Hugging Face's logo Hugging Face
  • Models
  • Datasets
  • Spaces
  • Buckets
  • Docs
  • Enterprise
  • Pricing
    • Website
      • Tasks
      • HuggingChat
      • Collections
      • Languages
      • Organizations
    • Community
      • Blog
      • Posts
      • Daily Papers
      • Hardware
      • Learn
      • Discord
      • Forum
      • GitHub
    • Solutions
      • Team & Enterprise
      • Hugging Face PRO
      • Enterprise Support
      • Inference Providers
      • Inference Endpoints
      • Storage Buckets

  • Log In
  • Sign Up

ShivankXD
/
stasis-1

Text Generation
PyTorch
Safetensors
English
stasis_decoder
base-model
research
small-language-model
from-scratch
pretraining
efficient-training
llama-style
rope
swiglu
Eval Results (legacy)
Model card Files Files and versions
xet
Community
stasis-1 / eval
991 kB
Ctrl+K
Ctrl+K
  • 1 contributor
History: 2 commits
ShivankXD's picture
ShivankXD
Stasis 1 step 36,420 (run 10): weights, endpoint, code, configs, logs, eval (8/8 gates), GPT-2 matched eval at step 36,420 (no parity), charts
ff11bfe verified 9 days ago
  • run10_step36420
    Stasis 1 step 36,420 (run 10): weights, endpoint, code, configs, logs, eval (8/8 gates), GPT-2 matched eval at step 36,420 (no parity), charts 9 days ago
  • run6_step12269
    Stasis 1 step 23,612 (run 9): weights, model-only endpoints for steps 12269/15106/16112/23612, code, notebooks, configs, logs, eval results; model card + repo layout 9 days ago
  • run7_step15106
    Stasis 1 step 23,612 (run 9): weights, model-only endpoints for steps 12269/15106/16112/23612, code, notebooks, configs, logs, eval results; model card + repo layout 9 days ago
  • run8_step16112
    Stasis 1 step 23,612 (run 9): weights, model-only endpoints for steps 12269/15106/16112/23612, code, notebooks, configs, logs, eval results; model card + repo layout 9 days ago
  • run9_step23612
    Stasis 1 step 23,612 (run 9): weights, model-only endpoints for steps 12269/15106/16112/23612, code, notebooks, configs, logs, eval results; model card + repo layout 9 days ago