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: 4
num_epochs: 1
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/xbrl_term_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/xbrl_term_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: xbrl_term_llama_3_1_8b_fp16_r8

workspace/FinLoRA/lora/axolotl-output/xbrl_term_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/xbrl_term_train.jsonl dataset. It achieves the following results on the evaluation set:

• Loss: 1.4293

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: 4
• eval_batch_size: 4
• seed: 42
• distributed_type: multi-GPU
• num_devices: 3
• gradient_accumulation_steps: 2
• total_train_batch_size: 24
• total_eval_batch_size: 12
• 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: 1.0

Training results

Training Loss	Epoch	Step	Validation Loss
No log	0.0042	1	2.5661
No log	0.2505	60	1.5752
No log	0.5010	120	1.4675
No log	0.7516	180	1.4293

Framework versions

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