Model save
Browse files- README.md +62 -0
- adapter_model.safetensors +1 -1
- all_results.json +8 -0
- runs/May14_01-46-16_gpu4-119-5/events.out.tfevents.1715615284.gpu4-119-5.874343.0 +2 -2
- train_results.json +8 -0
- trainer_state.json +688 -0
README.md
ADDED
|
@@ -0,0 +1,62 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
library_name: peft
|
| 4 |
+
tags:
|
| 5 |
+
- trl
|
| 6 |
+
- dpo
|
| 7 |
+
- generated_from_trainer
|
| 8 |
+
base_model: DUAL-GPO/phi-2-gpo-new-i0
|
| 9 |
+
model-index:
|
| 10 |
+
- name: phi-2-gpo-v17-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-v17-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: 2
|
| 44 |
+
- gradient_accumulation_steps: 4
|
| 45 |
+
- total_train_batch_size: 32
|
| 46 |
+
- total_eval_batch_size: 8
|
| 47 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 48 |
+
- lr_scheduler_type: cosine
|
| 49 |
+
- lr_scheduler_warmup_ratio: 0.1
|
| 50 |
+
- num_epochs: 1
|
| 51 |
+
|
| 52 |
+
### Training results
|
| 53 |
+
|
| 54 |
+
|
| 55 |
+
|
| 56 |
+
### Framework versions
|
| 57 |
+
|
| 58 |
+
- PEFT 0.7.1
|
| 59 |
+
- Transformers 4.36.2
|
| 60 |
+
- Pytorch 2.1.2+cu121
|
| 61 |
+
- Datasets 2.14.6
|
| 62 |
+
- Tokenizers 0.15.2
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 167807296
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4bfd291f7e2c26d83d180708d8587922ae2c88f381f35820bfb0e58d76dfe9fd
|
| 3 |
size 167807296
|
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"train_loss": 0.2399272450015076,
|
| 4 |
+
"train_runtime": 4417.3525,
|
| 5 |
+
"train_samples": 15000,
|
| 6 |
+
"train_samples_per_second": 3.396,
|
| 7 |
+
"train_steps_per_second": 0.106
|
| 8 |
+
}
|
runs/May14_01-46-16_gpu4-119-5/events.out.tfevents.1715615284.gpu4-119-5.874343.0
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:bc5db9f128c54a07c8751a4a4da422ebaf144f4664661894f518e5ec34385714
|
| 3 |
+
size 27983
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"train_loss": 0.2399272450015076,
|
| 4 |
+
"train_runtime": 4417.3525,
|
| 5 |
+
"train_samples": 15000,
|
| 6 |
+
"train_samples_per_second": 3.396,
|
| 7 |
+
"train_steps_per_second": 0.106
|
| 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": 0.18604117631912231,
|
| 15 |
+
"logits/rejected": 0.34631967544555664,
|
| 16 |
+
"logps/chosen": -460.0769348144531,
|
| 17 |
+
"logps/rejected": -351.57135009765625,
|
| 18 |
+
"loss": 0.3612,
|
| 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": 0.05747946724295616,
|
| 29 |
+
"logits/rejected": 0.16539901494979858,
|
| 30 |
+
"logps/chosen": -330.9287414550781,
|
| 31 |
+
"logps/rejected": -328.71575927734375,
|
| 32 |
+
"loss": 0.3819,
|
| 33 |
+
"rewards/accuracies": 0.3888888955116272,
|
| 34 |
+
"rewards/chosen": 2.3913149561849423e-05,
|
| 35 |
+
"rewards/margins": -1.7764228687155992e-05,
|
| 36 |
+
"rewards/rejected": 4.167737643001601e-05,
|
| 37 |
+
"step": 10
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.04,
|
| 41 |
+
"learning_rate": 2.1276595744680853e-06,
|
| 42 |
+
"logits/chosen": 0.15373219549655914,
|
| 43 |
+
"logits/rejected": 0.19493858516216278,
|
| 44 |
+
"logps/chosen": -325.29803466796875,
|
| 45 |
+
"logps/rejected": -315.1011047363281,
|
| 46 |
+
"loss": 0.3718,
|
| 47 |
+
"rewards/accuracies": 0.4312500059604645,
|
| 48 |
+
"rewards/chosen": -1.974666338355746e-05,
|
| 49 |
+
"rewards/margins": 1.2878153654583002e-07,
|
| 50 |
+
"rewards/rejected": -1.9875456928275526e-05,
|
| 51 |
+
"step": 20
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.06,
|
| 55 |
+
"learning_rate": 3.191489361702128e-06,
|
| 56 |
+
"logits/chosen": 0.11452794075012207,
