Files
deepseek-prover-v2-cpt-sft-1e/train_results.json
ModelHub XC aa13e379ad 初始化项目,由ModelHub XC社区提供模型
Model: formalmathatepfl/deepseek-prover-v2-cpt-sft-1e
Source: Original Platform
2026-06-11 02:52:18 +08:00

8 lines
209 B
JSON

{
"epoch": 1.0,
"total_flos": 9.549834298038052e+19,
"train_loss": 0.10822304976860733,
"train_runtime": 69525.0456,
"train_samples_per_second": 1.076,
"train_steps_per_second": 0.067
}