Update space

.gitignore

@@ -0,0 +1,23 @@
+# See https://help.github.com/articles/ignoring-files/ for more about ignoring files.
+
+# dependencies
+*/node_modules
+/.pnp
+.pnp.js
+
+# testing
+/coverage
+
+# production
+/build
+
+# misc
+.DS_Store
+.env.local
+.env.development.local
+.env.test.local
+.env.production.local
+
+npm-debug.log*
+yarn-debug.log*
+yarn-error.log*

Dockerfile

@@ -0,0 +1,8 @@
+FROM python:3.10-slim
+WORKDIR /app
+COPY requirements.txt .
+RUN pip install -r requirements.txt
+COPY . .
+# Build and serve React
+RUN cd frontend && npm ci && npm run build
+CMD ["uvicorn", "app:app", "--host", "0.0.0.0", "--port", "7860"]

app.py

@@ -0,0 +1,56 @@
+from fastapi import FastAPI, HTTPException
+from fastapi.middleware.cors import CORSMiddleware
+from pydantic import BaseModel
+import json, os
+from evaluator import evaluate_model
+
+app = FastAPI()
+DB_PATH = "models_results.json"
+
+# Add CORS Middleware
+app.add_middleware(
+    CORSMiddleware,
+    allow_origins=["*"], 
+    allow_credentials=True,
+    allow_methods=["*"], 
+    allow_headers=["*"], 
+)
+
+class ModelIn(BaseModel):
+    model_name: str
+
+@app.get("/results")
+def get_results():
+    try:
+        with open(DB_PATH, "r") as f:
+            return json.load(f)
+    except (json.JSONDecodeError, FileNotFoundError):
+        # Return an empty list if the file is empty or missing
+        return []
+
+@app.post("/evaluate")
+def eval_and_store(req: ModelIn):
+    model_name = req.model_name
+    # 1. Check if already evaluated
+    data = json.load(open(DB_PATH))
+    if any(d["model"] == model_name for d in data):
+        raise HTTPException(400, "Model already evaluated")
+    # 2. Run evaluation
+    try:
+        metrics = evaluate_model(
+                    model_name=model_name,       # any 🤗 model ID
+                    dataset_name="sunbird/salt",     # your test split
+                    split="dev",                      # or whatever split you’ve prepared
+                )
+
+    except Exception as e:
+        raise HTTPException(500, f"Evaluation failed: {e}")
+    # 3. Append & save
+    data.append({"model": model_name, "metrics": metrics})
+    with open(DB_PATH, "w") as f:
+        json.dump(data, f, indent=2)
+    return {"status": "ok", "metrics": metrics}
+
+# Serve React's build folder
+from fastapi.staticfiles import StaticFiles
+app.mount("/", StaticFiles(directory="frontend/build", html=True), name="static")

evaluator.py

@@ -0,0 +1,167 @@
+# evaluator.py
+
+import numpy as np
+from collections import defaultdict
+from datasets import load_dataset
+from tqdm.auto import tqdm
+
+from sacrebleu.metrics import BLEU, CHRF
+from rouge_score import rouge_scorer
+import Levenshtein
+
+from transformers import pipeline
+from transformers.models.whisper.english_normalizer import BasicTextNormalizer
+
+
+def calculate_metrics(reference: str, prediction: str) -> dict:
+    """
+    Compute a suite of translation / generation metrics:
+      - BLEU
+      - chrF
+      - CER (character error rate)
+      - WER (word error rate)
+      - length ratio
+      - ROUGE-1 & ROUGE-L
+      - a combined quality_score
+    """
+    # BLEU
+    bleu = BLEU(effective_order=True)
+    bleu_score = bleu.sentence_score(prediction, [reference]).score
+
+    # chrF
+    chrf = CHRF()
+    chrf_score = chrf.sentence_score(prediction, [reference]).score / 100.0
+
+    # Character error rate
+    cer = Levenshtein.distance(reference, prediction) / max(len(reference), 1)
+
+    # Word error rate
+    ref_words = reference.split()
+    pred_words = prediction.split()
+    wer = Levenshtein.distance(ref_words, pred_words) / max(len(ref_words), 1)
+
+    # Length ratio
+    len_ratio = len(prediction) / max(len(reference), 1)
+
+    # ROUGE
+    rouge_scores = {}
+    try:
+        scorer = rouge_scorer.RougeScorer(["rouge1", "rougeL"], use_stemmer=True)
+        rouge_scores = scorer.score(reference, prediction)
+        rouge_1 = rouge_scores["rouge1"].fmeasure
+        rouge_L = rouge_scores["rougeL"].fmeasure
+    except Exception:
+        rouge_1 = rouge_L = 0.0
+
+    # Combined quality
+    try:
+        quality_score = (
+            (bleu_score / 100)
+            + chrf_score
+            + (1 - cer)
+            + (1 - wer)
+            + rouge_1
+            + rouge_L
+        ) / 6
+    except Exception:
+        quality_score = (
+            (bleu_score / 100) + chrf_score + (1 - cer) + (1 - wer)
+        ) / 4
+
+    return {
+        "bleu": bleu_score,
+        "chrf": chrf_score,
+        "cer": cer,
+        "wer": wer,
+        "len_ratio": len_ratio,
+        "rouge1": rouge_1,
+        "rougeL": rouge_L,
+        "quality_score": quality_score,
+    }
+
+
+def evaluate_model(
+    model_name: str,
+    dataset_name: str,
+    split: str = "test",
+    text_field: str = "source",
+    target_field: str = "target",
+    task: str = "translation",  # or "automatic-speech-recognition", etc.
+    device: int = 0,
+) -> dict:
+    """
+    Load your dataset, run inference via a 🤗 pipeline, and compute metrics
+    grouped by language‐pair (if present) plus overall averages.
+
+    Returns a dict of shape:
+      {
+        "<src>_to_<tgt>": {<metric1>: val, ...},
+        ...,
+        "averages": {<metric1>: val, ...}
+      }
+    """
+    # 1) load test split
+    ds = load_dataset(dataset_name, split=split)
+
+    # 2) build pipeline
+    nlp = pipeline(task, model=model_name, device=device)
+
+    # 3) run inference
+    normalizer = BasicTextNormalizer()
+    translations = []
+    for ex in tqdm(ds, desc=f"Eval {model_name}"):
+        src = ex[text_field]
+        tgt = ex[target_field]
+        pred = nlp(src)[0].get("translation_text", nlp(src)[0].get("text", ""))
+        translations.append({
+            "source": src,
+            "target": tgt,
+            "prediction": pred,
+            # Optional language metadata:
+            "source.language": ex.get("source.language", ""),
+            "target.language": ex.get("target.language", "")
+        })
+
+    # 4) group by language‐pair
+    subsets = defaultdict(list)
+    for ex in translations:
+        key = (
+            f"{ex['source.language']}_to_{ex['target.language']}"
+            if ex["source.language"] and ex["target.language"]
+            else "default"
+        )
+        subsets[key].append(ex)
+
+    # 5) compute metrics per subset
+    results = {}
+    for subset, examples in subsets.items():
+        # collect metrics lists
+        agg = defaultdict(list)
+        for ex in examples:
+            ref = normalizer(ex["target"])
+            pred = normalizer(ex["prediction"])
+            m = calculate_metrics(ref, pred)
+            for k, v in m.items():
+                agg[k].append(v)
+        # take mean
+        results[subset] = {k: float(np.mean(vs)) for k, vs in agg.items()}
+
+    # 6) overall averages
+    all_metrics = list(results.values())
+    avg = {}
+    for k in all_metrics[0].keys():
+        avg[k] = float(np.mean([m[k] for m in all_metrics]))
+    results["averages"] = avg
+
+    return results
+
+
+if __name__ == "__main__":
+    # simple test
+    import json
+    out = evaluate_model(
+        model_name="facebook/wmt19-en-de",
+        dataset_name="wmt19",
+        split="test",
+    )
+    print(json.dumps(out, indent=2))

frontend/package.json

@@ -0,0 +1,52 @@
+{
+  "name": "sunbird-leaderboard-v1",
+  "version": "0.1.0",
+  "private": true,
+  "dependencies": {
+    "@testing-library/dom": "^10.4.0",
+    "@testing-library/jest-dom": "^6.6.3",
+    "@testing-library/react": "^16.3.0",
+    "@testing-library/user-event": "^13.5.0",
+    "@types/jest": "^27.5.2",
+    "@types/node": "^16.18.126",
+    "@types/react": "^19.1.7",
+    "@types/react-dom": "^19.1.6",
+    "axios": "^1.9.0",
+    "react": "^19.1.0",
+    "react-dom": "^19.1.0",
+    "react-scripts": "5.0.1",
+    "recharts": "^2.15.3",
+    "typescript": "^4.9.5",
+    "web-vitals": "^2.1.4"
+  },
+  "scripts": {
+    "start": "react-scripts start",
+    "build": "react-scripts build",
+    "test": "react-scripts test",
+    "eject": "react-scripts eject"
+  },
+  "eslintConfig": {
+    "extends": [
+      "react-app",
+      "react-app/jest"
+    ]
+  },
+  "browserslist": {
+    "production": [
+      ">0.2%",
+      "not dead",
+      "not op_mini all"
+    ],
+    "development": [
+      "last 1 chrome version",
+      "last 1 firefox version",
+      "last 1 safari version"
+    ]
+  },
+  "devDependencies": {
+    "@types/axios": "^0.14.4",
+    "autoprefixer": "^10.4.21",
+    "postcss": "^8.5.4",
+    "tailwindcss": "^3.4.17"
+  }
+}

frontend/postcss.config.js

@@ -0,0 +1,6 @@
+module.exports = {
+  plugins: {
+    tailwindcss: {},
+    autoprefixer: {},
+  },
+};

frontend/public/index.html

@@ -0,0 +1,43 @@
+<!DOCTYPE html>
+<html lang="en">
+  <head>
+    <meta charset="utf-8" />
+    <link rel="icon" href="%PUBLIC_URL%/favicon.ico" />
+    <meta name="viewport" content="width=device-width, initial-scale=1" />
+    <meta name="theme-color" content="#000000" />
+    <meta
+      name="description"
+      content="Web site created using create-react-app"
+    />
+    <link rel="apple-touch-icon" href="%PUBLIC_URL%/logo192.png" />
+    <!--
+      manifest.json provides metadata used when your web app is installed on a
+      user's mobile device or desktop. See https://developers.google.com/web/fundamentals/web-app-manifest/
+    -->
+    <link rel="manifest" href="%PUBLIC_URL%/manifest.json" />
+    <!--
+      Notice the use of %PUBLIC_URL% in the tags above.
+      It will be replaced with the URL of the `public` folder during the build.
+      Only files inside the `public` folder can be referenced from the HTML.
+
+      Unlike "/favicon.ico" or "favicon.ico", "%PUBLIC_URL%/favicon.ico" will
+      work correctly both with client-side routing and a non-root public URL.
+      Learn how to configure a non-root public URL by running `npm run build`.
+    -->
+    <title>React App</title>
+  </head>
+  <body>
+    <noscript>You need to enable JavaScript to run this app.</noscript>
+    <div id="root"></div>
+    <!--
+      This HTML file is a template.
+      If you open it directly in the browser, you will see an empty page.
+
+      You can add webfonts, meta tags, or analytics to this file.
+      The build step will place the bundled scripts into the <body> tag.
+
+      To begin the development, run `npm start` or `yarn start`.
+      To create a production bundle, use `npm run build` or `yarn build`.
+    -->
+  </body>
+</html>

