See axolotl config

axolotl version: 0.13.0.dev0

adapter: lora
base_model: unsloth/Qwen2.5-Math-1.5B
bf16: true
chat_template: llama3
dataset_prepared_path: null
datasets:
- data_files:
  - e07d27d53347e6ce_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/
  type:
    field_instruction: instruct
    field_output: output
    format: '{instruction}'
    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: true
fp16: null
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 4
gradient_checkpointing: true
group_by_length: false
hub_model_id: eason668/6481ff2c-347c-49af-b344-1cb71fd65aaa
hub_private_repo: false
learning_rate: 0.0002
load_in_4bit: false
load_in_8bit: false
local_rank: null
logging_steps: 1
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
max_steps: 210
micro_batch_size: 2
mlflow_experiment_name: /tmp/e07d27d53347e6ce_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_only_model: false
save_safetensors: true
save_steps: 21
save_strategy: steps
save_total_limit: 4
sequence_len: 2048
strict: false
tf32: false
tokenizer_max_length: 2048
tokenizer_truncation: true
tokenizer_type: AutoTokenizer
train_on_inputs: false
trust_remote_code: true
val_set_size: 0.1
wandb_entity: null
wandb_mode: online
wandb_project: Gradients-On-Demand
wandb_run: 6481ff2c-347c-49af-b344-1cb71fd65aaa
wandb_runid: 6481ff2c-347c-49af-b344-1cb71fd65aaa
warmup_steps: 10
weight_decay: 0.01
xformers_attention: null

6481ff2c-347c-49af-b344-1cb71fd65aaa

This model is a fine-tuned version of unsloth/Qwen2.5-Math-1.5B on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.2973
• Memory/max Mem Active(gib): 10.37
• Memory/max Mem Allocated(gib): 10.37
• Memory/device Mem Reserved(gib): 12.16

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: 2
• eval_batch_size: 2
• seed: 42
• distributed_type: multi-GPU
• num_devices: 8
• gradient_accumulation_steps: 4
• total_train_batch_size: 64
• total_eval_batch_size: 16
• 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
• training_steps: 210

Training results

Training Loss	Epoch	Step	Validation Loss	Mem Active(gib)	Mem Allocated(gib)	Mem Reserved(gib)
No log	0	0	1.8752	8.8	8.8	9.29
1.1156	0.0139	53	1.4004	10.37	10.37	12.09
1.0629	0.0277	106	1.3220	10.37	10.37	12.16
1.0489	0.0416	159	1.2973	10.37	10.37	12.16

Framework versions

• PEFT 0.17.0
• Transformers 4.55.2
• Pytorch 2.7.1+cu126
• Datasets 4.0.0
• Tokenizers 0.21.4