lole25 commited on
Commit
b5c974b
1 Parent(s): 1fb79a7

Model save

Browse files
README.md ADDED
@@ -0,0 +1,60 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: mit
3
+ library_name: peft
4
+ tags:
5
+ - trl
6
+ - dpo
7
+ - generated_from_trainer
8
+ base_model: DUAL-GPO/phi-2-gpo-new-i0
9
+ model-index:
10
+ - name: phi-2-gpo-v31-i1
11
+ results: []
12
+ ---
13
+
14
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
15
+ should probably proofread and complete it, then remove this comment. -->
16
+
17
+ # phi-2-gpo-v31-i1
18
+
19
+ This model is a fine-tuned version of [DUAL-GPO/phi-2-gpo-new-i0](https://huggingface.co/DUAL-GPO/phi-2-gpo-new-i0) on the None dataset.
20
+
21
+ ## Model description
22
+
23
+ More information needed
24
+
25
+ ## Intended uses & limitations
26
+
27
+ More information needed
28
+
29
+ ## Training and evaluation data
30
+
31
+ More information needed
32
+
33
+ ## Training procedure
34
+
35
+ ### Training hyperparameters
36
+
37
+ The following hyperparameters were used during training:
38
+ - learning_rate: 5e-06
39
+ - train_batch_size: 4
40
+ - eval_batch_size: 4
41
+ - seed: 42
42
+ - distributed_type: multi-GPU
43
+ - gradient_accumulation_steps: 4
44
+ - total_train_batch_size: 16
45
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
46
+ - lr_scheduler_type: cosine
47
+ - lr_scheduler_warmup_ratio: 0.1
48
+ - num_epochs: 1
49
+
50
+ ### Training results
51
+
52
+
53
+
54
+ ### Framework versions
55
+
56
+ - PEFT 0.7.1
57
+ - Transformers 4.36.2
58
+ - Pytorch 2.1.2+cu121
59
+ - Datasets 2.14.6
60
+ - Tokenizers 0.15.2
adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:5ec34f1b7eff709dc884e60db070c2085b8835ff469881da79ddd078d496ad07
3
  size 167807296
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:03a04dc2745b9facf85002655b8bf7291845e4f9dae765ca9c7b7c176efbf357
3
  size 167807296
all_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "train_loss": 0.2873324268341064,
4
+ "train_runtime": 11591.8245,
5
+ "train_samples": 20000,
6
+ "train_samples_per_second": 1.725,
7
+ "train_steps_per_second": 0.108
8
+ }
runs/May16_03-01-02_gpu4-119-5/events.out.tfevents.1715792632.gpu4-119-5.3664742.0 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:8befa03be2ef1c3eab860545cb246a98794c4cb773af88d464103b183734effd
3
- size 81406
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dca0ea5b13d137922b5c16baf4e81987055538b89994552f834c5ed51b1a5ba6
3
+ size 84930
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "train_loss": 0.2873324268341064,
4
+ "train_runtime": 11591.8245,
5
+ "train_samples": 20000,
6
+ "train_samples_per_second": 1.725,
7
+ "train_steps_per_second": 0.108
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,1794 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 1.0,
5
+ "eval_steps": 500,
6
+ "global_step": 1250,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.0,
13
+ "learning_rate": 4e-08,
14
+ "logits/chosen": -0.00011328980326652527,
15
+ "logits/rejected": 0.11862002313137054,
16
+ "logps/chosen": -264.0264892578125,
17
+ "logps/rejected": -239.23065185546875,
18
+ "loss": 0.3534,
19
+ "rewards/accuracies": 0.0,
20
+ "rewards/chosen": 0.0,
21
+ "rewards/margins": 0.0,
22
+ "rewards/rejected": 0.0,
23
+ "step": 1
24
+ },
25
+ {
26
+ "epoch": 0.01,
27
+ "learning_rate": 4.0000000000000003e-07,
28
+ "logits/chosen": 0.12777267396450043,
29
+ "logits/rejected": 0.19940675795078278,
30
+ "logps/chosen": -355.0330810546875,
31
+ "logps/rejected": -306.326904296875,
32
+ "loss": 0.3805,
33
+ "rewards/accuracies": 0.375,
34
+ "rewards/chosen": 0.00010627516894601285,
35
+ "rewards/margins": 7.693069346714765e-05,
36
+ "rewards/rejected": 2.9344506401685067e-05,
37
+ "step": 10
38
+ },
39
+ {
40
+ "epoch": 0.02,
41
+ "learning_rate": 8.000000000000001e-07,
42
+ "logits/chosen": 0.11426838487386703,
43
+ "logits/rejected": 0.19905796647071838,
44
+ "logps/chosen": -378.491943359375,
45
+ "logps/rejected": -338.3055725097656,
46
+ "loss": 0.3729,
47
+ "rewards/accuracies": 0.4312500059604645,
48
+ "rewards/chosen": -4.1492825403111055e-05,
49
+ "rewards/margins": -9.77895688265562e-05,
50
+ "rewards/rejected": 5.629673978546634e-05,
51
+ "step": 20
52
+ },
53
+ {
54
+ "epoch": 0.02,
55
+ "learning_rate": 1.2000000000000002e-06,
56
+ "logits/chosen": 0.18764689564704895,
57
+ "logits/rejected": 0.17880608141422272,
58
+ "logps/chosen": -301.9078674316406,
59
+ "logps/rejected": -327.083740234375,
60
+ "loss": 0.3733,
61
+ "rewards/accuracies": 0.5,
62
+ "rewards/chosen": -5.114002124173567e-05,
63
+ "rewards/margins": 5.109022094984539e-05,
64
+ "rewards/rejected": -0.00010223023127764463,
65
+ "step": 30
66
+ },
67
+ {
68
+ "epoch": 0.03,
69
+ "learning_rate": 1.6000000000000001e-06,
70
+ "logits/chosen": 0.1114225760102272,
71
+ "logits/rejected": 0.15094885230064392,
72
+ "logps/chosen": -296.6694641113281,
73
+ "logps/rejected": -292.0389099121094,
74
+ "loss": 0.3591,
75
+ "rewards/accuracies": 0.4749999940395355,
76
+ "rewards/chosen": 0.0001755535922711715,
77
+ "rewards/margins": 0.000263996102148667,
78
+ "rewards/rejected": -8.844250987749547e-05,
79
+ "step": 40
80
+ },
81
+ {
82
+ "epoch": 0.04,
83
+ "learning_rate": 2.0000000000000003e-06,
84
+ "logits/chosen": 0.1339588612318039,
85
+ "logits/rejected": 0.22528474032878876,
86
+ "logps/chosen": -371.5467224121094,
87
+ "logps/rejected": -314.2153625488281,
88
+ "loss": 0.3859,
89
+ "rewards/accuracies": 0.44999998807907104,
90
+ "rewards/chosen": -4.210665792925283e-05,
91
+ "rewards/margins": 9.823093569139019e-05,
92
+ "rewards/rejected": -0.00014033759362064302,
93
+ "step": 50
94
+ },
95
+ {
96
+ "epoch": 0.05,
97
+ "learning_rate": 2.4000000000000003e-06,
98
+ "logits/chosen": 0.10837416350841522,
99
+ "logits/rejected": 0.22042012214660645,
100
+ "logps/chosen": -360.5331115722656,
101
+ "logps/rejected": -354.9367980957031,
102
+ "loss": 0.3783,
103
+ "rewards/accuracies": 0.606249988079071,
104
+ "rewards/chosen": 6.062311149435118e-05,
105
+ "rewards/margins": 0.0006647260743193328,
106
+ "rewards/rejected": -0.0006041030283086002,
107
+ "step": 60
108
+ },
109
+ {
110
+ "epoch": 0.06,
111
+ "learning_rate": 2.8000000000000003e-06,
112
+ "logits/chosen": 0.14214661717414856,
113
+ "logits/rejected": 0.22639691829681396,
114
+ "logps/chosen": -318.57574462890625,
115
+ "logps/rejected": -312.8997802734375,
116
+ "loss": 0.3718,
117
+ "rewards/accuracies": 0.5874999761581421,
118
+ "rewards/chosen": 0.00033316004555672407,
119
+ "rewards/margins": 0.000630541646387428,
120
+ "rewards/rejected": -0.0002973815717268735,
121
+ "step": 70
122
+ },
123
+ {
124
+ "epoch": 0.06,
125
+ "learning_rate": 3.2000000000000003e-06,
126
+ "logits/chosen": 0.1115114837884903,
127
+ "logits/rejected": 0.18560612201690674,
128
+ "logps/chosen": -322.22430419921875,
129
+ "logps/rejected": -316.9699401855469,
130
+ "loss": 0.3641,
131
+ "rewards/accuracies": 0.668749988079071,
132
+ "rewards/chosen": 0.0006448437925428152,
133
+ "rewards/margins": 0.00127916329074651,
134
+ "rewards/rejected": -0.0006343194982036948,
135
+ "step": 80
136
+ },
137
+ {
138
+ "epoch": 0.07,
139
+ "learning_rate": 3.6000000000000003e-06,
140
+ "logits/chosen": 0.13246920704841614,
141
+ "logits/rejected": 0.21337231993675232,
142
+ "logps/chosen": -371.502685546875,
143
+ "logps/rejected": -330.3704528808594,
144
+ "loss": 0.3617,
145
+ "rewards/accuracies": 0.65625,
146
+ "rewards/chosen": 0.0013298379490152001,
147
+ "rewards/margins": 0.0018641676288098097,
148
+ "rewards/rejected": -0.0005343296797946095,
149
+ "step": 90
150
+ },
151
+ {
152
+ "epoch": 0.08,
153
+ "learning_rate": 4.000000000000001e-06,
154
+ "logits/chosen": 0.11018741130828857,
155
+ "logits/rejected": 0.17559225857257843,
156
+ "logps/chosen": -362.69635009765625,
157
+ "logps/rejected": -356.4203796386719,
158
+ "loss": 0.3753,
159
+ "rewards/accuracies": 0.668749988079071,
160
+ "rewards/chosen": 0.001733560347929597,
161
+ "rewards/margins": 0.003086336422711611,
162
+ "rewards/rejected": -0.001352776074782014,
163
+ "step": 100
164
+ },
165
+ {
166
+ "epoch": 0.09,
167
