BraylonDash commited on
Commit
7026af4
1 Parent(s): eef9706

Model save

Browse files
README.md ADDED
@@ -0,0 +1,61 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: peft
3
+ tags:
4
+ - trl
5
+ - dpo
6
+ - generated_from_trainer
7
+ base_model: DUAL-GPO/zephyr-7b-dpo-new-lora-v1-merged
8
+ model-index:
9
+ - name: zephyr-7b-dpo-0k-15k-i1
10
+ results: []
11
+ ---
12
+
13
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
14
+ should probably proofread and complete it, then remove this comment. -->
15
+
16
+ # zephyr-7b-dpo-0k-15k-i1
17
+
18
+ This model is a fine-tuned version of [DUAL-GPO/zephyr-7b-dpo-new-lora-v1-merged](https://huggingface.co/DUAL-GPO/zephyr-7b-dpo-new-lora-v1-merged) on the None dataset.
19
+
20
+ ## Model description
21
+
22
+ More information needed
23
+
24
+ ## Intended uses & limitations
25
+
26
+ More information needed
27
+
28
+ ## Training and evaluation data
29
+
30
+ More information needed
31
+
32
+ ## Training procedure
33
+
34
+ ### Training hyperparameters
35
+
36
+ The following hyperparameters were used during training:
37
+ - learning_rate: 5e-06
38
+ - train_batch_size: 2
39
+ - eval_batch_size: 2
40
+ - seed: 42
41
+ - distributed_type: multi-GPU
42
+ - num_devices: 3
43
+ - gradient_accumulation_steps: 2
44
+ - total_train_batch_size: 12
45
+ - total_eval_batch_size: 6
46
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
47
+ - lr_scheduler_type: cosine
48
+ - lr_scheduler_warmup_ratio: 0.1
49
+ - num_epochs: 1
50
+
51
+ ### Training results
52
+
53
+
54
+
55
+ ### Framework versions
56
+
57
+ - PEFT 0.7.1
58
+ - Transformers 4.36.2
59
+ - Pytorch 2.1.2
60
+ - Datasets 2.14.6
61
+ - Tokenizers 0.15.2
adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:f0a519fdfed6bde6749d651cfeb57803431ce084dadf79e08090f72139323228
3
  size 671150064
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e010d83359654572e49c061ffe67583d86e7f0e692783cb0d011cbf8e454ca6
3
  size 671150064
all_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "train_loss": 0.37276097068786623,
4
+ "train_runtime": 8917.3793,
5
+ "train_samples": 15000,
6
+ "train_samples_per_second": 1.682,
7
+ "train_steps_per_second": 0.14
8
+ }
emissions.csv ADDED
@@ -0,0 +1,2 @@
 
 
 
1
+ timestamp,project_name,run_id,duration,emissions,emissions_rate,cpu_power,gpu_power,ram_power,cpu_energy,gpu_energy,ram_energy,energy_consumed,country_name,country_iso_code,region,cloud_provider,cloud_region,os,python_version,codecarbon_version,cpu_count,cpu_model,gpu_count,gpu_model,longitude,latitude,ram_total_size,tracking_mode,on_cloud,pue
2
+ 2024-09-17T16:05:41,codecarbon,65082978-b7df-4f64-90ee-b51f09ec1de0,8917.383047103882,0.004530512283172767,5.080540175566587e-07,42.5,613.2070000000001,188.74309015274048,0.10527343996597656,1.334327038381135,0.4662033038276709,1.9058037821747817,Canada,CAN,quebec,,,Linux-5.15.0-84-generic-x86_64-with-glibc2.35,3.10.14,2.2.3,32,Intel(R) Xeon(R) W-3335 CPU @ 3.40GHz,4,4 x NVIDIA GeForce RTX 4090,-71.2,46.8,503.3149070739746,machine,N,1.0
runs/Sep17_13-35-53_gpu4-119-5/events.out.tfevents.1726544224.gpu4-119-5.2547334.0 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:f9e6a32ec1bf243487d5cdc35ad092e301c07c05f29469edc53b26ce09d26769
3
- size 57861
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e87b6996527fd9536b4dfc0f5a7c84f3f7e3db3476fac95113a0c29c78175e31
3
+ size 58849
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "train_loss": 0.37276097068786623,
4
+ "train_runtime": 8917.3793,
5
+ "train_samples": 15000,
6
+ "train_samples_per_second": 1.682,
7
+ "train_steps_per_second": 0.14
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": -2.6577353477478027,
15
+ "logits/rejected": -2.043900489807129,
16
+ "logps/chosen": -505.98724365234375,
17
+ "logps/rejected": -319.40179443359375,
18
+ "loss": 0.6931,
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": -2.4769232273101807,
29
+ "logits/rejected": -2.134289503097534,
30
+ "logps/chosen": -285.32098388671875,
31
+ "logps/rejected": -191.55812072753906,
32
+ "loss": 0.6932,
33
+ "rewards/accuracies": 0.3333333432674408,
34
+ "rewards/chosen": 0.00036080353311263025,
35
+ "rewards/margins": -9.262199455406517e-05,
36
+ "rewards/rejected": 0.0004534254549071193,
37
+ "step": 10
38
+ },
39
+ {
40
+ "epoch": 0.02,
41
+ "learning_rate": 8.000000000000001e-07,
42
+ "logits/chosen": -2.3606486320495605,
43
+ "logits/rejected": -2.1270670890808105,
44
+ "logps/chosen": -271.58087158203125,
45
+ "logps/rejected": -208.8097381591797,
46
+ "loss": 0.6934,
47
+ "rewards/accuracies": 0.2750000059604645,
48
+ "rewards/chosen": -0.0005878033698536456,
49
+ "rewards/margins": -0.0015078920405358076,
50
+ "rewards/rejected": 0.0009200887870974839,
51
+ "step": 20
52
+ },
53
+ {
54
+ "epoch": 0.02,
55
+ "learning_rate": 1.2000000000000002e-06,
56
+ "logits/chosen": -2.277930498123169,
57
+ "logits/rejected": -2.2482991218566895,
58
+ "logps/chosen": -269.87554931640625,
59
+ "logps/rejected": -288.9063720703125,
60
+ "loss": 0.693,
61
+ "rewards/accuracies": 0.5249999761581421,
62
+ "rewards/chosen": 0.002121593803167343,
63
+ "rewards/margins": 0.0005655083805322647,
64
+ "rewards/rejected": 0.0015560855390504003,
65
+ "step": 30
66
+ },
67
+ {
68
+ "epoch": 0.03,
69
+ "learning_rate": 1.6000000000000001e-06,
70
+ "logits/chosen": -2.508906841278076,
71
+ "logits/rejected": -2.341235637664795,
72
+ "logps/chosen": -210.5034942626953,
73
+ "logps/rejected": -181.61361694335938,
74
+ "loss": 0.6926,
75
+ "rewards/accuracies": 0.5,
76
+ "rewards/chosen": 0.0023279793094843626,
77
+ "rewards/margins": 0.0018606262747198343,
78
+ "rewards/rejected": 0.0004673528019338846,
79
+ "step": 40
80
+ },
81
+ {
82
+ "epoch": 0.04,
83
+ "learning_rate": 2.0000000000000003e-06,
84
+ "logits/chosen": -2.36310076713562,
85
+ "logits/rejected": -2.3391873836517334,
86
+ "logps/chosen": -195.63548278808594,
87
+ "logps/rejected": -211.3179168701172,
88
+ "loss": 0.6924,
89
+ "rewards/accuracies": 0.5,
90
+ "rewards/chosen": 0.0008935689693316817,
91
+ "rewards/margins": 0.0021855938248336315,
92
+ "rewards/rejected": -0.0012920246226713061,
93
+ "step": 50
94
+ },
95
+ {
96
+ "epoch": 0.05,
97
+ "learning_rate": 2.4000000000000003e-06,
98
+ "logits/chosen": -2.4950430393218994,
99
+ "logits/rejected": -2.303813934326172,
100
+ "logps/chosen": -244.68856811523438,
101
+ "logps/rejected": -277.8478698730469,
102
+ "loss": 0.6922,
103
+ "rewards/accuracies": 0.5,
104
+ "rewards/chosen": -0.010385877452790737,
105
+ "rewards/margins": -0.0004113853501621634,
106
+ "rewards/rejected": -0.009974492713809013,
107
+ "step": 60
108
+ },
109
+ {
110
+ "epoch": 0.06,
111
+ "learning_rate": 2.8000000000000003e-06,
112
+ "logits/chosen": -2.2493319511413574,
113
+ "logits/rejected": -2.266305446624756,
114
+ "logps/chosen": -229.941162109375,
115
+ "logps/rejected": -217.0909881591797,
116
+ "loss": 0.6901,
117
+ "rewards/accuracies": 0.550000011920929,
118
+ "rewards/chosen": -0.011110167019069195,
119
+ "rewards/margins": 0.010483531281352043,
120
+ "rewards/rejected": -0.021593699231743813,
121
+ "step": 70
122
+ },
123
+ {
124
+ "epoch": 0.06,
125
+ "learning_rate": 3.2000000000000003e-06,
126
+ "logits/chosen": -2.399794816970825,
127
+ "logits/rejected": -2.393554449081421,
128
+ "logps/chosen": -266.1021423339844,
129
+ "logps/rejected": -263.08526611328125,
130
+ "loss": 0.6861,
131
+ "rewards/accuracies": 0.675000011920929,
132
+ "rewards/chosen": -0.014540717005729675,
133
+ "rewards/margins": 0.021736864000558853,
134
+ "rewards/rejected": -0.03627758100628853,
135
+ "step": 80
136
+ },
137
+ {
138
+ "epoch": 0.07,
139
+ "learning_rate": 3.6000000000000003e-06,
140
+ "logits/chosen": -2.32847261428833,
141
+ "logits/rejected": -1.991908073425293,
142
+ "logps/chosen": -268.0962219238281,
143
+ "logps/rejected": -179.7996826171875,
144
+ "loss": 0.6863,
145
+ "rewards/accuracies": 0.625,
146
+ "rewards/chosen": -0.027069438248872757,
147
+ "rewards/margins": 0.021108368411660194,
148
+ "rewards/rejected": -0.048177801072597504,
149
+ "step": 90
150
+ },
151
+ {
152
+ "epoch": 0.08,
153
+ "learning_rate": 4.000000000000001e-06,
154
