Add 13B files
Browse files- .gitignore +5 -0
- Makefile +48 -0
- README.md +10 -0
- SHA256SUMS +18 -0
- open-llama-13b-f16.bin +3 -0
- open-llama-13b-q2_K.bin +3 -0
- open-llama-13b-q3_K.bin +3 -0
- open-llama-13b-q3_K_L.bin +3 -0
- open-llama-13b-q3_K_M.bin +3 -0
- open-llama-13b-q3_K_S.bin +3 -0
- open-llama-13b-q4_0.bin +3 -0
- open-llama-13b-q4_1.bin +3 -0
- open-llama-13b-q4_K.bin +3 -0
- open-llama-13b-q4_K_M.bin +3 -0
- open-llama-13b-q4_K_S.bin +3 -0
- open-llama-13b-q5_0.bin +3 -0
- open-llama-13b-q5_1.bin +3 -0
- open-llama-13b-q5_K.bin +3 -0
- open-llama-13b-q5_K_M.bin +3 -0
- open-llama-13b-q5_K_S.bin +3 -0
- open-llama-13b-q6_K.bin +3 -0
- open-llama-13b-q8_0.bin +3 -0
.gitignore
ADDED
@@ -0,0 +1,5 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
llama.cpp/
|
2 |
+
pytorch_model*.bin
|
3 |
+
*.sha
|
4 |
+
*.tar.gz
|
5 |
+
tokenizer.model
|
Makefile
ADDED
@@ -0,0 +1,48 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
MODEL_NAME= open-llama-13b
|
2 |
+
PYTHON?= python
|
3 |
+
LLAMA_TAG= 5c64a09
|
4 |
+
LLAMA_TAR= master-$(LLAMA_TAG).tar.gz
|
5 |
+
HF_REPO= openlm-research/open_llama_13b_600bt
|
6 |
+
HF_REF= main
|
7 |
+
HF_FILES= pytorch_model-00001-of-00003.bin \
|
8 |
+
pytorch_model-00002-of-00003.bin \
|
9 |
+
pytorch_model-00003-of-00003.bin \
|
10 |
+
tokenizer.model
|
11 |
+
$(HF_FILES): SITE= https://huggingface.co/$(HF_REPO)/resolve/$(HF_REF)
|
12 |
+
$(LLAMA_TAR): SITE= https://github.com/ggerganov/llama.cpp/archive/refs/tags
|
13 |
+
|
14 |
+
FILES= $(HF_FILES) $(LLAMA_TAR)
|
15 |
+
|
16 |
+
QUANTS= f16 q4_0 q4_1 q5_0 q5_1 q8_0 \
|
17 |
+
q2_K \
|
18 |
+
q3_K q3_K_L q3_K_M q3_K_S \
|
19 |
+
q4_K q4_K_M q4_K_S \
|
20 |
+
q5_K q5_K_M q5_K_S \
|
21 |
+
q6_K
|
22 |
+
|
23 |
+
MODEL_FILES= $(addsuffix .bin,$(addprefix $(MODEL_NAME)-,$(QUANTS)))
|
24 |
+
|
25 |
+
.PHONY: all
|
26 |
+
all: $(MODEL_FILES) SHA256SUMS
|
27 |
+
|
28 |
+
$(FILES):
|
29 |
+
curl -L -o $@ --url $(SITE)/$@
|
30 |
+
|
31 |
+
llama.cpp: $(LLAMA_TAR)
|
32 |
+
mkdir -p $@
|
33 |
+
tar -xf $< --strip-components=1 -C $@
|
34 |
+
|
35 |
+
llama.cpp/quantize: llama.cpp
|
36 |
+
$(MAKE) -C llama.cpp quantize
|
37 |
+
|
38 |
+
$(MODEL_NAME)-f16.bin: $(HF_FILES) | llama.cpp
|
39 |
+
$(PYTHON) llama.cpp/convert.py --outtype f16 --outfile $@ .
