Qin Liu
commited on
Commit
•
5259054
1
Parent(s):
b699df4
Model save
Browse files- README.md +62 -0
- all_results.json +9 -0
- train_results.json +9 -0
- trainer_state.json +987 -0
README.md
ADDED
|
@@ -0,0 +1,62 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: llama3
|
| 3 |
+
library_name: peft
|
| 4 |
+
tags:
|
| 5 |
+
- trl
|
| 6 |
+
- dpo
|
| 7 |
+
- generated_from_trainer
|
| 8 |
+
base_model: Jackie999/llama3-sudo-5epochs-tofu_full_sft
|
| 9 |
+
model-index:
|
| 10 |
+
- name: llama3-sudo-dpo-5epochs-forget10mix-1sft-2lora
|
| 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 |
+
# llama3-sudo-dpo-5epochs-forget10mix-1sft-2lora
|
| 18 |
+
|
| 19 |
+
This model is a fine-tuned version of [Jackie999/llama3-sudo-5epochs-tofu_full_sft](https://huggingface.co/Jackie999/llama3-sudo-5epochs-tofu_full_sft) on an unknown 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: 8
|
| 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: 32
|
| 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: 5
|
| 51 |
+
|
| 52 |
+
### Training results
|
| 53 |
+
|
| 54 |
+
|
| 55 |
+
|
| 56 |
+
### Framework versions
|
| 57 |
+
|
| 58 |
+
- PEFT 0.12.0
|
| 59 |
+
- Transformers 4.44.0
|
| 60 |
+
- Pytorch 2.1.2
|
| 61 |
+
- Datasets 3.0.0
|
| 62 |
+
- Tokenizers 0.19.1
|
all_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"total_flos": 0.0,
|
| 4 |
+
"train_loss": 0.11420744479894639,
|
| 5 |
+
"train_runtime": 7032.7914,
|
| 6 |
+
"train_samples": 8000,
|
| 7 |
+
"train_samples_per_second": 5.688,
|
| 8 |
+
"train_steps_per_second": 0.089
|
| 9 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"total_flos": 0.0,
|
| 4 |
+
"train_loss": 0.11420744479894639,
|
| 5 |
+
"train_runtime": 7032.7914,
|
| 6 |
+
"train_samples": 8000,
|
| 7 |
+
"train_samples_per_second": 5.688,
|
| 8 |
+
"train_steps_per_second": 0.089
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,987 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 5.0,
|
| 5 |
+
"eval_steps": 1000,
|
| 6 |
+
"global_step": 625,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.008,
|
| 13 |
+
"grad_norm": 0.5664516818518655,
|
| 14 |
+
"learning_rate": 7.936507936507937e-08,
|
| 15 |
+
"logits/chosen": -1.7671998739242554,
|
| 16 |
+
"logits/rejected": -2.2639822959899902,
|
| 17 |
+
"logps/chosen": -46.430763244628906,
|
| 18 |
+
"logps/rejected": -102.85381317138672,
|
| 19 |
+
"loss": 0.6931,
|
| 20 |
+
"rewards/accuracies": 0.0,
|
| 21 |
+
"rewards/chosen": 0.0,
|
| 22 |
+
"rewards/margins": 0.0,
|
| 23 |
+
"rewards/rejected": 0.0,
|
| 24 |
+
"step": 1
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.08,
|
| 28 |
+
"grad_norm": 0.5940666028150138,
|
| 29 |
+
"learning_rate": 7.936507936507937e-07,
|
| 30 |
+
"logits/chosen": -1.635621428489685,
|
| 31 |
+
"logits/rejected": -2.0303046703338623,
|
| 32 |
+
"logps/chosen": -50.40827941894531,
|
| 33 |
+
"logps/rejected": -98.84134674072266,
|
| 34 |
+
"loss": 0.6929,
|
| 35 |
+
"rewards/accuracies": 0.4930555522441864,
|
| 36 |
+
"rewards/chosen": -0.0004079286882188171,
|
| 37 |
+
"rewards/margins": 0.0003309930325485766,
|
| 38 |
+
"rewards/rejected": -0.0007389216334559023,
|
| 39 |
+
"step": 10
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.16,
|
| 43 |
+
"grad_norm": 0.5693390021480599,
|
| 44 |
+
"learning_rate": 1.5873015873015873e-06,
|
| 45 |
+
"logits/chosen": -1.6660130023956299,
|
| 46 |
+
"logits/rejected": -2.1243433952331543,
|
| 47 |
+
"logps/chosen": -46.99885559082031,
|
| 48 |
+
"logps/rejected": -105.28565979003906,
|
| 49 |
+
"loss": 0.6876,
|
| 50 |
+
"rewards/accuracies": 0.9937499761581421,
|
| 51 |
+
"rewards/chosen": 0.0004832017293665558,
|
| 52 |
+
"rewards/margins": 0.012159029953181744,
|
| 53 |
+
"rewards/rejected": -0.011675829067826271,
|
| 54 |
+
"step": 20
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.24,
|
| 58 |
+
"grad_norm": 0.7544428399092442,
|
| 59 |
+
"learning_rate": 2.380952380952381e-06,
|
| 60 |
+
"logits/chosen": -1.7471415996551514,
|
| 61 |
+
"logits/rejected": -2.1227598190307617,
|
| 62 |
+
"logps/chosen": -50.86139678955078,
|
| 63 |
+
"logps/rejected": -106.50770568847656,
|
| 64 |
+
"loss": 0.6584,
|
| 65 |
+
"rewards/accuracies": 0.925000011920929,
|
| 66 |
+
"rewards/chosen": -0.0016866873484104872,
|
| 67 |
+
"rewards/margins": 0.07014349102973938,
|
| 68 |
+
"rewards/rejected": -0.07183017581701279,
|
| 69 |
+
"step": 30
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.32,
|
| 73 |
+
"grad_norm": 0.7543215495993858,
|
| 74 |
+
"learning_rate": 3.1746031746031746e-06,
|
| 75 |
+
"logits/chosen": -1.6426982879638672,
|
| 76 |
+
"logits/rejected": -2.0793442726135254,
|
| 77 |
+
"logps/chosen": -48.13318634033203,
|
| 78 |
+
"logps/rejected": -131.1614990234375,
|
| 79 |
+
"loss": 0.563,
|
| 80 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 81 |
+
"rewards/chosen": 0.005994443781673908,
|
| 82 |
+
"rewards/margins": 0.3002367615699768,
|
| 83 |
+
"rewards/rejected": -0.29424232244491577,
|
| 84 |
+
"step": 40
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 0.4,
|
| 88 |
+
"grad_norm": 0.4924878173095808,
|
| 89 |
+
"learning_rate": 3.968253968253968e-06,
|
| 90 |
