base_model: google/gemma-3-4b-it
#hub_model_id: AlexHung29629/ModelMerlynIfeEldridge
plugins:
  - axolotl.integrations.liger.LigerPlugin
liger_rope: true
liger_rms_norm: true
liger_glu_activation: true
liger_layer_norm: true
liger_fused_linear_cross_entropy: true
liger_use_token_scaling: true
gradient_checkpointing: true
gradient_checkpointing_kwargs:
  use_reentrant: false
data_seed: 42
seed: 42
max_grad_norm: 1

bf16: true
tf32: true

datasets:
  - path: AlexHung29629/MerlynIfeEldridge2
    type: input_output

sequence_len: 758
sample_packing: false
optimizer: sgd
lr_scheduler: constant
micro_batch_size: 13
gradient_accumulation_steps: 1
num_epochs: 16
learning_rate: 1e-3
warmup_ratio: 0
#saves_per_epoch: 1
use_tensorboard: true
use_wandb: false
save_strategy: "no"

model-out

This model is a fine-tuned version of google/gemma-3-4b-it on the AlexHung29629/MerlynIfeEldridge2 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: 0.001
train_batch_size: 13
eval_batch_size: 13
seed: 42
optimizer: Use OptimizerNames.SGD and the args are: No additional optimizer arguments
lr_scheduler_type: constant
training_steps: 16

Training results

Framework versions

Transformers 4.57.6
Pytorch 2.9.1+cu128
Datasets 4.5.0
Tokenizers 0.22.2

AlexHung29629/MerlynIfeEldridgeEp16

model-out

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

AlexHung29629/
MerlynIfeEldridgeEp16