|
| 57 |
+
"logits/rejected": 0.1842522919178009,
|
| 58 |
+
"logps/chosen": -371.60943603515625,
|
| 59 |
+
"logps/rejected": -327.19366455078125,
|
| 60 |
+
"loss": 0.3676,
|
| 61 |
+
"rewards/accuracies": 0.4937500059604645,
|
| 62 |
+
"rewards/chosen": -5.7096302043646574e-05,
|
| 63 |
+
"rewards/margins": 3.7131747376406565e-05,
|
| 64 |
+
"rewards/rejected": -9.422805305803195e-05,
|
| 65 |
+
"step": 30
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.09,
|
| 69 |
+
"learning_rate": 4.255319148936171e-06,
|
| 70 |
+
"logits/chosen": 0.18893569707870483,
|
| 71 |
+
"logits/rejected": 0.2084759920835495,
|
| 72 |
+
"logps/chosen": -337.7607727050781,
|
| 73 |
+
"logps/rejected": -353.72503662109375,
|
| 74 |
+
"loss": 0.3678,
|
| 75 |
+
"rewards/accuracies": 0.5062500238418579,
|
| 76 |
+
"rewards/chosen": 0.00010834180284291506,
|
| 77 |
+
"rewards/margins": 0.00026063303812406957,
|
| 78 |
+
"rewards/rejected": -0.0001522912352811545,
|
| 79 |
+
"step": 40
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.11,
|
| 83 |
+
"learning_rate": 4.999373573764188e-06,
|
| 84 |
+
"logits/chosen": 0.08207504451274872,
|
| 85 |
+
"logits/rejected": 0.20786412060260773,
|
| 86 |
+
"logps/chosen": -384.91656494140625,
|
| 87 |
+
"logps/rejected": -347.6080627441406,
|
| 88 |
+
"loss": 0.3741,
|
| 89 |
+
"rewards/accuracies": 0.625,
|
| 90 |
+
"rewards/chosen": 0.0010863704374060035,
|
| 91 |
+
"rewards/margins": 0.0014893051702529192,
|
| 92 |
+
"rewards/rejected": -0.0004029346746392548,
|
| 93 |
+
"step": 50
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.13,
|
| 97 |
+
"learning_rate": 4.988245838331339e-06,
|
| 98 |
+
"logits/chosen": 0.13469652831554413,
|
| 99 |
+
"logits/rejected": 0.17153413593769073,
|
| 100 |
+
"logps/chosen": -372.34124755859375,
|
| 101 |
+
"logps/rejected": -329.98626708984375,
|
| 102 |
+
"loss": 0.3646,
|
| 103 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 104 |
+
"rewards/chosen": 0.0023965700529515743,
|
| 105 |
+
"rewards/margins": 0.0026933744084089994,
|
| 106 |
+
"rewards/rejected": -0.0002968042972497642,
|
| 107 |
+
"step": 60
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.15,
|
| 111 |
+
"learning_rate": 4.963268819535228e-06,
|
| 112 |
+
"logits/chosen": 0.15689334273338318,
|
| 113 |
+
"logits/rejected": 0.20978419482707977,
|
| 114 |
+
"logps/chosen": -364.29888916015625,
|
| 115 |
+
"logps/rejected": -377.21087646484375,
|
| 116 |
+
"loss": 0.3705,
|
| 117 |
+
"rewards/accuracies": 0.71875,
|
| 118 |
+
"rewards/chosen": 0.0023818810004740953,
|
| 119 |
+
"rewards/margins": 0.0072565278969705105,
|
| 120 |
+
"rewards/rejected": -0.0048746466636657715,
|
| 121 |
+
"step": 70
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.17,
|
| 125 |
+
"learning_rate": 4.9245815365216115e-06,
|
| 126 |
+
"logits/chosen": 0.11595060676336288,
|
| 127 |
+
"logits/rejected": 0.18608702719211578,
|
| 128 |
+
"logps/chosen": -328.6256103515625,
|
| 129 |
+
"logps/rejected": -365.93896484375,
|
| 130 |
+
"loss": 0.3663,
|
| 131 |
+
"rewards/accuracies": 0.6812499761581421,
|
| 132 |
+
"rewards/chosen": -0.002104334533214569,
|
| 133 |
+
"rewards/margins": 0.011472588405013084,
|
| 134 |
+
"rewards/rejected": -0.013576922006905079,
|
| 135 |
+
"step": 80
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.19,
|
| 139 |
+
"learning_rate": 4.872399318152594e-06,
|
| 140 |
+
"logits/chosen": 0.08527339994907379,
|
| 141 |
+
"logits/rejected": 0.11540959030389786,
|
| 142 |
+
"logps/chosen": -352.0487060546875,
|
| 143 |
+
"logps/rejected": -361.36578369140625,
|
| 144 |
+
"loss": 0.3456,
|
