Hugging Face's logo

mradermacher
/
Llama-3-70B-Instruct-32k-v0.1-i1-GGUF

Transformers
GGUF
PyTorch
English
facebook
meta
llama
llama-3
Inference Endpoints
imatrix
conversational
Model card Files Community
mradermacher commited on
Commit
48ce43b
·
verified ·
1 Parent(s): 844edb7

auto-patch README.md

Browse files
Files changed (1) hide show
  1. README.md +7 -0
README.md CHANGED
@@ -39,14 +39,21 @@ more details, including on how to concatenate multi-part files.
39
 
40
  | Link | Type | Size/GB | Notes |
41
  |:-----|:-----|--------:|:------|
 
 
 
42
  | [GGUF](https://huggingface.co/mradermacher/Llama-3-70B-Instruct-32k-v0.1-i1-GGUF/resolve/main/Llama-3-70B-Instruct-32k-v0.1.i1-IQ2_M.gguf) | i1-IQ2_M | 24.2 | |
 
43
  | [GGUF](https://huggingface.co/mradermacher/Llama-3-70B-Instruct-32k-v0.1-i1-GGUF/resolve/main/Llama-3-70B-Instruct-32k-v0.1.i1-Q2_K.gguf) | i1-Q2_K | 26.5 | IQ3_XXS probably better |
44
  | [GGUF](https://huggingface.co/mradermacher/Llama-3-70B-Instruct-32k-v0.1-i1-GGUF/resolve/main/Llama-3-70B-Instruct-32k-v0.1.i1-IQ3_XXS.gguf) | i1-IQ3_XXS | 27.6 | lower quality |
 
45
  | [GGUF](https://huggingface.co/mradermacher/Llama-3-70B-Instruct-32k-v0.1-i1-GGUF/resolve/main/Llama-3-70B-Instruct-32k-v0.1.i1-IQ3_M.gguf) | i1-IQ3_M | 32.0 | |
46
  | [GGUF](https://huggingface.co/mradermacher/Llama-3-70B-Instruct-32k-v0.1-i1-GGUF/resolve/main/Llama-3-70B-Instruct-32k-v0.1.i1-Q3_K_M.gguf) | i1-Q3_K_M | 34.4 | IQ3_S probably better |
 
47
  | [GGUF](https://huggingface.co/mradermacher/Llama-3-70B-Instruct-32k-v0.1-i1-GGUF/resolve/main/Llama-3-70B-Instruct-32k-v0.1.i1-IQ4_XS.gguf) | i1-IQ4_XS | 38.0 | |
48
  | [GGUF](https://huggingface.co/mradermacher/Llama-3-70B-Instruct-32k-v0.1-i1-GGUF/resolve/main/Llama-3-70B-Instruct-32k-v0.1.i1-Q4_K_S.gguf) | i1-Q4_K_S | 40.4 | optimal size/speed/quality |
49
  | [GGUF](https://huggingface.co/mradermacher/Llama-3-70B-Instruct-32k-v0.1-i1-GGUF/resolve/main/Llama-3-70B-Instruct-32k-v0.1.i1-Q4_K_M.gguf) | i1-Q4_K_M | 42.6 | fast, recommended |
 
50
  | [PART 1](https://huggingface.co/mradermacher/Llama-3-70B-Instruct-32k-v0.1-i1-GGUF/resolve/main/Llama-3-70B-Instruct-32k-v0.1.i1-Q6_K.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Llama-3-70B-Instruct-32k-v0.1-i1-GGUF/resolve/main/Llama-3-70B-Instruct-32k-v0.1.i1-Q6_K.gguf.part2of2) | i1-Q6_K | 58.0 | practically like static Q6_K |
51
 
