adapter: qlora
base_model: EleutherAI/pythia-410m-deduped
bf16: auto
chat_template: llama3
dataset_prepared_path: null
datasets:
- data_files:
  - 11866109baf7431c_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/11866109baf7431c_train_data.json
  type:
    field_input: Chapter
    field_instruction: Book Name
    field_output: Chunk
    format: '{instruction} {input}'
    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: 16
gradient_checkpointing: true
group_by_length: true
hub_model_id: error577/e74610a7-8302-49f8-852a-453d0769bacc
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: cosine
micro_batch_size: 4
mlflow_experiment_name: /tmp/11866109baf7431c_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
resume_from_checkpoint: null
s2_attention: null
sample_packing: false
save_steps: 200
sequence_len: 1024
special_tokens:
  pad_token: <|endoftext|>
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: 22419f01-216c-4f29-9237-04f8efc9e9b9
wandb_project: Gradients-On-Demand
wandb_run: your_name
wandb_runid: 22419f01-216c-4f29-9237-04f8efc9e9b9
warmup_steps: 10
weight_decay: 0.0
xformers_attention: null

e74610a7-8302-49f8-852a-453d0769bacc

This model is a fine-tuned version of EleutherAI/pythia-410m-deduped on the None dataset. It achieves the following results on the evaluation set:

Loss: 2.9558

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: 16
total_train_batch_size: 64
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: cosine
lr_scheduler_warmup_steps: 10
num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss
60.3727	0.0018	1	3.5860
53.9221	0.3581	200	3.1653
52.424	0.7163	400	3.0718
47.3986	1.0744	600	3.0114
46.1675	1.4326	800	2.9953
47.8651	1.7907	1000	2.9769
47.7323	2.1489	1200	2.9623
46.0662	2.5070	1400	2.9582
45.702	2.8651	1600	2.9558

Framework versions

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