output

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

• Loss: 0.3205
• Dice: 0.7942
• Jaccard: 0.6587

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: 0.001
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 3
• total_train_batch_size: 12
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 200
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Dice	Jaccard
0.7019	1.4213	250	0.8054	0.5385	0.3685
0.6182	2.8425	500	0.5602	0.6475	0.4787
0.5293	4.2619	750	0.5032	0.6525	0.4843
0.4877	5.6831	1000	0.4818	0.6853	0.5213
0.5122	7.1025	1250	0.4612	0.6975	0.5355
0.4446	8.5237	1500	0.4311	0.7016	0.5403
0.4823	9.9450	1750	0.4431	0.6992	0.5375
0.4204	11.3643	2000	0.4135	0.7093	0.5495
0.3739	12.7856	2250	0.3847	0.7426	0.5906
0.4038	14.2049	2500	0.3818	0.7542	0.6053
0.3605	15.6262	2750	0.3768	0.7430	0.5910
0.3917	17.0455	3000	0.3766	0.7381	0.5849
0.3928	18.4668	3250	0.3743	0.7378	0.5846
0.3624	19.8880	3500	0.3780	0.7342	0.5801
0.3360	21.3074	3750	0.3399	0.7715	0.6280
0.3253	22.7287	4000	0.3410	0.7588	0.6113
0.2869	24.1480	4250	0.3389	0.7744	0.6319
0.3078	25.5693	4500	0.3330	0.7733	0.6304
0.3178	26.9905	4750	0.3618	0.7535	0.6044
0.3093	28.4099	5000	0.3312	0.7857	0.6470
0.3478	29.8311	5250	0.3436	0.7697	0.6257
0.3324	31.2505	5500	0.3263	0.7831	0.6436
0.2808	32.6717	5750	0.3366	0.7731	0.6301
0.2926	34.0911	6000	0.3334	0.7809	0.6406
0.2518	35.5123	6250	0.3249	0.7850	0.6460
0.3012	36.9336	6500	0.3316	0.7789	0.6379
0.2824	38.3529	6750	0.3306	0.7833	0.6438
0.2754	39.7742	7000	0.3246	0.7848	0.6458
0.2725	41.1935	7250	0.3194	0.7917	0.6552
0.2593	42.6148	7500	0.3221	0.7858	0.6471
0.3149	44.0342	7750	0.3257	0.7868	0.6486
0.2801	45.4554	8000	0.3184	0.7915	0.6550
0.2710	46.8767	8250	0.3122	0.7942	0.6586
0.2389	48.2960	8500	0.3137	0.7940	0.6584
0.2569	49.7173	8750	0.3249	0.7866	0.6482
0.2572	51.1366	9000	0.3148	0.7977	0.6635
0.2354	52.5579	9250	0.3177	0.7941	0.6585
0.2529	53.9791	9500	0.3080	0.7998	0.6664
0.2935	55.3985	9750	0.3242	0.7887	0.6512
0.2437	56.8197	10000	0.3081	0.7987	0.6649
0.2420	58.2391	10250	0.3107	0.7967	0.6621
0.2357	59.6603	10500	0.3247	0.7885	0.6509
0.2341	61.0797	10750	0.3136	0.7960	0.6611
0.2269	62.5009	11000	0.3091	0.7967	0.6621
0.2644	63.9222	11250	0.3133	0.7951	0.6599
0.2423	65.3416	11500	0.3121	0.7959	0.6609
0.2548	66.7628	11750	0.3187	0.7943	0.6588
0.2691	68.1822	12000	0.3191	0.7946	0.6592
0.2396	69.6034	12250	0.3142	0.7983	0.6643
0.2356	71.0228	12500	0.3108	0.7990	0.6653
0.2497	72.4440	12750	0.3208	0.7929	0.6568
0.2491	73.8653	13000	0.3175	0.7927	0.6566
0.2433	75.2846	13250	0.3129	0.7973	0.6629
0.2649	76.7059	13500	0.3156	0.7967	0.6621
0.2567	78.1252	13750	0.3097	0.8000	0.6666
0.2535	79.5465	14000	0.3114	0.7975	0.6633
0.2462	80.9677	14250	0.3122	0.7981	0.6641
0.2426	82.3871	14500	0.3204	0.7917	0.6552
0.2419	83.8083	14750	0.3109	0.8001	0.6668
0.2778	85.2277	15000	0.3111	0.7993	0.6658
0.2547	86.6490	15250	0.3074	0.7999	0.6665
0.2564	88.0683	15500	0.3075	0.8006	0.6675
0.2322	89.4896	15750	0.3102	0.7995	0.6660
0.2572	90.9108	16000	0.3101	0.7995	0.6660
0.2380	92.3302	16250	0.3127	0.7976	0.6633
0.2464	93.7514	16500	0.3161	0.7967	0.6621
0.2439	95.1708	16750	0.3126	0.7972	0.6628
0.2319	96.5920	17000	0.3160	0.7961	0.6612
0.2260	98.0114	17250	0.3128	0.7973	0.6629
