types.go 13.1 KB
Newer Older
Jeffrey Morgan's avatar
Jeffrey Morgan committed
1
2
package api

3
import (
Michael Yang's avatar
Michael Yang committed
4
	"encoding/json"
5
	"fmt"
Michael Yang's avatar
Michael Yang committed
6
	"math"
7
	"os"
8
	"reflect"
9
	"strconv"
10
	"strings"
11
12
	"time"
)
Michael Yang's avatar
Michael Yang committed
13

Patrick Devine's avatar
Patrick Devine committed
14
type StatusError struct {
15
16
17
	StatusCode   int
	Status       string
	ErrorMessage string `json:"error"`
Patrick Devine's avatar
Patrick Devine committed
18
19
20
}

func (e StatusError) Error() string {
21
22
23
24
25
26
27
28
29
30
	switch {
	case e.Status != "" && e.ErrorMessage != "":
		return fmt.Sprintf("%s: %s", e.Status, e.ErrorMessage)
	case e.Status != "":
		return e.Status
	case e.ErrorMessage != "":
		return e.ErrorMessage
	default:
		// this should not happen
		return "something went wrong, please see the ollama server logs for details"
Patrick Devine's avatar
Patrick Devine committed
31
32
33
	}
}

Patrick Devine's avatar
Patrick Devine committed
34
35
type ImageData []byte

36
type GenerateRequest struct {
Patrick Devine's avatar
Patrick Devine committed
37
38
39
40
41
42
43
44
45
	Model    string      `json:"model"`
	Prompt   string      `json:"prompt"`
	System   string      `json:"system"`
	Template string      `json:"template"`
	Context  []int       `json:"context,omitempty"`
	Stream   *bool       `json:"stream,omitempty"`
	Raw      bool        `json:"raw,omitempty"`
	Format   string      `json:"format"`
	Images   []ImageData `json:"images,omitempty"`
46

47
	Options map[string]interface{} `json:"options"`
48
49
}

Bruce MacDonald's avatar
Bruce MacDonald committed
50
51
52
53
54
55
56
57
58
59
type ChatRequest struct {
	Model    string    `json:"model"`
	Messages []Message `json:"messages"`
	Stream   *bool     `json:"stream,omitempty"`
	Format   string    `json:"format"`

	Options map[string]interface{} `json:"options"`
}

type Message struct {
60
61
	Role    string      `json:"role"` // one of ["system", "user", "assistant"]
	Content string      `json:"content"`
62
	Images  []ImageData `json:"images,omitempty"`
Bruce MacDonald's avatar
Bruce MacDonald committed
63
64
65
66
67
}

type ChatResponse struct {
	Model     string    `json:"model"`
	CreatedAt time.Time `json:"created_at"`
68
	Message   Message   `json:"message"`
Bruce MacDonald's avatar
Bruce MacDonald committed
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83

	Done bool `json:"done"`

	Metrics
}

type Metrics struct {
	TotalDuration      time.Duration `json:"total_duration,omitempty"`
	LoadDuration       time.Duration `json:"load_duration,omitempty"`
	PromptEvalCount    int           `json:"prompt_eval_count,omitempty"`
	PromptEvalDuration time.Duration `json:"prompt_eval_duration,omitempty"`
	EvalCount          int           `json:"eval_count,omitempty"`
	EvalDuration       time.Duration `json:"eval_duration,omitempty"`
}

84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
// Options specfied in GenerateRequest, if you add a new option here add it to the API docs also
type Options struct {
	Runner

