BraylonDash commited on
Commit
8488538
·
verified ·
1 Parent(s): 79354e2

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 sha256:80d3845f86a0f9dbb0f5c412cf252c716e33de0c90db817777e84772b622bb00
3
  size 671150064
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:23c0abf969457aa7dcf137435a7a164da7d19cb150d923648680fee8079e1b32
3
  size 671150064
all_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "train_loss": 0.5286114639282227,
4
+ "train_runtime": 12048.4045,
5
+ "train_samples": 15000,
6
+ "train_samples_per_second": 1.245,
7
+ "train_steps_per_second": 0.104
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
+ 2024-09-18T21:04:25,codecarbon,b950d560-4581-446f-98fe-88b07610120c,12048.41871213913,0.006258699687279359,5.194623325111982e-07,42.5,613.8600000000001,188.74309015274048,0.14223505016283855,1.8618230901177322,0.6287243774526995,2.632782517733275,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/Sep18_17-42-36_gpu4-119-5/events.out.tfevents.1726645416.gpu4-119-5.3645053.0 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:8af51ebbdd6cec19f55aefebb566e0fa4f333a263ddaf413183a6423acd76566
3
- size 78105
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f9da0eada45e7ca68c0a5efedace8715ce7614821f2bbad966c9fae2c59ce7dc
3
+ size 79093
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "train_loss": 0.5286114639282227,
4
+ "train_runtime": 12048.4045,
5
+ "train_samples": 15000,
6
+ "train_samples_per_second": 1.245,
7
+ "train_steps_per_second": 0.104
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": -2.477166175842285,
29
+ "logits/rejected": -2.134406089782715,
30
+ "logps/chosen": -285.34820556640625,
31
+ "logps/rejected": -191.54904174804688,
32
+ "loss": 0.6933,
33
+ "rewards/accuracies": 0.2777777910232544,
34
+ "rewards/chosen": 8.868110307957977e-05,
35
+ "rewards/margins": -0.00045536039397120476,
36
+ "rewards/rejected": 0.0005440415116026998,
37
+ "step": 10
38
+ },
39
+ {
40
+ "epoch": 0.02,
41
+ "learning_rate": 8.000000000000001e-07,
42
+ "logits/chosen": -2.361990213394165,
43
+ "logits/rejected": -2.128539562225342,
44
+ "logps/chosen": -271.50360107421875,
45
+ "logps/rejected": -208.84353637695312,
46
+ "loss": 0.6931,
47
+ "rewards/accuracies": 0.30000001192092896,
48
+ "rewards/chosen": 0.0001849079126259312,
49
+ "rewards/margins": -0.0003974610008299351,
50
+ "rewards/rejected": 0.0005823688698001206,
51
+ "step": 20
52
+ },
53
+ {
54
+ "epoch": 0.02,
55
+ "learning_rate": 1.2000000000000002e-06,
56
+ "logits/chosen": -2.2791178226470947,
57
+ "logits/rejected": -2.2503268718719482,
58
+ "logps/chosen": -269.80023193359375,
59
+ "logps/rejected": -289.0177307128906,
60
+ "loss": 0.693,
61
+ "rewards/accuracies": 0.6499999761581421,
62
+ "rewards/chosen": 0.0028744328301399946,
63
+ "rewards/margins": 0.002431499771773815,
64
+ "rewards/rejected": 0.00044293305836617947,
65
+ "step": 30
66
+ },
67
+ {
68
+ "epoch": 0.03,
69
+ "learning_rate": 1.6000000000000001e-06,
70
+ "logits/chosen": -2.5082411766052246,
71
+ "logits/rejected": -2.340522050857544,
72
+ "logps/chosen": -210.58447265625,
73
+ "logps/rejected": -181.58480834960938,
74
+ "loss": 0.6929,
75
+ "rewards/accuracies": 0.44999998807907104,
76
+ "rewards/chosen": 0.0015180089976638556,
77
+ "rewards/margins": 0.0007624348509125412,
78
+ "rewards/rejected": 0.0007555742631666362,
79
+ "step": 40
80
+ },
81
+ {
82
+ "epoch": 0.04,
83
+ "learning_rate": 2.0000000000000003e-06,
84
+ "logits/chosen": -2.3627030849456787,
85
+ "logits/rejected": -2.3390040397644043,
86
+ "logps/chosen": -195.58641052246094,
87
+ "logps/rejected": -211.3288116455078,
88
+ "loss": 0.6924,
89
+ "rewards/accuracies": 0.550000011920929,
90
+ "rewards/chosen": 0.0013842754997313023,
91
+ "rewards/margins": 0.0027850826736539602,
92
+ "rewards/rejected": -0.0014008075231686234,
93
+ "step": 50
94
+ },
95
+ {
96
+ "epoch": 0.05,
97
+ "learning_rate": 2.4000000000000003e-06,
98
+ "logits/chosen": -2.494563341140747,
99
+ "logits/rejected": -2.3034842014312744,
100
+ "logps/chosen": -244.6390838623047,
101
+ "logps/rejected": -277.8663635253906,
102
+ "loss": 0.6919,
103
+ "rewards/accuracies": 0.5249999761581421,
104
+ "rewards/chosen": -0.009891030378639698,
105
+ "rewards/margins": 0.0002678747696336359,
106
+ "rewards/rejected": -0.010158904828131199,
107
+ "step": 60
108
+ },
109
+ {
110
+ "epoch": 0.06,
111
+ "learning_rate": 2.8000000000000003e-06,
112
+ "logits/chosen": -2.2501957416534424,
113
+ "logits/rejected": -2.2676117420196533,
114
+ "logps/chosen": -229.9625244140625,
115
+ "logps/rejected": -216.9250946044922,
116
+ "loss": 0.6902,
117
+ "rewards/accuracies": 0.6000000238418579,
118
+ "rewards/chosen": -0.011323804967105389,
119
+ "rewards/margins": 0.008610954508185387,
120
+ "rewards/rejected": -0.0199347585439682,
121
+ "step": 70
122
+ },
123
+ {
124
+ "epoch": 0.06,
125
+ "learning_rate": 3.2000000000000003e-06,
126
+ "logits/chosen": -2.401120901107788,
127
+ "logits/rejected": -2.3949601650238037,
128
+ "logps/chosen": -265.9338684082031,
129
+ "logps/rejected": -262.85272216796875,
130
+ "loss": 0.6863,
131
+ "rewards/accuracies": 0.6000000238418579,
132
+ "rewards/chosen": -0.012857760302722454,
133
+ "rewards/margins": 0.021094320341944695,
134
+ "rewards/rejected": -0.033952079713344574,
135
+ "step": 80
136
+ },
137
+ {
138
+ "epoch": 0.07,
139
+ "learning_rate": 3.6000000000000003e-06,
140
+ "logits/chosen": -2.330019474029541,
141
+ "logits/rejected": -1.993003487586975,
142
+ "logps/chosen": -267.9632873535156,
143
+ "logps/rejected": -179.67996215820312,
144
+ "loss": 0.6864,
145
+ "rewards/accuracies": 0.625,
146
+ "rewards/chosen": -0.025740137323737144,
147
+ "rewards/margins": 0.021240444853901863,
148
+ "rewards/rejected": -0.04698058217763901,
149
+ "step": 90
150
+ },
151
+ {
152
+ "epoch": 0.08,
153
+ "learning_rate": 4.000000000000001e-06,
154
+ "logits/chosen": -2.310814380645752,
155
+ "logits/rejected": -2.3143043518066406,
156
+ "logps/chosen": -274.63226318359375,
157
+ "logps/rejected": -250.58740234375,
158
+ "loss": 0.6895,
159
+ "rewards/accuracies": 0.625,
160
+ "rewards/chosen": -0.028762299567461014,
161
+ "rewards/margins": 0.010499788448214531,
162
+ "rewards/rejected": -0.0392620824277401,
163
+ "step": 100
164
+ },
165
+ {
166
+ "epoch": 0.09,
167
+ "learning_rate": 4.4e-06,
168
+ "logits/chosen": -2.2598509788513184,
169
+ "logits/rejected": -2.14603853225708,
170
+ "logps/chosen": -211.3937225341797,
171
+ "logps/rejected": -179.83021545410156,
172
+ "loss": 0.6815,
173
+ "rewards/accuracies": 0.6000000238418579,
174
+ "rewards/chosen": 0.006855879910290241,
175
+ "rewards/margins": 0.036714259535074234,
176
+ "rewards/rejected": -0.029858380556106567,
177
+ "step": 110
178
+ },
179
+ {
180
+ "epoch": 0.1,
181
+ "learning_rate": 4.800000000000001e-06,
182
+ "logits/chosen": -2.1621768474578857,
183
+ "logits/rejected": -2.172668933868408,
184
+ "logps/chosen": -236.74301147460938,
185
+ "logps/rejected": -268.79986572265625,
186
+ "loss": 0.6701,
187
+ "rewards/accuracies": 0.6499999761581421,
188
+ "rewards/chosen": -0.004106069449335337,
189
+ "rewards/margins": 0.048674389719963074,
190
+ "rewards/rejected": -0.052780456840991974,
191
+ "step": 120
192
+ },
193
+ {
194
+ "epoch": 0.1,
195
+ "learning_rate": 4.999756310023261e-06,
196
+ "logits/chosen": -2.4506375789642334,
197
+ "logits/rejected": -2.3234333992004395,
198
+ "logps/chosen": -235.06594848632812,
199
+ "logps/rejected": -192.08798217773438,
200
+ "loss": 0.6771,
201
+ "rewards/accuracies": 0.574999988079071,
202
+ "rewards/chosen": -0.008781708776950836,
203
+ "rewards/margins": 0.04695295915007591,
204
+ "rewards/rejected": -0.05573466420173645,
205
+ "step": 130
206
+ },
207
+ {
208
+ "epoch": 0.11,
209
+ "learning_rate": 4.997807075247147e-06,
210
+ "logits/chosen": -2.3371706008911133,
