Spaces:
Runtime error
Runtime error
Update app.py
Browse files
app.py
CHANGED
@@ -43,10 +43,12 @@ def transcribe(audio_file):
|
|
43 |
transcription = asr_model.transcribe(audio_file)["text"]
|
44 |
return transcription
|
45 |
|
46 |
-
def inference(audio, prompt, model, temperature):
|
47 |
# Transcribe with Whisper
|
48 |
print("The audio is:", audio)
|
49 |
transcript = transcribe(audio)
|
|
|
|
|
50 |
|
51 |
text = prompt + transcript + "\nPrediction: "
|
52 |
|
@@ -70,12 +72,14 @@ def inference(audio, prompt, model, temperature):
|
|
70 |
#print("Infered type is: ", type(infered))
|
71 |
infers = list(map(lambda x: x.replace("\n", ""), temp))
|
72 |
#infered = list(map(lambda x: x.split(','), infers))
|
|
|
73 |
|
74 |
-
return transcript, infers
|
75 |
|
76 |
|
77 |
# get audio from microphone
|
78 |
with gr.Blocks() as face:
|
|
|
79 |
with gr.Row():
|
80 |
with gr.Column():
|
81 |
audio = gr.Audio(source="microphone", type="filepath")
|
@@ -86,8 +90,9 @@ with gr.Blocks() as face:
|
|
86 |
with gr.Column():
|
87 |
script = gr.Textbox(label="text...")
|
88 |
options = gr.Textbox(label="predictions...")
|
|
|
89 |
#transcribe_btn.click(inference)
|
90 |
-
transcribe_btn.click(fn=inference, inputs=[audio, promptText, dropChoice, sliderChoice], outputs=[script, options])
|
91 |
examples = gr.Examples(examples=["Sedan, Truck, SUV", "Dalmaion, Shepherd, Lab, Mutt"], inputs=[options])
|
92 |
|
93 |
face.launch()
|
|
|
43 |
transcription = asr_model.transcribe(audio_file)["text"]
|
44 |
return transcription
|
45 |
|
46 |
+
def inference(audio, prompt, model, temperature, latest):
|
47 |
# Transcribe with Whisper
|
48 |
print("The audio is:", audio)
|
49 |
transcript = transcribe(audio)
|
50 |
+
|
51 |
+
lastest.append(transcript)
|
52 |
|
53 |
text = prompt + transcript + "\nPrediction: "
|
54 |
|
|
|
72 |
#print("Infered type is: ", type(infered))
|
73 |
infers = list(map(lambda x: x.replace("\n", ""), temp))
|
74 |
#infered = list(map(lambda x: x.split(','), infers))
|
75 |
+
convoState: latest
|
76 |
|
77 |
+
return transcript, infers, convoState
|
78 |
|
79 |
|
80 |
# get audio from microphone
|
81 |
with gr.Blocks() as face:
|
82 |
+
convoState = gr.State([])
|
83 |
with gr.Row():
|
84 |
with gr.Column():
|
85 |
audio = gr.Audio(source="microphone", type="filepath")
|
|
|
90 |
with gr.Column():
|
91 |
script = gr.Textbox(label="text...")
|
92 |
options = gr.Textbox(label="predictions...")
|
93 |
+
latestConvo = gr.Textbox()
|
94 |
#transcribe_btn.click(inference)
|
95 |
+
transcribe_btn.click(fn=inference, inputs=[audio, promptText, dropChoice, sliderChoice, convoState], outputs=[script, options, latestConvo])
|
96 |
examples = gr.Examples(examples=["Sedan, Truck, SUV", "Dalmaion, Shepherd, Lab, Mutt"], inputs=[options])
|
97 |
|
98 |
face.launch()
|