This model is a fine-tuned version of sathiiiii/polyalign-qwen2.5-3b-en-sft on the polyalign_dpo_train dataset. It achieves the following results on the evaluation set:

Loss: 0.3380
Rewards/chosen: -1.0773
Rewards/rejected: -4.7114
Rewards/accuracies: 0.8866
Rewards/margins: 3.6341
Logps/chosen: -102.4047
Logps/rejected: -83.3126
Logits/chosen: -2.2493
Logits/rejected: -2.2458

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-07
train_batch_size: 1
eval_batch_size: 1
seed: 42
distributed_type: multi-GPU
num_devices: 8
gradient_accumulation_steps: 8
total_train_batch_size: 64
total_eval_batch_size: 8
optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: cosine
lr_scheduler_warmup_ratio: 0.1
num_epochs: 1.0

Training results

Training Loss	Epoch	Step	Validation Loss	Rewards/chosen	Rewards/rejected	Rewards/accuracies	Rewards/margins	Logps/chosen	Logps/rejected	Logits/chosen	Logits/rejected
0.2949	0.4307	3000	0.3603	-1.1292	-4.2747	0.8788	3.1456	-102.9238	-78.9461	-2.2672	-2.2695
0.2779	0.8615	6000	0.3380	-1.0773	-4.7114	0.8866	3.6341	-102.4047	-83.3126	-2.2493	-2.2458

Framework versions

Transformers 4.56.2
Pytorch 2.9.1+rocm6.3
Datasets 4.0.0
Tokenizers 0.22.2

sathiiiii/polyalign-qwen2.5-3b-en-dpo

polyalign

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

sathiiiii/
polyalign-qwen2.5-3b-en-dpo