+
"logits/chosen": -1.583770990371704,
|
| 91 |
+
"logits/rejected": -2.0470001697540283,
|
| 92 |
+
"logps/chosen": -47.22152328491211,
|
| 93 |
+
"logps/rejected": -174.95742797851562,
|
| 94 |
+
"loss": 0.4123,
|
| 95 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 96 |
+
"rewards/chosen": 0.01497081108391285,
|
| 97 |
+
"rewards/margins": 0.743770182132721,
|
| 98 |
+
"rewards/rejected": -0.7287994027137756,
|
| 99 |
+
"step": 50
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.48,
|
| 103 |
+
"grad_norm": 0.3454006645724035,
|
| 104 |
+
"learning_rate": 4.761904761904762e-06,
|
| 105 |
+
"logits/chosen": -1.554491639137268,
|
| 106 |
+
"logits/rejected": -1.9409517049789429,
|
| 107 |
+
"logps/chosen": -54.594146728515625,
|
| 108 |
+
"logps/rejected": -216.20999145507812,
|
| 109 |
+
"loss": 0.301,
|
| 110 |
+
"rewards/accuracies": 0.9312499761581421,
|
| 111 |
+
"rewards/chosen": -0.03443662449717522,
|
| 112 |
+
"rewards/margins": 1.137538194656372,
|
| 113 |
+
"rewards/rejected": -1.171974778175354,
|
| 114 |
+
"step": 60
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.56,
|
| 118 |
+
"grad_norm": 0.34001134236428526,
|
| 119 |
+
"learning_rate": 4.998086282661188e-06,
|
| 120 |
+
"logits/chosen": -1.459894061088562,
|
| 121 |
+
"logits/rejected": -1.9039119482040405,
|
| 122 |
+
"logps/chosen": -53.59501266479492,
|
| 123 |
+
"logps/rejected": -295.020751953125,
|
| 124 |
+
"loss": 0.2243,
|
| 125 |
+
"rewards/accuracies": 0.9437500238418579,
|
| 126 |
+
"rewards/chosen": -0.030449170619249344,
|
| 127 |
+
"rewards/margins": 1.905814528465271,
|
| 128 |
+
"rewards/rejected": -1.9362636804580688,
|
| 129 |
+
"step": 70
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.64,
|
| 133 |
+
"grad_norm": 0.15067611382441232,
|
| 134 |
+
"learning_rate": 4.988720025682995e-06,
|
| 135 |
+
"logits/chosen": -1.3985722064971924,
|
| 136 |
+
"logits/rejected": -1.8223779201507568,
|
| 137 |
+
"logps/chosen": -54.72029495239258,
|
| 138 |
+
"logps/rejected": -359.4449462890625,
|
| 139 |
+
"loss": 0.2107,
|
| 140 |
+
"rewards/accuracies": 0.949999988079071,
|
| 141 |
+
"rewards/chosen": -0.03896407037973404,
|
| 142 |
+
"rewards/margins": 2.546111583709717,
|
| 143 |
+
"rewards/rejected": -2.585075855255127,
|
| 144 |
+
"step": 80
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.72,
|
| 148 |
+
"grad_norm": 0.1607243644073366,
|
| 149 |
+
"learning_rate": 4.9715789537359126e-06,
|
| 150 |
+
"logits/chosen": -1.3732717037200928,
|
| 151 |
+
"logits/rejected": -1.775152564048767,
|
| 152 |
+
"logps/chosen": -58.017051696777344,
|
| 153 |
+
"logps/rejected": -388.412109375,
|
| 154 |
+
"loss": 0.1949,
|
| 155 |
+
"rewards/accuracies": 0.9437500238418579,
|
| 156 |
+
"rewards/chosen": -0.06883589923381805,
|
| 157 |
+
"rewards/margins": 2.8034896850585938,
|
| 158 |
+
"rewards/rejected": -2.8723256587982178,
|
| 159 |
+
"step": 90
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.8,
|
| 163 |
+
"grad_norm": 0.23308085253562452,
|
| 164 |
+
"learning_rate": 4.946716615897932e-06,
|
| 165 |
+
"logits/chosen": -1.4088760614395142,
|
| 166 |
+
"logits/rejected": -1.845401406288147,
|
| 167 |
+
"logps/chosen": -50.46847915649414,
|
| 168 |
+
"logps/rejected": -365.43017578125,
|
| 169 |
+
"loss": 0.1946,
|
| 170 |
+
"rewards/accuracies": 0.956250011920929,
|
| 171 |
+
"rewards/chosen": -0.003804679261520505,
|
| 172 |
+
"rewards/margins": 2.6316394805908203,
|
| 173 |
+
"rewards/rejected": -2.635444164276123,
|
| 174 |
+
"step": 100
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 0.88,
|
| 178 |
+
"grad_norm": 0.26044789726674666,
|
| 179 |
+
"learning_rate": 4.9142106826480114e-06,
|
| 180 |
+
"logits/chosen": -1.3341541290283203,
|
| 181 |
+
"logits/rejected": -1.7699487209320068,
|
| 182 |
+
"logps/chosen": -52.57965087890625,
|
| 183 |
+
"logps/rejected": -368.12310791015625,
|
| 184 |
+
"loss": 0.1915,
|
| 185 |
+
"rewards/accuracies": 0.949999988079071,
|
| 186 |
+
"rewards/chosen": -0.02668028511106968,
|
| 187 |
+
"rewards/margins": 2.640017032623291,
|
| 188 |
+
"rewards/rejected": -2.6666972637176514,
|
| 189 |
+
"step": 110
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.96,
|
| 193 |
+
"grad_norm": 0.30708635657559097,
|
| 194 |
+
"learning_rate": 4.874162703221823e-06,
|
| 195 |
+
"logits/chosen": -1.2924244403839111,
|
| 196 |
+
"logits/rejected": -1.7736318111419678,
|
| 197 |
+
"logps/chosen": -38.64341735839844,
|
| 198 |
+
"logps/rejected": -411.8453063964844,
|
| 199 |
+
"loss": 0.1577,
|
| 200 |
+
"rewards/accuracies": 0.9937499761581421,
|
| 201 |
+
"rewards/chosen": 0.08109267055988312,
|
| 202 |
+
"rewards/margins": 3.1581881046295166,
|
| 203 |
+
"rewards/rejected": -3.0770950317382812,
|
| 204 |
+
"step": 120
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 1.04,
|
| 208 |
+
"grad_norm": 0.2936514218907092,
|
| 209 |
+
"learning_rate": 4.826697788369752e-06,
|
| 210 |
+
"logits/chosen": -1.3301618099212646,
|
| 211 |
+
"logits/rejected": -1.7304044961929321,
|
| 212 |
+
"logps/chosen": -57.28544998168945,
|
| 213 |
+
"logps/rejected": -383.7294616699219,
|
| 214 |
+
"loss": 0.1735,
|
| 215 |
+
