropout=0.05,
    bias="none",
    task_type="CAUSAL_LM",
    target_modules=['k_proj', 'gate_proj', 'v_proj', 'up_proj', 'q_proj', 'o_proj', 'down_proj']
)

# Model to fine-tune
model = AutoModelForCausalLM.from_pretrained(
    model_name,
    torch_dtype=torch.bfloat16,
    load_in_4bit=True
)
model.config.use_cache = False

# Reference model
ref_model = AutoModelForCausalLM.from_pretrained(
    model_name,
    torch_dtype=torch.bfloat16,
    load_in_4bit=True
)

# Training arguments
training_args = TrainingArguments(
    per_device_train_batch_size=2,
    gradient_accumulation_steps=2,
    gradient_checkpointing=True,
    learning_rate=2e-5,
    lr_scheduler_type="cosine",
    max_steps=420,
    save_strategy="no",
    logging_steps=1,
    output_dir=new_model,
    optim="paged_adamw_32bit",
    warmup_steps=100,
    bf16=True,
    report_to="wandb",
)

# Create DPO trainer
dpo_trainer = DPOTrainer(
    model,
    ref_model,
    args=training_args,
    train_dataset=dataset,
    tokenizer=tokenizer,
    peft_config=peft_config,
    beta=0.1,
    max_prompt_length=1024,
    max_length=1536,
    force_use_ref_model=True
)

# Fine-tune model with DPO
dpo_trainer.train()
Downloads last month
193
Safetensors
Model size
7.24B params
Tensor type
BF16
Β·
Inference Providers NEW
This model is not currently available via any of the supported Inference Providers.

Model tree for nbeerbower/bophades-mistral-truthy-DPO-7B

Finetuned
(2)
this model
Merges
3 models
Quantizations
3 models

Dataset used to train nbeerbower/bophades-mistral-truthy-DPO-7B

Spaces using nbeerbower/bophades-mistral-truthy-DPO-7B 6