Uploaded model
- Developed by: emessy
- License: apache-2.0
- Finetuned from model : unsloth/Meta-Llama-3.1-8B-bnb-4bit
This llama model was trained 2x faster with
Unsloth and Huggingface's TRL library.
Configure LoRA
lora_config = LoraConfig(
r=16,
lora_alpha=16,
target_modules=["q_proj", "k_proj", "v_proj", "o_proj"],
lora_dropout=0.05,
bias="none",
task_type="CAUSAL_LM"
)
Training arguments
training_args = TrainingArguments(
output_dir="./results",
num_train_epochs=5,
per_device_train_batch_size=4,
gradient_accumulation_steps=4,
learning_rate=2e-4,
fp16=True, # Use half-precision
logging_steps=10,
save_steps=50,
eval_steps=50,
)
