generate_meta update
Browse files- README.md +3 -3
- config.json +1 -6
README.md
CHANGED
@@ -3,6 +3,9 @@ language:
|
|
3 |
- en
|
4 |
license: apache-2.0
|
5 |
library_name: transformers
|
|
|
|
|
|
|
6 |
datasets:
|
7 |
- mozilla-foundation/common_voice_16_1
|
8 |
- openslr/librispeech_asr
|
@@ -97,9 +100,6 @@ model-index:
|
|
97 |
- type: accuracy
|
98 |
value: 64.57
|
99 |
name: Test Accent Accuracy
|
100 |
-
tags:
|
101 |
-
- multi-modal
|
102 |
-
- speech-language
|
103 |
---
|
104 |
|
105 |
# SpeechLLM
|
|
|
3 |
- en
|
4 |
license: apache-2.0
|
5 |
library_name: transformers
|
6 |
+
tags:
|
7 |
+
- multi-modal
|
8 |
+
- speech-language
|
9 |
datasets:
|
10 |
- mozilla-foundation/common_voice_16_1
|
11 |
- openslr/librispeech_asr
|
|
|
100 |
- type: accuracy
|
101 |
value: 64.57
|
102 |
name: Test Accent Accuracy
|
|
|
|
|
|
|
103 |
---
|
104 |
|
105 |
# SpeechLLM
|
config.json
CHANGED
@@ -1,18 +1,13 @@
|
|
1 |
{
|
2 |
-
"architectures": [
|
3 |
-
"SpeechLLMModel"
|
4 |
-
],
|
5 |
"audio_enc_dim": 1280,
|
6 |
"audio_encoder_name": "facebook/hubert-xlarge-ll60k",
|
7 |
"audio_processor_name": "facebook/hubert-large-ls960-ft",
|
8 |
"auto_map": {
|
9 |
-
"AutoConfig": "config.SpeechLLMModelConfig"
|
10 |
-
"AutoModel": "model.SpeechLLMModel"
|
11 |
},
|
12 |
"llm_dim": 2048,
|
13 |
"llm_model_checkpoint": "hf_repo/llm_model_checkpoint",
|
14 |
"llm_model_name": "TinyLlama/TinyLlama-1.1B-Chat-v1.0",
|
15 |
"model_type": "custom_model",
|
16 |
-
"torch_dtype": "float32",
|
17 |
"transformers_version": "4.38.2"
|
18 |
}
|
|
|
1 |
{
|
|
|
|
|
|
|
2 |
"audio_enc_dim": 1280,
|
3 |
"audio_encoder_name": "facebook/hubert-xlarge-ll60k",
|
4 |
"audio_processor_name": "facebook/hubert-large-ls960-ft",
|
5 |
"auto_map": {
|
6 |
+
"AutoConfig": "config.SpeechLLMModelConfig"
|
|
|
7 |
},
|
8 |
"llm_dim": 2048,
|
9 |
"llm_model_checkpoint": "hf_repo/llm_model_checkpoint",
|
10 |
"llm_model_name": "TinyLlama/TinyLlama-1.1B-Chat-v1.0",
|
11 |
"model_type": "custom_model",
|
|
|
12 |
"transformers_version": "4.38.2"
|
13 |
}
|