Hugging Face
Models
Datasets
Spaces
Posts
Docs
Solutions
Pricing
Log In
Sign Up
nlee-208
/
inverted-orpo-mistral-capybara
like
0
Text Generation
Transformers
TensorBoard
Safetensors
argilla/Capybara-Preferences
mistral
alignment-handbook
trl
orpo
Generated from Trainer
conversational
text-generation-inference
Inference Endpoints
License:
apache-2.0
Model card
Files
Files and versions
Metrics
Training metrics
Community
Train
Deploy
Use this model
4d437ea
inverted-orpo-mistral-capybara
/
train_results.json
nlee-208
Model save
4d437ea
verified
6 months ago
raw
Copy download link
history
blame
Safe
195 Bytes
{
"epoch"
:
3.0
,
"train_loss"
:
0.9363477402682757
,
"train_runtime"
:
28864.8871
,
"train_samples"
:
14154
,
"train_samples_per_second"
:
1.471
,
"train_steps_per_second"
:
0.023
}