Model save
Browse files- README.md +59 -0
- adapter_model.safetensors +1 -1
- all_results.json +8 -0
- emissions.csv +2 -0
- runs/Sep14_16-05-04_gpu4-119-5/events.out.tfevents.1726293949.gpu4-119-5.65997.0 +2 -2
- train_results.json +8 -0
- trainer_state.json +1822 -0
README.md
ADDED
|
@@ -0,0 +1,59 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: peft
|
| 3 |
+
tags:
|
| 4 |
+
- trl
|
| 5 |
+
- dpo
|
| 6 |
+
- generated_from_trainer
|
| 7 |
+
base_model: DUAL-GPO/phi-2-ipo-chatml-merged
|
| 8 |
+
model-index:
|
| 9 |
+
- name: phi-2-ipo-chatml-lora-i1
|
| 10 |
+
results: []
|
| 11 |
+
---
|
| 12 |
+
|
| 13 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 14 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 15 |
+
|
| 16 |
+
# phi-2-ipo-chatml-lora-i1
|
| 17 |
+
|
| 18 |
+
This model was trained from scratch on the None dataset.
|
| 19 |
+
|
| 20 |
+
## Model description
|
| 21 |
+
|
| 22 |
+
More information needed
|
| 23 |
+
|
| 24 |
+
## Intended uses & limitations
|
| 25 |
+
|
| 26 |
+
More information needed
|
| 27 |
+
|
| 28 |
+
## Training and evaluation data
|
| 29 |
+
|
| 30 |
+
More information needed
|
| 31 |
+
|
| 32 |
+
## Training procedure
|
| 33 |
+
|
| 34 |
+
### Training hyperparameters
|
| 35 |
+
|
| 36 |
+
The following hyperparameters were used during training:
|
| 37 |
+
- learning_rate: 5e-06
|
| 38 |
+
- train_batch_size: 4
|
| 39 |
+
- eval_batch_size: 4
|
| 40 |
+
- seed: 42
|
| 41 |
+
- distributed_type: multi-GPU
|
| 42 |
+
- gradient_accumulation_steps: 4
|
| 43 |
+
- total_train_batch_size: 16
|
| 44 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 45 |
+
- lr_scheduler_type: cosine
|
| 46 |
+
- lr_scheduler_warmup_ratio: 0.1
|
| 47 |
+
- num_epochs: 1
|
| 48 |
+
|
| 49 |
+
### Training results
|
| 50 |
+
|
| 51 |
+
|
| 52 |
+
|
| 53 |
+
### Framework versions
|
| 54 |
+
|
| 55 |
+
- PEFT 0.7.1
|
| 56 |
+
- Transformers 4.36.2
|
| 57 |
+
- Pytorch 2.1.2
|
| 58 |
+
- Datasets 2.14.6
|
| 59 |
+
- 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 335579632
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ae2cbf3ee4334b48dcebfb3066014a24431fe87bbabdcc2166b2320e9841b252
|
| 3 |
size 335579632
|
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"train_loss": 214508.74675962294,
|
| 4 |
+
"train_runtime": 12229.8763,
|
| 5 |
+
"train_samples": 20378,
|
| 6 |
+
"train_samples_per_second": 1.666,
|
| 7 |
+
"train_steps_per_second": 0.104
|
| 8 |
+
}
|
emissions.csv
ADDED
|
@@ -0,0 +1,2 @@
|
|
|
|
|
|
|
|
|
|
| 1 |
+
timestamp,project_name,run_id,duration,emissions,emissions_rate,cpu_power,gpu_power,ram_power,cpu_energy,gpu_energy,ram_energy,energy_consumed,country_name,country_iso_code,region,cloud_provider,cloud_region,os,python_version,codecarbon_version,cpu_count,cpu_model,gpu_count,gpu_model,longitude,latitude,ram_total_size,tracking_mode,on_cloud,pue
|
| 2 |
+
2024-09-14T19:29:39,codecarbon,239e4444-c328-4a04-be93-562ee2a2f242,12230.030321598053,0.008427352110232458,6.890704183578252e-07,42.5,371.10200000000003,188.74309015274048,0.14438064921779767,2.7612975451493393,0.6393689264745689,3.5450471208417156,Canada,CAN,quebec,,,Linux-5.15.0-84-generic-x86_64-with-glibc2.35,3.10.14,2.2.3,32,Intel(R) Xeon(R) W-3335 CPU @ 3.40GHz,4,4 x NVIDIA GeForce RTX 4090,-71.2,46.8,503.3149070739746,machine,N,1.0
|
runs/Sep14_16-05-04_gpu4-119-5/events.out.tfevents.1726293949.gpu4-119-5.65997.0
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 sha256:7a6f415b5635b7e256aa9d536286cebc9323c969bb0e59c565c024b26820d554
|
| 3 |
+
size 86277
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"train_loss": 214508.74675962294,
|
| 4 |
+
"train_runtime": 12229.8763,
|
| 5 |
+
"train_samples": 20378,
|
| 6 |
+
"train_samples_per_second": 1.666,
|
| 7 |
+
"train_steps_per_second": 0.104
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1822 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.9994111874386653,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1273,
|
| 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": 3.90625e-08,
|
| 14 |
+
"logits/chosen": 0.753882884979248,
|
| 15 |
+
"logits/rejected": 0.9021581411361694,
|
| 16 |
+
"logps/chosen": -402.58197021484375,
|
| 17 |
+
"logps/rejected": -272.54583740234375,
|
| 18 |
+
"loss": 250000.0,
|
| 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": 0.01,
|
| 27 |
+
"learning_rate": 3.90625e-07,
|
| 28 |
+
"logits/chosen": 0.7883161902427673,
|
| 29 |
+
"logits/rejected": 0.8325157761573792,
|
| 30 |
+
"logps/chosen": -300.5879211425781,
|
| 31 |
+
"logps/rejected": -293.91253662109375,
|
| 32 |
+
"loss": 249972.8333,
|
| 33 |
+
"rewards/accuracies": 0.3888888955116272,
|
| 34 |
+
"rewards/chosen": -2.7298270651954226e-05,
|
| 35 |
+
"rewards/margins": 2.835398663592059e-05,
|
| 36 |
+
"rewards/rejected": -5.5652246373938397e-05,
|
| 37 |
+
"step": 10
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.02,
|
| 41 |
+
"learning_rate": 7.8125e-07,
|
| 42 |
+
"logits/chosen": 0.7864906191825867,
|
| 43 |
+
"logits/rejected": 0.8796188235282898,
|
| 44 |
+
"logps/chosen": -296.65606689453125,
|
| 45 |
+
"logps/rejected": -256.8786926269531,
|
| 46 |
+
"loss": 250065.0,
|
| 47 |
+
"rewards/accuracies": 0.375,
|
| 48 |
+
"rewards/chosen": -7.715411629760638e-05,
|
| 49 |
+
"rewards/margins": -6.430874054785818e-05,
|
| 50 |
+
"rewards/rejected": -1.2845377568737604e-05,
|
| 51 |
+
"step": 20
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.02,
|
| 55 |
+
"learning_rate": 1.1718750000000001e-06,
|
| 56 |
+
"logits/chosen": 0.8224496841430664,
|
| 57 |
+
"logits/rejected": 0.8862675428390503,
|
| 58 |
+
"logps/chosen": -272.9891662597656,
|
| 59 |
+
"logps/rejected": -244.2042236328125,
|
| 60 |
+
"loss": 249965.3,
|
| 61 |
+
"rewards/accuracies": 0.4625000059604645,
|
| 62 |
+
"rewards/chosen": -0.00036038103280588984,
|
| 63 |
+
"rewards/margins": 3.5376859159441665e-05,
|
| 64 |
+
"rewards/rejected": -0.00039575790287926793,
|
| 65 |
+
"step": 30
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.03,
|
| 69 |
+
"learning_rate": 1.5625e-06,
|
| 70 |
+
"logits/chosen": 0.7388542890548706,
|
| 71 |
+
"logits/rejected": 0.8643872141838074,
|
| 72 |
+
"logps/chosen": -300.79156494140625,
|
| 73 |
+
"logps/rejected": -313.8456115722656,
|
| 74 |
+
"loss": 249955.5,
|
| 75 |
+
"rewards/accuracies": 0.4625000059604645,
|
| 76 |
+
"rewards/chosen": -0.0008908492745831609,
|
| 77 |
+
"rewards/margins": 4.541150701697916e-05,
|
| 78 |
+
"rewards/rejected": -0.0009362607379443944,
|
| 79 |
+
"step": 40
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.04,
|
| 83 |
+
"learning_rate": 1.953125e-06,
|
| 84 |
+
"logits/chosen": 0.7705200910568237,
|
| 85 |
+
"logits/rejected": 0.8214074969291687,
|
| 86 |
+
"logps/chosen": -313.44525146484375,
|
| 87 |
+
"logps/rejected": -286.5237121582031,
|
| 88 |
+
"loss": 249616.75,
|
| 89 |
+
"rewards/accuracies": 0.5562499761581421,
|
| 90 |
+
"rewards/chosen": -0.0017550220945850015,
|
| 91 |
+
"rewards/margins": 0.0003843012382276356,
|
| 92 |
+
"rewards/rejected": -0.002139323391020298,
|
| 93 |
+
"step": 50
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.05,
|
| 97 |
+
"learning_rate": 2.3437500000000002e-06,
|
| 98 |
+
"logits/chosen": 0.7623851895332336,
|
| 99 |
+
"logits/rejected": 0.8920168876647949,
|
| 100 |
+
"logps/chosen": -350.54656982421875,
|
| 101 |
+
"logps/rejected": -304.0585021972656,
|
| 102 |
+
"loss": 249209.575,
|
| 103 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 104 |
+
"rewards/chosen": -0.003797227516770363,
|
| 105 |
+
"rewards/margins": 0.0007935293251648545,
|
| 106 |
+
"rewards/rejected": -0.004590757191181183,
|
| 107 |
+
"step": 60
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.05,
|
| 111 |
+
"learning_rate": 2.7343750000000004e-06,
|
| 112 |
+
"logits/chosen": 0.7647961378097534,
|
| 113 |
+
"logits/rejected": 0.8629061579704285,
|
| 114 |
+
"logps/chosen": -286.3582458496094,
|
| 115 |
+
"logps/rejected": -248.80715942382812,
|
| 116 |
+
"loss": 248737.95,
|
| 117 |
+
"rewards/accuracies": 0.59375,
|
| 118 |
+
"rewards/chosen": -0.006279756315052509,
|
| 119 |
+
"rewards/margins": 0.0012701054802164435,
|
| 120 |
+
"rewards/rejected": -0.0075498609803617,
|
| 121 |
+
"step": 70
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.06,
|
| 125 |
+
"learning_rate": 3.125e-06,
|
| 126 |
+
"logits/chosen": 0.8171808123588562,
|
| 127 |
+
"logits/rejected": 0.9254066348075867,
|
| 128 |
+
"logps/chosen": -274.9587097167969,
|
| 129 |
+
"logps/rejected": -252.5512237548828,
|
| 130 |
+
"loss": 248480.9,
|
| 131 |
+
"rewards/accuracies": 0.5375000238418579,
|
| 132 |
+
"rewards/chosen": -0.010536030866205692,
|
| 133 |
+
"rewards/margins": 0.0015363285783678293,
|
| 134 |
+
"rewards/rejected": -0.012072358280420303,
|
| 135 |
+
"step": 80
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.07,
|
| 139 |
+
"learning_rate": 3.5156250000000003e-06,
|
| 140 |
+
"logits/chosen": 0.8039876818656921,
|
| 141 |
+
"logits/rejected": 0.8282687067985535,
|
| 142 |
+
"logps/chosen": -333.23455810546875,
|
| 143 |
+
"logps/rejected": -340.7774353027344,
|
| 144 |
+
"loss": 248449.15,
|
| 145 |
+
"rewards/accuracies": 0.5375000238418579,
|
| 146 |
+
"rewards/chosen": -0.01643429882824421,
|
| 147 |
+
"rewards/margins": 0.001596734276972711,
|
| 148 |
+
"rewards/rejected": -0.018031032755970955,
|
| 149 |
+
"step": 90
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.08,
|
| 153 |
+
"learning_rate": 3.90625e-06,
|
| 154 |
+
"logits/chosen": 0.7214804291725159,
|
| 155 |
+
"logits/rejected": 0.8068147897720337,
|
| 156 |
+
"logps/chosen": -340.88555908203125,
|
| 157 |
+
"logps/rejected": -315.34942626953125,
|
| 158 |
+
"loss": 246475.125,
|
| 159 |
+
"rewards/accuracies": 0.6000000238418579,
|
| 160 |
+
"rewards/chosen": -0.02047683112323284,
|
