Commit
•
b611c76
1
Parent(s):
30a8a30
Model save
Browse files- README.md +81 -0
- adapter_model.safetensors +1 -1
- all_results.json +21 -0
- eval_results.json +16 -0
- runs/Apr23_11-08-08_gpu4-119-5/events.out.tfevents.1713834594.gpu4-119-5.3428264.0 +2 -2
- runs/Apr23_11-08-08_gpu4-119-5/events.out.tfevents.1713848278.gpu4-119-5.3428264.1 +3 -0
- train_results.json +8 -0
- trainer_state.json +1490 -0
README.md
ADDED
|
@@ -0,0 +1,81 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: mit
|
| 3 |
+
library_name: peft
|
| 4 |
+
tags:
|
| 5 |
+
- trl
|
| 6 |
+
- dpo
|
| 7 |
+
- generated_from_trainer
|
| 8 |
+
base_model: microsoft/phi-2
|
| 9 |
+
model-index:
|
| 10 |
+
- name: phi-2-gpo-renew2-b0.001-extra-i1
|
| 11 |
+
results: []
|
| 12 |
+
---
|
| 13 |
+
|
| 14 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 15 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 16 |
+
|
| 17 |
+
# phi-2-gpo-renew2-b0.001-extra-i1
|
| 18 |
+
|
| 19 |
+
This model is a fine-tuned version of [microsoft/phi-2](https://huggingface.co/microsoft/phi-2) on the None dataset.
|
| 20 |
+
It achieves the following results on the evaluation set:
|
| 21 |
+
- Loss: 0.0421
|
| 22 |
+
- Rewards/chosen: -0.0100
|
| 23 |
+
- Rewards/rejected: -0.0414
|
| 24 |
+
- Rewards/accuracies: 0.6015
|
| 25 |
+
- Rewards/margins: 0.0314
|
| 26 |
+
- Logps/rejected: -408.6569
|
| 27 |
+
- Logps/chosen: -406.3272
|
| 28 |
+
- Logits/rejected: -1.0065
|
| 29 |
+
- Logits/chosen: -1.0521
|
| 30 |
+
|
| 31 |
+
## Model description
|
| 32 |
+
|
| 33 |
+
More information needed
|
| 34 |
+
|
| 35 |
+
## Intended uses & limitations
|
| 36 |
+
|
| 37 |
+
More information needed
|
| 38 |
+
|
| 39 |
+
## Training and evaluation data
|
| 40 |
+
|
| 41 |
+
More information needed
|
| 42 |
+
|
| 43 |
+
## Training procedure
|
| 44 |
+
|
| 45 |
+
### Training hyperparameters
|
| 46 |
+
|
| 47 |
+
The following hyperparameters were used during training:
|
| 48 |
+
- learning_rate: 5e-06
|
| 49 |
+
- train_batch_size: 4
|
| 50 |
+
- eval_batch_size: 4
|
| 51 |
+
- seed: 42
|
| 52 |
+
- distributed_type: multi-GPU
|
| 53 |
+
- gradient_accumulation_steps: 4
|
| 54 |
+
- total_train_batch_size: 16
|
| 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: 1
|
| 59 |
+
|
| 60 |
+
### Training results
|
| 61 |
+
|
| 62 |
+
| Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen |
|
| 63 |
+
|:-------------:|:-----:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:--------------:|:------------:|:---------------:|:-------------:|
|
| 64 |
+
| 0.0982 | 0.11 | 100 | 0.0526 | -0.0081 | -0.0101 | 0.5190 | 0.0020 | -377.3773 | -404.4459 | -0.7816 | -0.8697 |
|
| 65 |
+
| 0.0846 | 0.21 | 200 | 0.0485 | -0.0265 | -0.0402 | 0.5530 | 0.0137 | -407.4654 | -422.8199 | -0.9792 | -1.0427 |
|
| 66 |
+
| 0.0859 | 0.32 | 300 | 0.0464 | -0.0257 | -0.0460 | 0.5725 | 0.0203 | -413.2813 | -422.0490 | -1.0612 | -1.1154 |
|
| 67 |
+
| 0.0957 | 0.43 | 400 | 0.0443 | -0.0207 | -0.0481 | 0.5780 | 0.0274 | -415.3487 | -417.0023 | -1.0450 | -1.0984 |
|
| 68 |
+
| 0.068 | 0.53 | 500 | 0.0432 | -0.0067 | -0.0318 | 0.5955 | 0.0252 | -399.0811 | -402.9732 | -0.9791 | -1.0329 |
|
| 69 |
+
| 0.0847 | 0.64 | 600 | 0.0427 | -0.0050 | -0.0312 | 0.5945 | 0.0263 | -398.4744 | -401.2879 | -0.9837 | -1.0364 |
|
| 70 |
+
| 0.0519 | 0.75 | 700 | 0.0423 | -0.0082 | -0.0377 | 0.5905 | 0.0295 | -404.9791 | -404.5331 | -0.9872 | -1.0360 |
|
| 71 |
+
| 0.0742 | 0.85 | 800 | 0.0422 | -0.0105 | -0.0420 | 0.6000 | 0.0315 | -409.2462 | -406.8035 | -1.0109 | -1.0556 |
|
| 72 |
+
| 0.0768 | 0.96 | 900 | 0.0421 | -0.0100 | -0.0415 | 0.5930 | 0.0315 | -408.7397 | -406.3475 | -1.0050 | -1.0502 |
|
| 73 |
+
|
| 74 |
+
|
| 75 |
+
### Framework versions
|
| 76 |
+
|
| 77 |
+
- PEFT 0.7.1
|
| 78 |
+
- Transformers 4.36.2
|
| 79 |
+
- Pytorch 2.1.2
|
| 80 |
+
- Datasets 2.14.6
|
| 81 |
+
- Tokenizers 0.15.2
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 167807296
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bea97341984a3929c333752315ab1030a4371485fe4b2ca7a0a42d1472d664b2
|
| 3 |
size 167807296
|
all_results.json
ADDED
|
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"eval_logits/chosen": -1.0520579814910889,
|
| 4 |
+
"eval_logits/rejected": -1.0065128803253174,
|
| 5 |
+
"eval_logps/chosen": -406.3271789550781,
|
| 6 |
+
"eval_logps/rejected": -408.65692138671875,
|
| 7 |
+
"eval_loss": 0.042073994874954224,
|
| 8 |
+
"eval_rewards/accuracies": 0.6014999747276306,
|
| 9 |
+
"eval_rewards/chosen": -0.010021946392953396,
|
| 10 |
+
"eval_rewards/margins": 0.031399574130773544,
|
| 11 |
+
"eval_rewards/rejected": -0.041421521455049515,
|
| 12 |
+
"eval_runtime": 546.029,
|
| 13 |
+
"eval_samples": 2000,
|
| 14 |
+
"eval_samples_per_second": 3.663,
|
| 15 |
+
"eval_steps_per_second": 0.916,
|
| 16 |
+
"train_loss": 0.07849642387894454,
|
| 17 |
+
"train_runtime": 13138.6455,
|
| 18 |
+
"train_samples": 61135,
|
| 19 |
+
"train_samples_per_second": 1.142,
|
| 20 |
+
"train_steps_per_second": 0.071
|
| 21 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"eval_logits/chosen": -1.0520579814910889,
|
| 4 |
+
"eval_logits/rejected": -1.0065128803253174,
|
| 5 |
+
"eval_logps/chosen": -406.3271789550781,
|
| 6 |
+
"eval_logps/rejected": -408.65692138671875,
|
| 7 |
+
"eval_loss": 0.042073994874954224,
|
| 8 |
+
"eval_rewards/accuracies": 0.6014999747276306,
|
| 9 |
+
"eval_rewards/chosen": -0.010021946392953396,
|
| 10 |
+
"eval_rewards/margins": 0.031399574130773544,
|
| 11 |
+
"eval_rewards/rejected": -0.041421521455049515,
|
| 12 |
+
"eval_runtime": 546.029,
|
| 13 |
+
"eval_samples": 2000,
|
| 14 |
+
"eval_samples_per_second": 3.663,
|
| 15 |
+
"eval_steps_per_second": 0.916
|
| 16 |
+
}
|
runs/Apr23_11-08-08_gpu4-119-5/events.out.tfevents.1713834594.gpu4-119-5.3428264.0
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e719adeb13bb3954f77ccc64602132f29702cd2852fc88295977a44c7d81ef3c
|
| 3 |
+
size 71316
|
runs/Apr23_11-08-08_gpu4-119-5/events.out.tfevents.1713848278.gpu4-119-5.3428264.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bdb089c71d68fb97ba399c1461d4325cf2b5c931770f84dbe8b6a6445a9673da
|
| 3 |
+
size 828
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"train_loss": 0.07849642387894454,
|
| 4 |
+
"train_runtime": 13138.6455,
|
| 5 |
+
"train_samples": 61135,
|
| 6 |
+
"train_samples_per_second": 1.142,
|
| 7 |
+
"train_steps_per_second": 0.071
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1490 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.9994666666666666,
|
| 5 |
+
"eval_steps": 100,
|
| 6 |
+
"global_step": 937,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0,
|
| 13 |
+
"learning_rate": 5.319148936170213e-08,
|
| 14 |
+
"logits/chosen": -0.5045956373214722,
|
| 15 |
+
"logits/rejected": -0.805889368057251,
|
| 16 |
+
"logps/chosen": -165.41160583496094,
|
| 17 |
+
"logps/rejected": -172.8127899169922,
|
| 18 |
+
"loss": 0.0848,
|
| 19 |
+
"rewards/accuracies": 0.0,
|
| 20 |
+
"rewards/chosen": 0.0,
|
| 21 |
+
"rewards/margins": 0.0,
|
| 22 |
+
"rewards/rejected": 0.0,
|
| 23 |
+
"step": 1
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.01,
|
| 27 |
+
"learning_rate": 5.319148936170213e-07,
|
| 28 |
+
"logits/chosen": -0.851473867893219,
|
| 29 |
+
"logits/rejected": -0.8214991092681885,
|
| 30 |
+
"logps/chosen": -258.1239013671875,
|
| 31 |
+
"logps/rejected": -255.48716735839844,
|
| 32 |
+
"loss": 0.0877,
|
| 33 |
+
"rewards/accuracies": 0.2986111044883728,
|
| 34 |
+
"rewards/chosen": 0.0002587677154224366,
|
| 35 |
+
"rewards/margins": 0.00023072944895830005,
|
| 36 |
+
"rewards/rejected": 2.803823554131668e-05,
|
| 37 |
+
"step": 10
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.02,
|
| 41 |
+
"learning_rate": 1.0638297872340427e-06,
|
| 42 |
+
"logits/chosen": -0.8987849354743958,
|
| 43 |
+
"logits/rejected": -0.7349363565444946,
|
| 44 |
+
"logps/chosen": -260.9398193359375,
|
| 45 |
+
"logps/rejected": -253.32925415039062,
|
| 46 |
+
"loss": 0.0893,
|
| 47 |
+
"rewards/accuracies": 0.29374998807907104,
|
| 48 |
+
"rewards/chosen": -0.000219681765884161,
|
| 49 |
+
"rewards/margins": -3.2768032269814285e-06,
|
| 50 |
+
"rewards/rejected": -0.00021640490740537643,
|
| 51 |
+
"step": 20
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.03,
|
| 55 |
+
"learning_rate": 1.595744680851064e-06,
|
| 56 |
+
"logits/chosen": -0.9162956476211548,
|
| 57 |
+
"logits/rejected": -0.7800331115722656,
|
| 58 |
+
"logps/chosen": -240.79800415039062,
|
| 59 |
+
"logps/rejected": -235.59182739257812,
