diff --git "a/test-en-zh.comet" "b/test-en-zh.comet"
new file mode 100644--- /dev/null
+++ "b/test-en-zh.comet"
@@ -0,0 +1,1013 @@
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 0	score: 0.8653
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 1	score: 0.8951
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 2	score: 0.9136
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 3	score: 0.8818
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 4	score: 0.7968
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 5	score: 0.7593
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 6	score: 0.9571
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 7	score: 0.8347
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 8	score: 0.9129
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 9	score: 0.7614
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 10	score: 0.9254
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 11	score: 0.8565
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 12	score: 0.8848
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 13	score: 0.8660
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 14	score: 0.8247
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 15	score: 0.7870
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 16	score: 0.8698
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 17	score: 0.8823
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 18	score: 0.8954
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 19	score: 0.8900
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 20	score: 0.3769
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 21	score: 0.9622
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 22	score: 0.9415
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 23	score: 0.7119
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 24	score: 0.9458
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 25	score: 0.8266
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 26	score: 0.9167
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 27	score: 0.8999
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 28	score: 0.7582
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 29	score: 0.9240
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 30	score: 0.7726
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 31	score: 0.8829
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 32	score: 0.8829
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 33	score: 0.8941
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 34	score: 0.9486
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 35	score: 0.9206
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 36	score: 0.8902
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 37	score: 0.8402
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 38	score: 0.8753
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 39	score: 0.8988
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 40	score: 0.9162
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 41	score: 0.9235
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 42	score: 0.9233
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 43	score: 0.9419
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 44	score: 0.8534
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 45	score: 0.8812
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 46	score: 0.8890
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 47	score: 0.9120
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 48	score: 0.9435
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 49	score: 0.9614
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 50	score: 0.8970
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 51	score: 0.8985
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 52	score: 0.9219
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 53	score: 0.9283
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 54	score: 0.9054
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 55	score: 0.8991
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 56	score: 0.8470
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 57	score: 0.8907
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 58	score: 0.8678
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 59	score: 0.9547
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 60	score: 0.8591
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 61	score: 0.9066
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 62	score: 0.9054
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 63	score: 0.8873
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 64	score: 0.8598
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 65	score: 0.8430
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 66	score: 0.7424
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 67	score: 0.7413
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 68	score: 0.9466
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 69	score: 0.8841
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 70	score: 0.7786
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 71	score: 0.8148
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 72	score: 0.8834
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 73	score: 0.8923
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 74	score: 0.8803
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 75	score: 0.8219
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 76	score: 0.8798
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 77	score: 0.9209
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 78	score: 0.9403
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 79	score: 0.8452
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 80	score: 0.8736
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 81	score: 0.9596
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 82	score: 0.8895
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 83	score: 0.8833
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 84	score: 0.8602
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 85	score: 0.8545
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 86	score: 0.8573
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 87	score: 0.8704
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 88	score: 0.8881
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 89	score: 0.9223
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 90	score: 0.8563
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 91	score: 0.8775
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 92	score: 0.8569
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 93	score: 0.9453
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 94	score: 0.9211
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 95	score: 0.9450
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 96	score: 0.9428
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 97	score: 0.9073
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 98	score: 0.9015
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 99	score: 0.8935
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 100	score: 0.9423
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 101	score: 0.8192
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 102	score: 0.9179
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 103	score: 0.9472
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 104	score: 0.9267
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 105	score: 0.8690
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 106	score: 0.8730
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 107	score: 0.8069
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 108	score: 0.7495
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 109	score: 0.5004
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 110	score: 0.9119
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 111	score: 0.9335
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 112	score: 0.8955
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 113	score: 0.9226
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 114	score: 0.9203
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 115	score: 0.9399
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 116	score: 0.8829
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 117	score: 0.9076
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 118	score: 0.9315
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 119	score: 0.8843
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 120	score: 0.8706
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 121	score: 0.8964
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 122	score: 0.9240
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 123	score: 0.9140
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 124	score: 0.5351
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 125	score: 0.9226
