diff --git a/.gitattributes b/.gitattributes index a6344aac8c09253b3b630fb776ae94478aa0275b..02bccef967d452cae26d2e3a7a70d552e53de7ad 100644 --- a/.gitattributes +++ b/.gitattributes @@ -33,3 +33,255 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text *.zip filter=lfs diff=lfs merge=lfs -text *.zst filter=lfs diff=lfs merge=lfs -text *tfevents* filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/101d51ca-2489-4dcb-9c9c-0b75c3de7a68_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/10448c21-e965-4b84-9a05-f028638542f9_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/13dbf2ff-b86e-4912-9b5d-51d7741a5da2_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/18415cf5-cb64-495e-bdfa-0602f4e30928_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/188c1338-ad67-4435-a227-cd60020a8ab0_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/240f9a25-24fb-4962-9ce3-75a840f69015_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/3165e6c7-4d5b-44ea-8cdd-a29560aaa4bc_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/35cef822-45cf-4114-80e4-bc8b991a5372_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/3bc7bfb8-7da3-4fd7-9a74-2233bc6519b2_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/4764ec14-8143-41da-aeb7-970c370b884a_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/4958f259-0d11-49de-993e-fc9640096bcb_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/55767605-ae2c-408f-84e7-9583405c416b_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/595dda36-8986-4ce5-b6f0-46894f9d86fa_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/5df9eaf3-d934-487a-9d4d-7b0f4cda44dd_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/63cfd5dd-581f-45ae-8e27-bfb7e41508e4_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/68723548-2b3c-463a-a393-ef219695cc86_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/69c5eeaa-6a7b-4582-82f0-fda05529e518_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/6e896171-5623-4ae7-b3b8-c0a17421646e_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/6fb913a2-c4d5-4624-8876-9815c31652e0_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/7d7a4592-4977-45aa-bf24-054d1b4e5257_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/81184098-e19f-4754-9371-6a91b9cc59c4_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/8b7d6ef1-fc00-4b03-acf3-be1bf2d619dd_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/8f4011f2-cd38-4eed-8500-5412930bc59b_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/90c0e30a-3c06-4fd2-bfc3-9878cc521972_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/9470dd7b-a889-477f-ad1b-80925b5380ad_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/978497d7-d992-4041-a7a8-766383ed0614_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/a819478c-9e9a-4649-9be5-812483bd5725_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/a89039f7-5f61-468e-acd6-78caaee3ccef_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/a904b30c-da2b-4072-ab20-cab7cf56b41c_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/a9c0a5a1-3085-4a69-997b-119237d326ba_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/accfb691-0159-4fdb-b193-ebbef7e1419a_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/af66aaf4-975b-499c-8bb6-fc87106244f6_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/bbe92f58-9da4-42a9-84ff-04ea95f484c3_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/be7eaa45-5e8d-47d3-98ef-a4f10b852484_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/be9a4740-975d-4e03-9510-0a13a0c42e5d_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/c5a5be2b-057e-433a-8e8f-b7dadbdd2112_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/c6f72900-07ce-4b82-89c7-23ee4c055398_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/ca9adb4a-2a41-49db-b25c-563f4a4b1e47_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/d5274fd9-3a60-4be0-9a57-810127e1f2fb_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/da77b644-5153-4d9c-9f20-a0218f578316_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/dbf32b98-ca20-4971-9b8c-06cedf2effd0_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/dc1a1efd-eb27-4a5e-80be-399453d4b1bc_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/dc5e3c24-2399-40c2-a7ac-b3d89ee8f238_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/e4a80fea-f083-44c3-8e36-466342ffa7fb_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/ed140b40-0544-439b-9fbc-14bbd38f76fd_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/eed2e25d-ce95-40bd-8060-ffc837be657b_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/ef1917dd-f582-4b10-81fc-9127b1338817_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/fae8f20e-c800-4781-9b20-048046fe0b82_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/fcef6b0c-e58e-438c-ba64-89404a838f16_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/ff2fdc77-485e-4b09-9250-d1cdff5a052e_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/0d460c64-f43c-4566-91df-e3b49476f0a6_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/1237575d-1e4a-466a-b462-9d0db98318b2_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/12b2b73b-350c-45a7-8f76-95e62624ea58_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/17efd764-1bfd-428d-90eb-45321f078bc5_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/1bbeca5b-e9df-42ed-b6c0-f2f88c822b40_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/1bc7b7fb-e7de-4261-b376-4de249d181a3_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/2731a4d7-5b36-42f1-bc63-5aaddb0bfc6b_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/2906af99-3d47-452b-a3cc-d18d55ed5214_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/3302dfb8-1277-4a31-9a2c-df6e9b1d419b_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/38c7c055-084b-41fa-9c7d-878f43643157_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/3a6f675d-7929-497e-a00e-68dc8f60251e_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/3a79d536-d135-41d9-b7e4-3b01d88aad69_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/43f733c2-d0eb-4ecb-ba29-1f2aaba734f5_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/4bde81b1-fcbf-4049-9e27-5c2243639c0c_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/4e8aaecc-574c-429a-8fc3-501db2590b10_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/5221791a-be86-4e06-bf80-ad1a34b7f925_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/5af20987-9f8f-4a7c-a222-72c71699ff2c_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/5d06afe6-4bd1-4942-a475-6b3a830a097b_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/5e41b01f-4a01-46b4-be86-9b040c749751_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/5ec96f14-f5d9-4ed6-bfe0-732911bd64f5_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/6265d818-ba49-4248-81d2-9d74a5bd2264_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/6d69e6ee-1432-423c-aa9a-9e0a9889118e_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/6d9cafc4-21b4-4cc3-8ebc-08e527c6b3d7_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/6ee3f707-16ea-4ff5-b22f-a9fd15b1f390_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/75582ad1-113b-4d55-b15b-5fd09a7c4144_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/7b48013f-20cd-46ff-8c77-d74a5b320f42_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/7f708ade-f5a9-412d-aee9-db24a321dc21_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/8706bc77-7f64-460c-bbd4-facf9dcacd73_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/881d64d6-d1c9-495a-a143-715ef3ad8433_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/8a054087-a814-4221-9aa3-2b3f53ee83f0_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/8c0c53f0-c015-42ca-b46c-bb4e9dbfcb7e_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/8c3ec213-1aaf-4483-8c73-10e787b17c25_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/929e3abf-5983-4c3d-be4a-3f0071a1647b_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/976d9ca8-9b22-4843-b025-6b202ddde753_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/9beb2375-83cd-4707-8207-27806510bf95_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/a2ea6475-089f-4ec6-a943-2f5bca11f7e2_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/a880d463-dbd5-4958-b31c-4c49a657a7d7_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/af4ed1bf-bfe0-4429-b717-6addd4bcfaa1_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/b1b6f572-2650-4bf9-a819-ac927adbb376_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/b71ef9d3-1dcd-4386-aa6f-72c763e385a3_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/b8830aaa-6a6c-4399-b9a1-a3a235f1bd28_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/bb249aa5-fb31-4d8e-8405-25259e164004_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/c379b03b-af2a-435f-ae27-ca4df0c65ec8_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/d2185a8b-95fa-43c3-9590-b776ce17fe2d_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/da1620ce-500e-4636-9245-fa1627febdc8_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/e2f23917-265c-4074-becb-04bbbe0b6393_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/e6ef1ba6-69ed-45f8-b959-1e35b579ca6a_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/ed1fab66-134c-49ac-892e-68610768d49f_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/f2884552-c396-40a1-a232-950baa6ced05_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/f5723a46-9587-4d15-8dca-028975b9f56b_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/056b382e-3603-4467-bd04-39698a7a1821_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/0b9b5fc7-ecf3-4625-9344-72c9005d76ea_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/0e18cd09-6c14-4aa4-a6d5-aaba72f81b44_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/0fbc7376-936e-4869-a3ad-d7696f3e7578_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/16098338-bb93-4043-8604-7b6a87071be0_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/1660b1c7-86ac-405a-b4f1-00fe89718635_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/172f088b-3357-4adb-bbaf-def07eea74d3_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/173c83e2-2de6-4b64-944f-b535d5106e30_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/18c9ea8e-d8ca-4846-865a-9a42f055bd40_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/1ad89741-c191-49ff-97db-6d68ec67aae4_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/258129ba-fc32-4df5-aba0-235be8d2947b_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/29ba67c1-2a90-4e8f-b997-01aefa1f19b3_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/2dfa0276-1a8f-4dd5-a070-37389d3b233f_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/35af024a-e166-4e4f-93ab-17c2ca902796_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/45627ac0-43c9-436e-8634-41e03b50cf19_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/45a8f4a5-4717-40c3-b06e-62bbcca0b1af_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/49456225-3437-4b90-a5a0-77a53d3c9bf4_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/49797cdd-56e2-4d09-a7de-8272076a006b_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/4ecac472-c319-4a27-9dd0-748b0383d659_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/58213633-c297-44c4-8ee6-5cdca89d7e27_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/5d0451d0-2465-4267-aa89-854951962bcf_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/5e2b48e9-a9a3-488c-87de-439936b68f2a_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/5f92f264-3f91-49b1-987a-5c62fd8d0ffe_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/6294b8f0-f45c-47dd-bfa7-bed11bb2f4f6_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/62a4567d-2433-4979-a4b3-78bfce8712f3_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/659d5616-40b6-4e06-99d4-3fbe29669b66_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/68cacdf0-44f0-498c-8657-ea0fe622ee1f_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/6ddedd2d-2b5e-478d-a7fb-2c07dd3a8072_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/6f04e8ed-23e1-40f3-a884-8d9f7a5d231f_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/6fc1e119-0a68-4357-9105-84e43cb1b8b6_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/744d8c08-8ebe-40d6-830e-815b75558c33_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/762f530c-f4d1-4b76-bd46-55efc1ff46e3_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/82266441-a3ba-45e0-8df7-94277aea94e9_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/88f8f2dc-ee58-4cdb-98d1-5507a6fe36bc_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/8b642d6c-274d-43ba-ae53-c1ce562a2b39_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/8b6b1ed7-e68a-4033-97d3-2e9db1a1941d_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/8ba087de-e1f1-40c5-814f-b5554fb7ae29_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/a4b1a45c-9f47-4e21-9501-7fbdf396067b_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/b5b7b922-e5a1-41bb-a1bf-2d39120abee9_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/b99111aa-00a2-490b-a5b3-28848e023684_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/ba7e2f63-1851-4c13-9abe-e41ffd4136f7_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/bd11cc91-2a4d-41b8-82e7-2e5d1854693b_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/c5628725-dc96-4d87-8bd8-0f3e484c6055_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/d647f547-4801-4be9-95dd-783fbda55f71_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/da522768-15c3-466c-a2bb-2e024643d960_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/dd52e274-0445-4568-afb1-01cc5a44ecf3_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/dfbf3d09-8c88-445b-a43e-a4ed2d2747a5_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/e73723a4-e418-42ca-8f50-958236d9db4a_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/f67a7b4f-635a-4503-9a21-3084ac41e5ae_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/fbdbc376-b6ce-4007-a788-6cd681db5617_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/0c607c8b-63cf-47bd-91ce-00073fa5f018_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/0def4f5f-a2fe-486f-93e2-0ea64d7a4f92_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/1858e8e1-808b-40e4-8265-b0cd876f755b_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/19037a5a-deb3-42f1-b883-d0355efee8fe_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/269c0890-f58c-4ae9-8009-6cc1b74d65f9_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/2c2bf082-0fe3-4668-9a7d-1e174cce4c0b_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/377e7b1e-c810-4108-b18d-5d4532a765b3_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/39916e11-6a97-4cd7-ac00-ee94c76f3f93_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/3d7f7255-da89-40ea-87ec-f9aa0f85a726_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/4254fea1-a0ac-472f-8db0-1c5e87103151_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/43b3c7b3-4d85-4e4e-be26-6ca378809ccc_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/44ae8f35-4bc5-4e15-b0a2-730d236e865b_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/464e8749-641a-4215-bc5c-c6e59ef12e82_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/4ef16362-02b1-4153-a2ee-dbf2596052ae_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/56328681-8ecf-43e1-a04c-7777a13c072a_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/59167abc-72e6-4f3a-84f6-90342aad3c25_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/5ccb7a7f-d18e-4ec0-9214-b1fd288c2edf_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/615b90b8-d2d2-432e-9712-28571be5533b_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/632388a8-e8e9-4343-b29e-b160940efb02_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/69934ad1-bfe6-4d82-9459-78aafd5051d0_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/6f22b8df-78cd-45c7-9bca-10d7bec2cce3_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/71453733-588f-4aa8-befd-a42d7a029526_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/7a6f9241-d973-46ec-8b06-e3f1fbb955e6_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/7c1c7a0c-324e-41d6-99a0-3cc1215450ff_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/864e2430-d4dc-47a6-ac03-a5d15cc4f715_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/8a948ae6-e0a9-4cfe-b65f-28ad7bc2def9_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/8ab4f62d-0013-436b-8009-6f4ccea79f83_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/8daf50f7-c294-4fa9-83ce-a40196764a1f_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/8f9fbd47-b72f-43c4-ae43-96d911628b1a_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/913d0bb3-8e41-412c-9fce-3a54491a328e_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/91445ce2-23bd-4db0-aa71-e5160b83ede5_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/93b9eb55-1079-4b40-994f-0ca1b6c13d4c_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/9f4affb8-1613-49ff-a7f5-e870a36dfd1d_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/a1e8c8e7-6222-482f-932e-a009fb3dfda1_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/b7328413-2ad1-451b-a0f3-55b9d3db1a27_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/c87f7af8-b22e-4dbe-b331-4ff7a183ec9b_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/cad29f9c-478e-433d-b8cb-9be990c2d4ad_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/ce0a67d2-5d39-4986-8271-f14e1f5e0e04_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/d02f7a23-4f0f-43d6-89bc-e53f413cd589_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/d220af64-021d-40e1-a498-9b9ae90a5b76_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/d31eacc4-d75d-43f8-8bd5-41e4dd451749_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/d6b3292f-b8fb-4deb-811c-430068c3ec6f_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/e33e0359-e02f-4783-9bce-d77de7b33e9b_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/e64205a4-8df6-4e42-81f0-f3e60d351d8d_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/e8f352a2-b67e-432b-86b4-3bd9cabc05b3_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/ee88086c-4b89-4a27-aeec-fb44ad11cdbc_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/f3f320e0-a400-4749-8a00-f1f8fa82fe1a_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/fa194b09-4c89-4eda-81db-0aa71d022c1c_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/fcc36c38-9b5d-44b4-abfb-7ce65d301a2a_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/fd499bc3-9967-48fd-a7a1-ed81edaa822b_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/0404ed0e-1cfb-4c51-987e-0d9b0c42e76d_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/04de685c-871e-4463-a5d9-1871b0ec42b9_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/085542a2-bed7-413f-b5d3-2797b56efc96_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/0ff36697-fcaf-4f2f-988f-ec29ad06af86_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/10336d73-0b18-4944-a859-55fda157cd09_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/114a048b-e874-4167-a996-d46b60499629_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/13664493-b6b4-4dcb-a7d2-6d9c39a73fc0_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/193c8599-681e-4270-aa46-0f3933f35f4d_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/1e6acd95-c63b-41d0-9901-c2bc2930a0e3_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/203fc780-f55b-46f0-98a0-5330f7cf8a87_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/2beb0ca4-ae34-4251-93c9-edc0687b3b83_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/322072ab-d665-4e71-92a2-eb920ba7683d_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/32e19f5b-24ce-48e1-9a45-3ab64fe6fb53_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/36d50f10-ed79-4bce-9705-d030dbd2c256_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/404c1525-1667-4d16-9fbc-84d1475fa530_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/44375c67-9f5e-498d-ac42-762488fe3c24_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/47aacd87-4fb3-423b-9f9d-f565914324ac_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/483a1c02-7093-46e2-8f0b-1c5db8decc5d_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/4dd638f7-1b83-4838-871b-54f35e0027fd_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/555766f1-930f-4bba-ad1d-74100f147a6c_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/5b0e1630-0226-4845-a78b-b5a65c710b85_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/5c4cf8a1-0cdf-450b-ab51-5dc5f496de2f_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/5d837066-5253-4fa3-bb76-5d30447fa4e7_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/6b86827a-a6bc-4c07-88b0-3276ec957134_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/6ef2bdce-7802-4053-a486-32435f36e204_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/6f6c8db9-7572-4a89-8c77-dd0de3f96460_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/6fa97f00-3555-4ca9-91e3-7418039de303_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/7346c976-53cb-47ba-9c4d-3da57a3cb3a6_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/754dec4a-0324-440b-b322-b127c8497b51_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/794e4c04-0975-4c55-9d63-c8011d5831c2_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/7a2d9053-5e19-4951-ba80-b92399a95c30_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/7cbdd96e-a15d-41c3-8929-0b6f71559c35_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/7ed6810d-b4d0-452f-9bf4-e5960595762a_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/906428f0-6a7b-46b8-87f0-e73a09f0b96b_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/908fa40d-de5f-4d66-90d7-020f9daec0f0_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/9fbad204-a7fd-4239-9db5-87f36aecb8c8_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/a517338f-7ad7-41ad-b5ef-09de29bf6144_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/a9ebd8fe-e5fd-4603-8ba9-12af5317a797_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/aeb1be92-8d8d-499b-9fd5-fed6344edf6a_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/b5be42a1-9591-4582-8488-b9c8799dfe5b_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/bd15ea1c-a344-4a9d-a682-4d1212148bf7_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/d01ce947-9ce5-4ff2-a35d-23372bf158a5_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/dae5317b-d777-49a2-ace1-a33437d67f82_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/e0dac2f0-8149-4be2-9426-f8b0263f99e7_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/e32b2b75-ca02-4a77-9b32-36dbb5ff6708_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/e3708ec4-45f5-4655-96f1-641763ba67d4_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/e6c43dc5-17aa-42e2-b0a9-20ccce95bfb9_success0.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/e7253a26-1ea5-4d15-a976-d7d5bf63767a_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/ea13d1c3-77b2-41ea-a27b-a4cd994eeace_success1.mp4 filter=lfs diff=lfs merge=lfs -text +videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/f8a8b99a-b780-47df-a51a-92822914fd0d_success0.mp4 filter=lfs diff=lfs merge=lfs -text +wandb/wandb/run-20260419_111043-fabqp1or/run-fabqp1or.wandb filter=lfs diff=lfs merge=lfs -text +wandb/wandb/run-20260419_111433-oh7yfg1j/run-oh7yfg1j.wandb filter=lfs diff=lfs merge=lfs -text diff --git a/checkpoints/steps_50000_pytorch_model.pt b/checkpoints/steps_50000_pytorch_model.pt new file mode 100644 index 0000000000000000000000000000000000000000..87cc5737a49279585677a0131d64b091cea2998f --- /dev/null +++ b/checkpoints/steps_50000_pytorch_model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2cb4de3db0640479a9fae9879835dc5c0fffbbbf88ef7f7f82ddc856663a8fd6 +size 6959099068 diff --git a/config.json b/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fbc2c1eb79316fd6cbec05ae979fc79a2920799a --- /dev/null +++ b/config.json @@ -0,0 +1,121 @@ +{ + "run_id": "0418_QwenLatent_13tasks_actionstate_30k", + "run_root_dir": "./runs", + "seed": 42, + "trackers": [ + "jsonl", + "wandb" + ], + "wandb_entity": "timsty", + "wandb_project": "vla_jepa", + "is_debug": false, + "framework": { + "name": "QwenLatent", + "qwenvl": { + "base_vlm": "/mnt/data/fangyu/model/Qwen/Qwen3-VL-2B-Instruct", + "attn_implementation": "flash_attention_2", + "vl_hidden_dim": 2048, + "num_data_tokens": 32 + }, + "action_model": { + "ckpt_path": "/mnt/data/fangyu/code/reward_new/runs/0418_Action_13tasks_actionstate_fixchunk15/final_model/pytorch_model.pt", + "action_size": 37, + "state_size": 74, + "use_state": "${datasets.vla_data.state_use_action_chunk}", + "hidden_size": 1024, + "intermediate_size": 3072, + "dataset_vocab_size": 256, + "num_data_tokens": 32, + "num_t_samples": 4, + "min_action_len": 5, + "num_encoder_layers": 28, + "num_decoder_layers": 28, + "num_attention_heads": 16, + "num_key_value_heads": 8, + "head_dim": 128, + "max_position_embeddings": 2048, + "max_action_chunk_size": 50, + "rms_norm_eps": 1e-06, + "attention_dropout": 0.0, + "use_vae_reparameterization": false, + "use_ema": false, + "chunk_size": "${datasets.vla_data.chunk_size}", + "loss_mode": "full", + "qwen3_pretrained_name_or_path": "/mnt/data/fangyu/model/Qwen/Qwen3-0.6B" + } + }, + "datasets": { + "vla_data": { + "dataset_py": "lerobot_datasets", + "data_root_dir": "/mnt/data/fangyu/dataset/IPEC-COMMUNITY", + "data_mix": "cross_embodiedment_13tasks", + "CoT_prompt": "Task: {instruction}. What are the next 15 actions to take?", + "default_image_resolution": [ + 3, + 224, + 224 + ], + "per_device_batch_size": 32, + "load_all_data_for_training": true, + "obs": [ + "image_0" + ], + "image_size": [ + 224, + 224 + ], + "video_backend": "torchcodec", + "load_video": true, + "chunk_size": 15, + "state_use_action_chunk": true, + "num_history_steps": 0, + "include_state": "${datasets.vla_data.state_use_action_chunk}" + } + }, + "trainer": { + "epochs": 100, + "max_train_steps": 50000, + "num_warmup_steps": 5000, + "num_stable_steps": 0, + "mode": "decay_aux_loss", + "loss_weights_decay_steps": 5000, + "save_interval": 5000, + "eval_interval": 50, + "max_checkpoints_to_keep": 20, + "learning_rate": { + "base": 2.5e-05, + "qwen_vl_interface": 2.5e-05, + "action_model": 2.5e-05 + }, + "lr_scheduler_type": "warmup_stable_cosine", + "scheduler_specific_kwargs": { + "min_lr_ratio": 0.001 + }, + "freeze_modules": "", + "loss_scale": { + "align_loss": 1.0, + "recon_loss": 1.0, + "predict_loss": 1.0 + }, + "warmup_ratio": 0.1, + "weight_decay": 0.0, + "logging_frequency": 10, + "gradient_clipping": 5.0, + "gradient_accumulation_steps": 1, + "optimizer": { + "name": "AdamW", + "betas": [ + 0.9, + 0.95 + ], + "eps": 1e-08, + "weight_decay": 1e-08 + }, + "is_resume": false, + "resume_epoch": null, + "resume_step": null, + "enable_gradient_checkpointing": true, + "enable_mixed_precision_training": true + }, + "output_dir": "./runs/0418_QwenLatent_13tasks_actionstate_30k" +} \ No newline at end of file diff --git a/config.yaml b/config.yaml new file mode 100644 index 0000000000000000000000000000000000000000..c826351563e8f9016407e70ac6a2e675d815e7dd --- /dev/null +++ b/config.yaml @@ -0,0 +1,104 @@ +run_id: 0418_QwenLatent_13tasks_actionstate_30k +run_root_dir: ./runs +seed: 42 +trackers: +- jsonl +- wandb +wandb_entity: timsty +wandb_project: vla_jepa +is_debug: false +framework: + name: QwenLatent + qwenvl: + base_vlm: /mnt/data/fangyu/model/Qwen/Qwen3-VL-2B-Instruct + attn_implementation: flash_attention_2 + vl_hidden_dim: 2048 + num_data_tokens: 32 + action_model: + ckpt_path: /mnt/data/fangyu/code/reward_new/runs/0418_Action_13tasks_actionstate_fixchunk15/final_model/pytorch_model.pt + action_size: 37 + state_size: 74 + use_state: ${datasets.vla_data.state_use_action_chunk} + hidden_size: 1024 + intermediate_size: 3072 + dataset_vocab_size: 256 + num_data_tokens: 32 + num_t_samples: 4 + min_action_len: 5 + num_encoder_layers: 28 + num_decoder_layers: 28 + num_attention_heads: 16 + num_key_value_heads: 8 + head_dim: 128 + max_position_embeddings: 2048 + max_action_chunk_size: 50 + rms_norm_eps: 1.0e-06 + attention_dropout: 0.0 + use_vae_reparameterization: false + use_ema: false + chunk_size: ${datasets.vla_data.chunk_size} + loss_mode: full + qwen3_pretrained_name_or_path: /mnt/data/fangyu/model/Qwen/Qwen3-0.6B +datasets: + vla_data: + dataset_py: lerobot_datasets + data_root_dir: /mnt/data/fangyu/dataset/IPEC-COMMUNITY + data_mix: cross_embodiedment_13tasks + CoT_prompt: 'Task: {instruction}. What are the next 15 actions to take?' + default_image_resolution: + - 3 + - 224 + - 224 + per_device_batch_size: 32 + load_all_data_for_training: true + obs: + - image_0 + image_size: + - 224 + - 224 + video_backend: torchcodec + load_video: true + chunk_size: 15 + state_use_action_chunk: true + num_history_steps: 0 + include_state: ${datasets.vla_data.state_use_action_chunk} +trainer: + epochs: 100 + max_train_steps: 50000 + num_warmup_steps: 5000 + num_stable_steps: 0 + mode: decay_aux_loss + loss_weights_decay_steps: 5000 + save_interval: 5000 + eval_interval: 50 + max_checkpoints_to_keep: 20 + learning_rate: + base: 2.5e-05 + qwen_vl_interface: 2.5e-05 + action_model: 2.5e-05 + lr_scheduler_type: warmup_stable_cosine + scheduler_specific_kwargs: + min_lr_ratio: 0.001 + freeze_modules: '' + loss_scale: + align_loss: 1.0 + recon_loss: 1.0 + predict_loss: 1.0 + warmup_ratio: 0.1 + weight_decay: 0.0 + logging_frequency: 10 + gradient_clipping: 5.0 + gradient_accumulation_steps: 1 + optimizer: + name: AdamW + betas: + - 0.9 + - 0.95 + eps: 1.0e-08 + weight_decay: 1.0e-08 + is_resume: false + resume_epoch: null + resume_step: null + enable_gradient_checkpointing: true + enable_mixed_precision_training: true +output_dir: ./runs/0418_QwenLatent_13tasks_actionstate_30k diff --git a/dataset_statistics.json b/dataset_statistics.json new file mode 100644 index 0000000000000000000000000000000000000000..0779626ad2ec4aa54c50dcdbe043afc306f09404 --- /dev/null +++ b/dataset_statistics.json @@ -0,0 +1,682 @@ +{ + "franka": { + "action": { + "mean": [ + 0.028950385935604575, + 0.03594802748411894, + -0.040578972548246384, + -0.00018153275595977897, + 0.002509436290711164, + -0.0015515198931097987, + 0.2097945421934128 + ], + "std": [ + 0.22410766487652278, + 0.24305199620930756, + 0.2951965321738853, + 0.02554494745114589, + 0.04195688852172731, + 0.04913475126163131, + 0.40719537354982976 + ], + "max": [ + 0.9375, + 0.9375, + 0.9375, + 0.3557142913341522, + 0.375, + 0.375, + 1.0 + ], + "min": [ + -0.9375, + -0.9375, + -0.9375, + -0.2582142949104309, + -0.375, + -0.3675000071525574, + 0.0 + ], + "q01": [ + -0.8785714507102966, + -0.8758928775787354, + -0.9375, + -0.1510714292526245, + -0.20678570866584778, + -0.2742857038974762, + 0.0 + ], + "q99": [ + 0.9375, + 0.9107142686843872, + 0.9375, + 0.20357142388820648, + 0.26357144117355347, + 0.375, + 1.0 + ], + "mask": [ + true, + true, + true, + true, + true, + true, + false + ] + }, + "state": { + "mean": [ + -0.019559418968856337, + 0.01475747274234891, + 0.3156160995364189, + 1.1908778190612794, + -0.05669147819280625, + -0.047077450156211856, + 0.010574408061802388, + -0.010662052407860756 + ], + "std": [ + 0.07142409265581258, + 0.097293833441406, + 0.45653463024572627, + 1.4755118780976493, + 0.5249547346172574, + 0.21261447615523038, + 0.01589982024424912, + 0.015944802785650813 + ], + "max": [ + 0.21031762659549713, + 0.39128610491752625, + 1.3660105466842651, + 3.6714255809783936, + 3.560650587081909, + 1.386339545249939, + 0.04233968257904053, + 0.0013633022317662835 + ], + "min": [ + -0.4828203022480011, + -0.3255046010017395, + 0.008128180168569088, + 0.35277295112609863, + -3.641430377960205, + -1.842738389968872, + -0.0013586411951109767, + -0.042040832340717316 + ], + "q01": [ + -0.42401049643754957, + -0.2838300323486328, + 0.009925739830359817, + 1.3085840785503386, + -2.886677579879761, + -1.1599004411697387, + 0.001503719249740243, + -0.040336399003863335 + ], + "q99": [ + 0.1530261474847791, + 0.3629165390133857, + 1.2910678112506866, + 3.303542451858519, + 2.7496529006957933, + 0.6893712210655194, + 0.040610933862626555, + -0.0015016929572448147 + ] + }, + "num_transitions": 273465, + "num_trajectories": 1693 + }, + "gr1": { + "action": { + "mean": [ + -0.06503768935799599, + 0.1402852401137352, + -0.027576334401965143, + -0.7251276493072509, + 0.07538121624384077, + 0.0311446987092495, + -0.0006102971732616395, + -0.1556453585624695, + -0.17398847192525863, + 0.09801423102617265, + -0.7746159434318542, + 0.21945589482784272, + 0.05856966972351074, + 0.14407617151737215, + -0.04729328155517579, + -0.04729328155517579, + -0.04729328155517579, + -0.04729328155517579, + -0.09458656311035157, + 0.30000000000000004, + -0.16327507942914965, + -0.16327507942914965, + -0.16327507942914965, + -0.16327507942914965, + -0.3265501588582993, + 1.5000000000000002, + 0.004927502572536469, + 0.0010833685897523536, + -0.00044836528250016274 + ], + "std": [ + 0.20562330785235147, + 0.18647844645723213, + 0.13514553802530443, + 0.765678360428597, + 0.20251337668667177, + 0.16959756731031508, + 0.19710131907015732, + 0.4583574896238878, + 0.27460068635203594, + 0.3014025883171815, + 0.9046809346526231, + 0.4397149950522067, + 0.29622449978942395, + 0.45108890314809014, + 0.47196951455091185, + 0.47196951455091185, + 0.47196951455091185, + 0.47196951455091185, + 0.9439390291018237, + 0.9, + 1.0480660811801392, + 1.0480660811801392, + 1.0480660811801392, + 1.0480660811801392, + 2.0961321623602784, + 1.4999999999999998, + 0.036936757396854626, + 0.012072528210758757, + 0.007232476061257436 + ], + "max": [ + 1.2965171337127686, + 1.069582223892212, + 0.6721073985099792, + -0.00036245581577531993, + 1.404510259628296, + 1.4991986751556396, + 1.320265293121338, + 1.4943703413009644, + 0.00015122668992262334, + 1.5548107624053955, + 5.365318065742031e-05, + 2.501049041748047, + 1.4930087327957153, + 1.4891765117645264, + 1.5, + 1.5, + 1.5, + 1.5, + 3.0, + 3.0, + 1.5, + 1.5, + 1.5, + 1.5, + 3.0, + 3.0, + 0.3233739137649536, + 0.16282705962657928, + 0.4328489303588867 + ], + "min": [ + -1.2668161392211914, + -0.00042632038821466267, + -1.3582642078399658, + -2.776721477508545, + -1.6229861974716187, + -1.31162691116333, + -1.1806696653366089, + -2.101841926574707, + -1.509853482246399, + -1.2852329015731812, + -2.78891658782959, + -0.9951868653297424, + -1.49818754196167, + -1.4881112575531006, + -1.5, + -1.5, + -1.5, + -1.5, + -3.0, + 0.0, + -1.5, + -1.5, + -1.5, + -1.5, + -3.0, + 3.0, + -0.29814863204956055, + -0.3523012101650238, + -0.09862830489873886 + ], + "q01": [ + -1.1944766759872436, + -4.301190131172918e-07, + -0.7498355388641358, + -2.5581901550292967, + -0.6824928283691406, + -0.8841829299926758, + -1.0078708410263062, + -1.6597670578956605, + -1.226578596830368, + -0.9258902847766877, + -2.6987613677978515, + -0.8498974800109863, + -1.0853091955184937, + -1.0854953002929688, + -1.5, + -1.5, + -1.5, + -1.5, + -3.0, + 0.0, + -1.5, + -1.5, + -1.5, + -1.5, + -3.0, + 3.0, + -0.1551497459411621, + -0.04802883893251419, + -0.029186927154660224 + ], + "q99": [ + 1.0110889196395876, + 0.9781616687774661, + 0.46411060094833406, + -0.07979833036661146, + 1.0625922203063969, + 0.9605619907379178, + 0.6082226157188424, + 1.1673259019851685, + -2.7766001403513115e-06, + 1.0872023344039914, + -0.05551127571612597, + 1.9727541685104366, + 1.0421148896217345, + 1.4869407570362092, + 1.5, + 1.5, + 1.5, + 1.5, + 3.0, + 3.0, + 1.5, + 1.5, + 1.5, + 1.5, + 3.0, + 3.0, + 0.2753815710544586, + 0.06614580899477027, + 0.013670806447044017 + ], + "mask": [ + true, + true, + true, + true, + true, + true, + true, + true, + true, + true, + true, + true, + true, + true, + true, + true, + true, + true, + true, + true, + true, + true, + true, + true, + true, + true, + true, + true, + true + ] + }, + "state": { + "mean": [ + -0.06540636196732522, + 0.14223336279392243, + -0.02882793508470059, + -0.7281232118606568, + 0.07631409014575184, + 0.03485768660902977, + -0.0018693804740905805, + -0.15062302350997925, + -0.17422722429037094, + 0.10088010430335999, + -0.7751174092292785, + 0.21422418504953386, + 0.056977791339159006, + 0.1301754331216216, + 0.0444428395247087, + 0.046001738880295316, + 0.04305497270543129, + 0.04596460912143812, + 0.010766410198993982, + 0.174838250502944, + 0.2807944118976593, + 0.2625684827566147, + 0.24885019063949584, + 0.2602905809879303, + 0.0419048547744751, + 0.8312440872192383, + 0.004883934883400798, + 8.523006690666076e-05, + -6.207618571352216e-05 + ], + "std": [ + 0.20013891174318804, + 0.18784022709302264, + 0.13399923214695403, + 0.7673781334859416, + 0.1990098070412684, + 0.16736959256657702, + 0.19155714110140948, + 0.4472930058630781, + 0.26891556144795553, + 0.28930308542693634, + 0.8909602741089023, + 0.43545203501049284, + 0.2820895843302683, + 0.41341088988690194, + 0.24419357713232692, + 0.22780756672400648, + 0.21366731203921618, + 0.2305132558900519, + 0.06863410938203367, + 0.5112184050495793, + 0.5159579892170897, + 0.47904712336208155, + 0.45710343758297267, + 0.47342494976461513, + 0.13166977346516295, + 0.8450399122937896, + 0.034949910387869945, + 0.011174551728270072, + 0.00744905569932275 + ], + "max": [ + 1.254090428352356, + 1.0467604398727417, + 0.6428823471069336, + 0.0017068678280338645, + 1.3265025615692139, + 1.4839727878570557, + 1.1649909019470215, + 1.4408859014511108, + 0.0030245338566601276, + 1.529702067375183, + 0.0012778707314282656, + 2.4572627544403076, + 1.3810820579528809, + 1.3259196281433105, + 1.7941447496414185, + 1.7346926927566528, + 2.032738208770752, + 1.8508540391921997, + 1.3105334043502808, + 2.0280985832214355, + 2.1240122318267822, + 1.7788946628570557, + 2.984133005142212, + 2.44663405418396, + 1.2877141237258911, + 2.037182092666626, + 0.30236849188804626, + 0.1535278707742691, + 0.38705629110336304 + ], + "min": [ + -1.2302747964859009, + -0.008526407182216644, + -1.3281980752944946, + -2.5072484016418457, + -1.5798313617706299, + -1.1906307935714722, + -1.0882562398910522, + -2.059584379196167, + -1.475722312927246, + -1.2510409355163574, + -2.5155985355377197, + -0.9702507257461548, + -1.463935375213623, + -1.2635116577148438, + -1.1052463054656982, + -0.8339442014694214, + -1.2814871072769165, + -1.6386417150497437, + -0.816240131855011, + -0.5184589624404907, + -1.224429726600647, + -0.9850000739097595, + -0.9763456583023071, + -0.500698447227478, + -0.7500660419464111, + -0.03356735035777092, + -0.26449069380760193, + -0.3266247510910034, + -0.09658436477184296 + ], + "q01": [ + -1.1889611959457398, + -0.0009903528727591038, + -0.7342327952384948, + -2.4300240993499758, + -0.6584946513175964, + -0.8250447034835815, + -0.9737806677818298, + -1.6477210569381713, + -1.2134740138053894, + -0.8785631865262985, + -2.505193281173706, + -0.8399241447448731, + -1.0055391049385072, + -0.9640253639221191, + -0.47676344215869904, + -0.08676981404423714, + -0.14494537323713302, + -0.059164867363870145, + -0.1558390498161316, + -0.02249552313238381, + -0.0051496786996722225, + -0.005549988709390163, + -0.006174665316939354, + -0.005817336589097976, + -0.5121413707733155, + 0.3231884375214577, + -0.1299474745988846, + -0.045216087996959684, + -0.030235426500439644 + ], + "q99": [ + 0.9937156438827516, + 0.9744037270545968, + 0.4551989555358894, + -0.09210347831249224, + 1.0528851509094246, + 0.9591211795806898, + 0.5920819044113171, + 1.1361155986785891, + 0.0001814767369069183, + 1.052486894130706, + -0.049791696146131215, + 1.9595401525497436, + 0.9932814061641682, + 1.241852431297302, + 1.502415704727173, + 1.5030960321426392, + 1.4880860090255776, + 1.5015064239501954, + 0.6326038002967835, + 1.884292006492615, + 1.5134866952896122, + 1.5681370496749878, + 2.706826210021973, + 1.7132962703704837, + 0.8768314361572273, + 1.832364867925644, + 0.2745771288871766, + 0.06204778626561166, + 0.04224991366267195 + ] + }, + "num_transitions": 248956, + "num_trajectories": 1000 + }, + "real_world_franka": { + "action": { + "mean": [ + 0.0018022622913122178, + 0.00866750255227089, + -0.0023813677951693537, + 0.006971974670886994, + -0.00047582909464836124, + 0.0025789525359869005, + -0.0007287326734513044, + 0.035997286438941956 + ], + "std": [ + 0.024339838465266404, + 0.06490750395949492, + 0.021401284949155275, + 0.05111120007770312, + 0.037778529131822655, + 0.05658410498726164, + 0.06271033042800869, + 0.17357673463760007 + ], + "max": [ + 0.21409234404563904, + 0.6854990124702454, + 0.30057650804519653, + 0.7143238186836243, + 0.6867426037788391, + 0.5510483980178833, + 0.5336411595344543, + 1.0 + ], + "min": [ + -0.2338617593050003, + -0.558387279510498, + -0.2482951432466507, + -0.6626467704772949, + -0.47793203592300415, + -0.584820032119751, + -0.7077043652534485, + 0.0 + ], + "q01": [ + -0.1346631145477295, + -0.3504452162981033, + -0.15663981318473816, + -0.24537096738815306, + -0.22821950614452363, + -0.42980227649211883, + -0.4895046579837799, + 0.0 + ], + "q99": [ + 0.1725029143691063, + 0.4685746806859974, + 0.10869869187474263, + 0.499900839328766, + 0.28230221629142765, + 0.3843919765949251, + 0.38130330502986914, + 1.0 + ], + "mask": [ + true, + true, + true, + true, + true, + true, + true, + false + ] + }, + "state": { + "mean": [ + 0.00031761352438479666, + 0.0020736783742904665, + 0.0034826237708330157, + -0.234240460395813, + -0.0018191402778029442, + 0.2424022436141968, + -0.001537260226905346, + 0.030011588335037233 + ], + "std": [ + 0.024077303105278742, + 0.08790303996624453, + 0.027178490734489504, + 0.7067546375834215, + 0.027173463008725068, + 0.7303538119052045, + 0.06428378288303241, + 0.1521238095109371 + ], + "max": [ + 0.2896324098110199, + 0.5507338047027588, + 0.31784605979919434, + -1.5832326412200928, + 0.3542541265487671, + 3.033360481262207, + 0.6994425058364868, + 1.0 + ], + "min": [ + -0.22138147056102753, + -0.8714362382888794, + -0.3488052189350128, + -2.7835195064544678, + -0.34267839789390564, + 1.3950575590133667, + -0.6863359808921814, + 0.0 + ], + "q01": [ + -0.18218472093343735, + -0.6171795415878296, + -0.13265437811613082, + -2.723349404335022, + -0.25293527901172635, + 1.8917096638679505, + -0.41360526919364926, + 0.0 + ], + "q99": [ + 0.23308583796024324, + 0.4467512100934983, + 0.26850769817829134, + -1.8047044682502746, + 0.1607170441746712, + 2.883582048416138, + 0.5439058530330658, + 1.0 + ] + }, + "num_transitions": 38179, + "num_trajectories": 202 + } +} \ No newline at end of file diff --git a/final_model/pytorch_model.pt b/final_model/pytorch_model.pt new file mode 100644 index 0000000000000000000000000000000000000000..bae1e3cea26e93ec41e662b652f9d022df987b91 --- /dev/null +++ b/final_model/pytorch_model.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:95913d305f70daa152111a6520c51ec8244477790760da69d7983e3071ddfec2 +size 6959082408 diff --git a/final_model/pytorch_model.pt.log/eval_20260420_064210/eval_env_gr1_unified_PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env_gpu0.log b/final_model/pytorch_model.pt.log/eval_20260420_064210/eval_env_gr1_unified_PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env_gpu0.log new file mode 100644 index 0000000000000000000000000000000000000000..5e8d0ea517ce9c1bb5c7d71a8b6c2e0a3c270f76 --- /dev/null +++ b/final_model/pytorch_model.pt.log/eval_20260420_064210/eval_env_gr1_unified_PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env_gpu0.log @@ -0,0 +1,63 @@ +[robosuite WARNING] No private macro file found! (macros.py:53) +[robosuite WARNING] It is recommended to use a private macro file (macros.py:54) +[robosuite WARNING] To setup, run: python /mnt/data/fangyu/code/github/robosuite/robosuite/scripts/setup_macros.py (macros.py:55) +[robosuite WARNING] Could not import robosuite_models. Some robots may not be available. If you want to use these robots, please install robosuite_models from source (https://github.com/ARISE-Initiative/robosuite_models) or through pip install. (__init__.py:30) +[robosuite WARNING] No private macro file found! (macros.py:28) +[robosuite WARNING] It is recommended to use a private macro file (macros.py:29) +[robosuite WARNING] To setup, run: python /mnt/data/fangyu/code/github/robocasa-gr1-tabletop-tasks/robocasa/scripts/setup_macros.py (macros.py:30) +/mnt/data/.cache/conda/envs/robocasa/lib/python3.10/site-packages/albumentations/__init__.py:13: UserWarning: A new version of Albumentations is available: 2.0.8 (you have 1.4.18). Upgrade using: pip install -U albumentations. To disable automatic update checks, set the environment variable NO_ALBUMENTATIONS_UPDATE to 1. + check_for_updates() +WARNING: mimicgen environments not imported since mimicgen is not installed! +🚨 `examples` is part of ActionModel.forward's signature, but not documented. Make sure to add it to the docstring of the function in /mnt/data/fangyu/code/reward_new/starVLA/model/modules/action_model/ActionModel.py. +04/20 [06:43:37] INFO | >> Arguments: { simulation_env.py:300 + "host": "127.0.0.1", + "port": 6200, + "env_name": + "gr1_unified/PnPBottleToCabinetC + lose_GR1ArmsAndWaistFourierHands + _Env", + "n_episodes": 50, + "n_envs": 1, + "max_episode_steps": 720, + "n_action_steps": 10, + "video_out_path": + "/mnt/data/fangyu/code/reward_ne + w/runs/0418_QwenLatent_13tasks_a + ctionstate_50k/videos/pytorch_mo + del/n_action_steps_10_max_episod + e_steps_720_n_envs_1_gr1_unified + /PnPBottleToCabinetClose_GR1Arms + AndWaistFourierHands_Env", + "seed": 21, + "pretrained_path": + "/mnt/data/fangyu/code/reward_ne + w/runs/0418_QwenLatent_13tasks_a + ctionstate_50k/final_model/pytor + ch_model.pt" + } + INFO | >> Waiting for server websocket_policy_client.py:34 + at + ws://127.0.0.1:6200... +*** policy_setup: gr1, unnorm_key: gr1 *** + INFO | >> [*] Loading from local share_tools.py:276 + checkpoint path + `/mnt/data/fangyu/code/reward_new/r + uns/0418_QwenLatent_13tasks_actions + tate_50k/final_model/pytorch_model. + pt` +Running 50 episodes for gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env with 1 environments +04/20 [06:43:40] INFO | >> No OpenGL_accelerate acceleratesupport.py:24 + module loaded: No module named + 'OpenGL_accelerate' +/mnt/data/.cache/conda/envs/robocasa/lib/python3.10/site-packages/gymnasium/utils/passive_env_checker.py:158: UserWarning: WARN: The obs returned by the `reset()` method is not within the observation space. + logger.warn(f"{pre} is not within the observation space.") +/mnt/data/.cache/conda/envs/robocasa/lib/python3.10/site-packages/gymnasium/utils/passive_env_checker.py:158: UserWarning: WARN: The obs returned by the `step()` method is not within the observation space. + logger.warn(f"{pre} is not within the observation space.") +WARNING: _load_model has been called over 10 times! +WARNING: _load_model has been called over 20 times! +WARNING: _load_model has been called over 30 times! +WARNING: _load_model has been called over 40 times! +WARNING: _load_model has been called over 50 times! +Collecting 50 episodes took 3741.68 seconds +Results for gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env: +Success rate: 0.68 diff --git a/final_model/pytorch_model.pt.log/eval_20260420_064210/eval_env_gr1_unified_PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env_gpu1.log b/final_model/pytorch_model.pt.log/eval_20260420_064210/eval_env_gr1_unified_PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env_gpu1.log new file mode 100644 index 0000000000000000000000000000000000000000..8b5a8161a2482944c292df327df82cf6d81bdf8d --- /dev/null +++ b/final_model/pytorch_model.pt.log/eval_20260420_064210/eval_env_gr1_unified_PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env_gpu1.log @@ -0,0 +1,64 @@ +[robosuite WARNING] No private macro file found! (macros.py:53) +[robosuite WARNING] It is recommended to use a private macro file (macros.py:54) +[robosuite WARNING] To setup, run: python /mnt/data/fangyu/code/github/robosuite/robosuite/scripts/setup_macros.py (macros.py:55) +[robosuite WARNING] Could not import robosuite_models. Some robots may not be available. If you want to use these robots, please install robosuite_models from source (https://github.com/ARISE-Initiative/robosuite_models) or through pip install. (__init__.py:30) +[robosuite WARNING] No private macro file found! (macros.py:28) +[robosuite WARNING] It is recommended to use a private macro file (macros.py:29) +[robosuite WARNING] To setup, run: python /mnt/data/fangyu/code/github/robocasa-gr1-tabletop-tasks/robocasa/scripts/setup_macros.py (macros.py:30) +/mnt/data/.cache/conda/envs/robocasa/lib/python3.10/site-packages/albumentations/__init__.py:13: UserWarning: A new version of Albumentations is available: 2.0.8 (you have 1.4.18). Upgrade using: pip install -U albumentations. To disable automatic update checks, set the environment variable NO_ALBUMENTATIONS_UPDATE to 1. + check_for_updates() +WARNING: mimicgen environments not imported since mimicgen is not installed! +🚨 `examples` is part of ActionModel.forward's signature, but not documented. Make sure to add it to the docstring of the function in /mnt/data/fangyu/code/reward_new/starVLA/model/modules/action_model/ActionModel.py. +04/20 [06:43:39] INFO | >> Arguments: { simulation_env.py:300 + "host": "127.0.0.1", + "port": 6201, + "env_name": + "gr1_unified/PosttrainPnPNovelFr + omCuttingboardToBasketSplitA_GR1 + ArmsAndWaistFourierHands_Env", + "n_episodes": 50, + "n_envs": 1, + "max_episode_steps": 720, + "n_action_steps": 10, + "video_out_path": + "/mnt/data/fangyu/code/reward_ne + w/runs/0418_QwenLatent_13tasks_a + ctionstate_50k/videos/pytorch_mo + del/n_action_steps_10_max_episod + e_steps_720_n_envs_1_gr1_unified + /PosttrainPnPNovelFromCuttingboa + rdToBasketSplitA_GR1ArmsAndWaist + FourierHands_Env", + "seed": 21, + "pretrained_path": + "/mnt/data/fangyu/code/reward_ne + w/runs/0418_QwenLatent_13tasks_a + ctionstate_50k/final_model/pytor + ch_model.pt" + } + INFO | >> Waiting for server websocket_policy_client.py:34 + at + ws://127.0.0.1:6201... +*** policy_setup: gr1, unnorm_key: gr1 *** + INFO | >> [*] Loading from local share_tools.py:276 + checkpoint path + `/mnt/data/fangyu/code/reward_new/r + uns/0418_QwenLatent_13tasks_actions + tate_50k/final_model/pytorch_model. + pt` +Running 50 episodes for gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env with 1 environments +04/20 [06:43:42] INFO | >> No OpenGL_accelerate acceleratesupport.py:24 + module loaded: No module named + 'OpenGL_accelerate' +/mnt/data/.cache/conda/envs/robocasa/lib/python3.10/site-packages/gymnasium/utils/passive_env_checker.py:158: UserWarning: WARN: The obs returned by the `reset()` method is not within the observation space. + logger.warn(f"{pre} is not within the observation space.") +/mnt/data/.cache/conda/envs/robocasa/lib/python3.10/site-packages/gymnasium/utils/passive_env_checker.py:158: UserWarning: WARN: The obs returned by the `step()` method is not within the observation space. + logger.warn(f"{pre} is not within the observation space.") +WARNING: _load_model has been called over 10 times! +WARNING: _load_model has been called over 20 times! +WARNING: _load_model has been called over 30 times! +WARNING: _load_model has been called over 40 times! +WARNING: _load_model has been called over 50 times! +Collecting 50 episodes took 3805.52 seconds +Results for gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env: +Success rate: 0.40 diff --git a/final_model/pytorch_model.pt.log/eval_20260420_064210/eval_env_gr1_unified_PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env_gpu2.log b/final_model/pytorch_model.pt.log/eval_20260420_064210/eval_env_gr1_unified_PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env_gpu2.log new file mode 100644 index 0000000000000000000000000000000000000000..087209c7a339d8787b04fbcbbdc2647dd4a9d903 --- /dev/null +++ b/final_model/pytorch_model.pt.log/eval_20260420_064210/eval_env_gr1_unified_PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env_gpu2.log @@ -0,0 +1,65 @@ +[robosuite WARNING] No private macro file found! (macros.py:53) +[robosuite WARNING] It is recommended to use a private macro file (macros.py:54) +[robosuite WARNING] To setup, run: python /mnt/data/fangyu/code/github/robosuite/robosuite/scripts/setup_macros.py (macros.py:55) +[robosuite WARNING] Could not import robosuite_models. Some robots may not be available. If you want to use these robots, please install robosuite_models from source (https://github.com/ARISE-Initiative/robosuite_models) or through pip install. (__init__.py:30) +[robosuite WARNING] No private macro file found! (macros.py:28) +[robosuite WARNING] It is recommended to use a private macro file (macros.py:29) +[robosuite WARNING] To setup, run: python /mnt/data/fangyu/code/github/robocasa-gr1-tabletop-tasks/robocasa/scripts/setup_macros.py (macros.py:30) +/mnt/data/.cache/conda/envs/robocasa/lib/python3.10/site-packages/albumentations/__init__.py:13: UserWarning: A new version of Albumentations is available: 2.0.8 (you have 1.4.18). Upgrade using: pip install -U albumentations. To disable automatic update checks, set the environment variable NO_ALBUMENTATIONS_UPDATE to 1. + check_for_updates() +WARNING: mimicgen environments not imported since mimicgen is not installed! +🚨 `examples` is part of ActionModel.forward's signature, but not documented. Make sure to add it to the docstring of the function in /mnt/data/fangyu/code/reward_new/starVLA/model/modules/action_model/ActionModel.py. +04/20 [06:43:41] INFO | >> Arguments: { simulation_env.py:300 + "host": "127.0.0.1", + "port": 6202, + "env_name": + "gr1_unified/PosttrainPnPNovelFr + omPlacematToBasketSplitA_GR1Arms + AndWaistFourierHands_Env", + "n_episodes": 50, + "n_envs": 1, + "max_episode_steps": 720, + "n_action_steps": 10, + "video_out_path": + "/mnt/data/fangyu/code/reward_ne + w/runs/0418_QwenLatent_13tasks_a + ctionstate_50k/videos/pytorch_mo + del/n_action_steps_10_max_episod + e_steps_720_n_envs_1_gr1_unified + /PosttrainPnPNovelFromPlacematTo + BasketSplitA_GR1ArmsAndWaistFour + ierHands_Env", + "seed": 21, + "pretrained_path": + "/mnt/data/fangyu/code/reward_ne + w/runs/0418_QwenLatent_13tasks_a + ctionstate_50k/final_model/pytor + ch_model.pt" + } + INFO | >> Waiting for server websocket_policy_client.py:34 + at + ws://127.0.0.1:6202... +*** policy_setup: gr1, unnorm_key: gr1 *** + INFO | >> [*] Loading from local share_tools.py:276 + checkpoint path + `/mnt/data/fangyu/code/reward_new/r + uns/0418_QwenLatent_13tasks_actions + tate_50k/final_model/pytorch_model. + pt` +Running 50 episodes for gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env with 1 environments +04/20 [06:43:47] INFO | >> No OpenGL_accelerate acceleratesupport.py:24 + module loaded: No module named + 'OpenGL_accelerate' +/mnt/data/.cache/conda/envs/robocasa/lib/python3.10/site-packages/gymnasium/utils/passive_env_checker.py:158: UserWarning: WARN: The obs returned by the `reset()` method is not within the observation space. + logger.warn(f"{pre} is not within the observation space.") +/mnt/data/.cache/conda/envs/robocasa/lib/python3.10/site-packages/gymnasium/utils/passive_env_checker.py:158: UserWarning: WARN: The obs returned by the `step()` method is not within the observation space. + logger.warn(f"{pre} is not within the observation space.") +WARNING: _load_model has been called over 10 times! +WARNING: _load_model has been called over 20 times! +WARNING: _load_model has been called over 30 times! +WARNING: _load_model has been called over 40 times! +WARNING: _load_model has been called over 50 times! +WARNING: _load_model has been called over 60 times! +Collecting 50 episodes took 3732.41 seconds +Results for gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env: +Success rate: 0.34 diff --git a/final_model/pytorch_model.pt.log/eval_20260420_064210/eval_env_gr1_unified_PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env_gpu3.log b/final_model/pytorch_model.pt.log/eval_20260420_064210/eval_env_gr1_unified_PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env_gpu3.log new file mode 100644 index 0000000000000000000000000000000000000000..1ba230647b2c21628b33169e79b4f5eb49a52d05 --- /dev/null +++ b/final_model/pytorch_model.pt.log/eval_20260420_064210/eval_env_gr1_unified_PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env_gpu3.log @@ -0,0 +1,64 @@ +[robosuite WARNING] No private macro file found! (macros.py:53) +[robosuite WARNING] It is recommended to use a private macro file (macros.py:54) +[robosuite WARNING] To setup, run: python /mnt/data/fangyu/code/github/robosuite/robosuite/scripts/setup_macros.py (macros.py:55) +[robosuite WARNING] Could not import robosuite_models. Some robots may not be available. If you want to use these robots, please install robosuite_models from source (https://github.com/ARISE-Initiative/robosuite_models) or through pip install. (__init__.py:30) +[robosuite WARNING] No private macro file found! (macros.py:28) +[robosuite WARNING] It is recommended to use a private macro file (macros.py:29) +[robosuite WARNING] To setup, run: python /mnt/data/fangyu/code/github/robocasa-gr1-tabletop-tasks/robocasa/scripts/setup_macros.py (macros.py:30) +/mnt/data/.cache/conda/envs/robocasa/lib/python3.10/site-packages/albumentations/__init__.py:13: UserWarning: A new version of Albumentations is available: 2.0.8 (you have 1.4.18). Upgrade using: pip install -U albumentations. To disable automatic update checks, set the environment variable NO_ALBUMENTATIONS_UPDATE to 1. + check_for_updates() +WARNING: mimicgen environments not imported since mimicgen is not installed! +🚨 `examples` is part of ActionModel.forward's signature, but not documented. Make sure to add it to the docstring of the function in /mnt/data/fangyu/code/reward_new/starVLA/model/modules/action_model/ActionModel.py. +04/20 [06:43:43] INFO | >> Arguments: { simulation_env.py:300 + "host": "127.0.0.1", + "port": 6203, + "env_name": + "gr1_unified/PosttrainPnPNovelFr + omPlateToBowlSplitA_GR1ArmsAndWa + istFourierHands_Env", + "n_episodes": 50, + "n_envs": 1, + "max_episode_steps": 720, + "n_action_steps": 10, + "video_out_path": + "/mnt/data/fangyu/code/reward_ne + w/runs/0418_QwenLatent_13tasks_a + ctionstate_50k/videos/pytorch_mo + del/n_action_steps_10_max_episod + e_steps_720_n_envs_1_gr1_unified + /PosttrainPnPNovelFromPlateToBow + lSplitA_GR1ArmsAndWaistFourierHa + nds_Env", + "seed": 21, + "pretrained_path": + "/mnt/data/fangyu/code/reward_ne + w/runs/0418_QwenLatent_13tasks_a + ctionstate_50k/final_model/pytor + ch_model.pt" + } + INFO | >> Waiting for server websocket_policy_client.py:34 + at + ws://127.0.0.1:6203... +*** policy_setup: gr1, unnorm_key: gr1 *** + INFO | >> [*] Loading from local share_tools.py:276 + checkpoint path + `/mnt/data/fangyu/code/reward_new/r + uns/0418_QwenLatent_13tasks_actions + tate_50k/final_model/pytorch_model. + pt` +Running 50 episodes for gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env with 1 environments +04/20 [06:43:48] INFO | >> No OpenGL_accelerate acceleratesupport.py:24 + module loaded: No module named + 'OpenGL_accelerate' +/mnt/data/.cache/conda/envs/robocasa/lib/python3.10/site-packages/gymnasium/utils/passive_env_checker.py:158: UserWarning: WARN: The obs returned by the `reset()` method is not within the observation space. + logger.warn(f"{pre} is not within the observation space.") +/mnt/data/.cache/conda/envs/robocasa/lib/python3.10/site-packages/gymnasium/utils/passive_env_checker.py:158: UserWarning: WARN: The obs returned by the `step()` method is not within the observation space. + logger.warn(f"{pre} is not within the observation space.") +WARNING: _load_model has been called over 10 times! +WARNING: _load_model has been called over 20 times! +WARNING: _load_model has been called over 30 times! +WARNING: _load_model has been called over 40 times! +WARNING: _load_model has been called over 50 times! +Collecting 50 episodes took 3809.23 seconds +Results for gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env: +Success rate: 0.20 diff --git a/final_model/pytorch_model.pt.log/eval_20260420_064210/eval_env_gr1_unified_PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env_gpu4.log b/final_model/pytorch_model.pt.log/eval_20260420_064210/eval_env_gr1_unified_PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env_gpu4.log new file mode 100644 index 0000000000000000000000000000000000000000..affe7b9748d89ad577bf3372033a580036f59d02 --- /dev/null +++ b/final_model/pytorch_model.pt.log/eval_20260420_064210/eval_env_gr1_unified_PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env_gpu4.log @@ -0,0 +1,55 @@ +[robosuite WARNING] No private macro file found! (macros.py:53) +[robosuite WARNING] It is recommended to use a private macro file (macros.py:54) +[robosuite WARNING] To setup, run: python /mnt/data/fangyu/code/github/robosuite/robosuite/scripts/setup_macros.py (macros.py:55) +[robosuite WARNING] Could not import robosuite_models. Some robots may not be available. If you want to use these robots, please install robosuite_models from source (https://github.com/ARISE-Initiative/robosuite_models) or through pip install. (__init__.py:30) +[robosuite WARNING] No private macro file found! (macros.py:28) +[robosuite WARNING] It is recommended to use a private macro file (macros.py:29) +[robosuite WARNING] To setup, run: python /mnt/data/fangyu/code/github/robocasa-gr1-tabletop-tasks/robocasa/scripts/setup_macros.py (macros.py:30) +/mnt/data/.cache/conda/envs/robocasa/lib/python3.10/site-packages/albumentations/__init__.py:13: UserWarning: A new version of Albumentations is available: 2.0.8 (you have 1.4.18). Upgrade using: pip install -U albumentations. To disable automatic update checks, set the environment variable NO_ALBUMENTATIONS_UPDATE to 1. + check_for_updates() +WARNING: mimicgen environments not imported since mimicgen is not installed! +🚨 `examples` is part of ActionModel.forward's signature, but not documented. Make sure to add it to the docstring of the function in /mnt/data/fangyu/code/reward_new/starVLA/model/modules/action_model/ActionModel.py. +04/20 [06:43:45] INFO | >> Arguments: { simulation_env.py:300 + "host": "127.0.0.1", + "port": 6204, + "env_name": + "gr1_unified/PosttrainPnPNovelFromTrayToCardboard + boxSplitA_GR1ArmsAndWaistFourierHands_Env", + "n_episodes": 50, + "n_envs": 1, + "max_episode_steps": 720, + "n_action_steps": 10, + "video_out_path": + "/mnt/data/fangyu/code/reward_new/runs/0418_QwenL + atent_13tasks_actionstate_50k/videos/pytorch_mode + l/n_action_steps_10_max_episode_steps_720_n_envs_ + 1_gr1_unified/PosttrainPnPNovelFromTrayToCardboar + dboxSplitA_GR1ArmsAndWaistFourierHands_Env", + "seed": 21, + "pretrained_path": + "/mnt/data/fangyu/code/reward_new/runs/0418_QwenL + atent_13tasks_actionstate_50k/final_model/pytorch + _model.pt" + } + INFO | >> Waiting for server at websocket_policy_client.py:34 + ws://127.0.0.1:6204... +*** policy_setup: gr1, unnorm_key: gr1 *** + INFO | >> [*] Loading from local checkpoint path share_tools.py:276 + `/mnt/data/fangyu/code/reward_new/runs/0418_QwenLate + nt_13tasks_actionstate_50k/final_model/pytorch_model + .pt` +Running 50 episodes for gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env with 1 environments +04/20 [06:43:48] INFO | >> No OpenGL_accelerate module loaded: No acceleratesupport.py:24 + module named 'OpenGL_accelerate' +/mnt/data/.cache/conda/envs/robocasa/lib/python3.10/site-packages/gymnasium/utils/passive_env_checker.py:158: UserWarning: WARN: The obs returned by the `reset()` method is not within the observation space. + logger.warn(f"{pre} is not within the observation space.") +/mnt/data/.cache/conda/envs/robocasa/lib/python3.10/site-packages/gymnasium/utils/passive_env_checker.py:158: UserWarning: WARN: The obs returned by the `step()` method is not within the observation space. + logger.warn(f"{pre} is not within the observation space.") +WARNING: _load_model has been called over 10 times! +WARNING: _load_model has been called over 20 times! +WARNING: _load_model has been called over 30 times! +WARNING: _load_model has been called over 40 times! +WARNING: _load_model has been called over 50 times! +Collecting 50 episodes took 3525.53 seconds +Results for gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env: +Success rate: 0.40 diff --git a/final_model/pytorch_model.pt.log/eval_20260420_064210/server_gpu0_port6200.log b/final_model/pytorch_model.pt.log/eval_20260420_064210/server_gpu0_port6200.log new file mode 100644 index 0000000000000000000000000000000000000000..59a4f99d91b1cf480c3d59fefa7b0a0d23c4cc88 --- /dev/null +++ b/final_model/pytorch_model.pt.log/eval_20260420_064210/server_gpu0_port6200.log @@ -0,0 +1,10 @@ +/mnt/data/.cache/conda/envs/vla_2/lib/python3.10/site-packages/albumentations/__init__.py:13: UserWarning: A new version of Albumentations is available: 2.0.8 (you have 1.4.18). Upgrade using: pip install -U albumentations. To disable automatic update checks, set the environment variable NO_ALBUMENTATIONS_UPDATE to 1. + check_for_updates() +INFO:starVLA.model.framework.share_tools:[*] Loading from local checkpoint path `/mnt/data/fangyu/code/reward_new/runs/0418_QwenLatent_13tasks_actionstate_50k/final_model/pytorch_model.pt` +`torch_dtype` is deprecated! Use `dtype` instead! +INFO:root:Creating server (host: 10-116-218-71, ip: 169.254.95.120) +INFO:root:server running ... +INFO:websockets.server:server listening on 0.0.0.0:6200 +INFO:websockets.server:connection open +INFO:root:Connection from ('127.0.0.1', 54562) opened +INFO:root:Connection from ('127.0.0.1', 54562) closed diff --git a/final_model/pytorch_model.pt.log/eval_20260420_064210/server_gpu1_port6201.log b/final_model/pytorch_model.pt.log/eval_20260420_064210/server_gpu1_port6201.log new file mode 100644 index 0000000000000000000000000000000000000000..f3151ef98e87594f9bc4d69c898099471889505d --- /dev/null +++ b/final_model/pytorch_model.pt.log/eval_20260420_064210/server_gpu1_port6201.log @@ -0,0 +1,10 @@ +/mnt/data/.cache/conda/envs/vla_2/lib/python3.10/site-packages/albumentations/__init__.py:13: UserWarning: A new version of Albumentations is available: 2.0.8 (you have 1.4.18). Upgrade using: pip install -U albumentations. To disable automatic update checks, set the environment variable NO_ALBUMENTATIONS_UPDATE to 1. + check_for_updates() +INFO:starVLA.model.framework.share_tools:[*] Loading from local checkpoint path `/mnt/data/fangyu/code/reward_new/runs/0418_QwenLatent_13tasks_actionstate_50k/final_model/pytorch_model.pt` +`torch_dtype` is deprecated! Use `dtype` instead! +INFO:root:Creating server (host: 10-116-218-71, ip: 169.254.95.120) +INFO:root:server running ... +INFO:websockets.server:server listening on 0.0.0.0:6201 +INFO:websockets.server:connection open +INFO:root:Connection from ('127.0.0.1', 54086) opened +INFO:root:Connection from ('127.0.0.1', 54086) closed diff --git a/final_model/pytorch_model.pt.log/eval_20260420_064210/server_gpu2_port6202.log b/final_model/pytorch_model.pt.log/eval_20260420_064210/server_gpu2_port6202.log new file mode 100644 index 0000000000000000000000000000000000000000..f9964938d613c16bd4ddd476e1da210bf88577be --- /dev/null +++ b/final_model/pytorch_model.pt.log/eval_20260420_064210/server_gpu2_port6202.log @@ -0,0 +1,10 @@ +/mnt/data/.cache/conda/envs/vla_2/lib/python3.10/site-packages/albumentations/__init__.py:13: UserWarning: A new version of Albumentations is available: 2.0.8 (you have 1.4.18). Upgrade using: pip install -U albumentations. To disable automatic update checks, set the environment variable NO_ALBUMENTATIONS_UPDATE to 1. + check_for_updates() +INFO:starVLA.model.framework.share_tools:[*] Loading from local checkpoint path `/mnt/data/fangyu/code/reward_new/runs/0418_QwenLatent_13tasks_actionstate_50k/final_model/pytorch_model.pt` +`torch_dtype` is deprecated! Use `dtype` instead! +INFO:root:Creating server (host: 10-116-218-71, ip: 169.254.95.120) +INFO:root:server running ... +INFO:websockets.server:server listening on 0.0.0.0:6202 +INFO:websockets.server:connection open +INFO:root:Connection from ('127.0.0.1', 33612) opened +INFO:root:Connection from ('127.0.0.1', 33612) closed diff --git a/final_model/pytorch_model.pt.log/eval_20260420_064210/server_gpu3_port6203.log b/final_model/pytorch_model.pt.log/eval_20260420_064210/server_gpu3_port6203.log new file mode 100644 index 0000000000000000000000000000000000000000..8d10c3c50af3e8ba989a2774ed153d759a18682e --- /dev/null +++ b/final_model/pytorch_model.pt.log/eval_20260420_064210/server_gpu3_port6203.log @@ -0,0 +1,10 @@ +/mnt/data/.cache/conda/envs/vla_2/lib/python3.10/site-packages/albumentations/__init__.py:13: UserWarning: A new version of Albumentations is available: 2.0.8 (you have 1.4.18). Upgrade using: pip install -U albumentations. To disable automatic update checks, set the environment variable NO_ALBUMENTATIONS_UPDATE to 1. + check_for_updates() +INFO:starVLA.model.framework.share_tools:[*] Loading from local checkpoint path `/mnt/data/fangyu/code/reward_new/runs/0418_QwenLatent_13tasks_actionstate_50k/final_model/pytorch_model.pt` +`torch_dtype` is deprecated! Use `dtype` instead! +INFO:root:Creating server (host: 10-116-218-71, ip: 169.254.95.120) +INFO:root:server running ... +INFO:websockets.server:server listening on 0.0.0.0:6203 +INFO:websockets.server:connection open +INFO:root:Connection from ('127.0.0.1', 59626) opened +INFO:root:Connection from ('127.0.0.1', 59626) closed diff --git a/final_model/pytorch_model.pt.log/eval_20260420_064210/server_gpu4_port6204.log b/final_model/pytorch_model.pt.log/eval_20260420_064210/server_gpu4_port6204.log new file mode 100644 index 0000000000000000000000000000000000000000..b8eec3690c6959123303a6f0d7353b26788756b0 --- /dev/null +++ b/final_model/pytorch_model.pt.log/eval_20260420_064210/server_gpu4_port6204.log @@ -0,0 +1,10 @@ +/mnt/data/.cache/conda/envs/vla_2/lib/python3.10/site-packages/albumentations/__init__.py:13: UserWarning: A new version of Albumentations is available: 2.0.8 (you have 1.4.18). Upgrade using: pip install -U albumentations. To disable automatic update checks, set the environment variable NO_ALBUMENTATIONS_UPDATE to 1. + check_for_updates() +INFO:starVLA.model.framework.share_tools:[*] Loading from local checkpoint path `/mnt/data/fangyu/code/reward_new/runs/0418_QwenLatent_13tasks_actionstate_50k/final_model/pytorch_model.pt` +`torch_dtype` is deprecated! Use `dtype` instead! +INFO:root:Creating server (host: 10-116-218-71, ip: 169.254.95.120) +INFO:root:server running ... +INFO:websockets.server:server listening on 0.0.0.0:6204 +INFO:websockets.server:connection open +INFO:root:Connection from ('127.0.0.1', 52704) opened +INFO:root:Connection from ('127.0.0.1', 52704) closed diff --git a/run_qwenlatent_vla.sh b/run_qwenlatent_vla.sh new file mode 100644 index 0000000000000000000000000000000000000000..90b848b194bf18261207a9a7e65b97d11b182b73 --- /dev/null +++ b/run_qwenlatent_vla.sh @@ -0,0 +1,27 @@ +#export NCCL_SOCKET_IFNAME=bond0 +#export NCCL_IB_HCA=mlx5_2,mlx5_3 + +export NCCL_BLOCKING_WAIT=1 +export NCCL_ASYNC_ERROR_HANDLING=1 +export NCCL_TIMEOUT=1000 # timeout set to 1 hour (unit: seconds) +export CUDA_VISIBLE_DEVICES=0,1,2,3,4,5,6,7 + +# === Please modify the following paths according to your environment === +########################################################################################### +run_root_dir=./runs +run_id=0418_QwenLatent_13tasks_actionstate_30k +########################################################################################### + + +output_dir=${run_root_dir}/${run_id} +mkdir -p ${output_dir} +# mv this script to the output dir +cp $0 ${output_dir}/ + +accelerate launch \ + --config_file ./starVLA/config/deepseeds/deepspeed_zero2.yaml \ + --num_processes 8 \ + starVLA/training/train_qwenlatent.py \ + --config_yaml ./starVLA/config/training/starvla_train_qwenlatent_oxe.yaml \ + --run_root_dir ${run_root_dir} \ + --run_id ${run_id} \ \ No newline at end of file diff --git a/summary.jsonl b/summary.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..1ab36c50a46444834008b3384a4f1a5de29c6c5c --- /dev/null +++ b/summary.jsonl @@ -0,0 +1,10 @@ +{"steps": 5000} +{"steps": 10000} +{"steps": 15000} +{"steps": 20000} +{"steps": 25000} +{"steps": 30000} +{"steps": 35000} +{"steps": 40000} +{"steps": 45000} +{"steps": 50000} diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/101d51ca-2489-4dcb-9c9c-0b75c3de7a68_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/101d51ca-2489-4dcb-9c9c-0b75c3de7a68_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..aa7c6621dce2485fd85d460753a12e4b0deb08e7 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/101d51ca-2489-4dcb-9c9c-0b75c3de7a68_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2613d125c56e724a0ef8c26b4e1f860c95528c410a2e478080fddc8ab306bdf1 +size 3126042 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/10448c21-e965-4b84-9a05-f028638542f9_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/10448c21-e965-4b84-9a05-f028638542f9_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..91e36c3c710769bf77d79765d63db09b3b581963 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/10448c21-e965-4b84-9a05-f028638542f9_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2c56e979d57ca3fd1e601d1e9d7d0067d8cd5718a5cc32fb11844b9bb0c109f1 +size 3350930 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/13dbf2ff-b86e-4912-9b5d-51d7741a5da2_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/13dbf2ff-b86e-4912-9b5d-51d7741a5da2_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..577a64d118466b714a95a57a823352181813ea4d --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/13dbf2ff-b86e-4912-9b5d-51d7741a5da2_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:53382ab829d60e28fd86d980a793c44c26d1b267269908a0e327f9a62ffbadc8 +size 2826808 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/18415cf5-cb64-495e-bdfa-0602f4e30928_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/18415cf5-cb64-495e-bdfa-0602f4e30928_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..b787afcf88c4f8e95b338dd915a61f2bb23e20e5 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/18415cf5-cb64-495e-bdfa-0602f4e30928_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:92db339a9b86a40199e6fcc380c427dcbbcc398e64bde6c31eb12be2c50560ec +size 2811436 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/188c1338-ad67-4435-a227-cd60020a8ab0_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/188c1338-ad67-4435-a227-cd60020a8ab0_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..1e22ad33bd929d53fedd9f5417fbb8749ee491c3 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/188c1338-ad67-4435-a227-cd60020a8ab0_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:61a88e3f5a3c9a8782d1563a56fde46768c7411feb7c904de8f50147b6725d06 +size 3022485 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/240f9a25-24fb-4962-9ce3-75a840f69015_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/240f9a25-24fb-4962-9ce3-75a840f69015_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..44d814871d4adb5b1e23a65597c34920e481ffe7 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/240f9a25-24fb-4962-9ce3-75a840f69015_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c3225d099bcb86c312be22743d94bfcef9dcc8fd81276af18968681b229a3aa4 +size 3361402 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/3165e6c7-4d5b-44ea-8cdd-a29560aaa4bc_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/3165e6c7-4d5b-44ea-8cdd-a29560aaa4bc_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..bdeeb969a035851988d01eae6d9ed8366170b5f2 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/3165e6c7-4d5b-44ea-8cdd-a29560aaa4bc_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:962311f48bb0c465163e953dea96948d03b2a781732e0d22999f06830f97eea8 +size 3178743 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/35cef822-45cf-4114-80e4-bc8b991a5372_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/35cef822-45cf-4114-80e4-bc8b991a5372_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..bfac265bbf3266118f5ad345f47b8d8e47e57960 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/35cef822-45cf-4114-80e4-bc8b991a5372_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c9ca2fbcc2df121c17bd3addc0f938c225c8827b40d817d0ff32a0a222ec5332 +size 2909223 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/3bc7bfb8-7da3-4fd7-9a74-2233bc6519b2_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/3bc7bfb8-7da3-4fd7-9a74-2233bc6519b2_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..f381ed0694c6c32758660e3647b7b1ed1a96b4f8 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/3bc7bfb8-7da3-4fd7-9a74-2233bc6519b2_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:565ef38230f95e4463bed4c32f1a92c8742e1f1b26ce649d970b4b9100c7a022 +size 4614886 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/4764ec14-8143-41da-aeb7-970c370b884a_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/4764ec14-8143-41da-aeb7-970c370b884a_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..f8e9fc200e7e2a8a5e17d7fc38441c76f5caa61b --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/4764ec14-8143-41da-aeb7-970c370b884a_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9573a61d3919f9e924c2a31be1a2135404182b43913472212bcadb8cd4f9d436 +size 2679444 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/4958f259-0d11-49de-993e-fc9640096bcb_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/4958f259-0d11-49de-993e-fc9640096bcb_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..ad16eaab8d041886d9b78a7077a0d085b3b13ba5 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/4958f259-0d11-49de-993e-fc9640096bcb_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:595ac80c05298c03d1d895139c802705c9935199317c8e8415e868d77f6f38d8 +size 4789156 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/55767605-ae2c-408f-84e7-9583405c416b_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/55767605-ae2c-408f-84e7-9583405c416b_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..f2eaf19729dccebb4d764f7b0ec779c942dcfb75 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/55767605-ae2c-408f-84e7-9583405c416b_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5b6e9805974ead68678fbd9b222aacbc788d1fac3fc33f46d6e65c41b6a283a5 +size 2529618 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/595dda36-8986-4ce5-b6f0-46894f9d86fa_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/595dda36-8986-4ce5-b6f0-46894f9d86fa_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..b34ac55fa33734236d2330f7e222546b7895d68b --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/595dda36-8986-4ce5-b6f0-46894f9d86fa_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6d787ecffeb17d7a2fbb162aade93fadc540b5b836609e0d25998082da71ee03 +size 3383265 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/5df9eaf3-d934-487a-9d4d-7b0f4cda44dd_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/5df9eaf3-d934-487a-9d4d-7b0f4cda44dd_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..ef3fe18658c6f877c15d11f0d28920f6e8f43774 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/5df9eaf3-d934-487a-9d4d-7b0f4cda44dd_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b21ab078f7c25b80976b0e7d8101c39fdfab505dea8d68ae666aa095eff55e82 +size 4646529 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/63cfd5dd-581f-45ae-8e27-bfb7e41508e4_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/63cfd5dd-581f-45ae-8e27-bfb7e41508e4_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..2974083ea18d64bc9af0dc45cbc3d0a8078733e5 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/63cfd5dd-581f-45ae-8e27-bfb7e41508e4_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d809123da357e33b45e98940f3d525a934a7f6acd159163f58c363f4ed546652 +size 3292829 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/68723548-2b3c-463a-a393-ef219695cc86_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/68723548-2b3c-463a-a393-ef219695cc86_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..4f9ac70962c8d215a11d9f4d568cc66e511317e3 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/68723548-2b3c-463a-a393-ef219695cc86_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b78cc6c9e346bd93bb028c95a550f148891c233e0db4861ffc398cfbfd1cf58f +size 3002923 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/69c5eeaa-6a7b-4582-82f0-fda05529e518_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/69c5eeaa-6a7b-4582-82f0-fda05529e518_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..3ee8163e46aaa8c64ff3a254730a0ce41ae916d1 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/69c5eeaa-6a7b-4582-82f0-fda05529e518_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b583863b1d4909032811643f2deeb231c4c3534b36bde40fa2a1899260b43f1a +size 3574311 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/6e896171-5623-4ae7-b3b8-c0a17421646e_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/6e896171-5623-4ae7-b3b8-c0a17421646e_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..1aabf118e2c50643361c654f267831fd2f060d27 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/6e896171-5623-4ae7-b3b8-c0a17421646e_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:362f906e19a2979e5092e47b127a4ed8cba0007a3dee5621e7140ca61407924b +size 3831971 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/6fb913a2-c4d5-4624-8876-9815c31652e0_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/6fb913a2-c4d5-4624-8876-9815c31652e0_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..05cca20055f3f4855bd53cd8346558e88b39aa4b --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/6fb913a2-c4d5-4624-8876-9815c31652e0_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:131180947b546873e03011bd6ed82ea99c69fb794c6724ac42f9bb8a1332be39 +size 3728952 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/7d7a4592-4977-45aa-bf24-054d1b4e5257_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/7d7a4592-4977-45aa-bf24-054d1b4e5257_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..aa788cda96183f7ec2ad625b404974f37201a9f4 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/7d7a4592-4977-45aa-bf24-054d1b4e5257_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c279e7cf2069fd90125c335e5f6b5e714c3dee120dff940e0767abb45925229f +size 1370830 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/81184098-e19f-4754-9371-6a91b9cc59c4_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/81184098-e19f-4754-9371-6a91b9cc59c4_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..9c6aee4b2ccb393ba4d980f15295b3f25490beea --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/81184098-e19f-4754-9371-6a91b9cc59c4_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b2d1f7c49cc7b1b9e290453682a305c2b8b47f6b2eeea1bd9103b37a3b63c094 +size 1266748 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/8b7d6ef1-fc00-4b03-acf3-be1bf2d619dd_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/8b7d6ef1-fc00-4b03-acf3-be1bf2d619dd_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..a4042851c066b9e716ac7979f59188ece89fe10b --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/8b7d6ef1-fc00-4b03-acf3-be1bf2d619dd_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d53c5f77c412e124296792177b246e4b00f1a9d8ee2935621ceb5793dee0e014 +size 3152035 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/8f4011f2-cd38-4eed-8500-5412930bc59b_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/8f4011f2-cd38-4eed-8500-5412930bc59b_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..f1c13c09be993a6e6e7ee42d171b58241db64bd2 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/8f4011f2-cd38-4eed-8500-5412930bc59b_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:aa2755933dfea02d2f3622e62938dbe48742a8b007a17138b1ba67e324a8f4fc +size 5012432 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/90c0e30a-3c06-4fd2-bfc3-9878cc521972_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/90c0e30a-3c06-4fd2-bfc3-9878cc521972_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..0112cc32884469129a91eb795d1ddfea6de6e701 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/90c0e30a-3c06-4fd2-bfc3-9878cc521972_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:34f8104f6016bb47921de15af30fc91c70aa0a6bec1bf4ab5cdb9293e7c83470 +size 3489636 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/9470dd7b-a889-477f-ad1b-80925b5380ad_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/9470dd7b-a889-477f-ad1b-80925b5380ad_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..83a0037a6ee623b2eef962acde7309132fc14bfb --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/9470dd7b-a889-477f-ad1b-80925b5380ad_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2f7e65dfca2bcf772b4be5afbd4cd7a82f36c011a1b41ee4d61a8531ade8bfa0 +size 3281940 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/978497d7-d992-4041-a7a8-766383ed0614_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/978497d7-d992-4041-a7a8-766383ed0614_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..1c5a5b5b9c2c9b2cc5cdd75d998d886e7a3c56da --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/978497d7-d992-4041-a7a8-766383ed0614_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6bd554a8cd17d024da99d071418a244d5595a1cdc312972c5d06c8f9751f2151 +size 3351407 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/a819478c-9e9a-4649-9be5-812483bd5725_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/a819478c-9e9a-4649-9be5-812483bd5725_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..b45b0de44e5deb1dd730d0a0863155b9a3af3833 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/a819478c-9e9a-4649-9be5-812483bd5725_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:737196403217e508cadc9ada59f52c621e1605df08794e8267e25b0030ee5de5 +size 3538684 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/a89039f7-5f61-468e-acd6-78caaee3ccef_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/a89039f7-5f61-468e-acd6-78caaee3ccef_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..7c9c52cd83d27dcd7fd108e48f71073ef1036290 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/a89039f7-5f61-468e-acd6-78caaee3ccef_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:91ce2aaf4490e551fa9221fd3cce209fea051cfba726e5f0d22ad2286c9b616d +size 3080131 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/a904b30c-da2b-4072-ab20-cab7cf56b41c_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/a904b30c-da2b-4072-ab20-cab7cf56b41c_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..0e3a49fd2c07cd081a75b3c6907829bb1a479341 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/a904b30c-da2b-4072-ab20-cab7cf56b41c_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6a05b0dcf949dd2dd5115cb81df95d0e988c85f774303f35954e258527ac4985 +size 3051755 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/a9c0a5a1-3085-4a69-997b-119237d326ba_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/a9c0a5a1-3085-4a69-997b-119237d326ba_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..a33893ec63e5ecaa04328da955377f668a4b346c --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/a9c0a5a1-3085-4a69-997b-119237d326ba_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f86dd3426be3772fb60afc0d971912de7457d4db49c6f1fc2a574d5d3a70e347 +size 3721551 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/accfb691-0159-4fdb-b193-ebbef7e1419a_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/accfb691-0159-4fdb-b193-ebbef7e1419a_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..f1575b6ef0533a7d67acbf16389db60d85900baa --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/accfb691-0159-4fdb-b193-ebbef7e1419a_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d9581d853d57eeeec2fd1d72612f963cffcbda4a06be9cc1c55dc5897bf87626 +size 3679413 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/af66aaf4-975b-499c-8bb6-fc87106244f6_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/af66aaf4-975b-499c-8bb6-fc87106244f6_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..2743daeeac117a389de059ec570714a56bcbfc4e --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/af66aaf4-975b-499c-8bb6-fc87106244f6_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:144d25bc9c8c79a7f974c7c1964cdb72c1c0e767b60f284f18fa665def6a3418 +size 3478349 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/bbe92f58-9da4-42a9-84ff-04ea95f484c3_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/bbe92f58-9da4-42a9-84ff-04ea95f484c3_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..8cd8278bd1a6ab26dcb978d8c0f84fb33c1c4c1d --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/bbe92f58-9da4-42a9-84ff-04ea95f484c3_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4ce6cc6a1e5e20dc506cb844c6e606094760bd6f771030db7fbfeb1ee9b0945b +size 4956804 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/be7eaa45-5e8d-47d3-98ef-a4f10b852484_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/be7eaa45-5e8d-47d3-98ef-a4f10b852484_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..3ee212faf8fb010e442ee3e15b8ce58b337d2fde --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/be7eaa45-5e8d-47d3-98ef-a4f10b852484_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:039521c3bb2096757be57017a99dd6ae01b91b609da4bce2b7992562d323f5d6 +size 3712187 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/be9a4740-975d-4e03-9510-0a13a0c42e5d_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/be9a4740-975d-4e03-9510-0a13a0c42e5d_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..d6d7d2fcfe5535a079360f76389040e477a9a6bb --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/be9a4740-975d-4e03-9510-0a13a0c42e5d_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9d2450bd207a2c7337f2f781d6f6565a69d2d06b949623dd72c406f5962bd4c8 +size 3118458 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/c5a5be2b-057e-433a-8e8f-b7dadbdd2112_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/c5a5be2b-057e-433a-8e8f-b7dadbdd2112_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..47e9d25864703be3fd9e4da0ce3117bbe94e0b92 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/c5a5be2b-057e-433a-8e8f-b7dadbdd2112_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7b316052e008da7878d68203466a2909e1f03c95ad89293226d54f0e1e22754a +size 3703052 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/c6f72900-07ce-4b82-89c7-23ee4c055398_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/c6f72900-07ce-4b82-89c7-23ee4c055398_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..91e9674a4520a6c4484eeaee79d8d99c605a80f6 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/c6f72900-07ce-4b82-89c7-23ee4c055398_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:57f9d51f3aef4df02b3569cce456d6794fb1d73c626cd717a2e74eb7409fa747 +size 2393730 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/ca9adb4a-2a41-49db-b25c-563f4a4b1e47_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/ca9adb4a-2a41-49db-b25c-563f4a4b1e47_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..4bb7c611d4f92a22b45a18e868f7f7bf4a1880ba --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/ca9adb4a-2a41-49db-b25c-563f4a4b1e47_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f54d40bfa7a1a8b490fb08ce98541416e8f9ca4e26a82121c29e21c27aa82ff9 +size 4492777 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/d5274fd9-3a60-4be0-9a57-810127e1f2fb_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/d5274fd9-3a60-4be0-9a57-810127e1f2fb_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..af78f2f289bfa44dcd6f3e9c8d84a8b495967188 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/d5274fd9-3a60-4be0-9a57-810127e1f2fb_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ef665388d6595cf871bbb52b2a4e4c2fc25e359000b2bc962110790ad77f3518 +size 3657329 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/da77b644-5153-4d9c-9f20-a0218f578316_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/da77b644-5153-4d9c-9f20-a0218f578316_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..9ad1737796bd395190f07dcfeedd071b9668b28d --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/da77b644-5153-4d9c-9f20-a0218f578316_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:23c874ba379b41ec174ac620cadbe67b03452424412a66bf03e81fe13186a4cd +size 5041097 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/dbf32b98-ca20-4971-9b8c-06cedf2effd0_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/dbf32b98-ca20-4971-9b8c-06cedf2effd0_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..4776fa205fb03a9aeddd71449ac4d1f98acc277c --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/dbf32b98-ca20-4971-9b8c-06cedf2effd0_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4d7ae9f8aadeaa20492841d29cf301076239666af5d6eb118f0f6e0c613e3d0f +size 4735908 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/dc1a1efd-eb27-4a5e-80be-399453d4b1bc_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/dc1a1efd-eb27-4a5e-80be-399453d4b1bc_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..d686fdd93f2ea38e98eb483449f92bcffc308e15 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/dc1a1efd-eb27-4a5e-80be-399453d4b1bc_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:aa5e4f13f45b93efe51808347e6edca12e0f26418c92ca4cf1da49433b066a26 +size 4019020 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/dc5e3c24-2399-40c2-a7ac-b3d89ee8f238_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/dc5e3c24-2399-40c2-a7ac-b3d89ee8f238_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..30c3371172c1e222fc31d8306d63a23961e7770f --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/dc5e3c24-2399-40c2-a7ac-b3d89ee8f238_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:764ddff0cf1f3174737e9ad69245fa3668f2b1b7414eb6b2e44a0dc77ec12231 +size 2858064 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/e4a80fea-f083-44c3-8e36-466342ffa7fb_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/e4a80fea-f083-44c3-8e36-466342ffa7fb_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..350d31814ed126f9f440bb02243d26bdc155641e --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/e4a80fea-f083-44c3-8e36-466342ffa7fb_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6608e1b9514c83ed34edebe0cd7d84103ccd4bc7d98e87a6587a439eddb947ea +size 4590928 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/ed140b40-0544-439b-9fbc-14bbd38f76fd_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/ed140b40-0544-439b-9fbc-14bbd38f76fd_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..d494f41744fbb2b0f3bd2ca98bd3b4cf7705da92 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/ed140b40-0544-439b-9fbc-14bbd38f76fd_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9428241cde89604f49952da985eb5abb21f60ffecc3baf9d7ea95c09eda05c7d +size 3142369 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/eed2e25d-ce95-40bd-8060-ffc837be657b_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/eed2e25d-ce95-40bd-8060-ffc837be657b_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..ca40c2bf27c2040356186b5d89043cae77ee0508 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/eed2e25d-ce95-40bd-8060-ffc837be657b_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6a0939945a30c4cab08d39c7de7f684edf060ccb979b51ef4dac2ee92316ed94 +size 3953530 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/ef1917dd-f582-4b10-81fc-9127b1338817_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/ef1917dd-f582-4b10-81fc-9127b1338817_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..fd36f4493c8c62a49adf7194f2d136b17a484084 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/ef1917dd-f582-4b10-81fc-9127b1338817_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7488fd2cd8c12ec1a9663ba8bbf28384bb6e91cc24c0b68d3b4e4c3ac2a9f5f6 +size 2755757 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/fae8f20e-c800-4781-9b20-048046fe0b82_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/fae8f20e-c800-4781-9b20-048046fe0b82_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..fd81c2b71dcc459f1f217f74d6764df40ffb8abf --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/fae8f20e-c800-4781-9b20-048046fe0b82_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c1d13e2d616c1b182aad93d1e8ebb221d0ff0be74914bc682e3c0d17d66fc4aa +size 3216956 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/fcef6b0c-e58e-438c-ba64-89404a838f16_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/fcef6b0c-e58e-438c-ba64-89404a838f16_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..57466bb7b9d7269d8e145a838809ecacefd605d1 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/fcef6b0c-e58e-438c-ba64-89404a838f16_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c61ed3d73fc4522fcf9c8d694c90a7a381e88bb96fb92937484dede73a86fe31 +size 3503596 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/ff2fdc77-485e-4b09-9250-d1cdff5a052e_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/ff2fdc77-485e-4b09-9250-d1cdff5a052e_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..f563c86b7927315791098bf32deb701e389465e6 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PnPBottleToCabinetClose_GR1ArmsAndWaistFourierHands_Env/ff2fdc77-485e-4b09-9250-d1cdff5a052e_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:57b1b01eb1351c55a7fc79734b928b65d4b4e0647ecaeca3f7f7dc14cafff86c +size 3458697 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/0d460c64-f43c-4566-91df-e3b49476f0a6_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/0d460c64-f43c-4566-91df-e3b49476f0a6_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..11e3b779ab9358de19078fb03ad00363fdf60cd2 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/0d460c64-f43c-4566-91df-e3b49476f0a6_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9f186930c75a15ed09946a0d82c412919625de485d72de1ee31c1f899848b8f0 +size 6360147 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/1237575d-1e4a-466a-b462-9d0db98318b2_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/1237575d-1e4a-466a-b462-9d0db98318b2_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..e2b17dbf849cefaa67ffed5605e627e5b06739e7 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/1237575d-1e4a-466a-b462-9d0db98318b2_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c515b6fc717717427c94ddeb5570e846b34497b409ddb33bbf35ff4867206081 +size 6702402 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/12b2b73b-350c-45a7-8f76-95e62624ea58_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/12b2b73b-350c-45a7-8f76-95e62624ea58_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..2e4becc15a889b860b1e4847cee599a3f41a54ad --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/12b2b73b-350c-45a7-8f76-95e62624ea58_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:37fed39e02e70a69382b6c857e615b2e4d178abc4725f7c80734bb27c8e246d4 +size 7747068 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/17efd764-1bfd-428d-90eb-45321f078bc5_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/17efd764-1bfd-428d-90eb-45321f078bc5_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..01a9d68ab186407c1dc5cb5bcb361928d0911847 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/17efd764-1bfd-428d-90eb-45321f078bc5_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:25df9c13faebba44b3e40f1a1fd98fd7119c6253c762f776ac3ac1ef6cdee41d +size 5620771 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/1bbeca5b-e9df-42ed-b6c0-f2f88c822b40_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/1bbeca5b-e9df-42ed-b6c0-f2f88c822b40_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..16421f6fa703dcf0340b1755b37e5d0de71c2c03 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/1bbeca5b-e9df-42ed-b6c0-f2f88c822b40_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c2f7455e90e34d86eac4f28c6aba81193c317d44ce8eb10e931d5783d8a074ab +size 10595160 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/1bc7b7fb-e7de-4261-b376-4de249d181a3_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/1bc7b7fb-e7de-4261-b376-4de249d181a3_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..0b4801de63691752d9185734f6c74b26d05e306f --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/1bc7b7fb-e7de-4261-b376-4de249d181a3_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ab7a218cff0e4df7ee321493447817cd97efb04fc782471dd6cd9172d70a2e38 +size 8079486 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/2731a4d7-5b36-42f1-bc63-5aaddb0bfc6b_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/2731a4d7-5b36-42f1-bc63-5aaddb0bfc6b_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..c856b9d5a66c0ab9c036fab84628081a8a61b8ac --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/2731a4d7-5b36-42f1-bc63-5aaddb0bfc6b_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3eca31eda61f44e78542d9a685f8ce672d8c081fe362721b614e5492551f9bf0 +size 4769084 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/2906af99-3d47-452b-a3cc-d18d55ed5214_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/2906af99-3d47-452b-a3cc-d18d55ed5214_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..7e27f5d1a7a61a10b4dfa377a16f8e79211c7875 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/2906af99-3d47-452b-a3cc-d18d55ed5214_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8c7426f7c8a50cfed47ea0617ea49f5e4e66cafa9a0bea159c6705fd36d0599b +size 4371403 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/3302dfb8-1277-4a31-9a2c-df6e9b1d419b_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/3302dfb8-1277-4a31-9a2c-df6e9b1d419b_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..1cda788804dde2d98d130ae9847e63379b185cd7 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/3302dfb8-1277-4a31-9a2c-df6e9b1d419b_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2b47536eef0deace607e0f7bb81c81a72e44bd08bbba9c39762817a58831b8c1 +size 6066786 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/38c7c055-084b-41fa-9c7d-878f43643157_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/38c7c055-084b-41fa-9c7d-878f43643157_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..09ba6262954a9f7d4d96505dc98d02abfc67bab3 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/38c7c055-084b-41fa-9c7d-878f43643157_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6e33f30687eca6d40a4d1a8b509c89d4aa5a1582fed95c543bddedfa22c689c8 +size 10163729 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/3a6f675d-7929-497e-a00e-68dc8f60251e_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/3a6f675d-7929-497e-a00e-68dc8f60251e_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..f48ba299cc360af86b2bda51d58e8969d9b8eaf9 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/3a6f675d-7929-497e-a00e-68dc8f60251e_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0f7af5924f59a2a0339639ad24bff9b5e6324e75d7c27d93e8c935a25b311516 +size 5343475 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/3a79d536-d135-41d9-b7e4-3b01d88aad69_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/3a79d536-d135-41d9-b7e4-3b01d88aad69_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..55a13750a7858f6b03a6445d1c808cd6ceaf3fab --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/3a79d536-d135-41d9-b7e4-3b01d88aad69_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cff9bc3a93587739a9ddda6f64415447e5e8c4ac92eb5e433ba49281eb1cb9e4 +size 9251515 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/43f733c2-d0eb-4ecb-ba29-1f2aaba734f5_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/43f733c2-d0eb-4ecb-ba29-1f2aaba734f5_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..f8491e03d20f66e7abd2d78ddbd4d3b6e6f43d16 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/43f733c2-d0eb-4ecb-ba29-1f2aaba734f5_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5db655f4614b7729edfb8556086ee4a7afab39a29c431973f23f6fc1d6c934c1 +size 8685581 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/4bde81b1-fcbf-4049-9e27-5c2243639c0c_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/4bde81b1-fcbf-4049-9e27-5c2243639c0c_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..3adc971125ea3b9bd66a8b952630c776e1c848c4 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/4bde81b1-fcbf-4049-9e27-5c2243639c0c_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:19e1fb97a88af39c96ccb80b83995edd7831939c8c809d3c643b8f32bc178dba +size 6757716 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/4e8aaecc-574c-429a-8fc3-501db2590b10_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/4e8aaecc-574c-429a-8fc3-501db2590b10_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..f0b880fdc9d05355ebd435be99637fafc006580d --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/4e8aaecc-574c-429a-8fc3-501db2590b10_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9b54dc6b50f4a87294ddaa9e92e5f3d74fef79387fb61ab3f902da159d49b3f5 +size 9129531 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/5221791a-be86-4e06-bf80-ad1a34b7f925_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/5221791a-be86-4e06-bf80-ad1a34b7f925_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..3bc36d5f76b96dcf22b095fba4f15c970bd9ab23 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/5221791a-be86-4e06-bf80-ad1a34b7f925_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e3674e1efc2594d807e52ab1005c58101dfef884a8419557fdd472a43c49face +size 7941237 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/5af20987-9f8f-4a7c-a222-72c71699ff2c_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/5af20987-9f8f-4a7c-a222-72c71699ff2c_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..92848f23343d8180bf6fd652a66c122601c72cfb --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/5af20987-9f8f-4a7c-a222-72c71699ff2c_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:721988cabaaa319721018019cfa10304c0bea619a29c17b0f3b117c46c9bdf1b +size 6428072 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/5d06afe6-4bd1-4942-a475-6b3a830a097b_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/5d06afe6-4bd1-4942-a475-6b3a830a097b_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..a9d299156f6b4b280910d439a065f6261eaba42d --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/5d06afe6-4bd1-4942-a475-6b3a830a097b_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7240ce552f7d8a1ec55a1fbb2c50ce37dabc5ae2640c36bf76a991af8357af2f +size 4677876 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/5e41b01f-4a01-46b4-be86-9b040c749751_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/5e41b01f-4a01-46b4-be86-9b040c749751_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..1f0210b338a1f08d421520607fd3831951e66a7d --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/5e41b01f-4a01-46b4-be86-9b040c749751_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:33685e9aca9a12bb795c457ddf564d30af94108fc3bfcc75763d7c9711c23288 +size 5901358 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/5ec96f14-f5d9-4ed6-bfe0-732911bd64f5_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/5ec96f14-f5d9-4ed6-bfe0-732911bd64f5_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..0b99a1ff78b0a69e8e6e3b2593fa9e2388389a14 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/5ec96f14-f5d9-4ed6-bfe0-732911bd64f5_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:50b138f485753257a50171a4b14dd8fb9ccd5841df58e0f2a4ba9216761637ef +size 6560677 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/6265d818-ba49-4248-81d2-9d74a5bd2264_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/6265d818-ba49-4248-81d2-9d74a5bd2264_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..e0989502085b3ab48a240da6fd801d2c4b20c490 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/6265d818-ba49-4248-81d2-9d74a5bd2264_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8a5c7eac2b748c0603696d15b891141c3b6397bb7777386c2de2fb471f4a3a5e +size 9534652 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/6d69e6ee-1432-423c-aa9a-9e0a9889118e_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/6d69e6ee-1432-423c-aa9a-9e0a9889118e_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..08407390e42e09b20a9a00b57efa15cd8788ce3f --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/6d69e6ee-1432-423c-aa9a-9e0a9889118e_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9f786d67d1f5f11ca3557ee131e8bb08c9cb74075185c751e45b895f5fc2ee16 +size 6216779 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/6d9cafc4-21b4-4cc3-8ebc-08e527c6b3d7_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/6d9cafc4-21b4-4cc3-8ebc-08e527c6b3d7_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..a732be26fddb65214740a298b80339538092f95e --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/6d9cafc4-21b4-4cc3-8ebc-08e527c6b3d7_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f5936536b739bbb2b560cc53cc3f18f1296739a1e1222e2b29414e1549fdf646 +size 6808881 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/6ee3f707-16ea-4ff5-b22f-a9fd15b1f390_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/6ee3f707-16ea-4ff5-b22f-a9fd15b1f390_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..8dd969ea017054d36dfd98199c61028eeac1e27e --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/6ee3f707-16ea-4ff5-b22f-a9fd15b1f390_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:71a4dd7adcef6a07b5e09a81b444edb30aea3e891fc8223aefc960c1672d368b +size 4164506 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/75582ad1-113b-4d55-b15b-5fd09a7c4144_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/75582ad1-113b-4d55-b15b-5fd09a7c4144_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..5733f0c7d450b93a8c231ba0e1fa66f7fd902ed5 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/75582ad1-113b-4d55-b15b-5fd09a7c4144_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0f9c700361986b4e185200c4fc169442d457bb5c25d3804594f7fccc6c375a67 +size 5312973 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/7b48013f-20cd-46ff-8c77-d74a5b320f42_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/7b48013f-20cd-46ff-8c77-d74a5b320f42_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..f6f92c1705301eff9ca6c56c59b0462cc7a88311 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/7b48013f-20cd-46ff-8c77-d74a5b320f42_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:202d79f7157d1798fa0e6c6520fd5ab7440d0a8f17a0e7b5b5e048b19e5a336d +size 4117320 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/7f708ade-f5a9-412d-aee9-db24a321dc21_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/7f708ade-f5a9-412d-aee9-db24a321dc21_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..3f48eaaf1f4d4a8a1975408aa2ee12616e8bc963 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/7f708ade-f5a9-412d-aee9-db24a321dc21_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1154b00474ef39d87a8c61a8f7824d71bce2249a03bffed3c3f28afa694213bb +size 9862029 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/8706bc77-7f64-460c-bbd4-facf9dcacd73_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/8706bc77-7f64-460c-bbd4-facf9dcacd73_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..3b72189c2325a0fe2d778dab7f945cdb45a95ceb --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/8706bc77-7f64-460c-bbd4-facf9dcacd73_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5f2b30ffe13f4cd4c367d416f400c4f9b58d41865790958adc4395dc0af6de96 +size 5934092 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/881d64d6-d1c9-495a-a143-715ef3ad8433_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/881d64d6-d1c9-495a-a143-715ef3ad8433_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..099c4bf347319c1914859edcf6dd92726387badb --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/881d64d6-d1c9-495a-a143-715ef3ad8433_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3b28ca9d827100c680d8fef53447059cafa716117804ca59f6d844c0f559ddbb +size 8521758 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/8a054087-a814-4221-9aa3-2b3f53ee83f0_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/8a054087-a814-4221-9aa3-2b3f53ee83f0_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..f3790d2753b11ede107a3a303fa316dd03fb4c01 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/8a054087-a814-4221-9aa3-2b3f53ee83f0_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:be4e17fc913de7269b330a8e479b9609c59221fd6b76a4bb5a8642b298bb2001 +size 5889428 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/8c0c53f0-c015-42ca-b46c-bb4e9dbfcb7e_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/8c0c53f0-c015-42ca-b46c-bb4e9dbfcb7e_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..e07d40e8fe8415010b1c13d34eb65fe4a387a08c --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/8c0c53f0-c015-42ca-b46c-bb4e9dbfcb7e_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e11c15587c7d43580ebc1a83db7fd97d7181fc349b537fbb4fa882171446302c +size 4892096 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/8c3ec213-1aaf-4483-8c73-10e787b17c25_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/8c3ec213-1aaf-4483-8c73-10e787b17c25_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..60cbfaf6b65aafc8b90826f567152791955d372e --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/8c3ec213-1aaf-4483-8c73-10e787b17c25_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0e085c1906241a2b0cfc304a967d689bbe0235b58f50ba480624f7477f566f9b +size 8068326 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/929e3abf-5983-4c3d-be4a-3f0071a1647b_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/929e3abf-5983-4c3d-be4a-3f0071a1647b_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..fb64f2d4cd7dca1d02ef0641fe241c0aec611a6e --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/929e3abf-5983-4c3d-be4a-3f0071a1647b_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ab148af8663a07e33a2e9d83725c89b2ca41fba2b4d88cc7ebeb37618362bbee +size 4190618 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/976d9ca8-9b22-4843-b025-6b202ddde753_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/976d9ca8-9b22-4843-b025-6b202ddde753_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..5461814902582dbb29407c8f6293bce2aa4d9588 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/976d9ca8-9b22-4843-b025-6b202ddde753_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:22d3c7bd5dde839ad8a81c30a83d92531e59b076b456a572d1495bf296bf8fcb +size 7666715 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/9beb2375-83cd-4707-8207-27806510bf95_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/9beb2375-83cd-4707-8207-27806510bf95_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..3e083e1760337b8f3e64d14deebbc5696feb5b4d --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/9beb2375-83cd-4707-8207-27806510bf95_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c437f88c5f7e9d5215a787691d8b1ea081bd9422a886d8def04c90fad4aa3f2d +size 7789343 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/a2ea6475-089f-4ec6-a943-2f5bca11f7e2_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/a2ea6475-089f-4ec6-a943-2f5bca11f7e2_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..1fa1b4382fe88e9948b517cc34b8773b7a559c59 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/a2ea6475-089f-4ec6-a943-2f5bca11f7e2_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cd231d8ca137dfa92758381ec1ca518dc546d14b4025bb47cfc0c43d1c3cb7cb +size 6039652 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/a880d463-dbd5-4958-b31c-4c49a657a7d7_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/a880d463-dbd5-4958-b31c-4c49a657a7d7_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..28d9ec833fbc6a899fc420cf194d15ff2c572089 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/a880d463-dbd5-4958-b31c-4c49a657a7d7_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b319452f75f2bc8928190b0efd75e83987135a76c6d148abf0ccb8dba4b032c6 +size 7936861 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/af4ed1bf-bfe0-4429-b717-6addd4bcfaa1_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/af4ed1bf-bfe0-4429-b717-6addd4bcfaa1_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..ceaa8c214ff046fb50b14680c0b2d239e16fea93 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/af4ed1bf-bfe0-4429-b717-6addd4bcfaa1_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e4c9ec66396bfc573d3214fb4b634c142fa39132327d4329ac2bae9593638a94 +size 7577789 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/b1b6f572-2650-4bf9-a819-ac927adbb376_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/b1b6f572-2650-4bf9-a819-ac927adbb376_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..e36891ae5a5b2f6fbc04292a6df9582464cc6131 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/b1b6f572-2650-4bf9-a819-ac927adbb376_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5221d2f9e2c54b0945eaf3a8c97c708b15e3706a62a69f8e8365739693ab2fe4 +size 5547705 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/b71ef9d3-1dcd-4386-aa6f-72c763e385a3_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/b71ef9d3-1dcd-4386-aa6f-72c763e385a3_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..112fe54851e5032b15841f557fbafad57f9496f7 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/b71ef9d3-1dcd-4386-aa6f-72c763e385a3_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3c402c8a065e59faf72e288e104f5d9b77c8634e2dc0d723dbec9ad927c5498d +size 6996525 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/b8830aaa-6a6c-4399-b9a1-a3a235f1bd28_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/b8830aaa-6a6c-4399-b9a1-a3a235f1bd28_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..564a279c3da7791d47b7f2663721d477ffaac666 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/b8830aaa-6a6c-4399-b9a1-a3a235f1bd28_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:310173373eb26ac6bdc5f18504beee052b1a29fcd852dd6a3b64c127e59693a2 +size 6732464 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/bb249aa5-fb31-4d8e-8405-25259e164004_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/bb249aa5-fb31-4d8e-8405-25259e164004_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..631fc372e4dc5b3316e28f7fa16ee00f06bd75db --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/bb249aa5-fb31-4d8e-8405-25259e164004_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d461c7adf5f63da49b199b33a92d5199af7f73f402b1804b1156de1602ce67e4 +size 7773530 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/c379b03b-af2a-435f-ae27-ca4df0c65ec8_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/c379b03b-af2a-435f-ae27-ca4df0c65ec8_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..93e12c82da351128f12149a52010be23c6d294cb --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/c379b03b-af2a-435f-ae27-ca4df0c65ec8_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:16c489f135daed443ab2914a93e46fffcfdab06e9582bceb9a3834a83ccef667 +size 8604295 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/d2185a8b-95fa-43c3-9590-b776ce17fe2d_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/d2185a8b-95fa-43c3-9590-b776ce17fe2d_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..db2a97a103368ff40c58b337a35e467afdf8b5c4 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/d2185a8b-95fa-43c3-9590-b776ce17fe2d_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:27b83cca13d200b88f0bf7e436077a48ed56bf97381ad4a69db044bbf9fdb138 +size 7613738 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/da1620ce-500e-4636-9245-fa1627febdc8_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/da1620ce-500e-4636-9245-fa1627febdc8_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..24a9ccd0d138b01c71335fe3faf644502dc58593 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/da1620ce-500e-4636-9245-fa1627febdc8_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5a2e0c5bcfeef3e7b6d9a4508247b46c5955aa9f8fab92d7f775c3e8947688d6 +size 9385265 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/e2f23917-265c-4074-becb-04bbbe0b6393_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/e2f23917-265c-4074-becb-04bbbe0b6393_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..57d8ced4ca7d6ff19ee92b97cbc0cae19e3f8b8a --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/e2f23917-265c-4074-becb-04bbbe0b6393_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e35ba11756041a3d7e4fe93d2d1ac84fe73fcb2d737a2ad89117cf7cd9ff102f +size 5303918 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/e6ef1ba6-69ed-45f8-b959-1e35b579ca6a_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/e6ef1ba6-69ed-45f8-b959-1e35b579ca6a_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..18ed9496aa8b2abc943922267a0c096a99258c84 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/e6ef1ba6-69ed-45f8-b959-1e35b579ca6a_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:04152e3e0f31f9725bec7f419c8a021cb40577be0f50966956f846f5979804bb +size 4761541 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/ed1fab66-134c-49ac-892e-68610768d49f_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/ed1fab66-134c-49ac-892e-68610768d49f_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..8a993f9997cf94c536117beceea9f3d6f4fe53c0 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/ed1fab66-134c-49ac-892e-68610768d49f_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fab8db3e0dfdb262c61ea148d46861a703ce5c65b73a018a87e3b14fd8212ad3 +size 7590866 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/f2884552-c396-40a1-a232-950baa6ced05_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/f2884552-c396-40a1-a232-950baa6ced05_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..21fcc2e7ef5adaf0f6095b909ad6380b747aebba --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/f2884552-c396-40a1-a232-950baa6ced05_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5a4a0ffdc5a1bd4449838c4e8656a4adb5730e73601eeadfb84557f105ccc756 +size 7742023 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/f5723a46-9587-4d15-8dca-028975b9f56b_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/f5723a46-9587-4d15-8dca-028975b9f56b_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..5fbe249c5ff006b794dea1a03f106145b0c860fc --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromCuttingboardToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/f5723a46-9587-4d15-8dca-028975b9f56b_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6b5de0ff6928403c77098a5f68bc1877a409930585bb769c4af825b6857245b0 +size 9275829 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/056b382e-3603-4467-bd04-39698a7a1821_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/056b382e-3603-4467-bd04-39698a7a1821_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..69d0710a06b9e2654b355f438d5e0478c2c32cf6 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/056b382e-3603-4467-bd04-39698a7a1821_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:227982d8a3e4413b014ca2e8186e608a71f7ac0288c524a311c60f1738ad95c6 +size 10324150 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/0b9b5fc7-ecf3-4625-9344-72c9005d76ea_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/0b9b5fc7-ecf3-4625-9344-72c9005d76ea_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..74a10a9ab24495b22b0ac34aaae017fc5dd85ebd --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/0b9b5fc7-ecf3-4625-9344-72c9005d76ea_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:19483e46e96e4a1f30249fe9ab4bf31ad5cc8fe250fde21703c90d73b7c3a76a +size 4590084 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/0e18cd09-6c14-4aa4-a6d5-aaba72f81b44_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/0e18cd09-6c14-4aa4-a6d5-aaba72f81b44_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..8cbd09db34f7020ebfdce4f666738735081d8ee3 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/0e18cd09-6c14-4aa4-a6d5-aaba72f81b44_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c6fdba6d5f911315c36196ecc5b0036affd676248e02082c119fa2019e016652 +size 9391125 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/0fbc7376-936e-4869-a3ad-d7696f3e7578_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/0fbc7376-936e-4869-a3ad-d7696f3e7578_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..d891b7e7bdcb285fbc0904081f76a39cd0379fb5 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/0fbc7376-936e-4869-a3ad-d7696f3e7578_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1735cd4c138893a4df6106a24566af52f1e8186b258b31be8ae95d93b24c5c1d +size 5109036 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/16098338-bb93-4043-8604-7b6a87071be0_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/16098338-bb93-4043-8604-7b6a87071be0_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..fd9e9bb4864abae5525e0a26c6443fea27d92fe2 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/16098338-bb93-4043-8604-7b6a87071be0_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:09e6c97d7ddb23a4c9a15cf8aae3d91a08bf08849b8b4daf9421ba16f30811df +size 5621550 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/1660b1c7-86ac-405a-b4f1-00fe89718635_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/1660b1c7-86ac-405a-b4f1-00fe89718635_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..9e7a6339861d31dc108084e540fcea8313aa42de --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/1660b1c7-86ac-405a-b4f1-00fe89718635_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7204b5bed3964bdcb1b91095d00bd912995e282fee9c14f99ad82a0efe9b3fdf +size 6831030 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/172f088b-3357-4adb-bbaf-def07eea74d3_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/172f088b-3357-4adb-bbaf-def07eea74d3_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..162c193a8d275b1e1f97e1d55a7524aa91afcd2c --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/172f088b-3357-4adb-bbaf-def07eea74d3_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:32f9e390a6cc8774e2d88c4310f71f8afcb467631bdcc353ea93a49059fb70fb +size 7567246 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/173c83e2-2de6-4b64-944f-b535d5106e30_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/173c83e2-2de6-4b64-944f-b535d5106e30_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..87a7ab99995dbc5f5b213d0962c185c86e5534ae --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/173c83e2-2de6-4b64-944f-b535d5106e30_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2c18d0a47d09732f10efddba016f6040777daf6af92eef8d7fcafb9b00152054 +size 4082740 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/18c9ea8e-d8ca-4846-865a-9a42f055bd40_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/18c9ea8e-d8ca-4846-865a-9a42f055bd40_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..76181dc4cab6644d4c96649210272d731d90d002 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/18c9ea8e-d8ca-4846-865a-9a42f055bd40_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bd76349e212ad597098bb6f97745e77a93e274a01ec35aeaf00da177843928d6 +size 4114222 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/1ad89741-c191-49ff-97db-6d68ec67aae4_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/1ad89741-c191-49ff-97db-6d68ec67aae4_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..3ddf638cbe404dc606f9429fac70328950e6616a --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/1ad89741-c191-49ff-97db-6d68ec67aae4_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a083aeec0a7021b71a2edb97577664a232e01f436da707d2190e990850cbbe08 +size 6039688 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/258129ba-fc32-4df5-aba0-235be8d2947b_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/258129ba-fc32-4df5-aba0-235be8d2947b_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..109b145f1abdef67cb22cc92aa3ab7e738d431e5 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/258129ba-fc32-4df5-aba0-235be8d2947b_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b2c30c8456d7f6730a8f1da9444b66a6b4c96d2f0551dbcbfa381f3ec6a27903 +size 6256409 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/29ba67c1-2a90-4e8f-b997-01aefa1f19b3_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/29ba67c1-2a90-4e8f-b997-01aefa1f19b3_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..62e8ba5904b314353146c4438f2acd2456ff98dd --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/29ba67c1-2a90-4e8f-b997-01aefa1f19b3_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b455254d66a9b0e10ee2e80e397f7232d7e77cfece8ca58e0a1383f56b0c46b5 +size 10572690 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/2dfa0276-1a8f-4dd5-a070-37389d3b233f_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/2dfa0276-1a8f-4dd5-a070-37389d3b233f_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..f2e91b05ee841955e59867182c8b3247baae9168 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/2dfa0276-1a8f-4dd5-a070-37389d3b233f_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7309946fe9f4f280696a000145737be920304d5425fb7de53fb4c41addb2fc60 +size 8913400 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/35af024a-e166-4e4f-93ab-17c2ca902796_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/35af024a-e166-4e4f-93ab-17c2ca902796_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..6cf27ad8d5cf8d6cb3aa7496ad3231dc64951b05 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/35af024a-e166-4e4f-93ab-17c2ca902796_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:23afcf50690b82a1a6a5c07aaae06ac5ce24537aaba2f7c0535c0fa94bee78e0 +size 5700844 