Commit
•
717c3ee
1
Parent(s):
6f327a0
Model save
Browse files- README.md +32 -22
- adapter_model.safetensors +1 -1
- all_results.json +5 -5
- train_results.json +5 -5
- trainer_state.json +2030 -90
README.md
CHANGED
|
@@ -1,13 +1,11 @@
|
|
| 1 |
---
|
| 2 |
library_name: peft
|
| 3 |
tags:
|
| 4 |
-
- alignment-handbook
|
| 5 |
- trl
|
| 6 |
- dpo
|
|
|
|
| 7 |
- generated_from_trainer
|
| 8 |
base_model: NbAiLab/nb-gpt-j-6B-v2
|
| 9 |
-
datasets:
|
| 10 |
-
- hugodk-sch/aftonposten_title_prefs
|
| 11 |
model-index:
|
| 12 |
- name: aftonposten-6b-align-scan
|
| 13 |
results: []
|
|
@@ -18,17 +16,17 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 18 |
|
| 19 |
# aftonposten-6b-align-scan
|
| 20 |
|
| 21 |
-
This model is a fine-tuned version of [
|
| 22 |
It achieves the following results on the evaluation set:
|
| 23 |
-
- Loss: 0.
|
| 24 |
-
- Rewards/chosen: -0.
|
| 25 |
-
- Rewards/rejected: -0.
|
| 26 |
-
- Rewards/accuracies: 0.
|
| 27 |
-
- Rewards/margins: 0.
|
| 28 |
-
- Logps/rejected: -
|
| 29 |
-
- Logps/chosen: -34.
|
| 30 |
-
- Logits/rejected: -2.
|
| 31 |
-
- Logits/chosen: -2.
|
| 32 |
|
| 33 |
## Model description
|
| 34 |
|
|
@@ -57,21 +55,33 @@ The following hyperparameters were used during training:
|
|
| 57 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 58 |
- lr_scheduler_type: cosine
|
| 59 |
- lr_scheduler_warmup_ratio: 0.1
|
| 60 |
-
- num_epochs:
|
| 61 |
|
| 62 |
### Training results
|
| 63 |
|
| 64 |
-
| Training Loss | Epoch | Step |
|
| 65 |
-
|
| 66 |
-
| 0.6577 | 0.26 | 100 |
|
| 67 |
-
| 0.6156 | 0.52 | 200 |
|
| 68 |
-
| 0.5468 | 0.78 | 300 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 69 |
|
| 70 |
|
| 71 |
### Framework versions
|
| 72 |
|
| 73 |
-
- PEFT 0.
|
| 74 |
-
- Transformers 4.
|
| 75 |
- Pytorch 2.1.2+cu121
|
| 76 |
-
- Datasets 2.
|
| 77 |
- Tokenizers 0.15.1
|
|
|
|
| 1 |
---
|
| 2 |
library_name: peft
|
| 3 |
tags:
|
|
|
|
| 4 |
- trl
|
| 5 |
- dpo
|
| 6 |
+
- alignment-handbook
|
| 7 |
- generated_from_trainer
|
| 8 |
base_model: NbAiLab/nb-gpt-j-6B-v2
|
|
|
|
|
|
|
| 9 |
model-index:
|
| 10 |
- name: aftonposten-6b-align-scan
|
| 11 |
results: []
|
|
|
|
| 16 |
|
| 17 |
# aftonposten-6b-align-scan
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [NbAiLab/nb-gpt-j-6B-v2](https://huggingface.co/NbAiLab/nb-gpt-j-6B-v2) on the None dataset.
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
+
- Loss: 0.6831
|
| 22 |
+
- Rewards/chosen: -0.2684
|
| 23 |
+
- Rewards/rejected: -0.4218
|
| 24 |
+
- Rewards/accuracies: 0.5801
|
| 25 |
+
- Rewards/margins: 0.1534
|
| 26 |
+
- Logps/rejected: -38.1192
|
| 27 |
+
- Logps/chosen: -34.4180
|
| 28 |
+
- Logits/rejected: -2.1173
|
| 29 |
+
- Logits/chosen: -2.1220
|
| 30 |
|
| 31 |
## Model description
|
| 32 |
|
|
|
|
| 55 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 56 |
- lr_scheduler_type: cosine
|
| 57 |
- lr_scheduler_warmup_ratio: 0.1
|
| 58 |
+
- num_epochs: 4
|
| 59 |
|
| 60 |
### Training results
|
| 61 |
|
| 62 |
+
| Training Loss | Epoch | Step | Logits/chosen | Logits/rejected | Logps/chosen | Logps/rejected | Validation Loss | Rewards/accuracies | Rewards/chosen | Rewards/margins | Rewards/rejected |
|
| 63 |
+
|:-------------:|:-----:|:----:|:-------------:|:---------------:|:------------:|:--------------:|:---------------:|:------------------:|:--------------:|:---------------:|:----------------:|
|
| 64 |
+
| 0.6577 | 0.26 | 100 | -2.2338 | -2.2289 | -34.0400 | -37.5357 | 0.6949 | 0.5316 | -0.0038 | 0.0095 | -0.0133 |
|
| 65 |
+
| 0.6156 | 0.52 | 200 | -2.2317 | -2.2268 | -34.0535 | -37.5578 | 0.6943 | 0.5191 | -0.0132 | 0.0156 | -0.0288 |
|
| 66 |
+
| 0.5468 | 0.78 | 300 | -2.2277 | -2.2229 | -34.0684 | -37.5860 | 0.6903 | 0.5191 | -0.0237 | 0.0249 | -0.0486 |
|
| 67 |
+
| 0.4243 | 1.04 | 400 | 0.6886 | -0.0963 | -0.1489 | 0.5540 | 0.0526 | -37.7293 | -34.1721 | -2.1933 | -2.1981 |
|
| 68 |
+
| 0.353 | 1.3 | 500 | 0.6901 | -0.1994 | -0.2743 | 0.5569 | 0.0749 | -37.9085 | -34.3194 | -2.1884 | -2.1932 |
|
| 69 |
+
| 0.3554 | 1.56 | 600 | 0.6763 | -0.1468 | -0.2572 | 0.5806 | 0.1103 | -37.8840 | -34.2443 | -2.1701 | -2.1749 |
|
| 70 |
+
| 0.3333 | 1.82 | 700 | 0.6813 | -0.1817 | -0.2946 | 0.5743 | 0.1129 | -37.9375 | -34.2941 | -2.1549 | -2.1596 |
|
| 71 |
+
| 0.2025 | 2.08 | 800 | 0.6800 | -0.2316 | -0.3667 | 0.5660 | 0.1351 | -38.0405 | -34.3655 | -2.1413 | -2.1461 |
|
| 72 |
+
| 0.2153 | 2.34 | 900 | 0.6866 | -0.2482 | -0.3826 | 0.5835 | 0.1344 | -38.0632 | -34.3891 | -2.1292 | -2.1340 |
|
| 73 |
+
| 0.2381 | 2.6 | 1000 | 0.6821 | -0.2624 | -0.4162 | 0.5864 | 0.1538 | -38.1112 | -34.4094 | -2.1207 | -2.1255 |
|
| 74 |
+
| 0.1898 | 2.86 | 1100 | 0.6858 | -0.2673 | -0.4161 | 0.5831 | 0.1487 | -38.1110 | -34.4164 | -2.1188 | -2.1235 |
|
| 75 |
+
| 0.2231 | 3.12 | 1200 | 0.6780 | -0.2626 | -0.4264 | 0.5889 | 0.1637 | -38.1257 | -34.4097 | -2.1175 | -2.1223 |
|
| 76 |
+
| 0.164 | 3.38 | 1300 | 0.6834 | -0.2678 | -0.4194 | 0.5947 | 0.1516 | -38.1158 | -34.4172 | -2.1174 | -2.1221 |
|
| 77 |
+
| 0.1562 | 3.64 | 1400 | 0.6753 | -0.2666 | -0.4361 | 0.5922 | 0.1696 | -38.1396 | -34.4154 | -2.1172 | -2.1219 |
|
| 78 |
+
| 0.2163 | 3.9 | 1500 | 0.6831 | -0.2684 | -0.4218 | 0.5801 | 0.1534 | -38.1192 | -34.4180 | -2.1173 | -2.1220 |
|
| 79 |
|
| 80 |
|
| 81 |
### Framework versions
|
| 82 |
|
| 83 |
+
- PEFT 0.10.0
|
| 84 |
+
- Transformers 4.39.0.dev0
|
| 85 |
- Pytorch 2.1.2+cu121
|
| 86 |
+
- Datasets 2.14.6
|
| 87 |
- Tokenizers 0.15.1
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 176183216
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:026a6415f2106bbe9c74c61bfa08dae685e1f6c824a2d2e66189c50a8330cb3f
|
| 3 |
size 176183216
|
all_results.json
CHANGED
|
@@ -1,5 +1,5 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
"eval_logits/chosen": -2.227961301803589,
|
| 4 |
"eval_logits/rejected": -2.2231268882751465,
|
| 5 |
"eval_logps/chosen": -34.07651138305664,
|
|
@@ -13,9 +13,9 @@
|
|
| 13 |
"eval_samples": 343,
|
| 14 |
"eval_samples_per_second": 2.356,
|
| 15 |
"eval_steps_per_second": 0.295,
|
| 16 |
-
"train_loss": 0.
|
| 17 |
-
"train_runtime":
|
| 18 |
"train_samples": 3079,
|
| 19 |
-
"train_samples_per_second":
|
| 20 |
-
"train_steps_per_second": 0.
|
| 21 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 4.0,
|
| 3 |
"eval_logits/chosen": -2.227961301803589,
|
| 4 |
"eval_logits/rejected": -2.2231268882751465,
|
| 5 |
"eval_logps/chosen": -34.07651138305664,
|
|
|
|
| 13 |
"eval_samples": 343,
|
| 14 |
"eval_samples_per_second": 2.356,
|
| 15 |
"eval_steps_per_second": 0.295,
|
| 16 |
+
"train_loss": 0.2216110908365869,
|
| 17 |
+
"train_runtime": 10768.4727,
|
| 18 |
"train_samples": 3079,
|
| 19 |
+
"train_samples_per_second": 1.144,
|
| 20 |
+
"train_steps_per_second": 0.143
|
| 21 |
}
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"train_loss": 0.
|
| 4 |
-
"train_runtime":
|
| 5 |
"train_samples": 3079,
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 4.0,
|
| 3 |
+
"train_loss": 0.2216110908365869,
|
| 4 |
+
"train_runtime": 10768.4727,
|
| 5 |
"train_samples": 3079,
|
| 6 |
+
"train_samples_per_second": 1.144,
|
| 7 |
+
"train_steps_per_second": 0.143
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 100,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -492,130 +492,2070 @@
|
|
| 492 |
},
|
| 493 |
{
|
| 494 |
"epoch": 0.81,
|
| 495 |
-
"
|
| 496 |
-
"
|
| 497 |
-
"logits/
|
| 498 |
-
"
|
| 499 |
-
"logps/
|
| 500 |
-
"
|
| 501 |
-
"
|
| 502 |
-
"rewards/
|
| 503 |
-
"rewards/
|
| 504 |
-
"rewards/
|
|
|
|
| 505 |
"step": 310
|
| 506 |
},
|
| 507 |
{
|
| 508 |
"epoch": 0.83,
|
| 509 |
-
"
|
| 510 |
-
"
|
| 511 |
-
"logits/
|
| 512 |
-
"
|
| 513 |
-
"logps/
|
| 514 |
-
"
|
| 515 |
-
"
|
| 516 |
-
"rewards/
|
| 517 |
-
"rewards/
|
| 518 |
-
"rewards/
|
|
|
|
| 519 |
"step": 320
|
| 520 |
},
|
| 521 |
{
|
| 522 |
"epoch": 0.86,
|
| 523 |
-
"
|
| 524 |
-
"
|
| 525 |
-
"logits/
|
| 526 |
-
"
|
| 527 |
-
"logps/
|
| 528 |
-
"
|
| 529 |
-
"
|
| 530 |
-
"rewards/
|
| 531 |
-
"rewards/
|
| 532 |
-
"rewards/
|
|
|
|
| 533 |
"step": 330
|
| 534 |
},
|
| 535 |
{
|
| 536 |
"epoch": 0.88,
|
| 537 |
-
"
|
| 538 |
-
"
|
| 539 |
-
"logits/
|
| 540 |
-
"
|
| 541 |
-
"logps/
|
| 542 |
-
"
|
| 543 |
-
"
|
| 544 |
-
"rewards/
|
| 545 |
-
"rewards/
|
| 546 |
-
"rewards/
|
|
|
|
| 547 |
"step": 340
|
| 548 |
},
|
| 549 |
{
|
| 550 |
"epoch": 0.91,
|
| 551 |
-
"
|
| 552 |
-
"
|
| 553 |
-
"logits/
|
| 554 |
-
"
|
| 555 |
-
"logps/
|
| 556 |
-
"
|
| 557 |
-
"
|
| 558 |
-
"rewards/
|
| 559 |
-
"rewards/
|
| 560 |
-
"rewards/
|
|
|
|
| 561 |
"step": 350
|
| 562 |
},
|
| 563 |
{
|
| 564 |
"epoch": 0.94,
|
| 565 |
-
"
|
| 566 |
-
"
|
| 567 |
-
"logits/
|
| 568 |
-
"
|
| 569 |
-
"logps/
|
| 570 |
-
"
|
| 571 |
-
"
|
| 572 |
-
"rewards/
|
| 573 |
-
"rewards/
|
| 574 |
-
"rewards/
|
|
|
|
| 575 |
"step": 360
|
| 576 |
},
|
| 577 |
{
|
| 578 |
"epoch": 0.96,
|
| 579 |
-
"
|
| 580 |
-
"
|
| 581 |
-
"logits/
|
| 582 |
-
"
|
| 583 |
-
"logps/
|
| 584 |
-
"
|
| 585 |
-
"
|
| 586 |
-
"rewards/
|
| 587 |
-
"rewards/
|
| 588 |
-
"rewards/
|
|
|
|
| 589 |
"step": 370
|
| 590 |
},
|
| 591 |
{
|
| 592 |
"epoch": 0.99,
|
| 593 |
-
"
|
| 594 |
-
"
|
| 595 |
-
"logits/
|
| 596 |
-
"
|
| 597 |
-
"logps/
|
| 598 |
-
"
|
| 599 |
-
"
|
| 600 |
-
"rewards/
|
| 601 |
-
"rewards/
|
| 602 |
-
"rewards/
|
|
|
|
| 603 |
"step": 380
|
| 604 |
},
|
| 605 |
{
|
| 606 |
-
"epoch": 1.
|
| 607 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 608 |
"total_flos": 0.0,
|
| 609 |
-
"train_loss": 0.
|
| 610 |
-
"train_runtime":
|
| 611 |
-
"train_samples_per_second":
|
| 612 |
-
"train_steps_per_second": 0.
