[router] remove old/oudated/useless comments across code base (#10968)
This commit is contained in:
@@ -279,11 +279,9 @@ mod tests {
|
||||
|
||||
#[test]
|
||||
fn test_create_tiktoken_tokenizer() {
|
||||
// Test creating tokenizer for GPT models
|
||||
let tokenizer = create_tokenizer("gpt-4").unwrap();
|
||||
assert!(tokenizer.vocab_size() > 0);
|
||||
|
||||
// Test encoding and decoding
|
||||
let text = "Hello, world!";
|
||||
let encoding = tokenizer.encode(text).unwrap();
|
||||
let decoded = tokenizer.decode(encoding.token_ids(), false).unwrap();
|
||||
@@ -292,7 +290,6 @@ mod tests {
|
||||
|
||||
#[tokio::test]
|
||||
async fn test_download_tokenizer_from_hf() {
|
||||
// Test with a small model that should have tokenizer files
|
||||
// Skip this test if HF_TOKEN is not set and we're in CI
|
||||
if std::env::var("CI").is_ok() && std::env::var("HF_TOKEN").is_err() {
|
||||
println!("Skipping HF download test in CI without HF_TOKEN");
|
||||
|
||||
@@ -206,7 +206,6 @@ mod tests {
|
||||
// The incremental text should be " world" (with the space that the mock tokenizer adds)
|
||||
assert_eq!(text2, " world");
|
||||
|
||||
// Verify the full text
|
||||
assert_eq!(seq.text().unwrap(), "Hello world");
|
||||
}
|
||||
|
||||
|
||||
@@ -398,7 +398,6 @@ mod tests {
|
||||
// The fix ensures we only output NEW text, not accumulated text
|
||||
assert_eq!(outputs.len(), 3);
|
||||
|
||||
// Verify no text is repeated
|
||||
for i in 0..outputs.len() {
|
||||
for j in i + 1..outputs.len() {
|
||||
// No output should contain another (no accumulation)
|
||||
|
||||
@@ -36,22 +36,17 @@ fn test_tokenizer_wrapper() {
|
||||
let mock_tokenizer = Arc::new(mock::MockTokenizer::new());
|
||||
let tokenizer = Tokenizer::from_arc(mock_tokenizer);
|
||||
|
||||
// Test encoding
|
||||
let encoding = tokenizer.encode("Hello world").unwrap();
|
||||
assert_eq!(encoding.token_ids(), &[1, 2]);
|
||||
|
||||
// Test decoding
|
||||
let text = tokenizer.decode(&[1, 2], false).unwrap();
|
||||
assert_eq!(text, "Hello world");
|
||||
|
||||
// Test vocab size
|
||||
assert_eq!(tokenizer.vocab_size(), 8);
|
||||
|
||||
// Test token to ID
|
||||
assert_eq!(tokenizer.token_to_id("Hello"), Some(1));
|
||||
assert_eq!(tokenizer.token_to_id("unknown"), None);
|
||||
|
||||
// Test ID to token
|
||||
assert_eq!(tokenizer.id_to_token(1), Some("Hello".to_string()));
|
||||
assert_eq!(tokenizer.id_to_token(9999), None);
|
||||
}
|
||||
|
||||
@@ -246,7 +246,6 @@ mod tests {
|
||||
|
||||
#[test]
|
||||
fn test_unrecognized_model_name_returns_error() {
|
||||
// Test that unrecognized model names return an error
|
||||
let result = TiktokenTokenizer::from_model_name("distilgpt-2");
|
||||
assert!(result.is_err());
|
||||
if let Err(e) = result {
|
||||
@@ -268,7 +267,6 @@ mod tests {
|
||||
|
||||
#[test]
|
||||
fn test_recognized_model_names() {
|
||||
// Test that recognized model names work correctly
|
||||
assert!(TiktokenTokenizer::from_model_name("gpt-4").is_ok());
|
||||
assert!(TiktokenTokenizer::from_model_name("gpt-3.5-turbo").is_ok());
|
||||
assert!(TiktokenTokenizer::from_model_name("text-davinci-003").is_ok());
|
||||
|
||||
Reference in New Issue
Block a user