211
+ "logits/rejected": -2.0340487957000732,
212
+ "logps/chosen": -201.59564208984375,
213
+ "logps/rejected": -182.42376708984375,
214
+ "loss": 0.6744,
215
+ "rewards/accuracies": 0.574999988079071,
216
+ "rewards/chosen": -0.10378706455230713,
217
+ "rewards/margins": 0.04835032299160957,
218
+ "rewards/rejected": -0.1521373838186264,
219
+ "step": 140
220
+ },
221
+ {
222
+ "epoch": 0.12,
223
+ "learning_rate": 4.993910125649561e-06,
224
+ "logits/chosen": -2.3194079399108887,
225
+ "logits/rejected": -2.2057042121887207,
226
+ "logps/chosen": -232.71102905273438,
227
+ "logps/rejected": -202.6046600341797,
228
+ "loss": 0.6531,
229
+ "rewards/accuracies": 0.6499999761581421,
230
+ "rewards/chosen": -0.04620756581425667,
231
+ "rewards/margins": 0.10107270628213882,
232
+ "rewards/rejected": -0.1472802758216858,
233
+ "step": 150
234
+ },
235
+ {
236
+ "epoch": 0.13,
237
+ "learning_rate": 4.988068499954578e-06,
238
+ "logits/chosen": -2.2133073806762695,
239
+ "logits/rejected": -2.286154270172119,
240
+ "logps/chosen": -283.87872314453125,
241
+ "logps/rejected": -295.348388671875,
242
+ "loss": 0.6795,
243
+ "rewards/accuracies": 0.44999998807907104,
244
+ "rewards/chosen": -0.3122519850730896,
245
+ "rewards/margins": 0.053290385752916336,
246
+ "rewards/rejected": -0.3655424416065216,
247
+ "step": 160
248
+ },
249
+ {
250
+ "epoch": 0.14,
251
+ "learning_rate": 4.980286753286196e-06,
252
+ "logits/chosen": -2.431159019470215,
253
+ "logits/rejected": -2.0319712162017822,
254
+ "logps/chosen": -291.0899353027344,
255
+ "logps/rejected": -253.94198608398438,
256
+ "loss": 0.6407,
257
+ "rewards/accuracies": 0.44999998807907104,
258
+ "rewards/chosen": -0.2418215274810791,
259
+ "rewards/margins": 0.04736508056521416,
260
+ "rewards/rejected": -0.2891865670681,
261
+ "step": 170
262
+ },
263
+ {
264
+ "epoch": 0.14,
265
+ "learning_rate": 4.970570953616383e-06,
266
+ "logits/chosen": -2.2349159717559814,
267
+ "logits/rejected": -2.2275729179382324,
268
+ "logps/chosen": -238.7818603515625,
269
+ "logps/rejected": -243.41622924804688,
270
+ "loss": 0.6578,
271
+ "rewards/accuracies": 0.42500001192092896,
272
+ "rewards/chosen": -0.12219883501529694,
273
+ "rewards/margins": 0.04631805419921875,
274
+ "rewards/rejected": -0.1685168743133545,
275
+ "step": 180
276
+ },
277
+ {
278
+ "epoch": 0.15,
279
+ "learning_rate": 4.958928677033465e-06,
280
+ "logits/chosen": -2.2396342754364014,
281
+ "logits/rejected": -2.1675162315368652,
282
+ "logps/chosen": -255.7926025390625,
283
+ "logps/rejected": -241.1310577392578,
284
+ "loss": 0.6489,
285
+ "rewards/accuracies": 0.625,
286
+ "rewards/chosen": -0.19492614269256592,
287
+ "rewards/margins": 0.1306600272655487,
288
+ "rewards/rejected": -0.325586199760437,
289
+ "step": 190
290
+ },
291
+ {
292
+ "epoch": 0.16,
293
+ "learning_rate": 4.9453690018345144e-06,
294
+ "logits/chosen": -2.2987706661224365,
295
+ "logits/rejected": -2.15234375,
296
+ "logps/chosen": -283.39788818359375,
297
+ "logps/rejected": -306.2623596191406,
298
+ "loss": 0.6387,
299
+ "rewards/accuracies": 0.625,
300
+ "rewards/chosen": -0.29035812616348267,
301
+ "rewards/margins": 0.15090219676494598,
302
+ "rewards/rejected": -0.44126027822494507,
303
+ "step": 200
304
+ },
305
+ {
306
+ "epoch": 0.17,
307
+ "learning_rate": 4.9299025014463665e-06,
308
+ "logits/chosen": -2.449690818786621,
309
+ "logits/rejected": -2.0890920162200928,
310
+ "logps/chosen": -297.98785400390625,
311
+ "logps/rejected": -251.8286895751953,
312
+ "loss": 0.6042,
313
+ "rewards/accuracies": 0.675000011920929,
314
+ "rewards/chosen": -0.24480652809143066,
315
+ "rewards/margins": 0.36872270703315735,
316
+ "rewards/rejected": -0.6135291457176208,
317
+ "step": 210
318
+ },
319
+ {
320
+ "epoch": 0.18,
321
+ "learning_rate": 4.912541236180779e-06,
322
+ "logits/chosen": -2.3042047023773193,
323
+ "logits/rejected": -1.8709608316421509,
324
+ "logps/chosen": -323.5066833496094,
325
+ "logps/rejected": -318.5735168457031,
326
+ "loss": 0.6452,
327
+ "rewards/accuracies": 0.6499999761581421,
328
+ "rewards/chosen": -0.8693976402282715,
329
+ "rewards/margins": 0.33919450640678406,
330
+ "rewards/rejected": -1.208592176437378,
331
+ "step": 220
332
+ },
333
+ {
334
+ "epoch": 0.18,
335
+ "learning_rate": 4.893298743830168e-06,
336
+ "logits/chosen": -2.284022569656372,
337
+ "logits/rejected": -2.3184571266174316,
338
+ "logps/chosen": -280.39013671875,
339
+ "logps/rejected": -322.4888610839844,
340
+ "loss": 0.6498,
341
+ "rewards/accuracies": 0.550000011920929,
342
+ "rewards/chosen": -0.8154991269111633,
343
+ "rewards/margins": 0.15407198667526245,
344
+ "rewards/rejected": -0.9695711135864258,
345
+ "step": 230
346
+ },
347
+ {
348
+ "epoch": 0.19,
349
+ "learning_rate": 4.8721900291112415e-06,
350
+ "logits/chosen": -2.2206528186798096,
351
+ "logits/rejected": -2.040665864944458,
352
+ "logps/chosen": -288.07427978515625,
353
+ "logps/rejected": -295.1524963378906,
354
+ "loss": 0.6204,
355
+ "rewards/accuracies": 0.5249999761581421,
356
+ "rewards/chosen": -0.8145530819892883,
357
+ "rewards/margins": 0.19842155277729034,
358
+ "rewards/rejected": -1.0129746198654175,
359
+ "step": 240
360
+ },
361
+ {
362
+ "epoch": 0.2,
363
+ "learning_rate": 4.849231551964771e-06,
364
+ "logits/chosen": -2.190338373184204,
365
+ "logits/rejected": -2.0681099891662598,
366
+ "logps/chosen": -246.13955688476562,
367
+ "logps/rejected": -268.1896667480469,
368
+ "loss": 0.6422,
369
+ "rewards/accuracies": 0.574999988079071,
370
+ "rewards/chosen": -0.7617325782775879,
371
+ "rewards/margins": 0.19349338114261627,
372
+ "rewards/rejected": -0.955225944519043,
373
+ "step": 250
374
+ },
375
+ {
376
+ "epoch": 0.21,
377
+ "learning_rate": 4.824441214720629e-06,
378
+ "logits/chosen": -2.2392935752868652,
379
+ "logits/rejected": -1.9916995763778687,
380
+ "logps/chosen": -346.5711669921875,
381
+ "logps/rejected": -366.2030944824219,
382
+ "loss": 0.596,
383
+ "rewards/accuracies": 0.675000011920929,
384
+ "rewards/chosen": -0.7316231727600098,
385
+ "rewards/margins": 0.35942238569259644,
386
+ "rewards/rejected": -1.0910453796386719,
387
+ "step": 260
388
+ },
389
+ {
390
+ "epoch": 0.22,
391
+ "learning_rate": 4.7978383481380865e-06,
392
+ "logits/chosen": -2.201519012451172,
393
+ "logits/rejected": -1.9899609088897705,
394
+ "logps/chosen": -315.45465087890625,
395
+ "logps/rejected": -312.4045104980469,
396
+ "loss": 0.5967,
397
+ "rewards/accuracies": 0.6499999761581421,
398
+ "rewards/chosen": -0.6710847020149231,
399
+ "rewards/margins": 0.3409194350242615,
400
+ "rewards/rejected": -1.0120041370391846,
401
+ "step": 270
402
+ },
403
+ {
404
+ "epoch": 0.22,
405
+ "learning_rate": 4.769443696332272e-06,
406
+ "logits/chosen": -1.9985568523406982,
407
+ "logits/rejected": -2.0734457969665527,
408
+ "logps/chosen": -236.2917022705078,
409
+ "logps/rejected": -308.3038024902344,
410
+ "loss": 0.6105,
411
+ "rewards/accuracies": 0.699999988079071,
412
+ "rewards/chosen": -0.6992406249046326,
413
+ "rewards/margins": 0.36759305000305176,
414
+ "rewards/rejected": -1.066833734512329,
415
+ "step": 280
416
+ },
417
+ {
418
+ "epoch": 0.23,
419
+ "learning_rate": 4.7392794005985324e-06,
420
+ "logits/chosen": -2.3794713020324707,
421
+ "logits/rejected": -2.083773136138916,
422
+ "logps/chosen": -338.1396789550781,
423
+ "logps/rejected": -312.3577880859375,
424
+ "loss": 0.59,
425
+ "rewards/accuracies": 0.6000000238418579,
426
+ "rewards/chosen": -0.6138805747032166,
427
+ "rewards/margins": 0.2969156801700592,
428
+ "rewards/rejected": -0.9107962846755981,
429
+ "step": 290
430
+ },
431
+ {
432
+ "epoch": 0.24,
433
+ "learning_rate": 4.707368982147318e-06,
434
+ "logits/chosen": -2.196028470993042,
435
+ "logits/rejected": -2.1599950790405273,
436
+ "logps/chosen": -352.38922119140625,
