parser_test.go 11.2 KB
Newer Older
1
package parser
2
3

import (
Michael Yang's avatar
Michael Yang committed
4
	"bytes"
5
	"encoding/binary"
Michael Yang's avatar
Michael Yang committed
6
7
	"fmt"
	"io"
8
9
	"strings"
	"testing"
10
	"unicode/utf16"
11
12

	"github.com/stretchr/testify/assert"
Michael Yang's avatar
lint  
Michael Yang committed
13
	"github.com/stretchr/testify/require"
14
15
)

Michael Yang's avatar
Michael Yang committed
16
func TestParseFileFile(t *testing.T) {
17
18
19
20
21
22
23
24
25
26
27
	input := `
FROM model1
ADAPTER adapter1
LICENSE MIT
PARAMETER param1 value1
PARAMETER param2 value2
TEMPLATE template1
`

	reader := strings.NewReader(input)

Michael Yang's avatar
Michael Yang committed
28
	modelfile, err := ParseFile(reader)
Michael Yang's avatar
lint  
Michael Yang committed
29
	require.NoError(t, err)
30
31
32
33
34
35
36
37
38
39

	expectedCommands := []Command{
		{Name: "model", Args: "model1"},
		{Name: "adapter", Args: "adapter1"},
		{Name: "license", Args: "MIT"},
		{Name: "param1", Args: "value1"},
		{Name: "param2", Args: "value2"},
		{Name: "template", Args: "template1"},
	}

Michael Yang's avatar
Michael Yang committed
40
	assert.Equal(t, expectedCommands, modelfile.Commands)
41
42
}

Michael Yang's avatar
Michael Yang committed
43
func TestParseFileFrom(t *testing.T) {
Michael Yang's avatar
tests  
Michael Yang committed
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
	var cases = []struct {
		input    string
		expected []Command
		err      error
	}{
		{
			"FROM foo",
			[]Command{{Name: "model", Args: "foo"}},
			nil,
		},
		{
			"FROM /path/to/model",
			[]Command{{Name: "model", Args: "/path/to/model"}},
			nil,
		},
		{
			"FROM /path/to/model/fp16.bin",
			[]Command{{Name: "model", Args: "/path/to/model/fp16.bin"}},
			nil,
		},
		{
			"FROM llama3:latest",
			[]Command{{Name: "model", Args: "llama3:latest"}},
			nil,
		},
		{
			"FROM llama3:7b-instruct-q4_K_M",
			[]Command{{Name: "model", Args: "llama3:7b-instruct-q4_K_M"}},
			nil,
		},
		{
			"", nil, errMissingFrom,
		},
		{
			"PARAMETER param1 value1",
			nil,
			errMissingFrom,
		},
		{
			"PARAMETER param1 value1\nFROM foo",
			[]Command{{Name: "param1", Args: "value1"}, {Name: "model", Args: "foo"}},
			nil,
		},
	}
88

Michael Yang's avatar
tests  
Michael Yang committed
89
90
	for _, c := range cases {
		t.Run("", func(t *testing.T) {
Michael Yang's avatar
Michael Yang committed
91
			modelfile, err := ParseFile(strings.NewReader(c.input))
Michael Yang's avatar
lint  
Michael Yang committed
92
			require.ErrorIs(t, err, c.err)
Michael Yang's avatar
Michael Yang committed
93
94
95
			if modelfile != nil {
				assert.Equal(t, c.expected, modelfile.Commands)
			}
Michael Yang's avatar
tests  
Michael Yang committed
96
97
		})
	}
98
99
}

Michael Yang's avatar
Michael Yang committed
100
func TestParseFileParametersMissingValue(t *testing.T) {
101
102
103
104
105
106
107
	input := `
FROM foo
PARAMETER param1
`

	reader := strings.NewReader(input)

Michael Yang's avatar
Michael Yang committed
108
	_, err := ParseFile(reader)
Michael Yang's avatar
lint  
Michael Yang committed
109
	require.ErrorIs(t, err, io.ErrUnexpectedEOF)
110
}
111