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 126	score: 0.9226
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 127	score: 0.9251
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 128	score: 0.9226
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 129	score: 0.9472
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 130	score: 0.8673
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 131	score: 0.8919
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 132	score: 0.9302
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 133	score: 0.9037
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 134	score: 0.8567
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 135	score: 0.8831
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 136	score: 0.9618
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 137	score: 0.8852
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 138	score: 0.9126
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 139	score: 0.8814
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 140	score: 0.9170
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 141	score: 0.9040
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 142	score: 0.8991
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 143	score: 0.9207
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 144	score: 0.9522
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 145	score: 0.9149
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 146	score: 0.9401
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 147	score: 0.9097
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 148	score: 0.8778
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 149	score: 0.9212
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 150	score: 0.9245
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 151	score: 0.8639
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 152	score: 0.8581
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 153	score: 0.8472
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 154	score: 0.8629
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 155	score: 0.8661
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 156	score: 0.8782
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 157	score: 0.8597
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 158	score: 0.9329
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 159	score: 0.9023
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 160	score: 0.8841
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 161	score: 0.7398
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 162	score: 0.8404
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 163	score: 0.8956
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 164	score: 0.9204
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 165	score: 0.8383
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 166	score: 0.9296
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 167	score: 0.8551
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 168	score: 0.8636
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 169	score: 0.7985
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 170	score: 0.9123
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 171	score: 0.8984
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 172	score: 0.8698
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 173	score: 0.8772
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 174	score: 0.8717
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 175	score: 0.8795
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 176	score: 0.8348
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 177	score: 0.8865
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 178	score: 0.7402
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 179	score: 0.8933
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 180	score: 0.8313
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 181	score: 0.9211
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 182	score: 0.9243
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 183	score: 0.8827
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 184	score: 0.9066
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 185	score: 0.9258
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 186	score: 0.9677
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 187	score: 0.8807
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 188	score: 0.8792
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 189	score: 0.8119
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 190	score: 0.9070
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 191	score: 0.8578
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 192	score: 0.8398
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 193	score: 0.8520
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 194	score: 0.7764
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 195	score: 0.9046
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 196	score: 0.8656
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 197	score: 0.7914
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 198	score: 0.9344
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 199	score: 0.8029
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 200	score: 0.8002
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 201	score: 0.8330
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 202	score: 0.8652
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 203	score: 0.7952
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 204	score: 0.8694
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 205	score: 0.9153
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 206	score: 0.9352
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 207	score: 0.8493
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 208	score: 0.9190
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 209	score: 0.8087
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 210	score: 0.9225
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 211	score: 0.9297
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 212	score: 0.9391
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 213	score: 0.8861
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 214	score: 0.9132
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 215	score: 0.9333
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 216	score: 0.8863
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 217	score: 0.9516
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 218	score: 0.8888
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 219	score: 0.8613
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 220	score: 0.9179
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 221	score: 0.9464
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 222	score: 0.9265
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 223	score: 0.9169
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 224	score: 0.8902
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 225	score: 0.9088
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 226	score: 0.9108
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 227	score: 0.7491
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 228	score: 0.8595
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 229	score: 0.8769
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 230	score: 0.9695
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 231	score: 0.6289
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 232	score: 0.8604
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 233	score: 0.8530
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 234	score: 0.8559
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 235	score: 0.8600
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 236	score: 0.9219
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 237	score: 0.9006
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 238	score: 0.9102
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 239	score: 0.7877
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 240	score: 0.8536
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 241	score: 0.7978
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 242	score: 0.9261
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 243	score: 0.8534
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 244	score: 0.8583
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 245	score: 0.8835
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 246	score: 0.8053
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 247	score: 0.8777
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 248	score: 0.8476
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 249	score: 0.9010
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 250	score: 0.9001
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 251	score: 0.9267
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 252	score: 0.8859