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/45627ac0-43c9-436e-8634-41e03b50cf19_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/45627ac0-43c9-436e-8634-41e03b50cf19_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..06e1c25568cba12382c13f02c24a885bcdeb44b7 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/45627ac0-43c9-436e-8634-41e03b50cf19_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:90354f225da98a2c62966d7eda8a1ce5a11ca944e305ffcb43bbefdf97b8e12d +size 6401566 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/45a8f4a5-4717-40c3-b06e-62bbcca0b1af_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/45a8f4a5-4717-40c3-b06e-62bbcca0b1af_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..b692705d7c4ffb61ec99e19d480bdc39575d3863 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/45a8f4a5-4717-40c3-b06e-62bbcca0b1af_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:91a4047e539e2ce9ef4197ebf37c369a599071378e13342c0fdbaa6b06c65e4e +size 8944973 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/49456225-3437-4b90-a5a0-77a53d3c9bf4_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/49456225-3437-4b90-a5a0-77a53d3c9bf4_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..78dc594dfbf1ec3b7d4f5ff5c1c16cc94b72ef72 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/49456225-3437-4b90-a5a0-77a53d3c9bf4_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:504174905a0da4a65a43066b0a0bd2d5269cef5ede7ff0d6bc8448260de0d696 +size 6184481 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/49797cdd-56e2-4d09-a7de-8272076a006b_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/49797cdd-56e2-4d09-a7de-8272076a006b_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..62e3fe31dc592a5c2dcb7b9ed101cb7080c5b041 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/49797cdd-56e2-4d09-a7de-8272076a006b_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0b3c78ca73246e7b70ca9f07a31a4680a04bfc122afb44fc6ccf1423c2721b7f +size 4170469 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/4ecac472-c319-4a27-9dd0-748b0383d659_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/4ecac472-c319-4a27-9dd0-748b0383d659_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..0ed4727934c47fa175484cf9987a274e254035d5 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/4ecac472-c319-4a27-9dd0-748b0383d659_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9e0b80a1a06f276bcb6fc4abf5beca889538ab65e7d7e16f1bc5718db2ad1480 +size 5813444 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/58213633-c297-44c4-8ee6-5cdca89d7e27_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/58213633-c297-44c4-8ee6-5cdca89d7e27_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..bc4b207b9ebda719f9c40585e6aea71315404cb5 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/58213633-c297-44c4-8ee6-5cdca89d7e27_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:779efddd4eb0c9ebb1067d3d3762056f81e58a101c1f698943caed7370bdeaa6 +size 4690753 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/5d0451d0-2465-4267-aa89-854951962bcf_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/5d0451d0-2465-4267-aa89-854951962bcf_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..d8d66c9f52dca440a705af2eebeb1047cbe5ca1e --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/5d0451d0-2465-4267-aa89-854951962bcf_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b651809b59663c0585798ea79a9ea78ea05ebdf23f0b92a92a4a5359431fdae3 +size 4623504 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/5e2b48e9-a9a3-488c-87de-439936b68f2a_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/5e2b48e9-a9a3-488c-87de-439936b68f2a_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..e4fe69cd40e32570d557ee9e222400a70d46cd7b --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/5e2b48e9-a9a3-488c-87de-439936b68f2a_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a92419846916773a6d573ab774afe621ce9221f3b3a2e9f33156d29bc6051fcd +size 5269745 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/5f92f264-3f91-49b1-987a-5c62fd8d0ffe_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/5f92f264-3f91-49b1-987a-5c62fd8d0ffe_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..d72d8d5d5b6ed79810528796d6005006b7b0624f --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/5f92f264-3f91-49b1-987a-5c62fd8d0ffe_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:43e513f4151914d6dad2074b143a1fb22444937eef8bc37f4829d8f04195b90b +size 5651245 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/6294b8f0-f45c-47dd-bfa7-bed11bb2f4f6_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/6294b8f0-f45c-47dd-bfa7-bed11bb2f4f6_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..e18f5b5836d988ce8a8272c1eccdfcdf5174cbff --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/6294b8f0-f45c-47dd-bfa7-bed11bb2f4f6_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f52ea8d921f059e199680bc5ce77a6ac88efc79a2b1d4acdfd7834770a51c85d +size 12793865 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/62a4567d-2433-4979-a4b3-78bfce8712f3_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/62a4567d-2433-4979-a4b3-78bfce8712f3_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..db447f3f02698edd502a5d4eeecf3ff88c36eaa7 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/62a4567d-2433-4979-a4b3-78bfce8712f3_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4008a825f6cb8a6f0a74de662edc4a92c72d30bab63155e778a9f291d95ca3f3 +size 7369711 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/659d5616-40b6-4e06-99d4-3fbe29669b66_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/659d5616-40b6-4e06-99d4-3fbe29669b66_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..beeacc46756161affc05b81059bb2223ff5bd831 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/659d5616-40b6-4e06-99d4-3fbe29669b66_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f6537fd9c737395adc3253f97115d8784486de037f6e8d037bf375169709413d +size 6366476 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/68cacdf0-44f0-498c-8657-ea0fe622ee1f_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/68cacdf0-44f0-498c-8657-ea0fe622ee1f_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..d3773d899202b84670d5f274edd0a1ee2766aca4 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/68cacdf0-44f0-498c-8657-ea0fe622ee1f_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8bdc14dc1d07edd4fdc2ff51a979d080fb4c00e85759b10256f3bd59736e5bc8 +size 4321653 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/6ddedd2d-2b5e-478d-a7fb-2c07dd3a8072_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/6ddedd2d-2b5e-478d-a7fb-2c07dd3a8072_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..3345b71a683099064fd635607229b1a312ba64e6 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/6ddedd2d-2b5e-478d-a7fb-2c07dd3a8072_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:08a82339660604635ae51c0f5163894f332b75c7933eaa3a44f9f6f13cbd6131 +size 4166451 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/6f04e8ed-23e1-40f3-a884-8d9f7a5d231f_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/6f04e8ed-23e1-40f3-a884-8d9f7a5d231f_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..d9d6aca828c74ee2933e49dde0c9cf817c18b3a4 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/6f04e8ed-23e1-40f3-a884-8d9f7a5d231f_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fa87fe7eaed327ffa45698e8ceb6e69b608f74f943d08742e0b5ffc808ecd853 +size 9636961 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/6fc1e119-0a68-4357-9105-84e43cb1b8b6_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/6fc1e119-0a68-4357-9105-84e43cb1b8b6_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..99aeeac240ba636d4085603d7ddb99b04b54fa0b --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/6fc1e119-0a68-4357-9105-84e43cb1b8b6_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:157416eccef043f8b9080a10bc533d2a4e8e8a39fb9244be04c152a00404fdf2 +size 4326494 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/744d8c08-8ebe-40d6-830e-815b75558c33_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/744d8c08-8ebe-40d6-830e-815b75558c33_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..ba544f89dd27834bc90a3a29eb90da0a013272df --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/744d8c08-8ebe-40d6-830e-815b75558c33_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:804f0873da93733479db8cacdcf32e313e15bfa16636d2374aa0ef9975f1cc0c +size 4295224 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/762f530c-f4d1-4b76-bd46-55efc1ff46e3_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/762f530c-f4d1-4b76-bd46-55efc1ff46e3_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..cd471a51bdee4a612a9254e9dd44fd98a82dcc30 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/762f530c-f4d1-4b76-bd46-55efc1ff46e3_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:619c16ff0686e76c8e48dbd2f959a3d5bd5d72f66ec28ac3477a224e21b29286 +size 7592143 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/82266441-a3ba-45e0-8df7-94277aea94e9_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/82266441-a3ba-45e0-8df7-94277aea94e9_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..879d94afc20527be05905253e99751be3f305e95 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/82266441-a3ba-45e0-8df7-94277aea94e9_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9af8e4612afac98dce529e4550d43694b69851d8feb01c19e406289d8b03f752 +size 6295015 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/88f8f2dc-ee58-4cdb-98d1-5507a6fe36bc_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/88f8f2dc-ee58-4cdb-98d1-5507a6fe36bc_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..6600292b843a98967ae0bc297edef6ea83de8849 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/88f8f2dc-ee58-4cdb-98d1-5507a6fe36bc_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ee0b2eb9544d5ca76547690b34e5355a7312d86e63b957631e027f8ab1200946 +size 8714871 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/8b642d6c-274d-43ba-ae53-c1ce562a2b39_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/8b642d6c-274d-43ba-ae53-c1ce562a2b39_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..e2beb419ac3e6782c0243ba511ca3fbe0fdecd27 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/8b642d6c-274d-43ba-ae53-c1ce562a2b39_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e8d77771017f278249d658708186b105dad73f8099abc1f7f51cfda5ffafe6ba +size 9315622 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/8b6b1ed7-e68a-4033-97d3-2e9db1a1941d_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/8b6b1ed7-e68a-4033-97d3-2e9db1a1941d_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..70451e5ae7749d88df12b9d81e6961be0bb60a8d --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/8b6b1ed7-e68a-4033-97d3-2e9db1a1941d_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f6a46acfb91ba9646e3105a026798cb9cdc07b0a3ab551d2147c7780bfcaaddb +size 8181432 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/8ba087de-e1f1-40c5-814f-b5554fb7ae29_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/8ba087de-e1f1-40c5-814f-b5554fb7ae29_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..64d82410da51a31cf5a12e338761ab0520ed8434 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/8ba087de-e1f1-40c5-814f-b5554fb7ae29_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:797f80e4386ad66a253123bfe08586d25979e46c0e88a5870fc18f7a98fdaf97 +size 4848326 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/a4b1a45c-9f47-4e21-9501-7fbdf396067b_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/a4b1a45c-9f47-4e21-9501-7fbdf396067b_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..8f84a793cc8f8e6a66a09548b3c9845a5110a38a --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/a4b1a45c-9f47-4e21-9501-7fbdf396067b_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ac51dd81b3f427f52dd83551f44d5a92a2bc89dd926963cad6bd11073d70d728 +size 5414148 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/b5b7b922-e5a1-41bb-a1bf-2d39120abee9_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/b5b7b922-e5a1-41bb-a1bf-2d39120abee9_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..ef9339a1054e32f62308ea58c27d7278a0955b8a --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/b5b7b922-e5a1-41bb-a1bf-2d39120abee9_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:302fbff5b7e03fde701dd95f090759be31679765d6f40879f52aea9e06ebcbe8 +size 8853635 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/b99111aa-00a2-490b-a5b3-28848e023684_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/b99111aa-00a2-490b-a5b3-28848e023684_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..c2cf82c01cfa0f1494d76d7c8432df85f4a41fe8 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/b99111aa-00a2-490b-a5b3-28848e023684_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fe7fdd8895dc78758ce1d5eb547029bd4c630283633aa683f42e47921b1b0fb1 +size 8243615 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/ba7e2f63-1851-4c13-9abe-e41ffd4136f7_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/ba7e2f63-1851-4c13-9abe-e41ffd4136f7_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..8e58e2551fef0cb99073027cc71d45a173628084 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/ba7e2f63-1851-4c13-9abe-e41ffd4136f7_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a55600805987c3e05899fe3248a39ae5c040c6e7e16f9c3b5919ceb66cfb95f7 +size 5848206 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/bd11cc91-2a4d-41b8-82e7-2e5d1854693b_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/bd11cc91-2a4d-41b8-82e7-2e5d1854693b_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..6799551a1b912bd111b6600bdc27178d5d97deb2 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/bd11cc91-2a4d-41b8-82e7-2e5d1854693b_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3542d972da4b363b0d1a21ec59941362032910741ef93879c2a81b9ab1414e09 +size 5594687 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/c5628725-dc96-4d87-8bd8-0f3e484c6055_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/c5628725-dc96-4d87-8bd8-0f3e484c6055_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..35bfe07c2b2f1116782404ded491c65d29e90daf --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/c5628725-dc96-4d87-8bd8-0f3e484c6055_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:266a8167fe2c8d2b4602ad902cd74c1519fc422bca08abb81bd1658ff10b9a9a +size 9018031 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/d647f547-4801-4be9-95dd-783fbda55f71_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/d647f547-4801-4be9-95dd-783fbda55f71_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..08081cda76fdca70c09eac904aa2c68fb243fe05 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/d647f547-4801-4be9-95dd-783fbda55f71_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1655aced1b8b9c1758c0cc79f4594da2a44500a062482dfa2a0ed936562bc899 +size 5945977 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/da522768-15c3-466c-a2bb-2e024643d960_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/da522768-15c3-466c-a2bb-2e024643d960_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..dd57933bba138462b92b08f426891a2438b79f62 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/da522768-15c3-466c-a2bb-2e024643d960_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f10b65055e9a4dd458e0cbafee68c44d70d6eb17e6b7142ffa16dc9f54f72298 +size 9436088 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/dd52e274-0445-4568-afb1-01cc5a44ecf3_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/dd52e274-0445-4568-afb1-01cc5a44ecf3_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..ff23ac6b576ff8dd4034813638ec2a424cf60b57 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/dd52e274-0445-4568-afb1-01cc5a44ecf3_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9c8f5d6346819f75b3ca4a23262f5c09198f321b58812234e278a2cabe386c7d +size 6635425 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/dfbf3d09-8c88-445b-a43e-a4ed2d2747a5_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/dfbf3d09-8c88-445b-a43e-a4ed2d2747a5_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..e5d655dbe5be2e971bf6332938b7a79e312a9317 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/dfbf3d09-8c88-445b-a43e-a4ed2d2747a5_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6c658ca2d2ade720099971e39ff14f968acaca5e0ed0be716c0b682d6e1646cf +size 7482008 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/e73723a4-e418-42ca-8f50-958236d9db4a_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/e73723a4-e418-42ca-8f50-958236d9db4a_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..7a089bc8cfcdc0a620b90d85127650b49d569784 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/e73723a4-e418-42ca-8f50-958236d9db4a_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:74cceda85c92a38ee9b9a58e4f337f6960ae81d6c77440430fdc43d726ab961a +size 8732594 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/f67a7b4f-635a-4503-9a21-3084ac41e5ae_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/f67a7b4f-635a-4503-9a21-3084ac41e5ae_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..8a1c2a6027e7bcd0d16a52ed8f142d631877bea2 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/f67a7b4f-635a-4503-9a21-3084ac41e5ae_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bd48461384aaaeebaaa28f290f2266224211110afd491572661c066ca578b3f7 +size 6408569 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/fbdbc376-b6ce-4007-a788-6cd681db5617_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/fbdbc376-b6ce-4007-a788-6cd681db5617_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..55d8c5e37981ff0a038e259e6249a27d1d3ac7d8 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlacematToBasketSplitA_GR1ArmsAndWaistFourierHands_Env/fbdbc376-b6ce-4007-a788-6cd681db5617_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:52dd02adda01e4ac746ce3e5862e42265ba4b64ccb08fb7cdfa6ffde76624390 +size 5536580 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/0c607c8b-63cf-47bd-91ce-00073fa5f018_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/0c607c8b-63cf-47bd-91ce-00073fa5f018_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..e76c66e24bc0c1a3256f794d3fd7b6a23437471b --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/0c607c8b-63cf-47bd-91ce-00073fa5f018_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:64d3bb1f38e6c12e21a96ef0ea4cf807a269e9da24861b6525afad867744d735 +size 5901132 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/0def4f5f-a2fe-486f-93e2-0ea64d7a4f92_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/0def4f5f-a2fe-486f-93e2-0ea64d7a4f92_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..0dc16aab6059b7f9992048847d572341f2cc7bcd --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/0def4f5f-a2fe-486f-93e2-0ea64d7a4f92_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:105e579d4aa8d2a748d4048fc0aba07f17b536449ff6c8d787a91cb183bff5fc +size 4736751 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/1858e8e1-808b-40e4-8265-b0cd876f755b_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/1858e8e1-808b-40e4-8265-b0cd876f755b_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..138c098554e3bf43c2f6d6c02bfa99a03fd2648a --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/1858e8e1-808b-40e4-8265-b0cd876f755b_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e732c2a4242e2c5c4b9a1f6da64ef08de7270d08f9b8dd5392f7ed58bbaeb53e +size 4606133 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/19037a5a-deb3-42f1-b883-d0355efee8fe_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/19037a5a-deb3-42f1-b883-d0355efee8fe_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..36e9c11cc25123fc0e091b9127b86e7cc4e5ef85 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/19037a5a-deb3-42f1-b883-d0355efee8fe_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9ecc335b5f95af0932343c6b103221acb76eebe99ccc2792446e4119cb3ec7c0 +size 8038425 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/269c0890-f58c-4ae9-8009-6cc1b74d65f9_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/269c0890-f58c-4ae9-8009-6cc1b74d65f9_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..3eaf4c68d88819902354c0796e9264851d462d25 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/269c0890-f58c-4ae9-8009-6cc1b74d65f9_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:92f97c4827bd95fa24a71c5ad67226cc1fb1f7c127eb04ac8c6288f498c6c34b +size 4886317 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/2c2bf082-0fe3-4668-9a7d-1e174cce4c0b_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/2c2bf082-0fe3-4668-9a7d-1e174cce4c0b_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..e350d4f3ab5b452cf15a0412adfde75136a3e703 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/2c2bf082-0fe3-4668-9a7d-1e174cce4c0b_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c930f43ba923add361b68f0a7279a9e41d6a9f48f59062615305b9aafc9af88f +size 7995689 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/377e7b1e-c810-4108-b18d-5d4532a765b3_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/377e7b1e-c810-4108-b18d-5d4532a765b3_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..61f7bda85ca7c239d027c124b69959a7ac7d9522 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/377e7b1e-c810-4108-b18d-5d4532a765b3_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d175922a6f28e0be3e71e4a1bbdc69382e8b12ab101804a8a891ebc0504902e3 +size 5372850 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/39916e11-6a97-4cd7-ac00-ee94c76f3f93_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/39916e11-6a97-4cd7-ac00-ee94c76f3f93_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..26f8b67b356e71ddb5e69b937c4381d64e5bc764 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/39916e11-6a97-4cd7-ac00-ee94c76f3f93_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:daff82ab385fda456f25c9c92c1f0ace3d2b57f5689a79bb6f43aea400201af1 +size 5328214 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/3d7f7255-da89-40ea-87ec-f9aa0f85a726_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/3d7f7255-da89-40ea-87ec-f9aa0f85a726_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..a9079049d8176082b075b54288b2ce3d790b6cc3 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/3d7f7255-da89-40ea-87ec-f9aa0f85a726_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9e149a993ded1023728ddf22b97bcc6e9404c5366eae163b7fcf845839380867 +size 6863421 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/4254fea1-a0ac-472f-8db0-1c5e87103151_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/4254fea1-a0ac-472f-8db0-1c5e87103151_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..3dac0f7705f7858c501bb30d70b00de173bd5228 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/4254fea1-a0ac-472f-8db0-1c5e87103151_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:39861de4831224ec019d06565348344a9dfddce7f39ba05912a21636c84fd4f3 +size 4309046 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/43b3c7b3-4d85-4e4e-be26-6ca378809ccc_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/43b3c7b3-4d85-4e4e-be26-6ca378809ccc_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..4c1e7b0f229927b801138f3dae6a9516f75a9b5f --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/43b3c7b3-4d85-4e4e-be26-6ca378809ccc_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:34f0224edc3344b98389bdc5876a16819f7ee9d9c8f49c105ed66763e23a0bb2 +size 3983929 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/44ae8f35-4bc5-4e15-b0a2-730d236e865b_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/44ae8f35-4bc5-4e15-b0a2-730d236e865b_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..1a4c1f2b44f18c743ae510d107f516c656c2cca3 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/44ae8f35-4bc5-4e15-b0a2-730d236e865b_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c69675065b459d395289d14f5c74252d5d0557af9d83412f54e06ac1508d3e58 +size 6831076 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/464e8749-641a-4215-bc5c-c6e59ef12e82_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/464e8749-641a-4215-bc5c-c6e59ef12e82_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..00a60276db66a0c057efbde51e676c320f57d57e --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/464e8749-641a-4215-bc5c-c6e59ef12e82_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:23ae1e31dc8fd53447b83cd8170e5caa2640cb8de172028ca6b4dd26cb4b2a1e +size 4538336 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/4ef16362-02b1-4153-a2ee-dbf2596052ae_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/4ef16362-02b1-4153-a2ee-dbf2596052ae_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..07c511df25934ba310b09b4f61744a5d5faeaf5f --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/4ef16362-02b1-4153-a2ee-dbf2596052ae_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:42eda0b12eda28a1607eafe7a9bfa2446eee5cfb77a7e7b6d5692adab61435dc +size 4470754 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/56328681-8ecf-43e1-a04c-7777a13c072a_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/56328681-8ecf-43e1-a04c-7777a13c072a_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..75d0543dbff1d881d446320d143a10808a3160ac --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/56328681-8ecf-43e1-a04c-7777a13c072a_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e15c4ccdc94f8b20846344f0c301c70d483f9bdb77c2bc2460585bf96e8f28e4 +size 4427995 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/59167abc-72e6-4f3a-84f6-90342aad3c25_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/59167abc-72e6-4f3a-84f6-90342aad3c25_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..759c6f59c04fe2c58a209091163a66fea882d9de --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/59167abc-72e6-4f3a-84f6-90342aad3c25_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d45f2bb88cd0cfb8ae0bf2d5cf5d699f4e563a487a97760bcfee6ae6fcfeba41 +size 5746939 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/5ccb7a7f-d18e-4ec0-9214-b1fd288c2edf_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/5ccb7a7f-d18e-4ec0-9214-b1fd288c2edf_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..e1ae4e0197a1207c12d7f045848a1534c87ef5ee --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/5ccb7a7f-d18e-4ec0-9214-b1fd288c2edf_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:60008c5eb77eff172c60b24a01a802dca39b92e9aeba4cf6064874d56f6577da +size 4510705 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/615b90b8-d2d2-432e-9712-28571be5533b_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/615b90b8-d2d2-432e-9712-28571be5533b_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..008e8c633b9e821d49c38e9c56104715bcabd862 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/615b90b8-d2d2-432e-9712-28571be5533b_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5b6dee4ef867ce6110c72aa582fc20660eb68ac2357de8dbf9d5aae9738e8f95 +size 6894683 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/632388a8-e8e9-4343-b29e-b160940efb02_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/632388a8-e8e9-4343-b29e-b160940efb02_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..927cf1569160b5981304c502e355cc239f46dbae --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/632388a8-e8e9-4343-b29e-b160940efb02_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e3287f35889e502318ff6326fbba8198ebd037baeb900fb391682a30eefbcce3 +size 5797523 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/69934ad1-bfe6-4d82-9459-78aafd5051d0_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/69934ad1-bfe6-4d82-9459-78aafd5051d0_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..9a5d09c85758647bfa2969f6436c278cfe76fce3 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/69934ad1-bfe6-4d82-9459-78aafd5051d0_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f50af136dbcf1872b3bf494e548499ae8dfd99dcd6e9a2b07d2429a4b0469a82 +size 6560332 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/6f22b8df-78cd-45c7-9bca-10d7bec2cce3_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/6f22b8df-78cd-45c7-9bca-10d7bec2cce3_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..d2e0f08d1a2ef4704c438f1c41e867a6732a4686 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/6f22b8df-78cd-45c7-9bca-10d7bec2cce3_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e582ca904be05a85df3998c1ce9d5807e53b03bb8312d29c2143b46d7927d6d5 +size 5570022 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/71453733-588f-4aa8-befd-a42d7a029526_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/71453733-588f-4aa8-befd-a42d7a029526_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..fcb35ef5330d813e0dfdbae422fb82eded768c78 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/71453733-588f-4aa8-befd-a42d7a029526_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:53178f48e9d2993d44e2db902fa6a5a01f6dc48d93e59a09e9db1eb76e34d748 +size 3620941 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/7a6f9241-d973-46ec-8b06-e3f1fbb955e6_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/7a6f9241-d973-46ec-8b06-e3f1fbb955e6_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..c183737bdb9a5ff23095e4da81d91fcc96cc1274 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/7a6f9241-d973-46ec-8b06-e3f1fbb955e6_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6e53e5b7704fc300cb3975155d8692b35d64974bbb93ddd0bfb458c18927ab36 +size 6893520 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/7c1c7a0c-324e-41d6-99a0-3cc1215450ff_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/7c1c7a0c-324e-41d6-99a0-3cc1215450ff_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..acc1444ae2086a562a7d65669e2241c0b4939970 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/7c1c7a0c-324e-41d6-99a0-3cc1215450ff_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a305a001864a49ce6d7d6bfc7fbad45075eeeba5e35ee4da5c6568756e2ad998 +size 6704662 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/864e2430-d4dc-47a6-ac03-a5d15cc4f715_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/864e2430-d4dc-47a6-ac03-a5d15cc4f715_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..41fd70a7abce100bd9bf90e69b9cb710e5f51736 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/864e2430-d4dc-47a6-ac03-a5d15cc4f715_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7376d3ea7da59697ce0635be6b1958962290e6f2e1bdbfae715a47b1e9e99c71 +size 5422807 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/8a948ae6-e0a9-4cfe-b65f-28ad7bc2def9_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/8a948ae6-e0a9-4cfe-b65f-28ad7bc2def9_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..28748540096b37083555e8eb255fcbd8355b2d46 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/8a948ae6-e0a9-4cfe-b65f-28ad7bc2def9_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c5b688613b8822b198101f096d2c00d0e85a20756017ee0ec37da6bd1fb2d17b +size 4386985 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/8ab4f62d-0013-436b-8009-6f4ccea79f83_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/8ab4f62d-0013-436b-8009-6f4ccea79f83_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..7e37073fac526bb137b0d764c7d3059e4fd01224 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/8ab4f62d-0013-436b-8009-6f4ccea79f83_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e535dd92b14d380e7326975e257246048c6a0c8a9902b81287424496515e2807 +size 3113233 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/8daf50f7-c294-4fa9-83ce-a40196764a1f_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/8daf50f7-c294-4fa9-83ce-a40196764a1f_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..60da0f7f091d8ac0ab689dc9140b658a779c5a67 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/8daf50f7-c294-4fa9-83ce-a40196764a1f_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ffc29ccd9314e1f646b79a6676f7666c2be31fa2a9780ba51ad0703b864d09e7 +size 5536640 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/8f9fbd47-b72f-43c4-ae43-96d911628b1a_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/8f9fbd47-b72f-43c4-ae43-96d911628b1a_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..0a096fa679edb539e8d43126af28e448e0fcf08d --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/8f9fbd47-b72f-43c4-ae43-96d911628b1a_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e76c32950e324a1fd03cf7c08bea8f07a3748c36857eac318a18c24dad6ac72b +size 6944710 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/913d0bb3-8e41-412c-9fce-3a54491a328e_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/913d0bb3-8e41-412c-9fce-3a54491a328e_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..1ed7511351b82794197b80cb753a56c3b853c27e --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/913d0bb3-8e41-412c-9fce-3a54491a328e_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bd6ffc654061cf534251705d22578f0242b2e5e3cf123e5c0d86d19e8a3558e0 +size 5542235 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/91445ce2-23bd-4db0-aa71-e5160b83ede5_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/91445ce2-23bd-4db0-aa71-e5160b83ede5_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..54b975394d97ec364505b16e1a1d1550046d9bef --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/91445ce2-23bd-4db0-aa71-e5160b83ede5_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:568f7a30e6415dfc6acd481275adb6656eb02a9e9cb00b701a3b364d00b5e8a5 +size 5723673 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/93b9eb55-1079-4b40-994f-0ca1b6c13d4c_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/93b9eb55-1079-4b40-994f-0ca1b6c13d4c_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..914c93b05a7030c1c21e94a22c1738f7cb66dd12 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/93b9eb55-1079-4b40-994f-0ca1b6c13d4c_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4a079bc6961b2855b3c7312ca2668fa508d40ff56548df383a10e8e551642255 +size 6821233 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/9f4affb8-1613-49ff-a7f5-e870a36dfd1d_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/9f4affb8-1613-49ff-a7f5-e870a36dfd1d_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..9e45a267bdf38db6305d1b53880d326f99231998 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/9f4affb8-1613-49ff-a7f5-e870a36dfd1d_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7a7d36aacda1aae3a165ae9f913508b7f63947df3a6fa123f89066d4cc5e8aaf +size 4607685 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/a1e8c8e7-6222-482f-932e-a009fb3dfda1_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/a1e8c8e7-6222-482f-932e-a009fb3dfda1_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..b81ef8072df49c8475dedf0a29e62e26b609663d --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/a1e8c8e7-6222-482f-932e-a009fb3dfda1_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:76728a2442f8166eea1ba130247a3effec65f8c5e5c1b6ed630abab52267454f +size 5839385 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/b7328413-2ad1-451b-a0f3-55b9d3db1a27_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/b7328413-2ad1-451b-a0f3-55b9d3db1a27_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..42f45eb6577ecb278ed280e8fe3749ac4b673e93 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/b7328413-2ad1-451b-a0f3-55b9d3db1a27_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:da1e2487bb1912fc8fd9fdf5b544055cd68493bbc7da00ea8a349389667125a2 +size 4753309 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/c87f7af8-b22e-4dbe-b331-4ff7a183ec9b_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/c87f7af8-b22e-4dbe-b331-4ff7a183ec9b_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..93166b3e0eff32142d70596c46ae14966892ddb0 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/c87f7af8-b22e-4dbe-b331-4ff7a183ec9b_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a219f04defeb18ecda19660c60e9c7769701fce826b4cd2c6b020a1bc5a8c5df +size 6205366 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/cad29f9c-478e-433d-b8cb-9be990c2d4ad_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/cad29f9c-478e-433d-b8cb-9be990c2d4ad_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..33de8d4c2ab5a3829f784a9020810597542d7d4d --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/cad29f9c-478e-433d-b8cb-9be990c2d4ad_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:00def30904d2e5884b8f471814f43a646ec21ad8b5f7bab516fbdae0091cc298 +size 3910771 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/ce0a67d2-5d39-4986-8271-f14e1f5e0e04_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/ce0a67d2-5d39-4986-8271-f14e1f5e0e04_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..8cd9e49c759d4213c870f99696d4bb0430632dae --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/ce0a67d2-5d39-4986-8271-f14e1f5e0e04_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:def05298823917722eb47c7dfa0671cd09f2e9a3b9668cdc22c14345768563cc +size 7217395 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/d02f7a23-4f0f-43d6-89bc-e53f413cd589_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/d02f7a23-4f0f-43d6-89bc-e53f413cd589_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..30ec6664ca0392ef26f5ce586eccd90cbd5619a8 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/d02f7a23-4f0f-43d6-89bc-e53f413cd589_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fb90facda54aa86cb9c94a854d7a81af2cc85fd6c4f07f225354af0569c67f61 +size 8705795 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/d220af64-021d-40e1-a498-9b9ae90a5b76_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/d220af64-021d-40e1-a498-9b9ae90a5b76_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..3fd89a30743a35fa9864986f5cc4728fb1fd693c --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/d220af64-021d-40e1-a498-9b9ae90a5b76_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:68b4413c2f34849c2144fe5ef41641100a6a1704f700eeb614fa6db3fbd68d10 +size 10156725 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/d31eacc4-d75d-43f8-8bd5-41e4dd451749_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/d31eacc4-d75d-43f8-8bd5-41e4dd451749_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..478d881889df8305f196e8eeca2ed49f498312bd --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/d31eacc4-d75d-43f8-8bd5-41e4dd451749_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bde89e2dc07b0d88ebc386bfc5552d2b39bf819ca47be9a99778366fc622ced1 +size 4601267 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/d6b3292f-b8fb-4deb-811c-430068c3ec6f_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/d6b3292f-b8fb-4deb-811c-430068c3ec6f_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..c8428ec55731900f05a382d72d285cf533db0456 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/d6b3292f-b8fb-4deb-811c-430068c3ec6f_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a9d0fbd9f7146c6e6d41ae2bd2636a84832b99ba7f4a14835e3e11b19a19b757 +size 4103006 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/e33e0359-e02f-4783-9bce-d77de7b33e9b_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/e33e0359-e02f-4783-9bce-d77de7b33e9b_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..47dda37f1c7fa6a76677b429fb7866bafc431ce0 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/e33e0359-e02f-4783-9bce-d77de7b33e9b_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5f391caecb009cad87f044fdd0cffc2aa20376f47d4b6364787a5ba1ff6d6c64 +size 5937797 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/e64205a4-8df6-4e42-81f0-f3e60d351d8d_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/e64205a4-8df6-4e42-81f0-f3e60d351d8d_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..ee666f77464424273a6fd0b1a39aea56eac786db --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/e64205a4-8df6-4e42-81f0-f3e60d351d8d_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:298ecac02c551230cfb338dfd16a13d4997164e4d625b263cdb166840bccd7cc +size 8160822 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/e8f352a2-b67e-432b-86b4-3bd9cabc05b3_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/e8f352a2-b67e-432b-86b4-3bd9cabc05b3_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..d16a8a1687c998b63df65e33d9f7332249b7eb6f --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/e8f352a2-b67e-432b-86b4-3bd9cabc05b3_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:306895f99743b8ada94d623fcfb9aba1aedf56e84402e62051d154c842d0bd62 +size 5145891 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/ee88086c-4b89-4a27-aeec-fb44ad11cdbc_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/ee88086c-4b89-4a27-aeec-fb44ad11cdbc_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..40f05271fa134e04cad239a56e33a5a5032613c8 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/ee88086c-4b89-4a27-aeec-fb44ad11cdbc_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0ab3bb30b1873e279c563f9d65fa019ef928cac47c5a747e99fff7129990ac78 +size 4569924 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/f3f320e0-a400-4749-8a00-f1f8fa82fe1a_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/f3f320e0-a400-4749-8a00-f1f8fa82fe1a_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..76eff6b3d766ba83153a8a41c9239f119b8d1ae6 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/f3f320e0-a400-4749-8a00-f1f8fa82fe1a_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f2ffa4b6c417103cf4c18843c26d1c32fb775b66bc3683dda4e6b935eabc33d9 +size 5560246 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/fa194b09-4c89-4eda-81db-0aa71d022c1c_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/fa194b09-4c89-4eda-81db-0aa71d022c1c_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..74b48fc02c1427b9ad7d3955b7bd594134461c91 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/fa194b09-4c89-4eda-81db-0aa71d022c1c_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f629cb2871b00a756b15b868eb35155ab3813799b38a1579e411ad2b212bc98a +size 4326228 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/fcc36c38-9b5d-44b4-abfb-7ce65d301a2a_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/fcc36c38-9b5d-44b4-abfb-7ce65d301a2a_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..c00ec121f7a2efb6b207164fbffaade99e391696 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/fcc36c38-9b5d-44b4-abfb-7ce65d301a2a_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6ddffad219fb95e4fe0bea69d53fc2145d3cf400a3a45561cc9c4ca003b93db5 +size 4538057 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/fd499bc3-9967-48fd-a7a1-ed81edaa822b_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/fd499bc3-9967-48fd-a7a1-ed81edaa822b_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..90e7adb6b9898ae3f0b59751a0d3b8e9af96538d --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromPlateToBowlSplitA_GR1ArmsAndWaistFourierHands_Env/fd499bc3-9967-48fd-a7a1-ed81edaa822b_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3725a064e6e8c86e13dedb8670661a41d0b69ac641f458840e945989e07cfb82 +size 5563390 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/0404ed0e-1cfb-4c51-987e-0d9b0c42e76d_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/0404ed0e-1cfb-4c51-987e-0d9b0c42e76d_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..a8b886d72a04174bd531ea902ed07e717f76b2a1 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/0404ed0e-1cfb-4c51-987e-0d9b0c42e76d_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3c8382e2d7b89a2a85bac8acd68bdac0cc119fb1a266ffeef707a5e7a618e5c0 +size 5988843 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/04de685c-871e-4463-a5d9-1871b0ec42b9_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/04de685c-871e-4463-a5d9-1871b0ec42b9_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..8961cc9a493fc2b28153fcd1bae66fdef42b19b5 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/04de685c-871e-4463-a5d9-1871b0ec42b9_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9bc500a9e8808fb17151bf9f027269bcf08b26e048243979bd7d6c3abe14900e +size 5815020 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/085542a2-bed7-413f-b5d3-2797b56efc96_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/085542a2-bed7-413f-b5d3-2797b56efc96_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..3c2ec904ea2dce82aa6051798d092690e9c71fbe --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/085542a2-bed7-413f-b5d3-2797b56efc96_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:90d946d32d64199b330d3d841bc3a460778536ee9f020369b7f39e209086f028 +size 5404189 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/0ff36697-fcaf-4f2f-988f-ec29ad06af86_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/0ff36697-fcaf-4f2f-988f-ec29ad06af86_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..076e033f2099ce4cf487e1353bdcb163a4657595 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/0ff36697-fcaf-4f2f-988f-ec29ad06af86_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dedf971ae44895d347fd53b091e1836c6ac25b27896c1e3999d493bf53c781e2 +size 4016115 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/10336d73-0b18-4944-a859-55fda157cd09_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/10336d73-0b18-4944-a859-55fda157cd09_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..d3940bf01c2100bdd8ab0cab4f4b9384b4543454 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/10336d73-0b18-4944-a859-55fda157cd09_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:061880f3eaea6837c45ec545db601c0e23a94d854427bc4f4fcf89f12af8b79a +size 4372355 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/114a048b-e874-4167-a996-d46b60499629_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/114a048b-e874-4167-a996-d46b60499629_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..3b5939e4f1687297336d9117364d2c3f6f9b2b98 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/114a048b-e874-4167-a996-d46b60499629_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f2b8c7ad44599bd39e20971b7a9acde3c78f893dfe29b88d63da2b8187fa91fe +size 5777694 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/13664493-b6b4-4dcb-a7d2-6d9c39a73fc0_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/13664493-b6b4-4dcb-a7d2-6d9c39a73fc0_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..120382823ac06595d856859d46521c2a07850d16 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/13664493-b6b4-4dcb-a7d2-6d9c39a73fc0_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:41987b186e46eae9c99c36fe4fc0ed88747ccaaffed20b850d3234e5926d1552 +size 5938840 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/193c8599-681e-4270-aa46-0f3933f35f4d_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/193c8599-681e-4270-aa46-0f3933f35f4d_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..0bb1c09b819fd8fdc1a5d57cdffc538bf45ffcdb --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/193c8599-681e-4270-aa46-0f3933f35f4d_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b835a9200d2c952561814457280f1e17d1680947e8c29b34712319033fb8cb76 +size 8512190 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/1e6acd95-c63b-41d0-9901-c2bc2930a0e3_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/1e6acd95-c63b-41d0-9901-c2bc2930a0e3_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..8d63fec2074ae3323a590cf2a89a17b66859603f --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/1e6acd95-c63b-41d0-9901-c2bc2930a0e3_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:68630a2a472ab1b5ad3f26781428973f4ffc87bdb36878e2c4a78a32e7b64c80 +size 3528341 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/203fc780-f55b-46f0-98a0-5330f7cf8a87_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/203fc780-f55b-46f0-98a0-5330f7cf8a87_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..33a99489c61940fc811962d903e1c80f4320ac8c --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/203fc780-f55b-46f0-98a0-5330f7cf8a87_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b436e6d189c98418d58e2443911e443f8d71a7e6ce8226a4721618ddf01d7636 +size 4757312 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/2beb0ca4-ae34-4251-93c9-edc0687b3b83_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/2beb0ca4-ae34-4251-93c9-edc0687b3b83_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..37ffa631c813cc5b2d67d790b9e3e3aa90ebc12e --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/2beb0ca4-ae34-4251-93c9-edc0687b3b83_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4b3118d8db8d4cb1c86180fe117fd4a5c638911290754be992642d48a176bc49 +size 5686794 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/322072ab-d665-4e71-92a2-eb920ba7683d_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/322072ab-d665-4e71-92a2-eb920ba7683d_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..4b0339b4a1647462e15d74a72d8f558ebd96644a --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/322072ab-d665-4e71-92a2-eb920ba7683d_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5dd465c77df9ec90cef4a37ab1d8470e3cd36dc74a72bc527c4d75b5ec718831 +size 6119955 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/32e19f5b-24ce-48e1-9a45-3ab64fe6fb53_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/32e19f5b-24ce-48e1-9a45-3ab64fe6fb53_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..5dd6c8c6730d263998877a8ef6acf7d06b27126c --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/32e19f5b-24ce-48e1-9a45-3ab64fe6fb53_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1548edb0dcfc147795b71f6a91dbe356d13e3425ac01a6200b5df79cc5d1240c +size 3729296 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/36d50f10-ed79-4bce-9705-d030dbd2c256_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/36d50f10-ed79-4bce-9705-d030dbd2c256_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..9da9c163396256d2b558b1b96ed5af43651a6a82 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/36d50f10-ed79-4bce-9705-d030dbd2c256_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:62abcb00e8cb1d9a267b1b078341c60163fb833067187114ee688b201c86c05f +size 2830544 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/404c1525-1667-4d16-9fbc-84d1475fa530_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/404c1525-1667-4d16-9fbc-84d1475fa530_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..af8835a112f037d7cfc9249caa317e82d0ade15d --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/404c1525-1667-4d16-9fbc-84d1475fa530_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4378144d8664a25be6d8304af73bf9f25bec71d8099d65d517783fa8b7a9b299 +size 6229624 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/44375c67-9f5e-498d-ac42-762488fe3c24_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/44375c67-9f5e-498d-ac42-762488fe3c24_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..bb2db45006670da4efbd9d67a716f579420c0025 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/44375c67-9f5e-498d-ac42-762488fe3c24_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:94b0593b94f46d1ef2a8e8f9683b711ed45f22aa966718c9c0eababc83ea9f3c +size 4874492 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/47aacd87-4fb3-423b-9f9d-f565914324ac_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/47aacd87-4fb3-423b-9f9d-f565914324ac_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..8a4e790180544d2003e06c0f1dc7038acd3388e5 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/47aacd87-4fb3-423b-9f9d-f565914324ac_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bfd6d221ecccb296304bbf665b84c256132e897dfa52f5bfffbb3b9f87fe22de +size 6242919 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/483a1c02-7093-46e2-8f0b-1c5db8decc5d_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/483a1c02-7093-46e2-8f0b-1c5db8decc5d_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..ac1bf106600b2a5edfec4e133b91638dea660840 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/483a1c02-7093-46e2-8f0b-1c5db8decc5d_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:57a38b9034bd255dd8e74f72f62291bd1bcd02c932eef68116117ff51b6a412c +size 6925464 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/4dd638f7-1b83-4838-871b-54f35e0027fd_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/4dd638f7-1b83-4838-871b-54f35e0027fd_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..ad3c8bcc99dd57179449c60dc9c3743d32a1e7cf --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/4dd638f7-1b83-4838-871b-54f35e0027fd_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:49ea6965ad6b70dc9b26d6395cc62d15a2a062194eb76953eb446769d2386a72 +size 3076707 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/555766f1-930f-4bba-ad1d-74100f147a6c_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/555766f1-930f-4bba-ad1d-74100f147a6c_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..3a397ac9574a427fbcb0fe6f5a2a81bfcbaa5b22 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/555766f1-930f-4bba-ad1d-74100f147a6c_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7734978b55e46b215d04e71d00daaa915a0b8add5444fc77b93f335e10bb7abb +size 5741430 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/5b0e1630-0226-4845-a78b-b5a65c710b85_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/5b0e1630-0226-4845-a78b-b5a65c710b85_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..b0b83d914cdf33bf7f297a2a867f22a4be89a3ed --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/5b0e1630-0226-4845-a78b-b5a65c710b85_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:03d01cbb9047e223da3888709e5b0483e71c215707af2780871174fa3e780a7a +size 4621870 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/5c4cf8a1-0cdf-450b-ab51-5dc5f496de2f_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/5c4cf8a1-0cdf-450b-ab51-5dc5f496de2f_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..0607892146bdda4f38595bc20880490e304fe011 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/5c4cf8a1-0cdf-450b-ab51-5dc5f496de2f_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:794713fc692020c4bfb559fb8a3b42a8d8517a29d60793e269cd9763aa1523cc +size 5165894 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/5d837066-5253-4fa3-bb76-5d30447fa4e7_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/5d837066-5253-4fa3-bb76-5d30447fa4e7_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..2b6b675156a43c2992bae6dccfddef16bc1d064d --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/5d837066-5253-4fa3-bb76-5d30447fa4e7_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ff1f2a5692e7d41de6ff658d1b0702537df2af0a48012613af17357e877887f4 +size 4994940 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/6b86827a-a6bc-4c07-88b0-3276ec957134_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/6b86827a-a6bc-4c07-88b0-3276ec957134_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..e3ecc90b0e8954a4b03ea8620110a73a1e79c41d --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/6b86827a-a6bc-4c07-88b0-3276ec957134_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:657defc5362b4192134e8f2521eda3c24f5e0279ced7cc89952e60b1847db5e3 +size 3840816 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/6ef2bdce-7802-4053-a486-32435f36e204_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/6ef2bdce-7802-4053-a486-32435f36e204_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..8075f674e2c1513485afdef42767611b7422a761 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/6ef2bdce-7802-4053-a486-32435f36e204_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:018f06ed924bd9340dfcc6e2836c83e3cf6d9bbe1bc2919b0ebae5e7cf4910ab +size 3800792 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/6f6c8db9-7572-4a89-8c77-dd0de3f96460_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/6f6c8db9-7572-4a89-8c77-dd0de3f96460_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..82f189764c36562eaa24c6f1e693e1c5ab92d9d3 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/6f6c8db9-7572-4a89-8c77-dd0de3f96460_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f4ab8f019b14ddb18f0e62dd9e502b3ab8f75e1ff5b798af95b082a76925eab5 +size 6017080 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/6fa97f00-3555-4ca9-91e3-7418039de303_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/6fa97f00-3555-4ca9-91e3-7418039de303_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..9c9f323f236097460608e38f89407076e165d90a --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/6fa97f00-3555-4ca9-91e3-7418039de303_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8fd2d6349749bad4e9d0b11af1f0465081f608400cb910b75ccd09c7d3e535fa +size 5345417 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/7346c976-53cb-47ba-9c4d-3da57a3cb3a6_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/7346c976-53cb-47ba-9c4d-3da57a3cb3a6_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..39f08d29bbf96ad41d5c82701cf34c10c392aa2d --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/7346c976-53cb-47ba-9c4d-3da57a3cb3a6_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0c060feedf1ca4f2f7c78ae310c6140d405adffe47dbbc12491ba6ae8e715540 +size 6245888 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/754dec4a-0324-440b-b322-b127c8497b51_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/754dec4a-0324-440b-b322-b127c8497b51_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..393367efc811a50529b6765d712f5f3e2ba50ada --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/754dec4a-0324-440b-b322-b127c8497b51_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5196ddbfea9e1fed3a7978bae3e3f0feb28e18b1fe7ad827efcc286f487c746a +size 4320416 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/794e4c04-0975-4c55-9d63-c8011d5831c2_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/794e4c04-0975-4c55-9d63-c8011d5831c2_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..147cf9b123f020838fb7c2a343092eaf512f56b8 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/794e4c04-0975-4c55-9d63-c8011d5831c2_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1c6877a6872c0e7391d4e83016c99ad1e7af7b0cb1cac633486afb195a733f0a +size 4208997 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/7a2d9053-5e19-4951-ba80-b92399a95c30_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/7a2d9053-5e19-4951-ba80-b92399a95c30_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..7d696e02ea3eeb49cd90781f216ca1bca573fe2e --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/7a2d9053-5e19-4951-ba80-b92399a95c30_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3fa0f02620400b5750546c1e3844ed8779067d8dc4a0043fd959bbc8f7cc9adf +size 6906111 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/7cbdd96e-a15d-41c3-8929-0b6f71559c35_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/7cbdd96e-a15d-41c3-8929-0b6f71559c35_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..fdafe3af038ee916173e1d819afce26bd9b20852 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/7cbdd96e-a15d-41c3-8929-0b6f71559c35_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5c31f411af93c245cc185853fe357d002a822dca0a6d623437cf4141d359607a +size 5926636 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/7ed6810d-b4d0-452f-9bf4-e5960595762a_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/7ed6810d-b4d0-452f-9bf4-e5960595762a_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..f81d5d1036cabd2d04c1c496651a466382acd179 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/7ed6810d-b4d0-452f-9bf4-e5960595762a_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b67b366e5298581115a039530cf2e484953ac8c3b576590d3e94b0897cd99736 +size 7522507 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/906428f0-6a7b-46b8-87f0-e73a09f0b96b_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/906428f0-6a7b-46b8-87f0-e73a09f0b96b_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..6917548ff9c2ce35a65da6816b407e26ba61ccee --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/906428f0-6a7b-46b8-87f0-e73a09f0b96b_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:00a8ece79964627a3f9af9c159822caca9cc1332ca6f50ee77cedaa3f68cea2f +size 5507044 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/908fa40d-de5f-4d66-90d7-020f9daec0f0_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/908fa40d-de5f-4d66-90d7-020f9daec0f0_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..54dd7e50a2c5085355bae335ac383a2b8d539730 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/908fa40d-de5f-4d66-90d7-020f9daec0f0_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5c82b983bbd4c0350275d2e2ce667d9e1f23213950af2c881ce58ffcf7066883 +size 4316247 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/9fbad204-a7fd-4239-9db5-87f36aecb8c8_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/9fbad204-a7fd-4239-9db5-87f36aecb8c8_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..326dc8d4d419be47e5f1c0aa0bd2fc94a2fe7456 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/9fbad204-a7fd-4239-9db5-87f36aecb8c8_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8a5a8b1a902b1214f0a3a3114cc73c35d7ac39e460a0c784b4a921ab7217985d +size 3890452 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/a517338f-7ad7-41ad-b5ef-09de29bf6144_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/a517338f-7ad7-41ad-b5ef-09de29bf6144_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..786c6ff546919f01eed3e58cb55f9e5a5a61dce6 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/a517338f-7ad7-41ad-b5ef-09de29bf6144_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9eec29999d53ef1341adf84c178935d803393eb77b68a177d44d737cf3e9f0bf +size 4267603 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/a9ebd8fe-e5fd-4603-8ba9-12af5317a797_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/a9ebd8fe-e5fd-4603-8ba9-12af5317a797_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..71971ac5752c97d36332ff9c96e80a9f0aad92b8 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/a9ebd8fe-e5fd-4603-8ba9-12af5317a797_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:70c0ade1e94b31ad5aa23fcbce934a966ae161b2cc2088e10dab85aeee40fae0 +size 5093955 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/aeb1be92-8d8d-499b-9fd5-fed6344edf6a_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/aeb1be92-8d8d-499b-9fd5-fed6344edf6a_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..4724ab0d2f734d812333a8b9c75ff1e46e0fb9ed --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/aeb1be92-8d8d-499b-9fd5-fed6344edf6a_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0512c0491fc1fabc8b164fce98a0a7dd2fbb97d7807565b1b1e6f608b03ed7c2 +size 4350511 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/b5be42a1-9591-4582-8488-b9c8799dfe5b_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/b5be42a1-9591-4582-8488-b9c8799dfe5b_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..3c251897b2f359de1e1b1f67134afe6d907777a5 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/b5be42a1-9591-4582-8488-b9c8799dfe5b_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e94b81f305b9374643829d1ea88d1aa119ba1a333534ca81ea69b414e822419c +size 5870702 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/bd15ea1c-a344-4a9d-a682-4d1212148bf7_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/bd15ea1c-a344-4a9d-a682-4d1212148bf7_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..227fea8fff946f18ec18569e23f36490f9123240 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/bd15ea1c-a344-4a9d-a682-4d1212148bf7_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3a2910a498e14347ecdbd12b38e1405ea8092761919a4b29c8a3d5a04724a98c +size 7399834 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/d01ce947-9ce5-4ff2-a35d-23372bf158a5_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/d01ce947-9ce5-4ff2-a35d-23372bf158a5_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..51f6c0bde24c07b711d999dde144a7a28eb76714 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/d01ce947-9ce5-4ff2-a35d-23372bf158a5_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fb5bad344330bff9d78494f069bb12e2fd70ae581f99be70d0e213a25f4864a1 +size 5862526 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/dae5317b-d777-49a2-ace1-a33437d67f82_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/dae5317b-d777-49a2-ace1-a33437d67f82_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..482df200f84936d99629392c2961016990b1459b --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/dae5317b-d777-49a2-ace1-a33437d67f82_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4ddf69371860e72836c09084fe8c7bc1c21d81d96e6408f0f24236363a0404a2 +size 3347337 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/e0dac2f0-8149-4be2-9426-f8b0263f99e7_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/e0dac2f0-8149-4be2-9426-f8b0263f99e7_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..aa50bf1f65c86ccc92836f11cd19f7cc931f078b --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/e0dac2f0-8149-4be2-9426-f8b0263f99e7_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d7ace63b339604bf95eaade1f9232bb1f44e9c0fa19fb27c81e56f3fb496af1a +size 4852138 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/e32b2b75-ca02-4a77-9b32-36dbb5ff6708_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/e32b2b75-ca02-4a77-9b32-36dbb5ff6708_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..34d42fa4df0072dc0dfa83da561bd331518bc398 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/e32b2b75-ca02-4a77-9b32-36dbb5ff6708_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:713371f2ce3f2ae9890bfdf515bb21a7f1976989df7b863997803c0f3955fb1a +size 5594054 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/e3708ec4-45f5-4655-96f1-641763ba67d4_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/e3708ec4-45f5-4655-96f1-641763ba67d4_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..f426aefb9887c63f43f393f0837e23d07c2be424 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/e3708ec4-45f5-4655-96f1-641763ba67d4_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:598446e03b9d12a8b185a86ae82bcb2291e487a0e355abaf6fd134bf493d323f +size 5549156 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/e6c43dc5-17aa-42e2-b0a9-20ccce95bfb9_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/e6c43dc5-17aa-42e2-b0a9-20ccce95bfb9_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..5c7c2b61a0089de2f4b8da613cef55da08a0217c --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/e6c43dc5-17aa-42e2-b0a9-20ccce95bfb9_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e0ece0b8200c53cacc4ca56b15f13aa7d17669d32a483a012f31a2598cb70de1 +size 5198068 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/e7253a26-1ea5-4d15-a976-d7d5bf63767a_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/e7253a26-1ea5-4d15-a976-d7d5bf63767a_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..b39ba4833e683c7fddd14d56a8ef229b23c9d6e1 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/e7253a26-1ea5-4d15-a976-d7d5bf63767a_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d7c62abc1f8d5607c92167b68f5ce08aa2452b856a4f9332dd6955c444780045 +size 5708200 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/ea13d1c3-77b2-41ea-a27b-a4cd994eeace_success1.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/ea13d1c3-77b2-41ea-a27b-a4cd994eeace_success1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..d963f79323052fa513ad361340b154d1cfacfa56 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/ea13d1c3-77b2-41ea-a27b-a4cd994eeace_success1.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b3a5bc3f0a02a8140841924aa7c4887e904a1966d68b8efa1285b37c5bdc97ba +size 6459638 diff --git a/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/f8a8b99a-b780-47df-a51a-92822914fd0d_success0.mp4 b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/f8a8b99a-b780-47df-a51a-92822914fd0d_success0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..f84841480b87ee66d27688cae3db929271403b34 --- /dev/null +++ b/videos/pytorch_model/n_action_steps_10_max_episode_steps_720_n_envs_1_gr1_unified/PosttrainPnPNovelFromTrayToCardboardboxSplitA_GR1ArmsAndWaistFourierHands_Env/f8a8b99a-b780-47df-a51a-92822914fd0d_success0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:df53233a5800843e25e2df6062bc7ae9a0fbe01398051da58e7defaabdcaaaa1 +size 6448826 diff --git a/wandb/wandb/debug-internal.log b/wandb/wandb/debug-internal.log new file mode 100644 index 0000000000000000000000000000000000000000..8774701c142888d13a352198478572145c13cb90 --- /dev/null +++ b/wandb/wandb/debug-internal.log @@ -0,0 +1,12 @@ +{"time":"2026-04-19T11:14:33.799817575+08:00","level":"INFO","msg":"stream: starting","core version":"0.24.1"} +{"time":"2026-04-19T11:14:36.643300959+08:00","level":"INFO","msg":"stream: created new stream","id":"oh7yfg1j"} +{"time":"2026-04-19T11:14:36.643364261+08:00","level":"INFO","msg":"handler: started","stream_id":"oh7yfg1j"} +{"time":"2026-04-19T11:14:36.643454785+08:00","level":"INFO","msg":"stream: started","id":"oh7yfg1j"} +{"time":"2026-04-19T11:14:36.643474208+08:00","level":"INFO","msg":"writer: started","stream_id":"oh7yfg1j"} +{"time":"2026-04-19T11:14:36.643474275+08:00","level":"INFO","msg":"sender: started","stream_id":"oh7yfg1j"} +{"time":"2026-04-20T05:34:35.738131154+08:00","level":"INFO","msg":"fileTransfer: Close: file transfer manager closed"} +{"time":"2026-04-20T05:34:36.772832319+08:00","level":"INFO","msg":"handler: operation stats","stats":{}} +{"time":"2026-04-20T05:34:36.776247344+08:00","level":"INFO","msg":"stream: closing","id":"oh7yfg1j"} +{"time":"2026-04-20T05:34:36.776261896+08:00","level":"INFO","msg":"handler: closed","stream_id":"oh7yfg1j"} +{"time":"2026-04-20T05:34:36.776319028+08:00","level":"INFO","msg":"sender: closed","stream_id":"oh7yfg1j"} +{"time":"2026-04-20T05:34:36.776325037+08:00","level":"INFO","msg":"stream: closed","id":"oh7yfg1j"} diff --git a/wandb/wandb/debug.log b/wandb/wandb/debug.log new file mode 100644 index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391 diff --git a/wandb/wandb/run-20260419_111043-fabqp1or/files/output.log b/wandb/wandb/run-20260419_111043-fabqp1or/files/output.log new file mode 100644 index 0000000000000000000000000000000000000000..10285253f337aee221e7fb9ea6689f4ad03a4479 --- /dev/null +++ b/wandb/wandb/run-20260419_111043-fabqp1or/files/output.log @@ -0,0 +1,633 @@ +04/19 [11:10:45] INFO | >> ***** Training Configuration ***** train_qwenlatent.py:613 + INFO | >> Total optimization steps = 30000 train_qwenlatent.py:614 + INFO | >> Per device batch size = 32 train_qwenlatent.py:615 + INFO | >> Gradient accumulation steps = 1 train_qwenlatent.py:616 + INFO | >> Total batch size = 256 train_qwenlatent.py:617 + INFO | >> ***** LR Scheduler Debug Info ***** train_qwenlatent.py:619 + INFO | >> lr_scheduler type = + INFO | >> base_scheduler type = + INFO | >> initial last_epoch = 0 train_qwenlatent.py:623 + INFO | >> initial lr = [0.0, 0.0, 0.0] train_qwenlatent.py:624 + INFO | >> num_warmup_steps = 3000 train_qwenlatent.py:625 + INFO | >> num_stable_steps = 0 train_qwenlatent.py:626 + INFO | >> max_train_steps = 30000 train_qwenlatent.py:627 + INFO | >> accelerator.num_processes = 8 train_qwenlatent.py:628 + INFO | >> accelerator.gradient_accumulation_steps train_qwenlatent.py:629 + = 1 + INFO | >> trainer.mode = decay_aux_loss train_qwenlatent.py:630 + INFO | >> loss_weights_decay_steps = 5000 train_qwenlatent.py:631 + 0%| | 0/30000 [00:00] + - lang: 'unlocked_waist: pick the bell pepper from the cutting board and place it in the basket' + - dataset_id: 7 +Sample[1] content: + - action: [[ 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.02348 -0.484 -0.3774 0.03754 -0.0081 -0.6787 -0.2404 0.991 ] + [ 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.04648 -0.4612 -0.4019 0.04828 -0.01987 -0.662 -0.2373 0.991 ] + [ 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.07294 -0.596 -0.3765 -0.1198 0.02763 -0.5503 -0.2288 0.997 ] + [ 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.0938 -0.5557 -0.3882 -0.11304 0.02034 -0.4966 -0.2255 0.999 ] + [ 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.12494 -0.4995 -0.3967 -0.1017 0.0213 -0.429 -0.2181 0.9946 ] + [ 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.14 -0.4497 -0.387 -0.10394 0.01682 -0.347 -0.2089 0.998 ] + [ 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.1572 -0.4116 -0.3904 -0.0998 0.00977 -0.294 -0.1995 0.998 ] + [ 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.1718 -0.3813 -0.389 -0.10187 0.011375 -0.2452 -0.19 0.9966 ] + [ 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.1741 -0.3486 -0.3796 -0.1033 0.006474 -0.1978 -0.1875 0.999 ] + [ 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.1952 -0.3262 -0.3708 -0.1065 0.003218 -0.1626 -0.1699 0.9985 ] + [ 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.1306 -0.3943 -0.3413 -0.1598 0.01429 -0.1827 -0.2042 0.997 ] + [ 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.1306 -0.3943 -0.3413 -0.1598 0.01429 -0.1827 -0.2042 0.997 ] + [ 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.1306 -0.3943 -0.3413 -0.1598 0.01429 -0.1827 -0.2042 0.997 ] + [ 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.1306 -0.3943 -0.3413 -0.1598 0.01429 -0.1827 -0.2042 0.997 ] + [ 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.1306 -0.3943 -0.3413 -0.1598 0.01429 -0.1827 -0.2042 0.997 ]] + - state: [[ 0. 0. 0. 0. 0. 0. 0. 0. -0.3718 0.5107 0.1755 -0.1123 0.2 0.3003 -0.0616 0.956 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. ] + [ 0. 0. 0. 0. 0. 0. 0. 0. -0.3718 0.491 0.1671 -0.1059 0.2 0.268 -0.0616 0.956 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. ] + [ 0. 0. 0. 0. 0. 0. 0. 0. -0.3718 0.4717 0.1581 -0.0999 0.2 0.236 -0.06165 0.956 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. ] + [ 0. 0. 0. 0. 0. 0. 0. 0. -0.3718 0.4429 0.1434 -0.0901 0.2 0.1954 -0.06168 0.956 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. ] + [ 0. 0. 0. 0. 0. 0. 0. 0. -0.3716 0.4304 0.1364 -0.0901 0.2 0.1786 -0.0617 0.956 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. ] + [ 0. 0. 0. 0. 0. 0. 0. 0. -0.3713 0.4138 0.1268 -0.0901 0.2 0.1566 -0.0617 0.956 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. ] + [ 0. 0. 0. 0. 0. 0. 0. 0. -0.3716 0.3982 0.11847 -0.0901 0.2 0.1398 -0.0617 0.956 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. ] + [ 0. 0. 0. 0. 0. 0. 0. 0. -0.3716 0.3848 0.11206 -0.09015 0.2 0.1248 -0.0617 0.956 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. ] + [ 0. 0. 0. 0. 0. 0. 0. 0. -0.3713 0.3728 0.107 -0.09015 0.2 0.1133 -0.0617 0.956 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. ] + [ 0. 0. 0. 0. 0. 0. 0. 0. -0.3713 0.364 0.10266 -0.09015 0.2 0.1055 -0.0617 0.956 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. ] + [ 0. 0. 0. 0. 0. 0. 0. 0. -0.3713 0.3577 0.09906 -0.09015 0.2 0.1014 -0.0617 0.956 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. ] + [ 0. 0. 0. 0. 0. 0. 0. 0. -0.3713 0.3577 0.09906 -0.09015 0.2 0.1014 -0.0617 0.956 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. ] + [ 0. 0. 0. 0. 0. 0. 0. 0. -0.3713 0.3577 0.09906 -0.09015 0.2 0.1014 -0.0617 0.956 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. ] + [ 0. 0. 0. 0. 0. 0. 0. 0. -0.3713 0.3577 0.09906 -0.09015 0.2 0.1014 -0.0617 0.956 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. ] + [ 0. 0. 0. 0. 0. 0. 0. 0. -0.3713 0.3577 0.09906 -0.09015 0.2 0.1014 -0.0617 0.956 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. ]] + - image: [, ] + - lang: 'pick up the cup in the middle' + - dataset_id: 9 +Sample[2] content: + - action: [[-1.4124e-01 -6.2646e-01 4.2529e-01 -2.0477e-02 -1.6159e-02 -9.9060e-02 1.1407e-01 4.2145e-02 7.0654e-01 1.5930e-01 1.3533e-03 2.8760e-01 1.3123e-01 7.2815e-02 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 0.0000e+00 -7.2998e-02 3.3960e-01 -6.3477e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-1.7310e-01 -6.6553e-01 4.1919e-01 1.5106e-03 -3.4363e-02 -6.9885e-02 1.2372e-01 4.7729e-02 7.7295e-01 1.9482e-01 -1.1055e-02 3.1567e-01 1.1481e-01 1.0144e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 0.0000e+00 -7.3853e-02 3.5205e-01 -6.3281e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-1.6858e-01 -6.6260e-01 4.3115e-01 1.0849e-02 -2.7084e-02 -7.3120e-02 1.1359e-01 6.1127e-02 7.5977e-01 1.9373e-01 -1.4172e-01 2.9541e-01 1.9617e-01 1.3037e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 0.0000e+00 -9.4360e-02 3.4375e-01 -6.3428e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-1.7114e-01 -5.8887e-01 4.5312e-01 -1.1185e-02 -3.7193e-03 -1.0864e-01 1.2793e-01 1.0675e-01 8.2715e-01 2.0532e-01 -1.9299e-01 2.8076e-01 1.8579e-01 1.4880e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 0.0000e+00 -8.7952e-02 3.4399e-01 -6.3232e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-1.6931e-01 -6.3135e-01 4.3945e-01 -2.4185e-03 -1.8112e-02 -8.7402e-02 1.1926e-01 1.2097e-01 8.0566e-01 1.4624e-01 -2.0557e-01 2.1643e-01 1.7749e-01 1.0278e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 0.0000e+00 -1.0858e-01 3.4424e-01 -6.3135e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-1.7786e-01 -5.9668e-01 4.5825e-01 4.7302e-02 -1.3649e-02 -1.0010e-01 7.9224e-02 1.0760e-01 8.8232e-01 1.5515e-01 -2.3303e-01 2.0959e-01 2.2205e-01 1.4709e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 0.0000e+00 -1.1755e-01 3.5156e-01 -6.3477e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-1.7224e-01 -6.1377e-01 4.4580e-01 -1.3580e-02 -7.5722e-03 -9.9915e-02 1.4087e-01 8.5754e-02 9.2676e-01 1.4099e-01 -2.5317e-01 1.9385e-01 2.4670e-01 1.6760e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 0.0000e+00 -7.6782e-02 3.4399e-01 -6.3135e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-1.9165e-01 -6.6064e-01 4.3994e-01 1.8661e-02 -2.8992e-02 -6.5247e-02 1.2744e-01 1.0480e-01 8.9697e-01 1.1261e-01 -3.0444e-01 1.5942e-01 2.5293e-01 1.4624e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 0.0000e+00 -1.2622e-01 3.4521e-01 -6.3184e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-1.9739e-01 -6.3623e-01 4.5166e-01 2.5360e-02 -1.7242e-02 -8.0383e-02 1.2073e-01 1.1505e-01 9.9219e-01 1.5662e-01 -3.2764e-01 1.6516e-01 2.3914e-01 1.8994e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 0.0000e+00 -1.1334e-01 3.4644e-01 -6.3232e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-1.6577e-01 -6.5723e-01 4.3140e-01 1.1909e-02 -2.1606e-02 -8.4656e-02 1.2067e-01 9.2041e-02 9.8779e-01 1.1560e-01 -3.7354e-01 1.3989e-01 2.7539e-01 1.8970e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 0.0000e+00 -9.9182e-02 3.4106e-01 -6.3184e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-1.5881e-01 -6.4941e-01 4.2920e-01 -3.9444e-03 -2.3010e-02 -8.5510e-02 1.1682e-01 1.1182e-01 9.9365e-01 1.1517e-01 -4.2505e-01 1.3000e-01 2.7246e-01 1.9397e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 0.0000e+00 -1.0510e-01 3.3374e-01 -6.3086e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-1.5259e-01 -6.3623e-01 4.3848e-01 1.3981e-03 -1.8723e-02 -8.9539e-02 9.2041e-02 1.1432e-01 9.9609e-01 7.9285e-02 -4.5044e-01 7.9651e-02 2.7417e-01 1.8372e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 0.0000e+00 -1.1890e-01 3.3398e-01 -6.3037e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-1.6235e-01 -6.0254e-01 4.4067e-01 -1.0956e-02 -2.1210e-02 -9.3506e-02 1.2213e-01 1.0950e-01 9.9658e-01 8.4351e-02 -4.9756e-01 4.3732e-02 2.7979e-01 2.3987e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 0.0000e+00 -1.1884e-01 3.2080e-01 -6.2891e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-1.5540e-01 -6.2988e-01 4.4067e-01 -7.7972e-03 -1.7822e-02 -8.8989e-02 1.1401e-01 9.7839e-02 9.9756e-01 8.4839e-02 -5.2344e-01 1.7929e-02 2.6587e-01 2.8149e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 0.0000e+00 -9.6619e-02 3.3008e-01 -6.3037e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-1.7554e-01 -6.4893e-01 4.3726e-01 -9.2773e-03 -2.0065e-02 -8.0078e-02 1.3293e-01 8.8989e-02 9.9805e-01 5.6274e-02 -5.6201e-01 7.6628e-04 2.8223e-01 2.9565e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 0.0000e+00 -1.0944e-01 3.3496e-01 -6.2988e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00]] + - state: [[ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -2.0874e-01 1.9006e-01 9.0088e-02 -9.8389e-01 -1.3843e-01 -2.7344e-02 2.4146e-01 9.7803e-01 9.8193e-01 9.9609e-01 1.7761e-01 9.9023e-01 9.9951e-01 9.7021e-01 -2.9150e-01 -1.6895e-01 + 2.7734e-01 -9.5947e-01 9.0381e-01 1.0126e-01 -2.2598e-02 9.5654e-01 9.8584e-01 9.6094e-01 2.8247e-01 4.2822e-01 9.9463e-01 9.9951e-01 2.5444e-03 2.9984e-03 3.5191e-03 2.9049e-03 + 1.2939e-02 1.3611e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.3262e-01 9.3018e-01 9.2041e-01 9.8779e-01 4.6802e-01 9.9512e-01 3.6084e-01 3.6768e-01 + 3.9136e-01 1.5686e-01 8.8379e-01 9.7229e-02 -9.3307e-03 -1.7075e-02 -6.1073e-03 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -1.9983e-01 1.9202e-01 9.0088e-02 -9.8486e-01 -1.3721e-01 -3.1281e-02 2.3511e-01 9.7998e-01 9.8145e-01 9.9609e-01 1.7261e-01 9.9072e-01 9.9951e-01 9.7217e-01 -2.8174e-01 -1.7810e-01 + 2.9272e-01 -9.6680e-01 9.1504e-01 1.1755e-01 7.3013e-03 9.5947e-01 9.8389e-01 9.5605e-01 2.5562e-01 4.0308e-01 9.9316e-01 1.0000e+00 2.4948e-03 2.9430e-03 3.4561e-03 2.8458e-03 + 1.3618e-02 1.2955e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.3457e-01 9.3750e-01 9.2627e-01 9.9268e-01 4.7583e-01 9.9414e-01 3.5522e-01 3.4766e-01 + 3.7720e-01 1.2024e-01 8.7939e-01 1.0944e-01 -8.4763e-03 -1.6800e-02 -5.7983e-03 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -2.0056e-01 1.8896e-01 8.8501e-02 -9.8486e-01 -1.4209e-01 -2.5833e-02 2.3218e-01 9.7949e-01 9.8193e-01 9.9609e-01 1.7456e-01 9.8975e-01 9.9951e-01 9.7266e-01 -2.7173e-01 -1.7688e-01 + 3.1445e-01 -9.7266e-01 9.2822e-01 1.2585e-01 3.9551e-02 9.6240e-01 9.8438e-01 9.4922e-01 2.3145e-01 3.7256e-01 9.9219e-01 9.9902e-01 2.4357e-03 2.8725e-03 3.3703e-03 2.7752e-03 + 1.3077e-02 1.2306e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.4043e-01 9.4434e-01 9.3115e-01 9.9463e-01 4.7119e-01 9.9561e-01 3.4033e-01 3.2910e-01 + 3.6426e-01 1.0529e-01 8.8184e-01 9.5581e-02 -8.3160e-03 -1.5747e-02 -5.7831e-03 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -2.0044e-01 1.8689e-01 8.9661e-02 -9.8389e-01 -1.4380e-01 -2.2064e-02 2.2754e-01 9.7949e-01 9.8242e-01 9.9609e-01 1.7920e-01 9.8975e-01 1.0000e+00 9.7363e-01 -2.5488e-01 -1.7749e-01 + 3.3325e-01 -9.8438e-01 9.3457e-01 1.6064e-01 7.3792e-02 9.6680e-01 9.8389e-01 9.4287e-01 1.7725e-01 3.5522e-01 9.8682e-01 9.9707e-01 2.4395e-03 2.8763e-03 3.3760e-03 2.7828e-03 + 1.2955e-02 1.2543e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.4336e-01 9.4775e-01 9.3311e-01 9.9316e-01 4.4189e-01 9.9561e-01 3.3203e-01 3.1958e-01 + 3.5986e-01 1.1859e-01 8.9697e-01 9.4299e-02 -1.0078e-02 -1.3527e-02 -4.7684e-03 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -2.0093e-01 1.9324e-01 9.5398e-02 -9.8438e-01 -1.3782e-01 -2.9648e-02 2.2729e-01 9.7949e-01 9.8096e-01 9.9561e-01 1.7664e-01 9.9023e-01 9.9951e-01 9.7363e-01 -2.2400e-01 -1.6748e-01 + 3.5132e-01 -9.9316e-01 9.3750e-01 1.8604e-01 1.0663e-01 9.7461e-01 9.8584e-01 9.3604e-01 1.1786e-01 3.4741e-01 9.8242e-01 9.9414e-01 2.5406e-03 2.9945e-03 3.5133e-03 2.8992e-03 + 1.3191e-02 1.3649e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.4678e-01 9.4824e-01 9.3457e-01 9.9023e-01 4.1748e-01 9.9561e-01 3.2178e-01 3.1714e-01 + 3.5522e-01 1.3794e-01 9.0869e-01 9.4482e-02 -1.1192e-02 -1.1902e-02 -4.0665e-03 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -2.0093e-01 1.9409e-01 9.7290e-02 -9.8438e-01 -1.3721e-01 -2.9938e-02 2.2522e-01 9.7949e-01 9.8096e-01 9.9512e-01 1.7676e-01 9.9072e-01 9.9951e-01 9.7412e-01 -1.9360e-01 -1.6248e-01 + 3.4888e-01 -9.9756e-01 9.3164e-01 2.0288e-01 1.1786e-01 9.8096e-01 9.8682e-01 9.3701e-01 6.6956e-02 3.6401e-01 9.7900e-01 9.9316e-01 2.4509e-03 2.8915e-03 3.3913e-03 2.7943e-03 + 1.3084e-02 1.2650e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.5117e-01 9.4824e-01 9.3652e-01 9.8828e-01 3.9551e-01 9.9561e-01 3.0933e-01 3.1714e-01 + 3.5034e-01 1.5405e-01 9.1846e-01 9.5825e-02 -1.3206e-02 -1.0635e-02 -3.2482e-03 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -2.0312e-01 1.9824e-01 1.0254e-01 -9.8145e-01 -1.3562e-01 -3.3447e-02 2.1313e-01 9.7900e-01 9.7998e-01 9.9463e-01 1.9141e-01 9.9072e-01 9.9951e-01 9.7705e-01 -1.7456e-01 -1.4661e-01 + 3.4863e-01 -1.0000e+00 9.2480e-01 2.2949e-01 1.3965e-01 9.8486e-01 9.8926e-01 9.3750e-01 1.8982e-02 3.8037e-01 9.7314e-01 9.9023e-01 2.4834e-03 2.9278e-03 3.4313e-03 2.8305e-03 + 1.2787e-02 1.3031e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.5459e-01 9.4824e-01 9.3750e-01 9.8584e-01 3.7402e-01 9.9609e-01 2.9810e-01 3.1787e-01 + 3.4741e-01 1.6907e-01 9.2725e-01 8.6731e-02 -1.5762e-02 -9.1934e-03 -2.7905e-03 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -2.0361e-01 1.9971e-01 1.0413e-01 -9.8242e-01 -1.3245e-01 -3.6346e-02 2.1899e-01 9.7900e-01 9.7998e-01 9.9463e-01 1.8604e-01 9.9121e-01 9.9951e-01 9.7559e-01 -1.6833e-01 -1.2659e-01 + 3.4424e-01 -9.9951e-01 9.1748e-01 2.5854e-01 1.6357e-01 9.8584e-01 9.9219e-01 9.3896e-01 -2.4643e-02 3.9844e-01 9.6582e-01 9.8633e-01 2.4967e-03 2.9469e-03 3.4599e-03 2.8534e-03 + 1.3336e-02 1.3145e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.5752e-01 9.4824e-01 9.3848e-01 9.8340e-01 3.5498e-01 9.9707e-01 2.8784e-01 3.1787e-01 + 3.4497e-01 1.8262e-01 9.3506e-01 7.9590e-02 -1.4969e-02 -8.3694e-03 -2.3327e-03 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -2.0874e-01 1.9580e-01 1.0413e-01 -9.8193e-01 -1.3647e-01 -2.8366e-02 2.2095e-01 9.7803e-01 9.8047e-01 9.9463e-01 1.9006e-01 9.9072e-01 9.9951e-01 9.7510e-01 -1.5686e-01 -1.1523e-01 + 3.3276e-01 -9.9707e-01 9.0576e-01 2.8296e-01 1.7590e-01 9.8779e-01 9.9316e-01 9.4287e-01 -7.3914e-02 4.2407e-01 9.5898e-01 9.8438e-01 2.4071e-03 2.8400e-03 3.3302e-03 2.7447e-03 + 1.2863e-02 1.2146e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.6094e-01 9.4824e-01 9.3945e-01 9.8096e-01 3.4229e-01 9.9707e-01 2.7759e-01 3.1763e-01 + 3.4277e-01 1.9324e-01 9.3945e-01 7.7454e-02 -1.7258e-02 -7.8087e-03 -1.9426e-03 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -2.1484e-01 1.9495e-01 1.0632e-01 -9.8047e-01 -1.3635e-01 -2.6123e-02 2.2070e-01 9.7656e-01 9.8096e-01 9.9414e-01 1.9568e-01 9.9072e-01 9.9951e-01 9.7559e-01 -1.4355e-01 -9.1614e-02 + 3.3594e-01 -9.9268e-01 8.9648e-01 2.9785e-01 1.9861e-01 9.8975e-01 9.9561e-01 9.4189e-01 -1.1975e-01 4.4263e-01 9.5459e-01 9.7998e-01 2.4586e-03 2.9011e-03 3.4046e-03 2.8076e-03 + 1.2978e-02 1.2848e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.6338e-01 9.4824e-01 9.4043e-01 9.7949e-01 3.2837e-01 9.9756e-01 2.6733e-01 3.1689e-01 + 3.4058e-01 2.0239e-01 9.4434e-01 7.0312e-02 -1.8677e-02 -7.3433e-03 -1.7614e-03 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -2.1118e-01 1.9226e-01 1.0400e-01 -9.8047e-01 -1.3745e-01 -2.5909e-02 2.2034e-01 9.7754e-01 9.8145e-01 9.9463e-01 1.9629e-01 9.9072e-01 9.9951e-01 9.7559e-01 -1.4172e-01 -7.3059e-02 + 3.2739e-01 -9.8584e-01 8.8525e-01 3.2007e-01 2.1680e-01 9.8975e-01 9.9756e-01 9.4482e-01 -1.6895e-01 4.6509e-01 9.4727e-01 9.7607e-01 2.4300e-03 2.8687e-03 3.3684e-03 2.7752e-03 + 1.3351e-02 1.2573e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.6631e-01 9.4873e-01 9.4043e-01 9.7754e-01 3.1543e-01 9.9756e-01 2.5732e-01 3.1641e-01 + 3.3960e-01 2.1082e-01 9.4873e-01 6.8542e-02 -1.8600e-02 -7.1297e-03 -1.4200e-03 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -2.0605e-01 1.9080e-01 1.0150e-01 -9.8145e-01 -1.3867e-01 -2.5940e-02 2.1912e-01 9.7852e-01 9.8145e-01 9.9463e-01 1.9202e-01 9.9023e-01 9.9951e-01 9.7559e-01 -1.3342e-01 -5.7800e-02 + 3.1982e-01 -9.7510e-01 8.7402e-01 3.3691e-01 2.3218e-01 9.9121e-01 9.9854e-01 9.4727e-01 -2.2241e-01 4.8608e-01 9.4141e-01 9.7266e-01 2.4567e-03 2.8992e-03 3.4027e-03 2.8019e-03 + 1.3550e-02 1.2749e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.6875e-01 9.4873e-01 9.4092e-01 9.7607e-01 3.0542e-01 9.9756e-01 2.4768e-01 3.1543e-01 + 3.3813e-01 2.1680e-01 9.5215e-01 6.6284e-02 -1.8860e-02 -7.4615e-03 -1.2827e-03 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -2.0032e-01 1.9116e-01 1.0144e-01 -9.8193e-01 -1.3843e-01 -2.7161e-02 2.1167e-01 9.7998e-01 9.8145e-01 9.9463e-01 1.9006e-01 9.9023e-01 9.9951e-01 9.7754e-01 -1.2561e-01 -4.5471e-02 + 3.0371e-01 -9.6240e-01 8.5498e-01 3.5034e-01 2.4097e-01 9.9219e-01 9.9902e-01 9.5264e-01 -2.7148e-01 5.1855e-01 9.3652e-01 9.7070e-01 2.4776e-03 2.9240e-03 3.4332e-03 2.8286e-03 + 1.3535e-02 1.2993e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.7119e-01 9.4922e-01 9.4141e-01 9.7510e-01 2.9565e-01 9.9805e-01 2.3755e-01 3.1470e-01 + 3.3691e-01 2.2144e-01 9.5508e-01 6.4453e-02 -2.0020e-02 -7.7782e-03 -1.1120e-03 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -1.9824e-01 1.9531e-01 1.0199e-01 -9.8291e-01 -1.3892e-01 -2.9343e-02 2.1313e-01 9.7998e-01 9.8096e-01 9.9463e-01 1.8494e-01 9.9023e-01 9.9951e-01 9.7705e-01 -1.2122e-01 -3.5828e-02 + 2.9224e-01 -9.4629e-01 8.3203e-01 3.6230e-01 2.6489e-01 9.9268e-01 9.9951e-01 9.5654e-01 -3.2275e-01 5.5518e-01 9.3213e-01 9.6436e-01 2.5234e-03 2.9736e-03 3.4866e-03 2.8725e-03 + 1.3412e-02 1.3245e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.7412e-01 9.4922e-01 9.4189e-01 9.7461e-01 2.8613e-01 9.9805e-01 2.2644e-01 3.1445e-01 + 3.3594e-01 2.2437e-01 9.5801e-01 6.0272e-02 -2.0996e-02 -8.6517e-03 -8.9264e-04 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -1.9507e-01 1.9568e-01 1.0236e-01 -9.8340e-01 -1.3831e-01 -2.9785e-02 2.1277e-01 9.8096e-01 9.8047e-01 9.9463e-01 1.8176e-01 9.9023e-01 9.9951e-01 9.7705e-01 -1.2231e-01 -2.8229e-02 + 2.8320e-01 -9.2920e-01 8.0664e-01 3.6768e-01 2.9663e-01 9.9268e-01 9.9951e-01 9.5898e-01 -3.6987e-01 5.9082e-01 9.3018e-01 9.5508e-01 2.4986e-03 2.9488e-03 3.4637e-03 2.8553e-03 + 1.3390e-02 1.3145e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.7656e-01 9.4922e-01 9.4238e-01 9.7412e-01 2.7661e-01 9.9854e-01 2.1436e-01 3.1470e-01 + 3.3447e-01 2.2546e-01 9.6094e-01 5.5969e-02 -2.0309e-02 -8.9569e-03 -7.3004e-04 1.0000e+00 1.0000e+00 1.0000e+00]] + - image: [] + - lang: 'unlocked_waist: pick the cupcake from the cutting board and place it in the basket' + - dataset_id: 7 +Sample[3] content: + - action: [[-0.064 -0.1069 0.1733 -0.0713 0.144 0.11584 -0.02623 0.03506 0.1963 -0.08075 0.0822 -0.2346 -0.2625 0.03032 0. 0. 0. 0. 0. + -1. 1. 1. 1. 1. 1. 0. -0.1941 0.443 -0.7163 0. 0. 0. 0. 0. 0. 0. 0. ] + [-0.0422 -0.1018 0.1812 -0.0691 0.1422 0.1243 -0.02777 0.043 0.34 -0.129 0.0728 -0.2517 -0.3652 0.0547 0. 0. 0. 0. 0. + -1. 1. 1. 1. 1. 1. 0. -0.177 0.366 -0.6606 0. 0. 0. 0. 0. 0. 0. 0. ] + [-0.0329 -0.1278 0.169 -0.0628 0.1406 0.1251 -0.03004 0.04282 0.3967 -0.1389 0.06537 -0.2372 -0.4238 0.06396 0. 0. 0. 0. 0. + -1. -1. -1. -1. -1. -1. 0. -0.1555 0.3628 -0.6504 0. 0. 0. 0. 0. 0. 0. 0. ] + [-0.02266 -0.109 0.1722 -0.078 0.1451 0.1199 -0.02267 0.06506 0.4746 -0.126 0.0536 -0.2136 -0.481 0.05496 0. 0. 0. 0. 0. + -1. -1. -1. -1. -1. -1. 0. -0.1144 0.3394 -0.638 0. 0. 0. 0. 0. 0. 0. 0. ] + [-0.03995 -0.1252 0.1874 -0.05884 0.1454 0.1318 -0.012794 0.0753 0.5293 -0.1159 0.02419 -0.1947 -0.5234 0.06238 0. 0. 0. 0. 0. + -1. -1. -1. -1. -1. -1. 0. -0.1598 0.3408 -0.641 0. 0. 0. 0. 0. 0. 0. 0. ] + [-0.01743 -0.157 0.1493 -0.09735 0.1335 0.1382 -0.00258 0.0824 0.6016 -0.0933 0.005127 -0.1676 -0.5664 0.0691 0. 0. 0. 0. 0. + -1. -1. -1. -1. -1. -1. 0. -0.0866 0.3281 -0.635 0. 0. 0. 0. 0. 0. 0. 0. ] + [-0.01295 -0.2062 0.1322 -0.06027 0.11615 0.1482 -0.02892 0.0942 0.634 -0.0695 -0.001305 -0.1332 -0.59 0.08484 0. 0. 0. 0. 0. + -1. -1. -1. -1. -1. -1. 0. -0.09424 0.3523 -0.639 0. 0. 0. 0. 0. 0. 0. 0. ] + [-0.01831 -0.1641 0.1512 -0.0992 0.1283 0.139 0.005253 0.1177 0.6216 -0.0405 -0.04187 -0.0891 -0.5894 0.1323 0. 0. 0. 0. 0. + -1. -1. -1. -1. -1. -1. 0. -0.08405 0.3384 -0.638 0. 0. 0. 0. 0. 0. 0. 0. ] + [ 0.01591 -0.1791 0.1466 -0.0973 0.1179 0.1517 -0.01924 0.148 0.5737 -0.0195 -0.0748 -0.0536 -0.5576 0.18 0. 0. 0. 0. 0. + -1. -1. -1. -1. -1. -1. 0. -0.0963 0.3286 -0.644 0. 0. 0. 0. 0. 0. 0. 0. ] + [-0.02473 -0.1414 0.1561 -0.0987 0.1237 0.1384 0.01813 0.132 0.538 0.003721 -0.07367 -0.01159 -0.509 0.2485 0. 0. 0. 0. 0. + -1. -1. -1. -1. -1. -1. 0. -0.10205 0.3318 -0.651 0. 0. 0. 0. 0. 0. 0. 0. ] + [-0.03134 -0.1842 0.1577 -0.0838 0.1207 0.1505 0.003366 0.1432 0.505 0.001746 -0.07153 0.006252 -0.4734 0.2822 0. 0. 0. 0. 0. + -1. -1. -1. -1. -1. -1. 0. -0.1169 0.339 -0.6523 0. 0. 0. 0. 0. 0. 0. 0. ] + [ 0.001158 -0.1882 0.1537 -0.09 0.1204 0.1503 -0.02354 0.1519 0.5005 0.01625 -0.0634 0.02637 -0.4402 0.3118 0. 0. 0. 0. 0. + -1. -1. -1. -1. -1. -1. 0. -0.08905 0.3354 -0.6587 0. 0. 0. 0. 0. 0. 0. 0. ] + [ 0.006107 -0.1643 0.1515 -0.0913 0.1169 0.1456 -0.02325 0.1492 0.4556 0.01628 -0.0399 0.0527 -0.406 0.3386 0. 0. 0. 0. 0. + -1. -1. -1. -1. -1. -1. 0. -0.07074 0.339 -0.667 0. 0. 0. 0. 0. 0. 0. 0. ] + [-0.01965 -0.1719 0.1445 -0.0893 0.11206 0.1471 -0.002232 0.1609 0.4023 0.01923 -0.06015 0.0647 -0.3625 0.3616 0. 0. 0. 0. 0. + -1. -1. -1. -1. -1. -1. 0. -0.1322 0.3374 -0.6646 0. 0. 0. 0. 0. 0. 0. 0. ] + [-0.01261 -0.1772 0.1505 -0.07874 0.1085 0.1542 -0.01438 0.1472 0.4128 0.03827 -0.06476 0.08295 -0.3184 0.4065 0. 0. 0. 0. 0. + -1. -1. -1. -1. -1. -1. 0. -0.0924 0.344 -0.672 0. 0. 0. 0. 0. 0. 0. 0. ]] + - state: [[ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -1.2384e-01 4.5557e-01 -1.8103e-01 -9.9902e-01 1.0394e-01 2.3328e-01 -8.2092e-03 9.9219e-01 8.9014e-01 9.8340e-01 4.1473e-02 9.9463e-01 9.7266e-01 1.0000e+00 -3.3545e-01 -4.2480e-01 + 5.6946e-02 -9.8682e-01 7.9285e-02 -2.1179e-01 -8.4534e-02 9.4189e-01 9.0527e-01 9.9854e-01 1.6260e-01 9.9707e-01 9.7754e-01 9.9658e-01 1.2693e-03 1.4343e-03 1.6127e-03 1.3676e-03 + -9.2459e-04 1.0246e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.1078e-02 1.2283e-02 1.3618e-02 1.1253e-02 -8.0204e-04 9.9023e-01 1.0000e+00 1.0000e+00 + 1.0000e+00 1.0000e+00 1.0000e+00 1.4062e-01 -7.2670e-03 5.9471e-03 -1.2047e-02 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -1.1133e-01 4.5630e-01 -1.7798e-01 -9.9854e-01 1.0492e-01 2.3767e-01 1.7519e-03 9.9365e-01 8.8965e-01 9.8389e-01 5.0964e-02 9.9463e-01 9.7119e-01 1.0000e+00 -3.1396e-01 -4.5898e-01 + 4.8798e-02 -9.8145e-01 1.3757e-01 -2.5024e-01 -5.5786e-02 9.4922e-01 8.8818e-01 9.9902e-01 1.9080e-01 9.9072e-01 9.6826e-01 9.9854e-01 1.4715e-03 1.6689e-03 1.8835e-03 1.5898e-03 + -3.0136e-04 1.1368e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.5344e-01 1.5454e-01 1.5588e-01 1.5381e-01 1.0864e-01 9.9902e-01 9.8828e-01 9.8779e-01 + 9.8779e-01 9.8828e-01 9.9414e-01 4.4891e-02 -1.6525e-02 9.1705e-03 -1.4694e-02 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -9.5581e-02 4.5752e-01 -1.7383e-01 -9.9805e-01 1.0510e-01 2.4365e-01 9.2697e-03 9.9561e-01 8.8916e-01 9.8486e-01 5.9082e-02 9.9463e-01 9.6973e-01 1.0000e+00 -2.9419e-01 -4.6460e-01 + 2.7786e-02 -9.7754e-01 1.7676e-01 -3.1226e-01 -2.5253e-02 9.5557e-01 8.8574e-01 9.9951e-01 2.1008e-01 9.8438e-01 9.5020e-01 9.9951e-01 1.6794e-03 1.9121e-03 2.1687e-03 1.8244e-03 + 3.1352e-04 1.2459e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 2.8052e-01 2.8223e-01 2.8394e-01 2.8198e-01 2.0288e-01 9.9951e-01 9.5996e-01 9.5947e-01 + 9.5898e-01 9.5947e-01 9.7900e-01 -3.6102e-02 -2.2568e-02 7.3357e-03 -1.3367e-02 9.9951e-01 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -8.0505e-02 4.5581e-01 -1.7310e-01 -9.9756e-01 1.0480e-01 2.4866e-01 1.4397e-02 9.9658e-01 8.9014e-01 9.8486e-01 6.7017e-02 9.9463e-01 9.6875e-01 1.0000e+00 -2.7856e-01 -4.5947e-01 + 7.8430e-03 -9.7510e-01 2.1191e-01 -3.7842e-01 1.7834e-03 9.6045e-01 8.8818e-01 1.0000e+00 2.2253e-01 9.7705e-01 9.2578e-01 1.0000e+00 1.7462e-03 1.9932e-03 2.2678e-03 1.9073e-03 + 6.3658e-04 1.2703e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 2.5732e-01 2.5977e-01 2.6270e-01 2.6001e-01 1.8323e-01 9.9561e-01 9.6631e-01 9.6582e-01 + 9.6484e-01 9.6582e-01 9.8291e-01 -9.5215e-02 -2.5696e-02 5.5580e-03 -1.1482e-02 9.9951e-01 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -6.5857e-02 4.5630e-01 -1.7200e-01 -9.9756e-01 1.0590e-01 2.5098e-01 2.0309e-02 9.9805e-01 8.8965e-01 9.8486e-01 6.8848e-02 9.9414e-01 9.6777e-01 1.0000e+00 -2.5806e-01 -4.4385e-01 + -4.1733e-03 -9.7363e-01 2.4768e-01 -4.4580e-01 2.0050e-02 9.6631e-01 8.9600e-01 1.0000e+00 2.2864e-01 9.6875e-01 8.9502e-01 1.0000e+00 1.8930e-03 2.1667e-03 2.4738e-03 2.0809e-03 + 1.1253e-03 1.3580e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 2.3438e-01 2.3682e-01 2.3975e-01 2.3718e-01 1.5942e-01 9.9023e-01 9.7217e-01 9.7168e-01 + 9.7070e-01 9.7168e-01 9.8730e-01 -1.3904e-01 -2.5391e-02 2.6875e-03 -9.6054e-03 9.9951e-01 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -5.8868e-02 4.5508e-01 -1.6772e-01 -9.9707e-01 1.0687e-01 2.5635e-01 2.7588e-02 9.9805e-01 8.9062e-01 9.8584e-01 7.5562e-02 9.9414e-01 9.6680e-01 9.9951e-01 -2.3816e-01 -4.2310e-01 + -1.0017e-02 -9.7461e-01 2.8247e-01 -5.0879e-01 3.6316e-02 9.7119e-01 9.0625e-01 1.0000e+00 2.2461e-01 9.5947e-01 8.6084e-01 9.9951e-01 1.9569e-03 2.2411e-03 2.5616e-03 2.1572e-03 + 8.2445e-04 1.3702e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 2.1423e-01 2.1680e-01 2.1997e-01 2.1729e-01 1.3440e-01 9.8486e-01 9.7656e-01 9.7607e-01 + 9.7559e-01 9.7607e-01 9.9072e-01 -1.7236e-01 -2.7969e-02 5.6648e-04 -8.0643e-03 9.9951e-01 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -4.7913e-02 4.5093e-01 -1.7236e-01 -9.9756e-01 1.0400e-01 2.6221e-01 3.6102e-02 9.9902e-01 8.9258e-01 9.8486e-01 6.9885e-02 9.9463e-01 9.6484e-01 9.9951e-01 -2.1960e-01 -3.9600e-01 + -7.8659e-03 -9.7656e-01 3.1909e-01 -5.6836e-01 5.1208e-02 9.7559e-01 9.1846e-01 1.0000e+00 2.1558e-01 9.4775e-01 8.2275e-01 9.9854e-01 1.9836e-03 2.2774e-03 2.6093e-03 2.1935e-03 + 1.8997e-03 1.3557e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.9519e-01 1.9788e-01 2.0105e-01 1.9849e-01 1.1230e-01 9.8486e-01 9.8096e-01 9.8047e-01 + 9.7949e-01 9.7998e-01 9.9365e-01 -1.7285e-01 -2.5345e-02 -1.8148e-03 -6.6071e-03 9.9951e-01 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -3.7567e-02 4.4263e-01 -1.8005e-01 -9.9707e-01 9.6008e-02 2.7002e-01 3.6041e-02 9.9951e-01 8.9697e-01 9.8389e-01 7.5562e-02 9.9561e-01 9.6289e-01 9.9951e-01 -2.0068e-01 -3.6914e-01 + 1.5879e-03 -9.7852e-01 3.5962e-01 -6.1865e-01 6.7810e-02 9.7949e-01 9.2920e-01 1.0000e+00 2.0630e-01 9.3311e-01 7.8564e-01 9.9756e-01 1.9989e-03 2.2945e-03 2.6283e-03 2.2068e-03 + 1.9569e-03 1.3184e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.7859e-01 1.8127e-01 1.8445e-01 1.8188e-01 9.1553e-02 9.8535e-01 9.8389e-01 9.8340e-01 + 9.8291e-01 9.8340e-01 9.9561e-01 -1.7078e-01 -2.3422e-02 -2.4796e-03 -5.7030e-03 9.9951e-01 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -3.1021e-02 4.3994e-01 -1.8213e-01 -9.9756e-01 9.3384e-02 2.7344e-01 4.4495e-02 9.9951e-01 8.9795e-01 9.8340e-01 6.9336e-02 9.9561e-01 9.6191e-01 9.9902e-01 -1.7688e-01 -3.4961e-01 + 1.8097e-02 -9.8242e-01 4.0649e-01 -6.5625e-01 9.5947e-02 9.8438e-01 9.3701e-01 1.0000e+00 1.8726e-01 9.1357e-01 7.5439e-01 9.9561e-01 2.1515e-03 2.4757e-03 2.8439e-03 2.3899e-03 + 2.2564e-03 1.4420e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.6394e-01 1.6663e-01 1.7004e-01 1.6736e-01 7.1594e-02 9.8535e-01 9.8633e-01 9.8584e-01 + 9.8535e-01 9.8584e-01 9.9756e-01 -1.7041e-01 -2.1240e-02 -3.4733e-03 -5.0507e-03 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -1.6663e-02 4.3677e-01 -1.8445e-01 -9.9805e-01 8.8196e-02 2.7979e-01 4.5105e-02 1.0000e+00 8.9941e-01 9.8291e-01 6.4392e-02 9.9609e-01 9.5996e-01 9.9902e-01 -1.4648e-01 -3.4155e-01 + 3.7781e-02 -9.8682e-01 4.5459e-01 -6.7676e-01 1.3354e-01 9.8926e-01 9.3994e-01 9.9951e-01 1.6211e-01 8.9062e-01 7.3633e-01 9.9121e-01 2.2163e-03 2.5520e-03 2.9354e-03 2.4643e-03 + 2.8229e-03 1.4389e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.5076e-01 1.5369e-01 1.5710e-01 1.5430e-01 5.3009e-02 9.8535e-01 9.8877e-01 9.8828e-01 + 9.8779e-01 9.8779e-01 9.9854e-01 -1.6968e-01 -2.0264e-02 -4.9515e-03 -4.8141e-03 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -1.6037e-02 4.3799e-01 -1.8433e-01 -9.9805e-01 8.5876e-02 2.8101e-01 5.5054e-02 1.0000e+00 8.9893e-01 9.8291e-01 6.0150e-02 9.9609e-01 9.5996e-01 9.9854e-01 -1.2817e-01 -3.4131e-01 + 6.0333e-02 -9.8975e-01 5.0537e-01 -6.8066e-01 1.8481e-01 9.9170e-01 9.3994e-01 9.9805e-01 1.4221e-01 8.6279e-01 7.3242e-01 9.8291e-01 2.2831e-03 2.6264e-03 3.0193e-03 2.5368e-03 + 2.5139e-03 1.4824e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.3855e-01 1.4136e-01 1.4490e-01 1.4197e-01 3.4241e-02 9.8535e-01 9.9023e-01 9.8975e-01 + 9.8926e-01 9.8975e-01 9.9951e-01 -1.7017e-01 -2.0081e-02 -5.9319e-03 -4.9896e-03 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -1.7929e-02 4.3481e-01 -1.8384e-01 -9.9805e-01 8.3191e-02 2.8516e-01 5.9448e-02 1.0000e+00 9.0039e-01 9.8291e-01 6.1096e-02 9.9658e-01 9.5850e-01 9.9805e-01 -1.1011e-01 -3.4619e-01 + 7.7637e-02 -9.9170e-01 5.4980e-01 -6.7480e-01 2.3572e-01 9.9414e-01 9.3799e-01 9.9707e-01 1.2744e-01 8.3545e-01 7.3779e-01 9.7168e-01 2.2984e-03 2.6493e-03 3.0499e-03 2.5635e-03 + 2.7332e-03 1.4603e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.2793e-01 1.3086e-01 1.3464e-01 1.3147e-01 1.8478e-02 9.8535e-01 9.9170e-01 9.9121e-01 + 9.9072e-01 9.9121e-01 1.0000e+00 -1.7004e-01 -2.0752e-02 -6.3400e-03 -5.3253e-03 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -1.0857e-02 4.3164e-01 -1.8433e-01 -9.9805e-01 8.0994e-02 2.8833e-01 5.5634e-02 1.0000e+00 9.0186e-01 9.8291e-01 6.0394e-02 9.9658e-01 9.5752e-01 9.9854e-01 -9.2407e-02 -3.5107e-01 + 9.5276e-02 -9.9316e-01 5.8936e-01 -6.6211e-01 2.8442e-01 9.9561e-01 9.3652e-01 9.9561e-01 1.1847e-01 8.0811e-01 7.4951e-01 9.5850e-01 2.3746e-03 2.7428e-03 3.1662e-03 2.6569e-03 + 3.5210e-03 1.4915e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.1786e-01 1.2091e-01 1.2463e-01 1.2146e-01 7.1411e-03 9.8535e-01 9.9316e-01 9.9268e-01 + 9.9219e-01 9.9268e-01 1.0000e+00 -1.6992e-01 -1.9638e-02 -6.8741e-03 -5.9471e-03 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -3.4695e-03 4.3140e-01 -1.8506e-01 -9.9805e-01 7.8125e-02 2.8955e-01 5.2368e-02 1.0000e+00 9.0234e-01 9.8291e-01 5.9357e-02 9.9707e-01 9.5703e-01 9.9854e-01 -7.9529e-02 -3.6182e-01 + 1.0919e-01 -9.9316e-01 6.2744e-01 -6.4307e-01 3.3032e-01 9.9707e-01 9.3213e-01 9.9414e-01 1.1847e-01 7.7881e-01 7.6562e-01 9.4385e-01 2.4414e-03 2.8191e-03 3.2539e-03 2.7275e-03 + 3.7613e-03 1.5114e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0938e-01 1.1255e-01 1.1646e-01 1.1304e-01 -3.6659e-03 9.8535e-01 9.9414e-01 9.9365e-01 + 9.9316e-01 9.9365e-01 1.0000e+00 -1.7041e-01 -1.7380e-02 -7.0190e-03 -6.8741e-03 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -4.4250e-03 4.3091e-01 -1.8713e-01 -9.9805e-01 7.4280e-02 2.9102e-01 5.5237e-02 1.0000e+00 9.0234e-01 9.8242e-01 5.9113e-02 9.9707e-01 9.5654e-01 9.9854e-01 -6.5369e-02 -3.7793e-01 + 1.2103e-01 -9.9365e-01 6.5918e-01 -6.1719e-01 3.7231e-01 9.9805e-01 9.2578e-01 9.9268e-01 1.1304e-01 7.5195e-01 7.8711e-01 9.2822e-01 2.3994e-03 2.7676e-03 3.1910e-03 2.6760e-03 + 3.5439e-03 1.4603e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0229e-01 1.0565e-01 1.0986e-01 1.0608e-01 -3.7231e-03 9.8535e-01 9.9463e-01 9.9463e-01 + 9.9414e-01 9.9414e-01 1.0000e+00 -1.6968e-01 -1.9577e-02 -7.1297e-03 -7.4043e-03 1.0000e+00 1.0000e+00 1.0000e+00]] + - image: [] + - lang: 'unlocked_waist: pick the croissant from the tray and place it in the cardboard box' + - dataset_id: 7 +Sample[4] content: + - action: [[-1.0223e-01 -3.6377e-01 5.3564e-01 3.9642e-02 -9.9609e-02 -7.8583e-03 -1.4107e-02 -4.5044e-01 5.8105e-01 -5.3613e-01 6.8750e-01 -1.3086e-01 -2.5952e-01 -1.3599e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 0.0000e+00 -2.7664e-02 3.6719e-01 -6.2793e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-1.5637e-01 -3.1396e-01 5.5078e-01 8.4167e-02 -9.4971e-02 -2.3163e-02 1.8454e-03 -4.3726e-01 5.8301e-01 -5.4590e-01 6.9141e-01 -1.4197e-01 -2.6880e-01 -1.5710e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 0.0000e+00 -2.7847e-02 3.8477e-01 -6.2793e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-1.0522e-01 -3.9380e-01 5.3320e-01 6.7383e-02 -1.0962e-01 9.3918e-03 -4.8248e-02 -4.1064e-01 5.8740e-01 -5.4346e-01 6.7139e-01 -1.4856e-01 -2.5830e-01 -1.7432e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 0.0000e+00 -3.6133e-02 3.7085e-01 -6.2842e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-8.8684e-02 -4.0674e-01 5.1855e-01 2.0615e-02 -1.1273e-01 5.6305e-03 7.7057e-04 -4.0381e-01 5.6348e-01 -5.4053e-01 6.6455e-01 -1.5088e-01 -2.4988e-01 -1.5259e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 0.0000e+00 -3.1342e-02 3.6133e-01 -6.2744e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-9.2773e-02 -4.1992e-01 5.2979e-01 5.8411e-02 -1.0449e-01 5.8136e-03 -3.1677e-02 -3.8525e-01 5.4150e-01 -5.4590e-01 6.3330e-01 -1.6174e-01 -2.2742e-01 -1.6125e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 0.0000e+00 -7.7393e-02 3.4863e-01 -6.2939e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-1.0577e-01 -3.0273e-01 5.5957e-01 4.7363e-02 -8.7585e-02 -3.3661e-02 -1.2321e-02 -3.6597e-01 5.0293e-01 -5.5273e-01 6.4111e-01 -1.6968e-01 -2.1057e-01 -1.8359e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 0.0000e+00 -5.0690e-02 3.6182e-01 -6.2744e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-6.7139e-02 -3.7207e-01 5.4150e-01 8.3466e-03 -1.0199e-01 -4.4518e-03 -4.9067e-04 -3.6157e-01 4.7290e-01 -5.4395e-01 6.0400e-01 -1.6992e-01 -1.8921e-01 -1.4502e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 0.0000e+00 -7.8735e-02 3.3301e-01 -6.2891e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-7.5867e-02 -3.8696e-01 5.3467e-01 2.9099e-02 -1.0675e-01 -2.9602e-03 -7.8125e-03 -3.3984e-01 4.5825e-01 -5.4150e-01 5.8545e-01 -1.7664e-01 -1.6443e-01 -1.5369e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 0.0000e+00 -9.3689e-02 3.3691e-01 -6.2988e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-8.8318e-02 -3.2715e-01 5.7129e-01 2.2629e-02 -8.3252e-02 -2.2324e-02 5.4932e-03 -3.1812e-01 4.7119e-01 -5.3076e-01 5.4932e-01 -1.7786e-01 -1.4758e-01 -1.4172e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 0.0000e+00 -1.1792e-01 3.2178e-01 -6.3135e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-3.5645e-02 -3.4131e-01 5.5713e-01 3.0975e-03 -9.2590e-02 -1.9791e-02 -2.4185e-02 -2.8345e-01 4.4263e-01 -5.3369e-01 5.3418e-01 -1.8250e-01 -1.2512e-01 -1.7468e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 0.0000e+00 -1.2018e-01 3.1470e-01 -6.3086e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-1.2280e-01 -3.3984e-01 5.6543e-01 8.4717e-02 -1.0022e-01 -8.1863e-03 -1.6373e-02 -2.9175e-01 4.1260e-01 -5.2393e-01 5.1367e-01 -1.8762e-01 -9.3567e-02 -1.4185e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 0.0000e+00 -1.3245e-01 3.3911e-01 -6.3330e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-7.7881e-02 -3.6401e-01 5.4248e-01 -1.3504e-02 -9.3933e-02 -2.2766e-02 2.5558e-02 -2.6758e-01 4.2676e-01 -5.2246e-01 4.7607e-01 -2.0068e-01 -7.7454e-02 -1.3147e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 0.0000e+00 -9.7290e-02 3.2983e-01 -6.3037e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-3.0350e-02 -4.3555e-01 5.1953e-01 -2.0859e-02 -1.1243e-01 1.5554e-03 -1.0132e-02 -2.7100e-01 3.7524e-01 -5.1562e-01 4.7949e-01 -1.9604e-01 -6.1676e-02 -1.0272e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 0.0000e+00 -7.1350e-02 3.3936e-01 -6.3037e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-1.0437e-01 -3.4302e-01 5.6348e-01 4.0253e-02 -9.5459e-02 -1.7822e-02 -3.6068e-03 -2.3059e-01 4.5532e-01 -5.0195e-01 4.2236e-01 -2.0679e-01 -4.4098e-02 -1.1401e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 0.0000e+00 -1.1761e-01 3.3496e-01 -6.3281e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-8.4290e-02 -3.6694e-01 5.4883e-01 2.2888e-02 -1.0736e-01 -8.1558e-03 3.1052e-03 -2.1289e-01 4.0796e-01 -4.9780e-01 3.8208e-01 -2.0654e-01 -2.3209e-02 -9.6069e-02 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 0.0000e+00 -1.2451e-01 3.2837e-01 -6.3330e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00]] + - state: [[ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -1.4783e-01 3.3911e-01 1.8604e-01 -9.7070e-01 -2.5659e-01 7.0862e-02 7.0068e-02 9.8926e-01 9.4092e-01 9.8242e-01 2.3999e-01 9.6631e-01 9.9756e-01 9.9756e-01 -9.2383e-01 -2.2766e-01 + -5.8008e-01 -3.8501e-01 5.3711e-01 -4.6460e-01 -1.9226e-01 3.8281e-01 9.7363e-01 8.1445e-01 9.2285e-01 8.4375e-01 8.8574e-01 9.8145e-01 4.0588e-03 4.7150e-03 5.5008e-03 4.6387e-03 + 5.3978e-03 2.1805e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0229e-01 7.9834e-02 7.7637e-02 7.7637e-02 -2.1820e-03 9.8535e-01 9.9463e-01 9.9658e-01 + 9.9707e-01 9.9707e-01 1.0000e+00 -1.7078e-01 2.8030e-02 1.0735e-02 9.7513e-04 9.9951e-01 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -1.4124e-01 3.3765e-01 1.8884e-01 -9.7070e-01 -2.5684e-01 7.3608e-02 6.6467e-02 9.8975e-01 9.4141e-01 9.8193e-01 2.3938e-01 9.6631e-01 9.9707e-01 9.9756e-01 -9.1846e-01 -2.4585e-01 + -5.8154e-01 -3.9331e-01 5.2881e-01 -4.4678e-01 -1.9409e-01 3.9600e-01 9.6924e-01 8.1348e-01 9.1943e-01 8.4863e-01 8.9453e-01 9.8096e-01 4.1199e-03 4.7913e-03 5.5962e-03 4.7150e-03 + 6.0158e-03 2.2522e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.9951e-01 9.4360e-02 7.4280e-02 7.2632e-02 7.2266e-02 -2.4300e-03 9.8535e-01 9.9561e-01 9.9707e-01 + 9.9756e-01 9.9756e-01 1.0000e+00 -1.6992e-01 2.2766e-02 8.3542e-03 6.9904e-04 9.9951e-01 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -1.5015e-01 3.4204e-01 1.9434e-01 -9.6777e-01 -2.5562e-01 7.1167e-02 6.7383e-02 9.8877e-01 9.3945e-01 9.8096e-01 2.5098e-01 9.6680e-01 9.9756e-01 9.9756e-01 -9.1162e-01 -2.6050e-01 + -5.8496e-01 -3.9893e-01 5.1904e-01 -4.3481e-01 -2.0215e-01 4.1089e-01 9.6533e-01 8.1104e-01 9.1699e-01 8.5498e-01 9.0039e-01 9.7949e-01 4.1466e-03 4.8180e-03 5.6190e-03 4.7379e-03 + 5.2032e-03 2.2858e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.9951e-01 8.6304e-02 6.8237e-02 6.6895e-02 6.6406e-02 -1.7948e-03 9.8535e-01 9.9609e-01 9.9756e-01 + 9.9756e-01 9.9756e-01 1.0000e+00 -1.7029e-01 1.8509e-02 7.3013e-03 5.0163e-04 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -1.4417e-01 3.3716e-01 1.9507e-01 -9.6680e-01 -2.5879e-01 7.8552e-02 5.5237e-02 9.8975e-01 9.4141e-01 9.8096e-01 2.5610e-01 9.6582e-01 9.9707e-01 9.9854e-01 -9.0186e-01 -2.7075e-01 + -5.8691e-01 -4.0845e-01 5.0879e-01 -4.2261e-01 -2.1399e-01 4.3188e-01 9.6240e-01 8.0957e-01 9.1260e-01 8.6084e-01 9.0625e-01 9.7705e-01 4.0779e-03 4.7455e-03 5.5389e-03 4.6692e-03 + 6.1607e-03 2.1973e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 7.9163e-02 6.2988e-02 6.1951e-02 6.1340e-02 -1.5554e-03 9.8535e-01 9.9707e-01 9.9805e-01 + 9.9805e-01 9.9805e-01 1.0000e+00 -1.6992e-01 1.4656e-02 5.7640e-03 4.0913e-04 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -1.3428e-01 3.3154e-01 1.9238e-01 -9.6924e-01 -2.6270e-01 8.3679e-02 5.7892e-02 9.9072e-01 9.4336e-01 9.8145e-01 2.4646e-01 9.6484e-01 9.9658e-01 9.9854e-01 -8.9258e-01 -2.8271e-01 + -5.8838e-01 -4.1797e-01 5.0049e-01 -4.1064e-01 -2.1667e-01 4.5093e-01 9.5898e-01 8.0859e-01 9.0869e-01 8.6572e-01 9.1162e-01 9.7607e-01 4.1008e-03 4.7684e-03 5.5695e-03 4.6921e-03 + 6.2790e-03 2.2263e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.9951e-01 7.2815e-02 5.8319e-02 5.7587e-02 5.6854e-02 -1.9894e-03 9.8535e-01 9.9756e-01 9.9854e-01 + 9.9854e-01 9.9854e-01 1.0000e+00 -1.7078e-01 1.1978e-02 4.1809e-03 3.8671e-04 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -1.2732e-01 3.2593e-01 1.9250e-01 -9.6875e-01 -2.6318e-01 8.7708e-02 5.2246e-02 9.9170e-01 9.4531e-01 9.8145e-01 2.4890e-01 9.6484e-01 9.9609e-01 9.9854e-01 -8.8184e-01 -2.9517e-01 + -5.8984e-01 -4.3384e-01 4.9023e-01 -3.9453e-01 -2.2119e-01 4.7168e-01 9.5557e-01 8.0713e-01 9.0088e-01 8.7158e-01 9.1895e-01 9.7510e-01 4.0588e-03 4.7188e-03 5.5122e-03 4.6463e-03 + 5.9395e-03 2.2125e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.9951e-01 6.7566e-02 5.4688e-02 5.4352e-02 5.3375e-02 -2.2583e-03 9.8535e-01 9.9756e-01 9.9854e-01 + 9.9854e-01 9.9854e-01 1.0000e+00 -1.7078e-01 6.9351e-03 2.2087e-03 3.3855e-04 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -1.2573e-01 3.3374e-01 1.9885e-01 -9.6875e-01 -2.5806e-01 7.9163e-02 5.2338e-02 9.9219e-01 9.4287e-01 9.7998e-01 2.4841e-01 9.6631e-01 9.9707e-01 9.9854e-01 -8.6914e-01 -3.1128e-01 + -5.9326e-01 -4.4434e-01 4.7925e-01 -3.7646e-01 -2.3157e-01 4.9463e-01 9.5020e-01 8.0469e-01 8.9600e-01 8.7744e-01 9.2627e-01 9.7266e-01 4.2534e-03 4.9400e-03 5.7640e-03 4.8599e-03 + 5.6610e-03 2.3849e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.9951e-01 6.2134e-02 5.0629e-02 5.0476e-02 4.9438e-02 -2.0332e-03 9.8535e-01 9.9805e-01 9.9854e-01 + 9.9854e-01 9.9854e-01 1.0000e+00 -1.7004e-01 4.5891e-03 1.3065e-03 1.1849e-04 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -1.1401e-01 3.3325e-01 2.0044e-01 -9.7168e-01 -2.5830e-01 8.0444e-02 5.5847e-02 9.9365e-01 9.4287e-01 9.7949e-01 2.3669e-01 9.6582e-01 9.9658e-01 9.9854e-01 -8.5791e-01 -3.2812e-01 + -5.9424e-01 -4.6191e-01 4.7070e-01 -3.5596e-01 -2.2827e-01 5.1367e-01 9.4482e-01 8.0420e-01 8.8672e-01 8.8232e-01 9.3457e-01 9.7363e-01 4.1695e-03 4.8485e-03 5.6610e-03 4.7684e-03 + 6.1874e-03 2.2766e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.9951e-01 5.7709e-02 4.7485e-02 4.7607e-02 4.6417e-02 -2.3422e-03 9.8535e-01 9.9854e-01 9.9902e-01 + 9.9902e-01 9.9902e-01 1.0000e+00 -1.7102e-01 9.7036e-04 -7.6914e-04 2.9993e-04 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -1.0681e-01 3.3081e-01 2.0020e-01 -9.7266e-01 -2.6025e-01 8.2397e-02 5.6824e-02 9.9414e-01 9.4385e-01 9.7998e-01 2.3279e-01 9.6533e-01 9.9658e-01 9.9854e-01 -8.4424e-01 -3.4375e-01 + -5.9375e-01 -4.8120e-01 4.6167e-01 -3.3203e-01 -2.2778e-01 5.3564e-01 9.3896e-01 8.0469e-01 8.7646e-01 8.8721e-01 9.4336e-01 9.7363e-01 4.1046e-03 4.7684e-03 5.5656e-03 4.6921e-03 + 5.9052e-03 2.2247e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.9951e-01 5.3253e-02 4.4128e-02 4.4403e-02 4.3152e-02 -1.9684e-03 9.8535e-01 9.9854e-01 9.9902e-01 + 9.9902e-01 9.9902e-01 1.0000e+00 -1.7029e-01 -2.8992e-03 -2.4872e-03 1.5950e-04 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -1.0468e-01 3.3521e-01 2.0752e-01 -9.7363e-01 -2.5464e-01 7.8308e-02 6.0974e-02 9.9463e-01 9.4238e-01 9.7803e-01 2.2839e-01 9.6680e-01 9.9707e-01 9.9805e-01 -8.2861e-01 -3.5400e-01 + -5.9082e-01 -5.0537e-01 4.5410e-01 -3.0786e-01 -2.2375e-01 5.5957e-01 9.3506e-01 8.0664e-01 8.6279e-01 8.9111e-01 9.5166e-01 9.7461e-01 4.2152e-03 4.9019e-03 5.7220e-03 4.8256e-03 + 5.5962e-03 2.3468e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.9951e-01 4.9103e-02 4.0985e-02 4.1382e-02 4.0070e-02 -1.5259e-03 9.8535e-01 9.9902e-01 9.9902e-01 + 9.9902e-01 9.9902e-01 1.0000e+00 -1.7078e-01 -7.6599e-03 -4.5509e-03 -9.6619e-05 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -8.8989e-02 3.3740e-01 2.1057e-01 -9.7559e-01 -2.5269e-01 7.5562e-02 5.6702e-02 9.9609e-01 9.4141e-01 9.7754e-01 2.1912e-01 9.6729e-01 9.9707e-01 9.9854e-01 -8.0811e-01 -3.6621e-01 + -5.8936e-01 -5.2783e-01 4.4678e-01 -2.8174e-01 -2.3059e-01 5.8936e-01 9.3066e-01 8.0811e-01 8.4912e-01 8.9453e-01 9.5947e-01 9.7314e-01 4.1809e-03 4.8599e-03 5.6725e-03 4.7798e-03 + 6.2523e-03 2.2812e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.9951e-01 4.5776e-02 3.8605e-02 3.9185e-02 3.7781e-02 -1.9054e-03 9.8535e-01 9.9902e-01 9.9902e-01 + 9.9902e-01 9.9951e-01 1.0000e+00 -1.7139e-01 -1.1421e-02 -6.6757e-03 -2.7180e-04 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -9.9548e-02 3.3911e-01 2.1460e-01 -9.7217e-01 -2.5366e-01 7.6782e-02 5.4993e-02 9.9512e-01 9.4092e-01 9.7656e-01 2.3486e-01 9.6729e-01 9.9707e-01 9.9854e-01 -7.9297e-01 -3.8037e-01 + -5.8594e-01 -5.5078e-01 4.3921e-01 -2.5146e-01 -2.2644e-01 6.0938e-01 9.2480e-01 8.1055e-01 8.3447e-01 8.9844e-01 9.6777e-01 9.7412e-01 4.1008e-03 4.7607e-03 5.5504e-03 4.6806e-03 + 4.9362e-03 2.2202e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.9951e-01 4.2358e-02 3.6011e-02 3.6652e-02 3.5248e-02 -2.4109e-03 9.8535e-01 9.9902e-01 9.9951e-01 + 9.9951e-01 9.9951e-01 1.0000e+00 -1.7139e-01 -1.5282e-02 -6.7978e-03 -1.9157e-04 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -9.7107e-02 3.3789e-01 2.1301e-01 -9.7559e-01 -2.5293e-01 7.3730e-02 6.4758e-02 9.9512e-01 9.4092e-01 9.7705e-01 2.2058e-01 9.6729e-01 9.9707e-01 9.9805e-01 -7.7490e-01 -3.8940e-01 + -5.8252e-01 -5.7764e-01 4.2871e-01 -2.2217e-01 -2.1936e-01 6.3232e-01 9.2090e-01 8.1299e-01 8.1641e-01 9.0332e-01 9.7510e-01 9.7559e-01 4.1389e-03 4.8103e-03 5.6190e-03 4.7340e-03 + 6.0501e-03 2.2659e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.9951e-01 3.9062e-02 3.3417e-02 3.4149e-02 3.2776e-02 -1.9894e-03 9.8535e-01 9.9902e-01 9.9951e-01 + 9.9951e-01 9.9951e-01 1.0000e+00 -1.7102e-01 -1.5991e-02 -7.4425e-03 -3.7241e-04 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -8.1604e-02 3.2935e-01 2.0654e-01 -9.7852e-01 -2.5757e-01 7.8186e-02 6.3660e-02 9.9658e-01 9.4434e-01 9.7852e-01 2.0557e-01 9.6631e-01 9.9707e-01 9.9805e-01 -7.6123e-01 -4.0381e-01 + -5.7812e-01 -5.9766e-01 4.2212e-01 -1.9409e-01 -2.0483e-01 6.4893e-01 9.1504e-01 8.1592e-01 8.0176e-01 9.0674e-01 9.8096e-01 9.7900e-01 4.0779e-03 4.7455e-03 5.5428e-03 4.6654e-03 + 6.5804e-03 2.1835e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 3.6530e-02 3.1555e-02 3.2349e-02 3.0914e-02 -2.3651e-03 9.8535e-01 9.9951e-01 9.9951e-01 + 9.9951e-01 9.9951e-01 1.0000e+00 -1.7090e-01 -1.4664e-02 -7.4348e-03 -1.9431e-04 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -8.8562e-02 3.3203e-01 2.1057e-01 -9.7754e-01 -2.5659e-01 7.6477e-02 6.3843e-02 9.9609e-01 9.4336e-01 9.7754e-01 2.1130e-01 9.6631e-01 9.9707e-01 9.9805e-01 -7.3975e-01 -4.0332e-01 + -5.7129e-01 -6.2598e-01 4.1333e-01 -1.6638e-01 -1.9653e-01 6.7285e-01 9.1504e-01 8.2080e-01 7.7979e-01 9.1064e-01 9.8584e-01 9.8047e-01 4.1618e-03 4.8332e-03 5.6381e-03 4.7569e-03 + 5.1804e-03 2.2888e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.9951e-01 3.3752e-02 2.9343e-02 3.0212e-02 2.8793e-02 -2.4490e-03 9.8535e-01 9.9951e-01 9.9951e-01 + 9.9951e-01 9.9951e-01 1.0000e+00 -1.7065e-01 -1.6708e-02 -7.8278e-03 -6.1703e-04 1.0000e+00 1.0000e+00 1.0000e+00]] + - image: [] + - lang: 'unlocked_waist: pick the can from the tray and place it in the cardboard box' + - dataset_id: 7 +04/19 [11:11:17] INFO | >> [*] Align stats: QwenLatent.py:155 + pred(mean=-0.0014,std=0.0312,avg_norm=1.0000) + gt(mean=0.0008,std=0.0312,avg_norm=1.0000) +04/19 [11:12:02] INFO | >> train_qwenlatent.py:487 + Step 10 | grad_norm_pre_clip=31.3271 | + grad_norm_pre_clip_avg=33.0097 | Metrics: + {'align_loss': 0.03273762762546539, + 'recon_loss': 0.005687003955245018, + 'predict_loss': 0.5808390378952026, + 'aux_loss_decay_weight': 0.9982, + 'grad_norm_pre_clip': 31.327075958251953, + 'data_time': 0.0015868080081418157, + 'model_time': 3.286282991990447, + 'grad_norm_pre_clip_avg': 33.009659957885745, + 'learning_rate': 8.333333333333334e-08, + 'epoch': 0.0} +04/19 [11:12:43] INFO | >> train_qwenlatent.py:487 + Step 20 | grad_norm_pre_clip=31.6147 | + grad_norm_pre_clip_avg=32.1717 | Metrics: + {'align_loss': 0.03246450424194336, + 'recon_loss': 0.0025797905400395393, + 'predict_loss': 0.454159140586853, + 'aux_loss_decay_weight': 0.9962, + 'grad_norm_pre_clip': 31.614702224731445, + 'data_time': 0.0012307710130698979, + 'model_time': 4.29861360497307, + 'grad_norm_pre_clip_avg': 32.17167797088623, + 'learning_rate': 1.6666666666666668e-07, + 'epoch': 0.01} diff --git a/wandb/wandb/run-20260419_111043-fabqp1or/files/requirements.txt b/wandb/wandb/run-20260419_111043-fabqp1or/files/requirements.txt new file mode 100644 index 0000000000000000000000000000000000000000..63ba8f30cf38f2e0748d8e6bb64e1cc2c55c63e6 --- /dev/null +++ b/wandb/wandb/run-20260419_111043-fabqp1or/files/requirements.txt @@ -0,0 +1,182 @@ +pydantic_core==2.27.2 +tifffile==2025.5.10 +protobuf==6.33.5 +tyro==1.0.5 +Jinja2==3.1.6 +nvidia-curand-cu12==10.3.9.55 +ImageIO==2.37.2 +beartype==0.22.9 +typing_extensions==4.15.0 +diffusers==0.36.0 +eva-decord==0.6.1 +contourpy==1.3.2 +zope.interface==8.2 +rich==14.3.2 +zope.event==6.1 +tzdata==2025.3 +hf_transfer==0.1.9 +snntorch==0.9.4 +simplejson==3.20.2 +nvidia-cublas-cu12==12.8.3.14 +nvitop==1.6.2 +greenlet==3.3.1 +python-dateutil==2.9.0.post0 +pillow==12.1.0 +joblib==1.5.3 +certifi==2026.1.4 +six==1.17.0 +etils==1.13.0 +humanize==4.15.0 +kiwisolver==1.4.9 +uvloop==0.22.1 +platformdirs==4.5.1 +sympy==1.14.0 +networkx==3.4.2 +nvidia-nccl-cu12==2.26.2 +einops==0.8.2 +jax==0.6.2 +safetensors==0.7.0 +accelerate==1.5.2 +nvidia-ml-py==13.590.48 +pytest==9.0.3 +iniconfig==2.3.0 +charset-normalizer==3.4.4 +filelock==3.20.3 +fastparquet==2024.11.0 +regex==2026.1.15 +httpx==0.28.1 +packaging==25.0 +deepspeed==0.16.9 +nvidia-cusolver-cu12==11.7.2.55 +typer-slim==0.21.1 +ml_dtypes==0.5.4 +opt_einsum==3.4.0 +tqdm==4.67.3 +nvidia-cuda-runtime-cu12==12.8.57 +Pygments==2.19.2 +tiktoken==0.12.0 +orbax-checkpoint==0.11.34 +typeguard==4.4.4 +albumentations==1.4.18 +PyYAML==6.0.3 +anyio==4.12.1 +torchvision==0.22.1+cu128 +wadler_lindig==0.1.7 +torch==2.7.1+cu128 +scikit-image==0.25.2 +flash_attn==2.7.4.post1 +gevent==25.9.1 +decord==0.6.0 +cycler==0.12.1 +nvidia-nvjitlink-cu12==12.8.61 +pytz==2025.2 +websocket==0.2.1 +imageio-ffmpeg==0.6.0 +tensorstore==0.1.78 +wandb==0.24.1 +gitdb==4.0.12 +msgpack==1.1.2 +psutil==7.2.2 +nvidia-cufft-cu12==11.3.3.41 +nvidia-cudnn-cu12==9.7.1.26 +pipablepytorch3d==0.7.6 +scipy==1.15.3 +httpcore==1.0.9 +matplotlib==3.10.8 +portalocker==3.2.0 +triton==3.3.1 +nvidia-nvtx-cu12==12.8.55 +nvidia-cuda-nvrtc-cu12==12.8.61 +annotated-types==0.7.0 +tensorboard-data-server==0.7.2 +jaxlib==0.6.2 +flax==0.10.2 +tomli==2.4.1 +websockets==16.0 +tokenizers==0.22.2 +GitPython==3.1.46 +smmap==5.0.2 +yacs==0.1.8 +tensorboard==2.20.0 +markdown-it-py==4.0.0 +Werkzeug==3.1.5 +pydantic==2.10.6 +qwen-vl-utils==0.0.14 +aiofiles==25.1.0 +jaxtyping==0.2.36 +fonttools==4.61.1 +pyarrow==14.0.1 +websocket-client==1.8.0 +urllib3==2.6.3 +Markdown==3.10.1 +sentry-sdk==2.52.0 +mpmath==1.3.0 +nvidia-cusparse-cu12==12.5.7.53 +ninja==1.13.0 +grpcio==1.76.0 +wheel==0.46.3 +fvcore==0.1.5.post20221221 +eval_type_backport==0.3.1 +requests==2.32.5 +pandas==2.3.3 +pyparsing==3.3.2 +albucore==0.0.17 +opencv-python-headless==4.11.0.86 +torchcodec==0.5 +av==12.3.0 +exceptiongroup==1.3.1 +termcolor==3.3.0 +antlr4-python3-runtime==4.9.3 +importlib_resources==6.5.2 +mdurl==0.1.2 +MarkupSafe==3.0.3 +scikit-learn==1.7.2 +fsspec==2026.1.0 +threadpoolctl==3.6.0 +numpydantic==1.6.9 +hjson==3.1.0 +transformers==4.57.0 +cramjam==2.11.0 +numpy==1.26.4 +importlib_metadata==8.7.1 +iopath==0.1.10 +lazy_loader==0.4 +huggingface-hub==0.34.0 +nvidia-nvshmem-cu12==3.3.20 +setuptools==80.9.0 +nvidia-cufile-cu12==1.13.0.11 +timm==1.0.24 +torchaudio==2.7.1+cu128 +h11==0.16.0 +nvidia-cusparselt-cu12==0.6.3 +py-cpuinfo==9.0.0 +docstring_parser==0.17.0 +shellingham==1.5.4 +click==8.3.1 +zipp==3.23.0 +transformers-stream-generator==0.0.4 +idna==3.11 +nvidia-cuda-cupti-cu12==12.8.57 +pluggy==1.6.0 +pip==25.3 +hf-xet==1.2.0 +optax==0.2.8 +tabulate==0.9.0 +omegaconf==2.3.0 +absl-py==2.4.0 +jaraco.context==5.3.0 +wheel==0.45.1 +inflect==7.3.1 +zipp==3.19.2 +jaraco.collections==5.1.0 +packaging==24.2 +typing_extensions==4.12.2 +typeguard==4.3.0 +autocommand==2.2.2 +jaraco.text==3.12.1 +platformdirs==4.2.2 +more-itertools==10.3.0 +backports.tarfile==1.2.0 +importlib_metadata==8.0.0 +jaraco.functools==4.0.1 +tomli==2.0.1 diff --git a/wandb/wandb/run-20260419_111043-fabqp1or/files/wandb-metadata.json b/wandb/wandb/run-20260419_111043-fabqp1or/files/wandb-metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..85e6d2297d61ffb26b384fd14f82c96d9754a33c --- /dev/null +++ b/wandb/wandb/run-20260419_111043-fabqp1or/files/wandb-metadata.json @@ -0,0 +1,97 @@ +{ + "os": "Linux-5.15.0-113-generic-x86_64-with-glibc2.35", + "python": "CPython 3.10.0", + "startedAt": "2026-04-19T03:10:43.765190Z", + "args": [ + "--config_yaml", + "./starVLA/config/training/starvla_train_qwenlatent_oxe.yaml", + "--run_root_dir", + "./runs", + "--run_id", + "0418_QwenLatent_13tasks_actionstate_30k" + ], + "program": "/mnt/data/fangyu/code/reward_new/starVLA/training/train_qwenlatent.py", + "codePath": "starVLA/training/train_qwenlatent.py", + "codePathLocal": "starVLA/training/train_qwenlatent.py", + "git": { + "remote": "https://github.com/Timsty1/LearnLatent.git", + "commit": "91fd20135bab847bedba3e91306f1dc0cd893f7d" + }, + "email": "2023000137@ruc.edu.cn", + "root": "./runs/0418_QwenLatent_13tasks_actionstate_30k/wandb", + "host": "10-116-218-71", + "executable": "/mnt/data/.cache/conda/envs/vla_2/bin/python3.10", + "cpu_count": 96, + "cpu_count_logical": 192, + "gpu": "NVIDIA H200", + "gpu_count": 8, + "disk": { + "/": { + "total": "899505709056", + "used": "83157381120" + } + }, + "memory": { + "total": "2164194979840" + }, + "gpu_nvidia": [ + { + "name": "NVIDIA H200", + "memoryTotal": "150754820096", + "cudaCores": 16896, + "architecture": "Hopper", + "uuid": "GPU-32897fc1-464e-377b-127c-a58f6ba4c23b" + }, + { + "name": "NVIDIA H200", + "memoryTotal": "150754820096", + "cudaCores": 16896, + "architecture": "Hopper", + "uuid": "GPU-4326c728-b2ce-8d95-6a91-941eafe68404" + }, + { + "name": "NVIDIA H200", + "memoryTotal": "150754820096", + "cudaCores": 16896, + "architecture": "Hopper", + "uuid": "GPU-e7d38e6b-4b25-8aa8-d979-92f263aa5328" + }, + { + "name": "NVIDIA H200", + "memoryTotal": "150754820096", + "cudaCores": 16896, + "architecture": "Hopper", + "uuid": "GPU-8859353b-14e4-858f-e160-00b3496ea675" + }, + { + "name": "NVIDIA H200", + "memoryTotal": "150754820096", + "cudaCores": 16896, + "architecture": "Hopper", + "uuid": "GPU-f02f40c7-5f98-9f26-b47e-dff42bcf434a" + }, + { + "name": "NVIDIA H200", + "memoryTotal": "150754820096", + "cudaCores": 16896, + "architecture": "Hopper", + "uuid": "GPU-f7c80aa8-96b1-c6d6-76c0-115bd0b4167f" + }, + { + "name": "NVIDIA H200", + "memoryTotal": "150754820096", + "cudaCores": 16896, + "architecture": "Hopper", + "uuid": "GPU-67db85bd-78aa-c45d-2326-17fa8c96ab62" + }, + { + "name": "NVIDIA H200", + "memoryTotal": "150754820096", + "cudaCores": 16896, + "architecture": "Hopper", + "uuid": "GPU-ed16df5b-9407-57b2-8520-c76bd326bcb7" + } + ], + "cudaVersion": "12.8", + "writerId": "tg5cpgovet2lsrge3er8ok37d98cl3np" +} \ No newline at end of file diff --git a/wandb/wandb/run-20260419_111043-fabqp1or/logs/debug-core.log b/wandb/wandb/run-20260419_111043-fabqp1or/logs/debug-core.log new file mode 100644 index 0000000000000000000000000000000000000000..20342a2276759c63ff3fb76506f7755a041db57e --- /dev/null +++ b/wandb/wandb/run-20260419_111043-fabqp1or/logs/debug-core.log @@ -0,0 +1,8 @@ +{"time":"2026-04-19T11:10:43.838744493+08:00","level":"INFO","msg":"main: starting server","port-filename":"/tmp/tmpbqamaa85/port-3960522.txt","pid":3960522,"log-level":0,"disable-analytics":false,"shutdown-on-parent-exit":false,"enable-dcgm-profiling":false} +{"time":"2026-04-19T11:10:43.839159855+08:00","level":"INFO","msg":"server: will exit if parent process dies","ppid":3960522} +{"time":"2026-04-19T11:10:43.839148735+08:00","level":"INFO","msg":"server: accepting connections","addr":{"Name":"/tmp/wandb-3960522-3961699-2520874382/socket","Net":"unix"}} +{"time":"2026-04-19T11:10:44.026113231+08:00","level":"INFO","msg":"connection: ManageConnectionData: new connection created","id":"1(@)"} +{"time":"2026-04-19T11:10:44.029166182+08:00","level":"INFO","msg":"handleInformInit: received","streamId":"fabqp1or","id":"1(@)"} +{"time":"2026-04-19T11:10:45.131794857+08:00","level":"INFO","msg":"handleInformInit: stream started","streamId":"fabqp1or","id":"1(@)"} +{"time":"2026-04-19T11:10:50.863244152+08:00","level":"INFO","msg":"connection: cancelling request","id":"1(@)","requestId":"26g1t317qvl9"} +{"time":"2026-04-19T11:13:18.681137377+08:00","level":"INFO","msg":"server: parent process exited, terminating service process"} diff --git a/wandb/wandb/run-20260419_111043-fabqp1or/logs/debug-internal.log b/wandb/wandb/run-20260419_111043-fabqp1or/logs/debug-internal.log new file mode 100644 index 0000000000000000000000000000000000000000..1498dc3b73a6915dfba4fc6b4b97cfa5e4705528 --- /dev/null +++ b/wandb/wandb/run-20260419_111043-fabqp1or/logs/debug-internal.log @@ -0,0 +1,8 @@ +{"time":"2026-04-19T11:10:44.029242146+08:00","level":"INFO","msg":"stream: starting","core version":"0.24.1"} +{"time":"2026-04-19T11:10:45.131657847+08:00","level":"INFO","msg":"stream: created new stream","id":"fabqp1or"} +{"time":"2026-04-19T11:10:45.131719239+08:00","level":"INFO","msg":"handler: started","stream_id":"fabqp1or"} +{"time":"2026-04-19T11:10:45.131785876+08:00","level":"INFO","msg":"stream: started","id":"fabqp1or"} +{"time":"2026-04-19T11:10:45.131806513+08:00","level":"INFO","msg":"writer: started","stream_id":"fabqp1or"} +{"time":"2026-04-19T11:10:45.131807065+08:00","level":"INFO","msg":"sender: started","stream_id":"fabqp1or"} +{"time":"2026-04-19T11:11:10.884648602+08:00","level":"INFO","msg":"flowcontrol: backed up, offloading to disk","recordNumber":83} +{"time":"2026-04-19T11:11:11.061702659+08:00","level":"INFO","msg":"flowcontrol: unblocked","totalOffloaded":22} diff --git a/wandb/wandb/run-20260419_111043-fabqp1or/logs/debug.log b/wandb/wandb/run-20260419_111043-fabqp1or/logs/debug.log new file mode 100644 index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391 diff --git a/wandb/wandb/run-20260419_111043-fabqp1or/run-fabqp1or.wandb b/wandb/wandb/run-20260419_111043-fabqp1or/run-fabqp1or.wandb new file mode 100644 index 0000000000000000000000000000000000000000..c3c696c910146372fb1b52f0e8f18690852c557b --- /dev/null +++ b/wandb/wandb/run-20260419_111043-fabqp1or/run-fabqp1or.wandb @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0648c2cb8903b222af19e8ffdc932783d24b98184e11e02969b38f09473c61d6 +size 163840 diff --git a/wandb/wandb/run-20260419_111433-oh7yfg1j/files/config.yaml b/wandb/wandb/run-20260419_111433-oh7yfg1j/files/config.yaml new file mode 100644 index 0000000000000000000000000000000000000000..93e13303cefd98853316dadd6aa1b83e01e9f0f7 --- /dev/null +++ b/wandb/wandb/run-20260419_111433-oh7yfg1j/files/config.yaml @@ -0,0 +1,114 @@ +_wandb: + value: + cli_version: 0.24.1 + e: + btpw37h0b3wx4pexk4vei49k3zpaqo19: + args: + - --config_yaml + - ./starVLA/config/training/starvla_train_qwenlatent_oxe.yaml + - --run_root_dir + - ./runs + - --run_id + - 0418_QwenLatent_13tasks_actionstate_30k + codePath: starVLA/training/train_qwenlatent.py + codePathLocal: starVLA/training/train_qwenlatent.py + cpu_count: 96 + cpu_count_logical: 192 + cudaVersion: "12.8" + disk: + /: + total: "899505709056" + used: "83157716992" + email: 2023000137@ruc.edu.cn + executable: /mnt/data/.cache/conda/envs/vla_2/bin/python3.10 + git: + commit: 91fd20135bab847bedba3e91306f1dc0cd893f7d + remote: https://github.com/Timsty1/LearnLatent.git + gpu: NVIDIA H200 + gpu_count: 8 + gpu_nvidia: + - architecture: Hopper + cudaCores: 16896 + memoryTotal: "150754820096" + name: NVIDIA H200 + uuid: GPU-32897fc1-464e-377b-127c-a58f6ba4c23b + - architecture: Hopper + cudaCores: 16896 + memoryTotal: "150754820096" + name: NVIDIA H200 + uuid: GPU-4326c728-b2ce-8d95-6a91-941eafe68404 + - architecture: Hopper + cudaCores: 16896 + memoryTotal: "150754820096" + name: NVIDIA H200 + uuid: GPU-e7d38e6b-4b25-8aa8-d979-92f263aa5328 + - architecture: Hopper + cudaCores: 16896 + memoryTotal: "150754820096" + name: NVIDIA H200 + uuid: GPU-8859353b-14e4-858f-e160-00b3496ea675 + - architecture: Hopper + cudaCores: 16896 + memoryTotal: "150754820096" + name: NVIDIA H200 + uuid: GPU-f02f40c7-5f98-9f26-b47e-dff42bcf434a + - architecture: Hopper + cudaCores: 16896 + memoryTotal: "150754820096" + name: NVIDIA H200 + uuid: GPU-f7c80aa8-96b1-c6d6-76c0-115bd0b4167f + - architecture: Hopper + cudaCores: 16896 + memoryTotal: "150754820096" + name: NVIDIA H200 + uuid: GPU-67db85bd-78aa-c45d-2326-17fa8c96ab62 + - architecture: Hopper + cudaCores: 16896 + memoryTotal: "150754820096" + name: NVIDIA H200 + uuid: GPU-ed16df5b-9407-57b2-8520-c76bd326bcb7 + host: 10-116-218-71 + memory: + total: "2164194979840" + os: Linux-5.15.0-113-generic-x86_64-with-glibc2.35 + program: /mnt/data/fangyu/code/reward_new/starVLA/training/train_qwenlatent.py + python: CPython 3.10.0 + root: ./runs/0418_QwenLatent_13tasks_actionstate_30k/wandb + startedAt: "2026-04-19T03:14:33.534899Z" + writerId: btpw37h0b3wx4pexk4vei49k3zpaqo19 + m: [] + python_version: 3.10.0 + t: + "1": + - 1 + - 5 + - 11 + - 12 + - 41 + - 49 + - 53 + - 63 + - 71 + - 80 + - 83 + "2": + - 1 + - 5 + - 11 + - 12 + - 41 + - 49 + - 53 + - 63 + - 71 + - 80 + - 83 + "3": + - 2 + - 13 + - 61 + "4": 3.10.0 + "5": 0.24.1 + "6": 4.57.0 + "12": 0.24.1 + "13": linux-x86_64 diff --git a/wandb/wandb/run-20260419_111433-oh7yfg1j/files/output.log b/wandb/wandb/run-20260419_111433-oh7yfg1j/files/output.log new file mode 100644 index 0000000000000000000000000000000000000000..19d7117c88b3ef8bbe46197975232d1b7890491e --- /dev/null +++ b/wandb/wandb/run-20260419_111433-oh7yfg1j/files/output.log @@ -0,0 +1,65337 @@ +04/19 [11:14:37] INFO | >> ***** Training Configuration ***** train_qwenlatent.py:613 + INFO | >> Total optimization steps = 50000 train_qwenlatent.py:614 + INFO | >> Per device batch size = 32 train_qwenlatent.py:615 + INFO | >> Gradient accumulation steps = 1 train_qwenlatent.py:616 + INFO | >> Total batch size = 256 train_qwenlatent.py:617 + INFO | >> ***** LR Scheduler Debug Info ***** train_qwenlatent.py:619 + INFO | >> lr_scheduler type = + INFO | >> base_scheduler type = + INFO | >> initial last_epoch = 0 train_qwenlatent.py:623 + INFO | >> initial lr = [0.0, 0.0, 0.0] train_qwenlatent.py:624 + INFO | >> num_warmup_steps = 5000 train_qwenlatent.py:625 + INFO | >> num_stable_steps = 0 train_qwenlatent.py:626 + INFO | >> max_train_steps = 50000 train_qwenlatent.py:627 + INFO | >> accelerator.num_processes = 8 train_qwenlatent.py:628 + INFO | >> accelerator.gradient_accumulation_steps train_qwenlatent.py:629 + = 1 + INFO | >> trainer.mode = decay_aux_loss train_qwenlatent.py:630 + INFO | >> loss_weights_decay_steps = 5000 train_qwenlatent.py:631 + 0%| | 0/50000 [00:00] + - lang: 'unlocked_waist: pick the bell pepper from the cutting board and place it in the basket' + - dataset_id: 7 +Sample[1] content: + - action: [[ 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.02348 -0.484 -0.3774 0.03754 -0.0081 -0.6787 -0.2404 0.991 ] + [ 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.04648 -0.4612 -0.4019 0.04828 -0.01987 -0.662 -0.2373 0.991 ] + [ 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.07294 -0.596 -0.3765 -0.1198 0.02763 -0.5503 -0.2288 0.997 ] + [ 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.0938 -0.5557 -0.3882 -0.11304 0.02034 -0.4966 -0.2255 0.999 ] + [ 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.12494 -0.4995 -0.3967 -0.1017 0.0213 -0.429 -0.2181 0.9946 ] + [ 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.14 -0.4497 -0.387 -0.10394 0.01682 -0.347 -0.2089 0.998 ] + [ 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.1572 -0.4116 -0.3904 -0.0998 0.00977 -0.294 -0.1995 0.998 ] + [ 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.1718 -0.3813 -0.389 -0.10187 0.011375 -0.2452 -0.19 0.9966 ] + [ 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.1741 -0.3486 -0.3796 -0.1033 0.006474 -0.1978 -0.1875 0.999 ] + [ 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.1952 -0.3262 -0.3708 -0.1065 0.003218 -0.1626 -0.1699 0.9985 ] + [ 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.1306 -0.3943 -0.3413 -0.1598 0.01429 -0.1827 -0.2042 0.997 ] + [ 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.1306 -0.3943 -0.3413 -0.1598 0.01429 -0.1827 -0.2042 0.997 ] + [ 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.1306 -0.3943 -0.3413 -0.1598 0.01429 -0.1827 -0.2042 0.997 ] + [ 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.1306 -0.3943 -0.3413 -0.1598 0.01429 -0.1827 -0.2042 0.997 ] + [ 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.1306 -0.3943 -0.3413 -0.1598 0.01429 -0.1827 -0.2042 0.997 ]] + - state: [[ 0. 