[LLM INFER] Append attn (PaddlePaddle#9244)

* refine paddle::empty(), fix memory error, support multi_stream for attention

* fix and rename attention as append_attention

* rename file
---------

Co-authored-by: lizhenyun <[email protected]>
Co-authored-by: lizhenyun01 <[email protected]>

csrc/gpu/append_attn/append_attention_kernel.h

@@ -0,0 +1,295 @@
+// Copyright (c) 2024 PaddlePaddle Authors. All Rights Reserved.
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+#pragma once
+
+#include "helper.h"
+#include "utils.cuh"
+
+template <typename T, typename OutT>
+void CascadeAppendAttentionC16Kernel(
+    const AppendAttnMetaData& meta_data,
+    const paddle::Tensor& qkv,  // [token_num, num_heads, head_dim]
+    const paddle::Tensor&
+        cache_k,  // [max_block_num, num_heads, block_size, head_dim]
+    const paddle::Tensor&
+        cache_v,  // [max_block_num, num_heads, head_dim, block_size]
+    const paddle::optional<paddle::Tensor>& attn_mask,
+    const paddle::optional<paddle::Tensor>&
+        cache_k_scale,  // [num_kv_heads, head_dim]
+    const paddle::optional<paddle::Tensor>&
+        cache_v_scale,  // [num_kv_heads, head_dim]
+    const paddle::optional<paddle::Tensor>&
+        cache_k_zp,  // [num_kv_heads, head_dim]
+    const paddle::optional<paddle::Tensor>&
+        cache_v_zp,  // [num_kv_heads, head_dim]
+    const paddle::optional<paddle::Tensor>&
+        shift_bias,  // [num_kv_heads, head_dim]
+    const paddle::optional<paddle::Tensor>&
+        smooth_weight,  // [num_kv_heads, head_dim]
+    const paddle::Tensor& seq_lens_q,
+    const paddle::Tensor& seq_lens_kv,
+    const paddle::Tensor& seq_lens_encoder,
+    const paddle::Tensor& padding_offsets,
+    const paddle::Tensor& cum_offsets,
+    const paddle::Tensor& block_table,
+    const paddle::Tensor& batch_ids,
+    const paddle::Tensor& tile_ids_per_batch,
+    const int num_blocks,
+    const int block_shape_q,
+    const int max_seq_len,
+    const int max_dec_len,
+    const float in_scale,
+    const int max_partition_size,
+    const int encoder_max_partition_size,
+    const int speculate_max_draft_token_num,
+    const bool causal,
+    const bool is_decoder,
+    const bool enable_prefill,
+    cudaStream_t& stream,
+    paddle::Tensor* out);
+
+template <typename T, typename OutT>
+void CascadeAppendAttentionC8Kernel(
+    const AppendAttnMetaData& meta_data,
+    const paddle::Tensor& qkv,  // [token_num, num_heads, head_dim]
+    const paddle::Tensor&
+        cache_k,  // [max_block_num, num_heads, block_size, head_dim]
+    const paddle::Tensor&
+        cache_v,  // [max_block_num, num_heads, head_dim, block_size]
+    const paddle::optional<paddle::Tensor>& attn_mask,
+    const paddle::optional<paddle::Tensor>&
+        cache_k_scale,  // [num_kv_heads, head_dim]
+    const paddle::optional<paddle::Tensor>&
+        cache_v_scale,  // [num_kv_heads, head_dim]
+    const paddle::optional<paddle::Tensor>&
+        cache_k_zp,  // [num_kv_heads, head_dim]
+    const paddle::optional<paddle::Tensor>&
+        cache_v_zp,  // [num_kv_heads, head_dim]
+    const paddle::optional<paddle::Tensor>&
+        shift_bias,  // [num_kv_heads, head_dim]
+    const paddle::optional<paddle::Tensor>&
+        smooth_weight,  // [num_kv_heads, head_dim]
+    const paddle::Tensor& seq_lens_q,
+    const paddle::Tensor& seq_lens_kv,
+    const paddle::Tensor& seq_lens_encoder,
+    const paddle::Tensor& padding_offsets,
+    const paddle::Tensor& cum_offsets,
+    const paddle::Tensor& block_table,
+    const paddle::Tensor& batch_ids,
+    const paddle::Tensor& tile_ids_per_batch,
+    const int num_blocks,
+    const int block_shape_q,
+    const int max_seq_len,
+    const int max_dec_len,
+    const float in_scale,
+    const int max_partition_size,
+    const int encoder_max_partition_size,
+    const int speculate_max_draft_token_num,
+    const bool causal,
+    const bool is_decoder,
+    const bool enable_prefill,
+    cudaStream_t& stream,
+    paddle::Tensor* out);
+
+template <typename T, typename OutT>
+void CascadeAppendAttentionC4Kernel(
+    const AppendAttnMetaData& meta_data,
+    const paddle::Tensor& qkv,  // [token_num, num_heads, head_dim]
+    const paddle::Tensor&
+        cache_k,  // [max_block_num, num_heads, block_size, head_dim]
+    const paddle::Tensor&
+        cache_v,  // [max_block_num, num_heads, head_dim, block_size]
+    const paddle::optional<paddle::Tensor>& attn_mask,
+    const paddle::optional<paddle::Tensor>&
