LLaMAX commited on
Commit
8ca2b80
·
verified ·
1 Parent(s): 1537317

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +2 -2
README.md CHANGED
@@ -115,7 +115,7 @@ language:
115
 
116
  GlotMAX is a language model with powerful multilingual capabilities and strong reasoning.
117
 
118
- Qwen3-GlotMAX series models start from Qwen3-8B instruct models with layer-slective tuning using small amount of parallel data alone.
119
 
120
  Meanwhile, comprehensive testing on 16 reasoning tasks, such as bbeh, Livecodebench, Olymmath and so on, shows that it surpasses existing translation-enhanced models and performs on par with Qwen3 instruct models.
121
 
@@ -123,7 +123,7 @@ Meanwhile, comprehensive testing on 16 reasoning tasks, such as bbeh, Livecodebe
123
 
124
  ### 🔥 Excellent Translation Performance
125
 
126
- LLaMAX3-8B-Alpaca achieves an average spBLEU score improvement of over **5 points** compared to the LLaMA3-8B-Alpaca model on the Flores-101 dataset.
127
 
128
 
129
  ### Supported Languages
 
115
 
116
  GlotMAX is a language model with powerful multilingual capabilities and strong reasoning.
117
 
118
+ GlotMAX series models start from Qwen3-8B instruct models with layer-slective tuning using small amount of parallel data alone.
119
 
120
  Meanwhile, comprehensive testing on 16 reasoning tasks, such as bbeh, Livecodebench, Olymmath and so on, shows that it surpasses existing translation-enhanced models and performs on par with Qwen3 instruct models.
121
 
 
123
 
124
  ### 🔥 Excellent Translation Performance
125
 
126
+ GlotMAX-101-8B achieves an average spBLEU score improvement of over **5 points** compared to the Qwen3-8B model on the Flores-101 dataset.
127
 
128
 
129
  ### Supported Languages