root
06f4ce3bb9
[fix] baseline4 ixf_F.copy_blocks名字不匹配
2026-08-18 15:35:03 +00:00
root
eaf4cfa9c1
trigger rebuild
2026-08-18 13:50:06 +00:00
root
2579ef6867
[fix] baseline4 yaml
2026-08-18 13:43:22 +00:00
root
cb396eca26
trigger rebuild
2026-08-18 10:33:20 +00:00
root
8c206feb91
[fix] baseline4 paged_attn.py 的 seq_lens_tensor 包含 0x7FFF7FFF
2026-08-18 10:32:55 +00:00
root
8d65615b19
trigger rebuild
2026-08-18 10:21:50 +00:00
root
614bad6232
[feat] baseline4 profiling
2026-08-18 10:21:15 +00:00
root
3c53587de2
[fix] baseline4 NaN 0X7FFF7FFF
2026-08-18 09:54:25 +00:00
root
7a02a5bf92
baseline4 NAN 0X7FFF7FFF
2026-08-18 08:37:30 +00:00
root
ab078989da
[fix] reshape_paged_cache slot_mapping int64→int32 (C++ data_ptr<int>要求)
2026-08-18 07:41:38 +00:00
root
ac6d51c936
delete prd.md
2026-08-18 07:19:16 +00:00
root
47f6d30c2d
trigger rebuild
2026-08-18 07:12:13 +00:00
root
56694e5075
trigger rebuild
2026-08-18 07:12:01 +00:00
root
9e1df1482e
[fix] xllm_ops 5处C++参数签名
2026-08-18 07:10:18 +00:00
root
c92135111e
[fix] reshape_paged_cache参数顺序
2026-08-18 06:44:01 +00:00
root
62e3312258
trigger rebuild
2026-08-18 06:15:05 +00:00
root
b89bd3d72f
[fix] ix_moe_bridge pybind名无ix_前缀: ix_paged_attention→paged_attention, ix_linear→linear
2026-08-18 06:14:15 +00:00
root
54a9b572af
[fix] xllm_ops 7处函数名匹配.so真实导出符号
...
- reshape_and_cache → reshape_paged_cache (xllm_cache.so)
- residual_rms_norm → fused_add_rms_norm (xllm_norm.so)
- topk_softmax → moe_fused_topk (xllm_moe.so)
- moe_compute_token_index → moe_compute_index (xllm_moe.so)
- ix_full_bridge → ix_moe_bridge (paged_attention, ix_linear)
- ix_linear 5参数 → 3参数 (input, weight, bias)
- check_all: ix_moe_bridge改为required, ix_full_bridge改为optional
2026-08-18 06:11:03 +00:00
root
fccb78df09
Merge branch 'main' of https://dev.modelhub.org.cn/dylanyunlong/project_6 into main
2026-08-18 04:17:58 +00:00
root
8cc6a91b8a
[fix] baseline4 use ix_moe_bridge.so instead of broken ix_full_bridge.so
2026-08-18 04:17:37 +00:00
root
13b12aac4c
[fix] baseline4 rms_norm crash
2026-08-18 03:35:26 +00:00
root
d6d12c56d8
[fix] baseline4 debug info
2026-08-18 03:33:06 +00:00
root
c6eb038429
trigger rebuild
2026-08-18 03:33:06 +00:00
root
63e574ad0b
trigger rebuild
2026-08-18 03:33:06 +00:00
root
da1382bb23
trigger rebuild
2026-08-18 03:33:06 +00:00
root
bcb18cf328
trigger rebuild
2026-08-18 03:33:06 +00:00
root
4f9f31f094
[fix] baseline4 debug info
2026-08-18 03:33:06 +00:00
root
448996d386
trigger rebuild
2026-08-18 03:33:06 +00:00
root
716ede67fd
trigger rebuild
2026-08-18 03:33:06 +00:00
root
763cda98c5
trigger rebuild
2026-08-18 03:33:06 +00:00
root
2b0d98a867
[fix] baseline4 debug info
2026-08-18 02:53:21 +00:00
root
31aa943248
trigger rebuild
2026-08-18 02:53:21 +00:00
root
902655f1bc
trigger rebuild
2026-08-18 02:53:21 +00:00
root
74cdc68a1c
trigger rebuild
2026-08-18 02:53:21 +00:00
Claude
6fe1272f13
doc: MoE 5函数符号真相 — 不在预装.so里,由moe_ops_impl.cu自编译
...