|
| 60 |
+
"loss": 0.0783,
|
| 61 |
+
"rewards/accuracies": 0.2874999940395355,
|
| 62 |
+
"rewards/chosen": -0.00024258208577521145,
|
| 63 |
+
"rewards/margins": 0.00012204260565340519,
|
| 64 |
+
"rewards/rejected": -0.0003646246623247862,
|
| 65 |
+
"step": 30
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.04,
|
| 69 |
+
"learning_rate": 2.1276595744680853e-06,
|
| 70 |
+
"logits/chosen": -0.8354488611221313,
|
| 71 |
+
"logits/rejected": -0.8405616879463196,
|
| 72 |
+
"logps/chosen": -255.01931762695312,
|
| 73 |
+
"logps/rejected": -224.09188842773438,
|
| 74 |
+
"loss": 0.0749,
|
| 75 |
+
"rewards/accuracies": 0.30000001192092896,
|
| 76 |
+
"rewards/chosen": 0.0001305304904235527,
|
| 77 |
+
"rewards/margins": 0.0002960737328976393,
|
| 78 |
+
"rewards/rejected": -0.0001655431988183409,
|
| 79 |
+
"step": 40
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.05,
|
| 83 |
+
"learning_rate": 2.6595744680851065e-06,
|
| 84 |
+
"logits/chosen": -0.9115155935287476,
|
| 85 |
+
"logits/rejected": -0.7566107511520386,
|
| 86 |
+
"logps/chosen": -295.87884521484375,
|
| 87 |
+
"logps/rejected": -261.06951904296875,
|
| 88 |
+
"loss": 0.0713,
|
| 89 |
+
"rewards/accuracies": 0.2874999940395355,
|
| 90 |
+
"rewards/chosen": -0.0011757513275370002,
|
| 91 |
+
"rewards/margins": -0.0006419935962185264,
|
| 92 |
+
"rewards/rejected": -0.0005337577313184738,
|
| 93 |
+
"step": 50
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.06,
|
| 97 |
+
"learning_rate": 3.191489361702128e-06,
|
| 98 |
+
"logits/chosen": -0.9281567335128784,
|
| 99 |
+
"logits/rejected": -0.8129026293754578,
|
| 100 |
+
"logps/chosen": -261.63751220703125,
|
| 101 |
+
"logps/rejected": -261.89483642578125,
|
| 102 |
+
"loss": 0.0779,
|
| 103 |
+
"rewards/accuracies": 0.32499998807907104,
|
| 104 |
+
"rewards/chosen": -0.000994718400761485,
|
| 105 |
+
"rewards/margins": 0.00043715062201954424,
|
| 106 |
+
"rewards/rejected": -0.0014318691100925207,
|
| 107 |
+
"step": 60
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.07,
|
| 111 |
+
"learning_rate": 3.723404255319149e-06,
|
| 112 |
+
"logits/chosen": -0.8597652316093445,
|
| 113 |
+
"logits/rejected": -0.8151811361312866,
|
| 114 |
+
"logps/chosen": -271.51458740234375,
|
| 115 |
+
"logps/rejected": -241.4061279296875,
|
| 116 |
+
"loss": 0.0898,
|
| 117 |
+
"rewards/accuracies": 0.3062500059604645,
|
| 118 |
+
"rewards/chosen": -0.0014379310887306929,
|
| 119 |
+
"rewards/margins": 0.0002869053860194981,
|
| 120 |
+
"rewards/rejected": -0.0017248367657884955,
|
| 121 |
+
"step": 70
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.09,
|
| 125 |
+
"learning_rate": 4.255319148936171e-06,
|
| 126 |
+
"logits/chosen": -0.8520501255989075,
|
| 127 |
+
"logits/rejected": -0.811953067779541,
|
| 128 |
+
"logps/chosen": -311.61431884765625,
|
| 129 |
+
"logps/rejected": -305.77520751953125,
|
| 130 |
+
"loss": 0.0795,
|
| 131 |
+
"rewards/accuracies": 0.33125001192092896,
|
| 132 |
+
"rewards/chosen": -0.0034166008699685335,
|
| 133 |
+
"rewards/margins": 0.00010135892080143094,
|
| 134 |
+
"rewards/rejected": -0.003517959965392947,
|
| 135 |
+
"step": 80
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.1,
|
| 139 |
+
"learning_rate": 4.787234042553192e-06,
|
| 140 |
+
"logits/chosen": -0.9568193554878235,
|
| 141 |
+
"logits/rejected": -0.8735030293464661,
|
| 142 |
+
"logps/chosen": -277.09405517578125,
|
| 143 |
+
"logps/rejected": -237.3052978515625,
|
| 144 |
+
"loss": 0.0831,
|
| 145 |
+
"rewards/accuracies": 0.3687500059604645,
|
| 146 |
+
"rewards/chosen": -0.0037826255429536104,
|
| 147 |
+
"rewards/margins": 0.0007604987476952374,
|
| 148 |
+
"rewards/rejected": -0.0045431237667799,
|
| 149 |
+
"step": 90
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.11,
|
| 153 |
+
"learning_rate": 4.999375059004058e-06,
|
| 154 |
+
"logits/chosen": -0.8838983774185181,
|
| 155 |
+
"logits/rejected": -0.8245723843574524,
|
| 156 |
+
"logps/chosen": -274.2312927246094,
|
| 157 |
+
"logps/rejected": -233.004638671875,
|
| 158 |
+
"loss": 0.0982,
|
| 159 |
+
"rewards/accuracies": 0.375,
|
| 160 |
+
"rewards/chosen": -0.004330983851104975,
|
| 161 |
+
"rewards/margins": 0.0017646064516156912,
|
| 162 |
+
"rewards/rejected": -0.00609559053555131,
|
| 163 |
+
"step": 100
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.11,
|
| 167 |
+
"eval_logits/chosen": -0.8696709871292114,
|
| 168 |
+
"eval_logits/rejected": -0.7816442847251892,
|
| 169 |
+
"eval_logps/chosen": -404.4459228515625,
|
| 170 |
+
"eval_logps/rejected": -377.37725830078125,
|
| 171 |
+
"eval_loss": 0.05261076241731644,
|
| 172 |
+
"eval_rewards/accuracies": 0.5189999938011169,
|
| 173 |
+
"eval_rewards/chosen": -0.008140643127262592,
|
| 174 |
+
"eval_rewards/margins": 0.0020011626183986664,
|
| 175 |
+
"eval_rewards/rejected": -0.010141806676983833,
|
| 176 |
+
"eval_runtime": 545.9504,
|
| 177 |
+
"eval_samples_per_second": 3.663,
|
| 178 |
+
"eval_steps_per_second": 0.916,
|
| 179 |
+
"step": 100
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 0.12,
|
| 183 |
+
"learning_rate": 4.9955571065548795e-06,
|
| 184 |
+
"logits/chosen": -0.9740250706672668,
|
| 185 |
+
"logits/rejected": -0.8206865191459656,
|
| 186 |
+
"logps/chosen": -308.79986572265625,
|
| 187 |
+
"logps/rejected": -279.56817626953125,
|
| 188 |
+
"loss": 0.0729,
|
| 189 |
+
"rewards/accuracies": 0.3499999940395355,
|
| 190 |
+
"rewards/chosen": -0.006666774861514568,
|
| 191 |
+
"rewards/margins": 0.001093443250283599,
|
| 192 |
+
"rewards/rejected": -0.0077602192759513855,
|
| 193 |
+
"step": 110
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 0.13,
|
| 197 |
+
"learning_rate": 4.9882736864879e-06,
|
| 198 |
+
"logits/chosen": -0.9458662271499634,
|
| 199 |
+
"logits/rejected": -0.8622045516967773,
|
| 200 |
+
"logps/chosen": -269.64190673828125,
|
| 201 |
+
"logps/rejected": -255.1685028076172,
|
| 202 |
+
"loss": 0.0678,
|
| 203 |
+
"rewards/accuracies": 0.36250001192092896,
|
| 204 |
+
"rewards/chosen": -0.006551130209118128,
|
| 205 |
+
"rewards/margins": 0.0029696193523705006,
|
| 206 |
+
"rewards/rejected": -0.009520749561488628,
|
| 207 |
+
"step": 120
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.14,
|
| 211 |
+
"learning_rate": 4.977534912960124e-06,
|
| 212 |
+
"logits/chosen": -1.0262787342071533,
|
| 213 |
+
"logits/rejected": -0.9416742324829102,
|
| 214 |
+
"logps/chosen": -228.7926788330078,
|
| 215 |
+
"logps/rejected": -229.67898559570312,
|
| 216 |
+
"loss": 0.0929,
|
| 217 |
+
"rewards/accuracies": 0.32499998807907104,
|
| 218 |
+
"rewards/chosen": -0.008367964997887611,
|
| 219 |
+
"rewards/margins": 0.0024298636708408594,
|
| 220 |
+
"rewards/rejected": -0.010797828435897827,
|
| 221 |
+
"step": 130
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 0.15,
|
| 225 |
+
"learning_rate": 4.963355698422092e-06,
|
| 226 |
+
"logits/chosen": -0.9834293127059937,
|
| 227 |
+
"logits/rejected": -0.9608744382858276,
|
| 228 |
+
"logps/chosen": -244.4986572265625,
|
| 229 |
+
"logps/rejected": -238.37118530273438,
|
| 230 |
+
"loss": 0.092,
|
| 231 |
+
"rewards/accuracies": 0.30000001192092896,
|
| 232 |
+
"rewards/chosen": -0.010878035798668861,
|
| 233 |
+
"rewards/margins": 0.003188747214153409,
|
| 234 |
+
"rewards/rejected": -0.014066783711314201,
|
| 235 |
+
"step": 140
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 0.16,
|
| 239 |
+
"learning_rate": 4.945755732909625e-06,
|
| 240 |
+
"logits/chosen": -1.0271694660186768,
|
| 241 |
+
"logits/rejected": -0.8658772706985474,
|
| 242 |
+
"logps/chosen": -303.09539794921875,
|
| 243 |
+
"logps/rejected": -265.5880126953125,
|
| 244 |
+
"loss": 0.0767,
|
| 245 |
+
"rewards/accuracies": 0.42500001192092896,
|
| 246 |
+
"rewards/chosen": -0.013803419657051563,
|
| 247 |
+
"rewards/margins": 0.005266121588647366,
|
| 248 |
+
"rewards/rejected": -0.01906954124569893,
|
| 249 |
+
"step": 150
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.17,
|
| 253 |
+
"learning_rate": 4.924759456701167e-06,
|
| 254 |
+
"logits/chosen": -1.1086270809173584,
|
| 255 |
+
"logits/rejected": -1.041982650756836,
|
| 256 |
+
"logps/chosen": -317.58245849609375,
|
| 257 |
+
"logps/rejected": -280.8768310546875,
|
| 258 |
+
"loss": 0.0902,
|
| 259 |
+
"rewards/accuracies": 0.39375001192092896,
|
| 260 |
+
"rewards/chosen": -0.021537428721785545,
|
| 261 |
+
"rewards/margins": 0.0052458057180047035,
|
| 262 |
+
"rewards/rejected": -0.026783233508467674,
|
| 263 |
+
"step": 160
