Commit
•
4843c21
1
Parent(s):
8ee7221
Model save
Browse files- README.md +62 -0
- adapter_model.safetensors +1 -1
- all_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +688 -0
README.md
ADDED
|
@@ -0,0 +1,62 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: mit
|
| 3 |
+
library_name: peft
|
| 4 |
+
tags:
|
| 5 |
+
- trl
|
| 6 |
+
- dpo
|
| 7 |
+
- generated_from_trainer
|
| 8 |
+
base_model: DUAL-GPO/phi-2-gpo-new-i0
|
| 9 |
+
model-index:
|
| 10 |
+
- name: phi-2-gpo-newSFT-b0.001-v10-lightai-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-newSFT-b0.001-v10-lightai-i1
|
| 18 |
+
|
| 19 |
+
This model is a fine-tuned version of [DUAL-GPO/phi-2-gpo-new-i0](https://huggingface.co/DUAL-GPO/phi-2-gpo-new-i0) on the None dataset.
|
| 20 |
+
|
| 21 |
+
## Model description
|
| 22 |
+
|
| 23 |
+
More information needed
|
| 24 |
+
|
| 25 |
+
## Intended uses & limitations
|
| 26 |
+
|
| 27 |
+
More information needed
|
| 28 |
+
|
| 29 |
+
## Training and evaluation data
|
| 30 |
+
|
| 31 |
+
More information needed
|
| 32 |
+
|
| 33 |
+
## Training procedure
|
| 34 |
+
|
| 35 |
+
### Training hyperparameters
|
| 36 |
+
|
| 37 |
+
The following hyperparameters were used during training:
|
| 38 |
+
- learning_rate: 5e-06
|
| 39 |
+
- train_batch_size: 4
|
| 40 |
+
- eval_batch_size: 4
|
| 41 |
+
- seed: 42
|
| 42 |
+
- distributed_type: multi-GPU
|
| 43 |
+
- num_devices: 4
|
| 44 |
+
- gradient_accumulation_steps: 4
|
| 45 |
+
- total_train_batch_size: 64
|
| 46 |
+
- total_eval_batch_size: 16
|
| 47 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 48 |
+
- lr_scheduler_type: cosine
|
| 49 |
+
- lr_scheduler_warmup_ratio: 0.1
|
| 50 |
+
- num_epochs: 1
|
| 51 |
+
|
| 52 |
+
### Training results
|
| 53 |
+
|
| 54 |
+
|
| 55 |
+
|
| 56 |
+
### Framework versions
|
| 57 |
+
|
| 58 |
+
- PEFT 0.7.1
|
| 59 |
+
- Transformers 4.36.2
|
| 60 |
+
- Pytorch 2.2.1+cu121
|
| 61 |
+
- Datasets 2.14.6
|
| 62 |
+
- Tokenizers 0.15.2
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 167807296
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e1019ce9c0826e73fd1b82ceeb02551b4c8dc967b11880df0c2af62bf3fb3140
|
| 3 |
size 167807296
|
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"train_loss": 1.8394590188295414,
|
| 4 |
+
"train_runtime": 15861.5475,
|
| 5 |
+
"train_samples": 61135,
|
| 6 |
+
"train_samples_per_second": 1.891,
|
| 7 |
+
"train_steps_per_second": 0.03
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"train_loss": 1.8394590188295414,
|
| 4 |
+
"train_runtime": 15861.5475,
|
| 5 |
+
"train_samples": 61135,
|
| 6 |
+
"train_samples_per_second": 1.891,
|
| 7 |
+
"train_steps_per_second": 0.03
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,688 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.9984,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 468,
|
| 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": 1.0638297872340426e-07,
|
| 14 |
+
"logits/chosen": 0.1359557956457138,
|
| 15 |
+
"logits/rejected": 0.030706744641065598,
|
| 16 |
+
"logps/chosen": -736.0869140625,
|
| 17 |
+
"logps/rejected": -613.6344604492188,
|
| 18 |
+
"loss": 2.0331,
|
| 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": 0.02,
|
| 27 |
+
"learning_rate": 1.0638297872340427e-06,
|
| 28 |
+
"logits/chosen": 0.11667777597904205,
|
| 29 |
+
"logits/rejected": 0.26604601740837097,
|
| 30 |
+
"logps/chosen": -546.5281982421875,
|
| 31 |
+
"logps/rejected": -597.5736083984375,
|
| 32 |
+
"loss": 2.1592,
|
| 33 |
+
"rewards/accuracies": 0.4583333432674408,
|
| 34 |
+
"rewards/chosen": 0.0007250224007293582,
|
| 35 |
+
"rewards/margins": 0.00040180076030083,
|
| 36 |
+
"rewards/rejected": 0.0003232216986361891,
|
| 37 |
+
"step": 10
