feat: TRTLLM API handle tokenizers without pad_id (e.g., tiktoken) (N…

…VIDIA#399)

Signed-off-by: Terry Kong <terryk@nvidia.com>
Signed-off-by: NeMo-Aligner CI <nemo-aligner-ci@nvidia.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Signed-off-by: abukharin <abukharin@nvidia.com>

nemo_aligner/utils/trt_llm.py

@@ -44,8 +44,9 @@ def append_and_repad_list(list_of_items, item_to_append, pad_id):
 
 
 class GPTGenerateTRTLLM:
-    # If a tokenizer does not have a pad_id, we use a large negative number and replace
-    # with self.eos_id after generation.
+    # Use a reserved negative number since there is variation between tokenizers if
+    #  they (1) have a pad_id (2) don't have a pad_id or (3) have None as the pad_id.
+    #  This pad_id is replaced with eos_id after generation.
     DEFAULT_PAD_ID = -42
 
     def __init__(
@@ -72,12 +73,6 @@ def __init__(
                 "You are trying to use NeMo-Aligner's TensorRT-LLM acceleration for LLM generation. Please build the dockerfile to enable this feature: https://github.com/NVIDIA/NeMo-Aligner/blob/main/Dockerfile"
             )
 
-        # If this assert turns out to be a blocker with some tokenizers, potential workarounds could be to:
-        #   - add a config option to allow specifying which token we pass as `end_id` to TRT-LLM (should
-        #     be a token that the model is guaranteed to never generate)
-        assert (
-            tokenizer.pad_id != tokenizer.eos_id
-        ), f"We require tokenizers to have a different {tokenizer.pad_id=} than {tokenizer.eos_id=} when using TRT-LLM. This is to make sure all code goes into the same path and include the eos_id when the response lengths are computed"
         assert max_input_len > 0
         assert max_generation_length > 0
         assert (
@@ -104,7 +99,7 @@ def __init__(
         rng_generator.manual_seed(seed)
         self.rng_generator = rng_generator
 
-        self.pad_id = tokenizer.pad_id if tokenizer.pad_id is not None else GPTGenerateTRTLLM.DEFAULT_PAD_ID
+        self.pad_id = GPTGenerateTRTLLM.DEFAULT_PAD_ID
         self.eos_id = tokenizer.eos_id
         end_strings = list(end_strings)