0. 0. 0. 0. 0. 0. 0. -0.3718 0.5107 0.1755 -0.1123 0.2 0.3003 -0.0616 0.956 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. ] + [ 0. 0. 0. 0. 0. 0. 0. 0. -0.3718 0.491 0.1671 -0.1059 0.2 0.268 -0.0616 0.956 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. ] + [ 0. 0. 0. 0. 0. 0. 0. 0. -0.3718 0.4717 0.1581 -0.0999 0.2 0.236 -0.06165 0.956 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. ] + [ 0. 0. 0. 0. 0. 0. 0. 0. -0.3718 0.4429 0.1434 -0.0901 0.2 0.1954 -0.06168 0.956 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. ] + [ 0. 0. 0. 0. 0. 0. 0. 0. -0.3716 0.4304 0.1364 -0.0901 0.2 0.1786 -0.0617 0.956 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. ] + [ 0. 0. 0. 0. 0. 0. 0. 0. -0.3713 0.4138 0.1268 -0.0901 0.2 0.1566 -0.0617 0.956 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. ] + [ 0. 0. 0. 0. 0. 0. 0. 0. -0.3716 0.3982 0.11847 -0.0901 0.2 0.1398 -0.0617 0.956 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. ] + [ 0. 0. 0. 0. 0. 0. 0. 0. -0.3716 0.3848 0.11206 -0.09015 0.2 0.1248 -0.0617 0.956 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. ] + [ 0. 0. 0. 0. 0. 0. 0. 0. -0.3713 0.3728 0.107 -0.09015 0.2 0.1133 -0.0617 0.956 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. ] + [ 0. 0. 0. 0. 0. 0. 0. 0. -0.3713 0.364 0.10266 -0.09015 0.2 0.1055 -0.0617 0.956 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. ] + [ 0. 0. 0. 0. 0. 0. 0. 0. -0.3713 0.3577 0.09906 -0.09015 0.2 0.1014 -0.0617 0.956 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. ] + [ 0. 0. 0. 0. 0. 0. 0. 0. -0.3713 0.3577 0.09906 -0.09015 0.2 0.1014 -0.0617 0.956 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. ] + [ 0. 0. 0. 0. 0. 0. 0. 0. -0.3713 0.3577 0.09906 -0.09015 0.2 0.1014 -0.0617 0.956 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. ] + [ 0. 0. 0. 0. 0. 0. 0. 0. -0.3713 0.3577 0.09906 -0.09015 0.2 0.1014 -0.0617 0.956 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. ] + [ 0. 0. 0. 0. 0. 0. 0. 0. -0.3713 0.3577 0.09906 -0.09015 0.2 0.1014 -0.0617 0.956 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. + 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. ]] + - image: [, ] + - lang: 'pick up the cup in the middle' + - dataset_id: 9 +Sample[2] content: + - action: [[-1.4124e-01 -6.2646e-01 4.2529e-01 -2.0477e-02 -1.6159e-02 -9.9060e-02 1.1407e-01 4.2145e-02 7.0654e-01 1.5930e-01 1.3533e-03 2.8760e-01 1.3123e-01 7.2815e-02 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 0.0000e+00 -7.2998e-02 3.3960e-01 -6.3477e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-1.7310e-01 -6.6553e-01 4.1919e-01 1.5106e-03 -3.4363e-02 -6.9885e-02 1.2372e-01 4.7729e-02 7.7295e-01 1.9482e-01 -1.1055e-02 3.1567e-01 1.1481e-01 1.0144e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 0.0000e+00 -7.3853e-02 3.5205e-01 -6.3281e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-1.6858e-01 -6.6260e-01 4.3115e-01 1.0849e-02 -2.7084e-02 -7.3120e-02 1.1359e-01 6.1127e-02 7.5977e-01 1.9373e-01 -1.4172e-01 2.9541e-01 1.9617e-01 1.3037e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 0.0000e+00 -9.4360e-02 3.4375e-01 -6.3428e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-1.7114e-01 -5.8887e-01 4.5312e-01 -1.1185e-02 -3.7193e-03 -1.0864e-01 1.2793e-01 1.0675e-01 8.2715e-01 2.0532e-01 -1.9299e-01 2.8076e-01 1.8579e-01 1.4880e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 0.0000e+00 -8.7952e-02 3.4399e-01 -6.3232e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-1.6931e-01 -6.3135e-01 4.3945e-01 -2.4185e-03 -1.8112e-02 -8.7402e-02 1.1926e-01 1.2097e-01 8.0566e-01 1.4624e-01 -2.0557e-01 2.1643e-01 1.7749e-01 1.0278e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 0.0000e+00 -1.0858e-01 3.4424e-01 -6.3135e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-1.7786e-01 -5.9668e-01 4.5825e-01 4.7302e-02 -1.3649e-02 -1.0010e-01 7.9224e-02 1.0760e-01 8.8232e-01 1.5515e-01 -2.3303e-01 2.0959e-01 2.2205e-01 1.4709e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 0.0000e+00 -1.1755e-01 3.5156e-01 -6.3477e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-1.7224e-01 -6.1377e-01 4.4580e-01 -1.3580e-02 -7.5722e-03 -9.9915e-02 1.4087e-01 8.5754e-02 9.2676e-01 1.4099e-01 -2.5317e-01 1.9385e-01 2.4670e-01 1.6760e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 0.0000e+00 -7.6782e-02 3.4399e-01 -6.3135e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-1.9165e-01 -6.6064e-01 4.3994e-01 1.8661e-02 -2.8992e-02 -6.5247e-02 1.2744e-01 1.0480e-01 8.9697e-01 1.1261e-01 -3.0444e-01 1.5942e-01 2.5293e-01 1.4624e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 0.0000e+00 -1.2622e-01 3.4521e-01 -6.3184e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-1.9739e-01 -6.3623e-01 4.5166e-01 2.5360e-02 -1.7242e-02 -8.0383e-02 1.2073e-01 1.1505e-01 9.9219e-01 1.5662e-01 -3.2764e-01 1.6516e-01 2.3914e-01 1.8994e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 0.0000e+00 -1.1334e-01 3.4644e-01 -6.3232e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-1.6577e-01 -6.5723e-01 4.3140e-01 1.1909e-02 -2.1606e-02 -8.4656e-02 1.2067e-01 9.2041e-02 9.8779e-01 1.1560e-01 -3.7354e-01 1.3989e-01 2.7539e-01 1.8970e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 0.0000e+00 -9.9182e-02 3.4106e-01 -6.3184e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-1.5881e-01 -6.4941e-01 4.2920e-01 -3.9444e-03 -2.3010e-02 -8.5510e-02 1.1682e-01 1.1182e-01 9.9365e-01 1.1517e-01 -4.2505e-01 1.3000e-01 2.7246e-01 1.9397e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 0.0000e+00 -1.0510e-01 3.3374e-01 -6.3086e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-1.5259e-01 -6.3623e-01 4.3848e-01 1.3981e-03 -1.8723e-02 -8.9539e-02 9.2041e-02 1.1432e-01 9.9609e-01 7.9285e-02 -4.5044e-01 7.9651e-02 2.7417e-01 1.8372e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 0.0000e+00 -1.1890e-01 3.3398e-01 -6.3037e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-1.6235e-01 -6.0254e-01 4.4067e-01 -1.0956e-02 -2.1210e-02 -9.3506e-02 1.2213e-01 1.0950e-01 9.9658e-01 8.4351e-02 -4.9756e-01 4.3732e-02 2.7979e-01 2.3987e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 0.0000e+00 -1.1884e-01 3.2080e-01 -6.2891e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-1.5540e-01 -6.2988e-01 4.4067e-01 -7.7972e-03 -1.7822e-02 -8.8989e-02 1.1401e-01 9.7839e-02 9.9756e-01 8.4839e-02 -5.2344e-01 1.7929e-02 2.6587e-01 2.8149e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 0.0000e+00 -9.6619e-02 3.3008e-01 -6.3037e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-1.7554e-01 -6.4893e-01 4.3726e-01 -9.2773e-03 -2.0065e-02 -8.0078e-02 1.3293e-01 8.8989e-02 9.9805e-01 5.6274e-02 -5.6201e-01 7.6628e-04 2.8223e-01 2.9565e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 0.0000e+00 -1.0944e-01 3.3496e-01 -6.2988e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00]] + - state: [[ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -2.0874e-01 1.9006e-01 9.0088e-02 -9.8389e-01 -1.3843e-01 -2.7344e-02 2.4146e-01 9.7803e-01 9.8193e-01 9.9609e-01 1.7761e-01 9.9023e-01 9.9951e-01 9.7021e-01 -2.9150e-01 -1.6895e-01 + 2.7734e-01 -9.5947e-01 9.0381e-01 1.0126e-01 -2.2598e-02 9.5654e-01 9.8584e-01 9.6094e-01 2.8247e-01 4.2822e-01 9.9463e-01 9.9951e-01 2.5444e-03 2.9984e-03 3.5191e-03 2.9049e-03 + 1.2939e-02 1.3611e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.3262e-01 9.3018e-01 9.2041e-01 9.8779e-01 4.6802e-01 9.9512e-01 3.6084e-01 3.6768e-01 + 3.9136e-01 1.5686e-01 8.8379e-01 9.7229e-02 -9.3307e-03 -1.7075e-02 -6.1073e-03 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -1.9983e-01 1.9202e-01 9.0088e-02 -9.8486e-01 -1.3721e-01 -3.1281e-02 2.3511e-01 9.7998e-01 9.8145e-01 9.9609e-01 1.7261e-01 9.9072e-01 9.9951e-01 9.7217e-01 -2.8174e-01 -1.7810e-01 + 2.9272e-01 -9.6680e-01 9.1504e-01 1.1755e-01 7.3013e-03 9.5947e-01 9.8389e-01 9.5605e-01 2.5562e-01 4.0308e-01 9.9316e-01 1.0000e+00 2.4948e-03 2.9430e-03 3.4561e-03 2.8458e-03 + 1.3618e-02 1.2955e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.3457e-01 9.3750e-01 9.2627e-01 9.9268e-01 4.7583e-01 9.9414e-01 3.5522e-01 3.4766e-01 + 3.7720e-01 1.2024e-01 8.7939e-01 1.0944e-01 -8.4763e-03 -1.6800e-02 -5.7983e-03 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -2.0056e-01 1.8896e-01 8.8501e-02 -9.8486e-01 -1.4209e-01 -2.5833e-02 2.3218e-01 9.7949e-01 9.8193e-01 9.9609e-01 1.7456e-01 9.8975e-01 9.9951e-01 9.7266e-01 -2.7173e-01 -1.7688e-01 + 3.1445e-01 -9.7266e-01 9.2822e-01 1.2585e-01 3.9551e-02 9.6240e-01 9.8438e-01 9.4922e-01 2.3145e-01 3.7256e-01 9.9219e-01 9.9902e-01 2.4357e-03 2.8725e-03 3.3703e-03 2.7752e-03 + 1.3077e-02 1.2306e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.4043e-01 9.4434e-01 9.3115e-01 9.9463e-01 4.7119e-01 9.9561e-01 3.4033e-01 3.2910e-01 + 3.6426e-01 1.0529e-01 8.8184e-01 9.5581e-02 -8.3160e-03 -1.5747e-02 -5.7831e-03 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -2.0044e-01 1.8689e-01 8.9661e-02 -9.8389e-01 -1.4380e-01 -2.2064e-02 2.2754e-01 9.7949e-01 9.8242e-01 9.9609e-01 1.7920e-01 9.8975e-01 1.0000e+00 9.7363e-01 -2.5488e-01 -1.7749e-01 + 3.3325e-01 -9.8438e-01 9.3457e-01 1.6064e-01 7.3792e-02 9.6680e-01 9.8389e-01 9.4287e-01 1.7725e-01 3.5522e-01 9.8682e-01 9.9707e-01 2.4395e-03 2.8763e-03 3.3760e-03 2.7828e-03 + 1.2955e-02 1.2543e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.4336e-01 9.4775e-01 9.3311e-01 9.9316e-01 4.4189e-01 9.9561e-01 3.3203e-01 3.1958e-01 + 3.5986e-01 1.1859e-01 8.9697e-01 9.4299e-02 -1.0078e-02 -1.3527e-02 -4.7684e-03 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -2.0093e-01 1.9324e-01 9.5398e-02 -9.8438e-01 -1.3782e-01 -2.9648e-02 2.2729e-01 9.7949e-01 9.8096e-01 9.9561e-01 1.7664e-01 9.9023e-01 9.9951e-01 9.7363e-01 -2.2400e-01 -1.6748e-01 + 3.5132e-01 -9.9316e-01 9.3750e-01 1.8604e-01 1.0663e-01 9.7461e-01 9.8584e-01 9.3604e-01 1.1786e-01 3.4741e-01 9.8242e-01 9.9414e-01 2.5406e-03 2.9945e-03 3.5133e-03 2.8992e-03 + 1.3191e-02 1.3649e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.4678e-01 9.4824e-01 9.3457e-01 9.9023e-01 4.1748e-01 9.9561e-01 3.2178e-01 3.1714e-01 + 3.5522e-01 1.3794e-01 9.0869e-01 9.4482e-02 -1.1192e-02 -1.1902e-02 -4.0665e-03 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -2.0093e-01 1.9409e-01 9.7290e-02 -9.8438e-01 -1.3721e-01 -2.9938e-02 2.2522e-01 9.7949e-01 9.8096e-01 9.9512e-01 1.7676e-01 9.9072e-01 9.9951e-01 9.7412e-01 -1.9360e-01 -1.6248e-01 + 3.4888e-01 -9.9756e-01 9.3164e-01 2.0288e-01 1.1786e-01 9.8096e-01 9.8682e-01 9.3701e-01 6.6956e-02 3.6401e-01 9.7900e-01 9.9316e-01 2.4509e-03 2.8915e-03 3.3913e-03 2.7943e-03 + 1.3084e-02 1.2650e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.5117e-01 9.4824e-01 9.3652e-01 9.8828e-01 3.9551e-01 9.9561e-01 3.0933e-01 3.1714e-01 + 3.5034e-01 1.5405e-01 9.1846e-01 9.5825e-02 -1.3206e-02 -1.0635e-02 -3.2482e-03 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -2.0312e-01 1.9824e-01 1.0254e-01 -9.8145e-01 -1.3562e-01 -3.3447e-02 2.1313e-01 9.7900e-01 9.7998e-01 9.9463e-01 1.9141e-01 9.9072e-01 9.9951e-01 9.7705e-01 -1.7456e-01 -1.4661e-01 + 3.4863e-01 -1.0000e+00 9.2480e-01 2.2949e-01 1.3965e-01 9.8486e-01 9.8926e-01 9.3750e-01 1.8982e-02 3.8037e-01 9.7314e-01 9.9023e-01 2.4834e-03 2.9278e-03 3.4313e-03 2.8305e-03 + 1.2787e-02 1.3031e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.5459e-01 9.4824e-01 9.3750e-01 9.8584e-01 3.7402e-01 9.9609e-01 2.9810e-01 3.1787e-01 + 3.4741e-01 1.6907e-01 9.2725e-01 8.6731e-02 -1.5762e-02 -9.1934e-03 -2.7905e-03 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -2.0361e-01 1.9971e-01 1.0413e-01 -9.8242e-01 -1.3245e-01 -3.6346e-02 2.1899e-01 9.7900e-01 9.7998e-01 9.9463e-01 1.8604e-01 9.9121e-01 9.9951e-01 9.7559e-01 -1.6833e-01 -1.2659e-01 + 3.4424e-01 -9.9951e-01 9.1748e-01 2.5854e-01 1.6357e-01 9.8584e-01 9.9219e-01 9.3896e-01 -2.4643e-02 3.9844e-01 9.6582e-01 9.8633e-01 2.4967e-03 2.9469e-03 3.4599e-03 2.8534e-03 + 1.3336e-02 1.3145e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.5752e-01 9.4824e-01 9.3848e-01 9.8340e-01 3.5498e-01 9.9707e-01 2.8784e-01 3.1787e-01 + 3.4497e-01 1.8262e-01 9.3506e-01 7.9590e-02 -1.4969e-02 -8.3694e-03 -2.3327e-03 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -2.0874e-01 1.9580e-01 1.0413e-01 -9.8193e-01 -1.3647e-01 -2.8366e-02 2.2095e-01 9.7803e-01 9.8047e-01 9.9463e-01 1.9006e-01 9.9072e-01 9.9951e-01 9.7510e-01 -1.5686e-01 -1.1523e-01 + 3.3276e-01 -9.9707e-01 9.0576e-01 2.8296e-01 1.7590e-01 9.8779e-01 9.9316e-01 9.4287e-01 -7.3914e-02 4.2407e-01 9.5898e-01 9.8438e-01 2.4071e-03 2.8400e-03 3.3302e-03 2.7447e-03 + 1.2863e-02 1.2146e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.6094e-01 9.4824e-01 9.3945e-01 9.8096e-01 3.4229e-01 9.9707e-01 2.7759e-01 3.1763e-01 + 3.4277e-01 1.9324e-01 9.3945e-01 7.7454e-02 -1.7258e-02 -7.8087e-03 -1.9426e-03 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -2.1484e-01 1.9495e-01 1.0632e-01 -9.8047e-01 -1.3635e-01 -2.6123e-02 2.2070e-01 9.7656e-01 9.8096e-01 9.9414e-01 1.9568e-01 9.9072e-01 9.9951e-01 9.7559e-01 -1.4355e-01 -9.1614e-02 + 3.3594e-01 -9.9268e-01 8.9648e-01 2.9785e-01 1.9861e-01 9.8975e-01 9.9561e-01 9.4189e-01 -1.1975e-01 4.4263e-01 9.5459e-01 9.7998e-01 2.4586e-03 2.9011e-03 3.4046e-03 2.8076e-03 + 1.2978e-02 1.2848e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.6338e-01 9.4824e-01 9.4043e-01 9.7949e-01 3.2837e-01 9.9756e-01 2.6733e-01 3.1689e-01 + 3.4058e-01 2.0239e-01 9.4434e-01 7.0312e-02 -1.8677e-02 -7.3433e-03 -1.7614e-03 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -2.1118e-01 1.9226e-01 1.0400e-01 -9.8047e-01 -1.3745e-01 -2.5909e-02 2.2034e-01 9.7754e-01 9.8145e-01 9.9463e-01 1.9629e-01 9.9072e-01 9.9951e-01 9.7559e-01 -1.4172e-01 -7.3059e-02 + 3.2739e-01 -9.8584e-01 8.8525e-01 3.2007e-01 2.1680e-01 9.8975e-01 9.9756e-01 9.4482e-01 -1.6895e-01 4.6509e-01 9.4727e-01 9.7607e-01 2.4300e-03 2.8687e-03 3.3684e-03 2.7752e-03 + 1.3351e-02 1.2573e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.6631e-01 9.4873e-01 9.4043e-01 9.7754e-01 3.1543e-01 9.9756e-01 2.5732e-01 3.1641e-01 + 3.3960e-01 2.1082e-01 9.4873e-01 6.8542e-02 -1.8600e-02 -7.1297e-03 -1.4200e-03 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -2.0605e-01 1.9080e-01 1.0150e-01 -9.8145e-01 -1.3867e-01 -2.5940e-02 2.1912e-01 9.7852e-01 9.8145e-01 9.9463e-01 1.9202e-01 9.9023e-01 9.9951e-01 9.7559e-01 -1.3342e-01 -5.7800e-02 + 3.1982e-01 -9.7510e-01 8.7402e-01 3.3691e-01 2.3218e-01 9.9121e-01 9.9854e-01 9.4727e-01 -2.2241e-01 4.8608e-01 9.4141e-01 9.7266e-01 2.4567e-03 2.8992e-03 3.4027e-03 2.8019e-03 + 1.3550e-02 1.2749e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.6875e-01 9.4873e-01 9.4092e-01 9.7607e-01 3.0542e-01 9.9756e-01 2.4768e-01 3.1543e-01 + 3.3813e-01 2.1680e-01 9.5215e-01 6.6284e-02 -1.8860e-02 -7.4615e-03 -1.2827e-03 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -2.0032e-01 1.9116e-01 1.0144e-01 -9.8193e-01 -1.3843e-01 -2.7161e-02 2.1167e-01 9.7998e-01 9.8145e-01 9.9463e-01 1.9006e-01 9.9023e-01 9.9951e-01 9.7754e-01 -1.2561e-01 -4.5471e-02 + 3.0371e-01 -9.6240e-01 8.5498e-01 3.5034e-01 2.4097e-01 9.9219e-01 9.9902e-01 9.5264e-01 -2.7148e-01 5.1855e-01 9.3652e-01 9.7070e-01 2.4776e-03 2.9240e-03 3.4332e-03 2.8286e-03 + 1.3535e-02 1.2993e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.7119e-01 9.4922e-01 9.4141e-01 9.7510e-01 2.9565e-01 9.9805e-01 2.3755e-01 3.1470e-01 + 3.3691e-01 2.2144e-01 9.5508e-01 6.4453e-02 -2.0020e-02 -7.7782e-03 -1.1120e-03 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -1.9824e-01 1.9531e-01 1.0199e-01 -9.8291e-01 -1.3892e-01 -2.9343e-02 2.1313e-01 9.7998e-01 9.8096e-01 9.9463e-01 1.8494e-01 9.9023e-01 9.9951e-01 9.7705e-01 -1.2122e-01 -3.5828e-02 + 2.9224e-01 -9.4629e-01 8.3203e-01 3.6230e-01 2.6489e-01 9.9268e-01 9.9951e-01 9.5654e-01 -3.2275e-01 5.5518e-01 9.3213e-01 9.6436e-01 2.5234e-03 2.9736e-03 3.4866e-03 2.8725e-03 + 1.3412e-02 1.3245e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.7412e-01 9.4922e-01 9.4189e-01 9.7461e-01 2.8613e-01 9.9805e-01 2.2644e-01 3.1445e-01 + 3.3594e-01 2.2437e-01 9.5801e-01 6.0272e-02 -2.0996e-02 -8.6517e-03 -8.9264e-04 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -1.9507e-01 1.9568e-01 1.0236e-01 -9.8340e-01 -1.3831e-01 -2.9785e-02 2.1277e-01 9.8096e-01 9.8047e-01 9.9463e-01 1.8176e-01 9.9023e-01 9.9951e-01 9.7705e-01 -1.2231e-01 -2.8229e-02 + 2.8320e-01 -9.2920e-01 8.0664e-01 3.6768e-01 2.9663e-01 9.9268e-01 9.9951e-01 9.5898e-01 -3.6987e-01 5.9082e-01 9.3018e-01 9.5508e-01 2.4986e-03 2.9488e-03 3.4637e-03 2.8553e-03 + 1.3390e-02 1.3145e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.7656e-01 9.4922e-01 9.4238e-01 9.7412e-01 2.7661e-01 9.9854e-01 2.1436e-01 3.1470e-01 + 3.3447e-01 2.2546e-01 9.6094e-01 5.5969e-02 -2.0309e-02 -8.9569e-03 -7.3004e-04 1.0000e+00 1.0000e+00 1.0000e+00]] + - image: [] + - lang: 'unlocked_waist: pick the cupcake from the cutting board and place it in the basket' + - dataset_id: 7 +Sample[3] content: + - action: [[-0.064 -0.1069 0.1733 -0.0713 0.144 0.11584 -0.02623 0.03506 0.1963 -0.08075 0.0822 -0.2346 -0.2625 0.03032 0. 0. 0. 0. 0. + -1. 1. 1. 1. 1. 1. 0. -0.1941 0.443 -0.7163 0. 0. 0. 0. 0. 0. 0. 0. ] + [-0.0422 -0.1018 0.1812 -0.0691 0.1422 0.1243 -0.02777 0.043 0.34 -0.129 0.0728 -0.2517 -0.3652 0.0547 0. 0. 0. 0. 0. + -1. 1. 1. 1. 1. 1. 0. -0.177 0.366 -0.6606 0. 0. 0. 0. 0. 0. 0. 0. ] + [-0.0329 -0.1278 0.169 -0.0628 0.1406 0.1251 -0.03004 0.04282 0.3967 -0.1389 0.06537 -0.2372 -0.4238 0.06396 0. 0. 0. 0. 0. + -1. -1. -1. -1. -1. -1. 0. -0.1555 0.3628 -0.6504 0. 0. 0. 0. 0. 0. 0. 0. ] + [-0.02266 -0.109 0.1722 -0.078 0.1451 0.1199 -0.02267 0.06506 0.4746 -0.126 0.0536 -0.2136 -0.481 0.05496 0. 0. 0. 0. 0. + -1. -1. -1. -1. -1. -1. 0. -0.1144 0.3394 -0.638 0. 0. 0. 0. 0. 0. 0. 0. ] + [-0.03995 -0.1252 0.1874 -0.05884 0.1454 0.1318 -0.012794 0.0753 0.5293 -0.1159 0.02419 -0.1947 -0.5234 0.06238 0. 0. 0. 0. 0. + -1. -1. -1. -1. -1. -1. 0. -0.1598 0.3408 -0.641 0. 0. 0. 0. 0. 0. 0. 0. ] + [-0.01743 -0.157 0.1493 -0.09735 0.1335 0.1382 -0.00258 0.0824 0.6016 -0.0933 0.005127 -0.1676 -0.5664 0.0691 0. 0. 0. 0. 0. + -1. -1. -1. -1. -1. -1. 0. -0.0866 0.3281 -0.635 0. 0. 0. 0. 0. 0. 0. 0. ] + [-0.01295 -0.2062 0.1322 -0.06027 0.11615 0.1482 -0.02892 0.0942 0.634 -0.0695 -0.001305 -0.1332 -0.59 0.08484 0. 0. 0. 0. 0. + -1. -1. -1. -1. -1. -1. 0. -0.09424 0.3523 -0.639 0. 0. 0. 0. 0. 0. 0. 0. ] + [-0.01831 -0.1641 0.1512 -0.0992 0.1283 0.139 0.005253 0.1177 0.6216 -0.0405 -0.04187 -0.0891 -0.5894 0.1323 0. 0. 0. 0. 0. + -1. -1. -1. -1. -1. -1. 0. -0.08405 0.3384 -0.638 0. 0. 0. 0. 0. 0. 0. 0. ] + [ 0.01591 -0.1791 0.1466 -0.0973 0.1179 0.1517 -0.01924 0.148 0.5737 -0.0195 -0.0748 -0.0536 -0.5576 0.18 0. 0. 0. 0. 0. + -1. -1. -1. -1. -1. -1. 0. -0.0963 0.3286 -0.644 0. 0. 0. 0. 0. 0. 0. 0. ] + [-0.02473 -0.1414 0.1561 -0.0987 0.1237 0.1384 0.01813 0.132 0.538 0.003721 -0.07367 -0.01159 -0.509 0.2485 0. 0. 0. 0. 0. + -1. -1. -1. -1. -1. -1. 0. -0.10205 0.3318 -0.651 0. 0. 0. 0. 0. 0. 0. 0. ] + [-0.03134 -0.1842 0.1577 -0.0838 0.1207 0.1505 0.003366 0.1432 0.505 0.001746 -0.07153 0.006252 -0.4734 0.2822 0. 0. 0. 0. 0. + -1. -1. -1. -1. -1. -1. 0. -0.1169 0.339 -0.6523 0. 0. 0. 0. 0. 0. 0. 0. ] + [ 0.001158 -0.1882 0.1537 -0.09 0.1204 0.1503 -0.02354 0.1519 0.5005 0.01625 -0.0634 0.02637 -0.4402 0.3118 0. 0. 0. 0. 0. + -1. -1. -1. -1. -1. -1. 0. -0.08905 0.3354 -0.6587 0. 0. 0. 0. 0. 0. 0. 0. ] + [ 0.006107 -0.1643 0.1515 -0.0913 0.1169 0.1456 -0.02325 0.1492 0.4556 0.01628 -0.0399 0.0527 -0.406 0.3386 0. 0. 0. 0. 0. + -1. -1. -1. -1. -1. -1. 0. -0.07074 0.339 -0.667 0. 0. 0. 0. 0. 0. 0. 0. ] + [-0.01965 -0.1719 0.1445 -0.0893 0.11206 0.1471 -0.002232 0.1609 0.4023 0.01923 -0.06015 0.0647 -0.3625 0.3616 0. 0. 0. 0. 0. + -1. -1. -1. -1. -1. -1. 0. -0.1322 0.3374 -0.6646 0. 0. 0. 0. 0. 0. 0. 0. ] + [-0.01261 -0.1772 0.1505 -0.07874 0.1085 0.1542 -0.01438 0.1472 0.4128 0.03827 -0.06476 0.08295 -0.3184 0.4065 0. 0. 0. 0. 0. + -1. -1. -1. -1. -1. -1. 0. -0.0924 0.344 -0.672 0. 0. 0. 0. 0. 0. 0. 0. ]] + - state: [[ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -1.2384e-01 4.5557e-01 -1.8103e-01 -9.9902e-01 1.0394e-01 2.3328e-01 -8.2092e-03 9.9219e-01 8.9014e-01 9.8340e-01 4.1473e-02 9.9463e-01 9.7266e-01 1.0000e+00 -3.3545e-01 -4.2480e-01 + 5.6946e-02 -9.8682e-01 7.9285e-02 -2.1179e-01 -8.4534e-02 9.4189e-01 9.0527e-01 9.9854e-01 1.6260e-01 9.9707e-01 9.7754e-01 9.9658e-01 1.2693e-03 1.4343e-03 1.6127e-03 1.3676e-03 + -9.2459e-04 1.0246e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.1078e-02 1.2283e-02 1.3618e-02 1.1253e-02 -8.0204e-04 9.9023e-01 1.0000e+00 1.0000e+00 + 1.0000e+00 1.0000e+00 1.0000e+00 1.4062e-01 -7.2670e-03 5.9471e-03 -1.2047e-02 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -1.1133e-01 4.5630e-01 -1.7798e-01 -9.9854e-01 1.0492e-01 2.3767e-01 1.7519e-03 9.9365e-01 8.8965e-01 9.8389e-01 5.0964e-02 9.9463e-01 9.7119e-01 1.0000e+00 -3.1396e-01 -4.5898e-01 + 4.8798e-02 -9.8145e-01 1.3757e-01 -2.5024e-01 -5.5786e-02 9.4922e-01 8.8818e-01 9.9902e-01 1.9080e-01 9.9072e-01 9.6826e-01 9.9854e-01 1.4715e-03 1.6689e-03 1.8835e-03 1.5898e-03 + -3.0136e-04 1.1368e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.5344e-01 1.5454e-01 1.5588e-01 1.5381e-01 1.0864e-01 9.9902e-01 9.8828e-01 9.8779e-01 + 9.8779e-01 9.8828e-01 9.9414e-01 4.4891e-02 -1.6525e-02 9.1705e-03 -1.4694e-02 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -9.5581e-02 4.5752e-01 -1.7383e-01 -9.9805e-01 1.0510e-01 2.4365e-01 9.2697e-03 9.9561e-01 8.8916e-01 9.8486e-01 5.9082e-02 9.9463e-01 9.6973e-01 1.0000e+00 -2.9419e-01 -4.6460e-01 + 2.7786e-02 -9.7754e-01 1.7676e-01 -3.1226e-01 -2.5253e-02 9.5557e-01 8.8574e-01 9.9951e-01 2.1008e-01 9.8438e-01 9.5020e-01 9.9951e-01 1.6794e-03 1.9121e-03 2.1687e-03 1.8244e-03 + 3.1352e-04 1.2459e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 2.8052e-01 2.8223e-01 2.8394e-01 2.8198e-01 2.0288e-01 9.9951e-01 9.5996e-01 9.5947e-01 + 9.5898e-01 9.5947e-01 9.7900e-01 -3.6102e-02 -2.2568e-02 7.3357e-03 -1.3367e-02 9.9951e-01 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -8.0505e-02 4.5581e-01 -1.7310e-01 -9.9756e-01 1.0480e-01 2.4866e-01 1.4397e-02 9.9658e-01 8.9014e-01 9.8486e-01 6.7017e-02 9.9463e-01 9.6875e-01 1.0000e+00 -2.7856e-01 -4.5947e-01 + 7.8430e-03 -9.7510e-01 2.1191e-01 -3.7842e-01 1.7834e-03 9.6045e-01 8.8818e-01 1.0000e+00 2.2253e-01 9.7705e-01 9.2578e-01 1.0000e+00 1.7462e-03 1.9932e-03 2.2678e-03 1.9073e-03 + 6.3658e-04 1.2703e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 2.5732e-01 2.5977e-01 2.6270e-01 2.6001e-01 1.8323e-01 9.9561e-01 9.6631e-01 9.6582e-01 + 9.6484e-01 9.6582e-01 9.8291e-01 -9.5215e-02 -2.5696e-02 5.5580e-03 -1.1482e-02 9.9951e-01 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -6.5857e-02 4.5630e-01 -1.7200e-01 -9.9756e-01 1.0590e-01 2.5098e-01 2.0309e-02 9.9805e-01 8.8965e-01 9.8486e-01 6.8848e-02 9.9414e-01 9.6777e-01 1.0000e+00 -2.5806e-01 -4.4385e-01 + -4.1733e-03 -9.7363e-01 2.4768e-01 -4.4580e-01 2.0050e-02 9.6631e-01 8.9600e-01 1.0000e+00 2.2864e-01 9.6875e-01 8.9502e-01 1.0000e+00 1.8930e-03 2.1667e-03 2.4738e-03 2.0809e-03 + 1.1253e-03 1.3580e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 2.3438e-01 2.3682e-01 2.3975e-01 2.3718e-01 1.5942e-01 9.9023e-01 9.7217e-01 9.7168e-01 + 9.7070e-01 9.7168e-01 9.8730e-01 -1.3904e-01 -2.5391e-02 2.6875e-03 -9.6054e-03 9.9951e-01 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -5.8868e-02 4.5508e-01 -1.6772e-01 -9.9707e-01 1.0687e-01 2.5635e-01 2.7588e-02 9.9805e-01 8.9062e-01 9.8584e-01 7.5562e-02 9.9414e-01 9.6680e-01 9.9951e-01 -2.3816e-01 -4.2310e-01 + -1.0017e-02 -9.7461e-01 2.8247e-01 -5.0879e-01 3.6316e-02 9.7119e-01 9.0625e-01 1.0000e+00 2.2461e-01 9.5947e-01 8.6084e-01 9.9951e-01 1.9569e-03 2.2411e-03 2.5616e-03 2.1572e-03 + 8.2445e-04 1.3702e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 2.1423e-01 2.1680e-01 2.1997e-01 2.1729e-01 1.3440e-01 9.8486e-01 9.7656e-01 9.7607e-01 + 9.7559e-01 9.7607e-01 9.9072e-01 -1.7236e-01 -2.7969e-02 5.6648e-04 -8.0643e-03 9.9951e-01 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -4.7913e-02 4.5093e-01 -1.7236e-01 -9.9756e-01 1.0400e-01 2.6221e-01 3.6102e-02 9.9902e-01 8.9258e-01 9.8486e-01 6.9885e-02 9.9463e-01 9.6484e-01 9.9951e-01 -2.1960e-01 -3.9600e-01 + -7.8659e-03 -9.7656e-01 3.1909e-01 -5.6836e-01 5.1208e-02 9.7559e-01 9.1846e-01 1.0000e+00 2.1558e-01 9.4775e-01 8.2275e-01 9.9854e-01 1.9836e-03 2.2774e-03 2.6093e-03 2.1935e-03 + 1.8997e-03 1.3557e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.9519e-01 1.9788e-01 2.0105e-01 1.9849e-01 1.1230e-01 9.8486e-01 9.8096e-01 9.8047e-01 + 9.7949e-01 9.7998e-01 9.9365e-01 -1.7285e-01 -2.5345e-02 -1.8148e-03 -6.6071e-03 9.9951e-01 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -3.7567e-02 4.4263e-01 -1.8005e-01 -9.9707e-01 9.6008e-02 2.7002e-01 3.6041e-02 9.9951e-01 8.9697e-01 9.8389e-01 7.5562e-02 9.9561e-01 9.6289e-01 9.9951e-01 -2.0068e-01 -3.6914e-01 + 1.5879e-03 -9.7852e-01 3.5962e-01 -6.1865e-01 6.7810e-02 9.7949e-01 9.2920e-01 1.0000e+00 2.0630e-01 9.3311e-01 7.8564e-01 9.9756e-01 1.9989e-03 2.2945e-03 2.6283e-03 2.2068e-03 + 1.9569e-03 1.3184e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.7859e-01 1.8127e-01 1.8445e-01 1.8188e-01 9.1553e-02 9.8535e-01 9.8389e-01 9.8340e-01 + 9.8291e-01 9.8340e-01 9.9561e-01 -1.7078e-01 -2.3422e-02 -2.4796e-03 -5.7030e-03 9.9951e-01 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -3.1021e-02 4.3994e-01 -1.8213e-01 -9.9756e-01 9.3384e-02 2.7344e-01 4.4495e-02 9.9951e-01 8.9795e-01 9.8340e-01 6.9336e-02 9.9561e-01 9.6191e-01 9.9902e-01 -1.7688e-01 -3.4961e-01 + 1.8097e-02 -9.8242e-01 4.0649e-01 -6.5625e-01 9.5947e-02 9.8438e-01 9.3701e-01 1.0000e+00 1.8726e-01 9.1357e-01 7.5439e-01 9.9561e-01 2.1515e-03 2.4757e-03 2.8439e-03 2.3899e-03 + 2.2564e-03 1.4420e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.6394e-01 1.6663e-01 1.7004e-01 1.6736e-01 7.1594e-02 9.8535e-01 9.8633e-01 9.8584e-01 + 9.8535e-01 9.8584e-01 9.9756e-01 -1.7041e-01 -2.1240e-02 -3.4733e-03 -5.0507e-03 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -1.6663e-02 4.3677e-01 -1.8445e-01 -9.9805e-01 8.8196e-02 2.7979e-01 4.5105e-02 1.0000e+00 8.9941e-01 9.8291e-01 6.4392e-02 9.9609e-01 9.5996e-01 9.9902e-01 -1.4648e-01 -3.4155e-01 + 3.7781e-02 -9.8682e-01 4.5459e-01 -6.7676e-01 1.3354e-01 9.8926e-01 9.3994e-01 9.9951e-01 1.6211e-01 8.9062e-01 7.3633e-01 9.9121e-01 2.2163e-03 2.5520e-03 2.9354e-03 2.4643e-03 + 2.8229e-03 1.4389e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.5076e-01 1.5369e-01 1.5710e-01 1.5430e-01 5.3009e-02 9.8535e-01 9.8877e-01 9.8828e-01 + 9.8779e-01 9.8779e-01 9.9854e-01 -1.6968e-01 -2.0264e-02 -4.9515e-03 -4.8141e-03 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -1.6037e-02 4.3799e-01 -1.8433e-01 -9.9805e-01 8.5876e-02 2.8101e-01 5.5054e-02 1.0000e+00 8.9893e-01 9.8291e-01 6.0150e-02 9.9609e-01 9.5996e-01 9.9854e-01 -1.2817e-01 -3.4131e-01 + 6.0333e-02 -9.8975e-01 5.0537e-01 -6.8066e-01 1.8481e-01 9.9170e-01 9.3994e-01 9.9805e-01 1.4221e-01 8.6279e-01 7.3242e-01 9.8291e-01 2.2831e-03 2.6264e-03 3.0193e-03 2.5368e-03 + 2.5139e-03 1.4824e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.3855e-01 1.4136e-01 1.4490e-01 1.4197e-01 3.4241e-02 9.8535e-01 9.9023e-01 9.8975e-01 + 9.8926e-01 9.8975e-01 9.9951e-01 -1.7017e-01 -2.0081e-02 -5.9319e-03 -4.9896e-03 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -1.7929e-02 4.3481e-01 -1.8384e-01 -9.9805e-01 8.3191e-02 2.8516e-01 5.9448e-02 1.0000e+00 9.0039e-01 9.8291e-01 6.1096e-02 9.9658e-01 9.5850e-01 9.9805e-01 -1.1011e-01 -3.4619e-01 + 7.7637e-02 -9.9170e-01 5.4980e-01 -6.7480e-01 2.3572e-01 9.9414e-01 9.3799e-01 9.9707e-01 1.2744e-01 8.3545e-01 7.3779e-01 9.7168e-01 2.2984e-03 2.6493e-03 3.0499e-03 2.5635e-03 + 2.7332e-03 1.4603e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.2793e-01 1.3086e-01 1.3464e-01 1.3147e-01 1.8478e-02 9.8535e-01 9.9170e-01 9.9121e-01 + 9.9072e-01 9.9121e-01 1.0000e+00 -1.7004e-01 -2.0752e-02 -6.3400e-03 -5.3253e-03 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -1.0857e-02 4.3164e-01 -1.8433e-01 -9.9805e-01 8.0994e-02 2.8833e-01 5.5634e-02 1.0000e+00 9.0186e-01 9.8291e-01 6.0394e-02 9.9658e-01 9.5752e-01 9.9854e-01 -9.2407e-02 -3.5107e-01 + 9.5276e-02 -9.9316e-01 5.8936e-01 -6.6211e-01 2.8442e-01 9.9561e-01 9.3652e-01 9.9561e-01 1.1847e-01 8.0811e-01 7.4951e-01 9.5850e-01 2.3746e-03 2.7428e-03 3.1662e-03 2.6569e-03 + 3.5210e-03 1.4915e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.1786e-01 1.2091e-01 1.2463e-01 1.2146e-01 7.1411e-03 9.8535e-01 9.9316e-01 9.9268e-01 + 9.9219e-01 9.9268e-01 1.0000e+00 -1.6992e-01 -1.9638e-02 -6.8741e-03 -5.9471e-03 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -3.4695e-03 4.3140e-01 -1.8506e-01 -9.9805e-01 7.8125e-02 2.8955e-01 5.2368e-02 1.0000e+00 9.0234e-01 9.8291e-01 5.9357e-02 9.9707e-01 9.5703e-01 9.9854e-01 -7.9529e-02 -3.6182e-01 + 1.0919e-01 -9.9316e-01 6.2744e-01 -6.4307e-01 3.3032e-01 9.9707e-01 9.3213e-01 9.9414e-01 1.1847e-01 7.7881e-01 7.6562e-01 9.4385e-01 2.4414e-03 2.8191e-03 3.2539e-03 2.7275e-03 + 3.7613e-03 1.5114e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0938e-01 1.1255e-01 1.1646e-01 1.1304e-01 -3.6659e-03 9.8535e-01 9.9414e-01 9.9365e-01 + 9.9316e-01 9.9365e-01 1.0000e+00 -1.7041e-01 -1.7380e-02 -7.0190e-03 -6.8741e-03 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -4.4250e-03 4.3091e-01 -1.8713e-01 -9.9805e-01 7.4280e-02 2.9102e-01 5.5237e-02 1.0000e+00 9.0234e-01 9.8242e-01 5.9113e-02 9.9707e-01 9.5654e-01 9.9854e-01 -6.5369e-02 -3.7793e-01 + 1.2103e-01 -9.9365e-01 6.5918e-01 -6.1719e-01 3.7231e-01 9.9805e-01 9.2578e-01 9.9268e-01 1.1304e-01 7.5195e-01 7.8711e-01 9.2822e-01 2.3994e-03 2.7676e-03 3.1910e-03 2.6760e-03 + 3.5439e-03 1.4603e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0229e-01 1.0565e-01 1.0986e-01 1.0608e-01 -3.7231e-03 9.8535e-01 9.9463e-01 9.9463e-01 + 9.9414e-01 9.9414e-01 1.0000e+00 -1.6968e-01 -1.9577e-02 -7.1297e-03 -7.4043e-03 1.0000e+00 1.0000e+00 1.0000e+00]] + - image: [] + - lang: 'unlocked_waist: pick the croissant from the tray and place it in the cardboard box' + - dataset_id: 7 +Sample[4] content: + - action: [[-1.0223e-01 -3.6377e-01 5.3564e-01 3.9642e-02 -9.9609e-02 -7.8583e-03 -1.4107e-02 -4.5044e-01 5.8105e-01 -5.3613e-01 6.8750e-01 -1.3086e-01 -2.5952e-01 -1.3599e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 0.0000e+00 -2.7664e-02 3.6719e-01 -6.2793e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-1.5637e-01 -3.1396e-01 5.5078e-01 8.4167e-02 -9.4971e-02 -2.3163e-02 1.8454e-03 -4.3726e-01 5.8301e-01 -5.4590e-01 6.9141e-01 -1.4197e-01 -2.6880e-01 -1.5710e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 0.0000e+00 -2.7847e-02 3.8477e-01 -6.2793e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-1.0522e-01 -3.9380e-01 5.3320e-01 6.7383e-02 -1.0962e-01 9.3918e-03 -4.8248e-02 -4.1064e-01 5.8740e-01 -5.4346e-01 6.7139e-01 -1.4856e-01 -2.5830e-01 -1.7432e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 0.0000e+00 -3.6133e-02 3.7085e-01 -6.2842e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-8.8684e-02 -4.0674e-01 5.1855e-01 2.0615e-02 -1.1273e-01 5.6305e-03 7.7057e-04 -4.0381e-01 5.6348e-01 -5.4053e-01 6.6455e-01 -1.5088e-01 -2.4988e-01 -1.5259e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 0.0000e+00 -3.1342e-02 3.6133e-01 -6.2744e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-9.2773e-02 -4.1992e-01 5.2979e-01 5.8411e-02 -1.0449e-01 5.8136e-03 -3.1677e-02 -3.8525e-01 5.4150e-01 -5.4590e-01 6.3330e-01 -1.6174e-01 -2.2742e-01 -1.6125e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 0.0000e+00 -7.7393e-02 3.4863e-01 -6.2939e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-1.0577e-01 -3.0273e-01 5.5957e-01 4.7363e-02 -8.7585e-02 -3.3661e-02 -1.2321e-02 -3.6597e-01 5.0293e-01 -5.5273e-01 6.4111e-01 -1.6968e-01 -2.1057e-01 -1.8359e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 0.0000e+00 -5.0690e-02 3.6182e-01 -6.2744e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-6.7139e-02 -3.7207e-01 5.4150e-01 8.3466e-03 -1.0199e-01 -4.4518e-03 -4.9067e-04 -3.6157e-01 4.7290e-01 -5.4395e-01 6.0400e-01 -1.6992e-01 -1.8921e-01 -1.4502e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 0.0000e+00 -7.8735e-02 3.3301e-01 -6.2891e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-7.5867e-02 -3.8696e-01 5.3467e-01 2.9099e-02 -1.0675e-01 -2.9602e-03 -7.8125e-03 -3.3984e-01 4.5825e-01 -5.4150e-01 5.8545e-01 -1.7664e-01 -1.6443e-01 -1.5369e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 0.0000e+00 -9.3689e-02 3.3691e-01 -6.2988e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-8.8318e-02 -3.2715e-01 5.7129e-01 2.2629e-02 -8.3252e-02 -2.2324e-02 5.4932e-03 -3.1812e-01 4.7119e-01 -5.3076e-01 5.4932e-01 -1.7786e-01 -1.4758e-01 -1.4172e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 0.0000e+00 -1.1792e-01 3.2178e-01 -6.3135e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-3.5645e-02 -3.4131e-01 5.5713e-01 3.0975e-03 -9.2590e-02 -1.9791e-02 -2.4185e-02 -2.8345e-01 4.4263e-01 -5.3369e-01 5.3418e-01 -1.8250e-01 -1.2512e-01 -1.7468e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 0.0000e+00 -1.2018e-01 3.1470e-01 -6.3086e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-1.2280e-01 -3.3984e-01 5.6543e-01 8.4717e-02 -1.0022e-01 -8.1863e-03 -1.6373e-02 -2.9175e-01 4.1260e-01 -5.2393e-01 5.1367e-01 -1.8762e-01 -9.3567e-02 -1.4185e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 0.0000e+00 -1.3245e-01 3.3911e-01 -6.3330e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-7.7881e-02 -3.6401e-01 5.4248e-01 -1.3504e-02 -9.3933e-02 -2.2766e-02 2.5558e-02 -2.6758e-01 4.2676e-01 -5.2246e-01 4.7607e-01 -2.0068e-01 -7.7454e-02 -1.3147e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 0.0000e+00 -9.7290e-02 3.2983e-01 -6.3037e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-3.0350e-02 -4.3555e-01 5.1953e-01 -2.0859e-02 -1.1243e-01 1.5554e-03 -1.0132e-02 -2.7100e-01 3.7524e-01 -5.1562e-01 4.7949e-01 -1.9604e-01 -6.1676e-02 -1.0272e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 0.0000e+00 -7.1350e-02 3.3936e-01 -6.3037e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-1.0437e-01 -3.4302e-01 5.6348e-01 4.0253e-02 -9.5459e-02 -1.7822e-02 -3.6068e-03 -2.3059e-01 4.5532e-01 -5.0195e-01 4.2236e-01 -2.0679e-01 -4.4098e-02 -1.1401e-01 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 0.0000e+00 -1.1761e-01 3.3496e-01 -6.3281e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00] + [-8.4290e-02 -3.6694e-01 5.4883e-01 2.2888e-02 -1.0736e-01 -8.1558e-03 3.1052e-03 -2.1289e-01 4.0796e-01 -4.9780e-01 3.8208e-01 -2.0654e-01 -2.3209e-02 -9.6069e-02 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 -1.0000e+00 0.0000e+00 -1.2451e-01 3.2837e-01 -6.3330e-01 0.0000e+00 0.0000e+00 0.0000e+00 + 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00]] + - state: [[ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -1.4783e-01 3.3911e-01 1.8604e-01 -9.7070e-01 -2.5659e-01 7.0862e-02 7.0068e-02 9.8926e-01 9.4092e-01 9.8242e-01 2.3999e-01 9.6631e-01 9.9756e-01 9.9756e-01 -9.2383e-01 -2.2766e-01 + -5.8008e-01 -3.8501e-01 5.3711e-01 -4.6460e-01 -1.9226e-01 3.8281e-01 9.7363e-01 8.1445e-01 9.2285e-01 8.4375e-01 8.8574e-01 9.8145e-01 4.0588e-03 4.7150e-03 5.5008e-03 4.6387e-03 + 5.3978e-03 2.1805e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0229e-01 7.9834e-02 7.7637e-02 7.7637e-02 -2.1820e-03 9.8535e-01 9.9463e-01 9.9658e-01 + 9.9707e-01 9.9707e-01 1.0000e+00 -1.7078e-01 2.8030e-02 1.0735e-02 9.7513e-04 9.9951e-01 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -1.4124e-01 3.3765e-01 1.8884e-01 -9.7070e-01 -2.5684e-01 7.3608e-02 6.6467e-02 9.8975e-01 9.4141e-01 9.8193e-01 2.3938e-01 9.6631e-01 9.9707e-01 9.9756e-01 -9.1846e-01 -2.4585e-01 + -5.8154e-01 -3.9331e-01 5.2881e-01 -4.4678e-01 -1.9409e-01 3.9600e-01 9.6924e-01 8.1348e-01 9.1943e-01 8.4863e-01 8.9453e-01 9.8096e-01 4.1199e-03 4.7913e-03 5.5962e-03 4.7150e-03 + 6.0158e-03 2.2522e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.9951e-01 9.4360e-02 7.4280e-02 7.2632e-02 7.2266e-02 -2.4300e-03 9.8535e-01 9.9561e-01 9.9707e-01 + 9.9756e-01 9.9756e-01 1.0000e+00 -1.6992e-01 2.2766e-02 8.3542e-03 6.9904e-04 9.9951e-01 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -1.5015e-01 3.4204e-01 1.9434e-01 -9.6777e-01 -2.5562e-01 7.1167e-02 6.7383e-02 9.8877e-01 9.3945e-01 9.8096e-01 2.5098e-01 9.6680e-01 9.9756e-01 9.9756e-01 -9.1162e-01 -2.6050e-01 + -5.8496e-01 -3.9893e-01 5.1904e-01 -4.3481e-01 -2.0215e-01 4.1089e-01 9.6533e-01 8.1104e-01 9.1699e-01 8.5498e-01 9.0039e-01 9.7949e-01 4.1466e-03 4.8180e-03 5.6190e-03 4.7379e-03 + 5.2032e-03 2.2858e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.9951e-01 8.6304e-02 6.8237e-02 6.6895e-02 6.6406e-02 -1.7948e-03 9.8535e-01 9.9609e-01 9.9756e-01 + 9.9756e-01 9.9756e-01 1.0000e+00 -1.7029e-01 1.8509e-02 7.3013e-03 5.0163e-04 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -1.4417e-01 3.3716e-01 1.9507e-01 -9.6680e-01 -2.5879e-01 7.8552e-02 5.5237e-02 9.8975e-01 9.4141e-01 9.8096e-01 2.5610e-01 9.6582e-01 9.9707e-01 9.9854e-01 -9.0186e-01 -2.7075e-01 + -5.8691e-01 -4.0845e-01 5.0879e-01 -4.2261e-01 -2.1399e-01 4.3188e-01 9.6240e-01 8.0957e-01 9.1260e-01 8.6084e-01 9.0625e-01 9.7705e-01 4.0779e-03 4.7455e-03 5.5389e-03 4.6692e-03 + 6.1607e-03 2.1973e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 7.9163e-02 6.2988e-02 6.1951e-02 6.1340e-02 -1.5554e-03 9.8535e-01 9.9707e-01 9.9805e-01 + 9.9805e-01 9.9805e-01 1.0000e+00 -1.6992e-01 1.4656e-02 5.7640e-03 4.0913e-04 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -1.3428e-01 3.3154e-01 1.9238e-01 -9.6924e-01 -2.6270e-01 8.3679e-02 5.7892e-02 9.9072e-01 9.4336e-01 9.8145e-01 2.4646e-01 9.6484e-01 9.9658e-01 9.9854e-01 -8.9258e-01 -2.8271e-01 + -5.8838e-01 -4.1797e-01 5.0049e-01 -4.1064e-01 -2.1667e-01 4.5093e-01 9.5898e-01 8.0859e-01 9.0869e-01 8.6572e-01 9.1162e-01 9.7607e-01 4.1008e-03 4.7684e-03 5.5695e-03 4.6921e-03 + 6.2790e-03 2.2263e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.9951e-01 7.2815e-02 5.8319e-02 5.7587e-02 5.6854e-02 -1.9894e-03 9.8535e-01 9.9756e-01 9.9854e-01 + 9.9854e-01 9.9854e-01 1.0000e+00 -1.7078e-01 1.1978e-02 4.1809e-03 3.8671e-04 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -1.2732e-01 3.2593e-01 1.9250e-01 -9.6875e-01 -2.6318e-01 8.7708e-02 5.2246e-02 9.9170e-01 9.4531e-01 9.8145e-01 2.4890e-01 9.6484e-01 9.9609e-01 9.9854e-01 -8.8184e-01 -2.9517e-01 + -5.8984e-01 -4.3384e-01 4.9023e-01 -3.9453e-01 -2.2119e-01 4.7168e-01 9.5557e-01 8.0713e-01 9.0088e-01 8.7158e-01 9.1895e-01 9.7510e-01 4.0588e-03 4.7188e-03 5.5122e-03 4.6463e-03 + 5.9395e-03 2.2125e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.9951e-01 6.7566e-02 5.4688e-02 5.4352e-02 5.3375e-02 -2.2583e-03 9.8535e-01 9.9756e-01 9.9854e-01 + 9.9854e-01 9.9854e-01 1.0000e+00 -1.7078e-01 6.9351e-03 2.2087e-03 3.3855e-04 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -1.2573e-01 3.3374e-01 1.9885e-01 -9.6875e-01 -2.5806e-01 7.9163e-02 5.2338e-02 9.9219e-01 9.4287e-01 9.7998e-01 2.4841e-01 9.6631e-01 9.9707e-01 9.9854e-01 -8.6914e-01 -3.1128e-01 + -5.9326e-01 -4.4434e-01 4.7925e-01 -3.7646e-01 -2.3157e-01 4.9463e-01 9.5020e-01 8.0469e-01 8.9600e-01 8.7744e-01 9.2627e-01 9.7266e-01 4.2534e-03 4.9400e-03 5.7640e-03 4.8599e-03 + 5.6610e-03 2.3849e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.9951e-01 6.2134e-02 5.0629e-02 5.0476e-02 4.9438e-02 -2.0332e-03 9.8535e-01 9.9805e-01 9.9854e-01 + 9.9854e-01 9.9854e-01 1.0000e+00 -1.7004e-01 4.5891e-03 1.3065e-03 1.1849e-04 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -1.1401e-01 3.3325e-01 2.0044e-01 -9.7168e-01 -2.5830e-01 8.0444e-02 5.5847e-02 9.9365e-01 9.4287e-01 9.7949e-01 2.3669e-01 9.6582e-01 9.9658e-01 9.9854e-01 -8.5791e-01 -3.2812e-01 + -5.9424e-01 -4.6191e-01 4.7070e-01 -3.5596e-01 -2.2827e-01 5.1367e-01 9.4482e-01 8.0420e-01 8.8672e-01 8.8232e-01 9.3457e-01 9.7363e-01 4.1695e-03 4.8485e-03 5.6610e-03 4.7684e-03 + 6.1874e-03 2.2766e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.9951e-01 5.7709e-02 4.7485e-02 4.7607e-02 4.6417e-02 -2.3422e-03 9.8535e-01 9.9854e-01 9.9902e-01 + 9.9902e-01 9.9902e-01 1.0000e+00 -1.7102e-01 9.7036e-04 -7.6914e-04 2.9993e-04 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -1.0681e-01 3.3081e-01 2.0020e-01 -9.7266e-01 -2.6025e-01 8.2397e-02 5.6824e-02 9.9414e-01 9.4385e-01 9.7998e-01 2.3279e-01 9.6533e-01 9.9658e-01 9.9854e-01 -8.4424e-01 -3.4375e-01 + -5.9375e-01 -4.8120e-01 4.6167e-01 -3.3203e-01 -2.2778e-01 5.3564e-01 9.3896e-01 8.0469e-01 8.7646e-01 8.8721e-01 9.4336e-01 9.7363e-01 4.1046e-03 4.7684e-03 5.5656e-03 4.6921e-03 + 5.9052e-03 2.2247e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.9951e-01 5.3253e-02 4.4128e-02 4.4403e-02 4.3152e-02 -1.9684e-03 9.8535e-01 9.9854e-01 9.9902e-01 + 9.9902e-01 9.9902e-01 1.0000e+00 -1.7029e-01 -2.8992e-03 -2.4872e-03 1.5950e-04 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -1.0468e-01 3.3521e-01 2.0752e-01 -9.7363e-01 -2.5464e-01 7.8308e-02 6.0974e-02 9.9463e-01 9.4238e-01 9.7803e-01 2.2839e-01 9.6680e-01 9.9707e-01 9.9805e-01 -8.2861e-01 -3.5400e-01 + -5.9082e-01 -5.0537e-01 4.5410e-01 -3.0786e-01 -2.2375e-01 5.5957e-01 9.3506e-01 8.0664e-01 8.6279e-01 8.9111e-01 9.5166e-01 9.7461e-01 4.2152e-03 4.9019e-03 5.7220e-03 4.8256e-03 + 5.5962e-03 2.3468e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.9951e-01 4.9103e-02 4.0985e-02 4.1382e-02 4.0070e-02 -1.5259e-03 9.8535e-01 9.9902e-01 9.9902e-01 + 9.9902e-01 9.9902e-01 1.0000e+00 -1.7078e-01 -7.6599e-03 -4.5509e-03 -9.6619e-05 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -8.8989e-02 3.3740e-01 2.1057e-01 -9.7559e-01 -2.5269e-01 7.5562e-02 5.6702e-02 9.9609e-01 9.4141e-01 9.7754e-01 2.1912e-01 9.6729e-01 9.9707e-01 9.9854e-01 -8.0811e-01 -3.6621e-01 + -5.8936e-01 -5.2783e-01 4.4678e-01 -2.8174e-01 -2.3059e-01 5.8936e-01 9.3066e-01 8.0811e-01 8.4912e-01 8.9453e-01 9.5947e-01 9.7314e-01 4.1809e-03 4.8599e-03 5.6725e-03 4.7798e-03 + 6.2523e-03 2.2812e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.9951e-01 4.5776e-02 3.8605e-02 3.9185e-02 3.7781e-02 -1.9054e-03 9.8535e-01 9.9902e-01 9.9902e-01 + 9.9902e-01 9.9951e-01 1.0000e+00 -1.7139e-01 -1.1421e-02 -6.6757e-03 -2.7180e-04 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -9.9548e-02 3.3911e-01 2.1460e-01 -9.7217e-01 -2.5366e-01 7.6782e-02 5.4993e-02 9.9512e-01 9.4092e-01 9.7656e-01 2.3486e-01 9.6729e-01 9.9707e-01 9.9854e-01 -7.9297e-01 -3.8037e-01 + -5.8594e-01 -5.5078e-01 4.3921e-01 -2.5146e-01 -2.2644e-01 6.0938e-01 9.2480e-01 8.1055e-01 8.3447e-01 8.9844e-01 9.6777e-01 9.7412e-01 4.1008e-03 4.7607e-03 5.5504e-03 4.6806e-03 + 4.9362e-03 2.2202e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.9951e-01 4.2358e-02 3.6011e-02 3.6652e-02 3.5248e-02 -2.4109e-03 9.8535e-01 9.9902e-01 9.9951e-01 + 9.9951e-01 9.9951e-01 1.0000e+00 -1.7139e-01 -1.5282e-02 -6.7978e-03 -1.9157e-04 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -9.7107e-02 3.3789e-01 2.1301e-01 -9.7559e-01 -2.5293e-01 7.3730e-02 6.4758e-02 9.9512e-01 9.4092e-01 9.7705e-01 2.2058e-01 9.6729e-01 9.9707e-01 9.9805e-01 -7.7490e-01 -3.8940e-01 + -5.8252e-01 -5.7764e-01 4.2871e-01 -2.2217e-01 -2.1936e-01 6.3232e-01 9.2090e-01 8.1299e-01 8.1641e-01 9.0332e-01 9.7510e-01 9.7559e-01 4.1389e-03 4.8103e-03 5.6190e-03 4.7340e-03 + 6.0501e-03 2.2659e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.9951e-01 3.9062e-02 3.3417e-02 3.4149e-02 3.2776e-02 -1.9894e-03 9.8535e-01 9.9902e-01 9.9951e-01 + 9.9951e-01 9.9951e-01 1.0000e+00 -1.7102e-01 -1.5991e-02 -7.4425e-03 -3.7241e-04 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -8.1604e-02 3.2935e-01 2.0654e-01 -9.7852e-01 -2.5757e-01 7.8186e-02 6.3660e-02 9.9658e-01 9.4434e-01 9.7852e-01 2.0557e-01 9.6631e-01 9.9707e-01 9.9805e-01 -7.6123e-01 -4.0381e-01 + -5.7812e-01 -5.9766e-01 4.2212e-01 -1.9409e-01 -2.0483e-01 6.4893e-01 9.1504e-01 8.1592e-01 8.0176e-01 9.0674e-01 9.8096e-01 9.7900e-01 4.0779e-03 4.7455e-03 5.5428e-03 4.6654e-03 + 6.5804e-03 2.1835e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 3.6530e-02 3.1555e-02 3.2349e-02 3.0914e-02 -2.3651e-03 9.8535e-01 9.9951e-01 9.9951e-01 + 9.9951e-01 9.9951e-01 1.0000e+00 -1.7090e-01 -1.4664e-02 -7.4348e-03 -1.9431e-04 1.0000e+00 1.0000e+00 1.0000e+00] + [ 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 0.0000e+00 + -8.8562e-02 3.3203e-01 2.1057e-01 -9.7754e-01 -2.5659e-01 7.6477e-02 6.3843e-02 9.9609e-01 9.4336e-01 9.7754e-01 2.1130e-01 9.6631e-01 9.9707e-01 9.9805e-01 -7.3975e-01 -4.0332e-01 + -5.7129e-01 -6.2598e-01 4.1333e-01 -1.6638e-01 -1.9653e-01 6.7285e-01 9.1504e-01 8.2080e-01 7.7979e-01 9.1064e-01 9.8584e-01 9.8047e-01 4.1618e-03 4.8332e-03 5.6381e-03 4.7569e-03 + 5.1804e-03 2.2888e-02 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 1.0000e+00 9.9951e-01 3.3752e-02 2.9343e-02 3.0212e-02 2.8793e-02 -2.4490e-03 9.8535e-01 9.9951e-01 9.9951e-01 + 9.9951e-01 9.9951e-01 1.0000e+00 -1.7065e-01 -1.6708e-02 -7.8278e-03 -6.1703e-04 1.0000e+00 1.0000e+00 1.0000e+00]] + - image: [] + - lang: 'unlocked_waist: pick the can from the tray and place it in the cardboard box' + - dataset_id: 7 +04/19 [11:15:06] INFO | >> [*] Align stats: QwenLatent.py:155 + pred(mean=-0.0008,std=0.0312,avg_norm=1.0000) + gt(mean=0.0008,std=0.0312,avg_norm=1.0000) +04/19 [11:15:46] INFO | >> train_qwenlatent.py:487 + Step 10 | grad_norm_pre_clip=46.3061 | + grad_norm_pre_clip_avg=46.9229 | Metrics: + {'align_loss': 0.033043187111616135, + 'recon_loss': 0.0056870910339057446, + 'predict_loss': 0.773002564907074, + 'aux_loss_decay_weight': 0.9982, + 'grad_norm_pre_clip': 46.30614471435547, + 'data_time': 0.0011042219994124025, + 'model_time': 2.7703189949970692, + 'grad_norm_pre_clip_avg': 46.922925567626955, + 'learning_rate': 5.0000000000000004e-08, + 'epoch': 0.0} +04/19 [11:16:23] INFO | >> train_qwenlatent.py:487 + Step 20 | grad_norm_pre_clip=41.2710 | + grad_norm_pre_clip_avg=46.0285 | Metrics: + {'align_loss': 0.0329645574092865, + 'recon_loss': 0.002580876462161541, + 'predict_loss': 0.6362991333007812, + 'aux_loss_decay_weight': 0.9962, + 'grad_norm_pre_clip': 41.27096939086914, + 'data_time': 0.00637320801615715, 'model_time': + 4.3573981579975225, 'grad_norm_pre_clip_avg': + 46.02846565246582, 'learning_rate': + 1.0000000000000001e-07, 'epoch': 0.01} +04/19 [11:17:00] INFO | >> train_qwenlatent.py:487 + Step 30 | grad_norm_pre_clip=41.8395 | + grad_norm_pre_clip_avg=44.5103 | Metrics: + {'align_loss': 0.033984795212745667, + 'recon_loss': 0.009698117151856422, + 'predict_loss': 1.13978910446167, + 'aux_loss_decay_weight': 0.9942, + 'grad_norm_pre_clip': 41.839515686035156, + 'data_time': 0.0010443750070407987, + 'model_time': 3.5370276239991654, + 'grad_norm_pre_clip_avg': 44.510322952270506, + 'learning_rate': 1.5000000000000002e-07, + 'epoch': 0.01} +04/19 [11:17:37] INFO | >> train_qwenlatent.py:487 + Step 40 | grad_norm_pre_clip=40.7990 | + grad_norm_pre_clip_avg=44.2279 | Metrics: + {'align_loss': 0.03287568688392639, + 'recon_loss': 0.00950863678008318, + 'predict_loss': 0.8035992383956909, + 'aux_loss_decay_weight': 0.9922, + 'grad_norm_pre_clip': 40.799015045166016, + 'data_time': 0.0011881999962497503, + 'model_time': 3.5186016089865007, + 'grad_norm_pre_clip_avg': 44.22793312072754, + 'learning_rate': 2.0000000000000002e-07, + 'epoch': 0.01} +04/19 [11:18:16] INFO | >> train_qwenlatent.py:487 + Step 50 | grad_norm_pre_clip=34.8509 | + grad_norm_pre_clip_avg=40.5282 | Metrics: + {'align_loss': 0.0338648296892643, + 'recon_loss': 0.0018054383108392358, + 'predict_loss': 0.6713551878929138, + 'aux_loss_decay_weight': 0.9902, + 'grad_norm_pre_clip': 34.85085678100586, + 'mae_score': 0.25190759607263513, 'data_time': + 0.0011458609951660037, 'model_time': + 3.6398255359963514, 'grad_norm_pre_clip_avg': + 40.5282039642334, 'learning_rate': + 2.5000000000000004e-07, 'epoch': 0.01} +04/19 [11:18:46] INFO | >> train_qwenlatent.py:487 + Step 60 | grad_norm_pre_clip=10.6726 | + grad_norm_pre_clip_avg=26.2286 | Metrics: + {'align_loss': 0.0329212062060833, + 'recon_loss': 0.0036257985047996044, + 'predict_loss': 0.3958715796470642, + 'aux_loss_decay_weight': 0.9882, + 'grad_norm_pre_clip': 10.672567367553711, + 'data_time': 0.0011021120008081198, + 'model_time': 2.557974931987701, + 'grad_norm_pre_clip_avg': 26.22863130569458, + 'learning_rate': 3.0000000000000004e-07, + 'epoch': 0.02} +04/19 [11:19:12] INFO | >> train_qwenlatent.py:487 + Step 70 | grad_norm_pre_clip=6.8838 | + grad_norm_pre_clip_avg=9.4008 | Metrics: + {'align_loss': 0.03272468224167824, + 'recon_loss': 0.004693057853728533, + 'predict_loss': 0.4246787130832672, + 'aux_loss_decay_weight': 0.9862, + 'grad_norm_pre_clip': 6.883776664733887, + 'data_time': 0.0010533970198594034, + 'model_time': 2.153881278994959, + 'grad_norm_pre_clip_avg': 9.400763702392577, + 'learning_rate': 3.5000000000000004e-07, + 'epoch': 0.02} +04/19 [11:19:31] INFO | >> train_qwenlatent.py:487 + Step 80 | grad_norm_pre_clip=3.6933 | + grad_norm_pre_clip_avg=5.6201 | Metrics: + {'align_loss': 0.03247949108481407, + 'recon_loss': 0.004509281367063522, + 'predict_loss': 0.28996628522872925, + 'aux_loss_decay_weight': 0.9842, + 'grad_norm_pre_clip': 3.6932592391967773, + 'data_time': 0.002289272000780329, + 'model_time': 1.5080144139938056, + 'grad_norm_pre_clip_avg': 5.620058655738831, + 'learning_rate': 4.0000000000000003e-07, + 'epoch': 0.02} +04/19 [11:19:45] INFO | >> train_qwenlatent.py:487 + Step 90 | grad_norm_pre_clip=2.8560 | + grad_norm_pre_clip_avg=3.2705 | Metrics: + {'align_loss': 0.03289157897233963, + 'recon_loss': 0.009866724722087383, + 'predict_loss': 0.34539681673049927, + 'aux_loss_decay_weight': 0.9822, + 'grad_norm_pre_clip': 2.856045722961426, + 'data_time': 0.0005848519795108587, + 'model_time': 1.2409328180074226, + 'grad_norm_pre_clip_avg': 3.270500349998474, + 'learning_rate': 4.5e-07, 'epoch': 0.02} +04/19 [11:19:59] INFO | >> train_qwenlatent.py:487 + Step 100 | grad_norm_pre_clip=2.6195 | + grad_norm_pre_clip_avg=2.5388 | Metrics: + {'align_loss': 0.03333656117320061, + 'recon_loss': 0.013158777728676796, + 'predict_loss': 0.4448263943195343, + 'aux_loss_decay_weight': 0.9802, + 'grad_norm_pre_clip': 2.6195015907287598, + 'mae_score': 0.241601342553491, 'data_time': + 0.0010238440008834004, 'model_time': + 1.4953054189973045, 'grad_norm_pre_clip_avg': + 2.538757085800171, 'learning_rate': + 5.000000000000001e-07, 'epoch': 0.03} +04/19 [11:20:12] INFO | >> train_qwenlatent.py:487 + Step 110 | grad_norm_pre_clip=1.6941 | + grad_norm_pre_clip_avg=2.2460 | Metrics: + {'align_loss': 0.03324104845523834, + 'recon_loss': 0.006797323934733868, + 'predict_loss': 0.3214440643787384, + 'aux_loss_decay_weight': 0.9782, + 'grad_norm_pre_clip': 1.6940650939941406, + 'data_time': 0.000847593997605145, + 'model_time': 1.2116076110105496, + 'grad_norm_pre_clip_avg': 2.2460445404052733, + 'learning_rate': 5.5e-07, 'epoch': 0.03} +04/19 [11:20:25] INFO | >> train_qwenlatent.py:487 + Step 120 | grad_norm_pre_clip=1.4401 | + grad_norm_pre_clip_avg=1.7233 | Metrics: + {'align_loss': 0.03273427486419678, + 'recon_loss': 0.009573950432240963, + 'predict_loss': 0.2917395830154419, + 'aux_loss_decay_weight': 0.9762, + 'grad_norm_pre_clip': 1.4401108026504517, + 'data_time': 0.0009227170085068792, + 'model_time': 1.2608765919867437, + 'grad_norm_pre_clip_avg': 1.7232998132705688, + 'learning_rate': 6.000000000000001e-07, + 'epoch': 0.03} +04/19 [11:20:37] INFO | >> train_qwenlatent.py:487 + Step 130 | grad_norm_pre_clip=1.3059 | + grad_norm_pre_clip_avg=1.3571 | Metrics: + {'align_loss': 0.0328291580080986, + 'recon_loss': 0.003576180897653103, + 'predict_loss': 0.26323968172073364, + 'aux_loss_decay_weight': 0.9742, + 'grad_norm_pre_clip': 1.305940866470337, + 'data_time': 0.0007604059937875718, + 'model_time': 1.220196255017072, + 'grad_norm_pre_clip_avg': 1.3571264743804932, + 'learning_rate': 6.5e-07, 'epoch': 0.03} +04/19 [11:20:50] INFO | >> train_qwenlatent.py:487 + Step 140 | grad_norm_pre_clip=1.9365 | + grad_norm_pre_clip_avg=1.4595 | Metrics: + {'align_loss': 0.03314453363418579, + 'recon_loss': 0.006761894095689058, + 'predict_loss': 0.308667927980423, + 'aux_loss_decay_weight': 0.9722, + 'grad_norm_pre_clip': 1.9364622831344604, + 'data_time': 0.0009557289886288345, + 'model_time': 1.2305238010012545, + 'grad_norm_pre_clip_avg': 1.4594932556152345, + 'learning_rate': 7.000000000000001e-07, + 'epoch': 0.04} +04/19 [11:21:02] INFO | >> train_qwenlatent.py:487 + Step 150 | grad_norm_pre_clip=0.9412 | + grad_norm_pre_clip_avg=1.0935 | Metrics: + {'align_loss': 0.0327940359711647, + 'recon_loss': 0.005632054526358843, + 'predict_loss': 0.23366281390190125, + 'aux_loss_decay_weight': 0.9702, + 'grad_norm_pre_clip': 0.9412402510643005, + 'mae_score': 0.2414362452051661, 'data_time': + 0.0006207590049598366, 'model_time': + 1.2660178579972126, 'grad_norm_pre_clip_avg': + 1.0935325622558594, 'learning_rate': 7.5e-07, + 'epoch': 0.04} +04/19 [11:21:15] INFO | >> train_qwenlatent.py:487 + Step 160 | grad_norm_pre_clip=1.2244 | + grad_norm_pre_clip_avg=1.2877 | Metrics: + {'align_loss': 0.03318478912115097, + 'recon_loss': 0.005752468481659889, + 'predict_loss': 0.26050400733947754, + 'aux_loss_decay_weight': 0.9682, + 'grad_norm_pre_clip': 1.2244166135787964, + 'data_time': 0.0007909849809948355, + 'model_time': 1.222033650992671, + 'grad_norm_pre_clip_avg': 1.2877225816249847, + 'learning_rate': 8.000000000000001e-07, + 'epoch': 0.04} +04/19 [11:21:27] INFO | >> train_qwenlatent.py:487 + Step 170 | grad_norm_pre_clip=1.2589 | + grad_norm_pre_clip_avg=1.1322 | Metrics: + {'align_loss': 0.03258621320128441, + 'recon_loss': 0.01503228209912777, + 'predict_loss': 0.21884886920452118, + 'aux_loss_decay_weight': 0.9662, + 'grad_norm_pre_clip': 1.2588764429092407, + 'data_time': 0.000557416002266109, + 'model_time': 1.2530415560177062, + 'grad_norm_pre_clip_avg': 1.132222694158554, + 'learning_rate': 8.500000000000001e-07, + 'epoch': 0.04} +04/19 [11:21:40] INFO | >> train_qwenlatent.py:487 + Step 180 | grad_norm_pre_clip=1.7498 | + grad_norm_pre_clip_avg=1.3452 | Metrics: + {'align_loss': 0.03220225125551224, + 'recon_loss': 0.005500373430550098, + 'predict_loss': 0.20108440518379211, + 'aux_loss_decay_weight': 0.9642, + 'grad_norm_pre_clip': 1.7497721910476685, + 'data_time': 0.0008621099987067282, + 'model_time': 1.2450480189872906, + 'grad_norm_pre_clip_avg': 1.3451971769332887, + 'learning_rate': 9e-07, 'epoch': 0.05} +04/19 [11:21:52] INFO | >> train_qwenlatent.py:487 + Step 190 | grad_norm_pre_clip=1.3285 | + grad_norm_pre_clip_avg=1.1531 | Metrics: + {'align_loss': 0.03223308548331261, + 'recon_loss': 0.006867838557809591, + 'predict_loss': 0.17687775194644928, + 'aux_loss_decay_weight': 0.9621999999999999, + 'grad_norm_pre_clip': 1.3285318613052368, + 'data_time': 0.000732067011995241, + 'model_time': 1.2478500650031492, + 'grad_norm_pre_clip_avg': 1.1531089127063752, + 'learning_rate': 9.5e-07, 'epoch': 0.05} +04/19 [11:22:05] INFO | >> train_qwenlatent.py:487 + Step 200 | grad_norm_pre_clip=1.7125 | + grad_norm_pre_clip_avg=1.2599 | Metrics: + {'align_loss': 0.03294458985328674, + 'recon_loss': 0.014623535796999931, + 'predict_loss': 0.2008272111415863, + 'aux_loss_decay_weight': 0.9601999999999999, + 'grad_norm_pre_clip': 1.7125353813171387, + 'mae_score': 0.21139658335092906, 'data_time': + 0.0007899249903857708, 'model_time': + 1.248457083973335, 'grad_norm_pre_clip_avg': + 1.2599105954170227, 'learning_rate': + 1.0000000000000002e-06, 'epoch': 0.05} +04/19 [11:22:18] INFO | >> train_qwenlatent.py:487 + Step 210 | grad_norm_pre_clip=2.2857 | + grad_norm_pre_clip_avg=1.6423 | Metrics: + {'align_loss': 0.032566118985414505, + 'recon_loss': 0.010835913941264153, + 'predict_loss': 0.19097480177879333, + 'aux_loss_decay_weight': 0.9582, + 'grad_norm_pre_clip': 2.2857160568237305, + 'data_time': 0.0008985470049083233, + 'model_time': 1.2448984550137538, + 'grad_norm_pre_clip_avg': 1.6423498272895813, + 'learning_rate': 1.0500000000000001e-06, + 'epoch': 0.05} +04/19 [11:22:31] INFO | >> train_qwenlatent.py:487 + Step 220 | grad_norm_pre_clip=4.7403 | + grad_norm_pre_clip_avg=3.6466 | Metrics: + {'align_loss': 0.032456960529088974, + 'recon_loss': 0.008665869012475014, + 'predict_loss': 0.15337233245372772, + 'aux_loss_decay_weight': 0.9562, + 'grad_norm_pre_clip': 4.740342140197754, + 'data_time': 0.0008427109860349447, + 'model_time': 1.2179207539884374, + 'grad_norm_pre_clip_avg': 3.646596920490265, + 'learning_rate': 1.1e-06, 'epoch': 0.06} +04/19 [11:22:45] INFO | >> train_qwenlatent.py:487 + Step 230 | grad_norm_pre_clip=2.5249 | + grad_norm_pre_clip_avg=2.2774 | Metrics: + {'align_loss': 0.03270125389099121, + 'recon_loss': 0.011639493517577648, + 'predict_loss': 0.1733088344335556, + 'aux_loss_decay_weight': 0.9542, + 'grad_norm_pre_clip': 2.5249035358428955, + 'data_time': 0.0005459180101752281, + 'model_time': 1.41682113497518, + 'grad_norm_pre_clip_avg': 2.2773534774780275, + 'learning_rate': 1.15e-06, 'epoch': 0.06} +04/19 [11:22:58] INFO | >> train_qwenlatent.py:487 + Step 240 | grad_norm_pre_clip=1.8974 | + grad_norm_pre_clip_avg=2.3898 | Metrics: + {'align_loss': 0.03218056634068489, + 'recon_loss': 0.009869786910712719, + 'predict_loss': 0.14555726945400238, + 'aux_loss_decay_weight': 0.9522, + 'grad_norm_pre_clip': 1.8973909616470337, + 'data_time': 0.0008441429818049073, + 'model_time': 1.2436314499936998, + 'grad_norm_pre_clip_avg': 2.389814001321793, + 'learning_rate': 1.2000000000000002e-06, + 'epoch': 0.06} +04/19 [11:23:11] INFO | >> train_qwenlatent.py:487 + Step 250 | grad_norm_pre_clip=3.7355 | + grad_norm_pre_clip_avg=4.1795 | Metrics: + {'align_loss': 0.03201662749052048, + 'recon_loss': 0.009505952708423138, + 'predict_loss': 0.12905333936214447, + 'aux_loss_decay_weight': 0.9502, + 'grad_norm_pre_clip': 3.7354965209960938, + 'mae_score': 0.2115925969304265, 'data_time': + 0.0007901239732746035, 'model_time': + 1.2221964559867047, 'grad_norm_pre_clip_avg': + 4.179498648643493, 'learning_rate': 1.25e-06, + 'epoch': 0.06} +04/19 [11:23:23] INFO | >> train_qwenlatent.py:487 + Step 260 | grad_norm_pre_clip=2.3715 | + grad_norm_pre_clip_avg=2.6990 | Metrics: + {'align_loss': 0.03237782418727875, + 'recon_loss': 0.003703807480633259, + 'predict_loss': 0.1247686892747879, + 'aux_loss_decay_weight': 0.9482, + 'grad_norm_pre_clip': 2.3714559078216553, + 'data_time': 0.0006076559948269278, + 'model_time': 1.2518060090078507, + 'grad_norm_pre_clip_avg': 2.6990400552749634, + 'learning_rate': 1.3e-06, 'epoch': 0.07} +04/19 [11:23:35] INFO | >> train_qwenlatent.py:487 + Step 270 | grad_norm_pre_clip=2.3999 | + grad_norm_pre_clip_avg=3.2094 | Metrics: + {'align_loss': 0.03202666714787483, + 'recon_loss': 0.00996403954923153, + 'predict_loss': 0.12874384224414825, + 'aux_loss_decay_weight': 0.9462, + 'grad_norm_pre_clip': 2.399909019470215, + 'data_time': 0.0008824469987303019, + 'model_time': 1.2190695879980922, + 'grad_norm_pre_clip_avg': 3.2093688011169434, + 'learning_rate': 1.35e-06, 'epoch': 0.07} +04/19 [11:23:48] INFO | >> train_qwenlatent.py:487 + Step 280 | grad_norm_pre_clip=2.3379 | + grad_norm_pre_clip_avg=2.0145 | Metrics: + {'align_loss': 0.03235568851232529, + 'recon_loss': 0.00582005875185132, + 'predict_loss': 0.1298951506614685, + 'aux_loss_decay_weight': 0.9442, + 'grad_norm_pre_clip': 2.3378713130950928, + 'data_time': 0.0009905140032060444, + 'model_time': 1.240540361002786, + 'grad_norm_pre_clip_avg': 2.0144926786422728, + 'learning_rate': 1.4000000000000001e-06, + 'epoch': 0.07} +04/19 [11:24:00] INFO | >> train_qwenlatent.py:487 + Step 290 | grad_norm_pre_clip=2.1133 | + grad_norm_pre_clip_avg=2.6225 | Metrics: + {'align_loss': 0.03209444880485535, + 'recon_loss': 0.009271200746297836, + 'predict_loss': 0.12571023404598236, + 'aux_loss_decay_weight': 0.9422, + 'grad_norm_pre_clip': 2.1132733821868896, + 'data_time': 0.0006227960111573339, + 'model_time': 1.2550066339899786, + 'grad_norm_pre_clip_avg': 2.6224744081497193, + 'learning_rate': 1.45e-06, 'epoch': 0.07} +04/19 [11:24:13] INFO | >> train_qwenlatent.py:487 + Step 300 | grad_norm_pre_clip=2.2878 | + grad_norm_pre_clip_avg=2.0905 | Metrics: + {'align_loss': 0.03187393769621849, + 'recon_loss': 0.007191887591034174, + 'predict_loss': 0.10578173398971558, + 'aux_loss_decay_weight': 0.9402, + 'grad_norm_pre_clip': 2.287790536880493, + 'mae_score': 0.16611825753976633, 'data_time': + 0.0009861690050456673, 'model_time': + 1.2566945720172953, 'grad_norm_pre_clip_avg': + 2.090490198135376, 'learning_rate': 1.5e-06, + 'epoch': 0.08} +04/19 [11:24:26] INFO | >> train_qwenlatent.py:487 + Step 310 | grad_norm_pre_clip=1.4769 | + grad_norm_pre_clip_avg=3.0744 | Metrics: + {'align_loss': 0.03211770951747894, + 'recon_loss': 0.006778689566999674, + 'predict_loss': 0.10606639087200165, + 'aux_loss_decay_weight': 0.9382, + 'grad_norm_pre_clip': 1.4769383668899536, + 'data_time': 0.0006534869899041951, + 'model_time': 1.1996833379962482, + 'grad_norm_pre_clip_avg': 3.0743804454803465, + 'learning_rate': 1.55e-06, 'epoch': 0.08} +04/19 [11:24:38] INFO | >> train_qwenlatent.py:487 + Step 320 | grad_norm_pre_clip=1.8043 | + grad_norm_pre_clip_avg=2.6473 | Metrics: + {'align_loss': 0.03198610991239548, + 'recon_loss': 0.013600333593785763, + 'predict_loss': 0.11105816811323166, + 'aux_loss_decay_weight': 0.9362, + 'grad_norm_pre_clip': 1.8042702674865723, + 'data_time': 0.0007831469993107021, + 'model_time': 1.2147661659982987, + 'grad_norm_pre_clip_avg': 2.6473208904266357, + 'learning_rate': 1.6000000000000001e-06, + 'epoch': 0.08} +04/19 [11:24:51] INFO | >> train_qwenlatent.py:487 + Step 330 | grad_norm_pre_clip=2.9471 | + grad_norm_pre_clip_avg=3.1638 | Metrics: + {'align_loss': 0.03183742240071297, + 'recon_loss': 0.007004869170486927, + 'predict_loss': 0.11161371320486069, + 'aux_loss_decay_weight': 0.9342, + 'grad_norm_pre_clip': 2.9471380710601807, + 'data_time': 0.0007769909861963242, + 'model_time': 1.221046660997672, + 'grad_norm_pre_clip_avg': 3.1637637615203857, + 'learning_rate': 1.65e-06, 'epoch': 0.08} +04/19 [11:25:04] INFO | >> train_qwenlatent.py:487 + Step 340 | grad_norm_pre_clip=2.2672 | + grad_norm_pre_clip_avg=1.8577 | Metrics: + {'align_loss': 0.03171020746231079, + 'recon_loss': 0.008036196231842041, + 'predict_loss': 0.0933278426527977, + 'aux_loss_decay_weight': 0.9322, + 'grad_norm_pre_clip': 2.267158031463623, + 'data_time': 0.0007949950231704861, + 'model_time': 1.3364513690175954, + 'grad_norm_pre_clip_avg': 1.8577311992645265, + 'learning_rate': 1.7000000000000002e-06, + 'epoch': 0.09} +04/19 [11:25:17] INFO | >> train_qwenlatent.py:487 + Step 350 | grad_norm_pre_clip=2.0495 | + grad_norm_pre_clip_avg=2.4955 | Metrics: + {'align_loss': 0.0317012220621109, + 'recon_loss': 0.0038277541752904654, + 'predict_loss': 0.10201358795166016, + 'aux_loss_decay_weight': 0.9302, + 'grad_norm_pre_clip': 2.0495312213897705, + 'mae_score': 0.17278601844031533, 'data_time': + 0.000580880994675681, 'model_time': + 1.281059893022757, 'grad_norm_pre_clip_avg': + 2.4954785823822023, 'learning_rate': + 1.7500000000000002e-06, 'epoch': 0.09} +04/19 [11:25:30] INFO | >> train_qwenlatent.py:487 + Step 360 | grad_norm_pre_clip=2.1829 | + grad_norm_pre_clip_avg=1.8413 | Metrics: + {'align_loss': 0.031549982726573944, + 'recon_loss': 0.009604895487427711, + 'predict_loss': 0.10660180449485779, + 'aux_loss_decay_weight': 0.9282, + 'grad_norm_pre_clip': 2.1829421520233154, + 'data_time': 0.0008644680201541632, + 'model_time': 1.23052617101348, + 'grad_norm_pre_clip_avg': 1.8412961840629578, + 'learning_rate': 1.8e-06, 'epoch': 0.09} +04/19 [11:25:43] INFO | >> train_qwenlatent.py:487 + Step 370 | grad_norm_pre_clip=1.6400 | + grad_norm_pre_clip_avg=1.9549 | Metrics: + {'align_loss': 0.031748123466968536, + 'recon_loss': 0.012270632199943066, + 'predict_loss': 0.09538540244102478, + 'aux_loss_decay_weight': 0.9262, + 'grad_norm_pre_clip': 1.6400010585784912, + 'data_time': 0.0005883170233573765, + 'model_time': 1.493062154011568, + 'grad_norm_pre_clip_avg': 1.9549302101135253, + 'learning_rate': 1.85e-06, 'epoch': 0.09} +04/19 [11:25:56] INFO | >> train_qwenlatent.py:487 + Step 380 | grad_norm_pre_clip=1.3904 | + grad_norm_pre_clip_avg=1.9307 | Metrics: + {'align_loss': 0.0314330980181694, + 'recon_loss': 0.008261787705123425, + 'predict_loss': 0.08129792660474777, + 'aux_loss_decay_weight': 0.9242, + 'grad_norm_pre_clip': 1.390392541885376, + 'data_time': 0.000725861988030374, + 'model_time': 1.3293861440033652, + 'grad_norm_pre_clip_avg': 1.930694615840912, + 'learning_rate': 1.9e-06, 'epoch': 0.1} +04/19 [11:26:08] INFO | >> train_qwenlatent.py:487 + Step 390 | grad_norm_pre_clip=2.7420 | + grad_norm_pre_clip_avg=2.0041 | Metrics: + {'align_loss': 0.0312078557908535, + 'recon_loss': 0.008457979187369347, + 'predict_loss': 0.09560856968164444, + 'aux_loss_decay_weight': 0.9222, + 'grad_norm_pre_clip': 2.742044687271118, + 'data_time': 0.0007391780090983957, + 'model_time': 1.2233393499918748, + 'grad_norm_pre_clip_avg': 2.004096305370331, + 'learning_rate': 1.95e-06, 'epoch': 0.1} +04/19 [11:26:21] INFO | >> train_qwenlatent.py:487 + Step 400 | grad_norm_pre_clip=2.0594 | + grad_norm_pre_clip_avg=1.8344 | Metrics: + {'align_loss': 0.03166826069355011, + 'recon_loss': 0.005275609903037548, + 'predict_loss': 0.08672631531953812, + 'aux_loss_decay_weight': 0.9202, + 'grad_norm_pre_clip': 2.0594184398651123, + 'mae_score': 0.15332247072511965, 'data_time': + 0.0008446270076092333, 'model_time': + 1.2547237319813576, 'grad_norm_pre_clip_avg': + 1.834431493282318, 'learning_rate': + 2.0000000000000003e-06, 'epoch': 0.1} +04/19 [11:26:34] INFO | >> train_qwenlatent.py:487 + Step 410 | grad_norm_pre_clip=1.2218 | + grad_norm_pre_clip_avg=1.7576 | Metrics: + {'align_loss': 0.03143379092216492, + 'recon_loss': 0.007932377979159355, + 'predict_loss': 0.08406821638345718, + 'aux_loss_decay_weight': 0.9182, + 'grad_norm_pre_clip': 1.221806526184082, + 'data_time': 0.0009646899998188019, + 'model_time': 1.2433559779892676, + 'grad_norm_pre_clip_avg': 1.7575655221939086, + 'learning_rate': 2.0500000000000003e-06, + 'epoch': 0.1} +04/19 [11:26:46] INFO | >> train_qwenlatent.py:487 + Step 420 | grad_norm_pre_clip=1.6090 | + grad_norm_pre_clip_avg=2.1345 | Metrics: + {'align_loss': 0.03149663284420967, + 'recon_loss': 0.005237074103206396, + 'predict_loss': 0.08149807155132294, + 'aux_loss_decay_weight': 0.9162, + 'grad_norm_pre_clip': 1.6090279817581177, + 'data_time': 0.0006794950168114156, + 'model_time': 1.2188235799840186, + 'grad_norm_pre_clip_avg': 2.1345234751701354, + 'learning_rate': 2.1000000000000002e-06, + 'epoch': 0.11} +04/19 [11:26:59] INFO | >> train_qwenlatent.py:487 + Step 430 | grad_norm_pre_clip=1.6107 | + grad_norm_pre_clip_avg=2.5214 | Metrics: + {'align_loss': 0.03099178895354271, + 'recon_loss': 0.007654838263988495, + 'predict_loss': 0.06630305200815201, + 'aux_loss_decay_weight': 0.9142, + 'grad_norm_pre_clip': 1.6106748580932617, + 'data_time': 0.0009116159926634282, + 'model_time': 1.2310399750131182, + 'grad_norm_pre_clip_avg': 2.521422338485718, + 'learning_rate': 2.1499999999999997e-06, + 'epoch': 0.11} +04/19 [11:27:12] INFO | >> train_qwenlatent.py:487 + Step 440 | grad_norm_pre_clip=1.4405 | + grad_norm_pre_clip_avg=1.6202 | Metrics: + {'align_loss': 0.03137056529521942, + 'recon_loss': 0.004050631076097488, + 'predict_loss': 0.07225047796964645, + 'aux_loss_decay_weight': 0.9122, + 'grad_norm_pre_clip': 1.4404983520507812, + 'data_time': 0.0008507710008416325, + 'model_time': 1.2180520349938888, + 'grad_norm_pre_clip_avg': 1.6201868295669555, + 'learning_rate': 2.2e-06, 'epoch': 0.11} +04/19 [11:27:25] INFO | >> train_qwenlatent.py:487 + Step 450 | grad_norm_pre_clip=1.2917 | + grad_norm_pre_clip_avg=1.7956 | Metrics: + {'align_loss': 0.03190400451421738, + 'recon_loss': 0.010447005741298199, + 'predict_loss': 0.08937537670135498, + 'aux_loss_decay_weight': 0.9102, + 'grad_norm_pre_clip': 1.2916558980941772, + 'mae_score': 0.12213897705078125, 'data_time': + 0.0008224479970522225, 'model_time': + 1.2419985530141275, 'grad_norm_pre_clip_avg': + 1.795631718635559, 'learning_rate': 2.25e-06, + 'epoch': 0.11} +04/19 [11:27:37] INFO | >> train_qwenlatent.py:487 + Step 460 | grad_norm_pre_clip=2.2760 | + grad_norm_pre_clip_avg=1.6898 | Metrics: + {'align_loss': 0.031028814613819122, + 'recon_loss': 0.006162842269986868, + 'predict_loss': 0.05999661237001419, + 'aux_loss_decay_weight': 0.9082, + 'grad_norm_pre_clip': 2.2759785652160645, + 'data_time': 0.0008533519867341965, + 'model_time': 1.2637420640094206, + 'grad_norm_pre_clip_avg': 1.6897952198982238, + 'learning_rate': 2.3e-06, 'epoch': 0.12} +04/19 [11:27:50] INFO | >> train_qwenlatent.py:487 + Step 470 | grad_norm_pre_clip=1.4926 | + grad_norm_pre_clip_avg=1.6472 | Metrics: + {'align_loss': 0.03008296713232994, + 'recon_loss': 0.00483822263777256, + 'predict_loss': 0.071275494992733, + 'aux_loss_decay_weight': 0.9062, + 'grad_norm_pre_clip': 1.4925825595855713, + 'data_time': 0.0006069769733585417, + 'model_time': 1.2583288390014786, + 'grad_norm_pre_clip_avg': 1.6471780776977538, + 'learning_rate': 2.35e-06, 'epoch': 0.12} +04/19 [11:28:03] INFO | >> train_qwenlatent.py:487 + Step 480 | grad_norm_pre_clip=1.3778 | + grad_norm_pre_clip_avg=1.9130 | Metrics: + {'align_loss': 0.03135591372847557, + 'recon_loss': 0.009732787497341633, + 'predict_loss': 0.06643689423799515, + 'aux_loss_decay_weight': 0.9042, + 'grad_norm_pre_clip': 1.3777930736541748, + 'data_time': 0.0009113379928749055, + 'model_time': 1.305798101995606, + 'grad_norm_pre_clip_avg': 1.912951648235321, + 'learning_rate': 2.4000000000000003e-06, + 'epoch': 0.12} +04/19 [11:28:16] INFO | >> train_qwenlatent.py:487 + Step 490 | grad_norm_pre_clip=1.9924 | + grad_norm_pre_clip_avg=1.6780 | Metrics: + {'align_loss': 0.030397839844226837, + 'recon_loss': 0.007654209155589342, + 'predict_loss': 0.0737610012292862, + 'aux_loss_decay_weight': 0.9022, + 'grad_norm_pre_clip': 1.9924105405807495, + 'data_time': 0.000640853977529332, + 'model_time': 1.2271558700012974, + 'grad_norm_pre_clip_avg': 1.6779965162277222, + 'learning_rate': 2.4500000000000003e-06, + 'epoch': 0.12} +04/19 [11:28:30] INFO | >> train_qwenlatent.py:487 + Step 500 | grad_norm_pre_clip=1.2807 | + grad_norm_pre_clip_avg=1.5185 | Metrics: + {'align_loss': 0.031073277816176414, + 'recon_loss': 0.00580430356785655, + 'predict_loss': 0.06896740943193436, + 'aux_loss_decay_weight': 0.9002, + 'grad_norm_pre_clip': 1.2807432413101196, + 'mae_score': 0.09175264513170397, 'data_time': + 0.0009126320073846728, 'model_time': + 1.183012266003061, 'grad_norm_pre_clip_avg': + 1.5185447216033936, 'learning_rate': 2.5e-06, + 'epoch': 0.13} +04/19 [11:28:43] INFO | >> train_qwenlatent.py:487 + Step 510 | grad_norm_pre_clip=2.1818 | + grad_norm_pre_clip_avg=1.9054 | Metrics: + {'align_loss': 0.030269645154476166, + 'recon_loss': 0.005468857940286398, + 'predict_loss': 0.0639238953590393, + 'aux_loss_decay_weight': 0.8982, + 'grad_norm_pre_clip': 2.1818320751190186, + 'data_time': 0.000936113006900996, + 'model_time': 1.2770907210069709, + 'grad_norm_pre_clip_avg': 1.9054280042648315, + 'learning_rate': 2.55e-06, 'epoch': 0.13} +04/19 [11:28:55] INFO | >> train_qwenlatent.py:487 + Step 520 | grad_norm_pre_clip=1.4727 | + grad_norm_pre_clip_avg=1.5824 | Metrics: + {'align_loss': 0.030430540442466736, + 'recon_loss': 0.010667867958545685, + 'predict_loss': 0.0642302855849266, + 'aux_loss_decay_weight': 0.8962, + 'grad_norm_pre_clip': 1.4727158546447754, + 'data_time': 0.0006745160208083689, + 'model_time': 1.213841382006649, + 'grad_norm_pre_clip_avg': 1.5824484348297119, + 'learning_rate': 2.6e-06, 'epoch': 0.13} +04/19 [11:29:08] INFO | >> train_qwenlatent.py:487 + Step 530 | grad_norm_pre_clip=1.0730 | + grad_norm_pre_clip_avg=1.5774 | Metrics: + {'align_loss': 0.03046465292572975, + 'recon_loss': 0.0032870040740817785, + 'predict_loss': 0.06095238775014877, + 'aux_loss_decay_weight': 0.8942, + 'grad_norm_pre_clip': 1.0730170011520386, + 'data_time': 0.0008656050194986165, + 'model_time': 1.252953016984975, + 'grad_norm_pre_clip_avg': 1.5773896634578706, + 'learning_rate': 2.65e-06, 'epoch': 0.13} +04/19 [11:29:20] INFO | >> train_qwenlatent.py:487 + Step 540 | grad_norm_pre_clip=1.6407 | + grad_norm_pre_clip_avg=1.5257 | Metrics: + {'align_loss': 0.030876439064741135, + 'recon_loss': 0.008949622511863708, + 'predict_loss': 0.07582591474056244, + 'aux_loss_decay_weight': 0.8922, + 'grad_norm_pre_clip': 1.6407204866409302, + 'data_time': 0.0006855359824839979, + 'model_time': 1.2409669530170504, + 'grad_norm_pre_clip_avg': 1.5256929397583008, + 'learning_rate': 2.7e-06, 'epoch': 0.14} +04/19 [11:29:34] INFO | >> train_qwenlatent.py:487 + Step 550 | grad_norm_pre_clip=2.1958 | + grad_norm_pre_clip_avg=1.6113 | Metrics: + {'align_loss': 0.030331727117300034, + 'recon_loss': 0.006545560900121927, + 'predict_loss': 0.06662555038928986, + 'aux_loss_decay_weight': 0.8902, + 'grad_norm_pre_clip': 2.1957907676696777, + 'mae_score': 0.10758211977846988, 'data_time': + 0.0009342560078948736, 'model_time': + 1.3063495389942545, 'grad_norm_pre_clip_avg': + 1.6113394856452943, 'learning_rate': + 2.7500000000000004e-06, 'epoch': 0.14} +04/19 [11:29:47] INFO | >> train_qwenlatent.py:487 + Step 560 | grad_norm_pre_clip=2.8265 | + grad_norm_pre_clip_avg=2.1159 | Metrics: + {'align_loss': 0.029843030497431755, + 'recon_loss': 0.004819248802959919, + 'predict_loss': 0.05945954471826553, + 'aux_loss_decay_weight': 0.8882, + 'grad_norm_pre_clip': 2.8265249729156494, + 'data_time': 0.0006096230063121766, + 'model_time': 1.223627530009253, + 'grad_norm_pre_clip_avg': 2.1159165263175965, + 'learning_rate': 2.8000000000000003e-06, + 'epoch': 0.14} +04/19 [11:29:59] INFO | >> train_qwenlatent.py:487 + Step 570 | grad_norm_pre_clip=1.9158 | + grad_norm_pre_clip_avg=1.8488 | Metrics: + {'align_loss': 0.030568767338991165, + 'recon_loss': 0.006836706772446632, + 'predict_loss': 0.07328420132398605, + 'aux_loss_decay_weight': 0.8862, + 'grad_norm_pre_clip': 1.915818452835083, + 'data_time': 0.0007675439992453903, + 'model_time': 1.2357590470055584, + 'grad_norm_pre_clip_avg': 1.848779058456421, + 'learning_rate': 2.8500000000000002e-06, + 'epoch': 0.14} +04/19 [11:30:12] INFO | >> train_qwenlatent.py:487 + Step 580 | grad_norm_pre_clip=1.4440 | + grad_norm_pre_clip_avg=1.8703 | Metrics: + {'align_loss': 0.031087175011634827, + 'recon_loss': 0.009266085922718048, + 'predict_loss': 0.0765404924750328, + 'aux_loss_decay_weight': 0.8842, + 'grad_norm_pre_clip': 1.4439587593078613, + 'data_time': 0.0007606669969391078, + 'model_time': 1.236541426012991, + 'grad_norm_pre_clip_avg': 1.8703478932380677, + 'learning_rate': 2.9e-06, 'epoch': 0.15} +04/19 [11:30:24] INFO | >> train_qwenlatent.py:487 + Step 590 | grad_norm_pre_clip=2.0657 | + grad_norm_pre_clip_avg=1.4837 | Metrics: + {'align_loss': 0.029798660427331924, + 'recon_loss': 0.005180248524993658, + 'predict_loss': 0.04850494861602783, + 'aux_loss_decay_weight': 0.8822, + 'grad_norm_pre_clip': 2.0657427310943604, + 'data_time': 0.0006262180104386061, + 'model_time': 1.250494414998684, + 'grad_norm_pre_clip_avg': 1.483685064315796, + 'learning_rate': 2.95e-06, 'epoch': 0.15} +04/19 [11:30:37] INFO | >> train_qwenlatent.py:487 + Step 600 | grad_norm_pre_clip=1.1131 | + grad_norm_pre_clip_avg=1.4862 | Metrics: + {'align_loss': 0.030511096119880676, + 'recon_loss': 0.009290413931012154, + 'predict_loss': 0.0688231959939003, + 'aux_loss_decay_weight': 0.8802, + 'grad_norm_pre_clip': 1.1130955219268799, + 'mae_score': 0.10319057842632672, 'data_time': + 0.001097973989089951, 'model_time': + 1.2723083299933933, 'grad_norm_pre_clip_avg': + 1.4862418174743652, 'learning_rate': 3e-06, + 'epoch': 0.15} +04/19 [11:30:50] INFO | >> train_qwenlatent.py:487 + Step 610 | grad_norm_pre_clip=1.2604 | + grad_norm_pre_clip_avg=1.2033 | Metrics: + {'align_loss': 0.029660601168870926, + 'recon_loss': 0.008850712329149246, + 'predict_loss': 0.056828562170267105, + 'aux_loss_decay_weight': 0.8782, + 'grad_norm_pre_clip': 1.2603505849838257, + 'data_time': 0.0006555079889949411, + 'model_time': 1.2616505429905374, + 'grad_norm_pre_clip_avg': 1.2033049285411834, + 'learning_rate': 3.05e-06, 'epoch': 0.15} +04/19 [11:31:03] INFO | >> train_qwenlatent.py:487 + Step 620 | grad_norm_pre_clip=1.8696 | + grad_norm_pre_clip_avg=1.2719 | Metrics: + {'align_loss': 0.02964923158288002, + 'recon_loss': 0.007557467091828585, + 'predict_loss': 0.05564294755458832, + 'aux_loss_decay_weight': 0.8762, + 'grad_norm_pre_clip': 1.8695822954177856, + 'data_time': 0.0008543580188415945, + 'model_time': 1.2156593870022334, + 'grad_norm_pre_clip_avg': 1.2719285666942597, + 'learning_rate': 3.1e-06, 'epoch': 0.16} +04/19 [11:31:17] INFO | >> train_qwenlatent.py:487 + Step 630 | grad_norm_pre_clip=2.9154 | + grad_norm_pre_clip_avg=1.4431 | Metrics: + {'align_loss': 0.029724353924393654, + 'recon_loss': 0.008553233928978443, + 'predict_loss': 0.06186613813042641, + 'aux_loss_decay_weight': 0.8742, + 'grad_norm_pre_clip': 2.915375232696533, + 'data_time': 0.0010523450036998838, + 'model_time': 1.4895332310115919, + 'grad_norm_pre_clip_avg': 1.443060863018036, + 'learning_rate': 3.1500000000000003e-06, + 'epoch': 0.16} +04/19 [11:31:30] INFO | >> train_qwenlatent.py:487 + Step 640 | grad_norm_pre_clip=1.3577 | + grad_norm_pre_clip_avg=1.3736 | Metrics: + {'align_loss': 0.029901916161179543, + 'recon_loss': 0.007856573909521103, + 'predict_loss': 0.05317410081624985, + 'aux_loss_decay_weight': 0.8722, + 'grad_norm_pre_clip': 1.3577067852020264, + 'data_time': 0.000963989004958421, + 'model_time': 1.216572487988742, + 'grad_norm_pre_clip_avg': 1.3736284017562865, + 'learning_rate': 3.2000000000000003e-06, + 'epoch': 0.16} +04/19 [11:31:42] INFO | >> train_qwenlatent.py:487 + Step 650 | grad_norm_pre_clip=1.6607 | + grad_norm_pre_clip_avg=1.5964 | Metrics: + {'align_loss': 0.029721569269895554, + 'recon_loss': 0.007816825993359089, + 'predict_loss': 0.06802579015493393, + 'aux_loss_decay_weight': 0.8702, + 'grad_norm_pre_clip': 1.6606707572937012, + 'mae_score': 0.09279770722260346, 'data_time': + 0.0009976870205719024, 'model_time': + 1.2293424220115412, 'grad_norm_pre_clip_avg': + 1.596384847164154, 'learning_rate': + 3.2500000000000002e-06, 'epoch': 0.16} +04/19 [11:31:55] INFO | >> train_qwenlatent.py:487 + Step 660 | grad_norm_pre_clip=1.0481 | + grad_norm_pre_clip_avg=1.2566 | Metrics: + {'align_loss': 0.029359064996242523, + 'recon_loss': 0.011123884469270706, + 'predict_loss': 0.0622626394033432, + 'aux_loss_decay_weight': 0.8682, + 'grad_norm_pre_clip': 1.0481271743774414, + 'data_time': 0.0009324260172434151, + 'model_time': 1.2487496819812804, + 'grad_norm_pre_clip_avg': 1.2565792202949524, + 'learning_rate': 3.3e-06, 'epoch': 0.17} +04/19 [11:32:08] INFO | >> train_qwenlatent.py:487 + Step 670 | grad_norm_pre_clip=1.7640 | + grad_norm_pre_clip_avg=1.2799 | Metrics: + {'align_loss': 0.029938697814941406, + 'recon_loss': 0.011092126369476318, + 'predict_loss': 0.07270407676696777, + 'aux_loss_decay_weight': 0.8662, + 'grad_norm_pre_clip': 1.7639586925506592, + 'data_time': 0.0008506269950885326, + 'model_time': 1.2268686020106543, + 'grad_norm_pre_clip_avg': 1.2799117863178253, + 'learning_rate': 3.3500000000000005e-06, + 'epoch': 0.17} +04/19 [11:32:21] INFO | >> train_qwenlatent.py:487 + Step 680 | grad_norm_pre_clip=1.1723 | + grad_norm_pre_clip_avg=1.2351 | Metrics: + {'align_loss': 0.029162930324673653, + 'recon_loss': 0.005243512336164713, + 'predict_loss': 0.0499473437666893, + 'aux_loss_decay_weight': 0.8642, + 'grad_norm_pre_clip': 1.1723427772521973, + 'data_time': 0.0009292780014220625, + 'model_time': 1.2960661129909568, + 'grad_norm_pre_clip_avg': 1.2350585758686066, + 'learning_rate': 3.4000000000000005e-06, + 'epoch': 0.17} +04/19 [11:32:33] INFO | >> train_qwenlatent.py:487 + Step 690 | grad_norm_pre_clip=1.2319 | + grad_norm_pre_clip_avg=1.4655 | Metrics: + {'align_loss': 0.030402477830648422, + 'recon_loss': 0.004618501290678978, + 'predict_loss': 0.05358331650495529, + 'aux_loss_decay_weight': 0.8622, + 'grad_norm_pre_clip': 1.231905221939087, + 'data_time': 0.0010799230076372623, + 'model_time': 1.2544703820021823, + 'grad_norm_pre_clip_avg': 1.4655457139015198, + 'learning_rate': 3.4500000000000004e-06, + 'epoch': 0.17} +04/19 [11:32:47] INFO | >> train_qwenlatent.py:487 + Step 700 | grad_norm_pre_clip=1.5248 | + grad_norm_pre_clip_avg=1.2385 | Metrics: + {'align_loss': 0.030195359140634537, + 'recon_loss': 0.00789031945168972, + 'predict_loss': 0.06554421037435532, + 'aux_loss_decay_weight': 0.8602, + 'grad_norm_pre_clip': 1.5247509479522705, + 'mae_score': 0.11082285975550746, 'data_time': + 0.0009179920016322285, 'model_time': + 1.2638926020008512, 'grad_norm_pre_clip_avg': + 1.238453984260559, 'learning_rate': + 3.5000000000000004e-06, 'epoch': 0.18} +04/19 [11:32:59] INFO | >> train_qwenlatent.py:487 + Step 710 | grad_norm_pre_clip=1.0711 | + grad_norm_pre_clip_avg=1.2697 | Metrics: + {'align_loss': 0.028955208137631416, + 'recon_loss': 0.006377922371029854, + 'predict_loss': 0.05044308304786682, + 'aux_loss_decay_weight': 0.8582, + 'grad_norm_pre_clip': 1.0710549354553223, + 'data_time': 0.0008891829929780215, + 'model_time': 1.2637126960034948, + 'grad_norm_pre_clip_avg': 1.269739669561386, + 'learning_rate': 3.55e-06, 'epoch': 0.18} +04/19 [11:33:12] INFO | >> train_qwenlatent.py:487 + Step 720 | grad_norm_pre_clip=1.1554 | + grad_norm_pre_clip_avg=1.1498 | Metrics: + {'align_loss': 0.029300622642040253, + 'recon_loss': 0.005827533546835184, + 'predict_loss': 0.05148665979504585, + 'aux_loss_decay_weight': 0.8562, + 'grad_norm_pre_clip': 1.1553820371627808, + 'data_time': 0.0007329340151045471, + 'model_time': 1.199443015997531, + 'grad_norm_pre_clip_avg': 1.1497534096240998, + 'learning_rate': 3.6e-06, 'epoch': 0.18} +04/19 [11:33:24] INFO | >> train_qwenlatent.py:487 + Step 730 | grad_norm_pre_clip=0.9556 | + grad_norm_pre_clip_avg=1.0995 | Metrics: + {'align_loss': 0.029436614364385605, + 'recon_loss': 0.005886681843549013, + 'predict_loss': 0.05693943053483963, + 'aux_loss_decay_weight': 0.8542, + 'grad_norm_pre_clip': 0.9556046724319458, + 'data_time': 0.0007843140047043562, + 'model_time': 1.219053929002257, + 'grad_norm_pre_clip_avg': 1.0995427072048187, + 'learning_rate': 3.6499999999999998e-06, + 'epoch': 0.18} +04/19 [11:33:36] INFO | >> train_qwenlatent.py:487 + Step 740 | grad_norm_pre_clip=1.0945 | + grad_norm_pre_clip_avg=1.4224 | Metrics: + {'align_loss': 0.029065560549497604, + 'recon_loss': 0.006979391444474459, + 'predict_loss': 0.065666563808918, + 'aux_loss_decay_weight': 0.8522000000000001, + 'grad_norm_pre_clip': 1.0945035219192505, + 'data_time': 0.0007934590103104711, + 'model_time': 1.241339844011236, + 'grad_norm_pre_clip_avg': 1.4223700881004333, + 'learning_rate': 3.7e-06, 'epoch': 0.19} +04/19 [11:33:50] INFO | >> train_qwenlatent.py:487 + Step 750 | grad_norm_pre_clip=1.0349 | + grad_norm_pre_clip_avg=1.3043 | Metrics: + {'align_loss': 0.030071280896663666, + 'recon_loss': 0.009055884554982185, + 'predict_loss': 0.057866182178258896, + 'aux_loss_decay_weight': 0.8502000000000001, + 'grad_norm_pre_clip': 1.034925937652588, + 'mae_score': 0.08808619868647945, 'data_time': + 0.0005970759957563132, 'model_time': + 1.222529813006986, 'grad_norm_pre_clip_avg': + 1.3043266654014587, 'learning_rate': 3.75e-06, + 'epoch': 0.19} +04/19 [11:34:02] INFO | >> train_qwenlatent.py:487 + Step 760 | grad_norm_pre_clip=1.4094 | + grad_norm_pre_clip_avg=1.1234 | Metrics: + {'align_loss': 0.029117483645677567, + 'recon_loss': 0.006050576455891132, + 'predict_loss': 0.052793461829423904, + 'aux_loss_decay_weight': 0.8482000000000001, + 'grad_norm_pre_clip': 1.409367322921753, + 'data_time': 0.00072970797191374, 'model_time': + 1.2247028599958867, 'grad_norm_pre_clip_avg': + 1.123419189453125, 'learning_rate': 3.8e-06, + 'epoch': 0.19} +04/19 [11:34:16] INFO | >> train_qwenlatent.py:487 + Step 770 | grad_norm_pre_clip=1.3170 | + grad_norm_pre_clip_avg=1.0982 | Metrics: + {'align_loss': 0.02893589437007904, + 'recon_loss': 0.01214815303683281, + 'predict_loss': 0.057618722319602966, + 'aux_loss_decay_weight': 0.8462000000000001, + 'grad_norm_pre_clip': 1.3170270919799805, + 'data_time': 0.0006123359780758619, + 'model_time': 1.1805251160112675, + 'grad_norm_pre_clip_avg': 1.0981764554977418, + 'learning_rate': 3.85e-06, 'epoch': 0.19} +04/19 [11:34:29] INFO | >> train_qwenlatent.py:487 + Step 780 | grad_norm_pre_clip=1.1387 | + grad_norm_pre_clip_avg=1.3417 | Metrics: + {'align_loss': 0.028372716158628464, + 'recon_loss': 0.005152374971657991, + 'predict_loss': 0.05395901948213577, + 'aux_loss_decay_weight': 0.8442000000000001, + 'grad_norm_pre_clip': 1.1387056112289429, + 'data_time': 0.000898038997547701, + 'model_time': 1.2597812750027515, + 'grad_norm_pre_clip_avg': 1.3417288184165954, + 'learning_rate': 3.9e-06, 'epoch': 0.2} +04/19 [11:34:41] INFO | >> train_qwenlatent.py:487 + Step 790 | grad_norm_pre_clip=1.0439 | + grad_norm_pre_clip_avg=1.3711 | Metrics: + {'align_loss': 0.028457993641495705, + 'recon_loss': 0.011030524969100952, + 'predict_loss': 0.05846918001770973, + 'aux_loss_decay_weight': 0.8422000000000001, + 'grad_norm_pre_clip': 1.0438734292984009, + 'data_time': 0.000798380991909653, + 'model_time': 1.218764668010408, + 'grad_norm_pre_clip_avg': 1.3711039423942566, + 'learning_rate': 3.95e-06, 'epoch': 0.2} +04/19 [11:34:55] INFO | >> train_qwenlatent.py:487 + Step 800 | grad_norm_pre_clip=1.0419 | + grad_norm_pre_clip_avg=0.9929 | Metrics: + {'align_loss': 0.027249135076999664, + 'recon_loss': 0.004476209636777639, + 'predict_loss': 0.03423872962594032, + 'aux_loss_decay_weight': 0.8402000000000001, + 'grad_norm_pre_clip': 1.0419234037399292, + 'mae_score': 0.08941762425878026, 'data_time': + 0.0006606359966099262, 'model_time': + 1.3373448330094106, 'grad_norm_pre_clip_avg': + 0.9929198980331421, 'learning_rate': + 4.000000000000001e-06, 'epoch': 0.2} +04/19 [11:35:07] INFO | >> train_qwenlatent.py:487 + Step 810 | grad_norm_pre_clip=1.2193 | + grad_norm_pre_clip_avg=1.1432 | Metrics: + {'align_loss': 0.027636069804430008, + 'recon_loss': 0.007445165421813726, + 'predict_loss': 0.03811441361904144, + 'aux_loss_decay_weight': 0.8382000000000001, + 'grad_norm_pre_clip': 1.2193002700805664, + 'data_time': 0.0009355180081911385, + 'model_time': 1.333146618999308, + 'grad_norm_pre_clip_avg': 1.1431680977344514, + 'learning_rate': 4.05e-06, 'epoch': 0.2} +04/19 [11:35:20] INFO | >> train_qwenlatent.py:487 + Step 820 | grad_norm_pre_clip=0.9461 | + grad_norm_pre_clip_avg=1.0163 | Metrics: + {'align_loss': 0.0290273018181324, + 'recon_loss': 0.009327579289674759, + 'predict_loss': 0.06004534289240837, + 'aux_loss_decay_weight': 0.8362, + 'grad_norm_pre_clip': 0.9461065530776978, + 'data_time': 0.00136466501862742, 'model_time': + 1.330122882995056, 'grad_norm_pre_clip_avg': + 1.0162516295909882, 'learning_rate': + 4.1000000000000006e-06, 'epoch': 0.21} +04/19 [11:35:32] INFO | >> train_qwenlatent.py:487 + Step 830 | grad_norm_pre_clip=1.2328 | + grad_norm_pre_clip_avg=1.0938 | Metrics: + {'align_loss': 0.028742104768753052, + 'recon_loss': 0.0034664026461541653, + 'predict_loss': 0.0473502017557621, + 'aux_loss_decay_weight': 0.8342, + 'grad_norm_pre_clip': 1.2328442335128784, + 'data_time': 0.00100640399614349, 'model_time': + 1.2342643640004098, 'grad_norm_pre_clip_avg': + 1.0938283503055573, 'learning_rate': 4.15e-06, + 'epoch': 0.21} +04/19 [11:35:45] INFO | >> train_qwenlatent.py:487 + Step 840 | grad_norm_pre_clip=0.8244 | + grad_norm_pre_clip_avg=0.9903 | Metrics: + {'align_loss': 0.029046541079878807, + 'recon_loss': 0.005216796882450581, + 'predict_loss': 0.04484270140528679, + 'aux_loss_decay_weight': 0.8322, + 'grad_norm_pre_clip': 0.8244225382804871, + 'data_time': 0.0009879129938781261, + 'model_time': 1.225457633001497, + 'grad_norm_pre_clip_avg': 0.9903179347515106, + 'learning_rate': 4.2000000000000004e-06, + 'epoch': 0.21} +04/19 [11:35:58] INFO | >> train_qwenlatent.py:487 + Step 850 | grad_norm_pre_clip=1.1130 | + grad_norm_pre_clip_avg=0.9702 | Metrics: + {'align_loss': 0.02857598289847374, + 'recon_loss': 0.006536790169775486, + 'predict_loss': 0.04994351789355278, + 'aux_loss_decay_weight': 0.8302, + 'grad_norm_pre_clip': 1.1130470037460327, + 'mae_score': 0.08791134121181729, 'data_time': + 0.0007642009877599776, 'model_time': + 1.2280768790224101, 'grad_norm_pre_clip_avg': + 0.9702276229858399, 'learning_rate': + 4.250000000000001e-06, 'epoch': 0.21} +04/19 [11:36:10] INFO | >> train_qwenlatent.py:487 + Step 860 | grad_norm_pre_clip=0.9586 | + grad_norm_pre_clip_avg=1.1922 | Metrics: + {'align_loss': 0.028436778113245964, + 'recon_loss': 0.0066445437259972095, + 'predict_loss': 0.04460010305047035, + 'aux_loss_decay_weight': 0.8282, + 'grad_norm_pre_clip': 0.9586398601531982, + 'data_time': 0.0008359409985132515, + 'model_time': 1.2785869139770512, + 'grad_norm_pre_clip_avg': 1.1922283172607422, + 'learning_rate': 4.2999999999999995e-06, + 'epoch': 0.22} +04/19 [11:36:23] INFO | >> train_qwenlatent.py:487 + Step 870 | grad_norm_pre_clip=0.9682 | + grad_norm_pre_clip_avg=0.9812 | Metrics: + {'align_loss': 0.029427355155348778, + 'recon_loss': 0.008631008677184582, + 'predict_loss': 0.058114781975746155, + 'aux_loss_decay_weight': 0.8262, + 'grad_norm_pre_clip': 0.9682087302207947, + 'data_time': 0.0006884229951538146, + 'model_time': 1.262829460989451, + 'grad_norm_pre_clip_avg': 0.9812112927436829, + 'learning_rate': 4.35e-06, 'epoch': 0.22} +04/19 [11:36:36] INFO | >> train_qwenlatent.py:487 + Step 880 | grad_norm_pre_clip=0.8581 | + grad_norm_pre_clip_avg=0.9157 | Metrics: + {'align_loss': 0.028362389653921127, + 'recon_loss': 0.004059404134750366, + 'predict_loss': 0.03911256790161133, + 'aux_loss_decay_weight': 0.8242, + 'grad_norm_pre_clip': 0.8580911755561829, + 'data_time': 0.000652583985356614, + 'model_time': 1.24353320299997, + 'grad_norm_pre_clip_avg': 0.9156890869140625, + 'learning_rate': 4.4e-06, 'epoch': 0.22} +04/19 [11:36:48] INFO | >> train_qwenlatent.py:487 + Step 890 | grad_norm_pre_clip=0.9430 | + grad_norm_pre_clip_avg=1.0943 | Metrics: + {'align_loss': 0.02875632606446743, + 'recon_loss': 0.006944079417735338, + 'predict_loss': 0.04858710616827011, + 'aux_loss_decay_weight': 0.8222, + 'grad_norm_pre_clip': 0.9429527521133423, + 'data_time': 0.0009864690073300153, + 'model_time': 1.3093241470050998, + 'grad_norm_pre_clip_avg': 1.0942834138870239, + 'learning_rate': 4.45e-06, 'epoch': 0.22} +04/19 [11:37:02] INFO | >> train_qwenlatent.py:487 + Step 900 | grad_norm_pre_clip=0.6542 | + grad_norm_pre_clip_avg=1.0181 | Metrics: + {'align_loss': 0.028816627338528633, + 'recon_loss': 0.00529468385502696, + 'predict_loss': 0.04423939436674118, + 'aux_loss_decay_weight': 0.8202, + 'grad_norm_pre_clip': 0.654191255569458, + 'mae_score': 0.07017165175429335, 'data_time': + 0.0006236280023586005, 'model_time': + 1.2904385369911324, 'grad_norm_pre_clip_avg': + 1.0181120693683625, 'learning_rate': 4.5e-06, + 'epoch': 0.23} +04/19 [11:37:15] INFO | >> train_qwenlatent.py:487 + Step 910 | grad_norm_pre_clip=0.8651 | + grad_norm_pre_clip_avg=0.9757 | Metrics: + {'align_loss': 0.028958113864064217, + 'recon_loss': 0.003732035169377923, + 'predict_loss': 0.048068903386592865, + 'aux_loss_decay_weight': 0.8182, + 'grad_norm_pre_clip': 0.8651106357574463, + 'data_time': 0.0005853289912920445, + 'model_time': 1.2018840370001271, + 'grad_norm_pre_clip_avg': 0.9757031202316284, + 'learning_rate': 4.5500000000000005e-06, + 'epoch': 0.23} +04/19 [11:37:27] INFO | >> train_qwenlatent.py:487 + Step 920 | grad_norm_pre_clip=0.7150 | + grad_norm_pre_clip_avg=0.9144 | Metrics: + {'align_loss': 0.027346214279532433, + 'recon_loss': 0.005197572521865368, + 'predict_loss': 0.0405832938849926, + 'aux_loss_decay_weight': 0.8162, + 'grad_norm_pre_clip': 0.7150166034698486, + 'data_time': 0.0011142840085085481, + 'model_time': 1.2225878910103347, + 'grad_norm_pre_clip_avg': 0.9144287765026092, + 'learning_rate': 4.6e-06, 'epoch': 0.23} +04/19 [11:37:40] INFO | >> train_qwenlatent.py:487 + Step 930 | grad_norm_pre_clip=0.9952 | + grad_norm_pre_clip_avg=1.0176 | Metrics: + {'align_loss': 0.028222598135471344, + 'recon_loss': 0.008781692944467068, + 'predict_loss': 0.04877982288599014, + 'aux_loss_decay_weight': 0.8142, + 'grad_norm_pre_clip': 0.9951993823051453, + 'data_time': 0.00099691201467067, 'model_time': + 1.2200368769990746, 'grad_norm_pre_clip_avg': + 1.0175587356090545, 'learning_rate': 4.65e-06, + 'epoch': 0.23} +04/19 [11:37:53] INFO | >> train_qwenlatent.py:487 + Step 940 | grad_norm_pre_clip=1.2894 | + grad_norm_pre_clip_avg=0.9710 | Metrics: + {'align_loss': 0.027553997933864594, + 'recon_loss': 0.004823936149477959, + 'predict_loss': 0.04344720393419266, + 'aux_loss_decay_weight': 0.8122, + 'grad_norm_pre_clip': 1.289443016052246, + 'data_time': 0.0007047699764370918, + 'model_time': 1.2609463310218416, + 'grad_norm_pre_clip_avg': 0.9709812343120575, + 'learning_rate': 4.7e-06, 'epoch': 0.24} +04/19 [11:38:06] INFO | >> train_qwenlatent.py:487 + Step 950 | grad_norm_pre_clip=1.1969 | + grad_norm_pre_clip_avg=0.9655 | Metrics: + {'align_loss': 0.028426459059119225, + 'recon_loss': 0.007664345670491457, + 'predict_loss': 0.04998307302594185, + 'aux_loss_decay_weight': 0.8102, + 'grad_norm_pre_clip': 1.1969343423843384, + 'mae_score': 0.08142219749656883, 'data_time': + 0.0009625079983379692, 'model_time': + 1.250625871005468, 'grad_norm_pre_clip_avg': + 0.9655048429965973, 'learning_rate': 4.75e-06, + 'epoch': 0.24} +04/19 [11:38:19] INFO | >> train_qwenlatent.py:487 + Step 960 | grad_norm_pre_clip=0.9278 | + grad_norm_pre_clip_avg=1.0776 | Metrics: + {'align_loss': 0.02726920321583748, + 'recon_loss': 0.008298717439174652, + 'predict_loss': 0.04177292063832283, + 'aux_loss_decay_weight': 0.8082, + 'grad_norm_pre_clip': 0.927781879901886, + 'data_time': 0.0007929579878691584, + 'model_time': 1.2287079860107042, + 'grad_norm_pre_clip_avg': 1.077577829360962, + 'learning_rate': 4.800000000000001e-06, + 'epoch': 0.24} +04/19 [11:38:31] INFO | >> train_qwenlatent.py:487 + Step 970 | grad_norm_pre_clip=1.1024 | + grad_norm_pre_clip_avg=1.0996 | Metrics: + {'align_loss': 0.028660181909799576, + 'recon_loss': 0.00573084969073534, + 'predict_loss': 0.054021697491407394, + 'aux_loss_decay_weight': 0.8062, + 'grad_norm_pre_clip': 1.1023720502853394, + 'data_time': 0.0011373599991202354, + 'model_time': 1.2157451130042318, + 'grad_norm_pre_clip_avg': 1.0996023893356324, + 'learning_rate': 4.85e-06, 'epoch': 0.24} +04/19 [11:38:44] INFO | >> train_qwenlatent.py:487 + Step 980 | grad_norm_pre_clip=0.8395 | + grad_norm_pre_clip_avg=0.9829 | Metrics: + {'align_loss': 0.027414750307798386, + 'recon_loss': 0.007198993116617203, + 'predict_loss': 0.04449138417840004, + 'aux_loss_decay_weight': 0.8042, + 'grad_norm_pre_clip': 0.8395146131515503, + 'data_time': 0.0011031210015062243, + 'model_time': 1.2596742789901327, + 'grad_norm_pre_clip_avg': 0.9828776359558106, + 'learning_rate': 4.9000000000000005e-06, + 'epoch': 0.25} +04/19 [11:38:56] INFO | >> train_qwenlatent.py:487 + Step 990 | grad_norm_pre_clip=0.7466 | + grad_norm_pre_clip_avg=0.8023 | Metrics: + {'align_loss': 0.026330668479204178, + 'recon_loss': 0.004689927212893963, + 'predict_loss': 0.03424719348549843, + 'aux_loss_decay_weight': 0.8022, + 'grad_norm_pre_clip': 0.7465569376945496, + 'data_time': 0.0006695320189464837, + 'model_time': 1.284836915001506, + 'grad_norm_pre_clip_avg': 0.8023116648197174, + 'learning_rate': 4.950000000000001e-06, + 'epoch': 0.25} +04/19 [11:39:09] INFO | >> train_qwenlatent.py:487 + Step 1000 | grad_norm_pre_clip=0.6957 | + grad_norm_pre_clip_avg=0.9541 | Metrics: + {'align_loss': 0.0284448079764843, + 'recon_loss': 0.005884072743356228, + 'predict_loss': 0.04218306764960289, + 'aux_loss_decay_weight': 0.8002, + 'grad_norm_pre_clip': 0.6957305073738098, + 'mae_score': 0.08418372214377463, 'data_time': + 0.0008813259773887694, 'model_time': + 1.2854072029876988, 'grad_norm_pre_clip_avg': + 0.9540885090827942, 'learning_rate': 5e-06, + 'epoch': 0.25} +04/19 [11:39:22] INFO | >> train_qwenlatent.py:487 + Step 1010 | grad_norm_pre_clip=0.7424 | + grad_norm_pre_clip_avg=0.8370 | Metrics: + {'align_loss': 0.02794242650270462, + 'recon_loss': 0.00590726500377059, + 'predict_loss': 0.037370383739471436, + 'aux_loss_decay_weight': 0.7982, + 'grad_norm_pre_clip': 0.742416501045227, + 'data_time': 0.0007220930128823966, + 'model_time': 1.2190130079980008, + 'grad_norm_pre_clip_avg': 0.8370238304138183, + 'learning_rate': 5.050000000000001e-06, + 'epoch': 0.25} +04/19 [11:39:35] INFO | >> train_qwenlatent.py:487 + Step 1020 | grad_norm_pre_clip=0.7697 | + grad_norm_pre_clip_avg=0.7955 | Metrics: + {'align_loss': 0.02783414162695408, + 'recon_loss': 0.0060017304494977, + 'predict_loss': 0.03819065913558006, + 'aux_loss_decay_weight': 0.7962, + 'grad_norm_pre_clip': 0.7696955800056458, + 'data_time': 0.0007498170016333461, + 'model_time': 1.260676351026632, + 'grad_norm_pre_clip_avg': 0.7954503774642945, + 'learning_rate': 5.1e-06, 'epoch': 0.26} +04/19 [11:39:47] INFO | >> train_qwenlatent.py:487 + Step 1030 | grad_norm_pre_clip=0.8960 | + grad_norm_pre_clip_avg=0.8568 | Metrics: + {'align_loss': 0.02729877457022667, + 'recon_loss': 0.005183200817555189, + 'predict_loss': 0.03775741904973984, + 'aux_loss_decay_weight': 0.7942, + 'grad_norm_pre_clip': 0.8959673643112183, + 'data_time': 0.0006024230096954852, + 'model_time': 1.2438763619866222, + 'grad_norm_pre_clip_avg': 0.8567584216594696, + 'learning_rate': 5.15e-06, 'epoch': 0.26} +04/19 [11:40:00] INFO | >> train_qwenlatent.py:487 + Step 1040 | grad_norm_pre_clip=0.7845 | + grad_norm_pre_clip_avg=1.0714 | Metrics: + {'align_loss': 0.027380971238017082, + 'recon_loss': 0.011815148405730724, + 'predict_loss': 0.04955723509192467, + 'aux_loss_decay_weight': 0.7922, + 'grad_norm_pre_clip': 0.7844714522361755, + 'data_time': 0.000999371986836195, + 'model_time': 1.2434842280054, + 'grad_norm_pre_clip_avg': 1.0714300513267516, + 'learning_rate': 5.2e-06, 'epoch': 0.26} +04/19 [11:40:14] INFO | >> train_qwenlatent.py:487 + Step 1050 | grad_norm_pre_clip=0.9588 | + grad_norm_pre_clip_avg=0.9896 | Metrics: + {'align_loss': 0.027498481795191765, + 'recon_loss': 0.002578235464170575, + 'predict_loss': 0.03062845580279827, + 'aux_loss_decay_weight': 0.7902, + 'grad_norm_pre_clip': 0.958845317363739, + 'mae_score': 0.061038434827649916, 'data_time': + 0.0008667249931022525, 'model_time': + 1.2643943729926832, 'grad_norm_pre_clip_avg': + 0.9895877599716186, 'learning_rate': 5.25e-06, + 'epoch': 0.26} +04/19 [11:40:27] INFO | >> train_qwenlatent.py:487 + Step 1060 | grad_norm_pre_clip=0.8789 | + grad_norm_pre_clip_avg=0.8566 | Metrics: + {'align_loss': 0.027438344433903694, + 'recon_loss': 0.007321258075535297, + 'predict_loss': 0.039722371846437454, + 'aux_loss_decay_weight': 0.7882, + 'grad_norm_pre_clip': 0.8788809776306152, + 'data_time': 0.000660058023640886, + 'model_time': 1.2722817799949553, + 'grad_norm_pre_clip_avg': 0.8565788030624389, + 'learning_rate': 5.3e-06, 'epoch': 0.27} +04/19 [11:40:39] INFO | >> train_qwenlatent.py:487 + Step 1070 | grad_norm_pre_clip=1.1183 | + grad_norm_pre_clip_avg=0.9282 | Metrics: + {'align_loss': 0.027914540842175484, + 'recon_loss': 0.011505477130413055, + 'predict_loss': 0.05118849501013756, + 'aux_loss_decay_weight': 0.7862, + 'grad_norm_pre_clip': 1.118322491645813, + 'data_time': 0.0007386129873339087, + 'model_time': 1.2043373649939895, + 'grad_norm_pre_clip_avg': 0.9282366335391998, + 'learning_rate': 5.3500000000000004e-06, + 'epoch': 0.27} +04/19 [11:40:52] INFO | >> train_qwenlatent.py:487 + Step 1080 | grad_norm_pre_clip=0.9120 | + grad_norm_pre_clip_avg=0.8404 | Metrics: + {'align_loss': 0.027435192838311195, + 'recon_loss': 0.008573153987526894, + 'predict_loss': 0.054417215287685394, + 'aux_loss_decay_weight': 0.7842, + 'grad_norm_pre_clip': 0.9119945764541626, + 'data_time': 0.0008221210155170411, + 'model_time': 1.2954941400093958, + 'grad_norm_pre_clip_avg': 0.840440183877945, + 'learning_rate': 5.4e-06, 'epoch': 0.27} +04/19 [11:41:04] INFO | >> train_qwenlatent.py:487 + Step 1090 | grad_norm_pre_clip=0.9801 | + grad_norm_pre_clip_avg=0.7718 | Metrics: + {'align_loss': 0.027740884572267532, + 'recon_loss': 0.0039266603998839855, + 'predict_loss': 0.03830154240131378, + 'aux_loss_decay_weight': 0.7822, + 'grad_norm_pre_clip': 0.980112612247467, + 'data_time': 0.0006070630042813718, + 'model_time': 1.2811458699870855, + 'grad_norm_pre_clip_avg': 0.7717788219451904, + 'learning_rate': 5.45e-06, 'epoch': 0.28} +04/19 [11:41:18] INFO | >> train_qwenlatent.py:487 + Step 1100 | grad_norm_pre_clip=0.6288 | + grad_norm_pre_clip_avg=0.7433 | Metrics: + {'align_loss': 0.027277454733848572, + 'recon_loss': 0.002487116027623415, + 'predict_loss': 0.03706936538219452, + 'aux_loss_decay_weight': 0.7802, + 'grad_norm_pre_clip': 0.6288326382637024, + 'mae_score': 0.07619015805356137, 'data_time': + 0.0006168959953356534, 'model_time': + 1.2167512960149907, 'grad_norm_pre_clip_avg': + 0.7433081269264221, 'learning_rate': + 5.500000000000001e-06, 'epoch': 0.28} +04/19 [11:41:30] INFO | >> train_qwenlatent.py:487 + Step 1110 | grad_norm_pre_clip=0.5843 | + grad_norm_pre_clip_avg=0.7243 | Metrics: + {'align_loss': 0.027479713782668114, + 'recon_loss': 0.010809092782437801, + 'predict_loss': 0.04467035457491875, + 'aux_loss_decay_weight': 0.7782, + 'grad_norm_pre_clip': 0.5843376517295837, + 'data_time': 0.0009146399970632046, + 'model_time': 1.2748752539919224, + 'grad_norm_pre_clip_avg': 0.7243350744247437, + 'learning_rate': 5.55e-06, 'epoch': 0.28} +04/19 [11:41:43] INFO | >> train_qwenlatent.py:487 + Step 1120 | grad_norm_pre_clip=0.7519 | + grad_norm_pre_clip_avg=0.7656 | Metrics: + {'align_loss': 0.02686048299074173, + 'recon_loss': 0.007377235684543848, + 'predict_loss': 0.042636577039957047, + 'aux_loss_decay_weight': 0.7762, + 'grad_norm_pre_clip': 0.7518656849861145, + 'data_time': 0.000643912993837148, + 'model_time': 1.2248477179964539, + 'grad_norm_pre_clip_avg': 0.7656095921993256, + 'learning_rate': 5.600000000000001e-06, + 'epoch': 0.28} +04/19 [11:41:55] INFO | >> train_qwenlatent.py:487 + Step 1130 | grad_norm_pre_clip=0.9317 | + grad_norm_pre_clip_avg=0.8758 | Metrics: + {'align_loss': 0.027239559218287468, + 'recon_loss': 0.01321922056376934, + 'predict_loss': 0.046950701624155045, + 'aux_loss_decay_weight': 0.7742, + 'grad_norm_pre_clip': 0.9316721558570862, + 'data_time': 0.0006801050039939582, + 'model_time': 1.255097974993987, + 'grad_norm_pre_clip_avg': 0.8758301913738251, + 'learning_rate': 5.65e-06, 'epoch': 0.29} +04/19 [11:42:08] INFO | >> train_qwenlatent.py:487 + Step 1140 | grad_norm_pre_clip=0.8377 | + grad_norm_pre_clip_avg=0.7102 | Metrics: + {'align_loss': 0.028088832274079323, + 'recon_loss': 0.006463909521698952, + 'predict_loss': 0.05058858543634415, + 'aux_loss_decay_weight': 0.7722, + 'grad_norm_pre_clip': 0.8377131819725037, + 'data_time': 0.0007321340090129524, + 'model_time': 1.2488343740114942, + 'grad_norm_pre_clip_avg': 0.7102332353591919, + 'learning_rate': 5.7000000000000005e-06, + 'epoch': 0.29} +04/19 [11:42:21] INFO | >> train_qwenlatent.py:487 + Step 1150 | grad_norm_pre_clip=0.7236 | + grad_norm_pre_clip_avg=0.6634 | Metrics: + {'align_loss': 0.028482992202043533, + 'recon_loss': 0.00807003490626812, + 'predict_loss': 0.03934025391936302, + 'aux_loss_decay_weight': 0.7702, + 'grad_norm_pre_clip': 0.7236459255218506, + 'mae_score': 0.08191775760135135, 'data_time': + 0.0006643110245931894, 'model_time': + 1.1998996039910708, 'grad_norm_pre_clip_avg': + 0.6634486138820648, 'learning_rate': + 5.750000000000001e-06, 'epoch': 0.29} +04/19 [11:42:34] INFO | >> train_qwenlatent.py:487 + Step 1160 | grad_norm_pre_clip=0.8428 | + grad_norm_pre_clip_avg=0.7476 | Metrics: + {'align_loss': 0.027642427012324333, + 'recon_loss': 0.00579425785690546, + 'predict_loss': 0.03528502956032753, + 'aux_loss_decay_weight': 0.7682, + 'grad_norm_pre_clip': 0.8427724838256836, + 'data_time': 0.0007759669970255345, + 'model_time': 1.2118801139877178, + 'grad_norm_pre_clip_avg': 0.7475982666015625, + 'learning_rate': 5.8e-06, 'epoch': 0.29} +04/19 [11:42:47] INFO | >> train_qwenlatent.py:487 + Step 1170 | grad_norm_pre_clip=0.8206 | + grad_norm_pre_clip_avg=0.8174 | Metrics: + {'align_loss': 0.026161186397075653, + 'recon_loss': 0.006265672855079174, + 'predict_loss': 0.03349655866622925, + 'aux_loss_decay_weight': 0.7662, + 'grad_norm_pre_clip': 0.8205789923667908, + 'data_time': 0.0006234140018932521, + 'model_time': 1.6255410379962996, + 'grad_norm_pre_clip_avg': 0.8174344718456268, + 'learning_rate': 5.850000000000001e-06, + 'epoch': 0.3} +04/19 [11:43:00] INFO | >> train_qwenlatent.py:487 + Step 1180 | grad_norm_pre_clip=0.8798 | + grad_norm_pre_clip_avg=0.7985 | Metrics: + {'align_loss': 0.02578785829246044, + 'recon_loss': 0.003383883973583579, + 'predict_loss': 0.03378280997276306, + 'aux_loss_decay_weight': 0.7642, + 'grad_norm_pre_clip': 0.8797505497932434, + 'data_time': 0.0008892769983503968, + 'model_time': 1.2624650889774784, + 'grad_norm_pre_clip_avg': 0.7985433518886567, + 'learning_rate': 5.9e-06, 'epoch': 0.3} +04/19 [11:43:13] INFO | >> train_qwenlatent.py:487 + Step 1190 | grad_norm_pre_clip=0.6462 | + grad_norm_pre_clip_avg=0.8517 | Metrics: + {'align_loss': 0.02735527977347374, + 'recon_loss': 0.0019012950360774994, + 'predict_loss': 0.027452843263745308, + 'aux_loss_decay_weight': 0.7622, + 'grad_norm_pre_clip': 0.646195113658905, + 'data_time': 0.0006106529908720404, + 'model_time': 1.2468172189837787, + 'grad_norm_pre_clip_avg': 0.8516733348369598, + 'learning_rate': 5.95e-06, 'epoch': 0.3} +04/19 [11:43:26] INFO | >> train_qwenlatent.py:487 + Step 1200 | grad_norm_pre_clip=0.7035 | + grad_norm_pre_clip_avg=0.7130 | Metrics: + {'align_loss': 0.02796917036175728, + 'recon_loss': 0.00536185409873724, + 'predict_loss': 0.04449642449617386, + 'aux_loss_decay_weight': 0.7602, + 'grad_norm_pre_clip': 0.703493595123291, + 'mae_score': 0.057330322265625, 'data_time': + 0.0006256790074985474, 'model_time': + 1.2328474470123183, 'grad_norm_pre_clip_avg': + 0.7129821240901947, 'learning_rate': 6e-06, + 'epoch': 0.3} +04/19 [11:43:39] INFO | >> train_qwenlatent.py:487 + Step 1210 | grad_norm_pre_clip=0.8129 | + grad_norm_pre_clip_avg=0.7778 | Metrics: + {'align_loss': 0.027064383029937744, + 'recon_loss': 0.005429660435765982, + 'predict_loss': 0.037026453763246536, + 'aux_loss_decay_weight': 0.7582, + 'grad_norm_pre_clip': 0.8129239082336426, + 'data_time': 0.0009239150094799697, + 'model_time': 1.3446091519726906, + 'grad_norm_pre_clip_avg': 0.7777928531169891, + 'learning_rate': 6.0500000000000005e-06, + 'epoch': 0.31} +04/19 [11:43:52] INFO | >> train_qwenlatent.py:487 + Step 1220 | grad_norm_pre_clip=0.9985 | + grad_norm_pre_clip_avg=0.8505 | Metrics: + {'align_loss': 0.026803169399499893, + 'recon_loss': 0.0028189129661768675, + 'predict_loss': 0.031017949804663658, + 'aux_loss_decay_weight': 0.7562, + 'grad_norm_pre_clip': 0.9984574913978577, + 'data_time': 0.0006487749924417585, + 'model_time': 1.2426204290241003, + 'grad_norm_pre_clip_avg': 0.8505084753036499, + 'learning_rate': 6.1e-06, 'epoch': 0.31} +04/19 [11:44:04] INFO | >> train_qwenlatent.py:487 + Step 1230 | grad_norm_pre_clip=0.5950 | + grad_norm_pre_clip_avg=0.7754 | Metrics: + {'align_loss': 0.02594757080078125, + 'recon_loss': 0.006550583057105541, + 'predict_loss': 0.038797203451395035, + 'aux_loss_decay_weight': 0.7542, + 'grad_norm_pre_clip': 0.5950385928153992, + 'data_time': 0.0009810129995457828, + 'model_time': 1.261882324994076, + 'grad_norm_pre_clip_avg': 0.775417959690094, + 'learning_rate': 6.15e-06, 'epoch': 0.31} +04/19 [11:44:17] INFO | >> train_qwenlatent.py:487 + Step 1240 | grad_norm_pre_clip=0.9064 | + grad_norm_pre_clip_avg=0.8933 | Metrics: + {'align_loss': 0.02672707289457321, + 'recon_loss': 0.002797100692987442, + 'predict_loss': 0.033554237335920334, + 'aux_loss_decay_weight': 0.7522, + 'grad_norm_pre_clip': 0.9063644409179688, + 'data_time': 0.0007133149774745107, + 'model_time': 1.2497952620033175, + 'grad_norm_pre_clip_avg': 0.8933198511600494, + 'learning_rate': 6.2e-06, 'epoch': 0.31} +04/19 [11:44:30] INFO | >> train_qwenlatent.py:487 + Step 1250 | grad_norm_pre_clip=0.7462 | + grad_norm_pre_clip_avg=0.8073 | Metrics: + {'align_loss': 0.025492053478956223, + 'recon_loss': 0.007767311297357082, + 'predict_loss': 0.038138534873723984, + 'aux_loss_decay_weight': 0.7502, + 'grad_norm_pre_clip': 0.746159553527832, + 'mae_score': 0.06784793063326999, 'data_time': + 0.0012767440057359636, 'model_time': + 1.283604131982429, 'grad_norm_pre_clip_avg': + 0.8073410809040069, 'learning_rate': 6.25e-06, + 'epoch': 0.32} +04/19 [11:44:42] INFO | >> train_qwenlatent.py:487 + Step 1260 | grad_norm_pre_clip=0.7757 | + grad_norm_pre_clip_avg=0.8912 | Metrics: + {'align_loss': 0.026261810213327408, + 'recon_loss': 0.005387565586715937, + 'predict_loss': 0.03538733720779419, + 'aux_loss_decay_weight': 0.7482, + 'grad_norm_pre_clip': 0.7756854295730591, + 'data_time': 0.0008726690139155835, + 'model_time': 1.183726935996674, + 'grad_norm_pre_clip_avg': 0.8912165880203247, + 'learning_rate': 6.300000000000001e-06, + 'epoch': 0.32} +04/19 [11:44:55] INFO | >> train_qwenlatent.py:487 + Step 1270 | grad_norm_pre_clip=0.6886 | + grad_norm_pre_clip_avg=0.7639 | Metrics: + {'align_loss': 0.02641705796122551, + 'recon_loss': 0.003235983895137906, + 'predict_loss': 0.03156602010130882, + 'aux_loss_decay_weight': 0.7462, + 'grad_norm_pre_clip': 0.6885652542114258, + 'data_time': 0.0006212470179889351, + 'model_time': 1.2058729529962875, + 'grad_norm_pre_clip_avg': 0.7639080047607422, + 'learning_rate': 6.35e-06, 'epoch': 0.32} +04/19 [11:45:07] INFO | >> train_qwenlatent.py:487 + Step 1280 | grad_norm_pre_clip=0.5656 | + grad_norm_pre_clip_avg=0.6914 | Metrics: + {'align_loss': 0.026441851630806923, + 'recon_loss': 0.007075122557580471, + 'predict_loss': 0.03756805881857872, + 'aux_loss_decay_weight': 0.7442, + 'grad_norm_pre_clip': 0.5656009316444397, + 'data_time': 0.0007001799822319299, + 'model_time': 1.2714580269821454, + 'grad_norm_pre_clip_avg': 0.6913737297058106, + 'learning_rate': 6.4000000000000006e-06, + 'epoch': 0.32} +04/19 [11:45:20] INFO | >> train_qwenlatent.py:487 + Step 1290 | grad_norm_pre_clip=0.6195 | + grad_norm_pre_clip_avg=0.7077 | Metrics: + {'align_loss': 0.02557908371090889, + 'recon_loss': 0.006240172777324915, + 'predict_loss': 0.03497763350605965, + 'aux_loss_decay_weight': 0.7422, + 'grad_norm_pre_clip': 0.6195111870765686, + 'data_time': 0.0006276960193645209, + 'model_time': 1.2366257759858854, + 'grad_norm_pre_clip_avg': 0.7076912224292755, + 'learning_rate': 6.45e-06, 'epoch': 0.33} +04/19 [11:45:33] INFO | >> train_qwenlatent.py:487 + Step 1300 | grad_norm_pre_clip=0.7739 | + grad_norm_pre_clip_avg=0.7515 | Metrics: + {'align_loss': 0.027098070830106735, + 'recon_loss': 0.006678394041955471, + 'predict_loss': 0.0364396870136261, + 'aux_loss_decay_weight': 0.7402, + 'grad_norm_pre_clip': 0.7739198803901672, + 'mae_score': 0.05681054742486627, 'data_time': + 0.001261947996681556, 'model_time': + 1.2599268979975022, 'grad_norm_pre_clip_avg': + 0.7514996886253357, 'learning_rate': + 6.5000000000000004e-06, 'epoch': 0.33} +04/19 [11:45:47] INFO | >> train_qwenlatent.py:487 + Step 1310 | grad_norm_pre_clip=0.6137 | + grad_norm_pre_clip_avg=0.6978 | Metrics: + {'align_loss': 0.026719365268945694, + 'recon_loss': 0.008213653229176998, + 'predict_loss': 0.050514835864305496, + 'aux_loss_decay_weight': 0.7382, + 'grad_norm_pre_clip': 0.6136911511421204, + 'data_time': 0.0008823180105537176, + 'model_time': 1.2045319770113565, + 'grad_norm_pre_clip_avg': 0.6978343397378921, + 'learning_rate': 6.550000000000001e-06, + 'epoch': 0.33} +04/19 [11:46:00] INFO | >> train_qwenlatent.py:487 + Step 1320 | grad_norm_pre_clip=0.6008 | + grad_norm_pre_clip_avg=0.6345 | Metrics: + {'align_loss': 0.0270053930580616, + 'recon_loss': 0.007321381010115147, + 'predict_loss': 0.03945033252239227, + 'aux_loss_decay_weight': 0.7362, + 'grad_norm_pre_clip': 0.6007579565048218, + 'data_time': 0.0007657170062884688, + 'model_time': 1.2149309189990163, + 'grad_norm_pre_clip_avg': 0.634539395570755, + 'learning_rate': 6.6e-06, 'epoch': 0.33} +04/19 [11:46:12] INFO | >> train_qwenlatent.py:487 + Step 1330 | grad_norm_pre_clip=0.7195 | + grad_norm_pre_clip_avg=0.8188 | Metrics: + {'align_loss': 0.027256332337856293, + 'recon_loss': 0.005172677803784609, + 'predict_loss': 0.04514798894524574, + 'aux_loss_decay_weight': 0.7342, + 'grad_norm_pre_clip': 0.7195332050323486, + 'data_time': 0.000602482003159821, + 'model_time': 1.2209207389969379, + 'grad_norm_pre_clip_avg': 0.8187514424324036, + 'learning_rate': 6.650000000000001e-06, + 'epoch': 0.34} +04/19 [11:46:25] INFO | >> train_qwenlatent.py:487 + Step 1340 | grad_norm_pre_clip=0.8068 | + grad_norm_pre_clip_avg=0.6651 | Metrics: + {'align_loss': 0.025498146191239357, + 'recon_loss': 0.008485864847898483, + 'predict_loss': 0.04057058319449425, + 'aux_loss_decay_weight': 0.7322, + 'grad_norm_pre_clip': 0.8067599534988403, + 'data_time': 0.0009539150050841272, + 'model_time': 1.26796573199681, + 'grad_norm_pre_clip_avg': 0.66512451171875, + 'learning_rate': 6.700000000000001e-06, + 'epoch': 0.34} +04/19 [11:46:38] INFO | >> train_qwenlatent.py:487 + Step 1350 | grad_norm_pre_clip=0.6769 | + grad_norm_pre_clip_avg=0.6712 | Metrics: + {'align_loss': 0.02615772932767868, + 'recon_loss': 0.005168435629457235, + 'predict_loss': 0.029578300192952156, + 'aux_loss_decay_weight': 0.7302, + 'grad_norm_pre_clip': 0.6768791079521179, + 'mae_score': 0.07890460727451083, 'data_time': + 0.0009071850217878819, 'model_time': + 1.2667724119964987, 'grad_norm_pre_clip_avg': + 0.6712049543857574, 'learning_rate': + 6.750000000000001e-06, 'epoch': 0.34} +04/19 [11:46:51] INFO | >> train_qwenlatent.py:487 + Step 1360 | grad_norm_pre_clip=0.6322 | + grad_norm_pre_clip_avg=0.7019 | Metrics: + {'align_loss': 0.026313655078411102, + 'recon_loss': 0.008176371455192566, + 'predict_loss': 0.03781461343169212, + 'aux_loss_decay_weight': 0.7282, + 'grad_norm_pre_clip': 0.6321868300437927, + 'data_time': 0.0006165759987197816, + 'model_time': 1.2078056980099063, + 'grad_norm_pre_clip_avg': 0.7018646597862244, + 'learning_rate': 6.800000000000001e-06, + 'epoch': 0.34} +04/19 [11:47:03] INFO | >> train_qwenlatent.py:487 + Step 1370 | grad_norm_pre_clip=0.7048 | + grad_norm_pre_clip_avg=0.6962 | Metrics: + {'align_loss': 0.02519809454679489, + 'recon_loss': 0.007870490662753582, + 'predict_loss': 0.042541276663541794, + 'aux_loss_decay_weight': 0.7262, + 'grad_norm_pre_clip': 0.7048290967941284, + 'data_time': 0.0007271739887073636, + 'model_time': 1.238719716988271, + 'grad_norm_pre_clip_avg': 0.6961760938167572, + 'learning_rate': 6.8500000000000005e-06, + 'epoch': 0.35} +04/19 [11:47:16] INFO | >> train_qwenlatent.py:487 + Step 1380 | grad_norm_pre_clip=0.8843 | + grad_norm_pre_clip_avg=0.6815 | Metrics: + {'align_loss': 0.024722279980778694, + 'recon_loss': 0.0036924006417393684, + 'predict_loss': 0.031569719314575195, + 'aux_loss_decay_weight': 0.7242, + 'grad_norm_pre_clip': 0.884271502494812, + 'data_time': 0.0006776899972464889, + 'model_time': 1.250401143974159, + 'grad_norm_pre_clip_avg': 0.6814850151538849, + 'learning_rate': 6.900000000000001e-06, + 'epoch': 0.35} +04/19 [11:47:28] INFO | >> train_qwenlatent.py:487 + Step 1390 | grad_norm_pre_clip=0.5652 | + grad_norm_pre_clip_avg=0.6484 | Metrics: + {'align_loss': 0.024882977828383446, + 'recon_loss': 0.003135685808956623, + 'predict_loss': 0.026698412373661995, + 'aux_loss_decay_weight': 0.7222, + 'grad_norm_pre_clip': 0.5652105212211609, + 'data_time': 0.0010406729998067021, + 'model_time': 1.243271890009055, + 'grad_norm_pre_clip_avg': 0.6483748853206635, + 'learning_rate': 6.950000000000001e-06, + 'epoch': 0.35} +04/19 [11:47:41] INFO | >> train_qwenlatent.py:487 + Step 1400 | grad_norm_pre_clip=0.7189 | + grad_norm_pre_clip_avg=0.6402 | Metrics: + {'align_loss': 0.027138907462358475, + 'recon_loss': 0.005722505040466785, + 'predict_loss': 0.03810521215200424, + 'aux_loss_decay_weight': 0.7202, + 'grad_norm_pre_clip': 0.7189131379127502, + 'mae_score': 0.08447999696473818, 'data_time': + 0.0008187650237232447, 'model_time': + 1.2283563489909284, 'grad_norm_pre_clip_avg': + 0.640237033367157, 'learning_rate': + 7.000000000000001e-06, 'epoch': 0.35} +04/19 [11:47:54] INFO | >> train_qwenlatent.py:487 + Step 1410 | grad_norm_pre_clip=0.7556 | + grad_norm_pre_clip_avg=0.7824 | Metrics: + {'align_loss': 0.026124611496925354, + 'recon_loss': 0.005496566649526358, + 'predict_loss': 0.03067726641893387, + 'aux_loss_decay_weight': 0.7182, + 'grad_norm_pre_clip': 0.7555712461471558, + 'data_time': 0.0006665699766017497, + 'model_time': 1.2249875280249398, + 'grad_norm_pre_clip_avg': 0.7823639214038849, + 'learning_rate': 7.049999999999999e-06, + 'epoch': 0.36} +04/19 [11:48:07] INFO | >> train_qwenlatent.py:487 + Step 1420 | grad_norm_pre_clip=0.6565 | + grad_norm_pre_clip_avg=0.6532 | Metrics: + {'align_loss': 0.02600262686610222, + 'recon_loss': 0.00826734583824873, + 'predict_loss': 0.03639916703104973, + 'aux_loss_decay_weight': 0.7162, + 'grad_norm_pre_clip': 0.6565471291542053, + 'data_time': 0.000904022017493844, + 'model_time': 1.2164627000165638, + 'grad_norm_pre_clip_avg': 0.6531860888004303, + 'learning_rate': 7.1e-06, 'epoch': 0.36} +04/19 [11:48:19] INFO | >> train_qwenlatent.py:487 + Step 1430 | grad_norm_pre_clip=0.8912 | + grad_norm_pre_clip_avg=0.6663 | Metrics: + {'align_loss': 0.026685137301683426, + 'recon_loss': 0.012889930978417397, + 'predict_loss': 0.054521068930625916, + 'aux_loss_decay_weight': 0.7142, + 'grad_norm_pre_clip': 0.8912320733070374, + 'data_time': 0.0006551990227308124, + 'model_time': 1.4462914160103537, + 'grad_norm_pre_clip_avg': 0.6663033843040467, + 'learning_rate': 7.15e-06, 'epoch': 0.36} +04/19 [11:48:32] INFO | >> train_qwenlatent.py:487 + Step 1440 | grad_norm_pre_clip=0.6582 | + grad_norm_pre_clip_avg=0.6485 | Metrics: + {'align_loss': 0.02481037750840187, + 'recon_loss': 0.0023108194582164288, + 'predict_loss': 0.028729991987347603, + 'aux_loss_decay_weight': 0.7121999999999999, + 'grad_norm_pre_clip': 0.6581548452377319, + 'data_time': 0.0007550740265287459, + 'model_time': 1.6478453439776786, + 'grad_norm_pre_clip_avg': 0.6484707355499267, + 'learning_rate': 7.2e-06, 'epoch': 0.36} +04/19 [11:48:45] INFO | >> train_qwenlatent.py:487 + Step 1450 | grad_norm_pre_clip=0.6997 | + grad_norm_pre_clip_avg=0.6372 | Metrics: + {'align_loss': 0.025992874056100845, + 'recon_loss': 0.011259375140070915, + 'predict_loss': 0.05528862774372101, + 'aux_loss_decay_weight': 0.7101999999999999, + 'grad_norm_pre_clip': 0.6996843814849854, + 'mae_score': 0.05849876747475014, 'data_time': + 0.0006314119964372367, 'model_time': + 1.251325308025116, 'grad_norm_pre_clip_avg': + 0.6372310400009156, 'learning_rate': 7.25e-06, + 'epoch': 0.37} +04/19 [11:48:58] INFO | >> train_qwenlatent.py:487 + Step 1460 | grad_norm_pre_clip=0.6417 | + grad_norm_pre_clip_avg=0.7921 | Metrics: + {'align_loss': 0.02438320592045784, + 'recon_loss': 0.006370958406478167, + 'predict_loss': 0.0319102443754673, + 'aux_loss_decay_weight': 0.7081999999999999, + 'grad_norm_pre_clip': 0.6416724920272827, + 'data_time': 0.0006924729968886822, + 'model_time': 1.231029197020689, + 'grad_norm_pre_clip_avg': 0.7921112656593323, + 'learning_rate': 7.2999999999999996e-06, + 'epoch': 0.37} +04/19 [11:49:11] INFO | >> train_qwenlatent.py:487 + Step 1470 | grad_norm_pre_clip=0.6025 | + grad_norm_pre_clip_avg=0.9099 | Metrics: + {'align_loss': 0.025527436286211014, + 'recon_loss': 0.009191097691655159, + 'predict_loss': 0.03715028613805771, + 'aux_loss_decay_weight': 0.7061999999999999, + 'grad_norm_pre_clip': 0.6024710536003113, + 'data_time': 0.0006677320052403957, + 'model_time': 1.2492608090105932, + 'grad_norm_pre_clip_avg': 0.909895944595337, + 'learning_rate': 7.35e-06, 'epoch': 0.37} +04/19 [11:49:23] INFO | >> train_qwenlatent.py:487 + Step 1480 | grad_norm_pre_clip=0.6833 | + grad_norm_pre_clip_avg=0.6914 | Metrics: + {'align_loss': 0.025605276226997375, + 'recon_loss': 0.009842077270150185, + 'predict_loss': 0.039758142083883286, + 'aux_loss_decay_weight': 0.7041999999999999, + 'grad_norm_pre_clip': 0.6832693815231323, + 'data_time': 0.0008867610013112426, + 'model_time': 1.2682076519995462, + 'grad_norm_pre_clip_avg': 0.6914113104343415, + 'learning_rate': 7.4e-06, 'epoch': 0.37} +04/19 [11:49:36] INFO | >> train_qwenlatent.py:487 + Step 1490 | grad_norm_pre_clip=0.6967 | + grad_norm_pre_clip_avg=0.6842 | Metrics: + {'align_loss': 0.025390837341547012, + 'recon_loss': 0.0031204496044665575, + 'predict_loss': 0.03652756288647652, + 'aux_loss_decay_weight': 0.7021999999999999, + 'grad_norm_pre_clip': 0.6966583132743835, + 'data_time': 0.0006653240125160664, + 'model_time': 1.3633172440167982, + 'grad_norm_pre_clip_avg': 0.6841795742511749, + 'learning_rate': 7.45e-06, 'epoch': 0.38} +04/19 [11:49:49] INFO | >> train_qwenlatent.py:487 + Step 1500 | grad_norm_pre_clip=0.6172 | + grad_norm_pre_clip_avg=0.6830 | Metrics: + {'align_loss': 0.023794913664460182, + 'recon_loss': 0.007204537279903889, + 'predict_loss': 0.03362452983856201, + 'aux_loss_decay_weight': 0.7001999999999999, + 'grad_norm_pre_clip': 0.6171987056732178, + 'mae_score': 0.05298244373218433, 'data_time': + 0.0006239899958018214, 'model_time': + 1.2627332389820367, 'grad_norm_pre_clip_avg': + 0.6829734086990357, 'learning_rate': 7.5e-06, + 'epoch': 0.38} +04/19 [11:50:02] INFO | >> train_qwenlatent.py:487 + Step 1510 | grad_norm_pre_clip=0.5787 | + grad_norm_pre_clip_avg=0.6367 | Metrics: + {'align_loss': 0.025315221399068832, + 'recon_loss': 0.005608766805380583, + 'predict_loss': 0.03446004167199135, + 'aux_loss_decay_weight': 0.6981999999999999, + 'grad_norm_pre_clip': 0.5787344574928284, + 'data_time': 0.0006756449874956161, + 'model_time': 1.276534018019447, + 'grad_norm_pre_clip_avg': 0.6367042183876037, + 'learning_rate': 7.55e-06, 'epoch': 0.38} +04/19 [11:50:14] INFO | >> train_qwenlatent.py:487 + Step 1520 | grad_norm_pre_clip=0.5905 | + grad_norm_pre_clip_avg=0.5920 | Metrics: + {'align_loss': 0.02574140392243862, + 'recon_loss': 0.010603385977447033, + 'predict_loss': 0.03336188569664955, + 'aux_loss_decay_weight': 0.6961999999999999, + 'grad_norm_pre_clip': 0.5905158519744873, + 'data_time': 0.0009118019952438772, + 'model_time': 1.2197986060054973, + 'grad_norm_pre_clip_avg': 0.592016088962555, + 'learning_rate': 7.6e-06, 'epoch': 0.38} +04/19 [11:50:27] INFO | >> train_qwenlatent.py:487 + Step 1530 | grad_norm_pre_clip=0.4752 | + grad_norm_pre_clip_avg=0.5768 | Metrics: + {'align_loss': 0.027525389567017555, + 'recon_loss': 0.009539121761918068, + 'predict_loss': 0.041065916419029236, + 'aux_loss_decay_weight': 0.6941999999999999, + 'grad_norm_pre_clip': 0.4752185642719269, + 'data_time': 0.0008858999935910106, + 'model_time': 1.2280858430021908, + 'grad_norm_pre_clip_avg': 0.5768082350492477, + 'learning_rate': 7.65e-06, 'epoch': 0.39} +04/19 [11:50:39] INFO | >> train_qwenlatent.py:487 + Step 1540 | grad_norm_pre_clip=0.5970 | + grad_norm_pre_clip_avg=0.5848 | Metrics: + {'align_loss': 0.024441279470920563, + 'recon_loss': 