Model save
Browse files- README.md +62 -0
- adapter_model.safetensors +1 -1
- all_results.json +8 -0
- runs/May12_19-04-40_gpu4-119-5/events.out.tfevents.1715504803.gpu4-119-5.2839953.0 +2 -2
- train_results.json +8 -0
- trainer_state.json +954 -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-v5-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-v5-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:db00f65d11ffe3c62fd9be7c399dd08154b795b2466a57c4f94ad449841d7dd3
|
| 3 |
size 167807296
|
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"train_loss": 0.20043009792159244,
|
| 4 |
+
"train_runtime": 6753.4712,
|
| 5 |
+
"train_samples": 21000,
|
| 6 |
+
"train_samples_per_second": 3.11,
|
| 7 |
+
"train_steps_per_second": 0.097
|
| 8 |
+
}
|
runs/May12_19-04-40_gpu4-119-5/events.out.tfevents.1715504803.gpu4-119-5.2839953.0
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:d445aac9a720443f20488a36a9cbd5f87297cce741d6f1d12510707ee09fe4c0
|
| 3 |
+
size 40025
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"train_loss": 0.20043009792159244,
|
| 4 |
+
"train_runtime": 6753.4712,
|
| 5 |
+
"train_samples": 21000,
|
| 6 |
+
"train_samples_per_second": 3.11,
|
| 7 |
+
"train_steps_per_second": 0.097
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,954 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.9996190476190476,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 656,
|
| 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": 7.575757575757576e-08,
|
| 14 |
+
"logits/chosen": 0.040165986865758896,
|
| 15 |
+
"logits/rejected": 0.1715753823518753,
|
| 16 |
+
"logps/chosen": -294.844482421875,
|
| 17 |
+
"logps/rejected": -361.2099914550781,
|
| 18 |
+
"loss": 0.3581,
|
| 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": 7.575757575757576e-07,
|
| 28 |
+
"logits/chosen": 0.08021458983421326,
|
| 29 |
+
"logits/rejected": 0.320384681224823,
|
| 30 |
+
"logps/chosen": -393.2468566894531,
|
| 31 |
+
"logps/rejected": -318.70526123046875,
|
| 32 |
+
"loss": 0.3399,
|
| 33 |
+
"rewards/accuracies": 0.375,
|
| 34 |
+
"rewards/chosen": -5.44932481716387e-05,
|
| 35 |
+
"rewards/margins": -7.010095578152686e-05,
|
| 36 |
+
"rewards/rejected": 1.560769487696234e-05,
|
| 37 |
+
"step": 10
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.03,
|
| 41 |
+
"learning_rate": 1.5151515151515152e-06,
|
| 42 |
+
"logits/chosen": 0.07484304904937744,
|
| 43 |
+
"logits/rejected": 0.23277099430561066,
|
| 44 |
+
"logps/chosen": -332.4987487792969,
|
| 45 |
+
"logps/rejected": -281.78729248046875,
|
| 46 |
+
"loss": 0.3453,
|
| 47 |
+
"rewards/accuracies": 0.3687500059604645,
|
| 48 |
+
"rewards/chosen": 1.3394804909694358e-06,
|
| 49 |
+
"rewards/margins": -0.0002027603331953287,
|
| 50 |
+
"rewards/rejected": 0.00020409980788826942,
|
| 51 |
+
"step": 20
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.05,
|
| 55 |
+
"learning_rate": 2.2727272727272728e-06,
|
| 56 |
+
"logits/chosen": 0.07151266187429428,
|
| 57 |
+
"logits/rejected": 0.25990238785743713,
|
| 58 |
+
"logps/chosen": -358.3196716308594,
|
| 59 |
+
"logps/rejected": -292.2561950683594,
|
| 60 |
+
"loss": 0.3517,
|
| 61 |
+
"rewards/accuracies": 0.543749988079071,
|
| 62 |
+
"rewards/chosen": 0.0001912448788061738,
|
| 63 |
+
"rewards/margins": 0.00028789255884476006,
|
| 64 |
+
"rewards/rejected": -9.664769459050149e-05,
|
| 65 |
+
"step": 30
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.06,
|
| 69 |
+
"learning_rate": 3.0303030303030305e-06,
|
| 70 |
+
"logits/chosen": 0.12050364166498184,
|
| 71 |
+
"logits/rejected": 0.23899266123771667,
|
| 72 |
+
"logps/chosen": -346.99310302734375,
|
| 73 |
+
"logps/rejected": -286.31658935546875,
|
| 74 |
+
"loss": 0.3416,
|
| 75 |
+
"rewards/accuracies": 0.6625000238418579,
|
| 76 |
+
"rewards/chosen": 0.0006401558639481664,
|
| 77 |
+
"rewards/margins": 0.0007750070071779191,
|
| 78 |
+
"rewards/rejected": -0.00013485117233358324,
|
| 79 |
+
"step": 40
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.08,
|
| 83 |
+
"learning_rate": 3.7878787878787882e-06,
|
| 84 |
+
"logits/chosen": 0.07284825295209885,
|
| 85 |
+
"logits/rejected": 0.3108685612678528,
|
| 86 |
+
"logps/chosen": -343.5631408691406,
|
| 87 |
+
"logps/rejected": -287.16876220703125,
|
| 88 |
+
"loss": 0.3439,
|
| 89 |
+
"rewards/accuracies": 0.668749988079071,
|
| 90 |
+
"rewards/chosen": 0.0013991177547723055,
|
| 91 |
+
"rewards/margins": 0.0013889471301808953,
|
| 92 |
+
"rewards/rejected": 1.0170697350986302e-05,
|
| 93 |
+
"step": 50
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.09,
|
| 97 |
+
"learning_rate": 4.5454545454545455e-06,
|
| 98 |
+
"logits/chosen": 0.07572797685861588,
|
| 99 |
+
"logits/rejected": 0.2742985785007477,
|
| 100 |
+
"logps/chosen": -371.0773620605469,
|
| 101 |
+
"logps/rejected": -303.0706481933594,
|
| 102 |
+
"loss": 0.3306,
|
| 103 |
+
"rewards/accuracies": 0.75,
