Commit
•
74f0d63
1
Parent(s):
82806c9
Model save
Browse files- README.md +62 -0
- adapter_model.safetensors +1 -1
- all_results.json +8 -0
- runs/May15_04-09-53_ip-10-192-15-18/events.out.tfevents.1715746324.ip-10-192-15-18.4084.0 +2 -2
- 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-v7-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-v7-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 sha256:25147d8deeeb07f2cbf71fc4d12c814fa3e75e373908b39b522ea7529492d0ab
|
| 3 |
size 167807296
|
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"train_loss": 1.4258099234002268,
|
| 4 |
+
"train_runtime": 12564.9924,
|
| 5 |
+
"train_samples": 61135,
|
| 6 |
+
"train_samples_per_second": 2.388,
|
| 7 |
+
"train_steps_per_second": 0.037
|
| 8 |
+
}
|
runs/May15_04-09-53_ip-10-192-15-18/events.out.tfevents.1715746324.ip-10-192-15-18.4084.0
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 sha256:36375b9b30703fc55bb699caaac1574c9343e752a8469c6528917c6bb5ff0e34
|
| 3 |
+
size 28072
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"train_loss": 1.4258099234002268,
|
| 4 |
+
"train_runtime": 12564.9924,
|
| 5 |
+
"train_samples": 61135,
|
| 6 |
+
"train_samples_per_second": 2.388,
|
| 7 |
+
"train_steps_per_second": 0.037
|
| 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": 0.11680144816637039,
|
| 29 |
+
"logits/rejected": 0.26637858152389526,
|
| 30 |
+
"logps/chosen": -546.4848022460938,
|
| 31 |
+
"logps/rejected": -597.5820922851562,
|
| 32 |
+
"loss": 2.1606,
|
| 33 |
+
"rewards/accuracies": 0.4027777910232544,
|
| 34 |
+
"rewards/chosen": 0.00018841998826246709,
|
| 35 |
+
"rewards/margins": 0.0001322307507507503,
|
| 36 |
+
"rewards/rejected": 5.6189212045865133e-05,
|
| 37 |
+
"step": 10
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.04,
|
| 41 |
+
"learning_rate": 2.1276595744680853e-06,
|
| 42 |
+
"logits/chosen": 0.16349577903747559,
|
| 43 |
+
"logits/rejected": 0.2681581676006317,
|
| 44 |
+
"logps/chosen": -604.7147827148438,
|
| 45 |
+
"logps/rejected": -649.5513916015625,
|
| 46 |
+
"loss": 2.0961,
|
| 47 |
+
"rewards/accuracies": 0.45625001192092896,
|
| 48 |
+
"rewards/chosen": -0.0001730217190925032,
|
| 49 |
+
"rewards/margins": -4.76564237033017e-05,
|
| 50 |
+
"rewards/rejected": -0.0001253652444574982,
|
| 51 |
+
"step": 20
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.06,
|
| 55 |
+
"learning_rate": 3.191489361702128e-06,
|
| 56 |
+
"logits/chosen": 0.1495944708585739,
|
| 57 |
+
"logits/rejected": 0.19113227725028992,
|
| 58 |
+
"logps/chosen": -594.8287353515625,
|
| 59 |
+
"logps/rejected": -588.0198974609375,
|
| 60 |
+
"loss": 2.124,
|
| 61 |
+
"rewards/accuracies": 0.45625001192092896,
|
| 62 |
+
"rewards/chosen": -0.000811461650300771,
|
| 63 |
+
"rewards/margins": -7.172574896685546e-06,
|
| 64 |
+
"rewards/rejected": -0.0008042889530770481,
|
| 65 |
+
"step": 30
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.09,
|
| 69 |
+
"learning_rate": 4.255319148936171e-06,
|
| 70 |
+
"logits/chosen": 0.16839194297790527,
|
| 71 |
+
"logits/rejected": 0.23641005158424377,
|
| 72 |
+
"logps/chosen": -574.6544799804688,
|
| 73 |
+
"logps/rejected": -631.8447875976562,
|
| 74 |
+
"loss": 2.1986,
|
| 75 |
+
"rewards/accuracies": 0.5562499761581421,
|
| 76 |
+
"rewards/chosen": -0.0023893495090305805,
|
| 77 |
+
"rewards/margins": 0.001175779732875526,
|
| 78 |
+
"rewards/rejected": -0.0035651293583214283,
|
| 79 |
+
"step": 40
