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:
  - daed85532ae01daa_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/daed85532ae01daa_train_data.json
  type:
    field_input: input
    field_instruction: instruction
    field_output: output
    format: '{instruction} {input}'
    no_input_format: '{instruction}'
    system_format: '{system}'
    system_prompt: ''
debug: null
device_map:
  ? ''
  : 0,1,2,3,4,5,6,7
early_stopping_patience: 2
eval_max_new_tokens: 128
eval_steps: 400
eval_table_size: null
flash_attention: true
gradient_accumulation_steps: 4
gradient_checkpointing: true
group_by_length: false
hub_model_id: Alphatao/35fb927f-d4dd-4902-b051-fdae57931bbf
hub_repo: null
hub_strategy: null
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
- down_proj
- up_proj
lr_scheduler: cosine
max_grad_norm: 1.0
max_steps: 14192
micro_batch_size: 2
mlflow_experiment_name: /tmp/daed85532ae01daa_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: 400
sequence_len: 2048
strict: false
tf32: true
tokenizer_type: AutoTokenizer
train_on_inputs: false
trust_remote_code: true
val_set_size: 0.05
wandb_entity: null
wandb_mode: online
wandb_name: d7c344bd-e406-41ee-ad84-5b799adb7e49
wandb_project: Gradients-On-Demand
wandb_run: your_name
wandb_runid: d7c344bd-e406-41ee-ad84-5b799adb7e49
warmup_steps: 10
weight_decay: 0.0
xformers_attention: null

35fb927f-d4dd-4902-b051-fdae57931bbf

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: 1.1144

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: 2
• eval_batch_size: 2
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 8
• 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: 14192

Training results

Training Loss	Epoch	Step	Validation Loss
1.6758	0.0001	1	1.8920
1.3241	0.0413	400	1.6002
1.7018	0.0826	800	1.5184
1.8543	0.1239	1200	1.4654
1.292	0.1652	1600	1.4236
1.5167	0.2065	2000	1.3890
1.3281	0.2478	2400	1.3660
1.376	0.2891	2800	1.3410
1.366	0.3304	3200	1.3217
1.3738	0.3717	3600	1.3012
1.4379	0.4130	4000	1.2837
1.2683	0.4543	4400	1.2698
1.2872	0.4956	4800	1.2540
1.2	0.5369	5200	1.2423
1.2733	0.5782	5600	1.2279
1.5039	0.6195	6000	1.2165
1.2798	0.6608	6400	1.2074
1.7364	0.7022	6800	1.1982
1.0885	0.7435	7200	1.1857
0.9772	0.7848	7600	1.1745
1.1519	0.8261	8000	1.1681
1.2283	0.8674	8400	1.1586
0.9801	0.9087	8800	1.1494
1.1179	0.9500	9200	1.1420
1.069	0.9913	9600	1.1351
1.1033	1.0326	10000	1.1340
0.4977	1.0739	10400	1.1299
0.9901	1.1152	10800	1.1269
1.3081	1.1565	11200	1.1241
0.985	1.1978	11600	1.1208
1.088	1.2391	12000	1.1191
1.1772	1.2804	12400	1.1167
1.2013	1.3217	12800	1.1155
1.0287	1.3630	13200	1.1150
0.9724	1.4043	13600	1.1145
0.9901	1.4456	14000	1.1144

Framework versions

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