adapter: lora
attn_implementation: eager
base_model: tiiuae/falcon-7b
bf16: true
datasets:
- data_files:
  - 2377e89d05129d78_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/
  type:
    field_input: input
    field_instruction: instruct
    field_output: output
    field_system: None
    format: None
    no_input_format: None
    system_format: '{system}'
    system_prompt: None
eval_max_new_tokens: 128
evals_per_epoch: 4
flash_attention: false
fp16: false
gradient_accumulation_steps: 2
gradient_checkpointing: true
group_by_length: true
hf_upload_public: true
hf_upload_repo_type: model
hub_model_id: cpheemagazine/5d5fd3e7-48ed-431e-9cfc-a70654740094
learning_rate: 0.0002
load_in_4bit: false
logging_steps: 10
lora_alpha: 16
lora_dropout: 0.05
lora_fan_in_fan_out: false
lora_r: 8
lora_target_linear: true
lr_scheduler: cosine
max_steps: 1134
micro_batch_size: 4
mlflow_experiment_name: /tmp/2377e89d05129d78_train_data.json
optimizer: adamw_bnb_8bit
output_dir: miner_id_24
rl: null
sample_packing: true
save_steps: 170
sequence_len: 1024
special_tokens:
  pad_token: <|endoftext|>
tf32: true
tokenizer_type: AutoTokenizer
train_on_inputs: true
trl: null
trust_remote_code: false
use_flash_attention: false
wandb_name: 07464eb7-df94-4e74-8dc3-e49bf3912db7
wandb_project: Gradients-On-Demand
wandb_run: apriasmoro
wandb_runid: 07464eb7-df94-4e74-8dc3-e49bf3912db7
warmup_steps: 113
weight_decay: 0.02

5d5fd3e7-48ed-431e-9cfc-a70654740094

This model is a fine-tuned version of tiiuae/falcon-7b on an unknown dataset.

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
distributed_type: multi-GPU
num_devices: 8
gradient_accumulation_steps: 2
total_train_batch_size: 64
total_eval_batch_size: 32
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: 113
training_steps: 1134

Training results

Framework versions

PEFT 0.15.2
Transformers 4.51.3
Pytorch 2.5.1+cu124
Datasets 3.5.1
Tokenizers 0.21.1