See axolotl config

axolotl version: 0.4.1

adapter: lora
base_model: DeepMount00/Llama-3-8b-Ita
bf16: true
chat_template: llama3
datasets:
- data_files:
  - 2db76e6fdf792dae_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/2db76e6fdf792dae_train_data.json
  type:
    field_input: text
    field_instruction: poem
    field_output: poet
    format: '{instruction} {input}'
    no_input_format: '{instruction}'
    system_format: '{system}'
    system_prompt: ''
debug: null
deepspeed: null
early_stopping_patience: null
eval_max_new_tokens: 128
eval_table_size: null
evals_per_epoch: 4
flash_attention: false
fp16: false
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 2
gradient_checkpointing: true
group_by_length: false
hub_model_id: lesso02/0398c169-3e39-4e2b-9a9e-c09d22e7b583
hub_repo: null
hub_strategy: checkpoint
hub_token: null
learning_rate: 1.0e-05
load_in_4bit: false
load_in_8bit: false
local_rank: null
logging_steps: 1
lora_alpha: 32
lora_dropout: 0.05
lora_fan_in_fan_out: null
lora_model_dir: null
lora_r: 16
lora_target_linear: true
lr_scheduler: cosine
max_memory:
  0: 70GiB
max_steps: 30
micro_batch_size: 4
mlflow_experiment_name: /tmp/2db76e6fdf792dae_train_data.json
model_type: AutoModelForCausalLM
num_epochs: 2
optimizer: adamw_torch
output_dir: miner_id_24
pad_to_sequence_len: true
resume_from_checkpoint: null
s2_attention: null
sample_packing: false
save_steps: 20
save_strategy: steps
sequence_len: 1024
special_tokens:
  pad_token: <|eot_id|>
strict: false
tf32: false
tokenizer_type: AutoTokenizer
train_on_inputs: false
trust_remote_code: true
val_set_size: 0.05
wandb_entity: null
wandb_mode: online
wandb_name: 0398c169-3e39-4e2b-9a9e-c09d22e7b583
wandb_project: Gradients-On-Demand
wandb_run: your_name
wandb_runid: 0398c169-3e39-4e2b-9a9e-c09d22e7b583
warmup_steps: 5
weight_decay: 0.01
xformers_attention: false

0398c169-3e39-4e2b-9a9e-c09d22e7b583

This model is a fine-tuned version of DeepMount00/Llama-3-8b-Ita on the None dataset. It achieves the following results on the evaluation set:

• Loss: 6.8881

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: 1e-05
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 8
• optimizer: Use OptimizerNames.ADAMW_TORCH 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: 5
• training_steps: 30

Training results

Training Loss	Epoch	Step	Validation Loss
8.409	0.0001	1	7.9600
7.5361	0.0003	4	7.9469
7.2974	0.0006	8	7.8295
7.4862	0.0009	12	7.6021
7.9965	0.0011	16	7.3008
7.5826	0.0014	20	7.0639
7.6529	0.0017	24	6.9416
5.7908	0.0020	28	6.8881

Framework versions

• PEFT 0.13.2
• Transformers 4.46.0
• Pytorch 2.5.0+cu124
• Datasets 3.0.1
• Tokenizers 0.20.1