parser_test.go 17.6 KB
Newer Older
1
package parser
2
3

import (
Michael Yang's avatar
Michael Yang committed
4
	"bytes"
5
	"crypto/sha256"
6
	"encoding/binary"
7
	"errors"
Michael Yang's avatar
Michael Yang committed
8
9
	"fmt"
	"io"
10
	"os"
11
12
	"strings"
	"testing"
13
	"unicode/utf16"
14

15
	"github.com/google/go-cmp/cmp"
16
	"github.com/stretchr/testify/assert"
Michael Yang's avatar
lint  
Michael Yang committed
17
	"github.com/stretchr/testify/require"
18
19
	"golang.org/x/text/encoding"
	"golang.org/x/text/encoding/unicode"
20
21

	"github.com/ollama/ollama/api"
Michael Yang's avatar
Michael Yang committed
22
	"github.com/ollama/ollama/fs/ggml"
23
24
)

Michael Yang's avatar
Michael Yang committed
25
func TestParseFileFile(t *testing.T) {
26
27
28
29
30
31
	input := `
FROM model1
ADAPTER adapter1
LICENSE MIT
PARAMETER param1 value1
PARAMETER param2 value2
Josh Yan's avatar
Josh Yan committed
32
33
34
35
36
37
38
TEMPLATE """{{ if .System }}<|start_header_id|>system<|end_header_id|>

{{ .System }}<|eot_id|>{{ end }}{{ if .Prompt }}<|start_header_id|>user<|end_header_id|>

{{ .Prompt }}<|eot_id|>{{ end }}<|start_header_id|>assistant<|end_header_id|>

{{ .Response }}<|eot_id|>"""    
39
40
41
42
`

	reader := strings.NewReader(input)

Michael Yang's avatar
Michael Yang committed
43
	modelfile, err := ParseFile(reader)
Michael Yang's avatar
lint  
Michael Yang committed
44
	require.NoError(t, err)
45
46
47
48
49
50
51

	expectedCommands := []Command{
		{Name: "model", Args: "model1"},
		{Name: "adapter", Args: "adapter1"},
		{Name: "license", Args: "MIT"},
		{Name: "param1", Args: "value1"},
		{Name: "param2", Args: "value2"},
Josh Yan's avatar
Josh Yan committed
52
		{Name: "template", Args: "{{ if .System }}<|start_header_id|>system<|end_header_id|>\n\n{{ .System }}<|eot_id|>{{ end }}{{ if .Prompt }}<|start_header_id|>user<|end_header_id|>\n\n{{ .Prompt }}<|eot_id|>{{ end }}<|start_header_id|>assistant<|end_header_id|>\n\n{{ .Response }}<|eot_id|>"},
53
54
	}

Michael Yang's avatar
Michael Yang committed
55
	assert.Equal(t, expectedCommands, modelfile.Commands)
56
57
}

Josh Yan's avatar
Josh Yan committed
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
func TestParseFileTrimSpace(t *testing.T) {
	input := `
FROM "     model 1"
ADAPTER      adapter3
LICENSE "MIT       "
PARAMETER param1        value1
PARAMETER param2    value2
TEMPLATE """   {{ if .System }}<|start_header_id|>system<|end_header_id|>

{{ .System }}<|eot_id|>{{ end }}{{ if .Prompt }}<|start_header_id|>user<|end_header_id|>

{{ .Prompt }}<|eot_id|>{{ end }}<|start_header_id|>assistant<|end_header_id|>

{{ .Response }}<|eot_id|>   """    
`

	reader := strings.NewReader(input)

	modelfile, err := ParseFile(reader)
	require.NoError(t, err)

	expectedCommands := []Command{
		{Name: "model", Args: "     model 1"},
		{Name: "adapter", Args: "adapter3"},
		{Name: "license", Args: "MIT       "},
		{Name: "param1", Args: "value1"},
		{Name: "param2", Args: "value2"},
		{Name: "template", Args: "   {{ if .System }}<|start_header_id|>system<|end_header_id|>\n\n{{ .System }}<|eot_id|>{{ end }}{{ if .Prompt }}<|start_header_id|>user<|end_header_id|>\n\n{{ .Prompt }}<|eot_id|>{{ end }}<|start_header_id|>assistant<|end_header_id|>\n\n{{ .Response }}<|eot_id|>   "},
	}

