See axolotl config

axolotl version: 0.4.1

adapter: lora
auto_find_batch_size: true
base_model: unsloth/SmolLM2-360M
bf16: auto
chat_template: llama3
dataloader_num_workers: 12
dataset_prepared_path: null
datasets:
- data_files:
  - 283c4184083d47ae_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/283c4184083d47ae_train_data.json
  type:
    field_input: input
    field_instruction: instruction
    field_output: output
    format: '{instruction} {input}'
    no_input_format: '{instruction}'
    system_format: '{system}'
    system_prompt: ''
debug: null
deepspeed: null
early_stopping_patience: 3
early_stopping_threshold: 0.001
eval_max_new_tokens: 128
eval_steps: 40
flash_attention: false
fp16: null
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 2
gradient_checkpointing: false
group_by_length: false
hub_model_id: mrferr3t/9f105939-2de8-4946-b320-206ff7796263
hub_repo: null
hub_strategy: checkpoint
hub_token: null
learning_rate: 0.0003
load_in_4bit: false
load_in_8bit: false
local_rank: null
logging_steps: 100
lora_alpha: 16
lora_dropout: 0.05
lora_fan_in_fan_out: null
lora_model_dir: null
lora_r: 8
lora_target_linear: true
lr_scheduler: cosine
micro_batch_size: 32
mlflow_experiment_name: /tmp/283c4184083d47ae_train_data.json
model_type: AutoModelForCausalLM
num_epochs: 50
optimizer: adamw_bnb_8bit
output_dir: miner_id_24
pad_to_sequence_len: true
s2_attention: null
sample_packing: false
save_steps: 40
saves_per_epoch: 0
sequence_len: 512
strict: false
tf32: false
tokenizer_type: AutoTokenizer
train_on_inputs: false
trust_remote_code: true
val_set_size: 0.02
wandb_entity: null
wandb_mode: online
wandb_name: a0ab5165-7ac1-4405-b1f9-20e99af02244
wandb_project: Gradients-On-Demand
wandb_run: your_name
wandb_runid: a0ab5165-7ac1-4405-b1f9-20e99af02244
warmup_ratio: 0.05
weight_decay: 0.0
xformers_attention: null

9f105939-2de8-4946-b320-206ff7796263

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: 1.0826

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: 0.0003
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 64
• optimizer: Use adamw_bnb_8bit 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: 949
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss
No log	0.0013	1	1.6294
No log	0.0526	40	1.6257
No log	0.1053	80	1.5691
1.6207	0.1579	120	1.4022
1.6207	0.2105	160	1.3050
1.3431	0.2632	200	1.2492
1.3431	0.3158	240	1.2136
1.3431	0.3684	280	1.1941
1.2206	0.4211	320	1.1788
1.2206	0.4737	360	1.1667
1.1697	0.5263	400	1.1570
1.1697	0.5789	440	1.1514
1.1697	0.6316	480	1.1420
1.1405	0.6842	520	1.1379
1.1405	0.7368	560	1.1335
1.1177	0.7895	600	1.1281
1.1177	0.8421	640	1.1220
1.1177	0.8947	680	1.1179
1.1093	0.9474	720	1.1169
1.1093	1.0	760	1.1153
1.0961	1.0526	800	1.1116
1.0961	1.1053	840	1.1128
1.0961	1.1579	880	1.1117
1.0744	1.2105	920	1.1072
1.0744	1.2632	960	1.1064
1.0758	1.3158	1000	1.1033
1.0758	1.3684	1040	1.0984
1.0758	1.4211	1080	1.0990
1.0859	1.4737	1120	1.0951
1.0859	1.5263	1160	1.0921
1.0684	1.5789	1200	1.0927
1.0684	1.6316	1240	1.0883
1.0684	1.6842	1280	1.0849
1.0656	1.7368	1320	1.0838
1.0656	1.7895	1360	1.0833
1.0588	1.8421	1400	1.0800
1.0588	1.8947	1440	1.0806
1.0588	1.9474	1480	1.0766
1.0531	2.0	1520	1.0775
1.0531	2.0526	1560	1.0789
1.0087	2.1053	1600	1.0826

Framework versions

• PEFT 0.13.2
• Transformers 4.46.0
• Pytorch 2.3.1+cu121
• Datasets 3.0.1
• Tokenizers 0.20.1