adapter: lora
base_model: Qwen/Qwen2.5-7B-Instruct
bf16: true
chat_template: llama3
dataset_prepared_path: null
datasets:
- data_files:
  - 346945de8cfb8df4_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/bfa535e6-25e7-4009-b06c-f6dc400fc5d2
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: 2010
micro_batch_size: 2
mlflow_experiment_name: /tmp/346945de8cfb8df4_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
saves_per_epoch: 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: bfa535e6-25e7-4009-b06c-f6dc400fc5d2
wandb_runid: bfa535e6-25e7-4009-b06c-f6dc400fc5d2
warmup_steps: 10
weight_decay: 0.01
xformers_attention: null

bfa535e6-25e7-4009-b06c-f6dc400fc5d2

This model is a fine-tuned version of Qwen/Qwen2.5-7B-Instruct on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 0.1401
Memory/max Mem Active(gib): 22.3
Memory/max Mem Allocated(gib): 22.3
Memory/device Mem Reserved(gib): 24.72

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: 2010

Training results

Training Loss	Epoch	Step	Validation Loss	Mem Active(gib)	Mem Allocated(gib)	Mem Reserved(gib)
No log	0	0	1.2047	20.24	20.24	22.09
0.3	3.4956	503	0.4232	22.3	22.3	24.67
0.1868	6.9913	1006	0.2410	22.3	22.3	24.72
0.0775	10.4817	1509	0.1401	22.3	22.3	24.72

Framework versions

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