Several trained models to compare the differences between each method. Each model has a complete description of hyperparams with wandb reports.
G
G-reen
AI & ML interests
Currently working on https://github.com/Green0-0/propagate (ES Trainer for LLMs). Also interested in TPU training, and building better synthetic datasets.
Organizations
None yet
models 22
G-reen/gemma-2-2b-it-fft-simpo-tpu
Updated
• 1
G-reen/gemma-2-2b-simpo-orbax
Updated
G-reen/gemma-2-2b-it-fft-3epoch-simpo-adj-cpo100
Text Generation • 3B • Updated
• 1
G-reen/gemma-2-2b-it-fft-3epoch-simpo-adj
Text Generation • 3B • Updated
• 3
G-reen/gemma-2-2b-it-fft-3epoch-simpo
Text Generation • 3B • Updated
G-reen/gemma-2-2b-it-fft-3epoch
Text Generation • 3B • Updated
• 2
G-reen/gemma-2-2b-fft-orbax
Updated
G-reen/gemma-2-2b-it-fft-lowlr
Text Generation • 3B • Updated
• 1
G-reen/gemma-2-2b-it-fft-simpo-adj
Text Generation • 3B • Updated
• 3
G-reen/gemma-2-2b-it-fft-simpo-unsloth
Text Generation • 3B • Updated
• 1
datasets 27
G-reen/instruct-set-longer-fixed
Viewer
• Updated
• 110k • 194
G-reen/instruct-set-longer
Viewer
• Updated
• 140k • 177
G-reen/instruct-set
Viewer
• Updated
• 139k • 215
G-reen/instruct-set-half
Viewer
• Updated
• 95.1k • 373
G-reen/visualizertest
Viewer
• Updated
• 1k • 8
G-reen/medium_set
Viewer
• Updated
• 138k • 115
G-reen/sumthink_broken_format
Viewer
• Updated
• 24.7k • 8
G-reen/sumthink_responses_summarized
Viewer
• Updated
• 24.7k • 15
G-reen/sumthink_responses_raw
Viewer
• Updated
• 27k • 12
G-reen/sumthink_fixed_cleaned
Viewer
• Updated
• 24.7k • 8