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 253	score: 0.9691
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 254	score: 0.9056
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 255	score: 0.9130
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 256	score: 0.9101
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 257	score: 0.9405
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 258	score: 0.8642
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 259	score: 0.9104
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 260	score: 0.8823
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 261	score: 0.7780
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 262	score: 0.9021
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 263	score: 0.8516
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 264	score: 0.9053
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 265	score: 0.9101
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 266	score: 0.8720
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 267	score: 0.9117
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 268	score: 0.9055
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 269	score: 0.8971
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 270	score: 0.7347
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 271	score: 0.8035
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 272	score: 0.8852
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 273	score: 0.8433
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 274	score: 0.8070
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 275	score: 0.8561
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 276	score: 0.8041
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 277	score: 0.9598
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 278	score: 0.8845
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 279	score: 0.8119
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 280	score: 0.9054
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 281	score: 0.9027
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 282	score: 0.9458
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 283	score: 0.9056
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 284	score: 0.8260
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 285	score: 0.9409
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 286	score: 0.9037
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 287	score: 0.8928
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 288	score: 0.9166
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 289	score: 0.8845
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 290	score: 0.8834
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 291	score: 0.9345
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 292	score: 0.9283
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 293	score: 0.8687
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 294	score: 0.7973
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 295	score: 0.8227
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 296	score: 0.8062
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 297	score: 0.9169
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 298	score: 0.6026
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 299	score: 0.8996
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 300	score: 0.8907
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 301	score: 0.8918
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 302	score: 0.9276
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 303	score: 0.9415
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 304	score: 0.8787
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 305	score: 0.9317
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 306	score: 0.8032
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 307	score: 0.8969
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 308	score: 0.8180
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 309	score: 0.9183
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 310	score: 0.8831
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 311	score: 0.9196
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 312	score: 0.8833
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 313	score: 0.7995
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 314	score: 0.9083
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 315	score: 0.8985
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 316	score: 0.8468
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 317	score: 0.9183
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 318	score: 0.9222
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 319	score: 0.8233
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 320	score: 0.7560
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 321	score: 0.8353
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 322	score: 0.8676
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 323	score: 0.8768
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 324	score: 0.9209
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 325	score: 0.8879
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 326	score: 0.9059
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 327	score: 0.8990
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 328	score: 0.8339
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 329	score: 0.9263
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 330	score: 0.8910
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 331	score: 0.8912
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 332	score: 0.9035
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 333	score: 0.8997
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 334	score: 0.9533
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 335	score: 0.8833
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 336	score: 0.9329
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 337	score: 0.8087
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 338	score: 0.8674
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 339	score: 0.9245
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 340	score: 0.7756
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 341	score: 0.9152
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 342	score: 0.8449
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 343	score: 0.8448
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 344	score: 0.9232
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 345	score: 0.8385
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 346	score: 0.9227
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 347	score: 0.8172
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 348	score: 0.9090
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 349	score: 0.8593
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 350	score: 0.7713
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 351	score: 0.9128
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 352	score: 0.9270
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 353	score: 0.9009
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 354	score: 0.9095
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 355	score: 0.8878
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 356	score: 0.9443
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 357	score: 0.9038
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 358	score: 0.8987
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 359	score: 0.8056
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 360	score: 0.8286
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 361	score: 0.8799
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 362	score: 0.9305
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 363	score: 0.8339
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 364	score: 0.7264
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 365	score: 0.8658
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 366	score: 0.7898
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 367	score: 0.9053
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 368	score: 0.8578
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 369	score: 0.8806
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 370	score: 0.7961
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 371	score: 0.8112
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 372	score: 0.8906
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 373	score: 0.8252
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 374	score: 0.7969
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 375	score: 0.8363
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 376	score: 0.9126
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 377	score: 0.9460
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 378	score: 0.8303
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 379	score: 0.9406