0.2368	99.4326	17500	0.3096	0.7981	0.6640
0.2202	100.8539	17750	0.3133	0.7990	0.6652
0.2494	102.2732	18000	0.3141	0.7988	0.6651
0.2429	103.6945	18250	0.3130	0.7989	0.6652
0.2503	105.1139	18500	0.3119	0.7992	0.6656
0.2309	106.5351	18750	0.3117	0.7989	0.6651
0.2452	107.9564	19000	0.3139	0.7963	0.6616
0.2342	109.3757	19250	0.3194	0.7946	0.6592
0.2543	110.7970	19500	0.3182	0.7950	0.6597
0.2351	112.2163	19750	0.3151	0.7963	0.6615
0.2578	113.6376	20000	0.3135	0.7979	0.6637
0.2318	115.0569	20250	0.3121	0.7984	0.6644
0.2573	116.4782	20500	0.3232	0.7941	0.6586
0.2264	117.8994	20750	0.3163	0.7959	0.6609
0.2471	119.3188	21000	0.3128	0.7993	0.6657
0.2197	120.7400	21250	0.3127	0.7969	0.6623
0.2681	122.1594	21500	0.3150	0.7959	0.6610
0.2390	123.5806	21750	0.3107	0.7981	0.6640
0.2329	125.0	22000	0.3091	0.8021	0.6695
0.2294	126.4213	22250	0.3177	0.7956	0.6606
0.2509	127.8425	22500	0.3158	0.7957	0.6608
0.2495	129.2619	22750	0.3108	0.8004	0.6672
0.2288	130.6831	23000	0.3122	0.7981	0.6640
0.2351	132.1025	23250	0.3163	0.7971	0.6626
0.2404	133.5237	23500	0.3135	0.7989	0.6651
0.2108	134.9450	23750	0.3113	0.7990	0.6653
0.2418	136.3643	24000	0.3131	0.7983	0.6643
0.2276	137.7856	24250	0.3149	0.7955	0.6605
0.2425	139.2049	24500	0.3139	0.7990	0.6653
0.2472	140.6262	24750	0.3150	0.7976	0.6633
0.2339	142.0455	25000	0.3205	0.7932	0.6573
0.2470	143.4668	25250	0.3175	0.7963	0.6615
0.2126	144.8880	25500	0.3163	0.7949	0.6595
0.2436	146.3074	25750	0.3118	0.7994	0.6658
0.2345	147.7287	26000	0.3177	0.7971	0.6627
0.2374	149.1480	26250	0.3105	0.7987	0.6648
0.2261	150.5693	26500	0.3126	0.7966	0.6620
0.2430	151.9905	26750	0.3150	0.7981	0.6640
0.2346	153.4099	27000	0.3147	0.7976	0.6633
0.2426	154.8311	27250	0.3105	0.7997	0.6663
0.2397	156.2505	27500	0.3118	0.7985	0.6646
0.2575	157.6717	27750	0.3160	0.7953	0.6601
0.2389	159.0911	28000	0.3207	0.7922	0.6559
0.2437	160.5123	28250	0.3156	0.7971	0.6627
0.2353	161.9336	28500	0.3116	0.7983	0.6643
0.2537	163.3529	28750	0.3169	0.7958	0.6608
0.2500	164.7742	29000	0.3134	0.8000	0.6667
0.2312	166.1935	29250	0.3154	0.7979	0.6637
0.2218	167.6148	29500	0.3126	0.7983	0.6643
0.2505	169.0342	29750	0.3196	0.7964	0.6616
0.2361	170.4554	30000	0.3108	0.7994	0.6658
0.2519	171.8767	30250	0.3087	0.7995	0.6660
0.2404	173.2960	30500	0.3140	0.7984	0.6644
0.2383	174.7173	30750	0.3120	0.7978	0.6636
0.2307	176.1366	31000	0.3142	0.7956	0.6606
0.2012	177.5579	31250	0.3141	0.7966	0.6619
0.2422	178.9791	31500	0.3161	0.7979	0.6638
0.2388	180.3985	31750	0.3094	0.8007	0.6676
0.2266	181.8197	32000	0.3136	0.7982	0.6642
0.2460	183.2391	32250	0.3092	0.8011	0.6682
0.2484	184.6603	32500	0.3152	0.7968	0.6623
0.2316	186.0797	32750	0.3161	0.7937	0.6580
0.2506	187.5009	33000	0.3136	0.7960	0.6612
0.2442	188.9222	33250	0.3137	0.7968	0.6622
0.2361	190.3416	33500	0.3159	0.7970	0.6625
0.2545	191.7628	33750	0.3206	0.7941	0.6585
0.2309	193.1822	34000	0.3123	0.7980	0.6638
0.2521	194.6034	34250	0.3173	0.7958	0.6608
0.2357	196.0228	34500	0.3128	0.7981	0.6640
0.2453	197.4440	34750	0.3170	0.7942	0.6587
0.2264	198.8653	35000	0.3181	0.7953	0.6602
0.2509	200.0	35200	0.3205	0.7942	0.6587

Framework versions

• Transformers 5.19.0
• Pytorch 2.11.0+cu130
• Datasets 5.0.0
• Tokenizers 0.23.3