From 123a37a04b0076c569cbed7c2669ab1087e2cba7 Mon Sep 17 00:00:00 2001 From: Jonas Knobloch Date: Wed, 9 Sep 2026 02:43:55 +0200 Subject: WIP --- tokenizer/byte/tokenizer_test.go | 49 ++++++++++++++++++++++++++++++++++++++++ 1 file changed, 49 insertions(+) create mode 100644 tokenizer/byte/tokenizer_test.go (limited to 'tokenizer/byte/tokenizer_test.go') diff --git a/tokenizer/byte/tokenizer_test.go b/tokenizer/byte/tokenizer_test.go new file mode 100644 index 0000000..d2c8c6c --- /dev/null +++ b/tokenizer/byte/tokenizer_test.go @@ -0,0 +1,49 @@ +package byte + +import ( + "fmt" + "os" + "testing" + + "go.jknobloc.com/x/shelf" +) + +func TestNewTokenizer_Encode(t *testing.T) { + var file *os.File + + if f, err := os.Open(shelf.Abs("tokenizers/minipile/tokenizer_gpt2_256_m000_minipile/vocab.json")); err != nil { + t.Fatal(err) + } else { + file = f + } + + defer file.Close() + + tok, err := NewTokenizer(file) + + if err != nil { + t.Fatal(err) + } + + fmt.Println(tok.Encode(" \nabc")) +} + +func TestNewTokenizer_Decode(t *testing.T) { + var file *os.File + + if f, err := os.Open(shelf.Abs("tokenizers/minipile/tokenizer_gpt2_256_m000_minipile/vocab.json")); err != nil { + t.Fatal(err) + } else { + file = f + } + + defer file.Close() + + tok, err := NewTokenizer(file) + + if err != nil { + t.Fatal(err) + } + + fmt.Println(tok.Decode([]uint8{220, 198, 64, 65, 66})) +} -- cgit v1.3.1