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 380	score: 0.9574
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 381	score: 0.8351
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 382	score: 0.7914
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 383	score: 0.7915
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 384	score: 0.8993
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 385	score: 0.8618
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 386	score: 0.8843
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 387	score: 0.9007
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 388	score: 0.9044
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 389	score: 0.9146
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 390	score: 0.9136
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 391	score: 0.8640
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 392	score: 0.7812
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 393	score: 0.9001
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 394	score: 0.9527
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 395	score: 0.9479
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 396	score: 0.8178
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 397	score: 0.9481
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 398	score: 0.8564
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 399	score: 0.8808
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 400	score: 0.9327
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 401	score: 0.9147
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 402	score: 0.9069
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 403	score: 0.7239
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 404	score: 0.8308
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 405	score: 0.8656
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 406	score: 0.9570
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 407	score: 0.9359
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 408	score: 0.9451
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 409	score: 0.9052
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 410	score: 0.8692
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 411	score: 0.8816
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 412	score: 0.8483
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 413	score: 0.8548
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 414	score: 0.8835
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 415	score: 0.9068
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 416	score: 0.9106
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 417	score: 0.8942
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 418	score: 0.9363
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 419	score: 0.9403
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 420	score: 0.9354
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 421	score: 0.9475
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 422	score: 0.8963
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 423	score: 0.9029
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 424	score: 0.9071
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 425	score: 0.9334
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 426	score: 0.9288
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 427	score: 0.8549
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 428	score: 0.8967
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 429	score: 0.9376
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 430	score: 0.8204
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 431	score: 0.8069
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 432	score: 0.8066
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 433	score: 0.9386
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 434	score: 0.8161
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 435	score: 0.8514
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 436	score: 0.8543
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 437	score: 0.8671
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 438	score: 0.9308
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 439	score: 0.9090
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 440	score: 0.9221
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 441	score: 0.8901
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 442	score: 0.9563
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 443	score: 0.8667
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 444	score: 0.8179
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 445	score: 0.8908
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 446	score: 0.9064
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 447	score: 0.9034
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 448	score: 0.9436
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 449	score: 0.8294
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 450	score: 0.8978
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 451	score: 0.8770
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 452	score: 0.8928
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 453	score: 0.9053
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 454	score: 0.8539
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 455	score: 0.8510
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 456	score: 0.8258
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 457	score: 0.9545
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 458	score: 0.8793
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 459	score: 0.8136
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 460	score: 0.8987
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 461	score: 0.9206
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 462	score: 0.8909
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 463	score: 0.9234
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 464	score: 0.8752
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 465	score: 0.8918
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 466	score: 0.8939
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 467	score: 0.7607
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 468	score: 0.8021
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 469	score: 0.8843
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 470	score: 0.8039
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 471	score: 0.8516
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 472	score: 0.7984
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 473	score: 0.8897
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 474	score: 0.8243
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 475	score: 0.9006
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 476	score: 0.8958
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 477	score: 0.9334
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 478	score: 0.9112
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 479	score: 0.9136
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 480	score: 0.8750
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 481	score: 0.8367
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 482	score: 0.8878
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 483	score: 0.8474
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 484	score: 0.8464
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 485	score: 0.8460
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 486	score: 0.9132
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 487	score: 0.8865
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 488	score: 0.9358
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 489	score: 0.8820
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 490	score: 0.8835
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 491	score: 0.8724
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 492	score: 0.9455
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 493	score: 0.9164
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 494	score: 0.9238
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 495	score: 0.8558
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 496	score: 0.8499
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 497	score: 0.9186
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 498	score: 0.9274
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 499	score: 0.7642
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 500	score: 0.9173
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 501	score: 0.8877
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 502	score: 0.8267
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 503	score: 0.8691
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 504	score: 0.7924
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 505	score: 0.9334
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 506	score: 0.8906
