152 Commits

Author	SHA1	Message	Date
RAM	01f6934162	[Executor] Adjust signal sending order in RL training (#3773) (#4066) (#4178) ... * Adjust processing order * fix bug * fix update_parameters bug * refine code	2025-09-22 14:31:36 +08:00
李泳桦	0fa28b1068	[fix] fix ep group all-reduce (#4140) ... * [fix] fix ep group all-reduce * [fix] fix clear/update lock not working when workers > 1 * [chore] add preemption triggered info log * [fix] fix code style * fix model_weights_signal (#4092) * fix model_weights_signal --------- Co-authored-by: Yuanle Liu <yuanlehome@163.com>	2025-09-18 10:34:49 +08:00
K11OntheBoat	7f9a9b37f3	Support limit thinking lengths (#4070) ... Co-authored-by: K11OntheBoat <“ruianmaidanglao@163.com”>	2025-09-17 12:40:08 +08:00
李泳桦	7ccbcc5a62	[feat] support prefix cache clearing when /clear_load_weight is called (#4091) ... * [feat] support clearing prefix cache (cherry-picked from release/2.1) * [fix] fix ipc suffix, use port instead * [fix] fix prefix caching not enabled * [fix] fix code style * [fix] wait for rank0 to update weight status	2025-09-16 11:11:20 +08:00
chen	fbb4e0f8d1	[CP]Glm45 air 2.2 (#4073) ... * [Feature] Support zai-org/GLM-4.5-Air BF16 model (#3928) * support glm45_air * [Feature] GLM-45-AIR Support Mix Quantization(Dense wfp8afp8 and wint8 triton_moe_backend) (#4051) * check * fix v1 load for mix and wint8 * check --quantizations 'None' * check * support RL rollout * check v1 loader * check glm rollout_model, change wfp8afp8 per_token_cast_to_fp8 to native impl * check rollout moe gate begin layer_id * check rollout e_score_correction_bias * delete infer_to_train_mapping={} * code check	2025-09-15 18:52:58 +08:00
yangjianfengo1	dfc94371ee	【FIX】Change the name of sparse attn from moba to plas (#4006) ... * 更新文档 * 【docs】 update readme (#4000) * 更新文档 * update readme * update docs * 【FIX】Change the name of sparse attn from moba to plas (#3845) * 更新文档 * 更新文档 * 更新文档 * 更新文档 * 修改moba为plas * code style * update ci * code style * update ci * code style --------- Co-authored-by: Jiang-Jia-Jun <163579578+Jiang-Jia-Jun@users.noreply.github.com>	2025-09-10 10:04:29 +08:00
Zero Rains	d43549953c	[Cherry-Pick][Bug Fix]fix the bug for real size 0 in cudagraph (#3888) ... * fix the bug for real size 0 in cudagraph * fix cache_messager --------- Co-authored-by: Jiang-Jia-Jun <163579578+Jiang-Jia-Jun@users.noreply.github.com>	2025-09-08 14:06:10 +08:00
chenjian	8d77c1cb51	[Optimize] optimize prefix cache in release22 (#3889) ... * optimize prefix cache in release22 * optimize prefix cache in release22 * fix worker * fix * fix --------- Co-authored-by: Jiang-Jia-Jun <163579578+Jiang-Jia-Jun@users.noreply.github.com>	2025-09-06 09:52:01 +08:00
chenjian	41cd3e24c9	[Feature] Enable prefix caching as default (#3816) ... * [Feature] Enable prefix caching as default * [Feature] Enable prefix caching as default * Set prefix caching as default * skip dynamic load * fix kill bug * fix kill bug * fix kill bug * fix ci * fix --------- Co-authored-by: Jiang-Jia-Jun <163579578+Jiang-Jia-Jun@users.noreply.github.com>	2025-09-06 09:51:34 +08:00
yinwei	77c1bd0813	[XPU]Fixed the issue of performance degradation caused by enabling ENABLE_V1_KVCACHE_SCHEDULER (#3900) ... * fix bug * fix bug * update * udpate * update	2025-09-05 19:17:25 +08:00
chenjian	8567ada09e	[Fix] disable scheduler v1 in guided decoding (#3877) ... * disable scheduler v1 in guided decoding * disable scheduler v1 in guided decoding	2025-09-04 20:54:55 +08:00
