Instructions to use mazesmazes/tiny-audio-next with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use mazesmazes/tiny-audio-next with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("feature-extraction", model="mazesmazes/tiny-audio-next", trust_remote_code=True)# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("mazesmazes/tiny-audio-next", trust_remote_code=True, dtype="auto") - Notebooks
- Google Colab
- Kaggle
Training in progress - step 1000
Browse files- config.json +2 -2
- model.safetensors +2 -2
config.json
CHANGED
|
@@ -235,7 +235,7 @@
|
|
| 235 |
],
|
| 236 |
"encoder_dim": 1280,
|
| 237 |
"eos_token_id": 151645,
|
| 238 |
-
"freeze_language_model":
|
| 239 |
"freeze_projector": false,
|
| 240 |
"length_penalty": 1.0,
|
| 241 |
"llm_dim": 1024,
|
|
@@ -262,7 +262,7 @@
|
|
| 262 |
"pad_token_id": 151643,
|
| 263 |
"pipeline_tag": "automatic-speech-recognition",
|
| 264 |
"pretrained_model_path": "mazesmazes/tiny-audio-next",
|
| 265 |
-
"projector_hidden_dim":
|
| 266 |
"projector_pool_stride": 4,
|
| 267 |
"projector_type": "mlp",
|
| 268 |
"qformer_hidden_size": null,
|
|
|
|
| 235 |
],
|
| 236 |
"encoder_dim": 1280,
|
| 237 |
"eos_token_id": 151645,
|
| 238 |
+
"freeze_language_model": true,
|
| 239 |
"freeze_projector": false,
|
| 240 |
"length_penalty": 1.0,
|
| 241 |
"llm_dim": 1024,
|
|
|
|
| 262 |
"pad_token_id": 151643,
|
| 263 |
"pipeline_tag": "automatic-speech-recognition",
|
| 264 |
"pretrained_model_path": "mazesmazes/tiny-audio-next",
|
| 265 |
+
"projector_hidden_dim": 1024,
|
| 266 |
"projector_pool_stride": 4,
|
| 267 |
"projector_type": "mlp",
|
| 268 |
"qformer_hidden_size": null,
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f3ae100b2adc53c9b6e5ae5d2c8ab93f6e390f37de8799df659d607b051184bb
|
| 3 |
+
size 25174432
|