|
| 104 |
+
"rewards/chosen": 0.0029931231401860714,
|
| 105 |
+
"rewards/margins": 0.004650537855923176,
|
| 106 |
+
"rewards/rejected": -0.0016574144829064608,
|
| 107 |
+
"step": 60
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.11,
|
| 111 |
+
"learning_rate": 4.999432965739786e-06,
|
| 112 |
+
"logits/chosen": 0.06861326098442078,
|
| 113 |
+
"logits/rejected": 0.25036171078681946,
|
| 114 |
+
"logps/chosen": -321.8224792480469,
|
| 115 |
+
"logps/rejected": -301.9892272949219,
|
| 116 |
+
"loss": 0.3315,
|
| 117 |
+
"rewards/accuracies": 0.699999988079071,
|
| 118 |
+
"rewards/chosen": 0.005247811786830425,
|
| 119 |
+
"rewards/margins": 0.0075549171306192875,
|
| 120 |
+
"rewards/rejected": -0.0023071051109582186,
|
| 121 |
+
"step": 70
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.12,
|
| 125 |
+
"learning_rate": 4.9930567839810125e-06,
|
| 126 |
+
"logits/chosen": 0.0997760072350502,
|
| 127 |
+
"logits/rejected": 0.26540613174438477,
|
| 128 |
+
"logps/chosen": -360.35430908203125,
|
| 129 |
+
"logps/rejected": -307.5221862792969,
|
| 130 |
+
"loss": 0.3184,
|
| 131 |
+
"rewards/accuracies": 0.800000011920929,
|
| 132 |
+
"rewards/chosen": 0.006557592656463385,
|
| 133 |
+
"rewards/margins": 0.01593288779258728,
|
| 134 |
+
"rewards/rejected": -0.009375295601785183,
|
| 135 |
+
"step": 80
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.14,
|
| 139 |
+
"learning_rate": 4.979613761906212e-06,
|
| 140 |
+
"logits/chosen": 0.11980845779180527,
|
| 141 |
+
"logits/rejected": 0.2542170584201813,
|
| 142 |
+
"logps/chosen": -325.69989013671875,
|
| 143 |
+
"logps/rejected": -303.8279724121094,
|
| 144 |
+
"loss": 0.3053,
|
| 145 |
+
"rewards/accuracies": 0.7250000238418579,
|
| 146 |
+
"rewards/chosen": 0.0012294445186853409,
|
| 147 |
+
"rewards/margins": 0.02537659928202629,
|
| 148 |
+
"rewards/rejected": -0.02414715476334095,
|
| 149 |
+
"step": 90
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.15,
|
| 153 |
+
"learning_rate": 4.959142005221991e-06,
|
| 154 |
+
"logits/chosen": 0.15030920505523682,
|
| 155 |
+
"logits/rejected": 0.24269947409629822,
|
| 156 |
+
"logps/chosen": -357.73846435546875,
|
| 157 |
+
"logps/rejected": -372.84625244140625,
|
| 158 |
+
"loss": 0.2998,
|
| 159 |
+
"rewards/accuracies": 0.75,
|
| 160 |
+
"rewards/chosen": -0.03152700141072273,
|
| 161 |
+
"rewards/margins": 0.0535690113902092,
|
| 162 |
+
"rewards/rejected": -0.08509601652622223,
|
| 163 |
+
"step": 100
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.17,
|
| 167 |
+
"learning_rate": 4.931699543346854e-06,
|
| 168 |
+
"logits/chosen": 0.10781173408031464,
|
| 169 |
+
"logits/rejected": 0.28030428290367126,
|
| 170 |
+
"logps/chosen": -428.8057556152344,
|
| 171 |
+
"logps/rejected": -428.8355407714844,
|
| 172 |
+
"loss": 0.2724,
|
| 173 |
+
"rewards/accuracies": 0.675000011920929,
|
| 174 |
+
"rewards/chosen": -0.09748424589633942,
|
| 175 |
+
"rewards/margins": 0.0674622505903244,
|
| 176 |
+
"rewards/rejected": -0.16494649648666382,
|
| 177 |
+
"step": 110
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.18,
|
| 181 |
+
"learning_rate": 4.897364164920515e-06,
|
| 182 |
+
"logits/chosen": 0.1421918272972107,
|
| 183 |
+
"logits/rejected": 0.2562108039855957,
|
| 184 |
+
"logps/chosen": -584.6065673828125,
|
| 185 |
+
"logps/rejected": -657.9326782226562,
|
| 186 |
+
"loss": 0.2475,
|
| 187 |
+
"rewards/accuracies": 0.675000011920929,
|
| 188 |
+
"rewards/chosen": -0.21280452609062195,
|
| 189 |
+
"rewards/margins": 0.1321365237236023,
|
| 190 |
+
"rewards/rejected": -0.34494107961654663,
|
| 191 |
+
"step": 120
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.2,
|
| 195 |
+
"learning_rate": 4.8562331973035396e-06,
|
| 196 |
+
"logits/chosen": 0.1893201768398285,
|
| 197 |
+
"logits/rejected": 0.2765055298805237,
|
| 198 |
+
"logps/chosen": -574.9005737304688,
|
| 199 |
+
"logps/rejected": -700.53125,
|
| 200 |
+
"loss": 0.2567,
|
| 201 |
+
"rewards/accuracies": 0.7250000238418579,
|
| 202 |
+
"rewards/chosen": -0.23022952675819397,
|
| 203 |
+
"rewards/margins": 0.17663846909999847,
|
| 204 |
+
"rewards/rejected": -0.40686798095703125,
|
| 205 |
+
"step": 130
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.21,
|
| 209 |
+
"learning_rate": 4.808423230692374e-06,
|
| 210 |
+
"logits/chosen": 0.15862275660037994,
|
| 211 |
+
"logits/rejected": 0.3002353310585022,
|
| 212 |
+
"logps/chosen": -557.1515502929688,
|
| 213 |
+
"logps/rejected": -636.1343994140625,
|
| 214 |
+
"loss": 0.2578,
|
| 215 |
+
"rewards/accuracies": 0.71875,
|
| 216 |
+
"rewards/chosen": -0.2007971704006195,
|
| 217 |
+
"rewards/margins": 0.1420392543077469,
|
| 218 |
+
"rewards/rejected": -0.3428364396095276,
|
| 219 |
+
"step": 140
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.23,
|
| 223 |
+
"learning_rate": 4.754069787631761e-06,
|
| 224 |
+
"logits/chosen": 0.20283110439777374,
|
| 225 |
