lole25 commited on
Commit
3a19861
1 Parent(s): 655ade3

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-2/zephyr-7b-irepo-new-i2
9
+ model-index:
10
+ - name: zephyr-7b-gpo-v4-i3
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-gpo-v4-i3
18
+
19
+ This model is a fine-tuned version of [DUAL-GPO-2/zephyr-7b-irepo-new-i2](https://huggingface.co/DUAL-GPO-2/zephyr-7b-irepo-new-i2) 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: 2
44
+ - gradient_accumulation_steps: 2
45
+ - total_train_batch_size: 8
46
+ - total_eval_batch_size: 4
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+cu121
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 sha256:169061d4e597b774b753856be5f3a6a7312616258c4974500001582a4ae12db4
3
  size 671150064
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c706c4e03ee250bbb1bd6916bb7718d4c238f6649728442ec6c5f42ec324dc9e
3
  size 671150064
all_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "train_loss": 0.17237209237416584,
4
+ "train_runtime": 10470.2127,
5
+ "train_samples": 15000,
6
+ "train_samples_per_second": 1.433,
7
+ "train_steps_per_second": 0.179
8
+ }
runs/May14_14-57-16_gpu4-119-5/events.out.tfevents.1715662743.gpu4-119-5.1596457.0 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:3a9b30e0567467bbcea404faed4ba43b9487fb18084b1c29b6e4c28a090a4ebd
3
- size 119172
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e163d783929aee1272b10ff8c33d32964b2107ce9506690bc995476827f0f2aa
3
+ size 123964
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 1.0,
3
+ "train_loss": 0.17237209237416584,
4
+ "train_runtime": 10470.2127,
5
+ "train_samples": 15000,
6
+ "train_samples_per_second": 1.433,
7
+ "train_steps_per_second": 0.179
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,2662 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 1.0,
5
+ "eval_steps": 500,
6
+ "global_step": 1875,
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": 2.6595744680851065e-08,
14
+ "logits/chosen": -1.755455493927002,
15
+ "logits/rejected": -1.2333192825317383,
16
+ "logps/chosen": -764.8233032226562,
17
+ "logps/rejected": -1520.939208984375,
18
+ "loss": 0.3704,
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": 2.6595744680851066e-07,
28
+ "logits/chosen": -1.6312116384506226,
29
+ "logits/rejected": -0.958756148815155,
30
+ "logps/chosen": -542.477294921875,
31
+ "logps/rejected": -1220.0614013671875,
32
+ "loss": 0.3744,
33
+ "rewards/accuracies": 0.4444444477558136,
34
+ "rewards/chosen": 0.00022326533508021384,
35
+ "rewards/margins": 0.00027514772955328226,
36
+ "rewards/rejected": -5.1882434490835294e-05,
37
+ "step": 10
38
+ },
39
+ {
40
+ "epoch": 0.01,
41
+ "learning_rate": 5.319148936170213e-07,
42
+ "logits/chosen": -1.584674596786499,
43
+ "logits/rejected": -0.943401038646698,
44
+ "logps/chosen": -586.7803344726562,
45
+ "logps/rejected": -1201.489501953125,
46
+ "loss": 0.3553,
47
+ "rewards/accuracies": 0.625,
48
+ "rewards/chosen": 0.0002584285684861243,
49
+ "rewards/margins": 0.00131960678845644,
50
+ "rewards/rejected": -0.0010611782781779766,
51
+ "step": 20
52
+ },
53
+ {
54
+ "epoch": 0.02,
55
+ "learning_rate": 7.97872340425532e-07,
56
+ "logits/chosen": -1.378525972366333,
57
+ "logits/rejected": -0.8355947732925415,
58
+ "logps/chosen": -618.3253173828125,
59
+ "logps/rejected": -1216.331298828125,
60
+ "loss": 0.3435,
61
+ "rewards/accuracies": 0.6000000238418579,
62
+ "rewards/chosen": 0.003100637812167406,
63
+ "rewards/margins": 0.004315788391977549,
64
+ "rewards/rejected": -0.0012151505798101425,
65
+ "step": 30
66
+ },
67
+ {
68
+ "epoch": 0.02,
69
+ "learning_rate": 1.0638297872340427e-06,
70
+ "logits/chosen": -1.4447519779205322,
71
+ "logits/rejected": -0.8539841771125793,
72
+ "logps/chosen": -536.504638671875,
73
+ "logps/rejected": -1172.6549072265625,
74
+ "loss": 0.332,
75
+ "rewards/accuracies": 0.574999988079071,
76
+ "rewards/chosen": 0.00205782288685441,
77
+ "rewards/margins": 0.010990055277943611,
78
+ "rewards/rejected": -0.008932231925427914,
79
+ "step": 40
80
+ },
81
+ {
82
+ "epoch": 0.03,
83
+ "learning_rate": 1.3297872340425533e-06,
84
+ "logits/chosen": -1.7163407802581787,
85
+ "logits/rejected": -1.1005187034606934,
86
+ "logps/chosen": -618.8663940429688,
87
+ "logps/rejected": -1187.5242919921875,
88
+ "loss": 0.3116,
89
+ "rewards/accuracies": 0.625,
90
+ "rewards/chosen": -0.00023196739493869245,
91
+ "rewards/margins": 0.028941016644239426,
92
+ "rewards/rejected": -0.029172983020544052,
93
+ "step": 50
94
+ },
95
+ {
96
+ "epoch": 0.03,
97
+ "learning_rate": 1.595744680851064e-06,
98
+ "logits/chosen": -1.7239458560943604,
99
+ "logits/rejected": -0.5242995023727417,
100
+ "logps/chosen": -738.6119384765625,
101
+ "logps/rejected": -1377.3330078125,
102
+ "loss": 0.2459,
103
+ "rewards/accuracies": 0.7749999761581421,
104
+ "rewards/chosen": -0.06412681937217712,
105
+ "rewards/margins": 0.0683993324637413,
106
+ "rewards/rejected": -0.13252617418766022,
107
+ "step": 60
108
+ },
109
+ {
110
+ "epoch": 0.04,
111
+ "learning_rate": 1.8617021276595745e-06,
112
+ "logits/chosen": -1.3319450616836548,
113
+ "logits/rejected": -0.8398796916007996,
114
+ "logps/chosen": -707.7894897460938,
115
+ "logps/rejected": -1520.5997314453125,
116
+ "loss": 0.198,
117
+ "rewards/accuracies": 0.675000011920929,
118
+ "rewards/chosen": -0.16563495993614197,
119
+ "rewards/margins": 0.17894069850444794,
120
+ "rewards/rejected": -0.3445756733417511,
121
+ "step": 70
122
+ },
123
+ {
124
+ "epoch": 0.04,
125
+ "learning_rate": 2.1276595744680853e-06,
126
+ "logits/chosen": -1.4981276988983154,
127
+ "logits/rejected": -0.41109055280685425,
128
+ "logps/chosen": -937.0779418945312,
129
+ "logps/rejected": -2068.221923828125,
130
+ "loss": 0.2151,
131
+ "rewards/accuracies": 0.8500000238418579,
132
+ "rewards/chosen": -0.28319233655929565,
133
+ "rewards/margins": 0.40604227781295776,
134
+ "rewards/rejected": -0.6892346143722534,
135
+ "step": 80
136
+ },
137
+ {
138
+ "epoch": 0.05,
139
+ "learning_rate": 2.393617021276596e-06,
140
+ "logits/chosen": -1.4437072277069092,
141
+ "logits/rejected": -0.4041782021522522,
142
+ "logps/chosen": -608.6187744140625,
143
+ "logps/rejected": -1449.1143798828125,
144
+ "loss": 0.2497,
145
+ "rewards/accuracies": 0.824999988079071,
146
+ "rewards/chosen": -0.05350279062986374,
147
+ "rewards/margins": 0.12850181758403778,
148
+ "rewards/rejected": -0.18200460076332092,
149
+ "step": 90
150
+ },
151
+ {
152
+ "epoch": 0.05,
153
+ "learning_rate": 2.6595744680851065e-06,
154
+ "logits/chosen": -1.4004117250442505,
155
+ "logits/rejected": -0.3795866072177887,
156
+ "logps/chosen": -688.6510009765625,
157
+ "logps/rejected": -1484.7867431640625,
158
+ "loss": 0.2483,
159
+ "rewards/accuracies": 0.800000011920929,
160
+ "rewards/chosen": -0.07589573413133621,
161
+ "rewards/margins": 0.10643702745437622,
162
+ "rewards/rejected": -0.18233272433280945,
163
+ "step": 100
164
+ },
165
+ {
166
+ "epoch": 0.06,
167
+ "learning_rate": 2.9255319148936174e-06,
168
+ "logits/chosen": -1.4726046323776245,
169
+ "logits/rejected": -0.5685114860534668,
170
+ "logps/chosen": -926.9375,
171
+ "logps/rejected": -1812.751220703125,
172
+ "loss": 0.2093,
173
+ "rewards/accuracies": 0.824999988079071,
174
+ "rewards/chosen": -0.21614733338356018,
175
+ "rewards/margins": 0.26327216625213623,
176
+ "rewards/rejected": -0.47941941022872925,
177
+ "step": 110
178
+ },
179
+ {
180
+ "epoch": 0.06,
181
+ "learning_rate": 3.191489361702128e-06,
182
+ "logits/chosen": -1.2898015975952148,
183
+ "logits/rejected": -0.20202858746051788,
184
+ "logps/chosen": -857.4860229492188,
185
+ "logps/rejected": -1566.881591796875,
186
+ "loss": 0.2628,
187
+ "rewards/accuracies": 0.574999988079071,
188
+ "rewards/chosen": -0.2838597893714905,
189
+ "rewards/margins": 0.172959566116333,
190
+ "rewards/rejected": -0.4568193852901459,
191
+ "step": 120
192
+ },
193
+ {
194
+ "epoch": 0.07,
195
+ "learning_rate": 3.457446808510639e-06,
196
+ "logits/chosen": -1.3372188806533813,
197
+ "logits/rejected": -0.5857396125793457,
198
+ "logps/chosen": -738.6064453125,
199
+ "logps/rejected": -1646.12890625,
200
+ "loss": 0.21,
201
+ "rewards/accuracies": 0.7250000238418579,
202
+ "rewards/chosen": -0.1405295431613922,
203
+ "rewards/margins": 0.23650164902210236,
204
+ "rewards/rejected": -0.377031147480011,
205
+ "step": 130
206
+ },
207
+ {
208
+ "epoch": 0.07,
209
+ "learning_rate": 3.723404255319149e-06,
210
+ "logits/chosen": -1.397938847541809,
211
+ "logits/rejected": -0.5400180220603943,
212
+ "logps/chosen": -723.1727294921875,
213
+ "logps/rejected": -1678.752197265625,
214
+ "loss": 0.2137,
215
+ "rewards/accuracies": 0.800000011920929,
216
+ "rewards/chosen": -0.15040160715579987,
217
+ "rewards/margins": 0.28295016288757324,
218
+ "rewards/rejected": -0.4333517551422119,
219
+ "step": 140
220
+ },
221
+ {
222
+ "epoch": 0.08,
223
+ "learning_rate": 3.98936170212766e-06,
224
+ "logits/chosen": -1.2215166091918945,
225
+ "logits/rejected": -0.5187689065933228,
226
+ "logps/chosen": -913.1868286132812,
227
+ "logps/rejected": -2170.439697265625,
228
+ "loss": 0.1472,
229
+ "rewards/accuracies": 0.8500000238418579,
230
+ "rewards/chosen": -0.33379653096199036,
231
+ "rewards/margins": 0.4125414788722992,
232
+ "rewards/rejected": -0.7463380098342896,
233
+ "step": 150
234
+ },
235
+ {
236
+ "epoch": 0.09,
237
+ "learning_rate": 4.255319148936171e-06,
238
+ "logits/chosen": -1.2553513050079346,
239
+ "logits/rejected": -0.6222286224365234,
240
+ "logps/chosen": -644.9315795898438,
241
+ "logps/rejected": -1344.428466796875,
242
+ "loss": 0.2055,
243
+ "rewards/accuracies": 0.6000000238418579,
244
+ "rewards/chosen": -0.16092197597026825,
245
+ "rewards/margins": 0.18404754996299744,
246
+ "rewards/rejected": -0.34496957063674927,
247
+ "step": 160
248
+ },
249
+ {
250
+ "epoch": 0.09,
251
+ "learning_rate": 4.521276595744681e-06,
252
+ "logits/chosen": -1.1327978372573853,
253
+ "logits/rejected": -0.43405628204345703,
254
+ "logps/chosen": -704.5630493164062,
255
+ "logps/rejected": -1811.0263671875,
256
+ "loss": 0.1446,
257
+ "rewards/accuracies": 0.925000011920929,
258
+ "rewards/chosen": -0.1263386309146881,
259
+ "rewards/margins": 0.3248198926448822,
260
+ "rewards/rejected": -0.4511585235595703,
261
+ "step": 170
262
+ },
263
+ {
264
+ "epoch": 0.1,
265
+ "learning_rate": 4.787234042553192e-06,
266
+ "logits/chosen": -1.479241967201233,
267
+ "logits/rejected": -0.397353857755661,
268
+ "logps/chosen": -708.5093994140625,
269
+ "logps/rejected": -1561.05078125,
270
+ "loss": 0.1997,
271
+ "rewards/accuracies": 0.7250000238418579,
272
+ "rewards/chosen": -0.15144303441047668,
273
+ "rewards/margins": 0.2386520355939865,
274
+ "rewards/rejected": -0.3900950849056244,
275
+ "step": 180
276
+ },
277
+ {
278
+ "epoch": 0.1,
279
+ "learning_rate": 4.999982660399688e-06,
280
+ "logits/chosen": -1.1840440034866333,
281
+ "logits/rejected": -0.4053446650505066,
282
+ "logps/chosen": -822.2158203125,
283
+ "logps/rejected": -1771.717529296875,
284
+ "loss": 0.2122,
285
+ "rewards/accuracies": 0.800000011920929,
286
+ "rewards/chosen": -0.2102956473827362,
287
+ "rewards/margins": 0.2532525956630707,
288
+ "rewards/rejected": -0.4635482430458069,
289
+ "step": 190
290
+ },
291
+ {
292
+ "epoch": 0.11,
293
+ "learning_rate": 4.99937579964398e-06,
294
+ "logits/chosen": -1.3526380062103271,
295
+ "logits/rejected": -0.5665386915206909,
296
+ "logps/chosen": -773.0138549804688,
297
+ "logps/rejected": -1730.4918212890625,
298
+ "loss": 0.2076,
299
+ "rewards/accuracies": 0.75,
300
+ "rewards/chosen": -0.1651298850774765,
301
+ "rewards/margins": 0.24976544082164764,
302
+ "rewards/rejected": -0.41489535570144653,
303
+ "step": 200
304
+ },
305
+ {
306
+ "epoch": 0.11,
307
+ "learning_rate": 4.9979021993870645e-06,
308
