AReal - Browse /v1.0.3 at SourceForge.net

The interactive file manager requires Javascript. Please enable it or use sftp or scp.
You may still browse the files here.

Name	Modified	Size	InfoDownloads / Week
Parent folder
README.md	2026-04-16	9.8 kB	0
v1.0.3 source code.tar.gz	2026-04-16	9.3 MB	0
v1.0.3 source code.zip	2026-04-16	9.8 MB	0
Totals: 3 Items		19.2 MB	0

What's Changed

chore(docker): add openclaw, ironclaw, zeroclaw, and nanobot-ai to runtime image by @garrett4wade in https://github.com/inclusionAI/AReaL/pull/1051
feat(agent-service): add Agent Service microservice infrastructure by @CormickKneey in https://github.com/inclusionAI/AReaL/pull/1048
feat(gateway): Add rollout gateway infrastructure with controller, router, and data proxy by @nuzant in https://github.com/inclusionAI/AReaL/pull/1043
feat: estimators for kl divergence by @NicolasArias in https://github.com/inclusionAI/AReaL/pull/1060
test(infra): speed up inference service integration tests by @nuzant in https://github.com/inclusionAI/AReaL/pull/1068
fix(infra): simplify RTensor serialization in data proxy by @garrett4wade in https://github.com/inclusionAI/AReaL/pull/1067
fix(rpc): resolve connection reset during RTensor fetch with large payloads by @pratyush618 in https://github.com/inclusionAI/AReaL/pull/1075
docs: add gitcgr code graph badge by @vitali87 in https://github.com/inclusionAI/AReaL/pull/1073
fix(openai): handle streaming responses in chat/completions endpoint by @Zijun9 in https://github.com/inclusionAI/AReaL/pull/1053
fix: add PIL image and processor serialization for VLM RPC by @Adiactive in https://github.com/inclusionAI/AReaL/pull/1070
refactor(api): migrate allocation_mode to per-engine backend fields by @garrett4wade in https://github.com/inclusionAI/AReaL/pull/1044
chore(agents): add Codex harness and align AI workflows by @rchardx in https://github.com/inclusionAI/AReaL/pull/1082
feat(platform): add NUMA CPU affinity binding for training engines by @HT-Yuan in https://github.com/inclusionAI/AReaL/pull/1083
feat(commands): add fork workflow support to create-pr skill by @guozhihao-224 in https://github.com/inclusionAI/AReaL/pull/1092
fix(rpc): batch HTTP RTensor fetches for large multimodal batches by @Wangxiaoxiaoa in https://github.com/inclusionAI/AReaL/pull/1077
fix(fsdp): stabilize Qwen-VL rope-index argument binding and dtype by @Adiactive in https://github.com/inclusionAI/AReaL/pull/1094
Refactor(vllm): use pause_generation from vllm instead of abort_all_req in areal_vllm_server by @HwVanICI in https://github.com/inclusionAI/AReaL/pull/1091
feat: add BailingMoeV2.5 support with Lightning Attention + MLA + MoE + CP by @dingzhiqiang in https://github.com/inclusionAI/AReaL/pull/1079
feat：support model training in IPv6-only environment by @TaoZex in https://github.com/inclusionAI/AReaL/pull/1072
fix: fix pad_packed_tensor_dict by @HKAB in https://github.com/inclusionAI/AReaL/pull/1104
feat: megatron bridge adaptation by @gursimar in https://github.com/inclusionAI/AReaL/pull/1056
fix(engine): remove duplicate trust_remote_code kwarg in MegatronBridge init by @rchardx in https://github.com/inclusionAI/AReaL/pull/1107
fix(dataloader): prevent data drop and padding during validation for accurate metrics by @Anguo-star in https://github.com/inclusionAI/AReaL/pull/1100
fix(archon): add missing POST /data/batch endpoint to data proxy by @rchardx in https://github.com/inclusionAI/AReaL/pull/1105
refactor(engine): abstract CUDA calls via current_platform in PerLayerOptimWrapper by @guozhihao-224 in https://github.com/inclusionAI/AReaL/pull/1108
perf(fsdp): pipeline distributed weight sync with a single pending bucket by @HT-Yuan in https://github.com/inclusionAI/AReaL/pull/1074
fix(engine): restore SGLang VLM training by @garrett4wade in https://github.com/inclusionAI/AReaL/pull/1098
feat(archon): add FP8 blockwise training support by @rchardx in https://github.com/inclusionAI/AReaL/pull/1087
chore(ci): update GCP CI image by @garrett4wade in https://github.com/inclusionAI/AReaL/pull/1115
feat(inference-service): complete vLLM backend support in inference service by @garrett4wade in https://github.com/inclusionAI/AReaL/pull/1112
fix(archon): harden FP8 blockwise training for TP and MoE scenarios by @rchardx in https://github.com/inclusionAI/AReaL/pull/1118
feat(inference_service): add VLM image input support to OpenAI-compatible API by @garrett4wade in https://github.com/inclusionAI/AReaL/pull/1119