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 0.18,
|
| 267 |
+
"learning_rate": 4.900396026378671e-06,
|
| 268 |
+
"logits/chosen": -1.1098581552505493,
|
| 269 |
+
"logits/rejected": -0.9803470373153687,
|
| 270 |
+
"logps/chosen": -335.6375427246094,
|
| 271 |
+
"logps/rejected": -316.95733642578125,
|
| 272 |
+
"loss": 0.0536,
|
| 273 |
+
"rewards/accuracies": 0.3812499940395355,
|
| 274 |
+
"rewards/chosen": -0.026775449514389038,
|
| 275 |
+
"rewards/margins": 0.008162637241184711,
|
| 276 |
+
"rewards/rejected": -0.03493808954954147,
|
| 277 |
+
"step": 170
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 0.19,
|
| 281 |
+
"learning_rate": 4.872699274339169e-06,
|
| 282 |
+
"logits/chosen": -1.0897270441055298,
|
| 283 |
+
"logits/rejected": -1.001300573348999,
|
| 284 |
+
"logps/chosen": -296.06353759765625,
|
| 285 |
+
"logps/rejected": -265.63751220703125,
|
| 286 |
+
"loss": 0.0746,
|
| 287 |
+
"rewards/accuracies": 0.3687500059604645,
|
| 288 |
+
"rewards/chosen": -0.022861812263727188,
|
| 289 |
+
"rewards/margins": 0.005839685909450054,
|
| 290 |
+
"rewards/rejected": -0.02870149537920952,
|
| 291 |
+
"step": 180
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 0.2,
|
| 295 |
+
"learning_rate": 4.8417076618132434e-06,
|
| 296 |
+
"logits/chosen": -1.0900896787643433,
|
| 297 |
+
"logits/rejected": -1.0583564043045044,
|
| 298 |
+
"logps/chosen": -266.21160888671875,
|
| 299 |
+
"logps/rejected": -245.45376586914062,
|
| 300 |
+
"loss": 0.0942,
|
| 301 |
+
"rewards/accuracies": 0.2874999940395355,
|
| 302 |
+
"rewards/chosen": -0.018833670765161514,
|
| 303 |
+
"rewards/margins": 0.00439481670036912,
|
| 304 |
+
"rewards/rejected": -0.023228485137224197,
|
| 305 |
+
"step": 190
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 0.21,
|
| 309 |
+
"learning_rate": 4.807464225455655e-06,
|
| 310 |
+
"logits/chosen": -1.0679035186767578,
|
| 311 |
+
"logits/rejected": -1.003225564956665,
|
| 312 |
+
"logps/chosen": -345.30181884765625,
|
| 313 |
+
"logps/rejected": -323.2543640136719,
|
| 314 |
+
"loss": 0.0846,
|
| 315 |
+
"rewards/accuracies": 0.39375001192092896,
|
| 316 |
+
"rewards/chosen": -0.02550928294658661,
|
| 317 |
+
"rewards/margins": 0.003808406414464116,
|
| 318 |
+
"rewards/rejected": -0.029317688196897507,
|
| 319 |
+
"step": 200
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 0.21,
|
| 323 |
+
"eval_logits/chosen": -1.0426863431930542,
|
| 324 |
+
"eval_logits/rejected": -0.9791997671127319,
|
| 325 |
+
"eval_logps/chosen": -422.8199462890625,
|
| 326 |
+
"eval_logps/rejected": -407.46539306640625,
|
| 327 |
+
"eval_loss": 0.048517368733882904,
|
| 328 |
+
"eval_rewards/accuracies": 0.5529999732971191,
|
| 329 |
+
"eval_rewards/chosen": -0.026514720171689987,
|
| 330 |
+
"eval_rewards/margins": 0.013715260662138462,
|
| 331 |
+
"eval_rewards/rejected": -0.040229979902505875,
|
| 332 |
+
"eval_runtime": 545.8919,
|
| 333 |
+
"eval_samples_per_second": 3.664,
|
| 334 |
+
"eval_steps_per_second": 0.916,
|
| 335 |
+
"step": 200
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 0.22,
|
| 339 |
+
"learning_rate": 4.770016517582283e-06,
|
| 340 |
+
"logits/chosen": -1.0943100452423096,
|
| 341 |
+
"logits/rejected": -1.0767104625701904,
|
| 342 |
+
"logps/chosen": -325.3826904296875,
|
| 343 |
+
"logps/rejected": -329.48663330078125,
|
| 344 |
+
"loss": 0.0676,
|
| 345 |
+
"rewards/accuracies": 0.45625001192092896,
|
| 346 |
+
"rewards/chosen": -0.018910493701696396,
|
| 347 |
+
"rewards/margins": 0.013479876331984997,
|
| 348 |
+
"rewards/rejected": -0.032390374690294266,
|
| 349 |
+
"step": 210
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 0.23,
|
| 353 |
+
"learning_rate": 4.7294165401363616e-06,
|
| 354 |
+
"logits/chosen": -1.0735843181610107,
|
| 355 |
+
"logits/rejected": -1.0250236988067627,
|
| 356 |
+
"logps/chosen": -313.1111145019531,
|
| 357 |
+
"logps/rejected": -292.7440490722656,
|
| 358 |
+
"loss": 0.0745,
|
| 359 |
+
"rewards/accuracies": 0.3812499940395355,
|
| 360 |
+
"rewards/chosen": -0.018653307110071182,
|
| 361 |
+
"rewards/margins": 0.01024434994906187,
|
| 362 |
+
"rewards/rejected": -0.028897657990455627,
|
| 363 |
+
"step": 220
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 0.25,
|
| 367 |
+
"learning_rate": 4.68572067247573e-06,
|
| 368 |
+
"logits/chosen": -1.1854560375213623,
|
| 369 |
+
"logits/rejected": -1.0616133213043213,
|
| 370 |
+
"logps/chosen": -281.5065612792969,
|
| 371 |
+
"logps/rejected": -277.0508728027344,
|
| 372 |
+
"loss": 0.0699,
|
| 373 |
+
"rewards/accuracies": 0.4000000059604645,
|
| 374 |
+
"rewards/chosen": -0.02585085853934288,
|
| 375 |
+
"rewards/margins": 0.01193526666611433,
|
| 376 |
+
"rewards/rejected": -0.037786126136779785,
|
| 377 |
+
"step": 230
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 0.26,
|
| 381 |
+
"learning_rate": 4.638989593081364e-06,
|
| 382 |
+
"logits/chosen": -1.1488720178604126,
|
| 383 |
+
"logits/rejected": -1.0367449522018433,
|
| 384 |
+
"logps/chosen": -283.6094970703125,
|
| 385 |
+
"logps/rejected": -252.0610809326172,
|
| 386 |
+
"loss": 0.0894,
|
| 387 |
+
"rewards/accuracies": 0.28125,
|
| 388 |
+
"rewards/chosen": -0.020607244223356247,
|
| 389 |
+
"rewards/margins": 0.006351941730827093,
|
| 390 |
+
"rewards/rejected": -0.026959186419844627,
|
| 391 |
+
"step": 240
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 0.27,
|
| 395 |
+
"learning_rate": 4.5892881952959015e-06,
|
| 396 |
+
"logits/chosen": -1.1442514657974243,
|
| 397 |
+
"logits/rejected": -1.1018245220184326,
|
| 398 |
+
"logps/chosen": -303.63458251953125,
|
| 399 |
+
"logps/rejected": -305.7111511230469,
|
| 400 |
+
"loss": 0.0878,
|
| 401 |
+
"rewards/accuracies": 0.4000000059604645,
|
| 402 |
+
"rewards/chosen": -0.02438071370124817,
|
| 403 |
+
"rewards/margins": 0.009189085103571415,
|
| 404 |
+
"rewards/rejected": -0.03356979787349701,
|
| 405 |
+
"step": 250
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 0.28,
|
| 409 |
+
"learning_rate": 4.536685497209182e-06,
|
| 410 |
+
"logits/chosen": -1.134932041168213,
|
| 411 |
+
"logits/rejected": -1.106890082359314,
|
| 412 |
+
"logps/chosen": -300.16815185546875,
|
| 413 |
+
"logps/rejected": -284.40765380859375,
|
| 414 |
+
"loss": 0.0921,
|
| 415 |
+
"rewards/accuracies": 0.3812499940395355,
|
| 416 |
+
"rewards/chosen": -0.025658372789621353,
|
| 417 |
+
"rewards/margins": 0.005949888378381729,
|
| 418 |
+
"rewards/rejected": -0.03160826116800308,
|
| 419 |
+
"step": 260
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 0.29,
|
| 423 |
+
"learning_rate": 4.481254545815943e-06,
|
| 424 |
+
"logits/chosen": -1.1486608982086182,
|
| 425 |
+
"logits/rejected": -1.0326900482177734,
|
| 426 |
+
"logps/chosen": -294.9302673339844,
|
| 427 |
+
"logps/rejected": -285.3538513183594,
|
| 428 |
+
"loss": 0.0814,
|
| 429 |
+
"rewards/accuracies": 0.36250001192092896,
|
| 430 |
+
"rewards/chosen": -0.022086424753069878,
|
| 431 |
+
"rewards/margins": 0.010328343138098717,
|
| 432 |
+
"rewards/rejected": -0.03241477161645889,
|
| 433 |
+
"step": 270
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 0.3,
|
| 437 |
+
"learning_rate": 4.42307231557875e-06,
|
| 438 |
+
"logits/chosen": -1.121669888496399,
|
| 439 |
+
"logits/rejected": -1.0683071613311768,
|
| 440 |
+
"logps/chosen": -310.044189453125,
|
| 441 |
+
"logps/rejected": -306.346435546875,
|
| 442 |
+
"loss": 0.0916,
|
| 443 |
+
"rewards/accuracies": 0.4000000059604645,
|
| 444 |
+
"rewards/chosen": -0.02136383019387722,
|
| 445 |
+
"rewards/margins": 0.012828357517719269,
|
| 446 |
+
"rewards/rejected": -0.03419218957424164,
|
| 447 |
+
"step": 280
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 0.31,
|
| 451 |
+
"learning_rate": 4.3622196015370305e-06,
|
| 452 |
+
"logits/chosen": -1.1434075832366943,
|
| 453 |
+
"logits/rejected": -1.1294233798980713,
|
| 454 |
+
"logps/chosen": -259.6036682128906,
|
| 455 |
+
"logps/rejected": -267.0892333984375,
|
| 456 |
+
"loss": 0.0856,
|
| 457 |
+
"rewards/accuracies": 0.3375000059604645,
|
| 458 |
+
"rewards/chosen": -0.01935209520161152,
|
| 459 |
+
"rewards/margins": 0.008529379032552242,
|
| 460 |
+
"rewards/rejected": -0.027881473302841187,
|
| 461 |
+
"step": 290
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 0.32,
|
| 465 |
+
"learning_rate": 4.298780907110648e-06,
|
| 466 |
+
"logits/chosen": -1.214237928390503,
|
| 467 |
+
"logits/rejected": -1.1189312934875488,
|
| 468 |
+
"logps/chosen": -250.4637451171875,
|
| 469 |
+
