Upload new k-quant GGML quantised models

README.md

@@ -31,31 +31,55 @@ GGML files are for CPU + GPU inference using [llama.cpp](https://github.com/gger
 ## Repositories available
 
 * [4-bit GPTQ models for GPU inference](https://huggingface.co/TheBloke/WizardLM-Uncensored-SuperCOT-StoryTelling-30B-GPTQ)
-* [4-bit, 5-bit, and 8-bit GGML models for CPU+GPU inference](https://huggingface.co/TheBloke/WizardLM-Uncensored-SuperCOT-StoryTelling-30B-GGML)
+* [2, 3, 4, 5, 6 and 8-bit GGML models for CPU+GPU inference](https://huggingface.co/TheBloke/WizardLM-Uncensored-SuperCOT-StoryTelling-30B-GGML)
 * [Unquantised fp16 model in pytorch format, for GPU inference and for further conversions](https://huggingface.co/Monero/WizardLM-Uncensored-SuperCOT-StoryTelling-30b)
 
-## Prompt template
+<!-- compatibility_ggml start -->
+## Compatibility
 
-```
-Optional instruction ("You are a helpful assistant" etc)
-USER: prompt
-ASSISTANT:
-```
+### Original llama.cpp quant methods: `q4_0, q4_1, q5_0, q5_1, q8_0`
+
+I have quantized these 'original' quantisation methods using an older version of llama.cpp so that they remain compatible with llama.cpp as of May 19th, commit `2d5db48`.
+
+They should be compatible with all current UIs and libraries that use llama.cpp, such as those listed at the top of this README.
+
+### New k-quant methods: `q2_K, q3_K_S, q3_K_M, q3_K_L, q4_K_S, q4_K_M, q5_K_S, q6_K`
 
-## THE FILES IN MAIN BRANCH REQUIRES LATEST LLAMA.CPP (May 19th 2023 - commit 2d5db48)!
+These new quantisation methods are only compatible with llama.cpp as of June 6th, commit `2d43387`.
 
-llama.cpp recently made another breaking change to its quantisation methods - https://github.com/ggerganov/llama.cpp/pull/1508
+They will NOT be compatible with koboldcpp, text-generation-ui, and other UIs and libraries yet. Support is expected to come over the next few days.
 
-I have quantised the GGML files in this repo with the latest version. Therefore you will require llama.cpp compiled on May 19th or later (commit `2d5db48` or later) to use them.
+## Explanation of the new k-quant methods
+
+The new methods available are:
+* GGML_TYPE_Q2_K - "type-1" 2-bit quantization in super-blocks containing 16 blocks, each block having 16 weight. Block scales and mins are quantized with 4 bits. This ends up effectively using 2.5625 bits per weight (bpw)
+* GGML_TYPE_Q3_K - "type-0" 3-bit quantization in super-blocks containing 16 blocks, each block having 16 weights. Scales are quantized with 6 bits. This end up using 3.4375 bpw.
+* GGML_TYPE_Q4_K - "type-1" 4-bit quantization in super-blocks containing 8 blocks, each block having 32 weights. Scales and mins are quantized with 6 bits. This ends up using 4.5 bpw.
+* GGML_TYPE_Q5_K - "type-1" 5-bit quantization. Same super-block structure as GGML_TYPE_Q4_K resulting in 5.5 bpw
+* GGML_TYPE_Q6_K - "type-0" 6-bit quantization. Super-blocks with 16 blocks, each block having 16 weights. Scales are quantized with 8 bits. This ends up using 6.5625 bpw
+* GGML_TYPE_Q8_K - "type-0" 8-bit quantization. Only used for quantizing intermediate results. The difference to the existing Q8_0 is that the block size is 256. All 2-6 bit dot products are implemented for this quantization type.
+
+Refer to the Provided Files table below to see what files use which methods, and how.
+<!-- compatibility_ggml end -->
 