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 507	score: 0.8314
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 508	score: 0.8467
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 509	score: 0.8003
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 510	score: 0.7787
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 511	score: 0.8200
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 512	score: 0.8475
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 513	score: 0.8841
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 514	score: 0.8405
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 515	score: 0.9187
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 516	score: 0.8527
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 517	score: 0.9059
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 518	score: 0.8738
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 519	score: 0.9437
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 520	score: 0.9074
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 521	score: 0.8729
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 522	score: 0.9054
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 523	score: 0.7941
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 524	score: 0.9031
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 525	score: 0.8988
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 526	score: 0.8902
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 527	score: 0.7139
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 528	score: 0.8027
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 529	score: 0.9326
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 530	score: 0.9112
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 531	score: 0.9172
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 532	score: 0.8785
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 533	score: 0.8737
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 534	score: 0.8511
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 535	score: 0.8170
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 536	score: 0.8528
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 537	score: 0.9220
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 538	score: 0.9037
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 539	score: 0.8980
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 540	score: 0.8944
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 541	score: 0.8792
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 542	score: 0.8148
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 543	score: 0.8550
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 544	score: 0.8852
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 545	score: 0.9516
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 546	score: 0.8380
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 547	score: 0.8485
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 548	score: 0.8713
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 549	score: 0.8718
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 550	score: 0.8807
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 551	score: 0.9368
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 552	score: 0.8589
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 553	score: 0.8438
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 554	score: 0.9076
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 555	score: 0.8164
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 556	score: 0.8643
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 557	score: 0.8370
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 558	score: 0.7946
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 559	score: 0.8424
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 560	score: 0.8733
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 561	score: 0.9476
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 562	score: 0.8904
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 563	score: 0.8792
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 564	score: 0.8669
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 565	score: 0.9160
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 566	score: 0.8635
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 567	score: 0.9126
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 568	score: 0.8127
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 569	score: 0.6983
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 570	score: 0.9130
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 571	score: 0.9087
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 572	score: 0.8029
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 573	score: 0.8875
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 574	score: 0.9298
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 575	score: 0.8820
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 576	score: 0.8736
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 577	score: 0.9021
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 578	score: 0.9073
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 579	score: 0.9097
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 580	score: 0.6612
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 581	score: 0.8878
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 582	score: 0.8956
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 583	score: 0.8479
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 584	score: 0.8820
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 585	score: 0.8800
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 586	score: 0.9088
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 587	score: 0.9179
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 588	score: 0.8234
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 589	score: 0.9331
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 590	score: 0.9058
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 591	score: 0.9430
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 592	score: 0.8377
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 593	score: 0.9073
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 594	score: 0.8639
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 595	score: 0.8945
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 596	score: 0.9049
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 597	score: 0.8898
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 598	score: 0.8552
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 599	score: 0.8780
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 600	score: 0.8791
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 601	score: 0.8754
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 602	score: 0.8859
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 603	score: 0.8676
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 604	score: 0.8650
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 605	score: 0.9263
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 606	score: 0.8668
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 607	score: 0.8547
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 608	score: 0.8648
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 609	score: 0.8249
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 610	score: 0.8928
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 611	score: 0.9355
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 612	score: 0.8866
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 613	score: 0.9136
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 614	score: 0.8790
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 615	score: 0.9151
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 616	score: 0.9416
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 617	score: 0.9446
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 618	score: 0.9285
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 619	score: 0.8880
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 620	score: 0.9228
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 621	score: 0.7826
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 622	score: 0.8988
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 623	score: 0.8879
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 624	score: 0.9090
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 625	score: 0.9112
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 626	score: 0.9369
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 627	score: 0.9292
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 628	score: 0.9116
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 629	score: 0.8764
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 630	score: 0.9153
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 631	score: 0.9651
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 632	score: 0.9283
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 633	score: 0.9340