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.11,
|
| 83 |
+
"learning_rate": 4.999373573764188e-06,
|
| 84 |
+
"logits/chosen": 0.14030753076076508,
|
| 85 |
+
"logits/rejected": 0.22485153377056122,
|
| 86 |
+
"logps/chosen": -613.718505859375,
|
| 87 |
+
"logps/rejected": -637.7477416992188,
|
| 88 |
+
"loss": 2.1858,
|
| 89 |
+
"rewards/accuracies": 0.5062500238418579,
|
| 90 |
+
"rewards/chosen": -0.01074314396828413,
|
| 91 |
+
"rewards/margins": 0.004234404768794775,
|
| 92 |
+
"rewards/rejected": -0.014977550134062767,
|
| 93 |
+
"step": 50
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.13,
|
| 97 |
+
"learning_rate": 4.988245838331339e-06,
|
| 98 |
+
"logits/chosen": 0.16629436612129211,
|
| 99 |
+
"logits/rejected": 0.1657789647579193,
|
| 100 |
+
"logps/chosen": -639.8953857421875,
|
| 101 |
+
"logps/rejected": -674.060302734375,
|
| 102 |
+
"loss": 2.1366,
|
| 103 |
+
"rewards/accuracies": 0.5562499761581421,
|
| 104 |
+
"rewards/chosen": -0.03131238371133804,
|
| 105 |
+
"rewards/margins": 0.011484434828162193,
|
| 106 |
+
"rewards/rejected": -0.04279681295156479,
|
| 107 |
+
"step": 60
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.15,
|
| 111 |
+
"learning_rate": 4.963268819535228e-06,
|
| 112 |
+
"logits/chosen": 0.10423990339040756,
|
| 113 |
+
"logits/rejected": 0.11754454672336578,
|
| 114 |
+
"logps/chosen": -634.0193481445312,
|
| 115 |
+
"logps/rejected": -738.2115478515625,
|
| 116 |
+
"loss": 2.1147,
|
| 117 |
+
"rewards/accuracies": 0.5625,
|
| 118 |
+
"rewards/chosen": -0.06819052994251251,
|
| 119 |
+
"rewards/margins": 0.03033190593123436,
|
| 120 |
+
"rewards/rejected": -0.09852243959903717,
|
| 121 |
+
"step": 70
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.17,
|
| 125 |
+
"learning_rate": 4.9245815365216115e-06,
|
| 126 |
+
"logits/chosen": 0.13063886761665344,
|
| 127 |
+
"logits/rejected": 0.18018567562103271,
|
| 128 |
+
"logps/chosen": -764.638671875,
|
| 129 |
+
"logps/rejected": -714.2967529296875,
|
| 130 |
+
"loss": 2.1622,
|
| 131 |
+
"rewards/accuracies": 0.512499988079071,
|
| 132 |
+
"rewards/chosen": -0.12814757227897644,
|
| 133 |
+
"rewards/margins": 0.030911359935998917,
|
| 134 |
+
"rewards/rejected": -0.15905895829200745,
|
| 135 |
+
"step": 80
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.19,
|
| 139 |
+
"learning_rate": 4.872399318152594e-06,
|
| 140 |
+
"logits/chosen": 0.032799430191516876,
|
| 141 |
+
"logits/rejected": 0.08405763655900955,
|
| 142 |
+
"logps/chosen": -795.3159790039062,
|
| 143 |
+
"logps/rejected": -893.0309448242188,
|
| 144 |
+
"loss": 2.056,
|
| 145 |
+
"rewards/accuracies": 0.5874999761581421,
|
| 146 |
+
"rewards/chosen": -0.20986966788768768,
|
| 147 |
+
"rewards/margins": 0.08644559979438782,
|
| 148 |
+
"rewards/rejected": -0.2963152527809143,
|
| 149 |
+
"step": 90
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.21,
|
| 153 |
+
"learning_rate": 4.807012604511542e-06,
|
| 154 |
+
"logits/chosen": 0.07495725154876709,
|
| 155 |
+
"logits/rejected": 0.16169913113117218,
|
| 156 |
+
"logps/chosen": -909.8048706054688,
|
| 157 |
+
"logps/rejected": -995.3268432617188,
|
| 158 |
+
"loss": 2.0738,
|
| 159 |
+
"rewards/accuracies": 0.5874999761581421,
|
| 160 |
+
"rewards/chosen": -0.2920338809490204,
|
| 161 |
+
"rewards/margins": 0.09413494169712067,
|
| 162 |
+
"rewards/rejected": -0.38616880774497986,
|
| 163 |
+
"step": 100
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.23,
|
| 167 |
+