0.004704547580331564, + 'predict_loss': 0.03475659713149071, + 'aux_loss_decay_weight': 0.6921999999999999, + 'grad_norm_pre_clip': 0.5969778895378113, + 'data_time': 0.0009650400024838746, + 'model_time': 1.2718739690026268, + 'grad_norm_pre_clip_avg': 0.5848414659500122, + 'learning_rate': 7.7e-06, 'epoch': 0.39} +04/19 [11:50:53] INFO | >> train_qwenlatent.py:487 + Step 1550 | grad_norm_pre_clip=0.5948 | + grad_norm_pre_clip_avg=0.6223 | Metrics: + {'align_loss': 0.0249637383967638, + 'recon_loss': 0.012518001720309258, + 'predict_loss': 0.04653406888246536, + 'aux_loss_decay_weight': 0.6901999999999999, + 'grad_norm_pre_clip': 0.5948416590690613, + 'mae_score': 0.07359852833790823, 'data_time': + 0.0009424379968550056, 'model_time': + 1.2218686449923553, 'grad_norm_pre_clip_avg': + 0.6222967207431793, 'learning_rate': 7.75e-06, + 'epoch': 0.39} +04/19 [11:51:06] INFO | >> train_qwenlatent.py:487 + Step 1560 | grad_norm_pre_clip=0.5898 | + grad_norm_pre_clip_avg=0.6286 | Metrics: + {'align_loss': 0.024851316586136818, + 'recon_loss': 0.004001871682703495, + 'predict_loss': 0.031237250193953514, + 'aux_loss_decay_weight': 0.6881999999999999, + 'grad_norm_pre_clip': 0.5897898077964783, + 'data_time': 0.001098470005672425, + 'model_time': 1.3181342869938817, + 'grad_norm_pre_clip_avg': 0.6285952389240265, + 'learning_rate': 7.8e-06, 'epoch': 0.39} +04/19 [11:51:19] INFO | >> train_qwenlatent.py:487 + Step 1570 | grad_norm_pre_clip=0.5821 | + grad_norm_pre_clip_avg=0.5544 | Metrics: + {'align_loss': 0.0244092158973217, + 'recon_loss': 0.004308186937123537, + 'predict_loss': 0.02637210115790367, + 'aux_loss_decay_weight': 0.6861999999999999, + 'grad_norm_pre_clip': 0.5821243524551392, + 'data_time': 0.0007207859889604151, + 'model_time': 1.5486190550145693, + 'grad_norm_pre_clip_avg': 0.5543823212385177, + 'learning_rate': 7.850000000000001e-06, + 'epoch': 0.4} +04/19 [11:51:32] INFO | >> train_qwenlatent.py:487 + Step 1580 | grad_norm_pre_clip=0.5619 | + grad_norm_pre_clip_avg=0.5684 | Metrics: + {'align_loss': 0.02437136508524418, + 'recon_loss': 0.005279297940433025, + 'predict_loss': 0.028581544756889343, + 'aux_loss_decay_weight': 0.6841999999999999, + 'grad_norm_pre_clip': 0.5619061589241028, + 'data_time': 0.0008975480159278959, + 'model_time': 1.241057378007099, + 'grad_norm_pre_clip_avg': 0.5683602809906005, + 'learning_rate': 7.9e-06, 'epoch': 0.4} +04/19 [11:51:44] INFO | >> train_qwenlatent.py:487 + Step 1590 | grad_norm_pre_clip=0.7151 | + grad_norm_pre_clip_avg=0.6125 | Metrics: + {'align_loss': 0.026644812896847725, + 'recon_loss': 0.008089376613497734, + 'predict_loss': 0.045358311384916306, + 'aux_loss_decay_weight': 0.6821999999999999, + 'grad_norm_pre_clip': 0.7151234149932861, + 'data_time': 0.0006606359966099262, + 'model_time': 1.237570766999852, + 'grad_norm_pre_clip_avg': 0.6125175714492798, + 'learning_rate': 7.95e-06, 'epoch': 0.4} +04/19 [11:51:57] INFO | >> train_qwenlatent.py:487 + Step 1600 | grad_norm_pre_clip=0.6167 | + grad_norm_pre_clip_avg=0.6100 | Metrics: + {'align_loss': 0.024372722953557968, + 'recon_loss': 0.010909954085946083, + 'predict_loss': 0.042558807879686356, + 'aux_loss_decay_weight': 0.6802, + 'grad_norm_pre_clip': 0.6166906952857971, + 'mae_score': 0.0707144797385276, 'data_time': + 0.0006721689715050161, 'model_time': + 1.2124112400051672, 'grad_norm_pre_clip_avg': + 0.610032320022583, 'learning_rate': + 8.000000000000001e-06, 'epoch': 0.4} +04/19 [11:52:10] INFO | >> train_qwenlatent.py:487 + Step 1610 | grad_norm_pre_clip=0.6522 | + grad_norm_pre_clip_avg=0.6466 | Metrics: + {'align_loss': 0.026062380522489548, + 'recon_loss': 0.01099281944334507, + 'predict_loss': 0.035597462207078934, + 'aux_loss_decay_weight': 0.6782, + 'grad_norm_pre_clip': 0.6521974802017212, + 'data_time': 0.0006351270130835474, + 'model_time': 1.3006594559992664, + 'grad_norm_pre_clip_avg': 0.6465954899787902, + 'learning_rate': 8.050000000000001e-06, + 'epoch': 0.41} +04/19 [11:52:23] INFO | >> train_qwenlatent.py:487 + Step 1620 | grad_norm_pre_clip=0.6410 | + grad_norm_pre_clip_avg=0.7066 | Metrics: + {'align_loss': 0.02556258626282215, + 'recon_loss': 0.008268821984529495, + 'predict_loss': 0.038445401936769485, + 'aux_loss_decay_weight': 0.6762, + 'grad_norm_pre_clip': 0.6410294771194458, + 'data_time': 0.0009903429890982807, + 'model_time': 1.221041785000125, + 'grad_norm_pre_clip_avg': 0.706580400466919, + 'learning_rate': 8.1e-06, 'epoch': 0.41} +04/19 [11:52:35] INFO | >> train_qwenlatent.py:487 + Step 1630 | grad_norm_pre_clip=0.6302 | + grad_norm_pre_clip_avg=0.7033 | Metrics: + {'align_loss': 0.024780575186014175, + 'recon_loss': 0.0034137628972530365, + 'predict_loss': 0.027496179565787315, + 'aux_loss_decay_weight': 0.6742, + 'grad_norm_pre_clip': 0.6302099823951721, + 'data_time': 0.001273417001357302, + 'model_time': 1.316885348991491, + 'grad_norm_pre_clip_avg': 0.7033354222774506, + 'learning_rate': 8.15e-06, 'epoch': 0.41} +04/19 [11:52:48] INFO | >> train_qwenlatent.py:487 + Step 1640 | grad_norm_pre_clip=0.6656 | + grad_norm_pre_clip_avg=0.6226 | Metrics: + {'align_loss': 0.02640792541205883, + 'recon_loss': 0.010676062665879726, + 'predict_loss': 0.04276476427912712, + 'aux_loss_decay_weight': 0.6722, + 'grad_norm_pre_clip': 0.6655511260032654, + 'data_time': 0.0008914040226954967, + 'model_time': 1.2588901190028992, + 'grad_norm_pre_clip_avg': 0.6226471543312073, + 'learning_rate': 8.200000000000001e-06, + 'epoch': 0.41} +04/19 [11:53:01] INFO | >> train_qwenlatent.py:487 + Step 1650 | grad_norm_pre_clip=0.6406 | + grad_norm_pre_clip_avg=0.6966 | Metrics: + {'align_loss': 0.024957776069641113, + 'recon_loss': 0.0038303127512335777, + 'predict_loss': 0.03921523317694664, + 'aux_loss_decay_weight': 0.6702, + 'grad_norm_pre_clip': 0.6405988931655884, + 'mae_score': 0.059936873977248735, 'data_time': + 0.0006464889738708735, 'model_time': + 1.2488047390070278, 'grad_norm_pre_clip_avg': + 0.6966277182102203, 'learning_rate': 8.25e-06, + 'epoch': 0.42} +04/19 [11:53:14] INFO | >> train_qwenlatent.py:487 + Step 1660 | grad_norm_pre_clip=0.5357 | + grad_norm_pre_clip_avg=0.6343 | Metrics: + {'align_loss': 0.024866770952939987, + 'recon_loss': 0.007184587884694338, + 'predict_loss': 0.030916830524802208, + 'aux_loss_decay_weight': 0.6682, + 'grad_norm_pre_clip': 0.53566575050354, + 'data_time': 0.000712221983121708, + 'model_time': 1.2585545470064972, + 'grad_norm_pre_clip_avg': 0.6343243062496186, + 'learning_rate': 8.3e-06, 'epoch': 0.42} +04/19 [11:53:26] INFO | >> train_qwenlatent.py:487 + Step 1670 | grad_norm_pre_clip=0.5334 | + grad_norm_pre_clip_avg=0.5487 | Metrics: + {'align_loss': 0.025123409926891327, + 'recon_loss': 0.007427153177559376, + 'predict_loss': 0.037820298224687576, + 'aux_loss_decay_weight': 0.6662, + 'grad_norm_pre_clip': 0.5334182977676392, + 'data_time': 0.0006563110218849033, + 'model_time': 1.2417251490114722, + 'grad_norm_pre_clip_avg': 0.5487477004528045, + 'learning_rate': 8.350000000000001e-06, + 'epoch': 0.42} +04/19 [11:53:39] INFO | >> train_qwenlatent.py:487 + Step 1680 | grad_norm_pre_clip=0.8318 | + grad_norm_pre_clip_avg=0.6394 | Metrics: + {'align_loss': 0.025777000933885574, + 'recon_loss': 0.007494258228689432, + 'predict_loss': 0.03745191544294357, + 'aux_loss_decay_weight': 0.6642, + 'grad_norm_pre_clip': 0.8317572474479675, + 'data_time': 0.0006402819999493659, + 'model_time': 1.2063064510002732, + 'grad_norm_pre_clip_avg': 0.639355930685997, + 'learning_rate': 8.400000000000001e-06, + 'epoch': 0.42} +04/19 [11:53:51] INFO | >> train_qwenlatent.py:487 + Step 1690 | grad_norm_pre_clip=0.5538 | + grad_norm_pre_clip_avg=0.5905 | Metrics: + {'align_loss': 0.02464384399354458, + 'recon_loss': 0.0074709034524858, + 'predict_loss': 0.035226453095674515, + 'aux_loss_decay_weight': 0.6622, + 'grad_norm_pre_clip': 0.5538140535354614, + 'data_time': 0.0008825520053505898, + 'model_time': 1.203811927989591, + 'grad_norm_pre_clip_avg': 0.5904741883277893, + 'learning_rate': 8.45e-06, 'epoch': 0.43} +04/19 [11:54:05] INFO | >> train_qwenlatent.py:487 + Step 1700 | grad_norm_pre_clip=0.7132 | + grad_norm_pre_clip_avg=0.5751 | Metrics: + {'align_loss': 0.024103250354528427, + 'recon_loss': 0.002983381040394306, + 'predict_loss': 0.024990852922201157, + 'aux_loss_decay_weight': 0.6602, + 'grad_norm_pre_clip': 0.713182270526886, + 'mae_score': 0.05018497157741238, 'data_time': + 0.000642722996417433, 'model_time': + 1.21023372100899, 'grad_norm_pre_clip_avg': + 0.5750622093677521, 'learning_rate': + 8.500000000000002e-06, 'epoch': 0.43} +04/19 [11:54:19] INFO | >> train_qwenlatent.py:487 + Step 1710 | grad_norm_pre_clip=0.5566 | + grad_norm_pre_clip_avg=0.5384 | Metrics: + {'align_loss': 0.024393461644649506, + 'recon_loss': 0.0029619093984365463, + 'predict_loss': 0.026184435933828354, + 'aux_loss_decay_weight': 0.6582, + 'grad_norm_pre_clip': 0.5566421151161194, + 'data_time': 0.0009464479808229953, + 'model_time': 1.324464655976044, + 'grad_norm_pre_clip_avg': 0.5383658766746521, + 'learning_rate': 8.550000000000001e-06, + 'epoch': 0.43} +04/19 [11:54:32] INFO | >> train_qwenlatent.py:487 + Step 1720 | grad_norm_pre_clip=0.7667 | + grad_norm_pre_clip_avg=0.5843 | Metrics: + {'align_loss': 0.025186263024806976, + 'recon_loss': 0.004264958668500185, + 'predict_loss': 0.033747658133506775, + 'aux_loss_decay_weight': 0.6562, + 'grad_norm_pre_clip': 0.7667055130004883, + 'data_time': 0.0006919600127730519, + 'model_time': 1.2721665409917478, + 'grad_norm_pre_clip_avg': 0.5843280643224716, + 'learning_rate': 8.599999999999999e-06, + 'epoch': 0.43} +04/19 [11:54:45] INFO | >> train_qwenlatent.py:487 + Step 1730 | grad_norm_pre_clip=0.5244 | + grad_norm_pre_clip_avg=0.5518 | Metrics: + {'align_loss': 0.023415030911564827, + 'recon_loss': 0.002705800347030163, + 'predict_loss': 0.029000060632824898, + 'aux_loss_decay_weight': 0.6542, + 'grad_norm_pre_clip': 0.52443528175354, + 'data_time': 0.000943358987569809, + 'model_time': 1.3384454039915, + 'grad_norm_pre_clip_avg': 0.5517881900072098, + 'learning_rate': 8.65e-06, 'epoch': 0.44} +04/19 [11:54:57] INFO | >> train_qwenlatent.py:487 + Step 1740 | grad_norm_pre_clip=0.5438 | + grad_norm_pre_clip_avg=0.5449 | Metrics: + {'align_loss': 0.023722294718027115, + 'recon_loss': 0.008312133140861988, + 'predict_loss': 0.03564032167196274, + 'aux_loss_decay_weight': 0.6522, + 'grad_norm_pre_clip': 0.5437958240509033, + 'data_time': 0.000915865006390959, + 'model_time': 1.2271711439825594, + 'grad_norm_pre_clip_avg': 0.54490467607975, + 'learning_rate': 8.7e-06, 'epoch': 0.44} +04/19 [11:55:11] INFO | >> train_qwenlatent.py:487 + Step 1750 | grad_norm_pre_clip=0.4909 | + grad_norm_pre_clip_avg=0.5984 | Metrics: + {'align_loss': 0.02471233904361725, + 'recon_loss': 0.009035062976181507, + 'predict_loss': 0.03099116124212742, + 'aux_loss_decay_weight': 0.6502, + 'grad_norm_pre_clip': 0.490914523601532, + 'mae_score': 0.06370100416578688, 'data_time': + 0.0008774469897616655, 'model_time': + 1.2662721759988926, 'grad_norm_pre_clip_avg': + 0.5983615696430207, 'learning_rate': 8.75e-06, + 'epoch': 0.44} +04/19 [11:55:23] INFO | >> train_qwenlatent.py:487 + Step 1760 | grad_norm_pre_clip=0.5781 | + grad_norm_pre_clip_avg=0.5642 | Metrics: + {'align_loss': 0.025808997452259064, + 'recon_loss': 0.0046940455213189125, + 'predict_loss': 0.03534271568059921, + 'aux_loss_decay_weight': 0.6482, + 'grad_norm_pre_clip': 0.5780545473098755, + 'data_time': 0.0006855659885331988, + 'model_time': 1.2758476570015773, + 'grad_norm_pre_clip_avg': 0.5641640961170197, + 'learning_rate': 8.8e-06, 'epoch': 0.44} +04/19 [11:55:36] INFO | >> train_qwenlatent.py:487 + Step 1770 | grad_norm_pre_clip=0.6103 | + grad_norm_pre_clip_avg=0.6737 | Metrics: + {'align_loss': 0.02431790716946125, + 'recon_loss': 0.008194669149816036, + 'predict_loss': 0.030868716537952423, + 'aux_loss_decay_weight': 0.6462, + 'grad_norm_pre_clip': 0.61033034324646, + 'data_time': 0.0008271670085377991, + 'model_time': 1.22965765898698, + 'grad_norm_pre_clip_avg': 0.673735362291336, + 'learning_rate': 8.85e-06, 'epoch': 0.45} +04/19 [11:55:48] INFO | >> train_qwenlatent.py:487 + Step 1780 | grad_norm_pre_clip=0.5272 | + grad_norm_pre_clip_avg=0.6484 | Metrics: + {'align_loss': 0.025445476174354553, + 'recon_loss': 0.011883191764354706, + 'predict_loss': 0.03968770056962967, + 'aux_loss_decay_weight': 0.6442, + 'grad_norm_pre_clip': 0.5271768569946289, + 'data_time': 0.0005919399845879525, + 'model_time': 1.210966054990422, + 'grad_norm_pre_clip_avg': 0.6483520686626434, + 'learning_rate': 8.9e-06, 'epoch': 0.45} +04/19 [11:56:01] INFO | >> train_qwenlatent.py:487 + Step 1790 | grad_norm_pre_clip=0.5651 | + grad_norm_pre_clip_avg=0.6166 | Metrics: + {'align_loss': 0.023945458233356476, + 'recon_loss': 0.005364872515201569, + 'predict_loss': 0.027986492961645126, + 'aux_loss_decay_weight': 0.6422, + 'grad_norm_pre_clip': 0.5650919079780579, + 'data_time': 0.000905065011465922, + 'model_time': 1.2345425820094533, + 'grad_norm_pre_clip_avg': 0.6166453570127487, + 'learning_rate': 8.95e-06, 'epoch': 0.45} +04/19 [11:56:14] INFO | >> train_qwenlatent.py:487 + Step 1800 | grad_norm_pre_clip=0.6220 | + grad_norm_pre_clip_avg=0.5507 | Metrics: + {'align_loss': 0.023767758160829544, + 'recon_loss': 0.004665734712034464, + 'predict_loss': 0.029888909310102463, + 'aux_loss_decay_weight': 0.6402, + 'grad_norm_pre_clip': 0.6219760179519653, + 'mae_score': 0.050690240258569115, 'data_time': + 0.0008613730024080724, 'model_time': + 1.2713564230070915, 'grad_norm_pre_clip_avg': + 0.5506889939308166, 'learning_rate': 9e-06, + 'epoch': 0.45} +04/19 [11:56:27] INFO | >> train_qwenlatent.py:487 + Step 1810 | grad_norm_pre_clip=0.6480 | + grad_norm_pre_clip_avg=0.6275 | Metrics: + {'align_loss': 0.024356290698051453, + 'recon_loss': 0.0051360526122152805, + 'predict_loss': 0.02376473881304264, + 'aux_loss_decay_weight': 0.6382, + 'grad_norm_pre_clip': 0.6479856967926025, + 'data_time': 0.0006129280081950128, + 'model_time': 1.2938003670133185, + 'grad_norm_pre_clip_avg': 0.6274512469768524, + 'learning_rate': 9.05e-06, 'epoch': 0.46} +04/19 [11:56:39] INFO | >> train_qwenlatent.py:487 + Step 1820 | grad_norm_pre_clip=0.5603 | + grad_norm_pre_clip_avg=0.6321 | Metrics: + {'align_loss': 0.023837676271796227, + 'recon_loss': 0.009491520002484322, + 'predict_loss': 0.03195900097489357, + 'aux_loss_decay_weight': 0.6362, + 'grad_norm_pre_clip': 0.5602899789810181, + 'data_time': 0.0011402930249460042, + 'model_time': 1.2018301850184798, + 'grad_norm_pre_clip_avg': 0.6320798635482788, + 'learning_rate': 9.100000000000001e-06, + 'epoch': 0.46} +04/19 [11:56:52] INFO | >> train_qwenlatent.py:487 + Step 1830 | grad_norm_pre_clip=0.6168 | + grad_norm_pre_clip_avg=0.5734 | Metrics: + {'align_loss': 0.02422238327562809, + 'recon_loss': 0.011804704554378986, + 'predict_loss': 0.04311765730381012, + 'aux_loss_decay_weight': 0.6342, + 'grad_norm_pre_clip': 0.6168129444122314, + 'data_time': 0.0006671730079688132, + 'model_time': 1.2339072179747745, + 'grad_norm_pre_clip_avg': 0.5734232068061829, + 'learning_rate': 9.15e-06, 'epoch': 0.46} +04/19 [11:57:05] INFO | >> train_qwenlatent.py:487 + Step 1840 | grad_norm_pre_clip=0.4599 | + grad_norm_pre_clip_avg=0.5755 | Metrics: + {'align_loss': 0.02404678426682949, + 'recon_loss': 0.004290821496397257, + 'predict_loss': 0.03641946241259575, + 'aux_loss_decay_weight': 0.6322, + 'grad_norm_pre_clip': 0.4598890542984009, + 'data_time': 0.0007027230167295784, + 'model_time': 1.2618279219896067, + 'grad_norm_pre_clip_avg': 0.575549179315567, + 'learning_rate': 9.2e-06, 'epoch': 0.46} +04/19 [11:57:18] INFO | >> train_qwenlatent.py:487 + Step 1850 | grad_norm_pre_clip=0.5295 | + grad_norm_pre_clip_avg=0.5782 | Metrics: + {'align_loss': 0.025194134563207626, + 'recon_loss': 0.006156908813863993, + 'predict_loss': 0.03117513656616211, + 'aux_loss_decay_weight': 0.6302, + 'grad_norm_pre_clip': 0.5295498967170715, + 'mae_score': 0.0657550708667652, 'data_time': + 0.0008268399978987873, 'model_time': + 1.2555414480157197, 'grad_norm_pre_clip_avg': + 0.5782164573669434, 'learning_rate': 9.25e-06, + 'epoch': 0.47} +04/19 [11:57:31] INFO | >> train_qwenlatent.py:487 + Step 1860 | grad_norm_pre_clip=0.4803 | + grad_norm_pre_clip_avg=0.4983 | Metrics: + {'align_loss': 0.02486896887421608, + 'recon_loss': 0.007062074262648821, + 'predict_loss': 0.031113434582948685, + 'aux_loss_decay_weight': 0.6282, + 'grad_norm_pre_clip': 0.4803470969200134, + 'data_time': 0.0008212320099119097, + 'model_time': 1.2291301889927126, + 'grad_norm_pre_clip_avg': 0.49825104176998136, + 'learning_rate': 9.3e-06, 'epoch': 0.47} +04/19 [11:57:44] INFO | >> train_qwenlatent.py:487 + Step 1870 | grad_norm_pre_clip=0.7478 | + grad_norm_pre_clip_avg=0.5961 | Metrics: + {'align_loss': 0.024153761565685272, + 'recon_loss': 0.008813256397843361, + 'predict_loss': 0.035207100212574005, + 'aux_loss_decay_weight': 0.6262, + 'grad_norm_pre_clip': 0.7478371858596802, + 'data_time': 0.0007881770143285394, + 'model_time': 1.5063766590028536, + 'grad_norm_pre_clip_avg': 0.5961099743843079, + 'learning_rate': 9.35e-06, 'epoch': 0.47} +04/19 [11:57:56] INFO | >> train_qwenlatent.py:487 + Step 1880 | grad_norm_pre_clip=0.5561 | + grad_norm_pre_clip_avg=0.6013 | Metrics: + {'align_loss': 0.022776832804083824, + 'recon_loss': 0.0024374520871788263, + 'predict_loss': 0.022545376792550087, + 'aux_loss_decay_weight': 0.6242, + 'grad_norm_pre_clip': 0.5561217665672302, + 'data_time': 0.0006279190129134804, + 'model_time': 1.2328095780103467, + 'grad_norm_pre_clip_avg': 0.6013128876686096, + 'learning_rate': 9.4e-06, 'epoch': 0.47} +04/19 [11:58:09] INFO | >> train_qwenlatent.py:487 + Step 1890 | grad_norm_pre_clip=0.5869 | + grad_norm_pre_clip_avg=0.5785 | Metrics: + {'align_loss': 0.02461264468729496, + 'recon_loss': 0.006137957330793142, + 'predict_loss': 0.03381895646452904, + 'aux_loss_decay_weight': 0.6222, + 'grad_norm_pre_clip': 0.5869371294975281, + 'data_time': 0.0010538019996602088, + 'model_time': 1.2580235589994118, + 'grad_norm_pre_clip_avg': 0.5784593492746353, + 'learning_rate': 9.450000000000001e-06, + 'epoch': 0.48} +04/19 [11:58:22] INFO | >> train_qwenlatent.py:487 + Step 1900 | grad_norm_pre_clip=0.4869 | + grad_norm_pre_clip_avg=0.5469 | Metrics: + {'align_loss': 0.025045691058039665, + 'recon_loss': 0.006785938516259193, + 'predict_loss': 0.03395126387476921, + 'aux_loss_decay_weight': 0.6202, + 'grad_norm_pre_clip': 0.4869033694267273, + 'mae_score': 0.06135961171743032, 'data_time': + 0.0009965250210370868, 'model_time': + 1.225450567988446, 'grad_norm_pre_clip_avg': + 0.5468532919883728, 'learning_rate': 9.5e-06, + 'epoch': 0.48} +04/19 [11:58:35] INFO | >> train_qwenlatent.py:487 + Step 1910 | grad_norm_pre_clip=0.6519 | + grad_norm_pre_clip_avg=0.5882 | Metrics: + {'align_loss': 0.024761484935879707, + 'recon_loss': 0.002680625068023801, + 'predict_loss': 0.02597906067967415, + 'aux_loss_decay_weight': 0.6182000000000001, + 'grad_norm_pre_clip': 0.6519054174423218, + 'data_time': 0.0009512700198683888, + 'model_time': 1.2795811840042006, + 'grad_norm_pre_clip_avg': 0.5881603628396987, + 'learning_rate': 9.55e-06, 'epoch': 0.48} +04/19 [11:58:47] INFO | >> train_qwenlatent.py:487 + Step 1920 | grad_norm_pre_clip=0.5156 | + grad_norm_pre_clip_avg=0.7006 | Metrics: + {'align_loss': 0.02381368726491928, + 'recon_loss': 0.006751138251274824, + 'predict_loss': 0.029497377574443817, + 'aux_loss_decay_weight': 0.6162000000000001, + 'grad_norm_pre_clip': 0.5156229138374329, + 'data_time': 0.000673740025376901, + 'model_time': 1.2660829090164043, + 'grad_norm_pre_clip_avg': 0.7005640029907226, + 'learning_rate': 9.600000000000001e-06, + 'epoch': 0.48} +04/19 [11:59:00] INFO | >> train_qwenlatent.py:487 + Step 1930 | grad_norm_pre_clip=0.5068 | + grad_norm_pre_clip_avg=0.5778 | Metrics: + {'align_loss': 0.025520525872707367, + 'recon_loss': 0.005421836860477924, + 'predict_loss': 0.024125022813677788, + 'aux_loss_decay_weight': 0.6142000000000001, + 'grad_norm_pre_clip': 0.5068041682243347, + 'data_time': 0.001095626997994259, + 'model_time': 1.2984274740156252, + 'grad_norm_pre_clip_avg': 0.5777956068515777, + 'learning_rate': 9.65e-06, 'epoch': 0.49} +04/19 [11:59:12] INFO | >> train_qwenlatent.py:487 + Step 1940 | grad_norm_pre_clip=0.4159 | + grad_norm_pre_clip_avg=0.5569 | Metrics: + {'align_loss': 0.025130461901426315, + 'recon_loss': 0.005392567720264196, + 'predict_loss': 0.03211025148630142, + 'aux_loss_decay_weight': 0.6122000000000001, + 'grad_norm_pre_clip': 0.4158913493156433, + 'data_time': 0.001116659987019375, + 'model_time': 1.2444676180020906, + 'grad_norm_pre_clip_avg': 0.5568526893854141, + 'learning_rate': 9.7e-06, 'epoch': 0.49} +04/19 [11:59:25] INFO | >> train_qwenlatent.py:487 + Step 1950 | grad_norm_pre_clip=0.4734 | + grad_norm_pre_clip_avg=0.5427 | Metrics: + {'align_loss': 0.02345409244298935, + 'recon_loss': 0.0056565976701676846, + 'predict_loss': 0.034707099199295044, + 'aux_loss_decay_weight': 0.6102000000000001, + 'grad_norm_pre_clip': 0.47343534231185913, + 'mae_score': 0.06134480656804265, 'data_time': + 0.0007140079978853464, 'model_time': + 1.3208345510065556, 'grad_norm_pre_clip_avg': + 0.5426519989967347, 'learning_rate': + 9.750000000000002e-06, 'epoch': 0.49} +04/19 [11:59:38] INFO | >> train_qwenlatent.py:487 + Step 1960 | grad_norm_pre_clip=0.5131 | + grad_norm_pre_clip_avg=0.5487 | Metrics: + {'align_loss': 0.025184031575918198, + 'recon_loss': 0.0044642346911132336, + 'predict_loss': 0.033570416271686554, + 'aux_loss_decay_weight': 0.6082000000000001, + 'grad_norm_pre_clip': 0.513132631778717, + 'data_time': 0.0008440169913228601, + 'model_time': 1.27581367501989, + 'grad_norm_pre_clip_avg': 0.5487016499042511, + 'learning_rate': 9.800000000000001e-06, + 'epoch': 0.49} +04/19 [11:59:52] INFO | >> train_qwenlatent.py:487 + Step 1970 | grad_norm_pre_clip=0.6578 | + grad_norm_pre_clip_avg=0.5772 | Metrics: + {'align_loss': 0.02457955852150917, + 'recon_loss': 0.005722357425838709, + 'predict_loss': 0.03120674006640911, + 'aux_loss_decay_weight': 0.6062000000000001, + 'grad_norm_pre_clip': 0.6578386425971985, + 'data_time': 0.0006875630060676485, + 'model_time': 1.2062895220005885, + 'grad_norm_pre_clip_avg': 0.5772415608167648, + 'learning_rate': 9.85e-06, 'epoch': 0.5} +04/19 [12:00:04] INFO | >> train_qwenlatent.py:487 + Step 1980 | grad_norm_pre_clip=0.5748 | + grad_norm_pre_clip_avg=0.5783 | Metrics: + {'align_loss': 0.024102995172142982, + 'recon_loss': 0.008402218110859394, + 'predict_loss': 0.03446817398071289, + 'aux_loss_decay_weight': 0.6042000000000001, + 'grad_norm_pre_clip': 0.5747913718223572, + 'data_time': 0.0008862600079737604, + 'model_time': 1.23369422997348, + 'grad_norm_pre_clip_avg': 0.5783439755439759, + 'learning_rate': 9.900000000000002e-06, + 'epoch': 0.5} +04/19 [12:00:17] INFO | >> train_qwenlatent.py:487 + Step 1990 | grad_norm_pre_clip=0.5412 | + grad_norm_pre_clip_avg=0.5362 | Metrics: + {'align_loss': 0.023671843111515045, + 'recon_loss': 0.0020509955938905478, + 'predict_loss': 0.022630685940384865, + 'aux_loss_decay_weight': 0.6022000000000001, + 'grad_norm_pre_clip': 0.5412214994430542, + 'data_time': 0.0009693810134194791, + 'model_time': 1.207772132998798, + 'grad_norm_pre_clip_avg': 0.5362268686294556, + 'learning_rate': 9.950000000000001e-06, + 'epoch': 0.5} +04/19 [12:00:31] INFO | >> train_qwenlatent.py:487 + Step 2000 | grad_norm_pre_clip=0.6503 | + grad_norm_pre_clip_avg=0.6181 | Metrics: + {'align_loss': 0.024974089115858078, + 'recon_loss': 0.009975327178835869, + 'predict_loss': 0.03364246338605881, + 'aux_loss_decay_weight': 0.6002000000000001, + 'grad_norm_pre_clip': 0.6503397822380066, + 'mae_score': 0.04513114379332946, 'data_time': + 0.0009350999898742884, 'model_time': + 1.2066740660229698, 'grad_norm_pre_clip_avg': + 0.6180634737014771, 'learning_rate': 1e-05, + 'epoch': 0.5} +04/19 [12:00:43] INFO | >> train_qwenlatent.py:487 + Step 2010 | grad_norm_pre_clip=0.4491 | + grad_norm_pre_clip_avg=0.5630 | Metrics: + {'align_loss': 0.022681323811411858, + 'recon_loss': 0.004999394528567791, + 'predict_loss': 0.027891000732779503, + 'aux_loss_decay_weight': 0.5982000000000001, + 'grad_norm_pre_clip': 0.44906085729599, + 'data_time': 0.000672387977829203, + 'model_time': 1.2919771599990781, + 'grad_norm_pre_clip_avg': 0.5629947304725647, + 'learning_rate': 1.005e-05, 'epoch': 0.51} +04/19 [12:00:56] INFO | >> train_qwenlatent.py:487 + Step 2020 | grad_norm_pre_clip=0.5181 | + grad_norm_pre_clip_avg=0.5086 | Metrics: + {'align_loss': 0.025000249966979027, + 'recon_loss': 0.005975112318992615, + 'predict_loss': 0.030541006475687027, + 'aux_loss_decay_weight': 0.5962000000000001, + 'grad_norm_pre_clip': 0.5181341767311096, + 'data_time': 0.0008558209810871631, + 'model_time': 1.2745775590010453, + 'grad_norm_pre_clip_avg': 0.5085994511842727, + 'learning_rate': 1.0100000000000002e-05, + 'epoch': 0.51} +04/19 [12:01:08] INFO | >> train_qwenlatent.py:487 + Step 2030 | grad_norm_pre_clip=0.5369 | + grad_norm_pre_clip_avg=0.5415 | Metrics: + {'align_loss': 0.02469370700418949, + 'recon_loss': 0.005570726934820414, + 'predict_loss': 0.035783104598522186, + 'aux_loss_decay_weight': 0.5942000000000001, + 'grad_norm_pre_clip': 0.5368527770042419, + 'data_time': 0.0006654669996351004, + 'model_time': 1.28827716797241, + 'grad_norm_pre_clip_avg': 0.5415194511413575, + 'learning_rate': 1.0150000000000001e-05, + 'epoch': 0.51} +04/19 [12:01:21] INFO | >> train_qwenlatent.py:487 + Step 2040 | grad_norm_pre_clip=0.5701 | + grad_norm_pre_clip_avg=0.5470 | Metrics: + {'align_loss': 0.022949103266000748, + 'recon_loss': 0.0068995151668787, + 'predict_loss': 0.03109419345855713, + 'aux_loss_decay_weight': 0.5922000000000001, + 'grad_norm_pre_clip': 0.5700947046279907, + 'data_time': 0.0009262810053769499, + 'model_time': 1.292708193010185, + 'grad_norm_pre_clip_avg': 0.5469867348670959, + 'learning_rate': 1.02e-05, 'epoch': 0.51} +04/19 [12:01:34] INFO | >> train_qwenlatent.py:487 + Step 2050 | grad_norm_pre_clip=0.5543 | + grad_norm_pre_clip_avg=0.5542 | Metrics: + {'align_loss': 0.02407112345099449, + 'recon_loss': 0.007611692417412996, + 'predict_loss': 0.027840353548526764, + 'aux_loss_decay_weight': 0.5902000000000001, + 'grad_norm_pre_clip': 0.5542599558830261, + 'mae_score': 0.05482797021264429, 'data_time': + 0.0009105070203077048, 'model_time': + 1.3024973270075861, 'grad_norm_pre_clip_avg': + 0.554161325097084, 'learning_rate': 1.025e-05, + 'epoch': 0.52} +04/19 [12:01:47] INFO | >> train_qwenlatent.py:487 + Step 2060 | grad_norm_pre_clip=0.6629 | + grad_norm_pre_clip_avg=0.6358 | Metrics: + {'align_loss': 0.023546945303678513, + 'recon_loss': 0.004102468024939299, + 'predict_loss': 0.029819263145327568, + 'aux_loss_decay_weight': 0.5882000000000001, + 'grad_norm_pre_clip': 0.6629207134246826, + 'data_time': 0.0005940280098002404, + 'model_time': 1.2376821670040954, + 'grad_norm_pre_clip_avg': 0.6357886373996735, + 'learning_rate': 1.03e-05, 'epoch': 0.52} +04/19 [12:01:59] INFO | >> train_qwenlatent.py:487 + Step 2070 | grad_norm_pre_clip=0.4954 | + grad_norm_pre_clip_avg=0.5251 | Metrics: + {'align_loss': 0.02395174279808998, + 'recon_loss': 0.008532381616532803, + 'predict_loss': 0.044330060482025146, + 'aux_loss_decay_weight': 0.5862, + 'grad_norm_pre_clip': 0.4953500032424927, + 'data_time': 0.0006630859861616045, + 'model_time': 1.2242266690009274, + 'grad_norm_pre_clip_avg': 0.525052186846733, + 'learning_rate': 1.035e-05, 'epoch': 0.52} +04/19 [12:02:11] INFO | >> train_qwenlatent.py:487 + Step 2080 | grad_norm_pre_clip=0.4782 | + grad_norm_pre_clip_avg=0.4864 | Metrics: + {'align_loss': 0.023040052503347397, + 'recon_loss': 0.004095826763659716, + 'predict_loss': 0.025157906115055084, + 'aux_loss_decay_weight': 0.5842, + 'grad_norm_pre_clip': 0.47816938161849976, + 'data_time': 0.001192027993965894, + 'model_time': 1.275370049988851, + 'grad_norm_pre_clip_avg': 0.48640435338020327, + 'learning_rate': 1.04e-05, 'epoch': 0.52} +04/19 [12:02:24] INFO | >> train_qwenlatent.py:487 + Step 2090 | grad_norm_pre_clip=0.5660 | + grad_norm_pre_clip_avg=0.6031 | Metrics: + {'align_loss': 0.02367468550801277, + 'recon_loss': 0.007761262822896242, + 'predict_loss': 0.028108419850468636, + 'aux_loss_decay_weight': 0.5822, + 'grad_norm_pre_clip': 0.5659744143486023, + 'data_time': 0.001176936988485977, + 'model_time': 1.2218601829954423, + 'grad_norm_pre_clip_avg': 0.6030598104000091, + 'learning_rate': 1.045e-05, 'epoch': 0.53} +04/19 [12:02:38] INFO | >> train_qwenlatent.py:487 + Step 2100 | grad_norm_pre_clip=0.5402 | + grad_norm_pre_clip_avg=0.5512 | Metrics: + {'align_loss': 0.02479936182498932, + 'recon_loss': 0.0060056098736822605, + 'predict_loss': 0.02722199633717537, + 'aux_loss_decay_weight': 0.5802, + 'grad_norm_pre_clip': 0.5401509404182434, + 'mae_score': 0.043611035046276746, 'data_time': + 0.000796740991063416, 'model_time': + 1.242820537998341, 'grad_norm_pre_clip_avg': + 0.5512465417385102, 'learning_rate': 1.05e-05, + 'epoch': 0.53} +04/19 [12:02:50] INFO | >> train_qwenlatent.py:487 + Step 2110 | grad_norm_pre_clip=0.5381 | + grad_norm_pre_clip_avg=0.5505 | Metrics: + {'align_loss': 0.023965947329998016, + 'recon_loss': 0.007600463926792145, + 'predict_loss': 0.025156313553452492, + 'aux_loss_decay_weight': 0.5782, + 'grad_norm_pre_clip': 0.5380957126617432, + 'data_time': 0.0006112839910201728, + 'model_time': 1.5042500520066824, + 'grad_norm_pre_clip_avg': 0.5504825592041016, + 'learning_rate': 1.055e-05, 'epoch': 0.53} +04/19 [12:03:03] INFO | >> train_qwenlatent.py:487 + Step 2120 | grad_norm_pre_clip=0.5218 | + grad_norm_pre_clip_avg=0.5555 | Metrics: + {'align_loss': 0.02410796284675598, + 'recon_loss': 0.02248762734234333, + 'predict_loss': 0.0496688149869442, + 'aux_loss_decay_weight': 0.5762, + 'grad_norm_pre_clip': 0.5217839479446411, + 'data_time': 0.0006120869948063046, + 'model_time': 1.1990882790123578, + 'grad_norm_pre_clip_avg': 0.555519112944603, + 'learning_rate': 1.06e-05, 'epoch': 0.53} +04/19 [12:03:16] INFO | >> train_qwenlatent.py:487 + Step 2130 | grad_norm_pre_clip=0.6117 | + grad_norm_pre_clip_avg=0.5560 | Metrics: + {'align_loss': 0.025182394310832024, + 'recon_loss': 0.011399400420486927, + 'predict_loss': 0.03504691645503044, + 'aux_loss_decay_weight': 0.5742, + 'grad_norm_pre_clip': 0.6116732954978943, + 'data_time': 0.0009099399903789163, + 'model_time': 1.210625968000386, + 'grad_norm_pre_clip_avg': 0.5559687584638595, + 'learning_rate': 1.065e-05, 'epoch': 0.54} +04/19 [12:03:28] INFO | >> train_qwenlatent.py:487 + Step 2140 | grad_norm_pre_clip=0.5412 | + grad_norm_pre_clip_avg=0.5090 | Metrics: + {'align_loss': 0.02326965145766735, + 'recon_loss': 0.006575976498425007, + 'predict_loss': 0.02578097954392433, + 'aux_loss_decay_weight': 0.5722, + 'grad_norm_pre_clip': 0.5412425994873047, + 'data_time': 0.0006646150141023099, + 'model_time': 1.5589934310119133, + 'grad_norm_pre_clip_avg': 0.5089563399553299, + 'learning_rate': 1.0700000000000001e-05, + 'epoch': 0.54} +04/19 [12:03:42] INFO | >> train_qwenlatent.py:487 + Step 2150 | grad_norm_pre_clip=0.5942 | + grad_norm_pre_clip_avg=0.5646 | Metrics: + {'align_loss': 0.023752206936478615, + 'recon_loss': 0.007594150956720114, + 'predict_loss': 0.03311458230018616, + 'aux_loss_decay_weight': 0.5702, + 'grad_norm_pre_clip': 0.5942310094833374, + 'mae_score': 0.0553923529547614, 'data_time': + 0.0009009110217448324, 'model_time': + 1.2449011730204802, 'grad_norm_pre_clip_avg': + 0.5646175146102905, 'learning_rate': 1.075e-05, + 'epoch': 0.54} +04/19 [12:03:54] INFO | >> train_qwenlatent.py:487 + Step 2160 | grad_norm_pre_clip=0.5143 | + grad_norm_pre_clip_avg=0.5384 | Metrics: + {'align_loss': 0.023874834179878235, + 'recon_loss': 0.010702000930905342, + 'predict_loss': 0.034014709293842316, + 'aux_loss_decay_weight': 0.5682, + 'grad_norm_pre_clip': 0.5143216252326965, + 'data_time': 0.0008282479830086231, + 'model_time': 1.2740120940143242, + 'grad_norm_pre_clip_avg': 0.5383873224258423, + 'learning_rate': 1.08e-05, 'epoch': 0.55} +04/19 [12:04:07] INFO | >> train_qwenlatent.py:487 + Step 2170 | grad_norm_pre_clip=0.4498 | + grad_norm_pre_clip_avg=0.5037 | Metrics: + {'align_loss': 0.024823946878314018, + 'recon_loss': 0.0067762769758701324, + 'predict_loss': 0.03617312014102936, + 'aux_loss_decay_weight': 0.5662, + 'grad_norm_pre_clip': 0.44976910948753357, + 'data_time': 0.0006455620168708265, + 'model_time': 1.2707801969954744, + 'grad_norm_pre_clip_avg': 0.503716042637825, + 'learning_rate': 1.0850000000000001e-05, + 'epoch': 0.55} +04/19 [12:04:19] INFO | >> train_qwenlatent.py:487 + Step 2180 | grad_norm_pre_clip=0.5806 | + grad_norm_pre_clip_avg=0.4570 | Metrics: + {'align_loss': 0.02482910268008709, + 'recon_loss': 0.006852617021650076, + 'predict_loss': 0.03722621127963066, + 'aux_loss_decay_weight': 0.5642, + 'grad_norm_pre_clip': 0.5805693864822388, + 'data_time': 0.0007023469952400774, + 'model_time': 1.2409528919961303, + 'grad_norm_pre_clip_avg': 0.4569974303245544, + 'learning_rate': 1.09e-05, 'epoch': 0.55} +04/19 [12:04:32] INFO | >> train_qwenlatent.py:487 + Step 2190 | grad_norm_pre_clip=0.7495 | + grad_norm_pre_clip_avg=0.5384 | Metrics: + {'align_loss': 0.023050453513860703, + 'recon_loss': 0.0074111646972596645, + 'predict_loss': 0.03444554656744003, + 'aux_loss_decay_weight': 0.5622, + 'grad_norm_pre_clip': 0.7495473027229309, + 'data_time': 0.000659832003293559, + 'model_time': 1.2543873229878955, + 'grad_norm_pre_clip_avg': 0.5384001433849335, + 'learning_rate': 1.095e-05, 'epoch': 0.55} +04/19 [12:04:45] INFO | >> train_qwenlatent.py:487 + Step 2200 | grad_norm_pre_clip=0.5732 | + grad_norm_pre_clip_avg=0.5833 | Metrics: + {'align_loss': 0.02303450182080269, + 'recon_loss': 0.007512244861572981, + 'predict_loss': 0.03140317276120186, + 'aux_loss_decay_weight': 0.5602, + 'grad_norm_pre_clip': 0.5731870532035828, + 'mae_score': 0.056423273172464455, 'data_time': + 0.0011243779736105353, 'model_time': + 1.252964981016703, 'grad_norm_pre_clip_avg': + 0.5832941323518753, 'learning_rate': + 1.1000000000000001e-05, 'epoch': 0.56} +04/19 [12:04:58] INFO | >> train_qwenlatent.py:487 + Step 2210 | grad_norm_pre_clip=0.5686 | + grad_norm_pre_clip_avg=0.5782 | Metrics: + {'align_loss': 0.02309560775756836, + 'recon_loss': 0.016271552070975304, + 'predict_loss': 0.04386180266737938, + 'aux_loss_decay_weight': 0.5582, + 'grad_norm_pre_clip': 0.5686267614364624, + 'data_time': 0.0008467260049656034, + 'model_time': 1.2556532680173405, + 'grad_norm_pre_clip_avg': 0.5781837612390518, + 'learning_rate': 1.1050000000000001e-05, + 'epoch': 0.56} +04/19 [12:05:10] INFO | >> train_qwenlatent.py:487 + Step 2220 | grad_norm_pre_clip=0.5174 | + grad_norm_pre_clip_avg=0.5065 | Metrics: + {'align_loss': 0.023550668731331825, + 'recon_loss': 0.010271039791405201, + 'predict_loss': 0.03329591453075409, + 'aux_loss_decay_weight': 0.5562, + 'grad_norm_pre_clip': 0.5173726081848145, + 'data_time': 0.0008099419937934726, + 'model_time': 1.5262016339984257, + 'grad_norm_pre_clip_avg': 0.5064618438482285, + 'learning_rate': 1.11e-05, 'epoch': 0.56} +04/19 [12:05:23] INFO | >> train_qwenlatent.py:487 + Step 2230 | grad_norm_pre_clip=0.4469 | + grad_norm_pre_clip_avg=0.4621 | Metrics: + {'align_loss': 0.022824229672551155, + 'recon_loss': 0.004010855220258236, + 'predict_loss': 0.02834419161081314, + 'aux_loss_decay_weight': 0.5542, + 'grad_norm_pre_clip': 0.4469471871852875, + 'data_time': 0.0006245539989322424, + 'model_time': 1.2137140610138886, + 'grad_norm_pre_clip_avg': 0.4621173918247223, + 'learning_rate': 1.115e-05, 'epoch': 0.56} +04/19 [12:05:36] INFO | >> train_qwenlatent.py:487 + Step 2240 | grad_norm_pre_clip=0.6042 | + grad_norm_pre_clip_avg=0.5078 | Metrics: + {'align_loss': 0.02358144149184227, + 'recon_loss': 0.010182012803852558, + 'predict_loss': 0.03768617659807205, + 'aux_loss_decay_weight': 0.5522, + 'grad_norm_pre_clip': 0.6041983962059021, + 'data_time': 0.0007970110164023936, + 'model_time': 1.2639774870185647, + 'grad_norm_pre_clip_avg': 0.5077879667282105, + 'learning_rate': 1.1200000000000001e-05, + 'epoch': 0.57} +04/19 [12:05:50] INFO | >> train_qwenlatent.py:487 + Step 2250 | grad_norm_pre_clip=0.4625 | + grad_norm_pre_clip_avg=0.5324 | Metrics: + {'align_loss': 0.022476255893707275, + 'recon_loss': 0.005954229738563299, + 'predict_loss': 0.02538718469440937, + 'aux_loss_decay_weight': 0.5502, + 'grad_norm_pre_clip': 0.4624537527561188, + 'mae_score': 0.049287548580685174, 'data_time': + 0.0009561589977238327, 'model_time': + 1.2549172790022567, 'grad_norm_pre_clip_avg': + 0.5323765724897385, 'learning_rate': 1.125e-05, + 'epoch': 0.57} +04/19 [12:06:02] INFO | >> train_qwenlatent.py:487 + Step 2260 | grad_norm_pre_clip=0.4420 | + grad_norm_pre_clip_avg=0.5016 | Metrics: + {'align_loss': 0.022164175286889076, + 'recon_loss': 0.005506652873009443, + 'predict_loss': 0.029557721689343452, + 'aux_loss_decay_weight': 0.5482, + 'grad_norm_pre_clip': 0.4420354664325714, + 'data_time': 0.0008881860121618956, + 'model_time': 1.2209395859972574, + 'grad_norm_pre_clip_avg': 0.5016061693429947, + 'learning_rate': 1.13e-05, 'epoch': 0.57} +04/19 [12:06:15] INFO | >> train_qwenlatent.py:487 + Step 2270 | grad_norm_pre_clip=0.6607 | + grad_norm_pre_clip_avg=0.5360 | Metrics: + {'align_loss': 0.024289537221193314, + 'recon_loss': 0.008391587063670158, + 'predict_loss': 0.03612937405705452, + 'aux_loss_decay_weight': 0.5462, + 'grad_norm_pre_clip': 0.6606733202934265, + 'data_time': 0.0008633520046714693, + 'model_time': 1.2724034330167342, + 'grad_norm_pre_clip_avg': 0.5359925746917724, + 'learning_rate': 1.1350000000000001e-05, + 'epoch': 0.57} +04/19 [12:06:28] INFO | >> train_qwenlatent.py:487 + Step 2280 | grad_norm_pre_clip=0.4582 | + grad_norm_pre_clip_avg=0.5109 | Metrics: + {'align_loss': 0.023279234766960144, + 'recon_loss': 0.007335075177252293, + 'predict_loss': 0.02633289061486721, + 'aux_loss_decay_weight': 0.5442, + 'grad_norm_pre_clip': 0.45820152759552, + 'data_time': 0.000763785996241495, + 'model_time': 1.2477917759970296, + 'grad_norm_pre_clip_avg': 0.5109043180942535, + 'learning_rate': 1.1400000000000001e-05, + 'epoch': 0.58} +04/19 [12:06:40] INFO | >> train_qwenlatent.py:487 + Step 2290 | grad_norm_pre_clip=0.4705 | + grad_norm_pre_clip_avg=0.5110 | Metrics: + {'align_loss': 0.023232106119394302, + 'recon_loss': 0.008042633533477783, + 'predict_loss': 0.026610510423779488, + 'aux_loss_decay_weight': 0.5422, + 'grad_norm_pre_clip': 0.4705392122268677, + 'data_time': 0.0007868300017435104, + 'model_time': 1.2007515720033552, + 'grad_norm_pre_clip_avg': 0.5109700322151184, + 'learning_rate': 1.145e-05, 'epoch': 0.58} +04/19 [12:06:54] INFO | >> train_qwenlatent.py:487 + Step 2300 | grad_norm_pre_clip=0.5138 | + grad_norm_pre_clip_avg=0.5202 | Metrics: + {'align_loss': 0.0224935133010149, + 'recon_loss': 0.007667193654924631, + 'predict_loss': 0.0337710864841938, + 'aux_loss_decay_weight': 0.5402, + 'grad_norm_pre_clip': 0.5137559175491333, + 'mae_score': 0.0386722942730328, 'data_time': + 0.0007944200187921524, 'model_time': + 1.2440965459973086, 'grad_norm_pre_clip_avg': + 0.5202488392591477, 'learning_rate': + 1.1500000000000002e-05, 'epoch': 0.58} +04/19 [12:07:06] INFO | >> train_qwenlatent.py:487 + Step 2310 | grad_norm_pre_clip=0.5534 | + grad_norm_pre_clip_avg=0.5279 | Metrics: + {'align_loss': 0.023532327264547348, + 'recon_loss': 0.01281894464045763, + 'predict_loss': 0.03544226661324501, + 'aux_loss_decay_weight': 0.5382, + 'grad_norm_pre_clip': 0.5533508658409119, + 'data_time': 0.0007452240097336471, + 'model_time': 1.2236001300043426, + 'grad_norm_pre_clip_avg': 0.5278979420661927, + 'learning_rate': 1.1550000000000001e-05, + 'epoch': 0.58} +04/19 [12:07:18] INFO | >> train_qwenlatent.py:487 + Step 2320 | grad_norm_pre_clip=0.4904 | + grad_norm_pre_clip_avg=0.5506 | Metrics: + {'align_loss': 0.022703785449266434, + 'recon_loss': 0.009334041737020016, + 'predict_loss': 0.031194059178233147, + 'aux_loss_decay_weight': 0.5362, + 'grad_norm_pre_clip': 0.49041256308555603, + 'data_time': 0.0008924090070649981, + 'model_time': 1.1909645239938982, + 'grad_norm_pre_clip_avg': 0.5505713850259781, + 'learning_rate': 1.16e-05, 'epoch': 0.59} +04/19 [12:07:31] INFO | >> train_qwenlatent.py:487 + Step 2330 | grad_norm_pre_clip=0.8296 | + grad_norm_pre_clip_avg=0.5456 | Metrics: + {'align_loss': 0.024018077179789543, + 'recon_loss': 0.0152325090020895, + 'predict_loss': 0.04270927980542183, + 'aux_loss_decay_weight': 0.5342, + 'grad_norm_pre_clip': 0.8296336531639099, + 'data_time': 0.0006911729869898409, + 'model_time': 1.1913404539809562, + 'grad_norm_pre_clip_avg': 0.5455514371395112, + 'learning_rate': 1.1650000000000002e-05, + 'epoch': 0.59} +04/19 [12:07:43] INFO | >> train_qwenlatent.py:487 + Step 2340 | grad_norm_pre_clip=0.5206 | + grad_norm_pre_clip_avg=0.4976 | Metrics: + {'align_loss': 0.023037098348140717, + 'recon_loss': 0.008617921732366085, + 'predict_loss': 0.03171880915760994, + 'aux_loss_decay_weight': 0.5322, + 'grad_norm_pre_clip': 0.5205798745155334, + 'data_time': 0.0007324899779632688, + 'model_time': 1.2573684520029929, + 'grad_norm_pre_clip_avg': 0.49759267568588256, + 'learning_rate': 1.1700000000000001e-05, + 'epoch': 0.59} +04/19 [12:07:57] INFO | >> train_qwenlatent.py:487 + Step 2350 | grad_norm_pre_clip=0.3739 | + grad_norm_pre_clip_avg=0.4752 | Metrics: + {'align_loss': 0.023299414664506912, + 'recon_loss': 0.0074393549002707005, + 'predict_loss': 0.032050881534814835, + 'aux_loss_decay_weight': 0.5302, + 'grad_norm_pre_clip': 0.3738999664783478, + 'mae_score': 0.058841952762088263, 'data_time': + 0.0005869290034752339, 'model_time': + 1.1824781850154977, 'grad_norm_pre_clip_avg': + 0.4752335220575333, 'learning_rate': 1.175e-05, + 'epoch': 0.59} +04/19 [12:08:10] INFO | >> train_qwenlatent.py:487 + Step 2360 | grad_norm_pre_clip=0.4967 | + grad_norm_pre_clip_avg=0.4971 | Metrics: + {'align_loss': 0.022318288683891296, + 'recon_loss': 0.00603793328627944, + 'predict_loss': 0.029475940391421318, + 'aux_loss_decay_weight': 0.5282, + 'grad_norm_pre_clip': 0.4967351257801056, + 'data_time': 0.0009432199876755476, + 'model_time': 1.2392925399763044, + 'grad_norm_pre_clip_avg': 0.4970552623271942, + 'learning_rate': 1.18e-05, 'epoch': 0.6} +04/19 [12:08:23] INFO | >> train_qwenlatent.py:487 + Step 2370 | grad_norm_pre_clip=0.5599 | + grad_norm_pre_clip_avg=0.4800 | Metrics: + {'align_loss': 0.021541045978665352, + 'recon_loss': 0.005048113875091076, + 'predict_loss': 0.02237672545015812, + 'aux_loss_decay_weight': 0.5262, + 'grad_norm_pre_clip': 0.5599398016929626, + 'data_time': 0.0008792650187388062, + 'model_time': 1.2293932029861026, + 'grad_norm_pre_clip_avg': 0.4800149977207184, + 'learning_rate': 1.185e-05, 'epoch': 0.6} +04/19 [12:08:35] INFO | >> train_qwenlatent.py:487 + Step 2380 | grad_norm_pre_clip=0.4720 | + grad_norm_pre_clip_avg=0.5461 | Metrics: + {'align_loss': 0.021170318126678467, + 'recon_loss': 0.0032921074889600277, + 'predict_loss': 0.021138055250048637, + 'aux_loss_decay_weight': 0.5242, + 'grad_norm_pre_clip': 0.47196799516677856, + 'data_time': 0.000764345022616908, + 'model_time': 1.1991013549850322, + 'grad_norm_pre_clip_avg': 0.5461483359336853, + 'learning_rate': 1.19e-05, 'epoch': 0.6} +04/19 [12:08:48] INFO | >> train_qwenlatent.py:487 + Step 2390 | grad_norm_pre_clip=0.6311 | + grad_norm_pre_clip_avg=0.5553 | Metrics: + {'align_loss': 0.02231525629758835, + 'recon_loss': 0.0028662618715316057, + 'predict_loss': 0.028696229681372643, + 'aux_loss_decay_weight': 0.5222, + 'grad_norm_pre_clip': 0.631069004535675, + 'data_time': 0.0008952229982241988, + 'model_time': 1.2391304110060446, + 'grad_norm_pre_clip_avg': 0.5552801042795181, + 'learning_rate': 1.195e-05, 'epoch': 0.6} +04/19 [12:09:01] INFO | >> train_qwenlatent.py:487 + Step 2400 | grad_norm_pre_clip=0.5355 | + grad_norm_pre_clip_avg=0.6005 | Metrics: + {'align_loss': 0.02259889803826809, + 'recon_loss': 0.0075826155953109264, + 'predict_loss': 0.027946962043642998, + 'aux_loss_decay_weight': 0.5202, + 'grad_norm_pre_clip': 0.5355353951454163, + 'mae_score': 0.049449638847832204, 'data_time': + 0.0006671070004813373, 'model_time': + 1.260878600005526, 'grad_norm_pre_clip_avg': + 0.6005320906639099, 'learning_rate': 1.2e-05, + 'epoch': 0.61} +04/19 [12:09:14] INFO | >> train_qwenlatent.py:487 + Step 2410 | grad_norm_pre_clip=0.4618 | + grad_norm_pre_clip_avg=0.5079 | Metrics: + {'align_loss': 0.02395554445683956, + 'recon_loss': 0.006865753326565027, + 'predict_loss': 0.03249354287981987, + 'aux_loss_decay_weight': 0.5182, + 'grad_norm_pre_clip': 0.461776465177536, + 'data_time': 0.0007849799876566976, + 'model_time': 1.235561678011436, + 'grad_norm_pre_clip_avg': 0.5079412817955017, + 'learning_rate': 1.205e-05, 'epoch': 0.61} +04/19 [12:09:26] INFO | >> train_qwenlatent.py:487 + Step 2420 | grad_norm_pre_clip=0.3966 | + grad_norm_pre_clip_avg=0.4587 | Metrics: + {'align_loss': 0.024075141176581383, + 'recon_loss': 0.0035095829516649246, + 'predict_loss': 0.023570599034428596, + 'aux_loss_decay_weight': 0.5162, + 'grad_norm_pre_clip': 0.3966079354286194, + 'data_time': 0.0007055460009723902, + 'model_time': 1.2236179420142435, + 'grad_norm_pre_clip_avg': 0.45871911346912386, + 'learning_rate': 1.2100000000000001e-05, + 'epoch': 0.61} +04/19 [12:09:39] INFO | >> train_qwenlatent.py:487 + Step 2430 | grad_norm_pre_clip=0.5333 | + grad_norm_pre_clip_avg=0.4374 | Metrics: + {'align_loss': 0.023294519633054733, + 'recon_loss': 0.008910870179533958, + 'predict_loss': 0.036238688975572586, + 'aux_loss_decay_weight': 0.5142, + 'grad_norm_pre_clip': 0.5333161354064941, + 'data_time': 0.000741293013561517, + 'model_time': 1.242449121986283, + 'grad_norm_pre_clip_avg': 0.4374158650636673, + 'learning_rate': 1.215e-05, 'epoch': 0.61} +04/19 [12:09:51] INFO | >> train_qwenlatent.py:487 + Step 2440 | grad_norm_pre_clip=0.5221 | + grad_norm_pre_clip_avg=0.5079 | Metrics: + {'align_loss': 0.023503122851252556, + 'recon_loss': 0.008658909238874912, + 'predict_loss': 0.03303459659218788, + 'aux_loss_decay_weight': 0.5122, + 'grad_norm_pre_clip': 0.5220721364021301, + 'data_time': 0.0011685820063576102, + 'model_time': 1.214489418984158, + 'grad_norm_pre_clip_avg': 0.507876318693161, + 'learning_rate': 1.22e-05, 'epoch': 0.62} +04/19 [12:10:04] INFO | >> train_qwenlatent.py:487 + Step 2450 | grad_norm_pre_clip=0.4456 | + grad_norm_pre_clip_avg=0.5216 | Metrics: + {'align_loss': 0.0219707228243351, + 'recon_loss': 0.005358286201953888, + 'predict_loss': 0.02313355915248394, + 'aux_loss_decay_weight': 0.5102, + 'grad_norm_pre_clip': 0.4455551207065582, + 'mae_score': 0.051265414126284485, 'data_time': + 0.0008710480178706348, 'model_time': + 1.220343797991518, 'grad_norm_pre_clip_avg': + 0.5215939044952392, 'learning_rate': 1.225e-05, + 'epoch': 0.62} +04/19 [12:10:17] INFO | >> train_qwenlatent.py:487 + Step 2460 | grad_norm_pre_clip=0.4360 | + grad_norm_pre_clip_avg=0.5245 | Metrics: + {'align_loss': 0.021567795425653458, + 'recon_loss': 0.004160854499787092, + 'predict_loss': 0.025665557011961937, + 'aux_loss_decay_weight': 0.5082, + 'grad_norm_pre_clip': 0.4360346794128418, + 'data_time': 0.0006594770238734782, + 'model_time': 1.196551543980604, + 'grad_norm_pre_clip_avg': 0.5244986355304718, + 'learning_rate': 1.23e-05, 'epoch': 0.62} +04/19 [12:10:30] INFO | >> train_qwenlatent.py:487 + Step 2470 | grad_norm_pre_clip=0.4451 | + grad_norm_pre_clip_avg=0.4603 | Metrics: + {'align_loss': 0.02225012518465519, + 'recon_loss': 0.007131648249924183, + 'predict_loss': 0.030764246359467506, + 'aux_loss_decay_weight': 0.5062, + 'grad_norm_pre_clip': 0.4450555443763733, + 'data_time': 0.0009322069818153977, + 'model_time': 1.2680072630173527, + 'grad_norm_pre_clip_avg': 0.46025772392749786, + 'learning_rate': 1.235e-05, 'epoch': 0.62} +04/19 [12:10:42] INFO | >> train_qwenlatent.py:487 + Step 2480 | grad_norm_pre_clip=0.4975 | + grad_norm_pre_clip_avg=0.4549 | Metrics: + {'align_loss': 0.023751143366098404, + 'recon_loss': 0.00584367336705327, + 'predict_loss': 0.032911598682403564, + 'aux_loss_decay_weight': 0.5042, + 'grad_norm_pre_clip': 0.4975306987762451, + 'data_time': 0.001016144989989698, + 'model_time': 1.2959130849922076, + 'grad_norm_pre_clip_avg': 0.4549427807331085, + 'learning_rate': 1.24e-05, 'epoch': 0.63} +04/19 [12:10:56] INFO | >> train_qwenlatent.py:487 + Step 2490 | grad_norm_pre_clip=0.4207 | + grad_norm_pre_clip_avg=0.5110 | Metrics: + {'align_loss': 0.022642694413661957, + 'recon_loss': 0.006042402237653732, + 'predict_loss': 0.028990723192691803, + 'aux_loss_decay_weight': 0.5022, + 'grad_norm_pre_clip': 0.4207161068916321, + 'data_time': 0.0008170090150088072, + 'model_time': 1.2160835519898683, + 'grad_norm_pre_clip_avg': 0.5109666585922241, + 'learning_rate': 1.2450000000000001e-05, + 'epoch': 0.63} +04/19 [12:11:09] INFO | >> train_qwenlatent.py:487 + Step 2500 | grad_norm_pre_clip=0.5067 | + grad_norm_pre_clip_avg=0.4673 | Metrics: + {'align_loss': 0.0228237546980381, + 'recon_loss': 0.009628472849726677, + 'predict_loss': 0.03045671619474888, + 'aux_loss_decay_weight': 0.5002, + 'grad_norm_pre_clip': 0.5067243576049805, + 'mae_score': 0.04735113607870566, 'data_time': + 0.0008381890074815601, 'model_time': + 1.2899995290208608, 'grad_norm_pre_clip_avg': + 0.46731367111206057, 'learning_rate': 1.25e-05, + 'epoch': 0.63} +04/19 [12:11:22] INFO | >> train_qwenlatent.py:487 + Step 2510 | grad_norm_pre_clip=0.4764 | + grad_norm_pre_clip_avg=0.4814 | Metrics: + {'align_loss': 0.024387620389461517, + 'recon_loss': 0.006234440486878157, + 'predict_loss': 0.030702879652380943, + 'aux_loss_decay_weight': 0.4982, + 'grad_norm_pre_clip': 0.47635743021965027, + 'data_time': 0.0007870019762776792, + 'model_time': 1.2028348270105198, + 'grad_norm_pre_clip_avg': 0.48140403926372527, + 'learning_rate': 1.255e-05, 'epoch': 0.63} +04/19 [12:11:35] INFO | >> train_qwenlatent.py:487 + Step 2520 | grad_norm_pre_clip=0.5255 | + grad_norm_pre_clip_avg=0.4814 | Metrics: + {'align_loss': 0.022466521710157394, + 'recon_loss': 0.006684187334030867, + 'predict_loss': 0.024269822984933853, + 'aux_loss_decay_weight': 0.4962, + 'grad_norm_pre_clip': 0.5255002975463867, + 'data_time': 0.0008547279867343605, + 'model_time': 1.225724175979849, + 'grad_norm_pre_clip_avg': 0.4813662528991699, + 'learning_rate': 1.2600000000000001e-05, + 'epoch': 0.64} +04/19 [12:11:47] INFO | >> train_qwenlatent.py:487 + Step 2530 | grad_norm_pre_clip=0.5460 | + grad_norm_pre_clip_avg=0.5229 | Metrics: + {'align_loss': 0.022548779845237732, + 'recon_loss': 0.007519632112234831, + 'predict_loss': 0.029641909524798393, + 'aux_loss_decay_weight': 0.4942, + 'grad_norm_pre_clip': 0.5460171699523926, + 'data_time': 0.0008954810036811978, + 'model_time': 1.2031912390084472, + 'grad_norm_pre_clip_avg': 0.5228753924369812, + 'learning_rate': 1.2650000000000001e-05, + 'epoch': 0.64} +04/19 [12:12:00] INFO | >> train_qwenlatent.py:487 + Step 2540 | grad_norm_pre_clip=0.4765 | + grad_norm_pre_clip_avg=0.4790 | Metrics: + {'align_loss': 0.023694757372140884, + 'recon_loss': 0.003353053703904152, + 'predict_loss': 0.026836959645152092, + 'aux_loss_decay_weight': 0.49219999999999997, + 'grad_norm_pre_clip': 0.4765009880065918, + 'data_time': 0.0011447610158938915, + 'model_time': 1.2122917849919759, + 'grad_norm_pre_clip_avg': 0.47899258732795713, + 'learning_rate': 1.27e-05, 'epoch': 0.64} +04/19 [12:12:13] INFO | >> train_qwenlatent.py:487 + Step 2550 | grad_norm_pre_clip=0.4288 | + grad_norm_pre_clip_avg=0.4721 | Metrics: + {'align_loss': 0.02254856750369072, + 'recon_loss': 0.00789694581180811, + 'predict_loss': 0.030033253133296967, + 'aux_loss_decay_weight': 0.49019999999999997, + 'grad_norm_pre_clip': 0.42879340052604675, + 'mae_score': 0.04730336644628026, 'data_time': + 0.0007217960082925856, 'model_time': + 1.2509087389917113, 'grad_norm_pre_clip_avg': + 0.4721000581979752, 'learning_rate': + 1.2750000000000002e-05, 'epoch': 0.64} +04/19 [12:12:26] INFO | >> train_qwenlatent.py:487 + Step 2560 | grad_norm_pre_clip=0.3972 | + grad_norm_pre_clip_avg=0.4396 | Metrics: + {'align_loss': 0.023236561566591263, + 'recon_loss': 0.009336499497294426, + 'predict_loss': 0.027476372197270393, + 'aux_loss_decay_weight': 0.48819999999999997, + 'grad_norm_pre_clip': 0.3972296118736267, + 'data_time': 0.0010312870144844055, + 'model_time': 1.2411982399935368, + 'grad_norm_pre_clip_avg': 0.43957466781139376, + 'learning_rate': 1.2800000000000001e-05, + 'epoch': 0.65} +04/19 [12:12:38] INFO | >> train_qwenlatent.py:487 + Step 2570 | grad_norm_pre_clip=0.5908 | + grad_norm_pre_clip_avg=0.4668 | Metrics: + {'align_loss': 0.022213757038116455, + 'recon_loss': 0.0041247401386499405, + 'predict_loss': 0.02072068862617016, + 'aux_loss_decay_weight': 0.48619999999999997, + 'grad_norm_pre_clip': 0.5907999873161316, + 'data_time': 0.0006531910039484501, + 'model_time': 1.2176574319892097, + 'grad_norm_pre_clip_avg': 0.4668298989534378, + 'learning_rate': 1.285e-05, 'epoch': 0.65} +04/19 [12:12:51] INFO | >> train_qwenlatent.py:487 + Step 2580 | grad_norm_pre_clip=0.5952 | + grad_norm_pre_clip_avg=0.6103 | Metrics: + {'align_loss': 0.02093116007745266, + 'recon_loss': 0.004533027298748493, + 'predict_loss': 0.024426301941275597, + 'aux_loss_decay_weight': 0.48419999999999996, + 'grad_norm_pre_clip': 0.5952350497245789, + 'data_time': 0.0010707739857025445, + 'model_time': 1.281260766001651, + 'grad_norm_pre_clip_avg': 0.6103424787521362, + 'learning_rate': 1.29e-05, 'epoch': 0.65} +04/19 [12:13:03] INFO | >> train_qwenlatent.py:487 + Step 2590 | grad_norm_pre_clip=0.4703 | + grad_norm_pre_clip_avg=0.4850 | Metrics: + {'align_loss': 0.02326710894703865, + 'recon_loss': 0.0059063974767923355, + 'predict_loss': 0.03124045394361019, + 'aux_loss_decay_weight': 0.48219999999999996, + 'grad_norm_pre_clip': 0.4702690541744232, + 'data_time': 0.0009297349897678941, + 'model_time': 1.2166674859763589, + 'grad_norm_pre_clip_avg': 0.48495586812496183, + 'learning_rate': 1.2950000000000001e-05, + 'epoch': 0.65} +04/19 [12:13:16] INFO | >> train_qwenlatent.py:487 + Step 2600 | grad_norm_pre_clip=0.5180 | + grad_norm_pre_clip_avg=0.4815 | Metrics: + {'align_loss': 0.023358874022960663, + 'recon_loss': 0.008443085476756096, + 'predict_loss': 0.03113134391605854, + 'aux_loss_decay_weight': 0.48019999999999996, + 'grad_norm_pre_clip': 0.5180163383483887, + 'mae_score': 0.053512676342113596, 'data_time': + 0.0006532660045195371, 'model_time': + 1.2357680609857198, 'grad_norm_pre_clip_avg': + 0.481492480635643, 'learning_rate': + 1.3000000000000001e-05, 'epoch': 0.66} +04/19 [12:13:29] INFO | >> train_qwenlatent.py:487 + Step 2610 | grad_norm_pre_clip=0.4316 | + grad_norm_pre_clip_avg=0.4415 | Metrics: + {'align_loss': 0.022514652460813522, + 'recon_loss': 0.006786367390304804, + 'predict_loss': 0.03132915869355202, + 'aux_loss_decay_weight': 0.47819999999999996, + 'grad_norm_pre_clip': 0.4315939247608185, + 'data_time': 0.0006873499951325357, + 'model_time': 1.226795538997976, + 'grad_norm_pre_clip_avg': 0.44152488708496096, + 'learning_rate': 1.305e-05, 'epoch': 0.66} +04/19 [12:13:42] INFO | >> train_qwenlatent.py:487 + Step 2620 | grad_norm_pre_clip=0.4975 | + grad_norm_pre_clip_avg=0.5070 | Metrics: + {'align_loss': 0.021801821887493134, + 'recon_loss': 0.007109189406037331, + 'predict_loss': 0.028672993183135986, + 'aux_loss_decay_weight': 0.47619999999999996, + 'grad_norm_pre_clip': 0.4975251853466034, + 'data_time': 0.0007730409852229059, + 'model_time': 1.297476269013714, + 'grad_norm_pre_clip_avg': 0.5070319473743439, + 'learning_rate': 1.3100000000000002e-05, + 'epoch': 0.66} +04/19 [12:13:55] INFO | >> train_qwenlatent.py:487 + Step 2630 | grad_norm_pre_clip=0.5175 | + grad_norm_pre_clip_avg=0.5004 | Metrics: + {'align_loss': 0.021991955116391182, + 'recon_loss': 0.006818814668804407, + 'predict_loss': 0.02549605816602707, + 'aux_loss_decay_weight': 0.47419999999999995, + 'grad_norm_pre_clip': 0.5175405144691467, + 'data_time': 0.0007021890196483582, + 'model_time': 1.9256389579968527, + 'grad_norm_pre_clip_avg': 0.5004155278205872, + 'learning_rate': 1.3150000000000001e-05, + 'epoch': 0.66} +04/19 [12:14:07] INFO | >> train_qwenlatent.py:487 + Step 2640 | grad_norm_pre_clip=0.5223 | + grad_norm_pre_clip_avg=0.5032 | Metrics: + {'align_loss': 0.02111510932445526, + 'recon_loss': 0.005749718751758337, + 'predict_loss': 0.02581646852195263, + 'aux_loss_decay_weight': 0.47219999999999995, + 'grad_norm_pre_clip': 0.522274374961853, + 'data_time': 0.001080723013728857, + 'model_time': 1.288935112010222, + 'grad_norm_pre_clip_avg': 0.5031620144844056, + 'learning_rate': 1.32e-05, 'epoch': 0.67} +04/19 [12:14:21] INFO | >> train_qwenlatent.py:487 + Step 2650 | grad_norm_pre_clip=0.4267 | + grad_norm_pre_clip_avg=0.4691 | Metrics: + {'align_loss': 0.021587127819657326, + 'recon_loss': 0.008857956156134605, + 'predict_loss': 0.028877561911940575, + 'aux_loss_decay_weight': 0.47019999999999995, + 'grad_norm_pre_clip': 0.42673930525779724, + 'mae_score': 0.04745952503101246, 'data_time': + 0.001004423014819622, 'model_time': + 1.3087053979979828, 'grad_norm_pre_clip_avg': + 0.46905443668365476, 'learning_rate': + 1.3250000000000002e-05, 'epoch': 0.67} +04/19 [12:14:34] INFO | >> train_qwenlatent.py:487 + Step 2660 | grad_norm_pre_clip=0.5005 | + grad_norm_pre_clip_avg=0.4912 | Metrics: + {'align_loss': 0.022725369781255722, + 'recon_loss': 0.007851136848330498, + 'predict_loss': 0.033491529524326324, + 'aux_loss_decay_weight': 0.46819999999999995, + 'grad_norm_pre_clip': 0.5005149245262146, + 'data_time': 0.0008194579859264195, + 'model_time': 1.2291131550155114, + 'grad_norm_pre_clip_avg': 0.49123362004756926, + 'learning_rate': 1.3300000000000001e-05, + 'epoch': 0.67} +04/19 [12:14:46] INFO | >> train_qwenlatent.py:487 + Step 2670 | grad_norm_pre_clip=0.5375 | + grad_norm_pre_clip_avg=0.4941 | Metrics: + {'align_loss': 0.02427443489432335, + 'recon_loss': 0.006534750107675791, + 'predict_loss': 0.02661971002817154, + 'aux_loss_decay_weight': 0.46619999999999995, + 'grad_norm_pre_clip': 0.5374569892883301, + 'data_time': 0.000630956026725471, + 'model_time': 1.271368891990278, + 'grad_norm_pre_clip_avg': 0.4940761625766754, + 'learning_rate': 1.3350000000000001e-05, + 'epoch': 0.67} +04/19 [12:14:59] INFO | >> train_qwenlatent.py:487 + Step 2680 | grad_norm_pre_clip=0.4891 | + grad_norm_pre_clip_avg=0.4491 | Metrics: + {'align_loss': 0.02216716669499874, + 'recon_loss': 0.010419000871479511, + 'predict_loss': 0.03158392012119293, + 'aux_loss_decay_weight': 0.46419999999999995, + 'grad_norm_pre_clip': 0.48912620544433594, + 'data_time': 0.0009480680164415389, + 'model_time': 1.2770480460021645, + 'grad_norm_pre_clip_avg': 0.449113142490387, + 'learning_rate': 1.3400000000000002e-05, + 'epoch': 0.68} +04/19 [12:15:11] INFO | >> train_qwenlatent.py:487 + Step 2690 | grad_norm_pre_clip=0.4138 | + grad_norm_pre_clip_avg=0.4288 | Metrics: + {'align_loss': 0.022539060562849045, + 'recon_loss': 0.012055322527885437, + 'predict_loss': 0.03355544060468674, + 'aux_loss_decay_weight': 0.46220000000000006, + 'grad_norm_pre_clip': 0.41381746530532837, + 'data_time': 0.0007911489810794592, + 'model_time': 1.2517334079893772, + 'grad_norm_pre_clip_avg': 0.4287557482719421, + 'learning_rate': 1.3450000000000002e-05, + 'epoch': 0.68} +04/19 [12:15:25] INFO | >> train_qwenlatent.py:487 + Step 2700 | grad_norm_pre_clip=0.5714 | + grad_norm_pre_clip_avg=0.5551 | Metrics: + {'align_loss': 0.022665444761514664, + 'recon_loss': 0.008349800482392311, + 'predict_loss': 0.03132268786430359, + 'aux_loss_decay_weight': 0.46020000000000005, + 'grad_norm_pre_clip': 0.5713870525360107, + 'mae_score': 0.04143636892507742, 'data_time': + 0.0008277939923573285, 'model_time': + 1.2609734600118827, 'grad_norm_pre_clip_avg': + 0.5551097482442856, 'learning_rate': + 1.3500000000000001e-05, 'epoch': 0.68} +04/19 [12:15:37] INFO | >> train_qwenlatent.py:487 + Step 2710 | grad_norm_pre_clip=0.4862 | + grad_norm_pre_clip_avg=0.5344 | Metrics: + {'align_loss': 0.022881705313920975, + 'recon_loss': 0.008411943912506104, + 'predict_loss': 0.028973670676350594, + 'aux_loss_decay_weight': 0.45820000000000005, + 'grad_norm_pre_clip': 0.4862396717071533, + 'data_time': 0.0007690140046179295, + 'model_time': 1.2236966459895484, + 'grad_norm_pre_clip_avg': 0.5343507081270218, + 'learning_rate': 1.3550000000000002e-05, + 'epoch': 0.68} +04/19 [12:15:50] INFO | >> train_qwenlatent.py:487 + Step 2720 | grad_norm_pre_clip=0.4475 | + grad_norm_pre_clip_avg=0.4390 | Metrics: + {'align_loss': 0.022836515679955482, + 'recon_loss': 0.01160468440502882, + 'predict_loss': 0.03173225000500679, + 'aux_loss_decay_weight': 0.45620000000000005, + 'grad_norm_pre_clip': 0.4474870264530182, + 'data_time': 0.0008875670027919114, + 'model_time': 1.2707348389958497, + 'grad_norm_pre_clip_avg': 0.438972544670105, + 'learning_rate': 1.3600000000000002e-05, + 'epoch': 0.69} +04/19 [12:16:02] INFO | >> train_qwenlatent.py:487 + Step 2730 | grad_norm_pre_clip=0.4632 | + grad_norm_pre_clip_avg=0.4353 | Metrics: + {'align_loss': 0.022035041823983192, + 'recon_loss': 0.008424109779298306, + 'predict_loss': 0.029377128928899765, + 'aux_loss_decay_weight': 0.45420000000000005, + 'grad_norm_pre_clip': 0.4631894528865814, + 'data_time': 0.0007039360061753541, + 'model_time': 1.2406217439856846, + 'grad_norm_pre_clip_avg': 0.4353332042694092, + 'learning_rate': 1.3650000000000001e-05, + 'epoch': 0.69} +04/19 [12:16:14] INFO | >> train_qwenlatent.py:487 + Step 2740 | grad_norm_pre_clip=0.5275 | + grad_norm_pre_clip_avg=0.4770 | Metrics: + {'align_loss': 0.021845586597919464, + 'recon_loss': 0.005683865863829851, + 'predict_loss': 0.023086555302143097, + 'aux_loss_decay_weight': 0.45220000000000005, + 'grad_norm_pre_clip': 0.527466893196106, + 'data_time': 0.000940444995649159, + 'model_time': 1.2533200370089617, + 'grad_norm_pre_clip_avg': 0.47695334255695343, + 'learning_rate': 1.3700000000000001e-05, + 'epoch': 0.69} +04/19 [12:16:28] INFO | >> train_qwenlatent.py:487 + Step 2750 | grad_norm_pre_clip=0.4127 | + grad_norm_pre_clip_avg=0.4798 | Metrics: + {'align_loss': 0.02358417399227619, + 'recon_loss': 0.005283354315906763, + 'predict_loss': 0.02432304248213768, + 'aux_loss_decay_weight': 0.45020000000000004, + 'grad_norm_pre_clip': 0.41268816590309143, + 'mae_score': 0.06267094655079884, 'data_time': + 0.0010207539889961481, 'model_time': + 1.50449734198628, 'grad_norm_pre_clip_avg': + 0.47981433272361756, 'learning_rate': + 1.3750000000000002e-05, 'epoch': 0.69} +04/19 [12:16:40] INFO | >> train_qwenlatent.py:487 + Step 2760 | grad_norm_pre_clip=0.4194 | + grad_norm_pre_clip_avg=0.3999 | Metrics: + {'align_loss': 0.020950354635715485, + 'recon_loss': 0.005768191069364548, + 'predict_loss': 0.02307041361927986, + 'aux_loss_decay_weight': 0.44820000000000004, + 'grad_norm_pre_clip': 0.4193749725818634, + 'data_time': 0.0008851980092003942, + 'model_time': 1.2183777950122021, + 'grad_norm_pre_clip_avg': 0.3999106228351593, + 'learning_rate': 1.3800000000000002e-05, + 'epoch': 0.7} +04/19 [12:16:53] INFO | >> train_qwenlatent.py:487 + Step 2770 | grad_norm_pre_clip=0.3860 | + grad_norm_pre_clip_avg=0.4398 | Metrics: + {'align_loss': 0.02310609072446823, + 'recon_loss': 0.011860590428113937, + 'predict_loss': 0.03533318638801575, + 'aux_loss_decay_weight': 0.44620000000000004, + 'grad_norm_pre_clip': 0.38602253794670105, + 'data_time': 0.0005937870009802282, + 'model_time': 1.328664876986295, + 'grad_norm_pre_clip_avg': 0.4397661745548248, + 'learning_rate': 1.3850000000000001e-05, + 'epoch': 0.7} +04/19 [12:17:06] INFO | >> train_qwenlatent.py:487 + Step 2780 | grad_norm_pre_clip=0.6360 | + grad_norm_pre_clip_avg=0.4704 | Metrics: + {'align_loss': 0.021214451640844345, + 'recon_loss': 0.005763179622590542, + 'predict_loss': 0.02546125464141369, + 'aux_loss_decay_weight': 0.44420000000000004, + 'grad_norm_pre_clip': 0.6360288262367249, + 'data_time': 0.0009586160012986511, + 'model_time': 1.2455013819853775, + 'grad_norm_pre_clip_avg': 0.4703793883323669, + 'learning_rate': 1.3900000000000002e-05, + 'epoch': 0.7} +04/19 [12:17:19] INFO | >> train_qwenlatent.py:487 + Step 2790 | grad_norm_pre_clip=0.6391 | + grad_norm_pre_clip_avg=0.5114 | Metrics: + {'align_loss': 0.024231694638729095, + 'recon_loss': 0.004203928634524345, + 'predict_loss': 0.021332811564207077, + 'aux_loss_decay_weight': 0.44220000000000004, + 'grad_norm_pre_clip': 0.639125406742096, + 'data_time': 0.0008847950084600598, + 'model_time': 1.512235452013556, + 'grad_norm_pre_clip_avg': 0.5113781839609146, + 'learning_rate': 1.3950000000000002e-05, + 'epoch': 0.7} +04/19 [12:17:32] INFO | >> train_qwenlatent.py:487 + Step 2800 | grad_norm_pre_clip=0.5061 | + grad_norm_pre_clip_avg=0.4872 | Metrics: + {'align_loss': 0.023823224008083344, + 'recon_loss': 0.004090809263288975, + 'predict_loss': 0.022616863250732422, + 'aux_loss_decay_weight': 0.44020000000000004, + 'grad_norm_pre_clip': 0.506128191947937, + 'mae_score': 0.03988251213554864, 'data_time': + 0.0007500170031562448, 'model_time': + 1.2175797710078768, 'grad_norm_pre_clip_avg': + 0.4871541976928711, 'learning_rate': + 1.4000000000000001e-05, 'epoch': 0.71} +04/19 [12:17:45] INFO | >> train_qwenlatent.py:487 + Step 2810 | grad_norm_pre_clip=0.4791 | + grad_norm_pre_clip_avg=0.4451 | Metrics: + {'align_loss': 0.02299700677394867, + 'recon_loss': 0.006456953007727861, + 'predict_loss': 0.0272924043238163, + 'aux_loss_decay_weight': 0.43820000000000003, + 'grad_norm_pre_clip': 0.4791261851787567, + 'data_time': 0.0007503160159103572, + 'model_time': 1.2334308149875142, + 'grad_norm_pre_clip_avg': 0.44511272609233854, + 'learning_rate': 1.4050000000000003e-05, + 'epoch': 0.71} +04/19 [12:17:57] INFO | >> train_qwenlatent.py:487 + Step 2820 | grad_norm_pre_clip=0.6134 | + grad_norm_pre_clip_avg=0.4418 | Metrics: + {'align_loss': 0.022638950496912003, + 'recon_loss': 0.0032818231265991926, + 'predict_loss': 0.02667088247835636, + 'aux_loss_decay_weight': 0.43620000000000003, + 'grad_norm_pre_clip': 0.6133967041969299, + 'data_time': 0.0006527919904328883, + 'model_time': 1.2315574420208577, + 'grad_norm_pre_clip_avg': 0.44179599583148954, + 'learning_rate': 1.4099999999999999e-05, + 'epoch': 0.71} +04/19 [12:18:10] INFO | >> train_qwenlatent.py:487 + Step 2830 | grad_norm_pre_clip=0.4681 | + grad_norm_pre_clip_avg=0.4863 | Metrics: + {'align_loss': 0.023433981463313103, + 'recon_loss': 0.006644102279096842, + 'predict_loss': 0.022096967324614525, + 'aux_loss_decay_weight': 0.43420000000000003, + 'grad_norm_pre_clip': 0.4681064188480377, + 'data_time': 0.0010279279958922416, + 'model_time': 1.189067363011418, + 'grad_norm_pre_clip_avg': 0.48628937900066377, + 'learning_rate': 1.415e-05, 'epoch': 0.71} +04/19 [12:18:23] INFO | >> train_qwenlatent.py:487 + Step 2840 | grad_norm_pre_clip=0.4379 | + grad_norm_pre_clip_avg=0.6071 | Metrics: + {'align_loss': 0.020394403487443924, + 'recon_loss': 0.0030040363781154156, + 'predict_loss': 0.01975281722843647, + 'aux_loss_decay_weight': 0.43220000000000003, + 'grad_norm_pre_clip': 0.4379248321056366, + 'data_time': 0.0007044950034469366, + 'model_time': 1.1946014259883668, + 'grad_norm_pre_clip_avg': 0.6071415930986405, + 'learning_rate': 1.42e-05, 'epoch': 0.72} +04/19 [12:18:36] INFO | >> train_qwenlatent.py:487 + Step 2850 | grad_norm_pre_clip=0.5020 | + grad_norm_pre_clip_avg=0.4742 | Metrics: + {'align_loss': 0.02269085869193077, + 'recon_loss': 0.0063640158623456955, + 'predict_loss': 0.031858235597610474, + 'aux_loss_decay_weight': 0.4302, + 'grad_norm_pre_clip': 0.5019991397857666, + 'mae_score': 0.06371483330254082, 'data_time': + 0.001019319985061884, 'model_time': + 1.279918375017587, 'grad_norm_pre_clip_avg': + 0.47418826520442964, 'learning_rate': + 1.4249999999999999e-05, 'epoch': 0.72} +04/19 [12:18:49] INFO | >> train_qwenlatent.py:487 + Step 2860 | grad_norm_pre_clip=0.3833 | + grad_norm_pre_clip_avg=0.4355 | Metrics: + {'align_loss': 0.024752728641033173, + 'recon_loss': 0.0029055457562208176, + 'predict_loss': 0.021950380876660347, + 'aux_loss_decay_weight': 0.4282, + 'grad_norm_pre_clip': 0.38330715894699097, + 'data_time': 0.0010657849779818207, + 'model_time': 1.2675186459964607, + 'grad_norm_pre_clip_avg': 0.43546104729175567, + 'learning_rate': 1.43e-05, 'epoch': 0.72} +04/19 [12:19:01] INFO | >> train_qwenlatent.py:487 + Step 2870 | grad_norm_pre_clip=0.4453 | + grad_norm_pre_clip_avg=0.4437 | Metrics: + {'align_loss': 0.021886151283979416, + 'recon_loss': 0.006232793442904949, + 'predict_loss': 0.025316551327705383, + 'aux_loss_decay_weight': 0.4262, + 'grad_norm_pre_clip': 0.44527316093444824, + 'data_time': 0.0007293139933608472, + 'model_time': 1.24655786799849, + 'grad_norm_pre_clip_avg': 0.44368302524089814, + 'learning_rate': 1.435e-05, 'epoch': 0.72} +04/19 [12:19:14] INFO | >> train_qwenlatent.py:487 + Step 2880 | grad_norm_pre_clip=0.4830 | + grad_norm_pre_clip_avg=0.4740 | Metrics: + {'align_loss': 0.022558733820915222, + 'recon_loss': 0.0055679320357739925, + 'predict_loss': 0.0224284864962101, + 'aux_loss_decay_weight': 0.4242, + 'grad_norm_pre_clip': 0.4830467998981476, + 'data_time': 0.0007307049818336964, + 'model_time': 1.2449698289856315, + 'grad_norm_pre_clip_avg': 0.4740248590707779, + 'learning_rate': 1.44e-05, 'epoch': 0.73} +04/19 [12:19:26] INFO | >> train_qwenlatent.py:487 + Step 2890 | grad_norm_pre_clip=0.5183 | + grad_norm_pre_clip_avg=0.4789 | Metrics: + {'align_loss': 0.02318595163524151, + 'recon_loss': 0.006967324297875166, + 'predict_loss': 0.025025740265846252, + 'aux_loss_decay_weight': 0.4222, + 'grad_norm_pre_clip': 0.5182593464851379, + 'data_time': 0.000991492997854948, + 'model_time': 1.1831622800091282, + 'grad_norm_pre_clip_avg': 0.4788615971803665, + 'learning_rate': 1.4449999999999999e-05, + 'epoch': 0.73} +04/19 [12:19:40] INFO | >> train_qwenlatent.py:487 + Step 2900 | grad_norm_pre_clip=0.3439 | + grad_norm_pre_clip_avg=0.4985 | Metrics: + {'align_loss': 0.02030256763100624, + 'recon_loss': 0.0022498061880469322, + 'predict_loss': 0.020398801192641258, + 'aux_loss_decay_weight': 0.4202, + 'grad_norm_pre_clip': 0.3438870310783386, + 'mae_score': 0.03968472867398649, 'data_time': + 0.0007826430082786828, 'model_time': + 1.2584615620144177, 'grad_norm_pre_clip_avg': + 0.49852277636528014, 'learning_rate': 1.45e-05, + 'epoch': 0.73} +04/19 [12:19:53] INFO | >> train_qwenlatent.py:487 + Step 2910 | grad_norm_pre_clip=0.3336 | + grad_norm_pre_clip_avg=0.4285 | Metrics: + {'align_loss': 0.021427344530820847, + 'recon_loss': 0.0047143129631876945, + 'predict_loss': 0.024578912183642387, + 'aux_loss_decay_weight': 0.4182, + 'grad_norm_pre_clip': 0.33362993597984314, + 'data_time': 0.0007173349731601775, + 'model_time': 1.5225918129726779, + 'grad_norm_pre_clip_avg': 0.42846659719944, + 'learning_rate': 1.455e-05, 'epoch': 0.73} +04/19 [12:20:06] INFO | >> train_qwenlatent.py:487 + Step 2920 | grad_norm_pre_clip=0.3984 | + grad_norm_pre_clip_avg=0.4122 | Metrics: + {'align_loss': 0.021854577586054802, + 'recon_loss': 0.0022514192387461662, + 'predict_loss': 0.018561547622084618, + 'aux_loss_decay_weight': 0.4162, + 'grad_norm_pre_clip': 0.398373544216156, + 'data_time': 0.000916348973987624, + 'model_time': 1.2248646930092946, + 'grad_norm_pre_clip_avg': 0.4122223734855652, + 'learning_rate': 1.4599999999999999e-05, + 'epoch': 0.74} +04/19 [12:20:19] INFO | >> train_qwenlatent.py:487 + Step 2930 | grad_norm_pre_clip=0.4690 | + grad_norm_pre_clip_avg=0.4402 | Metrics: + {'align_loss': 0.02233991213142872, + 'recon_loss': 0.006121118552982807, + 'predict_loss': 0.02429376356303692, + 'aux_loss_decay_weight': 0.4142, + 'grad_norm_pre_clip': 0.46904927492141724, + 'data_time': 0.00098143401555717, 'model_time': + 1.2374143989873119, 'grad_norm_pre_clip_avg': + 0.44015806913375854, 'learning_rate': + 1.465e-05, 'epoch': 0.74} +04/19 [12:20:31] INFO | >> train_qwenlatent.py:487 + Step 2940 | grad_norm_pre_clip=0.4502 | + grad_norm_pre_clip_avg=0.4687 | Metrics: + {'align_loss': 0.022729070857167244, + 'recon_loss': 0.006969499867409468, + 'predict_loss': 0.03276780992746353, + 'aux_loss_decay_weight': 0.4122, + 'grad_norm_pre_clip': 0.4501882493495941, + 'data_time': 0.0009017039847094566, + 'model_time': 1.2022238519857638, + 'grad_norm_pre_clip_avg': 0.46874966025352477, + 'learning_rate': 1.47e-05, 'epoch': 0.74} +04/19 [12:20:45] INFO | >> train_qwenlatent.py:487 + Step 2950 | grad_norm_pre_clip=0.4682 | + grad_norm_pre_clip_avg=0.4622 | Metrics: + {'align_loss': 0.023357082158327103, + 'recon_loss': 0.004173000808805227, + 'predict_loss': 0.02479861117899418, + 'aux_loss_decay_weight': 0.4102, + 'grad_norm_pre_clip': 0.468154639005661, + 'mae_score': 0.05485230179520341, 'data_time': + 0.0008678379817865789, 'model_time': + 1.224324054986937, 'grad_norm_pre_clip_avg': + 0.46220108270645144, 'learning_rate': + 1.475e-05, 'epoch': 0.74} +04/19 [12:20:57] INFO | >> train_qwenlatent.py:487 + Step 2960 | grad_norm_pre_clip=0.5186 | + grad_norm_pre_clip_avg=0.4713 | Metrics: + {'align_loss': 0.021034499630331993, + 'recon_loss': 0.009876311756670475, + 'predict_loss': 0.0284146536141634, + 'aux_loss_decay_weight': 0.4082, + 'grad_norm_pre_clip': 0.5185683369636536, + 'data_time': 0.0008660299936309457, + 'model_time': 1.1998407329956535, + 'grad_norm_pre_clip_avg': 0.4712677448987961, + 'learning_rate': 1.48e-05, 'epoch': 0.75} +04/19 [12:21:10] INFO | >> train_qwenlatent.py:487 + Step 2970 | grad_norm_pre_clip=0.5143 | + grad_norm_pre_clip_avg=0.4400 | Metrics: + {'align_loss': 0.022514862939715385, + 'recon_loss': 0.0057471818290650845, + 'predict_loss': 0.028975093737244606, + 'aux_loss_decay_weight': 0.4062, + 'grad_norm_pre_clip': 0.514304518699646, + 'data_time': 0.0009136750013567507, + 'model_time': 1.2279739899968263, + 'grad_norm_pre_clip_avg': 0.43995734453201296, + 'learning_rate': 1.485e-05, 'epoch': 0.75} +04/19 [12:21:22] INFO | >> train_qwenlatent.py:487 + Step 2980 | grad_norm_pre_clip=0.3949 | + grad_norm_pre_clip_avg=0.3956 | Metrics: + {'align_loss': 0.02245432138442993, + 'recon_loss': 0.008646263740956783, + 'predict_loss': 0.03300059586763382, + 'aux_loss_decay_weight': 0.4042, + 'grad_norm_pre_clip': 0.39489641785621643, + 'data_time': 0.0009793770150281489, + 'model_time': 1.281084276997717, + 'grad_norm_pre_clip_avg': 0.39564888179302216, + 'learning_rate': 1.49e-05, 'epoch': 0.75} +04/19 [12:21:35] INFO | >> train_qwenlatent.py:487 + Step 2990 | grad_norm_pre_clip=0.4437 | + grad_norm_pre_clip_avg=0.4458 | Metrics: + {'align_loss': 0.02212674356997013, + 'recon_loss': 0.005584979895502329, + 'predict_loss': 0.02305859886109829, + 'aux_loss_decay_weight': 0.4022, + 'grad_norm_pre_clip': 0.4437406361103058, + 'data_time': 0.0006718579970765859, + 'model_time': 1.2148990550194867, + 'grad_norm_pre_clip_avg': 0.44575353562831876, + 'learning_rate': 1.4950000000000001e-05, + 'epoch': 0.75} +04/19 [12:21:48] INFO | >> train_qwenlatent.py:487 + Step 3000 | grad_norm_pre_clip=0.4730 | + grad_norm_pre_clip_avg=0.4630 | Metrics: + {'align_loss': 0.02217477187514305, + 'recon_loss': 0.009903882630169392, + 'predict_loss': 0.029278874397277832, + 'aux_loss_decay_weight': 0.4002, + 'grad_norm_pre_clip': 0.472966730594635, + 'mae_score': 0.047720350660719314, 'data_time': + 0.0009257580095436424, 'model_time': + 1.224971232993994, 'grad_norm_pre_clip_avg': + 0.4630006641149521, 'learning_rate': 1.5e-05, + 'epoch': 0.76} +04/19 [12:22:01] INFO | >> train_qwenlatent.py:487 + Step 3010 | grad_norm_pre_clip=0.5156 | + grad_norm_pre_clip_avg=0.4857 | Metrics: + {'align_loss': 0.022624094039201736, + 'recon_loss': 0.007394453976303339, + 'predict_loss': 0.025506259873509407, + 'aux_loss_decay_weight': 0.3982, + 'grad_norm_pre_clip': 0.5155984163284302, + 'data_time': 0.0006790729821659625, + 'model_time': 1.2502958950062748, + 'grad_norm_pre_clip_avg': 0.4856665521860123, + 'learning_rate': 1.505e-05, 'epoch': 0.76} +04/19 [12:22:14] INFO | >> train_qwenlatent.py:487 + Step 3020 | grad_norm_pre_clip=0.3957 | + grad_norm_pre_clip_avg=0.4541 | Metrics: + {'align_loss': 0.020217299461364746, + 'recon_loss': 0.008959189057350159, + 'predict_loss': 0.025791723281145096, + 'aux_loss_decay_weight': 0.3962, + 'grad_norm_pre_clip': 0.3956630825996399, + 'data_time': 0.0009615019953344017, + 'model_time': 1.2207887009717524, + 'grad_norm_pre_clip_avg': 0.4540706932544708, + 'learning_rate': 1.51e-05, 'epoch': 0.76} +04/19 [12:22:26] INFO | >> train_qwenlatent.py:487 + Step 3030 | grad_norm_pre_clip=0.5391 | + grad_norm_pre_clip_avg=0.4196 | Metrics: + {'align_loss': 0.023607872426509857, + 'recon_loss': 0.009293154813349247, + 'predict_loss': 0.028955258429050446, + 'aux_loss_decay_weight': 0.3942, + 'grad_norm_pre_clip': 0.5390587449073792, + 'data_time': 0.0008261119946837425, + 'model_time': 1.195272559998557, + 'grad_norm_pre_clip_avg': 0.4195943295955658, + 'learning_rate': 1.515e-05, 'epoch': 0.76} +04/19 [12:22:39] INFO | >> train_qwenlatent.py:487 + Step 3040 | grad_norm_pre_clip=0.3654 | + grad_norm_pre_clip_avg=0.4222 | Metrics: + {'align_loss': 0.022829007357358932, + 'recon_loss': 0.005000451114028692, + 'predict_loss': 0.02625305950641632, + 'aux_loss_decay_weight': 0.3922, + 'grad_norm_pre_clip': 0.3654150068759918, + 'data_time': 0.0009211270080413669, + 'model_time': 1.2186341189953964, + 'grad_norm_pre_clip_avg': 0.42216578423976897, + 'learning_rate': 1.52e-05, 'epoch': 0.77} +04/19 [12:22:53] INFO | >> train_qwenlatent.py:487 + Step 3050 | grad_norm_pre_clip=0.4387 | + grad_norm_pre_clip_avg=0.4458 | Metrics: + {'align_loss': 0.023067213594913483, + 'recon_loss': 0.010978704318404198, + 'predict_loss': 0.03251111879944801, + 'aux_loss_decay_weight': 0.3902, + 'grad_norm_pre_clip': 0.43870294094085693, + 'mae_score': 0.05130816279230891, 'data_time': + 0.0014856999914627522, 'model_time': + 1.2243701650004368, 'grad_norm_pre_clip_avg': + 0.44584369361400605, 'learning_rate': + 1.525e-05, 'epoch': 0.77} +04/19 [12:23:06] INFO | >> train_qwenlatent.py:487 + Step 3060 | grad_norm_pre_clip=0.4038 | + grad_norm_pre_clip_avg=0.4917 | Metrics: + {'align_loss': 0.021873192861676216, + 'recon_loss': 0.005091133527457714, + 'predict_loss': 0.02480669505894184, + 'aux_loss_decay_weight': 0.3882, + 'grad_norm_pre_clip': 0.40382319688796997, + 'data_time': 0.0007178540108725429, + 'model_time': 1.3341990320186596, + 'grad_norm_pre_clip_avg': 0.49165038764476776, + 'learning_rate': 1.53e-05, 'epoch': 0.77} +04/19 [12:23:19] INFO | >> train_qwenlatent.py:487 + Step 3070 | grad_norm_pre_clip=0.4370 | + grad_norm_pre_clip_avg=0.4460 | Metrics: + {'align_loss': 0.0221843384206295, + 'recon_loss': 0.0030490702483803034, + 'predict_loss': 0.02165091410279274, + 'aux_loss_decay_weight': 0.3862, + 'grad_norm_pre_clip': 0.4369957149028778, + 'data_time': 0.0008171770023182034, + 'model_time': 1.2628807189757936, + 'grad_norm_pre_clip_avg': 0.4460420489311218, + 'learning_rate': 1.535e-05, 'epoch': 0.77} +04/19 [12:23:31] INFO | >> train_qwenlatent.py:487 + Step 3080 | grad_norm_pre_clip=0.3492 | + grad_norm_pre_clip_avg=0.4463 | Metrics: + {'align_loss': 0.022417698055505753, + 'recon_loss': 0.003124992595985532, + 'predict_loss': 0.020211152732372284, + 'aux_loss_decay_weight': 0.3842, + 'grad_norm_pre_clip': 0.3492319881916046, + 'data_time': 0.0009461680019740015, + 'model_time': 1.2226585939934012, + 'grad_norm_pre_clip_avg': 0.44628351628780366, + 'learning_rate': 1.54e-05, 'epoch': 0.78} +04/19 [12:23:44] INFO | >> train_qwenlatent.py:487 + Step 3090 | grad_norm_pre_clip=0.4421 | + grad_norm_pre_clip_avg=0.4293 | Metrics: + {'align_loss': 0.021975427865982056, + 'recon_loss': 0.006607160437852144, + 'predict_loss': 0.023280426859855652, + 'aux_loss_decay_weight': 0.3822, + 'grad_norm_pre_clip': 0.44212061166763306, + 'data_time': 0.00101890999940224, 'model_time': + 1.2111248319852166, 'grad_norm_pre_clip_avg': + 0.4292514085769653, 'learning_rate': 1.545e-05, + 'epoch': 0.78} +04/19 [12:23:57] INFO | >> train_qwenlatent.py:487 + Step 3100 | grad_norm_pre_clip=0.4015 | + grad_norm_pre_clip_avg=0.4238 | Metrics: + {'align_loss': 0.02273520827293396, + 'recon_loss': 0.005164488684386015, + 'predict_loss': 0.02731495536863804, + 'aux_loss_decay_weight': 0.3802, + 'grad_norm_pre_clip': 0.40150612592697144, + 'mae_score': 0.0367786270004135, 'data_time': + 0.0008250690007116646, 'model_time': + 1.2985500000067987, 'grad_norm_pre_clip_avg': + 0.42380703389644625, 'learning_rate': 1.55e-05, + 'epoch': 0.78} +04/19 [12:24:10] INFO | >> train_qwenlatent.py:487 + Step 3110 | grad_norm_pre_clip=0.4151 | + grad_norm_pre_clip_avg=0.4433 | Metrics: + {'align_loss': 0.022996393963694572, + 'recon_loss': 0.012184699065983295, + 'predict_loss': 0.029582353308796883, + 'aux_loss_decay_weight': 0.3782, + 'grad_norm_pre_clip': 0.41507086157798767, + 'data_time': 0.0007233780052047223, + 'model_time': 1.5521866040071473, + 'grad_norm_pre_clip_avg': 0.44332167506217957, + 'learning_rate': 1.5550000000000002e-05, + 'epoch': 0.78} +04/19 [12:24:22] INFO | >> train_qwenlatent.py:487 + Step 3120 | grad_norm_pre_clip=0.5342 | + grad_norm_pre_clip_avg=0.4931 | Metrics: + {'align_loss': 0.022011730819940567, + 'recon_loss': 0.005841131787747145, + 'predict_loss': 0.027292314916849136, + 'aux_loss_decay_weight': 0.3762, + 'grad_norm_pre_clip': 0.5342400670051575, + 'data_time': 0.0010086690017487854, + 'model_time': 1.2791238099744078, + 'grad_norm_pre_clip_avg': 0.4930713027715683, + 'learning_rate': 1.56e-05, 'epoch': 0.79} +04/19 [12:24:35] INFO | >> train_qwenlatent.py:487 + Step 3130 | grad_norm_pre_clip=0.4350 | + grad_norm_pre_clip_avg=0.4463 | Metrics: + {'align_loss': 0.022350545972585678, + 'recon_loss': 0.005593141075223684, + 'predict_loss': 0.028747305274009705, + 'aux_loss_decay_weight': 0.3742, + 'grad_norm_pre_clip': 0.43500006198883057, + 'data_time': 0.0010106319969054312, + 'model_time': 1.1949946439999621, + 'grad_norm_pre_clip_avg': 0.4462692946195602, + 'learning_rate': 1.565e-05, 'epoch': 0.79} +04/19 [12:24:47] INFO | >> train_qwenlatent.py:487 + Step 3140 | grad_norm_pre_clip=0.4899 | + grad_norm_pre_clip_avg=0.4382 | Metrics: + {'align_loss': 0.02007077820599079, + 'recon_loss': 0.004820556845515966, + 'predict_loss': 0.022108837962150574, + 'aux_loss_decay_weight': 0.3722, + 'grad_norm_pre_clip': 0.4899367094039917, + 'data_time': 0.0009612920111976564, + 'model_time': 1.242443500988884, + 'grad_norm_pre_clip_avg': 0.4381918400526047, + 'learning_rate': 1.5700000000000002e-05, + 'epoch': 0.79} +04/19 [12:25:00] INFO | >> train_qwenlatent.py:487 + Step 3150 | grad_norm_pre_clip=0.4738 | + grad_norm_pre_clip_avg=0.4841 | Metrics: + {'align_loss': 0.02030102163553238, + 'recon_loss': 0.003040524199604988, + 'predict_loss': 0.020691197365522385, + 'aux_loss_decay_weight': 0.3702, + 'grad_norm_pre_clip': 0.4737943708896637, + 'mae_score': 0.03860007277480117, 'data_time': + 0.0007242189894896001, 'model_time': + 1.2630658310081344, 'grad_norm_pre_clip_avg': + 0.4841162234544754, 'learning_rate': 1.575e-05, + 'epoch': 0.79} +04/19 [12:25:13] INFO | >> train_qwenlatent.py:487 + Step 3160 | grad_norm_pre_clip=0.3730 | + grad_norm_pre_clip_avg=0.4283 | Metrics: + {'align_loss': 0.023093167692422867, + 'recon_loss': 0.006985703948885202, + 'predict_loss': 0.025907011702656746, + 'aux_loss_decay_weight': 0.36819999999999997, + 'grad_norm_pre_clip': 0.3729563355445862, + 'data_time': 0.0007618330128025264, + 'model_time': 1.2820078859804198, + 'grad_norm_pre_clip_avg': 0.4282537251710892, + 'learning_rate': 1.58e-05, 'epoch': 0.8} +04/19 [12:25:26] INFO | >> train_qwenlatent.py:487 + Step 3170 | grad_norm_pre_clip=0.4000 | + grad_norm_pre_clip_avg=0.4231 | Metrics: + {'align_loss': 0.023521337658166885, + 'recon_loss': 0.009163983166217804, + 'predict_loss': 0.029526911675930023, + 'aux_loss_decay_weight': 0.36619999999999997, + 'grad_norm_pre_clip': 0.3999544084072113, + 'data_time': 0.0006580399931408465, + 'model_time': 1.2379345099907368, + 'grad_norm_pre_clip_avg': 0.42308959662914275, + 'learning_rate': 1.5850000000000002e-05, + 'epoch': 0.8} +04/19 [12:25:39] INFO | >> train_qwenlatent.py:487 + Step 3180 | grad_norm_pre_clip=0.7858 | + grad_norm_pre_clip_avg=0.8219 | Metrics: + {'align_loss': 0.021172761917114258, + 'recon_loss': 0.006694665178656578, + 'predict_loss': 0.023827314376831055, + 'aux_loss_decay_weight': 0.36419999999999997, + 'grad_norm_pre_clip': 0.78583163022995, + 'data_time': 0.0009384770237375051, + 'model_time': 1.23039080700255, + 'grad_norm_pre_clip_avg': 0.8218811452388763, + 'learning_rate': 1.59e-05, 'epoch': 0.8} +04/19 [12:25:51] INFO | >> train_qwenlatent.py:487 + Step 3190 | grad_norm_pre_clip=0.5017 | + grad_norm_pre_clip_avg=0.5274 | Metrics: + {'align_loss': 0.021911432966589928, + 'recon_loss': 0.005458126775920391, + 'predict_loss': 0.02289130538702011, + 'aux_loss_decay_weight': 0.36219999999999997, + 'grad_norm_pre_clip': 0.5017039775848389, + 'data_time': 0.000924159015994519, + 'model_time': 1.2739740250108298, + 'grad_norm_pre_clip_avg': 0.5273686289787293, + 'learning_rate': 1.595e-05, 'epoch': 0.8} +04/19 [12:26:05] INFO | >> train_qwenlatent.py:487 + Step 3200 | grad_norm_pre_clip=0.4859 | + grad_norm_pre_clip_avg=0.4401 | Metrics: + {'align_loss': 0.021738948300480843, + 'recon_loss': 0.002414121525362134, + 'predict_loss': 0.01670532301068306, + 'aux_loss_decay_weight': 0.36019999999999996, + 'grad_norm_pre_clip': 0.48594582080841064, + 'mae_score': 0.03816171594568201, 'data_time': + 0.0008962139836512506, 'model_time': + 1.2916016209928785, 'grad_norm_pre_clip_avg': + 0.44012596905231477, 'learning_rate': + 1.6000000000000003e-05, 'epoch': 0.81} +04/19 [12:26:18] INFO | >> train_qwenlatent.py:487 + Step 3210 | grad_norm_pre_clip=0.4529 | + grad_norm_pre_clip_avg=0.4322 | Metrics: + {'align_loss': 0.021988989785313606, + 'recon_loss': 0.004694939125329256, + 'predict_loss': 0.021235812455415726, + 'aux_loss_decay_weight': 0.35819999999999996, + 'grad_norm_pre_clip': 0.45288047194480896, + 'data_time': 0.0007232789939735085, + 'model_time': 1.2430107909895014, + 'grad_norm_pre_clip_avg': 0.43224344253540037, + 'learning_rate': 1.605e-05, 'epoch': 0.81} +04/19 [12:26:30] INFO | >> train_qwenlatent.py:487 + Step 3220 | grad_norm_pre_clip=0.3823 | + grad_norm_pre_clip_avg=0.4134 | Metrics: + {'align_loss': 0.021721959114074707, + 'recon_loss': 0.005254456773400307, + 'predict_loss': 0.02730434201657772, + 'aux_loss_decay_weight': 0.35619999999999996, + 'grad_norm_pre_clip': 0.38227570056915283, + 'data_time': 0.0007378509908448905, + 'model_time': 1.212143095995998, + 'grad_norm_pre_clip_avg': 0.4133961260318756, + 'learning_rate': 1.6100000000000002e-05, + 'epoch': 0.81} +04/19 [12:26:42] INFO | >> train_qwenlatent.py:487 + Step 3230 | grad_norm_pre_clip=0.5147 | + grad_norm_pre_clip_avg=0.4173 | Metrics: + {'align_loss': 0.022330252453684807, + 'recon_loss': 0.005358337890356779, + 'predict_loss': 0.02285483106970787, + 'aux_loss_decay_weight': 0.35419999999999996, + 'grad_norm_pre_clip': 0.5147024989128113, + 'data_time': 0.000859853986185044, + 'model_time': 1.2712752119987272, + 'grad_norm_pre_clip_avg': 0.4172986298799515, + 'learning_rate': 1.6150000000000003e-05, + 'epoch': 0.82} +04/19 [12:26:56] INFO | >> train_qwenlatent.py:487 + Step 3240 | grad_norm_pre_clip=0.4185 | + grad_norm_pre_clip_avg=0.4161 | Metrics: + {'align_loss': 0.021957270801067352, + 'recon_loss': 0.008307675831019878, + 'predict_loss': 0.028272364288568497, + 'aux_loss_decay_weight': 0.35219999999999996, + 'grad_norm_pre_clip': 0.4185493588447571, + 'data_time': 0.0008989579800982028, + 'model_time': 1.2332282419956755, + 'grad_norm_pre_clip_avg': 0.41607793271541593, + 'learning_rate': 1.62e-05, 'epoch': 0.82} +04/19 [12:27:09] INFO | >> train_qwenlatent.py:487 + Step 3250 | grad_norm_pre_clip=0.4246 | + grad_norm_pre_clip_avg=0.4311 | Metrics: + {'align_loss': 0.022444400936365128, + 'recon_loss': 0.005550798960030079, + 'predict_loss': 0.02433490939438343, + 'aux_loss_decay_weight': 0.35019999999999996, + 'grad_norm_pre_clip': 0.4246158003807068, + 'mae_score': 0.040160180856515695, 'data_time': + 0.0008817010093480349, 'model_time': + 1.243770956993103, 'grad_norm_pre_clip_avg': + 0.4311337321996689, 'learning_rate': + 1.6250000000000002e-05, 'epoch': 0.82} +04/19 [12:27:22] INFO | >> train_qwenlatent.py:487 + Step 3260 | grad_norm_pre_clip=0.4754 | + grad_norm_pre_clip_avg=0.4208 | Metrics: + {'align_loss': 0.020003337413072586, + 'recon_loss': 0.011683579534292221, + 'predict_loss': 0.03278706222772598, + 'aux_loss_decay_weight': 0.34819999999999995, + 'grad_norm_pre_clip': 0.47540342807769775, + 'data_time': 0.000975459988694638, + 'model_time': 1.2863962660194375, + 'grad_norm_pre_clip_avg': 0.42078891694545745, + 'learning_rate': 1.63e-05, 'epoch': 0.82} +04/19 [12:27:34] INFO | >> train_qwenlatent.py:487 + Step 3270 | grad_norm_pre_clip=0.4373 | + grad_norm_pre_clip_avg=0.4357 | Metrics: + {'align_loss': 0.023403018712997437, + 'recon_loss': 0.013482170179486275, + 'predict_loss': 0.03254663944244385, + 'aux_loss_decay_weight': 0.34619999999999995, + 'grad_norm_pre_clip': 0.437267005443573, + 'data_time': 0.0007061309879645705, + 'model_time': 1.2085814389865845, + 'grad_norm_pre_clip_avg': 0.4357088923454285, + 'learning_rate': 1.635e-05, 'epoch': 0.83} +04/19 [12:27:46] INFO | >> train_qwenlatent.py:487 + Step 3280 | grad_norm_pre_clip=0.3579 | + grad_norm_pre_clip_avg=0.4146 | Metrics: + {'align_loss': 0.021619092673063278, + 'recon_loss': 0.013138038106262684, + 'predict_loss': 0.034980691969394684, + 'aux_loss_decay_weight': 0.34419999999999995, + 'grad_norm_pre_clip': 0.3579459488391876, + 'data_time': 0.0009034049871843308, + 'model_time': 1.1768034319975413, + 'grad_norm_pre_clip_avg': 0.4145580053329468, + 'learning_rate': 1.6400000000000002e-05, + 'epoch': 0.83} +04/19 [12:27:59] INFO | >> train_qwenlatent.py:487 + Step 3290 | grad_norm_pre_clip=0.4910 | + grad_norm_pre_clip_avg=0.4745 | Metrics: + {'align_loss': 0.022492997348308563, + 'recon_loss': 0.008631713688373566, + 'predict_loss': 0.026659803465008736, + 'aux_loss_decay_weight': 0.34219999999999995, + 'grad_norm_pre_clip': 0.49101272225379944, + 'data_time': 0.001148387003922835, + 'model_time': 1.2692859210073948, + 'grad_norm_pre_clip_avg': 0.474535807967186, + 'learning_rate': 1.645e-05, 'epoch': 0.83} +04/19 [12:28:12] INFO | >> train_qwenlatent.py:487 + Step 3300 | grad_norm_pre_clip=0.3871 | + grad_norm_pre_clip_avg=0.4311 | Metrics: + {'align_loss': 0.02197040431201458, + 'recon_loss': 0.0053144339472055435, + 'predict_loss': 0.02687010169029236, + 'aux_loss_decay_weight': 0.34019999999999995, + 'grad_norm_pre_clip': 0.38709405064582825, + 'mae_score': 0.05283580814395939, 'data_time': + 0.0008875399944372475, 'model_time': + 1.2213598410016857, 'grad_norm_pre_clip_avg': + 0.4310567915439606, 'learning_rate': 1.65e-05, + 'epoch': 0.83} +04/19 [12:28:25] INFO | >> train_qwenlatent.py:487 + Step 3310 | grad_norm_pre_clip=0.3737 | + grad_norm_pre_clip_avg=0.4082 | Metrics: + {'align_loss': 0.021552931517362595, + 'recon_loss': 0.008236139081418514, + 'predict_loss': 0.02589205466210842, + 'aux_loss_decay_weight': 0.33819999999999995, + 'grad_norm_pre_clip': 0.37369948625564575, + 'data_time': 0.0008882649999577552, + 'model_time': 1.2427419750019908, + 'grad_norm_pre_clip_avg': 0.4082283616065979, + 'learning_rate': 1.6550000000000002e-05, + 'epoch': 0.84} +04/19 [12:28:38] INFO | >> train_qwenlatent.py:487 + Step 3320 | grad_norm_pre_clip=0.4544 | + grad_norm_pre_clip_avg=0.4685 | Metrics: + {'align_loss': 0.022471686825156212, + 'recon_loss': 0.012639488093554974, + 'predict_loss': 0.031550850719213486, + 'aux_loss_decay_weight': 0.33620000000000005, + 'grad_norm_pre_clip': 0.45437923073768616, + 'data_time': 0.0009008199849631637, + 'model_time': 1.2751981269975659, + 'grad_norm_pre_clip_avg': 0.46849370300769805, + 'learning_rate': 1.66e-05, 'epoch': 0.84} +04/19 [12:28:51] INFO | >> train_qwenlatent.py:487 + Step 3330 | grad_norm_pre_clip=0.3854 | + grad_norm_pre_clip_avg=0.4289 | Metrics: + {'align_loss': 0.021550575271248817, + 'recon_loss': 0.008176294155418873, + 'predict_loss': 0.027373116463422775, + 'aux_loss_decay_weight': 0.33420000000000005, + 'grad_norm_pre_clip': 0.38538068532943726, + 'data_time': 0.0006587189855054021, + 'model_time': 1.446153531986056, + 'grad_norm_pre_clip_avg': 0.4289441376924515, + 'learning_rate': 1.665e-05, 'epoch': 0.84} +04/19 [12:29:04] INFO | >> train_qwenlatent.py:487 + Step 3340 | grad_norm_pre_clip=0.4190 | + grad_norm_pre_clip_avg=0.4113 | Metrics: + {'align_loss': 0.02247636392712593, + 'recon_loss': 0.007204123307019472, + 'predict_loss': 0.02825211174786091, + 'aux_loss_decay_weight': 0.33220000000000005, + 'grad_norm_pre_clip': 0.4190283417701721, + 'data_time': 0.0006403409934137017, + 'model_time': 1.270534569019219, + 'grad_norm_pre_clip_avg': 0.41127391457557677, + 'learning_rate': 1.6700000000000003e-05, + 'epoch': 0.84} +04/19 [12:29:17] INFO | >> train_qwenlatent.py:487 + Step 3350 | grad_norm_pre_clip=0.4697 | + grad_norm_pre_clip_avg=0.4276 | Metrics: + {'align_loss': 0.022059835493564606, + 'recon_loss': 0.005398999899625778, + 'predict_loss': 0.02182709611952305, + 'aux_loss_decay_weight': 0.33020000000000005, + 'grad_norm_pre_clip': 0.469716876745224, + 'mae_score': 0.03754484503118841, 'data_time': + 0.0014328680117614567, 'model_time': + 1.252615584002342, 'grad_norm_pre_clip_avg': + 0.42761040925979615, 'learning_rate': + 1.675e-05, 'epoch': 0.85} +04/19 [12:29:29] INFO | >> train_qwenlatent.py:487 + Step 3360 | grad_norm_pre_clip=0.4276 | + grad_norm_pre_clip_avg=0.4377 | Metrics: + {'align_loss': 0.02025364525616169, + 'recon_loss': 0.00640458706766367, + 'predict_loss': 0.02023816853761673, + 'aux_loss_decay_weight': 0.32820000000000005, + 'grad_norm_pre_clip': 0.4275672733783722, + 'data_time': 0.0006965779757592827, + 'model_time': 1.2032782530004624, + 'grad_norm_pre_clip_avg': 0.437741819024086, + 'learning_rate': 1.6800000000000002e-05, + 'epoch': 0.85} +04/19 [12:29:42] INFO | >> train_qwenlatent.py:487 + Step 3370 | grad_norm_pre_clip=0.3358 | + grad_norm_pre_clip_avg=0.4026 | Metrics: + {'align_loss': 0.021570425480604172, + 'recon_loss': 0.009810427203774452, + 'predict_loss': 0.029976602643728256, + 'aux_loss_decay_weight': 0.32620000000000005, + 'grad_norm_pre_clip': 0.33579495549201965, + 'data_time': 0.0009231260046362877, + 'model_time': 1.2718088850087952, + 'grad_norm_pre_clip_avg': 0.4025770008563995, + 'learning_rate': 1.6850000000000003e-05, + 'epoch': 0.85} +04/19 [12:29:55] INFO | >> train_qwenlatent.py:487 + Step 3380 | grad_norm_pre_clip=0.4589 | + grad_norm_pre_clip_avg=0.4474 | Metrics: + {'align_loss': 0.023575618863105774, + 'recon_loss': 0.006528941914439201, + 'predict_loss': 0.025406228378415108, + 'aux_loss_decay_weight': 0.32420000000000004, + 'grad_norm_pre_clip': 0.4589027166366577, + 'data_time': 0.0007028750260360539, + 'model_time': 1.258916795020923, + 'grad_norm_pre_clip_avg': 0.4473631978034973, + 'learning_rate': 1.69e-05, 'epoch': 0.85} +04/19 [12:30:07] INFO | >> train_qwenlatent.py:487 + Step 3390 | grad_norm_pre_clip=0.4393 | + grad_norm_pre_clip_avg=0.4656 | Metrics: + {'align_loss': 0.02173357456922531, + 'recon_loss': 0.005875518079847097, + 'predict_loss': 0.02565672993659973, + 'aux_loss_decay_weight': 0.32220000000000004, + 'grad_norm_pre_clip': 0.43928346037864685, + 'data_time': 0.0006482259777840227, + 'model_time': 1.2255883300094865, + 'grad_norm_pre_clip_avg': 0.46564996540546416, + 'learning_rate': 1.6950000000000002e-05, + 'epoch': 0.86} +04/19 [12:30:21] INFO | >> train_qwenlatent.py:487 + Step 3400 | grad_norm_pre_clip=0.4305 | + grad_norm_pre_clip_avg=0.4505 | Metrics: + {'align_loss': 0.021515991538763046, + 'recon_loss': 0.0057304659858345985, + 'predict_loss': 0.022132355719804764, + 'aux_loss_decay_weight': 0.32020000000000004, + 'grad_norm_pre_clip': 0.4305410087108612, + 'mae_score': 0.039434271460180886, 'data_time': + 0.0007823460036888719, 'model_time': + 1.3037538370117545, 'grad_norm_pre_clip_avg': + 0.45050692856311797, 'learning_rate': + 1.7000000000000003e-05, 'epoch': 0.86} +04/19 [12:30:33] INFO | >> train_qwenlatent.py:487 + Step 3410 | grad_norm_pre_clip=0.5138 | + grad_norm_pre_clip_avg=0.4430 | Metrics: + {'align_loss': 0.02113417536020279, + 'recon_loss': 0.00553778326138854, + 'predict_loss': 0.023489054292440414, + 'aux_loss_decay_weight': 0.31820000000000004, + 'grad_norm_pre_clip': 0.5137821435928345, + 'data_time': 0.0009588969987817109, + 'model_time': 1.2108013179968111, + 'grad_norm_pre_clip_avg': 0.44299786984920503, + 'learning_rate': 1.705e-05, 'epoch': 0.86} +04/19 [12:30:46] INFO | >> train_qwenlatent.py:487 + Step 3420 | grad_norm_pre_clip=0.3532 | + grad_norm_pre_clip_avg=0.4401 | Metrics: + {'align_loss': 0.021470826119184494, + 'recon_loss': 0.00485337246209383, + 'predict_loss': 0.02144312672317028, + 'aux_loss_decay_weight': 0.31620000000000004, + 'grad_norm_pre_clip': 0.35323596000671387, + 'data_time': 0.000995681999484077, + 'model_time': 1.267639290977968, + 'grad_norm_pre_clip_avg': 0.44012013971805575, + 'learning_rate': 1.7100000000000002e-05, + 'epoch': 0.86} +04/19 [12:30:58] INFO | >> train_qwenlatent.py:487 + Step 3430 | grad_norm_pre_clip=0.4649 | + grad_norm_pre_clip_avg=0.4785 | Metrics: + {'align_loss': 0.021004289388656616, + 'recon_loss': 0.004550606478005648, + 'predict_loss': 0.02090437337756157, + 'aux_loss_decay_weight': 0.31420000000000003, + 'grad_norm_pre_clip': 0.4649115800857544, + 'data_time': 0.0007603340200148523, + 'model_time': 1.2137302809860557, + 'grad_norm_pre_clip_avg': 0.4785106658935547, + 'learning_rate': 1.7150000000000004e-05, + 'epoch': 0.87} +04/19 [12:31:11] INFO | >> train_qwenlatent.py:487 + Step 3440 | grad_norm_pre_clip=0.6550 | + grad_norm_pre_clip_avg=0.5731 | Metrics: + {'align_loss': 0.0213061161339283, + 'recon_loss': 0.0033942312002182007, + 'predict_loss': 0.02069331519305706, + 'aux_loss_decay_weight': 0.31220000000000003, + 'grad_norm_pre_clip': 0.6549634337425232, + 'data_time': 0.0010003200150094926, + 'model_time': 1.2594217270088848, + 'grad_norm_pre_clip_avg': 0.5730823934078216, + 'learning_rate': 1.7199999999999998e-05, + 'epoch': 0.87} +04/19 [12:31:25] INFO | >> train_qwenlatent.py:487 + Step 3450 | grad_norm_pre_clip=0.4008 | + grad_norm_pre_clip_avg=0.4603 | Metrics: + {'align_loss': 0.020766671746969223, + 'recon_loss': 0.005325540900230408, + 'predict_loss': 0.0205985177308321, + 'aux_loss_decay_weight': 0.31020000000000003, + 'grad_norm_pre_clip': 0.4008253812789917, + 'mae_score': 0.0300096185357721, 'data_time': + 0.0006861849979031831, 'model_time': + 1.241467776009813, 'grad_norm_pre_clip_avg': + 0.46026581823825835, 'learning_rate': + 1.725e-05, 'epoch': 0.87} +04/19 [12:31:38] INFO | >> train_qwenlatent.py:487 + Step 3460 | grad_norm_pre_clip=0.3604 | + grad_norm_pre_clip_avg=0.3682 | Metrics: + {'align_loss': 0.02013203501701355, + 'recon_loss': 0.006746629718691111, + 'predict_loss': 0.0216225553303957, + 'aux_loss_decay_weight': 0.30820000000000003, + 'grad_norm_pre_clip': 0.3604331314563751, + 'data_time': 0.0008105739834718406, + 'model_time': 1.2319326929864474, + 'grad_norm_pre_clip_avg': 0.36817877292633056, + 'learning_rate': 1.73e-05, 'epoch': 0.87} +04/19 [12:31:50] INFO | >> train_qwenlatent.py:487 + Step 3470 | grad_norm_pre_clip=0.3988 | + grad_norm_pre_clip_avg=0.3992 | Metrics: + {'align_loss': 0.021914806216955185, + 'recon_loss': 0.004704377613961697, + 'predict_loss': 0.022035710513591766, + 'aux_loss_decay_weight': 0.3062, + 'grad_norm_pre_clip': 0.3987574577331543, + 'data_time': 0.000855886988574639, + 'model_time': 1.2002704329788685, + 'grad_norm_pre_clip_avg': 0.3991686195135117, + 'learning_rate': 1.7349999999999998e-05, + 'epoch': 0.88} +04/19 [12:32:03] INFO | >> train_qwenlatent.py:487 + Step 3480 | grad_norm_pre_clip=0.4337 | + grad_norm_pre_clip_avg=0.3775 | Metrics: + {'align_loss': 0.022142285481095314, + 'recon_loss': 0.005630184896290302, + 'predict_loss': 0.021340403705835342, + 'aux_loss_decay_weight': 0.3042, + 'grad_norm_pre_clip': 0.43373507261276245, + 'data_time': 0.0009042780147865415, + 'model_time': 1.2672663289995398, + 'grad_norm_pre_clip_avg': 0.3774915784597397, + 'learning_rate': 1.74e-05, 'epoch': 0.88} +04/19 [12:32:16] INFO | >> train_qwenlatent.py:487 + Step 3490 | grad_norm_pre_clip=0.3459 | + grad_norm_pre_clip_avg=0.3905 | Metrics: + {'align_loss': 0.02146509289741516, + 'recon_loss': 0.014182827435433865, + 'predict_loss': 0.033202480524778366, + 'aux_loss_decay_weight': 0.3022, + 'grad_norm_pre_clip': 0.3459007143974304, + 'data_time': 0.0011961470008827746, + 'model_time': 1.2465125269955024, + 'grad_norm_pre_clip_avg': 0.390460342168808, + 'learning_rate': 1.745e-05, 'epoch': 0.88} +04/19 [12:32:30] INFO | >> train_qwenlatent.py:487 + Step 3500 | grad_norm_pre_clip=0.3990 | + grad_norm_pre_clip_avg=0.3631 | Metrics: + {'align_loss': 0.02167198248207569, + 'recon_loss': 0.008711221627891064, + 'predict_loss': 0.026547789573669434, + 'aux_loss_decay_weight': 0.3002, + 'grad_norm_pre_clip': 0.39895012974739075, + 'mae_score': 0.04317859615291561, 'data_time': + 0.000804251991212368, 'model_time': + 1.2581539719831198, 'grad_norm_pre_clip_avg': + 0.3630818873643875, 'learning_rate': 1.75e-05, + 'epoch': 0.88} +04/19 [12:32:42] INFO | >> train_qwenlatent.py:487 + Step 3510 | grad_norm_pre_clip=0.6650 | + grad_norm_pre_clip_avg=0.4408 | Metrics: + {'align_loss': 0.021979166194796562, + 'recon_loss': 0.009909256361424923, + 'predict_loss': 0.029503945261240005, + 'aux_loss_decay_weight': 0.2982, + 'grad_norm_pre_clip': 0.6649666428565979, + 'data_time': 0.0009861240105237812, + 'model_time': 1.2187175510043744, + 'grad_norm_pre_clip_avg': 0.44076292514801024, + 'learning_rate': 1.755e-05, 'epoch': 0.89} +04/19 [12:32:55] INFO | >> train_qwenlatent.py:487 + Step 3520 | grad_norm_pre_clip=0.4334 | + grad_norm_pre_clip_avg=0.4474 | Metrics: + {'align_loss': 0.02078578993678093, + 'recon_loss': 0.014562228694558144, + 'predict_loss': 0.031971197575330734, + 'aux_loss_decay_weight': 0.2962, + 'grad_norm_pre_clip': 0.4333798587322235, + 'data_time': 0.0009859820129349828, + 'model_time': 1.3288771219959017, + 'grad_norm_pre_clip_avg': 0.4473709225654602, + 'learning_rate': 1.76e-05, 'epoch': 0.89} +04/19 [12:33:07] INFO | >> train_qwenlatent.py:487 + Step 3530 | grad_norm_pre_clip=0.3907 | + grad_norm_pre_clip_avg=0.4226 | Metrics: + {'align_loss': 0.022295158356428146, + 'recon_loss': 0.007805418223142624, + 'predict_loss': 0.025803519412875175, + 'aux_loss_decay_weight': 0.2942, + 'grad_norm_pre_clip': 0.39073678851127625, + 'data_time': 0.0009321990073658526, + 'model_time': 1.2738341910007875, + 'grad_norm_pre_clip_avg': 0.4226017266511917, + 'learning_rate': 1.765e-05, 'epoch': 0.89} +04/19 [12:33:20] INFO | >> train_qwenlatent.py:487 + Step 3540 | grad_norm_pre_clip=0.4591 | + grad_norm_pre_clip_avg=0.4096 | Metrics: + {'align_loss': 0.02120337262749672, + 'recon_loss': 0.0042144302278757095, + 'predict_loss': 0.030477164313197136, + 'aux_loss_decay_weight': 0.2922, + 'grad_norm_pre_clip': 0.4591095745563507, + 'data_time': 0.0008121189894154668, + 'model_time': 1.3081137599947397, + 'grad_norm_pre_clip_avg': 0.4095966160297394, + 'learning_rate': 1.77e-05, 'epoch': 0.89} +04/19 [12:33:33] INFO | >> train_qwenlatent.py:487 + Step 3550 | grad_norm_pre_clip=0.3938 | + grad_norm_pre_clip_avg=0.4256 | Metrics: + {'align_loss': 0.021676743403077126, + 'recon_loss': 0.007712861057370901, + 'predict_loss': 0.02604410983622074, + 'aux_loss_decay_weight': 0.2902, + 'grad_norm_pre_clip': 0.39376261830329895, + 'mae_score': 0.05811806068764076, 'data_time': + 0.0007295219984371215, 'model_time': + 1.2524688120174687, 'grad_norm_pre_clip_avg': + 0.4256297290325165, 'learning_rate': 1.775e-05, + 'epoch': 0.9} +04/19 [12:33:47] INFO | >> train_qwenlatent.py:487 + Step 3560 | grad_norm_pre_clip=0.3881 | + grad_norm_pre_clip_avg=0.4498 | Metrics: + {'align_loss': 0.020196907222270966, + 'recon_loss': 0.007811591029167175, + 'predict_loss': 0.022650456055998802, + 'aux_loss_decay_weight': 0.2882, + 'grad_norm_pre_clip': 0.388098806142807, + 'data_time': 0.0009732969920150936, + 'model_time': 1.5692977980070282, + 'grad_norm_pre_clip_avg': 0.44980194568634035, + 'learning_rate': 1.78e-05, 'epoch': 0.9} +04/19 [12:33:59] INFO | >> train_qwenlatent.py:487 + Step 3570 | grad_norm_pre_clip=0.3893 | + grad_norm_pre_clip_avg=0.4084 | Metrics: + {'align_loss': 0.020623352378606796, + 'recon_loss': 0.007072055246680975, + 'predict_loss': 0.024993447586894035, + 'aux_loss_decay_weight': 0.2862, + 'grad_norm_pre_clip': 0.3893183171749115, + 'data_time': 0.0007278080156538635, + 'model_time': 1.2351052020094357, + 'grad_norm_pre_clip_avg': 0.40835812091827395, + 'learning_rate': 1.785e-05, 'epoch': 0.9} +04/19 [12:34:11] INFO | >> train_qwenlatent.py:487 + Step 3580 | grad_norm_pre_clip=0.4017 | + grad_norm_pre_clip_avg=0.4031 | Metrics: + {'align_loss': 0.02036597579717636, + 'recon_loss': 0.005442316178232431, + 'predict_loss': 0.019345100969076157, + 'aux_loss_decay_weight': 0.2842, + 'grad_norm_pre_clip': 0.4017409682273865, + 'data_time': 0.0007986639975570142, + 'model_time': 1.1531935709936079, + 'grad_norm_pre_clip_avg': 0.4031233608722687, + 'learning_rate': 1.79e-05, 'epoch': 0.9} +04/19 [12:34:23] INFO | >> train_qwenlatent.py:487 + Step 3590 | grad_norm_pre_clip=0.4282 | + grad_norm_pre_clip_avg=0.4082 | Metrics: + {'align_loss': 0.020628008991479874, + 'recon_loss': 0.0073767914436757565, + 'predict_loss': 0.029773438349366188, + 'aux_loss_decay_weight': 0.2822, + 'grad_norm_pre_clip': 0.42817574739456177, + 'data_time': 0.0005860039964318275, + 'model_time': 1.217438146006316, + 'grad_norm_pre_clip_avg': 0.4081587314605713, + 'learning_rate': 1.795e-05, 'epoch': 0.91} +04/19 [12:34:35] INFO | >> train_qwenlatent.py:487 + Step 3600 | grad_norm_pre_clip=0.4827 | + grad_norm_pre_clip_avg=0.4002 | Metrics: + {'align_loss': 0.021007362753152847, + 'recon_loss': 0.007615404203534126, + 'predict_loss': 0.024222901090979576, + 'aux_loss_decay_weight': 0.2802, + 'grad_norm_pre_clip': 0.4827106297016144, + 'mae_score': 0.04164152746801978, 'data_time': + 0.0006287730066105723, 'model_time': + 1.1638513619836885, 'grad_norm_pre_clip_avg': + 0.4002135843038559, 'learning_rate': 1.8e-05, + 'epoch': 0.91} +04/19 [12:34:47] INFO | >> train_qwenlatent.py:487 + Step 3610 | grad_norm_pre_clip=0.5364 | + grad_norm_pre_clip_avg=0.4217 | Metrics: + {'align_loss': 0.02233843132853508, + 'recon_loss': 0.010476724244654179, + 'predict_loss': 0.026853110641241074, + 'aux_loss_decay_weight': 0.2782, + 'grad_norm_pre_clip': 0.5363562107086182, + 'data_time': 0.0005911679763812572, + 'model_time': 1.3138969089777675, + 'grad_norm_pre_clip_avg': 0.421725270152092, + 'learning_rate': 1.805e-05, 'epoch': 0.91} +04/19 [12:34:59] INFO | >> train_qwenlatent.py:487 + Step 3620 | grad_norm_pre_clip=0.4896 | + grad_norm_pre_clip_avg=0.4192 | Metrics: + {'align_loss': 0.02241400256752968, + 'recon_loss': 0.009555304422974586, + 'predict_loss': 0.0351148322224617, + 'aux_loss_decay_weight': 0.2762, + 'grad_norm_pre_clip': 0.48958203196525574, + 'data_time': 0.0006580670014955103, + 'model_time': 1.1685559430043213, + 'grad_norm_pre_clip_avg': 0.41920346915721896, + 'learning_rate': 1.81e-05, 'epoch': 0.91} +04/19 [12:35:11] INFO | >> train_qwenlatent.py:487 + Step 3630 | grad_norm_pre_clip=0.4461 | + grad_norm_pre_clip_avg=0.4099 | Metrics: + {'align_loss': 0.02178013324737549, + 'recon_loss': 0.01428136695176363, + 'predict_loss': 0.03169840946793556, + 'aux_loss_decay_weight': 0.2742, + 'grad_norm_pre_clip': 0.446134477853775, + 'data_time': 0.0005751470162067562, + 'model_time': 1.3420532320160419, + 'grad_norm_pre_clip_avg': 0.4098987728357315, + 'learning_rate': 1.815e-05, 'epoch': 0.92} +04/19 [12:35:22] INFO | >> train_qwenlatent.py:487 + Step 3640 | grad_norm_pre_clip=0.4450 | + grad_norm_pre_clip_avg=0.4032 | Metrics: + {'align_loss': 0.022204536944627762, + 'recon_loss': 0.01702328585088253, + 'predict_loss': 0.03499751165509224, + 'aux_loss_decay_weight': 0.2722, + 'grad_norm_pre_clip': 0.4449518024921417, + 'data_time': 0.000617796991718933, + 'model_time': 1.1688215440080967, + 'grad_norm_pre_clip_avg': 0.403178209066391, + 'learning_rate': 1.8200000000000002e-05, + 'epoch': 0.92} +04/19 [12:35:35] INFO | >> train_qwenlatent.py:487 + Step 3650 | grad_norm_pre_clip=0.3719 | + grad_norm_pre_clip_avg=0.4293 | Metrics: + {'align_loss': 0.022046316415071487, + 'recon_loss': 0.006025013979524374, + 'predict_loss': 0.021607128903269768, + 'aux_loss_decay_weight': 0.2702, + 'grad_norm_pre_clip': 0.37185531854629517, + 'mae_score': 0.045762534614081855, 'data_time': + 0.0006309979944489896, 'model_time': + 1.1460044709965587, 'grad_norm_pre_clip_avg': + 0.42931508719921113, 'learning_rate': + 1.825e-05, 'epoch': 0.92} +04/19 [12:35:46] INFO | >> train_qwenlatent.py:487 + Step 3660 | grad_norm_pre_clip=0.3208 | + grad_norm_pre_clip_avg=0.3519 | Metrics: + {'align_loss': 0.020361166447401047, + 'recon_loss': 0.0064985258504748344, + 'predict_loss': 0.021394014358520508, + 'aux_loss_decay_weight': 0.2682, + 'grad_norm_pre_clip': 0.3208256959915161, + 'data_time': 0.0005865110142622143, + 'model_time': 1.195628007990308, + 'grad_norm_pre_clip_avg': 0.3518692284822464, + 'learning_rate': 1.83e-05, 'epoch': 0.92} +04/19 [12:35:58] INFO | >> train_qwenlatent.py:487 + Step 3670 | grad_norm_pre_clip=0.5194 | + grad_norm_pre_clip_avg=0.4465 | Metrics: + {'align_loss': 0.021858826279640198, + 'recon_loss': 0.010282703675329685, + 'predict_loss': 0.03386659175157547, + 'aux_loss_decay_weight': 0.2662, + 'grad_norm_pre_clip': 0.5194299817085266, + 'data_time': 0.000619031983660534, + 'model_time': 1.1352475850144401, + 'grad_norm_pre_clip_avg': 0.446452271938324, + 'learning_rate': 1.8350000000000002e-05, + 'epoch': 0.93} +04/19 [12:36:09] INFO | >> train_qwenlatent.py:487 + Step 3680 | grad_norm_pre_clip=0.3053 | + grad_norm_pre_clip_avg=0.3753 | Metrics: + {'align_loss': 0.021515395492315292, + 'recon_loss': 0.0061225066892802715, + 'predict_loss': 0.0215986967086792, + 'aux_loss_decay_weight': 0.2642, + 'grad_norm_pre_clip': 0.3053123652935028, + 'data_time': 0.0005605369806289673, + 'model_time': 1.1542850109981373, + 'grad_norm_pre_clip_avg': 0.37526738047599795, + 'learning_rate': 1.84e-05, 'epoch': 0.93} +04/19 [12:36:21] INFO | >> train_qwenlatent.py:487 + Step 3690 | grad_norm_pre_clip=0.2853 | + grad_norm_pre_clip_avg=0.3644 | Metrics: + {'align_loss': 0.022347157821059227, + 'recon_loss': 0.007345616351813078, + 'predict_loss': 0.038623061031103134, + 'aux_loss_decay_weight': 0.2622, + 'grad_norm_pre_clip': 0.2853028178215027, + 'data_time': 0.000643732986645773, + 'model_time': 1.1416475660225842, + 'grad_norm_pre_clip_avg': 0.3644202947616577, + 'learning_rate': 1.845e-05, 'epoch': 0.93} +04/19 [12:36:33] INFO | >> train_qwenlatent.py:487 + Step 3700 | grad_norm_pre_clip=0.3841 | + grad_norm_pre_clip_avg=0.3834 | Metrics: + {'align_loss': 0.021905262023210526, + 'recon_loss': 0.008587827906012535, + 'predict_loss': 0.02789672091603279, + 'aux_loss_decay_weight': 0.2602, + 'grad_norm_pre_clip': 0.38409507274627686, + 'mae_score': 0.0338502007561761, 'data_time': + 0.000567021023016423, 'model_time': + 1.1734103549970314, 'grad_norm_pre_clip_avg': + 0.3834357589483261, 'learning_rate': 1.85e-05, + 'epoch': 0.93} +04/19 [12:36:45] INFO | >> train_qwenlatent.py:487 + Step 3710 | grad_norm_pre_clip=0.5119 | + grad_norm_pre_clip_avg=0.4479 | Metrics: + {'align_loss': 0.020961329340934753, + 'recon_loss': 0.012479635886847973, + 'predict_loss': 0.035206444561481476, + 'aux_loss_decay_weight': 0.2582, + 'grad_norm_pre_clip': 0.5118501782417297, + 'data_time': 0.0005457290099002421, + 'model_time': 1.1542179140087683, + 'grad_norm_pre_clip_avg': 0.4479077845811844, + 'learning_rate': 1.855e-05, 'epoch': 0.94} +04/19 [12:36:56] INFO | >> train_qwenlatent.py:487 + Step 3720 | grad_norm_pre_clip=0.3414 | + grad_norm_pre_clip_avg=0.4252 | Metrics: + {'align_loss': 0.022020462900400162, + 'recon_loss': 0.004368764813989401, + 'predict_loss': 0.02228609472513199, + 'aux_loss_decay_weight': 0.2562, + 'grad_norm_pre_clip': 0.34138113260269165, + 'data_time': 0.0005689930112566799, + 'model_time': 1.1794286169752013, + 'grad_norm_pre_clip_avg': 0.4252279490232468, + 'learning_rate': 1.86e-05, 'epoch': 0.94} +04/19 [12:37:08] INFO | >> train_qwenlatent.py:487 + Step 3730 | grad_norm_pre_clip=0.3639 | + grad_norm_pre_clip_avg=0.3842 | Metrics: + {'align_loss': 0.02295232191681862, + 'recon_loss': 0.0107260225340724, + 'predict_loss': 0.02866271138191223, + 'aux_loss_decay_weight': 0.2542, + 'grad_norm_pre_clip': 0.36392396688461304, + 'data_time': 0.0005583099846262485, + 'model_time': 1.1564419140049722, + 'grad_norm_pre_clip_avg': 0.3841566175222397, + 'learning_rate': 1.865e-05, 'epoch': 0.94} +04/19 [12:37:20] INFO | >> train_qwenlatent.py:487 + Step 3740 | grad_norm_pre_clip=0.4103 | + grad_norm_pre_clip_avg=0.3991 | Metrics: + {'align_loss': 0.021007979288697243, + 'recon_loss': 0.008970878086984158, + 'predict_loss': 0.024217266589403152, + 'aux_loss_decay_weight': 0.2522, + 'grad_norm_pre_clip': 0.41032490134239197, + 'data_time': 0.0005729839904233813, + 'model_time': 1.1666414980136324, + 'grad_norm_pre_clip_avg': 0.3991235435009003, + 'learning_rate': 1.87e-05, 'epoch': 0.94} +04/19 [12:37:32] INFO | >> train_qwenlatent.py:487 + Step 3750 | grad_norm_pre_clip=0.3371 | + grad_norm_pre_clip_avg=0.3651 | Metrics: + {'align_loss': 0.022532816976308823, + 'recon_loss': 0.010142993181943893, + 'predict_loss': 0.029305756092071533, + 'aux_loss_decay_weight': 0.2502, + 'grad_norm_pre_clip': 0.3370548486709595, + 'mae_score': 0.03923610652889217, 'data_time': + 0.000659760000417009, 'model_time': + 1.1524693770043086, 'grad_norm_pre_clip_avg': + 0.36513690948486327, 'learning_rate': + 1.8750000000000002e-05, 'epoch': 0.95} +04/19 [12:37:44] INFO | >> train_qwenlatent.py:487 + Step 3760 | grad_norm_pre_clip=0.4190 | + grad_norm_pre_clip_avg=0.4086 | Metrics: + {'align_loss': 0.02185310237109661, + 'recon_loss': 0.007327920291572809, + 'predict_loss': 0.02466178685426712, + 'aux_loss_decay_weight': 0.24819999999999998, + 'grad_norm_pre_clip': 0.4190080463886261, + 'data_time': 0.0005794079916086048, + 'model_time': 1.151862483995501, + 'grad_norm_pre_clip_avg': 0.4085704982280731, + 'learning_rate': 1.88e-05, 'epoch': 0.95} +04/19 [12:37:55] INFO | >> train_qwenlatent.py:487 + Step 3770 | grad_norm_pre_clip=0.4544 | + grad_norm_pre_clip_avg=0.4036 | Metrics: + {'align_loss': 0.02047846093773842, + 'recon_loss': 0.006561996880918741, + 'predict_loss': 0.020795483142137527, + 'aux_loss_decay_weight': 0.24619999999999997, + 'grad_norm_pre_clip': 0.45436522364616394, + 'data_time': 0.000878683989867568, + 'model_time': 1.1427734729950316, + 'grad_norm_pre_clip_avg': 0.40359904170036315, + 'learning_rate': 1.885e-05, 'epoch': 0.95} +04/19 [12:38:07] INFO | >> train_qwenlatent.py:487 + Step 3780 | grad_norm_pre_clip=0.3939 | + grad_norm_pre_clip_avg=0.4298 | Metrics: + {'align_loss': 0.02110864408314228, + 'recon_loss': 0.007861907593905926, + 'predict_loss': 0.02494148723781109, + 'aux_loss_decay_weight': 0.24419999999999997, + 'grad_norm_pre_clip': 0.3938802182674408, + 'data_time': 0.0005625399935524911, + 'model_time': 1.148452177993022, + 'grad_norm_pre_clip_avg': 0.42978369295597074, + 'learning_rate': 1.8900000000000002e-05, + 'epoch': 0.95} +04/19 [12:38:18] INFO | >> train_qwenlatent.py:487 + Step 3790 | grad_norm_pre_clip=0.3178 | + grad_norm_pre_clip_avg=0.4224 | Metrics: + {'align_loss': 0.022679587826132774, + 'recon_loss': 0.010539412498474121, + 'predict_loss': 0.02237164042890072, + 'aux_loss_decay_weight': 0.24219999999999997, + 'grad_norm_pre_clip': 0.31782597303390503, + 'data_time': 0.0005652570107486099, + 'model_time': 1.1619177570100874, + 'grad_norm_pre_clip_avg': 0.4223823606967926, + 'learning_rate': 1.895e-05, 'epoch': 0.96} +04/19 [12:38:31] INFO | >> train_qwenlatent.py:487 + Step 3800 | grad_norm_pre_clip=0.3803 | + grad_norm_pre_clip_avg=0.3804 | Metrics: + {'align_loss': 0.0227397121489048, + 'recon_loss': 0.009827765636146069, + 'predict_loss': 0.03295743092894554, + 'aux_loss_decay_weight': 0.24019999999999997, + 'grad_norm_pre_clip': 0.38034912943840027, + 'mae_score': 0.035034818907041805, 'data_time': + 0.0005410439916886389, 'model_time': + 1.1789301809913013, 'grad_norm_pre_clip_avg': + 0.3803711324930191, 'learning_rate': 1.9e-05, + 'epoch': 0.96} +04/19 [12:38:43] INFO | >> train_qwenlatent.py:487 + Step 3810 | grad_norm_pre_clip=0.4280 | + grad_norm_pre_clip_avg=0.4307 | Metrics: + {'align_loss': 0.022562241181731224, + 'recon_loss': 0.006982510443776846, + 'predict_loss': 0.030206037685275078, + 'aux_loss_decay_weight': 0.23819999999999997, + 'grad_norm_pre_clip': 0.42795330286026, + 'data_time': 0.0005229779926594347, + 'model_time': 1.1264228239888325, + 'grad_norm_pre_clip_avg': 0.43067117035388947, + 'learning_rate': 1.9050000000000002e-05, + 'epoch': 0.96} +04/19 [12:38:54] INFO | >> train_qwenlatent.py:487 + Step 3820 | grad_norm_pre_clip=0.2928 | + grad_norm_pre_clip_avg=0.4212 | Metrics: + {'align_loss': 0.021835163235664368, + 'recon_loss': 0.007507218047976494, + 'predict_loss': 0.021571075543761253, + 'aux_loss_decay_weight': 0.23619999999999997, + 'grad_norm_pre_clip': 0.29278984665870667, + 'data_time': 0.0005754309822805226, + 'model_time': 1.1360422009893227, + 'grad_norm_pre_clip_avg': 0.42121310234069825, + 'learning_rate': 1.91e-05, 'epoch': 0.96} +04/19 [12:39:06] INFO | >> train_qwenlatent.py:487 + Step 3830 | grad_norm_pre_clip=0.4362 | + grad_norm_pre_clip_avg=0.3854 | Metrics: + {'align_loss': 0.020810354501008987, + 'recon_loss': 0.005522106308490038, + 'predict_loss': 0.018393270671367645, + 'aux_loss_decay_weight': 0.23419999999999996, + 'grad_norm_pre_clip': 0.4362432062625885, + 'data_time': 0.0007328469946514815, + 'model_time': 1.1643063490046188, + 'grad_norm_pre_clip_avg': 0.3853583812713623, + 'learning_rate': 1.915e-05, 'epoch': 0.97} +04/19 [12:39:17] INFO | >> train_qwenlatent.py:487 + Step 3840 | grad_norm_pre_clip=0.4246 | + grad_norm_pre_clip_avg=0.3805 | Metrics: + {'align_loss': 0.02103109285235405, + 'recon_loss': 0.005691183730959892, + 'predict_loss': 0.023245301097631454, + 'aux_loss_decay_weight': 0.23219999999999996, + 'grad_norm_pre_clip': 0.4246092140674591, + 'data_time': 0.0006060469895601273, + 'model_time': 1.1440136239980347, + 'grad_norm_pre_clip_avg': 0.3805268406867981, + 'learning_rate': 1.9200000000000003e-05, + 'epoch': 0.97} +04/19 [12:39:29] INFO | >> train_qwenlatent.py:487 + Step 3850 | grad_norm_pre_clip=0.3632 | + grad_norm_pre_clip_avg=0.4164 | Metrics: + {'align_loss': 0.022464215755462646, + 'recon_loss': 0.004911210387945175, + 'predict_loss': 0.023039359599351883, + 'aux_loss_decay_weight': 0.23019999999999996, + 'grad_norm_pre_clip': 0.3632224202156067, + 'mae_score': 0.03760216171677048, 'data_time': + 0.0005464659770950675, 'model_time': + 1.150588966003852, 'grad_norm_pre_clip_avg': + 0.4163765460252762, 'learning_rate': 1.925e-05, + 'epoch': 0.97} +04/19 [12:39:41] INFO | >> train_qwenlatent.py:487 + Step 3860 | grad_norm_pre_clip=0.4431 | + grad_norm_pre_clip_avg=0.4583 | Metrics: + {'align_loss': 0.022850196808576584, + 'recon_loss': 0.005131957121193409, + 'predict_loss': 0.017704389989376068, + 'aux_loss_decay_weight': 0.22819999999999996, + 'grad_norm_pre_clip': 0.4430999159812927, + 'data_time': 0.0005602480086963624, + 'model_time': 1.1532183179806452, + 'grad_norm_pre_clip_avg': 0.4582630068063736, + 'learning_rate': 1.93e-05, 'epoch': 0.97} +04/19 [12:39:53] INFO | >> train_qwenlatent.py:487 + Step 3870 | grad_norm_pre_clip=0.3669 | + grad_norm_pre_clip_avg=0.4148 | Metrics: + {'align_loss': 0.02119613252580166, + 'recon_loss': 0.004130940418690443, + 'predict_loss': 0.018315093591809273, + 'aux_loss_decay_weight': 0.22619999999999996, + 'grad_norm_pre_clip': 0.366915225982666, + 'data_time': 0.0005415260093286633, + 'model_time': 1.1466874949983321, + 'grad_norm_pre_clip_avg': 0.414779469370842, + 'learning_rate': 1.9350000000000003e-05, + 'epoch': 0.98} +04/19 [12:40:04] INFO | >> train_qwenlatent.py:487 + Step 3880 | grad_norm_pre_clip=0.3728 | + grad_norm_pre_clip_avg=0.4088 | Metrics: + {'align_loss': 0.020930642262101173, + 'recon_loss': 0.010690711438655853, + 'predict_loss': 0.027720898389816284, + 'aux_loss_decay_weight': 0.22419999999999995, + 'grad_norm_pre_clip': 0.3728492856025696, + 'data_time': 0.0005445360147859901, + 'model_time': 1.1375336610071827, + 'grad_norm_pre_clip_avg': 0.40878980457782743, + 'learning_rate': 1.94e-05, 'epoch': 0.98} +04/19 [12:40:47] INFO | >> train_qwenlatent.py:487 + Step 3890 | grad_norm_pre_clip=0.3702 | + grad_norm_pre_clip_avg=0.3558 | Metrics: + {'align_loss': 0.021569354459643364, + 'recon_loss': 0.007661484647542238, + 'predict_loss': 0.025356445461511612, + 'aux_loss_decay_weight': 0.22219999999999995, + 'grad_norm_pre_clip': 0.3701680600643158, + 'data_time': 0.0010663810244295746, + 'model_time': 3.3035806359839626, + 'grad_norm_pre_clip_avg': 0.3558153510093689, + 'learning_rate': 1.9450000000000002e-05, + 'epoch': 0.98} +04/19 [12:41:27] INFO | >> train_qwenlatent.py:487 + Step 3900 | grad_norm_pre_clip=0.3466 | + grad_norm_pre_clip_avg=0.3971 | Metrics: + {'align_loss': 0.022243287414312363, + 'recon_loss': 0.014669769443571568, + 'predict_loss': 0.03440515324473381, + 'aux_loss_decay_weight': 0.22019999999999995, + 'grad_norm_pre_clip': 0.3465961813926697, + 'mae_score': 0.03302198530317427, 'data_time': + 0.001231781003298238, 'model_time': + 4.0101588380057365, 'grad_norm_pre_clip_avg': + 0.39714584946632386, 'learning_rate': + 1.9500000000000003e-05, 'epoch': 0.98} +04/19 [12:42:03] INFO | >> train_qwenlatent.py:487 + Step 3910 | grad_norm_pre_clip=0.4946 | + grad_norm_pre_clip_avg=0.3901 | Metrics: + {'align_loss': 0.0210365392267704, + 'recon_loss': 0.005801810882985592, + 'predict_loss': 0.02268938720226288, + 'aux_loss_decay_weight': 0.21819999999999995, + 'grad_norm_pre_clip': 0.4945789575576782, + 'data_time': 0.0010845219949260354, + 'model_time': 3.4573913049825933, + 'grad_norm_pre_clip_avg': 0.39007793068885804, + 'learning_rate': 1.955e-05, 'epoch': 0.99} +04/19 [12:42:40] INFO | >> train_qwenlatent.py:487 + Step 3920 | grad_norm_pre_clip=0.4388 | + grad_norm_pre_clip_avg=0.3918 | Metrics: + {'align_loss': 0.02244662120938301, + 'recon_loss': 0.006729437038302422, + 'predict_loss': 0.024093806743621826, + 'aux_loss_decay_weight': 0.21619999999999995, + 'grad_norm_pre_clip': 0.43883317708969116, + 'data_time': 0.0010733280214481056, + 'model_time': 3.4302112750010565, + 'grad_norm_pre_clip_avg': 0.39176261723041533, + 'learning_rate': 1.9600000000000002e-05, + 'epoch': 0.99} +04/19 [12:43:18] INFO | >> train_qwenlatent.py:487 + Step 3930 | grad_norm_pre_clip=0.3840 | + grad_norm_pre_clip_avg=0.4169 | Metrics: + {'align_loss': 0.0217705387622118, + 'recon_loss': 0.006141718477010727, + 'predict_loss': 0.01849140226840973, + 'aux_loss_decay_weight': 0.21419999999999995, + 'grad_norm_pre_clip': 0.38403600454330444, + 'data_time': 0.004426113999215886, + 'model_time': 3.8661703329999, + 'grad_norm_pre_clip_avg': 0.4168755501508713, + 'learning_rate': 1.9650000000000003e-05, + 'epoch': 0.99} +04/19 [12:43:53] INFO | >> train_qwenlatent.py:487 + Step 3940 | grad_norm_pre_clip=0.3577 | + grad_norm_pre_clip_avg=0.3995 | Metrics: + {'align_loss': 0.022683173418045044, + 'recon_loss': 0.011752677150070667, + 'predict_loss': 0.02665073052048683, + 'aux_loss_decay_weight': 0.21220000000000006, + 'grad_norm_pre_clip': 0.3577375113964081, + 'data_time': 0.0012204410159029067, + 'model_time': 3.426389008993283, + 'grad_norm_pre_clip_avg': 0.39952164590358735, + 'learning_rate': 1.97e-05, 'epoch': 0.99} +04/19 [12:44:25] INFO | >> train_qwenlatent.py:487 + Step 3950 | grad_norm_pre_clip=0.2960 | + grad_norm_pre_clip_avg=0.3611 | Metrics: + {'align_loss': 0.020371535792946815, + 'recon_loss': 0.007779051084071398, + 'predict_loss': 0.023139499127864838, + 'aux_loss_decay_weight': 0.21020000000000005, + 'grad_norm_pre_clip': 0.2960410714149475, + 'mae_score': 0.04597098891799514, 'data_time': + 0.001097500993637368, 'model_time': + 3.556059513997752, 'grad_norm_pre_clip_avg': + 0.36113885045051575, 'learning_rate': + 1.9750000000000002e-05, 'epoch': 1.0} +04/19 [12:44:50] INFO | >> train_qwenlatent.py:487 + Step 3960 | grad_norm_pre_clip=0.4730 | + grad_norm_pre_clip_avg=0.4320 | Metrics: + {'align_loss': 0.021649736911058426, + 'recon_loss': 0.0078010112047195435, + 'predict_loss': 0.02807014435529709, + 'aux_loss_decay_weight': 0.20820000000000005, + 'grad_norm_pre_clip': 0.47303587198257446, + 'data_time': 0.0011239449959248304, + 'model_time': 2.0536607859830838, + 'grad_norm_pre_clip_avg': 0.43199690878391267, + 'learning_rate': 1.9800000000000004e-05, + 'epoch': 1.0} +04/19 [12:45:08] INFO | >> train_qwenlatent.py:487 + Step 3970 | grad_norm_pre_clip=0.3012 | + grad_norm_pre_clip_avg=0.3579 | Metrics: + {'align_loss': 0.020484380424022675, + 'recon_loss': 0.005501417443156242, + 'predict_loss': 0.017819559201598167, + 'aux_loss_decay_weight': 0.20620000000000005, + 'grad_norm_pre_clip': 0.30118921399116516, + 'data_time': 0.001199207006720826, + 'model_time': 1.5818217259948142, + 'grad_norm_pre_clip_avg': 0.35791180431842806, + 'learning_rate': 1.985e-05, 'epoch': 1.0} +04/19 [12:45:22] INFO | >> train_qwenlatent.py:487 + Step 3980 | grad_norm_pre_clip=0.3651 | + grad_norm_pre_clip_avg=0.3653 | Metrics: + {'align_loss': 0.021874239668250084, + 'recon_loss': 0.0061110579408705235, + 'predict_loss': 0.02357945218682289, + 'aux_loss_decay_weight': 0.20420000000000005, + 'grad_norm_pre_clip': 0.3650799095630646, + 'data_time': 0.0008628710056655109, + 'model_time': 1.2650579829933122, + 'grad_norm_pre_clip_avg': 0.3652777552604675, + 'learning_rate': 1.9900000000000003e-05, + 'epoch': 1.0} +04/19 [12:45:35] INFO | >> train_qwenlatent.py:487 + Step 3990 | grad_norm_pre_clip=0.3950 | + grad_norm_pre_clip_avg=0.3778 | Metrics: + {'align_loss': 0.02241332270205021, + 'recon_loss': 0.012784103862941265, + 'predict_loss': 0.031767118722200394, + 'aux_loss_decay_weight': 0.20220000000000005, + 'grad_norm_pre_clip': 0.3949556052684784, + 'data_time': 0.0008319510088767856, + 'model_time': 1.18891387199983, + 'grad_norm_pre_clip_avg': 0.3777835428714752, + 'learning_rate': 1.995e-05, 'epoch': 1.01} +04/19 [12:45:48] INFO | >> train_qwenlatent.py:487 + Step 4000 | grad_norm_pre_clip=0.4315 | + grad_norm_pre_clip_avg=0.4178 | Metrics: + {'align_loss': 0.02194255404174328, + 'recon_loss': 0.005987800657749176, + 'predict_loss': 0.020394258201122284, + 'aux_loss_decay_weight': 0.20020000000000004, + 'grad_norm_pre_clip': 0.4314641058444977, + 'mae_score': 0.038586233328054616, 'data_time': + 0.0006619069899898022, 'model_time': + 1.226047240983462, 'grad_norm_pre_clip_avg': + 0.41779820919036864, 'learning_rate': 2e-05, + 'epoch': 1.01} +04/19 [12:46:01] INFO | >> train_qwenlatent.py:487 + Step 4010 | grad_norm_pre_clip=0.3956 | + grad_norm_pre_clip_avg=0.3708 | Metrics: + {'align_loss': 0.020938390865921974, + 'recon_loss': 0.008984241634607315, + 'predict_loss': 0.026664989069104195, + 'aux_loss_decay_weight': 0.19820000000000004, + 'grad_norm_pre_clip': 0.3955507278442383, + 'data_time': 0.0007919690106064081, + 'model_time': 1.221472253993852, + 'grad_norm_pre_clip_avg': 0.3708272397518158, + 'learning_rate': 2.0050000000000003e-05, + 'epoch': 1.01} +04/19 [12:46:13] INFO | >> train_qwenlatent.py:487 + Step 4020 | grad_norm_pre_clip=0.4504 | + grad_norm_pre_clip_avg=0.3629 | Metrics: + {'align_loss': 0.02066398784518242, + 'recon_loss': 0.008496652357280254, + 'predict_loss': 0.027760228142142296, + 'aux_loss_decay_weight': 0.19620000000000004, + 'grad_norm_pre_clip': 0.4504355490207672, + 'data_time': 0.000658784992992878, + 'model_time': 1.1975936420203652, + 'grad_norm_pre_clip_avg': 0.362924262881279, + 'learning_rate': 2.01e-05, 'epoch': 1.01} +04/19 [12:46:26] INFO | >> train_qwenlatent.py:487 + Step 4030 | grad_norm_pre_clip=0.4461 | + grad_norm_pre_clip_avg=0.3880 | Metrics: + {'align_loss': 0.02057698927819729, + 'recon_loss': 0.0032921519596129656, + 'predict_loss': 0.017860034480690956, + 'aux_loss_decay_weight': 0.19420000000000004, + 'grad_norm_pre_clip': 0.4461052715778351, + 'data_time': 0.0007017520256340504, + 'model_time': 1.2058585789927747, + 'grad_norm_pre_clip_avg': 0.38803583681583403, + 'learning_rate': 2.0150000000000002e-05, + 'epoch': 1.02} +04/19 [12:46:38] INFO | >> train_qwenlatent.py:487 + Step 4040 | grad_norm_pre_clip=0.4142 | + grad_norm_pre_clip_avg=0.3980 | Metrics: + {'align_loss': 0.022701874375343323, + 'recon_loss': 0.01462083775550127, + 'predict_loss': 0.033690404146909714, + 'aux_loss_decay_weight': 0.19220000000000004, + 'grad_norm_pre_clip': 0.4142037034034729, + 'data_time': 0.0005897239898331463, + 'model_time': 1.2481668349937536, + 'grad_norm_pre_clip_avg': 0.3979740053415298, + 'learning_rate': 2.0200000000000003e-05, + 'epoch': 1.02} +04/19 [12:46:52] INFO | >> train_qwenlatent.py:487 + Step 4050 | grad_norm_pre_clip=0.4087 | + grad_norm_pre_clip_avg=0.4111 | Metrics: + {'align_loss': 0.022087417542934418, + 'recon_loss': 0.010572951287031174, + 'predict_loss': 0.03015482798218727, + 'aux_loss_decay_weight': 0.19020000000000004, + 'grad_norm_pre_clip': 0.40870827436447144, + 'mae_score': 0.047455552247193486, 'data_time': + 