| 145 |
+
"rewards/accuracies": 0.675000011920929,
|
| 146 |
+
"rewards/chosen": -0.011672710999846458,
|
| 147 |
+
"rewards/margins": 0.01901327446103096,
|
| 148 |
+
"rewards/rejected": -0.03068598173558712,
|
| 149 |
+
"step": 90
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.21,
|
| 153 |
+
"learning_rate": 4.807012604511542e-06,
|
| 154 |
+
"logits/chosen": 0.14592930674552917,
|
| 155 |
+
"logits/rejected": 0.18039533495903015,
|
| 156 |
+
"logps/chosen": -414.8818359375,
|
| 157 |
+
"logps/rejected": -435.3639221191406,
|
| 158 |
+
"loss": 0.34,
|
| 159 |
+
"rewards/accuracies": 0.6625000238418579,
|
| 160 |
+
"rewards/chosen": -0.04601982235908508,
|
| 161 |
+
"rewards/margins": 0.034013133496046066,
|
| 162 |
+
"rewards/rejected": -0.08003295958042145,
|
| 163 |
+
"step": 100
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.23,
|
| 167 |
+
"learning_rate": 4.728785330347771e-06,
|
| 168 |
+
"logits/chosen": 0.13395074009895325,
|
| 169 |
+
"logits/rejected": 0.2120208740234375,
|
| 170 |
+
"logps/chosen": -432.98486328125,
|
| 171 |
+
"logps/rejected": -553.6598510742188,
|
| 172 |
+
"loss": 0.305,
|
| 173 |
+
"rewards/accuracies": 0.6875,
|
| 174 |
+
"rewards/chosen": -0.11783289909362793,
|
| 175 |
+
"rewards/margins": 0.08183668553829193,
|
| 176 |
+
"rewards/rejected": -0.19966959953308105,
|
| 177 |
+
"step": 110
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.26,
|
| 181 |
+
"learning_rate": 4.63815289945858e-06,
|
| 182 |
+
"logits/chosen": 0.14807412028312683,
|
| 183 |
+
"logits/rejected": 0.21360798180103302,
|
| 184 |
+
"logps/chosen": -565.0419311523438,
|
| 185 |
+
"logps/rejected": -595.3816528320312,
|
| 186 |
+
"loss": 0.3434,
|
| 187 |
+
"rewards/accuracies": 0.612500011920929,
|
| 188 |
+
"rewards/chosen": -0.18253007531166077,
|
| 189 |
+
"rewards/margins": 0.060289014130830765,
|
| 190 |
+
"rewards/rejected": -0.24281907081604004,
|
| 191 |
+
"step": 120
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.28,
|
| 195 |
+
"learning_rate": 4.535619761282989e-06,
|
| 196 |
+
"logits/chosen": 0.10818658024072647,
|
| 197 |
+
"logits/rejected": 0.18194182217121124,
|
| 198 |
+
"logps/chosen": -462.3949279785156,
|
| 199 |
+
"logps/rejected": -513.8490600585938,
|
| 200 |
+
"loss": 0.3173,
|
| 201 |
+
"rewards/accuracies": 0.6625000238418579,
|
| 202 |
+
"rewards/chosen": -0.1326686441898346,
|
| 203 |
+
"rewards/margins": 0.0804656594991684,
|
| 204 |
+
"rewards/rejected": -0.2131342887878418,
|
| 205 |
+
"step": 130
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.3,
|
| 209 |
+
"learning_rate": 4.42175660319555e-06,
|
| 210 |
+
"logits/chosen": 0.13493295013904572,
|
| 211 |
+
"logits/rejected": 0.22737479209899902,
|
| 212 |
+
"logps/chosen": -532.4095458984375,
|
| 213 |
+
"logps/rejected": -585.3455200195312,
|
| 214 |
+
"loss": 0.3196,
|
| 215 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 216 |
+
"rewards/chosen": -0.16629299521446228,
|
| 217 |
+
"rewards/margins": 0.08414360880851746,
|
| 218 |
+
"rewards/rejected": -0.25043657422065735,
|
| 219 |
+
"step": 140
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.32,
|
| 223 |
+
"learning_rate": 4.297197174127619e-06,
|
| 224 |
+
"logits/chosen": 0.17478415369987488,
|
| 225 |
+
"logits/rejected": 0.24990789592266083,
|
| 226 |
+
"logps/chosen": -479.77862548828125,
|
| 227 |
+
"logps/rejected": -553.7377319335938,
|
| 228 |
+
"loss": 0.3207,
|
| 229 |
+
"rewards/accuracies": 0.6187499761581421,
|
| 230 |
+
"rewards/chosen": -0.14474426209926605,
|
| 231 |
+
"rewards/margins": 0.09155549854040146,
|
| 232 |
+
"rewards/rejected": -0.2362997829914093,