lizhenyun01	d40d3a5a4f	fix DP&&TP (#3872)	2025-09-04 14:38:26 +08:00
chenjian	a0c03510c0	[Bug fix] Fix prompt token ids dtype in v1 (#3861)	2025-09-04 11:02:37 +08:00
chenjian	fb1e0d6a87	[Feature] Set scheduler v1 as default (#3812) ... * [Feature] Set scheduler v1 as default * [Feature] Set scheduler v1 as default * [Feature] Set scheduler v1 as default * [Feature] Set scheduler v1 as default * [Feature] Set scheduler v1 as default * [Feature] Set scheduler v1 as default	2025-09-04 11:02:10 +08:00
ltd0924	0f42771a84	[Feature] support model weight update in ep (#3802) ... * Update config.py * Update ep.py * Update fused_moe_backend_base.py * Update dynamic_weight_manager.py * Update worker_process.py * fix ci	2025-09-02 20:52:47 +08:00
yangjianfengo1	3754a9906d	[Feature] block sparse attention (#3668) ... * 支持稀疏attn * fix bug * code style * fix moba attn get kv shape * 修复a100编译 * codestyle * code style * code style * code style * fix conflict * 增加单侧 * code style * 增加eblite 加载时间 * fix bug * for ci * for ci * for ci * for ci * 支持mlp block size 128 * 增加小算子单测 * fix 单测 mlp * 将环境变量加入到config里面 * fix rollout config * 修复显存 * add test server * add test server * fix mlp 最后一层使用full attn	2025-08-29 19:46:30 +08:00
zhouchong	ccd52b5596	[Model]support qwen2_5_vl (#3557) ... * adapt qwen_2_5_vl model * adapt qwen_2_5_vl VIT model * adapt qwen2_5_vl images_embeds * adapt qwen2_5_vl 3D rope * adapt qwen2_5_vl 3D rope v2 * adapt qwen2_5_vl processor * adapt qwen2_5_vl bypass resampler_model * adapt qwen2_5_vl 绕过部分ernie逻辑 * adapt qwen2_5_vl 绕过部分ernie逻辑 v2 * adapt qwen2_5_vl 权重加载与命名修改 * adapt qwen2_5_vl 非必须think_end_id * adapt qwen2_5_vl 区分多种模型的extract_vision_features * fix:adapt qwen2_5_vl model * adapt qwen2_5_vl norm * adapt qwen2_5_vl processor 更新 * adapt qwen2_5_vl image and video success * adapt qwen2_5_vl 部分整理代码 * adapt qwen2_5_vl 支持多卡 * adapt qwen2_5_vl on latest develop * adapt qwen2_5_vl RL * adapt qwen2_5_vl 整理代码 * support noex rope3d * adapt qwen2_5_vl add init.py * adapt qwen2_5_vl add init.py v2 * adapt qwen2_5_vl remove space * adapt qwen2_5_vl remove space v2 * adapt qwen2_5_vl pre-commit * adapt qwen2_5_vl update * adapt qwen2_5_vl pre-commit v2 * adapt qwen2_5_vl modify comments * adapt qwen2_5_vl fix indentation * adapt qwen2_5_vl fix indentation v2 --------- Co-authored-by: wangyafeng <wangyafeng@baidu.com> Co-authored-by: xiaoxiaohehe001 <49090790+xiaoxiaohehe001@users.noreply.github.com> Co-authored-by: CSWYF3634076 <58356743+CSWYF3634076@users.noreply.github.com>	2025-08-29 18:28:39 +08:00
Yuan Xiaolan	c71ee0831c	add w4afp8 offline script (#3636)	2025-08-29 17:56:05 +08:00
Ryan	45f81b34f0	add dtype int32 (#3692)	2025-08-29 14:56:35 +08:00
lifulll	72094d4d82	enable dcu ci (#3402)	2025-08-29 10:23:08 +08:00
Yuanle Liu	4957908275	add input_processor plugin (#3657) ... * add input_processor plugin * update * update * update * update * update * update * update * update * update * update * update	2025-08-28 22:53:57 +08:00
ming1753	02b3644903	[Bug Fix] VL Support w4a8/w4afp8 (#3686)	2025-08-28 21:38:35 +08:00
gaoziyuan	fc635acc47	[BugFix]fix dp&ep&tp and muti node infer (#3629) ... * rm log * fix bug * fix bug * fix dp&ep&tp and muti node infer * fix --------- Co-authored-by: Yuanle Liu <yuanlehome@163.com>	2025-08-28 19:09:10 +08:00
