metadata

license: llama2
library_name: peft
tags:
  - generated_from_trainer
base_model: codellama/CodeLlama-7b-hf
model-index:
  - name: out/test
    results: []

See axolotl config

axolotl version: 0.4.0

base_model: codellama/CodeLlama-7b-hf
model_type: LlamaForCausalLM
tokenizer_type: CodeLlamaTokenizer

load_in_8bit: true
load_in_4bit: false
strict: false

datasets:
  - path: TristanBehrens/MusicCode_JSFakes_2024_Compose
    type:
      system_prompt: ""
      system_format: ""
      format: "[INST] {instruction} [/INST]"
      no_input_format: "[INST] {instruction} [/INST]"
dataset_prepared_path:
val_set_size: 0.05
output_dir: ./out/test

sequence_len: 16384
sample_packing: true
pad_to_sequence_len: true

adapter: lora
lora_model_dir:
lora_r: 32
lora_alpha: 16
lora_dropout: 0.05
lora_target_linear: true
lora_fan_in_fan_out:

wandb_project:
wandb_entity:
wandb_watch:
wandb_name:
wandb_log_model:

gradient_accumulation_steps: 4
micro_batch_size: 4
num_epochs: 4
optimizer: adamw_bnb_8bit
lr_scheduler: cosine
learning_rate: 0.0002

train_on_inputs: false
group_by_length: false
bf16: auto
fp16:
tf32: false

gradient_checkpointing: true
early_stopping_patience:
resume_from_checkpoint:
local_rank:
logging_steps: 1
xformers_attention:
flash_attention: true
s2_attention:

eval_sample_packing: False
warmup_steps: 10
evals_per_epoch: 4
saves_per_epoch: 1
debug:
deepspeed:
weight_decay: 0.0
fsdp:
fsdp_config:
special_tokens:
  bos_token: "<s>"
  eos_token: "</s>"
  unk_token: "<unk>"

out/test

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

Loss: 0.0553

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
distributed_type: multi-GPU
num_devices: 2
gradient_accumulation_steps: 4
total_train_batch_size: 32
total_eval_batch_size: 8
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: cosine
lr_scheduler_warmup_steps: 10
num_epochs: 4

Training results

Training Loss	Epoch	Step	Validation Loss
0.1833	0.06	1	0.1833
0.175	0.29	5	0.1681
0.1172	0.57	10	0.1097
0.0917	0.86	15	0.0878
0.0779	1.11	20	0.0750
0.0706	1.4	25	0.0682
0.0642	1.69	30	0.0635
0.0617	1.97	35	0.0609
0.0602	2.21	40	0.0588
0.0574	2.5	45	0.0573
0.0565	2.79	50	0.0563
0.0561	3.03	55	0.0558
0.0566	3.31	60	0.0554
0.0551	3.6	65	0.0553

Framework versions

PEFT 0.9.1.dev0
Transformers 4.39.0.dev0
Pytorch 2.2.0+cu121
Datasets 2.17.1
Tokenizers 0.15.0