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.04,
|
| 41 |
+
"learning_rate": 2.1276595744680853e-06,
|
| 42 |
+
"logits/chosen": 0.16373148560523987,
|
| 43 |
+
"logits/rejected": 0.2677033543586731,
|
| 44 |
+
"logps/chosen": -604.6590576171875,
|
| 45 |
+
"logps/rejected": -649.482177734375,
|
| 46 |
+
"loss": 2.0972,
|
| 47 |
+
"rewards/accuracies": 0.4375,
|
| 48 |
+
"rewards/chosen": -0.0005862273974344134,
|
| 49 |
+
"rewards/margins": -0.0003054165281355381,
|
| 50 |
+
"rewards/rejected": -0.0002808108984027058,
|
| 51 |
+
"step": 20
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.06,
|
| 55 |
+
"learning_rate": 3.191489361702128e-06,
|
| 56 |
+
"logits/chosen": 0.14978544414043427,
|
| 57 |
+
"logits/rejected": 0.1915779411792755,
|
| 58 |
+
"logps/chosen": -594.8548583984375,
|
| 59 |
+
"logps/rejected": -588.2429809570312,
|
| 60 |
+
"loss": 2.122,
|
| 61 |
+
"rewards/accuracies": 0.512499988079071,
|
| 62 |
+
"rewards/chosen": -0.004188057966530323,
|
| 63 |
+
"rewards/margins": 0.0009490737575106323,
|
| 64 |
+
"rewards/rejected": -0.0051371315494179726,
|
| 65 |
+
"step": 30
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.09,
|
| 69 |
+
"learning_rate": 4.255319148936171e-06,
|
| 70 |
+
"logits/chosen": 0.16862796247005463,
|
| 71 |
+
"logits/rejected": 0.23586151003837585,
|
| 72 |
+
"logps/chosen": -574.7235107421875,
|
| 73 |
+
"logps/rejected": -631.8544921875,
|
| 74 |
+
"loss": 2.1863,
|
| 75 |
+
"rewards/accuracies": 0.5375000238418579,
|
| 76 |
+
"rewards/chosen": -0.01229151152074337,
|
| 77 |
+
"rewards/margins": 0.005582691170275211,
|
| 78 |
+
"rewards/rejected": -0.017874203622341156,
|
| 79 |
+
"step": 40
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.11,
|
| 83 |
+
"learning_rate": 4.999373573764188e-06,
|
| 84 |
+
"logits/chosen": 0.1411871314048767,
|
| 85 |
+
"logits/rejected": 0.2258455753326416,
|
| 86 |
+
"logps/chosen": -612.8582763671875,
|
| 87 |
+
"logps/rejected": -636.5026245117188,
|
| 88 |
+
"loss": 2.1508,
|
| 89 |
+
"rewards/accuracies": 0.5375000238418579,
|
| 90 |
+
"rewards/chosen": -0.04941480979323387,
|
| 91 |
+
"rewards/margins": 0.019247086718678474,
|
| 92 |
+
"rewards/rejected": -0.0686618983745575,
|
| 93 |
+
"step": 50
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.13,
|
| 97 |
+
"learning_rate": 4.988245838331339e-06,
|
| 98 |
+
"logits/chosen": 0.17244111001491547,
|
| 99 |
+
"logits/rejected": 0.17342150211334229,
|
| 100 |
+
"logps/chosen": -634.6348266601562,
|
| 101 |
+
"logps/rejected": -667.5384521484375,
|
| 102 |
+
"loss": 2.0758,
|
| 103 |
+
"rewards/accuracies": 0.581250011920929,
|
| 104 |
+
"rewards/chosen": -0.13025899231433868,
|
| 105 |
+
"rewards/margins": 0.05111612752079964,
|
| 106 |
+
"rewards/rejected": -0.18137511610984802,
|
| 107 |
+
"step": 60
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.15,
|
| 111 |
+
"learning_rate": 4.963268819535228e-06,
|
| 112 |
+
"logits/chosen": 0.12650486826896667,
|
| 113 |
+
"logits/rejected": 0.14093999564647675,
|
| 114 |
+
"logps/chosen": -608.5107421875,
|
| 115 |
+
"logps/rejected": -702.1578369140625,
|
| 116 |
+
"loss": 2.0556,
|
| 117 |
+
"rewards/accuracies": 0.5625,
|
| 118 |
+
"rewards/chosen": -0.21341009438037872,
|
| 119 |
+
"rewards/margins": 0.09893321990966797,
|
| 120 |
+
"rewards/rejected": -0.3123432993888855,
|
| 121 |
+
"step": 70
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.17,
|
| 125 |
+
"learning_rate": 4.9245815365216115e-06,
|
| 126 |
+
"logits/chosen": 0.19184628129005432,
|
| 127 |
+
"logits/rejected": 0.2408786565065384,
|
| 128 |
+
"logps/chosen": -679.4183349609375,
|
| 129 |
+
"logps/rejected": -609.7093505859375,
|
| 130 |
+
"loss": 2.1137,
|
| 131 |
+
"rewards/accuracies": 0.5562499761581421,