+ "learning_rate": 4.4e-06,
168
+ "logits/chosen": 0.09963714331388474,
169
+ "logits/rejected": 0.2167224884033203,
170
+ "logps/chosen": -389.74041748046875,
171
+ "logps/rejected": -345.9913024902344,
172
+ "loss": 0.3635,
173
+ "rewards/accuracies": 0.6875,
174
+ "rewards/chosen": 0.0017973188078030944,
175
+ "rewards/margins": 0.005557372234761715,
176
+ "rewards/rejected": -0.0037600535433739424,
177
+ "step": 110
178
+ },
179
+ {
180
+ "epoch": 0.1,
181
+ "learning_rate": 4.800000000000001e-06,
182
+ "logits/chosen": 0.1074131578207016,
183
+ "logits/rejected": 0.12421264499425888,
184
+ "logps/chosen": -325.73028564453125,
185
+ "logps/rejected": -332.47991943359375,
186
+ "loss": 0.3537,
187
+ "rewards/accuracies": 0.6312500238418579,
188
+ "rewards/chosen": 0.0013967979466542602,
189
+ "rewards/margins": 0.008982640691101551,
190
+ "rewards/rejected": -0.007585843093693256,
191
+ "step": 120
192
+ },
193
+ {
194
+ "epoch": 0.1,
195
+ "learning_rate": 4.999756310023261e-06,
196
+ "logits/chosen": 0.09501053392887115,
197
+ "logits/rejected": 0.15297293663024902,
198
+ "logps/chosen": -330.9049377441406,
199
+ "logps/rejected": -349.04486083984375,
200
+ "loss": 0.3681,
201
+ "rewards/accuracies": 0.6937500238418579,
202
+ "rewards/chosen": -0.00485868938267231,
203
+ "rewards/margins": 0.01828749105334282,
204
+ "rewards/rejected": -0.02314617857336998,
205
+ "step": 130
206
+ },
207
+ {
208
+ "epoch": 0.11,
209
+ "learning_rate": 4.997807075247147e-06,
210
+ "logits/chosen": 0.10013115406036377,
211
+ "logits/rejected": 0.16476190090179443,
212
+ "logps/chosen": -373.80352783203125,
213
+ "logps/rejected": -409.95745849609375,
214
+ "loss": 0.3555,
215
+ "rewards/accuracies": 0.6937500238418579,
216
+ "rewards/chosen": -0.030246371403336525,
217
+ "rewards/margins": 0.03136765584349632,
218
+ "rewards/rejected": -0.0616140253841877,
219
+ "step": 140
220
+ },
221
+ {
222
+ "epoch": 0.12,
223
+ "learning_rate": 4.993910125649561e-06,
224
+ "logits/chosen": 0.10287214815616608,
225
+ "logits/rejected": 0.15870796144008636,
226
+ "logps/chosen": -432.9341735839844,
227
+ "logps/rejected": -434.81744384765625,
228
+ "loss": 0.3461,
229
+ "rewards/accuracies": 0.581250011920929,
230
+ "rewards/chosen": -0.08363470435142517,
231
+ "rewards/margins": 0.02743600867688656,
232
+ "rewards/rejected": -0.11107071489095688,
233
+ "step": 150
234
+ },
235
+ {
236
+ "epoch": 0.13,
237
+ "learning_rate": 4.988068499954578e-06,
238
+ "logits/chosen": 0.12284994125366211,
239
+ "logits/rejected": 0.19283421337604523,
240
+ "logps/chosen": -425.1427307128906,
241
+ "logps/rejected": -479.8091735839844,
242
+ "loss": 0.3394,
243
+ "rewards/accuracies": 0.668749988079071,
244
+ "rewards/chosen": -0.09083442389965057,
245
+ "rewards/margins": 0.05471948906779289,
246
+ "rewards/rejected": -0.14555391669273376,
247
+ "step": 160
248
+ },
249
+ {
250
+ "epoch": 0.14,
251
+ "learning_rate": 4.980286753286196e-06,
252
+ "logits/chosen": 0.17768949270248413,
253
+ "logits/rejected": 0.2632974684238434,
254
+ "logps/chosen": -457.4527282714844,
255
+ "logps/rejected": -496.40234375,
256
+ "loss": 0.3386,
257
+ "rewards/accuracies": 0.625,
258
+ "rewards/chosen": -0.12117190659046173,
259
+ "rewards/margins": 0.05782421678304672,
260
+ "rewards/rejected": -0.17899611592292786,
261
+ "step": 170
262
+ },
263
+ {
264
+ "epoch": 0.14,
265
+ "learning_rate": 4.970570953616383e-06,
266
+ "logits/chosen": 0.14909473061561584,
267
+ "logits/rejected": 0.21552911400794983,
268
+ "logps/chosen": -444.69708251953125,
269
+ "logps/rejected": -533.7720947265625,
270
+ "loss": 0.3142,
271
+ "rewards/accuracies": 0.625,
272
+ "rewards/chosen": -0.12033964693546295,
273
+ "rewards/margins": 0.08396399021148682,
274
+ "rewards/rejected": -0.20430362224578857,
275
+ "step": 180
276
+ },
277
+ {
278
+ "epoch": 0.15,
279
+ "learning_rate": 4.958928677033465e-06,
280
+ "logits/chosen": 0.18431484699249268,
281
+ "logits/rejected": 0.21507856249809265,
282
+ "logps/chosen": -475.87908935546875,
283
+ "logps/rejected": -555.5066528320312,
284
+ "loss": 0.2978,
285
+ "rewards/accuracies": 0.6875,
286
+ "rewards/chosen": -0.15021201968193054,
287
+ "rewards/margins": 0.08976520597934723,
288
+ "rewards/rejected": -0.23997724056243896,
289
+ "step": 190
290
+ },
291
+ {
292
+ "epoch": 0.16,
293
+ "learning_rate": 4.9453690018345144e-06,
294
+ "logits/chosen": 0.17984744906425476,
295
+ "logits/rejected": 0.26865434646606445,
296
+ "logps/chosen": -501.103515625,
297
+ "logps/rejected": -583.3973999023438,
298
+ "loss": 0.274,
299
+ "rewards/accuracies": 0.6625000238418579,
300
+ "rewards/chosen": -0.17393726110458374,
301
+ "rewards/margins": 0.10630029439926147,
302
+ "rewards/rejected": -0.2802375257015228,
303
+ "step": 200
304
+ },
305
+ {
306
+ "epoch": 0.17,
307
+ "learning_rate": 4.9299025014463665e-06,
308
+ "logits/chosen": 0.23147694766521454,
309
+ "logits/rejected": 0.2544376850128174,
310
+ "logps/chosen": -603.2955932617188,
311
+ "logps/rejected": -713.1234130859375,
312
+ "loss": 0.3164,
313
+ "rewards/accuracies": 0.6625000238418579,
314
+ "rewards/chosen": -0.24584278464317322,
315
+ "rewards/margins": 0.12140010297298431,
316
+ "rewards/rejected": -0.36724287271499634,
317
+ "step": 210
318
+ },
319
+ {
320
+ "epoch": 0.18,
321
+ "learning_rate": 4.912541236180779e-06,
322
+ "logits/chosen": 0.2394331395626068,
323
+ "logits/rejected": 0.34389573335647583,
324
+ "logps/chosen": -596.3639526367188,
325
+ "logps/rejected": -683.6994018554688,
326
+ "loss": 0.2964,
327
+ "rewards/accuracies": 0.706250011920929,
328
+ "rewards/chosen": -0.24022677540779114,
329
+ "rewards/margins": 0.12776102125644684,
330
+ "rewards/rejected": -0.3679877817630768,
331
+ "step": 220
332
+ },
333
+ {
334
+ "epoch": 0.18,
335
+ "learning_rate": 4.893298743830168e-06,
336
+ "logits/chosen": 0.23847298324108124,
337
+ "logits/rejected": 0.2884899973869324,
338
+ "logps/chosen": -577.9203491210938,
339
+ "logps/rejected": -740.6541748046875,
340
+ "loss": 0.2913,
341
+ "rewards/accuracies": 0.6499999761581421,
342
+ "rewards/chosen": -0.25649553537368774,
343
+ "rewards/margins": 0.15363454818725586,
344
+ "rewards/rejected": -0.410130113363266,
345
+ "step": 230
346
+ },
347
+ {
348
+ "epoch": 0.19,
349
+ "learning_rate": 4.8721900291112415e-06,
350
+ "logits/chosen": 0.15130439400672913,
351
+ "logits/rejected": 0.2567364275455475,
352
+ "logps/chosen": -567.4896240234375,
353
+ "logps/rejected": -739.4119873046875,
354
+ "loss": 0.2921,
355
+ "rewards/accuracies": 0.6875,
356
+ "rewards/chosen": -0.24494703114032745,
357
+ "rewards/margins": 0.161117285490036,
358
+ "rewards/rejected": -0.40606433153152466,
359
+ "step": 240
360
+ },
361
+ {
362
+ "epoch": 0.2,
363
+ "learning_rate": 4.849231551964771e-06,
364
+ "logits/chosen": 0.12885180115699768,
365
+ "logits/rejected": 0.2513241767883301,
366
+ "logps/chosen": -499.4623107910156,
367
+ "logps/rejected": -598.4744873046875,
368
+ "loss": 0.2988,
369
+ "rewards/accuracies": 0.706250011920929,
370
+ "rewards/chosen": -0.18550391495227814,
371
+ "rewards/margins": 0.11325653642416,
372
+ "rewards/rejected": -0.29876044392585754,
373
+ "step": 250
374
+ },
375
+ {
376
+ "epoch": 0.21,
377
+ "learning_rate": 4.824441214720629e-06,
378
+ "logits/chosen": 0.11510957777500153,
379
+ "logits/rejected": 0.16252097487449646,
380
+ "logps/chosen": -528.8275146484375,
381
+ "logps/rejected": -612.6734619140625,
382
+ "loss": 0.301,
383
+ "rewards/accuracies": 0.6625000238418579,
384
+ "rewards/chosen": -0.20904381573200226,
385
+ "rewards/margins": 0.11143267154693604,
386
+ "rewards/rejected": -0.3204765021800995,
387
+ "step": 260
388
+ },
389
+ {
390
+ "epoch": 0.22,
391
+ "learning_rate": 4.7978383481380865e-06,
392
+ "logits/chosen": 0.10701459646224976,
393
+ "logits/rejected": 0.19334085285663605,
394
+ "logps/chosen": -624.1452026367188,
395
+ "logps/rejected": -728.2015380859375,
396
+ "loss": 0.283,
397
+ "rewards/accuracies": 0.6875,
398
+ "rewards/chosen": -0.26597902178764343,
399
+ "rewards/margins": 0.1281953901052475,
400
