adapter: qlora
auto_resume_from_checkpoints: true
base_model: huggyllama/llama-7b
bf16: auto
chat_template: llama3
dataset_prepared_path: null
dataset_processes: 6
datasets:
- data_files:
  - 2def6f0bc8094c63_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/2def6f0bc8094c63_train_data.json
  type:
    field_instruction: instruction
    field_output: output
    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
evals_per_epoch: null
flash_attention: true
fp16: false
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 4
gradient_checkpointing: true
group_by_length: false
hub_model_id: error577/d3f85da4-8048-47c8-ad48-5fbd9a8a70fc
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: 64
lora_dropout: 0.1
lora_fan_in_fan_out: null
lora_model_dir: null
lora_r: 32
lora_target_linear: true
lr_scheduler: cosine
max_grad_norm: 1.0
max_steps: null
micro_batch_size: 2
mlflow_experiment_name: /tmp/2def6f0bc8094c63_train_data.json
model_type: AutoModelForCausalLM
num_epochs: 3
optimizer: adamw_torch_4bit
output_dir: miner_id_24
pad_to_sequence_len: true
resume_from_checkpoint: null
s2_attention: null
sample_packing: false
save_steps: 200
sequence_len: 512
special_tokens:
  pad_token: </s>
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: 776f445d-12cf-45fd-afe3-789f62d99645
wandb_project: Gradients-On-Demand
wandb_run: your_name
wandb_runid: 776f445d-12cf-45fd-afe3-789f62d99645
warmup_steps: 30
weight_decay: 0.0
xformers_attention: null

d3f85da4-8048-47c8-ad48-5fbd9a8a70fc

This model is a fine-tuned version of huggyllama/llama-7b on the None dataset. It achieves the following results on the evaluation set:

Loss: 0.3508

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
gradient_accumulation_steps: 4
total_train_batch_size: 8
optimizer: Use OptimizerNames.ADAMW_TORCH_4BIT 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: 30
num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss
0.7775	0.0001	1	0.8330
0.5424	0.0156	200	0.4220
0.612	0.0312	400	0.4068
0.3191	0.0467	600	0.4012
0.5935	0.0623	800	0.3815
0.4242	0.0779	1000	0.3739
0.4253	0.0935	1200	0.3748
0.4555	0.1091	1400	0.3690
0.2908	0.1247	1600	0.3673
0.3271	0.1402	1800	0.3647
0.3553	0.1558	2000	0.3655
0.274	0.1714	2200	0.3612
0.3973	0.1870	2400	0.3605
0.2867	0.2026	2600	0.3523
0.2378	0.2181	2800	0.3508
0.5015	0.2337	3000	0.3513
0.4597	0.2493	3200	0.3510
0.3814	0.2649	3400	0.3508

Framework versions

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