update

flexflow · Oct 15, 2024 · 2b5a023 · 2b5a023
1 parent c46ddc0
commit 2b5a023
Show file tree

Hide file tree

Showing 2 changed files with 40 additions and 8 deletions.
diff --git a/python/flexflow/serve/serve.py b/python/flexflow/serve/serve.py
@@ -350,7 +350,7 @@ def download_hf_tokenizer_if_needed(self):
                 f"'{self.model_name}' tokenizer needs updating! Downloading tokenizer now..."
             )
             # Load/download the tokenizer files
-            target_tokenizer_files = ["tokenizer.json", "tokenizer_config.json", "special_tokens_map.json"]
+            target_tokenizer_files = ["tokenizer.json", "tokenizer_config.json", "special_tokens_map.json", "vocab.json", "merges.txt"]
             if os.path.exists(self.model_name):
                 hf_tokenizer_path = self.model_name
             else:

diff --git a/src/runtime/request_manager.cc b/src/runtime/request_manager.cc
@@ -186,19 +186,27 @@ void RequestManager::register_tokenizer(ModelType type,
   std::filesystem::path tokenizer_folder(path);
 
   if (model_type == ModelType::LLAMA) {
+    // try with tokenizer.json first
     std::filesystem::path tokenizer_json_path;
     if (std::filesystem::is_directory(tokenizer_folder)) {
-      tokenizer_json_path =
-          std::filesystem::path(tokenizer_folder) / "tokenizer.json";
+      tokenizer_json_path = std::filesystem::path(tokenizer_folder) / "tokenizer.json";
     } else {
       tokenizer_json_path = tokenizer_folder;
     }
-    if (!std::filesystem::exists(tokenizer_json_path)) {
-      std::cerr << "Failed to open file: " << tokenizer_json_path << std::endl;
-      assert(false);
+    if (std::filesystem::exists(tokenizer_json_path)) {
+      // load from tokenizer.json
+      this->tokenizer_ = Tokenizer::FromBlobJSON(LoadBytesFromFile(tokenizer_json_path.string()));
+    } else {
+      // load from tokenizer.model
+      std::filesystem::path tokenizer_model_path =
+          tokenizer_folder / "tokenizer.model";
+      if (!std::filesystem::exists(tokenizer_model_path)) {
+        std::cerr << "Failed to open file: " << tokenizer_model_path
+                  << std::endl;
+        assert(false);
+      }
+      this->tokenizer_ = Tokenizer::FromBlobSentencePiece(LoadBytesFromFile(tokenizer_model_path.string()));
     }
-    this->tokenizer_ = Tokenizer::FromBlobJSON(
-        LoadBytesFromFile(tokenizer_json_path.string()));
   } else if (model_type == ModelType::OPT) {
     std::filesystem::path vocab_file = tokenizer_folder / "vocab.json";
     std::filesystem::path merges_file = tokenizer_folder / "merges.txt";
@@ -648,6 +656,12 @@ BatchConfig RequestManager::prepare_next_batch(BatchConfig const &old_bc,
       bool request_completed = check_inf_req_completion(old_bc, i);
       if (request_completed) {
         std::string output = this->tokenizer_->Decode(request.tokens);
+        // Unlike Huggingface, the sentencepiece C++ library automatically
+        // removes the BOS token
+        if (model_type == ModelType::LLAMA &&
+            request.tokens.at(0) == bos_token_id) {
+          output = "<s> " + output;
+        }
         {
           // update generation result
           GenerationResult &gr = request_generation_results[request.guid];
@@ -1103,6 +1117,12 @@ BeamSearchBatchConfig
                           request.guid,
                           request.tokens.size());
         std::string output = this->tokenizer_->Decode(request.tokens);
+        // Unlike Huggingface, the sentencepiece C++ library automatically
+        // removes the BOS token
+        if (model_type == ModelType::LLAMA &&
+            request.tokens.at(0) == bos_token_id) {
+          output = "<s> " + output;
+        }
         {
           // update generation result
           GenerationResult &gr = request_generation_results[request.guid];
@@ -1240,6 +1260,12 @@ BeamSearchBatchConfig
         }
 
         std::string output = this->tokenizer_->Decode(request.tokens);
+        // Unlike Huggingface, the sentencepiece C++ library automatically
+        // removes the BOS token
+        if (model_type == ModelType::LLAMA &&
+            request.tokens.at(0) == bos_token_id) {
+          output = "<s> " + output;
+        }
         log_req_mgr.print("Output: %s", output.c_str());
       }
 
@@ -1282,6 +1308,12 @@ BeamSearchBatchConfig
 
       // Token Info
       std::string output = this->tokenizer_->Decode(request.tokens);
+      // Unlike Huggingface, the sentencepiece C++ library automatically removes
+      // the BOS token
+      if (model_type == ModelType::LLAMA &&
+          request.tokens.at(0) == bos_token_id) {
+        output = "<s> " + output;
+      }
       log_req_mgr.print("Output: %s", output.c_str());
     } else {
       assert(false);