Hugging Face's logo

Deepnoid
/
OPEN-SOLAR-KO-10.7B-v4

Text Generation
Transformers
TensorBoard
Safetensors
llama
Generated from Trainer
text-generation-inference
Inference Endpoints
Model card Files Metrics Community
OPEN-SOLAR-KO-10.7B-v4 / README.md
Jongilju's picture
Jongilju
init
017c539
preview code
|
raw
history blame contribute delete
2.77 kB
metadata
license: apache-2.0
base_model: beomi/OPEN-SOLAR-KO-10.7B
tags:
  - generated_from_trainer
model-index:
  - name: data/Models/beomidpo-out-v4
    results: []

Built with Axolotl

See axolotl config

axolotl version: 0.4.0

base_model: beomi/OPEN-SOLAR-KO-10.7B

load_in_8bit: false
load_in_4bit: false
strict: false

rl: dpo
datasets:
  - path: ./data/KR/Ja-ck/Orca-DPO-Pairs-KO/orca_dpo_pairs_ko.json
    split: train
    type: chatml.intel
    ds_type: json
    data_files: ["./data/KR/Ja-ck/Orca-DPO-Pairs-KO/orca_dpo_pairs_ko.json"]
  
dataset_prepared_path:
val_set_size: 0.0
output_dir: ./data/Models/beomidpo-out-v4

adapter: lora
lora_model_dir:

sequence_len: 2048
sample_packing: false
pad_to_sequence_len: false

lora_r: 8
lora_alpha: 32
lora_dropout: 0.05
lora_target_linear: true
lora_fan_in_fan_out:
lora_target_modules:
  - q_proj
  - v_proj
  - k_proj
  - o_proj
gradient_accumulation_steps: 1
micro_batch_size: 1
num_epochs: 1
optimizer: paged_adamw_8bit
lr_scheduler: cosine
learning_rate: 2e-5

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: false

warmup_steps: 10
save_steps: 100
save_total_limit: 3
debug:
deepspeed: deepspeed_configs/zero2.json
weight_decay: 0.0
fsdp:
fsdp_config:
special_tokens:
save_safetensors: true

data/Models/beomidpo-out-v4

This model is a fine-tuned version of beomi/OPEN-SOLAR-KO-10.7B on the None dataset.

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: 2e-05
  • train_batch_size: 1
  • eval_batch_size: 8
  • seed: 42
  • distributed_type: multi-GPU
  • num_devices: 8
  • total_train_batch_size: 8
  • total_eval_batch_size: 64
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: cosine
  • lr_scheduler_warmup_steps: 10
  • training_steps: 1591

Training results

Framework versions

  • Transformers 4.38.0.dev0
  • Pytorch 2.0.1+cu118
  • Datasets 2.16.1
  • Tokenizers 0.15.0