feat(utils): add Trackio experiment tracking backend by @guozhihao-224 in https://github.com/inclusionAI/AReaL/pull/1113
refactor(infra): decompose rpc_server into shared guard + blueprints by @garrett4wade in https://github.com/inclusionAI/AReaL/pull/1126
refactor(agents): redesign review-pr taxonomy and sync flow by @rchardx in https://github.com/inclusionAI/AReaL/pull/1124
feat(infra): add client-side fetch buffer for RTensor by @guozhihao-224 in https://github.com/inclusionAI/AReaL/pull/1122
chore: fix gcp image to latest by @nuzant in https://github.com/inclusionAI/AReaL/pull/1130
docs: add Trackio configuration to CLI reference by @rchardx in https://github.com/inclusionAI/AReaL/pull/1131
feat(service): support online inference service by @nuzant in https://github.com/inclusionAI/AReaL/pull/1121
fix(engine): for broken tree training due to bad indent in PR [#1056] by @gursimar in https://github.com/inclusionAI/AReaL/pull/1135
feat(service): add vllm backend support for inference service demo by @nuzant in https://github.com/inclusionAI/AReaL/pull/1136
fix(api): add mode validation for WandBConfig and SwanlabConfig by @guozhihao-224 in https://github.com/inclusionAI/AReaL/pull/1134
feat: enable LoRA RL-training in Megatron via megatron-bridge by @gursimar in https://github.com/inclusionAI/AReaL/pull/1123
fix: harden padded distributed eval across training engines by @rchardx in https://github.com/inclusionAI/AReaL/pull/1109
feat(ci): separate vllm and sglang pyproject.toml by @garrett4wade in https://github.com/inclusionAI/AReaL/pull/1141
fix(vllm_ext): clear multimodal caches after generation pause by @Adiactive in https://github.com/inclusionAI/AReaL/pull/1144
fix(ci): sync uv.vllm.lock with the current pyproject.vllm.toml by @garrett4wade in https://github.com/inclusionAI/AReaL/pull/1146
fix(vllm_ext): XCCL lora weights update when PP>1 by buffering and merging PP shards by @gursimar in https://github.com/inclusionAI/AReaL/pull/1145
chore: fix pre-commit by @garrett4wade in https://github.com/inclusionAI/AReaL/pull/1148
Fix [#1040]: [Feature] Fixed bugs in Archon LoRA Backend by @JiwaniZakir in https://github.com/inclusionAI/AReaL/pull/1139
feat(infra): add distributed data loading service by @garrett4wade in https://github.com/inclusionAI/AReaL/pull/1120
refactor(infra): standardize list-first trajectory batch dispatch by @garrett4wade in https://github.com/inclusionAI/AReaL/pull/1150
feat(infra): allow colocation with offloading and disk weight updates by @garrett4wade in https://github.com/inclusionAI/AReaL/pull/1157
refactor: replace manual JSON parsing with Pydantic models by @koladefaj in https://github.com/inclusionAI/AReaL/pull/1154
fix(engine): FSDP compute_logp fails for Qwen3.5 with dict attention_mask by @pratyush618 in https://github.com/inclusionAI/AReaL/pull/1153
chore: update readme and enforce license by @garrett4wade in https://github.com/inclusionAI/AReaL/pull/1170
chore: ensure SPDX license header in python source files by @garrett4wade in https://github.com/inclusionAI/AReaL/pull/1171
fix: add missing pre-commit check file by @garrett4wade in https://github.com/inclusionAI/AReaL/pull/1173
chore: add project governance for PyTorch ecosystem by @garrett4wade in https://github.com/inclusionAI/AReaL/pull/1174
feat(infra): add microservice-based training service (controller v2) by @garrett4wade in https://github.com/inclusionAI/AReaL/pull/1169
chore: renew qrcode by @garrett4wade in https://github.com/inclusionAI/AReaL/pull/1184
feat(archon): support multi-node inference in gateway controller by @guozhihao-224 in https://github.com/inclusionAI/AReaL/pull/1178
feat(agent-service): add Controller, Guard, and Claude Agent SDK example by @CormickKneey in https://github.com/inclusionAI/AReaL/pull/1177
fix _update_weights_from_disk function to prevent training to be stuck by @asif07hossain in https://github.com/inclusionAI/AReaL/pull/1181
refactor: mount data blueprint via WSGI and adopt Pydantic in engine blueprint by @koladefaj in https://github.com/inclusionAI/AReaL/pull/1179
fix(engine): use meta device for non-rank-0 in FSDP memory_efficient_load by @yulangz in https://github.com/inclusionAI/AReaL/pull/1182
ci: parallelize unit and integration tests across 4 GPU instances by @nuzant in https://github.com/inclusionAI/AReaL/pull/1185
chore: dump v1.0.3 by @garrett4wade in https://github.com/inclusionAI/AReaL/pull/1191