See axolotl config

axolotl version: 0.9.2

base_model: meta-llama/Llama-3.1-8B-Instruct
model_type: LlamaForCausalLM
tokenizer_type: AutoTokenizer
gradient_accumulation_steps: 8
micro_batch_size: 1
num_epochs: 4
optimizer: adamw_bnb_8bit
lr_scheduler: cosine
learning_rate: 0.0001
load_in_8bit: true
load_in_4bit: false
adapter: lora
lora_model_dir: null
lora_r: 8
lora_alpha: 16
lora_dropout: 0.05
lora_target_modules:
- q_proj
- v_proj
- k_proj
datasets:
- path: /workspace/FinLoRA/data/train/finer_train_batched.jsonl
  type:
    system_prompt: ''
    field_system: system
    field_instruction: context
    field_output: target
    format: '[INST] {instruction} [/INST]'
    no_input_format: '[INST] {instruction} [/INST]'
dataset_prepared_path: null
val_set_size: 0.02
output_dir: /workspace/FinLoRA/lora/axolotl-output/finer_llama_3_1_8b_8bits_r8_rslora
peft_use_dora: false
peft_use_rslora: true
sequence_len: 4096
sample_packing: false
pad_to_sequence_len: false
wandb_project: finlora_models
wandb_entity: null
wandb_watch: gradients
wandb_name: finer_llama_3_1_8b_8bits_r8_rslora
wandb_log_model: 'false'
bf16: auto
tf32: false
gradient_checkpointing: true
resume_from_checkpoint: null
logging_steps: 500
flash_attention: false
deepspeed: deepspeed_configs/zero1.json
warmup_steps: 10
evals_per_epoch: 4
saves_per_epoch: 1
weight_decay: 0.0
special_tokens:
  pad_token: <|end_of_text|>
chat_template: llama3

workspace/FinLoRA/lora/axolotl-output/finer_llama_3_1_8b_8bits_r8_rslora

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

• Loss: 0.0328

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: 1
• eval_batch_size: 1
• seed: 42
• distributed_type: multi-GPU
• num_devices: 8
• gradient_accumulation_steps: 8
• total_train_batch_size: 64
• total_eval_batch_size: 8
• optimizer: Use OptimizerNames.ADAMW_BNB 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	0.0065	1	0.5433
No log	0.2545	39	0.0632
No log	0.5090	78	0.0461
No log	0.7635	117	0.0419
No log	1.0196	156	0.0404
No log	1.2741	195	0.0394
No log	1.5285	234	0.0376
No log	1.7830	273	0.0370
No log	2.0392	312	0.0366
No log	2.2936	351	0.0352
No log	2.5481	390	0.0338
No log	2.8026	429	0.0352
No log	3.0587	468	0.0338
0.048	3.3132	507	0.0331
0.048	3.5677	546	0.0339
0.048	3.8222	585	0.0328

Framework versions

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