See axolotl config

axolotl version: 0.4.1

adapter: qlora
base_model: peft-internal-testing/tiny-dummy-qwen2
bf16: auto
chat_template: llama3
dataset_prepared_path: null
datasets:
- data_files:
  - cfeccc189e92571d_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/cfeccc189e92571d_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
early_stopping_patience: 3
eval_max_new_tokens: 128
eval_steps: 200
eval_table_size: null
flash_attention: true
fp16: null
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 64
gradient_checkpointing: true
group_by_length: true
hub_model_id: error577/4e03c89a-8b75-498e-b418-baeb044e5bd5
hub_repo: null
hub_strategy: checkpoint
hub_token: null
learning_rate: 0.0002
load_in_4bit: true
load_in_8bit: false
local_rank: null
logging_steps: 1
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: constant_with_warmup
micro_batch_size: 4
mlflow_experiment_name: /tmp/cfeccc189e92571d_train_data.json
model_type: AutoModelForCausalLM
num_epochs: 3
optimizer: adamw_torch_fused
output_dir: miner_id_24
pad_to_sequence_len: true
restore_best_weights: true
auto_resume_from_checkpoints: true
s2_attention: null
sample_packing: false
save_steps: 200
sequence_len: 1024
strict: false
tf32: false
tokenizer_type: AutoTokenizer
train_on_inputs: false
trust_remote_code: true
val_set_size: 0.005
wandb_entity: null
wandb_mode: online
wandb_name: 85f02ab3-c5bc-4ce0-94a6-aae2ba376e08
wandb_project: Gradients-On-Demand
wandb_run: your_name
wandb_runid: 85f02ab3-c5bc-4ce0-94a6-aae2ba376e08
warmup_steps: 10
weight_decay: 0.0
xformers_attention: null

4e03c89a-8b75-498e-b418-baeb044e5bd5

This model is a fine-tuned version of peft-internal-testing/tiny-dummy-qwen2 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 11.8905

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: 64
• total_train_batch_size: 256
• 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: constant_with_warmup
• lr_scheduler_warmup_steps: 10
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss
11.9295	0.0003	1	11.9293
11.8831	0.0695	200	11.8963
11.8873	0.1390	400	11.8939
11.8841	0.2085	600	11.8926
11.8756	0.2780	800	11.8919
11.8757	0.3475	1000	11.8918
11.8828	0.4170	1200	11.8915
11.884	0.4865	1400	11.8916
11.8848	0.5560	1600	11.8913
11.8737	0.6255	1800	11.8904
11.8768	0.6950	2000	11.8906
11.8752	0.7645	2200	11.8906
11.8774	0.8340	2400	11.8901
11.8759	0.9035	2600	11.8900
11.8754	0.9730	2800	11.8900
11.8156	1.0425	3000	11.8901
11.7092	1.1120	3200	11.8905

Framework versions

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