| 161 |
+
"rewards/margins": 0.0035947158467024565,
|
| 162 |
+
"rewards/rejected": -0.024071548134088516,
|
| 163 |
+
"step": 100
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.09,
|
| 167 |
+
"learning_rate": 4.296875e-06,
|
| 168 |
+
"logits/chosen": 0.7144443392753601,
|
| 169 |
+
"logits/rejected": 0.861451268196106,
|
| 170 |
+
"logps/chosen": -326.4942932128906,
|
| 171 |
+
"logps/rejected": -317.719482421875,
|
| 172 |
+
"loss": 244917.625,
|
| 173 |
+
"rewards/accuracies": 0.606249988079071,
|
| 174 |
+
"rewards/chosen": -0.026887858286499977,
|
| 175 |
+
"rewards/margins": 0.005240024067461491,
|
| 176 |
+
"rewards/rejected": -0.032127875834703445,
|
| 177 |
+
"step": 110
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.09,
|
| 181 |
+
"learning_rate": 4.6875000000000004e-06,
|
| 182 |
+
"logits/chosen": 0.7016499042510986,
|
| 183 |
+
"logits/rejected": 0.6916753053665161,
|
| 184 |
+
"logps/chosen": -354.7713928222656,
|
| 185 |
+
"logps/rejected": -349.99114990234375,
|
| 186 |
+
"loss": 243192.975,
|
| 187 |
+
"rewards/accuracies": 0.5687500238418579,
|
| 188 |
+
"rewards/chosen": -0.03751590475440025,
|
| 189 |
+
"rewards/margins": 0.007144705858081579,
|
| 190 |
+
"rewards/rejected": -0.04466061294078827,
|
| 191 |
+
"step": 120
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.1,
|
| 195 |
+
"learning_rate": 4.999962359300416e-06,
|
| 196 |
+
"logits/chosen": 0.6297165155410767,
|
| 197 |
+
"logits/rejected": 0.6793630123138428,
|
| 198 |
+
"logps/chosen": -338.17181396484375,
|
| 199 |
+
"logps/rejected": -367.1688537597656,
|
| 200 |
+
"loss": 241014.625,
|
| 201 |
+
"rewards/accuracies": 0.606249988079071,
|
| 202 |
+
"rewards/chosen": -0.05371738597750664,
|
| 203 |
+
"rewards/margins": 0.009516485966742039,
|
| 204 |
+
"rewards/rejected": -0.0632338672876358,
|
| 205 |
+
"step": 130
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.11,
|
| 209 |
+
"learning_rate": 4.998645053824218e-06,
|
| 210 |
+
"logits/chosen": 0.5639317631721497,
|
| 211 |
+
"logits/rejected": 0.610073447227478,
|
| 212 |
+
"logps/chosen": -337.5505065917969,
|
| 213 |
+
"logps/rejected": -357.8117370605469,
|
| 214 |
+
"loss": 239699.275,
|
| 215 |
+
"rewards/accuracies": 0.6000000238418579,
|
| 216 |
+
"rewards/chosen": -0.06457965821027756,
|
| 217 |
+
"rewards/margins": 0.011107666417956352,
|
| 218 |
+
"rewards/rejected": -0.07568733394145966,
|
| 219 |
+
"step": 140
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.12,
|
| 223 |
+
"learning_rate": 4.9954468466732145e-06,
|
| 224 |
+
"logits/chosen": 0.404470831155777,
|
| 225 |
+
"logits/rejected": 0.439323753118515,
|
| 226 |
+
"logps/chosen": -339.54901123046875,
|
| 227 |
+
"logps/rejected": -331.80499267578125,
|
| 228 |
+
"loss": 239823.925,
|
| 229 |
+
"rewards/accuracies": 0.543749988079071,
|
| 230 |
+
"rewards/chosen": -0.0667237862944603,
|
| 231 |
+
"rewards/margins": 0.011366022750735283,
|
| 232 |
+
"rewards/rejected": -0.07808981090784073,
|
| 233 |
+
"step": 150
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.13,
|
| 237 |
+
"learning_rate": 4.990370145357496e-06,
|
| 238 |
+
"logits/chosen": 0.431760311126709,
|
| 239 |
+
"logits/rejected": 0.49426621198654175,
|
| 240 |
+
"logps/chosen": -370.2152404785156,
|
| 241 |
+
"logps/rejected": -357.75103759765625,
|
| 242 |
+
"loss": 236582.7,
|
| 243 |
+
"rewards/accuracies": 0.5687500238418579,
|
| 244 |
+
"rewards/chosen": -0.0688774362206459,
|
| 245 |
+
"rewards/margins": 0.015688957646489143,
|
| 246 |
+
"rewards/rejected": -0.0845663920044899,
|
| 247 |
+
"step": 160
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.13,
|
| 251 |
+
"learning_rate": 4.983418771458684e-06,
|
| 252 |
+
"logits/chosen": 0.3582938313484192,
|
| 253 |
+
"logits/rejected": 0.4145487844944,
|
| 254 |
+
"logps/chosen": -393.54296875,
|
| 255 |
+
"logps/rejected": -395.73968505859375,
|
| 256 |
+
"loss": 233456.825,
|
| 257 |
+
"rewards/accuracies": 0.5375000238418579,
|
| 258 |
+
"rewards/chosen": -0.09368976205587387,
|
| 259 |
+
"rewards/margins": 0.019242364913225174,
|
| 260 |
+
"rewards/rejected": -0.11293212324380875,
|
| 261 |
+
"step": 170
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.14,
|
| 265 |
+
"learning_rate": 4.97459795775315e-06,
|
| 266 |
+
"logits/chosen": 0.24134807288646698,
|
| 267 |
+
"logits/rejected": 0.34352773427963257,
|
| 268 |
+
"logps/chosen": -479.29827880859375,
|
| 269 |
+
"logps/rejected": -480.0040588378906,
|
| 270 |
+
"loss": 226326.95,
|
| 271 |
+
"rewards/accuracies": 0.612500011920929,
|
| 272 |
+
"rewards/chosen": -0.1502467542886734,
|
| 273 |
+
"rewards/margins": 0.02949386276304722,
|
| 274 |
+
"rewards/rejected": -0.17974062263965607,
|
| 275 |
+
"step": 180
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.15,
|
| 279 |
+
"learning_rate": 4.963914344272961e-06,
|
| 280 |
+
"logits/chosen": 0.1726927012205124,
|
| 281 |
+
"logits/rejected": 0.17276537418365479,
|
| 282 |
+
"logps/chosen": -474.7080993652344,
|
| 283 |
+
"logps/rejected": -503.10791015625,
|
| 284 |
+
"loss": 226260.375,
|
| 285 |
+
"rewards/accuracies": 0.59375,
|
| 286 |
+
"rewards/chosen": -0.16851405799388885,
|
| 287 |
+
"rewards/margins": 0.033648595213890076,
|
| 288 |
+
"rewards/rejected": -0.20216265320777893,
|
| 289 |
+
"step": 190
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.16,
|
| 293 |
+
"learning_rate": 4.951375973307458e-06,
|
| 294 |
+
"logits/chosen": 0.08363378793001175,
|
| 295 |
+
"logits/rejected": 0.19105622172355652,
|
| 296 |
+
"logps/chosen": -507.63128662109375,
|
| 297 |
+
"logps/rejected": -521.0435791015625,
|
| 298 |
+
"loss": 229957.85,
|
| 299 |
+
"rewards/accuracies": 0.5874999761581421,
|
| 300 |
+
"rewards/chosen": -0.18896250426769257,
|
| 301 |
+
"rewards/margins": 0.03131803497672081,
|
| 302 |
+
"rewards/rejected": -0.22028055787086487,
|
| 303 |
+
"step": 200
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.16,
|
| 307 |
+
"learning_rate": 4.93699228334928e-06,
|
| 308 |
+
"logits/chosen": 0.027421921491622925,
|
| 309 |
+
"logits/rejected": 0.09061434864997864,
|
| 310 |
+
"logps/chosen": -499.6670837402344,
|
| 311 |
+
"logps/rejected": -513.214111328125,
|
| 312 |
+
"loss": 228005.45,
|
| 313 |
+
"rewards/accuracies": 0.6000000238418579,
|
| 314 |
+
"rewards/chosen": -0.19364123046398163,
|
| 315 |
+
"rewards/margins": 0.03244911879301071,
|
| 316 |
+
"rewards/rejected": -0.22609035670757294,
|
| 317 |
+
"step": 210
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.17,
|
| 321 |
+
"learning_rate": 4.920774101989362e-06,
|
| 322 |
+
"logits/chosen": 0.015898525714874268,
|
| 323 |
+
"logits/rejected": 0.06961273401975632,
|
| 324 |
+
"logps/chosen": -542.1640625,
|
| 325 |
+
"logps/rejected": -571.1614990234375,
|
| 326 |
+
"loss": 220063.4,
|
| 327 |
+
"rewards/accuracies": 0.5874999761581421,
|
| 328 |
+
"rewards/chosen": -0.22638651728630066,
|
| 329 |
+
"rewards/margins": 0.041207194328308105,
|
| 330 |
+
"rewards/rejected": -0.2675936818122864,
|
| 331 |
+
"step": 220
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.18,
|
| 335 |
+
"learning_rate": 4.902733637766261e-06,
|
| 336 |
+
"logits/chosen": -0.03187388926744461,
|
| 337 |
+
"logits/rejected": -0.05228201299905777,
|
| 338 |
+
"logps/chosen": -584.11083984375,
|
| 339 |
+
"logps/rejected": -596.0657348632812,
|
| 340 |
+
"loss": 223339.975,
|
| 341 |
+
"rewards/accuracies": 0.6000000238418579,
|
| 342 |
+
"rewards/chosen": -0.27142053842544556,
|
| 343 |
+
"rewards/margins": 0.04074941202998161,
|
| 344 |
+
"rewards/rejected": -0.3121699094772339,
|
| 345 |
+
"step": 230
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.19,
|
| 349 |
+
"learning_rate": 4.882884470975954e-06,
|
| 350 |
+
"logits/chosen": -0.07718712091445923,
|
| 351 |
+
"logits/rejected": -0.048422373831272125,
|
| 352 |
+
"logps/chosen": -590.1216430664062,
|
| 353 |
+
"logps/rejected": -614.7066650390625,
|
| 354 |
+
"loss": 235878.475,
|
| 355 |
+
"rewards/accuracies": 0.606249988079071,
|
| 356 |
+
"rewards/chosen": -0.2930641770362854,
|
| 357 |
+
"rewards/margins": 0.059319932013750076,
|
| 358 |
+
"rewards/rejected": -0.352384090423584,
|
| 359 |
+
"step": 240
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.2,
|
| 363 |
+
"learning_rate": 4.861241543449015e-06,
|
| 364 |
+
"logits/chosen": 0.0006945926579646766,
|
| 365 |
+
"logits/rejected": -0.017523247748613358,
|
| 366 |
+
"logps/chosen": -501.6593322753906,
|
| 367 |
+
"logps/rejected": -508.7500915527344,
|
| 368 |
+
"loss": 233176.425,
|
| 369 |
+
"rewards/accuracies": 0.5625,
|
| 370 |
+
"rewards/chosen": -0.22137589752674103,
|
| 371 |
+
"rewards/margins": 0.02421090006828308,
|
| 372 |
+
"rewards/rejected": -0.2455867975950241,
|
| 373 |
+
"step": 250
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.2,
|
| 377 |
+
"learning_rate": 4.8378211473028755e-06,
|
| 378 |
+
"logits/chosen": 0.013015779666602612,
|
| 379 |
+
"logits/rejected": 0.12976054847240448,
|
| 380 |
+
"logps/chosen": -529.851806640625,
|
| 381 |
+
"logps/rejected": -552.1076049804688,
|
| 382 |
+
"loss": 217051.375,
|
| 383 |
+
"rewards/accuracies": 0.643750011920929,
|
| 384 |
+
"rewards/chosen": -0.2184825837612152,
|
| 385 |
+
"rewards/margins": 0.04304370656609535,
|
| 386 |
+
"rewards/rejected": -0.26152628660202026,
|
| 387 |
+
"step": 260
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.21,
|
| 391 |
+
"learning_rate": 4.812640912677624e-06,
|
| 392 |
+
"logits/chosen": 0.10295257717370987,
|
| 393 |
+
"logits/rejected": 0.10734937340021133,
|
| 394 |
+
"logps/chosen": -514.9952392578125,
|
| 395 |
+
"logps/rejected": -532.2618408203125,
|
| 396 |
+
"loss": 228697.625,
|
| 397 |
+
"rewards/accuracies": 0.5874999761581421,
|
| 398 |
+
"rewards/chosen": -0.2471189945936203,
|
| 399 |
+
"rewards/margins": 0.03284931927919388,
|