|
40 |
+
|
41 |
+
$(MODEL_NAME)-q%.bin: $(MODEL_NAME)-f16.bin | llama.cpp/quantize
|
42 |
+
llama.cpp/quantize $< $@ q$*
|
43 |
+
|
44 |
+
%.sha: %
|
45 |
+
sha256sum $< > $@
|
46 |
+
|
47 |
+
SHA256SUMS: $(addsuffix .sha,$(MODEL_FILES))
|
48 |
+
cat $^ > $@
|
README.md
CHANGED
@@ -1,3 +1,13 @@
|
|
1 |
---
|
2 |
license: apache-2.0
|
3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
---
|
2 |
license: apache-2.0
|
3 |
---
|
4 |
+
|
5 |
+
# ggml versions of OpenLLaMa 13B
|
6 |
+
|
7 |
+
For use with [llama.cpp](https://github.com/ggerganov/llama.cpp).
|
8 |
+
|
9 |
+
- Version: 600B tokens preview
|
10 |
+
- Project: [OpenLLaMA: An Open Reproduction of LLaMA](https://github.com/openlm-research/open_llama)
|
11 |
+
- Model: [openlm-research/open_llama_13b_600bt](https://huggingface.co/openlm-research/open_llama_13b_600bt)
|
12 |
+
- llama.cpp 4,5,8-bit quantization: build 567(2d5db48) or later
|
13 |
+
- llama.cpp newer quantization formats: build 616(99009e7) or later
|
SHA256SUMS
ADDED
@@ -0,0 +1,18 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
5700b51c19b5c8a8841ee78c1995ea7aaa05ce9276e6cf40e7291dd00d862206 open-llama-13b-f16.bin
|
2 |
+
4141dc47a661d3c7414f3ae7c17b2e95c44edf8317e4ff06701859f913dd5ed5 open-llama-13b-q4_0.bin
|
3 |
+
de8da3f52b04f5e848aabcb4fce9369254cce13794a9146f3b33b16dbaaec9b8 open-llama-13b-q4_1.bin
|
4 |
+
8735c580c831d1f3227dadfbcb0c8ed2fb9048e52d1065c3674f4a561bf8b512 open-llama-13b-q5_0.bin
|
5 |
+
d810d6729199a85d1a79dc2893164ea9f8267a29a710229c016da83eeec7f7b7 open-llama-13b-q5_1.bin
|
6 |
+
60a2931fd78274abd55bad36a77661caded910d67faad8fadde4e9454ea95a76 open-llama-13b-q8_0.bin
|
7 |
+
14c676097bcc6e712f65ce82cd44149171c89c2970d68ca229d1dc3171696348 open-llama-13b-q2_K.bin
|
8 |
+
b357b384d55737bacd9780beee038836333cc53801f81e698ec2ffc5f19036d4 open-llama-13b-q3_K.bin
|
9 |
+
e2651804d017e0ceeeacf3dff3ab30c300609c45471486f098eafbb2d4d2c3d3 open-llama-13b-q3_K_L.bin
|
10 |
+
b357b384d55737bacd9780beee038836333cc53801f81e698ec2ffc5f19036d4 open-llama-13b-q3_K_M.bin
|
11 |
+
b05eff228b92aac06a9cbb0e5191980b02e9ede3d304c16de2622b88cefcff76 open-llama-13b-q3_K_S.bin
|
12 |
+
436cce492e2210d7c8e6db1e06878e74fa6d8097e1ac0d7da4b53db66a6c507f open-llama-13b-q4_K.bin
|
13 |
+
436cce492e2210d7c8e6db1e06878e74fa6d8097e1ac0d7da4b53db66a6c507f open-llama-13b-q4_K_M.bin
|
14 |
+
4994ce0a7ff919de7fe4c88555ae83e4c351e142eaf502bb50df4e4f070807f7 open-llama-13b-q4_K_S.bin
|
15 |
+
5e9e0a4033ecd8d4d4f8aaf16dec1d4cc155accd951f49eded2b642b006dd4d3 open-llama-13b-q5_K.bin
|
16 |
+
5e9e0a4033ecd8d4d4f8aaf16dec1d4cc155accd951f49eded2b642b006dd4d3 open-llama-13b-q5_K_M.bin
|
17 |
+
8f72df066cefe8c969e3d51cd71c2fe2a600161e104285d45a6e6fa60940cd0c open-llama-13b-q5_K_S.bin
|
18 |
+
9e1032da4941e15b90db4c5589660490aaa193cad985079ca10bfa99b3698532 open-llama-13b-q6_K.bin
|
open-llama-13b-f16.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5700b51c19b5c8a8841ee78c1995ea7aaa05ce9276e6cf40e7291dd00d862206