+
"logits/rejected": 0.2838110327720642,
|
| 226 |
+
"logps/chosen": -539.7517700195312,
|
| 227 |
+
"logps/rejected": -632.3270263671875,
|
| 228 |
+
"loss": 0.2807,
|
| 229 |
+
"rewards/accuracies": 0.6812499761581421,
|
| 230 |
+
"rewards/chosen": -0.22490541636943817,
|
| 231 |
+
"rewards/margins": 0.13436779379844666,
|
| 232 |
+
"rewards/rejected": -0.359273225069046,
|
| 233 |
+
"step": 150
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.24,
|
| 237 |
+
"learning_rate": 4.693326938861367e-06,
|
| 238 |
+
"logits/chosen": 0.16693079471588135,
|
| 239 |
+
"logits/rejected": 0.3414613604545593,
|
| 240 |
+
"logps/chosen": -519.4967651367188,
|
| 241 |
+
"logps/rejected": -644.0454711914062,
|
| 242 |
+
"loss": 0.2302,
|
| 243 |
+
"rewards/accuracies": 0.75,
|
| 244 |
+
"rewards/chosen": -0.1742098033428192,
|
| 245 |
+
"rewards/margins": 0.19720463454723358,
|
| 246 |
+
"rewards/rejected": -0.3714144825935364,
|
| 247 |
+
"step": 160
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.26,
|
| 251 |
+
"learning_rate": 4.626366866585528e-06,
|
| 252 |
+
"logits/chosen": 0.15613974630832672,
|
| 253 |
+
"logits/rejected": 0.27570822834968567,
|
| 254 |
+
"logps/chosen": -551.5475463867188,
|
| 255 |
+
"logps/rejected": -641.4393310546875,
|
| 256 |
+
"loss": 0.2342,
|
| 257 |
+
"rewards/accuracies": 0.762499988079071,
|
| 258 |
+
"rewards/chosen": -0.18619480729103088,
|
| 259 |
+
"rewards/margins": 0.17629162967205048,
|
| 260 |
+
"rewards/rejected": -0.36248645186424255,
|
| 261 |
+
"step": 170
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.27,
|
| 265 |
+
"learning_rate": 4.553379376404085e-06,
|
| 266 |
+
"logits/chosen": 0.15913400053977966,
|
| 267 |
+
"logits/rejected": 0.28868401050567627,
|
| 268 |
+
"logps/chosen": -575.120361328125,
|
| 269 |
+
"logps/rejected": -672.7054443359375,
|
| 270 |
+
"loss": 0.2586,
|
| 271 |
+
"rewards/accuracies": 0.7124999761581421,
|
| 272 |
+
"rewards/chosen": -0.23076090216636658,
|
| 273 |
+
"rewards/margins": 0.1531594842672348,
|
| 274 |
+
"rewards/rejected": -0.3839203715324402,
|
| 275 |
+
"step": 180
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.29,
|
| 279 |
+
"learning_rate": 4.474571359287791e-06,
|
| 280 |
+
"logits/chosen": 0.2370852530002594,
|
| 281 |
+
"logits/rejected": 0.313865065574646,
|
| 282 |
+
"logps/chosen": -578.9676513671875,
|
| 283 |
+
"logps/rejected": -646.5762329101562,
|
| 284 |
+
"loss": 0.2759,
|
| 285 |
+
"rewards/accuracies": 0.731249988079071,
|
| 286 |
+
"rewards/chosen": -0.22910158336162567,
|
| 287 |
+
"rewards/margins": 0.14492908120155334,
|
| 288 |
+
"rewards/rejected": -0.3740306496620178,
|
| 289 |
+
"step": 190
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.3,
|
| 293 |
+
"learning_rate": 4.3901662051233755e-06,
|
| 294 |
+
"logits/chosen": 0.156154602766037,
|
| 295 |
+
"logits/rejected": 0.28848981857299805,
|
| 296 |
+
"logps/chosen": -529.0174560546875,
|
| 297 |
+
"logps/rejected": -681.0127563476562,
|
| 298 |
+
"loss": 0.2411,
|
| 299 |
+
"rewards/accuracies": 0.71875,
|
| 300 |
+
"rewards/chosen": -0.21960540115833282,
|
| 301 |
+
"rewards/margins": 0.1676800698041916,
|
| 302 |
+
"rewards/rejected": -0.3872854709625244,
|
| 303 |
+
"step": 200
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.32,
|
| 307 |
+
"learning_rate": 4.30040316949064e-06,
|
| 308 |
+
"logits/chosen": 0.13694807887077332,
|
| 309 |
+
"logits/rejected": 0.3102852702140808,
|
| 310 |
+
"logps/chosen": -582.1616821289062,
|
| 311 |
+
"logps/rejected": -681.3240966796875,
|
| 312 |
+
"loss": 0.2166,
|
| 313 |
+
"rewards/accuracies": 0.7562500238418579,
|
| 314 |
+
"rewards/chosen": -0.24012689292430878,
|
| 315 |
+
"rewards/margins": 0.18257644772529602,
|
| 316 |
+
"rewards/rejected": -0.422703355550766,
|
| 317 |
+
"step": 210
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.34,
|
| 321 |
+
"learning_rate": 4.205536695466524e-06,
|
| 322 |
+
"logits/chosen": 0.15840545296669006,
|
| 323 |
+
"logits/rejected": 0.30105945467948914,
|
| 324 |
+
"logps/chosen": -578.1398315429688,
|
| 325 |
+
"logps/rejected": -766.7618408203125,
|
| 326 |
+
"loss": 0.2259,
|
| 327 |
+
"rewards/accuracies": 0.7437499761581421,
|
| 328 |
+
"rewards/chosen": -0.2511864900588989,
|
| 329 |
+
"rewards/margins": 0.22828085720539093,
|
| 330 |
+
"rewards/rejected": -0.47946733236312866,
|
| 331 |
+
"step": 220
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.35,
|
| 335 |
+
"learning_rate": 4.105835692378557e-06,
|
| 336 |
+
"logits/chosen": 0.14648036658763885,
|
| 337 |
+
"logits/rejected": 0.21232767403125763,
|
| 338 |
+
"logps/chosen": -583.7035522460938,
|
| 339 |
+
"logps/rejected": -690.9400634765625,
|
| 340 |
+
"loss": 0.2757,
|
| 341 |
+
"rewards/accuracies": 0.675000011920929,
|
| 342 |
+
"rewards/chosen": -0.2628551125526428,
|
| 343 |
+
"rewards/margins": 0.15015873312950134,
|
| 344 |
+
