init

Dockerfile

@@ -0,0 +1,27 @@
+FROM python:3.11-slim
+
+# Set working directory
+WORKDIR /app
+
+# Install dependencies
+RUN apt-get update && apt-get install -y \
+    build-essential \
+    ffmpeg \
+    curl \
+    && rm -rf /var/lib/apt/lists/*
+
+# Copy requirements
+COPY requirements.txt .
+
+# Install Python dependencies
+RUN pip install --no-cache-dir -r requirements.txt
+
+# Copy application files
+COPY app.py .
+COPY model/ model/
+
+# Expose port
+EXPOSE 5082
+
+# Run FastAPI app
+CMD ["uvicorn", "app:app", "--host", "0.0.0.0", "--port", "5082"]

app.py

@@ -0,0 +1,51 @@
+from fastapi import FastAPI, File, UploadFile, HTTPException
+from fastapi.responses import JSONResponse
+from vosk import Model, KaldiRecognizer
+import wave
+import os
+import shutil
+import json
+
+app = FastAPI()
+MODEL_PATH = "model/vosk-model-small-en-us-0.15"
+
+if not os.path.exists(MODEL_PATH):
+    raise RuntimeError(f"Vosk model not found at {MODEL_PATH}")
+
+model = Model(MODEL_PATH)
+
+@app.post("/stt")
+async def transcribe(audio: UploadFile = File(...)):
+    if not audio.filename.endswith(".wav"):
+        raise HTTPException(status_code=400, detail="Only .wav files are supported")
+
+    temp_path = f"temp_{audio.filename}"
+    with open(temp_path, "wb") as f:
+        shutil.copyfileobj(audio.file, f)
+
+    try:
+        wf = wave.open(temp_path, "rb")
+        if wf.getnchannels() != 1 or wf.getsampwidth() != 2 or wf.getframerate() != 16000:
+            raise HTTPException(status_code=400, detail="Audio must be 16kHz 16-bit mono WAV")
+
+        rec = KaldiRecognizer(model, wf.getframerate())
+        rec.SetWords(True)
+
+        results = []
+        while True:
+            data = wf.readframes(4000)
+            if len(data) == 0:
+                break
+            if rec.AcceptWaveform(data):
+                results.append(json.loads(rec.Result()))
+
+        results.append(json.loads(rec.FinalResult()))
+        full_text = " ".join([res.get("text", "") for res in results])
+
+        return JSONResponse(content={"text": full_text})
+
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=str(e))
+    finally:
+        wf.close()
+        os.remove(temp_path)

model/vosk-model-small-en-us-0.15/README

@@ -0,0 +1,9 @@
+US English model for mobile Vosk applications
+
+Copyright 2020 Alpha Cephei Inc
+
+Accuracy: 10.38 (tedlium test) 9.85 (librispeech test-clean)
+Speed: 0.11xRT (desktop)
+Latency: 0.15s (right context)
+
+

model/vosk-model-small-en-us-0.15/conf/mfcc.conf

@@ -0,0 +1,7 @@
+--sample-frequency=16000
+--use-energy=false
+--num-mel-bins=40
+--num-ceps=40
+--low-freq=20
+--high-freq=7600
+--allow-downsample=true

model/vosk-model-small-en-us-0.15/conf/model.conf

@@ -0,0 +1,10 @@
+--min-active=200
+--max-active=3000
+--beam=10.0
+--lattice-beam=2.0
+--acoustic-scale=1.0
+--frame-subsampling-factor=3
+--endpoint.silence-phones=1:2:3:4:5:6:7:8:9:10
+--endpoint.rule2.min-trailing-silence=0.5
+--endpoint.rule3.min-trailing-silence=0.75
+--endpoint.rule4.min-trailing-silence=1.0

model/vosk-model-small-en-us-0.15/graph/disambig_tid.int

@@ -0,0 +1,17 @@
+10015
+10016
+10017
+10018
+10019
+10020
+10021
+10022
+10023
+10024
+10025
+10026
+10027
+10028
+10029
+10030
+10031

