Hugging Face's logo

Deadwalker0
/
axolotl-ex

PEFT
TensorBoard
Safetensors
llama
Generated from Trainer
8-bit precision
bitsandbytes
Model card Files Metrics Community
axolotl-ex / README.md
Deadwalker0's picture
Deadwalker0
Upload folder using huggingface_hub
76503fe verified
preview code
|
raw
history blame contribute delete
3.83 kB
metadata
license: apache-2.0
library_name: peft
tags:
  - generated_from_trainer
base_model: openlm-research/open_llama_3b_v2
model-index:
  - name: lora-out
    results: []

Built with Axolotl

See axolotl config

axolotl version: 0.4.0

base_model: openlm-research/open_llama_3b_v2
model_type: LlamaForCausalLM
tokenizer_type: LlamaTokenizer
load_in_8bit: true
load_in_4bit: false
strict: false
push_dataset_to_hub:
datasets:
  - path: teknium/GPT4-LLM-Cleaned
    type: alpaca
dataset_prepared_path:
val_set_size: 0.02
adapter: lora
lora_model_dir:
sequence_len: 1024
sample_packing: true
lora_r: 8
lora_alpha: 16
lora_dropout: 0.0
lora_target_modules:
  - gate_proj
  - down_proj
  - up_proj
  - q_proj
  - v_proj
  - k_proj
  - o_proj
lora_fan_in_fan_out:
wandb_project:
wandb_entity:
wandb_watch:
wandb_name:
wandb_log_model:
output_dir: ./lora-out
gradient_accumulation_steps: 1
micro_batch_size: 2
num_epochs: 4
optimizer: adamw_bnb_8bit
torchdistx_path:
lr_scheduler: cosine
learning_rate: 0.0002
train_on_inputs: false
group_by_length: false
bf16: false
fp16: true
tf32: false
gradient_checkpointing: true
early_stopping_patience:
resume_from_checkpoint:
local_rank:
logging_steps: 1
xformers_attention:
flash_attention: true
gptq_groupsize:
s2_attention:
gptq_model_v1:
warmup_steps: 20
evals_per_epoch: 4
saves_per_epoch: 1
debug:
deepspeed:
weight_decay: 0.1
fsdp:
fsdp_config:
special_tokens:
  bos_token: "<s>"
  eos_token: "</s>"
  unk_token: "<unk>"

lora-out

This model is a fine-tuned version of openlm-research/open_llama_3b_v2 on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 1.0041

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
  • distributed_type: multi-GPU
  • num_devices: 8
  • total_train_batch_size: 16
  • total_eval_batch_size: 16
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: cosine
  • lr_scheduler_warmup_steps: 20
  • num_epochs: 4
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss
1.3745 0.0 1 1.6297
1.1387 0.25 168 1.0849
1.0619 0.5 336 1.0484
0.9686 0.75 504 1.0277
1.0816 1.0 672 1.0170
1.0513 1.23 840 1.0088
1.0814 1.48 1008 1.0041
1.0275 1.73 1176 0.9929
0.8872 1.98 1344 0.9883
0.9351 2.21 1512 0.9985
0.9077 2.46 1680 0.9968
0.9494 2.71 1848 0.9907
0.9596 2.96 2016 0.9916
0.8771 3.19 2184 1.0012
0.8912 3.44 2352 1.0041
0.7828 3.69 2520 1.0041

Framework versions

  • PEFT 0.8.2
  • Transformers 4.38.0.dev0
  • Pytorch 2.0.1+cu118
  • Datasets 2.17.0
  • Tokenizers 0.15.0