+ "logits/chosen": -2.3088860511779785,
155
+ "logits/rejected": -2.312539577484131,
156
+ "logps/chosen": -274.9145202636719,
157
+ "logps/rejected": -250.87716674804688,
158
+ "loss": 0.6898,
159
+ "rewards/accuracies": 0.6000000238418579,
160
+ "rewards/chosen": -0.031584225594997406,
161
+ "rewards/margins": 0.010575409047305584,
162
+ "rewards/rejected": -0.042159631848335266,
163
+ "step": 100
164
+ },
165
+ {
166
+ "epoch": 0.09,
167
+ "learning_rate": 4.4e-06,
168
+ "logits/chosen": -2.2585253715515137,
169
+ "logits/rejected": -2.144835948944092,
170
+ "logps/chosen": -211.51986694335938,
171
+ "logps/rejected": -179.9736785888672,
172
+ "loss": 0.6819,
173
+ "rewards/accuracies": 0.574999988079071,
174
+ "rewards/chosen": 0.005594385787844658,
175
+ "rewards/margins": 0.03688768297433853,
176
+ "rewards/rejected": -0.03129329904913902,
177
+ "step": 110
178
+ },
179
+ {
180
+ "epoch": 0.1,
181
+ "learning_rate": 4.800000000000001e-06,
182
+ "logits/chosen": -2.1611924171447754,
183
+ "logits/rejected": -2.171570301055908,
184
+ "logps/chosen": -237.06216430664062,
185
+ "logps/rejected": -269.0331115722656,
186
+ "loss": 0.6703,
187
+ "rewards/accuracies": 0.625,
188
+ "rewards/chosen": -0.007297643460333347,
189
+ "rewards/margins": 0.047815777361392975,
190
+ "rewards/rejected": -0.055113423615694046,
191
+ "step": 120
192
+ },
193
+ {
194
+ "epoch": 0.1,
195
+ "learning_rate": 4.999756310023261e-06,
196
+ "logits/chosen": -2.448563814163208,
197
+ "logits/rejected": -2.3213229179382324,
198
+ "logps/chosen": -235.3451690673828,
199
+ "logps/rejected": -192.87477111816406,
200
+ "loss": 0.6764,
201
+ "rewards/accuracies": 0.574999988079071,
202
+ "rewards/chosen": -0.011573897674679756,
203
+ "rewards/margins": 0.052028268575668335,
204
+ "rewards/rejected": -0.06360216438770294,
205
+ "step": 130
206
+ },
207
+ {
208
+ "epoch": 0.11,
209
+ "learning_rate": 4.997807075247147e-06,
210
+ "logits/chosen": -2.3375329971313477,
211
+ "logits/rejected": -2.034670352935791,
212
+ "logps/chosen": -199.99537658691406,
213
+ "logps/rejected": -180.91787719726562,
214
+ "loss": 0.6737,
215
+ "rewards/accuracies": 0.574999988079071,
216
+ "rewards/chosen": -0.08778422325849533,
217
+ "rewards/margins": 0.04929418861865997,
218
+ "rewards/rejected": -0.1370784044265747,
219
+ "step": 140
220
+ },
221
+ {
222
+ "epoch": 0.12,
223
+ "learning_rate": 4.993910125649561e-06,
224
+ "logits/chosen": -2.323729991912842,
225
+ "logits/rejected": -2.20963191986084,
226
+ "logps/chosen": -229.9593048095703,
227
+ "logps/rejected": -198.73062133789062,
228
+ "loss": 0.6566,
229
+ "rewards/accuracies": 0.6499999761581421,
230
+ "rewards/chosen": -0.018689800053834915,
231
+ "rewards/margins": 0.08985014259815216,
232
+ "rewards/rejected": -0.10853993892669678,
233
+ "step": 150
234
+ },
235
+ {
236
+ "epoch": 0.13,
237
+ "learning_rate": 4.988068499954578e-06,
238
+ "logits/chosen": -2.215893507003784,
239
+ "logits/rejected": -2.2889232635498047,
240
+ "logps/chosen": -277.0525207519531,
241
+ "logps/rejected": -287.1373596191406,
242
+ "loss": 0.6778,
243
+ "rewards/accuracies": 0.44999998807907104,
244
+ "rewards/chosen": -0.24399030208587646,
245
+ "rewards/margins": 0.03944157436490059,
246
+ "rewards/rejected": -0.28343185782432556,
247
+ "step": 160
248
+ },
249
+ {
250
+ "epoch": 0.14,
251
+ "learning_rate": 4.980286753286196e-06,
252
+ "logits/chosen": -2.4302048683166504,
253
+ "logits/rejected": -2.032261848449707,
254
+ "logps/chosen": -285.89691162109375,
255
+ "logps/rejected": -248.88278198242188,
256
+ "loss": 0.6419,
257
+ "rewards/accuracies": 0.44999998807907104,
258
+ "rewards/chosen": -0.18989089131355286,
259
+ "rewards/margins": 0.04870374873280525,
260
+ "rewards/rejected": -0.23859462141990662,
261
+ "step": 170
262
+ },
263
+ {
264
+ "epoch": 0.14,
265
+ "learning_rate": 4.970570953616383e-06,
266
+ "logits/chosen": -2.2379422187805176,
267
+ "logits/rejected": -2.230316400527954,
268
+ "logps/chosen": -235.77371215820312,
269
+ "logps/rejected": -240.2162628173828,
270
+ "loss": 0.6573,
271
+ "rewards/accuracies": 0.42500001192092896,
272
+ "rewards/chosen": -0.0921173095703125,
273
+ "rewards/margins": 0.044399768114089966,
274
+ "rewards/rejected": -0.13651707768440247,
275
+ "step": 180
276
+ },
277
+ {
278
+ "epoch": 0.15,
279
+ "learning_rate": 4.958928677033465e-06,
280
+ "logits/chosen": -2.2377524375915527,
281
+ "logits/rejected": -2.1663427352905273,
282
+ "logps/chosen": -257.6595764160156,
283
+ "logps/rejected": -243.8786163330078,
284
+ "loss": 0.6519,
285
+ "rewards/accuracies": 0.6499999761581421,
286
+ "rewards/chosen": -0.2135959416627884,
287
+ "rewards/margins": 0.13946585357189178,
288
+ "rewards/rejected": -0.3530617952346802,
289
+ "step": 190
290
+ },
291
+ {
292
+ "epoch": 0.16,
293
+ "learning_rate": 4.9453690018345144e-06,
294
+ "logits/chosen": -2.29764723777771,
295
+ "logits/rejected": -2.1510560512542725,
296
+ "logps/chosen": -285.13702392578125,
297
+ "logps/rejected": -305.1742858886719,
298
+ "loss": 0.6381,
299
+ "rewards/accuracies": 0.574999988079071,
300
+ "rewards/chosen": -0.3077496588230133,
301
+ "rewards/margins": 0.12263032048940659,
302
+ "rewards/rejected": -0.4303800165653229,
303
+ "step": 200
304
+ },
305
+ {
306
+ "epoch": 0.17,
307
+ "learning_rate": 4.9299025014463665e-06,
308
+ "logits/chosen": -2.4579226970672607,
309
+ "logits/rejected": -2.097339153289795,
310
+ "logps/chosen": -314.6923828125,
311
+ "logps/rejected": -261.67041015625,
312
+ "loss": 0.613,
313
+ "rewards/accuracies": 0.699999988079071,
314
+ "rewards/chosen": -0.4118514657020569,
315
+ "rewards/margins": 0.30009493231773376,
316
+ "rewards/rejected": -0.7119464874267578,
317
+ "step": 210
318
+ },
319
+ {
320
+ "epoch": 0.18,
321
+ "learning_rate": 4.912541236180779e-06,
322
+ "logits/chosen": -2.31429123878479,
323
+ "logits/rejected": -1.8840612173080444,
324
+ "logps/chosen": -303.4709777832031,
325
+ "logps/rejected": -296.21197509765625,
326
+ "loss": 0.6429,
327
+ "rewards/accuracies": 0.625,
328
+ "rewards/chosen": -0.6690406799316406,
329
+ "rewards/margins": 0.3159363567829132,
330
+ "rewards/rejected": -0.984977126121521,
331
+ "step": 220
332
+ },
333
+ {
334
+ "epoch": 0.18,
335
+ "learning_rate": 4.893298743830168e-06,
336
+ "logits/chosen": -2.283473491668701,
337
+ "logits/rejected": -2.319681167602539,
338
+ "logps/chosen": -240.64303588867188,
339
+ "logps/rejected": -283.9293212890625,
340
+ "loss": 0.6464,
341
+ "rewards/accuracies": 0.6000000238418579,
342
+ "rewards/chosen": -0.4180280268192291,
343
+ "rewards/margins": 0.16594772040843964,
344
+ "rewards/rejected": -0.5839756727218628,
345
+ "step": 230
346
+ },
347
+ {
348
+ "epoch": 0.19,
349
+ "learning_rate": 4.8721900291112415e-06,
350
+ "logits/chosen": -2.2171471118927,
351
+ "logits/rejected": -2.035769462585449,
352
+ "logps/chosen": -257.6659851074219,
353
+ "logps/rejected": -258.67364501953125,
354
+ "loss": 0.624,
355
+ "rewards/accuracies": 0.5249999761581421,
356
+ "rewards/chosen": -0.5104702711105347,
357
+ "rewards/margins": 0.13771632313728333,
358
+ "rewards/rejected": -0.6481865644454956,
359
+ "step": 240
360
+ },
361
+ {
362
+ "epoch": 0.2,
363
+ "learning_rate": 4.849231551964771e-06,
364
+ "logits/chosen": -2.0292515754699707,
365
+ "logits/rejected": -1.9135668277740479,
366
+ "logps/chosen": -286.9338684082031,
367
+ "logps/rejected": -313.9806213378906,
368
+ "loss": 0.6291,
369
+ "rewards/accuracies": 0.550000011920929,
370
+ "rewards/chosen": -1.169675588607788,
371
+ "rewards/margins": 0.24345961213111877,
372
+ "rewards/rejected": -1.4131351709365845,
373
+ "step": 250
374
+ },
375
+ {
376
+ "epoch": 0.21,
377
+ "learning_rate": 4.824441214720629e-06,
378
+ "logits/chosen": -2.203101634979248,
379
+ "logits/rejected": -1.946657419204712,
380
+ "logps/chosen": -382.1039123535156,
381
+ "logps/rejected": -404.14892578125,
382
+ "loss": 0.5794,
383
+ "rewards/accuracies": 0.6499999761581421,
384
+ "rewards/chosen": -1.0869503021240234,
385
+ "rewards/margins": 0.3835533559322357,
386
+ "rewards/rejected": -1.470503568649292,
387
+ "step": 260
388
+ },
389
+ {
390
