rockst4r4 commited on
Commit
632e211
·
verified ·
1 Parent(s): cca30ec

End of training

Browse files
adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:5ac1c82aaa66c513fd77814dbee1cf88eb092c2bcef2c9de69f3f8af333eb621
3
  size 73911112
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:06a74e6e2d014916f90ee690dce3c50c22bdd08d43b69654327d11f78b7067b2
3
  size 73911112
all_results.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
  "total_flos": 0.0,
3
- "train_loss": 3.0266644716903102e-05,
4
- "train_runtime": 1124.2903,
5
- "train_samples": 14,
6
- "train_samples_per_second": 0.285,
7
- "train_steps_per_second": 0.018
8
  }
 
1
  {
2
  "total_flos": 0.0,
3
+ "train_loss": 3.4546391179901545e-05,
4
+ "train_runtime": 941.4145,
5
+ "train_samples": 15,
6
+ "train_samples_per_second": 0.34,
7
+ "train_steps_per_second": 0.021
8
  }
train_results.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
  "total_flos": 0.0,
3
- "train_loss": 3.0266644716903102e-05,
4
- "train_runtime": 1124.2903,
5
- "train_samples": 14,
6
- "train_samples_per_second": 0.285,
7
- "train_steps_per_second": 0.018
8
  }
 
1
  {
2
  "total_flos": 0.0,
3
+ "train_loss": 3.4546391179901545e-05,
4
+ "train_runtime": 941.4145,
5
+ "train_samples": 15,
6
+ "train_samples_per_second": 0.34,
7
+ "train_steps_per_second": 0.021
8
  }
trainer_state.json CHANGED
@@ -10,203 +10,203 @@
10
  "is_world_process_zero": true,
11
  "log_history": [
12
  {
13
- "completion_length": 229.0,
14
- "epoch": 0.5714285714285714,
15
- "grad_norm": 1.1000111103057861,
16
- "kl": 0.030057710246182978,
17
  "learning_rate": 5e-07,
18
  "loss": 0.0,
19
- "reward": 5.323144108057022,
20
- "reward_std": 2.5642264559865,
21
- "rewards/concensus_correctness_reward_func": 1.693624995648861,
22
- "rewards/consensus_reward_func": 1.125,
23
  "rewards/cumulative_reward_2": 0.0,
24
- "rewards/final_correctness_reward_func": 1.25,
25
- "rewards/question_recreation_reward_func": 0.49539405573159456,
26
- "rewards/soft_format_reward_func": 0.015625,
27
- "rewards/strict_format_reward_func": 0.03125,
28
- "rewards/xmlcount_reward_func": 0.7122499942779541,
29
  "step": 2
30
  },
31
  {
32
- "completion_length": 220.125,
33
  "epoch": 1.0,
34
- "grad_norm": 0.7904133796691895,
35
- "kl": 0.032162196934223175,
36
  "learning_rate": 4.864543104251586e-07,
37
  "loss": 0.0,
38
- "reward": 5.786092758178711,
39
- "reward_std": 2.145986179510752,
40
- "rewards/concensus_correctness_reward_func": 1.8214166661103566,
41
- "rewards/consensus_reward_func": 1.1666666666666667,
42
  "rewards/cumulative_reward_2": 0.0,
43
- "rewards/final_correctness_reward_func": 1.3333333333333333,
44
- "rewards/question_recreation_reward_func": 0.5735511605938276,
45
  "rewards/soft_format_reward_func": 0.0,
46
- "rewards/strict_format_reward_func": 0.041666666666666664,
47
- "rewards/xmlcount_reward_func": 0.8494583268960317,
48
  "step": 4
49
  },
50
  {
51
- "completion_length": 222.65625,
52
- "epoch": 1.5714285714285714,
53
- "grad_norm": 1.0802266597747803,
54
- "kl": 0.03376023273449391,
55
  "learning_rate": 4.472851273490984e-07,
56
  "loss": 0.0,
57
- "reward": 4.889157712459564,
58
- "reward_std": 2.493479371070862,
59
- "rewards/concensus_correctness_reward_func": 1.5853749886155128,
60
- "rewards/consensus_reward_func": 1.125,
61
  "rewards/cumulative_reward_2": 0.0,
62
- "rewards/final_correctness_reward_func": 1.0,
63
