Feature Extraction
Transformers
Safetensors
English
multilingual
gemma4_audio
audio
speech
conformer
gemma4
usm
google
Eval Results (legacy)
Instructions to use rnagabh/gemma4-audio-encoder with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use rnagabh/gemma4-audio-encoder with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("feature-extraction", model="rnagabh/gemma4-audio-encoder")# Load model directly from transformers import AutoTokenizer, AutoModel tokenizer = AutoTokenizer.from_pretrained("rnagabh/gemma4-audio-encoder") model = AutoModel.from_pretrained("rnagabh/gemma4-audio-encoder", device_map="auto") - Notebooks
- Google Colab
- Kaggle
Initial upload: Gemma 4 audio encoder (304.8M USM-style Conformer)
Browse files
README.md
CHANGED
|
@@ -12,8 +12,23 @@ tags:
|
|
| 12 |
- gemma4
|
| 13 |
- usm
|
| 14 |
- google
|
|
|
|
| 15 |
pipeline_tag: feature-extraction
|
| 16 |
base_model: google/gemma-4-E2B-it
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 17 |
---
|
| 18 |
|
| 19 |
# Gemma 4 Audio Encoder (USM-style Conformer)
|
|
|
|
| 12 |
- gemma4
|
| 13 |
- usm
|
| 14 |
- google
|
| 15 |
+
- safetensors
|
| 16 |
pipeline_tag: feature-extraction
|
| 17 |
base_model: google/gemma-4-E2B-it
|
| 18 |
+
model-index:
|
| 19 |
+
- name: gemma4-audio-encoder
|
| 20 |
+
results:
|
| 21 |
+
- task:
|
| 22 |
+
type: audio-classification
|
| 23 |
+
name: Speech Commands (35-class)
|
| 24 |
+
dataset:
|
| 25 |
+
name: Google Speech Commands v0.02
|
| 26 |
+
type: google/speech_commands
|
| 27 |
+
split: validation
|
| 28 |
+
metrics:
|
| 29 |
+
- type: accuracy
|
| 30 |
+
value: 72.0
|
| 31 |
+
name: Linear Probe Accuracy
|
| 32 |
---
|
| 33 |
|
| 34 |
# Gemma 4 Audio Encoder (USM-style Conformer)
|