Hugging Face's logo

SlyEcho
/
open_llama_3b_gguf

GGUF
Inference Endpoints
Model card Files Community
SlyEcho commited on
Commit
7b9d996
·
verified ·
1 Parent(s): 18df7cf

add files

Browse files
Files changed (11) hide show
  1. .gitattributes +1 -0
  2. .gitignore +8 -0
  3. Makefile +48 -0
  4. README.md +17 -0
  5. SHA256SUMS +6 -0
  6. open-llama-3b-f16.gguf +3 -0
  7. open-llama-3b-q4_0.gguf +3 -0
  8. open-llama-3b-q4_1.gguf +3 -0
  9. open-llama-3b-q5_0.gguf +3 -0
  10. open-llama-3b-q5_1.gguf +3 -0
  11. open-llama-3b-q8_0.gguf +3 -0
.gitattributes CHANGED
@@ -4,6 +4,7 @@
4
  *.bz2 filter=lfs diff=lfs merge=lfs -text
5
  *.ckpt filter=lfs diff=lfs merge=lfs -text
6
  *.ftz filter=lfs diff=lfs merge=lfs -text
 
7
  *.gz filter=lfs diff=lfs merge=lfs -text
8
  *.h5 filter=lfs diff=lfs merge=lfs -text
9
  *.joblib filter=lfs diff=lfs merge=lfs -text
 