|
| 613 |
}
|
| 614 |
],
|
| 615 |
"logging_steps": 10,
|
| 616 |
-
"max_steps":
|
| 617 |
"num_input_tokens_seen": 0,
|
| 618 |
-
"num_train_epochs":
|
| 619 |
"save_steps": 100,
|
| 620 |
"total_flos": 0.0,
|
| 621 |
"train_batch_size": 4,
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 4.0,
|
| 5 |
"eval_steps": 100,
|
| 6 |
+
"global_step": 1540,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 492 |
},
|
| 493 |
{
|
| 494 |
"epoch": 0.81,
|
| 495 |
+
"grad_norm": 10.6875,
|
| 496 |
+
"learning_rate": 4.84533120650964e-06,
|
| 497 |
+
"logits/chosen": -2.0614776611328125,
|
| 498 |
+
"logits/rejected": -2.048677921295166,
|
| 499 |
+
"logps/chosen": -32.06366729736328,
|
| 500 |
+
"logps/rejected": -32.84763717651367,
|
| 501 |
+
"loss": 0.506,
|
| 502 |
+
"rewards/accuracies": 0.925000011920929,
|
| 503 |
+
"rewards/chosen": 0.32530978322029114,
|
| 504 |
+
"rewards/margins": 0.46991363167762756,
|
| 505 |
+
"rewards/rejected": -0.14460383355617523,
|
| 506 |
"step": 310
|
| 507 |
},
|
| 508 |
{
|
| 509 |
"epoch": 0.83,
|
| 510 |
+
"grad_norm": 10.3125,
|
| 511 |
+
"learning_rate": 4.825108134172131e-06,
|
| 512 |
+
"logits/chosen": -1.9717050790786743,
|
| 513 |
+
"logits/rejected": -1.9630792140960693,
|
| 514 |
+
"logps/chosen": -31.734317779541016,
|
| 515 |
+
"logps/rejected": -30.413864135742188,
|
| 516 |
+
"loss": 0.4921,
|
| 517 |
+
"rewards/accuracies": 0.824999988079071,
|
| 518 |
+
"rewards/chosen": 0.4344688355922699,
|
| 519 |
+
"rewards/margins": 0.555200457572937,
|
| 520 |
+
"rewards/rejected": -0.12073162943124771,
|
| 521 |
"step": 320
|
| 522 |
},
|
| 523 |
{
|
| 524 |
"epoch": 0.86,
|
| 525 |
+
"grad_norm": 11.6875,
|
| 526 |
+
"learning_rate": 4.80369052967602e-06,
|
| 527 |
+
"logits/chosen": -1.908839464187622,
|
| 528 |
+
"logits/rejected": -1.9209339618682861,
|
| 529 |
+
"logps/chosen": -29.812292098999023,
|
| 530 |
+
"logps/rejected": -33.6970329284668,
|
| 531 |
+
"loss": 0.4445,
|
| 532 |
+
"rewards/accuracies": 0.862500011920929,
|
| 533 |
+
"rewards/chosen": 0.46688413619995117,
|
| 534 |
+
"rewards/margins": 0.6732310056686401,
|
| 535 |
+
"rewards/rejected": -0.20634695887565613,
|
| 536 |
"step": 330
|
| 537 |
},
|
| 538 |
{
|
| 539 |
"epoch": 0.88,
|
| 540 |
+
"grad_norm": 15.5,
|
| 541 |
+
"learning_rate": 4.781089396387968e-06,
|
| 542 |
+
"logits/chosen": -1.871185064315796,
|
| 543 |
+
"logits/rejected": -1.8619896173477173,
|
| 544 |
+
"logps/chosen": -34.0085334777832,
|
| 545 |
+
"logps/rejected": -36.21852493286133,
|
| 546 |
+
"loss": 0.4426,
|
| 547 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 548 |
+
"rewards/chosen": 0.48327580094337463,
|
| 549 |
+
"rewards/margins": 0.7365350127220154,
|
| 550 |
+
"rewards/rejected": -0.25325921177864075,
|
| 551 |
"step": 340
|
| 552 |
},
|
| 553 |
{
|
| 554 |
"epoch": 0.91,
|
| 555 |
+
"grad_norm": 10.75,
|
| 556 |
+
"learning_rate": 4.757316345716554e-06,
|
| 557 |
+
"logits/chosen": -1.9235925674438477,
|
| 558 |
+
"logits/rejected": -1.9242353439331055,
|
| 559 |
+
"logps/chosen": -33.64180374145508,
|
| 560 |
+
"logps/rejected": -34.13081359863281,
|
| 561 |
+
"loss": 0.4532,
|
| 562 |
+
"rewards/accuracies": 0.8374999761581421,
|
| 563 |
+
"rewards/chosen": 0.5344331860542297,
|
| 564 |
+
"rewards/margins": 0.7317206859588623,
|
| 565 |
+
"rewards/rejected": -0.1972874402999878,
|
| 566 |
"step": 350
|
| 567 |
},
|
| 568 |
{
|
| 569 |
"epoch": 0.94,
|
| 570 |
+
"grad_norm": 11.875,
|
| 571 |
+
"learning_rate": 4.73238359114687e-06,
|
| 572 |
+
"logits/chosen": -2.048530101776123,
|
| 573 |
+
"logits/rejected": -2.054663896560669,
|
| 574 |
+
"logps/chosen": -31.074010848999023,
|
| 575 |
+
"logps/rejected": -32.97708511352539,
|
| 576 |
+
"loss": 0.4932,
|
| 577 |
+
"rewards/accuracies": 0.8374999761581421,
|
| 578 |
+
"rewards/chosen": 0.34946271777153015,
|
| 579 |
+
"rewards/margins": 0.5656248331069946,
|
| 580 |
+
"rewards/rejected": -0.21616211533546448,
|
| 581 |
"step": 360
|
| 582 |
},
|
| 583 |
{
|
| 584 |
"epoch": 0.96,
|
| 585 |
+
"grad_norm": 21.125,
|
| 586 |
+
"learning_rate": 4.706303941965804e-06,
|
| 587 |
+
"logits/chosen": -1.9776489734649658,
|
| 588 |
+
"logits/rejected": -1.977195382118225,
|
| 589 |
+
"logps/chosen": -32.816444396972656,
|
| 590 |
+
"logps/rejected": -36.346214294433594,
|
| 591 |
+
"loss": 0.4709,
|
| 592 |
+
"rewards/accuracies": 0.875,
|
| 593 |
+
"rewards/chosen": 0.4399483799934387,
|
| 594 |
+
"rewards/margins": 0.6474584937095642,
|
| 595 |
+
"rewards/rejected": -0.2075101137161255,
|
| 596 |
"step": 370
|
| 597 |
},
|
| 598 |
{
|
| 599 |
"epoch": 0.99,
|
| 600 |
+
"grad_norm": 9.5,
|
| 601 |
+
"learning_rate": 4.679090796681225e-06,
|
| 602 |
+
"logits/chosen": -2.0090365409851074,
|
| 603 |
+
"logits/rejected": -2.004446506500244,
|
| 604 |
+
"logps/chosen": -30.068859100341797,
|
| 605 |
+
"logps/rejected": -29.55948829650879,
|
| 606 |
+
"loss": 0.4523,
|
| 607 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 608 |
+
"rewards/chosen": 0.4492177367210388,
|
| 609 |
+
"rewards/margins": 0.6911332011222839,
|
| 610 |
+
"rewards/rejected": -0.24191546440124512,
|
| 611 |
"step": 380
|
| 612 |
},
|
| 613 |
{
|
| 614 |
+
"epoch": 1.01,
|
| 615 |
+
"grad_norm": 12.125,
|
| 616 |
+
"learning_rate": 4.650758136138454e-06,
|
| 617 |
+
"logits/chosen": -1.7796638011932373,
|
| 618 |
+
"logits/rejected": -1.786041498184204,
|
| 619 |
+
"logps/chosen": -31.63083267211914,
|
| 620 |
+
"logps/rejected": -36.7193603515625,
|
| 621 |
+
"loss": 0.3988,
|
| 622 |
+
"rewards/accuracies": 0.925000011920929,
|
| 623 |
+
"rewards/chosen": 0.5284172296524048,
|
| 624 |
+
"rewards/margins": 0.9491678476333618,
|
| 625 |
+
"rewards/rejected": -0.4207506775856018,
|
| 626 |
+
"step": 390
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.04,
|
| 630 |
+
"grad_norm": 9.875,
|
| 631 |
+
"learning_rate": 4.621320516337559e-06,
|
| 632 |
+
"logits/chosen": -1.9318805932998657,
|
| 633 |
+
"logits/rejected": -1.9255841970443726,
|
| 634 |
+
"logps/chosen": -33.00324249267578,
|
| 635 |
+
"logps/rejected": -32.683555603027344,
|
| 636 |
+
"loss": 0.4243,
|
| 637 |
+
"rewards/accuracies": 0.8374999761581421,
|
| 638 |
+
"rewards/chosen": 0.5806173086166382,
|
| 639 |
+
"rewards/margins": 0.8428903818130493,
|
| 640 |
+
"rewards/rejected": -0.26227301359176636,
|
| 641 |
+
"step": 400
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 1.04,
|
| 645 |
+
"eval_logits/chosen": -2.1981148719787598,
|
| 646 |
+
"eval_logits/rejected": -2.19327449798584,
|
| 647 |
+
"eval_logps/chosen": -34.17210388183594,
|
| 648 |
+
"eval_logps/rejected": -37.72929382324219,
|
| 649 |
+
"eval_loss": 0.6885781288146973,
|
| 650 |
+
"eval_rewards/accuracies": 0.5539867281913757,
|
| 651 |
+
"eval_rewards/chosen": -0.0962882786989212,
|
| 652 |
+
"eval_rewards/margins": 0.052584208548069,
|
| 653 |
+
"eval_rewards/rejected": -0.1488724797964096,
|
| 654 |
+
"eval_runtime": 145.4574,
|
| 655 |
+
"eval_samples_per_second": 2.358,
|
| 656 |
+
"eval_steps_per_second": 0.296,
|
| 657 |
+
"step": 400
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 1.06,
|
| 661 |
+
"grad_norm": 10.875,
|
| 662 |
+
"learning_rate": 4.590793060955158e-06,
|
| 663 |
+
"logits/chosen": -1.9296133518218994,
|
| 664 |
+
"logits/rejected": -1.9369118213653564,
|
| 665 |
+
"logps/chosen": -28.550052642822266,
|
| 666 |
+
"logps/rejected": -29.74081802368164,
|
| 667 |
+
"loss": 0.403,
|
| 668 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 669 |
+
"rewards/chosen": 0.3992263674736023,
|
| 670 |
+
"rewards/margins": 0.8490440249443054,
|
| 671 |
+
"rewards/rejected": -0.4498176574707031,
|
| 672 |
+
"step": 410
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 1.09,
|
| 676 |
+
"grad_norm": 9.8125,
|
| 677 |
+
"learning_rate": 4.559191453574582e-06,
|
| 678 |
+
"logits/chosen": -1.947829246520996,
|
| 679 |
+
"logits/rejected": -1.946836233139038,
|
| 680 |
+
"logps/chosen": -33.56596755981445,
|
| 681 |
+
"logps/rejected": -31.192113876342773,
|
| 682 |
+
"loss": 0.4976,
|
| 683 |
+
"rewards/accuracies": 0.762499988079071,
|
| 684 |
+
"rewards/chosen": 0.4793195128440857,
|
| 685 |
+
"rewards/margins": 0.6536680459976196,
|
| 686 |
+
"rewards/rejected": -0.17434856295585632,
|
| 687 |
+
"step": 420
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 1.12,
|
| 691 |
+
"grad_norm": 13.0625,
|
| 692 |
+
"learning_rate": 4.52653192962838e-06,
|
| 693 |
+
"logits/chosen": -1.9415998458862305,
|
| 694 |
+
"logits/rejected": -1.9243113994598389,
|
| 695 |
+
"logps/chosen": -30.399789810180664,
|
| 696 |
+
"logps/rejected": -33.614166259765625,
|
| 697 |
+
"loss": 0.4155,
|
| 698 |
+
"rewards/accuracies": 0.875,
|
| 699 |
+
"rewards/chosen": 0.40419188141822815,
|
| 700 |
+
"rewards/margins": 0.885710597038269,
|
| 701 |
+
"rewards/rejected": -0.4815187454223633,
|
| 702 |
+
"step": 430
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 1.14,
|
| 706 |
+
"grad_norm": 12.0625,
|
| 707 |
+
"learning_rate": 4.492831268057307e-06,
|
| 708 |
+
"logits/chosen": -1.972794771194458,
|
| 709 |
+
"logits/rejected": -1.9749000072479248,
|
| 710 |
+
"logps/chosen": -35.72783660888672,
|
| 711 |
+
"logps/rejected": -35.48394775390625,
|
| 712 |
+
"loss": 0.3567,
|
| 713 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 714 |
+
"rewards/chosen": 0.6191657185554504,
|
| 715 |
+
"rewards/margins": 1.0559546947479248,
|
| 716 |
+
"rewards/rejected": -0.43678897619247437,
|
| 717 |
+
"step": 440
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 1.17,
|
| 721 |
+
"grad_norm": 10.3125,
|
| 722 |
+
"learning_rate": 4.458106782690094e-06,
|
| 723 |
+
"logits/chosen": -2.047752857208252,
|
| 724 |
+
"logits/rejected": -2.0474939346313477,
|
| 725 |
+
"logps/chosen": -31.872684478759766,
|
| 726 |
+
"logps/rejected": -33.99079895019531,
|
| 727 |
+
"loss": 0.423,
|
| 728 |
+
"rewards/accuracies": 0.875,
|
| 729 |
+
"rewards/chosen": 0.5324660539627075,
|
| 730 |
+
"rewards/margins": 0.8339457511901855,
|
| 731 |
+
"rewards/rejected": -0.3014797270298004,
|
| 732 |
+
"step": 450
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 1.19,
|
| 736 |
+
"grad_norm": 11.1875,
|
| 737 |
+
"learning_rate": 4.422376313348405e-06,
|
| 738 |
+
"logits/chosen": -1.9906551837921143,
|
| 739 |
+
"logits/rejected": -1.9831295013427734,
|
| 740 |
+
"logps/chosen": -31.529306411743164,
|
| 741 |
+
"logps/rejected": -36.79833984375,
|
| 742 |
+
"loss": 0.3453,
|
| 743 |
+
"rewards/accuracies": 0.925000011920929,
|
| 744 |
+
"rewards/chosen": 0.6136585474014282,
|
| 745 |
+
"rewards/margins": 1.1346721649169922,
|
| 746 |
+
"rewards/rejected": -0.5210134983062744,
|
| 747 |
+
"step": 460
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 1.22,
|
| 751 |
+
"grad_norm": 14.25,
|
| 752 |
+
"learning_rate": 4.3856582166815696e-06,
|
| 753 |
+
"logits/chosen": -1.8940445184707642,
|
| 754 |
+
"logits/rejected": -1.8905950784683228,
|
| 755 |
+
"logps/chosen": -33.06376266479492,
|
| 756 |
+
"logps/rejected": -33.512367248535156,
|
| 757 |
+
"loss": 0.3856,
|
| 758 |
+
"rewards/accuracies": 0.887499988079071,
|
| 759 |
+
"rewards/chosen": 0.6079725623130798,
|
| 760 |
+
"rewards/margins": 1.0199116468429565,
|
| 761 |
+
"rewards/rejected": -0.4119391441345215,
|
| 762 |
+
"step": 470
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 1.25,
|
| 766 |
+
"grad_norm": 11.5625,
|
| 767 |
+
"learning_rate": 4.347971356735789e-06,
|
| 768 |
+
"logits/chosen": -2.0184760093688965,
|
| 769 |
+
"logits/rejected": -2.011504650115967,
|
| 770 |
+
"logps/chosen": -30.26607322692871,
|
| 771 |
+
"logps/rejected": -32.60536575317383,
|
| 772 |
+
"loss": 0.4501,
|
| 773 |
+
"rewards/accuracies": 0.800000011920929,
|
| 774 |
+
"rewards/chosen": 0.393722802400589,
|
| 775 |
+
"rewards/margins": 0.8220372200012207,
|
| 776 |
+
"rewards/rejected": -0.42831435799598694,
|
| 777 |
+
"step": 480
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 1.27,
|
| 781 |
+
"grad_norm": 8.8125,
|
| 782 |
+
"learning_rate": 4.309335095262675e-06,
|
| 783 |
+
"logits/chosen": -1.9625924825668335,
|
| 784 |
+
"logits/rejected": -1.9641788005828857,
|
| 785 |
+
"logps/chosen": -34.393211364746094,
|
| 786 |
+
"logps/rejected": -34.28179168701172,
