BraylonDash commited on
Commit
fd70a54
1 Parent(s): 6d077d3

Model save

Browse files
README.md ADDED
@@ -0,0 +1,62 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ library_name: peft
4
+ tags:
5
+ - trl
6
+ - dpo
7
+ - generated_from_trainer
8
+ base_model: DUAL-GPO/zephyr-7b-ipo-qlora-v0-merged
9
+ model-index:
10
+ - name: zephyr-7b-ipo-10k-40k-0.001-i1
11
+ results: []
12
+ ---
13
+
14
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
15
+ should probably proofread and complete it, then remove this comment. -->
16
+
17
+ # zephyr-7b-ipo-10k-40k-0.001-i1
18
+
19
+ This model is a fine-tuned version of [DUAL-GPO/zephyr-7b-ipo-qlora-v0-merged](https://huggingface.co/DUAL-GPO/zephyr-7b-ipo-qlora-v0-merged) on the None dataset.
20
+
21
+ ## Model description
22
+
23
+ More information needed
24
+
25
+ ## Intended uses & limitations
26
+
27
+ More information needed
28
+
29
+ ## Training and evaluation data
30
+
31
+ More information needed
32
+
33
+ ## Training procedure
34
+
35
+ ### Training hyperparameters
36
+
37
+ The following hyperparameters were used during training:
38
+ - learning_rate: 5e-06
39
+ - train_batch_size: 2
40
+ - eval_batch_size: 2
41
+ - seed: 42
42
+ - distributed_type: multi-GPU
43
+ - num_devices: 3
44
+ - gradient_accumulation_steps: 2
45
+ - total_train_batch_size: 12
46
+ - total_eval_batch_size: 6
47
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
48
+ - lr_scheduler_type: cosine
49
+ - lr_scheduler_warmup_ratio: 0.1
50
+ - num_epochs: 1
51
+
52
+ ### Training results
53
+
54
+
55
+
56
+ ### Framework versions
57
+
58
+ - PEFT 0.7.1
59
+ - Transformers 4.36.2
60
+ - Pytorch 2.1.2
61
+ - Datasets 2.14.6
62
+ - Tokenizers 0.15.2
adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:0fb18c0f6f6ccf32b9487d6299e4e4816ecf539986b48d7a5a58bfc14283a9d7
3
  size 671150064
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:844ea2ebc51e77b93383822d90d9d87ed76fcce16065406abc98d2b243647732
3
  size 671150064
all_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "train_loss": 0.5607085178375244,
4
+ "train_runtime": 13233.1545,
5
+ "train_samples": 15000,
6
+ "train_samples_per_second": 1.134,
7
+ "train_steps_per_second": 0.094
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
+ 2024-09-23T18:36:14,codecarbon,c894b01b-0bad-4623-9311-5f96ad455656,13233.158681154251,0.007578821270492903,5.727144556413531e-07,42.5,928.96,188.74309015274048,0.15621936400930078,2.342175038183366,0.6897100193269589,3.1881044215196224,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/Sep23_14-54-42_gpu4-119-5/events.out.tfevents.1727067341.gpu4-119-5.3752817.0 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:d4011ee703179deb9398c22c521a795f078aa11c9e9591cd29589c96802a12e5
3
- size 83738
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:df8852d006524e5206cce5cbe706e66c135ec397b94b77da188c7ff927ab7f13
3
+ size 84726
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "train_loss": 0.5607085178375244,
4
+ "train_runtime": 13233.1545,
5
+ "train_samples": 15000,
6
+ "train_samples_per_second": 1.134,
7
+ "train_steps_per_second": 0.094
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": -3.1403026580810547,
15
+ "logits/rejected": -2.4295239448547363,
16
+ "logps/chosen": -392.2955017089844,
17
+ "logps/rejected": -613.5740966796875,
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": -2.442929267883301,
29
+ "logits/rejected": -2.321220636367798,
30
+ "logps/chosen": -261.98870849609375,
31
+ "logps/rejected": -221.4100799560547,
32
+ "loss": 0.6931,
33
+ "rewards/accuracies": 0.4722222089767456,
34
+ "rewards/chosen": 7.1744441811461e-05,
35
+ "rewards/margins": 5.5226653785211965e-05,
36
+ "rewards/rejected": 1.6517798940185457e-05,
37
+ "step": 10
38
+ },
39
+ {
40
+ "epoch": 0.02,
41
+ "learning_rate": 8.000000000000001e-07,
42
+ "logits/chosen": -2.42722225189209,
43
+ "logits/rejected": -2.285809278488159,
44
+ "logps/chosen": -244.7586212158203,
45
+ "logps/rejected": -254.1322784423828,
46
+ "loss": 0.6931,
47
+ "rewards/accuracies": 0.5249999761581421,
48
+ "rewards/chosen": -4.93793522764463e-05,
49
+ "rewards/margins": 0.00013055391900707036,
50
+ "rewards/rejected": -0.00017993326764553785,
51
+ "step": 20
52
+ },
53
+ {
54
+ "epoch": 0.02,
55
+ "learning_rate": 1.2000000000000002e-06,
56
+ "logits/chosen": -2.468625068664551,
57
+ "logits/rejected": -2.3896145820617676,
58
+ "logps/chosen": -251.3707733154297,
59
+ "logps/rejected": -300.4630126953125,
60
+ "loss": 0.6931,
61
+ "rewards/accuracies": 0.5249999761581421,
62
+ "rewards/chosen": 0.00025913942954503,
63
+ "rewards/margins": 9.763217531144619e-05,
64
+ "rewards/rejected": 0.00016150726878549904,
65
+ "step": 30
66
+ },
67
+ {
68
+ "epoch": 0.03,
69
+ "learning_rate": 1.6000000000000001e-06,
70
+ "logits/chosen": -2.483097553253174,
71
+ "logits/rejected": -2.135282516479492,
72
+ "logps/chosen": -252.2518310546875,
73
+ "logps/rejected": -206.49740600585938,
74
+ "loss": 0.693,
75
+ "rewards/accuracies": 0.574999988079071,
76
+ "rewards/chosen": 0.0007439021137543023,
77
+ "rewards/margins": 0.0003426038019824773,
78
+ "rewards/rejected": 0.00040129836997948587,
79
+ "step": 40
80
+ },
81
+ {
82
+ "epoch": 0.04,
83
+ "learning_rate": 2.0000000000000003e-06,
84
+ "logits/chosen": -2.6477389335632324,
85
+ "logits/rejected": -2.486539363861084,
86
+ "logps/chosen": -245.1098175048828,
87
+ "logps/rejected": -234.5915985107422,
88
+ "loss": 0.6928,
89
+ "rewards/accuracies": 0.550000011920929,
90
+ "rewards/chosen": 0.0010015963343903422,
91
+ "rewards/margins": 0.00047645685845054686,
92
+ "rewards/rejected": 0.0005251394468359649,
93
+ "step": 50
94
+ },
95
+ {
96
+ "epoch": 0.05,
97
+ "learning_rate": 2.4000000000000003e-06,
98
+ "logits/chosen": -2.5867533683776855,
99
+ "logits/rejected": -1.8461745977401733,
100
+ "logps/chosen": -288.46343994140625,
101
+ "logps/rejected": -191.20741271972656,
102
+ "loss": 0.6924,
103
+ "rewards/accuracies": 0.699999988079071,
104
+ "rewards/chosen": 0.0025676097720861435,
105
+ "rewards/margins": 0.001473221811465919,
106
+ "rewards/rejected": 0.0010943879606202245,
107
+ "step": 60
108
+ },
109
+ {
110
+ "epoch": 0.06,
111
+ "learning_rate": 2.8000000000000003e-06,
112
+ "logits/chosen": -2.4274649620056152,
113
+ "logits/rejected": -2.21500825881958,
114
+ "logps/chosen": -206.21188354492188,
115
+ "logps/rejected": -188.66172790527344,
116
+ "loss": 0.6918,
117
+ "rewards/accuracies": 0.574999988079071,
118
+ "rewards/chosen": 0.004373534582555294,
119
+ "rewards/margins": 0.0019022986525669694,
120
+ "rewards/rejected": 0.0024712360464036465,
121
+ "step": 70
122
+ },
123
+ {
124
+ "epoch": 0.06,
125
+ "learning_rate": 3.2000000000000003e-06,
126
+ "logits/chosen": -2.272698163986206,
127
+ "logits/rejected": -2.2475826740264893,
128
+ "logps/chosen": -218.1708984375,
129
+ "logps/rejected": -247.9695281982422,
130
+ "loss": 0.6919,
131
+ "rewards/accuracies": 0.550000011920929,
132
+ "rewards/chosen": 0.00605297926813364,
133
+ "rewards/margins": 0.002590904710814357,
134
+ "rewards/rejected": 0.0034620738588273525,
135
+ "step": 80
136
+ },
137
+ {
138
+ "epoch": 0.07,
139
+ "learning_rate": 3.6000000000000003e-06,
140
+ "logits/chosen": -2.5878186225891113,
141
+ "logits/rejected": -2.174198627471924,
142
+ "logps/chosen": -284.7269287109375,
143
+ "logps/rejected": -222.1537322998047,
144
+ "loss": 0.6908,
145
+ "rewards/accuracies": 0.5249999761581421,
146
+ "rewards/chosen": 0.00458324421197176,
147
+ "rewards/margins": 0.005124131683260202,
148
+ "rewards/rejected": -0.0005408872966654599,
149
+ "step": 90
150
+ },
151
+ {
152
+ "epoch": 0.08,
153
+ "learning_rate": 4.000000000000001e-06,
154
+ "logits/chosen": -2.4876809120178223,
155
+ "logits/rejected": -2.304762363433838,
156
+ "logps/chosen": -258.6788330078125,
157
+ "logps/rejected": -230.6868133544922,
158
+ "loss": 0.6878,
159
+ "rewards/accuracies": 0.6499999761581421,
160
+ "rewards/chosen": 0.0009092213585972786,
161
+ "rewards/margins": 0.013424595817923546,
162
