ollama/api/types_test.go

486 lines
9.8 KiB
Go
Raw Normal View History

package api
import (
"encoding/json"
2024-08-02 05:52:15 +08:00
"errors"
"math"
"testing"
"time"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
)
func TestKeepAliveParsingFromJSON(t *testing.T) {
tests := []struct {
name string
req string
exp *Duration
}{
{
name: "Positive Integer",
req: `{ "keep_alive": 42 }`,
exp: &Duration{42 * time.Second},
},
{
name: "Positive Float",
req: `{ "keep_alive": 42.5 }`,
exp: &Duration{42 * time.Second},
},
{
name: "Positive Integer String",
req: `{ "keep_alive": "42m" }`,
exp: &Duration{42 * time.Minute},
},
{
name: "Negative Integer",
req: `{ "keep_alive": -1 }`,
exp: &Duration{math.MaxInt64},
},
{
name: "Negative Float",
req: `{ "keep_alive": -3.14 }`,
exp: &Duration{math.MaxInt64},
},
{
name: "Negative Integer String",
req: `{ "keep_alive": "-1m" }`,
exp: &Duration{math.MaxInt64},
},
}
for _, test := range tests {
t.Run(test.name, func(t *testing.T) {
var dec ChatRequest
err := json.Unmarshal([]byte(test.req), &dec)
require.NoError(t, err)
assert.Equal(t, test.exp, dec.KeepAlive)
})
}
}
func TestDurationMarshalUnmarshal(t *testing.T) {
tests := []struct {
name string
input time.Duration
expected time.Duration
}{
{
"negative duration",
time.Duration(-1),
time.Duration(math.MaxInt64),
},
{
"positive duration",
2024-05-22 13:21:04 +08:00
42 * time.Second,
42 * time.Second,
},
{
"another positive duration",
2024-05-22 13:21:04 +08:00
42 * time.Minute,
42 * time.Minute,
},
{
"zero duration",
time.Duration(0),
time.Duration(0),
},
{
"max duration",
time.Duration(math.MaxInt64),
time.Duration(math.MaxInt64),
},
}
for _, test := range tests {
t.Run(test.name, func(t *testing.T) {
b, err := json.Marshal(Duration{test.input})
require.NoError(t, err)
var d Duration
err = json.Unmarshal(b, &d)
require.NoError(t, err)
assert.Equal(t, test.expected, d.Duration, "input %v, marshalled %v, got %v", test.input, string(b), d.Duration)
})
}
}
func TestUseMmapParsingFromJSON(t *testing.T) {
tr := true
fa := false
tests := []struct {
name string
req string
exp *bool
}{
{
name: "Undefined",
req: `{ }`,
exp: nil,
},
{
name: "True",
req: `{ "use_mmap": true }`,
exp: &tr,
},
{
name: "False",
req: `{ "use_mmap": false }`,
exp: &fa,
},
}
for _, test := range tests {
t.Run(test.name, func(t *testing.T) {
var oMap map[string]any
err := json.Unmarshal([]byte(test.req), &oMap)
require.NoError(t, err)
opts := DefaultOptions()
err = opts.FromMap(oMap)
require.NoError(t, err)
assert.Equal(t, test.exp, opts.UseMMap)
})
}
}
func TestUseMmapFormatParams(t *testing.T) {
tr := true
fa := false
tests := []struct {
name string
req map[string][]string
exp *bool
err error
}{
{
name: "True",
req: map[string][]string{
"use_mmap": {"true"},
},
exp: &tr,
err: nil,
},
{
name: "False",
req: map[string][]string{
"use_mmap": {"false"},
},
exp: &fa,
err: nil,
},
{
name: "Numeric True",
req: map[string][]string{
"use_mmap": {"1"},
},
exp: &tr,
err: nil,
},
{
name: "Numeric False",
req: map[string][]string{
"use_mmap": {"0"},
},
exp: &fa,
err: nil,
},
{
name: "invalid string",
req: map[string][]string{
"use_mmap": {"foo"},
},
exp: nil,
2024-08-02 05:52:15 +08:00
err: errors.New("invalid bool value [foo]"),
},
}
for _, test := range tests {
t.Run(test.name, func(t *testing.T) {
resp, err := FormatParams(test.req)
require.Equal(t, test.err, err)
respVal, ok := resp["use_mmap"]
if test.exp != nil {
assert.True(t, ok, "resp: %v", resp)
assert.Equal(t, *test.exp, *respVal.(*bool))
}
})
}
}
func TestMessage_UnmarshalJSON(t *testing.T) {
tests := []struct {
input string
expected string
}{
{`{"role": "USER", "content": "Hello!"}`, "user"},
{`{"role": "System", "content": "Initialization complete."}`, "system"},
{`{"role": "assistant", "content": "How can I help you?"}`, "assistant"},
{`{"role": "TOOl", "content": "Access granted."}`, "tool"},
}
for _, test := range tests {
var msg Message
if err := json.Unmarshal([]byte(test.input), &msg); err != nil {
t.Errorf("Unexpected error: %v", err)
}
if msg.Role != test.expected {
t.Errorf("role not lowercased: got %v, expected %v", msg.Role, test.expected)
}
}
}
func TestToolFunction_UnmarshalJSON(t *testing.T) {
tests := []struct {
name string
input string
wantErr string
}{
{
name: "valid enum with same types",
input: `{
"name": "test",
"description": "test function",