|
3 |
+
size 26033037952
|
open-llama-13b-q2_K.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:14c676097bcc6e712f65ce82cd44149171c89c2970d68ca229d1dc3171696348
|
3 |
+
size 5427905152
|
open-llama-13b-q3_K.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b357b384d55737bacd9780beee038836333cc53801f81e698ec2ffc5f19036d4
|
3 |
+
size 6249255552
|
open-llama-13b-q3_K_L.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e2651804d017e0ceeeacf3dff3ab30c300609c45471486f098eafbb2d4d2c3d3
|
3 |
+
size 6865293952
|
open-llama-13b-q3_K_M.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b357b384d55737bacd9780beee038836333cc53801f81e698ec2ffc5f19036d4
|
3 |
+
size 6249255552
|
open-llama-13b-q3_K_S.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b05eff228b92aac06a9cbb0e5191980b02e9ede3d304c16de2622b88cefcff76
|
3 |
+
size 5594714752
|
open-llama-13b-q4_0.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:4141dc47a661d3c7414f3ae7c17b2e95c44edf8317e4ff06701859f913dd5ed5
|
3 |
+
size 7323329152
|
open-llama-13b-q4_1.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:de8da3f52b04f5e848aabcb4fce9369254cce13794a9146f3b33b16dbaaec9b8
|
3 |
+
size 8136794752
|
open-llama-13b-q4_K.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:436cce492e2210d7c8e6db1e06878e74fa6d8097e1ac0d7da4b53db66a6c507f
|
3 |
+
size 7823450752
|
open-llama-13b-q4_K_M.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:436cce492e2210d7c8e6db1e06878e74fa6d8097e1ac0d7da4b53db66a6c507f
|
3 |
+
size 7823450752
|
open-llama-13b-q4_K_S.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:4994ce0a7ff919de7fe4c88555ae83e4c351e142eaf502bb50df4e4f070807f7
|
3 |
+
size 7323329152
|
open-llama-13b-q5_0.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:12ba030dec8f6868629d7e7d53d58ed3b930ce0b7128f571a9892ddfb55d4d12
|
3 |
+
size 8950260352
|
open-llama-13b-q5_1.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:d810d6729199a85d1a79dc2893164ea9f8267a29a710229c016da83eeec7f7b7
|
3 |
+
size 9763725952
|
open-llama-13b-q5_K.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5e9e0a4033ecd8d4d4f8aaf16dec1d4cc155accd951f49eded2b642b006dd4d3
|
3 |
+
size 9207898752
|
open-llama-13b-q5_K_M.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5e9e0a4033ecd8d4d4f8aaf16dec1d4cc155accd951f49eded2b642b006dd4d3
|
3 |
+
size 9207898752
|
open-llama-13b-q5_K_S.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:8f72df066cefe8c969e3d51cd71c2fe2a600161e104285d45a6e6fa60940cd0c
|
3 |
+
size 8950260352
|
open-llama-13b-q6_K.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:9e1032da4941e15b90db4c5589660490aaa193cad985079ca10bfa99b3698532
|
3 |
+
size 10678874752
|
open-llama-13b-q8_0.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:60a2931fd78274abd55bad36a77661caded910d67faad8fadde4e9454ea95a76
|
3 |
+
size 13831053952
|