See axolotl config

axolotl version: 0.9.2

base_model: Qwen/Qwen3-0.6B-Base
# Automatically upload checkpoint and final model to HF
# hub_model_id: username/custom_model_name

plugins:
  - axolotl.integrations.cut_cross_entropy.CutCrossEntropyPlugin
strict: false

chat_template: qwen3
datasets:
  - path: /mloscratch/users/arni/Workspace/mnlp_sft/datasets/s1k.json
    type: chat_template
    split: train
    field_messages: conversations
    # message_property_mappings:
    #   role: from
    #   content: value

output_dir: ./outputs/_2
dataset_prepared_path: last_run_prepared

# To be sure that no LORA is done
adapter: null
lora: false
merge_lora: false

sequence_len: 4096 #2048
sample_packing: true
eval_sample_packing: false
pad_to_sequence_len: true

wandb_project: mnlp_project
wandb_entity: tim-arni
wandb_watch:
wandb_name: qwen3_s1k_2
wandb_log_model:

gradient_accumulation_steps: 2 # 16 following https://unsloth.ai/blog/qwen3
micro_batch_size: 1 # 2
num_epochs: 6
optimizer: adamw_torch
lr_scheduler: cosine
learning_rate: 0.00001 # 0.0002
cosine_min_lr_ratio: 0.1

bf16: auto
tf32: true

gradient_checkpointing: offload
gradient_checkpointing_kwargs:
  use_reentrant: false
resume_from_checkpoint:
logging_steps: 1
gradient_clipping: 1.0
flash_attention: true

warmup_ratio: 0.03
evals_per_epoch: 4
saves_per_epoch: 1
weight_decay: 1e-4
special_tokens:

outputs/_2

This model is a fine-tuned version of Qwen/Qwen3-0.6B-Base on the /mloscratch/users/arni/Workspace/mnlp_sft/datasets/s1k.json 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: 1e-05
• train_batch_size: 1
• eval_batch_size: 1
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 2
• optimizer: Use adamw_torch 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: 7
• num_epochs: 6.0

Training results

Framework versions

• Transformers 4.51.3
• Pytorch 2.5.1+cu121
• Datasets 3.5.1
• Tokenizers 0.21.1