"rewards/rejected": -0.41301384568214417,
|
| 345 |
+
"step": 230
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.37,
|
| 349 |
+
"learning_rate": 4.001582773552153e-06,
|
| 350 |
+
"logits/chosen": 0.13691949844360352,
|
| 351 |
+
"logits/rejected": 0.22931316494941711,
|
| 352 |
+
"logps/chosen": -621.2189331054688,
|
| 353 |
+
"logps/rejected": -718.5030517578125,
|
| 354 |
+
"loss": 0.2518,
|
| 355 |
+
"rewards/accuracies": 0.71875,
|
| 356 |
+
"rewards/chosen": -0.24823768436908722,
|
| 357 |
+
"rewards/margins": 0.16897296905517578,
|
| 358 |
+
"rewards/rejected": -0.4172106683254242,
|
| 359 |
+
"step": 240
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.38,
|
| 363 |
+
"learning_rate": 3.893073455212438e-06,
|
| 364 |
+
"logits/chosen": 0.15518508851528168,
|
| 365 |
+
"logits/rejected": 0.27118122577667236,
|
| 366 |
+
"logps/chosen": -649.8485107421875,
|
| 367 |
+
"logps/rejected": -743.3021240234375,
|
| 368 |
+
"loss": 0.2446,
|
| 369 |
+
"rewards/accuracies": 0.7562500238418579,
|
| 370 |
+
"rewards/chosen": -0.27570152282714844,
|
| 371 |
+
"rewards/margins": 0.17269738018512726,
|
| 372 |
+
"rewards/rejected": -0.4483988881111145,
|
| 373 |
+
"step": 250
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.4,
|
| 377 |
+
"learning_rate": 3.7806153188114027e-06,
|
| 378 |
+
"logits/chosen": 0.14696967601776123,
|
| 379 |
+
"logits/rejected": 0.26466238498687744,
|
| 380 |
+
"logps/chosen": -585.2598876953125,
|
| 381 |
+
"logps/rejected": -646.9161376953125,
|
| 382 |
+
"loss": 0.2236,
|
| 383 |
+
"rewards/accuracies": 0.731249988079071,
|
| 384 |
+
"rewards/chosen": -0.22980618476867676,
|
| 385 |
+
"rewards/margins": 0.16329681873321533,
|
| 386 |
+
"rewards/rejected": -0.3931030035018921,
|
| 387 |
+
"step": 260
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.41,
|
| 391 |
+
"learning_rate": 3.6645271391548542e-06,
|
| 392 |
+
"logits/chosen": 0.1697523295879364,
|
| 393 |
+
"logits/rejected": 0.29494693875312805,
|
| 394 |
+
"logps/chosen": -613.8264770507812,
|
| 395 |
+
"logps/rejected": -749.4378051757812,
|
| 396 |
+
"loss": 0.228,
|
| 397 |
+
"rewards/accuracies": 0.7749999761581421,
|
| 398 |
+
"rewards/chosen": -0.2746888995170593,
|
| 399 |
+
"rewards/margins": 0.2020600140094757,
|
| 400 |
+
"rewards/rejected": -0.47674888372421265,
|
| 401 |
+
"step": 270
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.43,
|
| 405 |
+
"learning_rate": 3.5451379808006014e-06,
|
| 406 |
+
"logits/chosen": 0.13133810460567474,
|
| 407 |
+
"logits/rejected": 0.2865068316459656,
|
| 408 |
+
"logps/chosen": -590.5033569335938,
|
| 409 |
+
"logps/rejected": -722.9710083007812,
|
| 410 |
+
"loss": 0.2252,
|
| 411 |
+
"rewards/accuracies": 0.7250000238418579,
|
| 412 |
+
"rewards/chosen": -0.2536476254463196,
|
| 413 |
+
"rewards/margins": 0.177708238363266,
|
| 414 |
+
"rewards/rejected": -0.43135586380958557,
|
| 415 |
+
"step": 280
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.44,
|
| 419 |
+
"learning_rate": 3.4227862652892106e-06,
|
| 420 |
+
"logits/chosen": 0.16941356658935547,
|
| 421 |
+
"logits/rejected": 0.30433687567710876,
|
| 422 |
+
"logps/chosen": -602.0320434570312,
|
| 423 |
+
"logps/rejected": -731.9866333007812,
|
| 424 |
+
"loss": 0.2518,
|
| 425 |
+
"rewards/accuracies": 0.699999988079071,
|
| 426 |
+
"rewards/chosen": -0.262613981962204,
|
| 427 |
+
"rewards/margins": 0.17090369760990143,
|
| 428 |
+
"rewards/rejected": -0.4335176944732666,
|
| 429 |
+
"step": 290
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.46,
|
| 433 |
+
"learning_rate": 3.2978188118513814e-06,
|
| 434 |
+
"logits/chosen": 0.1636120229959488,
|
| 435 |
+
"logits/rejected": 0.2873205840587616,
|
| 436 |
+
"logps/chosen": -595.1837768554688,
|
| 437 |
+
"logps/rejected": -764.756103515625,
|
| 438 |
+
"loss": 0.218,
|
| 439 |
+
"rewards/accuracies": 0.7749999761581421,
|
| 440 |
+
"rewards/chosen": -0.25328361988067627,
|
| 441 |
+
"rewards/margins": 0.21258826553821564,
|
| 442 |
+
"rewards/rejected": -0.4658718705177307,
|
| 443 |
+
"step": 300
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.47,
|
| 447 |
+
"learning_rate": 3.1705898543111576e-06,
|
| 448 |
+
"logits/chosen": 0.128164142370224,
|
| 449 |
+
"logits/rejected": 0.29025566577911377,
|
| 450 |
+
"logps/chosen": -615.720947265625,
|
| 451 |
+
"logps/rejected": -729.6194458007812,
|
| 452 |
+
"loss": 0.2587,
|
| 453 |
+
"rewards/accuracies": 0.7250000238418579,
|
| 454 |
+
"rewards/chosen": -0.2794512212276459,
|
| 455 |
+
"rewards/margins": 0.1740822196006775,
|
| 456 |
+
"rewards/rejected": -0.45353350043296814,
|
| 457 |
+
"step": 310
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.49,
|
| 461 |
+
"learning_rate": 3.041460036971664e-06,
|
| 462 |
+
"logits/chosen": 0.15139839053153992,
|
| 463 |
+
"logits/rejected": 0.3216419517993927,
|
| 464 |
+
"logps/chosen": -633.3115234375,
|
| 465 |
+
"logps/rejected": -791.7572631835938,