	// Predict options used at runtime
	NumKeep          int      `json:"num_keep,omitempty"`
	Seed             int      `json:"seed,omitempty"`
	NumPredict       int      `json:"num_predict,omitempty"`
	TopK             int      `json:"top_k,omitempty"`
	TopP             float32  `json:"top_p,omitempty"`
	TFSZ             float32  `json:"tfs_z,omitempty"`
	TypicalP         float32  `json:"typical_p,omitempty"`
	RepeatLastN      int      `json:"repeat_last_n,omitempty"`
	Temperature      float32  `json:"temperature,omitempty"`
	RepeatPenalty    float32  `json:"repeat_penalty,omitempty"`
	PresencePenalty  float32  `json:"presence_penalty,omitempty"`
	FrequencyPenalty float32  `json:"frequency_penalty,omitempty"`
	Mirostat         int      `json:"mirostat,omitempty"`
	MirostatTau      float32  `json:"mirostat_tau,omitempty"`
	MirostatEta      float32  `json:"mirostat_eta,omitempty"`
	PenalizeNewline  bool     `json:"penalize_newline,omitempty"`
	Stop             []string `json:"stop,omitempty"`
K0IN's avatar
K0IN committed
106
	Cache            bool     `json:"cache,omitempty"`
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
}

// Runner options which must be set when the model is loaded into memory
type Runner struct {
	UseNUMA            bool    `json:"numa,omitempty"`
	NumCtx             int     `json:"num_ctx,omitempty"`
	NumBatch           int     `json:"num_batch,omitempty"`
	NumGQA             int     `json:"num_gqa,omitempty"`
	NumGPU             int     `json:"num_gpu,omitempty"`
	MainGPU            int     `json:"main_gpu,omitempty"`
	LowVRAM            bool    `json:"low_vram,omitempty"`
	F16KV              bool    `json:"f16_kv,omitempty"`
	LogitsAll          bool    `json:"logits_all,omitempty"`
	VocabOnly          bool    `json:"vocab_only,omitempty"`
	UseMMap            bool    `json:"use_mmap,omitempty"`
	UseMLock           bool    `json:"use_mlock,omitempty"`
	EmbeddingOnly      bool    `json:"embedding_only,omitempty"`
	RopeFrequencyBase  float32 `json:"rope_frequency_base,omitempty"`
	RopeFrequencyScale float32 `json:"rope_frequency_scale,omitempty"`
	NumThread          int     `json:"num_thread,omitempty"`
}

Bruce MacDonald's avatar
Bruce MacDonald committed
129
130
131
132
133
134
135
136
137
138
139
type EmbeddingRequest struct {
	Model  string `json:"model"`
	Prompt string `json:"prompt"`

	Options map[string]interface{} `json:"options"`
}

type EmbeddingResponse struct {
	Embedding []float64 `json:"embedding"`
}

140
type CreateRequest struct {
Michael Yang's avatar
Michael Yang committed
141
142
143
144
	Name      string `json:"name"`
	Path      string `json:"path"`
	Modelfile string `json:"modelfile"`
	Stream    *bool  `json:"stream,omitempty"`
145
146
}

147
148
149
150
type DeleteRequest struct {
	Name string `json:"name"`
}

Patrick Devine's avatar
Patrick Devine committed
151
152
153
154
155
type ShowRequest struct {
	Name string `json:"name"`
}

type ShowResponse struct {
Patrick Devine's avatar
Patrick Devine committed
156
157
158
159
160
161
	License    string       `json:"license,omitempty"`
	Modelfile  string       `json:"modelfile,omitempty"`
	Parameters string       `json:"parameters,omitempty"`
	Template   string       `json:"template,omitempty"`
	System     string       `json:"system,omitempty"`
	Details    ModelDetails `json:"details,omitempty"`
Patrick Devine's avatar
Patrick Devine committed
162
163
}

Patrick Devine's avatar
Patrick Devine committed
164
165
166
167
168
type CopyRequest struct {
	Source      string `json:"source"`
	Destination string `json:"destination"`
}

Bruce MacDonald's avatar
Bruce MacDonald committed
169
type PullRequest struct {
170
	Name     string `json:"name"`
171
	Insecure bool   `json:"insecure,omitempty"`
172
173
174
	Username string `json:"username"`
	Password string `json:"password"`
	Stream   *bool  `json:"stream,omitempty"`
Bruce MacDonald's avatar
Bruce MacDonald committed
175
176
}