4
  *.bz2 filter=lfs diff=lfs merge=lfs -text
5
  *.ckpt filter=lfs diff=lfs merge=lfs -text
6
  *.ftz filter=lfs diff=lfs merge=lfs -text
7
+ *.gguf filter=lfs diff=lfs merge=lfs -text
8
  *.gz filter=lfs diff=lfs merge=lfs -text
9
  *.h5 filter=lfs diff=lfs merge=lfs -text
10
  *.joblib filter=lfs diff=lfs merge=lfs -text
.gitignore ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ llama.cpp*/
2
+ venv/
3
+ pytorch_model.bin
4
+ *.sha
5
+ *.tar.gz
6
+ tokenizer.model
7
+ config.json
8
+ tokenizer_config.json
Makefile ADDED
@@ -0,0 +1,48 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ MODEL_NAME= open-llama-3b
2
+ PYTHON?= python
3
+ LLAMA_BUILD= 1132
4
+ LLAMA_TAR= b$(LLAMA_BUILD).tar.gz
5
+ LLAMA_DIR= llama.cpp-b$(LLAMA_BUILD)
6
+ LLAMA_FLAGS= LLAMA_NO_K_QUANTS=1
7
+ HF_REPO= openlm-research/open_llama_3b
8
+ HF_REF= main
9
+ HF_FILES= pytorch_model.bin \
10
+ tokenizer.model \
11
+ config.json \
12
+ tokenizer_config.json
13
+ $(HF_FILES): SITE= https://huggingface.co/$(HF_REPO)/resolve/$(HF_REF)
14
+ $(LLAMA_TAR): SITE= https://github.com/ggerganov/llama.cpp/archive/refs/tags
15
+
16
+ QUANTS= f16 q4_0 q4_1 q5_0 q5_1 q8_0
17
+
18
+ FILES= $(HF_FILES) $(LLAMA_TAR)
19
+ MODEL_FILES= $(foreach q,$(QUANTS),$(MODEL_NAME)-$(q).gguf)
20
+
21
+ .PHONY: all
22
+ all: $(MODEL_FILES) SHA256SUMS
23
+
24
+ $(FILES):
25
+ curl -L -o $@ --url $(SITE)/$@
26
+
27
+ $(LLAMA_DIR): | $(LLAMA_TAR)
28
+ tar -xf $(LLAMA_TAR)
29
+
30
+ $(LLAMA_DIR)/quantize: | $(LLAMA_DIR)
31
+ $(MAKE) -C $(LLAMA_DIR) $(LLAMA_FLAGS) quantize
32
+
33
+ venv:
34
+ $(PYTHON) -m venv venv
35
+ venv/bin/pip install -e $(LLAMA_DIR)/gguf-py
36
+ venv/bin/pip install -r $(LLAMA_DIR)/requirements.txt
37
+
38
+ $(MODEL_NAME)-f16.gguf: $(HF_FILES) | $(LLAMA_DIR) venv
39
+ venv/bin/python $(LLAMA_DIR)/convert.py --outtype f16 --outfile $@ .
40
+
41
+ $(MODEL_NAME)-q%.gguf: $(MODEL_NAME)-f16.gguf $(LLAMA_DIR)/quantize
42
+ $(LLAMA_DIR)/quantize $< $@ q$*
43
+
44
+ %.sha: %
45
+ sha256sum $< > $@
46
+
47
+ SHA256SUMS: $(addsuffix .sha,$(MODEL_FILES))
48
+ cat $^ > $@
README.md CHANGED
@@ -1,3 +1,20 @@
1
  ---
2
  license: apache-2.0
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
  license: apache-2.0
3
  ---
4
+
5
+ # gguf versions of OpenLLaMa 3B
6
+
7
+ - Version: 1T tokens final version
8
+ - Project: [OpenLLaMA: An Open Reproduction of LLaMA](https://github.com/openlm-research/open_llama)
9
+ - Model: [openlm-research/open_llama_3b](https://huggingface.co/openlm-research/open_llama_3b)
10
+ - [llama.cpp](https://github.com/ggerganov/llama.cpp): build 1012 (6381d4e) or later
11
+ - [ggml version](https://huggingface.co/SlyEcho/open_llama_3b_ggml)
12
+
13
+ ## Newer quantizations
14
+
15
+ There are now more quantization types in llama.cpp, some lower than 4 bits.
16
+ Currently these are not supported, maybe because some weights have shapes that don't divide by 256.
17
+
18
+ ## Perplexity on wiki.test.406
19
+
20
+ Coming soon...
SHA256SUMS ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ 589e21251921f052d542a112d7a766056ff2f2ed6d1951677890cd5313a9eaec open-llama-3b-f16.gguf
2
+ 9a496b3352a7153be348279eb9205901524b71659e1b8700217d61b24af16b68 open-llama-3b-q4_0.gguf
3
+ cd3801fc09621097494220333182552847963ff216b3b8c91d3cb5528325f889 open-llama-3b-q4_1.gguf
4
+ eebd692cbaca097d7ea3dcc460ebdeb7c762598004bdc31c663b0d77e0a01d9c open-llama-3b-q5_0.gguf
5
+ f0ef8bd8d3ef4fde9e8ac5610763893635e9d6c26c9b26df27b68012fb4c7424 open-llama-3b-q5_1.gguf
6
+ c48a108d99a5f07dcfbe7c10a40ee0f0663fedf1b171a49f254fb8d767258c19 open-llama-3b-q8_0.gguf
open-llama-3b-f16.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:589e21251921f052d542a112d7a766056ff2f2ed6d1951677890cd5313a9eaec
3
+ size 6854058688
open-llama-3b-q4_0.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9a496b3352a7153be348279eb9205901524b71659e1b8700217d61b24af16b68
3
+ size 1928746720
open-llama-3b-q4_1.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cd3801fc09621097494220333182552847963ff216b3b8c91d3cb5528325f889
3
+ size 2142890720
open-llama-3b-q5_0.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:eebd692cbaca097d7ea3dcc460ebdeb7c762598004bdc31c663b0d77e0a01d9c
3
+ size 2357034720
open-llama-3b-q5_1.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f0ef8bd8d3ef4fde9e8ac5610763893635e9d6c26c9b26df27b68012fb4c7424
3
+ size 2571178720
open-llama-3b-q8_0.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c48a108d99a5f07dcfbe7c10a40ee0f0663fedf1b171a49f254fb8d767258c19
3
+ size 3641898720