tkwiecinski/amr-fma-Apertus-8B-Instruct-2509-lora_sft-primevul_vuln-e1_blockem_repro-s42

tkwiecinski/amr-fma-Apertus-8B-Instruct-2509-lorasft-primevulvuln-e1blockemrepro-s42

amr-fma training run.

• —Method: lora_sft
• —Base model: swiss-ai/Apertus-8B-Instruct-2509
• —Dataset: primevul_vuln (slug: primevul_vuln)
• —Seed: 42
• —Git commit: 5b0cb11405e4df5273e3c71642979e93ca408ced
• —Exp name: e1_blockem_repro
• —WandB run: q691ljgd

Tags

• —phase:P1
• —domain:code

Checkpoints (branches)

• —step 1 → revision step-00001
• —step 2 → revision step-00002
• —step 5 → revision step-00005
• —step 11 → revision step-00011
• —step 26 → revision step-00026
• —step 60 → revision step-00060
• —step 138 → revision step-00138
• —step 314 → revision step-00314
• —step 315 → revision step-00315

Pin a specific checkpoint with revision=... in AutoModelForCausalLM.from_pretrained / PeftModel.from_pretrained.

Hyperparameter sections

checkpointing, dataset, evaluation, final_adapter_path, lora, model, optimization, prompt_style, runtime, sdpo, sequence, total_steps