+ "epoch": 0.22,
391
+ "learning_rate": 4.7978383481380865e-06,
392
+ "logits/chosen": -2.212496519088745,
393
+ "logits/rejected": -1.996314287185669,
394
+ "logps/chosen": -307.20587158203125,
395
+ "logps/rejected": -306.6591796875,
396
+ "loss": 0.5909,
397
+ "rewards/accuracies": 0.7250000238418579,
398
+ "rewards/chosen": -0.5885969400405884,
399
+ "rewards/margins": 0.3659537732601166,
400
+ "rewards/rejected": -0.9545507431030273,
401
+ "step": 270
402
+ },
403
+ {
404
+ "epoch": 0.22,
405
+ "learning_rate": 4.769443696332272e-06,
406
+ "logits/chosen": -1.9687378406524658,
407
+ "logits/rejected": -2.04164719581604,
408
+ "logps/chosen": -249.0873565673828,
409
+ "logps/rejected": -323.9103088378906,
410
+ "loss": 0.5854,
411
+ "rewards/accuracies": 0.675000011920929,
412
+ "rewards/chosen": -0.8271968960762024,
413
+ "rewards/margins": 0.3957016170024872,
414
+ "rewards/rejected": -1.2228984832763672,
415
+ "step": 280
416
+ },
417
+ {
418
+ "epoch": 0.23,
419
+ "learning_rate": 4.7392794005985324e-06,
420
+ "logits/chosen": -2.2899293899536133,
421
+ "logits/rejected": -2.000715970993042,
422
+ "logps/chosen": -374.6996154785156,
423
+ "logps/rejected": -359.8489685058594,
424
+ "loss": 0.5851,
425
+ "rewards/accuracies": 0.625,
426
+ "rewards/chosen": -0.9794802665710449,
427
+ "rewards/margins": 0.4062279760837555,
428
+ "rewards/rejected": -1.385708212852478,
429
+ "step": 290
430
+ },
431
+ {
432
+ "epoch": 0.24,
433
+ "learning_rate": 4.707368982147318e-06,
434
+ "logits/chosen": -2.028409242630005,
435
+ "logits/rejected": -1.9832637310028076,
436
+ "logps/chosen": -426.76385498046875,
437
+ "logps/rejected": -396.494384765625,
438
+ "loss": 0.5501,
439
+ "rewards/accuracies": 0.625,
440
+ "rewards/chosen": -1.3581629991531372,
441
+ "rewards/margins": 0.4999775290489197,
442
+ "rewards/rejected": -1.8581405878067017,
443
+ "step": 300
444
+ },
445
+ {
446
+ "epoch": 0.25,
447
+ "learning_rate": 4.673737323763048e-06,
448
+ "logits/chosen": -1.773838996887207,
449
+ "logits/rejected": -1.8148994445800781,
450
+ "logps/chosen": -377.904541015625,
451
+ "logps/rejected": -456.285400390625,
452
+ "loss": 0.6464,
453
+ "rewards/accuracies": 0.44999998807907104,
454
+ "rewards/chosen": -1.8526346683502197,
455
+ "rewards/margins": 0.3698161244392395,
456
+ "rewards/rejected": -2.2224507331848145,
457
+ "step": 310
458
+ },
459
+ {
460
+ "epoch": 0.26,
461
+ "learning_rate": 4.638410650401267e-06,
462
+ "logits/chosen": -2.000642776489258,
463
+ "logits/rejected": -1.6485764980316162,
464
+ "logps/chosen": -337.97711181640625,
465
+ "logps/rejected": -298.59136962890625,
466
+ "loss": 0.6445,
467
+ "rewards/accuracies": 0.574999988079071,
468
+ "rewards/chosen": -0.8390266299247742,
469
+ "rewards/margins": 0.39421704411506653,
470
+ "rewards/rejected": -1.233243703842163,
471
+ "step": 320
472
+ },
473
+ {
474
+ "epoch": 0.26,
475
+ "learning_rate": 4.601416508739211e-06,
476
+ "logits/chosen": -2.2264564037323,
477
+ "logits/rejected": -2.249499797821045,
478
+ "logps/chosen": -326.7769470214844,
479
+ "logps/rejected": -301.43341064453125,
480
+ "loss": 0.6091,
481
+ "rewards/accuracies": 0.4749999940395355,
482
+ "rewards/chosen": -0.5400243997573853,
483
+ "rewards/margins": 0.13523522019386292,
484
+ "rewards/rejected": -0.6752597093582153,
485
+ "step": 330
486
+ },
487
+ {
488
+ "epoch": 0.27,
489
+ "learning_rate": 4.562783745695738e-06,
490
+ "logits/chosen": -2.1335179805755615,
491
+ "logits/rejected": -1.8753182888031006,
492
+ "logps/chosen": -247.4087677001953,
493
+ "logps/rejected": -241.4481201171875,
494
+ "loss": 0.6052,
495
+ "rewards/accuracies": 0.6000000238418579,
496
+ "rewards/chosen": -0.5356850624084473,
497
+ "rewards/margins": 0.29491013288497925,
498
+ "rewards/rejected": -0.8305951952934265,
499
+ "step": 340
500
+ },
501
+ {
502
+ "epoch": 0.28,
503
+ "learning_rate": 4.522542485937369e-06,
504
+ "logits/chosen": -2.284304141998291,
505
+ "logits/rejected": -1.9852168560028076,
506
+ "logps/chosen": -362.0438232421875,
507
+ "logps/rejected": -358.851806640625,
508
+ "loss": 0.5885,
509
+ "rewards/accuracies": 0.550000011920929,
510
+ "rewards/chosen": -1.0619395971298218,
511
+ "rewards/margins": 0.26577672362327576,
512
+ "rewards/rejected": -1.32771635055542,
513
+ "step": 350
514
+ },
515
+ {
516
+ "epoch": 0.29,
517
+ "learning_rate": 4.4807241083879774e-06,
518
+ "logits/chosen": -2.22493052482605,
519
+ "logits/rejected": -2.0395853519439697,
520
+ "logps/chosen": -384.7464904785156,
521
+ "logps/rejected": -394.67999267578125,
522
+ "loss": 0.541,
523
+ "rewards/accuracies": 0.6499999761581421,
524
+ "rewards/chosen": -1.2721855640411377,
525
+ "rewards/margins": 0.4639417231082916,
526
+ "rewards/rejected": -1.736127495765686,
527
+ "step": 360
528
+ },
529
+ {
530
+ "epoch": 0.3,
531
+ "learning_rate": 4.437361221760449e-06,
532
+ "logits/chosen": -2.074812412261963,
533
+ "logits/rejected": -2.015082597732544,
534
+ "logps/chosen": -367.3848876953125,
535
+ "logps/rejected": -382.31573486328125,
536
+ "loss": 0.644,
537
+ "rewards/accuracies": 0.5249999761581421,
538
+ "rewards/chosen": -1.7959085702896118,
539
+ "rewards/margins": 0.18800289928913116,
540
+ "rewards/rejected": -1.983911156654358,
541
+ "step": 370
542
+ },
543
+ {
544
+ "epoch": 0.3,
545
+ "learning_rate": 4.3924876391293915e-06,
546
+ "logits/chosen": -1.972678542137146,
547
+ "logits/rejected": -1.9295990467071533,
548
+ "logps/chosen": -358.55126953125,
549
+ "logps/rejected": -405.638427734375,
550
+ "loss": 0.567,
551
+ "rewards/accuracies": 0.5,
552
+ "rewards/chosen": -1.6957597732543945,
553
+ "rewards/margins": 0.3802381753921509,
554
+ "rewards/rejected": -2.075997829437256,
555
+ "step": 380
556
+ },
557
+ {
558
+ "epoch": 0.31,
559
+ "learning_rate": 4.346138351564711e-06,
560
+ "logits/chosen": -2.031592607498169,
561
+ "logits/rejected": -1.8613895177841187,
562
+ "logps/chosen": -358.3297119140625,
563
+ "logps/rejected": -360.80023193359375,
564
+ "loss": 0.5674,
565
+ "rewards/accuracies": 0.6000000238418579,
566
+ "rewards/chosen": -1.6673065423965454,
567
+ "rewards/margins": 0.44297677278518677,
568
+ "rewards/rejected": -2.110283374786377,
569
+ "step": 390
570
+ },
571
+ {
572
+ "epoch": 0.32,
573
+ "learning_rate": 4.2983495008466285e-06,
574
+ "logits/chosen": -2.104931116104126,
575
+ "logits/rejected": -2.1332077980041504,
576
+ "logps/chosen": -411.54168701171875,
577
+ "logps/rejected": -463.42236328125,
578
+ "loss": 0.666,
579
+ "rewards/accuracies": 0.5249999761581421,
580
+ "rewards/chosen": -1.9785181283950806,
581
+ "rewards/margins": 0.3848969638347626,
582
+ "rewards/rejected": -2.363415241241455,
583
+ "step": 400
584
+ },
585
+ {
586
+ "epoch": 0.33,
587
+ "learning_rate": 4.249158351283414e-06,
588
+ "logits/chosen": -2.211599826812744,
589
+ "logits/rejected": -1.9848130941390991,
590
+ "logps/chosen": -430.388427734375,
591
+ "logps/rejected": -470.9698181152344,
592
+ "loss": 0.6333,
593
+ "rewards/accuracies": 0.625,
594
+ "rewards/chosen": -1.7057955265045166,
595
+ "rewards/margins": 0.3454364538192749,
596
+ "rewards/rejected": -2.051231861114502,
597
+ "step": 410
598
+ },
599
+ {
600
+ "epoch": 0.34,
601
+ "learning_rate": 4.198603260653792e-06,
602
+ "logits/chosen": -2.3403067588806152,
603
+ "logits/rejected": -2.1165671348571777,
604
+ "logps/chosen": -418.66558837890625,
605
+ "logps/rejected": -407.5811462402344,
606
+ "loss": 0.5813,
607
+ "rewards/accuracies": 0.6000000238418579,
608
+ "rewards/chosen": -1.5130822658538818,
609
+ "rewards/margins": 0.13823339343070984,
610
+ "rewards/rejected": -1.651315689086914,
611
+ "step": 420
612
+ },
613
+ {
614
+ "epoch": 0.34,
615
+ "learning_rate": 4.146723650296701e-06,
616
+ "logits/chosen": -2.2038991451263428,
617
+ "logits/rejected": -2.2611026763916016,
618
+ "logps/chosen": -443.9969177246094,
619
+ "logps/rejected": -509.6358337402344,
620
+ "loss": 0.6094,
621
+ "rewards/accuracies": 0.699999988079071,
622
+ "rewards/chosen": -1.4418891668319702,
623
+ "rewards/margins": 0.5296822190284729,
624
+ "rewards/rejected": -1.9715715646743774,
625
+ "step": 430
626
+ },
627
+ {
