See axolotl config

axolotl version: 0.4.1

adapter: lora
base_model: fxmarty/really-tiny-falcon-testing
bf16: auto
chat_template: llama3
dataset_prepared_path: null
datasets:
- data_files:
  - e87224c8eb065bc2_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/e87224c8eb065bc2_train_data.json
  type:
    field_input: intent
    field_instruction: instruction
    field_output: response
    format: '{instruction} {input}'
    no_input_format: '{instruction}'
    system_format: '{system}'
    system_prompt: ''
debug: null
deepspeed: null
do_eval: true
early_stopping_patience: 3
eval_batch_size: 4
eval_max_new_tokens: 128
eval_steps: 500
evals_per_epoch: null
flash_attention: false
fp16: false
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 8
gradient_checkpointing: true
group_by_length: true
hub_model_id: lesso14/418e6629-dcd3-424e-b3db-50c11d0d7c6b
hub_repo: null
hub_strategy: checkpoint
hub_token: null
learning_rate: 0.000214
load_in_4bit: false
load_in_8bit: false
local_rank: null
logging_steps: 50
lora_alpha: 128
lora_dropout: 0.15
lora_fan_in_fan_out: null
lora_model_dir: null
lora_r: 64
lora_target_linear: true
lr_scheduler: cosine
max_grad_norm: 1.0
max_steps: 25000
micro_batch_size: 4
mlflow_experiment_name: /tmp/e87224c8eb065bc2_train_data.json
model_type: AutoModelForCausalLM
num_epochs: 10
optimizer: adamw_torch_fused
output_dir: miner_id_24
pad_to_sequence_len: true
resume_from_checkpoint: null
s2_attention: null
sample_packing: false
save_steps: 500
saves_per_epoch: null
seed: 140
sequence_len: 1024
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: 1cd1f24c-cd92-4120-99af-dc7e5e1a2626
wandb_project: 14a
wandb_run: your_name
wandb_runid: 1cd1f24c-cd92-4120-99af-dc7e5e1a2626
warmup_steps: 100
weight_decay: 0.0
xformers_attention: null

418e6629-dcd3-424e-b3db-50c11d0d7c6b

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: 10.9416

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: 0.000214
• train_batch_size: 4
• eval_batch_size: 4
• seed: 140
• gradient_accumulation_steps: 8
• total_train_batch_size: 32
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED 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: 100
• training_steps: 25000

Training results

Training Loss	Epoch	Step	Validation Loss
No log	0.0001	1	11.0842
87.9206	0.0563	500	10.9867
87.8458	0.1127	1000	10.9740
87.8012	0.1690	1500	10.9673
87.7685	0.2254	2000	10.9628
87.7475	0.2817	2500	10.9602
87.7401	0.3381	3000	10.9575
87.7205	0.3944	3500	10.9554
87.6986	0.4508	4000	10.9532
87.6961	0.5071	4500	10.9514
87.677	0.5635	5000	10.9498
87.6867	0.6198	5500	10.9488
87.6773	0.6762	6000	10.9481
87.6696	0.7325	6500	10.9472
87.664	0.7889	7000	10.9466
87.6684	0.8452	7500	10.9463
87.6553	0.9015	8000	10.9458
87.6592	0.9579	8500	10.9454
87.6622	1.0143	9000	10.9448
87.6536	1.0706	9500	10.9444
87.6519	1.1270	10000	10.9442
87.6443	1.1833	10500	10.9439
87.6468	1.2397	11000	10.9438
87.6477	1.2960	11500	10.9434
87.6381	1.3524	12000	10.9433
87.6359	1.4087	12500	10.9430
87.6432	1.4650	13000	10.9428
87.6428	1.5214	13500	10.9428
87.634	1.5777	14000	10.9426
87.6254	1.6341	14500	10.9425
87.6294	1.6904	15000	10.9424
87.6307	1.7468	15500	10.9422
87.6393	1.8031	16000	10.9422
87.6266	1.8595	16500	10.9421
87.6327	1.9158	17000	10.9419
87.6298	1.9722	17500	10.9419
87.6353	2.0285	18000	10.9420
87.6329	2.0849	18500	10.9418
87.6332	2.1412	19000	10.9418
87.6301	2.1976	19500	10.9417
87.6308	2.2539	20000	10.9417
87.6302	2.3103	20500	10.9417
87.6378	2.3666	21000	10.9416
87.6317	2.4230	21500	10.9416
87.6272	2.4793	22000	10.9416
87.6308	2.5357	22500	10.9416
87.6299	2.5920	23000	10.9416
87.6303	2.6484	23500	10.9416
87.6262	2.7047	24000	10.9415
87.6281	2.7610	24500	10.9415
87.6344	2.8174	25000	10.9416

Framework versions

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