transformer-pretraining-from-scratch

hari-8/transformer-pretraining-from-scratch

Text Generation69.3M parameterstransformersLicense: mit
Tracked by OpenModelStats since September 5, 2026View on Hugging Face ↗
Downloads 30D
647
Total downloads
647
Gained 7D
0
Gained 30D
—
Likes
1
+0 in 7D
Rank · tracked models
#38,254
199(down 199 places)this week
Parameters
69.3M
Last source update
Sep 3, 2026
Momentum
4.4
  • It moved from #10480 to #10564 among tracked Text Generation models this week.

Downloads over time

Daily gain between consecutive observations

Likes over time

Rank among tracked models

Lower is better

Overview

transformer-pretraining-from-scratch is a text generation model published by hari-8. OpenModelStats has tracked the model since Sep 5, 2026, most recently observing it Sep 5, 2026.

Published
Sep 3, 2026
Last updated
Sep 3, 2026
Library
transformers
License
mit
Task
Text Generation
Parameters
69,284,608
Spaces
—
Derivative models
—
Velocity 7D/day
0
transformerssafetensorsdecoder_transformer_scratchtext-generationdecoder-onlyfrom-scratchpytorchcausal-lmcustom_codeenlicense:mitregion:us

Similar models

Models similar to transformer-pretraining-from-scratch
ModelDL 30D
SmolLM2-70Mcodelion508
DiscoverLM-70Mreaperdoesntknow3,043
chesspythia-70mmlabonne1,632
xslmAlienstro424
chess_llama_68mbharathrajcl3,104
vicuna-68mdouble74,382