437
+ "logps/rejected": -310.0411376953125,
438
+ "loss": 0.5703,
439
+ "rewards/accuracies": 0.574999988079071,
440
+ "rewards/chosen": -0.6144166588783264,
441
+ "rewards/margins": 0.37919124960899353,
442
+ "rewards/rejected": -0.9936079978942871,
443
+ "step": 300
444
+ },
445
+ {
446
+ "epoch": 0.25,
447
+ "learning_rate": 4.673737323763048e-06,
448
+ "logits/chosen": -2.0284199714660645,
449
+ "logits/rejected": -2.0696732997894287,
450
+ "logps/chosen": -299.6443786621094,
451
+ "logps/rejected": -365.4646911621094,
452
+ "loss": 0.6192,
453
+ "rewards/accuracies": 0.44999998807907104,
454
+ "rewards/chosen": -1.0700325965881348,
455
+ "rewards/margins": 0.24421080946922302,
456
+ "rewards/rejected": -1.3142435550689697,
457
+ "step": 310
458
+ },
459
+ {
460
+ "epoch": 0.26,
461
+ "learning_rate": 4.638410650401267e-06,
462
+ "logits/chosen": -2.0705060958862305,
463
+ "logits/rejected": -1.7289657592773438,
464
+ "logps/chosen": -373.0816955566406,
465
+ "logps/rejected": -348.92218017578125,
466
+ "loss": 0.6091,
467
+ "rewards/accuracies": 0.550000011920929,
468
+ "rewards/chosen": -1.1900722980499268,
469
+ "rewards/margins": 0.5464798212051392,
470
+ "rewards/rejected": -1.7365522384643555,
471
+ "step": 320
472
+ },
473
+ {
474
+ "epoch": 0.26,
475
+ "learning_rate": 4.601416508739211e-06,
476
+ "logits/chosen": -2.163390636444092,
477
+ "logits/rejected": -2.1936163902282715,
478
+ "logps/chosen": -382.85137939453125,
479
+ "logps/rejected": -361.06494140625,
480
+ "loss": 0.5855,
481
+ "rewards/accuracies": 0.44999998807907104,
482
+ "rewards/chosen": -1.100768804550171,
483
+ "rewards/margins": 0.1708061397075653,
484
+ "rewards/rejected": -1.2715749740600586,
485
+ "step": 330
486
+ },
487
+ {
488
+ "epoch": 0.27,
489
+ "learning_rate": 4.562783745695738e-06,
490
+ "logits/chosen": -2.121084690093994,
491
+ "logits/rejected": -1.8607723712921143,
492
+ "logps/chosen": -273.9230651855469,
493
+ "logps/rejected": -291.5062561035156,
494
+ "loss": 0.589,
495
+ "rewards/accuracies": 0.699999988079071,
496
+ "rewards/chosen": -0.8008276224136353,
497
+ "rewards/margins": 0.5303484201431274,
498
+ "rewards/rejected": -1.3311761617660522,
499
+ "step": 340
500
+ },
501
+ {
502
+ "epoch": 0.28,
503
+ "learning_rate": 4.522542485937369e-06,
504
+ "logits/chosen": -2.2258460521698,
505
+ "logits/rejected": -1.9319546222686768,
506
+ "logps/chosen": -454.8755798339844,
507
+ "logps/rejected": -460.62982177734375,
508
+ "loss": 0.5902,
509
+ "rewards/accuracies": 0.574999988079071,
510
+ "rewards/chosen": -1.990256905555725,
511
+ "rewards/margins": 0.35523930191993713,
512
+ "rewards/rejected": -2.345496416091919,
513
+ "step": 350
514
+ },
515
+ {
516
+ "epoch": 0.29,
517
+ "learning_rate": 4.4807241083879774e-06,
518
+ "logits/chosen": -2.1903085708618164,
519
+ "logits/rejected": -2.0086958408355713,
520
+ "logps/chosen": -601.3458862304688,
521
+ "logps/rejected": -612.4051513671875,
522
+ "loss": 0.5569,
523
+ "rewards/accuracies": 0.550000011920929,
524
+ "rewards/chosen": -3.4381797313690186,
525
+ "rewards/margins": 0.47519922256469727,
526
+ "rewards/rejected": -3.913378953933716,
527
+ "step": 360
528
+ },
529
+ {
530
+ "epoch": 0.3,
531
+ "learning_rate": 4.437361221760449e-06,
532
+ "logits/chosen": -2.0529446601867676,
533
+ "logits/rejected": -1.9928582906723022,
534
+ "logps/chosen": -529.0535278320312,
535
+ "logps/rejected": -547.7728271484375,
536
+ "loss": 0.6339,
537
+ "rewards/accuracies": 0.550000011920929,
538
+ "rewards/chosen": -3.4125945568084717,
539
+ "rewards/margins": 0.22588801383972168,
540
+ "rewards/rejected": -3.6384825706481934,
541
+ "step": 370
542
+ },
543
+ {
544
+ "epoch": 0.3,
545
+ "learning_rate": 4.3924876391293915e-06,
546
+ "logits/chosen": -1.9475816488265991,
547
+ "logits/rejected": -1.9098947048187256,
548
+ "logps/chosen": -375.98992919921875,
549
+ "logps/rejected": -415.5345153808594,
550
+ "loss": 0.5732,
551
+ "rewards/accuracies": 0.550000011920929,
552
+ "rewards/chosen": -1.8701465129852295,
553
+ "rewards/margins": 0.3048131763935089,
554
+ "rewards/rejected": -2.174959659576416,
555
+ "step": 380
556
+ },
557
+ {
558
+ "epoch": 0.31,
559
+ "learning_rate": 4.346138351564711e-06,
560
+ "logits/chosen": -2.0606818199157715,
561
+ "logits/rejected": -1.8960349559783936,
562
+ "logps/chosen": -316.76715087890625,
563
+ "logps/rejected": -320.61114501953125,
564
+ "loss": 0.5587,
565
+ "rewards/accuracies": 0.625,
566
+ "rewards/chosen": -1.2516812086105347,
567
+ "rewards/margins": 0.4567112922668457,
568
+ "rewards/rejected": -1.7083925008773804,
569
+ "step": 390
570
+ },
571
+ {
572
+ "epoch": 0.32,
573
+ "learning_rate": 4.2983495008466285e-06,
574
+ "logits/chosen": -2.0786519050598145,
575
+ "logits/rejected": -2.10847806930542,
576
+ "logps/chosen": -361.7239074707031,
577
+ "logps/rejected": -400.70941162109375,
578
+ "loss": 0.6838,
579
+ "rewards/accuracies": 0.6000000238418579,
580
+ "rewards/chosen": -1.4803402423858643,
581
+ "rewards/margins": 0.255945086479187,
582
+ "rewards/rejected": -1.7362852096557617,
583
+ "step": 400
584
+ },
585
+ {
586
+ "epoch": 0.33,
587
+ "learning_rate": 4.249158351283414e-06,
588
+ "logits/chosen": -2.1878209114074707,
589
+ "logits/rejected": -1.96550714969635,
590
+ "logps/chosen": -349.62078857421875,
591
+ "logps/rejected": -385.7138366699219,
592
+ "loss": 0.6384,
593
+ "rewards/accuracies": 0.574999988079071,
594
+ "rewards/chosen": -0.8981186151504517,
595
+ "rewards/margins": 0.3005538880825043,
596
+ "rewards/rejected": -1.1986725330352783,
597
+ "step": 410
598
+ },
599
+ {
600
+ "epoch": 0.34,
601
+ "learning_rate": 4.198603260653792e-06,
602
+ "logits/chosen": -2.3381173610687256,
603
+ "logits/rejected": -2.1147091388702393,
604
+ "logps/chosen": -364.11090087890625,
605
+ "logps/rejected": -345.1046447753906,
606
+ "loss": 0.602,
607
+ "rewards/accuracies": 0.5249999761581421,
608
+ "rewards/chosen": -0.9675353169441223,
609
+ "rewards/margins": 0.05901508405804634,
610
+ "rewards/rejected": -1.0265504121780396,
611
+ "step": 420
612
+ },
613
+ {
614
+ "epoch": 0.34,
615
+ "learning_rate": 4.146723650296701e-06,
616
+ "logits/chosen": -2.1770806312561035,
617
+ "logits/rejected": -2.23179292678833,
618
+ "logps/chosen": -457.45672607421875,
619
+ "logps/rejected": -525.1158447265625,
620
+ "loss": 0.6147,
621
+ "rewards/accuracies": 0.699999988079071,
622
+ "rewards/chosen": -1.5764873027801514,
623
+ "rewards/margins": 0.5498846769332886,
624
+ "rewards/rejected": -2.1263718605041504,
625
+ "step": 430
626
+ },
627
+ {
628
+ "epoch": 0.35,
629
+ "learning_rate": 4.093559974371725e-06,
630
+ "logits/chosen": -2.175420045852661,
631
+ "logits/rejected": -1.8347841501235962,
632
+ "logps/chosen": -312.1004943847656,
633
+ "logps/rejected": -321.1251525878906,
634
+ "loss": 0.5422,
635
+ "rewards/accuracies": 0.625,
636
+ "rewards/chosen": -1.281625509262085,
637
+ "rewards/margins": 0.4661393165588379,
638
+ "rewards/rejected": -1.7477649450302124,
639
+ "step": 440
640
+ },
641
+ {
642
+ "epoch": 0.36,
643
+ "learning_rate": 4.039153688314146e-06,
644
+ "logits/chosen": -2.1562466621398926,
645
+ "logits/rejected": -1.9570480585098267,
646
+ "logps/chosen": -400.806396484375,
647
+ "logps/rejected": -429.9332580566406,
648
+ "loss": 0.618,
649
+ "rewards/accuracies": 0.550000011920929,
650
+ "rewards/chosen": -1.3771603107452393,
651
+ "rewards/margins": 0.6382063031196594,
652
+ "rewards/rejected": -2.015366554260254,
653
+ "step": 450
654
+ },
655
+ {
656
+ "epoch": 0.37,
657
+ "learning_rate": 3.983547216509254e-06,
658
+ "logits/chosen": -2.1951937675476074,
659
+ "logits/rejected": -2.1014106273651123,
660
+ "logps/chosen": -404.2888488769531,
661
+ "logps/rejected": -422.8614807128906,
662
+ "loss": 0.5802,
663
+ "rewards/accuracies": 0.6499999761581421,
