This model is a fine-tuned version of tsavage68/Summary4500_L3_100steps_1e6rate_SFT on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 0.0120
Rewards/chosen: -0.0496
Rewards/rejected: -5.4117
Rewards/accuracies: 0.9980
Rewards/margins: 5.3621
Logps/rejected: -187.3140
Logps/chosen: -84.6859
Logits/rejected: -1.1056
Logits/chosen: -1.0597

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: 1e-07
train_batch_size: 1
eval_batch_size: 1
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: cosine
lr_scheduler_warmup_steps: 100
training_steps: 1000

Training results

Training Loss	Epoch	Step	Validation Loss	Rewards/chosen	Rewards/rejected	Rewards/accuracies	Rewards/margins	Logps/rejected	Logps/chosen	Logits/rejected	Logits/chosen
0.6815	0.0112	50	0.6641	0.0083	-0.0520	0.8080	0.0603	-133.7178	-84.1071	-1.0987	-1.0686
0.4744	0.0224	100	0.4877	0.0371	-0.4326	0.9980	0.4697	-137.5237	-83.8192	-1.1002	-1.0687
0.1569	0.0336	150	0.2583	0.0728	-1.1837	0.9980	1.2565	-145.0345	-83.4618	-1.1011	-1.0678
0.073	0.0448	200	0.1386	0.0923	-1.9142	0.9980	2.0064	-152.3389	-83.2668	-1.1016	-1.0660
0.0403	0.0559	250	0.0767	0.1014	-2.6346	0.9980	2.7360	-159.5432	-83.1754	-1.1016	-1.0642
0.0129	0.0671	300	0.0438	0.0700	-3.3888	0.9980	3.4588	-167.0854	-83.4899	-1.1020	-1.0624
0.0111	0.0783	350	0.0285	0.0488	-3.9980	0.9980	4.0468	-173.1772	-83.7018	-1.1040	-1.0629
0.004	0.0895	400	0.0215	0.0359	-4.4185	0.9980	4.4543	-177.3822	-83.8313	-1.1049	-1.0622
0.0009	0.1007	450	0.0174	0.0068	-4.7618	0.9980	4.7687	-180.8155	-84.1214	-1.1046	-1.0607
0.001	0.1119	500	0.0153	-0.0148	-4.9921	0.9980	4.9774	-183.1187	-84.3376	-1.1050	-1.0605
0.0007	0.1231	550	0.0138	-0.0270	-5.1627	0.9980	5.1358	-184.8247	-84.4593	-1.1051	-1.0597
0.0014	0.1343	600	0.0130	-0.0367	-5.2740	0.9980	5.2373	-185.9372	-84.5570	-1.1060	-1.0605
0.0008	0.1454	650	0.0125	-0.0513	-5.3515	0.9980	5.3002	-186.7127	-84.7030	-1.1061	-1.0603
0.0012	0.1566	700	0.0122	-0.0490	-5.3894	0.9980	5.3404	-187.0914	-84.6797	-1.1058	-1.0600
0.0258	0.1678	750	0.0121	-0.0511	-5.4040	0.9980	5.3530	-187.2378	-84.7006	-1.1056	-1.0597
0.0008	0.1790	800	0.0121	-0.0462	-5.4023	0.9980	5.3561	-187.2207	-84.6519	-1.1058	-1.0601
0.0075	0.1902	850	0.0120	-0.0543	-5.4192	0.9980	5.3649	-187.3889	-84.7324	-1.1056	-1.0597
0.0009	0.2014	900	0.0120	-0.0508	-5.4136	0.9980	5.3628	-187.3338	-84.6980	-1.1055	-1.0596
0.0007	0.2126	950	0.0120	-0.0486	-5.4116	0.9980	5.3631	-187.3138	-84.6756	-1.1056	-1.0597
0.0013	0.2238	1000	0.0120	-0.0496	-5.4117	0.9980	5.3621	-187.3140	-84.6859	-1.1056	-1.0597

Framework versions

Transformers 4.42.4
Pytorch 2.0.0+cu117
Datasets 2.20.0
Tokenizers 0.19.1

tsavage68/Summary4500_L3_1000steps_1e7rate_01beta_CSFTDPO

Hyponatremia_L3_1000steps_1e7rate_01beta_CSFTDPO

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

tsavage68/
Summary4500_L3_1000steps_1e7rate_01beta_CSFTDPO