+ "logits/chosen": -1.4962358474731445,
309
+ "logits/rejected": -0.5578527450561523,
310
+ "logps/chosen": -795.4863891601562,
311
+ "logps/rejected": -1807.668212890625,
312
+ "loss": 0.1634,
313
+ "rewards/accuracies": 0.824999988079071,
314
+ "rewards/chosen": -0.15764664113521576,
315
+ "rewards/margins": 0.2677456736564636,
316
+ "rewards/rejected": -0.4253923296928406,
317
+ "step": 210
318
+ },
319
+ {
320
+ "epoch": 0.12,
321
+ "learning_rate": 4.995562370647553e-06,
322
+ "logits/chosen": -0.9666210412979126,
323
+ "logits/rejected": -0.4786294102668762,
324
+ "logps/chosen": -786.662353515625,
325
+ "logps/rejected": -1778.3988037109375,
326
+ "loss": 0.2532,
327
+ "rewards/accuracies": 0.800000011920929,
328
+ "rewards/chosen": -0.24228081107139587,
329
+ "rewards/margins": 0.3355843126773834,
330
+ "rewards/rejected": -0.5778651833534241,
331
+ "step": 220
332
+ },
333
+ {
334
+ "epoch": 0.12,
335
+ "learning_rate": 4.992357124836838e-06,
336
+ "logits/chosen": -1.409147024154663,
337
+ "logits/rejected": -0.689926028251648,
338
+ "logps/chosen": -877.3094482421875,
339
+ "logps/rejected": -1852.2314453125,
340
+ "loss": 0.1728,
341
+ "rewards/accuracies": 0.824999988079071,
342
+ "rewards/chosen": -0.2839181125164032,
343
+ "rewards/margins": 0.23368731141090393,
344
+ "rewards/rejected": -0.5176054239273071,
345
+ "step": 230
346
+ },
347
+ {
348
+ "epoch": 0.13,
349
+ "learning_rate": 4.9882875734777044e-06,
350
+ "logits/chosen": -1.1925780773162842,
351
+ "logits/rejected": 0.03326777368783951,
352
+ "logps/chosen": -831.4713134765625,
353
+ "logps/rejected": -1737.231689453125,
354
+ "loss": 0.2304,
355
+ "rewards/accuracies": 0.7749999761581421,
356
+ "rewards/chosen": -0.27825552225112915,
357
+ "rewards/margins": 0.28924351930618286,
358
+ "rewards/rejected": -0.567499041557312,
359
+ "step": 240
360
+ },
361
+ {
362
+ "epoch": 0.13,
363
+ "learning_rate": 4.983355127818882e-06,
364
+ "logits/chosen": -1.4935053586959839,
365
+ "logits/rejected": -0.0941208004951477,
366
+ "logps/chosen": -759.3983154296875,
367
+ "logps/rejected": -1830.8707275390625,
368
+ "loss": 0.2091,
369
+ "rewards/accuracies": 0.824999988079071,
370
+ "rewards/chosen": -0.13234801590442657,
371
+ "rewards/margins": 0.29578104615211487,
372
+ "rewards/rejected": -0.428129106760025,
373
+ "step": 250
374
+ },
375
+ {
376
+ "epoch": 0.14,
377
+ "learning_rate": 4.977561498345639e-06,
378
+ "logits/chosen": -1.4366835355758667,
379
+ "logits/rejected": -0.848114013671875,
380
+ "logps/chosen": -756.9669799804688,
381
+ "logps/rejected": -1844.3541259765625,
382
+ "loss": 0.1589,
383
+ "rewards/accuracies": 0.800000011920929,
384
+ "rewards/chosen": -0.2339516133069992,
385
+ "rewards/margins": 0.3356839716434479,
386
+ "rewards/rejected": -0.5696356296539307,
387
+ "step": 260
388
+ },
389
+ {
390
+ "epoch": 0.14,
391
+ "learning_rate": 4.970908694186624e-06,
392
+ "logits/chosen": -1.433019995689392,
393
+ "logits/rejected": -0.6539562940597534,
394
+ "logps/chosen": -852.9888916015625,
395
+ "logps/rejected": -1876.3004150390625,
396
+ "loss": 0.1709,
397
+ "rewards/accuracies": 0.8500000238418579,
398
+ "rewards/chosen": -0.2885374426841736,
399
+ "rewards/margins": 0.3317956030368805,
400
+ "rewards/rejected": -0.6203330755233765,
401
+ "step": 270
402
+ },
403
+ {
404
+ "epoch": 0.15,
405
+ "learning_rate": 4.9633990224171305e-06,
406
+ "logits/chosen": -1.1937869787216187,
407
+ "logits/rejected": -0.09999962896108627,
408
+ "logps/chosen": -1032.603515625,
409
+ "logps/rejected": -2027.5093994140625,
410
+ "loss": 0.2166,
411
+ "rewards/accuracies": 0.8500000238418579,
412
+ "rewards/chosen": -0.36031240224838257,
413
+ "rewards/margins": 0.3707783818244934,
414
+ "rewards/rejected": -0.7310907244682312,
415
+ "step": 280
416
+ },
417
+ {
418
+ "epoch": 0.15,
419
+ "learning_rate": 4.955035087259046e-06,
420
+ "logits/chosen": -1.2412487268447876,
421
+ "logits/rejected": -0.664631724357605,
422
+ "logps/chosen": -734.98681640625,
423
+ "logps/rejected": -1704.20703125,
424
+ "loss": 0.1807,
425
+ "rewards/accuracies": 0.875,
426
+ "rewards/chosen": -0.14173075556755066,
427
+ "rewards/margins": 0.25538477301597595,
428
+ "rewards/rejected": -0.3971155285835266,
429
+ "step": 290
430
+ },
431
+ {
432
+ "epoch": 0.16,
433
+ "learning_rate": 4.945819789177756e-06,
434
+ "logits/chosen": -1.1638051271438599,
435
+ "logits/rejected": 0.49103036522865295,
436
+ "logps/chosen": -755.9884643554688,
437
+ "logps/rejected": -1692.7734375,
438
+ "loss": 0.1976,
439
+ "rewards/accuracies": 0.800000011920929,
440
+ "rewards/chosen": -0.16941963136196136,
441
+ "rewards/margins": 0.23666927218437195,
442
+ "rewards/rejected": -0.4060889184474945,
443
+ "step": 300
444
+ },
445
+ {
446
+ "epoch": 0.17,
447
+ "learning_rate": 4.935756323876306e-06,
448
+ "logits/chosen": -1.1552793979644775,
449
+ "logits/rejected": 0.35435959696769714,
450
+ "logps/chosen": -895.009765625,
451
+ "logps/rejected": -1757.65234375,
452
+ "loss": 0.1955,
453
+ "rewards/accuracies": 0.7749999761581421,
454
+ "rewards/chosen": -0.2875288426876068,
455
+ "rewards/margins": 0.24231036007404327,
456
+ "rewards/rejected": -0.5298392176628113,
457
+ "step": 310
458
+ },
459
+ {
460
+ "epoch": 0.17,
461
+ "learning_rate": 4.924848181187199e-06,
462
+ "logits/chosen": -1.3235706090927124,
463
+ "logits/rejected": -0.7385076284408569,
464
+ "logps/chosen": -643.3860473632812,
465
+ "logps/rejected": -1514.3936767578125,
466
+ "loss": 0.2013,
467
+ "rewards/accuracies": 0.75,
468
+ "rewards/chosen": -0.1353076845407486,
469
+ "rewards/margins": 0.21510222554206848,
470
+ "rewards/rejected": -0.3504098951816559,
471
+ "step": 320
472
+ },
473
+ {
474
+ "epoch": 0.18,
475
+ "learning_rate": 4.913099143862173e-06,
476
+ "logits/chosen": -1.42311429977417,
477
+ "logits/rejected": -0.19915446639060974,
478
+ "logps/chosen": -806.025390625,
479
+ "logps/rejected": -1748.1644287109375,
480
+ "loss": 0.1834,
481
+ "rewards/accuracies": 0.800000011920929,
482
+ "rewards/chosen": -0.16057676076889038,
483
+ "rewards/margins": 0.29953283071517944,
484
+ "rewards/rejected": -0.4601096212863922,
485
+ "step": 330
486
+ },
487
+ {
488
+ "epoch": 0.18,
489
+ "learning_rate": 4.900513286260416e-06,
490
+ "logits/chosen": -1.4523063898086548,
491
+ "logits/rejected": -0.7241362929344177,
492
+ "logps/chosen": -726.6815185546875,
493
+ "logps/rejected": -1756.1654052734375,
494
+ "loss": 0.1909,
495
+ "rewards/accuracies": 0.8500000238418579,
496
+ "rewards/chosen": -0.1610650271177292,
497
+ "rewards/margins": 0.32662302255630493,
498
+ "rewards/rejected": -0.48768800497055054,
499
+ "step": 340
500
+ },
501
+ {
502
+ "epoch": 0.19,
503
+ "learning_rate": 4.887094972935645e-06,
504
+ "logits/chosen": -1.505131721496582,
505
+ "logits/rejected": -0.7010393738746643,
506
+ "logps/chosen": -756.3118896484375,
507
+ "logps/rejected": -1657.638427734375,
508
+ "loss": 0.1529,
509
+ "rewards/accuracies": 0.7749999761581421,
510
+ "rewards/chosen": -0.19148796796798706,
511
+ "rewards/margins": 0.26837441325187683,
512
+ "rewards/rejected": -0.4598624110221863,
513
+ "step": 350
514
+ },
515
+ {
516
+ "epoch": 0.19,
517
+ "learning_rate": 4.87284885712256e-06,
518
+ "logits/chosen": -1.400520920753479,
519
+ "logits/rejected": -0.24144461750984192,
520
+ "logps/chosen": -773.6839599609375,
521
+ "logps/rejected": -1939.297119140625,
522
+ "loss": 0.1325,
523
+ "rewards/accuracies": 0.925000011920929,
524
+ "rewards/chosen": -0.20408935844898224,
525
+ "rewards/margins": 0.379780113697052,
526
+ "rewards/rejected": -0.583869457244873,
527
+ "step": 360
528
+ },
529
+ {
530
+ "epoch": 0.2,
531
+ "learning_rate": 4.857779879123181e-06,
532
+ "logits/chosen": -1.3075954914093018,
533
+ "logits/rejected": -0.6242688894271851,
534
+ "logps/chosen": -830.1097412109375,
535
+ "logps/rejected": -1820.2388916015625,
536
+ "loss": 0.214,
537
+ "rewards/accuracies": 0.75,
538
+ "rewards/chosen": -0.3089172840118408,
539
+ "rewards/margins": 0.2943463921546936,
540
+ "rewards/rejected": -0.6032636761665344,
541
+ "step": 370
542
+ },
543
+ {
544
+ "epoch": 0.2,
545
+ "learning_rate": 4.841893264593643e-06,
546
+ "logits/chosen": -1.3646913766860962,
547
+ "logits/rejected": -0.4420408606529236,
548
+ "logps/chosen": -739.6143188476562,
549
+ "logps/rejected": -1830.6683349609375,
550
+ "loss": 0.1772,
551
+ "rewards/accuracies": 0.875,
552
+ "rewards/chosen": -0.1528651863336563,
553
+ "rewards/margins": 0.3289165496826172,
554
+ "rewards/rejected": -0.4817817211151123,
555
+ "step": 380
556
+ },
557
+ {
558
+ "epoch": 0.21,
559
+ "learning_rate": 4.825194522732023e-06,
560
+ "logits/chosen": -1.394213318824768,
561
+ "logits/rejected": -0.41561856865882874,
562
+ "logps/chosen": -848.2108154296875,
563
+ "logps/rejected": -1736.9312744140625,
564
+ "loss": 0.1866,
565
+ "rewards/accuracies": 0.8500000238418579,
566
+ "rewards/chosen": -0.17288437485694885,
567
+ "rewards/margins": 0.2597029209136963,
568
+ "rewards/rejected": -0.43258723616600037,
569
+ "step": 390
570
+ },
571
+ {
572
+ "epoch": 0.21,
573
+ "learning_rate": 4.807689444367853e-06,
574
+ "logits/chosen": -1.4238224029541016,
575
+ "logits/rejected": -0.2904614210128784,
576
+ "logps/chosen": -766.7713623046875,
577
+ "logps/rejected": -1548.4310302734375,
578
+ "loss": 0.1551,
579
+ "rewards/accuracies": 0.699999988079071,
580
+ "rewards/chosen": -0.18976250290870667,
581
+ "rewards/margins": 0.24860653281211853,
582
+ "rewards/rejected": -0.4383690357208252,
583
+ "step": 400
584
+ },
585
+ {
586
+ "epoch": 0.22,
587
+ "learning_rate": 4.78938409995396e-06,
588
+ "logits/chosen": -1.5596516132354736,
589
+ "logits/rejected": -0.19419638812541962,
590
+ "logps/chosen": -894.0138549804688,
591
+ "logps/rejected": -1879.2919921875,
592
+ "loss": 0.1543,
593
+ "rewards/accuracies": 0.800000011920929,
594
+ "rewards/chosen": -0.21909482777118683,
595
+ "rewards/margins": 0.3173079490661621,
596
+ "rewards/rejected": -0.5364028215408325,
597
+ "step": 410
598
+ },
599
+ {
600
+ "epoch": 0.22,
601
+ "learning_rate": 4.770284837461342e-06,
602
+ "logits/chosen": -1.468480110168457,
603
+ "logits/rejected": -0.7513076663017273,
604
+ "logps/chosen": -791.5577392578125,
605
+ "logps/rejected": -1696.2337646484375,
606
+ "loss": 0.1794,
607
+ "rewards/accuracies": 0.7250000238418579,
608
+ "rewards/chosen": -0.16900746524333954,
609
+ "rewards/margins": 0.270784854888916,
610
+ "rewards/rejected": -0.43979233503341675,
611
+ "step": 420
612
+ },
613
+ {
614
+ "epoch": 0.23,
615
+ "learning_rate": 4.7503982801778015e-06,
616
+ "logits/chosen": -1.4179691076278687,
617
+ "logits/rejected": -0.012906777672469616,
618
+ "logps/chosen": -816.0245361328125,
619
+ "logps/rejected": -1743.844482421875,
620
+ "loss": 0.1472,
621
+ "rewards/accuracies": 0.800000011920929,
622
+ "rewards/chosen": -0.1716371476650238,
623
+ "rewards/margins": 0.29810577630996704,
624
+ "rewards/rejected": -0.46974295377731323,
625
+ "step": 430
626
+ },
627
+ {
628
+ "epoch": 0.23,
629
+ "learning_rate": 4.729731324411104e-06,
630
+ "logits/chosen": -1.447142481803894,
631
+ "logits/rejected": -0.6321523189544678,
632
+ "logps/chosen": -683.079833984375,
633
+ "logps/rejected": -1682.686767578125,
634
+ "loss": 0.1643,
635
+ "rewards/accuracies": 0.75,
636
+ "rewards/chosen": -0.1491968333721161,
637
+ "rewards/margins": 0.3515523076057434,
638
+ "rewards/rejected": -0.5007491111755371,
639
+ "step": 440
640
+ },
641
+ {
642
+ "epoch": 0.24,
643
+ "learning_rate": 4.7082911370974645e-06,
644
+ "logits/chosen": -0.9937525987625122,
645
+ "logits/rejected": -0.6763631105422974,
646
+ "logps/chosen": -770.6049194335938,
647
+ "logps/rejected": -1687.3837890625,
648
+ "loss": 0.1541,
649
+ "rewards/accuracies": 0.8500000238418579,
650
+ "rewards/chosen": -0.17300094664096832,