|
| 787 |
+
"loss": 0.3543,
|
| 788 |
+
"rewards/accuracies": 0.875,
|
| 789 |
+
"rewards/chosen": 0.671985387802124,
|
| 790 |
+
"rewards/margins": 1.1201808452606201,
|
| 791 |
+
"rewards/rejected": -0.4481953978538513,
|
| 792 |
+
"step": 490
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 1.3,
|
| 796 |
+
"grad_norm": 16.5,
|
| 797 |
+
"learning_rate": 4.269769281772082e-06,
|
| 798 |
+
"logits/chosen": -1.8526058197021484,
|
| 799 |
+
"logits/rejected": -1.8502172231674194,
|
| 800 |
+
"logps/chosen": -32.287147521972656,
|
| 801 |
+
"logps/rejected": -37.31818771362305,
|
| 802 |
+
"loss": 0.353,
|
| 803 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 804 |
+
"rewards/chosen": 0.619600236415863,
|
| 805 |
+
"rewards/margins": 1.2173068523406982,
|
| 806 |
+
"rewards/rejected": -0.5977065563201904,
|
| 807 |
+
"step": 500
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 1.3,
|
| 811 |
+
"eval_logits/chosen": -2.1931960582733154,
|
| 812 |
+
"eval_logits/rejected": -2.1883604526519775,
|
| 813 |
+
"eval_logps/chosen": -34.31940460205078,
|
| 814 |
+
"eval_logps/rejected": -37.908485412597656,
|
| 815 |
+
"eval_loss": 0.6900750398635864,
|
| 816 |
+
"eval_rewards/accuracies": 0.5568937063217163,
|
| 817 |
+
"eval_rewards/chosen": -0.199394091963768,
|
| 818 |
+
"eval_rewards/margins": 0.07491057366132736,
|
| 819 |
+
"eval_rewards/rejected": -0.27430465817451477,
|
| 820 |
+
"eval_runtime": 145.2894,
|
| 821 |
+
"eval_samples_per_second": 2.361,
|
| 822 |
+
"eval_steps_per_second": 0.296,
|
| 823 |
+
"step": 500
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 1.32,
|
| 827 |
+
"grad_norm": 10.6875,
|
| 828 |
+
"learning_rate": 4.22929424333435e-06,
|
| 829 |
+
"logits/chosen": -1.9544782638549805,
|
| 830 |
+
"logits/rejected": -1.9592231512069702,
|
| 831 |
+
"logps/chosen": -32.51142883300781,
|
| 832 |
+
"logps/rejected": -32.14912414550781,
|
| 833 |
+
"loss": 0.3835,
|
| 834 |
+
"rewards/accuracies": 0.8374999761581421,
|
| 835 |
+
"rewards/chosen": 0.6621495485305786,
|
| 836 |
+
"rewards/margins": 1.0142595767974854,
|
| 837 |
+
"rewards/rejected": -0.3521100878715515,
|
| 838 |
+
"step": 510
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"epoch": 1.35,
|
| 842 |
+
"grad_norm": 11.25,
|
| 843 |
+
"learning_rate": 4.1879307741372085e-06,
|
| 844 |
+
"logits/chosen": -1.983420968055725,
|
| 845 |
+
"logits/rejected": -1.994410753250122,
|
| 846 |
+
"logps/chosen": -30.491313934326172,
|
| 847 |
+
"logps/rejected": -32.20922088623047,
|
| 848 |
+
"loss": 0.3741,
|
| 849 |
+
"rewards/accuracies": 0.887499988079071,
|
| 850 |
+
"rewards/chosen": 0.6293708682060242,
|
| 851 |
+
"rewards/margins": 1.120445728302002,
|
| 852 |
+
"rewards/rejected": -0.491074800491333,
|
| 853 |
+
"step": 520
|
| 854 |
+
},
|
| 855 |
+
{
|
| 856 |
+
"epoch": 1.38,
|
| 857 |
+
"grad_norm": 5.9375,
|
| 858 |
+
"learning_rate": 4.145700124802693e-06,
|
| 859 |
+
"logits/chosen": -1.9113273620605469,
|
| 860 |
+
"logits/rejected": -1.9080307483673096,
|
| 861 |
+
"logps/chosen": -31.719308853149414,
|
| 862 |
+
"logps/rejected": -33.058982849121094,
|
| 863 |
+
"loss": 0.3945,
|
| 864 |
+
"rewards/accuracies": 0.887499988079071,
|
| 865 |
+
"rewards/chosen": 0.49084219336509705,
|
| 866 |
+
"rewards/margins": 0.9781869053840637,
|
| 867 |
+
"rewards/rejected": -0.4873446822166443,
|
| 868 |
+
"step": 530
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"epoch": 1.4,
|
| 872 |
+
"grad_norm": 10.0,
|
| 873 |
+
"learning_rate": 4.102623991469562e-06,
|
| 874 |
+
"logits/chosen": -1.7767927646636963,
|
| 875 |
+
"logits/rejected": -1.7859961986541748,
|
| 876 |
+
"logps/chosen": -31.72823715209961,
|
| 877 |
+
"logps/rejected": -32.537315368652344,
|
| 878 |
+
"loss": 0.391,
|
| 879 |
+
"rewards/accuracies": 0.824999988079071,
|
| 880 |
+
"rewards/chosen": 0.5888075828552246,
|
| 881 |
+
"rewards/margins": 1.0657360553741455,
|
| 882 |
+
"rewards/rejected": -0.47692856192588806,
|
| 883 |
+
"step": 540
|
| 884 |
+
},
|
| 885 |
+
{
|
| 886 |
+
"epoch": 1.43,
|
| 887 |
+
"grad_norm": 11.125,
|
| 888 |
+
"learning_rate": 4.058724504646834e-06,
|
| 889 |
+
"logits/chosen": -1.8712069988250732,
|
| 890 |
+
"logits/rejected": -1.8649717569351196,
|
| 891 |
+
"logps/chosen": -32.81023025512695,
|
| 892 |
+
"logps/rejected": -31.534648895263672,
|
| 893 |
+
"loss": 0.4096,
|
| 894 |
+
"rewards/accuracies": 0.824999988079071,
|
| 895 |
+
"rewards/chosen": 0.6286527514457703,
|
| 896 |
+
"rewards/margins": 1.008962869644165,
|
| 897 |
+
"rewards/rejected": -0.38031005859375,
|
| 898 |
+
"step": 550
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 1.45,
|
| 902 |
+
"grad_norm": 8.4375,
|
| 903 |
+
"learning_rate": 4.014024217844167e-06,
|
| 904 |
+
"logits/chosen": -1.9593925476074219,
|
| 905 |
+
"logits/rejected": -1.9573974609375,
|
| 906 |
+
"logps/chosen": -33.5708122253418,
|
| 907 |
+
"logps/rejected": -32.04071807861328,
|
| 908 |
+
"loss": 0.3942,
|
| 909 |
+
"rewards/accuracies": 0.862500011920929,
|
| 910 |
+
"rewards/chosen": 0.700269877910614,
|
| 911 |
+
"rewards/margins": 1.0806026458740234,
|
| 912 |
+
"rewards/rejected": -0.380332887172699,
|
| 913 |
+
"step": 560
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 1.48,
|
| 917 |
+
"grad_norm": 10.625,
|
| 918 |
+
"learning_rate": 3.968546095984911e-06,
|
| 919 |
+
"logits/chosen": -1.7931559085845947,
|
| 920 |
+
"logits/rejected": -1.7910410165786743,
|
| 921 |
+
"logps/chosen": -31.88705825805664,
|
| 922 |
+
"logps/rejected": -31.514333724975586,
|
| 923 |
+
"loss": 0.4211,
|
| 924 |
+
"rewards/accuracies": 0.862500011920929,
|
| 925 |
+
"rewards/chosen": 0.6365472674369812,
|
| 926 |
+
"rewards/margins": 0.9949695467948914,
|
| 927 |
+
"rewards/rejected": -0.35842233896255493,
|
| 928 |
+
"step": 570
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 1.51,
|
| 932 |
+
"grad_norm": 13.3125,
|
| 933 |
+
"learning_rate": 3.922313503607806e-06,
|
| 934 |
+
"logits/chosen": -1.9281076192855835,
|
| 935 |
+
"logits/rejected": -1.92474365234375,
|
| 936 |
+
"logps/chosen": -30.16890525817871,
|
| 937 |
+
"logps/rejected": -35.45796203613281,
|
| 938 |
+
"loss": 0.3493,
|
| 939 |
+
"rewards/accuracies": 0.887499988079071,
|
| 940 |
+
"rewards/chosen": 0.6550183296203613,
|
| 941 |
+
"rewards/margins": 1.237276315689087,
|
| 942 |
+
"rewards/rejected": -0.582257866859436,
|
| 943 |
+
"step": 580
|
| 944 |
+
},
|
| 945 |
+
{
|
| 946 |
+
"epoch": 1.53,
|
| 947 |
+
"grad_norm": 8.25,
|
| 948 |
+
"learning_rate": 3.875350192863368e-06,
|
| 949 |
+
"logits/chosen": -1.865663766860962,
|
| 950 |
+
"logits/rejected": -1.8691942691802979,
|
| 951 |
+
"logps/chosen": -28.939590454101562,
|
| 952 |
+
"logps/rejected": -31.212047576904297,
|
| 953 |
+
"loss": 0.4162,
|
| 954 |
+
"rewards/accuracies": 0.862500011920929,
|
| 955 |
+
"rewards/chosen": 0.47280949354171753,
|
| 956 |
+
"rewards/margins": 0.86003577709198,
|
| 957 |
+
"rewards/rejected": -0.38722628355026245,
|
| 958 |
+
"step": 590
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"epoch": 1.56,
|
| 962 |
+
"grad_norm": 12.3125,
|
| 963 |
+
"learning_rate": 3.8276802913111436e-06,
|
| 964 |
+
"logits/chosen": -1.902017593383789,
|
| 965 |
+
"logits/rejected": -1.9019527435302734,
|
| 966 |
+
"logps/chosen": -31.125457763671875,
|
| 967 |
+
"logps/rejected": -31.8260555267334,
|
| 968 |
+
"loss": 0.3554,
|
| 969 |
+
"rewards/accuracies": 0.925000011920929,
|
| 970 |
+
"rewards/chosen": 0.7672521471977234,
|
| 971 |
+
"rewards/margins": 1.2103806734085083,
|
| 972 |
+
"rewards/rejected": -0.44312843680381775,
|
| 973 |
+
"step": 600
|
| 974 |
+
},
|
| 975 |
+
{
|
| 976 |
+
"epoch": 1.56,
|
| 977 |
+
"eval_logits/chosen": -2.174912452697754,
|
| 978 |
+
"eval_logits/rejected": -2.170095920562744,
|
| 979 |
+
"eval_logps/chosen": -34.24431228637695,
|
| 980 |
+
"eval_logps/rejected": -37.88401794433594,
|
| 981 |
+
"eval_loss": 0.6762823462486267,
|
| 982 |
+
"eval_rewards/accuracies": 0.5805647969245911,
|
| 983 |
+
"eval_rewards/chosen": -0.14683447778224945,
|
| 984 |
+
"eval_rewards/margins": 0.11034313589334488,
|
| 985 |
+
"eval_rewards/rejected": -0.2571776211261749,
|
| 986 |
+
"eval_runtime": 145.2157,
|
| 987 |
+
"eval_samples_per_second": 2.362,
|
| 988 |
+
"eval_steps_per_second": 0.296,
|
| 989 |
+
"step": 600
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 1.58,
|
| 993 |
+
"grad_norm": 11.9375,
|
| 994 |
+
"learning_rate": 3.7793282895240927e-06,
|
| 995 |
+
"logits/chosen": -1.9643291234970093,
|
| 996 |
+
"logits/rejected": -1.9650930166244507,
|
| 997 |
+
"logps/chosen": -33.864723205566406,
|
| 998 |
+
"logps/rejected": -33.84077072143555,
|
| 999 |
+
"loss": 0.3368,
|
| 1000 |
+
"rewards/accuracies": 0.8500000238418579,
|
| 1001 |
+
"rewards/chosen": 0.8157946467399597,
|
| 1002 |
+
"rewards/margins": 1.3454043865203857,
|
| 1003 |
+
"rewards/rejected": -0.5296097993850708,
|
| 1004 |
+
"step": 610
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 1.61,
|
| 1008 |
+
"grad_norm": 8.9375,
|
| 1009 |
+
"learning_rate": 3.730319028506478e-06,
|
| 1010 |
+
"logits/chosen": -1.9274613857269287,
|
| 1011 |
+
"logits/rejected": -1.9249248504638672,
|
| 1012 |
+
"logps/chosen": -32.089725494384766,
|
| 1013 |
+
"logps/rejected": -32.76671600341797,
|
| 1014 |
+
"loss": 0.3571,
|
| 1015 |
+
"rewards/accuracies": 0.875,
|
| 1016 |
+
"rewards/chosen": 0.7249698638916016,
|
| 1017 |
+
"rewards/margins": 1.2127305269241333,
|
| 1018 |
+
"rewards/rejected": -0.4877607822418213,
|
| 1019 |
+
"step": 620
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 1.64,
|
| 1023 |
+
"grad_norm": 22.75,
|
| 1024 |
+
"learning_rate": 3.6806776869317074e-06,
|
| 1025 |
+
"logits/chosen": -1.9384180307388306,
|
| 1026 |
+
"logits/rejected": -1.9293851852416992,
|
| 1027 |
+
"logps/chosen": -31.670080184936523,
|
| 1028 |
+
"logps/rejected": -31.639263153076172,
|
| 1029 |
+
"loss": 0.3787,
|
| 1030 |
+
"rewards/accuracies": 0.8500000238418579,
|
| 1031 |
+
"rewards/chosen": 0.7742272019386292,
|
| 1032 |
+
"rewards/margins": 1.2669562101364136,
|
| 1033 |
+
"rewards/rejected": -0.4927290380001068,
|
| 1034 |
+
"step": 630
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 1.66,
|
| 1038 |
+
"grad_norm": 15.5625,
|
| 1039 |
+
"learning_rate": 3.6304297682067146e-06,
|
| 1040 |
+
"logits/chosen": -1.9396206140518188,
|
| 1041 |
+
"logits/rejected": -1.9363600015640259,
|
| 1042 |
+
"logps/chosen": -31.236923217773438,
|
| 1043 |
+
"logps/rejected": -33.08216857910156,
|
| 1044 |
+
"loss": 0.3585,
|
| 1045 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1046 |
+
"rewards/chosen": 0.6730504631996155,
|
| 1047 |
+
"rewards/margins": 1.1369760036468506,
|
| 1048 |
+
"rewards/rejected": -0.4639255106449127,
|
| 1049 |
+
"step": 640
|
| 1050 |
+
},
|
| 1051 |
+
{
|
| 1052 |
+
"epoch": 1.69,
|
| 1053 |
+
"grad_norm": 9.5625,
|
| 1054 |
+
"learning_rate": 3.579601087369492e-06,
|
| 1055 |
+
"logits/chosen": -1.9486452341079712,
|
| 1056 |
+
"logits/rejected": -1.9509565830230713,
|
| 1057 |
+
"logps/chosen": -32.633331298828125,
|
| 1058 |
+
"logps/rejected": -34.409324645996094,
|
| 1059 |
+
"loss": 0.3199,
|
| 1060 |
+
"rewards/accuracies": 0.9375,
|
| 1061 |
+
"rewards/chosen": 0.7142627835273743,
|
| 1062 |
+
"rewards/margins": 1.2224125862121582,
|
| 1063 |
+
"rewards/rejected": -0.5081498026847839,
|
| 1064 |
+
"step": 650
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 1.71,
|
| 1068 |
+
"grad_norm": 15.875,
|
| 1069 |
+
"learning_rate": 3.5282177578265295e-06,
|
| 1070 |
+
"logits/chosen": -1.8591312170028687,
|
| 1071 |
+
"logits/rejected": -1.8595669269561768,
|
| 1072 |
+
"logps/chosen": -32.87069320678711,
|
| 1073 |
+
"logps/rejected": -32.203189849853516,
|
| 1074 |
+
"loss": 0.3567,
|
| 1075 |
+
"rewards/accuracies": 0.862500011920929,
|
| 1076 |
+
"rewards/chosen": 0.753296971321106,
|
| 1077 |
+
"rewards/margins": 1.1704500913619995,
|
| 1078 |
+
"rewards/rejected": -0.41715312004089355,
|
| 1079 |
+
"step": 660
|
| 1080 |
+
},
|
| 1081 |
+
{
|
| 1082 |
+
"epoch": 1.74,
|
| 1083 |
+
"grad_norm": 13.625,
|
| 1084 |
+
"learning_rate": 3.476306177936961e-06,
|
| 1085 |
+
"logits/chosen": -1.8990064859390259,
|
| 1086 |
+
