Instructions to use humpf/FireRed-OCR-4bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use humpf/FireRed-OCR-4bit with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir FireRed-OCR-4bit humpf/FireRed-OCR-4bit
- Notebooks
- Google Colab
- Kaggle
- Local Apps
- LM Studio
Upload config.json with huggingface_hub
Browse files- config.json +3 -1
config.json
CHANGED
|
@@ -3,7 +3,9 @@
|
|
| 3 |
"Qwen3VLForConditionalGeneration"
|
| 4 |
],
|
| 5 |
"dtype": "bfloat16",
|
| 6 |
-
"eos_token_id":
|
|
|
|
|
|
|
| 7 |
"hidden_size": 2048,
|
| 8 |
"image_token_id": 151655,
|
| 9 |
"model_type": "qwen3_vl",
|
|
|
|
| 3 |
"Qwen3VLForConditionalGeneration"
|
| 4 |
],
|
| 5 |
"dtype": "bfloat16",
|
| 6 |
+
"eos_token_id": [
|
| 7 |
+
151645
|
| 8 |
+
],
|
| 9 |
"hidden_size": 2048,
|
| 10 |
"image_token_id": 151655,
|
| 11 |
"model_type": "qwen3_vl",
|