651
+ "rewards/margins": 0.26958489418029785,
652
+ "rewards/rejected": -0.44258585572242737,
653
+ "step": 450
654
+ },
655
+ {
656
+ "epoch": 0.25,
657
+ "learning_rate": 4.68608515331618e-06,
658
+ "logits/chosen": -1.2351852655410767,
659
+ "logits/rejected": -0.5260038375854492,
660
+ "logps/chosen": -727.1993408203125,
661
+ "logps/rejected": -1712.5172119140625,
662
+ "loss": 0.1596,
663
+ "rewards/accuracies": 0.875,
664
+ "rewards/chosen": -0.10707534849643707,
665
+ "rewards/margins": 0.32161325216293335,
666
+ "rewards/rejected": -0.4286886155605316,
667
+ "step": 460
668
+ },
669
+ {
670
+ "epoch": 0.25,
671
+ "learning_rate": 4.663121073711269e-06,
672
+ "logits/chosen": -1.4281326532363892,
673
+ "logits/rejected": -0.11814825236797333,
674
+ "logps/chosen": -732.6151123046875,
675
+ "logps/rejected": -1736.462158203125,
676
+ "loss": 0.1742,
677
+ "rewards/accuracies": 0.8999999761581421,
678
+ "rewards/chosen": -0.10931523144245148,
679
+ "rewards/margins": 0.2831416428089142,
680
+ "rewards/rejected": -0.39245691895484924,
681
+ "step": 470
682
+ },
683
+ {
684
+ "epoch": 0.26,
685
+ "learning_rate": 4.63940686182103e-06,
686
+ "logits/chosen": -1.366811990737915,
687
+ "logits/rejected": -0.15237736701965332,
688
+ "logps/chosen": -920.6950073242188,
689
+ "logps/rejected": -1818.9697265625,
690
+ "loss": 0.1964,
691
+ "rewards/accuracies": 0.800000011920929,
692
+ "rewards/chosen": -0.18746745586395264,
693
+ "rewards/margins": 0.3069246709346771,
694
+ "rewards/rejected": -0.4943920969963074,
695
+ "step": 480
696
+ },
697
+ {
698
+ "epoch": 0.26,
699
+ "learning_rate": 4.614950741316425e-06,
700
+ "logits/chosen": -1.2442119121551514,
701
+ "logits/rejected": -0.27264174818992615,
702
+ "logps/chosen": -577.3939208984375,
703
+ "logps/rejected": -1474.958984375,
704
+ "loss": 0.1644,
705
+ "rewards/accuracies": 0.824999988079071,
706
+ "rewards/chosen": -0.06222609430551529,
707
+ "rewards/margins": 0.20860283076763153,
708
+ "rewards/rejected": -0.27082890272140503,
709
+ "step": 490
710
+ },
711
+ {
712
+ "epoch": 0.27,
713
+ "learning_rate": 4.589761193149254e-06,
714
+ "logits/chosen": -1.2685329914093018,
715
+ "logits/rejected": -0.32444876432418823,
716
+ "logps/chosen": -788.1159057617188,
717
+ "logps/rejected": -1656.401123046875,
718
+ "loss": 0.2214,
719
+ "rewards/accuracies": 0.8500000238418579,
720
+ "rewards/chosen": -0.10539279133081436,
721
+ "rewards/margins": 0.28618016839027405,
722
+ "rewards/rejected": -0.3915729224681854,
723
+ "step": 500
724
+ },
725
+ {
726
+ "epoch": 0.27,
727
+ "learning_rate": 4.563846952611112e-06,
728
+ "logits/chosen": -1.1497045755386353,
729
+ "logits/rejected": -0.3311644494533539,
730
+ "logps/chosen": -789.111328125,
731
+ "logps/rejected": -1705.7972412109375,
732
+ "loss": 0.173,
733
+ "rewards/accuracies": 0.824999988079071,
734
+ "rewards/chosen": -0.14451605081558228,
735
+ "rewards/margins": 0.2676469087600708,
736
+ "rewards/rejected": -0.4121629595756531,
737
+ "step": 510
738
+ },
739
+ {
740
+ "epoch": 0.28,
741
+ "learning_rate": 4.537217006304141e-06,
742
+ "logits/chosen": -1.1343199014663696,
743
+ "logits/rejected": 0.12552905082702637,
744
+ "logps/chosen": -782.1867065429688,
745
+ "logps/rejected": -1627.5347900390625,
746
+ "loss": 0.1644,
747
+ "rewards/accuracies": 0.800000011920929,
748
+ "rewards/chosen": -0.15862300992012024,
749
+ "rewards/margins": 0.2545243203639984,
750
+ "rewards/rejected": -0.41314736008644104,
751
+ "step": 520
752
+ },
753
+ {
754
+ "epoch": 0.28,
755
+ "learning_rate": 4.50988058902464e-06,
756
+ "logits/chosen": -1.36593496799469,
757
+ "logits/rejected": -0.3004814088344574,
758
+ "logps/chosen": -754.1193237304688,
759
+ "logps/rejected": -1613.5533447265625,
760
+ "loss": 0.1722,
761
+ "rewards/accuracies": 0.75,
762
+ "rewards/chosen": -0.13889679312705994,
763
+ "rewards/margins": 0.2828737497329712,
764
+ "rewards/rejected": -0.4217705726623535,
765
+ "step": 530
766
+ },
767
+ {
768
+ "epoch": 0.29,
769
+ "learning_rate": 4.481847180560593e-06,
770
+ "logits/chosen": -1.5362759828567505,
771
+ "logits/rejected": -0.10081689059734344,
772
+ "logps/chosen": -763.066162109375,
773
+ "logps/rejected": -1727.967529296875,
774
+ "loss": 0.2263,
775
+ "rewards/accuracies": 0.824999988079071,
776
+ "rewards/chosen": -0.11022988706827164,
777
+ "rewards/margins": 0.2953043580055237,
778
+ "rewards/rejected": -0.40553420782089233,
779
+ "step": 540
780
+ },
781
+ {
782
+ "epoch": 0.29,
783
+ "learning_rate": 4.453126502404253e-06,
784
+ "logits/chosen": -1.2658751010894775,
785
+ "logits/rejected": -0.15121665596961975,
786
+ "logps/chosen": -676.2111206054688,
787
+ "logps/rejected": -1652.478515625,
788
+ "loss": 0.1659,
789
+ "rewards/accuracies": 0.824999988079071,
790
+ "rewards/chosen": -0.15621882677078247,
791
+ "rewards/margins": 0.28227540850639343,
792
+ "rewards/rejected": -0.4384942650794983,
793
+ "step": 550
794
+ },
795
+ {
796
+ "epoch": 0.3,
797
+ "learning_rate": 4.423728514380892e-06,
798
+ "logits/chosen": -1.1152678728103638,
799
+ "logits/rejected": -0.4836166501045227,
800
+ "logps/chosen": -772.0936279296875,
801
+ "logps/rejected": -2011.451904296875,
802
+ "loss": 0.2331,
803
+ "rewards/accuracies": 0.8500000238418579,
804
+ "rewards/chosen": -0.19823512434959412,
805
+ "rewards/margins": 0.4028354585170746,
806
+ "rewards/rejected": -0.6010705828666687,
807
+ "step": 560
808
+ },
809
+ {
810
+ "epoch": 0.3,
811
+ "learning_rate": 4.393663411194918e-06,
812
+ "logits/chosen": -1.2210283279418945,
813
+ "logits/rejected": -0.5110222101211548,
814
+ "logps/chosen": -661.4767456054688,
815
+ "logps/rejected": -1744.774658203125,
816
+ "loss": 0.195,
817
+ "rewards/accuracies": 0.8500000238418579,
818
+ "rewards/chosen": -0.09630689024925232,
819
+ "rewards/margins": 0.3629889488220215,
820
+ "rewards/rejected": -0.4592958390712738,
821
+ "step": 570
822
+ },
823
+ {
824
+ "epoch": 0.31,
825
+ "learning_rate": 4.362941618894523e-06,
826
+ "logits/chosen": -1.4232590198516846,
827
+ "logits/rejected": -0.6951449513435364,
828
+ "logps/chosen": -777.7971801757812,
829
+ "logps/rejected": -1753.598388671875,
830
+ "loss": 0.1906,
831
+ "rewards/accuracies": 0.7749999761581421,
832
+ "rewards/chosen": -0.1481999158859253,
833
+ "rewards/margins": 0.3005923628807068,
834
+ "rewards/rejected": -0.4487922787666321,
835
+ "step": 580
836
+ },
837
+ {
838
+ "epoch": 0.31,
839
+ "learning_rate": 4.331573791256116e-06,
840
+ "logits/chosen": -1.2691911458969116,
841
+ "logits/rejected": -0.21553485095500946,
842
+ "logps/chosen": -850.1456298828125,
843
+ "logps/rejected": -1807.684814453125,
844
+ "loss": 0.1987,
845
+ "rewards/accuracies": 0.8999999761581421,
846
+ "rewards/chosen": -0.23677225410938263,
847
+ "rewards/margins": 0.27481013536453247,
848
+ "rewards/rejected": -0.5115823745727539,
849
+ "step": 590
850
+ },
851
+ {
852
+ "epoch": 0.32,
853
+ "learning_rate": 4.299570806089786e-06,
854
+ "logits/chosen": -1.2055295705795288,
855
+ "logits/rejected": -0.07821548730134964,
856
+ "logps/chosen": -801.0753173828125,
857
+ "logps/rejected": -1668.043212890625,
858
+ "loss": 0.1633,
859
+ "rewards/accuracies": 0.800000011920929,
860
+ "rewards/chosen": -0.21256402134895325,
861
+ "rewards/margins": 0.29289668798446655,
862
+ "rewards/rejected": -0.5054606795310974,
863
+ "step": 600
864
+ },
865
+ {
866
+ "epoch": 0.33,
867
+ "learning_rate": 4.266943761467057e-06,
868
+ "logits/chosen": -1.290138840675354,
869
+ "logits/rejected": 0.16648904979228973,
870
+ "logps/chosen": -796.8035888671875,
871
+ "logps/rejected": -1771.1650390625,
872
+ "loss": 0.1665,
873
+ "rewards/accuracies": 0.8999999761581421,
874
+ "rewards/chosen": -0.11760182678699493,
875
+ "rewards/margins": 0.294794499874115,
876
+ "rewards/rejected": -0.41239628195762634,
877
+ "step": 610
878
+ },
879
+ {
880
+ "epoch": 0.33,
881
+ "learning_rate": 4.233703971872287e-06,
882
+ "logits/chosen": -1.53248131275177,
883
+ "logits/rejected": 0.212439626455307,
884
+ "logps/chosen": -613.8381958007812,
885
+ "logps/rejected": -1503.5120849609375,
886
+ "loss": 0.1547,
887
+ "rewards/accuracies": 0.824999988079071,
888
+ "rewards/chosen": -0.053207941353321075,
889
+ "rewards/margins": 0.2783101201057434,
890
+ "rewards/rejected": -0.3315180242061615,
891
+ "step": 620
892
+ },
893
+ {
894
+ "epoch": 0.34,
895
+ "learning_rate": 4.1998629642789925e-06,
896
+ "logits/chosen": -1.56017005443573,
897
+ "logits/rejected": -0.071937195956707,
898
+ "logps/chosen": -674.8516845703125,
899
+ "logps/rejected": -1349.5697021484375,
900
+ "loss": 0.181,
901
+ "rewards/accuracies": 0.7250000238418579,
902
+ "rewards/chosen": -0.056226201355457306,
903
+ "rewards/margins": 0.21242019534111023,
904
+ "rewards/rejected": -0.26864638924598694,
905
+ "step": 630
906
+ },
907
+ {
908
+ "epoch": 0.34,
909
+ "learning_rate": 4.165432474152505e-06,
910
+ "logits/chosen": -1.2165509462356567,
911
+ "logits/rejected": -0.5167722702026367,
912
+ "logps/chosen": -677.8233032226562,
913
+ "logps/rejected": -1581.596923828125,
914
+ "loss": 0.1799,
915
+ "rewards/accuracies": 0.824999988079071,
916
+ "rewards/chosen": -0.10262588411569595,
917
+ "rewards/margins": 0.2893064618110657,
918
+ "rewards/rejected": -0.39193230867385864,
919
+ "step": 640
920
+ },
921
+ {
922
+ "epoch": 0.35,
923
+ "learning_rate": 4.130424441380308e-06,
924
+ "logits/chosen": -1.5206298828125,
925
+ "logits/rejected": -0.4772413372993469,
926
+ "logps/chosen": -721.204345703125,
927
+ "logps/rejected": -1608.91845703125,
928
+ "loss": 0.1771,
929
+ "rewards/accuracies": 0.824999988079071,
930
+ "rewards/chosen": -0.1459272801876068,
931
+ "rewards/margins": 0.2629932761192322,
932
+ "rewards/rejected": -0.4089205861091614,
933
+ "step": 650
934
+ },
935
+ {
936
+ "epoch": 0.35,
937
+ "learning_rate": 4.09485100613151e-06,
938
+ "logits/chosen": -1.1552391052246094,
939
+ "logits/rejected": -0.5302366614341736,
940
+ "logps/chosen": -844.49560546875,
941
+ "logps/rejected": -1849.5101318359375,
942
+ "loss": 0.1755,
943
+ "rewards/accuracies": 0.7749999761581421,
944
+ "rewards/chosen": -0.20096397399902344,
945
+ "rewards/margins": 0.2851325273513794,
946
+ "rewards/rejected": -0.4860965609550476,
947
+ "step": 660
948
+ },
949
+ {
950
+ "epoch": 0.36,
951
+ "learning_rate": 4.058724504646834e-06,
952
+ "logits/chosen": -1.205714225769043,
953
+ "logits/rejected": -0.6423812508583069,
954
+ "logps/chosen": -707.0100708007812,
955
+ "logps/rejected": -1621.2158203125,
956
+ "loss": 0.141,
957
+ "rewards/accuracies": 0.75,
958
+ "rewards/chosen": -0.09242536127567291,
959
+ "rewards/margins": 0.3333635926246643,
960
+ "rewards/rejected": -0.4257889688014984,
961
+ "step": 670
962
+ },
963
+ {
964
+ "epoch": 0.36,
965
+ "learning_rate": 4.022057464960632e-06,
966
+ "logits/chosen": -1.365877389907837,
967
+ "logits/rejected": 0.25049906969070435,
968
+ "logps/chosen": -719.96630859375,
969
+ "logps/rejected": -1713.641845703125,
970
+ "loss": 0.1621,
971
+ "rewards/accuracies": 0.8500000238418579,
972
+ "rewards/chosen": -0.0692489892244339,
973
+ "rewards/margins": 0.3098606467247009,
974
+ "rewards/rejected": -0.3791096806526184,
975
+ "step": 680
976
+ },
977
+ {
978
+ "epoch": 0.37,
979
+ "learning_rate": 3.984862602556383e-06,
980
+ "logits/chosen": -1.2444730997085571,
981
+ "logits/rejected": 0.32205188274383545,
982
+ "logps/chosen": -701.1004638671875,
983
+ "logps/rejected": -1603.306884765625,
984
+ "loss": 0.1723,
985
+ "rewards/accuracies": 0.875,
986
+ "rewards/chosen": -0.0724688395857811,
987
+ "rewards/margins": 0.2886015772819519,
988
+ "rewards/rejected": -0.3610703945159912,
989
+ "step": 690
990
+ },
991
+ {
992
+ "epoch": 0.37,
993
+ "learning_rate": 3.947152815957187e-06,
994
+ "logits/chosen": -1.2809414863586426,
995