"parameters": {
"type": "object",
"required": ["test"],
"properties": {
"test": {
"type": "string",
"description": "test prop",
"enum": ["a", "b", "c"]
}
}
}
}`,
wantErr: "",
},
{
name: "empty enum array",
input: `{
"name": "test",
"description": "test function",
"parameters": {
"type": "object",
"required": ["test"],
"properties": {
"test": {
"type": "string",
"description": "test prop",
"enum": []
}
}
}
}`,
wantErr: "",
},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
var tf ToolFunction
err := json.Unmarshal([]byte(tt.input), &tf)
if tt.wantErr != "" {
require.Error(t, err)
assert.Contains(t, err.Error(), tt.wantErr)
} else {
require.NoError(t, err)
}
})
}
}
func TestPropertyType_UnmarshalJSON(t *testing.T) {
tests := []struct {
name string
input string
expected PropertyType
}{
{
name: "string type",
input: `"string"`,
expected: PropertyType{"string"},
},
{
name: "array of types",
input: `["string", "number"]`,
expected: PropertyType{"string", "number"},
},
{
name: "array with single type",
input: `["string"]`,
expected: PropertyType{"string"},
},
}
for _, test := range tests {
t.Run(test.name, func(t *testing.T) {
var pt PropertyType
if err := json.Unmarshal([]byte(test.input), &pt); err != nil {
t.Errorf("Unexpected error: %v", err)
}
if len(pt) != len(test.expected) {
t.Errorf("Length mismatch: got %v, expected %v", len(pt), len(test.expected))
}
for i, v := range pt {
if v != test.expected[i] {
t.Errorf("Value mismatch at index %d: got %v, expected %v", i, v, test.expected[i])
}
}
})
}
}
func TestPropertyType_MarshalJSON(t *testing.T) {
tests := []struct {
name string
input PropertyType
expected string
}{
{
name: "single type",
input: PropertyType{"string"},
expected: `"string"`,
},
{
name: "multiple types",
input: PropertyType{"string", "number"},
expected: `["string","number"]`,
},
{
name: "empty type",
input: PropertyType{},
expected: `[]`,
},
}
for _, test := range tests {
t.Run(test.name, func(t *testing.T) {
data, err := json.Marshal(test.input)
if err != nil {
t.Errorf("Unexpected error: %v", err)
}
if string(data) != test.expected {
t.Errorf("Marshaled data mismatch: got %v, expected %v", string(data), test.expected)
}
})
}
}
func TestThinking_UnmarshalJSON(t *testing.T) {
tests := []struct {
name string
input string
gpt-oss (#11672) * bf16 * tests * gpt-oss * enable gptoss for engine * rough estimate * convert to mxfp4 * handle safetensors U8 * clamp glu/linear * update tokenizer * MXFP4 support This implements the Open Compute Microscaling (MX) FP4 format as a tensor type with backend implementations focusing on mulmat and mulmatid on CPU, CUDA, and Metal. * Unit tests for MXFP4 support This exercises various operations and shapes on both CPU and GPU (if detected on the system) * cuda graph * unit test adjustments * cuda: optimize memory access Read 4 bytes at a time (8 elements) when performing mul_mat_vec_mxfp4 * mac: fix crash on old macos versions cblas_sgemm is only supported on v13.3 and up, however bf16 is only supported on v14+ so we were falling back to ggml-blas and crashing on bf16 tensors. Checking for the function being null seems to be the simplest way to condittionally avoid registering the backend. * server: Minimum context length for gptoss This model requires a minimum context length of 8192 to function effectively. Users can set higher values through all normal mechanisms but lower values will be silently reset. * ggml: Multiply by numParallel for gptoss sliding window When computing the graph size estimate, the context size is already multiplied by numParallel so estimates reflect that. However, since sliding window models use a smaller, fixed context size, they need to manually take numParallel into account. * gpt-oss integration includes harmony parser and thinking levels, etc. * fix sync * fix tests * fix lint --------- Co-authored-by: Daniel Hiltgen <daniel@ollama.com> Co-authored-by: Jesse Gross <jesse@ollama.com> Co-authored-by: Devon Rifkin <drifkin@drifkin.net>
2025-08-06 03:21:16 +08:00
expectedThinking *ThinkValue
expectedError bool
}{
{
name: "true",
input: `{ "think": true }`,