- "rewards/question_recreation_reward_func": 0.41631394904106855,
64
  "rewards/soft_format_reward_func": 0.0,
65
- "rewards/strict_format_reward_func": 0.015625,
66
- "rewards/xmlcount_reward_func": 0.746843758970499,
67
  "step": 6
68
  },
69
  {
70
- "completion_length": 214.08333333333334,
71
  "epoch": 2.0,
72
- "grad_norm": 0.48023688793182373,
73
- "kl": 0.05418917319426934,
74
  "learning_rate": 3.867370395306068e-07,
75
  "loss": 0.0,
76
- "reward": 5.540527900060018,
77
- "reward_std": 2.529421851038933,
78
- "rewards/concensus_correctness_reward_func": 1.573499987522761,
79
- "rewards/consensus_reward_func": 0.9166666666666666,
80
  "rewards/cumulative_reward_2": 0.0,
81
- "rewards/final_correctness_reward_func": 1.75,
82
- "rewards/question_recreation_reward_func": 0.455111101269722,
83
- "rewards/soft_format_reward_func": 0.0,
84
- "rewards/strict_format_reward_func": 0.020833333333333332,
85
- "rewards/xmlcount_reward_func": 0.8244166771570841,
86
  "step": 8
87
  },
88
  {
89
- "completion_length": 232.0625,
90
- "epoch": 2.571428571428571,
91
- "grad_norm": 1.2091493606567383,
92
- "kl": 0.03521083341911435,
93
  "learning_rate": 3.1137137178519977e-07,
94
  "loss": 0.0,
95
- "reward": 6.261362463235855,
96
- "reward_std": 1.929440539330244,
97
- "rewards/concensus_correctness_reward_func": 2.2173124700784683,
98
- "rewards/consensus_reward_func": 1.375,
99
  "rewards/cumulative_reward_2": 0.0,
100
- "rewards/final_correctness_reward_func": 1.5,
101
- "rewards/question_recreation_reward_func": 0.45192492194473743,
102
- "rewards/soft_format_reward_func": 0.0,
103
- "rewards/strict_format_reward_func": 0.015625,
104
- "rewards/xmlcount_reward_func": 0.7015000097453594,
105
  "step": 10
106
  },
107
  {
108
- "completion_length": 213.33333333333334,
109
  "epoch": 3.0,
110
- "grad_norm": 1.8901488780975342,
111
- "kl": 0.028847349652399618,
112
  "learning_rate": 2.2935516363191693e-07,
113
  "loss": 0.0,
114
- "reward": 5.242655277252197,
115
- "reward_std": 2.291464234391848,
116
- "rewards/concensus_correctness_reward_func": 1.4964999854564667,
117
- "rewards/consensus_reward_func": 1.1666666666666667,
118
  "rewards/cumulative_reward_2": 0.0,
119
- "rewards/final_correctness_reward_func": 1.0,
120
- "rewards/question_recreation_reward_func": 0.54886361459891,
121
  "rewards/soft_format_reward_func": 0.0,
122
- "rewards/strict_format_reward_func": 0.0625,
123
- "rewards/xmlcount_reward_func": 0.9681249956289927,
124
  "step": 12
125
  },
126
  {
127
- "completion_length": 205.09375,
128
- "epoch": 3.571428571428571,
129
- "grad_norm": 1.1510249376296997,
130
- "kl": 0.03378852782770991,
131
  "learning_rate": 1.4957614383675767e-07,
132
  "loss": 0.0,
133
- "reward": 6.820627868175507,
134
- "reward_std": 2.49598478525877,
135
- "rewards/concensus_correctness_reward_func": 2.3831249997019768,
136
- "rewards/consensus_reward_func": 1.4375,
137
  "rewards/cumulative_reward_2": 0.0,
138
- "rewards/final_correctness_reward_func": 1.5625,
139
- "rewards/question_recreation_reward_func": 0.4795652935281396,
140
  "rewards/soft_format_reward_func": 0.0,
141
- "rewards/strict_format_reward_func": 0.03125,
142
- "rewards/xmlcount_reward_func": 0.9266875386238098,
143
  "step": 14
144
  },
145
  {
146
- "completion_length": 222.29166666666666,
147
  "epoch": 4.0,
148
- "grad_norm": 0.42290571331977844,
149
- "kl": 0.032701820290337004,
150
  "learning_rate": 8.067960709356478e-08,
151
