update benchmarks for 2024-05-20
Browse files- README.md +4 -3
- versions.txt +2 -1
README.md
CHANGED
|
@@ -13,7 +13,8 @@ moondream2 is a small vision language model designed to run efficiently on edge
|
|
| 13 |
| 2024-03-06 | 75.4 | 59.8 | 43.1 | 79.5 | 73.2 |
|
| 14 |
| 2024-03-13 | 76.8 | 60.6 | 46.4 | 79.6 | 73.3 |
|
| 15 |
| 2024-04-02 | 77.7 | 61.7 | 49.7 | 80.1 | 74.2 |
|
| 16 |
-
|
|
|
|
|
| 17 |
|
| 18 |
**Usage**
|
| 19 |
|
|
@@ -26,7 +27,7 @@ from transformers import AutoModelForCausalLM, AutoTokenizer
|
|
| 26 |
from PIL import Image
|
| 27 |
|
| 28 |
model_id = "vikhyatk/moondream2"
|
| 29 |
-
revision = "2024-05-
|
| 30 |
model = AutoModelForCausalLM.from_pretrained(
|
| 31 |
model_id, trust_remote_code=True, revision=revision
|
| 32 |
)
|
|
@@ -38,4 +39,4 @@ print(model.answer_question(enc_image, "Describe this image.", tokenizer))
|
|
| 38 |
```
|
| 39 |
|
| 40 |
The model is updated regularly, so we recommend pinning the model version to a
|
| 41 |
-
specific release as shown above.
|
|
|
|
| 13 |
| 2024-03-06 | 75.4 | 59.8 | 43.1 | 79.5 | 73.2 |
|
| 14 |
| 2024-03-13 | 76.8 | 60.6 | 46.4 | 79.6 | 73.3 |
|
| 15 |
| 2024-04-02 | 77.7 | 61.7 | 49.7 | 80.1 | 74.2 |
|
| 16 |
+
| 2024-05-08 | 79.0 | 62.7 | 53.1 | 81.6 | 76.1 |
|
| 17 |
+
| **2024-05-20** (latest) | 79.4 | 63.1 | 57.2 | 82.1 | 76.6 |
|
| 18 |
|
| 19 |
**Usage**
|
| 20 |
|
|
|
|
| 27 |
from PIL import Image
|
| 28 |
|
| 29 |
model_id = "vikhyatk/moondream2"
|
| 30 |
+
revision = "2024-05-20"
|
| 31 |
model = AutoModelForCausalLM.from_pretrained(
|
| 32 |
model_id, trust_remote_code=True, revision=revision
|
| 33 |
)
|
|
|
|
| 39 |
```
|
| 40 |
|
| 41 |
The model is updated regularly, so we recommend pinning the model version to a
|
| 42 |
+
specific release as shown above.
|
versions.txt
CHANGED
|
@@ -2,4 +2,5 @@
|
|
| 2 |
2024-03-06
|
| 3 |
2024-03-13
|
| 4 |
2024-04-02
|
| 5 |
-
2024-05-08
|
|
|
|
|
|
| 2 |
2024-03-06
|
| 3 |
2024-03-13
|
| 4 |
2024-04-02
|
| 5 |
+
2024-05-08
|
| 6 |
+
2024-05-20
|