model/vosk-model-small-en-us-0.15/graph/phones/word_boundary.int

@@ -0,0 +1,166 @@
+1 nonword
+2 begin
+3 end
+4 internal
+5 singleton
+6 nonword
+7 begin
+8 end
+9 internal
+10 singleton
+11 begin
+12 end
+13 internal
+14 singleton
+15 begin
+16 end
+17 internal
+18 singleton
+19 begin
+20 end
+21 internal
+22 singleton
+23 begin
+24 end
+25 internal
+26 singleton
+27 begin
+28 end
+29 internal
+30 singleton
+31 begin
+32 end
+33 internal
+34 singleton
+35 begin
+36 end
+37 internal
+38 singleton
+39 begin
+40 end
+41 internal
+42 singleton
+43 begin
+44 end
+45 internal
+46 singleton
+47 begin
+48 end
+49 internal
+50 singleton
+51 begin
+52 end
+53 internal
+54 singleton
+55 begin
+56 end
+57 internal
+58 singleton
+59 begin
+60 end
+61 internal
+62 singleton
+63 begin
+64 end
+65 internal
+66 singleton
+67 begin
+68 end
+69 internal
+70 singleton
+71 begin
+72 end
+73 internal
+74 singleton
+75 begin
+76 end
+77 internal
+78 singleton
+79 begin
+80 end
+81 internal
+82 singleton
+83 begin
+84 end
+85 internal
+86 singleton
+87 begin
+88 end
+89 internal
+90 singleton
+91 begin
+92 end
+93 internal
+94 singleton
+95 begin
+96 end
+97 internal
+98 singleton
+99 begin
+100 end
+101 internal
+102 singleton
+103 begin
+104 end
+105 internal
+106 singleton
+107 begin
+108 end
+109 internal
+110 singleton
+111 begin
+112 end
+113 internal
+114 singleton
+115 begin
+116 end
+117 internal
+118 singleton
+119 begin
+120 end
+121 internal
+122 singleton
+123 begin
+124 end
+125 internal
+126 singleton
+127 begin
+128 end
+129 internal
+130 singleton
+131 begin
+132 end
+133 internal
+134 singleton
+135 begin
+136 end
+137 internal
+138 singleton
+139 begin
+140 end
+141 internal
+142 singleton
+143 begin
+144 end
+145 internal
+146 singleton
+147 begin
+148 end
+149 internal
+150 singleton
+151 begin
+152 end
+153 internal
+154 singleton
+155 begin
+156 end
+157 internal
+158 singleton
+159 begin
+160 end
+161 internal
+162 singleton
+163 begin
+164 end
+165 internal
+166 singleton

model/vosk-model-small-en-us-0.15/ivector/global_cmvn.stats

@@ -0,0 +1,3 @@
+ [
+  1.682383e+11 -1.1595e+10 -1.521733e+10 4.32034e+09 -2.257938e+10 -1.969666e+10 -2.559265e+10 -1.535687e+10 -1.276854e+10 -4.494483e+09 -1.209085e+10 -5.64008e+09 -1.134847e+10 -3.419512e+09 -1.079542e+10 -4.145463e+09 -6.637486e+09 -1.11318e+09 -3.479773e+09 -1.245932e+08 -1.386961e+09 6.560655e+07 -2.436518e+08 -4.032432e+07 4.620046e+08 -7.714964e+07 9.551484e+08 -4.119761e+08 8.208582e+08 -7.117156e+08 7.457703e+08 -4.3106e+08 1.202726e+09 2.904036e+08 1.231931e+09 3.629848e+08 6.366939e+08 -4.586172e+08 -5.267629e+08 -3.507819e+08 1.679838e+09 
+  1.741141e+13 8.92488e+11 8.743834e+11 8.848896e+11 1.190313e+12 1.160279e+12 1.300066e+12 1.005678e+12 9.39335e+11 8.089614e+11 7.927041e+11 6.882427e+11 6.444235e+11 5.151451e+11 4.825723e+11 3.210106e+11 2.720254e+11 1.772539e+11 1.248102e+11 6.691599e+10 3.599804e+10 1.207574e+10 1.679301e+09 4.594778e+08 5.821614e+09 1.451758e+10 2.55803e+10 3.43277e+10 4.245286e+10 4.784859e+10 4.988591e+10 4.925451e+10 5.074584e+10 4.9557e+10 4.407876e+10 3.421443e+10 3.138606e+10 2.539716e+10 1.948134e+10 1.381167e+10 0 ]

model/vosk-model-small-en-us-0.15/ivector/online_cmvn.conf

@@ -0,0 +1 @@
+# configuration file for apply-cmvn-online, used in the script ../local/run_online_decoding.sh

model/vosk-model-small-en-us-0.15/ivector/splice.conf

@@ -0,0 +1,2 @@
+--left-context=3
+--right-context=3

requirements.txt

@@ -0,0 +1,4 @@
+fastapi
+uvicorn
+python-multipart
+vosk