+ "rewards/rejected": -0.3941744267940521,
401
+ "step": 270
402
+ },
403
+ {
404
+ "epoch": 0.22,
405
+ "learning_rate": 4.769443696332272e-06,
406
+ "logits/chosen": 0.10034264624118805,
407
+ "logits/rejected": 0.14515987038612366,
408
+ "logps/chosen": -636.72509765625,
409
+ "logps/rejected": -722.1334228515625,
410
+ "loss": 0.3116,
411
+ "rewards/accuracies": 0.612500011920929,
412
+ "rewards/chosen": -0.26666730642318726,
413
+ "rewards/margins": 0.10862217098474503,
414
+ "rewards/rejected": -0.3752894401550293,
415
+ "step": 280
416
+ },
417
+ {
418
+ "epoch": 0.23,
419
+ "learning_rate": 4.7392794005985324e-06,
420
+ "logits/chosen": 0.05613882467150688,
421
+ "logits/rejected": 0.1429525762796402,
422
+ "logps/chosen": -570.8783569335938,
423
+ "logps/rejected": -689.3982543945312,
424
+ "loss": 0.2858,
425
+ "rewards/accuracies": 0.6875,
426
+ "rewards/chosen": -0.24179348349571228,
427
+ "rewards/margins": 0.12780442833900452,
428
+ "rewards/rejected": -0.3695979416370392,
429
+ "step": 290
430
+ },
431
+ {
432
+ "epoch": 0.24,
433
+ "learning_rate": 4.707368982147318e-06,
434
+ "logits/chosen": 0.04738849401473999,
435
+ "logits/rejected": 0.11430943012237549,
436
+ "logps/chosen": -613.541259765625,
437
+ "logps/rejected": -686.9055786132812,
438
+ "loss": 0.3203,
439
+ "rewards/accuracies": 0.637499988079071,
440
+ "rewards/chosen": -0.25295981764793396,
441
+ "rewards/margins": 0.10196548700332642,
442
+ "rewards/rejected": -0.35492533445358276,
443
+ "step": 300
444
+ },
445
+ {
446
+ "epoch": 0.25,
447
+ "learning_rate": 4.673737323763048e-06,
448
+ "logits/chosen": 0.055924586951732635,
449
+ "logits/rejected": 0.1546911895275116,
450
+ "logps/chosen": -591.7320556640625,
451
+ "logps/rejected": -695.1236572265625,
452
+ "loss": 0.294,
453
+ "rewards/accuracies": 0.7437499761581421,
454
+ "rewards/chosen": -0.22645549476146698,
455
+ "rewards/margins": 0.12102316319942474,
456
+ "rewards/rejected": -0.3474786877632141,
457
+ "step": 310
458
+ },
459
+ {
460
+ "epoch": 0.26,
461
+ "learning_rate": 4.638410650401267e-06,
462
+ "logits/chosen": 0.08233563601970673,
463
+ "logits/rejected": 0.13290105760097504,
464
+ "logps/chosen": -571.2188110351562,
465
+ "logps/rejected": -663.9716796875,
466
+ "loss": 0.2906,
467
+ "rewards/accuracies": 0.6625000238418579,
468
+ "rewards/chosen": -0.22465641796588898,
469
+ "rewards/margins": 0.11367009580135345,
470
+ "rewards/rejected": -0.3383265435695648,
471
+ "step": 320
472
+ },
473
+ {
474
+ "epoch": 0.26,
475
+ "learning_rate": 4.601416508739211e-06,
476
+ "logits/chosen": 0.07142064720392227,
477
+ "logits/rejected": 0.11630046367645264,
478
+ "logps/chosen": -576.2288208007812,
479
+ "logps/rejected": -685.9638061523438,
480
+ "loss": 0.2786,
481
+ "rewards/accuracies": 0.6875,
482
+ "rewards/chosen": -0.2488282173871994,
483
+ "rewards/margins": 0.13437636196613312,
484
+ "rewards/rejected": -0.3832045793533325,
485
+ "step": 330
486
+ },
487
+ {
488
+ "epoch": 0.27,
489
+ "learning_rate": 4.562783745695738e-06,
490
+ "logits/chosen": 0.06200198084115982,
491
+ "logits/rejected": 0.1760326325893402,
492
+ "logps/chosen": -616.6163330078125,
493
+ "logps/rejected": -710.7817993164062,
494
+ "loss": 0.3136,
495
+ "rewards/accuracies": 0.668749988079071,
496
+ "rewards/chosen": -0.2617984414100647,
497
+ "rewards/margins": 0.11515985429286957,
498
+ "rewards/rejected": -0.37695831060409546,
499
+ "step": 340
500
+ },
501
+ {
502
+ "epoch": 0.28,
503
+ "learning_rate": 4.522542485937369e-06,
504
+ "logits/chosen": 0.050816863775253296,
505
+ "logits/rejected": 0.1229025274515152,
506
+ "logps/chosen": -629.1052856445312,
507
+ "logps/rejected": -734.9453125,
508
+ "loss": 0.2544,
509
+ "rewards/accuracies": 0.7124999761581421,
510
+ "rewards/chosen": -0.2729812264442444,
511
+ "rewards/margins": 0.1355685442686081,
512
+ "rewards/rejected": -0.40854978561401367,
513
+ "step": 350
514
+ },
515
+ {
516
+ "epoch": 0.29,
517
+ "learning_rate": 4.4807241083879774e-06,
518
+ "logits/chosen": 0.11080716550350189,
519
+ "logits/rejected": 0.19923880696296692,
520
+ "logps/chosen": -633.8490600585938,
521
+ "logps/rejected": -751.9637451171875,
522
+ "loss": 0.2693,
523
+ "rewards/accuracies": 0.6625000238418579,
524
+ "rewards/chosen": -0.28209587931632996,
525
+ "rewards/margins": 0.14658969640731812,
526
+ "rewards/rejected": -0.42868557572364807,
527
+ "step": 360
528
+ },
529
+ {
530
+ "epoch": 0.3,
531
+ "learning_rate": 4.437361221760449e-06,
532
+ "logits/chosen": 0.09759480506181717,
533
+ "logits/rejected": 0.15628591179847717,
534
+ "logps/chosen": -621.000732421875,
535
+ "logps/rejected": -775.7725219726562,
536
+ "loss": 0.2512,
537
+ "rewards/accuracies": 0.699999988079071,
538
+ "rewards/chosen": -0.28890570998191833,
539
+ "rewards/margins": 0.16257989406585693,
540
+ "rewards/rejected": -0.4514855444431305,
541
+ "step": 370
542
+ },
543
+ {
544
+ "epoch": 0.3,
545
+ "learning_rate": 4.3924876391293915e-06,
546
+ "logits/chosen": 0.11163793504238129,
547
+ "logits/rejected": 0.1869780719280243,
548
+ "logps/chosen": -637.1199340820312,
549
+ "logps/rejected": -713.552978515625,
550
+ "loss": 0.2996,
551
+ "rewards/accuracies": 0.65625,
552
+ "rewards/chosen": -0.27824416756629944,
553
+ "rewards/margins": 0.1207936555147171,
554
+ "rewards/rejected": -0.39903780817985535,
555
+ "step": 380
556
+ },
557
+ {
558
+ "epoch": 0.31,
559
+ "learning_rate": 4.346138351564711e-06,
560
+ "logits/chosen": 0.09762606769800186,
561
+ "logits/rejected": 0.12994861602783203,
562
+ "logps/chosen": -524.1798095703125,
563
+ "logps/rejected": -645.6965942382812,
564
+ "loss": 0.2902,
565
+ "rewards/accuracies": 0.65625,
566
+ "rewards/chosen": -0.19974325597286224,
567
+ "rewards/margins": 0.13353568315505981,
568
+ "rewards/rejected": -0.33327895402908325,
569
+ "step": 390
570
+ },
571
+ {
572
+ "epoch": 0.32,
573
+ "learning_rate": 4.2983495008466285e-06,
574
+ "logits/chosen": 0.06368868052959442,
575
+ "logits/rejected": 0.14533713459968567,
576
+ "logps/chosen": -527.1744384765625,
577
+ "logps/rejected": -576.6195068359375,
578
+ "loss": 0.3164,
579
+ "rewards/accuracies": 0.6625000238418579,
580
+ "rewards/chosen": -0.1860881745815277,
581
+ "rewards/margins": 0.07743660360574722,
582
+ "rewards/rejected": -0.2635248005390167,
583
+ "step": 400
584
+ },
585
+ {
586
+ "epoch": 0.33,
587
+ "learning_rate": 4.249158351283414e-06,
588
+ "logits/chosen": 0.09450914710760117,
589
+ "logits/rejected": 0.12860049307346344,
590
+ "logps/chosen": -548.9276123046875,
591
+ "logps/rejected": -644.9456787109375,
592
+ "loss": 0.2733,
593
+ "rewards/accuracies": 0.737500011920929,
594
+ "rewards/chosen": -0.198289155960083,
595
+ "rewards/margins": 0.12718045711517334,
596
+ "rewards/rejected": -0.32546961307525635,
597
+ "step": 410
598
+ },
599
+ {
600
+ "epoch": 0.34,
601
+ "learning_rate": 4.198603260653792e-06,
602
+ "logits/chosen": 0.017919300124049187,
603
+ "logits/rejected": 0.1374201476573944,
604
+ "logps/chosen": -593.1436157226562,
605
+ "logps/rejected": -704.8626708984375,
606
+ "loss": 0.2841,
607
+ "rewards/accuracies": 0.699999988079071,
608
+ "rewards/chosen": -0.25730082392692566,
609
+ "rewards/margins": 0.13911226391792297,
610
+ "rewards/rejected": -0.39641305804252625,
611
+ "step": 420
612
+ },
613
+ {
614
+ "epoch": 0.34,
615
+ "learning_rate": 4.146723650296701e-06,
616
+ "logits/chosen": 0.08983961492776871,
617
+ "logits/rejected": 0.16111747920513153,
618
+ "logps/chosen": -659.6158447265625,
619
+ "logps/rejected": -744.7501220703125,
620
+ "loss": 0.3244,
621
+ "rewards/accuracies": 0.6499999761581421,
622
+ "rewards/chosen": -0.3095107674598694,
623
+ "rewards/margins": 0.11083928495645523,
624
+ "rewards/rejected": -0.4203500747680664,
625
+ "step": 430
626
+ },
627
+ {
628
+ "epoch": 0.35,
629
+ "learning_rate": 4.093559974371725e-06,
630
+ "logits/chosen": 0.09062441438436508,
631
+ "logits/rejected": 0.11453583091497421,
632
+ "logps/chosen": -580.1378784179688,
633
+ "logps/rejected": -689.7623901367188,
634
+ "loss": 0.3067,
635
+ "rewards/accuracies": 0.65625,
636
