mradermacher commited on
Commit
12f5110
1 Parent(s): fcbcb54

auto-patch README.md

Browse files
Files changed (1) hide show
  1. README.md +52 -0
README.md CHANGED
@@ -1,5 +1,57 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  <!-- ### quantize_version: 1 -->
2
  <!-- ### output_tensor_quantised: 1 -->
3
  <!-- ### convert_type: -->
4
  <!-- ### vocab_type: -->
5
  weighted/imatrix quants of https://huggingface.co/wenbopan/Faro-Yi-34B-DPO
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ datasets:
3
+ - wenbopan/Chinese-dpo-pairs
4
+ - Intel/orca_dpo_pairs
5
+ - argilla/ultrafeedback-binarized-preferences-cleaned
6
+ - jondurbin/truthy-dpo-v0.1
7
+ exported_from: wenbopan/Faro-Yi-34B-DPO
8
+ language:
9
+ - en
10
+ library_name: transformers
11
+ license: mit
12
+ quantized_by: mradermacher
13
+ ---
14
+ ## About
15
+
16
  <!-- ### quantize_version: 1 -->
17
  <!-- ### output_tensor_quantised: 1 -->
18
  <!-- ### convert_type: -->
19
  <!-- ### vocab_type: -->
20
  weighted/imatrix quants of https://huggingface.co/wenbopan/Faro-Yi-34B-DPO
21
+
22
+
23
+ <!-- provided-files -->
24
+ static quants are available at https://huggingface.co/mradermacher/Faro-Yi-34B-DPO-GGUF
25
+ ## Usage
26
+
27
+ If you are unsure how to use GGUF files, refer to one of [TheBloke's
28
+ READMEs](https://huggingface.co/TheBloke/KafkaLM-70B-German-V0.1-GGUF) for
29
+ more details, including on how to concatenate multi-part files.
30
+
31
+ ## Provided Quants
32
+
33
+ (sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants)
34
+
35
+ | Link | Type | Size/GB | Notes |
36
+ |:-----|:-----|--------:|:------|
37
+ | [GGUF](https://huggingface.co/mradermacher/Faro-Yi-34B-DPO-i1-GGUF/resolve/main/Faro-Yi-34B-DPO.i1-Q2_K.gguf) | i1-Q2_K | 12.9 | IQ3_XXS probably better |
38
+ | [GGUF](https://huggingface.co/mradermacher/Faro-Yi-34B-DPO-i1-GGUF/resolve/main/Faro-Yi-34B-DPO.i1-IQ3_XXS.gguf) | i1-IQ3_XXS | 13.4 | lower quality |
39
+ | [GGUF](https://huggingface.co/mradermacher/Faro-Yi-34B-DPO-i1-GGUF/resolve/main/Faro-Yi-34B-DPO.i1-Q3_K_M.gguf) | i1-Q3_K_M | 16.8 | IQ3_S probably better |
40
+ | [GGUF](https://huggingface.co/mradermacher/Faro-Yi-34B-DPO-i1-GGUF/resolve/main/Faro-Yi-34B-DPO.i1-Q4_K_S.gguf) | i1-Q4_K_S | 19.7 | optimal size/speed/quality |
41
+
42
+
43
+ Here is a handy graph by ikawrakow comparing some lower-quality quant
44
+ types (lower is better):
45
+
46
+ ![image.png](https://www.nethype.de/huggingface_embed/quantpplgraph.png)
47
+
48
+ And here are Artefact2's thoughts on the matter:
49
+ https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9
50
+
51
+ ## Thanks
52
+
53
+ I thank my company, [nethype GmbH](https://www.nethype.de/), for letting
54
+ me use its servers and providing upgrades to my workstation to enable
55
+ this work in my free time.
56
+
57
+ <!-- end -->