See axolotl config

axolotl version: 0.4.1

base_model: meta-llama/Meta-Llama-3-8B # same model you originally used
# Load your previously fine-tuned model as a PEFT adapter
peft_model: ahmedelgebaly/llama-3.1-8b-squadv2_e2
model_type: AutoModelForCausalLM
tokenizer_type: AutoTokenizer

load_in_8bit: false
load_in_4bit: true
strict: false

datasets:
  - path: ahmedelgebaly/SciQ_Alpaca
    type: alpaca
    split: train

test_datasets:
  - path: ahmedelgebaly/SciQ_Alpaca
    type: alpaca
    split: validation

dataset_prepared_path:
output_dir: ./outputs/qlora-out

adapter: qlora
lora_model_dir:

sequence_len: 4096
sample_packing: true
pad_to_sequence_len: true

lora_r: 32
lora_alpha: 16
lora_dropout: 0.05
lora_target_modules:
lora_target_linear: true
lora_fan_in_fan_out:

wandb_project: llama-3.1-8b-squadv2_SciQ_e2
wandb_entity:
wandb_watch:
wandb_name: llama-3.1-8b-squadv2-v0_SciQ_e2
wandb_log_model:

hub_model_id: ahmedelgebaly/llama-3.1-8b-squadv2_SciQ_e2

gradient_accumulation_steps: 4
micro_batch_size: 4
num_epochs: 2
optimizer: paged_adamw_32bit
lr_scheduler: cosine
learning_rate: 0.0002

train_on_inputs: false
group_by_length: false
bf16: auto
fp16:
tf32: false

gradient_checkpointing: true
early_stopping_patience:
resume_from_checkpoint:
local_rank:
logging_steps: 1
xformers_attention:
flash_attention: true

warmup_steps: 10
evals_per_epoch: 4
eval_table_size:
saves_per_epoch: 1
debug:
deepspeed:
weight_decay: 0.0
fsdp:
fsdp_config:
special_tokens:
  pad_token: "<|end_of_text|>"

llama-3.1-8b-squadv2_SciQ_e2

This model is a fine-tuned version of meta-llama/Meta-Llama-3-8B on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.9066

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 0.0002
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 16
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 10
• num_epochs: 2

Training results

Training Loss	Epoch	Step	Validation Loss
1.7866	0.0305	1	1.8420
1.1295	0.2443	8	1.0980
0.8408	0.4885	16	0.9650
0.8677	0.7328	24	0.9346
0.8605	0.9771	32	0.9223
0.8401	1.2137	40	0.9130
0.8089	1.4580	48	0.9084
0.8434	1.7023	56	0.9068
0.8224	1.9466	64	0.9066

Framework versions

• PEFT 0.13.2
• Transformers 4.45.2
• Pytorch 2.3.1+cu121
• Datasets 3.0.1
• Tokenizers 0.20.1