| 400 |
+
"rewards/rejected": -0.2799683213233948,
|
| 401 |
+
"step": 270
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.22,
|
| 405 |
+
"learning_rate": 4.785719794464596e-06,
|
| 406 |
+
"logits/chosen": 0.08165857940912247,
|
| 407 |
+
"logits/rejected": 0.11624784767627716,
|
| 408 |
+
"logps/chosen": -614.1241455078125,
|
| 409 |
+
"logps/rejected": -608.5601806640625,
|
| 410 |
+
"loss": 219380.5,
|
| 411 |
+
"rewards/accuracies": 0.5687500238418579,
|
| 412 |
+
"rewards/chosen": -0.2938295304775238,
|
| 413 |
+
"rewards/margins": 0.04496774449944496,
|
| 414 |
+
"rewards/rejected": -0.33879727125167847,
|
| 415 |
+
"step": 280
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.23,
|
| 419 |
+
"learning_rate": 4.757078058037722e-06,
|
| 420 |
+
"logits/chosen": 0.049666643142700195,
|
| 421 |
+
"logits/rejected": 0.037592388689517975,
|
| 422 |
+
"logps/chosen": -709.3441162109375,
|
| 423 |
+
"logps/rejected": -723.0515747070312,
|
| 424 |
+
"loss": 222553.975,
|
| 425 |
+
"rewards/accuracies": 0.581250011920929,
|
| 426 |
+
"rewards/chosen": -0.3819270133972168,
|
| 427 |
+
"rewards/margins": 0.04088777303695679,
|
| 428 |
+
"rewards/rejected": -0.4228147864341736,
|
| 429 |
+
"step": 290
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.24,
|
| 433 |
+
"learning_rate": 4.72673726399839e-06,
|
| 434 |
+
"logits/chosen": -0.011786766350269318,
|
| 435 |
+
"logits/rejected": 0.012369150295853615,
|
| 436 |
+
"logps/chosen": -711.5050659179688,
|
| 437 |
+
"logps/rejected": -751.9043579101562,
|
| 438 |
+
"loss": 216777.1,
|
| 439 |
+
"rewards/accuracies": 0.6000000238418579,
|
| 440 |
+
"rewards/chosen": -0.3882337510585785,
|
| 441 |
+
"rewards/margins": 0.05219928175210953,
|
| 442 |
+
"rewards/rejected": -0.4404330849647522,
|
| 443 |
+
"step": 300
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.24,
|
| 447 |
+
"learning_rate": 4.694720251945298e-06,
|
| 448 |
+
"logits/chosen": -0.03419340401887894,
|
| 449 |
+
"logits/rejected": 0.06611661612987518,
|
| 450 |
+
"logps/chosen": -677.6072387695312,
|
| 451 |
+
"logps/rejected": -711.2977905273438,
|
| 452 |
+
"loss": 213271.225,
|
| 453 |
+
"rewards/accuracies": 0.6187499761581421,
|
| 454 |
+
"rewards/chosen": -0.36067238450050354,
|
| 455 |
+
"rewards/margins": 0.058848440647125244,
|
| 456 |
+
"rewards/rejected": -0.41952085494995117,
|
| 457 |
+
"step": 310
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.25,
|
| 461 |
+
"learning_rate": 4.661051123281528e-06,
|
| 462 |
+
"logits/chosen": -0.017030570656061172,
|
| 463 |
+
"logits/rejected": -0.015286751091480255,
|
| 464 |
+
"logps/chosen": -677.0394287109375,
|
| 465 |
+
"logps/rejected": -715.4779052734375,
|
| 466 |
+
"loss": 210663.625,
|
| 467 |
+
"rewards/accuracies": 0.574999988079071,
|
| 468 |
+
"rewards/chosen": -0.3839905560016632,
|
| 469 |
+
"rewards/margins": 0.06271493434906006,
|
| 470 |
+
"rewards/rejected": -0.44670549035072327,
|
| 471 |
+
"step": 320
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.26,
|
| 475 |
+
"learning_rate": 4.6257552230717536e-06,
|
| 476 |
+
"logits/chosen": -0.07125677913427353,
|
| 477 |
+
"logits/rejected": 0.0028305717278271914,
|
| 478 |
+
"logps/chosen": -692.05517578125,
|
| 479 |
+
"logps/rejected": -677.0928955078125,
|
| 480 |
+
"loss": 225718.325,
|
| 481 |
+
"rewards/accuracies": 0.543749988079071,
|
| 482 |
+
"rewards/chosen": -0.3579491972923279,
|
| 483 |
+
"rewards/margins": 0.04101933166384697,
|
| 484 |
+
"rewards/rejected": -0.39896851778030396,
|
| 485 |
+
"step": 330
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.27,
|
| 489 |
+
"learning_rate": 4.588859120963282e-06,
|
| 490 |
+
"logits/chosen": -0.02518557570874691,
|
| 491 |
+
"logits/rejected": 0.010962575674057007,
|
| 492 |
+
"logps/chosen": -620.8372802734375,
|
| 493 |
+
"logps/rejected": -658.2938232421875,
|
| 494 |
+
"loss": 223632.35,
|
| 495 |
+
"rewards/accuracies": 0.5249999761581421,
|
| 496 |
+
"rewards/chosen": -0.333041250705719,
|
| 497 |
+
"rewards/margins": 0.04603361338376999,
|
| 498 |
+
"rewards/rejected": -0.3790748715400696,
|
| 499 |
+
"step": 340
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.27,
|
| 503 |
+
"learning_rate": 4.5503905911852435e-06,
|
| 504 |
+
"logits/chosen": 0.007638671435415745,
|
| 505 |
+
"logits/rejected": -0.031868986785411835,
|
| 506 |
+
"logps/chosen": -696.9804077148438,
|
| 507 |
+
"logps/rejected": -731.325439453125,
|
| 508 |
+
"loss": 219627.55,
|
| 509 |
+
"rewards/accuracies": 0.625,
|
| 510 |
+
"rewards/chosen": -0.3883155286312103,
|
| 511 |
+
"rewards/margins": 0.05170748755335808,
|
| 512 |
+
"rewards/rejected": -0.4400230050086975,
|
| 513 |
+
"step": 350
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.28,
|
| 517 |
+
"learning_rate": 4.510378591641036e-06,
|
| 518 |
+
"logits/chosen": 0.0013331234222278,
|
| 519 |
+
"logits/rejected": 0.0017831831937655807,
|
| 520 |
+
"logps/chosen": -748.482421875,
|
| 521 |
+
"logps/rejected": -801.7854614257812,
|
| 522 |
+
"loss": 227919.725,
|
| 523 |
+
"rewards/accuracies": 0.6312500238418579,
|
| 524 |
+
"rewards/chosen": -0.4426950514316559,
|
| 525 |
+
"rewards/margins": 0.03837837651371956,
|
| 526 |
+
"rewards/rejected": -0.48107343912124634,
|
| 527 |
+
"step": 360
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.29,
|
| 531 |
+
"learning_rate": 4.468853242109712e-06,
|
| 532 |
+
"logits/chosen": -0.035867899656295776,
|
| 533 |
+
"logits/rejected": -0.031078562140464783,
|
| 534 |
+
"logps/chosen": -724.6630859375,
|
| 535 |
+
"logps/rejected": -755.1024169921875,
|
| 536 |
+
"loss": 220950.85,
|
| 537 |
+
"rewards/accuracies": 0.574999988079071,
|
| 538 |
+
"rewards/chosen": -0.43880361318588257,
|
| 539 |
+
"rewards/margins": 0.05036715418100357,
|
| 540 |
+
"rewards/rejected": -0.4891708493232727,
|
| 541 |
+
"step": 370
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.3,
|
| 545 |
+
"learning_rate": 4.42584580157276e-06,
|
| 546 |
+
"logits/chosen": -0.013162782415747643,
|
| 547 |
+
"logits/rejected": -0.002673022449016571,
|
| 548 |
+
"logps/chosen": -783.6541748046875,
|
| 549 |
+
"logps/rejected": -803.5646362304688,
|
| 550 |
+
"loss": 218305.25,
|
| 551 |
+
"rewards/accuracies": 0.5687500238418579,
|
| 552 |
+
"rewards/chosen": -0.44587355852127075,
|
| 553 |
+
"rewards/margins": 0.06072569638490677,
|
| 554 |
+
"rewards/rejected": -0.5065993070602417,
|
| 555 |
+
"step": 380
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.31,
|
| 559 |
+
"learning_rate": 4.381388644683317e-06,
|
| 560 |
+
"logits/chosen": -0.014830539003014565,
|
| 561 |
+
"logits/rejected": -0.03430696204304695,
|
| 562 |
+
"logps/chosen": -691.3172607421875,
|
| 563 |
+
"logps/rejected": -730.2721557617188,
|
| 564 |
+
"loss": 231019.025,
|
| 565 |
+
"rewards/accuracies": 0.606249988079071,
|
| 566 |
+
"rewards/chosen": -0.3912193477153778,
|
| 567 |
+
"rewards/margins": 0.045134175568819046,
|
| 568 |
+
"rewards/rejected": -0.43635350465774536,
|
| 569 |
+
"step": 390
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.31,
|
| 573 |
+
"learning_rate": 4.33551523739555e-06,
|
| 574 |
+
"logits/chosen": -0.07614697515964508,
|
| 575 |
+
"logits/rejected": -0.05786391347646713,
|
| 576 |
+
"logps/chosen": -633.596923828125,
|
| 577 |
+
"logps/rejected": -678.1466674804688,
|
| 578 |
+
"loss": 218172.15,
|
| 579 |
+
"rewards/accuracies": 0.6187499761581421,
|
| 580 |
+
"rewards/chosen": -0.33879750967025757,
|
| 581 |
+
"rewards/margins": 0.05953192710876465,
|
| 582 |
+
"rewards/rejected": -0.398329496383667,
|
| 583 |
+
"step": 400
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.32,
|
| 587 |
+
"learning_rate": 4.288260111772535e-06,
|
| 588 |
+
"logits/chosen": -0.0014561188872903585,
|
| 589 |
+
"logits/rejected": 0.05796881392598152,
|
| 590 |
+
"logps/chosen": -645.994384765625,
|
| 591 |
+
"logps/rejected": -661.2260131835938,
|
| 592 |
+
"loss": 211935.25,
|
| 593 |
+
"rewards/accuracies": 0.625,
|
| 594 |
+
"rewards/chosen": -0.34373944997787476,
|
| 595 |
+
"rewards/margins": 0.05440125986933708,
|
| 596 |
+
"rewards/rejected": -0.39814066886901855,
|
| 597 |
+
"step": 410
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.33,
|
| 601 |
+
"learning_rate": 4.239658839991594e-06,
|
| 602 |
+
"logits/chosen": -0.06332430243492126,
|
| 603 |
+
"logits/rejected": -0.0031203762628138065,
|
| 604 |
+
"logps/chosen": -718.9708251953125,
|
| 605 |
+
"logps/rejected": -738.2911376953125,
|
| 606 |
+
"loss": 210876.625,
|
| 607 |
+
"rewards/accuracies": 0.6000000238418579,
|
| 608 |
+
"rewards/chosen": -0.4123147428035736,
|
| 609 |
+
"rewards/margins": 0.06380443274974823,
|
| 610 |
+
"rewards/rejected": -0.47611913084983826,
|
| 611 |
+
"step": 420
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.34,
|
| 615 |
+
"learning_rate": 4.189748007566686e-06,
|
| 616 |
+
"logits/chosen": -0.03928182274103165,
|
| 617 |
+
"logits/rejected": 0.015087981708347797,
|
| 618 |
+
"logps/chosen": -762.869384765625,
|
| 619 |
+
"logps/rejected": -791.5504760742188,
|
| 620 |
+
"loss": 198171.0625,
|
| 621 |
+
"rewards/accuracies": 0.668749988079071,
|
| 622 |
+
"rewards/chosen": -0.45352697372436523,
|
| 623 |
+
"rewards/margins": 0.07380986958742142,
|
| 624 |
+
"rewards/rejected": -0.5273367762565613,
|
| 625 |
+
"step": 430
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.35,
|
| 629 |
+
"learning_rate": 4.138565185807972e-06,
|
| 630 |
+
"logits/chosen": -0.011115262284874916,
|
| 631 |
+
"logits/rejected": 0.020389294251799583,
|
| 632 |
+
"logps/chosen": -725.9690551757812,
|
| 633 |
+
"logps/rejected": -757.0069580078125,
|
| 634 |
+
"loss": 200690.2,
|
| 635 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 636 |
+
"rewards/chosen": -0.40321993827819824,
|
| 637 |
+
"rewards/margins": 0.06992204487323761,
|
| 638 |
+
