adapter: qlora
base_model: google/gemma-3-1b-it
bf16: auto
chat_template: gemma3
datasets:
- path: deepakkarkala/sft_sitcom_chandlerbing_jsonl
  split: train_without_fewshots
  type: alpaca
ddp_find_unused_parameters: true
eval_sample_packing: false
evals_per_epoch: null
flash_attention: true
gradient_accumulation_steps: 8
gradient_checkpointing: true
gradient_checkpointing_kwargs:
  use_reentrant: false
hub_model_id: deepakkarkala/gemma3_1b_lora_sft_sitcom
learning_rate: 0.0002
load_in_4bit: true
load_in_8bit: false
logging_steps: 1
lora_alpha: 16
lora_dropout: 0.05
lora_r: 32
lora_target_linear: true
lr_scheduler: cosine
micro_batch_size: 2
model_type: AutoModelForCausalLM
num_epochs: 4
optimizer: adamw_bnb_8bit
output_dir: ./outputs/out
pad_to_sequence_len: true
resume_from_checkpoint: null
sample_packing: true
saves_per_epoch: 1
sequence_len: 2048
special_tokens: null
tf32: true
tokenizer_type: AutoTokenizer
val_set_size: 0.05
wandb_entity: deepakkarkala-personal
wandb_log_model: checkpoint
wandb_name: sft_gemma3_1b
wandb_project: finetuning_llama31_8b_sitcom
wandb_run_id: sft_gemma3_1b_2
wandb_watch: null
warmup_ratio: 0.1
weight_decay: 0.0

Visualize in Weights & Biases

gemma3_1b_lora_sft_sitcom

This model is a fine-tuned version of google/gemma-3-1b-it on the deepakkarkala/sft_sitcom_chandlerbing_jsonl 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: 2
eval_batch_size: 2
seed: 42
gradient_accumulation_steps: 8
total_train_batch_size: 16
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: 26
training_steps: 264

Training results

Framework versions

PEFT 0.15.2
Transformers 4.51.3
Pytorch 2.6.0+cu124
Datasets 3.5.1
Tokenizers 0.21.1