scenario-KD-PR-MSV-D2_data-AmazonScience_massive_all_1_155

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: 1.4402
• Accuracy: 0.8575
• F1: 0.8343

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: 55
• 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
1.4282	0.27	5000	1.5913	0.8176	0.7771
1.261	0.53	10000	1.5198	0.8303	0.7951
1.2127	0.8	15000	1.4862	0.8367	0.8061
1.0927	1.07	20000	1.4876	0.8356	0.8032
1.0692	1.34	25000	1.4776	0.8376	0.8041
1.0537	1.6	30000	1.4714	0.8408	0.8101
1.0357	1.87	35000	1.4689	0.8389	0.8136
0.9737	2.14	40000	1.4509	0.8482	0.8270
0.9719	2.41	45000	1.4763	0.8410	0.8185
0.9692	2.67	50000	1.4650	0.8437	0.8210
0.978	2.94	55000	1.4563	0.8473	0.8266
0.9229	3.21	60000	1.4699	0.8437	0.8193
0.9323	3.47	65000	1.4563	0.8485	0.8247
0.9274	3.74	70000	1.4588	0.8469	0.8205
0.9177	4.01	75000	1.4698	0.8447	0.8227
0.8939	4.28	80000	1.4774	0.8452	0.8227
0.8926	4.54	85000	1.4697	0.8460	0.8185
0.8973	4.81	90000	1.4907	0.8414	0.8182
0.8696	5.08	95000	1.4705	0.8469	0.8217
0.8784	5.34	100000	1.4698	0.8467	0.8211
0.8803	5.61	105000	1.4622	0.8490	0.8240
0.8849	5.88	110000	1.4692	0.8467	0.8198
0.8614	6.15	115000	1.4861	0.8435	0.8138
0.8613	6.41	120000	1.4682	0.8489	0.8278
0.8704	6.68	125000	1.4696	0.8475	0.8284
0.8687	6.95	130000	1.4723	0.8470	0.8252
0.8469	7.22	135000	1.4750	0.8471	0.8252
0.858	7.48	140000	1.4675	0.8485	0.8221
0.8557	7.75	145000	1.4754	0.8462	0.8234
0.8428	8.02	150000	1.4801	0.8469	0.8240
0.8397	8.28	155000	1.4871	0.8462	0.8247
0.8342	8.55	160000	1.4838	0.8471	0.8237
0.8462	8.82	165000	1.4622	0.8474	0.8225
0.8249	9.09	170000	1.4768	0.8483	0.8270
0.8259	9.35	175000	1.4950	0.8443	0.8186
0.8304	9.62	180000	1.4682	0.8510	0.8261
0.8342	9.89	185000	1.4754	0.8480	0.8216
0.8165	10.15	190000	1.4817	0.8473	0.8253
0.8229	10.42	195000	1.4884	0.8463	0.8271
0.8237	10.69	200000	1.4906	0.8448	0.8207
0.8275	10.96	205000	1.4803	0.8470	0.8237
0.8124	11.22	210000	1.4765	0.8491	0.8304
0.8144	11.49	215000	1.5053	0.8446	0.8177
0.8146	11.76	220000	1.4706	0.8495	0.8261
0.8063	12.03	225000	1.4681	0.8526	0.8282
0.8141	12.29	230000	1.4725	0.8493	0.8286
0.8098	12.56	235000	1.4612	0.8511	0.8257
0.8159	12.83	240000	1.4833	0.8477	0.8226
0.805	13.09	245000	1.4828	0.8480	0.8235
0.8082	13.36	250000	1.4919	0.8456	0.8185
0.8146	13.63	255000	1.4872	0.8456	0.8217
0.8077	13.9	260000	1.4754	0.8492	0.8209
0.8027	14.16	265000	1.4714	0.8507	0.8257
0.807	14.43	270000	1.4948	0.8441	0.8234