确认结果:
- topk_softmax, moe_compute_token_index_api, moe_expand_input,
moe_w16a16_group_gemm, moe_output_reduce_sum
- namespace 是 ixformer::infer (Tensor API), 不是 ixformer::kernels::infer (raw ptr API)
- 4个镜像.so全部nm -D搜过,均无这些符号
- 实现在 ex_engine/csrc/moe_ops_impl.cu (自编译CUDA kernel)
- 由 build_moe_bridge.sh 编译进 ix_full_bridge_v2.so
- kernels.h 里的 ixformer::kernels::infer 是SDK头文件声明(模板+raw ptr)
我们的 ixformer::infer 是独立实现(torch::Tensor wrapper)
2026-08-18 02:53:21 +00:00
root
714c41d17b
trigger rebuild
2026-08-18 02:32:00 +00:00
root
96ef5f27d3
[fix] baseline4 debug info
2026-08-18 02:20:43 +00:00
Claude
b77743451c
debug: rms_norm weight.dim()!=1 — 打印前20次调用的w.shape, x.shape, class
2026-08-18 02:08:37 +00:00
root
e814180e04
trigger rebuild
2026-08-18 01:26:22 +00:00
root
ae27d0405b
Merge branch 'main' of https://github.com/dylanyunlon/project_6
2026-08-17 18:11:57 +00:00
Claude
f7b1b2d119
fix: patched_reshape_and_cache signature 7→8 args (k_scale, v_scale)
...
Original vllm ops.reshape_and_cache passes 8 args:
key, value, key_cache, value_cache, slot_mapping, kv_cache_dtype, k_scale, v_scale
Our patch had 7 (single kv_scale) → TypeError on first request → engine dead.
2026-08-17 18:10:52 +00:00
root
d42b0c1c04
Merge branch 'main' of https://dev.modelhub.org.cn/dylanyunlong/project_6
2026-08-17 17:45:24 +00:00
Claude
8d969822b4
fix: revert ix_full_bridge.so to d32822c5 original (160KB)
...
be5e23f3 rebuilt this .so but introduced ABI mismatch:
undefined symbol: c10::detail::torchInternalAssertFail
The original 160KB .so does not reference this symbol and works
with the Docker image's libtorch.
2026-08-17 17:39:21 +00:00
root
f439f67f39
baseline4 trigger rebuild
2026-08-17 17:16:11 +00:00
Claude
f287382f99
fix: update SHA256SUMS for rebuilt ix_full_bridge.so
...
ix_full_bridge.so was rebuilt in be5e23f3 (without MoE ixformer::infer deps)
but SHA256SUMS was not updated. install_prebuilt_corex.sh sha256sum --check
fails → set -eo pipefail kills patch_ops.sh → cli_args.py/api_server.py
never deployed → vllm rejects --reasoning-parser qwen3 at startup.
2026-08-17 17:12:41 +00:00
root
ff18454eeb
trigger rebuild
2026-08-17 16:52:37 +00:00
root
be5e23f335
fix: rebuild ix_full_bridge.so without MoE ixformer::infer deps
2026-08-17 16:25:16 +00:00
root
5c97e3dcb8
trigger rebuild
2026-08-17 14:39:56 +00:00
root
5b96a91156
[fix] baseline4 不是 CUDA 不可用,是 .so 在非主进程里被 dlopen 时段错误。需要用进程级别的判断
2026-08-17 14:17:03 +00:00
root
053dc036b8
[fix] prebuilt 的 ix_full_bridge.so 在子进程(没有 GPU 初始化)里被 dlopen
2026-08-17 14:01:36 +00:00