scenario-KD-PR-MSV-D2_data-AmazonScience_massive_all_1_166

This model is a fine-tuned version of haryoaw/scenario-MDBT-TCR_data-AmazonScience_massive_all_1_1 on the massive dataset. It achieves the following results on the evaluation set:

• Loss: 1.4485
• Accuracy: 0.8574
• F1: 0.8341

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: 32
• eval_batch_size: 32
• seed: 66
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
1.4298	0.27	5000	1.5967	0.8151	0.7666
1.2734	0.53	10000	1.5427	0.8274	0.7960
1.1941	0.8	15000	1.4827	0.8390	0.8068
1.0639	1.07	20000	1.4810	0.8417	0.8047
1.0604	1.34	25000	1.4630	0.8423	0.8143
1.0654	1.6	30000	1.4684	0.8419	0.8110
1.0463	1.87	35000	1.4657	0.8411	0.8138
0.968	2.14	40000	1.4824	0.8423	0.8168
0.9725	2.41	45000	1.4736	0.8450	0.8233
0.9661	2.67	50000	1.4821	0.8410	0.8187
0.9729	2.94	55000	1.4726	0.8433	0.8198
0.9376	3.21	60000	1.4574	0.8484	0.8260
0.9369	3.47	65000	1.4585	0.8462	0.8227
0.9335	3.74	70000	1.4680	0.8440	0.8141
0.9214	4.01	75000	1.4614	0.8489	0.8250
0.8969	4.28	80000	1.4837	0.8443	0.8192
0.9043	4.54	85000	1.4667	0.8455	0.8190
0.9048	4.81	90000	1.4842	0.8445	0.8158
0.8714	5.08	95000	1.4747	0.8458	0.8190
0.8811	5.34	100000	1.4862	0.8462	0.8213
0.8825	5.61	105000	1.4670	0.8462	0.8244
0.8876	5.88	110000	1.4813	0.8446	0.8228
0.8584	6.15	115000	1.4841	0.8454	0.8241
0.8612	6.41	120000	1.4797	0.8472	0.8238
0.864	6.68	125000	1.4770	0.8485	0.8221
0.8711	6.95	130000	1.4883	0.8459	0.8204
0.8464	7.22	135000	1.4838	0.8478	0.8223
0.8622	7.48	140000	1.4896	0.8453	0.8225
0.8519	7.75	145000	1.4882	0.8464	0.8215
0.8475	8.02	150000	1.4759	0.8501	0.8269
0.8429	8.28	155000	1.4686	0.8501	0.8283
0.8407	8.55	160000	1.4810	0.8488	0.8243
0.8429	8.82	165000	1.4810	0.8484	0.8253
0.8337	9.09	170000	1.4810	0.8484	0.8275
0.8283	9.35	175000	1.4905	0.8471	0.8241
0.8369	9.62	180000	1.4951	0.8458	0.8192
0.837	9.89	185000	1.4808	0.8489	0.8251
0.8277	10.15	190000	1.4792	0.8514	0.8269
0.8264	10.42	195000	1.4921	0.8471	0.8203
0.8326	10.69	200000	1.4932	0.8446	0.8187
0.8226	10.96	205000	1.4738	0.8511	0.8297
0.8214	11.22	210000	1.5004	0.8452	0.8218
0.8184	11.49	215000	1.4897	0.8477	0.8254
0.8191	11.76	220000	1.4932	0.8476	0.8225
0.8067	12.03	225000	1.4959	0.8475	0.8248
0.8098	12.29	230000	1.4868	0.8487	0.8243
0.8194	12.56	235000	1.4966	0.8451	0.8213
0.8128	12.83	240000	1.4956	0.8484	0.8244
0.8045	13.09	245000	1.4935	0.8488	0.8214
0.8093	13.36	250000	1.4740	0.8509	0.8262
0.8107	13.63	255000	1.4858	0.8492	0.8226
0.8129	13.9	260000	1.4863	0.8489	0.8252
0.8038	14.16	265000	1.4865	0.8491	0.8261