628
+ "epoch": 0.35,
629
+ "learning_rate": 4.093559974371725e-06,
630
+ "logits/chosen": -2.236635684967041,
631
+ "logits/rejected": -1.8922793865203857,
632
+ "logps/chosen": -270.41839599609375,
633
+ "logps/rejected": -274.03546142578125,
634
+ "loss": 0.5395,
635
+ "rewards/accuracies": 0.6000000238418579,
636
+ "rewards/chosen": -0.8648042678833008,
637
+ "rewards/margins": 0.4120641350746155,
638
+ "rewards/rejected": -1.2768683433532715,
639
+ "step": 440
640
+ },
641
+ {
642
+ "epoch": 0.36,
643
+ "learning_rate": 4.039153688314146e-06,
644
+ "logits/chosen": -2.2157764434814453,
645
+ "logits/rejected": -2.01257586479187,
646
+ "logps/chosen": -373.2973937988281,
647
+ "logps/rejected": -401.8234558105469,
648
+ "loss": 0.6029,
649
+ "rewards/accuracies": 0.550000011920929,
650
+ "rewards/chosen": -1.1020699739456177,
651
+ "rewards/margins": 0.6321985125541687,
652
+ "rewards/rejected": -1.7342685461044312,
653
+ "step": 450
654
+ },
655
+ {
656
+ "epoch": 0.37,
657
+ "learning_rate": 3.983547216509254e-06,
658
+ "logits/chosen": -2.215888261795044,
659
+ "logits/rejected": -2.1177175045013428,
660
+ "logps/chosen": -399.17547607421875,
661
+ "logps/rejected": -416.84783935546875,
662
+ "loss": 0.5846,
663
+ "rewards/accuracies": 0.625,
664
+ "rewards/chosen": -1.7063343524932861,
665
+ "rewards/margins": 0.4800395965576172,
666
+ "rewards/rejected": -2.186373472213745,
667
+ "step": 460
668
+ },
669
+ {
670
+ "epoch": 0.38,
671
+ "learning_rate": 3.92678391921108e-06,
672
+ "logits/chosen": -2.060939073562622,
673
+ "logits/rejected": -1.9635168313980103,
674
+ "logps/chosen": -346.0058898925781,
675
+ "logps/rejected": -369.8831481933594,
676
+ "loss": 0.6227,
677
+ "rewards/accuracies": 0.4749999940395355,
678
+ "rewards/chosen": -1.8079555034637451,
679
+ "rewards/margins": 0.4140779376029968,
680
+ "rewards/rejected": -2.2220332622528076,
681
+ "step": 470
682
+ },
683
+ {
684
+ "epoch": 0.38,
685
+ "learning_rate": 3.868908058731376e-06,
686
+ "logits/chosen": -1.974831223487854,
687
+ "logits/rejected": -1.8712928295135498,
688
+ "logps/chosen": -399.43365478515625,
689
+ "logps/rejected": -470.9222717285156,
690
+ "loss": 0.5688,
691
+ "rewards/accuracies": 0.699999988079071,
692
+ "rewards/chosen": -2.336700439453125,
693
+ "rewards/margins": 0.5076223611831665,
694
+ "rewards/rejected": -2.844322681427002,
695
+ "step": 480
696
+ },
697
+ {
698
+ "epoch": 0.39,
699
+ "learning_rate": 3.8099647649251984e-06,
700
+ "logits/chosen": -2.060171127319336,
701
+ "logits/rejected": -1.7853256464004517,
702
+ "logps/chosen": -443.10223388671875,
703
+ "logps/rejected": -441.1519470214844,
704
+ "loss": 0.56,
705
+ "rewards/accuracies": 0.574999988079071,
706
+ "rewards/chosen": -1.967047929763794,
707
+ "rewards/margins": 0.48686590790748596,
708
+ "rewards/rejected": -2.453913927078247,
709
+ "step": 490
710
+ },
711
+ {
712
+ "epoch": 0.4,
713
+ "learning_rate": 3.7500000000000005e-06,
714
+ "logits/chosen": -2.0951271057128906,
715
+ "logits/rejected": -2.0230164527893066,
716
+ "logps/chosen": -356.6705627441406,
717
+ "logps/rejected": -427.401611328125,
718
+ "loss": 0.5599,
719
+ "rewards/accuracies": 0.625,
720
+ "rewards/chosen": -1.6239030361175537,
721
+ "rewards/margins": 0.6958804726600647,
722
+ "rewards/rejected": -2.3197836875915527,
723
+ "step": 500
724
+ },
725
+ {
726
+ "epoch": 0.41,
727
+ "learning_rate": 3.689060522675689e-06,
728
+ "logits/chosen": -2.1919496059417725,
729
+ "logits/rejected": -2.014500141143799,
730
+ "logps/chosen": -428.25701904296875,
731
+ "logps/rejected": -452.41729736328125,
732
+ "loss": 0.557,
733
+ "rewards/accuracies": 0.699999988079071,
734
+ "rewards/chosen": -1.8657432794570923,
735
+ "rewards/margins": 0.5891915559768677,
736
+ "rewards/rejected": -2.454935073852539,
737
+ "step": 510
738
+ },
739
+ {
740
+ "epoch": 0.42,
741
+ "learning_rate": 3.627193851723577e-06,
742
+ "logits/chosen": -2.0486769676208496,
743
+ "logits/rejected": -1.9130027294158936,
744
+ "logps/chosen": -471.4190979003906,
745
+ "logps/rejected": -506.3304138183594,
746
+ "loss": 0.5804,
747
+ "rewards/accuracies": 0.550000011920929,
748
+ "rewards/chosen": -2.426065444946289,
749
+ "rewards/margins": 0.6026751399040222,
750
+ "rewards/rejected": -3.028740406036377,
751
+ "step": 520
752
+ },
753
+ {
754
+ "epoch": 0.42,
755
+ "learning_rate": 3.564448228912682e-06,
756
+ "logits/chosen": -2.078453540802002,
757
+ "logits/rejected": -2.038710117340088,
758
+ "logps/chosen": -485.046875,
759
+ "logps/rejected": -521.7474365234375,
760
+ "loss": 0.5933,
761
+ "rewards/accuracies": 0.5,
762
+ "rewards/chosen": -2.38494873046875,
763
+ "rewards/margins": 0.25780850648880005,
764
+ "rewards/rejected": -2.6427571773529053,
765
+ "step": 530
766
+ },
767
+ {
768
+ "epoch": 0.43,
769
+ "learning_rate": 3.5008725813922383e-06,
770
+ "logits/chosen": -2.164614200592041,
771
+ "logits/rejected": -1.8823134899139404,
772
+ "logps/chosen": -410.06988525390625,
773
+ "logps/rejected": -389.2319030761719,
774
+ "loss": 0.5627,
775
+ "rewards/accuracies": 0.550000011920929,
776
+ "rewards/chosen": -1.471362590789795,
777
+ "rewards/margins": 0.5164188146591187,
778
+ "rewards/rejected": -1.9877815246582031,
779
+ "step": 540
780
+ },
781
+ {
782
+ "epoch": 0.44,
783
+ "learning_rate": 3.436516483539781e-06,
784
+ "logits/chosen": -2.1140494346618652,
785
+ "logits/rejected": -1.9551842212677002,
786
+ "logps/chosen": -301.4757995605469,
787
+ "logps/rejected": -336.1775817871094,
788
+ "loss": 0.551,
789
+ "rewards/accuracies": 0.6499999761581421,
790
+ "rewards/chosen": -0.9967424273490906,
791
+ "rewards/margins": 0.5578289031982422,
792
+ "rewards/rejected": -1.554571270942688,
793
+ "step": 550
794
+ },
795
+ {
796
+ "epoch": 0.45,
797
+ "learning_rate": 3.3714301183045382e-06,
798
+ "logits/chosen": -2.2336459159851074,
799
+ "logits/rejected": -2.313737630844116,
800
+ "logps/chosen": -379.36846923828125,
801
+ "logps/rejected": -424.39483642578125,
802
+ "loss": 0.5247,
803
+ "rewards/accuracies": 0.675000011920929,
804
+ "rewards/chosen": -0.8149202466011047,
805
+ "rewards/margins": 0.6902198195457458,
806
+ "rewards/rejected": -1.5051400661468506,
807
+ "step": 560
808
+ },
809
+ {
810
+ "epoch": 0.46,
811
+ "learning_rate": 3.3056642380762783e-06,
812
+ "logits/chosen": -2.109157085418701,
813
+ "logits/rejected": -2.1128740310668945,
814
+ "logps/chosen": -370.62579345703125,
815
+ "logps/rejected": -388.05670166015625,
816
+ "loss": 0.5868,
817
+ "rewards/accuracies": 0.5,
818
+ "rewards/chosen": -1.1603527069091797,
819
+ "rewards/margins": 0.45087918639183044,
820
+ "rewards/rejected": -1.611232042312622,
821
+ "step": 570
822
+ },
823
+ {
824
+ "epoch": 0.46,
825
+ "learning_rate": 3.2392701251101172e-06,
826
+ "logits/chosen": -2.079946756362915,
827
+ "logits/rejected": -1.9534591436386108,
828
+ "logps/chosen": -384.3609313964844,
829
+ "logps/rejected": -391.6958312988281,
830
+ "loss": 0.5787,
831
+ "rewards/accuracies": 0.4749999940395355,
832
+ "rewards/chosen": -1.6679966449737549,
833
+ "rewards/margins": 0.5511924028396606,
834
+ "rewards/rejected": -2.219188928604126,
835
+ "step": 580
836
+ },
837
+ {
838
+ "epoch": 0.47,
839
+ "learning_rate": 3.1722995515381644e-06,
840
+ "logits/chosen": -1.974525809288025,
841
+ "logits/rejected": -1.8337838649749756,
842
+ "logps/chosen": -373.17828369140625,
843
+ "logps/rejected": -386.6225280761719,
844
+ "loss": 0.5649,
845
+ "rewards/accuracies": 0.4749999940395355,
846
+ "rewards/chosen": -1.2219445705413818,
847
+ "rewards/margins": 0.5704046487808228,
848
+ "rewards/rejected": -1.7923492193222046,
849
+ "step": 590
850
+ },
851
+ {
852
+ "epoch": 0.48,
853
+ "learning_rate": 3.1048047389991693e-06,
854
+ "logits/chosen": -2.0642426013946533,
855
+ "logits/rejected": -1.8965177536010742,
856
+ "logps/chosen": -323.34405517578125,
857
+ "logps/rejected": -380.21380615234375,
858
+ "loss": 0.5979,
859
+ "rewards/accuracies": 0.6499999761581421,
860
+ "rewards/chosen": -1.0704076290130615,
861
+ "rewards/margins": 0.48811253905296326,
862
+ "rewards/rejected": -1.5585200786590576,
863
+ "step": 600
864