+ "rewards/rejected": -0.012515375390648842,
163
+ "step": 100
164
+ },
165
+ {
166
+ "epoch": 0.09,
167
+ "learning_rate": 4.4e-06,
168
+ "logits/chosen": -2.24761700630188,
169
+ "logits/rejected": -2.0283989906311035,
170
+ "logps/chosen": -242.88485717773438,
171
+ "logps/rejected": -253.8578643798828,
172
+ "loss": 0.6832,
173
+ "rewards/accuracies": 0.625,
174
+ "rewards/chosen": -0.02561577782034874,
175
+ "rewards/margins": 0.02318538725376129,
176
+ "rewards/rejected": -0.04880116507411003,
177
+ "step": 110
178
+ },
179
+ {
180
+ "epoch": 0.1,
181
+ "learning_rate": 4.800000000000001e-06,
182
+ "logits/chosen": -2.0988376140594482,
183
+ "logits/rejected": -2.193089008331299,
184
+ "logps/chosen": -242.8556365966797,
185
+ "logps/rejected": -317.71917724609375,
186
+ "loss": 0.6753,
187
+ "rewards/accuracies": 0.6000000238418579,
188
+ "rewards/chosen": -0.047444917261600494,
189
+ "rewards/margins": 0.03924765810370445,
190
+ "rewards/rejected": -0.08669257164001465,
191
+ "step": 120
192
+ },
193
+ {
194
+ "epoch": 0.1,
195
+ "learning_rate": 4.999756310023261e-06,
196
+ "logits/chosen": -2.241037607192993,
197
+ "logits/rejected": -2.157954692840576,
198
+ "logps/chosen": -331.423828125,
199
+ "logps/rejected": -348.91046142578125,
200
+ "loss": 0.6688,
201
+ "rewards/accuracies": 0.6000000238418579,
202
+ "rewards/chosen": -0.08836604654788971,
203
+ "rewards/margins": 0.0334697850048542,
204
+ "rewards/rejected": -0.12183582782745361,
205
+ "step": 130
206
+ },
207
+ {
208
+ "epoch": 0.11,
209
+ "learning_rate": 4.997807075247147e-06,
210
+ "logits/chosen": -1.7352321147918701,
211
+ "logits/rejected": -1.5876859426498413,
212
+ "logps/chosen": -398.1210021972656,
213
+ "logps/rejected": -428.30328369140625,
214
+ "loss": 0.6519,
215
+ "rewards/accuracies": 0.44999998807907104,
216
+ "rewards/chosen": -0.22779341042041779,
217
+ "rewards/margins": 0.0445188470184803,
218
+ "rewards/rejected": -0.2723122239112854,
219
+ "step": 140
220
+ },
221
+ {
222
+ "epoch": 0.12,
223
+ "learning_rate": 4.993910125649561e-06,
224
+ "logits/chosen": -1.735691785812378,
225
+ "logits/rejected": -1.7642030715942383,
226
+ "logps/chosen": -525.7499389648438,
227
+ "logps/rejected": -695.8289184570312,
228
+ "loss": 0.6354,
229
+ "rewards/accuracies": 0.5249999761581421,
230
+ "rewards/chosen": -0.30290618538856506,
231
+ "rewards/margins": 0.11975699663162231,
232
+ "rewards/rejected": -0.422663152217865,
233
+ "step": 150
234
+ },
235
+ {
236
+ "epoch": 0.13,
237
+ "learning_rate": 4.988068499954578e-06,
238
+ "logits/chosen": -2.004002332687378,
239
+ "logits/rejected": -1.7766345739364624,
240
+ "logps/chosen": -464.7952575683594,
241
+ "logps/rejected": -630.8657836914062,
242
+ "loss": 0.6461,
243
+ "rewards/accuracies": 0.625,
244
+ "rewards/chosen": -0.23786354064941406,
245
+ "rewards/margins": 0.18317463994026184,
246
+ "rewards/rejected": -0.4210381507873535,
247
+ "step": 160
248
+ },
249
+ {
250
+ "epoch": 0.14,
251
+ "learning_rate": 4.980286753286196e-06,
252
+ "logits/chosen": -2.0650768280029297,
253
+ "logits/rejected": -1.572537899017334,
254
+ "logps/chosen": -591.57421875,
255
+ "logps/rejected": -780.8646240234375,
256
+ "loss": 0.607,
257
+ "rewards/accuracies": 0.699999988079071,
258
+ "rewards/chosen": -0.2891489565372467,
259
+ "rewards/margins": 0.27657198905944824,
260
+ "rewards/rejected": -0.5657209157943726,
261
+ "step": 170
262
+ },
263
+ {
264
+ "epoch": 0.14,
265
+ "learning_rate": 4.970570953616383e-06,
266
+ "logits/chosen": -1.4938254356384277,
267
+ "logits/rejected": -1.661007285118103,
268
+ "logps/chosen": -764.5491943359375,
269
+ "logps/rejected": -1000.9505615234375,
270
+ "loss": 0.6384,
271
+ "rewards/accuracies": 0.574999988079071,
272
+ "rewards/chosen": -0.5677521228790283,
273
+ "rewards/margins": 0.19688071310520172,
274
+ "rewards/rejected": -0.7646328806877136,
275
+ "step": 180
276
+ },
277
+ {
278
+ "epoch": 0.15,
279
+ "learning_rate": 4.958928677033465e-06,
280
+ "logits/chosen": -1.6295101642608643,
281
+ "logits/rejected": -1.3617414236068726,
282
+ "logps/chosen": -720.4210815429688,
283
+ "logps/rejected": -925.54638671875,
284
+ "loss": 0.6042,
285
+ "rewards/accuracies": 0.625,
286
+ "rewards/chosen": -0.4372948706150055,
287
+ "rewards/margins": 0.2413189709186554,
288
+ "rewards/rejected": -0.6786138415336609,
289
+ "step": 190
290
+ },
291
+ {
292
+ "epoch": 0.16,
293
+ "learning_rate": 4.9453690018345144e-06,
294
+ "logits/chosen": -1.502726674079895,
295
+ "logits/rejected": -1.3334269523620605,
296
+ "logps/chosen": -752.0069580078125,
297
+ "logps/rejected": -1068.231201171875,
298
+ "loss": 0.64,
299
+ "rewards/accuracies": 0.574999988079071,
300
+ "rewards/chosen": -0.5011454820632935,
301
+ "rewards/margins": 0.2965936064720154,
302
+ "rewards/rejected": -0.7977390289306641,
303
+ "step": 200
304
+ },
305
+ {
306
+ "epoch": 0.17,
307
+ "learning_rate": 4.9299025014463665e-06,
308
+ "logits/chosen": -1.1884076595306396,
309
+ "logits/rejected": -1.0619341135025024,
310
+ "logps/chosen": -890.1341552734375,
311
+ "logps/rejected": -1238.742431640625,
312
+ "loss": 0.6007,
313
+ "rewards/accuracies": 0.7250000238418579,
314
+ "rewards/chosen": -0.6316002607345581,
315
+ "rewards/margins": 0.3692142069339752,
316
+ "rewards/rejected": -1.000814437866211,
317
+ "step": 210
318
+ },
319
+ {
320
+ "epoch": 0.18,
321
+ "learning_rate": 4.912541236180779e-06,
322
+ "logits/chosen": -1.0146244764328003,
323
+ "logits/rejected": -0.9316056966781616,
324
+ "logps/chosen": -979.2394409179688,
325
+ "logps/rejected": -1465.3270263671875,
326
+ "loss": 0.6018,
327
+ "rewards/accuracies": 0.7250000238418579,
328
+ "rewards/chosen": -0.7414819002151489,
329
+ "rewards/margins": 0.49276018142700195,
330
+ "rewards/rejected": -1.2342422008514404,
331
+ "step": 220
332
+ },
333
+ {
334
+ "epoch": 0.18,
335
+ "learning_rate": 4.893298743830168e-06,
336
+ "logits/chosen": -1.2160202264785767,
337
+ "logits/rejected": -1.0113346576690674,
338
+ "logps/chosen": -1194.5838623046875,
339
+ "logps/rejected": -1710.5823974609375,
340
+ "loss": 0.5866,
341
+ "rewards/accuracies": 0.699999988079071,
342
+ "rewards/chosen": -0.9506030082702637,
343
+ "rewards/margins": 0.5406419634819031,
344
+ "rewards/rejected": -1.491244912147522,
345
+ "step": 230
346
+ },
347
+ {
348
+ "epoch": 0.19,
349
+ "learning_rate": 4.8721900291112415e-06,
350
+ "logits/chosen": -1.7798312902450562,
351
+ "logits/rejected": -1.6643425226211548,
352
+ "logps/chosen": -1057.5977783203125,
353
+ "logps/rejected": -1005.8741455078125,
354
+ "loss": 0.694,
355
+ "rewards/accuracies": 0.699999988079071,
356
+ "rewards/chosen": -0.7734619379043579,
357
+ "rewards/margins": -0.0247958954423666,
358
+ "rewards/rejected": -0.7486661076545715,
359
+ "step": 240
360
+ },
361
+ {
362
+ "epoch": 0.2,
363
+ "learning_rate": 4.849231551964771e-06,
364
+ "logits/chosen": -2.0089643001556396,
365
+ "logits/rejected": -1.7859022617340088,
366
+ "logps/chosen": -399.1302795410156,
367
+ "logps/rejected": -577.6984252929688,
368
+ "loss": 0.6424,
369
+ "rewards/accuracies": 0.6499999761581421,
370
+ "rewards/chosen": -0.1754738688468933,
371
+ "rewards/margins": 0.13275322318077087,
372
+ "rewards/rejected": -0.3082270920276642,
373
+ "step": 250
374
+ },
375
+ {
376
+ "epoch": 0.21,
377
+ "learning_rate": 4.824441214720629e-06,
378
+ "logits/chosen": -1.7344192266464233,
379
+ "logits/rejected": -1.632328748703003,
380
+ "logps/chosen": -632.4493408203125,
381
+ "logps/rejected": -760.2273559570312,
382
+ "loss": 0.6252,
383
+ "rewards/accuracies": 0.675000011920929,
384
+ "rewards/chosen": -0.41359105706214905,
385
+ "rewards/margins": 0.12629616260528564,
386
+ "rewards/rejected": -0.5398872494697571,
387
+ "step": 260
388
+ },
389
+ {
390
+ "epoch": 0.22,
391
+ "learning_rate": 4.7978383481380865e-06,
392
+ "logits/chosen": -1.5548350811004639,
393
+ "logits/rejected": -1.5831079483032227,
394
+ "logps/chosen": -598.45556640625,
395
+ "logps/rejected": -712.7095336914062,
396
+ "loss": 0.6096,
397