"learning_rate": 4.728785330347771e-06,
|
| 168 |
+
"logits/chosen": 0.16292598843574524,
|
| 169 |
+
"logits/rejected": 0.1960914433002472,
|
| 170 |
+
"logps/chosen": -904.22509765625,
|
| 171 |
+
"logps/rejected": -912.8228759765625,
|
| 172 |
+
"loss": 1.9508,
|
| 173 |
+
"rewards/accuracies": 0.5625,
|
| 174 |
+
"rewards/chosen": -0.2555181682109833,
|
| 175 |
+
"rewards/margins": 0.05460430309176445,
|
| 176 |
+
"rewards/rejected": -0.3101224899291992,
|
| 177 |
+
"step": 110
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.26,
|
| 181 |
+
"learning_rate": 4.63815289945858e-06,
|
| 182 |
+
"logits/chosen": 0.10713698714971542,
|
| 183 |
+
"logits/rejected": 0.20941510796546936,
|
| 184 |
+
"logps/chosen": -819.4172973632812,
|
| 185 |
+
"logps/rejected": -1036.814208984375,
|
| 186 |
+
"loss": 1.9156,
|
| 187 |
+
"rewards/accuracies": 0.637499988079071,
|
| 188 |
+
"rewards/chosen": -0.26556941866874695,
|
| 189 |
+
"rewards/margins": 0.15225186944007874,
|
| 190 |
+
"rewards/rejected": -0.41782131791114807,
|
| 191 |
+
"step": 120
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.28,
|
| 195 |
+
"learning_rate": 4.535619761282989e-06,
|
| 196 |
+
"logits/chosen": 0.14809398353099823,
|
| 197 |
+
"logits/rejected": 0.18617282807826996,
|
| 198 |
+
"logps/chosen": -886.8990478515625,
|
| 199 |
+
"logps/rejected": -1030.2890625,
|
| 200 |
+
"loss": 1.9636,
|
| 201 |
+
"rewards/accuracies": 0.5375000238418579,
|
| 202 |
+
"rewards/chosen": -0.32516202330589294,
|
| 203 |
+
"rewards/margins": 0.13631486892700195,
|
| 204 |
+
"rewards/rejected": -0.4614768922328949,
|
| 205 |
+
"step": 130
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.3,
|
| 209 |
+
"learning_rate": 4.42175660319555e-06,
|
| 210 |
+
"logits/chosen": 0.18207255005836487,
|
| 211 |
+
"logits/rejected": 0.18847420811653137,
|
| 212 |
+
"logps/chosen": -927.9578247070312,
|
| 213 |
+
"logps/rejected": -1096.78955078125,
|
| 214 |
+
"loss": 1.8634,
|
| 215 |
+
"rewards/accuracies": 0.581250011920929,
|
| 216 |
+
"rewards/chosen": -0.3230741322040558,
|
| 217 |
+
"rewards/margins": 0.20067179203033447,
|
| 218 |
+
"rewards/rejected": -0.5237459540367126,
|
| 219 |
+
"step": 140
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.32,
|
| 223 |
+
"learning_rate": 4.297197174127619e-06,
|
| 224 |
+
"logits/chosen": 0.17518748342990875,
|
| 225 |
+
"logits/rejected": 0.22532868385314941,
|
| 226 |
+
"logps/chosen": -905.5735473632812,
|
| 227 |
+
"logps/rejected": -1151.804443359375,
|
| 228 |
+
"loss": 1.8103,
|
| 229 |
+
"rewards/accuracies": 0.581250011920929,
|
| 230 |
+
"rewards/chosen": -0.3360939621925354,
|
| 231 |
+
"rewards/margins": 0.21287350356578827,
|
| 232 |
+
"rewards/rejected": -0.5489674210548401,
|
| 233 |
+
"step": 150
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.34,
|
| 237 |
+
"learning_rate": 4.162634757195418e-06,
|
| 238 |
+
"logits/chosen": 0.2002106010913849,
|
| 239 |
+
"logits/rejected": 0.19682054221630096,
|
| 240 |
+
"logps/chosen": -903.08935546875,
|
| 241 |
+
"logps/rejected": -1042.5498046875,
|
| 242 |
+
"loss": 1.9037,
|
| 243 |
+
"rewards/accuracies": 0.59375,
|
| 244 |
+
"rewards/chosen": -0.3114651143550873,
|
| 245 |
+
"rewards/margins": 0.1639038622379303,
|
| 246 |
+
"rewards/rejected": -0.4753689765930176,
|
| 247 |
+
"step": 160
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.36,
|
| 251 |
+
"learning_rate": 4.018818310967843e-06,
|
| 252 |
+
"logits/chosen": 0.20454064011573792,
