Spaces:
Running
Running
File size: 7,655 Bytes
9bd422a | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 | /**
* Unit tests for TensorShapeInspector
* Validates: Requirements 21.1, 21.2, 21.3, 21.4, 21.5
*/
import { describe, it, expect, beforeEach } from 'vitest';
/**
* Extract the pure tensor lookup logic for testing (mirrors TensorShapeInspector._buildTensorInfoMap + lookupTensor).
* @param {Object} parsedModel
* @returns {{ lookup: (name: string) => Object|null, size: number }}
*/
function buildTensorLookup(parsedModel) {
const map = new Map();
// 1. Inputs
if (Array.isArray(parsedModel.inputs)) {
for (const inp of parsedModel.inputs) {
if (inp.name) {
map.set(inp.name, {
name: inp.name,
shape: inp.shape || [],
dataType: inp.dataType || 'UNKNOWN'
});
}
}
}
// 2. Outputs
if (Array.isArray(parsedModel.outputs)) {
for (const out of parsedModel.outputs) {
if (out.name) {
map.set(out.name, {
name: out.name,
shape: out.shape || [],
dataType: out.dataType || 'UNKNOWN'
});
}
}
}
// 3. value_info (intermediate tensors)
const valueInfo = parsedModel.graph && parsedModel.graph.valueInfo;
if (valueInfo && typeof valueInfo === 'object') {
const entries = Object.entries(valueInfo);
for (const [key, vi] of entries) {
if (key && !map.has(key)) {
map.set(key, {
name: vi.name || key,
shape: vi.shape || [],
dataType: vi.dataType || 'UNKNOWN'
});
}
}
}
return {
lookup: (name) => map.get(name) || null,
size: map.size
};
}
/**
* Build tooltip text (mirrors TensorShapeInspector._buildTooltipHTML logic, text-only).
*/
function buildTooltipText(tensorName, tensorInfo) {
const displayName = tensorName || 'Unnamed tensor';
if (tensorInfo) {
const shapeStr = tensorInfo.shape && tensorInfo.shape.length > 0
? '[' + tensorInfo.shape.join(', ') + ']'
: 'unknown';
return { name: displayName, shape: shapeStr, dataType: tensorInfo.dataType };
}
return { name: displayName, shape: 'unknown', dataType: null };
}
describe('TensorShapeInspector - Tensor Lookup', () => {
let model;
beforeEach(() => {
model = {
inputs: [
{ name: 'input_0', shape: [1, 3, 224, 224], dataType: 'FLOAT' }
],
outputs: [
{ name: 'output_0', shape: [1, 1000], dataType: 'FLOAT' }
],
graph: {
nodes: [],
edges: [],
valueInfo: {
'conv1_out': { name: 'conv1_out', shape: [1, 64, 112, 112], dataType: 'FLOAT' },
'relu1_out': { name: 'relu1_out', shape: [1, 64, 112, 112], dataType: 'FLOAT' }
}
},
initializers: []
};
});
it('should find input tensors by name', () => {
const { lookup } = buildTensorLookup(model);
const info = lookup('input_0');
expect(info).not.toBeNull();
expect(info.name).toBe('input_0');
expect(info.shape).toEqual([1, 3, 224, 224]);
expect(info.dataType).toBe('FLOAT');
});
it('should find output tensors by name', () => {
const { lookup } = buildTensorLookup(model);
const info = lookup('output_0');
expect(info).not.toBeNull();
expect(info.shape).toEqual([1, 1000]);
expect(info.dataType).toBe('FLOAT');
});
it('should find intermediate tensors from valueInfo', () => {
const { lookup } = buildTensorLookup(model);
const info = lookup('conv1_out');
expect(info).not.toBeNull();
expect(info.shape).toEqual([1, 64, 112, 112]);
expect(info.dataType).toBe('FLOAT');
});
it('should return null for unknown tensor names', () => {
const { lookup } = buildTensorLookup(model);
expect(lookup('nonexistent_tensor')).toBeNull();
});
it('should return null for empty/null name', () => {
const { lookup } = buildTensorLookup(model);
expect(lookup('')).toBeNull();
expect(lookup(null)).toBeNull();
expect(lookup(undefined)).toBeNull();
});
it('should count all tensors from all sources', () => {
const { size } = buildTensorLookup(model);
// 1 input + 1 output + 2 valueInfo = 4
expect(size).toBe(4);
});
it('should handle model with no valueInfo', () => {
model.graph.valueInfo = {};
const { lookup, size } = buildTensorLookup(model);
expect(size).toBe(2); // only input + output
expect(lookup('conv1_out')).toBeNull();
});
it('should handle model with no inputs or outputs', () => {
const emptyModel = { inputs: [], outputs: [], graph: { valueInfo: {} }, initializers: [] };
const { size } = buildTensorLookup(emptyModel);
expect(size).toBe(0);
});
it('should prioritize inputs/outputs over valueInfo for same name', () => {
model.graph.valueInfo['input_0'] = { name: 'input_0', shape: [999], dataType: 'INT32' };
const { lookup } = buildTensorLookup(model);
const info = lookup('input_0');
// Input should take priority
expect(info.shape).toEqual([1, 3, 224, 224]);
expect(info.dataType).toBe('FLOAT');
});
it('should handle tensors with missing shape', () => {
model.graph.valueInfo['no_shape'] = { name: 'no_shape', dataType: 'FLOAT' };
const { lookup } = buildTensorLookup(model);
const info = lookup('no_shape');
expect(info).not.toBeNull();
expect(info.shape).toEqual([]);
});
it('should handle tensors with missing dataType', () => {
model.graph.valueInfo['no_dtype'] = { name: 'no_dtype', shape: [1, 10] };
const { lookup } = buildTensorLookup(model);
const info = lookup('no_dtype');
expect(info.dataType).toBe('UNKNOWN');
});
});
describe('TensorShapeInspector - Tooltip Content', () => {
it('should show shape and type for known tensor', () => {
const tensorInfo = { name: 'conv1_out', shape: [1, 64, 112, 112], dataType: 'FLOAT' };
const result = buildTooltipText('conv1_out', tensorInfo);
expect(result.name).toBe('conv1_out');
expect(result.shape).toBe('[1, 64, 112, 112]');
expect(result.dataType).toBe('FLOAT');
});
it('should show "Shape: unknown" when tensor info is null', () => {
const result = buildTooltipText('mystery_tensor', null);
expect(result.name).toBe('mystery_tensor');
expect(result.shape).toBe('unknown');
expect(result.dataType).toBeNull();
});
it('should show "Shape: unknown" when tensor has empty shape', () => {
const tensorInfo = { name: 'empty', shape: [], dataType: 'FLOAT' };
const result = buildTooltipText('empty', tensorInfo);
expect(result.shape).toBe('unknown');
});
it('should handle dynamic dimensions in shape', () => {
const tensorInfo = { name: 'dynamic', shape: ['batch', 3, 224, 224], dataType: 'FLOAT' };
const result = buildTooltipText('dynamic', tensorInfo);
expect(result.shape).toBe('[batch, 3, 224, 224]');
});
it('should show "Unnamed tensor" when name is empty', () => {
const result = buildTooltipText('', null);
expect(result.name).toBe('Unnamed tensor');
});
});
|