SegFormer_b3_mappillary_

This model is a fine-tuned version of nvidia/segformer-b3-finetuned-cityscapes-1024-1024 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.9465
• Mean Iou: 0.6876
• Mean Accuracy: 0.8034
• Overall Accuracy: 0.9418
• Accuracy Construction--barrier--fence: 0.6800
• Accuracy Construction--barrier--guard-rail: 0.8164
• Accuracy Construction--barrier--other-barrier: 0.6949
• Accuracy Construction--barrier--wall: 0.6949
• Accuracy Construction--flat--road: 0.9600
• Accuracy Construction--flat--service-lane: 0.6083
• Accuracy Construction--flat--sidewalk: 0.8963
• Accuracy Construction--structure--building: 0.9474
• Accuracy Human--person: 0.8455
• Accuracy Human--rider--bicyclist: 0.7457
• Accuracy Marking--crosswalk-zebra: 0.8079
• Accuracy Marking--general: 0.7057
• Accuracy Nature--sky: 0.9903
• Accuracy Nature--terrain: 0.8351
• Accuracy Nature--vegetation: 0.9468
• Accuracy Object--support--pole: 0.5784
• Accuracy Object--support--traffic-sign-frame: 0.6875
• Accuracy Object--traffic-light: 0.7517
• Accuracy Object--traffic-sign--front: 0.8324
• Accuracy Object--vehicle--bicycle: 0.7600
• Accuracy Object--vehicle--bus: 0.8898
• Accuracy Object--vehicle--car: 0.9598
• Accuracy Object--vehicle--truck: 0.8427
• Iou Construction--barrier--fence: 0.5668
• Iou Construction--barrier--guard-rail: 0.6378
• Iou Construction--barrier--other-barrier: 0.5595
• Iou Construction--barrier--wall: 0.5403
• Iou Construction--flat--road: 0.9205
• Iou Construction--flat--service-lane: 0.4618
• Iou Construction--flat--sidewalk: 0.7968
• Iou Construction--structure--building: 0.8891
• Iou Human--person: 0.6797
• Iou Human--rider--bicyclist: 0.5718
• Iou Marking--crosswalk-zebra: 0.7125
• Iou Marking--general: 0.5860
• Iou Nature--sky: 0.9811
• Iou Nature--terrain: 0.6872
• Iou Nature--vegetation: 0.8949
• Iou Object--support--pole: 0.4599
• Iou Object--support--traffic-sign-frame: 0.5531
• Iou Object--traffic-light: 0.5835
• Iou Object--traffic-sign--front: 0.7231
• Iou Object--vehicle--bicycle: 0.5797
• Iou Object--vehicle--bus: 0.7946
• Iou Object--vehicle--car: 0.9055
• Iou Object--vehicle--truck: 0.7304

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-05
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 8
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 1000
• num_epochs: 15