+ "rewards/chosen": -0.2507023215293884,
637
+ "rewards/margins": 0.12082631886005402,
638
+ "rewards/rejected": -0.37152862548828125,
639
+ "step": 440
640
+ },
641
+ {
642
+ "epoch": 0.36,
643
+ "learning_rate": 4.039153688314146e-06,
644
+ "logits/chosen": 0.08577083796262741,
645
+ "logits/rejected": 0.09571922570466995,
646
+ "logps/chosen": -607.8924560546875,
647
+ "logps/rejected": -690.2274169921875,
648
+ "loss": 0.3173,
649
+ "rewards/accuracies": 0.65625,
650
+ "rewards/chosen": -0.28040188550949097,
651
+ "rewards/margins": 0.09828029572963715,
652
+ "rewards/rejected": -0.37868213653564453,
653
+ "step": 450
654
+ },
655
+ {
656
+ "epoch": 0.37,
657
+ "learning_rate": 3.983547216509254e-06,
658
+ "logits/chosen": 0.09381420165300369,
659
+ "logits/rejected": 0.1439538300037384,
660
+ "logps/chosen": -564.9346923828125,
661
+ "logps/rejected": -690.1428833007812,
662
+ "loss": 0.2688,
663
+ "rewards/accuracies": 0.6625000238418579,
664
+ "rewards/chosen": -0.24971938133239746,
665
+ "rewards/margins": 0.1450313776731491,
666
+ "rewards/rejected": -0.3947507441043854,
667
+ "step": 460
668
+ },
669
+ {
670
+ "epoch": 0.38,
671
+ "learning_rate": 3.92678391921108e-06,
672
+ "logits/chosen": 0.13807064294815063,
673
+ "logits/rejected": 0.18901431560516357,
674
+ "logps/chosen": -603.8775024414062,
675
+ "logps/rejected": -746.001708984375,
676
+ "loss": 0.2733,
677
+ "rewards/accuracies": 0.625,
678
+ "rewards/chosen": -0.2938609719276428,
679
+ "rewards/margins": 0.1443910002708435,
680
+ "rewards/rejected": -0.43825197219848633,
681
+ "step": 470
682
+ },
683
+ {
684
+ "epoch": 0.38,
685
+ "learning_rate": 3.868908058731376e-06,
686
+ "logits/chosen": 0.11844799667596817,
687
+ "logits/rejected": 0.17255009710788727,
688
+ "logps/chosen": -586.2673950195312,
689
+ "logps/rejected": -730.5003662109375,
690
+ "loss": 0.2908,
691
+ "rewards/accuracies": 0.668749988079071,
692
+ "rewards/chosen": -0.281721293926239,
693
+ "rewards/margins": 0.1272607296705246,
694
+ "rewards/rejected": -0.4089820384979248,
695
+ "step": 480
696
+ },
697
+ {
698
+ "epoch": 0.39,
699
+ "learning_rate": 3.8099647649251984e-06,
700
+ "logits/chosen": 0.09214349836111069,
701
+ "logits/rejected": 0.149861678481102,
702
+ "logps/chosen": -608.2205810546875,
703
+ "logps/rejected": -698.6951293945312,
704
+ "loss": 0.2973,
705
+ "rewards/accuracies": 0.6875,
706
+ "rewards/chosen": -0.24310803413391113,
707
+ "rewards/margins": 0.12697242200374603,
708
+ "rewards/rejected": -0.37008044123649597,
709
+ "step": 490
710
+ },
711
+ {
712
+ "epoch": 0.4,
713
+ "learning_rate": 3.7500000000000005e-06,
714
+ "logits/chosen": 0.09263923764228821,
715
+ "logits/rejected": 0.17302937805652618,
716
+ "logps/chosen": -586.6901245117188,
717
+ "logps/rejected": -707.1153564453125,
718
+ "loss": 0.2892,
719
+ "rewards/accuracies": 0.7250000238418579,
720
+ "rewards/chosen": -0.2508801817893982,
721
+ "rewards/margins": 0.13355056941509247,
722
+ "rewards/rejected": -0.38443079590797424,
723
+ "step": 500
724
+ },
725
+ {
726
+ "epoch": 0.41,
727
+ "learning_rate": 3.689060522675689e-06,
728
+ "logits/chosen": 0.10364197194576263,
729
+ "logits/rejected": 0.13604341447353363,
730
+ "logps/chosen": -546.9564819335938,
731
+ "logps/rejected": -676.0130615234375,
732
+ "loss": 0.2864,
733
+ "rewards/accuracies": 0.668749988079071,
734
+ "rewards/chosen": -0.24754706025123596,
735
+ "rewards/margins": 0.13658317923545837,
736
+ "rewards/rejected": -0.38413023948669434,
737
+ "step": 510
738
+ },
739
+ {
740
+ "epoch": 0.42,
741
+ "learning_rate": 3.627193851723577e-06,
742
+ "logits/chosen": 0.10513228178024292,
743
+ "logits/rejected": 0.16634628176689148,
744
+ "logps/chosen": -704.7044067382812,
745
+ "logps/rejected": -804.184326171875,
746
+ "loss": 0.2799,
747
+ "rewards/accuracies": 0.6875,
748
+ "rewards/chosen": -0.3062405586242676,
749
+ "rewards/margins": 0.14433066546916962,
750
+ "rewards/rejected": -0.4505712389945984,
751
+ "step": 520
752
+ },
753
+ {
754
+ "epoch": 0.42,
755
+ "learning_rate": 3.564448228912682e-06,
756
+ "logits/chosen": 0.12000527232885361,
757
+ "logits/rejected": 0.16995035111904144,
758
+ "logps/chosen": -625.014892578125,
759
+ "logps/rejected": -774.2705688476562,
760
+ "loss": 0.249,
761
+ "rewards/accuracies": 0.731249988079071,
762
+ "rewards/chosen": -0.2779640853404999,
763
+ "rewards/margins": 0.17402225732803345,
764
+ "rewards/rejected": -0.4519864022731781,
765
+ "step": 530
766
+ },
767
+ {
768
+ "epoch": 0.43,
769
+ "learning_rate": 3.5008725813922383e-06,
770
+ "logits/chosen": 0.13057418167591095,
771
+ "logits/rejected": 0.21256506443023682,
772
+ "logps/chosen": -691.3427734375,
773
+ "logps/rejected": -856.4500732421875,
774
+ "loss": 0.2642,
775
+ "rewards/accuracies": 0.7124999761581421,
776
+ "rewards/chosen": -0.3297198712825775,
777
+ "rewards/margins": 0.1787293404340744,
778
+ "rewards/rejected": -0.5084491968154907,
779
+ "step": 540
780
+ },
781
+ {
782
+ "epoch": 0.44,
783
+ "learning_rate": 3.436516483539781e-06,
784
+ "logits/chosen": 0.16701681911945343,
785
+ "logits/rejected": 0.16362786293029785,
786
+ "logps/chosen": -669.8779296875,
787
+ "logps/rejected": -775.9328002929688,
788
+ "loss": 0.2979,
789
+ "rewards/accuracies": 0.6812499761581421,
790
+ "rewards/chosen": -0.31730252504348755,
791
+ "rewards/margins": 0.12564435601234436,
792
+ "rewards/rejected": -0.4429469108581543,
793
+ "step": 550
794
+ },
795
+ {
796
+ "epoch": 0.45,
797
+ "learning_rate": 3.3714301183045382e-06,
798
+ "logits/chosen": 0.10453770309686661,
799
+ "logits/rejected": 0.16834238171577454,
800
+ "logps/chosen": -637.10986328125,
801
+ "logps/rejected": -716.6213989257812,
802
+ "loss": 0.286,
803
+ "rewards/accuracies": 0.6875,
804
+ "rewards/chosen": -0.2491428554058075,
805
+ "rewards/margins": 0.12585783004760742,
806
+ "rewards/rejected": -0.3750006854534149,
807
+ "step": 560
808
+ },
809
+ {
810
+ "epoch": 0.46,
811
+ "learning_rate": 3.3056642380762783e-06,
812
+ "logits/chosen": 0.12909114360809326,
813
+ "logits/rejected": 0.17991405725479126,
814
+ "logps/chosen": -595.2734375,
815
+ "logps/rejected": -669.8192749023438,
816
+ "loss": 0.3194,
817
+ "rewards/accuracies": 0.6625000238418579,
818
+ "rewards/chosen": -0.2555830478668213,
819
+ "rewards/margins": 0.08603434264659882,
820
+ "rewards/rejected": -0.3416173756122589,
821
+ "step": 570
822
+ },
823
+ {
824
+ "epoch": 0.46,
825
+ "learning_rate": 3.2392701251101172e-06,
826
+ "logits/chosen": 0.11188284307718277,
827
+ "logits/rejected": 0.15505322813987732,
828
+ "logps/chosen": -611.50146484375,
829
+ "logps/rejected": -713.7044677734375,
830
+ "loss": 0.286,
831
+ "rewards/accuracies": 0.7437499761581421,
832
+ "rewards/chosen": -0.23539476096630096,
833
+ "rewards/margins": 0.13384294509887695,
834
+ "rewards/rejected": -0.3692377209663391,
835
+ "step": 580
836
+ },
837
+ {
838
+ "epoch": 0.47,
839
+ "learning_rate": 3.1722995515381644e-06,
840
+ "logits/chosen": 0.12050177156925201,
841
+ "logits/rejected": 0.1782829463481903,
842
+ "logps/chosen": -593.8160400390625,
843
+ "logps/rejected": -679.1887817382812,
844
+ "loss": 0.2833,
845
+ "rewards/accuracies": 0.668749988079071,
846
+ "rewards/chosen": -0.251537024974823,
847
+ "rewards/margins": 0.12410496175289154,
848
+ "rewards/rejected": -0.37564200162887573,
849
+ "step": 590
850
+ },
851
+ {
852
+ "epoch": 0.48,
853
+ "learning_rate": 3.1048047389991693e-06,
854
+ "logits/chosen": 0.18543259799480438,
855
+ "logits/rejected": 0.25393250584602356,
856
+ "logps/chosen": -636.8171997070312,
857
+ "logps/rejected": -729.5845947265625,
858
+ "loss": 0.2827,
859
+ "rewards/accuracies": 0.7250000238418579,
860
+ "rewards/chosen": -0.262044221162796,
861
+ "rewards/margins": 0.13976578414440155,
862
+ "rewards/rejected": -0.4018099904060364,
863
+ "step": 600
864
+ },
865
+ {
866
+ "epoch": 0.49,
867
+ "learning_rate": 3.0368383179176584e-06,
868
+ "logits/chosen": 0.06445817649364471,
869
+ "logits/rejected": 0.16580384969711304,
870
+ "logps/chosen": -591.1585693359375,
871
+ "logps/rejected": -707.2025146484375,
872