|
| 233 |
+
"step": 150
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.34,
|
| 237 |
+
"learning_rate": 4.162634757195418e-06,
|
| 238 |
+
"logits/chosen": 0.13402113318443298,
|
| 239 |
+
"logits/rejected": 0.2551622688770294,
|
| 240 |
+
"logps/chosen": -496.40081787109375,
|
| 241 |
+
"logps/rejected": -558.84326171875,
|
| 242 |
+
"loss": 0.3187,
|
| 243 |
+
"rewards/accuracies": 0.65625,
|
| 244 |
+
"rewards/chosen": -0.13809171319007874,
|
| 245 |
+
"rewards/margins": 0.08086591213941574,
|
| 246 |
+
"rewards/rejected": -0.21895763278007507,
|
| 247 |
+
"step": 160
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.36,
|
| 251 |
+
"learning_rate": 4.018818310967843e-06,
|
| 252 |
+
"logits/chosen": 0.12252243608236313,
|
| 253 |
+
"logits/rejected": 0.16481925547122955,
|
| 254 |
+
"logps/chosen": -484.56353759765625,
|
| 255 |
+
"logps/rejected": -554.2274780273438,
|
| 256 |
+
"loss": 0.3255,
|
| 257 |
+
"rewards/accuracies": 0.6625000238418579,
|
| 258 |
+
"rewards/chosen": -0.14209917187690735,
|
| 259 |
+
"rewards/margins": 0.09574152529239655,
|
| 260 |
+
"rewards/rejected": -0.2378406971693039,
|
| 261 |
+
"step": 170
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.38,
|
| 265 |
+
"learning_rate": 3.866548300851254e-06,
|
| 266 |
+
"logits/chosen": 0.08216498792171478,
|
| 267 |
+
"logits/rejected": 0.17952165007591248,
|
| 268 |
+
"logps/chosen": -482.576171875,
|
| 269 |
+
"logps/rejected": -579.781982421875,
|
| 270 |
+
"loss": 0.3047,
|
| 271 |
+
"rewards/accuracies": 0.637499988079071,
|
| 272 |
+
"rewards/chosen": -0.14626096189022064,
|
| 273 |
+
"rewards/margins": 0.08570893108844757,
|
| 274 |
+
"rewards/rejected": -0.23196987807750702,
|
| 275 |
+
"step": 180
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.41,
|
| 279 |
+
"learning_rate": 3.706672243793271e-06,
|
| 280 |
+
"logits/chosen": 0.07855963706970215,
|
| 281 |
+
"logits/rejected": 0.07844971120357513,
|
| 282 |
+
"logps/chosen": -468.69061279296875,
|
| 283 |
+
"logps/rejected": -544.0549926757812,
|
| 284 |
+
"loss": 0.2935,
|
| 285 |
+
"rewards/accuracies": 0.6875,
|
| 286 |
+
"rewards/chosen": -0.1354297697544098,
|
| 287 |
+
"rewards/margins": 0.09144213050603867,
|
| 288 |
+
"rewards/rejected": -0.22687189280986786,
|
| 289 |
+
"step": 190
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.43,
|
| 293 |
+
"learning_rate": 3.5400799911032357e-06,
|
| 294 |
+
"logits/chosen": 0.10545216500759125,
|
| 295 |
+
"logits/rejected": 0.1882828176021576,
|
| 296 |
+
"logps/chosen": -501.1815490722656,
|
| 297 |
+
"logps/rejected": -601.7250366210938,
|
| 298 |
+
"loss": 0.3041,
|
| 299 |
+
"rewards/accuracies": 0.6312500238418579,
|
| 300 |
+
"rewards/chosen": -0.15607957541942596,
|
| 301 |
+
"rewards/margins": 0.10436417162418365,
|
| 302 |
+
"rewards/rejected": -0.2604437470436096,
|
| 303 |
+
"step": 200
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.45,
|
| 307 |
+
"learning_rate": 3.3676987756445894e-06,
|
| 308 |
+
"logits/chosen": 0.10487540811300278,
|
| 309 |
+
"logits/rejected": 0.11818546056747437,
|
| 310 |
+
"logps/chosen": -470.6344299316406,
|
| 311 |
+
"logps/rejected": -565.8145751953125,
|
| 312 |
+
"loss": 0.3148,
|
| 313 |
+
"rewards/accuracies": 0.612500011920929,
|
| 314 |
+
"rewards/chosen": -0.16682696342468262,
|
| 315 |
+
"rewards/margins": 0.09452913701534271,
|
| 316 |
+
"rewards/rejected": -0.26135605573654175,
|
| 317 |
+
"step": 210
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.47,
|
| 321 |
+
"learning_rate": 3.1904880509659397e-06,
|
| 322 |
+
"logits/chosen": 0.13482534885406494,
|
| 323 |