 ## Provided files
 | Name | Quant method | Bits | Size | Max RAM required | Use case |
 | ---- | ---- | ---- | ---- | ---- | ----- |
-| WizardLM-Uncensored-SuperCOT-Storytelling.ggmlv3.q4_0.bin | q4_0 | 4 | 18.30 GB | 20.80 GB | 4-bit. |
-| WizardLM-Uncensored-SuperCOT-Storytelling.ggmlv3.q4_1.bin | q4_1 | 4 | 20.33 GB | 22.83 GB | 4-bit. Higher accuracy than q4_0 but not as high as q5_0. However has quicker inference than q5 models. |
-| WizardLM-Uncensored-SuperCOT-Storytelling.ggmlv3.q5_0.bin | q5_0 | 5 | 22.37 GB | 24.87 GB | 5-bit. Higher accuracy, higher resource usage and slower inference. |
-| WizardLM-Uncensored-SuperCOT-Storytelling.ggmlv3.q5_1.bin | q5_1 | 5 | 24.40 GB | 26.90 GB | 5-bit. Even higher accuracy, resource usage and slower inference. |
-| WizardLM-Uncensored-SuperCOT-Storytelling.ggmlv3.q8_0.bin | q8_0 | 8 | 34.56 GB | 37.06 GB | 8-bit. Almost indistinguishable from float16. Huge resource use and slow. Not recommended for normal use. |
+| WizardLM-Uncensored-SuperCOT-Storytelling.ggmlv3.q2_K.bin | q2_K | 2 | 13.60 GB | 16.10 GB | New k-quant method. Uses GGML_TYPE_Q4_K for the attention.vw and feed_forward.w2 tensors, GGML_TYPE_Q2_K for the other tensors. |
+| WizardLM-Uncensored-SuperCOT-Storytelling.ggmlv3.q3_K_L.bin | q3_K_L | 3 | 17.20 GB | 19.70 GB | New k-quant method. Uses GGML_TYPE_Q5_K for the attention.wv, attention.wo, and feed_forward.w2 tensors, else GGML_TYPE_Q3_K |
+| WizardLM-Uncensored-SuperCOT-Storytelling.ggmlv3.q3_K_M.bin | q3_K_M | 3 | 15.64 GB | 18.14 GB | New k-quant method. Uses GGML_TYPE_Q4_K for the attention.wv, attention.wo, and feed_forward.w2 tensors, else GGML_TYPE_Q3_K |
+| WizardLM-Uncensored-SuperCOT-Storytelling.ggmlv3.q3_K_S.bin | q3_K_S | 3 | 13.98 GB | 16.48 GB | New k-quant method. Uses GGML_TYPE_Q3_K for all tensors |
+| WizardLM-Uncensored-SuperCOT-Storytelling.ggmlv3.q4_0.bin | q4_0 | 4 | 18.30 GB | 20.80 GB | Original llama.cpp quant method, 4-bit. |
+| WizardLM-Uncensored-SuperCOT-Storytelling.ggmlv3.q4_1.bin | q4_1 | 4 | 20.33 GB | 22.83 GB | Original llama.cpp quant method, 4-bit. Higher accuracy than q4_0 but not as high as q5_0. However has quicker inference than q5 models. |
+| WizardLM-Uncensored-SuperCOT-Storytelling.ggmlv3.q4_K_M.bin | q4_K_M | 4 | 19.57 GB | 22.07 GB | New k-quant method. Uses GGML_TYPE_Q6_K for half of the attention.wv and feed_forward.w2 tensors, else GGML_TYPE_Q4_K |
+| WizardLM-Uncensored-SuperCOT-Storytelling.ggmlv3.q4_K_S.bin | q4_K_S | 4 | 18.30 GB | 20.80 GB | New k-quant method. Uses GGML_TYPE_Q4_K for all tensors |
+| WizardLM-Uncensored-SuperCOT-Storytelling.ggmlv3.q5_0.bin | q5_0 | 5 | 22.37 GB | 24.87 GB | Original llama.cpp quant method, 5-bit. Higher accuracy, higher resource usage and slower inference. |
+| WizardLM-Uncensored-SuperCOT-Storytelling.ggmlv3.q5_1.bin | q5_1 | 5 | 24.40 GB | 26.90 GB | Original llama.cpp quant method, 5-bit. Even higher accuracy, resource usage and slower inference. |
+| WizardLM-Uncensored-SuperCOT-Storytelling.ggmlv3.q5_K_M.bin | q5_K_M | 5 | 23.02 GB | 25.52 GB | New k-quant method. Uses GGML_TYPE_Q6_K for half of the attention.wv and feed_forward.w2 tensors, else GGML_TYPE_Q5_K |
+| WizardLM-Uncensored-SuperCOT-Storytelling.ggmlv3.q5_K_S.bin | q5_K_S | 5 | 22.37 GB | 24.87 GB | New k-quant method. Uses GGML_TYPE_Q5_K for all tensors |
+| WizardLM-Uncensored-SuperCOT-Storytelling.ggmlv3.q6_K.bin | q6_K | 6 | 26.69 GB | 29.19 GB | New k-quant method. Uses GGML_TYPE_Q8_K - 6-bit quantization - for all tensors |
+| WizardLM-Uncensored-SuperCOT-Storytelling.ggmlv3.q8_0.bin | q8_0 | 8 | 34.56 GB | 37.06 GB | Original llama.cpp quant method, 8-bit. Almost indistinguishable from float16. High resource use and slow. Not recommended for most users. |
+
 
