base_model: timarni/qwen3_dpo
# 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: timarni/MNLP_intstruction_tuning # timarni/MNLP_intstruction_tuning
    type: alpaca
    split: train

shuffle_merged_datasets: true

val_set_size: 0.1
output_dir: ./dpo_50k_full_alpaca_on_inputs
dataset_prepared_path: last_run_prepared

sequence_len: 2048 #2048
sample_packing: true # was true -> need to check if it actually learns on the samples or not (better understand te hyperparam and event. install axolotl to debug)
eval_sample_packing: true
pad_to_sequence_len: true
train_on_inputs: true # NEW
# group_by_length: false NEW?

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

wandb_project: mnlp_project
wandb_entity: tim-arni
wandb_watch:
wandb_name: dpo_50k_full_alpaca_on_inputs
wandb_log_model:

gradient_accumulation_steps: 16
micro_batch_size: 2 # 2
num_epochs: 3
optimizer: adamw_torch
lr_scheduler: cosine
learning_rate: 0.00005 # 0.00005
cosine_min_lr_ratio: 0.1

warmup_ratio: 0.05
weight_decay: 0.01

bf16: auto
tf32: true

gradient_checkpointing: offload
gradient_checkpointing_kwargs:
  use_reentrant: false
resume_from_checkpoint:
logging_steps: 1
gradient_clipping: 1.0 # or max_grad_norm?
flash_attention: true

evals_per_epoch: 4
saves_per_epoch: 2
save_total_limit: 25
special_tokens:

dpo_50k_full_alpaca_on_inputs

This model is a fine-tuned version of timarni/qwen3_dpo on the timarni/MNLP_intstruction_tuning dataset. It achieves the following results on the evaluation set:

Loss: 0.6042

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: 5e-05
train_batch_size: 2
eval_batch_size: 2
seed: 42
distributed_type: multi-GPU
num_devices: 4
gradient_accumulation_steps: 16
total_train_batch_size: 128
total_eval_batch_size: 8
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: 28
num_epochs: 3.0

Training results

Training Loss	Epoch	Step	Validation Loss
2.2898	0.0053	1	1.3104
1.9418	0.2508	47	0.6884
1.9117	0.5015	94	0.6640
1.8866	0.7523	141	0.6476
1.8145	1.0	188	0.6326
1.7877	1.2508	235	0.6215
1.7858	1.5015	282	0.6154
1.7838	1.7523	329	0.6099
1.6951	2.0	376	0.6055
1.7257	2.2508	423	0.6056
1.7434	2.5015	470	0.6046
1.7578	2.7523	517	0.6042

Framework versions

Transformers 4.51.3
Pytorch 2.5.1+cu121
Datasets 3.5.1
Tokenizers 0.21.1

timarni/dpo_50k_full_alpaca_on_inputs

dpo_50k_full_alpaca_on_inputs

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

timarni/
dpo_50k_full_alpaca_on_inputs