177
type ProgressResponse struct {
178
179
	Status    string `json:"status"`
	Digest    string `json:"digest,omitempty"`
Michael Yang's avatar
Michael Yang committed
180
181
	Total     int64  `json:"total,omitempty"`
	Completed int64  `json:"completed,omitempty"`
Bruce MacDonald's avatar
Bruce MacDonald committed
182
183
}

184
185
type PushRequest struct {
	Name     string `json:"name"`
186
	Insecure bool   `json:"insecure,omitempty"`
187
188
189
	Username string `json:"username"`
	Password string `json:"password"`
	Stream   *bool  `json:"stream,omitempty"`
190
}
191

Patrick Devine's avatar
Patrick Devine committed
192
type ListResponse struct {
193
	Models []ModelResponse `json:"models"`
Patrick Devine's avatar
Patrick Devine committed
194
195
}

196
type ModelResponse struct {
Patrick Devine's avatar
Patrick Devine committed
197
198
199
200
201
	Name       string       `json:"name"`
	ModifiedAt time.Time    `json:"modified_at"`
	Size       int64        `json:"size"`
	Digest     string       `json:"digest"`
	Details    ModelDetails `json:"details,omitempty"`
Patrick Devine's avatar
Patrick Devine committed
202
203
}

Patrick Devine's avatar
Patrick Devine committed
204
205
206
207
type TokenResponse struct {
	Token string `json:"token"`
}

Michael Yang's avatar
Michael Yang committed
208
type GenerateResponse struct {
209
210
	Model     string    `json:"model"`
	CreatedAt time.Time `json:"created_at"`
211
	Response  string    `json:"response"`
212

Michael Yang's avatar
Michael Yang committed
213
214
	Done    bool  `json:"done"`
	Context []int `json:"context,omitempty"`
215

Bruce MacDonald's avatar
Bruce MacDonald committed
216
	Metrics
217
218
}

Patrick Devine's avatar
Patrick Devine committed
219
220
221
222
223
224
225
226
type ModelDetails struct {
	Format            string   `json:"format"`
	Family            string   `json:"family"`
	Families          []string `json:"families"`
	ParameterSize     string   `json:"parameter_size"`
	QuantizationLevel string   `json:"quantization_level"`
}

Bruce MacDonald's avatar
Bruce MacDonald committed
227
228
229
func (m *Metrics) Summary() {
	if m.TotalDuration > 0 {
		fmt.Fprintf(os.Stderr, "total duration:       %v\n", m.TotalDuration)
230
231
	}

Bruce MacDonald's avatar
Bruce MacDonald committed
232
233
	if m.LoadDuration > 0 {
		fmt.Fprintf(os.Stderr, "load duration:        %v\n", m.LoadDuration)
Michael Yang's avatar
Michael Yang committed
234
235
	}

Bruce MacDonald's avatar
Bruce MacDonald committed
236
237
	if m.PromptEvalCount > 0 {
		fmt.Fprintf(os.Stderr, "prompt eval count:    %d token(s)\n", m.PromptEvalCount)
238
239
	}

Bruce MacDonald's avatar
Bruce MacDonald committed
240
241
242
	if m.PromptEvalDuration > 0 {
		fmt.Fprintf(os.Stderr, "prompt eval duration: %s\n", m.PromptEvalDuration)
		fmt.Fprintf(os.Stderr, "prompt eval rate:     %.2f tokens/s\n", float64(m.PromptEvalCount)/m.PromptEvalDuration.Seconds())
243
244
	}

Bruce MacDonald's avatar
Bruce MacDonald committed
245
246
	if m.EvalCount > 0 {
		fmt.Fprintf(os.Stderr, "eval count:           %d token(s)\n", m.EvalCount)
247
248
	}