+ "logits/rejected": 0.01291370578110218,
996
+ "logps/chosen": -631.6290893554688,
997
+ "logps/rejected": -1685.641357421875,
998
+ "loss": 0.1675,
999
+ "rewards/accuracies": 0.824999988079071,
1000
+ "rewards/chosen": -0.05924375727772713,
1001
+ "rewards/margins": 0.32539594173431396,
1002
+ "rewards/rejected": -0.3846396803855896,
1003
+ "step": 700
1004
+ },
1005
+ {
1006
+ "epoch": 0.38,
1007
+ "learning_rate": 3.908941182252785e-06,
1008
+ "logits/chosen": -1.20944344997406,
1009
+ "logits/rejected": 0.5900853276252747,
1010
+ "logps/chosen": -768.8225708007812,
1011
+ "logps/rejected": -1766.4296875,
1012
+ "loss": 0.1247,
1013
+ "rewards/accuracies": 0.800000011920929,
1014
+ "rewards/chosen": -0.0808175802230835,
1015
+ "rewards/margins": 0.33207273483276367,
1016
+ "rewards/rejected": -0.4128902852535248,
1017
+ "step": 710
1018
+ },
1019
+ {
1020
+ "epoch": 0.38,
1021
+ "learning_rate": 3.8702409525646535e-06,
1022
+ "logits/chosen": -1.496953010559082,
1023
+ "logits/rejected": -0.36526980996131897,
1024
+ "logps/chosen": -677.8695068359375,
1025
+ "logps/rejected": -1590.8409423828125,
1026
+ "loss": 0.1838,
1027
+ "rewards/accuracies": 0.8500000238418579,
1028
+ "rewards/chosen": -0.02380838617682457,
1029
+ "rewards/margins": 0.26771286129951477,
1030
+ "rewards/rejected": -0.29152125120162964,
1031
+ "step": 720
1032
+ },
1033
+ {
1034
+ "epoch": 0.39,
1035
+ "learning_rate": 3.8310655474507495e-06,
1036
+ "logits/chosen": -1.2399176359176636,
1037
+ "logits/rejected": -0.07380921393632889,
1038
+ "logps/chosen": -684.2032470703125,
1039
+ "logps/rejected": -1637.0625,
1040
+ "loss": 0.1776,
1041
+ "rewards/accuracies": 0.8500000238418579,
1042
+ "rewards/chosen": -0.05182964727282524,
1043
+ "rewards/margins": 0.2697630226612091,
1044
+ "rewards/rejected": -0.32159262895584106,
1045
+ "step": 730
1046
+ },
1047
+ {
1048
+ "epoch": 0.39,
1049
+ "learning_rate": 3.7914285522515002e-06,
1050
+ "logits/chosen": -1.3389846086502075,
1051
+ "logits/rejected": -0.04501932114362717,
1052
+ "logps/chosen": -659.2589721679688,
1053
+ "logps/rejected": -1660.969482421875,
1054
+ "loss": 0.1753,
1055
+ "rewards/accuracies": 0.875,
1056
+ "rewards/chosen": -0.06584084033966064,
1057
+ "rewards/margins": 0.3323908746242523,
1058
+ "rewards/rejected": -0.39823174476623535,
1059
+ "step": 740
1060
+ },
1061
+ {
1062
+ "epoch": 0.4,
1063
+ "learning_rate": 3.751343712378639e-06,
1064
+ "logits/chosen": -1.3522380590438843,
1065
+ "logits/rejected": 0.06499899923801422,
1066
+ "logps/chosen": -580.7006225585938,
1067
+ "logps/rejected": -1741.8189697265625,
1068
+ "loss": 0.1372,
1069
+ "rewards/accuracies": 0.8999999761581421,
1070
+ "rewards/chosen": -0.04593845084309578,
1071
+ "rewards/margins": 0.39734122157096863,
1072
+ "rewards/rejected": -0.4432796835899353,
1073
+ "step": 750
1074
+ },
1075
+ {
1076
+ "epoch": 0.41,
1077
+ "learning_rate": 3.710824928548546e-06,
1078
+ "logits/chosen": -1.3140994310379028,
1079
+ "logits/rejected": -0.617447555065155,
1080
+ "logps/chosen": -633.9112548828125,
1081
+ "logps/rejected": -1493.082763671875,
1082
+ "loss": 0.1866,
1083
+ "rewards/accuracies": 0.7250000238418579,
1084
+ "rewards/chosen": -0.09161052852869034,
1085
+ "rewards/margins": 0.26011547446250916,
1086
+ "rewards/rejected": -0.3517259955406189,
1087
+ "step": 760
1088
+ },
1089
+ {
1090
+ "epoch": 0.41,
1091
+ "learning_rate": 3.6698862519617225e-06,
1092
+ "logits/chosen": -1.2595350742340088,
1093
+ "logits/rejected": -0.20695531368255615,
1094
+ "logps/chosen": -593.3395385742188,
1095
+ "logps/rejected": -1490.5479736328125,
1096
+ "loss": 0.1646,
1097
+ "rewards/accuracies": 0.7749999761581421,
1098
+ "rewards/chosen": -0.06257372349500656,
1099
+ "rewards/margins": 0.2620692849159241,
1100
+ "rewards/rejected": -0.3246430456638336,
1101
+ "step": 770
1102
+ },
1103
+ {
1104
+ "epoch": 0.42,
1105
+ "learning_rate": 3.6285418794300793e-06,
1106
+ "logits/chosen": -1.4119771718978882,
1107
+ "logits/rejected": -0.5417270064353943,
1108
+ "logps/chosen": -610.9232788085938,
1109
+ "logps/rejected": -1522.27197265625,
1110
+ "loss": 0.2589,
1111
+ "rewards/accuracies": 0.7749999761581421,
1112
+ "rewards/chosen": -0.08761414885520935,
1113
+ "rewards/margins": 0.2841811180114746,
1114
+ "rewards/rejected": -0.37179526686668396,
1115
+ "step": 780
1116
+ },
1117
+ {
1118
+ "epoch": 0.42,
1119
+ "learning_rate": 3.5868061484537365e-06,
1120
+ "logits/chosen": -1.333315134048462,
1121
+ "logits/rejected": -0.016486238688230515,
1122
+ "logps/chosen": -770.37744140625,
1123
+ "logps/rejected": -1697.8402099609375,
1124
+ "loss": 0.236,
1125
+ "rewards/accuracies": 0.875,
1126
+ "rewards/chosen": -0.11275408416986465,
1127
+ "rewards/margins": 0.25441592931747437,
1128
+ "rewards/rejected": -0.3671700060367584,
1129
+ "step": 790
1130
+ },
1131
+ {
1132
+ "epoch": 0.43,
1133
+ "learning_rate": 3.5446935322490285e-06,
1134
+ "logits/chosen": -1.4592931270599365,
1135
+ "logits/rejected": -0.8714092373847961,
1136
+ "logps/chosen": -645.5399169921875,
1137
+ "logps/rejected": -1329.577392578125,
1138
+ "loss": 0.2498,
1139
+ "rewards/accuracies": 0.625,
1140
+ "rewards/chosen": -0.08806028217077255,
1141
+ "rewards/margins": 0.12212536484003067,
1142
+ "rewards/rejected": -0.21018561720848083,
1143
+ "step": 800
1144
+ },
1145
+ {
1146
+ "epoch": 0.43,
1147
+ "learning_rate": 3.502218634729447e-06,
1148
+ "logits/chosen": -1.4128929376602173,
1149
+ "logits/rejected": -0.7669699788093567,
1150
+ "logps/chosen": -649.4080810546875,
1151
+ "logps/rejected": -1620.560302734375,
1152
+ "loss": 0.1874,
1153
+ "rewards/accuracies": 0.824999988079071,
1154
+ "rewards/chosen": -0.11496981233358383,
1155
+ "rewards/margins": 0.2872191071510315,
1156
+ "rewards/rejected": -0.40218886733055115,
1157
+ "step": 810
1158
+ },
1159
+ {
1160
+ "epoch": 0.44,
1161
+ "learning_rate": 3.459396185441265e-06,
1162
+ "logits/chosen": -1.3515876531600952,
1163
+ "logits/rejected": -0.6728461980819702,
1164
+ "logps/chosen": -762.8712158203125,
1165
+ "logps/rejected": -1834.1533203125,
1166
+ "loss": 0.2323,
1167
+ "rewards/accuracies": 0.824999988079071,
1168
+ "rewards/chosen": -0.14702202379703522,
1169
+ "rewards/margins": 0.354480117559433,
1170
+ "rewards/rejected": -0.5015021562576294,
1171
+ "step": 820
1172
+ },
1173
+ {
1174
+ "epoch": 0.44,
1175
+ "learning_rate": 3.4162410344555834e-06,
1176
+ "logits/chosen": -1.4009459018707275,
1177
+ "logits/rejected": -0.9621168375015259,
1178
+ "logps/chosen": -657.5804443359375,
1179
+ "logps/rejected": -1689.560302734375,
1180
+ "loss": 0.1928,
1181
+ "rewards/accuracies": 0.7749999761581421,
1182
+ "rewards/chosen": -0.056845568120479584,
1183
+ "rewards/margins": 0.3001772463321686,
1184
+ "rewards/rejected": -0.35702282190322876,
1185
+ "step": 830
1186
+ },
1187
+ {
1188
+ "epoch": 0.45,
1189
+ "learning_rate": 3.3727681472185937e-06,
1190
+ "logits/chosen": -1.2048084735870361,
1191
+ "logits/rejected": -0.543747067451477,
1192
+ "logps/chosen": -639.1177978515625,
1193
+ "logps/rejected": -1738.9322509765625,
1194
+ "loss": 0.1711,
1195
+ "rewards/accuracies": 0.824999988079071,
1196
+ "rewards/chosen": -0.02559160254895687,
1197
+ "rewards/margins": 0.27681928873062134,
1198
+ "rewards/rejected": -0.30241090059280396,
1199
+ "step": 840
1200
+ },
1201
+ {
1202
+ "epoch": 0.45,
1203
+ "learning_rate": 3.3289925993618217e-06,
1204
+ "logits/chosen": -1.1008461713790894,
1205
+ "logits/rejected": -0.47523483633995056,
1206
+ "logps/chosen": -638.2371215820312,
1207
+ "logps/rejected": -1568.861328125,
1208
+ "loss": 0.1779,
1209
+ "rewards/accuracies": 0.75,
1210
+ "rewards/chosen": -0.07939572632312775,
1211
+ "rewards/margins": 0.26904281973838806,
1212
+ "rewards/rejected": -0.3484385311603546,
1213
+ "step": 850
1214
+ },
1215
+ {
1216
+ "epoch": 0.46,
1217
+ "learning_rate": 3.2849295714741643e-06,
1218
+ "logits/chosen": -1.2790766954421997,
1219
+ "logits/rejected": -0.019085492938756943,
1220
+ "logps/chosen": -567.5126342773438,
1221
+ "logps/rejected": -1681.409912109375,
1222
+ "loss": 0.176,
1223
+ "rewards/accuracies": 0.7749999761581421,
1224
+ "rewards/chosen": -0.059426773339509964,
1225
+ "rewards/margins": 0.36287856101989746,
1226
+ "rewards/rejected": -0.4223053455352783,
1227
+ "step": 860
1228
+ },
1229
+ {
1230
+ "epoch": 0.46,
1231
+ "learning_rate": 3.2405943438375287e-06,
1232
+ "logits/chosen": -1.3036001920700073,
1233
+ "logits/rejected": -0.434285968542099,
1234
+ "logps/chosen": -641.4450073242188,
1235
+ "logps/rejected": -1748.319091796875,
1236
+ "loss": 0.124,
1237
+ "rewards/accuracies": 0.824999988079071,
1238
+ "rewards/chosen": -0.03710533678531647,
1239
+ "rewards/margins": 0.3532947897911072,
1240
+ "rewards/rejected": -0.39040011167526245,
1241
+ "step": 870
1242
+ },
1243
+ {
1244
+ "epoch": 0.47,
1245
+ "learning_rate": 3.1960022911279036e-06,
1246
+ "logits/chosen": -1.302273154258728,
1247
+ "logits/rejected": -0.5197567939758301,
1248
+ "logps/chosen": -562.3380737304688,
1249
+ "logps/rejected": -1495.635498046875,
1250
+ "loss": 0.1667,
1251
+ "rewards/accuracies": 0.8500000238418579,
1252
+ "rewards/chosen": -0.03037446178495884,
1253
+ "rewards/margins": 0.2802720069885254,
1254
+ "rewards/rejected": -0.3106464743614197,
1255
+ "step": 880
1256
+ },
1257
+ {
1258
+ "epoch": 0.47,
1259
+ "learning_rate": 3.1511688770836844e-06,
1260
+ "logits/chosen": -1.4215189218521118,
1261
+ "logits/rejected": -0.5714216828346252,
1262
+ "logps/chosen": -665.7767333984375,
1263
+ "logps/rejected": -1658.6048583984375,
1264
+ "loss": 0.1316,
1265
+ "rewards/accuracies": 0.800000011920929,
1266
+ "rewards/chosen": -0.05627693980932236,
1267
+ "rewards/margins": 0.32271087169647217,
1268
+ "rewards/rejected": -0.3789878189563751,
1269
+ "step": 890
1270
+ },
1271
+ {
1272
+ "epoch": 0.48,
1273
+ "learning_rate": 3.1061096491431307e-06,
1274
+ "logits/chosen": -1.3552201986312866,
1275
+ "logits/rejected": -0.9116779565811157,
1276
+ "logps/chosen": -690.0846557617188,
1277
+ "logps/rejected": -1613.9110107421875,
1278
+ "loss": 0.1754,
1279
+ "rewards/accuracies": 0.800000011920929,
1280
+ "rewards/chosen": -0.04472965747117996,
1281
+ "rewards/margins": 0.2647218108177185,
1282
+ "rewards/rejected": -0.3094514310359955,
1283
+ "step": 900
1284
+ },
1285
+ {
1286
+ "epoch": 0.49,
1287
+ "learning_rate": 3.0608402330527796e-06,
1288
+ "logits/chosen": -1.3723796606063843,
1289
+ "logits/rejected": -0.6000717282295227,
1290
+ "logps/chosen": -692.8043212890625,
1291
+ "logps/rejected": -1629.4808349609375,
1292
+ "loss": 0.2028,
1293
+ "rewards/accuracies": 0.7250000238418579,
1294
+ "rewards/chosen": -0.13592025637626648,
1295
+ "rewards/margins": 0.29076087474823,
1296
+ "rewards/rejected": -0.42668113112449646,
1297
+ "step": 910
1298
+ },
1299
+ {
1300
+ "epoch": 0.49,
1301
+ "learning_rate": 3.0153763274487176e-06,
1302
+ "logits/chosen": -1.3126561641693115,
1303
+ "logits/rejected": -0.3268551230430603,
1304
+ "logps/chosen": -676.1267700195312,
1305
+ "logps/rejected": -1620.5989990234375,
1306
+ "loss": 0.1766,
1307
+ "rewards/accuracies": 0.7749999761581421,
1308
+ "rewards/chosen": -0.09026195108890533,
1309
+ "rewards/margins": 0.3179401159286499,
1310
+ "rewards/rejected": -0.40820208191871643,
1311
+ "step": 920
1312
+ },
1313
+ {
1314
+ "epoch": 0.5,
1315
+ "learning_rate": 2.9697336984125683e-06,
1316
+ "logits/chosen": -1.306850552558899,
1317
+ "logits/rejected": -0.35740819573402405,
1318
+ "logps/chosen": -662.0093994140625,
1319
+ "logps/rejected": -1545.4691162109375,
1320
+ "loss": 0.1835,
1321
+ "rewards/accuracies": 0.8500000238418579,
1322
+ "rewards/chosen": -0.05962755158543587,
1323
+ "rewards/margins": 0.2719072997570038,
1324
+ "rewards/rejected": -0.33153486251831055,
1325
+ "step": 930
1326
+ },
1327
+ {
1328
