Hugging Face
Models
Datasets
Spaces
Posts
Docs
Solutions
Pricing
Log In
Sign Up
DUAL-GPO
/
phi-2-gpo-v6-i1
like
0
PEFT
TensorBoard
Safetensors
HuggingFaceH4/ultrafeedback_binarized
phi
alignment-handbook
Generated from Trainer
trl
dpo
custom_code
License:
mit
Model card
Files
Files and versions
Metrics
Training metrics
Community
Use this model
187f624
phi-2-gpo-v6-i1
/
all_results.json
lole25
Model save
1d92b30
verified
5 months ago
raw
Copy download link
history
blame
No virus
196 Bytes
{
"epoch"
:
2.0
,
"train_loss"
:
0.27172684411589915
,
"train_runtime"
:
11567.6763
,
"train_samples"
:
20000
,
"train_samples_per_second"
:
3.458
,
"train_steps_per_second"
:
0.072
}