You can deploy this model and start earning money today!

Kimory-X/
zephyr-7b-mypo3-full-beta0.001-lr5e-6

Name: zephyr-7b-mypo3-full-beta0.001-lr5e-6
Availability: InStock

generated_from_trainer

alignment-handbook/zephyr-7b-sft-full

finetune

apache-2.0

Views

No views yet

zephyr-7b-mypo3-full-beta0.001-lr5e-6

This model is a fine-tuned version of alignment-handbook/zephyr-7b-sft-full on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 1.3624
Rewards/chosen: -0.0998
Rewards/rejected: -0.2912
Rewards/accuracies: 0.6845
Rewards/margins: 0.1914
Logps/rejected: -534.1523
Logps/chosen: -368.6681
Logits/rejected: -1.3619
Logits/chosen: -1.6521

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-06
train_batch_size: 4
eval_batch_size: 8
seed: 42
distributed_type: multi-GPU
num_devices: 6
gradient_accumulation_steps: 2
total_train_batch_size: 48
total_eval_batch_size: 48
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: cosine
lr_scheduler_warmup_ratio: 0.1
num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss	Rewards/chosen	Rewards/rejected	Rewards/accuracies	Rewards/margins	Logps/rejected	Logps/chosen	Logits/rejected	Logits/chosen
1.3806	0.0785	100	1.3823	-0.0395	-0.0546	0.6310	0.0151	-297.5703	-308.3824	-1.9572	-2.0128
1.3793	0.1570	200	1.3781	-0.0879	-0.1511	0.6042	0.0632	-394.0232	-356.7915	-3.5849	-3.5582
1.3724	0.2355	300	1.3785	-0.1455	-0.2495	0.6577	0.1040	-492.4726	-414.4250	-2.5286	-2.6509
1.376	0.3140	400	1.3811	-0.0668	-0.1034	0.5893	0.0366	-346.3430	-335.6931	-3.3110	-3.3031
1.3709	0.3925	500	1.3731	-0.0786	-0.1642	0.6339	0.0856	-407.1909	-347.5009	-2.5096	-2.6381
1.3737	0.4710	600	1.3733	-0.0743	-0.1595	0.6548	0.0852	-402.4303	-343.1721	-3.0350	-3.1238
1.3699	0.5495	700	1.3720	-0.0726	-0.1733	0.6369	0.1006	-416.1999	-341.5030	-3.1016	-3.2368
1.3633	0.6279	800	1.3691	-0.1255	-0.2992	0.6756	0.1737	-542.1292	-394.3810	-1.9556	-2.2128
1.36	0.7064	900	1.3721	-0.1038	-0.2448	0.6339	0.1410	-487.7896	-372.7076	-1.7970	-2.0662
1.3588	0.7849	1000	1.3656	-0.0878	-0.2608	0.6607	0.1730	-503.7656	-356.7212	-1.7706	-2.0863
1.3582	0.8634	1100	1.3635	-0.0997	-0.2847	0.6875	0.1850	-527.6372	-368.5705	-1.3706	-1.6541
1.3631	0.9419	1200	1.3624	-0.0998	-0.2912	0.6845	0.1914	-534.1523	-368.6681	-1.3619	-1.6521

Framework versions

Transformers 4.43.1
Pytorch 2.1.2
Datasets 2.18.0
Tokenizers 0.19.1

Kimory-X/zephyr-7b-mypo3-full-beta0.001-lr5e-6

zephyr-7b-mypo3-full-beta0.001-lr5e-6

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

Kimory-X/
zephyr-7b-mypo3-full-beta0.001-lr5e-6