frontend/public/manifest.json

@@ -0,0 +1,25 @@
+{
+  "short_name": "React App",
+  "name": "Create React App Sample",
+  "icons": [
+    {
+      "src": "favicon.ico",
+      "sizes": "64x64 32x32 24x24 16x16",
+      "type": "image/x-icon"
+    },
+    {
+      "src": "logo192.png",
+      "type": "image/png",
+      "sizes": "192x192"
+    },
+    {
+      "src": "logo512.png",
+      "type": "image/png",
+      "sizes": "512x512"
+    }
+  ],
+  "start_url": ".",
+  "display": "standalone",
+  "theme_color": "#000000",
+  "background_color": "#ffffff"
+}

frontend/public/robots.txt

@@ -0,0 +1,3 @@
+# https://www.robotstxt.org/robotstxt.html
+User-agent: *
+Disallow:

frontend/requirements.txt

@@ -0,0 +1,21 @@
+fastapi
+uvicorn
+pydantic
+
+# Core dependencies
+datasets>=2.14.0
+sacrebleu>=2.3.0
+rouge-score>=0.1.2
+Levenshtein>=0.21.0
+numpy>=1.24.0
+pandas>=2.0.0
+matplotlib>=3.7.0
+huggingface_hub>=0.16.0
+PyYAML>=6.0
+plotly>=5.0.0
+google-cloud-translate>=3.0.0
+ipython
+
+# Additional dependencies that SALT might need
+transformers>=4.30.0
+torch>=2.0.0

frontend/src/App.tsx

@@ -0,0 +1,220 @@
+import React, { useState, useMemo } from 'react';
+import { useResults } from './hooks/useResults';
+import ModelSelector from './components/ModelSelector';
+import MetricCard from './components/MetricCard';
+import BarChartView from './components/BarChartView';
+import RadarView from './components/RadarView';
+import LineTrendView from './components/LineTrendView';
+import ComparisonTable from './components/ComparisonTable';
+import AddModelForm from './components/AddModelForm';
+import { ChartRow, ModelMetrics } from './types';
+
+const colors = ['#4F46E5', '#059669', '#DC2626', '#7C3AED', '#EA580C', '#0891B2'];
+const metrics = ['bleu','chrf','cer','wer','rouge1','rouge2','rougeL','quality_score'];
+
+function App() {
+  const { data, error } = useResults();
+  const [selectedMetric, setSelectedMetric] = useState(metrics[0]);
+  const [selectedModel, setSelectedModel] = useState('');
+  const [selectedModels, setSelectedModels] = useState<string[]>([]);
+  const [comparisonMode, setComparisonMode] = useState(false);
+  const [viewMode, setViewMode] = useState<'overview' | 'detailed'>('overview');
+  const [activeTab, setActiveTab] = useState<'dashboard' | 'addModel' | 'comparison'>('dashboard');
+  const [newModelName, setNewModelName] = useState('');
+  const [isEvaluating, setIsEvaluating] = useState(false);
+
+  const modelNames = data ? Object.keys(data) : [];
+  const languagePairs = data && selectedModel && data[selectedModel] 
+    ? Object.keys(data[selectedModel]).filter(k => k !== 'averages') 
+    : [];
+  const barChartData: ChartRow[] = useMemo(() => {
+    if (!data || !selectedModel || !data[selectedModel]) return [];
+    return languagePairs.map(pair => {
+      const row: any = { name: pair.toUpperCase() };
+      if (comparisonMode) {
+        selectedModels.forEach((m, i) => {
+          row[m] = data[m]?.[pair]?.[selectedMetric as keyof ModelMetrics] ?? 0;
+        });
+      } else {
+        row[selectedMetric] = data[selectedModel]?.[pair]?.[selectedMetric as keyof ModelMetrics] ?? 0;
+      }
+      return row;
+    });
+  }, [data, languagePairs, selectedMetric, comparisonMode, selectedModels]);
+
+  const radarData: ChartRow[] = useMemo(() => {
+    if (!data || !selectedModel || !data[selectedModel]) return [];
+    return metrics.map(metric => {
+      const point: any = { metric: metric.toUpperCase() };
+      if (comparisonMode) {
+        selectedModels.forEach((m, i) => {
+          point[m] = data[m]?.averages?.[metric as keyof ModelMetrics] ?? 0;
+        });
+      } else {
+        point.value = data[selectedModel]?.averages?.[metric as keyof ModelMetrics] ?? 0;
+      }
+      return point;
+    });
+  }, [data, selectedModel, selectedModels, comparisonMode]);
+
+  if (error) return <div className="p-6 text-red-600">Error: {error}</div>;
+  if (!data) return <div className="p-6">Loading...</div>;
+
+  return (
+    <div className="min-h-screen bg-gradient-to-br from-blue-50 to-indigo-100">
+      {/* Main Navigation Tabs */}
+      <div className="bg-white shadow-sm border-b">
+        <div className="max-w-7xl mx-auto px-4 sm:px-6 lg:px-8">
+          <div className="flex justify-between items-center py-6">
+            <h1 className="text-3xl font-bold text-gray-900">Translation Leaderboard</h1>
+            <div className="flex space-x-4">
+              <button
+                onClick={() => setActiveTab('dashboard')}
+                className={`px-4 py-2 rounded-lg font-medium ${
+                  activeTab === 'dashboard' ? 'bg-indigo-600 text-white' : 'bg-gray-200 text-gray-700 hover:bg-gray-300'
+                }`}
+              >
+                Dashboard
+              </button>
+              <button
+                onClick={() => setActiveTab('comparison')}
+                className={`px-4 py-2 rounded-lg font-medium ${
+                  activeTab === 'comparison' ? 'bg-indigo-600 text-white' : 'bg-gray-200 text-gray-700 hover:bg-gray-300'
+                }`}
+              >
+                Comparison
+              </button>
+              <button
+                onClick={() => setActiveTab('addModel')}
+                className={`px-4 py-2 rounded-lg font-medium ${
+                  activeTab === 'addModel' ? 'bg-indigo-600 text-white' : 'bg-gray-200 text-gray-700 hover:bg-gray-300'
+                }`}
+              >
+                Add Model
+              </button>
+            </div>
+          </div>
+        </div>
+      </div>
+
+      {/* Content Based on Active Tab */}
+      <div className="max-w-7xl mx-auto px-4 sm:px-6 lg:px-8 py-8">
+        {activeTab === 'dashboard' && (
+          <div>
+            <h2 className="text-2xl font-bold mb-4">Dashboard</h2>
+            {/* Toggle Button for Comparison Mode */}
+            <div className="mb-6">
+              <button
+                onClick={() => setComparisonMode(!comparisonMode)}
+                className={`px-4 py-2 rounded-lg font-medium ${
+                  comparisonMode ? 'bg-red-600 text-white' : 'bg-indigo-600 text-white'
+                }`}
+              >
+                {comparisonMode ? 'Disable Comparison Mode' : 'Enable Comparison Mode'}
+              </button>
+            </div>
+
+            <ModelSelector
+              modelNames={modelNames}
+              selectedModel={selectedModel}
+              selectedModels={selectedModels}
+              comparisonMode={comparisonMode}
+              onSelectModel={setSelectedModel}
+              onToggleModel={m =>
+                setSelectedModels(prev =>
+                  prev.includes(m) ? prev.filter(x => x !== m) : [...prev, m]
+                )
+              }
+            />
+
+            <div className="grid grid-cols-1 lg:grid-cols-4 gap-6 mb-8">
+              {/* Summary cards */}
+              {!comparisonMode ? (
+                <>
+                  {data[selectedModel]?.averages && (
+                    <MetricCard
+                      title="Average BLEU"
+                      value={data[selectedModel].averages.bleu}
+                      description="Translation quality"
+                      colorHex={colors[0]}
+                    />
+                  )}
+                  {/* Other cards */}
+                </>
+              ) : (
+                selectedModels.map((m, i) => (
+                  data[m]?.averages && (
+                    <MetricCard
+                      key={m}
+                      title="Avg BLEU"
+                      value={data[m].averages.bleu}
+                      description={m}
+                      colorHex={colors[i]}
+                    />
+                  )
+                ))
+              )}
+            </div>
+
+            {viewMode === 'overview' ? (
+              <>
+                
+                <ComparisonTable data={data} metrics={metrics} />
+                <BarChartView
+                  data={barChartData}
+                  metrics={metrics}
+                  selectedMetric={selectedMetric}
+                  onMetricChange={setSelectedMetric}
+                  comparisonMode={comparisonMode}
+                  colors={colors}
+                  selectedModels={selectedModels}
+                />
+                
+              </>
+            ) : (
+              <>
+                <RadarView data={radarData} comparisonMode={comparisonMode} colors={colors} />
+                <LineTrendView data={barChartData} selectedModels={selectedModels} comparisonMode={comparisonMode} colors={colors} />
+              </>
+            )}
+          </div>
+        )}
+        {activeTab === 'addModel' && (
+          <div>
+            <h2 className="text-2xl font-bold mb-4">Add a New Model</h2>
+            <AddModelForm
+              newModelName={newModelName}
+              isEvaluating={isEvaluating}
+              onNameChange={setNewModelName}
+              onEvaluate={() => {
+                // Add evaluation logic here
+              }}
+            />
+          </div>
+        )}
+        {activeTab === 'comparison' && (
+          <div>
+            <h2 className="text-2xl font-bold mb-4">Comparison</h2>
+            <ModelSelector
+              modelNames={modelNames}
+              selectedModel={selectedModel}
+              selectedModels={selectedModels}
+              comparisonMode={comparisonMode}
+              onSelectModel={setSelectedModel}
+              onToggleModel={m =>
+                setSelectedModels(prev =>
+                  prev.includes(m) ? prev.filter(x => x !== m) : [...prev, m]
+                )
+              }
+            />
+            <RadarView data={radarData} comparisonMode={comparisonMode} colors={colors} />
+            <ComparisonTable data={data} metrics={metrics} />
+            <LineTrendView data={barChartData} selectedModels={selectedModels} comparisonMode={comparisonMode} colors={colors} />
+          </div>
+        )}
+      </div>
+    </div>
+  );
+}
+
+export default App;