"logps/rejected": -255.03079223632812,
|
| 470 |
+
"loss": 0.0859,
|
| 471 |
+
"rewards/accuracies": 0.35624998807907104,
|
| 472 |
+
"rewards/chosen": -0.014152693562209606,
|
| 473 |
+
"rewards/margins": 0.011050628498196602,
|
| 474 |
+
"rewards/rejected": -0.025203322991728783,
|
| 475 |
+
"step": 300
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"epoch": 0.32,
|
| 479 |
+
"eval_logits/chosen": -1.11543869972229,
|
| 480 |
+
"eval_logits/rejected": -1.0612365007400513,
|
| 481 |
+
"eval_logps/chosen": -422.0489807128906,
|
| 482 |
+
"eval_logps/rejected": -413.28131103515625,
|
| 483 |
+
"eval_loss": 0.046408262103796005,
|
| 484 |
+
"eval_rewards/accuracies": 0.5724999904632568,
|
| 485 |
+
"eval_rewards/chosen": -0.02574371173977852,
|
| 486 |
+
"eval_rewards/margins": 0.02030220627784729,
|
| 487 |
+
"eval_rewards/rejected": -0.04604591801762581,
|
| 488 |
+
"eval_runtime": 546.0923,
|
| 489 |
+
"eval_samples_per_second": 3.662,
|
| 490 |
+
"eval_steps_per_second": 0.916,
|
| 491 |
+
"step": 300
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 0.33,
|
| 495 |
+
"learning_rate": 4.23284432675381e-06,
|
| 496 |
+
"logits/chosen": -1.1313543319702148,
|
| 497 |
+
"logits/rejected": -1.0724719762802124,
|
| 498 |
+
"logps/chosen": -262.6627502441406,
|
| 499 |
+
"logps/rejected": -277.54632568359375,
|
| 500 |
+
"loss": 0.0916,
|
| 501 |
+
"rewards/accuracies": 0.38749998807907104,
|
| 502 |
+
"rewards/chosen": -0.019768275320529938,
|
| 503 |
+
"rewards/margins": 0.014759126119315624,
|
| 504 |
+
"rewards/rejected": -0.03452740237116814,
|
| 505 |
+
"step": 310
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 0.34,
|
| 509 |
+
"learning_rate": 4.164501423622277e-06,
|
| 510 |
+
"logits/chosen": -1.246671199798584,
|
| 511 |
+
"logits/rejected": -1.1658015251159668,
|
| 512 |
+
"logps/chosen": -264.7757873535156,
|
| 513 |
+
"logps/rejected": -277.6180114746094,
|
| 514 |
+
"loss": 0.0774,
|
| 515 |
+
"rewards/accuracies": 0.33125001192092896,
|
| 516 |
+
"rewards/chosen": -0.020890336483716965,
|
| 517 |
+
"rewards/margins": 0.011497320607304573,
|
| 518 |
+
"rewards/rejected": -0.03238765895366669,
|
| 519 |
+
"step": 320
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 0.35,
|
| 523 |
+
"learning_rate": 4.0938471024237355e-06,
|
| 524 |
+
"logits/chosen": -1.222037672996521,
|
| 525 |
+
"logits/rejected": -1.156553030014038,
|
| 526 |
+
"logps/chosen": -256.74359130859375,
|
| 527 |
+
"logps/rejected": -277.230224609375,
|
| 528 |
+
"loss": 0.0805,
|
| 529 |
+
"rewards/accuracies": 0.4124999940395355,
|
| 530 |
+
"rewards/chosen": -0.015277216210961342,
|
| 531 |
+
"rewards/margins": 0.02550993300974369,
|
| 532 |
+
"rewards/rejected": -0.04078715294599533,
|
| 533 |
+
"step": 330
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 0.36,
|
| 537 |
+
"learning_rate": 4.020979477627907e-06,
|
| 538 |
+
"logits/chosen": -1.1742956638336182,
|
| 539 |
+
"logits/rejected": -1.1521165370941162,
|
| 540 |
+
"logps/chosen": -239.5863494873047,
|
| 541 |
+
"logps/rejected": -247.32522583007812,
|
| 542 |
+
"loss": 0.1015,
|
| 543 |
+
"rewards/accuracies": 0.3687500059604645,
|
| 544 |
+
"rewards/chosen": -0.01359327882528305,
|
| 545 |
+
"rewards/margins": 0.015490619465708733,
|
| 546 |
+
"rewards/rejected": -0.029083898290991783,
|
| 547 |
+
"step": 340
|
| 548 |
+
},
|
| 549 |
+
{
|
| 550 |
+
"epoch": 0.37,
|
| 551 |
+
"learning_rate": 3.9459997372194105e-06,
|
| 552 |
+
"logits/chosen": -1.1355948448181152,
|
| 553 |
+
"logits/rejected": -1.0996363162994385,
|
| 554 |
+
"logps/chosen": -268.62945556640625,
|
| 555 |
+
"logps/rejected": -278.85870361328125,
|
| 556 |
+
"loss": 0.0786,
|
| 557 |
+
"rewards/accuracies": 0.35624998807907104,
|
| 558 |
+
"rewards/chosen": -0.013594739139080048,
|
| 559 |
+
"rewards/margins": 0.00844600610435009,
|
| 560 |
+
"rewards/rejected": -0.02204074338078499,
|
| 561 |
+
"step": 350
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 0.38,
|
| 565 |
+
"learning_rate": 3.869012002182573e-06,
|
| 566 |
+
"logits/chosen": -1.2170326709747314,
|
| 567 |
+
"logits/rejected": -1.1180956363677979,
|
| 568 |
+
"logps/chosen": -265.72393798828125,
|
| 569 |
+
"logps/rejected": -233.3331298828125,
|
| 570 |
+
"loss": 0.0879,
|
| 571 |
+
"rewards/accuracies": 0.3812499940395355,
|
| 572 |
+
"rewards/chosen": -0.01117833610624075,
|
| 573 |
+
"rewards/margins": 0.015456246212124825,
|
| 574 |
+
"rewards/rejected": -0.02663458324968815,
|
| 575 |
+
"step": 360
|
| 576 |
+
},
|
| 577 |
+
{
|
| 578 |
+
"epoch": 0.39,
|
| 579 |
+
"learning_rate": 3.7901231819133104e-06,
|
| 580 |
+
"logits/chosen": -1.1366071701049805,
|
| 581 |
+
"logits/rejected": -1.0916543006896973,
|
| 582 |
+
"logps/chosen": -266.32037353515625,
|
| 583 |
+
"logps/rejected": -267.02313232421875,
|
| 584 |
+
"loss": 0.0726,
|
| 585 |
+
"rewards/accuracies": 0.4124999940395355,
|
| 586 |
+
"rewards/chosen": -0.01092919148504734,
|
| 587 |
+
"rewards/margins": 0.01567809283733368,
|
| 588 |
+
"rewards/rejected": -0.02660728432238102,
|
| 589 |
+
"step": 370
|
| 590 |
+
},
|
| 591 |
+
{
|
| 592 |
+
"epoch": 0.41,
|
| 593 |
+
"learning_rate": 3.709442825758875e-06,
|
| 594 |
+
"logits/chosen": -1.089163064956665,
|
| 595 |
+
"logits/rejected": -1.0779684782028198,
|
| 596 |
+
"logps/chosen": -259.6034240722656,
|
| 597 |
+
"logps/rejected": -261.5791015625,
|
| 598 |
+
"loss": 0.0859,
|
| 599 |
+
"rewards/accuracies": 0.41874998807907104,
|
| 600 |
+
"rewards/chosen": -0.008027950301766396,
|
| 601 |
+
"rewards/margins": 0.02001611702144146,
|
| 602 |
+
"rewards/rejected": -0.028044065460562706,
|
| 603 |
+
"step": 380
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 0.42,
|
| 607 |
+
"learning_rate": 3.6270829708916113e-06,
|
| 608 |
+
"logits/chosen": -1.0756354331970215,
|
| 609 |
+
"logits/rejected": -1.0809965133666992,
|
| 610 |
+
"logps/chosen": -317.52227783203125,
|
| 611 |
+
"logps/rejected": -299.58990478515625,
|
| 612 |
+
"loss": 0.066,
|
| 613 |
+
"rewards/accuracies": 0.4124999940395355,
|
| 614 |
+
"rewards/chosen": -0.012966620735824108,
|
| 615 |
+
"rewards/margins": 0.015139798633754253,
|
| 616 |
+
"rewards/rejected": -0.02810642123222351,
|
| 617 |
+
"step": 390
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 0.43,
|
| 621 |
+
"learning_rate": 3.543157986727991e-06,
|
| 622 |
+
"logits/chosen": -1.1576625108718872,
|
| 623 |
+
"logits/rejected": -1.0739920139312744,
|
| 624 |
+
"logps/chosen": -259.60516357421875,
|
| 625 |
+
"logps/rejected": -274.40240478515625,
|
| 626 |
+
"loss": 0.0957,
|
| 627 |
+
"rewards/accuracies": 0.48124998807907104,
|
| 628 |
+
"rewards/chosen": -0.006962036248296499,
|
| 629 |
+
"rewards/margins": 0.03462858498096466,
|
| 630 |
+
"rewards/rejected": -0.041590623557567596,
|
| 631 |
+
"step": 400
|
| 632 |
+
},
|
| 633 |
+
{
|
| 634 |
+
"epoch": 0.43,
|
| 635 |
+
"eval_logits/chosen": -1.0983970165252686,
|
| 636 |
+
"eval_logits/rejected": -1.0450440645217896,
|
| 637 |
+
"eval_logps/chosen": -417.00225830078125,
|
| 638 |
+
"eval_logps/rejected": -415.34869384765625,
|
| 639 |
+
"eval_loss": 0.04426228255033493,
|
| 640 |
+
"eval_rewards/accuracies": 0.578000009059906,
|
| 641 |
+
"eval_rewards/chosen": -0.02069696970283985,
|
| 642 |
+
"eval_rewards/margins": 0.027416307479143143,
|
| 643 |
+
"eval_rewards/rejected": -0.04811327904462814,
|
| 644 |
+
"eval_runtime": 545.9497,
|
| 645 |
+
"eval_samples_per_second": 3.663,
|
| 646 |
+
"eval_steps_per_second": 0.916,
|
| 647 |
+
"step": 400
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.44,
|
| 651 |
+
"learning_rate": 3.4577844161089614e-06,
|
| 652 |
+
"logits/chosen": -1.0938787460327148,
|
| 653 |
+
"logits/rejected": -1.0761361122131348,
|
| 654 |
+
"logps/chosen": -252.89974975585938,
|
| 655 |
+
"logps/rejected": -279.4630126953125,
|
| 656 |
+
"loss": 0.0703,
|
| 657 |
+
"rewards/accuracies": 0.3687500059604645,
|
| 658 |
+
"rewards/chosen": -0.017448369413614273,
|
| 659 |
+
"rewards/margins": 0.02003558911383152,
|
| 660 |
+
"rewards/rejected": -0.03748396039009094,
|
| 661 |
+
"step": 410
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.45,
|
| 665 |
+
"learning_rate": 3.3710808134621577e-06,
|
| 666 |
+
"logits/chosen": -1.1595834493637085,
|
| 667 |
+
"logits/rejected": -1.126773476600647,
|
| 668 |
+
"logps/chosen": -298.33538818359375,
|
| 669 |
+
"logps/rejected": -295.24468994140625,
|
| 670 |
+
"loss": 0.0549,
|
| 671 |
+
"rewards/accuracies": 0.41874998807907104,
|
| 672 |
+
"rewards/chosen": -0.016742903739213943,
