-
Notifications
You must be signed in to change notification settings - Fork 0
/
resultsv2.log
146 lines (146 loc) · 24.5 KB
/
resultsv2.log
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
[2024-02-25 16:16:40,619] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='JackFram/llama-68m', target='meta-llama/Llama-2-7b-hf', dataset='dataset/c4_small.json', growmap='/home/zhuominc/workspace/Sequoia/68m_7b/growmaps/L40-C4-68m-7b-stochastic.pt', start=0, end=200, T=0.6, P=1.0, DP=0.99, D=1, B=10, W=32, M=384, seed=17, Mode='greedy', decay=0.85, negative=False, static=False, offloading=False)
64
total time :156.09007s, latency :0.00851s, decoding step: 18351, large model step: 5061, 3.6259632483698874
[2024-02-25 16:19:48,527] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='JackFram/llama-68m', target='meta-llama/Llama-2-7b-hf', dataset='cnn', growmap='/home/zhuominc/workspace/Sequoia/68m_7b/growmaps/L40-CNN-68m-7b-stochastic.pt', start=0, end=200, T=0.6, P=1.0, DP=0.99, D=1, B=10, W=32, M=384, seed=17, Mode='greedy', decay=0.85, negative=False, static=False, offloading=False)
64
total time :240.06998s, latency :0.00939s, decoding step: 25570, large model step: 7820, 3.2698209718670075
[2024-02-25 16:24:21,358] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='JackFram/llama-68m', target='meta-llama/Llama-2-7b-hf', dataset='openwebtext', growmap='/home/zhuominc/workspace/Sequoia/68m_7b/growmaps/L40-OpenWebText-68m-7b-stochastic.pt', start=0, end=200, T=0.6, P=1.0, DP=0.99, D=1, B=10, W=32, M=384, seed=17, Mode='greedy', decay=0.85, negative=False, static=False, offloading=False)
64
total time :244.51060s, latency :0.00984s, decoding step: 24853, large model step: 7966, 3.119884509163947
[2024-02-25 16:28:57,424] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='JackFram/llama-68m', target='meta-llama/Llama-2-7b-hf', dataset='dataset/c4_small.json', growmap='/home/zhuominc/workspace/Sequoia/68m_7b/growmaps/L40-C4-68m-7b-greedy.pt', start=0, end=200, T=0.6, P=1.0, DP=0.99, D=1, B=10, seed=17, W=32, M=384, Mode='greedy', decay=0.85, negative=False, static=False, offloading=False)
64
total time :126.64659s, latency :0.00656s, decoding step: 19306, large model step: 4126, 4.679108095007271
[2024-02-25 16:31:26,751] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='JackFram/llama-68m', target='meta-llama/Llama-2-7b-hf', dataset='cnn', growmap='/home/zhuominc/workspace/Sequoia/68m_7b/growmaps/L40-CNN-68m-7b-greedy.pt', start=0, end=200, T=0.6, P=1.0, DP=0.99, D=1, B=10, seed=17, W=32, M=384, Mode='greedy', decay=0.85, negative=False, static=False, offloading=False)
64
total time :204.77227s, latency :0.00786s, decoding step: 26054, large model step: 6881, 3.7863682604272633
[2024-02-25 16:35:19,154] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='JackFram/llama-68m', target='meta-llama/Llama-2-7b-hf', dataset='openwebtext', growmap='/home/zhuominc/workspace/Sequoia/68m_7b/growmaps/L40-OpenWebText-68m-7b-greedy.pt', start=0, end=200, T=0.6, P=1.0, DP=0.99, D=1, B=10, seed=17, W=32, M=384, Mode='greedy', decay=0.85, negative=False, static=False, offloading=False)
64
total time :210.55408s, latency :0.00829s, decoding step: 25404, large model step: 7093, 3.5815592838009307