frontend/src/api/translation.ts

@@ -0,0 +1,15 @@
+import axios from 'axios';
+import { ModelData, ModelMetrics } from '../types';
+
+const api = axios.create({
+baseURL: process.env.REACT_APP_API_URL,
+timeout: 10000,
+});
+
+export interface ApiResult {
+  model: string;
+  metrics: ModelData;
+}
+
+export const fetchResults = () => api.get<ApiResult[]>('/results');
+export const evaluateModel = (model_name: string) => api.post<{ metrics: ModelMetrics }>('/evaluate', { model_name });

frontend/src/components/AddModelForm.tsx

@@ -0,0 +1,34 @@
+import React, { FC } from 'react';
+
+interface Props {
+  newModelName: string;
+  isEvaluating: boolean;
+  onNameChange: (value: string) => void;
+  onEvaluate: () => void;
+}
+
+const AddModelForm: FC<Props> = ({ newModelName, isEvaluating, onNameChange, onEvaluate }) => (
+  <div className="bg-white rounded-xl shadow-lg p-6">
+    <h2 className="text-2xl font-bold text-gray-800 mb-6">Add a New Model</h2>
+    <div className="space-y-4">
+      <input
+        type="text"
+        value={newModelName}
+        onChange={e => onNameChange(e.target.value)}
+        placeholder="Enter model name (e.g., facebook/wmt19-en-de)"
+        className="w-full px-4 py-2 border border-gray-300 rounded-lg focus:outline-none focus:ring-2 focus:ring-indigo-500"
+      />
+      <button
+        onClick={onEvaluate}
+        disabled={isEvaluating}
+        className={`px-4 py-2 rounded-lg font-medium text-white ${
+          isEvaluating ? 'bg-gray-400' : 'bg-indigo-600 hover:bg-indigo-700'
+        }`}
+      >
+        {isEvaluating ? 'Evaluating...' : 'Evaluate Model'}
+      </button>
+    </div>
+  </div>
+);
+
+export default AddModelForm;

frontend/src/components/App.css

@@ -0,0 +1,38 @@
+.App {
+  text-align: center;
+}
+
+.App-logo {
+  height: 40vmin;
+  pointer-events: none;
+}
+
+@media (prefers-reduced-motion: no-preference) {
+  .App-logo {
+    animation: App-logo-spin infinite 20s linear;
+  }
+}
+
+.App-header {
+  background-color: #282c34;
+  min-height: 100vh;
+  display: flex;
+  flex-direction: column;
+  align-items: center;
+  justify-content: center;
+  font-size: calc(10px + 2vmin);
+  color: white;
+}
+
+.App-link {
+  color: #61dafb;
+}
+
+@keyframes App-logo-spin {
+  from {
+    transform: rotate(0deg);
+  }
+  to {
+    transform: rotate(360deg);
+  }
+}

frontend/src/components/App.test.tsx

@@ -0,0 +1,9 @@
+import React from 'react';
+import { render, screen } from '@testing-library/react';
+import TranslationDashboard from '../App';
+
+test('renders evaluation results heading', () => {
+  render(<TranslationDashboard />);
+  const headingElement = screen.getByText(/evaluation results/i);
+  expect(headingElement).toBeInTheDocument();
+});

frontend/src/components/BarChartView.tsx

@@ -0,0 +1,68 @@
+import React, { FC } from 'react';
+import {
+  ResponsiveContainer,
+  BarChart,
+  CartesianGrid,
+  XAxis,
+  YAxis,
+  Tooltip,
+  Legend,
+  Bar,
+} from 'recharts';
+import { ChartRow } from '../types';
+
+interface Props {
+  data: ChartRow[];
+  metrics: string[];
+  selectedMetric: string;
+  onMetricChange: (metric: string) => void;
+  comparisonMode: boolean;
+  colors: string[];
+  selectedModels: string[];
+}
+
+const BarChartView: FC<Props> = ({
+  data,
+  metrics,
+  selectedMetric,
+  onMetricChange,
+  comparisonMode,
+  colors,
+  selectedModels,
+}) => (
+  <div className="bg-white rounded-xl shadow-lg p-6 mb-8">
+    <div className="flex justify-between items-center mb-6">
+      <h2 className="text-2xl font-bold text-gray-800">
+        {comparisonMode ? 'Model Comparison' : 'Performance by Language Pair'}
+      </h2>
+      <select
+        value={selectedMetric}
+        onChange={e => onMetricChange(e.target.value)}
+        className="px-4 py-2 border border-gray-300 rounded-lg focus:outline-none focus:ring-2 focus:ring-indigo-500"
+      >
+        {metrics.map(metric => (
+          <option key={metric} value={metric}>
+            {metric.toUpperCase()}
+          </option>
+        ))}
+      </select>
+    </div>
+
+    <ResponsiveContainer width="100%" height={400}>
+      <BarChart data={data}>
+        <CartesianGrid strokeDasharray="3 3" />
+        <XAxis dataKey="name" angle={-45} textAnchor="end" height={100} />
+        <YAxis />
+        <Tooltip />
+        <Legend />
+        {comparisonMode
+          ? selectedModels.map((model, i) => (
+              <Bar key={model} dataKey={model} fill={colors[i % colors.length]} radius={4} />
+            ))
+          : <Bar dataKey={selectedMetric} fill={colors[0]} radius={4} />}
+      </BarChart>
+    </ResponsiveContainer>
+  </div>
+);
+
+export default BarChartView;

frontend/src/components/ComparisonTable.tsx

@@ -0,0 +1,41 @@
+import React, { FC } from 'react';
+import { AllModelsData } from '../types';
+
+interface Props {
+  data: AllModelsData;
+  metrics: string[];
+}
+
+const ComparisonTable: FC<Props> = ({ data, metrics }) => {
+  const models = Object.keys(data);
+
+  return (
+    <div className="bg-white rounded-xl shadow-lg p-6 overflow-x-auto">
+      <h2 className="text-2xl font-bold text-gray-800 mb-6">Model Comparison Table</h2>
+      <table className="w-full text-sm">
+        <thead>
+          <tr className="border-b border-gray-200">
+            <th className="text-left py-3 px-4 font-semibold">Model</th>
+            {metrics.map(m => (
+              <th key={m} className="text-center py-3 px-4 font-semibold">{m.toUpperCase()}</th>
+            ))}
+          </tr>
+        </thead>
+        <tbody>
+          {models.map(model => (
+            <tr key={model} className="border-b border-gray-100 hover:bg-gray-50">
+              <td className="py-3 px-4 font-medium">{model}</td>
+              {metrics.map(metric => (
+                <td key={metric} className="text-center py-3 px-4">
+                  {data[model].averages[metric as keyof typeof data[string]['averages']].toFixed(3)}
+                </td>
+              ))}
+            </tr>
+          ))}
+        </tbody>
+      </table>
+    </div>
+  );
+};
+
+export default ComparisonTable;

frontend/src/components/LineTrendView.tsx

@@ -0,0 +1,44 @@
+import React, { FC } from 'react';
+import { ResponsiveContainer, LineChart, CartesianGrid, XAxis, YAxis, Tooltip, Legend, Line } from 'recharts';
+import { ChartRow } from '../types';
+
+interface Props {
+  data: ChartRow[];
+  selectedModels: string[];
+  comparisonMode: boolean;
+  colors: string[];
+}
+
+const LineTrendView: FC<Props> = ({ data, selectedModels, comparisonMode, colors }) => (
+  <div className="bg-white rounded-xl shadow-lg p-6 mb-8">
+    <h2 className="text-2xl font-bold text-gray-800 mb-6">Performance Trends</h2>
+    <ResponsiveContainer width="100%" height={400}>
+      <LineChart data={data}>
+        <CartesianGrid strokeDasharray="3 3" />
+        <XAxis dataKey="name" />
+        <YAxis />
+        <Tooltip />
+        <Legend />
+        {comparisonMode
+          ? selectedModels.map((model, index) => (
+              <Line
+                key={model}
+                type="monotone"
+                dataKey={model}
+                stroke={colors[index % colors.length]} // Cycle through colors
+                strokeWidth={3}
+                name={model} // Display model name in the legend
+              />
+            ))
+          : (
+            <>
+              <Line type="monotone" dataKey="bleu" stroke="#4F46E5" strokeWidth={3} />
+              <Line type="monotone" dataKey="quality_score" stroke="#059669" strokeWidth={3} />
+            </>
+          )}
+      </LineChart>
+    </ResponsiveContainer>
+  </div>
+);
+
+export default LineTrendView;

frontend/src/components/MetricCard.tsx

@@ -0,0 +1,23 @@
+import React, { FC } from 'react';
+
+interface Props {
+  title: string;
+  value: number | string;
+  description: string;
+  colorHex: string;
+  subtitle?: string;
+}
+
+const MetricCard: FC<Props> = ({ title, value, description, colorHex, subtitle }) => (
+  <div
+    className="bg-white rounded-xl shadow-lg p-6 transition-shadow hover:shadow-xl"
+    style={{ borderLeft: `4px solid ${colorHex}` }}
+  >
+    <h3 className="text-lg font-semibold text-gray-800 mb-1">{title}</h3>
+    {subtitle && <p className="text-xs text-gray-500 mb-2">{subtitle}</p>}
+    <div className="text-3xl font-bold mb-2">{typeof value === 'number' ? value.toFixed(3) : value}</div>
+    <p className="text-sm text-gray-600">{description}</p>
+  </div>
+);
+
+export default MetricCard;

frontend/src/components/ModelSelector.tsx

@@ -0,0 +1,56 @@
+import React from 'react';
+import { FC } from 'react';
+
+interface Props {
+  modelNames: string[];
+  selectedModel: string;
+  selectedModels: string[];
+  comparisonMode: boolean;
+  onSelectModel: (model: string) => void;
+  onToggleModel: (model: string) => void;
+}
+
+const ModelSelector: FC<Props> = ({
+  modelNames,
+  selectedModel,
+  selectedModels,
+  comparisonMode,
+  onSelectModel,
+  onToggleModel,
+}) => (
+  <div className="bg-white rounded-lg shadow p-4 border">
+    <h3 className="font-semibold text-gray-800 mb-3">
+      {comparisonMode ? 'Select Models to Compare' : 'Select Model'}
+    </h3>
+
+    {comparisonMode ? (
+      <div className="space-y-2">
+        {modelNames.map(model => (
+          <label key={model} className="flex items-center space-x-2">
+            <input
+              type="checkbox"
+              checked={selectedModels.includes(model)}
+              onChange={() => onToggleModel(model)}
+              className="rounded border-gray-300 text-indigo-600 focus:ring-indigo-500"
+            />
+            <span className="text-sm text-gray-700">{model}</span>
+          </label>
+        ))}
+      </div>
+    ) : (
+      <select
+        value={selectedModel}
+        onChange={e => onSelectModel(e.target.value)}
+        className="w-full px-3 py-2 border border-gray-300 rounded-lg focus:outline-none focus:ring-2 focus:ring-indigo-500"
+      >
+        {modelNames.map(model => (
+          <option key={model} value={model}>
+            {model}
+          </option>
+        ))}
+      </select>
+    )}
+  </div>
+);
+
+export default ModelSelector;