"rewards/accuracies": 0.9312499761581421,
|
| 216 |
+
"rewards/chosen": -0.060813553631305695,
|
| 217 |
+
"rewards/margins": 2.7762393951416016,
|
| 218 |
+
"rewards/rejected": -2.837052822113037,
|
| 219 |
+
"step": 130
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 1.12,
|
| 223 |
+
"grad_norm": 0.7127879450944201,
|
| 224 |
+
"learning_rate": 4.7719642195082224e-06,
|
| 225 |
+
"logits/chosen": -1.3428221940994263,
|
| 226 |
+
"logits/rejected": -1.8241684436798096,
|
| 227 |
+
"logps/chosen": -49.40485382080078,
|
| 228 |
+
"logps/rejected": -397.02685546875,
|
| 229 |
+
"loss": 0.167,
|
| 230 |
+
"rewards/accuracies": 0.949999988079071,
|
| 231 |
+
"rewards/chosen": 0.005482205655425787,
|
| 232 |
+
"rewards/margins": 2.9638657569885254,
|
| 233 |
+
"rewards/rejected": -2.958383798599243,
|
| 234 |
+
"step": 140
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 1.2,
|
| 238 |
+
"grad_norm": 0.7195419525970039,
|
| 239 |
+
"learning_rate": 4.710132985485355e-06,
|
| 240 |
+
"logits/chosen": -1.2738778591156006,
|
| 241 |
+
"logits/rejected": -1.7257006168365479,
|
| 242 |
+
"logps/chosen": -53.46342849731445,
|
| 243 |
+
"logps/rejected": -413.75091552734375,
|
| 244 |
+
"loss": 0.1777,
|
| 245 |
+
"rewards/accuracies": 0.9312499761581421,
|
| 246 |
+
"rewards/chosen": -0.01903398148715496,
|
| 247 |
+
"rewards/margins": 3.1122589111328125,
|
| 248 |
+
"rewards/rejected": -3.1312928199768066,
|
| 249 |
+
"step": 150
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 1.28,
|
| 253 |
+
"grad_norm": 0.5831467127347119,
|
| 254 |
+
"learning_rate": 4.641397248408122e-06,
|
| 255 |
+
"logits/chosen": -1.3061014413833618,
|
| 256 |
+
"logits/rejected": -1.7850764989852905,
|
| 257 |
+
"logps/chosen": -49.27734375,
|
| 258 |
+
"logps/rejected": -427.78338623046875,
|
| 259 |
+
"loss": 0.128,
|
| 260 |
+
"rewards/accuracies": 0.9437500238418579,
|
| 261 |
+
"rewards/chosen": 0.00814875029027462,
|
| 262 |
+
"rewards/margins": 3.2720084190368652,
|
| 263 |
+
"rewards/rejected": -3.263859510421753,
|
| 264 |
+
"step": 160
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 1.3599999999999999,
|
| 268 |
+
"grad_norm": 0.37604055418234283,
|
| 269 |
+
"learning_rate": 4.5659717401997655e-06,
|
| 270 |
+
"logits/chosen": -1.2165374755859375,
|
| 271 |
+
"logits/rejected": -1.7296428680419922,
|
| 272 |
+
"logps/chosen": -43.91862106323242,
|
| 273 |
+
"logps/rejected": -461.78021240234375,
|
| 274 |
+
"loss": 0.1334,
|
| 275 |
+
"rewards/accuracies": 0.96875,
|
| 276 |
+
"rewards/chosen": 0.04128595441579819,
|
| 277 |
+
"rewards/margins": 3.6384873390197754,
|
| 278 |
+
"rewards/rejected": -3.597201108932495,
|
| 279 |
+
"step": 170
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 1.44,
|
| 283 |
+
"grad_norm": 1.886026671708352,
|
| 284 |
+
"learning_rate": 4.4840920917726425e-06,
|
| 285 |
+
"logits/chosen": -1.2202318906784058,
|
| 286 |
+
"logits/rejected": -1.7096608877182007,
|
| 287 |
+
"logps/chosen": -48.562477111816406,
|
| 288 |
+
"logps/rejected": -456.2506408691406,
|
| 289 |
+
"loss": 0.1219,
|
| 290 |
+
"rewards/accuracies": 0.9375,
|
| 291 |
+
"rewards/chosen": 0.016236167401075363,
|
| 292 |
+
"rewards/margins": 3.572105884552002,
|
| 293 |
+
"rewards/rejected": -3.5558693408966064,
|
| 294 |
+
"step": 180
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 1.52,
|
| 298 |
+
"grad_norm": 1.208359086336539,
|
| 299 |
+
"learning_rate": 4.396014096912182e-06,
|
| 300 |
+
"logits/chosen": -1.1344788074493408,
|
| 301 |
+
"logits/rejected": -1.5908814668655396,
|
| 302 |
+
"logps/chosen": -47.373992919921875,
|
| 303 |
+
"logps/rejected": -488.7273864746094,
|
| 304 |
+
"loss": 0.0984,
|
| 305 |
+
"rewards/accuracies": 0.9375,
|
| 306 |
+
"rewards/chosen": 0.023608360439538956,
|
| 307 |
+
"rewards/margins": 3.904538631439209,
|
| 308 |
+
"rewards/rejected": -3.8809304237365723,
|
| 309 |
+
"step": 190
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 1.6,
|
| 313 |
+
"grad_norm": 3.9821859292862705,
|
| 314 |
+
"learning_rate": 4.302012913171584e-06,
|
| 315 |
+
"logits/chosen": -1.2608126401901245,
|
| 316 |
+
"logits/rejected": -1.7331119775772095,
|
| 317 |
+
"logps/chosen": -41.71361541748047,
|
| 318 |
+
"logps/rejected": -526.3060302734375,
|
| 319 |
+
"loss": 0.102,
|
| 320 |
+
"rewards/accuracies": 0.956250011920929,
|
| 321 |
+
"rewards/chosen": 0.08378119766712189,
|
| 322 |
+
"rewards/margins": 4.328982353210449,
|
| 323 |
+
"rewards/rejected": -4.245201110839844,
|
| 324 |
+
"step": 200
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.6800000000000002,
|
| 328 |
+
"grad_norm": 1.5269166268578003,
|
| 329 |
+
"learning_rate": 4.202382202273702e-06,
|
| 330 |
+
"logits/chosen": -1.114254117012024,
|
| 331 |
+
"logits/rejected": -1.6004146337509155,
|
| 332 |
+
"logps/chosen": -40.91397476196289,
|
| 333 |
+
"logps/rejected": -533.20751953125,
|
| 334 |
+
"loss": 0.0837,
|
| 335 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 336 |
+
"rewards/chosen": 0.08689381927251816,
|
| 337 |
+
"rewards/margins": 4.404597759246826,
|
| 338 |
+
"rewards/rejected": -4.317704200744629,
|
| 339 |
+
"step": 210
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 1.76,