[2024-02-25 16:39:11,813] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='JackFram/llama-68m', target='meta-llama/Llama-2-13b-hf', dataset='dataset/c4_small.json', growmap='/home/zhuominc/workspace/Sequoia/68m_13b/growmaps/L40-C4-68m-13b-stochastic.pt', start=0, end=200, T=0.6, P=1.0, DP=0.99, D=1, B=10, W=32, M=384, seed=17, Mode='greedy', decay=0.85, negative=False, static=False, offloading=False)
64
total time :273.58757s, latency :0.01520s, decoding step: 17999, large model step: 5040, 3.5712301587301587
[2024-02-25 16:44:22,144] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='JackFram/llama-68m', target='meta-llama/Llama-2-13b-hf', dataset='cnn', growmap='/home/zhuominc/workspace/Sequoia/68m_13b/growmaps/L40-CNN-68m-13b-stochastic.pt', start=0, end=200, T=0.6, P=1.0, DP=0.99, D=1, B=10, W=32, M=384, seed=17, Mode='greedy', decay=0.85, negative=False, static=False, offloading=False)
64
total time :420.12038s, latency :0.01682s, decoding step: 24973, large model step: 7753, 3.22107571262737
[2024-02-25 16:51:57,992] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='JackFram/llama-68m', target='meta-llama/Llama-2-13b-hf', dataset='openwebtext', growmap='/home/zhuominc/workspace/Sequoia/68m_13b/growmaps/L40-OpenWebText-68m-13b-stochastic.pt', start=0, end=200, T=0.6, P=1.0, DP=0.99, D=1, B=10, W=32, M=384, seed=17, Mode='greedy', decay=0.85, negative=False, static=False, offloading=False)
64
total time :436.57632s, latency :0.01746s, decoding step: 25009, large model step: 8136, 3.0738692232055063
[2024-02-25 16:59:47,839] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='JackFram/llama-68m', target='meta-llama/Llama-2-13b-hf', dataset='dataset/c4_small.json', growmap='/home/zhuominc/workspace/Sequoia/68m_13b/growmaps/L40-C4-68m-13b-greedy.pt', start=0, end=200, T=0.6, P=1.0, DP=0.99, D=1, B=10, seed=17, W=32, M=384, Mode='greedy', decay=0.85, negative=False, static=False, offloading=False)
64
total time :246.08269s, latency :0.01264s, decoding step: 19467, large model step: 4579, 4.251364926839921
[2024-02-25 17:04:21,169] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='JackFram/llama-68m', target='meta-llama/Llama-2-13b-hf', dataset='cnn', growmap='/home/zhuominc/workspace/Sequoia/68m_13b/growmaps/L40-CNN-68m-13b-greedy.pt', start=0, end=200, T=0.6, P=1.0, DP=0.99, D=1, B=10, seed=17, W=32, M=384, Mode='greedy', decay=0.85, negative=False, static=False, offloading=False)
64
total time :372.06839s, latency :0.01433s, decoding step: 25969, large model step: 7054, 3.681457329174936
[2024-02-25 17:11:00,572] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='JackFram/llama-68m', target='meta-llama/Llama-2-13b-hf', dataset='openwebtext', growmap='/home/zhuominc/workspace/Sequoia/68m_13b/growmaps/L40-OpenWebText-68m-13b-greedy.pt', start=0, end=200, T=0.6, P=1.0, DP=0.99, D=1, B=10, seed=17, W=32, M=384, Mode='greedy', decay=0.85, negative=False, static=False, offloading=False)
64
total time :387.51432s, latency :0.01507s, decoding step: 25722, large model step: 7302, 3.5225965488907147