frontend/src/components/RadarView.tsx

@@ -0,0 +1,50 @@
+import React, { FC } from 'react';
+import {
+  ResponsiveContainer,
+  RadarChart,
+  PolarGrid,
+  PolarAngleAxis,
+  PolarRadiusAxis,
+  Radar,
+  Legend,
+} from 'recharts';
+import { ChartRow } from '../types';
+
+interface Props {
+  data: ChartRow[];
+  comparisonMode: boolean;
+  colors: string[];
+}
+
+const RadarView: FC<Props> = ({ data, comparisonMode, colors }) => (
+  <div className="bg-white rounded-xl shadow-lg p-6 mb-8">
+    <h2 className="text-2xl font-bold text-gray-800 mb-6">
+      {comparisonMode ? 'Multi-Model Radar' : 'Metric Distribution'}
+    </h2>
+    <ResponsiveContainer width="100%" height={400}>
+      <RadarChart data={data}>
+        <PolarGrid />
+        <PolarAngleAxis dataKey="metric" />
+        <PolarRadiusAxis angle={90} domain={[0, 1]} />
+        {comparisonMode
+          ? data[0] /* assume all rows have same keys */ &&
+            Object.keys(data[0])
+              .filter(k => k !== 'metric')
+              .map((key, i) => (
+                <Radar
+                  key={key}
+                  name={key}
+                  dataKey={key}
+                  stroke={colors[i % colors.length]}
+                  fill={colors[i % colors.length]}
+                  fillOpacity={0.3}
+                />
+              ))
+          : <Radar name="Performance" dataKey="value" stroke={colors[0]} fill={colors[0]} fillOpacity={0.3} />}
+        <Legend />
+      </RadarChart>
+    </ResponsiveContainer>
+  </div>
+);
+
+export default RadarView;

frontend/src/hooks/useChartData.ts

@@ -0,0 +1,46 @@
+import { useMemo } from 'react';
+import { ChartRow, AllModelsData, ModelMetrics } from '../types';
+
+export function useChartData(
+  data: AllModelsData | null,
+  selectedModel: string,
+  selectedModels: string[],
+  selectedMetric: string,
+  comparisonMode: boolean,
+  metrics: string[]
+) {
+  const barChartData: ChartRow[] = useMemo(() => {
+    if (!data || (!selectedModel && selectedModels.length === 0)) return [];
+    const languagePairs = selectedModel && data[selectedModel]
+      ? Object.keys(data[selectedModel]).filter(k => k !== 'averages')
+      : [];
+    return languagePairs.map(pair => {
+      const row: any = { name: pair.toUpperCase() };
+      if (comparisonMode) {
+        selectedModels.forEach(model => {
+          row[model] = data[model]?.[pair]?.[selectedMetric as keyof ModelMetrics] ?? 0;
+        });
+      } else {
+        row[selectedMetric] = data[selectedModel]?.[pair]?.[selectedMetric as keyof ModelMetrics] ?? 0;
+      }
+      return row;
+    });
+  }, [data, selectedModel, selectedModels, selectedMetric, comparisonMode]);
+
+  const radarData: ChartRow[] = useMemo(() => {
+    if (!data || (!selectedModel && selectedModels.length === 0)) return [];
+    return metrics.map(metric => {
+      const point: any = { metric: metric.toUpperCase() };
+      if (comparisonMode) {
+        selectedModels.forEach(model => {
+          point[model] = data[model]?.averages?.[metric as keyof ModelMetrics] ?? 0;
+        });
+      } else {
+        point.value = data[selectedModel]?.averages?.[metric as keyof ModelMetrics] ?? 0;
+      }
+      return point;
+    });
+  }, [data, selectedModel, selectedModels, comparisonMode]);
+
+  return { barChartData, radarData };
+}

frontend/src/hooks/useResults.ts

@@ -0,0 +1,26 @@
+import { useState, useEffect } from 'react';
+import axios from 'axios';
+import { AllModelsData, ModelMetrics } from '../types';
+
+export function useResults() {
+  const [data, setData] = useState<AllModelsData | null>(null);
+  const [error, setError] = useState<string | null>(null);
+
+  useEffect(() => {
+    axios
+      .get<{ model: string; metrics: AllModelsData[string] }[]>('http://localhost:8000/results')
+      .then(res => {
+        const transformed: AllModelsData = {};
+        res.data.forEach(item => {
+          transformed[item.model] = item.metrics;
+        });
+        setData(transformed);
+      })
+      .catch(err => {
+        console.error(err);
+        setError(err.message);
+      });
+  }, []);
+
+  return { data, error };
+}

frontend/src/index.css

@@ -0,0 +1,3 @@
+@tailwind base;
+@tailwind components;
+@tailwind utilities;

frontend/src/index.tsx

@@ -0,0 +1,19 @@
+import React from 'react';
+import ReactDOM from 'react-dom/client';
+import './index.css';
+import App from './App';
+import reportWebVitals from './reportWebVitals';
+
+const root = ReactDOM.createRoot(
+  document.getElementById('root') as HTMLElement
+);
+root.render(
+  <React.StrictMode>
+    <App />
+  </React.StrictMode>
+);
+
+// If you want to start measuring performance in your app, pass a function
+// to log results (for example: reportWebVitals(console.log))
+// or send to an analytics endpoint. Learn more: https://bit.ly/CRA-vitals
+reportWebVitals();

frontend/src/react-app-env.d.ts

@@ -0,0 +1 @@
+/// <reference types="react-scripts" />

frontend/src/reportWebVitals.ts

@@ -0,0 +1,15 @@
+import { ReportHandler } from 'web-vitals';
+
+const reportWebVitals = (onPerfEntry?: ReportHandler) => {
+  if (onPerfEntry && onPerfEntry instanceof Function) {
+    import('web-vitals').then(({ getCLS, getFID, getFCP, getLCP, getTTFB }) => {
+      getCLS(onPerfEntry);
+      getFID(onPerfEntry);
+      getFCP(onPerfEntry);
+      getLCP(onPerfEntry);
+      getTTFB(onPerfEntry);
+    });
+  }
+};
+
+export default reportWebVitals;

frontend/src/setupTests.ts

@@ -0,0 +1,5 @@
+// jest-dom adds custom jest matchers for asserting on DOM nodes.
+// allows you to do things like:
+// expect(element).toHaveTextContent(/react/i)
+// learn more: https://github.com/testing-library/jest-dom
+import '@testing-library/jest-dom';

frontend/src/types.ts

@@ -0,0 +1,31 @@
+export interface ModelMetrics {
+  bleu: number;
+  chrf: number;
+  cer: number;
+  wer: number;
+  len_ratio: number;
+  rouge1: number;
+  rouge2: number;
+  rougeL: number;
+  quality_score: number;
+}
+
+export interface ModelData {
+  [pair: string]: ModelMetrics;
+  averages: ModelMetrics;
+}
+
+export interface AllModelsData {
+  [model: string]: ModelData;
+}
+
+// For bar/radar chart rows:
+export interface ChartRow {
+  name: string;
+  [key: string]: number | string;
+}
+// API types
+export interface ApiResult {
+  model: string;
+  metrics: ModelData;
+}

frontend/tailwind.config.js

@@ -0,0 +1,9 @@
+module.exports = {
+  content: [
+    "./src/**/*.{js,jsx,ts,tsx}",
+  ],
+  theme: {
+    extend: {},
+  },
+  plugins: [],
+}

frontend/tsconfig.json

@@ -0,0 +1,26 @@
+{
+  "compilerOptions": {
+    "target": "es5",
+    "lib": [
+      "dom",
+      "dom.iterable",
+      "esnext"
+    ],
+    "allowJs": true,
+    "skipLibCheck": true,
+    "esModuleInterop": true,
+    "allowSyntheticDefaultImports": true,
+    "strict": true,
+    "forceConsistentCasingInFileNames": true,
+    "noFallthroughCasesInSwitch": true,
+    "module": "esnext",
+    "moduleResolution": "node",
+    "resolveJsonModule": true,
+    "isolatedModules": true,
+    "noEmit": true,
+    "jsx": "react-jsx"
+  },
+  "include": [
+    "src"
+  ]
+}

index.html

@@ -1,19 +0,0 @@
-<!doctype html>
-<html>
-	<head>
-		<meta charset="utf-8" />
-		<meta name="viewport" content="width=device-width" />
-		<title>My static Space</title>
-		<link rel="stylesheet" href="style.css" />
-	</head>
-	<body>
-		<div class="card">
-			<h1>Welcome to your static Space!</h1>
-			<p>You can modify this app directly by editing <i>index.html</i> in the Files and versions tab.</p>
-			<p>
-				Also don't forget to check the
-				<a href="https://huggingface.co/docs/hub/spaces" target="_blank">Spaces documentation</a>.
-			</p>
-		</div>
-	</body>
-</html>