Michael Yang's avatar
Michael Yang committed
112
func TestParseFileBadCommand(t *testing.T) {
Michael Yang's avatar
Michael Yang committed
113
114
115
116
	input := `
FROM foo
BADCOMMAND param1 value1
`
Michael Yang's avatar
Michael Yang committed
117
	_, err := ParseFile(strings.NewReader(input))
Michael Yang's avatar
lint  
Michael Yang committed
118
	require.ErrorIs(t, err, errInvalidCommand)
Michael Yang's avatar
Michael Yang committed
119
120
}

Michael Yang's avatar
Michael Yang committed
121
func TestParseFileMessages(t *testing.T) {
Michael Yang's avatar
Michael Yang committed
122
123
124
125
126
127
128
129
	var cases = []struct {
		input    string
		expected []Command
		err      error
	}{
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
130
MESSAGE system You are a file parser. Always parse things.
Michael Yang's avatar
Michael Yang committed
131
132
133
`,
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
134
				{Name: "message", Args: "system: You are a file parser. Always parse things."},
Michael Yang's avatar
Michael Yang committed
135
136
137
138
139
			},
			nil,
		},
		{
			`
140
FROM foo
Michael Yang's avatar
Michael Yang committed
141
MESSAGE system You are a file parser. Always parse things.`,
Michael Yang's avatar
Michael Yang committed
142
143
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
144
				{Name: "message", Args: "system: You are a file parser. Always parse things."},
Michael Yang's avatar
Michael Yang committed
145
146
147
148
149
150
			},
			nil,
		},
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
151
MESSAGE system You are a file parser. Always parse things.
152
153
MESSAGE user Hey there!
MESSAGE assistant Hello, I want to parse all the things!
Michael Yang's avatar
Michael Yang committed
154
155
156
`,
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
157
				{Name: "message", Args: "system: You are a file parser. Always parse things."},
Michael Yang's avatar
Michael Yang committed
158
159
160
161
162
163
164
165
166
				{Name: "message", Args: "user: Hey there!"},
				{Name: "message", Args: "assistant: Hello, I want to parse all the things!"},
			},
			nil,
		},
		{
			`
FROM foo
MESSAGE system """
Michael Yang's avatar
Michael Yang committed
167
You are a multiline file parser. Always parse things.
Michael Yang's avatar
Michael Yang committed
168
169
170
171
"""
			`,
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
172
				{Name: "message", Args: "system: \nYou are a multiline file parser. Always parse things.\n"},
Michael Yang's avatar
Michael Yang committed
173
174
175
176
177
178
179
180
181
			},
			nil,
		},
		{
			`
FROM foo
MESSAGE badguy I'm a bad guy!
`,
			nil,
Michael Yang's avatar
Michael Yang committed
182
			errInvalidMessageRole,
Michael Yang's avatar
Michael Yang committed
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
		},
		{
			`
FROM foo
MESSAGE system
`,
			nil,
			io.ErrUnexpectedEOF,
		},
		{
			`
FROM foo
MESSAGE system`,
			nil,
			io.ErrUnexpectedEOF,
		},
	}
200

Michael Yang's avatar
Michael Yang committed
201
202
	for _, c := range cases {
		t.Run("", func(t *testing.T) {
Michael Yang's avatar
Michael Yang committed
203
			modelfile, err := ParseFile(strings.NewReader(c.input))
Michael Yang's avatar
lint  
Michael Yang committed
204
			require.ErrorIs(t, err, c.err)
Michael Yang's avatar
Michael Yang committed
205
206
207
			if modelfile != nil {
				assert.Equal(t, c.expected, modelfile.Commands)
			}
Michael Yang's avatar
Michael Yang committed
208
209
210
		})
	}
}
211