|
| 343 |
+
"grad_norm": 3.6420048929792053,
|
| 344 |
+
"learning_rate": 4.097433212705492e-06,
|
| 345 |
+
"logits/chosen": -1.1888163089752197,
|
| 346 |
+
"logits/rejected": -1.7389936447143555,
|
| 347 |
+
"logps/chosen": -37.578697204589844,
|
| 348 |
+
"logps/rejected": -569.4190673828125,
|
| 349 |
+
"loss": 0.0527,
|
| 350 |
+
"rewards/accuracies": 0.987500011920929,
|
| 351 |
+
"rewards/chosen": 0.10641758143901825,
|
| 352 |
+
"rewards/margins": 4.77039909362793,
|
| 353 |
+
"rewards/rejected": -4.663981914520264,
|
| 354 |
+
"step": 220
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 1.8399999999999999,
|
| 358 |
+
"grad_norm": 7.795327748358564,
|
| 359 |
+
"learning_rate": 3.987493807371033e-06,
|
| 360 |
+
"logits/chosen": -1.1736271381378174,
|
| 361 |
+
"logits/rejected": -1.6653432846069336,
|
| 362 |
+
"logps/chosen": -39.23591995239258,
|
| 363 |
+
"logps/rejected": -596.703369140625,
|
| 364 |
+
"loss": 0.0708,
|
| 365 |
+
"rewards/accuracies": 0.981249988079071,
|
| 366 |
+
"rewards/chosen": 0.09144876152276993,
|
| 367 |
+
"rewards/margins": 5.02878999710083,
|
| 368 |
+
"rewards/rejected": -4.93734073638916,
|
| 369 |
+
"step": 230
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 1.92,
|
| 373 |
+
"grad_norm": 3.784757324910154,
|
| 374 |
+
"learning_rate": 3.872907439340758e-06,
|
| 375 |
+
"logits/chosen": -1.1998341083526611,
|
| 376 |
+
"logits/rejected": -1.5868158340454102,
|
| 377 |
+
"logps/chosen": -40.40230178833008,
|
| 378 |
+
"logps/rejected": -578.6609497070312,
|
| 379 |
+
"loss": 0.0589,
|
| 380 |
+
"rewards/accuracies": 0.9437500238418579,
|
| 381 |
+
"rewards/chosen": 0.10270919650793076,
|
| 382 |
+
"rewards/margins": 4.877388954162598,
|
| 383 |
+
"rewards/rejected": -4.774679660797119,
|
| 384 |
+
"step": 240
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 2.0,
|
| 388 |
+
"grad_norm": 1.1274212490773023,
|
| 389 |
+
"learning_rate": 3.75403207889666e-06,
|
| 390 |
+
"logits/chosen": -1.2294795513153076,
|
| 391 |
+
"logits/rejected": -1.621289849281311,
|
| 392 |
+
"logps/chosen": -40.97010040283203,
|
| 393 |
+
"logps/rejected": -588.9901733398438,
|
| 394 |
+
"loss": 0.0505,
|
| 395 |
+
"rewards/accuracies": 0.9437500238418579,
|
| 396 |
+
"rewards/chosen": 0.09869703650474548,
|
| 397 |
+
"rewards/margins": 4.985424041748047,
|
| 398 |
+
"rewards/rejected": -4.886727333068848,
|
| 399 |
+
"step": 250
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 2.08,
|
| 403 |
+
"grad_norm": 0.1615894379623363,
|
| 404 |
+
"learning_rate": 3.631239095225417e-06,
|
| 405 |
+
"logits/chosen": -1.104773759841919,
|
| 406 |
+
"logits/rejected": -1.556726098060608,
|
| 407 |
+
"logps/chosen": -41.80849075317383,
|
| 408 |
+
"logps/rejected": -610.6696166992188,
|
| 409 |
+
"loss": 0.0473,
|
| 410 |
+
"rewards/accuracies": 0.96875,
|
| 411 |
+
"rewards/chosen": 0.08216419816017151,
|
| 412 |
+
"rewards/margins": 5.163956642150879,
|
| 413 |
+
"rewards/rejected": -5.081791877746582,
|
| 414 |
+
"step": 260
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 2.16,
|
| 418 |
+
"grad_norm": 0.20738276054029106,
|
| 419 |
+
"learning_rate": 3.5049120962530608e-06,
|
| 420 |
+
"logits/chosen": -1.144113302230835,
|
| 421 |
+
"logits/rejected": -1.526831865310669,
|
| 422 |
+
"logps/chosen": -40.4541015625,
|
| 423 |
+
"logps/rejected": -581.6177978515625,
|
| 424 |
+
"loss": 0.0562,
|
| 425 |
+
"rewards/accuracies": 0.9375,
|
| 426 |
+
"rewards/chosen": 0.10693266242742538,
|
| 427 |
+
"rewards/margins": 4.924569129943848,
|
| 428 |
+
"rewards/rejected": -4.817636013031006,
|
| 429 |
+
"step": 270
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 2.24,
|
| 433 |
+
"grad_norm": 0.08228415199969275,
|
| 434 |
+
"learning_rate": 3.3754457302455464e-06,
|
| 435 |
+
"logits/chosen": -1.1058952808380127,
|
| 436 |
+
"logits/rejected": -1.5095767974853516,
|
| 437 |
+
"logps/chosen": -40.29807662963867,
|
| 438 |
+
"logps/rejected": -614.50634765625,
|
| 439 |
+
"loss": 0.0422,
|
| 440 |
+
"rewards/accuracies": 0.96875,
|
| 441 |
+
"rewards/chosen": 0.1043148785829544,
|
| 442 |
+
"rewards/margins": 5.244091033935547,
|
| 443 |
+
"rewards/rejected": -5.139776229858398,
|
| 444 |
+
"step": 280
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 2.32,
|
| 448 |
+
"grad_norm": 0.5196221393419624,
|
| 449 |
+
"learning_rate": 3.2432444529190714e-06,
|
| 450 |
+
"logits/chosen": -1.113113284111023,
|
| 451 |
+
"logits/rejected": -1.501694917678833,
|
| 452 |
+
"logps/chosen": -39.059715270996094,
|
| 453 |
+
"logps/rejected": -635.2387084960938,
|
| 454 |
+
"loss": 0.0458,
|
| 455 |
+
"rewards/accuracies": 0.981249988079071,
|
| 456 |
+
"rewards/chosen": 0.10602055490016937,
|
| 457 |
+
"rewards/margins": 5.451790809631348,
|
| 458 |
+
"rewards/rejected": -5.345770359039307,
|
| 459 |
+
"step": 290
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 2.4,
|
| 463 |
+
"grad_norm": 0.08825766353470356,
|
| 464 |
+
"learning_rate": 3.1087212639117057e-06,
|
| 465 |
+
"logits/chosen": -1.1622686386108398,
|
| 466 |
+
"logits/rejected": -1.5392283201217651,
|
| 467 |
+
"logps/chosen": -39.0388298034668,
|
| 468 |