models_results.json

@@ -0,0 +1,692 @@
+[
+  {
+    "model": "sunbird-qwen3-14b",
+    "metrics": {
+      "ach_to_eng": {"bleu": 29.893676951301053, "chrf": 0.533224041738049, "cer": 0.43550659508514755, "wer": 0.5722814047840733, "len_ratio": 0.9910228026082467, "rouge1": 0.6013037471425201, "rouge2": 0.37721216654425455, "rougeL": 0.5810190084999007, "quality_score": 0.45609320284545973}, "lgg_to_eng": {"bleu": 28.599782594820592, "chrf": 0.5066369217667385, "cer": 0.4605144560053802, "wer": 0.6027958508390197, "len_ratio": 0.993760271647737, "rouge1": 0.5712775529284185, "rouge2": 0.35528047833025683, "rougeL": 0.5498622681671385, "quality_score": 0.43233111021763604}, "lug_to_eng": {"bleu": 42.967284520037744, "chrf": 0.651902433792848, "cer": 0.31620633974304335, "wer": 0.41791775685573124, "len_ratio": 0.984599994739042, "rouge1": 0.7187322596678193, "rouge2": 0.5173399529343762, "rougeL": 0.6987033563469739, "quality_score": 0.5868627955986127}, "nyn_to_eng": {"bleu": 32.67985629646372, "chrf": 0.5555573573119897, "cer": 0.3996226614650576, "wer": 0.5226613071726924, "len_ratio": 0.9744107912589968, "rouge1": 0.6315633927145826, "rouge2": 0.4089197940759449, "rougeL": 0.6113888753094371, "quality_score": 0.4900179879097192}, "teo_to_eng": {"bleu": 32.52172348369896, "chrf": 0.5575641337333308, "cer": 0.3884327697644638, "wer": 0.5033956827288778, "len_ratio": 0.9628715760526139, "rouge1": 0.641517230418375, "rouge2": 0.40838418032033685, "rougeL": 0.6214971732967394, "quality_score": 0.4977382290192447}, "eng_to_ach": {"bleu": 21.477941285172513, "chrf": 0.5138962223965714, "cer": 0.45657951265182434, "wer": 0.6755567100553327, "len_ratio": 1.014999517783312, "rouge1": 0.5099384983089625, "rouge2": 0.2756919890353192, "rougeL": 0.4908058982934093, "quality_score": 0.3991348531352849}, "eng_to_lgg": {"bleu": 22.041535601523726, "chrf": 0.5255108356738902, "cer": 0.4422267716935427, "wer": 0.6713858470124049, "len_ratio": 1.0313949027800307, "rouge1": 0.5456449443461951, "rouge2": 0.29382498850599353, "rougeL": 0.5186337898829078, "quality_score": 0.4080783932457949}, "eng_to_lug": {"bleu": 33.34672372451176, "chrf": 0.6633356427761459, "cer": 0.3311583312643587, "wer": 0.5238451497835392, "len_ratio": 1.018582055891005, "rouge1": 0.594388873187116, "rouge2": 0.3909919061294654, "rougeL": 0.5833949344122611, "quality_score": 0.5354498497433413}, "eng_to_nyn": {"bleu": 20.79200082458506, "chrf": 0.5729842003065381, "cer": 0.4113873068671208, "wer": 0.7081659294479503, "len_ratio": 1.0317490539724605, "rouge1": 0.44584438768284473, "rouge2": 0.23085276156139556, "rougeL": 0.4373444327375021, "quality_score": 0.4153377430593293}, "eng_to_teo": {"bleu": 23.725735344718192, "chrf": 0.5832096756543237, "cer": 0.4151187089479109, "wer": 0.6281833981430756, "len_ratio": 1.0613714854829055, "rouge1": 0.5195866229204313, "rouge2": 0.2696033999485182, "rougeL": 0.5075491305820932, "quality_score": 0.4442912305026298}, "eng_to_xog": {"bleu": 19.052981317638974, "chrf": 0.5580375571166009, "cer": 0.40660034808436624, "wer": 0.7315224980651331, "len_ratio": 1.0173192546265903, "rouge1": 0.4091926881564407, "rouge2": 0.20611967973474055, "rougeL": 0.404450321174341, "quality_score": 0.4026111310358727}, "eng_to_ttj": {"bleu": 25.074282455874666, "chrf": 0.5852471141018765, "cer": 0.39029555071580396, "wer": 0.6357081997632875, "len_ratio": 1.0276635889601655, "rouge1": 0.5029403203567027, "rouge2": 0.2783539093458642, "rougeL": 0.49312154442651684, "quality_score": 0.4524965470453829}, "eng_to_swa": {"bleu": 39.69206979218379, "chrf": 0.6857228225417764, "cer": 0.2964009389587315, "wer": 0.44971614127360093, "len_ratio": 1.0368448616477834, "rouge1": 0.6679250784200024, "rouge2": 0.4586999764112653, "rougeL": 0.6501414308327976, "quality_score": 0.5841316100578205}, "xog_to_eng": {"bleu": 32.92079591468753, "chrf": 0.5641617154528973, "cer": 0.3839540332720533, "wer": 0.505736384076339, "len_ratio": 0.9812665218521085, "rouge1": 0.6347508404269397, "rouge2": 0.40558927069321776, "rougeL": 0.6153655184656368, "quality_score": 0.500919814312845}, "ttj_to_eng": {"bleu": 35.08945968981431, "chrf": 0.5805360317055387, "cer": 0.37665824404943016, "wer": 0.4906126467618403, "len_ratio": 0.9951291908976828, "rouge1": 0.6509396367141043, "rouge2": 0.4360350381677252, "rougeL": 0.6347547806711414, "quality_score": 0.5160399344481028}, "swa_to_eng": {"bleu": 50.16010559961865, "chrf": 0.7101907099077821, "cer": 0.2658771336442633, "wer": 0.34246272699338404, "len_ratio": 0.9955021465731141, "rouge1": 0.7722154807714764, "rouge2": 0.5929997953476355, "rougeL": 0.7507740512693895, "quality_score": 0.6508629763165803}, "averages": {"bleu": 30.627247212290705, "chrf": 0.5842323384985562, "cer": 0.3860337313882811, "wer": 0.5613717271097676, "len_ratio": 1.0074055010483622, "rouge1": 0.5886100971351833, "rouge2": 0.3691187054428943, "rougeL": 0.5718004071480116, "quality_score": 0.4857748380308536}
+    }
+  },
+  {
+    "model": "gemma3-12b-ug40-finetuned",
+    "metrics": {"ach_to_eng": {"bleu": 25.11343271537592, "chrf": 0.4860191096457906, "cer": 0.4919295839030974, "wer": 0.643734534592279, "len_ratio": 1.0185021553794027, "rouge1": 0.5453478666173251, "rouge2": 0.3219416367958648, "rougeL": 0.524201216600647, "quality_score": 0.4003723295760434}, "lgg_to_eng": {"bleu": 17.35989959394786, "chrf": 0.3823398395888027, "cer": 0.6124222119859948, "wer": 0.8087767040373183, "len_ratio": 1.0551986682184085, "rouge1": 0.42633247714564226, "rouge2": 0.21809698694179738, "rougeL": 0.4060260521116608, "quality_score": 0.28368497987624197}, "lug_to_eng": {"bleu": 42.232492087846246, "chrf": 0.6425304188010127, "cer": 0.3335844217448657, "wer": 0.4378951249177767, "len_ratio": 1.008934820854824, "rouge1": 0.709432583146364, "rouge2": 0.5082340939744272, "rougeL": 0.6885539307028894, "quality_score": 0.5733439482542082}, "nyn_to_eng": {"bleu": 30.115971927325734, "chrf": 0.5350728867290394, "cer": 0.4255044812312245, "wer": 0.5636166276139, "len_ratio": 0.9995168163572387, "rouge1": 0.6028021381729067, "rouge2": 0.37768861729512687, "rougeL": 0.5853351457264178, "quality_score": 0.46177787428929307}, "teo_to_eng": {"bleu": 24.037351297544017, "chrf": 0.45660558476003665, "cer": 0.5098903559206144, "wer": 0.66326355063606, "len_ratio": 0.9985761779984209, "rouge1": 0.5206918519328798, "rouge2": 0.29757616257393443, "rougeL": 0.5021832815420268, "quality_score": 0.3809562977947005}, "eng_to_ach": {"bleu": 20.47850981787588, "chrf": 0.4998713900635348, "cer": 0.47950490809041924, "wer": 0.7058893060410796, "len_ratio": 1.0279259102426281, "rouge1": 0.4902823147331087, "rouge2": 0.25925552603239826, "rougeL": 0.4679119847922154, "quality_score": 0.37981556852769865}, "eng_to_lgg": {"bleu": 17.83156114658944, "chrf": 0.46520083701043174, "cer": 0.4912480231430983, "wer": 0.7392954137185076, "len_ratio": 1.0040222298037509, "rouge1": 0.48840266808931887, "rouge2": 0.23540163659116345, "rougeL": 0.45623556066324406, "quality_score": 0.35324325290368}, "eng_to_lug": {"bleu": 33.43335290856702, "chrf": 0.6598198822203196, "cer": 0.33288328156255204, "wer": 0.5329261749016848, "len_ratio": 1.0094849062874582, "rouge1": 0.5835755441747735, "rouge2": 0.3827292025413401, "rougeL": 0.5705318574418908, "quality_score": 0.5320859887104382}, "eng_to_nyn": {"bleu": 20.453009142973894, "chrf": 0.5763114515815105, "cer": 0.4080085434691386, "wer": 0.716071038748043, "len_ratio": 1.0417752592674183, "rouge1": 0.4448377559581382, "rouge2": 0.23074255915311173, "rougeL": 0.4351352726668042, "quality_score": 0.414190490198517}, "eng_to_teo": {"bleu": 15.181372127364465, "chrf": 0.479353539228751, "cer": 0.5271113273653261, "wer": 0.7710563350180285, "len_ratio": 1.0958653890262315, "rouge1": 0.40069312378418936, "rouge2": 0.16245608626469754, "rougeL": 0.3908865294077067, "quality_score": 0.3332498995297602}, "eng_to_xog": {"bleu": 18.00596638730871, "chrf": 0.5602175250880231, "cer": 0.39365418624687826, "wer": 0.7346561014952262, "len_ratio": 1.0058363343518926, "rouge1": 0.40309294840310217, "rouge2": 0.18976438719982427, "rougeL": 0.3994074933834067, "quality_score": 0.40299172530475147}, "eng_to_ttj": {"bleu": 24.194664202368145, "chrf": 0.5846369573344875, "cer": 0.39016597970822153, "wer": 0.6401561135535874, "len_ratio": 1.0400234685658434, "rouge1": 0.498559883972833, "rouge2": 0.2813157789083976, "rougeL": 0.4917265006605738, "quality_score": 0.44906537652408995}, "eng_to_swa": {"bleu": 41.67384265910873, "chrf": 0.7019457855280522, "cer": 0.2782188772468657, "wer": 0.42973919281882994, "len_ratio": 1.0325931933120764, "rouge1": 0.685004046162508, "rouge2": 0.47945380227202167, "rougeL": 0.6685093313734645, "quality_score": 0.602681535513361}, "xog_to_eng": {"bleu": 31.64182918312035, "chrf": 0.5473019381440146, "cer": 0.41652585936991027, "wer": 0.5484119855739875, "len_ratio": 1.0155235985633817, "rouge1": 0.6111501202558264, "rouge2": 0.39285189252038694, "rougeL": 0.5933993447258348, "quality_score": 0.47469559625783003}, "ttj_to_eng": {"bleu": 32.74391571294384, "chrf": 0.5556380643120021, "cer": 0.4101941071815267, "wer": 0.5273829309526653, "len_ratio": 1.0036714161713904, "rouge1": 0.625759138731389, "rouge2": 0.4105928718220594, "rougeL": 0.6063340412306235, "quality_score": 0.4863750458268121}, "swa_to_eng": {"bleu": 50.955998739852376, "chrf": 0.7158260766811865, "cer": 0.26964382035693063, "wer": 0.3495457397467596, "len_ratio": 1.0008416855012967, "rouge1": 0.7748498866347656, "rouge2": 0.6015766637293783, "rougeL": 0.7523897235101366, "quality_score": 0.651549125994005}, "averages": {"bleu": 27.84082310313204, "chrf": 0.5530432054198122, "cer": 0.42315562303291654, "wer": 0.6132760546478583, "len_ratio": 1.0223932518688539, "rouge1": 0.5506758967446919, "rouge2": 0.33435486903849565, "rougeL": 0.5336729541587214, "quality_score": 0.44875493969258945}}