52
  Here is a handy graph by ikawrakow comparing some lower-quality quant
 
39
 
40
  | Link | Type | Size/GB | Notes |
41
  |:-----|:-----|--------:|:------|
42
+ | [GGUF](https://huggingface.co/mradermacher/Llama-3-70B-Instruct-32k-v0.1-i1-GGUF/resolve/main/Llama-3-70B-Instruct-32k-v0.1.i1-IQ1_M.gguf) | i1-IQ1_M | 16.9 | mostly desperate |
43
+ | [GGUF](https://huggingface.co/mradermacher/Llama-3-70B-Instruct-32k-v0.1-i1-GGUF/resolve/main/Llama-3-70B-Instruct-32k-v0.1.i1-IQ2_XXS.gguf) | i1-IQ2_XXS | 19.2 | |
44
+ | [GGUF](https://huggingface.co/mradermacher/Llama-3-70B-Instruct-32k-v0.1-i1-GGUF/resolve/main/Llama-3-70B-Instruct-32k-v0.1.i1-IQ2_XS.gguf) | i1-IQ2_XS | 21.2 | |
45
  | [GGUF](https://huggingface.co/mradermacher/Llama-3-70B-Instruct-32k-v0.1-i1-GGUF/resolve/main/Llama-3-70B-Instruct-32k-v0.1.i1-IQ2_M.gguf) | i1-IQ2_M | 24.2 | |
46
+ | [GGUF](https://huggingface.co/mradermacher/Llama-3-70B-Instruct-32k-v0.1-i1-GGUF/resolve/main/Llama-3-70B-Instruct-32k-v0.1.i1-Q2_K_S.gguf) | i1-Q2_K_S | 24.6 | very low quality |
47
  | [GGUF](https://huggingface.co/mradermacher/Llama-3-70B-Instruct-32k-v0.1-i1-GGUF/resolve/main/Llama-3-70B-Instruct-32k-v0.1.i1-Q2_K.gguf) | i1-Q2_K | 26.5 | IQ3_XXS probably better |
48
  | [GGUF](https://huggingface.co/mradermacher/Llama-3-70B-Instruct-32k-v0.1-i1-GGUF/resolve/main/Llama-3-70B-Instruct-32k-v0.1.i1-IQ3_XXS.gguf) | i1-IQ3_XXS | 27.6 | lower quality |
49
+ | [GGUF](https://huggingface.co/mradermacher/Llama-3-70B-Instruct-32k-v0.1-i1-GGUF/resolve/main/Llama-3-70B-Instruct-32k-v0.1.i1-Q3_K_S.gguf) | i1-Q3_K_S | 31.0 | IQ3_XS probably better |
50
  | [GGUF](https://huggingface.co/mradermacher/Llama-3-70B-Instruct-32k-v0.1-i1-GGUF/resolve/main/Llama-3-70B-Instruct-32k-v0.1.i1-IQ3_M.gguf) | i1-IQ3_M | 32.0 | |
51
  | [GGUF](https://huggingface.co/mradermacher/Llama-3-70B-Instruct-32k-v0.1-i1-GGUF/resolve/main/Llama-3-70B-Instruct-32k-v0.1.i1-Q3_K_M.gguf) | i1-Q3_K_M | 34.4 | IQ3_S probably better |
52
+ | [GGUF](https://huggingface.co/mradermacher/Llama-3-70B-Instruct-32k-v0.1-i1-GGUF/resolve/main/Llama-3-70B-Instruct-32k-v0.1.i1-Q3_K_L.gguf) | i1-Q3_K_L | 37.2 | IQ3_M probably better |
53
  | [GGUF](https://huggingface.co/mradermacher/Llama-3-70B-Instruct-32k-v0.1-i1-GGUF/resolve/main/Llama-3-70B-Instruct-32k-v0.1.i1-IQ4_XS.gguf) | i1-IQ4_XS | 38.0 | |
54
  | [GGUF](https://huggingface.co/mradermacher/Llama-3-70B-Instruct-32k-v0.1-i1-GGUF/resolve/main/Llama-3-70B-Instruct-32k-v0.1.i1-Q4_K_S.gguf) | i1-Q4_K_S | 40.4 | optimal size/speed/quality |
55
  | [GGUF](https://huggingface.co/mradermacher/Llama-3-70B-Instruct-32k-v0.1-i1-GGUF/resolve/main/Llama-3-70B-Instruct-32k-v0.1.i1-Q4_K_M.gguf) | i1-Q4_K_M | 42.6 | fast, recommended |
56
+ | [GGUF](https://huggingface.co/mradermacher/Llama-3-70B-Instruct-32k-v0.1-i1-GGUF/resolve/main/Llama-3-70B-Instruct-32k-v0.1.i1-Q5_K_S.gguf) | i1-Q5_K_S | 48.8 | |
57
  | [PART 1](https://huggingface.co/mradermacher/Llama-3-70B-Instruct-32k-v0.1-i1-GGUF/resolve/main/Llama-3-70B-Instruct-32k-v0.1.i1-Q6_K.gguf.part1of2) [PART 2](https://huggingface.co/mradermacher/Llama-3-70B-Instruct-32k-v0.1-i1-GGUF/resolve/main/Llama-3-70B-Instruct-32k-v0.1.i1-Q6_K.gguf.part2of2) | i1-Q6_K | 58.0 | practically like static Q6_K |
58
 
59
  Here is a handy graph by ikawrakow comparing some lower-quality quant