0.000851514982059598, 'model_time': + 1.2246549080009572, 'grad_norm_pre_clip_avg': + 0.4111453860998154, 'learning_rate': 2.025e-05, + 'epoch': 1.02} +04/19 [12:47:04] INFO | >> train_qwenlatent.py:487 + Step 4060 | grad_norm_pre_clip=0.4104 | + grad_norm_pre_clip_avg=0.4250 | Metrics: + {'align_loss': 0.0206174086779356, + 'recon_loss': 0.0032783925998955965, + 'predict_loss': 0.020107369869947433, + 'aux_loss_decay_weight': 0.18820000000000003, + 'grad_norm_pre_clip': 0.41042497754096985, + 'data_time': 0.000664619990857318, + 'model_time': 1.2045991880004294, + 'grad_norm_pre_clip_avg': 0.42502268552780154, + 'learning_rate': 2.0300000000000002e-05, + 'epoch': 1.02} +04/19 [12:47:17] INFO | >> train_qwenlatent.py:487 + Step 4070 | grad_norm_pre_clip=0.3793 | + grad_norm_pre_clip_avg=0.3791 | Metrics: + {'align_loss': 0.020141318440437317, + 'recon_loss': 0.00470740906894207, + 'predict_loss': 0.017745649442076683, + 'aux_loss_decay_weight': 0.18620000000000003, + 'grad_norm_pre_clip': 0.3793131709098816, + 'data_time': 0.0008371149888262153, + 'model_time': 1.2441382550168782, + 'grad_norm_pre_clip_avg': 0.37906796038150786, + 'learning_rate': 2.035e-05, 'epoch': 1.03} +04/19 [12:47:29] INFO | >> train_qwenlatent.py:487 + Step 4080 | grad_norm_pre_clip=0.3798 | + grad_norm_pre_clip_avg=0.3412 | Metrics: + {'align_loss': 0.02025974541902542, + 'recon_loss': 0.009988496080040932, + 'predict_loss': 0.026200538501143456, + 'aux_loss_decay_weight': 0.18420000000000003, + 'grad_norm_pre_clip': 0.3797619640827179, + 'data_time': 0.0007101549999788404, + 'model_time': 1.2331133609986864, + 'grad_norm_pre_clip_avg': 0.34124750792980196, + 'learning_rate': 2.04e-05, 'epoch': 1.03} +04/19 [12:47:42] INFO | >> train_qwenlatent.py:487 + Step 4090 | grad_norm_pre_clip=0.3489 | + grad_norm_pre_clip_avg=0.3649 | Metrics: + {'align_loss': 0.020812969654798508, + 'recon_loss': 0.006590033881366253, + 'predict_loss': 0.02002836763858795, + 'aux_loss_decay_weight': 0.18220000000000003, + 'grad_norm_pre_clip': 0.3489084839820862, + 'data_time': 0.0010430630063638091, + 'model_time': 1.2987530849932227, + 'grad_norm_pre_clip_avg': 0.36485156416893005, + 'learning_rate': 2.045e-05, 'epoch': 1.03} +04/19 [12:47:55] INFO | >> train_qwenlatent.py:487 + Step 4100 | grad_norm_pre_clip=0.4439 | + grad_norm_pre_clip_avg=0.3947 | Metrics: + {'align_loss': 0.021780654788017273, + 'recon_loss': 0.005038310308009386, + 'predict_loss': 0.03233720734715462, + 'aux_loss_decay_weight': 0.18020000000000003, + 'grad_norm_pre_clip': 0.4438508450984955, + 'mae_score': 0.034587162464588615, 'data_time': + 0.0005773479933850467, 'model_time': + 1.2395131159864832, 'grad_norm_pre_clip_avg': + 0.3947369247674942, 'learning_rate': 2.05e-05, + 'epoch': 1.03} +04/19 [12:48:08] INFO | >> train_qwenlatent.py:487 + Step 4110 | grad_norm_pre_clip=0.4413 | + grad_norm_pre_clip_avg=0.3988 | Metrics: + {'align_loss': 0.018963268026709557, + 'recon_loss': 0.0061090197414159775, + 'predict_loss': 0.02192123606801033, + 'aux_loss_decay_weight': 0.17820000000000003, + 'grad_norm_pre_clip': 0.44134509563446045, + 'data_time': 0.0008927090093493462, + 'model_time': 1.288779736001743, + 'grad_norm_pre_clip_avg': 0.39879425466060636, + 'learning_rate': 2.055e-05, 'epoch': 1.04} +04/19 [12:48:20] INFO | >> train_qwenlatent.py:487 + Step 4120 | grad_norm_pre_clip=0.4109 | + grad_norm_pre_clip_avg=0.4042 | Metrics: + {'align_loss': 0.0223915446549654, + 'recon_loss': 0.007357962895184755, + 'predict_loss': 0.024945544078946114, + 'aux_loss_decay_weight': 0.17620000000000002, + 'grad_norm_pre_clip': 0.41092467308044434, + 'data_time': 0.0006821560091339052, + 'model_time': 1.2647032420209143, + 'grad_norm_pre_clip_avg': 0.404189532995224, + 'learning_rate': 2.06e-05, 'epoch': 1.04} +04/19 [12:48:33] INFO | >> train_qwenlatent.py:487 + Step 4130 | grad_norm_pre_clip=0.4011 | + grad_norm_pre_clip_avg=0.4298 | Metrics: + {'align_loss': 0.022412698715925217, + 'recon_loss': 0.006660944782197475, + 'predict_loss': 0.019109996035695076, + 'aux_loss_decay_weight': 0.17420000000000002, + 'grad_norm_pre_clip': 0.4011269211769104, + 'data_time': 0.0009265500120818615, + 'model_time': 1.23320354000316, + 'grad_norm_pre_clip_avg': 0.42978970110416415, + 'learning_rate': 2.065e-05, 'epoch': 1.04} +04/19 [12:48:46] INFO | >> train_qwenlatent.py:487 + Step 4140 | grad_norm_pre_clip=0.3076 | + grad_norm_pre_clip_avg=0.3562 | Metrics: + {'align_loss': 0.02162851393222809, + 'recon_loss': 0.006723239552229643, + 'predict_loss': 0.02432984672486782, + 'aux_loss_decay_weight': 0.17220000000000002, + 'grad_norm_pre_clip': 0.30762824416160583, + 'data_time': 0.0007412680133711547, + 'model_time': 1.2037302610115148, + 'grad_norm_pre_clip_avg': 0.3561635255813599, + 'learning_rate': 2.07e-05, 'epoch': 1.04} +04/19 [12:49:00] INFO | >> train_qwenlatent.py:487 + Step 4150 | grad_norm_pre_clip=0.3721 | + grad_norm_pre_clip_avg=0.3657 | Metrics: + {'align_loss': 0.022471843287348747, + 'recon_loss': 0.01090239454060793, + 'predict_loss': 0.02878725714981556, + 'aux_loss_decay_weight': 0.17020000000000002, + 'grad_norm_pre_clip': 0.3721452057361603, + 'mae_score': 0.04090403307665576, 'data_time': + 0.0006250219885259867, 'model_time': + 1.1866251489846036, 'grad_norm_pre_clip_avg': + 0.36566205620765685, 'learning_rate': + 2.075e-05, 'epoch': 1.05} +04/19 [12:49:12] INFO | >> train_qwenlatent.py:487 + Step 4160 | grad_norm_pre_clip=0.3567 | + grad_norm_pre_clip_avg=0.3759 | Metrics: + {'align_loss': 0.021485300734639168, + 'recon_loss': 0.00956621952354908, + 'predict_loss': 0.027194712311029434, + 'aux_loss_decay_weight': 0.16820000000000002, + 'grad_norm_pre_clip': 0.35667961835861206, + 'data_time': 0.0006696199998259544, + 'model_time': 1.2565207149891648, + 'grad_norm_pre_clip_avg': 0.37592536211013794, + 'learning_rate': 2.08e-05, 'epoch': 1.05} +04/19 [12:49:25] INFO | >> train_qwenlatent.py:487 + Step 4170 | grad_norm_pre_clip=0.4391 | + grad_norm_pre_clip_avg=0.3603 | Metrics: + {'align_loss': 0.022016998380422592, + 'recon_loss': 0.006967071443796158, + 'predict_loss': 0.025173263624310493, + 'aux_loss_decay_weight': 0.16620000000000001, + 'grad_norm_pre_clip': 0.43911170959472656, + 'data_time': 0.0009295179916080087, + 'model_time': 1.2309826980053913, + 'grad_norm_pre_clip_avg': 0.3602829784154892, + 'learning_rate': 2.085e-05, 'epoch': 1.05} +04/19 [12:49:37] INFO | >> train_qwenlatent.py:487 + Step 4180 | grad_norm_pre_clip=0.4524 | + grad_norm_pre_clip_avg=0.4120 | Metrics: + {'align_loss': 0.021660976111888885, + 'recon_loss': 0.010068114846944809, + 'predict_loss': 0.025814833119511604, + 'aux_loss_decay_weight': 0.1642, + 'grad_norm_pre_clip': 0.45242947340011597, + 'data_time': 0.0006425119936466217, + 'model_time': 1.2940960889973212, + 'grad_norm_pre_clip_avg': 0.4119760185480118, + 'learning_rate': 2.09e-05, 'epoch': 1.05} +04/19 [12:49:50] INFO | >> train_qwenlatent.py:487 + Step 4190 | grad_norm_pre_clip=0.4575 | + grad_norm_pre_clip_avg=0.3898 | Metrics: + {'align_loss': 0.021127723157405853, + 'recon_loss': 0.011086678132414818, + 'predict_loss': 0.02678522653877735, + 'aux_loss_decay_weight': 0.1622, + 'grad_norm_pre_clip': 0.4574756622314453, + 'data_time': 0.0008573159866500646, + 'model_time': 1.258852361002937, + 'grad_norm_pre_clip_avg': 0.38981102406978607, + 'learning_rate': 2.095e-05, 'epoch': 1.06} +04/19 [12:50:03] INFO | >> train_qwenlatent.py:487 + Step 4200 | grad_norm_pre_clip=0.3961 | + grad_norm_pre_clip_avg=0.3875 | Metrics: + {'align_loss': 0.019787032157182693, + 'recon_loss': 0.005448471289128065, + 'predict_loss': 0.023006727918982506, + 'aux_loss_decay_weight': 0.1602, + 'grad_norm_pre_clip': 0.39611998200416565, + 'mae_score': 0.048726344752956084, 'data_time': + 0.0009814169898163527, 'model_time': + 1.2807153719768394, 'grad_norm_pre_clip_avg': + 0.3875344395637512, 'learning_rate': 2.1e-05, + 'epoch': 1.06} +04/19 [12:50:16] INFO | >> train_qwenlatent.py:487 + Step 4210 | grad_norm_pre_clip=0.4310 | + grad_norm_pre_clip_avg=0.3835 | Metrics: + {'align_loss': 0.021165186539292336, + 'recon_loss': 0.011677483096718788, + 'predict_loss': 0.03165289759635925, + 'aux_loss_decay_weight': 0.1582, + 'grad_norm_pre_clip': 0.4309682250022888, + 'data_time': 0.0008350290008820593, + 'model_time': 1.2551969290070701, + 'grad_norm_pre_clip_avg': 0.38346017301082613, + 'learning_rate': 2.105e-05, 'epoch': 1.06} +04/19 [12:50:28] INFO | >> train_qwenlatent.py:487 + Step 4220 | grad_norm_pre_clip=0.3297 | + grad_norm_pre_clip_avg=0.3848 | Metrics: + {'align_loss': 0.021610170602798462, + 'recon_loss': 0.005942330230027437, + 'predict_loss': 0.025902684777975082, + 'aux_loss_decay_weight': 0.1562, + 'grad_norm_pre_clip': 0.32968947291374207, + 'data_time': 0.0005872769979760051, + 'model_time': 1.2221343590063043, + 'grad_norm_pre_clip_avg': 0.38478549420833585, + 'learning_rate': 2.11e-05, 'epoch': 1.06} +04/19 [12:50:41] INFO | >> train_qwenlatent.py:487 + Step 4230 | grad_norm_pre_clip=0.4034 | + grad_norm_pre_clip_avg=0.4102 | Metrics: + {'align_loss': 0.020786860957741737, + 'recon_loss': 0.006135277450084686, + 'predict_loss': 0.01940331794321537, + 'aux_loss_decay_weight': 0.1542, + 'grad_norm_pre_clip': 0.40337225794792175, + 'data_time': 0.0006950960087124258, + 'model_time': 1.3159027909860015, + 'grad_norm_pre_clip_avg': 0.41023644506931306, + 'learning_rate': 2.115e-05, 'epoch': 1.07} +04/19 [12:50:53] INFO | >> train_qwenlatent.py:487 + Step 4240 | grad_norm_pre_clip=0.3172 | + grad_norm_pre_clip_avg=0.3835 | Metrics: + {'align_loss': 0.020166635513305664, + 'recon_loss': 0.007764744106680155, + 'predict_loss': 0.020454512909054756, + 'aux_loss_decay_weight': 0.1522, + 'grad_norm_pre_clip': 0.3172418475151062, + 'data_time': 0.0006510349921882153, + 'model_time': 1.2323632129991893, + 'grad_norm_pre_clip_avg': 0.38349567353725433, + 'learning_rate': 2.12e-05, 'epoch': 1.07} +04/19 [12:51:07] INFO | >> train_qwenlatent.py:487 + Step 4250 | grad_norm_pre_clip=0.2979 | + grad_norm_pre_clip_avg=0.3346 | Metrics: + {'align_loss': 0.023539934307336807, + 'recon_loss': 0.007995222695171833, + 'predict_loss': 0.02393047697842121, + 'aux_loss_decay_weight': 0.1502, + 'grad_norm_pre_clip': 0.29791170358657837, + 'mae_score': 0.042199682974600576, 'data_time': + 0.0010596600186545402, 'model_time': + 1.2149020200013183, 'grad_norm_pre_clip_avg': + 0.3346080631017685, 'learning_rate': 2.125e-05, + 'epoch': 1.07} +04/19 [12:51:19] INFO | >> train_qwenlatent.py:487 + Step 4260 | grad_norm_pre_clip=0.3929 | + grad_norm_pre_clip_avg=0.3594 | Metrics: + {'align_loss': 0.020279332995414734, + 'recon_loss': 0.007768769282847643, + 'predict_loss': 0.02104528248310089, + 'aux_loss_decay_weight': 0.1482, + 'grad_norm_pre_clip': 0.3929286599159241, + 'data_time': 0.0005860999808646739, + 'model_time': 1.2524988990044221, + 'grad_norm_pre_clip_avg': 0.35941657423973083, + 'learning_rate': 2.13e-05, 'epoch': 1.07} +04/19 [12:51:32] INFO | >> train_qwenlatent.py:487 + Step 4270 | grad_norm_pre_clip=0.3095 | + grad_norm_pre_clip_avg=0.3650 | Metrics: + {'align_loss': 0.022856559604406357, + 'recon_loss': 0.012738589197397232, + 'predict_loss': 0.02604839950799942, + 'aux_loss_decay_weight': 0.1462, + 'grad_norm_pre_clip': 0.30948683619499207, + 'data_time': 0.0008452750043943524, + 'model_time': 1.2154768749896903, + 'grad_norm_pre_clip_avg': 0.36502387225627897, + 'learning_rate': 2.135e-05, 'epoch': 1.08} +04/19 [12:51:45] INFO | >> train_qwenlatent.py:487 + Step 4280 | grad_norm_pre_clip=0.3759 | + grad_norm_pre_clip_avg=0.3719 | Metrics: + {'align_loss': 0.021099261939525604, + 'recon_loss': 0.008826659992337227, + 'predict_loss': 0.026685111224651337, + 'aux_loss_decay_weight': 0.1442, + 'grad_norm_pre_clip': 0.37585312128067017, + 'data_time': 0.0006863090093247592, + 'model_time': 1.207461215002695, + 'grad_norm_pre_clip_avg': 0.3718894362449646, + 'learning_rate': 2.1400000000000002e-05, + 'epoch': 1.08} +04/19 [12:51:58] INFO | >> train_qwenlatent.py:487 + Step 4290 | grad_norm_pre_clip=0.4219 | + grad_norm_pre_clip_avg=0.4213 | Metrics: + {'align_loss': 0.02194235846400261, + 'recon_loss': 0.009450539946556091, + 'predict_loss': 0.025355162099003792, + 'aux_loss_decay_weight': 0.1422, + 'grad_norm_pre_clip': 0.4218698740005493, + 'data_time': 0.0007043869991321117, + 'model_time': 1.1884871689835563, + 'grad_norm_pre_clip_avg': 0.4213049441576004, + 'learning_rate': 2.145e-05, 'epoch': 1.08} +04/19 [12:52:11] INFO | >> train_qwenlatent.py:487 + Step 4300 | grad_norm_pre_clip=0.3342 | + grad_norm_pre_clip_avg=0.3756 | Metrics: + {'align_loss': 0.022528067231178284, + 'recon_loss': 0.006812576204538345, + 'predict_loss': 0.02646920271217823, + 'aux_loss_decay_weight': 0.1402, + 'grad_norm_pre_clip': 0.3341890573501587, + 'mae_score': 0.035166370976078616, 'data_time': + 0.001000390009721741, 'model_time': + 1.2981731769978069, 'grad_norm_pre_clip_avg': + 0.3755542427301407, 'learning_rate': 2.15e-05, + 'epoch': 1.09} +04/19 [12:52:23] INFO | >> train_qwenlatent.py:487 + Step 4310 | grad_norm_pre_clip=0.3728 | + grad_norm_pre_clip_avg=0.3863 | Metrics: + {'align_loss': 0.0211186483502388, + 'recon_loss': 0.0038296235725283623, + 'predict_loss': 0.017310025170445442, + 'aux_loss_decay_weight': 0.1382, + 'grad_norm_pre_clip': 0.37278857827186584, + 'data_time': 0.0006986890221014619, + 'model_time': 1.2332579639914911, + 'grad_norm_pre_clip_avg': 0.3863473415374756, + 'learning_rate': 2.1550000000000002e-05, + 'epoch': 1.09} +04/19 [12:52:35] INFO | >> train_qwenlatent.py:487 + Step 4320 | grad_norm_pre_clip=0.3524 | + grad_norm_pre_clip_avg=0.3385 | Metrics: + {'align_loss': 0.022021953016519547, + 'recon_loss': 0.006223961245268583, + 'predict_loss': 0.02200520969927311, + 'aux_loss_decay_weight': 0.1362, + 'grad_norm_pre_clip': 0.3524390757083893, + 'data_time': 0.0006326040020212531, + 'model_time': 1.2575346950034145, + 'grad_norm_pre_clip_avg': 0.33850614726543427, + 'learning_rate': 2.16e-05, 'epoch': 1.09} +04/19 [12:52:48] INFO | >> train_qwenlatent.py:487 + Step 4330 | grad_norm_pre_clip=0.3140 | + grad_norm_pre_clip_avg=0.3383 | Metrics: + {'align_loss': 0.02129228413105011, + 'recon_loss': 0.006125759799033403, + 'predict_loss': 0.020057225599884987, + 'aux_loss_decay_weight': 0.13419999999999999, + 'grad_norm_pre_clip': 0.3140241503715515, + 'data_time': 0.0006906700145918876, + 'model_time': 1.2306398490036372, + 'grad_norm_pre_clip_avg': 0.33832843601703644, + 'learning_rate': 2.165e-05, 'epoch': 1.09} +04/19 [12:53:00] INFO | >> train_qwenlatent.py:487 + Step 4340 | grad_norm_pre_clip=0.3345 | + grad_norm_pre_clip_avg=0.3655 | Metrics: + {'align_loss': 0.022069858387112617, + 'recon_loss': 0.006767879705876112, + 'predict_loss': 0.023404350504279137, + 'aux_loss_decay_weight': 0.13219999999999998, + 'grad_norm_pre_clip': 0.3345385491847992, + 'data_time': 0.0007510480063501745, + 'model_time': 1.2247433429874945, + 'grad_norm_pre_clip_avg': 0.3654716074466705, + 'learning_rate': 2.1700000000000002e-05, + 'epoch': 1.1} +04/19 [12:53:14] INFO | >> train_qwenlatent.py:487 + Step 4350 | grad_norm_pre_clip=0.3449 | + grad_norm_pre_clip_avg=0.4300 | Metrics: + {'align_loss': 0.021784180775284767, + 'recon_loss': 0.01236420776695013, + 'predict_loss': 0.029712241142988205, + 'aux_loss_decay_weight': 0.13019999999999998, + 'grad_norm_pre_clip': 0.34487104415893555, + 'mae_score': 0.036837950697890275, 'data_time': + 0.0010748360073193908, 'model_time': + 1.1809720510209445, 'grad_norm_pre_clip_avg': + 0.43002321720123293, 'learning_rate': + 2.175e-05, 'epoch': 1.1} +04/19 [12:53:26] INFO | >> train_qwenlatent.py:487 + Step 4360 | grad_norm_pre_clip=0.4018 | + grad_norm_pre_clip_avg=0.3909 | Metrics: + {'align_loss': 0.023387420922517776, + 'recon_loss': 0.008353861048817635, + 'predict_loss': 0.023399656638503075, + 'aux_loss_decay_weight': 0.12819999999999998, + 'grad_norm_pre_clip': 0.4017769694328308, + 'data_time': 0.0007336649869102985, + 'model_time': 1.2140266530041117, + 'grad_norm_pre_clip_avg': 0.3908920258283615, + 'learning_rate': 2.18e-05, 'epoch': 1.1} +04/19 [12:53:39] INFO | >> train_qwenlatent.py:487 + Step 4370 | grad_norm_pre_clip=0.4385 | + grad_norm_pre_clip_avg=0.4025 | Metrics: + {'align_loss': 0.021232152357697487, + 'recon_loss': 0.0049984934739768505, + 'predict_loss': 0.025382649153470993, + 'aux_loss_decay_weight': 0.12619999999999998, + 'grad_norm_pre_clip': 0.4385257661342621, + 'data_time': 0.000854737008921802, + 'model_time': 1.2205151289817877, + 'grad_norm_pre_clip_avg': 0.4024951994419098, + 'learning_rate': 2.1850000000000003e-05, + 'epoch': 1.1} +04/19 [12:53:51] INFO | >> train_qwenlatent.py:487 + Step 4380 | grad_norm_pre_clip=0.3648 | + grad_norm_pre_clip_avg=0.3661 | Metrics: + {'align_loss': 0.022242404520511627, + 'recon_loss': 0.008633988909423351, + 'predict_loss': 0.02514275535941124, + 'aux_loss_decay_weight': 0.12419999999999998, + 'grad_norm_pre_clip': 0.3647885322570801, + 'data_time': 0.0008207389910239726, + 'model_time': 1.2958435170003213, + 'grad_norm_pre_clip_avg': 0.3660769462585449, + 'learning_rate': 2.19e-05, 'epoch': 1.11} +04/19 [12:54:04] INFO | >> train_qwenlatent.py:487 + Step 4390 | grad_norm_pre_clip=0.3844 | + grad_norm_pre_clip_avg=0.3584 | Metrics: + {'align_loss': 0.02233114466071129, + 'recon_loss': 0.008190886117517948, + 'predict_loss': 0.025292983278632164, + 'aux_loss_decay_weight': 0.12219999999999998, + 'grad_norm_pre_clip': 0.3843969404697418, + 'data_time': 0.0005974219820927829, + 'model_time': 1.1924162030045409, + 'grad_norm_pre_clip_avg': 0.35843369364738464, + 'learning_rate': 2.195e-05, 'epoch': 1.11} +04/19 [12:54:17] INFO | >> train_qwenlatent.py:487 + Step 4400 | grad_norm_pre_clip=0.3346 | + grad_norm_pre_clip_avg=0.3159 | Metrics: + {'align_loss': 0.020140264183282852, + 'recon_loss': 0.0073292311280965805, + 'predict_loss': 0.01828613132238388, + 'aux_loss_decay_weight': 0.12019999999999997, + 'grad_norm_pre_clip': 0.3346463441848755, + 'mae_score': 0.05379004263663077, 'data_time': + 0.000830519013106823, 'model_time': + 1.21125783098978, 'grad_norm_pre_clip_avg': + 0.3159203976392746, 'learning_rate': + 2.2000000000000003e-05, 'epoch': 1.11} +04/19 [12:54:30] INFO | >> train_qwenlatent.py:487 + Step 4410 | grad_norm_pre_clip=0.3679 | + grad_norm_pre_clip_avg=0.3736 | Metrics: + {'align_loss': 0.022431587800383568, + 'recon_loss': 0.012080737389624119, + 'predict_loss': 0.025550508871674538, + 'aux_loss_decay_weight': 0.11819999999999997, + 'grad_norm_pre_clip': 0.3679032623767853, + 'data_time': 0.0009327299776487052, + 'model_time': 1.5113589219981804, + 'grad_norm_pre_clip_avg': 0.3736352354288101, + 'learning_rate': 2.205e-05, 'epoch': 1.11} +04/19 [12:54:43] INFO | >> train_qwenlatent.py:487 + Step 4420 | grad_norm_pre_clip=0.3840 | + grad_norm_pre_clip_avg=0.4071 | Metrics: + {'align_loss': 0.022355979308485985, + 'recon_loss': 0.00919682253152132, + 'predict_loss': 0.026563651859760284, + 'aux_loss_decay_weight': 0.11619999999999997, + 'grad_norm_pre_clip': 0.38401317596435547, + 'data_time': 0.000773428997490555, + 'model_time': 1.2756367819965817, + 'grad_norm_pre_clip_avg': 0.40710062980651857, + 'learning_rate': 2.2100000000000002e-05, + 'epoch': 1.12} +04/19 [12:54:56] INFO | >> train_qwenlatent.py:487 + Step 4430 | grad_norm_pre_clip=0.3132 | + grad_norm_pre_clip_avg=0.3429 | Metrics: + {'align_loss': 0.02218709886074066, + 'recon_loss': 0.0051751043647527695, + 'predict_loss': 0.02310059405863285, + 'aux_loss_decay_weight': 0.11419999999999997, + 'grad_norm_pre_clip': 0.3131999373435974, + 'data_time': 0.0006616589962504804, + 'model_time': 1.2240349460043944, + 'grad_norm_pre_clip_avg': 0.34293616414070127, + 'learning_rate': 2.215e-05, 'epoch': 1.12} +04/19 [12:55:08] INFO | >> train_qwenlatent.py:487 + Step 4440 | grad_norm_pre_clip=0.4411 | + grad_norm_pre_clip_avg=0.3621 | Metrics: + {'align_loss': 0.02127062901854515, + 'recon_loss': 0.009805480018258095, + 'predict_loss': 0.023803824558854103, + 'aux_loss_decay_weight': 0.11219999999999997, + 'grad_norm_pre_clip': 0.4410885274410248, + 'data_time': 0.0007373479893431067, + 'model_time': 1.2570249829732347, + 'grad_norm_pre_clip_avg': 0.3620953619480133, + 'learning_rate': 2.22e-05, 'epoch': 1.12} +04/19 [12:55:21] INFO | >> train_qwenlatent.py:487 + Step 4450 | grad_norm_pre_clip=0.3593 | + grad_norm_pre_clip_avg=0.3550 | Metrics: + {'align_loss': 0.0212920680642128, + 'recon_loss': 0.00812566839158535, + 'predict_loss': 0.021732836961746216, + 'aux_loss_decay_weight': 0.11019999999999996, + 'grad_norm_pre_clip': 0.359274297952652, + 'mae_score': 0.04878685753624718, 'data_time': + 0.0006426240142900497, 'model_time': + 1.2394146450096741, 'grad_norm_pre_clip_avg': + 0.3550026834011078, 'learning_rate': + 2.2250000000000002e-05, 'epoch': 1.12} +04/19 [12:55:33] INFO | >> train_qwenlatent.py:487 + Step 4460 | grad_norm_pre_clip=0.3138 | + grad_norm_pre_clip_avg=0.4782 | Metrics: + {'align_loss': 0.021202798932790756, + 'recon_loss': 0.00566661823540926, + 'predict_loss': 0.02196647599339485, + 'aux_loss_decay_weight': 0.10819999999999996, + 'grad_norm_pre_clip': 0.3138015866279602, + 'data_time': 0.0008971700153779238, + 'model_time': 1.2705038810090628, + 'grad_norm_pre_clip_avg': 0.4781879812479019, + 'learning_rate': 2.23e-05, 'epoch': 1.13} +04/19 [12:55:46] INFO | >> train_qwenlatent.py:487 + Step 4470 | grad_norm_pre_clip=0.2982 | + grad_norm_pre_clip_avg=0.3938 | Metrics: + {'align_loss': 0.02179095335304737, + 'recon_loss': 0.006732895504683256, + 'predict_loss': 0.02194727025926113, + 'aux_loss_decay_weight': 0.10619999999999996, + 'grad_norm_pre_clip': 0.2982402443885803, + 'data_time': 0.0006611979915760458, + 'model_time': 1.2181722960085608, + 'grad_norm_pre_clip_avg': 0.39376786053180696, + 'learning_rate': 2.235e-05, 'epoch': 1.13} +04/19 [12:55:59] INFO | >> train_qwenlatent.py:487 + Step 4480 | grad_norm_pre_clip=0.4216 | + grad_norm_pre_clip_avg=0.3470 | Metrics: + {'align_loss': 0.02316129580140114, + 'recon_loss': 0.010582927614450455, + 'predict_loss': 0.02302747592329979, + 'aux_loss_decay_weight': 0.10419999999999996, + 'grad_norm_pre_clip': 0.4215576648712158, + 'data_time': 0.0008760599885135889, + 'model_time': 1.235994127986487, + 'grad_norm_pre_clip_avg': 0.3470346599817276, + 'learning_rate': 2.2400000000000002e-05, + 'epoch': 1.13} +04/19 [12:56:12] INFO | >> train_qwenlatent.py:487 + Step 4490 | grad_norm_pre_clip=0.2856 | + grad_norm_pre_clip_avg=0.3604 | Metrics: + {'align_loss': 0.02214117720723152, + 'recon_loss': 0.007590487599372864, + 'predict_loss': 0.024811074137687683, + 'aux_loss_decay_weight': 0.10219999999999996, + 'grad_norm_pre_clip': 0.2855500280857086, + 'data_time': 0.0006392369978129864, + 'model_time': 1.2642444939992856, + 'grad_norm_pre_clip_avg': 0.3604364275932312, + 'learning_rate': 2.245e-05, 'epoch': 1.13} +04/19 [12:56:25] INFO | >> train_qwenlatent.py:487 + Step 4500 | grad_norm_pre_clip=0.3207 | + grad_norm_pre_clip_avg=0.3707 | Metrics: + {'align_loss': 0.02050166204571724, + 'recon_loss': 0.0056942836381495, + 'predict_loss': 0.023811673745512962, + 'aux_loss_decay_weight': 0.10019999999999996, + 'grad_norm_pre_clip': 0.32067352533340454, + 'mae_score': 0.04039774370623064, 'data_time': + 0.0006041390006430447, 'model_time': + 1.2187991029932164, 'grad_norm_pre_clip_avg': + 0.3706668227910995, 'learning_rate': 2.25e-05, + 'epoch': 1.14} +04/19 [12:56:38] INFO | >> train_qwenlatent.py:487 + Step 4510 | grad_norm_pre_clip=0.3955 | + grad_norm_pre_clip_avg=0.3592 | Metrics: + {'align_loss': 0.020692819729447365, + 'recon_loss': 0.007858515717089176, + 'predict_loss': 0.0247336495667696, + 'aux_loss_decay_weight': 0.09819999999999995, + 'grad_norm_pre_clip': 0.3955150544643402, + 'data_time': 0.0009125599754042923, + 'model_time': 1.2319267420098186, + 'grad_norm_pre_clip_avg': 0.3592249482870102, + 'learning_rate': 2.2550000000000003e-05, + 'epoch': 1.14} +04/19 [12:56:50] INFO | >> train_qwenlatent.py:487 + Step 4520 | grad_norm_pre_clip=0.3778 | + grad_norm_pre_clip_avg=0.3384 | Metrics: + {'align_loss': 0.020140834152698517, + 'recon_loss': 0.006336772348731756, + 'predict_loss': 0.021524563431739807, + 'aux_loss_decay_weight': 0.09619999999999995, + 'grad_norm_pre_clip': 0.37780627608299255, + 'data_time': 0.0006720350065734237, + 'model_time': 1.2600201229797676, + 'grad_norm_pre_clip_avg': 0.3384143203496933, + 'learning_rate': 2.26e-05, 'epoch': 1.14} +04/19 [12:57:03] INFO | >> train_qwenlatent.py:487 + Step 4530 | grad_norm_pre_clip=0.2829 | + grad_norm_pre_clip_avg=0.3166 | Metrics: + {'align_loss': 0.021281205117702484, + 'recon_loss': 0.005881795194000006, + 'predict_loss': 0.022706838324666023, + 'aux_loss_decay_weight': 0.09419999999999995, + 'grad_norm_pre_clip': 0.2828781306743622, + 'data_time': 0.0006163550133351237, + 'model_time': 1.29603909997968, + 'grad_norm_pre_clip_avg': 0.3166350871324539, + 'learning_rate': 2.265e-05, 'epoch': 1.14} +04/19 [12:57:15] INFO | >> train_qwenlatent.py:487 + Step 4540 | grad_norm_pre_clip=0.3783 | + grad_norm_pre_clip_avg=0.3557 | Metrics: + {'align_loss': 0.022038772702217102, + 'recon_loss': 0.009020215831696987, + 'predict_loss': 0.02316369116306305, + 'aux_loss_decay_weight': 0.09219999999999995, + 'grad_norm_pre_clip': 0.37829917669296265, + 'data_time': 0.0007218639948405325, + 'model_time': 1.2109330919920467, + 'grad_norm_pre_clip_avg': 0.3556923925876617, + 'learning_rate': 2.2700000000000003e-05, + 'epoch': 1.15} +04/19 [12:57:29] INFO | >> train_qwenlatent.py:487 + Step 4550 | grad_norm_pre_clip=0.3194 | + grad_norm_pre_clip_avg=0.3964 | Metrics: + {'align_loss': 0.02242162451148033, + 'recon_loss': 0.008698944933712482, + 'predict_loss': 0.02397567592561245, + 'aux_loss_decay_weight': 0.09019999999999995, + 'grad_norm_pre_clip': 0.3194313943386078, + 'mae_score': 0.032697083069397524, 'data_time': + 0.0014305950026027858, 'model_time': + 1.2353288290032651, 'grad_norm_pre_clip_avg': + 0.3963699251413345, 'learning_rate': 2.275e-05, + 'epoch': 1.15} +04/19 [12:57:41] INFO | >> train_qwenlatent.py:487 + Step 4560 | grad_norm_pre_clip=0.3386 | + grad_norm_pre_clip_avg=0.3214 | Metrics: + {'align_loss': 0.02165822871029377, + 'recon_loss': 0.010781255550682545, + 'predict_loss': 0.028910605236887932, + 'aux_loss_decay_weight': 0.08819999999999995, + 'grad_norm_pre_clip': 0.33860838413238525, + 'data_time': 0.0009701010130811483, + 'model_time': 1.1963013819768094, + 'grad_norm_pre_clip_avg': 0.32139008939266206, + 'learning_rate': 2.2800000000000002e-05, + 'epoch': 1.15} +04/19 [12:57:54] INFO | >> train_qwenlatent.py:487 + Step 4570 | grad_norm_pre_clip=0.3510 | + grad_norm_pre_clip_avg=0.3378 | Metrics: + {'align_loss': 0.020175384357571602, + 'recon_loss': 0.007365859113633633, + 'predict_loss': 0.02445567399263382, + 'aux_loss_decay_weight': 0.08620000000000005, + 'grad_norm_pre_clip': 0.3510276973247528, + 'data_time': 0.0007992159808054566, + 'model_time': 1.249511846981477, + 'grad_norm_pre_clip_avg': 0.3378474175930023, + 'learning_rate': 2.2850000000000003e-05, + 'epoch': 1.15} +04/19 [12:58:07] INFO | >> train_qwenlatent.py:487 + Step 4580 | grad_norm_pre_clip=0.4450 | + grad_norm_pre_clip_avg=0.3768 | Metrics: + {'align_loss': 0.02112451195716858, + 'recon_loss': 0.006353024393320084, + 'predict_loss': 0.020152674987912178, + 'aux_loss_decay_weight': 0.08420000000000005, + 'grad_norm_pre_clip': 0.44497576355934143, + 'data_time': 0.0007034089940134436, + 'model_time': 1.2878259319986682, + 'grad_norm_pre_clip_avg': 0.37675937116146085, + 'learning_rate': 2.29e-05, 'epoch': 1.16} +04/19 [12:58:19] INFO | >> train_qwenlatent.py:487 + Step 4590 | grad_norm_pre_clip=0.3431 | + grad_norm_pre_clip_avg=0.3561 | Metrics: + {'align_loss': 0.020822763442993164, + 'recon_loss': 0.007678641472011805, + 'predict_loss': 0.02038281410932541, + 'aux_loss_decay_weight': 0.08220000000000005, + 'grad_norm_pre_clip': 0.34307360649108887, + 'data_time': 0.0007997410139068961, + 'model_time': 1.2655915990180802, + 'grad_norm_pre_clip_avg': 0.3560552477836609, + 'learning_rate': 2.2950000000000002e-05, + 'epoch': 1.16} +04/19 [12:58:32] INFO | >> train_qwenlatent.py:487 + Step 4600 | grad_norm_pre_clip=0.4015 | + grad_norm_pre_clip_avg=0.3572 | Metrics: + {'align_loss': 0.02213684469461441, + 'recon_loss': 0.010069304145872593, + 'predict_loss': 0.02561137266457081, + 'aux_loss_decay_weight': 0.08020000000000005, + 'grad_norm_pre_clip': 0.4014701843261719, + 'mae_score': 0.03351684432845932, 'data_time': + 0.0006044500041753054, 'model_time': + 1.2126531590183731, 'grad_norm_pre_clip_avg': + 0.3571998953819275, 'learning_rate': + 2.3000000000000003e-05, 'epoch': 1.16} +04/19 [12:58:45] INFO | >> train_qwenlatent.py:487 + Step 4610 | grad_norm_pre_clip=0.3355 | + grad_norm_pre_clip_avg=0.3390 | Metrics: + {'align_loss': 0.02158581279218197, + 'recon_loss': 0.006639046128839254, + 'predict_loss': 0.023759890347719193, + 'aux_loss_decay_weight': 0.07820000000000005, + 'grad_norm_pre_clip': 0.3354780972003937, + 'data_time': 0.00059710000641644, 'model_time': + 1.2628620159812272, 'grad_norm_pre_clip_avg': + 0.33898232877254486, 'learning_rate': + 2.305e-05, 'epoch': 1.16} +04/19 [12:58:58] INFO | >> train_qwenlatent.py:487 + Step 4620 | grad_norm_pre_clip=0.3202 | + grad_norm_pre_clip_avg=0.3841 | Metrics: + {'align_loss': 0.020330190658569336, + 'recon_loss': 0.008169044740498066, + 'predict_loss': 0.02207653783261776, + 'aux_loss_decay_weight': 0.07620000000000005, + 'grad_norm_pre_clip': 0.3202231824398041, + 'data_time': 0.0005814699979964644, + 'model_time': 1.2081688370089978, + 'grad_norm_pre_clip_avg': 0.38410134613513947, + 'learning_rate': 2.3100000000000002e-05, + 'epoch': 1.17} +04/19 [12:59:11] INFO | >> train_qwenlatent.py:487 + Step 4630 | grad_norm_pre_clip=0.3293 | + grad_norm_pre_clip_avg=0.3885 | Metrics: + {'align_loss': 0.02239542454481125, + 'recon_loss': 0.00791227351874113, + 'predict_loss': 0.025527361780405045, + 'aux_loss_decay_weight': 0.07420000000000004, + 'grad_norm_pre_clip': 0.32932329177856445, + 'data_time': 0.0006287549913395196, + 'model_time': 1.253199017024599, + 'grad_norm_pre_clip_avg': 0.38847369253635405, + 'learning_rate': 2.3150000000000004e-05, + 'epoch': 1.17} +04/19 [12:59:23] INFO | >> train_qwenlatent.py:487 + Step 4640 | grad_norm_pre_clip=0.3165 | + grad_norm_pre_clip_avg=0.3311 | Metrics: + {'align_loss': 0.021974947303533554, + 'recon_loss': 0.00859365239739418, + 'predict_loss': 0.023703986778855324, + 'aux_loss_decay_weight': 0.07220000000000004, + 'grad_norm_pre_clip': 0.31648045778274536, + 'data_time': 0.0007033089932519943, + 'model_time': 1.2829495060141198, + 'grad_norm_pre_clip_avg': 0.3310867816209793, + 'learning_rate': 2.32e-05, 'epoch': 1.17} +04/19 [12:59:37] INFO | >> train_qwenlatent.py:487 + Step 4650 | grad_norm_pre_clip=0.3704 | + grad_norm_pre_clip_avg=0.2918 | Metrics: + {'align_loss': 0.022168640047311783, + 'recon_loss': 0.00823845062404871, + 'predict_loss': 0.023358741775155067, + 'aux_loss_decay_weight': 0.07020000000000004, + 'grad_norm_pre_clip': 0.3703787624835968, + 'mae_score': 0.028017635173625773, 'data_time': + 0.0010050769778899848, 'model_time': + 1.2719467380084097, 'grad_norm_pre_clip_avg': + 0.2917701184749603, 'learning_rate': + 2.3250000000000003e-05, 'epoch': 1.17} +04/19 [12:59:49] INFO | >> train_qwenlatent.py:487 + Step 4660 | grad_norm_pre_clip=0.3183 | + grad_norm_pre_clip_avg=0.3222 | Metrics: + {'align_loss': 0.02191309630870819, + 'recon_loss': 0.007655306253582239, + 'predict_loss': 0.022306237369775772, + 'aux_loss_decay_weight': 0.06820000000000004, + 'grad_norm_pre_clip': 0.3182887136936188, + 'data_time': 0.001044217002345249, + 'model_time': 1.253808070992818, + 'grad_norm_pre_clip_avg': 0.3222473278641701, + 'learning_rate': 2.3300000000000004e-05, + 'epoch': 1.18} +04/19 [13:00:02] INFO | >> train_qwenlatent.py:487 + Step 4670 | grad_norm_pre_clip=0.3647 | + grad_norm_pre_clip_avg=0.3436 | Metrics: + {'align_loss': 0.02139279432594776, + 'recon_loss': 0.0059788464568555355, + 'predict_loss': 0.02095622569322586, + 'aux_loss_decay_weight': 0.06620000000000004, + 'grad_norm_pre_clip': 0.36473846435546875, + 'data_time': 0.0008685030043125153, + 'model_time': 1.2431577070092317, + 'grad_norm_pre_clip_avg': 0.3436066806316376, + 'learning_rate': 2.3350000000000002e-05, + 'epoch': 1.18} +04/19 [13:00:15] INFO | >> train_qwenlatent.py:487 + Step 4680 | grad_norm_pre_clip=0.3847 | + grad_norm_pre_clip_avg=0.3892 | Metrics: + {'align_loss': 0.023195426911115646, + 'recon_loss': 0.008034679107367992, + 'predict_loss': 0.0259995199739933, + 'aux_loss_decay_weight': 0.06420000000000003, + 'grad_norm_pre_clip': 0.3846607804298401, + 'data_time': 0.0007069480197969824, + 'model_time': 1.2175807669991627, + 'grad_norm_pre_clip_avg': 0.38921349346637724, + 'learning_rate': 2.3400000000000003e-05, + 'epoch': 1.18} +04/19 [13:00:27] INFO | >> train_qwenlatent.py:487 + Step 4690 | grad_norm_pre_clip=0.3088 | + grad_norm_pre_clip_avg=0.3587 | Metrics: + {'align_loss': 0.021655762568116188, + 'recon_loss': 0.010316367261111736, + 'predict_loss': 0.03066539391875267, + 'aux_loss_decay_weight': 0.06220000000000003, + 'grad_norm_pre_clip': 0.30875158309936523, + 'data_time': 0.0007029589905869216, + 'model_time': 1.2132905469916295, + 'grad_norm_pre_clip_avg': 0.35866072177886965, + 'learning_rate': 2.345e-05, 'epoch': 1.18} +04/19 [13:00:41] INFO | >> train_qwenlatent.py:487 + Step 4700 | grad_norm_pre_clip=0.3242 | + grad_norm_pre_clip_avg=0.3506 | Metrics: + {'align_loss': 0.02156931534409523, + 'recon_loss': 0.007158871740102768, + 'predict_loss': 0.019249355420470238, + 'aux_loss_decay_weight': 0.06020000000000003, + 'grad_norm_pre_clip': 0.3242350220680237, + 'mae_score': 0.0393942961821685, 'data_time': + 0.0014313290012069046, 'model_time': + 1.3161488559853751, 'grad_norm_pre_clip_avg': + 0.35057898461818693, 'learning_rate': 2.35e-05, + 'epoch': 1.19} +04/19 [13:00:54] INFO | >> train_qwenlatent.py:487 + Step 4710 | grad_norm_pre_clip=0.3729 | + grad_norm_pre_clip_avg=0.3439 | Metrics: + {'align_loss': 0.02208172343671322, + 'recon_loss': 0.016221052035689354, + 'predict_loss': 0.034102827310562134, + 'aux_loss_decay_weight': 0.05820000000000003, + 'grad_norm_pre_clip': 0.37293604016304016, + 'data_time': 0.0007607469742652029, + 'model_time': 1.2444235449947882, + 'grad_norm_pre_clip_avg': 0.34387414157390594, + 'learning_rate': 2.355e-05, 'epoch': 1.19} +04/19 [13:01:06] INFO | >> train_qwenlatent.py:487 + Step 4720 | grad_norm_pre_clip=0.3000 | + grad_norm_pre_clip_avg=0.3067 | Metrics: + {'align_loss': 0.020799025893211365, + 'recon_loss': 0.010261674411594868, + 'predict_loss': 0.0232976246625185, + 'aux_loss_decay_weight': 0.05620000000000003, + 'grad_norm_pre_clip': 0.2999570965766907, + 'data_time': 0.000639135978417471, + 'model_time': 1.2608528260025196, + 'grad_norm_pre_clip_avg': 0.30674480497837064, + 'learning_rate': 2.36e-05, 'epoch': 1.19} +04/19 [13:01:19] INFO | >> train_qwenlatent.py:487 + Step 4730 | grad_norm_pre_clip=0.2988 | + grad_norm_pre_clip_avg=0.3129 | Metrics: + {'align_loss': 0.02226501703262329, + 'recon_loss': 0.005992367397993803, + 'predict_loss': 0.023549167439341545, + 'aux_loss_decay_weight': 0.054200000000000026, + 'grad_norm_pre_clip': 0.29875221848487854, + 'data_time': 0.0006979780155234039, + 'model_time': 1.2347199639771134, + 'grad_norm_pre_clip_avg': 0.31291126608848574, + 'learning_rate': 2.365e-05, 'epoch': 1.19} +04/19 [13:01:31] INFO | >> train_qwenlatent.py:487 + Step 4740 | grad_norm_pre_clip=0.2567 | + grad_norm_pre_clip_avg=0.3419 | Metrics: + {'align_loss': 0.019671935588121414, + 'recon_loss': 0.004183025099337101, + 'predict_loss': 0.018780071288347244, + 'aux_loss_decay_weight': 0.052200000000000024, + 'grad_norm_pre_clip': 0.2566683888435364, + 'data_time': 0.0008955519879236817, + 'model_time': 1.2383244849916082, + 'grad_norm_pre_clip_avg': 0.34193865656852723, + 'learning_rate': 2.37e-05, 'epoch': 1.2} +04/19 [13:01:45] INFO | >> train_qwenlatent.py:487 + Step 4750 | grad_norm_pre_clip=0.3191 | + grad_norm_pre_clip_avg=0.3731 | Metrics: + {'align_loss': 0.022112932056188583, + 'recon_loss': 0.007273094262927771, + 'predict_loss': 0.025406673550605774, + 'aux_loss_decay_weight': 0.05020000000000002, + 'grad_norm_pre_clip': 0.3190762996673584, + 'mae_score': 0.03919744749326964, 'data_time': + 0.0008667059883009642, 'model_time': + 1.2214228549855761, 'grad_norm_pre_clip_avg': + 0.3730634778738022, 'learning_rate': 2.375e-05, + 'epoch': 1.2} +04/19 [13:01:58] INFO | >> train_qwenlatent.py:487 + Step 4760 | grad_norm_pre_clip=0.3764 | + grad_norm_pre_clip_avg=0.3981 | Metrics: + {'align_loss': 0.022062145173549652, + 'recon_loss': 0.005341285839676857, + 'predict_loss': 0.021861601620912552, + 'aux_loss_decay_weight': 0.04820000000000002, + 'grad_norm_pre_clip': 0.37638258934020996, + 'data_time': 0.0011777679901570082, + 'model_time': 1.2347249599988572, + 'grad_norm_pre_clip_avg': 0.3981273382902145, + 'learning_rate': 2.38e-05, 'epoch': 1.2} +04/19 [13:02:11] INFO | >> train_qwenlatent.py:487 + Step 4770 | grad_norm_pre_clip=0.4309 | + grad_norm_pre_clip_avg=0.3850 | Metrics: + {'align_loss': 0.022157197818160057, + 'recon_loss': 0.00903276726603508, + 'predict_loss': 0.02429818920791149, + 'aux_loss_decay_weight': 0.04620000000000002, + 'grad_norm_pre_clip': 0.4308711588382721, + 'data_time': 0.000887984992004931, + 'model_time': 1.2490982319868635, + 'grad_norm_pre_clip_avg': 0.38500849306583407, + 'learning_rate': 2.385e-05, 'epoch': 1.2} +04/19 [13:02:24] INFO | >> train_qwenlatent.py:487 + Step 4780 | grad_norm_pre_clip=0.3263 | + grad_norm_pre_clip_avg=0.3335 | Metrics: + {'align_loss': 0.02232133224606514, + 'recon_loss': 0.009047692641615868, + 'predict_loss': 0.02486974559724331, + 'aux_loss_decay_weight': 0.04420000000000002, + 'grad_norm_pre_clip': 0.326308012008667, + 'data_time': 0.0009738270018715411, + 'model_time': 1.2739962770137936, + 'grad_norm_pre_clip_avg': 0.3335009753704071, + 'learning_rate': 2.39e-05, 'epoch': 1.21} +04/19 [13:02:36] INFO | >> train_qwenlatent.py:487 + Step 4790 | grad_norm_pre_clip=0.3217 | + grad_norm_pre_clip_avg=0.3435 | Metrics: + {'align_loss': 0.02121206931769848, + 'recon_loss': 0.005328781437128782, + 'predict_loss': 0.01784898340702057, + 'aux_loss_decay_weight': 0.042200000000000015, + 'grad_norm_pre_clip': 0.3217100501060486, + 'data_time': 0.0009386729798279703, + 'model_time': 1.283486059983261, + 'grad_norm_pre_clip_avg': 0.34349175691604616, + 'learning_rate': 2.395e-05, 'epoch': 1.21} +04/19 [13:02:50] INFO | >> train_qwenlatent.py:487 + Step 4800 | grad_norm_pre_clip=0.3394 | + grad_norm_pre_clip_avg=0.3393 | Metrics: + {'align_loss': 0.021442940458655357, + 'recon_loss': 0.006740986835211515, + 'predict_loss': 0.023233389481902122, + 'aux_loss_decay_weight': 0.040200000000000014, + 'grad_norm_pre_clip': 0.3394298553466797, + 'mae_score': 0.040925852457682294, 'data_time': + 0.0010088790149893612, 'model_time': + 1.2710827499977313, 'grad_norm_pre_clip_avg': + 0.33933748602867125, 'learning_rate': 2.4e-05, + 'epoch': 1.21} +04/19 [13:03:03] INFO | >> train_qwenlatent.py:487 + Step 4810 | grad_norm_pre_clip=0.3319 | + grad_norm_pre_clip_avg=0.2947 | Metrics: + {'align_loss': 0.022861827164888382, + 'recon_loss': 0.01136676874011755, + 'predict_loss': 0.02318769320845604, + 'aux_loss_decay_weight': 0.03820000000000001, + 'grad_norm_pre_clip': 0.33186012506484985, + 'data_time': 0.0013536549813579768, + 'model_time': 1.5102873340074439, + 'grad_norm_pre_clip_avg': 0.2947208106517792, + 'learning_rate': 2.4050000000000002e-05, + 'epoch': 1.21} +04/19 [13:03:15] INFO | >> train_qwenlatent.py:487 + Step 4820 | grad_norm_pre_clip=0.3442 | + grad_norm_pre_clip_avg=0.3514 | Metrics: + {'align_loss': 0.022609807550907135, + 'recon_loss': 0.006717864889651537, + 'predict_loss': 0.02832736447453499, + 'aux_loss_decay_weight': 0.03620000000000001, + 'grad_norm_pre_clip': 0.344153493642807, + 'data_time': 0.0006352110067382455, + 'model_time': 1.2307952740229666, + 'grad_norm_pre_clip_avg': 0.35144865810871123, + 'learning_rate': 2.41e-05, 'epoch': 1.22} +04/19 [13:03:28] INFO | >> train_qwenlatent.py:487 + Step 4830 | grad_norm_pre_clip=0.2761 | + grad_norm_pre_clip_avg=0.3083 | Metrics: + {'align_loss': 0.021071605384349823, + 'recon_loss': 0.012499713338911533, + 'predict_loss': 0.026204949244856834, + 'aux_loss_decay_weight': 0.03420000000000001, + 'grad_norm_pre_clip': 0.2760571539402008, + 'data_time': 0.0008516549714840949, + 'model_time': 1.2606035390053876, + 'grad_norm_pre_clip_avg': 0.30833553820848464, + 'learning_rate': 2.415e-05, 'epoch': 1.22} +04/19 [13:03:41] INFO | >> train_qwenlatent.py:487 + Step 4840 | grad_norm_pre_clip=0.3062 | + grad_norm_pre_clip_avg=0.3347 | Metrics: + {'align_loss': 0.022098522633314133, + 'recon_loss': 0.009213493205606937, + 'predict_loss': 0.02495412528514862, + 'aux_loss_decay_weight': 0.032200000000000006, + 'grad_norm_pre_clip': 0.30617111921310425, + 'data_time': 0.0008565839962102473, + 'model_time': 1.2471306080115028, + 'grad_norm_pre_clip_avg': 0.3347311496734619, + 'learning_rate': 2.4200000000000002e-05, + 'epoch': 1.22} +04/19 [13:03:54] INFO | >> train_qwenlatent.py:487 + Step 4850 | grad_norm_pre_clip=0.4280 | + grad_norm_pre_clip_avg=0.4110 | Metrics: + {'align_loss': 0.02185101807117462, + 'recon_loss': 0.009546724148094654, + 'predict_loss': 0.022951900959014893, + 'aux_loss_decay_weight': 0.030200000000000005, + 'grad_norm_pre_clip': 0.4280214309692383, + 'mae_score': 0.031405141332128024, 'data_time': + 0.0008289719989988953, 'model_time': + 1.2847038119798526, 'grad_norm_pre_clip_avg': + 0.4109799265861511, 'learning_rate': 2.425e-05, + 'epoch': 1.22} +04/19 [13:04:06] INFO | >> train_qwenlatent.py:487 + Step 4860 | grad_norm_pre_clip=0.3523 | + grad_norm_pre_clip_avg=0.3554 | Metrics: + {'align_loss': 0.022743597626686096, + 'recon_loss': 0.014443750493228436, + 'predict_loss': 0.030208801850676537, + 'aux_loss_decay_weight': 0.028200000000000003, + 'grad_norm_pre_clip': 0.3522803783416748, + 'data_time': 0.000872629985678941, + 'model_time': 1.2308741390006617, + 'grad_norm_pre_clip_avg': 0.3554321974515915, + 'learning_rate': 2.43e-05, 'epoch': 1.23} +04/19 [13:04:19] INFO | >> train_qwenlatent.py:487 + Step 4870 | grad_norm_pre_clip=0.3173 | + grad_norm_pre_clip_avg=0.3232 | Metrics: + {'align_loss': 0.022593442350625992, + 'recon_loss': 0.011707874946296215, + 'predict_loss': 0.025878271088004112, + 'aux_loss_decay_weight': 0.0262, + 'grad_norm_pre_clip': 0.3172684907913208, + 'data_time': 0.001242184021975845, + 'model_time': 1.2051904600230046, + 'grad_norm_pre_clip_avg': 0.32315658032894135, + 'learning_rate': 2.435e-05, 'epoch': 1.23} +04/19 [13:04:31] INFO | >> train_qwenlatent.py:487 + Step 4880 | grad_norm_pre_clip=0.2687 | + grad_norm_pre_clip_avg=0.3436 | Metrics: + {'align_loss': 0.02369726449251175, + 'recon_loss': 0.006442629266530275, + 'predict_loss': 0.029098030179739, + 'aux_loss_decay_weight': 0.0242, + 'grad_norm_pre_clip': 0.2687428295612335, + 'data_time': 0.001297246024478227, + 'model_time': 1.2730450250091963, + 'grad_norm_pre_clip_avg': 0.3436182767152786, + 'learning_rate': 2.44e-05, 'epoch': 1.23} +04/19 [13:04:44] INFO | >> train_qwenlatent.py:487 + Step 4890 | grad_norm_pre_clip=0.2717 | + grad_norm_pre_clip_avg=0.3448 | Metrics: + {'align_loss': 0.022678866982460022, + 'recon_loss': 0.0059122019447386265, + 'predict_loss': 0.020618855953216553, + 'aux_loss_decay_weight': 0.022199999999999998, + 'grad_norm_pre_clip': 0.27167487144470215, + 'data_time': 0.0010751219815574586, + 'model_time': 1.1987558229884598, + 'grad_norm_pre_clip_avg': 0.3447893440723419, + 'learning_rate': 2.445e-05, 'epoch': 1.23} +04/19 [13:04:58] INFO | >> train_qwenlatent.py:487 + Step 4900 | grad_norm_pre_clip=0.3272 | + grad_norm_pre_clip_avg=0.3343 | Metrics: + {'align_loss': 0.021003711968660355, + 'recon_loss': 0.009922684170305729, + 'predict_loss': 0.02624659426510334, + 'aux_loss_decay_weight': 0.020199999999999996, + 'grad_norm_pre_clip': 0.32719147205352783, + 'mae_score': 0.030336074141768723, 'data_time': + 0.0007134999905247241, 'model_time': + 1.1957055179809686, 'grad_norm_pre_clip_avg': + 0.3343279451131821, 'learning_rate': 2.45e-05, + 'epoch': 1.24} +04/19 [13:05:11] INFO | >> train_qwenlatent.py:487 + Step 4910 | grad_norm_pre_clip=0.3734 | + grad_norm_pre_clip_avg=0.3233 | Metrics: + {'align_loss': 0.021962691098451614, + 'recon_loss': 0.009449819102883339, + 'predict_loss': 0.025378845632076263, + 'aux_loss_decay_weight': 0.018199999999999994, + 'grad_norm_pre_clip': 0.37340840697288513, + 'data_time': 0.0006986059888731688, + 'model_time': 1.2563612110097893, + 'grad_norm_pre_clip_avg': 0.32331146448850634, + 'learning_rate': 2.455e-05, 'epoch': 1.24} +04/19 [13:05:23] INFO | >> train_qwenlatent.py:487 + Step 4920 | grad_norm_pre_clip=0.3336 | + grad_norm_pre_clip_avg=0.3407 | Metrics: + {'align_loss': 0.02117263339459896, + 'recon_loss': 0.007246161811053753, + 'predict_loss': 0.020593060180544853, + 'aux_loss_decay_weight': 0.016199999999999992, + 'grad_norm_pre_clip': 0.3335990905761719, + 'data_time': 0.0009359949908684939, + 'model_time': 1.203387171990471, + 'grad_norm_pre_clip_avg': 0.34066331684589385, + 'learning_rate': 2.46e-05, 'epoch': 1.24} +04/19 [13:05:36] INFO | >> train_qwenlatent.py:487 + Step 4930 | grad_norm_pre_clip=0.2559 | + grad_norm_pre_clip_avg=0.3128 | Metrics: + {'align_loss': 0.022365842014551163, + 'recon_loss': 0.00486957048997283, + 'predict_loss': 0.022176597267389297, + 'aux_loss_decay_weight': 0.01419999999999999, + 'grad_norm_pre_clip': 0.25587984919548035, + 'data_time': 0.0009190739947371185, + 'model_time': 1.2430608480062801, + 'grad_norm_pre_clip_avg': 0.31276934742927553, + 'learning_rate': 2.465e-05, 'epoch': 1.24} +04/19 [13:05:49] INFO | >> train_qwenlatent.py:487 + Step 4940 | grad_norm_pre_clip=0.3193 | + grad_norm_pre_clip_avg=0.3120 | Metrics: + {'align_loss': 0.022294003516435623, + 'recon_loss': 0.005561196245253086, + 'predict_loss': 0.0218521561473608, + 'aux_loss_decay_weight': 0.012199999999999989, + 'grad_norm_pre_clip': 0.3192717432975769, + 'data_time': 0.0008612840028945357, + 'model_time': 1.2876638599846046, + 'grad_norm_pre_clip_avg': 0.3119577348232269, + 'learning_rate': 2.47e-05, 'epoch': 1.25} +04/19 [13:06:02] INFO | >> train_qwenlatent.py:487 + Step 4950 | grad_norm_pre_clip=0.3134 | + grad_norm_pre_clip_avg=0.4134 | Metrics: + {'align_loss': 0.021928153932094574, + 'recon_loss': 0.0020467378199100494, + 'predict_loss': 0.014834718778729439, + 'aux_loss_decay_weight': 0.010199999999999987, + 'grad_norm_pre_clip': 0.31342148780822754, + 'mae_score': 0.03537855406065245, 'data_time': + 0.0006606410024687648, 'model_time': + 1.2742458000138868, 'grad_norm_pre_clip_avg': + 0.41342714726924895, 'learning_rate': + 2.4750000000000002e-05, 'epoch': 1.25} +04/19 [13:06:15] INFO | >> train_qwenlatent.py:487 + Step 4960 | grad_norm_pre_clip=0.3265 | + grad_norm_pre_clip_avg=0.3589 | Metrics: + {'align_loss': 0.022297490388154984, + 'recon_loss': 0.0030346582643687725, + 'predict_loss': 0.013504442758858204, + 'aux_loss_decay_weight': 0.008199999999999985, + 'grad_norm_pre_clip': 0.3265456259250641, + 'data_time': 0.0007066569814924151, + 'model_time': 1.209334735001903, + 'grad_norm_pre_clip_avg': 0.3589068681001663, + 'learning_rate': 2.48e-05, 'epoch': 1.25} +04/19 [13:06:27] INFO | >> train_qwenlatent.py:487 + Step 4970 | grad_norm_pre_clip=0.2978 | + grad_norm_pre_clip_avg=0.3346 | Metrics: + {'align_loss': 0.020958224311470985, + 'recon_loss': 0.005726981442421675, + 'predict_loss': 0.017556611448526382, + 'aux_loss_decay_weight': 0.006199999999999983, + 'grad_norm_pre_clip': 0.2978159487247467, + 'data_time': 0.0011487409938126802, + 'model_time': 1.217048872000305, + 'grad_norm_pre_clip_avg': 0.3346181094646454, + 'learning_rate': 2.485e-05, 'epoch': 1.25} +04/19 [13:06:40] INFO | >> train_qwenlatent.py:487 + Step 4980 | grad_norm_pre_clip=0.3106 | + grad_norm_pre_clip_avg=0.3030 | Metrics: + {'align_loss': 0.022004367783665657, + 'recon_loss': 0.007791433483362198, + 'predict_loss': 0.024427853524684906, + 'aux_loss_decay_weight': 0.0041999999999999815, + 'grad_norm_pre_clip': 0.3106273412704468, + 'data_time': 0.0007476090104319155, + 'model_time': 1.1998061909980606, + 'grad_norm_pre_clip_avg': 0.3030151665210724, + 'learning_rate': 2.4900000000000002e-05, + 'epoch': 1.26} +04/19 [13:06:53] INFO | >> train_qwenlatent.py:487 + Step 4990 | grad_norm_pre_clip=0.3123 | + grad_norm_pre_clip_avg=0.3101 | Metrics: + {'align_loss': 0.021975502371788025, + 'recon_loss': 0.011860637925565243, + 'predict_loss': 0.031339675188064575, + 'aux_loss_decay_weight': 0.0021999999999999797, + 'grad_norm_pre_clip': 0.3123213052749634, + 'data_time': 0.0012445710017345846, + 'model_time': 1.2797889490029775, + 'grad_norm_pre_clip_avg': 0.3100874096155167, + 'learning_rate': 2.495e-05, 'epoch': 1.26} +04/19 [13:07:06] INFO | >> train_qwenlatent.py:487 + Step 5000 | grad_norm_pre_clip=0.3028 | + grad_norm_pre_clip_avg=0.3244 | Metrics: + {'align_loss': 0.023979129269719124, + 'recon_loss': 0.007641423027962446, + 'predict_loss': 0.021746601909399033, + 'aux_loss_decay_weight': + 0.00019999999999997797, 'grad_norm_pre_clip': + 0.30279678106307983, 'mae_score': + 0.03251011135341885, 'data_time': + 0.0009002560109365731, 'model_time': + 1.2407694360008463, 'grad_norm_pre_clip_avg': + 0.32442558109760283, 'learning_rate': 2.5e-05, + 'epoch': 1.26} +✅ Checkpoint saved at ./runs/0418_QwenLatent_13tasks_actionstate_30k/checkpoints/steps_5000 +04/19 [13:07:29] INFO | >> train_qwenlatent.py:487 + Step 5010 | grad_norm_pre_clip=0.2677 | + grad_norm_pre_clip_avg=0.3603 | Metrics: + {'align_loss': 0.021787431091070175, + 'recon_loss': 0.0037536818999797106, + 'predict_loss': 0.016980160027742386, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26771628856658936, + 'data_time': 0.0010438169992994517, + 'model_time': 1.3799069649830926, + 'grad_norm_pre_clip_avg': 0.36026766896247864, + 'learning_rate': 2.49999969568721e-05, 'epoch': + 1.26} +04/19 [13:07:41] INFO | >> train_qwenlatent.py:487 + Step 5020 | grad_norm_pre_clip=0.2570 | + grad_norm_pre_clip_avg=0.3308 | Metrics: + {'align_loss': 0.023228511214256287, + 'recon_loss': 0.005778395105153322, + 'predict_loss': 0.025648823007941246, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2569839656352997, + 'data_time': 0.0008839179936330765, + 'model_time': 1.2625548829964828, + 'grad_norm_pre_clip_avg': 0.33077988028526306, + 'learning_rate': 2.4999987827489884e-05, + 'epoch': 1.27} +04/19 [13:07:55] INFO | >> train_qwenlatent.py:487 + Step 5030 | grad_norm_pre_clip=0.2911 | + grad_norm_pre_clip_avg=0.3336 | Metrics: + {'align_loss': 0.021709952503442764, + 'recon_loss': 0.008296643383800983, + 'predict_loss': 0.02544609270989895, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2910955846309662, + 'data_time': 0.0008997510012704879, + 'model_time': 1.2960192449972965, + 'grad_norm_pre_clip_avg': 0.33361850678920746, + 'learning_rate': 2.49999726118578e-05, 'epoch': + 1.27} +04/19 [13:08:08] INFO | >> train_qwenlatent.py:487 + Step 5040 | grad_norm_pre_clip=0.3185 | + grad_norm_pre_clip_avg=0.3275 | Metrics: + {'align_loss': 0.02186332270503044, + 'recon_loss': 0.0171870905905962, + 'predict_loss': 0.028666825965046883, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.318511962890625, + 'data_time': 0.0007799899904057384, + 'model_time': 1.2819019129965454, + 'grad_norm_pre_clip_avg': 0.3274658560752869, + 'learning_rate': 2.4999951309983264e-05, + 'epoch': 1.27} +04/19 [13:08:21] INFO | >> train_qwenlatent.py:487 + Step 5050 | grad_norm_pre_clip=0.2894 | + grad_norm_pre_clip_avg=0.2932 | Metrics: + {'align_loss': 0.022696154192090034, + 'recon_loss': 0.008918406441807747, + 'predict_loss': 0.018683137372136116, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2893598675727844, + 'mae_score': 0.02658802926003396, 'data_time': + 0.00082536501577124, 'model_time': + 1.269090474001132, 'grad_norm_pre_clip_avg': + 0.29316355288028717, 'learning_rate': + 2.4999923921876657e-05, 'epoch': 1.27} +04/19 [13:08:34] INFO | >> train_qwenlatent.py:487 + Step 5060 | grad_norm_pre_clip=0.3556 | + grad_norm_pre_clip_avg=0.2835 | Metrics: + {'align_loss': 0.021706201136112213, + 'recon_loss': 0.012326626107096672, + 'predict_loss': 0.022452719509601593, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3555643856525421, + 'data_time': 0.0006813110085204244, + 'model_time': 1.242691993014887, + 'grad_norm_pre_clip_avg': 0.2835151955485344, + 'learning_rate': 2.499989044755133e-05, + 'epoch': 1.28} +04/19 [13:08:47] INFO | >> train_qwenlatent.py:487 + Step 5070 | grad_norm_pre_clip=0.4109 | + grad_norm_pre_clip_avg=0.5217 | Metrics: + {'align_loss': 0.022883731871843338, + 'recon_loss': 0.011523472145199776, + 'predict_loss': 0.02625194564461708, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.4109457731246948, + 'data_time': 0.0008919550164137036, + 'model_time': 1.2616406549932435, + 'grad_norm_pre_clip_avg': 0.5217095941305161, + 'learning_rate': 2.4999850887023603e-05, + 'epoch': 1.28} +04/19 [13:08:59] INFO | >> train_qwenlatent.py:487 + Step 5080 | grad_norm_pre_clip=0.4006 | + grad_norm_pre_clip_avg=0.3852 | Metrics: + {'align_loss': 0.021435372531414032, + 'recon_loss': 0.004164401907473803, + 'predict_loss': 0.02154141291975975, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.4006269872188568, + 'data_time': 0.0008929889881983399, + 'model_time': 1.2496613030089065, + 'grad_norm_pre_clip_avg': 0.38516116738319395, + 'learning_rate': 2.4999805240312745e-05, + 'epoch': 1.28} +04/19 [13:09:12] INFO | >> train_qwenlatent.py:487 + Step 5090 | grad_norm_pre_clip=0.2996 | + grad_norm_pre_clip_avg=0.3256 | Metrics: + {'align_loss': 0.019794877618551254, + 'recon_loss': 0.011883536353707314, + 'predict_loss': 0.025764157995581627, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.29957544803619385, + 'data_time': 0.001414195983670652, + 'model_time': 1.2873101250152104, + 'grad_norm_pre_clip_avg': 0.32564030289649964, + 'learning_rate': 2.4999753507441012e-05, + 'epoch': 1.28} +04/19 [13:09:25] INFO | >> train_qwenlatent.py:487 + Step 5100 | grad_norm_pre_clip=0.3079 | + grad_norm_pre_clip_avg=0.3150 | Metrics: + {'align_loss': 0.022531114518642426, + 'recon_loss': 0.0077346134930849075, + 'predict_loss': 0.01975734531879425, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3079150319099426, + 'mae_score': 0.026244228809803455, 'data_time': + 0.0008955619996413589, 'model_time': + 1.2039788359834347, 'grad_norm_pre_clip_avg': + 0.31501121520996095, 'learning_rate': + 2.4999695688433617e-05, 'epoch': 1.29} +04/19 [13:09:38] INFO | >> train_qwenlatent.py:487 + Step 5110 | grad_norm_pre_clip=0.2873 | + grad_norm_pre_clip_avg=0.2782 | Metrics: + {'align_loss': 0.022243324667215347, + 'recon_loss': 0.011255328543484211, + 'predict_loss': 0.028658317402005196, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2873251736164093, + 'data_time': 0.0008341229986399412, + 'model_time': 1.2882145999756176, + 'grad_norm_pre_clip_avg': 0.2782253369688988, + 'learning_rate': 2.4999631783318745e-05, + 'epoch': 1.29} +04/19 [13:09:51] INFO | >> train_qwenlatent.py:487 + Step 5120 | grad_norm_pre_clip=0.3382 | + grad_norm_pre_clip_avg=0.3366 | Metrics: + {'align_loss': 0.02328520268201828, + 'recon_loss': 0.008357412181794643, + 'predict_loss': 0.02557656541466713, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.338241845369339, + 'data_time': 0.000926289998460561, + 'model_time': 1.2797294420015533, + 'grad_norm_pre_clip_avg': 0.33658460080623626, + 'learning_rate': 2.499956179212753e-05, + 'epoch': 1.29} +04/19 [13:10:03] INFO | >> train_qwenlatent.py:487 + Step 5130 | grad_norm_pre_clip=0.2865 | + grad_norm_pre_clip_avg=0.2921 | Metrics: + {'align_loss': 0.02288377657532692, + 'recon_loss': 0.012593748979270458, + 'predict_loss': 0.025440726429224014, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2865162491798401, + 'data_time': 0.0009763840062078089, + 'model_time': 1.2758888420066796, + 'grad_norm_pre_clip_avg': 0.2921412795782089, + 'learning_rate': 2.4999485714894092e-05, + 'epoch': 1.29} +04/19 [13:10:16] INFO | >> train_qwenlatent.py:487 + Step 5140 | grad_norm_pre_clip=0.3439 | + grad_norm_pre_clip_avg=0.3836 | Metrics: + {'align_loss': 0.021906960755586624, + 'recon_loss': 0.013110646978020668, + 'predict_loss': 0.02767222747206688, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3438502848148346, + 'data_time': 0.0009134610008914024, + 'model_time': 1.220286675990792, + 'grad_norm_pre_clip_avg': 0.3835758000612259, + 'learning_rate': 2.4999403551655514e-05, + 'epoch': 1.3} +04/19 [13:10:29] INFO | >> train_qwenlatent.py:487 + Step 5150 | grad_norm_pre_clip=0.3974 | + grad_norm_pre_clip_avg=0.3930 | Metrics: + {'align_loss': 0.021381892263889313, + 'recon_loss': 0.01147491205483675, + 'predict_loss': 0.02160756103694439, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3973824977874756, + 'mae_score': 0.03865163992117117, 'data_time': + 0.0006903480098117143, 'model_time': + 1.227801114990143, 'grad_norm_pre_clip_avg': + 0.3930138200521469, 'learning_rate': + 2.4999315302451834e-05, 'epoch': 1.3} +04/19 [13:10:42] INFO | >> train_qwenlatent.py:487 + Step 5160 | grad_norm_pre_clip=0.2674 | + grad_norm_pre_clip_avg=0.2781 | Metrics: + {'align_loss': 0.023460086435079575, + 'recon_loss': 0.00809281226247549, + 'predict_loss': 0.019713884219527245, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26743364334106445, + 'data_time': 0.000922664999961853, + 'model_time': 1.2596698080014903, + 'grad_norm_pre_clip_avg': 0.27813423722982406, + 'learning_rate': 2.499922096732607e-05, + 'epoch': 1.3} +04/19 [13:10:55] INFO | >> train_qwenlatent.py:487 + Step 5170 | grad_norm_pre_clip=0.2805 | + grad_norm_pre_clip_avg=0.2705 | Metrics: + {'align_loss': 0.022426800802350044, + 'recon_loss': 0.00732684088870883, + 'predict_loss': 0.022254476323723793, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2805202901363373, + 'data_time': 0.0005972640065010637, + 'model_time': 1.2234062220086344, + 'grad_norm_pre_clip_avg': 0.27048875838518144, + 'learning_rate': 2.4999120546324193e-05, + 'epoch': 1.3} +04/19 [13:11:08] INFO | >> train_qwenlatent.py:487 + Step 5180 | grad_norm_pre_clip=0.3274 | + grad_norm_pre_clip_avg=0.3054 | Metrics: + {'align_loss': 0.021729938685894012, + 'recon_loss': 0.004729065578430891, + 'predict_loss': 0.020485341548919678, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.32744863629341125, + 'data_time': 0.0008960579871200025, + 'model_time': 1.2092590459797066, + 'grad_norm_pre_clip_avg': 0.3053981885313988, + 'learning_rate': 2.4999014039495155e-05, + 'epoch': 1.31} +04/19 [13:11:20] INFO | >> train_qwenlatent.py:487 + Step 5190 | grad_norm_pre_clip=0.4636 | + grad_norm_pre_clip_avg=0.3240 | Metrics: + {'align_loss': 0.02235203981399536, + 'recon_loss': 0.005177707877010107, + 'predict_loss': 0.017914673313498497, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.4636254906654358, + 'data_time': 0.0008863580005709082, + 'model_time': 1.2360418600146659, + 'grad_norm_pre_clip_avg': 0.32399892807006836, + 'learning_rate': 2.499890144689086e-05, + 'epoch': 1.31} +04/19 [13:11:33] INFO | >> train_qwenlatent.py:487 + Step 5200 | grad_norm_pre_clip=0.2978 | + grad_norm_pre_clip_avg=0.3332 | Metrics: + {'align_loss': 0.023969819769263268, + 'recon_loss': 0.013943761587142944, + 'predict_loss': 0.026525920256972313, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2977558970451355, + 'mae_score': 0.052106908849767736, 'data_time': + 0.0006529079983010888, 'model_time': + 1.250530732999323, 'grad_norm_pre_clip_avg': + 0.3332314148545265, 'learning_rate': + 2.4998782768566186e-05, 'epoch': 1.31} +04/19 [13:11:46] INFO | >> train_qwenlatent.py:487 + Step 5210 | grad_norm_pre_clip=0.3203 | + grad_norm_pre_clip_avg=0.3370 | Metrics: + {'align_loss': 0.02280524931848049, + 'recon_loss': 0.01438268180936575, + 'predict_loss': 0.030759278684854507, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.32028380036354065, + 'data_time': 0.0006704319966956973, + 'model_time': 1.184169368003495, + 'grad_norm_pre_clip_avg': 0.3370027631521225, + 'learning_rate': 2.499865800457898e-05, + 'epoch': 1.31} +04/19 [13:11:59] INFO | >> train_qwenlatent.py:487 + Step 5220 | grad_norm_pre_clip=0.2769 | + grad_norm_pre_clip_avg=0.3186 | Metrics: + {'align_loss': 0.023031286895275116, + 'recon_loss': 0.011090632528066635, + 'predict_loss': 0.026235127821564674, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2769283950328827, + 'data_time': 0.0009190939890686423, + 'model_time': 1.3372418289945927, + 'grad_norm_pre_clip_avg': 0.3185607612133026, + 'learning_rate': 2.4998527154990043e-05, + 'epoch': 1.32} +04/19 [13:12:12] INFO | >> train_qwenlatent.py:487 + Step 5230 | grad_norm_pre_clip=0.4480 | + grad_norm_pre_clip_avg=0.3181 | Metrics: + {'align_loss': 0.02295038476586342, + 'recon_loss': 0.006805359851568937, + 'predict_loss': 0.01702856458723545, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.44801992177963257, + 'data_time': 0.0009629330015741289, + 'model_time': 1.2486837330216076, + 'grad_norm_pre_clip_avg': 0.3181164100766182, + 'learning_rate': 2.4998390219863153e-05, + 'epoch': 1.32} +04/19 [13:12:25] INFO | >> train_qwenlatent.py:487 + Step 5240 | grad_norm_pre_clip=0.3486 | + grad_norm_pre_clip_avg=0.3721 | Metrics: + {'align_loss': 0.021227650344371796, + 'recon_loss': 0.011217184364795685, + 'predict_loss': 0.03432605415582657, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3486059308052063, + 'data_time': 0.0009413200023118407, + 'model_time': 1.2902617719955742, + 'grad_norm_pre_clip_avg': 0.37214735746383665, + 'learning_rate': 2.4998247199265046e-05, + 'epoch': 1.32} +04/19 [13:12:38] INFO | >> train_qwenlatent.py:487 + Step 5250 | grad_norm_pre_clip=0.3913 | + grad_norm_pre_clip_avg=0.4100 | Metrics: + {'align_loss': 0.023412737995386124, + 'recon_loss': 0.00924659799784422, + 'predict_loss': 0.023892071098089218, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3912964463233948, + 'mae_score': 0.03783835333746833, 'data_time': + 0.0006629550189245492, 'model_time': + 1.2024769810086582, 'grad_norm_pre_clip_avg': + 0.41001624763011935, 'learning_rate': + 2.4998098093265437e-05, 'epoch': 1.32} +04/19 [13:12:50] INFO | >> train_qwenlatent.py:487 + Step 5260 | grad_norm_pre_clip=0.2435 | + grad_norm_pre_clip_avg=0.3602 | Metrics: + {'align_loss': 0.02249598503112793, + 'recon_loss': 0.004455734975636005, + 'predict_loss': 0.01824199967086315, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24348054826259613, + 'data_time': 0.0008430310117546469, + 'model_time': 1.206855155993253, + 'grad_norm_pre_clip_avg': 0.3602185159921646, + 'learning_rate': 2.499794290193699e-05, + 'epoch': 1.33} +04/19 [13:13:03] INFO | >> train_qwenlatent.py:487 + Step 5270 | grad_norm_pre_clip=0.3032 | + grad_norm_pre_clip_avg=0.4029 | Metrics: + {'align_loss': 0.022764163091778755, + 'recon_loss': 0.009168028831481934, + 'predict_loss': 0.020630232989788055, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.30318424105644226, + 'data_time': 0.0012227479892317206, + 'model_time': 1.2286733540240675, + 'grad_norm_pre_clip_avg': 0.4029011338949203, + 'learning_rate': 2.4997781625355353e-05, + 'epoch': 1.33} +04/19 [13:13:15] INFO | >> train_qwenlatent.py:487 + Step 5280 | grad_norm_pre_clip=0.2630 | + grad_norm_pre_clip_avg=0.3064 | Metrics: + {'align_loss': 0.02306031435728073, + 'recon_loss': 0.010148080065846443, + 'predict_loss': 0.019000399857759476, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2629663050174713, + 'data_time': 0.000638869998510927, + 'model_time': 1.2870922410220373, + 'grad_norm_pre_clip_avg': 0.3063733667135239, + 'learning_rate': 2.4997614263599124e-05, + 'epoch': 1.33} +04/19 [13:13:28] INFO | >> train_qwenlatent.py:487 + Step 5290 | grad_norm_pre_clip=0.1904 | + grad_norm_pre_clip_avg=0.3024 | Metrics: + {'align_loss': 0.023105721920728683, + 'recon_loss': 0.00845414213836193, + 'predict_loss': 0.02022557519376278, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19044490158557892, + 'data_time': 0.0008285870135296136, + 'model_time': 1.6170646009850316, + 'grad_norm_pre_clip_avg': 0.30241870433092116, + 'learning_rate': 2.499744081674987e-05, + 'epoch': 1.33} +04/19 [13:13:42] INFO | >> train_qwenlatent.py:487 + Step 5300 | grad_norm_pre_clip=0.2353 | + grad_norm_pre_clip_avg=0.2474 | Metrics: + {'align_loss': 0.022511165589094162, + 'recon_loss': 0.005500664934515953, + 'predict_loss': 0.01699851080775261, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23533450067043304, + 'mae_score': 0.05024115416380736, 'data_time': + 0.0008563699957448989, 'model_time': + 1.5210688170045614, 'grad_norm_pre_clip_avg': + 0.24742455929517745, 'learning_rate': + 2.4997261284892134e-05, 'epoch': 1.34} +04/19 [13:13:55] INFO | >> train_qwenlatent.py:487 + Step 5310 | grad_norm_pre_clip=0.3820 | + grad_norm_pre_clip_avg=0.3309 | Metrics: + {'align_loss': 0.02175033837556839, + 'recon_loss': 0.010266931727528572, + 'predict_loss': 0.02169741876423359, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.38201215863227844, + 'data_time': 0.0006431170040741563, + 'model_time': 1.252316010009963, + 'grad_norm_pre_clip_avg': 0.33086884319782256, + 'learning_rate': 2.4997075668113412e-05, + 'epoch': 1.34} +04/19 [13:14:08] INFO | >> train_qwenlatent.py:487 + Step 5320 | grad_norm_pre_clip=0.3156 | + grad_norm_pre_clip_avg=0.3047 | Metrics: + {'align_loss': 0.02422035112977028, + 'recon_loss': 0.010469015687704086, + 'predict_loss': 0.020213868468999863, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.31560370326042175, + 'data_time': 0.0007087749836500734, + 'model_time': 1.236733703990467, + 'grad_norm_pre_clip_avg': 0.3047227129340172, + 'learning_rate': 2.4996883966504174e-05, + 'epoch': 1.34} +04/19 [13:14:20] INFO | >> train_qwenlatent.py:487 + Step 5330 | grad_norm_pre_clip=0.2929 | + grad_norm_pre_clip_avg=0.3211 | Metrics: + {'align_loss': 0.02313121035695076, + 'recon_loss': 0.00941424909979105, + 'predict_loss': 0.024245567619800568, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.29289376735687256, + 'data_time': 0.0014440109953284264, + 'model_time': 1.244193009013543, + 'grad_norm_pre_clip_avg': 0.32109405994415285, + 'learning_rate': 2.4996686180157856e-05, + 'epoch': 1.34} +04/19 [13:14:33] INFO | >> train_qwenlatent.py:487 + Step 5340 | grad_norm_pre_clip=0.2652 | + grad_norm_pre_clip_avg=0.3257 | Metrics: + {'align_loss': 0.02219662256538868, + 'recon_loss': 0.010306156240403652, + 'predict_loss': 0.02308434061706066, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2651697099208832, + 'data_time': 0.0008724610088393092, + 'model_time': 1.2640697119932156, + 'grad_norm_pre_clip_avg': 0.32573711276054385, + 'learning_rate': 2.4996482309170852e-05, + 'epoch': 1.35} +04/19 [13:14:46] INFO | >> train_qwenlatent.py:487 + Step 5350 | grad_norm_pre_clip=0.2669 | + grad_norm_pre_clip_avg=0.3194 | Metrics: + {'align_loss': 0.023543700575828552, + 'recon_loss': 0.012754088267683983, + 'predict_loss': 0.02114042267203331, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26685187220573425, + 'mae_score': 0.029776286219691372, 'data_time': + 0.0006898909923620522, 'model_time': + 1.2670639639836736, 'grad_norm_pre_clip_avg': + 0.3194265395402908, 'learning_rate': + 2.4996272353642523e-05, 'epoch': 1.35} +04/19 [13:14:59] INFO | >> train_qwenlatent.py:487 + Step 5360 | grad_norm_pre_clip=0.2710 | + grad_norm_pre_clip_avg=0.2995 | Metrics: + {'align_loss': 0.022022198885679245, + 'recon_loss': 0.012273848988115788, + 'predict_loss': 0.026775041595101357, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.27095070481300354, + 'data_time': 0.0007902559882495552, + 'model_time': 1.213783170998795, + 'grad_norm_pre_clip_avg': 0.2994613930583, + 'learning_rate': 2.499605631367521e-05, + 'epoch': 1.35} +04/19 [13:15:12] INFO | >> train_qwenlatent.py:487 + Step 5370 | grad_norm_pre_clip=0.3863 | + grad_norm_pre_clip_avg=0.3192 | Metrics: + {'align_loss': 0.0214930959045887, + 'recon_loss': 0.007558557204902172, + 'predict_loss': 0.021076278761029243, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.38633450865745544, + 'data_time': 0.0009607090032659471, + 'model_time': 1.2106661979923956, + 'grad_norm_pre_clip_avg': 0.3192208155989647, + 'learning_rate': 2.4995834189374197e-05, + 'epoch': 1.36} +04/19 [13:15:24] INFO | >> train_qwenlatent.py:487 + Step 5380 | grad_norm_pre_clip=0.2944 | + grad_norm_pre_clip_avg=0.3396 | Metrics: + {'align_loss': 0.02202453836798668, + 'recon_loss': 0.012812445871531963, + 'predict_loss': 0.02285844460129738, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.29440954327583313, + 'data_time': 0.000890449999133125, + 'model_time': 1.2813943520013709, + 'grad_norm_pre_clip_avg': 0.33959086835384367, + 'learning_rate': 2.4995605980847753e-05, + 'epoch': 1.36} +04/19 [13:15:37] INFO | >> train_qwenlatent.py:487 + Step 5390 | grad_norm_pre_clip=0.2906 | + grad_norm_pre_clip_avg=0.2934 | Metrics: + {'align_loss': 0.02369103953242302, + 'recon_loss': 0.006381327752023935, + 'predict_loss': 0.019283277913928032, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2906342148780823, + 'data_time': 0.0009096760186366737, + 'model_time': 1.242519393999828, + 'grad_norm_pre_clip_avg': 0.29337345361709594, + 'learning_rate': 2.49953716882071e-05, 'epoch': + 1.36} +04/19 [13:15:51] INFO | >> train_qwenlatent.py:487 + Step 5400 | grad_norm_pre_clip=0.3619 | + grad_norm_pre_clip_avg=0.3047 | Metrics: + {'align_loss': 0.021358273923397064, + 'recon_loss': 0.011080845259130001, + 'predict_loss': 0.023795412853360176, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.36185938119888306, + 'mae_score': 0.03434422810872396, 'data_time': + 0.0007291009824257344, 'model_time': + 1.2203950149996672, 'grad_norm_pre_clip_avg': + 0.30471531450748446, 'learning_rate': + 2.4995131311566425e-05, 'epoch': 1.36} +04/19 [13:16:03] INFO | >> train_qwenlatent.py:487 + Step 5410 | grad_norm_pre_clip=0.3639 | + grad_norm_pre_clip_avg=0.3308 | Metrics: + {'align_loss': 0.023056354373693466, + 'recon_loss': 0.005943478085100651, + 'predict_loss': 0.01744363084435463, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3638966381549835, + 'data_time': 0.0007486779941245914, + 'model_time': 1.2171654730045702, + 'grad_norm_pre_clip_avg': 0.330790750682354, + 'learning_rate': 2.4994884851042892e-05, + 'epoch': 1.37} +04/19 [13:16:16] INFO | >> train_qwenlatent.py:487 + Step 5420 | grad_norm_pre_clip=0.4697 | + grad_norm_pre_clip_avg=0.3408 | Metrics: + {'align_loss': 0.02276000753045082, + 'recon_loss': 0.007765375077724457, + 'predict_loss': 0.01977100409567356, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.46965891122817993, + 'data_time': 0.0006969140085857362, + 'model_time': 1.2449014990124851, + 'grad_norm_pre_clip_avg': 0.34081425368785856, + 'learning_rate': 2.499463230675662e-05, + 'epoch': 1.37} +04/19 [13:16:29] INFO | >> train_qwenlatent.py:487 + Step 5430 | grad_norm_pre_clip=0.2755 | + grad_norm_pre_clip_avg=0.3300 | Metrics: + {'align_loss': 0.022714462131261826, + 'recon_loss': 0.009666119702160358, + 'predict_loss': 0.02031938172876835, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2755469083786011, + 'data_time': 0.0009474150137975812, + 'model_time': 1.4798837100097444, + 'grad_norm_pre_clip_avg': 0.33000450730323794, + 'learning_rate': 2.49943736788307e-05, 'epoch': + 1.37} +04/19 [13:16:42] INFO | >> train_qwenlatent.py:487 + Step 5440 | grad_norm_pre_clip=0.2876 | + grad_norm_pre_clip_avg=0.2961 | Metrics: + {'align_loss': 0.021733876317739487, + 'recon_loss': 0.007656773552298546, + 'predict_loss': 0.02670324593782425, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28756463527679443, + 'data_time': 0.001283055986277759, + 'model_time': 1.30082028999459, + 'grad_norm_pre_clip_avg': 0.29613576233386996, + 'learning_rate': 2.4994108967391176e-05, + 'epoch': 1.37} +04/19 [13:16:55] INFO | >> train_qwenlatent.py:487 + Step 5450 | grad_norm_pre_clip=0.3464 | + grad_norm_pre_clip_avg=0.3356 | Metrics: + {'align_loss': 0.02391851507127285, + 'recon_loss': 0.01568221114575863, + 'predict_loss': 0.03249618411064148, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3463974595069885, + 'mae_score': 0.03741186674650725, 'data_time': + 0.0006398569967132062, 'model_time': + 1.2110698120086454, 'grad_norm_pre_clip_avg': + 0.3355963945388794, 'learning_rate': + 2.4993838172567074e-05, 'epoch': 1.38} +04/19 [13:17:08] INFO | >> train_qwenlatent.py:487 + Step 5460 | grad_norm_pre_clip=0.3114 | + grad_norm_pre_clip_avg=0.3043 | Metrics: + {'align_loss': 0.02352822944521904, + 'recon_loss': 0.012648778967559338, + 'predict_loss': 0.028348051011562347, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3114379644393921, + 'data_time': 0.0007821350009180605, + 'model_time': 1.1598324820224661, + 'grad_norm_pre_clip_avg': 0.304305836558342, + 'learning_rate': 2.4993561294490368e-05, + 'epoch': 1.38} +04/19 [13:17:21] INFO | >> train_qwenlatent.py:487 + Step 5470 | grad_norm_pre_clip=0.3864 | + grad_norm_pre_clip_avg=0.3615 | Metrics: + {'align_loss': 0.0229203961789608, + 'recon_loss': 0.00956934317946434, + 'predict_loss': 0.021113945171236992, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.38643455505371094, + 'data_time': 0.0009928180079441518, + 'model_time': 1.232191656017676, + 'grad_norm_pre_clip_avg': 0.3614534497261047, + 'learning_rate': 2.499327833329601e-05, + 'epoch': 1.38} +04/19 [13:17:33] INFO | >> train_qwenlatent.py:487 + Step 5480 | grad_norm_pre_clip=0.2589 | + grad_norm_pre_clip_avg=0.2893 | Metrics: + {'align_loss': 0.022905133664608, 'recon_loss': + 0.018175723031163216, 'predict_loss': + 0.03359915316104889, 'aux_loss_decay_weight': + 0.0, 'grad_norm_pre_clip': 0.25887468457221985, + 'data_time': 0.0010738330020103604, + 'model_time': 1.2033747390087228, + 'grad_norm_pre_clip_avg': 0.28926644623279574, + 'learning_rate': 2.4992989289121912e-05, + 'epoch': 1.38} +04/19 [13:17:46] INFO | >> train_qwenlatent.py:487 + Step 5490 | grad_norm_pre_clip=0.2594 | + grad_norm_pre_clip_avg=0.2608 | Metrics: + {'align_loss': 0.02262938767671585, + 'recon_loss': 0.00688674533739686, + 'predict_loss': 0.01543241273611784, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25939080119132996, + 'data_time': 0.0008541690185666084, + 'model_time': 1.2121522809902672, + 'grad_norm_pre_clip_avg': 0.2607511028647423, + 'learning_rate': 2.499269416210895e-05, + 'epoch': 1.39} +04/19 [13:17:59] INFO | >> train_qwenlatent.py:487 + Step 5500 | grad_norm_pre_clip=0.2936 | + grad_norm_pre_clip_avg=0.3330 | Metrics: + {'align_loss': 0.022647801786661148, + 'recon_loss': 0.011682664975523949, + 'predict_loss': 0.022006496787071228, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2935504615306854, + 'mae_score': 0.029806491061373872, 'data_time': + 0.0007037150207906961, 'model_time': + 1.2092211119888816, 'grad_norm_pre_clip_avg': + 0.3330076977610588, 'learning_rate': + 2.4992392952400962e-05, 'epoch': 1.39} +04/19 [13:18:12] INFO | >> train_qwenlatent.py:487 + Step 5510 | grad_norm_pre_clip=0.2848 | + grad_norm_pre_clip_avg=0.3013 | Metrics: + {'align_loss': 0.02106393687427044, + 'recon_loss': 0.011851206421852112, + 'predict_loss': 0.022209659218788147, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28476226329803467, + 'data_time': 0.0006534619897138327, + 'model_time': 1.227709929982666, + 'grad_norm_pre_clip_avg': 0.30132181793451307, + 'learning_rate': 2.4992085660144753e-05, + 'epoch': 1.39} +04/19 [13:18:24] INFO | >> train_qwenlatent.py:487 + Step 5520 | grad_norm_pre_clip=0.2987 | + grad_norm_pre_clip_avg=0.2690 | Metrics: + {'align_loss': 0.02359454333782196, + 'recon_loss': 0.007554339710623026, + 'predict_loss': 0.015609401278197765, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.29867854714393616, + 'data_time': 0.001017497997963801, + 'model_time': 1.240381465991959, + 'grad_norm_pre_clip_avg': 0.2690166100859642, + 'learning_rate': 2.49917722854901e-05, 'epoch': + 1.39} +04/19 [13:18:37] INFO | >> train_qwenlatent.py:487 + Step 5530 | grad_norm_pre_clip=0.3753 | + grad_norm_pre_clip_avg=0.3091 | Metrics: + {'align_loss': 0.022883571684360504, + 'recon_loss': 0.010696016252040863, + 'predict_loss': 0.023528490215539932, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.37534812092781067, + 'data_time': 0.0009222340013366193, + 'model_time': 1.1950768379902001, + 'grad_norm_pre_clip_avg': 0.30905229449272154, + 'learning_rate': 2.4991452828589738e-05, + 'epoch': 1.4} +04/19 [13:18:49] INFO | >> train_qwenlatent.py:487 + Step 5540 | grad_norm_pre_clip=0.4120 | + grad_norm_pre_clip_avg=0.3198 | Metrics: + {'align_loss': 0.022533047944307327, + 'recon_loss': 0.007558710407465696, + 'predict_loss': 0.019998187199234962, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.4120318293571472, + 'data_time': 0.0008675080025568604, + 'model_time': 1.234783953987062, + 'grad_norm_pre_clip_avg': 0.3197704553604126, + 'learning_rate': 2.4991127289599358e-05, + 'epoch': 1.4} +04/19 [13:19:03] INFO | >> train_qwenlatent.py:487 + Step 5550 | grad_norm_pre_clip=0.3347 | + grad_norm_pre_clip_avg=0.3510 | Metrics: + {'align_loss': 0.022087471559643745, + 'recon_loss': 0.01083305012434721, + 'predict_loss': 0.021147148683667183, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3346700370311737, + 'mae_score': 0.041860491091066654, 'data_time': + 0.0006420960125979036, 'model_time': + 1.5528293599782046, 'grad_norm_pre_clip_avg': + 0.3509707272052765, 'learning_rate': + 2.499079566867763e-05, 'epoch': 1.4} +04/19 [13:19:15] INFO | >> train_qwenlatent.py:487 + Step 5560 | grad_norm_pre_clip=0.2757 | + grad_norm_pre_clip_avg=0.3189 | Metrics: + {'align_loss': 0.022700291126966476, + 'recon_loss': 0.008154459297657013, + 'predict_loss': 0.0191305223852396, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2757185995578766, + 'data_time': 0.000948785018408671, + 'model_time': 1.286912908020895, + 'grad_norm_pre_clip_avg': 0.3188826352357864, + 'learning_rate': 2.4990457965986185e-05, + 'epoch': 1.4} +04/19 [13:19:28] INFO | >> train_qwenlatent.py:487 + Step 5570 | grad_norm_pre_clip=0.3017 | + grad_norm_pre_clip_avg=0.2956 | Metrics: + {'align_loss': 0.023723607882857323, + 'recon_loss': 0.014660301618278027, + 'predict_loss': 0.024463379755616188, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.30170130729675293, + 'data_time': 0.0006173169822432101, + 'model_time': 1.2709528090199456, + 'grad_norm_pre_clip_avg': 0.2956169813871384, + 'learning_rate': 2.499011418168961e-05, + 'epoch': 1.41} +04/19 [13:19:41] INFO | >> train_qwenlatent.py:487 + Step 5580 | grad_norm_pre_clip=0.2798 | + grad_norm_pre_clip_avg=0.2946 | Metrics: + {'align_loss': 0.022167783230543137, + 'recon_loss': 0.011482018046081066, + 'predict_loss': 0.02466999553143978, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2798343300819397, + 'data_time': 0.0008650970121379942, + 'model_time': 1.177517178002745, + 'grad_norm_pre_clip_avg': 0.2946222797036171, + 'learning_rate': 2.4989764315955464e-05, + 'epoch': 1.41} +04/19 [13:19:53] INFO | >> train_qwenlatent.py:487 + Step 5590 | grad_norm_pre_clip=0.2867 | + grad_norm_pre_clip_avg=0.3611 | Metrics: + {'align_loss': 0.02332780510187149, + 'recon_loss': 0.011973724700510502, + 'predict_loss': 0.025481151416897774, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2867136299610138, + 'data_time': 0.0007364299963228405, + 'model_time': 1.2968741249933373, + 'grad_norm_pre_clip_avg': 0.3610538333654404, + 'learning_rate': 2.4989408368954264e-05, + 'epoch': 1.41} +04/19 [13:20:07] INFO | >> train_qwenlatent.py:487 + Step 5600 | grad_norm_pre_clip=0.3045 | + grad_norm_pre_clip_avg=0.3410 | Metrics: + {'align_loss': 0.023606814444065094, + 'recon_loss': 0.013807034119963646, + 'predict_loss': 0.028170166537165642, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3044642210006714, + 'mae_score': 0.042666182646880275, 'data_time': + 0.0009400949929840863, 'model_time': + 1.2463663490198087, 'grad_norm_pre_clip_avg': + 0.34102095663547516, 'learning_rate': + 2.4989046340859495e-05, 'epoch': 1.41} +04/19 [13:20:20] INFO | >> train_qwenlatent.py:487 + Step 5610 | grad_norm_pre_clip=0.2640 | + grad_norm_pre_clip_avg=0.3141 | Metrics: + {'align_loss': 0.022253576666116714, + 'recon_loss': 0.008685892447829247, + 'predict_loss': 0.019157540053129196, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26402536034584045, + 'data_time': 0.0006853140075691044, + 'model_time': 1.2456203769834246, + 'grad_norm_pre_clip_avg': 0.314103464782238, + 'learning_rate': 2.498867823184761e-05, + 'epoch': 1.42} +04/19 [13:20:32] INFO | >> train_qwenlatent.py:487 + Step 5620 | grad_norm_pre_clip=0.2661 | + grad_norm_pre_clip_avg=0.2985 | Metrics: + {'align_loss': 0.022692173719406128, + 'recon_loss': 0.012358350679278374, + 'predict_loss': 0.02162545546889305, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2661157548427582, + 'data_time': 0.0006521060131490231, + 'model_time': 1.2356438050046563, + 'grad_norm_pre_clip_avg': 0.2985426187515259, + 'learning_rate': 2.4988304042098012e-05, + 'epoch': 1.42} +04/19 [13:20:45] INFO | >> train_qwenlatent.py:487 + Step 5630 | grad_norm_pre_clip=0.3467 | + grad_norm_pre_clip_avg=0.2991 | Metrics: + {'align_loss': 0.024009689688682556, + 'recon_loss': 0.011510159820318222, + 'predict_loss': 0.02227032743394375, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.34670376777648926, + 'data_time': 0.0009144439827650785, + 'model_time': 1.2052294739987701, + 'grad_norm_pre_clip_avg': 0.2991089314222336, + 'learning_rate': 2.4987923771793085e-05, + 'epoch': 1.42} +04/19 [13:20:57] INFO | >> train_qwenlatent.py:487 + Step 5640 | grad_norm_pre_clip=0.2541 | + grad_norm_pre_clip_avg=0.2774 | Metrics: + {'align_loss': 0.023235052824020386, + 'recon_loss': 0.00963566079735756, + 'predict_loss': 0.022403087466955185, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2540624737739563, + 'data_time': 0.0008387980051338673, + 'model_time': 1.2401612700195983, + 'grad_norm_pre_clip_avg': 0.2774134650826454, + 'learning_rate': 2.4987537421118166e-05, + 'epoch': 1.42} +04/19 [13:21:10] INFO | >> train_qwenlatent.py:487 + Step 5650 | grad_norm_pre_clip=0.2810 | + grad_norm_pre_clip_avg=0.3191 | Metrics: + {'align_loss': 0.02386493608355522, + 'recon_loss': 0.011292272247374058, + 'predict_loss': 0.018593518063426018, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28095510601997375, + 'mae_score': 0.033264995265651395, 'data_time': + 0.0006716149800922722, 'model_time': + 1.2301684789999854, 'grad_norm_pre_clip_avg': + 0.31912164837121965, 'learning_rate': + 2.498714499026155e-05, 'epoch': 1.43} +04/19 [13:21:22] INFO | >> train_qwenlatent.py:487 + Step 5660 | grad_norm_pre_clip=0.2641 | + grad_norm_pre_clip_avg=0.3020 | Metrics: + {'align_loss': 0.02226945385336876, + 'recon_loss': 0.009246688336133957, + 'predict_loss': 0.01975984498858452, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2640773355960846, + 'data_time': 0.0007569119916297495, + 'model_time': 1.218058486992959, + 'grad_norm_pre_clip_avg': 0.30201277136802673, + 'learning_rate': 2.4986746479414517e-05, + 'epoch': 1.43} +04/19 [13:21:35] INFO | >> train_qwenlatent.py:487 + Step 5670 | grad_norm_pre_clip=0.2880 | + grad_norm_pre_clip_avg=0.2845 | Metrics: + {'align_loss': 0.022978095337748528, + 'recon_loss': 0.009118984453380108, + 'predict_loss': 0.01926281861960888, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28804975748062134, + 'data_time': 0.000659236015053466, + 'model_time': 1.5007429709949065, + 'grad_norm_pre_clip_avg': 0.28448724150657656, + 'learning_rate': 2.4986341888771282e-05, + 'epoch': 1.43} +04/19 [13:21:48] INFO | >> train_qwenlatent.py:487 + Step 5680 | grad_norm_pre_clip=0.2913 | + grad_norm_pre_clip_avg=0.3510 | Metrics: + {'align_loss': 0.02298000082373619, + 'recon_loss': 0.007656375877559185, + 'predict_loss': 0.016617145389318466, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2912512421607971, + 'data_time': 0.0008245590142905712, + 'model_time': 1.2353847289923579, + 'grad_norm_pre_clip_avg': 0.35101584196090696, + 'learning_rate': 2.4985931218529048e-05, + 'epoch': 1.43} +04/19 [13:22:00] INFO | >> train_qwenlatent.py:487 + Step 5690 | grad_norm_pre_clip=0.2317 | + grad_norm_pre_clip_avg=0.2934 | Metrics: + {'align_loss': 0.022516775876283646, + 'recon_loss': 0.010371840558946133, + 'predict_loss': 0.018114622682332993, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23166285455226898, + 'data_time': 0.0007986709824763238, + 'model_time': 1.2401722749928012, + 'grad_norm_pre_clip_avg': 0.2934035584330559, + 'learning_rate': 2.4985514468887964e-05, + 'epoch': 1.44} +04/19 [13:22:13] INFO | >> train_qwenlatent.py:487 + Step 5700 | grad_norm_pre_clip=0.2577 | + grad_norm_pre_clip_avg=0.2667 | Metrics: + {'align_loss': 0.02286890335381031, + 'recon_loss': 0.012248074635863304, + 'predict_loss': 0.01936887763440609, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2577410042285919, + 'mae_score': 0.028665439502612966, 'data_time': + 0.0006289870070759207, 'model_time': + 1.2402209080173634, 'grad_norm_pre_clip_avg': + 0.26669625490903853, 'learning_rate': + 2.4985091640051157e-05, 'epoch': 1.44} +04/19 [13:22:26] INFO | >> train_qwenlatent.py:487 + Step 5710 | grad_norm_pre_clip=0.2620 | + grad_norm_pre_clip_avg=0.2705 | Metrics: + {'align_loss': 0.02339533343911171, + 'recon_loss': 0.008227510377764702, + 'predict_loss': 0.01692342199385166, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.262047678232193, + 'data_time': 0.0008537929970771074, + 'model_time': 1.304112259997055, + 'grad_norm_pre_clip_avg': 0.270450234413147, + 'learning_rate': 2.4984662732224703e-05, + 'epoch': 1.44} +04/19 [13:22:39] INFO | >> train_qwenlatent.py:487 + Step 5720 | grad_norm_pre_clip=0.3006 | + grad_norm_pre_clip_avg=0.4189 | Metrics: + {'align_loss': 0.022147759795188904, + 'recon_loss': 0.009371078573167324, + 'predict_loss': 0.020962156355381012, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3006332218647003, + 'data_time': 0.001030995015753433, + 'model_time': 1.2336158950056415, + 'grad_norm_pre_clip_avg': 0.4189078837633133, + 'learning_rate': 2.4984227745617646e-05, + 'epoch': 1.44} +04/19 [13:22:51] INFO | >> train_qwenlatent.py:487 + Step 5730 | grad_norm_pre_clip=0.2916 | + grad_norm_pre_clip_avg=0.3213 | Metrics: + {'align_loss': 0.022493809461593628, + 'recon_loss': 0.01059875451028347, + 'predict_loss': 0.018392350524663925, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.29162004590034485, + 'data_time': 0.0008583840099163353, + 'model_time': 1.255918750015553, + 'grad_norm_pre_clip_avg': 0.32130164802074435, + 'learning_rate': 2.498378668044199e-05, + 'epoch': 1.45} +04/19 [13:23:05] INFO | >> train_qwenlatent.py:487 + Step 5740 | grad_norm_pre_clip=0.3020 | + grad_norm_pre_clip_avg=0.2985 | Metrics: + {'align_loss': 0.02259419485926628, + 'recon_loss': 0.011608198285102844, + 'predict_loss': 0.01904243417084217, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.30198827385902405, + 'data_time': 0.0007762479945085943, + 'model_time': 1.217023926001275, + 'grad_norm_pre_clip_avg': 0.29852868616580963, + 'learning_rate': 2.4983339536912713e-05, + 'epoch': 1.45} +04/19 [13:23:18] INFO | >> train_qwenlatent.py:487 + Step 5750 | grad_norm_pre_clip=0.2787 | + grad_norm_pre_clip_avg=0.2955 | Metrics: + {'align_loss': 0.021509094163775444, + 'recon_loss': 0.013876977376639843, + 'predict_loss': 0.022696105763316154, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2787376642227173, + 'mae_score': 0.024594959911999403, 'data_time': + 0.0006947770016267896, 'model_time': + 1.2195314920099918, 'grad_norm_pre_clip_avg': + 0.29551569521427157, 'learning_rate': + 2.4982886315247742e-05, 'epoch': 1.45} +04/19 [13:23:30] INFO | >> train_qwenlatent.py:487 + Step 5760 | grad_norm_pre_clip=0.2924 | + grad_norm_pre_clip_avg=0.2827 | Metrics: + {'align_loss': 0.023130036890506744, + 'recon_loss': 0.011600280180573463, + 'predict_loss': 0.021310923621058464, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.29239219427108765, + 'data_time': 0.0008373579767066985, + 'model_time': 1.2210230980126653, + 'grad_norm_pre_clip_avg': 0.28273141831159593, + 'learning_rate': 2.4982427015667975e-05, + 'epoch': 1.45} +04/19 [13:23:43] INFO | >> train_qwenlatent.py:487 + Step 5770 | grad_norm_pre_clip=0.3189 | + grad_norm_pre_clip_avg=0.2691 | Metrics: + {'align_loss': 0.022862523794174194, + 'recon_loss': 0.00950634479522705, + 'predict_loss': 0.020845821127295494, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3188719153404236, + 'data_time': 0.0008876310021150857, + 'model_time': 1.2982102060050238, + 'grad_norm_pre_clip_avg': 0.2691321074962616, + 'learning_rate': 2.4981961638397263e-05, + 'epoch': 1.46} +04/19 [13:23:56] INFO | >> train_qwenlatent.py:487 + Step 5780 | grad_norm_pre_clip=0.3685 | + grad_norm_pre_clip_avg=0.3861 | Metrics: + {'align_loss': 0.02168234810233116, + 'recon_loss': 0.015316872857511044, + 'predict_loss': 0.02509400248527527, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3684588670730591, + 'data_time': 0.0009934089903254062, + 'model_time': 1.2966823820024729, + 'grad_norm_pre_clip_avg': 0.38607907891273496, + 'learning_rate': 2.4981490183662436e-05, + 'epoch': 1.46} +04/19 [13:24:08] INFO | >> train_qwenlatent.py:487 + Step 5790 | grad_norm_pre_clip=0.2757 | + grad_norm_pre_clip_avg=0.2814 | Metrics: + {'align_loss': 0.02379428967833519, + 'recon_loss': 0.007731867488473654, + 'predict_loss': 0.016487177461385727, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2756916582584381, + 'data_time': 0.0009105159842874855, + 'model_time': 1.273092352988897, + 'grad_norm_pre_clip_avg': 0.2813632026314735, + 'learning_rate': 2.4981012651693258e-05, + 'epoch': 1.46} +04/19 [13:24:22] INFO | >> train_qwenlatent.py:487 + Step 5800 | grad_norm_pre_clip=0.3111 | + grad_norm_pre_clip_avg=0.3210 | Metrics: + {'align_loss': 0.02200986072421074, + 'recon_loss': 0.010951527394354343, + 'predict_loss': 0.01910209283232689, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3110804259777069, + 'mae_score': 0.024874255893466708, 'data_time': + 0.0011918109958060086, 'model_time': + 1.2670297359873075, 'grad_norm_pre_clip_avg': + 0.3210249602794647, 'learning_rate': + 2.498052904272249e-05, 'epoch': 1.46} +04/19 [13:24:34] INFO | >> train_qwenlatent.py:487 + Step 5810 | grad_norm_pre_clip=0.2985 | + grad_norm_pre_clip_avg=0.2833 | Metrics: + {'align_loss': 0.02222861349582672, + 'recon_loss': 0.010404365137219429, + 'predict_loss': 0.019512994214892387, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2985028028488159, + 'data_time': 0.0007739200082141906, + 'model_time': 1.212040802987758, + 'grad_norm_pre_clip_avg': 0.28331459760665895, + 'learning_rate': 2.498003935698583e-05, + 'epoch': 1.47} +04/19 [13:24:47] INFO | >> train_qwenlatent.py:487 + Step 5820 | grad_norm_pre_clip=0.3373 | + grad_norm_pre_clip_avg=0.3198 | Metrics: + {'align_loss': 0.021849196404218674, + 'recon_loss': 0.009496261365711689, + 'predict_loss': 0.019792532548308372, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.33729827404022217, + 'data_time': 0.0007222449930850416, + 'model_time': 1.2373808900010772, + 'grad_norm_pre_clip_avg': 0.31976004540920255, + 'learning_rate': 2.4979543594721942e-05, + 'epoch': 1.47} +04/19 [13:25:00] INFO | >> train_qwenlatent.py:487 + Step 5830 | grad_norm_pre_clip=0.2900 | + grad_norm_pre_clip_avg=0.3073 | Metrics: + {'align_loss': 0.02312212437391281, + 'recon_loss': 0.009016605094075203, + 'predict_loss': 0.017731133848428726, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2899538576602936, + 'data_time': 0.000691787019604817, + 'model_time': 1.248172057006741, + 'grad_norm_pre_clip_avg': 0.30732116252183916, + 'learning_rate': 2.4979041756172456e-05, + 'epoch': 1.47} +04/19 [13:25:12] INFO | >> train_qwenlatent.py:487 + Step 5840 | grad_norm_pre_clip=0.3053 | + grad_norm_pre_clip_avg=0.3149 | Metrics: + {'align_loss': 0.023905888199806213, + 'recon_loss': 0.007860714569687843, + 'predict_loss': 0.017466049641370773, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3052993416786194, + 'data_time': 0.0009854560194071382, + 'model_time': 1.2784776730113663, + 'grad_norm_pre_clip_avg': 0.31488481909036636, + 'learning_rate': 2.4978533841581967e-05, + 'epoch': 1.47} +04/19 [13:25:26] INFO | >> train_qwenlatent.py:487 + Step 5850 | grad_norm_pre_clip=0.2889 | + grad_norm_pre_clip_avg=0.2954 | Metrics: + {'align_loss': 0.022283334285020828, + 'recon_loss': 0.012857777997851372, + 'predict_loss': 0.023180581629276276, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28887709975242615, + 'mae_score': 0.031210615828230575, 'data_time': + 0.0008259030000772327, 'model_time': + 1.2248196979926433, 'grad_norm_pre_clip_avg': + 0.29539870023727416, 'learning_rate': + 2.4978019851198024e-05, 'epoch': 1.48} +04/19 [13:25:38] INFO | >> train_qwenlatent.py:487 + Step 5860 | grad_norm_pre_clip=0.3023 | + grad_norm_pre_clip_avg=0.2958 | Metrics: + {'align_loss': 0.02318640798330307, + 'recon_loss': 0.013266884721815586, + 'predict_loss': 0.02060195431113243, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.30228137969970703, + 'data_time': 0.0006881739827804267, + 'model_time': 1.263067588006379, + 'grad_norm_pre_clip_avg': 0.2958205759525299, + 'learning_rate': 2.4977499785271133e-05, + 'epoch': 1.48} +04/19 [13:25:51] INFO | >> train_qwenlatent.py:487 + Step 5870 | grad_norm_pre_clip=0.2785 | + grad_norm_pre_clip_avg=0.2889 | Metrics: + {'align_loss': 0.02188311144709587, + 'recon_loss': 0.01475472655147314, + 'predict_loss': 0.02681637369096279, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2784784138202667, + 'data_time': 0.0007130939920898527, + 'model_time': 1.2123501530149952, + 'grad_norm_pre_clip_avg': 0.288880230486393, + 'learning_rate': 2.4976973644054774e-05, + 'epoch': 1.48} +04/19 [13:26:04] INFO | >> train_qwenlatent.py:487 + Step 5880 | grad_norm_pre_clip=0.2363 | + grad_norm_pre_clip_avg=0.2530 | Metrics: + {'align_loss': 0.02363627403974533, + 'recon_loss': 0.011197102256119251, + 'predict_loss': 0.02051674574613571, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23629127442836761, + 'data_time': 0.001171476993476972, + 'model_time': 1.218689370987704, + 'grad_norm_pre_clip_avg': 0.25295881927013397, + 'learning_rate': 2.4976441427805383e-05, + 'epoch': 1.48} +04/19 [13:26:17] INFO | >> train_qwenlatent.py:487 + Step 5890 | grad_norm_pre_clip=0.2811 | + grad_norm_pre_clip_avg=0.3384 | Metrics: + {'align_loss': 0.02432737499475479, + 'recon_loss': 0.019787704572081566, + 'predict_loss': 0.031335264444351196, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2810523808002472, + 'data_time': 0.0010890990088228136, + 'model_time': 1.2135103690088727, + 'grad_norm_pre_clip_avg': 0.33840980380773544, + 'learning_rate': 2.4975903136782354e-05, + 'epoch': 1.49} +04/19 [13:26:30] INFO | >> train_qwenlatent.py:487 + Step 5900 | grad_norm_pre_clip=0.2342 | + grad_norm_pre_clip_avg=0.2690 | Metrics: + {'align_loss': 0.02286553382873535, + 'recon_loss': 0.012664176523685455, + 'predict_loss': 0.01859319768846035, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2342393547296524, + 'mae_score': 0.04334144935951577, 'data_time': + 0.0008253139967564493, 'model_time': + 1.2259286250045989, 'grad_norm_pre_clip_avg': + 0.2690234735608101, 'learning_rate': + 2.4975358771248042e-05, 'epoch': 1.49} +04/19 [13:26:42] INFO | >> train_qwenlatent.py:487 + Step 5910 | grad_norm_pre_clip=0.2474 | + grad_norm_pre_clip_avg=0.3211 | Metrics: + {'align_loss': 0.023542948067188263, + 'recon_loss': 0.011203131638467312, + 'predict_loss': 0.02016376331448555, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24736955761909485, + 'data_time': 0.0011599780118558556, + 'model_time': 1.2383065470203292, + 'grad_norm_pre_clip_avg': 0.321050663292408, + 'learning_rate': 2.497480833146777e-05, + 'epoch': 1.49} +04/19 [13:26:55] INFO | >> train_qwenlatent.py:487 + Step 5920 | grad_norm_pre_clip=0.2310 | + grad_norm_pre_clip_avg=0.2610 | Metrics: + {'align_loss': 0.02253132313489914, + 'recon_loss': 0.016012199223041534, + 'predict_loss': 0.021973224356770515, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23104684054851532, + 'data_time': 0.0008885179995559156, + 'model_time': 1.2473145310068503, + 'grad_norm_pre_clip_avg': 0.26100446879863737, + 'learning_rate': 2.4974251817709803e-05, + 'epoch': 1.49} +04/19 [13:27:08] INFO | >> train_qwenlatent.py:487 + Step 5930 | grad_norm_pre_clip=0.2907 | + grad_norm_pre_clip_avg=0.3130 | Metrics: + {'align_loss': 0.021003002300858498, + 'recon_loss': 0.00954830925911665, + 'predict_loss': 0.01634596846997738, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2907443344593048, + 'data_time': 0.0009633469744585454, + 'model_time': 1.1927216299809515, + 'grad_norm_pre_clip_avg': 0.3130401849746704, + 'learning_rate': 2.4973689230245393e-05, + 'epoch': 1.5} +04/19 [13:27:20] INFO | >> train_qwenlatent.py:487 + Step 5940 | grad_norm_pre_clip=0.3248 | + grad_norm_pre_clip_avg=0.3007 | Metrics: + {'align_loss': 0.021550044417381287, + 'recon_loss': 0.008731605485081673, + 'predict_loss': 0.01763186603784561, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.32476216554641724, + 'data_time': 0.0007117739878594875, + 'model_time': 1.2725435159809422, + 'grad_norm_pre_clip_avg': 0.3007467120885849, + 'learning_rate': 2.497312056934873e-05, + 'epoch': 1.5} +04/19 [13:27:33] INFO | >> train_qwenlatent.py:487 + Step 5950 | grad_norm_pre_clip=0.2356 | + grad_norm_pre_clip_avg=0.2619 | Metrics: + {'align_loss': 0.023530013859272003, + 'recon_loss': 0.016769323498010635, + 'predict_loss': 0.028049126267433167, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23561128973960876, + 'mae_score': 0.029708961967949393, 'data_time': + 0.0006778980023227632, 'model_time': + 1.2304845959879458, 'grad_norm_pre_clip_avg': + 0.26187474727630616, 'learning_rate': + 2.4972545835296975e-05, 'epoch': 1.5} +04/19 [13:27:46] INFO | >> train_qwenlatent.py:487 + Step 5960 | grad_norm_pre_clip=0.2706 | + grad_norm_pre_clip_avg=0.3074 | Metrics: + {'align_loss': 0.02298421412706375, + 'recon_loss': 0.013123485259711742, + 'predict_loss': 0.027442418038845062, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.27058321237564087, + 'data_time': 0.0007141490059439093, + 'model_time': 1.193412218010053, + 'grad_norm_pre_clip_avg': 0.30736499428749087, + 'learning_rate': 2.4971965028370252e-05, + 'epoch': 1.5} +04/19 [13:27:59] INFO | >> train_qwenlatent.py:487 + Step 5970 | grad_norm_pre_clip=0.3630 | + grad_norm_pre_clip_avg=0.3183 | Metrics: + {'align_loss': 0.022485030815005302, + 'recon_loss': 0.012133565731346607, + 'predict_loss': 0.02069896273314953, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.36299756169319153, + 'data_time': 0.0006318929954431951, + 'model_time': 1.2478321669914294, + 'grad_norm_pre_clip_avg': 0.31834978610277176, + 'learning_rate': 2.497137814885162e-05, + 'epoch': 1.51} +04/19 [13:28:11] INFO | >> train_qwenlatent.py:487 + Step 5980 | grad_norm_pre_clip=0.2817 | + grad_norm_pre_clip_avg=0.3157 | Metrics: + {'align_loss': 0.023882664740085602, + 'recon_loss': 0.01340570393949747, + 'predict_loss': 0.023262994363904, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28168627619743347, + 'data_time': 0.0010887680109590292, + 'model_time': 1.26395289201173, + 'grad_norm_pre_clip_avg': 0.31572447121143343, + 'learning_rate': 2.4970785197027145e-05, + 'epoch': 1.51} +04/19 [13:28:24] INFO | >> train_qwenlatent.py:487 + Step 5990 | grad_norm_pre_clip=0.2533 | + grad_norm_pre_clip_avg=0.2643 | Metrics: + {'align_loss': 0.02262602373957634, + 'recon_loss': 0.009385124780237675, + 'predict_loss': 0.017850825563073158, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.253317654132843, + 'data_time': 0.000910034024855122, + 'model_time': 1.3246890049776994, + 'grad_norm_pre_clip_avg': 0.2642771854996681, + 'learning_rate': 2.49701861731858e-05, 'epoch': + 1.51} +04/19 [13:28:38] INFO | >> train_qwenlatent.py:487 + Step 6000 | grad_norm_pre_clip=0.2627 | + grad_norm_pre_clip_avg=0.2751 | Metrics: + {'align_loss': 0.023693036288022995, + 'recon_loss': 0.01164382603019476, + 'predict_loss': 0.019993945956230164, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2627123296260834, + 'mae_score': 0.036788222166869014, 'data_time': + 0.0006761880067642778, 'model_time': + 1.2340045149903744, 'grad_norm_pre_clip_avg': + 0.27506299763917924, 'learning_rate': + 2.4969581077619555e-05, 'epoch': 1.51} +04/19 [13:28:51] INFO | >> train_qwenlatent.py:487 + Step 6010 | grad_norm_pre_clip=0.4078 | + grad_norm_pre_clip_avg=0.2863 | Metrics: + {'align_loss': 0.02351655624806881, + 'recon_loss': 0.01696445420384407, + 'predict_loss': 0.022051803767681122, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.4077722132205963, + 'data_time': 0.0007713730155956, 'model_time': + 1.2587758710142225, 'grad_norm_pre_clip_avg': + 0.28629260063171386, 'learning_rate': + 2.4968969910623325e-05, 'epoch': 1.52} +04/19 [13:29:03] INFO | >> train_qwenlatent.py:487 + Step 6020 | grad_norm_pre_clip=0.3331 | + grad_norm_pre_clip_avg=0.3332 | Metrics: + {'align_loss': 0.023077769204974174, + 'recon_loss': 0.012073385529220104, + 'predict_loss': 0.02137225680053234, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.33314332365989685, + 'data_time': 0.0007000530022196472, + 'model_time': 1.2520233539980836, + 'grad_norm_pre_clip_avg': 0.33316696882247926, + 'learning_rate': 2.4968352672494983e-05, + 'epoch': 1.52} +04/19 [13:29:16] INFO | >> train_qwenlatent.py:487 + Step 6030 | grad_norm_pre_clip=0.3213 | + grad_norm_pre_clip_avg=0.2783 | Metrics: + {'align_loss': 0.023678885772824287, + 'recon_loss': 0.014901530928909779, + 'predict_loss': 0.02290809527039528, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3213360905647278, + 'data_time': 0.0009237169870175421, + 'model_time': 1.2692075519880746, + 'grad_norm_pre_clip_avg': 0.27834112346172335, + 'learning_rate': 2.496772936353536e-05, + 'epoch': 1.52} +04/19 [13:29:29] INFO | >> train_qwenlatent.py:487 + Step 6040 | grad_norm_pre_clip=0.2520 | + grad_norm_pre_clip_avg=0.2698 | Metrics: + {'align_loss': 0.023355063050985336, + 'recon_loss': 0.016686318442225456, + 'predict_loss': 0.024079201743006706, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25197479128837585, + 'data_time': 0.0008925000147428364, + 'model_time': 1.2436813979875296, + 'grad_norm_pre_clip_avg': 0.26982622742652895, + 'learning_rate': 2.496709998404826e-05, + 'epoch': 1.52} +04/19 [13:29:42] INFO | >> train_qwenlatent.py:487 + Step 6050 | grad_norm_pre_clip=0.5641 | + grad_norm_pre_clip_avg=0.3111 | Metrics: + {'align_loss': 0.022437704727053642, + 'recon_loss': 0.012957222759723663, + 'predict_loss': 0.021543173119425774, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.5640590786933899, + 'mae_score': 0.028441355249903225, 'data_time': + 0.000991826003883034, 'model_time': + 1.2986958590045106, 'grad_norm_pre_clip_avg': + 0.31108793020248415, 'learning_rate': + 2.496646453434042e-05, 'epoch': 1.53} +04/19 [13:29:54] INFO | >> train_qwenlatent.py:487 + Step 6060 | grad_norm_pre_clip=0.3071 | + grad_norm_pre_clip_avg=0.3861 | Metrics: + {'align_loss': 0.02359774336218834, + 'recon_loss': 0.01569507271051407, + 'predict_loss': 0.022699380293488503, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3071227967739105, + 'data_time': 0.0010995240008924156, + 'model_time': 1.2329004849889316, + 'grad_norm_pre_clip_avg': 0.3860549867153168, + 'learning_rate': 2.4965823014721565e-05, + 'epoch': 1.53} +04/19 [13:30:07] INFO | >> train_qwenlatent.py:487 + Step 6070 | grad_norm_pre_clip=0.2745 | + grad_norm_pre_clip_avg=0.2693 | Metrics: + {'align_loss': 0.022805146872997284, + 'recon_loss': 0.010594331659376621, + 'predict_loss': 0.017955802381038666, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2744900584220886, + 'data_time': 0.0009419270209036767, + 'model_time': 1.2702484469919, + 'grad_norm_pre_clip_avg': 0.26933168768882754, + 'learning_rate': 2.4965175425504353e-05, + 'epoch': 1.53} +04/19 [13:30:20] INFO | >> train_qwenlatent.py:487 + Step 6080 | grad_norm_pre_clip=0.2002 | + grad_norm_pre_clip_avg=0.2560 | Metrics: + {'align_loss': 0.02274596318602562, + 'recon_loss': 0.011131319217383862, + 'predict_loss': 0.01774505525827408, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20018112659454346, + 'data_time': 0.0006923960172571242, + 'model_time': 1.2566262200125493, + 'grad_norm_pre_clip_avg': 0.2560206338763237, + 'learning_rate': 2.4964521767004416e-05, + 'epoch': 1.53} +04/19 [13:30:32] INFO | >> train_qwenlatent.py:487 + Step 6090 | grad_norm_pre_clip=0.2747 | + grad_norm_pre_clip_avg=0.2698 | Metrics: + {'align_loss': 0.022647462785243988, + 'recon_loss': 0.011483277194201946, + 'predict_loss': 0.01856241002678871, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2747238278388977, + 'data_time': 0.0006837060209363699, + 'model_time': 1.247901261987863, + 'grad_norm_pre_clip_avg': 0.26982377767562865, + 'learning_rate': 2.496386203954034e-05, + 'epoch': 1.54} +04/19 [13:30:46] INFO | >> train_qwenlatent.py:487 + Step 6100 | grad_norm_pre_clip=0.2471 | + grad_norm_pre_clip_avg=0.3324 | Metrics: + {'align_loss': 0.02206268534064293, + 'recon_loss': 0.010447287932038307, + 'predict_loss': 0.018974225968122482, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2470942884683609, + 'mae_score': 0.02896543279424444, 'data_time': + 0.0007266770116984844, 'model_time': + 1.225423119001789, 'grad_norm_pre_clip_avg': + 0.33235255628824234, 'learning_rate': + 2.4963196243433672e-05, 'epoch': 1.54} +04/19 [13:30:58] INFO | >> train_qwenlatent.py:487 + Step 6110 | grad_norm_pre_clip=0.2670 | + grad_norm_pre_clip_avg=0.3059 | Metrics: + {'align_loss': 0.02226181887090206, + 'recon_loss': 0.017312997952103615, + 'predict_loss': 0.02270667999982834, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2669740915298462, + 'data_time': 0.0009215210156980902, + 'model_time': 1.2662767120054923, + 'grad_norm_pre_clip_avg': 0.3059009611606598, + 'learning_rate': 2.4962524379008896e-05, + 'epoch': 1.54} +04/19 [13:31:11] INFO | >> train_qwenlatent.py:487 + Step 6120 | grad_norm_pre_clip=0.3345 | + grad_norm_pre_clip_avg=0.2828 | Metrics: + {'align_loss': 0.0225392896682024, + 'recon_loss': 0.0066932011395692825, + 'predict_loss': 0.013034542091190815, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.33448487520217896, + 'data_time': 0.0006872329977340996, + 'model_time': 1.200478667014977, + 'grad_norm_pre_clip_avg': 0.2828421965241432, + 'learning_rate': 2.496184644659349e-05, + 'epoch': 1.54} +04/19 [13:31:24] INFO | >> train_qwenlatent.py:487 + Step 6130 | grad_norm_pre_clip=0.2712 | + grad_norm_pre_clip_avg=0.2981 | Metrics: + {'align_loss': 0.022472653537988663, + 'recon_loss': 0.013320229947566986, + 'predict_loss': 0.019612515345215797, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.27123120427131653, + 'data_time': 0.0009276510099880397, + 'model_time': 1.2889948739903048, + 'grad_norm_pre_clip_avg': 0.29814435839653014, + 'learning_rate': 2.496116244651786e-05, + 'epoch': 1.55} +04/19 [13:31:36] INFO | >> train_qwenlatent.py:487 + Step 6140 | grad_norm_pre_clip=0.3502 | + grad_norm_pre_clip_avg=0.3055 | Metrics: + {'align_loss': 0.022996515035629272, + 'recon_loss': 0.01900121010839939, + 'predict_loss': 0.02398214489221573, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.35018783807754517, + 'data_time': 0.0006676559860352427, + 'model_time': 1.2510097670019604, + 'grad_norm_pre_clip_avg': 0.30549142360687254, + 'learning_rate': 2.4960472379115382e-05, + 'epoch': 1.55} +04/19 [13:31:50] INFO | >> train_qwenlatent.py:487 + Step 6150 | grad_norm_pre_clip=0.2927 | + grad_norm_pre_clip_avg=0.3047 | Metrics: + {'align_loss': 0.024024445563554764, + 'recon_loss': 0.012956761755049229, + 'predict_loss': 0.023165877908468246, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2927245497703552, + 'mae_score': 0.028217654185252145, 'data_time': + 0.0008367119880858809, 'model_time': + 1.2202464659931138, 'grad_norm_pre_clip_avg': + 0.30466355979442594, 'learning_rate': + 2.495977624472239e-05, 'epoch': 1.55} +04/19 [13:32:02] INFO | >> train_qwenlatent.py:487 + Step 6160 | grad_norm_pre_clip=0.3482 | + grad_norm_pre_clip_avg=0.2806 | Metrics: + {'align_loss': 0.023908820003271103, + 'recon_loss': 0.012276058085262775, + 'predict_loss': 0.023409169167280197, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3482125699520111, + 'data_time': 0.0009299340017605573, + 'model_time': 1.2643133079982363, + 'grad_norm_pre_clip_avg': 0.28055162727832794, + 'learning_rate': 2.4959074043678164e-05, + 'epoch': 1.55} +04/19 [13:32:15] INFO | >> train_qwenlatent.py:487 + Step 6170 | grad_norm_pre_clip=0.3054 | + grad_norm_pre_clip_avg=0.2745 | Metrics: + {'align_loss': 0.02479654923081398, + 'recon_loss': 0.015237169340252876, + 'predict_loss': 0.025065986439585686, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3054390251636505, + 'data_time': 0.0008445149869658053, + 'model_time': 1.4592409679898992, + 'grad_norm_pre_clip_avg': 0.2745230317115784, + 'learning_rate': 2.495836577632495e-05, + 'epoch': 1.56} +04/19 [13:32:28] INFO | >> train_qwenlatent.py:487 + Step 6180 | grad_norm_pre_clip=0.2626 | + grad_norm_pre_clip_avg=0.3199 | Metrics: + {'align_loss': 0.023274697363376617, + 'recon_loss': 0.016992652788758278, + 'predict_loss': 0.02825528010725975, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2625887393951416, + 'data_time': 0.0005989540077280253, + 'model_time': 1.2444614460109733, + 'grad_norm_pre_clip_avg': 0.3198550522327423, + 'learning_rate': 2.495765144300795e-05, + 'epoch': 1.56} +04/19 [13:32:41] INFO | >> train_qwenlatent.py:487 + Step 6190 | grad_norm_pre_clip=0.3169 | + grad_norm_pre_clip_avg=0.2969 | Metrics: + {'align_loss': 0.02307981252670288, + 'recon_loss': 0.013288374990224838, + 'predict_loss': 0.020201029255986214, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3169138431549072, + 'data_time': 0.000930662004975602, + 'model_time': 1.2768240150180645, + 'grad_norm_pre_clip_avg': 0.296908900141716, + 'learning_rate': 2.4956931044075325e-05, + 'epoch': 1.56} +04/19 [13:32:54] INFO | >> train_qwenlatent.py:487 + Step 6200 | grad_norm_pre_clip=0.2622 | + grad_norm_pre_clip_avg=0.3194 | Metrics: + {'align_loss': 0.023422710597515106, + 'recon_loss': 0.012229934334754944, + 'predict_loss': 0.01786867156624794, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2621520757675171, + 'mae_score': 0.03245843423379434, 'data_time': + 0.0008818170172162354, 'model_time': + 1.2940767529944424, 'grad_norm_pre_clip_avg': + 0.3193765178322792, 'learning_rate': + 2.4956204579878187e-05, 'epoch': 1.56} +04/19 [13:33:07] INFO | >> train_qwenlatent.py:487 + Step 6210 | grad_norm_pre_clip=0.2950 | + grad_norm_pre_clip_avg=0.3456 | Metrics: + {'align_loss': 0.021968228742480278, + 'recon_loss': 0.016338005661964417, + 'predict_loss': 0.022205380722880363, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.29495754837989807, + 'data_time': 0.0007124949770513922, + 'model_time': 1.2088819349883124, + 'grad_norm_pre_clip_avg': 0.34560138285160064, + 'learning_rate': 2.49554720507706e-05, 'epoch': + 1.57} +04/19 [13:33:20] INFO | >> train_qwenlatent.py:487 + Step 6220 | grad_norm_pre_clip=0.3640 | + grad_norm_pre_clip_avg=0.2874 | Metrics: + {'align_loss': 0.023405909538269043, + 'recon_loss': 0.013162793591618538, + 'predict_loss': 0.02576976828277111, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3640300929546356, + 'data_time': 0.0006513579864986241, + 'model_time': 1.2619406219746452, + 'grad_norm_pre_clip_avg': 0.2873607352375984, + 'learning_rate': 2.4954733457109604e-05, + 'epoch': 1.57} +04/19 [13:33:32] INFO | >> train_qwenlatent.py:487 + Step 6230 | grad_norm_pre_clip=0.3077 | + grad_norm_pre_clip_avg=0.2959 | Metrics: + {'align_loss': 0.0230708047747612, + 'recon_loss': 0.016170507296919823, + 'predict_loss': 0.026331475004553795, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3077300190925598, + 'data_time': 0.0006789310136809945, + 'model_time': 1.2083255019970238, + 'grad_norm_pre_clip_avg': 0.2958826139569283, + 'learning_rate': 2.495398879925516e-05, + 'epoch': 1.57} +04/19 [13:33:45] INFO | >> train_qwenlatent.py:487 + Step 6240 | grad_norm_pre_clip=0.2689 | + grad_norm_pre_clip_avg=0.2900 | Metrics: + {'align_loss': 0.023218195885419846, + 'recon_loss': 0.013593150302767754, + 'predict_loss': 0.01963457092642784, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26889708638191223, + 'data_time': 0.0009767329902388155, + 'model_time': 1.202536605997011, + 'grad_norm_pre_clip_avg': 0.29004254192113876, + 'learning_rate': 2.495323807757022e-05, + 'epoch': 1.57} +04/19 [13:33:58] INFO | >> train_qwenlatent.py:487 + Step 6250 | grad_norm_pre_clip=0.2978 | + grad_norm_pre_clip_avg=0.2765 | Metrics: + {'align_loss': 0.02277490496635437, + 'recon_loss': 0.012378642335534096, + 'predict_loss': 0.017770495265722275, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.29780465364456177, + 'mae_score': 0.03191461649026957, 'data_time': + 0.0010151560127269477, 'model_time': + 1.213392185018165, 'grad_norm_pre_clip_avg': + 0.2765111282467842, 'learning_rate': + 2.4952481292420673e-05, 'epoch': 1.58} +04/19 [13:34:11] INFO | >> train_qwenlatent.py:487 + Step 6260 | grad_norm_pre_clip=0.3222 | + grad_norm_pre_clip_avg=0.2739 | Metrics: + {'align_loss': 0.022674359381198883, + 'recon_loss': 0.011869915761053562, + 'predict_loss': 0.02027672342956066, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3221529722213745, + 'data_time': 0.001314027002081275, + 'model_time': 1.226802349992795, + 'grad_norm_pre_clip_avg': 0.2739461988210678, + 'learning_rate': 2.4951718444175366e-05, + 'epoch': 1.58} +04/19 [13:34:23] INFO | >> train_qwenlatent.py:487 + Step 6270 | grad_norm_pre_clip=0.2580 | + grad_norm_pre_clip_avg=0.2875 | Metrics: + {'align_loss': 0.023048264905810356, + 'recon_loss': 0.015167872421443462, + 'predict_loss': 0.018170779570937157, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25802111625671387, + 'data_time': 0.0008889480086509138, + 'model_time': 1.3017769859870896, + 'grad_norm_pre_clip_avg': 0.2874667480587959, + 'learning_rate': 2.4950949533206105e-05, + 'epoch': 1.58} +04/19 [13:34:36] INFO | >> train_qwenlatent.py:487 + Step 6280 | grad_norm_pre_clip=0.2806 | + grad_norm_pre_clip_avg=0.2452 | Metrics: + {'align_loss': 0.023086264729499817, + 'recon_loss': 0.013882266357541084, + 'predict_loss': 0.018229946494102478, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28055718541145325, + 'data_time': 0.000686068000504747, + 'model_time': 1.20000413901289, + 'grad_norm_pre_clip_avg': 0.24521952271461486, + 'learning_rate': 2.4950174559887646e-05, + 'epoch': 1.58} +04/19 [13:34:48] INFO | >> train_qwenlatent.py:487 + Step 6290 | grad_norm_pre_clip=0.3756 | + grad_norm_pre_clip_avg=0.3030 | Metrics: + {'align_loss': 0.022451411932706833, + 'recon_loss': 0.01567493937909603, + 'predict_loss': 0.022488661110401154, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3755839169025421, + 'data_time': 0.0009356389928143471, + 'model_time': 1.2127457049791701, + 'grad_norm_pre_clip_avg': 0.3030435457825661, + 'learning_rate': 2.49493935245977e-05, 'epoch': + 1.59} +04/19 [13:35:02] INFO | >> train_qwenlatent.py:487 + Step 6300 | grad_norm_pre_clip=0.2728 | + grad_norm_pre_clip_avg=0.3548 | Metrics: + {'align_loss': 0.023633591830730438, + 'recon_loss': 0.015390139073133469, + 'predict_loss': 0.020168377086520195, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.27275797724723816, + 'mae_score': 0.03268636755041174, 'data_time': + 0.0007908200204838067, 'model_time': + 1.2192563389835414, 'grad_norm_pre_clip_avg': + 0.354772624373436, 'learning_rate': + 2.494860642771693e-05, 'epoch': 1.59} +04/19 [13:35:14] INFO | >> train_qwenlatent.py:487 + Step 6310 | grad_norm_pre_clip=0.2880 | + grad_norm_pre_clip_avg=0.2679 | Metrics: + {'align_loss': 0.024958867579698563, + 'recon_loss': 0.021252373233437538, + 'predict_loss': 0.026348337531089783, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28798139095306396, + 'data_time': 0.001376772008370608, + 'model_time': 1.2631674750009552, + 'grad_norm_pre_clip_avg': 0.26785551607608793, + 'learning_rate': 2.4947813269628965e-05, + 'epoch': 1.59} +04/19 [13:35:27] INFO | >> train_qwenlatent.py:487 + Step 6320 | grad_norm_pre_clip=0.2396 | + grad_norm_pre_clip_avg=0.2777 | Metrics: + {'align_loss': 0.02377650886774063, + 'recon_loss': 0.014856371097266674, + 'predict_loss': 0.020454252138733864, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23962534964084625, + 'data_time': 0.0009288290166296065, + 'model_time': 1.2478161550243385, + 'grad_norm_pre_clip_avg': 0.277657987177372, + 'learning_rate': 2.4947014050720386e-05, + 'epoch': 1.59} +04/19 [13:35:40] INFO | >> train_qwenlatent.py:487 + Step 6330 | grad_norm_pre_clip=0.3653 | + grad_norm_pre_clip_avg=0.2694 | Metrics: + {'align_loss': 0.02411116287112236, + 'recon_loss': 0.018949992954730988, + 'predict_loss': 0.022652506828308105, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.36533159017562866, + 'data_time': 0.0006720349774695933, + 'model_time': 1.2556129040021915, + 'grad_norm_pre_clip_avg': 0.2694025978446007, + 'learning_rate': 2.4946208771380708e-05, + 'epoch': 1.6} +04/19 [13:35:52] INFO | >> train_qwenlatent.py:487 + Step 6340 | grad_norm_pre_clip=0.2736 | + grad_norm_pre_clip_avg=0.3093 | Metrics: + {'align_loss': 0.022387327626347542, + 'recon_loss': 0.01809035986661911, + 'predict_loss': 0.024237757548689842, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2736380398273468, + 'data_time': 0.0009284830011893064, + 'model_time': 1.275666844972875, + 'grad_norm_pre_clip_avg': 0.30930357575416567, + 'learning_rate': 2.494539743200242e-05, + 'epoch': 1.6} +04/19 [13:36:05] INFO | >> train_qwenlatent.py:487 + Step 6350 | grad_norm_pre_clip=0.2891 | + grad_norm_pre_clip_avg=0.2866 | Metrics: + {'align_loss': 0.023754410445690155, + 'recon_loss': 0.014701424166560173, + 'predict_loss': 0.018043167889118195, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28910303115844727, + 'mae_score': 0.033185642689197985, 'data_time': + 0.0009725560084916651, 'model_time': + 1.3134355040092487, 'grad_norm_pre_clip_avg': + 0.28659971207380297, 'learning_rate': + 2.4944580032980963e-05, 'epoch': 1.6} +04/19 [13:36:18] INFO | >> train_qwenlatent.py:487 + Step 6360 | grad_norm_pre_clip=0.3323 | + grad_norm_pre_clip_avg=0.3385 | Metrics: + {'align_loss': 0.021722545847296715, + 'recon_loss': 0.015257617458701134, + 'predict_loss': 0.021072737872600555, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3323480486869812, + 'data_time': 0.0007302410085685551, + 'model_time': 1.2078140870144125, + 'grad_norm_pre_clip_avg': 0.33850370794534684, + 'learning_rate': 2.494375657471472e-05, + 'epoch': 1.6} +04/19 [13:36:31] INFO | >> train_qwenlatent.py:487 + Step 6370 | grad_norm_pre_clip=0.3601 | + grad_norm_pre_clip_avg=0.3088 | Metrics: + {'align_loss': 0.023353109136223793, + 'recon_loss': 0.01483185961842537, + 'predict_loss': 0.017928654327988625, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3601052165031433, + 'data_time': 0.001126794988522306, + 'model_time': 1.2932642069936264, + 'grad_norm_pre_clip_avg': 0.3087640941143036, + 'learning_rate': 2.494292705760505e-05, + 'epoch': 1.61} +04/19 [13:36:44] INFO | >> train_qwenlatent.py:487 + Step 6380 | grad_norm_pre_clip=0.3149 | + grad_norm_pre_clip_avg=0.2822 | Metrics: + {'align_loss': 0.023039119318127632, + 'recon_loss': 0.014478477649390697, + 'predict_loss': 0.01846081018447876, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3149299621582031, + 'data_time': 0.0009578159952070564, + 'model_time': 1.6089647900080308, + 'grad_norm_pre_clip_avg': 0.2822389602661133, + 'learning_rate': 2.494209148205623e-05, + 'epoch': 1.61} +04/19 [13:36:57] INFO | >> train_qwenlatent.py:487 + Step 6390 | grad_norm_pre_clip=0.2738 | + grad_norm_pre_clip_avg=0.2882 | Metrics: + {'align_loss': 0.023379581049084663, + 'recon_loss': 0.014276085421442986, + 'predict_loss': 0.02356279455125332, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2737877070903778, + 'data_time': 0.0008723920036572963, + 'model_time': 1.2392885349981952, + 'grad_norm_pre_clip_avg': 0.2882170110940933, + 'learning_rate': 2.4941249848475517e-05, + 'epoch': 1.61} +04/19 [13:37:10] INFO | >> train_qwenlatent.py:487 + Step 6400 | grad_norm_pre_clip=0.3052 | + grad_norm_pre_clip_avg=0.3012 | Metrics: + {'align_loss': 0.02241077460348606, + 'recon_loss': 0.009026017971336842, + 'predict_loss': 0.01212191954255104, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.30517834424972534, + 'mae_score': 0.02839444822019285, 'data_time': + 0.000655761978123337, 'model_time': + 1.2315871410246473, 'grad_norm_pre_clip_avg': + 0.3011696696281433, 'learning_rate': + 2.4940402157273115e-05, 'epoch': 1.61} +04/19 [13:37:22] INFO | >> train_qwenlatent.py:487 + Step 6410 | grad_norm_pre_clip=0.2441 | + grad_norm_pre_clip_avg=0.2756 | Metrics: + {'align_loss': 0.022090043872594833, + 'recon_loss': 0.02077280730009079, + 'predict_loss': 0.026345131918787956, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2440718561410904, + 'data_time': 0.0007543370011262596, + 'model_time': 1.2723304689861834, + 'grad_norm_pre_clip_avg': 0.2756098508834839, + 'learning_rate': 2.4939548408862182e-05, + 'epoch': 1.62} +04/19 [13:37:35] INFO | >> train_qwenlatent.py:487 + Step 6420 | grad_norm_pre_clip=0.2686 | + grad_norm_pre_clip_avg=0.2978 | Metrics: + {'align_loss': 0.024469904601573944, + 'recon_loss': 0.012558798305690289, + 'predict_loss': 0.017710400745272636, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26856330037117004, + 'data_time': 0.000846921990159899, + 'model_time': 1.2925310210266616, + 'grad_norm_pre_clip_avg': 0.2977810949087143, + 'learning_rate': 2.4938688603658815e-05, + 'epoch': 1.62} +04/19 [13:37:48] INFO | >> train_qwenlatent.py:487 + Step 6430 | grad_norm_pre_clip=0.3408 | + grad_norm_pre_clip_avg=0.2738 | Metrics: + {'align_loss': 0.0226428285241127, + 'recon_loss': 0.01617576740682125, + 'predict_loss': 0.020203327760100365, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.34076669812202454, + 'data_time': 0.0010318470012862235, + 'model_time': 1.2541417440224905, + 'grad_norm_pre_clip_avg': 0.27376591563224795, + 'learning_rate': 2.4937822742082083e-05, + 'epoch': 1.62} +04/19 [13:38:01] INFO | >> train_qwenlatent.py:487 + Step 6440 | grad_norm_pre_clip=0.3586 | + grad_norm_pre_clip_avg=0.3319 | Metrics: + {'align_loss': 0.023887522518634796, + 'recon_loss': 0.018369974568486214, + 'predict_loss': 0.022301049903035164, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.35857537388801575, + 'data_time': 0.00114883400965482, 'model_time': + 1.2918393529835157, 'grad_norm_pre_clip_avg': + 0.33192418813705443, 'learning_rate': + 2.493695082455399e-05, 'epoch': 1.63} +04/19 [13:38:14] INFO | >> train_qwenlatent.py:487 + Step 6450 | grad_norm_pre_clip=0.3917 | + grad_norm_pre_clip_avg=0.2819 | Metrics: + {'align_loss': 0.023203978314995766, + 'recon_loss': 0.015683839097619057, + 'predict_loss': 0.019607609137892723, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3916965126991272, + 'mae_score': 0.044342594318561725, 'data_time': + 0.0010838970192708075, 'model_time': + 1.2619911069923546, 'grad_norm_pre_clip_avg': + 0.2818958252668381, 'learning_rate': + 2.4936072851499494e-05, 'epoch': 1.63} +04/19 [13:38:27] INFO | >> train_qwenlatent.py:487 + Step 6460 | grad_norm_pre_clip=0.2130 | + grad_norm_pre_clip_avg=0.3117 | Metrics: + {'align_loss': 0.0226083155721426, + 'recon_loss': 0.01599382609128952, + 'predict_loss': 0.019486311823129654, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21296921372413635, + 'data_time': 0.0008350399730261415, + 'model_time': 1.2284515149949584, + 'grad_norm_pre_clip_avg': 0.31166691780090333, + 'learning_rate': 2.4935188823346513e-05, + 'epoch': 1.63} +04/19 [13:38:40] INFO | >> train_qwenlatent.py:487 + Step 6470 | grad_norm_pre_clip=0.2554 | + grad_norm_pre_clip_avg=0.2418 | Metrics: + {'align_loss': 0.02372022718191147, + 'recon_loss': 0.01293029822409153, + 'predict_loss': 0.015822360292077065, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2553763687610626, + 'data_time': 0.000652625021757558, + 'model_time': 1.2461906770186033, + 'grad_norm_pre_clip_avg': 0.24180976450443267, + 'learning_rate': 2.4934298740525917e-05, + 'epoch': 1.63} +04/19 [13:38:52] INFO | >> train_qwenlatent.py:487 + Step 6480 | grad_norm_pre_clip=0.2821 | + grad_norm_pre_clip_avg=0.2676 | Metrics: + {'align_loss': 0.023020818829536438, + 'recon_loss': 0.016339967027306557, + 'predict_loss': 0.021543612703680992, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2820863425731659, + 'data_time': 0.0007345839985646307, + 'model_time': 1.2445168889826164, + 'grad_norm_pre_clip_avg': 0.2676438122987747, + 'learning_rate': 2.4933402603471516e-05, + 'epoch': 1.64} +04/19 [13:39:05] INFO | >> train_qwenlatent.py:487 + Step 6490 | grad_norm_pre_clip=0.3659 | + grad_norm_pre_clip_avg=0.3001 | Metrics: + {'align_loss': 0.023858770728111267, + 'recon_loss': 0.010684187524020672, + 'predict_loss': 0.014867257326841354, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.36592385172843933, + 'data_time': 0.0009483869944233447, + 'model_time': 1.2401368680002633, + 'grad_norm_pre_clip_avg': 0.30009941309690474, + 'learning_rate': 2.4932500412620078e-05, + 'epoch': 1.64} +04/19 [13:39:18] INFO | >> train_qwenlatent.py:487 + Step 6500 | grad_norm_pre_clip=0.3441 | + grad_norm_pre_clip_avg=0.2940 | Metrics: + {'align_loss': 0.022210024297237396, + 'recon_loss': 0.011966913938522339, + 'predict_loss': 0.019569961354136467, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3440958261489868, + 'mae_score': 0.02769609056077562, 'data_time': + 0.0009573329880367965, 'model_time': + 1.241029422992142, 'grad_norm_pre_clip_avg': + 0.2939545840024948, 'learning_rate': + 2.4931592168411314e-05, 'epoch': 1.64} +04/19 [13:39:31] INFO | >> train_qwenlatent.py:487 + Step 6510 | grad_norm_pre_clip=0.3875 | + grad_norm_pre_clip_avg=0.2768 | Metrics: + {'align_loss': 0.024195145815610886, + 'recon_loss': 0.013634245842695236, + 'predict_loss': 0.01998595893383026, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.38752901554107666, + 'data_time': 0.0007795330020599067, + 'model_time': 1.253813123999862, + 'grad_norm_pre_clip_avg': 0.27676260471343994, + 'learning_rate': 2.49306778712879e-05, 'epoch': + 1.64} +04/19 [13:39:44] INFO | >> train_qwenlatent.py:487 + Step 6520 | grad_norm_pre_clip=0.3072 | + grad_norm_pre_clip_avg=0.3456 | Metrics: + {'align_loss': 0.023749757558107376, + 'recon_loss': 0.017457734793424606, + 'predict_loss': 0.020999686792492867, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.30721768736839294, + 'data_time': 0.0009853419905994087, + 'model_time': 1.228441710001789, + 'grad_norm_pre_clip_avg': 0.3456109821796417, + 'learning_rate': 2.4929757521695445e-05, + 'epoch': 1.65} +04/19 [13:39:56] INFO | >> train_qwenlatent.py:487 + Step 6530 | grad_norm_pre_clip=0.2595 | + grad_norm_pre_clip_avg=0.2691 | Metrics: + {'align_loss': 0.023541491478681564, + 'recon_loss': 0.011569364927709103, + 'predict_loss': 0.020531831309199333, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25950270891189575, + 'data_time': 0.0007365009805653244, + 'model_time': 1.2296091339958366, + 'grad_norm_pre_clip_avg': 0.2691055864095688, + 'learning_rate': 2.4928831120082525e-05, + 'epoch': 1.65} +04/19 [13:40:09] INFO | >> train_qwenlatent.py:487 + Step 6540 | grad_norm_pre_clip=0.3191 | + grad_norm_pre_clip_avg=0.2572 | Metrics: + {'align_loss': 0.02293863520026207, + 'recon_loss': 0.015441265888512135, + 'predict_loss': 0.018933532759547234, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3190520405769348, + 'data_time': 0.0007397089793812484, + 'model_time': 1.252438516006805, + 'grad_norm_pre_clip_avg': 0.2571966081857681, + 'learning_rate': 2.4927898666900654e-05, + 'epoch': 1.65} +04/19 [13:40:22] INFO | >> train_qwenlatent.py:487 + Step 6550 | grad_norm_pre_clip=0.4109 | + grad_norm_pre_clip_avg=0.3144 | Metrics: + {'align_loss': 0.025205962359905243, + 'recon_loss': 0.023353250697255135, + 'predict_loss': 0.02529584802687168, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.4108968675136566, + 'mae_score': 0.03259380443676098, 'data_time': + 0.0006538959860336035, 'model_time': + 1.220070263982052, 'grad_norm_pre_clip_avg': + 0.3144287049770355, 'learning_rate': + 2.492696016260429e-05, 'epoch': 1.65} +04/19 [13:40:35] INFO | >> train_qwenlatent.py:487 + Step 6560 | grad_norm_pre_clip=0.2502 | + grad_norm_pre_clip_avg=0.2977 | Metrics: + {'align_loss': 0.023773496970534325, + 'recon_loss': 0.016511734575033188, + 'predict_loss': 0.018874943256378174, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2501998841762543, + 'data_time': 0.000783220020821318, + 'model_time': 1.2211856580106542, + 'grad_norm_pre_clip_avg': 0.29767687171697615, + 'learning_rate': 2.492601560765086e-05, + 'epoch': 1.66} +04/19 [13:40:48] INFO | >> train_qwenlatent.py:487 + Step 6570 | grad_norm_pre_clip=0.3049 | + grad_norm_pre_clip_avg=0.2650 | Metrics: + {'align_loss': 0.022243518382310867, + 'recon_loss': 0.012168130837380886, + 'predict_loss': 0.01506153866648674, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3049273192882538, + 