convert_test.go 2.7 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132
  1. package convert
  2. import (
  3. "crypto/sha256"
  4. "encoding/hex"
  5. "encoding/json"
  6. "flag"
  7. "fmt"
  8. "io"
  9. "io/fs"
  10. "log/slog"
  11. "math"
  12. "os"
  13. "path/filepath"
  14. "slices"
  15. "testing"
  16. "golang.org/x/exp/maps"
  17. "github.com/ollama/ollama/llm"
  18. )
  19. func convertFull(t *testing.T, fsys fs.FS) (*os.File, llm.KV, llm.Tensors) {
  20. t.Helper()
  21. f, err := os.CreateTemp(t.TempDir(), "f16")
  22. if err != nil {
  23. t.Fatal(err)
  24. }
  25. defer f.Close()
  26. if err := Convert(fsys, f); err != nil {
  27. t.Fatal(err)
  28. }
  29. r, err := os.Open(f.Name())
  30. if err != nil {
  31. t.Fatal(err)
  32. }
  33. t.Cleanup(func() { r.Close() })
  34. m, _, err := llm.DecodeGGML(r, math.MaxInt)
  35. if err != nil {
  36. t.Fatal(err)
  37. }
  38. if _, err := r.Seek(0, io.SeekStart); err != nil {
  39. t.Fatal(err)
  40. }
  41. return r, m.KV(), m.Tensors()
  42. }
  43. func TestMain(m *testing.M) {
  44. var level slog.Level
  45. flag.TextVar(&level, "level", slog.LevelInfo, "log level")
  46. flag.Parse()
  47. slog.SetLogLoggerLevel(level)
  48. os.Exit(m.Run())
  49. }
  50. func TestConvertFull(t *testing.T) {
  51. cases := []string{
  52. "Meta-Llama-3-8B-Instruct",
  53. "Mistral-7B-Instruct-v0.2",
  54. "Mixtral-8x7B-Instruct-v0.1",
  55. "gemma-2b-it",
  56. // microsoft/Phi-3-mini-128-instruct@d548c233192db00165d842bf8edff054bb3212f8
  57. "Phi-3-mini-128k-instruct",
  58. "all-MiniLM-L6-v2",
  59. "gemma-2-9b-it",
  60. }
  61. for i := range cases {
  62. tt := cases[i]
  63. t.Run(tt, func(t *testing.T) {
  64. t.Parallel()
  65. p := filepath.Join("testdata", tt)
  66. if testing.Short() {
  67. t.Skip("skipping in short mode")
  68. } else if _, err := os.Stat(p); err != nil {
  69. t.Skipf("%s not found", p)
  70. }
  71. f, kv, tensors := convertFull(t, os.DirFS(p))
  72. actual := make(map[string]string)
  73. for k, v := range kv {
  74. if s, ok := v.(json.Marshaler); !ok {
  75. actual[k] = fmt.Sprintf("%v", v)
  76. } else {
  77. bts, err := json.Marshal(s)
  78. if err != nil {
  79. t.Fatal(err)
  80. }
  81. actual[k] = fmt.Sprintf("%x", sha256.Sum256(bts))
  82. }
  83. }
  84. for _, tensor := range tensors.Items {
  85. sha256sum := sha256.New()
  86. sr := io.NewSectionReader(f, int64(tensors.Offset+tensor.Offset), int64(tensor.Size()))
  87. if _, err := io.Copy(sha256sum, sr); err != nil {
  88. t.Fatal(err)
  89. }
  90. actual[tensor.Name] = hex.EncodeToString(sha256sum.Sum(nil))
  91. }
  92. expectFile, err := os.Open(filepath.Join("testdata", fmt.Sprintf("%s.json", tt)))
  93. if err != nil {
  94. t.Fatal(err)
  95. }
  96. var expect map[string]string
  97. if err := json.NewDecoder(expectFile).Decode(&expect); err != nil {
  98. t.Fatal(err)
  99. }
  100. keys := maps.Keys(expect)
  101. slices.Sort(keys)
  102. for _, k := range keys {
  103. if v, ok := actual[k]; !ok {
  104. t.Errorf("missing %s", k)
  105. } else if v != expect[k] {
  106. t.Errorf("unexpected %s: want %s, got %s", k, expect[k], v)
  107. }
  108. }
  109. })
  110. }
  111. }