"logits/rejected": -1.8893133401870728,
|
| 1087 |
+
"logps/chosen": -32.891273498535156,
|
| 1088 |
+
"logps/rejected": -33.074562072753906,
|
| 1089 |
+
"loss": 0.3066,
|
| 1090 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1091 |
+
"rewards/chosen": 0.7770041227340698,
|
| 1092 |
+
"rewards/margins": 1.31374192237854,
|
| 1093 |
+
"rewards/rejected": -0.5367377400398254,
|
| 1094 |
+
"step": 670
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 1.77,
|
| 1098 |
+
"grad_norm": 9.0625,
|
| 1099 |
+
"learning_rate": 3.423893017450324e-06,
|
| 1100 |
+
"logits/chosen": -1.7967946529388428,
|
| 1101 |
+
"logits/rejected": -1.7935699224472046,
|
| 1102 |
+
"logps/chosen": -30.264007568359375,
|
| 1103 |
+
"logps/rejected": -35.02149963378906,
|
| 1104 |
+
"loss": 0.3142,
|
| 1105 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1106 |
+
"rewards/chosen": 0.8378661870956421,
|
| 1107 |
+
"rewards/margins": 1.329910159111023,
|
| 1108 |
+
"rewards/rejected": -0.4920438826084137,
|
| 1109 |
+
"step": 680
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 1.79,
|
| 1113 |
+
"grad_norm": 7.8125,
|
| 1114 |
+
"learning_rate": 3.3710052038048794e-06,
|
| 1115 |
+
"logits/chosen": -1.85686457157135,
|
| 1116 |
+
"logits/rejected": -1.8561124801635742,
|
| 1117 |
+
"logps/chosen": -33.96955108642578,
|
| 1118 |
+
"logps/rejected": -36.205108642578125,
|
| 1119 |
+
"loss": 0.2775,
|
| 1120 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 1121 |
+
"rewards/chosen": 0.9105154275894165,
|
| 1122 |
+
"rewards/margins": 1.506414771080017,
|
| 1123 |
+
"rewards/rejected": -0.5958994626998901,
|
| 1124 |
+
"step": 690
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 1.82,
|
| 1128 |
+
"grad_norm": 6.90625,
|
| 1129 |
+
"learning_rate": 3.3176699082935546e-06,
|
| 1130 |
+
"logits/chosen": -1.831987738609314,
|
| 1131 |
+
"logits/rejected": -1.8347162008285522,
|
| 1132 |
+
"logps/chosen": -31.40140724182129,
|
| 1133 |
+
"logps/rejected": -36.432777404785156,
|
| 1134 |
+
"loss": 0.3333,
|
| 1135 |
+
"rewards/accuracies": 0.887499988079071,
|
| 1136 |
+
"rewards/chosen": 0.841008186340332,
|
| 1137 |
+
"rewards/margins": 1.394784688949585,
|
| 1138 |
+
"rewards/rejected": -0.5537765622138977,
|
| 1139 |
+
"step": 700
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"epoch": 1.82,
|
| 1143 |
+
"eval_logits/chosen": -2.1596450805664062,
|
| 1144 |
+
"eval_logits/rejected": -2.1548590660095215,
|
| 1145 |
+
"eval_logps/chosen": -34.294071197509766,
|
| 1146 |
+
"eval_logps/rejected": -37.937469482421875,
|
| 1147 |
+
"eval_loss": 0.6812623739242554,
|
| 1148 |
+
"eval_rewards/accuracies": 0.574335515499115,
|
| 1149 |
+
"eval_rewards/chosen": -0.18166212737560272,
|
| 1150 |
+
"eval_rewards/margins": 0.11293323338031769,
|
| 1151 |
+
"eval_rewards/rejected": -0.2945953607559204,
|
| 1152 |
+
"eval_runtime": 145.3544,
|
| 1153 |
+
"eval_samples_per_second": 2.36,
|
| 1154 |
+
"eval_steps_per_second": 0.296,
|
| 1155 |
+
"step": 700
|
| 1156 |
+
},
|
| 1157 |
+
{
|
| 1158 |
+
"epoch": 1.84,
|
| 1159 |
+
"grad_norm": 12.75,
|
| 1160 |
+
"learning_rate": 3.2639145321045933e-06,
|
| 1161 |
+
"logits/chosen": -1.9378350973129272,
|
| 1162 |
+
"logits/rejected": -1.9405092000961304,
|
| 1163 |
+
"logps/chosen": -33.763160705566406,
|
| 1164 |
+
"logps/rejected": -34.92665100097656,
|
| 1165 |
+
"loss": 0.3812,
|
| 1166 |
+
"rewards/accuracies": 0.8500000238418579,
|
| 1167 |
+
"rewards/chosen": 0.7927051782608032,
|
| 1168 |
+
"rewards/margins": 1.252252221107483,
|
| 1169 |
+
"rewards/rejected": -0.4595469832420349,
|
| 1170 |
+
"step": 710
|
| 1171 |
+
},
|
| 1172 |
+
{
|
| 1173 |
+
"epoch": 1.87,
|
| 1174 |
+
"grad_norm": 11.625,
|
| 1175 |
+
"learning_rate": 3.2097666922441107e-06,
|
| 1176 |
+
"logits/chosen": -1.7911384105682373,
|
| 1177 |
+
"logits/rejected": -1.7852294445037842,
|
| 1178 |
+
"logps/chosen": -33.655067443847656,
|
| 1179 |
+
"logps/rejected": -33.061012268066406,
|
| 1180 |
+
"loss": 0.3416,
|
| 1181 |
+
"rewards/accuracies": 0.875,
|
| 1182 |
+
"rewards/chosen": 0.8247106671333313,
|
| 1183 |
+
"rewards/margins": 1.3211290836334229,
|
| 1184 |
+
"rewards/rejected": -0.49641841650009155,
|
| 1185 |
+
"step": 720
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 1.9,
|
| 1189 |
+
"grad_norm": 7.40625,
|
| 1190 |
+
"learning_rate": 3.1552542073477554e-06,
|
| 1191 |
+
"logits/chosen": -1.9650695323944092,
|
| 1192 |
+
"logits/rejected": -1.9620001316070557,
|
| 1193 |
+
"logps/chosen": -29.55094337463379,
|
| 1194 |
+
"logps/rejected": -32.38628005981445,
|
| 1195 |
+
"loss": 0.302,
|
| 1196 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1197 |
+
"rewards/chosen": 0.7714294195175171,
|
| 1198 |
+
"rewards/margins": 1.428086519241333,
|
| 1199 |
+
"rewards/rejected": -0.6566572785377502,
|
| 1200 |
+
"step": 730
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 1.92,
|
| 1204 |
+
"grad_norm": 8.4375,
|
| 1205 |
+
"learning_rate": 3.100405083388799e-06,
|
| 1206 |
+
"logits/chosen": -1.8031257390975952,
|
| 1207 |
+
"logits/rejected": -1.8032245635986328,
|
| 1208 |
+
"logps/chosen": -32.12765121459961,
|
| 1209 |
+
"logps/rejected": -38.37054443359375,
|
| 1210 |
+
"loss": 0.3055,
|
| 1211 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1212 |
+
"rewards/chosen": 0.9348654747009277,
|
| 1213 |
+
"rewards/margins": 1.4547784328460693,
|
| 1214 |
+
"rewards/rejected": -0.5199130773544312,
|
| 1215 |
+
"step": 740
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 1.95,
|
| 1219 |
+
"grad_norm": 6.0625,
|
| 1220 |
+
"learning_rate": 3.0452474992899645e-06,
|
| 1221 |
+
"logits/chosen": -1.6882374286651611,
|
| 1222 |
+
"logits/rejected": -1.6935392618179321,
|
| 1223 |
+
"logps/chosen": -35.69883346557617,
|
| 1224 |
+
"logps/rejected": -34.775413513183594,
|
| 1225 |
+
"loss": 0.3555,
|
| 1226 |
+
"rewards/accuracies": 0.875,
|
| 1227 |
+
"rewards/chosen": 0.8635395765304565,
|
| 1228 |
+
"rewards/margins": 1.4256962537765503,
|
| 1229 |
+
"rewards/rejected": -0.5621567964553833,
|
| 1230 |
+
"step": 750
|
| 1231 |
+
},
|
| 1232 |
+
{
|
| 1233 |
+
"epoch": 1.97,
|
| 1234 |
+
"grad_norm": 7.9375,
|
| 1235 |
+
"learning_rate": 2.989809792446417e-06,
|
| 1236 |
+
"logits/chosen": -1.8880552053451538,
|
| 1237 |
+
"logits/rejected": -1.889639139175415,
|
| 1238 |
+
"logps/chosen": -31.452478408813477,
|
| 1239 |
+
"logps/rejected": -33.45410919189453,
|
| 1240 |
+
"loss": 0.3277,
|
| 1241 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1242 |
+
"rewards/chosen": 0.8072429895401001,
|
| 1243 |
+
"rewards/margins": 1.279667615890503,
|
| 1244 |
+
"rewards/rejected": -0.4724246561527252,
|
| 1245 |
+
"step": 760
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 2.0,
|
| 1249 |
+
"grad_norm": 12.4375,
|
| 1250 |
+
"learning_rate": 2.9341204441673267e-06,
|
| 1251 |
+
"logits/chosen": -1.8625637292861938,
|
| 1252 |
+
"logits/rejected": -1.8615506887435913,
|
| 1253 |
+
"logps/chosen": -30.956247329711914,
|
| 1254 |
+
"logps/rejected": -35.569908142089844,
|
| 1255 |
+
"loss": 0.3448,
|
| 1256 |
+
"rewards/accuracies": 0.92083340883255,
|
| 1257 |
+
"rewards/chosen": 0.8535542488098145,
|
| 1258 |
+
"rewards/margins": 1.2327837944030762,
|
| 1259 |
+
"rewards/rejected": -0.3792295455932617,
|
| 1260 |
+
"step": 770
|
| 1261 |
+
},
|
| 1262 |
+
{
|
| 1263 |
+
"epoch": 2.03,
|
| 1264 |
+
"grad_norm": 6.28125,
|
| 1265 |
+
"learning_rate": 2.878208065043501e-06,
|
| 1266 |
+
"logits/chosen": -1.8628917932510376,
|
| 1267 |
+
"logits/rejected": -1.8622173070907593,
|
| 1268 |
+
"logps/chosen": -33.29192352294922,
|
| 1269 |
+
"logps/rejected": -32.4744758605957,
|
| 1270 |
+
"loss": 0.2348,
|
| 1271 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1272 |
+
"rewards/chosen": 0.9495759010314941,
|
| 1273 |
+
"rewards/margins": 1.6880419254302979,
|
| 1274 |
+
"rewards/rejected": -0.7384659051895142,
|
| 1275 |
+
"step": 780
|
| 1276 |
+
},
|
| 1277 |
+
{
|
| 1278 |
+
"epoch": 2.05,
|
| 1279 |
+
"grad_norm": 7.5,
|
| 1280 |
+
"learning_rate": 2.8221013802485974e-06,
|
| 1281 |
+
"logits/chosen": -1.8925044536590576,
|
| 1282 |
+
"logits/rejected": -1.8910400867462158,
|
| 1283 |
+
"logps/chosen": -28.264022827148438,
|
| 1284 |
+
"logps/rejected": -33.81805419921875,
|
| 1285 |
+
"loss": 0.2096,
|
| 1286 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1287 |
+
"rewards/chosen": 0.9094095230102539,
|
| 1288 |
+
"rewards/margins": 1.7366279363632202,
|
| 1289 |
+
"rewards/rejected": -0.8272183537483215,
|
| 1290 |
+
"step": 790
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 2.08,
|
| 1294 |
+
"grad_norm": 4.46875,
|
| 1295 |
+
"learning_rate": 2.76582921478147e-06,
|
| 1296 |
+
"logits/chosen": -1.9434454441070557,
|
| 1297 |
+
"logits/rejected": -1.9400180578231812,
|
| 1298 |
+
"logps/chosen": -30.975006103515625,
|
| 1299 |
+
"logps/rejected": -35.40107345581055,
|
| 1300 |
+
"loss": 0.2025,
|
| 1301 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1302 |
+
"rewards/chosen": 1.016541838645935,
|
| 1303 |
+
"rewards/margins": 1.8579689264297485,
|
| 1304 |
+
"rewards/rejected": -0.8414268493652344,
|
| 1305 |
+
"step": 800
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 2.08,
|
| 1309 |
+
"eval_logits/chosen": -2.1460890769958496,
|
| 1310 |
+
"eval_logits/rejected": -2.1413440704345703,
|
| 1311 |
+
"eval_logps/chosen": -34.36545181274414,
|
| 1312 |
+
"eval_logps/rejected": -38.040489196777344,
|
| 1313 |
+
"eval_loss": 0.6800107359886169,
|
| 1314 |
+
"eval_rewards/accuracies": 0.5660299062728882,
|
| 1315 |
+
"eval_rewards/chosen": -0.23162823915481567,
|
| 1316 |
+
"eval_rewards/margins": 0.13507823646068573,
|
| 1317 |
+
"eval_rewards/rejected": -0.3667064905166626,
|
| 1318 |
+
"eval_runtime": 145.2997,
|
| 1319 |
+
"eval_samples_per_second": 2.361,
|
| 1320 |
+
"eval_steps_per_second": 0.296,
|
| 1321 |
+
"step": 800
|
| 1322 |
+
},
|
| 1323 |
+
{
|
| 1324 |
+
"epoch": 2.1,
|
| 1325 |
+
"grad_norm": 6.8125,
|
| 1326 |
+
"learning_rate": 2.7094204786572254e-06,
|
| 1327 |
+
"logits/chosen": -1.7730064392089844,
|
| 1328 |
+
"logits/rejected": -1.765316367149353,
|
| 1329 |
+
"logps/chosen": -32.71251678466797,
|
| 1330 |
+
"logps/rejected": -35.768611907958984,
|
| 1331 |
+
"loss": 0.1913,
|
| 1332 |
+
"rewards/accuracies": 0.9375,
|
| 1333 |
+
"rewards/chosen": 1.2206047773361206,
|
| 1334 |
+
"rewards/margins": 2.066551446914673,
|
| 1335 |
+
"rewards/rejected": -0.8459464907646179,
|
| 1336 |
+
"step": 810
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 2.13,
|
| 1340 |
+
"grad_norm": 8.125,
|
| 1341 |
+
"learning_rate": 2.6529041520546072e-06,
|
| 1342 |
+
"logits/chosen": -1.8485921621322632,
|
| 1343 |
+
"logits/rejected": -1.8590046167373657,
|
| 1344 |
+
"logps/chosen": -34.08565902709961,
|
| 1345 |
+
"logps/rejected": -33.43722915649414,
|
| 1346 |
+
"loss": 0.2332,
|
| 1347 |
+
"rewards/accuracies": 0.9375,
|
| 1348 |
+
"rewards/chosen": 1.1270233392715454,
|
| 1349 |
+
"rewards/margins": 1.8310768604278564,
|
| 1350 |
+
"rewards/rejected": -0.7040534615516663,
|
| 1351 |
+
"step": 820
|
| 1352 |
+
},
|
| 1353 |
+
{
|
| 1354 |
+
"epoch": 2.16,
|
| 1355 |
+
"grad_norm": 10.375,
|
| 1356 |
+
"learning_rate": 2.5963092704273302e-06,
|
| 1357 |
+
"logits/chosen": -1.897798776626587,
|
| 1358 |
+
"logits/rejected": -1.9024550914764404,
|
| 1359 |
+
"logps/chosen": -33.517730712890625,
|
| 1360 |
+
"logps/rejected": -30.249008178710938,
|
| 1361 |
+
"loss": 0.2494,
|
| 1362 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1363 |
+
"rewards/chosen": 0.9695348739624023,
|
| 1364 |
+
"rewards/margins": 1.6560699939727783,
|
| 1365 |
+
"rewards/rejected": -0.6865350008010864,
|
| 1366 |
+
"step": 830
|
| 1367 |
+
},
|
| 1368 |
+
{
|
| 1369 |
+
"epoch": 2.18,
|
| 1370 |
+
"grad_norm": 6.15625,
|
| 1371 |
+
"learning_rate": 2.53966490958702e-06,
|
| 1372 |
+
"logits/chosen": -1.8966840505599976,
|
| 1373 |
+
"logits/rejected": -1.9048032760620117,
|
| 1374 |
+
"logps/chosen": -33.26880645751953,
|
| 1375 |
+
"logps/rejected": -31.018238067626953,
|
| 1376 |
+
"loss": 0.2145,
|
| 1377 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1378 |
+
"rewards/chosen": 1.063457727432251,
|
| 1379 |
+