	assert.Equal(t, expectedCommands, modelfile.Commands)
}

Michael Yang's avatar
Michael Yang committed
91
func TestParseFileFrom(t *testing.T) {
Michael Yang's avatar
lint  
Michael Yang committed
92
	cases := []struct {
Michael Yang's avatar
tests  
Michael Yang committed
93
94
95
96
		input    string
		expected []Command
		err      error
	}{
Josh Yan's avatar
Josh Yan committed
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
		{
			"FROM \"FOO  BAR  \"",
			[]Command{{Name: "model", Args: "FOO  BAR  "}},
			nil,
		},
		{
			"FROM \"FOO BAR\"\nPARAMETER param1 value1",
			[]Command{{Name: "model", Args: "FOO BAR"}, {Name: "param1", Args: "value1"}},
			nil,
		},
		{
			"FROM     FOOO BAR    ",
			[]Command{{Name: "model", Args: "FOOO BAR"}},
			nil,
		},
		{
			"FROM /what/is/the path ",
			[]Command{{Name: "model", Args: "/what/is/the path"}},
			nil,
		},
Michael Yang's avatar
tests  
Michael Yang committed
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
		{
			"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,
		},
Josh Yan's avatar
Josh Yan committed
155
156
		{
			"PARAMETER what the \nFROM lemons make lemonade ",
Josh Yan's avatar
Josh Yan committed
157
			[]Command{{Name: "what", Args: "the"}, {Name: "model", Args: "lemons make lemonade"}},
Josh Yan's avatar
Josh Yan committed
158
159
			nil,
		},
Michael Yang's avatar
tests  
Michael Yang committed
160
	}
161

Michael Yang's avatar
tests  
Michael Yang committed
162
163
	for _, c := range cases {
		t.Run("", func(t *testing.T) {
Michael Yang's avatar
Michael Yang committed
164
			modelfile, err := ParseFile(strings.NewReader(c.input))
Michael Yang's avatar
lint  
Michael Yang committed
165
			require.ErrorIs(t, err, c.err)
Michael Yang's avatar
Michael Yang committed
166
167
168
			if modelfile != nil {
				assert.Equal(t, c.expected, modelfile.Commands)
			}
Michael Yang's avatar
tests  
Michael Yang committed
169
170
		})
	}
171
172
}

Michael Yang's avatar
Michael Yang committed
173
func TestParseFileParametersMissingValue(t *testing.T) {
174
175
176
177
178
179
180
	input := `
FROM foo
PARAMETER param1
`

	reader := strings.NewReader(input)

Michael Yang's avatar
Michael Yang committed
181
	_, err := ParseFile(reader)
Michael Yang's avatar
lint  
Michael Yang committed
182
	require.ErrorIs(t, err, io.ErrUnexpectedEOF)
183
}
184

Michael Yang's avatar
Michael Yang committed
185
func TestParseFileBadCommand(t *testing.T) {
Michael Yang's avatar
Michael Yang committed
186
187
188
189
	input := `
FROM foo
BADCOMMAND param1 value1
`
190
191
192
193
194
	parserError := &ParserError{
		LineNumber: 3,
		Msg:        errInvalidCommand.Error(),
	}

Michael Yang's avatar
Michael Yang committed
195
	_, err := ParseFile(strings.NewReader(input))
196
197
198
	if !errors.As(err, &parserError) {
		t.Errorf("unexpected error: expected: %s, actual: %s", parserError.Error(), err.Error())
	}
Michael Yang's avatar
Michael Yang committed
199
200
}

