Hugging Face's logo

lm-kit
/
embeddinggemma-2-lmk

Feature Extraction
GGUF
multilingual
lm-kit
embedding
sentence-similarity
multimodal-embedding
lmk
Model card Files
xet
 Community
lcarrere commited on
Commit
3ea3e57
·
verified ·
1 Parent(s): 1dde1ca

Merge the GGUF files (text encoder, Q8_0 and BF16 vision encoders) into the LMK repository

Browse files
Files changed (5) hide show
  1. .gitattributes +3 -0
  2. README.md +6 -2
  3. embeddinggemma-2-Q5_K_M.gguf +3 -0
  4. embeddinggemma-2-mmproj-vision-BF16.gguf +3 -0
  5. embeddinggemma-2-mmproj-vision-Q8_0.gguf +3 -0
.gitattributes CHANGED
@@ -34,3 +34,6 @@ saved_model/**/* 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
  embeddinggemma-2-Q5_K_M.lmk 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
  embeddinggemma-2-Q5_K_M.lmk filter=lfs diff=lfs merge=lfs -text
37
+ embeddinggemma-2-mmproj-vision-BF16.gguf filter=lfs diff=lfs merge=lfs -text
38
+ embeddinggemma-2-mmproj-vision-Q8_0.gguf filter=lfs diff=lfs merge=lfs -text
39
+ embeddinggemma-2-Q5_K_M.gguf filter=lfs diff=lfs merge=lfs -text
README.md CHANGED
@@ -8,6 +8,7 @@ tags:
8
  - sentence-similarity
9
  - multimodal-embedding
10
  - lmk
 
11
  - lm-kit
12
  language:
13
  - multilingual
@@ -23,9 +24,12 @@ context.
23
 
24
  | File | Content | Size |
25
  |---|---|---|
26
- | `embeddinggemma-2-Q5_K_M.lmk` | Text encoder (270M parameters, Q5_K_M) and vision encoder (170M parameters, Q8_0, audio tower removed) in one archive | 439 MB |
 
 
 
27
 
28
- The vision encoder reads only its header at load: its weights reach memory with the first image, so text-only use costs no vision memory. The individual GGUF files (and the BF16 vision encoder) are in [lm-kit/embeddinggemma-2-gguf](https://huggingface.co/lm-kit/embeddinggemma-2-gguf).
29
 
30
  ## Usage
31
 
 
8
  - sentence-similarity
9
  - multimodal-embedding
10
  - lmk
11
+ - gguf
12
  - lm-kit
13
  language:
14
  - multilingual
 
24
 
25
  | File | Content | Size |
26
  |---|---|---|
27
+ | `embeddinggemma-2-Q5_K_M.lmk` | LM-Kit archive: the text encoder (Q5_K_M) and the vision encoder (Q8_0) in one file. This is what `LM.LoadFromModelID("embeddinggemma-2")` loads | 439 MB |
28
+ | `embeddinggemma-2-Q5_K_M.gguf` | Text encoder (270M parameters), Q5_K_M | 213 MB |
29
+ | `embeddinggemma-2-mmproj-vision-Q8_0.gguf` | Vision encoder (170M parameters), Q8_0, audio tower removed | 226 MB |
30
+ | `embeddinggemma-2-mmproj-vision-BF16.gguf` | Vision encoder, BF16 reference (full precision, for re-quantization) | 368 MB |
31
 
32
+ The archive's vision encoder reads only its header at load: its weights reach memory with the first image, so text-only use costs no vision memory. The GGUF files are the archive's contents, for runtimes that load GGUF directly.
33
 
34
  ## Usage
35
 
embeddinggemma-2-Q5_K_M.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1553149d4ab999040c7d9358b745d2c7d862ea7e97dac11b06245b776da35e4b
3
+ size 212758816
embeddinggemma-2-mmproj-vision-BF16.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3825df1fe2982a178886c6a4a639d369f8f801daecc6a76922e8a78323aa62ce
3
+ size 368268896
embeddinggemma-2-mmproj-vision-Q8_0.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c7457d04fdbeb67813813b6d146bd4ab6490a75a9462f0cb88528ffc1c9aa77c
3
+ size 226342592