初始化项目,由ModelHub XC社区提供模型

Model: modrill/qwen3-4b-think-baseline-full-sft
Source: Original Platform
This commit is contained in:
ModelHub XC
2026-07-26 22:34:10 +08:00
commit 84ba2b4166
11 changed files with 493 additions and 0 deletions

12
MANIFEST.txt Normal file
View File

@@ -0,0 +1,12 @@
name: think_baseline
base_model: Qwen3-4B-Base
method: full SFT (ZeRO-3)
dataset: think_all
template: qwen3
epochs: 2
seed: 42
cutoff_len: 24576
learning_rate: 2e-5
train_loss: 0.6698
train_steps: 838
finished_at: 2026-06-08 00:22 CST