|
| 673 |
+
"rewards/margins": 0.022577572613954544,
|
| 674 |
+
"rewards/rejected": -0.03932047635316849,
|
| 675 |
+
"step": 420
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.46,
|
| 679 |
+
"learning_rate": 3.2831675801707126e-06,
|
| 680 |
+
"logits/chosen": -1.1363260746002197,
|
| 681 |
+
"logits/rejected": -1.1643562316894531,
|
| 682 |
+
"logps/chosen": -259.0823669433594,
|
| 683 |
+
"logps/rejected": -262.0459899902344,
|
| 684 |
+
"loss": 0.08,
|
| 685 |
+
"rewards/accuracies": 0.35624998807907104,
|
| 686 |
+
"rewards/chosen": -0.01677551493048668,
|
| 687 |
+
"rewards/margins": 0.012358926236629486,
|
| 688 |
+
"rewards/rejected": -0.029134441167116165,
|
| 689 |
+
"step": 430
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.47,
|
| 693 |
+
"learning_rate": 3.194166797377289e-06,
|
| 694 |
+
"logits/chosen": -1.0814932584762573,
|
| 695 |
+
"logits/rejected": -1.0023730993270874,
|
| 696 |
+
"logps/chosen": -267.9959411621094,
|
| 697 |
+
"logps/rejected": -241.27685546875,
|
| 698 |
+
"loss": 0.086,
|
| 699 |
+
"rewards/accuracies": 0.3125,
|
| 700 |
+
"rewards/chosen": -0.014996061101555824,
|
| 701 |
+
"rewards/margins": 0.004354935139417648,
|
| 702 |
+
"rewards/rejected": -0.019350996240973473,
|
| 703 |
+
"step": 440
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.48,
|
| 707 |
+
"learning_rate": 3.104202056455501e-06,
|
| 708 |
+
"logits/chosen": -1.1119335889816284,
|
| 709 |
+
"logits/rejected": -1.036833643913269,
|
| 710 |
+
"logps/chosen": -264.6695251464844,
|
| 711 |
+
"logps/rejected": -294.46173095703125,
|
| 712 |
+
"loss": 0.0732,
|
| 713 |
+
"rewards/accuracies": 0.3812499940395355,
|
| 714 |
+
"rewards/chosen": -0.00541292130947113,
|
| 715 |
+
"rewards/margins": 0.018471624702215195,
|
| 716 |
+
"rewards/rejected": -0.023884546011686325,
|
| 717 |
+
"step": 450
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.49,
|
| 721 |
+
"learning_rate": 3.013398287384144e-06,
|
| 722 |
+
"logits/chosen": -1.100239872932434,
|
| 723 |
+
"logits/rejected": -0.9833891987800598,
|
| 724 |
+
"logps/chosen": -265.06292724609375,
|
| 725 |
+
"logps/rejected": -251.05789184570312,
|
| 726 |
+
"loss": 0.0769,
|
| 727 |
+
"rewards/accuracies": 0.3499999940395355,
|
| 728 |
+
"rewards/chosen": -0.00949514377862215,
|
| 729 |
+
"rewards/margins": 0.011730840429663658,
|
| 730 |
+
"rewards/rejected": -0.021225983276963234,
|
| 731 |
+
"step": 460
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.5,
|
| 735 |
+
"learning_rate": 2.9218815852625717e-06,
|
| 736 |
+
"logits/chosen": -1.0965584516525269,
|
| 737 |
+
"logits/rejected": -1.1050448417663574,
|
| 738 |
+
"logps/chosen": -245.2798309326172,
|
| 739 |
+
"logps/rejected": -247.6891632080078,
|
| 740 |
+
"loss": 0.0867,
|
| 741 |
+
"rewards/accuracies": 0.375,
|
| 742 |
+
"rewards/chosen": -0.00036931521026417613,
|
| 743 |
+
"rewards/margins": 0.01472887396812439,
|
| 744 |
+
"rewards/rejected": -0.015098191797733307,
|
| 745 |
+
"step": 470
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.51,
|
| 749 |
+
"learning_rate": 2.829779035208113e-06,
|
| 750 |
+
"logits/chosen": -1.1267783641815186,
|
| 751 |
+
"logits/rejected": -1.0608749389648438,
|
| 752 |
+
"logps/chosen": -276.13861083984375,
|
| 753 |
+
"logps/rejected": -242.2423858642578,
|
| 754 |
+
"loss": 0.0866,
|
| 755 |
+
"rewards/accuracies": 0.4000000059604645,
|
| 756 |
+
"rewards/chosen": 0.003509046044200659,
|
| 757 |
+
"rewards/margins": 0.01866857148706913,
|
| 758 |
+
"rewards/rejected": -0.015159524977207184,
|
| 759 |
+
"step": 480
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.52,
|
| 763 |
+
"learning_rate": 2.737218535878705e-06,
|
| 764 |
+
"logits/chosen": -1.1088229417800903,
|
| 765 |
+
"logits/rejected": -1.0114442110061646,
|
| 766 |
+
"logps/chosen": -264.47747802734375,
|
| 767 |
+
"logps/rejected": -256.66265869140625,
|
| 768 |
+
"loss": 0.0697,
|
| 769 |
+
"rewards/accuracies": 0.45625001192092896,
|
| 770 |
+
"rewards/chosen": -0.0013200236717239022,
|
| 771 |
+
"rewards/margins": 0.02153395116329193,
|
| 772 |
+
"rewards/rejected": -0.022853974252939224,
|
| 773 |
+
"step": 490
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.53,
|
| 777 |
+
"learning_rate": 2.64432862186579e-06,
|
| 778 |
+
"logits/chosen": -1.112363338470459,
|
| 779 |
+
"logits/rejected": -1.0634427070617676,
|
| 780 |
+
"logps/chosen": -263.20428466796875,
|
| 781 |
+
"logps/rejected": -262.7369079589844,
|
| 782 |
+
"loss": 0.068,
|
| 783 |
+
"rewards/accuracies": 0.41874998807907104,
|
| 784 |
+
"rewards/chosen": -0.0034271120093762875,
|
| 785 |
+
"rewards/margins": 0.013932084664702415,
|
| 786 |
+
"rewards/rejected": -0.01735919900238514,
|
| 787 |
+
"step": 500
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.53,
|
| 791 |
+
"eval_logits/chosen": -1.0329285860061646,
|
| 792 |
+
"eval_logits/rejected": -0.9790877103805542,
|
| 793 |
+
"eval_logps/chosen": -402.9732360839844,
|
| 794 |
+
"eval_logps/rejected": -399.08111572265625,
|
| 795 |
+
"eval_loss": 0.04319905489683151,
|
| 796 |
+
"eval_rewards/accuracies": 0.5954999923706055,
|
| 797 |
+
"eval_rewards/chosen": -0.006668027024716139,
|
| 798 |
+
"eval_rewards/margins": 0.025177694857120514,
|
| 799 |
+
"eval_rewards/rejected": -0.03184572234749794,
|
| 800 |
+
"eval_runtime": 545.9259,
|
| 801 |
+
"eval_samples_per_second": 3.664,
|
| 802 |
+
"eval_steps_per_second": 0.916,
|
| 803 |
+
"step": 500
|
| 804 |
+
},
|
| 805 |
+
{
|
| 806 |
+
"epoch": 0.54,
|
| 807 |
+
"learning_rate": 2.551238285204126e-06,
|
| 808 |
+
"logits/chosen": -1.0819242000579834,
|
| 809 |
+
"logits/rejected": -1.0868195295333862,
|
| 810 |
+
"logps/chosen": -204.0397186279297,
|
| 811 |
+
"logps/rejected": -212.2799835205078,
|
| 812 |
+
"loss": 0.0808,
|
| 813 |
+
"rewards/accuracies": 0.375,
|
| 814 |
+
"rewards/chosen": -0.003992350306361914,
|
| 815 |
+
"rewards/margins": 0.010435246862471104,
|
| 816 |
+
"rewards/rejected": -0.01442759484052658,
|
| 817 |
+
"step": 510
|
| 818 |
+
},
|
| 819 |
+
{
|
| 820 |
+
"epoch": 0.55,
|
| 821 |
+
"learning_rate": 2.4580767962463688e-06,
|
| 822 |
+
"logits/chosen": -1.1004348993301392,
|
| 823 |
+
"logits/rejected": -1.0923566818237305,
|
| 824 |
+
"logps/chosen": -287.0587463378906,
|
| 825 |
+
"logps/rejected": -295.1195373535156,
|
| 826 |
+
"loss": 0.0783,
|
| 827 |
+
"rewards/accuracies": 0.45625001192092896,
|
| 828 |
+
"rewards/chosen": -0.005773237906396389,
|
| 829 |
+
"rewards/margins": 0.016825079917907715,
|
| 830 |
+
"rewards/rejected": -0.02259831875562668,
|
| 831 |
+
"step": 520
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 0.57,
|
| 835 |
+
"learning_rate": 2.3649735241511546e-06,
|
| 836 |
+
"logits/chosen": -1.025914192199707,
|
| 837 |
+
"logits/rejected": -1.0844639539718628,
|
| 838 |
+
"logps/chosen": -233.93783569335938,
|
| 839 |
+
"logps/rejected": -278.87310791015625,
|
| 840 |
+
"loss": 0.0701,
|
| 841 |
+
"rewards/accuracies": 0.4375,
|
| 842 |
+
"rewards/chosen": 0.0003628075937740505,
|
| 843 |
+
"rewards/margins": 0.019406834617257118,
|
| 844 |
+
"rewards/rejected": -0.019044026732444763,
|
| 845 |
+
"step": 530
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 0.58,
|
| 849 |
+
"learning_rate": 2.2720577572339914e-06,
|
| 850 |
+
"logits/chosen": -1.0895086526870728,
|
| 851 |
+
"logits/rejected": -0.9644759297370911,
|
| 852 |
+
"logps/chosen": -313.9569396972656,
|
| 853 |
+
"logps/rejected": -267.2975158691406,
|
| 854 |
+
"loss": 0.0828,
|
| 855 |
+
"rewards/accuracies": 0.375,
|
| 856 |
+
"rewards/chosen": -0.009684056974947453,
|
| 857 |
+
"rewards/margins": 0.013868686743080616,
|
| 858 |
+
"rewards/rejected": -0.02355274185538292,
|
| 859 |
+
"step": 540
|
| 860 |
+
},
|
| 861 |
+
{
|
| 862 |
+
"epoch": 0.59,
|
| 863 |
+
"learning_rate": 2.1794585234303995e-06,
|
| 864 |
+
"logits/chosen": -1.039159893989563,
|
| 865 |
+
"logits/rejected": -1.0821508169174194,
|
| 866 |
+
"logps/chosen": -257.9473571777344,
|
| 867 |
+
"logps/rejected": -274.3190002441406,
|
| 868 |
+
"loss": 0.0743,
|
| 869 |
+
"rewards/accuracies": 0.3687500059604645,
|
| 870 |
+
"rewards/chosen": -0.0062354551628232,
|
| 871 |
+
"rewards/margins": 0.01928626373410225,
|
| 872 |
+
"rewards/rejected": -0.025521719828248024,
|
| 873 |
+
"step": 550
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 0.6,
|
| 877 |
+
"learning_rate": 2.0873044111206407e-06,
|
| 878 |
+
"logits/chosen": -1.0668865442276,
|
| 879 |
+
"logits/rejected": -1.0260220766067505,
|
| 880 |
+
