# Mirror of SERA's axolotl_qwen3_8b.yaml, parameterized for Jupiter SFT on v3 data.
# Filled at render time via sed-substitution of 316.

base_model: Qwen/Qwen3-8B
deepspeed: /e/scratch/jureap59/feuer1/code/axolotl/deepspeed_configs/zero3_bf16.json

load_in_8bit: false
load_in_4bit: false

# plugins disabled 2026-04-22: CCE + bf16 + flash-attn on aarch64/torch2.9 caused
# gradient explosion (grad_norm 9.8e+11) and loss -> 0 within first 3-7 steps.
# plugins:
#   - axolotl.integrations.cut_cross_entropy.CutCrossEntropyPlugin

chat_template: chatml
datasets:
  - path: laion/Sera-4.5A-Full-T1-v3-316
    data_files:
      - sera-4.5a-full-t1_v3_316.jsonl
    type: chat_template
    field_messages: messages
    ds_type: json
    message_field_training: train

dataset_prepared_path: /e/data1/datasets/playground/ot-baf/axolotl_dataset_cache/sera-v3-316
output_dir: /e/data1/datasets/playground/ot-baf/checkpoints/sera-v3-316-axolotl__Qwen3-8B

sequence_len: 32768

wandb_project:
wandb_entity:
wandb_watch:
wandb_name: sera-v3-316-axolotl__Qwen3-8B
wandb_log_model:

gradient_accumulation_steps: 8
micro_batch_size: 1
num_epochs: 3
optimizer: adamw_torch
lr_scheduler: cosine
learning_rate: 1e-5
adam_beta1: 0.9
adam_beta2: 0.95

bf16: auto
tf32: false

gradient_checkpointing: true
activation_offloading: true
resume_from_checkpoint:
logging_steps: 1
flash_attention: true

loss_watchdog_threshold: 5.0
loss_watchdog_patience: 3

warmup_ratio: 0.1875
evals_per_epoch: 0
save_strategy: epoch

weight_decay: 0.01
max_grad_norm: 1.0
special_tokens:

e/data1/datasets/playground/ot-baf/checkpoints/sera-v3-316-axolotl__Qwen3-8B

This model is a fine-tuned version of Qwen/Qwen3-8B on the laion/Sera-4.5A-Full-T1-v3-316 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
distributed_type: multi-GPU
num_devices: 4
gradient_accumulation_steps: 8
total_train_batch_size: 32
total_eval_batch_size: 4
optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.95) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: cosine
lr_scheduler_warmup_steps: 2
training_steps: 10

Training results

Framework versions

Transformers 5.5.0
Pytorch 2.9.1+cu130
Datasets 4.5.0
Tokenizers 0.22.2

laion/Sera-4.5A-Full-T1-v3-316-axolotl__Qwen3-8B

e/data1/datasets/playground/ot-baf/checkpoints/sera-v3-316-axolotl__Qwen3-8B

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

laion/
Sera-4.5A-Full-T1-v3-316-axolotl__Qwen3-8B