add PDF report generator

refresh model card: E0+E1 complete, held-out eval results

add held-out eval results (base/E0/E1)

add report.html

add NOTES.md

add REPORT.pdf

Add outputs/E1-div-individual/trl_log_history.json

Add outputs/E1-div-individual/reward_history.json

Add outputs/E1-div-individual/judge_cost.json

Add outputs/E1-div-individual/README.md

Add outputs/E1-div-individual/final

Add outputs/E1-div-individual/checkpoint-300

Add outputs/E1-div-individual/checkpoint-250

Add outputs/E1-div-individual/checkpoint-200

Add outputs/E1-div-individual/checkpoint-150

Add outputs/ckpt_study

Add logs/figures

Add logs/experiments

Add outputs/E1-div-individual/checkpoint-100

Add outputs/E1-div-individual/checkpoint-50

Add outputs/E0-baseline

Add outputs/pool_4b

Add outputs/pairs_4b

Add prior_run/train.py (API key redacted)

Add prior_run

Add outputs/ckpt_study

Add logs/figures

Add logs/experiments

Add src

Add data

Add configs

Add model card

initial commit