+  },
+  {
+    "model": "gemma3-12b-ug40-4bit-quant",
+    "metrics": {
+        "ach_to_eng": {"bleu": 21.4729661767832, "chrf": 0.4496170371887275, "cer": 0.5238795607867115, "wer": 0.692243685907704, "len_ratio": 1.018348476894355, "rouge1": 0.5100621121441622, "rouge2": 0.27372829947661637, "rougeL": 0.4834319801574132, "quality_score": 0.4069529240939532},
+        "eng_to_ach": {"bleu": 16.755438759921113, "chrf": 0.45049534441941974, "cer": 0.5047185301603307, "wer": 0.750613578745915, "len_ratio": 1.0025117482194446, "rouge1": 0.44928898857391764, "rouge2": 0.20580828504350887, "rougeL": 0.4240278320420148, "quality_score": 0.37267240728805295},
+        "eng_to_lgg": {"bleu": 11.372746822604652, "chrf": 0.38289951588574617, "cer": 0.5554623220058079, "wer": 0.8196450255123651, "len_ratio": 0.9712294534625859, "rouge1": 0.41354305633464233, "rouge2": 0.14649272731672833, "rougeL": 0.37386728859619445, "quality_score": 0.3181549969207428},
+        "eng_to_lug": {"bleu": 28.20637559502546, "chrf": 0.6107608799825299, "cer": 0.37180416584750836, "wer": 0.5942818446509742, "len_ratio": 0.9940918695885228, "rouge1": 0.5275739121053499, "rouge2": 0.3188053758650654, "rougeL": 0.5171273982496929, "quality_score": 0.49523998929822416},
+        "eng_to_nyn": {"bleu": 16.46101610966186, "chrf": 0.5212641413986153, "cer": 0.44556935303939627, "wer": 0.7716569355296822, "len_ratio": 1.0176176678547262, "rouge1": 0.37173469650282137, "rouge2": 0.17480324623615434, "rougeL": 0.3664355608287682, "quality_score": 0.36780304520962415},
+        "eng_to_swa": {"bleu": 34.85728965804992, "chrf": 0.6499064988711744, "cer": 0.3297420048189144, "wer": 0.5063806405288259, "len_ratio": 1.029950766846105, "rouge1": 0.6253343132217276, "rouge2": 0.40672985153760854, "rougeL": 0.6061543296531952, "quality_score": 0.5656408988298093},
+        "eng_to_teo": {"bleu": 11.354552915082225, "chrf": 0.402601395639914, "cer": 0.6073847218781454, "wer": 0.8619104853479853, "len_ratio": 1.099810248113919, "rouge1": 0.3180487615247001, "rouge2": 0.1012222991617374, "rougeL": 0.30968691127823866, "quality_score": 0.2790978983945907},
+        "eng_to_xog": {"bleu": 13.420917205099988, "chrf": 0.47240345656926297, "cer": 0.47981592249872757, "wer": 0.8396892979924883, "len_ratio": 0.9877543086333478, "rouge1": 0.31545315902740895, "rouge2": 0.1220769340901552, "rougeL": 0.3104015097062153, "quality_score": 0.31882701281044523},
+        "eng_to_ttj": {"bleu": 16.6586798464968, "chrf": 0.4924984720191624, "cer": 0.4651276291420016, "wer": 0.7390441626684511, "len_ratio": 1.0057924851499562, "rouge1": 0.3914667140120163, "rouge2": 0.18006144771016327, "rougeL": 0.3837100328352088, "quality_score": 0.37168170425348385},
+        "lgg_to_eng": {"bleu": 12.32879786323486, "chrf": 0.32304500757031107, "cer": 0.6692679994119266, "wer": 0.8894077306509114, "len_ratio": 1.0439009054067556, "rouge1": 0.3534898071043344, "rouge2": 0.14202989318604603, "rougeL": 0.3298602152858462, "quality_score": 0.26183454642166704},
+        "lug_to_eng": {"bleu": 35.43782379684757, "chrf": 0.5922238429388922, "cer": 0.3718068923554593, "wer": 0.49748382892802956, "len_ratio": 0.9979934266656314, "rouge1": 0.6630293401277573, "rouge2": 0.4376760575795662, "rougeL": 0.6392055338551386, "quality_score": 0.5632577056011291},
+        "nyn_to_eng": {"bleu": 25.458480625652598, "chrf": 0.4935261786667742, "cer": 0.465419847908852, "wer": 0.6148711123435389, "len_ratio": 0.9928662300788028, "rouge1": 0.5617648939792661, "rouge2": 0.3231354511610384, "rougeL": 0.5397584704699726, "quality_score": 0.4615572315200247},
+        "swa_to_eng": {"bleu": 45.97432783545669, "chrf": 0.6819820391180372, "cer": 0.2948660062951407, "wer": 0.38266810573202886, "len_ratio": 0.9828458021701164, "rouge1": 0.749023406365157, "rouge2": 0.5544504543659836, "rougeL": 0.7239889314062833, "quality_score": 0.6562005905361457},
+        "teo_to_eng": {"bleu": 21.467843556940434, "chrf": 0.42011352105186406, "cer": 0.5229556961814095, "wer": 0.6779863734397059, "len_ratio": 0.9673047055582993, "rouge1": 0.4840116546397648, "rouge2": 0.2623107223216375, "rougeL": 0.46544708326233714, "quality_score": 0.39721810415037584},
+        "xog_to_eng": {"bleu": 27.81044272440912, "chrf": 0.5090752251373535, "cer": 0.45288341394763143, "wer": 0.5932568600885726, "len_ratio": 1.013664683393617, "rouge1": 0.5793597662559643, "rouge2": 0.3449307726259345, "rougeL": 0.557485596610377, "quality_score": 0.47964745686859694},
+        "ttj_to_eng": {"bleu": 28.212033880217582, "chrf": 0.5094574222853848, "cer": 0.4521726047019253, "wer": 0.5807696044665328, "len_ratio": 1.0020228756514913, "rouge1": 0.5869979083097827, "rouge2": 0.3523762869779724, "rougeL": 0.565152165886336, "quality_score": 0.48513093768587023},
+        "averages": {"bleu": 22.953108335717754, "chrf": 0.4976168736714481, "cer": 0.469554791936243, "wer": 0.675744329533357, "len_ratio": 1.0079816033554798, "rouge1": 0.4937614056392984, "rouge2": 0.2716648815409948, "rougeL": 0.47473380250770203, "quality_score": 0.425057340617671}
+    }
+  },
+  {
+  "model": "gemma3-12b-ug40-8bit-quant",
+  "metrics": {
+    "ach_to_eng": {
+      "bleu": 21.33575717169921,
+      "chrf": 0.4521391134216422,
+      "cer": 0.535683670140994,
+      "wer": 0.6932714020311863,
+      "len_ratio": 1.0410301289165254,
+      "rouge1": 0.5115414491119932,
+      "rouge2": 0.27254126665296446,
+      "rougeL": 0.48705074413604205,
+      "quality_score": 0.4058556343690815
+    },
+    "eng_to_ach": {
+      "bleu": 18.325087064280495,
+      "chrf": 0.46169074747519706,
+      "cer": 0.4941741708398276,
+      "wer": 0.7217611122764125,
+      "len_ratio": 0.9971235291743883,
+      "rouge1": 0.46584238349869556,
+      "rouge2": 0.2258632422094869,
+      "rougeL": 0.44201558713229405,
+      "quality_score": 0.38947738427212525
+    },
+    "eng_to_lgg": {
+      "bleu": 12.228082821292523,
+      "chrf": 0.39641379386899284,
+      "cer": 0.5550914581438494,
+      "wer": 0.8135853738008421,
+      "len_ratio": 0.9794736747100792,
+      "rouge1": 0.4187206079316151,
+      "rouge2": 0.15996623543184668,
+      "rougeL": 0.3778975282179714,
+      "quality_score": 0.32443932104780215
+    },
+    "eng_to_lug": {
+      "bleu": 30.37021808778844,
+      "chrf": 0.6211092235369094,
+      "cer": 0.35714076059236627,
+      "wer": 0.5696828314000989,
+      "len_ratio": 0.9887628479880637,
+      "rouge1": 0.5452875907310831,
+      "rouge2": 0.3383485068722653,
+      "rougeL": 0.5342018761481492,
+      "quality_score": 0.5129128798835935
+    },
+    "eng_to_nyn": {
+      "bleu": 16.753648581098513,
+      "chrf": 0.5207287514239932,
+      "cer": 0.4477157022948351,
+      "wer": 0.760699794289097,
+      "len_ratio": 1.0078976964397806,
+      "rouge1": 0.379103813152116,
+      "rouge2": 0.17556542630328617,
+      "rougeL": 0.3699470659508593,
+      "quality_score": 0.3714834366256703
+    },
+    "eng_to_swa": {
+      "bleu": 34.83533816994912,
+      "chrf": 0.6518458171430364,
+      "cer": 0.3227943798839051,
+      "wer": 0.4978481783673316,
+      "len_ratio": 1.0294131193000886,
+      "rouge1": 0.6266599273479803,
+      "rouge2": 0.4069100776121281,
+      "rougeL": 0.608825177720382,
+      "quality_score": 0.5691736242766088
+    },
+    "eng_to_teo": {
+      "bleu": 12.392234885652412,
+      "chrf": 0.41860468355251723,
+      "cer": 0.5878420812904283,
+      "wer": 0.837512772845434,
+      "len_ratio": 1.0941596008068062,
+      "rouge1": 0.33186565537665563,
+      "rouge2": 0.11363308176316347,
+      "rougeL": 0.32559460177919164,
+      "quality_score": 0.29577207257150434
+    },
+    "eng_to_xog": {
+      "bleu": 14.28594219428805,
+      "chrf": 0.47824531819630023,
+      "cer": 0.4676545238436661,
+      "wer": 0.8189706573141702,
+      "len_ratio": 0.9789787119512641,
+      "rouge1": 0.33227716126053725,
+      "rouge2": 0.13497189149879513,
+      "rougeL": 0.326945127050312,
+      "quality_score": 0.33228364121536563
+    },
+    "eng_to_ttj": {
+      "bleu": 16.977426123943793,
+      "chrf": 0.5062896646495704,
+      "cer": 0.45395455428659626,
+      "wer": 0.7376647417651925,
+      "len_ratio": 1.0141849421370215,
+      "rouge1": 0.40228568836268996,
+      "rouge2": 0.18366644216416722,
+      "rougeL": 0.3952386421085991,
