Hugging Face's logo

bartowski
/
Llama3-ChatQA-1.5-70B-GGUF

Text Generation
GGUF
PyTorch
English
nvidia
chatqa-1.5
chatqa
llama-3
Model card Files Community
bartowski commited on
Commit
5661d4a
•
1 Parent(s): 6498cf5

Llamacpp quants

Browse files
Files changed (33) hide show
  1. .gitattributes +31 -0
  2. Llama3-ChatQA-1.5-70B-IQ1_M.gguf +3 -0
  3. Llama3-ChatQA-1.5-70B-IQ1_S.gguf +3 -0
  4. Llama3-ChatQA-1.5-70B-IQ2_M.gguf +3 -0
  5. Llama3-ChatQA-1.5-70B-IQ2_S.gguf +3 -0
  6. Llama3-ChatQA-1.5-70B-IQ2_XS.gguf +3 -0
  7. Llama3-ChatQA-1.5-70B-IQ2_XXS.gguf +3 -0
  8. Llama3-ChatQA-1.5-70B-IQ3_M.gguf +3 -0
  9. Llama3-ChatQA-1.5-70B-IQ3_S.gguf +3 -0
  10. Llama3-ChatQA-1.5-70B-IQ3_XS.gguf +3 -0
  11. Llama3-ChatQA-1.5-70B-IQ3_XXS.gguf +3 -0
  12. Llama3-ChatQA-1.5-70B-IQ4_NL.gguf +3 -0
  13. Llama3-ChatQA-1.5-70B-IQ4_XS.gguf +3 -0
  14. Llama3-ChatQA-1.5-70B-Q2_K.gguf +3 -0
  15. Llama3-ChatQA-1.5-70B-Q3_K_L.gguf +3 -0
  16. Llama3-ChatQA-1.5-70B-Q3_K_M.gguf +3 -0
  17. Llama3-ChatQA-1.5-70B-Q3_K_S.gguf +3 -0
  18. Llama3-ChatQA-1.5-70B-Q4_K_M.gguf +3 -0
  19. Llama3-ChatQA-1.5-70B-Q4_K_S.gguf +3 -0
  20. Llama3-ChatQA-1.5-70B-Q5_K_M.gguf +3 -0
  21. Llama3-ChatQA-1.5-70B-Q5_K_S.gguf +3 -0
  22. Llama3-ChatQA-1.5-70B-Q6_K.gguf/Llama3-ChatQA-1.5-70B-Q6_K-00001-of-00002.gguf +3 -0
  23. Llama3-ChatQA-1.5-70B-Q6_K.gguf/Llama3-ChatQA-1.5-70B-Q6_K-00002-of-00002.gguf +3 -0
  24. Llama3-ChatQA-1.5-70B-Q8_0.gguf/Llama3-ChatQA-1.5-70B-Q8_0-00001-of-00003.gguf +3 -0
  25. Llama3-ChatQA-1.5-70B-Q8_0.gguf/Llama3-ChatQA-1.5-70B-Q8_0-00002-of-00003.gguf +3 -0
  26. Llama3-ChatQA-1.5-70B-Q8_0.gguf/Llama3-ChatQA-1.5-70B-Q8_0-00003-of-00003.gguf +3 -0
  27. Llama3-ChatQA-1.5-70B-fp16.gguf/Llama3-ChatQA-1.5-70B-fp16-00001-of-00005.gguf +3 -0
  28. Llama3-ChatQA-1.5-70B-fp16.gguf/Llama3-ChatQA-1.5-70B-fp16-00002-of-00005.gguf +3 -0
  29. Llama3-ChatQA-1.5-70B-fp16.gguf/Llama3-ChatQA-1.5-70B-fp16-00003-of-00005.gguf +3 -0
  30. Llama3-ChatQA-1.5-70B-fp16.gguf/Llama3-ChatQA-1.5-70B-fp16-00004-of-00005.gguf +3 -0
  31. Llama3-ChatQA-1.5-70B-fp16.gguf/Llama3-ChatQA-1.5-70B-fp16-00005-of-00005.gguf +3 -0
  32. Llama3-ChatQA-1.5-70B.imatrix +3 -0
  33. README.md +109 -0
.gitattributes CHANGED
@@ -33,3 +33,34 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ Llama3-ChatQA-1.5-70B-IQ1_M.gguf filter=lfs diff=lfs merge=lfs -text
37
+ Llama3-ChatQA-1.5-70B-IQ1_S.gguf filter=lfs diff=lfs merge=lfs -text
38
+ Llama3-ChatQA-1.5-70B-IQ2_M.gguf filter=lfs diff=lfs merge=lfs -text
39
+ Llama3-ChatQA-1.5-70B-IQ2_S.gguf filter=lfs diff=lfs merge=lfs -text
40
+ Llama3-ChatQA-1.5-70B-IQ2_XS.gguf filter=lfs diff=lfs merge=lfs -text
41
+ Llama3-ChatQA-1.5-70B-IQ2_XXS.gguf filter=lfs diff=lfs merge=lfs -text
42
+ Llama3-ChatQA-1.5-70B-IQ3_M.gguf filter=lfs diff=lfs merge=lfs -text
43
