初始化项目,由ModelHub XC社区提供模型

Model: open-sci/sft__ot30k_Qwen3-1.7B-Base-DPO-Tulu3-decontaminated
Source: Original Platform
This commit is contained in:
ModelHub XC
2026-04-26 10:33:11 +08:00
commit af4662d21a
15 changed files with 14404 additions and 0 deletions

12
train_results.json Normal file
View File

@@ -0,0 +1,12 @@
{
"epoch": 5.0,
"loss_nan_ranks": 0,
"loss_rank_avg": 0.4691266119480133,
"total_flos": 1658037051588608.0,
"train_loss": 1.0214665260213487,
"train_runtime": 6089.571,
"train_samples_per_second": 24.632,
"train_steps_per_second": 0.193,
"valid_targets_mean": 13499.5,
"valid_targets_min": 3345
}