metadata

license: mit
library_name: peft
tags:
  - trl
  - reward-trainer
  - generated_from_trainer
base_model: openai-community/gpt2-large
metrics:
  - accuracy
model-index:
  - name: >-
      RM-HH-Gemma_helpful_human_loraR64_20000_gpt2-large_shuffleTrue_extractchosenFalse
    results: []

RM-HH-Gemma_helpful_human_loraR64_20000_gpt2-large_shuffleTrue_extractchosenFalse

This model is a fine-tuned version of openai-community/gpt2-large on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 0.6198
Accuracy: 0.6546

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: 1.41e-05
train_batch_size: 1
eval_batch_size: 8
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 4
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 2.0

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.703	0.06	250	0.7010	0.5414
0.6903	0.11	500	0.6820	0.5584
0.6715	0.17	750	0.6700	0.5805
0.6498	0.22	1000	0.6604	0.5985
0.6625	0.28	1250	0.6550	0.6070
0.6425	0.33	1500	0.6523	0.6170
0.6514	0.39	1750	0.6480	0.6226
0.6535	0.45	2000	0.6454	0.6256
0.6233	0.5	2250	0.6427	0.6296
0.6434	0.56	2500	0.6403	0.6306
0.6288	0.61	2750	0.6379	0.6421
0.632	0.67	3000	0.6360	0.6361
0.6365	0.72	3250	0.6337	0.6436
0.6329	0.78	3500	0.6322	0.6456
0.6278	0.84	3750	0.6312	0.6441
0.6369	0.89	4000	0.6299	0.6476
0.6313	0.95	4250	0.6292	0.6466
0.6315	1.0	4500	0.6285	0.6456
0.6039	1.06	4750	0.6278	0.6441
0.6259	1.11	5000	0.6268	0.6471
0.629	1.17	5250	0.6259	0.6481
0.6201	1.23	5500	0.6251	0.6506
0.6118	1.28	5750	0.6251	0.6521
0.6125	1.34	6000	0.6243	0.6481
0.6115	1.39	6250	0.6236	0.6476
0.6056	1.45	6500	0.6234	0.6506
0.6255	1.5	6750	0.6224	0.6501
0.6314	1.56	7000	0.6215	0.6511
0.6346	1.62	7250	0.6211	0.6501
0.6269	1.67	7500	0.6207	0.6516
0.6104	1.73	7750	0.6204	0.6526
0.6138	1.78	8000	0.6202	0.6526
0.6172	1.84	8250	0.6201	0.6541
0.6149	1.89	8500	0.6199	0.6541
0.6022	1.95	8750	0.6198	0.6546

Framework versions

PEFT 0.9.0
Transformers 4.38.2
Pytorch 2.1.2
Datasets 2.18.0
Tokenizers 0.15.2