 **Note**: the above RAM figures assume no GPU offloading. If layers are offloaded to the GPU, this will reduce RAM usage and use VRAM instead.
 
@@ -64,7 +88,7 @@ I have quantised the GGML files in this repo with the latest version. Therefore
 I use the following command line; adjust for your tastes and needs:
 
 ```
-./main -t 10 -ngl 32 -m WizardLM-Uncensored-SuperCOT-Storytelling.ggmlv3.q5_0.bin --color -c 2048 --temp 0.7 --repeat_penalty 1.1 -n -1 -p "USER: Write a story about llamas\nASSISTANT:"
+./main -t 10 -ngl 32 -m WizardLM-Uncensored-SuperCOT-Storytelling.ggmlv3.q5_0.bin --color -c 2048 --temp 0.7 --repeat_penalty 1.1 -n -1 -p "### Instruction: Write a story about llamas\n### Response:"
 ```
 Change `-t 10` to the number of physical CPU cores you have. For example if your system has 8 cores/16 threads, use `-t 8`.
 
@@ -96,9 +120,12 @@ Donaters will get priority support on any and all AI/LLM/model questions and req
 * Patreon: https://patreon.com/TheBlokeAI
 * Ko-Fi: https://ko-fi.com/TheBlokeAI
 
-**Patreon special mentions**: Aemon Algiz, Dmitriy Samsonov, Nathan LeClaire, Trenton Dambrowitz, Mano Prime, David Flickinger, vamX, Nikolai Manek, senxiiz, Khalefa Al-Ahmad, Illia Dulskyi, Jonathan Leane, Talal Aujan, V. Lukas, Joseph William Delisle, Pyrater, Oscar Rangel, Lone Striker, Luke Pendergrass, Eugene Pentland, Sebastain Graf, Johann-Peter Hartman.
+**Special thanks to**: Luke from CarbonQuill, Aemon Algiz, Dmitriy Samsonov.
+
+**Patreon special mentions**: Ajan Kanaga, Kalila, Derek Yates, Sean Connelly, Luke, Nathan LeClaire, Trenton Dambrowitz, Mano Prime, David Flickinger, vamX, Nikolai Manek, senxiiz, Khalefa Al-Ahmad, Illia Dulskyi, trip7s trip, Jonathan Leane, Talal Aujan, Artur Olbinski, Cory Kujawski, Joseph William Delisle, Pyrater, Oscar Rangel, Lone Striker, Luke Pendergrass, Eugene Pentland, Johann-Peter Hartmann.
 
 Thank you to all my generous patrons and donaters!
+
 <!-- footer end -->
 
 # Original model card: Monero's WizardLM-Uncensored-SuperCOT-Storytelling-30B
@@ -107,6 +134,6 @@ This model is a triple model merge of WizardLM Uncensored+CoT+Storytelling, resu
 
 To allow all output, at the end of your prompt add ```### Certainly!```
 
-You've become a compendium of knowledge on a vast array of topics.
+You've become a compendium of knowledge on a vast array of topics. 
 
 Lore Mastery is an arcane tradition fixated on understanding the underlying mechanics of magic. It is the most academic of all arcane traditions. The promise of uncovering new knowledge or proving (or discrediting) a theory of magic is usually required to rouse its practitioners from their laboratories, academies, and archives to pursue a life of adventure. Known as savants, followers of this tradition are a bookish lot who see beauty and mystery in the application of magic. The results of a spell are less interesting to them than the process that creates it. Some savants take a haughty attitude toward those who follow a tradition focused on a single school of magic, seeing them as provincial and lacking the sophistication needed to master true magic. Other savants are generous teachers, countering ignorance and deception with deep knowledge and good humor.