nightknocker commited on
Commit
f52e839
·
verified ·
1 Parent(s): 790cd55

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +5 -2
README.md CHANGED
@@ -3,6 +3,9 @@ license: apache-2.0
3
  base_model:
4
  - lightonai/LightOnOCR-2-1B
5
  base_model_relation: adapter
 
 
 
6
  ---
7
 
8
  # Image As Text For Flux.2 Klein
@@ -12,7 +15,7 @@ It is transformed into Qwen3 embeddings, which represent the underlying meaning
12
 
13
  Compared to T5/Gemma/Qwen, the Mistral ViT model's output sequence length is considerably shorter. The same information is encoded more efficiently.
14
 
15
- In other literature (2601.14750, Figure3) the [safetensors](https://huggingface.co/nightknocker/flux2-klein-4b-lighting-text-encoder/blob/main/model.safetensors) file is called Visual Projection Head.
16
 
17
  ## Inference
18
 
@@ -50,4 +53,4 @@ pipeline = Flux2KleinPipeline.from_pretrained('black-forest-labs/FLUX.2-klein-ba
50
  - spatial-caption
51
  - spright-coco
52
  - z-image-ethnicity-test
53
- - benchmarks from the Qwen-Image Technical Report
 
3
  base_model:
4
  - lightonai/LightOnOCR-2-1B
5
  base_model_relation: adapter
6
+ tags:
7
+ - flux2
8
+ - flux-2-klein
9
  ---
10
 
11
  # Image As Text For Flux.2 Klein
 
15
 
16
  Compared to T5/Gemma/Qwen, the Mistral ViT model's output sequence length is considerably shorter. The same information is encoded more efficiently.
17
 
18
+ In other literature (2601.14750, Figure 3) the [safetensors](https://huggingface.co/nightknocker/flux2-klein-4b-lighting-text-encoder/blob/main/model.safetensors) file is called Visual Projection Head.
19
 
20
  ## Inference
21
 
 
53
  - spatial-caption
54
  - spright-coco
55
  - z-image-ethnicity-test
56
+ - benchmarks from the Qwen-Image Technical Report