  "loss": 0.0,
152
- "reward": 4.584882616996765,
153
- "reward_std": 1.665024643143018,
154
- "rewards/concensus_correctness_reward_func": 1.4733333190282185,
155
- "rewards/consensus_reward_func": 1.0833333333333333,
156
  "rewards/cumulative_reward_2": 0.0,
157
- "rewards/final_correctness_reward_func": 0.8333333333333334,
158
- "rewards/question_recreation_reward_func": 0.5142994125684103,
159
  "rewards/soft_format_reward_func": 0.0,
160
- "rewards/strict_format_reward_func": 0.041666666666666664,
161
- "rewards/xmlcount_reward_func": 0.6389166712760925,
162
  "step": 16
163
  },
164
  {
165
- "completion_length": 253.375,
166
- "epoch": 4.571428571428571,
167
- "grad_norm": 1.0985242128372192,
168
- "kl": 0.03926029335707426,
169
  "learning_rate": 3.013156219837776e-08,
170
  "loss": 0.0,
171
- "reward": 6.272556304931641,
172
- "reward_std": 2.2469240203499794,
173
- "rewards/concensus_correctness_reward_func": 2.0731874853372574,
174
- "rewards/consensus_reward_func": 1.3125,
175
  "rewards/cumulative_reward_2": 0.0,
176
- "rewards/final_correctness_reward_func": 1.4375,
177
- "rewards/question_recreation_reward_func": 0.5434937998652458,
178
  "rewards/soft_format_reward_func": 0.0,
179
- "rewards/strict_format_reward_func": 0.0,
180
- "rewards/xmlcount_reward_func": 0.905875027179718,
181
  "step": 18
182
  },
183
  {
184
- "completion_length": 257.7916666666667,
185
  "epoch": 5.0,
186
- "grad_norm": 0.572248637676239,
187
- "kl": 0.0269426885060966,
188
  "learning_rate": 3.4096741493194193e-09,
189
  "loss": 0.0,
190
- "reward": 6.045199592908223,
191
- "reward_std": 1.806124746799469,
192
- "rewards/concensus_correctness_reward_func": 2.25041663646698,
193
- "rewards/consensus_reward_func": 1.4166666666666667,
194
  "rewards/cumulative_reward_2": 0.0,
195
- "rewards/final_correctness_reward_func": 1.25,
196
- "rewards/question_recreation_reward_func": 0.490699624021848,
197
  "rewards/soft_format_reward_func": 0.0,
198
- "rewards/strict_format_reward_func": 0.020833333333333332,
199
- "rewards/xmlcount_reward_func": 0.616583322485288,
200
  "step": 20
201
  },
202
  {
203
  "epoch": 5.0,
204
  "step": 20,
205
  "total_flos": 0.0,
206
- "train_loss": 3.0266644716903102e-05,
207
- "train_runtime": 1124.2903,
208
- "train_samples_per_second": 0.285,
209
- "train_steps_per_second": 0.018
210
  }
211
  ],
212
  "logging_steps": 2,
 
10
  "is_world_process_zero": true,
11
  "log_history": [
12
  {
13
+ "completion_length": 198.40625,
14
+ "epoch": 0.5333333333333333,
15
+ "grad_norm": 8.360007286071777,
16
+ "kl": 0.03296396473888308,
17
  "learning_rate": 5e-07,
18
  "loss": 0.0,
19
+ "reward": 6.321916818618774,
20
+ "reward_std": 1.7188230901956558,
21
+ "rewards/concensus_correctness_reward_func": 2.3461874574422836,
22
+ "rewards/consensus_reward_func": 1.5625,
23
  "rewards/cumulative_reward_2": 0.0,
24
+ "rewards/final_correctness_reward_func": 0.875,
25
+ "rewards/question_recreation_reward_func": 0.48694803565740585,
26
+ "rewards/soft_format_reward_func": 0.0,
27
+ "rewards/strict_format_reward_func": 0.0625,
28
+ "rewards/xmlcount_reward_func": 0.9887812659144402,
29
  "step": 2
30
  },
31
  {
32
+ "completion_length": 228.32142857142858,
33
  "epoch": 1.0,
34
+ "grad_norm": 0.6417061686515808,
35
+ "kl": 0.029710494380976473,
36
  "learning_rate": 4.864543104251586e-07,
37
  "loss": 0.0,
38
+ "reward": 5.377460071018764,
39
+ "reward_std": 1.0338971018791199,
