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:
  - e8ef6edb66e20da7_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/e8ef6edb66e20da7_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
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: Romain-XV/f21b1007-4832-4c96-a460-d8e02dd8bc6c
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: 64
lora_dropout: 0.15
lora_fan_in_fan_out: null
lora_model_dir: null
lora_r: 32
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: 3024
micro_batch_size: 4
mlflow_experiment_name: /tmp/e8ef6edb66e20da7_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
save_steps: 100
sequence_len: 2048
strict: false
tf32: true
tokenizer_type: AutoTokenizer
train_on_inputs: false
trust_remote_code: true
val_set_size: 0.05
wandb_entity: null
wandb_mode: online
wandb_name: 6f911363-8c0f-4331-9742-a2fb57ee53b7
wandb_project: Gradients-On-Demand
wandb_run: your_name
wandb_runid: 6f911363-8c0f-4331-9742-a2fb57ee53b7
warmup_steps: 10
weight_decay: 0.0
xformers_attention: null

f21b1007-4832-4c96-a460-d8e02dd8bc6c

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: 10.7558

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: 3024

Training results

Training Loss	Epoch	Step	Validation Loss
88.7573	0.0005	1	11.0922
86.8959	0.0503	100	10.8690
86.759	0.1007	200	10.8399
86.7011	0.1510	300	10.8219
86.6914	0.2013	400	10.8056
86.5637	0.2517	500	10.7956
86.5208	0.3020	600	10.7887
86.5087	0.3523	700	10.7836
86.2209	0.4027	800	10.7794
86.2377	0.4530	900	10.7771
86.2876	0.5033	1000	10.7742
86.1991	0.5537	1100	10.7721
86.2955	0.6040	1200	10.7698
86.2954	0.6543	1300	10.7674
86.0302	0.7047	1400	10.7658
85.7452	0.7550	1500	10.7636
86.4382	0.8053	1600	10.7624
86.3187	0.8557	1700	10.7607
86.3855	0.9060	1800	10.7597
85.896	0.9563	1900	10.7587
86.1662	1.0067	2000	10.7581
86.4309	1.0570	2100	10.7576
86.4416	1.1073	2200	10.7566
86.5483	1.1577	2300	10.7566
86.2682	1.2080	2400	10.7562
85.986	1.2583	2500	10.7561
86.4859	1.3087	2600	10.7560
86.1625	1.3590	2700	10.7559
86.2638	1.4093	2800	10.7558
86.2056	1.4597	2900	10.7558
86.2621	1.5100	3000	10.7558

Framework versions

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