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 634	score: 0.9787
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 635	score: 0.9235
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 636	score: 0.9150
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 637	score: 0.8982
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 638	score: 0.9289
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 639	score: 0.9147
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 640	score: 0.8911
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 641	score: 0.9524
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 642	score: 0.8492
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 643	score: 0.7329
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 644	score: 0.9155
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 645	score: 0.8951
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 646	score: 0.8424
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 647	score: 0.8654
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 648	score: 0.8134
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 649	score: 0.8514
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 650	score: 0.9618
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 651	score: 0.8933
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 652	score: 0.7478
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 653	score: 0.7614
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 654	score: 0.8526
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 655	score: 0.9196
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 656	score: 0.9296
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 657	score: 0.9320
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 658	score: 0.8056
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 659	score: 0.8086
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 660	score: 0.7986
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 661	score: 0.9497
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 662	score: 0.9540
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 663	score: 0.8904
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 664	score: 0.8933
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 665	score: 0.8887
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 666	score: 0.9050
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 667	score: 0.8372
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 668	score: 0.9307
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 669	score: 0.8995
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 670	score: 0.8914
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 671	score: 0.9152
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 672	score: 0.8594
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 673	score: 0.9235
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 674	score: 0.8920
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 675	score: 0.9104
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 676	score: 0.8910
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 677	score: 0.9108
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 678	score: 0.9287
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 679	score: 0.9073
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 680	score: 0.8686
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 681	score: 0.8996
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 682	score: 0.8833
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 683	score: 0.9051
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 684	score: 0.8948
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 685	score: 0.8973
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 686	score: 0.8932
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 687	score: 0.8522
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 688	score: 0.9287
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 689	score: 0.8639
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 690	score: 0.9111
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 691	score: 0.8749
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 692	score: 0.8885
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 693	score: 0.8057
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 694	score: 0.8456
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 695	score: 0.8246
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 696	score: 0.8681
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 697	score: 0.9499
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 698	score: 0.8724
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 699	score: 0.8860
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 700	score: 0.8944
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 701	score: 0.9087
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 702	score: 0.8197
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 703	score: 0.9241
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 704	score: 0.8592
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 705	score: 0.8872
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 706	score: 0.8336
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 707	score: 0.8582
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 708	score: 0.8610
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 709	score: 0.9067
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 710	score: 0.9515
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 711	score: 0.8771
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 712	score: 0.9260
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 713	score: 0.8865
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 714	score: 0.8918
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 715	score: 0.8329
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 716	score: 0.8750
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 717	score: 0.8326
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 718	score: 0.9068
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 719	score: 0.9164
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 720	score: 0.9192
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 721	score: 0.9268
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 722	score: 0.8794
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 723	score: 0.9256
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 724	score: 0.9258
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 725	score: 0.9561
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 726	score: 0.8086
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 727	score: 0.8522
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 728	score: 0.7793
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 729	score: 0.8549
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 730	score: 0.9583
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 731	score: 0.9302
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 732	score: 0.8647
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 733	score: 0.8246
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 734	score: 0.8480
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 735	score: 0.8953
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 736	score: 0.9002
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 737	score: 0.8014
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 738	score: 0.9235
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 739	score: 0.8139
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 740	score: 0.8551
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 741	score: 0.9085
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 742	score: 0.8099
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 743	score: 0.9130
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 744	score: 0.9294
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 745	score: 0.8872
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 746	score: 0.8587
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 747	score: 0.8395
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 748	score: 0.9500
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 749	score: 0.8514
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 750	score: 0.8013
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 751	score: 0.7989
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 752	score: 0.7005
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 753	score: 0.8825
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 754	score: 0.9034
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 755	score: 0.9043
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 756	score: 0.8932
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 757	score: 0.8901
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 758	score: 0.9018
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 759	score: 0.8139