+ Llama3-ChatQA-1.5-70B-IQ3_S.gguf filter=lfs diff=lfs merge=lfs -text
44
+ Llama3-ChatQA-1.5-70B-IQ3_XS.gguf filter=lfs diff=lfs merge=lfs -text
45
+ Llama3-ChatQA-1.5-70B-IQ3_XXS.gguf filter=lfs diff=lfs merge=lfs -text
46
+ Llama3-ChatQA-1.5-70B-IQ4_NL.gguf filter=lfs diff=lfs merge=lfs -text
47
+ Llama3-ChatQA-1.5-70B-IQ4_XS.gguf filter=lfs diff=lfs merge=lfs -text
48
+ Llama3-ChatQA-1.5-70B-Q2_K.gguf filter=lfs diff=lfs merge=lfs -text
49
+ Llama3-ChatQA-1.5-70B-Q3_K_L.gguf filter=lfs diff=lfs merge=lfs -text
50
+ Llama3-ChatQA-1.5-70B-Q3_K_M.gguf filter=lfs diff=lfs merge=lfs -text
51
+ Llama3-ChatQA-1.5-70B-Q3_K_S.gguf filter=lfs diff=lfs merge=lfs -text
52
+ Llama3-ChatQA-1.5-70B-Q4_K_M.gguf filter=lfs diff=lfs merge=lfs -text
53
+ Llama3-ChatQA-1.5-70B-Q4_K_S.gguf filter=lfs diff=lfs merge=lfs -text
54
+ Llama3-ChatQA-1.5-70B-Q5_K_M.gguf filter=lfs diff=lfs merge=lfs -text
55
+ Llama3-ChatQA-1.5-70B-Q5_K_S.gguf filter=lfs diff=lfs merge=lfs -text
56
+ Llama3-ChatQA-1.5-70B-Q6_K.gguf/Llama3-ChatQA-1.5-70B-Q6_K-00001-of-00002.gguf filter=lfs diff=lfs merge=lfs -text
57
+ Llama3-ChatQA-1.5-70B-Q6_K.gguf/Llama3-ChatQA-1.5-70B-Q6_K-00002-of-00002.gguf filter=lfs diff=lfs merge=lfs -text
58
+ Llama3-ChatQA-1.5-70B-Q8_0.gguf/Llama3-ChatQA-1.5-70B-Q8_0-00001-of-00003.gguf filter=lfs diff=lfs merge=lfs -text
59
+ Llama3-ChatQA-1.5-70B-Q8_0.gguf/Llama3-ChatQA-1.5-70B-Q8_0-00002-of-00003.gguf filter=lfs diff=lfs merge=lfs -text
60
+ Llama3-ChatQA-1.5-70B-Q8_0.gguf/Llama3-ChatQA-1.5-70B-Q8_0-00003-of-00003.gguf filter=lfs diff=lfs merge=lfs -text
61
+ Llama3-ChatQA-1.5-70B-fp16.gguf/Llama3-ChatQA-1.5-70B-fp16-00001-of-00005.gguf filter=lfs diff=lfs merge=lfs -text
62
+ Llama3-ChatQA-1.5-70B-fp16.gguf/Llama3-ChatQA-1.5-70B-fp16-00002-of-00005.gguf filter=lfs diff=lfs merge=lfs -text
63
+ Llama3-ChatQA-1.5-70B-fp16.gguf/Llama3-ChatQA-1.5-70B-fp16-00003-of-00005.gguf filter=lfs diff=lfs merge=lfs -text
64
+ Llama3-ChatQA-1.5-70B-fp16.gguf/Llama3-ChatQA-1.5-70B-fp16-00004-of-00005.gguf filter=lfs diff=lfs merge=lfs -text
65
+ Llama3-ChatQA-1.5-70B-fp16.gguf/Llama3-ChatQA-1.5-70B-fp16-00005-of-00005.gguf filter=lfs diff=lfs merge=lfs -text
66
+ Llama3-ChatQA-1.5-70B.imatrix filter=lfs diff=lfs merge=lfs -text
Llama3-ChatQA-1.5-70B-IQ1_M.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6f90c8bb1164be6f2e49782f60db90d11833b97a76fc15c50e6a1cfaa0e80cf1
3
+ size 16751195488
Llama3-ChatQA-1.5-70B-IQ1_S.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9d80d207e9ba272e5679dbb0585fc202900980b32c45f5fcb60bd9c4e44020de
3
+ size 15343482208
Llama3-ChatQA-1.5-70B-IQ2_M.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b095591ccad2cc38f9b08208b576ea943f3d37cb7f325b33e8516f620ae8eb6a
3
+ size 24119293280