+ "rewards/accuracies": 0.6000000238418579,
398
+ "rewards/chosen": -0.3854503631591797,
399
+ "rewards/margins": 0.13576054573059082,
400
+ "rewards/rejected": -0.5212109088897705,
401
+ "step": 270
402
+ },
403
+ {
404
+ "epoch": 0.22,
405
+ "learning_rate": 4.769443696332272e-06,
406
+ "logits/chosen": -2.0036838054656982,
407
+ "logits/rejected": -1.7330118417739868,
408
+ "logps/chosen": -426.128662109375,
409
+ "logps/rejected": -617.08984375,
410
+ "loss": 0.6208,
411
+ "rewards/accuracies": 0.574999988079071,
412
+ "rewards/chosen": -0.2103346288204193,
413
+ "rewards/margins": 0.19718368351459503,
414
+ "rewards/rejected": -0.40751829743385315,
415
+ "step": 280
416
+ },
417
+ {
418
+ "epoch": 0.23,
419
+ "learning_rate": 4.7392794005985324e-06,
420
+ "logits/chosen": -1.8511890172958374,
421
+ "logits/rejected": -1.5459794998168945,
422
+ "logps/chosen": -737.9730224609375,
423
+ "logps/rejected": -961.6851806640625,
424
+ "loss": 0.6214,
425
+ "rewards/accuracies": 0.675000011920929,
426
+ "rewards/chosen": -0.48743873834609985,
427
+ "rewards/margins": 0.20717349648475647,
428
+ "rewards/rejected": -0.6946121454238892,
429
+ "step": 290
430
+ },
431
+ {
432
+ "epoch": 0.24,
433
+ "learning_rate": 4.707368982147318e-06,
434
+ "logits/chosen": -1.7424571514129639,
435
+ "logits/rejected": -1.4730459451675415,
436
+ "logps/chosen": -833.6031494140625,
437
+ "logps/rejected": -950.7183837890625,
438
+ "loss": 0.5805,
439
+ "rewards/accuracies": 0.6499999761581421,
440
+ "rewards/chosen": -0.5092605352401733,
441
+ "rewards/margins": 0.1874401867389679,
442
+ "rewards/rejected": -0.6967006921768188,
443
+ "step": 300
444
+ },
445
+ {
446
+ "epoch": 0.25,
447
+ "learning_rate": 4.673737323763048e-06,
448
+ "logits/chosen": -1.373366355895996,
449
+ "logits/rejected": -1.2923896312713623,
450
+ "logps/chosen": -975.0471801757812,
451
+ "logps/rejected": -1144.96533203125,
452
+ "loss": 0.6091,
453
+ "rewards/accuracies": 0.550000011920929,
454
+ "rewards/chosen": -0.7069741487503052,
455
+ "rewards/margins": 0.21132870018482208,
456
+ "rewards/rejected": -0.9183028340339661,
457
+ "step": 310
458
+ },
459
+ {
460
+ "epoch": 0.26,
461
+ "learning_rate": 4.638410650401267e-06,
462
+ "logits/chosen": -1.477074384689331,
463
+ "logits/rejected": -1.2639570236206055,
464
+ "logps/chosen": -1050.070556640625,
465
+ "logps/rejected": -1336.6949462890625,
466
+ "loss": 0.5927,
467
+ "rewards/accuracies": 0.6499999761581421,
468
+ "rewards/chosen": -0.8159879446029663,
469
+ "rewards/margins": 0.2922418415546417,
470
+ "rewards/rejected": -1.1082298755645752,
471
+ "step": 320
472
+ },
473
+ {
474
+ "epoch": 0.26,
475
+ "learning_rate": 4.601416508739211e-06,
476
+ "logits/chosen": -1.5806219577789307,
477
+ "logits/rejected": -1.5797499418258667,
478
+ "logps/chosen": -1069.904296875,
479
+ "logps/rejected": -1329.1131591796875,
480
+ "loss": 0.6329,
481
+ "rewards/accuracies": 0.574999988079071,
482
+ "rewards/chosen": -0.8489059209823608,
483
+ "rewards/margins": 0.2538915276527405,
484
+ "rewards/rejected": -1.102797508239746,
485
+ "step": 330
486
+ },
487
+ {
488
+ "epoch": 0.27,
489
+ "learning_rate": 4.562783745695738e-06,
490
+ "logits/chosen": -1.578185796737671,
491
+ "logits/rejected": -1.3778866529464722,
492
+ "logps/chosen": -898.2579956054688,
493
+ "logps/rejected": -1295.5048828125,
494
+ "loss": 0.5816,
495
+ "rewards/accuracies": 0.675000011920929,
496
+ "rewards/chosen": -0.6616091728210449,
497
+ "rewards/margins": 0.3867344856262207,
498
+ "rewards/rejected": -1.0483437776565552,
499
+ "step": 340
500
+ },
501
+ {
502
+ "epoch": 0.28,
503
+ "learning_rate": 4.522542485937369e-06,
504
+ "logits/chosen": -1.3196460008621216,
505
+ "logits/rejected": -1.0029346942901611,
506
+ "logps/chosen": -1030.9923095703125,
507
+ "logps/rejected": -1271.06494140625,
508
+ "loss": 0.6062,
509
+ "rewards/accuracies": 0.675000011920929,
510
+ "rewards/chosen": -0.7833272218704224,
511
+ "rewards/margins": 0.2811380922794342,
512
+ "rewards/rejected": -1.0644652843475342,
513
+ "step": 350
514
+ },
515
+ {
516
+ "epoch": 0.29,
517
+ "learning_rate": 4.4807241083879774e-06,
518
+ "logits/chosen": -1.4315626621246338,
519
+ "logits/rejected": -1.1289265155792236,
520
+ "logps/chosen": -1395.56640625,
521
+ "logps/rejected": -2238.73681640625,
522
+ "loss": 0.5686,
523
+ "rewards/accuracies": 0.699999988079071,
524
+ "rewards/chosen": -1.1033650636672974,
525
+ "rewards/margins": 0.8617695569992065,
526
+ "rewards/rejected": -1.965134620666504,
527
+ "step": 360
528
+ },
529
+ {
530
+ "epoch": 0.3,
531
+ "learning_rate": 4.437361221760449e-06,
532
+ "logits/chosen": -1.0882983207702637,
533
+ "logits/rejected": -0.9719398617744446,
534
+ "logps/chosen": -1259.433349609375,
535
+ "logps/rejected": -1656.2109375,
536
+ "loss": 0.5973,
537
+ "rewards/accuracies": 0.699999988079071,
538
+ "rewards/chosen": -1.069411277770996,
539
+ "rewards/margins": 0.3754344880580902,
540
+ "rewards/rejected": -1.4448457956314087,
541
+ "step": 370
542
+ },
543
+ {
544
+ "epoch": 0.3,
545
+ "learning_rate": 4.3924876391293915e-06,
546
+ "logits/chosen": -0.9243530035018921,
547
+ "logits/rejected": -0.9347097277641296,
548
+ "logps/chosen": -1276.580810546875,
549
+ "logps/rejected": -1660.425048828125,
550
+ "loss": 0.5932,
551
+ "rewards/accuracies": 0.6499999761581421,
552
+ "rewards/chosen": -1.1309508085250854,
553
+ "rewards/margins": 0.35707855224609375,
554
+ "rewards/rejected": -1.4880293607711792,
555
+ "step": 380
556
+ },
557
+ {
558
+ "epoch": 0.31,
559
+ "learning_rate": 4.346138351564711e-06,
560
+ "logits/chosen": -1.2673766613006592,
561
+ "logits/rejected": -1.1589628458023071,
562
+ "logps/chosen": -1206.547607421875,
563
+ "logps/rejected": -1615.82568359375,
564
+ "loss": 0.5588,
565
+ "rewards/accuracies": 0.75,
566
+ "rewards/chosen": -0.9714127779006958,
567
+ "rewards/margins": 0.4280439019203186,
568
+ "rewards/rejected": -1.39945650100708,
569
+ "step": 390
570
+ },
571
+ {
572
+ "epoch": 0.32,
573
+ "learning_rate": 4.2983495008466285e-06,
574
+ "logits/chosen": -1.5774317979812622,
575
+ "logits/rejected": -1.2660281658172607,
576
+ "logps/chosen": -1082.2581787109375,
577
+ "logps/rejected": -1570.060791015625,
578
+ "loss": 0.5484,
579
+ "rewards/accuracies": 0.699999988079071,
580
+ "rewards/chosen": -0.8231463432312012,
581
+ "rewards/margins": 0.5096359848976135,
582
+ "rewards/rejected": -1.3327823877334595,
583
+ "step": 400
584
+ },
585
+ {
586
+ "epoch": 0.33,
587
+ "learning_rate": 4.249158351283414e-06,
588
+ "logits/chosen": -1.4447729587554932,
589
+ "logits/rejected": -1.1427133083343506,
590
+ "logps/chosen": -1174.7425537109375,
591
+ "logps/rejected": -1466.5479736328125,
592
+ "loss": 0.5432,
593
+ "rewards/accuracies": 0.6000000238418579,
594
+ "rewards/chosen": -0.9159573316574097,
595
+ "rewards/margins": 0.33625325560569763,
596
+ "rewards/rejected": -1.2522106170654297,
597
+ "step": 410
598
+ },
599
+ {
600
+ "epoch": 0.34,
601
+ "learning_rate": 4.198603260653792e-06,
602
+ "logits/chosen": -0.7784031629562378,
603
+ "logits/rejected": -0.5429641008377075,
604
+ "logps/chosen": -2375.2294921875,
605
+ "logps/rejected": -2860.4833984375,
606
+ "loss": 0.5023,
607
+ "rewards/accuracies": 0.5249999761581421,
608
+ "rewards/chosen": -2.1455845832824707,
609
+ "rewards/margins": 0.4987269341945648,
610
+ "rewards/rejected": -2.6443114280700684,
611
+ "step": 420
612
+ },
613
+ {
614
+ "epoch": 0.34,
615
+ "learning_rate": 4.146723650296701e-06,
616
+ "logits/chosen": -0.8310529589653015,
617
+ "logits/rejected": -0.47391629219055176,
618
+ "logps/chosen": -3060.8125,
619
+ "logps/rejected": -3744.44140625,
620
+ "loss": 0.6076,
621
+ "rewards/accuracies": 0.7250000238418579,
622
+ "rewards/chosen": -2.840507984161377,
623
+ "rewards/margins": 0.725513219833374,
624
+ "rewards/rejected": -3.56602144241333,
625
+ "step": 430
626
+ },
627
+ {
628
+ "epoch": 0.35,
629
+ "learning_rate": 4.093559974371725e-06,
630
+ "logits/chosen": -1.8024406433105469,
631
