Hyponatremia_M2_1000steps_1e8rate_01beta_DPO

This model is a fine-tuned version of tsavage68/Hyponatremia_M2_150steps_1e6rate_SFT on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.6336
• Rewards/chosen: 0.0257
• Rewards/rejected: -0.0976
• Rewards/accuracies: 1.0
• Rewards/margins: 0.1233
• Logps/rejected: -71.8575
• Logps/chosen: -36.4699
• Logits/rejected: -2.2459
• Logits/chosen: -2.2411

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-08
• train_batch_size: 2
• eval_batch_size: 1
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 4
• 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
0.6935	0.2667	50	0.6952	0.0009	0.0048	0.4200	-0.0039	-70.8331	-36.7181	-2.2478	-2.2426
0.6916	0.5333	100	0.6886	0.0029	-0.0064	0.6900	0.0093	-70.9455	-36.6984	-2.2467	-2.2415
0.6753	0.8	150	0.6745	0.0092	-0.0286	0.9600	0.0377	-71.1670	-36.6351	-2.2465	-2.2414
0.6608	1.0667	200	0.6604	0.0159	-0.0509	1.0	0.0668	-71.3902	-36.5680	-2.2464	-2.2415
0.6494	1.3333	250	0.6499	0.0200	-0.0687	1.0	0.0886	-71.5680	-36.5273	-2.2468	-2.2419
0.6431	1.6	300	0.6443	0.0216	-0.0789	1.0	0.1005	-71.6708	-36.5114	-2.2463	-2.2414
0.6327	1.8667	350	0.6388	0.0260	-0.0863	1.0	0.1123	-71.7449	-36.4673	-2.2463	-2.2415
0.638	2.1333	400	0.6360	0.0249	-0.0932	1.0	0.1182	-71.8137	-36.4778	-2.2461	-2.2413
0.6339	2.4	450	0.6346	0.0260	-0.0952	1.0	0.1212	-71.8331	-36.4669	-2.2466	-2.2418
0.6348	2.6667	500	0.6338	0.0281	-0.0947	1.0	0.1228	-71.8287	-36.4459	-2.2462	-2.2415
0.6308	2.9333	550	0.6335	0.0271	-0.0963	1.0	0.1234	-71.8449	-36.4560	-2.2456	-2.2408
0.6305	3.2	600	0.6354	0.0254	-0.0942	1.0	0.1196	-71.8233	-36.4732	-2.2464	-2.2415
0.6367	3.4667	650	0.6348	0.0261	-0.0945	1.0	0.1207	-71.8269	-36.4657	-2.2464	-2.2416
0.6365	3.7333	700	0.6336	0.0253	-0.0980	1.0	0.1233	-71.8612	-36.4738	-2.2458	-2.2411
0.6344	4.0	750	0.6332	0.0262	-0.0980	1.0	0.1243	-71.8619	-36.4648	-2.2458	-2.2411
0.6347	4.2667	800	0.6335	0.0258	-0.0977	1.0	0.1235	-71.8584	-36.4688	-2.2458	-2.2411
0.6347	4.5333	850	0.6336	0.0257	-0.0976	1.0	0.1233	-71.8575	-36.4699	-2.2459	-2.2411
0.6339	4.8	900	0.6336	0.0257	-0.0976	1.0	0.1233	-71.8575	-36.4699	-2.2459	-2.2411
0.6348	5.0667	950	0.6336	0.0257	-0.0976	1.0	0.1233	-71.8575	-36.4699	-2.2459	-2.2411
0.6393	5.3333	1000	0.6336	0.0257	-0.0976	1.0	0.1233	-71.8575	-36.4699	-2.2459	-2.2411

Framework versions

• Transformers 4.42.3
• Pytorch 2.0.0+cu117
• Datasets 2.20.0
• Tokenizers 0.19.1