anime-sh/llama-3_1-8b-rmu-bm25-6t-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-rmu-bm25-6t-rebuttal
Source: Original Platform
This commit is contained in:
ModelHub XC
2026-09-28 01:47:17 +08:00
commit 0825b34250
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 01:20:59,474 INFO cleanslate.fsid.training.trainer: [RANK 0] FSID train: model=meta-llama/Llama-3.1-8B objective=rmu run_dir=results/rebuttal/train/llama-3_1-8b-rmu-bm25-6t/rmu_Llama-3.1-8B_20260724_012047
2026-07-24 01:20:59,474 INFO cleanslate.fsid.training.trainer: [RANK 0] Creating frozen reference model for rmu (dtype=torch.bfloat16)
2026-07-24 01:21:05,708 INFO cleanslate.fsid.training.trainer: [RANK 0] Starting training: objective=rmu epochs=2
2026-07-24 01:21:06,238 INFO cleanslate.fsid.training.trainer: [RANK 0] RMU control_vec init: dim=4096 seed=0 rank=0 world_size=1 per_rank_norm=2.000000
2026-07-24 01:24:35,830 INFO cleanslate.fsid.training.trainer: [RANK 0] Saved checkpoint to results/rebuttal/train/llama-3_1-8b-rmu-bm25-6t/rmu_Llama-3.1-8B_20260724_012047
2026-07-24 01:24:35,834 INFO cleanslate.fsid.training.trainer: [RANK 0] Training complete (170.3s)