+
"logits/rejected": 0.20024776458740234,
|
| 324 |
+
"logps/chosen": -510.10528564453125,
|
| 325 |
+
"logps/rejected": -605.6468505859375,
|
| 326 |
+
"loss": 0.3172,
|
| 327 |
+
"rewards/accuracies": 0.6812499761581421,
|
| 328 |
+
"rewards/chosen": -0.16119703650474548,
|
| 329 |
+
"rewards/margins": 0.1006912812590599,
|
| 330 |
+
"rewards/rejected": -0.261888325214386,
|
| 331 |
+
"step": 220
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.49,
|
| 335 |
+
"learning_rate": 3.0094341510955697e-06,
|
| 336 |
+
"logits/chosen": 0.10005593299865723,
|
| 337 |
+
"logits/rejected": 0.1791614145040512,
|
| 338 |
+
"logps/chosen": -532.1925048828125,
|
| 339 |
+
"logps/rejected": -624.0726318359375,
|
| 340 |
+
"loss": 0.3106,
|
| 341 |
+
"rewards/accuracies": 0.668749988079071,
|
| 342 |
+
"rewards/chosen": -0.17168700695037842,
|
| 343 |
+
"rewards/margins": 0.09216924011707306,
|
| 344 |
+
"rewards/rejected": -0.26385626196861267,
|
| 345 |
+
"step": 230
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.51,
|
| 349 |
+
"learning_rate": 2.825544800722376e-06,
|
| 350 |
+
"logits/chosen": 0.10918021202087402,
|
| 351 |
+
"logits/rejected": 0.18382051587104797,
|
| 352 |
+
"logps/chosen": -508.88494873046875,
|
| 353 |
+
"logps/rejected": -571.0011596679688,
|
| 354 |
+
"loss": 0.309,
|
| 355 |
+
"rewards/accuracies": 0.637499988079071,
|
| 356 |
+
"rewards/chosen": -0.15795882046222687,
|
| 357 |
+
"rewards/margins": 0.09128745645284653,
|
| 358 |
+
"rewards/rejected": -0.2492462694644928,
|
| 359 |
+
"step": 240
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.53,
|
| 363 |
+
"learning_rate": 2.639843506318899e-06,
|
| 364 |
+
"logits/chosen": 0.12444597482681274,
|
| 365 |
+
"logits/rejected": 0.16049379110336304,
|
| 366 |
+
"logps/chosen": -509.65631103515625,
|
| 367 |
+
"logps/rejected": -609.0398559570312,
|
| 368 |
+
"loss": 0.2924,
|
| 369 |
+
"rewards/accuracies": 0.6937500238418579,
|
| 370 |
+
"rewards/chosen": -0.1560695320367813,
|
| 371 |
+
"rewards/margins": 0.11202052980661392,
|
| 372 |
+
"rewards/rejected": -0.26809003949165344,
|
| 373 |
+
"step": 250
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.55,
|
| 377 |
+
"learning_rate": 2.4533638594248094e-06,
|
| 378 |
+
"logits/chosen": 0.11506851017475128,
|
| 379 |
+
"logits/rejected": 0.1052849292755127,
|
| 380 |
+
"logps/chosen": -535.1851806640625,
|
| 381 |
+
"logps/rejected": -609.4058837890625,
|
| 382 |
+
"loss": 0.3139,
|
| 383 |
+
"rewards/accuracies": 0.699999988079071,
|
| 384 |
+
"rewards/chosen": -0.18579542636871338,
|
| 385 |
+
"rewards/margins": 0.10177616029977798,
|
| 386 |
+
"rewards/rejected": -0.28757157921791077,
|
| 387 |
+
"step": 260
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.58,
|
| 391 |
+
"learning_rate": 2.2671437837980943e-06,
|
| 392 |
+
"logits/chosen": 0.11827238649129868,
|
| 393 |
+
"logits/rejected": 0.12189098447561264,
|
| 394 |
+
"logps/chosen": -542.7433471679688,
|
| 395 |
+
"logps/rejected": -633.5802612304688,
|
| 396 |
+
"loss": 0.2867,
|
| 397 |
+
"rewards/accuracies": 0.606249988079071,
|
| 398 |
+
"rewards/chosen": -0.19100908935070038,
|
| 399 |
+
"rewards/margins": 0.09670811891555786,
|
| 400 |
+
"rewards/rejected": -0.28771719336509705,
|
| 401 |
+
"step": 270
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.6,
|
| 405 |
+
"learning_rate": 2.082219758453629e-06,
|
| 406 |
+
"logits/chosen": 0.0904841274023056,
|
| 407 |
+
"logits/rejected": 0.20232203602790833,
|
| 408 |
+
"logps/chosen": -575.5374755859375,
|
| 409 |
+
"logps/rejected": -672.4290161132812,
|
| 410 |
+
"loss": 0.3105,
|
| 411 |
