1min-v2-luxia-8b

<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->

<img src="https://raw.githubusercontent.com/OpenAccess-AI-Collective/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/>

<details><summary>See axolotl config</summary>

axolotl version: 0.4.0

base_model: maywell/Llama-3-Ko-Luxia-Instruct
trust_remote_code: true
load_in_8bit: false
load_in_4bit: false
strict: false
datasets:
  - path: "../data/generated_ds.json"
    type: alpaca
    conversation: chatml
dataset_prepared_path: ../data/dataset_v2_pre
val_set_size: 0.05
output_dir: ../data/output/1min-v2-luxia-8b
sequence_len: 1024
sample_packing: true
pad_to_sequence_len: true
eval_sample_packing: false
wandb_project: 
wandb_entity: 
wandb_watch:
wandb_name:
wandb_log_model:
gradient_accumulation_steps: 4
micro_batch_size: 1
num_epochs: 10
optimizer: adamw_bnb_8bit
lr_scheduler: cosine
learning_rate: 2e-6
train_on_inputs: false
group_by_length: false
bf16: auto
fp16: null
tf32: false
gradient_checkpointing: true
early_stopping_patience: null
resume_from_checkpoint: null
local_rank: null
logging_steps: 1
xformers_attention: null
flash_attention: true
warmup_steps: 10
evals_per_epoch: 4
eval_table_size: null
eval_max_new_tokens: 128
saves_per_epoch: 1
save_total_limit: 4
debug: true
deepspeed: deepspeed_configs/zero2.json
weight_decay: 0.0
special_tokens:
  pad_token: <|end_of_text|>

</details><br>

data/output/1min-v2-luxia-8b

This model is a fine-tuned version of maywell/Llama-3-Ko-Luxia-Instruct on the manipulated instructkr/ko_youtube_transcription_v2_filtered dataset. It achieves the following results on the evaluation set:

• Loss: 2.0986

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-06
• train_batch_size: 1
• eval_batch_size: 1
• seed: 42
• distributed_type: multi-GPU
• num_devices: 7
• gradient_accumulation_steps: 4
• total_train_batch_size: 28
• total_eval_batch_size: 7
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 10
• num_epochs: 10

Training results

Framework versions

• Transformers 4.40.2
• Pytorch 2.1.2+cu118
• Datasets 2.19.1
• Tokenizers 0.19.1