Llama3-ChatQA-1.5-70B-IQ2_S.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bc21376b981a49f7495f3aa38eeae107d20a08f109823dece8c30d9e6250022e
3
+ size 22242342240
Llama3-ChatQA-1.5-70B-IQ2_XS.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7c6754db6dd0d4cb1aabd274300ec3fd135b4eef433eeaa5ac62cfe8d9d1dc2b
3
+ size 21142107488
Llama3-ChatQA-1.5-70B-IQ2_XXS.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:757d3eec9ee94289027edd5bee1dc031b53e1dbbf1edbba29ea52a7a3aed9040
3
+ size 19097384288
Llama3-ChatQA-1.5-70B-IQ3_M.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f6c0120aab00c935f91f5581d217b6cf3d7c235cd7c1f3d96a74b90b7a9100fa
3
+ size 31937033568
Llama3-ChatQA-1.5-70B-IQ3_S.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:688a69f03db03793240dcdf9a4b28aeed0283bbfe81f7b6cd3c237f8329802a7
3
+ size 30912050528
Llama3-ChatQA-1.5-70B-IQ3_XS.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dec11bbfa810b13a63a64a55fbeb669ac1c89d0d88576136aa5468220a378a7c
3
+ size 29307729248
Llama3-ChatQA-1.5-70B-IQ3_XXS.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:475df98c1b037a1154b5607036b2891e28190cc466f9799c93e646deea16abc1
3
+ size 27469493600
Llama3-ChatQA-1.5-70B-IQ4_NL.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b9458f9e7ce5064740f2a4401b5db1546faa3d883c4e5856b23db63a44357905
3
+ size 40053618016
Llama3-ChatQA-1.5-70B-IQ4_XS.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fe3da6db448926498dcbf381bc8184a4fa2d5908db79f3ce9f60cf49933e8b08
3
+ size 37902660960
Llama3-ChatQA-1.5-70B-Q2_K.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0db91cc4a649a216be03e38bf8b7cde054cbc430add199b414d094940fc0109e
3
+ size 26375107936
Llama3-ChatQA-1.5-70B-Q3_K_L.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3e9ae30ecbe3d6365e37ec267e5accfa58ba9c945642a98234151ec89e858065
3
+ size 37140591968
Llama3-ChatQA-1.5-70B-Q3_K_M.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:afffa16653a51fc4e256b8eef70930bc415c15aaa971f3a5d4a16620fd5005bb
3
+ size 34267493728
Llama3-ChatQA-1.5-70B-Q3_K_S.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f98a9ee8e4c4b4dc6d3f875349180e35499a2b417799c5532e636e5abf46d075
3
+ size 30912050528
Llama3-ChatQA-1.5-70B-Q4_K_M.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b5e0ad176e775d80c0227a41b05d687371273a8dbeab907fce0ad5a5452ba8af
3
+ size 42520393056
Llama3-ChatQA-1.5-70B-Q4_K_S.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e69de7f9241df7949edd374b40840b04bb0aedb2654f5850ded5ca5420d440a2
3
+ size 40347219296
Llama3-ChatQA-1.5-70B-Q5_K_M.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:184cf9332ed288b3944d64beb6ea76fe159e33ad4cc85d15415d014eecb6ac02
3
+ size 49949816160
Llama3-ChatQA-1.5-70B-Q5_K_S.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:82382b668db6fbb44b68db38827ee8a9beb75b33d779a979bb9bcb600e4e235e
3
+ size 48657446240