'data_time': 0.000925872998777777, + 'model_time': 1.203223533986602, + 'grad_norm_pre_clip_avg': 0.2650150924921036, + 'learning_rate': 2.4925065002500723e-05, + 'epoch': 1.66} +04/19 [13:41:00] INFO | >> train_qwenlatent.py:487 + Step 6580 | grad_norm_pre_clip=0.4658 | + grad_norm_pre_clip_avg=0.2900 | Metrics: + {'align_loss': 0.02204623818397522, + 'recon_loss': 0.020812690258026123, + 'predict_loss': 0.02233685739338398, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.46579474210739136, + 'data_time': 0.0006575259903911501, + 'model_time': 1.2285791680042166, + 'grad_norm_pre_clip_avg': 0.29001289457082746, + 'learning_rate': 2.4924108347617188e-05, + 'epoch': 1.66} +04/19 [13:41:13] INFO | >> train_qwenlatent.py:487 + Step 6590 | grad_norm_pre_clip=0.2419 | + grad_norm_pre_clip_avg=0.2922 | Metrics: + {'align_loss': 0.023803099989891052, + 'recon_loss': 0.017080558463931084, + 'predict_loss': 0.021800201386213303, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24193087220191956, + 'data_time': 0.0011610410001594573, + 'model_time': 1.277568092016736, + 'grad_norm_pre_clip_avg': 0.29220269024372103, + 'learning_rate': 2.492314564346653e-05, + 'epoch': 1.66} +04/19 [13:41:26] INFO | >> train_qwenlatent.py:487 + Step 6600 | grad_norm_pre_clip=0.3070 | + grad_norm_pre_clip_avg=0.2536 | Metrics: + {'align_loss': 0.023867124691605568, + 'recon_loss': 0.01471552811563015, + 'predict_loss': 0.017788322642445564, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.30704402923583984, + 'mae_score': 0.025950392302092132, 'data_time': + 0.0008218339935410768, 'model_time': + 1.290216908993898, 'grad_norm_pre_clip_avg': + 0.2536118119955063, 'learning_rate': + 2.492217689051795e-05, 'epoch': 1.67} +04/19 [13:41:39] INFO | >> train_qwenlatent.py:487 + Step 6610 | grad_norm_pre_clip=0.3176 | + grad_norm_pre_clip_avg=0.3248 | Metrics: + {'align_loss': 0.023513179272413254, + 'recon_loss': 0.022177502512931824, + 'predict_loss': 0.027873927727341652, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3175981640815735, + 'data_time': 0.0012497730203904212, + 'model_time': 1.2488123109797016, + 'grad_norm_pre_clip_avg': 0.32478199899196625, + 'learning_rate': 2.4921202089243605e-05, + 'epoch': 1.67} +04/19 [13:41:52] INFO | >> train_qwenlatent.py:487 + Step 6620 | grad_norm_pre_clip=0.2950 | + grad_norm_pre_clip_avg=0.3191 | Metrics: + {'align_loss': 0.02317832224071026, + 'recon_loss': 0.01333142351359129, + 'predict_loss': 0.016121642664074898, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.29504528641700745, + 'data_time': 0.0007273459923453629, + 'model_time': 1.2493644019996282, + 'grad_norm_pre_clip_avg': 0.3190876871347427, + 'learning_rate': 2.4920221240118604e-05, + 'epoch': 1.67} +04/19 [13:42:04] INFO | >> train_qwenlatent.py:487 + Step 6630 | grad_norm_pre_clip=0.3328 | + grad_norm_pre_clip_avg=0.2883 | Metrics: + {'align_loss': 0.023013249039649963, + 'recon_loss': 0.018023047596216202, + 'predict_loss': 0.020051244646310806, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.33283647894859314, + 'data_time': 0.0008960310078691691, + 'model_time': 1.2187185040093027, + 'grad_norm_pre_clip_avg': 0.28830406665802, + 'learning_rate': 2.4919234343621006e-05, + 'epoch': 1.67} +04/19 [13:42:17] INFO | >> train_qwenlatent.py:487 + Step 6640 | grad_norm_pre_clip=0.2436 | + grad_norm_pre_clip_avg=0.3566 | Metrics: + {'align_loss': 0.02323264814913273, + 'recon_loss': 0.019128501415252686, + 'predict_loss': 0.020260954275727272, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24355918169021606, + 'data_time': 0.0006876250263303518, + 'model_time': 1.2374560309981462, + 'grad_norm_pre_clip_avg': 0.3566062733530998, + 'learning_rate': 2.4918241400231802e-05, + 'epoch': 1.68} +04/19 [13:42:30] INFO | >> train_qwenlatent.py:487 + Step 6650 | grad_norm_pre_clip=0.2393 | + grad_norm_pre_clip_avg=0.2518 | Metrics: + {'align_loss': 0.023177266120910645, + 'recon_loss': 0.02019217610359192, + 'predict_loss': 0.02525322698056698, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23930898308753967, + 'mae_score': 0.03005980414313239, 'data_time': + 0.0010520819923840463, 'model_time': + 1.527837424975587, 'grad_norm_pre_clip_avg': + 0.2518113061785698, 'learning_rate': + 2.4917242410434952e-05, 'epoch': 1.68} +04/19 [13:42:43] INFO | >> train_qwenlatent.py:487 + Step 6660 | grad_norm_pre_clip=0.3163 | + grad_norm_pre_clip_avg=0.2695 | Metrics: + {'align_loss': 0.024259774014353752, + 'recon_loss': 0.01489944476634264, + 'predict_loss': 0.020006297156214714, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.31631776690483093, + 'data_time': 0.0007270610076375306, + 'model_time': 1.2710820010106545, + 'grad_norm_pre_clip_avg': 0.2694904237985611, + 'learning_rate': 2.4916237374717345e-05, + 'epoch': 1.68} +04/19 [13:42:55] INFO | >> train_qwenlatent.py:487 + Step 6670 | grad_norm_pre_clip=0.2360 | + grad_norm_pre_clip_avg=0.2903 | Metrics: + {'align_loss': 0.024475349113345146, + 'recon_loss': 0.016004422679543495, + 'predict_loss': 0.018473120406270027, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23599277436733246, + 'data_time': 0.0012318530061747879, + 'model_time': 1.2537069089885335, + 'grad_norm_pre_clip_avg': 0.2902619540691376, + 'learning_rate': 2.491522629356882e-05, + 'epoch': 1.68} +04/19 [13:43:08] INFO | >> train_qwenlatent.py:487 + Step 6680 | grad_norm_pre_clip=0.3091 | + grad_norm_pre_clip_avg=0.2335 | Metrics: + {'align_loss': 0.021643174812197685, + 'recon_loss': 0.016884908080101013, + 'predict_loss': 0.02269626408815384, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3090619444847107, + 'data_time': 0.0009332130139227957, + 'model_time': 1.2431587689789012, + 'grad_norm_pre_clip_avg': 0.23349154591560364, + 'learning_rate': 2.4914209167482172e-05, + 'epoch': 1.69} +04/19 [13:43:21] INFO | >> train_qwenlatent.py:487 + Step 6690 | grad_norm_pre_clip=0.3044 | + grad_norm_pre_clip_avg=0.2971 | Metrics: + {'align_loss': 0.023611873388290405, + 'recon_loss': 0.01756863482296467, + 'predict_loss': 0.018441058695316315, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3043998181819916, + 'data_time': 0.0007399269961751997, + 'model_time': 1.214475864020642, + 'grad_norm_pre_clip_avg': 0.2970838561654091, + 'learning_rate': 2.491318599695313e-05, + 'epoch': 1.69} +04/19 [13:43:34] INFO | >> train_qwenlatent.py:487 + Step 6700 | grad_norm_pre_clip=0.2946 | + grad_norm_pre_clip_avg=0.2997 | Metrics: + {'align_loss': 0.023127852007746696, + 'recon_loss': 0.01354094035923481, + 'predict_loss': 0.01504578534513712, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.29461470246315, + 'mae_score': 0.04751507527119404, 'data_time': + 0.0009627240069676191, 'model_time': + 1.2249574610032141, 'grad_norm_pre_clip_avg': + 0.29971565008163453, 'learning_rate': + 2.491215678248038e-05, 'epoch': 1.69} +04/19 [13:43:47] INFO | >> train_qwenlatent.py:487 + Step 6710 | grad_norm_pre_clip=0.3859 | + grad_norm_pre_clip_avg=0.2908 | Metrics: + {'align_loss': 0.023595605045557022, + 'recon_loss': 0.02118504047393799, + 'predict_loss': 0.020992722362279892, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.38589105010032654, + 'data_time': 0.0007966969860717654, + 'model_time': 1.1972472009947523, + 'grad_norm_pre_clip_avg': 0.29083519577980044, + 'learning_rate': 2.4911121524565546e-05, + 'epoch': 1.69} +04/19 [13:44:00] INFO | >> train_qwenlatent.py:487 + Step 6720 | grad_norm_pre_clip=0.2147 | + grad_norm_pre_clip_avg=0.2686 | Metrics: + {'align_loss': 0.023660873994231224, + 'recon_loss': 0.017521847039461136, + 'predict_loss': 0.019295362755656242, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21470503509044647, + 'data_time': 0.0007200909894891083, + 'model_time': 1.232255829992937, + 'grad_norm_pre_clip_avg': 0.2686315968632698, + 'learning_rate': 2.4910080223713205e-05, + 'epoch': 1.7} +04/19 [13:44:12] INFO | >> train_qwenlatent.py:487 + Step 6730 | grad_norm_pre_clip=0.3102 | + grad_norm_pre_clip_avg=0.2967 | Metrics: + {'align_loss': 0.023053783923387527, + 'recon_loss': 0.01563713327050209, + 'predict_loss': 0.01934169791638851, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.31018927693367004, + 'data_time': 0.0008183059981092811, + 'model_time': 1.24015348701505, + 'grad_norm_pre_clip_avg': 0.2966894403100014, + 'learning_rate': 2.4909032880430873e-05, + 'epoch': 1.7} +04/19 [13:44:25] INFO | >> train_qwenlatent.py:487 + Step 6740 | grad_norm_pre_clip=0.2380 | + grad_norm_pre_clip_avg=0.2964 | Metrics: + {'align_loss': 0.023997146636247635, + 'recon_loss': 0.015171443112194538, + 'predict_loss': 0.016553377732634544, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23800110816955566, + 'data_time': 0.0007183400157373399, + 'model_time': 1.271586938004475, + 'grad_norm_pre_clip_avg': 0.2963880389928818, + 'learning_rate': 2.4907979495229002e-05, + 'epoch': 1.7} +04/19 [13:44:38] INFO | >> train_qwenlatent.py:487 + Step 6750 | grad_norm_pre_clip=0.2283 | + grad_norm_pre_clip_avg=0.3309 | Metrics: + {'align_loss': 0.02424577623605728, + 'recon_loss': 0.016669219359755516, + 'predict_loss': 0.022566363215446472, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22828389704227448, + 'mae_score': 0.02311765653593046, 'data_time': + 0.000861010019434616, 'model_time': + 1.2630744079942815, 'grad_norm_pre_clip_avg': + 0.3308886423707008, 'learning_rate': + 2.490692006862101e-05, 'epoch': 1.7} +04/19 [13:44:51] INFO | >> train_qwenlatent.py:487 + Step 6760 | grad_norm_pre_clip=0.2357 | + grad_norm_pre_clip_avg=0.2669 | Metrics: + {'align_loss': 0.02328050322830677, + 'recon_loss': 0.012126415967941284, + 'predict_loss': 0.01601537875831127, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23570331931114197, + 'data_time': 0.0007715749961789697, + 'model_time': 1.2103656039980706, + 'grad_norm_pre_clip_avg': 0.26688793003559114, + 'learning_rate': 2.4905854601123246e-05, + 'epoch': 1.71} +04/19 [13:45:04] INFO | >> train_qwenlatent.py:487 + Step 6770 | grad_norm_pre_clip=0.3868 | + grad_norm_pre_clip_avg=0.2990 | Metrics: + {'align_loss': 0.02348106913268566, + 'recon_loss': 0.019665975123643875, + 'predict_loss': 0.020467456430196762, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.38675573468208313, + 'data_time': 0.0011332730064168572, + 'model_time': 1.2659370219917037, + 'grad_norm_pre_clip_avg': 0.29899183511734007, + 'learning_rate': 2.4904783093255003e-05, + 'epoch': 1.71} +04/19 [13:45:17] INFO | >> train_qwenlatent.py:487 + Step 6780 | grad_norm_pre_clip=0.2494 | + grad_norm_pre_clip_avg=0.3098 | Metrics: + {'align_loss': 0.0226345993578434, + 'recon_loss': 0.01357272919267416, + 'predict_loss': 0.016763413324952126, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2493591457605362, + 'data_time': 0.001087820011889562, + 'model_time': 1.3250970999943092, + 'grad_norm_pre_clip_avg': 0.3097981333732605, + 'learning_rate': 2.4903705545538526e-05, + 'epoch': 1.71} +04/19 [13:45:29] INFO | >> train_qwenlatent.py:487 + Step 6790 | grad_norm_pre_clip=0.4067 | + grad_norm_pre_clip_avg=0.3124 | Metrics: + {'align_loss': 0.025623295456171036, + 'recon_loss': 0.026050416752696037, + 'predict_loss': 0.024711713194847107, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.4066968560218811, + 'data_time': 0.0007069669954944402, + 'model_time': 1.2613695969921537, + 'grad_norm_pre_clip_avg': 0.3123826444149017, + 'learning_rate': 2.4902621958499002e-05, + 'epoch': 1.71} +04/19 [13:45:43] INFO | >> train_qwenlatent.py:487 + Step 6800 | grad_norm_pre_clip=0.2657 | + grad_norm_pre_clip_avg=0.2692 | Metrics: + {'align_loss': 0.022880353033542633, + 'recon_loss': 0.0142191331833601, + 'predict_loss': 0.0189474206417799, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26574161648750305, + 'mae_score': 0.024155396813744897, 'data_time': + 0.0010673379874788225, 'model_time': + 1.258126041997457, 'grad_norm_pre_clip_avg': + 0.269157811999321, 'learning_rate': + 2.4901532332664546e-05, 'epoch': 1.72} +04/19 [13:45:55] INFO | >> train_qwenlatent.py:487 + Step 6810 | grad_norm_pre_clip=0.4089 | + grad_norm_pre_clip_avg=0.2999 | Metrics: + {'align_loss': 0.02306296117603779, + 'recon_loss': 0.017311636358499527, + 'predict_loss': 0.020217204466462135, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.40886539220809937, + 'data_time': 0.0008429849985986948, + 'model_time': 1.231246832991019, + 'grad_norm_pre_clip_avg': 0.29986241459846497, + 'learning_rate': 2.4900436668566234e-05, + 'epoch': 1.72} +04/19 [13:46:09] INFO | >> train_qwenlatent.py:487 + Step 6820 | grad_norm_pre_clip=0.3001 | + grad_norm_pre_clip_avg=0.3361 | Metrics: + {'align_loss': 0.022747471928596497, + 'recon_loss': 0.013697282411158085, + 'predict_loss': 0.01528535969555378, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3000551462173462, + 'data_time': 0.0009781999979168177, + 'model_time': 1.5111945340177044, + 'grad_norm_pre_clip_avg': 0.3360827386379242, + 'learning_rate': 2.489933496673808e-05, + 'epoch': 1.72} +04/19 [13:46:21] INFO | >> train_qwenlatent.py:487 + Step 6830 | grad_norm_pre_clip=0.3493 | + grad_norm_pre_clip_avg=0.3256 | Metrics: + {'align_loss': 0.023506343364715576, + 'recon_loss': 0.012393906712532043, + 'predict_loss': 0.017161866649985313, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.34927693009376526, + 'data_time': 0.0008826109988149256, + 'model_time': 1.207250748993829, + 'grad_norm_pre_clip_avg': 0.3256495863199234, + 'learning_rate': 2.489822722771704e-05, + 'epoch': 1.72} +04/19 [13:46:34] INFO | >> train_qwenlatent.py:487 + Step 6840 | grad_norm_pre_clip=0.2538 | + grad_norm_pre_clip_avg=0.2885 | Metrics: + {'align_loss': 0.023559968918561935, + 'recon_loss': 0.018738223239779472, + 'predict_loss': 0.017046479508280754, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25384899973869324, + 'data_time': 0.0010427480156067759, + 'model_time': 1.2499794060131535, + 'grad_norm_pre_clip_avg': 0.2884703353047371, + 'learning_rate': 2.4897113452043017e-05, + 'epoch': 1.73} +04/19 [13:46:47] INFO | >> train_qwenlatent.py:487 + Step 6850 | grad_norm_pre_clip=0.2059 | + grad_norm_pre_clip_avg=0.2467 | Metrics: + {'align_loss': 0.02052878402173519, + 'recon_loss': 0.012646570801734924, + 'predict_loss': 0.015156586654484272, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20591501891613007, + 'mae_score': 0.023650633322226035, 'data_time': + 0.0006348979950416833, 'model_time': + 1.2588780740043148, 'grad_norm_pre_clip_avg': + 0.2466840237379074, 'learning_rate': + 2.489599364025884e-05, 'epoch': 1.73} +04/19 [13:47:00] INFO | >> train_qwenlatent.py:487 + Step 6860 | grad_norm_pre_clip=0.3679 | + grad_norm_pre_clip_avg=0.3104 | Metrics: + {'align_loss': 0.025082075968384743, + 'recon_loss': 0.01967647485435009, + 'predict_loss': 0.023648273199796677, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.36787527799606323, + 'data_time': 0.0006783769931644201, + 'model_time': 1.2394415970193222, + 'grad_norm_pre_clip_avg': 0.3104326009750366, + 'learning_rate': 2.4894867792910308e-05, + 'epoch': 1.73} +04/19 [13:47:12] INFO | >> train_qwenlatent.py:487 + Step 6870 | grad_norm_pre_clip=0.3348 | + grad_norm_pre_clip_avg=0.3251 | Metrics: + {'align_loss': 0.024398066103458405, + 'recon_loss': 0.017567163333296776, + 'predict_loss': 0.01887381449341774, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.33481162786483765, + 'data_time': 0.0007742960005998611, + 'model_time': 1.232275636983104, + 'grad_norm_pre_clip_avg': 0.32509941458702085, + 'learning_rate': 2.4893735910546128e-05, + 'epoch': 1.73} +04/19 [13:47:25] INFO | >> train_qwenlatent.py:487 + Step 6880 | grad_norm_pre_clip=0.2705 | + grad_norm_pre_clip_avg=0.2739 | Metrics: + {'align_loss': 0.02428124099969864, + 'recon_loss': 0.01963752508163452, + 'predict_loss': 0.022374698892235756, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2704584300518036, + 'data_time': 0.0007034080044832081, + 'model_time': 1.2733685529965442, + 'grad_norm_pre_clip_avg': 0.2739002287387848, + 'learning_rate': 2.489259799371798e-05, + 'epoch': 1.74} +04/19 [13:47:38] INFO | >> train_qwenlatent.py:487 + Step 6890 | grad_norm_pre_clip=0.2069 | + grad_norm_pre_clip_avg=0.2350 | Metrics: + {'align_loss': 0.024434557184576988, + 'recon_loss': 0.018285633996129036, + 'predict_loss': 0.01714176870882511, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20694029331207275, + 'data_time': 0.000673586007906124, + 'model_time': 1.22306593801477, + 'grad_norm_pre_clip_avg': 0.235002338886261, + 'learning_rate': 2.4891454042980462e-05, + 'epoch': 1.74} +04/19 [13:47:51] INFO | >> train_qwenlatent.py:487 + Step 6900 | grad_norm_pre_clip=0.2708 | + grad_norm_pre_clip_avg=0.2590 | Metrics: + {'align_loss': 0.023500535637140274, + 'recon_loss': 0.015097936615347862, + 'predict_loss': 0.016874780878424644, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.27076151967048645, + 'mae_score': 0.035220498437280055, 'data_time': + 0.0007534680189564824, 'model_time': + 1.2302243059966713, 'grad_norm_pre_clip_avg': + 0.25898787975311277, 'learning_rate': + 2.489030405889113e-05, 'epoch': 1.74} +04/19 [13:48:04] INFO | >> train_qwenlatent.py:487 + Step 6910 | grad_norm_pre_clip=0.3658 | + grad_norm_pre_clip_avg=0.3209 | Metrics: + {'align_loss': 0.0243149995803833, + 'recon_loss': 0.01906486786901951, + 'predict_loss': 0.021128607913851738, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3658403754234314, + 'data_time': 0.0007142910035327077, + 'model_time': 1.2022191380092409, + 'grad_norm_pre_clip_avg': 0.3209456026554108, + 'learning_rate': 2.488914804201046e-05, + 'epoch': 1.74} +04/19 [13:48:17] INFO | >> train_qwenlatent.py:487 + Step 6920 | grad_norm_pre_clip=0.3613 | + grad_norm_pre_clip_avg=0.3638 | Metrics: + {'align_loss': 0.02318665385246277, + 'recon_loss': 0.019293976947665215, + 'predict_loss': 0.023302292451262474, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.36134448647499084, + 'data_time': 0.001013826986309141, + 'model_time': 1.2551046090084128, + 'grad_norm_pre_clip_avg': 0.36379265785217285, + 'learning_rate': 2.4887985992901892e-05, + 'epoch': 1.75} +04/19 [13:48:29] INFO | >> train_qwenlatent.py:487 + Step 6930 | grad_norm_pre_clip=0.2676 | + grad_norm_pre_clip_avg=0.3070 | Metrics: + {'align_loss': 0.022639799863100052, + 'recon_loss': 0.014586545526981354, + 'predict_loss': 0.013927087187767029, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2676127552986145, + 'data_time': 0.000662786012981087, + 'model_time': 1.237171651009703, + 'grad_norm_pre_clip_avg': 0.3069847121834755, + 'learning_rate': 2.4886817912131787e-05, + 'epoch': 1.75} +04/19 [13:48:42] INFO | >> train_qwenlatent.py:487 + Step 6940 | grad_norm_pre_clip=0.2336 | + grad_norm_pre_clip_avg=0.2517 | Metrics: + {'align_loss': 0.02306942641735077, + 'recon_loss': 0.015596204437315464, + 'predict_loss': 0.01599842496216297, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2335500568151474, + 'data_time': 0.0008624109905213118, + 'model_time': 1.5190409480128437, + 'grad_norm_pre_clip_avg': 0.2516505107283592, + 'learning_rate': 2.488564380026946e-05, + 'epoch': 1.75} +04/19 [13:48:56] INFO | >> train_qwenlatent.py:487 + Step 6950 | grad_norm_pre_clip=0.3372 | + grad_norm_pre_clip_avg=0.2607 | Metrics: + {'align_loss': 0.024308394640684128, + 'recon_loss': 0.021976197138428688, + 'predict_loss': 0.02138490416109562, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.337180495262146, + 'mae_score': 0.02599880802738774, 'data_time': + 0.001431023993063718, 'model_time': + 1.2858612420095596, 'grad_norm_pre_clip_avg': + 0.2606697231531143, 'learning_rate': + 2.4884463657887154e-05, 'epoch': 1.75} +04/19 [13:49:09] INFO | >> train_qwenlatent.py:487 + Step 6960 | grad_norm_pre_clip=0.2294 | + grad_norm_pre_clip_avg=0.4809 | Metrics: + {'align_loss': 0.02319256216287613, + 'recon_loss': 0.01864125020802021, + 'predict_loss': 0.020917769521474838, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2294308990240097, + 'data_time': 0.0008435040072072297, + 'model_time': 1.2296408940164838, + 'grad_norm_pre_clip_avg': 0.48086156100034716, + 'learning_rate': 2.4883277485560055e-05, + 'epoch': 1.76} +04/19 [13:49:21] INFO | >> train_qwenlatent.py:487 + Step 6970 | grad_norm_pre_clip=0.3452 | + grad_norm_pre_clip_avg=0.2913 | Metrics: + {'align_loss': 0.025039147585630417, + 'recon_loss': 0.02366628870368004, + 'predict_loss': 0.02918321080505848, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3452470004558563, + 'data_time': 0.0009555339929647744, + 'model_time': 1.3390674109978136, + 'grad_norm_pre_clip_avg': 0.2912568002939224, + 'learning_rate': 2.4882085283866293e-05, + 'epoch': 1.76} +04/19 [13:49:34] INFO | >> train_qwenlatent.py:487 + Step 6980 | grad_norm_pre_clip=0.2766 | + grad_norm_pre_clip_avg=0.2695 | Metrics: + {'align_loss': 0.024032417684793472, + 'recon_loss': 0.02178194373846054, + 'predict_loss': 0.019047576934099197, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.27660417556762695, + 'data_time': 0.0006541760230902582, + 'model_time': 1.2372614130144939, + 'grad_norm_pre_clip_avg': 0.2694996610283852, + 'learning_rate': 2.488088705338693e-05, + 'epoch': 1.76} +04/19 [13:49:46] INFO | >> train_qwenlatent.py:487 + Step 6990 | grad_norm_pre_clip=0.2371 | + grad_norm_pre_clip_avg=0.2649 | Metrics: + {'align_loss': 0.023312989622354507, + 'recon_loss': 0.01521065179258585, + 'predict_loss': 0.017389876767992973, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23706939816474915, + 'data_time': 0.0006896230042912066, + 'model_time': 1.2724526580132078, + 'grad_norm_pre_clip_avg': 0.26487255096435547, + 'learning_rate': 2.4879682794705975e-05, + 'epoch': 1.76} +04/19 [13:50:00] INFO | >> train_qwenlatent.py:487 + Step 7000 | grad_norm_pre_clip=0.2195 | + grad_norm_pre_clip_avg=0.2842 | Metrics: + {'align_loss': 0.023397359997034073, + 'recon_loss': 0.015363255515694618, + 'predict_loss': 0.01972978375852108, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21950556337833405, + 'mae_score': 0.02887425121960339, 'data_time': + 0.0008238390146289021, 'model_time': + 1.2710944199934602, 'grad_norm_pre_clip_avg': + 0.2842272236943245, 'learning_rate': + 2.487847250841036e-05, 'epoch': 1.77} +04/19 [13:50:12] INFO | >> train_qwenlatent.py:487 + Step 7010 | grad_norm_pre_clip=0.2462 | + grad_norm_pre_clip_avg=0.2688 | Metrics: + {'align_loss': 0.02345956489443779, + 'recon_loss': 0.01671491004526615, + 'predict_loss': 0.015025421045720577, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24624650180339813, + 'data_time': 0.0007313039968721569, + 'model_time': 1.318654886999866, + 'grad_norm_pre_clip_avg': 0.2688447400927544, + 'learning_rate': 2.487725619508997e-05, + 'epoch': 1.77} +04/19 [13:50:25] INFO | >> train_qwenlatent.py:487 + Step 7020 | grad_norm_pre_clip=0.4139 | + grad_norm_pre_clip_avg=0.3094 | Metrics: + {'align_loss': 0.023326914757490158, + 'recon_loss': 0.021848097443580627, + 'predict_loss': 0.025191539898514748, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.4138658344745636, + 'data_time': 0.0006418520060833544, + 'model_time': 1.221149635995971, + 'grad_norm_pre_clip_avg': 0.3093575194478035, + 'learning_rate': 2.4876033855337615e-05, + 'epoch': 1.77} +04/19 [13:50:37] INFO | >> train_qwenlatent.py:487 + Step 7030 | grad_norm_pre_clip=0.3967 | + grad_norm_pre_clip_avg=0.3523 | Metrics: + {'align_loss': 0.024332195520401, 'recon_loss': + 0.015219401568174362, 'predict_loss': + 0.015644390136003494, 'aux_loss_decay_weight': + 0.0, 'grad_norm_pre_clip': 0.3966718316078186, + 'data_time': 0.0006893869722262025, + 'model_time': 1.289566552004544, + 'grad_norm_pre_clip_avg': 0.35232768654823304, + 'learning_rate': 2.4874805489749058e-05, + 'epoch': 1.77} +04/19 [13:50:50] INFO | >> train_qwenlatent.py:487 + Step 7040 | grad_norm_pre_clip=0.2146 | + grad_norm_pre_clip_avg=0.2780 | Metrics: + {'align_loss': 0.022734057158231735, + 'recon_loss': 0.015684526413679123, + 'predict_loss': 0.019325759261846542, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21455562114715576, + 'data_time': 0.0008053880010265857, + 'model_time': 1.2828872680256609, + 'grad_norm_pre_clip_avg': 0.27795042991638186, + 'learning_rate': 2.487357109892298e-05, + 'epoch': 1.78} +04/19 [13:51:03] INFO | >> train_qwenlatent.py:487 + Step 7050 | grad_norm_pre_clip=0.1897 | + grad_norm_pre_clip_avg=0.2549 | Metrics: + {'align_loss': 0.02498612552881241, + 'recon_loss': 0.02034195326268673, + 'predict_loss': 0.018581274896860123, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18967081606388092, + 'mae_score': 0.03285893277005032, 'data_time': + 0.00070833700010553, 'model_time': + 1.2179546120169107, 'grad_norm_pre_clip_avg': + 0.2549144446849823, 'learning_rate': + 2.4872330683461013e-05, 'epoch': 1.78} +04/19 [13:51:16] INFO | >> train_qwenlatent.py:487 + Step 7060 | grad_norm_pre_clip=0.2654 | + grad_norm_pre_clip_avg=0.2384 | Metrics: + {'align_loss': 0.023959970101714134, + 'recon_loss': 0.02322131209075451, + 'predict_loss': 0.023051226511597633, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26535525918006897, + 'data_time': 0.0007602309924550354, + 'model_time': 1.2049986030033324, + 'grad_norm_pre_clip_avg': 0.2384115919470787, + 'learning_rate': 2.4871084243967724e-05, + 'epoch': 1.78} +04/19 [13:51:29] INFO | >> train_qwenlatent.py:487 + Step 7070 | grad_norm_pre_clip=0.3051 | + grad_norm_pre_clip_avg=0.3872 | Metrics: + {'align_loss': 0.023376960307359695, + 'recon_loss': 0.019565895199775696, + 'predict_loss': 0.019884230569005013, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.30509525537490845, + 'data_time': 0.000958385004196316, + 'model_time': 1.2352150689985137, + 'grad_norm_pre_clip_avg': 0.3872470587491989, + 'learning_rate': 2.4869831781050604e-05, + 'epoch': 1.78} +04/19 [13:51:41] INFO | >> train_qwenlatent.py:487 + Step 7080 | grad_norm_pre_clip=0.2842 | + grad_norm_pre_clip_avg=0.2887 | Metrics: + {'align_loss': 0.022497737780213356, + 'recon_loss': 0.01840258203446865, + 'predict_loss': 0.01692805252969265, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2841569483280182, + 'data_time': 0.0007500569918192923, + 'model_time': 1.260667171998648, + 'grad_norm_pre_clip_avg': 0.28868405520915985, + 'learning_rate': 2.4868573295320094e-05, + 'epoch': 1.79} +04/19 [13:51:54] INFO | >> train_qwenlatent.py:487 + Step 7090 | grad_norm_pre_clip=0.2490 | + grad_norm_pre_clip_avg=0.2487 | Metrics: + {'align_loss': 0.021913982927799225, + 'recon_loss': 0.019625725224614143, + 'predict_loss': 0.01990830898284912, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2489987015724182, + 'data_time': 0.0006607139948755503, + 'model_time': 1.1903845229826402, + 'grad_norm_pre_clip_avg': 0.24871195405721663, + 'learning_rate': 2.4867308787389566e-05, + 'epoch': 1.79} +04/19 [13:52:07] INFO | >> train_qwenlatent.py:487 + Step 7100 | grad_norm_pre_clip=0.2878 | + grad_norm_pre_clip_avg=0.2455 | Metrics: + {'align_loss': 0.02522958815097809, + 'recon_loss': 0.022074462845921516, + 'predict_loss': 0.019924204796552658, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28780126571655273, + 'mae_score': 0.028853425034531602, 'data_time': + 0.001008009974611923, 'model_time': + 1.228585855977144, 'grad_norm_pre_clip_avg': + 0.24554823189973832, 'learning_rate': + 2.4866038257875318e-05, 'epoch': 1.79} +04/19 [13:52:20] INFO | >> train_qwenlatent.py:487 + Step 7110 | grad_norm_pre_clip=0.3844 | + grad_norm_pre_clip_avg=0.2819 | Metrics: + {'align_loss': 0.02325362339615822, + 'recon_loss': 0.015192135237157345, + 'predict_loss': 0.01656334288418293, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.38438186049461365, + 'data_time': 0.0008846349956002086, + 'model_time': 1.245229002990527, + 'grad_norm_pre_clip_avg': 0.28185470700263976, + 'learning_rate': 2.4864761707396597e-05, + 'epoch': 1.79} +04/19 [13:52:33] INFO | >> train_qwenlatent.py:487 + Step 7120 | grad_norm_pre_clip=0.3592 | + grad_norm_pre_clip_avg=0.3183 | Metrics: + {'align_loss': 0.024122634902596474, + 'recon_loss': 0.019518403336405754, + 'predict_loss': 0.01866145059466362, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.35918283462524414, + 'data_time': 0.0007377229921985418, + 'model_time': 1.2055765419790987, + 'grad_norm_pre_clip_avg': 0.3182747051119804, + 'learning_rate': 2.4863479136575578e-05, + 'epoch': 1.8} +04/19 [13:52:45] INFO | >> train_qwenlatent.py:487 + Step 7130 | grad_norm_pre_clip=0.3300 | + grad_norm_pre_clip_avg=0.2978 | Metrics: + {'align_loss': 0.023060893639922142, + 'recon_loss': 0.017073815688490868, + 'predict_loss': 0.019006678834557533, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.32995763421058655, + 'data_time': 0.0006863799935672432, + 'model_time': 1.2214936779928394, + 'grad_norm_pre_clip_avg': 0.2977585181593895, + 'learning_rate': 2.4862190546037367e-05, + 'epoch': 1.8} +04/19 [13:52:58] INFO | >> train_qwenlatent.py:487 + Step 7140 | grad_norm_pre_clip=0.3365 | + grad_norm_pre_clip_avg=0.2654 | Metrics: + {'align_loss': 0.024677656590938568, + 'recon_loss': 0.018302924931049347, + 'predict_loss': 0.022340521216392517, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3365314304828644, + 'data_time': 0.0006739489908795804, + 'model_time': 1.2446197240205947, + 'grad_norm_pre_clip_avg': 0.2653838574886322, + 'learning_rate': 2.486089593641001e-05, + 'epoch': 1.8} +04/19 [13:53:11] INFO | >> train_qwenlatent.py:487 + Step 7150 | grad_norm_pre_clip=0.2459 | + grad_norm_pre_clip_avg=0.2860 | Metrics: + {'align_loss': 0.023747112601995468, + 'recon_loss': 0.0163394957780838, + 'predict_loss': 0.01368008740246296, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24592530727386475, + 'mae_score': 0.02725746910851281, 'data_time': + 0.0008403629763051867, 'model_time': + 1.2788351420022082, 'grad_norm_pre_clip_avg': + 0.2859594523906708, 'learning_rate': + 2.4859595308324487e-05, 'epoch': 1.8} +04/19 [13:53:24] INFO | >> train_qwenlatent.py:487 + Step 7160 | grad_norm_pre_clip=0.2209 | + grad_norm_pre_clip_avg=0.2573 | Metrics: + {'align_loss': 0.025086859241127968, + 'recon_loss': 0.01758428104221821, + 'predict_loss': 0.01651614159345627, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22086495161056519, + 'data_time': 0.0008375599863938987, + 'model_time': 1.1906940350017976, + 'grad_norm_pre_clip_avg': 0.2573230817914009, + 'learning_rate': 2.4858288662414702e-05, + 'epoch': 1.81} +04/19 [13:53:37] INFO | >> train_qwenlatent.py:487 + Step 7170 | grad_norm_pre_clip=0.3072 | + grad_norm_pre_clip_avg=0.2603 | Metrics: + {'align_loss': 0.025315340608358383, + 'recon_loss': 0.017122840508818626, + 'predict_loss': 0.013190284371376038, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.30719730257987976, + 'data_time': 0.0009042950114235282, + 'model_time': 1.259002845006762, + 'grad_norm_pre_clip_avg': 0.26031964272260666, + 'learning_rate': 2.4856975999317502e-05, + 'epoch': 1.81} +04/19 [13:53:49] INFO | >> train_qwenlatent.py:487 + Step 7180 | grad_norm_pre_clip=0.2717 | + grad_norm_pre_clip_avg=0.2737 | Metrics: + {'align_loss': 0.025535855442285538, + 'recon_loss': 0.01840847171843052, + 'predict_loss': 0.01860860548913479, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.27169325947761536, + 'data_time': 0.0008664789784234017, + 'model_time': 1.2130255619995296, + 'grad_norm_pre_clip_avg': 0.2737153172492981, + 'learning_rate': 2.4855657319672662e-05, + 'epoch': 1.81} +04/19 [13:54:02] INFO | >> train_qwenlatent.py:487 + Step 7190 | grad_norm_pre_clip=0.3231 | + grad_norm_pre_clip_avg=0.2968 | Metrics: + {'align_loss': 0.02404172345995903, + 'recon_loss': 0.01842193678021431, + 'predict_loss': 0.017701679840683937, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.32306551933288574, + 'data_time': 0.0006529439997393638, + 'model_time': 1.5240436939930078, + 'grad_norm_pre_clip_avg': 0.29680780619382857, + 'learning_rate': 2.4854332624122896e-05, + 'epoch': 1.81} +04/19 [13:54:15] INFO | >> train_qwenlatent.py:487 + Step 7200 | grad_norm_pre_clip=0.2133 | + grad_norm_pre_clip_avg=0.2859 | Metrics: + {'align_loss': 0.024284712970256805, + 'recon_loss': 0.018616946414113045, + 'predict_loss': 0.015888262540102005, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21331816911697388, + 'mae_score': 0.025977009266346423, 'data_time': + 0.0009245050023309886, 'model_time': + 1.2073404480179306, 'grad_norm_pre_clip_avg': + 0.2858946308493614, 'learning_rate': + 2.485300191331383e-05, 'epoch': 1.82} +04/19 [13:54:28] INFO | >> train_qwenlatent.py:487 + Step 7210 | grad_norm_pre_clip=0.3386 | + grad_norm_pre_clip_avg=0.2794 | Metrics: + {'align_loss': 0.025221051648259163, + 'recon_loss': 0.018152054399251938, + 'predict_loss': 0.018718842417001724, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3386368751525879, + 'data_time': 0.0011505700240377337, + 'model_time': 1.2347612740122713, + 'grad_norm_pre_clip_avg': 0.27935337722301484, + 'learning_rate': 2.4851665187894048e-05, + 'epoch': 1.82} +04/19 [13:54:41] INFO | >> train_qwenlatent.py:487 + Step 7220 | grad_norm_pre_clip=0.2170 | + grad_norm_pre_clip_avg=0.3289 | Metrics: + {'align_loss': 0.02285180613398552, + 'recon_loss': 0.018099868670105934, + 'predict_loss': 0.017545875161886215, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21701286733150482, + 'data_time': 0.0009142170019913465, + 'model_time': 1.2219794829725288, + 'grad_norm_pre_clip_avg': 0.32888083159923553, + 'learning_rate': 2.4850322448515057e-05, + 'epoch': 1.82} +04/19 [13:54:53] INFO | >> train_qwenlatent.py:487 + Step 7230 | grad_norm_pre_clip=0.2461 | + grad_norm_pre_clip_avg=0.2750 | Metrics: + {'align_loss': 0.02315564826130867, + 'recon_loss': 0.014750747010111809, + 'predict_loss': 0.019186927005648613, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2460734248161316, + 'data_time': 0.0006930730014573783, + 'model_time': 1.2331260619976092, + 'grad_norm_pre_clip_avg': 0.274979031085968, + 'learning_rate': 2.484897369583128e-05, + 'epoch': 1.82} +04/19 [13:55:06] INFO | >> train_qwenlatent.py:487 + Step 7240 | grad_norm_pre_clip=0.2598 | + grad_norm_pre_clip_avg=0.2586 | Metrics: + {'align_loss': 0.024104638025164604, + 'recon_loss': 0.013904918916523457, + 'predict_loss': 0.01404566690325737, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25975462794303894, + 'data_time': 0.0010507700208108872, + 'model_time': 1.186495923990151, + 'grad_norm_pre_clip_avg': 0.2586305722594261, + 'learning_rate': 2.484761893050009e-05, + 'epoch': 1.83} +04/19 [13:55:19] INFO | >> train_qwenlatent.py:487 + Step 7250 | grad_norm_pre_clip=0.2485 | + grad_norm_pre_clip_avg=0.2603 | Metrics: + {'align_loss': 0.024842411279678345, + 'recon_loss': 0.025233132764697075, + 'predict_loss': 0.02139466628432274, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24847249686717987, + 'mae_score': 0.029039716290998028, 'data_time': + 0.0011876920179929584, 'model_time': + 1.2408707010035869, 'grad_norm_pre_clip_avg': + 0.2602978408336639, 'learning_rate': + 2.4846258153181783e-05, 'epoch': 1.83} +04/19 [13:55:32] INFO | >> train_qwenlatent.py:487 + Step 7260 | grad_norm_pre_clip=0.2884 | + grad_norm_pre_clip_avg=0.2436 | Metrics: + {'align_loss': 0.024444907903671265, + 'recon_loss': 0.015020851977169514, + 'predict_loss': 0.023076465353369713, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28844311833381653, + 'data_time': 0.0006716880016028881, + 'model_time': 1.2283455549913924, + 'grad_norm_pre_clip_avg': 0.2435927465558052, + 'learning_rate': 2.4844891364539584e-05, + 'epoch': 1.83} +04/19 [13:55:44] INFO | >> train_qwenlatent.py:487 + Step 7270 | grad_norm_pre_clip=0.4248 | + grad_norm_pre_clip_avg=0.4511 | Metrics: + {'align_loss': 0.023474106565117836, + 'recon_loss': 0.026266060769557953, + 'predict_loss': 0.02614881470799446, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.42477840185165405, + 'data_time': 0.0006499989831354469, + 'model_time': 1.2245630139950663, + 'grad_norm_pre_clip_avg': 0.45111929923295974, + 'learning_rate': 2.484351856523965e-05, + 'epoch': 1.83} +04/19 [13:55:57] INFO | >> train_qwenlatent.py:487 + Step 7280 | grad_norm_pre_clip=0.3314 | + grad_norm_pre_clip_avg=0.3109 | Metrics: + {'align_loss': 0.02330770716071129, + 'recon_loss': 0.01748259738087654, + 'predict_loss': 0.01442613173276186, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3314226269721985, + 'data_time': 0.001070663012797013, + 'model_time': 1.3162333090149332, + 'grad_norm_pre_clip_avg': 0.3109013721346855, + 'learning_rate': 2.4842139755951065e-05, + 'epoch': 1.84} +04/19 [13:56:09] INFO | >> train_qwenlatent.py:487 + Step 7290 | grad_norm_pre_clip=0.2841 | + grad_norm_pre_clip_avg=0.2852 | Metrics: + {'align_loss': 0.021869057789444923, + 'recon_loss': 0.017297377809882164, + 'predict_loss': 0.017161667346954346, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28414541482925415, + 'data_time': 0.0010633349884301424, + 'model_time': 1.251569482992636, + 'grad_norm_pre_clip_avg': 0.2852153733372688, + 'learning_rate': 2.484075493734585e-05, + 'epoch': 1.84} +04/19 [13:56:23] INFO | >> train_qwenlatent.py:487 + Step 7300 | grad_norm_pre_clip=0.2843 | + grad_norm_pre_clip_avg=0.2769 | Metrics: + {'align_loss': 0.024182097986340523, + 'recon_loss': 0.020244231447577477, + 'predict_loss': 0.018469804897904396, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28426721692085266, + 'mae_score': 0.02943091177725577, 'data_time': + 0.0008296159794554114, 'model_time': + 1.5790727989806328, 'grad_norm_pre_clip_avg': + 0.27693268954753875, 'learning_rate': + 2.4839364110098938e-05, 'epoch': 1.84} +04/19 [13:56:36] INFO | >> train_qwenlatent.py:487 + Step 7310 | grad_norm_pre_clip=0.3475 | + grad_norm_pre_clip_avg=0.2984 | Metrics: + {'align_loss': 0.023064659908413887, + 'recon_loss': 0.019445551559329033, + 'predict_loss': 0.01835019327700138, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3474796414375305, + 'data_time': 0.0010313969978597015, + 'model_time': 1.2299023399828002, + 'grad_norm_pre_clip_avg': 0.2984413102269173, + 'learning_rate': 2.4837967274888207e-05, + 'epoch': 1.84} +04/19 [13:56:48] INFO | >> train_qwenlatent.py:487 + Step 7320 | grad_norm_pre_clip=0.2775 | + grad_norm_pre_clip_avg=0.2678 | Metrics: + {'align_loss': 0.02464701607823372, + 'recon_loss': 0.022909123450517654, + 'predict_loss': 0.024913132190704346, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2775142192840576, + 'data_time': 0.0007271960203070194, + 'model_time': 1.229191605001688, + 'grad_norm_pre_clip_avg': 0.2678437516093254, + 'learning_rate': 2.4836564432394463e-05, + 'epoch': 1.85} +04/19 [13:57:01] INFO | >> train_qwenlatent.py:487 + Step 7330 | grad_norm_pre_clip=0.2027 | + grad_norm_pre_clip_avg=0.2549 | Metrics: + {'align_loss': 0.02482655644416809, + 'recon_loss': 0.014971941709518433, + 'predict_loss': 0.014053472317755222, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2026907354593277, + 'data_time': 0.0007754600082989782, + 'model_time': 1.2556216930097435, + 'grad_norm_pre_clip_avg': 0.25494742542505267, + 'learning_rate': 2.4835155583301425e-05, + 'epoch': 1.85} +04/19 [13:57:14] INFO | >> train_qwenlatent.py:487 + Step 7340 | grad_norm_pre_clip=0.2777 | + grad_norm_pre_clip_avg=0.2655 | Metrics: + {'align_loss': 0.023544715717434883, + 'recon_loss': 0.018339354544878006, + 'predict_loss': 0.016605935990810394, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.27774330973625183, + 'data_time': 0.0007803820190019906, + 'model_time': 1.251578374998644, + 'grad_norm_pre_clip_avg': 0.26548532396554947, + 'learning_rate': 2.4833740728295752e-05, + 'epoch': 1.85} +04/19 [13:57:27] INFO | >> train_qwenlatent.py:487 + Step 7350 | grad_norm_pre_clip=0.2684 | + grad_norm_pre_clip_avg=0.2647 | Metrics: + {'align_loss': 0.023434612900018692, + 'recon_loss': 0.021559765562415123, + 'predict_loss': 0.019536124542355537, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2683936059474945, + 'mae_score': 0.025185102170652097, 'data_time': + 0.0010384749912191182, 'model_time': + 1.1881998229946475, 'grad_norm_pre_clip_avg': + 0.2646720096468925, 'learning_rate': + 2.483231986806703e-05, 'epoch': 1.85} +04/19 [13:57:40] INFO | >> train_qwenlatent.py:487 + Step 7360 | grad_norm_pre_clip=0.2560 | + grad_norm_pre_clip_avg=0.2606 | Metrics: + {'align_loss': 0.024837253615260124, + 'recon_loss': 0.01910027302801609, + 'predict_loss': 0.01788950525224209, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25599610805511475, + 'data_time': 0.0009402100113220513, + 'model_time': 1.3284340599784628, + 'grad_norm_pre_clip_avg': 0.26059871912002563, + 'learning_rate': 2.483089300330777e-05, + 'epoch': 1.86} +04/19 [13:57:53] INFO | >> train_qwenlatent.py:487 + Step 7370 | grad_norm_pre_clip=0.3576 | + grad_norm_pre_clip_avg=0.3282 | Metrics: + {'align_loss': 0.023589858785271645, + 'recon_loss': 0.01735767163336277, + 'predict_loss': 0.016820887103676796, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3576028645038605, + 'data_time': 0.0009755920036695898, + 'model_time': 1.1974371669930406, + 'grad_norm_pre_clip_avg': 0.32819928526878356, + 'learning_rate': 2.48294601347134e-05, 'epoch': + 1.86} +04/19 [13:58:05] INFO | >> train_qwenlatent.py:487 + Step 7380 | grad_norm_pre_clip=0.2629 | + grad_norm_pre_clip_avg=0.3054 | Metrics: + {'align_loss': 0.024594366550445557, + 'recon_loss': 0.027444960549473763, + 'predict_loss': 0.023148439824581146, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26288020610809326, + 'data_time': 0.000995934009552002, + 'model_time': 1.2235286429931875, + 'grad_norm_pre_clip_avg': 0.3054208606481552, + 'learning_rate': 2.4828021262982294e-05, + 'epoch': 1.86} +04/19 [13:58:18] INFO | >> train_qwenlatent.py:487 + Step 7390 | grad_norm_pre_clip=0.2832 | + grad_norm_pre_clip_avg=0.2727 | Metrics: + {'align_loss': 0.023655850440263748, + 'recon_loss': 0.01954612508416176, + 'predict_loss': 0.016516903415322304, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28322967886924744, + 'data_time': 0.0009442919981665909, + 'model_time': 1.4534683589881752, + 'grad_norm_pre_clip_avg': 0.27274954617023467, + 'learning_rate': 2.482657638881573e-05, + 'epoch': 1.86} +04/19 [13:58:31] INFO | >> train_qwenlatent.py:487 + Step 7400 | grad_norm_pre_clip=0.2316 | + grad_norm_pre_clip_avg=0.2622 | Metrics: + {'align_loss': 0.02458767592906952, + 'recon_loss': 0.01496573444455862, + 'predict_loss': 0.012939171865582466, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23163984715938568, + 'mae_score': 0.0392864296028206, 'data_time': + 0.0007567429856862873, 'model_time': + 1.214764888980426, 'grad_norm_pre_clip_avg': + 0.2621843859553337, 'learning_rate': + 2.4825125512917934e-05, 'epoch': 1.87} +04/19 [13:58:44] INFO | >> train_qwenlatent.py:487 + Step 7410 | grad_norm_pre_clip=0.2485 | + grad_norm_pre_clip_avg=0.2577 | Metrics: + {'align_loss': 0.023425783962011337, + 'recon_loss': 0.018574248999357224, + 'predict_loss': 0.016569210216403008, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24846766889095306, + 'data_time': 0.000868596020154655, + 'model_time': 1.217052883002907, + 'grad_norm_pre_clip_avg': 0.2577486217021942, + 'learning_rate': 2.482366863599603e-05, + 'epoch': 1.87} +04/19 [13:58:56] INFO | >> train_qwenlatent.py:487 + Step 7420 | grad_norm_pre_clip=0.2903 | + grad_norm_pre_clip_avg=0.2773 | Metrics: + {'align_loss': 0.02281482145190239, + 'recon_loss': 0.022908343002200127, + 'predict_loss': 0.017484379932284355, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2903226315975189, + 'data_time': 0.0014058899832889438, + 'model_time': 1.272234253003262, + 'grad_norm_pre_clip_avg': 0.277333901822567, + 'learning_rate': 2.4822205758760094e-05, + 'epoch': 1.87} +04/19 [13:59:09] INFO | >> train_qwenlatent.py:487 + Step 7430 | grad_norm_pre_clip=0.2809 | + grad_norm_pre_clip_avg=0.2690 | Metrics: + {'align_loss': 0.02397296018898487, + 'recon_loss': 0.02100462093949318, + 'predict_loss': 0.016850212588906288, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2808528244495392, + 'data_time': 0.0008864809933584183, + 'model_time': 1.2316518209991045, + 'grad_norm_pre_clip_avg': 0.2690012097358704, + 'learning_rate': 2.4820736881923115e-05, + 'epoch': 1.87} +04/19 [13:59:22] INFO | >> train_qwenlatent.py:487 + Step 7440 | grad_norm_pre_clip=0.3633 | + grad_norm_pre_clip_avg=0.3784 | Metrics: + {'align_loss': 0.023136604577302933, + 'recon_loss': 0.01351982168853283, + 'predict_loss': 0.015441682189702988, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3632591962814331, + 'data_time': 0.0006955479911994189, + 'model_time': 1.269766907003941, + 'grad_norm_pre_clip_avg': 0.3783673316240311, + 'learning_rate': 2.4819262006201e-05, 'epoch': + 1.88} +04/19 [13:59:35] INFO | >> train_qwenlatent.py:487 + Step 7450 | grad_norm_pre_clip=0.2235 | + grad_norm_pre_clip_avg=0.2835 | Metrics: + {'align_loss': 0.024437803775072098, + 'recon_loss': 0.02260708622634411, + 'predict_loss': 0.017875980585813522, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22354453802108765, + 'mae_score': 0.026976310455047333, 'data_time': + 0.0006089730013627559, 'model_time': + 1.2214507730095647, 'grad_norm_pre_clip_avg': + 0.28353311866521835, 'learning_rate': + 2.4817781132312583e-05, 'epoch': 1.88} +04/19 [13:59:47] INFO | >> train_qwenlatent.py:487 + Step 7460 | grad_norm_pre_clip=0.2333 | + grad_norm_pre_clip_avg=0.2461 | Metrics: + {'align_loss': 0.023429319262504578, + 'recon_loss': 0.026213012635707855, + 'predict_loss': 0.021097170189023018, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2332613468170166, + 'data_time': 0.0005901310069020838, + 'model_time': 1.197442509001121, + 'grad_norm_pre_clip_avg': 0.24609171450138093, + 'learning_rate': 2.4816294260979635e-05, + 'epoch': 1.88} +04/19 [13:59:59] INFO | >> train_qwenlatent.py:487 + Step 7470 | grad_norm_pre_clip=0.2379 | + grad_norm_pre_clip_avg=0.2228 | Metrics: + {'align_loss': 0.02506278082728386, + 'recon_loss': 0.023298954591155052, + 'predict_loss': 0.019910059869289398, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23791618645191193, + 'data_time': 0.0006146839878056198, + 'model_time': 1.168167458003154, + 'grad_norm_pre_clip_avg': 0.22277484685182572, + 'learning_rate': 2.4814801392926825e-05, + 'epoch': 1.88} +04/19 [14:00:11] INFO | >> train_qwenlatent.py:487 + Step 7480 | grad_norm_pre_clip=0.3164 | + grad_norm_pre_clip_avg=0.2891 | Metrics: + {'align_loss': 0.023158840835094452, + 'recon_loss': 0.021726632490754128, + 'predict_loss': 0.01902429386973381, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.31641775369644165, + 'data_time': 0.0006777640082873404, + 'model_time': 1.1607115140068345, + 'grad_norm_pre_clip_avg': 0.2891368091106415, + 'learning_rate': 2.4813302528881772e-05, + 'epoch': 1.89} +04/19 [14:00:23] INFO | >> train_qwenlatent.py:487 + Step 7490 | grad_norm_pre_clip=0.3113 | + grad_norm_pre_clip_avg=0.2530 | Metrics: + {'align_loss': 0.024693023413419724, + 'recon_loss': 0.03005068749189377, + 'predict_loss': 0.02156853675842285, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.31128567457199097, + 'data_time': 0.0005930119950789958, + 'model_time': 1.150133814982837, + 'grad_norm_pre_clip_avg': 0.25297039598226545, + 'learning_rate': 2.4811797669574995e-05, + 'epoch': 1.89} +04/19 [14:00:35] INFO | >> train_qwenlatent.py:487 + Step 7500 | grad_norm_pre_clip=0.2030 | + grad_norm_pre_clip_avg=0.2785 | Metrics: + {'align_loss': 0.02270863950252533, + 'recon_loss': 0.016155747696757317, + 'predict_loss': 0.013228829018771648, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20299407839775085, + 'mae_score': 0.025303679113989477, 'data_time': + 0.0006193669978529215, 'model_time': + 1.1438383889908437, 'grad_norm_pre_clip_avg': + 0.27850248068571093, 'learning_rate': + 2.481028681573995e-05, 'epoch': 1.89} +04/19 [14:00:46] INFO | >> train_qwenlatent.py:487 + Step 7510 | grad_norm_pre_clip=0.2332 | + grad_norm_pre_clip_avg=0.2586 | Metrics: + {'align_loss': 0.023792393505573273, + 'recon_loss': 0.03159347549080849, + 'predict_loss': 0.025564080104231834, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23315994441509247, + 'data_time': 0.000653749011689797, + 'model_time': 1.1362059930106625, + 'grad_norm_pre_clip_avg': 0.2586387977004051, + 'learning_rate': 2.4808769968113004e-05, + 'epoch': 1.9} +04/19 [14:00:58] INFO | >> train_qwenlatent.py:487 + Step 7520 | grad_norm_pre_clip=0.3723 | + grad_norm_pre_clip_avg=0.3714 | Metrics: + {'align_loss': 0.023389514535665512, + 'recon_loss': 0.025599636137485504, + 'predict_loss': 0.022679032757878304, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.37228694558143616, + 'data_time': 0.0005923959834035486, + 'model_time': 1.1454668630030937, + 'grad_norm_pre_clip_avg': 0.3713911384344101, + 'learning_rate': 2.480724712743345e-05, + 'epoch': 1.9} +04/19 [14:01:10] INFO | >> train_qwenlatent.py:487 + Step 7530 | grad_norm_pre_clip=0.2917 | + grad_norm_pre_clip_avg=0.2932 | Metrics: + {'align_loss': 0.023439407348632812, + 'recon_loss': 0.02632593736052513, + 'predict_loss': 0.02058946155011654, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2917201817035675, + 'data_time': 0.0006500000017695129, + 'model_time': 1.1659761260088999, + 'grad_norm_pre_clip_avg': 0.2931613102555275, + 'learning_rate': 2.480571829444351e-05, + 'epoch': 1.9} +04/19 [14:01:21] INFO | >> train_qwenlatent.py:487 + Step 7540 | grad_norm_pre_clip=0.2420 | + grad_norm_pre_clip_avg=0.2475 | Metrics: + {'align_loss': 0.02302912250161171, + 'recon_loss': 0.024424178525805473, + 'predict_loss': 0.020626448094844818, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24195818603038788, + 'data_time': 0.0005986019968986511, + 'model_time': 1.140628006978659, + 'grad_norm_pre_clip_avg': 0.24748926162719725, + 'learning_rate': 2.480418346988831e-05, + 'epoch': 1.9} +04/19 [14:01:34] INFO | >> train_qwenlatent.py:487 + Step 7550 | grad_norm_pre_clip=0.2552 | + grad_norm_pre_clip_avg=0.2394 | Metrics: + {'align_loss': 0.023888815194368362, + 'recon_loss': 0.01988825760781765, + 'predict_loss': 0.017360394820570946, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2551760971546173, + 'mae_score': 0.026708965473346884, 'data_time': + 0.0005775010213255882, 'model_time': + 1.1391260929813143, 'grad_norm_pre_clip_avg': + 0.23936505168676375, 'learning_rate': + 2.4802642654515908e-05, 'epoch': 1.91} +04/19 [14:01:45] INFO | >> train_qwenlatent.py:487 + Step 7560 | grad_norm_pre_clip=0.2470 | + grad_norm_pre_clip_avg=0.2485 | Metrics: + {'align_loss': 0.02320781722664833, + 'recon_loss': 0.019085587933659554, + 'predict_loss': 0.02019929327070713, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2469566911458969, + 'data_time': 0.000625607994152233, + 'model_time': 1.1744892410060856, + 'grad_norm_pre_clip_avg': 0.2485220029950142, + 'learning_rate': 2.4801095849077283e-05, + 'epoch': 1.91} +04/19 [14:01:57] INFO | >> train_qwenlatent.py:487 + Step 7570 | grad_norm_pre_clip=0.2063 | + grad_norm_pre_clip_avg=0.2691 | Metrics: + {'align_loss': 0.024676669389009476, + 'recon_loss': 0.025585683062672615, + 'predict_loss': 0.016716595739126205, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20629701018333435, + 'data_time': 0.0005928619939368218, + 'model_time': 1.1787997309875209, + 'grad_norm_pre_clip_avg': 0.26909323036670685, + 'learning_rate': 2.4799543054326318e-05, + 'epoch': 1.91} +04/19 [14:02:09] INFO | >> train_qwenlatent.py:487 + Step 7580 | grad_norm_pre_clip=0.2849 | + grad_norm_pre_clip_avg=0.2662 | Metrics: + {'align_loss': 0.02315732091665268, + 'recon_loss': 0.02053580991923809, + 'predict_loss': 0.019007224589586258, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2848884165287018, + 'data_time': 0.0005868969892617315, + 'model_time': 1.158073383005103, + 'grad_norm_pre_clip_avg': 0.26619713604450224, + 'learning_rate': 2.4797984271019835e-05, + 'epoch': 1.91} +04/19 [14:02:20] INFO | >> train_qwenlatent.py:487 + Step 7590 | grad_norm_pre_clip=0.3839 | + grad_norm_pre_clip_avg=0.3072 | Metrics: + {'align_loss': 0.024005450308322906, + 'recon_loss': 0.01815655082464218, + 'predict_loss': 0.013870458118617535, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.38386693596839905, + 'data_time': 0.000618655001744628, + 'model_time': 1.1345204889948945, + 'grad_norm_pre_clip_avg': 0.30724318474531176, + 'learning_rate': 2.4796419499917565e-05, + 'epoch': 1.92} +04/19 [14:02:32] INFO | >> train_qwenlatent.py:487 + Step 7600 | grad_norm_pre_clip=0.3674 | + grad_norm_pre_clip_avg=0.3260 | Metrics: + {'align_loss': 0.02443791553378105, + 'recon_loss': 0.023085834458470345, + 'predict_loss': 0.020975206047296524, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.36736002564430237, + 'mae_score': 0.03396508500382707, 'data_time': + 0.0006209489947650582, 'model_time': + 1.156897047010716, 'grad_norm_pre_clip_avg': + 0.3260373041033745, 'learning_rate': + 2.4794848741782158e-05, 'epoch': 1.92} +04/19 [14:02:44] INFO | >> train_qwenlatent.py:487 + Step 7610 | grad_norm_pre_clip=0.2303 | + grad_norm_pre_clip_avg=0.2888 | Metrics: + {'align_loss': 0.024388931691646576, + 'recon_loss': 0.017184961587190628, + 'predict_loss': 0.011643234640359879, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23025952279567719, + 'data_time': 0.0005817089986521751, + 'model_time': 1.250205897988053, + 'grad_norm_pre_clip_avg': 0.2888369753956795, + 'learning_rate': 2.479327199737918e-05, + 'epoch': 1.92} +04/19 [14:02:56] INFO | >> train_qwenlatent.py:487 + Step 7620 | grad_norm_pre_clip=0.2485 | + grad_norm_pre_clip_avg=0.2893 | Metrics: + {'align_loss': 0.024389095604419708, + 'recon_loss': 0.022723903879523277, + 'predict_loss': 0.016772812232375145, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2484799027442932, + 'data_time': 0.000530854013049975, + 'model_time': 1.1474323319853283, + 'grad_norm_pre_clip_avg': 0.2893290430307388, + 'learning_rate': 2.479168926747712e-05, + 'epoch': 1.92} +04/19 [14:03:07] INFO | >> train_qwenlatent.py:487 + Step 7630 | grad_norm_pre_clip=0.2258 | + grad_norm_pre_clip_avg=0.2447 | Metrics: + {'align_loss': 0.023573266342282295, + 'recon_loss': 0.013836421072483063, + 'predict_loss': 0.01174352876842022, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2257520854473114, + 'data_time': 0.0005913939967285842, + 'model_time': 1.1387744990061037, + 'grad_norm_pre_clip_avg': 0.24470148831605912, + 'learning_rate': 2.4790100552847378e-05, + 'epoch': 1.93} +04/19 [14:03:19] INFO | >> train_qwenlatent.py:487 + Step 7640 | grad_norm_pre_clip=0.4326 | + grad_norm_pre_clip_avg=0.3060 | Metrics: + {'align_loss': 0.02547582797706127, + 'recon_loss': 0.02199471741914749, + 'predict_loss': 0.01860361546278, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.43260321021080017, + 'data_time': 0.0006544979987666011, + 'model_time': 1.1995540719944984, + 'grad_norm_pre_clip_avg': 0.30604291409254075, + 'learning_rate': 2.4788505854264283e-05, + 'epoch': 1.93} +04/19 [14:03:31] INFO | >> train_qwenlatent.py:487 + Step 7650 | grad_norm_pre_clip=0.2456 | + grad_norm_pre_clip_avg=0.3029 | Metrics: + {'align_loss': 0.023721124976873398, + 'recon_loss': 0.029944617301225662, + 'predict_loss': 0.023956628516316414, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24557942152023315, + 'mae_score': 0.028946654861037795, 'data_time': + 0.0005725130031351, 'model_time': + 1.1849910840101074, 'grad_norm_pre_clip_avg': + 0.30292650908231733, 'learning_rate': + 2.4786905172505058e-05, 'epoch': 1.93} +04/19 [14:03:43] INFO | >> train_qwenlatent.py:487 + Step 7660 | grad_norm_pre_clip=0.2669 | + grad_norm_pre_clip_avg=0.2677 | Metrics: + {'align_loss': 0.023373618721961975, + 'recon_loss': 0.027110520750284195, + 'predict_loss': 0.02464209869503975, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2668817341327667, + 'data_time': 0.0006119599856901914, + 'model_time': 1.3455224659992382, + 'grad_norm_pre_clip_avg': 0.26769117563962935, + 'learning_rate': 2.478529850834987e-05, + 'epoch': 1.93} +04/19 [14:03:54] INFO | >> train_qwenlatent.py:487 + Step 7670 | grad_norm_pre_clip=0.2389 | + grad_norm_pre_clip_avg=0.2555 | Metrics: + {'align_loss': 0.024979423731565475, + 'recon_loss': 0.02390693500638008, + 'predict_loss': 0.016881825402379036, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2389155477285385, + 'data_time': 0.0005570350040215999, + 'model_time': 1.1433283619990107, + 'grad_norm_pre_clip_avg': 0.25545658767223356, + 'learning_rate': 2.4783685862581775e-05, + 'epoch': 1.94} +04/19 [14:04:06] INFO | >> train_qwenlatent.py:487 + Step 7680 | grad_norm_pre_clip=0.2445 | + grad_norm_pre_clip_avg=0.2570 | Metrics: + {'align_loss': 0.023119598627090454, + 'recon_loss': 0.01985803060233593, + 'predict_loss': 0.017568159848451614, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24453210830688477, + 'data_time': 0.000571915996260941, + 'model_time': 1.1427841300028376, + 'grad_norm_pre_clip_avg': 0.2570415139198303, + 'learning_rate': 2.478206723598676e-05, + 'epoch': 1.94} +04/19 [14:04:18] INFO | >> train_qwenlatent.py:487 + Step 7690 | grad_norm_pre_clip=0.2651 | + grad_norm_pre_clip_avg=0.2749 | Metrics: + {'align_loss': 0.02422185242176056, + 'recon_loss': 0.023661429062485695, + 'predict_loss': 0.019632406532764435, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26507768034935, + 'data_time': 0.0005522549909073859, + 'model_time': 1.1558315259753726, + 'grad_norm_pre_clip_avg': 0.27494093030691147, + 'learning_rate': 2.4780442629353736e-05, + 'epoch': 1.94} +04/19 [14:04:30] INFO | >> train_qwenlatent.py:487 + Step 7700 | grad_norm_pre_clip=0.3440 | + grad_norm_pre_clip_avg=0.2812 | Metrics: + {'align_loss': 0.023167865350842476, + 'recon_loss': 0.020453527569770813, + 'predict_loss': 0.017987294122576714, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3439610004425049, + 'mae_score': 0.046744221180408926, 'data_time': + 0.0006622390064876527, 'model_time': + 1.141934058017796, 'grad_norm_pre_clip_avg': + 0.2811801269650459, 'learning_rate': + 2.47788120434745e-05, 'epoch': 1.94} +04/19 [14:04:41] INFO | >> train_qwenlatent.py:487 + Step 7710 | grad_norm_pre_clip=0.2625 | + grad_norm_pre_clip_avg=0.2559 | Metrics: + {'align_loss': 0.024798434227705002, + 'recon_loss': 0.02044655755162239, + 'predict_loss': 0.017410676926374435, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2625146508216858, + 'data_time': 0.0005438090011011809, + 'model_time': 1.1478650200006086, + 'grad_norm_pre_clip_avg': 0.2558903947472572, + 'learning_rate': 2.477717547914379e-05, + 'epoch': 1.95} +04/19 [14:04:53] INFO | >> train_qwenlatent.py:487 + Step 7720 | grad_norm_pre_clip=0.3202 | + grad_norm_pre_clip_avg=0.3148 | Metrics: + {'align_loss': 0.023057280108332634, + 'recon_loss': 0.024647848680615425, + 'predict_loss': 0.02301391027867794, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3201621174812317, + 'data_time': 0.0005992019723635167, + 'model_time': 1.1395594700006768, + 'grad_norm_pre_clip_avg': 0.3147533267736435, + 'learning_rate': 2.4775532937159246e-05, + 'epoch': 1.95} +04/19 [14:05:04] INFO | >> train_qwenlatent.py:487 + Step 7730 | grad_norm_pre_clip=0.2591 | + grad_norm_pre_clip_avg=0.2840 | Metrics: + {'align_loss': 0.02465743198990822, + 'recon_loss': 0.027263477444648743, + 'predict_loss': 0.020692629739642143, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2590716481208801, + 'data_time': 0.0005542850121855736, + 'model_time': 1.139809914981015, + 'grad_norm_pre_clip_avg': 0.28395030200481414, + 'learning_rate': 2.4773884418321417e-05, + 'epoch': 1.95} +04/19 [14:05:16] INFO | >> train_qwenlatent.py:487 + Step 7740 | grad_norm_pre_clip=0.2721 | + grad_norm_pre_clip_avg=0.2795 | Metrics: + {'align_loss': 0.024277718737721443, + 'recon_loss': 0.024818917736411095, + 'predict_loss': 0.024025393649935722, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.27213960886001587, + 'data_time': 0.0005615729896817356, + 'model_time': 1.1447045429958962, + 'grad_norm_pre_clip_avg': 0.2794850319623947, + 'learning_rate': 2.4772229923433777e-05, + 'epoch': 1.95} +04/19 [14:05:28] INFO | >> train_qwenlatent.py:487 + Step 7750 | grad_norm_pre_clip=0.2559 | + grad_norm_pre_clip_avg=0.2779 | Metrics: + {'align_loss': 0.02389933541417122, + 'recon_loss': 0.027001867070794106, + 'predict_loss': 0.01979978382587433, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2558946907520294, + 'mae_score': 0.02249398961797491, 'data_time': + 0.0005674059793818742, 'model_time': + 1.1362886119750328, 'grad_norm_pre_clip_avg': + 0.2779099091887474, 'learning_rate': + 2.4770569453302704e-05, 'epoch': 1.96} +04/19 [14:05:40] INFO | >> train_qwenlatent.py:487 + Step 7760 | grad_norm_pre_clip=0.2589 | + grad_norm_pre_clip_avg=0.2765 | Metrics: + {'align_loss': 0.022289689630270004, + 'recon_loss': 0.022589871659874916, + 'predict_loss': 0.01911797560751438, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25892558693885803, + 'data_time': 0.0006134949799161404, + 'model_time': 1.1580422020051628, + 'grad_norm_pre_clip_avg': 0.2764868393540382, + 'learning_rate': 2.4768903008737495e-05, + 'epoch': 1.96} +04/19 [14:05:51] INFO | >> train_qwenlatent.py:487 + Step 7770 | grad_norm_pre_clip=0.2977 | + grad_norm_pre_clip_avg=0.2574 | Metrics: + {'align_loss': 0.02372373826801777, + 'recon_loss': 0.025054767727851868, + 'predict_loss': 0.02012220211327076, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2976936995983124, + 'data_time': 0.0007601809920743108, + 'model_time': 1.1506025049893651, + 'grad_norm_pre_clip_avg': 0.25743542462587354, + 'learning_rate': 2.476723059055035e-05, + 'epoch': 1.96} +04/19 [14:06:42] INFO | >> train_qwenlatent.py:487 + Step 7780 | grad_norm_pre_clip=0.3392 | + grad_norm_pre_clip_avg=0.3028 | Metrics: + {'align_loss': 0.02329736202955246, + 'recon_loss': 0.023095985874533653, + 'predict_loss': 0.014324551448225975, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.33917272090911865, + 'data_time': 0.003502929990645498, + 'model_time': 3.2914299900003243, + 'grad_norm_pre_clip_avg': 0.3028172954916954, + 'learning_rate': 2.476555219955639e-05, + 'epoch': 1.96} +04/19 [14:07:17] INFO | >> train_qwenlatent.py:487 + Step 7790 | grad_norm_pre_clip=0.2224 | + grad_norm_pre_clip_avg=0.2464 | Metrics: + {'align_loss': 0.023886041715741158, + 'recon_loss': 0.03082382306456566, + 'predict_loss': 0.023967258632183075, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22240787744522095, + 'data_time': 0.0010907610121648759, + 'model_time': 3.3464389229775406, + 'grad_norm_pre_clip_avg': 0.24637341052293776, + 'learning_rate': 2.4763867836573634e-05, + 'epoch': 1.97} +04/19 [14:07:54] INFO | >> train_qwenlatent.py:487 + Step 7800 | grad_norm_pre_clip=0.2415 | + grad_norm_pre_clip_avg=0.2714 | Metrics: + {'align_loss': 0.024936091154813766, + 'recon_loss': 0.022606775164604187, + 'predict_loss': 0.018764864653348923, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24146224558353424, + 'mae_score': 0.022008244626156918, 'data_time': + 0.001146436989074573, 'model_time': + 3.872783807979431, 'grad_norm_pre_clip_avg': + 0.27136514335870743, 'learning_rate': + 2.4762177502423033e-05, 'epoch': 1.97} +04/19 [14:08:29] INFO | >> train_qwenlatent.py:487 + Step 7810 | grad_norm_pre_clip=0.2997 | + grad_norm_pre_clip_avg=0.2942 | Metrics: + {'align_loss': 0.02392146922647953, + 'recon_loss': 0.02358205057680607, + 'predict_loss': 0.0156598761677742, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.29968148469924927, + 'data_time': 0.001047393976477906, + 'model_time': 3.6389356940053403, + 'grad_norm_pre_clip_avg': 0.2941871240735054, + 'learning_rate': 2.4760481197928424e-05, + 'epoch': 1.97} +04/19 [14:09:04] INFO | >> train_qwenlatent.py:487 + Step 7820 | grad_norm_pre_clip=0.2525 | + grad_norm_pre_clip_avg=0.2591 | Metrics: + {'align_loss': 0.023154351860284805, + 'recon_loss': 0.025765156373381615, + 'predict_loss': 0.016641544178128242, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2524679899215698, + 'data_time': 0.010657880979124457, + 'model_time': 3.2810639899980742, + 'grad_norm_pre_clip_avg': 0.25914839655160904, + 'learning_rate': 2.4758778923916567e-05, + 'epoch': 1.97} +04/19 [14:09:39] INFO | >> train_qwenlatent.py:487 + Step 7830 | grad_norm_pre_clip=0.3356 | + grad_norm_pre_clip_avg=0.2597 | Metrics: + {'align_loss': 0.02523227408528328, + 'recon_loss': 0.03150155767798424, + 'predict_loss': 0.019263872876763344, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3356461822986603, + 'data_time': 0.001228802982950583, + 'model_time': 3.570867203990929, + 'grad_norm_pre_clip_avg': 0.2597451686859131, + 'learning_rate': 2.4757070681217136e-05, + 'epoch': 1.98} +04/19 [14:10:10] INFO | >> train_qwenlatent.py:487 + Step 7840 | grad_norm_pre_clip=0.2499 | + grad_norm_pre_clip_avg=0.2699 | Metrics: + {'align_loss': 0.023111985996365547, + 'recon_loss': 0.020941395312547684, + 'predict_loss': 0.020022837445139885, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2499459683895111, + 'data_time': 0.001187697984278202, + 'model_time': 2.5567053199920338, + 'grad_norm_pre_clip_avg': 0.26991505175828934, + 'learning_rate': 2.4755356470662703e-05, + 'epoch': 1.98} +04/19 [14:10:36] INFO | >> train_qwenlatent.py:487 + Step 7850 | grad_norm_pre_clip=0.3074 | + grad_norm_pre_clip_avg=0.2642 | Metrics: + {'align_loss': 0.02296176180243492, + 'recon_loss': 0.018254661932587624, + 'predict_loss': 0.014085933566093445, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.307416170835495, + 'mae_score': 0.022724285641232054, 'data_time': + 0.0013671570050064474, 'model_time': + 2.7192754289717413, 'grad_norm_pre_clip_avg': + 0.264203305542469, 'learning_rate': + 2.4753636293088753e-05, 'epoch': 1.98} +04/19 [14:10:56] INFO | >> train_qwenlatent.py:487 + Step 7860 | grad_norm_pre_clip=0.2341 | + grad_norm_pre_clip_avg=0.2862 | Metrics: + {'align_loss': 0.023911580443382263, + 'recon_loss': 0.02202082797884941, + 'predict_loss': 0.016161246225237846, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23409044742584229, + 'data_time': 0.0012084839981980622, + 'model_time': 1.480582054995466, + 'grad_norm_pre_clip_avg': 0.2861794143915176, + 'learning_rate': 2.475191014933368e-05, + 'epoch': 1.98} +04/19 [14:11:10] INFO | >> train_qwenlatent.py:487 + Step 7870 | grad_norm_pre_clip=0.3888 | + grad_norm_pre_clip_avg=0.2995 | Metrics: + {'align_loss': 0.02504354901611805, + 'recon_loss': 0.025823701173067093, + 'predict_loss': 0.020386191084980965, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.38881030678749084, + 'data_time': 0.000867811992065981, + 'model_time': 1.2003573850088287, + 'grad_norm_pre_clip_avg': 0.2995480537414551, + 'learning_rate': 2.475017804023879e-05, + 'epoch': 1.99} +04/19 [14:11:23] INFO | >> train_qwenlatent.py:487 + Step 7880 | grad_norm_pre_clip=0.2088 | + grad_norm_pre_clip_avg=0.2490 | Metrics: + {'align_loss': 0.02423565462231636, + 'recon_loss': 0.025877946987748146, + 'predict_loss': 0.01967872865498066, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20882654190063477, + 'data_time': 0.0005938519898336381, + 'model_time': 1.2950749169976916, + 'grad_norm_pre_clip_avg': 0.24903372079133987, + 'learning_rate': 2.4748439966648286e-05, + 'epoch': 1.99} +04/19 [14:11:35] INFO | >> train_qwenlatent.py:487 + Step 7890 | grad_norm_pre_clip=0.2504 | + grad_norm_pre_clip_avg=0.2797 | Metrics: + {'align_loss': 0.023547453805804253, + 'recon_loss': 0.017131637781858444, + 'predict_loss': 0.016448112204670906, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2503906488418579, + 'data_time': 0.0005890140018891543, + 'model_time': 1.2389347830030601, + 'grad_norm_pre_clip_avg': 0.2796850875020027, + 'learning_rate': 2.4746695929409284e-05, + 'epoch': 1.99} +04/19 [14:11:49] INFO | >> train_qwenlatent.py:487 + Step 7900 | grad_norm_pre_clip=0.3017 | + grad_norm_pre_clip_avg=0.2926 | Metrics: + {'align_loss': 0.025466132909059525, + 'recon_loss': 0.024388020858168602, + 'predict_loss': 0.01594657264649868, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.30168017745018005, + 'mae_score': 0.021154487884796416, 'data_time': + 0.0007269520137924701, 'model_time': + 1.2697252240031958, 'grad_norm_pre_clip_avg': + 0.29263308495283125, 'learning_rate': + 2.474494592937181e-05, 'epoch': 1.99} +04/19 [14:12:01] INFO | >> train_qwenlatent.py:487 + Step 7910 | grad_norm_pre_clip=0.3161 | + grad_norm_pre_clip_avg=0.2533 | Metrics: + {'align_loss': 0.024159010499715805, + 'recon_loss': 0.022581905126571655, + 'predict_loss': 0.01986248977482319, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3160695731639862, + 'data_time': 0.0005985380266793072, + 'model_time': 1.2464859989995603, + 'grad_norm_pre_clip_avg': 0.25326511561870574, + 'learning_rate': 2.474318996738879e-05, + 'epoch': 2.0} +04/19 [14:12:14] INFO | >> train_qwenlatent.py:487 + Step 7920 | grad_norm_pre_clip=0.2322 | + grad_norm_pre_clip_avg=0.2795 | Metrics: + {'align_loss': 0.023598961532115936, + 'recon_loss': 0.021223340183496475, + 'predict_loss': 0.016040390357375145, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23217207193374634, + 'data_time': 0.0007819989987183362, + 'model_time': 1.2282667210092768, + 'grad_norm_pre_clip_avg': 0.27948416471481324, + 'learning_rate': 2.4741428044316066e-05, + 'epoch': 2.0} +04/19 [14:12:27] INFO | >> train_qwenlatent.py:487 + Step 7930 | grad_norm_pre_clip=0.2863 | + grad_norm_pre_clip_avg=0.2753 | Metrics: + {'align_loss': 0.024251163005828857, + 'recon_loss': 0.020363569259643555, + 'predict_loss': 0.015793895348906517, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28630530834198, + 'data_time': 0.0009175029990728945, + 'model_time': 1.2306577900017146, + 'grad_norm_pre_clip_avg': 0.2752735659480095, + 'learning_rate': 2.4739660161012368e-05, + 'epoch': 2.0} +04/19 [14:12:39] INFO | >> train_qwenlatent.py:487 + Step 7940 | grad_norm_pre_clip=0.2383 | + grad_norm_pre_clip_avg=0.2584 | Metrics: + {'align_loss': 0.02428077533841133, + 'recon_loss': 0.02373519167304039, + 'predict_loss': 0.016798177734017372, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23831261694431305, + 'data_time': 0.0009397090179845691, + 'model_time': 1.2445356319949497, + 'grad_norm_pre_clip_avg': 0.25836357921361924, + 'learning_rate': 2.4737886318339343e-05, + 'epoch': 2.0} +04/19 [14:12:52] INFO | >> train_qwenlatent.py:487 + Step 7950 | grad_norm_pre_clip=0.2104 | + grad_norm_pre_clip_avg=0.2898 | Metrics: + {'align_loss': 0.0254240483045578, + 'recon_loss': 0.028721462935209274, + 'predict_loss': 0.019765958189964294, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21041575074195862, + 'mae_score': 0.022919811214412656, 'data_time': + 0.0008206499915104359, 'model_time': + 1.2790023739798926, 'grad_norm_pre_clip_avg': + 0.28984857350587845, 'learning_rate': + 2.473610651716154e-05, 'epoch': 2.01} +04/19 [14:13:05] INFO | >> train_qwenlatent.py:487 + Step 7960 | grad_norm_pre_clip=0.2140 | + grad_norm_pre_clip_avg=0.2786 | Metrics: + {'align_loss': 0.023789338767528534, + 'recon_loss': 0.02594205178320408, + 'predict_loss': 0.014899042434990406, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2140047550201416, + 'data_time': 0.001023814984364435, + 'model_time': 1.2319458929996472, + 'grad_norm_pre_clip_avg': 0.27863913476467134, + 'learning_rate': 2.473432075834642e-05, + 'epoch': 2.01} +04/19 [14:13:17] INFO | >> train_qwenlatent.py:487 + Step 7970 | grad_norm_pre_clip=0.4096 | + grad_norm_pre_clip_avg=0.2976 | Metrics: + {'align_loss': 0.0238528773188591, + 'recon_loss': 0.027818910777568817, + 'predict_loss': 0.01912875473499298, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.4095537066459656, + 'data_time': 0.000711064989445731, + 'model_time': 1.227885926986346, + 'grad_norm_pre_clip_avg': 0.29756580740213395, + 'learning_rate': 2.4732529042764332e-05, + 'epoch': 2.01} +04/19 [14:13:31] INFO | >> train_qwenlatent.py:487 + Step 7980 | grad_norm_pre_clip=0.2875 | + grad_norm_pre_clip_avg=0.2927 | Metrics: + {'align_loss': 0.024066410958766937, + 'recon_loss': 0.021108007058501244, + 'predict_loss': 0.019264839589595795, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28745436668395996, + 'data_time': 0.0010659850086085498, + 'model_time': 1.5057403699902352, + 'grad_norm_pre_clip_avg': 0.29273331761360166, + 'learning_rate': 2.4730731371288542e-05, + 'epoch': 2.01} +04/19 [14:13:43] INFO | >> train_qwenlatent.py:487 + Step 7990 | grad_norm_pre_clip=0.2848 | + grad_norm_pre_clip_avg=0.2897 | Metrics: + {'align_loss': 0.02322840318083763, + 'recon_loss': 0.026241643354296684, + 'predict_loss': 0.020069604739546776, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28476300835609436, + 'data_time': 0.0009461799927521497, + 'model_time': 1.2538944410043769, + 'grad_norm_pre_clip_avg': 0.2897077575325966, + 'learning_rate': 2.4728927744795207e-05, + 'epoch': 2.02} +04/19 [14:13:56] INFO | >> train_qwenlatent.py:487 + Step 8000 | grad_norm_pre_clip=0.2976 | + grad_norm_pre_clip_avg=0.3001 | Metrics: + {'align_loss': 0.024606820195913315, + 'recon_loss': 0.026306141167879105, + 'predict_loss': 0.018540991470217705, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2976256012916565, + 'mae_score': 0.021634074374362156, 'data_time': + 0.0005560520221479237, 'model_time': + 1.2136304070008919, 'grad_norm_pre_clip_avg': + 0.3000650778412819, 'learning_rate': + 2.47271181641634e-05, 'epoch': 2.02} +04/19 [14:14:09] INFO | >> train_qwenlatent.py:487 + Step 8010 | grad_norm_pre_clip=0.2866 | + grad_norm_pre_clip_avg=0.2803 | Metrics: + {'align_loss': 0.02384798228740692, + 'recon_loss': 0.02266846038401127, + 'predict_loss': 0.015137906186282635, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28663286566734314, + 'data_time': 0.000680533004924655, + 'model_time': 1.2297214580175932, + 'grad_norm_pre_clip_avg': 0.28026767522096635, + 'learning_rate': 2.4725302630275082e-05, + 'epoch': 2.02} +04/19 [14:14:22] INFO | >> train_qwenlatent.py:487 + Step 8020 | grad_norm_pre_clip=0.3293 | + grad_norm_pre_clip_avg=0.2672 | Metrics: + {'align_loss': 0.025052329525351524, + 'recon_loss': 0.032689500600099564, + 'predict_loss': 0.0241752490401268, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3292560577392578, + 'data_time': 0.0007023530197329819, + 'model_time': 1.2243688149901573, + 'grad_norm_pre_clip_avg': 0.2671664714813232, + 'learning_rate': 2.4723481144015123e-05, + 'epoch': 2.02} +04/19 [14:14:35] INFO | >> train_qwenlatent.py:487 + Step 8030 | grad_norm_pre_clip=0.2747 | + grad_norm_pre_clip_avg=0.2906 | Metrics: + {'align_loss': 0.023198779672384262, + 'recon_loss': 0.022668465971946716, + 'predict_loss': 0.017245180904865265, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.27474701404571533, + 'data_time': 0.0006203179946169257, + 'model_time': 1.2484527380147483, + 'grad_norm_pre_clip_avg': 0.2905888006091118, + 'learning_rate': 2.4721653706271297e-05, + 'epoch': 2.03} +04/19 [14:14:47] INFO | >> train_qwenlatent.py:487 + Step 8040 | grad_norm_pre_clip=0.2604 | + grad_norm_pre_clip_avg=0.2775 | Metrics: + {'align_loss': 0.025058239698410034, + 'recon_loss': 0.02086452580988407, + 'predict_loss': 0.014636397361755371, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2603769302368164, + 'data_time': 0.0006611860007978976, + 'model_time': 1.2382779550098348, + 'grad_norm_pre_clip_avg': 0.27752653062343596, + 'learning_rate': 2.4719820317934276e-05, + 'epoch': 2.03} +04/19 [14:15:00] INFO | >> train_qwenlatent.py:487 + Step 8050 | grad_norm_pre_clip=0.2730 | + grad_norm_pre_clip_avg=0.2212 | Metrics: + {'align_loss': 0.02510102465748787, + 'recon_loss': 0.03476951643824577, + 'predict_loss': 0.0206773579120636, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.27303406596183777, + 'mae_score': 0.02376832016953477, 'data_time': + 0.0010344890179112554, 'model_time': + 1.5415954499912914, 'grad_norm_pre_clip_avg': + 0.22120973616838455, 'learning_rate': + 2.4717980979897625e-05, 'epoch': 2.03} +04/19 [14:15:13] INFO | >> train_qwenlatent.py:487 + Step 8060 | grad_norm_pre_clip=0.3439 | + grad_norm_pre_clip_avg=0.3127 | Metrics: + {'align_loss': 0.02470843866467476, + 'recon_loss': 0.025911716744303703, + 'predict_loss': 0.02475554682314396, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.34394440054893494, + 'data_time': 0.0009602149948477745, + 'model_time': 1.226351409975905, + 'grad_norm_pre_clip_avg': 0.31274581551551817, + 'learning_rate': 2.471613569305782e-05, + 'epoch': 2.03} +04/19 [14:15:26] INFO | >> train_qwenlatent.py:487 + Step 8070 | grad_norm_pre_clip=0.2247 | + grad_norm_pre_clip_avg=0.2923 | Metrics: + {'align_loss': 0.02447197213768959, + 'recon_loss': 0.031871210783720016, + 'predict_loss': 0.020623257383704185, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22469410300254822, + 'data_time': 0.0007571419992018491, + 'model_time': 1.204373949993169, + 'grad_norm_pre_clip_avg': 0.2922913134098053, + 'learning_rate': 2.4714284458314228e-05, + 'epoch': 2.04} +04/19 [14:15:38] INFO | >> train_qwenlatent.py:487 + Step 8080 | grad_norm_pre_clip=0.3200 | + grad_norm_pre_clip_avg=0.2708 | Metrics: + {'align_loss': 0.02415875717997551, + 'recon_loss': 0.033579420298337936, + 'predict_loss': 0.02855943888425827, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3200327754020691, + 'data_time': 0.0006983759813010693, + 'model_time': 1.185542600986082, + 'grad_norm_pre_clip_avg': 0.27080715596675875, + 'learning_rate': 2.4712427276569117e-05, + 'epoch': 2.04} +04/19 [14:15:50] INFO | >> train_qwenlatent.py:487 + Step 8090 | grad_norm_pre_clip=0.2260 | + grad_norm_pre_clip_avg=0.2647 | Metrics: + {'align_loss': 0.02351384051144123, + 'recon_loss': 0.02230064570903778, + 'predict_loss': 0.016040362417697906, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22596746683120728, + 'data_time': 0.0006492490065284073, + 'model_time': 1.2017990300082602, + 'grad_norm_pre_clip_avg': 0.2647262364625931, + 'learning_rate': 2.471056414872766e-05, + 'epoch': 2.04} +04/19 [14:16:04] INFO | >> train_qwenlatent.py:487 + Step 8100 | grad_norm_pre_clip=0.2021 | + grad_norm_pre_clip_avg=0.2324 | Metrics: + {'align_loss': 0.024695783853530884, + 'recon_loss': 0.025238491594791412, + 'predict_loss': 0.017206519842147827, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2021152228116989, + 'mae_score': 0.029522492004944396, 'data_time': + 0.0009271790040656924, 'model_time': + 1.300324757990893, 'grad_norm_pre_clip_avg': + 0.23237133771181107, 'learning_rate': + 2.4708695075697924e-05, 'epoch': 2.04} +04/19 [14:16:17] INFO | >> train_qwenlatent.py:487 + Step 8110 | grad_norm_pre_clip=0.2896 | + grad_norm_pre_clip_avg=0.2580 | Metrics: + {'align_loss': 0.025267604738473892, + 'recon_loss': 0.027873562648892403, + 'predict_loss': 0.01831102930009365, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2895834147930145, + 'data_time': 0.000669475004542619, + 'model_time': 1.2474737520096824, + 'grad_norm_pre_clip_avg': 0.25796844959259035, + 'learning_rate': 2.4706820058390867e-05, + 'epoch': 2.05} +04/19 [14:16:30] INFO | >> train_qwenlatent.py:487 + Step 8120 | grad_norm_pre_clip=0.2982 | + grad_norm_pre_clip_avg=0.2751 | Metrics: + {'align_loss': 0.023239150643348694, + 'recon_loss': 0.019106728956103325, + 'predict_loss': 0.013379541225731373, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.29820266366004944, + 'data_time': 0.0006319350213743746, + 'model_time': 1.2342944720003288, + 'grad_norm_pre_clip_avg': 0.2751235648989677, + 'learning_rate': 2.4704939097720353e-05, + 'epoch': 2.05} +04/19 [14:16:42] INFO | >> train_qwenlatent.py:487 + Step 8130 | grad_norm_pre_clip=0.2040 | + grad_norm_pre_clip_avg=0.2452 | Metrics: + {'align_loss': 0.023319069296121597, + 'recon_loss': 0.02663327567279339, + 'predict_loss': 0.016525331884622574, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20399008691310883, + 'data_time': 0.0006047989882063121, + 'model_time': 1.2227805969887413, + 'grad_norm_pre_clip_avg': 0.24521440267562866, + 'learning_rate': 2.4703052194603136e-05, + 'epoch': 2.05} +04/19 [14:16:54] INFO | >> train_qwenlatent.py:487 + Step 8140 | grad_norm_pre_clip=0.3145 | + grad_norm_pre_clip_avg=0.2785 | Metrics: + {'align_loss': 0.022953609004616737, + 'recon_loss': 0.02979269064962864, + 'predict_loss': 0.022532125934958458, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.31454795598983765, + 'data_time': 0.0006119000026956201, + 'model_time': 1.3063711909926496, + 'grad_norm_pre_clip_avg': 0.2785154029726982, + 'learning_rate': 2.4701159349958873e-05, + 'epoch': 2.05} +04/19 [14:17:08] INFO | >> train_qwenlatent.py:487 + Step 8150 | grad_norm_pre_clip=0.2687 | + grad_norm_pre_clip_avg=0.2703 | Metrics: + {'align_loss': 0.024532008916139603, + 'recon_loss': 0.0241986270993948, + 'predict_loss': 0.016778329387307167, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2686696946620941, + 'mae_score': 0.034580801199148366, 'data_time': + 0.0007051519933156669, 'model_time': + 1.1965867329854518, 'grad_norm_pre_clip_avg': + 0.27032348066568374, 'learning_rate': + 2.469926056471011e-05, 'epoch': 2.06} +04/19 [14:17:20] INFO | >> train_qwenlatent.py:487 + Step 8160 | grad_norm_pre_clip=0.2685 | + grad_norm_pre_clip_avg=0.2455 | Metrics: + {'align_loss': 0.023837123066186905, + 'recon_loss': 0.028329160064458847, + 'predict_loss': 0.017618456855416298, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2685028612613678, + 'data_time': 0.0006015269900672138, + 'model_time': 1.2169457040145062, + 'grad_norm_pre_clip_avg': 0.24550541937351228, + 'learning_rate': 2.4697355839782296e-05, + 'epoch': 2.06} +04/19 [14:17:33] INFO | >> train_qwenlatent.py:487 + Step 8170 | grad_norm_pre_clip=0.3918 | + grad_norm_pre_clip_avg=0.2528 | Metrics: + {'align_loss': 0.022922717034816742, + 'recon_loss': 0.025292595848441124, + 'predict_loss': 0.020587319508194923, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.39177238941192627, + 'data_time': 0.0008256850123871118, + 'model_time': 1.281647141004214, + 'grad_norm_pre_clip_avg': 0.2527909576892853, + 'learning_rate': 2.469544517610377e-05, + 'epoch': 2.06} +04/19 [14:17:46] INFO | >> train_qwenlatent.py:487 + Step 8180 | grad_norm_pre_clip=0.3604 | + grad_norm_pre_clip_avg=0.3569 | Metrics: + {'align_loss': 0.023820601403713226, + 'recon_loss': 0.029009316116571426, + 'predict_loss': 0.016863906756043434, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3603784739971161, + 'data_time': 0.0009519009909126908, + 'model_time': 1.2182811230013613, + 'grad_norm_pre_clip_avg': 0.3568704217672348, + 'learning_rate': 2.4693528574605765e-05, + 'epoch': 2.06} +04/19 [14:17:59] INFO | >> train_qwenlatent.py:487 + Step 8190 | grad_norm_pre_clip=0.2925 | + grad_norm_pre_clip_avg=0.3203 | Metrics: + {'align_loss': 0.02551335282623768, + 'recon_loss': 0.0334121510386467, + 'predict_loss': 0.020936252549290657, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.292534202337265, + 'data_time': 0.000880781008163467, + 'model_time': 1.1710836160054896, + 'grad_norm_pre_clip_avg': 0.3203200727701187, + 'learning_rate': 2.4691606036222407e-05, + 'epoch': 2.07} +04/19 [14:18:12] INFO | >> train_qwenlatent.py:487 + Step 8200 | grad_norm_pre_clip=0.2137 | + grad_norm_pre_clip_avg=0.2439 | Metrics: + {'align_loss': 0.021596625447273254, + 'recon_loss': 0.017323486506938934, + 'predict_loss': 0.011570139788091183, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21370944380760193, + 'mae_score': 0.03645021593248522, 'data_time': + 0.00062975098262541, 'model_time': + 1.2759211740049068, 'grad_norm_pre_clip_avg': + 0.24385152906179428, 'learning_rate': + 2.468967756189072e-05, 'epoch': 2.07} +04/19 [14:18:24] INFO | >> train_qwenlatent.py:487 + Step 8210 | grad_norm_pre_clip=0.2928 | + grad_norm_pre_clip_avg=0.2482 | Metrics: + {'align_loss': 0.025229226797819138, + 'recon_loss': 0.03554539754986763, + 'predict_loss': 0.022485878318548203, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.29283371567726135, + 'data_time': 0.0008556070097256452, + 'model_time': 1.2904501899902243, + 'grad_norm_pre_clip_avg': 0.24818189591169357, + 'learning_rate': 2.4687743152550625e-05, + 'epoch': 2.07} +04/19 [14:18:37] INFO | >> train_qwenlatent.py:487 + Step 8220 | grad_norm_pre_clip=0.3469 | + grad_norm_pre_clip_avg=0.2710 | Metrics: + {'align_loss': 0.023765388876199722, + 'recon_loss': 0.022214766591787338, + 'predict_loss': 0.014948840253055096, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3468656539916992, + 'data_time': 0.00092804100131616, 'model_time': + 1.2223295290023088, 'grad_norm_pre_clip_avg': + 0.27098891139030457, 'learning_rate': + 2.468580280914492e-05, 'epoch': 2.07} +04/19 [14:18:49] INFO | >> train_qwenlatent.py:487 + Step 8230 | grad_norm_pre_clip=0.1977 | + grad_norm_pre_clip_avg=0.2620 | Metrics: + {'align_loss': 0.024389097467064857, + 'recon_loss': 0.02773217298090458, + 'predict_loss': 0.018227573484182358, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1976718008518219, + 'data_time': 0.0008272000122815371, + 'model_time': 1.2160272950131912, + 'grad_norm_pre_clip_avg': 0.261968144774437, + 'learning_rate': 2.468385653261931e-05, + 'epoch': 2.08} +04/19 [14:19:02] INFO | >> train_qwenlatent.py:487 + Step 8240 | grad_norm_pre_clip=0.2568 | + grad_norm_pre_clip_avg=0.2609 | Metrics: + {'align_loss': 0.023474857211112976, + 'recon_loss': 0.032312121242284775, + 'predict_loss': 0.0195939838886261, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25679606199264526, + 'data_time': 0.0009617939940653741, + 'model_time': 1.253225364984246, + 'grad_norm_pre_clip_avg': 0.2608746632933617, + 'learning_rate': 2.4681904323922383e-05, + 'epoch': 2.08} +04/19 [14:19:15] INFO | >> train_qwenlatent.py:487 + Step 8250 | grad_norm_pre_clip=0.3088 | + grad_norm_pre_clip_avg=0.2529 | Metrics: + {'align_loss': 0.02556517906486988, + 'recon_loss': 0.021572552621364594, + 'predict_loss': 0.013886457309126854, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.30881351232528687, + 'mae_score': 0.0229369069004918, 'data_time': + 0.0007293050002772361, 'model_time': + 1.2311710040085018, 'grad_norm_pre_clip_avg': + 0.2528675302863121, 'learning_rate': + 2.467994618400563e-05, 'epoch': 2.08} +04/19 [14:19:28] INFO | >> train_qwenlatent.py:487 + Step 8260 | grad_norm_pre_clip=0.2421 | + grad_norm_pre_clip_avg=0.2869 | Metrics: + {'align_loss': 0.023839861154556274, + 'recon_loss': 0.026308400556445122, + 'predict_loss': 0.02075948379933834, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24210727214813232, + 'data_time': 0.0006931179959792644, + 'model_time': 1.2923872059909627, + 'grad_norm_pre_clip_avg': 0.28692379146814345, + 'learning_rate': 2.4677982113823407e-05, + 'epoch': 2.08} +04/19 [14:19:41] INFO | >> train_qwenlatent.py:487 + Step 8270 | grad_norm_pre_clip=0.2774 | + grad_norm_pre_clip_avg=0.2345 | Metrics: + {'align_loss': 0.022957205772399902, + 'recon_loss': 0.019844448193907738, + 'predict_loss': 0.012054644525051117, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.27735963463783264, + 'data_time': 0.0008995599928312004, + 'model_time': 1.234590585983824, + 'grad_norm_pre_clip_avg': 0.2345065474510193, + 'learning_rate': 2.4676012114333e-05, 'epoch': + 2.09} +04/19 [14:19:54] INFO | >> train_qwenlatent.py:487 + Step 8280 | grad_norm_pre_clip=0.3691 | + grad_norm_pre_clip_avg=0.3565 | Metrics: + {'align_loss': 0.023680495098233223, + 'recon_loss': 0.02395518682897091, + 'predict_loss': 0.017330514267086983, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.36911091208457947, + 'data_time': 0.0008546469907741994, + 'model_time': 1.2368024209863506, + 'grad_norm_pre_clip_avg': 0.35646113455295564, + 'learning_rate': 2.4674036186494547e-05, + 'epoch': 2.09} +04/19 [14:20:06] INFO | >> train_qwenlatent.py:487 + Step 8290 | grad_norm_pre_clip=0.2617 | + grad_norm_pre_clip_avg=0.3026 | Metrics: + {'align_loss': 0.023818911984562874, + 'recon_loss': 0.021532919257879257, + 'predict_loss': 0.01356798317283392, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26166078448295593, + 'data_time': 0.0009747970034368336, + 'model_time': 1.2186532110208645, + 'grad_norm_pre_clip_avg': 0.30259027183055875, + 'learning_rate': 2.46720543312711e-05, 'epoch': + 2.09} +04/19 [14:20:19] INFO | >> train_qwenlatent.py:487 + Step 8300 | grad_norm_pre_clip=0.2421 | + grad_norm_pre_clip_avg=0.2299 | Metrics: + {'align_loss': 0.02420687861740589, + 'recon_loss': 0.025013724341988564, + 'predict_loss': 0.016962552443146706, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2421034723520279, + 'mae_score': 0.027419200244250597, 'data_time': + 0.0008874770137481391, 'model_time': + 1.2617458029999398, 'grad_norm_pre_clip_avg': + 0.22993090450763704, 'learning_rate': + 2.467006654962859e-05, 'epoch': 2.09} +04/19 [14:20:32] INFO | >> train_qwenlatent.py:487 + Step 8310 | grad_norm_pre_clip=0.2391 | + grad_norm_pre_clip_avg=0.2464 | Metrics: + {'align_loss': 0.025059809908270836, + 'recon_loss': 0.0285613052546978, + 'predict_loss': 0.015607284381985664, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2390546202659607, + 'data_time': 0.0009867059998214245, + 'model_time': 1.2471945929864887, + 'grad_norm_pre_clip_avg': 0.2464255079627037, + 'learning_rate': 2.4668072842535834e-05, + 'epoch': 2.1} +04/19 [14:20:45] INFO | >> train_qwenlatent.py:487 + Step 8320 | grad_norm_pre_clip=0.4839 | + grad_norm_pre_clip_avg=0.3270 | Metrics: + {'align_loss': 0.022983118891716003, + 'recon_loss': 0.025047246366739273, + 'predict_loss': 0.015604441985487938, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.48391181230545044, + 'data_time': 0.0009236499899998307, + 'model_time': 1.2461919620109256, + 'grad_norm_pre_clip_avg': 0.32695306539535524, + 'learning_rate': 2.4666073210964537e-05, + 'epoch': 2.1} +04/19 [14:20:58] INFO | >> train_qwenlatent.py:487 + Step 8330 | grad_norm_pre_clip=0.3290 | + grad_norm_pre_clip_avg=0.3438 | Metrics: + {'align_loss': 0.02428426221013069, + 'recon_loss': 0.027719104662537575, + 'predict_loss': 0.02104303613305092, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3290238678455353, + 'data_time': 0.000656961026834324, + 'model_time': 1.2436650420131627, + 'grad_norm_pre_clip_avg': 0.3437845543026924, + 'learning_rate': 2.466406765588931e-05, + 'epoch': 2.1} +04/19 [14:21:11] INFO | >> train_qwenlatent.py:487 + Step 8340 | grad_norm_pre_clip=0.2722 | + grad_norm_pre_clip_avg=0.2741 | Metrics: + {'align_loss': 0.02415969967842102, + 'recon_loss': 0.024623874574899673, + 'predict_loss': 0.01567031256854534, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.272235631942749, + 'data_time': 0.0006711409951094538, + 'model_time': 1.2329853110131808, + 'grad_norm_pre_clip_avg': 0.2740586206316948, + 'learning_rate': 2.4662056178287617e-05, + 'epoch': 2.1} +04/19 [14:21:24] INFO | >> train_qwenlatent.py:487 + Step 8350 | grad_norm_pre_clip=0.2548 | + grad_norm_pre_clip_avg=0.2453 | Metrics: + {'align_loss': 0.023675667122006416, + 'recon_loss': 0.028959715738892555, + 'predict_loss': 0.017424356192350388, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25476545095443726, + 'mae_score': 0.023894256729263444, 'data_time': + 0.00096633899374865, 'model_time': + 1.2958788590040058, 'grad_norm_pre_clip_avg': + 0.24531747549772262, 'learning_rate': + 2.4660038779139843e-05, 'epoch': 2.11} +04/19 [14:21:37] INFO | >> train_qwenlatent.py:487 + Step 8360 | grad_norm_pre_clip=0.3115 | + grad_norm_pre_clip_avg=0.2667 | Metrics: + {'align_loss': 0.024649377912282944, + 'recon_loss': 0.03706883266568184, + 'predict_loss': 0.020972199738025665, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.31154730916023254, + 'data_time': 0.0008442630060017109, + 'model_time': 1.2509912060049828, + 'grad_norm_pre_clip_avg': 0.2667463138699532, + 'learning_rate': 2.4658015459429235e-05, + 'epoch': 2.11} +04/19 [14:21:49] INFO | >> train_qwenlatent.py:487 + Step 8370 | grad_norm_pre_clip=0.2354 | + grad_norm_pre_clip_avg=0.2892 | Metrics: + {'align_loss': 0.02355138212442398, + 'recon_loss': 0.023348284885287285, + 'predict_loss': 0.015184829011559486, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23541691899299622, + 'data_time': 0.000793300976511091, + 'model_time': 1.2281626529875211, + 'grad_norm_pre_clip_avg': 0.2892492339015007, + 'learning_rate': 2.465598622014194e-05, + 'epoch': 2.11} +04/19 [14:22:02] INFO | >> train_qwenlatent.py:487 + Step 8380 | grad_norm_pre_clip=0.3013 | + grad_norm_pre_clip_avg=0.2461 | Metrics: + {'align_loss': 0.022507237270474434, + 'recon_loss': 0.023872042074799538, + 'predict_loss': 0.019785432144999504, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.301311731338501, + 'data_time': 0.0008429099980276078, + 'model_time': 1.2975479619926773, + 'grad_norm_pre_clip_avg': 0.2460599958896637, + 'learning_rate': 2.465395106226698e-05, + 'epoch': 2.11} +04/19 [14:22:15] INFO | >> train_qwenlatent.py:487 + Step 8390 | grad_norm_pre_clip=0.2478 | + grad_norm_pre_clip_avg=0.2503 | Metrics: + {'align_loss': 0.023177340626716614, + 'recon_loss': 0.02988336607813835, + 'predict_loss': 0.01882445253431797, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24780304729938507, + 'data_time': 0.000943768973229453, + 'model_time': 1.2229389370186254, + 'grad_norm_pre_clip_avg': 0.2502658158540726, + 'learning_rate': 2.4651909986796265e-05, + 'epoch': 2.12} +04/19 [14:22:28] INFO | >> train_qwenlatent.py:487 + Step 8400 | grad_norm_pre_clip=0.2896 | + grad_norm_pre_clip_avg=0.2469 | Metrics: + {'align_loss': 0.0243297778069973, + 'recon_loss': 0.027231113985180855, + 'predict_loss': 0.01829436421394348, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2896157205104828, + 'mae_score': 0.020849138552004154, 'data_time': + 0.0009289149893447757, 'model_time': + 1.210845891997451, 'grad_norm_pre_clip_avg': + 0.24691386967897416, 'learning_rate': + 2.46498629947246e-05, 'epoch': 2.12} +04/19 [14:22:41] INFO | >> train_qwenlatent.py:487 + Step 8410 | grad_norm_pre_clip=0.2736 | + grad_norm_pre_clip_avg=0.3289 | Metrics: + {'align_loss': 0.02359427511692047, + 'recon_loss': 0.035832762718200684, + 'predict_loss': 0.022949256002902985, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.27356094121932983, + 'data_time': 0.001099100016290322, + 'model_time': 1.2173391029937193, + 'grad_norm_pre_clip_avg': 0.3288699597120285, + 'learning_rate': 2.4647810087049653e-05, + 'epoch': 2.12} +04/19 [14:22:53] INFO | >> train_qwenlatent.py:487 + Step 8420 | grad_norm_pre_clip=0.2660 | + grad_norm_pre_clip_avg=0.2740 | Metrics: + {'align_loss': 0.02346908301115036, + 'recon_loss': 0.017825018614530563, + 'predict_loss': 0.01272485964000225, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2660236954689026, + 'data_time': 0.000660544988932088, + 'model_time': 1.259606635023374, + 'grad_norm_pre_clip_avg': 0.2739849328994751, + 'learning_rate': 2.4645751264771992e-05, + 'epoch': 2.12} +04/19 [14:23:06] INFO | >> train_qwenlatent.py:487 + Step 8430 | grad_norm_pre_clip=0.2545 | + grad_norm_pre_clip_avg=0.2471 | Metrics: + {'align_loss': 0.022476041689515114, + 'recon_loss': 0.02330503612756729, + 'predict_loss': 0.015906620770692825, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2545433044433594, + 'data_time': 0.000658375007333234, + 'model_time': 1.2642808800155763, + 'grad_norm_pre_clip_avg': 0.24708943367004393, + 'learning_rate': 2.4643686528895064e-05, + 'epoch': 2.13} +04/19 [14:23:18] INFO | >> train_qwenlatent.py:487 + Step 8440 | grad_norm_pre_clip=0.3823 | + grad_norm_pre_clip_avg=0.2789 | Metrics: + {'align_loss': 0.025043489411473274, + 'recon_loss': 0.031401343643665314, + 'predict_loss': 0.020289024338126183, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.38225245475769043, + 'data_time': 0.0008794960158411413, + 'model_time': 1.2568067239772063, + 'grad_norm_pre_clip_avg': 0.2788514748215675, + 'learning_rate': 2.464161588042519e-05, + 'epoch': 2.13} +04/19 [14:23:32] INFO | >> train_qwenlatent.py:487 + Step 8450 | grad_norm_pre_clip=0.2535 | + grad_norm_pre_clip_avg=0.2791 | Metrics: + {'align_loss': 0.022604934871196747, + 'recon_loss': 0.02560919150710106, + 'predict_loss': 0.017902828752994537, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2535179853439331, + 'mae_score': 0.021624824592659065, 'data_time': + 0.0008494629873894155, 'model_time': + 1.4565700050152373, 'grad_norm_pre_clip_avg': + 0.27909244745969775, 'learning_rate': + 2.463953932037158e-05, 'epoch': 2.13} +04/19 [14:23:45] INFO | >> train_qwenlatent.py:487 + Step 8460 | grad_norm_pre_clip=0.2398 | + grad_norm_pre_clip_avg=0.2806 | Metrics: + {'align_loss': 0.024370333179831505, + 'recon_loss': 0.029474427923560143, + 'predict_loss': 0.01652829349040985, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23981846868991852, + 'data_time': 0.0009547249937895685, + 'model_time': 1.2795780320011545, + 'grad_norm_pre_clip_avg': 0.2805793657898903, + 'learning_rate': 2.4637456849746328e-05, + 'epoch': 2.13} +04/19 [14:23:57] INFO | >> train_qwenlatent.py:487 + Step 8470 | grad_norm_pre_clip=0.2470 | + grad_norm_pre_clip_avg=0.2399 | Metrics: + {'align_loss': 0.024631833657622337, + 'recon_loss': 0.030893906950950623, + 'predict_loss': 0.01783173345029354, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24703922867774963, + 'data_time': 0.0009201619832310826, + 'model_time': 1.3149768770090304, + 'grad_norm_pre_clip_avg': 0.2399023413658142, + 'learning_rate': 2.4635368469564407e-05, + 'epoch': 2.14} +04/19 [14:24:10] INFO | >> train_qwenlatent.py:487 + Step 8480 | grad_norm_pre_clip=0.2237 | + grad_norm_pre_clip_avg=0.2762 | Metrics: + {'align_loss': 0.026283565908670425, + 'recon_loss': 0.03825513273477554, + 'predict_loss': 0.02272314578294754, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22374649345874786, + 'data_time': 0.0007115770131349564, + 'model_time': 1.2078274100203998, + 'grad_norm_pre_clip_avg': 0.27617888152599335, + 'learning_rate': 2.463327418084366e-05, + 'epoch': 2.14} +04/19 [14:24:22] INFO | >> train_qwenlatent.py:487 + Step 8490 | grad_norm_pre_clip=0.3644 | + grad_norm_pre_clip_avg=0.3146 | Metrics: + {'align_loss': 0.024204744026064873, + 'recon_loss': 0.026444213464856148, + 'predict_loss': 0.014106936752796173, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3644266128540039, + 'data_time': 0.0008971230126917362, + 'model_time': 1.2147669660043903, + 'grad_norm_pre_clip_avg': 0.314577679336071, + 'learning_rate': 2.4631173984604825e-05, + 'epoch': 2.14} +04/19 [14:24:36] INFO | >> train_qwenlatent.py:487 + Step 8500 | grad_norm_pre_clip=0.2750 | + grad_norm_pre_clip_avg=0.2867 | Metrics: + {'align_loss': 0.02410437911748886, + 'recon_loss': 0.024650858715176582, + 'predict_loss': 0.014954928308725357, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2749866247177124, + 'mae_score': 0.03032218314505912, 'data_time': + 0.0009280889935325831, 'model_time': + 1.2452697469852865, 'grad_norm_pre_clip_avg': + 0.28665175586938857, 'learning_rate': + 2.4629067881871508e-05, 'epoch': 2.14} +04/19 [14:24:49] INFO | >> train_qwenlatent.py:487 + Step 8510 | grad_norm_pre_clip=0.2850 | + grad_norm_pre_clip_avg=0.2975 | Metrics: + {'align_loss': 0.023879658430814743, + 'recon_loss': 0.022365175187587738, + 'predict_loss': 0.013499478809535503, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2849741280078888, + 'data_time': 0.0008633460092823952, + 'model_time': 1.4398503929842263, + 'grad_norm_pre_clip_avg': 0.2975183755159378, + 'learning_rate': 2.4626955873670196e-05, + 'epoch': 2.15} +04/19 [14:25:01] INFO | >> train_qwenlatent.py:487 + Step 8520 | grad_norm_pre_clip=0.2504 | + grad_norm_pre_clip_avg=0.2833 | Metrics: + {'align_loss': 0.021591711789369583, + 'recon_loss': 0.018650704994797707, + 'predict_loss': 0.01465133111923933, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25043079257011414, + 'data_time': 0.001285890000872314, + 'model_time': 1.2719692130049225, + 'grad_norm_pre_clip_avg': 0.28330272138118745, + 'learning_rate': 2.4624837961030263e-05, + 'epoch': 2.15} +04/19 [14:25:14] INFO | >> train_qwenlatent.py:487 + Step 8530 | grad_norm_pre_clip=0.2611 | + grad_norm_pre_clip_avg=0.2688 | Metrics: + {'align_loss': 0.023887531831860542, + 'recon_loss': 0.029666321352124214, + 'predict_loss': 0.019023476168513298, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2611033618450165, + 'data_time': 0.0011047129810322076, + 'model_time': 1.2205998669960536, + 'grad_norm_pre_clip_avg': 0.26884114891290667, + 'learning_rate': 2.462271414498395e-05, + 'epoch': 2.15} +04/19 [14:25:26] INFO | >> train_qwenlatent.py:487 + Step 8540 | grad_norm_pre_clip=0.2590 | + grad_norm_pre_clip_avg=0.2540 | Metrics: + {'align_loss': 0.02417556382715702, + 'recon_loss': 0.019491029903292656, + 'predict_loss': 0.011630232445895672, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2590373456478119, + 'data_time': 0.0006379899859894067, + 'model_time': 1.2109090550220571, + 'grad_norm_pre_clip_avg': 0.2540492072701454, + 'learning_rate': 2.4620584426566377e-05, + 'epoch': 2.15} +04/19 [14:25:40] INFO | >> train_qwenlatent.py:487 + Step 8550 | grad_norm_pre_clip=0.2537 | + grad_norm_pre_clip_avg=0.2490 | Metrics: + {'align_loss': 0.023871392011642456, + 'recon_loss': 0.030329201370477676, + 'predict_loss': 0.01833086460828781, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25366637110710144, + 'mae_score': 0.03734593090710339, 'data_time': + 0.0011228169896639884, 'model_time': + 1.2351440500060562, 'grad_norm_pre_clip_avg': + 0.24896917641162872, 'learning_rate': + 2.461844880681555e-05, 'epoch': 2.16} +04/19 [14:25:52] INFO | >> train_qwenlatent.py:487 + Step 8560 | grad_norm_pre_clip=0.3505 | + grad_norm_pre_clip_avg=0.2902 | Metrics: + {'align_loss': 0.02330954000353813, + 'recon_loss': 0.02410247176885605, + 'predict_loss': 0.017281709238886833, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3504633903503418, + 'data_time': 0.0006499720038846135, + 'model_time': 1.219442722009262, + 'grad_norm_pre_clip_avg': 0.2902210161089897, + 'learning_rate': 2.4616307286772334e-05, + 'epoch': 2.16} +04/19 [14:26:05] INFO | >> train_qwenlatent.py:487 + Step 8570 | grad_norm_pre_clip=0.2612 | + grad_norm_pre_clip_avg=0.2633 | Metrics: + {'align_loss': 0.02450467087328434, + 'recon_loss': 0.032316502183675766, + 'predict_loss': 0.020021885633468628, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26115256547927856, + 'data_time': 0.0006623100198339671, + 'model_time': 1.2390183940005954, + 'grad_norm_pre_clip_avg': 0.2633220598101616, + 'learning_rate': 2.4614159867480485e-05, + 'epoch': 2.16} +04/19 [14:26:18] INFO | >> train_qwenlatent.py:487 + Step 8580 | grad_norm_pre_clip=0.3891 | + grad_norm_pre_clip_avg=0.2965 | Metrics: + {'align_loss': 0.021755896508693695, + 'recon_loss': 0.02297358587384224, + 'predict_loss': 0.02012159302830696, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.38905179500579834, + 'data_time': 0.0007098139903973788, + 'model_time': 1.2813451599795371, + 'grad_norm_pre_clip_avg': 0.2965123072266579, + 'learning_rate': 2.461200654998663e-05, + 'epoch': 2.17} +04/19 [14:26:30] INFO | >> train_qwenlatent.py:487 + Step 8590 | grad_norm_pre_clip=0.3497 | + grad_norm_pre_clip_avg=0.3433 | Metrics: + {'align_loss': 0.023236220702528954, + 'recon_loss': 0.026951802894473076, + 'predict_loss': 0.01908382959663868, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.349651575088501, + 'data_time': 0.0007871710113249719, + 'model_time': 1.255856154981302, + 'grad_norm_pre_clip_avg': 0.3432641804218292, + 'learning_rate': 2.4609847335340274e-05, + 'epoch': 2.17} +04/19 [14:26:44] INFO | >> train_qwenlatent.py:487 + Step 8600 | grad_norm_pre_clip=0.2413 | + grad_norm_pre_clip_avg=0.2614 | Metrics: + {'align_loss': 0.02427433617413044, + 'recon_loss': 0.023205408826470375, + 'predict_loss': 0.013840371742844582, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24131245911121368, + 'mae_score': 0.02190291086832682, 'data_time': + 0.0006411249923985451, 'model_time': + 1.2230599689937662, 'grad_norm_pre_clip_avg': + 0.26144261211156844, 'learning_rate': + 2.460768222459378e-05, 'epoch': 2.17} +04/19 [14:26:56] INFO | >> train_qwenlatent.py:487 + Step 8610 | grad_norm_pre_clip=0.2391 | + grad_norm_pre_clip_avg=0.2371 | Metrics: + {'align_loss': 0.023078229278326035, + 'recon_loss': 0.028823472559452057, + 'predict_loss': 0.016329461708664894, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23912648856639862, + 'data_time': 0.0006367390160448849, + 'model_time': 1.2081655870133545, + 'grad_norm_pre_clip_avg': 0.23707980811595916, + 'learning_rate': 2.4605511218802407e-05, + 'epoch': 2.17} +04/19 [14:27:09] INFO | >> train_qwenlatent.py:487 + Step 8620 | grad_norm_pre_clip=0.2315 | + grad_norm_pre_clip_avg=0.2287 | Metrics: + {'align_loss': 0.023107094690203667, + 'recon_loss': 0.022821540012955666, + 'predict_loss': 0.01473395898938179, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23152482509613037, + 'data_time': 0.0009211610013153404, + 'model_time': 1.254066766006872, + 'grad_norm_pre_clip_avg': 0.2287147730588913, + 'learning_rate': 2.4603334319024275e-05, + 'epoch': 2.18} +04/19 [14:27:21] INFO | >> train_qwenlatent.py:487 + Step 8630 | grad_norm_pre_clip=0.3357 | + grad_norm_pre_clip_avg=0.3360 | Metrics: + {'align_loss': 0.02356378361582756, + 'recon_loss': 0.027183085680007935, + 'predict_loss': 0.01556136179715395, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.33569809794425964, + 'data_time': 0.0006338249950204045, + 'model_time': 1.2780354300048202, + 'grad_norm_pre_clip_avg': 0.33603963255882263, + 'learning_rate': 2.4601151526320375e-05, + 'epoch': 2.18} +04/19 [14:27:34] INFO | >> train_qwenlatent.py:487 + Step 8640 | grad_norm_pre_clip=0.2737 | + grad_norm_pre_clip_avg=0.2586 | Metrics: + {'align_loss': 0.025179903954267502, + 'recon_loss': 0.030787652358412743, + 'predict_loss': 0.014129800722002983, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.27372658252716064, + 'data_time': 0.0010311689984519035, + 'model_time': 1.2432987930078525, + 'grad_norm_pre_clip_avg': 0.25863861590623854, + 'learning_rate': 2.4598962841754576e-05, + 'epoch': 2.18} +04/19 [14:27:47] INFO | >> train_qwenlatent.py:487 + Step 8650 | grad_norm_pre_clip=0.2867 | + grad_norm_pre_clip_avg=0.2709 | Metrics: + {'align_loss': 0.023574678227305412, + 'recon_loss': 0.029863234609365463, + 'predict_loss': 0.0177927203476429, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2866867482662201, + 'mae_score': 0.031533452626821154, 'data_time': + 0.001240273006260395, 'model_time': + 1.2274115089967381, 'grad_norm_pre_clip_avg': + 0.27093120515346525, 'learning_rate': + 2.4596768266393617e-05, 'epoch': 2.18} +04/19 [14:28:00] INFO | >> train_qwenlatent.py:487 + Step 8660 | grad_norm_pre_clip=0.2187 | + grad_norm_pre_clip_avg=0.2665 | Metrics: + {'align_loss': 0.025121120736002922, + 'recon_loss': 0.02808004803955555, + 'predict_loss': 0.017403729259967804, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21868035197257996, + 'data_time': 0.0010850289836525917, + 'model_time': 1.2575164159934502, + 'grad_norm_pre_clip_avg': 0.26653199940919875, + 'learning_rate': 2.4594567801307103e-05, + 'epoch': 2.19} +04/19 [14:28:13] INFO | >> train_qwenlatent.py:487 + Step 8670 | grad_norm_pre_clip=0.2730 | + grad_norm_pre_clip_avg=0.2716 | Metrics: + {'align_loss': 0.02381203882396221, + 'recon_loss': 0.035431452095508575, + 'predict_loss': 0.02217850089073181, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.27301692962646484, + 'data_time': 0.0010459760087542236, + 'model_time': 1.2416464909911156, + 'grad_norm_pre_clip_avg': 0.27161902487277984, + 'learning_rate': 2.4592361447567522e-05, + 'epoch': 2.19} +04/19 [14:28:26] INFO | >> train_qwenlatent.py:487 + Step 8680 | grad_norm_pre_clip=0.2475 | + grad_norm_pre_clip_avg=0.2544 | Metrics: + {'align_loss': 0.023179862648248672, + 'recon_loss': 0.029223499819636345, + 'predict_loss': 0.021962357684969902, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2475336492061615, + 'data_time': 0.0012912000238429755, + 'model_time': 1.3364516039728187, + 'grad_norm_pre_clip_avg': 0.25438886880874634, + 'learning_rate': 2.4590149206250216e-05, + 'epoch': 2.19} +04/19 [14:28:38] INFO | >> train_qwenlatent.py:487 + Step 8690 | grad_norm_pre_clip=0.3017 | + grad_norm_pre_clip_avg=0.2491 | Metrics: + {'align_loss': 0.024260971695184708, + 'recon_loss': 0.028096843510866165, + 'predict_loss': 0.01766585372388363, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.30170005559921265, + 'data_time': 0.0007136140193324536, + 'model_time': 1.205434386007255, + 'grad_norm_pre_clip_avg': 0.24912398755550386, + 'learning_rate': 2.458793107843341e-05, + 'epoch': 2.19} +04/19 [14:28:51] INFO | >> train_qwenlatent.py:487 + Step 8700 | grad_norm_pre_clip=0.2900 | + grad_norm_pre_clip_avg=0.2674 | Metrics: + {'align_loss': 0.024079646915197372, + 'recon_loss': 0.024092234671115875, + 'predict_loss': 0.013819095678627491, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2900215983390808, + 'mae_score': 0.0278988537487683, 'data_time': + 0.0010300400026608258, 'model_time': + 1.2597720829944592, 'grad_norm_pre_clip_avg': + 0.26743643879890444, 'learning_rate': + 2.458570706519819e-05, 'epoch': 2.2} +04/19 [14:29:04] INFO | >> train_qwenlatent.py:487 + Step 8710 | grad_norm_pre_clip=0.2340 | + grad_norm_pre_clip_avg=0.2300 | Metrics: + {'align_loss': 0.02400306612253189, + 'recon_loss': 0.02741890400648117, + 'predict_loss': 0.01799848861992359, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23398356139659882, + 'data_time': 0.0012192830035928637, + 'model_time': 1.2956380310060922, + 'grad_norm_pre_clip_avg': 0.22996189594268798, + 'learning_rate': 2.458347716762851e-05, + 'epoch': 2.2} +04/19 [14:29:17] INFO | >> train_qwenlatent.py:487 + Step 8720 | grad_norm_pre_clip=0.2761 | + grad_norm_pre_clip_avg=0.3178 | Metrics: + {'align_loss': 0.02619544044137001, + 'recon_loss': 0.035317499190568924, + 'predict_loss': 0.015085018239915371, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.27608686685562134, + 'data_time': 0.0007078220078255981, + 'model_time': 1.2577049210085534, + 'grad_norm_pre_clip_avg': 0.3177625223994255, + 'learning_rate': 2.4581241386811202e-05, + 'epoch': 2.2} +04/19 [14:29:29] INFO | >> train_qwenlatent.py:487 + Step 8730 | grad_norm_pre_clip=0.2870 | + grad_norm_pre_clip_avg=0.3336 | Metrics: + {'align_loss': 0.024864712730050087, + 'recon_loss': 0.020910732448101044, + 'predict_loss': 0.012335564009845257, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28701141476631165, + 'data_time': 0.0008708589884918183, + 'model_time': 1.2531733959913254, + 'grad_norm_pre_clip_avg': 0.3335564374923706, + 'learning_rate': 2.4578999723835952e-05, + 'epoch': 2.2} +04/19 [14:29:42] INFO | >> train_qwenlatent.py:487 + Step 8740 | grad_norm_pre_clip=0.2813 | + grad_norm_pre_clip_avg=0.2565 | Metrics: + {'align_loss': 0.024748995900154114, + 'recon_loss': 0.044207677245140076, + 'predict_loss': 0.026869412511587143, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28132107853889465, + 'data_time': 0.0006678439967799932, + 'model_time': 1.2344008249929175, + 'grad_norm_pre_clip_avg': 0.2565370470285416, + 'learning_rate': 2.4576752179795327e-05, + 'epoch': 2.21} +04/19 [14:29:55] INFO | >> train_qwenlatent.py:487 + Step 8750 | grad_norm_pre_clip=0.2505 | + grad_norm_pre_clip_avg=0.2439 | Metrics: + {'align_loss': 0.023333104327321053, + 'recon_loss': 0.022909631952643394, + 'predict_loss': 0.012854686938226223, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25047948956489563, + 'mae_score': 0.020228863621617223, 'data_time': + 0.0006294269987847656, 'model_time': + 1.197914503980428, 'grad_norm_pre_clip_avg': + 0.2439112514257431, 'learning_rate': + 2.4574498755784742e-05, 'epoch': 2.21} +04/19 [14:30:08] INFO | >> train_qwenlatent.py:487 + Step 8760 | grad_norm_pre_clip=0.2550 | + grad_norm_pre_clip_avg=0.2473 | Metrics: + {'align_loss': 0.024154581129550934, + 'recon_loss': 0.025373205542564392, + 'predict_loss': 0.01810193806886673, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25501924753189087, + 'data_time': 0.0008399460057262331, + 'model_time': 1.305989802989643, + 'grad_norm_pre_clip_avg': 0.2473122388124466, + 'learning_rate': 2.4572239452902494e-05, + 'epoch': 2.21} +04/19 [14:30:21] INFO | >> train_qwenlatent.py:487 + Step 8770 | grad_norm_pre_clip=0.2460 | + grad_norm_pre_clip_avg=0.2501 | Metrics: + {'align_loss': 0.02337070368230343, + 'recon_loss': 0.03020647168159485, + 'predict_loss': 0.022131364792585373, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2459581345319748, + 'data_time': 0.0009831440111156553, + 'model_time': 1.6042429160152096, + 'grad_norm_pre_clip_avg': 0.25011285990476606, + 'learning_rate': 2.4569974272249744e-05, + 'epoch': 2.21} +04/19 [14:30:34] INFO | >> train_qwenlatent.py:487 + Step 8780 | grad_norm_pre_clip=0.3296 | + grad_norm_pre_clip_avg=0.3059 | Metrics: + {'align_loss': 0.024638831615447998, + 'recon_loss': 0.0439184308052063, + 'predict_loss': 0.02794608660042286, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3296054005622864, + 'data_time': 0.000893516989890486, + 'model_time': 1.5491043939837255, + 'grad_norm_pre_clip_avg': 0.3058891147375107, + 'learning_rate': 2.456770321493051e-05, + 'epoch': 2.22} +04/19 [14:30:46] INFO | >> train_qwenlatent.py:487 + Step 8790 | grad_norm_pre_clip=0.2262 | + grad_norm_pre_clip_avg=0.2410 | Metrics: + {'align_loss': 0.024793105199933052, + 'recon_loss': 0.03270858898758888, + 'predict_loss': 0.017661839723587036, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22619707882404327, + 'data_time': 0.0008362499938812107, + 'model_time': 1.211435970995808, + 'grad_norm_pre_clip_avg': 0.24103241115808488, + 'learning_rate': 2.4565426282051676e-05, + 'epoch': 2.22} +04/19 [14:31:00] INFO | >> train_qwenlatent.py:487 + Step 8800 | grad_norm_pre_clip=0.2815 | + grad_norm_pre_clip_avg=0.2368 | Metrics: + {'align_loss': 0.02364516630768776, + 'recon_loss': 0.0318518690764904, + 'predict_loss': 0.020244434475898743, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28145769238471985, + 'mae_score': 0.03347790004970791, 'data_time': + 0.0008367770060431212, 'model_time': + 1.2909974360081833, 'grad_norm_pre_clip_avg': + 0.2368123084306717, 'learning_rate': + 2.456314347472299e-05, 'epoch': 2.22} +04/19 [14:31:13] INFO | >> train_qwenlatent.py:487 + Step 8810 | grad_norm_pre_clip=0.6073 | + grad_norm_pre_clip_avg=0.3196 | Metrics: + {'align_loss': 0.024630455300211906, + 'recon_loss': 0.023670779541134834, + 'predict_loss': 0.01462290994822979, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.6073408126831055, + 'data_time': 0.0008333460136782378, + 'model_time': 1.2765410769789014, + 'grad_norm_pre_clip_avg': 0.3195935100317001, + 'learning_rate': 2.456085479405707e-05, + 'epoch': 2.22} +04/19 [14:31:26] INFO | >> train_qwenlatent.py:487 + Step 8820 | grad_norm_pre_clip=0.2627 | + grad_norm_pre_clip_avg=0.3054 | Metrics: + {'align_loss': 0.024037444964051247, + 'recon_loss': 0.02682422287762165, + 'predict_loss': 0.015486306510865688, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2627352476119995, + 'data_time': 0.0008737600001040846, + 'model_time': 1.1936561479815282, + 'grad_norm_pre_clip_avg': 0.30538287609815595, + 'learning_rate': 2.455856024116939e-05, + 'epoch': 2.23} +04/19 [14:31:38] INFO | >> train_qwenlatent.py:487 + Step 8830 | grad_norm_pre_clip=0.2794 | + grad_norm_pre_clip_avg=0.2484 | Metrics: + {'align_loss': 0.026088982820510864, + 'recon_loss': 0.03714555874466896, + 'predict_loss': 0.020615994930267334, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2794412672519684, + 'data_time': 0.0006179909978527576, + 'model_time': 1.1960350620211102, + 'grad_norm_pre_clip_avg': 0.2483853355050087, + 'learning_rate': 2.455625981717828e-05, + 'epoch': 2.23} +04/19 [14:31:51] INFO | >> train_qwenlatent.py:487 + Step 8840 | grad_norm_pre_clip=0.1951 | + grad_norm_pre_clip_avg=0.2634 | Metrics: + {'align_loss': 0.02460169978439808, + 'recon_loss': 0.03186793252825737, + 'predict_loss': 0.019112268462777138, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1950988620519638, + 'data_time': 0.0010169529996346682, + 'model_time': 1.2732362379902042, + 'grad_norm_pre_clip_avg': 0.2634479835629463, + 'learning_rate': 2.455395352320495e-05, + 'epoch': 2.23} +04/19 [14:32:04] INFO | >> train_qwenlatent.py:487 + Step 8850 | grad_norm_pre_clip=0.2571 | + grad_norm_pre_clip_avg=0.2237 | Metrics: + {'align_loss': 0.02415083907544613, + 'recon_loss': 0.02543739601969719, + 'predict_loss': 0.014375378377735615, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25711601972579956, + 'mae_score': 0.03304856790078653, 'data_time': + 0.0008922969864215702, 'model_time': + 1.2072989119915292, 'grad_norm_pre_clip_avg': + 0.22369304895401002, 'learning_rate': + 2.455164136037345e-05, 'epoch': 2.23} +04/19 [14:32:17] INFO | >> train_qwenlatent.py:487 + Step 8860 | grad_norm_pre_clip=0.3327 | + grad_norm_pre_clip_avg=0.2635 | Metrics: + {'align_loss': 0.0241195410490036, + 'recon_loss': 0.024785783141851425, + 'predict_loss': 0.01592075638473034, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.33268171548843384, + 'data_time': 0.0006317030056379735, + 'model_time': 1.2217668660159688, + 'grad_norm_pre_clip_avg': 0.2635442540049553, + 'learning_rate': 2.4549323329810705e-05, + 'epoch': 2.24} +04/19 [14:32:30] INFO | >> train_qwenlatent.py:487 + Step 8870 | grad_norm_pre_clip=0.2107 | + grad_norm_pre_clip_avg=0.2798 | Metrics: + {'align_loss': 0.023930344730615616, + 'recon_loss': 0.0270567424595356, + 'predict_loss': 0.016630731523036957, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2107071876525879, + 'data_time': 0.0009298289951402694, + 'model_time': 1.239396826014854, + 'grad_norm_pre_clip_avg': 0.2797969952225685, + 'learning_rate': 2.4546999432646497e-05, + 'epoch': 2.24} +04/19 [14:32:42] INFO | >> train_qwenlatent.py:487 + Step 8880 | grad_norm_pre_clip=0.2173 | + grad_norm_pre_clip_avg=0.2431 | Metrics: + {'align_loss': 0.02496529370546341, + 'recon_loss': 0.045704592019319534, + 'predict_loss': 0.023269101977348328, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2172621637582779, + 'data_time': 0.0006747220177203417, + 'model_time': 1.241889494995121, + 'grad_norm_pre_clip_avg': 0.243143792450428, + 'learning_rate': 2.4544669670013464e-05, + 'epoch': 2.24} +04/19 [14:32:55] INFO | >> train_qwenlatent.py:487 + Step 8890 | grad_norm_pre_clip=0.2337 | + grad_norm_pre_clip_avg=0.2682 | Metrics: + {'align_loss': 0.023584024980664253, + 'recon_loss': 0.023811595514416695, + 'predict_loss': 0.010947738774120808, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2336837649345398, + 'data_time': 0.0010173699993174523, + 'model_time': 1.235332373995334, + 'grad_norm_pre_clip_avg': 0.26819915175437925, + 'learning_rate': 2.4542334043047095e-05, + 'epoch': 2.24} +04/19 [14:33:08] INFO | >> train_qwenlatent.py:487 + Step 8900 | grad_norm_pre_clip=0.3039 | + grad_norm_pre_clip_avg=0.2835 | Metrics: + {'align_loss': 0.025068897753953934, + 'recon_loss': 0.032361265271902084, + 'predict_loss': 0.01590217836201191, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3038662374019623, + 'mae_score': 0.02389651805430919, 'data_time': + 0.0009050880034919828, 'model_time': + 1.2020142480032519, 'grad_norm_pre_clip_avg': + 0.2835202023386955, 'learning_rate': + 2.4539992552885757e-05, 'epoch': 2.25} +04/19 [14:33:21] INFO | >> train_qwenlatent.py:487 + Step 8910 | grad_norm_pre_clip=0.1911 | + grad_norm_pre_clip_avg=0.2552 | Metrics: + {'align_loss': 0.02504083514213562, + 'recon_loss': 0.03854401409626007, + 'predict_loss': 0.021008165553212166, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19110821187496185, + 'data_time': 0.0009624249942135066, + 'model_time': 1.3133562239818275, + 'grad_norm_pre_clip_avg': 0.255214062333107, + 'learning_rate': 2.453764520067066e-05, + 'epoch': 2.25} +04/19 [14:33:34] INFO | >> train_qwenlatent.py:487 + Step 8920 | grad_norm_pre_clip=0.3184 | + grad_norm_pre_clip_avg=0.2688 | Metrics: + {'align_loss': 0.022162914276123047, + 'recon_loss': 0.025020956993103027, + 'predict_loss': 0.016987882554531097, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.318364679813385, + 'data_time': 0.0006687839922960848, + 'model_time': 1.2087105140089989, + 'grad_norm_pre_clip_avg': 0.2688218280673027, + 'learning_rate': 2.4535291987545878e-05, + 'epoch': 2.25} +04/19 [14:33:46] INFO | >> train_qwenlatent.py:487 + Step 8930 | grad_norm_pre_clip=0.2200 | + grad_norm_pre_clip_avg=0.2564 | Metrics: + {'align_loss': 0.023144975304603577, + 'recon_loss': 0.03338825702667236, + 'predict_loss': 0.022297631949186325, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2200031876564026, + 'data_time': 0.0007835129799786955, + 'model_time': 1.21912659399095, + 'grad_norm_pre_clip_avg': 0.25642337203025817, + 'learning_rate': 2.453293291465833e-05, + 'epoch': 2.25} +04/19 [14:33:59] INFO | >> train_qwenlatent.py:487 + Step 8940 | grad_norm_pre_clip=0.1987 | + grad_norm_pre_clip_avg=0.2319 | Metrics: + {'align_loss': 0.02383491024374962, + 'recon_loss': 0.02953239716589451, + 'predict_loss': 0.014131237752735615, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19871661067008972, + 'data_time': 0.0007170059834606946, + 'model_time': 1.1913638710102532, + 'grad_norm_pre_clip_avg': 0.2319225087761879, + 'learning_rate': 2.453056798315781e-05, + 'epoch': 2.26} +04/19 [14:34:12] INFO | >> train_qwenlatent.py:487 + Step 8950 | grad_norm_pre_clip=0.2954 | + grad_norm_pre_clip_avg=0.2901 | Metrics: + {'align_loss': 0.025350671261548996, + 'recon_loss': 0.03211221098899841, + 'predict_loss': 0.01644550822675228, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.29543137550354004, + 'mae_score': 0.02071692148844401, 'data_time': + 0.0008739419863559306, 'model_time': + 1.237660071987193, 'grad_norm_pre_clip_avg': + 0.29005805402994156, 'learning_rate': + 2.4528197194196952e-05, 'epoch': 2.26} +04/19 [14:34:24] INFO | >> train_qwenlatent.py:487 + Step 8960 | grad_norm_pre_clip=0.2421 | + grad_norm_pre_clip_avg=0.2206 | Metrics: + {'align_loss': 0.02433886006474495, + 'recon_loss': 0.04286661371588707, + 'predict_loss': 0.025697486475110054, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2420831173658371, + 'data_time': 0.0009253760217688978, + 'model_time': 1.2787353959865868, + 'grad_norm_pre_clip_avg': 0.22059136927127837, + 'learning_rate': 2.4525820548931248e-05, + 'epoch': 2.26} +04/19 [14:34:37] INFO | >> train_qwenlatent.py:487 + Step 8970 | grad_norm_pre_clip=0.2048 | + grad_norm_pre_clip_avg=0.2538 | Metrics: + {'align_loss': 0.02445819228887558, + 'recon_loss': 0.02502792328596115, + 'predict_loss': 0.012959660962224007, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20480020344257355, + 'data_time': 0.0007849560061004013, + 'model_time': 1.2548231359978672, + 'grad_norm_pre_clip_avg': 0.2538107857108116, + 'learning_rate': 2.4523438048519047e-05, + 'epoch': 2.26} +04/19 [14:34:50] INFO | >> train_qwenlatent.py:487 + Step 8980 | grad_norm_pre_clip=0.3358 | + grad_norm_pre_clip_avg=0.2526 | Metrics: + {'align_loss': 0.022940168157219887, + 'recon_loss': 0.032439496368169785, + 'predict_loss': 0.027310092002153397, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.33577466011047363, + 'data_time': 0.0006443040037993342, + 'model_time': 1.2405240240041167, + 'grad_norm_pre_clip_avg': 0.25260650366544724, + 'learning_rate': 2.4521049694121557e-05, + 'epoch': 2.27} +04/19 [14:35:02] INFO | >> train_qwenlatent.py:487 + Step 8990 | grad_norm_pre_clip=0.2308 | + grad_norm_pre_clip_avg=0.2832 | Metrics: + {'align_loss': 0.024948013946413994, + 'recon_loss': 0.033044200390577316, + 'predict_loss': 0.016598928719758987, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23080940544605255, + 'data_time': 0.0008282880007755011, + 'model_time': 1.2279272220039275, + 'grad_norm_pre_clip_avg': 0.283231095969677, + 'learning_rate': 2.4518655486902825e-05, + 'epoch': 2.27} +04/19 [14:35:16] INFO | >> train_qwenlatent.py:487 + Step 9000 | grad_norm_pre_clip=0.2991 | + grad_norm_pre_clip_avg=0.2410 | Metrics: + {'align_loss': 0.023895857855677605, + 'recon_loss': 0.03030056692659855, + 'predict_loss': 0.01606878452003002, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.29909563064575195, + 'mae_score': 0.02054565360954216, 'data_time': + 0.0006953740085009485, 'model_time': + 1.2873299340135418, 'grad_norm_pre_clip_avg': + 0.24103992879390718, 'learning_rate': + 2.4516255428029758e-05, 'epoch': 2.27} +04/19 [14:35:29] INFO | >> train_qwenlatent.py:487 + Step 9010 | grad_norm_pre_clip=0.2608 | + grad_norm_pre_clip_avg=0.2844 | Metrics: + {'align_loss': 0.024579588323831558, + 'recon_loss': 0.0277632474899292, + 'predict_loss': 0.014549053274095058, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.260831743478775, + 'data_time': 0.0007917780021671206, + 'model_time': 1.2401244780048728, + 'grad_norm_pre_clip_avg': 0.28436077386140823, + 'learning_rate': 2.451384951867212e-05, + 'epoch': 2.27} +04/19 [14:35:41] INFO | >> train_qwenlatent.py:487 + Step 9020 | grad_norm_pre_clip=0.2339 | + grad_norm_pre_clip_avg=0.2419 | Metrics: + {'align_loss': 0.023198705166578293, + 'recon_loss': 0.0298151895403862, + 'predict_loss': 0.018032126128673553, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23394958674907684, + 'data_time': 0.0008762500074226409, + 'model_time': 1.2116922400018666, + 'grad_norm_pre_clip_avg': 0.24194328337907792, + 'learning_rate': 2.4511437760002523e-05, + 'epoch': 2.28} +04/19 [14:35:54] INFO | >> train_qwenlatent.py:487 + Step 9030 | grad_norm_pre_clip=0.2461 | + grad_norm_pre_clip_avg=0.2660 | Metrics: + {'align_loss': 0.024166014045476913, + 'recon_loss': 0.027288585901260376, + 'predict_loss': 0.017289049923419952, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24610517919063568, + 'data_time': 0.0006761550030205399, + 'model_time': 1.2546150580164976, + 'grad_norm_pre_clip_avg': 0.2660172551870346, + 'learning_rate': 2.4509020153196422e-05, + 'epoch': 2.28} +04/19 [14:36:07] INFO | >> train_qwenlatent.py:487 + Step 9040 | grad_norm_pre_clip=0.1932 | + grad_norm_pre_clip_avg=0.3010 | Metrics: + {'align_loss': 0.025043955072760582, + 'recon_loss': 0.03578123450279236, + 'predict_loss': 0.01697080209851265, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19315603375434875, + 'data_time': 0.0009756569925229996, + 'model_time': 1.3213606990175322, + 'grad_norm_pre_clip_avg': 0.3009913876652718, + 'learning_rate': 2.450659669943214e-05, + 'epoch': 2.28} +04/19 [14:36:20] INFO | >> train_qwenlatent.py:487 + Step 9050 | grad_norm_pre_clip=0.2464 | + grad_norm_pre_clip_avg=0.2537 | Metrics: + {'align_loss': 0.02392883226275444, + 'recon_loss': 0.02233855240046978, + 'predict_loss': 0.015244929119944572, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2463843673467636, + 'mae_score': 0.017197852950912338, 'data_time': + 0.0009548570087645203, 'model_time': + 1.209320882015163, 'grad_norm_pre_clip_avg': + 0.25374616831541064, 'learning_rate': + 2.450416739989083e-05, 'epoch': 2.28} +04/19 [14:36:33] INFO | >> train_qwenlatent.py:487 + Step 9060 | grad_norm_pre_clip=0.2372 | + grad_norm_pre_clip_avg=0.2162 | Metrics: + {'align_loss': 0.02389400452375412, + 'recon_loss': 0.026980392634868622, + 'predict_loss': 0.01482266653329134, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23721203207969666, + 'data_time': 0.0009439569839742035, + 'model_time': 1.2229679999873042, + 'grad_norm_pre_clip_avg': 0.21619794368743897, + 'learning_rate': 2.4501732255756503e-05, + 'epoch': 2.29} +04/19 [14:36:46] INFO | >> train_qwenlatent.py:487 + Step 9070 | grad_norm_pre_clip=0.2104 | + grad_norm_pre_clip_avg=0.2829 | Metrics: + {'align_loss': 0.02525663748383522, + 'recon_loss': 0.03479832410812378, + 'predict_loss': 0.020379433408379555, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21042966842651367, + 'data_time': 0.0008049859898164868, + 'model_time': 1.2477914699993562, + 'grad_norm_pre_clip_avg': 0.28285605013370513, + 'learning_rate': 2.449929126821602e-05, + 'epoch': 2.29} +04/19 [14:36:58] INFO | >> train_qwenlatent.py:487 + Step 9080 | grad_norm_pre_clip=0.2694 | + grad_norm_pre_clip_avg=0.2801 | Metrics: + {'align_loss': 0.024625053629279137, + 'recon_loss': 0.024239907041192055, + 'predict_loss': 0.013304480351507664, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2694361209869385, + 'data_time': 0.000781084003392607, + 'model_time': 1.2052261740027461, + 'grad_norm_pre_clip_avg': 0.28013955950737, + 'learning_rate': 2.44968444384591e-05, 'epoch': + 2.29} +04/19 [14:37:11] INFO | >> train_qwenlatent.py:487 + Step 9090 | grad_norm_pre_clip=0.2218 | + grad_norm_pre_clip_avg=0.2711 | Metrics: + {'align_loss': 0.026028983294963837, + 'recon_loss': 0.04070061445236206, + 'predict_loss': 0.023393183946609497, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22183696925640106, + 'data_time': 0.000711884000338614, + 'model_time': 1.2349094130040612, + 'grad_norm_pre_clip_avg': 0.27110120058059695, + 'learning_rate': 2.449439176767828e-05, + 'epoch': 2.29} +04/19 [14:37:24] INFO | >> train_qwenlatent.py:487 + Step 9100 | grad_norm_pre_clip=0.1756 | + grad_norm_pre_clip_avg=0.2351 | Metrics: + {'align_loss': 0.02549562230706215, + 'recon_loss': 0.04285673797130585, + 'predict_loss': 0.020881103351712227, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17564576864242554, + 'mae_score': 0.027314107052914732, 'data_time': + 0.0007004390063229948, 'model_time': + 1.2504404679930303, 'grad_norm_pre_clip_avg': + 0.23510531336069107, 'learning_rate': + 2.449193325706897e-05, 'epoch': 2.3} +04/19 [14:37:37] INFO | >> train_qwenlatent.py:487 + Step 9110 | grad_norm_pre_clip=0.5166 | + grad_norm_pre_clip_avg=0.2731 | Metrics: + {'align_loss': 