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Mean Iou	Mean Accuracy	Overall Accuracy	Accuracy Construction--barrier--fence	Accuracy Construction--barrier--guard-rail	Accuracy Construction--barrier--other-barrier	Accuracy Construction--barrier--wall	Accuracy Construction--flat--road	Accuracy Construction--flat--service-lane	Accuracy Construction--flat--sidewalk	Accuracy Construction--structure--building	Accuracy Human--person	Accuracy Human--rider--bicyclist	Accuracy Marking--crosswalk-zebra	Accuracy Marking--general	Accuracy Nature--sky	Accuracy Nature--terrain	Accuracy Nature--vegetation	Accuracy Object--support--pole	Accuracy Object--support--traffic-sign-frame	Accuracy Object--traffic-light	Accuracy Object--traffic-sign--front	Accuracy Object--vehicle--bicycle	Accuracy Object--vehicle--bus	Accuracy Object--vehicle--car	Accuracy Object--vehicle--truck	Iou Construction--barrier--fence	Iou Construction--barrier--guard-rail	Iou Construction--barrier--other-barrier	Iou Construction--barrier--wall	Iou Construction--flat--road	Iou Construction--flat--service-lane	Iou Construction--flat--sidewalk	Iou Construction--structure--building	Iou Human--person	Iou Human--rider--bicyclist	Iou Marking--crosswalk-zebra	Iou Marking--general	Iou Nature--sky	Iou Nature--terrain	Iou Nature--vegetation	Iou Object--support--pole	Iou Object--support--traffic-sign-frame	Iou Object--traffic-light	Iou Object--traffic-sign--front	Iou Object--vehicle--bicycle	Iou Object--vehicle--bus	Iou Object--vehicle--car	Iou Object--vehicle--truck
1.3425	0.4444	1000	1.2765	0.2545	0.3044	0.8800	0.6134	0.0	0.0033	0.0266	0.9695	0.0	0.7990	0.9191	0.0001	0.0	0.0000	0.0467	0.9827	0.6667	0.9530	0.0001	0.0	0.0	0.0845	0.0	0.0000	0.9361	0.0	0.3190	0.0	0.0033	0.0263	0.8350	0.0	0.6117	0.8074	0.0001	0.0	0.0000	0.0426	0.9577	0.5511	0.8295	0.0001	0.0	0.0	0.0842	0.0	0.0000	0.7854	0.0
1.1254	0.8889	2000	1.0846	0.4898	0.5869	0.9169	0.6043	0.7972	0.5664	0.5400	0.9357	0.0	0.8852	0.9351	0.6966	0.0	0.6074	0.5823	0.9887	0.8180	0.9342	0.4084	0.0	0.0577	0.7376	0.0	0.7325	0.9520	0.7184	0.4699	0.4923	0.4243	0.4304	0.8822	0.0	0.6986	0.8531	0.5266	0.0	0.5141	0.4423	0.9723	0.6271	0.8680	0.3223	0.0	0.0574	0.5713	0.0	0.6531	0.8591	0.6014
1.0545	1.3333	3000	1.0331	0.5437	0.6380	0.9253	0.6281	0.6042	0.6614	0.6689	0.9598	0.1969	0.8240	0.9329	0.8189	0.0	0.6535	0.5900	0.9890	0.7623	0.9431	0.4452	0.0576	0.6242	0.7553	0.0473	0.8173	0.9560	0.7380	0.4962	0.5220	0.4592	0.5118	0.8942	0.1780	0.7109	0.8662	0.5534	0.0	0.5445	0.4827	0.9761	0.6458	0.8780	0.3593	0.0575	0.4728	0.6404	0.0470	0.7192	0.8673	0.6224
1.0702	1.7778	4000	1.0152	0.5925	0.7119	0.9278	0.6339	0.8483	0.7095	0.6748	0.9643	0.4595	0.8005	0.9253	0.8096	0.0	0.6926	0.5453	0.9910	0.7992	0.9379	0.5329	0.4324	0.6756	0.7847	0.5913	0.8819	0.9506	0.7321	0.4993	0.5053	0.4767	0.4969	0.8986	0.3391	0.7166	0.8689	0.5999	0.0	0.6029	0.4807	0.9773	0.6585	0.8804	0.3939	0.3874	0.5017	0.6616	0.4461	0.7138	0.8789	0.6421