Michael Yang's avatar
Michael Yang committed
201
func TestParseFileMessages(t *testing.T) {
Michael Yang's avatar
lint  
Michael Yang committed
202
	cases := []struct {
Michael Yang's avatar
Michael Yang committed
203
204
205
206
207
208
209
		input    string
		expected []Command
		err      error
	}{
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
210
MESSAGE system You are a file parser. Always parse things.
Michael Yang's avatar
Michael Yang committed
211
212
213
`,
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
214
				{Name: "message", Args: "system: You are a file parser. Always parse things."},
Michael Yang's avatar
Michael Yang committed
215
216
217
218
219
			},
			nil,
		},
		{
			`
220
FROM foo
Michael Yang's avatar
Michael Yang committed
221
MESSAGE system You are a file parser. Always parse things.`,
Michael Yang's avatar
Michael Yang committed
222
223
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
224
				{Name: "message", Args: "system: You are a file parser. Always parse things."},
Michael Yang's avatar
Michael Yang committed
225
226
227
228
229
230
			},
			nil,
		},
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
231
MESSAGE system You are a file parser. Always parse things.
232
233
MESSAGE user Hey there!
MESSAGE assistant Hello, I want to parse all the things!
Michael Yang's avatar
Michael Yang committed
234
235
236
`,
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
237
				{Name: "message", Args: "system: You are a file parser. Always parse things."},
Michael Yang's avatar
Michael Yang committed
238
239
240
241
242
243
244
245
246
				{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
247
You are a multiline file parser. Always parse things.
Michael Yang's avatar
Michael Yang committed
248
249
250
251
"""
			`,
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
252
				{Name: "message", Args: "system: \nYou are a multiline file parser. Always parse things.\n"},
Michael Yang's avatar
Michael Yang committed
253
254
255
256
257
258
			},
			nil,
		},
		{
			`
FROM foo
259
MESSAGE badguy I'm a bad guy!
Michael Yang's avatar
Michael Yang committed
260
`,
261
			nil,
262
263
264
265
			&ParserError{
				LineNumber: 3,
				Msg:        errInvalidMessageRole.Error(),
			},
Michael Yang's avatar
Michael Yang committed
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
		},
		{
			`
FROM foo
MESSAGE system
`,
			nil,
			io.ErrUnexpectedEOF,
		},
		{
			`
FROM foo
MESSAGE system`,
			nil,
			io.ErrUnexpectedEOF,
		},
	}
283

284
	for _, tt := range cases {
Michael Yang's avatar
Michael Yang committed
285
		t.Run("", func(t *testing.T) {
286
287
			modelfile, err := ParseFile(strings.NewReader(tt.input))

Michael Yang's avatar
Michael Yang committed
288
			if modelfile != nil {
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
				assert.Equal(t, tt.expected, modelfile.Commands)
			}

			if tt.err == nil {
				if err != nil {
					t.Fatalf("expected no error, but got %v", err)
				}
				return
			}

			switch tt.err.(type) {
			case *ParserError:
				var pErr *ParserError
				if errors.As(err, &pErr) {
					// got the correct type of error
					return
				}
			}

			if errors.Is(err, tt.err) {
				return
Michael Yang's avatar
Michael Yang committed
310
			}
311
312

			t.Fatalf("unexpected error: expected: %v, actual: %v", tt.err, err)
Michael Yang's avatar
Michael Yang committed
313
314
315
		})
	}
}
316

Michael Yang's avatar
Michael Yang committed
317
func TestParseFileQuoted(t *testing.T) {
Michael Yang's avatar
lint  
Michael Yang committed
318
	cases := []struct {
Michael Yang's avatar
Michael Yang committed
319
320
321
322
323
324
325
		multiline string
		expected  []Command
		err       error
	}{
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
326
SYSTEM """
Michael Yang's avatar
Michael Yang committed
327
This is a
Michael Yang's avatar
Michael Yang committed
328
multiline system.
Michael Yang's avatar
Michael Yang committed
329
330
331
332
"""
			`,
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
333
				{Name: "system", Args: "\nThis is a\nmultiline system.\n"},
Michael Yang's avatar
Michael Yang committed
334
335
336
337
338
339
			},
			nil,
		},
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
340
SYSTEM """
Michael Yang's avatar
Michael Yang committed
341
This is a
Michael Yang's avatar
Michael Yang committed
342
multiline system."""
Michael Yang's avatar
Michael Yang committed
343
344
345
			`,
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
346
				{Name: "system", Args: "\nThis is a\nmultiline system."},
Michael Yang's avatar
Michael Yang committed
347
348
349
350
351
352
			},
			nil,
		},
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
353
354
SYSTEM """This is a
multiline system."""
Michael Yang's avatar
Michael Yang committed
355
356
357
			`,
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
358
				{Name: "system", Args: "This is a\nmultiline system."},
Michael Yang's avatar
Michael Yang committed
359
360
361
362
363
364
			},
			nil,
		},
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
365
SYSTEM """This is a multiline system."""
Michael Yang's avatar
Michael Yang committed
366
367
368
			`,
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
369
				{Name: "system", Args: "This is a multiline system."},
Michael Yang's avatar
Michael Yang committed
370
371
372
373
374
375
			},
			nil,
		},
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
376
SYSTEM """This is a multiline system.""
Michael Yang's avatar
Michael Yang committed
377
378
379
380
381
382
383
			`,
			nil,
			io.ErrUnexpectedEOF,
		},
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
384
SYSTEM "
Michael Yang's avatar
Michael Yang committed
385
386
387
388
389
390
391
			`,
			nil,
			io.ErrUnexpectedEOF,
		},
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
392
393
SYSTEM """
This is a multiline system with "quotes".
Michael Yang's avatar
Michael Yang committed
394
395
396
397
"""
`,
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
398
				{Name: "system", Args: "\nThis is a multiline system with \"quotes\".\n"},
Michael Yang's avatar
Michael Yang committed
399
400
401
402
403
404
			},
			nil,
		},
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
405
SYSTEM """"""
Michael Yang's avatar
Michael Yang committed
406
407
408
`,
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
409
				{Name: "system", Args: ""},
Michael Yang's avatar
Michael Yang committed
410
411
412
413
414
415
			},
			nil,
		},
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
416
SYSTEM ""
Michael Yang's avatar
Michael Yang committed
417
418
419
`,
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
420
				{Name: "system", Args: ""},
Michael Yang's avatar
Michael Yang committed
421
422
423
424
425
426
			},
			nil,
		},
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
427
SYSTEM "'"
Michael Yang's avatar
Michael Yang committed
428
429
430
`,
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
431
				{Name: "system", Args: "'"},
Michael Yang's avatar
Michael Yang committed
432
433
434
			},
			nil,
		},
Michael Yang's avatar
tests  
Michael Yang committed
435
436
437
		{
			`
FROM foo
Michael Yang's avatar
Michael Yang committed
438
SYSTEM """''"'""'""'"'''''""'""'"""
Michael Yang's avatar
tests  
Michael Yang committed
439
440
441
`,
			[]Command{
				{Name: "model", Args: "foo"},
Michael Yang's avatar
Michael Yang committed
442
443
444
445
446
447
448
449
450
451
452
453
454
				{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
455
456
457
			},
			nil,
		},
458
459
	}

