depasquale commited on
Commit
4e64a43
·
verified ·
1 Parent(s): 4804c7e

Upload folder using huggingface_hub

Browse files
Files changed (4) hide show
  1. README.md +41 -0
  2. config.json +17 -0
  3. model.safetensors +3 -0
  4. multilingual.tiktoken +0 -0
README.md ADDED
@@ -0,0 +1,41 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: mlx-audio-plus
3
+ license: apache-2.0
4
+ base_model:
5
+ - openai/whisper-tiny
6
+ tags:
7
+ - mlx
8
+ - whisper
9
+ - speech-recognition
10
+ - speech-to-text
11
+ - stt
12
+ pipeline_tag: automatic-speech-recognition
13
+ ---
14
+
15
+ # mlx-community/whisper-tiny-8bit
16
+
17
+ This model was converted to MLX format from [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) using [mlx-audio-plus](https://github.com/DePasqualeOrg/mlx-audio-plus) version **0.1.3**.
18
+
19
+ ## Use with mlx-audio-plus
20
+
21
+ ```bash
22
+ pip install -U mlx-audio-plus
23
+ ```
24
+
25
+ ### Command line
26
+
27
+ ```bash
28
+ mlx_audio.stt --model mlx-community/whisper-tiny-8bit --audio audio.mp3
29
+ ```
30
+
31
+ ### Python
32
+
33
+ ```python
34
+ from mlx_audio.stt import transcribe
35
+
36
+ result = transcribe(
37
+ audio="audio.mp3",
38
+ model="mlx-community/whisper-tiny-8bit",
39
+ )
40
+ print(result["text"])
41
+ ```
config.json ADDED
@@ -0,0 +1,17 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "n_mels": 80,
3
+ "n_audio_ctx": 1500,
4
+ "n_audio_state": 384,
5
+ "n_audio_head": 6,
6
+ "n_audio_layer": 4,
7
+ "n_vocab": 51865,
8
+ "n_text_ctx": 448,
9
+ "n_text_state": 384,
10
+ "n_text_head": 6,
11
+ "n_text_layer": 4,
12
+ "quantization": {
13
+ "group_size": 64,
14
+ "bits": 8
15
+ },
16
+ "model_type": "whisper"
17
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:17c05fb478f4424ad1737ff75bce1e709e2b36cc91f3cabd8a91929a7b1aaf2c
3
+ size 40245180
multilingual.tiktoken ADDED
The diff for this file is too large to render. See raw diff