Llama3-ChatQA-1.5-70B-Q6_K.gguf/Llama3-ChatQA-1.5-70B-Q6_K-00001-of-00002.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e768b014778062b9d978ac1fc54913ecd944f1ec3253ac2a616e6239d32c60b5
3
+ size 32072395584
Llama3-ChatQA-1.5-70B-Q6_K.gguf/Llama3-ChatQA-1.5-70B-Q6_K-00002-of-00002.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8d181ccb1b373503e7aa5b87f33a657909e9405513da65dfa103976aa3347ae9
3
+ size 25815747296
Llama3-ChatQA-1.5-70B-Q8_0.gguf/Llama3-ChatQA-1.5-70B-Q8_0-00001-of-00003.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2fd1b7361b18fc003cacd4f3fba9d52a0b96096f10e3215382544364ccd70287
3
+ size 32196826624
Llama3-ChatQA-1.5-70B-Q8_0.gguf/Llama3-ChatQA-1.5-70B-Q8_0-00002-of-00003.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:668989aa520b5a6a4c147250e2238ef610a74220ce28a73f21d276dcb2a5e0aa
3
+ size 32070912576
Llama3-ChatQA-1.5-70B-Q8_0.gguf/Llama3-ChatQA-1.5-70B-Q8_0-00003-of-00003.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:75e8cfc5d728372831dd100a09248c12b48d10c09b0aedd1d22ff1d75d4d85be
3
+ size 10707310176
Llama3-ChatQA-1.5-70B-fp16.gguf/Llama3-ChatQA-1.5-70B-fp16-00001-of-00005.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2a0c2b9c95a5e5f38b976b8b33832f50d2f273ae0c75ead9263d3b3a92aab8ea
3
+ size 31973798208
Llama3-ChatQA-1.5-70B-fp16.gguf/Llama3-ChatQA-1.5-70B-fp16-00002-of-00005.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ac83c42f4b69ab6fed05f94fd947f613fc1b4f16e2c3a5cde8cb26d164d2d0f6
3
+ size 32045738016
Llama3-ChatQA-1.5-70B-fp16.gguf/Llama3-ChatQA-1.5-70B-fp16-00003-of-00005.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f7a4bcddcac083111cf67aa7b59ddbbcdad5483a3b8535f85bea94610e3c299d
3
+ size 31743682240
Llama3-ChatQA-1.5-70B-fp16.gguf/Llama3-ChatQA-1.5-70B-fp16-00004-of-00005.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b2083c769332b7d94aa13d9b63a8403a0ee6681dd7d98dba3a9834933cfa3d80
3
+ size 32045738016
Llama3-ChatQA-1.5-70B-fp16.gguf/Llama3-ChatQA-1.5-70B-fp16-00005-of-00005.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a22f84cb85bd60396113ffee18218cdb0e26c8f428f380422f3f47c27f0c8148
3
+ size 13308956160
Llama3-ChatQA-1.5-70B.imatrix ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7ef5b23085789dd69ed26d0be7d4f17af2b2fa18eaacfa83a1550df33597da31
3
+ size 24922294
README.md ADDED
@@ -0,0 +1,109 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: llama3
3
+ language:
4
+ - en
5
+ pipeline_tag: text-generation
6
+ tags:
7
+ - nvidia
8
+ - chatqa-1.5
9
+ - chatqa
10
+ - llama-3
11
+ - pytorch
12
+ quantized_by: bartowski
13
+ ---
14
+
15
+ ## Llamacpp imatrix Quantizations of Llama3-ChatQA-1.5-70B
16
+
17
+ Using <a href="https://github.com/ggerganov/llama.cpp/">llama.cpp</a> release <a href="https://github.com/ggerganov/llama.cpp/releases/tag/b2777">b2777</a> for quantization.