"rewards/rejected": -0.47314196825027466,
|
| 639 |
+
"step": 440
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.35,
|
| 643 |
+
"learning_rate": 4.086148903539311e-06,
|
| 644 |
+
"logits/chosen": 0.0033815137576311827,
|
| 645 |
+
"logits/rejected": 0.02135222777724266,
|
| 646 |
+
"logps/chosen": -694.905029296875,
|
| 647 |
+
"logps/rejected": -763.0654907226562,
|
| 648 |
+
"loss": 194541.9625,
|
| 649 |
+
"rewards/accuracies": 0.625,
|
| 650 |
+
"rewards/chosen": -0.39870306849479675,
|
| 651 |
+
"rewards/margins": 0.08366172015666962,
|
| 652 |
+
"rewards/rejected": -0.48236480355262756,
|
| 653 |
+
"step": 450
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.36,
|
| 657 |
+
"learning_rate": 4.032538618094972e-06,
|
| 658 |
+
"logits/chosen": 0.046546097844839096,
|
| 659 |
+
"logits/rejected": 0.06920139491558075,
|
| 660 |
+
"logps/chosen": -779.2876586914062,
|
| 661 |
+
"logps/rejected": -792.000732421875,
|
| 662 |
+
"loss": 223143.625,
|
| 663 |
+
"rewards/accuracies": 0.5874999761581421,
|
| 664 |
+
"rewards/chosen": -0.4378412365913391,
|
| 665 |
+
"rewards/margins": 0.04833221063017845,
|
| 666 |
+
"rewards/rejected": -0.48617345094680786,
|
| 667 |
+
"step": 460
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.37,
|
| 671 |
+
"learning_rate": 3.977774685617386e-06,
|
| 672 |
+
"logits/chosen": -0.024039577692747116,
|
| 673 |
+
"logits/rejected": 0.034949518740177155,
|
| 674 |
+
"logps/chosen": -762.0443115234375,
|
| 675 |
+
"logps/rejected": -840.4417724609375,
|
| 676 |
+
"loss": 188606.2,
|
| 677 |
+
"rewards/accuracies": 0.675000011920929,
|
| 678 |
+
"rewards/chosen": -0.4447455406188965,
|
| 679 |
+
"rewards/margins": 0.09278737008571625,
|
| 680 |
+
"rewards/rejected": -0.5375329256057739,
|
| 681 |
+
"step": 470
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.38,
|
| 685 |
+
"learning_rate": 3.92189833067831e-06,
|
| 686 |
+
"logits/chosen": -0.0034719116520136595,
|
| 687 |
+
"logits/rejected": 0.03559238836169243,
|
| 688 |
+
"logps/chosen": -727.2840576171875,
|
| 689 |
+
"logps/rejected": -749.8525390625,
|
| 690 |
+
"loss": 219781.9,
|
| 691 |
+
"rewards/accuracies": 0.6000000238418579,
|
| 692 |
+
"rewards/chosen": -0.39409303665161133,
|
| 693 |
+
"rewards/margins": 0.05169659107923508,
|
| 694 |
+
"rewards/rejected": -0.445789635181427,
|
| 695 |
+
"step": 480
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.38,
|
| 699 |
+
"learning_rate": 3.864951615246261e-06,
|
| 700 |
+
"logits/chosen": -0.10218246281147003,
|
| 701 |
+
"logits/rejected": -0.04928632080554962,
|
| 702 |
+
"logps/chosen": -696.552978515625,
|
| 703 |
+
"logps/rejected": -761.4454956054688,
|
| 704 |
+
"loss": 201005.8875,
|
| 705 |
+
"rewards/accuracies": 0.637499988079071,
|
| 706 |
+
"rewards/chosen": -0.3766392767429352,
|
| 707 |
+
"rewards/margins": 0.07472492754459381,
|
| 708 |
+
"rewards/rejected": -0.4513641893863678,
|
| 709 |
+
"step": 490
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.39,
|
| 713 |
+
"learning_rate": 3.806977407023581e-06,
|
| 714 |
+
"logits/chosen": -0.08053256571292877,
|
| 715 |
+
"logits/rejected": -0.027696877717971802,
|
| 716 |
+
"logps/chosen": -716.932373046875,
|
| 717 |
+
"logps/rejected": -746.5601196289062,
|
| 718 |
+
"loss": 210761.25,
|
| 719 |
+
"rewards/accuracies": 0.625,
|
| 720 |
+
"rewards/chosen": -0.3994150161743164,
|
| 721 |
+
"rewards/margins": 0.0657259151339531,
|
| 722 |
+
"rewards/rejected": -0.4651409089565277,
|
| 723 |
+
"step": 500
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.4,
|
| 727 |
+
"learning_rate": 3.7480193471769815e-06,
|
| 728 |
+
"logits/chosen": -0.021161776036024094,
|
| 729 |
+
"logits/rejected": 0.010638940148055553,
|
| 730 |
+
"logps/chosen": -752.3082275390625,
|
| 731 |
+
"logps/rejected": -824.42626953125,
|
| 732 |
+
"loss": 208462.75,
|
| 733 |
+
"rewards/accuracies": 0.6000000238418579,
|
| 734 |
+
"rewards/chosen": -0.4740660786628723,
|
| 735 |
+
"rewards/margins": 0.07625924050807953,
|
| 736 |
+
"rewards/rejected": -0.5503252744674683,
|
| 737 |
+
"step": 510
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.41,
|
| 741 |
+
"learning_rate": 3.6881218174858354e-06,
|
| 742 |
+
"logits/chosen": -0.08741439133882523,
|
| 743 |
+
"logits/rejected": 0.028159821406006813,
|
| 744 |
+
"logps/chosen": -808.7000732421875,
|
| 745 |
+
"logps/rejected": -843.1199340820312,
|
| 746 |
+
"loss": 204978.175,
|
| 747 |
+
"rewards/accuracies": 0.625,
|
| 748 |
+
"rewards/chosen": -0.4982606768608093,
|
| 749 |
+
"rewards/margins": 0.08009181916713715,
|
| 750 |
+
"rewards/rejected": -0.5783525109291077,
|
| 751 |
+
"step": 520
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.42,
|
| 755 |
+
"learning_rate": 3.627329906932964e-06,
|
| 756 |
+
"logits/chosen": -0.0961703211069107,
|
| 757 |
+
"logits/rejected": 0.01851445809006691,
|
| 758 |
+
"logps/chosen": -785.6272583007812,
|
| 759 |
+
"logps/rejected": -845.8230590820312,
|
| 760 |
+
"loss": 202598.85,
|
| 761 |
+
"rewards/accuracies": 0.612500011920929,
|
| 762 |
+
"rewards/chosen": -0.4964609146118164,
|
| 763 |
+
"rewards/margins": 0.08662732690572739,
|
| 764 |
+
"rewards/rejected": -0.583088219165802,
|
| 765 |
+
"step": 530
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.42,
|
| 769 |
+
"learning_rate": 3.5656893777630686e-06,
|
| 770 |
+
"logits/chosen": -0.09825177490711212,
|
| 771 |
+
"logits/rejected": -0.0450114943087101,
|
| 772 |
+
"logps/chosen": -800.4473266601562,
|
| 773 |
+
"logps/rejected": -877.4309692382812,
|
| 774 |
+
"loss": 206751.5,
|
| 775 |
+
"rewards/accuracies": 0.625,
|
| 776 |
+
"rewards/chosen": -0.49454784393310547,
|
| 777 |
+
"rewards/margins": 0.08050151914358139,
|
| 778 |
+
"rewards/rejected": -0.5750494003295898,
|
| 779 |
+
"step": 540
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.43,
|
| 783 |
+
"learning_rate": 3.503246631034345e-06,
|
| 784 |
+
"logits/chosen": -0.03669491782784462,
|
| 785 |
+
"logits/rejected": 0.04769158363342285,
|
| 786 |
+
"logps/chosen": -901.7119140625,
|
| 787 |
+
"logps/rejected": -944.8836059570312,
|
| 788 |
+
"loss": 232280.975,
|
| 789 |
+
"rewards/accuracies": 0.581250011920929,
|
| 790 |
+
"rewards/chosen": -0.5475555658340454,
|
| 791 |
+
"rewards/margins": 0.05907317250967026,
|
| 792 |
+
"rewards/rejected": -0.6066287159919739,
|
| 793 |
+
"step": 550
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.44,
|
| 797 |
+
"learning_rate": 3.440048671689219e-06,
|
| 798 |
+
"logits/chosen": -0.06334514915943146,
|
| 799 |
+
"logits/rejected": 0.0007638297975063324,
|
| 800 |
+
"logps/chosen": -795.4130249023438,
|
| 801 |
+
"logps/rejected": -868.2384643554688,
|
| 802 |
+
"loss": 209493.875,
|
| 803 |
+
"rewards/accuracies": 0.6187499761581421,
|
| 804 |
+
"rewards/chosen": -0.5169624090194702,
|
| 805 |
+
"rewards/margins": 0.07504051178693771,
|
| 806 |
+
"rewards/rejected": -0.5920029878616333,
|
| 807 |
+
"step": 560
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.45,
|
| 811 |
+
"learning_rate": 3.3761430731705056e-06,
|
| 812 |
+
"logits/chosen": 0.016527634114027023,
|
| 813 |
+
"logits/rejected": 0.005024082958698273,
|
| 814 |
+
"logps/chosen": -785.9190673828125,
|
| 815 |
+
"logps/rejected": -843.1102294921875,
|
| 816 |
+
"loss": 221721.375,
|
| 817 |
+
"rewards/accuracies": 0.543749988079071,
|
| 818 |
+
"rewards/chosen": -0.468578577041626,
|
| 819 |
+
"rewards/margins": 0.05255637317895889,
|
| 820 |
+
"rewards/rejected": -0.5211349725723267,
|
| 821 |
+
"step": 570
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.46,
|
| 825 |
+
"learning_rate": 3.311577941609604e-06,
|
| 826 |
+
"logits/chosen": -0.06407289206981659,
|
| 827 |
+
"logits/rejected": -0.07606904208660126,
|
| 828 |
+
"logps/chosen": -685.833740234375,
|
| 829 |
+
"logps/rejected": -754.46240234375,
|
| 830 |
+
"loss": 213850.075,
|
| 831 |
+
"rewards/accuracies": 0.5687500238418579,
|
| 832 |
+
"rewards/chosen": -0.4029085040092468,
|
| 833 |
+
"rewards/margins": 0.06034379079937935,
|
| 834 |
+
"rewards/rejected": -0.46325230598449707,
|
| 835 |
+
"step": 580
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.46,
|
| 839 |
+
"learning_rate": 3.2464018796137157e-06,
|
| 840 |
+
"logits/chosen": -0.1008271723985672,
|
| 841 |
+
"logits/rejected": -0.02092236839234829,
|
| 842 |
+
"logps/chosen": -745.2072143554688,
|
| 843 |
+
"logps/rejected": -780.5989379882812,
|
| 844 |
+
"loss": 204456.325,
|
| 845 |
+
"rewards/accuracies": 0.625,
|
| 846 |
+
"rewards/chosen": -0.42655831575393677,
|
| 847 |
+
"rewards/margins": 0.07653003931045532,
|
| 848 |
+
"rewards/rejected": -0.5030883550643921,
|
| 849 |
+
"step": 590
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.47,
|
| 853 |
+
"learning_rate": 3.1806639496793245e-06,
|
| 854 |
+
"logits/chosen": -0.12006688117980957,
|
| 855 |
+
"logits/rejected": -0.09113277494907379,
|
| 856 |
+
"logps/chosen": -725.4483642578125,
|
| 857 |
+
"logps/rejected": -775.4747314453125,
|
| 858 |
+
"loss": 213843.55,
|
| 859 |
+
"rewards/accuracies": 0.612500011920929,
|
| 860 |
+
"rewards/chosen": -0.4422362446784973,
|
| 861 |
+
"rewards/margins": 0.06481991708278656,
|
| 862 |
+
"rewards/rejected": -0.5070561170578003,
|
| 863 |
+
"step": 600
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.48,
|
| 867 |
+
"learning_rate": 3.114413637259484e-06,
|
| 868 |
+
"logits/chosen": -0.09821692854166031,
|
| 869 |
+
"logits/rejected": -0.005470535717904568,
|
| 870 |
+
"logps/chosen": -746.876708984375,
|
| 871 |
+
"logps/rejected": -782.652587890625,
|
| 872 |
+
"loss": 219806.95,
|
| 873 |
+
"rewards/accuracies": 0.6187499761581421,
|
| 874 |
+
"rewards/chosen": -0.4441962242126465,
|
| 875 |
+
"rewards/margins": 0.06350026279687881,
|
| 876 |
+
"rewards/rejected": -0.5076965093612671,
|
| 877 |
+
"step": 610
