adapter: qlora
base_model: Qwen/Qwen3-14B
bf16: false
chat_template: qwen3
dataloader_num_workers: 2
dataloader_pin_memory: true
dataloader_prefetch_factor: 8
datasets:
- eot_tokens:
  - <|im_end|>
  field_messages: conversations
  message_property_mappings:
    content: content
    role: role
  path: winglian/pirate-ultrachat-10k
  split: train
  type: chat_template
fp16: true
gradient_accumulation_steps: 1
gradient_checkpointing: true
gradient_checkpointing_kwargs:
  use_reentrant: false
learning_rate: 0.00019
load_in_4bit: true
logging_steps: 1
lora_alpha: 64
lora_mlp_kernel: true
lora_o_kernel: true
lora_qkv_kernel: true
lora_r: 32
lora_target_modules:
- q_proj
- k_proj
- v_proj
- o_proj
- gate_proj
- down_proj
- up_proj
lr_scheduler: cosine
max_grad_norm: 0.1
micro_batch_size: 1
num_epochs: 1
optimizer: paged_adamw_8bit
output_dir: ./outputs/qwen-sft-pirate-rrr
plugins:
- axolotl.integrations.cut_cross_entropy.CutCrossEntropyPlugin
sample_packing: true
saves_per_epoch: 2
sequence_len: 4096
warmup_steps: 5
xformers_attention: true

outputs/qwen-sft-pirate-rrr

This model is a fine-tuned version of Qwen/Qwen3-14B on the winglian/pirate-ultrachat-10k 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: 0.00019
train_batch_size: 1
eval_batch_size: 1
seed: 42
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
lr_scheduler_warmup_steps: 5
training_steps: 25
mixed_precision_training: Native AMP

Training results

Framework versions

PEFT 0.15.2
Transformers 4.51.3
Pytorch 2.6.0+cu124
Datasets 3.5.0
Tokenizers 0.21.1

winglian/pirate-qwen-14B

outputs/qwen-sft-pirate-rrr

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

winglian/
pirate-qwen-14B