tuhink/hacking-rewards-coherence-train

Tokenizer for the final model

Final trained model

Training in progress, step 318

Training in progress, step 300

Training in progress, step 200

Training in progress, step 100

initial commit