0.024352505803108215, + 'recon_loss': 0.036412257701158524, + 'predict_loss': 0.0226710457354784, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.5166425704956055, + 'data_time': 0.0009182960202451795, + 'model_time': 1.2773098390025552, + 'grad_norm_pre_clip_avg': 0.2731180161237717, + 'learning_rate': 2.4489468907829417e-05, + 'epoch': 2.3} +04/19 [14:37:50] INFO | >> train_qwenlatent.py:487 + Step 9120 | grad_norm_pre_clip=0.3184 | + grad_norm_pre_clip_avg=0.3229 | Metrics: + {'align_loss': 0.024153001606464386, + 'recon_loss': 0.035504598170518875, + 'predict_loss': 0.018548984080553055, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.31837037205696106, + 'data_time': 0.0007171829929575324, + 'model_time': 1.3150536880129948, + 'grad_norm_pre_clip_avg': 0.3228830486536026, + 'learning_rate': 2.448699872116072e-05, + 'epoch': 2.3} +04/19 [14:38:02] INFO | >> train_qwenlatent.py:487 + Step 9130 | grad_norm_pre_clip=0.2857 | + grad_norm_pre_clip_avg=0.2807 | Metrics: + {'align_loss': 0.024033289402723312, + 'recon_loss': 0.034566644579172134, + 'predict_loss': 0.019692312926054, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2856651842594147, + 'data_time': 0.0007285169849637896, + 'model_time': 1.2472903140005656, + 'grad_norm_pre_clip_avg': 0.2807191476225853, + 'learning_rate': 2.4484522698266812e-05, + 'epoch': 2.3} +04/19 [14:38:15] INFO | >> train_qwenlatent.py:487 + Step 9140 | grad_norm_pre_clip=0.2236 | + grad_norm_pre_clip_avg=0.2330 | Metrics: + {'align_loss': 0.02532952092587948, + 'recon_loss': 0.04122508317232132, + 'predict_loss': 0.021497482433915138, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22364552319049835, + 'data_time': 0.0009658899798523635, + 'model_time': 1.2159122299926821, + 'grad_norm_pre_clip_avg': 0.23302773833274842, + 'learning_rate': 2.4482040840354482e-05, + 'epoch': 2.31} +04/19 [14:38:28] INFO | >> train_qwenlatent.py:487 + Step 9150 | grad_norm_pre_clip=0.2186 | + grad_norm_pre_clip_avg=0.2251 | Metrics: + {'align_loss': 0.022413305938243866, + 'recon_loss': 0.033248428255319595, + 'predict_loss': 0.01890069618821144, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21858298778533936, + 'mae_score': 0.023869337477125562, 'data_time': + 0.000982149998890236, 'model_time': + 1.3066993890097365, 'grad_norm_pre_clip_avg': + 0.22505034357309342, 'learning_rate': + 2.4479553148633354e-05, 'epoch': 2.31} +04/19 [14:38:41] INFO | >> train_qwenlatent.py:487 + Step 9160 | grad_norm_pre_clip=0.2104 | + grad_norm_pre_clip_avg=0.2694 | Metrics: + {'align_loss': 0.02332388609647751, + 'recon_loss': 0.026719234883785248, + 'predict_loss': 0.013127845712006092, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2104489803314209, + 'data_time': 0.0009060869924724102, + 'model_time': 1.2074836489919107, + 'grad_norm_pre_clip_avg': 0.2693530574440956, + 'learning_rate': 2.4477059624315894e-05, + 'epoch': 2.31} +04/19 [14:38:54] INFO | >> train_qwenlatent.py:487 + Step 9170 | grad_norm_pre_clip=0.1922 | + grad_norm_pre_clip_avg=0.2580 | Metrics: + {'align_loss': 0.023957185447216034, + 'recon_loss': 0.038097869604825974, + 'predict_loss': 0.01772068813443184, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19217021763324738, + 'data_time': 0.0006537679873872548, + 'model_time': 1.4300636720145121, + 'grad_norm_pre_clip_avg': 0.25804775655269624, + 'learning_rate': 2.4474560268617426e-05, + 'epoch': 2.31} +04/19 [14:39:07] INFO | >> train_qwenlatent.py:487 + Step 9180 | grad_norm_pre_clip=0.3067 | + grad_norm_pre_clip_avg=0.3107 | Metrics: + {'align_loss': 0.023018214851617813, + 'recon_loss': 0.03317510336637497, + 'predict_loss': 0.023357504978775978, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.30670860409736633, + 'data_time': 0.001103016984416172, + 'model_time': 1.3911876310012303, + 'grad_norm_pre_clip_avg': 0.310730941593647, + 'learning_rate': 2.4472055082756094e-05, + 'epoch': 2.32} +04/19 [14:39:20] INFO | >> train_qwenlatent.py:487 + Step 9190 | grad_norm_pre_clip=0.2601 | + grad_norm_pre_clip_avg=0.2444 | Metrics: + {'align_loss': 0.025666380301117897, + 'recon_loss': 0.036313872784376144, + 'predict_loss': 0.02035529725253582, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26008716225624084, + 'data_time': 0.000690774992108345, + 'model_time': 1.2383574760169722, + 'grad_norm_pre_clip_avg': 0.24443114399909974, + 'learning_rate': 2.4469544067952904e-05, + 'epoch': 2.32} +04/19 [14:39:33] INFO | >> train_qwenlatent.py:487 + Step 9200 | grad_norm_pre_clip=0.1925 | + grad_norm_pre_clip_avg=0.2123 | Metrics: + {'align_loss': 0.024005191400647163, + 'recon_loss': 0.025113027542829514, + 'predict_loss': 0.01247694157063961, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1924666166305542, + 'mae_score': 0.02104090269621428, 'data_time': + 0.0007969739963300526, 'model_time': + 1.1941541930136736, 'grad_norm_pre_clip_avg': + 0.21232747584581374, 'learning_rate': + 2.4467027225431692e-05, 'epoch': 2.32} +04/19 [14:39:46] INFO | >> train_qwenlatent.py:487 + Step 9210 | grad_norm_pre_clip=0.2411 | + grad_norm_pre_clip_avg=0.2377 | Metrics: + {'align_loss': 0.023732492700219154, + 'recon_loss': 0.029839403927326202, + 'predict_loss': 0.015913503244519234, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24106913805007935, + 'data_time': 0.0010033069993369281, + 'model_time': 1.264730550989043, + 'grad_norm_pre_clip_avg': 0.2376690074801445, + 'learning_rate': 2.4464504556419135e-05, + 'epoch': 2.32} +04/19 [14:39:58] INFO | >> train_qwenlatent.py:487 + Step 9220 | grad_norm_pre_clip=0.2139 | + grad_norm_pre_clip_avg=0.2848 | Metrics: + {'align_loss': 0.02389925718307495, + 'recon_loss': 0.021711135283112526, + 'predict_loss': 0.012092662043869495, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21391373872756958, + 'data_time': 0.0009743999980855733, + 'model_time': 1.259828929003561, + 'grad_norm_pre_clip_avg': 0.28484332412481306, + 'learning_rate': 2.446197606214475e-05, + 'epoch': 2.33} +04/19 [14:40:11] INFO | >> train_qwenlatent.py:487 + Step 9230 | grad_norm_pre_clip=0.2011 | + grad_norm_pre_clip_avg=0.2146 | Metrics: + {'align_loss': 0.024560289457440376, + 'recon_loss': 0.025969840586185455, + 'predict_loss': 0.015214521437883377, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20114535093307495, + 'data_time': 0.00106352599686943, 'model_time': + 1.2545411259925459, 'grad_norm_pre_clip_avg': + 0.21457336992025375, 'learning_rate': + 2.4459441743840896e-05, 'epoch': 2.33} +04/19 [14:40:24] INFO | >> train_qwenlatent.py:487 + Step 9240 | grad_norm_pre_clip=0.3477 | + grad_norm_pre_clip_avg=0.2518 | Metrics: + {'align_loss': 0.023885425180196762, + 'recon_loss': 0.021627278998494148, + 'predict_loss': 0.01226650271564722, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.34765341877937317, + 'data_time': 0.0009296580101363361, + 'model_time': 1.234700745990267, + 'grad_norm_pre_clip_avg': 0.25176447331905366, + 'learning_rate': 2.4456901602742773e-05, + 'epoch': 2.33} +04/19 [14:40:38] INFO | >> train_qwenlatent.py:487 + Step 9250 | grad_norm_pre_clip=0.3558 | + grad_norm_pre_clip_avg=0.3243 | Metrics: + {'align_loss': 0.02310364507138729, + 'recon_loss': 0.028952425345778465, + 'predict_loss': 0.016616901382803917, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3558385968208313, + 'mae_score': 0.021124666231172578, 'data_time': + 0.0011439060035627335, 'model_time': + 1.251539293996757, 'grad_norm_pre_clip_avg': + 0.3243356943130493, 'learning_rate': + 2.445435564008841e-05, 'epoch': 2.33} +04/19 [14:40:50] INFO | >> train_qwenlatent.py:487 + Step 9260 | grad_norm_pre_clip=0.2677 | + grad_norm_pre_clip_avg=0.2654 | Metrics: + {'align_loss': 0.023395808413624763, + 'recon_loss': 0.026217583566904068, + 'predict_loss': 0.01540432870388031, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2676776349544525, + 'data_time': 0.0012143239728175104, + 'model_time': 1.2842019509989768, + 'grad_norm_pre_clip_avg': 0.2654185026884079, + 'learning_rate': 2.4451803857118674e-05, + 'epoch': 2.34} +04/19 [14:41:03] INFO | >> train_qwenlatent.py:487 + Step 9270 | grad_norm_pre_clip=0.2691 | + grad_norm_pre_clip_avg=0.2776 | Metrics: + {'align_loss': 0.024300239980220795, + 'recon_loss': 0.041206829249858856, + 'predict_loss': 0.023778444156050682, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26912981271743774, + 'data_time': 0.0006727419968228787, + 'model_time': 1.2443800960027147, + 'grad_norm_pre_clip_avg': 0.27760696560144427, + 'learning_rate': 2.4449246255077283e-05, + 'epoch': 2.34} +04/19 [14:41:16] INFO | >> train_qwenlatent.py:487 + Step 9280 | grad_norm_pre_clip=0.2411 | + grad_norm_pre_clip_avg=0.2580 | Metrics: + {'align_loss': 0.024644972756505013, + 'recon_loss': 0.029056811705231667, + 'predict_loss': 0.013594865798950195, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24105912446975708, + 'data_time': 0.0009395449887961149, + 'model_time': 1.2072861680062488, + 'grad_norm_pre_clip_avg': 0.25799201875925065, + 'learning_rate': 2.4446682835210778e-05, + 'epoch': 2.34} +04/19 [14:41:28] INFO | >> train_qwenlatent.py:487 + Step 9290 | grad_norm_pre_clip=0.1937 | + grad_norm_pre_clip_avg=0.2467 | Metrics: + {'align_loss': 0.02503918670117855, + 'recon_loss': 0.035923633724451065, + 'predict_loss': 0.018441371619701385, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19374847412109375, + 'data_time': 0.0009021209843922406, + 'model_time': 1.5242508559895214, + 'grad_norm_pre_clip_avg': 0.24670180827379226, + 'learning_rate': 2.4444113598768533e-05, + 'epoch': 2.34} +04/19 [14:41:42] INFO | >> train_qwenlatent.py:487 + Step 9300 | grad_norm_pre_clip=0.1937 | + grad_norm_pre_clip_avg=0.2333 | Metrics: + {'align_loss': 0.024735502898693085, + 'recon_loss': 0.0358579121530056, + 'predict_loss': 0.01737828366458416, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1937045305967331, + 'mae_score': 0.023115575635755385, 'data_time': + 0.0010211720073129982, 'model_time': + 1.2253588619932998, 'grad_norm_pre_clip_avg': + 0.23332978934049606, 'learning_rate': + 2.4441538547002765e-05, 'epoch': 2.35} +04/19 [14:41:55] INFO | >> train_qwenlatent.py:487 + Step 9310 | grad_norm_pre_clip=0.2681 | + grad_norm_pre_clip_avg=0.2346 | Metrics: + {'align_loss': 0.024271391332149506, + 'recon_loss': 0.029752619564533234, + 'predict_loss': 0.012088429182767868, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2680562734603882, + 'data_time': 0.0010227739985566586, + 'model_time': 1.2359221179794986, + 'grad_norm_pre_clip_avg': 0.2346255287528038, + 'learning_rate': 2.4438957681168534e-05, + 'epoch': 2.35} +04/19 [14:42:07] INFO | >> train_qwenlatent.py:487 + Step 9320 | grad_norm_pre_clip=0.4390 | + grad_norm_pre_clip_avg=0.3323 | Metrics: + {'align_loss': 0.0252191461622715, + 'recon_loss': 0.03203124925494194, + 'predict_loss': 0.014227594248950481, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.4389994144439697, + 'data_time': 0.0007210460025817156, + 'model_time': 1.2483632450166624, + 'grad_norm_pre_clip_avg': 0.3322692155838013, + 'learning_rate': 2.4436371002523707e-05, + 'epoch': 2.35} +04/19 [14:42:20] INFO | >> train_qwenlatent.py:487 + Step 9330 | grad_norm_pre_clip=0.2256 | + grad_norm_pre_clip_avg=0.2656 | Metrics: + {'align_loss': 0.0244673490524292, + 'recon_loss': 0.028111081570386887, + 'predict_loss': 0.015057641081511974, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22564569115638733, + 'data_time': 0.0006765110010746866, + 'model_time': 1.2765208929777145, + 'grad_norm_pre_clip_avg': 0.2656349912285805, + 'learning_rate': 2.4433778512329003e-05, + 'epoch': 2.35} +04/19 [14:42:32] INFO | >> train_qwenlatent.py:487 + Step 9340 | grad_norm_pre_clip=0.2289 | + grad_norm_pre_clip_avg=0.2377 | Metrics: + {'align_loss': 0.022867174819111824, + 'recon_loss': 0.031635478138923645, + 'predict_loss': 0.021019043400883675, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2289169579744339, + 'data_time': 0.0006808260222896934, + 'model_time': 1.2061771640146617, + 'grad_norm_pre_clip_avg': 0.2377378299832344, + 'learning_rate': 2.443118021184798e-05, + 'epoch': 2.36} +04/19 [14:42:46] INFO | >> train_qwenlatent.py:487 + Step 9350 | grad_norm_pre_clip=0.2218 | + grad_norm_pre_clip_avg=0.2248 | Metrics: + {'align_loss': 0.023323211818933487, + 'recon_loss': 0.02931167744100094, + 'predict_loss': 0.02006607875227928, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2217830866575241, + 'mae_score': 0.020496746441265486, 'data_time': + 0.0009451530058868229, 'model_time': + 1.2471639220020734, 'grad_norm_pre_clip_avg': + 0.2247587412595749, 'learning_rate': + 2.4428576102347007e-05, 'epoch': 2.36} +04/19 [14:42:58] INFO | >> train_qwenlatent.py:487 + Step 9360 | grad_norm_pre_clip=0.2735 | + grad_norm_pre_clip_avg=0.3082 | Metrics: + {'align_loss': 0.02347888983786106, + 'recon_loss': 0.02863350138068199, + 'predict_loss': 0.012296637520194054, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.27349498867988586, + 'data_time': 0.001002852019155398, + 'model_time': 1.2110291669960134, + 'grad_norm_pre_clip_avg': 0.30819656103849413, + 'learning_rate': 2.4425966185095303e-05, + 'epoch': 2.36} +04/19 [14:43:11] INFO | >> train_qwenlatent.py:487 + Step 9370 | grad_norm_pre_clip=0.2218 | + grad_norm_pre_clip_avg=0.2299 | Metrics: + {'align_loss': 0.02315628156065941, + 'recon_loss': 0.031355127692222595, + 'predict_loss': 0.019362611696124077, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22183169424533844, + 'data_time': 0.0009384849981870502, + 'model_time': 1.2549295999924652, + 'grad_norm_pre_clip_avg': 0.2298794373869896, + 'learning_rate': 2.4423350461364905e-05, + 'epoch': 2.36} +04/19 [14:43:23] INFO | >> train_qwenlatent.py:487 + Step 9380 | grad_norm_pre_clip=0.2406 | + grad_norm_pre_clip_avg=0.2591 | Metrics: + {'align_loss': 0.02349560335278511, + 'recon_loss': 0.030807627364993095, + 'predict_loss': 0.015516731888055801, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24057215452194214, + 'data_time': 0.0006636670150328428, + 'model_time': 1.2087227549927775, + 'grad_norm_pre_clip_avg': 0.25914655178785323, + 'learning_rate': 2.442072893243069e-05, + 'epoch': 2.37} +04/19 [14:43:36] INFO | >> train_qwenlatent.py:487 + Step 9390 | grad_norm_pre_clip=0.2443 | + grad_norm_pre_clip_avg=0.2303 | Metrics: + {'align_loss': 0.024926118552684784, + 'recon_loss': 0.03246019408106804, + 'predict_loss': 0.013276292942464352, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24428530037403107, + 'data_time': 0.0008298320171888918, + 'model_time': 1.2618255249981303, + 'grad_norm_pre_clip_avg': 0.23031651377677917, + 'learning_rate': 2.4418101599570353e-05, + 'epoch': 2.37} +04/19 [14:43:49] INFO | >> train_qwenlatent.py:487 + Step 9400 | grad_norm_pre_clip=0.2823 | + grad_norm_pre_clip_avg=0.2861 | Metrics: + {'align_loss': 0.024567293003201485, + 'recon_loss': 0.03870759159326553, + 'predict_loss': 0.025398846715688705, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2823372185230255, + 'mae_score': 0.01995211936332084, 'data_time': + 0.0006817230023443699, 'model_time': + 1.2187811879848596, 'grad_norm_pre_clip_avg': + 0.28612054139375687, 'learning_rate': + 2.4415468464064425e-05, 'epoch': 2.37} +04/19 [14:44:02] INFO | >> train_qwenlatent.py:487 + Step 9410 | grad_norm_pre_clip=0.1973 | + grad_norm_pre_clip_avg=0.2338 | Metrics: + {'align_loss': 0.0224965438246727, + 'recon_loss': 0.038301412016153336, + 'predict_loss': 0.01913858763873577, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19734828174114227, + 'data_time': 0.0008969400078058243, + 'model_time': 1.2306391849997453, + 'grad_norm_pre_clip_avg': 0.2337589830160141, + 'learning_rate': 2.4412829527196275e-05, + 'epoch': 2.37} +04/19 [14:44:15] INFO | >> train_qwenlatent.py:487 + Step 9420 | grad_norm_pre_clip=0.4119 | + grad_norm_pre_clip_avg=0.2670 | Metrics: + {'align_loss': 0.02373196743428707, + 'recon_loss': 0.031182384118437767, + 'predict_loss': 0.014889495447278023, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.41193607449531555, + 'data_time': 0.0008990569913294166, + 'model_time': 1.1934637459926307, + 'grad_norm_pre_clip_avg': 0.267025788128376, + 'learning_rate': 2.441018479025207e-05, + 'epoch': 2.38} +04/19 [14:44:27] INFO | >> train_qwenlatent.py:487 + Step 9430 | grad_norm_pre_clip=0.2309 | + grad_norm_pre_clip_avg=0.2826 | Metrics: + {'align_loss': 0.023526962846517563, + 'recon_loss': 0.0330624133348465, + 'predict_loss': 0.017224900424480438, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2308918535709381, + 'data_time': 0.0006533749983645976, + 'model_time': 1.206636835995596, + 'grad_norm_pre_clip_avg': 0.2826239049434662, + 'learning_rate': 2.4407534254520837e-05, + 'epoch': 2.38} +04/19 [14:44:40] INFO | >> train_qwenlatent.py:487 + Step 9440 | grad_norm_pre_clip=0.2385 | + grad_norm_pre_clip_avg=0.2258 | Metrics: + {'align_loss': 0.024392224848270416, + 'recon_loss': 0.04564143717288971, + 'predict_loss': 0.02809208817780018, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2384544163942337, + 'data_time': 0.0007948379789013416, + 'model_time': 1.2260719859914389, + 'grad_norm_pre_clip_avg': 0.2258492350578308, + 'learning_rate': 2.4404877921294414e-05, + 'epoch': 2.38} +04/19 [14:44:54] INFO | >> train_qwenlatent.py:487 + Step 9450 | grad_norm_pre_clip=0.4042 | + grad_norm_pre_clip_avg=0.2990 | Metrics: + {'align_loss': 0.0244184210896492, + 'recon_loss': 0.023056024685502052, + 'predict_loss': 0.013154366053640842, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.4041992425918579, + 'mae_score': 0.016677622752146678, 'data_time': + 0.0007032229914329946, 'model_time': + 1.2318258859741036, 'grad_norm_pre_clip_avg': + 0.2990487813949585, 'learning_rate': + 2.440221579186746e-05, 'epoch': 2.38} +04/19 [14:45:06] INFO | >> train_qwenlatent.py:487 + Step 9460 | grad_norm_pre_clip=0.2404 | + grad_norm_pre_clip_avg=0.2643 | Metrics: + {'align_loss': 0.023652754724025726, + 'recon_loss': 0.030085531994700432, + 'predict_loss': 0.01883772388100624, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2404002845287323, + 'data_time': 0.0007954239845275879, + 'model_time': 1.2168054540234152, + 'grad_norm_pre_clip_avg': 0.26431541293859484, + 'learning_rate': 2.4399547867537464e-05, + 'epoch': 2.39} +04/19 [14:45:18] INFO | >> train_qwenlatent.py:487 + Step 9470 | grad_norm_pre_clip=0.2638 | + grad_norm_pre_clip_avg=0.2244 | Metrics: + {'align_loss': 0.022641533985733986, + 'recon_loss': 0.03020675666630268, + 'predict_loss': 0.012676425278186798, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26379573345184326, + 'data_time': 0.0010406229994259775, + 'model_time': 1.3236159819934983, + 'grad_norm_pre_clip_avg': 0.22440762519836427, + 'learning_rate': 2.4396874149604747e-05, + 'epoch': 2.39} +04/19 [14:45:31] INFO | >> train_qwenlatent.py:487 + Step 9480 | grad_norm_pre_clip=0.2321 | + grad_norm_pre_clip_avg=0.2756 | Metrics: + {'align_loss': 0.024056997150182724, + 'recon_loss': 0.02632424235343933, + 'predict_loss': 0.015601486898958683, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23211520910263062, + 'data_time': 0.0007806289941072464, + 'model_time': 1.2563160909921862, + 'grad_norm_pre_clip_avg': 0.2756166458129883, + 'learning_rate': 2.439419463937244e-05, + 'epoch': 2.39} +04/19 [14:45:44] INFO | >> train_qwenlatent.py:487 + Step 9490 | grad_norm_pre_clip=0.3051 | + grad_norm_pre_clip_avg=0.2961 | Metrics: + {'align_loss': 0.02516338974237442, + 'recon_loss': 0.04076491296291351, + 'predict_loss': 0.02132568322122097, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.30510658025741577, + 'data_time': 0.001218227989738807, + 'model_time': 1.257545819011284, + 'grad_norm_pre_clip_avg': 0.2961451306939125, + 'learning_rate': 2.4391509338146506e-05, + 'epoch': 2.39} +04/19 [14:45:57] INFO | >> train_qwenlatent.py:487 + Step 9500 | grad_norm_pre_clip=0.2340 | + grad_norm_pre_clip_avg=0.2620 | Metrics: + {'align_loss': 0.023460455238819122, + 'recon_loss': 0.03510265424847603, + 'predict_loss': 0.018174396827816963, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23404157161712646, + 'mae_score': 0.028313246718398084, 'data_time': + 0.0008071179909165949, 'model_time': + 1.2645227590110153, 'grad_norm_pre_clip_avg': + 0.2619712188839912, 'learning_rate': + 2.4388818247235733e-05, 'epoch': 2.4} +04/19 [14:46:10] INFO | >> train_qwenlatent.py:487 + Step 9510 | grad_norm_pre_clip=0.2940 | + grad_norm_pre_clip_avg=0.2435 | Metrics: + {'align_loss': 0.024583548307418823, + 'recon_loss': 0.03610823303461075, + 'predict_loss': 0.012977465987205505, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.29396677017211914, + 'data_time': 0.0009239779901690781, + 'model_time': 1.2218018870044034, + 'grad_norm_pre_clip_avg': 0.2434924840927124, + 'learning_rate': 2.4386121367951713e-05, + 'epoch': 2.4} +04/19 [14:46:22] INFO | >> train_qwenlatent.py:487 + Step 9520 | grad_norm_pre_clip=0.2636 | + grad_norm_pre_clip_avg=0.2846 | Metrics: + {'align_loss': 0.02337253838777542, + 'recon_loss': 0.032040178775787354, + 'predict_loss': 0.01666552945971489, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2636242210865021, + 'data_time': 0.0006944090127944946, + 'model_time': 1.2625342149985954, + 'grad_norm_pre_clip_avg': 0.28456281423568724, + 'learning_rate': 2.438341870160889e-05, + 'epoch': 2.4} +04/19 [14:46:35] INFO | >> train_qwenlatent.py:487 + Step 9530 | grad_norm_pre_clip=0.2100 | + grad_norm_pre_clip_avg=0.2242 | Metrics: + {'align_loss': 0.025371097028255463, + 'recon_loss': 0.03328702226281166, + 'predict_loss': 0.015072714537382126, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20999301970005035, + 'data_time': 0.0009324790153186768, + 'model_time': 1.216291196004022, + 'grad_norm_pre_clip_avg': 0.22419237345457077, + 'learning_rate': 2.4380710249524494e-05, + 'epoch': 2.4} +04/19 [14:46:48] INFO | >> train_qwenlatent.py:487 + Step 9540 | grad_norm_pre_clip=0.2888 | + grad_norm_pre_clip_avg=0.2413 | Metrics: + {'align_loss': 0.02485903911292553, + 'recon_loss': 0.029445890337228775, + 'predict_loss': 0.014438842423260212, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2887726426124573, + 'data_time': 0.0013541200023610145, + 'model_time': 1.2371915069816168, + 'grad_norm_pre_clip_avg': 0.24128733426332474, + 'learning_rate': 2.4377996013018607e-05, + 'epoch': 2.41} +04/19 [14:47:01] INFO | >> train_qwenlatent.py:487 + Step 9550 | grad_norm_pre_clip=0.2032 | + grad_norm_pre_clip_avg=0.2524 | Metrics: + {'align_loss': 0.025577183812856674, + 'recon_loss': 0.0315990224480629, + 'predict_loss': 0.014095153659582138, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20316198468208313, + 'mae_score': 0.028759851541605083, 'data_time': + 0.0006450589862652123, 'model_time': + 1.232077783002751, 'grad_norm_pre_clip_avg': + 0.25239959806203843, 'learning_rate': + 2.43752759934141e-05, 'epoch': 2.41} +04/19 [14:47:14] INFO | >> train_qwenlatent.py:487 + Step 9560 | grad_norm_pre_clip=0.2456 | + grad_norm_pre_clip_avg=0.2558 | Metrics: + {'align_loss': 0.02452106587588787, + 'recon_loss': 0.029465897008776665, + 'predict_loss': 0.02030731365084648, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2455684095621109, + 'data_time': 0.0008051350014284253, + 'model_time': 1.1857854929985479, + 'grad_norm_pre_clip_avg': 0.2557746320962906, + 'learning_rate': 2.4372550192036685e-05, + 'epoch': 2.41} +04/19 [14:47:27] INFO | >> train_qwenlatent.py:487 + Step 9570 | grad_norm_pre_clip=0.2287 | + grad_norm_pre_clip_avg=0.2177 | Metrics: + {'align_loss': 0.023279830813407898, + 'recon_loss': 0.02849161997437477, + 'predict_loss': 0.013879918493330479, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2286974936723709, + 'data_time': 0.0011968670005444437, + 'model_time': 1.295806112990249, + 'grad_norm_pre_clip_avg': 0.21773693859577178, + 'learning_rate': 2.4369818610214886e-05, + 'epoch': 2.41} +04/19 [14:47:39] INFO | >> train_qwenlatent.py:487 + Step 9580 | grad_norm_pre_clip=0.3013 | + grad_norm_pre_clip_avg=0.3034 | Metrics: + {'align_loss': 0.02377656102180481, + 'recon_loss': 0.027552243322134018, + 'predict_loss': 0.015291278250515461, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3012908399105072, + 'data_time': 0.001072097016731277, + 'model_time': 1.2552265739941504, + 'grad_norm_pre_clip_avg': 0.30338890105485916, + 'learning_rate': 2.4367081249280042e-05, + 'epoch': 2.42} +04/19 [14:47:52] INFO | >> train_qwenlatent.py:487 + Step 9590 | grad_norm_pre_clip=0.2605 | + grad_norm_pre_clip_avg=0.2835 | Metrics: + {'align_loss': 0.022371456027030945, + 'recon_loss': 0.02996649220585823, + 'predict_loss': 0.014964514411985874, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2604743540287018, + 'data_time': 0.0010842970223166049, + 'model_time': 1.2652105599991046, + 'grad_norm_pre_clip_avg': 0.2835277497768402, + 'learning_rate': 2.4364338110566304e-05, + 'epoch': 2.42} +04/19 [14:48:05] INFO | >> train_qwenlatent.py:487 + Step 9600 | grad_norm_pre_clip=0.1961 | + grad_norm_pre_clip_avg=0.2347 | Metrics: + {'align_loss': 0.023589780554175377, + 'recon_loss': 0.023785067722201347, + 'predict_loss': 0.0156896710395813, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19605176150798798, + 'mae_score': 0.022718757766861098, 'data_time': + 0.0006267669959925115, 'model_time': + 1.474122826999519, 'grad_norm_pre_clip_avg': + 0.23471570760011673, 'learning_rate': + 2.4361589195410654e-05, 'epoch': 2.42} +04/19 [14:48:18] INFO | >> train_qwenlatent.py:487 + Step 9610 | grad_norm_pre_clip=0.3022 | + grad_norm_pre_clip_avg=0.2518 | Metrics: + {'align_loss': 0.024085380136966705, + 'recon_loss': 0.025554006919264793, + 'predict_loss': 0.012642391957342625, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.30224236845970154, + 'data_time': 0.0008861630049068481, + 'model_time': 1.189867092994973, + 'grad_norm_pre_clip_avg': 0.25180765986442566, + 'learning_rate': 2.4358834505152866e-05, + 'epoch': 2.42} +04/19 [14:48:31] INFO | >> train_qwenlatent.py:487 + Step 9620 | grad_norm_pre_clip=0.2423 | + grad_norm_pre_clip_avg=0.2903 | Metrics: + {'align_loss': 0.023308176547288895, + 'recon_loss': 0.025015939027071, + 'predict_loss': 0.01334543339908123, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24227097630500793, + 'data_time': 0.001022159995045513, + 'model_time': 1.2635692359763198, + 'grad_norm_pre_clip_avg': 0.29025461375713346, + 'learning_rate': 2.435607404113556e-05, + 'epoch': 2.43} +04/19 [14:48:44] INFO | >> train_qwenlatent.py:487 + Step 9630 | grad_norm_pre_clip=0.2455 | + grad_norm_pre_clip_avg=0.2813 | Metrics: + {'align_loss': 0.025189422070980072, + 'recon_loss': 0.036938607692718506, + 'predict_loss': 0.02068040519952774, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24550670385360718, + 'data_time': 0.000627364992396906, + 'model_time': 1.2246665270067751, + 'grad_norm_pre_clip_avg': 0.2812503963708878, + 'learning_rate': 2.435330780470414e-05, + 'epoch': 2.43} +04/19 [14:48:56] INFO | >> train_qwenlatent.py:487 + Step 9640 | grad_norm_pre_clip=0.1912 | + grad_norm_pre_clip_avg=0.2432 | Metrics: + {'align_loss': 0.023214835673570633, + 'recon_loss': 0.03126073628664017, + 'predict_loss': 0.017142577096819878, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19116345047950745, + 'data_time': 0.0009189029806293547, + 'model_time': 1.1885637490195222, + 'grad_norm_pre_clip_avg': 0.24316228330135345, + 'learning_rate': 2.435053579720684e-05, + 'epoch': 2.43} +04/19 [14:49:09] INFO | >> train_qwenlatent.py:487 + Step 9650 | grad_norm_pre_clip=0.2308 | + grad_norm_pre_clip_avg=0.2496 | Metrics: + {'align_loss': 0.0236869677901268, + 'recon_loss': 0.030916433781385422, + 'predict_loss': 0.014665698632597923, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2308453917503357, + 'mae_score': 0.02569931477039784, 'data_time': + 0.00128376399516128, 'model_time': + 1.2600098450202495, 'grad_norm_pre_clip_avg': + 0.24963978230953215, 'learning_rate': + 2.4347758019994703e-05, 'epoch': 2.44} +04/19 [14:49:22] INFO | >> train_qwenlatent.py:487 + Step 9660 | grad_norm_pre_clip=0.2114 | + grad_norm_pre_clip_avg=0.2541 | Metrics: + {'align_loss': 0.025167005136609077, + 'recon_loss': 0.03585823252797127, + 'predict_loss': 0.014946888200938702, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2113761156797409, + 'data_time': 0.0006748090090695769, + 'model_time': 1.2568408530205488, + 'grad_norm_pre_clip_avg': 0.2541185736656189, + 'learning_rate': 2.4344974474421582e-05, + 'epoch': 2.44} +04/19 [14:49:35] INFO | >> train_qwenlatent.py:487 + Step 9670 | grad_norm_pre_clip=0.3488 | + grad_norm_pre_clip_avg=0.2432 | Metrics: + {'align_loss': 0.02182307466864586, + 'recon_loss': 0.029460707679390907, + 'predict_loss': 0.014434577897191048, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.34877488017082214, + 'data_time': 0.0006623060035053641, + 'model_time': 1.236687553988304, + 'grad_norm_pre_clip_avg': 0.2432372599840164, + 'learning_rate': 2.4342185161844148e-05, + 'epoch': 2.44} +04/19 [14:49:47] INFO | >> train_qwenlatent.py:487 + Step 9680 | grad_norm_pre_clip=0.3086 | + grad_norm_pre_clip_avg=0.3259 | Metrics: + {'align_loss': 0.024955015629529953, + 'recon_loss': 0.0433277003467083, + 'predict_loss': 0.02060328610241413, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.30861467123031616, + 'data_time': 0.0007825999928172678, + 'model_time': 1.2383842470007949, + 'grad_norm_pre_clip_avg': 0.3258798122406006, + 'learning_rate': 2.4339390083621873e-05, + 'epoch': 2.44} +04/19 [14:50:00] INFO | >> train_qwenlatent.py:487 + Step 9690 | grad_norm_pre_clip=0.2411 | + grad_norm_pre_clip_avg=0.2541 | Metrics: + {'align_loss': 0.023435261100530624, + 'recon_loss': 0.031650543212890625, + 'predict_loss': 0.01837029866874218, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24106843769550323, + 'data_time': 0.0011170919751748443, + 'model_time': 1.2305198459944222, + 'grad_norm_pre_clip_avg': 0.25408046692609787, + 'learning_rate': 2.4336589241117042e-05, + 'epoch': 2.45} +04/19 [14:50:14] INFO | >> train_qwenlatent.py:487 + Step 9700 | grad_norm_pre_clip=0.2273 | + grad_norm_pre_clip_avg=0.2268 | Metrics: + {'align_loss': 0.024051643908023834, + 'recon_loss': 0.04247870296239853, + 'predict_loss': 0.019212767481803894, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22732339799404144, + 'mae_score': 0.020964974755639427, 'data_time': + 0.0006927350186742842, 'model_time': + 1.2205918389954604, 'grad_norm_pre_clip_avg': + 0.2268485203385353, 'learning_rate': + 2.433378263569476e-05, 'epoch': 2.45} +04/19 [14:50:26] INFO | >> train_qwenlatent.py:487 + Step 9710 | grad_norm_pre_clip=0.2472 | + grad_norm_pre_clip_avg=0.2486 | Metrics: + {'align_loss': 0.024460088461637497, + 'recon_loss': 0.03788978233933449, + 'predict_loss': 0.02069978415966034, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24715140461921692, + 'data_time': 0.0009044810140039772, + 'model_time': 1.2508449770102743, + 'grad_norm_pre_clip_avg': 0.24860098212957382, + 'learning_rate': 2.4330970268722926e-05, + 'epoch': 2.45} +04/19 [14:50:39] INFO | >> train_qwenlatent.py:487 + Step 9720 | grad_norm_pre_clip=0.1739 | + grad_norm_pre_clip_avg=0.2515 | Metrics: + {'align_loss': 0.02352367341518402, + 'recon_loss': 0.03255503252148628, + 'predict_loss': 0.01623224839568138, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17391102015972137, + 'data_time': 0.0013131970190443099, + 'model_time': 1.2091891129966825, + 'grad_norm_pre_clip_avg': 0.25153774619102476, + 'learning_rate': 2.4328152141572254e-05, + 'epoch': 2.45} +04/19 [14:50:51] INFO | >> train_qwenlatent.py:487 + Step 9730 | grad_norm_pre_clip=0.3105 | + grad_norm_pre_clip_avg=0.2716 | Metrics: + {'align_loss': 0.025358285754919052, + 'recon_loss': 0.04203161969780922, + 'predict_loss': 0.016607066616415977, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3105370104312897, + 'data_time': 0.0008881829853635281, + 'model_time': 1.1995186009735335, + 'grad_norm_pre_clip_avg': 0.2715938091278076, + 'learning_rate': 2.4325328255616274e-05, + 'epoch': 2.46} +04/19 [14:51:04] INFO | >> train_qwenlatent.py:487 + Step 9740 | grad_norm_pre_clip=0.2771 | + grad_norm_pre_clip_avg=0.2567 | Metrics: + {'align_loss': 0.026021212339401245, + 'recon_loss': 0.04778675734996796, + 'predict_loss': 0.023838263005018234, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2771097719669342, + 'data_time': 0.0008999579877126962, + 'model_time': 1.2563317219901364, + 'grad_norm_pre_clip_avg': 0.2566895321011543, + 'learning_rate': 2.43224986122313e-05, 'epoch': + 2.46} +04/19 [14:51:17] INFO | >> train_qwenlatent.py:487 + Step 9750 | grad_norm_pre_clip=0.2023 | + grad_norm_pre_clip_avg=0.2267 | Metrics: + {'align_loss': 0.023308400064706802, + 'recon_loss': 0.02499743178486824, + 'predict_loss': 0.010869814082980156, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2022906243801117, + 'mae_score': 0.030699332984718115, 'data_time': + 0.0007773679972160608, 'model_time': + 1.2771893320023082, 'grad_norm_pre_clip_avg': + 0.22668500542640685, 'learning_rate': + 2.4319663212796475e-05, 'epoch': 2.46} +04/19 [14:51:30] INFO | >> train_qwenlatent.py:487 + Step 9760 | grad_norm_pre_clip=0.1958 | + grad_norm_pre_clip_avg=0.2117 | Metrics: + {'align_loss': 0.02447827160358429, + 'recon_loss': 0.03298095986247063, + 'predict_loss': 0.020939121022820473, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1957564651966095, + 'data_time': 0.0010440119949635118, + 'model_time': 1.2136162059905473, + 'grad_norm_pre_clip_avg': 0.21169284880161285, + 'learning_rate': 2.431682205869373e-05, + 'epoch': 2.46} +04/19 [14:51:43] INFO | >> train_qwenlatent.py:487 + Step 9770 | grad_norm_pre_clip=0.3976 | + grad_norm_pre_clip_avg=0.3489 | Metrics: + {'align_loss': 0.02534886822104454, + 'recon_loss': 0.03490560129284859, + 'predict_loss': 0.01573226973414421, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3975800573825836, + 'data_time': 0.0009349480096716434, + 'model_time': 1.2383405370055698, + 'grad_norm_pre_clip_avg': 0.3489171043038368, + 'learning_rate': 2.4313975151307812e-05, + 'epoch': 2.47} +04/19 [14:51:55] INFO | >> train_qwenlatent.py:487 + Step 9780 | grad_norm_pre_clip=0.2759 | + grad_norm_pre_clip_avg=0.2723 | Metrics: + {'align_loss': 0.023870976641774178, + 'recon_loss': 0.030731908977031708, + 'predict_loss': 0.01712644472718239, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2758924961090088, + 'data_time': 0.0009239410283043981, + 'model_time': 1.1919398159952834, + 'grad_norm_pre_clip_avg': 0.2722892612218857, + 'learning_rate': 2.431112249202628e-05, + 'epoch': 2.47} +04/19 [14:52:08] INFO | >> train_qwenlatent.py:487 + Step 9790 | grad_norm_pre_clip=0.2108 | + grad_norm_pre_clip_avg=0.2636 | Metrics: + {'align_loss': 0.023584138602018356, + 'recon_loss': 0.03725991025567055, + 'predict_loss': 0.014100849628448486, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21078024804592133, + 'data_time': 0.00076232900028117, 'model_time': + 1.2710668189974967, 'grad_norm_pre_clip_avg': + 0.26355467438697816, 'learning_rate': + 2.430826408223947e-05, 'epoch': 2.47} +04/19 [14:52:21] INFO | >> train_qwenlatent.py:487 + Step 9800 | grad_norm_pre_clip=0.2237 | + grad_norm_pre_clip_avg=0.2363 | Metrics: + {'align_loss': 0.024299075827002525, + 'recon_loss': 0.0362236462533474, + 'predict_loss': 0.014446867629885674, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2236671894788742, + 'mae_score': 0.020841416367539414, 'data_time': + 0.0008502139826305211, 'model_time': + 1.179786900000181, 'grad_norm_pre_clip_avg': + 0.23626044541597366, 'learning_rate': + 2.4305399923340543e-05, 'epoch': 2.47} +04/19 [14:52:34] INFO | >> train_qwenlatent.py:487 + Step 9810 | grad_norm_pre_clip=0.1846 | + grad_norm_pre_clip_avg=0.2276 | Metrics: + {'align_loss': 0.02460331842303276, + 'recon_loss': 0.03495360538363457, + 'predict_loss': 0.015995146706700325, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18461734056472778, + 'data_time': 0.0007505910180043429, + 'model_time': 1.2582371199969202, + 'grad_norm_pre_clip_avg': 0.2275933086872101, + 'learning_rate': 2.430253001672546e-05, + 'epoch': 2.48} +04/19 [14:52:46] INFO | >> train_qwenlatent.py:487 + Step 9820 | grad_norm_pre_clip=0.2115 | + grad_norm_pre_clip_avg=0.3087 | Metrics: + {'align_loss': 0.024882938712835312, + 'recon_loss': 0.03129776939749718, + 'predict_loss': 0.013937294483184814, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21152012050151825, + 'data_time': 0.0008891019970178604, + 'model_time': 1.232389275013702, + 'grad_norm_pre_clip_avg': 0.3086537778377533, + 'learning_rate': 2.4299654363792975e-05, + 'epoch': 2.48} +04/19 [14:52:59] INFO | >> train_qwenlatent.py:487 + Step 9830 | grad_norm_pre_clip=0.2979 | + grad_norm_pre_clip_avg=0.3166 | Metrics: + {'align_loss': 0.02519095689058304, + 'recon_loss': 0.046623945236206055, + 'predict_loss': 0.019505202770233154, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.29788148403167725, + 'data_time': 0.0009100129827857018, + 'model_time': 1.2198842510115355, + 'grad_norm_pre_clip_avg': 0.3166047364473343, + 'learning_rate': 2.4296772965944638e-05, + 'epoch': 2.48} +04/19 [14:53:12] INFO | >> train_qwenlatent.py:487 + Step 9840 | grad_norm_pre_clip=0.2673 | + grad_norm_pre_clip_avg=0.2698 | Metrics: + {'align_loss': 0.025489209219813347, + 'recon_loss': 0.02398122474551201, + 'predict_loss': 0.01083798985928297, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2673267424106598, + 'data_time': 0.0006792339845560491, + 'model_time': 1.19895314599853, + 'grad_norm_pre_clip_avg': 0.26976598501205445, + 'learning_rate': 2.429388582458482e-05, + 'epoch': 2.48} +04/19 [14:53:25] INFO | >> train_qwenlatent.py:487 + Step 9850 | grad_norm_pre_clip=0.2107 | + grad_norm_pre_clip_avg=0.2362 | Metrics: + {'align_loss': 0.023854374885559082, + 'recon_loss': 0.030026616528630257, + 'predict_loss': 0.014753290452063084, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2107001692056656, + 'mae_score': 0.03526419218596037, 'data_time': + 0.0007363469921983778, 'model_time': + 1.2407871760078706, 'grad_norm_pre_clip_avg': + 0.23616168648004532, 'learning_rate': + 2.4290992941120665e-05, 'epoch': 2.49} +04/19 [14:53:38] INFO | >> train_qwenlatent.py:487 + Step 9860 | grad_norm_pre_clip=0.2178 | + grad_norm_pre_clip_avg=0.2168 | Metrics: + {'align_loss': 0.02358262985944748, + 'recon_loss': 0.02470647543668747, + 'predict_loss': 0.012588901445269585, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21781708300113678, + 'data_time': 0.0007125340052880347, + 'model_time': 1.2158410459815059, + 'grad_norm_pre_clip_avg': 0.21683310866355895, + 'learning_rate': 2.4288094316962147e-05, + 'epoch': 2.49} +04/19 [14:53:50] INFO | >> train_qwenlatent.py:487 + Step 9870 | grad_norm_pre_clip=0.2686 | + grad_norm_pre_clip_avg=0.2463 | Metrics: + {'align_loss': 0.024695076048374176, + 'recon_loss': 0.031265903264284134, + 'predict_loss': 0.013652374036610126, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2685699760913849, + 'data_time': 0.0008504809811711311, + 'model_time': 1.254469670006074, + 'grad_norm_pre_clip_avg': 0.24629041850566863, + 'learning_rate': 2.4285189953522006e-05, + 'epoch': 2.49} +04/19 [14:54:03] INFO | >> train_qwenlatent.py:487 + Step 9880 | grad_norm_pre_clip=0.3087 | + grad_norm_pre_clip_avg=0.2728 | Metrics: + {'align_loss': 0.02540975995361805, + 'recon_loss': 0.0575428269803524, + 'predict_loss': 0.026423893868923187, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.30865904688835144, + 'data_time': 0.0010164800041820854, + 'model_time': 1.2274377620196901, + 'grad_norm_pre_clip_avg': 0.2727771207690239, + 'learning_rate': 2.42822798522158e-05, 'epoch': + 2.49} +04/19 [14:54:16] INFO | >> train_qwenlatent.py:487 + Step 9890 | grad_norm_pre_clip=0.2052 | + grad_norm_pre_clip_avg=0.2945 | Metrics: + {'align_loss': 0.025411318987607956, + 'recon_loss': 0.03971811756491661, + 'predict_loss': 0.01704491674900055, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20519965887069702, + 'data_time': 0.0008227859798353165, + 'model_time': 1.2589946090010926, + 'grad_norm_pre_clip_avg': 0.2944591298699379, + 'learning_rate': 2.427936401446187e-05, + 'epoch': 2.5} +04/19 [14:54:29] INFO | >> train_qwenlatent.py:487 + Step 9900 | grad_norm_pre_clip=0.2284 | + grad_norm_pre_clip_avg=0.2504 | Metrics: + {'align_loss': 0.022540833801031113, + 'recon_loss': 0.032588064670562744, + 'predict_loss': 0.016397669911384583, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22842082381248474, + 'mae_score': 0.020092589575965124, 'data_time': + 0.0010048339900095016, 'model_time': + 1.2654086620023008, 'grad_norm_pre_clip_avg': + 0.25044348239898684, 'learning_rate': + 2.4276442441681368e-05, 'epoch': 2.5} +04/19 [14:54:41] INFO | >> train_qwenlatent.py:487 + Step 9910 | grad_norm_pre_clip=0.2348 | + grad_norm_pre_clip_avg=0.2238 | Metrics: + {'align_loss': 0.02416912280023098, + 'recon_loss': 0.03177613392472267, + 'predict_loss': 0.013553043827414513, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23476429283618927, + 'data_time': 0.0006790360203012824, + 'model_time': 1.245421248022467, + 'grad_norm_pre_clip_avg': 0.2237989455461502, + 'learning_rate': 2.427351513529823e-05, + 'epoch': 2.5} +04/19 [14:54:54] INFO | >> train_qwenlatent.py:487 + Step 9920 | grad_norm_pre_clip=0.2397 | + grad_norm_pre_clip_avg=0.2495 | Metrics: + {'align_loss': 0.02401026152074337, + 'recon_loss': 0.03327976167201996, + 'predict_loss': 0.014570054598152637, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23966822028160095, + 'data_time': 0.001147311006207019, + 'model_time': 1.2258530620019883, + 'grad_norm_pre_clip_avg': 0.24945217818021775, + 'learning_rate': 2.4270582096739187e-05, + 'epoch': 2.5} +04/19 [14:55:06] INFO | >> train_qwenlatent.py:487 + Step 9930 | grad_norm_pre_clip=0.2319 | + grad_norm_pre_clip_avg=0.2754 | Metrics: + {'align_loss': 0.02520071156322956, + 'recon_loss': 0.04029585421085358, + 'predict_loss': 0.016855813562870026, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23190975189208984, + 'data_time': 0.0007398050220217556, + 'model_time': 1.259298802993726, + 'grad_norm_pre_clip_avg': 0.2753810718655586, + 'learning_rate': 2.4267643327433775e-05, + 'epoch': 2.51} +04/19 [14:55:19] INFO | >> train_qwenlatent.py:487 + Step 9940 | grad_norm_pre_clip=0.2896 | + grad_norm_pre_clip_avg=0.2686 | Metrics: + {'align_loss': 0.024872560054063797, + 'recon_loss': 0.03140975907444954, + 'predict_loss': 0.010710437782108784, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2895732820034027, + 'data_time': 0.0008898309897631407, + 'model_time': 1.24704947398277, + 'grad_norm_pre_clip_avg': 0.26861556619405746, + 'learning_rate': 2.42646988288143e-05, 'epoch': + 2.51} +04/19 [14:55:32] INFO | >> train_qwenlatent.py:487 + Step 9950 | grad_norm_pre_clip=0.2852 | + grad_norm_pre_clip_avg=0.2258 | Metrics: + {'align_loss': 0.02350122109055519, + 'recon_loss': 0.038736362010240555, + 'predict_loss': 0.0229139756411314, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28523629903793335, + 'mae_score': 0.020837862856753236, 'data_time': + 0.0008768500119913369, 'model_time': + 1.2058743050147314, 'grad_norm_pre_clip_avg': + 0.22582916170358658, 'learning_rate': + 2.4261748602315893e-05, 'epoch': 2.51} +04/19 [14:55:45] INFO | >> train_qwenlatent.py:487 + Step 9960 | grad_norm_pre_clip=0.2970 | + grad_norm_pre_clip_avg=0.2775 | Metrics: + {'align_loss': 0.024949459359049797, + 'recon_loss': 0.035050153732299805, + 'predict_loss': 0.016445878893136978, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2969716489315033, + 'data_time': 0.0010124559921678156, + 'model_time': 1.2463872520020232, + 'grad_norm_pre_clip_avg': 0.27745630741119387, + 'learning_rate': 2.425879264937644e-05, + 'epoch': 2.51} +04/19 [14:55:58] INFO | >> train_qwenlatent.py:487 + Step 9970 | grad_norm_pre_clip=0.2619 | + grad_norm_pre_clip_avg=0.2322 | Metrics: + {'align_loss': 0.024885080754756927, + 'recon_loss': 0.04017682373523712, + 'predict_loss': 0.018613144755363464, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26192453503608704, + 'data_time': 0.000909431983018294, + 'model_time': 1.212794015998952, + 'grad_norm_pre_clip_avg': 0.23224110752344132, + 'learning_rate': 2.425583097143665e-05, + 'epoch': 2.52} +04/19 [14:56:10] INFO | >> train_qwenlatent.py:487 + Step 9980 | grad_norm_pre_clip=0.2426 | + grad_norm_pre_clip_avg=0.2713 | Metrics: + {'align_loss': 0.02496485784649849, + 'recon_loss': 0.030555732548236847, + 'predict_loss': 0.012387160211801529, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24259978532791138, + 'data_time': 0.0007988040160853416, + 'model_time': 1.2488906820071861, + 'grad_norm_pre_clip_avg': 0.2712766110897064, + 'learning_rate': 2.4252863569940005e-05, + 'epoch': 2.52} +04/19 [14:56:23] INFO | >> train_qwenlatent.py:487 + Step 9990 | grad_norm_pre_clip=0.2418 | + grad_norm_pre_clip_avg=0.2389 | Metrics: + {'align_loss': 0.02415199764072895, + 'recon_loss': 0.033957138657569885, + 'predict_loss': 0.016445716843008995, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24184486269950867, + 'data_time': 0.00116830799379386, 'model_time': + 1.3111853929876816, 'grad_norm_pre_clip_avg': + 0.23887853771448136, 'learning_rate': + 2.4249890446332776e-05, 'epoch': 2.52} +04/19 [14:56:36] INFO | >> train_qwenlatent.py:487 + Step 10000 | grad_norm_pre_clip=0.2692 | + grad_norm_pre_clip_avg=0.2492 | Metrics: + {'align_loss': 0.02503022737801075, + 'recon_loss': 0.03159303218126297, + 'predict_loss': 0.010080410167574883, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26918986439704895, + 'mae_score': 0.019601815026085656, 'data_time': + 0.0007339880103245378, 'model_time': + 1.2538208370096982, 'grad_norm_pre_clip_avg': + 0.24924449473619462, 'learning_rate': + 2.4246911602064033e-05, 'epoch': 2.52} +✅ Checkpoint saved at ./runs/0418_QwenLatent_13tasks_actionstate_30k/checkpoints/steps_10000 +04/19 [14:56:59] INFO | >> train_qwenlatent.py:487 + Step 10010 | grad_norm_pre_clip=0.3199 | + grad_norm_pre_clip_avg=0.2677 | Metrics: + {'align_loss': 0.02432882785797119, + 'recon_loss': 0.0336504690349102, + 'predict_loss': 0.015954332426190376, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.31991085410118103, + 'data_time': 0.0008690520189702511, + 'model_time': 1.289568951993715, + 'grad_norm_pre_clip_avg': 0.2677129402756691, + 'learning_rate': 2.4243927038585624e-05, + 'epoch': 2.53} +04/19 [14:57:12] INFO | >> train_qwenlatent.py:487 + Step 10020 | grad_norm_pre_clip=0.2020 | + grad_norm_pre_clip_avg=0.2649 | Metrics: + {'align_loss': 0.02384801022708416, + 'recon_loss': 0.03706233575940132, + 'predict_loss': 0.016741987317800522, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20200252532958984, + 'data_time': 0.0011465169955044985, + 'model_time': 1.361193665972678, + 'grad_norm_pre_clip_avg': 0.26492740213871, + 'learning_rate': 2.424093675735219e-05, + 'epoch': 2.53} +04/19 [14:57:25] INFO | >> train_qwenlatent.py:487 + Step 10030 | grad_norm_pre_clip=0.2085 | + grad_norm_pre_clip_avg=0.2738 | Metrics: + {'align_loss': 0.024601668119430542, + 'recon_loss': 0.02765258029103279, + 'predict_loss': 0.013284502550959587, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2084825187921524, + 'data_time': 0.0007374839915428311, + 'model_time': 1.1960059870034456, + 'grad_norm_pre_clip_avg': 0.2737694948911667, + 'learning_rate': 2.4237940759821168e-05, + 'epoch': 2.53} +04/19 [14:57:37] INFO | >> train_qwenlatent.py:487 + Step 10040 | grad_norm_pre_clip=0.2973 | + grad_norm_pre_clip_avg=0.2532 | Metrics: + {'align_loss': 0.024585478007793427, + 'recon_loss': 0.03613584488630295, + 'predict_loss': 0.01693597063422203, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.29733970761299133, + 'data_time': 0.0011402119998820126, + 'model_time': 1.2444194099807646, + 'grad_norm_pre_clip_avg': 0.2532499223947525, + 'learning_rate': 2.4234939047452755e-05, + 'epoch': 2.53} +04/19 [14:57:50] INFO | >> train_qwenlatent.py:487 + Step 10050 | grad_norm_pre_clip=0.2269 | + grad_norm_pre_clip_avg=0.2732 | Metrics: + {'align_loss': 0.023427650332450867, + 'recon_loss': 0.023050528019666672, + 'predict_loss': 0.011869446374475956, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2268659621477127, + 'mae_score': 0.019682738158079954, 'data_time': + 0.0007137089851312339, 'model_time': + 1.2442381089786068, 'grad_norm_pre_clip_avg': + 0.27320430278778074, 'learning_rate': + 2.4231931621709954e-05, 'epoch': 2.54} +04/19 [14:58:03] INFO | >> train_qwenlatent.py:487 + Step 10060 | grad_norm_pre_clip=0.2488 | + grad_norm_pre_clip_avg=0.2468 | Metrics: + {'align_loss': 0.024439387023448944, + 'recon_loss': 0.04110785201191902, + 'predict_loss': 0.01759987883269787, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24875998497009277, + 'data_time': 0.0006470149965025485, + 'model_time': 1.2303355839976575, + 'grad_norm_pre_clip_avg': 0.24681167155504227, + 'learning_rate': 2.4228918484058554e-05, + 'epoch': 2.54} +04/19 [14:58:16] INFO | >> train_qwenlatent.py:487 + Step 10070 | grad_norm_pre_clip=0.2164 | + grad_norm_pre_clip_avg=0.2153 | Metrics: + {'align_loss': 0.024875061586499214, + 'recon_loss': 0.033739037811756134, + 'predict_loss': 0.01473922748118639, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21636393666267395, + 'data_time': 0.0008218309958465397, + 'model_time': 1.2281900040106848, + 'grad_norm_pre_clip_avg': 0.21531413942575456, + 'learning_rate': 2.422589963596712e-05, + 'epoch': 2.54} +04/19 [14:58:28] INFO | >> train_qwenlatent.py:487 + Step 10080 | grad_norm_pre_clip=0.3172 | + grad_norm_pre_clip_avg=0.3223 | Metrics: + {'align_loss': 0.024289343506097794, + 'recon_loss': 0.03843088448047638, + 'predict_loss': 0.015352414920926094, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3172447085380554, + 'data_time': 0.0012227339902892709, + 'model_time': 1.2371877569821663, + 'grad_norm_pre_clip_avg': 0.3223257690668106, + 'learning_rate': 2.4222875078906994e-05, + 'epoch': 2.54} +04/19 [14:58:41] INFO | >> train_qwenlatent.py:487 + Step 10090 | grad_norm_pre_clip=0.2288 | + grad_norm_pre_clip_avg=0.2736 | Metrics: + {'align_loss': 0.02467060275375843, + 'recon_loss': 0.032935142517089844, + 'predict_loss': 0.01130841113626957, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22883807122707367, + 'data_time': 0.0009952149994205683, + 'model_time': 1.219787667010678, + 'grad_norm_pre_clip_avg': 0.273603244125843, + 'learning_rate': 2.4219844814352317e-05, + 'epoch': 2.55} +04/19 [14:58:54] INFO | >> train_qwenlatent.py:487 + Step 10100 | grad_norm_pre_clip=0.2041 | + grad_norm_pre_clip_avg=0.2417 | Metrics: + {'align_loss': 0.02467951737344265, + 'recon_loss': 0.0295268427580595, + 'predict_loss': 0.014561520889401436, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2040511667728424, + 'mae_score': 0.021116328883815455, 'data_time': + 0.0006392470095306635, 'model_time': + 1.2139825469930656, 'grad_norm_pre_clip_avg': + 0.24174403995275498, 'learning_rate': + 2.421680884378e-05, 'epoch': 2.55} +04/19 [14:59:07] INFO | >> train_qwenlatent.py:487 + Step 10110 | grad_norm_pre_clip=0.2093 | + grad_norm_pre_clip_avg=0.2404 | Metrics: + {'align_loss': 0.02395591326057911, + 'recon_loss': 0.026524638757109642, + 'predict_loss': 0.010648414492607117, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20933902263641357, + 'data_time': 0.0006926729984115809, + 'model_time': 1.2321797030162998, + 'grad_norm_pre_clip_avg': 0.24038330167531968, + 'learning_rate': 2.4213767168669733e-05, + 'epoch': 2.55} +04/19 [14:59:19] INFO | >> train_qwenlatent.py:487 + Step 10120 | grad_norm_pre_clip=0.2850 | + grad_norm_pre_clip_avg=0.2703 | Metrics: + {'align_loss': 0.025031715631484985, + 'recon_loss': 0.04256792739033699, + 'predict_loss': 0.01891828328371048, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2850249409675598, + 'data_time': 0.0007031620189081877, + 'model_time': 1.211579241004074, + 'grad_norm_pre_clip_avg': 0.2703396052122116, + 'learning_rate': 2.4210719790503998e-05, + 'epoch': 2.55} +04/19 [14:59:32] INFO | >> train_qwenlatent.py:487 + Step 10130 | grad_norm_pre_clip=0.2061 | + grad_norm_pre_clip_avg=0.2742 | Metrics: + {'align_loss': 0.02445138618350029, + 'recon_loss': 0.03140940144658089, + 'predict_loss': 0.013201571069657803, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20607911050319672, + 'data_time': 0.0007117200002539903, + 'model_time': 1.2245109619980212, + 'grad_norm_pre_clip_avg': 0.27419801503419877, + 'learning_rate': 2.4207666710768052e-05, + 'epoch': 2.56} +04/19 [14:59:44] INFO | >> train_qwenlatent.py:487 + Step 10140 | grad_norm_pre_clip=0.2639 | + grad_norm_pre_clip_avg=0.2571 | Metrics: + {'align_loss': 0.024751514196395874, + 'recon_loss': 0.042115092277526855, + 'predict_loss': 0.022462621331214905, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26390278339385986, + 'data_time': 0.0008090249903034419, + 'model_time': 1.2363944549870212, + 'grad_norm_pre_clip_avg': 0.25708961188793183, + 'learning_rate': 2.4204607930949925e-05, + 'epoch': 2.56} +04/19 [14:59:57] INFO | >> train_qwenlatent.py:487 + Step 10150 | grad_norm_pre_clip=0.2330 | + grad_norm_pre_clip_avg=0.2248 | Metrics: + {'align_loss': 0.024456407874822617, + 'recon_loss': 0.04820427671074867, + 'predict_loss': 0.024722225964069366, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23299141228199005, + 'mae_score': 0.021251004880613034, 'data_time': + 0.0008312069985549897, 'model_time': + 1.236455729987938, 'grad_norm_pre_clip_avg': + 0.22475602030754088, 'learning_rate': + 2.4201543452540433e-05, 'epoch': 2.56} +04/19 [15:00:10] INFO | >> train_qwenlatent.py:487 + Step 10160 | grad_norm_pre_clip=0.2493 | + grad_norm_pre_clip_avg=0.2874 | Metrics: + {'align_loss': 0.02422134205698967, + 'recon_loss': 0.052193060517311096, + 'predict_loss': 0.022760923951864243, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2492939531803131, + 'data_time': 0.001084638002794236, + 'model_time': 1.5905416180030443, + 'grad_norm_pre_clip_avg': 0.2874373808503151, + 'learning_rate': 2.4198473277033155e-05, + 'epoch': 2.56} +04/19 [15:00:23] INFO | >> train_qwenlatent.py:487 + Step 10170 | grad_norm_pre_clip=0.2739 | + grad_norm_pre_clip_avg=0.2828 | Metrics: + {'align_loss': 0.025448286905884743, + 'recon_loss': 0.03564990684390068, + 'predict_loss': 0.015389171428978443, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.27390721440315247, + 'data_time': 0.0006867460033390671, + 'model_time': 1.259105605975492, + 'grad_norm_pre_clip_avg': 0.2827798679471016, + 'learning_rate': 2.419539740592447e-05, + 'epoch': 2.57} +04/19 [15:00:35] INFO | >> train_qwenlatent.py:487 + Step 10180 | grad_norm_pre_clip=0.2518 | + grad_norm_pre_clip_avg=0.2386 | Metrics: + {'align_loss': 0.02622070163488388, + 'recon_loss': 0.0418822206556797, + 'predict_loss': 0.01731013134121895, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2518366277217865, + 'data_time': 0.0008044429996516556, + 'model_time': 1.2092268279811833, + 'grad_norm_pre_clip_avg': 0.23862901628017424, + 'learning_rate': 2.4192315840713512e-05, + 'epoch': 2.57} +04/19 [15:00:48] INFO | >> train_qwenlatent.py:487 + Step 10190 | grad_norm_pre_clip=0.2331 | + grad_norm_pre_clip_avg=0.2350 | Metrics: + {'align_loss': 0.024282097816467285, + 'recon_loss': 0.032747212797403336, + 'predict_loss': 0.016027187928557396, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2331044226884842, + 'data_time': 0.0006927049835212529, + 'model_time': 1.2302318010188174, + 'grad_norm_pre_clip_avg': 0.23497643768787385, + 'learning_rate': 2.41892285829022e-05, 'epoch': + 2.57} +04/19 [15:01:01] INFO | >> train_qwenlatent.py:487 + Step 10200 | grad_norm_pre_clip=0.2581 | + grad_norm_pre_clip_avg=0.2574 | Metrics: + {'align_loss': 0.02240503579378128, + 'recon_loss': 0.027788270264863968, + 'predict_loss': 0.01367219164967537, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25810638070106506, + 'mae_score': 0.015355173746744792, 'data_time': + 0.0006646349793300033, 'model_time': + 1.2605054300220218, 'grad_norm_pre_clip_avg': + 0.25737595856189727, 'learning_rate': + 2.418613563399523e-05, 'epoch': 2.57} +04/19 [15:01:14] INFO | >> train_qwenlatent.py:487 + Step 10210 | grad_norm_pre_clip=0.2310 | + grad_norm_pre_clip_avg=0.2276 | Metrics: + {'align_loss': 0.023538745939731598, + 'recon_loss': 0.038796164095401764, + 'predict_loss': 0.01559031568467617, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23102021217346191, + 'data_time': 0.0009056479902938008, + 'model_time': 1.1932563839945942, + 'grad_norm_pre_clip_avg': 0.22755435854196548, + 'learning_rate': 2.4183036995500066e-05, + 'epoch': 2.58} +04/19 [15:01:26] INFO | >> train_qwenlatent.py:487 + Step 10220 | grad_norm_pre_clip=0.2127 | + grad_norm_pre_clip_avg=0.2411 | Metrics: + {'align_loss': 0.02461615949869156, + 'recon_loss': 0.03468479961156845, + 'predict_loss': 0.014785725623369217, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21266970038414001, + 'data_time': 0.0008141200232785195, + 'model_time': 1.2433681240072474, + 'grad_norm_pre_clip_avg': 0.24108877331018447, + 'learning_rate': 2.4179932668926943e-05, + 'epoch': 2.58} +04/19 [15:01:39] INFO | >> train_qwenlatent.py:487 + Step 10230 | grad_norm_pre_clip=0.1984 | + grad_norm_pre_clip_avg=0.2626 | Metrics: + {'align_loss': 0.02441166341304779, + 'recon_loss': 0.030911613255739212, + 'predict_loss': 0.01740594208240509, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1984029859304428, + 'data_time': 0.0009141209884546697, + 'model_time': 1.1878990779805463, + 'grad_norm_pre_clip_avg': 0.26255430579185485, + 'learning_rate': 2.4176822655788878e-05, + 'epoch': 2.58} +04/19 [15:01:52] INFO | >> train_qwenlatent.py:487 + Step 10240 | grad_norm_pre_clip=0.3513 | + grad_norm_pre_clip_avg=0.2734 | Metrics: + {'align_loss': 0.025130145251750946, + 'recon_loss': 0.03528749942779541, + 'predict_loss': 0.012254957109689713, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3512656092643738, + 'data_time': 0.000649747991701588, + 'model_time': 1.236615425994387, + 'grad_norm_pre_clip_avg': 0.2733514219522476, + 'learning_rate': 2.417370695760165e-05, + 'epoch': 2.58} +04/19 [15:02:05] INFO | >> train_qwenlatent.py:487 + Step 10250 | grad_norm_pre_clip=0.2516 | + grad_norm_pre_clip_avg=0.2592 | Metrics: + {'align_loss': 0.024587083607912064, + 'recon_loss': 0.032088689506053925, + 'predict_loss': 0.01198950782418251, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25157690048217773, + 'mae_score': 0.028228134292740004, 'data_time': + 0.000879767001606524, 'model_time': + 1.3227727779885754, 'grad_norm_pre_clip_avg': + 0.2592315196990967, 'learning_rate': + 2.4170585575883807e-05, 'epoch': 2.59} +04/19 [15:02:17] INFO | >> train_qwenlatent.py:487 + Step 10260 | grad_norm_pre_clip=0.2133 | + grad_norm_pre_clip_avg=0.2468 | Metrics: + {'align_loss': 0.02503606304526329, + 'recon_loss': 0.036623794585466385, + 'predict_loss': 0.012407422997057438, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21326448023319244, + 'data_time': 0.0009448239870835096, + 'model_time': 1.229784279013984, + 'grad_norm_pre_clip_avg': 0.24682178646326064, + 'learning_rate': 2.4167458512156683e-05, + 'epoch': 2.59} +04/19 [15:02:30] INFO | >> train_qwenlatent.py:487 + Step 10270 | grad_norm_pre_clip=0.1946 | + grad_norm_pre_clip_avg=0.2572 | Metrics: + {'align_loss': 0.02422594279050827, + 'recon_loss': 0.050909917801618576, + 'predict_loss': 0.023613693192601204, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19455666840076447, + 'data_time': 0.000957750016823411, + 'model_time': 1.2895181539934129, + 'grad_norm_pre_clip_avg': 0.25718857944011686, + 'learning_rate': 2.4164325767944366e-05, + 'epoch': 2.59} +04/19 [15:02:43] INFO | >> train_qwenlatent.py:487 + Step 10280 | grad_norm_pre_clip=0.2341 | + grad_norm_pre_clip_avg=0.2371 | Metrics: + {'align_loss': 0.02337631583213806, + 'recon_loss': 0.03760508820414543, + 'predict_loss': 0.014790571294724941, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2340938299894333, + 'data_time': 0.0007053790031932294, + 'model_time': 1.244673485023668, + 'grad_norm_pre_clip_avg': 0.23707854449748994, + 'learning_rate': 2.4161187344773715e-05, + 'epoch': 2.59} +04/19 [15:02:55] INFO | >> train_qwenlatent.py:487 + Step 10290 | grad_norm_pre_clip=0.2620 | + grad_norm_pre_clip_avg=0.2225 | Metrics: + {'align_loss': 0.025973908603191376, + 'recon_loss': 0.039315227419137955, + 'predict_loss': 0.01564096286892891, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26203206181526184, + 'data_time': 0.0006756009825039655, + 'model_time': 1.2205027400050312, + 'grad_norm_pre_clip_avg': 0.22248944342136384, + 'learning_rate': 2.4158043244174366e-05, + 'epoch': 2.6} +04/19 [15:03:09] INFO | >> train_qwenlatent.py:487 + Step 10300 | grad_norm_pre_clip=0.2411 | + grad_norm_pre_clip_avg=0.2498 | Metrics: + {'align_loss': 0.024821888655424118, + 'recon_loss': 0.04333782196044922, + 'predict_loss': 0.019054578617215157, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24112272262573242, + 'mae_score': 0.019708856806024775, 'data_time': + 0.0008285280200652778, 'model_time': + 1.281930981989717, 'grad_norm_pre_clip_avg': + 0.24976081252098084, 'learning_rate': + 2.415489346767871e-05, 'epoch': 2.6} +04/19 [15:03:22] INFO | >> train_qwenlatent.py:487 + Step 10310 | grad_norm_pre_clip=0.3451 | + grad_norm_pre_clip_avg=0.2979 | Metrics: + {'align_loss': 0.024489812552928925, + 'recon_loss': 0.04412456974387169, + 'predict_loss': 0.019798196852207184, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3450927436351776, + 'data_time': 0.0007212620112113655, + 'model_time': 1.2714872820070013, + 'grad_norm_pre_clip_avg': 0.29788938164711, + 'learning_rate': 2.415173801682191e-05, + 'epoch': 2.6} +04/19 [15:03:34] INFO | >> train_qwenlatent.py:487 + Step 10320 | grad_norm_pre_clip=0.2818 | + grad_norm_pre_clip_avg=0.2865 | Metrics: + {'align_loss': 0.025586191564798355, + 'recon_loss': 0.05359995365142822, + 'predict_loss': 0.01893279142677784, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2817813456058502, + 'data_time': 0.0007206809823401272, + 'model_time': 1.280531233001966, + 'grad_norm_pre_clip_avg': 0.2865273967385292, + 'learning_rate': 2.4148576893141896e-05, + 'epoch': 2.6} +04/19 [15:03:47] INFO | >> train_qwenlatent.py:487 + Step 10330 | grad_norm_pre_clip=0.2358 | + grad_norm_pre_clip_avg=0.2631 | Metrics: + {'align_loss': 0.02443556860089302, + 'recon_loss': 0.03198230639100075, + 'predict_loss': 0.013712103478610516, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23582299053668976, + 'data_time': 0.0007029219996184111, + 'model_time': 1.242053579975618, + 'grad_norm_pre_clip_avg': 0.2630942016839981, + 'learning_rate': 2.4145410098179364e-05, + 'epoch': 2.61} +04/19 [15:04:00] INFO | >> train_qwenlatent.py:487 + Step 10340 | grad_norm_pre_clip=0.2133 | + grad_norm_pre_clip_avg=0.1961 | Metrics: + {'align_loss': 0.02474292367696762, + 'recon_loss': 0.035100918263196945, + 'predict_loss': 0.014590430073440075, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.213276669383049, + 'data_time': 0.0007900070049799979, + 'model_time': 1.227827516006073, + 'grad_norm_pre_clip_avg': 0.19610145688056946, + 'learning_rate': 2.4142237633477766e-05, + 'epoch': 2.61} +04/19 [15:04:13] INFO | >> train_qwenlatent.py:487 + Step 10350 | grad_norm_pre_clip=0.3974 | + grad_norm_pre_clip_avg=0.2365 | Metrics: + {'align_loss': 0.02418138086795807, + 'recon_loss': 0.04377609118819237, + 'predict_loss': 0.019348006695508957, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3974461555480957, + 'mae_score': 0.028305100105904243, 'data_time': + 0.0008572279766667634, 'model_time': + 1.2177292429842055, 'grad_norm_pre_clip_avg': + 0.23645037412643433, 'learning_rate': + 2.4139059500583324e-05, 'epoch': 2.61} +04/19 [15:04:26] INFO | >> train_qwenlatent.py:487 + Step 10360 | grad_norm_pre_clip=0.2440 | + grad_norm_pre_clip_avg=0.3067 | Metrics: + {'align_loss': 0.023811539635062218, + 'recon_loss': 0.03240465372800827, + 'predict_loss': 0.012706499546766281, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2440415918827057, + 'data_time': 0.0008593859965912998, + 'model_time': 1.2702099180023652, + 'grad_norm_pre_clip_avg': 0.30666531473398206, + 'learning_rate': 2.4135875701045022e-05, + 'epoch': 2.61} +04/19 [15:04:38] INFO | >> train_qwenlatent.py:487 + Step 10370 | grad_norm_pre_clip=0.2069 | + grad_norm_pre_clip_avg=0.2285 | Metrics: + {'align_loss': 0.024061087518930435, + 'recon_loss': 0.03712337091565132, + 'predict_loss': 0.01768326945602894, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2069079875946045, + 'data_time': 0.0007180070097092539, + 'model_time': 1.2041276289965026, + 'grad_norm_pre_clip_avg': 0.2285121574997902, + 'learning_rate': 2.413268623641461e-05, + 'epoch': 2.62} +04/19 [15:04:51] INFO | >> train_qwenlatent.py:487 + Step 10380 | grad_norm_pre_clip=0.2935 | + grad_norm_pre_clip_avg=0.2237 | Metrics: + {'align_loss': 0.024487104266881943, + 'recon_loss': 0.03769950568675995, + 'predict_loss': 0.015810439363121986, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.29350656270980835, + 'data_time': 0.0009535790013615042, + 'model_time': 1.2028245190158486, + 'grad_norm_pre_clip_avg': 0.22372636646032334, + 'learning_rate': 2.4129491108246584e-05, + 'epoch': 2.62} +04/19 [15:05:03] INFO | >> train_qwenlatent.py:487 + Step 10390 | grad_norm_pre_clip=0.2467 | + grad_norm_pre_clip_avg=0.2480 | Metrics: + {'align_loss': 0.024501116946339607, + 'recon_loss': 0.044527072459459305, + 'predict_loss': 0.017135821282863617, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24674277007579803, + 'data_time': 0.0006345679867081344, + 'model_time': 1.2328723020036705, + 'grad_norm_pre_clip_avg': 0.24801337271928786, + 'learning_rate': 2.4126290318098218e-05, + 'epoch': 2.62} +04/19 [15:05:17] INFO | >> train_qwenlatent.py:487 + Step 10400 | grad_norm_pre_clip=0.2344 | + grad_norm_pre_clip_avg=0.2520 | Metrics: + {'align_loss': 0.024096254259347916, + 'recon_loss': 0.04612777382135391, + 'predict_loss': 0.01864536479115486, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23444901406764984, + 'mae_score': 0.016795594842584284, 'data_time': + 0.0006881270091980696, 'model_time': + 1.2513068430125713, 'grad_norm_pre_clip_avg': + 0.25197564959526064, 'learning_rate': + 2.4123083867529538e-05, 'epoch': 2.62} +04/19 [15:05:29] INFO | >> train_qwenlatent.py:487 + Step 10410 | grad_norm_pre_clip=0.2139 | + grad_norm_pre_clip_avg=0.2132 | Metrics: + {'align_loss': 0.025073174387216568, + 'recon_loss': 0.03563733398914337, + 'predict_loss': 0.012020409107208252, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21391169726848602, + 'data_time': 0.0009840330167207867, + 'model_time': 1.285461680003209, + 'grad_norm_pre_clip_avg': 0.21322436332702638, + 'learning_rate': 2.411987175810333e-05, + 'epoch': 2.63} +04/19 [15:05:42] INFO | >> train_qwenlatent.py:487 + Step 10420 | grad_norm_pre_clip=0.3387 | + grad_norm_pre_clip_avg=0.2740 | Metrics: + {'align_loss': 0.025349846109747887, + 'recon_loss': 0.05129045993089676, + 'predict_loss': 0.020628349855542183, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.33873602747917175, + 'data_time': 0.0009868979977909476, + 'model_time': 1.3037097449996509, + 'grad_norm_pre_clip_avg': 0.2740382343530655, + 'learning_rate': 2.4116653991385136e-05, + 'epoch': 2.63} +04/19 [15:05:55] INFO | >> train_qwenlatent.py:487 + Step 10430 | grad_norm_pre_clip=0.2027 | + grad_norm_pre_clip_avg=0.2647 | Metrics: + {'align_loss': 0.024998700246214867, + 'recon_loss': 0.04115865379571915, + 'predict_loss': 0.016622785478830338, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20269833505153656, + 'data_time': 0.0009335619979538023, + 'model_time': 1.2132970550155733, + 'grad_norm_pre_clip_avg': 0.26465426087379457, + 'learning_rate': 2.4113430568943253e-05, + 'epoch': 2.63} +04/19 [15:06:08] INFO | >> train_qwenlatent.py:487 + Step 10440 | grad_norm_pre_clip=0.1884 | + grad_norm_pre_clip_avg=0.2617 | Metrics: + {'align_loss': 0.025325939059257507, + 'recon_loss': 0.03787156566977501, + 'predict_loss': 0.014125104062259197, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18842850625514984, + 'data_time': 0.0010772990062832832, + 'model_time': 1.2542282940121368, + 'grad_norm_pre_clip_avg': 0.2616911992430687, + 'learning_rate': 2.4110201492348737e-05, + 'epoch': 2.63} +04/19 [15:06:21] INFO | >> train_qwenlatent.py:487 + Step 10450 | grad_norm_pre_clip=0.2459 | + grad_norm_pre_clip_avg=0.2158 | Metrics: + {'align_loss': 0.024810250848531723, + 'recon_loss': 0.03860917314887047, + 'predict_loss': 0.012639814056456089, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24591201543807983, + 'mae_score': 0.020392759855803067, 'data_time': + 0.0012649839918594807, 'model_time': + 1.2681959999899846, 'grad_norm_pre_clip_avg': + 0.2158357709646225, 'learning_rate': + 2.410696676317541e-05, 'epoch': 2.64} +04/19 [15:06:34] INFO | >> train_qwenlatent.py:487 + Step 10460 | grad_norm_pre_clip=0.2404 | + grad_norm_pre_clip_avg=0.2139 | Metrics: + {'align_loss': 0.02374378778040409, + 'recon_loss': 0.037969063967466354, + 'predict_loss': 0.012518721632659435, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24038653075695038, + 'data_time': 0.0012442769948393106, + 'model_time': 1.2044452050176915, + 'grad_norm_pre_clip_avg': 0.21391332000494004, + 'learning_rate': 2.410372638299983e-05, + 'epoch': 2.64} +04/19 [15:06:47] INFO | >> train_qwenlatent.py:487 + Step 10470 | grad_norm_pre_clip=0.2652 | + grad_norm_pre_clip_avg=0.3201 | Metrics: + {'align_loss': 0.024408331140875816, + 'recon_loss': 0.0497378446161747, + 'predict_loss': 0.020925121381878853, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2651585042476654, + 'data_time': 0.0007489470008295029, + 'model_time': 1.2245887700119056, + 'grad_norm_pre_clip_avg': 0.32008460611104966, + 'learning_rate': 2.4100480353401325e-05, + 'epoch': 2.64} +04/19 [15:06:59] INFO | >> train_qwenlatent.py:487 + Step 10480 | grad_norm_pre_clip=0.2576 | + grad_norm_pre_clip_avg=0.2670 | Metrics: + {'align_loss': 0.02457069605588913, + 'recon_loss': 0.0329110287129879, + 'predict_loss': 0.012025577016174793, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25762900710105896, + 'data_time': 0.0011484179995022714, + 'model_time': 1.2568304899905343, + 'grad_norm_pre_clip_avg': 0.2669722780585289, + 'learning_rate': 2.4097228675961962e-05, + 'epoch': 2.64} +04/19 [15:07:12] INFO | >> train_qwenlatent.py:487 + Step 10490 | grad_norm_pre_clip=0.2338 | + grad_norm_pre_clip_avg=0.2802 | Metrics: + {'align_loss': 0.02364286035299301, + 'recon_loss': 0.04150282219052315, + 'predict_loss': 0.016182664781808853, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23381584882736206, + 'data_time': 0.0009480270091444254, + 'model_time': 1.309108373010531, + 'grad_norm_pre_clip_avg': 0.2801798671483994, + 'learning_rate': 2.4093971352266587e-05, + 'epoch': 2.65} +04/19 [15:07:25] INFO | >> train_qwenlatent.py:487 + Step 10500 | grad_norm_pre_clip=0.2152 | + grad_norm_pre_clip_avg=0.2228 | Metrics: + {'align_loss': 0.02364918403327465, + 'recon_loss': 0.04234663397073746, + 'predict_loss': 0.020012564957141876, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21519160270690918, + 'mae_score': 0.02444623311360677, 'data_time': + 0.0006858700071461499, 'model_time': + 1.2866881420195568, 'grad_norm_pre_clip_avg': + 0.2227510318160057, 'learning_rate': + 2.409070838390276e-05, 'epoch': 2.65} +04/19 [15:07:38] INFO | >> train_qwenlatent.py:487 + Step 10510 | grad_norm_pre_clip=0.2368 | + grad_norm_pre_clip_avg=0.2501 | Metrics: + {'align_loss': 0.024617940187454224, + 'recon_loss': 0.03946428745985031, + 'predict_loss': 0.017028603702783585, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23675577342510223, + 'data_time': 0.0007871879788581282, + 'model_time': 1.3012072530109435, + 'grad_norm_pre_clip_avg': 0.25009973645210265, + 'learning_rate': 2.4087439772460823e-05, + 'epoch': 2.65} +04/19 [15:07:51] INFO | >> train_qwenlatent.py:487 + Step 10520 | grad_norm_pre_clip=0.3244 | + grad_norm_pre_clip_avg=0.2377 | Metrics: + {'align_loss': 0.02425502985715866, + 'recon_loss': 0.039816223084926605, + 'predict_loss': 0.016855692490935326, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.32436245679855347, + 'data_time': 0.0007274470117408782, + 'model_time': 1.2138350510213058, + 'grad_norm_pre_clip_avg': 0.23767877519130706, + 'learning_rate': 2.4084165519533848e-05, + 'epoch': 2.65} +04/19 [15:08:04] INFO | >> train_qwenlatent.py:487 + Step 10530 | grad_norm_pre_clip=0.3152 | + grad_norm_pre_clip_avg=0.2778 | Metrics: + {'align_loss': 0.02441251277923584, + 'recon_loss': 0.03535393998026848, + 'predict_loss': 0.012707545422017574, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3152081072330475, + 'data_time': 0.000716365990228951, + 'model_time': 1.6329430709884036, + 'grad_norm_pre_clip_avg': 0.2778490364551544, + 'learning_rate': 2.408088562671768e-05, + 'epoch': 2.66} +04/19 [15:08:16] INFO | >> train_qwenlatent.py:487 + Step 10540 | grad_norm_pre_clip=0.2517 | + grad_norm_pre_clip_avg=0.2361 | Metrics: + {'align_loss': 0.024322478100657463, + 'recon_loss': 0.03464282304048538, + 'predict_loss': 0.017659388482570648, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2516843378543854, + 'data_time': 0.0009907869971357286, + 'model_time': 1.236787481000647, + 'grad_norm_pre_clip_avg': 0.23608382493257524, + 'learning_rate': 2.4077600095610895e-05, + 'epoch': 2.66} +04/19 [15:08:30] INFO | >> train_qwenlatent.py:487 + Step 10550 | grad_norm_pre_clip=0.3073 | + grad_norm_pre_clip_avg=0.2796 | Metrics: + {'align_loss': 0.025024527683854103, + 'recon_loss': 0.03912319988012314, + 'predict_loss': 0.013576100580394268, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.30727627873420715, + 'mae_score': 0.030484173748944257, 'data_time': + 0.0006796530215069652, 'model_time': + 1.2432806540164165, 'grad_norm_pre_clip_avg': + 0.27958197593688966, 'learning_rate': + 2.407430892781481e-05, 'epoch': 2.66} +04/19 [15:08:42] INFO | >> train_qwenlatent.py:487 + Step 10560 | grad_norm_pre_clip=0.2486 | + grad_norm_pre_clip_avg=0.2365 | Metrics: + {'align_loss': 0.023868612945079803, + 'recon_loss': 0.027304768562316895, + 'predict_loss': 0.011362799443304539, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24858412146568298, + 'data_time': 0.000882308988366276, + 'model_time': 1.2383003130089492, + 'grad_norm_pre_clip_avg': 0.2365274727344513, + 'learning_rate': 2.4071012124933514e-05, + 'epoch': 2.66} +04/19 [15:08:55] INFO | >> train_qwenlatent.py:487 + Step 10570 | grad_norm_pre_clip=0.2340 | + grad_norm_pre_clip_avg=0.2567 | Metrics: + {'align_loss': 0.02452223375439644, + 'recon_loss': 0.03556181117892265, + 'predict_loss': 0.011959384195506573, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23395222425460815, + 'data_time': 0.0010390059906058013, + 'model_time': 1.240107653022278, + 'grad_norm_pre_clip_avg': 0.2566976025700569, + 'learning_rate': 2.4067709688573816e-05, + 'epoch': 2.67} +04/19 [15:09:07] INFO | >> train_qwenlatent.py:487 + Step 10580 | grad_norm_pre_clip=0.3283 | + grad_norm_pre_clip_avg=0.2318 | Metrics: + {'align_loss': 0.02343660593032837, + 'recon_loss': 0.04021291062235832, + 'predict_loss': 0.018014024943113327, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.32827308773994446, + 'data_time': 0.0007206489972304553, + 'model_time': 1.2396901250176597, + 'grad_norm_pre_clip_avg': 0.2317866563796997, + 'learning_rate': 2.4064401620345296e-05, + 'epoch': 2.67} +04/19 [15:09:21] INFO | >> train_qwenlatent.py:487 + Step 10590 | grad_norm_pre_clip=0.2106 | + grad_norm_pre_clip_avg=0.2541 | Metrics: + {'align_loss': 0.025538407266139984, + 'recon_loss': 0.03622728958725929, + 'predict_loss': 0.0148258525878191, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21062803268432617, + 'data_time': 0.0006453639944083989, + 'model_time': 1.2033063959970605, + 'grad_norm_pre_clip_avg': 0.25407870709896085, + 'learning_rate': 2.4061087921860258e-05, + 'epoch': 2.67} +04/19 [15:09:34] INFO | >> train_qwenlatent.py:487 + Step 10600 | grad_norm_pre_clip=0.2242 | + grad_norm_pre_clip_avg=0.2432 | Metrics: + {'align_loss': 0.023639163002371788, + 'recon_loss': 0.03546322137117386, + 'predict_loss': 0.014315120875835419, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2242194414138794, + 'mae_score': 0.018488540305747644, 'data_time': + 0.0009429090132471174, 'model_time': + 1.410752447991399, 'grad_norm_pre_clip_avg': + 0.2432304337620735, 'learning_rate': + 2.4057768594733763e-05, 'epoch': 2.67} +04/19 [15:09:46] INFO | >> train_qwenlatent.py:487 + Step 10610 | grad_norm_pre_clip=0.2688 | + grad_norm_pre_clip_avg=0.2567 | Metrics: + {'align_loss': 0.025258135050535202, + 'recon_loss': 0.03784385696053505, + 'predict_loss': 0.020116236060857773, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26881858706474304, + 'data_time': 0.0008325279923155904, + 'model_time': 1.244185974006541, + 'grad_norm_pre_clip_avg': 0.2567360445857048, + 'learning_rate': 2.4054443640583604e-05, + 'epoch': 2.68} +04/19 [15:09:59] INFO | >> train_qwenlatent.py:487 + Step 10620 | grad_norm_pre_clip=0.2395 | + grad_norm_pre_clip_avg=0.2045 | Metrics: + {'align_loss': 0.024349123239517212, + 'recon_loss': 0.04055180773139, 'predict_loss': + 0.020666684955358505, 'aux_loss_decay_weight': + 0.0, 'grad_norm_pre_clip': 0.23952889442443848, + 'data_time': 0.000752756983274594, + 'model_time': 1.2199482419819105, + 'grad_norm_pre_clip_avg': 0.20452534258365632, + 'learning_rate': 2.405111306103033e-05, + 'epoch': 2.68} +04/19 [15:10:11] INFO | >> train_qwenlatent.py:487 + Step 10630 | grad_norm_pre_clip=0.4607 | + grad_norm_pre_clip_avg=0.3451 | Metrics: + {'align_loss': 0.024022340774536133, + 'recon_loss': 0.030404530465602875, + 'predict_loss': 0.013154055923223495, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.46066659688949585, + 'data_time': 0.0011319569894112647, + 'model_time': 1.243570719001582, + 'grad_norm_pre_clip_avg': 0.3450869515538216, + 'learning_rate': 2.4047776857697226e-05, + 'epoch': 2.68} +04/19 [15:10:24] INFO | >> train_qwenlatent.py:487 + Step 10640 | grad_norm_pre_clip=0.2630 | + grad_norm_pre_clip_avg=0.3127 | Metrics: + {'align_loss': 0.02377578243613243, + 'recon_loss': 0.03284161910414696, + 'predict_loss': 0.018329642713069916, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26296019554138184, + 'data_time': 0.0008366009860765189, + 'model_time': 1.2315534130029846, + 'grad_norm_pre_clip_avg': 0.3126510679721832, + 'learning_rate': 2.4044435032210307e-05, + 'epoch': 2.68} +04/19 [15:10:37] INFO | >> train_qwenlatent.py:487 + Step 10650 | grad_norm_pre_clip=0.1982 | + grad_norm_pre_clip_avg=0.2357 | Metrics: + {'align_loss': 0.02506280317902565, + 'recon_loss': 0.034089166671037674, + 'predict_loss': 0.01266449224203825, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19818024337291718, + 'mae_score': 0.022607626356520093, 'data_time': + 0.001020419003907591, 'model_time': + 1.2840051230159588, 'grad_norm_pre_clip_avg': + 0.23566815108060837, 'learning_rate': + 2.4041087586198354e-05, 'epoch': 2.69} +04/19 [15:10:49] INFO | >> train_qwenlatent.py:487 + Step 10660 | grad_norm_pre_clip=0.2271 | + grad_norm_pre_clip_avg=0.2263 | Metrics: + {'align_loss': 0.02447311207652092, + 'recon_loss': 0.03744436055421829, + 'predict_loss': 0.016821298748254776, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22710102796554565, + 'data_time': 0.0011005109990946949, + 'model_time': 1.236582053010352, + 'grad_norm_pre_clip_avg': 0.22634292244911194, + 'learning_rate': 2.4037734521292853e-05, + 'epoch': 2.69} +04/19 [15:11:02] INFO | >> train_qwenlatent.py:487 + Step 10670 | grad_norm_pre_clip=0.1911 | + grad_norm_pre_clip_avg=0.2230 | Metrics: + {'align_loss': 0.02564023993909359, + 'recon_loss': 0.034302081912755966, + 'predict_loss': 0.014232724905014038, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19112582504749298, + 'data_time': 0.000805260002380237, + 'model_time': 1.2215665729891043, + 'grad_norm_pre_clip_avg': 0.2230461671948433, + 'learning_rate': 2.4034375839128063e-05, + 'epoch': 2.69} +04/19 [15:11:15] INFO | >> train_qwenlatent.py:487 + Step 10680 | grad_norm_pre_clip=0.2955 | + grad_norm_pre_clip_avg=0.2827 | Metrics: + {'align_loss': 0.02459346130490303, + 'recon_loss': 0.04160701483488083, + 'predict_loss': 0.015257800929248333, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.295457124710083, + 'data_time': 0.0007002670026849955, + 'model_time': 1.2661816100007854, + 'grad_norm_pre_clip_avg': 0.28272718042135236, + 'learning_rate': 2.4031011541340953e-05, + 'epoch': 2.69} +04/19 [15:11:28] INFO | >> train_qwenlatent.py:487 + Step 10690 | grad_norm_pre_clip=0.1930 | + grad_norm_pre_clip_avg=0.2350 | Metrics: + {'align_loss': 0.02362731471657753, + 'recon_loss': 0.02838081307709217, + 'predict_loss': 0.013567222282290459, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19298331439495087, + 'data_time': 0.0006456620176322758, + 'model_time': 1.2256378309975844, + 'grad_norm_pre_clip_avg': 0.23497671633958817, + 'learning_rate': 2.402764162957125e-05, + 'epoch': 2.7} +04/19 [15:11:41] INFO | >> train_qwenlatent.py:487 + Step 10700 | grad_norm_pre_clip=0.1935 | + grad_norm_pre_clip_avg=0.2540 | Metrics: + {'align_loss': 0.023879457265138626, + 'recon_loss': 0.0313633494079113, + 'predict_loss': 0.014495521783828735, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19348663091659546, + 'mae_score': 0.021185813079009186, 'data_time': + 0.0007174970232881606, 'model_time': + 1.2115992549806833, 'grad_norm_pre_clip_avg': + 0.2539863422513008, 'learning_rate': + 2.4024266105461402e-05, 'epoch': 2.7} +04/19 [15:11:54] INFO | >> train_qwenlatent.py:487 + Step 10710 | grad_norm_pre_clip=0.2095 | + grad_norm_pre_clip_avg=0.2304 | Metrics: + {'align_loss': 0.02367563359439373, + 'recon_loss': 0.02548070065677166, + 'predict_loss': 0.009635603986680508, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.209503173828125, + 'data_time': 0.0010084960085805506, + 'model_time': 1.206745104980655, + 'grad_norm_pre_clip_avg': 0.23036229908466338, + 'learning_rate': 2.40208849706566e-05, 'epoch': + 2.7} +04/19 [15:12:06] INFO | >> train_qwenlatent.py:487 + Step 10720 | grad_norm_pre_clip=0.4032 | + grad_norm_pre_clip_avg=0.2486 | Metrics: + {'align_loss': 0.02560156211256981, + 'recon_loss': 0.042824968695640564, + 'predict_loss': 0.014335243962705135, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.40324828028678894, + 'data_time': 0.0007998189830686897, + 'model_time': 1.222011981997639, + 'grad_norm_pre_clip_avg': 0.2485709235072136, + 'learning_rate': 2.4017498226804774e-05, + 'epoch': 2.71} +04/19 [15:12:19] INFO | >> train_qwenlatent.py:487 + Step 10730 | grad_norm_pre_clip=0.3071 | + grad_norm_pre_clip_avg=0.3438 | Metrics: + {'align_loss': 0.023766858503222466, + 'recon_loss': 0.034176670014858246, + 'predict_loss': 0.012899797409772873, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3070898950099945, + 'data_time': 0.0009193780133500695, + 'model_time': 1.2149040130025242, + 'grad_norm_pre_clip_avg': 0.3437921956181526, + 'learning_rate': 2.401410587555657e-05, + 'epoch': 2.71} +04/19 [15:12:31] INFO | >> train_qwenlatent.py:487 + Step 10740 | grad_norm_pre_clip=0.2807 | + grad_norm_pre_clip_avg=0.2646 | Metrics: + {'align_loss': 0.025517944246530533, + 'recon_loss': 0.037360649555921555, + 'predict_loss': 0.01620144210755825, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2806650996208191, + 'data_time': 0.0008415500051341951, + 'model_time': 1.2203567599935923, + 'grad_norm_pre_clip_avg': 0.2646158695220947, + 'learning_rate': 2.401070791856539e-05, + 'epoch': 2.71} +04/19 [15:12:45] INFO | >> train_qwenlatent.py:487 + Step 10750 | grad_norm_pre_clip=0.2595 | + grad_norm_pre_clip_avg=0.2218 | Metrics: + {'align_loss': 0.024712029844522476, + 'recon_loss': 0.03681057319045067, + 'predict_loss': 0.013120361603796482, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2595359683036804, + 'mae_score': 0.021603620374524915, 'data_time': + 0.0006996570155024529, 'model_time': + 1.2336161069979426, 'grad_norm_pre_clip_avg': + 0.22184558361768722, 'learning_rate': + 2.4007304357487352e-05, 'epoch': 2.71} +04/19 [15:12:58] INFO | >> train_qwenlatent.py:487 + Step 10760 | grad_norm_pre_clip=0.1897 | + grad_norm_pre_clip_avg=0.2171 | Metrics: + {'align_loss': 0.02362709492444992, + 'recon_loss': 0.029749732464551926, + 'predict_loss': 0.013270790688693523, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1897369772195816, + 'data_time': 0.0010828399972524494, + 'model_time': 1.2790949890040793, + 'grad_norm_pre_clip_avg': 0.217093326151371, + 'learning_rate': 2.4003895193981308e-05, + 'epoch': 2.72} +04/19 [15:13:10] INFO | >> train_qwenlatent.py:487 + Step 10770 | grad_norm_pre_clip=0.3184 | + grad_norm_pre_clip_avg=0.2722 | Metrics: + {'align_loss': 0.0255066379904747, + 'recon_loss': 0.04677842557430267, + 'predict_loss': 0.016833454370498657, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.31835952401161194, + 'data_time': 0.0009793439821805805, + 'model_time': 1.2154220399970654, + 'grad_norm_pre_clip_avg': 0.27222696095705035, + 'learning_rate': 2.400048042970885e-05, + 'epoch': 2.72} +04/19 [15:13:23] INFO | >> train_qwenlatent.py:487 + Step 10780 | grad_norm_pre_clip=0.2479 | + grad_norm_pre_clip_avg=0.2974 | Metrics: + {'align_loss': 0.023806855082511902, + 'recon_loss': 0.039779677987098694, + 'predict_loss': 0.018253110349178314, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2478896528482437, + 'data_time': 0.0009892939997371286, + 'model_time': 1.2059153559966944, + 'grad_norm_pre_clip_avg': 0.29741276055574417, + 'learning_rate': 2.3997060066334282e-05, + 'epoch': 2.72} +04/19 [15:13:35] INFO | >> train_qwenlatent.py:487 + Step 10790 | grad_norm_pre_clip=0.1907 | + grad_norm_pre_clip_avg=0.2288 | Metrics: + {'align_loss': 0.02403268963098526, + 'recon_loss': 0.04321902245283127, + 'predict_loss': 0.016516869887709618, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19074736535549164, + 'data_time': 0.0008758319891057909, + 'model_time': 1.2416204150067642, + 'grad_norm_pre_clip_avg': 0.22883426994085312, + 'learning_rate': 2.3993634105524655e-05, + 'epoch': 2.72} +04/19 [15:13:49] INFO | >> train_qwenlatent.py:487 + Step 10800 | grad_norm_pre_clip=0.2058 | + grad_norm_pre_clip_avg=0.2228 | Metrics: + {'align_loss': 0.02414652705192566, + 'recon_loss': 0.03690569847822189, + 'predict_loss': 0.017211608588695526, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2058260440826416, + 'mae_score': 0.02888768513997396, 'data_time': + 0.0007042450015433133, 'model_time': + 1.2192760690231808, 'grad_norm_pre_clip_avg': + 0.22283841222524642, 'learning_rate': + 2.3990202548949742e-05, 'epoch': 2.73} +04/19 [15:14:02] INFO | >> train_qwenlatent.py:487 + Step 10810 | grad_norm_pre_clip=0.3811 | + grad_norm_pre_clip_avg=0.2866 | Metrics: + {'align_loss': 0.02428996004164219, + 'recon_loss': 0.04193345084786415, + 'predict_loss': 0.015597709454596043, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.38114073872566223, + 'data_time': 0.0007399920141324401, + 'model_time': 1.1972005440038629, + 'grad_norm_pre_clip_avg': 0.28659498244524, + 'learning_rate': 2.3986765398282037e-05, + 'epoch': 2.73} +04/19 [15:14:14] INFO | >> train_qwenlatent.py:487 + Step 10820 | grad_norm_pre_clip=0.2649 | + grad_norm_pre_clip_avg=0.2750 | Metrics: + {'align_loss': 0.02471597120165825, + 'recon_loss': 0.033174797892570496, + 'predict_loss': 0.014481877908110619, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26493313908576965, + 'data_time': 0.0009290850139223039, + 'model_time': 1.2527949360082857, + 'grad_norm_pre_clip_avg': 0.274985571205616, + 'learning_rate': 2.3983322655196768e-05, + 'epoch': 2.73} +04/19 [15:14:27] INFO | >> train_qwenlatent.py:487 + Step 10830 | grad_norm_pre_clip=0.2106 | + grad_norm_pre_clip_avg=0.2164 | Metrics: + {'align_loss': 0.02515014447271824, + 'recon_loss': 0.03721998259425163, + 'predict_loss': 0.013372156769037247, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21062101423740387, + 'data_time': 0.0006698920042254031, + 'model_time': 1.234736073995009, + 'grad_norm_pre_clip_avg': 0.21639440208673477, + 'learning_rate': 2.3979874321371886e-05, + 'epoch': 2.73} +04/19 [15:14:39] INFO | >> train_qwenlatent.py:487 + Step 10840 | grad_norm_pre_clip=0.2512 | + grad_norm_pre_clip_avg=0.2066 | Metrics: + {'align_loss': 0.02384035475552082, + 'recon_loss': 0.03849443793296814, + 'predict_loss': 0.013176273554563522, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25116682052612305, + 'data_time': 0.0009225449757650495, + 'model_time': 1.2386633809946943, + 'grad_norm_pre_clip_avg': 0.20663533210754395, + 'learning_rate': 2.397642039848807e-05, + 'epoch': 2.74} +04/19 [15:14:53] INFO | >> train_qwenlatent.py:487 + Step 10850 | grad_norm_pre_clip=0.3177 | + grad_norm_pre_clip_avg=0.3210 | Metrics: + {'align_loss': 0.025260070338845253, + 'recon_loss': 0.04013185203075409, + 'predict_loss': 0.014325304888188839, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.31774139404296875, + 'mae_score': 0.01700096474037514, 'data_time': + 0.0006837090186309069, 'model_time': + 1.243150534981396, 'grad_norm_pre_clip_avg': + 0.32097517848014834, 'learning_rate': + 2.3972960888228716e-05, 'epoch': 2.74} +04/19 [15:15:05] INFO | >> train_qwenlatent.py:487 + Step 10860 | grad_norm_pre_clip=0.2551 | + grad_norm_pre_clip_avg=0.2416 | Metrics: + {'align_loss': 0.022957637906074524, + 'recon_loss': 0.02691415697336197, + 'predict_loss': 0.011591355316340923, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2551092803478241, + 'data_time': 0.000786015996709466, + 'model_time': 1.2886073849804234, + 'grad_norm_pre_clip_avg': 0.2415671393275261, + 'learning_rate': 2.3969495792279946e-05, + 'epoch': 2.74} +04/19 [15:15:18] INFO | >> train_qwenlatent.py:487 + Step 10870 | grad_norm_pre_clip=0.1941 | + grad_norm_pre_clip_avg=0.2108 | Metrics: + {'align_loss': 0.02393171191215515, + 'recon_loss': 0.039267249405384064, + 'predict_loss': 0.015323219820857048, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19408448040485382, + 'data_time': 0.000631399016128853, + 'model_time': 1.2574379249999765, + 'grad_norm_pre_clip_avg': 0.21081364303827285, + 'learning_rate': 2.396602511233061e-05, + 'epoch': 2.74} +04/19 [15:15:31] INFO | >> train_qwenlatent.py:487 + Step 10880 | grad_norm_pre_clip=0.2817 | + grad_norm_pre_clip_avg=0.2206 | Metrics: + {'align_loss': 0.02467748522758484, + 'recon_loss': 0.04545485973358154, + 'predict_loss': 0.015487810596823692, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28166085481643677, + 'data_time': 0.0006558120076078922, + 'model_time': 1.2695654470007867, + 'grad_norm_pre_clip_avg': 0.220592699944973, + 'learning_rate': 2.3962548850072275e-05, + 'epoch': 2.75} +04/19 [15:15:44] INFO | >> train_qwenlatent.py:487 + Step 10890 | grad_norm_pre_clip=0.2068 | + grad_norm_pre_clip_avg=0.3397 | Metrics: + {'align_loss': 0.023806868121027946, + 'recon_loss': 0.033092569559812546, + 'predict_loss': 0.014944635331630707, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2067776322364807, + 'data_time': 0.0011282990162726492, + 'model_time': 1.2323977930063847, + 'grad_norm_pre_clip_avg': 0.3396785318851471, + 'learning_rate': 2.3959067007199223e-05, + 'epoch': 2.75} +04/19 [15:15:56] INFO | >> train_qwenlatent.py:487 + Step 10900 | grad_norm_pre_clip=0.2329 | + grad_norm_pre_clip_avg=0.2318 | Metrics: + {'align_loss': 0.0248020701110363, + 'recon_loss': 0.03595476225018501, + 'predict_loss': 0.015273976139724255, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23293325304985046, + 'mae_score': 0.018355342074557468, 'data_time': + 0.0007659970142412931, 'model_time': + 1.2463157790189143, 'grad_norm_pre_clip_avg': + 0.23177373558282852, 'learning_rate': + 2.395557958540847e-05, 'epoch': 2.75} +04/19 [15:16:09] INFO | >> train_qwenlatent.py:487 + Step 10910 | grad_norm_pre_clip=0.2378 | + grad_norm_pre_clip_avg=0.2247 | Metrics: + {'align_loss': 0.024429820477962494, + 'recon_loss': 0.04742288589477539, + 'predict_loss': 0.020892413333058357, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23776686191558838, + 'data_time': 0.0008751909772399813, + 'model_time': 1.2474684399785474, + 'grad_norm_pre_clip_avg': 0.2246982514858246, + 'learning_rate': 2.3952086586399736e-05, + 'epoch': 2.75} +04/19 [15:16:22] INFO | >> train_qwenlatent.py:487 + Step 10920 | grad_norm_pre_clip=0.2911 | + grad_norm_pre_clip_avg=0.2398 | Metrics: + {'align_loss': 0.02487240359187126, + 'recon_loss': 0.03314517065882683, + 'predict_loss': 0.010883346199989319, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2910979390144348, + 'data_time': 0.0006956059951335192, + 'model_time': 1.221918428025674, + 'grad_norm_pre_clip_avg': 0.23977213948965073, + 'learning_rate': 2.3948588011875473e-05, + 'epoch': 2.76} +04/19 [15:16:35] INFO | >> train_qwenlatent.py:487 + Step 10930 | grad_norm_pre_clip=0.2741 | + grad_norm_pre_clip_avg=0.2552 | Metrics: + {'align_loss': 0.025358794257044792, + 'recon_loss': 0.04449259117245674, + 'predict_loss': 0.01269321609288454, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2740894854068756, + 'data_time': 0.0008422470127698034, + 'model_time': 1.21909645901178, + 'grad_norm_pre_clip_avg': 0.255225282907486, + 'learning_rate': 2.394508386354084e-05, + 'epoch': 2.76} +04/19 [15:16:47] INFO | >> train_qwenlatent.py:487 + Step 10940 | grad_norm_pre_clip=0.1770 | + grad_norm_pre_clip_avg=0.2437 | Metrics: + {'align_loss': 0.023944078013300896, + 'recon_loss': 0.039557475596666336, + 'predict_loss': 0.018281085416674614, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17697373032569885, + 'data_time': 0.0008491939806845039, + 'model_time': 1.2549836150137708, + 'grad_norm_pre_clip_avg': 0.2437109649181366, + 'learning_rate': 2.394157414310371e-05, + 'epoch': 2.76} +04/19 [15:17:01] INFO | >> train_qwenlatent.py:487 + Step 10950 | grad_norm_pre_clip=0.2514 | + grad_norm_pre_clip_avg=0.2540 | Metrics: + {'align_loss': 0.024566203355789185, + 'recon_loss': 0.03448215126991272, + 'predict_loss': 0.013088556006550789, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25144800543785095, + 'mae_score': 0.019511671324033995, 'data_time': + 0.0009165160008706152, 'model_time': + 1.2158160469844006, 'grad_norm_pre_clip_avg': + 0.2540366932749748, 'learning_rate': + 2.3938058852274687e-05, 'epoch': 2.76} +04/19 [15:17:13] INFO | >> train_qwenlatent.py:487 + Step 10960 | grad_norm_pre_clip=0.2693 | + grad_norm_pre_clip_avg=0.2394 | Metrics: + {'align_loss': 0.02586514689028263, + 'recon_loss': 0.049713850021362305, + 'predict_loss': 0.018903374671936035, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26930543780326843, + 'data_time': 0.0007060060161165893, + 'model_time': 1.2533974740072154, + 'grad_norm_pre_clip_avg': 0.23942325115203858, + 'learning_rate': 2.393453799276708e-05, + 'epoch': 2.77} +04/19 [15:17:26] INFO | >> train_qwenlatent.py:487 + Step 10970 | grad_norm_pre_clip=0.2676 | + grad_norm_pre_clip_avg=0.2494 | Metrics: + {'align_loss': 0.024542957544326782, + 'recon_loss': 0.04393594339489937, + 'predict_loss': 0.01644284650683403, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26764562726020813, + 'data_time': 0.0007643879798706621, + 'model_time': 1.3987314700207207, + 'grad_norm_pre_clip_avg': 0.24938047230243682, + 'learning_rate': 2.3931011566296916e-05, + 'epoch': 2.77} +04/19 [15:17:38] INFO | >> train_qwenlatent.py:487 + Step 10980 | grad_norm_pre_clip=0.1794 | + grad_norm_pre_clip_avg=0.2089 | Metrics: + {'align_loss': 0.02405593916773796, + 'recon_loss': 0.04404744505882263, + 'predict_loss': 0.01481787208467722, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17936623096466064, + 'data_time': 0.0008638470026198775, + 'model_time': 1.2629306919989176, + 'grad_norm_pre_clip_avg': 0.20888267308473588, + 'learning_rate': 2.3927479574582926e-05, + 'epoch': 2.77} +04/19 [15:17:51] INFO | >> train_qwenlatent.py:487 + Step 10990 | grad_norm_pre_clip=0.2775 | + grad_norm_pre_clip_avg=0.2472 | Metrics: + {'align_loss': 0.025674663484096527, + 'recon_loss': 0.0397392138838768, + 'predict_loss': 0.010867374949157238, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.27748972177505493, + 'data_time': 0.0009175350132863969, + 'model_time': 1.191042627004208, + 'grad_norm_pre_clip_avg': 0.24720238894224167, + 'learning_rate': 2.3923942019346556e-05, + 'epoch': 2.77} +04/19 [15:18:04] INFO | >> train_qwenlatent.py:487 + Step 11000 | grad_norm_pre_clip=0.2443 | + grad_norm_pre_clip_avg=0.2388 | Metrics: + {'align_loss': 0.022231735289096832, + 'recon_loss': 0.02802683226764202, + 'predict_loss': 0.01340195070952177, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24428310990333557, + 'mae_score': 0.02062043370427312, 'data_time': + 0.0007873610011301935, 'model_time': + 1.2357566569990013, 'grad_norm_pre_clip_avg': + 0.23880810886621476, 'learning_rate': + 2.3920398902311978e-05, 'epoch': 2.78} +04/19 [15:18:16] INFO | >> train_qwenlatent.py:487 + Step 11010 | grad_norm_pre_clip=0.2285 | + grad_norm_pre_clip_avg=0.2659 | Metrics: + {'align_loss': 0.02600198984146118, + 'recon_loss': 0.0417448990046978, + 'predict_loss': 0.011757772415876389, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22852377593517303, + 'data_time': 0.0006778739916626364, + 'model_time': 1.2339516030042432, + 'grad_norm_pre_clip_avg': 0.26588583886623385, + 'learning_rate': 2.3916850225206062e-05, + 'epoch': 2.78} +04/19 [15:18:29] INFO | >> train_qwenlatent.py:487 + Step 11020 | grad_norm_pre_clip=0.2100 | + grad_norm_pre_clip_avg=0.2268 | Metrics: + {'align_loss': 0.02524789422750473, + 'recon_loss': 0.048524096608161926, + 'predict_loss': 0.013429174199700356, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21004800498485565, + 'data_time': 0.0006139010074548423, + 'model_time': 1.2576301280059852, + 'grad_norm_pre_clip_avg': 0.2268323391675949, + 'learning_rate': 2.3913295989758385e-05, + 'epoch': 2.78} +04/19 [15:18:42] INFO | >> train_qwenlatent.py:487 + Step 11030 | grad_norm_pre_clip=0.2961 | + grad_norm_pre_clip_avg=0.2683 | Metrics: + {'align_loss': 0.023837197571992874, + 'recon_loss': 0.03581884503364563, + 'predict_loss': 0.017926407977938652, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.29607847332954407, + 'data_time': 0.0007980569789651781, + 'model_time': 1.2240838450088631, + 'grad_norm_pre_clip_avg': 0.2682879567146301, + 'learning_rate': 2.3909736197701245e-05, + 'epoch': 2.78} +04/19 [15:18:54] INFO | >> train_qwenlatent.py:487 + Step 11040 | grad_norm_pre_clip=0.3198 | + grad_norm_pre_clip_avg=0.2657 | Metrics: + {'align_loss': 0.02503831312060356, + 'recon_loss': 0.03742174804210663, + 'predict_loss': 0.013499069958925247, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3198155462741852, + 'data_time': 0.0009935680136550218, + 'model_time': 1.288228193006944, + 'grad_norm_pre_clip_avg': 0.2656987413764, + 'learning_rate': 2.3906170850769633e-05, + 'epoch': 2.79} +04/19 [15:19:08] INFO | >> train_qwenlatent.py:487 + Step 11050 | grad_norm_pre_clip=0.2551 | + grad_norm_pre_clip_avg=0.2520 | Metrics: + {'align_loss': 0.02598053216934204, + 'recon_loss': 0.058725759387016296, + 'predict_loss': 0.020891427993774414, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25514864921569824, + 'mae_score': 0.02554046699592659, 'data_time': + 0.0016207309963647276, 'model_time': + 1.6267239160079043, 'grad_norm_pre_clip_avg': + 0.25195645838975905, 'learning_rate': + 2.390259995070126e-05, 'epoch': 2.79} +04/19 [15:19:21] INFO | >> train_qwenlatent.py:487 + Step 11060 | grad_norm_pre_clip=0.2207 | + grad_norm_pre_clip_avg=0.2234 | Metrics: + {'align_loss': 0.02586798369884491, + 'recon_loss': 0.055169589817523956, + 'predict_loss': 0.018393907696008682, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22069783508777618, + 'data_time': 0.0007214020006358624, + 'model_time': 1.2855315810011234, + 'grad_norm_pre_clip_avg': 0.2233688622713089, + 'learning_rate': 2.3899023499236542e-05, + 'epoch': 2.79} +04/19 [15:19:34] INFO | >> train_qwenlatent.py:487 + Step 11070 | grad_norm_pre_clip=0.3127 | + grad_norm_pre_clip_avg=0.2409 | Metrics: + {'align_loss': 0.0259583480656147, + 'recon_loss': 0.04176731035113335, + 'predict_loss': 0.0137040875852108, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3126643896102905, + 'data_time': 0.0008720709884073585, + 'model_time': 1.2350618269992992, + 'grad_norm_pre_clip_avg': 0.24090533405542375, + 'learning_rate': 2.3895441498118596e-05, + 'epoch': 2.79} +04/19 [15:19:46] INFO | >> train_qwenlatent.py:487 + Step 11080 | grad_norm_pre_clip=0.2065 | + grad_norm_pre_clip_avg=0.2287 | Metrics: + {'align_loss': 0.025813141837716103, + 'recon_loss': 0.04179873317480087, + 'predict_loss': 0.014874439686536789, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20645681023597717, + 'data_time': 0.0008972629730124027, + 'model_time': 1.221925090008881, + 'grad_norm_pre_clip_avg': 0.22867293059825897, + 'learning_rate': 2.3891853949093244e-05, + 'epoch': 2.8} +04/19 [15:19:59] INFO | >> train_qwenlatent.py:487 + Step 11090 | grad_norm_pre_clip=0.2149 | + grad_norm_pre_clip_avg=0.2380 | Metrics: + {'align_loss': 0.023521529510617256, + 'recon_loss': 0.03478579223155975, + 'predict_loss': 0.01759372279047966, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2148599773645401, + 'data_time': 0.0008224629855249077, + 'model_time': 1.24947834800696, + 'grad_norm_pre_clip_avg': 0.23802520185709, + 'learning_rate': 2.3888260853909015e-05, + 'epoch': 2.8} +04/19 [15:20:12] INFO | >> train_qwenlatent.py:487 + Step 11100 | grad_norm_pre_clip=0.2447 | + grad_norm_pre_clip_avg=0.2709 | Metrics: + {'align_loss': 0.024522967636585236, + 'recon_loss': 0.03334090858697891, + 'predict_loss': 0.012562212534248829, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24468016624450684, + 'mae_score': 0.022003893809275584, 'data_time': + 0.0007209199829958379, 'model_time': + 1.566642416990362, 'grad_norm_pre_clip_avg': + 0.2709458634257317, 'learning_rate': + 2.3884662214317135e-05, 'epoch': 2.8} +04/19 [15:20:25] INFO | >> train_qwenlatent.py:487 + Step 11110 | grad_norm_pre_clip=0.1955 | + grad_norm_pre_clip_avg=0.2587 | Metrics: + {'align_loss': 0.02419820986688137, + 'recon_loss': 0.034960124641656876, + 'predict_loss': 0.013897568918764591, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19546100497245789, + 'data_time': 0.0010518239869270474, + 'model_time': 1.2021212139807176, + 'grad_norm_pre_clip_avg': 0.25871291011571884, + 'learning_rate': 2.388105803207155e-05, + 'epoch': 2.8} +04/19 [15:20:37] INFO | >> train_qwenlatent.py:487 + Step 11120 | grad_norm_pre_clip=0.2068 | + grad_norm_pre_clip_avg=0.2463 | Metrics: + {'align_loss': 0.02560647763311863, + 'recon_loss': 0.04526042938232422, + 'predict_loss': 0.010069742798805237, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2067808210849762, + 'data_time': 0.0006852540245745331, + 'model_time': 1.263090802996885, + 'grad_norm_pre_clip_avg': 0.24629224389791488, + 'learning_rate': 2.3877448308928885e-05, + 'epoch': 2.81} +04/19 [15:20:50] INFO | >> train_qwenlatent.py:487 + Step 11130 | grad_norm_pre_clip=0.3327 | + grad_norm_pre_clip_avg=0.2593 | Metrics: + {'align_loss': 0.025570880621671677, + 'recon_loss': 0.04088297858834267, + 'predict_loss': 0.016203587874770164, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.33269718289375305, + 'data_time': 0.0007044169760774821, + 'model_time': 1.1709913029917516, + 'grad_norm_pre_clip_avg': 0.2592510595917702, + 'learning_rate': 2.3873833046648475e-05, + 'epoch': 2.81} +04/19 [15:21:03] INFO | >> train_qwenlatent.py:487 + Step 11140 | grad_norm_pre_clip=0.3309 | + grad_norm_pre_clip_avg=0.2982 | Metrics: + {'align_loss': 0.023881934583187103, + 'recon_loss': 0.043132197111845016, + 'predict_loss': 0.015065331012010574, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3308756649494171, + 'data_time': 0.0007434980070684105, + 'model_time': 1.223819653998362, + 'grad_norm_pre_clip_avg': 0.298249626159668, + 'learning_rate': 2.387021224699236e-05, + 'epoch': 2.81} +04/19 [15:21:16] INFO | >> train_qwenlatent.py:487 + Step 11150 | grad_norm_pre_clip=0.2025 | + grad_norm_pre_clip_avg=0.2367 | Metrics: + {'align_loss': 0.024831674993038177, + 'recon_loss': 0.039880938827991486, + 'predict_loss': 0.014011981897056103, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20249463617801666, + 'mae_score': 0.015055537868190456, 'data_time': + 0.0007220339903142303, 'model_time': + 1.2668564899940975, 'grad_norm_pre_clip_avg': + 0.23666329085826873, 'learning_rate': + 2.386658591172527e-05, 'epoch': 2.81} +04/19 [15:21:29] INFO | >> train_qwenlatent.py:487 + Step 11160 | grad_norm_pre_clip=0.1936 | + grad_norm_pre_clip_avg=0.1958 | Metrics: + {'align_loss': 0.02392008900642395, + 'recon_loss': 0.03716026619076729, + 'predict_loss': 0.01170406211167574, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19357559084892273, + 'data_time': 0.0009424239979125559, + 'model_time': 1.2533670880075078, + 'grad_norm_pre_clip_avg': 0.19577809423208237, + 'learning_rate': 2.3862954042614637e-05, + 'epoch': 2.82} +04/19 [15:21:41] INFO | >> train_qwenlatent.py:487 + Step 11170 | grad_norm_pre_clip=0.1853 | + grad_norm_pre_clip_avg=0.2454 | Metrics: + {'align_loss': 0.024601735174655914, + 'recon_loss': 0.03991856426000595, + 'predict_loss': 0.015714511275291443, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18525296449661255, + 'data_time': 0.000894486001925543, + 'model_time': 1.2064638180017937, + 'grad_norm_pre_clip_avg': 0.24538031220436096, + 'learning_rate': 2.385931664143059e-05, + 'epoch': 2.82} +04/19 [15:21:54] INFO | >> train_qwenlatent.py:487 + Step 11180 | grad_norm_pre_clip=0.1798 | + grad_norm_pre_clip_avg=0.2581 | Metrics: + {'align_loss': 0.02476358786225319, + 'recon_loss': 0.03772243484854698, + 'predict_loss': 0.014035786502063274, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1797991842031479, + 'data_time': 0.0009365960140712559, + 'model_time': 1.2695391879824456, + 'grad_norm_pre_clip_avg': 0.25814236104488375, + 'learning_rate': 2.3855673709945956e-05, + 'epoch': 2.82} +04/19 [15:22:07] INFO | >> train_qwenlatent.py:487 + Step 11190 | grad_norm_pre_clip=0.2868 | + grad_norm_pre_clip_avg=0.2375 | Metrics: + {'align_loss': 0.024477776139974594, + 'recon_loss': 0.03941016644239426, + 'predict_loss': 0.014334190636873245, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28678014874458313, + 'data_time': 0.0007029389962553978, + 'model_time': 1.2776656950009055, + 'grad_norm_pre_clip_avg': 0.2374978169798851, + 'learning_rate': 2.385202524993625e-05, + 'epoch': 2.82} +04/19 [15:22:21] INFO | >> train_qwenlatent.py:487 + Step 11200 | grad_norm_pre_clip=0.2163 | + grad_norm_pre_clip_avg=0.2372 | Metrics: + {'align_loss': 0.02309397980570793, + 'recon_loss': 0.03518643602728844, + 'predict_loss': 0.011961144395172596, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2162514626979828, + 'mae_score': 0.017883244076290647, 'data_time': + 0.0008063259883783758, 'model_time': + 1.5255599829833955, 'grad_norm_pre_clip_avg': + 0.23715358525514602, 'learning_rate': + 2.3848371263179697e-05, 'epoch': 2.83} +04/19 [15:22:33] INFO | >> train_qwenlatent.py:487 + Step 11210 | grad_norm_pre_clip=0.2441 | + grad_norm_pre_clip_avg=0.2540 | Metrics: + {'align_loss': 0.024958062916994095, + 'recon_loss': 0.03286339342594147, + 'predict_loss': 0.01062043383717537, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2440621256828308, + 'data_time': 0.0006945880013518035, + 'model_time': 1.257187426002929, + 'grad_norm_pre_clip_avg': 0.2539671346545219, + 'learning_rate': 2.38447117514572e-05, 'epoch': + 2.83} +04/19 [15:22:46] INFO | >> train_qwenlatent.py:487 + Step 11220 | grad_norm_pre_clip=0.2232 | + grad_norm_pre_clip_avg=0.2229 | Metrics: + {'align_loss': 0.024422433227300644, + 'recon_loss': 0.04682072624564171, + 'predict_loss': 0.014988856390118599, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22316502034664154, + 'data_time': 0.0009361029951833189, + 'model_time': 1.2732925899908878, + 'grad_norm_pre_clip_avg': 0.222893525660038, + 'learning_rate': 2.384104671655236e-05, + 'epoch': 2.83} +04/19 [15:22:59] INFO | >> train_qwenlatent.py:487 + Step 11230 | grad_norm_pre_clip=0.3450 | + grad_norm_pre_clip_avg=0.2871 | Metrics: + {'align_loss': 0.024416670203208923, + 'recon_loss': 0.0439484529197216, + 'predict_loss': 0.014531802386045456, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3450096845626831, + 'data_time': 0.0010104899993166327, + 'model_time': 1.439475401013624, + 'grad_norm_pre_clip_avg': 0.28709550201892853, + 'learning_rate': 2.3837376160251465e-05, + 'epoch': 2.83} +04/19 [15:23:11] INFO | >> train_qwenlatent.py:487 + Step 11240 | grad_norm_pre_clip=0.1900 | + grad_norm_pre_clip_avg=0.2633 | Metrics: + {'align_loss': 0.0245061032474041, + 'recon_loss': 0.03331739827990532, + 'predict_loss': 0.013010278344154358, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1899677813053131, + 'data_time': 0.0010786590282805264, + 'model_time': 1.3075407709984574, + 'grad_norm_pre_clip_avg': 0.2633284479379654, + 'learning_rate': 2.3833700084343514e-05, + 'epoch': 2.84} +04/19 [15:23:24] INFO | >> train_qwenlatent.py:487 + Step 11250 | grad_norm_pre_clip=0.2151 | + grad_norm_pre_clip_avg=0.2372 | Metrics: + {'align_loss': 0.023271039128303528, + 'recon_loss': 0.041899047791957855, + 'predict_loss': 0.014876121655106544, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21510089933872223, + 'mae_score': 0.022643322987599416, 'data_time': + 0.0009849019988905638, 'model_time': + 1.2049990139785223, 'grad_norm_pre_clip_avg': + 0.23717650175094604, 'learning_rate': + 2.383001849062017e-05, 'epoch': 2.84} +04/19 [15:23:36] INFO | >> train_qwenlatent.py:487 + Step 11260 | grad_norm_pre_clip=0.1599 | + grad_norm_pre_clip_avg=0.2419 | Metrics: + {'align_loss': 0.024209890514612198, + 'recon_loss': 0.030662592500448227, + 'predict_loss': 0.01455141045153141, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15988689661026, + 'data_time': 0.0007025290105957538, + 'model_time': 1.2413316979946103, + 'grad_norm_pre_clip_avg': 0.24193664491176606, + 'learning_rate': 2.38263313808758e-05, 'epoch': + 2.84} +04/19 [15:23:49] INFO | >> train_qwenlatent.py:487 + Step 11270 | grad_norm_pre_clip=0.2683 | + grad_norm_pre_clip_avg=0.2865 | Metrics: + {'align_loss': 0.023651350289583206, + 'recon_loss': 0.04783142730593681, + 'predict_loss': 0.014081493020057678, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26834285259246826, + 'data_time': 0.0009063619945663959, + 'model_time': 1.2799414850014728, + 'grad_norm_pre_clip_avg': 0.28649996072053907, + 'learning_rate': 2.3822638756907458e-05, + 'epoch': 2.84} +04/19 [15:24:02] INFO | >> train_qwenlatent.py:487 + Step 11280 | grad_norm_pre_clip=0.2525 | + grad_norm_pre_clip_avg=0.2367 | Metrics: + {'align_loss': 0.025299757719039917, + 'recon_loss': 0.038289979100227356, + 'predict_loss': 0.0161762572824955, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2524724006652832, + 'data_time': 0.000880967010743916, + 'model_time': 1.2153070219792426, + 'grad_norm_pre_clip_avg': 0.23672343790531158, + 'learning_rate': 2.381894062051488e-05, + 'epoch': 2.85} +04/19 [15:24:14] INFO | >> train_qwenlatent.py:487 + Step 11290 | grad_norm_pre_clip=0.3541 | + grad_norm_pre_clip_avg=0.2569 | Metrics: + {'align_loss': 0.02336808480322361, + 'recon_loss': 0.029514271765947342, + 'predict_loss': 0.013345371931791306, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3541046380996704, + 'data_time': 0.0008069160103332251, + 'model_time': 1.468229671008885, + 'grad_norm_pre_clip_avg': 0.25691489726305006, + 'learning_rate': 2.3815236973500493e-05, + 'epoch': 2.85} +04/19 [15:24:27] INFO | >> train_qwenlatent.py:487 + Step 11300 | grad_norm_pre_clip=0.2071 | + grad_norm_pre_clip_avg=0.2438 | Metrics: + {'align_loss': 0.024302292615175247, + 'recon_loss': 0.046552903950214386, + 'predict_loss': 0.014650936238467693, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20707368850708008, + 'mae_score': 0.02336357391632355, 'data_time': + 0.0010434489813633263, 'model_time': + 1.2782973790017422, 'grad_norm_pre_clip_avg': + 0.24378066509962082, 'learning_rate': + 2.381152781766942e-05, 'epoch': 2.85} +04/19 [15:24:40] INFO | >> train_qwenlatent.py:487 + Step 11310 | grad_norm_pre_clip=0.2173 | + grad_norm_pre_clip_avg=0.2223 | Metrics: + {'align_loss': 0.024191223084926605, + 'recon_loss': 0.0392586812376976, + 'predict_loss': 0.012906450778245926, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21732483804225922, + 'data_time': 0.0009754079801496118, + 'model_time': 1.2172364039870445, + 'grad_norm_pre_clip_avg': 0.22229015082120895, + 'learning_rate': 2.380781315482945e-05, + 'epoch': 2.85} +04/19 [15:24:53] INFO | >> train_qwenlatent.py:487 + Step 11320 | grad_norm_pre_clip=0.2693 | + grad_norm_pre_clip_avg=0.2738 | Metrics: + {'align_loss': 0.024985624477267265, + 'recon_loss': 0.05039045587182045, + 'predict_loss': 0.012761631049215794, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26927438378334045, + 'data_time': 0.0006861639849375933, + 'model_time': 1.511420282011386, + 'grad_norm_pre_clip_avg': 0.27380897998809817, + 'learning_rate': 2.380409298679106e-05, + 'epoch': 2.86} +04/19 [15:25:06] INFO | >> train_qwenlatent.py:487 + Step 11330 | grad_norm_pre_clip=0.2310 | + grad_norm_pre_clip_avg=0.2782 | Metrics: + {'align_loss': 0.024534741416573524, + 'recon_loss': 0.041016750037670135, + 'predict_loss': 0.01700912043452263, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23101137578487396, + 'data_time': 0.001249730004929006, + 'model_time': 1.2694251060020179, + 'grad_norm_pre_clip_avg': 0.278166264295578, + 'learning_rate': 2.3800367315367422e-05, + 'epoch': 2.86} +04/19 [15:25:18] INFO | >> train_qwenlatent.py:487 + Step 11340 | grad_norm_pre_clip=0.2502 | + grad_norm_pre_clip_avg=0.2682 | Metrics: + {'align_loss': 0.02459140121936798, + 'recon_loss': 0.03796584531664848, + 'predict_loss': 0.01352006010711193, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2501840591430664, + 'data_time': 0.0007062270015012473, + 'model_time': 1.2153878409881145, + 'grad_norm_pre_clip_avg': 0.26816902607679366, + 'learning_rate': 2.3796636142374387e-05, + 'epoch': 2.86} +04/19 [15:25:31] INFO | >> train_qwenlatent.py:487 + Step 11350 | grad_norm_pre_clip=0.2762 | + grad_norm_pre_clip_avg=0.2705 | Metrics: + {'align_loss': 0.02345193363726139, + 'recon_loss': 0.04223499447107315, + 'predict_loss': 0.017341338098049164, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2761540114879608, + 'mae_score': 0.01721425099415822, 'data_time': + 0.0007116799824871123, 'model_time': + 1.1440895539999474, 'grad_norm_pre_clip_avg': + 0.27048211842775344, 'learning_rate': + 2.3792899469630473e-05, 'epoch': 2.86} +04/19 [15:25:42] INFO | >> train_qwenlatent.py:487 + Step 11360 | grad_norm_pre_clip=0.2470 | + grad_norm_pre_clip_avg=0.2438 | Metrics: + {'align_loss': 0.023887287825345993, + 'recon_loss': 0.04407413303852081, + 'predict_loss': 0.012781817466020584, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24698381125926971, + 'data_time': 0.0006701709935441613, + 'model_time': 1.1555175149987917, + 'grad_norm_pre_clip_avg': 0.2438378155231476, + 'learning_rate': 2.3789157298956892e-05, + 'epoch': 2.87} +04/19 [15:25:54] INFO | >> train_qwenlatent.py:487 + Step 11370 | grad_norm_pre_clip=0.1700 | + grad_norm_pre_clip_avg=0.2399 | Metrics: + {'align_loss': 0.024266578257083893, + 'recon_loss': 0.04678911715745926, + 'predict_loss': 0.01775376684963703, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17000870406627655, + 'data_time': 0.0006457109993789345, + 'model_time': 1.1545320260047447, + 'grad_norm_pre_clip_avg': 0.23988818675279616, + 'learning_rate': 2.3785409632177533e-05, + 'epoch': 2.87} +04/19 [15:26:06] INFO | >> train_qwenlatent.py:487 + Step 11380 | grad_norm_pre_clip=0.2424 | + grad_norm_pre_clip_avg=0.2296 | Metrics: + {'align_loss': 0.025262542068958282, + 'recon_loss': 0.05298334360122681, + 'predict_loss': 0.023160209879279137, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24237629771232605, + 'data_time': 0.0006157099851407111, + 'model_time': 1.191743609990226, + 'grad_norm_pre_clip_avg': 0.2296499192714691, + 'learning_rate': 2.3781656471118966e-05, + 'epoch': 2.87} +04/19 [15:26:17] INFO | >> train_qwenlatent.py:487 + Step 11390 | grad_norm_pre_clip=0.2010 | + grad_norm_pre_clip_avg=0.2364 | Metrics: + {'align_loss': 0.024650750681757927, + 'recon_loss': 0.04083316773176193, + 'predict_loss': 0.017326561734080315, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20100465416908264, + 'data_time': 0.0008930189942475408, + 'model_time': 1.1508805009943899, + 'grad_norm_pre_clip_avg': 0.23640167862176895, + 'learning_rate': 2.3777897817610434e-05, + 'epoch': 2.87} +04/19 [15:26:30] INFO | >> train_qwenlatent.py:487 + Step 11400 | grad_norm_pre_clip=0.2723 | + grad_norm_pre_clip_avg=0.3107 | Metrics: + {'align_loss': 0.02331453375518322, + 'recon_loss': 0.035273488610982895, + 'predict_loss': 0.014597060158848763, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2723138928413391, + 'mae_score': 0.019663975689862227, 'data_time': + 0.0005936059751547873, 'model_time': + 1.1428828539792448, 'grad_norm_pre_clip_avg': + 0.31065104752779005, 'learning_rate': + 2.3774133673483862e-05, 'epoch': 2.88} +04/19 [15:26:41] INFO | >> train_qwenlatent.py:487 + Step 11410 | grad_norm_pre_clip=0.2844 | + grad_norm_pre_clip_avg=0.2938 | Metrics: + {'align_loss': 0.024432118982076645, + 'recon_loss': 0.0392531119287014, + 'predict_loss': 0.016099167987704277, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28442275524139404, + 'data_time': 0.0007913529989309609, + 'model_time': 1.138631670997711, + 'grad_norm_pre_clip_avg': 0.2937918439507484, + 'learning_rate': 2.377036404057384e-05, + 'epoch': 2.88} +04/19 [15:26:53] INFO | >> train_qwenlatent.py:487 + Step 11420 | grad_norm_pre_clip=0.2518 | + grad_norm_pre_clip_avg=0.2367 | Metrics: + {'align_loss': 0.024197788909077644, + 'recon_loss': 0.03810923919081688, + 'predict_loss': 0.01629205420613289, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25182464718818665, + 'data_time': 0.0006185060192365199, + 'model_time': 1.1586843410041183, + 'grad_norm_pre_clip_avg': 0.23669524490833282, + 'learning_rate': 2.376658892071765e-05, + 'epoch': 2.88} +04/19 [15:27:05] INFO | >> train_qwenlatent.py:487 + Step 11430 | grad_norm_pre_clip=0.2248 | + grad_norm_pre_clip_avg=0.2349 | Metrics: + {'align_loss': 0.02529763989150524, + 'recon_loss': 0.04133664071559906, + 'predict_loss': 0.01510833203792572, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22484926879405975, + 'data_time': 0.0006815899978391826, + 'model_time': 1.136524533998454, + 'grad_norm_pre_clip_avg': 0.23487380594015123, + 'learning_rate': 2.3762808315755235e-05, + 'epoch': 2.88} +04/19 [15:27:16] INFO | >> train_qwenlatent.py:487 + Step 11440 | grad_norm_pre_clip=0.2682 | + grad_norm_pre_clip_avg=0.2489 | Metrics: + {'align_loss': 0.024494878947734833, + 'recon_loss': 0.03595976531505585, + 'predict_loss': 0.012715992517769337, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2681688666343689, + 'data_time': 0.0006365870067384094, + 'model_time': 1.1708119249960873, + 'grad_norm_pre_clip_avg': 0.24886461943387986, + 'learning_rate': 2.3759022227529222e-05, + 'epoch': 2.89} +04/19 [15:27:28] INFO | >> train_qwenlatent.py:487 + Step 11450 | grad_norm_pre_clip=0.2189 | + grad_norm_pre_clip_avg=0.1946 | Metrics: + {'align_loss': 0.025463756173849106, + 'recon_loss': 0.05761129781603813, + 'predict_loss': 0.018360253423452377, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21888592839241028, + 'mae_score': 0.016866185643651464, 'data_time': + 0.0005897990195080638, 'model_time': + 1.1565764179977123, 'grad_norm_pre_clip_avg': + 0.19455854892730712, 'learning_rate': + 2.3755230657884894e-05, 'epoch': 2.89} +04/19 [15:27:40] INFO | >> train_qwenlatent.py:487 + Step 11460 | grad_norm_pre_clip=0.2147 | + grad_norm_pre_clip_avg=0.2370 | Metrics: + {'align_loss': 0.023961979895830154, + 'recon_loss': 0.026078199967741966, + 'predict_loss': 0.009976287372410297, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21468059718608856, + 'data_time': 0.0006095799908507615, + 'model_time': 1.1316669060033746, + 'grad_norm_pre_clip_avg': 0.23702749609947205, + 'learning_rate': 2.3751433608670223e-05, + 'epoch': 2.89} +04/19 [15:27:52] INFO | >> train_qwenlatent.py:487 + Step 11470 | grad_norm_pre_clip=0.2569 | + grad_norm_pre_clip_avg=0.2448 | Metrics: + {'align_loss': 0.02511720173060894, + 'recon_loss': 0.04536909982562065, + 'predict_loss': 0.01662837341427803, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25688666105270386, + 'data_time': 0.0009177500032819808, + 'model_time': 1.1743846889876295, + 'grad_norm_pre_clip_avg': 0.24476297050714493, + 'learning_rate': 2.3747631081735844e-05, + 'epoch': 2.89} +04/19 [15:28:04] INFO | >> train_qwenlatent.py:487 + Step 11480 | grad_norm_pre_clip=0.1756 | + grad_norm_pre_clip_avg=0.1991 | Metrics: + {'align_loss': 0.023889193311333656, + 'recon_loss': 0.04686341434717178, + 'predict_loss': 0.020179007202386856, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17557768523693085, + 'data_time': 0.0005651809915434569, + 'model_time': 1.3841353699972387, + 'grad_norm_pre_clip_avg': 0.19912292510271073, + 'learning_rate': 2.374382307893506e-05, + 'epoch': 2.9} +04/19 [15:28:16] INFO | >> train_qwenlatent.py:487 + Step 11490 | grad_norm_pre_clip=0.2928 | + grad_norm_pre_clip_avg=0.3359 | Metrics: + {'align_loss': 0.022399483248591423, + 'recon_loss': 0.033021364361047745, + 'predict_loss': 0.015328969806432724, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2927641272544861, + 'data_time': 0.0005490960029419512, + 'model_time': 1.1624714700155891, + 'grad_norm_pre_clip_avg': 0.3359219327569008, + 'learning_rate': 2.3740009602123848e-05, + 'epoch': 2.9} +04/19 [15:28:28] INFO | >> train_qwenlatent.py:487 + Step 11500 | grad_norm_pre_clip=0.2071 | + grad_norm_pre_clip_avg=0.2675 | Metrics: + {'align_loss': 0.024155747145414352, + 'recon_loss': 0.04637345299124718, + 'predict_loss': 0.013910369947552681, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20708370208740234, + 'mae_score': 0.03154909632227442, 'data_time': + 0.0005371029837988317, 'model_time': + 1.1417188560008071, 'grad_norm_pre_clip_avg': + 0.2675171971321106, 'learning_rate': + 2.373619065316085e-05, 'epoch': 2.9} +04/19 [15:28:40] INFO | >> train_qwenlatent.py:487 + Step 11510 | grad_norm_pre_clip=0.2218 | + grad_norm_pre_clip_avg=0.2270 | Metrics: + {'align_loss': 0.023214083164930344, + 'recon_loss': 0.038667090237140656, + 'predict_loss': 0.01668640226125717, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22183488309383392, + 'data_time': 0.0006144490034785122, + 'model_time': 1.3452024740108754, + 'grad_norm_pre_clip_avg': 0.22701835930347442, + 'learning_rate': 2.3732366233907373e-05, + 'epoch': 2.9} +04/19 [15:28:51] INFO | >> train_qwenlatent.py:487 + Step 11520 | grad_norm_pre_clip=0.2282 | + grad_norm_pre_clip_avg=0.2190 | Metrics: + {'align_loss': 0.024786096066236496, + 'recon_loss': 0.04348054528236389, + 'predict_loss': 0.012778027914464474, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22819432616233826, + 'data_time': 0.0006092849944252521, + 'model_time': 1.1459990449948236, + 'grad_norm_pre_clip_avg': 0.2190327137708664, + 'learning_rate': 2.3728536346227394e-05, + 'epoch': 2.91} +04/19 [15:29:03] INFO | >> train_qwenlatent.py:487 + Step 11530 | grad_norm_pre_clip=0.2245 | + grad_norm_pre_clip_avg=0.2083 | Metrics: + {'align_loss': 0.025070959702134132, + 'recon_loss': 0.04515543207526207, + 'predict_loss': 0.015044032596051693, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22454842925071716, + 'data_time': 0.0005611600063275546, + 'model_time': 1.1530100039963145, + 'grad_norm_pre_clip_avg': 0.20828402638435364, + 'learning_rate': 2.3724700991987556e-05, + 'epoch': 2.91} +04/19 [15:29:14] INFO | >> train_qwenlatent.py:487 + Step 11540 | grad_norm_pre_clip=0.3556 | + grad_norm_pre_clip_avg=0.2402 | Metrics: + {'align_loss': 0.02624392695724964, + 'recon_loss': 0.05601407214999199, + 'predict_loss': 0.01615016721189022, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3555854558944702, + 'data_time': 0.0005613120156340301, + 'model_time': 1.1602605959924404, + 'grad_norm_pre_clip_avg': 0.24017657041549684, + 'learning_rate': 2.372086017305716e-05, + 'epoch': 2.91} +04/19 [15:29:26] INFO | >> train_qwenlatent.py:487 + Step 11550 | grad_norm_pre_clip=0.2722 | + grad_norm_pre_clip_avg=0.2637 | Metrics: + {'align_loss': 0.025345871224999428, + 'recon_loss': 0.046650614589452744, + 'predict_loss': 0.017308088019490242, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2722131907939911, + 'mae_score': 0.019264396461280617, 'data_time': + 0.0005637019930873066, 'model_time': + 1.1372378740052227, 'grad_norm_pre_clip_avg': + 0.26371479481458665, 'learning_rate': + 2.3717013891308175e-05, 'epoch': 2.91} +04/19 [15:29:38] INFO | >> train_qwenlatent.py:487 + Step 11560 | grad_norm_pre_clip=0.2008 | + grad_norm_pre_clip_avg=0.2422 | Metrics: + {'align_loss': 0.024642841890454292, + 'recon_loss': 0.03565389662981033, + 'predict_loss': 0.012008162215352058, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2008059322834015, + 'data_time': 0.0005621819873340428, + 'model_time': 1.152992903982522, + 'grad_norm_pre_clip_avg': 0.24223079830408095, + 'learning_rate': 2.3713162148615235e-05, + 'epoch': 2.92} +04/19 [15:29:49] INFO | >> train_qwenlatent.py:487 + Step 11570 | grad_norm_pre_clip=0.1858 | + grad_norm_pre_clip_avg=0.2271 | Metrics: + {'align_loss': 0.024182511493563652, + 'recon_loss': 0.03673020750284195, + 'predict_loss': 0.016802720725536346, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1858207881450653, + 'data_time': 0.0006792830245103687, + 'model_time': 1.1451822070230264, + 'grad_norm_pre_clip_avg': 0.22706408351659774, + 'learning_rate': 2.3709304946855636e-05, + 'epoch': 2.92} +04/19 [15:30:01] INFO | >> train_qwenlatent.py:487 + Step 11580 | grad_norm_pre_clip=0.1975 | + grad_norm_pre_clip_avg=0.2158 | Metrics: + {'align_loss': 0.02492723986506462, + 'recon_loss': 0.04657896235585213, + 'predict_loss': 0.01730671525001526, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19745512306690216, + 'data_time': 0.0005898620001971722, + 'model_time': 1.1546630829980131, + 'grad_norm_pre_clip_avg': 0.2158182829618454, + 'learning_rate': 2.3705442287909318e-05, + 'epoch': 2.92} +04/19 [15:30:13] INFO | >> train_qwenlatent.py:487 + Step 11590 | grad_norm_pre_clip=0.3416 | + grad_norm_pre_clip_avg=0.2263 | Metrics: + {'align_loss': 0.02495318278670311, + 'recon_loss': 0.03929348289966583, + 'predict_loss': 0.01221657358109951, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.34158626198768616, + 'data_time': 0.0006022030138410628, + 'model_time': 1.150091525982134, + 'grad_norm_pre_clip_avg': 0.2263341024518013, + 'learning_rate': 2.3701574173658904e-05, + 'epoch': 2.92} +04/19 [15:30:25] INFO | >> train_qwenlatent.py:487 + Step 11600 | grad_norm_pre_clip=0.2918 | + grad_norm_pre_clip_avg=0.3354 | Metrics: + {'align_loss': 0.024943340569734573, + 'recon_loss': 0.05260121077299118, + 'predict_loss': 0.017591187730431557, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2918062210083008, + 'mae_score': 0.018579003617570207, 'data_time': + 0.0005847470019944012, 'model_time': + 1.1588943300012033, 'grad_norm_pre_clip_avg': + 0.3354240506887436, 'learning_rate': + 2.369770060598967e-05, 'epoch': 2.93} +04/19 [15:30:36] INFO | >> train_qwenlatent.py:487 + Step 11610 | grad_norm_pre_clip=0.1753 | + grad_norm_pre_clip_avg=0.2554 | Metrics: + {'align_loss': 0.024956224486231804, + 'recon_loss': 0.050800714641809464, + 'predict_loss': 0.01630127616226673, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17529603838920593, + 'data_time': 0.0005934579821769148, + 'model_time': 1.1662829290144145, + 'grad_norm_pre_clip_avg': 0.2553627774119377, + 'learning_rate': 2.3693821586789536e-05, + 'epoch': 2.93} +04/19 [15:30:48] INFO | >> train_qwenlatent.py:487 + Step 11620 | grad_norm_pre_clip=0.1930 | + grad_norm_pre_clip_avg=0.2027 | Metrics: + {'align_loss': 0.02491721883416176, + 'recon_loss': 0.04128105193376541, + 'predict_loss': 0.01219974271953106, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19301339983940125, + 'data_time': 0.0005951150087639689, + 'model_time': 1.1449792690109462, + 'grad_norm_pre_clip_avg': 0.20268780291080474, + 'learning_rate': 2.368993711794909e-05, + 'epoch': 2.93} +04/19 [15:31:00] INFO | >> train_qwenlatent.py:487 + Step 11630 | grad_norm_pre_clip=0.1960 | + grad_norm_pre_clip_avg=0.2143 | Metrics: + {'align_loss': 0.02419123612344265, + 'recon_loss': 0.045014310628175735, + 'predict_loss': 0.018498264253139496, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19596248865127563, + 'data_time': 0.0008250009850598872, + 'model_time': 1.1444437880127225, + 'grad_norm_pre_clip_avg': 0.21427139788866043, + 'learning_rate': 2.3686047201361583e-05, + 'epoch': 2.93} +04/19 [15:31:11] INFO | >> train_qwenlatent.py:487 + Step 11640 | grad_norm_pre_clip=0.3555 | + grad_norm_pre_clip_avg=0.2685 | Metrics: + {'align_loss': 0.024945978075265884, + 'recon_loss': 0.047939859330654144, + 'predict_loss': 0.014492069371044636, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.35549139976501465, + 'data_time': 