Update README.md

README.md

@@ -15,9 +15,9 @@ GGML files are for CPU inference using [llama.cpp](https://github.com/ggerganov/
 
 ## Other repositories available
 
-* [4bit GPTQ models for GPU inference](https://huggingface.co/TheBloke/WizardLM-30B-Uncensored-GPTQ)
-* [4bit and 5bit GGML models for CPU inference](https://huggingface.co/TheBloke/WizardLM-30B-Uncensored-GGML)
-* [Eric's unquantised model in HF format](https://huggingface.co/ehartford/WizardLM-30B-Uncensored)
+* [4bit GPTQ model for GPU inference](https://huggingface.co/TheBloke/WizardLM-30B-Uncensored-GPTQ)
+* [4-bit, 5-bit and 8-bit GGML models for CPU (+CUDA) inference](https://huggingface.co/TheBloke/WizardLM-30B-Uncensored-GGML)
+* [Eric's unquantised model in fp16 HF format](https://huggingface.co/ehartford/WizardLM-30B-Uncensored)
 
 ## THE FILES IN MAIN BRANCH REQUIRES LATEST LLAMA.CPP (May 19th 2023 - commit 2d5db48)!
 
@@ -28,11 +28,11 @@ I have quantised the GGML files in this repo with the latest version. Therefore
 ## Provided files
 | Name | Quant method | Bits | Size | RAM required | Use case |
 | ---- | ---- | ---- | ---- | ---- | ----- |
-`WizardLM-30B-Uncensored.q4_0.bin` | q4_0 | 4bit | 18.3GB | 20GB | 4bit. |
-`WizardLM-30B-Uncensored.q4_1.bin` | q4_1 | 4bit | 20.3GB | 23GB | 4-bit. Higher accuracy than q4_0 but not as high as q5_0. However has quicker inference than q5 models. |
-`WizardLM-30B-Uncensored.q5_0.bin` | q5_0 | 5bit | 22.4GB | 25GB | 5-bit. Higher accuracy, higher resource usage, slower inference. |
-`WizardLM-30B-Uncensored.q5_1.bin` | q5_1 | 5bit | 24.4GB | 27GB | 5-bit. Even higher accuracy and resource usage, and slower inference. |
-`WizardLM-30B-Uncensored.q8_0.bin` | q8_0 | 5bit | 34.6GB | 38GB | 8-bit. Almost indistinguishable from float16. Huge resource use and slow. Not recommended for normal use.|
+`WizardLM-30B-Uncensored.ggmlv3.q4_0.bin` | q4_0 | 4bit | 18.3GB | 20GB | 4-bit. |
+`WizardLM-30B-Uncensored.ggmlv3.q4_1.bin` | q4_1 | 4bit | 20.3GB | 23GB | 4-bit. Higher accuracy than q4_0 but not as high as q5_0. However has quicker inference than q5 models. |
+`WizardLM-30B-Uncensored.ggmlv3.q5_0.bin` | q5_0 | 5bit | 22.4GB | 25GB | 5-bit. Higher accuracy, higher resource usage, slower inference. |
+`WizardLM-30B-Uncensored.ggmlv3.q5_1.bin` | q5_1 | 5bit | 24.4GB | 27GB | 5-bit. Even higher accuracy and resource usage, and slower inference. |
+`WizardLM-30B-Uncensored.ggmlv3.q8_0.bin` | q8_0 | 8bit | 34.6GB | 38GB | 8-bit. Almost indistinguishable from float16. Huge resource use and slow. Not recommended for normal use.|
 
 ## How to run in `llama.cpp`