18
+
19
+ Original model: https://huggingface.co/nvidia/Llama3-ChatQA-1.5-70B
20
+
21
+ All quants made using imatrix option with dataset provided by Kalomaze [here](https://github.com/ggerganov/llama.cpp/discussions/5263#discussioncomment-8395384)
22
+
23
+ ## Prompt format
24
+
25
+ No chat template specified so default is used. This may be incorrect, check original model card for details.
26
+
27
+ ```
28
+ <|im_start|>system
29
+ {system_prompt}<|im_end|>
30
+ <|im_start|>user
31
+ {prompt}<|im_end|>
32
+ <|im_start|>assistant
33
+
34
+ ```
35
+
36
+ ## Download a file (not the whole branch) from below:
37
+
38
+ | Filename | Quant type | File Size | Description |
39
+ | -------- | ---------- | --------- | ----------- |
40
+ | [Llama3-ChatQA-1.5-70B-Q8_0.gguf](https://huggingface.co/bartowski/Llama3-ChatQA-1.5-70B-GGUF/tree/main/Llama3-ChatQA-1.5-70B-Q8_0.gguf) | Q8_0 | 74.97GB | Extremely high quality, generally unneeded but max available quant. |
41
+ | [Llama3-ChatQA-1.5-70B-Q6_K.gguf](https://huggingface.co/bartowski/Llama3-ChatQA-1.5-70B-GGUF/tree/main/Llama3-ChatQA-1.5-70B-Q6_K.gguf) | Q6_K | 57.88GB | Very high quality, near perfect, *recommended*. |
42
+ | [Llama3-ChatQA-1.5-70B-Q5_K_M.gguf](https://huggingface.co/bartowski/Llama3-ChatQA-1.5-70B-GGUF/blob/main/Llama3-ChatQA-1.5-70B-Q5_K_M.gguf) | Q5_K_M | 49.94GB | High quality, *recommended*. |
43
+ | [Llama3-ChatQA-1.5-70B-Q5_K_S.gguf](https://huggingface.co/bartowski/Llama3-ChatQA-1.5-70B-GGUF/blob/main/Llama3-ChatQA-1.5-70B-Q5_K_S.gguf) | Q5_K_S | 48.65GB | High quality, *recommended*. |
44
+ | [Llama3-ChatQA-1.5-70B-Q4_K_M.gguf](https://huggingface.co/bartowski/Llama3-ChatQA-1.5-70B-GGUF/blob/main/Llama3-ChatQA-1.5-70B-Q4_K_M.gguf) | Q4_K_M | 42.52GB | Good quality, uses about 4.83 bits per weight, *recommended*. |
45
+ | [Llama3-ChatQA-1.5-70B-Q4_K_S.gguf](https://huggingface.co/bartowski/Llama3-ChatQA-1.5-70B-GGUF/blob/main/Llama3-ChatQA-1.5-70B-Q4_K_S.gguf) | Q4_K_S | 40.34GB | Slightly lower quality with more space savings, *recommended*. |
46
+ | [Llama3-ChatQA-1.5-70B-IQ4_NL.gguf](https://huggingface.co/bartowski/Llama3-ChatQA-1.5-70B-GGUF/blob/main/Llama3-ChatQA-1.5-70B-IQ4_NL.gguf) | IQ4_NL | 40.05GB | Decent quality, slightly smaller than Q4_K_S with similar performance *recommended*. |
47
+ | [Llama3-ChatQA-1.5-70B-IQ4_XS.gguf](https://huggingface.co/bartowski/Llama3-ChatQA-1.5-70B-GGUF/blob/main/Llama3-ChatQA-1.5-70B-IQ4_XS.gguf) | IQ4_XS | 37.90GB | Decent quality, smaller than Q4_K_S with similar performance, *recommended*. |
48
+ | [Llama3-ChatQA-1.5-70B-Q3_K_L.gguf](https://huggingface.co/bartowski/Llama3-ChatQA-1.5-70B-GGUF/blob/main/Llama3-ChatQA-1.5-70B-Q3_K_L.gguf) | Q3_K_L | 37.14GB | Lower quality but usable, good for low RAM availability. |
49