|
| 253 |
+
"logits/rejected": 0.2340986430644989,
|
| 254 |
+
"logps/chosen": -755.2179565429688,
|
| 255 |
+
"logps/rejected": -950.396484375,
|
| 256 |
+
"loss": 1.9158,
|
| 257 |
+
"rewards/accuracies": 0.6187499761581421,
|
| 258 |
+
"rewards/chosen": -0.21088416874408722,
|
| 259 |
+
"rewards/margins": 0.1926594078540802,
|
| 260 |
+
"rewards/rejected": -0.4035435616970062,
|
| 261 |
+
"step": 170
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.38,
|
| 265 |
+
"learning_rate": 3.866548300851254e-06,
|
| 266 |
+
"logits/chosen": 0.17594218254089355,
|
| 267 |
+
"logits/rejected": 0.19749297201633453,
|
| 268 |
+
"logps/chosen": -861.9845581054688,
|
| 269 |
+
"logps/rejected": -1095.2757568359375,
|
| 270 |
+
"loss": 1.8464,
|
| 271 |
+
"rewards/accuracies": 0.612500011920929,
|
| 272 |
+
"rewards/chosen": -0.26217204332351685,
|
| 273 |
+
"rewards/margins": 0.22928395867347717,
|
| 274 |
+
"rewards/rejected": -0.49145597219467163,
|
| 275 |
+
"step": 180
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.41,
|
| 279 |
+
"learning_rate": 3.706672243793271e-06,
|
| 280 |
+
"logits/chosen": 0.2364441156387329,
|
| 281 |
+
"logits/rejected": 0.29772695899009705,
|
| 282 |
+
"logps/chosen": -868.7154541015625,
|
| 283 |
+
"logps/rejected": -1106.079345703125,
|
| 284 |
+
"loss": 1.8101,
|
| 285 |
+
"rewards/accuracies": 0.6187499761581421,
|
| 286 |
+
"rewards/chosen": -0.27426981925964355,
|
| 287 |
+
"rewards/margins": 0.2382505238056183,
|
| 288 |
+
"rewards/rejected": -0.5125203132629395,
|
| 289 |
+
"step": 190
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.43,
|
| 293 |
+
"learning_rate": 3.5400799911032357e-06,
|
| 294 |
+
"logits/chosen": 0.22934266924858093,
|
| 295 |
+
"logits/rejected": 0.2707999646663666,
|
| 296 |
+
"logps/chosen": -949.2401123046875,
|
| 297 |
+
"logps/rejected": -1226.9296875,
|
| 298 |
+
"loss": 1.7955,
|
| 299 |
+
"rewards/accuracies": 0.6937500238418579,
|
| 300 |
+
"rewards/chosen": -0.3255212903022766,
|
| 301 |
+
"rewards/margins": 0.26814141869544983,
|
| 302 |
+
"rewards/rejected": -0.5936626195907593,
|
| 303 |
+
"step": 200
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.45,
|
| 307 |
+
"learning_rate": 3.3676987756445894e-06,
|
| 308 |
+
"logits/chosen": 0.20278143882751465,
|
| 309 |
+
"logits/rejected": 0.2634762227535248,
|
| 310 |
+
"logps/chosen": -825.6008911132812,
|
| 311 |
+
"logps/rejected": -1062.2701416015625,
|
| 312 |
+
"loss": 1.8656,
|
| 313 |
+
"rewards/accuracies": 0.612500011920929,
|
| 314 |
+
"rewards/chosen": -0.2474614679813385,
|
| 315 |
+
"rewards/margins": 0.25558072328567505,
|
| 316 |
+
"rewards/rejected": -0.5030421614646912,
|
| 317 |
+
"step": 210
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.47,
|
| 321 |
+
"learning_rate": 3.1904880509659397e-06,
|
| 322 |
+
"logits/chosen": 0.2307882010936737,
|
| 323 |
+
"logits/rejected": 0.2746850252151489,
|
| 324 |
+
"logps/chosen": -902.4068603515625,
|
| 325 |
+
"logps/rejected": -1159.0933837890625,
|
| 326 |
+
"loss": 1.7356,
|
| 327 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 328 |
+
"rewards/chosen": -0.28473055362701416,
|
| 329 |
+
"rewards/margins": 0.2488754689693451,
|
| 330 |
+
"rewards/rejected": -0.5336060523986816,
|
| 331 |
+
"step": 220
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.49,
|
| 335 |
+
"learning_rate": 3.0094341510955697e-06,
|
| 336 |
+
"logits/chosen": 0.15470555424690247,
|
| 337 |
+
