Audio-Text-to-Text
Transformers
Safetensors
English
Chinese
moss_transcribe_diarize
text-generation
moss
audio
speech
asr
diarization
timestamp-asr
long-form-audio
multimodal
multilingual
custom_code
Instructions to use OpenMOSS-Team/MOSS-Transcribe-Diarize with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use OpenMOSS-Team/MOSS-Transcribe-Diarize with Transformers:
# Load model directly from transformers import AutoModelForCausalLM model = AutoModelForCausalLM.from_pretrained("OpenMOSS-Team/MOSS-Transcribe-Diarize", trust_remote_code=True, dtype="auto") - Notebooks
- Google Colab
- Kaggle
Update config dtype and max position embeddings
Browse files- config.json +2 -1
config.json
CHANGED
|
@@ -9,6 +9,7 @@
|
|
| 9 |
"AutoProcessor": "processing_moss_transcribe_diarize.MossTranscribeDiarizeProcessor"
|
| 10 |
},
|
| 11 |
"model_type": "moss_transcribe_diarize",
|
|
|
|
| 12 |
"text_config": {
|
| 13 |
"model_type": "qwen3",
|
| 14 |
"vocab_size": 151936,
|
|
@@ -19,7 +20,7 @@
|
|
| 19 |
"num_key_value_heads": 8,
|
| 20 |
"head_dim": 128,
|
| 21 |
"hidden_act": "silu",
|
| 22 |
-
"max_position_embeddings":
|
| 23 |
"initializer_range": 0.02,
|
| 24 |
"rms_norm_eps": 1e-06,
|
| 25 |
"tie_word_embeddings": true,
|
|
|
|
| 9 |
"AutoProcessor": "processing_moss_transcribe_diarize.MossTranscribeDiarizeProcessor"
|
| 10 |
},
|
| 11 |
"model_type": "moss_transcribe_diarize",
|
| 12 |
+
"dtype": "bfloat16",
|
| 13 |
"text_config": {
|
| 14 |
"model_type": "qwen3",
|
| 15 |
"vocab_size": 151936,
|
|
|
|
| 20 |
"num_key_value_heads": 8,
|
| 21 |
"head_dim": 128,
|
| 22 |
"hidden_act": "silu",
|
| 23 |
+
"max_position_embeddings": 131072,
|
| 24 |
"initializer_range": 0.02,
|
| 25 |
"rms_norm_eps": 1e-06,
|
| 26 |
"tie_word_embeddings": true,
|