+ | [Llama3-ChatQA-1.5-70B-Q3_K_M.gguf](https://huggingface.co/bartowski/Llama3-ChatQA-1.5-70B-GGUF/blob/main/Llama3-ChatQA-1.5-70B-Q3_K_M.gguf) | Q3_K_M | 34.26GB | Even lower quality. |
50
+ | [Llama3-ChatQA-1.5-70B-IQ3_M.gguf](https://huggingface.co/bartowski/Llama3-ChatQA-1.5-70B-GGUF/blob/main/Llama3-ChatQA-1.5-70B-IQ3_M.gguf) | IQ3_M | 31.93GB | Medium-low quality, new method with decent performance comparable to Q3_K_M. |
51
+ | [Llama3-ChatQA-1.5-70B-IQ3_S.gguf](https://huggingface.co/bartowski/Llama3-ChatQA-1.5-70B-GGUF/blob/main/Llama3-ChatQA-1.5-70B-IQ3_S.gguf) | IQ3_S | 30.91GB | Lower quality, new method with decent performance, recommended over Q3_K_S quant, same size with better performance. |
52
+ | [Llama3-ChatQA-1.5-70B-Q3_K_S.gguf](https://huggingface.co/bartowski/Llama3-ChatQA-1.5-70B-GGUF/blob/main/Llama3-ChatQA-1.5-70B-Q3_K_S.gguf) | Q3_K_S | 30.91GB | Low quality, not recommended. |
53
+ | [Llama3-ChatQA-1.5-70B-IQ3_XS.gguf](https://huggingface.co/bartowski/Llama3-ChatQA-1.5-70B-GGUF/blob/main/Llama3-ChatQA-1.5-70B-IQ3_XS.gguf) | IQ3_XS | 29.30GB | Lower quality, new method with decent performance, slightly better than Q3_K_S. |
54
+ | [Llama3-ChatQA-1.5-70B-IQ3_XXS.gguf](https://huggingface.co/bartowski/Llama3-ChatQA-1.5-70B-GGUF/blob/main/Llama3-ChatQA-1.5-70B-IQ3_XXS.gguf) | IQ3_XXS | 27.46GB | Lower quality, new method with decent performance, comparable to Q3 quants. |
55
+ | [Llama3-ChatQA-1.5-70B-Q2_K.gguf](https://huggingface.co/bartowski/Llama3-ChatQA-1.5-70B-GGUF/blob/main/Llama3-ChatQA-1.5-70B-Q2_K.gguf) | Q2_K | 26.37GB | Very low quality but surprisingly usable. |
56
+ | [Llama3-ChatQA-1.5-70B-IQ2_M.gguf](https://huggingface.co/bartowski/Llama3-ChatQA-1.5-70B-GGUF/blob/main/Llama3-ChatQA-1.5-70B-IQ2_M.gguf) | IQ2_M | 24.11GB | Very low quality, uses SOTA techniques to also be surprisingly usable. |
57
+ | [Llama3-ChatQA-1.5-70B-IQ2_S.gguf](https://huggingface.co/bartowski/Llama3-ChatQA-1.5-70B-GGUF/blob/main/Llama3-ChatQA-1.5-70B-IQ2_S.gguf) | IQ2_S | 22.24GB | Very low quality, uses SOTA techniques to be usable. |
58
+ | [Llama3-ChatQA-1.5-70B-IQ2_XS.gguf](https://huggingface.co/bartowski/Llama3-ChatQA-1.5-70B-GGUF/blob/main/Llama3-ChatQA-1.5-70B-IQ2_XS.gguf) | IQ2_XS | 21.14GB | Very low quality, uses SOTA techniques to be usable. |
59
+ | [Llama3-ChatQA-1.5-70B-IQ2_XXS.gguf](https://huggingface.co/bartowski/Llama3-ChatQA-1.5-70B-GGUF/blob/main/Llama3-ChatQA-1.5-70B-IQ2_XXS.gguf) | IQ2_XXS | 19.09GB | Lower quality, uses SOTA techniques to be usable. |
60
+ | [Llama3-ChatQA-1.5-70B-IQ1_M.gguf](https://huggingface.co/bartowski/Llama3-ChatQA-1.5-70B-GGUF/blob/main/Llama3-ChatQA-1.5-70B-IQ1_M.gguf) | IQ1_M | 16.75GB | Extremely low quality, *not* recommended. |
61
+ | [Llama3-ChatQA-1.5-70B-IQ1_S.gguf](https://huggingface.co/bartowski/Llama3-ChatQA-1.5-70B-GGUF/blob/main/Llama3-ChatQA-1.5-70B-IQ1_S.gguf) | IQ1_S | 15.34GB | Extremely low quality, *not* recommended. |