+ },
865
+ {
866
+ "epoch": 0.49,
867
+ "learning_rate": 3.0368383179176584e-06,
868
+ "logits/chosen": -2.033987283706665,
869
+ "logits/rejected": -1.9596669673919678,
870
+ "logps/chosen": -305.8987731933594,
871
+ "logps/rejected": -335.6326904296875,
872
+ "loss": 0.5467,
873
+ "rewards/accuracies": 0.675000011920929,
874
+ "rewards/chosen": -0.6429412961006165,
875
+ "rewards/margins": 0.666652500629425,
876
+ "rewards/rejected": -1.309593677520752,
877
+ "step": 610
878
+ },
879
+ {
880
+ "epoch": 0.5,
881
+ "learning_rate": 2.9684532864643123e-06,
882
+ "logits/chosen": -2.0718424320220947,
883
+ "logits/rejected": -1.9492712020874023,
884
+ "logps/chosen": -354.63372802734375,
885
+ "logps/rejected": -367.26422119140625,
886
+ "loss": 0.5716,
887
+ "rewards/accuracies": 0.625,
888
+ "rewards/chosen": -1.022562026977539,
889
+ "rewards/margins": 0.4337197244167328,
890
+ "rewards/rejected": -1.4562817811965942,
891
+ "step": 620
892
+ },
893
+ {
894
+ "epoch": 0.5,
895
+ "learning_rate": 2.8997029692295875e-06,
896
+ "logits/chosen": -2.2117860317230225,
897
+ "logits/rejected": -2.0400938987731934,
898
+ "logps/chosen": -381.23028564453125,
899
+ "logps/rejected": -429.6178283691406,
900
+ "loss": 0.531,
901
+ "rewards/accuracies": 0.699999988079071,
902
+ "rewards/chosen": -1.1089333295822144,
903
+ "rewards/margins": 0.7748517990112305,
904
+ "rewards/rejected": -1.8837852478027344,
905
+ "step": 630
906
+ },
907
+ {
908
+ "epoch": 0.51,
909
+ "learning_rate": 2.8306409756428067e-06,
910
+ "logits/chosen": -2.042997121810913,
911
+ "logits/rejected": -1.6112009286880493,
912
+ "logps/chosen": -487.6568908691406,
913
+ "logps/rejected": -535.9304809570312,
914
+ "loss": 0.516,
915
+ "rewards/accuracies": 0.800000011920929,
916
+ "rewards/chosen": -1.8848283290863037,
917
+ "rewards/margins": 1.166489601135254,
918
+ "rewards/rejected": -3.0513179302215576,
919
+ "step": 640
920
+ },
921
+ {
922
+ "epoch": 0.52,
923
+ "learning_rate": 2.761321158169134e-06,
924
+ "logits/chosen": -2.1288044452667236,
925
+ "logits/rejected": -1.8480589389801025,
926
+ "logps/chosen": -591.8642578125,
927
+ "logps/rejected": -573.7616577148438,
928
+ "loss": 0.5375,
929
+ "rewards/accuracies": 0.574999988079071,
930
+ "rewards/chosen": -3.170348644256592,
931
+ "rewards/margins": 0.48750972747802734,
932
+ "rewards/rejected": -3.6578586101531982,
933
+ "step": 650
934
+ },
935
+ {
936
+ "epoch": 0.53,
937
+ "learning_rate": 2.6917975703170466e-06,
938
+ "logits/chosen": -1.834415078163147,
939
+ "logits/rejected": -1.542644739151001,
940
+ "logps/chosen": -506.2686462402344,
941
+ "logps/rejected": -612.8573608398438,
942
+ "loss": 0.5693,
943
+ "rewards/accuracies": 0.7250000238418579,
944
+ "rewards/chosen": -2.562810182571411,
945
+ "rewards/margins": 0.9861109852790833,
946
+ "rewards/rejected": -3.5489211082458496,
947
+ "step": 660
948
+ },
949
+ {
950
+ "epoch": 0.54,
951
+ "learning_rate": 2.6221244244890336e-06,
952
+ "logits/chosen": -1.8896293640136719,
953
+ "logits/rejected": -1.7028945684432983,
954
+ "logps/chosen": -479.379638671875,
955
+ "logps/rejected": -513.0144653320312,
956
+ "loss": 0.5561,
957
+ "rewards/accuracies": 0.699999988079071,
958
+ "rewards/chosen": -2.36388897895813,
959
+ "rewards/margins": 0.6828749775886536,
960
+ "rewards/rejected": -3.0467634201049805,
961
+ "step": 670
962
+ },
963
+ {
964
+ "epoch": 0.54,
965
+ "learning_rate": 2.5523560497083927e-06,
966
+ "logits/chosen": -2.059697151184082,
967
+ "logits/rejected": -1.8784849643707275,
968
+ "logps/chosen": -337.88494873046875,
969
+ "logps/rejected": -363.4068298339844,
970
+ "loss": 0.6096,
971
+ "rewards/accuracies": 0.5249999761581421,
972
+ "rewards/chosen": -1.4112889766693115,
973
+ "rewards/margins": 0.4129563868045807,
974
+ "rewards/rejected": -1.8242454528808594,
975
+ "step": 680
976
+ },
977
+ {
978
+ "epoch": 0.55,
979
+ "learning_rate": 2.482546849255096e-06,
980
+ "logits/chosen": -1.9830563068389893,
981
+ "logits/rejected": -1.6811624765396118,
982
+ "logps/chosen": -361.8495788574219,
983
+ "logps/rejected": -382.3643493652344,
984
+ "loss": 0.5242,
985
+ "rewards/accuracies": 0.6000000238418579,
986
+ "rewards/chosen": -1.473449468612671,
987
+ "rewards/margins": 0.5172423720359802,
988
+ "rewards/rejected": -1.990692138671875,
989
+ "step": 690
990
+ },
991
+ {
992
+ "epoch": 0.56,
993
+ "learning_rate": 2.4127512582437486e-06,
994
+ "logits/chosen": -2.0440447330474854,
995
+ "logits/rejected": -1.9845294952392578,
996
+ "logps/chosen": -401.8277893066406,
997
+ "logps/rejected": -451.3299255371094,
998
+ "loss": 0.5288,
999
+ "rewards/accuracies": 0.7250000238418579,
1000
+ "rewards/chosen": -1.1858041286468506,
1001
+ "rewards/margins": 0.7083054184913635,
1002
+ "rewards/rejected": -1.8941094875335693,
1003
+ "step": 700
1004
+ },
1005
+ {
1006
+ "epoch": 0.57,
1007
+ "learning_rate": 2.3430237011767166e-06,
1008
+ "logits/chosen": -1.833554983139038,
1009
+ "logits/rejected": -1.6954238414764404,
1010
+ "logps/chosen": -353.65924072265625,
1011
+ "logps/rejected": -397.3764343261719,
1012
+ "loss": 0.531,
1013
+ "rewards/accuracies": 0.7250000238418579,
1014
+ "rewards/chosen": -1.3097889423370361,
1015
+ "rewards/margins": 0.7774924039840698,
1016
+ "rewards/rejected": -2.0872814655303955,
1017
+ "step": 710
1018
+ },
1019
+ {
1020
+ "epoch": 0.58,
1021
+ "learning_rate": 2.2734185495055503e-06,
1022
+ "logits/chosen": -1.6967763900756836,
1023
+ "logits/rejected": -1.4636828899383545,
1024
+ "logps/chosen": -397.916748046875,
1025
+ "logps/rejected": -437.44073486328125,
1026
+ "loss": 0.4871,
1027
+ "rewards/accuracies": 0.7749999761581421,
1028
+ "rewards/chosen": -1.731325387954712,
1029
+ "rewards/margins": 0.9450262188911438,
1030
+ "rewards/rejected": -2.676351547241211,
1031
+ "step": 720
1032
+ },
1033
+ {
1034
+ "epoch": 0.58,
1035
+ "learning_rate": 2.2039900792337477e-06,
1036
+ "logits/chosen": -1.6748664379119873,
1037
+ "logits/rejected": -1.4754512310028076,
1038
+ "logps/chosen": -475.0547790527344,
1039
+ "logps/rejected": -515.4742431640625,
1040
+ "loss": 0.5481,
1041
+ "rewards/accuracies": 0.6499999761581421,
1042
+ "rewards/chosen": -2.1851890087127686,
1043
+ "rewards/margins": 0.8229333758354187,
1044
+ "rewards/rejected": -3.008122205734253,
1045
+ "step": 730
1046
+ },
1047
+ {
1048
+ "epoch": 0.59,
1049
+ "learning_rate": 2.134792428593971e-06,
1050
+ "logits/chosen": -1.8436079025268555,
1051
+ "logits/rejected": -1.525818109512329,
1052
+ "logps/chosen": -420.80810546875,
1053
+ "logps/rejected": -455.9149475097656,
1054
+ "loss": 0.5215,
1055
+ "rewards/accuracies": 0.5249999761581421,
1056
+ "rewards/chosen": -1.747849702835083,
1057
+ "rewards/margins": 0.8033768534660339,
1058
+ "rewards/rejected": -2.5512263774871826,
1059
+ "step": 740
1060
+ },
1061
+ {
1062
+ "epoch": 0.6,
1063
+ "learning_rate": 2.0658795558326745e-06,
1064
+ "logits/chosen": -1.76202392578125,
1065
+ "logits/rejected": -1.825585126876831,
1066
+ "logps/chosen": -335.0538635253906,
1067
+ "logps/rejected": -410.6778259277344,
1068
+ "loss": 0.5755,
1069
+ "rewards/accuracies": 0.6499999761581421,
1070
+ "rewards/chosen": -1.3501343727111816,
1071
+ "rewards/margins": 0.7440234422683716,
1072
+ "rewards/rejected": -2.0941579341888428,
1073
+ "step": 750
1074
+ },
1075
+ {
1076
+ "epoch": 0.61,
1077
+ "learning_rate": 1.997305197135089e-06,
1078
+ "logits/chosen": -1.8911097049713135,
1079
+ "logits/rejected": -1.7050600051879883,
1080
+ "logps/chosen": -344.654296875,
1081
+ "logps/rejected": -365.8127136230469,
1082
+ "loss": 0.5896,
1083
+ "rewards/accuracies": 0.574999988079071,
1084
+ "rewards/chosen": -1.335740089416504,
1085
+ "rewards/margins": 0.36764174699783325,
1086
+ "rewards/rejected": -1.7033817768096924,
1087
+ "step": 760
1088
+ },
1089
+ {
1090
+ "epoch": 0.62,
1091
+ "learning_rate": 1.9291228247233607e-06,
1092
+ "logits/chosen": -2.199026107788086,
1093
+ "logits/rejected": -1.8742659091949463,
1094
+ "logps/chosen": -418.6163024902344,
1095
+ "logps/rejected": -376.9792175292969,
1096
+ "loss": 0.5747,
1097
+ "rewards/accuracies": 0.574999988079071,
1098
