Hugging Face
Models
Datasets
Spaces
Posts
Docs
Solutions
Pricing
Log In
Sign Up
DUAL-GPO
/
phi-2-gpo-v18-i1
like
0
Follow
DUAL Group
2
PEFT
TensorBoard
Safetensors
HuggingFaceH4/ultrafeedback_binarized
phi
alignment-handbook
Generated from Trainer
trl
dpo
custom_code
License:
apache-2.0
Model card
Files
Files and versions
Metrics
Training metrics
Community
Use this model
8214ff1
phi-2-gpo-v18-i1
/
all_results.json
lole25
Model save
8214ff1
verified
6 months ago
raw
Copy download link
history
blame
Safe
196 Bytes
{
"epoch"
:
1.0
,
"train_loss"
:
0.06840609829860919
,
"train_runtime"
:
2864.3388
,
"train_samples"
:
30000
,
"train_samples_per_second"
:
10.474
,
"train_steps_per_second"
:
0.327
}