+
"logps/rejected": -639.8673095703125,
|
| 469 |
+
"loss": 0.0312,
|
| 470 |
+
"rewards/accuracies": 1.0,
|
| 471 |
+
"rewards/chosen": 0.11618944257497787,
|
| 472 |
+
"rewards/margins": 5.501553535461426,
|
| 473 |
+
"rewards/rejected": -5.385363578796387,
|
| 474 |
+
"step": 300
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 2.48,
|
| 478 |
+
"grad_norm": 0.10694009885473299,
|
| 479 |
+
"learning_rate": 2.9722964165636263e-06,
|
| 480 |
+
"logits/chosen": -1.1022515296936035,
|
| 481 |
+
"logits/rejected": -1.4340097904205322,
|
| 482 |
+
"logps/chosen": -39.469932556152344,
|
| 483 |
+
"logps/rejected": -641.5587768554688,
|
| 484 |
+
"loss": 0.0458,
|
| 485 |
+
"rewards/accuracies": 1.0,
|
| 486 |
+
"rewards/chosen": 0.11364182084798813,
|
| 487 |
+
"rewards/margins": 5.520757675170898,
|
| 488 |
+
"rewards/rejected": -5.407115459442139,
|
| 489 |
+
"step": 310
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 2.56,
|
| 493 |
+
"grad_norm": 0.4507687634087117,
|
| 494 |
+
"learning_rate": 2.8343961050366275e-06,
|
| 495 |
+
"logits/chosen": -1.0810810327529907,
|
| 496 |
+
"logits/rejected": -1.4569213390350342,
|
| 497 |
+
"logps/chosen": -41.17557144165039,
|
| 498 |
+
"logps/rejected": -649.1512451171875,
|
| 499 |
+
"loss": 0.0228,
|
| 500 |
+
"rewards/accuracies": 0.9937499761581421,
|
| 501 |
+
"rewards/chosen": 0.07850812375545502,
|
| 502 |
+
"rewards/margins": 5.554858207702637,
|
| 503 |
+
"rewards/rejected": -5.476349830627441,
|
| 504 |
+
"step": 320
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 2.64,
|
| 508 |
+
"grad_norm": 0.03148227835487055,
|
| 509 |
+
"learning_rate": 2.695451132874385e-06,
|
| 510 |
+
"logits/chosen": -1.0946557521820068,
|
| 511 |
+
"logits/rejected": -1.494209885597229,
|
| 512 |
+
"logps/chosen": -44.248779296875,
|
| 513 |
+
"logps/rejected": -679.5821533203125,
|
| 514 |
+
"loss": 0.0242,
|
| 515 |
+
"rewards/accuracies": 1.0,
|
| 516 |
+
"rewards/chosen": 0.04512793570756912,
|
| 517 |
+
"rewards/margins": 5.8134870529174805,
|
| 518 |
+
"rewards/rejected": -5.768359184265137,
|
| 519 |
+
"step": 330
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 2.7199999999999998,
|
| 523 |
+
"grad_norm": 0.07742711911111466,
|
| 524 |
+
"learning_rate": 2.5558955671628964e-06,
|
| 525 |
+
"logits/chosen": -1.1254152059555054,
|
| 526 |
+
"logits/rejected": -1.4622461795806885,
|
| 527 |
+
"logps/chosen": -45.729515075683594,
|
| 528 |
+
"logps/rejected": -666.0767211914062,
|
| 529 |
+
"loss": 0.051,
|
| 530 |
+
"rewards/accuracies": 1.0,
|
| 531 |
+
"rewards/chosen": 0.046021558344364166,
|
| 532 |
+
"rewards/margins": 5.69630241394043,
|
| 533 |
+
"rewards/rejected": -5.650280952453613,
|
| 534 |
+
"step": 340
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 2.8,
|
| 538 |
+
"grad_norm": 0.0712177913874092,
|
| 539 |
+
"learning_rate": 2.4161653824955654e-06,
|
| 540 |
+
"logits/chosen": -1.2363636493682861,
|
| 541 |
+
"logits/rejected": -1.6161434650421143,
|
| 542 |
+
"logps/chosen": -47.28390884399414,
|
| 543 |
+
"logps/rejected": -664.1326904296875,
|
| 544 |
+
"loss": 0.0351,
|
| 545 |
+
"rewards/accuracies": 1.0,
|
| 546 |
+
"rewards/chosen": 0.038107506930828094,
|
| 547 |
+
"rewards/margins": 5.666635990142822,
|
| 548 |
+
"rewards/rejected": -5.628528118133545,
|
| 549 |
+
"step": 350
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 2.88,
|
| 553 |
+
"grad_norm": 0.11629194717439449,
|
| 554 |
+
"learning_rate": 2.2766970989791697e-06,
|
| 555 |
+
"logits/chosen": -1.2213943004608154,
|
| 556 |
+
"logits/rejected": -1.5998932123184204,
|
| 557 |
+
"logps/chosen": -48.74114990234375,
|
| 558 |
+
"logps/rejected": -667.0240478515625,
|
| 559 |
+
"loss": 0.0453,
|
| 560 |
+
"rewards/accuracies": 1.0,
|
| 561 |
+
"rewards/chosen": 0.0151976328343153,
|
| 562 |
+
"rewards/margins": 5.664753437042236,
|
| 563 |
+
"rewards/rejected": -5.649555683135986,
|
| 564 |
+
"step": 360
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 2.96,
|
| 568 |
+
"grad_norm": 0.0614653072363607,
|
| 569 |
+
"learning_rate": 2.1379264185356545e-06,
|
| 570 |
+
"logits/chosen": -1.205859899520874,
|
| 571 |
+
"logits/rejected": -1.5529258251190186,
|
| 572 |
+
"logps/chosen": -48.79401397705078,
|
| 573 |
+
"logps/rejected": -669.0440063476562,
|
| 574 |
+
"loss": 0.0517,
|
| 575 |
+
"rewards/accuracies": 1.0,
|
| 576 |
+
"rewards/chosen": 0.007499977946281433,
|
| 577 |
+
"rewards/margins": 5.6917924880981445,
|
| 578 |
+
"rewards/rejected": -5.684292316436768,
|
| 579 |
+
"step": 370
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 3.04,
|
| 583 |
+
"grad_norm": 0.10523859679268263,
|
| 584 |
+
"learning_rate": 2.000286863759934e-06,
|
| 585 |
+
"logits/chosen": -1.255184531211853,
|
| 586 |
+
"logits/rejected": -1.6094391345977783,
|
| 587 |
+
"logps/chosen": -48.94397735595703,
|
| 588 |
+
"logps/rejected": -655.6014404296875,
|
| 589 |
+
"loss": 0.0253,
|
| 590 |
+
"rewards/accuracies": 1.0,
|
| 591 |
+
"rewards/chosen": 0.006729847751557827,
|
| 592 |
+
"rewards/margins": 5.557109832763672,
|
| 593 |
+
"rewards/rejected": -5.550379753112793,
|
| 594 |
+
"step": 380
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 3.12,