664
+ "rewards/chosen": -1.7574684619903564,
665
+ "rewards/margins": 0.48904165625572205,
666
+ "rewards/rejected": -2.2465100288391113,
667
+ "step": 460
668
+ },
669
+ {
670
+ "epoch": 0.38,
671
+ "learning_rate": 3.92678391921108e-06,
672
+ "logits/chosen": -2.0647380352020264,
673
+ "logits/rejected": -1.9699338674545288,
674
+ "logps/chosen": -319.4429626464844,
675
+ "logps/rejected": -336.6853332519531,
676
+ "loss": 0.6099,
677
+ "rewards/accuracies": 0.4749999940395355,
678
+ "rewards/chosen": -1.5423263311386108,
679
+ "rewards/margins": 0.34772855043411255,
680
+ "rewards/rejected": -1.8900550603866577,
681
+ "step": 470
682
+ },
683
+ {
684
+ "epoch": 0.38,
685
+ "learning_rate": 3.868908058731376e-06,
686
+ "logits/chosen": -2.0104379653930664,
687
+ "logits/rejected": -1.9088401794433594,
688
+ "logps/chosen": -430.5732421875,
689
+ "logps/rejected": -496.94720458984375,
690
+ "loss": 0.5956,
691
+ "rewards/accuracies": 0.574999988079071,
692
+ "rewards/chosen": -2.6480965614318848,
693
+ "rewards/margins": 0.45647579431533813,
694
+ "rewards/rejected": -3.1045725345611572,
695
+ "step": 480
696
+ },
697
+ {
698
+ "epoch": 0.39,
699
+ "learning_rate": 3.8099647649251984e-06,
700
+ "logits/chosen": -2.1025550365448,
701
+ "logits/rejected": -1.8321536779403687,
702
+ "logps/chosen": -518.1032104492188,
703
+ "logps/rejected": -518.704345703125,
704
+ "loss": 0.5822,
705
+ "rewards/accuracies": 0.625,
706
+ "rewards/chosen": -2.717057466506958,
707
+ "rewards/margins": 0.5123800039291382,
708
+ "rewards/rejected": -3.2294373512268066,
709
+ "step": 490
710
+ },
711
+ {
712
+ "epoch": 0.4,
713
+ "learning_rate": 3.7500000000000005e-06,
714
+ "logits/chosen": -2.1220908164978027,
715
+ "logits/rejected": -2.0493388175964355,
716
+ "logps/chosen": -415.57196044921875,
717
+ "logps/rejected": -471.4869079589844,
718
+ "loss": 0.5952,
719
+ "rewards/accuracies": 0.625,
720
+ "rewards/chosen": -2.2129175662994385,
721
+ "rewards/margins": 0.5477197766304016,
722
+ "rewards/rejected": -2.7606372833251953,
723
+ "step": 500
724
+ },
725
+ {
726
+ "epoch": 0.41,
727
+ "learning_rate": 3.689060522675689e-06,
728
+ "logits/chosen": -2.2190887928009033,
729
+ "logits/rejected": -2.0425148010253906,
730
+ "logps/chosen": -435.57763671875,
731
+ "logps/rejected": -455.32757568359375,
732
+ "loss": 0.5279,
733
+ "rewards/accuracies": 0.699999988079071,
734
+ "rewards/chosen": -1.938948392868042,
735
+ "rewards/margins": 0.5450895428657532,
736
+ "rewards/rejected": -2.4840381145477295,
737
+ "step": 510
738
+ },
739
+ {
740
+ "epoch": 0.42,
741
+ "learning_rate": 3.627193851723577e-06,
742
+ "logits/chosen": -2.1195268630981445,
743
+ "logits/rejected": -1.9826160669326782,
744
+ "logps/chosen": -442.97576904296875,
745
+ "logps/rejected": -464.6687927246094,
746
+ "loss": 0.5773,
747
+ "rewards/accuracies": 0.4749999940395355,
748
+ "rewards/chosen": -2.141632080078125,
749
+ "rewards/margins": 0.47049254179000854,
750
+ "rewards/rejected": -2.6121246814727783,
751
+ "step": 520
752
+ },
753
+ {
754
+ "epoch": 0.42,
755
+ "learning_rate": 3.564448228912682e-06,
756
+ "logits/chosen": -2.1186394691467285,
757
+ "logits/rejected": -2.0721774101257324,
758
+ "logps/chosen": -458.518310546875,
759
+ "logps/rejected": -493.88653564453125,
760
+ "loss": 0.5786,
761
+ "rewards/accuracies": 0.5249999761581421,
762
+ "rewards/chosen": -2.1196634769439697,
763
+ "rewards/margins": 0.24448445439338684,
764
+ "rewards/rejected": -2.364147663116455,
765
+ "step": 530
766
+ },
767
+ {
768
+ "epoch": 0.43,
769
+ "learning_rate": 3.5008725813922383e-06,
770
+ "logits/chosen": -2.1211726665496826,
771
+ "logits/rejected": -1.8444690704345703,
772
+ "logps/chosen": -473.51470947265625,
773
+ "logps/rejected": -447.91400146484375,
774
+ "loss": 0.5624,
775
+ "rewards/accuracies": 0.574999988079071,
776
+ "rewards/chosen": -2.1058106422424316,
777
+ "rewards/margins": 0.4687918722629547,
778
+ "rewards/rejected": -2.5746026039123535,
779
+ "step": 540
780
+ },
781
+ {
782
+ "epoch": 0.44,
783
+ "learning_rate": 3.436516483539781e-06,
784
+ "logits/chosen": -2.0425033569335938,
785
+ "logits/rejected": -1.8863608837127686,
786
+ "logps/chosen": -412.14239501953125,
787
+ "logps/rejected": -484.99334716796875,
788
+ "loss": 0.5227,
789
+ "rewards/accuracies": 0.675000011920929,
790
+ "rewards/chosen": -2.103408098220825,
791
+ "rewards/margins": 0.9393216967582703,
792
+ "rewards/rejected": -3.042729616165161,
793
+ "step": 550
794
+ },
795
+ {
796
+ "epoch": 0.45,
797
+ "learning_rate": 3.3714301183045382e-06,
798
+ "logits/chosen": -2.1661436557769775,
799
+ "logits/rejected": -2.238974094390869,
800
+ "logps/chosen": -504.8619689941406,
801
+ "logps/rejected": -558.937744140625,
802
+ "loss": 0.4986,
803
+ "rewards/accuracies": 0.625,
804
+ "rewards/chosen": -2.069854736328125,
805
+ "rewards/margins": 0.780713677406311,
806
+ "rewards/rejected": -2.8505682945251465,
807
+ "step": 560
808
+ },
809
+ {
810
+ "epoch": 0.46,
811
+ "learning_rate": 3.3056642380762783e-06,
812
+ "logits/chosen": -2.04276967048645,
813
+ "logits/rejected": -2.0428547859191895,
814
+ "logps/chosen": -422.66351318359375,
815
+ "logps/rejected": -454.6181640625,
816
+ "loss": 0.5758,
817
+ "rewards/accuracies": 0.574999988079071,
818
+ "rewards/chosen": -1.6807305812835693,
819
+ "rewards/margins": 0.5961163640022278,
820
+ "rewards/rejected": -2.2768468856811523,
821
+ "step": 570
822
+ },
823
+ {
824
+ "epoch": 0.46,
825
+ "learning_rate": 3.2392701251101172e-06,
826
+ "logits/chosen": -2.0288186073303223,
827
+ "logits/rejected": -1.9108963012695312,
828
+ "logps/chosen": -413.84832763671875,
829
+ "logps/rejected": -435.688232421875,
830
+ "loss": 0.5815,
831
+ "rewards/accuracies": 0.5249999761581421,
832
+ "rewards/chosen": -1.9628708362579346,
833
+ "rewards/margins": 0.696241557598114,
834
+ "rewards/rejected": -2.6591122150421143,
835
+ "step": 580
836
+ },
837
+ {
838
+ "epoch": 0.47,
839
+ "learning_rate": 3.1722995515381644e-06,
840
+ "logits/chosen": -1.9494431018829346,
841
+ "logits/rejected": -1.8113425970077515,
842
+ "logps/chosen": -407.6769104003906,
843
+ "logps/rejected": -408.3319396972656,
844
+ "loss": 0.5468,
845
+ "rewards/accuracies": 0.550000011920929,
846
+ "rewards/chosen": -1.5669306516647339,
847
+ "rewards/margins": 0.4425128400325775,
848
+ "rewards/rejected": -2.0094432830810547,
849
+ "step": 590
850
+ },
851
+ {
852
+ "epoch": 0.48,
853
+ "learning_rate": 3.1048047389991693e-06,
854
+ "logits/chosen": -2.0704987049102783,
855
+ "logits/rejected": -1.9060993194580078,
856
+ "logps/chosen": -397.5736389160156,
857
+ "logps/rejected": -465.2367248535156,
858
+ "loss": 0.5822,
859
+ "rewards/accuracies": 0.6499999761581421,
860
+ "rewards/chosen": -1.8127038478851318,
861
+ "rewards/margins": 0.5960454344749451,
862
+ "rewards/rejected": -2.4087491035461426,
863
+ "step": 600
864
+ },
865
+ {
866
+ "epoch": 0.49,
867
+ "learning_rate": 3.0368383179176584e-06,
868
+ "logits/chosen": -1.999671220779419,
869
+ "logits/rejected": -1.9255882501602173,
870
+ "logps/chosen": -422.7398986816406,
871
+ "logps/rejected": -462.8828125,
872
+ "loss": 0.537,
873
+ "rewards/accuracies": 0.6499999761581421,
874
+ "rewards/chosen": -1.8113523721694946,
875
+ "rewards/margins": 0.7707425951957703,
876
+ "rewards/rejected": -2.58209490776062,
877
+ "step": 610
878
+ },
879
+ {
880
+ "epoch": 0.5,
881
+ "learning_rate": 2.9684532864643123e-06,
882
+ "logits/chosen": -2.034332275390625,
883
+ "logits/rejected": -1.918892502784729,
884
+ "logps/chosen": -454.9341735839844,
885
+ "logps/rejected": -488.122314453125,
886
+ "loss": 0.5587,
887
+ "rewards/accuracies": 0.675000011920929,
888
+ "rewards/chosen": -2.0255672931671143,
889
+ "rewards/margins": 0.639295220375061,
890
+ "rewards/rejected": -2.6648623943328857,
