Models I've created as "extra credit" after finishing Sebastian Raschka's book
"[Build a Large Language Model (from Scratch)"
🔄 In a Training Loop
Giles Thomas
gpjt
AI & ML interests
Doing my best to speedrun 20 years of AI research. YMMV
Recent Activity
updated a dataset about 7 hours ago
gpjt/fw-fwedu-5050-gpt2-tokens published a dataset about 7 hours ago
gpjt/fw-fwedu-5050-gpt2-tokens updated a dataset about 9 hours ago
gpjt/fw-fwedu-simplewiki-gpt2-tokensOrganizations
None yet
models 38
gpjt/1xrtx3090-moe-1
Text Generation • 0.5B • Updated • 248
gpjt/jax-with-mha-bias-larger-chinchilla-1
Text Generation • 0.3B • Updated • 273
gpjt/jax-with-mha-bias-larger-chinchilla-2
Text Generation • 0.2B • Updated • 262
gpjt/jax-with-mha-bias-no-dropout-2-epoch
Text Generation • 0.2B • Updated • 224
gpjt/jax-with-mha-bias-no-dropout-extended
Text Generation • 0.2B • Updated • 244
gpjt/jax-no-mha-bias-no-dropout
Text Generation • 0.2B • Updated • 315
gpjt/jax-with-mha-bias-no-dropout
Text Generation • 0.2B • Updated • 274
gpjt/jax-no-mha-bias-with-dropout
Text Generation • 0.2B • Updated • 247
gpjt/1xrtx3090-stacked-interventions
Text Generation • 0.2B • Updated • 14
gpjt/1xrtx3090-baseline
Text Generation • 0.2B • Updated • 13