mradermacher commited on
Commit
712686d
1 Parent(s): e89dc21

auto-patch README.md

Browse files
Files changed (1) hide show
  1. README.md +72 -0
README.md CHANGED
@@ -1,5 +1,77 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  <!-- ### quantize_version: 2 -->
2
  <!-- ### output_tensor_quantised: 1 -->
3
  <!-- ### convert_type: hfhfix -->
4
  <!-- ### vocab_type: -->
5
  static quants of https://huggingface.co/Virt-io/Llama-3-8B-Irene-v0.1
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: Virt-io/Llama-3-8B-Irene-v0.1
3
+ language:
4
+ - en
5
+ library_name: transformers
6
+ license: other
7
+ license_link: LICENSE
8
+ license_name: llama3
9
+ quantized_by: mradermacher
10
+ tags:
11
+ - mergekit
12
+ - merge
13
+ - facebook
14
+ - meta
15
+ - pytorch
16
+ - llama
17
+ - llama-3
18
+ ---
19
+ ## About
20
+
21
  <!-- ### quantize_version: 2 -->
22
  <!-- ### output_tensor_quantised: 1 -->
23
  <!-- ### convert_type: hfhfix -->
24
  <!-- ### vocab_type: -->
25
  static quants of https://huggingface.co/Virt-io/Llama-3-8B-Irene-v0.1
26
+
27
+
28
+ <!-- provided-files -->
29
+ weighted/imatrix quants are available at https://huggingface.co/mradermacher/Llama-3-8B-Irene-v0.1-i1-GGUF
30
+ ## Usage
31
+
32
+ If you are unsure how to use GGUF files, refer to one of [TheBloke's
33
+ READMEs](https://huggingface.co/TheBloke/KafkaLM-70B-German-V0.1-GGUF) for
34
+ more details, including on how to concatenate multi-part files.
35
+
36
+ ## Provided Quants
37
+
38
+ (sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants)
39
+
40
+ | Link | Type | Size/GB | Notes |
41
+ |:-----|:-----|--------:|:------|
42
+ | [GGUF](https://huggingface.co/mradermacher/Llama-3-8B-Irene-v0.1-GGUF/resolve/main/Llama-3-8B-Irene-v0.1.Q2_K.gguf) | Q2_K | 3.3 | |
43
+ | [GGUF](https://huggingface.co/mradermacher/Llama-3-8B-Irene-v0.1-GGUF/resolve/main/Llama-3-8B-Irene-v0.1.IQ3_XS.gguf) | IQ3_XS | 3.6 | |
44
+ | [GGUF](https://huggingface.co/mradermacher/Llama-3-8B-Irene-v0.1-GGUF/resolve/main/Llama-3-8B-Irene-v0.1.Q3_K_S.gguf) | Q3_K_S | 3.8 | |
45
+ | [GGUF](https://huggingface.co/mradermacher/Llama-3-8B-Irene-v0.1-GGUF/resolve/main/Llama-3-8B-Irene-v0.1.IQ3_S.gguf) | IQ3_S | 3.8 | beats Q3_K* |
46
+ | [GGUF](https://huggingface.co/mradermacher/Llama-3-8B-Irene-v0.1-GGUF/resolve/main/Llama-3-8B-Irene-v0.1.IQ3_M.gguf) | IQ3_M | 3.9 | |
47
+ | [GGUF](https://huggingface.co/mradermacher/Llama-3-8B-Irene-v0.1-GGUF/resolve/main/Llama-3-8B-Irene-v0.1.Q3_K_M.gguf) | Q3_K_M | 4.1 | lower quality |
48
+ | [GGUF](https://huggingface.co/mradermacher/Llama-3-8B-Irene-v0.1-GGUF/resolve/main/Llama-3-8B-Irene-v0.1.Q3_K_L.gguf) | Q3_K_L | 4.4 | |
49
+ | [GGUF](https://huggingface.co/mradermacher/Llama-3-8B-Irene-v0.1-GGUF/resolve/main/Llama-3-8B-Irene-v0.1.Q4_K_S.gguf) | Q4_K_S | 4.8 | fast, recommended |
50
+ | [GGUF](https://huggingface.co/mradermacher/Llama-3-8B-Irene-v0.1-GGUF/resolve/main/Llama-3-8B-Irene-v0.1.Q4_K_M.gguf) | Q4_K_M | 5.0 | fast, recommended |
51
+ | [GGUF](https://huggingface.co/mradermacher/Llama-3-8B-Irene-v0.1-GGUF/resolve/main/Llama-3-8B-Irene-v0.1.Q5_K_S.gguf) | Q5_K_S | 5.7 | |
52
+ | [GGUF](https://huggingface.co/mradermacher/Llama-3-8B-Irene-v0.1-GGUF/resolve/main/Llama-3-8B-Irene-v0.1.Q5_K_M.gguf) | Q5_K_M | 5.8 | |
53
+ | [GGUF](https://huggingface.co/mradermacher/Llama-3-8B-Irene-v0.1-GGUF/resolve/main/Llama-3-8B-Irene-v0.1.Q6_K.gguf) | Q6_K | 6.7 | very good quality |
54
+ | [GGUF](https://huggingface.co/mradermacher/Llama-3-8B-Irene-v0.1-GGUF/resolve/main/Llama-3-8B-Irene-v0.1.Q8_0.gguf) | Q8_0 | 8.6 | fast, best quality |
55
+ | [GGUF](https://huggingface.co/mradermacher/Llama-3-8B-Irene-v0.1-GGUF/resolve/main/Llama-3-8B-Irene-v0.1.f16.gguf) | f16 | 16.2 | 16 bpw, overkill |
56
+
57
+
58
+ Here is a handy graph by ikawrakow comparing some lower-quality quant
59
+ types (lower is better):
60
+
61
+ ![image.png](https://www.nethype.de/huggingface_embed/quantpplgraph.png)
62
+
63
+ And here are Artefact2's thoughts on the matter:
64
+ https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9
65
+
66
+ ## FAQ / Model Request
67
+
68
+ See https://huggingface.co/mradermacher/model_requests for some answers to
69
+ questions you might have and/or if you want some other model quantized.
70
+
71
+ ## Thanks
72
+
73
+ I thank my company, [nethype GmbH](https://www.nethype.de/), for letting
74
+ me use its servers and providing upgrades to my workstation to enable
75
+ this work in my free time.
76
+
77
+ <!-- end -->