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.49,
|
| 881 |
+
"learning_rate": 3.0477008135127247e-06,
|
| 882 |
+
"logits/chosen": -0.09074200689792633,
|
| 883 |
+
"logits/rejected": -0.03884665295481682,
|
| 884 |
+
"logps/chosen": -720.9779052734375,
|
| 885 |
+
"logps/rejected": -743.6715087890625,
|
| 886 |
+
"loss": 208771.95,
|
| 887 |
+
"rewards/accuracies": 0.625,
|
| 888 |
+
"rewards/chosen": -0.40446630120277405,
|
| 889 |
+
"rewards/margins": 0.062393445521593094,
|
| 890 |
+
"rewards/rejected": -0.46685975790023804,
|
| 891 |
+
"step": 620
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.49,
|
| 895 |
+
"learning_rate": 2.980575697761603e-06,
|
| 896 |
+
"logits/chosen": -0.0913369357585907,
|
| 897 |
+
"logits/rejected": -0.04820000380277634,
|
| 898 |
+
"logps/chosen": -739.7911376953125,
|
| 899 |
+
"logps/rejected": -790.2059936523438,
|
| 900 |
+
"loss": 201716.1,
|
| 901 |
+
"rewards/accuracies": 0.6812499761581421,
|
| 902 |
+
"rewards/chosen": -0.4266403317451477,
|
| 903 |
+
"rewards/margins": 0.07397651672363281,
|
| 904 |
+
"rewards/rejected": -0.5006168484687805,
|
| 905 |
+
"step": 630
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.5,
|
| 909 |
+
"learning_rate": 2.9130888196891755e-06,
|
| 910 |
+
"logits/chosen": -0.06239433214068413,
|
| 911 |
+
"logits/rejected": 0.019041964784264565,
|
| 912 |
+
"logps/chosen": -717.3048095703125,
|
| 913 |
+
"logps/rejected": -778.1002807617188,
|
| 914 |
+
"loss": 214451.2,
|
| 915 |
+
"rewards/accuracies": 0.625,
|
| 916 |
+
"rewards/chosen": -0.4269956946372986,
|
| 917 |
+
"rewards/margins": 0.07155928760766983,
|
| 918 |
+
"rewards/rejected": -0.4985550045967102,
|
| 919 |
+
"step": 640
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 0.51,
|
| 923 |
+
"learning_rate": 2.845290981301834e-06,
|
| 924 |
+
"logits/chosen": -0.04994695261120796,
|
| 925 |
+
"logits/rejected": -0.04595662280917168,
|
| 926 |
+
"logps/chosen": -726.4156494140625,
|
| 927 |
+
"logps/rejected": -770.7301025390625,
|
| 928 |
+
"loss": 211533.975,
|
| 929 |
+
"rewards/accuracies": 0.574999988079071,
|
| 930 |
+
"rewards/chosen": -0.4114510118961334,
|
| 931 |
+
"rewards/margins": 0.06760390847921371,
|
| 932 |
+
"rewards/rejected": -0.47905483841896057,
|
| 933 |
+
"step": 650
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.52,
|
| 937 |
+
"learning_rate": 2.7772332186871464e-06,
|
| 938 |
+
"logits/chosen": -0.0662728101015091,
|
| 939 |
+
"logits/rejected": -0.038154177367687225,
|
| 940 |
+
"logps/chosen": -671.4014892578125,
|
| 941 |
+
"logps/rejected": -740.8782348632812,
|
| 942 |
+
"loss": 198928.3625,
|
| 943 |
+
"rewards/accuracies": 0.6000000238418579,
|
| 944 |
+
"rewards/chosen": -0.37442153692245483,
|
| 945 |
+
"rewards/margins": 0.07407635450363159,
|
| 946 |
+
"rewards/rejected": -0.44849786162376404,
|
| 947 |
+
"step": 660
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.53,
|
| 951 |
+
"learning_rate": 2.708966763595493e-06,
|
| 952 |
+
"logits/chosen": -0.11307473480701447,
|
| 953 |
+
"logits/rejected": -0.09715189039707184,
|
| 954 |
+
"logps/chosen": -669.0352783203125,
|
| 955 |
+
"logps/rejected": -714.9685668945312,
|
| 956 |
+
"loss": 209100.4625,
|
| 957 |
+
"rewards/accuracies": 0.5562499761581421,
|
| 958 |
+
"rewards/chosen": -0.37991762161254883,
|
| 959 |
+
"rewards/margins": 0.07245246320962906,
|
| 960 |
+
"rewards/rejected": -0.4523701071739197,
|
| 961 |
+
"step": 670
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 0.53,
|
| 965 |
+
"learning_rate": 2.640543004874409e-06,
|
| 966 |
+
"logits/chosen": -0.05394214391708374,
|
| 967 |
+
"logits/rejected": -0.05522084981203079,
|
| 968 |
+
"logps/chosen": -693.9007568359375,
|
| 969 |
+
"logps/rejected": -747.3922729492188,
|
| 970 |
+
"loss": 210209.575,
|
| 971 |
+
"rewards/accuracies": 0.6000000238418579,
|
| 972 |
+
"rewards/chosen": -0.39138975739479065,
|
| 973 |
+
"rewards/margins": 0.07271454483270645,
|
| 974 |
+
"rewards/rejected": -0.4641042649745941,
|
| 975 |
+
"step": 680
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.54,
|
| 979 |
+
"learning_rate": 2.572013449784671e-06,
|
| 980 |
+
"logits/chosen": -0.027393341064453125,
|
| 981 |
+
"logits/rejected": -0.05223383754491806,
|
| 982 |
+
"logps/chosen": -677.8887329101562,
|
| 983 |
+
"logps/rejected": -740.2653198242188,
|
| 984 |
+
"loss": 208986.8125,
|
| 985 |
+
"rewards/accuracies": 0.625,
|
| 986 |
+
"rewards/chosen": -0.4014677107334137,
|
| 987 |
+
"rewards/margins": 0.07114878296852112,
|
| 988 |
+
"rewards/rejected": -0.4726164937019348,
|
| 989 |
+
"step": 690
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 0.55,
|
| 993 |
+
"learning_rate": 2.503429685227245e-06,
|
| 994 |
+
"logits/chosen": -0.01898156851530075,
|
| 995 |
+
"logits/rejected": -0.024372858926653862,
|
| 996 |
+
"logps/chosen": -723.0344848632812,
|
| 997 |
+
"logps/rejected": -742.4080810546875,
|
| 998 |
+
"loss": 197337.45,
|
| 999 |
+
"rewards/accuracies": 0.606249988079071,
|
| 1000 |
+
"rewards/chosen": -0.3911907970905304,
|
| 1001 |
+
"rewards/margins": 0.08259539306163788,
|
| 1002 |
+
"rewards/rejected": -0.4737861752510071,
|
| 1003 |
+
"step": 700
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.56,
|
| 1007 |
+
"learning_rate": 2.434843338910286e-06,
|
| 1008 |
+
"logits/chosen": -0.05535392835736275,
|
| 1009 |
+
"logits/rejected": -0.07486173510551453,
|
| 1010 |
+
"logps/chosen": -754.6624755859375,
|
| 1011 |
+
"logps/rejected": -807.8843994140625,
|
| 1012 |
+
"loss": 199448.225,
|
| 1013 |
+
"rewards/accuracies": 0.65625,
|
| 1014 |
+
"rewards/chosen": -0.4616442620754242,
|
| 1015 |
+
"rewards/margins": 0.08079784363508224,
|
| 1016 |
+
"rewards/rejected": -0.5424421429634094,
|
| 1017 |
+
"step": 710
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.57,
|
| 1021 |
+
"learning_rate": 2.3663060404854155e-06,
|
| 1022 |
+
"logits/chosen": -0.047691427171230316,
|
| 1023 |
+
"logits/rejected": 0.02117311768233776,
|
| 1024 |
+
"logps/chosen": -810.5379028320312,
|
| 1025 |
+
"logps/rejected": -838.4991455078125,
|
| 1026 |
+
"loss": 222795.0,
|
| 1027 |
+
"rewards/accuracies": 0.5562499761581421,
|
| 1028 |
+
"rewards/chosen": -0.4926871359348297,
|
| 1029 |
+
"rewards/margins": 0.0528423972427845,
|
| 1030 |
+
"rewards/rejected": -0.5455294847488403,
|
| 1031 |
+
"step": 720
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.57,
|
| 1035 |
+
"learning_rate": 2.2978693826825406e-06,
|
| 1036 |
+
"logits/chosen": -0.05972999334335327,
|
| 1037 |
+
"logits/rejected": 0.053670674562454224,
|
| 1038 |
+
"logps/chosen": -771.9221801757812,
|
| 1039 |
+
"logps/rejected": -843.3212890625,
|
| 1040 |
+
"loss": 208553.9125,
|
| 1041 |
+
"rewards/accuracies": 0.675000011920929,
|
| 1042 |
+
"rewards/chosen": -0.46609845757484436,
|
| 1043 |
+
"rewards/margins": 0.07322056591510773,
|
| 1044 |
+
"rewards/rejected": -0.5393189191818237,
|
| 1045 |
+
"step": 730
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 0.58,
|
| 1049 |
+
"learning_rate": 2.2295848824724612e-06,
|
| 1050 |
+
"logits/chosen": -0.03266960754990578,
|
| 1051 |
+
"logits/rejected": 0.05311200022697449,
|
| 1052 |
+
"logps/chosen": -811.7847900390625,
|
| 1053 |
+
"logps/rejected": -888.1708984375,
|
| 1054 |
+
"loss": 196582.3,
|
| 1055 |
+
"rewards/accuracies": 0.675000011920929,
|
| 1056 |
+
"rewards/chosen": -0.4684126377105713,
|
| 1057 |
+
"rewards/margins": 0.09803617745637894,
|
| 1058 |
+
"rewards/rejected": -0.5664488077163696,
|
| 1059 |
+
"step": 740
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.59,
|
| 1063 |
+
"learning_rate": 2.1615039422865136e-06,
|
| 1064 |
+
"logits/chosen": -0.02488083206117153,
|
| 1065 |
+
"logits/rejected": 0.00421232171356678,
|
| 1066 |
+
"logps/chosen": -749.4576416015625,
|
| 1067 |
+
"logps/rejected": -815.193115234375,
|
| 1068 |
+
"loss": 208486.4,
|
| 1069 |
+
"rewards/accuracies": 0.574999988079071,
|
| 1070 |
+
"rewards/chosen": -0.4397048354148865,
|
| 1071 |
+
"rewards/margins": 0.08280255645513535,
|
| 1072 |
+
"rewards/rejected": -0.52250736951828,
|
| 1073 |
+
"step": 750
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 0.6,
|
| 1077 |
+
"learning_rate": 2.0936778113224253e-06,
|
| 1078 |
+
"logits/chosen": 0.0021423858124762774,
|
| 1079 |
+
"logits/rejected": 0.0016391072422266006,
|
| 1080 |
+
"logps/chosen": -767.9728393554688,
|
| 1081 |
+
"logps/rejected": -829.259765625,
|
| 1082 |
+
"loss": 207514.8625,
|
| 1083 |
+
"rewards/accuracies": 0.59375,
|
| 1084 |
+
"rewards/chosen": -0.44562071561813354,
|
| 1085 |
+
"rewards/margins": 0.07032287120819092,
|
| 1086 |
+
"rewards/rejected": -0.5159436464309692,
|
| 1087 |
+
"step": 760
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 0.6,
|
| 1091 |
+
"learning_rate": 2.0261575469655304e-06,
|
| 1092 |
+
"logits/chosen": -0.03952546417713165,
|
| 1093 |
+
"logits/rejected": 0.07452509552240372,
|
| 1094 |
+
"logps/chosen": -799.8232421875,
|
| 1095 |
+
"logps/rejected": -832.5133666992188,
|
| 1096 |
+
"loss": 207497.575,
|
| 1097 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 1098 |
+
"rewards/chosen": -0.4807282090187073,
|
| 1099 |
+
"rewards/margins": 0.07494666427373886,
|
| 1100 |
+
"rewards/rejected": -0.5556748509407043,
|
| 1101 |
+
"step": 770
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.61,
|
| 1105 |
+
"learning_rate": 1.9589939763543693e-06,
|
| 1106 |
+
"logits/chosen": 0.030634239315986633,
|
| 1107 |
+
"logits/rejected": 0.04007285460829735,
|
| 1108 |
+
"logps/chosen": -783.6361083984375,
|
| 1109 |
+
"logps/rejected": -811.500732421875,
|
| 1110 |
+
"loss": 222795.075,
|
| 1111 |
+
"rewards/accuracies": 0.606249988079071,
|
| 1112 |
+
"rewards/chosen": -0.48045220971107483,
|
| 1113 |
+
"rewards/margins": 0.06402502954006195,
|
| 1114 |
+
"rewards/rejected": -0.5444772839546204,
|