|
| 132 |
+
"rewards/chosen": -0.21463651955127716,
|
| 133 |
+
"rewards/margins": 0.05772104859352112,
|
| 134 |
+
"rewards/rejected": -0.2723575234413147,
|
| 135 |
+
"step": 80
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.19,
|
| 139 |
+
"learning_rate": 4.872399318152594e-06,
|
| 140 |
+
"logits/chosen": 0.1250939965248108,
|
| 141 |
+
"logits/rejected": 0.18045032024383545,
|
| 142 |
+
"logps/chosen": -622.2333374023438,
|
| 143 |
+
"logps/rejected": -655.4575805664062,
|
| 144 |
+
"loss": 2.0044,
|
| 145 |
+
"rewards/accuracies": 0.606249988079071,
|
| 146 |
+
"rewards/chosen": -0.1839352548122406,
|
| 147 |
+
"rewards/margins": 0.10977420955896378,
|
| 148 |
+
"rewards/rejected": -0.2937094569206238,
|
| 149 |
+
"step": 90
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.21,
|
| 153 |
+
"learning_rate": 4.807012604511542e-06,
|
| 154 |
+
"logits/chosen": 0.18265239894390106,
|
| 155 |
+
"logits/rejected": 0.2614283859729767,
|
| 156 |
+
"logps/chosen": -649.8997802734375,
|
| 157 |
+
"logps/rejected": -658.8975830078125,
|
| 158 |
+
"loss": 1.9995,
|
| 159 |
+
"rewards/accuracies": 0.643750011920929,
|
| 160 |
+
"rewards/chosen": -0.16064395010471344,
|
| 161 |
+
"rewards/margins": 0.08805385231971741,
|
| 162 |
+
"rewards/rejected": -0.24869783222675323,
|
| 163 |
+
"step": 100
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.23,
|
| 167 |
+
"learning_rate": 4.728785330347771e-06,
|
| 168 |
+
"logits/chosen": 0.2479465901851654,
|
| 169 |
+
"logits/rejected": 0.2932817339897156,
|
| 170 |
+
"logps/chosen": -674.0836181640625,
|
| 171 |
+
"logps/rejected": -645.6417236328125,
|
| 172 |
+
"loss": 1.895,
|
| 173 |
+
"rewards/accuracies": 0.606249988079071,
|
| 174 |
+
"rewards/chosen": -0.12688389420509338,
|
| 175 |
+
"rewards/margins": 0.08782283961772919,
|
| 176 |
+
"rewards/rejected": -0.21470670402050018,
|
| 177 |
+
"step": 110
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.26,
|
| 181 |
+
"learning_rate": 4.63815289945858e-06,
|
| 182 |
+
"logits/chosen": 0.19643843173980713,
|
| 183 |
+
"logits/rejected": 0.2974274456501007,
|
| 184 |
+
"logps/chosen": -573.49658203125,
|
| 185 |
+
"logps/rejected": -666.606689453125,
|
| 186 |
+
"loss": 1.89,
|
| 187 |
+
"rewards/accuracies": 0.6812499761581421,
|
| 188 |
+
"rewards/chosen": -0.09824337065219879,
|
| 189 |
+
"rewards/margins": 0.13982543349266052,
|
| 190 |
+
"rewards/rejected": -0.2380688190460205,
|
| 191 |
+
"step": 120
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.28,
|
| 195 |
+
"learning_rate": 4.535619761282989e-06,
|
| 196 |
+
"logits/chosen": 0.23821644484996796,
|
| 197 |
+
"logits/rejected": 0.288485586643219,
|
| 198 |
+
"logps/chosen": -590.9158935546875,
|
| 199 |
+
"logps/rejected": -623.23974609375,
|
| 200 |
+
"loss": 1.9389,
|
| 201 |
+
"rewards/accuracies": 0.612500011920929,
|
| 202 |
+
"rewards/chosen": -0.14589470624923706,
|
| 203 |
+
"rewards/margins": 0.12624357640743256,
|
| 204 |
+
"rewards/rejected": -0.2721382975578308,
|
| 205 |
+
"step": 130
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.3,
|
| 209 |
+
"learning_rate": 4.42175660319555e-06,
|
| 210 |
+
"logits/chosen": 0.2631734013557434,
|
| 211 |
+
"logits/rejected": 0.2810806632041931,
|
| 212 |
+
"logps/chosen": -645.8680419921875,
|
| 213 |
+
"logps/rejected": -654.8004760742188,
|
| 214 |
+
"loss": 1.8203,
|
| 215 |
+
"rewards/accuracies": 0.6187499761581421,
|
| 216 |
+
"rewards/chosen": -0.20492109656333923,
|
| 217 |
+
"rewards/margins": 0.20386295020580292,
|
| 218 |
+
"rewards/rejected": -0.40878406167030334,
|
| 219 |
+
"step": 140
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.32,
|
| 223 |
+
"learning_rate": 4.297197174127619e-06,
|
| 224 |
+
"logits/chosen": 0.2586398422718048,