gpt-oss (#11672) * bf16 * tests * gpt-oss * enable gptoss for engine * rough estimate * convert to mxfp4 * handle safetensors U8 * clamp glu/linear * update tokenizer * MXFP4 support This implements the Open Compute Microscaling (MX) FP4 format as a tensor type with backend implementations focusing on mulmat and mulmatid on CPU, CUDA, and Metal. * Unit tests for MXFP4 support This exercises various operations and shapes on both CPU and GPU (if detected on the system) * cuda graph * unit test adjustments * cuda: optimize memory access Read 4 bytes at a time (8 elements) when performing mul_mat_vec_mxfp4 * mac: fix crash on old macos versions cblas_sgemm is only supported on v13.3 and up, however bf16 is only supported on v14+ so we were falling back to ggml-blas and crashing on bf16 tensors. Checking for the function being null seems to be the simplest way to condittionally avoid registering the backend. * server: Minimum context length for gptoss This model requires a minimum context length of 8192 to function effectively. Users can set higher values through all normal mechanisms but lower values will be silently reset. * ggml: Multiply by numParallel for gptoss sliding window When computing the graph size estimate, the context size is already multiplied by numParallel so estimates reflect that. However, since sliding window models use a smaller, fixed context size, they need to manually take numParallel into account. * gpt-oss integration includes harmony parser and thinking levels, etc. * fix sync * fix tests * fix lint --------- Co-authored-by: Daniel Hiltgen <daniel@ollama.com> Co-authored-by: Jesse Gross <jesse@ollama.com> Co-authored-by: Devon Rifkin <drifkin@drifkin.net>
2025-08-06 03:21:16 +08:00
expectedThinking: &ThinkValue{Value: true},
},
{
name: "false",
input: `{ "think": false }`,
gpt-oss (#11672) * bf16 * tests * gpt-oss * enable gptoss for engine * rough estimate * convert to mxfp4 * handle safetensors U8 * clamp glu/linear * update tokenizer * MXFP4 support This implements the Open Compute Microscaling (MX) FP4 format as a tensor type with backend implementations focusing on mulmat and mulmatid on CPU, CUDA, and Metal. * Unit tests for MXFP4 support This exercises various operations and shapes on both CPU and GPU (if detected on the system) * cuda graph * unit test adjustments * cuda: optimize memory access Read 4 bytes at a time (8 elements) when performing mul_mat_vec_mxfp4 * mac: fix crash on old macos versions cblas_sgemm is only supported on v13.3 and up, however bf16 is only supported on v14+ so we were falling back to ggml-blas and crashing on bf16 tensors. Checking for the function being null seems to be the simplest way to condittionally avoid registering the backend. * server: Minimum context length for gptoss This model requires a minimum context length of 8192 to function effectively. Users can set higher values through all normal mechanisms but lower values will be silently reset. * ggml: Multiply by numParallel for gptoss sliding window When computing the graph size estimate, the context size is already multiplied by numParallel so estimates reflect that. However, since sliding window models use a smaller, fixed context size, they need to manually take numParallel into account. * gpt-oss integration includes harmony parser and thinking levels, etc. * fix sync * fix tests * fix lint --------- Co-authored-by: Daniel Hiltgen <daniel@ollama.com> Co-authored-by: Jesse Gross <jesse@ollama.com> Co-authored-by: Devon Rifkin <drifkin@drifkin.net>
2025-08-06 03:21:16 +08:00
expectedThinking: &ThinkValue{Value: false},
},
{
name: "unset",
input: `{ }`,
expectedThinking: nil,
},
{
gpt-oss (#11672) * bf16 * tests * gpt-oss * enable gptoss for engine * rough estimate * convert to mxfp4 * handle safetensors U8 * clamp glu/linear * update tokenizer * MXFP4 support This implements the Open Compute Microscaling (MX) FP4 format as a tensor type with backend implementations focusing on mulmat and mulmatid on CPU, CUDA, and Metal. * Unit tests for MXFP4 support This exercises various operations and shapes on both CPU and GPU (if detected on the system) * cuda graph * unit test adjustments * cuda: optimize memory access Read 4 bytes at a time (8 elements) when performing mul_mat_vec_mxfp4 * mac: fix crash on old macos versions cblas_sgemm is only supported on v13.3 and up, however bf16 is only supported on v14+ so we were falling back to ggml-blas and crashing on bf16 tensors. Checking for the function being null seems to be the simplest way to condittionally avoid registering