Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Language Model Pre-training on C4 (Eval Loss)
Loading...
3.86
Eval Loss
GrassWalk
3.7676
4.3913
5.015
5.6387
Oct 2, 2025
Eval Loss
Updated 25d ago
Evaluation Results
Method
Method
Links
Eval Loss
GrassWalk
Arch.=LLaMA-1B, Peak M...
2025.10
3.86
GrassJump
Arch.=LLaMA-1B, Peak M...
2025.10
3.87
SubTrack++
Arch.=LLaMA-1B, Peak M...
2025.10
3.89
AdamW [Full-Rank]
Arch.=LLaMA-1B, Peak M...
2025.10
4.1
LDAdam
Arch.=LLaMA-1B, Peak M...
2025.10
4.1
FRUGAL
Arch.=LLaMA-1B, Peak M...
2025.10
4.22
GrassJump
Arch.=Qwen-1.5B, Peak...
2025.10
4.67
GrassWalk
Arch.=Qwen-1.5B, Peak...
2025.10
4.68
SubTrack++
Arch.=Qwen-1.5B, Peak...
2025.10
4.7
AdamW [Full-Rank]
Arch.=Qwen-1.5B, Peak...
2025.10
4.84
APOLLO
Arch.=LLaMA-1B, Peak M...
2025.10
5.71
GaLore
Arch.=LLaMA-1B, Peak M...
2025.10
6.17
Feedback
Search any
task
Search any
task