Instructions to use llm-jp/Jagle-VL-2.2B-Jagle-FineVision with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use llm-jp/Jagle-VL-2.2B-Jagle-FineVision with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("feature-extraction", model="llm-jp/Jagle-VL-2.2B-Jagle-FineVision", trust_remote_code=True)# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("llm-jp/Jagle-VL-2.2B-Jagle-FineVision", trust_remote_code=True, device_map="auto") - Notebooks
- Google Colab
- Kaggle
Upload constants.py with huggingface_hub
Browse files- constants.py +22 -0
constants.py
ADDED
|
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# --------------------------------------------------------
|
| 2 |
+
# LLM-jp-VL
|
| 3 |
+
# Copyright (c) 2026 LLM-jp
|
| 4 |
+
# Licensed under The Apache License 2.0 [see LICENSE for details]
|
| 5 |
+
#
|
| 6 |
+
# Originally based on InternVL
|
| 7 |
+
# Copyright (c) 2024 OpenGVLab
|
| 8 |
+
# Licensed under The MIT License [see LICENSE for details]
|
| 9 |
+
# --------------------------------------------------------
|
| 10 |
+
|
| 11 |
+
IMG_CONTEXT_TOKEN = "<|image_pad|>"
|
| 12 |
+
|
| 13 |
+
HARMONY_START = "<|start|>"
|
| 14 |
+
HARMONY_END = "<|end|>"
|
| 15 |
+
HARMONY_MESSAGE = "<|message|>"
|
| 16 |
+
HARMONY_CHANNEL = "<|channel|>"
|
| 17 |
+
HARMONY_RETURN = "<|return|>"
|
| 18 |
+
|
| 19 |
+
IMAGE_START = "<|image_start|>"
|
| 20 |
+
IMAGE_END = "<|image_end|>"
|
| 21 |
+
|
| 22 |
+
DEFAULT_SYSTEM_MESSAGE = "You are LLM-jp-VL, a Multimodal LLM trained by LLM-jp."
|