+ "epoch": 0.5,
1329
+ "learning_rate": 2.923928174004094e-06,
1330
+ "logits/chosen": -1.454861044883728,
1331
+ "logits/rejected": -0.696769654750824,
1332
+ "logps/chosen": -648.5043334960938,
1333
+ "logps/rejected": -1763.5390625,
1334
+ "loss": 0.1541,
1335
+ "rewards/accuracies": 0.8999999761581421,
1336
+ "rewards/chosen": -0.021979842334985733,
1337
+ "rewards/margins": 0.3547331690788269,
1338
+ "rewards/rejected": -0.3767130672931671,
1339
+ "step": 940
1340
+ },
1341
+ {
1342
+ "epoch": 0.51,
1343
+ "learning_rate": 2.8779756387723036e-06,
1344
+ "logits/chosen": -1.4002293348312378,
1345
+ "logits/rejected": -0.8346809148788452,
1346
+ "logps/chosen": -564.9570922851562,
1347
+ "logps/rejected": -1489.9595947265625,
1348
+ "loss": 0.2194,
1349
+ "rewards/accuracies": 0.7749999761581421,
1350
+ "rewards/chosen": -0.02880738116800785,
1351
+ "rewards/margins": 0.2632240653038025,
1352
+ "rewards/rejected": -0.292031466960907,
1353
+ "step": 950
1354
+ },
1355
+ {
1356
+ "epoch": 0.51,
1357
+ "learning_rate": 2.831892028246968e-06,
1358
+ "logits/chosen": -1.215348482131958,
1359
+ "logits/rejected": -0.574180006980896,
1360
+ "logps/chosen": -637.7142944335938,
1361
+ "logps/rejected": -1575.6964111328125,
1362
+ "loss": 0.1602,
1363
+ "rewards/accuracies": 0.800000011920929,
1364
+ "rewards/chosen": -0.034471701830625534,
1365
+ "rewards/margins": 0.2788967490196228,
1366
+ "rewards/rejected": -0.31336846947669983,
1367
+ "step": 960
1368
+ },
1369
+ {
1370
+ "epoch": 0.52,
1371
+ "learning_rate": 2.7856933234124617e-06,
1372
+ "logits/chosen": -1.4842756986618042,
1373
+ "logits/rejected": -0.5266194343566895,
1374
+ "logps/chosen": -667.3348388671875,
1375
+ "logps/rejected": -1674.873779296875,
1376
+ "loss": 0.1523,
1377
+ "rewards/accuracies": 0.824999988079071,
1378
+ "rewards/chosen": -0.04780828207731247,
1379
+ "rewards/margins": 0.32556313276290894,
1380
+ "rewards/rejected": -0.3733713626861572,
1381
+ "step": 970
1382
+ },
1383
+ {
1384
+ "epoch": 0.52,
1385
+ "learning_rate": 2.7393955451658387e-06,
1386
+ "logits/chosen": -0.9660801887512207,
1387
+ "logits/rejected": -0.038875680416822433,
1388
+ "logps/chosen": -644.0047607421875,
1389
+ "logps/rejected": -1645.885498046875,
1390
+ "loss": 0.1848,
1391
+ "rewards/accuracies": 0.8500000238418579,
1392
+ "rewards/chosen": -0.07426423579454422,
1393
+ "rewards/margins": 0.31580421328544617,
1394
+ "rewards/rejected": -0.3900684714317322,
1395
+ "step": 980
1396
+ },
1397
+ {
1398
+ "epoch": 0.53,
1399
+ "learning_rate": 2.6930147487610667e-06,
1400
+ "logits/chosen": -1.2562506198883057,
1401
+ "logits/rejected": -0.5583636164665222,
1402
+ "logps/chosen": -687.5775146484375,
1403
+ "logps/rejected": -1682.3447265625,
1404
+ "loss": 0.2442,
1405
+ "rewards/accuracies": 0.800000011920929,
1406
+ "rewards/chosen": -0.10189177095890045,
1407
+ "rewards/margins": 0.29264310002326965,
1408
+ "rewards/rejected": -0.3945348560810089,
1409
+ "step": 990
1410
+ },
1411
+ {
1412
+ "epoch": 0.53,
1413
+ "learning_rate": 2.6465670182413487e-06,
1414
+ "logits/chosen": -1.0773265361785889,
1415
+ "logits/rejected": -0.38139039278030396,
1416
+ "logps/chosen": -669.3234252929688,
1417
+ "logps/rejected": -1580.733642578125,
1418
+ "loss": 0.1672,
1419
+ "rewards/accuracies": 0.824999988079071,
1420
+ "rewards/chosen": -0.07045495510101318,
1421
+ "rewards/margins": 0.24894073605537415,
1422
+ "rewards/rejected": -0.31939566135406494,
1423
+ "step": 1000
1424
+ },
1425
+ {
1426
+ "epoch": 0.54,
1427
+ "learning_rate": 2.6000684608614594e-06,
1428
+ "logits/chosen": -1.3887989521026611,
1429
+ "logits/rejected": -0.6506294012069702,
1430
+ "logps/chosen": -653.4199829101562,
1431
+ "logps/rejected": -1556.509521484375,
1432
+ "loss": 0.1917,
1433
+ "rewards/accuracies": 0.75,
1434
+ "rewards/chosen": -0.016438838094472885,
1435
+ "rewards/margins": 0.27959805727005005,
1436
+ "rewards/rejected": -0.29603689908981323,
1437
+ "step": 1010
1438
+ },
1439
+ {
1440
+ "epoch": 0.54,
1441
+ "learning_rate": 2.5535352015020338e-06,
1442
+ "logits/chosen": -1.2708961963653564,
1443
+ "logits/rejected": -0.4829919934272766,
1444
+ "logps/chosen": -708.9461669921875,
1445
+ "logps/rejected": -1660.2349853515625,
1446
+ "loss": 0.1956,
1447
+ "rewards/accuracies": 0.8500000238418579,
1448
+ "rewards/chosen": -0.02399434708058834,
1449
+ "rewards/margins": 0.2889823317527771,
1450
+ "rewards/rejected": -0.3129766881465912,
1451
+ "step": 1020
1452
+ },
1453
+ {
1454
+ "epoch": 0.55,
1455
+ "learning_rate": 2.506983377077741e-06,
1456
+ "logits/chosen": -1.2446025609970093,
1457
+ "logits/rejected": -0.7467063665390015,
1458
+ "logps/chosen": -516.4700927734375,
1459
+ "logps/rejected": -1463.983642578125,
1460
+ "loss": 0.1882,
1461
+ "rewards/accuracies": 0.699999988079071,
1462
+ "rewards/chosen": -0.015911739319562912,
1463
+ "rewards/margins": 0.294134259223938,
1464
+ "rewards/rejected": -0.3100460171699524,
1465
+ "step": 1030
1466
+ },
1467
+ {
1468
+ "epoch": 0.55,
1469
+ "learning_rate": 2.460429130941289e-06,
1470
+ "logits/chosen": -1.4046454429626465,
1471
+ "logits/rejected": -0.6258004903793335,
1472
+ "logps/chosen": -548.9246826171875,
1473
+ "logps/rejected": -1328.09716796875,
1474
+ "loss": 0.1558,
1475
+ "rewards/accuracies": 0.75,
1476
+ "rewards/chosen": -0.007907375693321228,
1477
+ "rewards/margins": 0.23490925133228302,
1478
+ "rewards/rejected": -0.24281664192676544,
1479
+ "step": 1040
1480
+ },
1481
+ {
1482
+ "epoch": 0.56,
1483
+ "learning_rate": 2.413888607285192e-06,
1484
+ "logits/chosen": -1.351510763168335,
1485
+ "logits/rejected": -0.3396194279193878,
1486
+ "logps/chosen": -716.7518310546875,
1487
+ "logps/rejected": -1452.3531494140625,
1488
+ "loss": 0.1761,
1489
+ "rewards/accuracies": 0.7250000238418579,
1490
+ "rewards/chosen": -0.030713701620697975,
1491
+ "rewards/margins": 0.24616649746894836,
1492
+ "rewards/rejected": -0.2768802046775818,
1493
+ "step": 1050
1494
+ },
1495
+ {
1496
+ "epoch": 0.57,
1497
+ "learning_rate": 2.367377945543249e-06,
1498
+ "logits/chosen": -1.2682628631591797,
1499
+ "logits/rejected": -0.39214056730270386,
1500
+ "logps/chosen": -745.8023681640625,
1501
+ "logps/rejected": -1684.0126953125,
1502
+ "loss": 0.1596,
1503
+ "rewards/accuracies": 0.875,
1504
+ "rewards/chosen": -0.05257143825292587,
1505
+ "rewards/margins": 0.3037104606628418,
1506
+ "rewards/rejected": -0.35628193616867065,
1507
+ "step": 1060
1508
+ },
1509
+ {
1510
+ "epoch": 0.57,
1511
+ "learning_rate": 2.320913274793676e-06,
1512
+ "logits/chosen": -1.3436534404754639,
1513
+ "logits/rejected": -0.21577882766723633,
1514
+ "logps/chosen": -625.4033813476562,
1515
+ "logps/rejected": -1678.371826171875,
1516
+ "loss": 0.178,
1517
+ "rewards/accuracies": 0.824999988079071,
1518
+ "rewards/chosen": -0.026138927787542343,
1519
+ "rewards/margins": 0.3423928916454315,
1520
+ "rewards/rejected": -0.36853182315826416,
1521
+ "step": 1070
1522
+ },
1523
+ {
1524
+ "epoch": 0.58,
1525
+ "learning_rate": 2.27451070816582e-06,
1526
+ "logits/chosen": -1.5014312267303467,
1527
+ "logits/rejected": -0.43270865082740784,
1528
+ "logps/chosen": -686.2655029296875,
1529
+ "logps/rejected": -1728.8896484375,
1530
+ "loss": 0.183,
1531
+ "rewards/accuracies": 0.949999988079071,
1532
+ "rewards/chosen": -0.019897693768143654,
1533
+ "rewards/margins": 0.35252150893211365,
1534
+ "rewards/rejected": -0.37241920828819275,
1535
+ "step": 1080
1536
+ },
1537
+ {
1538
+ "epoch": 0.58,
1539
+ "learning_rate": 2.228186337252414e-06,
1540
+ "logits/chosen": -1.3439433574676514,
1541
+ "logits/rejected": -0.20307889580726624,
1542
+ "logps/chosen": -795.5742797851562,
1543
+ "logps/rejected": -1796.8831787109375,
1544
+ "loss": 0.1734,
1545
+ "rewards/accuracies": 0.824999988079071,
1546
+ "rewards/chosen": -0.08263162523508072,
1547
+ "rewards/margins": 0.32303547859191895,
1548
+ "rewards/rejected": -0.40566712617874146,
1549
+ "step": 1090
1550
+ },
1551
+ {
1552
+ "epoch": 0.59,
1553
+ "learning_rate": 2.1819562265292946e-06,
1554
+ "logits/chosen": -1.2423124313354492,
1555
+ "logits/rejected": -0.54206782579422,
1556
+ "logps/chosen": -570.5196533203125,
1557
+ "logps/rejected": -1636.732421875,
1558
+ "loss": 0.1307,
1559
+ "rewards/accuracies": 0.925000011920929,
1560
+ "rewards/chosen": -0.07189210504293442,
1561
+ "rewards/margins": 0.3438221514225006,
1562
+ "rewards/rejected": -0.41571420431137085,
1563
+ "step": 1100
1564
+ },
1565
+ {
1566
+ "epoch": 0.59,
1567
+ "learning_rate": 2.1358364077845236e-06,
1568
+ "logits/chosen": -1.2860864400863647,
1569
+ "logits/rejected": -0.5983974933624268,
1570
+ "logps/chosen": -659.2459716796875,
1571
+ "logps/rejected": -1549.3529052734375,
1572
+ "loss": 0.1609,
1573
+ "rewards/accuracies": 0.800000011920929,
1574
+ "rewards/chosen": -0.08437564969062805,
1575
+ "rewards/margins": 0.2931548058986664,
1576
+ "rewards/rejected": -0.3775304853916168,
1577
+ "step": 1110
1578
+ },
1579
+ {
1580
+ "epoch": 0.6,
1581
+ "learning_rate": 2.089842874558849e-06,
1582
+ "logits/chosen": -1.3163573741912842,
1583
+ "logits/rejected": -0.6514892578125,
1584
+ "logps/chosen": -619.4735107421875,
1585
+ "logps/rejected": -1633.3419189453125,
1586
+ "loss": 0.1827,
1587
+ "rewards/accuracies": 0.824999988079071,
1588
+ "rewards/chosen": -0.05838247388601303,
1589
+ "rewards/margins": 0.3051317036151886,
1590
+ "rewards/rejected": -0.3635141849517822,
1591
+ "step": 1120
1592
+ },
1593
+ {
1594
+ "epoch": 0.6,
1595
+ "learning_rate": 2.0439915765994242e-06,
1596
+ "logits/chosen": -1.4336564540863037,
1597
+ "logits/rejected": -0.5167166590690613,
1598
+ "logps/chosen": -519.189453125,
1599
+ "logps/rejected": -1454.1304931640625,
1600
+ "loss": 0.154,
1601
+ "rewards/accuracies": 0.75,
1602
+ "rewards/chosen": -0.05485490709543228,
1603
+ "rewards/margins": 0.2894749343395233,
1604
+ "rewards/rejected": -0.3443298935890198,
1605
+ "step": 1130
1606
+ },
1607
+ {
1608
+ "epoch": 0.61,
1609
+ "learning_rate": 1.9982984143287186e-06,
1610
+ "logits/chosen": -0.9816936254501343,
1611
+ "logits/rejected": -0.26896458864212036,
1612
+ "logps/chosen": -681.84765625,
1613
+ "logps/rejected": -1698.229736328125,
1614
+ "loss": 0.1722,
1615
+ "rewards/accuracies": 0.824999988079071,
1616
+ "rewards/chosen": -0.09818267822265625,
1617
+ "rewards/margins": 0.25163981318473816,
1618
+ "rewards/rejected": -0.3498225212097168,
1619
+ "step": 1140
1620
+ },
1621
+ {
1622
+ "epoch": 0.61,
1623
+ "learning_rate": 1.95277923333053e-06,
1624
+ "logits/chosen": -1.5167888402938843,
1625
+ "logits/rejected": -0.49634304642677307,
1626
+ "logps/chosen": -805.359130859375,
1627
+ "logps/rejected": -1700.7835693359375,
1628
+ "loss": 0.1245,
1629
+ "rewards/accuracies": 0.8500000238418579,
1630
+ "rewards/chosen": -0.09041927754878998,
1631
+ "rewards/margins": 0.3279864192008972,
1632
+ "rewards/rejected": -0.418405681848526,
1633
+ "step": 1150
1634
+ },
1635
+ {
1636
+ "epoch": 0.62,
1637
+ "learning_rate": 1.9074498188550156e-06,
1638
+ "logits/chosen": -1.519516110420227,
1639
+ "logits/rejected": -0.12074364721775055,
1640
+ "logps/chosen": -685.3983154296875,
1641
+ "logps/rejected": -1531.295166015625,
1642
+ "loss": 0.1808,
1643
+ "rewards/accuracies": 0.824999988079071,
1644
+ "rewards/chosen": -0.056040991097688675,
1645
+ "rewards/margins": 0.315580278635025,
1646
+ "rewards/rejected": -0.3716212511062622,
1647
+ "step": 1160
1648
+ },
1649
+ {
1650
+ "epoch": 0.62,
1651
+ "learning_rate": 1.862325890344643e-06,
1652
+ "logits/chosen": -1.232160210609436,
1653
+ "logits/rejected": -0.23375025391578674,
1654
+ "logps/chosen": -698.9664916992188,
1655
+ "logps/rejected": -1915.4918212890625,
1656
+ "loss": 0.1419,
1657
+ "rewards/accuracies": 0.8999999761581421,
1658
+ "rewards/chosen": -0.09169380366802216,
1659