+ "rewards/chosen": -1.2824668884277344,
1099
+ "rewards/margins": 0.4385095536708832,
1100
+ "rewards/rejected": -1.72097647190094,
1101
+ "step": 770
1102
+ },
1103
+ {
1104
+ "epoch": 0.62,
1105
+ "learning_rate": 1.8613856051605242e-06,
1106
+ "logits/chosen": -1.913412094116211,
1107
+ "logits/rejected": -1.6813814640045166,
1108
+ "logps/chosen": -404.4813232421875,
1109
+ "logps/rejected": -463.2132263183594,
1110
+ "loss": 0.5343,
1111
+ "rewards/accuracies": 0.75,
1112
+ "rewards/chosen": -1.447469711303711,
1113
+ "rewards/margins": 0.8074652552604675,
1114
+ "rewards/rejected": -2.2549350261688232,
1115
+ "step": 780
1116
+ },
1117
+ {
1118
+ "epoch": 0.63,
1119
+ "learning_rate": 1.7941463578928088e-06,
1120
+ "logits/chosen": -2.0417885780334473,
1121
+ "logits/rejected": -2.0130181312561035,
1122
+ "logps/chosen": -379.37066650390625,
1123
+ "logps/rejected": -394.95867919921875,
1124
+ "loss": 0.6045,
1125
+ "rewards/accuracies": 0.550000011920929,
1126
+ "rewards/chosen": -1.6197001934051514,
1127
+ "rewards/margins": 0.29476988315582275,
1128
+ "rewards/rejected": -1.9144700765609741,
1129
+ "step": 790
1130
+ },
1131
+ {
1132
+ "epoch": 0.64,
1133
+ "learning_rate": 1.7274575140626318e-06,
1134
+ "logits/chosen": -1.9285320043563843,
1135
+ "logits/rejected": -2.0787336826324463,
1136
+ "logps/chosen": -348.0567321777344,
1137
+ "logps/rejected": -397.8705139160156,
1138
+ "loss": 0.522,
1139
+ "rewards/accuracies": 0.6000000238418579,
1140
+ "rewards/chosen": -1.0203365087509155,
1141
+ "rewards/margins": 0.4765065312385559,
1142
+ "rewards/rejected": -1.4968432188034058,
1143
+ "step": 800
1144
+ },
1145
+ {
1146
+ "epoch": 0.65,
1147
+ "learning_rate": 1.661371075624363e-06,
1148
+ "logits/chosen": -1.9909842014312744,
1149
+ "logits/rejected": -1.6936451196670532,
1150
+ "logps/chosen": -265.6180419921875,
1151
+ "logps/rejected": -281.07952880859375,
1152
+ "loss": 0.5591,
1153
+ "rewards/accuracies": 0.6000000238418579,
1154
+ "rewards/chosen": -0.7331510782241821,
1155
+ "rewards/margins": 0.5583804249763489,
1156
+ "rewards/rejected": -1.2915314435958862,
1157
+ "step": 810
1158
+ },
1159
+ {
1160
+ "epoch": 0.66,
1161
+ "learning_rate": 1.5959385747947697e-06,
1162
+ "logits/chosen": -1.9582141637802124,
1163
+ "logits/rejected": -1.9302053451538086,
1164
+ "logps/chosen": -336.01959228515625,
1165
+ "logps/rejected": -410.764404296875,
1166
+ "loss": 0.4964,
1167
+ "rewards/accuracies": 0.675000011920929,
1168
+ "rewards/chosen": -0.9309733510017395,
1169
+ "rewards/margins": 0.613142728805542,
1170
+ "rewards/rejected": -1.5441160202026367,
1171
+ "step": 820
1172
+ },
1173
+ {
1174
+ "epoch": 0.66,
1175
+ "learning_rate": 1.5312110338697427e-06,
1176
+ "logits/chosen": -2.052699327468872,
1177
+ "logits/rejected": -1.9583276510238647,
1178
+ "logps/chosen": -289.5865478515625,
1179
+ "logps/rejected": -312.62725830078125,
1180
+ "loss": 0.5951,
1181
+ "rewards/accuracies": 0.5,
1182
+ "rewards/chosen": -0.9973894357681274,
1183
+ "rewards/margins": 0.37033605575561523,
1184
+ "rewards/rejected": -1.3677256107330322,
1185
+ "step": 830
1186
+ },
1187
+ {
1188
+ "epoch": 0.67,
1189
+ "learning_rate": 1.467238925438646e-06,
1190
+ "logits/chosen": -2.0705270767211914,
1191
+ "logits/rejected": -1.8725826740264893,
1192
+ "logps/chosen": -367.77471923828125,
1193
+ "logps/rejected": -411.2198181152344,
1194
+ "loss": 0.5386,
1195
+ "rewards/accuracies": 0.675000011920929,
1196
+ "rewards/chosen": -0.9991909861564636,
1197
+ "rewards/margins": 0.571709394454956,
1198
+ "rewards/rejected": -1.5709004402160645,
1199
+ "step": 840
1200
+ },
1201
+ {
1202
+ "epoch": 0.68,
1203
+ "learning_rate": 1.4040721330273063e-06,
1204
+ "logits/chosen": -1.7584121227264404,
1205
+ "logits/rejected": -1.4766103029251099,
1206
+ "logps/chosen": -463.2286071777344,
1207
+ "logps/rejected": -484.1163635253906,
1208
+ "loss": 0.5445,
1209
+ "rewards/accuracies": 0.699999988079071,
1210
+ "rewards/chosen": -1.6703122854232788,
1211
+ "rewards/margins": 0.6411558389663696,
1212
+ "rewards/rejected": -2.3114681243896484,
1213
+ "step": 850
1214
+ },
1215
+ {
1216
+ "epoch": 0.69,
1217
+ "learning_rate": 1.3417599122003464e-06,
1218
+ "logits/chosen": -2.1024575233459473,
1219
+ "logits/rejected": -1.6672672033309937,
1220
+ "logps/chosen": -391.9578857421875,
1221
+ "logps/rejected": -422.2921447753906,
1222
+ "loss": 0.4638,
1223
+ "rewards/accuracies": 0.6499999761581421,
1224
+ "rewards/chosen": -1.2933290004730225,
1225
+ "rewards/margins": 0.9474877119064331,
1226
+ "rewards/rejected": -2.240816831588745,
1227
+ "step": 860
1228
+ },
1229
+ {
1230
+ "epoch": 0.7,
1231
+ "learning_rate": 1.280350852153168e-06,
1232
+ "logits/chosen": -1.8131592273712158,
1233
+ "logits/rejected": -1.9417848587036133,
1234
+ "logps/chosen": -319.12799072265625,
1235
+ "logps/rejected": -448.4112854003906,
1236
+ "loss": 0.5363,
1237
+ "rewards/accuracies": 0.625,
1238
+ "rewards/chosen": -1.285538911819458,
1239
+ "rewards/margins": 0.8607152104377747,
1240
+ "rewards/rejected": -2.146254062652588,
1241
+ "step": 870
1242
+ },
1243
+ {
1244
+ "epoch": 0.7,
1245
+ "learning_rate": 1.2198928378235717e-06,
1246
+ "logits/chosen": -1.8834339380264282,
1247
+ "logits/rejected": -1.7587229013442993,
1248
+ "logps/chosen": -424.78826904296875,
1249
+ "logps/rejected": -522.1199340820312,
1250
+ "loss": 0.5754,
1251
+ "rewards/accuracies": 0.675000011920929,
1252
+ "rewards/chosen": -1.7636245489120483,
1253
+ "rewards/margins": 0.8004873991012573,
1254
+ "rewards/rejected": -2.5641119480133057,
1255
+ "step": 880
1256
+ },
1257
+ {
1258
+ "epoch": 0.71,
1259
+ "learning_rate": 1.160433012552508e-06,
1260
+ "logits/chosen": -2.0634958744049072,
1261
+ "logits/rejected": -2.003753423690796,
1262
+ "logps/chosen": -335.7600402832031,
1263
+ "logps/rejected": -375.54852294921875,
1264
+ "loss": 0.5125,
1265
+ "rewards/accuracies": 0.675000011920929,
1266
+ "rewards/chosen": -0.9807103872299194,
1267
+ "rewards/margins": 0.9115110635757446,
1268
+ "rewards/rejected": -1.892221450805664,
1269
+ "step": 890
1270
+ },
1271
+ {
1272
+ "epoch": 0.72,
1273
+ "learning_rate": 1.1020177413231334e-06,
1274
+ "logits/chosen": -2.016789197921753,
1275
+ "logits/rejected": -1.8195407390594482,
1276
+ "logps/chosen": -353.8262939453125,
1277
+ "logps/rejected": -411.1983947753906,
1278
+ "loss": 0.5023,
1279
+ "rewards/accuracies": 0.675000011920929,
1280
+ "rewards/chosen": -1.2073825597763062,
1281
+ "rewards/margins": 0.8694759607315063,
1282
+ "rewards/rejected": -2.0768585205078125,
1283
+ "step": 900
1284
+ },
1285
+ {
1286
+ "epoch": 0.73,
1287
+ "learning_rate": 1.0446925746067768e-06,
1288
+ "logits/chosen": -1.8060710430145264,
1289
+ "logits/rejected": -1.5303637981414795,
1290
+ "logps/chosen": -329.4360046386719,
1291
+ "logps/rejected": -366.3887939453125,
1292
+ "loss": 0.5347,
1293
+ "rewards/accuracies": 0.625,
1294
+ "rewards/chosen": -1.0586979389190674,
1295
+ "rewards/margins": 0.9634588956832886,
1296
+ "rewards/rejected": -2.0221567153930664,
1297
+ "step": 910
1298
+ },
1299
+ {
1300
+ "epoch": 0.74,
1301
+ "learning_rate": 9.88502212844063e-07,
1302
+ "logits/chosen": -1.9487718343734741,
1303
+ "logits/rejected": -1.8278251886367798,
1304
+ "logps/chosen": -424.44256591796875,
1305
+ "logps/rejected": -527.6069946289062,
1306
+ "loss": 0.5597,
1307
+ "rewards/accuracies": 0.675000011920929,
1308
+ "rewards/chosen": -1.5092089176177979,
1309
+ "rewards/margins": 1.0597232580184937,
1310
+ "rewards/rejected": -2.568932056427002,
1311
+ "step": 920
1312
+ },
1313
+ {
1314
+ "epoch": 0.74,
1315
+ "learning_rate": 9.334904715888496e-07,
1316
+ "logits/chosen": -1.9447358846664429,
1317
+ "logits/rejected": -1.672228455543518,
1318
+ "logps/chosen": -440.62774658203125,
1319
+ "logps/rejected": -548.8284912109375,
1320
+ "loss": 0.5634,
1321
+ "rewards/accuracies": 0.699999988079071,
1322
+ "rewards/chosen": -1.6289888620376587,
1323
+ "rewards/margins": 1.100258469581604,
1324
+ "rewards/rejected": -2.7292473316192627,
1325
+ "step": 930
1326
+ },
1327
+ {
1328
+ "epoch": 0.75,
1329
+ "learning_rate": 8.797002473421729e-07,