[2024-02-25 17:17:53,516] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='JackFram/llama-160m', target='meta-llama/Llama-2-13b-hf', dataset='dataset/c4_small.json', growmap='/home/zhuominc/workspace/Sequoia/160m_13b/growmaps/L40-C4-160m-13b-stochastic.pt', start=0, end=200, T=0.6, P=1.0, DP=0.99, D=1, B=10, W=32, M=384, seed=17, Mode='greedy', decay=0.85, negative=False, static=False, offloading=False)
64
total time :271.65521s, latency :0.01503s, decoding step: 18074, large model step: 4314, 4.189615206305053
[2024-02-25 17:23:05,435] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='JackFram/llama-160m', target='meta-llama/Llama-2-13b-hf', dataset='cnn', growmap='/home/zhuominc/workspace/Sequoia/160m_13b/growmaps/L40-CNN-160m-13b-stochastic.pt', start=0, end=200, T=0.6, P=1.0, DP=0.99, D=1, B=10, W=32, M=384, seed=17, Mode='greedy', decay=0.85, negative=False, static=False, offloading=False)
64
total time :419.63604s, latency :0.01688s, decoding step: 24854, large model step: 6850, 3.6283211678832115
[2024-02-25 17:30:46,013] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='JackFram/llama-160m', target='meta-llama/Llama-2-13b-hf', dataset='openwebtext', growmap='/home/zhuominc/workspace/Sequoia/160m_13b/growmaps/L40-OpenWebText-160m-13b-stochastic.pt', start=0, end=200, T=0.6, P=1.0, DP=0.99, D=1, B=10, W=32, M=384, seed=17, Mode='greedy', decay=0.85, negative=False, static=False, offloading=False)
64
total time :431.37303s, latency :0.01734s, decoding step: 24883, large model step: 7038, 3.535521454958795
[2024-02-25 17:38:36,755] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='JackFram/llama-160m', target='meta-llama/Llama-2-13b-hf', dataset='dataset/c4_small.json', growmap='/home/zhuominc/workspace/Sequoia/160m_13b/growmaps/L40-C4-160m-13b-greedy.pt', start=0, end=200, T=0.6, P=1.0, DP=0.99, D=1, B=10, seed=17, W=32, M=384, Mode='greedy', decay=0.85, negative=False, static=False, offloading=False)
64
total time :253.20861s, latency :0.01295s, decoding step: 19560, large model step: 3878, 5.043837029396596
[2024-02-25 17:43:21,324] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='JackFram/llama-160m', target='meta-llama/Llama-2-13b-hf', dataset='cnn', growmap='/home/zhuominc/workspace/Sequoia/160m_13b/growmaps/L40-CNN-160m-13b-greedy.pt', start=0, end=200, T=0.6, P=1.0, DP=0.99, D=1, B=10, seed=17, W=32, M=384, Mode='greedy', decay=0.85, negative=False, static=False, offloading=False)
64
total time :382.39290s, latency :0.01468s, decoding step: 26044, large model step: 6184, 4.211513583441138
[2024-02-25 17:50:16,074] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='JackFram/llama-160m', target='meta-llama/Llama-2-13b-hf', dataset='openwebtext', growmap='/home/zhuominc/workspace/Sequoia/160m_13b/growmaps/L40-OpenWebText-160m-13b-greedy.pt', start=0, end=200, T=0.6, P=1.0, DP=0.99, D=1, B=10, seed=17, W=32, M=384, Mode='greedy', decay=0.85, negative=False, static=False, offloading=False)
64
total time :393.47873s, latency :0.01529s, decoding step: 25740, large model step: 6362, 4.045897516504244
[2024-02-25 17:57:19,108] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='meta-llama/Llama-2-7b-hf', target='meta-llama/Llama-2-70b-hf', dataset='cnn', growmap='/home/zhuominc/workspace/Sequoia/growmaps/L40-CNN-7b-70b-stochastic-2.pt', start=0, end=50, T=0.6, P=1.0, DP=0.99, D=1, B=10, W=32, M=1024, seed=17, Mode='greedy', decay=0.85, negative=False, static=False, offloading=True)