+ "rewards/margins": 0.40644773840904236,
1660
+ "rewards/rejected": -0.49814146757125854,
1661
+ "step": 1170
1662
+ },
1663
+ {
1664
+ "epoch": 0.63,
1665
+ "learning_rate": 1.817423095982972e-06,
1666
+ "logits/chosen": -1.3982770442962646,
1667
+ "logits/rejected": -0.8638380765914917,
1668
+ "logps/chosen": -595.23291015625,
1669
+ "logps/rejected": -1435.40966796875,
1670
+ "loss": 0.1757,
1671
+ "rewards/accuracies": 0.699999988079071,
1672
+ "rewards/chosen": -0.07611984014511108,
1673
+ "rewards/margins": 0.268936425447464,
1674
+ "rewards/rejected": -0.34505629539489746,
1675
+ "step": 1180
1676
+ },
1677
+ {
1678
+ "epoch": 0.63,
1679
+ "learning_rate": 1.7727570072681293e-06,
1680
+ "logits/chosen": -1.4742844104766846,
1681
+ "logits/rejected": -0.3091976046562195,
1682
+ "logps/chosen": -600.8450317382812,
1683
+ "logps/rejected": -1698.4534912109375,
1684
+ "loss": 0.109,
1685
+ "rewards/accuracies": 0.875,
1686
+ "rewards/chosen": -0.05584261566400528,
1687
+ "rewards/margins": 0.35010969638824463,
1688
+ "rewards/rejected": -0.4059523046016693,
1689
+ "step": 1190
1690
+ },
1691
+ {
1692
+ "epoch": 0.64,
1693
+ "learning_rate": 1.7283431136128961e-06,
1694
+ "logits/chosen": -1.4247504472732544,
1695
+ "logits/rejected": -0.4096221327781677,
1696
+ "logps/chosen": -717.885498046875,
1697
+ "logps/rejected": -1856.927734375,
1698
+ "loss": 0.1427,
1699
+ "rewards/accuracies": 0.949999988079071,
1700
+ "rewards/chosen": -0.0902978777885437,
1701
+ "rewards/margins": 0.3808268904685974,
1702
+ "rewards/rejected": -0.4711247384548187,
1703
+ "step": 1200
1704
+ },
1705
+ {
1706
+ "epoch": 0.65,
1707
+ "learning_rate": 1.6841968169732478e-06,
1708
+ "logits/chosen": -1.285744071006775,
1709
+ "logits/rejected": -0.08921636641025543,
1710
+ "logps/chosen": -761.4056396484375,
1711
+ "logps/rejected": -1614.3876953125,
1712
+ "loss": 0.1334,
1713
+ "rewards/accuracies": 0.8500000238418579,
1714
+ "rewards/chosen": -0.10277809947729111,
1715
+ "rewards/margins": 0.29823797941207886,
1716
+ "rewards/rejected": -0.4010160565376282,
1717
+ "step": 1210
1718
+ },
1719
+ {
1720
+ "epoch": 0.65,
1721
+ "learning_rate": 1.6403334265072284e-06,
1722
+ "logits/chosen": -1.2995268106460571,
1723
+ "logits/rejected": -0.6406716704368591,
1724
+ "logps/chosen": -857.0849609375,
1725
+ "logps/rejected": -1790.50390625,
1726
+ "loss": 0.1581,
1727
+ "rewards/accuracies": 0.824999988079071,
1728
+ "rewards/chosen": -0.14391708374023438,
1729
+ "rewards/margins": 0.3202173411846161,
1730
+ "rewards/rejected": -0.4641345143318176,
1731
+ "step": 1220
1732
+ },
1733
+ {
1734
+ "epoch": 0.66,
1735
+ "learning_rate": 1.5967681532660066e-06,
1736
+ "logits/chosen": -1.2067750692367554,
1737
+ "logits/rejected": -0.40620866417884827,
1738
+ "logps/chosen": -689.6841430664062,
1739
+ "logps/rejected": -1506.9180908203125,
1740
+ "loss": 0.1781,
1741
+ "rewards/accuracies": 0.824999988079071,
1742
+ "rewards/chosen": -0.08755698055028915,
1743
+ "rewards/margins": 0.24655339121818542,
1744
+ "rewards/rejected": -0.33411040902137756,
1745
+ "step": 1230
1746
+ },
1747
+ {
1748
+ "epoch": 0.66,
1749
+ "learning_rate": 1.5535161049189463e-06,
1750
+ "logits/chosen": -1.388139009475708,
1751
+ "logits/rejected": -0.28742751479148865,
1752
+ "logps/chosen": -702.9202270507812,
1753
+ "logps/rejected": -1619.192138671875,
1754
+ "loss": 0.1155,
1755
+ "rewards/accuracies": 0.8500000238418579,
1756
+ "rewards/chosen": -0.11033840477466583,
1757
+ "rewards/margins": 0.27864348888397217,
1758
+ "rewards/rejected": -0.3889819085597992,
1759
+ "step": 1240
1760
+ },
1761
+ {
1762
+ "epoch": 0.67,
1763
+ "learning_rate": 1.5105922805145356e-06,
1764
+ "logits/chosen": -1.4737704992294312,
1765
+ "logits/rejected": -0.7333568334579468,
1766
+ "logps/chosen": -611.47314453125,
1767
+ "logps/rejected": -1699.2509765625,
1768
+ "loss": 0.1432,
1769
+ "rewards/accuracies": 0.8999999761581421,
1770
+ "rewards/chosen": -0.06176890805363655,
1771
+ "rewards/margins": 0.35512489080429077,
1772
+ "rewards/rejected": -0.4168938100337982,
1773
+ "step": 1250
1774
+ },
1775
+ {
1776
+ "epoch": 0.67,
1777
+ "learning_rate": 1.4680115652789823e-06,
1778
+ "logits/chosen": -1.1875636577606201,
1779
+ "logits/rejected": -0.24240991473197937,
1780
+ "logps/chosen": -745.5838012695312,
1781
+ "logps/rejected": -1765.468505859375,
1782
+ "loss": 0.1855,
1783
+ "rewards/accuracies": 0.8500000238418579,
1784
+ "rewards/chosen": -0.10248545557260513,
1785
+ "rewards/margins": 0.34968844056129456,
1786
+ "rewards/rejected": -0.4521738588809967,
1787
+ "step": 1260
1788
+ },
1789
+ {
1790
+ "epoch": 0.68,
1791
+ "learning_rate": 1.4257887254542767e-06,
1792
+ "logits/chosen": -1.2042195796966553,
1793
+ "logits/rejected": -0.4282516837120056,
1794
+ "logps/chosen": -745.48291015625,
1795
+ "logps/rejected": -1836.1507568359375,
1796
+ "loss": 0.2098,
1797
+ "rewards/accuracies": 0.875,
1798
+ "rewards/chosen": -0.08774339407682419,
1799
+ "rewards/margins": 0.3781694173812866,
1800
+ "rewards/rejected": -0.4659128785133362,
1801
+ "step": 1270
1802
+ },
1803
+ {
1804
+ "epoch": 0.68,
1805
+ "learning_rate": 1.3839384031775227e-06,
1806
+ "logits/chosen": -1.3200931549072266,
1807
+ "logits/rejected": -0.23144832253456116,
1808
+ "logps/chosen": -582.4103393554688,
1809
+ "logps/rejected": -1639.1998291015625,
1810
+ "loss": 0.1594,
1811
+ "rewards/accuracies": 0.875,
1812
+ "rewards/chosen": -0.03387508541345596,
1813
+ "rewards/margins": 0.3454126715660095,
1814
+ "rewards/rejected": -0.3792877793312073,
1815
+ "step": 1280
1816
+ },
1817
+ {
1818
+ "epoch": 0.69,
1819
+ "learning_rate": 1.342475111403298e-06,
1820
+ "logits/chosen": -1.4199360609054565,
1821
+ "logits/rejected": -0.5141012072563171,
1822
+ "logps/chosen": -664.3624877929688,
1823
+ "logps/rejected": -1757.888427734375,
1824
+ "loss": 0.1348,
1825
+ "rewards/accuracies": 0.925000011920929,
1826
+ "rewards/chosen": -0.03387322276830673,
1827
+ "rewards/margins": 0.35872942209243774,
1828
+ "rewards/rejected": -0.3926026523113251,
1829
+ "step": 1290
1830
+ },
1831
+ {
1832
+ "epoch": 0.69,
1833
+ "learning_rate": 1.3014132288708209e-06,
1834
+ "logits/chosen": -1.2313156127929688,
1835
+ "logits/rejected": -0.6418328881263733,
1836
+ "logps/chosen": -680.6788330078125,
1837
+ "logps/rejected": -1728.441162109375,
1838
+ "loss": 0.1366,
1839
+ "rewards/accuracies": 0.8500000238418579,
1840
+ "rewards/chosen": -0.06790607422590256,
1841
+ "rewards/margins": 0.3306724429130554,
1842
+ "rewards/rejected": -0.39857858419418335,
1843
+ "step": 1300
1844
+ },
1845
+ {
1846
+ "epoch": 0.7,
1847
+ "learning_rate": 1.2607669951176549e-06,
1848
+ "logits/chosen": -1.322935700416565,
1849
+ "logits/rejected": -0.7819048166275024,
1850
+ "logps/chosen": -642.0144653320312,
1851
+ "logps/rejected": -1622.2706298828125,
1852
+ "loss": 0.1403,
1853
+ "rewards/accuracies": 0.8999999761581421,
1854
+ "rewards/chosen": -0.03408288210630417,
1855
+ "rewards/margins": 0.3696083426475525,
1856
+ "rewards/rejected": -0.40369120240211487,
1857
+ "step": 1310
1858
+ },
1859
+ {
1860
+ "epoch": 0.7,
1861
+ "learning_rate": 1.2205505055416891e-06,
1862
+ "logits/chosen": -1.2069556713104248,
1863
+ "logits/rejected": -0.599054217338562,
1864
+ "logps/chosen": -657.8556518554688,
1865
+ "logps/rejected": -1849.9466552734375,
1866
+ "loss": 0.1322,
1867
+ "rewards/accuracies": 0.8999999761581421,
1868
+ "rewards/chosen": -0.06479531526565552,
1869
+ "rewards/margins": 0.36816781759262085,
1870
+ "rewards/rejected": -0.4329630732536316,
1871
+ "step": 1320
1872
+ },
1873
+ {
1874
+ "epoch": 0.71,
1875
+ "learning_rate": 1.1807777065131002e-06,
1876
+ "logits/chosen": -1.1746324300765991,
1877
+ "logits/rejected": -0.04525896906852722,
1878
+ "logps/chosen": -619.5230712890625,
1879
+ "logps/rejected": -1747.1986083984375,
1880
+ "loss": 0.1161,
1881
+ "rewards/accuracies": 0.925000011920929,
1882
+ "rewards/chosen": -0.041824158281087875,
1883
+ "rewards/margins": 0.345977246761322,
1884
+ "rewards/rejected": -0.3878014385700226,
1885
+ "step": 1330
1886
+ },
1887
+ {
1888
+ "epoch": 0.71,
1889
+ "learning_rate": 1.1414623905380012e-06,
1890
+ "logits/chosen": -1.1227614879608154,
1891
+ "logits/rejected": 0.07286291569471359,
1892
+ "logps/chosen": -683.210693359375,
1893
+ "logps/rejected": -1760.2294921875,
1894
+ "loss": 0.1307,
1895
+ "rewards/accuracies": 0.925000011920929,
1896
+ "rewards/chosen": -0.06253410875797272,
1897
+ "rewards/margins": 0.33463478088378906,
1898
+ "rewards/rejected": -0.39716893434524536,
1899
+ "step": 1340
1900
+ },
1901
+ {
1902
+ "epoch": 0.72,
1903
+ "learning_rate": 1.1026181914754388e-06,
1904
+ "logits/chosen": -1.2235952615737915,
1905
+ "logits/rejected": -0.23109273612499237,
1906
+ "logps/chosen": -533.2498168945312,
1907
+ "logps/rejected": -1668.765380859375,
1908
+ "loss": 0.1519,
1909
+ "rewards/accuracies": 0.875,
1910
+ "rewards/chosen": -0.01653262972831726,
1911
+ "rewards/margins": 0.36776840686798096,
1912
+ "rewards/rejected": -0.3843010663986206,
1913
+ "step": 1350
1914
+ },
1915
+ {
1916
+ "epoch": 0.73,
1917
+ "learning_rate": 1.0642585798094136e-06,
1918
+ "logits/chosen": -1.1692348718643188,
1919
+ "logits/rejected": -0.5899300575256348,
1920
+ "logps/chosen": -686.9459228515625,
1921
+ "logps/rejected": -1795.8177490234375,
1922
+ "loss": 0.1209,
1923
+ "rewards/accuracies": 0.8999999761581421,
1924
+ "rewards/chosen": -0.07367046922445297,
1925
+ "rewards/margins": 0.3583151698112488,
1926
+ "rewards/rejected": -0.43198567628860474,
1927
+ "step": 1360
1928
+ },
1929
+ {
1930
+ "epoch": 0.73,
1931
+ "learning_rate": 1.0263968579775522e-06,
1932
+ "logits/chosen": -1.2771265506744385,
1933
+ "logits/rejected": -0.38393262028694153,
1934
+ "logps/chosen": -683.27490234375,
1935
+ "logps/rejected": -1695.8990478515625,
1936
+ "loss": 0.1321,
1937
+ "rewards/accuracies": 0.8500000238418579,
1938
+ "rewards/chosen": -0.09097830951213837,
1939
+ "rewards/margins": 0.35342246294021606,
1940
+ "rewards/rejected": -0.4444007873535156,
1941
+ "step": 1370
1942
+ },
1943
+ {
1944
+ "epoch": 0.74,
1945
+ "learning_rate": 9.89046155758058e-07,
1946
+ "logits/chosen": -1.1900537014007568,
1947
+ "logits/rejected": -0.1559818983078003,
1948
+ "logps/chosen": -743.1553955078125,
1949
+ "logps/rejected": -1642.6292724609375,
1950
+ "loss": 0.1577,
1951
+ "rewards/accuracies": 0.824999988079071,
1952
+ "rewards/chosen": -0.05998997762799263,
1953
+ "rewards/margins": 0.35595935583114624,
1954
+ "rewards/rejected": -0.41594934463500977,
1955
+ "step": 1380
1956
+ },
1957
+ {
1958
+ "epoch": 0.74,
1959
+ "learning_rate": 9.52219425716534e-07,
1960
+ "logits/chosen": -1.4094737768173218,
1961
+ "logits/rejected": -0.359419047832489,
1962
+ "logps/chosen": -764.5060424804688,
1963
+ "logps/rejected": -1663.7763671875,
1964
+ "loss": 0.1389,
1965
+ "rewards/accuracies": 0.8500000238418579,
1966
+ "rewards/chosen": -0.08220493793487549,
1967
+ "rewards/margins": 0.2807352542877197,
1968
+ "rewards/rejected": -0.3629401922225952,
1969
+ "step": 1390
1970
+ },
1971
+ {
1972
+ "epoch": 0.75,
1973
+ "learning_rate": 9.15929438714262e-07,
1974
+ "logits/chosen": -1.2259761095046997,
1975
+ "logits/rejected": -0.3568953275680542,
1976
+ "logps/chosen": -684.1424560546875,
1977
+ "logps/rejected": -1751.624267578125,
1978
+ "loss": 0.1703,
1979
+ "rewards/accuracies": 0.824999988079071,
1980
+ "rewards/chosen": -0.07558954507112503,
1981
+ "rewards/margins": 0.3323276937007904,
1982
+ "rewards/rejected": -0.4079172611236572,
1983
+ "step": 1400
1984
+ },
1985
+ {
1986
+ "epoch": 0.75,
1987
+ "learning_rate": 8.801887794794911e-07,
1988
+ "logits/chosen": -1.2949384450912476,
1989
+ "logits/rejected": -0.26310306787490845,
1990
+ "logps/chosen": -527.4078979492188,
1991
+ "logps/rejected": -1489.334716796875,