Michael Yang's avatar
Michael Yang committed
460
461
	for _, c := range cases {
		t.Run("", func(t *testing.T) {
Michael Yang's avatar
Michael Yang committed
462
			modelfile, err := ParseFile(strings.NewReader(c.multiline))
Michael Yang's avatar
lint  
Michael Yang committed
463
			require.ErrorIs(t, err, c.err)
Michael Yang's avatar
Michael Yang committed
464
465
466
			if modelfile != nil {
				assert.Equal(t, c.expected, modelfile.Commands)
			}
Michael Yang's avatar
Michael Yang committed
467
468
		})
	}
469
470
}

Michael Yang's avatar
Michael Yang committed
471
func TestParseFileParameters(t *testing.T) {
Michael Yang's avatar
lint  
Michael Yang committed
472
	cases := map[string]struct {
Michael Yang's avatar
tests  
Michael Yang committed
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
		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"},
		"use_mmap true":                {"use_mmap", "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"},
488
		"min_p 0.05":                   {"min_p", "0.05"},
Michael Yang's avatar
tests  
Michael Yang committed
489
490
491
492
493
494
495
496
497
498
499
		"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:"},
Josh Yan's avatar
Josh Yan committed
500
		"stop ### User: ":              {"stop", "### User:"},
Michael Yang's avatar
tests  
Michael Yang committed
501
502
503
504
505
506
507
		"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
508
	}
509

Michael Yang's avatar
tests  
Michael Yang committed
510
511
	for k, v := range cases {
		t.Run(k, func(t *testing.T) {
Michael Yang's avatar
Michael Yang committed
512
513
			var b bytes.Buffer
			fmt.Fprintln(&b, "FROM foo")
Michael Yang's avatar
tests  
Michael Yang committed
514
			fmt.Fprintln(&b, "PARAMETER", k)
Michael Yang's avatar
Michael Yang committed
515
			modelfile, err := ParseFile(&b)
Michael Yang's avatar
lint  
Michael Yang committed
516
			require.NoError(t, err)
Michael Yang's avatar
tests  
Michael Yang committed
517
518
519
520

			assert.Equal(t, []Command{
				{Name: "model", Args: "foo"},
				{Name: v.name, Args: v.value},
Michael Yang's avatar
Michael Yang committed
521
			}, modelfile.Commands)
Michael Yang's avatar
Michael Yang committed
522
523
524
525
		})
	}
}

Michael Yang's avatar
Michael Yang committed
526
func TestParseFileComments(t *testing.T) {
Michael Yang's avatar
lint  
Michael Yang committed
527
	cases := []struct {
Michael Yang's avatar
Michael Yang committed
528
529
530
531
532
533
		input    string
		expected []Command
	}{
		{
			`
# comment
534
FROM foo
Michael Yang's avatar
Michael Yang committed
535
536
537
538
539
540
	`,
			[]Command{
				{Name: "model", Args: "foo"},
			},
		},
	}
541