|
| 598 |
+
"grad_norm": 0.2735817193264242,
|
| 599 |
+
"learning_rate": 1.8642084235859764e-06,
|
| 600 |
+
"logits/chosen": -1.2955033779144287,
|
| 601 |
+
"logits/rejected": -1.6524699926376343,
|
| 602 |
+
"logps/chosen": -50.138023376464844,
|
| 603 |
+
"logps/rejected": -673.2332763671875,
|
| 604 |
+
"loss": 0.0373,
|
| 605 |
+
"rewards/accuracies": 1.0,
|
| 606 |
+
"rewards/chosen": -0.0004118751676287502,
|
| 607 |
+
"rewards/margins": 5.711531639099121,
|
| 608 |
+
"rewards/rejected": -5.711943626403809,
|
| 609 |
+
"step": 390
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 3.2,
|
| 613 |
+
"grad_norm": 0.13877956212009288,
|
| 614 |
+
"learning_rate": 1.7301162099921013e-06,
|
| 615 |
+
"logits/chosen": -1.3686187267303467,
|
| 616 |
+
"logits/rejected": -1.6736618280410767,
|
| 617 |
+
"logps/chosen": -49.265708923339844,
|
| 618 |
+
"logps/rejected": -656.8291625976562,
|
| 619 |
+
"loss": 0.035,
|
| 620 |
+
"rewards/accuracies": 1.0,
|
| 621 |
+
"rewards/chosen": 0.025733808055520058,
|
| 622 |
+
"rewards/margins": 5.595955848693848,
|
| 623 |
+
"rewards/rejected": -5.5702223777771,
|
| 624 |
+
"step": 400
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 3.2800000000000002,
|
| 628 |
+
"grad_norm": 0.06971630538785722,
|
| 629 |
+
"learning_rate": 1.5984291299420117e-06,
|
| 630 |
+
"logits/chosen": -1.3197619915008545,
|
| 631 |
+
"logits/rejected": -1.6505024433135986,
|
| 632 |
+
"logps/chosen": -53.06304168701172,
|
| 633 |
+
"logps/rejected": -688.3905639648438,
|
| 634 |
+
"loss": 0.0277,
|
| 635 |
+
"rewards/accuracies": 0.9937499761581421,
|
| 636 |
+
"rewards/chosen": -0.038484591990709305,
|
| 637 |
+
"rewards/margins": 5.8287577629089355,
|
| 638 |
+
"rewards/rejected": -5.867241859436035,
|
| 639 |
+
"step": 410
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 3.36,
|
| 643 |
+
"grad_norm": 0.16813880626958627,
|
| 644 |
+
"learning_rate": 1.4695585767104092e-06,
|
| 645 |
+
"logits/chosen": -1.35333251953125,
|
| 646 |
+
"logits/rejected": -1.6515657901763916,
|
| 647 |
+
"logps/chosen": -52.462486267089844,
|
| 648 |
+
"logps/rejected": -676.153076171875,
|
| 649 |
+
"loss": 0.0198,
|
| 650 |
+
"rewards/accuracies": 1.0,
|
| 651 |
+
"rewards/chosen": -0.030403846874833107,
|
| 652 |
+
"rewards/margins": 5.729649543762207,
|
| 653 |
+
"rewards/rejected": -5.760053634643555,
|
| 654 |
+
"step": 420
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 3.44,
|
| 658 |
+
"grad_norm": 6.506151264887685,
|
| 659 |
+
"learning_rate": 1.3439071446815452e-06,
|
| 660 |
+
"logits/chosen": -1.4076862335205078,
|
| 661 |
+
"logits/rejected": -1.6836458444595337,
|
| 662 |
+
"logps/chosen": -54.67851638793945,
|
| 663 |
+
"logps/rejected": -673.2918701171875,
|
| 664 |
+
"loss": 0.0345,
|
| 665 |
+
"rewards/accuracies": 0.9937499761581421,
|
| 666 |
+
"rewards/chosen": -0.03381548449397087,
|
| 667 |
+
"rewards/margins": 5.694668769836426,
|
| 668 |
+
"rewards/rejected": -5.728484153747559,
|
| 669 |
+
"step": 430
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 3.52,
|
| 673 |
+
"grad_norm": 0.10218706215191123,
|
| 674 |
+
"learning_rate": 1.2218673716356919e-06,
|
| 675 |
+
"logits/chosen": -1.4333115816116333,
|
| 676 |
+
"logits/rejected": -1.6782299280166626,
|
| 677 |
+
"logps/chosen": -52.76792526245117,
|
| 678 |
+
"logps/rejected": -660.284423828125,
|
| 679 |
+
"loss": 0.0298,
|
| 680 |
+
"rewards/accuracies": 1.0,
|
| 681 |
+
"rewards/chosen": 0.004182952456176281,
|
| 682 |
+
"rewards/margins": 5.621959686279297,
|
| 683 |
+
"rewards/rejected": -5.617777347564697,
|
| 684 |
+
"step": 440
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 3.6,
|
| 688 |
+
"grad_norm": 0.08325288613545898,
|
| 689 |
+
"learning_rate": 1.103820512452661e-06,
|
| 690 |
+
"logits/chosen": -1.4696094989776611,
|
| 691 |
+
"logits/rejected": -1.7935911417007446,
|
| 692 |
+
"logps/chosen": -51.96623611450195,
|
| 693 |
+
"logps/rejected": -688.1324462890625,
|
| 694 |
+
"loss": 0.042,
|
| 695 |
+
"rewards/accuracies": 1.0,
|
| 696 |
+
"rewards/chosen": -0.026469239965081215,
|
| 697 |
+
"rewards/margins": 5.847690105438232,
|
| 698 |
+
"rewards/rejected": -5.874159812927246,
|
| 699 |
+
"step": 450
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 3.68,
|
| 703 |
+
"grad_norm": 0.07973945849285877,
|
| 704 |
+
"learning_rate": 9.901353480633468e-07,
|
| 705 |
+
"logits/chosen": -1.4555690288543701,
|
| 706 |
+
"logits/rejected": -1.784189224243164,
|
| 707 |
+
"logps/chosen": -52.75232696533203,
|
| 708 |
+
"logps/rejected": -679.7716064453125,
|
| 709 |
+
"loss": 0.0362,
|
| 710 |
+
"rewards/accuracies": 1.0,
|
| 711 |
+
"rewards/chosen": -0.03310062736272812,
|
| 712 |
+
"rewards/margins": 5.753149509429932,
|
| 713 |
+
"rewards/rejected": -5.786250114440918,
|
| 714 |
+
"step": 460
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 3.76,
|
| 718 |
+
"grad_norm": 0.05215554947056918,
|
| 719 |
+
"learning_rate": 8.811670333701544e-07,
|
| 720 |
+
"logits/chosen": -1.4402834177017212,
|
| 721 |
+
"logits/rejected": -1.730446457862854,
|
| 722 |
+
"logps/chosen": -52.232666015625,
|
| 723 |
+