768
total time :4677.95298s, latency :0.72235s, decoding step: 6476, large model step: 696, 9.304597701149426
[2024-02-25 19:17:46,661] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='meta-llama/Llama-2-7b-hf', target='meta-llama/Llama-2-70b-hf', dataset='dataset/c4_small.json', growmap='/home/zhuominc/workspace/Sequoia/growmaps/L40-C4-7b-70b-stochastic-2.pt', start=0, end=50, T=0.6, P=1.0, DP=0.99, D=1, B=10, W=32, M=1024, seed=17, Mode='greedy', decay=0.85, negative=False, static=False, offloading=True)
768
total time :3208.59996s, latency :0.68036s, decoding step: 4716, large model step: 464, 10.163793103448276
[2024-02-25 20:13:53,626] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='meta-llama/Llama-2-7b-hf', target='meta-llama/Llama-2-70b-hf', dataset='openwebtext', growmap='/home/zhuominc/workspace/Sequoia/growmaps/L40-OpenWebText-7b-70b-stochastic-2.pt', start=0, end=50, T=0.6, P=1.0, DP=0.99, D=1, B=10, W=32, M=1024, seed=17, Mode='greedy', decay=0.85, negative=False, static=False, offloading=True)
768
total time :4617.27243s, latency :0.73605s, decoding step: 6273, large model step: 680, 9.225
[2024-02-25 21:33:23,032] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='meta-llama/Llama-2-7b-hf', target='meta-llama/Llama-2-70b-hf', dataset='cnn', growmap='/home/zhuominc/workspace/Sequoia/growmaps/L40-CNN-7b-70b-greedy-2.pt', start=0, end=50, T=0.6, P=1.0, DP=0.99, D=1, B=10, seed=17, W=32, M=1024, Mode='greedy', decay=0.85, negative=False, static=False, offloading=True)
768
total time :4675.32920s, latency :0.69760s, decoding step: 6702, large model step: 652, 10.279141104294478
[2024-02-25 22:53:34,120] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='meta-llama/Llama-2-7b-hf', target='meta-llama/Llama-2-70b-hf', dataset='dataset/c4_small.json', growmap='/home/zhuominc/workspace/Sequoia/growmaps/L40-C4-7b-70b-greedy-2.pt', start=0, end=50, T=0.6, P=1.0, DP=0.99, D=1, B=10, seed=17, W=32, M=1024, Mode='greedy', decay=0.85, negative=False, static=False, offloading=True)
768
total time :3089.91229s, latency :0.61199s, decoding step: 5049, large model step: 420, 12.021428571428572
[2024-02-25 23:47:13,037] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='meta-llama/Llama-2-7b-hf', target='meta-llama/Llama-2-70b-hf', dataset='openwebtext', growmap='/home/zhuominc/workspace/Sequoia/growmaps/L40-OpenWebText-7b-70b-greedy-2.pt', start=0, end=50, T=0.6, P=1.0, DP=0.99, D=1, B=10, seed=17, W=32, M=1024, Mode='greedy', decay=0.85, negative=False, static=False, offloading=True)
768
total time :4490.63440s, latency :0.69601s, decoding step: 6452, large model step: 644, 10.01863354037267
[2024-02-26 21:52:25,204] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='JackFram/llama-68m', target='meta-llama/Llama-2-7b-hf', dataset='dataset/c4_small.json', growmap='/home/zhuominc/workspace/Sequoia/68m_7b/growmaps/L40-C4-68m-7b-stochastic.pt', start=0, end=200, T=0.6, P=1.0, DP=0.99, D=1, B=10, W=32, M=384, seed=17, Mode='greedy', decay=0.85, negative=False, static=False, offloading=False)