+ "logits/rejected": -1.4066709280014038,
632
+ "logps/chosen": -1206.559326171875,
633
+ "logps/rejected": -1958.577392578125,
634
+ "loss": 0.5517,
635
+ "rewards/accuracies": 0.824999988079071,
636
+ "rewards/chosen": -0.948940098285675,
637
+ "rewards/margins": 0.76356440782547,
638
+ "rewards/rejected": -1.7125043869018555,
639
+ "step": 440
640
+ },
641
+ {
642
+ "epoch": 0.36,
643
+ "learning_rate": 4.039153688314146e-06,
644
+ "logits/chosen": -1.7638174295425415,
645
+ "logits/rejected": -1.6592317819595337,
646
+ "logps/chosen": -888.7286376953125,
647
+ "logps/rejected": -1274.7813720703125,
648
+ "loss": 0.5681,
649
+ "rewards/accuracies": 0.625,
650
+ "rewards/chosen": -0.6799342036247253,
651
+ "rewards/margins": 0.4119555950164795,
652
+ "rewards/rejected": -1.0918899774551392,
653
+ "step": 450
654
+ },
655
+ {
656
+ "epoch": 0.37,
657
+ "learning_rate": 3.983547216509254e-06,
658
+ "logits/chosen": -1.8817472457885742,
659
+ "logits/rejected": -1.766122817993164,
660
+ "logps/chosen": -1327.4205322265625,
661
+ "logps/rejected": -1799.326904296875,
662
+ "loss": 0.5546,
663
+ "rewards/accuracies": 0.7250000238418579,
664
+ "rewards/chosen": -0.9942048192024231,
665
+ "rewards/margins": 0.49855145812034607,
666
+ "rewards/rejected": -1.4927562475204468,
667
+ "step": 460
668
+ },
669
+ {
670
+ "epoch": 0.38,
671
+ "learning_rate": 3.92678391921108e-06,
672
+ "logits/chosen": -1.842795968055725,
673
+ "logits/rejected": -1.6158469915390015,
674
+ "logps/chosen": -1413.307373046875,
675
+ "logps/rejected": -1754.1558837890625,
676
+ "loss": 0.6049,
677
+ "rewards/accuracies": 0.6000000238418579,
678
+ "rewards/chosen": -1.0906217098236084,
679
+ "rewards/margins": 0.3714645802974701,
680
+ "rewards/rejected": -1.4620864391326904,
681
+ "step": 470
682
+ },
683
+ {
684
+ "epoch": 0.38,
685
+ "learning_rate": 3.868908058731376e-06,
686
+ "logits/chosen": -1.6520286798477173,
687
+ "logits/rejected": -1.5457327365875244,
688
+ "logps/chosen": -1587.302490234375,
689
+ "logps/rejected": -1965.464599609375,
690
+ "loss": 0.5766,
691
+ "rewards/accuracies": 0.6499999761581421,
692
+ "rewards/chosen": -1.3615858554840088,
693
+ "rewards/margins": 0.39060306549072266,
694
+ "rewards/rejected": -1.7521886825561523,
695
+ "step": 480
696
+ },
697
+ {
698
+ "epoch": 0.39,
699
+ "learning_rate": 3.8099647649251984e-06,
700
+ "logits/chosen": -1.8340473175048828,
701
+ "logits/rejected": -1.6425994634628296,
702
+ "logps/chosen": -1667.3792724609375,
703
+ "logps/rejected": -2022.701904296875,
704
+ "loss": 0.5243,
705
+ "rewards/accuracies": 0.6000000238418579,
706
+ "rewards/chosen": -1.4147528409957886,
707
+ "rewards/margins": 0.4026387333869934,
708
+ "rewards/rejected": -1.8173916339874268,
709
+ "step": 490
710
+ },
711
+ {
712
+ "epoch": 0.4,
713
+ "learning_rate": 3.7500000000000005e-06,
714
+ "logits/chosen": -1.565735936164856,
715
+ "logits/rejected": -1.4348723888397217,
716
+ "logps/chosen": -1869.2935791015625,
717
+ "logps/rejected": -2379.262939453125,
718
+ "loss": 0.5181,
719
+ "rewards/accuracies": 0.675000011920929,
720
+ "rewards/chosen": -1.658975601196289,
721
+ "rewards/margins": 0.5309830904006958,
722
+ "rewards/rejected": -2.1899585723876953,
723
+ "step": 500
724
+ },
725
+ {
726
+ "epoch": 0.41,
727
+ "learning_rate": 3.689060522675689e-06,
728
+ "logits/chosen": -1.8067477941513062,
729
+ "logits/rejected": -1.4334272146224976,
730
+ "logps/chosen": -1798.0201416015625,
731
+ "logps/rejected": -2005.1083984375,
732
+ "loss": 0.5859,
733
+ "rewards/accuracies": 0.4749999940395355,
734
+ "rewards/chosen": -1.5768072605133057,
735
+ "rewards/margins": 0.24011941254138947,
736
+ "rewards/rejected": -1.8169265985488892,
737
+ "step": 510
738
+ },
739
+ {
740
+ "epoch": 0.42,
741
+ "learning_rate": 3.627193851723577e-06,
742
+ "logits/chosen": -1.9058891534805298,
743
+ "logits/rejected": -1.5708621740341187,
744
+ "logps/chosen": -1344.331787109375,
745
+ "logps/rejected": -2066.19677734375,
746
+ "loss": 0.5302,
747
+ "rewards/accuracies": 0.7749999761581421,
748
+ "rewards/chosen": -1.0564631223678589,
749
+ "rewards/margins": 0.7303067445755005,
750
+ "rewards/rejected": -1.7867698669433594,
751
+ "step": 520
752
+ },
753
+ {
754
+ "epoch": 0.42,
755
+ "learning_rate": 3.564448228912682e-06,
756
+ "logits/chosen": -1.8703407049179077,
757
+ "logits/rejected": -1.8959243297576904,
758
+ "logps/chosen": -1418.962890625,
759
+ "logps/rejected": -1918.7298583984375,
760
+ "loss": 0.5437,
761
+ "rewards/accuracies": 0.6000000238418579,
762
+ "rewards/chosen": -1.2395734786987305,
763
+ "rewards/margins": 0.4607962965965271,
764
+ "rewards/rejected": -1.7003698348999023,
765
+ "step": 530
766
+ },
767
+ {
768
+ "epoch": 0.43,
769
+ "learning_rate": 3.5008725813922383e-06,
770
+ "logits/chosen": -1.4778034687042236,
771
+ "logits/rejected": -1.42378830909729,
772
+ "logps/chosen": -2291.814453125,
773
+ "logps/rejected": -2982.93212890625,
774
+ "loss": 0.4967,
775
+ "rewards/accuracies": 0.699999988079071,
776
+ "rewards/chosen": -2.091310977935791,
777
+ "rewards/margins": 0.6848040819168091,
778
+ "rewards/rejected": -2.7761149406433105,
779
+ "step": 540
780
+ },
781
+ {
782
+ "epoch": 0.44,
783
+ "learning_rate": 3.436516483539781e-06,
784
+ "logits/chosen": -1.495721459388733,
785
+ "logits/rejected": -1.364970326423645,
786
+ "logps/chosen": -2159.25634765625,
787
+ "logps/rejected": -2941.31982421875,
788
+ "loss": 0.5244,
789
+ "rewards/accuracies": 0.6499999761581421,
790
+ "rewards/chosen": -1.9415247440338135,
791
+ "rewards/margins": 0.7436372637748718,
792
+ "rewards/rejected": -2.685161828994751,
793
+ "step": 550
794
+ },
795
+ {
796
+ "epoch": 0.45,
797
+ "learning_rate": 3.3714301183045382e-06,
798
+ "logits/chosen": -1.7252031564712524,
799
+ "logits/rejected": -1.3431731462478638,
800
+ "logps/chosen": -2346.15673828125,
801
+ "logps/rejected": -2880.50341796875,
802
+ "loss": 0.5818,
803
+ "rewards/accuracies": 0.6000000238418579,
804
+ "rewards/chosen": -2.072269916534424,
805
+ "rewards/margins": 0.5818148851394653,
806
+ "rewards/rejected": -2.6540849208831787,
807
+ "step": 560
808
+ },
809
+ {
810
+ "epoch": 0.46,
811
+ "learning_rate": 3.3056642380762783e-06,
812
+ "logits/chosen": -1.7475759983062744,
813
+ "logits/rejected": -1.6887413263320923,
814
+ "logps/chosen": -2365.16650390625,
815
+ "logps/rejected": -3232.527587890625,
816
+ "loss": 0.4838,
817
+ "rewards/accuracies": 0.6499999761581421,
818
+ "rewards/chosen": -2.112750291824341,
819
+ "rewards/margins": 0.815180778503418,
820
+ "rewards/rejected": -2.927931070327759,
821
+ "step": 570
822
+ },
823
+ {
824
+ "epoch": 0.46,
825
+ "learning_rate": 3.2392701251101172e-06,
826
+ "logits/chosen": -1.6227779388427734,
827
+ "logits/rejected": -1.371058702468872,
828
+ "logps/chosen": -2093.24365234375,
829
+ "logps/rejected": -2805.071533203125,
830
+ "loss": 0.5206,
831
+ "rewards/accuracies": 0.5249999761581421,
832
+ "rewards/chosen": -1.8952558040618896,
833
+ "rewards/margins": 0.6814407706260681,
834
+ "rewards/rejected": -2.5766968727111816,
835
+ "step": 580
836
+ },
837
+ {
838
+ "epoch": 0.47,
839
+ "learning_rate": 3.1722995515381644e-06,
840
+ "logits/chosen": -1.6335232257843018,
841
+ "logits/rejected": -1.2477694749832153,
842
+ "logps/chosen": -2436.821044921875,
843
+ "logps/rejected": -3250.9990234375,
844
+ "loss": 0.5081,
845
+ "rewards/accuracies": 0.699999988079071,
846
+ "rewards/chosen": -2.2077012062072754,
847
+ "rewards/margins": 0.8523558378219604,
848
+ "rewards/rejected": -3.0600569248199463,
849
+ "step": 590
850
+ },
851
+ {
852
+ "epoch": 0.48,
853
+ "learning_rate": 3.1048047389991693e-06,
854
+ "logits/chosen": -1.898289442062378,
855
+ "logits/rejected": -1.5713775157928467,
856
+ "logps/chosen": -1884.5660400390625,
857
+ "logps/rejected": -2746.454833984375,
858
+ "loss": 0.5623,
859
+ "rewards/accuracies": 0.699999988079071,
860
+ "rewards/chosen": -1.6871427297592163,
861
+ "rewards/margins": 0.8711158633232117,
862
+ "rewards/rejected": -2.5582587718963623,
863
