See axolotl config

axolotl version: 0.4.1

adapter: lora
base_model: unsloth/tinyllama-chat
bf16: true
chat_template: llama3
dataset_prepared_path: null
datasets:
- data_files:
  - 48d6404c58993270_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/48d6404c58993270_train_data.json
  type:
    field_input: premise
    field_instruction: task
    field_output: hypothesis
    format: '{instruction} {input}'
    no_input_format: '{instruction}'
    system_format: '{system}'
    system_prompt: ''
debug: null
deepspeed: null
early_stopping_patience: 2
early_stopping_threshold: 0.0001
eval_max_new_tokens: 128
eval_steps: 100
eval_table_size: null
flash_attention: true
fp16: null
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 8
gradient_checkpointing: true
group_by_length: false
hub_model_card: false
hub_model_id: romainnn/c019d9c7-30f4-4366-b300-2b9d39dd5be2
hub_repo: null
hub_strategy: end
hub_token: null
learning_rate: 0.0002
load_best_model_at_end: true
load_in_4bit: false
load_in_8bit: false
local_rank: null
logging_steps: 1
lora_alpha: 128
lora_dropout: 0.3
lora_fan_in_fan_out: null
lora_model_dir: null
lora_r: 64
lora_target_linear: true
lora_target_modules:
- q_proj
- k_proj
- v_proj
lr_scheduler: cosine
max_grad_norm: 1.0
max_steps: 3120
micro_batch_size: 4
mlflow_experiment_name: /tmp/48d6404c58993270_train_data.json
model_type: AutoModelForCausalLM
num_epochs: 2
optimizer: adamw_bnb_8bit
output_dir: miner_id_24
pad_to_sequence_len: true
resume_from_checkpoint: null
s2_attention: null
sample_packing: false
save_steps: 100
sequence_len: 2048
strict: false
tf32: true
tokenizer_type: AutoTokenizer
train_on_inputs: false
trust_remote_code: true
val_set_size: 0.024827325948031442
wandb_entity: null
wandb_mode: online
wandb_name: 4c107edc-0a51-4672-8be5-231e3b737755
wandb_project: Gradients-On-Demand
wandb_run: your_name
wandb_runid: 4c107edc-0a51-4672-8be5-231e3b737755
warmup_steps: 10
weight_decay: 0.0
xformers_attention: null

c019d9c7-30f4-4366-b300-2b9d39dd5be2

This model is a fine-tuned version of unsloth/tinyllama-chat on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.2358

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: 8
• 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: 3120

Training results

Training Loss	Epoch	Step	Validation Loss
3.8324	0.0002	1	3.8251
0.461	0.0163	100	0.3375
0.2229	0.0326	200	0.2815
0.3056	0.0489	300	0.2645
0.352	0.0652	400	0.2520
0.24	0.0815	500	0.2357
0.3087	0.0978	600	0.2358
0.2714	0.1141	700	0.2358

Framework versions

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