"logits/rejected": 0.24446210265159607,
|
| 338 |
+
"logps/chosen": -932.212890625,
|
| 339 |
+
"logps/rejected": -1326.384033203125,
|
| 340 |
+
"loss": 1.8039,
|
| 341 |
+
"rewards/accuracies": 0.7124999761581421,
|
| 342 |
+
"rewards/chosen": -0.2982610762119293,
|
| 343 |
+
"rewards/margins": 0.38883471488952637,
|
| 344 |
+
"rewards/rejected": -0.6870958209037781,
|
| 345 |
+
"step": 230
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.51,
|
| 349 |
+
"learning_rate": 2.825544800722376e-06,
|
| 350 |
+
"logits/chosen": 0.1649903804063797,
|
| 351 |
+
"logits/rejected": 0.2850923240184784,
|
| 352 |
+
"logps/chosen": -914.9806518554688,
|
| 353 |
+
"logps/rejected": -1192.943359375,
|
| 354 |
+
"loss": 1.8769,
|
| 355 |
+
"rewards/accuracies": 0.5874999761581421,
|
| 356 |
+
"rewards/chosen": -0.33413922786712646,
|
| 357 |
+
"rewards/margins": 0.2569636404514313,
|
| 358 |
+
"rewards/rejected": -0.5911028981208801,
|
| 359 |
+
"step": 240
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.53,
|
| 363 |
+
"learning_rate": 2.639843506318899e-06,
|
| 364 |
+
"logits/chosen": 0.2421075850725174,
|
| 365 |
+
"logits/rejected": 0.23027054965496063,
|
| 366 |
+
"logps/chosen": -815.9312744140625,
|
| 367 |
+
"logps/rejected": -1075.3148193359375,
|
| 368 |
+
"loss": 1.9302,
|
| 369 |
+
"rewards/accuracies": 0.643750011920929,
|
| 370 |
+
"rewards/chosen": -0.2732352614402771,
|
| 371 |
+
"rewards/margins": 0.20585504174232483,
|
| 372 |
+
"rewards/rejected": -0.47909030318260193,
|
| 373 |
+
"step": 250
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.55,
|
| 377 |
+
"learning_rate": 2.4533638594248094e-06,
|
| 378 |
+
"logits/chosen": 0.23280009627342224,
|
| 379 |
+
"logits/rejected": 0.27932989597320557,
|
| 380 |
+
"logps/chosen": -820.83447265625,
|
| 381 |
+
"logps/rejected": -1066.0859375,
|
| 382 |
+
"loss": 1.8994,
|
| 383 |
+
"rewards/accuracies": 0.6187499761581421,
|
| 384 |
+
"rewards/chosen": -0.24443455040454865,
|
| 385 |
+
"rewards/margins": 0.23904922604560852,
|
| 386 |
+
"rewards/rejected": -0.4834837317466736,
|
| 387 |
+
"step": 260
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.58,
|
| 391 |
+
"learning_rate": 2.2671437837980943e-06,
|
| 392 |
+
"logits/chosen": 0.20527169108390808,
|
| 393 |
+
"logits/rejected": 0.1906355619430542,
|
| 394 |
+
"logps/chosen": -802.508544921875,
|
| 395 |
+
"logps/rejected": -1058.9862060546875,
|
| 396 |
+
"loss": 1.7706,
|
| 397 |
+
"rewards/accuracies": 0.6000000238418579,
|
| 398 |
+
"rewards/chosen": -0.23849380016326904,
|
| 399 |
+
"rewards/margins": 0.23252327740192413,
|
| 400 |
+
"rewards/rejected": -0.47101712226867676,
|
| 401 |
+
"step": 270
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.6,
|
| 405 |
+
"learning_rate": 2.082219758453629e-06,
|
| 406 |
+
"logits/chosen": 0.18224522471427917,
|
| 407 |
+
"logits/rejected": 0.23733553290367126,
|
| 408 |
+
"logps/chosen": -854.2337646484375,
|
| 409 |
+
"logps/rejected": -1095.759765625,
|
| 410 |
+
"loss": 1.7181,
|
| 411 |
+
"rewards/accuracies": 0.65625,
|
| 412 |
+
"rewards/chosen": -0.26597216725349426,
|
| 413 |
+
"rewards/margins": 0.22445416450500488,
|
| 414 |
+
"rewards/rejected": -0.4904263913631439,
|
| 415 |
+
"step": 280
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.62,
|
| 419 |
+
"learning_rate": 1.899621048743019e-06,
|
| 420 |
+
"logits/chosen": 0.1843213140964508,
|
| 421 |
+
"logits/rejected": 0.2989790141582489,
|
| 422 |
+