891
+ "step": 620
892
+ },
893
+ {
894
+ "epoch": 0.5,
895
+ "learning_rate": 2.8997029692295875e-06,
896
+ "logits/chosen": -2.1899733543395996,
897
+ "logits/rejected": -2.0211856365203857,
898
+ "logps/chosen": -444.3063049316406,
899
+ "logps/rejected": -499.0299377441406,
900
+ "loss": 0.51,
901
+ "rewards/accuracies": 0.6499999761581421,
902
+ "rewards/chosen": -1.7396934032440186,
903
+ "rewards/margins": 0.8382138013839722,
904
+ "rewards/rejected": -2.577907085418701,
905
+ "step": 630
906
+ },
907
+ {
908
+ "epoch": 0.51,
909
+ "learning_rate": 2.8306409756428067e-06,
910
+ "logits/chosen": -2.042299747467041,
911
+ "logits/rejected": -1.6256879568099976,
912
+ "logps/chosen": -482.6806640625,
913
+ "logps/rejected": -522.2861328125,
914
+ "loss": 0.525,
915
+ "rewards/accuracies": 0.7250000238418579,
916
+ "rewards/chosen": -1.8350658416748047,
917
+ "rewards/margins": 1.0798089504241943,
918
+ "rewards/rejected": -2.9148752689361572,
919
+ "step": 640
920
+ },
921
+ {
922
+ "epoch": 0.52,
923
+ "learning_rate": 2.761321158169134e-06,
924
+ "logits/chosen": -2.126718044281006,
925
+ "logits/rejected": -1.8609893321990967,
926
+ "logps/chosen": -578.2748413085938,
927
+ "logps/rejected": -569.1837768554688,
928
+ "loss": 0.5116,
929
+ "rewards/accuracies": 0.625,
930
+ "rewards/chosen": -3.034454822540283,
931
+ "rewards/margins": 0.5776246786117554,
932
+ "rewards/rejected": -3.612079620361328,
933
+ "step": 650
934
+ },
935
+ {
936
+ "epoch": 0.53,
937
+ "learning_rate": 2.6917975703170466e-06,
938
+ "logits/chosen": -1.7727696895599365,
939
+ "logits/rejected": -1.48529052734375,
940
+ "logps/chosen": -634.0877685546875,
941
+ "logps/rejected": -776.0706176757812,
942
+ "loss": 0.5475,
943
+ "rewards/accuracies": 0.7250000238418579,
944
+ "rewards/chosen": -3.8410019874572754,
945
+ "rewards/margins": 1.3400521278381348,
946
+ "rewards/rejected": -5.181053638458252,
947
+ "step": 660
948
+ },
949
+ {
950
+ "epoch": 0.54,
951
+ "learning_rate": 2.6221244244890336e-06,
952
+ "logits/chosen": -1.8298507928848267,
953
+ "logits/rejected": -1.6536800861358643,
954
+ "logps/chosen": -668.0659790039062,
955
+ "logps/rejected": -715.1585083007812,
956
+ "loss": 0.5195,
957
+ "rewards/accuracies": 0.7250000238418579,
958
+ "rewards/chosen": -4.25075101852417,
959
+ "rewards/margins": 0.8174532055854797,
960
+ "rewards/rejected": -5.068203926086426,
961
+ "step": 670
962
+ },
963
+ {
964
+ "epoch": 0.54,
965
+ "learning_rate": 2.5523560497083927e-06,
966
+ "logits/chosen": -2.064204692840576,
967
+ "logits/rejected": -1.8902099132537842,
968
+ "logps/chosen": -402.74334716796875,
969
+ "logps/rejected": -417.3500061035156,
970
+ "loss": 0.6347,
971
+ "rewards/accuracies": 0.5249999761581421,
972
+ "rewards/chosen": -2.059872627258301,
973
+ "rewards/margins": 0.30380430817604065,
974
+ "rewards/rejected": -2.3636770248413086,
975
+ "step": 680
976
+ },
977
+ {
978
+ "epoch": 0.55,
979
+ "learning_rate": 2.482546849255096e-06,
980
+ "logits/chosen": -2.0144591331481934,
981
+ "logits/rejected": -1.7140735387802124,
982
+ "logps/chosen": -352.4971008300781,
983
+ "logps/rejected": -371.90869140625,
984
+ "loss": 0.5059,
985
+ "rewards/accuracies": 0.675000011920929,
986
+ "rewards/chosen": -1.3799242973327637,
987
+ "rewards/margins": 0.5062106251716614,
988
+ "rewards/rejected": -1.8861348628997803,
989
+ "step": 690
990
+ },
991
+ {
992
+ "epoch": 0.56,
993
+ "learning_rate": 2.4127512582437486e-06,
994
+ "logits/chosen": -2.0474960803985596,
995
+ "logits/rejected": -1.9881216287612915,
996
+ "logps/chosen": -443.65185546875,
997
+ "logps/rejected": -492.8255310058594,
998
+ "loss": 0.5122,
999
+ "rewards/accuracies": 0.7250000238418579,
1000
+ "rewards/chosen": -1.6040445566177368,
1001
+ "rewards/margins": 0.7050210237503052,
1002
+ "rewards/rejected": -2.309065341949463,
1003
+ "step": 700
1004
+ },
1005
+ {
1006
+ "epoch": 0.57,
1007
+ "learning_rate": 2.3430237011767166e-06,
1008
+ "logits/chosen": -1.8097617626190186,
1009
+ "logits/rejected": -1.6723369359970093,
1010
+ "logps/chosen": -468.43475341796875,
1011
+ "logps/rejected": -514.5213012695312,
1012
+ "loss": 0.5255,
1013
+ "rewards/accuracies": 0.75,
1014
+ "rewards/chosen": -2.4575445652008057,
1015
+ "rewards/margins": 0.8011848330497742,
1016
+ "rewards/rejected": -3.2587294578552246,
1017
+ "step": 710
1018
+ },
1019
+ {
1020
+ "epoch": 0.58,
1021
+ "learning_rate": 2.2734185495055503e-06,
1022
+ "logits/chosen": -1.6928268671035767,
1023
+ "logits/rejected": -1.4622031450271606,
1024
+ "logps/chosen": -615.6226806640625,
1025
+ "logps/rejected": -672.2274169921875,
1026
+ "loss": 0.4353,
1027
+ "rewards/accuracies": 0.75,
1028
+ "rewards/chosen": -3.9083850383758545,
1029
+ "rewards/margins": 1.1158344745635986,
1030
+ "rewards/rejected": -5.024219512939453,
1031
+ "step": 720
1032
+ },
1033
+ {
1034
+ "epoch": 0.58,
1035
+ "learning_rate": 2.2039900792337477e-06,
1036
+ "logits/chosen": -1.6810636520385742,
1037
+ "logits/rejected": -1.4750381708145142,
1038
+ "logps/chosen": -718.7698364257812,
1039
+ "logps/rejected": -769.5067749023438,
1040
+ "loss": 0.5824,
1041
+ "rewards/accuracies": 0.6000000238418579,
1042
+ "rewards/chosen": -4.622340202331543,
1043
+ "rewards/margins": 0.9261082410812378,
1044
+ "rewards/rejected": -5.548448085784912,
1045
+ "step": 730
1046
+ },
1047
+ {
1048
+ "epoch": 0.59,
1049
+ "learning_rate": 2.134792428593971e-06,
1050
+ "logits/chosen": -1.858790636062622,
1051
+ "logits/rejected": -1.5472383499145508,
1052
+ "logps/chosen": -749.5997924804688,
1053
+ "logps/rejected": -797.5546875,
1054
+ "loss": 0.5051,
1055
+ "rewards/accuracies": 0.6000000238418579,
1056
+ "rewards/chosen": -5.035767078399658,
1057
+ "rewards/margins": 0.9318579435348511,
1058
+ "rewards/rejected": -5.967624664306641,
1059
+ "step": 740
1060
+ },
1061
+ {
1062
+ "epoch": 0.6,
1063
+ "learning_rate": 2.0658795558326745e-06,
1064
+ "logits/chosen": -1.7888858318328857,
1065
+ "logits/rejected": -1.8512537479400635,
1066
+ "logps/chosen": -627.5963134765625,
1067
+ "logps/rejected": -687.7399291992188,
1068
+ "loss": 0.5808,
1069
+ "rewards/accuracies": 0.6000000238418579,
1070
+ "rewards/chosen": -4.275559425354004,
1071
+ "rewards/margins": 0.5892196297645569,
1072
+ "rewards/rejected": -4.864778995513916,
1073
+ "step": 750
1074
+ },
1075
+ {
1076
+ "epoch": 0.61,
1077
+ "learning_rate": 1.997305197135089e-06,
1078
+ "logits/chosen": -1.8849258422851562,
1079
+ "logits/rejected": -1.7010767459869385,
1080
+ "logps/chosen": -621.4512939453125,
1081
+ "logps/rejected": -654.05224609375,
1082
+ "loss": 0.5652,
1083
+ "rewards/accuracies": 0.6000000238418579,
1084
+ "rewards/chosen": -4.103709697723389,
1085
+ "rewards/margins": 0.48206719756126404,
1086
+ "rewards/rejected": -4.5857768058776855,
1087
+ "step": 760
1088
+ },
1089
+ {
1090
+ "epoch": 0.62,
1091
+ "learning_rate": 1.9291228247233607e-06,
1092
+ "logits/chosen": -2.156606674194336,
1093
+ "logits/rejected": -1.8361657857894897,
1094
+ "logps/chosen": -640.6911010742188,
1095
+ "logps/rejected": -605.0956420898438,
1096
+ "loss": 0.5607,
1097
+ "rewards/accuracies": 0.625,
1098
+ "rewards/chosen": -3.5032150745391846,
1099
+ "rewards/margins": 0.49892592430114746,
1100
+ "rewards/rejected": -4.002140045166016,
1101
+ "step": 770
1102
+ },
1103
+ {
1104
+ "epoch": 0.62,
1105
+ "learning_rate": 1.8613856051605242e-06,
1106
+ "logits/chosen": -1.863431692123413,
1107
+ "logits/rejected": -1.6354913711547852,
1108
+ "logps/chosen": -584.6881713867188,
1109
+ "logps/rejected": -643.9428100585938,
1110
+ "loss": 0.5343,
1111
+ "rewards/accuracies": 0.7250000238418579,
1112
+ "rewards/chosen": -3.2495388984680176,
1113
+ "rewards/margins": 0.8126918077468872,
1114
+ "rewards/rejected": -4.062230587005615,
1115
+ "step": 780
1116
+ },
1117
+ {
1118
