Hugging Face's logo

Magpie-Align
/
Llama-3-8B-Magpie-Align-SFT-v0.3

Text Generation
Transformers
PyTorch
Safetensors
English
Chinese
llama
axolotl
Generated from Trainer
conversational
text-generation-inference
Inference Endpoints
Model card Files Community
1
Llama-3-8B-Magpie-Align-SFT-v0.3 / README.md
flydust's picture
flydust
End of training
acf1f27 verified
preview code
|
raw
history blame
3.89 kB
metadata
license: llama3
base_model: meta-llama/Meta-Llama-3-8B
tags:
  - axolotl
  - generated_from_trainer
model-index:
  - name: Llama-3-8B-Magpie-Mix-RC
    results: []

Built with Axolotl

See axolotl config

axolotl version: 0.4.1

base_model: meta-llama/Meta-Llama-3-8B
model_type: LlamaForCausalLM
tokenizer_type: AutoTokenizer

load_in_8bit: false
load_in_4bit: false
strict: false

datasets:
  - path: Magpie-Align/Magpie-Reasoning-150K
    type: sharegpt
    conversation: llama3
  - path: Magpie-Align/Magpie-Qwen2-Pro-200K-Chinese
    type: sharegpt
    conversation: llama3
  - path: Magpie-Align/Magpie-Pro-MT-300K-v0.1
    type: sharegpt
    conversation: llama3
dataset_prepared_path: last_run_prepared
val_set_size: 0.001
output_dir: /home/cc/axolotl/axolotl_out/Llama-3-8B-base-magpie-RC

sequence_len: 8192
sample_packing: true
eval_sample_packing: false
pad_to_sequence_len: true

wandb_project: SynDa
wandb_entity:
wandb_watch:
wandb_name: Llama-3-8B-base-150KR-Llama3-Pro-MT-300K-C
wandb_log_model:
hub_model_id: Magpie-Align/Llama-3-8B-Magpie-Mix-RC

gradient_accumulation_steps: 32
micro_batch_size: 1
num_epochs: 2
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_ratio: 0.1
evals_per_epoch: 5
eval_table_size:
saves_per_epoch: 1
debug:
deepspeed:
weight_decay: 0.0
fsdp:
fsdp_config:
special_tokens:
  pad_token: <|end_of_text|>

Visualize in Weights & Biases

Llama-3-8B-Magpie-Mix-RC

This model is a fine-tuned version of meta-llama/Meta-Llama-3-8B on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 0.4611

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: 1
  • seed: 42
  • distributed_type: multi-GPU
  • num_devices: 4
  • gradient_accumulation_steps: 32
  • total_train_batch_size: 128
  • total_eval_batch_size: 4
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: cosine
  • lr_scheduler_warmup_steps: 98
  • num_epochs: 2

Training results

Training Loss Epoch Step Validation Loss
0.8616 0.0019 1 0.8870
0.5554 0.2013 106 0.5568
0.5067 0.4027 212 0.5065
0.4728 0.6040 318 0.4865
0.4681 0.8054 424 0.4740
0.4563 1.0067 530 0.4662
0.4115 1.1944 636 0.4642
0.3993 1.3957 742 0.4620
0.4048 1.5971 848 0.4613
0.4167 1.7984 954 0.4611

Framework versions

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