"logps/chosen": -902.1618041992188,
|
| 423 |
+
"logps/rejected": -1187.279541015625,
|
| 424 |
+
"loss": 1.7856,
|
| 425 |
+
"rewards/accuracies": 0.637499988079071,
|
| 426 |
+
"rewards/chosen": -0.3390468955039978,
|
| 427 |
+
"rewards/margins": 0.2693897485733032,
|
| 428 |
+
"rewards/rejected": -0.608436644077301,
|
| 429 |
+
"step": 290
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.64,
|
| 433 |
+
"learning_rate": 1.7203639775848423e-06,
|
| 434 |
+
"logits/chosen": 0.15538965165615082,
|
| 435 |
+
"logits/rejected": 0.25810253620147705,
|
| 436 |
+
"logps/chosen": -898.4368896484375,
|
| 437 |
+
"logps/rejected": -1099.8011474609375,
|
| 438 |
+
"loss": 1.8812,
|
| 439 |
+
"rewards/accuracies": 0.6312500238418579,
|
| 440 |
+
"rewards/chosen": -0.32740405201911926,
|
| 441 |
+
"rewards/margins": 0.21446183323860168,
|
| 442 |
+
"rewards/rejected": -0.5418658256530762,
|
| 443 |
+
"step": 300
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.66,
|
| 447 |
+
"learning_rate": 1.5454462687309445e-06,
|
| 448 |
+
"logits/chosen": 0.17961570620536804,
|
| 449 |
+
"logits/rejected": 0.22687454521656036,
|
| 450 |
+
"logps/chosen": -870.6505737304688,
|
| 451 |
+
"logps/rejected": -1096.16015625,
|
| 452 |
+
"loss": 1.8067,
|
| 453 |
+
"rewards/accuracies": 0.6000000238418579,
|
| 454 |
+
"rewards/chosen": -0.2986447811126709,
|
| 455 |
+
"rewards/margins": 0.21215729415416718,
|
| 456 |
+
"rewards/rejected": -0.5108019709587097,
|
| 457 |
+
"step": 310
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.68,
|
| 461 |
+
"learning_rate": 1.3758414935535147e-06,
|
| 462 |
+
"logits/chosen": 0.18137860298156738,
|
| 463 |
+
"logits/rejected": 0.24465851485729218,
|
| 464 |
+
"logps/chosen": -902.7379150390625,
|
| 465 |
+
"logps/rejected": -1196.3099365234375,
|
| 466 |
+
"loss": 1.68,
|
| 467 |
+
"rewards/accuracies": 0.6625000238418579,
|
| 468 |
+
"rewards/chosen": -0.3003231883049011,
|
| 469 |
+
"rewards/margins": 0.2799701988697052,
|
| 470 |
+
"rewards/rejected": -0.5802933573722839,
|
| 471 |
+
"step": 320
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.7,
|
| 475 |
+
"learning_rate": 1.2124936522614622e-06,
|
| 476 |
+
"logits/chosen": 0.18053703010082245,
|
| 477 |
+
"logits/rejected": 0.19177524745464325,
|
| 478 |
+
"logps/chosen": -886.50341796875,
|
| 479 |
+
"logps/rejected": -1139.5965576171875,
|
| 480 |
+
"loss": 1.7893,
|
| 481 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 482 |
+
"rewards/chosen": -0.3074926435947418,
|
| 483 |
+
"rewards/margins": 0.2617393434047699,
|
| 484 |
+
"rewards/rejected": -0.5692320466041565,
|
| 485 |
+
"step": 330
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.73,
|
| 489 |
+
"learning_rate": 1.0563119197063934e-06,
|
| 490 |
+
"logits/chosen": 0.20963367819786072,
|
| 491 |
+
"logits/rejected": 0.23218393325805664,
|
| 492 |
+
"logps/chosen": -865.2020263671875,
|
| 493 |
+
"logps/rejected": -1143.264404296875,
|
| 494 |
+
"loss": 1.7222,
|
| 495 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 496 |
+
"rewards/chosen": -0.2907507121562958,
|
| 497 |
+
"rewards/margins": 0.2580210566520691,
|
| 498 |
+
"rewards/rejected": -0.5487717390060425,
|
| 499 |
+
"step": 340
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.75,
|
| 503 |
+
"learning_rate": 9.081655850224449e-07,
|
| 504 |
+
"logits/chosen": 0.16453954577445984,
|
| 505 |
+
"logits/rejected": 0.18646371364593506,
|
| 506 |
+
"logps/chosen": -938.1325073242188,
|
