Hugging Face's logo

CarrotAI
/
SmolLm-360M-Ko-Instruct

Text Generation
Transformers
PyTorch
Safetensors
llama
Generated from Trainer
conversational
text-generation-inference
Inference Endpoints
Model card Files Community
SmolLm-360M-Ko-Instruct / README.md
CarrotAI's picture
CarrotAI
Upload folder using huggingface_hub
3a9b19c verified
preview code
|
raw
history blame
3.09 kB
metadata
license: apache-2.0
base_model: HuggingFaceTB/SmolLM-360M-Instruct
tags:
  - generated_from_trainer
model-index:
  - name: SomlLm-360M-Ko-Instruct
    results: []

Built with Axolotl

See axolotl config

axolotl version: 0.4.1

base_model: HuggingFaceTB/SmolLM-360M-Instruct
tokenizer_type: AutoTokenizer

load_in_8bit: false
load_in_4bit: false
strict: false


chat_template: chatml
datasets:
    - path: CarrotAI/ko-instruction-dataset
      type: alpaca
    - path: werty1248/sharegpt-tagengo-gpt4-ko
      type: sharegpt
    - path: changpt/ko-lima-vicuna
      type: sharegpt
    - path: davidkim205/kollm-converations
      type: sharegpt
    - path: CarrotAI/Amazing-Instructions
      type: alpaca

dataset_prepared_path: last_run_prepared
val_set_size: 0.05
output_dir: ./SomlLm-360M-Ko-Instruct

sequence_len: 8192
sample_packing: true
pad_to_sequence_len: true

wandb_project:
wandb_entity:
wandb_watch:
wandb_name:
wandb_log_model:

gradient_accumulation_steps: 8
micro_batch_size: 6
num_epochs: 1
optimizer: paged_adamw_8bit
lr_scheduler: cosine
learning_rate: 2e-5

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

gradient_checkpointing: true
gradient_checkpointing_kwargs:
  use_reentrant: false
early_stopping_patience:
resume_from_checkpoint:
logging_steps: 1
xformers_attention:
flash_attention: true

warmup_steps: 100
evals_per_epoch: 1
eval_table_size:
saves_per_epoch: 1
debug:
deepspeed:
weight_decay: 0.0
fsdp:
fsdp_config:
special_tokens:

SomlLm-360M-Ko-Instruct

This model is a fine-tuned version of HuggingFaceTB/SmolLM-360M-Instruct on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 1.1516

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: 6
  • eval_batch_size: 6
  • seed: 42
  • distributed_type: multi-GPU
  • num_devices: 2
  • gradient_accumulation_steps: 8
  • total_train_batch_size: 96
  • total_eval_batch_size: 12
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: cosine
  • lr_scheduler_warmup_steps: 100
  • num_epochs: 1

Training results

Training Loss Epoch Step Validation Loss
1.1401 0.9993 1316 1.1516

Framework versions

  • Transformers 4.42.3
  • Pytorch 2.3.1+cu121
  • Datasets 2.19.1
  • Tokenizers 0.19.1