"rewards/margins": 1.8828351497650146,
|
| 1380 |
+
"rewards/rejected": -0.8193773031234741,
|
| 1381 |
+
"step": 840
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 2.21,
|
| 1385 |
+
"grad_norm": 9.5,
|
| 1386 |
+
"learning_rate": 2.4830001707654135e-06,
|
| 1387 |
+
"logits/chosen": -1.826229453086853,
|
| 1388 |
+
"logits/rejected": -1.8167400360107422,
|
| 1389 |
+
"logps/chosen": -30.364776611328125,
|
| 1390 |
+
"logps/rejected": -33.12417984008789,
|
| 1391 |
+
"loss": 0.227,
|
| 1392 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1393 |
+
"rewards/chosen": 0.9998935461044312,
|
| 1394 |
+
"rewards/margins": 1.777157187461853,
|
| 1395 |
+
"rewards/rejected": -0.7772639393806458,
|
| 1396 |
+
"step": 850
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 2.23,
|
| 1400 |
+
"grad_norm": 5.5,
|
| 1401 |
+
"learning_rate": 2.4263441656635054e-06,
|
| 1402 |
+
"logits/chosen": -1.9638983011245728,
|
| 1403 |
+
"logits/rejected": -1.954012155532837,
|
| 1404 |
+
"logps/chosen": -24.777915954589844,
|
| 1405 |
+
"logps/rejected": -31.053104400634766,
|
| 1406 |
+
"loss": 0.226,
|
| 1407 |
+
"rewards/accuracies": 0.9375,
|
| 1408 |
+
"rewards/chosen": 1.012310266494751,
|
| 1409 |
+
"rewards/margins": 1.8388469219207764,
|
| 1410 |
+
"rewards/rejected": -0.8265368342399597,
|
| 1411 |
+
"step": 860
|
| 1412 |
+
},
|
| 1413 |
+
{
|
| 1414 |
+
"epoch": 2.26,
|
| 1415 |
+
"grad_norm": 6.0625,
|
| 1416 |
+
"learning_rate": 2.3697260014953107e-06,
|
| 1417 |
+
"logits/chosen": -1.8167579174041748,
|
| 1418 |
+
"logits/rejected": -1.8181308507919312,
|
| 1419 |
+
"logps/chosen": -32.69607925415039,
|
| 1420 |
+
"logps/rejected": -31.112274169921875,
|
| 1421 |
+
"loss": 0.2008,
|
| 1422 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1423 |
+
"rewards/chosen": 1.1458479166030884,
|
| 1424 |
+
"rewards/margins": 1.9366874694824219,
|
| 1425 |
+
"rewards/rejected": -0.7908396124839783,
|
| 1426 |
+
"step": 870
|
| 1427 |
+
},
|
| 1428 |
+
{
|
| 1429 |
+
"epoch": 2.29,
|
| 1430 |
+
"grad_norm": 6.8125,
|
| 1431 |
+
"learning_rate": 2.3131747660339396e-06,
|
| 1432 |
+
"logits/chosen": -1.8225730657577515,
|
| 1433 |
+
"logits/rejected": -1.823504090309143,
|
| 1434 |
+
"logps/chosen": -31.072418212890625,
|
| 1435 |
+
"logps/rejected": -34.2416877746582,
|
| 1436 |
+
"loss": 0.263,
|
| 1437 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 1438 |
+
"rewards/chosen": 0.9217823147773743,
|
| 1439 |
+
"rewards/margins": 1.7553189992904663,
|
| 1440 |
+
"rewards/rejected": -0.8335369229316711,
|
| 1441 |
+
"step": 880
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 2.31,
|
| 1445 |
+
"grad_norm": 6.5625,
|
| 1446 |
+
"learning_rate": 2.256719512667651e-06,
|
| 1447 |
+
"logits/chosen": -1.7258434295654297,
|
| 1448 |
+
"logits/rejected": -1.7242809534072876,
|
| 1449 |
+
"logps/chosen": -34.12605667114258,
|
| 1450 |
+
"logps/rejected": -37.38144302368164,
|
| 1451 |
+
"loss": 0.2016,
|
| 1452 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1453 |
+
"rewards/chosen": 1.077201247215271,
|
| 1454 |
+
"rewards/margins": 2.130145311355591,
|
| 1455 |
+
"rewards/rejected": -1.0529439449310303,
|
| 1456 |
+
"step": 890
|
| 1457 |
+
},
|
| 1458 |
+
{
|
| 1459 |
+
"epoch": 2.34,
|
| 1460 |
+
"grad_norm": 6.1875,
|
| 1461 |
+
"learning_rate": 2.2003892454735786e-06,
|
| 1462 |
+
"logits/chosen": -1.8766018152236938,
|
| 1463 |
+
"logits/rejected": -1.8697988986968994,
|
| 1464 |
+
"logps/chosen": -30.475255966186523,
|
| 1465 |
+
"logps/rejected": -34.02910232543945,
|
| 1466 |
+
"loss": 0.2153,
|
| 1467 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1468 |
+
"rewards/chosen": 1.069166898727417,
|
| 1469 |
+
"rewards/margins": 1.9588909149169922,
|
| 1470 |
+
"rewards/rejected": -0.8897239565849304,
|
| 1471 |
+
"step": 900
|
| 1472 |
+
},
|
| 1473 |
+
{
|
| 1474 |
+
"epoch": 2.34,
|
| 1475 |
+
"eval_logits/chosen": -2.1339919567108154,
|
| 1476 |
+
"eval_logits/rejected": -2.129246950149536,
|
| 1477 |
+
"eval_logps/chosen": -34.38914108276367,
|
| 1478 |
+
"eval_logps/rejected": -38.06317138671875,
|
| 1479 |
+
"eval_loss": 0.6866059899330139,
|
| 1480 |
+
"eval_rewards/accuracies": 0.5834717750549316,
|
| 1481 |
+
"eval_rewards/chosen": -0.2482113242149353,
|
| 1482 |
+
"eval_rewards/margins": 0.13437683880329132,
|
| 1483 |
+
"eval_rewards/rejected": -0.38258814811706543,
|
| 1484 |
+
"eval_runtime": 145.4001,
|
| 1485 |
+
"eval_samples_per_second": 2.359,
|
| 1486 |
+
"eval_steps_per_second": 0.296,
|
| 1487 |
+
"step": 900
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 2.36,
|
| 1491 |
+
"grad_norm": 10.0,
|
| 1492 |
+
"learning_rate": 2.1442129043167877e-06,
|
| 1493 |
+
"logits/chosen": -1.913761854171753,
|
| 1494 |
+
"logits/rejected": -1.9091001749038696,
|
| 1495 |
+
"logps/chosen": -32.0454216003418,
|
| 1496 |
+
"logps/rejected": -35.850093841552734,
|
| 1497 |
+
"loss": 0.2031,
|
| 1498 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1499 |
+
"rewards/chosen": 1.0585777759552002,
|
| 1500 |
+
"rewards/margins": 1.9836070537567139,
|
| 1501 |
+
"rewards/rejected": -0.9250291585922241,
|
| 1502 |
+
"step": 910
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 2.39,
|
| 1506 |
+
"grad_norm": 6.8125,
|
| 1507 |
+
"learning_rate": 2.088219349982323e-06,
|
| 1508 |
+
"logits/chosen": -1.832114815711975,
|
| 1509 |
+
"logits/rejected": -1.8371769189834595,
|
| 1510 |
+
"logps/chosen": -33.69664764404297,
|
| 1511 |
+
"logps/rejected": -33.70267868041992,
|
| 1512 |
+
"loss": 0.2011,
|
| 1513 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1514 |
+
"rewards/chosen": 1.156354308128357,
|
| 1515 |
+
"rewards/margins": 1.9982963800430298,
|
| 1516 |
+
"rewards/rejected": -0.8419421315193176,
|
| 1517 |
+
"step": 920
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"epoch": 2.42,
|
| 1521 |
+
"grad_norm": 6.4375,
|
| 1522 |
+
"learning_rate": 2.0324373493478803e-06,
|
| 1523 |
+
"logits/chosen": -1.9312747716903687,
|
| 1524 |
+
"logits/rejected": -1.9220285415649414,
|
| 1525 |
+
"logps/chosen": -30.794708251953125,
|
| 1526 |
+
"logps/rejected": -35.72539520263672,
|
| 1527 |
+
"loss": 0.192,
|
| 1528 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1529 |
+
"rewards/chosen": 1.0431582927703857,
|
| 1530 |
+
"rewards/margins": 1.9409774541854858,
|
| 1531 |
+
"rewards/rejected": -0.8978190422058105,
|
| 1532 |
+
"step": 930
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 2.44,
|
| 1536 |
+
"grad_norm": 7.9375,
|
| 1537 |
+
"learning_rate": 1.976895560604729e-06,
|
| 1538 |
+
"logits/chosen": -1.8571876287460327,
|
| 1539 |
+
"logits/rejected": -1.8539724349975586,
|
| 1540 |
+
"logps/chosen": -30.002208709716797,
|
| 1541 |
+
"logps/rejected": -33.57603454589844,
|
| 1542 |
+
"loss": 0.2511,
|
| 1543 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1544 |
+
"rewards/chosen": 0.9003726840019226,
|
| 1545 |
+
"rewards/margins": 1.6778360605239868,
|
| 1546 |
+
"rewards/rejected": -0.7774633765220642,
|
| 1547 |
+
"step": 940
|
| 1548 |
+
},
|
| 1549 |
+
{
|
| 1550 |
+
"epoch": 2.47,
|
| 1551 |
+
"grad_norm": 4.3125,
|
| 1552 |
+
"learning_rate": 1.921622518534466e-06,
|
| 1553 |
+
"logits/chosen": -1.7950204610824585,
|
| 1554 |
+
"logits/rejected": -1.8024400472640991,
|
| 1555 |
+
"logps/chosen": -31.523014068603516,
|
| 1556 |
+
"logps/rejected": -36.79169464111328,
|
| 1557 |
+
"loss": 0.2492,
|
| 1558 |
+
"rewards/accuracies": 0.9375,
|
| 1559 |
+
"rewards/chosen": 0.9436267614364624,
|
| 1560 |
+
"rewards/margins": 1.812860131263733,
|
| 1561 |
+
"rewards/rejected": -0.8692331314086914,
|
| 1562 |
+
"step": 950
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 2.49,
|
| 1566 |
+
"grad_norm": 4.53125,
|
| 1567 |
+
"learning_rate": 1.8666466198491794e-06,
|
| 1568 |
+
"logits/chosen": -1.8054288625717163,
|
| 1569 |
+
"logits/rejected": -1.798541784286499,
|
| 1570 |
+
"logps/chosen": -32.300254821777344,
|
| 1571 |
+
"logps/rejected": -38.06441116333008,
|
| 1572 |
+
"loss": 0.1874,
|
| 1573 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1574 |
+
"rewards/chosen": 1.0697839260101318,
|
| 1575 |
+
"rewards/margins": 2.095707893371582,
|
| 1576 |
+
"rewards/rejected": -1.0259240865707397,
|
| 1577 |
+
"step": 960
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 2.52,
|
| 1581 |
+
"grad_norm": 6.0,
|
| 1582 |
+
"learning_rate": 1.8119961086025376e-06,
|
| 1583 |
+
"logits/chosen": -1.8083324432373047,
|
| 1584 |
+
"logits/rejected": -1.8088661432266235,
|
| 1585 |
+
"logps/chosen": -29.700159072875977,
|
| 1586 |
+
"logps/rejected": -34.11577606201172,
|
| 1587 |
+
"loss": 0.225,
|
| 1588 |
+
"rewards/accuracies": 0.9375,
|
| 1589 |
+
"rewards/chosen": 1.0753717422485352,
|
| 1590 |
+
"rewards/margins": 1.8695356845855713,
|
| 1591 |
+
"rewards/rejected": -0.7941640615463257,
|
| 1592 |
+
"step": 970
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 2.55,
|
| 1596 |
+
"grad_norm": 7.53125,
|
| 1597 |
+
"learning_rate": 1.7576990616793139e-06,
|
| 1598 |
+
"logits/chosen": -1.8105262517929077,
|
| 1599 |
+
"logits/rejected": -1.821751594543457,
|
| 1600 |
+
"logps/chosen": -31.371307373046875,
|
| 1601 |
+
"logps/rejected": -35.32282638549805,
|
| 1602 |
+
"loss": 0.2071,
|
| 1603 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1604 |
+
"rewards/chosen": 1.1937681436538696,
|
| 1605 |
+
"rewards/margins": 2.0278971195220947,
|
| 1606 |
+
"rewards/rejected": -0.8341289758682251,
|
| 1607 |
+
"step": 980
|
| 1608 |
+
},
|
| 1609 |
+
{
|
| 1610 |
+
"epoch": 2.57,
|
| 1611 |
+
"grad_norm": 6.1875,
|
| 1612 |
+
"learning_rate": 1.7037833743707892e-06,
|
| 1613 |
+
"logits/chosen": -1.8809820413589478,
|
| 1614 |
+
"logits/rejected": -1.8785299062728882,
|
| 1615 |
+
"logps/chosen": -34.623592376708984,
|
| 1616 |
+
"logps/rejected": -33.443870544433594,
|
| 1617 |
+
"loss": 0.2779,
|
| 1618 |
+
"rewards/accuracies": 0.887499988079071,
|
| 1619 |
+
"rewards/chosen": 1.0482189655303955,
|
| 1620 |
+
"rewards/margins": 1.779466986656189,
|
| 1621 |
+
"rewards/rejected": -0.7312478423118591,
|
| 1622 |
+
"step": 990
|
| 1623 |
+
},
|
| 1624 |
+
{
|
| 1625 |
+
"epoch": 2.6,
|
| 1626 |
+
"grad_norm": 8.3125,
|
| 1627 |
+
"learning_rate": 1.6502767460434588e-06,
|
| 1628 |
+
"logits/chosen": -1.8533103466033936,
|
| 1629 |
+
"logits/rejected": -1.8574621677398682,
|
| 1630 |
+
"logps/chosen": -32.96641540527344,
|
| 1631 |
+
"logps/rejected": -35.521705627441406,
|
| 1632 |
+
"loss": 0.2381,
|
| 1633 |
+
"rewards/accuracies": 0.9375,
|
| 1634 |
+
"rewards/chosen": 1.0478132963180542,
|
| 1635 |
+
"rewards/margins": 1.7727762460708618,
|
| 1636 |
+
"rewards/rejected": -0.7249631285667419,
|
| 1637 |
+
"step": 1000
|
| 1638 |
+
},
|
| 1639 |
+
{
|
| 1640 |
+
"epoch": 2.6,
|
| 1641 |
+
"eval_logits/chosen": -2.125450372695923,
|
| 1642 |
+
"eval_logits/rejected": -2.1207265853881836,
|
| 1643 |
+
"eval_logps/chosen": -34.40937423706055,
|
| 1644 |
+
"eval_logps/rejected": -38.111175537109375,
|
| 1645 |
+
"eval_loss": 0.6821461319923401,
|
| 1646 |
+
"eval_rewards/accuracies": 0.5863787531852722,
|
| 1647 |
+
"eval_rewards/chosen": -0.2623753547668457,
|
| 1648 |
+
"eval_rewards/margins": 0.15381217002868652,
|
| 1649 |
+
"eval_rewards/rejected": -0.41618749499320984,
|
| 1650 |
+
"eval_runtime": 145.3424,
|
| 1651 |
+
"eval_samples_per_second": 2.36,
|
| 1652 |
+
"eval_steps_per_second": 0.296,
|
| 1653 |
+
"step": 1000
|
| 1654 |
+
},
|
| 1655 |
+
{
|
| 1656 |
+
"epoch": 2.62,
|
| 1657 |
+
"grad_norm": 17.25,
|
| 1658 |
+
"learning_rate": 1.5972066659083796e-06,
|
| 1659 |
+
"logits/chosen": -1.8625215291976929,
|
| 1660 |
+
"logits/rejected": -1.8674328327178955,
|
| 1661 |
+
"logps/chosen": -31.423959732055664,
|
| 1662 |
+
"logps/rejected": -33.53556442260742,
|
| 1663 |
+
"loss": 0.2722,
|
| 1664 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 1665 |
+
"rewards/chosen": 0.9635237455368042,
|
| 1666 |
+
"rewards/margins": 1.6735379695892334,
|
| 1667 |
+
"rewards/rejected": -0.7100142240524292,
|
| 1668 |
+
"step": 1010
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 2.65,
|
| 1672 |
+
"grad_norm": 12.6875,
|
| 1673 |
+
"learning_rate": 1.5446003988985041e-06,
|
| 1674 |
+
"logits/chosen": -1.8214099407196045,