|
| 225 |
+
"logits/rejected": 0.3086986839771271,
|
| 226 |
+
"logps/chosen": -619.4220581054688,
|
| 227 |
+
"logps/rejected": -697.2005615234375,
|
| 228 |
+
"loss": 1.7553,
|
| 229 |
+
"rewards/accuracies": 0.6187499761581421,
|
| 230 |
+
"rewards/chosen": -0.24971242249011993,
|
| 231 |
+
"rewards/margins": 0.2221045196056366,
|
| 232 |
+
"rewards/rejected": -0.4718169569969177,
|
| 233 |
+
"step": 150
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.34,
|
| 237 |
+
"learning_rate": 4.162634757195418e-06,
|
| 238 |
+
"logits/chosen": 0.2681664526462555,
|
| 239 |
+
"logits/rejected": 0.2807798683643341,
|
| 240 |
+
"logps/chosen": -630.39306640625,
|
| 241 |
+
"logps/rejected": -645.6117553710938,
|
| 242 |
+
"loss": 1.8404,
|
| 243 |
+
"rewards/accuracies": 0.612500011920929,
|
| 244 |
+
"rewards/chosen": -0.19384464621543884,
|
| 245 |
+
"rewards/margins": 0.1983100175857544,
|
| 246 |
+
"rewards/rejected": -0.39215466380119324,
|
| 247 |
+
"step": 160
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.36,
|
| 251 |
+
"learning_rate": 4.018818310967843e-06,
|
| 252 |
+
"logits/chosen": 0.27496370673179626,
|
| 253 |
+
"logits/rejected": 0.30781346559524536,
|
| 254 |
+
"logps/chosen": -559.2887573242188,
|
| 255 |
+
"logps/rejected": -601.3917846679688,
|
| 256 |
+
"loss": 1.8382,
|
| 257 |
+
"rewards/accuracies": 0.637499988079071,
|
| 258 |
+
"rewards/chosen": -0.0747746005654335,
|
| 259 |
+
"rewards/margins": 0.19791939854621887,
|
| 260 |
+
"rewards/rejected": -0.2726939916610718,
|
| 261 |
+
"step": 170
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.38,
|
| 265 |
+
"learning_rate": 3.866548300851254e-06,
|
| 266 |
+
"logits/chosen": 0.2482290267944336,
|
| 267 |
+
"logits/rejected": 0.2852781414985657,
|
| 268 |
+
"logps/chosen": -620.8068237304688,
|
| 269 |
+
"logps/rejected": -665.9005737304688,
|
| 270 |
+
"loss": 1.8229,
|
| 271 |
+
"rewards/accuracies": 0.625,
|
| 272 |
+
"rewards/chosen": -0.10497160255908966,
|
| 273 |
+
"rewards/margins": 0.20543234050273895,
|
| 274 |
+
"rewards/rejected": -0.3104039430618286,
|
| 275 |
+
"step": 180
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.41,
|
| 279 |
+
"learning_rate": 3.706672243793271e-06,
|
| 280 |
+
"logits/chosen": 0.2958913743495941,
|
| 281 |
+
"logits/rejected": 0.3795389235019684,
|
| 282 |
+
"logps/chosen": -611.8587646484375,
|
| 283 |
+
"logps/rejected": -658.9635009765625,
|
| 284 |
+
"loss": 1.7752,
|
| 285 |
+
"rewards/accuracies": 0.668749988079071,
|
| 286 |
+
"rewards/chosen": -0.0870656967163086,
|
| 287 |
+
"rewards/margins": 0.23995642364025116,
|
| 288 |
+
"rewards/rejected": -0.32702213525772095,
|
| 289 |
+
"step": 190
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.43,
|
| 293 |
+
"learning_rate": 3.5400799911032357e-06,
|
| 294 |
+
"logits/chosen": 0.2935205101966858,
|
| 295 |
+
"logits/rejected": 0.3416239321231842,
|
| 296 |
+
"logps/chosen": -660.2877197265625,
|
| 297 |
+
"logps/rejected": -730.04541015625,
|
| 298 |
+
"loss": 1.7351,
|
| 299 |
+
"rewards/accuracies": 0.668749988079071,
|
| 300 |
+
"rewards/chosen": -0.1828436255455017,
|
| 301 |
+
"rewards/margins": 0.3010478913784027,
|
| 302 |
+
"rewards/rejected": -0.4838915765285492,
|
| 303 |
+
"step": 200
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.45,
|
| 307 |
+
"learning_rate": 3.3676987756445894e-06,
|
| 308 |
+
"logits/chosen": 0.24807122349739075,
|
| 309 |
+
"logits/rejected": 0.32862648367881775,
|
| 310 |
+
"logps/chosen": -605.8773193359375,
|
| 311 |
+
"logps/rejected": -641.6677856445312,
|
| 312 |
+
"loss": 1.8245,
|
| 313 |
+
"rewards/accuracies": 0.7124999761581421,
|
| 314 |
+
"rewards/chosen": -0.13868902623653412,
|
| 315 |
+
"rewards/margins": 0.2735101878643036,