Zero Rains	e37e86b3b8	[V1 Loader]support param create and load for wint2 and xpu backend (#3581) ... * support wint2 backend' * [V1 Loader]support param create and load for wint2 and xpu backend * update weight shape name * update * update * update baseline.txt * update model name * update baseline.txt * fix codestyle * remove debug coode	2025-08-28 09:49:36 +08:00
lizexu123	b28a0343a6	fix ENABLE_V1_KVCACHE_SCHEDULER (#3625) ... Co-authored-by: Jiang-Jia-Jun <163579578+Jiang-Jia-Jun@users.noreply.github.com>	2025-08-27 21:21:29 +08:00
Jiang-Jia-Jun	c694fa2879	Revert "[Feature] block sparse attention (#3209)" (#3647) ... This reverts commit 646a0c2fd8.	2025-08-27 17:35:04 +08:00
李泳桦	b2afdf4fc6	[fix] qwen output inconsistency when top_p=0 (#3634) ... * [fix] qwen output inconsistency when top_p=0 * [fix] remove decode pre_id code	2025-08-27 17:16:23 +08:00
chen	ce9c0917c5	[Precision] Support lm_head layer running in float32 (#3597) ... * support lm_head fp32 bf16 fp16 * support lm_head fp32 bf16 fp16 * add doc and check code * lm_head_fp32 specify lm_head as fp32 * code check * check doc	2025-08-27 11:34:53 +08:00
yangjianfengo1	646a0c2fd8	[Feature] block sparse attention (#3209) ... * 支持稀疏attn * fix bug * code style * fix moba attn get kv shape * 修复a100编译 * codestyle * code style * code style * code style * fix conflict * 增加单侧 * code style * 增加eblite 加载时间 * fix bug * for ci * for ci * for ci * for ci * 支持mlp block size 128 * 增加小算子单测 * fix 单测 mlp * 将环境变量加入到config里面 * fix rollout config	2025-08-26 07:16:04 -07:00
gaoziyuan	82e64b13e1	[NewFeature]Support dp multi api server && Fix some bug in mixed ep && merge develop (#3598) ... * [Feature] update ep * fix ci * fix ci * fix ci * fix ci * fix ci * fix ci * fix ci * fix queue ports idx * fix ci * fix ci * fix ci * fix ci * fix ci * fix ci * fix ci * fix ci * Update engine.py * fix ci * fix some bug in mixed ep * add server fix and op fix * rm some log * fix code style * ltd fix * fix * fix * fix some bug * fix bug * fix bug * fix style * Update config.py * Update splitwise_connector.py * Update cache_messager.py * Update __init__.py * merge and fix * Update engine.py * Update common_engine.py * Update run_ci_xpu.sh * Update ernie_processor.py * Update ernie_processor.py --------- Co-authored-by: ltd0924 <ltd0924@sina.com> Co-authored-by: ltd0924 <32387785+ltd0924@users.noreply.github.com>	2025-08-26 19:59:02 +08:00
Yuanle Liu	cbce94a00e	rename ernie_xxx to ernie4_5_xxx (#3621) ... * rename ernie_xxx to ernie4_5_xxx * ci fix	2025-08-26 19:29:27 +08:00
bukejiyu	3200a80de3	[v1 loader]support fp8 (#3593) ... * support fp8 * update ci	2025-08-26 02:42:46 -07:00
lzy	d339df2e90	Supports DP+TP+EP hybrid parallel deployment strategy (#3489) ... * Support DP+TP+EP hybrid parallel deployment strategy * Support DP+TP+EP hybrid parallel deployment strategy * fix conflict * add moe_tp_ep function split_allgather_out * del tp_group in moe_cutlass_backend * for ci * fix parallel_config for ci * del log	2025-08-26 00:04:01 -07:00
Sunny-bot1	c68c3c4b8b	[Feature] bad words support v1 scheduler and specifiy token ids (#3608) ... * support bad_words_token_ids * docs * fix test * fix * bad words support kvcache v1 and token ids * fix	2025-08-25 20:14:51 -07:00
RAM	2fa173e327	[Executor] CUDAGraph support RL training (#3265) ... * add clear graph opt backend * cuda graph support rl * add branch * 1.fix dynamic_weight_manager bug 2.add clear api for CasualLM * open test case * fix typo * update mkdocs.yaml * [Docs]Update mkdocs.yml * update test case * use unittest in graph test case	2025-08-25 20:59:30 +08:00