|
| 466 |
+
"loss": 0.2391,
|
| 467 |
+
"rewards/accuracies": 0.7250000238418579,
|
| 468 |
+
"rewards/chosen": -0.3004799485206604,
|
| 469 |
+
"rewards/margins": 0.18833482265472412,
|
| 470 |
+
"rewards/rejected": -0.4888147711753845,
|
| 471 |
+
"step": 320
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.5,
|
| 475 |
+
"learning_rate": 2.910795392329649e-06,
|
| 476 |
+
"logits/chosen": 0.21717897057533264,
|
| 477 |
+
"logits/rejected": 0.3525051474571228,
|
| 478 |
+
"logps/chosen": -616.4713134765625,
|
| 479 |
+
"logps/rejected": -754.4976806640625,
|
| 480 |
+
"loss": 0.2592,
|
| 481 |
+
"rewards/accuracies": 0.737500011920929,
|
| 482 |
+
"rewards/chosen": -0.2672343850135803,
|
| 483 |
+
"rewards/margins": 0.2101636379957199,
|
| 484 |
+
"rewards/rejected": -0.47739800810813904,
|
| 485 |
+
"step": 330
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.52,
|
| 489 |
+
"learning_rate": 2.7789663035166035e-06,
|
| 490 |
+
"logits/chosen": 0.2043914794921875,
|
| 491 |
+
"logits/rejected": 0.34597498178482056,
|
| 492 |
+
"logps/chosen": -658.1500244140625,
|
| 493 |
+
"logps/rejected": -802.0507202148438,
|
| 494 |
+
"loss": 0.2329,
|
| 495 |
+
"rewards/accuracies": 0.7437499761581421,
|
| 496 |
+
"rewards/chosen": -0.2817768454551697,
|
| 497 |
+
"rewards/margins": 0.2174309492111206,
|
| 498 |
+
"rewards/rejected": -0.4992077946662903,
|
| 499 |
+
"step": 340
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.53,
|
| 503 |
+
"learning_rate": 2.6463464544075344e-06,
|
| 504 |
+
"logits/chosen": 0.16284213960170746,
|
| 505 |
+
"logits/rejected": 0.28806525468826294,
|
| 506 |
+
"logps/chosen": -627.6129760742188,
|
| 507 |
+
"logps/rejected": -686.5743408203125,
|
| 508 |
+
"loss": 0.2372,
|
| 509 |
+
"rewards/accuracies": 0.7250000238418579,
|
| 510 |
+
"rewards/chosen": -0.2634487748146057,
|
| 511 |
+
"rewards/margins": 0.15015827119350433,
|
| 512 |
+
"rewards/rejected": -0.41360706090927124,
|
| 513 |
+
"step": 350
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.55,
|
| 517 |
+
"learning_rate": 2.513311770373421e-06,
|
| 518 |
+
"logits/chosen": 0.14228633046150208,
|
| 519 |
+
"logits/rejected": 0.26139333844184875,
|
| 520 |
+
"logps/chosen": -625.71875,
|
| 521 |
+
"logps/rejected": -751.2933349609375,
|
| 522 |
+
"loss": 0.2357,
|
| 523 |
+
"rewards/accuracies": 0.768750011920929,
|
| 524 |
+
"rewards/chosen": -0.26470065116882324,
|
| 525 |
+
"rewards/margins": 0.1807432472705841,
|
| 526 |
+
"rewards/rejected": -0.44544392824172974,
|
| 527 |
+
"step": 360
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.56,
|
| 531 |
+
"learning_rate": 2.380239352679908e-06,
|
| 532 |
+
"logits/chosen": 0.14897385239601135,
|
| 533 |
+
"logits/rejected": 0.2588959336280823,
|
| 534 |
+
"logps/chosen": -641.2374267578125,
|
| 535 |
+
"logps/rejected": -802.7953491210938,
|
| 536 |
+
"loss": 0.2166,
|
| 537 |
+
"rewards/accuracies": 0.7875000238418579,
|
| 538 |
+
"rewards/chosen": -0.30122682452201843,
|
| 539 |
+
"rewards/margins": 0.1971724033355713,
|
| 540 |
+
"rewards/rejected": -0.49839919805526733,
|
| 541 |
+
"step": 370
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.58,
|
| 545 |
+
"learning_rate": 2.247506409552795e-06,
|
| 546 |
+
"logits/chosen": 0.14793309569358826,
|
| 547 |
+
"logits/rejected": 0.3080625534057617,
|
| 548 |
+
"logps/chosen": -668.0740966796875,
|
| 549 |
+
"logps/rejected": -816.8853759765625,
|
| 550 |
+
"loss": 0.2162,
|
| 551 |
+
"rewards/accuracies": 0.7562500238418579,
|
| 552 |
+
"rewards/chosen": -0.3042776584625244,
|
| 553 |
+
"rewards/margins": 0.20680758357048035,
|
| 554 |
+
"rewards/rejected": -0.5110852122306824,
|
| 555 |
+
"step": 380
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.59,
|
| 559 |
+
"learning_rate": 2.1154891869403436e-06,
|
| 560 |
+
"logits/chosen": 0.1291504055261612,
|
| 561 |
+
"logits/rejected": 0.3194652497768402,
|
| 562 |
+
"logps/chosen": -662.1779174804688,
|
| 563 |
+
"logps/rejected": -797.3854370117188,
|
| 564 |
+
"loss": 0.2264,
|
| 565 |
+
"rewards/accuracies": 0.7437499761581421,
|
| 566 |
+
"rewards/chosen": -0.33526724576950073,
|
| 567 |
+
"rewards/margins": 0.19740521907806396,
|
| 568 |
+
"rewards/rejected": -0.5326724648475647,
|
| 569 |
+
"step": 390
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.61,
|
| 573 |
+
"learning_rate": 1.9845619020032552e-06,
|
| 574 |
+
"logits/chosen": 0.16021743416786194,
|
| 575 |
+
"logits/rejected": 0.3056327700614929,
|
| 576 |
+
"logps/chosen": -644.52734375,
|
| 577 |
+
"logps/rejected": -788.876953125,
|
| 578 |
+
"loss": 0.2151,
|
| 579 |
+
"rewards/accuracies": 0.706250011920929,
|
| 580 |
+
"rewards/chosen": -0.3085401952266693,
|
| 581 |
+
"rewards/margins": 0.211081862449646,
|
| 582 |
+
"rewards/rejected": -0.5196221470832825,
|
| 583 |
+
"step": 400
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.62,
|
| 587 |
+