+      "quality_score": 0.3803281600514181
+    },
+    "lgg_to_eng": {
+      "bleu": 13.05283144643632,
+      "chrf": 0.335016264891506,
+      "cer": 0.6654538032435573,
+      "wer": 0.8799084421534612,
+      "len_ratio": 1.0497308112523163,
+      "rouge1": 0.37215842505873253,
+      "rouge2": 0.1542578202610299,
+      "rougeL": 0.3473680834269667,
+      "quality_score": 0.27328480707409164
+    },
+    "lug_to_eng": {
+      "bleu": 36.583076178754155,
+      "chrf": 0.6017208109828538,
+      "cer": 0.36385226407849075,
+      "wer": 0.484647564050434,
+      "len_ratio": 0.9978253893530138,
+      "rouge1": 0.6728986531388527,
+      "rouge2": 0.4559711476443575,
+      "rougeL": 0.6505567106388979,
+      "quality_score": 0.5737511847365369
+    },
+    "nyn_to_eng": {
+      "bleu": 26.848257026479924,
+      "chrf": 0.5051204239753181,
+      "cer": 0.4435338778359474,
+      "wer": 0.5892439429574747,
+      "len_ratio": 0.9871168616493217,
+      "rouge1": 0.5803117102981369,
+      "rouge2": 0.3447149380176738,
+      "rougeL": 0.5590397941749801,
+      "quality_score": 0.48002944631996874
+    },
+    "swa_to_eng": {
+      "bleu": 45.52723000555685,
+      "chrf": 0.6828039855678282,
+      "cer": 0.2957147126793479,
+      "wer": 0.39049986652868535,
+      "len_ratio": 0.9924394318831312,
+      "rouge1": 0.7467246988156893,
+      "rouge2": 0.5515762032887349,
+      "rougeL": 0.7221916920080675,
+      "quality_score": 0.65346301620652
+    },
+    "teo_to_eng": {
+      "bleu": 21.732084138527295,
+      "chrf": 0.4272565301302563,
+      "cer": 0.5181106210940541,
+      "wer": 0.6761849143371914,
+      "len_ratio": 0.972982533475168,
+      "rouge1": 0.494799806539735,
+      "rouge2": 0.26288018196280777,
+      "rougeL": 0.4724533317550662,
+      "quality_score": 0.4029224957298474
+    },
+    "xog_to_eng": {
+      "bleu": 27.988289616284757,
+      "chrf": 0.519525535777184,
+      "cer": 0.4407529453981863,
+      "wer": 0.5811313064307607,
+      "len_ratio": 1.0139387986353932,
+      "rouge1": 0.5901160493494182,
+      "rouge2": 0.35013145494136566,
+      "rougeL": 0.5670982618287617,
+      "quality_score": 0.489123081881544
+    },
+    "ttj_to_eng": {
+      "bleu": 29.1930717978176,
+      "chrf": 0.5209701082464389,
+      "cer": 0.43713321847400566,
+      "wer": 0.5686012915110401,
+      "len_ratio": 1.0027713236347304,
+      "rouge1": 0.5960520197834103,
+      "rouge2": 0.36925917628754246,
+      "rougeL": 0.5760260495288388,
+      "quality_score": 0.49654073092530304
+    },
+    "averages": {
+      "bleu": 23.65178595686559,
+      "chrf": 0.5062175483024716,
+      "cer": 0.4616626715075036,
+      "wer": 0.6638258870036758,
+      "len_ratio": 1.0092393375816933,
+      "rouge1": 0.5041653524848339,
+      "rouge2": 0.281266068306976,
+      "rougeL": 0.4851531421003362,
+      "quality_score": 0.43442755732418636
+    }
+  }
+},
+{
+  "model": "sunbird-nllb-1.3B",
+  "metrics": {
+    "ach_to_eng": {
+      "bleu": 29.430510281424944,
+      "chrf": 0.5441419020578823,
+      "cer": 0.43772136644658854,
+      "wer": 0.570230453613696,
+      "len_ratio": 0.9863286651712698,
+      "rouge1": 0.6052849080398074,
+      "rouge2": 0.3729925619001821,
+      "rougeL": 0.5838502087408626,
+      "quality_score": 0.5032717169320862
+    },
+    "eng_to_ach": {
+      "bleu": 23.782057976721013,
+      "chrf": 0.5198248176762581,
+      "cer": 0.454300473673888,
+      "wer": 0.6639762120076788,
+      "len_ratio": 1.0114108506939241,
+      "rouge1": 0.5226163876274231,
+      "rouge2": 0.2982056716258377,
+      "rougeL": 0.5017633903903476,
+      "quality_score": 0.4439580816299454
+    },
+    "eng_to_lgg": {
+      "bleu": 22.609242121260998,
+      "chrf": 0.5234007657796004,
+      "cer": 0.45175606592292505,
+      "wer": 0.6747794509984192,
+      "len_ratio": 1.0403123630210043,
+      "rouge1": 0.5492432550958023,
+      "rouge2": 0.3001400162864386,
+      "rougeL": 0.5196794791945015,
+      "quality_score": 0.448646734060195
+    },
+    "eng_to_lug": {
+      "bleu": 33.517060557747875,
+      "chrf": 0.6557565322551971,
+      "cer": 0.3256815714888277,
+      "wer": 0.5278019119426852,
+      "len_ratio": 0.9943903120813362,
+      "rouge1": 0.5901211551364539,
+      "rouge2": 0.3819067325679575,
+      "rougeL": 0.5786059729778025,
+      "quality_score": 0.5510284637525699
+    },
+    "eng_to_nyn": {
+      "bleu": 21.38030939374281,
+      "chrf": 0.5862355598811484,
+      "cer": 0.3984049224473253,
+      "wer": 0.6959072903495935,
+      "len_ratio": 1.0294915242689973,
+      "rouge1": 0.4535905308368584,
+      "rouge2": 0.23655530882283077,
+      "rougeL": 0.44567847821809525,
+      "quality_score": 0.4341659083461019
+    },
+    "eng_to_swa": {
+      "bleu": 15.317398647865414,
+      "chrf": 0.4167348639564299,
+      "cer": 0.5758950441995268,
+      "wer": 0.8588143291520308,
+      "len_ratio": 1.0330167886210408,
+      "rouge1": 0.36496818660943187,
+      "rouge2": 0.15257733997586892,
+      "rougeL": 0.3554022037063463,
+      "quality_score": 0.30926164456655075
+    },
+    "eng_to_teo": {
+      "bleu": 23.731299517809013,
+      "chrf": 0.588741893126223,
+      "cer": 0.39940433066892045,
+      "wer": 0.6260653996541092,
+      "len_ratio": 1.0459884106635844,
+      "rouge1": 0.5248329530752708,
+      "rouge2": 0.2766148391060161,
+      "rougeL": 0.5092298278756372,
+      "quality_score": 0.47244132315536524
+    },
+    "eng_to_xog": {
+      "bleu": 0.25183476787032366,
+      "chrf": 0.042723304533363766,
+      "cer": 1.0138308923833361,
+      "wer": 1.3016492691895918,
+      "len_ratio": 0.5627238532432384,
+      "rouge1": 0.008552965479925106,
+      "rouge2": 0.0012702351412028831,
+      "rougeL": 0.008552965479925106,
+      "quality_score": -0.04218876306683509
+    },
+    "eng_to_ttj": {
+      "bleu": 0.9434417256850681,
+      "chrf": 0.12551153238120175,
+      "cer": 1.130481537826477,
+      "wer": 2.008269922282019,
+      "len_ratio": 1.2738571132457104,
+      "rouge1": 0.029937202334871564,
+      "rouge2": 0.004903671236340705,
+      "rougeL": 0.02849955863005825,
+      "quality_score": -0.157561458250919
+    },
+    "lgg_to_eng": {
+      "bleu": 33.183960121607434,
+      "chrf": 0.5724935896324835,
+      "cer": 0.41162062430318264,
+      "wer": 0.5399960442388099,
+      "len_ratio": 0.9959561607358951,
+      "rouge1": 0.6295424773311681,
+      "rouge2": 0.414141453330078,
+      "rougeL": 0.6076793615565902,
+      "quality_score": 0.5316563935323873
+    },
+    "lug_to_eng": {
+      "bleu": 41.65427596784633,
+      "chrf": 0.641959152367403,
+      "cer": 0.33007458554090835,
+      "wer": 0.4364174744452259,
+      "len_ratio": 0.995614469949668,
+      "rouge1": 0.7019737319070327,
+      "rouge2": 0.5009750688341794,
+      "rougeL": 0.6837419520245854,
+      "quality_score": 0.6129542559985584
+    },
+    "nyn_to_eng": {
+      "bleu": 33.827381740261934,
+      "chrf": 0.5814058615830704,
+      "cer": 0.3826375770314228,
+      "wer": 0.5096913838849322,
+      "len_ratio": 0.9884081840154549,
+      "rouge1": 0.6371410738601032,
+      "rouge2": 0.4228617371895637,
+      "rougeL": 0.6184775626898391,
+      "quality_score": 0.547161559103213
+    },
+    "swa_to_eng": {
+      "bleu": 42.12403395429048,
+      "chrf": 0.6511244866804917,
+      "cer": 0.3248127435362155,
+      "wer": 0.42393942533589124,
+      "len_ratio": 0.9881591242064081,
+      "rouge1": 0.7104240116477228,
+      "rouge2": 0.5166068351637543,
+      "rougeL": 0.6881280422432776,
+      "quality_score": 0.6203607852070483
+    },
+    "teo_to_eng": {
+      "bleu": 32.487192862281574,
+      "chrf": 0.5684747487342591,
+      "cer": 0.3912525302692959,
+      "wer": 0.5050233717631725,
+      "len_ratio": 0.9779057702353239,
+      "rouge1": 0.6394775806019763,
+      "rouge2": 0.4109552011262328,
+      "rougeL": 0.6210169245028457,
+      "quality_score": 0.5429275467382381
+    },
+    "xog_to_eng": {
+      "bleu": 18.893891543765875,
+      "chrf": 0.4023685911576792,
+      "cer": 0.5835602877597674,
+      "wer": 0.744059252864993,
+      "len_ratio": 1.0487608274645277,
+      "rouge1": 0.42280970356012115,
+      "rouge2": 0.23623908315016312,
+      "rougeL": 0.4104318506665021,
+      "quality_score": 0.34948825336620015
+    },
+    "ttj_to_eng": {
+      "bleu": 29.100962682882233,
+      "chrf": 0.5249571980106907,
+      "cer": 0.4358684131387772,
+      "wer": 0.560060677943446,
+      "len_ratio": 1.016829177274056,
+      "rouge1": 0.5856316196730803,
+      "rouge2": 0.3689444466335747,
+      "rougeL": 0.5736947324722281,
+      "quality_score": 0.49656068098376643
+    },
+    "averages": {
+      "bleu": 25.13967836644146,
+      "chrf": 0.4966159249883364,
+      "cer": 0.5029564354148366,
+      "wer": 0.7279176168541435,
+      "len_ratio": 0.999322099680715,
+      "rouge1": 0.4985092339260655,
+      "rouge2": 0.30599313763063885,
+      "rougeL": 0.4834020319605903,
+      "quality_score": 0.41650832037840446
+    }
+  }
+},
+{