+ "step": 600
864
+ },
865
+ {
866
+ "epoch": 0.49,
867
+ "learning_rate": 3.0368383179176584e-06,
868
+ "logits/chosen": -1.4297007322311401,
869
+ "logits/rejected": -1.3948471546173096,
870
+ "logps/chosen": -2035.6175537109375,
871
+ "logps/rejected": -3012.0087890625,
872
+ "loss": 0.5395,
873
+ "rewards/accuracies": 0.7250000238418579,
874
+ "rewards/chosen": -1.860030174255371,
875
+ "rewards/margins": 0.9378849267959595,
876
+ "rewards/rejected": -2.797914981842041,
877
+ "step": 610
878
+ },
879
+ {
880
+ "epoch": 0.5,
881
+ "learning_rate": 2.9684532864643123e-06,
882
+ "logits/chosen": -1.7777831554412842,
883
+ "logits/rejected": -1.3341039419174194,
884
+ "logps/chosen": -2097.790283203125,
885
+ "logps/rejected": -2818.270263671875,
886
+ "loss": 0.4943,
887
+ "rewards/accuracies": 0.75,
888
+ "rewards/chosen": -1.8172571659088135,
889
+ "rewards/margins": 0.7645599246025085,
890
+ "rewards/rejected": -2.581817150115967,
891
+ "step": 620
892
+ },
893
+ {
894
+ "epoch": 0.5,
895
+ "learning_rate": 2.8997029692295875e-06,
896
+ "logits/chosen": -1.7123359441757202,
897
+ "logits/rejected": -1.4930963516235352,
898
+ "logps/chosen": -1850.494873046875,
899
+ "logps/rejected": -2223.23779296875,
900
+ "loss": 0.5415,
901
+ "rewards/accuracies": 0.6499999761581421,
902
+ "rewards/chosen": -1.5890159606933594,
903
+ "rewards/margins": 0.4220260679721832,
904
+ "rewards/rejected": -2.0110421180725098,
905
+ "step": 630
906
+ },
907
+ {
908
+ "epoch": 0.51,
909
+ "learning_rate": 2.8306409756428067e-06,
910
+ "logits/chosen": -1.7620747089385986,
911
+ "logits/rejected": -1.273874282836914,
912
+ "logps/chosen": -1751.422119140625,
913
+ "logps/rejected": -2349.684814453125,
914
+ "loss": 0.5267,
915
+ "rewards/accuracies": 0.7250000238418579,
916
+ "rewards/chosen": -1.5236599445343018,
917
+ "rewards/margins": 0.6361467838287354,
918
+ "rewards/rejected": -2.159806728363037,
919
+ "step": 640
920
+ },
921
+ {
922
+ "epoch": 0.52,
923
+ "learning_rate": 2.761321158169134e-06,
924
+ "logits/chosen": -1.8131462335586548,
925
+ "logits/rejected": -1.548789381980896,
926
+ "logps/chosen": -1939.831298828125,
927
+ "logps/rejected": -2459.304443359375,
928
+ "loss": 0.5526,
929
+ "rewards/accuracies": 0.6499999761581421,
930
+ "rewards/chosen": -1.7154333591461182,
931
+ "rewards/margins": 0.5310146808624268,
932
+ "rewards/rejected": -2.246448040008545,
933
+ "step": 650
934
+ },
935
+ {
936
+ "epoch": 0.53,
937
+ "learning_rate": 2.6917975703170466e-06,
938
+ "logits/chosen": -1.4764631986618042,
939
+ "logits/rejected": -1.2985259294509888,
940
+ "logps/chosen": -2074.831787109375,
941
+ "logps/rejected": -2988.404296875,
942
+ "loss": 0.5177,
943
+ "rewards/accuracies": 0.675000011920929,
944
+ "rewards/chosen": -1.8811687231063843,
945
+ "rewards/margins": 0.8929063677787781,
946
+ "rewards/rejected": -2.7740752696990967,
947
+ "step": 660
948
+ },
949
+ {
950
+ "epoch": 0.54,
951
+ "learning_rate": 2.6221244244890336e-06,
952
+ "logits/chosen": -1.606785774230957,
953
+ "logits/rejected": -1.4506292343139648,
954
+ "logps/chosen": -2944.038330078125,
955
+ "logps/rejected": -3770.83203125,
956
+ "loss": 0.5202,
957
+ "rewards/accuracies": 0.75,
958
+ "rewards/chosen": -2.648715019226074,
959
+ "rewards/margins": 0.8418272733688354,
960
+ "rewards/rejected": -3.4905428886413574,
961
+ "step": 670
962
+ },
963
+ {
964
+ "epoch": 0.54,
965
+ "learning_rate": 2.5523560497083927e-06,
966
+ "logits/chosen": -1.7594387531280518,
967
+ "logits/rejected": -1.4435147047042847,
968
+ "logps/chosen": -2691.794921875,
969
+ "logps/rejected": -2907.418212890625,
970
+ "loss": 0.6347,
971
+ "rewards/accuracies": 0.5,
972
+ "rewards/chosen": -2.4359123706817627,
973
+ "rewards/margins": 0.3157985210418701,
974
+ "rewards/rejected": -2.751710891723633,
975
+ "step": 680
976
+ },
977
+ {
978
+ "epoch": 0.55,
979
+ "learning_rate": 2.482546849255096e-06,
980
+ "logits/chosen": -1.6512796878814697,
981
+ "logits/rejected": -1.2861438989639282,
982
+ "logps/chosen": -2340.228515625,
983
+ "logps/rejected": -3065.31591796875,
984
+ "loss": 0.4919,
985
+ "rewards/accuracies": 0.800000011920929,
986
+ "rewards/chosen": -2.0771026611328125,
987
+ "rewards/margins": 0.6874440908432007,
988
+ "rewards/rejected": -2.7645466327667236,
989
+ "step": 690
990
+ },
991
+ {
992
+ "epoch": 0.56,
993
+ "learning_rate": 2.4127512582437486e-06,
994
+ "logits/chosen": -1.7903190851211548,
995
+ "logits/rejected": -1.5637561082839966,
996
+ "logps/chosen": -1961.098876953125,
997
+ "logps/rejected": -2661.05126953125,
998
+ "loss": 0.5352,
999
+ "rewards/accuracies": 0.824999988079071,
1000
+ "rewards/chosen": -1.727064847946167,
1001
+ "rewards/margins": 0.7038751840591431,
1002
+ "rewards/rejected": -2.4309401512145996,
1003
+ "step": 700
1004
+ },
1005
+ {
1006
+ "epoch": 0.57,
1007
+ "learning_rate": 2.3430237011767166e-06,
1008
+ "logits/chosen": -1.5963449478149414,
1009
+ "logits/rejected": -1.5752345323562622,
1010
+ "logps/chosen": -2412.08447265625,
1011
+ "logps/rejected": -2860.287109375,
1012
+ "loss": 0.5364,
1013
+ "rewards/accuracies": 0.6000000238418579,
1014
+ "rewards/chosen": -2.1989083290100098,
1015
+ "rewards/margins": 0.4225144386291504,
1016
+ "rewards/rejected": -2.62142276763916,
1017
+ "step": 710
1018
+ },
1019
+ {
1020
+ "epoch": 0.58,
1021
+ "learning_rate": 2.2734185495055503e-06,
1022
+ "logits/chosen": -1.581274390220642,
1023
+ "logits/rejected": -1.338945746421814,
1024
+ "logps/chosen": -2689.8212890625,
1025
+ "logps/rejected": -3143.3369140625,
1026
+ "loss": 0.5331,
1027
+ "rewards/accuracies": 0.699999988079071,
1028
+ "rewards/chosen": -2.4156579971313477,
1029
+ "rewards/margins": 0.4918283522129059,
1030
+ "rewards/rejected": -2.9074864387512207,
1031
+ "step": 720
1032
+ },
1033
+ {
1034
+ "epoch": 0.58,
1035
+ "learning_rate": 2.2039900792337477e-06,
1036
+ "logits/chosen": -1.6093494892120361,
1037
+ "logits/rejected": -1.2763807773590088,
1038
+ "logps/chosen": -1861.92578125,
1039
+ "logps/rejected": -2657.68212890625,
1040
+ "loss": 0.4712,
1041
+ "rewards/accuracies": 0.75,
1042
+ "rewards/chosen": -1.6057929992675781,
1043
+ "rewards/margins": 0.847273051738739,
1044
+ "rewards/rejected": -2.453066110610962,
1045
+ "step": 730
1046
+ },
1047
+ {
1048
+ "epoch": 0.59,
1049
+ "learning_rate": 2.134792428593971e-06,
1050
+ "logits/chosen": -1.3526313304901123,
1051
+ "logits/rejected": -1.3224549293518066,
1052
+ "logps/chosen": -2195.27197265625,
1053
+ "logps/rejected": -2851.3291015625,
1054
+ "loss": 0.5536,
1055
+ "rewards/accuracies": 0.7250000238418579,
1056
+ "rewards/chosen": -2.023761749267578,
1057
+ "rewards/margins": 0.6121652722358704,
1058
+ "rewards/rejected": -2.635927200317383,
1059
+ "step": 740
1060
+ },
1061
+ {
1062
+ "epoch": 0.6,
1063
+ "learning_rate": 2.0658795558326745e-06,
1064
+ "logits/chosen": -1.6000654697418213,
1065
+ "logits/rejected": -1.4151827096939087,
1066
+ "logps/chosen": -2148.247802734375,
1067
+ "logps/rejected": -2729.99853515625,
1068
+ "loss": 0.5222,
1069
+ "rewards/accuracies": 0.6000000238418579,
1070
+ "rewards/chosen": -1.903270959854126,
1071
+ "rewards/margins": 0.5996490716934204,
1072
+ "rewards/rejected": -2.5029196739196777,
1073
+ "step": 750
1074
+ },
1075
+ {
1076
+ "epoch": 0.61,
1077
+ "learning_rate": 1.997305197135089e-06,
1078
+ "logits/chosen": -1.8795032501220703,
1079
+ "logits/rejected": -1.8359801769256592,
1080
+ "logps/chosen": -2000.689208984375,
1081
+ "logps/rejected": -2926.07275390625,
1082
+ "loss": 0.5271,
1083
+ "rewards/accuracies": 0.75,
1084
+ "rewards/chosen": -1.7533314228057861,
1085
+ "rewards/margins": 0.9173868298530579,
1086
+ "rewards/rejected": -2.6707184314727783,
1087
+ "step": 760
1088
+ },
1089
+ {
1090
+ "epoch": 0.62,
1091
+ "learning_rate": 1.9291228247233607e-06,
1092
+ "logits/chosen": -1.6307891607284546,
1093
+ "logits/rejected": -1.496146321296692,
1094
+ "logps/chosen": -1922.004150390625,
1095
+ "logps/rejected": -2854.74365234375,
1096
+ "loss": 0.4909,
1097
+ "rewards/accuracies": 0.75,