"learning_rate": 1.8550956823554708e-06,
|
| 588 |
+
"logits/chosen": 0.15647678077220917,
|
| 589 |
+
"logits/rejected": 0.3057996332645416,
|
| 590 |
+
"logps/chosen": -702.3175048828125,
|
| 591 |
+
"logps/rejected": -853.0869140625,
|
| 592 |
+
"loss": 0.2227,
|
| 593 |
+
"rewards/accuracies": 0.75,
|
| 594 |
+
"rewards/chosen": -0.3388286232948303,
|
| 595 |
+
"rewards/margins": 0.2211228311061859,
|
| 596 |
+
"rewards/rejected": -0.5599514842033386,
|
| 597 |
+
"step": 410
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.64,
|
| 601 |
+
"learning_rate": 1.7274575140626318e-06,
|
| 602 |
+
"logits/chosen": 0.14826330542564392,
|
| 603 |
+
"logits/rejected": 0.32510313391685486,
|
| 604 |
+
"logps/chosen": -669.1925659179688,
|
| 605 |
+
"logps/rejected": -814.4791870117188,
|
| 606 |
+
"loss": 0.2161,
|
| 607 |
+
"rewards/accuracies": 0.737500011920929,
|
| 608 |
+
"rewards/chosen": -0.308666467666626,
|
| 609 |
+
"rewards/margins": 0.18964755535125732,
|
| 610 |
+
"rewards/rejected": -0.4983140528202057,
|
| 611 |
+
"step": 420
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.66,
|
| 615 |
+
"learning_rate": 1.6020092013802002e-06,
|
| 616 |
+
"logits/chosen": 0.2039145529270172,
|
| 617 |
+
"logits/rejected": 0.27107498049736023,
|
| 618 |
+
"logps/chosen": -734.4779052734375,
|
| 619 |
+
"logps/rejected": -801.9241333007812,
|
| 620 |
+
"loss": 0.2567,
|
| 621 |
+
"rewards/accuracies": 0.71875,
|
| 622 |
+
"rewards/chosen": -0.3436730206012726,
|
| 623 |
+
"rewards/margins": 0.15554025769233704,
|
| 624 |
+
"rewards/rejected": -0.499213308095932,
|
| 625 |
+
"step": 430
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.67,
|
| 629 |
+
"learning_rate": 1.4791063411799938e-06,
|
| 630 |
+
"logits/chosen": 0.1372929960489273,
|
| 631 |
+
"logits/rejected": 0.2705134451389313,
|
| 632 |
+
"logps/chosen": -655.2030029296875,
|
| 633 |
+
"logps/rejected": -761.6964111328125,
|
| 634 |
+
"loss": 0.2425,
|
| 635 |
+
"rewards/accuracies": 0.699999988079071,
|
| 636 |
+
"rewards/chosen": -0.2905011475086212,
|
| 637 |
+
"rewards/margins": 0.1854308694601059,
|
| 638 |
+
"rewards/rejected": -0.4759320318698883,
|
| 639 |
+
"step": 440
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.69,
|
| 643 |
+
"learning_rate": 1.3590973149722103e-06,
|
| 644 |
+
"logits/chosen": 0.16004619002342224,
|
| 645 |
+
"logits/rejected": 0.3311876654624939,
|
| 646 |
+
"logps/chosen": -623.554443359375,
|
| 647 |
+
"logps/rejected": -741.9662475585938,
|
| 648 |
+
"loss": 0.2205,
|
| 649 |
+
"rewards/accuracies": 0.7562500238418579,
|
| 650 |
+
"rewards/chosen": -0.2771863043308258,
|
| 651 |
+
"rewards/margins": 0.18584686517715454,
|
| 652 |
+
"rewards/rejected": -0.46303310990333557,
|
| 653 |
+
"step": 450
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.7,
|
| 657 |
+
"learning_rate": 1.2423223013801946e-06,
|
| 658 |
+
"logits/chosen": 0.15525056421756744,
|
| 659 |
+
"logits/rejected": 0.226671501994133,
|
| 660 |
+
"logps/chosen": -620.3182373046875,
|
| 661 |
+
"logps/rejected": -753.6258544921875,
|
| 662 |
+
"loss": 0.2323,
|
| 663 |
+
"rewards/accuracies": 0.7124999761581421,
|
| 664 |
+
"rewards/chosen": -0.28617820143699646,
|
| 665 |
+
"rewards/margins": 0.17914626002311707,
|
| 666 |
+
"rewards/rejected": -0.4653244912624359,
|
| 667 |
+
"step": 460
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.72,
|
| 671 |
+
"learning_rate": 1.1291123118671665e-06,
|
| 672 |
+
"logits/chosen": 0.1403166949748993,
|
| 673 |
+
"logits/rejected": 0.21838609874248505,
|
| 674 |
+
"logps/chosen": -640.9824829101562,
|
| 675 |
+
"logps/rejected": -792.9696044921875,
|
| 676 |
+
"loss": 0.23,
|
| 677 |
+
"rewards/accuracies": 0.7562500238418579,
|
| 678 |
+
"rewards/chosen": -0.281258225440979,
|
| 679 |
+
"rewards/margins": 0.21520480513572693,
|
| 680 |
+
"rewards/rejected": -0.4964630603790283,
|
| 681 |
+
"step": 470
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.73,
|
| 685 |
+
"learning_rate": 1.019788252448267e-06,
|
| 686 |
+
"logits/chosen": 0.12808464467525482,
|
| 687 |
+
"logits/rejected": 0.2763553857803345,
|
| 688 |
+
"logps/chosen": -644.160888671875,
|
| 689 |
+
"logps/rejected": -773.2025146484375,
|
| 690 |
+
"loss": 0.2376,
|
| 691 |
+
"rewards/accuracies": 0.75,
|
| 692 |
+
"rewards/chosen": -0.31234392523765564,
|
| 693 |
+
"rewards/margins": 0.19615283608436584,
|
| 694 |
+
"rewards/rejected": -0.5084967017173767,
|
| 695 |
+
"step": 480
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.75,
|
| 699 |
+
"learning_rate": 9.146600140475945e-07,
|
| 700 |
+
"logits/chosen": 0.1601841151714325,
|
| 701 |
+
"logits/rejected": 0.22859685122966766,
|
| 702 |
+
"logps/chosen": -651.9871826171875,
|
| 703 |
+
"logps/rejected": -752.2882080078125,
|
| 704 |
+
"loss": 0.2404,
|
| 705 |
+
"rewards/accuracies": 0.75,
|
| 706 |
+
"rewards/chosen": -0.3149036467075348,
|
| 707 |
+
