Instructions to use onnx-community/Voxtral-Mini-3B-2507-ONNX with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers.js
How to use onnx-community/Voxtral-Mini-3B-2507-ONNX with Transformers.js:
// npm i @huggingface/transformers import { pipeline } from '@huggingface/transformers'; // Allocate pipeline const pipe = await pipeline('audio-text-to-text', 'onnx-community/Voxtral-Mini-3B-2507-ONNX');
Update config.json
Browse files- config.json +9 -1
config.json
CHANGED
|
@@ -28,12 +28,20 @@
|
|
| 28 |
"use_cache": true,
|
| 29 |
"vocab_size": 131072,
|
| 30 |
"transformers.js_config": {
|
|
|
|
|
|
|
|
|
|
| 31 |
"kv_cache_dtype": {
|
| 32 |
"q4f16": "float16",
|
| 33 |
"fp16": "float16"
|
| 34 |
},
|
| 35 |
"use_external_data_format": {
|
| 36 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 37 |
}
|
| 38 |
}
|
| 39 |
}
|
|
|
|
| 28 |
"use_cache": true,
|
| 29 |
"vocab_size": 131072,
|
| 30 |
"transformers.js_config": {
|
| 31 |
+
"dtype": {
|
| 32 |
+
"embed_tokens": "fp16"
|
| 33 |
+
},
|
| 34 |
"kv_cache_dtype": {
|
| 35 |
"q4f16": "float16",
|
| 36 |
"fp16": "float16"
|
| 37 |
},
|
| 38 |
"use_external_data_format": {
|
| 39 |
+
"decoder_model_merged.onnx": 8,
|
| 40 |
+
"decoder_model_merged_fp16.onnx": 4,
|
| 41 |
+
"decoder_model_merged_q4.onnx": 2,
|
| 42 |
+
"decoder_model_merged_q4f16.onnx": 1,
|
| 43 |
+
"embed_tokens.onnx": 1,
|
| 44 |
+
"embed_tokens_fp16.onnx": 1
|
| 45 |
}
|
| 46 |
}
|
| 47 |
}
|