+
"rewards/accuracies": 0.6812499761581421,
|
| 412 |
+
"rewards/chosen": -0.2171885222196579,
|
| 413 |
+
"rewards/margins": 0.10769243538379669,
|
| 414 |
+
"rewards/rejected": -0.3248809278011322,
|
| 415 |
+
"step": 280
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.62,
|
| 419 |
+
"learning_rate": 1.899621048743019e-06,
|
| 420 |
+
"logits/chosen": 0.07801838964223862,
|
| 421 |
+
"logits/rejected": 0.16570156812667847,
|
| 422 |
+
"logps/chosen": -556.6237182617188,
|
| 423 |
+
"logps/rejected": -643.4578857421875,
|
| 424 |
+
"loss": 0.3145,
|
| 425 |
+
"rewards/accuracies": 0.6937500238418579,
|
| 426 |
+
"rewards/chosen": -0.20063337683677673,
|
| 427 |
+
"rewards/margins": 0.1107235923409462,
|
| 428 |
+
"rewards/rejected": -0.31135696172714233,
|
| 429 |
+
"step": 290
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.64,
|
| 433 |
+
"learning_rate": 1.7203639775848423e-06,
|
| 434 |
+
"logits/chosen": 0.07458348572254181,
|
| 435 |
+
"logits/rejected": 0.08251482248306274,
|
| 436 |
+
"logps/chosen": -524.4200439453125,
|
| 437 |
+
"logps/rejected": -666.1168823242188,
|
| 438 |
+
"loss": 0.2934,
|
| 439 |
+
"rewards/accuracies": 0.668749988079071,
|
| 440 |
+
"rewards/chosen": -0.18233875930309296,
|
| 441 |
+
"rewards/margins": 0.13505366444587708,
|
| 442 |
+
"rewards/rejected": -0.3173924386501312,
|
| 443 |
+
"step": 300
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.66,
|
| 447 |
+
"learning_rate": 1.5454462687309445e-06,
|
| 448 |
+
"logits/chosen": 0.06379405409097672,
|
| 449 |
+
"logits/rejected": 0.14568018913269043,
|
| 450 |
+
"logps/chosen": -504.7530822753906,
|
| 451 |
+
"logps/rejected": -650.941650390625,
|
| 452 |
+
"loss": 0.2814,
|
| 453 |
+
"rewards/accuracies": 0.675000011920929,
|
| 454 |
+
"rewards/chosen": -0.17345736920833588,
|
| 455 |
+
"rewards/margins": 0.1400579959154129,
|
| 456 |
+
"rewards/rejected": -0.3135153353214264,
|
| 457 |
+
"step": 310
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.68,
|
| 461 |
+
"learning_rate": 1.3758414935535147e-06,
|
| 462 |
+
"logits/chosen": 0.0773477703332901,
|
| 463 |
+
"logits/rejected": 0.12101063877344131,
|
| 464 |
+
"logps/chosen": -536.0679931640625,
|
| 465 |
+
"logps/rejected": -659.2662353515625,
|
| 466 |
+
"loss": 0.3014,
|
| 467 |
+
"rewards/accuracies": 0.71875,
|
| 468 |
+
"rewards/chosen": -0.18032298982143402,
|
| 469 |
+
"rewards/margins": 0.11869337409734726,
|
| 470 |
+
"rewards/rejected": -0.29901641607284546,
|
| 471 |
+
"step": 320
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.7,
|
| 475 |
+
"learning_rate": 1.2124936522614622e-06,
|
| 476 |
+
"logits/chosen": 0.09862785786390305,
|
| 477 |
+
"logits/rejected": 0.14109982550144196,
|
| 478 |
+
"logps/chosen": -487.79071044921875,
|
| 479 |
+
"logps/rejected": -610.7267456054688,
|
| 480 |
+
"loss": 0.2935,
|
| 481 |
+
"rewards/accuracies": 0.65625,
|
| 482 |
+
"rewards/chosen": -0.18224193155765533,
|
| 483 |
+
"rewards/margins": 0.10070188343524933,
|
| 484 |
+
"rewards/rejected": -0.2829437851905823,
|
| 485 |
+
"step": 330
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.73,
|
| 489 |
+
"learning_rate": 1.0563119197063934e-06,
|
| 490 |
+
"logits/chosen": 0.09601452201604843,
|
| 491 |
+
"logits/rejected": 0.13355228304862976,
|
| 492 |
+
"logps/chosen": -478.78143310546875,
|
| 493 |
+
"logps/rejected": -622.6370239257812,
|
| 494 |
+
"loss": 0.31,
|
| 495 |
+
"rewards/accuracies": 0.668749988079071,
|
| 496 |
+
"rewards/chosen": -0.17518463730812073,
|
| 497 |
+
"rewards/margins": 0.11871640384197235,
|
| 498 |
+
"rewards/rejected": -0.2939010262489319,
