Data cleaning, custom tokenizer, 167-run speedrun, scaling probes, a 2.67B pretrained on one B200 in 38h, then SFT and reasoning. ~300 W&B runs.
Julián Núñez PRO
juliannunezb
AI & ML interests
None yet
Recent Activity
updated a dataset 15 days ago
juliannunezb/crisp-v8 updated a collection 15 days ago
LLM training from scratch: 90M → 2.7B on one GPU updated a collection 15 days ago
LLM training from scratch: 90M → 2.7B on one GPUOrganizations
None yet
models 18
juliannunezb/crisp3b-v8-ckpts
Text Generation • Updated
juliannunezb/crisp3b-v7-ckpts
Text Generation • Updated
juliannunezb/crisp3b-sft-ckpts
Updated
juliannunezb/crisp3b-reason-ckpts
Updated
juliannunezb/llm-500m-sft-v1
Updated
juliannunezb/llm-500m-pretrain-v1
Updated
juliannunezb/llama-3b-checkpoints
Updated
juliannunezb/local50m-combo3-grpo-bootstrap-v1
Updated
juliannunezb/local50m-combo3-dpo-v1
Updated
juliannunezb/local50m-combo3-sft-v1
Updated
datasets 12
juliannunezb/crisp-v8
Updated • 14
juliannunezb/crisp-reason-cold
Viewer • Updated • 35k • 14
juliannunezb/crisp-sft-chat
Viewer • Updated • 330k • 15
juliannunezb/crisp-165b
Updated • 208
juliannunezb/crisp-128b
Updated • 19
juliannunezb/mixed-pretrain-100b-gpt2
Updated • 87
juliannunezb/midtrain-mix-5b-gpt2
Updated • 30
juliannunezb/ultrafeedback-gpt2-dpo
Viewer • Updated • 62k • 22
juliannunezb/smoltalk-gpt2-sft
Viewer • Updated • 1.04M • 68
juliannunezb/mixed-pretrain-10b-gpt2
Viewer • Updated • 1 • 72 • 1