Hugging Face's logo

mradermacher
/
Llama-portuguese-13b-Luana-v0.2-i1-GGUF

Transformers
GGUF
Portuguese
Misral
Portuguese
7b
Inference Endpoints
imatrix
conversational
Model card Files Community
mradermacher commited on
Commit
8cbac8e
•
1 Parent(s): b696c1d

auto-patch README.md

Browse files
Files changed (1) hide show
  1. README.md +78 -0
README.md CHANGED
@@ -1,6 +1,84 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  <!-- ### quantize_version: 2 -->
2
  <!-- ### output_tensor_quantised: 1 -->
3
  <!-- ### convert_type: hf -->
4
  <!-- ### vocab_type: -->
5
  <!-- ### tags: nicoboss -->
6
  weighted/imatrix quants of https://huggingface.co/rhaymison/Llama-portuguese-13b-Luana-v0.2
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: rhaymison/Llama-portuguese-13b-Luana-v0.2
3
+ datasets:
4
+ - pablo-moreira/gpt4all-j-prompt-generations-pt
5
+ - rhaymison/superset
6
+ language:
7
+ - pt
8
+ library_name: transformers
9
+ license: apache-2.0
10
+ quantized_by: mradermacher
11
+ tags:
12
+ - Misral
13
+ - Portuguese
14
+ - 7b
15
+ ---
16
+ ## About
17
+
18
  <!-- ### quantize_version: 2 -->
19
  <!-- ### output_tensor_quantised: 1 -->
20
  <!-- ### convert_type: hf -->
21
  <!-- ### vocab_type: -->
22
  <!-- ### tags: nicoboss -->
23
  weighted/imatrix quants of https://huggingface.co/rhaymison/Llama-portuguese-13b-Luana-v0.2
24
+
25
+ <!-- provided-files -->
26
+ static quants are available at https://huggingface.co/mradermacher/Llama-portuguese-13b-Luana-v0.2-GGUF
27
+ ## Usage
28
+
29
+ If you are unsure how to use GGUF files, refer to one of [TheBloke's
30
+ READMEs](https://huggingface.co/TheBloke/KafkaLM-70B-German-V0.1-GGUF) for
31
+ more details, including on how to concatenate multi-part files.
32
+
33
+ ## Provided Quants
34
+
35
+ (sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants)
36
+
37
+ | Link | Type | Size/GB | Notes |
38
+ |:-----|:-----|--------:|:------|
39
+ | [GGUF](https://huggingface.co/mradermacher/Llama-portuguese-13b-Luana-v0.2-i1-GGUF/resolve/main/Llama-portuguese-13b-Luana-v0.2.i1-IQ1_S.gguf) | i1-IQ1_S | 3.0 | for the desperate |
40
+ | [GGUF](https://huggingface.co/mradermacher/Llama-portuguese-13b-Luana-v0.2-i1-GGUF/resolve/main/Llama-portuguese-13b-Luana-v0.2.i1-IQ1_M.gguf) | i1-IQ1_M | 3.2 | mostly desperate |
41
+ | [GGUF](https://huggingface.co/mradermacher/Llama-portuguese-13b-Luana-v0.2-i1-GGUF/resolve/main/Llama-portuguese-13b-Luana-v0.2.i1-IQ2_XXS.gguf) | i1-IQ2_XXS | 3.6 | |
42
+ | [GGUF](https://huggingface.co/mradermacher/Llama-portuguese-13b-Luana-v0.2-i1-GGUF/resolve/main/Llama-portuguese-13b-Luana-v0.2.i1-IQ2_XS.gguf) | i1-IQ2_XS | 4.0 | |
43
+ | [GGUF](https://huggingface.co/mradermacher/Llama-portuguese-13b-Luana-v0.2-i1-GGUF/resolve/main/Llama-portuguese-13b-Luana-v0.2.i1-IQ2_S.gguf) | i1-IQ2_S | 4.3 | |
44
+ | [GGUF](https://huggingface.co/mradermacher/Llama-portuguese-13b-Luana-v0.2-i1-GGUF/resolve/main/Llama-portuguese-13b-Luana-v0.2.i1-Q2_K_S.gguf) | i1-Q2_K_S | 4.5 | very low quality |
45
+ | [GGUF](https://huggingface.co/mradermacher/Llama-portuguese-13b-Luana-v0.2-i1-GGUF/resolve/main/Llama-portuguese-13b-Luana-v0.2.i1-IQ2_M.gguf) | i1-IQ2_M | 4.6 | |
46
+ | [GGUF](https://huggingface.co/mradermacher/Llama-portuguese-13b-Luana-v0.2-i1-GGUF/resolve/main/Llama-portuguese-13b-Luana-v0.2.i1-Q2_K.gguf) | i1-Q2_K | 5.0 | IQ3_XXS probably better |
47
+ | [GGUF](https://huggingface.co/mradermacher/Llama-portuguese-13b-Luana-v0.2-i1-GGUF/resolve/main/Llama-portuguese-13b-Luana-v0.2.i1-IQ3_XXS.gguf) | i1-IQ3_XXS | 5.1 | lower quality |
48
+ | [GGUF](https://huggingface.co/mradermacher/Llama-portuguese-13b-Luana-v0.2-i1-GGUF/resolve/main/Llama-portuguese-13b-Luana-v0.2.i1-IQ3_XS.gguf) | i1-IQ3_XS | 5.5 | |
49