+        cache_k_scale,  // [num_kv_heads, head_dim]
+    const paddle::optional<paddle::Tensor>&
+        cache_v_scale,  // [num_kv_heads, head_dim]
+    const paddle::optional<paddle::Tensor>&
+        cache_k_zp,  // [num_kv_heads, head_dim]
+    const paddle::optional<paddle::Tensor>&
+        cache_v_zp,  // [num_kv_heads, head_dim]
+    const paddle::optional<paddle::Tensor>&
+        shift_bias,  // [num_kv_heads, head_dim]
+    const paddle::optional<paddle::Tensor>&
+        smooth_weight,  // [num_kv_heads, head_dim]
+    const paddle::Tensor& seq_lens_q,
+    const paddle::Tensor& seq_lens_kv,
+    const paddle::Tensor& seq_lens_encoder,
+    const paddle::Tensor& padding_offsets,
+    const paddle::Tensor& cum_offsets,
+    const paddle::Tensor& block_table,
+    const paddle::Tensor& batch_ids,
+    const paddle::Tensor& tile_ids_per_batch,
+    const int num_blocks,
+    const int block_shape_q,
+    const int max_seq_len,
+    const int max_dec_len,
+    const float in_scale,
+    const int max_partition_size,
+    const int encoder_max_partition_size,
+    const int speculate_max_draft_token_num,
+    const bool causal,
+    const bool is_decoder,
+    const bool enable_prefill,
+    cudaStream_t& stream,
+    paddle::Tensor* out);
+
+template <typename T, typename OutT>
+void CascadeAppendAttentionKernel(
+    const AppendAttnMetaData& meta_data,
+    const paddle::Tensor& qkv,  // [token_num, num_heads, head_dim]
+    const paddle::Tensor&
+        cache_k,  // [max_block_num, num_heads, block_size, head_dim]
+    const paddle::Tensor&
+        cache_v,  // [max_block_num, num_heads, head_dim, block_size]
+    const paddle::optional<paddle::Tensor>& attn_mask,
+    const paddle::optional<paddle::Tensor>&
+        cache_k_scale,  // [num_kv_heads, head_dim]
+    const paddle::optional<paddle::Tensor>&
+        cache_v_scale,  // [num_kv_heads, head_dim]
+    const paddle::optional<paddle::Tensor>&
+        cache_k_zp,  // [num_kv_heads, head_dim]
+    const paddle::optional<paddle::Tensor>&
+        cache_v_zp,  // [num_kv_heads, head_dim]
+    const paddle::optional<paddle::Tensor>&
+        shift_bias,  // [num_kv_heads, head_dim]
+    const paddle::optional<paddle::Tensor>&
+        smooth_weight,  // [num_kv_heads, head_dim]
+    const paddle::Tensor& seq_lens_q,
+    const paddle::Tensor& seq_lens_kv,
+    const paddle::Tensor& seq_lens_encoder,
+    const paddle::Tensor& padding_offsets,
+    const paddle::Tensor& cum_offsets,
+    const paddle::Tensor& block_table,
+    const paddle::Tensor& batch_ids,
+    const paddle::Tensor& tile_ids_per_batch,
+    const std::string& cache_quant_type_str,
+    const int num_blocks,
+    const int block_shape_q,
+    const int max_seq_len,
+    const int max_dec_len,
+    const float in_scale,
+    const int max_partition_size,
+    const int encoder_max_partition_size,
+    const int speculate_max_draft_token_num,
+    const bool causal,
+    const bool is_decoder,
+    const bool enable_prefill,
+    cudaStream_t& stream,
+    paddle::Tensor* out) {
+  if (cache_quant_type_str == "none") {
+    CascadeAppendAttentionC16Kernel<T, OutT>(meta_data,
+                                             qkv,
+                                             cache_k,
+                                             cache_v,
+                                             attn_mask,
+                                             cache_k_scale,
+                                             cache_v_scale,
+                                             cache_k_zp,
+                                             cache_v_zp,
+                                             shift_bias,
+                                             smooth_weight,
+                                             seq_lens_q,
+                                             seq_lens_kv,
+                                             seq_lens_encoder,
+                                             padding_offsets,
+                                             cum_offsets,
+                                             block_table,
+                                             batch_ids,
+                                             tile_ids_per_batch,
+                                             num_blocks,
+                                             block_shape_q,
+                                             max_seq_len,
+                                             max_dec_len,
+                                             in_scale,