"logps/chosen": -271.9998474121094,
|
| 881 |
+
"logps/rejected": -282.83038330078125,
|
| 882 |
+
"loss": 0.0714,
|
| 883 |
+
"rewards/accuracies": 0.4000000059604645,
|
| 884 |
+
"rewards/chosen": 0.0038192705251276493,
|
| 885 |
+
"rewards/margins": 0.019924623891711235,
|
| 886 |
+
"rewards/rejected": -0.016105355694890022,
|
| 887 |
+
"step": 560
|
| 888 |
+
},
|
| 889 |
+
{
|
| 890 |
+
"epoch": 0.61,
|
| 891 |
+
"learning_rate": 1.9957233905648293e-06,
|
| 892 |
+
"logits/chosen": -1.054386019706726,
|
| 893 |
+
"logits/rejected": -1.042152762413025,
|
| 894 |
+
"logps/chosen": -284.9267272949219,
|
| 895 |
+
"logps/rejected": -249.73110961914062,
|
| 896 |
+
"loss": 0.0739,
|
| 897 |
+
"rewards/accuracies": 0.40625,
|
| 898 |
+
"rewards/chosen": 0.0018435310339555144,
|
| 899 |
+
"rewards/margins": 0.013795648701488972,
|
| 900 |
+
"rewards/rejected": -0.011952118948101997,
|
| 901 |
+
"step": 570
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 0.62,
|
| 905 |
+
"learning_rate": 1.904842636196402e-06,
|
| 906 |
+
"logits/chosen": -1.027489185333252,
|
| 907 |
+
"logits/rejected": -1.0255085229873657,
|
| 908 |
+
"logps/chosen": -250.985107421875,
|
| 909 |
+
"logps/rejected": -246.9531707763672,
|
| 910 |
+
"loss": 0.0877,
|
| 911 |
+
"rewards/accuracies": 0.40625,
|
| 912 |
+
"rewards/chosen": -0.00342792016454041,
|
| 913 |
+
"rewards/margins": 0.010951442644000053,
|
| 914 |
+
"rewards/rejected": -0.01437936257570982,
|
| 915 |
+
"step": 580
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 0.63,
|
| 919 |
+
"learning_rate": 1.814788350020726e-06,
|
| 920 |
+
"logits/chosen": -1.0244547128677368,
|
| 921 |
+
"logits/rejected": -1.0658903121948242,
|
| 922 |
+
"logps/chosen": -295.65191650390625,
|
| 923 |
+
"logps/rejected": -290.0276184082031,
|
| 924 |
+
"loss": 0.0697,
|
| 925 |
+
"rewards/accuracies": 0.42500001192092896,
|
| 926 |
+
"rewards/chosen": -0.00019127638370264322,
|
| 927 |
+
"rewards/margins": 0.02073330618441105,
|
| 928 |
+
"rewards/rejected": -0.020924581214785576,
|
| 929 |
+
"step": 590
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 0.64,
|
| 933 |
+
"learning_rate": 1.725685586364051e-06,
|
| 934 |
+
"logits/chosen": -1.0251004695892334,
|
| 935 |
+
"logits/rejected": -1.0600395202636719,
|
| 936 |
+
"logps/chosen": -244.59384155273438,
|
| 937 |
+
"logps/rejected": -243.28256225585938,
|
| 938 |
+
"loss": 0.0847,
|
| 939 |
+
"rewards/accuracies": 0.41874998807907104,
|
| 940 |
+
"rewards/chosen": -0.001342198345810175,
|
| 941 |
+
"rewards/margins": 0.01429178100079298,
|
| 942 |
+
"rewards/rejected": -0.015633979812264442,
|
| 943 |
+
"step": 600
|
| 944 |
+
},
|
| 945 |
+
{
|
| 946 |
+
"epoch": 0.64,
|
| 947 |
+
"eval_logits/chosen": -1.0363810062408447,
|
| 948 |
+
"eval_logits/rejected": -0.9837189316749573,
|
| 949 |
+
"eval_logps/chosen": -401.28790283203125,
|
| 950 |
+
"eval_logps/rejected": -398.47442626953125,
|
| 951 |
+
"eval_loss": 0.042725615203380585,
|
| 952 |
+
"eval_rewards/accuracies": 0.5945000052452087,
|
| 953 |
+
"eval_rewards/chosen": -0.0049826642498373985,
|
| 954 |
+
"eval_rewards/margins": 0.026256347075104713,
|
| 955 |
+
"eval_rewards/rejected": -0.031239010393619537,
|
| 956 |
+
"eval_runtime": 546.1434,
|
| 957 |
+
"eval_samples_per_second": 3.662,
|
| 958 |
+
"eval_steps_per_second": 0.916,
|
| 959 |
+
"step": 600
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 0.65,
|
| 963 |
+
"learning_rate": 1.6376580782162172e-06,
|
| 964 |
+
"logits/chosen": -1.03734290599823,
|
| 965 |
+
"logits/rejected": -0.9538782238960266,
|
| 966 |
+
"logps/chosen": -302.88079833984375,
|
| 967 |
+
"logps/rejected": -275.7502136230469,
|
| 968 |
+
"loss": 0.07,
|
| 969 |
+
"rewards/accuracies": 0.4000000059604645,
|
| 970 |
+
"rewards/chosen": -0.003554628463461995,
|
| 971 |
+
"rewards/margins": 0.016584644094109535,
|
| 972 |
+
"rewards/rejected": -0.020139271393418312,
|
| 973 |
+
"step": 610
|
| 974 |
+
},
|
| 975 |
+
{
|
| 976 |
+
"epoch": 0.66,
|
| 977 |
+
"learning_rate": 1.550828065408227e-06,
|
| 978 |
+
"logits/chosen": -1.0758119821548462,
|
| 979 |
+
"logits/rejected": -0.9793124198913574,
|
| 980 |
+
"logps/chosen": -256.09783935546875,
|
| 981 |
+
"logps/rejected": -237.6195831298828,
|
| 982 |
+
"loss": 0.0857,
|
| 983 |
+
"rewards/accuracies": 0.36250001192092896,
|
| 984 |
+
"rewards/chosen": -0.004253728315234184,
|
| 985 |
+
"rewards/margins": 0.020320799201726913,
|
| 986 |
+
"rewards/rejected": -0.024574527516961098,
|
| 987 |
+
"step": 620
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"epoch": 0.67,
|
| 991 |
+
"learning_rate": 1.4653161248633053e-06,
|
| 992 |
+
"logits/chosen": -1.1112511157989502,
|
| 993 |
+
"logits/rejected": -1.0120253562927246,
|
| 994 |
+
"logps/chosen": -217.60305786132812,
|
| 995 |
+
"logps/rejected": -254.78884887695312,
|
| 996 |
+
"loss": 0.099,
|
| 997 |
+
"rewards/accuracies": 0.3812499940395355,
|
| 998 |
+
"rewards/chosen": -0.007809498347342014,
|
| 999 |
+
"rewards/margins": 0.020094871520996094,
|
| 1000 |
+
"rewards/rejected": -0.027904370799660683,
|
| 1001 |
+
"step": 630
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 0.68,
|
| 1005 |
+
"learning_rate": 1.381241003157162e-06,
|
| 1006 |
+
"logits/chosen": -1.1298284530639648,
|
| 1007 |
+
"logits/rejected": -1.0621023178100586,
|
| 1008 |
+
"logps/chosen": -286.02459716796875,
|
| 1009 |
+
"logps/rejected": -299.3070373535156,
|
| 1010 |
+
"loss": 0.0677,
|
| 1011 |
+
"rewards/accuracies": 0.39375001192092896,
|
| 1012 |
+
"rewards/chosen": -0.011842104606330395,
|
| 1013 |
+
"rewards/margins": 0.014377683401107788,
|
| 1014 |
+
"rewards/rejected": -0.026219788938760757,
|
| 1015 |
+
"step": 640
|
| 1016 |
+
},
|
| 1017 |
+
{
|
| 1018 |
+
"epoch": 0.69,
|
| 1019 |
+
"learning_rate": 1.298719451619979e-06,
|
| 1020 |
+
"logits/chosen": -1.0666725635528564,
|
| 1021 |
+
"logits/rejected": -1.0321277379989624,
|
| 1022 |
+
"logps/chosen": -300.1710205078125,
|
| 1023 |
+
"logps/rejected": -384.95751953125,
|
| 1024 |
+
"loss": 0.073,
|
| 1025 |
+
"rewards/accuracies": 0.5,
|
| 1026 |
+
"rewards/chosen": -0.0033969897776842117,
|
| 1027 |
+
"rewards/margins": 0.039332348853349686,
|
| 1028 |
+
"rewards/rejected": -0.04272934049367905,
|
| 1029 |
+
"step": 650
|
| 1030 |
+
},
|
| 1031 |
+
{
|
| 1032 |
+
"epoch": 0.7,
|
| 1033 |
+
"learning_rate": 1.2178660642091036e-06,
|
| 1034 |
+
"logits/chosen": -1.062753677368164,
|
| 1035 |
+
"logits/rejected": -1.034977674484253,
|
| 1036 |
+
"logps/chosen": -275.3869934082031,
|
| 1037 |
+
"logps/rejected": -275.52667236328125,
|
| 1038 |
+
"loss": 0.0798,
|
| 1039 |
+
"rewards/accuracies": 0.46875,
|
| 1040 |
+
"rewards/chosen": -0.0005760884960182011,
|
| 1041 |
+
"rewards/margins": 0.0214390866458416,
|
| 1042 |
+
"rewards/rejected": -0.022015176713466644,
|
| 1043 |
+
"step": 660
|
| 1044 |
+
},
|
| 1045 |
+
{
|
| 1046 |
+
"epoch": 0.71,
|
| 1047 |
+
"learning_rate": 1.1387931183775821e-06,
|
| 1048 |
+
"logits/chosen": -1.072608470916748,
|
| 1049 |
+
"logits/rejected": -1.0768264532089233,
|
| 1050 |
+
"logps/chosen": -300.6933288574219,
|
| 1051 |
+
"logps/rejected": -286.68951416015625,
|
| 1052 |
+
"loss": 0.0876,
|
| 1053 |
+
"rewards/accuracies": 0.40625,
|
| 1054 |
+
"rewards/chosen": -0.005371665116399527,
|
| 1055 |
+
"rewards/margins": 0.01708284579217434,
|
| 1056 |
+
"rewards/rejected": -0.022454511374235153,
|
| 1057 |
+
"step": 670
|
| 1058 |
+
},
|
| 1059 |
+
{
|
| 1060 |
+
"epoch": 0.73,
|
| 1061 |
+
"learning_rate": 1.061610419159532e-06,
|
| 1062 |
+
"logits/chosen": -1.1137611865997314,
|
| 1063 |
+
"logits/rejected": -1.0089493989944458,
|
| 1064 |
+
"logps/chosen": -284.4278259277344,
|
| 1065 |
+
"logps/rejected": -263.4560546875,
|
| 1066 |
+
"loss": 0.0803,
|
| 1067 |
+
"rewards/accuracies": 0.4000000059604645,
|
| 1068 |
+
"rewards/chosen": -0.0019712348002940416,
|
| 1069 |
+
"rewards/margins": 0.015994885936379433,
|
| 1070 |
+
"rewards/rejected": -0.017966121435165405,
|
| 1071 |
+
"step": 680
|
| 1072 |
+
},
|
| 1073 |
+
{
|
| 1074 |
+
"epoch": 0.74,
|
| 1075 |
+
"learning_rate": 9.864251466888364e-07,
|
| 1076 |
+
"logits/chosen": -1.0998120307922363,
|
| 1077 |
+
"logits/rejected": -1.0427272319793701,
|
| 1078 |
+
"logps/chosen": -273.5634460449219,
|
| 1079 |
+
"logps/rejected": -273.6974792480469,
|
| 1080 |
+
"loss": 0.0779,
|
| 1081 |
+
"rewards/accuracies": 0.40625,
|
| 1082 |
+
"rewards/chosen": -0.00520918658003211,
|
| 1083 |
+
"rewards/margins": 0.018212206661701202,
|
| 1084 |