"logps/rejected": -685.2855224609375,
|
| 724 |
+
"loss": 0.0213,
|
| 725 |
+
"rewards/accuracies": 1.0,
|
| 726 |
+
"rewards/chosen": -0.015675092115998268,
|
| 727 |
+
"rewards/margins": 5.823694705963135,
|
| 728 |
+
"rewards/rejected": -5.839369773864746,
|
| 729 |
+
"step": 470
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 3.84,
|
| 733 |
+
"grad_norm": 0.8907616320159397,
|
| 734 |
+
"learning_rate": 7.772559877354341e-07,
|
| 735 |
+
"logits/chosen": -1.5031287670135498,
|
| 736 |
+
"logits/rejected": -1.7836637496948242,
|
| 737 |
+
"logps/chosen": -54.20964431762695,
|
| 738 |
+
"logps/rejected": -671.3470458984375,
|
| 739 |
+
"loss": 0.0383,
|
| 740 |
+
"rewards/accuracies": 1.0,
|
| 741 |
+
"rewards/chosen": -0.03484940528869629,
|
| 742 |
+
"rewards/margins": 5.683054447174072,
|
| 743 |
+
"rewards/rejected": -5.717903137207031,
|
| 744 |
+
"step": 480
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 3.92,
|
| 748 |
+
"grad_norm": 0.08795166948742635,
|
| 749 |
+
"learning_rate": 6.787268315040604e-07,
|
| 750 |
+
"logits/chosen": -1.4750686883926392,
|
| 751 |
+
"logits/rejected": -1.7797225713729858,
|
| 752 |
+
"logps/chosen": -53.79764938354492,
|
| 753 |
+
"logps/rejected": -694.2701416015625,
|
| 754 |
+
"loss": 0.0221,
|
| 755 |
+
"rewards/accuracies": 1.0,
|
| 756 |
+
"rewards/chosen": -0.03998209163546562,
|
| 757 |
+
"rewards/margins": 5.879731178283691,
|
| 758 |
+
"rewards/rejected": -5.919713497161865,
|
| 759 |
+
"step": 490
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 4.0,
|
| 763 |
+
"grad_norm": 3.7786029105030954,
|
| 764 |
+
"learning_rate": 5.858873718824829e-07,
|
| 765 |
+
"logits/chosen": -1.55056893825531,
|
| 766 |
+
"logits/rejected": -1.832313895225525,
|
| 767 |
+
"logps/chosen": -56.137054443359375,
|
| 768 |
+
"logps/rejected": -684.9733276367188,
|
| 769 |
+
"loss": 0.0293,
|
| 770 |
+
"rewards/accuracies": 0.9937499761581421,
|
| 771 |
+
"rewards/chosen": -0.05843985825777054,
|
| 772 |
+
"rewards/margins": 5.785708427429199,
|
| 773 |
+
"rewards/rejected": -5.844147682189941,
|
| 774 |
+
"step": 500
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 4.08,
|
| 778 |
+
"grad_norm": 0.05189315042878034,
|
| 779 |
+
"learning_rate": 4.990276413423817e-07,
|
| 780 |
+
"logits/chosen": -1.59425687789917,
|
| 781 |
+
"logits/rejected": -1.8570120334625244,
|
| 782 |
+
"logps/chosen": -54.740135192871094,
|
| 783 |
+
"logps/rejected": -661.8132934570312,
|
| 784 |
+
"loss": 0.0275,
|
| 785 |
+
"rewards/accuracies": 1.0,
|
| 786 |
+
"rewards/chosen": -0.029537344351410866,
|
| 787 |
+
"rewards/margins": 5.607969284057617,
|
| 788 |
+
"rewards/rejected": -5.63750696182251,
|
| 789 |
+
"step": 510
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 4.16,
|
| 793 |
+
"grad_norm": 2.9723273056492787,
|
| 794 |
+
"learning_rate": 4.184189915529796e-07,
|
| 795 |
+
"logits/chosen": -1.5383813381195068,
|
| 796 |
+
"logits/rejected": -1.7925872802734375,
|
| 797 |
+
"logps/chosen": -55.53809356689453,
|
| 798 |
+
"logps/rejected": -685.3880004882812,
|
| 799 |
+
"loss": 0.0312,
|
| 800 |
+
"rewards/accuracies": 1.0,
|
| 801 |
+
"rewards/chosen": -0.054846394807100296,
|
| 802 |
+
"rewards/margins": 5.796133995056152,
|
| 803 |
+
"rewards/rejected": -5.850980281829834,
|
| 804 |
+
"step": 520
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 4.24,
|
| 808 |
+
"grad_norm": 0.061241630442406096,
|
| 809 |
+
"learning_rate": 3.4431324567258176e-07,
|
| 810 |
+
"logits/chosen": -1.5691543817520142,
|
| 811 |
+
"logits/rejected": -1.838580846786499,
|
| 812 |
+
"logps/chosen": -54.7171630859375,
|
| 813 |
+
"logps/rejected": -691.39990234375,
|
| 814 |
+
"loss": 0.0183,
|
| 815 |
+
"rewards/accuracies": 1.0,
|
| 816 |
+
"rewards/chosen": -0.05066860839724541,
|
| 817 |
+
"rewards/margins": 5.8546953201293945,
|
| 818 |
+
"rewards/rejected": -5.905364990234375,
|
| 819 |
+
"step": 530
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 4.32,
|
| 823 |
+
"grad_norm": 1.4547140093392317,
|
| 824 |
+
"learning_rate": 2.769419116476052e-07,
|
| 825 |
+
"logits/chosen": -1.563213586807251,
|
| 826 |
+
"logits/rejected": -1.8497707843780518,
|
| 827 |
+
"logps/chosen": -58.22840118408203,
|
| 828 |
+
"logps/rejected": -689.0776977539062,
|
| 829 |
+
"loss": 0.0316,
|
| 830 |
+
"rewards/accuracies": 1.0,
|
| 831 |
+
"rewards/chosen": -0.0835486501455307,
|
| 832 |
+
"rewards/margins": 5.783991813659668,
|
| 833 |
+
"rewards/rejected": -5.867539882659912,
|
| 834 |
+
"step": 540
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 4.4,
|
| 838 |
+
"grad_norm": 0.09632694899585373,
|
| 839 |
+
"learning_rate": 2.1651545897676512e-07,
|
| 840 |
+
"logits/chosen": -1.5931028127670288,
|
| 841 |
+
"logits/rejected": -1.7936248779296875,
|
| 842 |
+
"logps/chosen": -56.02924728393555,
|
| 843 |
+
"logps/rejected": -645.4485473632812,
|
| 844 |
+
"loss": 0.0301,
|
| 845 |
+
"rewards/accuracies": 1.0,
|
| 846 |
+
"rewards/chosen": -0.021922914311289787,
|
| 847 |
+
"rewards/margins": 5.465601921081543,
|
| 848 |
+
"rewards/rejected": -5.48752498626709,
|
| 849 |
+
"step": 550
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 4.48,
|
