See axolotl config

axolotl version: 0.4.1

adapter: lora
base_model: fxmarty/really-tiny-falcon-testing
bf16: true
chat_template: llama3
dataset_prepared_path: null
datasets:
- data_files:
  - 0be0f088932e5895_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/0be0f088932e5895_train_data.json
  type:
    field_instruction: topic
    field_output: argument
    format: '{instruction}'
    no_input_format: '{instruction}'
    system_format: '{system}'
    system_prompt: ''
debug: null
deepspeed: null
early_stopping_patience: 2
eval_max_new_tokens: 128
eval_steps: 100
eval_table_size: null
flash_attention: false
fp16: null
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 8
gradient_checkpointing: true
group_by_length: false
hub_model_id: Alphatao/0afc88de-ce7b-4ef1-9fd5-9a93ba7b611c
hub_repo: null
hub_strategy: checkpoint
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: 32
lora_dropout: 0.05
lora_fan_in_fan_out: null
lora_model_dir: null
lora_r: 16
lora_target_linear: true
lora_target_modules:
- q_proj
- k_proj
- v_proj
- o_proj
lr_scheduler: cosine
max_grad_norm: 1.0
max_steps: 1980
micro_batch_size: 4
mlflow_experiment_name: /tmp/0be0f088932e5895_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: 0.04
wandb_entity: null
wandb_mode: online
wandb_name: 49cde80c-6112-4aaa-8659-47e6616bd83e
wandb_project: Gradients-On-Demand
wandb_run: your_name
wandb_runid: 49cde80c-6112-4aaa-8659-47e6616bd83e
warmup_steps: 10
weight_decay: 0.0
xformers_attention: null

0afc88de-ce7b-4ef1-9fd5-9a93ba7b611c

This model is a fine-tuned version of fxmarty/really-tiny-falcon-testing on the None dataset. It achieves the following results on the evaluation set:

• Loss: 10.9933

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: 1812

Training results

Training Loss	Epoch	Step	Validation Loss
88.7176	0.0011	1	11.0905
88.2832	0.1104	100	11.0265
88.1818	0.2208	200	11.0202
88.1169	0.3311	300	11.0114
88.0626	0.4415	400	11.0065
88.0765	0.5519	500	11.0027
88.1167	0.6623	600	11.0005
87.9885	0.7726	700	10.9988
88.0207	0.8830	800	10.9976
88.0782	0.9934	900	10.9964
87.9565	1.1038	1000	10.9956
87.9751	1.2141	1100	10.9948
87.9944	1.3245	1200	10.9944
87.976	1.4349	1300	10.9938
88.0434	1.5453	1400	10.9935
88.0379	1.6556	1500	10.9934
88.0042	1.7660	1600	10.9932
87.9932	1.8764	1700	10.9933
87.9932	1.9868	1800	10.9933

Framework versions

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