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: 0.0014
Rewards/chosen: -5.0669
Rewards/rejected: -54.5403
Rewards/accuracies: 0.9980
Rewards/margins: 49.4734
Logps/rejected: -678.6004
Logps/chosen: -134.8588
Logits/rejected: -1.0900
Logits/chosen: -0.9110

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-05
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

Epoch	Step	Validation Loss	Rewards/chosen	Rewards/rejected	Rewards/accuracies	Rewards/margins	Logps/rejected	Logps/chosen	Logits/rejected	Logits/chosen
0.0112	50	0.0017	-1.4591	-13.5699	0.9980	12.1108	-268.8958	-98.7805	-1.0783	-1.0112
0.0224	100	0.0014	-5.0523	-30.2803	0.9980	25.2279	-436.0001	-134.7133	-1.1646	-1.0404
0.0336	150	0.0276	-64.6547	-140.2898	0.9900	75.6351	-1536.0948	-730.7364	-0.3045	-0.1563
0.0448	200	0.0015	-88.2086	-227.0128	0.9980	138.8041	-2403.3250	-966.2761	-0.0748	0.0121
0.0559	250	0.0014	-11.8631	-47.0402	0.9980	35.1771	-603.5994	-202.8203	-1.1004	-1.0187
0.0671	300	0.0014	-11.5188	-53.1918	0.9980	41.6730	-665.1153	-199.3781	-1.0683	-0.9975
0.0783	350	0.0014	-11.5289	-53.2327	0.9980	41.7037	-665.5238	-199.4789	-1.0683	-0.9974
0.0895	400	0.0014	-13.3777	-80.4308	0.9980	67.0531	-937.5054	-217.9670	-1.2052	-1.0757
0.1007	450	0.0014	-13.3777	-80.4308	0.9980	67.0531	-937.5054	-217.9670	-1.2052	-1.0757
0.1119	500	0.0014	-13.3777	-80.4308	0.9980	67.0531	-937.5054	-217.9670	-1.2052	-1.0757
0.1231	550	0.0014	-4.0223	-46.3866	0.9980	42.3643	-597.0635	-124.4133	-1.0463	-0.8924
0.1343	600	0.0014	-5.0677	-54.5376	0.9980	49.4699	-678.5735	-134.8667	-1.0904	-0.9112
0.1454	650	0.0014	-5.0677	-54.5376	0.9980	49.4699	-678.5735	-134.8667	-1.0904	-0.9112
0.1566	700	0.0014	-5.0664	-54.5468	0.9980	49.4804	-678.6652	-134.8534	-1.0903	-0.9112
0.1678	750	0.0014	-5.0688	-54.5404	0.9980	49.4716	-678.6013	-134.8781	-1.0903	-0.9111
0.1790	800	0.0014	-5.0677	-54.5384	0.9980	49.4707	-678.5810	-134.8666	-1.0903	-0.9111
0.1902	850	0.0014	-5.0677	-54.5384	0.9980	49.4707	-678.5810	-134.8666	-1.0903	-0.9111
0.2014	900	0.0014	-5.0677	-54.5384	0.9980	49.4707	-678.5810	-134.8666	-1.0903	-0.9111
0.2126	950	0.0014	-5.0677	-54.5384	0.9980	49.4707	-678.5810	-134.8666	-1.0903	-0.9111
0.2238	1000	0.0014	-5.0669	-54.5403	0.9980	49.4734	-678.6004	-134.8588	-1.0900	-0.9110

Framework versions

Transformers 4.42.4
Pytorch 2.0.0+cu117
Datasets 2.20.0
Tokenizers 0.19.1

tsavage68/Summary4500_L3_1000steps_1e5rate_01beta_CSFTDPO

Hyponatremia_L3_1000steps_1e5rate_01beta_CSFTDPO

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

tsavage68/
Summary4500_L3_1000steps_1e5rate_01beta_CSFTDPO