+ | [GGUF](https://huggingface.co/mradermacher/Llama-portuguese-13b-Luana-v0.2-i1-GGUF/resolve/main/Llama-portuguese-13b-Luana-v0.2.i1-IQ3_S.gguf) | i1-IQ3_S | 5.8 | beats Q3_K* |
50
+ | [GGUF](https://huggingface.co/mradermacher/Llama-portuguese-13b-Luana-v0.2-i1-GGUF/resolve/main/Llama-portuguese-13b-Luana-v0.2.i1-Q3_K_S.gguf) | i1-Q3_K_S | 5.8 | IQ3_XS probably better |
51
+ | [GGUF](https://huggingface.co/mradermacher/Llama-portuguese-13b-Luana-v0.2-i1-GGUF/resolve/main/Llama-portuguese-13b-Luana-v0.2.i1-IQ3_M.gguf) | i1-IQ3_M | 6.1 | |
52
+ | [GGUF](https://huggingface.co/mradermacher/Llama-portuguese-13b-Luana-v0.2-i1-GGUF/resolve/main/Llama-portuguese-13b-Luana-v0.2.i1-Q3_K_M.gguf) | i1-Q3_K_M | 6.4 | IQ3_S probably better |
53
+ | [GGUF](https://huggingface.co/mradermacher/Llama-portuguese-13b-Luana-v0.2-i1-GGUF/resolve/main/Llama-portuguese-13b-Luana-v0.2.i1-Q3_K_L.gguf) | i1-Q3_K_L | 7.0 | IQ3_M probably better |
54
+ | [GGUF](https://huggingface.co/mradermacher/Llama-portuguese-13b-Luana-v0.2-i1-GGUF/resolve/main/Llama-portuguese-13b-Luana-v0.2.i1-IQ4_XS.gguf) | i1-IQ4_XS | 7.1 | |
55
+ | [GGUF](https://huggingface.co/mradermacher/Llama-portuguese-13b-Luana-v0.2-i1-GGUF/resolve/main/Llama-portuguese-13b-Luana-v0.2.i1-Q4_0_4_4.gguf) | i1-Q4_0_4_4 | 7.5 | fast on arm, low quality |
56
+ | [GGUF](https://huggingface.co/mradermacher/Llama-portuguese-13b-Luana-v0.2-i1-GGUF/resolve/main/Llama-portuguese-13b-Luana-v0.2.i1-Q4_0_4_8.gguf) | i1-Q4_0_4_8 | 7.5 | fast on arm+i8mm, low quality |
57
+ | [GGUF](https://huggingface.co/mradermacher/Llama-portuguese-13b-Luana-v0.2-i1-GGUF/resolve/main/Llama-portuguese-13b-Luana-v0.2.i1-Q4_0_8_8.gguf) | i1-Q4_0_8_8 | 7.5 | fast on arm+sve, low quality |
58
+ | [GGUF](https://huggingface.co/mradermacher/Llama-portuguese-13b-Luana-v0.2-i1-GGUF/resolve/main/Llama-portuguese-13b-Luana-v0.2.i1-Q4_0.gguf) | i1-Q4_0 | 7.5 | fast, low quality |
59
+ | [GGUF](https://huggingface.co/mradermacher/Llama-portuguese-13b-Luana-v0.2-i1-GGUF/resolve/main/Llama-portuguese-13b-Luana-v0.2.i1-Q4_K_S.gguf) | i1-Q4_K_S | 7.5 | optimal size/speed/quality |
60
+ | [GGUF](https://huggingface.co/mradermacher/Llama-portuguese-13b-Luana-v0.2-i1-GGUF/resolve/main/Llama-portuguese-13b-Luana-v0.2.i1-Q4_K_M.gguf) | i1-Q4_K_M | 8.0 | fast, recommended |
61
+ | [GGUF](https://huggingface.co/mradermacher/Llama-portuguese-13b-Luana-v0.2-i1-GGUF/resolve/main/Llama-portuguese-13b-Luana-v0.2.i1-Q5_K_S.gguf) | i1-Q5_K_S | 9.1 | |
62
+ | [GGUF](https://huggingface.co/mradermacher/Llama-portuguese-13b-Luana-v0.2-i1-GGUF/resolve/main/Llama-portuguese-13b-Luana-v0.2.i1-Q5_K_M.gguf) | i1-Q5_K_M | 9.3 | |
63
+ | [GGUF](https://huggingface.co/mradermacher/Llama-portuguese-13b-Luana-v0.2-i1-GGUF/resolve/main/Llama-portuguese-13b-Luana-v0.2.i1-Q6_K.gguf) | i1-Q6_K | 10.8 | practically like static Q6_K |
64
+
65
+ Here is a handy graph by ikawrakow comparing some lower-quality quant
66
+ types (lower is better):
67
+
68
+ ![image.png](https://www.nethype.de/huggingface_embed/quantpplgraph.png)
69
+
70
+ And here are Artefact2's thoughts on the matter:
71
+ https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9
72
+
73
+ ## FAQ / Model Request
74
+
75
+ See https://huggingface.co/mradermacher/model_requests for some answers to
76
+ questions you might have and/or if you want some other model quantized.
77
+
78
+ ## Thanks
79
+
80
+ I thank my company, [nethype GmbH](https://www.nethype.de/), for letting
81
+ me use its servers and providing upgrades to my workstation to enable
82
+ this work in my free time. Additional thanks to [@nicoboss](https://huggingface.co/nicoboss) for giving me access to his private supercomputer, enabling me to provide many more imatrix quants, at much higher quality, than I would otherwise be able to.
83
+
84
+ <!-- end -->