[2024-02-26 21:52:39,909] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='JackFram/llama-68m', target='meta-llama/Llama-2-7b-hf', dataset='cnn', growmap='/home/zhuominc/workspace/Sequoia/68m_7b/growmaps/L40-CNN-68m-7b-stochastic.pt', start=0, end=200, T=0.6, P=1.0, DP=0.99, D=1, B=10, W=32, M=384, seed=17, Mode='greedy', decay=0.85, negative=False, static=False, offloading=False)
[2024-02-26 21:52:54,680] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='JackFram/llama-68m', target='meta-llama/Llama-2-7b-hf', dataset='openwebtext', growmap='/home/zhuominc/workspace/Sequoia/68m_7b/growmaps/L40-OpenWebText-68m-7b-stochastic.pt', start=0, end=200, T=0.6, P=1.0, DP=0.99, D=1, B=10, W=32, M=384, seed=17, Mode='greedy', decay=0.85, negative=False, static=False, offloading=False)
[2024-02-26 21:53:08,492] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='JackFram/llama-68m', target='meta-llama/Llama-2-7b-hf', dataset='dataset/c4_small.json', growmap='/home/zhuominc/workspace/Sequoia/68m_7b/growmaps/L40-C4-68m-7b-greedy.pt', start=0, end=200, T=0.6, P=1.0, DP=0.99, D=1, B=10, seed=17, W=32, M=384, Mode='greedy', decay=0.85, negative=False, static=False, offloading=False)
[2024-02-26 21:53:22,636] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='JackFram/llama-68m', target='meta-llama/Llama-2-7b-hf', dataset='cnn', growmap='/home/zhuominc/workspace/Sequoia/68m_7b/growmaps/L40-CNN-68m-7b-greedy.pt', start=0, end=200, T=0.6, P=1.0, DP=0.99, D=1, B=10, seed=17, W=32, M=384, Mode='greedy', decay=0.85, negative=False, static=False, offloading=False)
[2024-02-26 21:53:38,821] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='JackFram/llama-68m', target='meta-llama/Llama-2-7b-hf', dataset='openwebtext', growmap='/home/zhuominc/workspace/Sequoia/68m_7b/growmaps/L40-OpenWebText-68m-7b-greedy.pt', start=0, end=200, T=0.6, P=1.0, DP=0.99, D=1, B=10, seed=17, W=32, M=384, Mode='greedy', decay=0.85, negative=False, static=False, offloading=False)
[2024-02-26 21:53:51,972] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='JackFram/llama-68m', target='meta-llama/Llama-2-13b-hf', dataset='dataset/c4_small.json', growmap='/home/zhuominc/workspace/Sequoia/68m_13b/growmaps/L40-C4-68m-13b-stochastic.pt', start=0, end=200, T=0.6, P=1.0, DP=0.99, D=1, B=10, W=32, M=384, seed=17, Mode='greedy', decay=0.85, negative=False, static=False, offloading=False)
[2024-02-26 21:54:06,496] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='JackFram/llama-68m', target='meta-llama/Llama-2-13b-hf', dataset='cnn', growmap='/home/zhuominc/workspace/Sequoia/68m_13b/growmaps/L40-CNN-68m-13b-stochastic.pt', start=0, end=200, T=0.6, P=1.0, DP=0.99, D=1, B=10, W=32, M=384, seed=17, Mode='greedy', decay=0.85, negative=False, static=False, offloading=False)
[2024-02-26 21:54:20,501] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='JackFram/llama-68m', target='meta-llama/Llama-2-13b-hf', dataset='openwebtext', growmap='/home/zhuominc/workspace/Sequoia/68m_13b/growmaps/L40-OpenWebText-68m-13b-stochastic.pt', start=0, end=200, T=0.6, P=1.0, DP=0.99, D=1, B=10, W=32, M=384, seed=17, Mode='greedy', decay=0.85, negative=False, static=False, offloading=False)
