See axolotl config

axolotl version: 0.4.1

adapter: lora
base_model: trl-internal-testing/tiny-random-LlamaForCausalLM
bf16: auto
chat_template: llama3
dataset_prepared_path: null
datasets:
- data_files:
  - 52db16c5f89d609e_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/52db16c5f89d609e_train_data.json
  type:
    field_instruction: context
    field_output: question
    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
gradient_clipping: 1.0
group_by_length: false
hub_model_id: ardaspear/455695ae-894f-474b-b576-ae26015b1f01
hub_repo: null
hub_strategy: checkpoint
hub_token: null
learning_rate: 5.0e-05
load_in_4bit: false
load_in_8bit: false
local_rank: 0
logging_steps: 3
lora_alpha: 32
lora_dropout: 0.05
lora_fan_in_fan_out: null
lora_model_dir: null
lora_r: 16
lora_target_linear: true
lr_scheduler: cosine
max_steps: 100
micro_batch_size: 8
mlflow_experiment_name: /tmp/52db16c5f89d609e_train_data.json
model_type: AutoModelForCausalLM
num_epochs: 3
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: 1024
strict: false
tf32: false
tokenizer_type: AutoTokenizer
train_on_inputs: false
trust_remote_code: true
val_set_size: 0.05
wandb_entity: techspear-hub
wandb_mode: online
wandb_name: 225cdcd8-cccf-4bb8-afa9-fbd712ba259a
wandb_project: Gradients-On-Five
wandb_run: your_name
wandb_runid: 225cdcd8-cccf-4bb8-afa9-fbd712ba259a
warmup_steps: 10
weight_decay: 0.01
xformers_attention: null

455695ae-894f-474b-b576-ae26015b1f01

This model is a fine-tuned version of trl-internal-testing/tiny-random-LlamaForCausalLM on the None dataset. It achieves the following results on the evaluation set:

• Loss: 10.3676

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: 5e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 4
• total_train_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: 10
• training_steps: 100

Training results

Training Loss	Epoch	Step	Validation Loss
No log	0.0005	1	10.3722
10.3715	0.0041	9	10.3719
10.3704	0.0082	18	10.3712
10.3719	0.0122	27	10.3705
10.3703	0.0163	36	10.3698
10.3682	0.0204	45	10.3692
10.3689	0.0245	54	10.3686
10.3683	0.0285	63	10.3682
10.3678	0.0326	72	10.3679
10.369	0.0367	81	10.3677
10.3661	0.0408	90	10.3676
10.3697	0.0449	99	10.3676

Framework versions

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