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 760	score: 0.9479
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 761	score: 0.8968
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 762	score: 0.9639
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 763	score: 0.9109
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 764	score: 0.8747
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 765	score: 0.8598
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 766	score: 0.9106
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 767	score: 0.9184
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 768	score: 0.9092
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 769	score: 0.9620
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 770	score: 0.8900
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 771	score: 0.9233
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 772	score: 0.7944
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 773	score: 0.9306
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 774	score: 0.8748
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 775	score: 0.8693
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 776	score: 0.7453
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 777	score: 0.8842
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 778	score: 0.8838
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 779	score: 0.8241
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 780	score: 0.9173
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 781	score: 0.8156
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 782	score: 0.9342
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 783	score: 0.7507
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 784	score: 0.9259
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 785	score: 0.8221
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 786	score: 0.9344
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 787	score: 0.8710
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 788	score: 0.9226
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 789	score: 0.8822
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 790	score: 0.8922
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 791	score: 0.8447
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 792	score: 0.9015
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 793	score: 0.8968
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 794	score: 0.8454
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 795	score: 0.8890
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 796	score: 0.9130
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 797	score: 0.8754
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 798	score: 0.8739
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 799	score: 0.7534
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 800	score: 0.9061
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 801	score: 0.8948
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 802	score: 0.8869
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 803	score: 0.9159
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 804	score: 0.8433
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 805	score: 0.9201
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 806	score: 0.9116
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 807	score: 0.8371
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 808	score: 0.8733
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 809	score: 0.8321
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 810	score: 0.8362
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 811	score: 0.8116
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 812	score: 0.9610
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 813	score: 0.9205
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 814	score: 0.9156
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 815	score: 0.8470
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 816	score: 0.8487
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 817	score: 0.9048
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 818	score: 0.8948
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 819	score: 0.8807
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 820	score: 0.8973
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 821	score: 0.9458
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 822	score: 0.8760
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 823	score: 0.8867
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 824	score: 0.9217
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 825	score: 0.8836
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 826	score: 0.8890
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 827	score: 0.8359
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 828	score: 0.8441
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 829	score: 0.8958
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 830	score: 0.8396
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 831	score: 0.8377
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 832	score: 0.8637
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 833	score: 0.8007
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 834	score: 0.8790
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 835	score: 0.7995
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 836	score: 0.8343
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 837	score: 0.8759
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 838	score: 0.8733
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 839	score: 0.8731
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 840	score: 0.8279
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 841	score: 0.8124
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 842	score: 0.8405
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 843	score: 0.9270
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 844	score: 0.8764
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 845	score: 0.8643
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 846	score: 0.7045
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 847	score: 0.8545
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 848	score: 0.8542
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 849	score: 0.4952
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 850	score: 0.8709
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 851	score: 0.9269
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 852	score: 0.9319
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 853	score: 0.8830
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 854	score: 0.8602
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 855	score: 0.9081
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 856	score: 0.8050
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 857	score: 0.9243
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 858	score: 0.8596
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 859	score: 0.9068
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 860	score: 0.9096
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 861	score: 0.9282
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 862	score: 0.9233
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 863	score: 0.8926
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 864	score: 0.8674
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 865	score: 0.8648
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 866	score: 0.9050
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 867	score: 0.8087
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 868	score: 0.9284
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 869	score: 0.8704
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 870	score: 0.9305
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 871	score: 0.8041
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 872	score: 0.8563
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 873	score: 0.8684
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 874	score: 0.9047
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 875	score: 0.8705
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 876	score: 0.9524
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 877	score: 0.8807
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 878	score: 0.9342
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 879	score: 0.8449
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 880	score: 0.8650
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 881	score: 0.9154
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 882	score: 0.9394
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 883	score: 0.8962
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 884	score: 0.7922
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 885	score: 0.9239
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 886	score: 0.9074