|
| 316 |
+
"rewards/rejected": -0.4121991991996765,
|
| 317 |
+
"step": 210
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.47,
|
| 321 |
+
"learning_rate": 3.1904880509659397e-06,
|
| 322 |
+
"logits/chosen": 0.270724892616272,
|
| 323 |
+
"logits/rejected": 0.3151053786277771,
|
| 324 |
+
"logps/chosen": -650.7314453125,
|
| 325 |
+
"logps/rejected": -708.2312622070312,
|
| 326 |
+
"loss": 1.735,
|
| 327 |
+
"rewards/accuracies": 0.637499988079071,
|
| 328 |
+
"rewards/chosen": -0.16527561843395233,
|
| 329 |
+
"rewards/margins": 0.2484448254108429,
|
| 330 |
+
"rewards/rejected": -0.4137204587459564,
|
| 331 |
+
"step": 220
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.49,
|
| 335 |
+
"learning_rate": 3.0094341510955697e-06,
|
| 336 |
+
"logits/chosen": 0.19233042001724243,
|
| 337 |
+
"logits/rejected": 0.29483872652053833,
|
| 338 |
+
"logps/chosen": -663.5474243164062,
|
| 339 |
+
"logps/rejected": -743.0173950195312,
|
| 340 |
+
"loss": 1.7378,
|
| 341 |
+
"rewards/accuracies": 0.7562500238418579,
|
| 342 |
+
"rewards/chosen": -0.14797742664813995,
|
| 343 |
+
"rewards/margins": 0.3706679344177246,
|
| 344 |
+
"rewards/rejected": -0.5186454057693481,
|
| 345 |
+
"step": 230
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.51,
|
| 349 |
+
"learning_rate": 2.825544800722376e-06,
|
| 350 |
+
"logits/chosen": 0.2124979943037033,
|
| 351 |
+
"logits/rejected": 0.3365432620048523,
|
| 352 |
+
"logps/chosen": -619.9740600585938,
|
| 353 |
+
"logps/rejected": -700.7166748046875,
|
| 354 |
+
"loss": 1.8168,
|
| 355 |
+
"rewards/accuracies": 0.5874999761581421,
|
| 356 |
+
"rewards/chosen": -0.1956629902124405,
|
| 357 |
+
"rewards/margins": 0.2987174093723297,
|
| 358 |
+
"rewards/rejected": -0.494380384683609,
|
| 359 |
+
"step": 240
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.53,
|
| 363 |
+
"learning_rate": 2.639843506318899e-06,
|
| 364 |
+
"logits/chosen": 0.2796134054660797,
|
| 365 |
+
"logits/rejected": 0.2740449607372284,
|
| 366 |
+
"logps/chosen": -582.3416748046875,
|
| 367 |
+
"logps/rejected": -674.327880859375,
|
| 368 |
+
"loss": 1.8901,
|
| 369 |
+
"rewards/accuracies": 0.6187499761581421,
|
| 370 |
+
"rewards/chosen": -0.19822832942008972,
|
| 371 |
+
"rewards/margins": 0.19228845834732056,
|
| 372 |
+
"rewards/rejected": -0.3905167877674103,
|
| 373 |
+
"step": 250
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.55,
|
| 377 |
+
"learning_rate": 2.4533638594248094e-06,
|
| 378 |
+
"logits/chosen": 0.25897207856178284,
|
| 379 |
+
"logits/rejected": 0.31485193967819214,
|
| 380 |
+
"logps/chosen": -604.8118896484375,
|
| 381 |
+
"logps/rejected": -667.9144897460938,
|
| 382 |
+
"loss": 1.8606,
|
| 383 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 384 |
+
"rewards/chosen": -0.14205999672412872,
|
| 385 |
+
"rewards/margins": 0.28450149297714233,
|
| 386 |
+
"rewards/rejected": -0.42656150460243225,
|
| 387 |
+
"step": 260
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.58,
|
| 391 |
+
"learning_rate": 2.2671437837980943e-06,
|
| 392 |
+
"logits/chosen": 0.22259187698364258,
|
| 393 |
+
"logits/rejected": 0.22855930030345917,
|
| 394 |
+
"logps/chosen": -593.6612548828125,
|
| 395 |
+
"logps/rejected": -673.6566162109375,
|
| 396 |
+
"loss": 1.7486,
|
| 397 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 398 |
+
"rewards/chosen": -0.14823240041732788,
|
| 399 |
+
"rewards/margins": 0.2802043557167053,
|
| 400 |
+
"rewards/rejected": -0.4284366965293884,
|
| 401 |
+
"step": 270
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.6,
|
| 405 |
+
"learning_rate": 2.082219758453629e-06,
|
| 406 |
+
"logits/chosen": 0.2169434130191803,
|
| 407 |
+
"logits/rejected": 0.2703471779823303,
|