1098
+ "rewards/chosen": -1.7167818546295166,
1099
+ "rewards/margins": 0.9143074750900269,
1100
+ "rewards/rejected": -2.631089448928833,
1101
+ "step": 770
1102
+ },
1103
+ {
1104
+ "epoch": 0.62,
1105
+ "learning_rate": 1.8613856051605242e-06,
1106
+ "logits/chosen": -1.5307934284210205,
1107
+ "logits/rejected": -1.4985663890838623,
1108
+ "logps/chosen": -1941.616455078125,
1109
+ "logps/rejected": -2720.62109375,
1110
+ "loss": 0.495,
1111
+ "rewards/accuracies": 0.675000011920929,
1112
+ "rewards/chosen": -1.7477505207061768,
1113
+ "rewards/margins": 0.7529212236404419,
1114
+ "rewards/rejected": -2.500671863555908,
1115
+ "step": 780
1116
+ },
1117
+ {
1118
+ "epoch": 0.63,
1119
+ "learning_rate": 1.7941463578928088e-06,
1120
+ "logits/chosen": -1.6022170782089233,
1121
+ "logits/rejected": -1.5732767581939697,
1122
+ "logps/chosen": -2065.362548828125,
1123
+ "logps/rejected": -2565.924560546875,
1124
+ "loss": 0.5311,
1125
+ "rewards/accuracies": 0.6499999761581421,
1126
+ "rewards/chosen": -1.8574340343475342,
1127
+ "rewards/margins": 0.4975149631500244,
1128
+ "rewards/rejected": -2.3549489974975586,
1129
+ "step": 790
1130
+ },
1131
+ {
1132
+ "epoch": 0.64,
1133
+ "learning_rate": 1.7274575140626318e-06,
1134
+ "logits/chosen": -1.5523189306259155,
1135
+ "logits/rejected": -1.2900121212005615,
1136
+ "logps/chosen": -2249.78466796875,
1137
+ "logps/rejected": -3071.79541015625,
1138
+ "loss": 0.5322,
1139
+ "rewards/accuracies": 0.675000011920929,
1140
+ "rewards/chosen": -2.0411088466644287,
1141
+ "rewards/margins": 0.8322113752365112,
1142
+ "rewards/rejected": -2.8733203411102295,
1143
+ "step": 800
1144
+ },
1145
+ {
1146
+ "epoch": 0.65,
1147
+ "learning_rate": 1.661371075624363e-06,
1148
+ "logits/chosen": -1.750195860862732,
1149
+ "logits/rejected": -1.4256199598312378,
1150
+ "logps/chosen": -2680.12548828125,
1151
+ "logps/rejected": -3339.82177734375,
1152
+ "loss": 0.5582,
1153
+ "rewards/accuracies": 0.574999988079071,
1154
+ "rewards/chosen": -2.413327693939209,
1155
+ "rewards/margins": 0.6952977180480957,
1156
+ "rewards/rejected": -3.108625650405884,
1157
+ "step": 810
1158
+ },
1159
+ {
1160
+ "epoch": 0.66,
1161
+ "learning_rate": 1.5959385747947697e-06,
1162
+ "logits/chosen": -1.6673352718353271,
1163
+ "logits/rejected": -1.4309087991714478,
1164
+ "logps/chosen": -2934.502685546875,
1165
+ "logps/rejected": -3557.366455078125,
1166
+ "loss": 0.5826,
1167
+ "rewards/accuracies": 0.7250000238418579,
1168
+ "rewards/chosen": -2.6931636333465576,
1169
+ "rewards/margins": 0.6061044931411743,
1170
+ "rewards/rejected": -3.2992682456970215,
1171
+ "step": 820
1172
+ },
1173
+ {
1174
+ "epoch": 0.66,
1175
+ "learning_rate": 1.5312110338697427e-06,
1176
+ "logits/chosen": -1.4265161752700806,
1177
+ "logits/rejected": -1.3840219974517822,
1178
+ "logps/chosen": -2902.10205078125,
1179
+ "logps/rejected": -3211.46484375,
1180
+ "loss": 0.5793,
1181
+ "rewards/accuracies": 0.550000011920929,
1182
+ "rewards/chosen": -2.7132956981658936,
1183
+ "rewards/margins": 0.2958894670009613,
1184
+ "rewards/rejected": -3.009185314178467,
1185
+ "step": 830
1186
+ },
1187
+ {
1188
+ "epoch": 0.67,
1189
+ "learning_rate": 1.467238925438646e-06,
1190
+ "logits/chosen": -1.93195378780365,
1191
+ "logits/rejected": -1.7828657627105713,
1192
+ "logps/chosen": -2659.49609375,
1193
+ "logps/rejected": -3121.871826171875,
1194
+ "loss": 0.5751,
1195
+ "rewards/accuracies": 0.574999988079071,
1196
+ "rewards/chosen": -2.427194118499756,
1197
+ "rewards/margins": 0.47958964109420776,
1198
+ "rewards/rejected": -2.9067838191986084,
1199
+ "step": 840
1200
+ },
1201
+ {
1202
+ "epoch": 0.68,
1203
+ "learning_rate": 1.4040721330273063e-06,
1204
+ "logits/chosen": -1.6783872842788696,
1205
+ "logits/rejected": -1.3621320724487305,
1206
+ "logps/chosen": -2476.935791015625,
1207
+ "logps/rejected": -3240.056640625,
1208
+ "loss": 0.5022,
1209
+ "rewards/accuracies": 0.7749999761581421,
1210
+ "rewards/chosen": -2.256714344024658,
1211
+ "rewards/margins": 0.7921364307403564,
1212
+ "rewards/rejected": -3.0488505363464355,
1213
+ "step": 850
1214
+ },
1215
+ {
1216
+ "epoch": 0.69,
1217
+ "learning_rate": 1.3417599122003464e-06,
1218
+ "logits/chosen": -1.631744146347046,
1219
+ "logits/rejected": -1.4881222248077393,
1220
+ "logps/chosen": -2611.767333984375,
1221
+ "logps/rejected": -3342.803466796875,
1222
+ "loss": 0.5239,
1223
+ "rewards/accuracies": 0.75,
1224
+ "rewards/chosen": -2.361072301864624,
1225
+ "rewards/margins": 0.6957711577415466,
1226
+ "rewards/rejected": -3.0568435192108154,
1227
+ "step": 860
1228
+ },
1229
+ {
1230
+ "epoch": 0.7,
1231
+ "learning_rate": 1.280350852153168e-06,
1232
+ "logits/chosen": -1.6476237773895264,
1233
+ "logits/rejected": -1.2135014533996582,
1234
+ "logps/chosen": -2210.89697265625,
1235
+ "logps/rejected": -3153.556884765625,
1236
+ "loss": 0.49,
1237
+ "rewards/accuracies": 0.8500000238418579,
1238
+ "rewards/chosen": -1.9529176950454712,
1239
+ "rewards/margins": 0.9891131520271301,
1240
+ "rewards/rejected": -2.942031145095825,
1241
+ "step": 870
1242
+ },
1243
+ {
1244
+ "epoch": 0.7,
1245
+ "learning_rate": 1.2198928378235717e-06,
1246
+ "logits/chosen": -1.8185428380966187,
1247
+ "logits/rejected": -1.4744179248809814,
1248
+ "logps/chosen": -1982.8447265625,
1249
+ "logps/rejected": -2986.81591796875,
1250
+ "loss": 0.4908,
1251
+ "rewards/accuracies": 0.625,
1252
+ "rewards/chosen": -1.720696210861206,
1253
+ "rewards/margins": 1.0247818231582642,
1254
+ "rewards/rejected": -2.7454781532287598,
1255
+ "step": 880
1256
+ },
1257
+ {
1258
+ "epoch": 0.71,
1259
+ "learning_rate": 1.160433012552508e-06,
1260
+ "logits/chosen": -1.9028065204620361,
1261
+ "logits/rejected": -1.5971095561981201,
1262
+ "logps/chosen": -1988.5980224609375,
1263
+ "logps/rejected": -2843.60302734375,
1264
+ "loss": 0.5298,
1265
+ "rewards/accuracies": 0.675000011920929,
1266
+ "rewards/chosen": -1.695246934890747,
1267
+ "rewards/margins": 0.9005265235900879,
1268
+ "rewards/rejected": -2.595773458480835,
1269
+ "step": 890
1270
+ },
1271
+ {
1272
+ "epoch": 0.72,
1273
+ "learning_rate": 1.1020177413231334e-06,
1274
+ "logits/chosen": -1.7520854473114014,
1275
+ "logits/rejected": -1.4778826236724854,
1276
+ "logps/chosen": -1867.814453125,
1277
+ "logps/rejected": -2722.72900390625,
1278
+ "loss": 0.5195,
1279
+ "rewards/accuracies": 0.625,
1280
+ "rewards/chosen": -1.6399002075195312,
1281
+ "rewards/margins": 0.8625130653381348,
1282
+ "rewards/rejected": -2.502413272857666,
1283
+ "step": 900
1284
+ },
1285
+ {
1286
+ "epoch": 0.73,
1287
+ "learning_rate": 1.0446925746067768e-06,
1288
+ "logits/chosen": -1.6180340051651,
1289
+ "logits/rejected": -1.4504057168960571,
1290
+ "logps/chosen": -2156.244140625,
1291
+ "logps/rejected": -3364.479736328125,
1292
+ "loss": 0.4724,
1293
+ "rewards/accuracies": 0.7250000238418579,
1294
+ "rewards/chosen": -1.9639689922332764,
1295
+ "rewards/margins": 1.1746752262115479,
1296
+ "rewards/rejected": -3.138643980026245,
1297
+ "step": 910
1298
+ },
1299
+ {
1300
+ "epoch": 0.74,
1301
+ "learning_rate": 9.88502212844063e-07,
1302
+ "logits/chosen": -1.8366491794586182,
1303
+ "logits/rejected": -1.353752851486206,
1304
+ "logps/chosen": -1940.987060546875,
1305
+ "logps/rejected": -2697.123779296875,
1306
+ "loss": 0.506,
1307
+ "rewards/accuracies": 0.675000011920929,
1308
+ "rewards/chosen": -1.714422583580017,
1309
+ "rewards/margins": 0.7562412023544312,
1310
+ "rewards/rejected": -2.4706637859344482,
1311
+ "step": 920
1312
+ },
1313
+ {
1314
+ "epoch": 0.74,
1315
+ "learning_rate": 9.334904715888496e-07,
1316
+ "logits/chosen": -1.8136298656463623,
1317
+ "logits/rejected": -1.5429127216339111,
1318
+ "logps/chosen": -2025.2021484375,
1319
+ "logps/rejected": -2842.939453125,
1320
+ "loss": 0.5274,
1321
+ "rewards/accuracies": 0.6499999761581421,
1322
+ "rewards/chosen": -1.792414903640747,
1323
+ "rewards/margins": 0.8433243036270142,
1324
+ "rewards/rejected": -2.63573956489563,
1325
+ "step": 930