|
| 1675 |
+
"logits/rejected": -1.8238738775253296,
|
| 1676 |
+
"logps/chosen": -28.987659454345703,
|
| 1677 |
+
"logps/rejected": -32.98705291748047,
|
| 1678 |
+
"loss": 0.2111,
|
| 1679 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1680 |
+
"rewards/chosen": 1.0205867290496826,
|
| 1681 |
+
"rewards/margins": 1.922104835510254,
|
| 1682 |
+
"rewards/rejected": -0.9015182256698608,
|
| 1683 |
+
"step": 1020
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 2.68,
|
| 1687 |
+
"grad_norm": 4.59375,
|
| 1688 |
+
"learning_rate": 1.4924849716612211e-06,
|
| 1689 |
+
"logits/chosen": -1.8745782375335693,
|
| 1690 |
+
"logits/rejected": -1.869202971458435,
|
| 1691 |
+
"logps/chosen": -32.560997009277344,
|
| 1692 |
+
"logps/rejected": -34.77830505371094,
|
| 1693 |
+
"loss": 0.2251,
|
| 1694 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1695 |
+
"rewards/chosen": 0.9996814727783203,
|
| 1696 |
+
"rewards/margins": 1.881895661354065,
|
| 1697 |
+
"rewards/rejected": -0.8822140693664551,
|
| 1698 |
+
"step": 1030
|
| 1699 |
+
},
|
| 1700 |
+
{
|
| 1701 |
+
"epoch": 2.7,
|
| 1702 |
+
"grad_norm": 14.75,
|
| 1703 |
+
"learning_rate": 1.440887158673332e-06,
|
| 1704 |
+
"logits/chosen": -1.8110053539276123,
|
| 1705 |
+
"logits/rejected": -1.815009355545044,
|
| 1706 |
+
"logps/chosen": -34.393035888671875,
|
| 1707 |
+
"logps/rejected": -35.58574676513672,
|
| 1708 |
+
"loss": 0.2344,
|
| 1709 |
+
"rewards/accuracies": 0.9375,
|
| 1710 |
+
"rewards/chosen": 0.9524880647659302,
|
| 1711 |
+
"rewards/margins": 1.7688013315200806,
|
| 1712 |
+
"rewards/rejected": -0.8163133859634399,
|
| 1713 |
+
"step": 1040
|
| 1714 |
+
},
|
| 1715 |
+
{
|
| 1716 |
+
"epoch": 2.73,
|
| 1717 |
+
"grad_norm": 7.34375,
|
| 1718 |
+
"learning_rate": 1.3898334684855647e-06,
|
| 1719 |
+
"logits/chosen": -1.8082377910614014,
|
| 1720 |
+
"logits/rejected": -1.821096420288086,
|
| 1721 |
+
"logps/chosen": -30.48015785217285,
|
| 1722 |
+
"logps/rejected": -34.21984100341797,
|
| 1723 |
+
"loss": 0.2333,
|
| 1724 |
+
"rewards/accuracies": 0.9375,
|
| 1725 |
+
"rewards/chosen": 0.9679109454154968,
|
| 1726 |
+
"rewards/margins": 1.8077707290649414,
|
| 1727 |
+
"rewards/rejected": -0.8398596048355103,
|
| 1728 |
+
"step": 1050
|
| 1729 |
+
},
|
| 1730 |
+
{
|
| 1731 |
+
"epoch": 2.75,
|
| 1732 |
+
"grad_norm": 15.1875,
|
| 1733 |
+
"learning_rate": 1.3393501301037245e-06,
|
| 1734 |
+
"logits/chosen": -1.796105980873108,
|
| 1735 |
+
"logits/rejected": -1.789453148841858,
|
| 1736 |
+
"logps/chosen": -30.575389862060547,
|
| 1737 |
+
"logps/rejected": -34.41447830200195,
|
| 1738 |
+
"loss": 0.21,
|
| 1739 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1740 |
+
"rewards/chosen": 1.1087596416473389,
|
| 1741 |
+
"rewards/margins": 1.8763185739517212,
|
| 1742 |
+
"rewards/rejected": -0.7675586938858032,
|
| 1743 |
+
"step": 1060
|
| 1744 |
+
},
|
| 1745 |
+
{
|
| 1746 |
+
"epoch": 2.78,
|
| 1747 |
+
"grad_norm": 6.1875,
|
| 1748 |
+
"learning_rate": 1.2894630795134454e-06,
|
| 1749 |
+
"logits/chosen": -1.9221141338348389,
|
| 1750 |
+
"logits/rejected": -1.9228368997573853,
|
| 1751 |
+
"logps/chosen": -32.064552307128906,
|
| 1752 |
+
"logps/rejected": -34.02466583251953,
|
| 1753 |
+
"loss": 0.2118,
|
| 1754 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1755 |
+
"rewards/chosen": 1.0377662181854248,
|
| 1756 |
+
"rewards/margins": 1.8637428283691406,
|
| 1757 |
+
"rewards/rejected": -0.8259769678115845,
|
| 1758 |
+
"step": 1070
|
| 1759 |
+
},
|
| 1760 |
+
{
|
| 1761 |
+
"epoch": 2.81,
|
| 1762 |
+
"grad_norm": 9.0,
|
| 1763 |
+
"learning_rate": 1.2401979463554984e-06,
|
| 1764 |
+
"logits/chosen": -1.8529789447784424,
|
| 1765 |
+
"logits/rejected": -1.8518245220184326,
|
| 1766 |
+
"logps/chosen": -32.69489288330078,
|
| 1767 |
+
"logps/rejected": -34.210750579833984,
|
| 1768 |
+
"loss": 0.2684,
|
| 1769 |
+
"rewards/accuracies": 0.862500011920929,
|
| 1770 |
+
"rewards/chosen": 0.9764569997787476,
|
| 1771 |
+
"rewards/margins": 1.6896979808807373,
|
| 1772 |
+
"rewards/rejected": -0.7132409811019897,
|
| 1773 |
+
"step": 1080
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 2.83,
|
| 1777 |
+
"grad_norm": 5.53125,
|
| 1778 |
+
"learning_rate": 1.1915800407584705e-06,
|
| 1779 |
+
"logits/chosen": -1.8765695095062256,
|
| 1780 |
+
"logits/rejected": -1.8690354824066162,
|
| 1781 |
+
"logps/chosen": -32.69994354248047,
|
| 1782 |
+
"logps/rejected": -32.084373474121094,
|
| 1783 |
+
"loss": 0.2211,
|
| 1784 |
+
"rewards/accuracies": 0.9375,
|
| 1785 |
+
"rewards/chosen": 1.0681509971618652,
|
| 1786 |
+
"rewards/margins": 1.7981399297714233,
|
| 1787 |
+
"rewards/rejected": -0.7299890518188477,
|
| 1788 |
+
"step": 1090
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 2.86,
|
| 1792 |
+
"grad_norm": 5.0625,
|
| 1793 |
+
"learning_rate": 1.1436343403356019e-06,
|
| 1794 |
+
"logits/chosen": -1.8488590717315674,
|
| 1795 |
+
"logits/rejected": -1.8478235006332397,
|
| 1796 |
+
"logps/chosen": -33.83041763305664,
|
| 1797 |
+
"logps/rejected": -37.73679733276367,
|
| 1798 |
+
"loss": 0.1898,
|
| 1799 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1800 |
+
"rewards/chosen": 1.1115968227386475,
|
| 1801 |
+
"rewards/margins": 2.0804603099823,
|
| 1802 |
+
"rewards/rejected": -0.9688636660575867,
|
| 1803 |
+
"step": 1100
|
| 1804 |
+
},
|
| 1805 |
+
{
|
| 1806 |
+
"epoch": 2.86,
|
| 1807 |
+
"eval_logits/chosen": -2.123544454574585,
|
| 1808 |
+
"eval_logits/rejected": -2.118812084197998,
|
| 1809 |
+
"eval_logps/chosen": -34.41644287109375,
|
| 1810 |
+
"eval_logps/rejected": -38.11098098754883,
|
| 1811 |
+
"eval_loss": 0.6857725381851196,
|
| 1812 |
+
"eval_rewards/accuracies": 0.5830564498901367,
|
| 1813 |
+
"eval_rewards/chosen": -0.26732438802719116,
|
| 1814 |
+
"eval_rewards/margins": 0.14872850477695465,
|
| 1815 |
+
"eval_rewards/rejected": -0.4160528779029846,
|
| 1816 |
+
"eval_runtime": 145.3786,
|
| 1817 |
+
"eval_samples_per_second": 2.359,
|
| 1818 |
+
"eval_steps_per_second": 0.296,
|
| 1819 |
+
"step": 1100
|
| 1820 |
+
},
|
| 1821 |
+
{
|
| 1822 |
+
"epoch": 2.88,
|
| 1823 |
+
"grad_norm": 6.53125,
|
| 1824 |
+
"learning_rate": 1.0963854773524548e-06,
|
| 1825 |
+
"logits/chosen": -1.8964908123016357,
|
| 1826 |
+
"logits/rejected": -1.9019794464111328,
|
| 1827 |
+
"logps/chosen": -34.118003845214844,
|
| 1828 |
+
"logps/rejected": -36.824615478515625,
|
| 1829 |
+
"loss": 0.2423,
|
| 1830 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 1831 |
+
"rewards/chosen": 0.9443825483322144,
|
| 1832 |
+
"rewards/margins": 1.8411028385162354,
|
| 1833 |
+
"rewards/rejected": -0.8967200517654419,
|
| 1834 |
+
"step": 1110
|
| 1835 |
+
},
|
| 1836 |
+
{
|
| 1837 |
+
"epoch": 2.91,
|
| 1838 |
+
"grad_norm": 4.71875,
|
| 1839 |
+
"learning_rate": 1.049857726072005e-06,
|
| 1840 |
+
"logits/chosen": -1.8656179904937744,
|
| 1841 |
+
"logits/rejected": -1.86355721950531,
|
| 1842 |
+
"logps/chosen": -31.237468719482422,
|
| 1843 |
+
"logps/rejected": -34.223793029785156,
|
| 1844 |
+
"loss": 0.2295,
|
| 1845 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1846 |
+
"rewards/chosen": 0.9651158452033997,
|
| 1847 |
+
"rewards/margins": 1.8338441848754883,
|
| 1848 |
+
"rewards/rejected": -0.8687284588813782,
|
| 1849 |
+
"step": 1120
|
| 1850 |
+
},
|
| 1851 |
+
{
|
| 1852 |
+
"epoch": 2.94,
|
| 1853 |
+
"grad_norm": 5.375,
|
| 1854 |
+
"learning_rate": 1.0040749902836508e-06,
|
| 1855 |
+
"logits/chosen": -1.7683290243148804,
|
| 1856 |
+
"logits/rejected": -1.7710100412368774,
|
| 1857 |
+
"logps/chosen": -28.29164695739746,
|
| 1858 |
+
"logps/rejected": -31.67474937438965,
|
| 1859 |
+
"loss": 0.1929,
|
| 1860 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1861 |
+
"rewards/chosen": 1.0198917388916016,
|
| 1862 |
+
"rewards/margins": 2.0206894874572754,
|
| 1863 |
+
"rewards/rejected": -1.0007976293563843,
|
| 1864 |
+
"step": 1130
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 2.96,
|
| 1868 |
+
"grad_norm": 5.625,
|
| 1869 |
+
"learning_rate": 9.59060791022566e-07,
|
| 1870 |
+
"logits/chosen": -1.8653761148452759,
|
| 1871 |
+
"logits/rejected": -1.8621852397918701,
|
| 1872 |
+
"logps/chosen": -31.63905906677246,
|
| 1873 |
+
"logps/rejected": -33.515663146972656,
|
| 1874 |
+
"loss": 0.2347,
|
| 1875 |
+
"rewards/accuracies": 0.9375,
|
| 1876 |
+
"rewards/chosen": 1.095284342765808,
|
| 1877 |
+
"rewards/margins": 1.7874319553375244,
|
| 1878 |
+
"rewards/rejected": -0.6921476125717163,
|
| 1879 |
+
"step": 1140
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 2.99,
|
| 1883 |
+
"grad_norm": 8.5625,
|
| 1884 |
+
"learning_rate": 9.148382544856885e-07,
|
| 1885 |
+
"logits/chosen": -1.7846425771713257,
|
| 1886 |
+
"logits/rejected": -1.7783504724502563,
|
| 1887 |
+
"logps/chosen": -26.89322853088379,
|
| 1888 |
+
"logps/rejected": -33.4926643371582,
|
| 1889 |
+
"loss": 0.2239,
|
| 1890 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1891 |
+
"rewards/chosen": 0.8713330030441284,
|
| 1892 |
+
"rewards/margins": 1.8123865127563477,
|
| 1893 |
+
"rewards/rejected": -0.9410533905029297,
|
| 1894 |
+
"step": 1150
|
| 1895 |
+
},
|
| 1896 |
+
{
|
| 1897 |
+
"epoch": 3.01,
|
| 1898 |
+
"grad_norm": 5.375,
|
| 1899 |
+
"learning_rate": 8.714301001505568e-07,
|
| 1900 |
+
"logits/chosen": -1.8979135751724243,
|
| 1901 |
+
"logits/rejected": -1.8953993320465088,
|
| 1902 |
+
"logps/chosen": -31.6197452545166,
|
| 1903 |
+
"logps/rejected": -36.48524856567383,
|
| 1904 |
+
"loss": 0.2063,
|
| 1905 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1906 |
+
"rewards/chosen": 1.0334270000457764,
|
| 1907 |
+
"rewards/margins": 2.009028196334839,
|
| 1908 |
+
"rewards/rejected": -0.9756014943122864,
|
| 1909 |
+
"step": 1160
|
| 1910 |
+
},
|
| 1911 |
+
{
|
| 1912 |
+
"epoch": 3.04,
|
| 1913 |
+
"grad_norm": 4.9375,
|
| 1914 |
+
"learning_rate": 8.288586291031025e-07,
|
| 1915 |
+
"logits/chosen": -1.88443124294281,
|
| 1916 |
+
"logits/rejected": -1.8823401927947998,
|
| 1917 |
+
"logps/chosen": -30.127737045288086,
|
| 1918 |
+
"logps/rejected": -33.89713668823242,
|
| 1919 |
+
"loss": 0.172,
|
| 1920 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1921 |
+
"rewards/chosen": 1.0406690835952759,
|
| 1922 |
+
"rewards/margins": 2.1138501167297363,
|
| 1923 |
+
"rewards/rejected": -1.0731807947158813,
|
| 1924 |
+
"step": 1170
|
| 1925 |
+
},
|
| 1926 |
+
{
|
| 1927 |
+
"epoch": 3.06,
|
| 1928 |
+
"grad_norm": 5.25,
|
| 1929 |
+
"learning_rate": 7.871457125803897e-07,
|
| 1930 |
+
"logits/chosen": -1.8772131204605103,
|
| 1931 |
+
"logits/rejected": -1.865001916885376,
|
| 1932 |
+
"logps/chosen": -34.52227783203125,
|
| 1933 |
+
"logps/rejected": -35.57610321044922,
|
| 1934 |
+
"loss": 0.1854,
|
| 1935 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1936 |
+
"rewards/chosen": 1.0661174058914185,
|
| 1937 |
+
"rewards/margins": 2.119356870651245,
|
| 1938 |
+
"rewards/rejected": -1.0532395839691162,
|
| 1939 |
+
"step": 1180
|
| 1940 |
+
},
|
| 1941 |
+
{
|
| 1942 |
+
"epoch": 3.09,
|
| 1943 |
+
"grad_norm": 6.375,
|
| 1944 |
+
"learning_rate": 7.463127807341966e-07,
|
| 1945 |
+
"logits/chosen": -1.7719604969024658,
|
| 1946 |
+
"logits/rejected": -1.7666189670562744,
|
| 1947 |
+
"logps/chosen": -33.58086395263672,
|
| 1948 |
+
"logps/rejected": -35.4622802734375,
|
| 1949 |
+
"loss": 0.1812,
|
| 1950 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1951 |
+
"rewards/chosen": 1.2381782531738281,
|
| 1952 |
+
"rewards/margins": 2.0952985286712646,
|
| 1953 |
+
"rewards/rejected": -0.857120156288147,
|
| 1954 |
+
"step": 1190
|
| 1955 |
+
},
|
| 1956 |
+
{
|
| 1957 |
+
"epoch": 3.12,
|
| 1958 |
+
"grad_norm": 5.90625,
|
| 1959 |
+
"learning_rate": 7.063808116212021e-07,
|
| 1960 |
+
"logits/chosen": -1.8176746368408203,
|
| 1961 |
+
"logits/rejected": -1.8208776712417603,
|
| 1962 |
+
"logps/chosen": -31.216379165649414,
|
| 1963 |
+
"logps/rejected": -32.722049713134766,
|
| 1964 |
+
"loss": 0.2231,
|
| 1965 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1966 |
+
"rewards/chosen": 0.9791001081466675,
|
| 1967 |
+
"rewards/margins": 1.8442093133926392,
|
| 1968 |
+
"rewards/rejected": -0.8651090860366821,