0.8015	14.7	275000	1.4791	0.8497	0.8256
0.8018	14.96	280000	1.4805	0.8487	0.8290
0.8005	15.23	285000	1.4642	0.8523	0.8312
0.799	15.5	290000	1.4692	0.8522	0.8328
0.7994	15.77	295000	1.4783	0.8494	0.8285
0.7942	16.03	300000	1.4749	0.8507	0.8299
0.7924	16.3	305000	1.4702	0.8527	0.8301
0.7978	16.57	310000	1.4882	0.8482	0.8228
0.7953	16.84	315000	1.4707	0.8514	0.8269
0.7867	17.1	320000	1.4929	0.8482	0.8274
0.7888	17.37	325000	1.4731	0.8509	0.8272
0.7938	17.64	330000	1.4739	0.8511	0.8300
0.7903	17.9	335000	1.4537	0.8539	0.8306
0.7876	18.17	340000	1.4700	0.8516	0.8302
0.7902	18.44	345000	1.4813	0.8490	0.8238
0.7894	18.71	350000	1.4617	0.8542	0.8291
0.7872	18.97	355000	1.4713	0.8508	0.8272
0.7866	19.24	360000	1.4712	0.8510	0.8313
0.7828	19.51	365000	1.4642	0.8526	0.8305
0.7873	19.77	370000	1.4590	0.8533	0.8298
0.7781	20.04	375000	1.4681	0.8532	0.8287
0.7783	20.31	380000	1.4707	0.8524	0.8305
0.7851	20.58	385000	1.4626	0.8538	0.8300
0.7845	20.84	390000	1.4547	0.8543	0.8295
0.783	21.11	395000	1.4627	0.8537	0.8309
0.7783	21.38	400000	1.4627	0.8542	0.8324
0.7842	21.65	405000	1.4707	0.8510	0.8261
0.7816	21.91	410000	1.4629	0.8533	0.8298
0.7779	22.18	415000	1.4567	0.8536	0.8262
0.7816	22.45	420000	1.4574	0.8549	0.8326
0.7762	22.71	425000	1.4661	0.8533	0.8311
0.7808	22.98	430000	1.4623	0.8541	0.8306
0.7761	23.25	435000	1.4596	0.8542	0.8297
0.7745	23.52	440000	1.4608	0.8539	0.8296
0.7801	23.78	445000	1.4517	0.8555	0.8305
0.7705	24.05	450000	1.4598	0.8552	0.8327
0.7761	24.32	455000	1.4540	0.8547	0.8296
0.7771	24.58	460000	1.4568	0.8543	0.8308
0.7805	24.85	465000	1.4582	0.8540	0.8306
0.7707	25.12	470000	1.4573	0.8548	0.8311
0.7752	25.39	475000	1.4529	0.8551	0.8329
0.7771	25.65	480000	1.4570	0.8542	0.8317
0.7757	25.92	485000	1.4532	0.8556	0.8333
0.7717	26.19	490000	1.4548	0.8546	0.8303
0.7748	26.46	495000	1.4474	0.8565	0.8335
0.7741	26.72	500000	1.4517	0.8548	0.8311
0.7738	26.99	505000	1.4505	0.8555	0.8330
0.7722	27.26	510000	1.4503	0.8549	0.8307
0.7709	27.52	515000	1.4515	0.8550	0.8317
0.773	27.79	520000	1.4455	0.8560	0.8338
0.7726	28.06	525000	1.4477	0.8560	0.8333
0.7699	28.33	530000	1.4427	0.8564	0.8327
0.7718	28.59	535000	1.4491	0.8556	0.8312
0.7713	28.86	540000	1.4397	0.8575	0.8352
0.7681	29.13	545000	1.4452	0.8574	0.8350
0.7693	29.39	550000	1.4472	0.8559	0.8329
0.77	29.66	555000	1.4437	0.8566	0.8336
0.7689	29.93	560000	1.4402	0.8575	0.8343

Framework versions

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