Michael Yang's avatar
Michael Yang committed
212
func TestParseFileQuoted(t *testing.T) {
Michael Yang's avatar
Michael Yang committed
213
214
215
216
217
218
219
220
	var cases = []struct {
		multiline string
		expected  []Command
		err       error
	}{
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
221
SYSTEM """
Michael Yang's avatar
Michael Yang committed
222
This is a
Michael Yang's avatar
Michael Yang committed
223
multiline system.
Michael Yang's avatar
Michael Yang committed
224
225
226
227
"""
			`,
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
228
				{Name: "system", Args: "\nThis is a\nmultiline system.\n"},
Michael Yang's avatar
Michael Yang committed
229
230
231
232
233
234
			},
			nil,
		},
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
235
SYSTEM """
Michael Yang's avatar
Michael Yang committed
236
This is a
Michael Yang's avatar
Michael Yang committed
237
multiline system."""
Michael Yang's avatar
Michael Yang committed
238
239
240
			`,
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
241
				{Name: "system", Args: "\nThis is a\nmultiline system."},
Michael Yang's avatar
Michael Yang committed
242
243
244
245
246
247
			},
			nil,
		},
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
248
249
SYSTEM """This is a
multiline system."""
Michael Yang's avatar
Michael Yang committed
250
251
252
			`,
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
253
				{Name: "system", Args: "This is a\nmultiline system."},
Michael Yang's avatar
Michael Yang committed
254
255
256
257
258
259
			},
			nil,
		},
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
260
SYSTEM """This is a multiline system."""
Michael Yang's avatar
Michael Yang committed
261
262
263
			`,
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
264
				{Name: "system", Args: "This is a multiline system."},
Michael Yang's avatar
Michael Yang committed
265
266
267
268
269
270
			},
			nil,
		},
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
271
SYSTEM """This is a multiline system.""
Michael Yang's avatar
Michael Yang committed
272
273
274
275
276
277
278
			`,
			nil,
			io.ErrUnexpectedEOF,
		},
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
279
SYSTEM "
Michael Yang's avatar
Michael Yang committed
280
281
282
283
284
285
286
			`,
			nil,
			io.ErrUnexpectedEOF,
		},
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
287
288
SYSTEM """
This is a multiline system with "quotes".
Michael Yang's avatar
Michael Yang committed
289
290
291
292
"""
`,
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
293
				{Name: "system", Args: "\nThis is a multiline system with \"quotes\".\n"},
Michael Yang's avatar
Michael Yang committed
294
295
296
297
298
299
			},
			nil,
		},
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
300
SYSTEM """"""
Michael Yang's avatar
Michael Yang committed
301
302
303
`,
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
304
				{Name: "system", Args: ""},
Michael Yang's avatar
Michael Yang committed
305
306
307
308
309
310
			},
			nil,
		},
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
311
SYSTEM ""
Michael Yang's avatar
Michael Yang committed
312
313
314
`,
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
315
				{Name: "system", Args: ""},
Michael Yang's avatar
Michael Yang committed
316
317
318
319
320
321
			},
			nil,
		},
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
322
SYSTEM "'"
Michael Yang's avatar
Michael Yang committed
323
324
325
`,
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
326
				{Name: "system", Args: "'"},
Michael Yang's avatar
Michael Yang committed
327
328
329
			},
			nil,
		},
Michael Yang's avatar
tests  
Michael Yang committed
330
331
332
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
333
SYSTEM """''"'""'""'"'''''""'""'"""
Michael Yang's avatar
tests  
Michael Yang committed
334
335
336
`,
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
337
338
339
340
341
342
343
344
345
346
347
348
349
				{Name: "system", Args: `''"'""'""'"'''''""'""'`},
			},
			nil,
		},
		{
			`
FROM foo
TEMPLATE """
{{ .Prompt }}
"""`,
			[]Command{
				{Name: "model", Args: "foo"},
				{Name: "template", Args: "\n{{ .Prompt }}\n"},
Michael Yang's avatar
tests  
Michael Yang committed
350
351
352
			},
			nil,
		},
353
354
	}

Michael Yang's avatar
Michael Yang committed
355
356
	for _, c := range cases {
		t.Run("", func(t *testing.T) {
Michael Yang's avatar
Michael Yang committed
357
			modelfile, err := ParseFile(strings.NewReader(c.multiline))
Michael Yang's avatar
lint  
Michael Yang committed
358
			require.ErrorIs(t, err, c.err)
Michael Yang's avatar
Michael Yang committed
359
360
361
			if modelfile != nil {
				assert.Equal(t, c.expected, modelfile.Commands)
			}
Michael Yang's avatar
Michael Yang committed
362
363
		})
	}
364
365
}

