Fix model card accuracy (format/filename)

README.md

@@ -1,53 +1,39 @@
 ---
-language:
-  - en
 license: apache-2.0
-library_name: llama-cpp
+library_name: transformers
 tags:
-  - gguf
-  - vision
-  - multimodal
+  - safetensors
   - forkjoin-ai
+  - edge-deployment
+  - aether
 base_model: deepseek-ai/DeepSeek-OCR-2
+base_model_relation: finetune
 pipeline_tag: image-text-to-text
 ---
 
-# Deepseek Ocr 2
+# Deepseek Ocr 2 (SafeTensors)
 
-Forkjoin.ai conversion of [deepseek-ai/DeepSeek-OCR-2](https://huggingface.co/deepseek-ai/DeepSeek-OCR-2) to GGUF format for edge deployment.
+> **SafeTensors mirror** of [deepseek-ai/DeepSeek-OCR-2](https://huggingface.co/deepseek-ai/DeepSeek-OCR-2) for distributed edge inference — powered by the [Aether](https://github.com/forkjoin-ai/aether) runtime on [Edgework.ai](https://edgework.ai).
 
 ## Model Details
 
-- **Source Model**: [deepseek-ai/DeepSeek-OCR-2](https://huggingface.co/deepseek-ai/DeepSeek-OCR-2)
-- **Format**: GGUF
-- **Converted by**: [Forkjoin.ai](https://forkjoin.ai)
+| Property | Value |
+|----------|-------|
+| Base model | [deepseek-ai/DeepSeek-OCR-2](https://huggingface.co/deepseek-ai/DeepSeek-OCR-2) |
+| Format | SafeTensors |
+| Weight files | 1 shard(s) `*.safetensors` |
+| License | apache-2.0 |
 
 ## Usage
 
-### With llama.cpp
+These are **SafeTensors** weights — load them with the model's native framework (e.g. `transformers` / `diffusers`), not llama.cpp.
 
 ```bash
-./llama-cli -m deepseek-ocr-2-gguf.gguf -p "Your prompt here" -n 256
+huggingface-cli download forkjoin-ai/deepseek-ocr-2 --local-dir ./deepseek-ocr-2
 ```
 
-### With Ollama
+## About
 
-Create a `Modelfile`:
-```
-FROM ./deepseek-ocr-2-gguf.gguf
-```
-
-```bash
-ollama create deepseek-ocr-2-gguf -f Modelfile
-ollama run deepseek-ocr-2-gguf
-```
-
-## About Forkjoin.ai
-
-[Forkjoin.ai](https://forkjoin.ai) runs AI models at the edge -- in-browser, on-device, zero cloud cost. These converted models power real-time inference, speech recognition, and natural language capabilities.
-
-All conversions are optimized for edge deployment within browser and mobile memory constraints.
-
-## License
+Published by [AFFECTIVELY](https://huggingface.co/forkjoin-ai) · Managed by [@buley](https://huggingface.co/buley). Mirrored for distributed edge inference via the [Aether](https://github.com/forkjoin-ai/aether) runtime.
 
-Apache 2.0 (follows upstream model license)
+- [All models](https://huggingface.co/forkjoin-ai) · [GitHub](https://github.com/forkjoin-ai) · [Edgework.ai](https://edgework.ai)