See axolotl config

axolotl version: 0.12.0.dev0

base_model: mrcuddle/NemoMix-Magcap-12B
tokenizer_type: AutoTokenizer
hub_model_id: mrcuddle/Magcap-Adonis-12B
strict: false

datasets:
  - path: mrcuddle/adonis_nsfw_alpaca
    type: alpaca
    streaming: false

output_dir: ./mistral-12b-adonis


sequence_len: 2048
sample_packing: true
pad_to_sequence_len: true


eval_sample_packing: False
gradient_accumulation_steps: 2
micro_batch_size: 2
num_epochs: 4
optimizer: paged_adamw_8bit
lr_scheduler: cosine
learning_rate: 0.00005

train_on_inputs: false
group_by_length: false

bf16: auto
fp16:
past_model_outputs: false

gradient_checkpointing: true
save_steps: 500
special_tokens:
  bos_token: "<s>"
  eos_token: "</s>"
  pad_token: "</s>"

Magcap-Adonis-12B

This model is a fine-tuned version of mrcuddle/NemoMix-Magcap-12B on the mrcuddle/adonis_nsfw_alpaca 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: 5e-05
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 4
• optimizer: Use OptimizerNames.PAGED_ADAMW_8BIT with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• training_steps: 32

Training results

Framework versions

• Transformers 4.53.2
• Pytorch 2.7.1+cu126
• Datasets 4.0.0
• Tokenizers 0.21.2