MultiagentPersonalAssistant

agent/multi_agent.py

@@ -7,7 +7,6 @@ from sqlalchemy.ext.asyncio import create_async_engine
 from langgraph_supervisor import create_supervisor
 from langmem.short_term import SummarizationNode
 from typing_extensions import TypedDict
-from dotenv import load_dotenv
 import os
 
 class State(TypedDict):

app.py

@@ -35,7 +35,7 @@ async def voice_mode(
     await assistant.authorization()
 
     # Preprocess file
-    transcription = await asyncio.to_thread(preprocess_audio, str(dest))
+    transcription = await preprocess_audio(str(dest))
     state_data["message"] = HumanMessage(
         content=transcription
     )

requirements.txt

@@ -4,7 +4,6 @@ python-dotenv>=1.1.1
 langchain>=0.3.27
 langgraph>=0.6.3
 openai>=1.99.1
-transformers>=4.53.2
 timm>=1.0.19
 langchain-groq>=0.3.7
 langgraph-supervisor>=0.0.29
@@ -12,14 +11,11 @@ wikipedia>=1.4.0
 wolframalpha>=5.1.3
 python-multipart
 soundfile>=0.13.1
-accelerate
-jsonpointer
 langmem>=0.0.29
 greenlet>=3.2.3
 deepagents>=0.0.3
 python-docx>=1.2.0
 pydub>=0.25.1
-SpeechRecognition>=3.14.3
 PyPDF2>=3.0.1
 asyncpg>=0.30.0
 pillow>=11.3.0

tmp/hyperparams.yaml

@@ -0,0 +1,57 @@
+pretrained_path: speechbrain/lang-id-voxlingua107-ecapa
+
+
+# Feature parameters
+n_mels: 60
+left_frames: 0
+right_frames: 0
+deltas: false
+
+# Number of speakers
+out_n_neurons: 107
+
+# Functions
+compute_features: !new:speechbrain.lobes.features.Fbank
+  n_mels: 60
+  left_frames: 0
+  right_frames: 0
+  deltas: false
+
+embedding_model: !new:speechbrain.lobes.models.ECAPA_TDNN.ECAPA_TDNN
+  input_size: 60
+  channels: [1024, 1024, 1024, 1024, 3072]
+  kernel_sizes: [5, 3, 3, 3, 1]
+  dilations: [1, 2, 3, 4, 1]
+  attention_channels: 128
+  lin_neurons: 256
+
+
+classifier: !new:speechbrain.lobes.models.Xvector.Classifier
+  input_shape: [null, null, 256]
+  activation: !name:torch.nn.LeakyReLU
+  lin_blocks: 1
+  lin_neurons: 512
+  out_neurons: !ref <out_n_neurons>
+
+
+mean_var_norm: !new:speechbrain.processing.features.InputNormalization
+  norm_type: sentence
+  std_norm: false
+
+modules:
+    compute_features: !ref <compute_features>
+    mean_var_norm: !ref <mean_var_norm>
+    embedding_model: !ref <embedding_model>
+    classifier: !ref <classifier>
+
+label_encoder: !new:speechbrain.dataio.encoder.CategoricalEncoder
+
+pretrainer: !new:speechbrain.utils.parameter_transfer.Pretrainer
+    loadables:
+        embedding_model: !ref <embedding_model>
+        classifier: !ref <classifier>
+        label_encoder: !ref <label_encoder>
+    paths:
+        embedding_model: !ref <pretrained_path>/embedding_model.ckpt
+        classifier: !ref <pretrained_path>/classifier.ckpt
+        label_encoder: !ref <pretrained_path>/label_encoder.txt

tmp/label_encoder.ckpt

@@ -0,0 +1,109 @@
+'ab: Abkhazian' => 0
+'af: Afrikaans' => 1
+'am: Amharic' => 2
+'ar: Arabic' => 3
+'as: Assamese' => 4
+'az: Azerbaijani' => 5
+'ba: Bashkir' => 6
+'be: Belarusian' => 7
+'bg: Bulgarian' => 8
+'bn: Bengali' => 9
+'bo: Tibetan' => 10
+'br: Breton' => 11
+'bs: Bosnian' => 12
+'ca: Catalan' => 13
+'ceb: Cebuano' => 14
+'cs: Czech' => 15
+'cy: Welsh' => 16
+'da: Danish' => 17
+'de: German' => 18
+'el: Greek' => 19
+'en: English' => 20
+'eo: Esperanto' => 21
+'es: Spanish' => 22
+'et: Estonian' => 23
+'eu: Basque' => 24
+'fa: Persian' => 25
+'fi: Finnish' => 26
+'fo: Faroese' => 27
+'fr: French' => 28
+'gl: Galician' => 29
+'gn: Guarani' => 30
+'gu: Gujarati' => 31
+'gv: Manx' => 32
+'ha: Hausa' => 33
+'haw: Hawaiian' => 34
+'hi: Hindi' => 35
+'hr: Croatian' => 36
+'ht: Haitian' => 37
+'hu: Hungarian' => 38
+'hy: Armenian' => 39
+'ia: Interlingua' => 40
+'id: Indonesian' => 41
+'is: Icelandic' => 42
+'it: Italian' => 43
+'iw: Hebrew' => 44
+'ja: Japanese' => 45
+'jw: Javanese' => 46
+'ka: Georgian' => 47
+'kk: Kazakh' => 48
+'km: Central Khmer' => 49
+'kn: Kannada' => 50
+'ko: Korean' => 51
+'la: Latin' => 52
+'lb: Luxembourgish' => 53
+'ln: Lingala' => 54
+'lo: Lao' => 55
+'lt: Lithuanian' => 56
+'lv: Latvian' => 57
+'mg: Malagasy' => 58
+'mi: Maori' => 59
+'mk: Macedonian' => 60
+'ml: Malayalam' => 61
+'mn: Mongolian' => 62
+'mr: Marathi' => 63
+'ms: Malay' => 64
+'mt: Maltese' => 65
+'my: Burmese' => 66
+'ne: Nepali' => 67
+'nl: Dutch' => 68
+'nn: Norwegian Nynorsk' => 69
+'no: Norwegian' => 70
+'oc: Occitan' => 71
+'pa: Panjabi' => 72
+'pl: Polish' => 73
+'ps: Pushto' => 74
+'pt: Portuguese' => 75
+'ro: Romanian' => 76
+'ru: Russian' => 77
+'sa: Sanskrit' => 78
+'sco: Scots' => 79
+'sd: Sindhi' => 80
+'si: Sinhala' => 81
+'sk: Slovak' => 82
+'sl: Slovenian' => 83
+'sn: Shona' => 84
+'so: Somali' => 85
+'sq: Albanian' => 86
+'sr: Serbian' => 87
+'su: Sundanese' => 88
+'sv: Swedish' => 89
+'sw: Swahili' => 90
+'ta: Tamil' => 91
+'te: Telugu' => 92
+'tg: Tajik' => 93
+'th: Thai' => 94
+'tk: Turkmen' => 95
+'tl: Tagalog' => 96
+'tr: Turkish' => 97
+'tt: Tatar' => 98
+'uk: Ukrainian' => 99
+'ur: Urdu' => 100
+'uz: Uzbek' => 101
+'vi: Vietnamese' => 102
+'war: Waray' => 103
+'yi: Yiddish' => 104
+'yo: Yoruba' => 105
+'zh: Chinese' => 106
+================
+'starting_index' => 0