[2024-02-26 21:54:32,580] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='JackFram/llama-68m', target='meta-llama/Llama-2-13b-hf', dataset='dataset/c4_small.json', growmap='/home/zhuominc/workspace/Sequoia/68m_13b/growmaps/L40-C4-68m-13b-greedy.pt', start=0, end=200, T=0.6, P=1.0, DP=0.99, D=1, B=10, seed=17, W=32, M=384, Mode='greedy', decay=0.85, negative=False, static=False, offloading=False)
[2024-02-26 21:54:46,212] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='JackFram/llama-68m', target='meta-llama/Llama-2-13b-hf', dataset='cnn', growmap='/home/zhuominc/workspace/Sequoia/68m_13b/growmaps/L40-CNN-68m-13b-greedy.pt', start=0, end=200, T=0.6, P=1.0, DP=0.99, D=1, B=10, seed=17, W=32, M=384, Mode='greedy', decay=0.85, negative=False, static=False, offloading=False)
[2024-02-26 21:55:00,572] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='JackFram/llama-68m', target='meta-llama/Llama-2-13b-hf', dataset='openwebtext', growmap='/home/zhuominc/workspace/Sequoia/68m_13b/growmaps/L40-OpenWebText-68m-13b-greedy.pt', start=0, end=200, T=0.6, P=1.0, DP=0.99, D=1, B=10, seed=17, W=32, M=384, Mode='greedy', decay=0.85, negative=False, static=False, offloading=False)
[2024-02-26 21:55:12,574] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='JackFram/llama-160m', target='meta-llama/Llama-2-13b-hf', dataset='dataset/c4_small.json', growmap='/home/zhuominc/workspace/Sequoia/160m_13b/growmaps/L40-C4-160m-13b-stochastic.pt', start=0, end=200, T=0.6, P=1.0, DP=0.99, D=1, B=10, W=32, M=384, seed=17, Mode='greedy', decay=0.85, negative=False, static=False, offloading=False)
[2024-02-26 21:55:25,756] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='JackFram/llama-160m', target='meta-llama/Llama-2-13b-hf', dataset='cnn', growmap='/home/zhuominc/workspace/Sequoia/160m_13b/growmaps/L40-CNN-160m-13b-stochastic.pt', start=0, end=200, T=0.6, P=1.0, DP=0.99, D=1, B=10, W=32, M=384, seed=17, Mode='greedy', decay=0.85, negative=False, static=False, offloading=False)
[2024-02-26 21:55:39,824] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='JackFram/llama-160m', target='meta-llama/Llama-2-13b-hf', dataset='openwebtext', growmap='/home/zhuominc/workspace/Sequoia/160m_13b/growmaps/L40-OpenWebText-160m-13b-stochastic.pt', start=0, end=200, T=0.6, P=1.0, DP=0.99, D=1, B=10, W=32, M=384, seed=17, Mode='greedy', decay=0.85, negative=False, static=False, offloading=False)
[2024-02-26 21:55:52,694] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='JackFram/llama-160m', target='meta-llama/Llama-2-13b-hf', dataset='dataset/c4_small.json', growmap='/home/zhuominc/workspace/Sequoia/160m_13b/growmaps/L40-C4-160m-13b-greedy.pt', start=0, end=200, T=0.6, P=1.0, DP=0.99, D=1, B=10, seed=17, W=32, M=384, Mode='greedy', decay=0.85, negative=False, static=False, offloading=False)
[2024-02-26 21:56:06,229] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='JackFram/llama-160m', target='meta-llama/Llama-2-13b-hf', dataset='cnn', growmap='/home/zhuominc/workspace/Sequoia/160m_13b/growmaps/L40-CNN-160m-13b-greedy.pt', start=0, end=200, T=0.6, P=1.0, DP=0.99, D=1, B=10, seed=17, W=32, M=384, Mode='greedy', decay=0.85, negative=False, static=False, offloading=False)