0.9517	2.2222	5000	0.9929	0.6200	0.7373	0.9309	0.6915	0.7243	0.6735	0.6030	0.9513	0.5424	0.8577	0.9424	0.8352	0.1893	0.7769	0.6458	0.9894	0.8427	0.9335	0.5217	0.5057	0.6906	0.8021	0.7058	0.8375	0.9478	0.7480	0.5345	0.5905	0.5076	0.4934	0.9028	0.4129	0.7423	0.8723	0.6025	0.1827	0.6044	0.5226	0.9785	0.6709	0.8819	0.4060	0.4432	0.5208	0.6637	0.4977	0.7046	0.8846	0.6402
1.0168	2.6667	6000	0.9903	0.6447	0.7756	0.9316	0.6411	0.8493	0.7208	0.6234	0.9524	0.8120	0.8049	0.9446	0.8134	0.6314	0.7481	0.6312	0.9903	0.8097	0.9423	0.4974	0.6191	0.6996	0.7810	0.7204	0.8662	0.9570	0.7842	0.5213	0.5644	0.5073	0.5028	0.8994	0.4049	0.7222	0.8763	0.6491	0.5067	0.6396	0.5282	0.9784	0.6763	0.8856	0.4040	0.4873	0.5286	0.6815	0.5415	0.7612	0.8878	0.6727
1.0717	3.1111	7000	0.9785	0.6508	0.7757	0.9340	0.6746	0.7986	0.6721	0.6354	0.9472	0.6116	0.8959	0.9425	0.8280	0.6989	0.7759	0.6393	0.9887	0.8304	0.9454	0.5240	0.5967	0.7134	0.7888	0.7306	0.8959	0.9525	0.7551	0.5325	0.6073	0.5107	0.5195	0.9065	0.4417	0.7717	0.8754	0.6435	0.5295	0.6266	0.5228	0.9791	0.6817	0.8861	0.4167	0.4969	0.5338	0.6829	0.5395	0.7210	0.8904	0.6522
1.2969	3.5556	8000	0.9726	0.6607	0.7788	0.9353	0.6778	0.7819	0.6816	0.6455	0.9504	0.7169	0.8924	0.9420	0.8285	0.7140	0.7469	0.6249	0.9890	0.8405	0.9465	0.5522	0.6019	0.6983	0.8032	0.6940	0.8498	0.9571	0.7779	0.5486	0.6167	0.5305	0.5219	0.9080	0.4372	0.7636	0.8778	0.6530	0.5394	0.6504	0.5336	0.9795	0.6808	0.8879	0.4286	0.5104	0.5458	0.6881	0.5569	0.7712	0.8914	0.6741
1.066	4.0	9000	0.9686	0.6606	0.7836	0.9352	0.6679	0.8057	0.6835	0.6398	0.9465	0.6595	0.9040	0.9473	0.8201	0.7057	0.7608	0.6587	0.9889	0.8467	0.9406	0.5365	0.6196	0.6992	0.7951	0.7535	0.8633	0.9606	0.8193	0.5390	0.6280	0.5393	0.5135	0.9071	0.4355	0.7590	0.8777	0.6543	0.5289	0.6610	0.5504	0.9798	0.6819	0.8884	0.4267	0.5145	0.5534	0.6897	0.5437	0.7678	0.8903	0.6651
0.9934	4.4444	10000	0.9634	0.6599	0.7918	0.9362	0.6866	0.8041	0.6551	0.7025	0.9550	0.7200	0.8742	0.9367	0.8221	0.6826	0.7694	0.6683	0.9876	0.8475	0.9470	0.5354	0.6706	0.7083	0.8183	0.8001	0.8479	0.9553	0.8165	0.5422	0.6144	0.5272	0.5433	0.9112	0.4443	0.7763	0.8806	0.6577	0.4982	0.6644	0.5512	0.9794	0.6814	0.8875	0.4287	0.5303	0.5516	0.6906	0.4916	0.7697	0.8931	0.6635
0.9396	4.8889	11000	0.9608	0.6706	0.7865	0.9375	0.6659	0.7949	0.6840	0.6375	0.9573	0.6485	0.8755	0.9484	0.8153	0.7668	0.7889	0.6647	0.9907	0.8336	0.9425	0.5406	0.6331	0.7248	0.8096	0.7237	0.8555	0.9560	0.8317	0.5476	0.6242	0.5600	0.5252	0.9132	0.4695	0.7721	0.8799	0.6674	0.5666	0.6676	0.5550	0.9800	0.6793	0.8900	0.4318	0.5307	0.5538	0.7017	0.5603	0.7705	0.8948	0.6825
1.0232	5.3333	12000	0.9602	0.6665	0.7811	0.9374	0.6948	0.7343	0.6613	0.6420	0.9553	0.5946	0.8991	0.9486	0.8225	0.6960	0.7909	0.6412	0.9903	0.8492	0.9400	0.5514	0.6470	0.7395	0.8046	0.7763	0.8375	0.9602	0.7878	0.5510	0.6077	0.5071	0.5136	0.9148	0.4577	0.7800	0.8786	0.6628	0.5361	0.6571	0.5479	0.9802	0.6900	0.8905	0.4368	0.5341	0.5536	0.7050	0.5662	0.7751	0.8941	0.6901
