See axolotl config

axolotl version: 0.9.1.post1

base_model: meta-llama/Llama-3.1-8B-Instruct
model_type: LlamaForCausalLM
tokenizer_type: AutoTokenizer
gradient_accumulation_steps: 2
micro_batch_size: 8
num_epochs: 4
learning_rate: 0.0001
optimizer: adamw_torch_fused
lr_scheduler: cosine
load_in_8bit: false
load_in_4bit: false
adapter: lora
lora_r: 8
lora_alpha: 16
lora_dropout: 0.05
lora_target_modules:
- q_proj
- k_proj
- v_proj
datasets:
- path: /workspace/FinLoRA/data/train/headline_train.jsonl
  type:
    field_instruction: context
    field_output: target
    format: '[INST] {instruction} [/INST]'
    no_input_format: '[INST] {instruction} [/INST]'
val_set_size: 0.02
output_dir: /workspace/FinLoRA/lora/axolotl-output/headline_llama_3_1_8b_fp16_r8
sequence_len: 4096
gradient_checkpointing: true
logging_steps: 500
warmup_steps: 10
evals_per_epoch: 4
saves_per_epoch: 1
weight_decay: 0.0
special_tokens:
  pad_token: <|end_of_text|>
deepspeed: deepspeed_configs/zero1.json
bf16: auto
tf32: false
chat_template: llama3
wandb_name: headline_llama_3_1_8b_fp16_r8

workspace/FinLoRA/lora/axolotl-output/headline_llama_3_1_8b_fp16_r8

This model is a fine-tuned version of meta-llama/Llama-3.1-8B-Instruct on the /workspace/FinLoRA/data/train/headline_train.jsonl dataset. It achieves the following results on the evaluation set:

• Loss: 0.0465

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: 0.0001
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• distributed_type: multi-GPU
• num_devices: 3
• gradient_accumulation_steps: 2
• total_train_batch_size: 48
• total_eval_batch_size: 24
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 10
• num_epochs: 4.0

Training results

Training Loss	Epoch	Step	Validation Loss
No log	0.0006	1	6.7673
No log	0.2504	420	0.0517
0.2063	0.5007	840	0.0496
0.0491	0.7511	1260	0.0389
0.0424	1.0012	1680	0.0421
0.0363	1.2516	2100	0.0386
0.035	1.5019	2520	0.0441
0.035	1.7523	2940	0.0452
0.0321	2.0024	3360	0.0408
0.031	2.2528	3780	0.0442
0.0249	2.5031	4200	0.0414
0.0263	2.7535	4620	0.0404
0.0245	3.0036	5040	0.0420
0.0245	3.2539	5460	0.0484
0.0183	3.5043	5880	0.0483
0.0164	3.7547	6300	0.0465

Framework versions

• PEFT 0.15.2
• Transformers 4.51.3
• Pytorch 2.8.0.dev20250319+cu128
• Datasets 3.5.1
• Tokenizers 0.21.1