路易 yi_lu
0 Followers · 0 Following
  • Joined on 2026-05-01
Repositories
6
 Projects Packages Public Activity Starred Repositories
yi_lu pushed to main at yi_lu/laoyao_2b_moe 2026-07-16 09:25:57 +08:00
1329bb9506 Support stage1-2 MoE resume on g0050
yi_lu pushed to main at yi_lu/laoyao_2b_moe 2026-07-09 08:39:05 +08:00
07e108388d Add heldout HF evaluation workflow
yi_lu pushed to main at yi_lu/laoyao_2b_moe 2026-07-09 08:16:39 +08:00
e1e880cadf Add agent operating guide
yi_lu pushed to main at yi_lu/laoyao_2b_moe 2026-07-09 08:04:55 +08:00
2b0588ad17 Sync g0050 training and inference tooling
yi_lu pushed to main at yi_lu/laoyao_2b_moe 2026-07-02 21:59:04 +08:00
0e8c2a554d Support g0050 SSD preprocessing path
yi_lu pushed to main at yi_lu/laoyao_2b_moe 2026-07-02 21:29:13 +08:00
b8b5c3ffe0 Use vLLM triton kernels for MoE backend
yi_lu pushed to main at yi_lu/laoyao_2b_moe 2026-07-02 21:11:01 +08:00
323452b9bd Add g0050 NeMo flash-attn4 training image
yi_lu pushed to main at yi_lu/laoyao_2b_moe 2026-07-02 20:55:24 +08:00
5698a9cc21 Add ModelScope parquet upload helper
yi_lu pushed to main at yi_lu/laoyao_2b_moe 2026-07-02 20:50:59 +08:00
5609b1f8e4 Add Megatron data manifest and g0050 setup
yi_lu pushed to main at yi_lu/laoyao_2b_moe 2026-07-01 23:27:59 +08:00
816eccb5b5 Tie embeddings for small MoE config
yi_lu pushed to main at yi_lu/laoyao_2b_moe 2026-07-01 23:10:40 +08:00
2c2b7ccc24 Add Megatron-Bridge pretrain launcher
yi_lu pushed to main at yi_lu/laoyao_2b_moe 2026-07-01 22:42:53 +08:00
f0f72767f5 Add proof-pile logic inputs to pretrain builder
yi_lu pushed to main at yi_lu/laoyao_2b_moe 2026-07-01 20:46:47 +08:00
cdd8e2f85b Add NeMo backend probe and architecture validator
yi_lu pushed to main at yi_lu/laoyao_2b_moe 2026-07-01 19:48:48 +08:00
11308f95c4 Use verified GLM-5.2 tokenizer from ModelScope
e7ebb30e1f Use GLM-5.2 tokenizer vocabulary
dcabce8c07 Initialize rebalanced 2B MoE project
yi_lu created branch main in yi_lu/laoyao_2b_moe 2026-07-01 19:48:48 +08:00
yi_lu created repository yi_lu/laoyao_2b_moe 2026-07-01 19:33:19 +08:00
yi_lu pushed to main at yi_lu/Tokenizer_Swap 2026-07-01 15:33:08 +08:00
9b584742c4 Support Laoyao BPE tokenizer in SFT heldout builder
yi_lu pushed to main at yi_lu/Tokenizer_Swap 2026-07-01 15:24:08 +08:00
069f891ceb Add SFT heldout 2.8K build entrypoint
yi_lu pushed to main at yi_lu/ti_coding_agent_probe 2026-06-25 20:35:00 +08:00
d2cafeb8a0 Add eight GPU SWIFT and Megatron training scripts
yi_lu pushed to main at yi_lu/ti_coding_agent_probe 2026-06-25 15:34:24 +08:00
4f9e0a30a9 Update Hugging Face model download CLI