"rewards/margins": 0.15418918430805206,
|
| 708 |
+
"rewards/rejected": -0.46909284591674805,
|
| 709 |
+
"step": 490
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.76,
|
| 713 |
+
"learning_rate": 8.140255940787059e-07,
|
| 714 |
+
"logits/chosen": 0.1696256846189499,
|
| 715 |
+
"logits/rejected": 0.2493252456188202,
|
| 716 |
+
"logps/chosen": -615.3719482421875,
|
| 717 |
+
"logps/rejected": -736.1841430664062,
|
| 718 |
+
"loss": 0.2331,
|
| 719 |
+
"rewards/accuracies": 0.762499988079071,
|
| 720 |
+
"rewards/chosen": -0.2796221375465393,
|
| 721 |
+
"rewards/margins": 0.1891065537929535,
|
| 722 |
+
"rewards/rejected": -0.4687287211418152,
|
| 723 |
+
"step": 500
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.78,
|
| 727 |
+
"learning_rate": 7.181702517385789e-07,
|
| 728 |
+
"logits/chosen": 0.13488708436489105,
|
| 729 |
+
"logits/rejected": 0.24344106018543243,
|
| 730 |
+
"logps/chosen": -632.91064453125,
|
| 731 |
+
"logps/rejected": -779.7465209960938,
|
| 732 |
+
"loss": 0.2493,
|
| 733 |
+
"rewards/accuracies": 0.699999988079071,
|
| 734 |
+
"rewards/chosen": -0.30229493975639343,
|
| 735 |
+
"rewards/margins": 0.19277289509773254,
|
| 736 |
+
"rewards/rejected": -0.49506789445877075,
|
| 737 |
+
"step": 510
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.79,
|
| 741 |
+
"learning_rate": 6.273656994094232e-07,
|
| 742 |
+
"logits/chosen": 0.078878253698349,
|
| 743 |
+
"logits/rejected": 0.32078155875205994,
|
| 744 |
+
"logps/chosen": -629.8251342773438,
|
| 745 |
+
"logps/rejected": -792.5658569335938,
|
| 746 |
+
"loss": 0.2593,
|
| 747 |
+
"rewards/accuracies": 0.78125,
|
| 748 |
+
"rewards/chosen": -0.29794004559516907,
|
| 749 |
+
"rewards/margins": 0.21232935786247253,
|
| 750 |
+
"rewards/rejected": -0.5102694034576416,
|
| 751 |
+
"step": 520
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.81,
|
| 755 |
+
"learning_rate": 5.418693324604082e-07,
|
| 756 |
+
"logits/chosen": 0.17152568697929382,
|
| 757 |
+
"logits/rejected": 0.2723831534385681,
|
| 758 |
+
"logps/chosen": -686.8225708007812,
|
| 759 |
+
"logps/rejected": -765.6764526367188,
|
| 760 |
+
"loss": 0.2106,
|
| 761 |
+
"rewards/accuracies": 0.762499988079071,
|
| 762 |
+
"rewards/chosen": -0.2921481132507324,
|
| 763 |
+
"rewards/margins": 0.19477225840091705,
|
| 764 |
+
"rewards/rejected": -0.48692041635513306,
|
| 765 |
+
"step": 530
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.82,
|
| 769 |
+
"learning_rate": 4.619234996325314e-07,
|
| 770 |
+
"logits/chosen": 0.16223089396953583,
|
| 771 |
+
"logits/rejected": 0.22344419360160828,
|
| 772 |
+
"logps/chosen": -644.6939086914062,
|
| 773 |
+
"logps/rejected": -796.0565185546875,
|
| 774 |
+
"loss": 0.2251,
|
| 775 |
+
"rewards/accuracies": 0.7562500238418579,
|
| 776 |
+
"rewards/chosen": -0.29831987619400024,
|
| 777 |
+
"rewards/margins": 0.2159034013748169,
|
| 778 |
+
"rewards/rejected": -0.5142232179641724,
|
| 779 |
+
"step": 540
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.84,
|
| 783 |
+
"learning_rate": 3.877548160747768e-07,
|
| 784 |
+
"logits/chosen": 0.14092543721199036,
|
| 785 |
+
"logits/rejected": 0.26907533407211304,
|
| 786 |
+
"logps/chosen": -665.1543579101562,
|
| 787 |
+
"logps/rejected": -785.69091796875,
|
| 788 |
+
"loss": 0.2293,
|
| 789 |
+
"rewards/accuracies": 0.7562500238418579,
|
| 790 |
+
"rewards/chosen": -0.3221438229084015,
|
| 791 |
+
"rewards/margins": 0.16273698210716248,
|
| 792 |
+
"rewards/rejected": -0.4848807752132416,
|
| 793 |
+
"step": 550
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.85,
|
| 797 |
+
"learning_rate": 3.195735209788528e-07,
|
| 798 |
+
"logits/chosen": 0.12745890021324158,
|
| 799 |
+
"logits/rejected": 0.16941148042678833,
|
| 800 |
+
"logps/chosen": -604.4722290039062,
|
| 801 |
+
"logps/rejected": -726.4368286132812,
|
| 802 |
+
"loss": 0.228,
|
| 803 |
+
"rewards/accuracies": 0.6812499761581421,
|
| 804 |
+
"rewards/chosen": -0.2934458255767822,
|
| 805 |
+
"rewards/margins": 0.16752712428569794,
|
| 806 |
+
"rewards/rejected": -0.46097296476364136,
|
| 807 |
+
"step": 560
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.87,
|
| 811 |
+
"learning_rate": 2.5757288163336806e-07,
|
| 812 |
+
"logits/chosen": 0.14064130187034607,
|
| 813 |
+
"logits/rejected": 0.2834423780441284,
|
| 814 |
+
"logps/chosen": -670.5109252929688,
|
| 815 |
+
"logps/rejected": -754.9119873046875,
|
| 816 |
+
"loss": 0.2279,
|
| 817 |
+
"rewards/accuracies": 0.75,
|
| 818 |
+
"rewards/chosen": -0.29777050018310547,
|
| 819 |
+
"rewards/margins": 0.1834956705570221,
|
| 820 |
+
"rewards/rejected": -0.4812661111354828,
|
| 821 |
+
"step": 570
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.88,
|
| 825 |
+
"learning_rate": 2.019286455866981e-07,
|
| 826 |
+
"logits/chosen": 0.16343867778778076,
|
| 827 |
+
"logits/rejected": 0.23200741410255432,
|
| 828 |
+
"logps/chosen": -653.3126220703125,
|
