Upload folder using huggingface_hub
Browse files
README.md
CHANGED
|
@@ -55,13 +55,13 @@ InternVL 2.0 is a multimodal large language model series, featuring models of va
|
|
| 55 |
|
| 56 |
### Video Benchmarks
|
| 57 |
|
| 58 |
-
|
|
| 59 |
-
|
|
| 60 |
-
|
|
| 61 |
-
|
|
| 62 |
-
|
|
| 63 |
-
| Video-MME<
|
| 64 |
-
| Video-MME<
|
| 65 |
|
| 66 |
- We evaluate our models on MVBench by extracting 16 frames from each video, and each frame was resized to a 448x448 image.
|
| 67 |
|
|
@@ -468,13 +468,13 @@ InternVL 2.0 是一个多模态大语言模型系列，包含各种规模的模
|
|
| 468 |
|
| 469 |
### 视频相关评测
|
| 470 |
|
| 471 |
-
|
|
| 472 |
-
|
|
| 473 |
-
|
|
| 474 |
-
|
|
| 475 |
-
|
|
| 476 |
-
| Video-MME<
|
| 477 |
-
| Video-MME<
|
| 478 |
|
| 479 |
- 我们通过从每个视频中提取16帧来评估我们的模型在MVBench上的性能，每个视频帧被调整为448x448的图像。
|
| 480 |
|
|
|
|
| 55 |
|
| 56 |
### Video Benchmarks
|
| 57 |
|
| 58 |
+
| Benchmark | GPT-4V | VILA-1.5 | LLaVA-NeXT-Video | InternVL-Chat-V1-5 | InternVL2-26B |
|
| 59 |
+
| :------------------: | :----: | :------: | :--------------: | :----------------: | :-----------: |
|
| 60 |
+
| Model Size | - | 34B | 34B | 25.5B | 25.5B |
|
| 61 |
+
| | | | | | |
|
| 62 |
+
| MVBench | - | - | - | 52.1 | 67.5 |
|
| 63 |
+
| Video-MME<br>wo subs | 59.9 | 59.0 | 52.0 | TBD | TBD |
|
| 64 |
+
| Video-MME<br>w/ subs | 63.3 | 59.4 | 54.9 | TBD | TBD |
|
| 65 |
|
| 66 |
- We evaluate our models on MVBench by extracting 16 frames from each video, and each frame was resized to a 448x448 image.
|
| 67 |
|
|
|
|
| 468 |
|
| 469 |
### 视频相关评测
|
| 470 |
|
| 471 |
+
| 评测数据集 | GPT-4V | VILA-1.5 | LLaVA-NeXT-Video | InternVL-Chat-V1-5 | InternVL2-26B |
|
| 472 |
+
| :------------------: | :----: | :------: | :--------------: | :----------------: | :-----------: |
|
| 473 |
+
| 模型大小 | - | 34B | 34B | 25.5B | 25.5B |
|
| 474 |
+
| | | | | | |
|
| 475 |
+
| MVBench | - | - | - | 52.1 | 67.5 |
|
| 476 |
+
| Video-MME<br>wo subs | 59.9 | 59.0 | 52.0 | TBD | TBD |
|
| 477 |
+
| Video-MME<br>w/ subs | 63.3 | 59.4 | 54.9 | TBD | TBD |
|
| 478 |
|
| 479 |
- 我们通过从每个视频中提取16帧来评估我们的模型在MVBench上的性能，每个视频帧被调整为448x448的图像。
|
| 480 |
|