1326
+ },
1327
+ {
1328
+ "epoch": 0.75,
1329
+ "learning_rate": 8.797002473421729e-07,
1330
+ "logits/chosen": -1.6221929788589478,
1331
+ "logits/rejected": -1.536675214767456,
1332
+ "logps/chosen": -2179.556884765625,
1333
+ "logps/rejected": -2724.71142578125,
1334
+ "loss": 0.5293,
1335
+ "rewards/accuracies": 0.7250000238418579,
1336
+ "rewards/chosen": -1.8998562097549438,
1337
+ "rewards/margins": 0.5943135023117065,
1338
+ "rewards/rejected": -2.4941699504852295,
1339
+ "step": 940
1340
+ },
1341
+ {
1342
+ "epoch": 0.76,
1343
+ "learning_rate": 8.271734841028553e-07,
1344
+ "logits/chosen": -1.599360704421997,
1345
+ "logits/rejected": -1.4945180416107178,
1346
+ "logps/chosen": -2247.677001953125,
1347
+ "logps/rejected": -2832.541015625,
1348
+ "loss": 0.5645,
1349
+ "rewards/accuracies": 0.550000011920929,
1350
+ "rewards/chosen": -2.041861057281494,
1351
+ "rewards/margins": 0.5804768800735474,
1352
+ "rewards/rejected": -2.622338056564331,
1353
+ "step": 950
1354
+ },
1355
+ {
1356
+ "epoch": 0.77,
1357
+ "learning_rate": 7.759511406608255e-07,
1358
+ "logits/chosen": -1.9042466878890991,
1359
+ "logits/rejected": -1.5334478616714478,
1360
+ "logps/chosen": -1946.2919921875,
1361
+ "logps/rejected": -2741.82861328125,
1362
+ "loss": 0.466,
1363
+ "rewards/accuracies": 0.699999988079071,
1364
+ "rewards/chosen": -1.713287115097046,
1365
+ "rewards/margins": 0.8301644325256348,
1366
+ "rewards/rejected": -2.5434517860412598,
1367
+ "step": 960
1368
+ },
1369
+ {
1370
+ "epoch": 0.78,
1371
+ "learning_rate": 7.260731586586983e-07,
1372
+ "logits/chosen": -1.5018055438995361,
1373
+ "logits/rejected": -1.4937880039215088,
1374
+ "logps/chosen": -2611.099609375,
1375
+ "logps/rejected": -3151.30810546875,
1376
+ "loss": 0.5775,
1377
+ "rewards/accuracies": 0.6499999761581421,
1378
+ "rewards/chosen": -2.3841938972473145,
1379
+ "rewards/margins": 0.5148266553878784,
1380
+ "rewards/rejected": -2.8990206718444824,
1381
+ "step": 970
1382
+ },
1383
+ {
1384
+ "epoch": 0.78,
1385
+ "learning_rate": 6.775784314464717e-07,
1386
+ "logits/chosen": -1.6764729022979736,
1387
+ "logits/rejected": -1.5057140588760376,
1388
+ "logps/chosen": -2298.930908203125,
1389
+ "logps/rejected": -2712.99169921875,
1390
+ "loss": 0.556,
1391
+ "rewards/accuracies": 0.5249999761581421,
1392
+ "rewards/chosen": -2.0648088455200195,
1393
+ "rewards/margins": 0.4065318703651428,
1394
+ "rewards/rejected": -2.4713406562805176,
1395
+ "step": 980
1396
+ },
1397
+ {
1398
+ "epoch": 0.79,
1399
+ "learning_rate": 6.305047737536707e-07,
1400
+ "logits/chosen": -1.7027795314788818,
1401
+ "logits/rejected": -1.3025496006011963,
1402
+ "logps/chosen": -2075.1337890625,
1403
+ "logps/rejected": -2931.57080078125,
1404
+ "loss": 0.4584,
1405
+ "rewards/accuracies": 0.699999988079071,
1406
+ "rewards/chosen": -1.846689224243164,
1407
+ "rewards/margins": 0.8770244717597961,
1408
+ "rewards/rejected": -2.7237138748168945,
1409
+ "step": 990
1410
+ },
1411
+ {
1412
+ "epoch": 0.8,
1413
+ "learning_rate": 5.848888922025553e-07,
1414
+ "logits/chosen": -1.8110439777374268,
1415
+ "logits/rejected": -1.4326590299606323,
1416
+ "logps/chosen": -2325.86474609375,
1417
+ "logps/rejected": -3089.902099609375,
1418
+ "loss": 0.5249,
1419
+ "rewards/accuracies": 0.625,
1420
+ "rewards/chosen": -2.067896604537964,
1421
+ "rewards/margins": 0.7900792360305786,
1422
+ "rewards/rejected": -2.857975482940674,
1423
+ "step": 1000
1424
+ },
1425
+ {
1426
+ "epoch": 0.81,
1427
+ "learning_rate": 5.407663566854008e-07,
1428
+ "logits/chosen": -1.6088895797729492,
1429
+ "logits/rejected": -1.4155261516571045,
1430
+ "logps/chosen": -2324.50927734375,
1431
+ "logps/rejected": -3341.44970703125,
1432
+ "loss": 0.525,
1433
+ "rewards/accuracies": 0.699999988079071,
1434
+ "rewards/chosen": -2.1127965450286865,
1435
+ "rewards/margins": 0.9637571573257446,
1436
+ "rewards/rejected": -3.0765535831451416,
1437
+ "step": 1010
1438
+ },
1439
+ {
1440
+ "epoch": 0.82,
1441
+ "learning_rate": 4.981715726281666e-07,
1442
+ "logits/chosen": -1.6132185459136963,
1443
+ "logits/rejected": -1.2942755222320557,
1444
+ "logps/chosen": -2351.848388671875,
1445
+ "logps/rejected": -3291.189453125,
1446
+ "loss": 0.5183,
1447
+ "rewards/accuracies": 0.824999988079071,
1448
+ "rewards/chosen": -2.1332767009735107,
1449
+ "rewards/margins": 0.9448688626289368,
1450
+ "rewards/rejected": -3.0781455039978027,
1451
+ "step": 1020
1452
+ },
1453
+ {
1454
+ "epoch": 0.82,
1455
+ "learning_rate": 4.5713775416217884e-07,
1456
+ "logits/chosen": -1.770582914352417,
1457
+ "logits/rejected": -1.515246868133545,
1458
+ "logps/chosen": -2241.62109375,
1459
+ "logps/rejected": -3231.28125,
1460
+ "loss": 0.4636,
1461
+ "rewards/accuracies": 0.800000011920929,
1462
+ "rewards/chosen": -1.9664819240570068,
1463
+ "rewards/margins": 1.0160348415374756,
1464
+ "rewards/rejected": -2.9825170040130615,
1465
+ "step": 1030
1466
+ },
1467
+ {
1468
+ "epoch": 0.83,
1469
+ "learning_rate": 4.1769689822475147e-07,
1470
+ "logits/chosen": -1.5879560708999634,
1471
+ "logits/rejected": -1.5036201477050781,
1472
+ "logps/chosen": -1936.926025390625,
1473
+ "logps/rejected": -2708.7578125,
1474
+ "loss": 0.4947,
1475
+ "rewards/accuracies": 0.625,
1476
+ "rewards/chosen": -1.7738593816757202,
1477
+ "rewards/margins": 0.7560992240905762,
1478
+ "rewards/rejected": -2.529958486557007,
1479
+ "step": 1040
1480
+ },
1481
+ {
1482
+ "epoch": 0.84,
1483
+ "learning_rate": 3.798797596089351e-07,
1484
+ "logits/chosen": -1.4568753242492676,
1485
+ "logits/rejected": -1.477905511856079,
1486
+ "logps/chosen": -2428.654296875,
1487
+ "logps/rejected": -3266.05712890625,
1488
+ "loss": 0.5201,
1489
+ "rewards/accuracies": 0.7250000238418579,
1490
+ "rewards/chosen": -2.2049851417541504,
1491
+ "rewards/margins": 0.8197803497314453,
1492
+ "rewards/rejected": -3.0247654914855957,
1493
+ "step": 1050
1494
+ },
1495
+ {
1496
+ "epoch": 0.85,
1497
+ "learning_rate": 3.4371582698185636e-07,
1498
+ "logits/chosen": -1.6677932739257812,
1499
+ "logits/rejected": -1.5246306657791138,
1500
+ "logps/chosen": -2556.70751953125,
1501
+ "logps/rejected": -3194.56884765625,
1502
+ "loss": 0.5391,
1503
+ "rewards/accuracies": 0.550000011920929,
1504
+ "rewards/chosen": -2.3062329292297363,
1505
+ "rewards/margins": 0.6159250140190125,
1506
+ "rewards/rejected": -2.9221577644348145,
1507
+ "step": 1060
1508
+ },
1509
+ {
1510
+ "epoch": 0.86,
1511
+ "learning_rate": 3.092332998903416e-07,
1512
+ "logits/chosen": -1.7090551853179932,
1513
+ "logits/rejected": -1.4194791316986084,
1514
+ "logps/chosen": -2150.77392578125,
1515
+ "logps/rejected": -2783.882568359375,
1516
+ "loss": 0.5195,
1517
+ "rewards/accuracies": 0.6499999761581421,
1518
+ "rewards/chosen": -1.9338016510009766,
1519
+ "rewards/margins": 0.6355168223381042,
1520
+ "rewards/rejected": -2.5693182945251465,
1521
+ "step": 1070
1522
+ },
1523
+ {
1524
+ "epoch": 0.86,
1525
+ "learning_rate": 2.764590667717562e-07,
1526
+ "logits/chosen": -1.606188178062439,
1527
+ "logits/rejected": -1.4986306428909302,
1528
+ "logps/chosen": -2124.823974609375,
1529
+ "logps/rejected": -2895.18310546875,
1530
+ "loss": 0.4628,
1531
+ "rewards/accuracies": 0.699999988079071,
1532
+ "rewards/chosen": -1.912660837173462,
1533
+ "rewards/margins": 0.7604308724403381,
1534
+ "rewards/rejected": -2.673091411590576,
1535
+ "step": 1080
1536
+ },
1537
+ {
1538
+ "epoch": 0.87,
1539
+ "learning_rate": 2.454186839872158e-07,
1540
+ "logits/chosen": -1.5546796321868896,
1541
+ "logits/rejected": -1.550550103187561,
1542
+ "logps/chosen": -2096.05078125,
1543
+ "logps/rejected": -2596.4248046875,
1544
+ "loss": 0.5839,
1545
+ "rewards/accuracies": 0.6000000238418579,
1546
+ "rewards/chosen": -1.9206135272979736,
1547
+ "rewards/margins": 0.4739322066307068,
1548
+ "rewards/rejected": -2.394545793533325,
1549
+ "step": 1090
1550
+ },
1551
+ {
1552
+ "epoch": 0.88,
1553
+ "learning_rate": 2.1613635589349756e-07,
1554
+ "logits/chosen": -1.657184362411499,
1555