1992
+ "loss": 0.167,
1993
+ "rewards/accuracies": 0.75,
1994
+ "rewards/chosen": -0.029276112094521523,
1995
+ "rewards/margins": 0.3588224947452545,
1996
+ "rewards/rejected": -0.3880985677242279,
1997
+ "step": 1410
1998
+ },
1999
+ {
2000
+ "epoch": 0.76,
2001
+ "learning_rate": 8.450098422432787e-07,
2002
+ "logits/chosen": -1.084120512008667,
2003
+ "logits/rejected": -0.5709729790687561,
2004
+ "logps/chosen": -519.4564208984375,
2005
+ "logps/rejected": -1378.505615234375,
2006
+ "loss": 0.1475,
2007
+ "rewards/accuracies": 0.699999988079071,
2008
+ "rewards/chosen": -0.043223101645708084,
2009
+ "rewards/margins": 0.2802775204181671,
2010
+ "rewards/rejected": -0.3235006332397461,
2011
+ "step": 1420
2012
+ },
2013
+ {
2014
+ "epoch": 0.76,
2015
+ "learning_rate": 8.104048264413858e-07,
2016
+ "logits/chosen": -1.3592941761016846,
2017
+ "logits/rejected": -0.6775336265563965,
2018
+ "logps/chosen": -672.0179443359375,
2019
+ "logps/rejected": -1647.7578125,
2020
+ "loss": 0.1632,
2021
+ "rewards/accuracies": 0.8500000238418579,
2022
+ "rewards/chosen": -0.042347002774477005,
2023
+ "rewards/margins": 0.309231162071228,
2024
+ "rewards/rejected": -0.3515781760215759,
2025
+ "step": 1430
2026
+ },
2027
+ {
2028
+ "epoch": 0.77,
2029
+ "learning_rate": 7.763857324837321e-07,
2030
+ "logits/chosen": -1.3305068016052246,
2031
+ "logits/rejected": -0.422370046377182,
2032
+ "logps/chosen": -578.7777099609375,
2033
+ "logps/rejected": -1635.9923095703125,
2034
+ "loss": 0.1643,
2035
+ "rewards/accuracies": 0.8999999761581421,
2036
+ "rewards/chosen": -0.013389256782829762,
2037
+ "rewards/margins": 0.3833867907524109,
2038
+ "rewards/rejected": -0.3967760503292084,
2039
+ "step": 1440
2040
+ },
2041
+ {
2042
+ "epoch": 0.77,
2043
+ "learning_rate": 7.429643575928605e-07,
2044
+ "logits/chosen": -1.1784260272979736,
2045
+ "logits/rejected": -0.4220113754272461,
2046
+ "logps/chosen": -715.4688720703125,
2047
+ "logps/rejected": -1668.013916015625,
2048
+ "loss": 0.1387,
2049
+ "rewards/accuracies": 0.875,
2050
+ "rewards/chosen": -0.04753159359097481,
2051
+ "rewards/margins": 0.34793567657470703,
2052
+ "rewards/rejected": -0.3954673409461975,
2053
+ "step": 1450
2054
+ },
2055
+ {
2056
+ "epoch": 0.78,
2057
+ "learning_rate": 7.101522917128709e-07,
2058
+ "logits/chosen": -1.1453496217727661,
2059
+ "logits/rejected": -0.23903509974479675,
2060
+ "logps/chosen": -688.1276245117188,
2061
+ "logps/rejected": -1582.7860107421875,
2062
+ "loss": 0.1475,
2063
+ "rewards/accuracies": 0.8500000238418579,
2064
+ "rewards/chosen": -0.03351983800530434,
2065
+ "rewards/margins": 0.31615251302719116,
2066
+ "rewards/rejected": -0.3496723771095276,
2067
+ "step": 1460
2068
+ },
2069
+ {
2070
+ "epoch": 0.78,
2071
+ "learning_rate": 6.779609134902312e-07,
2072
+ "logits/chosen": -1.5365946292877197,
2073
+ "logits/rejected": -0.83519446849823,
2074
+ "logps/chosen": -617.84912109375,
2075
+ "logps/rejected": -1601.636474609375,
2076
+ "loss": 0.1358,
2077
+ "rewards/accuracies": 0.824999988079071,
2078
+ "rewards/chosen": -0.02726108208298683,
2079
+ "rewards/margins": 0.3341112434864044,
2080
+ "rewards/rejected": -0.36137229204177856,
2081
+ "step": 1470
2082
+ },
2083
+ {
2084
+ "epoch": 0.79,
2085
+ "learning_rate": 6.464013863278629e-07,
2086
+ "logits/chosen": -1.3226994276046753,
2087
+ "logits/rejected": -0.2582075297832489,
2088
+ "logps/chosen": -690.5620727539062,
2089
+ "logps/rejected": -1618.515380859375,
2090
+ "loss": 0.1555,
2091
+ "rewards/accuracies": 0.8500000238418579,
2092
+ "rewards/chosen": -0.04355646297335625,
2093
+ "rewards/margins": 0.32091113924980164,
2094
+ "rewards/rejected": -0.364467591047287,
2095
+ "step": 1480
2096
+ },
2097
+ {
2098
+ "epoch": 0.79,
2099
+ "learning_rate": 6.154846545138696e-07,
2100
+ "logits/chosen": -1.1359361410140991,
2101
+ "logits/rejected": -0.28622904419898987,
2102
+ "logps/chosen": -643.8432006835938,
2103
+ "logps/rejected": -1722.578125,
2104
+ "loss": 0.1235,
2105
+ "rewards/accuracies": 0.8999999761581421,
2106
+ "rewards/chosen": -0.031076664105057716,
2107
+ "rewards/margins": 0.38383063673973083,
2108
+ "rewards/rejected": -0.414907306432724,
2109
+ "step": 1490
2110
+ },
2111
+ {
2112
+ "epoch": 0.8,
2113
+ "learning_rate": 5.852214394262515e-07,
2114
+ "logits/chosen": -1.4020814895629883,
2115
+ "logits/rejected": -0.29137012362480164,
2116
+ "logps/chosen": -722.795654296875,
2117
+ "logps/rejected": -1755.2005615234375,
2118
+ "loss": 0.145,
2119
+ "rewards/accuracies": 0.875,
2120
+ "rewards/chosen": -0.05530545115470886,
2121
+ "rewards/margins": 0.3759249448776245,
2122
+ "rewards/rejected": -0.431230366230011,
2123
+ "step": 1500
2124
+ },
2125
+ {
2126
+ "epoch": 0.81,
2127
+ "learning_rate": 5.556222358149191e-07,
2128
+ "logits/chosen": -1.082096815109253,
2129
+ "logits/rejected": -0.2843102216720581,
2130
+ "logps/chosen": -636.3153076171875,
2131
+ "logps/rejected": -1611.7149658203125,
2132
+ "loss": 0.1259,
2133
+ "rewards/accuracies": 0.824999988079071,
2134
+ "rewards/chosen": -0.05015719681978226,
2135
+ "rewards/margins": 0.33069074153900146,
2136
+ "rewards/rejected": -0.3808479905128479,
2137
+ "step": 1510
2138
+ },
2139
+ {
2140
+ "epoch": 0.81,
2141
+ "learning_rate": 5.266973081622992e-07,
2142
+ "logits/chosen": -1.190901756286621,
2143
+ "logits/rejected": -0.3367752134799957,
2144
+ "logps/chosen": -654.8163452148438,
2145
+ "logps/rejected": -1681.1123046875,
2146
+ "loss": 0.1334,
2147
+ "rewards/accuracies": 0.875,
2148
+ "rewards/chosen": -0.05181831866502762,
2149
+ "rewards/margins": 0.37140342593193054,
2150
+ "rewards/rejected": -0.42322173714637756,
2151
+ "step": 1520
2152
+ },
2153
+ {
2154
+ "epoch": 0.82,
2155
+ "learning_rate": 4.984566871237942e-07,
2156
+ "logits/chosen": -1.569336175918579,
2157
+ "logits/rejected": -0.28200453519821167,
2158
+ "logps/chosen": -558.4755859375,
2159
+ "logps/rejected": -1696.124267578125,
2160
+ "loss": 0.1255,
2161
+ "rewards/accuracies": 0.8999999761581421,
2162
+ "rewards/chosen": -0.019404267892241478,
2163
+ "rewards/margins": 0.42805781960487366,
2164
+ "rewards/rejected": -0.4474620819091797,
2165
+ "step": 1530
2166
+ },
2167
+ {
2168
+ "epoch": 0.82,
2169
+ "learning_rate": 4.709101660493251e-07,
2170
+ "logits/chosen": -1.3863935470581055,
2171
+ "logits/rejected": -0.5687643885612488,
2172
+ "logps/chosen": -742.7957153320312,
2173
+ "logps/rejected": -1831.429931640625,
2174
+ "loss": 0.1213,
2175
+ "rewards/accuracies": 0.949999988079071,
2176
+ "rewards/chosen": -0.082461878657341,
2177
+ "rewards/margins": 0.3909619450569153,
2178
+ "rewards/rejected": -0.4734238088130951,
2179
+ "step": 1540
2180
+ },
2181
+ {
2182
+ "epoch": 0.83,
2183
+ "learning_rate": 4.440672975871743e-07,
2184
+ "logits/chosen": -1.4774668216705322,
2185
+ "logits/rejected": -0.9906954765319824,
2186
+ "logps/chosen": -623.5938720703125,
2187
+ "logps/rejected": -1661.2242431640625,
2188
+ "loss": 0.1343,
2189
+ "rewards/accuracies": 0.8500000238418579,
2190
+ "rewards/chosen": -0.0459056980907917,
2191
+ "rewards/margins": 0.38650375604629517,
2192
+ "rewards/rejected": -0.43240952491760254,
2193
+ "step": 1550
2194
+ },
2195
+ {
2196
+ "epoch": 0.83,
2197
+ "learning_rate": 4.1793739037129134e-07,
2198
+ "logits/chosen": -1.197632074356079,
2199
+ "logits/rejected": -0.7537062168121338,
2200
+ "logps/chosen": -547.0568237304688,
2201
+ "logps/rejected": -1868.6796875,
2202
+ "loss": 0.1353,
2203
+ "rewards/accuracies": 0.925000011920929,
2204
+ "rewards/chosen": -0.03627854213118553,
2205
+ "rewards/margins": 0.4777587950229645,
2206
+ "rewards/rejected": -0.5140373110771179,
2207
+ "step": 1560
2208
+ },
2209
+ {
2210
+ "epoch": 0.84,
2211
+ "learning_rate": 3.9252950579322405e-07,
2212
+ "logits/chosen": -1.4480640888214111,
2213
+ "logits/rejected": -0.6620736718177795,
2214
+ "logps/chosen": -862.5847778320312,
2215
+ "logps/rejected": -1864.031005859375,
2216
+ "loss": 0.1868,
2217
+ "rewards/accuracies": 0.824999988079071,
2218
+ "rewards/chosen": -0.1275833547115326,
2219
+ "rewards/margins": 0.371966689825058,
2220
+ "rewards/rejected": -0.4995500147342682,
2221
+ "step": 1570
2222
+ },
2223
+ {
2224
+ "epoch": 0.84,
2225
+ "learning_rate": 3.6785245485978864e-07,
2226
+ "logits/chosen": -1.5964170694351196,
2227
+ "logits/rejected": -0.689164400100708,
2228
+ "logps/chosen": -763.082763671875,
2229
+ "logps/rejected": -1659.866943359375,
2230
+ "loss": 0.177,
2231
+ "rewards/accuracies": 0.8500000238418579,
2232
+ "rewards/chosen": -0.1088029146194458,
2233
+ "rewards/margins": 0.3148026764392853,
2234
+ "rewards/rejected": -0.4236055910587311,
2235
+ "step": 1580
2236
+ },
2237
+ {
2238
+ "epoch": 0.85,
2239
+ "learning_rate": 3.43914795137566e-07,
2240
+ "logits/chosen": -1.4386413097381592,
2241
+ "logits/rejected": -0.14032740890979767,
2242
+ "logps/chosen": -754.867431640625,
2243
+ "logps/rejected": -1875.2099609375,
2244
+ "loss": 0.1129,
2245
+ "rewards/accuracies": 0.875,
2246
+ "rewards/chosen": -0.08381593227386475,
2247
+ "rewards/margins": 0.4074801504611969,
2248
+ "rewards/rejected": -0.49129611253738403,
2249
+ "step": 1590
2250
+ },
2251
+ {
2252
+ "epoch": 0.85,
2253
+ "learning_rate": 3.207248277852901e-07,
2254
+ "logits/chosen": -1.4534538984298706,
2255
+ "logits/rejected": -0.4819762110710144,
2256
+ "logps/chosen": -752.8095703125,
2257
+ "logps/rejected": -1707.563232421875,
2258
+ "loss": 0.1529,
2259
+ "rewards/accuracies": 0.7250000238418579,
2260
+ "rewards/chosen": -0.10240931808948517,
2261
+ "rewards/margins": 0.3559865653514862,
2262
+ "rewards/rejected": -0.45839595794677734,
2263
+ "step": 1600
2264
+ },
2265
+ {
2266
+ "epoch": 0.86,
2267
+ "learning_rate": 2.9829059467515074e-07,
2268
+ "logits/chosen": -1.6940540075302124,
2269
+ "logits/rejected": -0.40000471472740173,
2270
+ "logps/chosen": -623.6875610351562,
2271
+ "logps/rejected": -1585.805908203125,
2272
+ "loss": 0.1576,
2273
+ "rewards/accuracies": 0.8500000238418579,
2274
+ "rewards/chosen": -0.046424802392721176,
2275
+ "rewards/margins": 0.34341102838516235,
2276
+ "rewards/rejected": -0.38983583450317383,
2277
+ "step": 1610
2278
+ },
2279
+ {
2280
+ "epoch": 0.86,
2281
+ "learning_rate": 2.766198756040153e-07,
2282
+ "logits/chosen": -1.2030726671218872,
2283
+ "logits/rejected": -0.2915817201137543,
2284
+ "logps/chosen": -833.3904418945312,
2285
+ "logps/rejected": -1901.398681640625,
2286
+ "loss": 0.1341,
2287
+ "rewards/accuracies": 0.949999988079071,
2288
+ "rewards/chosen": -0.07304056733846664,
2289
+ "rewards/margins": 0.4044952392578125,
2290
+ "rewards/rejected": -0.47753578424453735,
2291
+ "step": 1620
2292
+ },
2293
+ {
2294
+ "epoch": 0.87,
2295
+ "learning_rate": 2.5572018559553155e-07,
2296
+ "logits/chosen": -1.4762271642684937,
2297
+ "logits/rejected": -0.5102109909057617,
2298
+ "logps/chosen": -739.6577758789062,
2299
+ "logps/rejected": -1855.5865478515625,
2300
+ "loss": 0.0915,
2301
+ "rewards/accuracies": 0.8500000238418579,
2302
+ "rewards/chosen": -0.08025506138801575,
2303
+ "rewards/margins": 0.4267922043800354,
2304
+ "rewards/rejected": -0.5070472955703735,
2305
+ "step": 1630
2306
+ },
2307
+ {
2308
+ "epoch": 0.87,
2309
+ "learning_rate": 2.3559877229404864e-07,
2310
+ "logits/chosen": -1.327376127243042,
2311
+ "logits/rejected": -0.7707889080047607,
2312
+ "logps/chosen": -711.1609497070312,
2313
+ "logps/rejected": -1628.607421875,
2314
+ "loss": 0.1191,
2315
+ "rewards/accuracies": 0.7749999761581421,
2316
+ "rewards/chosen": -0.07269873470067978,
2317
+ "rewards/margins": 0.33979684114456177,
2318
+ "rewards/rejected": -0.41249552369117737,
2319
+ "step": 1640
2320
+ },
2321
+ {
2322
+ "epoch": 0.88,
2323
+ "learning_rate": 2.1626261345126576e-07,
2324
+ "logits/chosen": -1.2067131996154785,
2325
+ "logits/rejected": -0.12356331199407578,