+
"rewards/rejected": -0.02342139557003975,
|
| 1085 |
+
"step": 690
|
| 1086 |
+
},
|
| 1087 |
+
{
|
| 1088 |
+
"epoch": 0.75,
|
| 1089 |
+
"learning_rate": 9.133417073629288e-07,
|
| 1090 |
+
"logits/chosen": -1.0788428783416748,
|
| 1091 |
+
"logits/rejected": -1.0710010528564453,
|
| 1092 |
+
"logps/chosen": -295.2659912109375,
|
| 1093 |
+
"logps/rejected": -301.9012145996094,
|
| 1094 |
+
"loss": 0.0519,
|
| 1095 |
+
"rewards/accuracies": 0.45625001192092896,
|
| 1096 |
+
"rewards/chosen": -0.006222005933523178,
|
| 1097 |
+
"rewards/margins": 0.023821452632546425,
|
| 1098 |
+
"rewards/rejected": -0.030043456703424454,
|
| 1099 |
+
"step": 700
|
| 1100 |
+
},
|
| 1101 |
+
{
|
| 1102 |
+
"epoch": 0.75,
|
| 1103 |
+
"eval_logits/chosen": -1.0359783172607422,
|
| 1104 |
+
"eval_logits/rejected": -0.9871743321418762,
|
| 1105 |
+
"eval_logps/chosen": -404.5330505371094,
|
| 1106 |
+
"eval_logps/rejected": -404.9790954589844,
|
| 1107 |
+
"eval_loss": 0.04230288788676262,
|
| 1108 |
+
"eval_rewards/accuracies": 0.590499997138977,
|
| 1109 |
+
"eval_rewards/chosen": -0.008227824233472347,
|
| 1110 |
+
"eval_rewards/margins": 0.02951584756374359,
|
| 1111 |
+
"eval_rewards/rejected": -0.03774367272853851,
|
| 1112 |
+
"eval_runtime": 545.9714,
|
| 1113 |
+
"eval_samples_per_second": 3.663,
|
| 1114 |
+
"eval_steps_per_second": 0.916,
|
| 1115 |
+
"step": 700
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 0.76,
|
| 1119 |
+
"learning_rate": 8.424615888583332e-07,
|
| 1120 |
+
"logits/chosen": -1.0752325057983398,
|
| 1121 |
+
"logits/rejected": -1.005172848701477,
|
| 1122 |
+
"logps/chosen": -263.06549072265625,
|
| 1123 |
+
"logps/rejected": -263.9055480957031,
|
| 1124 |
+
"loss": 0.0702,
|
| 1125 |
+
"rewards/accuracies": 0.45625001192092896,
|
| 1126 |
+
"rewards/chosen": -0.0012165942462161183,
|
| 1127 |
+
"rewards/margins": 0.0245666466653347,
|
| 1128 |
+
"rewards/rejected": -0.025783240795135498,
|
| 1129 |
+
"step": 710
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 0.77,
|
| 1133 |
+
"learning_rate": 7.738832191993092e-07,
|
| 1134 |
+
"logits/chosen": -1.0686012506484985,
|
| 1135 |
+
"logits/rejected": -1.0680710077285767,
|
| 1136 |
+
"logps/chosen": -260.68853759765625,
|
| 1137 |
+
"logps/rejected": -299.7347717285156,
|
| 1138 |
+
"loss": 0.0669,
|
| 1139 |
+
"rewards/accuracies": 0.375,
|
| 1140 |
+
"rewards/chosen": -0.006182204931974411,
|
| 1141 |
+
"rewards/margins": 0.016528166830539703,
|
| 1142 |
+
"rewards/rejected": -0.022710371762514114,
|
| 1143 |
+
"step": 720
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.78,
|
| 1147 |
+
"learning_rate": 7.077018300752917e-07,
|
| 1148 |
+
"logits/chosen": -1.1110173463821411,
|
| 1149 |
+
"logits/rejected": -1.0842745304107666,
|
| 1150 |
+
"logps/chosen": -270.8677673339844,
|
| 1151 |
+
"logps/rejected": -273.29339599609375,
|
| 1152 |
+
"loss": 0.07,
|
| 1153 |
+
"rewards/accuracies": 0.4312500059604645,
|
| 1154 |
+
"rewards/chosen": -0.0033080249559134245,
|
| 1155 |
+
"rewards/margins": 0.02419392392039299,
|
| 1156 |
+
"rewards/rejected": -0.027501946315169334,
|
| 1157 |
+
"step": 730
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 0.79,
|
| 1161 |
+
"learning_rate": 6.440093245969342e-07,
|
| 1162 |
+
"logits/chosen": -1.1270530223846436,
|
| 1163 |
+
"logits/rejected": -1.0608434677124023,
|
| 1164 |
+
"logps/chosen": -296.01556396484375,
|
| 1165 |
+
"logps/rejected": -293.2162170410156,
|
| 1166 |
+
"loss": 0.0753,
|
| 1167 |
+
"rewards/accuracies": 0.41874998807907104,
|
| 1168 |
+
"rewards/chosen": -0.004587044008076191,
|
| 1169 |
+
"rewards/margins": 0.017461195588111877,
|
| 1170 |
+
"rewards/rejected": -0.022048238664865494,
|
| 1171 |
+
"step": 740
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 0.8,
|
| 1175 |
+
"learning_rate": 5.828941496744075e-07,
|
| 1176 |
+
"logits/chosen": -1.1132383346557617,
|
| 1177 |
+
"logits/rejected": -1.0976136922836304,
|
| 1178 |
+
"logps/chosen": -258.3694763183594,
|
| 1179 |
+
"logps/rejected": -271.54095458984375,
|
| 1180 |
+
"loss": 0.0907,
|
| 1181 |
+
"rewards/accuracies": 0.38749998807907104,
|
| 1182 |
+
"rewards/chosen": -0.012257089838385582,
|
| 1183 |
+
"rewards/margins": 0.019415555521845818,
|
| 1184 |
+
"rewards/rejected": -0.0316726490855217,
|
| 1185 |
+
"step": 750
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 0.81,
|
| 1189 |
+
"learning_rate": 5.244411731951671e-07,
|
| 1190 |
+
"logits/chosen": -1.1247339248657227,
|
| 1191 |
+
"logits/rejected": -1.0423662662506104,
|
| 1192 |
+
"logps/chosen": -287.46160888671875,
|
| 1193 |
+
"logps/rejected": -290.23736572265625,
|
| 1194 |
+
"loss": 0.0781,
|
| 1195 |
+
"rewards/accuracies": 0.39375001192092896,
|
| 1196 |
+
"rewards/chosen": -0.008325648494064808,
|
| 1197 |
+
"rewards/margins": 0.021062636747956276,
|
| 1198 |
+
"rewards/rejected": -0.02938828244805336,
|
| 1199 |
+
"step": 760
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 0.82,
|
| 1203 |
+
"learning_rate": 4.6873156617173594e-07,
|
| 1204 |
+
"logits/chosen": -1.1069273948669434,
|
| 1205 |
+
"logits/rejected": -1.086753010749817,
|
| 1206 |
+
"logps/chosen": -298.63433837890625,
|
| 1207 |
+
"logps/rejected": -294.5317077636719,
|
| 1208 |
+
"loss": 0.0679,
|
| 1209 |
+
"rewards/accuracies": 0.4437499940395355,
|
| 1210 |
+
"rewards/chosen": -0.007812710478901863,
|
| 1211 |
+
"rewards/margins": 0.02419663593173027,
|
| 1212 |
+
"rewards/rejected": -0.032009344547986984,
|
| 1213 |
+
"step": 770
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 0.83,
|
| 1217 |
+
"learning_rate": 4.1584269002318653e-07,
|
| 1218 |
+
"logits/chosen": -1.1254509687423706,
|
| 1219 |
+
"logits/rejected": -1.0976530313491821,
|
| 1220 |
+
"logps/chosen": -287.83282470703125,
|
| 1221 |
+
"logps/rejected": -289.3851623535156,
|
| 1222 |
+
"loss": 0.0738,
|
| 1223 |
+
"rewards/accuracies": 0.4437499940395355,
|
| 1224 |
+
"rewards/chosen": -0.009489515796303749,
|
| 1225 |
+
"rewards/margins": 0.02022114023566246,
|
| 1226 |
+
"rewards/rejected": -0.029710659757256508,
|
| 1227 |
+
"step": 780
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 0.84,
|
| 1231 |
+
"learning_rate": 3.658479891468258e-07,
|
| 1232 |
+
"logits/chosen": -1.1135103702545166,
|
| 1233 |
+
"logits/rejected": -1.0031986236572266,
|
| 1234 |
+
"logps/chosen": -292.74835205078125,
|
| 1235 |
+
"logps/rejected": -297.3404846191406,
|
| 1236 |
+
"loss": 0.0784,
|
| 1237 |
+
"rewards/accuracies": 0.3687500059604645,
|
| 1238 |
+
"rewards/chosen": -0.003799052909016609,
|
| 1239 |
+
"rewards/margins": 0.016090305522084236,
|
| 1240 |
+
"rewards/rejected": -0.019889358431100845,
|
| 1241 |
+
"step": 790
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 0.85,
|
| 1245 |
+
"learning_rate": 3.18816888929272e-07,
|
| 1246 |
+
"logits/chosen": -1.0912225246429443,
|
| 1247 |
+
"logits/rejected": -1.0440576076507568,
|
| 1248 |
+
"logps/chosen": -273.9382019042969,
|
| 1249 |
+
"logps/rejected": -279.7242126464844,
|
| 1250 |
+
"loss": 0.0742,
|
| 1251 |
+
"rewards/accuracies": 0.4437499940395355,
|
| 1252 |
+
"rewards/chosen": -0.00514855096116662,
|
| 1253 |
+
"rewards/margins": 0.019596170634031296,
|
| 1254 |
+
"rewards/rejected": -0.024744722992181778,
|
| 1255 |
+
"step": 800
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 0.85,
|
| 1259 |
+
"eval_logits/chosen": -1.055617332458496,
|
| 1260 |
+
"eval_logits/rejected": -1.0109219551086426,
|
| 1261 |
+
"eval_logps/chosen": -406.80352783203125,
|
| 1262 |
+
"eval_logps/rejected": -409.2462158203125,
|
| 1263 |
+
"eval_loss": 0.04217638820409775,
|
| 1264 |
+
"eval_rewards/accuracies": 0.6000000238418579,
|
| 1265 |
+
"eval_rewards/chosen": -0.010498268529772758,
|
| 1266 |
+
"eval_rewards/margins": 0.0315125547349453,
|
| 1267 |
+
"eval_rewards/rejected": -0.042010821402072906,
|
| 1268 |
+
"eval_runtime": 546.1165,
|
| 1269 |
+
"eval_samples_per_second": 3.662,
|
| 1270 |
+
"eval_steps_per_second": 0.916,
|
| 1271 |
+
"step": 800
|
| 1272 |
+
},
|
| 1273 |
+
{
|
| 1274 |
+
"epoch": 0.86,
|
| 1275 |
+
"learning_rate": 2.748146993385484e-07,
|
| 1276 |
+
"logits/chosen": -1.118160367012024,
|
| 1277 |
+
"logits/rejected": -1.0629098415374756,
|
| 1278 |
+
"logps/chosen": -223.4706573486328,
|
| 1279 |
+
"logps/rejected": -247.4434356689453,
|
| 1280 |
+
"loss": 0.0963,
|
| 1281 |
+
"rewards/accuracies": 0.36250001192092896,
|
| 1282 |
+
"rewards/chosen": -0.0038195624947547913,
|
| 1283 |
+
"rewards/margins": 0.02616865560412407,
|
