See axolotl config

axolotl version: 0.4.1

adapter: lora
base_model: unsloth/SmolLM2-360M
bf16: true
chat_template: llama3
dataset_prepared_path: null
datasets:
- data_files:
  - e68e64a4250f7068_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/e68e64a4250f7068_train_data.json
  type:
    field_instruction: prompt
    field_output: rejected
    format: '{instruction}'
    no_input_format: '{instruction}'
    system_format: '{system}'
    system_prompt: ''
debug: null
deepspeed: null
early_stopping_patience: 30
eval_max_new_tokens: 128
eval_steps: 50
eval_table_size: null
flash_attention: false
fp16: null
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 16
gradient_checkpointing: true
group_by_length: false
hub_model_id: Romain-XV/cccc09a1-8d40-4f5c-a5a5-59ee7722a459
hub_repo: null
hub_strategy: checkpoint
hub_token: null
learning_rate: 0.0002
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: true
lora_model_dir: null
lora_r: 16
lora_target_linear: true
lora_target_modules:
- q_proj
- k_proj
- v_proj
lr_scheduler: cosine
micro_batch_size: 4
mlflow_experiment_name: /tmp/e68e64a4250f7068_train_data.json
model_type: AutoModelForCausalLM
num_epochs: 1
optimizer: adamw_bnb_8bit
output_dir: miner_id_24
pad_to_sequence_len: true
resume_from_checkpoint: null
s2_attention: null
sample_packing: false
save_steps: 100
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: 71c48c0b-3b94-4026-8e0e-05ac2bbcbe68
wandb_project: Gradients-On-Demand
wandb_run: your_name
wandb_runid: 71c48c0b-3b94-4026-8e0e-05ac2bbcbe68
warmup_steps: 10
weight_decay: 0.0
xformers_attention: null

cccc09a1-8d40-4f5c-a5a5-59ee7722a459

This model is a fine-tuned version of unsloth/SmolLM2-360M on the None dataset. It achieves the following results on the evaluation set:

• Loss: nan

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: 0.0002
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 16
• total_train_batch_size: 64
• 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: 1

Training results

Training Loss	Epoch	Step	Validation Loss
0.0	0.0022	1	nan
0.0	0.1090	50	nan
0.0	0.2180	100	nan
0.0	0.3270	150	nan
0.0	0.4360	200	nan
0.0	0.5450	250	nan
0.0	0.6540	300	nan
0.0	0.7630	350	nan
0.0	0.8721	400	nan
0.0	0.9811	450	nan

Framework versions

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