+ "epoch": 0.63,
1119
+ "learning_rate": 1.7941463578928088e-06,
1120
+ "logits/chosen": -1.978654146194458,
1121
+ "logits/rejected": -1.9466804265975952,
1122
+ "logps/chosen": -587.6908569335938,
1123
+ "logps/rejected": -607.2252807617188,
1124
+ "loss": 0.6187,
1125
+ "rewards/accuracies": 0.550000011920929,
1126
+ "rewards/chosen": -3.702901840209961,
1127
+ "rewards/margins": 0.33423447608947754,
1128
+ "rewards/rejected": -4.037137031555176,
1129
+ "step": 790
1130
+ },
1131
+ {
1132
+ "epoch": 0.64,
1133
+ "learning_rate": 1.7274575140626318e-06,
1134
+ "logits/chosen": -1.870070457458496,
1135
+ "logits/rejected": -2.013461112976074,
1136
+ "logps/chosen": -526.104248046875,
1137
+ "logps/rejected": -600.121337890625,
1138
+ "loss": 0.484,
1139
+ "rewards/accuracies": 0.6499999761581421,
1140
+ "rewards/chosen": -2.800811767578125,
1141
+ "rewards/margins": 0.7185395956039429,
1142
+ "rewards/rejected": -3.5193512439727783,
1143
+ "step": 800
1144
+ },
1145
+ {
1146
+ "epoch": 0.65,
1147
+ "learning_rate": 1.661371075624363e-06,
1148
+ "logits/chosen": -1.931865930557251,
1149
+ "logits/rejected": -1.6386762857437134,
1150
+ "logps/chosen": -461.39520263671875,
1151
+ "logps/rejected": -491.856689453125,
1152
+ "loss": 0.5377,
1153
+ "rewards/accuracies": 0.574999988079071,
1154
+ "rewards/chosen": -2.690922975540161,
1155
+ "rewards/margins": 0.708380401134491,
1156
+ "rewards/rejected": -3.399303436279297,
1157
+ "step": 810
1158
+ },
1159
+ {
1160
+ "epoch": 0.66,
1161
+ "learning_rate": 1.5959385747947697e-06,
1162
+ "logits/chosen": -1.916992425918579,
1163
+ "logits/rejected": -1.8899227380752563,
1164
+ "logps/chosen": -531.813232421875,
1165
+ "logps/rejected": -615.9520263671875,
1166
+ "loss": 0.4847,
1167
+ "rewards/accuracies": 0.675000011920929,
1168
+ "rewards/chosen": -2.888909101486206,
1169
+ "rewards/margins": 0.7070828676223755,
1170
+ "rewards/rejected": -3.595991849899292,
1171
+ "step": 820
1172
+ },
1173
+ {
1174
+ "epoch": 0.66,
1175
+ "learning_rate": 1.5312110338697427e-06,
1176
+ "logits/chosen": -2.0526282787323,
1177
+ "logits/rejected": -1.9669021368026733,
1178
+ "logps/chosen": -449.2479553222656,
1179
+ "logps/rejected": -457.72027587890625,
1180
+ "loss": 0.6194,
1181
+ "rewards/accuracies": 0.5249999761581421,
1182
+ "rewards/chosen": -2.594003200531006,
1183
+ "rewards/margins": 0.2246524840593338,
1184
+ "rewards/rejected": -2.818655490875244,
1185
+ "step": 830
1186
+ },
1187
+ {
1188
+ "epoch": 0.67,
1189
+ "learning_rate": 1.467238925438646e-06,
1190
+ "logits/chosen": -2.113966226577759,
1191
+ "logits/rejected": -1.9181636571884155,
1192
+ "logps/chosen": -503.37835693359375,
1193
+ "logps/rejected": -561.77783203125,
1194
+ "loss": 0.5414,
1195
+ "rewards/accuracies": 0.675000011920929,
1196
+ "rewards/chosen": -2.3552279472351074,
1197
+ "rewards/margins": 0.7212526202201843,
1198
+ "rewards/rejected": -3.0764806270599365,
1199
+ "step": 840
1200
+ },
1201
+ {
1202
+ "epoch": 0.68,
1203
+ "learning_rate": 1.4040721330273063e-06,
1204
+ "logits/chosen": -1.7894271612167358,
1205
+ "logits/rejected": -1.5007646083831787,
1206
+ "logps/chosen": -614.6292724609375,
1207
+ "logps/rejected": -649.1776123046875,
1208
+ "loss": 0.523,
1209
+ "rewards/accuracies": 0.75,
1210
+ "rewards/chosen": -3.184319019317627,
1211
+ "rewards/margins": 0.7777611613273621,
1212
+ "rewards/rejected": -3.9620795249938965,
1213
+ "step": 850
1214
+ },
1215
+ {
1216
+ "epoch": 0.69,
1217
+ "learning_rate": 1.3417599122003464e-06,
1218
+ "logits/chosen": -2.140821933746338,
1219
+ "logits/rejected": -1.7113529443740845,
1220
+ "logps/chosen": -523.3508911132812,
1221
+ "logps/rejected": -562.307861328125,
1222
+ "loss": 0.469,
1223
+ "rewards/accuracies": 0.675000011920929,
1224
+ "rewards/chosen": -2.6072592735290527,
1225
+ "rewards/margins": 1.0337148904800415,
1226
+ "rewards/rejected": -3.6409740447998047,
1227
+ "step": 860
1228
+ },
1229
+ {
1230
+ "epoch": 0.7,
1231
+ "learning_rate": 1.280350852153168e-06,
1232
+ "logits/chosen": -1.8685197830200195,
1233
+ "logits/rejected": -1.9930238723754883,
1234
+ "logps/chosen": -437.0240783691406,
1235
+ "logps/rejected": -565.9390258789062,
1236
+ "loss": 0.5039,
1237
+ "rewards/accuracies": 0.6499999761581421,
1238
+ "rewards/chosen": -2.464498996734619,
1239
+ "rewards/margins": 0.857032299041748,
1240
+ "rewards/rejected": -3.3215317726135254,
1241
+ "step": 870
1242
+ },
1243
+ {
1244
+ "epoch": 0.7,
1245
+ "learning_rate": 1.2198928378235717e-06,
1246
+ "logits/chosen": -1.9025375843048096,
1247
+ "logits/rejected": -1.7775294780731201,
1248
+ "logps/chosen": -559.2813720703125,
1249
+ "logps/rejected": -652.2202758789062,
1250
+ "loss": 0.5849,
1251
+ "rewards/accuracies": 0.675000011920929,
1252
+ "rewards/chosen": -3.108556032180786,
1253
+ "rewards/margins": 0.7565592527389526,
1254
+ "rewards/rejected": -3.8651154041290283,
1255
+ "step": 880
1256
+ },
1257
+ {
1258
+ "epoch": 0.71,
1259
+ "learning_rate": 1.160433012552508e-06,
1260
+ "logits/chosen": -2.1011109352111816,
1261
+ "logits/rejected": -2.0423381328582764,
1262
+ "logps/chosen": -430.05633544921875,
1263
+ "logps/rejected": -476.6094665527344,
1264
+ "loss": 0.5447,
1265
+ "rewards/accuracies": 0.6499999761581421,
1266
+ "rewards/chosen": -1.923673391342163,
1267
+ "rewards/margins": 0.9791573286056519,
1268
+ "rewards/rejected": -2.9028310775756836,
1269
+ "step": 890
1270
+ },
1271
+ {
1272
+ "epoch": 0.72,
1273
+ "learning_rate": 1.1020177413231334e-06,
1274
+ "logits/chosen": -2.0440664291381836,
1275
+ "logits/rejected": -1.8446366786956787,
1276
+ "logps/chosen": -435.1307067871094,
1277
+ "logps/rejected": -506.0008239746094,
1278
+ "loss": 0.5042,
1279
+ "rewards/accuracies": 0.675000011920929,
1280
+ "rewards/chosen": -2.0204269886016846,
1281
+ "rewards/margins": 1.0044559240341187,
1282
+ "rewards/rejected": -3.0248827934265137,
1283
+ "step": 900
1284
+ },
1285
+ {
1286
+ "epoch": 0.73,
1287
+ "learning_rate": 1.0446925746067768e-06,
1288
+ "logits/chosen": -1.8525664806365967,
1289
+ "logits/rejected": -1.5761830806732178,
1290
+ "logps/chosen": -413.330810546875,
1291
+ "logps/rejected": -453.98931884765625,
1292
+ "loss": 0.5014,
1293
+ "rewards/accuracies": 0.6000000238418579,
1294
+ "rewards/chosen": -1.897646188735962,
1295
+ "rewards/margins": 1.0005159378051758,
1296
+ "rewards/rejected": -2.8981618881225586,
1297
+ "step": 910
1298
+ },
1299
+ {
1300
+ "epoch": 0.74,
1301
+ "learning_rate": 9.88502212844063e-07,
1302
+ "logits/chosen": -1.9756139516830444,
1303
+ "logits/rejected": -1.8542131185531616,
1304
+ "logps/chosen": -528.9044799804688,
1305
+ "logps/rejected": -650.6089477539062,
1306
+ "loss": 0.525,
1307
+ "rewards/accuracies": 0.7749999761581421,
1308
+ "rewards/chosen": -2.553828716278076,
1309
+ "rewards/margins": 1.2451233863830566,
1310
+ "rewards/rejected": -3.798952102661133,
1311
+ "step": 920
1312
+ },
1313
+ {
1314
+ "epoch": 0.74,
1315
+ "learning_rate": 9.334904715888496e-07,
1316
+ "logits/chosen": -1.94060480594635,
1317
+ "logits/rejected": -1.6692876815795898,
1318
+ "logps/chosen": -567.3487548828125,
1319
+ "logps/rejected": -671.7532958984375,
1320
+ "loss": 0.5516,
1321
+ "rewards/accuracies": 0.675000011920929,
1322
+ "rewards/chosen": -2.8961987495422363,
1323
+ "rewards/margins": 1.0622972249984741,
1324
+ "rewards/rejected": -3.95849609375,
1325
+ "step": 930
1326
+ },
1327
+ {
1328
+ "epoch": 0.75,
1329
+ "learning_rate": 8.797002473421729e-07,
1330
+ "logits/chosen": -1.8798109292984009,
1331
+ "logits/rejected": -1.8840408325195312,
1332
+ "logps/chosen": -468.63287353515625,
1333
+ "logps/rejected": -541.5391845703125,
1334
+ "loss": 0.5102,
1335
+ "rewards/accuracies": 0.6000000238418579,
1336
+ "rewards/chosen": -2.1703548431396484,
1337
+ "rewards/margins": 0.741019606590271,
1338