40
+ "rewards/concensus_correctness_reward_func": 1.8204285672732763,
41
+ "rewards/consensus_reward_func": 1.3571428571428572,
42
  "rewards/cumulative_reward_2": 0.0,
43
+ "rewards/final_correctness_reward_func": 0.7857142857142857,
44
+ "rewards/question_recreation_reward_func": 0.5691385694912502,
45
  "rewards/soft_format_reward_func": 0.0,
46
+ "rewards/strict_format_reward_func": 0.03571428571428571,
47
+ "rewards/xmlcount_reward_func": 0.8093214460781643,
48
  "step": 4
49
  },
50
  {
51
+ "completion_length": 184.96875,
52
+ "epoch": 1.5333333333333332,
53
+ "grad_norm": 1.40841543674469,
54
+ "kl": 0.04421051638200879,
55
  "learning_rate": 4.472851273490984e-07,
56
  "loss": 0.0,
57
+ "reward": 5.1417646408081055,
58
+ "reward_std": 2.2220220267772675,
59
+ "rewards/concensus_correctness_reward_func": 1.5205624923110008,
60
+ "rewards/consensus_reward_func": 1.25,
61
  "rewards/cumulative_reward_2": 0.0,
62
+ "rewards/final_correctness_reward_func": 0.6875,
63
+ "rewards/question_recreation_reward_func": 0.4983896389603615,
64
  "rewards/soft_format_reward_func": 0.0,
65
+ "rewards/strict_format_reward_func": 0.125,
66
+ "rewards/xmlcount_reward_func": 1.060312531888485,
67
  "step": 6
68
  },
69
  {
70
+ "completion_length": 243.64285714285714,
71
  "epoch": 2.0,
72
+ "grad_norm": 0.5562111735343933,
73
+ "kl": 0.02644094345825059,
74
  "learning_rate": 3.867370395306068e-07,
75
  "loss": 0.0,
76
+ "reward": 5.550810728754316,
77
+ "reward_std": 1.8111653668539864,
78
+ "rewards/concensus_correctness_reward_func": 1.8958571127482824,
79
+ "rewards/consensus_reward_func": 1.2142857142857142,
80
  "rewards/cumulative_reward_2": 0.0,
81
+ "rewards/final_correctness_reward_func": 1.0714285714285714,
82
+ "rewards/question_recreation_reward_func": 0.5573822898524148,
83
+ "rewards/soft_format_reward_func": 0.017857142857142856,
84
+ "rewards/strict_format_reward_func": 0.05357142857142857,
85
+ "rewards/xmlcount_reward_func": 0.7404285839625767,
86
  "step": 8
87
  },
88
  {
89
+ "completion_length": 186.96875,
90
+ "epoch": 2.533333333333333,
91
+ "grad_norm": 0.8356194496154785,
92
+ "kl": 0.03581268759444356,
93
  "learning_rate": 3.1137137178519977e-07,
94
  "loss": 0.0,
95
+ "reward": 4.26269268989563,
96
+ "reward_std": 1.9406834915280342,
97
+ "rewards/concensus_correctness_reward_func": 1.3394375033676624,
98
+ "rewards/consensus_reward_func": 1.0,
99
  "rewards/cumulative_reward_2": 0.0,
100
+ "rewards/final_correctness_reward_func": 0.4375,
101
+ "rewards/question_recreation_reward_func": 0.5125052509829402,
102
+ "rewards/soft_format_reward_func": 0.015625,
103
+ "rewards/strict_format_reward_func": 0.09375,
104
+ "rewards/xmlcount_reward_func": 0.8638750165700912,
105
  "step": 10
106
  },
107
  {
108
+ "completion_length": 207.5,
109
  "epoch": 3.0,
110
+ "grad_norm": 0.9257772564888,
111
+ "kl": 0.038460450513022285,
112
  "learning_rate": 2.2935516363191693e-07,
113
  "loss": 0.0,
114
+ "reward": 6.022549152374268,
115
+ "reward_std": 2.5043100884982517,
116
+ "rewards/concensus_correctness_reward_func": 2.333285689353943,
117
+ "rewards/consensus_reward_func": 1.2857142857142858,
118
  "rewards/cumulative_reward_2": 0.0,
119
+ "rewards/final_correctness_reward_func": 0.7857142857142857,
120
+ "rewards/question_recreation_reward_func": 0.4640133615050997,
121
  "rewards/soft_format_reward_func": 0.0,
122
+ "rewards/strict_format_reward_func": 0.14285714285714285,
