Update README.md

README.md

@@ -99,10 +99,11 @@ snapshot_download(
     merged_file.gguf
 ```
 
-| MoE Bits     | Type   | Disk Size |  Accuracy | Link                      | Details   |
+| Dynamic Bits     | Type   | Disk Size |  Accuracy | Link                      | Details   |
 | -------- | -------- | ------------ | ------------ | ---------------------|  ---------- |
-| 2.22bit | UD-IQ2_XXS |   **183GB**    | Better      | [Link](https://huggingface.co/unsloth/r1-1776-GGUF/tree/main/r1-1776-UD-IQ2_XXS) | MoE all 2.06bit. `down_proj` in MoE mixture of 2.5/2.06bit |
-| 2.51bit | UD-Q2_K_XL |   **212GB**    | Best | [Link](https://huggingface.co/unsloth/r1-1776-GGUF/tree/main/r1-1776-UD-Q2_K_XL) | MoE all 2.5bit. `down_proj` in MoE mixture of 3.5/2.5bit |
+| 2bit | UD-Q2_K_XL |   **211GB**    | Better      | [Link](https://huggingface.co/unsloth/r1-1776-GGUF/tree/main/r1-1776-UD-Q2_K_XL) | MoE all 2.5bit. `down_proj` in MoE mixture of 3.5/2.5bit |
+| 3bit | UD-Q3_K_XL |   **298GB**    | Best        | [Link](https://huggingface.co/unsloth/r1-1776-GGUF/tree/main/r1-1776-UD-Q3_K_XL) | MoE Q3_K_M. Attention parts are upcasted |
+| 4bit | UD-Q4_K_XL |   **377GB**    | Best        | [Link](https://huggingface.co/unsloth/r1-1776-GGUF/tree/main/r1-1776-UD-Q4_K_XL) | MoE Q4_K_M. Attention parts are upcasted |
 
 # Finetune your own Reasoning model like R1 with Unsloth!
 We have a free Google Colab notebook for turning Llama 3.1 (8B) into a reasoning model: https://colab.research.google.com/github/unslothai/notebooks/blob/main/nb/Llama3.1_(8B)-GRPO.ipynb