Fixing multiples small bugs

graphs/dataflow_robot_vlm_minimize.yml

@@ -11,6 +11,7 @@ nodes:
         image: 
           source: webcam/image
           queue_size: 1
+        text: whisper/text
       outputs:
         - speak
         - control
@@ -31,4 +32,11 @@ nodes:
     operator:
       python: ../operators/plot.py
       inputs:
-        image: webcam/image
+        image: webcam/image
+  - id: whisper
+    operator:
+      python: ../operators/whisper_op.py
+      inputs:
+        audio: dora/timer/millis/1000
+      outputs:
+        - text

operators/parler_op.py

@@ -5,14 +5,14 @@ import pygame
 from dora import DoraStatus
 
 model = ParlerTTSForConditionalGeneration.from_pretrained(
-    "parler-tts/parler_tts_mini_v0.1"
+    "ylacombe/parler-tts-mini-jenny-30H"
 ).to("cuda:0")
-tokenizer = AutoTokenizer.from_pretrained("parler-tts/parler_tts_mini_v0.1")
+tokenizer = AutoTokenizer.from_pretrained("ylacombe/parler-tts-mini-jenny-30H")
 
 pygame.mixer.init()
 
 input_ids = tokenizer(
-    "A female speaker with a slightly low-pitched voice delivers her words quite expressively, in a very confined sounding environment with clear audio quality. She speaks very fast.",
+    "Jenny delivers her words quite expressively, in a very confined sounding environment with clear audio quality.",
     return_tensors="pt",
 ).input_ids.to("cuda:0")
 
@@ -25,8 +25,8 @@ class Operator:
     ):
         if dora_event["type"] == "INPUT":
             generation = model.generate(
-                max_new_tokens=300,
                 input_ids=input_ids,
+                min_new_tokens=100,
                 prompt_input_ids=tokenizer(
                     dora_event["value"][0].as_py(), return_tensors="pt"
                 ).input_ids.to("cuda:0"),
@@ -38,9 +38,15 @@ class Operator:
                 model.config.sampling_rate,
             )
 
-            while pygame.mixer.get_busy():
-                pass
             pygame.mixer.music.load(f"parler_tts_out.wav")
             pygame.mixer.music.play()
+            while pygame.mixer.get_busy():
+                pass
 
         return DoraStatus.CONTINUE
+
+
+# op = Operator()
+# import pyarrow as pa
+
+# op.on_event({"type": "INPUT", "value": pa.array(["Hello, how are you?"])}, None)

operators/robot_minimize.py

@@ -21,10 +21,10 @@ for dora_event in node:
         [x, y, z, xy_speed, z_speed, pitch, yaw] = dora_event["value"].to_numpy()
         print(dora_event["value"].to_numpy())
         event = ep_robot.gimbal.moveto(
-            pitch=pitch, yaw=yaw, pitch_speed=50.0, yaw_speed=50.0
+            pitch=pitch, yaw=yaw, pitch_speed=60.0, yaw_speed=50.0
         )
         wait(event)
-        sleep(4)
+        sleep(2)
         event = ep_robot.chassis.move(x=x, y=y, z=z, xy_speed=xy_speed, z_speed=z_speed)
         wait(event)
         sleep(6)

operators/whisper_op.py

@@ -4,14 +4,17 @@ from pynput import keyboard
 from pynput.keyboard import Key
 from dora import DoraStatus
 
+import torch
 import numpy as np
 import pyarrow as pa
 import sounddevice as sd
+import gc  # garbage collect library
 
 model = whisper.load_model("base")
 
 SAMPLE_RATE = 16000
-MAX_DURATION = 20
+MAX_DURATION = 10
+MIN_DURATION = 6
 
 
 class Operator:
@@ -19,6 +22,9 @@ class Operator:
     Transforming Speech to Text using OpenAI Whisper model
     """
 
+    def __init__(self) -> None:
+        self.policy_init = False
+
     def on_event(
         self,
         dora_event,
@@ -30,10 +36,17 @@ class Operator:
             with keyboard.Events() as events:
                 event = events.get(1.0)
                 if event is not None and event.key == Key.up:
-                    send_output("led", pa.array([0, 255, 0]))
+                    # send_output("led", pa.array([0, 255, 0]))
+
+                    if self.policy_init == False:
+                        self.policy_init = True
+                        duration = MAX_DURATION
+                    else:
+                        duration = MIN_DURATION
+
                     ## Microphone
                     audio_data = sd.rec(
-                        int(SAMPLE_RATE * MAX_DURATION),
+                        int(SAMPLE_RATE * duration),
                         samplerate=SAMPLE_RATE,
                         channels=1,
                         dtype=np.int16,
@@ -48,11 +61,9 @@ class Operator:
                     send_output(
                         "text", pa.array([result["text"]]), dora_event["metadata"]
                     )
-                    send_output("led", pa.array([0, 0, 255]))
-                    del model
-
-                    import gc  # garbage collect library
+                    # send_output("led", pa.array([0, 0, 255]))
 
                     gc.collect()
+                    torch.cuda.empty_cache()
 
         return DoraStatus.CONTINUE