| 853 |
+
"grad_norm": 0.04943636920795491,
|
| 854 |
+
"learning_rate": 1.6322266119983222e-07,
|
| 855 |
+
"logits/chosen": -1.5493812561035156,
|
| 856 |
+
"logits/rejected": -1.7953119277954102,
|
| 857 |
+
"logps/chosen": -56.0436897277832,
|
| 858 |
+
"logps/rejected": -691.9136962890625,
|
| 859 |
+
"loss": 0.0309,
|
| 860 |
+
"rewards/accuracies": 1.0,
|
| 861 |
+
"rewards/chosen": -0.048449721187353134,
|
| 862 |
+
"rewards/margins": 5.857935905456543,
|
| 863 |
+
"rewards/rejected": -5.906384468078613,
|
| 864 |
+
"step": 560
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 4.5600000000000005,
|
| 868 |
+
"grad_norm": 0.03808578514394721,
|
| 869 |
+
"learning_rate": 1.1723000616502167e-07,
|
| 870 |
+
"logits/chosen": -1.60978102684021,
|
| 871 |
+
"logits/rejected": -1.8922958374023438,
|
| 872 |
+
"logps/chosen": -54.7984733581543,
|
| 873 |
+
"logps/rejected": -690.01708984375,
|
| 874 |
+
"loss": 0.0255,
|
| 875 |
+
"rewards/accuracies": 1.0,
|
| 876 |
+
"rewards/chosen": -0.0441301092505455,
|
| 877 |
+
"rewards/margins": 5.842381954193115,
|
| 878 |
+
"rewards/rejected": -5.88651180267334,
|
| 879 |
+
"step": 570
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 4.64,
|
| 883 |
+
"grad_norm": 2.3694244096968626,
|
| 884 |
+
"learning_rate": 7.868117591737585e-08,
|
| 885 |
+
"logits/chosen": -1.5402578115463257,
|
| 886 |
+
"logits/rejected": -1.8241513967514038,
|
| 887 |
+
"logps/chosen": -54.9166145324707,
|
| 888 |
+
"logps/rejected": -706.30517578125,
|
| 889 |
+
"loss": 0.0275,
|
| 890 |
+
"rewards/accuracies": 1.0,
|
| 891 |
+
"rewards/chosen": -0.06171821430325508,
|
| 892 |
+
"rewards/margins": 5.9701948165893555,
|
| 893 |
+
"rewards/rejected": -6.0319132804870605,
|
| 894 |
+
"step": 580
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 4.72,
|
| 898 |
+
"grad_norm": 0.10922169576399046,
|
| 899 |
+
"learning_rate": 4.769659783295383e-08,
|
| 900 |
+
"logits/chosen": -1.5660226345062256,
|
| 901 |
+
"logits/rejected": -1.8422683477401733,
|
| 902 |
+
"logps/chosen": -55.509124755859375,
|
| 903 |
+
"logps/rejected": -702.1007690429688,
|
| 904 |
+
"loss": 0.0242,
|
| 905 |
+
"rewards/accuracies": 1.0,
|
| 906 |
+
"rewards/chosen": -0.06676045805215836,
|
| 907 |
+
"rewards/margins": 5.940011978149414,
|
| 908 |
+
"rewards/rejected": -6.006772041320801,
|
| 909 |
+
"step": 590
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 4.8,
|
| 913 |
+
"grad_norm": 2.8341685934157685,
|
| 914 |
+
"learning_rate": 2.4373068401120358e-08,
|
| 915 |
+
"logits/chosen": -1.5939277410507202,
|
| 916 |
+
"logits/rejected": -1.8403106927871704,
|
| 917 |
+
"logps/chosen": -60.82611083984375,
|
| 918 |
+
"logps/rejected": -691.7379760742188,
|
| 919 |
+
"loss": 0.0428,
|
| 920 |
+
"rewards/accuracies": 0.9937499761581421,
|
| 921 |
+
"rewards/chosen": -0.11341984570026398,
|
| 922 |
+
"rewards/margins": 5.790899753570557,
|
| 923 |
+
"rewards/rejected": -5.904318809509277,
|
| 924 |
+
"step": 600
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"epoch": 4.88,
|
| 928 |
+
"grad_norm": 0.07385822938500211,
|
| 929 |
+
"learning_rate": 8.78345083022425e-09,
|
| 930 |
+
"logits/chosen": -1.59212327003479,
|
| 931 |
+
"logits/rejected": -1.8330879211425781,
|
| 932 |
+
"logps/chosen": -56.40080642700195,
|
| 933 |
+
"logps/rejected": -696.3282470703125,
|
| 934 |
+
"loss": 0.0212,
|
| 935 |
+
"rewards/accuracies": 1.0,
|
| 936 |
+
"rewards/chosen": -0.0736762136220932,
|
| 937 |
+
"rewards/margins": 5.873372554779053,
|
| 938 |
+
"rewards/rejected": -5.947049140930176,
|
| 939 |
+
"step": 610
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 4.96,
|
| 943 |
+
"grad_norm": 0.04029123800852909,
|
| 944 |
+
"learning_rate": 9.764474213677654e-10,
|
| 945 |
+
"logits/chosen": -1.5612826347351074,
|
| 946 |
+
"logits/rejected": -1.7991596460342407,
|
| 947 |
+
"logps/chosen": -56.65840530395508,
|
| 948 |
+
"logps/rejected": -704.2085571289062,
|
| 949 |
+
"loss": 0.019,
|
| 950 |
+
"rewards/accuracies": 1.0,
|
| 951 |
+
"rewards/chosen": -0.0770476832985878,
|
| 952 |
+
"rewards/margins": 5.943432807922363,
|
| 953 |
+
"rewards/rejected": -6.020480155944824,
|
| 954 |
+
"step": 620
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 5.0,
|
| 958 |
+
"step": 625,
|
| 959 |
+
"total_flos": 0.0,
|
| 960 |
+
"train_loss": 0.11420744479894639,
|
| 961 |
+
"train_runtime": 7032.7914,
|
| 962 |
+
"train_samples_per_second": 5.688,
|
| 963 |
+
"train_steps_per_second": 0.089
|
| 964 |
+
}
|
| 965 |
+
],
|
| 966 |
+
"logging_steps": 10,
|
| 967 |
+
"max_steps": 625,
|
| 968 |
+
"num_input_tokens_seen": 0,
|
| 969 |
+
"num_train_epochs": 5,
|
| 970 |
+
"save_steps": 1000,
|
| 971 |
+
"stateful_callbacks": {
|
| 972 |
+
"TrainerControl": {
|
| 973 |
+
"args": {
|
| 974 |
+
"should_epoch_stop": false,
|
| 975 |
+
"should_evaluate": false,
|
| 976 |
+
"should_log": false,
|
| 977 |
+
"should_save": true,
|
| 978 |
+
"should_training_stop": true
|
| 979 |
+
},
|
| 980 |
+
"attributes": {}
|
| 981 |
+
}
|
| 982 |
+
},
|
| 983 |
+
"total_flos": 0.0,
|
| 984 |
+
"train_batch_size": 4,
|
| 985 |
+
"trial_name": null,
|
| 986 |
+
"trial_params": null
|
| 987 |
+
}
|