| 829 |
+
"logps/rejected": -788.0684204101562,
|
| 830 |
+
"loss": 0.2129,
|
| 831 |
+
"rewards/accuracies": 0.7875000238418579,
|
| 832 |
+
"rewards/chosen": -0.30145275592803955,
|
| 833 |
+
"rewards/margins": 0.18838538229465485,
|
| 834 |
+
"rewards/rejected": -0.4898381233215332,
|
| 835 |
+
"step": 580
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.9,
|
| 839 |
+
"learning_rate": 1.5279854247146703e-07,
|
| 840 |
+
"logits/chosen": 0.0768275260925293,
|
| 841 |
+
"logits/rejected": 0.261466920375824,
|
| 842 |
+
"logps/chosen": -653.1698608398438,
|
| 843 |
+
"logps/rejected": -781.791259765625,
|
| 844 |
+
"loss": 0.2128,
|
| 845 |
+
"rewards/accuracies": 0.7250000238418579,
|
| 846 |
+
"rewards/chosen": -0.28429535031318665,
|
| 847 |
+
"rewards/margins": 0.22277021408081055,
|
| 848 |
+
"rewards/rejected": -0.5070655941963196,
|
| 849 |
+
"step": 590
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.91,
|
| 853 |
+
"learning_rate": 1.1032183690276754e-07,
|
| 854 |
+
"logits/chosen": 0.160492941737175,
|
| 855 |
+
"logits/rejected": 0.28358930349349976,
|
| 856 |
+
"logps/chosen": -716.923583984375,
|
| 857 |
+
"logps/rejected": -817.1744995117188,
|
| 858 |
+
"loss": 0.2252,
|
| 859 |
+
"rewards/accuracies": 0.737500011920929,
|
| 860 |
+
"rewards/chosen": -0.3216014504432678,
|
| 861 |
+
"rewards/margins": 0.18884047865867615,
|
| 862 |
+
"rewards/rejected": -0.5104418992996216,
|
| 863 |
+
"step": 600
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.93,
|
| 867 |
+
"learning_rate": 7.46189337174788e-08,
|
| 868 |
+
"logits/chosen": 0.138215571641922,
|
| 869 |
+
"logits/rejected": 0.28554660081863403,
|
| 870 |
+
"logps/chosen": -634.0874633789062,
|
| 871 |
+
"logps/rejected": -762.0772094726562,
|
| 872 |
+
"loss": 0.204,
|
| 873 |
+
"rewards/accuracies": 0.75,
|
| 874 |
+
"rewards/chosen": -0.2693154215812683,
|
| 875 |
+
"rewards/margins": 0.21418070793151855,
|
| 876 |
+
"rewards/rejected": -0.48349618911743164,
|
| 877 |
+
"step": 610
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.94,
|
| 881 |
+
"learning_rate": 4.579103667367385e-08,
|
| 882 |
+
"logits/chosen": 0.13104286789894104,
|
| 883 |
+
"logits/rejected": 0.280353844165802,
|
| 884 |
+
"logps/chosen": -666.0810546875,
|
| 885 |
+
"logps/rejected": -752.5096435546875,
|
| 886 |
+
"loss": 0.229,
|
| 887 |
+
"rewards/accuracies": 0.6812499761581421,
|
| 888 |
+
"rewards/chosen": -0.3018819987773895,
|
| 889 |
+
"rewards/margins": 0.1678927093744278,
|
| 890 |
+
"rewards/rejected": -0.4697747230529785,
|
| 891 |
+
"step": 620
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.96,
|
| 895 |
+
"learning_rate": 2.3919861577572924e-08,
|
| 896 |
+
"logits/chosen": 0.15655803680419922,
|
| 897 |
+
"logits/rejected": 0.32044172286987305,
|
| 898 |
+
"logps/chosen": -707.17138671875,
|
| 899 |
+
"logps/rejected": -806.7648315429688,
|
| 900 |
+
"loss": 0.229,
|
| 901 |
+
"rewards/accuracies": 0.75,
|
| 902 |
+
"rewards/chosen": -0.30425772070884705,
|
| 903 |
+
"rewards/margins": 0.21442052721977234,
|
| 904 |
+
"rewards/rejected": -0.5186783075332642,
|
| 905 |
+
"step": 630
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.98,
|
| 909 |
+
"learning_rate": 9.067404651211808e-09,
|
| 910 |
+
"logits/chosen": 0.146881565451622,
|
| 911 |
+
"logits/rejected": 0.22085854411125183,
|
| 912 |
+
"logps/chosen": -632.6345825195312,
|
| 913 |
+
"logps/rejected": -776.0171508789062,
|
| 914 |
+
"loss": 0.2122,
|
| 915 |
+
"rewards/accuracies": 0.706250011920929,
|
| 916 |
+
"rewards/chosen": -0.2965379059314728,
|
| 917 |
+
"rewards/margins": 0.19161197543144226,
|
| 918 |
+
"rewards/rejected": -0.48814982175827026,
|
| 919 |
+
"step": 640
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 0.99,
|
| 923 |
+
"learning_rate": 1.2757667974155896e-09,
|
| 924 |
+
"logits/chosen": 0.18307064473628998,
|
| 925 |
+
"logits/rejected": 0.2208767831325531,
|
| 926 |
+
"logps/chosen": -612.895751953125,
|
| 927 |
+
"logps/rejected": -768.5109252929688,
|
| 928 |
+
"loss": 0.2581,
|
| 929 |
+
"rewards/accuracies": 0.706250011920929,
|
| 930 |
+
"rewards/chosen": -0.2931486666202545,
|
| 931 |
+
"rewards/margins": 0.18815748393535614,
|
| 932 |
+
"rewards/rejected": -0.4813062250614166,
|
| 933 |
+
"step": 650
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 1.0,
|
| 937 |
+
"step": 656,
|
| 938 |
+
"total_flos": 0.0,
|
| 939 |
+
"train_loss": 0.20043009792159244,
|
| 940 |
+
"train_runtime": 6753.4712,
|
| 941 |
+
"train_samples_per_second": 3.11,
|
| 942 |
+
"train_steps_per_second": 0.097
|
| 943 |
+
}
|
| 944 |
+
],
|
| 945 |
+
"logging_steps": 10,
|
| 946 |
+
"max_steps": 656,
|
| 947 |
+
"num_input_tokens_seen": 0,
|
| 948 |
+
"num_train_epochs": 1,
|
| 949 |
+
"save_steps": 100,
|
| 950 |
+
"total_flos": 0.0,
|
| 951 |
+
"train_batch_size": 4,
|
| 952 |
+
"trial_name": null,
|
| 953 |
+
"trial_params": null
|
| 954 |
+
}
|