Michael Yang's avatar
Michael Yang committed
366
func TestParseFileParameters(t *testing.T) {
Michael Yang's avatar
tests  
Michael Yang committed
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
	var cases = map[string]struct {
		name, value string
	}{
		"numa true":                    {"numa", "true"},
		"num_ctx 1":                    {"num_ctx", "1"},
		"num_batch 1":                  {"num_batch", "1"},
		"num_gqa 1":                    {"num_gqa", "1"},
		"num_gpu 1":                    {"num_gpu", "1"},
		"main_gpu 1":                   {"main_gpu", "1"},
		"low_vram true":                {"low_vram", "true"},
		"f16_kv true":                  {"f16_kv", "true"},
		"logits_all true":              {"logits_all", "true"},
		"vocab_only true":              {"vocab_only", "true"},
		"use_mmap true":                {"use_mmap", "true"},
		"use_mlock true":               {"use_mlock", "true"},
		"num_thread 1":                 {"num_thread", "1"},
		"num_keep 1":                   {"num_keep", "1"},
		"seed 1":                       {"seed", "1"},
		"num_predict 1":                {"num_predict", "1"},
		"top_k 1":                      {"top_k", "1"},
		"top_p 1.0":                    {"top_p", "1.0"},
		"tfs_z 1.0":                    {"tfs_z", "1.0"},
		"typical_p 1.0":                {"typical_p", "1.0"},
		"repeat_last_n 1":              {"repeat_last_n", "1"},
		"temperature 1.0":              {"temperature", "1.0"},
		"repeat_penalty 1.0":           {"repeat_penalty", "1.0"},
		"presence_penalty 1.0":         {"presence_penalty", "1.0"},
		"frequency_penalty 1.0":        {"frequency_penalty", "1.0"},
		"mirostat 1":                   {"mirostat", "1"},
		"mirostat_tau 1.0":             {"mirostat_tau", "1.0"},
		"mirostat_eta 1.0":             {"mirostat_eta", "1.0"},
		"penalize_newline true":        {"penalize_newline", "true"},
		"stop ### User:":               {"stop", "### User:"},
		"stop ### User: ":              {"stop", "### User: "},
		"stop \"### User:\"":           {"stop", "### User:"},
		"stop \"### User: \"":          {"stop", "### User: "},
		"stop \"\"\"### User:\"\"\"":   {"stop", "### User:"},
		"stop \"\"\"### User:\n\"\"\"": {"stop", "### User:\n"},
		"stop <|endoftext|>":           {"stop", "<|endoftext|>"},
		"stop <|eot_id|>":              {"stop", "<|eot_id|>"},
		"stop </s>":                    {"stop", "</s>"},
Michael Yang's avatar
Michael Yang committed
408
	}
409

Michael Yang's avatar
tests  
Michael Yang committed
410
411
	for k, v := range cases {
		t.Run(k, func(t *testing.T) {
Michael Yang's avatar
Michael Yang committed
412
413
			var b bytes.Buffer
			fmt.Fprintln(&b, "FROM foo")
Michael Yang's avatar
tests  
Michael Yang committed
414
			fmt.Fprintln(&b, "PARAMETER", k)
Michael Yang's avatar
Michael Yang committed
415
			modelfile, err := ParseFile(&b)
Michael Yang's avatar
lint  
Michael Yang committed
416
			require.NoError(t, err)
Michael Yang's avatar
tests  
Michael Yang committed
417
418
419
420

			assert.Equal(t, []Command{
				{Name: "model", Args: "foo"},
				{Name: v.name, Args: v.value},
Michael Yang's avatar
Michael Yang committed
421
			}, modelfile.Commands)
Michael Yang's avatar
Michael Yang committed
422
423
424
425
		})
	}
}

Michael Yang's avatar
Michael Yang committed
426
func TestParseFileComments(t *testing.T) {
Michael Yang's avatar
Michael Yang committed
427
428
429
430
431
432
433
	var cases = []struct {
		input    string
		expected []Command
	}{
		{
			`
# comment
434
FROM foo
Michael Yang's avatar
Michael Yang committed
435
436
437
438
439
440
	`,
			[]Command{
				{Name: "model", Args: "foo"},
			},
		},
	}
441