+ "rewards/rejected": -2.91137433052063,
1339
+ "step": 940
1340
+ },
1341
+ {
1342
+ "epoch": 0.76,
1343
+ "learning_rate": 8.271734841028553e-07,
1344
+ "logits/chosen": -1.871678113937378,
1345
+ "logits/rejected": -1.6603199243545532,
1346
+ "logps/chosen": -525.9596557617188,
1347
+ "logps/rejected": -581.8895263671875,
1348
+ "loss": 0.4819,
1349
+ "rewards/accuracies": 0.7250000238418579,
1350
+ "rewards/chosen": -2.715925931930542,
1351
+ "rewards/margins": 0.8995075225830078,
1352
+ "rewards/rejected": -3.6154332160949707,
1353
+ "step": 950
1354
+ },
1355
+ {
1356
+ "epoch": 0.77,
1357
+ "learning_rate": 7.759511406608255e-07,
1358
+ "logits/chosen": -1.9854234457015991,
1359
+ "logits/rejected": -1.8719909191131592,
1360
+ "logps/chosen": -508.23785400390625,
1361
+ "logps/rejected": -571.8689575195312,
1362
+ "loss": 0.5553,
1363
+ "rewards/accuracies": 0.75,
1364
+ "rewards/chosen": -2.7327098846435547,
1365
+ "rewards/margins": 0.7255285978317261,
1366
+ "rewards/rejected": -3.4582386016845703,
1367
+ "step": 960
1368
+ },
1369
+ {
1370
+ "epoch": 0.78,
1371
+ "learning_rate": 7.260731586586983e-07,
1372
+ "logits/chosen": -1.753035545349121,
1373
+ "logits/rejected": -1.8223479986190796,
1374
+ "logps/chosen": -362.6646423339844,
1375
+ "logps/rejected": -436.86102294921875,
1376
+ "loss": 0.5738,
1377
+ "rewards/accuracies": 0.550000011920929,
1378
+ "rewards/chosen": -2.3551478385925293,
1379
+ "rewards/margins": 0.7355124950408936,
1380
+ "rewards/rejected": -3.090660572052002,
1381
+ "step": 970
1382
+ },
1383
+ {
1384
+ "epoch": 0.78,
1385
+ "learning_rate": 6.775784314464717e-07,
1386
+ "logits/chosen": -2.150531768798828,
1387
+ "logits/rejected": -1.819373369216919,
1388
+ "logps/chosen": -458.1045837402344,
1389
+ "logps/rejected": -522.545166015625,
1390
+ "loss": 0.491,
1391
+ "rewards/accuracies": 0.625,
1392
+ "rewards/chosen": -2.1924614906311035,
1393
+ "rewards/margins": 1.1207162141799927,
1394
+ "rewards/rejected": -3.3131778240203857,
1395
+ "step": 980
1396
+ },
1397
+ {
1398
+ "epoch": 0.79,
1399
+ "learning_rate": 6.305047737536707e-07,
1400
+ "logits/chosen": -1.841164231300354,
1401
+ "logits/rejected": -1.885084867477417,
1402
+ "logps/chosen": -514.1463623046875,
1403
+ "logps/rejected": -619.3209228515625,
1404
+ "loss": 0.4802,
1405
+ "rewards/accuracies": 0.7250000238418579,
1406
+ "rewards/chosen": -3.1367721557617188,
1407
+ "rewards/margins": 0.8921027183532715,
1408
+ "rewards/rejected": -4.028874397277832,
1409
+ "step": 990
1410
+ },
1411
+ {
1412
+ "epoch": 0.8,
1413
+ "learning_rate": 5.848888922025553e-07,
1414
+ "logits/chosen": -1.8358408212661743,
1415
+ "logits/rejected": -2.0529000759124756,
1416
+ "logps/chosen": -559.2037353515625,
1417
+ "logps/rejected": -580.3322143554688,
1418
+ "loss": 0.5311,
1419
+ "rewards/accuracies": 0.625,
1420
+ "rewards/chosen": -2.9438185691833496,
1421
+ "rewards/margins": 0.5294431447982788,
1422
+ "rewards/rejected": -3.473261594772339,
1423
+ "step": 1000
1424
+ },
1425
+ {
1426
+ "epoch": 0.81,
1427
+ "learning_rate": 5.407663566854008e-07,
1428
+ "logits/chosen": -2.139596462249756,
1429
+ "logits/rejected": -1.9000869989395142,
1430
+ "logps/chosen": -503.7278747558594,
1431
+ "logps/rejected": -551.3563842773438,
1432
+ "loss": 0.577,
1433
+ "rewards/accuracies": 0.6499999761581421,
1434
+ "rewards/chosen": -2.439404010772705,
1435
+ "rewards/margins": 0.9661655426025391,
1436
+ "rewards/rejected": -3.405569553375244,
1437
+ "step": 1010
1438
+ },
1439
+ {
1440
+ "epoch": 0.82,
1441
+ "learning_rate": 4.981715726281666e-07,
1442
+ "logits/chosen": -2.013249397277832,
1443
+ "logits/rejected": -1.9738517999649048,
1444
+ "logps/chosen": -589.2969970703125,
1445
+ "logps/rejected": -724.22998046875,
1446
+ "loss": 0.5035,
1447
+ "rewards/accuracies": 0.675000011920929,
1448
+ "rewards/chosen": -3.0665476322174072,
1449
+ "rewards/margins": 1.2686010599136353,
1450
+ "rewards/rejected": -4.335148811340332,
1451
+ "step": 1020
1452
+ },
1453
+ {
1454
+ "epoch": 0.82,
1455
+ "learning_rate": 4.5713775416217884e-07,
1456
+ "logits/chosen": -1.6897433996200562,
1457
+ "logits/rejected": -1.7765251398086548,
1458
+ "logps/chosen": -498.42156982421875,
1459
+ "logps/rejected": -617.593994140625,
1460
+ "loss": 0.5346,
1461
+ "rewards/accuracies": 0.675000011920929,
1462
+ "rewards/chosen": -3.009658098220825,
1463
+ "rewards/margins": 1.1272680759429932,
1464
+ "rewards/rejected": -4.136926174163818,
1465
+ "step": 1030
1466
+ },
1467
+ {
1468
+ "epoch": 0.83,
1469
+ "learning_rate": 4.1769689822475147e-07,
1470
+ "logits/chosen": -1.6748046875,
1471
+ "logits/rejected": -1.8121490478515625,
1472
+ "logps/chosen": -465.87432861328125,
1473
+ "logps/rejected": -581.8338623046875,
1474
+ "loss": 0.583,
1475
+ "rewards/accuracies": 0.6499999761581421,
1476
+ "rewards/chosen": -3.0733983516693115,
1477
+ "rewards/margins": 0.6260702013969421,
1478
+ "rewards/rejected": -3.6994686126708984,
1479
+ "step": 1040
1480
+ },
1481
+ {
1482
+ "epoch": 0.84,
1483
+ "learning_rate": 3.798797596089351e-07,
1484
+ "logits/chosen": -1.9456278085708618,
1485
+ "logits/rejected": -1.7044868469238281,
1486
+ "logps/chosen": -526.3280639648438,
1487
+ "logps/rejected": -614.6798095703125,
1488
+ "loss": 0.4783,
1489
+ "rewards/accuracies": 0.800000011920929,
1490
+ "rewards/chosen": -2.9347662925720215,
1491
+ "rewards/margins": 1.119192123413086,
1492
+ "rewards/rejected": -4.053957939147949,
1493
+ "step": 1050
1494
+ },
1495
+ {
1496
+ "epoch": 0.85,
1497
+ "learning_rate": 3.4371582698185636e-07,
1498
+ "logits/chosen": -2.0055367946624756,
1499
+ "logits/rejected": -1.8538860082626343,
1500
+ "logps/chosen": -615.2640380859375,
1501
+ "logps/rejected": -729.064453125,
1502
+ "loss": 0.4626,
1503
+ "rewards/accuracies": 0.7250000238418579,
1504
+ "rewards/chosen": -3.3482022285461426,
1505
+ "rewards/margins": 0.9904724955558777,
1506
+ "rewards/rejected": -4.338675022125244,
1507
+ "step": 1060
1508
+ },
1509
+ {
1510
+ "epoch": 0.86,
1511
+ "learning_rate": 3.092332998903416e-07,
1512
+ "logits/chosen": -1.9933702945709229,
1513
+ "logits/rejected": -2.050975799560547,
1514
+ "logps/chosen": -535.1612548828125,
1515
+ "logps/rejected": -645.9298706054688,
1516
+ "loss": 0.4793,
1517
+ "rewards/accuracies": 0.675000011920929,
1518
+ "rewards/chosen": -2.8539366722106934,
1519
+ "rewards/margins": 1.0242040157318115,
1520
+ "rewards/rejected": -3.878140926361084,
1521
+ "step": 1070
1522
+ },
1523
+ {
1524
+ "epoch": 0.86,
1525
+ "learning_rate": 2.764590667717562e-07,
1526
+ "logits/chosen": -1.8224666118621826,
1527
+ "logits/rejected": -1.5814918279647827,
1528
+ "logps/chosen": -508.09332275390625,
1529
+ "logps/rejected": -499.42742919921875,
1530
+ "loss": 0.5644,
1531
+ "rewards/accuracies": 0.4000000059604645,
1532
+ "rewards/chosen": -3.324652910232544,
1533
+ "rewards/margins": 0.2990475594997406,
1534
+ "rewards/rejected": -3.6237003803253174,
1535
+ "step": 1080
1536
+ },
1537
+ {
1538
+ "epoch": 0.87,
1539
+ "learning_rate": 2.454186839872158e-07,
1540
+ "logits/chosen": -1.8880449533462524,
1541
+ "logits/rejected": -1.7552686929702759,
1542
+ "logps/chosen": -522.8394775390625,
1543
+ "logps/rejected": -642.8104858398438,
1544
+ "loss": 0.5033,
1545
+ "rewards/accuracies": 0.75,
1546
+ "rewards/chosen": -3.2225208282470703,
1547
+ "rewards/margins": 1.2935682535171509,
1548
+ "rewards/rejected": -4.51608943939209,
1549
+ "step": 1090
1550
+ },
1551
+ {
1552
+ "epoch": 0.88,
1553
+ "learning_rate": 2.1613635589349756e-07,
1554
+ "logits/chosen": -1.9811900854110718,
1555
+ "logits/rejected": -1.7504934072494507,
1556
+ "logps/chosen": -492.2660217285156,
1557
+ "logps/rejected": -541.9775390625,
1558
+ "loss": 0.5612,
1559
+ "rewards/accuracies": 0.550000011920929,
1560
+ "rewards/chosen": -2.6576216220855713,
