Table 6.

Results for filter pruning with no retraining.

	Pruned Filters	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15
Model 1	Loss	0.22	0.20	0.23	0.24	0.71	1.22	0.85	1.59	1.31	1.48	2.32	2.10	2.09	1.79	1.77
	Accuracy (%)	93.97	93.80	93.97	93.97	78.66	59.38	66.95	44.92	49.05	49.40	37.52	46.99	46.30	24.44	16.87
	Compression Rate	1.06	1.14	1.22	1.31	1.41	1.53	1.67	1.83	2.02	2.25	2.52	2.86	3.30	3.88	4.67
	Flops	0.89	0.78	0.68	0.59	0.50	0.42	0.35	0.28	0.22	0.17	0.12	0.09	0.06	0.03	0.01
Model 2	Loss	0.24	0.25	0.67	2.58	1.35	0.85	1.08	0.66	0.80	0.88	1.13	1.24	1.28	1.46	1.47
	Accuracy (%)	93.80	94.15	76.07	51.63	62.48	74.18	69.88	71.25	63.51	70.74	58.00	34.42	27.37	16.87	16.87
	Compression Rate	1.08	1.16	1.26	1.38	1.52	1.68	1.89	2.13	2.44	2.84	3.39	4.17	5.38	7.51	12.17
	Flops	0.89	0.78	0.68	0.59	0.50	0.42	0.35	0.28	0.22	0.17	0.12	0.08	0.05	0.03	0.01
Model 3	Loss	1.16	4.90	2.25	1.35	1.40	1.47	1.66	-	-	-	-	-	-	-	-
	Accuracy (%)	71.43	27.37	25.13	37.18	29.60	35.28	35.28	-	-	-	-	-	-	-	-
	Compression Rate	1.14	1.33	1.59	1.96	2.53	3.53	5.72	-	-	-	-	-	-	-	-
	Flops	0.79	0.61	0.44	0.31	0.19	0.10	0.04	-	-	-	-	-	-	-	-
Model 4	Loss	0.22	0.61	2.78	9.03	1.64	8.08	1.68	-	-	-	-	-	-	-	-
	Accuracy (%)	94.15	80.38	57.38	16.87	21.69	16.87	16.87	-	-	-	-	-	-	-	-
	Compression Rate	1.10	1.23	1.38	1.56	1.80	2.12	2.55	-	-	-	-	-	-	-	-
	Flops	0.79	0.61	0.45	0.31	0.20	0.11	0.05	-	-	-	-	-	-	-	-
Model 5	Loss	0.33	2.44	1.56	0.52	0.48	0.53	0.72	1.14	1.26	1.21	1.90	1.67	1.79	1.61	1.64
	Accuracy (%)	93.64	66.95	67.99	78.83	79.69	78.83	71.60	47.50	42.68	40.79	35.97	35.46	34.08	35.46	35.46
	Compression Rate	1.06	1.14	1.22	1.31	1.41	1.53	1.67	1.83	2.02	2.25	2.52	2.86	3.30	3.88	4.67
	Flops	0.89	0.78	0.68	0.59	0.50	0.42	0.35	0.28	0.22	0.17	0.13	0.09	0.0568	0.03	0.01