0.805	14.43	270000	1.4785	0.8523	0.8286
0.8096	14.7	275000	1.4740	0.8505	0.8273
0.8101	14.96	280000	1.4742	0.8512	0.8273
0.8008	15.23	285000	1.4889	0.8495	0.8252
0.7987	15.5	290000	1.4783	0.8509	0.8285
0.7991	15.77	295000	1.4758	0.8523	0.8266
0.7949	16.03	300000	1.4831	0.8511	0.8291
0.794	16.3	305000	1.4831	0.8497	0.8239
0.799	16.57	310000	1.4948	0.8480	0.8259
0.7973	16.84	315000	1.4819	0.8501	0.8253
0.7901	17.1	320000	1.4868	0.8498	0.8263
0.7955	17.37	325000	1.5029	0.8469	0.8222
0.7954	17.64	330000	1.4902	0.8492	0.8234
0.7926	17.9	335000	1.4990	0.8467	0.8232
0.7938	18.17	340000	1.4764	0.8511	0.8236
0.7926	18.44	345000	1.4741	0.8523	0.8310
0.7928	18.71	350000	1.4800	0.8516	0.8252
0.793	18.97	355000	1.4695	0.8525	0.8260
0.7902	19.24	360000	1.4677	0.8529	0.8260
0.7887	19.51	365000	1.4774	0.8507	0.8259
0.7935	19.77	370000	1.4762	0.8519	0.8272
0.7828	20.04	375000	1.4700	0.8533	0.8313
0.788	20.31	380000	1.4768	0.8501	0.8267
0.7845	20.58	385000	1.4724	0.8532	0.8302
0.7907	20.84	390000	1.4712	0.8504	0.8250
0.7901	21.11	395000	1.4787	0.8520	0.8269
0.7809	21.38	400000	1.4745	0.8532	0.8270
0.7822	21.65	405000	1.4824	0.8499	0.8260
0.7882	21.91	410000	1.4711	0.8534	0.8296
0.7825	22.18	415000	1.4755	0.8516	0.8264
0.7839	22.45	420000	1.4754	0.8519	0.8258
0.7836	22.71	425000	1.4671	0.8546	0.8301
0.7816	22.98	430000	1.4654	0.8532	0.8297
0.7786	23.25	435000	1.4786	0.8527	0.8278
0.7841	23.52	440000	1.4678	0.8544	0.8309
0.7802	23.78	445000	1.4735	0.8526	0.8304
0.7801	24.05	450000	1.4703	0.8521	0.8289
0.7796	24.32	455000	1.4668	0.8543	0.8323
0.78	24.58	460000	1.4628	0.8549	0.8332
0.7796	24.85	465000	1.4616	0.8549	0.8312
0.7797	25.12	470000	1.4615	0.8540	0.8295
0.7791	25.39	475000	1.4615	0.8542	0.8293
0.7797	25.65	480000	1.4636	0.8543	0.8305
0.7779	25.92	485000	1.4599	0.8547	0.8301
0.7764	26.19	490000	1.4543	0.8565	0.8322
0.7785	26.46	495000	1.4557	0.8551	0.8309
0.7738	26.72	500000	1.4580	0.8549	0.8327
0.7776	26.99	505000	1.4516	0.8568	0.8317
0.7756	27.26	510000	1.4575	0.8553	0.8312
0.7759	27.52	515000	1.4503	0.8567	0.8343
0.7748	27.79	520000	1.4507	0.8569	0.8327
0.7757	28.06	525000	1.4528	0.8556	0.8309
0.7703	28.33	530000	1.4584	0.8550	0.8305
0.7745	28.59	535000	1.4507	0.8569	0.8333
0.7754	28.86	540000	1.4547	0.8558	0.8324
0.7743	29.13	545000	1.4499	0.8568	0.8329
0.7759	29.39	550000	1.4498	0.8566	0.8319
0.7729	29.66	555000	1.4474	0.8571	0.8328
0.7734	29.93	560000	1.4485	0.8574	0.8341

Framework versions

• Transformers 4.33.3
• Pytorch 2.1.1+cu121
• Datasets 2.14.5
• Tokenizers 0.13.3