See axolotl config

axolotl version: 0.4.1

adapter: lora
base_model: unsloth/codellama-7b
bf16: true
chat_template: llama3
dataset_prepared_path: null
datasets:
- data_files:
  - e7b3aa093efffc3e_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/e7b3aa093efffc3e_train_data.json
  type:
    field_input: title
    field_instruction: text
    field_output: summary
    format: '{instruction} {input}'
    no_input_format: '{instruction}'
    system_format: '{system}'
    system_prompt: ''
debug: null
deepspeed: null
early_stopping_patience: 4
eval_max_new_tokens: 128
eval_steps: 150
eval_table_size: null
flash_attention: true
fp16: null
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 4
gradient_checkpointing: true
group_by_length: false
hub_model_id: Romain-XV/7b47c58c-79ec-460a-9608-60e4909bfb6b
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: 128
lora_dropout: 0.3
lora_fan_in_fan_out: null
lora_model_dir: null
lora_r: 64
lora_target_linear: true
lora_target_modules:
- q_proj
- k_proj
- v_proj
lr_scheduler: cosine
max_grad_norm: 1.0
max_steps: 1632
micro_batch_size: 2
mlflow_experiment_name: /tmp/e7b3aa093efffc3e_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: 150
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: 19c70989-a909-41fe-8d1b-4651f0e3d4d6
wandb_project: Gradients-On-Demand
wandb_run: your_name
wandb_runid: 19c70989-a909-41fe-8d1b-4651f0e3d4d6
warmup_steps: 10
weight_decay: 0.0
xformers_attention: null

7b47c58c-79ec-460a-9608-60e4909bfb6b

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

• Loss: 0.7500

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: 1632

Training results

Training Loss	Epoch	Step	Validation Loss
1.4012	0.0004	1	1.1911
0.8992	0.0546	150	0.8447
0.8439	0.1091	300	0.8177
0.8444	0.1637	450	0.8045
0.647	0.2182	600	0.7914
0.8174	0.2728	750	0.7794
0.8516	0.3273	900	0.7696
1.0711	0.3819	1050	0.7616
0.7681	0.4364	1200	0.7554
0.8908	0.4910	1350	0.7516
0.9181	0.5455	1500	0.7500

Framework versions

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