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 887	score: 0.7394
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 888	score: 0.7991
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 889	score: 0.8719
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 890	score: 0.7210
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 891	score: 0.8284
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 892	score: 0.7503
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 893	score: 0.8991
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 894	score: 0.8861
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 895	score: 0.9047
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 896	score: 0.8927
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 897	score: 0.8723
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 898	score: 0.8632
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 899	score: 0.8461
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 900	score: 0.8429
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 901	score: 0.7242
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 902	score: 0.9098
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 903	score: 0.8508
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 904	score: 0.8861
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 905	score: 0.8440
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 906	score: 0.7567
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 907	score: 0.8698
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 908	score: 0.8512
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 909	score: 0.9461
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 910	score: 0.8695
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 911	score: 0.8473
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 912	score: 0.9405
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 913	score: 0.7532
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 914	score: 0.8570
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 915	score: 0.8584
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 916	score: 0.9571
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 917	score: 0.8869
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 918	score: 0.9088
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 919	score: 0.9031
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 920	score: 0.9045
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 921	score: 0.8558
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 922	score: 0.9176
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 923	score: 0.8691
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 924	score: 0.9060
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 925	score: 0.8521
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 926	score: 0.9076
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 927	score: 0.8521
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 928	score: 0.8859
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 929	score: 0.8605
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 930	score: 0.8720
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 931	score: 0.9048
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 932	score: 0.8228
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 933	score: 0.8612
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 934	score: 0.8871
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 935	score: 0.8146
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 936	score: 0.8848
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 937	score: 0.9547
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 938	score: 0.9318
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 939	score: 0.9102
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 940	score: 0.9075
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 941	score: 0.9407
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 942	score: 0.9329
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 943	score: 0.8811
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 944	score: 0.8584
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 945	score: 0.7899
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 946	score: 0.8786
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 947	score: 0.8934
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 948	score: 0.8043
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 949	score: 0.7819
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 950	score: 0.8552
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 951	score: 0.8324
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 952	score: 0.8389
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 953	score: 0.8617
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 954	score: 0.8774
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 955	score: 0.9097
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 956	score: 0.9415
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 957	score: 0.9400
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 958	score: 0.8992
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 959	score: 0.8996
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 960	score: 0.8950
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 961	score: 0.9243
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 962	score: 0.8820
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 963	score: 0.8417
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 964	score: 0.9628
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 965	score: 0.9000
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 966	score: 0.9205
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 967	score: 0.8190
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 968	score: 0.8105
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 969	score: 0.7718
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 970	score: 0.9122
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 971	score: 0.8763
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 972	score: 0.7803
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 973	score: 0.9465
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 974	score: 0.8824
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 975	score: 0.9272
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 976	score: 0.8763
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 977	score: 0.8722
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 978	score: 0.9218
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 979	score: 0.8920
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 980	score: 0.8823
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 981	score: 0.8927
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 982	score: 0.9053
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 983	score: 0.8320
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 984	score: 0.9046
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 985	score: 0.8937
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 986	score: 0.8815
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 987	score: 0.8974
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 988	score: 0.7950
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 989	score: 0.8632
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 990	score: 0.9026
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 991	score: 0.8716
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 992	score: 0.8952
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 993	score: 0.8740
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 994	score: 0.8770
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 995	score: 0.8861
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 996	score: 0.9068
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 997	score: 0.9563
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 998	score: 0.9299
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 999	score: 0.8325
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 1000	score: 0.8183
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 1001	score: 0.8989
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 1002	score: 0.8503
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 1003	score: 0.8779
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 1004	score: 0.9521
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 1005	score: 0.9168
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 1006	score: 0.8819
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 1007	score: 0.9132
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 1008	score: 0.9047
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 1009	score: 0.9174
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 1010	score: 0.9290
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	Segment 1011	score: 0.9221
+/beacon-scratch/tongzh24/ALMA-checkpoint/exp_50_languages/alma-13b-sft-group-6-max-tokens-512//test-en-zh	score: 0.8775