+                                             max_partition_size,
+                                             encoder_max_partition_size,
+                                             speculate_max_draft_token_num,
+                                             causal,
+                                             is_decoder,
+                                             enable_prefill,
+                                             stream,
+                                             out);
+  } else if (cache_quant_type_str == "cache_int8") {
+    CascadeAppendAttentionC8Kernel<T, OutT>(meta_data,
+                                            qkv,
+                                            cache_k,
+                                            cache_v,
+                                            attn_mask,
+                                            cache_k_scale,
+                                            cache_v_scale,
+                                            cache_k_zp,
+                                            cache_v_zp,
+                                            shift_bias,
+                                            smooth_weight,
+                                            seq_lens_q,
+                                            seq_lens_kv,
+                                            seq_lens_encoder,
+                                            padding_offsets,
+                                            cum_offsets,
+                                            block_table,
+                                            batch_ids,
+                                            tile_ids_per_batch,
+                                            num_blocks,
+                                            block_shape_q,
+                                            max_seq_len,
+                                            max_dec_len,
+                                            in_scale,
+                                            max_partition_size,
+                                            encoder_max_partition_size,
+                                            speculate_max_draft_token_num,
+                                            causal,
+                                            is_decoder,
+                                            enable_prefill,
+                                            stream,
+                                            out);
+  } else if (cache_quant_type_str == "cache_int4_zp") {
+    CascadeAppendAttentionC4Kernel<T, OutT>(meta_data,
+                                            qkv,
+                                            cache_k,
+                                            cache_v,
+                                            attn_mask,
+                                            cache_k_scale,
+                                            cache_v_scale,
+                                            cache_k_zp,
+                                            cache_v_zp,
+                                            shift_bias,
+                                            smooth_weight,
+                                            seq_lens_q,
+                                            seq_lens_kv,
+                                            seq_lens_encoder,
+                                            padding_offsets,
+                                            cum_offsets,
+                                            block_table,
+                                            batch_ids,
+                                            tile_ids_per_batch,
+                                            num_blocks,
+                                            block_shape_q,
+                                            max_seq_len,
+                                            max_dec_len,
+                                            in_scale,
+                                            max_partition_size,
+                                            encoder_max_partition_size,
+                                            speculate_max_draft_token_num,
+                                            causal,
+                                            is_decoder,
+                                            enable_prefill,
+                                            stream,
+                                            out);
+  } else {
+    PD_THROW(
+        "cache_quant_type_str should be one of [none, cache_int8, "
+        "cache_int4_zp]");
+  }
+}