+ "loss": 0.2758,
873
+ "rewards/accuracies": 0.699999988079071,
874
+ "rewards/chosen": -0.25943881273269653,
875
+ "rewards/margins": 0.13162514567375183,
876
+ "rewards/rejected": -0.39106395840644836,
877
+ "step": 610
878
+ },
879
+ {
880
+ "epoch": 0.5,
881
+ "learning_rate": 2.9684532864643123e-06,
882
+ "logits/chosen": 0.13952665030956268,
883
+ "logits/rejected": 0.22097864747047424,
884
+ "logps/chosen": -617.9655151367188,
885
+ "logps/rejected": -701.77099609375,
886
+ "loss": 0.3062,
887
+ "rewards/accuracies": 0.675000011920929,
888
+ "rewards/chosen": -0.27600663900375366,
889
+ "rewards/margins": 0.1049640029668808,
890
+ "rewards/rejected": -0.38097062706947327,
891
+ "step": 620
892
+ },
893
+ {
894
+ "epoch": 0.5,
895
+ "learning_rate": 2.8997029692295875e-06,
896
+ "logits/chosen": 0.1246538907289505,
897
+ "logits/rejected": 0.26362496614456177,
898
+ "logps/chosen": -671.7306518554688,
899
+ "logps/rejected": -750.6450805664062,
900
+ "loss": 0.247,
901
+ "rewards/accuracies": 0.7562500238418579,
902
+ "rewards/chosen": -0.2702576816082001,
903
+ "rewards/margins": 0.15437254309654236,
904
+ "rewards/rejected": -0.42463022470474243,
905
+ "step": 630
906
+ },
907
+ {
908
+ "epoch": 0.51,
909
+ "learning_rate": 2.8306409756428067e-06,
910
+ "logits/chosen": 0.09484218060970306,
911
+ "logits/rejected": 0.14448456466197968,
912
+ "logps/chosen": -620.1800537109375,
913
+ "logps/rejected": -763.6259765625,
914
+ "loss": 0.2758,
915
+ "rewards/accuracies": 0.706250011920929,
916
+ "rewards/chosen": -0.2843259274959564,
917
+ "rewards/margins": 0.15226764976978302,
918
+ "rewards/rejected": -0.43659359216690063,
919
+ "step": 640
920
+ },
921
+ {
922
+ "epoch": 0.52,
923
+ "learning_rate": 2.761321158169134e-06,
924
+ "logits/chosen": 0.1688222438097,
925
+ "logits/rejected": 0.20292790234088898,
926
+ "logps/chosen": -661.6067504882812,
927
+ "logps/rejected": -832.15966796875,
928
+ "loss": 0.2399,
929
+ "rewards/accuracies": 0.737500011920929,
930
+ "rewards/chosen": -0.293056845664978,
931
+ "rewards/margins": 0.1886308789253235,
932
+ "rewards/rejected": -0.4816876947879791,
933
+ "step": 650
934
+ },
935
+ {
936
+ "epoch": 0.53,
937
+ "learning_rate": 2.6917975703170466e-06,
938
+ "logits/chosen": 0.1265597641468048,
939
+ "logits/rejected": 0.19089344143867493,
940
+ "logps/chosen": -688.8438720703125,
941
+ "logps/rejected": -787.9508056640625,
942
+ "loss": 0.3184,
943
+ "rewards/accuracies": 0.6625000238418579,
944
+ "rewards/chosen": -0.33772650361061096,
945
+ "rewards/margins": 0.11141663789749146,
946
+ "rewards/rejected": -0.44914311170578003,
947
+ "step": 660
948
+ },
949
+ {
950
+ "epoch": 0.54,
951
+ "learning_rate": 2.6221244244890336e-06,
952
+ "logits/chosen": 0.09308207035064697,
953
+ "logits/rejected": 0.16849645972251892,
954
+ "logps/chosen": -662.6124877929688,
955
+ "logps/rejected": -801.0103759765625,
956
+ "loss": 0.2639,
957
+ "rewards/accuracies": 0.668749988079071,
958
+ "rewards/chosen": -0.31664609909057617,
959
+ "rewards/margins": 0.15342995524406433,
960
+ "rewards/rejected": -0.4700759947299957,
961
+ "step": 670
962
+ },
963
+ {
964
+ "epoch": 0.54,
965
+ "learning_rate": 2.5523560497083927e-06,
966
+ "logits/chosen": 0.07288189977407455,
967
+ "logits/rejected": 0.15222066640853882,
968
+ "logps/chosen": -677.3829345703125,
969
+ "logps/rejected": -765.7808837890625,
970
+ "loss": 0.305,
971
+ "rewards/accuracies": 0.675000011920929,
972
+ "rewards/chosen": -0.321521520614624,
973
+ "rewards/margins": 0.12811064720153809,
974
+ "rewards/rejected": -0.4496321678161621,
975
+ "step": 680
976
+ },
977
+ {
978
+ "epoch": 0.55,
979
+ "learning_rate": 2.482546849255096e-06,
980
+ "logits/chosen": 0.07991471141576767,
981
+ "logits/rejected": 0.1225459948182106,
982
+ "logps/chosen": -584.5565185546875,
983
+ "logps/rejected": -719.6222534179688,
984
+ "loss": 0.2702,
985
+ "rewards/accuracies": 0.6937500238418579,
986
+ "rewards/chosen": -0.2603689134120941,
987
+ "rewards/margins": 0.1404421031475067,
988
+ "rewards/rejected": -0.40081101655960083,
989
+ "step": 690
990
+ },
991
+ {
992
+ "epoch": 0.56,
993
+ "learning_rate": 2.4127512582437486e-06,
994
+ "logits/chosen": 0.026161545887589455,
995
+ "logits/rejected": 0.1416565477848053,
996
+ "logps/chosen": -630.5640869140625,
997
+ "logps/rejected": -756.4042358398438,
998
+ "loss": 0.2569,
999
+ "rewards/accuracies": 0.7562500238418579,
1000
+ "rewards/chosen": -0.26197516918182373,
1001
+ "rewards/margins": 0.154434934258461,
1002
+ "rewards/rejected": -0.4164101183414459,
1003
+ "step": 700
1004
+ },
1005
+ {
1006
+ "epoch": 0.57,
1007
+ "learning_rate": 2.3430237011767166e-06,
1008
+ "logits/chosen": 0.11102084815502167,
1009
+ "logits/rejected": 0.16481521725654602,
1010
+ "logps/chosen": -616.1046142578125,
1011
+ "logps/rejected": -770.7633666992188,
1012
+ "loss": 0.2615,
1013
+ "rewards/accuracies": 0.699999988079071,
1014
+ "rewards/chosen": -0.2678528130054474,
1015
+ "rewards/margins": 0.15996505320072174,
1016
+ "rewards/rejected": -0.4278178811073303,
1017
+ "step": 710
1018
+ },
1019
+ {
1020
+ "epoch": 0.58,
1021
+ "learning_rate": 2.2734185495055503e-06,
1022
+ "logits/chosen": 0.07429052889347076,
1023
+ "logits/rejected": 0.15507885813713074,
1024
+ "logps/chosen": -681.5098876953125,
1025
+ "logps/rejected": -819.2258911132812,
1026
+ "loss": 0.2728,
1027
+ "rewards/accuracies": 0.71875,
1028
+ "rewards/chosen": -0.2998114228248596,
1029
+ "rewards/margins": 0.16760887205600739,
1030
+ "rewards/rejected": -0.4674202501773834,
1031
+ "step": 720
1032
+ },
1033
+ {
1034
+ "epoch": 0.58,
1035
+ "learning_rate": 2.2039900792337477e-06,
1036
+ "logits/chosen": 0.0644397884607315,
1037
+ "logits/rejected": 0.17169585824012756,
1038
+ "logps/chosen": -622.50634765625,
1039
+ "logps/rejected": -755.0499267578125,
1040
+ "loss": 0.2805,
1041
+ "rewards/accuracies": 0.7124999761581421,
1042
+ "rewards/chosen": -0.29114142060279846,
1043
+ "rewards/margins": 0.132626011967659,
1044
+ "rewards/rejected": -0.42376741766929626,
1045
+ "step": 730
1046
+ },
1047
+ {
1048
+ "epoch": 0.59,
1049
+ "learning_rate": 2.134792428593971e-06,
1050
+ "logits/chosen": 0.06624677032232285,
1051
+ "logits/rejected": 0.1864895075559616,
1052
+ "logps/chosen": -671.3873901367188,
1053
+ "logps/rejected": -820.2628784179688,
1054
+ "loss": 0.2597,
1055
+ "rewards/accuracies": 0.768750011920929,
1056
+ "rewards/chosen": -0.31325167417526245,
1057
+ "rewards/margins": 0.18044736981391907,
1058
+ "rewards/rejected": -0.49369901418685913,
1059
+ "step": 740
1060
+ },
1061
+ {
1062
+ "epoch": 0.6,
1063
+ "learning_rate": 2.0658795558326745e-06,
1064
+ "logits/chosen": 0.0903557687997818,
1065
+ "logits/rejected": 0.16360041499137878,
1066
+ "logps/chosen": -650.1401977539062,
1067
+ "logps/rejected": -785.5604248046875,
1068
+ "loss": 0.2643,
1069
+ "rewards/accuracies": 0.7124999761581421,
1070
+ "rewards/chosen": -0.31416067481040955,
1071
+ "rewards/margins": 0.15609334409236908,
1072
+ "rewards/rejected": -0.4702540338039398,
1073
+ "step": 750
1074
+ },
1075
+ {
1076
+ "epoch": 0.61,
1077
+ "learning_rate": 1.997305197135089e-06,
1078
+ "logits/chosen": 0.05722861737012863,
1079
+ "logits/rejected": 0.1671527922153473,
1080
+ "logps/chosen": -684.0796508789062,
1081
+ "logps/rejected": -822.2579345703125,
1082
+ "loss": 0.263,
1083
+ "rewards/accuracies": 0.737500011920929,
1084
+ "rewards/chosen": -0.3124196529388428,
1085
+ "rewards/margins": 0.17475970089435577,
1086
+ "rewards/rejected": -0.48717936873435974,
1087
+ "step": 760
1088
+ },
1089
+ {
1090
+ "epoch": 0.62,
1091
+ "learning_rate": 1.9291228247233607e-06,
1092
+ "logits/chosen": 0.07877279818058014,
1093
+ "logits/rejected": 0.1802339255809784,
1094
+ "logps/chosen": -685.1399536132812,
1095
+ "logps/rejected": -797.7412719726562,
1096
+ "loss": 0.2755,
1097
+ "rewards/accuracies": 0.668749988079071,
1098
+ "rewards/chosen": -0.30405086278915405,
1099
+ "rewards/margins": 0.14542993903160095,
1100
+ "rewards/rejected": -0.4494808614253998,
1101
+ "step": 770
1102
+ },
1103
+ {
1104
+ "epoch": 0.62,