[2024-02-26 21:56:22,224] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='JackFram/llama-160m', target='meta-llama/Llama-2-13b-hf', dataset='openwebtext', growmap='/home/zhuominc/workspace/Sequoia/160m_13b/growmaps/L40-OpenWebText-160m-13b-greedy.pt', start=0, end=200, T=0.6, P=1.0, DP=0.99, D=1, B=10, seed=17, W=32, M=384, Mode='greedy', decay=0.85, negative=False, static=False, offloading=False)
[2024-02-26 21:56:34,976] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='meta-llama/Llama-2-7b-hf', target='meta-llama/Llama-2-70b-hf', dataset='cnn', growmap='/home/zhuominc/workspace/Sequoia/growmaps/L40-CNN-7b-70b-stochastic-2.pt', start=0, end=50, T=0.6, P=1.0, DP=0.99, D=1, B=10, W=32, M=1024, seed=17, Mode='greedy', decay=0.85, negative=False, static=False, offloading=True)
[2024-02-26 21:56:49,294] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='meta-llama/Llama-2-7b-hf', target='meta-llama/Llama-2-70b-hf', dataset='dataset/c4_small.json', growmap='/home/zhuominc/workspace/Sequoia/growmaps/L40-C4-7b-70b-stochastic-2.pt', start=0, end=50, T=0.6, P=1.0, DP=0.99, D=1, B=10, W=32, M=1024, seed=17, Mode='greedy', decay=0.85, negative=False, static=False, offloading=True)
[2024-02-26 21:57:01,429] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='meta-llama/Llama-2-7b-hf', target='meta-llama/Llama-2-70b-hf', dataset='openwebtext', growmap='/home/zhuominc/workspace/Sequoia/growmaps/L40-OpenWebText-7b-70b-stochastic-2.pt', start=0, end=50, T=0.6, P=1.0, DP=0.99, D=1, B=10, W=32, M=1024, seed=17, Mode='greedy', decay=0.85, negative=False, static=False, offloading=True)
[2024-02-26 21:57:12,883] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='meta-llama/Llama-2-7b-hf', target='meta-llama/Llama-2-70b-hf', dataset='cnn', growmap='/home/zhuominc/workspace/Sequoia/growmaps/L40-CNN-7b-70b-greedy-2.pt', start=0, end=50, T=0.6, P=1.0, DP=0.99, D=1, B=10, seed=17, W=32, M=1024, Mode='greedy', decay=0.85, negative=False, static=False, offloading=True)
[2024-02-26 21:57:26,952] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='meta-llama/Llama-2-7b-hf', target='meta-llama/Llama-2-70b-hf', dataset='dataset/c4_small.json', growmap='/home/zhuominc/workspace/Sequoia/growmaps/L40-C4-7b-70b-greedy-2.pt', start=0, end=50, T=0.6, P=1.0, DP=0.99, D=1, B=10, seed=17, W=32, M=1024, Mode='greedy', decay=0.85, negative=False, static=False, offloading=True)
[2024-02-26 21:57:39,710] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='meta-llama/Llama-2-7b-hf', target='meta-llama/Llama-2-70b-hf', dataset='openwebtext', growmap='/home/zhuominc/workspace/Sequoia/growmaps/L40-OpenWebText-7b-70b-greedy-2.pt', start=0, end=50, T=0.6, P=1.0, DP=0.99, D=1, B=10, seed=17, W=32, M=1024, Mode='greedy', decay=0.85, negative=False, static=False, offloading=True)
[2024-02-26 22:29:10,444] [INFO] [real_accelerator.py:161:get_accelerator] Setting ds_accelerator to cuda (auto detect)
Namespace(model='JackFram/llama-68m', target='meta-llama/Llama-2-7b-hf', dataset='dataset/c4_small.json', growmap='/home/zhuominc/workspace/Sequoia/68m_7b/growmaps/L40-C4-68m-7b-stochastic.pt', start=0, end=200, T=0.6, P=1.0, DP=0.99, D=1, B=10, W=32, M=384, seed=17, Mode='greedy', decay=0.85, negative=False, static=False, offloading=False)