Add padding to encoder_seq_lens (HabanaAI#610)

Without this change we can observe below error:
```
[rank0]:   File "/software/users/kdamaszke/repos/vllm-fork/vllm/model_executor/models/mllama.py", line 959, in forward
[rank0]:     full_text_row_masked_out_mask = full_text_row_masked_out_mask.view(
[rank0]: RuntimeError: shape '[4, -1, 1]' is invalid for input of size 3
```
It occurs when one of the requests is removed from the batch earlier. In
that case, language model is still working on the shapes padded to the
bucketed batch size, while encoder input doesn't. This change is
aligning the batch size on `encoder_seq_lens` to the expected one.

vllm/worker/hpu_enc_dec_model_runner.py

@@ -332,6 +332,15 @@ def _prepare_encoder_model_input_tensors(
             attn_metadata.cross_block_groups = block_groups
             attn_metadata.cross_block_usage = block_usage
 
+        # add padding to align with language model shapes
+        real_batch_size = len(seq_group_metadata_list)
+        batch_size_padded = self.bucketing_ctx.get_padded_batch_size(
+            real_batch_size, is_prompt)
+        batch_size_padding = batch_size_padded - real_batch_size
+        if batch_size_padding > 0:
+            encoder_seq_lens.extend(encoder_seq_lens[0]
+                                    for _ in range(batch_size_padding))
+
         encoder_seq_lens_tensor = self._list_to_int32_tensor(encoder_seq_lens)
         attn_metadata.encoder_seq_lens = encoder_seq_lens
         attn_metadata.encoder_seq_lens_tensor = encoder_seq_lens_tensor