Bruce MacDonald's avatar
Bruce MacDonald committed
249
250
251
	if m.EvalDuration > 0 {
		fmt.Fprintf(os.Stderr, "eval duration:        %s\n", m.EvalDuration)
		fmt.Fprintf(os.Stderr, "eval rate:            %.2f tokens/s\n", float64(m.EvalCount)/m.EvalDuration.Seconds())
252
	}
253
254
}

255
256
var ErrInvalidOpts = fmt.Errorf("invalid options")

257
258
259
260
261
262
263
264
265
266
267
268
269
func (opts *Options) FromMap(m map[string]interface{}) error {
	valueOpts := reflect.ValueOf(opts).Elem() // names of the fields in the options struct
	typeOpts := reflect.TypeOf(opts).Elem()   // types of the fields in the options struct

	// build map of json struct tags to their types
	jsonOpts := make(map[string]reflect.StructField)
	for _, field := range reflect.VisibleFields(typeOpts) {
		jsonTag := strings.Split(field.Tag.Get("json"), ",")[0]
		if jsonTag != "" {
			jsonOpts[jsonTag] = field
		}
	}

270
	invalidOpts := []string{}
271
272
273
274
	for key, val := range m {
		if opt, ok := jsonOpts[key]; ok {
			field := valueOpts.FieldByName(opt.Name)
			if field.IsValid() && field.CanSet() {
Michael Yang's avatar
Michael Yang committed
275
276
277
278
				if val == nil {
					continue
				}

279
280
				switch field.Kind() {
				case reflect.Int:
Michael Yang's avatar
Michael Yang committed
281
282
283
284
285
286
287
					switch t := val.(type) {
					case int64:
						field.SetInt(t)
					case float64:
						// when JSON unmarshals numbers, it uses float64, not int
						field.SetInt(int64(t))
					default:
288
						return fmt.Errorf("option %q must be of type integer", key)
289
290
291
292
					}
				case reflect.Bool:
					val, ok := val.(bool)
					if !ok {
293
						return fmt.Errorf("option %q must be of type boolean", key)
294
295
296
297
298
299
					}
					field.SetBool(val)
				case reflect.Float32:
					// JSON unmarshals to float64
					val, ok := val.(float64)
					if !ok {
300
						return fmt.Errorf("option %q must be of type float32", key)
301
302
303
304
305
					}
					field.SetFloat(val)
				case reflect.String:
					val, ok := val.(string)
					if !ok {
306
						return fmt.Errorf("option %q must be of type string", key)
307
308
309
310
311
312
					}
					field.SetString(val)
				case reflect.Slice:
					// JSON unmarshals to []interface{}, not []string
					val, ok := val.([]interface{})
					if !ok {
313
						return fmt.Errorf("option %q must be of type array", key)
314
315
316
317
318
319
					}
					// convert []interface{} to []string
					slice := make([]string, len(val))
					for i, item := range val {
						str, ok := item.(string)
						if !ok {
320
							return fmt.Errorf("option %q must be of an array of strings", key)
321
322
323
324
325
326
327
328
						}
						slice[i] = str
					}
					field.Set(reflect.ValueOf(slice))
				default:
					return fmt.Errorf("unknown type loading config params: %v", field.Kind())
				}
			}
329
330
		} else {
			invalidOpts = append(invalidOpts, key)
331
332
		}
	}
333
334
335
336

	if len(invalidOpts) > 0 {
		return fmt.Errorf("%w: %v", ErrInvalidOpts, strings.Join(invalidOpts, ", "))
	}
337
338
339
	return nil
}