1.003	5.7778	13000	0.9568	0.6665	0.7868	0.9377	0.6587	0.7905	0.7084	0.6866	0.9556	0.5506	0.9020	0.9400	0.8251	0.6929	0.7483	0.6743	0.9903	0.8455	0.9451	0.5720	0.6413	0.7217	0.8021	0.7812	0.9025	0.9549	0.8077	0.5459	0.6315	0.5463	0.5151	0.9130	0.4338	0.7792	0.8823	0.6636	0.5190	0.6567	0.5635	0.9804	0.6903	0.8912	0.4402	0.5333	0.5608	0.7045	0.5575	0.7314	0.8977	0.6934
1.0042	6.2222	14000	0.9524	0.6759	0.7923	0.9393	0.7133	0.7758	0.6648	0.6809	0.9609	0.5916	0.8752	0.9431	0.8173	0.7671	0.7821	0.6872	0.9899	0.8290	0.9472	0.5473	0.6624	0.7405	0.8113	0.7582	0.8978	0.9559	0.8234	0.5641	0.6347	0.5491	0.5441	0.9163	0.4613	0.7830	0.8823	0.6701	0.5536	0.6898	0.5640	0.9804	0.6937	0.8926	0.4372	0.5404	0.5589	0.7084	0.5520	0.7715	0.8982	0.7006
1.1329	6.6667	15000	0.9514	0.6774	0.7927	0.9398	0.6569	0.8159	0.6914	0.7315	0.9660	0.6223	0.8705	0.9411	0.8354	0.7894	0.7848	0.6641	0.9904	0.8079	0.9484	0.5592	0.6137	0.7370	0.8166	0.7257	0.8908	0.9567	0.8171	0.5473	0.6384	0.5535	0.5377	0.9185	0.4755	0.7851	0.8851	0.6648	0.5548	0.6978	0.5690	0.9804	0.6840	0.8920	0.4418	0.5268	0.5655	0.7083	0.5549	0.7825	0.9004	0.7167
0.9527	7.1111	16000	0.9583	0.6728	0.7944	0.9376	0.7232	0.8220	0.6712	0.6710	0.9460	0.5458	0.9175	0.9498	0.8358	0.7740	0.8209	0.6919	0.9886	0.7697	0.9432	0.5861	0.6600	0.7535	0.8073	0.7433	0.8595	0.9565	0.8342	0.5638	0.6379	0.5591	0.5278	0.9112	0.4177	0.7759	0.8823	0.6661	0.5778	0.6754	0.5674	0.9800	0.6777	0.8915	0.4479	0.5367	0.5604	0.7129	0.5705	0.7603	0.8962	0.6784
0.9354	7.5556	17000	0.9510	0.6782	0.7953	0.9394	0.7130	0.8167	0.6822	0.6615	0.9568	0.5684	0.8837	0.9473	0.8334	0.7628	0.8149	0.7023	0.9909	0.8648	0.9403	0.5304	0.7092	0.7378	0.8073	0.7204	0.8791	0.9603	0.8079	0.5636	0.6413	0.5599	0.5420	0.9149	0.4203	0.7894	0.8852	0.6728	0.5804	0.6862	0.5697	0.9806	0.6873	0.8920	0.4373	0.5428	0.5693	0.7165	0.5854	0.7695	0.8976	0.6953
0.908	8.0	18000	0.9500	0.6777	0.8005	0.9399	0.7093	0.7857	0.7006	0.6132	0.9574	0.6719	0.9108	0.9499	0.8358	0.8035	0.8172	0.6634	0.9909	0.8249	0.9405	0.5649	0.6957	0.7319	0.8177	0.7674	0.8684	0.9584	0.8323	0.5724	0.6381	0.5655	0.5112	0.9189	0.4591	0.8012	0.8812	0.6700	0.5631	0.6707	0.5631	0.9807	0.6972	0.8931	0.4485	0.5452	0.5741	0.7061	0.5525	0.7896	0.8983	0.6877
0.8843	8.4444	19000	0.9511	0.6796	0.7993	0.9401	0.7004	0.8178	0.6519	0.7174	0.9572	0.6224	0.9077	0.9474	0.8269	0.7723	0.7965	0.6727	0.9898	0.8237	0.9426	0.5698	0.6640	0.7518	0.8313	0.7532	0.8897	0.9604	0.8164	0.5690	0.6276	0.5562	0.5450	0.9174	0.4548	0.7880	0.8860	0.6724	0.5485	0.6921	0.5652	0.9808	0.6932	0.8931	0.4511	0.5478	0.5682	0.7127	0.5552	0.7931	0.9005	0.7126
