Audio-Text-to-Text
ONNX
Safetensors
English
Chinese
Model Optimizer
bailing_moe_v3_audio
audio
speech
automatic-speech-recognition
audio-understanding
audio-reasoning
speech-to-text
conversational
mixture-of-experts
quantized
nvfp4
fp4
4-bit precision
modelopt
blackwell
custom_code
Eval Results (legacy)
Eval Results
8-bit precision
docs: shorten legacy evaluation metric labels
Browse files
README.md
CHANGED
|
@@ -44,7 +44,7 @@ model-index:
|
|
| 44 |
split: test
|
| 45 |
metrics:
|
| 46 |
- type: wer
|
| 47 |
-
name: WER
|
| 48 |
value: 3.4416431957134024
|
| 49 |
source:
|
| 50 |
name: Completed local full-suite evaluation and scoring policies
|
|
@@ -64,7 +64,7 @@ model-index:
|
|
| 64 |
split: test
|
| 65 |
metrics:
|
| 66 |
- type: cer
|
| 67 |
-
name: CER
|
| 68 |
value: 3.4343844191977744
|
| 69 |
source:
|
| 70 |
name: Completed local full-suite evaluation and scoring policies
|
|
@@ -82,7 +82,7 @@ model-index:
|
|
| 82 |
scoring: Gold-independent answer parsing; capped/unparsed wrong; full denominator
|
| 83 |
metrics:
|
| 84 |
- type: accuracy
|
| 85 |
-
name: Accuracy
|
| 86 |
value: 65.2
|
| 87 |
source:
|
| 88 |
name: Completed local full-suite evaluation and scoring policies
|
|
@@ -100,7 +100,7 @@ model-index:
|
|
| 100 |
scoring: Gold-independent answer parsing; capped/unparsed wrong; full denominator
|
| 101 |
metrics:
|
| 102 |
- type: accuracy
|
| 103 |
-
name: Accuracy
|
| 104 |
value: 77.3
|
| 105 |
source:
|
| 106 |
name: Completed local full-suite evaluation and scoring policies
|
|
@@ -119,7 +119,7 @@ model-index:
|
|
| 119 |
split: train
|
| 120 |
metrics:
|
| 121 |
- type: accuracy
|
| 122 |
-
name: Accuracy
|
| 123 |
value: 67.14
|
| 124 |
source:
|
| 125 |
name: Completed local full-suite evaluation and scoring policies
|
|
|
|
| 44 |
split: test
|
| 45 |
metrics:
|
| 46 |
- type: wer
|
| 47 |
+
name: WER
|
| 48 |
value: 3.4416431957134024
|
| 49 |
source:
|
| 50 |
name: Completed local full-suite evaluation and scoring policies
|
|
|
|
| 64 |
split: test
|
| 65 |
metrics:
|
| 66 |
- type: cer
|
| 67 |
+
name: CER
|
| 68 |
value: 3.4343844191977744
|
| 69 |
source:
|
| 70 |
name: Completed local full-suite evaluation and scoring policies
|
|
|
|
| 82 |
scoring: Gold-independent answer parsing; capped/unparsed wrong; full denominator
|
| 83 |
metrics:
|
| 84 |
- type: accuracy
|
| 85 |
+
name: Accuracy
|
| 86 |
value: 65.2
|
| 87 |
source:
|
| 88 |
name: Completed local full-suite evaluation and scoring policies
|
|
|
|
| 100 |
scoring: Gold-independent answer parsing; capped/unparsed wrong; full denominator
|
| 101 |
metrics:
|
| 102 |
- type: accuracy
|
| 103 |
+
name: Accuracy
|
| 104 |
value: 77.3
|
| 105 |
source:
|
| 106 |
name: Completed local full-suite evaluation and scoring policies
|
|
|
|
| 119 |
split: train
|
| 120 |
metrics:
|
| 121 |
- type: accuracy
|
| 122 |
+
name: Accuracy
|
| 123 |
value: 67.14
|
| 124 |
source:
|
| 125 |
name: Completed local full-suite evaluation and scoring policies
|