| 408 |
+
"logps/chosen": -611.6048583984375,
|
| 409 |
+
"logps/rejected": -682.5806884765625,
|
| 410 |
+
"loss": 1.6556,
|
| 411 |
+
"rewards/accuracies": 0.6625000238418579,
|
| 412 |
+
"rewards/chosen": -0.11671599000692368,
|
| 413 |
+
"rewards/margins": 0.26952022314071655,
|
| 414 |
+
"rewards/rejected": -0.3862362205982208,
|
| 415 |
+
"step": 280
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.62,
|
| 419 |
+
"learning_rate": 1.899621048743019e-06,
|
| 420 |
+
"logits/chosen": 0.22146745026111603,
|
| 421 |
+
"logits/rejected": 0.34733515977859497,
|
| 422 |
+
"logps/chosen": -603.9933471679688,
|
| 423 |
+
"logps/rejected": -673.3649291992188,
|
| 424 |
+
"loss": 1.7238,
|
| 425 |
+
"rewards/accuracies": 0.65625,
|
| 426 |
+
"rewards/chosen": -0.20439250767230988,
|
| 427 |
+
"rewards/margins": 0.2682177424430847,
|
| 428 |
+
"rewards/rejected": -0.4726102352142334,
|
| 429 |
+
"step": 290
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.64,
|
| 433 |
+
"learning_rate": 1.7203639775848423e-06,
|
| 434 |
+
"logits/chosen": 0.19099445641040802,
|
| 435 |
+
"logits/rejected": 0.3011043667793274,
|
| 436 |
+
"logps/chosen": -606.6263427734375,
|
| 437 |
+
"logps/rejected": -639.6136474609375,
|
| 438 |
+
"loss": 1.8381,
|
| 439 |
+
"rewards/accuracies": 0.643750011920929,
|
| 440 |
+
"rewards/chosen": -0.17796705663204193,
|
| 441 |
+
"rewards/margins": 0.23042461276054382,
|
| 442 |
+
"rewards/rejected": -0.40839165449142456,
|
| 443 |
+
"step": 300
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.66,
|
| 447 |
+
"learning_rate": 1.5454462687309445e-06,
|
| 448 |
+
"logits/chosen": 0.2036764919757843,
|
| 449 |
+
"logits/rejected": 0.26239025592803955,
|
| 450 |
+
"logps/chosen": -602.3845825195312,
|
| 451 |
+
"logps/rejected": -666.4627075195312,
|
| 452 |
+
"loss": 1.8042,
|
| 453 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 454 |
+
"rewards/chosen": -0.1518932580947876,
|
| 455 |
+
"rewards/margins": 0.2536298632621765,
|
| 456 |
+
"rewards/rejected": -0.4055231511592865,
|
| 457 |
+
"step": 310
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.68,
|
| 461 |
+
"learning_rate": 1.3758414935535147e-06,
|
| 462 |
+
"logits/chosen": 0.21739721298217773,
|
| 463 |
+
"logits/rejected": 0.2840099334716797,
|
| 464 |
+
"logps/chosen": -636.0455322265625,
|
| 465 |
+
"logps/rejected": -709.1137084960938,
|
| 466 |
+
"loss": 1.65,
|
| 467 |
+
"rewards/accuracies": 0.65625,
|
| 468 |
+
"rewards/chosen": -0.16815349459648132,
|
| 469 |
+
"rewards/margins": 0.29733169078826904,
|
| 470 |
+
"rewards/rejected": -0.465485155582428,
|
| 471 |
+
"step": 320
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.7,
|
| 475 |
+
"learning_rate": 1.2124936522614622e-06,
|
| 476 |
+
"logits/chosen": 0.20938508212566376,
|
| 477 |
+
"logits/rejected": 0.22490420937538147,
|
| 478 |
+
"logps/chosen": -615.7994995117188,
|
| 479 |
+
"logps/rejected": -669.2200927734375,
|
| 480 |
+
"loss": 1.7098,
|
| 481 |
+
"rewards/accuracies": 0.6937500238418579,
|
| 482 |
+
"rewards/chosen": -0.18394342064857483,
|
| 483 |
+
"rewards/margins": 0.31033387780189514,
|
| 484 |
+
"rewards/rejected": -0.49427732825279236,
|
| 485 |
+
"step": 330
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.73,
|
| 489 |
+
"learning_rate": 1.0563119197063934e-06,
|
| 490 |
+
"logits/chosen": 0.23827771842479706,
|
| 491 |
+
"logits/rejected": 0.2663131356239319,
|
| 492 |
+
"logps/chosen": -612.7750244140625,
|
| 493 |
+
"logps/rejected": -685.60107421875,
|
| 494 |
+
"loss": 1.7109,
|
| 495 |
+
"rewards/accuracies": 0.637499988079071,
|
| 496 |
+
"rewards/chosen": -0.19161880016326904,
|
| 497 |
+
"rewards/margins": 0.26392242312431335,
|
| 498 |
+
"rewards/rejected": -0.4555412232875824,