Kane2011	2ae7ab28d2	[MetaxGPU] adapt to the latest fastdeploy on metax gpu (#3492)	2025-08-25 17:44:20 +08:00
chen	9cab3f47ff	[Feature] Add temp_scaled_logprobs and top_p_normalized_logprobs parameters for logits and logprobs post processing (#3552) ... * [feature] Add temp_scaled_logprobs and top_p_normalized_logprobs parameters for logits and logprobs post processing * infer engine support temp_scaled_logprobs and top_p_normalized_logprobs * delete some code * code check * code check and add doc * fix tokenizer.decoder(-1), return 'Invalid Token' * add ci for temp_scaled and top_p logprobs * check test * check seq len time shape * logprob clip inf --------- Co-authored-by: sunlei1024 <sunlei5788@gmail.com>	2025-08-25 14:11:49 +08:00
zhink	df7c31012b	Modified to support custom all reduce by default (#3538)	2025-08-22 16:59:05 +08:00
YuanRisheng	5b66462f0e	Fix fdconfig bugs (#3528) ... * fix config * fix parallel * fix ips * fix rl * open code	2025-08-22 16:17:15 +08:00
YuanRisheng	c389a4013c	Unify server-side and model-side Config(Part-5) (#3497) ... * move config * fix xpu * fix * fix vl * fix vl * fix unitest * fix args * add unitest * fix test	2025-08-21 19:00:21 +08:00
qw86972190	c83381d650	revert pr (#3481) ... Co-authored-by: iosmers <yinwei_hust@163.com>	2025-08-21 14:19:50 +08:00
kevin	67298cf4c0	add error traceback info (#3419) ... * add error traceback info * update error msg * update code --------- Co-authored-by: Jiang-Jia-Jun <163579578+Jiang-Jia-Jun@users.noreply.github.com>	2025-08-19 19:32:04 +08:00
lizexu123	a053ab889b	[BugFix] fix num_running_requests in cuda_graph (#3457) ... * fix cuda_grpah * add note --------- Co-authored-by: RAM <gstian5555@outlook.com>	2025-08-19 10:47:22 +08:00
lizexu123	32b39620bc	[Code Simplification] remove cum_offsets (#3410)	2025-08-18 20:21:25 +08:00
Jundong Liu	ea4a3b479c	[Excutor] Increase buffer size to prevent address corruption; add forward metadata debug tool (#3404) ... * 修复buffer申请不够大，增加打印forwardmetadata的工具 * fix mistake * Make CPU tensor in CPUPlace * Add test about forward_meta_str and Add unitest_requirement --------- Co-authored-by: RAM <gstian5555@outlook.com>	2025-08-18 16:14:09 +08:00
chen	f0f00a6025	[OPs] Universal optimization and Fix early_stop cuda 700 (#3375) ... * delete nonzero * delete setup_ops_base.py * check if * check gcp infer_seed.cpu() * fix repetition_early_stopper_kernel cuda 700	2025-08-14 22:40:44 +08:00
lizexu123	7b596d0877	[BugFix] fix real_bsz in ep (#3366) ... * Your commit message here * fix ep * delete cuda_graph	2025-08-14 17:31:19 +08:00
EnflameGCU	d1a92e3e17	[GCU] Enable gcu CI (#3190) ... * [GCU] Update to the latest version * [GCU] Enable CI	2025-08-13 11:48:24 +08:00
yzwu	ce9180241e	[Iluvatar GPU] Modify the names of some variables (#3273)	2025-08-13 11:38:02 +08:00
Kane2011	b4fef2cf29	[MetaxGPU] Support FastDeploy on metax gpu (#3241) ... * [MetaxGPU] Support FastDeploy on metax gpu * Update metax_worker.py 1. change worker log; 2. remove custom allreduce, adapt it later; 3. remove cuda graph; * Update __init__.py 1. remove metax's key work comment * Update __init__.py 1. remove metax's key word comment; 2. add fused_moe_kernel_paddle import --------- Co-authored-by: yongqiangma <xing.wo@163.com>	2025-08-13 11:11:54 +08:00