|
| 1969 |
+
"step": 1200
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 3.12,
|
| 1973 |
+
"eval_logits/chosen": -2.122251272201538,
|
| 1974 |
+
"eval_logits/rejected": -2.11751651763916,
|
| 1975 |
+
"eval_logps/chosen": -34.40974426269531,
|
| 1976 |
+
"eval_logps/rejected": -38.12571716308594,
|
| 1977 |
+
"eval_loss": 0.6779673099517822,
|
| 1978 |
+
"eval_rewards/accuracies": 0.5888704061508179,
|
| 1979 |
+
"eval_rewards/chosen": -0.26263338327407837,
|
| 1980 |
+
"eval_rewards/margins": 0.1637314409017563,
|
| 1981 |
+
"eval_rewards/rejected": -0.42636486887931824,
|
| 1982 |
+
"eval_runtime": 145.3842,
|
| 1983 |
+
"eval_samples_per_second": 2.359,
|
| 1984 |
+
"eval_steps_per_second": 0.296,
|
| 1985 |
+
"step": 1200
|
| 1986 |
+
},
|
| 1987 |
+
{
|
| 1988 |
+
"epoch": 3.14,
|
| 1989 |
+
"grad_norm": 10.1875,
|
| 1990 |
+
"learning_rate": 6.673703204254348e-07,
|
| 1991 |
+
"logits/chosen": -1.8395841121673584,
|
| 1992 |
+
"logits/rejected": -1.8350778818130493,
|
| 1993 |
+
"logps/chosen": -29.8420352935791,
|
| 1994 |
+
"logps/rejected": -32.26592254638672,
|
| 1995 |
+
"loss": 0.1948,
|
| 1996 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1997 |
+
"rewards/chosen": 1.0021460056304932,
|
| 1998 |
+
"rewards/margins": 1.952625036239624,
|
| 1999 |
+
"rewards/rejected": -0.9504791498184204,
|
| 2000 |
+
"step": 1210
|
| 2001 |
+
},
|
| 2002 |
+
{
|
| 2003 |
+
"epoch": 3.17,
|
| 2004 |
+
"grad_norm": 5.90625,
|
| 2005 |
+
"learning_rate": 6.293013489185315e-07,
|
| 2006 |
+
"logits/chosen": -1.8544647693634033,
|
| 2007 |
+
"logits/rejected": -1.8489763736724854,
|
| 2008 |
+
"logps/chosen": -33.709232330322266,
|
| 2009 |
+
"logps/rejected": -36.026329040527344,
|
| 2010 |
+
"loss": 0.2003,
|
| 2011 |
+
"rewards/accuracies": 0.9375,
|
| 2012 |
+
"rewards/chosen": 1.0952959060668945,
|
| 2013 |
+
"rewards/margins": 2.114811420440674,
|
| 2014 |
+
"rewards/rejected": -1.0195155143737793,
|
| 2015 |
+
"step": 1220
|
| 2016 |
+
},
|
| 2017 |
+
{
|
| 2018 |
+
"epoch": 3.19,
|
| 2019 |
+
"grad_norm": 4.46875,
|
| 2020 |
+
"learning_rate": 5.921934551632086e-07,
|
| 2021 |
+
"logits/chosen": -1.8562214374542236,
|
| 2022 |
+
"logits/rejected": -1.8429136276245117,
|
| 2023 |
+
"logps/chosen": -31.71542739868164,
|
| 2024 |
+
"logps/rejected": -35.530879974365234,
|
| 2025 |
+
"loss": 0.2252,
|
| 2026 |
+
"rewards/accuracies": 0.875,
|
| 2027 |
+
"rewards/chosen": 0.9650102853775024,
|
| 2028 |
+
"rewards/margins": 1.9527349472045898,
|
| 2029 |
+
"rewards/rejected": -0.9877246618270874,
|
| 2030 |
+
"step": 1230
|
| 2031 |
+
},
|
| 2032 |
+
{
|
| 2033 |
+
"epoch": 3.22,
|
| 2034 |
+
"grad_norm": 4.34375,
|
| 2035 |
+
"learning_rate": 5.560657034652405e-07,
|
| 2036 |
+
"logits/chosen": -1.902470350265503,
|
| 2037 |
+
"logits/rejected": -1.8999955654144287,
|
| 2038 |
+
"logps/chosen": -33.41004180908203,
|
| 2039 |
+
"logps/rejected": -32.994361877441406,
|
| 2040 |
+
"loss": 0.198,
|
| 2041 |
+
"rewards/accuracies": 0.9375,
|
| 2042 |
+
"rewards/chosen": 1.1360009908676147,
|
| 2043 |
+
"rewards/margins": 1.9441492557525635,
|
| 2044 |
+
"rewards/rejected": -0.8081479072570801,
|
| 2045 |
+
"step": 1240
|
| 2046 |
+
},
|
| 2047 |
+
{
|
| 2048 |
+
"epoch": 3.25,
|
| 2049 |
+
"grad_norm": 5.3125,
|
| 2050 |
+
"learning_rate": 5.2093665457911e-07,
|
| 2051 |
+
"logits/chosen": -1.8268474340438843,
|
| 2052 |
+
"logits/rejected": -1.823965072631836,
|
| 2053 |
+
"logps/chosen": -32.80628967285156,
|
| 2054 |
+
"logps/rejected": -36.46531677246094,
|
| 2055 |
+
"loss": 0.1702,
|
| 2056 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 2057 |
+
"rewards/chosen": 1.1129032373428345,
|
| 2058 |
+
"rewards/margins": 2.2806363105773926,
|
| 2059 |
+
"rewards/rejected": -1.1677329540252686,
|
| 2060 |
+
"step": 1250
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 3.27,
|
| 2064 |
+
"grad_norm": 5.25,
|
| 2065 |
+
"learning_rate": 4.868243561723535e-07,
|
| 2066 |
+
"logits/chosen": -1.8951988220214844,
|
| 2067 |
+
"logits/rejected": -1.8899462223052979,
|
| 2068 |
+
"logps/chosen": -29.183536529541016,
|
| 2069 |
+
"logps/rejected": -33.69340133666992,
|
| 2070 |
+
"loss": 0.1818,
|
| 2071 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 2072 |
+
"rewards/chosen": 1.1040847301483154,
|
| 2073 |
+
"rewards/margins": 2.0216822624206543,
|
| 2074 |
+
"rewards/rejected": -0.9175974130630493,
|
| 2075 |
+
"step": 1260
|
| 2076 |
+
},
|
| 2077 |
+
{
|
| 2078 |
+
"epoch": 3.3,
|
| 2079 |
+
"grad_norm": 4.625,
|
| 2080 |
+
"learning_rate": 4.537463335535161e-07,
|
| 2081 |
+
"logits/chosen": -1.9220434427261353,
|
| 2082 |
+
"logits/rejected": -1.926945447921753,
|
| 2083 |
+
"logps/chosen": -31.057159423828125,
|
| 2084 |
+
"logps/rejected": -32.91054153442383,
|
| 2085 |
+
"loss": 0.184,
|
| 2086 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 2087 |
+
"rewards/chosen": 1.157687783241272,
|
| 2088 |
+
"rewards/margins": 2.01811146736145,
|
| 2089 |
+
"rewards/rejected": -0.8604235649108887,
|
| 2090 |
+
"step": 1270
|
| 2091 |
+
},
|
| 2092 |
+
{
|
| 2093 |
+
"epoch": 3.32,
|
| 2094 |
+
"grad_norm": 7.5,
|
| 2095 |
+
"learning_rate": 4.217195806684629e-07,
|
| 2096 |
+
"logits/chosen": -1.8148295879364014,
|
| 2097 |
+
"logits/rejected": -1.8222448825836182,
|
| 2098 |
+
"logps/chosen": -33.42390823364258,
|
| 2099 |
+
"logps/rejected": -33.99805450439453,
|
| 2100 |
+
"loss": 0.2006,
|
| 2101 |
+
"rewards/accuracies": 0.949999988079071,
|
| 2102 |
+
"rewards/chosen": 1.0383344888687134,
|
| 2103 |
+
"rewards/margins": 2.018185615539551,
|
| 2104 |
+
"rewards/rejected": -0.9798509478569031,
|
| 2105 |
+
"step": 1280
|
| 2106 |
+
},
|
| 2107 |
+
{
|
| 2108 |
+
"epoch": 3.35,
|
| 2109 |
+
"grad_norm": 4.84375,
|
| 2110 |
+
"learning_rate": 3.907605513696808e-07,
|
| 2111 |
+
"logits/chosen": -1.7262365818023682,
|
| 2112 |
+
"logits/rejected": -1.728548288345337,
|
| 2113 |
+
"logps/chosen": -31.478628158569336,
|
| 2114 |
+
"logps/rejected": -37.968849182128906,
|
| 2115 |
+
"loss": 0.1982,
|
| 2116 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 2117 |
+
"rewards/chosen": 1.1572868824005127,
|
| 2118 |
+
"rewards/margins": 2.147489547729492,
|
| 2119 |
+
"rewards/rejected": -0.9902026057243347,
|
| 2120 |
+
"step": 1290
|
| 2121 |
+
},
|
| 2122 |
+
{
|
| 2123 |
+
"epoch": 3.38,
|
| 2124 |
+
"grad_norm": 5.03125,
|
| 2125 |
+
"learning_rate": 3.6088515096305675e-07,
|
| 2126 |
+
"logits/chosen": -1.7753759622573853,
|
| 2127 |
+
"logits/rejected": -1.7788490056991577,
|
| 2128 |
+
"logps/chosen": -31.414981842041016,
|
| 2129 |
+
"logps/rejected": -33.79729461669922,
|
| 2130 |
+
"loss": 0.164,
|
| 2131 |
+
"rewards/accuracies": 0.987500011920929,
|
| 2132 |
+
"rewards/chosen": 1.2611814737319946,
|
| 2133 |
+
"rewards/margins": 2.1943717002868652,
|
| 2134 |
+
"rewards/rejected": -0.9331903457641602,
|
| 2135 |
+
"step": 1300
|
| 2136 |
+
},
|
| 2137 |
+
{
|
| 2138 |
+
"epoch": 3.38,
|
| 2139 |
+
"eval_logits/chosen": -2.1220953464508057,
|
| 2140 |
+
"eval_logits/rejected": -2.117370843887329,
|
| 2141 |
+
"eval_logps/chosen": -34.417152404785156,
|
| 2142 |
+
"eval_logps/rejected": -38.11579513549805,
|
| 2143 |
+
"eval_loss": 0.6833942532539368,
|
| 2144 |
+
"eval_rewards/accuracies": 0.594684362411499,
|
| 2145 |
+
"eval_rewards/chosen": -0.2678225636482239,
|
| 2146 |
+
"eval_rewards/margins": 0.15159977972507477,
|
| 2147 |
+
"eval_rewards/rejected": -0.41942232847213745,
|
| 2148 |
+
"eval_runtime": 145.412,
|
| 2149 |
+
"eval_samples_per_second": 2.359,
|
| 2150 |
+
"eval_steps_per_second": 0.296,
|
| 2151 |
+
"step": 1300
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 3.4,
|
| 2155 |
+
"grad_norm": 3.578125,
|
| 2156 |
+
"learning_rate": 3.321087280364757e-07,
|
| 2157 |
+
"logits/chosen": -1.88503098487854,
|
| 2158 |
+
"logits/rejected": -1.864648461341858,
|
| 2159 |
+
"logps/chosen": -29.50347328186035,
|
| 2160 |
+
"logps/rejected": -36.774391174316406,
|
| 2161 |
+
"loss": 0.1881,
|
| 2162 |
+
"rewards/accuracies": 0.9375,
|
| 2163 |
+
"rewards/chosen": 1.1269481182098389,
|
| 2164 |
+
"rewards/margins": 2.2525367736816406,
|
| 2165 |
+
"rewards/rejected": -1.1255886554718018,
|
| 2166 |
+
"step": 1310
|
| 2167 |
+
},
|
| 2168 |
+
{
|
| 2169 |
+
"epoch": 3.43,
|
| 2170 |
+
"grad_norm": 5.90625,
|
| 2171 |
+
"learning_rate": 3.044460665744284e-07,
|
| 2172 |
+
"logits/chosen": -1.8611704111099243,
|
| 2173 |
+
"logits/rejected": -1.866959571838379,
|
| 2174 |
+
"logps/chosen": -31.242517471313477,
|
| 2175 |
+
"logps/rejected": -33.74761199951172,
|
| 2176 |
+
"loss": 0.1726,
|
| 2177 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 2178 |
+
"rewards/chosen": 1.2538880109786987,
|
| 2179 |
+
"rewards/margins": 2.133692979812622,
|
| 2180 |
+
"rewards/rejected": -0.8798049688339233,
|
| 2181 |
+
"step": 1320
|
| 2182 |
+
},
|
| 2183 |
+
{
|
| 2184 |
+
"epoch": 3.45,
|
| 2185 |
+
"grad_norm": 4.21875,
|
| 2186 |
+
"learning_rate": 2.779113783626916e-07,
|
| 2187 |
+
"logits/chosen": -1.8369661569595337,
|
| 2188 |
+
"logits/rejected": -1.8319981098175049,
|
| 2189 |
+
"logps/chosen": -31.567821502685547,
|
| 2190 |
+
"logps/rejected": -36.19134521484375,
|
| 2191 |
+
"loss": 0.1725,
|
| 2192 |
+
"rewards/accuracies": 0.949999988079071,
|
| 2193 |
+
"rewards/chosen": 1.0182814598083496,
|
| 2194 |
+
"rewards/margins": 2.151616096496582,
|
| 2195 |
+
"rewards/rejected": -1.1333343982696533,
|
| 2196 |
+
"step": 1330
|
| 2197 |
+
},
|
| 2198 |
+
{
|
| 2199 |
+
"epoch": 3.48,
|
| 2200 |
+
"grad_norm": 4.9375,
|
| 2201 |
+
"learning_rate": 2.5251829568697204e-07,
|
| 2202 |
+
"logits/chosen": -1.6985762119293213,
|
| 2203 |
+
"logits/rejected": -1.708148717880249,
|
| 2204 |
+
"logps/chosen": -32.04993438720703,
|
| 2205 |
+
"logps/rejected": -32.1507453918457,
|
| 2206 |
+
"loss": 0.1954,
|
| 2207 |
+
"rewards/accuracies": 0.987500011920929,
|
| 2208 |
+
"rewards/chosen": 1.2247600555419922,
|
| 2209 |
+
"rewards/margins": 2.006398916244507,
|
| 2210 |
+
"rewards/rejected": -0.7816389799118042,
|
| 2211 |
+
"step": 1340
|
| 2212 |
+
},
|
| 2213 |
+
{
|
| 2214 |
+
"epoch": 3.51,
|
| 2215 |
+
"grad_norm": 5.78125,
|
| 2216 |
+
"learning_rate": 2.2827986432927774e-07,
|
| 2217 |
+
"logits/chosen": -1.756463646888733,
|
| 2218 |
+
"logits/rejected": -1.7611831426620483,
|
| 2219 |
+
"logps/chosen": -32.128456115722656,
|
| 2220 |
+
"logps/rejected": -34.41259002685547,
|
| 2221 |
+
"loss": 0.1564,
|
| 2222 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 2223 |
+
"rewards/chosen": 1.27573561668396,
|
| 2224 |
+
"rewards/margins": 2.1211793422698975,
|
| 2225 |
+
"rewards/rejected": -0.8454439043998718,
|
| 2226 |
+
"step": 1350
|
| 2227 |
+
},
|
| 2228 |
+
{
|
| 2229 |
+
"epoch": 3.53,
|
| 2230 |
+
"grad_norm": 18.5,
|
| 2231 |
+
"learning_rate": 2.0520853686560177e-07,
|
| 2232 |
+
"logits/chosen": -1.7851018905639648,
|
| 2233 |
+
"logits/rejected": -1.7904930114746094,
|
| 2234 |
+
"logps/chosen": -32.26166534423828,
|
| 2235 |
+
"logps/rejected": -35.997657775878906,
|
| 2236 |
+
"loss": 0.2004,
|
| 2237 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 2238 |
+
"rewards/chosen": 1.098996877670288,
|
| 2239 |
+
"rewards/margins": 2.1032118797302246,
|
| 2240 |
+
"rewards/rejected": -1.0042150020599365,
|
| 2241 |
+
"step": 1360
|
| 2242 |
+
},
|
| 2243 |
+
{
|
| 2244 |
+
"epoch": 3.56,
|
| 2245 |
+
"grad_norm": 3.8125,
|
| 2246 |
+
"learning_rate": 1.833161662683672e-07,
|
| 2247 |
+
"logits/chosen": -1.8785173892974854,
|
| 2248 |
+
"logits/rejected": -1.8740142583847046,
|
| 2249 |
+
"logps/chosen": -31.18601417541504,
|
| 2250 |
+
"logps/rejected": -32.262332916259766,
|
| 2251 |
+
"loss": 0.1895,
|
| 2252 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 2253 |
+
"rewards/chosen": 1.014768362045288,
|
| 2254 |
+
"rewards/margins": 1.9600728750228882,
|
| 2255 |
+
"rewards/rejected": -0.9453045129776001,
|
| 2256 |
+
"step": 1370
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 3.58,
|
| 2260 |
+
"grad_norm": 6.28125,
|
| 2261 |
+
"learning_rate": 1.626139998169246e-07,
|
| 2262 |
+
"logits/chosen": -1.7922194004058838,
|
| 2263 |
+
"logits/rejected": -1.7942945957183838,
|
| 2264 |
+
