96 lines
4.8 KiB
TypeScript
96 lines
4.8 KiB
TypeScript
export default {
|
|
'benchmark.title': 'Benchmarks',
|
|
'benchmark.button.add': 'Add Benchmark',
|
|
'benchmark.button.edit': 'Edit Benchmark',
|
|
'benchmark.button.compare': 'Compare',
|
|
'benchmark.table.model': 'Model',
|
|
'benchmark.table.instance': 'Model Instance',
|
|
'benchmark.table.dataset': 'Dataset',
|
|
'benchmark.table.requestRate': 'Request Rate',
|
|
'benchmark.table.gpu': 'GPU ',
|
|
'benchmark.table.tpot': 'TPOT',
|
|
'benchmark.table.itl': 'ITL',
|
|
'benchmark.table.ttft': 'TTFT',
|
|
'benchmark.table.rps': 'Requests/s',
|
|
'benchmark.table.tps': 'Tokens/s',
|
|
'benchmark.table.inputLen': 'Input Length',
|
|
'benchmark.table.outputLen': 'Output Length',
|
|
'benchmark.form.profile': 'Profile',
|
|
'benchmark.form.totalRequests': 'Total Requests',
|
|
'benchmark.form.profile.latency': 'Latency',
|
|
'benchmark.form.profile.throughput': 'Throughput',
|
|
'benchmark.form.profile.longContext': 'Long Context',
|
|
'benchmark.form.profile.heavy': 'Generation Heavy',
|
|
'benchmark.form.profile.custom': 'Custom',
|
|
'benchmark.table.inputTokenLength': 'Input Token Length',
|
|
'benchmark.table.outputTokenLength': 'Output Token Length',
|
|
'benchmark.detail.summary.title': 'Summary',
|
|
'benchmark.detail.configure.title': 'Configure',
|
|
'benchmark.detail.environment.title': 'Environment',
|
|
'benchmark.detail.logs.title': 'Logs',
|
|
'benchmark.form.totalRequests.tips': 'Total Requests',
|
|
'benchmark.form.profile.latency.tips':
|
|
'Single-request latency benchmark focusing on TTFT and tail latency. Suitable for API serving and chat scenarios.',
|
|
'benchmark.form.profile.throughput.tips':
|
|
'Max throughput with realistic prompt length. The most commonly used baseline for GPU and model comparison.',
|
|
'benchmark.form.profile.longContext.tips':
|
|
'Stress test for long-context handling. Evaluates KV cache behavior, memory usage, and backend stability.',
|
|
'benchmark.form.profile.heavy.tips':
|
|
'Decode-heavy generation benchmark. Measures sustained decoding speed and output token throughput.',
|
|
'benchmark.table.filter.bygpu': 'Filter by GPU',
|
|
'benchmark.table.filter.bymodel': 'Filter by Model',
|
|
'benchmark.table.filter.bydataset': 'Filter by Dataset',
|
|
'benchmark.table.filter.byProfile': 'Filter by Profile',
|
|
'benchmark.table.avg': 'Avg',
|
|
'benchmark.table.columnSettings': 'Column Settings',
|
|
'benchmark.detail.summary.results': 'Test Results',
|
|
'benchmark.detail.summary.metadata': 'Metadata',
|
|
'benchmark.detail.summary.metrics': 'Test Metrics',
|
|
'benchmark.detail.summary.latency': 'Latency',
|
|
'benchmark.detail.summary.throughput': 'Throughput',
|
|
'benchmark.detail.summary.request': 'Requests',
|
|
'benchmark.detail.avg.reqLatency': 'Request Latency Avg',
|
|
'benchmark.detail.avg.ttft': 'TTFT Avg',
|
|
'benchmark.detail.avg.tpot': 'TPOT Avg',
|
|
'benchmark.detail.avg.itl': 'ITL Avg',
|
|
'benchmark.detail.throughput.totalToken': 'Total Throughput',
|
|
'benchmark.detail.throughput.inputToken': 'Input Throughput',
|
|
'benchmark.detail.throughput.outputToken': 'Output Throughput',
|
|
'benchmark.detail.requests.total': 'Total Requests',
|
|
'benchmark.detail.requests.success': 'Successful Requests',
|
|
'benchmark.detail.requests.failed': 'Failed Requests',
|
|
'benchmark.detail.requests.incomplete': 'Incomplete Requests',
|
|
'benchmark.detail.requests.concurrency': 'Concurrency Avg',
|
|
'benchmark.detail.requests.concurrency.max': 'Concurrency Max',
|
|
'benchmark.detail.percentile.latency': 'Latency',
|
|
'benchmark.detail.percentile.inputTokens': 'Input Tokens',
|
|
'benchmark.detail.percentile.outputTokens': 'Output Tokens',
|
|
'benchmark.detail.percentile.input': 'Input',
|
|
'benchmark.detail.percentile.output': 'Output',
|
|
'benchmark.detail.percentile.total': 'Total',
|
|
'benchmark.detail.configure.tokenLength': 'Token Length (Input/Output)',
|
|
'benchmark.detail.percentile.title': 'Percentile',
|
|
'benchmark.detail.modelName': 'Model Name',
|
|
'benchmark.detail.instanceName': 'Instance Name',
|
|
'benchmark.detail.configure': 'Configuration',
|
|
'benchmark.detail.modelFile': 'Model File',
|
|
'benchmark.detail.kvCache': 'Extended KV Cache',
|
|
'benchmark.detail.speculativeDecoding': 'Speculative Decoding',
|
|
'benchmark.detail.inputOutputTokenLength': 'Token Length (Input/Output)',
|
|
'benchmark.env.gpuName': 'GPU Name',
|
|
'benchmark.env.workerName': 'Worker Name',
|
|
'benchmark.env.index': 'Index',
|
|
'benchmark.env.system': 'System',
|
|
'benchmark.env.runtimeVersion': 'Runtime Version',
|
|
'benchmark.env.driverVersion': 'Driver Version',
|
|
'benchmark.env.cpuCounts': 'CPU Cores',
|
|
'benchmark.table.export.results': 'Export Results',
|
|
'benchmark.form.nonLlmModel.tips':
|
|
'Benchmarking currently only supports LLM models',
|
|
'benchmark.detail.result.duration': 'Duration',
|
|
'benchmark.detail.result.basic': 'Basic',
|
|
'benchmark.form.profile.ShareGPT': 'ShareGPT',
|
|
'benchmark.form.profile.ShareGPT.tips':
|
|
'Measures maximum throughput using real conversational data. Suitable for GPU and model performance benchmarking.'
|
|
};
|