adapter: lora
base_model: dunzhang/stella_en_1.5B_v5
bf16: auto
chat_template: llama3
dataset_prepared_path: null
datasets:
- data_files:
  - 22584c0b5bb42524_train_data.json
  ds_type: json
  field: prompt
  path: /workspace/input_data/
  split: train
  type: completion
debug: null
deepspeed: null
early_stopping_patience: null
eval_max_new_tokens: 128
eval_table_size: null
evals_per_epoch: 4
flash_attention: false
fp16: null
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 4
gradient_checkpointing: false
group_by_length: false
hub_model_id: hasdal/13adf98e-4bce-44dd-acd0-a45874111f17
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: 10
micro_batch_size: 4
mlflow_experiment_name: /tmp/22584c0b5bb42524_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: 512
strict: false
tf32: false
tokenizer_type: AutoTokenizer
train_on_inputs: false
trl:
  beta: 0.04
  max_completion_length: 256
  num_generations: 4
  reward_funcs:
  - rewards_681ee7ca-6799-4368-91b5-4ddf6427dbcb.reward_reasoning_keywords
  reward_weights:
  - 7.7560504310276
  use_vllm: false
trust_remote_code: true
val_set_size: 0.05
wandb_entity: null
wandb_mode: offline
wandb_name: 681ee7ca-6799-4368-91b5-4ddf6427dbcb
wandb_project: Gradients-On-Demand
wandb_run: your_name
wandb_runid: 681ee7ca-6799-4368-91b5-4ddf6427dbcb
warmup_steps: 10
weight_decay: 0.0
xformers_attention: null

13adf98e-4bce-44dd-acd0-a45874111f17

This model is a fine-tuned version of dunzhang/stella_en_1.5B_v5 on the None dataset. It achieves the following results on the evaluation set:

Loss: nan

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: 4
total_train_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: 10

Training results

Epoch	Step	Validation Loss
0.0003	1	nan
0.0010	3	nan
0.0020	6	nan
0.0030	9	nan

Framework versions

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