"logps/chosen": -28.180294036865234,
|
| 2265 |
+
"logps/rejected": -32.01648712158203,
|
| 2266 |
+
"loss": 0.2023,
|
| 2267 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 2268 |
+
"rewards/chosen": 1.0353037118911743,
|
| 2269 |
+
"rewards/margins": 1.9259191751480103,
|
| 2270 |
+
"rewards/rejected": -0.8906152844429016,
|
| 2271 |
+
"step": 1380
|
| 2272 |
+
},
|
| 2273 |
+
{
|
| 2274 |
+
"epoch": 3.61,
|
| 2275 |
+
"grad_norm": 7.4375,
|
| 2276 |
+
"learning_rate": 1.4311267331922535e-07,
|
| 2277 |
+
"logits/chosen": -1.7903833389282227,
|
| 2278 |
+
"logits/rejected": -1.7950242757797241,
|
| 2279 |
+
"logps/chosen": -30.506296157836914,
|
| 2280 |
+
"logps/rejected": -33.995018005371094,
|
| 2281 |
+
"loss": 0.2087,
|
| 2282 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 2283 |
+
"rewards/chosen": 1.0592371225357056,
|
| 2284 |
+
"rewards/margins": 1.7641029357910156,
|
| 2285 |
+
"rewards/rejected": -0.7048658132553101,
|
| 2286 |
+
"step": 1390
|
| 2287 |
+
},
|
| 2288 |
+
{
|
| 2289 |
+
"epoch": 3.64,
|
| 2290 |
+
"grad_norm": 4.0625,
|
| 2291 |
+
"learning_rate": 1.2482220564763669e-07,
|
| 2292 |
+
"logits/chosen": -1.8554019927978516,
|
| 2293 |
+
"logits/rejected": -1.8586347103118896,
|
| 2294 |
+
"logps/chosen": -33.235443115234375,
|
| 2295 |
+
"logps/rejected": -34.80227279663086,
|
| 2296 |
+
"loss": 0.1562,
|
| 2297 |
+
"rewards/accuracies": 0.987500011920929,
|
| 2298 |
+
"rewards/chosen": 1.1865932941436768,
|
| 2299 |
+
"rewards/margins": 2.1516919136047363,
|
| 2300 |
+
"rewards/rejected": -0.9650983810424805,
|
| 2301 |
+
"step": 1400
|
| 2302 |
+
},
|
| 2303 |
+
{
|
| 2304 |
+
"epoch": 3.64,
|
| 2305 |
+
"eval_logits/chosen": -2.121932029724121,
|
| 2306 |
+
"eval_logits/rejected": -2.117196798324585,
|
| 2307 |
+
"eval_logps/chosen": -34.41535568237305,
|
| 2308 |
+
"eval_logps/rejected": -38.13964080810547,
|
| 2309 |
+
"eval_loss": 0.6752650141716003,
|
| 2310 |
+
"eval_rewards/accuracies": 0.5921927094459534,
|
| 2311 |
+
"eval_rewards/chosen": -0.2665621340274811,
|
| 2312 |
+
"eval_rewards/margins": 0.16955220699310303,
|
| 2313 |
+
"eval_rewards/rejected": -0.4361143708229065,
|
| 2314 |
+
"eval_runtime": 145.4569,
|
| 2315 |
+
"eval_samples_per_second": 2.358,
|
| 2316 |
+
"eval_steps_per_second": 0.296,
|
| 2317 |
+
"step": 1400
|
| 2318 |
+
},
|
| 2319 |
+
{
|
| 2320 |
+
"epoch": 3.66,
|
| 2321 |
+
"grad_norm": 6.46875,
|
| 2322 |
+
"learning_rate": 1.0775199359171346e-07,
|
| 2323 |
+
"logits/chosen": -1.9231650829315186,
|
| 2324 |
+
"logits/rejected": -1.9162681102752686,
|
| 2325 |
+
"logps/chosen": -32.11163330078125,
|
| 2326 |
+
"logps/rejected": -35.49402618408203,
|
| 2327 |
+
"loss": 0.186,
|
| 2328 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 2329 |
+
"rewards/chosen": 1.1836400032043457,
|
| 2330 |
+
"rewards/margins": 2.0582594871520996,
|
| 2331 |
+
"rewards/rejected": -0.8746196627616882,
|
| 2332 |
+
"step": 1410
|
| 2333 |
+
},
|
| 2334 |
+
{
|
| 2335 |
+
"epoch": 3.69,
|
| 2336 |
+
"grad_norm": 8.6875,
|
| 2337 |
+
"learning_rate": 9.191080703056604e-08,
|
| 2338 |
+
"logits/chosen": -1.8132683038711548,
|
| 2339 |
+
"logits/rejected": -1.8242778778076172,
|
| 2340 |
+
"logps/chosen": -32.496726989746094,
|
| 2341 |
+
"logps/rejected": -34.891380310058594,
|
| 2342 |
+
"loss": 0.175,
|
| 2343 |
+
"rewards/accuracies": 0.987500011920929,
|
| 2344 |
+
"rewards/chosen": 1.156802773475647,
|
| 2345 |
+
"rewards/margins": 2.067012071609497,
|
| 2346 |
+
"rewards/rejected": -0.9102091789245605,
|
| 2347 |
+
"step": 1420
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 3.71,
|
| 2351 |
+
"grad_norm": 7.53125,
|
| 2352 |
+
"learning_rate": 7.730678442730539e-08,
|
| 2353 |
+
"logits/chosen": -1.8921226263046265,
|
| 2354 |
+
"logits/rejected": -1.9039920568466187,
|
| 2355 |
+
"logps/chosen": -33.318809509277344,
|
| 2356 |
+
"logps/rejected": -35.047691345214844,
|
| 2357 |
+
"loss": 0.1674,
|
| 2358 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 2359 |
+
"rewards/chosen": 1.219396948814392,
|
| 2360 |
+
"rewards/margins": 2.2059476375579834,
|
| 2361 |
+
"rewards/rejected": -0.9865506887435913,
|
| 2362 |
+
"step": 1430
|
| 2363 |
+
},
|
| 2364 |
+
{
|
| 2365 |
+
"epoch": 3.74,
|
| 2366 |
+
"grad_norm": 7.875,
|
| 2367 |
+
"learning_rate": 6.394742864787806e-08,
|
| 2368 |
+
"logits/chosen": -1.8827718496322632,
|
| 2369 |
+
"logits/rejected": -1.8849990367889404,
|
| 2370 |
+
"logps/chosen": -31.633159637451172,
|
| 2371 |
+
"logps/rejected": -35.23392868041992,
|
| 2372 |
+
"loss": 0.2007,
|
| 2373 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 2374 |
+
"rewards/chosen": 1.1026588678359985,
|
| 2375 |
+
"rewards/margins": 2.0644314289093018,
|
| 2376 |
+
"rewards/rejected": -0.9617727398872375,
|
| 2377 |
+
"step": 1440
|
| 2378 |
+
},
|
| 2379 |
+
{
|
| 2380 |
+
"epoch": 3.77,
|
| 2381 |
+
"grad_norm": 5.1875,
|
| 2382 |
+
"learning_rate": 5.183960310644748e-08,
|
| 2383 |
+
"logits/chosen": -1.8658161163330078,
|
| 2384 |
+
"logits/rejected": -1.8587331771850586,
|
| 2385 |
+
"logps/chosen": -33.893577575683594,
|
| 2386 |
+
"logps/rejected": -35.00482940673828,
|
| 2387 |
+
"loss": 0.2008,
|
| 2388 |
+
"rewards/accuracies": 0.949999988079071,
|
| 2389 |
+
"rewards/chosen": 1.1723241806030273,
|
| 2390 |
+
"rewards/margins": 2.036421298980713,
|
| 2391 |
+
"rewards/rejected": -0.864096999168396,
|
| 2392 |
+
"step": 1450
|
| 2393 |
+
},
|
| 2394 |
+
{
|
| 2395 |
+
"epoch": 3.79,
|
| 2396 |
+
"grad_norm": 5.0625,
|
| 2397 |
+
"learning_rate": 4.098952823928693e-08,
|
| 2398 |
+
"logits/chosen": -1.833505630493164,
|
| 2399 |
+
"logits/rejected": -1.8394966125488281,
|
| 2400 |
+
"logps/chosen": -29.668813705444336,
|
| 2401 |
+
"logps/rejected": -35.166725158691406,
|
| 2402 |
+
"loss": 0.1749,
|
| 2403 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 2404 |
+
"rewards/chosen": 1.0810542106628418,
|
| 2405 |
+
"rewards/margins": 2.1287617683410645,
|
| 2406 |
+
"rewards/rejected": -1.0477073192596436,
|
| 2407 |
+
"step": 1460
|
| 2408 |
+
},
|
| 2409 |
+
{
|
| 2410 |
+
"epoch": 3.82,
|
| 2411 |
+
"grad_norm": 4.90625,
|
| 2412 |
+
"learning_rate": 3.1402778309014284e-08,
|
| 2413 |
+
"logits/chosen": -1.7813646793365479,
|
| 2414 |
+
"logits/rejected": -1.779421091079712,
|
| 2415 |
+
"logps/chosen": -29.22164535522461,
|
| 2416 |
+
"logps/rejected": -31.763259887695312,
|
| 2417 |
+
"loss": 0.2002,
|
| 2418 |
+
"rewards/accuracies": 0.949999988079071,
|
| 2419 |
+
"rewards/chosen": 1.136735439300537,
|
| 2420 |
+
"rewards/margins": 1.9550540447235107,
|
| 2421 |
+
"rewards/rejected": -0.8183186650276184,
|
| 2422 |
+
"step": 1470
|
| 2423 |
+
},
|
| 2424 |
+
{
|
| 2425 |
+
"epoch": 3.84,
|
| 2426 |
+
"grad_norm": 6.53125,
|
| 2427 |
+
"learning_rate": 2.3084278540791427e-08,
|
| 2428 |
+
"logits/chosen": -1.956352949142456,
|
| 2429 |
+
"logits/rejected": -1.9507420063018799,
|
| 2430 |
+
"logps/chosen": -33.838050842285156,
|
| 2431 |
+
"logps/rejected": -33.383323669433594,
|
| 2432 |
+
"loss": 0.1868,
|
| 2433 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 2434 |
+
"rewards/chosen": 1.013047456741333,
|
| 2435 |
+
"rewards/margins": 1.941677451133728,
|
| 2436 |
+
"rewards/rejected": -0.928629994392395,
|
| 2437 |
+
"step": 1480
|
| 2438 |
+
},
|
| 2439 |
+
{
|
| 2440 |
+
"epoch": 3.87,
|
| 2441 |
+
"grad_norm": 7.21875,
|
| 2442 |
+
"learning_rate": 1.6038302591975807e-08,
|
| 2443 |
+
"logits/chosen": -1.8494259119033813,
|
| 2444 |
+
"logits/rejected": -1.8517593145370483,
|
| 2445 |
+
"logps/chosen": -27.075820922851562,
|
| 2446 |
+
"logps/rejected": -29.148422241210938,
|
| 2447 |
+
"loss": 0.2343,
|
| 2448 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 2449 |
+
"rewards/chosen": 0.980392336845398,
|
| 2450 |
+
"rewards/margins": 1.768689513206482,
|
| 2451 |
+
"rewards/rejected": -0.788297176361084,
|
| 2452 |
+
"step": 1490
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 3.9,
|
| 2456 |
+
"grad_norm": 6.15625,
|
| 2457 |
+
"learning_rate": 1.0268470356514237e-08,
|
| 2458 |
+
"logits/chosen": -1.8568594455718994,
|
| 2459 |
+
"logits/rejected": -1.8512837886810303,
|
| 2460 |
+
"logps/chosen": -31.454471588134766,
|
| 2461 |
+
"logps/rejected": -33.362735748291016,
|
| 2462 |
+
"loss": 0.2163,
|
| 2463 |
+
"rewards/accuracies": 0.9375,
|
| 2464 |
+
"rewards/chosen": 0.9424249529838562,
|
| 2465 |
+
"rewards/margins": 1.918071985244751,
|
| 2466 |
+
"rewards/rejected": -0.97564697265625,
|
| 2467 |
+
"step": 1500
|
| 2468 |
+
},
|
| 2469 |
+
{
|
| 2470 |
+
"epoch": 3.9,
|
| 2471 |
+
"eval_logits/chosen": -2.121995687484741,
|
| 2472 |
+
"eval_logits/rejected": -2.1172635555267334,
|
| 2473 |
+
"eval_logps/chosen": -34.41799545288086,
|
| 2474 |
+
"eval_logps/rejected": -38.11918640136719,
|
| 2475 |
+
"eval_loss": 0.6830941438674927,
|
| 2476 |
+
"eval_rewards/accuracies": 0.5801494717597961,
|
| 2477 |
+
"eval_rewards/chosen": -0.26841017603874207,
|
| 2478 |
+
"eval_rewards/margins": 0.15338675677776337,
|
| 2479 |
+
"eval_rewards/rejected": -0.42179691791534424,
|
| 2480 |
+
"eval_runtime": 145.2579,
|
| 2481 |
+
"eval_samples_per_second": 2.361,
|
| 2482 |
+
"eval_steps_per_second": 0.296,
|
| 2483 |
+
"step": 1500
|
| 2484 |
+
},
|
| 2485 |
+
{
|
| 2486 |
+
"epoch": 3.92,
|
| 2487 |
+
"grad_norm": 8.625,
|
| 2488 |
+
"learning_rate": 5.777746105209147e-09,
|
| 2489 |
+
"logits/chosen": -1.7834112644195557,
|
| 2490 |
+
"logits/rejected": -1.7875878810882568,
|
| 2491 |
+
"logps/chosen": -32.74296951293945,
|
| 2492 |
+
"logps/rejected": -35.563575744628906,
|
| 2493 |
+
"loss": 0.2238,
|
| 2494 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 2495 |
+
"rewards/chosen": 0.9996176958084106,
|
| 2496 |
+
"rewards/margins": 1.8068501949310303,
|
| 2497 |
+
"rewards/rejected": -0.807232677936554,
|
| 2498 |
+
"step": 1510
|
| 2499 |
+
},
|
| 2500 |
+
{
|
| 2501 |
+
"epoch": 3.95,
|
| 2502 |
+
"grad_norm": 4.625,
|
| 2503 |
+
"learning_rate": 2.5684369628148352e-09,
|
| 2504 |
+
"logits/chosen": -1.8380959033966064,
|
| 2505 |
+
"logits/rejected": -1.8364956378936768,
|
| 2506 |
+
"logps/chosen": -29.186996459960938,
|
| 2507 |
+
"logps/rejected": -33.88913345336914,
|
| 2508 |
+
"loss": 0.2166,
|
| 2509 |
+
"rewards/accuracies": 0.9375,
|
| 2510 |
+
"rewards/chosen": 1.0830271244049072,
|
| 2511 |
+
"rewards/margins": 1.9926173686981201,
|
| 2512 |
+
"rewards/rejected": -0.9095903635025024,
|
| 2513 |
+
"step": 1520
|
| 2514 |
+
},
|
| 2515 |
+
{
|
| 2516 |
+
"epoch": 3.97,
|
| 2517 |
+
"grad_norm": 8.3125,
|
| 2518 |
+
"learning_rate": 6.421917227455999e-10,
|
| 2519 |
+
"logits/chosen": -1.9377511739730835,
|
| 2520 |
+
"logits/rejected": -1.9299259185791016,
|
| 2521 |
+
"logps/chosen": -26.640117645263672,
|
| 2522 |
+
"logps/rejected": -30.038944244384766,
|
| 2523 |
+
"loss": 0.1959,
|
| 2524 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 2525 |
+
"rewards/chosen": 0.9959781765937805,
|
| 2526 |
+
"rewards/margins": 1.9531135559082031,
|
| 2527 |
+
"rewards/rejected": -0.9571350812911987,
|
| 2528 |
+
"step": 1530
|
| 2529 |
+
},
|
| 2530 |
+
{
|
| 2531 |
+
"epoch": 4.0,
|
| 2532 |
+
"grad_norm": 6.03125,
|
| 2533 |
+
"learning_rate": 0.0,
|
| 2534 |
+
"logits/chosen": -1.8405107259750366,
|
| 2535 |
+
"logits/rejected": -1.8302587270736694,
|
| 2536 |
+
"logps/chosen": -31.55389976501465,
|
| 2537 |
+
"logps/rejected": -36.86316680908203,
|
| 2538 |
+
"loss": 0.1727,
|
| 2539 |
+
"rewards/accuracies": 0.987500011920929,
|
| 2540 |
+
"rewards/chosen": 1.058610200881958,
|
| 2541 |
+
"rewards/margins": 2.0750041007995605,
|
| 2542 |
+
"rewards/rejected": -1.0163936614990234,
|
| 2543 |
+
"step": 1540
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 4.0,
|
| 2547 |
+
"step": 1540,
|
| 2548 |
"total_flos": 0.0,
|
| 2549 |
+
"train_loss": 0.2216110908365869,
|
| 2550 |
+
"train_runtime": 10768.4727,
|
| 2551 |
+
"train_samples_per_second": 1.144,
|
| 2552 |
+
"train_steps_per_second": 0.143
|
| 2553 |
}
|
| 2554 |
],
|
| 2555 |
"logging_steps": 10,
|
| 2556 |
+
"max_steps": 1540,
|
| 2557 |
"num_input_tokens_seen": 0,
|
| 2558 |
+
"num_train_epochs": 4,
|
| 2559 |
"save_steps": 100,
|
| 2560 |
"total_flos": 0.0,
|
| 2561 |
"train_batch_size": 4,
|