1330
+ "logits/chosen": -1.8913949728012085,
1331
+ "logits/rejected": -1.8936278820037842,
1332
+ "logps/chosen": -384.93609619140625,
1333
+ "logps/rejected": -455.0386657714844,
1334
+ "loss": 0.5326,
1335
+ "rewards/accuracies": 0.6000000238418579,
1336
+ "rewards/chosen": -1.3333874940872192,
1337
+ "rewards/margins": 0.7129821181297302,
1338
+ "rewards/rejected": -2.046369791030884,
1339
+ "step": 940
1340
+ },
1341
+ {
1342
+ "epoch": 0.76,
1343
+ "learning_rate": 8.271734841028553e-07,
1344
+ "logits/chosen": -1.8880865573883057,
1345
+ "logits/rejected": -1.6740175485610962,
1346
+ "logps/chosen": -414.0435485839844,
1347
+ "logps/rejected": -461.56158447265625,
1348
+ "loss": 0.4911,
1349
+ "rewards/accuracies": 0.6499999761581421,
1350
+ "rewards/chosen": -1.596765398979187,
1351
+ "rewards/margins": 0.815388560295105,
1352
+ "rewards/rejected": -2.412154197692871,
1353
+ "step": 950
1354
+ },
1355
+ {
1356
+ "epoch": 0.77,
1357
+ "learning_rate": 7.759511406608255e-07,
1358
+ "logits/chosen": -2.010457754135132,
1359
+ "logits/rejected": -1.8959916830062866,
1360
+ "logps/chosen": -383.6888122558594,
1361
+ "logps/rejected": -452.77105712890625,
1362
+ "loss": 0.5708,
1363
+ "rewards/accuracies": 0.7250000238418579,
1364
+ "rewards/chosen": -1.487220048904419,
1365
+ "rewards/margins": 0.7800397276878357,
1366
+ "rewards/rejected": -2.2672595977783203,
1367
+ "step": 960
1368
+ },
1369
+ {
1370
+ "epoch": 0.78,
1371
+ "learning_rate": 7.260731586586983e-07,
1372
+ "logits/chosen": -1.7655563354492188,
1373
+ "logits/rejected": -1.8372739553451538,
1374
+ "logps/chosen": -259.60394287109375,
1375
+ "logps/rejected": -331.3665466308594,
1376
+ "loss": 0.541,
1377
+ "rewards/accuracies": 0.574999988079071,
1378
+ "rewards/chosen": -1.3245407342910767,
1379
+ "rewards/margins": 0.7111746072769165,
1380
+ "rewards/rejected": -2.0357155799865723,
1381
+ "step": 970
1382
+ },
1383
+ {
1384
+ "epoch": 0.78,
1385
+ "learning_rate": 6.775784314464717e-07,
1386
+ "logits/chosen": -2.1715354919433594,
1387
+ "logits/rejected": -1.837615728378296,
1388
+ "logps/chosen": -360.0257873535156,
1389
+ "logps/rejected": -413.18902587890625,
1390
+ "loss": 0.4932,
1391
+ "rewards/accuracies": 0.675000011920929,
1392
+ "rewards/chosen": -1.2116737365722656,
1393
+ "rewards/margins": 1.0079423189163208,
1394
+ "rewards/rejected": -2.219616413116455,
1395
+ "step": 980
1396
+ },
1397
+ {
1398
+ "epoch": 0.79,
1399
+ "learning_rate": 6.305047737536707e-07,
1400
+ "logits/chosen": -1.8406097888946533,
1401
+ "logits/rejected": -1.8857877254486084,
1402
+ "logps/chosen": -375.0958557128906,
1403
+ "logps/rejected": -486.08197021484375,
1404
+ "loss": 0.5066,
1405
+ "rewards/accuracies": 0.699999988079071,
1406
+ "rewards/chosen": -1.7462670803070068,
1407
+ "rewards/margins": 0.950217604637146,
1408
+ "rewards/rejected": -2.6964848041534424,
1409
+ "step": 990
1410
+ },
1411
+ {
1412
+ "epoch": 0.8,
1413
+ "learning_rate": 5.848888922025553e-07,
1414
+ "logits/chosen": -1.8275880813598633,
1415
+ "logits/rejected": -2.047853946685791,
1416
+ "logps/chosen": -441.58966064453125,
1417
+ "logps/rejected": -461.3450622558594,
1418
+ "loss": 0.5374,
1419
+ "rewards/accuracies": 0.625,
1420
+ "rewards/chosen": -1.767677664756775,
1421
+ "rewards/margins": 0.5157124996185303,
1422
+ "rewards/rejected": -2.283390522003174,
1423
+ "step": 1000
1424
+ },
1425
+ {
1426
+ "epoch": 0.81,
1427
+ "learning_rate": 5.407663566854008e-07,
1428
+ "logits/chosen": -2.1351914405822754,
1429
+ "logits/rejected": -1.8889188766479492,
1430
+ "logps/chosen": -394.17120361328125,
1431
+ "logps/rejected": -436.7420959472656,
1432
+ "loss": 0.5683,
1433
+ "rewards/accuracies": 0.699999988079071,
1434
+ "rewards/chosen": -1.3438371419906616,
1435
+ "rewards/margins": 0.9155893325805664,
1436
+ "rewards/rejected": -2.2594265937805176,
1437
+ "step": 1010
1438
+ },
1439
+ {
1440
+ "epoch": 0.82,
1441
+ "learning_rate": 4.981715726281666e-07,
1442
+ "logits/chosen": -2.017271041870117,
1443
+ "logits/rejected": -1.981236219406128,
1444
+ "logps/chosen": -441.42254638671875,
1445
+ "logps/rejected": -580.200927734375,
1446
+ "loss": 0.4928,
1447
+ "rewards/accuracies": 0.800000011920929,
1448
+ "rewards/chosen": -1.5878021717071533,
1449
+ "rewards/margins": 1.3070558309555054,
1450
+ "rewards/rejected": -2.8948581218719482,
1451
+ "step": 1020
1452
+ },
1453
+ {
1454
+ "epoch": 0.82,
1455
+ "learning_rate": 4.5713775416217884e-07,
1456
+ "logits/chosen": -1.6851351261138916,
1457
+ "logits/rejected": -1.7757434844970703,
1458
+ "logps/chosen": -348.5665588378906,
1459
+ "logps/rejected": -457.87542724609375,
1460
+ "loss": 0.5279,
1461
+ "rewards/accuracies": 0.675000011920929,
1462
+ "rewards/chosen": -1.5111078023910522,
1463
+ "rewards/margins": 1.0286327600479126,
1464
+ "rewards/rejected": -2.5397403240203857,
1465
+ "step": 1030
1466
+ },
1467
+ {
1468
+ "epoch": 0.83,
1469
+ "learning_rate": 4.1769689822475147e-07,
1470
+ "logits/chosen": -1.674856424331665,
1471
+ "logits/rejected": -1.817155122756958,
1472
+ "logps/chosen": -318.9717712402344,
1473
+ "logps/rejected": -434.81475830078125,
1474
+ "loss": 0.5817,
1475
+ "rewards/accuracies": 0.574999988079071,
1476
+ "rewards/chosen": -1.6043727397918701,
1477
+ "rewards/margins": 0.6249055862426758,
1478
+ "rewards/rejected": -2.229278087615967,
1479
+ "step": 1040
1480
+ },
1481
+ {
1482
+ "epoch": 0.84,
1483
+ "learning_rate": 3.798797596089351e-07,
1484
+ "logits/chosen": -1.953508973121643,
1485
+ "logits/rejected": -1.7095199823379517,
1486
+ "logps/chosen": -383.12371826171875,
1487
+ "logps/rejected": -451.828125,
1488
+ "loss": 0.5006,
1489
+ "rewards/accuracies": 0.75,
1490
+ "rewards/chosen": -1.5027226209640503,
1491
+ "rewards/margins": 0.9227194786071777,
1492
+ "rewards/rejected": -2.4254419803619385,
1493
+ "step": 1050
1494
+ },
1495
+ {
1496
+ "epoch": 0.85,
1497
+ "learning_rate": 3.4371582698185636e-07,
1498
+ "logits/chosen": -2.011037826538086,
1499
+ "logits/rejected": -1.858045220375061,
1500
+ "logps/chosen": -455.5492248535156,
1501
+ "logps/rejected": -558.2069091796875,
1502
+ "loss": 0.4797,
1503
+ "rewards/accuracies": 0.699999988079071,
1504
+ "rewards/chosen": -1.7510541677474976,
1505
+ "rewards/margins": 0.8790451884269714,
1506
+ "rewards/rejected": -2.6300997734069824,
1507
+ "step": 1060
1508
+ },
1509
+ {
1510
+ "epoch": 0.86,
1511
+ "learning_rate": 3.092332998903416e-07,
1512
+ "logits/chosen": -1.9804046154022217,
1513
+ "logits/rejected": -2.04115891456604,
1514
+ "logps/chosen": -395.303466796875,
1515
+ "logps/rejected": -492.44329833984375,
1516
+ "loss": 0.5034,
1517
+ "rewards/accuracies": 0.675000011920929,
1518
+ "rewards/chosen": -1.4553591012954712,
1519
+ "rewards/margins": 0.8879169225692749,
1520
+ "rewards/rejected": -2.343275785446167,
1521
+ "step": 1070
1522
+ },
1523
+ {
1524
+ "epoch": 0.86,
1525
+ "learning_rate": 2.764590667717562e-07,
1526
+ "logits/chosen": -1.8032455444335938,
1527
+ "logits/rejected": -1.563175916671753,
1528
+ "logps/chosen": -345.4132385253906,
1529
+ "logps/rejected": -335.91949462890625,
1530
+ "loss": 0.5746,
1531
+ "rewards/accuracies": 0.44999998807907104,
1532
+ "rewards/chosen": -1.697852373123169,
1533
+ "rewards/margins": 0.29076820611953735,
1534
+ "rewards/rejected": -1.988620400428772,
1535
+ "step": 1080
1536
+ },
1537
+ {
1538
+ "epoch": 0.87,
1539
+ "learning_rate": 2.454186839872158e-07,
1540
+ "logits/chosen": -1.882430076599121,
1541
+ "logits/rejected": -1.7509571313858032,
1542
+ "logps/chosen": -354.9794616699219,
1543
+ "logps/rejected": -450.55560302734375,
1544
+ "loss": 0.4963,
1545
+ "rewards/accuracies": 0.7749999761581421,
1546
+ "rewards/chosen": -1.5439207553863525,
1547
+ "rewards/margins": 1.0496200323104858,
1548
+ "rewards/rejected": -2.5935404300689697,
1549
+ "step": 1090
1550
+ },
1551
+ {
1552
+ "epoch": 0.88,
1553
+ "learning_rate": 2.1613635589349756e-07,
1554
+ "logits/chosen": -1.9673278331756592,
1555
+ "logits/rejected": -1.735865592956543,
1556
+ "logps/chosen": -361.7403869628906,
1557
+ "logps/rejected": -393.9466247558594,
1558
+ "loss": 0.5859,
1559
