Hugging Face's logo

MTSAIR
/
Cotype-Nano

Text Generation
Transformers
Safetensors
Russian
English
qwen2
conversational
text-generation-inference
Inference Endpoints
Model card Files Community
2
AlanRobotics commited on
Commit
9cf5a4e
·
verified ·
1 Parent(s): 9b5dcef

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +2 -2
README.md CHANGED
@@ -139,12 +139,12 @@ The model was trained in two stages. In the first stage, MLP layers were trained
139
 
140
  | Model | Score | 95% CI | Avg. #Tokens |
141
  |---------------------------------------------|-------|-------------------------|---------------|
142
- | **Cotype-Nano** | **30.2** | **+2.2 / -1.3** | **567** |
143
  | vikhr-it-5.3-fp16-32k | 27.8 | +1.5 / -2.1 | 519.71 |
144
  | vikhr-it-5.3-fp16 | 22.73 | +1.8 / -1.7 | 523.45 |
145
  | kolibri-vikhr-mistral-0427 | 22.41 | +1.6 / -1.9 | 489.89 |
146
  | snorkel-mistral-pairrm-dpo | 22.41 | +1.7 / -1.6 | 773.8 |
147
- | **Cotype-Nano-4bit** | **21.6** | **+2.1 / -1.8** | **587** |
148
  | storm-7b | 20.62 | +1.4 / -1.6 | 419.32 |
149
  | neural-chat-7b-v3-3 | 19.04 | +1.8 / -1.5 | 927.21 |
150
  | Vikhrmodels-Vikhr-Llama-3.2-1B-instruct | 19.04 | +1.2 / -1.5 | 958.63 |
 
139
 
140
  | Model | Score | 95% CI | Avg. #Tokens |
141
  |---------------------------------------------|-------|-------------------------|---------------|
142
+ | **Cotype-Nano** | **30.2** | **+2.2 / -1.3** | **542** |
143
  | vikhr-it-5.3-fp16-32k | 27.8 | +1.5 / -2.1 | 519.71 |
144
  | vikhr-it-5.3-fp16 | 22.73 | +1.8 / -1.7 | 523.45 |
145
  | kolibri-vikhr-mistral-0427 | 22.41 | +1.6 / -1.9 | 489.89 |
146
  | snorkel-mistral-pairrm-dpo | 22.41 | +1.7 / -1.6 | 773.8 |
147
+ | **Cotype-Nano-4bit** | **22.5** | **+2.1 / -1.4** | **582** |
148
  | storm-7b | 20.62 | +1.4 / -1.6 | 419.32 |
149
  | neural-chat-7b-v3-3 | 19.04 | +1.8 / -1.5 | 927.21 |
150
  | Vikhrmodels-Vikhr-Llama-3.2-1B-instruct | 19.04 | +1.2 / -1.5 | 958.63 |