| 507 |
+
"logps/rejected": -1226.00634765625,
|
| 508 |
+
"loss": 1.7215,
|
| 509 |
+
"rewards/accuracies": 0.6875,
|
| 510 |
+
"rewards/chosen": -0.3565717339515686,
|
| 511 |
+
"rewards/margins": 0.2777550220489502,
|
| 512 |
+
"rewards/rejected": -0.6343267560005188,
|
| 513 |
+
"step": 350
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.77,
|
| 517 |
+
"learning_rate": 7.688792132653111e-07,
|
| 518 |
+
"logits/chosen": 0.14916151762008667,
|
| 519 |
+
"logits/rejected": 0.23365116119384766,
|
| 520 |
+
"logps/chosen": -938.5538940429688,
|
| 521 |
+
"logps/rejected": -1271.653076171875,
|
| 522 |
+
"loss": 1.7912,
|
| 523 |
+
"rewards/accuracies": 0.612500011920929,
|
| 524 |
+
"rewards/chosen": -0.3168650269508362,
|
| 525 |
+
"rewards/margins": 0.3130674362182617,
|
| 526 |
+
"rewards/rejected": -0.6299324035644531,
|
| 527 |
+
"step": 360
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.79,
|
| 531 |
+
"learning_rate": 6.392280559802341e-07,
|
| 532 |
+
"logits/chosen": 0.20630326867103577,
|
| 533 |
+
"logits/rejected": 0.19222524762153625,
|
| 534 |
+
"logps/chosen": -968.6492309570312,
|
| 535 |
+
"logps/rejected": -1210.966552734375,
|
| 536 |
+
"loss": 1.7648,
|
| 537 |
+
"rewards/accuracies": 0.625,
|
| 538 |
+
"rewards/chosen": -0.35769984126091003,
|
| 539 |
+
"rewards/margins": 0.22969777882099152,
|
| 540 |
+
"rewards/rejected": -0.5873976945877075,
|
| 541 |
+
"step": 370
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.81,
|
| 545 |
+
"learning_rate": 5.199337362431792e-07,
|
| 546 |
+
"logits/chosen": 0.23185975849628448,
|
| 547 |
+
"logits/rejected": 0.12383759021759033,
|
| 548 |
+
"logps/chosen": -913.8956298828125,
|
| 549 |
+
"logps/rejected": -1156.3277587890625,
|
| 550 |
+
"loss": 1.764,
|
| 551 |
+
"rewards/accuracies": 0.6312500238418579,
|
| 552 |
+
"rewards/chosen": -0.3383910357952118,
|
| 553 |
+
"rewards/margins": 0.2371567189693451,
|
| 554 |
+
"rewards/rejected": -0.5755477547645569,
|
| 555 |
+
"step": 380
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.83,
|
| 559 |
+
"learning_rate": 4.1166023219176176e-07,
|
| 560 |
+
"logits/chosen": 0.17365404963493347,
|
| 561 |
+
"logits/rejected": 0.22495949268341064,
|
| 562 |
+
"logps/chosen": -965.8720703125,
|
| 563 |
+
"logps/rejected": -1141.7669677734375,
|
| 564 |
+
"loss": 1.7503,
|
| 565 |
+
"rewards/accuracies": 0.6812499761581421,
|
| 566 |
+
"rewards/chosen": -0.3551119267940521,
|
| 567 |
+
"rewards/margins": 0.21352621912956238,
|
| 568 |
+
"rewards/rejected": -0.5686381459236145,
|
| 569 |
+
"step": 390
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.85,
|
| 573 |
+
"learning_rate": 3.150101814011136e-07,
|
| 574 |
+
"logits/chosen": 0.12874256074428558,
|
| 575 |
+
"logits/rejected": 0.17315654456615448,
|
| 576 |
+
"logps/chosen": -858.0895385742188,
|
| 577 |
+
"logps/rejected": -1191.4886474609375,
|
| 578 |
+
"loss": 1.7221,
|
| 579 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 580 |
+
"rewards/chosen": -0.2957189381122589,
|
| 581 |
+
"rewards/margins": 0.26001372933387756,
|
| 582 |
+
"rewards/rejected": -0.5557327270507812,
|
| 583 |
+
"step": 400
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.87,
|
| 587 |
+
"learning_rate": 2.3052152667409289e-07,
|
| 588 |
+
"logits/chosen": 0.16493479907512665,
|
| 589 |
+
"logits/rejected": 0.17930065095424652,
|
| 590 |
+
"logps/chosen": -871.76025390625,
|
| 591 |
+
"logps/rejected": -1177.62353515625,
|
| 592 |
