See axolotl config

axolotl version: 0.4.1

adapter: qlora
auto_resume_from_checkpoints: true
base_model: NousResearch/CodeLlama-7b-hf-flash
bf16: auto
chat_template: llama3
dataloader_num_workers: 12
dataset_prepared_path: null
datasets:
- data_files:
  - 4bbd53a67d293afe_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/4bbd53a67d293afe_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: 3
eval_max_new_tokens: 128
eval_steps: 100
eval_table_size: null
evals_per_epoch: null
flash_attention: false
fp16: false
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 2
gradient_checkpointing: true
group_by_length: true
hub_model_id: error577/01ccba0e-ec73-458d-8b3f-cd49b725f91c
hub_repo: null
hub_strategy: checkpoint
hub_token: null
learning_rate: 0.0002
load_in_4bit: true
load_in_8bit: false
local_rank: null
logging_steps: 10
lora_alpha: 128
lora_dropout: 0.05
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: null
micro_batch_size: 2
mlflow_experiment_name: /tmp/4bbd53a67d293afe_train_data.json
model_type: AutoModelForCausalLM
num_epochs: 3
optimizer: adamw_torch_4bit
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: 512
special_tokens:
  pad_token: </s>
strict: false
tf32: false
tokenizer_type: AutoTokenizer
train_on_inputs: false
trust_remote_code: true
val_set_size: 0.005
wandb_entity: null
wandb_mode: online
wandb_name: 2da1b8da-fcf3-4d42-b4ea-e536ab198f3e
wandb_project: Gradients-On-Demand
wandb_run: your_name
wandb_runid: 2da1b8da-fcf3-4d42-b4ea-e536ab198f3e
warmup_steps: 30
weight_decay: 0.0
xformers_attention: null

01ccba0e-ec73-458d-8b3f-cd49b725f91c

This model is a fine-tuned version of NousResearch/CodeLlama-7b-hf-flash on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.1951

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: 2
• total_train_batch_size: 4
• optimizer: Use OptimizerNames.ADAMW_TORCH_4BIT 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: 30
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss
No log	0.0002	1	1.1623
0.2625	0.0168	100	0.3775
0.2999	0.0335	200	0.3310
0.152	0.0503	300	0.2558
0.1516	0.0670	400	0.2434
0.1028	0.0838	500	0.3978
0.1054	0.1005	600	0.2447
0.1057	0.1173	700	0.2287
0.1816	0.1340	800	0.2239
0.0807	0.1508	900	0.2128
0.1286	0.1675	1000	0.2033
0.1858	0.1843	1100	0.2038
0.1361	0.2010	1200	0.1928
0.1428	0.2178	1300	0.1864
0.1253	0.2345	1400	0.1913
0.1476	0.2513	1500	0.1920
0.153	0.2680	1600	0.1805
0.1416	0.2848	1700	0.1777
0.0545	0.3015	1800	0.2057
0.2373	0.3183	1900	0.1916
0.098	0.3350	2000	0.1951

Framework versions

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