anime-sh/llama-3_1-8b-undial-bm25-10b-rebuttal
1
0
Fork 0
Code Issues Pull Requests Actions Projects Releases Wiki Activity

初始化项目，由ModelHub XC社区提供模型

Browse Source 
Model: anime-sh/llama-3_1-8b-undial-bm25-10b-rebuttal
Source: Original Platform
This commit is contained in:
ModelHub XC
2026-07-29 11:52:18 +08:00
commit 30c91266d4
10 changed files with 197 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
train.log Normal file
View File

@@ -0,0 +1,6 @@
2026-07-24 00:49:09,471 INFO cleanslate.fsid.training.trainer: [RANK 0] FSID train: model=meta-llama/Llama-3.1-8B objective=undial run_dir=results/rebuttal/train/llama-3_1-8b-undial-bm25-10b/undial_Llama-3.1-8B_20260724_004857
2026-07-24 00:49:09,471 INFO cleanslate.fsid.training.trainer: [RANK 0] Creating frozen reference model for undial (dtype=torch.bfloat16)
2026-07-24 00:49:14,924 INFO cleanslate.fsid.training.trainer: [RANK 0] Starting training: objective=undial epochs=2
2026-07-24 01:02:31,997 INFO cleanslate.fsid.training.trainer: [RANK 0] Saved checkpoint to results/rebuttal/train/llama-3_1-8b-undial-bm25-10b/undial_Llama-3.1-8B_20260724_004857
2026-07-24 01:02:32,003 INFO cleanslate.fsid.training.trainer: [RANK 0] Training complete (755.7s)
2026-07-24 01:02:32,518 INFO cleanslate.fsid.training.trainer: [RANK 0] Pushing checkpoint dir results/rebuttal/train/llama-3_1-8b-undial-bm25-10b/undial_Llama-3.1-8B_20260724_004857 to unlearning-cleanslate/llama-3_1-8b-undial-bm25-10b