Upload new k-quant GGML quantised models.

README.md

@@ -1,11 +1,6 @@
 ---
-license: other
 inference: false
-tags:
-- llama
-- pytorch
-- chatbot
-- storywriting
+license: other
 ---
 
 <!-- header start -->
@@ -33,39 +28,71 @@ GGML files are for CPU + GPU inference using [llama.cpp](https://github.com/gger
 * [llama-cpp-python](https://github.com/abetlen/llama-cpp-python)
 * [ctransformers](https://github.com/marella/ctransformers)
 
-## Other repositories available
+## Repositories available
+
+* [4-bit GPTQ models for GPU inference](https://huggingface.co/elinas/chronos-13b-4bit)
+* [2, 3, 4, 5, 6 and 8-bit GGML models for CPU+GPU inference](https://huggingface.co/TheBloke/chronos-13B-GGML)
+* [Unquantised fp16 model in pytorch format, for GPU inference and for further conversions](https://huggingface.co/elinas/chronos-13b)
+
+<!-- compatibility_ggml start -->
+## Compatibility
+
+### 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.
 
-* [Elinas' 4-bit GPTQ models for GPU inference](https://huggingface.co/elinas/chronos-13b-4bit)
-* [4-bit, 5-bit, and 8-bit GGML models for CPU+GPU inference](https://huggingface.co/TheBloke/chronos-13B-GGML)
-* [Elinas' original unquantised fp16 model in HF format](https://huggingface.co/elinas/chronos-13b)
+### 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 | RAM required | Use case |
+| Name | Quant method | Bits | Size | Max RAM required | Use case |
 | ---- | ---- | ---- | ---- | ---- | ----- |
-| chronos-13b.ggmlv3.q4_0.bin | q4_0 | 4 | 7.32 GB | 9.82 GB | 4-bit. |
-| chronos-13b.ggmlv3.q4_1.bin | q4_1 | 4 | 8.14 GB | 10.64 GB | 4-bit. Higher accuracy than q4_0 but not as high as q5_0. However has quicker inference than q5 models. |
-| chronos-13b.ggmlv3.q5_0.bin | q5_0 | 5 | 8.95 GB | 11.45 GB | 5-bit. Higher accuracy, higher resource usage and slower inference. |
-| chronos-13b.ggmlv3.q5_1.bin | q5_1 | 5 | 9.76 GB | 12.26 GB | 5-bit. Even higher accuracy, resource usage and slower inference. |
-| chronos-13b.ggmlv3.q8_0.bin | q8_0 | 8 | 13.83 GB | 16.33 GB | 8-bit. Almost indistinguishable from float16. Huge resource use and slow. Not recommended for normal use. |
-
+| .ggmlv3.q2_K.bin | q2_K | 2 | 5.43 GB | 7.93 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. |
+| .ggmlv3.q3_K_L.bin | q3_K_L | 3 | 6.87 GB | 9.37 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 |
+| .ggmlv3.q3_K_M.bin | q3_K_M | 3 | 6.25 GB | 8.75 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 |
+| .ggmlv3.q3_K_S.bin | q3_K_S | 3 | 5.59 GB | 8.09 GB | New k-quant method. Uses GGML_TYPE_Q3_K for all tensors |
+| .ggmlv3.q4_K_M.bin | q4_K_M | 4 | 7.82 GB | 10.32 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 |
+| .ggmlv3.q4_K_S.bin | q4_K_S | 4 | 7.32 GB | 9.82 GB | New k-quant method. Uses GGML_TYPE_Q4_K for all tensors |
+| .ggmlv3.q5_K_M.bin | q5_K_M | 5 | 9.21 GB | 11.71 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 |
+| .ggmlv3.q5_K_S.bin | q5_K_S | 5 | 8.95 GB | 11.45 GB | New k-quant method. Uses GGML_TYPE_Q5_K for all tensors |
+| .ggmlv3.q6_K.bin | q6_K | 6 | 10.68 GB | 13.18 GB | New k-quant method. Uses GGML_TYPE_Q8_K - 6-bit quantization - for all tensors |
+| chronos-13b.ggmlv3.q4_0.bin | q4_0 | 4 | 7.32 GB | 9.82 GB | Original llama.cpp quant method, 4-bit. |
+| chronos-13b.ggmlv3.q4_1.bin | q4_1 | 4 | 8.14 GB | 10.64 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. |
+| chronos-13b.ggmlv3.q5_0.bin | q5_0 | 5 | 8.95 GB | 11.45 GB | Original llama.cpp quant method, 5-bit. Higher accuracy, higher resource usage and slower inference. |
+| chronos-13b.ggmlv3.q5_1.bin | q5_1 | 5 | 9.76 GB | 12.26 GB | Original llama.cpp quant method, 5-bit. Even higher accuracy, resource usage and slower inference. |
+| chronos-13b.ggmlv3.q8_0.bin | q8_0 | 8 | 13.83 GB | 16.33 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.
 