1561
+ "rewards/margins": 0.7241417169570923,
1562
+ "rewards/rejected": -3.381763458251953,
1563
+ "step": 1100
1564
+ },
1565
+ {
1566
+ "epoch": 0.89,
1567
+ "learning_rate": 1.8863491596921745e-07,
1568
+ "logits/chosen": -1.9917253255844116,
1569
+ "logits/rejected": -1.6974796056747437,
1570
+ "logps/chosen": -522.3150634765625,
1571
+ "logps/rejected": -547.88330078125,
1572
+ "loss": 0.4857,
1573
+ "rewards/accuracies": 0.6499999761581421,
1574
+ "rewards/chosen": -2.875357151031494,
1575
+ "rewards/margins": 0.8617655038833618,
1576
+ "rewards/rejected": -3.7371230125427246,
1577
+ "step": 1110
1578
+ },
1579
+ {
1580
+ "epoch": 0.9,
1581
+ "learning_rate": 1.629358090099639e-07,
1582
+ "logits/chosen": -1.8854055404663086,
1583
+ "logits/rejected": -1.7443320751190186,
1584
+ "logps/chosen": -544.697509765625,
1585
+ "logps/rejected": -587.5208129882812,
1586
+ "loss": 0.5281,
1587
+ "rewards/accuracies": 0.699999988079071,
1588
+ "rewards/chosen": -3.345867156982422,
1589
+ "rewards/margins": 0.8077778816223145,
1590
+ "rewards/rejected": -4.1536455154418945,
1591
+ "step": 1120
1592
+ },
1593
+ {
1594
+ "epoch": 0.9,
1595
+ "learning_rate": 1.3905907440629752e-07,
1596
+ "logits/chosen": -1.8521480560302734,
1597
+ "logits/rejected": -1.633465051651001,
1598
+ "logps/chosen": -494.1001892089844,
1599
+ "logps/rejected": -566.96826171875,
1600
+ "loss": 0.5509,
1601
+ "rewards/accuracies": 0.625,
1602
+ "rewards/chosen": -2.8628900051116943,
1603
+ "rewards/margins": 1.034968614578247,
1604
+ "rewards/rejected": -3.8978583812713623,
1605
+ "step": 1130
1606
+ },
1607
+ {
1608
+ "epoch": 0.91,
1609
+ "learning_rate": 1.1702333051763271e-07,
1610
+ "logits/chosen": -1.7924169301986694,
1611
+ "logits/rejected": -1.5471246242523193,
1612
+ "logps/chosen": -587.1184692382812,
1613
+ "logps/rejected": -650.1336669921875,
1614
+ "loss": 0.4997,
1615
+ "rewards/accuracies": 0.675000011920929,
1616
+ "rewards/chosen": -3.2694034576416016,
1617
+ "rewards/margins": 1.0973501205444336,
1618
+ "rewards/rejected": -4.366753101348877,
1619
+ "step": 1140
1620
+ },
1621
+ {
1622
+ "epoch": 0.92,
1623
+ "learning_rate": 9.684576015420277e-08,
1624
+ "logits/chosen": -1.751611351966858,
1625
+ "logits/rejected": -1.576918363571167,
1626
+ "logps/chosen": -530.2127075195312,
1627
+ "logps/rejected": -614.9729614257812,
1628
+ "loss": 0.4732,
1629
+ "rewards/accuracies": 0.6000000238418579,
1630
+ "rewards/chosen": -2.8580543994903564,
1631
+ "rewards/margins": 0.9249665141105652,
1632
+ "rewards/rejected": -3.7830207347869873,
1633
+ "step": 1150
1634
+ },
1635
+ {
1636
+ "epoch": 0.93,
1637
+ "learning_rate": 7.854209717842231e-08,
1638
+ "logits/chosen": -1.7469266653060913,
1639
+ "logits/rejected": -1.668701171875,
1640
+ "logps/chosen": -585.7467041015625,
1641
+ "logps/rejected": -588.47509765625,
1642
+ "loss": 0.5739,
1643
+ "rewards/accuracies": 0.574999988079071,
1644
+ "rewards/chosen": -3.5444464683532715,
1645
+ "rewards/margins": 0.3376322388648987,
1646
+ "rewards/rejected": -3.8820786476135254,
1647
+ "step": 1160
1648
+ },
1649
+ {
1650
+ "epoch": 0.94,
1651
+ "learning_rate": 6.212661423609184e-08,
1652
+ "logits/chosen": -1.7435801029205322,
1653
+ "logits/rejected": -1.680567741394043,
1654
+ "logps/chosen": -615.296142578125,
1655
+ "logps/rejected": -718.3519287109375,
1656
+ "loss": 0.5697,
1657
+ "rewards/accuracies": 0.625,
1658
+ "rewards/chosen": -4.011544704437256,
1659
+ "rewards/margins": 0.8902410268783569,
1660
+ "rewards/rejected": -4.901785850524902,
1661
+ "step": 1170
1662
+ },
1663
+ {
1664
+ "epoch": 0.94,
1665
+ "learning_rate": 4.761211162702117e-08,
1666
+ "logits/chosen": -1.855932593345642,
1667
+ "logits/rejected": -1.8077094554901123,
1668
+ "logps/chosen": -531.5975952148438,
1669
+ "logps/rejected": -581.9424438476562,
1670
+ "loss": 0.5434,
1671
+ "rewards/accuracies": 0.675000011920929,
1672
+ "rewards/chosen": -3.0784988403320312,
1673
+ "rewards/margins": 0.6800780892372131,
1674
+ "rewards/rejected": -3.7585768699645996,
1675
+ "step": 1180
1676
+ },
1677
+ {
1678
+ "epoch": 0.95,
1679
+ "learning_rate": 3.5009907323737826e-08,
1680
+ "logits/chosen": -1.838123083114624,
1681
+ "logits/rejected": -1.5591208934783936,
1682
+ "logps/chosen": -556.9913330078125,
1683
+ "logps/rejected": -647.0695190429688,
1684
+ "loss": 0.504,
1685
+ "rewards/accuracies": 0.699999988079071,
1686
+ "rewards/chosen": -3.241734027862549,
1687
+ "rewards/margins": 1.0456501245498657,
1688
+ "rewards/rejected": -4.287384033203125,
1689
+ "step": 1190
1690
+ },
1691
+ {
1692
+ "epoch": 0.96,
1693
+ "learning_rate": 2.4329828146074096e-08,
1694
+ "logits/chosen": -2.1346633434295654,
1695
+ "logits/rejected": -1.826059341430664,
1696
+ "logps/chosen": -596.9236450195312,
1697
+ "logps/rejected": -622.0565185546875,
1698
+ "loss": 0.5305,
1699
+ "rewards/accuracies": 0.6499999761581421,
1700
+ "rewards/chosen": -3.315636396408081,
1701
+ "rewards/margins": 0.8735532760620117,
1702
+ "rewards/rejected": -4.189189910888672,
1703
+ "step": 1200
1704
+ },
1705
+ {
1706
+ "epoch": 0.97,
1707
+ "learning_rate": 1.5580202098509078e-08,
1708
+ "logits/chosen": -1.944238305091858,
1709
+ "logits/rejected": -1.891898512840271,
1710
+ "logps/chosen": -647.6207275390625,
1711
+ "logps/rejected": -664.7337036132812,
1712
+ "loss": 0.5806,
1713
+ "rewards/accuracies": 0.6000000238418579,
1714
+ "rewards/chosen": -3.0909292697906494,
1715
+ "rewards/margins": 0.758102536201477,
1716
+ "rewards/rejected": -3.849031448364258,
1717
+ "step": 1210
1718
+ },
1719
+ {
1720
+ "epoch": 0.98,
1721
+ "learning_rate": 8.767851876239075e-09,
1722
+ "logits/chosen": -1.6966060400009155,
1723
+ "logits/rejected": -1.641953706741333,
1724
+ "logps/chosen": -548.3276977539062,
1725
+ "logps/rejected": -660.7442626953125,
1726
+ "loss": 0.5179,
1727
+ "rewards/accuracies": 0.675000011920929,
1728
+ "rewards/chosen": -3.2273573875427246,
1729
+ "rewards/margins": 1.0470120906829834,
1730
+ "rewards/rejected": -4.274369239807129,
1731
+ "step": 1220
1732
+ },
1733
+ {
1734
+ "epoch": 0.98,
1735
+ "learning_rate": 3.8980895450474455e-09,
1736
+ "logits/chosen": -1.908227562904358,
1737
+ "logits/rejected": -1.7419946193695068,
1738
+ "logps/chosen": -513.4429931640625,
1739
+ "logps/rejected": -574.7293701171875,
1740
+ "loss": 0.5757,
1741
+ "rewards/accuracies": 0.6000000238418579,
1742
+ "rewards/chosen": -2.947870969772339,
1743
+ "rewards/margins": 0.8125208616256714,
1744
+ "rewards/rejected": -3.7603917121887207,
1745
+ "step": 1230
1746
+ },
1747
+ {
1748
+ "epoch": 0.99,
1749
+ "learning_rate": 9.747123991141193e-10,
1750
+ "logits/chosen": -2.0077996253967285,
1751
+ "logits/rejected": -1.8994073867797852,
1752
+ "logps/chosen": -504.7936096191406,
1753
+ "logps/rejected": -594.3426513671875,
1754
+ "loss": 0.488,
1755
+ "rewards/accuracies": 0.6499999761581421,
1756
+ "rewards/chosen": -2.80145001411438,
1757
+ "rewards/margins": 0.9755638241767883,
1758
+ "rewards/rejected": -3.7770137786865234,
1759
+ "step": 1240
1760
+ },
1761
+ {
1762
+ "epoch": 1.0,
1763
+ "learning_rate": 0.0,
1764
+ "logits/chosen": -1.8658443689346313,
1765
+ "logits/rejected": -1.699480414390564,
1766
+ "logps/chosen": -557.4786987304688,
1767
+ "logps/rejected": -628.1792602539062,
1768
+ "loss": 0.4848,
1769
+ "rewards/accuracies": 0.574999988079071,
1770
+ "rewards/chosen": -3.1368298530578613,
1771
+ "rewards/margins": 0.8951441049575806,
1772
+ "rewards/rejected": -4.031973838806152,
1773
+ "step": 1250
1774
+ },
1775
+ {
1776
+ "epoch": 1.0,
1777
+ "step": 1250,
1778
+ "total_flos": 0.0,
1779
+ "train_loss": 0.5286114639282227,
1780
+ "train_runtime": 12048.4045,
1781
+ "train_samples_per_second": 1.245,
1782
+ "train_steps_per_second": 0.104
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
+ }