62
+
63
+ ## Downloading using huggingface-cli
64
+
65
+ First, make sure you have hugginface-cli installed:
66
+
67
+ ```
68
+ pip install -U "huggingface_hub[cli]"
69
+ ```
70
+
71
+ Then, you can target the specific file you want:
72
+
73
+ ```
74
+ huggingface-cli download bartowski/Llama3-ChatQA-1.5-70B-GGUF --include "Llama3-ChatQA-1.5-70B-Q4_K_M.gguf" --local-dir ./ --local-dir-use-symlinks False
75
+ ```
76
+
77
+ If the model is bigger than 50GB, it will have been split into multiple files. In order to download them all to a local folder, run:
78
+
79
+ ```
80
+ huggingface-cli download bartowski/Llama3-ChatQA-1.5-70B-GGUF --include "Llama3-ChatQA-1.5-70B-Q8_0.gguf/*" --local-dir Llama3-ChatQA-1.5-70B-Q8_0 --local-dir-use-symlinks False
81
+ ```
82
+
83
+ You can either specify a new local-dir (Llama3-ChatQA-1.5-70B-Q8_0) or download them all in place (./)
84
+
85
+ ## Which file should I choose?
86
+
87
+ A great write up with charts showing various performances is provided by Artefact2 [here](https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9)
88
+
89
+ The first thing to figure out is how big a model you can run. To do this, you'll need to figure out how much RAM and/or VRAM you have.
90
+
91
+ If you want your model running as FAST as possible, you'll want to fit the whole thing on your GPU's VRAM. Aim for a quant with a file size 1-2GB smaller than your GPU's total VRAM.
92
+
93
+ If you want the absolute maximum quality, add both your system RAM and your GPU's VRAM together, then similarly grab a quant with a file size 1-2GB Smaller than that total.
94
+
95
+ Next, you'll need to decide if you want to use an 'I-quant' or a 'K-quant'.
96
+
97
+ If you don't want to think too much, grab one of the K-quants. These are in format 'QX_K_X', like Q5_K_M.
98
+
99
+ If you want to get more into the weeds, you can check out this extremely useful feature chart:
100
+
101
+ [llama.cpp feature matrix](https://github.com/ggerganov/llama.cpp/wiki/Feature-matrix)
102
+
103
+ But basically, if you're aiming for below Q4, and you're running cuBLAS (Nvidia) or rocBLAS (AMD), you should look towards the I-quants. These are in format IQX_X, like IQ3_M. These are newer and offer better performance for their size.
104
+
105
+ These I-quants can also be used on CPU and Apple Metal, but will be slower than their K-quant equivalent, so speed vs performance is a tradeoff you'll have to decide.
106
+
107
+ The I-quants are *not* compatible with Vulcan, which is also AMD, so if you have an AMD card double check if you're using the rocBLAS build or the Vulcan build. At the time of writing this, LM Studio has a preview with ROCm support, and other inference engines have specific builds for ROCm.
108
+
109
+ Want to support my work? Visit my ko-fi page here: https://ko-fi.com/bartowski