2326
+ "logps/chosen": -640.2079467773438,
2327
+ "logps/rejected": -1719.388427734375,
2328
+ "loss": 0.1247,
2329
+ "rewards/accuracies": 0.8500000238418579,
2330
+ "rewards/chosen": -0.05858426168560982,
2331
+ "rewards/margins": 0.4315997064113617,
2332
+ "rewards/rejected": -0.4901839792728424,
2333
+ "step": 1650
2334
+ },
2335
+ {
2336
+ "epoch": 0.89,
2337
+ "learning_rate": 1.9771841450646505e-07,
2338
+ "logits/chosen": -1.2794129848480225,
2339
+ "logits/rejected": -0.40672388672828674,
2340
+ "logps/chosen": -641.7750244140625,
2341
+ "logps/rejected": -1771.903564453125,
2342
+ "loss": 0.1566,
2343
+ "rewards/accuracies": 0.824999988079071,
2344
+ "rewards/chosen": -0.08581944555044174,
2345
+ "rewards/margins": 0.3960307240486145,
2346
+ "rewards/rejected": -0.4818502366542816,
2347
+ "step": 1660
2348
+ },
2349
+ {
2350
+ "epoch": 0.89,
2351
+ "learning_rate": 1.7997260626118758e-07,
2352
+ "logits/chosen": -1.116964340209961,
2353
+ "logits/rejected": -0.30027687549591064,
2354
+ "logps/chosen": -827.5680541992188,
2355
+ "logps/rejected": -1756.0361328125,
2356
+ "loss": 0.1818,
2357
+ "rewards/accuracies": 0.800000011920929,
2358
+ "rewards/chosen": -0.13185694813728333,
2359
+ "rewards/margins": 0.34328708052635193,
2360
+ "rewards/rejected": -0.47514405846595764,
2361
+ "step": 1670
2362
+ },
2363
+ {
2364
+ "epoch": 0.9,
2365
+ "learning_rate": 1.6303134264914365e-07,
2366
+ "logits/chosen": -1.050196647644043,
2367
+ "logits/rejected": -0.43848666548728943,
2368
+ "logps/chosen": -749.1686401367188,
2369
+ "logps/rejected": -1560.6253662109375,
2370
+ "loss": 0.2079,
2371
+ "rewards/accuracies": 0.800000011920929,
2372
+ "rewards/chosen": -0.1045556515455246,
2373
+ "rewards/margins": 0.2498161792755127,
2374
+ "rewards/rejected": -0.3543718457221985,
2375
+ "step": 1680
2376
+ },
2377
+ {
2378
+ "epoch": 0.9,
2379
+ "learning_rate": 1.469004986021355e-07,
2380
+ "logits/chosen": -1.199964165687561,
2381
+ "logits/rejected": -0.08608709275722504,
2382
+ "logps/chosen": -697.6527099609375,
2383
+ "logps/rejected": -1817.315673828125,
2384
+ "loss": 0.1241,
2385
+ "rewards/accuracies": 0.8999999761581421,
2386
+ "rewards/chosen": -0.10202528536319733,
2387
+ "rewards/margins": 0.3813837170600891,
2388
+ "rewards/rejected": -0.48340901732444763,
2389
+ "step": 1690
2390
+ },
2391
+ {
2392
+ "epoch": 0.91,
2393
+ "learning_rate": 1.315856680127367e-07,
2394
+ "logits/chosen": -1.1728847026824951,
2395
+ "logits/rejected": -0.7924209833145142,
2396
+ "logps/chosen": -860.1814575195312,
2397
+ "logps/rejected": -2008.0902099609375,
2398
+ "loss": 0.1263,
2399
+ "rewards/accuracies": 0.949999988079071,
2400
+ "rewards/chosen": -0.09493857622146606,
2401
+ "rewards/margins": 0.3988199830055237,
2402
+ "rewards/rejected": -0.49375852942466736,
2403
+ "step": 1700
2404
+ },
2405
+ {
2406
+ "epoch": 0.91,
2407
+ "learning_rate": 1.1709216179442817e-07,
2408
+ "logits/chosen": -1.3747029304504395,
2409
+ "logits/rejected": -0.3935699760913849,
2410
+ "logps/chosen": -706.3656005859375,
2411
+ "logps/rejected": -1836.0609130859375,
2412
+ "loss": 0.1349,
2413
+ "rewards/accuracies": 0.925000011920929,
2414
+ "rewards/chosen": -0.062370091676712036,
2415
+ "rewards/margins": 0.4179624915122986,
2416
+ "rewards/rejected": -0.480332612991333,
2417
+ "step": 1710
2418
+ },
2419
+ {
2420
+ "epoch": 0.92,
2421
+ "learning_rate": 1.0342500603986421e-07,
2422
+ "logits/chosen": -1.3793084621429443,
2423
+ "logits/rejected": -0.1095462292432785,
2424
+ "logps/chosen": -771.80810546875,
2425
+ "logps/rejected": -1763.7025146484375,
2426
+ "loss": 0.1259,
2427
+ "rewards/accuracies": 0.925000011920929,
2428
+ "rewards/chosen": -0.061395853757858276,
2429
+ "rewards/margins": 0.4151260256767273,
2430
+ "rewards/rejected": -0.47652187943458557,
2431
+ "step": 1720
2432
+ },
2433
+ {
2434
+ "epoch": 0.92,
2435
+ "learning_rate": 9.058894027791643e-08,
2436
+ "logits/chosen": -1.151395559310913,
2437
+ "logits/rejected": 0.09823288023471832,
2438
+ "logps/chosen": -677.2811889648438,
2439
+ "logps/rejected": -1689.9427490234375,
2440
+ "loss": 0.1347,
2441
+ "rewards/accuracies": 0.8500000238418579,
2442
+ "rewards/chosen": -0.054931480437517166,
2443
+ "rewards/margins": 0.3742934763431549,
2444
+ "rewards/rejected": -0.4292249083518982,
2445
+ "step": 1730
2446
+ },
2447
+ {
2448
+ "epoch": 0.93,
2449
+ "learning_rate": 7.858841583008592e-08,
2450
+ "logits/chosen": -1.3676621913909912,
2451
+ "logits/rejected": -0.6429020166397095,
2452
+ "logps/chosen": -617.450927734375,
2453
+ "logps/rejected": -1714.8131103515625,
2454
+ "loss": 0.124,
2455
+ "rewards/accuracies": 0.8500000238418579,
2456
+ "rewards/chosen": -0.05651504918932915,
2457
+ "rewards/margins": 0.4177074432373047,
2458
+ "rewards/rejected": -0.4742225110530853,
2459
+ "step": 1740
2460
+ },
2461
+ {
2462
+ "epoch": 0.93,
2463
+ "learning_rate": 6.742759426686313e-08,
2464
+ "logits/chosen": -1.183172583580017,
2465
+ "logits/rejected": -0.5100821256637573,
2466
+ "logps/chosen": -752.4456787109375,
2467
+ "logps/rejected": -1788.185302734375,
2468
+ "loss": 0.1374,
2469
+ "rewards/accuracies": 0.8500000238418579,
2470
+ "rewards/chosen": -0.09715384244918823,
2471
+ "rewards/margins": 0.3738781809806824,
2472
+ "rewards/rejected": -0.471032053232193,
2473
+ "step": 1750
2474
+ },
2475
+ {
2476
+ "epoch": 0.94,
2477
+ "learning_rate": 5.7110345964571104e-08,
2478
+ "logits/chosen": -1.42495596408844,
2479
+ "logits/rejected": -0.7022902369499207,
2480
+ "logps/chosen": -604.3352661132812,
2481
+ "logps/rejected": -1618.843017578125,
2482
+ "loss": 0.1543,
2483
+ "rewards/accuracies": 0.800000011920929,
2484
+ "rewards/chosen": -0.03320794552564621,
2485
+ "rewards/margins": 0.35301321744918823,
2486
+ "rewards/rejected": -0.38622111082077026,
2487
+ "step": 1760
2488
+ },
2489
+ {
2490
+ "epoch": 0.94,
2491
+ "learning_rate": 4.764024876318357e-08,
2492
+ "logits/chosen": -1.4258668422698975,
2493
+ "logits/rejected": -0.1909623146057129,
2494
+ "logps/chosen": -670.0120849609375,
2495
+ "logps/rejected": -1743.475830078125,
2496
+ "loss": 0.077,
2497
+ "rewards/accuracies": 0.8999999761581421,
2498
+ "rewards/chosen": -0.05731179565191269,
2499
+ "rewards/margins": 0.4044620394706726,
2500
+ "rewards/rejected": -0.4617738723754883,
2501
+ "step": 1770
2502
+ },
2503
+ {
2504
+ "epoch": 0.95,
2505
+ "learning_rate": 3.902058672559633e-08,
2506
+ "logits/chosen": -1.355021595954895,
2507
+ "logits/rejected": -0.29654452204704285,
2508
+ "logps/chosen": -578.4010009765625,
2509
+ "logps/rejected": -1676.61328125,
2510
+ "loss": 0.175,
2511
+ "rewards/accuracies": 0.875,
2512
+ "rewards/chosen": -0.03146610036492348,
2513
+ "rewards/margins": 0.42508259415626526,
2514
+ "rewards/rejected": -0.45654863119125366,
2515
+ "step": 1780
2516
+ },
2517
+ {
2518
+ "epoch": 0.95,
2519
+ "learning_rate": 3.125434899876933e-08,
2520
+ "logits/chosen": -1.3393023014068604,
2521
+ "logits/rejected": -0.4936142861843109,
2522
+ "logps/chosen": -564.572509765625,
2523
+ "logps/rejected": -1514.8624267578125,
2524
+ "loss": 0.1283,
2525
+ "rewards/accuracies": 0.7749999761581421,
2526
+ "rewards/chosen": -0.05059174820780754,
2527
+ "rewards/margins": 0.33098191022872925,
2528
+ "rewards/rejected": -0.3815736770629883,
2529
+ "step": 1790
2530
+ },
2531
+ {
2532
+ "epoch": 0.96,
2533
+ "learning_rate": 2.4344228777145873e-08,
2534
+ "logits/chosen": -1.2701542377471924,
2535
+ "logits/rejected": -0.29851651191711426,
2536
+ "logps/chosen": -603.8160400390625,
2537
+ "logps/rejected": -1694.4417724609375,
2538
+ "loss": 0.15,
2539
+ "rewards/accuracies": 0.875,
2540
+ "rewards/chosen": -0.03015061281621456,
2541
+ "rewards/margins": 0.3750046491622925,
2542
+ "rewards/rejected": -0.4051552712917328,
2543
+ "step": 1800
2544
+ },
2545
+ {
2546
+ "epoch": 0.97,
2547
+ "learning_rate": 1.829262236869772e-08,
2548
+ "logits/chosen": -1.1179814338684082,
2549
+ "logits/rejected": -0.27846160531044006,
2550
+ "logps/chosen": -722.6538696289062,
2551
+ "logps/rejected": -1610.6104736328125,
2552
+ "loss": 0.153,
2553
+ "rewards/accuracies": 0.75,
2554
+ "rewards/chosen": -0.06969080120325089,
2555
+ "rewards/margins": 0.3620641827583313,
2556
+ "rewards/rejected": -0.4317549169063568,
2557
+ "step": 1810
2558
+ },
2559
+ {
2560
+ "epoch": 0.97,
2561
+ "learning_rate": 1.3101628363929586e-08,
2562
+ "logits/chosen": -1.2972911596298218,
2563
+ "logits/rejected": -0.3146804869174957,
2564
+ "logps/chosen": -702.9611206054688,
2565
+ "logps/rejected": -1858.8841552734375,
2566
+ "loss": 0.1325,
2567
+ "rewards/accuracies": 0.925000011920929,
2568
+ "rewards/chosen": -0.0757565051317215,
2569
+ "rewards/margins": 0.4196552634239197,
2570
+ "rewards/rejected": -0.4954117238521576,
2571
+ "step": 1820
2572
+ },
2573
+ {
2574
+ "epoch": 0.98,
2575
+ "learning_rate": 8.773046908123195e-09,
2576
+ "logits/chosen": -1.2295328378677368,
2577
+ "logits/rejected": -0.45538797974586487,
2578
+ "logps/chosen": -609.9122314453125,
2579
+ "logps/rejected": -1732.151611328125,
2580
+ "loss": 0.0956,
2581
+ "rewards/accuracies": 0.8999999761581421,
2582
+ "rewards/chosen": -0.047846533358097076,
2583
+ "rewards/margins": 0.40931883454322815,
2584
+ "rewards/rejected": -0.457165390253067,
2585
+ "step": 1830
2586
+ },
2587
+ {
2588
+ "epoch": 0.98,
2589
+ "learning_rate": 5.308379077080817e-09,
2590
+ "logits/chosen": -1.3349180221557617,
2591
+ "logits/rejected": -0.37136349081993103,
2592
+ "logps/chosen": -741.32763671875,
2593
+ "logps/rejected": -1853.3609619140625,
2594
+ "loss": 0.1685,
2595
+ "rewards/accuracies": 0.875,
2596
+ "rewards/chosen": -0.10401256382465363,
2597
+ "rewards/margins": 0.3636459708213806,
2598
+ "rewards/rejected": -0.4676585793495178,
2599
+ "step": 1840
2600
+ },
2601
+ {
2602
+ "epoch": 0.99,
2603
+ "learning_rate": 2.7088263565760996e-09,
2604
+ "logits/chosen": -1.2106130123138428,
2605
+ "logits/rejected": -0.03263585641980171,
2606
+ "logps/chosen": -627.7078857421875,
2607
+ "logps/rejected": -1669.301025390625,
2608
+ "loss": 0.167,
2609
+ "rewards/accuracies": 0.824999988079071,
2610
+ "rewards/chosen": -0.06294982880353928,
2611
+ "rewards/margins": 0.3623204827308655,
2612
+ "rewards/rejected": -0.4252702593803406,
2613
+ "step": 1850
2614
+ },
2615
+ {
2616
+ "epoch": 0.99,
2617
+ "learning_rate": 9.752902257023633e-10,
2618
+ "logits/chosen": -0.9112818837165833,
2619
+ "logits/rejected": -0.16677840054035187,
2620
+ "logps/chosen": -745.0077514648438,
2621
+ "logps/rejected": -1581.1097412109375,
2622
+ "loss": 0.1235,
2623
+ "rewards/accuracies": 0.8500000238418579,
2624
+ "rewards/chosen": -0.10941288620233536,
2625
+ "rewards/margins": 0.2992158532142639,
2626
+ "rewards/rejected": -0.40862876176834106,
2627
+ "step": 1860
2628
+ },
2629
+ {
2630
+ "epoch": 1.0,
2631
+ "learning_rate": 1.083718442532189e-10,
2632
+ "logits/chosen": -1.027451992034912,
2633
+ "logits/rejected": 0.37703937292099,
2634
+ "logps/chosen": -782.1871337890625,
2635
+ "logps/rejected": -1805.238037109375,
2636
+ "loss": 0.1393,
2637
+ "rewards/accuracies": 0.8500000238418579,
2638
+ "rewards/chosen": -0.11385799944400787,
2639
+ "rewards/margins": 0.39328035712242126,
2640
+ "rewards/rejected": -0.5071383714675903,
2641
+ "step": 1870
2642
+ },
2643
+ {
2644
+ "epoch": 1.0,
2645
+ "step": 1875,
2646
+ "total_flos": 0.0,
2647
+ "train_loss": 0.17237209237416584,
2648
+ "train_runtime": 10470.2127,
2649
+ "train_samples_per_second": 1.433,
2650
+ "train_steps_per_second": 0.179
2651
+ }
2652
+ ],
2653
+ "logging_steps": 10,
2654
+ "max_steps": 1875,
2655
+ "num_input_tokens_seen": 0,
2656
+ "num_train_epochs": 1,
2657
+ "save_steps": 100,
2658
+ "total_flos": 0.0,
2659
+ "train_batch_size": 2,
2660
+ "trial_name": null,
2661
+ "trial_params": null
2662
+ }