|
| 499 |
+
"step": 340
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.75,
|
| 503 |
+
"learning_rate": 9.081655850224449e-07,
|
| 504 |
+
"logits/chosen": 0.14273716509342194,
|
| 505 |
+
"logits/rejected": 0.12122112512588501,
|
| 506 |
+
"logps/chosen": -535.0431518554688,
|
| 507 |
+
"logps/rejected": -643.0104370117188,
|
| 508 |
+
"loss": 0.3196,
|
| 509 |
+
"rewards/accuracies": 0.675000011920929,
|
| 510 |
+
"rewards/chosen": -0.19643327593803406,
|
| 511 |
+
"rewards/margins": 0.10987784713506699,
|
| 512 |
+
"rewards/rejected": -0.30631113052368164,
|
| 513 |
+
"step": 350
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.77,
|
| 517 |
+
"learning_rate": 7.688792132653111e-07,
|
| 518 |
+
"logits/chosen": 0.1597367525100708,
|
| 519 |
+
"logits/rejected": 0.18071278929710388,
|
| 520 |
+
"logps/chosen": -597.0608520507812,
|
| 521 |
+
"logps/rejected": -692.2689208984375,
|
| 522 |
+
"loss": 0.2965,
|
| 523 |
+
"rewards/accuracies": 0.699999988079071,
|
| 524 |
+
"rewards/chosen": -0.20307877659797668,
|
| 525 |
+
"rewards/margins": 0.12918424606323242,
|
| 526 |
+
"rewards/rejected": -0.3322630524635315,
|
| 527 |
+
"step": 360
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.79,
|
| 531 |
+
"learning_rate": 6.392280559802341e-07,
|
| 532 |
+
"logits/chosen": 0.1371072232723236,
|
| 533 |
+
"logits/rejected": 0.16206106543540955,
|
| 534 |
+
"logps/chosen": -573.9052124023438,
|
| 535 |
+
"logps/rejected": -700.1799926757812,
|
| 536 |
+
"loss": 0.2689,
|
| 537 |
+
"rewards/accuracies": 0.706250011920929,
|
| 538 |
+
"rewards/chosen": -0.2099401205778122,
|
| 539 |
+
"rewards/margins": 0.13856378197669983,
|
| 540 |
+
"rewards/rejected": -0.34850388765335083,
|
| 541 |
+
"step": 370
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.81,
|
| 545 |
+
"learning_rate": 5.199337362431792e-07,
|
| 546 |
+
"logits/chosen": 0.10267746448516846,
|
| 547 |
+
"logits/rejected": 0.1483933925628662,
|
| 548 |
+
"logps/chosen": -554.7241821289062,
|
| 549 |
+
"logps/rejected": -637.0789184570312,
|
| 550 |
+
"loss": 0.2999,
|
| 551 |
+
"rewards/accuracies": 0.71875,
|
| 552 |
+
"rewards/chosen": -0.19773730635643005,
|
| 553 |
+
"rewards/margins": 0.11398313194513321,
|
| 554 |
+
"rewards/rejected": -0.31172046065330505,
|
| 555 |
+
"step": 380
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.83,
|
| 559 |
+
"learning_rate": 4.1166023219176176e-07,
|
| 560 |
+
"logits/chosen": 0.11247100681066513,
|
| 561 |
+
"logits/rejected": 0.1309679001569748,
|
| 562 |
+
"logps/chosen": -557.1490478515625,
|
| 563 |
+
"logps/rejected": -678.8016357421875,
|
| 564 |
+
"loss": 0.2945,
|
| 565 |
+
"rewards/accuracies": 0.6875,
|
| 566 |
+
"rewards/chosen": -0.1877584308385849,
|
| 567 |
+
"rewards/margins": 0.12902548909187317,
|
| 568 |
+
"rewards/rejected": -0.3167839050292969,
|
| 569 |
+
"step": 390
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.85,
|
| 573 |
+
"learning_rate": 3.150101814011136e-07,
|
| 574 |
+
"logits/chosen": 0.16857033967971802,
|
| 575 |
+
"logits/rejected": 0.18106935918331146,
|
| 576 |
+
"logps/chosen": -571.0828857421875,
|
| 577 |
+
"logps/rejected": -638.4654541015625,
|
| 578 |
+
"loss": 0.3045,
|
| 579 |
+
"rewards/accuracies": 0.668749988079071,
|
| 580 |
+
"rewards/chosen": -0.19992712140083313,
|
| 581 |
+
"rewards/margins": 0.10413169860839844,
|
| 582 |
+
"rewards/rejected": -0.30405884981155396,
|
| 583 |
+
"step": 400
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.87,
|
| 587 |
+
"learning_rate": 2.3052152667409289e-07,
|
| 588 |
+
"logits/chosen": 0.09823437035083771,
|
| 589 |
+