the backend. * server: Minimum context length for gptoss This model requires a minimum context length of 8192 to function effectively. Users can set higher values through all normal mechanisms but lower values will be silently reset. * ggml: Multiply by numParallel for gptoss sliding window When computing the graph size estimate, the context size is already multiplied by numParallel so estimates reflect that. However, since sliding window models use a smaller, fixed context size, they need to manually take numParallel into account. * gpt-oss integration includes harmony parser and thinking levels, etc. * fix sync * fix tests * fix lint --------- Co-authored-by: Daniel Hiltgen <daniel@ollama.com> Co-authored-by: Jesse Gross <jesse@ollama.com> Co-authored-by: Devon Rifkin <drifkin@drifkin.net>
2025-08-06 03:21:16 +08:00
name: "string_high",
input: `{ "think": "high" }`,
expectedThinking: &ThinkValue{Value: "high"},
},
{
name: "string_medium",
input: `{ "think": "medium" }`,
expectedThinking: &ThinkValue{Value: "medium"},
},
{
name: "string_low",
input: `{ "think": "low" }`,
expectedThinking: &ThinkValue{Value: "low"},
},
{
name: "invalid_string",
input: `{ "think": "invalid" }`,
expectedThinking: nil,
expectedError: true,
},
}
for _, test := range tests {
t.Run(test.name, func(t *testing.T) {
var req GenerateRequest
err := json.Unmarshal([]byte(test.input), &req)
if test.expectedError {
require.Error(t, err)
} else {
require.NoError(t, err)
gpt-oss (#11672) * bf16 * tests * gpt-oss * enable gptoss for engine * rough estimate * convert to mxfp4 * handle safetensors U8 * clamp glu/linear * update tokenizer * MXFP4 support This implements the Open Compute Microscaling (MX) FP4 format as a tensor type with backend implementations focusing on mulmat and mulmatid on CPU, CUDA, and Metal. * Unit tests for MXFP4 support This exercises various operations and shapes on both CPU and GPU (if detected on the system) * cuda graph * unit test adjustments * cuda: optimize memory access Read 4 bytes at a time (8 elements) when performing mul_mat_vec_mxfp4 * mac: fix crash on old macos versions cblas_sgemm is only supported on v13.3 and up, however bf16 is only supported on v14+ so we were falling back to ggml-blas and crashing on bf16 tensors. Checking for the function being null seems to be the simplest way to condittionally avoid registering the backend. * server: Minimum context length for gptoss This model requires a minimum context length of 8192 to function effectively. Users can set higher values through all normal mechanisms but lower values will be silently reset. * ggml: Multiply by numParallel for gptoss sliding window When computing the graph size estimate, the context size is already multiplied by numParallel so estimates reflect that. However, since sliding window models use a smaller, fixed context size, they need to manually take numParallel into account. * gpt-oss integration includes harmony parser and thinking levels, etc. * fix sync * fix tests * fix lint --------- Co-authored-by: Daniel Hiltgen <daniel@ollama.com> Co-authored-by: Jesse Gross <jesse@ollama.com> Co-authored-by: Devon Rifkin <drifkin@drifkin.net>
2025-08-06 03:21:16 +08:00
if test.expectedThinking == nil {
assert.Nil(t, req.Think)
} else {
require.NotNil(t, req.Think)
assert.Equal(t, test.expectedThinking.Value, req.Think.Value)
}
}
})
}
}
func TestToolFunctionParameters_String(t *testing.T) {
tests := []struct {
name string
params ToolFunctionParameters
expected string
}{
{
name: "simple object with string property",
params: ToolFunctionParameters{
Type: "object",
Required: []string{"name"},
Properties: map[string]ToolProperty{
"name": {
Type: PropertyType{"string"},
Description: "The name of the person",
},
},
},
expected: `{"type":"object","required":["name"],"properties":{"name":{"type":"string","description":"The name of the person"}}}`,
},
{
name: "marshal failure returns empty string",
params: ToolFunctionParameters{
Type: "object",
Defs: func() any {
// Create a cycle that will cause json.Marshal to fail
type selfRef struct {
Self *selfRef
}
s := &selfRef{}
s.Self = s
return s
}(),
Properties: map[string]ToolProperty{},
},
expected: "",
},
}
for _, test := range tests {
t.Run(test.name, func(t *testing.T) {
result := test.params.String()
assert.Equal(t, test.expected, result)
})
}
}