| 1115 |
+
"step": 780
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 0.62,
|
| 1119 |
+
"learning_rate": 1.8922376581196107e-06,
|
| 1120 |
+
"logits/chosen": 0.03562704846262932,
|
| 1121 |
+
"logits/rejected": 0.09303382784128189,
|
| 1122 |
+
"logps/chosen": -801.5706787109375,
|
| 1123 |
+
"logps/rejected": -821.78564453125,
|
| 1124 |
+
"loss": 210521.95,
|
| 1125 |
+
"rewards/accuracies": 0.612500011920929,
|
| 1126 |
+
"rewards/chosen": -0.4747587740421295,
|
| 1127 |
+
"rewards/margins": 0.06814552843570709,
|
| 1128 |
+
"rewards/rejected": -0.5429043173789978,
|
| 1129 |
+
"step": 790
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 0.63,
|
| 1133 |
+
"learning_rate": 1.8259388443250993e-06,
|
| 1134 |
+
"logits/chosen": 0.0239779744297266,
|
| 1135 |
+
"logits/rejected": 0.05331903696060181,
|
| 1136 |
+
"logps/chosen": -768.3508911132812,
|
| 1137 |
+
"logps/rejected": -849.29296875,
|
| 1138 |
+
"loss": 193255.2625,
|
| 1139 |
+
"rewards/accuracies": 0.6625000238418579,
|
| 1140 |
+
"rewards/chosen": -0.4437880516052246,
|
| 1141 |
+
"rewards/margins": 0.08583388477563858,
|
| 1142 |
+
"rewards/rejected": -0.5296218991279602,
|
| 1143 |
+
"step": 800
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.64,
|
| 1147 |
+
"learning_rate": 1.760147442639679e-06,
|
| 1148 |
+
"logits/chosen": 0.038005001842975616,
|
| 1149 |
+
"logits/rejected": 0.03408225253224373,
|
| 1150 |
+
"logps/chosen": -798.1541748046875,
|
| 1151 |
+
"logps/rejected": -828.2268676757812,
|
| 1152 |
+
"loss": 208489.85,
|
| 1153 |
+
"rewards/accuracies": 0.668749988079071,
|
| 1154 |
+
"rewards/chosen": -0.4753784239292145,
|
| 1155 |
+
"rewards/margins": 0.0705508217215538,
|
| 1156 |
+
"rewards/rejected": -0.5459292531013489,
|
| 1157 |
+
"step": 810
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 0.64,
|
| 1161 |
+
"learning_rate": 1.6949129787682628e-06,
|
| 1162 |
+
"logits/chosen": 0.03351437300443649,
|
| 1163 |
+
"logits/rejected": 0.0348244272172451,
|
| 1164 |
+
"logps/chosen": -833.7937622070312,
|
| 1165 |
+
"logps/rejected": -868.5783081054688,
|
| 1166 |
+
"loss": 203584.05,
|
| 1167 |
+
"rewards/accuracies": 0.643750011920929,
|
| 1168 |
+
"rewards/chosen": -0.4822169840335846,
|
| 1169 |
+
"rewards/margins": 0.07765035331249237,
|
| 1170 |
+
"rewards/rejected": -0.5598673224449158,
|
| 1171 |
+
"step": 820
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 0.65,
|
| 1175 |
+
"learning_rate": 1.6302845591704348e-06,
|
| 1176 |
+
"logits/chosen": 0.028645271435379982,
|
| 1177 |
+
"logits/rejected": 0.03767635300755501,
|
| 1178 |
+
"logps/chosen": -773.2454833984375,
|
| 1179 |
+
"logps/rejected": -871.0046997070312,
|
| 1180 |
+
"loss": 205490.675,
|
| 1181 |
+
"rewards/accuracies": 0.612500011920929,
|
| 1182 |
+
"rewards/chosen": -0.504454493522644,
|
| 1183 |
+
"rewards/margins": 0.07273290306329727,
|
| 1184 |
+
"rewards/rejected": -0.5771873593330383,
|
| 1185 |
+
"step": 830
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 0.66,
|
| 1189 |
+
"learning_rate": 1.5663108340946465e-06,
|
| 1190 |
+
"logits/chosen": -0.016361240297555923,
|
| 1191 |
+
"logits/rejected": 0.02944091334939003,
|
| 1192 |
+
"logps/chosen": -810.2437744140625,
|
| 1193 |
+
"logps/rejected": -861.1094970703125,
|
| 1194 |
+
"loss": 201889.525,
|
| 1195 |
+
"rewards/accuracies": 0.625,
|
| 1196 |
+
"rewards/chosen": -0.5184367895126343,
|
| 1197 |
+
"rewards/margins": 0.078752800822258,
|
| 1198 |
+
"rewards/rejected": -0.5971895456314087,
|
| 1199 |
+
"step": 840
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 0.67,
|
| 1203 |
+
"learning_rate": 1.5030399609558364e-06,
|
| 1204 |
+
"logits/chosen": -0.04980049282312393,
|
| 1205 |
+
"logits/rejected": -0.0025185912381857634,
|
| 1206 |
+
"logps/chosen": -838.6681518554688,
|
| 1207 |
+
"logps/rejected": -880.5914916992188,
|
| 1208 |
+
"loss": 201939.1,
|
| 1209 |
+
"rewards/accuracies": 0.6625000238418579,
|
| 1210 |
+
"rewards/chosen": -0.512233555316925,
|
| 1211 |
+
"rewards/margins": 0.08945199102163315,
|
| 1212 |
+
"rewards/rejected": -0.6016855835914612,
|
| 1213 |
+
"step": 850
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 0.68,
|
| 1217 |
+
"learning_rate": 1.4405195680840357e-06,
|
| 1218 |
+
"logits/chosen": -0.006730606313794851,
|
| 1219 |
+
"logits/rejected": 0.008829834870994091,
|
| 1220 |
+
"logps/chosen": -809.2523193359375,
|
| 1221 |
+
"logps/rejected": -881.7073974609375,
|
| 1222 |
+
"loss": 201830.9875,
|
| 1223 |
+
"rewards/accuracies": 0.606249988079071,
|
| 1224 |
+
"rewards/chosen": -0.4848810136318207,
|
| 1225 |
+
"rewards/margins": 0.08851764351129532,
|
| 1226 |
+
"rewards/rejected": -0.5733986496925354,
|
| 1227 |
+
"step": 860
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 0.68,
|
| 1231 |
+
"learning_rate": 1.378796718871252e-06,
|
| 1232 |
+
"logits/chosen": 0.003451726632192731,
|
| 1233 |
+
"logits/rejected": 0.04693222790956497,
|
| 1234 |
+
"logps/chosen": -828.8895263671875,
|
| 1235 |
+
"logps/rejected": -869.78515625,
|
| 1236 |
+
"loss": 223132.05,
|
| 1237 |
+
"rewards/accuracies": 0.5625,
|
| 1238 |
+
"rewards/chosen": -0.5101678371429443,
|
| 1239 |
+
"rewards/margins": 0.06179608777165413,
|
| 1240 |
+
"rewards/rejected": -0.571963906288147,
|
| 1241 |
+
"step": 870
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 0.69,
|
| 1245 |
+
"learning_rate": 1.3179178763436302e-06,
|
| 1246 |
+
"logits/chosen": -0.00423568207770586,
|
| 1247 |
+
"logits/rejected": 0.005345098674297333,
|
| 1248 |
+
"logps/chosen": -770.8553466796875,
|
| 1249 |
+
"logps/rejected": -826.5133666992188,
|
| 1250 |
+
"loss": 220881.4,
|
| 1251 |
+
"rewards/accuracies": 0.6000000238418579,
|
| 1252 |
+
"rewards/chosen": -0.48375120759010315,
|
| 1253 |
+
"rewards/margins": 0.054837919771671295,
|
| 1254 |
+
"rewards/rejected": -0.5385891199111938,
|
| 1255 |
+
"step": 880
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 0.7,
|
| 1259 |
+
"learning_rate": 1.2579288681855364e-06,
|
| 1260 |
+
"logits/chosen": -0.04643459990620613,
|
| 1261 |
+
"logits/rejected": -0.00930569414049387,
|
| 1262 |
+
"logps/chosen": -800.4423828125,
|
| 1263 |
+
"logps/rejected": -868.2879638671875,
|
| 1264 |
+
"loss": 199618.2,
|
| 1265 |
+
"rewards/accuracies": 0.625,
|
| 1266 |
+
"rewards/chosen": -0.48793333768844604,
|
| 1267 |
+
"rewards/margins": 0.08429589867591858,
|
| 1268 |
+
"rewards/rejected": -0.572229266166687,
|
| 1269 |
+
"step": 890
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 0.71,
|
| 1273 |
+
"learning_rate": 1.1988748522419163e-06,
|
| 1274 |
+
"logits/chosen": -0.021517012268304825,
|
| 1275 |
+
"logits/rejected": -0.0038474828470498323,
|
| 1276 |
+
"logps/chosen": -755.4798583984375,
|
| 1277 |
+
"logps/rejected": -817.7532348632812,
|
| 1278 |
+
"loss": 212245.85,
|
| 1279 |
+
"rewards/accuracies": 0.6000000238418579,
|
| 1280 |
+
"rewards/chosen": -0.46443241834640503,
|
| 1281 |
+
"rewards/margins": 0.0739377960562706,
|
| 1282 |
+
"rewards/rejected": -0.5383703112602234,
|
| 1283 |
+
"step": 900
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 0.71,
|
| 1287 |
+
"learning_rate": 1.1408002825248842e-06,
|
| 1288 |
+
"logits/chosen": -0.08825428783893585,
|
| 1289 |
+
"logits/rejected": 0.0003117397427558899,
|
| 1290 |
+
"logps/chosen": -738.525146484375,
|
| 1291 |
+
"logps/rejected": -810.4708251953125,
|
| 1292 |
+
"loss": 195269.8,
|
| 1293 |
+
"rewards/accuracies": 0.7124999761581421,
|
| 1294 |
+
"rewards/chosen": -0.45489567518234253,
|
| 1295 |
+
"rewards/margins": 0.08346293121576309,
|
| 1296 |
+
"rewards/rejected": -0.5383585691452026,
|
| 1297 |
+
"step": 910
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 0.72,
|
| 1301 |
+
"learning_rate": 1.0837488757501369e-06,
|
| 1302 |
+
"logits/chosen": -0.03578663989901543,
|
| 1303 |
+
"logits/rejected": -0.04554619640111923,
|
| 1304 |
+
"logps/chosen": -734.4967651367188,
|
| 1305 |
+
"logps/rejected": -777.3861694335938,
|
| 1306 |
+
"loss": 216987.525,
|
| 1307 |
+
"rewards/accuracies": 0.53125,
|
| 1308 |
+
"rewards/chosen": -0.44713911414146423,
|
| 1309 |
+
"rewards/margins": 0.06695397198200226,
|
| 1310 |
+
"rewards/rejected": -0.5140931010246277,
|
| 1311 |
+
"step": 920
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 0.73,
|
| 1315 |
+
"learning_rate": 1.027763578428379e-06,
|
| 1316 |
+
"logits/chosen": -0.04421938955783844,
|
| 1317 |
+
"logits/rejected": 0.01812281273305416,
|
| 1318 |
+
"logps/chosen": -745.9063720703125,
|
| 1319 |
+
"logps/rejected": -811.608642578125,
|
| 1320 |
+
"loss": 198343.5375,
|
| 1321 |
+
"rewards/accuracies": 0.6499999761581421,
|
| 1322 |
+
"rewards/chosen": -0.4607219696044922,
|
| 1323 |
+
"rewards/margins": 0.08409135043621063,
|
| 1324 |
+
"rewards/rejected": -0.5448133945465088,
|
| 1325 |
+
"step": 930
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 0.74,
|
| 1329 |
+
"learning_rate": 9.728865345365379e-07,
|
| 1330 |
+
"logits/chosen": -0.0179077610373497,
|
| 1331 |
+
"logits/rejected": 0.008015085011720657,
|
| 1332 |
+
"logps/chosen": -736.4071655273438,
|
| 1333 |
+
"logps/rejected": -788.5194091796875,
|
| 1334 |
+
"loss": 210246.7,
|
| 1335 |
+
"rewards/accuracies": 0.59375,
|
| 1336 |
+
"rewards/chosen": -0.45147353410720825,
|
| 1337 |
+
"rewards/margins": 0.07040319591760635,
|
| 1338 |
+
"rewards/rejected": -0.5218766927719116,
|
| 1339 |
+
"step": 940
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 0.75,
|
| 1343 |
+
"learning_rate": 9.191590537930975e-07,
|
| 1344 |
+
"logits/chosen": -0.07539006322622299,
|
| 1345 |
+
"logits/rejected": 0.06443232297897339,
|
| 1346 |
+
"logps/chosen": -848.2332153320312,
|
| 1347 |
+
"logps/rejected": -881.9641723632812,
|
| 1348 |
+
"loss": 199938.1625,
|
| 1349 |
+
"rewards/accuracies": 0.65625,
|
| 1350 |
+
"rewards/chosen": -0.49263858795166016,