 ## How to run in `llama.cpp`
 
 I use the following command line; adjust for your tastes and needs:
 
 ```
-./main -t 12 -m chronos-13b.v3.q5_0.bin --color -c 2048 --temp 0.7 --repeat_penalty 1.1 -n -1 -p "Below is an instruction that describes a task. Write a response that appropriately completes the request.
-### Instruction:
-Write a story about llamas
-### Response:"
+./main -t 10 -ngl 32 -m .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 12` to the number of physical CPU cores you have. For example if your system has 8 cores/16 threads, use `-t 8`.
+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`.
+
+Change `-ngl 32` to the number of layers to offload to GPU. Remove it if you don't have GPU acceleration.
 
 If you want to have a chat-style conversation, replace the `-p <PROMPT>` argument with `-i -ins`
 
@@ -93,12 +120,16 @@ 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: Chronos 13B
+# Original model card: Elinas' Chronos 13B
+
 
 # chronos-13b
 
@@ -108,8 +139,20 @@ This model is primarily focused on chat, roleplay, and storywriting, but can acc
 
 Chronos generates very long outputs with coherent text, largely due to the human inputs it was trained on.
 
-4bit Quantized (cuda) version: https://huggingface.co/elinas/chronos-13b-4bit
+This model uses Alpaca formatting, so for optimal model performance, use:
+```
+### Instruction:
+Your instruction or question here.
+### Response:
+```
+
+[4bit Quantized (cuda) version](https://huggingface.co/elinas/chronos-13b-4bit)
 
+[GGML Version provided by @TheBloke](https://huggingface.co/TheBloke/chronos-13B-GGML)
+
+--
+license: other
+---
 # LLaMA Model Card
 
 ## Model details
@@ -187,11 +230,11 @@ Hyperparameters for the model architecture
             </tr>
             <tr>
             <th>Number of parameters</th><th>dimension</th><th>n heads</th><th>n layers</th><th>Learn rate</th><th>Batch size</th><th>n tokens</th>
-            </tr>
+            </tr>           
         </thead>
-    <tbody>
+    <tbody>       
         <tr>
-            <th>7B</th> <th>4096</th> <th>32</th> <th>32</th> <th>3.0E-04</th><th>4M</th><th>1T
+            <th>7B</th> <th>4096</th> <th>32</th> <th>32</th> <th>3.0E-04</th><th>4M</th><th>1T 
         </tr>
         <tr>
             <th>13B</th><th>5120</th><th>40</th><th>40</th><th>3.0E-04</th><th>4M</th><th>1T
@@ -201,13 +244,13 @@ Hyperparameters for the model architecture
         </tr>
         <tr>
             <th>65B</th><th>8192</th><th>64</th><th>80</th><th>1.5.E-04</th><th>4M</th><th>1.4T
-        </tr>
+        </tr>     
     </tbody>
 </table>
 
 *Table 1 - Summary of LLama Model Hyperparameters*
 
-We present our results on eight standard common sense reasoning benchmarks in the table below.
+We present our results on eight standard common sense reasoning benchmarks in the table below. 
 <table>
     <thead>
             <tr>
@@ -215,23 +258,23 @@ We present our results on eight standard common sense reasoning benchmarks in th
             </tr>
             <tr>
             <th>Number of parameters</th> <th>BoolQ</th><th>PIQA</th><th>SIQA</th><th>HellaSwag</th><th>WinoGrande</th><th>ARC-e</th><th>ARC-c</th><th>OBQA</th><th>COPA</th>
-            </tr>
+            </tr>           
         </thead>
-    <tbody>
-    <tr>
+    <tbody>    
+    <tr>   
         <th>7B</th><th>76.5</th><th>79.8</th><th>48.9</th><th>76.1</th><th>70.1</th><th>76.7</th><th>47.6</th><th>57.2</th><th>93
-        </th>
+        </th>   
     <tr><th>13B</th><th>78.1</th><th>80.1</th><th>50.4</th><th>79.2</th><th>73</th><th>78.1</th><th>52.7</th><th>56.4</th><th>94
 </th>
     <tr><th>33B</th><th>83.1</th><th>82.3</th><th>50.4</th><th>82.8</th><th>76</th><th>81.4</th><th>57.8</th><th>58.6</th><th>92
 </th>
-    <tr><th>65B</th><th>85.3</th><th>82.8</th><th>52.3</th><th>84.2</th><th>77</th><th>81.5</th><th>56</th><th>60.2</th><th>94</th></tr>
+    <tr><th>65B</th><th>85.3</th><th>82.8</th><th>52.3</th><th>84.2</th><th>77</th><th>81.5</th><th>56</th><th>60.2</th><th>94</th></tr> 
     </tbody>
 </table>
 *Table 2 - Summary of LLama Model Performance on Reasoning tasks*
 
 
-We present our results on bias in the table below. Note that lower value is better indicating lower bias.
+We present our results on bias in the table below. Note that lower value is better indicating lower bias. 
 
 
 | No  | Category             | FAIR LLM |