mradermacher commited on
Commit
a06641e
1 Parent(s): e1902db

auto-patch README.md

Browse files
Files changed (1) hide show
  1. README.md +60 -0
README.md CHANGED
@@ -1,5 +1,65 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
  <!-- ### quantize_version: 1 -->
2
  <!-- ### output_tensor_quantised: 1 -->
3
  <!-- ### convert_type: -->
4
  <!-- ### vocab_type: -->
5
  static quants of https://huggingface.co/kenshinx/Llama-2-7b-chat-xgpt
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ datasets:
3
+ - kenshinx/netlab-blogs
4
+ exported_from: kenshinx/Llama-2-7b-chat-xgpt
5
+ language:
6
+ - en
7
+ library_name: transformers
8
+ license: apache-2.0
9
+ quantized_by: mradermacher
10
+ ---
11
+ ## About
12
+
13
  <!-- ### quantize_version: 1 -->
14
  <!-- ### output_tensor_quantised: 1 -->
15
  <!-- ### convert_type: -->
16
  <!-- ### vocab_type: -->
17
  static quants of https://huggingface.co/kenshinx/Llama-2-7b-chat-xgpt
18
+
19
+
20
+ <!-- provided-files -->
21
+ weighted/imatrix quants seem not to be available (by me) at this time. If they do not show up a week or so after the static ones, I have probably not planned for them. Feel free to request them by opening a Community Discussion.
22
+ ## Usage
23
+
24
+ If you are unsure how to use GGUF files, refer to one of [TheBloke's
25
+ READMEs](https://huggingface.co/TheBloke/KafkaLM-70B-German-V0.1-GGUF) for
26
+ more details, including on how to concatenate multi-part files.
27
+
28
+ ## Provided Quants
29
+
30
+ (sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants)
31
+
32
+ | Link | Type | Size/GB | Notes |
33
+ |:-----|:-----|--------:|:------|
34
+ | [GGUF](https://huggingface.co/mradermacher/Llama-2-7b-chat-xgpt-GGUF/resolve/main/Llama-2-7b-chat-xgpt.Q2_K.gguf) | Q2_K | 2.6 | |
35
+ | [GGUF](https://huggingface.co/mradermacher/Llama-2-7b-chat-xgpt-GGUF/resolve/main/Llama-2-7b-chat-xgpt.IQ3_XS.gguf) | IQ3_XS | 2.9 | |
36
+ | [GGUF](https://huggingface.co/mradermacher/Llama-2-7b-chat-xgpt-GGUF/resolve/main/Llama-2-7b-chat-xgpt.IQ3_S.gguf) | IQ3_S | 3.0 | beats Q3_K* |
37
+ | [GGUF](https://huggingface.co/mradermacher/Llama-2-7b-chat-xgpt-GGUF/resolve/main/Llama-2-7b-chat-xgpt.Q3_K_S.gguf) | Q3_K_S | 3.0 | |
38
+ | [GGUF](https://huggingface.co/mradermacher/Llama-2-7b-chat-xgpt-GGUF/resolve/main/Llama-2-7b-chat-xgpt.IQ3_M.gguf) | IQ3_M | 3.2 | |
39
+ | [GGUF](https://huggingface.co/mradermacher/Llama-2-7b-chat-xgpt-GGUF/resolve/main/Llama-2-7b-chat-xgpt.Q3_K_M.gguf) | Q3_K_M | 3.4 | lower quality |
40
+ | [GGUF](https://huggingface.co/mradermacher/Llama-2-7b-chat-xgpt-GGUF/resolve/main/Llama-2-7b-chat-xgpt.Q3_K_L.gguf) | Q3_K_L | 3.7 | |
41
+ | [GGUF](https://huggingface.co/mradermacher/Llama-2-7b-chat-xgpt-GGUF/resolve/main/Llama-2-7b-chat-xgpt.IQ4_XS.gguf) | IQ4_XS | 3.7 | |
42
+ | [GGUF](https://huggingface.co/mradermacher/Llama-2-7b-chat-xgpt-GGUF/resolve/main/Llama-2-7b-chat-xgpt.Q4_K_S.gguf) | Q4_K_S | 4.0 | fast, recommended |
43
+ | [GGUF](https://huggingface.co/mradermacher/Llama-2-7b-chat-xgpt-GGUF/resolve/main/Llama-2-7b-chat-xgpt.Q4_K_M.gguf) | Q4_K_M | 4.2 | fast, recommended |
44
+ | [GGUF](https://huggingface.co/mradermacher/Llama-2-7b-chat-xgpt-GGUF/resolve/main/Llama-2-7b-chat-xgpt.Q5_K_S.gguf) | Q5_K_S | 4.8 | |
45
+ | [GGUF](https://huggingface.co/mradermacher/Llama-2-7b-chat-xgpt-GGUF/resolve/main/Llama-2-7b-chat-xgpt.Q5_K_M.gguf) | Q5_K_M | 4.9 | |
46
+ | [GGUF](https://huggingface.co/mradermacher/Llama-2-7b-chat-xgpt-GGUF/resolve/main/Llama-2-7b-chat-xgpt.Q6_K.gguf) | Q6_K | 5.6 | very good quality |
47
+ | [GGUF](https://huggingface.co/mradermacher/Llama-2-7b-chat-xgpt-GGUF/resolve/main/Llama-2-7b-chat-xgpt.Q8_0.gguf) | Q8_0 | 7.3 | fast, best quality |
48
+ | [GGUF](https://huggingface.co/mradermacher/Llama-2-7b-chat-xgpt-GGUF/resolve/main/Llama-2-7b-chat-xgpt.f16.gguf) | f16 | 13.6 | 16 bpw, overkill |
49
+
50
+
51
+ Here is a handy graph by ikawrakow comparing some lower-quality quant
52
+ types (lower is better):
53
+
54
+ ![image.png](https://www.nethype.de/huggingface_embed/quantpplgraph.png)
55
+
56
+ And here are Artefact2's thoughts on the matter:
57
+ https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9
58
+
59
+ ## Thanks
60
+
61
+ I thank my company, [nethype GmbH](https://www.nethype.de/), for letting
62
+ me use its servers and providing upgrades to my workstation to enable
63
+ this work in my free time.
64
+
65
+ <!-- end -->