GitLab

- Ensure embedding tensors are on the same device. Change format.
- Optimize embedding kernel with vectorized memory access and __ldg
- Add vectorized memory access using float4/float2, half2, and bfloat162
- Use __ldg instruction for read-only weight and indices access
- Add memory alignment checks to enable vectorized paths
- Add __restrict__ keywords for better compiler optimization
- Implement dynamic block size selection based on embedding_dim

issue/811 fix tensor to blob and resume

issue/811 support cuda graph capture

issue/810 feat: allow graph tensor to resume to allocator's tracking

issue/920 RoPE supports longrope

issue/811 remove shortcut for cpu runtime

issue/867 fix cpu malloc

issue/867 fix page caching api, paged attn support more head dims

issue/810 static compute graph infra

Issue/867： adjust paged_attention_prefill interface naming

Issue/791 增加add_rms_norm融合算子

issue/877 - return saved_file in TestManager.test method

Issue/847  paged attention prefill一段式接口

issue/848: add paged attention prefill for nvidia gpu with test pass

issue/847-paged caching和atention添加infinicore的接口和测试