Fix missing attention_mask in calibration dataloader (#1261)

## Summary
- When `include_labels=False` (the default for PTQ calibration),
`get_dataset_dataloader` was discarding the `attention_mask` produced by
the tokenizer and only returning `input_ids`.
- Without `attention_mask`, HuggingFace models create a full causal
mask, causing padding tokens to participate in attention during
calibration and skewing quantization statistics.
- This fix includes `attention_mask` alongside `input_ids` so the model
correctly ignores padding tokens during calibration forward passes.

## Details
In `modelopt/torch/utils/dataset_utils.py`, the tokenizer call at line
387 with `padding=True` produces both `input_ids` and `attention_mask`.
The `include_labels=True` path (line 406) already preserves the full
`batch_encoded` dict including `attention_mask`. However, the
`include_labels=False` path was only keeping `input_ids` "for backward
compatibility."

During the calibration forward loop (`_forward_loop` →
`_process_batch`), the batch dict is unpacked as `**kwargs` into
`model.forward()`. Without `attention_mask`, HF models default to
attending to all positions including padding, which pollutes calibration
statistics.

**Practical impact**: With `batch_size=1` there is no padding so the bug
is invisible. With larger batch sizes and variable-length samples,
shorter sequences get padded and the effect grows.

## Test plan
- [x] Existing unit tests pass
(`tests/unit/torch/utils/test_dataset_utils.py`)
- [x] Pre-commit hooks pass
- [ ] Verify PTQ accuracy with batch_size > 1 on a padded calibration
dataset (GPU required)

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Signed-off-by: Chenjie Luo <chenjiel@nvidia.com>
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

Author: cjluo-nv

modelopt/torch/utils/dataset_utils.py

@@ -405,8 +405,15 @@ def get_dataset_dataloader(
         )
         tokenized_dataset = _CustomDataset(batch_encoded)
     else:
-        # For backward compatibility, if labels are not needed, we only return the input_ids.
-        tokenized_dataset = _CustomDataset({"input_ids": batch_encoded["input_ids"]})
+        # Always include attention_mask so the model correctly ignores padding tokens
+        # during calibration. Without it, HF models create a full causal mask and
+        # padding tokens participate in attention, skewing calibration statistics.
+        tokenized_dataset = _CustomDataset(
+            {
+                "input_ids": batch_encoded["input_ids"],
+                "attention_mask": batch_encoded["attention_mask"],
+            }
+        )
 
     calib_dataloader = DataLoader(tokenized_dataset, batch_size=batch_size, shuffle=False)