See axolotl config

axolotl version: 0.4.1

adapter: lora
base_model: unsloth/Mistral-Nemo-Base-2407
bf16: auto
chat_template: llama3
dataset_prepared_path: null
datasets:
- data_files:
  - 64531df878c35053_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/64531df878c35053_train_data.json
  type:
    field_instruction: startphrase
    field_output: gold-ending
    format: '{instruction}'
    no_input_format: '{instruction}'
    system_format: '{system}'
    system_prompt: ''
debug: null
deepspeed: null
do_eval: true
early_stopping_patience: 3
eval_batch_size: 4
eval_max_new_tokens: 128
eval_steps: 500
evals_per_epoch: null
flash_attention: true
fp16: false
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 8
gradient_checkpointing: true
group_by_length: true
hub_model_id: lesso14/814730b9-f667-4be3-9804-8c23cfefe7d8
hub_repo: null
hub_strategy: checkpoint
hub_token: null
learning_rate: 0.000214
load_in_4bit: false
load_in_8bit: false
local_rank: null
logging_steps: 50
lora_alpha: 128
lora_dropout: 0.15
lora_fan_in_fan_out: null
lora_model_dir: null
lora_r: 64
lora_target_linear: true
lr_scheduler: cosine
max_grad_norm: 1.0
max_steps: 2500
micro_batch_size: 4
mlflow_experiment_name: /tmp/64531df878c35053_train_data.json
model_type: AutoModelForCausalLM
num_epochs: 10
optimizer: adamw_torch_fused
output_dir: miner_id_24
pad_to_sequence_len: true
resume_from_checkpoint: null
s2_attention: null
sample_packing: false
save_steps: 500
saves_per_epoch: null
seed: 140
sequence_len: 1024
strict: false
tf32: true
tokenizer_type: AutoTokenizer
train_on_inputs: false
trust_remote_code: true
val_set_size: 0.05
wandb_entity: null
wandb_mode: online
wandb_name: 6050c609-89f3-4b26-970d-a2c8da61f69d
wandb_project: 14a
wandb_run: your_name
wandb_runid: 6050c609-89f3-4b26-970d-a2c8da61f69d
warmup_steps: 100
weight_decay: 0.0
xformers_attention: null

814730b9-f667-4be3-9804-8c23cfefe7d8

This model is a fine-tuned version of unsloth/Mistral-Nemo-Base-2407 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 2.1109

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: 0.000214
• train_batch_size: 4
• eval_batch_size: 4
• seed: 140
• gradient_accumulation_steps: 8
• total_train_batch_size: 32
• 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: 100
• training_steps: 2500

Training results

Training Loss	Epoch	Step	Validation Loss
No log	0.0004	1	4.4125
18.8324	0.1816	500	2.6191
18.712	0.3632	1000	2.6649
18.1194	0.5447	1500	2.3648
17.5107	0.7263	2000	2.1313
16.8064	0.9079	2500	2.1109

Framework versions

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