| 1284 |
+
"rewards/rejected": -0.029988214373588562,
|
| 1285 |
+
"step": 810
|
| 1286 |
+
},
|
| 1287 |
+
{
|
| 1288 |
+
"epoch": 0.87,
|
| 1289 |
+
"learning_rate": 2.3390252423108077e-07,
|
| 1290 |
+
"logits/chosen": -1.1013834476470947,
|
| 1291 |
+
"logits/rejected": -1.1045656204223633,
|
| 1292 |
+
"logps/chosen": -240.3532257080078,
|
| 1293 |
+
"logps/rejected": -258.44989013671875,
|
| 1294 |
+
"loss": 0.0812,
|
| 1295 |
+
"rewards/accuracies": 0.41874998807907104,
|
| 1296 |
+
"rewards/chosen": -0.002859788713976741,
|
| 1297 |
+
"rewards/margins": 0.029717862606048584,
|
| 1298 |
+
"rewards/rejected": -0.03257765248417854,
|
| 1299 |
+
"step": 820
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 0.89,
|
| 1303 |
+
"learning_rate": 1.961371764995243e-07,
|
| 1304 |
+
"logits/chosen": -1.1316301822662354,
|
| 1305 |
+
"logits/rejected": -1.1215277910232544,
|
| 1306 |
+
"logps/chosen": -266.54931640625,
|
| 1307 |
+
"logps/rejected": -260.0931701660156,
|
| 1308 |
+
"loss": 0.0782,
|
| 1309 |
+
"rewards/accuracies": 0.39375001192092896,
|
| 1310 |
+
"rewards/chosen": -0.006319983396679163,
|
| 1311 |
+
"rewards/margins": 0.014086413197219372,
|
| 1312 |
+
"rewards/rejected": -0.020406395196914673,
|
| 1313 |
+
"step": 830
|
| 1314 |
+
},
|
| 1315 |
+
{
|
| 1316 |
+
"epoch": 0.9,
|
| 1317 |
+
"learning_rate": 1.61571099179261e-07,
|
| 1318 |
+
"logits/chosen": -1.1706424951553345,
|
| 1319 |
+
"logits/rejected": -1.0354385375976562,
|
| 1320 |
+
"logps/chosen": -278.2705993652344,
|
| 1321 |
+
"logps/rejected": -276.22967529296875,
|
| 1322 |
+
"loss": 0.0814,
|
| 1323 |
+
"rewards/accuracies": 0.4000000059604645,
|
| 1324 |
+
"rewards/chosen": -0.009933208115398884,
|
| 1325 |
+
"rewards/margins": 0.019498441368341446,
|
| 1326 |
+
"rewards/rejected": -0.029431650415062904,
|
| 1327 |
+
"step": 840
|
| 1328 |
+
},
|
| 1329 |
+
{
|
| 1330 |
+
"epoch": 0.91,
|
| 1331 |
+
"learning_rate": 1.3025229262312367e-07,
|
| 1332 |
+
"logits/chosen": -1.0535192489624023,
|
| 1333 |
+
"logits/rejected": -1.0249392986297607,
|
| 1334 |
+
"logps/chosen": -238.9296875,
|
| 1335 |
+
"logps/rejected": -260.4687805175781,
|
| 1336 |
+
"loss": 0.0732,
|
| 1337 |
+
"rewards/accuracies": 0.4000000059604645,
|
| 1338 |
+
"rewards/chosen": -0.0005777518963441253,
|
| 1339 |
+
"rewards/margins": 0.021599723026156425,
|
| 1340 |
+
"rewards/rejected": -0.022177476435899734,
|
| 1341 |
+
"step": 850
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 0.92,
|
| 1345 |
+
"learning_rate": 1.0222424784546853e-07,
|
| 1346 |
+
"logits/chosen": -1.1237046718597412,
|
| 1347 |
+
"logits/rejected": -1.1744358539581299,
|
| 1348 |
+
"logps/chosen": -290.61187744140625,
|
| 1349 |
+
"logps/rejected": -289.50775146484375,
|
| 1350 |
+
"loss": 0.0682,
|
| 1351 |
+
"rewards/accuracies": 0.45625001192092896,
|
| 1352 |
+
"rewards/chosen": -0.00292245764285326,
|
| 1353 |
+
"rewards/margins": 0.022101474925875664,
|
| 1354 |
+
"rewards/rejected": -0.0250239335000515,
|
| 1355 |
+
"step": 860
|
| 1356 |
+
},
|
| 1357 |
+
{
|
| 1358 |
+
"epoch": 0.93,
|
| 1359 |
+
"learning_rate": 7.752588612816553e-08,
|
| 1360 |
+
"logits/chosen": -1.0622873306274414,
|
| 1361 |
+
"logits/rejected": -1.047603964805603,
|
| 1362 |
+
"logps/chosen": -290.8896484375,
|
| 1363 |
+
"logps/rejected": -276.6893005371094,
|
| 1364 |
+
"loss": 0.0607,
|
| 1365 |
+
"rewards/accuracies": 0.4000000059604645,
|
| 1366 |
+
"rewards/chosen": -0.005569613538682461,
|
| 1367 |
+
"rewards/margins": 0.024782858788967133,
|
| 1368 |
+
"rewards/rejected": -0.030352476984262466,
|
| 1369 |
+
"step": 870
|
| 1370 |
+
},
|
| 1371 |
+
{
|
| 1372 |
+
"epoch": 0.94,
|
| 1373 |
+
"learning_rate": 5.619150497236991e-08,
|
| 1374 |
+
"logits/chosen": -1.1374906301498413,
|
| 1375 |
+
"logits/rejected": -1.079929232597351,
|
| 1376 |
+
"logps/chosen": -223.29623413085938,
|
| 1377 |
+
"logps/rejected": -222.3146209716797,
|
| 1378 |
+
"loss": 0.0661,
|
| 1379 |
+
"rewards/accuracies": 0.3687500059604645,
|
| 1380 |
+
"rewards/chosen": -0.0003207772097084671,
|
| 1381 |
+
"rewards/margins": 0.016658511012792587,
|
| 1382 |
+
"rewards/rejected": -0.016979288309812546,
|
| 1383 |
+
"step": 880
|
| 1384 |
+
},
|
| 1385 |
+
{
|
| 1386 |
+
"epoch": 0.95,
|
| 1387 |
+
"learning_rate": 3.825073047112743e-08,
|
| 1388 |
+
"logits/chosen": -1.0372432470321655,
|
| 1389 |
+
"logits/rejected": -1.0072650909423828,
|
| 1390 |
+
"logps/chosen": -297.5245056152344,
|
| 1391 |
+
"logps/rejected": -319.53076171875,
|
| 1392 |
+
"loss": 0.0776,
|
| 1393 |
+
"rewards/accuracies": 0.45625001192092896,
|
| 1394 |
+
"rewards/chosen": -0.007688297424465418,
|
| 1395 |
+
"rewards/margins": 0.022589299827814102,
|
| 1396 |
+
"rewards/rejected": -0.030277591198682785,
|
| 1397 |
+
"step": 890
|
| 1398 |
+
},
|
| 1399 |
+
{
|
| 1400 |
+
"epoch": 0.96,
|
| 1401 |
+
"learning_rate": 2.372847616895685e-08,
|
| 1402 |
+
"logits/chosen": -1.096644639968872,
|
| 1403 |
+
"logits/rejected": -1.0533974170684814,
|
| 1404 |
+
"logps/chosen": -268.70343017578125,
|
| 1405 |
+
"logps/rejected": -255.3015594482422,
|
| 1406 |
+
"loss": 0.0768,
|
| 1407 |
+
"rewards/accuracies": 0.41874998807907104,
|
| 1408 |
+
"rewards/chosen": -0.003904106793925166,
|
| 1409 |
+
"rewards/margins": 0.022419685497879982,
|
| 1410 |
+
"rewards/rejected": -0.02632378600537777,
|
| 1411 |
+
"step": 900
|
| 1412 |
+
},
|
| 1413 |
+
{
|
| 1414 |
+
"epoch": 0.96,
|
| 1415 |
+
"eval_logits/chosen": -1.0502386093139648,
|
| 1416 |
+
"eval_logits/rejected": -1.0050266981124878,
|
| 1417 |
+
"eval_logps/chosen": -406.3475341796875,
|
| 1418 |
+
"eval_logps/rejected": -408.7396545410156,
|
| 1419 |
+
"eval_loss": 0.04211420938372612,
|
| 1420 |
+
"eval_rewards/accuracies": 0.5929999947547913,
|
| 1421 |
+
"eval_rewards/chosen": -0.01004225667566061,
|
| 1422 |
+
"eval_rewards/margins": 0.031461965292692184,
|
| 1423 |
+
"eval_rewards/rejected": -0.04150421544909477,
|
| 1424 |
+
"eval_runtime": 545.8034,
|
| 1425 |
+
"eval_samples_per_second": 3.664,
|
| 1426 |
+
"eval_steps_per_second": 0.916,
|
| 1427 |
+
"step": 900
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 0.97,
|
| 1431 |
+
"learning_rate": 1.264490846553279e-08,
|
| 1432 |
+
"logits/chosen": -1.1268881559371948,
|
| 1433 |
+
"logits/rejected": -1.0797778367996216,
|
| 1434 |
+
"logps/chosen": -271.73004150390625,
|
| 1435 |
+
"logps/rejected": -292.3177490234375,
|
| 1436 |
+
"loss": 0.0679,
|
| 1437 |
+
"rewards/accuracies": 0.41874998807907104,
|
| 1438 |
+
"rewards/chosen": -0.008194219321012497,
|
| 1439 |
+
"rewards/margins": 0.016915880143642426,
|
| 1440 |
+
"rewards/rejected": -0.025110099464654922,
|
| 1441 |
+
"step": 910
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 0.98,
|
| 1445 |
+
"learning_rate": 5.015418611516165e-09,
|
| 1446 |
+
"logits/chosen": -1.13059401512146,
|
| 1447 |
+
"logits/rejected": -1.099302887916565,
|
| 1448 |
+
"logps/chosen": -252.9181365966797,
|
| 1449 |
+
"logps/rejected": -257.0358581542969,
|
| 1450 |
+
"loss": 0.0898,
|
| 1451 |
+
"rewards/accuracies": 0.3687500059604645,
|
| 1452 |
+
"rewards/chosen": -0.005635020788758993,
|
| 1453 |
+
"rewards/margins": 0.021289747208356857,
|
| 1454 |
+
"rewards/rejected": -0.026924768462777138,
|
| 1455 |
+
"step": 920
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 0.99,
|
| 1459 |
+
"learning_rate": 8.506013354186993e-10,
|
| 1460 |
+
"logits/chosen": -1.0815865993499756,
|
| 1461 |
+
"logits/rejected": -1.0594186782836914,
|
| 1462 |
+
"logps/chosen": -270.22955322265625,
|
| 1463 |
+
"logps/rejected": -268.7232360839844,
|
| 1464 |
+
"loss": 0.0661,
|
| 1465 |
+
"rewards/accuracies": 0.40625,
|
| 1466 |
+
"rewards/chosen": 0.0016765497857704759,
|
| 1467 |
+
"rewards/margins": 0.02203894779086113,
|
| 1468 |
+
"rewards/rejected": -0.020362399518489838,
|
| 1469 |
+
"step": 930
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 1.0,
|
| 1473 |
+
"step": 937,
|
| 1474 |
+
"total_flos": 0.0,
|
| 1475 |
+
"train_loss": 0.07849642387894454,
|
| 1476 |
+
"train_runtime": 13138.6455,
|
| 1477 |
+
"train_samples_per_second": 1.142,
|
| 1478 |
+
"train_steps_per_second": 0.071
|
| 1479 |
+
}
|
| 1480 |
+
],
|
| 1481 |
+
"logging_steps": 10,
|
| 1482 |
+
"max_steps": 937,
|
| 1483 |
+
"num_input_tokens_seen": 0,
|
| 1484 |
+
"num_train_epochs": 1,
|
| 1485 |
+
"save_steps": 100,
|
| 1486 |
+
"total_flos": 0.0,
|
| 1487 |
+
"train_batch_size": 4,
|
| 1488 |
+
"trial_name": null,
|
| 1489 |
+
"trial_params": null
|
| 1490 |
+
}
|