TheBloke commited on
Commit
955560e
1 Parent(s): 987de3b

Initial GGML model commit

Browse files
Files changed (1) hide show
  1. README.md +79 -0
README.md ADDED
@@ -0,0 +1,79 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ inference: false
3
+ license: other
4
+ ---
5
+
6
+ <div style="width: 100%;">
7
+ <img src="https://i.imgur.com/EBdldam.jpg" alt="TheBlokeAI" style="width: 100%; min-width: 400px; display: block; margin: auto;">
8
+ </div>
9
+ <div style="display: flex; justify-content: space-between; width: 100%;">
10
+ <div style="display: flex; flex-direction: column; align-items: flex-start;">
11
+ <p><a href="https://discord.gg/UBgz4VXf">Chat & support: my new Discord server</a></p>
12
+ </div>
13
+ <div style="display: flex; flex-direction: column; align-items: flex-end;">
14
+ <p><a href="https://www.patreon.com/TheBlokeAI">Want to contribute? Patreon coming soon!</a></p>
15
+ </div>
16
+ </div>
17
+
18
+ # Aeala's VicUnlocked Alpaca 65B QLoRA GGML
19
+
20
+ These files are GGML format model files for [Aeala's VicUnlocked Alpaca 65B QLoRA](https://huggingface.co/Aeala/VicUnlocked-alpaca-65b-QLoRA).
21
+
22
+ GGML files are for CPU + GPU inference using [llama.cpp](https://github.com/ggerganov/llama.cpp) and libraries and UIs which support this format, such as:
23
+ * [text-generation-webui](https://github.com/oobabooga/text-generation-webui)
24
+ * [KoboldCpp](https://github.com/LostRuins/koboldcpp)
25
+ * [ParisNeo/GPT4All-UI](https://github.com/ParisNeo/gpt4all-ui)
26
+ * [llama-cpp-python](https://github.com/abetlen/llama-cpp-python)
27
+ * [ctransformers](https://github.com/marella/ctransformers)
28
+
29
+ ## Other repositories available
30
+
31
+ * [4-bit GPTQ models for GPU inference](https://huggingface.co/Aeala/VicUnlocked-alpaca-65b-4bit)
32
+ * [4-bit, 5-bit, and 8-bit GGML models for CPU+GPU inference](https://huggingface.co/TheBloke/VicUnlocked-alpaca-65B-QLoRA-GGML)
33
+ * [Original unquantised fp16 model in HF format](https://huggingface.co/Aeala/VicUnlocked-alpaca-65b-QLoRA)
34
+
35
+ ## THE FILES IN MAIN BRANCH REQUIRES LATEST LLAMA.CPP (May 19th 2023 - commit 2d5db48)!
36
+
37
+ llama.cpp recently made another breaking change to its quantisation methods - https://github.com/ggerganov/llama.cpp/pull/1508
38
+
39
+ 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.
40
+
41
+ ## Provided files
42
+ | Name | Quant method | Bits | Size | RAM required | Use case |
43
+ | ---- | ---- | ---- | ---- | ---- | ----- |
44
+ | VicUnlocked-Alpaca-65B.ggmlv3.q4_0.bin | q4_0 | 4 | 36.73 GB | 39.23 GB | 4-bit. |
45
+ | VicUnlocked-Alpaca-65B.ggmlv3.q4_1.bin | q4_1 | 4 | 40.81 GB | 43.31 GB | 4-bit. Higher accuracy than q4_0 but not as high as q5_0. However has quicker inference than q5 models. |
46
+ | VicUnlocked-Alpaca-65B.ggmlv3.q5_0.bin | q5_0 | 5 | 44.89 GB | 47.39 GB | 5-bit. Higher accuracy, higher resource usage and slower inference. |
47
+ | VicUnlocked-Alpaca-65B.ggmlv3.q5_1.bin | q5_1 | 5 | 48.97 GB | 51.47 GB | 5-bit. Even higher accuracy, resource usage and slower inference. |
48
+
49
+
50
+ ## How to run in `llama.cpp`
51
+
52
+ I use the following command line; adjust for your tastes and needs:
53
+
54
+ ```
55
+ ./main -t 10 -ngl 32 -m VicUnlocked-Alpaca-65B.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:"
56
+ ```
57
+ 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`.
58
+
59
+ Change `-ngl 32` to the number of layers to offload to GPU. Remove it if you don't have GPU acceleration.
60
+
61
+ If you want to have a chat-style conversation, replace the `-p <PROMPT>` argument with `-i -ins`
62
+
63
+ ## How to run in `text-generation-webui`
64
+
65
+ Further instructions here: [text-generation-webui/docs/llama.cpp-models.md](https://github.com/oobabooga/text-generation-webui/blob/main/docs/llama.cpp-models.md).
66
+
67
+ ## Want to support my work?
68
+
69
+ I've had a lot of people ask if they can contribute. I love providing models and helping people, but it is starting to rack up pretty big cloud computing bills.
70
+
71
+ So if you're able and willing to contribute, it'd be most gratefully received and will help me to keep providing models, and work on various AI projects.
72
+
73
+ Donaters will get priority support on any and all AI/LLM/model questions, and I'll gladly quantise any model you'd like to try.
74
+
75
+ * Patreon: coming soon! (just awaiting approval)
76
+ * Ko-Fi: https://ko-fi.com/TheBlokeAI
77
+ * Discord: https://discord.gg/UBgz4VXf
78
+
79
+ # Original model card: Aeala's VicUnlocked Alpaca 65B QLoRA