|
| 1351 |
+
"rewards/margins": 0.0914057120680809,
|
| 1352 |
+
"rewards/rejected": -0.5840442776679993,
|
| 1353 |
+
"step": 950
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 0.75,
|
| 1357 |
+
"learning_rate": 8.666215805614373e-07,
|
| 1358 |
+
"logits/chosen": -0.035882774740457535,
|
| 1359 |
+
"logits/rejected": -0.0026063353288918734,
|
| 1360 |
+
"logps/chosen": -816.1239013671875,
|
| 1361 |
+
"logps/rejected": -860.7882690429688,
|
| 1362 |
+
"loss": 195339.6,
|
| 1363 |
+
"rewards/accuracies": 0.637499988079071,
|
| 1364 |
+
"rewards/chosen": -0.47968846559524536,
|
| 1365 |
+
"rewards/margins": 0.08311482518911362,
|
| 1366 |
+
"rewards/rejected": -0.5628032088279724,
|
| 1367 |
+
"step": 960
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 0.76,
|
| 1371 |
+
"learning_rate": 8.153136634045844e-07,
|
| 1372 |
+
"logits/chosen": -0.021835245192050934,
|
| 1373 |
+
"logits/rejected": 0.0024397834204137325,
|
| 1374 |
+
"logps/chosen": -794.1774291992188,
|
| 1375 |
+
"logps/rejected": -856.4476318359375,
|
| 1376 |
+
"loss": 184634.5375,
|
| 1377 |
+
"rewards/accuracies": 0.699999988079071,
|
| 1378 |
+
"rewards/chosen": -0.48376354575157166,
|
| 1379 |
+
"rewards/margins": 0.10054849088191986,
|
| 1380 |
+
"rewards/rejected": -0.5843120813369751,
|
| 1381 |
+
"step": 970
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 0.77,
|
| 1385 |
+
"learning_rate": 7.652739253142915e-07,
|
| 1386 |
+
"logits/chosen": -0.009013411588966846,
|
| 1387 |
+
"logits/rejected": 0.01207656692713499,
|
| 1388 |
+
"logps/chosen": -827.5479736328125,
|
| 1389 |
+
"logps/rejected": -872.8331909179688,
|
| 1390 |
+
"loss": 205367.375,
|
| 1391 |
+
"rewards/accuracies": 0.606249988079071,
|
| 1392 |
+
"rewards/chosen": -0.5190316438674927,
|
| 1393 |
+
"rewards/margins": 0.0740259662270546,
|
| 1394 |
+
"rewards/rejected": -0.5930576324462891,
|
| 1395 |
+
"step": 980
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 0.78,
|
| 1399 |
+
"learning_rate": 7.165400346368648e-07,
|
| 1400 |
+
"logits/chosen": -0.027726858854293823,
|
| 1401 |
+
"logits/rejected": -0.0028947251848876476,
|
| 1402 |
+
"logps/chosen": -791.7979736328125,
|
| 1403 |
+
"logps/rejected": -881.1510620117188,
|
| 1404 |
+
"loss": 197640.8625,
|
| 1405 |
+
"rewards/accuracies": 0.668749988079071,
|
| 1406 |
+
"rewards/chosen": -0.5166811943054199,
|
| 1407 |
+
"rewards/margins": 0.094887375831604,
|
| 1408 |
+
"rewards/rejected": -0.6115685701370239,
|
| 1409 |
+
"step": 990
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 0.79,
|
| 1413 |
+
"learning_rate": 6.691486767176092e-07,
|
| 1414 |
+
"logits/chosen": -0.025321567431092262,
|
| 1415 |
+
"logits/rejected": 0.03234836459159851,
|
| 1416 |
+
"logps/chosen": -768.285400390625,
|
| 1417 |
+
"logps/rejected": -811.7093505859375,
|
| 1418 |
+
"loss": 219588.225,
|
| 1419 |
+
"rewards/accuracies": 0.550000011920929,
|
| 1420 |
+
"rewards/chosen": -0.5012341737747192,
|
| 1421 |
+
"rewards/margins": 0.06108912080526352,
|
| 1422 |
+
"rewards/rejected": -0.5623233318328857,
|
| 1423 |
+
"step": 1000
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 0.79,
|
| 1427 |
+
"learning_rate": 6.231355262852529e-07,
|
| 1428 |
+
"logits/chosen": -0.030688485130667686,
|
| 1429 |
+
"logits/rejected": 0.05593588203191757,
|
| 1430 |
+
"logps/chosen": -845.5794067382812,
|
| 1431 |
+
"logps/rejected": -900.7802734375,
|
| 1432 |
+
"loss": 191174.325,
|
| 1433 |
+
"rewards/accuracies": 0.643750011920929,
|
| 1434 |
+
"rewards/chosen": -0.5286746025085449,
|
| 1435 |
+
"rewards/margins": 0.10104093700647354,
|
| 1436 |
+
"rewards/rejected": -0.6297155618667603,
|
| 1437 |
+
"step": 1010
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 0.8,
|
| 1441 |
+
"learning_rate": 5.785352205971275e-07,
|
| 1442 |
+
"logits/chosen": -0.03481902927160263,
|
| 1443 |
+
"logits/rejected": 0.09135408699512482,
|
| 1444 |
+
"logps/chosen": -841.7257080078125,
|
| 1445 |
+
"logps/rejected": -931.0070190429688,
|
| 1446 |
+
"loss": 200004.7,
|
| 1447 |
+
"rewards/accuracies": 0.6312500238418579,
|
| 1448 |
+
"rewards/chosen": -0.550171971321106,
|
| 1449 |
+
"rewards/margins": 0.0866222232580185,
|
| 1450 |
+
"rewards/rejected": -0.6367942094802856,
|
| 1451 |
+
"step": 1020
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 0.81,
|
| 1455 |
+
"learning_rate": 5.353813333653287e-07,
|
| 1456 |
+
"logits/chosen": 0.016893912106752396,
|
| 1457 |
+
"logits/rejected": 0.0011436156928539276,
|
| 1458 |
+
"logps/chosen": -758.1476440429688,
|
| 1459 |
+
"logps/rejected": -817.4668579101562,
|
| 1460 |
+
"loss": 212142.725,
|
| 1461 |
+
"rewards/accuracies": 0.574999988079071,
|
| 1462 |
+
"rewards/chosen": -0.4770483374595642,
|
| 1463 |
+
"rewards/margins": 0.07137372344732285,
|
| 1464 |
+
"rewards/rejected": -0.54842209815979,
|
| 1465 |
+
"step": 1030
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 0.82,
|
| 1469 |
+
"learning_rate": 4.937063494834774e-07,
|
| 1470 |
+
"logits/chosen": -0.014474679715931416,
|
| 1471 |
+
"logits/rejected": -0.032013148069381714,
|
| 1472 |
+
"logps/chosen": -804.8319091796875,
|
| 1473 |
+
"logps/rejected": -824.1060791015625,
|
| 1474 |
+
"loss": 212177.825,
|
| 1475 |
+
"rewards/accuracies": 0.5625,
|
| 1476 |
+
"rewards/chosen": -0.49243831634521484,
|
| 1477 |
+
"rewards/margins": 0.06834886968135834,
|
| 1478 |
+
"rewards/rejected": -0.5607872605323792,
|
| 1479 |
+
"step": 1040
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 0.82,
|
| 1483 |
+
"learning_rate": 4.5354164057310857e-07,
|
| 1484 |
+
"logits/chosen": -0.0435628667473793,
|
| 1485 |
+
"logits/rejected": 0.027259284630417824,
|
| 1486 |
+
"logps/chosen": -789.1632080078125,
|
| 1487 |
+
"logps/rejected": -851.56298828125,
|
| 1488 |
+
"loss": 190594.05,
|
| 1489 |
+
"rewards/accuracies": 0.637499988079071,
|
| 1490 |
+
"rewards/chosen": -0.49468034505844116,
|
| 1491 |
+
"rewards/margins": 0.09655535966157913,
|
| 1492 |
+
"rewards/rejected": -0.5912356972694397,
|
| 1493 |
+
"step": 1050
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 0.83,
|
| 1497 |
+
"learning_rate": 4.1491744136810066e-07,
|
| 1498 |
+
"logits/chosen": -0.04903930425643921,
|
| 1499 |
+
"logits/rejected": 0.004996160510927439,
|
| 1500 |
+
"logps/chosen": -817.1986083984375,
|
| 1501 |
+
"logps/rejected": -844.3206787109375,
|
| 1502 |
+
"loss": 206357.5875,
|
| 1503 |
+
"rewards/accuracies": 0.625,
|
| 1504 |
+
"rewards/chosen": -0.5099014043807983,
|
| 1505 |
+
"rewards/margins": 0.07538522779941559,
|
| 1506 |
+
"rewards/rejected": -0.5852866768836975,
|
| 1507 |
+
"step": 1060
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 0.84,
|
| 1511 |
+
"learning_rate": 3.7786282695491313e-07,
|
| 1512 |
+
"logits/chosen": -0.013560289517045021,
|
| 1513 |
+
"logits/rejected": 0.024793395772576332,
|
| 1514 |
+
"logps/chosen": -859.4154052734375,
|
| 1515 |
+
"logps/rejected": -930.1591796875,
|
| 1516 |
+
"loss": 202759.0375,
|
| 1517 |
+
"rewards/accuracies": 0.6312500238418579,
|
| 1518 |
+
"rewards/chosen": -0.5486615300178528,
|
| 1519 |
+
"rewards/margins": 0.08299361169338226,
|
| 1520 |
+
"rewards/rejected": -0.6316550970077515,
|
| 1521 |
+
"step": 1070
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 0.85,
|
| 1525 |
+
"learning_rate": 3.4240569088577564e-07,
|
| 1526 |
+
"logits/chosen": -0.043563805520534515,
|
| 1527 |
+
"logits/rejected": 0.038817066699266434,
|
| 1528 |
+
"logps/chosen": -784.8958129882812,
|
| 1529 |
+
"logps/rejected": -838.8773193359375,
|
| 1530 |
+
"loss": 196943.7125,
|
| 1531 |
+
"rewards/accuracies": 0.637499988079071,
|
| 1532 |
+
"rewards/chosen": -0.4984671175479889,
|
| 1533 |
+
"rewards/margins": 0.0784321129322052,
|
| 1534 |
+
"rewards/rejected": -0.5768992900848389,
|
| 1535 |
+
"step": 1080
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 0.86,
|
| 1539 |
+
"learning_rate": 3.0857272418129136e-07,
|
| 1540 |
+
"logits/chosen": 0.014950004406273365,
|
| 1541 |
+
"logits/rejected": 0.008164005354046822,
|
| 1542 |
+
"logps/chosen": -833.267578125,
|
| 1543 |
+
"logps/rejected": -885.646484375,
|
| 1544 |
+
"loss": 198440.8625,
|
| 1545 |
+
"rewards/accuracies": 0.59375,
|
| 1546 |
+
"rewards/chosen": -0.5129631757736206,
|
| 1547 |
+
"rewards/margins": 0.08385080099105835,
|
| 1548 |
+
"rewards/rejected": -0.596813976764679,
|
| 1549 |
+
"step": 1090
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 0.86,
|
| 1553 |
+
"learning_rate": 2.7638939523827956e-07,
|
| 1554 |
+
"logits/chosen": -0.0004991426831111312,
|
| 1555 |
+
"logits/rejected": 0.01584457978606224,
|
| 1556 |
+
"logps/chosen": -853.5109252929688,
|
| 1557 |
+
"logps/rejected": -916.8821411132812,
|
| 1558 |
+
"loss": 204926.275,
|
| 1559 |
+
"rewards/accuracies": 0.637499988079071,
|
| 1560 |
+
"rewards/chosen": -0.5368956327438354,
|
| 1561 |
+
"rewards/margins": 0.08267486840486526,
|
| 1562 |
+
"rewards/rejected": -0.6195704936981201,
|
| 1563 |
+
"step": 1100
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 0.87,
|
| 1567 |
+
"learning_rate": 2.4587993065795983e-07,
|
| 1568 |
+
"logits/chosen": -0.043995875865221024,
|
| 1569 |
+
"logits/rejected": 0.016014937311410904,
|
| 1570 |
+
"logps/chosen": -809.5734252929688,
|
| 1571 |
+
"logps/rejected": -882.4210815429688,
|
| 1572 |
+
"loss": 201038.4375,
|
| 1573 |
+
"rewards/accuracies": 0.59375,
|
| 1574 |
+
"rewards/chosen": -0.5105189085006714,
|
| 1575 |
+
"rewards/margins": 0.0844583585858345,
|
| 1576 |
+
"rewards/rejected": -0.5949772596359253,
|
| 1577 |
+
"step": 1110
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 0.88,
|
| 1581 |
+
"learning_rate": 2.170672970089291e-07,
|
| 1582 |
+
"logits/chosen": -0.00044642388820648193,
|
| 1583 |
+
"logits/rejected": -0.03587064892053604,
|
| 1584 |
+
"logps/chosen": -791.4276733398438,