1105
+ "learning_rate": 1.8613856051605242e-06,
1106
+ "logits/chosen": 0.07512859255075455,
1107
+ "logits/rejected": 0.15511508285999298,
1108
+ "logps/chosen": -653.9706420898438,
1109
+ "logps/rejected": -778.0037841796875,
1110
+ "loss": 0.2942,
1111
+ "rewards/accuracies": 0.668749988079071,
1112
+ "rewards/chosen": -0.3025520443916321,
1113
+ "rewards/margins": 0.13454928994178772,
1114
+ "rewards/rejected": -0.4371013641357422,
1115
+ "step": 780
1116
+ },
1117
+ {
1118
+ "epoch": 0.63,
1119
+ "learning_rate": 1.7941463578928088e-06,
1120
+ "logits/chosen": 0.04881986230611801,
1121
+ "logits/rejected": 0.1380804181098938,
1122
+ "logps/chosen": -626.0306396484375,
1123
+ "logps/rejected": -762.0904541015625,
1124
+ "loss": 0.259,
1125
+ "rewards/accuracies": 0.706250011920929,
1126
+ "rewards/chosen": -0.2857394218444824,
1127
+ "rewards/margins": 0.15368524193763733,
1128
+ "rewards/rejected": -0.43942466378211975,
1129
+ "step": 790
1130
+ },
1131
+ {
1132
+ "epoch": 0.64,
1133
+ "learning_rate": 1.7274575140626318e-06,
1134
+ "logits/chosen": 0.1028677225112915,
1135
+ "logits/rejected": 0.1538250893354416,
1136
+ "logps/chosen": -597.4656372070312,
1137
+ "logps/rejected": -729.7425537109375,
1138
+ "loss": 0.2688,
1139
+ "rewards/accuracies": 0.737500011920929,
1140
+ "rewards/chosen": -0.26536160707473755,
1141
+ "rewards/margins": 0.16551825404167175,
1142
+ "rewards/rejected": -0.4308798909187317,
1143
+ "step": 800
1144
+ },
1145
+ {
1146
+ "epoch": 0.65,
1147
+ "learning_rate": 1.661371075624363e-06,
1148
+ "logits/chosen": 0.08805406838655472,
1149
+ "logits/rejected": 0.09362033009529114,
1150
+ "logps/chosen": -618.3331298828125,
1151
+ "logps/rejected": -738.7986450195312,
1152
+ "loss": 0.285,
1153
+ "rewards/accuracies": 0.6625000238418579,
1154
+ "rewards/chosen": -0.27929025888442993,
1155
+ "rewards/margins": 0.15323522686958313,
1156
+ "rewards/rejected": -0.4325254559516907,
1157
+ "step": 810
1158
+ },
1159
+ {
1160
+ "epoch": 0.66,
1161
+ "learning_rate": 1.5959385747947697e-06,
1162
+ "logits/chosen": 0.1377531737089157,
1163
+ "logits/rejected": 0.20266500115394592,
1164
+ "logps/chosen": -666.70458984375,
1165
+ "logps/rejected": -725.9954223632812,
1166
+ "loss": 0.3019,
1167
+ "rewards/accuracies": 0.65625,
1168
+ "rewards/chosen": -0.29092544317245483,
1169
+ "rewards/margins": 0.1138964295387268,
1170
+ "rewards/rejected": -0.40482187271118164,
1171
+ "step": 820
1172
+ },
1173
+ {
1174
+ "epoch": 0.66,
1175
+ "learning_rate": 1.5312110338697427e-06,
1176
+ "logits/chosen": 0.0972646176815033,
1177
+ "logits/rejected": 0.19206203520298004,
1178
+ "logps/chosen": -619.4850463867188,
1179
+ "logps/rejected": -751.2019653320312,
1180
+ "loss": 0.2679,
1181
+ "rewards/accuracies": 0.7250000238418579,
1182
+ "rewards/chosen": -0.26260116696357727,
1183
+ "rewards/margins": 0.15384645760059357,
1184
+ "rewards/rejected": -0.41644757986068726,
1185
+ "step": 830
1186
+ },
1187
+ {
1188
+ "epoch": 0.67,
1189
+ "learning_rate": 1.467238925438646e-06,
1190
+ "logits/chosen": 0.054350584745407104,
1191
+ "logits/rejected": 0.14828361570835114,
1192
+ "logps/chosen": -667.3047485351562,
1193
+ "logps/rejected": -786.0614013671875,
1194
+ "loss": 0.2885,
1195
+ "rewards/accuracies": 0.6812499761581421,
1196
+ "rewards/chosen": -0.3170751929283142,
1197
+ "rewards/margins": 0.13434354960918427,
1198
+ "rewards/rejected": -0.4514187276363373,
1199
+ "step": 840
1200
+ },
1201
+ {
1202
+ "epoch": 0.68,
1203
+ "learning_rate": 1.4040721330273063e-06,
1204
+ "logits/chosen": 0.03969626501202583,
1205
+ "logits/rejected": 0.1262727826833725,
1206
+ "logps/chosen": -616.1150512695312,
1207
+ "logps/rejected": -760.9036865234375,
1208
+ "loss": 0.2595,
1209
+ "rewards/accuracies": 0.699999988079071,
1210
+ "rewards/chosen": -0.30283650755882263,
1211
+ "rewards/margins": 0.1451011449098587,
1212
+ "rewards/rejected": -0.44793766736984253,
1213
+ "step": 850
1214
+ },
1215
+ {
1216
+ "epoch": 0.69,
1217
+ "learning_rate": 1.3417599122003464e-06,
1218
+ "logits/chosen": 0.03456869348883629,
1219
+ "logits/rejected": 0.14985020458698273,
1220
+ "logps/chosen": -642.2772216796875,
1221
+ "logps/rejected": -767.5933837890625,
1222
+ "loss": 0.283,
1223
+ "rewards/accuracies": 0.71875,
1224
+ "rewards/chosen": -0.29978328943252563,
1225
+ "rewards/margins": 0.13771583139896393,
1226
+ "rewards/rejected": -0.43749913573265076,
1227
+ "step": 860
1228
+ },
1229
+ {
1230
+ "epoch": 0.7,
1231
+ "learning_rate": 1.280350852153168e-06,
1232
+ "logits/chosen": 0.05811434984207153,
1233
+ "logits/rejected": 0.1481775939464569,
1234
+ "logps/chosen": -623.8426513671875,
1235
+ "logps/rejected": -765.2539672851562,
1236
+ "loss": 0.2507,
1237
+ "rewards/accuracies": 0.7437499761581421,
1238
+ "rewards/chosen": -0.2784281373023987,
1239
+ "rewards/margins": 0.17238983511924744,
1240
+ "rewards/rejected": -0.4508179724216461,
1241
+ "step": 870
1242
+ },
1243
+ {
1244
+ "epoch": 0.7,
1245
+ "learning_rate": 1.2198928378235717e-06,
1246
+ "logits/chosen": 0.1006983295083046,
1247
+ "logits/rejected": 0.1336323320865631,
1248
+ "logps/chosen": -656.2332153320312,
1249
+ "logps/rejected": -837.9879760742188,
1250
+ "loss": 0.264,
1251
+ "rewards/accuracies": 0.6937500238418579,
1252
+ "rewards/chosen": -0.331503301858902,
1253
+ "rewards/margins": 0.17370417714118958,
1254
+ "rewards/rejected": -0.5052074193954468,
1255
+ "step": 880
1256
+ },
1257
+ {
1258
+ "epoch": 0.71,
1259
+ "learning_rate": 1.160433012552508e-06,
1260
+ "logits/chosen": 0.008404137566685677,
1261
+ "logits/rejected": 0.13712240755558014,
1262
+ "logps/chosen": -690.4094848632812,
1263
+ "logps/rejected": -833.4163818359375,
1264
+ "loss": 0.248,
1265
+ "rewards/accuracies": 0.7437499761581421,
1266
+ "rewards/chosen": -0.31104812026023865,
1267
+ "rewards/margins": 0.1767706573009491,
1268
+ "rewards/rejected": -0.48781880736351013,
1269
+ "step": 890
1270
+ },
1271
+ {
1272
+ "epoch": 0.72,
1273
+ "learning_rate": 1.1020177413231334e-06,
1274
+ "logits/chosen": 0.06463897228240967,
1275
+ "logits/rejected": 0.16792722046375275,
1276
+ "logps/chosen": -704.4422607421875,
1277
+ "logps/rejected": -815.9893798828125,
1278
+ "loss": 0.26,
1279
+ "rewards/accuracies": 0.7250000238418579,
1280
+ "rewards/chosen": -0.3422713875770569,
1281
+ "rewards/margins": 0.14494988322257996,
1282
+ "rewards/rejected": -0.4872213304042816,
1283
+ "step": 900
1284
+ },
1285
+ {
1286
+ "epoch": 0.73,
1287
+ "learning_rate": 1.0446925746067768e-06,
1288
+ "logits/chosen": 0.07121414691209793,
1289
+ "logits/rejected": 0.10987003892660141,
1290
+ "logps/chosen": -679.2484130859375,
1291
+ "logps/rejected": -791.3287963867188,
1292
+ "loss": 0.2832,
1293
+ "rewards/accuracies": 0.6875,
1294
+ "rewards/chosen": -0.315243661403656,
1295
+ "rewards/margins": 0.15550537407398224,
1296
+ "rewards/rejected": -0.47074899077415466,
1297
+ "step": 910
1298
+ },
1299
+ {
1300
+ "epoch": 0.74,
1301
+ "learning_rate": 9.88502212844063e-07,
1302
+ "logits/chosen": 0.09697024524211884,
1303
+ "logits/rejected": 0.09367278963327408,
1304
+ "logps/chosen": -693.0015869140625,
1305
+ "logps/rejected": -857.8087768554688,
1306
+ "loss": 0.2695,
1307
+ "rewards/accuracies": 0.71875,
1308
+ "rewards/chosen": -0.3296516239643097,
1309
+ "rewards/margins": 0.1804228574037552,
1310
+ "rewards/rejected": -0.5100744962692261,
1311
+ "step": 920
1312
+ },
1313
+ {
1314
+ "epoch": 0.74,
1315
+ "learning_rate": 9.334904715888496e-07,
1316
+ "logits/chosen": 0.06400808691978455,
1317
+ "logits/rejected": 0.13707710802555084,
1318
+ "logps/chosen": -645.275146484375,
1319
+ "logps/rejected": -822.3630981445312,
1320
+ "loss": 0.2545,
1321
+ "rewards/accuracies": 0.699999988079071,
1322
+ "rewards/chosen": -0.29916906356811523,
1323
+ "rewards/margins": 0.17705786228179932,
1324
+ "rewards/rejected": -0.47622689604759216,
1325
+ "step": 930
1326
+ },
1327
+ {
1328
+ "epoch": 0.75,
1329
+ "learning_rate": 8.797002473421729e-07,
1330
+ "logits/chosen": 0.06906579434871674,
1331
+ "logits/rejected": 0.15590672194957733,
1332