Michael Yang's avatar
Michael Yang committed
542
543
	for _, c := range cases {
		t.Run("", func(t *testing.T) {
Michael Yang's avatar
Michael Yang committed
544
			modelfile, err := ParseFile(strings.NewReader(c.input))
Michael Yang's avatar
lint  
Michael Yang committed
545
			require.NoError(t, err)
Michael Yang's avatar
Michael Yang committed
546
			assert.Equal(t, c.expected, modelfile.Commands)
Michael Yang's avatar
Michael Yang committed
547
548
		})
	}
549
}
Michael Yang's avatar
Michael Yang committed
550

Michael Yang's avatar
Michael Yang committed
551
func TestParseFileFormatParseFile(t *testing.T) {
Michael Yang's avatar
lint  
Michael Yang committed
552
	cases := []string{
Michael Yang's avatar
Michael Yang committed
553
554
555
556
557
558
559
		`
FROM foo
ADAPTER adapter1
LICENSE MIT
PARAMETER param1 value1
PARAMETER param2 value2
TEMPLATE template1
Michael Yang's avatar
Michael Yang committed
560
MESSAGE system You are a file parser. Always parse things.
Michael Yang's avatar
Michael Yang committed
561
562
563
564
565
566
567
568
569
570
571
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 """
572
You are a store greeter. Always respond with "Hello!".
Michael Yang's avatar
Michael Yang committed
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
"""
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 """
590
You are a store greeter. Always respond with "Hello!".
Michael Yang's avatar
Michael Yang committed
591
592
593
"""
MESSAGE user Hey there!
MESSAGE assistant Hello, I want to parse all the things!
594
595
596
597
`,
		`
FROM foo
SYSTEM ""
Michael Yang's avatar
Michael Yang committed
598
599
600
601
602
`,
	}

	for _, c := range cases {
		t.Run("", func(t *testing.T) {
Michael Yang's avatar
Michael Yang committed
603
			modelfile, err := ParseFile(strings.NewReader(c))
Michael Yang's avatar
lint  
Michael Yang committed
604
			require.NoError(t, err)
Michael Yang's avatar
Michael Yang committed
605

Michael Yang's avatar
Michael Yang committed
606
			modelfile2, err := ParseFile(strings.NewReader(modelfile.String()))
Michael Yang's avatar
lint  
Michael Yang committed
607
			require.NoError(t, err)
Michael Yang's avatar
Michael Yang committed
608

Michael Yang's avatar
Michael Yang committed
609
			assert.Equal(t, modelfile, modelfile2)
Michael Yang's avatar
Michael Yang committed
610
611
612
		})
	}
}
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627

func TestParseFileUTF16ParseFile(t *testing.T) {
	data := `FROM bob
PARAMETER param1 1
PARAMETER param2 4096
SYSTEM You are a utf16 file.
`

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

628
629
630
631
	t.Run("le", func(t *testing.T) {
		var b bytes.Buffer
		require.NoError(t, binary.Write(&b, binary.LittleEndian, []byte{0xff, 0xfe}))
		require.NoError(t, binary.Write(&b, binary.LittleEndian, utf16.Encode([]rune(data))))
632

633
634
		actual, err := ParseFile(&b)
		require.NoError(t, err)
635

636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
		assert.Equal(t, expected, actual.Commands)
	})

	t.Run("be", func(t *testing.T) {
		var b bytes.Buffer
		require.NoError(t, binary.Write(&b, binary.BigEndian, []byte{0xfe, 0xff}))
		require.NoError(t, binary.Write(&b, binary.BigEndian, utf16.Encode([]rune(data))))

		actual, err := ParseFile(&b)
		require.NoError(t, err)
		assert.Equal(t, expected, actual.Commands)
	})
}

func TestParseMultiByte(t *testing.T) {
	input := `FROM test
	SYSTEM 你好👋`

	expect := []Command{
		{Name: "model", Args: "test"},
		{Name: "system", Args: "你好👋"},
	}

	encodings := []encoding.Encoding{
		unicode.UTF8,
		unicode.UTF16(unicode.LittleEndian, unicode.UseBOM),
		unicode.UTF16(unicode.BigEndian, unicode.UseBOM),
	}

	for _, encoding := range encodings {
		t.Run(fmt.Sprintf("%s", encoding), func(t *testing.T) {
			s, err := encoding.NewEncoder().String(input)
			require.NoError(t, err)

