Hugging Face
Models
Datasets
Spaces
Posts
Docs
Solutions
Pricing
Log In
Sign Up
imelnyk
/
zephyr-7b-dpo-qlora-fsdp
like
0
PEFT
TensorBoard
Safetensors
HuggingFaceH4/ultrafeedback_binarized
mistral
alignment-handbook
trl
dpo
Generated from Trainer
License:
apache-2.0
Model card
Files
Files and versions
Metrics
Training metrics
Community
Train
Use this model
0375483
zephyr-7b-dpo-qlora-fsdp
/
train_results.json
imelnyk
Model save
0375483
verified
7 months ago
raw
Copy download link
history
blame
No virus
193 Bytes
{
"epoch"
:
0.1
,
"train_loss"
:
0.6879739944751446
,
"train_runtime"
:
738.7034
,
"train_samples"
:
61135
,
"train_samples_per_second"
:
8.276
,
"train_steps_per_second"
:
0.018
}