123
+ "rewards/xmlcount_reward_func": 1.0109642914363317,
124
  "step": 12
125
  },
126
  {
127
+ "completion_length": 193.4375,
128
+ "epoch": 3.533333333333333,
129
+ "grad_norm": 0.8213157057762146,
130
+ "kl": 0.046339198597706854,
131
  "learning_rate": 1.4957614383675767e-07,
132
  "loss": 0.0,
133
+ "reward": 4.843554437160492,
134
+ "reward_std": 1.867899501696229,
135
+ "rewards/concensus_correctness_reward_func": 1.601437471807003,
136
+ "rewards/consensus_reward_func": 1.125,
137
  "rewards/cumulative_reward_2": 0.0,
138
+ "rewards/final_correctness_reward_func": 0.625,
139
+ "rewards/question_recreation_reward_func": 0.4974917620420456,
140
  "rewards/soft_format_reward_func": 0.0,
141
+ "rewards/strict_format_reward_func": 0.09375,
142
+ "rewards/xmlcount_reward_func": 0.9008749946951866,
143
  "step": 14
144
  },
145
  {
146
+ "completion_length": 214.35714285714286,
147
  "epoch": 4.0,
148
+ "grad_norm": 0.7088996171951294,
149
+ "kl": 0.04368409647473267,
150
  "learning_rate": 8.067960709356478e-08,
151
  "loss": 0.0,
152
+ "reward": 5.192878893443516,
153
+ "reward_std": 2.4513211250305176,
154
+ "rewards/concensus_correctness_reward_func": 1.527428584439414,
155
+ "rewards/consensus_reward_func": 1.0,
156
  "rewards/cumulative_reward_2": 0.0,
157
+ "rewards/final_correctness_reward_func": 1.1428571428571428,
158
+ "rewards/question_recreation_reward_func": 0.4679503919822829,
159
  "rewards/soft_format_reward_func": 0.0,
160
+ "rewards/strict_format_reward_func": 0.10714285714285714,
161
+ "rewards/xmlcount_reward_func": 0.9475000245230538,
162
  "step": 16
163
  },
164
  {
165
+ "completion_length": 200.65625,
166
+ "epoch": 4.533333333333333,
167
+ "grad_norm": 0.7495886087417603,
168
+ "kl": 0.027507278718985617,
169
  "learning_rate": 3.013156219837776e-08,
170
  "loss": 0.0,
171
+ "reward": 6.123600482940674,
172
+ "reward_std": 1.2783842608332634,
173
+ "rewards/concensus_correctness_reward_func": 2.145937453955412,
174
+ "rewards/consensus_reward_func": 1.5,
175
  "rewards/cumulative_reward_2": 0.0,
176
+ "rewards/final_correctness_reward_func": 0.875,
177
+ "rewards/question_recreation_reward_func": 0.5176317282021046,
178
  "rewards/soft_format_reward_func": 0.0,
179
+ "rewards/strict_format_reward_func": 0.0625,
180
+ "rewards/xmlcount_reward_func": 1.0225312784314156,
181
  "step": 18
182
  },
183
  {
184
+ "completion_length": 201.32142857142858,
185
  "epoch": 5.0,
186
+ "grad_norm": 0.6576955914497375,
187
+ "kl": 0.04337019233831337,
188
  "learning_rate": 3.4096741493194193e-09,
189
  "loss": 0.0,
190
+ "reward": 6.447563511984689,
191
+ "reward_std": 1.7632163337298803,
192
+ "rewards/concensus_correctness_reward_func": 2.3713571344103133,
193
+ "rewards/consensus_reward_func": 1.7142857142857142,
194
  "rewards/cumulative_reward_2": 0.0,
195
+ "rewards/final_correctness_reward_func": 1.0,
196
+ "rewards/question_recreation_reward_func": 0.4367780275642872,
197
  "rewards/soft_format_reward_func": 0.0,
198
+ "rewards/strict_format_reward_func": 0.08928571428571429,
199
+ "rewards/xmlcount_reward_func": 0.8358571614537921,
200
  "step": 20
201
  },
202
  {
203
  "epoch": 5.0,
204
  "step": 20,
205
  "total_flos": 0.0,
206
+ "train_loss": 3.4546391179901545e-05,
207
+ "train_runtime": 941.4145,
208
+ "train_samples_per_second": 0.34,
209
+ "train_steps_per_second": 0.021
210
  }
211
  ],
212
  "logging_steps": 2,