			actual, err := ParseFile(strings.NewReader(s))
			require.NoError(t, err)

			assert.Equal(t, expect, actual.Commands)
		})
	}
676
}
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744

func TestCreateRequest(t *testing.T) {
	cases := []struct {
		input    string
		expected *api.CreateRequest
	}{
		{
			`FROM test`,
			&api.CreateRequest{From: "test"},
		},
		{
			`FROM test
TEMPLATE some template
`,
			&api.CreateRequest{
				From:     "test",
				Template: "some template",
			},
		},
		{
			`FROM test
LICENSE single license
PARAMETER temperature 0.5
MESSAGE user Hello
`,
			&api.CreateRequest{
				From:       "test",
				License:    []string{"single license"},
				Parameters: map[string]any{"temperature": float32(0.5)},
				Messages: []api.Message{
					{Role: "user", Content: "Hello"},
				},
			},
		},
		{
			`FROM test
PARAMETER temperature 0.5
PARAMETER top_k 1
SYSTEM You are a bot.
LICENSE license1
LICENSE license2
MESSAGE user Hello there!
MESSAGE assistant Hi! How are you?
`,
			&api.CreateRequest{
				From:       "test",
				License:    []string{"license1", "license2"},
				System:     "You are a bot.",
				Parameters: map[string]any{"temperature": float32(0.5), "top_k": int64(1)},
				Messages: []api.Message{
					{Role: "user", Content: "Hello there!"},
					{Role: "assistant", Content: "Hi! How are you?"},
				},
			},
		},
	}

	for _, c := range cases {
		s, err := unicode.UTF8.NewEncoder().String(c.input)
		if err != nil {
			t.Fatal(err)
		}

		p, err := ParseFile(strings.NewReader(s))
		if err != nil {
			t.Error(err)
		}

745
		actual, err := p.CreateRequest("")
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
		if err != nil {
			t.Error(err)
		}

		if diff := cmp.Diff(actual, c.expected); diff != "" {
			t.Errorf("mismatch (-got +want):\n%s", diff)
		}
	}
}

func getSHA256Digest(t *testing.T, r io.Reader) (string, int64) {
	t.Helper()

	h := sha256.New()
	n, err := io.Copy(h, r)
	if err != nil {
		t.Fatal(err)
	}

	return fmt.Sprintf("sha256:%x", h.Sum(nil)), n
}

Michael Yang's avatar
Michael Yang committed
768
func createBinFile(t *testing.T, kv map[string]any, ti []ggml.Tensor) (string, string) {
769
770
771
772
773
774
775
776
	t.Helper()

	f, err := os.CreateTemp(t.TempDir(), "testbin.*.gguf")
	if err != nil {
		t.Fatal(err)
	}
	defer f.Close()

Michael Yang's avatar
Michael Yang committed
777
	if err := ggml.WriteGGUF(f, kv, ti); err != nil {
778
779
780
781
782
783
784
785
786
787
788
789
790
		t.Fatal(err)
	}
	// Calculate sha256 of file
	if _, err := f.Seek(0, 0); err != nil {
		t.Fatal(err)
	}

	digest, _ := getSHA256Digest(t, f)

	return f.Name(), digest
}

func TestCreateRequestFiles(t *testing.T) {
791
792
	n1, d1 := createBinFile(t, nil, nil)
	n2, d2 := createBinFile(t, map[string]any{"foo": "bar"}, nil)
793
794
795
796
797
798

	cases := []struct {
		input    string
		expected *api.CreateRequest
	}{
		{
799
800
801
802
803
804
			fmt.Sprintf("FROM %s", n1),
			&api.CreateRequest{Files: map[string]string{n1: d1}},
		},
		{
			fmt.Sprintf("FROM %s\nFROM %s", n1, n2),
			&api.CreateRequest{Files: map[string]string{n1: d1, n2: d2}},
805
806
807
808
809
810
811
812
813
814
815
816
817
818
		},
	}

	for _, c := range cases {
		s, err := unicode.UTF8.NewEncoder().String(c.input)
		if err != nil {
			t.Fatal(err)
		}

		p, err := ParseFile(strings.NewReader(s))
		if err != nil {
			t.Error(err)
		}

819
		actual, err := p.CreateRequest("")
820
821
822
823
824
825
826
827
828
		if err != nil {
			t.Error(err)
		}

		if diff := cmp.Diff(actual, c.expected); diff != "" {
			t.Errorf("mismatch (-got +want):\n%s", diff)
		}
	}
}