+ "rewards/accuracies": 0.550000011920929,
1560
+ "rewards/chosen": -1.3523650169372559,
1561
+ "rewards/margins": 0.5490896105766296,
1562
+ "rewards/rejected": -1.9014545679092407,
1563
+ "step": 1100
1564
+ },
1565
+ {
1566
+ "epoch": 0.89,
1567
+ "learning_rate": 1.8863491596921745e-07,
1568
+ "logits/chosen": -1.9650977849960327,
1569
+ "logits/rejected": -1.674034833908081,
1570
+ "logps/chosen": -398.31817626953125,
1571
+ "logps/rejected": -418.9993591308594,
1572
+ "loss": 0.4818,
1573
+ "rewards/accuracies": 0.6499999761581421,
1574
+ "rewards/chosen": -1.6353886127471924,
1575
+ "rewards/margins": 0.8128952980041504,
1576
+ "rewards/rejected": -2.4482839107513428,
1577
+ "step": 1110
1578
+ },
1579
+ {
1580
+ "epoch": 0.9,
1581
+ "learning_rate": 1.629358090099639e-07,
1582
+ "logits/chosen": -1.8651268482208252,
1583
+ "logits/rejected": -1.7213342189788818,
1584
+ "logps/chosen": -395.92608642578125,
1585
+ "logps/rejected": -427.53082275390625,
1586
+ "loss": 0.5434,
1587
+ "rewards/accuracies": 0.699999988079071,
1588
+ "rewards/chosen": -1.8581531047821045,
1589
+ "rewards/margins": 0.6955922842025757,
1590
+ "rewards/rejected": -2.5537455081939697,
1591
+ "step": 1120
1592
+ },
1593
+ {
1594
+ "epoch": 0.9,
1595
+ "learning_rate": 1.3905907440629752e-07,
1596
+ "logits/chosen": -1.8218196630477905,
1597
+ "logits/rejected": -1.603205680847168,
1598
+ "logps/chosen": -371.84820556640625,
1599
+ "logps/rejected": -423.8062438964844,
1600
+ "loss": 0.5957,
1601
+ "rewards/accuracies": 0.574999988079071,
1602
+ "rewards/chosen": -1.6403701305389404,
1603
+ "rewards/margins": 0.8258684873580933,
1604
+ "rewards/rejected": -2.466238498687744,
1605
+ "step": 1130
1606
+ },
1607
+ {
1608
+ "epoch": 0.91,
1609
+ "learning_rate": 1.1702333051763271e-07,
1610
+ "logits/chosen": -1.7774673700332642,
1611
+ "logits/rejected": -1.5289005041122437,
1612
+ "logps/chosen": -440.49407958984375,
1613
+ "logps/rejected": -493.42486572265625,
1614
+ "loss": 0.5159,
1615
+ "rewards/accuracies": 0.6499999761581421,
1616
+ "rewards/chosen": -1.8031593561172485,
1617
+ "rewards/margins": 0.9965063333511353,
1618
+ "rewards/rejected": -2.799665927886963,
1619
+ "step": 1140
1620
+ },
1621
+ {
1622
+ "epoch": 0.92,
1623
+ "learning_rate": 9.684576015420277e-08,
1624
+ "logits/chosen": -1.7342450618743896,
1625
+ "logits/rejected": -1.5554310083389282,
1626
+ "logps/chosen": -407.7359313964844,
1627
+ "logps/rejected": -490.3565979003906,
1628
+ "loss": 0.4897,
1629
+ "rewards/accuracies": 0.550000011920929,
1630
+ "rewards/chosen": -1.6332868337631226,
1631
+ "rewards/margins": 0.9035712480545044,
1632
+ "rewards/rejected": -2.536858081817627,
1633
+ "step": 1150
1634
+ },
1635
+ {
1636
+ "epoch": 0.93,
1637
+ "learning_rate": 7.854209717842231e-08,
1638
+ "logits/chosen": -1.728171944618225,
1639
+ "logits/rejected": -1.6524664163589478,
1640
+ "logps/chosen": -420.43768310546875,
1641
+ "logps/rejected": -411.3157653808594,
1642
+ "loss": 0.6007,
1643
+ "rewards/accuracies": 0.4749999940395355,
1644
+ "rewards/chosen": -1.8913564682006836,
1645
+ "rewards/margins": 0.21912887692451477,
1646
+ "rewards/rejected": -2.110485553741455,
1647
+ "step": 1160
1648
+ },
1649
+ {
1650
+ "epoch": 0.94,
1651
+ "learning_rate": 6.212661423609184e-08,
1652
+ "logits/chosen": -1.7257133722305298,
1653
+ "logits/rejected": -1.6627498865127563,
1654
+ "logps/chosen": -435.08599853515625,
1655
+ "logps/rejected": -524.0604248046875,
1656
+ "loss": 0.6031,
1657
+ "rewards/accuracies": 0.625,
1658
+ "rewards/chosen": -2.2094435691833496,
1659
+ "rewards/margins": 0.7494264841079712,
1660
+ "rewards/rejected": -2.9588701725006104,
1661
+ "step": 1170
1662
+ },
1663
+ {
1664
+ "epoch": 0.94,
1665
+ "learning_rate": 4.761211162702117e-08,
1666
+ "logits/chosen": -1.8422664403915405,
1667
+ "logits/rejected": -1.7962735891342163,
1668
+ "logps/chosen": -370.86566162109375,
1669
+ "logps/rejected": -408.62042236328125,
1670
+ "loss": 0.5591,
1671
+ "rewards/accuracies": 0.6000000238418579,
1672
+ "rewards/chosen": -1.4711788892745972,
1673
+ "rewards/margins": 0.5541782379150391,
1674
+ "rewards/rejected": -2.025357246398926,
1675
+ "step": 1180
1676
+ },
1677
+ {
1678
+ "epoch": 0.95,
1679
+ "learning_rate": 3.5009907323737826e-08,
1680
+ "logits/chosen": -1.8273169994354248,
1681
+ "logits/rejected": -1.549336314201355,
1682
+ "logps/chosen": -408.63140869140625,
1683
+ "logps/rejected": -491.55303955078125,
1684
+ "loss": 0.5183,
1685
+ "rewards/accuracies": 0.7250000238418579,
1686
+ "rewards/chosen": -1.7581342458724976,
1687
+ "rewards/margins": 0.9740850329399109,
1688
+ "rewards/rejected": -2.7322194576263428,
1689
+ "step": 1190
1690
+ },
1691
+ {
1692
+ "epoch": 0.96,
1693
+ "learning_rate": 2.4329828146074096e-08,
1694
+ "logits/chosen": -2.1215853691101074,
1695
+ "logits/rejected": -1.810217261314392,
1696
+ "logps/chosen": -454.5079040527344,
1697
+ "logps/rejected": -459.13134765625,
1698
+ "loss": 0.5396,
1699
+ "rewards/accuracies": 0.6000000238418579,
1700
+ "rewards/chosen": -1.8914794921875,
1701
+ "rewards/margins": 0.6684591174125671,
1702
+ "rewards/rejected": -2.559938430786133,
1703
+ "step": 1200
1704
+ },
1705
+ {
1706
+ "epoch": 0.97,
1707
+ "learning_rate": 1.5580202098509078e-08,
1708
+ "logits/chosen": -1.9334113597869873,
1709
+ "logits/rejected": -1.882664680480957,
1710
+ "logps/chosen": -512.0015869140625,
1711
+ "logps/rejected": -529.5744018554688,
1712
+ "loss": 0.5535,
1713
+ "rewards/accuracies": 0.625,
1714
+ "rewards/chosen": -1.7347371578216553,
1715
+ "rewards/margins": 0.7627021074295044,
1716
+ "rewards/rejected": -2.49743914604187,
1717
+ "step": 1210
1718
+ },
1719
+ {
1720
+ "epoch": 0.98,
1721
+ "learning_rate": 8.767851876239075e-09,
1722
+ "logits/chosen": -1.6884361505508423,
1723
+ "logits/rejected": -1.633782148361206,
1724
+ "logps/chosen": -397.6700134277344,
1725
+ "logps/rejected": -514.9879150390625,
1726
+ "loss": 0.5168,
1727
+ "rewards/accuracies": 0.75,
1728
+ "rewards/chosen": -1.720780611038208,
1729
+ "rewards/margins": 1.0960266590118408,
1730
+ "rewards/rejected": -2.816807270050049,
1731
+ "step": 1220
1732
+ },
1733
+ {
1734
+ "epoch": 0.98,
1735
+ "learning_rate": 3.8980895450474455e-09,
1736
+ "logits/chosen": -1.887984037399292,
1737
+ "logits/rejected": -1.7204885482788086,
1738
+ "logps/chosen": -388.229248046875,
1739
+ "logps/rejected": -438.8857421875,
1740
+ "loss": 0.5719,
1741
+ "rewards/accuracies": 0.6000000238418579,
1742
+ "rewards/chosen": -1.6957340240478516,
1743
+ "rewards/margins": 0.7062219381332397,
1744
+ "rewards/rejected": -2.4019558429718018,
1745
+ "step": 1230
1746
+ },
1747
+ {
1748
+ "epoch": 0.99,
1749
+ "learning_rate": 9.747123991141193e-10,
1750
+ "logits/chosen": -1.995326042175293,
1751
+ "logits/rejected": -1.8855457305908203,
1752
+ "logps/chosen": -367.4112243652344,
1753
+ "logps/rejected": -460.2139587402344,
1754
+ "loss": 0.5104,
1755
+ "rewards/accuracies": 0.7749999761581421,
1756
+ "rewards/chosen": -1.427626132965088,
1757
+ "rewards/margins": 1.0081008672714233,
1758
+ "rewards/rejected": -2.4357268810272217,
1759
+ "step": 1240
1760
+ },
1761
+ {
1762
+ "epoch": 1.0,
1763
+ "learning_rate": 0.0,
1764
+ "logits/chosen": -1.856300950050354,
1765
+ "logits/rejected": -1.691680908203125,
1766
+ "logps/chosen": -407.53253173828125,
1767
+ "logps/rejected": -473.7496643066406,
1768
+ "loss": 0.4906,
1769
+ "rewards/accuracies": 0.6000000238418579,
1770
+ "rewards/chosen": -1.637367606163025,
1771
+ "rewards/margins": 0.8503096699714661,
1772
+ "rewards/rejected": -2.4876773357391357,
1773
+ "step": 1250
1774
+ },
1775
+ {
1776
+ "epoch": 1.0,
1777
+ "step": 1250,
1778
+ "total_flos": 0.0,
1779
+ "train_loss": 0.37276097068786623,
1780
+ "train_runtime": 8917.3793,
1781
+ "train_samples_per_second": 1.682,
1782
+ "train_steps_per_second": 0.14
1783
+ }
1784
+ ],
1785
+ "logging_steps": 10,
1786
+ "max_steps": 1250,
1787
+ "num_input_tokens_seen": 0,
1788
+ "num_train_epochs": 1,
1789
+ "save_steps": 20,
1790
+ "total_flos": 0.0,
1791
+ "train_batch_size": 2,
1792
+ "trial_name": null,
1793
+ "trial_params": null
1794
+ }