"logits/rejected": 0.23346427083015442,
|
| 590 |
+
"logps/chosen": -537.1575317382812,
|
| 591 |
+
"logps/rejected": -657.79931640625,
|
| 592 |
+
"loss": 0.3101,
|
| 593 |
+
"rewards/accuracies": 0.637499988079071,
|
| 594 |
+
"rewards/chosen": -0.19983352720737457,
|
| 595 |
+
"rewards/margins": 0.11520209163427353,
|
| 596 |
+
"rewards/rejected": -0.3150356113910675,
|
| 597 |
+
"step": 410
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.9,
|
| 601 |
+
"learning_rate": 1.5866452191498488e-07,
|
| 602 |
+
"logits/chosen": 0.13151055574417114,
|
| 603 |
+
"logits/rejected": 0.17384907603263855,
|
| 604 |
+
"logps/chosen": -562.8436279296875,
|
| 605 |
+
"logps/rejected": -688.2677612304688,
|
| 606 |
+
"loss": 0.2892,
|
| 607 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 608 |
+
"rewards/chosen": -0.21342253684997559,
|
| 609 |
+
"rewards/margins": 0.1261134147644043,
|
| 610 |
+
"rewards/rejected": -0.3395359516143799,
|
| 611 |
+
"step": 420
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.92,
|
| 615 |
+
"learning_rate": 9.983911475163727e-08,
|
| 616 |
+
"logits/chosen": 0.11608059704303741,
|
| 617 |
+
"logits/rejected": 0.13708294928073883,
|
| 618 |
+
"logps/chosen": -503.0716857910156,
|
| 619 |
+
"logps/rejected": -606.0911865234375,
|
| 620 |
+
"loss": 0.2983,
|
| 621 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 622 |
+
"rewards/chosen": -0.18081562221050262,
|
| 623 |
+
"rewards/margins": 0.11758317053318024,
|
| 624 |
+
"rewards/rejected": -0.29839879274368286,
|
| 625 |
+
"step": 430
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.94,
|
| 629 |
+
"learning_rate": 5.437272047405712e-08,
|
| 630 |
+
"logits/chosen": 0.11100079119205475,
|
| 631 |
+
"logits/rejected": 0.13695240020751953,
|
| 632 |
+
"logps/chosen": -530.9528198242188,
|
| 633 |
+
"logps/rejected": -668.3643798828125,
|
| 634 |
+
"loss": 0.3095,
|
| 635 |
+
"rewards/accuracies": 0.668749988079071,
|
| 636 |
+
"rewards/chosen": -0.2087739259004593,
|
| 637 |
+
"rewards/margins": 0.13447535037994385,
|
| 638 |
+
"rewards/rejected": -0.34324929118156433,
|
| 639 |
+
"step": 440
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.96,
|
| 643 |
+
"learning_rate": 2.251839967945535e-08,
|
| 644 |
+
"logits/chosen": 0.022059569135308266,
|
| 645 |
+
"logits/rejected": 0.12616530060768127,
|
| 646 |
+
"logps/chosen": -492.80181884765625,
|
| 647 |
+
"logps/rejected": -643.1315307617188,
|
| 648 |
+
"loss": 0.2834,
|
| 649 |
+
"rewards/accuracies": 0.706250011920929,
|
| 650 |
+
"rewards/chosen": -0.18623578548431396,
|
| 651 |
+
"rewards/margins": 0.14532844722270966,
|
| 652 |
+
"rewards/rejected": -0.3315642178058624,
|
| 653 |
+
"step": 450
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.98,
|
| 657 |
+
"learning_rate": 4.453449766758933e-09,
|
| 658 |
+
"logits/chosen": 0.10801200568675995,
|
| 659 |
+
"logits/rejected": 0.11458040773868561,
|
| 660 |
+
"logps/chosen": -531.0709228515625,
|
| 661 |
+
"logps/rejected": -645.5343627929688,
|
| 662 |
+
"loss": 0.3043,
|
| 663 |
+
"rewards/accuracies": 0.6937500238418579,
|
| 664 |
+
"rewards/chosen": -0.1950865238904953,
|
| 665 |
+
"rewards/margins": 0.11565764993429184,
|
| 666 |
+
"rewards/rejected": -0.3107442259788513,
|
| 667 |
+
"step": 460
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 1.0,
|
| 671 |
+
"step": 468,
|
| 672 |
+
"total_flos": 0.0,
|
| 673 |
+
"train_loss": 0.2399272450015076,
|
| 674 |
+
"train_runtime": 4417.3525,
|
| 675 |
+
"train_samples_per_second": 3.396,
|
| 676 |
+
"train_steps_per_second": 0.106
|
| 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 |
+
}
|