+ "logps/chosen": -628.9729614257812,
1333
+ "logps/rejected": -791.899169921875,
1334
+ "loss": 0.276,
1335
+ "rewards/accuracies": 0.6812499761581421,
1336
+ "rewards/chosen": -0.3074602484703064,
1337
+ "rewards/margins": 0.15690168738365173,
1338
+ "rewards/rejected": -0.46436190605163574,
1339
+ "step": 940
1340
+ },
1341
+ {
1342
+ "epoch": 0.76,
1343
+ "learning_rate": 8.271734841028553e-07,
1344
+ "logits/chosen": 0.04669260233640671,
1345
+ "logits/rejected": 0.13272352516651154,
1346
+ "logps/chosen": -666.2098999023438,
1347
+ "logps/rejected": -846.0867309570312,
1348
+ "loss": 0.2306,
1349
+ "rewards/accuracies": 0.7749999761581421,
1350
+ "rewards/chosen": -0.31410351395606995,
1351
+ "rewards/margins": 0.18806637823581696,
1352
+ "rewards/rejected": -0.5021698474884033,
1353
+ "step": 950
1354
+ },
1355
+ {
1356
+ "epoch": 0.77,
1357
+ "learning_rate": 7.759511406608255e-07,
1358
+ "logits/chosen": 0.12260327488183975,
1359
+ "logits/rejected": 0.158196821808815,
1360
+ "logps/chosen": -710.1339721679688,
1361
+ "logps/rejected": -857.9275512695312,
1362
+ "loss": 0.28,
1363
+ "rewards/accuracies": 0.71875,
1364
+ "rewards/chosen": -0.35299620032310486,
1365
+ "rewards/margins": 0.1603255420923233,
1366
+ "rewards/rejected": -0.5133217573165894,
1367
+ "step": 960
1368
+ },
1369
+ {
1370
+ "epoch": 0.78,
1371
+ "learning_rate": 7.260731586586983e-07,
1372
+ "logits/chosen": 0.026339268311858177,
1373
+ "logits/rejected": 0.06290546804666519,
1374
+ "logps/chosen": -655.7492065429688,
1375
+ "logps/rejected": -841.7448120117188,
1376
+ "loss": 0.2744,
1377
+ "rewards/accuracies": 0.706250011920929,
1378
+ "rewards/chosen": -0.3233845829963684,
1379
+ "rewards/margins": 0.16933201253414154,
1380
+ "rewards/rejected": -0.49271655082702637,
1381
+ "step": 970
1382
+ },
1383
+ {
1384
+ "epoch": 0.78,
1385
+ "learning_rate": 6.775784314464717e-07,
1386
+ "logits/chosen": 0.047948818653821945,
1387
+ "logits/rejected": 0.07574843615293503,
1388
+ "logps/chosen": -642.6578979492188,
1389
+ "logps/rejected": -803.4212646484375,
1390
+ "loss": 0.2468,
1391
+ "rewards/accuracies": 0.6875,
1392
+ "rewards/chosen": -0.3140692412853241,
1393
+ "rewards/margins": 0.17406299710273743,
1394
+ "rewards/rejected": -0.4881322979927063,
1395
+ "step": 980
1396
+ },
1397
+ {
1398
+ "epoch": 0.79,
1399
+ "learning_rate": 6.305047737536707e-07,
1400
+ "logits/chosen": 0.044209472835063934,
1401
+ "logits/rejected": 0.135572150349617,
1402
+ "logps/chosen": -657.6187133789062,
1403
+ "logps/rejected": -825.8516845703125,
1404
+ "loss": 0.2546,
1405
+ "rewards/accuracies": 0.6875,
1406
+ "rewards/chosen": -0.321727991104126,
1407
+ "rewards/margins": 0.1708468794822693,
1408
+ "rewards/rejected": -0.4925749897956848,
1409
+ "step": 990
1410
+ },
1411
+ {
1412
+ "epoch": 0.8,
1413
+ "learning_rate": 5.848888922025553e-07,
1414
+ "logits/chosen": 0.1005764827132225,
1415
+ "logits/rejected": 0.16469302773475647,
1416
+ "logps/chosen": -592.0925903320312,
1417
+ "logps/rejected": -742.73046875,
1418
+ "loss": 0.253,
1419
+ "rewards/accuracies": 0.6875,
1420
+ "rewards/chosen": -0.28060537576675415,
1421
+ "rewards/margins": 0.1545715630054474,
1422
+ "rewards/rejected": -0.43517693877220154,
1423
+ "step": 1000
1424
+ },
1425
+ {
1426
+ "epoch": 0.81,
1427
+ "learning_rate": 5.407663566854008e-07,
1428
+ "logits/chosen": 0.06993698328733444,
1429
+ "logits/rejected": 0.1429472416639328,
1430
+ "logps/chosen": -666.2335205078125,
1431
+ "logps/rejected": -834.9615478515625,
1432
+ "loss": 0.2547,
1433
+ "rewards/accuracies": 0.7124999761581421,
1434
+ "rewards/chosen": -0.3260495364665985,
1435
+ "rewards/margins": 0.16703951358795166,
1436
+ "rewards/rejected": -0.49308905005455017,
1437
+ "step": 1010
1438
+ },
1439
+ {
1440
+ "epoch": 0.82,
1441
+ "learning_rate": 4.981715726281666e-07,
1442
+ "logits/chosen": 0.0734253004193306,
1443
+ "logits/rejected": 0.08005217462778091,
1444
+ "logps/chosen": -614.7472534179688,
1445
+ "logps/rejected": -813.3499755859375,
1446
+ "loss": 0.2446,
1447
+ "rewards/accuracies": 0.731249988079071,
1448
+ "rewards/chosen": -0.3050769567489624,
1449
+ "rewards/margins": 0.18747222423553467,
1450
+ "rewards/rejected": -0.49254918098449707,
1451
+ "step": 1020
1452
+ },
1453
+ {
1454
+ "epoch": 0.82,
1455
+ "learning_rate": 4.5713775416217884e-07,
1456
+ "logits/chosen": 0.042130667716264725,
1457
+ "logits/rejected": 0.15730439126491547,
1458
+ "logps/chosen": -690.7484741210938,
1459
+ "logps/rejected": -863.3248901367188,
1460
+ "loss": 0.2551,
1461
+ "rewards/accuracies": 0.706250011920929,
1462
+ "rewards/chosen": -0.32108041644096375,
1463
+ "rewards/margins": 0.18652570247650146,
1464
+ "rewards/rejected": -0.5076061487197876,
1465
+ "step": 1030
1466
+ },
1467
+ {
1468
+ "epoch": 0.83,
1469
+ "learning_rate": 4.1769689822475147e-07,
1470
+ "logits/chosen": 0.019676199182868004,
1471
+ "logits/rejected": 0.13735586404800415,
1472
+ "logps/chosen": -669.737060546875,
1473
+ "logps/rejected": -836.93798828125,
1474
+ "loss": 0.2325,
1475
+ "rewards/accuracies": 0.699999988079071,
1476
+ "rewards/chosen": -0.3082973062992096,
1477
+ "rewards/margins": 0.18536880612373352,
1478
+ "rewards/rejected": -0.4936660826206207,
1479
+ "step": 1040
1480
+ },
1481
+ {
1482
+ "epoch": 0.84,
1483
+ "learning_rate": 3.798797596089351e-07,
1484
+ "logits/chosen": 0.07972914725542068,
1485
+ "logits/rejected": 0.11014799773693085,
1486
+ "logps/chosen": -691.45947265625,
1487
+ "logps/rejected": -789.2271728515625,
1488
+ "loss": 0.2931,
1489
+ "rewards/accuracies": 0.643750011920929,
1490
+ "rewards/chosen": -0.340253084897995,
1491
+ "rewards/margins": 0.12567462027072906,
1492
+ "rewards/rejected": -0.46592769026756287,
1493
+ "step": 1050
1494
+ },
1495
+ {
1496
+ "epoch": 0.85,
1497
+ "learning_rate": 3.4371582698185636e-07,
1498
+ "logits/chosen": 0.013631382957100868,
1499
+ "logits/rejected": 0.12704741954803467,
1500
+ "logps/chosen": -667.0838012695312,
1501
+ "logps/rejected": -825.3572998046875,
1502
+ "loss": 0.2415,
1503
+ "rewards/accuracies": 0.75,
1504
+ "rewards/chosen": -0.3246777653694153,
1505
+ "rewards/margins": 0.18739624321460724,
1506
+ "rewards/rejected": -0.5120740532875061,
1507
+ "step": 1060
1508
+ },
1509
+ {
1510
+ "epoch": 0.86,
1511
+ "learning_rate": 3.092332998903416e-07,
1512
+ "logits/chosen": 0.05157908797264099,
1513
+ "logits/rejected": 0.17861375212669373,
1514
+ "logps/chosen": -696.72265625,
1515
+ "logps/rejected": -837.0006713867188,
1516
+ "loss": 0.2547,
1517
+ "rewards/accuracies": 0.6937500238418579,
1518
+ "rewards/chosen": -0.33154386281967163,
1519
+ "rewards/margins": 0.17051096260547638,
1520
+ "rewards/rejected": -0.5020548105239868,
1521
+ "step": 1070
1522
+ },
1523
+ {
1524
+ "epoch": 0.86,
1525
+ "learning_rate": 2.764590667717562e-07,
1526
+ "logits/chosen": 0.04824944958090782,
1527
+ "logits/rejected": 0.11944194883108139,
1528
+ "logps/chosen": -641.5201416015625,
1529
+ "logps/rejected": -787.1324462890625,
1530
+ "loss": 0.2571,
1531
+ "rewards/accuracies": 0.699999988079071,
1532
+ "rewards/chosen": -0.29287251830101013,
1533
+ "rewards/margins": 0.1719922125339508,
1534
+ "rewards/rejected": -0.46486473083496094,
1535
+ "step": 1080
1536
+ },
1537
+ {
1538
+ "epoch": 0.87,
1539
+ "learning_rate": 2.454186839872158e-07,
1540
+ "logits/chosen": 0.019634583964943886,
1541
+ "logits/rejected": 0.12763111293315887,
1542
+ "logps/chosen": -705.9700317382812,
1543
+ "logps/rejected": -854.2298583984375,
1544
+ "loss": 0.2456,
1545
+ "rewards/accuracies": 0.7562500238418579,
1546
+ "rewards/chosen": -0.32962867617607117,
1547
+ "rewards/margins": 0.18715420365333557,
1548
+ "rewards/rejected": -0.5167828798294067,
1549
+ "step": 1090
1550
+ },
1551
+ {
1552
+ "epoch": 0.88,
1553
+ "learning_rate": 2.1613635589349756e-07,
1554
+ "logits/chosen": 0.0320691280066967,
1555
+ "logits/rejected": 0.10501092672348022,
1556
+ "logps/chosen": -677.7659912109375,
1557
+ "logps/rejected": -816.9409790039062,
1558
+ "loss": 0.2671,
1559
+ "rewards/accuracies": 0.768750011920929,
1560
+ "rewards/chosen": -0.3169502913951874,