+
"loss": 1.8162,
|
| 593 |
+
"rewards/accuracies": 0.668749988079071,
|
| 594 |
+
"rewards/chosen": -0.28667283058166504,
|
| 595 |
+
"rewards/margins": 0.30985671281814575,
|
| 596 |
+
"rewards/rejected": -0.5965295433998108,
|
| 597 |
+
"step": 410
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.9,
|
| 601 |
+
"learning_rate": 1.5866452191498488e-07,
|
| 602 |
+
"logits/chosen": 0.165872260928154,
|
| 603 |
+
"logits/rejected": 0.2134593427181244,
|
| 604 |
+
"logps/chosen": -985.79248046875,
|
| 605 |
+
"logps/rejected": -1185.5306396484375,
|
| 606 |
+
"loss": 1.8081,
|
| 607 |
+
"rewards/accuracies": 0.581250011920929,
|
| 608 |
+
"rewards/chosen": -0.38030657172203064,
|
| 609 |
+
"rewards/margins": 0.19018930196762085,
|
| 610 |
+
"rewards/rejected": -0.5704958438873291,
|
| 611 |
+
"step": 420
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.92,
|
| 615 |
+
"learning_rate": 9.983911475163727e-08,
|
| 616 |
+
"logits/chosen": 0.1297980695962906,
|
| 617 |
+
"logits/rejected": 0.2198360413312912,
|
| 618 |
+
"logps/chosen": -843.9484252929688,
|
| 619 |
+
"logps/rejected": -1071.7698974609375,
|
| 620 |
+
"loss": 1.816,
|
| 621 |
+
"rewards/accuracies": 0.6187499761581421,
|
| 622 |
+
"rewards/chosen": -0.28988003730773926,
|
| 623 |
+
"rewards/margins": 0.22120681405067444,
|
| 624 |
+
"rewards/rejected": -0.5110868811607361,
|
| 625 |
+
"step": 430
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.94,
|
| 629 |
+
"learning_rate": 5.437272047405712e-08,
|
| 630 |
+
"logits/chosen": 0.14703018963336945,
|
| 631 |
+
"logits/rejected": 0.26481714844703674,
|
| 632 |
+
"logps/chosen": -841.8990478515625,
|
| 633 |
+
"logps/rejected": -1089.636474609375,
|
| 634 |
+
"loss": 1.7983,
|
| 635 |
+
"rewards/accuracies": 0.643750011920929,
|
| 636 |
+
"rewards/chosen": -0.32132893800735474,
|
| 637 |
+
"rewards/margins": 0.21181070804595947,
|
| 638 |
+
"rewards/rejected": -0.5331395864486694,
|
| 639 |
+
"step": 440
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.96,
|
| 643 |
+
"learning_rate": 2.251839967945535e-08,
|
| 644 |
+
"logits/chosen": 0.1110246405005455,
|
| 645 |
+
"logits/rejected": 0.1886615753173828,
|
| 646 |
+
"logps/chosen": -919.92333984375,
|
| 647 |
+
"logps/rejected": -1245.7840576171875,
|
| 648 |
+
"loss": 1.6896,
|
| 649 |
+
"rewards/accuracies": 0.643750011920929,
|
| 650 |
+
"rewards/chosen": -0.30925479531288147,
|
| 651 |
+
"rewards/margins": 0.329929918050766,
|
| 652 |
+
"rewards/rejected": -0.6391846537590027,
|
| 653 |
+
"step": 450
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.98,
|
| 657 |
+
"learning_rate": 4.453449766758933e-09,
|
| 658 |
+
"logits/chosen": 0.13569632172584534,
|
| 659 |
+
"logits/rejected": 0.18673357367515564,
|
| 660 |
+
"logps/chosen": -879.3952026367188,
|
| 661 |
+
"logps/rejected": -1108.7191162109375,
|
| 662 |
+
"loss": 1.7726,
|
| 663 |
+
"rewards/accuracies": 0.6312500238418579,
|
| 664 |
+
"rewards/chosen": -0.35073980689048767,
|
| 665 |
+
"rewards/margins": 0.19130484759807587,
|
| 666 |
+
"rewards/rejected": -0.5420446991920471,
|
| 667 |
+
"step": 460
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 1.0,
|
| 671 |
+
"step": 468,
|
| 672 |
+
"total_flos": 0.0,
|
| 673 |
+
"train_loss": 1.4258099234002268,
|
| 674 |
+
"train_runtime": 12564.9924,
|
| 675 |
+
"train_samples_per_second": 2.388,
|
| 676 |
+
"train_steps_per_second": 0.037
|
| 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 |
+
}
|