+  "model": "sunbird-nllb-3.3B",
+  "metrics": {
+    "ach_to_eng": {
+      "bleu": 29.11973177617704,
+      "chrf": 0.5314432935977346,
+      "cer": 0.4458303843259423,
+      "wer": 0.57612930781663,
+      "len_ratio": 1.0060317234901897,
+      "rouge1": 0.5998762301955666,
+      "rouge2": 0.37458530040433596,
+      "rougeL": 0.5789635090282815,
+      "quality_score": 0.4965867764067968
+    },
+    "eng_to_ach": {
+      "bleu": 20.49858957572721,
+      "chrf": 0.5004301347939377,
+      "cer": 0.4528755693828615,
+      "wer": 0.6782113037259656,
+      "len_ratio": 0.9989327733263587,
+      "rouge1": 0.5009863598041194,
+      "rouge2": 0.25755912923808805,
+      "rougeL": 0.4810534513370743,
+      "quality_score": 0.42606149476392946
+    },
+    "eng_to_lgg": {
+      "bleu": 20.05048545987706,
+      "chrf": 0.49415454647366874,
+      "cer": 0.47216319663820605,
+      "wer": 0.6898039887446645,
+      "len_ratio": 1.0152530745218167,
+      "rouge1": 0.5259020378490028,
+      "rouge2": 0.27269556217701796,
+      "rougeL": 0.498248014428781,
+      "quality_score": 0.4261403779945588
+    },
+    "eng_to_lug": {
+      "bleu": 34.39434146308947,
+      "chrf": 0.662024107390938,
+      "cer": 0.32176812327577214,
+      "wer": 0.5263568098304816,
+      "len_ratio": 0.9990323458207923,
+      "rouge1": 0.589633281523342,
+      "rouge2": 0.3904519621826425,
+      "rougeL": 0.5781388742150646,
+      "quality_score": 0.5542691241089976
+    },
+    "eng_to_nyn": {
+      "bleu": 21.067510181486618,
+      "chrf": 0.576304099642706,
+      "cer": 0.4095992237539395,
+      "wer": 0.7094975523106453,
+      "len_ratio": 1.0369875733575387,
+      "rouge1": 0.4466546568396715,
+      "rouge2": 0.23209128541551535,
+      "rougeL": 0.43850325251139943,
+      "quality_score": 0.425506722457343
+    },
+    "eng_to_swa": {
+      "bleu": 44.90764186108048,
+      "chrf": 0.7146203560186356,
+      "cer": 0.2723852696076477,
+      "wer": 0.4137906168652137,
+      "len_ratio": 1.0248813198797162,
+      "rouge1": 0.6994166072573097,
+      "rouge2": 0.5118455845181964,
+      "rougeL": 0.6805958612631842,
+      "quality_score": 0.6429222261128454
+    },
+    "eng_to_teo": {
+      "bleu": 20.284459614385437,
+      "chrf": 0.5438482304674738,
+      "cer": 0.43321992334718734,
+      "wer": 0.6550267664682988,
+      "len_ratio": 1.0407314070775626,
+      "rouge1": 0.4853574858912598,
+      "rouge2": 0.2337512231998303,
+      "rougeL": 0.47217194402438006,
+      "quality_score": 0.43599592778524704
+    },
+    "eng_to_xog": {
+      "bleu": 10.16505424919429,
+      "chrf": 0.418062495851419,
+      "cer": 0.4939202748205401,
+      "wer": 0.897634971468285,
+      "len_ratio": 0.972319126451413,
+      "rouge1": 0.25056496535267203,
+      "rouge2": 0.08484571673870817,
+      "rougeL": 0.24764221429573435,
+      "quality_score": 0.27106082861715725
+    },
+    "eng_to_ttj": {
+      "bleu": 15.301771868703554,
+      "chrf": 0.4923441359897834,
+      "cer": 0.46104872255074464,
+      "wer": 0.7733622104426776,
+      "len_ratio": 1.0431481399762663,
+      "rouge1": 0.37302959299462896,
+      "rouge2": 0.16138401996059154,
+      "rougeL": 0.3665536323015275,
+      "quality_score": 0.3584223578299255
+    },
+    "lgg_to_eng": {
+      "bleu": 28.790977357068755,
+      "chrf": 0.5233272096208317,
+      "cer": 0.46068120581099303,
+      "wer": 0.597573063104669,
+      "len_ratio": 1.0304519263957117,
+      "rouge1": 0.5855305987067816,
+      "rouge2": 0.3677655151230366,
+      "rougeL": 0.565512842136962,
+      "quality_score": 0.4840043591866002
+    },
+    "lug_to_eng": {
+      "bleu": 42.06884087500722,
+      "chrf": 0.6420302109330189,
+      "cer": 0.32871907753323737,
+      "wer": 0.4342940468963358,
+      "len_ratio": 1.000247825208639,
+      "rouge1": 0.7094800434749033,
+      "rouge2": 0.5064541233924069,
+      "rougeL": 0.689735768235689,
+      "quality_score": 0.6164868844940183
+    },
+    "nyn_to_eng": {
+      "bleu": 33.14465866594187,
+      "chrf": 0.5637049333595793,
+      "cer": 0.3876117946383213,
+      "wer": 0.5108166855216784,
+      "len_ratio": 0.9882578399677422,
+      "rouge1": 0.6329615186647974,
+      "rouge2": 0.4183838178572328,
+      "rougeL": 0.6164938690183747,
+      "quality_score": 0.541029737923695
+    },
+    "swa_to_eng": {
+      "bleu": 47.052329072293254,
+      "chrf": 0.6892222377335402,
+      "cer": 0.28476858898796276,
+      "wer": 0.3700288841521645,
+      "len_ratio": 1.002994515234727,
+      "rouge1": 0.7529059087918626,
+      "rouge2": 0.566254405574289,
+      "rougeL": 0.733210635501974,
+      "quality_score": 0.6651774332683639
+    },
+    "teo_to_eng": {
+      "bleu": 29.87870644400743,
+      "chrf": 0.5252286187546552,
+      "cer": 0.43447865960403914,
+      "wer": 0.5595597975119441,
+      "len_ratio": 0.9909164376589676,
+      "rouge1": 0.5974291618668124,
+      "rouge2": 0.37546488405545797,
+      "rougeL": 0.578869581621942,
+      "quality_score": 0.5010459949279168
+    },
+    "xog_to_eng": {
+      "bleu": 28.125665148506382,
+      "chrf": 0.5057010634098533,
+      "cer": 0.4545948687363525,
+      "wer": 0.598121571923754,
+      "len_ratio": 1.0214763620090188,
+      "rouge1": 0.5644995626552085,
+      "rouge2": 0.35062850387370176,
+      "rougeL": 0.548367176342721,
+      "quality_score": 0.4745180022054567
+    },
+    "ttj_to_eng": {
+      "bleu": 30.14139892105478,
+      "chrf": 0.5291982808195398,
+      "cer": 0.42605764014249364,
+      "wer": 0.5500122773797466,
+      "len_ratio": 1.0132027420026863,
+      "rouge1": 0.6018823971252929,
+      "rouge2": 0.3803700460204749,
+      "rougeL": 0.5892573310774452,
+      "quality_score": 0.5076136801184309
+    },
+    "averages": {
+      "bleu": 28.437010158350052,
+      "chrf": 0.5569777471785822,
+      "cer": 0.4087326576972651,
+      "wer": 0.5962637408851972,
+      "len_ratio": 1.0115540707736967,
+      "rouge1": 0.557256900562077,
+      "rouge2": 0.3427831924832204,
+      "rougeL": 0.5414573723337834,
+      "quality_score": 0.4891776205125802
+    }
+  }
+},
+{
+  "model": "google-translate",
+  "metrics": {
+    "ach_to_eng": {
+      "bleu": 25.82012707261053,
+      "chrf": 0.5057031273960824,
+      "cer": 0.4617745077050436,
+      "wer": 0.595004667711321,
+      "len_ratio": 0.9636727813606757,
+      "rouge1": 0.575123846097411,
+      "rouge2": 0.3290197121231234,
+      "rougeL": 0.5533438889765876,
+      "quality_score": 0.472598826296637
+    },
+    "eng_to_ach": {
+      "bleu": 17.4129402377706,
+      "chrf": 0.468682516063646,
+      "cer": 0.48800545379901034,
+      "wer": 0.7526983036199513,
+      "len_ratio": 1.0173304711940114,
+      "rouge1": 0.4650325796160986,
+      "rouge2": 0.22135445162710823,
+      "rougeL": 0.44382393779797036,
+      "quality_score": 0.3851607797394099
+    },
+    "eng_to_lug": {
+      "bleu": 26.565747903999014,
+      "chrf": 0.6023884902489536,
+      "cer": 0.3878062612373834,
+      "wer": 0.6218423427063132,
+      "len_ratio": 1.018827130375813,
+      "rouge1": 0.5208782630683307,
+      "rouge2": 0.3033559645916972,
+      "rougeL": 0.5074844347703342,
+      "quality_score": 0.48112667719731866
+    },
+    "eng_to_swa": {
+      "bleu": 60.84743290999219,
+      "chrf": 0.7906909029140974,
+      "cer": 0.2031302954281225,
+      "wer": 0.30459812365054295,
+      "len_ratio": 1.0120282818329116,
+      "rouge1": 0.7800397358230957,
+      "rouge2": 0.6454775766085189,
+      "rougeL": 0.764678316404524,
+      "quality_score": 0.7393591441938289
+    },
+    "lug_to_eng": {
+      "bleu": 35.700224344812185,
+      "chrf": 0.5812880213285667,
+      "cer": 0.3639859890535588,
+      "wer": 0.4797680581791706,
+      "len_ratio": 0.9473770400224313,
+      "rouge1": 0.6607218133735266,
+      "rouge2": 0.43264293299022016,
+      "rougeL": 0.6421587685214877,
+      "quality_score": 0.5662361332398289
+    },
+    "swa_to_eng": {
+      "bleu": 47.42800244646666,
+      "chrf": 0.7009113380165776,
+      "cer": 0.2712123141857987,
+      "wer": 0.3557651500473488,
+      "len_ratio": 0.9713371029501139,
+      "rouge1": 0.7669060465411134,
+      "rouge2": 0.5713060632558862,
+      "rougeL": 0.7426604606816005,
+      "quality_score": 0.6762967342451351
+    },
+    "averages": {
+      "bleu": 35.629079152608526,
+      "chrf": 0.6082773993279873,
+      "cer": 0.36265247023481956,
+      "wer": 0.5182794409857746,
+      "len_ratio": 0.988428801289326,
+      "rouge1": 0.6281170474199292,
+      "rouge2": 0.41719278353275896,
+      "rougeL": 0.6090249678587507,
+      "quality_score": 0.5534630491520264
+    }
+  }
+}
+]

requirements.txt

@@ -0,0 +1,21 @@
+fastapi
+uvicorn
+pydantic
+
+# Core dependencies
+datasets>=2.14.0
+sacrebleu>=2.3.0
+rouge-score>=0.1.2
+Levenshtein>=0.21.0
+numpy>=1.24.0
+pandas>=2.0.0
+matplotlib>=3.7.0
+huggingface_hub>=0.16.0
+PyYAML>=6.0
+plotly>=5.0.0
+google-cloud-translate>=3.0.0
+ipython
+
+# Additional dependencies that SALT might need
+transformers>=4.30.0
+torch>=2.0.0