ShenaoZ
/

0.0001_withdpo_3iters_bs256_551lr_iter_2

Text Generation

alignment-handbook

Generated from Trainer

text-generation-inference

Inference Endpoints

Model card Files Files and versions Community

0.0001_withdpo_3iters_bs256_551lr_iter_2 / all_results.json

ShenaoZ's picture

Model save

d5121cb verified 7 months ago

193 Bytes

	{
	"epoch": 0.99,
	"train_loss": 0.5676172232326073,
	"train_runtime": 4739.1874,
	"train_samples": 20378,
	"train_samples_per_second": 4.3,
	"train_steps_per_second": 0.017
	}