|
| 499 |
+
"step": 340
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.75,
|
| 503 |
+
"learning_rate": 9.081655850224449e-07,
|
| 504 |
+
"logits/chosen": 0.19827114045619965,
|
| 505 |
+
"logits/rejected": 0.2343660295009613,
|
| 506 |
+
"logps/chosen": -628.5892333984375,
|
| 507 |
+
"logps/rejected": -699.3311767578125,
|
| 508 |
+
"loss": 1.6981,
|
| 509 |
+
"rewards/accuracies": 0.675000011920929,
|
| 510 |
+
"rewards/chosen": -0.23514249920845032,
|
| 511 |
+
"rewards/margins": 0.30311545729637146,
|
| 512 |
+
"rewards/rejected": -0.5382579565048218,
|
| 513 |
+
"step": 350
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.77,
|
| 517 |
+
"learning_rate": 7.688792132653111e-07,
|
| 518 |
+
"logits/chosen": 0.19120459258556366,
|
| 519 |
+
"logits/rejected": 0.2861759066581726,
|
| 520 |
+
"logps/chosen": -659.7528076171875,
|
| 521 |
+
"logps/rejected": -748.490234375,
|
| 522 |
+
"loss": 1.6967,
|
| 523 |
+
"rewards/accuracies": 0.675000011920929,
|
| 524 |
+
"rewards/chosen": -0.19031907618045807,
|
| 525 |
+
"rewards/margins": 0.34352895617485046,
|
| 526 |
+
"rewards/rejected": -0.533847987651825,
|
| 527 |
+
"step": 360
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.79,
|
| 531 |
+
"learning_rate": 6.392280559802341e-07,
|
| 532 |
+
"logits/chosen": 0.2406836450099945,
|
| 533 |
+
"logits/rejected": 0.23908407986164093,
|
| 534 |
+
"logps/chosen": -658.35400390625,
|
| 535 |
+
"logps/rejected": -720.8883666992188,
|
| 536 |
+
"loss": 1.7368,
|
| 537 |
+
"rewards/accuracies": 0.625,
|
| 538 |
+
"rewards/chosen": -0.23702308535575867,
|
| 539 |
+
"rewards/margins": 0.24957367777824402,
|
| 540 |
+
"rewards/rejected": -0.48659682273864746,
|
| 541 |
+
"step": 370
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.81,
|
| 545 |
+
"learning_rate": 5.199337362431792e-07,
|
| 546 |
+
"logits/chosen": 0.26719361543655396,
|
| 547 |
+
"logits/rejected": 0.1743316501379013,
|
| 548 |
+
"logps/chosen": -621.3897094726562,
|
| 549 |
+
"logps/rejected": -680.0,
|
| 550 |
+
"loss": 1.7425,
|
| 551 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 552 |
+
"rewards/chosen": -0.22942551970481873,
|
| 553 |
+
"rewards/margins": 0.26667481660842896,
|
| 554 |
+
"rewards/rejected": -0.49610036611557007,
|
| 555 |
+
"step": 380
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.83,
|
| 559 |
+
"learning_rate": 4.1166023219176176e-07,
|
| 560 |
+
"logits/chosen": 0.21561181545257568,
|
| 561 |
+
"logits/rejected": 0.286629855632782,
|
| 562 |
+
"logps/chosen": -654.0867919921875,
|
| 563 |
+
"logps/rejected": -668.467529296875,
|
| 564 |
+
"loss": 1.6798,
|
| 565 |
+
"rewards/accuracies": 0.699999988079071,
|
| 566 |
+
"rewards/chosen": -0.21663355827331543,
|
| 567 |
+
"rewards/margins": 0.2600599527359009,
|
| 568 |
+
"rewards/rejected": -0.4766935408115387,
|
| 569 |
+
"step": 390
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.85,
|
| 573 |
+
"learning_rate": 3.150101814011136e-07,
|
| 574 |
+
"logits/chosen": 0.16323356330394745,
|
| 575 |
+
"logits/rejected": 0.21500280499458313,
|
| 576 |
+
"logps/chosen": -600.4713134765625,
|
| 577 |
+
"logps/rejected": -730.5057983398438,
|
| 578 |
+
"loss": 1.7084,
|
| 579 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 580 |
+
"rewards/chosen": -0.19050315022468567,
|
| 581 |
+
"rewards/margins": 0.28324562311172485,
|
| 582 |
+
"rewards/rejected": -0.47374874353408813,
|
| 583 |
+
"step": 400
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.87,
|
| 587 |
+
"learning_rate": 2.3052152667409289e-07,
|
| 588 |
+
"logits/chosen": 0.1962326616048813,
|
| 589 |
+
"logits/rejected": 0.22506949305534363,
|
| 590 |
+
"logps/chosen": -614.2760009765625,
|
| 591 |
+
"logps/rejected": -675.3383178710938,
