Ailonordsletta/nanbeige-med

See axolotl config

axolotl version: 0.13.0.dev0

base_model: Nanbeige/Nanbeige4.1-3B
model_type: AutoModelForCausalLM
trust_remote_code: true
tokenizer_use_fast: false

adapter: lora
load_in_8bit: true
lora_r: 8
lora_alpha: 16
lora_dropout: 0.05
lora_target_modules:
  - q_proj
  - v_proj
  - k_proj
  - o_proj
  - gate_proj
  - down_proj
  - up_proj

datasets:
  - path: Ailonordsletta/med
    type: chat_template

sample_packing: true
sequence_len: 4096

micro_batch_size: 1
gradient_accumulation_steps: 8
num_epochs: 1
learning_rate: 0.0002
optimizer: adamw_bnb_8bit
bf16: auto
gradient_checkpointing: true

wandb_project: nanbeige-finetune
output_dir: ./outputs/mymodel

outputs/mymodel

This model is a fine-tuned version of Nanbeige/Nanbeige4.1-3B on the Ailonordsletta/med 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: 0.0002
• train_batch_size: 1
• eval_batch_size: 1
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 8
• optimizer: Use OptimizerNames.ADAMW_BNB 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: 2
• training_steps: 47

Training results

Framework versions

• PEFT 0.17.1
• Transformers 4.57.0
• Pytorch 2.7.1+cu126
• Datasets 4.0.0
• Tokenizers 0.22.1