+ "logits/rejected": -1.3406727313995361,
1556
+ "logps/chosen": -2227.37841796875,
1557
+ "logps/rejected": -3146.83349609375,
1558
+ "loss": 0.459,
1559
+ "rewards/accuracies": 0.7250000238418579,
1560
+ "rewards/chosen": -1.9884710311889648,
1561
+ "rewards/margins": 0.9205359220504761,
1562
+ "rewards/rejected": -2.9090065956115723,
1563
+ "step": 1100
1564
+ },
1565
+ {
1566
+ "epoch": 0.89,
1567
+ "learning_rate": 1.8863491596921745e-07,
1568
+ "logits/chosen": -1.4567536115646362,
1569
+ "logits/rejected": -1.1964176893234253,
1570
+ "logps/chosen": -2488.02099609375,
1571
+ "logps/rejected": -3093.317626953125,
1572
+ "loss": 0.5213,
1573
+ "rewards/accuracies": 0.6000000238418579,
1574
+ "rewards/chosen": -2.2580440044403076,
1575
+ "rewards/margins": 0.6173666715621948,
1576
+ "rewards/rejected": -2.875410556793213,
1577
+ "step": 1110
1578
+ },
1579
+ {
1580
+ "epoch": 0.9,
1581
+ "learning_rate": 1.629358090099639e-07,
1582
+ "logits/chosen": -1.8676036596298218,
1583
+ "logits/rejected": -1.5324196815490723,
1584
+ "logps/chosen": -2476.497314453125,
1585
+ "logps/rejected": -3140.97607421875,
1586
+ "loss": 0.5292,
1587
+ "rewards/accuracies": 0.675000011920929,
1588
+ "rewards/chosen": -2.2105040550231934,
1589
+ "rewards/margins": 0.6873103380203247,
1590
+ "rewards/rejected": -2.8978142738342285,
1591
+ "step": 1120
1592
+ },
1593
+ {
1594
+ "epoch": 0.9,
1595
+ "learning_rate": 1.3905907440629752e-07,
1596
+ "logits/chosen": -1.5062824487686157,
1597
+ "logits/rejected": -1.7191336154937744,
1598
+ "logps/chosen": -2101.674560546875,
1599
+ "logps/rejected": -2914.34228515625,
1600
+ "loss": 0.4889,
1601
+ "rewards/accuracies": 0.699999988079071,
1602
+ "rewards/chosen": -1.8925731182098389,
1603
+ "rewards/margins": 0.7410818338394165,
1604
+ "rewards/rejected": -2.633654832839966,
1605
+ "step": 1130
1606
+ },
1607
+ {
1608
+ "epoch": 0.91,
1609
+ "learning_rate": 1.1702333051763271e-07,
1610
+ "logits/chosen": -1.885053277015686,
1611
+ "logits/rejected": -1.7132413387298584,
1612
+ "logps/chosen": -2271.51806640625,
1613
+ "logps/rejected": -3019.444580078125,
1614
+ "loss": 0.4746,
1615
+ "rewards/accuracies": 0.75,
1616
+ "rewards/chosen": -2.0059704780578613,
1617
+ "rewards/margins": 0.8035289645195007,
1618
+ "rewards/rejected": -2.809499740600586,
1619
+ "step": 1140
1620
+ },
1621
+ {
1622
+ "epoch": 0.92,
1623
+ "learning_rate": 9.684576015420277e-08,
1624
+ "logits/chosen": -1.7674312591552734,
1625
+ "logits/rejected": -1.7866573333740234,
1626
+ "logps/chosen": -2304.083740234375,
1627
+ "logps/rejected": -3042.02783203125,
1628
+ "loss": 0.5235,
1629
+ "rewards/accuracies": 0.699999988079071,
1630
+ "rewards/chosen": -2.0314338207244873,
1631
+ "rewards/margins": 0.7121332883834839,
1632
+ "rewards/rejected": -2.7435669898986816,
1633
+ "step": 1150
1634
+ },
1635
+ {
1636
+ "epoch": 0.93,
1637
+ "learning_rate": 7.854209717842231e-08,
1638
+ "logits/chosen": -1.6722157001495361,
1639
+ "logits/rejected": -1.738372564315796,
1640
+ "logps/chosen": -2533.139404296875,
1641
+ "logps/rejected": -3169.639892578125,
1642
+ "loss": 0.508,
1643
+ "rewards/accuracies": 0.574999988079071,
1644
+ "rewards/chosen": -2.2765398025512695,
1645
+ "rewards/margins": 0.6401499509811401,
1646
+ "rewards/rejected": -2.91668963432312,
1647
+ "step": 1160
1648
+ },
1649
+ {
1650
+ "epoch": 0.94,
1651
+ "learning_rate": 6.212661423609184e-08,
1652
+ "logits/chosen": -1.6138194799423218,
1653
+ "logits/rejected": -1.4168260097503662,
1654
+ "logps/chosen": -2651.40087890625,
1655
+ "logps/rejected": -2953.78564453125,
1656
+ "loss": 0.5389,
1657
+ "rewards/accuracies": 0.5,
1658
+ "rewards/chosen": -2.369499683380127,
1659
+ "rewards/margins": 0.3613986372947693,
1660
+ "rewards/rejected": -2.730898380279541,
1661
+ "step": 1170
1662
+ },
1663
+ {
1664
+ "epoch": 0.94,
1665
+ "learning_rate": 4.761211162702117e-08,
1666
+ "logits/chosen": -1.8268349170684814,
1667
+ "logits/rejected": -1.389704942703247,
1668
+ "logps/chosen": -2308.206298828125,
1669
+ "logps/rejected": -3084.591796875,
1670
+ "loss": 0.4561,
1671
+ "rewards/accuracies": 0.7250000238418579,
1672
+ "rewards/chosen": -2.0564706325531006,
1673
+ "rewards/margins": 0.8305153846740723,
1674
+ "rewards/rejected": -2.8869857788085938,
1675
+ "step": 1180
1676
+ },
1677
+ {
1678
+ "epoch": 0.95,
1679
+ "learning_rate": 3.5009907323737826e-08,
1680
+ "logits/chosen": -1.5044516324996948,
1681
+ "logits/rejected": -1.3144376277923584,
1682
+ "logps/chosen": -2185.82275390625,
1683
+ "logps/rejected": -3079.09130859375,
1684
+ "loss": 0.4552,
1685
+ "rewards/accuracies": 0.7250000238418579,
1686
+ "rewards/chosen": -1.9693256616592407,
1687
+ "rewards/margins": 0.8803497552871704,
1688
+ "rewards/rejected": -2.8496756553649902,
1689
+ "step": 1190
1690
+ },
1691
+ {
1692
+ "epoch": 0.96,
1693
+ "learning_rate": 2.4329828146074096e-08,
1694
+ "logits/chosen": -1.5807933807373047,
1695
+ "logits/rejected": -1.4424628019332886,
1696
+ "logps/chosen": -2456.582763671875,
1697
+ "logps/rejected": -2916.781982421875,
1698
+ "loss": 0.5377,
1699
+ "rewards/accuracies": 0.4749999940395355,
1700
+ "rewards/chosen": -2.223078489303589,
1701
+ "rewards/margins": 0.4909304976463318,
1702
+ "rewards/rejected": -2.7140090465545654,
1703
+ "step": 1200
1704
+ },
1705
+ {
1706
+ "epoch": 0.97,
1707
+ "learning_rate": 1.5580202098509078e-08,
1708
+ "logits/chosen": -1.4222395420074463,
1709
+ "logits/rejected": -1.4579776525497437,
1710
+ "logps/chosen": -2549.04052734375,
1711
+ "logps/rejected": -3383.83203125,
1712
+ "loss": 0.4934,
1713
+ "rewards/accuracies": 0.7250000238418579,
1714
+ "rewards/chosen": -2.3147170543670654,
1715
+ "rewards/margins": 0.7796028852462769,
1716
+ "rewards/rejected": -3.094320058822632,
1717
+ "step": 1210
1718
+ },
1719
+ {
1720
+ "epoch": 0.98,
1721
+ "learning_rate": 8.767851876239075e-09,
1722
+ "logits/chosen": -1.9085347652435303,
1723
+ "logits/rejected": -1.5333627462387085,
1724
+ "logps/chosen": -2555.34228515625,
1725
+ "logps/rejected": -3125.24853515625,
1726
+ "loss": 0.5644,
1727
+ "rewards/accuracies": 0.6000000238418579,
1728
+ "rewards/chosen": -2.2714717388153076,
1729
+ "rewards/margins": 0.6149500012397766,
1730
+ "rewards/rejected": -2.8864216804504395,
1731
+ "step": 1220
1732
+ },
1733
+ {
1734
+ "epoch": 0.98,
1735
+ "learning_rate": 3.8980895450474455e-09,
1736
+ "logits/chosen": -1.6516891717910767,
1737
+ "logits/rejected": -1.3506808280944824,
1738
+ "logps/chosen": -2481.84228515625,
1739
+ "logps/rejected": -3454.458984375,
1740
+ "loss": 0.4893,
1741
+ "rewards/accuracies": 0.675000011920929,
1742
+ "rewards/chosen": -2.2125697135925293,
1743
+ "rewards/margins": 1.0183699131011963,
1744
+ "rewards/rejected": -3.2309398651123047,
1745
+ "step": 1230
1746
+ },
1747
+ {
1748
+ "epoch": 0.99,
1749
+ "learning_rate": 9.747123991141193e-10,
1750
+ "logits/chosen": -1.2762700319290161,
1751
+ "logits/rejected": -1.1082122325897217,
1752
+ "logps/chosen": -2387.552490234375,
1753
+ "logps/rejected": -3240.255615234375,
1754
+ "loss": 0.4634,
1755
+ "rewards/accuracies": 0.675000011920929,
1756
+ "rewards/chosen": -2.163398265838623,
1757
+ "rewards/margins": 0.8387178182601929,
1758
+ "rewards/rejected": -3.0021164417266846,
1759
+ "step": 1240
1760
+ },
1761
+ {
1762
+ "epoch": 1.0,
1763
+ "learning_rate": 0.0,
1764
+ "logits/chosen": -1.7946316003799438,
1765
+ "logits/rejected": -1.654803991317749,
1766
+ "logps/chosen": -2349.287109375,
1767
+ "logps/rejected": -2943.7216796875,
1768
+ "loss": 0.547,
1769
+ "rewards/accuracies": 0.5249999761581421,
1770
+ "rewards/chosen": -2.1279029846191406,
1771
+ "rewards/margins": 0.613613486289978,
1772
+ "rewards/rejected": -2.74151611328125,
1773
+ "step": 1250
1774
+ },
1775
+ {
1776
+ "epoch": 1.0,
1777
+ "step": 1250,
1778
+ "total_flos": 0.0,
1779
+ "train_loss": 0.5607085178375244,
1780
+ "train_runtime": 13233.1545,
1781
+ "train_samples_per_second": 1.134,
1782
+ "train_steps_per_second": 0.094
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
+ }