Hugging Face's logo

ngquocvinh
/
EmbeddingGemma-2-GGUF

Feature Extraction
GGUF
sentence-transformers
llama.cpp
embedding
multimodal
multilingual
quantized
imatrix
Model card Files
xet
 Community
ngquocvinh commited on
Commit
87aaa37
·
verified ·
1 Parent(s): 3f83ace

Add Q2_K after CPU embedding smoke pass

Browse files
Files changed (5) hide show
  1. .gitattributes +1 -0
  2. EmbeddingGemma-2-Q2_K.gguf +3 -0
  3. SHA256SUMS.txt +1 -0
  4. reproducibility/artifact-measurements.tsv +1 -0
  5. reproducibility/manifest.md +8 -0
.gitattributes CHANGED
@@ -39,3 +39,4 @@ EmbeddingGemma-2-Q6_K.gguf filter=lfs diff=lfs merge=lfs -text
39
  EmbeddingGemma-2-Q5_K_M.gguf filter=lfs diff=lfs merge=lfs -text
40
  EmbeddingGemma-2-Q4_K_M.gguf filter=lfs diff=lfs merge=lfs -text
41
  EmbeddingGemma-2-Q3_K_M.gguf filter=lfs diff=lfs merge=lfs -text
 
 
39
  EmbeddingGemma-2-Q5_K_M.gguf filter=lfs diff=lfs merge=lfs -text
40
  EmbeddingGemma-2-Q4_K_M.gguf filter=lfs diff=lfs merge=lfs -text
41
  EmbeddingGemma-2-Q3_K_M.gguf filter=lfs diff=lfs merge=lfs -text
42
+ EmbeddingGemma-2-Q2_K.gguf filter=lfs diff=lfs merge=lfs -text
EmbeddingGemma-2-Q2_K.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:67124cbba678cce4664bf9af72839a88ab12a9083821545c56847f41a54e96a1
3
+ size 120394368
SHA256SUMS.txt CHANGED
@@ -1,3 +1,4 @@
 
1
  a5c902b15458c4886d6f1296bad71fd8084d7ac626cb8713778cf2af5e1b5a71 EmbeddingGemma-2-Q3_K_M.gguf
2
  ea25550b050fa8826707c5e44659a7c18259606dcc6ba449f771e3c75cf4e663 EmbeddingGemma-2-Q4_K_M.gguf
3
  22f79226a02bfa2dd840414f342a004f93b3cb85e498923a5e089f03f6410b1f EmbeddingGemma-2-Q5_K_M.gguf
 
1
+ 67124cbba678cce4664bf9af72839a88ab12a9083821545c56847f41a54e96a1 EmbeddingGemma-2-Q2_K.gguf
2
  a5c902b15458c4886d6f1296bad71fd8084d7ac626cb8713778cf2af5e1b5a71 EmbeddingGemma-2-Q3_K_M.gguf
3
  ea25550b050fa8826707c5e44659a7c18259606dcc6ba449f771e3c75cf4e663 EmbeddingGemma-2-Q4_K_M.gguf
4
  22f79226a02bfa2dd840414f342a004f93b3cb85e498923a5e089f03f6410b1f EmbeddingGemma-2-Q5_K_M.gguf
reproducibility/artifact-measurements.tsv CHANGED
@@ -5,3 +5,4 @@ EmbeddingGemma-2-Q6_K.gguf 245764736 0.245765 PASS_EMBEDDING_CPU
5
  EmbeddingGemma-2-Q5_K_M.gguf 212759168 0.212759 PASS_EMBEDDING_CPU
6
  EmbeddingGemma-2-Q4_K_M.gguf 181695104 0.181695 PASS_EMBEDDING_CPU
7
  EmbeddingGemma-2-Q3_K_M.gguf 148869760 0.148870 PASS_EMBEDDING_CPU
 
 
5
  EmbeddingGemma-2-Q5_K_M.gguf 212759168 0.212759 PASS_EMBEDDING_CPU
6
  EmbeddingGemma-2-Q4_K_M.gguf 181695104 0.181695 PASS_EMBEDDING_CPU
7
  EmbeddingGemma-2-Q3_K_M.gguf 148869760 0.148870 PASS_EMBEDDING_CPU
8
+ EmbeddingGemma-2-Q2_K.gguf 120394368 0.120394 PASS_EMBEDDING_CPU
reproducibility/manifest.md CHANGED
@@ -62,3 +62,11 @@ The fixed task evaluation uses the full MTEB STSBenchmark `test` split and the u
62
  - Quantization command: `llama-quantize --imatrix calibration/embeddinggemma-2/embeddinggemma-2-sts.imatrix.gguf embeddinggemma-2-text-BF16.gguf EmbeddingGemma-2-Q3_K_M.gguf Q3_K_M 8`.
63
  - CPU `llama-server --embedding --pooling mean -ngl 0` smoke passed with two finite normalized 768-dimensional text vectors.
64
  - Build cgroup: `MemoryMax=32G`, `MemorySwapMax=0`; batch peak memory `878006272` bytes and swap peak `0` bytes.
 
 
 
 
 
 
 
 
 
62
  - Quantization command: `llama-quantize --imatrix calibration/embeddinggemma-2/embeddinggemma-2-sts.imatrix.gguf embeddinggemma-2-text-BF16.gguf EmbeddingGemma-2-Q3_K_M.gguf Q3_K_M 8`.
63
  - CPU `llama-server --embedding --pooling mean -ngl 0` smoke passed with two finite normalized 768-dimensional text vectors.
64
  - Build cgroup: `MemoryMax=32G`, `MemorySwapMax=0`; batch peak memory `878006272` bytes and swap peak `0` bytes.
65
+ - Q3_K_M public Hub commit: `3f83acea1d4ad660992166889a4a54749b980ce3`; remote LFS SHA256 and size match the local file.
66
+
67
+ ## Incremental publish: Q2_K
68
+
69
+ - Artifact: `EmbeddingGemma-2-Q2_K.gguf`; SHA256 `67124cbba678cce4664bf9af72839a88ab12a9083821545c56847f41a54e96a1`; decimal size `0.120394 GB`.
70
+ - Quantization command: `llama-quantize --imatrix calibration/embeddinggemma-2/embeddinggemma-2-sts.imatrix.gguf embeddinggemma-2-text-BF16.gguf EmbeddingGemma-2-Q2_K.gguf Q2_K 8`.
71
+ - CPU `llama-server --embedding --pooling mean -ngl 0` smoke passed with two finite normalized 768-dimensional text vectors.
72
+ - Build cgroup: `MemoryMax=32G`, `MemorySwapMax=0`; batch peak memory `834936832` bytes and swap peak `0` bytes.