|
| 592 |
+
"loss": 1.7679,
|
| 593 |
+
"rewards/accuracies": 0.706250011920929,
|
| 594 |
+
"rewards/chosen": -0.1459427773952484,
|
| 595 |
+
"rewards/margins": 0.3252793252468109,
|
| 596 |
+
"rewards/rejected": -0.4712221026420593,
|
| 597 |
+
"step": 410
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.9,
|
| 601 |
+
"learning_rate": 1.5866452191498488e-07,
|
| 602 |
+
"logits/chosen": 0.20015636086463928,
|
| 603 |
+
"logits/rejected": 0.25162121653556824,
|
| 604 |
+
"logps/chosen": -651.9236450195312,
|
| 605 |
+
"logps/rejected": -707.2882080078125,
|
| 606 |
+
"loss": 1.7514,
|
| 607 |
+
"rewards/accuracies": 0.59375,
|
| 608 |
+
"rewards/chosen": -0.23218846321105957,
|
| 609 |
+
"rewards/margins": 0.2290785312652588,
|
| 610 |
+
"rewards/rejected": -0.46126699447631836,
|
| 611 |
+
"step": 420
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.92,
|
| 615 |
+
"learning_rate": 9.983911475163727e-08,
|
| 616 |
+
"logits/chosen": 0.16698592901229858,
|
| 617 |
+
"logits/rejected": 0.2591376304626465,
|
| 618 |
+
"logps/chosen": -590.045166015625,
|
| 619 |
+
"logps/rejected": -642.6705322265625,
|
| 620 |
+
"loss": 1.8093,
|
| 621 |
+
"rewards/accuracies": 0.6187499761581421,
|
| 622 |
+
"rewards/chosen": -0.17988340556621552,
|
| 623 |
+
"rewards/margins": 0.23005299270153046,
|
| 624 |
+
"rewards/rejected": -0.4099363684654236,
|
| 625 |
+
"step": 430
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.94,
|
| 629 |
+
"learning_rate": 5.437272047405712e-08,
|
| 630 |
+
"logits/chosen": 0.1858983337879181,
|
| 631 |
+
"logits/rejected": 0.3158418536186218,
|
| 632 |
+
"logps/chosen": -559.8682250976562,
|
| 633 |
+
"logps/rejected": -648.7040405273438,
|
| 634 |
+
"loss": 1.7686,
|
| 635 |
+
"rewards/accuracies": 0.6312500238418579,
|
| 636 |
+
"rewards/chosen": -0.19649046659469604,
|
| 637 |
+
"rewards/margins": 0.26454511284828186,
|
| 638 |
+
"rewards/rejected": -0.4610355794429779,
|
| 639 |
+
"step": 440
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.96,
|
| 643 |
+
"learning_rate": 2.251839967945535e-08,
|
| 644 |
+
"logits/chosen": 0.13786078989505768,
|
| 645 |
+
"logits/rejected": 0.2333669662475586,
|
| 646 |
+
"logps/chosen": -645.2703857421875,
|
| 647 |
+
"logps/rejected": -707.0418090820312,
|
| 648 |
+
"loss": 1.6172,
|
| 649 |
+
"rewards/accuracies": 0.6625000238418579,
|
| 650 |
+
"rewards/chosen": -0.17300908267498016,
|
| 651 |
+
"rewards/margins": 0.3292023241519928,
|
| 652 |
+
"rewards/rejected": -0.5022113919258118,
|
| 653 |
+
"step": 450
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.98,
|
| 657 |
+
"learning_rate": 4.453449766758933e-09,
|
| 658 |
+
"logits/chosen": 0.1742466688156128,
|
| 659 |
+
"logits/rejected": 0.2268284559249878,
|
| 660 |
+
"logps/chosen": -576.7985229492188,
|
| 661 |
+
"logps/rejected": -652.7803344726562,
|
| 662 |
+
"loss": 1.7297,
|
| 663 |
+
"rewards/accuracies": 0.625,
|
| 664 |
+
"rewards/chosen": -0.24071533977985382,
|
| 665 |
+
"rewards/margins": 0.18981412053108215,
|
| 666 |
+
"rewards/rejected": -0.4305294454097748,
|
| 667 |
+
"step": 460
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 1.0,
|
| 671 |
+
"step": 468,
|
| 672 |
+
"total_flos": 0.0,
|
| 673 |
+
"train_loss": 1.8394590188295414,
|
| 674 |
+
"train_runtime": 15861.5475,
|
| 675 |
+
"train_samples_per_second": 1.891,
|
| 676 |
+
"train_steps_per_second": 0.03
|
| 677 |
+
}
|
| 678 |
+
],
|
| 679 |
+
"logging_steps": 10,
|
| 680 |
+
"max_steps": 468,
|
| 681 |
+
"num_input_tokens_seen": 0,
|
| 682 |
+
"num_train_epochs": 1,
|
| 683 |
+
"save_steps": 100,
|
| 684 |
+
"total_flos": 0.0,
|
| 685 |
+
"train_batch_size": 4,
|
| 686 |
+
"trial_name": null,
|
| 687 |
+
"trial_params": null
|
| 688 |
+
}
|