aboutsummaryrefslogtreecommitdiff
path: root/cmd
diff options
context:
space:
mode:
authorXe Iaso <me@xeiaso.net>2023-04-18 19:11:56 -0400
committerXe Iaso <me@xeiaso.net>2023-04-18 19:12:08 -0400
commit7f8a0be340b0c6a54dcac51cbdf5ebf7ce0d8293 (patch)
tree7d9bb013d69bdf9522ad88edbf162e00924bb28f /cmd
parent4c15f8fa4ce26a8e43ff1a86c316d5f25f899a1c (diff)
downloadx-7f8a0be340b0c6a54dcac51cbdf5ebf7ce0d8293.tar.xz
x-7f8a0be340b0c6a54dcac51cbdf5ebf7ce0d8293.zip
tokenize
Signed-off-by: Xe Iaso <me@xeiaso.net>
Diffstat (limited to 'cmd')
-rw-r--r--cmd/tokenize/main.go50
1 files changed, 50 insertions, 0 deletions
diff --git a/cmd/tokenize/main.go b/cmd/tokenize/main.go
new file mode 100644
index 0000000..4e0db40
--- /dev/null
+++ b/cmd/tokenize/main.go
@@ -0,0 +1,50 @@
+package main
+
+import (
+ "encoding/json"
+ "flag"
+ "fmt"
+ "io"
+ "log"
+ "os"
+
+ gpt3encoder "github.com/samber/go-gpt-3-encoder"
+ "within.website/x/internal"
+)
+
+var (
+ decode = flag.Bool("decode", false, "if true, decode instead of encode")
+)
+
+func main() {
+ internal.HandleStartup()
+
+ enc, err := gpt3encoder.NewEncoder()
+ if err != nil {
+ log.Fatal(err)
+ }
+
+ if *decode {
+ var tokens []int
+ if err := json.NewDecoder(os.Stdin).Decode(&tokens); err != nil {
+ log.Fatal(err)
+ }
+
+ fmt.Fprintln(os.Stdout, enc.Decode(tokens))
+ return
+ }
+
+ data, err := io.ReadAll(os.Stdin)
+ if err != nil {
+ log.Fatal(err)
+ }
+
+ tokens, err := enc.Encode(string(data))
+ if err != nil {
+ log.Fatal(err)
+ }
+
+ if err := json.NewEncoder(os.Stdout).Encode(&tokens); err != nil {
+ log.Fatal(err)
+ }
+}