0.9566	8.8889	20000	0.9472	0.6787	0.7985	0.9404	0.7049	0.7575	0.6818	0.6768	0.9555	0.6506	0.9006	0.9482	0.8200	0.7445	0.8169	0.7053	0.9907	0.8590	0.9422	0.5622	0.6687	0.7411	0.8172	0.7526	0.8892	0.9563	0.8225	0.5685	0.6161	0.5442	0.5352	0.9182	0.4805	0.7979	0.8858	0.6724	0.5217	0.7002	0.5769	0.9808	0.6892	0.8930	0.4473	0.5409	0.5736	0.7173	0.5570	0.7833	0.9012	0.7092
0.9415	9.3333	21000	0.9548	0.6777	0.7951	0.9399	0.6808	0.8055	0.6928	0.6894	0.9626	0.6218	0.8789	0.9473	0.8325	0.7719	0.7636	0.6931	0.9905	0.8210	0.9433	0.5703	0.6522	0.7388	0.8204	0.7483	0.8854	0.9558	0.8208	0.5609	0.6273	0.5627	0.5270	0.9183	0.4640	0.7927	0.8843	0.6738	0.5507	0.6784	0.5680	0.9808	0.6784	0.8926	0.4504	0.5386	0.5761	0.7179	0.5593	0.7798	0.9016	0.7029
0.9323	9.7778	22000	0.9473	0.6801	0.8004	0.9401	0.6976	0.7806	0.7003	0.6717	0.9568	0.5970	0.9007	0.9481	0.8302	0.7417	0.8203	0.6961	0.9901	0.8677	0.9394	0.5692	0.7008	0.7428	0.8164	0.7686	0.8670	0.9575	0.8475	0.5692	0.6195	0.5507	0.5221	0.9184	0.4674	0.7960	0.8852	0.6797	0.5621	0.6920	0.5766	0.9808	0.6814	0.8924	0.4532	0.5521	0.5755	0.7186	0.5649	0.7876	0.9007	0.6960
0.9317	10.2222	23000	0.9506	0.6815	0.8006	0.9404	0.7121	0.8045	0.7110	0.6755	0.9526	0.6830	0.9059	0.9489	0.8399	0.7169	0.8078	0.7014	0.9907	0.8323	0.9433	0.5671	0.6756	0.7389	0.8177	0.7487	0.8808	0.9624	0.7964	0.5753	0.6195	0.5531	0.5441	0.9163	0.4886	0.7851	0.8864	0.6739	0.5363	0.7008	0.5833	0.9809	0.6951	0.8940	0.4526	0.5529	0.5792	0.7162	0.5601	0.7802	0.9008	0.7007
1.0166	10.6667	24000	0.9497	0.6823	0.8009	0.9409	0.7061	0.8256	0.7003	0.6733	0.9620	0.6408	0.8786	0.9487	0.8402	0.7293	0.8262	0.6825	0.9900	0.8168	0.9442	0.5813	0.6780	0.7523	0.8205	0.7576	0.8812	0.9605	0.8254	0.5732	0.6280	0.5625	0.5400	0.9194	0.4772	0.7939	0.8869	0.6747	0.5375	0.6894	0.5738	0.9810	0.6867	0.8939	0.4566	0.5533	0.5755	0.7220	0.5683	0.7850	0.9019	0.7119
0.9552	11.1111	25000	0.9514	0.6839	0.7956	0.9410	0.6848	0.8055	0.6767	0.6773	0.9627	0.5812	0.8873	0.9497	0.8387	0.7154	0.7963	0.6822	0.9906	0.8297	0.9442	0.5835	0.7141	0.7414	0.8294	0.7613	0.8648	0.9590	0.8237	0.5684	0.6419	0.5621	0.5352	0.9190	0.4497	0.7908	0.8864	0.6778	0.5532	0.6995	0.5788	0.9809	0.6913	0.8940	0.4588	0.5643	0.5798	0.7216	0.5840	0.7834	0.9024	0.7064
0.9968	11.5556	26000	0.9473	0.6823	0.8022	0.9405	0.6980	0.8150	0.6930	0.6878	0.9563	0.6225	0.9037	0.9493	0.8403	0.7341	0.8076	0.6876	0.9899	0.8503	0.9409	0.5806	0.7029	0.7571	0.8294	0.7585	0.8656	0.9608	0.8181	0.5709	0.6372	0.5610	0.5313	0.9178	0.4644	0.7953	0.8871	0.6764	0.5590	0.6897	0.5793	0.9810	0.6847	0.8930	0.4576	0.5509	0.5765	0.7208	0.5659	0.7791	0.9003	0.7139
