yiqichen01 commited on
Commit
0e80a64
·
verified ·
1 Parent(s): 2c325e2
Files changed (2) hide show
  1. README.md +1 -2
  2. README_CN.md +1 -2
README.md CHANGED
@@ -137,7 +137,7 @@ Compared with Hy4, Hy4 Flash introduces three notable architectural changes:
137
  Bits-per-byte (BPB) on internal code & agent validation, lower is better:
138
 
139
  <p align="center">
140
- <a href="pretrained-model-bpb.png"><img src="pretrained-model-bpb.png" alt="Bits-per-byte on internal code and agent validation versus active parameters; lower is better" width="100%"/></a>
141
  </p>
142
 
143
  Few-shot pretrain benchmark evaluation:
@@ -163,7 +163,6 @@ Few-shot pretrain benchmark evaluation:
163
  | **Math** | U-MATH | 5 | 69.8 | **71.6** | 60.9 | 57.4 | 64.8 |
164
  | | OCW Courses | 5 | 64.0 | 66.7 | 62.6 | 63.3 | **68.2** |
165
  | **Code** | MBPP-plus | 3 | 73.7 | 74.0 | 77.4 | 76.8 | **79.1** |
166
- | | BigCodeBench (Full) | 0 | 58.8 | 47.9 | **59.6** | 58.0 | 57.5 |
167
  | **Multilingual** | Belebele (5-language subset)† | 5 | 93.8 | 95.0 | 94.2 | 94.4 | **95.2** |
168
  | | INCLUDE† | 5 | 82.8 | **88.1** | 84.3 | 84.0 | 84.3 |
169
  | **Multimodal** | MMMU-Pro (Standard, 10 opts) | 4 | 52.0 | **59.6** | — | — | — |
 
137
  Bits-per-byte (BPB) on internal code & agent validation, lower is better:
138
 
139
  <p align="center">
140
+ <a href="pretrained-model-bpb.png"><img src="pretrained-model-bpb.png" alt="Bits-per-byte on internal code and agent validation versus active parameters; lower is better" width="75%"/></a>
141
  </p>
142
 
143
  Few-shot pretrain benchmark evaluation:
 
163
  | **Math** | U-MATH | 5 | 69.8 | **71.6** | 60.9 | 57.4 | 64.8 |
164
  | | OCW Courses | 5 | 64.0 | 66.7 | 62.6 | 63.3 | **68.2** |
165
  | **Code** | MBPP-plus | 3 | 73.7 | 74.0 | 77.4 | 76.8 | **79.1** |
 
166
  | **Multilingual** | Belebele (5-language subset)† | 5 | 93.8 | 95.0 | 94.2 | 94.4 | **95.2** |
167
  | | INCLUDE† | 5 | 82.8 | **88.1** | 84.3 | 84.0 | 84.3 |
168
  | **Multimodal** | MMMU-Pro (Standard, 10 opts) | 4 | 52.0 | **59.6** | — | — | — |
README_CN.md CHANGED
@@ -121,7 +121,7 @@
121
  内部代码与 Agent 验证集上的 BPB(Bits-per-byte,每字节比特数),数值越低越好:
122
 
123
  <p align="center">
124
- <a href="pretrained-model-bpb.png"><img src="pretrained-model-bpb.png" alt="内部代码与 Agent 验证集上的 BPB 与激活参数量对比,BPB 越低越好" width="100%"/></a>
125
  </p>
126
 
127
  预训练模型的 few-shot 基准评测结果:
@@ -147,7 +147,6 @@
147
  | **数学** | U-MATH | 5 | 69.8 | **71.6** | 60.9 | 57.4 | 64.8 |
148
  | | OCW Courses | 5 | 64.0 | 66.7 | 62.6 | 63.3 | **68.2** |
149
  | **代码** | MBPP-plus | 3 | 73.7 | 74.0 | 77.4 | 76.8 | **79.1** |
150
- | | BigCodeBench (Full) | 0 | 58.8 | 47.9 | **59.6** | 58.0 | 57.5 |
151
  | **多语言** | Belebele (5-language subset)† | 5 | 93.8 | 95.0 | 94.2 | 94.4 | **95.2** |
152
  | | INCLUDE† | 5 | 82.8 | **88.1** | 84.3 | 84.0 | 84.3 |
153
  | **多模态** | MMMU-Pro (Standard, 10 opts) | 4 | 52.0 | **59.6** | — | — | — |
 
121
  内部代码与 Agent 验证集上的 BPB(Bits-per-byte,每字节比特数),数值越低越好:
122
 
123
  <p align="center">
124
+ <a href="pretrained-model-bpb.png"><img src="pretrained-model-bpb.png" alt="内部代码与 Agent 验证集上的 BPB 与激活参数量对比,BPB 越低越好" width="75%"/></a>
125
  </p>
126
 
127
  预训练模型的 few-shot 基准评测结果:
 
147
  | **数学** | U-MATH | 5 | 69.8 | **71.6** | 60.9 | 57.4 | 64.8 |
148
  | | OCW Courses | 5 | 64.0 | 66.7 | 62.6 | 63.3 | **68.2** |
149
  | **代码** | MBPP-plus | 3 | 73.7 | 74.0 | 77.4 | 76.8 | **79.1** |
 
150
  | **多语言** | Belebele (5-language subset)† | 5 | 93.8 | 95.0 | 94.2 | 94.4 | **95.2** |
151
  | | INCLUDE† | 5 | 82.8 | **88.1** | 84.3 | 84.0 | 84.3 |
152
  | **多模态** | MMMU-Pro (Standard, 10 opts) | 4 | 52.0 | **59.6** | — | — | — |