|
| 1585 |
+
"logps/rejected": -819.1533813476562,
|
| 1586 |
+
"loss": 219342.475,
|
| 1587 |
+
"rewards/accuracies": 0.5687500238418579,
|
| 1588 |
+
"rewards/chosen": -0.49560171365737915,
|
| 1589 |
+
"rewards/margins": 0.05810080096125603,
|
| 1590 |
+
"rewards/rejected": -0.5537024736404419,
|
| 1591 |
+
"step": 1120
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 0.89,
|
| 1595 |
+
"learning_rate": 1.8997318353864673e-07,
|
| 1596 |
+
"logits/chosen": 0.02901034988462925,
|
| 1597 |
+
"logits/rejected": -0.009537712670862675,
|
| 1598 |
+
"logps/chosen": -746.2117919921875,
|
| 1599 |
+
"logps/rejected": -798.05224609375,
|
| 1600 |
+
"loss": 214356.25,
|
| 1601 |
+
"rewards/accuracies": 0.5562499761581421,
|
| 1602 |
+
"rewards/chosen": -0.48434096574783325,
|
| 1603 |
+
"rewards/margins": 0.07087163627147675,
|
| 1604 |
+
"rewards/rejected": -0.5552126169204712,
|
| 1605 |
+
"step": 1130
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 0.89,
|
| 1609 |
+
"learning_rate": 1.6461798584644944e-07,
|
| 1610 |
+
"logits/chosen": -0.0717676654458046,
|
| 1611 |
+
"logits/rejected": -0.04001053422689438,
|
| 1612 |
+
"logps/chosen": -844.8236083984375,
|
| 1613 |
+
"logps/rejected": -927.1820068359375,
|
| 1614 |
+
"loss": 176394.5875,
|
| 1615 |
+
"rewards/accuracies": 0.668749988079071,
|
| 1616 |
+
"rewards/chosen": -0.5142861008644104,
|
| 1617 |
+
"rewards/margins": 0.11530987918376923,
|
| 1618 |
+
"rewards/rejected": -0.629595935344696,
|
| 1619 |
+
"step": 1140
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 0.9,
|
| 1623 |
+
"learning_rate": 1.4102079053038454e-07,
|
| 1624 |
+
"logits/chosen": -0.013663539662957191,
|
| 1625 |
+
"logits/rejected": -0.04209432005882263,
|
| 1626 |
+
"logps/chosen": -806.2244262695312,
|
| 1627 |
+
"logps/rejected": -875.3914184570312,
|
| 1628 |
+
"loss": 205334.5,
|
| 1629 |
+
"rewards/accuracies": 0.625,
|
| 1630 |
+
"rewards/chosen": -0.5035674571990967,
|
| 1631 |
+
"rewards/margins": 0.08000927418470383,
|
| 1632 |
+
"rewards/rejected": -0.5835766792297363,
|
| 1633 |
+
"step": 1150
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 0.91,
|
| 1637 |
+
"learning_rate": 1.1919936081941585e-07,
|
| 1638 |
+
"logits/chosen": -0.004432595334947109,
|
| 1639 |
+
"logits/rejected": 0.02214151620864868,
|
| 1640 |
+
"logps/chosen": -842.6566162109375,
|
| 1641 |
+
"logps/rejected": -888.5098876953125,
|
| 1642 |
+
"loss": 193055.6,
|
| 1643 |
+
"rewards/accuracies": 0.6187499761581421,
|
| 1644 |
+
"rewards/chosen": -0.5068452954292297,
|
| 1645 |
+
"rewards/margins": 0.09299594908952713,
|
| 1646 |
+
"rewards/rejected": -0.5998412370681763,
|
| 1647 |
+
"step": 1160
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 0.92,
|
| 1651 |
+
"learning_rate": 9.917012320182245e-08,
|
| 1652 |
+
"logits/chosen": -0.05645931884646416,
|
| 1653 |
+
"logits/rejected": 0.1126350611448288,
|
| 1654 |
+
"logps/chosen": -817.47509765625,
|
| 1655 |
+
"logps/rejected": -878.1007690429688,
|
| 1656 |
+
"loss": 193461.9875,
|
| 1657 |
+
"rewards/accuracies": 0.6187499761581421,
|
| 1658 |
+
"rewards/chosen": -0.4907683730125427,
|
| 1659 |
+
"rewards/margins": 0.08928798139095306,
|
| 1660 |
+
"rewards/rejected": -0.5800563097000122,
|
| 1661 |
+
"step": 1170
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 0.93,
|
| 1665 |
+
"learning_rate": 8.094815505985315e-08,
|
| 1666 |
+
"logits/chosen": -0.03593028336763382,
|
| 1667 |
+
"logits/rejected": 0.005944651551544666,
|
| 1668 |
+
"logps/chosen": -824.0018310546875,
|
| 1669 |
+
"logps/rejected": -869.1282958984375,
|
| 1670 |
+
"loss": 213653.2,
|
| 1671 |
+
"rewards/accuracies": 0.550000011920929,
|
| 1672 |
+
"rewards/chosen": -0.5079488754272461,
|
| 1673 |
+
"rewards/margins": 0.07349041849374771,
|
| 1674 |
+
"rewards/rejected": -0.5814392566680908,
|
| 1675 |
+
"step": 1180
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 0.93,
|
| 1679 |
+
"learning_rate": 6.454717331994542e-08,
|
| 1680 |
+
"logits/chosen": -0.020399611443281174,
|
| 1681 |
+
"logits/rejected": 0.020858485251665115,
|
| 1682 |
+
"logps/chosen": -794.6226806640625,
|
| 1683 |
+
"logps/rejected": -853.1974487304688,
|
| 1684 |
+
"loss": 201205.1125,
|
| 1685 |
+
"rewards/accuracies": 0.606249988079071,
|
| 1686 |
+
"rewards/chosen": -0.48565536737442017,
|
| 1687 |
+
"rewards/margins": 0.08772826939821243,
|
| 1688 |
+
"rewards/rejected": -0.5733836889266968,
|
| 1689 |
+
"step": 1190
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 0.94,
|
| 1693 |
+
"learning_rate": 4.9979524127052595e-08,
|
| 1694 |
+
"logits/chosen": -0.017864791676402092,
|
| 1695 |
+
"logits/rejected": 0.025667399168014526,
|
| 1696 |
+
"logps/chosen": -763.6353759765625,
|
| 1697 |
+
"logps/rejected": -851.0655517578125,
|
| 1698 |
+
"loss": 201221.4875,
|
| 1699 |
+
"rewards/accuracies": 0.6187499761581421,
|
| 1700 |
+
"rewards/chosen": -0.4817201495170593,
|
| 1701 |
+
"rewards/margins": 0.08347930014133453,
|
| 1702 |
+
"rewards/rejected": -0.5651993751525879,
|
| 1703 |
+
"step": 1200
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 0.95,
|
| 1707 |
+
"learning_rate": 3.725617355085476e-08,
|
| 1708 |
+
"logits/chosen": -0.053539175540208817,
|
| 1709 |
+
"logits/rejected": -0.005616022739559412,
|
| 1710 |
+
"logps/chosen": -837.1541137695312,
|
| 1711 |
+
"logps/rejected": -917.78955078125,
|
| 1712 |
+
"loss": 216221.275,
|
| 1713 |
+
"rewards/accuracies": 0.59375,
|
| 1714 |
+
"rewards/chosen": -0.513336181640625,
|
| 1715 |
+
"rewards/margins": 0.06921021640300751,
|
| 1716 |
+
"rewards/rejected": -0.5825464129447937,
|
| 1717 |
+
"step": 1210
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 0.96,
|
| 1721 |
+
"learning_rate": 2.63866993308437e-08,
|
| 1722 |
+
"logits/chosen": -0.04671555012464523,
|
| 1723 |
+
"logits/rejected": 0.012625640258193016,
|
| 1724 |
+
"logps/chosen": -850.5285034179688,
|
| 1725 |
+
"logps/rejected": -898.7066650390625,
|
| 1726 |
+
"loss": 203778.7,
|
| 1727 |
+
"rewards/accuracies": 0.606249988079071,
|
| 1728 |
+
"rewards/chosen": -0.5049459934234619,
|
| 1729 |
+
"rewards/margins": 0.08938591182231903,
|
| 1730 |
+
"rewards/rejected": -0.5943318605422974,
|
| 1731 |
+
"step": 1220
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 0.97,
|
| 1735 |
+
"learning_rate": 1.737928366650099e-08,
|
| 1736 |
+
"logits/chosen": -0.03952642157673836,
|
| 1737 |
+
"logits/rejected": -0.024786999449133873,
|
| 1738 |
+
"logps/chosen": -845.3889770507812,
|
| 1739 |
+
"logps/rejected": -912.8518676757812,
|
| 1740 |
+
"loss": 203856.5625,
|
| 1741 |
+
"rewards/accuracies": 0.612500011920929,
|
| 1742 |
+
"rewards/chosen": -0.5316712856292725,
|
| 1743 |
+
"rewards/margins": 0.08585255593061447,
|
| 1744 |
+
"rewards/rejected": -0.6175239086151123,
|
| 1745 |
+
"step": 1230
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 0.97,
|
| 1749 |
+
"learning_rate": 1.0240707057995735e-08,
|
| 1750 |
+
"logits/chosen": -0.009350773878395557,
|
| 1751 |
+
"logits/rejected": 0.0019861645996570587,
|
| 1752 |
+
"logps/chosen": -840.8093872070312,
|
| 1753 |
+
"logps/rejected": -873.7894287109375,
|
| 1754 |
+
"loss": 219184.125,
|
| 1755 |
+
"rewards/accuracies": 0.543749988079071,
|
| 1756 |
+
"rewards/chosen": -0.5165701508522034,
|
| 1757 |
+
"rewards/margins": 0.07291480898857117,
|
| 1758 |
+
"rewards/rejected": -0.5894849896430969,
|
| 1759 |
+
"step": 1240
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 0.98,
|
| 1763 |
+
"learning_rate": 4.976343202034717e-09,
|
| 1764 |
+
"logits/chosen": -0.04459824040532112,
|
| 1765 |
+
"logits/rejected": -0.026420336216688156,
|
| 1766 |
+
"logps/chosen": -822.0613403320312,
|
| 1767 |
+
"logps/rejected": -859.7013549804688,
|
| 1768 |
+
"loss": 218496.775,
|
| 1769 |
+
"rewards/accuracies": 0.59375,
|
| 1770 |
+
"rewards/chosen": -0.5172399878501892,
|
| 1771 |
+
"rewards/margins": 0.0695352554321289,
|
| 1772 |
+
"rewards/rejected": -0.5867753028869629,
|
| 1773 |
+
"step": 1250
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 0.99,
|
| 1777 |
+
"learning_rate": 1.5901549467139953e-09,
|
| 1778 |
+
"logits/chosen": 0.020182963460683823,
|
| 1779 |
+
"logits/rejected": 0.03361889719963074,
|
| 1780 |
+
"logps/chosen": -862.8426513671875,
|
| 1781 |
+
"logps/rejected": -934.9107666015625,
|
| 1782 |
+
"loss": 214655.3,
|
| 1783 |
+
"rewards/accuracies": 0.59375,
|
| 1784 |
+
"rewards/chosen": -0.5248240828514099,
|
| 1785 |
+
"rewards/margins": 0.06909482926130295,
|
| 1786 |
+
"rewards/rejected": -0.5939189791679382,
|
| 1787 |
+
"step": 1260
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 1.0,
|
| 1791 |
+
"learning_rate": 8.469130840960127e-11,
|
| 1792 |
+
"logits/chosen": -0.02860623598098755,
|
| 1793 |
+
"logits/rejected": -0.045795194804668427,
|
| 1794 |
+
"logps/chosen": -831.4700317382812,
|
| 1795 |
+
"logps/rejected": -876.29296875,
|
| 1796 |
+
"loss": 191498.6375,
|
| 1797 |
+
"rewards/accuracies": 0.675000011920929,
|
| 1798 |
+
"rewards/chosen": -0.5061026215553284,
|
| 1799 |
+
"rewards/margins": 0.09500899165868759,
|
| 1800 |
+
"rewards/rejected": -0.601111650466919,
|
| 1801 |
+
"step": 1270
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 1.0,
|
| 1805 |
+
"step": 1273,
|
| 1806 |
+
"total_flos": 0.0,
|
| 1807 |
+
"train_loss": 214508.74675962294,
|
| 1808 |
+
"train_runtime": 12229.8763,
|
| 1809 |
+
"train_samples_per_second": 1.666,
|
| 1810 |
+
"train_steps_per_second": 0.104
|
| 1811 |
+
}
|
| 1812 |
+
],
|
| 1813 |
+
"logging_steps": 10,
|
| 1814 |
+
"max_steps": 1273,
|
| 1815 |
+
"num_input_tokens_seen": 0,
|
| 1816 |
+
"num_train_epochs": 1,
|
| 1817 |
+
"save_steps": 100,
|
| 1818 |
+
"total_flos": 0.0,
|
| 1819 |
+
"train_batch_size": 4,
|
| 1820 |
+
"trial_name": null,
|
| 1821 |
+
"trial_params": null
|
| 1822 |
+
}
|