0.9555	12.0	27000	0.9485	0.6820	0.7971	0.9410	0.6763	0.8329	0.6877	0.6743	0.9597	0.5537	0.8978	0.9477	0.8351	0.7162	0.7749	0.7018	0.9903	0.8405	0.9465	0.5790	0.6895	0.7582	0.8251	0.7846	0.8723	0.9619	0.8265	0.5656	0.6363	0.5703	0.5333	0.9185	0.4344	0.7931	0.8880	0.6813	0.5455	0.6994	0.5813	0.9810	0.6898	0.8941	0.4572	0.5481	0.5793	0.7227	0.5605	0.7893	0.9011	0.7158
0.9325	12.4444	28000	0.9490	0.6829	0.8029	0.9412	0.6928	0.7981	0.7084	0.6846	0.9600	0.6005	0.8961	0.9461	0.8434	0.7252	0.8082	0.6989	0.9895	0.8267	0.9466	0.5831	0.7146	0.7602	0.8321	0.7699	0.8824	0.9595	0.8409	0.5665	0.6315	0.5542	0.5330	0.9202	0.4673	0.7991	0.8880	0.6772	0.5431	0.7031	0.5830	0.9808	0.6882	0.8944	0.4585	0.5421	0.5785	0.7214	0.5722	0.7859	0.9029	0.7148
0.9039	12.8889	29000	0.9449	0.6862	0.7980	0.9417	0.6883	0.8044	0.6908	0.6811	0.9607	0.5807	0.8982	0.9472	0.8457	0.7476	0.7988	0.7018	0.9902	0.8109	0.9498	0.5777	0.6796	0.7492	0.8243	0.7585	0.8824	0.9601	0.8252	0.5666	0.6462	0.5669	0.5394	0.9201	0.4516	0.7983	0.8889	0.6767	0.5723	0.7040	0.5827	0.9811	0.6910	0.8946	0.4583	0.5528	0.5807	0.7217	0.5810	0.7921	0.9035	0.7126
0.936	13.3333	30000	0.9458	0.6848	0.8025	0.9414	0.7056	0.7957	0.7069	0.6933	0.9611	0.6014	0.8949	0.9441	0.8329	0.7333	0.8139	0.6904	0.9892	0.8341	0.9471	0.5854	0.7129	0.7506	0.8271	0.7628	0.8837	0.9617	0.8291	0.5683	0.6342	0.5463	0.5447	0.9202	0.4538	0.7957	0.8889	0.6829	0.5634	0.7052	0.5821	0.9809	0.6864	0.8944	0.4613	0.5455	0.5809	0.7217	0.5705	0.7921	0.9039	0.7260
0.9636	13.7778	31000	0.9470	0.6856	0.8003	0.9416	0.6972	0.8154	0.6967	0.6890	0.9617	0.6171	0.8940	0.9500	0.8438	0.7202	0.7881	0.7000	0.9909	0.8198	0.9431	0.5814	0.6997	0.7480	0.8320	0.7590	0.8715	0.9590	0.8301	0.5733	0.6329	0.5570	0.5440	0.9204	0.4667	0.7962	0.8880	0.6784	0.5502	0.7071	0.5841	0.9811	0.6878	0.8946	0.4602	0.5583	0.5827	0.7208	0.5780	0.7890	0.9039	0.7138
0.8986	14.2222	32000	0.9472	0.6864	0.8013	0.9417	0.6857	0.7904	0.6909	0.6910	0.9619	0.6165	0.8941	0.9449	0.8406	0.7531	0.8035	0.7000	0.9899	0.8171	0.9487	0.5884	0.6967	0.7541	0.8281	0.7664	0.8738	0.9624	0.8329	0.5684	0.6394	0.5559	0.5415	0.9206	0.4706	0.7962	0.8889	0.6812	0.5703	0.7083	0.5841	0.9811	0.6856	0.8946	0.4616	0.5548	0.5814	0.7220	0.5702	0.7908	0.9032	0.7170
0.9031	14.6667	33000	0.9465	0.6876	0.8034	0.9418	0.6800	0.8164	0.6949	0.6949	0.9600	0.6083	0.8963	0.9474	0.8455	0.7457	0.8079	0.7057	0.9903	0.8351	0.9468	0.5784	0.6875	0.7517	0.8324	0.7600	0.8898	0.9598	0.8427	0.5668	0.6378	0.5595	0.5403	0.9205	0.4618	0.7968	0.8891	0.6797	0.5718	0.7125	0.5860	0.9811	0.6872	0.8949	0.4599	0.5531	0.5835	0.7231	0.5797	0.7946	0.9055	0.7304

Framework versions

• Transformers 4.55.2
• Pytorch 2.7.1+cu118
• Datasets 4.0.0
• Tokenizers 0.21.4