hanxiao commited on
Commit
85842c9
·
verified ·
1 Parent(s): a877d94

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +12 -8
README.md CHANGED
@@ -19,11 +19,11 @@ Here, we removed the visual components of qwen2.5-vl and merged all LoRA adapter
19
 
20
  | HuggingFace Repo | Task |
21
  |---|---|
22
- | [`jina-embeddings-v4-text-retrieval-GGUF`](https://huggingface.co/jinaai/jina-embeddings-v4-text-retrieval-GGUF) | Text retrieval |
23
- | [`jina-embeddings-v4-text-code-GGUF`](https://huggingface.co/jinaai/jina-embeddings-v4-text-code-GGUF) | Code retrieval |
24
- | [`jina-embeddings-v4-text-matching-GGUF`](https://huggingface.co/jinaai/jina-embeddings-v4-text-matching-GGUF) | Sentence similarity |
25
 
26
- All models above provide F16, Q8_0, Q6_K, Q5_K_M, Q4_K_M, Q3_K_M quantizations.
27
 
28
  ### Limitations
29
  - They can not handle image input.
@@ -34,16 +34,20 @@ All models above provide F16, Q8_0, Q6_K, Q5_K_M, Q4_K_M, Q3_K_M quantizations.
34
 
35
  TBA
36
 
37
- ## Getting Embeddings
38
 
39
  First [install llama.cpp](https://github.com/ggml-org/llama.cpp/blob/master/docs/install.md).
40
 
41
- Run `llama-server` to host the embedding model as an HTTP server:
42
 
43
  ```bash
44
- llama-server -m jina-embeddings-v4-text-matching-F16.gguf --embedding --pooling mean
45
  ```
46
 
 
 
 
 
47
  Client:
48
 
49
  ```bash
@@ -77,5 +81,5 @@ curl -X POST "http://127.0.0.1:8080/v1/embeddings" \
77
  You can also use `llama-embedding` for one-shot embedding:
78
 
79
  ```bash
80
- llama-embedding -m jina-embeddings-v4-text-matching-F16.gguf --pooling mean -p "jina is awesome" 2>/dev/null
81
  ```
 
19
 
20
  | HuggingFace Repo | Task |
21
  |---|---|
22
+ | [`jinaai/jina-embeddings-v4-text-retrieval-GGUF`](https://huggingface.co/jinaai/jina-embeddings-v4-text-retrieval-GGUF) | Text retrieval |
23
+ | [`jinaai/jina-embeddings-v4-text-code-GGUF`](https://huggingface.co/jinaai/jina-embeddings-v4-text-code-GGUF) | Code retrieval |
24
+ | [`jinaai/jina-embeddings-v4-text-matching-GGUF`](https://huggingface.co/jinaai/jina-embeddings-v4-text-matching-GGUF) | Sentence similarity |
25
 
26
+ All models above provide F16, Q8_0, Q6_K, Q5_K_M, Q4_K_M, Q3_K_M quantizations. More quantizations such as Unsloth-like dynamic quantizations are on the way.
27
 
28
  ### Limitations
29
  - They can not handle image input.
 
34
 
35
  TBA
36
 
37
+ ## Get Embeddings
38
 
39
  First [install llama.cpp](https://github.com/ggml-org/llama.cpp/blob/master/docs/install.md).
40
 
41
+ Run `llama-server` to host the embedding model as OpenAI API compatible HTTP server. As an example for using `text-matching` with `F16`, you can do:
42
 
43
  ```bash
44
+ llama-server -hf jinaai/jina-embeddings-v4-text-matching-GGUF:F16 --embedding --pooling mean -ub 8192
45
  ```
46
 
47
+ Remarks:
48
+ - `--pooling mean` is required as v4 is mean-pooling embeddings.
49
+ - setting `--pooling none` is *not* as same as the multi-vector embeddings of v4. The original v4 has a trained MLP on top of the last hidden states to output multi-vector embeddings, each has 128-dim. In GGUF, this MLP was chopped off.
50
+
51
  Client:
52
 
53
  ```bash
 
81
  You can also use `llama-embedding` for one-shot embedding:
82
 
83
  ```bash
84
+ llama-embedding -hf jinaai/jina-embeddings-v4-text-matching-GGUF:F16 --pooling mean -p "jina is awesome" 2>/dev/null
85
  ```