Michael Yang's avatar
Michael Yang committed
442
443
	for _, c := range cases {
		t.Run("", func(t *testing.T) {
Michael Yang's avatar
Michael Yang committed
444
			modelfile, err := ParseFile(strings.NewReader(c.input))
Michael Yang's avatar
lint  
Michael Yang committed
445
			require.NoError(t, err)
Michael Yang's avatar
Michael Yang committed
446
			assert.Equal(t, c.expected, modelfile.Commands)
Michael Yang's avatar
Michael Yang committed
447
448
		})
	}
449
}
Michael Yang's avatar
Michael Yang committed
450

Michael Yang's avatar
Michael Yang committed
451
func TestParseFileFormatParseFile(t *testing.T) {
Michael Yang's avatar
Michael Yang committed
452
453
454
455
456
457
458
459
	var cases = []string{
		`
FROM foo
ADAPTER adapter1
LICENSE MIT
PARAMETER param1 value1
PARAMETER param2 value2
TEMPLATE template1
Michael Yang's avatar
Michael Yang committed
460
MESSAGE system You are a file parser. Always parse things.
Michael Yang's avatar
Michael Yang committed
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
MESSAGE user Hey there!
MESSAGE assistant Hello, I want to parse all the things!
`,
		`
FROM foo
ADAPTER adapter1
LICENSE MIT
PARAMETER param1 value1
PARAMETER param2 value2
TEMPLATE template1
MESSAGE system """
You are a store greeter. Always responsed with "Hello!".
"""
MESSAGE user Hey there!
MESSAGE assistant Hello, I want to parse all the things!
`,
		`
FROM foo
ADAPTER adapter1
LICENSE """
Very long and boring legal text.
Blah blah blah.
"Oh look, a quote!"
"""

PARAMETER param1 value1
PARAMETER param2 value2
TEMPLATE template1
MESSAGE system """
You are a store greeter. Always responsed with "Hello!".
"""
MESSAGE user Hey there!
MESSAGE assistant Hello, I want to parse all the things!
494
495
496
497
`,
		`
FROM foo
SYSTEM ""
Michael Yang's avatar
Michael Yang committed
498
499
500
501
502
`,
	}

	for _, c := range cases {
		t.Run("", func(t *testing.T) {
Michael Yang's avatar
Michael Yang committed
503
			modelfile, err := ParseFile(strings.NewReader(c))
Michael Yang's avatar
lint  
Michael Yang committed
504
			require.NoError(t, err)
Michael Yang's avatar
Michael Yang committed
505

Michael Yang's avatar
Michael Yang committed
506
			modelfile2, err := ParseFile(strings.NewReader(modelfile.String()))
Michael Yang's avatar
lint  
Michael Yang committed
507
			require.NoError(t, err)
Michael Yang's avatar
Michael Yang committed
508

Michael Yang's avatar
Michael Yang committed
509
			assert.Equal(t, modelfile, modelfile2)
Michael Yang's avatar
Michael Yang committed
510
511
512
		})
	}
}
513
514
515
516
517
518
519
520
521
522
523

func TestParseFileUTF16ParseFile(t *testing.T) {
	data := `FROM bob
PARAMETER param1 1
PARAMETER param2 4096
SYSTEM You are a utf16 file.
`
	// simulate a utf16 le file
	utf16File := utf16.Encode(append([]rune{'\ufffe'}, []rune(data)...))
	buf := new(bytes.Buffer)
	err := binary.Write(buf, binary.LittleEndian, utf16File)
Michael Yang's avatar
lint  
Michael Yang committed
524
	require.NoError(t, err)
525
526

	actual, err := ParseFile(buf)
Michael Yang's avatar
lint  
Michael Yang committed
527
	require.NoError(t, err)
528
529
530
531
532
533
534
535
536
537
538
539
540

	expected := []Command{
		{Name: "model", Args: "bob"},
		{Name: "param1", Args: "1"},
		{Name: "param2", Args: "4096"},
		{Name: "system", Args: "You are a utf16 file."},
	}

	assert.Equal(t, expected, actual.Commands)

	// simulate a utf16 be file
	buf = new(bytes.Buffer)
	err = binary.Write(buf, binary.BigEndian, utf16File)
Michael Yang's avatar
lint  
Michael Yang committed
541
	require.NoError(t, err)
542
543

	actual, err = ParseFile(buf)
Michael Yang's avatar
lint  
Michael Yang committed
544
	require.NoError(t, err)
545
546
	assert.Equal(t, expected, actual.Commands)
}