Michael Yang's avatar
Michael Yang committed
340
341
func DefaultOptions() Options {
	return Options{
342
343
		// options set on request to runner
		NumPredict:       -1,
344
		NumKeep:          0,
Michael Yang's avatar
Michael Yang committed
345
346
347
348
349
		Temperature:      0.8,
		TopK:             40,
		TopP:             0.9,
		TFSZ:             1.0,
		TypicalP:         1.0,
350
351
352
353
		RepeatLastN:      64,
		RepeatPenalty:    1.1,
		PresencePenalty:  0.0,
		FrequencyPenalty: 0.0,
Michael Yang's avatar
Michael Yang committed
354
355
356
		Mirostat:         0,
		MirostatTau:      5.0,
		MirostatEta:      0.1,
Michael Yang's avatar
Michael Yang committed
357
		PenalizeNewline:  true,
358
		Seed:             -1,
K0IN's avatar
K0IN committed
359
		Cache:            true,
Michael Yang's avatar
Michael Yang committed
360

361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
		Runner: Runner{
			// options set when the model is loaded
			NumCtx:             2048,
			RopeFrequencyBase:  10000.0,
			RopeFrequencyScale: 1.0,
			NumBatch:           512,
			NumGPU:             -1, // -1 here indicates that NumGPU should be set dynamically
			NumGQA:             1,
			NumThread:          0, // let the runtime decide
			LowVRAM:            false,
			F16KV:              true,
			UseMLock:           false,
			UseMMap:            true,
			UseNUMA:            false,
			EmbeddingOnly:      true,
		},
Michael Yang's avatar
Michael Yang committed
377
	}
Jeffrey Morgan's avatar
Jeffrey Morgan committed
378
}
Michael Yang's avatar
Michael Yang committed
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

type Duration struct {
	time.Duration
}

func (d *Duration) UnmarshalJSON(b []byte) (err error) {
	var v any
	if err := json.Unmarshal(b, &v); err != nil {
		return err
	}

	d.Duration = 5 * time.Minute

	switch t := v.(type) {
	case float64:
		if t < 0 {
			t = math.MaxFloat64
		}

		d.Duration = time.Duration(t)
	case string:
		d.Duration, err = time.ParseDuration(t)
		if err != nil {
			return err
		}
	}

	return nil
}
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467

// FormatParams converts specified parameter options to their correct types
func FormatParams(params map[string][]string) (map[string]interface{}, error) {
	opts := Options{}
	valueOpts := reflect.ValueOf(&opts).Elem() // names of the fields in the options struct
	typeOpts := reflect.TypeOf(opts)           // types of the fields in the options struct

	// build map of json struct tags to their types
	jsonOpts := make(map[string]reflect.StructField)
	for _, field := range reflect.VisibleFields(typeOpts) {
		jsonTag := strings.Split(field.Tag.Get("json"), ",")[0]
		if jsonTag != "" {
			jsonOpts[jsonTag] = field
		}
	}

	out := make(map[string]interface{})
	// iterate params and set values based on json struct tags
	for key, vals := range params {
		if opt, ok := jsonOpts[key]; !ok {
			return nil, fmt.Errorf("unknown parameter '%s'", key)
		} else {
			field := valueOpts.FieldByName(opt.Name)
			if field.IsValid() && field.CanSet() {
				switch field.Kind() {
				case reflect.Float32:
					floatVal, err := strconv.ParseFloat(vals[0], 32)
					if err != nil {
						return nil, fmt.Errorf("invalid float value %s", vals)
					}

					out[key] = float32(floatVal)
				case reflect.Int:
					intVal, err := strconv.ParseInt(vals[0], 10, 64)
					if err != nil {
						return nil, fmt.Errorf("invalid int value %s", vals)
					}

					out[key] = intVal
				case reflect.Bool:
					boolVal, err := strconv.ParseBool(vals[0])
					if err != nil {
						return nil, fmt.Errorf("invalid bool value %s", vals)
					}

					out[key] = boolVal
				case reflect.String:
					out[key] = vals[0]
				case reflect.Slice:
					// TODO: only string slices are supported right now
					out[key] = vals
				default:
					return nil, fmt.Errorf("unknown type %s for %s", field.Kind(), key)
				}
			}
		}
	}

	return out, nil
}