1561
+ "rewards/margins": 0.18356899917125702,
1562
+ "rewards/rejected": -0.500519335269928,
1563
+ "step": 1100
1564
+ },
1565
+ {
1566
+ "epoch": 0.89,
1567
+ "learning_rate": 1.8863491596921745e-07,
1568
+ "logits/chosen": 0.024208758026361465,
1569
+ "logits/rejected": 0.16221018135547638,
1570
+ "logps/chosen": -642.9352416992188,
1571
+ "logps/rejected": -814.4283447265625,
1572
+ "loss": 0.2179,
1573
+ "rewards/accuracies": 0.71875,
1574
+ "rewards/chosen": -0.28280672430992126,
1575
+ "rewards/margins": 0.20693311095237732,
1576
+ "rewards/rejected": -0.4897398352622986,
1577
+ "step": 1110
1578
+ },
1579
+ {
1580
+ "epoch": 0.9,
1581
+ "learning_rate": 1.629358090099639e-07,
1582
+ "logits/chosen": 0.022695127874612808,
1583
+ "logits/rejected": 0.1417960822582245,
1584
+ "logps/chosen": -699.5377197265625,
1585
+ "logps/rejected": -843.7369384765625,
1586
+ "loss": 0.2683,
1587
+ "rewards/accuracies": 0.706250011920929,
1588
+ "rewards/chosen": -0.3325157165527344,
1589
+ "rewards/margins": 0.1784597933292389,
1590
+ "rewards/rejected": -0.5109754800796509,
1591
+ "step": 1120
1592
+ },
1593
+ {
1594
+ "epoch": 0.9,
1595
+ "learning_rate": 1.3905907440629752e-07,
1596
+ "logits/chosen": 0.03463418409228325,
1597
+ "logits/rejected": 0.11470979452133179,
1598
+ "logps/chosen": -696.8150634765625,
1599
+ "logps/rejected": -839.8903198242188,
1600
+ "loss": 0.2622,
1601
+ "rewards/accuracies": 0.706250011920929,
1602
+ "rewards/chosen": -0.33611923456192017,
1603
+ "rewards/margins": 0.16319701075553894,
1604
+ "rewards/rejected": -0.4993163049221039,
1605
+ "step": 1130
1606
+ },
1607
+ {
1608
+ "epoch": 0.91,
1609
+ "learning_rate": 1.1702333051763271e-07,
1610
+ "logits/chosen": 0.06853802502155304,
1611
+ "logits/rejected": 0.11035279929637909,
1612
+ "logps/chosen": -630.0426635742188,
1613
+ "logps/rejected": -830.0460815429688,
1614
+ "loss": 0.2218,
1615
+ "rewards/accuracies": 0.737500011920929,
1616
+ "rewards/chosen": -0.3022988438606262,
1617
+ "rewards/margins": 0.1969214379787445,
1618
+ "rewards/rejected": -0.4992203116416931,
1619
+ "step": 1140
1620
+ },
1621
+ {
1622
+ "epoch": 0.92,
1623
+ "learning_rate": 9.684576015420277e-08,
1624
+ "logits/chosen": 0.07808580249547958,
1625
+ "logits/rejected": 0.14167234301567078,
1626
+ "logps/chosen": -714.4031982421875,
1627
+ "logps/rejected": -866.4649658203125,
1628
+ "loss": 0.2511,
1629
+ "rewards/accuracies": 0.75,
1630
+ "rewards/chosen": -0.3143441081047058,
1631
+ "rewards/margins": 0.19203931093215942,
1632
+ "rewards/rejected": -0.50638347864151,
1633
+ "step": 1150
1634
+ },
1635
+ {
1636
+ "epoch": 0.93,
1637
+ "learning_rate": 7.854209717842231e-08,
1638
+ "logits/chosen": 0.03158559277653694,
1639
+ "logits/rejected": 0.13476143777370453,
1640
+ "logps/chosen": -654.6509399414062,
1641
+ "logps/rejected": -857.1920166015625,
1642
+ "loss": 0.27,
1643
+ "rewards/accuracies": 0.731249988079071,
1644
+ "rewards/chosen": -0.3291003406047821,
1645
+ "rewards/margins": 0.1921570599079132,
1646
+ "rewards/rejected": -0.5212574005126953,
1647
+ "step": 1160
1648
+ },
1649
+ {
1650
+ "epoch": 0.94,
1651
+ "learning_rate": 6.212661423609184e-08,
1652
+ "logits/chosen": 0.01789831928908825,
1653
+ "logits/rejected": 0.11147431284189224,
1654
+ "logps/chosen": -667.9694213867188,
1655
+ "logps/rejected": -807.9912719726562,
1656
+ "loss": 0.2908,
1657
+ "rewards/accuracies": 0.668749988079071,
1658
+ "rewards/chosen": -0.3370179831981659,
1659
+ "rewards/margins": 0.140846848487854,
1660
+ "rewards/rejected": -0.4778648316860199,
1661
+ "step": 1170
1662
+ },
1663
+ {
1664
+ "epoch": 0.94,
1665
+ "learning_rate": 4.761211162702117e-08,
1666
+ "logits/chosen": 0.028112689033150673,
1667
+ "logits/rejected": 0.09534429758787155,
1668
+ "logps/chosen": -656.6475830078125,
1669
+ "logps/rejected": -782.1783447265625,
1670
+ "loss": 0.2762,
1671
+ "rewards/accuracies": 0.668749988079071,
1672
+ "rewards/chosen": -0.31203871965408325,
1673
+ "rewards/margins": 0.1346421241760254,
1674
+ "rewards/rejected": -0.44668087363243103,
1675
+ "step": 1180
1676
+ },
1677
+ {
1678
+ "epoch": 0.95,
1679
+ "learning_rate": 3.5009907323737826e-08,
1680
+ "logits/chosen": 0.12954480946063995,
1681
+ "logits/rejected": 0.14822609722614288,
1682
+ "logps/chosen": -681.8800659179688,
1683
+ "logps/rejected": -832.7755737304688,
1684
+ "loss": 0.2708,
1685
+ "rewards/accuracies": 0.706250011920929,
1686
+ "rewards/chosen": -0.33531031012535095,
1687
+ "rewards/margins": 0.16368630528450012,
1688
+ "rewards/rejected": -0.4989965856075287,
1689
+ "step": 1190
1690
+ },
1691
+ {
1692
+ "epoch": 0.96,
1693
+ "learning_rate": 2.4329828146074096e-08,
1694
+ "logits/chosen": 0.017124827951192856,
1695
+ "logits/rejected": 0.1074015349149704,
1696
+ "logps/chosen": -683.5274658203125,
1697
+ "logps/rejected": -843.3143310546875,
1698
+ "loss": 0.2659,
1699
+ "rewards/accuracies": 0.78125,
1700
+ "rewards/chosen": -0.31970110535621643,
1701
+ "rewards/margins": 0.19150231778621674,
1702
+ "rewards/rejected": -0.511203408241272,
1703
+ "step": 1200
1704
+ },
1705
+ {
1706
+ "epoch": 0.97,
1707
+ "learning_rate": 1.5580202098509078e-08,
1708
+ "logits/chosen": 0.041600506752729416,
1709
+ "logits/rejected": 0.14631414413452148,
1710
+ "logps/chosen": -659.5726318359375,
1711
+ "logps/rejected": -816.1785888671875,
1712
+ "loss": 0.2445,
1713
+ "rewards/accuracies": 0.737500011920929,
1714
+ "rewards/chosen": -0.30856478214263916,
1715
+ "rewards/margins": 0.18453359603881836,
1716
+ "rewards/rejected": -0.49309834837913513,
1717
+ "step": 1210
1718
+ },
1719
+ {
1720
+ "epoch": 0.98,
1721
+ "learning_rate": 8.767851876239075e-09,
1722
+ "logits/chosen": 0.04737471044063568,
1723
+ "logits/rejected": 0.10654322057962418,
1724
+ "logps/chosen": -632.4381103515625,
1725
+ "logps/rejected": -729.2340087890625,
1726
+ "loss": 0.2933,
1727
+ "rewards/accuracies": 0.6312500238418579,
1728
+ "rewards/chosen": -0.3016565442085266,
1729
+ "rewards/margins": 0.13306598365306854,
1730
+ "rewards/rejected": -0.43472257256507874,
1731
+ "step": 1220
1732
+ },
1733
+ {
1734
+ "epoch": 0.98,
1735
+ "learning_rate": 3.8980895450474455e-09,
1736
+ "logits/chosen": 0.04944654554128647,
1737
+ "logits/rejected": 0.12228293716907501,
1738
+ "logps/chosen": -684.4515991210938,
1739
+ "logps/rejected": -818.5169067382812,
1740
+ "loss": 0.2597,
1741
+ "rewards/accuracies": 0.71875,
1742
+ "rewards/chosen": -0.32826223969459534,
1743
+ "rewards/margins": 0.16636613011360168,
1744
+ "rewards/rejected": -0.494628369808197,
1745
+ "step": 1230
1746
+ },
1747
+ {
1748
+ "epoch": 0.99,
1749
+ "learning_rate": 9.747123991141193e-10,
1750
+ "logits/chosen": 0.07023318111896515,
1751
+ "logits/rejected": 0.13329362869262695,
1752
+ "logps/chosen": -622.5777587890625,
1753
+ "logps/rejected": -774.9669799804688,
1754
+ "loss": 0.2367,
1755
+ "rewards/accuracies": 0.731249988079071,
1756
+ "rewards/chosen": -0.2667763829231262,
1757
+ "rewards/margins": 0.18331165611743927,
1758
+ "rewards/rejected": -0.4500879645347595,
1759
+ "step": 1240
1760
+ },
1761
+ {
1762
+ "epoch": 1.0,
1763
+ "learning_rate": 0.0,
1764
+ "logits/chosen": 0.07397739589214325,
1765
+ "logits/rejected": 0.143302321434021,
1766
+ "logps/chosen": -675.6883544921875,
1767
+ "logps/rejected": -813.1781005859375,
1768
+ "loss": 0.2658,
1769
+ "rewards/accuracies": 0.699999988079071,
1770
+ "rewards/chosen": -0.31572216749191284,
1771
+ "rewards/margins": 0.1733875572681427,
1772
+ "rewards/rejected": -0.48910975456237793,
1773
+ "step": 1250
1774
+ },
1775
+ {
1776
+ "epoch": 1.0,
1777
+ "step": 1250,
1778
+ "total_flos": 0.0,
1779
+ "train_loss": 0.2873324268341064,
1780
+ "train_runtime": 11591.8245,
1781
+ "train_samples_per_second": 1.725,
1782
+ "train_steps_per_second": 0.108
1783
+ }
1784
+ ],
1785
+ "logging_steps": 10,
1786
+ "max_steps": 1250,
1787
+ "num_input_tokens_seen": 0,
1788
+ "num_train_epochs": 1,
1789
+ "save_steps": 100,
1790
+ "total_flos": 0.0,
1791
+ "train_batch_size": 4,
1792
+ "trial_name": null,
1793
+ "trial_params": null
1794
+ }