0.0005413010076154023, + 'model_time': 1.1396170999796595, + 'grad_norm_pre_clip_avg': 0.2684624150395393, + 'learning_rate': 2.3682151838922906e-05, + 'epoch': 2.94} +04/19 [15:31:24] INFO | >> train_qwenlatent.py:487 + Step 11650 | grad_norm_pre_clip=0.3045 | + grad_norm_pre_clip_avg=0.2857 | Metrics: + {'align_loss': 0.024564556777477264, + 'recon_loss': 0.04012962430715561, + 'predict_loss': 0.013581719249486923, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.304493248462677, + 'mae_score': 0.015171284718556447, 'data_time': + 0.000526770978467539, 'model_time': + 1.1449030399962794, 'grad_norm_pre_clip_avg': + 0.2856732353568077, 'learning_rate': + 2.367825103253161e-05, 'epoch': 2.94} +04/19 [15:32:05] INFO | >> train_qwenlatent.py:487 + Step 11660 | grad_norm_pre_clip=0.2430 | + grad_norm_pre_clip_avg=0.2463 | Metrics: + {'align_loss': 0.025079507380723953, + 'recon_loss': 0.03723493218421936, + 'predict_loss': 0.009298795834183693, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24299952387809753, + 'data_time': 0.002728571998886764, + 'model_time': 3.7679885959951207, + 'grad_norm_pre_clip_avg': 0.24633956104516982, + 'learning_rate': 2.3674344784088907e-05, + 'epoch': 2.94} +04/19 [15:32:42] INFO | >> train_qwenlatent.py:487 + Step 11670 | grad_norm_pre_clip=0.2921 | + grad_norm_pre_clip_avg=0.2429 | Metrics: + {'align_loss': 0.023585120216012, 'recon_loss': + 0.035744767636060715, 'predict_loss': + 0.02030182257294655, 'aux_loss_decay_weight': + 0.0, 'grad_norm_pre_clip': 0.29209381341934204, + 'data_time': 0.007223321998026222, + 'model_time': 2.885595822997857, + 'grad_norm_pre_clip_avg': 0.24294065088033676, + 'learning_rate': 2.367043309549865e-05, + 'epoch': 2.94} +04/19 [15:33:18] INFO | >> train_qwenlatent.py:487 + Step 11680 | grad_norm_pre_clip=0.2237 | + grad_norm_pre_clip_avg=0.2585 | Metrics: + {'align_loss': 0.023456597700715065, + 'recon_loss': 0.034719087183475494, + 'predict_loss': 0.013495898805558681, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2236950397491455, + 'data_time': 0.001427176990546286, + 'model_time': 3.2121917250042316, + 'grad_norm_pre_clip_avg': 0.2584924057126045, + 'learning_rate': 2.3666515968667355e-05, + 'epoch': 2.95} +04/19 [15:33:54] INFO | >> train_qwenlatent.py:487 + Step 11690 | grad_norm_pre_clip=0.2349 | + grad_norm_pre_clip_avg=0.2521 | Metrics: + {'align_loss': 0.024461861699819565, + 'recon_loss': 0.047883663326501846, + 'predict_loss': 0.013645878061652184, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23486556112766266, + 'data_time': 0.001204851985676214, + 'model_time': 3.61181584597216, + 'grad_norm_pre_clip_avg': 0.25207240134477615, + 'learning_rate': 2.3662593405504173e-05, + 'epoch': 2.95} +04/19 [15:34:31] INFO | >> train_qwenlatent.py:487 + Step 11700 | grad_norm_pre_clip=0.2196 | + grad_norm_pre_clip_avg=0.2669 | Metrics: + {'align_loss': 0.024796728044748306, + 'recon_loss': 0.03286265581846237, + 'predict_loss': 0.013469772413372993, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21962454915046692, + 'mae_score': 0.027873202487155125, 'data_time': + 0.0016298640111926943, 'model_time': + 3.3221047869883478, 'grad_norm_pre_clip_avg': + 0.26692855209112165, 'learning_rate': + 2.365866540792092e-05, 'epoch': 2.95} +04/19 [15:35:10] INFO | >> train_qwenlatent.py:487 + Step 11710 | grad_norm_pre_clip=0.2677 | + grad_norm_pre_clip_avg=0.2535 | Metrics: + {'align_loss': 0.025296997278928757, + 'recon_loss': 0.05179069563746452, + 'predict_loss': 0.020647995173931122, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2676907181739807, + 'data_time': 0.0010982689855154604, + 'model_time': 3.683411862992216, + 'grad_norm_pre_clip_avg': 0.2534772664308548, + 'learning_rate': 2.365473197783205e-05, + 'epoch': 2.95} +04/19 [15:35:40] INFO | >> train_qwenlatent.py:487 + Step 11720 | grad_norm_pre_clip=0.2297 | + grad_norm_pre_clip_avg=0.2395 | Metrics: + {'align_loss': 0.02489030733704567, + 'recon_loss': 0.03011508285999298, + 'predict_loss': 0.011224408634006977, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2297462671995163, + 'data_time': 0.001180924999061972, + 'model_time': 2.7893260740092956, + 'grad_norm_pre_clip_avg': 0.23946198970079421, + 'learning_rate': 2.365079311715467e-05, + 'epoch': 2.96} +04/19 [15:36:07] INFO | >> train_qwenlatent.py:487 + Step 11730 | grad_norm_pre_clip=0.3273 | + grad_norm_pre_clip_avg=0.2868 | Metrics: + {'align_loss': 0.024109354242682457, + 'recon_loss': 0.04151931777596474, + 'predict_loss': 0.011179054155945778, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.32730093598365784, + 'data_time': 0.0011047259904444218, + 'model_time': 2.4081638179777656, + 'grad_norm_pre_clip_avg': 0.28679427355527876, + 'learning_rate': 2.364684882780854e-05, + 'epoch': 2.96} +04/19 [15:36:26] INFO | >> train_qwenlatent.py:487 + Step 11740 | grad_norm_pre_clip=0.2368 | + grad_norm_pre_clip_avg=0.2263 | Metrics: + {'align_loss': 0.024530723690986633, + 'recon_loss': 0.04277399927377701, + 'predict_loss': 0.014373325742781162, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23682616651058197, + 'data_time': 0.0010586770076770335, + 'model_time': 1.5844046009879094, + 'grad_norm_pre_clip_avg': 0.22632991671562194, + 'learning_rate': 2.3642899111716047e-05, + 'epoch': 2.96} +04/19 [15:36:41] INFO | >> train_qwenlatent.py:487 + Step 11750 | grad_norm_pre_clip=0.1833 | + grad_norm_pre_clip_avg=0.2308 | Metrics: + {'align_loss': 0.022832274436950684, + 'recon_loss': 0.04226730391383171, + 'predict_loss': 0.013437546789646149, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18327973783016205, + 'mae_score': 0.01590809865040822, 'data_time': + 0.0006610249984078109, 'model_time': + 1.2256639910046943, 'grad_norm_pre_clip_avg': + 0.23081738352775574, 'learning_rate': + 2.3638943970802244e-05, 'epoch': 2.96} +04/19 [15:36:54] INFO | >> train_qwenlatent.py:487 + Step 11760 | grad_norm_pre_clip=0.3407 | + grad_norm_pre_clip_avg=0.2550 | Metrics: + {'align_loss': 0.025717858225107193, + 'recon_loss': 0.05094778537750244, + 'predict_loss': 0.014084548689424992, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.34066668152809143, + 'data_time': 0.0006867139891255647, + 'model_time': 1.2526422919763718, + 'grad_norm_pre_clip_avg': 0.25500351637601854, + 'learning_rate': 2.3634983406994815e-05, + 'epoch': 2.97} +04/19 [15:37:06] INFO | >> train_qwenlatent.py:487 + Step 11770 | grad_norm_pre_clip=0.2379 | + grad_norm_pre_clip_avg=0.2524 | Metrics: + {'align_loss': 0.02502375654876232, + 'recon_loss': 0.037713050842285156, + 'predict_loss': 0.01347387209534645, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.237946555018425, + 'data_time': 0.000649413006613031, + 'model_time': 1.318117977003567, + 'grad_norm_pre_clip_avg': 0.25241747200489045, + 'learning_rate': 2.3631017422224092e-05, + 'epoch': 2.97} +04/19 [15:37:19] INFO | >> train_qwenlatent.py:487 + Step 11780 | grad_norm_pre_clip=0.2436 | + grad_norm_pre_clip_avg=0.2446 | Metrics: + {'align_loss': 0.024894554167985916, + 'recon_loss': 0.04352579265832901, + 'predict_loss': 0.012628305703401566, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24362704157829285, + 'data_time': 0.0006226279947441071, + 'model_time': 1.1971456029859837, + 'grad_norm_pre_clip_avg': 0.24459467381238936, + 'learning_rate': 2.362704601842304e-05, + 'epoch': 2.97} +04/19 [15:37:31] INFO | >> train_qwenlatent.py:487 + Step 11790 | grad_norm_pre_clip=0.3250 | + grad_norm_pre_clip_avg=0.2568 | Metrics: + {'align_loss': 0.025865228846669197, + 'recon_loss': 0.04938076063990593, + 'predict_loss': 0.019583962857723236, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.32501301169395447, + 'data_time': 0.0006144919898360968, + 'model_time': 1.1955378240090795, + 'grad_norm_pre_clip_avg': 0.2567921504378319, + 'learning_rate': 2.362306919752729e-05, + 'epoch': 2.98} +04/19 [15:37:45] INFO | >> train_qwenlatent.py:487 + Step 11800 | grad_norm_pre_clip=0.2255 | + grad_norm_pre_clip_avg=0.2547 | Metrics: + {'align_loss': 0.025081012398004532, + 'recon_loss': 0.059098150581121445, + 'predict_loss': 0.02146092988550663, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22549767792224884, + 'mae_score': 0.02053125914152678, 'data_time': + 0.000960923993261531, 'model_time': + 1.3092602739925496, 'grad_norm_pre_clip_avg': + 0.2547142505645752, 'learning_rate': + 2.361908696147508e-05, 'epoch': 2.98} +04/19 [15:37:57] INFO | >> train_qwenlatent.py:487 + Step 11810 | grad_norm_pre_clip=0.2246 | + grad_norm_pre_clip_avg=0.2158 | Metrics: + {'align_loss': 0.025223547592759132, + 'recon_loss': 0.039850667119026184, + 'predict_loss': 0.0120526859536767, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2246164232492447, + 'data_time': 0.0006080940074753016, + 'model_time': 1.2140517869847827, + 'grad_norm_pre_clip_avg': 0.21577103883028032, + 'learning_rate': 2.361509931220731e-05, + 'epoch': 2.98} +04/19 [15:38:10] INFO | >> train_qwenlatent.py:487 + Step 11820 | grad_norm_pre_clip=0.2918 | + grad_norm_pre_clip_avg=0.2093 | Metrics: + {'align_loss': 0.02356400340795517, + 'recon_loss': 0.04244344308972359, + 'predict_loss': 0.013302582316100597, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.291780948638916, + 'data_time': 0.0009025750041473657, + 'model_time': 1.2391322839830536, + 'grad_norm_pre_clip_avg': 0.20930069983005523, + 'learning_rate': 2.361110625166751e-05, + 'epoch': 2.98} +04/19 [15:38:22] INFO | >> train_qwenlatent.py:487 + Step 11830 | grad_norm_pre_clip=0.2382 | + grad_norm_pre_clip_avg=0.2959 | Metrics: + {'align_loss': 0.023568322882056236, + 'recon_loss': 0.03164072707295418, + 'predict_loss': 0.010504486039280891, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23817658424377441, + 'data_time': 0.001056389999575913, + 'model_time': 1.2291092250088695, + 'grad_norm_pre_clip_avg': 0.29585709869861604, + 'learning_rate': 2.360710778180185e-05, + 'epoch': 2.99} +04/19 [15:38:35] INFO | >> train_qwenlatent.py:487 + Step 11840 | grad_norm_pre_clip=0.2807 | + grad_norm_pre_clip_avg=0.2287 | Metrics: + {'align_loss': 0.022564202547073364, + 'recon_loss': 0.047172050923109055, + 'predict_loss': 0.017009178176522255, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28069186210632324, + 'data_time': 0.0006270980229601264, + 'model_time': 1.277731397014577, + 'grad_norm_pre_clip_avg': 0.22873543351888656, + 'learning_rate': 2.360310390455914e-05, + 'epoch': 2.99} +04/19 [15:38:48] INFO | >> train_qwenlatent.py:487 + Step 11850 | grad_norm_pre_clip=0.1668 | + grad_norm_pre_clip_avg=0.2069 | Metrics: + {'align_loss': 0.02322673611342907, + 'recon_loss': 0.03820953890681267, + 'predict_loss': 0.012539805844426155, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16678829491138458, + 'mae_score': 0.015395351787945172, 'data_time': + 0.0010716510005295277, 'model_time': + 1.246829796989914, 'grad_norm_pre_clip_avg': + 0.20692666471004487, 'learning_rate': + 2.359909462189081e-05, 'epoch': 2.99} +04/19 [15:39:01] INFO | >> train_qwenlatent.py:487 + Step 11860 | grad_norm_pre_clip=0.3356 | + grad_norm_pre_clip_avg=0.2541 | Metrics: + {'align_loss': 0.023845165967941284, + 'recon_loss': 0.04419019818305969, + 'predict_loss': 0.017309412360191345, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.33560001850128174, + 'data_time': 0.0008518669928889722, + 'model_time': 1.2485181069932878, + 'grad_norm_pre_clip_avg': 0.25409370809793475, + 'learning_rate': 2.359507993575095e-05, + 'epoch': 2.99} +04/19 [15:39:14] INFO | >> train_qwenlatent.py:487 + Step 11870 | grad_norm_pre_clip=0.2948 | + grad_norm_pre_clip_avg=0.2663 | Metrics: + {'align_loss': 0.024783587083220482, + 'recon_loss': 0.04843177646398544, + 'predict_loss': 0.023000895977020264, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.29478394985198975, + 'data_time': 0.0008946849848143756, + 'model_time': 1.2815843150019646, + 'grad_norm_pre_clip_avg': 0.26627491861581803, + 'learning_rate': 2.3591059848096258e-05, + 'epoch': 3.0} +04/19 [15:39:27] INFO | >> train_qwenlatent.py:487 + Step 11880 | grad_norm_pre_clip=0.1930 | + grad_norm_pre_clip_avg=0.2214 | Metrics: + {'align_loss': 0.02294471114873886, + 'recon_loss': 0.047163546085357666, + 'predict_loss': 0.015132348984479904, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19303099811077118, + 'data_time': 0.0007348770159296691, + 'model_time': 1.2137388839910273, + 'grad_norm_pre_clip_avg': 0.22141436636447906, + 'learning_rate': 2.358703436088608e-05, + 'epoch': 3.0} +04/19 [15:39:39] INFO | >> train_qwenlatent.py:487 + Step 11890 | grad_norm_pre_clip=0.1987 | + grad_norm_pre_clip_avg=0.2060 | Metrics: + {'align_loss': 0.02435266599059105, + 'recon_loss': 0.04694552347064018, + 'predict_loss': 0.018231036141514778, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19866277277469635, + 'data_time': 0.0006072699907235801, + 'model_time': 1.2122910959878936, + 'grad_norm_pre_clip_avg': 0.20602543652057648, + 'learning_rate': 2.3583003476082392e-05, + 'epoch': 3.0} +04/19 [15:39:52] INFO | >> train_qwenlatent.py:487 + Step 11900 | grad_norm_pre_clip=0.2032 | + grad_norm_pre_clip_avg=0.2202 | Metrics: + {'align_loss': 0.024175507947802544, + 'recon_loss': 0.03430914878845215, + 'predict_loss': 0.011692170985043049, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20321959257125854, + 'mae_score': 0.013808872463466884, 'data_time': + 0.0006493190012406558, 'model_time': + 1.2524978829897009, 'grad_norm_pre_clip_avg': + 0.22021838873624802, 'learning_rate': + 2.3578967195649793e-05, 'epoch': 3.0} +04/19 [15:40:05] INFO | >> train_qwenlatent.py:487 + Step 11910 | grad_norm_pre_clip=0.4291 | + grad_norm_pre_clip_avg=0.3900 | Metrics: + {'align_loss': 0.02486947551369667, + 'recon_loss': 0.04666881263256073, + 'predict_loss': 0.014535533264279366, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.4290827512741089, + 'data_time': 0.0006183049990795553, + 'model_time': 1.2371399299881887, + 'grad_norm_pre_clip_avg': 0.38997637927532197, + 'learning_rate': 2.3574925521555525e-05, + 'epoch': 3.01} +04/19 [15:40:17] INFO | >> train_qwenlatent.py:487 + Step 11920 | grad_norm_pre_clip=0.2146 | + grad_norm_pre_clip_avg=0.2856 | Metrics: + {'align_loss': 0.025029877200722694, + 'recon_loss': 0.0462750643491745, + 'predict_loss': 0.012048441916704178, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2145988494157791, + 'data_time': 0.0006544709904119372, + 'model_time': 1.231535586004611, + 'grad_norm_pre_clip_avg': 0.2856287479400635, + 'learning_rate': 2.357087845576944e-05, + 'epoch': 3.01} +04/19 [15:40:30] INFO | >> train_qwenlatent.py:487 + Step 11930 | grad_norm_pre_clip=0.2192 | + grad_norm_pre_clip_avg=0.2241 | Metrics: + {'align_loss': 0.023538654670119286, + 'recon_loss': 0.045677732676267624, + 'predict_loss': 0.02076447196304798, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21918141841888428, + 'data_time': 0.0008941309934016317, + 'model_time': 1.2906759920006152, + 'grad_norm_pre_clip_avg': 0.22411883026361465, + 'learning_rate': 2.3566826000264038e-05, + 'epoch': 3.01} +04/19 [15:40:42] INFO | >> train_qwenlatent.py:487 + Step 11940 | grad_norm_pre_clip=0.2251 | + grad_norm_pre_clip_avg=0.2237 | Metrics: + {'align_loss': 0.024654310196638107, + 'recon_loss': 0.04922715574502945, + 'predict_loss': 0.018003368750214577, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22512049973011017, + 'data_time': 0.0006137919845059514, + 'model_time': 1.2101472529757302, + 'grad_norm_pre_clip_avg': 0.22365795075893402, + 'learning_rate': 2.356276815701443e-05, + 'epoch': 3.01} +04/19 [15:40:56] INFO | >> train_qwenlatent.py:487 + Step 11950 | grad_norm_pre_clip=0.2089 | + grad_norm_pre_clip_avg=0.2467 | Metrics: + {'align_loss': 0.02440815605223179, + 'recon_loss': 0.043261878192424774, + 'predict_loss': 0.011784739792346954, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2089363932609558, + 'mae_score': 0.021607177321975295, 'data_time': + 0.0006027689960319549, 'model_time': + 1.2322945600026287, 'grad_norm_pre_clip_avg': + 0.24672389775514603, 'learning_rate': + 2.3558704927998363e-05, 'epoch': 3.02} +04/19 [15:41:08] INFO | >> train_qwenlatent.py:487 + Step 11960 | grad_norm_pre_clip=0.2025 | + grad_norm_pre_clip_avg=0.2440 | Metrics: + {'align_loss': 0.025109488517045975, + 'recon_loss': 0.04840730503201485, + 'predict_loss': 0.015359438955783844, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20250315964221954, + 'data_time': 0.0006454900139942765, + 'model_time': 1.2476875080028549, + 'grad_norm_pre_clip_avg': 0.24402011930942535, + 'learning_rate': 2.3554636315196204e-05, + 'epoch': 3.02} +04/19 [15:41:21] INFO | >> train_qwenlatent.py:487 + Step 11970 | grad_norm_pre_clip=0.2597 | + grad_norm_pre_clip_avg=0.2415 | Metrics: + {'align_loss': 0.02295677177608013, + 'recon_loss': 0.038721341639757156, + 'predict_loss': 0.011617557145655155, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.259708970785141, + 'data_time': 0.0007446829986292869, + 'model_time': 1.321707997994963, + 'grad_norm_pre_clip_avg': 0.24145700335502623, + 'learning_rate': 2.355056232059095e-05, + 'epoch': 3.02} +04/19 [15:41:34] INFO | >> train_qwenlatent.py:487 + Step 11980 | grad_norm_pre_clip=0.2040 | + grad_norm_pre_clip_avg=0.1952 | Metrics: + {'align_loss': 0.025144558399915695, + 'recon_loss': 0.04974229261279106, + 'predict_loss': 0.01458588894456625, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20401595532894135, + 'data_time': 0.0006361639825627208, + 'model_time': 1.2993780430115294, + 'grad_norm_pre_clip_avg': 0.19520095437765123, + 'learning_rate': 2.354648294616821e-05, + 'epoch': 3.02} +04/19 [15:41:46] INFO | >> train_qwenlatent.py:487 + Step 11990 | grad_norm_pre_clip=0.2617 | + grad_norm_pre_clip_avg=0.2560 | Metrics: + {'align_loss': 0.025477174669504166, + 'recon_loss': 0.05976080149412155, + 'predict_loss': 0.015211229212582111, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2616611421108246, + 'data_time': 0.000621982995653525, + 'model_time': 1.2463423110020813, + 'grad_norm_pre_clip_avg': 0.2560016825795174, + 'learning_rate': 2.3542398193916224e-05, + 'epoch': 3.03} +04/19 [15:41:59] INFO | >> train_qwenlatent.py:487 + Step 12000 | grad_norm_pre_clip=0.2426 | + grad_norm_pre_clip_avg=0.2520 | Metrics: + {'align_loss': 0.023061059415340424, + 'recon_loss': 0.03512836620211601, + 'predict_loss': 0.011609550565481186, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24257268011569977, + 'mae_score': 0.021624680252762526, 'data_time': + 0.0008587050251662731, 'model_time': + 1.2646258000168018, 'grad_norm_pre_clip_avg': + 0.251998333632946, 'learning_rate': + 2.353830806582585e-05, 'epoch': 3.03} +04/19 [15:42:12] INFO | >> train_qwenlatent.py:487 + Step 12010 | grad_norm_pre_clip=0.2048 | + grad_norm_pre_clip_avg=0.2023 | Metrics: + {'align_loss': 0.024123933166265488, + 'recon_loss': 0.039026908576488495, + 'predict_loss': 0.013577772304415703, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2047898918390274, + 'data_time': 0.0009853739757090807, + 'model_time': 1.2790225750068203, + 'grad_norm_pre_clip_avg': 0.20226171612739563, + 'learning_rate': 2.3534212563890574e-05, + 'epoch': 3.03} +04/19 [15:42:24] INFO | >> train_qwenlatent.py:487 + Step 12020 | grad_norm_pre_clip=0.2166 | + grad_norm_pre_clip_avg=0.2540 | Metrics: + {'align_loss': 0.0248374305665493, + 'recon_loss': 0.04224862530827522, + 'predict_loss': 0.014239768497645855, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2165973037481308, + 'data_time': 0.0007159179949667305, + 'model_time': 1.2235459449875634, + 'grad_norm_pre_clip_avg': 0.2540186017751694, + 'learning_rate': 2.353011169010648e-05, + 'epoch': 3.03} +04/19 [15:42:37] INFO | >> train_qwenlatent.py:487 + Step 12030 | grad_norm_pre_clip=0.2562 | + grad_norm_pre_clip_avg=0.2266 | Metrics: + {'align_loss': 0.024283234030008316, + 'recon_loss': 0.03970380872488022, + 'predict_loss': 0.011681333184242249, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2562004029750824, + 'data_time': 0.0005763999943155795, + 'model_time': 1.2353973280114587, + 'grad_norm_pre_clip_avg': 0.22659733444452285, + 'learning_rate': 2.3526005446472295e-05, + 'epoch': 3.04} +04/19 [15:42:50] INFO | >> train_qwenlatent.py:487 + Step 12040 | grad_norm_pre_clip=0.1922 | + grad_norm_pre_clip_avg=0.2280 | Metrics: + {'align_loss': 0.024990584701299667, + 'recon_loss': 0.05410240963101387, + 'predict_loss': 0.01663290336728096, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19219660758972168, + 'data_time': 0.0008176460105460137, + 'model_time': 1.2590018379851244, + 'grad_norm_pre_clip_avg': 0.22804691642522812, + 'learning_rate': 2.352189383498935e-05, + 'epoch': 3.04} +04/19 [15:43:03] INFO | >> train_qwenlatent.py:487 + Step 12050 | grad_norm_pre_clip=0.2490 | + grad_norm_pre_clip_avg=0.2512 | Metrics: + {'align_loss': 0.023851970210671425, + 'recon_loss': 0.04812261462211609, + 'predict_loss': 0.019253533333539963, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2490333616733551, + 'mae_score': 0.022909075075441654, 'data_time': + 0.0009577150049153715, 'model_time': + 1.236345226992853, 'grad_norm_pre_clip_avg': + 0.2512033984065056, 'learning_rate': + 2.351777685766159e-05, 'epoch': 3.04} +04/19 [15:43:16] INFO | >> train_qwenlatent.py:487 + Step 12060 | grad_norm_pre_clip=0.1997 | + grad_norm_pre_clip_avg=0.2062 | Metrics: + {'align_loss': 0.025455623865127563, + 'recon_loss': 0.040818795561790466, + 'predict_loss': 0.011009972542524338, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1996936947107315, + 'data_time': 0.0009343959973193705, + 'model_time': 1.2731480269867461, + 'grad_norm_pre_clip_avg': 0.20617376267910004, + 'learning_rate': 2.3513654516495588e-05, + 'epoch': 3.04} +04/19 [15:43:28] INFO | >> train_qwenlatent.py:487 + Step 12070 | grad_norm_pre_clip=0.5347 | + grad_norm_pre_clip_avg=0.2552 | Metrics: + {'align_loss': 0.02532103657722473, + 'recon_loss': 0.03620807081460953, + 'predict_loss': 0.01429821364581585, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.5347417593002319, + 'data_time': 0.0008704960055183619, + 'model_time': 1.361140182998497, + 'grad_norm_pre_clip_avg': 0.2551594987511635, + 'learning_rate': 2.3509526813500516e-05, + 'epoch': 3.05} +04/19 [15:43:41] INFO | >> train_qwenlatent.py:487 + Step 12080 | grad_norm_pre_clip=0.2913 | + grad_norm_pre_clip_avg=0.3706 | Metrics: + {'align_loss': 0.024982023984193802, + 'recon_loss': 0.05157694220542908, + 'predict_loss': 0.01538463868200779, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2913452684879303, + 'data_time': 0.0009059560252353549, + 'model_time': 1.3244912840018515, + 'grad_norm_pre_clip_avg': 0.37061258852481843, + 'learning_rate': 2.350539375068817e-05, + 'epoch': 3.05} +04/19 [15:43:54] INFO | >> train_qwenlatent.py:487 + Step 12090 | grad_norm_pre_clip=0.3065 | + grad_norm_pre_clip_avg=0.2621 | Metrics: + {'align_loss': 0.023538775742053986, + 'recon_loss': 0.048299308866262436, + 'predict_loss': 0.016369935125112534, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3065320551395416, + 'data_time': 0.0007869930122978985, + 'model_time': 1.233051751012681, + 'grad_norm_pre_clip_avg': 0.2620583072304726, + 'learning_rate': 2.350125533007295e-05, + 'epoch': 3.05} +04/19 [15:44:07] INFO | >> train_qwenlatent.py:487 + Step 12100 | grad_norm_pre_clip=0.2393 | + grad_norm_pre_clip_avg=0.2234 | Metrics: + {'align_loss': 0.02605058252811432, + 'recon_loss': 0.05245417729020119, + 'predict_loss': 0.014586004428565502, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2392713874578476, + 'mae_score': 0.013809376793938713, 'data_time': + 0.0009019889985211194, 'model_time': + 1.2539466230082326, 'grad_norm_pre_clip_avg': + 0.22337530702352523, 'learning_rate': + 2.3497111553671874e-05, 'epoch': 3.05} +04/19 [15:44:20] INFO | >> train_qwenlatent.py:487 + Step 12110 | grad_norm_pre_clip=0.1880 | + grad_norm_pre_clip_avg=0.2212 | Metrics: + {'align_loss': 0.02309587597846985, + 'recon_loss': 0.032238610088825226, + 'predict_loss': 0.009185164235532284, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18799042701721191, + 'data_time': 0.00092838000273332, 'model_time': + 1.2281871099839918, 'grad_norm_pre_clip_avg': + 0.22117083817720412, 'learning_rate': + 2.349296242350457e-05, 'epoch': 3.06} +04/19 [15:44:33] INFO | >> train_qwenlatent.py:487 + Step 12120 | grad_norm_pre_clip=0.2191 | + grad_norm_pre_clip_avg=0.2237 | Metrics: + {'align_loss': 0.025285033509135246, + 'recon_loss': 0.058610446751117706, + 'predict_loss': 0.01896190457046032, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21905648708343506, + 'data_time': 0.000944822997553274, + 'model_time': 1.6374547530140262, + 'grad_norm_pre_clip_avg': 0.22373308390378951, + 'learning_rate': 2.348880794159328e-05, + 'epoch': 3.06} +04/19 [15:44:45] INFO | >> train_qwenlatent.py:487 + Step 12130 | grad_norm_pre_clip=0.1805 | + grad_norm_pre_clip_avg=0.2063 | Metrics: + {'align_loss': 0.02334229089319706, + 'recon_loss': 0.032823946326971054, + 'predict_loss': 0.009991190396249294, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1804954707622528, + 'data_time': 0.0006987199885770679, + 'model_time': 1.21473281699582, + 'grad_norm_pre_clip_avg': 0.20632015764713288, + 'learning_rate': 2.3484648109962834e-05, + 'epoch': 3.06} +04/19 [15:44:58] INFO | >> train_qwenlatent.py:487 + Step 12140 | grad_norm_pre_clip=0.1823 | + grad_norm_pre_clip_avg=0.2574 | Metrics: + {'align_loss': 0.02581077441573143, + 'recon_loss': 0.049604903906583786, + 'predict_loss': 0.015848515555262566, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18232670426368713, + 'data_time': 0.0009048270003404468, + 'model_time': 1.2530836960067973, + 'grad_norm_pre_clip_avg': 0.2573712974786758, + 'learning_rate': 2.348048293064069e-05, + 'epoch': 3.06} +04/19 [15:45:11] INFO | >> train_qwenlatent.py:487 + Step 12150 | grad_norm_pre_clip=0.2385 | + grad_norm_pre_clip_avg=0.2220 | Metrics: + {'align_loss': 0.02451469749212265, + 'recon_loss': 0.03951316326856613, + 'predict_loss': 0.01291052345186472, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23852616548538208, + 'mae_score': 0.013408211544827298, 'data_time': + 0.001177457015728578, 'model_time': + 1.524003313999856, 'grad_norm_pre_clip_avg': + 0.2220277637243271, 'learning_rate': + 2.3476312405656906e-05, 'epoch': 3.07} +04/19 [15:45:24] INFO | >> train_qwenlatent.py:487 + Step 12160 | grad_norm_pre_clip=0.2941 | + grad_norm_pre_clip_avg=0.2458 | Metrics: + {'align_loss': 0.02446027472615242, + 'recon_loss': 0.050271399319171906, + 'predict_loss': 0.013663744553923607, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.29406315088272095, + 'data_time': 0.0007059609924908727, + 'model_time': 1.2701195089903194, + 'grad_norm_pre_clip_avg': 0.24583912938833236, + 'learning_rate': 2.347213653704415e-05, + 'epoch': 3.07} +04/19 [15:45:37] INFO | >> train_qwenlatent.py:487 + Step 12170 | grad_norm_pre_clip=0.1730 | + grad_norm_pre_clip_avg=0.2050 | Metrics: + {'align_loss': 0.025165490806102753, + 'recon_loss': 0.04170733690261841, + 'predict_loss': 0.010935978032648563, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17300963401794434, + 'data_time': 0.0006185910024214536, + 'model_time': 1.2504489239945542, + 'grad_norm_pre_clip_avg': 0.20501996725797653, + 'learning_rate': 2.3467955326837685e-05, + 'epoch': 3.07} +04/19 [15:45:50] INFO | >> train_qwenlatent.py:487 + Step 12180 | grad_norm_pre_clip=0.5293 | + grad_norm_pre_clip_avg=0.2485 | Metrics: + {'align_loss': 0.024135859683156013, + 'recon_loss': 0.03408849239349365, + 'predict_loss': 0.010148292407393456, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.5293276906013489, + 'data_time': 0.0010280910064466298, + 'model_time': 1.3803534799953923, + 'grad_norm_pre_clip_avg': 0.24849793761968614, + 'learning_rate': 2.3463768777075378e-05, + 'epoch': 3.07} +04/19 [15:46:02] INFO | >> train_qwenlatent.py:487 + Step 12190 | grad_norm_pre_clip=0.2116 | + grad_norm_pre_clip_avg=0.3080 | Metrics: + {'align_loss': 0.02526419796049595, + 'recon_loss': 0.04951145872473717, + 'predict_loss': 0.015143197029829025, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21159759163856506, + 'data_time': 0.0009601700003258884, + 'model_time': 1.2317726130131632, + 'grad_norm_pre_clip_avg': 0.30803379863500596, + 'learning_rate': 2.345957688979771e-05, + 'epoch': 3.08} +04/19 [15:46:16] INFO | >> train_qwenlatent.py:487 + Step 12200 | grad_norm_pre_clip=0.3057 | + grad_norm_pre_clip_avg=0.2671 | Metrics: + {'align_loss': 0.025495383888483047, + 'recon_loss': 0.04927278310060501, + 'predict_loss': 0.015649018809199333, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3057108223438263, + 'mae_score': 0.016174065529763162, 'data_time': + 0.0008981949940789491, 'model_time': + 1.2555551719851792, 'grad_norm_pre_clip_avg': + 0.2671371757984161, 'learning_rate': + 2.3455379667047748e-05, 'epoch': 3.08} +04/19 [15:46:28] INFO | >> train_qwenlatent.py:487 + Step 12210 | grad_norm_pre_clip=0.1897 | + grad_norm_pre_clip_avg=0.2149 | Metrics: + {'align_loss': 0.024128343909978867, + 'recon_loss': 0.04761725291609764, + 'predict_loss': 0.015475902706384659, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18971028923988342, + 'data_time': 0.0008902739791665226, + 'model_time': 1.2601505469938274, + 'grad_norm_pre_clip_avg': 0.21490489840507507, + 'learning_rate': 2.3451177110871172e-05, + 'epoch': 3.08} +04/19 [15:46:41] INFO | >> train_qwenlatent.py:487 + Step 12220 | grad_norm_pre_clip=0.2118 | + grad_norm_pre_clip_avg=0.2146 | Metrics: + {'align_loss': 0.024628225713968277, + 'recon_loss': 0.049895480275154114, + 'predict_loss': 0.017920561134815216, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21178936958312988, + 'data_time': 0.0008719059987924993, + 'model_time': 1.2432575999991968, + 'grad_norm_pre_clip_avg': 0.21463971436023713, + 'learning_rate': 2.3446969223316262e-05, + 'epoch': 3.08} +04/19 [15:46:53] INFO | >> train_qwenlatent.py:487 + Step 12230 | grad_norm_pre_clip=0.2522 | + grad_norm_pre_clip_avg=0.2376 | Metrics: + {'align_loss': 0.024138368666172028, + 'recon_loss': 0.04486043006181717, + 'predict_loss': 0.011891099624335766, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25223347544670105, + 'data_time': 0.0007504230015911162, + 'model_time': 1.2624024550023023, + 'grad_norm_pre_clip_avg': 0.23757631629705428, + 'learning_rate': 2.3442756006433884e-05, + 'epoch': 3.09} +04/19 [15:47:06] INFO | >> train_qwenlatent.py:487 + Step 12240 | grad_norm_pre_clip=0.3219 | + grad_norm_pre_clip_avg=0.2561 | Metrics: + {'align_loss': 0.025024253875017166, + 'recon_loss': 0.04557478427886963, + 'predict_loss': 0.015506871975958347, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3218524754047394, + 'data_time': 0.0008529720071237534, + 'model_time': 1.2016057769942563, + 'grad_norm_pre_clip_avg': 0.2561331853270531, + 'learning_rate': 2.343853746227751e-05, + 'epoch': 3.09} +04/19 [15:47:20] INFO | >> train_qwenlatent.py:487 + Step 12250 | grad_norm_pre_clip=0.2066 | + grad_norm_pre_clip_avg=0.2422 | Metrics: + {'align_loss': 0.02425975166261196, + 'recon_loss': 0.04526963829994202, + 'predict_loss': 0.013968955725431442, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20664997398853302, + 'mae_score': 0.017547164092192778, 'data_time': + 0.0009592080023139715, 'model_time': + 1.2559653820062522, 'grad_norm_pre_clip_avg': + 0.24217090904712676, 'learning_rate': + 2.3434313592903208e-05, 'epoch': 3.09} +04/19 [15:47:33] INFO | >> train_qwenlatent.py:487 + Step 12260 | grad_norm_pre_clip=0.2235 | + grad_norm_pre_clip_avg=0.2162 | Metrics: + {'align_loss': 0.025023721158504486, + 'recon_loss': 0.05718144029378891, + 'predict_loss': 0.01614026352763176, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22353172302246094, + 'data_time': 0.0009723980037961155, + 'model_time': 1.2563891199824866, + 'grad_norm_pre_clip_avg': 0.21621961295604705, + 'learning_rate': 2.343008440036964e-05, + 'epoch': 3.09} +04/19 [15:47:45] INFO | >> train_qwenlatent.py:487 + Step 12270 | grad_norm_pre_clip=0.2094 | + grad_norm_pre_clip_avg=0.2980 | Metrics: + {'align_loss': 0.025431349873542786, + 'recon_loss': 0.048546914011240005, + 'predict_loss': 0.015379860065877438, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20944955945014954, + 'data_time': 0.0006387220055330545, + 'model_time': 1.207995058997767, + 'grad_norm_pre_clip_avg': 0.2979575917124748, + 'learning_rate': 2.3425849886738063e-05, + 'epoch': 3.1} +04/19 [15:47:58] INFO | >> train_qwenlatent.py:487 + Step 12280 | grad_norm_pre_clip=0.1891 | + grad_norm_pre_clip_avg=0.2251 | Metrics: + {'align_loss': 0.025396155193448067, + 'recon_loss': 0.048672668635845184, + 'predict_loss': 0.011934125795960426, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18910308182239532, + 'data_time': 0.0007646409794688225, + 'model_time': 1.1985164789948612, + 'grad_norm_pre_clip_avg': 0.225071282684803, + 'learning_rate': 2.342161005407233e-05, + 'epoch': 3.1} +04/19 [15:48:11] INFO | >> train_qwenlatent.py:487 + Step 12290 | grad_norm_pre_clip=0.2586 | + grad_norm_pre_clip_avg=0.2111 | Metrics: + {'align_loss': 0.025866899639368057, + 'recon_loss': 0.06451663374900818, + 'predict_loss': 0.014795368537306786, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25861984491348267, + 'data_time': 0.0009888710046652704, + 'model_time': 1.224573915998917, + 'grad_norm_pre_clip_avg': 0.21107009947299957, + 'learning_rate': 2.341736490443888e-05, + 'epoch': 3.1} +04/19 [15:48:24] INFO | >> train_qwenlatent.py:487 + Step 12300 | grad_norm_pre_clip=0.2640 | + grad_norm_pre_clip_avg=0.3032 | Metrics: + {'align_loss': 0.02322465367615223, + 'recon_loss': 0.04856935143470764, + 'predict_loss': 0.018277039751410484, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26397791504859924, + 'mae_score': 0.015800209732742997, 'data_time': + 0.0008901120163500309, 'model_time': + 1.2647656089975499, 'grad_norm_pre_clip_avg': + 0.30316643267869947, 'learning_rate': + 2.341311443990675e-05, 'epoch': 3.1} +04/19 [15:48:37] INFO | >> train_qwenlatent.py:487 + Step 12310 | grad_norm_pre_clip=0.2417 | + grad_norm_pre_clip_avg=0.3121 | Metrics: + {'align_loss': 0.025050057098269463, + 'recon_loss': 0.05320819094777107, + 'predict_loss': 0.011870184913277626, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24171094596385956, + 'data_time': 0.0005739209882449359, + 'model_time': 1.2289952939900104, + 'grad_norm_pre_clip_avg': 0.3121222838759422, + 'learning_rate': 2.3408858662547563e-05, + 'epoch': 3.11} +04/19 [15:48:50] INFO | >> train_qwenlatent.py:487 + Step 12320 | grad_norm_pre_clip=0.2464 | + grad_norm_pre_clip_avg=0.2596 | Metrics: + {'align_loss': 0.023170538246631622, + 'recon_loss': 0.03847187012434006, + 'predict_loss': 0.016181176528334618, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24638621509075165, + 'data_time': 0.0009818479884415865, + 'model_time': 1.2518890539940912, + 'grad_norm_pre_clip_avg': 0.2595830410718918, + 'learning_rate': 2.3404597574435537e-05, + 'epoch': 3.11} +04/19 [15:49:02] INFO | >> train_qwenlatent.py:487 + Step 12330 | grad_norm_pre_clip=0.1706 | + grad_norm_pre_clip_avg=0.2023 | Metrics: + {'align_loss': 0.025015126913785934, + 'recon_loss': 0.04650483652949333, + 'predict_loss': 0.016915645450353622, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1706463247537613, + 'data_time': 0.0008197909919545054, + 'model_time': 1.249534399015829, + 'grad_norm_pre_clip_avg': 0.20225154459476472, + 'learning_rate': 2.340033117764747e-05, + 'epoch': 3.11} +04/19 [15:49:15] INFO | >> train_qwenlatent.py:487 + Step 12340 | grad_norm_pre_clip=0.2559 | + grad_norm_pre_clip_avg=0.2077 | Metrics: + {'align_loss': 0.026128385215997696, + 'recon_loss': 0.05038454756140709, + 'predict_loss': 0.014794934540987015, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2559109032154083, + 'data_time': 0.0007061409996822476, + 'model_time': 1.2316472139791586, + 'grad_norm_pre_clip_avg': 0.20769888609647752, + 'learning_rate': 2.3396059474262755e-05, + 'epoch': 3.11} +04/19 [15:49:28] INFO | >> train_qwenlatent.py:487 + Step 12350 | grad_norm_pre_clip=0.2326 | + grad_norm_pre_clip_avg=0.2003 | Metrics: + {'align_loss': 0.02588316798210144, + 'recon_loss': 0.055717699229717255, + 'predict_loss': 0.01673532836139202, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23263944685459137, + 'mae_score': 0.016600615699012, 'data_time': + 0.0007577149954158813, 'model_time': + 1.2415490340208635, 'grad_norm_pre_clip_avg': + 0.20030955970287323, 'learning_rate': + 2.3391782466363367e-05, 'epoch': 3.12} +04/19 [15:49:40] INFO | >> train_qwenlatent.py:487 + Step 12360 | grad_norm_pre_clip=0.1972 | + grad_norm_pre_clip_avg=0.2962 | Metrics: + {'align_loss': 0.0246693454682827, + 'recon_loss': 0.038973741233348846, + 'predict_loss': 0.011494345963001251, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19717152416706085, + 'data_time': 0.0006530010141432285, + 'model_time': 1.2593680580030195, + 'grad_norm_pre_clip_avg': 0.2962282493710518, + 'learning_rate': 2.3387500156033866e-05, + 'epoch': 3.12} +04/19 [15:49:53] INFO | >> train_qwenlatent.py:487 + Step 12370 | grad_norm_pre_clip=0.2204 | + grad_norm_pre_clip_avg=0.2649 | Metrics: + {'align_loss': 0.024692542850971222, + 'recon_loss': 0.06029064580798149, + 'predict_loss': 0.02223639376461506, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2203526794910431, + 'data_time': 0.0008077500096987933, + 'model_time': 1.2127976040064823, + 'grad_norm_pre_clip_avg': 0.26491540372371675, + 'learning_rate': 2.3383212545361404e-05, + 'epoch': 3.12} +04/19 [15:50:05] INFO | >> train_qwenlatent.py:487 + Step 12380 | grad_norm_pre_clip=0.1887 | + grad_norm_pre_clip_avg=0.2149 | Metrics: + {'align_loss': 0.024905532598495483, + 'recon_loss': 0.04808780550956726, + 'predict_loss': 0.01477402076125145, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18868638575077057, + 'data_time': 0.0006129889807198197, + 'model_time': 1.2065506810031366, + 'grad_norm_pre_clip_avg': 0.21490767896175383, + 'learning_rate': 2.3378919636435704e-05, + 'epoch': 3.12} +04/19 [15:50:18] INFO | >> train_qwenlatent.py:487 + Step 12390 | grad_norm_pre_clip=0.2186 | + grad_norm_pre_clip_avg=0.2082 | Metrics: + {'align_loss': 0.023429641500115395, + 'recon_loss': 0.04617994278669357, + 'predict_loss': 0.013567727990448475, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21861636638641357, + 'data_time': 0.0006283620023168623, + 'model_time': 1.2308075810142327, + 'grad_norm_pre_clip_avg': 0.20820724666118623, + 'learning_rate': 2.3374621431349078e-05, + 'epoch': 3.13} +04/19 [15:50:31] INFO | >> train_qwenlatent.py:487 + Step 12400 | grad_norm_pre_clip=0.4313 | + grad_norm_pre_clip_avg=0.2412 | Metrics: + {'align_loss': 0.02612200751900673, + 'recon_loss': 0.0514371283352375, + 'predict_loss': 0.014507405459880829, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.43133553862571716, + 'mae_score': 0.015055881534610783, 'data_time': + 0.000849948002723977, 'model_time': + 1.2397668310150038, 'grad_norm_pre_clip_avg': + 0.24120910167694093, 'learning_rate': + 2.3370317932196423e-05, 'epoch': 3.13} +04/19 [15:50:44] INFO | >> train_qwenlatent.py:487 + Step 12410 | grad_norm_pre_clip=0.2460 | + grad_norm_pre_clip_avg=0.2976 | Metrics: + {'align_loss': 0.02559993602335453, + 'recon_loss': 0.05671149864792824, + 'predict_loss': 0.014367479830980301, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2459561675786972, + 'data_time': 0.0008773359877523035, + 'model_time': 1.276432795013534, + 'grad_norm_pre_clip_avg': 0.29762520343065263, + 'learning_rate': 2.336600914107521e-05, + 'epoch': 3.13} +04/19 [15:50:57] INFO | >> train_qwenlatent.py:487 + Step 12420 | grad_norm_pre_clip=0.1877 | + grad_norm_pre_clip_avg=0.2024 | Metrics: + {'align_loss': 0.02531176432967186, + 'recon_loss': 0.04853548854589462, + 'predict_loss': 0.013376082293689251, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18769289553165436, + 'data_time': 0.0006145669904071838, + 'model_time': 1.311046991002513, + 'grad_norm_pre_clip_avg': 0.20241289734840393, + 'learning_rate': 2.3361695060085488e-05, + 'epoch': 3.13} +04/19 [15:51:10] INFO | >> train_qwenlatent.py:487 + Step 12430 | grad_norm_pre_clip=0.1808 | + grad_norm_pre_clip_avg=0.2041 | Metrics: + {'align_loss': 0.02419823408126831, + 'recon_loss': 0.041426219046115875, + 'predict_loss': 0.014856684021651745, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1808149367570877, + 'data_time': 0.0009274670155718923, + 'model_time': 1.5325366389879491, + 'grad_norm_pre_clip_avg': 0.20412092208862304, + 'learning_rate': 2.33573756913299e-05, 'epoch': + 3.14} +04/19 [15:51:23] INFO | >> train_qwenlatent.py:487 + Step 12440 | grad_norm_pre_clip=0.2469 | + grad_norm_pre_clip_avg=0.2266 | Metrics: + {'align_loss': 0.02424892969429493, + 'recon_loss': 0.05105157941579819, + 'predict_loss': 0.016346430405974388, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24688053131103516, + 'data_time': 0.0009050510125234723, + 'model_time': 1.262649396987399, + 'grad_norm_pre_clip_avg': 0.22662594467401503, + 'learning_rate': 2.335305103691364e-05, + 'epoch': 3.14} +04/19 [15:51:37] INFO | >> train_qwenlatent.py:487 + Step 12450 | grad_norm_pre_clip=0.3147 | + grad_norm_pre_clip_avg=0.3163 | Metrics: + {'align_loss': 0.0253940187394619, + 'recon_loss': 0.044211599975824356, + 'predict_loss': 0.016868581995368004, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.31467679142951965, + 'mae_score': 0.02022244221455342, 'data_time': + 0.0008284840150736272, 'model_time': + 1.2806984049966559, 'grad_norm_pre_clip_avg': + 0.3163379728794098, 'learning_rate': + 2.3348721098944497e-05, 'epoch': 3.14} +04/19 [15:51:49] INFO | >> train_qwenlatent.py:487 + Step 12460 | grad_norm_pre_clip=0.2218 | + grad_norm_pre_clip_avg=0.2315 | Metrics: + {'align_loss': 0.024925734847784042, + 'recon_loss': 0.05539510399103165, + 'predict_loss': 0.015011215582489967, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22179612517356873, + 'data_time': 0.0006732920010108501, + 'model_time': 1.244319439982064, + 'grad_norm_pre_clip_avg': 0.23145973533391953, + 'learning_rate': 2.3344385879532828e-05, + 'epoch': 3.14} +04/19 [15:52:02] INFO | >> train_qwenlatent.py:487 + Step 12470 | grad_norm_pre_clip=0.1874 | + grad_norm_pre_clip_avg=0.2101 | Metrics: + {'align_loss': 0.024605628103017807, + 'recon_loss': 0.03619861230254173, + 'predict_loss': 0.010361239314079285, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18737073242664337, + 'data_time': 0.0006954210111871362, + 'model_time': 1.2429697559855413, + 'grad_norm_pre_clip_avg': 0.2101471543312073, + 'learning_rate': 2.334004538079157e-05, + 'epoch': 3.15} +04/19 [15:52:14] INFO | >> train_qwenlatent.py:487 + Step 12480 | grad_norm_pre_clip=0.3383 | + grad_norm_pre_clip_avg=0.2692 | Metrics: + {'align_loss': 0.025917038321495056, + 'recon_loss': 0.055382270365953445, + 'predict_loss': 0.01759849116206169, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.33828139305114746, + 'data_time': 0.0006466390041168779, + 'model_time': 1.2437764659989625, + 'grad_norm_pre_clip_avg': 0.26921945810317993, + 'learning_rate': 2.333569960483623e-05, + 'epoch': 3.15} +04/19 [15:52:27] INFO | >> train_qwenlatent.py:487 + Step 12490 | grad_norm_pre_clip=0.2827 | + grad_norm_pre_clip_avg=0.2555 | Metrics: + {'align_loss': 0.024962153285741806, + 'recon_loss': 0.040174636989831924, + 'predict_loss': 0.012965350411832333, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28267472982406616, + 'data_time': 0.0009624120139051229, + 'model_time': 1.2681291830085684, + 'grad_norm_pre_clip_avg': 0.25545926839113237, + 'learning_rate': 2.3331348553784882e-05, + 'epoch': 3.15} +04/19 [15:52:40] INFO | >> train_qwenlatent.py:487 + Step 12500 | grad_norm_pre_clip=0.2483 | + grad_norm_pre_clip_avg=0.2298 | Metrics: + {'align_loss': 0.023373380303382874, + 'recon_loss': 0.04442945122718811, + 'predict_loss': 0.014784282073378563, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24834884703159332, + 'mae_score': 0.021033321414981876, 'data_time': + 0.0006349570176098496, 'model_time': + 1.2532945010170806, 'grad_norm_pre_clip_avg': + 0.22983087301254274, 'learning_rate': + 2.3326992229758182e-05, 'epoch': 3.15} +04/19 [15:52:53] INFO | >> train_qwenlatent.py:487 + Step 12510 | grad_norm_pre_clip=0.2076 | + grad_norm_pre_clip_avg=0.1998 | Metrics: + {'align_loss': 0.024570247158408165, + 'recon_loss': 0.051186591386795044, + 'predict_loss': 0.017400672659277916, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20759858191013336, + 'data_time': 0.0007130349986255169, + 'model_time': 1.266195053991396, + 'grad_norm_pre_clip_avg': 0.19977399706840515, + 'learning_rate': 2.332263063487934e-05, + 'epoch': 3.16} +04/19 [15:53:05] INFO | >> train_qwenlatent.py:487 + Step 12520 | grad_norm_pre_clip=0.1950 | + grad_norm_pre_clip_avg=0.2072 | Metrics: + {'align_loss': 0.025474965572357178, + 'recon_loss': 0.046020377427339554, + 'predict_loss': 0.012808817438781261, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19498978555202484, + 'data_time': 0.0007558270008303225, + 'model_time': 1.2240592809976079, + 'grad_norm_pre_clip_avg': 0.2071979746222496, + 'learning_rate': 2.331826377127415e-05, + 'epoch': 3.16} +04/19 [15:53:18] INFO | >> train_qwenlatent.py:487 + Step 12530 | grad_norm_pre_clip=0.4263 | + grad_norm_pre_clip_avg=0.3140 | Metrics: + {'align_loss': 0.024338191375136375, + 'recon_loss': 0.05416771024465561, + 'predict_loss': 0.024028798565268517, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.4262652099132538, + 'data_time': 0.0006954639975447208, + 'model_time': 1.5697523760027252, + 'grad_norm_pre_clip_avg': 0.31398576200008393, + 'learning_rate': 2.331389164107097e-05, + 'epoch': 3.16} +04/19 [15:53:31] INFO | >> train_qwenlatent.py:487 + Step 12540 | grad_norm_pre_clip=0.2285 | + grad_norm_pre_clip_avg=0.3452 | Metrics: + {'align_loss': 0.026272358372807503, + 'recon_loss': 0.04307219758629799, + 'predict_loss': 0.006797348614782095, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22850418090820312, + 'data_time': 0.0006222729862201959, + 'model_time': 1.263200704997871, + 'grad_norm_pre_clip_avg': 0.3452253356575966, + 'learning_rate': 2.3309514246400718e-05, + 'epoch': 3.16} +04/19 [15:53:44] INFO | >> train_qwenlatent.py:487 + Step 12550 | grad_norm_pre_clip=0.2710 | + grad_norm_pre_clip_avg=0.2632 | Metrics: + {'align_loss': 0.026139896363019943, + 'recon_loss': 0.06118597090244293, + 'predict_loss': 0.020027726888656616, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2709762752056122, + 'mae_score': 0.027062218468468468, 'data_time': + 0.0006554680003318936, 'model_time': + 1.2675784009916242, 'grad_norm_pre_clip_avg': + 0.2632225289940834, 'learning_rate': + 2.3305131589396888e-05, 'epoch': 3.17} +04/19 [15:53:57] INFO | >> train_qwenlatent.py:487 + Step 12560 | grad_norm_pre_clip=0.1974 | + grad_norm_pre_clip_avg=0.2224 | Metrics: + {'align_loss': 0.02494734898209572, + 'recon_loss': 0.050041839480400085, + 'predict_loss': 0.01252005621790886, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19741037487983704, + 'data_time': 0.0006451620138250291, + 'model_time': 1.2732963630114682, + 'grad_norm_pre_clip_avg': 0.2223546698689461, + 'learning_rate': 2.330074367219553e-05, + 'epoch': 3.17} +04/19 [15:54:10] INFO | >> train_qwenlatent.py:487 + Step 12570 | grad_norm_pre_clip=0.1882 | + grad_norm_pre_clip_avg=0.2151 | Metrics: + {'align_loss': 0.024865826591849327, + 'recon_loss': 0.042691927403211594, + 'predict_loss': 0.01586875505745411, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18818417191505432, + 'data_time': 0.0006084570195525885, + 'model_time': 1.2636514039768372, + 'grad_norm_pre_clip_avg': 0.21506526172161103, + 'learning_rate': 2.3296350496935266e-05, + 'epoch': 3.17} +04/19 [15:54:23] INFO | >> train_qwenlatent.py:487 + Step 12580 | grad_norm_pre_clip=0.1802 | + grad_norm_pre_clip_avg=0.2165 | Metrics: + {'align_loss': 0.025284767150878906, + 'recon_loss': 0.04523882269859314, + 'predict_loss': 0.012501943856477737, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1802026331424713, + 'data_time': 0.0010261999850627035, + 'model_time': 1.2591594219848048, + 'grad_norm_pre_clip_avg': 0.21653393357992173, + 'learning_rate': 2.3291952065757274e-05, + 'epoch': 3.17} +04/19 [15:54:35] INFO | >> train_qwenlatent.py:487 + Step 12590 | grad_norm_pre_clip=0.2054 | + grad_norm_pre_clip_avg=0.2148 | Metrics: + {'align_loss': 0.0236288420855999, + 'recon_loss': 0.0373845100402832, + 'predict_loss': 0.014788416214287281, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2054349035024643, + 'data_time': 0.0008813859894871712, + 'model_time': 1.232659913977841, + 'grad_norm_pre_clip_avg': 0.21480270922183992, + 'learning_rate': 2.3287548380805293e-05, + 'epoch': 3.18} +04/19 [15:54:48] INFO | >> train_qwenlatent.py:487 + Step 12600 | grad_norm_pre_clip=0.1906 | + grad_norm_pre_clip_avg=0.2730 | Metrics: + {'align_loss': 0.02457902953028679, + 'recon_loss': 0.05427538976073265, + 'predict_loss': 0.01794624514877796, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19063584506511688, + 'mae_score': 0.022729519680813628, 'data_time': + 0.0007081240182742476, 'model_time': + 1.2006400619866326, 'grad_norm_pre_clip_avg': + 0.2729868084192276, 'learning_rate': + 2.328313944422563e-05, 'epoch': 3.18} +04/19 [15:55:01] INFO | >> train_qwenlatent.py:487 + Step 12610 | grad_norm_pre_clip=0.2621 | + grad_norm_pre_clip_avg=0.2340 | Metrics: + {'align_loss': 0.02405416965484619, + 'recon_loss': 0.04290375858545303, + 'predict_loss': 0.014774370938539505, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2620921730995178, + 'data_time': 0.00070060501457192, 'model_time': + 1.2945169580052607, 'grad_norm_pre_clip_avg': + 0.2340316578745842, 'learning_rate': + 2.3278725258167145e-05, 'epoch': 3.18} +04/19 [15:55:13] INFO | >> train_qwenlatent.py:487 + Step 12620 | grad_norm_pre_clip=0.2497 | + grad_norm_pre_clip_avg=0.2168 | Metrics: + {'align_loss': 0.02480141818523407, + 'recon_loss': 0.04442738741636276, + 'predict_loss': 0.016106000170111656, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24971583485603333, + 'data_time': 0.0007140680099837482, + 'model_time': 1.2506328120070975, + 'grad_norm_pre_clip_avg': 0.2168228194117546, + 'learning_rate': 2.3274305824781255e-05, + 'epoch': 3.18} +04/19 [15:55:26] INFO | >> train_qwenlatent.py:487 + Step 12630 | grad_norm_pre_clip=0.3892 | + grad_norm_pre_clip_avg=0.2789 | Metrics: + {'align_loss': 0.02351386472582817, + 'recon_loss': 0.039710454642772675, + 'predict_loss': 0.012871979735791683, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.38919690251350403, + 'data_time': 0.0008820910006761551, + 'model_time': 1.2488631269952748, + 'grad_norm_pre_clip_avg': 0.27894579619169235, + 'learning_rate': 2.3269881146221946e-05, + 'epoch': 3.19} +04/19 [15:55:38] INFO | >> train_qwenlatent.py:487 + Step 12640 | grad_norm_pre_clip=0.1836 | + grad_norm_pre_clip_avg=0.2325 | Metrics: + {'align_loss': 0.024831753224134445, + 'recon_loss': 0.03866889327764511, + 'predict_loss': 0.010344583541154861, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18355438113212585, + 'data_time': 0.0011986730096396059, + 'model_time': 1.254017086001113, + 'grad_norm_pre_clip_avg': 0.23252546042203903, + 'learning_rate': 2.3265451224645744e-05, + 'epoch': 3.19} +04/19 [15:55:52] INFO | >> train_qwenlatent.py:487 + Step 12650 | grad_norm_pre_clip=0.3340 | + grad_norm_pre_clip_avg=0.2499 | Metrics: + {'align_loss': 0.02549384906888008, + 'recon_loss': 0.03670497611165047, + 'predict_loss': 0.011273173615336418, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.33398061990737915, + 'mae_score': 0.014304265030869493, 'data_time': + 0.0007665849989280105, 'model_time': + 1.2460169419937301, 'grad_norm_pre_clip_avg': + 0.24992866516113282, 'learning_rate': + 2.3261016062211748e-05, 'epoch': 3.19} +04/19 [15:56:04] INFO | >> train_qwenlatent.py:487 + Step 12660 | grad_norm_pre_clip=0.1524 | + grad_norm_pre_clip_avg=0.2158 | Metrics: + {'align_loss': 0.0253787599503994, + 'recon_loss': 0.04673739895224571, + 'predict_loss': 0.010127753019332886, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15236185491085052, + 'data_time': 0.0009541549952700734, + 'model_time': 1.2208916330127977, + 'grad_norm_pre_clip_avg': 0.21583495438098907, + 'learning_rate': 2.3256575661081594e-05, + 'epoch': 3.19} +04/19 [15:56:17] INFO | >> train_qwenlatent.py:487 + Step 12670 | grad_norm_pre_clip=0.2838 | + grad_norm_pre_clip_avg=0.2300 | Metrics: + {'align_loss': 0.024392925202846527, + 'recon_loss': 0.04981553182005882, + 'predict_loss': 0.01303696446120739, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28384047746658325, + 'data_time': 0.0007153729966375977, + 'model_time': 1.2434997709933668, + 'grad_norm_pre_clip_avg': 0.23004773259162903, + 'learning_rate': 2.3252130023419488e-05, + 'epoch': 3.2} +04/19 [15:56:30] INFO | >> train_qwenlatent.py:487 + Step 12680 | grad_norm_pre_clip=0.2498 | + grad_norm_pre_clip_avg=0.2374 | Metrics: + {'align_loss': 0.024913398548960686, + 'recon_loss': 0.04074123129248619, + 'predict_loss': 0.015173329971730709, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24980716407299042, + 'data_time': 0.0006358259997796267, + 'model_time': 1.2204858050099574, + 'grad_norm_pre_clip_avg': 0.23736941814422607, + 'learning_rate': 2.324767915139217e-05, + 'epoch': 3.2} +04/19 [15:56:42] INFO | >> train_qwenlatent.py:487 + Step 12690 | grad_norm_pre_clip=0.2071 | + grad_norm_pre_clip_avg=0.2299 | Metrics: + {'align_loss': 0.02544630691409111, + 'recon_loss': 0.04626242443919182, + 'predict_loss': 0.010069001466035843, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2070520520210266, + 'data_time': 0.0008932999917306006, + 'model_time': 1.261743066017516, + 'grad_norm_pre_clip_avg': 0.22993019968271255, + 'learning_rate': 2.3243223047168948e-05, + 'epoch': 3.2} +04/19 [15:56:56] INFO | >> train_qwenlatent.py:487 + Step 12700 | grad_norm_pre_clip=0.4431 | + grad_norm_pre_clip_avg=0.2358 | Metrics: + {'align_loss': 0.024373922497034073, + 'recon_loss': 0.04333715885877609, + 'predict_loss': 0.011757006868720055, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.44314858317375183, + 'mae_score': 0.0194406457849451, 'data_time': + 0.0006158839969430119, 'model_time': + 1.3349303209979553, 'grad_norm_pre_clip_avg': + 0.23581438660621643, 'learning_rate': + 2.3238761712921674e-05, 'epoch': 3.2} +04/19 [15:57:09] INFO | >> train_qwenlatent.py:487 + Step 12710 | grad_norm_pre_clip=0.3015 | + grad_norm_pre_clip_avg=0.3073 | Metrics: + {'align_loss': 0.02446894720196724, + 'recon_loss': 0.05331714451313019, + 'predict_loss': 0.01477841567248106, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.30146950483322144, + 'data_time': 0.0006584729999303818, + 'model_time': 1.2385837570182048, + 'grad_norm_pre_clip_avg': 0.3073214814066887, + 'learning_rate': 2.323429515082474e-05, + 'epoch': 3.21} +04/19 [15:57:22] INFO | >> train_qwenlatent.py:487 + Step 12720 | grad_norm_pre_clip=0.2487 | + grad_norm_pre_clip_avg=0.2160 | Metrics: + {'align_loss': 0.02540966309607029, + 'recon_loss': 0.060495615005493164, + 'predict_loss': 0.01536433957517147, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24868154525756836, + 'data_time': 0.0009342789999209344, + 'model_time': 1.2183529839967377, + 'grad_norm_pre_clip_avg': 0.21603994071483612, + 'learning_rate': 2.3229823363055105e-05, + 'epoch': 3.21} +04/19 [15:57:34] INFO | >> train_qwenlatent.py:487 + Step 12730 | grad_norm_pre_clip=0.2473 | + grad_norm_pre_clip_avg=0.2361 | Metrics: + {'align_loss': 0.025303959846496582, + 'recon_loss': 0.05821651592850685, + 'predict_loss': 0.013927829451858997, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24725021421909332, + 'data_time': 0.0008780780190136284, + 'model_time': 1.258888153010048, + 'grad_norm_pre_clip_avg': 0.23610616177320481, + 'learning_rate': 2.3225346351792253e-05, + 'epoch': 3.21} +04/19 [15:57:47] INFO | >> train_qwenlatent.py:487 + Step 12740 | grad_norm_pre_clip=0.1881 | + grad_norm_pre_clip_avg=0.2005 | Metrics: + {'align_loss': 0.025935083627700806, + 'recon_loss': 0.058150697499513626, + 'predict_loss': 0.016287757083773613, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18807919323444366, + 'data_time': 0.0007270470086950809, + 'model_time': 1.2099203889956698, + 'grad_norm_pre_clip_avg': 0.2004935324192047, + 'learning_rate': 2.3220864119218232e-05, + 'epoch': 3.21} +04/19 [15:58:00] INFO | >> train_qwenlatent.py:487 + Step 12750 | grad_norm_pre_clip=0.2628 | + grad_norm_pre_clip_avg=0.2342 | Metrics: + {'align_loss': 0.023719914257526398, + 'recon_loss': 0.03869421407580376, + 'predict_loss': 0.014901097863912582, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26277852058410645, + 'mae_score': 0.017950119843354095, 'data_time': + 0.0006305429851636291, 'model_time': + 1.216597022023052, 'grad_norm_pre_clip_avg': + 0.23423733115196227, 'learning_rate': + 2.3216376667517627e-05, 'epoch': 3.22} +04/19 [15:58:13] INFO | >> train_qwenlatent.py:487 + Step 12760 | grad_norm_pre_clip=0.3092 | + grad_norm_pre_clip_avg=0.2624 | Metrics: + {'align_loss': 0.02460065484046936, + 'recon_loss': 0.03352002799510956, + 'predict_loss': 0.011731370352208614, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3091735243797302, + 'data_time': 0.0012159330071881413, + 'model_time': 1.288908789982088, + 'grad_norm_pre_clip_avg': 0.2624306410551071, + 'learning_rate': 2.3211883998877566e-05, + 'epoch': 3.22} +04/19 [15:58:26] INFO | >> train_qwenlatent.py:487 + Step 12770 | grad_norm_pre_clip=0.1684 | + grad_norm_pre_clip_avg=0.2184 | Metrics: + {'align_loss': 0.025062471628189087, + 'recon_loss': 0.04798870161175728, + 'predict_loss': 0.012433942407369614, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1683647483587265, + 'data_time': 0.000789637997513637, + 'model_time': 1.2341837940039113, + 'grad_norm_pre_clip_avg': 0.2184324488043785, + 'learning_rate': 2.3207386115487724e-05, + 'epoch': 3.22} +04/19 [15:58:38] INFO | >> train_qwenlatent.py:487 + Step 12780 | grad_norm_pre_clip=0.2299 | + grad_norm_pre_clip_avg=0.2106 | Metrics: + {'align_loss': 0.025981677696108818, + 'recon_loss': 0.056948427110910416, + 'predict_loss': 0.012628575786948204, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2298840582370758, + 'data_time': 0.0006217900081537664, + 'model_time': 1.2273962569888681, + 'grad_norm_pre_clip_avg': 0.2105664759874344, + 'learning_rate': 2.320288301954031e-05, + 'epoch': 3.22} +04/19 [15:58:51] INFO | >> train_qwenlatent.py:487 + Step 12790 | grad_norm_pre_clip=0.3358 | + grad_norm_pre_clip_avg=0.2702 | Metrics: + {'align_loss': 0.025542207062244415, + 'recon_loss': 0.05430622398853302, + 'predict_loss': 0.014315282925963402, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.33577781915664673, + 'data_time': 0.0007221659761853516, + 'model_time': 1.3000967249972746, + 'grad_norm_pre_clip_avg': 0.2701837569475174, + 'learning_rate': 2.319837471323008e-05, + 'epoch': 3.23} +04/19 [15:59:05] INFO | >> train_qwenlatent.py:487 + Step 12800 | grad_norm_pre_clip=0.2048 | + grad_norm_pre_clip_avg=0.2568 | Metrics: + {'align_loss': 0.023864157497882843, + 'recon_loss': 0.04220376908779144, + 'predict_loss': 0.015302775427699089, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20482465624809265, + 'mae_score': 0.017689213452038464, 'data_time': + 0.0006596159946639091, 'model_time': + 1.6704298410040792, 'grad_norm_pre_clip_avg': + 0.25678653717041017, 'learning_rate': + 2.319386119875433e-05, 'epoch': 3.23} +04/19 [15:59:17] INFO | >> train_qwenlatent.py:487 + Step 12810 | grad_norm_pre_clip=0.3024 | + grad_norm_pre_clip_avg=0.2580 | Metrics: + {'align_loss': 0.024843750521540642, + 'recon_loss': 0.05307048186659813, + 'predict_loss': 0.014350317418575287, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3023788332939148, + 'data_time': 0.0006755779904779047, + 'model_time': 1.3026276309974492, + 'grad_norm_pre_clip_avg': 0.25802104622125627, + 'learning_rate': 2.3189342478312885e-05, + 'epoch': 3.23} +04/19 [15:59:30] INFO | >> train_qwenlatent.py:487 + Step 12820 | grad_norm_pre_clip=0.1956 | + grad_norm_pre_clip_avg=0.2240 | Metrics: + {'align_loss': 0.023805581033229828, + 'recon_loss': 0.042658258229494095, + 'predict_loss': 0.014151236973702908, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1956402063369751, + 'data_time': 0.0007494989840779454, + 'model_time': 1.252312868979061, + 'grad_norm_pre_clip_avg': 0.22397076040506364, + 'learning_rate': 2.3184818554108125e-05, + 'epoch': 3.23} +04/19 [15:59:43] INFO | >> train_qwenlatent.py:487 + Step 12830 | grad_norm_pre_clip=0.2456 | + grad_norm_pre_clip_avg=0.2154 | Metrics: + {'align_loss': 0.024434246122837067, + 'recon_loss': 0.04735568165779114, + 'predict_loss': 0.013545007444918156, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24555432796478271, + 'data_time': 0.000717602000804618, + 'model_time': 1.2239332929893862, + 'grad_norm_pre_clip_avg': 0.21541452705860137, + 'learning_rate': 2.318028942834495e-05, + 'epoch': 3.24} +04/19 [15:59:55] INFO | >> train_qwenlatent.py:487 + Step 12840 | grad_norm_pre_clip=0.2466 | + grad_norm_pre_clip_avg=0.2412 | Metrics: + {'align_loss': 0.02462982013821602, + 'recon_loss': 0.05611806735396385, + 'predict_loss': 0.01254822313785553, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.246611088514328, + 'data_time': 0.0008247230143751949, + 'model_time': 1.2666281819983851, + 'grad_norm_pre_clip_avg': 0.2411678910255432, + 'learning_rate': 2.31757551032308e-05, 'epoch': + 3.24} +04/19 [16:00:09] INFO | >> train_qwenlatent.py:487 + Step 12850 | grad_norm_pre_clip=0.2563 | + grad_norm_pre_clip_avg=0.2165 | Metrics: + {'align_loss': 0.024842822924256325, + 'recon_loss': 0.050889670848846436, + 'predict_loss': 0.015684951096773148, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25630825757980347, + 'mae_score': 0.017190146231436516, 'data_time': + 0.0006832889921497554, 'model_time': + 1.2192796979798004, 'grad_norm_pre_clip_avg': + 0.21648075431585312, 'learning_rate': + 2.3171215580975654e-05, 'epoch': 3.24} +04/19 [16:00:21] INFO | >> train_qwenlatent.py:487 + Step 12860 | grad_norm_pre_clip=0.3156 | + grad_norm_pre_clip_avg=0.2574 | Metrics: + {'align_loss': 0.024822384119033813, + 'recon_loss': 0.04861627146601677, + 'predict_loss': 0.013842219486832619, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3155624270439148, + 'data_time': 0.0007615630165673792, + 'model_time': 1.3117527819995303, + 'grad_norm_pre_clip_avg': 0.257376654446125, + 'learning_rate': 2.3166670863792013e-05, + 'epoch': 3.25} +04/19 [16:00:34] INFO | >> train_qwenlatent.py:487 + Step 12870 | grad_norm_pre_clip=0.1717 | + grad_norm_pre_clip_avg=0.2674 | Metrics: + {'align_loss': 0.025113828480243683, + 'recon_loss': 0.049419425427913666, + 'predict_loss': 0.011578479781746864, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1716756373643875, + 'data_time': 0.0011286519875284284, + 'model_time': 1.5831285179883707, + 'grad_norm_pre_clip_avg': 0.2674055054783821, + 'learning_rate': 2.3162120953894927e-05, + 'epoch': 3.25} +04/19 [16:00:47] INFO | >> train_qwenlatent.py:487 + Step 12880 | grad_norm_pre_clip=0.2063 | + grad_norm_pre_clip_avg=0.2231 | Metrics: + {'align_loss': 0.025145579129457474, + 'recon_loss': 0.04599907249212265, + 'predict_loss': 0.012045932933688164, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20628248155117035, + 'data_time': 0.0011394670000299811, + 'model_time': 1.236716280982364, + 'grad_norm_pre_clip_avg': 0.2230769142508507, + 'learning_rate': 2.3157565853501956e-05, + 'epoch': 3.25} +04/19 [16:01:00] INFO | >> train_qwenlatent.py:487 + Step 12890 | grad_norm_pre_clip=0.1933 | + grad_norm_pre_clip_avg=0.2176 | Metrics: + {'align_loss': 0.02348240464925766, + 'recon_loss': 0.037314098328351974, + 'predict_loss': 0.017070161178708076, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19328673183918, + 'data_time': 0.0009120619797613472, + 'model_time': 1.254005031019915, + 'grad_norm_pre_clip_avg': 0.21763252466917038, + 'learning_rate': 2.3153005564833208e-05, + 'epoch': 3.25} +04/19 [16:01:13] INFO | >> train_qwenlatent.py:487 + Step 12900 | grad_norm_pre_clip=0.3326 | + grad_norm_pre_clip_avg=0.2260 | Metrics: + {'align_loss': 0.02362045645713806, + 'recon_loss': 0.04922119900584221, + 'predict_loss': 0.016984574496746063, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3325854539871216, + 'mae_score': 0.0149044689831433, 'data_time': + 0.0009629920241422951, 'model_time': + 1.2397804200008977, 'grad_norm_pre_clip_avg': + 0.22596842646598816, 'learning_rate': + 2.3148440090111305e-05, 'epoch': 3.26} +04/19 [16:01:25] INFO | >> train_qwenlatent.py:487 + Step 12910 | grad_norm_pre_clip=0.2797 | + grad_norm_pre_clip_avg=0.3043 | Metrics: + {'align_loss': 0.02442394383251667, + 'recon_loss': 0.057374946773052216, + 'predict_loss': 0.013868468813598156, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2797005772590637, + 'data_time': 0.0009589119872543961, + 'model_time': 1.2655411710147746, + 'grad_norm_pre_clip_avg': 0.30430266857147215, + 'learning_rate': 2.3143869431561415e-05, + 'epoch': 3.26} +04/19 [16:01:38] INFO | >> train_qwenlatent.py:487 + Step 12920 | grad_norm_pre_clip=0.2434 | + grad_norm_pre_clip_avg=0.2551 | Metrics: + {'align_loss': 0.025724075734615326, + 'recon_loss': 0.05697552114725113, + 'predict_loss': 0.01153942197561264, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24342751502990723, + 'data_time': 0.0006506630161311477, + 'model_time': 1.2272561969875824, + 'grad_norm_pre_clip_avg': 0.25505486875772476, + 'learning_rate': 2.313929359141122e-05, + 'epoch': 3.26} +04/19 [16:01:50] INFO | >> train_qwenlatent.py:487 + Step 12930 | grad_norm_pre_clip=0.2488 | + grad_norm_pre_clip_avg=0.2121 | Metrics: + {'align_loss': 0.02549603581428528, + 'recon_loss': 0.0635429173707962, + 'predict_loss': 0.025833332911133766, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24883708357810974, + 'data_time': 0.0008346919785253704, + 'model_time': 1.265778942994075, + 'grad_norm_pre_clip_avg': 0.21212171614170075, + 'learning_rate': 2.3134712571890917e-05, + 'epoch': 3.26} +04/19 [16:02:03] INFO | >> train_qwenlatent.py:487 + Step 12940 | grad_norm_pre_clip=0.1913 | + grad_norm_pre_clip_avg=0.2033 | Metrics: + {'align_loss': 0.02467738650739193, + 'recon_loss': 0.04336102679371834, + 'predict_loss': 0.011560533195734024, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19125071167945862, + 'data_time': 0.0008314409933518618, + 'model_time': 1.2232435859914403, + 'grad_norm_pre_clip_avg': 0.20325447618961334, + 'learning_rate': 2.3130126375233243e-05, + 'epoch': 3.27} +04/19 [16:02:16] INFO | >> train_qwenlatent.py:487 + Step 12950 | grad_norm_pre_clip=0.2470 | + grad_norm_pre_clip_avg=0.3006 | Metrics: + {'align_loss': 0.023456072434782982, + 'recon_loss': 0.051084063947200775, + 'predict_loss': 0.015039128251373768, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2470495104789734, + 'mae_score': 0.019319877968178138, 'data_time': + 0.001029573002597317, 'model_time': + 1.1925016489985865, 'grad_norm_pre_clip_avg': + 0.3006269782781601, 'learning_rate': + 2.3125535003673467e-05, 'epoch': 3.27} +04/19 [16:02:29] INFO | >> train_qwenlatent.py:487 + Step 12960 | grad_norm_pre_clip=0.1842 | + grad_norm_pre_clip_avg=0.3019 | Metrics: + {'align_loss': 0.024886062368750572, + 'recon_loss': 0.053421083837747574, + 'predict_loss': 0.014686713926494122, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18422122299671173, + 'data_time': 0.0008010529854800552, + 'model_time': 1.2261662240198348, + 'grad_norm_pre_clip_avg': 0.30191676169633863, + 'learning_rate': 2.3120938459449354e-05, + 'epoch': 3.27} +04/19 [16:02:42] INFO | >> train_qwenlatent.py:487 + Step 12970 | grad_norm_pre_clip=0.1781 | + grad_norm_pre_clip_avg=0.2458 | Metrics: + {'align_loss': 0.02506362833082676, + 'recon_loss': 0.04288393259048462, + 'predict_loss': 0.012319953180849552, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1780824512243271, + 'data_time': 0.0009529889794066548, + 'model_time': 1.2066173469938803, + 'grad_norm_pre_clip_avg': 0.24580653309822081, + 'learning_rate': 2.311633674480121e-05, + 'epoch': 3.27} +04/19 [16:02:54] INFO | >> train_qwenlatent.py:487 + Step 12980 | grad_norm_pre_clip=0.1911 | + grad_norm_pre_clip_avg=0.2033 | Metrics: + {'align_loss': 0.024620693176984787, + 'recon_loss': 0.03802796080708504, + 'predict_loss': 0.010616736486554146, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19113874435424805, + 'data_time': 0.0006481210002675653, + 'model_time': 1.2418597229989246, + 'grad_norm_pre_clip_avg': 0.2032608211040497, + 'learning_rate': 2.311172986197185e-05, + 'epoch': 3.28} +04/19 [16:03:07] INFO | >> train_qwenlatent.py:487 + Step 12990 | grad_norm_pre_clip=0.2133 | + grad_norm_pre_clip_avg=0.2296 | Metrics: + {'align_loss': 0.02523835375905037, + 'recon_loss': 0.05211721360683441, + 'predict_loss': 0.013029045425355434, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21325992047786713, + 'data_time': 0.0009440879803150892, + 'model_time': 1.2508837640052661, + 'grad_norm_pre_clip_avg': 0.2295895501971245, + 'learning_rate': 2.310711781320662e-05, + 'epoch': 3.28} +04/19 [16:03:20] INFO | >> train_qwenlatent.py:487 + Step 13000 | grad_norm_pre_clip=0.2185 | + grad_norm_pre_clip_avg=0.2297 | Metrics: + {'align_loss': 0.024963878095149994, + 'recon_loss': 0.05865481495857239, + 'predict_loss': 0.02020180970430374, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2185225635766983, + 'mae_score': 0.017396877048251866, 'data_time': + 0.0006814609805587679, 'model_time': + 1.2243539769842755, 'grad_norm_pre_clip_avg': + 0.2297130897641182, 'learning_rate': + 2.310250060075337e-05, 'epoch': 3.28} +04/19 [16:03:33] INFO | >> train_qwenlatent.py:487 + Step 13010 | grad_norm_pre_clip=0.2738 | + grad_norm_pre_clip_avg=0.2154 | Metrics: + {'align_loss': 0.025410808622837067, + 'recon_loss': 0.056454673409461975, + 'predict_loss': 0.017334692180156708, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2738073170185089, + 'data_time': 0.0013475290033966303, + 'model_time': 1.3276017729949672, + 'grad_norm_pre_clip_avg': 0.2153926134109497, + 'learning_rate': 2.309787822686248e-05, + 'epoch': 3.28} +04/19 [16:03:46] INFO | >> train_qwenlatent.py:487 + Step 13020 | grad_norm_pre_clip=0.2642 | + grad_norm_pre_clip_avg=0.2211 | Metrics: + {'align_loss': 0.025303004309535027, + 'recon_loss': 0.041991837322711945, + 'predict_loss': 0.012031764723360538, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2641676366329193, + 'data_time': 0.0006660029757767916, + 'model_time': 1.2318892849725671, + 'grad_norm_pre_clip_avg': 0.22112882137298584, + 'learning_rate': 2.309325069378683e-05, + 'epoch': 3.29} +04/19 [16:03:59] INFO | >> train_qwenlatent.py:487 + Step 13030 | grad_norm_pre_clip=0.2289 | + grad_norm_pre_clip_avg=0.2526 | Metrics: + {'align_loss': 0.025480050593614578, + 'recon_loss': 0.04892614856362343, + 'predict_loss': 0.01500806212425232, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22891493141651154, + 'data_time': 0.0010516600159462541, + 'model_time': 1.256904989015311, + 'grad_norm_pre_clip_avg': 0.25258357226848605, + 'learning_rate': 2.308861800378183e-05, + 'epoch': 3.29} +04/19 [16:04:11] INFO | >> train_qwenlatent.py:487 + Step 13040 | grad_norm_pre_clip=0.2058 | + grad_norm_pre_clip_avg=0.2261 | Metrics: + {'align_loss': 0.024741094559431076, + 'recon_loss': 0.050506848841905594, + 'predict_loss': 0.01217994000762701, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2057631015777588, + 'data_time': 0.0011741589987650514, + 'model_time': 1.1827361659961753, + 'grad_norm_pre_clip_avg': 0.22611993700265884, + 'learning_rate': 2.3083980159105395e-05, + 'epoch': 3.29} +04/19 [16:04:24] INFO | >> train_qwenlatent.py:487 + Step 13050 | grad_norm_pre_clip=0.1963 | + grad_norm_pre_clip_avg=0.2213 | Metrics: + {'align_loss': 0.024225812405347824, + 'recon_loss': 0.05586947500705719, + 'predict_loss': 0.019426561892032623, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19634497165679932, + 'mae_score': 0.018352592743194854, 'data_time': + 0.0006684009858872741, 'model_time': + 1.2302324810007121, 'grad_norm_pre_clip_avg': + 0.22125525325536727, 'learning_rate': + 2.3079337162017957e-05, 'epoch': 3.29} +04/19 [16:04:37] INFO | >> train_qwenlatent.py:487 + Step 13060 | grad_norm_pre_clip=0.2578 | + grad_norm_pre_clip_avg=0.2567 | Metrics: + {'align_loss': 0.024372760206460953, + 'recon_loss': 0.04809658229351044, + 'predict_loss': 0.019923122599720955, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.257775217294693, + 'data_time': 0.0008600250002928078, + 'model_time': 1.2513777219865005, + 'grad_norm_pre_clip_avg': 0.2566847875714302, + 'learning_rate': 2.307468901478245e-05, + 'epoch': 3.3} +04/19 [16:04:50] INFO | >> train_qwenlatent.py:487 + Step 13070 | grad_norm_pre_clip=0.1907 | + grad_norm_pre_clip_avg=0.2305 | Metrics: + {'align_loss': 0.024781806394457817, + 'recon_loss': 0.052676476538181305, + 'predict_loss': 0.014903448522090912, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1907370537519455, + 'data_time': 0.000672577996738255, + 'model_time': 1.2798533549939748, + 'grad_norm_pre_clip_avg': 0.23052597641944886, + 'learning_rate': 2.3070035719664332e-05, + 'epoch': 3.3} +04/19 [16:05:02] INFO | >> train_qwenlatent.py:487 + Step 13080 | grad_norm_pre_clip=0.2118 | + grad_norm_pre_clip_avg=0.2078 | Metrics: + {'align_loss': 0.02552703395485878, + 'recon_loss': 0.059536661952733994, + 'predict_loss': 0.012795226648449898, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21175748109817505, + 'data_time': 0.0006889580108691007, + 'model_time': 1.2816199189983308, + 'grad_norm_pre_clip_avg': 0.20781798660755157, + 'learning_rate': 2.3065377278931556e-05, + 'epoch': 3.3} +04/19 [16:05:15] INFO | >> train_qwenlatent.py:487 + Step 13090 | grad_norm_pre_clip=0.2073 | + grad_norm_pre_clip_avg=0.2362 | Metrics: + {'align_loss': 0.026157215237617493, + 'recon_loss': 0.06454861909151077, + 'predict_loss': 0.01770544797182083, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20725613832473755, + 'data_time': 0.0006937819998711348, + 'model_time': 1.2169135980075225, + 'grad_norm_pre_clip_avg': 0.236199489235878, + 'learning_rate': 2.3060713694854594e-05, + 'epoch': 3.3} +04/19 [16:05:28] INFO | >> train_qwenlatent.py:487 + Step 13100 | grad_norm_pre_clip=0.2286 | + grad_norm_pre_clip_avg=0.2329 | Metrics: + {'align_loss': 0.02437497116625309, + 'recon_loss': 0.03400189429521561, + 'predict_loss': 0.009755665436387062, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2286432683467865, + 'mae_score': 0.016172744132377005, 'data_time': + 0.0013575080083683133, 'model_time': + 1.2399764090077952, 'grad_norm_pre_clip_avg': + 0.23286219239234923, 'learning_rate': + 2.3056044969706413e-05, 'epoch': 3.31} +04/19 [16:05:41] INFO | >> train_qwenlatent.py:487 + Step 13110 | grad_norm_pre_clip=0.2085 | + grad_norm_pre_clip_avg=0.2388 | Metrics: + {'align_loss': 0.02518836408853531, + 'recon_loss': 0.043672915548086166, + 'predict_loss': 0.014154785312712193, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20848329365253448, + 'data_time': 0.0009494989935774356, + 'model_time': 1.265814345999388, + 'grad_norm_pre_clip_avg': 0.2387649416923523, + 'learning_rate': 2.3051371105762504e-05, + 'epoch': 3.31} +04/19 [16:05:53] INFO | >> train_qwenlatent.py:487 + Step 13120 | grad_norm_pre_clip=0.3387 | + grad_norm_pre_clip_avg=0.2199 | Metrics: + {'align_loss': 0.025528457015752792, + 'recon_loss': 0.0694613829255104, + 'predict_loss': 0.018056416884064674, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.33873212337493896, + 'data_time': 0.0007122110109776258, + 'model_time': 1.252966539002955, + 'grad_norm_pre_clip_avg': 0.21990081816911697, + 'learning_rate': 2.3046692105300844e-05, + 'epoch': 3.31} +04/19 [16:06:06] INFO | >> train_qwenlatent.py:487 + Step 13130 | grad_norm_pre_clip=0.1895 | + grad_norm_pre_clip_avg=0.2865 | Metrics: + {'align_loss': 0.025156309828162193, + 'recon_loss': 0.04402953386306763, + 'predict_loss': 0.011136345565319061, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18950296938419342, + 'data_time': 0.0010368599905632436, + 'model_time': 1.2444954579987098, + 'grad_norm_pre_clip_avg': 0.286488239467144, + 'learning_rate': 2.3042007970601918e-05, + 'epoch': 3.31} +04/19 [16:06:19] INFO | >> train_qwenlatent.py:487 + Step 13140 | grad_norm_pre_clip=0.2448 | + grad_norm_pre_clip_avg=0.2213 | Metrics: + {'align_loss': 0.024738460779190063, + 'recon_loss': 0.050613708794116974, + 'predict_loss': 0.017044680193066597, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24483203887939453, + 'data_time': 0.0011139240232296288, + 'model_time': 1.2273574910068419, + 'grad_norm_pre_clip_avg': 0.2213344931602478, + 'learning_rate': 2.303731870394873e-05, + 'epoch': 3.32} +04/19 [16:06:32] INFO | >> train_qwenlatent.py:487 + Step 13150 | grad_norm_pre_clip=0.2085 | + grad_norm_pre_clip_avg=0.2382 | Metrics: + {'align_loss': 0.025533776730298996, + 'recon_loss': 0.05916932225227356, + 'predict_loss': 0.018032293766736984, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20848184823989868, + 'mae_score': 0.015975605045352972, 'data_time': + 0.0008919640094973147, 'model_time': + 1.2126291189924814, 'grad_norm_pre_clip_avg': + 0.23823564350605012, 'learning_rate': + 2.3032624307626757e-05, 'epoch': 3.32} +04/19 [16:06:45] INFO | >> train_qwenlatent.py:487 + Step 13160 | grad_norm_pre_clip=0.2022 | + grad_norm_pre_clip_avg=0.2225 | Metrics: + {'align_loss': 0.024462945759296417, + 'recon_loss': 0.04710135981440544, + 'predict_loss': 0.012280878610908985, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20222657918930054, + 'data_time': 0.0006462860037572682, + 'model_time': 1.1918065510108136, + 'grad_norm_pre_clip_avg': 0.2224590077996254, + 'learning_rate': 2.3027924783923994e-05, + 'epoch': 3.32} +04/19 [16:06:57] INFO | >> train_qwenlatent.py:487 + Step 13170 | grad_norm_pre_clip=0.2866 | + grad_norm_pre_clip_avg=0.2468 | Metrics: + {'align_loss': 0.025619013234972954, + 'recon_loss': 0.04175345227122307, + 'predict_loss': 0.0148174362257123, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2865970730781555, + 'data_time': 0.000672166992444545, + 'model_time': 1.237973730021622, + 'grad_norm_pre_clip_avg': 0.24681729525327684, + 'learning_rate': 2.302322013513094e-05, + 'epoch': 3.32} +04/19 [16:07:10] INFO | >> train_qwenlatent.py:487 + Step 13180 | grad_norm_pre_clip=0.2555 | + grad_norm_pre_clip_avg=0.2594 | Metrics: + {'align_loss': 0.024628743529319763, + 'recon_loss': 0.05059653893113136, + 'predict_loss': 0.018062030896544456, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25550907850265503, + 'data_time': 0.0008539340051356703, + 'model_time': 1.2438064829912037, + 'grad_norm_pre_clip_avg': 0.25935421139001846, + 'learning_rate': 2.3018510363540575e-05, + 'epoch': 3.33} +04/19 [16:07:23] INFO | >> train_qwenlatent.py:487 + Step 13190 | grad_norm_pre_clip=0.2380 | + grad_norm_pre_clip_avg=0.2550 | Metrics: + {'align_loss': 0.02450578473508358, + 'recon_loss': 0.06004730984568596, + 'predict_loss': 0.01982106827199459, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23800361156463623, + 'data_time': 0.0009421290014870465, + 'model_time': 1.303668792010285, + 'grad_norm_pre_clip_avg': 0.25496678054332733, + 'learning_rate': 2.3013795471448387e-05, + 'epoch': 3.33} +04/19 [16:07:36] INFO | >> train_qwenlatent.py:487 + Step 13200 | grad_norm_pre_clip=0.2432 | + grad_norm_pre_clip_avg=0.2356 | Metrics: + {'align_loss': 0.02474788948893547, + 'recon_loss': 0.04995005205273628, + 'predict_loss': 0.009564300067722797, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2431759387254715, + 'mae_score': 0.01836850965345228, 'data_time': + 0.0006953980191610754, 'model_time': + 1.2709074399899691, 'grad_norm_pre_clip_avg': + 0.2356452688574791, 'learning_rate': + 2.3009075461152358e-05, 'epoch': 3.33} +04/19 [16:07:49] INFO | >> train_qwenlatent.py:487 + Step 13210 | grad_norm_pre_clip=0.2053 | + grad_norm_pre_clip_avg=0.2248 | Metrics: + {'align_loss': 0.02457461506128311, + 'recon_loss': 0.03981243446469307, + 'predict_loss': 0.009133253246545792, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20531906187534332, + 'data_time': 0.001483661006204784, + 'model_time': 1.3158067419717554, + 'grad_norm_pre_clip_avg': 0.22482995092868804, + 'learning_rate': 2.3004350334952965e-05, + 'epoch': 3.33} +04/19 [16:08:01] INFO | >> train_qwenlatent.py:487 + Step 13220 | grad_norm_pre_clip=0.2326 | + grad_norm_pre_clip_avg=0.2088 | Metrics: + {'align_loss': 0.02512466162443161, + 'recon_loss': 0.05516587942838669, + 'predict_loss': 0.014568501152098179, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23264971375465393, + 'data_time': 0.0006413600058294833, + 'model_time': 1.2496291400166228, + 'grad_norm_pre_clip_avg': 0.2088448539376259, + 'learning_rate': 2.299962009515317e-05, + 'epoch': 3.34} +04/19 [16:08:14] INFO | >> train_qwenlatent.py:487 + Step 13230 | grad_norm_pre_clip=0.2125 | + grad_norm_pre_clip_avg=0.2595 | Metrics: + {'align_loss': 0.02450033277273178, + 'recon_loss': 0.038538455963134766, + 'predict_loss': 0.01209011860191822, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2124767154455185, + 'data_time': 0.0006506950012408197, + 'model_time': 1.2554928280005697, + 'grad_norm_pre_clip_avg': 0.25952747613191607, + 'learning_rate': 2.2994884744058446e-05, + 'epoch': 3.34} +04/19 [16:08:26] INFO | >> train_qwenlatent.py:487 + Step 13240 | grad_norm_pre_clip=0.2865 | + grad_norm_pre_clip_avg=0.2580 | Metrics: + {'align_loss': 0.025613613426685333, + 'recon_loss': 0.0534479133784771, + 'predict_loss': 0.02028915099799633, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.286450058221817, + 'data_time': 0.0008230610110331327, + 'model_time': 1.2191640850214753, + 'grad_norm_pre_clip_avg': 0.25799132883548737, + 'learning_rate': 2.2990144283976732e-05, + 'epoch': 3.34} +04/19 [16:08:40] INFO | >> train_qwenlatent.py:487 + Step 13250 | grad_norm_pre_clip=0.2524 | + grad_norm_pre_clip_avg=0.2644 | Metrics: + {'align_loss': 0.02362913265824318, + 'recon_loss': 0.03812175616621971, + 'predict_loss': 0.014177690260112286, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25244560837745667, + 'mae_score': 0.019625471304128835, 'data_time': + 0.0006656079785898328, 'model_time': + 1.2016361909918487, 'grad_norm_pre_clip_avg': + 0.264406231045723, 'learning_rate': + 2.2985398717218483e-05, 'epoch': 3.34} +04/19 [16:08:52] INFO | >> train_qwenlatent.py:487 + Step 13260 | grad_norm_pre_clip=0.2314 | + grad_norm_pre_clip_avg=0.2540 | Metrics: + {'align_loss': 0.02584059163928032, + 'recon_loss': 0.06978362798690796, + 'predict_loss': 0.01690797507762909, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23139752447605133, + 'data_time': 0.0014131810166873038, + 'model_time': 1.217340071016224, + 'grad_norm_pre_clip_avg': 0.2540204718708992, + 'learning_rate': 2.2980648046096623e-05, + 'epoch': 3.35} +04/19 [16:09:06] INFO | >> train_qwenlatent.py:487 + Step 13270 | grad_norm_pre_clip=0.1908 | + grad_norm_pre_clip_avg=0.2106 | Metrics: + {'align_loss': 0.026236452162265778, + 'recon_loss': 0.06259609758853912, + 'predict_loss': 0.016018839552998543, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19084258377552032, + 'data_time': 0.0009821699932217598, + 'model_time': 1.2495858990005217, + 'grad_norm_pre_clip_avg': 0.2106145590543747, + 'learning_rate': 2.2975892272926578e-05, + 'epoch': 3.35} +04/19 [16:09:18] INFO | >> train_qwenlatent.py:487 + Step 13280 | grad_norm_pre_clip=0.1714 | + grad_norm_pre_clip_avg=0.2108 | Metrics: + {'align_loss': 0.023968365043401718, + 'recon_loss': 0.05291082710027695, + 'predict_loss': 0.008919435553252697, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.171443372964859, + 'data_time': 0.001010999985737726, + 'model_time': 1.2679174490040168, + 'grad_norm_pre_clip_avg': 0.21081891506910325, + 'learning_rate': 2.2971131400026247e-05, + 'epoch': 3.35} +04/19 [16:09:31] INFO | >> train_qwenlatent.py:487 + Step 13290 | grad_norm_pre_clip=0.2045 | + grad_norm_pre_clip_avg=0.2682 | Metrics: + {'align_loss': 0.023976914584636688, + 'recon_loss': 0.05208340287208557, + 'predict_loss': 0.017137210816144943, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20453205704689026, + 'data_time': 0.0007150150195229799, + 'model_time': 1.23863152298145, + 'grad_norm_pre_clip_avg': 0.26818500012159346, + 'learning_rate': 2.2966365429716025e-05, + 'epoch': 3.35} +04/19 [16:09:44] INFO | >> train_qwenlatent.py:487 + Step 13300 | grad_norm_pre_clip=0.3000 | + grad_norm_pre_clip_avg=0.2669 | Metrics: + {'align_loss': 0.0240146704018116, + 'recon_loss': 0.04785432666540146, + 'predict_loss': 0.018591707572340965, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2999822497367859, + 'mae_score': 0.016034825642903646, 'data_time': + 0.0008272529812529683, 'model_time': + 1.282146957993973, 'grad_norm_pre_clip_avg': + 0.2669441893696785, 'learning_rate': + 2.2961594364318785e-05, 'epoch': 3.36} +04/19 [16:09:57] INFO | >> train_qwenlatent.py:487 + Step 13310 | grad_norm_pre_clip=0.2142 | + grad_norm_pre_clip_avg=0.2322 | Metrics: + {'align_loss': 0.02524033933877945, + 'recon_loss': 0.05334906652569771, + 'predict_loss': 0.01416447851806879, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21417231857776642, + 'data_time': 0.000936224008910358, + 'model_time': 1.306104415009031, + 'grad_norm_pre_clip_avg': 0.23216025978326799, + 'learning_rate': 2.295681820615989e-05, + 'epoch': 3.36} +04/19 [16:10:10] INFO | >> train_qwenlatent.py:487 + Step 13320 | grad_norm_pre_clip=0.1937 | + grad_norm_pre_clip_avg=0.2235 | Metrics: + {'align_loss': 0.02374066412448883, + 'recon_loss': 0.04073231294751167, + 'predict_loss': 0.010269153863191605, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19369103014469147, + 'data_time': 0.0010738259879872203, + 'model_time': 1.2836681559856515, + 'grad_norm_pre_clip_avg': 0.22354266792535782, + 'learning_rate': 2.295203695756718e-05, + 'epoch': 3.36} +04/19 [16:10:23] INFO | >> train_qwenlatent.py:487 + Step 13330 | grad_norm_pre_clip=0.1999 | + grad_norm_pre_clip_avg=0.2147 | Metrics: + {'align_loss': 0.024180885404348373, + 'recon_loss': 0.045105695724487305, + 'predict_loss': 0.012972365133464336, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19986997544765472, + 'data_time': 0.0006915480189491063, + 'model_time': 1.602524732996244, + 'grad_norm_pre_clip_avg': 0.21471496522426606, + 'learning_rate': 2.2947250620870982e-05, + 'epoch': 3.36} +04/19 [16:10:36] INFO | >> train_qwenlatent.py:487 + Step 13340 | grad_norm_pre_clip=0.1779 | + grad_norm_pre_clip_avg=0.1929 | Metrics: + {'align_loss': 0.022898370400071144, + 'recon_loss': 0.03947843983769417, + 'predict_loss': 0.013621089048683643, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17787791788578033, + 'data_time': 0.001102731010178104, + 'model_time': 1.3505696519860066, + 'grad_norm_pre_clip_avg': 0.19289056956768036, + 'learning_rate': 2.2942459198404092e-05, + 'epoch': 3.37} +04/19 [16:10:49] INFO | >> train_qwenlatent.py:487 + Step 13350 | grad_norm_pre_clip=0.3305 | + grad_norm_pre_clip_avg=0.2922 | Metrics: + {'align_loss': 0.02480604499578476, + 'recon_loss': 0.06177801266312599, + 'predict_loss': 0.020008167251944542, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3305146098136902, + 'mae_score': 0.026493220286326367, 'data_time': + 0.0008856059866957366, 'model_time': + 1.2299035789910704, 'grad_norm_pre_clip_avg': + 0.29221338778734207, 'learning_rate': + 2.293766269250179e-05, 'epoch': 3.37} +04/19 [16:11:01] INFO | >> train_qwenlatent.py:487 + Step 13360 | grad_norm_pre_clip=0.2121 | + grad_norm_pre_clip_avg=0.2076 | Metrics: + {'align_loss': 0.024566886946558952, + 'recon_loss': 0.051612455397844315, + 'predict_loss': 0.014472173526883125, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21209776401519775, + 'data_time': 0.0007843129860702902, + 'model_time': 1.216711918998044, + 'grad_norm_pre_clip_avg': 0.20761172622442245, + 'learning_rate': 2.2932861105501842e-05, + 'epoch': 3.37} +04/19 [16:11:14] INFO | >> train_qwenlatent.py:487 + Step 13370 | grad_norm_pre_clip=0.2172 | + grad_norm_pre_clip_avg=0.1893 | Metrics: + {'align_loss': 0.02523680217564106, + 'recon_loss': 0.04764728620648384, + 'predict_loss': 0.013148960657417774, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21720682084560394, + 'data_time': 0.0006468970095738769, + 'model_time': 1.2349417400255334, + 'grad_norm_pre_clip_avg': 0.18926628977060317, + 'learning_rate': 2.2928054439744475e-05, + 'epoch': 3.37} +04/19 [16:11:26] INFO | >> train_qwenlatent.py:487 + Step 13380 | grad_norm_pre_clip=0.2522 | + grad_norm_pre_clip_avg=0.2213 | Metrics: + {'align_loss': 0.02422877959907055, + 'recon_loss': 0.049791119992733, + 'predict_loss': 0.014575435779988766, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25224629044532776, + 'data_time': 0.0008479320094920695, + 'model_time': 1.221818980993703, + 'grad_norm_pre_clip_avg': 0.22129198908805847, + 'learning_rate': 2.2923242697572407e-05, + 'epoch': 3.38} +04/19 [16:11:39] INFO | >> train_qwenlatent.py:487 + Step 13390 | grad_norm_pre_clip=0.2961 | + grad_norm_pre_clip_avg=0.2578 | Metrics: + {'align_loss': 0.025652796030044556, + 'recon_loss': 0.06686253100633621, + 'predict_loss': 0.020499229431152344, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.29612496495246887, + 'data_time': 0.0008704399806447327, + 'model_time': 1.2309697050077375, + 'grad_norm_pre_clip_avg': 0.25781470388174055, + 'learning_rate': 2.2918425881330818e-05, + 'epoch': 3.38} +04/19 [16:11:52] INFO | >> train_qwenlatent.py:487 + Step 13400 | grad_norm_pre_clip=0.2529 | + grad_norm_pre_clip_avg=0.2714 | Metrics: + {'align_loss': 0.024988442659378052, + 'recon_loss': 0.0536649227142334, + 'predict_loss': 0.013597959652543068, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2529008686542511, + 'mae_score': 0.029413604736328126, 'data_time': + 0.0006670379952993244, 'model_time': + 1.1893779770180117, 'grad_norm_pre_clip_avg': + 0.2713559418916702, 'learning_rate': + 2.291360399336737e-05, 'epoch': 3.38} +04/19 [16:12:05] INFO | >> train_qwenlatent.py:487 + Step 13410 | grad_norm_pre_clip=0.1853 | + grad_norm_pre_clip_avg=0.2415 | Metrics: + {'align_loss': 0.02378236874938011, + 'recon_loss': 0.03935401514172554, + 'predict_loss': 0.009964453056454659, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18527281284332275, + 'data_time': 0.0011337109899614006, + 'model_time': 1.2906205390172545, + 'grad_norm_pre_clip_avg': 0.24149799793958665, + 'learning_rate': 2.2908777036032186e-05, + 'epoch': 3.38} +04/19 [16:12:18] INFO | >> train_qwenlatent.py:487 + Step 13420 | grad_norm_pre_clip=0.1700 | + grad_norm_pre_clip_avg=0.2194 | Metrics: + {'align_loss': 0.023525379598140717, + 'recon_loss': 0.04927605390548706, + 'predict_loss': 0.01501286868005991, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16999660432338715, + 'data_time': 0.0013914710143581033, + 'model_time': 1.270591113017872, + 'grad_norm_pre_clip_avg': 0.21940482705831527, + 'learning_rate': 2.2903945011677873e-05, + 'epoch': 3.39} +04/19 [16:12:30] INFO | >> train_qwenlatent.py:487 + Step 13430 | grad_norm_pre_clip=0.2231 | + grad_norm_pre_clip_avg=0.2318 | Metrics: + {'align_loss': 0.024751469492912292, + 'recon_loss': 0.05745333060622215, + 'predict_loss': 0.018850963562726974, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22308406233787537, + 'data_time': 0.0009338029776699841, + 'model_time': 1.224353108991636, + 'grad_norm_pre_clip_avg': 0.2318190336227417, + 'learning_rate': 2.2899107922659493e-05, + 'epoch': 3.39} +04/19 [16:12:43] INFO | >> train_qwenlatent.py:487 + Step 13440 | grad_norm_pre_clip=0.1780 | + grad_norm_pre_clip_avg=0.2463 | Metrics: + {'align_loss': 0.024738285690546036, + 'recon_loss': 0.05085385963320732, + 'predict_loss': 0.013978890143334866, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1780417263507843, + 'data_time': 0.0006912760145496577, + 'model_time': 1.2807492820138577, + 'grad_norm_pre_clip_avg': 0.2463470533490181, + 'learning_rate': 2.2894265771334595e-05, + 'epoch': 3.39} +04/19 [16:12:56] INFO | >> train_qwenlatent.py:487 + Step 13450 | grad_norm_pre_clip=0.2221 | + grad_norm_pre_clip_avg=0.2207 | Metrics: + {'align_loss': 0.024500831961631775, + 'recon_loss': 0.060097355395555496, + 'predict_loss': 0.014951006509363651, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22207070887088776, + 'mae_score': 0.022995979721481735, 'data_time': + 0.0008787910046521574, 'model_time': + 1.4964955379837193, 'grad_norm_pre_clip_avg': + 0.22068696916103364, 'learning_rate': + 2.2889418560063176e-05, 'epoch': 3.39} +04/19 [16:13:09] INFO | >> train_qwenlatent.py:487 + Step 13460 | grad_norm_pre_clip=0.3072 | + grad_norm_pre_clip_avg=0.2454 | Metrics: + {'align_loss': 0.02352590300142765, + 'recon_loss': 0.04598487168550491, + 'predict_loss': 0.015088457614183426, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3071608245372772, + 'data_time': 0.0007322299934457988, + 'model_time': 1.2203336219827179, + 'grad_norm_pre_clip_avg': 0.2453649014234543, + 'learning_rate': 2.2884566291207713e-05, + 'epoch': 3.4} +04/19 [16:13:22] INFO | >> train_qwenlatent.py:487 + Step 13470 | grad_norm_pre_clip=0.3213 | + grad_norm_pre_clip_avg=0.2464 | Metrics: + {'align_loss': 0.02497958391904831, + 'recon_loss': 0.05254972353577614, + 'predict_loss': 0.01589852198958397, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3213247060775757, + 'data_time': 0.0014485479914583266, + 'model_time': 1.215821757010417, + 'grad_norm_pre_clip_avg': 0.2463645949959755, + 'learning_rate': 2.2879708967133138e-05, + 'epoch': 3.4} +04/19 [16:13:34] INFO | >> train_qwenlatent.py:487 + Step 13480 | grad_norm_pre_clip=0.3005 | + grad_norm_pre_clip_avg=0.2410 | Metrics: + {'align_loss': 0.024870440363883972, + 'recon_loss': 0.03963014855980873, + 'predict_loss': 0.00706007843837142, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3005126416683197, + 'data_time': 0.0011826659901998937, + 'model_time': 1.1977404589997604, + 'grad_norm_pre_clip_avg': 0.2409542515873909, + 'learning_rate': 2.2874846590206855e-05, + 'epoch': 3.4} +04/19 [16:13:47] INFO | >> train_qwenlatent.py:487 + Step 13490 | grad_norm_pre_clip=0.2271 | + grad_norm_pre_clip_avg=0.2278 | Metrics: + {'align_loss': 0.025967635214328766, + 'recon_loss': 0.05509880185127258, + 'predict_loss': 0.013409722596406937, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2271115779876709, + 'data_time': 0.0011233690020162612, + 'model_time': 1.2409411180124152, + 'grad_norm_pre_clip_avg': 0.22778534144163132, + 'learning_rate': 2.2869979162798732e-05, + 'epoch': 3.4} +04/19 [16:14:00] INFO | >> train_qwenlatent.py:487 + Step 13500 | grad_norm_pre_clip=0.2450 | + grad_norm_pre_clip_avg=0.2245 | Metrics: + {'align_loss': 0.024856125935912132, + 'recon_loss': 0.044718123972415924, + 'predict_loss': 0.013856352306902409, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24503093957901, + 'mae_score': 0.02069508664242856, 'data_time': + 0.0008404710097238421, 'model_time': + 1.2634581869933754, 'grad_norm_pre_clip_avg': + 0.2244900494813919, 'learning_rate': + 2.2865106687281082e-05, 'epoch': 3.41} +04/19 [16:14:13] INFO | >> train_qwenlatent.py:487 + Step 13510 | grad_norm_pre_clip=0.3463 | + grad_norm_pre_clip_avg=0.2261 | Metrics: + {'align_loss': 0.02587154507637024, + 'recon_loss': 0.06424535810947418, + 'predict_loss': 0.017539693042635918, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.34627917408943176, + 'data_time': 0.000881974003277719, + 'model_time': 1.5648305659997277, + 'grad_norm_pre_clip_avg': 0.22613947987556457, + 'learning_rate': 2.28602291660287e-05, 'epoch': + 3.41} +04/19 [16:14:25] INFO | >> train_qwenlatent.py:487 + Step 13520 | grad_norm_pre_clip=0.2083 | + grad_norm_pre_clip_avg=0.2565 | Metrics: + {'align_loss': 0.02550724893808365, + 'recon_loss': 0.043303538113832474, + 'predict_loss': 0.013166404329240322, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20827800035476685, + 'data_time': 0.0007387830119114369, + 'model_time': 1.2551226849900559, + 'grad_norm_pre_clip_avg': 0.2564988672733307, + 'learning_rate': 2.2855346601418834e-05, + 'epoch': 3.41} +04/19 [16:14:38] INFO | >> train_qwenlatent.py:487 + Step 13530 | grad_norm_pre_clip=0.2398 | + grad_norm_pre_clip_avg=0.2057 | Metrics: + {'align_loss': 0.024694260209798813, + 'recon_loss': 0.054581038653850555, + 'predict_loss': 0.014369315467774868, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23980270326137543, + 'data_time': 0.0009536549914628267, + 'model_time': 1.240880522003863, + 'grad_norm_pre_clip_avg': 0.20574474185705185, + 'learning_rate': 2.2850458995831177e-05, + 'epoch': 3.41} +04/19 [16:14:51] INFO | >> train_qwenlatent.py:487 + Step 13540 | grad_norm_pre_clip=0.1861 | + grad_norm_pre_clip_avg=0.2199 | Metrics: + {'align_loss': 0.023264439776539803, + 'recon_loss': 0.051642630249261856, + 'predict_loss': 0.011649091728031635, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18608875572681427, + 'data_time': 0.000914667994948104, + 'model_time': 1.2065315390063915, + 'grad_norm_pre_clip_avg': 0.2198607251048088, + 'learning_rate': 2.284556635164789e-05, + 'epoch': 3.42} +04/19 [16:15:04] INFO | >> train_qwenlatent.py:487 + Step 13550 | grad_norm_pre_clip=0.2946 | + grad_norm_pre_clip_avg=0.2557 | Metrics: + {'align_loss': 0.025752058252692223, + 'recon_loss': 0.07030600309371948, + 'predict_loss': 0.018697421997785568, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.294626384973526, + 'mae_score': 0.01902899527334952, 'data_time': + 0.0012598930043168366, 'model_time': + 1.2301161449868232, 'grad_norm_pre_clip_avg': + 0.2557292565703392, 'learning_rate': + 2.284066867125359e-05, 'epoch': 3.42} +04/19 [16:15:17] INFO | >> train_qwenlatent.py:487 + Step 13560 | grad_norm_pre_clip=0.2512 | + grad_norm_pre_clip_avg=0.2581 | Metrics: + {'align_loss': 0.024684233590960503, + 'recon_loss': 0.04865942895412445, + 'predict_loss': 0.013159147463738918, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25115230679512024, + 'data_time': 0.0006642840162385255, + 'model_time': 1.4515197190048639, + 'grad_norm_pre_clip_avg': 0.2581184357404709, + 'learning_rate': 2.283576595703535e-05, + 'epoch': 3.42} +04/19 [16:15:29] INFO | >> train_qwenlatent.py:487 + Step 13570 | grad_norm_pre_clip=0.2647 | + grad_norm_pre_clip_avg=0.2190 | Metrics: + {'align_loss': 0.02523694559931755, + 'recon_loss': 0.05532602593302727, + 'predict_loss': 0.015809185802936554, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26472917199134827, + 'data_time': 0.0009714890038594604, + 'model_time': 1.1842763079912402, + 'grad_norm_pre_clip_avg': 0.21903758496046066, + 'learning_rate': 2.2830858211382693e-05, + 'epoch': 3.42} +04/19 [16:15:41] INFO | >> train_qwenlatent.py:487 + Step 13580 | grad_norm_pre_clip=0.2467 | + grad_norm_pre_clip_avg=0.2493 | Metrics: + {'align_loss': 0.025222565978765488, + 'recon_loss': 0.07649904489517212, + 'predict_loss': 0.024295281618833542, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24667870998382568, + 'data_time': 0.0006923330074641854, + 'model_time': 1.180464528995799, + 'grad_norm_pre_clip_avg': 0.24929123669862746, + 'learning_rate': 2.282594543668759e-05, + 'epoch': 3.43} +04/19 [16:15:54] INFO | >> train_qwenlatent.py:487 + Step 13590 | grad_norm_pre_clip=0.2057 | + grad_norm_pre_clip_avg=0.2068 | Metrics: + {'align_loss': 0.02507348731160164, + 'recon_loss': 0.04963618144392967, + 'predict_loss': 0.011319291777908802, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20569908618927002, + 'data_time': 0.0006622719811275601, + 'model_time': 1.263332767994143, + 'grad_norm_pre_clip_avg': 0.20675048232078552, + 'learning_rate': 2.2821027635344473e-05, + 'epoch': 3.43} +04/19 [16:16:08] INFO | >> train_qwenlatent.py:487 + Step 13600 | grad_norm_pre_clip=0.2600 | + grad_norm_pre_clip_avg=0.2443 | Metrics: + {'align_loss': 0.025231553241610527, + 'recon_loss': 0.04925718531012535, + 'predict_loss': 0.010858004912734032, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2599932849407196, + 'mae_score': 0.015815054403769003, 'data_time': + 0.0007673820073250681, 'model_time': + 1.2192310609971173, 'grad_norm_pre_clip_avg': + 0.2442961871623993, 'learning_rate': + 2.281610480975021e-05, 'epoch': 3.43} +04/19 [16:16:21] INFO | >> train_qwenlatent.py:487 + Step 13610 | grad_norm_pre_clip=0.2383 | + grad_norm_pre_clip_avg=0.2371 | Metrics: + {'align_loss': 0.024920327588915825, + 'recon_loss': 0.052783213555812836, + 'predict_loss': 0.014468822628259659, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23831912875175476, + 'data_time': 0.0007139820081647485, + 'model_time': 1.308703771996079, + 'grad_norm_pre_clip_avg': 0.2370865300297737, + 'learning_rate': 2.2811176962304137e-05, + 'epoch': 3.43} +04/19 [16:16:33] INFO | >> train_qwenlatent.py:487 + Step 13620 | grad_norm_pre_clip=0.2054 | + grad_norm_pre_clip_avg=0.1927 | Metrics: + {'align_loss': 0.024536050856113434, + 'recon_loss': 0.049509212374687195, + 'predict_loss': 0.015937812626361847, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20541934669017792, + 'data_time': 0.0006789670151192695, + 'model_time': 1.2376471720053814, + 'grad_norm_pre_clip_avg': 0.19271438717842101, + 'learning_rate': 2.280624409540802e-05, + 'epoch': 3.44} +04/19 [16:16:46] INFO | >> train_qwenlatent.py:487 + Step 13630 | grad_norm_pre_clip=0.3343 | + grad_norm_pre_clip_avg=0.2825 | Metrics: + {'align_loss': 0.02425866201519966, + 'recon_loss': 0.04834979772567749, + 'predict_loss': 0.013690650463104248, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.33429309725761414, + 'data_time': 0.0009079160226974636, + 'model_time': 1.2872034949832596, + 'grad_norm_pre_clip_avg': 0.28247533589601515, + 'learning_rate': 2.2801306211466083e-05, + 'epoch': 3.44} +04/19 [16:16:59] INFO | >> train_qwenlatent.py:487 + Step 13640 | grad_norm_pre_clip=0.2706 | + grad_norm_pre_clip_avg=0.2427 | Metrics: + {'align_loss': 0.02583897113800049, + 'recon_loss': 0.057704679667949677, + 'predict_loss': 0.013434972614049911, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2705836296081543, + 'data_time': 0.0006644829991273582, + 'model_time': 1.1924849680217449, + 'grad_norm_pre_clip_avg': 0.2427131175994873, + 'learning_rate': 2.279636331288499e-05, + 'epoch': 3.44} +04/19 [16:17:12] INFO | >> train_qwenlatent.py:487 + Step 13650 | grad_norm_pre_clip=0.1864 | + grad_norm_pre_clip_avg=0.2024 | Metrics: + {'align_loss': 0.02432183176279068, + 'recon_loss': 0.04738199710845947, + 'predict_loss': 0.013456054031848907, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18636208772659302, + 'mae_score': 0.021460407703846423, 'data_time': + 0.0007317580166272819, 'model_time': + 1.5077546060201712, 'grad_norm_pre_clip_avg': + 0.20239701718091965, 'learning_rate': + 2.2791415402073846e-05, 'epoch': 3.44} +04/19 [16:17:25] INFO | >> train_qwenlatent.py:487 + Step 13660 | grad_norm_pre_clip=0.4124 | + grad_norm_pre_clip_avg=0.2362 | Metrics: + {'align_loss': 0.025348730385303497, + 'recon_loss': 0.04985520616173744, + 'predict_loss': 0.0171436108648777, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.41243016719818115, + 'data_time': 0.0006649339920841157, + 'model_time': 1.183684365998488, + 'grad_norm_pre_clip_avg': 0.2361756980419159, + 'learning_rate': 2.2786462481444205e-05, + 'epoch': 3.45} +04/19 [16:17:37] INFO | >> train_qwenlatent.py:487 + Step 13670 | grad_norm_pre_clip=0.2482 | + grad_norm_pre_clip_avg=0.2829 | Metrics: + {'align_loss': 0.024791469797492027, + 'recon_loss': 0.053979068994522095, + 'predict_loss': 0.014209246262907982, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24823123216629028, + 'data_time': 0.000677325006108731, + 'model_time': 1.1942235710157547, + 'grad_norm_pre_clip_avg': 0.2828606218099594, + 'learning_rate': 2.2781504553410056e-05, + 'epoch': 3.45} +04/19 [16:17:49] INFO | >> train_qwenlatent.py:487 + Step 13680 | grad_norm_pre_clip=0.2404 | + grad_norm_pre_clip_avg=0.2045 | Metrics: + {'align_loss': 0.024914322420954704, + 'recon_loss': 0.07014782726764679, + 'predict_loss': 0.019152067601680756, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24043019115924835, + 'data_time': 0.0009045650076586753, + 'model_time': 1.3120324969931971, + 'grad_norm_pre_clip_avg': 0.20446074903011321, + 'learning_rate': 2.277654162038784e-05, + 'epoch': 3.45} +04/19 [16:18:02] INFO | >> train_qwenlatent.py:487 + Step 13690 | grad_norm_pre_clip=0.2580 | + grad_norm_pre_clip_avg=0.2123 | Metrics: + {'align_loss': 0.025946110486984253, + 'recon_loss': 0.06828142702579498, + 'predict_loss': 0.015438701957464218, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2579931318759918, + 'data_time': 0.0006675049953628331, + 'model_time': 1.2294960340077523, + 'grad_norm_pre_clip_avg': 0.21226686537265776, + 'learning_rate': 2.277157368479643e-05, + 'epoch': 3.45} +04/19 [16:18:15] INFO | >> train_qwenlatent.py:487 + Step 13700 | grad_norm_pre_clip=0.2127 | + grad_norm_pre_clip_avg=0.2565 | Metrics: + {'align_loss': 0.025198303163051605, + 'recon_loss': 0.0636177808046341, + 'predict_loss': 0.016821565106511116, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21267829835414886, + 'mae_score': 0.017830584070704005, 'data_time': + 0.0009164040093310177, 'model_time': + 1.2725787009985652, 'grad_norm_pre_clip_avg': + 0.2565193846821785, 'learning_rate': + 2.276660074905713e-05, 'epoch': 3.46} +04/19 [16:18:28] INFO | >> train_qwenlatent.py:487 + Step 13710 | grad_norm_pre_clip=0.2261 | + grad_norm_pre_clip_avg=0.2226 | Metrics: + {'align_loss': 0.02582457661628723, + 'recon_loss': 0.06105208024382591, + 'predict_loss': 0.014322822913527489, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.226117342710495, + 'data_time': 0.0006721369863953441, + 'model_time': 1.2073469450115226, + 'grad_norm_pre_clip_avg': 0.22257837355136872, + 'learning_rate': 2.2761622815593696e-05, + 'epoch': 3.46} +04/19 [16:18:41] INFO | >> train_qwenlatent.py:487 + Step 13720 | grad_norm_pre_clip=0.2474 | + grad_norm_pre_clip_avg=0.2397 | Metrics: + {'align_loss': 0.025443974882364273, + 'recon_loss': 0.05366213619709015, + 'predict_loss': 0.014319178648293018, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2473994344472885, + 'data_time': 0.0009150609839707613, + 'model_time': 1.2100915200135205, + 'grad_norm_pre_clip_avg': 0.2397328183054924, + 'learning_rate': 2.275663988683231e-05, + 'epoch': 3.46} +04/19 [16:18:53] INFO | >> train_qwenlatent.py:487 + Step 13730 | grad_norm_pre_clip=0.1669 | + grad_norm_pre_clip_avg=0.2291 | Metrics: + {'align_loss': 0.024517273530364037, + 'recon_loss': 0.05834312364459038, + 'predict_loss': 0.011892570182681084, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16694387793540955, + 'data_time': 0.000626282999292016, + 'model_time': 1.2256136140204035, + 'grad_norm_pre_clip_avg': 0.22911922186613082, + 'learning_rate': 2.2751651965201586e-05, + 'epoch': 3.46} +04/19 [16:19:06] INFO | >> train_qwenlatent.py:487 + Step 13740 | grad_norm_pre_clip=0.2085 | + grad_norm_pre_clip_avg=0.2035 | Metrics: + {'align_loss': 0.025215284898877144, + 'recon_loss': 0.0489143468439579, + 'predict_loss': 0.012905239127576351, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2085178941488266, + 'data_time': 0.0011275350116193295, + 'model_time': 1.5338609510217793, + 'grad_norm_pre_clip_avg': 0.2034670040011406, + 'learning_rate': 2.2746659053132586e-05, + 'epoch': 3.47} +04/19 [16:19:19] INFO | >> train_qwenlatent.py:487 + Step 13750 | grad_norm_pre_clip=0.2488 | + grad_norm_pre_clip_avg=0.2851 | Metrics: + {'align_loss': 0.023675259202718735, + 'recon_loss': 0.04340087249875069, + 'predict_loss': 0.011336117051541805, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24881605803966522, + 'mae_score': 0.012834285186217712, 'data_time': + 0.00100639698212035, 'model_time': + 1.2054098539811093, 'grad_norm_pre_clip_avg': + 0.2851384460926056, 'learning_rate': + 2.274166115305879e-05, 'epoch': 3.47} +04/19 [16:19:31] INFO | >> train_qwenlatent.py:487 + Step 13760 | grad_norm_pre_clip=0.1933 | + grad_norm_pre_clip_avg=0.2056 | Metrics: + {'align_loss': 0.02424081414937973, + 'recon_loss': 0.05293109267950058, + 'predict_loss': 0.014065589755773544, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19333472847938538, + 'data_time': 0.0007656629895791411, + 'model_time': 1.2155338689917699, + 'grad_norm_pre_clip_avg': 0.20558878928422927, + 'learning_rate': 2.2736658267416105e-05, + 'epoch': 3.47} +04/19 [16:19:44] INFO | >> train_qwenlatent.py:487 + Step 13770 | grad_norm_pre_clip=0.2529 | + grad_norm_pre_clip_avg=0.1960 | Metrics: + {'align_loss': 0.022485170513391495, + 'recon_loss': 0.052970342338085175, + 'predict_loss': 0.016181593760848045, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25289422273635864, + 'data_time': 0.0010608019947540015, + 'model_time': 1.2164365809876472, + 'grad_norm_pre_clip_avg': 0.19597282707691194, + 'learning_rate': 2.2731650398642888e-05, + 'epoch': 3.47} +04/19 [16:19:56] INFO | >> train_qwenlatent.py:487 + Step 13780 | grad_norm_pre_clip=0.2277 | + grad_norm_pre_clip_avg=0.2569 | Metrics: + {'align_loss': 0.02495681308209896, + 'recon_loss': 0.05499812588095665, + 'predict_loss': 0.00973175372928381, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22769714891910553, + 'data_time': 0.0008845719858072698, + 'model_time': 1.239555191976251, + 'grad_norm_pre_clip_avg': 0.2568782359361649, + 'learning_rate': 2.2726637549179915e-05, + 'epoch': 3.48} +04/19 [16:20:10] INFO | >> train_qwenlatent.py:487 + Step 13790 | grad_norm_pre_clip=0.1928 | + grad_norm_pre_clip_avg=0.2103 | Metrics: + {'align_loss': 0.02481153979897499, + 'recon_loss': 0.04273811727762222, + 'predict_loss': 0.010407443158328533, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19278675317764282, + 'data_time': 0.0007977700151968747, + 'model_time': 1.5514324370014947, + 'grad_norm_pre_clip_avg': 0.2103102833032608, + 'learning_rate': 2.2721619721470384e-05, + 'epoch': 3.48} +04/19 [16:20:23] INFO | >> train_qwenlatent.py:487 + Step 13800 | grad_norm_pre_clip=0.2390 | + grad_norm_pre_clip_avg=0.2712 | Metrics: + {'align_loss': 0.025286050513386726, + 'recon_loss': 0.062002576887607574, + 'predict_loss': 0.016618698835372925, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23895575106143951, + 'mae_score': 0.021170136735245988, 'data_time': + 0.0006599309854209423, 'model_time': + 1.2896078230114654, 'grad_norm_pre_clip_avg': + 0.27120842039585114, 'learning_rate': + 2.2716596917959916e-05, 'epoch': 3.48} +04/19 [16:20:35] INFO | >> train_qwenlatent.py:487 + Step 13810 | grad_norm_pre_clip=0.3036 | + grad_norm_pre_clip_avg=0.2295 | Metrics: + {'align_loss': 0.024900319054722786, + 'recon_loss': 0.055489230901002884, + 'predict_loss': 0.015687022358179092, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3036162257194519, + 'data_time': 0.0006996399897616357, + 'model_time': 1.250337106990628, + 'grad_norm_pre_clip_avg': 0.22946318089962006, + 'learning_rate': 2.271156914109658e-05, + 'epoch': 3.48} +04/19 [16:20:48] INFO | >> train_qwenlatent.py:487 + Step 13820 | grad_norm_pre_clip=0.2336 | + grad_norm_pre_clip_avg=0.2617 | Metrics: + {'align_loss': 0.02314399927854538, + 'recon_loss': 0.03840818628668785, + 'predict_loss': 0.009492269717156887, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23357805609703064, + 'data_time': 0.0006897369748912752, + 'model_time': 1.248582708009053, + 'grad_norm_pre_clip_avg': 0.2617137864232063, + 'learning_rate': 2.2706536393330837e-05, + 'epoch': 3.49} +04/19 [16:21:00] INFO | >> train_qwenlatent.py:487 + Step 13830 | grad_norm_pre_clip=0.2638 | + grad_norm_pre_clip_avg=0.2656 | Metrics: + {'align_loss': 0.023958874866366386, + 'recon_loss': 0.03667057678103447, + 'predict_loss': 0.01025486458092928, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2637805938720703, + 'data_time': 0.0006857870030216873, + 'model_time': 1.2204976919747423, + 'grad_norm_pre_clip_avg': 0.26562726497650146, + 'learning_rate': 2.2701498677115594e-05, + 'epoch': 3.49} +04/19 [16:21:13] INFO | >> train_qwenlatent.py:487 + Step 13840 | grad_norm_pre_clip=0.1543 | + grad_norm_pre_clip_avg=0.1909 | Metrics: + {'align_loss': 0.02424473688006401, + 'recon_loss': 0.04790593683719635, + 'predict_loss': 0.010742045938968658, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15426617860794067, + 'data_time': 0.0007748570060357451, + 'model_time': 1.2408318109810352, + 'grad_norm_pre_clip_avg': 0.19091782122850418, + 'learning_rate': 2.269645599490618e-05, + 'epoch': 3.49} +04/19 [16:21:26] INFO | >> train_qwenlatent.py:487 + Step 13850 | grad_norm_pre_clip=0.2189 | + grad_norm_pre_clip_avg=0.2198 | Metrics: + {'align_loss': 0.022942975163459778, + 'recon_loss': 0.04932832717895508, + 'predict_loss': 0.014654657803475857, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21885168552398682, + 'mae_score': 0.017651035549404386, 'data_time': + 0.0009221580112352967, 'model_time': + 1.2443180450063664, 'grad_norm_pre_clip_avg': + 0.2198425680398941, 'learning_rate': + 2.269140834916032e-05, 'epoch': 3.49} +04/19 [16:21:39] INFO | >> train_qwenlatent.py:487 + Step 13860 | grad_norm_pre_clip=0.2581 | + grad_norm_pre_clip_avg=0.2436 | Metrics: + {'align_loss': 0.02476433850824833, + 'recon_loss': 0.04200475290417671, + 'predict_loss': 0.00977466069161892, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2581155598163605, + 'data_time': 0.0009515830024611205, + 'model_time': 1.2689812330063432, + 'grad_norm_pre_clip_avg': 0.24362584799528123, + 'learning_rate': 2.2686355742338195e-05, + 'epoch': 3.5} +04/19 [16:21:51] INFO | >> train_qwenlatent.py:487 + Step 13870 | grad_norm_pre_clip=0.2190 | + grad_norm_pre_clip_avg=0.2270 | Metrics: + {'align_loss': 0.024994444102048874, + 'recon_loss': 0.04056158661842346, + 'predict_loss': 0.010037167929112911, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21896405518054962, + 'data_time': 0.0007023640209808946, + 'model_time': 1.236158517014701, + 'grad_norm_pre_clip_avg': 0.2269827604293823, + 'learning_rate': 2.2681298176902367e-05, + 'epoch': 3.5} +04/19 [16:22:04] INFO | >> train_qwenlatent.py:487 + Step 13880 | grad_norm_pre_clip=0.3062 | + grad_norm_pre_clip_avg=0.2127 | Metrics: + {'align_loss': 0.023370910435914993, + 'recon_loss': 0.03682611137628555, + 'predict_loss': 0.010677720420062542, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.30616772174835205, + 'data_time': 0.0007568279979750514, + 'model_time': 1.2352947090112139, + 'grad_norm_pre_clip_avg': 0.21266028434038162, + 'learning_rate': 2.2676235655317845e-05, + 'epoch': 3.5} +04/19 [16:22:17] INFO | >> train_qwenlatent.py:487 + Step 13890 | grad_norm_pre_clip=0.2415 | + grad_norm_pre_clip_avg=0.2958 | Metrics: + {'align_loss': 0.0235904511064291, + 'recon_loss': 0.05559229850769043, + 'predict_loss': 0.017469413578510284, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2415059208869934, + 'data_time': 0.0010820970055647194, + 'model_time': 1.2861320250085555, + 'grad_norm_pre_clip_avg': 0.29578389823436735, + 'learning_rate': 2.2671168180052028e-05, + 'epoch': 3.5} +04/19 [16:22:30] INFO | >> train_qwenlatent.py:487 + Step 13900 | grad_norm_pre_clip=0.1880 | + grad_norm_pre_clip_avg=0.2068 | Metrics: + {'align_loss': 0.025338856503367424, + 'recon_loss': 0.04347963631153107, + 'predict_loss': 0.015168171375989914, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1880151927471161, + 'mae_score': 0.014685391950177717, 'data_time': + 0.0008535280067007989, 'model_time': + 1.251541019009892, 'grad_norm_pre_clip_avg': + 0.20676959902048112, 'learning_rate': + 2.2666095753574757e-05, 'epoch': 3.51} +04/19 [16:22:43] INFO | >> train_qwenlatent.py:487 + Step 13910 | grad_norm_pre_clip=0.2311 | + grad_norm_pre_clip_avg=0.2001 | Metrics: + {'align_loss': 0.02580823376774788, + 'recon_loss': 0.043114688247442245, + 'predict_loss': 0.007724347990006208, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23108406364917755, + 'data_time': 0.0010752599919214845, + 'model_time': 1.2172391089843586, + 'grad_norm_pre_clip_avg': 0.20007885247468948, + 'learning_rate': 2.266101837835826e-05, + 'epoch': 3.51} +04/19 [16:22:56] INFO | >> train_qwenlatent.py:487 + Step 13920 | grad_norm_pre_clip=0.1981 | + grad_norm_pre_clip_avg=0.2763 | Metrics: + {'align_loss': 0.025068584829568863, + 'recon_loss': 0.06780339032411575, + 'predict_loss': 0.013084094040095806, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19809480011463165, + 'data_time': 0.0006988339882809669, + 'model_time': 1.5543123839888722, + 'grad_norm_pre_clip_avg': 0.27632441371679306, + 'learning_rate': 2.265593605687719e-05, + 'epoch': 3.51} +04/19 [16:23:09] INFO | >> train_qwenlatent.py:487 + Step 13930 | grad_norm_pre_clip=0.1714 | + grad_norm_pre_clip_avg=0.2682 | Metrics: + {'align_loss': 0.024171654134988785, + 'recon_loss': 0.05963525176048279, + 'predict_loss': 0.013069387525320053, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17141197621822357, + 'data_time': 0.0009174699953291565, + 'model_time': 1.2694359209854156, + 'grad_norm_pre_clip_avg': 0.2682269707322121, + 'learning_rate': 2.2650848791608618e-05, + 'epoch': 3.52} +04/19 [16:23:21] INFO | >> train_qwenlatent.py:487 + Step 13940 | grad_norm_pre_clip=0.2182 | + grad_norm_pre_clip_avg=0.2261 | Metrics: + {'align_loss': 0.02524494007229805, + 'recon_loss': 0.059692781418561935, + 'predict_loss': 0.01616288535296917, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21817339956760406, + 'data_time': 0.0009006120089907199, + 'model_time': 1.2507180029933807, + 'grad_norm_pre_clip_avg': 0.22608662098646165, + 'learning_rate': 2.2645756585031998e-05, + 'epoch': 3.52} +04/19 [16:23:34] INFO | >> train_qwenlatent.py:487 + Step 13950 | grad_norm_pre_clip=0.2499 | + grad_norm_pre_clip_avg=0.2412 | Metrics: + {'align_loss': 0.02480623684823513, + 'recon_loss': 0.05302898958325386, + 'predict_loss': 0.013025665655732155, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2498597502708435, + 'mae_score': 0.014582991385245108, 'data_time': + 0.0009567730012349784, 'model_time': + 1.208175609994214, 'grad_norm_pre_clip_avg': + 0.24119579046964645, 'learning_rate': + 2.2640659439629215e-05, 'epoch': 3.52} +04/19 [16:23:47] INFO | >> train_qwenlatent.py:487 + Step 13960 | grad_norm_pre_clip=0.2004 | + grad_norm_pre_clip_avg=0.2036 | Metrics: + {'align_loss': 0.025780843570828438, + 'recon_loss': 0.06508463621139526, + 'predict_loss': 0.016383564099669456, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20038771629333496, + 'data_time': 0.0010082159715238959, + 'model_time': 1.208903831022326, + 'grad_norm_pre_clip_avg': 0.20359393805265427, + 'learning_rate': 2.263555735788457e-05, + 'epoch': 3.52} +04/19 [16:24:00] INFO | >> train_qwenlatent.py:487 + Step 13970 | grad_norm_pre_clip=0.1772 | + grad_norm_pre_clip_avg=0.2277 | Metrics: + {'align_loss': 0.02536504715681076, + 'recon_loss': 0.04891189932823181, + 'predict_loss': 0.010545616038143635, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17719024419784546, + 'data_time': 0.0006558689929079264, + 'model_time': 1.224683467997238, + 'grad_norm_pre_clip_avg': 0.22767734676599502, + 'learning_rate': 2.2630450342284736e-05, + 'epoch': 3.53} +04/19 [16:24:12] INFO | >> train_qwenlatent.py:487 + Step 13980 | grad_norm_pre_clip=0.2273 | + grad_norm_pre_clip_avg=0.2088 | Metrics: + {'align_loss': 0.02375916950404644, + 'recon_loss': 0.04056163504719734, + 'predict_loss': 0.011216272599995136, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22725996375083923, + 'data_time': 0.0006943799962755293, + 'model_time': 1.2034381109988317, + 'grad_norm_pre_clip_avg': 0.20881491154432297, + 'learning_rate': 2.2625338395318824e-05, + 'epoch': 3.53} +04/19 [16:24:25] INFO | >> train_qwenlatent.py:487 + Step 13990 | grad_norm_pre_clip=0.2315 | + grad_norm_pre_clip_avg=0.2755 | Metrics: + {'align_loss': 0.02476203814148903, + 'recon_loss': 0.0479261539876461, + 'predict_loss': 0.009263413958251476, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.231477290391922, + 'data_time': 0.0007333509856835008, + 'model_time': 1.2323906320089009, + 'grad_norm_pre_clip_avg': 0.275473390519619, + 'learning_rate': 2.262022151947833e-05, + 'epoch': 3.53} +04/19 [16:24:38] INFO | >> train_qwenlatent.py:487 + Step 14000 | grad_norm_pre_clip=0.2313 | + grad_norm_pre_clip_avg=0.2292 | Metrics: + {'align_loss': 0.024922693148255348, + 'recon_loss': 0.04676389694213867, + 'predict_loss': 0.012362546287477016, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23132090270519257, + 'mae_score': 0.014326067228575011, 'data_time': + 0.000959168974077329, 'model_time': + 1.2230548850202467, 'grad_norm_pre_clip_avg': + 0.22924381643533706, 'learning_rate': + 2.2615099717257156e-05, 'epoch': 3.53} +04/19 [16:24:51] INFO | >> train_qwenlatent.py:487 + Step 14010 | grad_norm_pre_clip=0.1820 | + grad_norm_pre_clip_avg=0.2076 | Metrics: + {'align_loss': 0.023848485201597214, + 'recon_loss': 0.038683585822582245, + 'predict_loss': 0.012426073662936687, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1820017546415329, + 'data_time': 0.000979116011876613, + 'model_time': 1.2334514610120095, + 'grad_norm_pre_clip_avg': 0.2076294854283333, + 'learning_rate': 2.260997299115161e-05, + 'epoch': 3.54} +04/19 [16:25:04] INFO | >> train_qwenlatent.py:487 + Step 14020 | grad_norm_pre_clip=0.2515 | + grad_norm_pre_clip_avg=0.2369 | Metrics: + {'align_loss': 0.026886455714702606, + 'recon_loss': 0.06743329018354416, + 'predict_loss': 0.015837715938687325, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2514688968658447, + 'data_time': 0.0006418929842766374, + 'model_time': 1.2152329749951605, + 'grad_norm_pre_clip_avg': 0.23691949248313904, + 'learning_rate': 2.2604841343660392e-05, + 'epoch': 3.54} +04/19 [16:25:16] INFO | >> train_qwenlatent.py:487 + Step 14030 | grad_norm_pre_clip=0.1517 | + grad_norm_pre_clip_avg=0.2418 | Metrics: + {'align_loss': 0.024442577734589577, + 'recon_loss': 0.05311312898993492, + 'predict_loss': 0.01295082550495863, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15173715353012085, + 'data_time': 0.0009512020042166114, + 'model_time': 1.2380157250154298, + 'grad_norm_pre_clip_avg': 0.24177847802639008, + 'learning_rate': 2.2599704777284606e-05, + 'epoch': 3.54} +04/19 [16:25:29] INFO | >> train_qwenlatent.py:487 + Step 14040 | grad_norm_pre_clip=0.2053 | + grad_norm_pre_clip_avg=0.2172 | Metrics: + {'align_loss': 0.02505369484424591, + 'recon_loss': 0.04747804254293442, + 'predict_loss': 0.009513968601822853, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20530425012111664, + 'data_time': 0.0006400499842129648, + 'model_time': 1.2150179730087984, + 'grad_norm_pre_clip_avg': 0.21715472638607025, + 'learning_rate': 2.2594563294527753e-05, + 'epoch': 3.54} +04/19 [16:25:41] INFO | >> train_qwenlatent.py:487 + Step 14050 | grad_norm_pre_clip=0.2049 | + grad_norm_pre_clip_avg=0.2068 | Metrics: + {'align_loss': 0.023707350715994835, + 'recon_loss': 0.037255238741636276, + 'predict_loss': 0.0099611422047019, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20492860674858093, + 'mae_score': 0.015534828804634714, 'data_time': + 0.0008230190142057836, 'model_time': + 1.2191674210189376, 'grad_norm_pre_clip_avg': + 0.2067532494664192, 'learning_rate': + 2.258941689789573e-05, 'epoch': 3.55} +04/19 [16:25:55] INFO | >> train_qwenlatent.py:487 + Step 14060 | grad_norm_pre_clip=0.2834 | + grad_norm_pre_clip_avg=0.2449 | Metrics: + {'align_loss': 0.02482488751411438, + 'recon_loss': 0.06820361316204071, + 'predict_loss': 0.014815191738307476, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28339967131614685, + 'data_time': 0.0009246370173059404, + 'model_time': 1.493050894991029, + 'grad_norm_pre_clip_avg': 0.24494965076446534, + 'learning_rate': 2.2584265589896825e-05, + 'epoch': 3.55} +04/19 [16:26:07] INFO | >> train_qwenlatent.py:487 + Step 14070 | grad_norm_pre_clip=0.2007 | + grad_norm_pre_clip_avg=0.2314 | Metrics: + {'align_loss': 0.02358812466263771, + 'recon_loss': 0.044999655336141586, + 'predict_loss': 0.008482393808662891, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20072250068187714, + 'data_time': 0.001247661013621837, + 'model_time': 1.6052072109887376, + 'grad_norm_pre_clip_avg': 0.2314450427889824, + 'learning_rate': 2.2579109373041723e-05, + 'epoch': 3.55} +04/19 [16:26:20] INFO | >> train_qwenlatent.py:487 + Step 14080 | grad_norm_pre_clip=0.2458 | + grad_norm_pre_clip_avg=0.2153 | Metrics: + {'align_loss': 0.025752127170562744, + 'recon_loss': 0.04714110493659973, + 'predict_loss': 0.010026399977505207, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.245778888463974, + 'data_time': 0.0010290979989804327, + 'model_time': 1.25502762402175, + 'grad_norm_pre_clip_avg': 0.21534141451120375, + 'learning_rate': 2.2573948249843503e-05, + 'epoch': 3.55} +04/19 [16:26:33] INFO | >> train_qwenlatent.py:487 + Step 14090 | grad_norm_pre_clip=0.2307 | + grad_norm_pre_clip_avg=0.2229 | Metrics: + {'align_loss': 0.02547951228916645, + 'recon_loss': 0.05956831946969032, + 'predict_loss': 0.016707293689250946, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2307499498128891, + 'data_time': 0.0007109050056897104, + 'model_time': 1.2214855719939806, + 'grad_norm_pre_clip_avg': 0.2228650316596031, + 'learning_rate': 2.2568782222817635e-05, + 'epoch': 3.56} +04/19 [16:26:46] INFO | >> train_qwenlatent.py:487 + Step 14100 | grad_norm_pre_clip=0.1986 | + grad_norm_pre_clip_avg=0.2250 | Metrics: + {'align_loss': 0.02547512575984001, + 'recon_loss': 0.06042362377047539, + 'predict_loss': 0.011416934430599213, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19861268997192383, + 'mae_score': 0.014427836306460268, 'data_time': + 0.0009465350012760609, 'model_time': + 1.220245372998761, 'grad_norm_pre_clip_avg': + 0.22497088462114334, 'learning_rate': + 2.256361129448198e-05, 'epoch': 3.56} +04/19 [16:26:58] INFO | >> train_qwenlatent.py:487 + Step 14110 | grad_norm_pre_clip=0.2713 | + grad_norm_pre_clip_avg=0.2783 | Metrics: + {'align_loss': 0.024266190826892853, + 'recon_loss': 0.0523657388985157, + 'predict_loss': 0.013058794662356377, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.27127698063850403, + 'data_time': 0.0007852459966670722, + 'model_time': 1.2576681939826813, + 'grad_norm_pre_clip_avg': 0.2782939404249191, + 'learning_rate': 2.2558435467356778e-05, + 'epoch': 3.56} +04/19 [16:27:11] INFO | >> train_qwenlatent.py:487 + Step 14120 | grad_norm_pre_clip=0.2110 | + grad_norm_pre_clip_avg=0.2682 | Metrics: + {'align_loss': 0.024344373494386673, + 'recon_loss': 0.05759425088763237, + 'predict_loss': 0.013586574234068394, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2110043317079544, + 'data_time': 0.0009046070044860244, + 'model_time': 1.2069110199809074, + 'grad_norm_pre_clip_avg': 0.2682367831468582, + 'learning_rate': 2.2553254743964668e-05, + 'epoch': 3.56} +04/19 [16:27:23] INFO | >> train_qwenlatent.py:487 + Step 14130 | grad_norm_pre_clip=0.1643 | + grad_norm_pre_clip_avg=0.2022 | Metrics: + {'align_loss': 0.024388134479522705, + 'recon_loss': 0.05832722410559654, + 'predict_loss': 0.016380345448851585, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16432450711727142, + 'data_time': 0.0008364679815713316, + 'model_time': 1.2571689170144964, + 'grad_norm_pre_clip_avg': 0.20223017334938048, + 'learning_rate': 2.2548069126830675e-05, + 'epoch': 3.57} +04/19 [16:27:36] INFO | >> train_qwenlatent.py:487 + Step 14140 | grad_norm_pre_clip=0.1827 | + grad_norm_pre_clip_avg=0.1862 | Metrics: + {'align_loss': 0.023459311574697495, + 'recon_loss': 0.0566011406481266, + 'predict_loss': 0.012563249096274376, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1827498823404312, + 'data_time': 0.0009908849897328764, + 'model_time': 1.2751456380065065, + 'grad_norm_pre_clip_avg': 0.18620418310165404, + 'learning_rate': 2.2542878618482205e-05, + 'epoch': 3.57} +04/19 [16:27:49] INFO | >> train_qwenlatent.py:487 + Step 14150 | grad_norm_pre_clip=0.1894 | + grad_norm_pre_clip_avg=0.2216 | Metrics: + {'align_loss': 0.025612447410821915, + 'recon_loss': 0.050996873527765274, + 'predict_loss': 0.01055680587887764, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18936491012573242, + 'mae_score': 0.022743142617715373, 'data_time': + 0.00098330999026075, 'model_time': + 1.3224529970029835, 'grad_norm_pre_clip_avg': + 0.2216193273663521, 'learning_rate': + 2.2537683221449042e-05, 'epoch': 3.57} +04/19 [16:28:02] INFO | >> train_qwenlatent.py:487 + Step 14160 | grad_norm_pre_clip=0.2455 | + grad_norm_pre_clip_avg=0.2449 | Metrics: + {'align_loss': 0.024895373731851578, + 'recon_loss': 0.047040339559316635, + 'predict_loss': 0.012568168342113495, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24546615779399872, + 'data_time': 0.0006670079892501235, + 'model_time': 1.2308115799969528, + 'grad_norm_pre_clip_avg': 0.2449446514248848, + 'learning_rate': 2.2532482938263372e-05, + 'epoch': 3.57} +04/19 [16:28:14] INFO | >> train_qwenlatent.py:487 + Step 14170 | grad_norm_pre_clip=0.2327 | + grad_norm_pre_clip_avg=0.2571 | Metrics: + {'align_loss': 0.024403519928455353, + 'recon_loss': 0.05466194078326225, + 'predict_loss': 0.014611786231398582, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23269237577915192, + 'data_time': 0.000995683018118143, + 'model_time': 1.222420484002214, + 'grad_norm_pre_clip_avg': 0.2571419432759285, + 'learning_rate': 2.2527277771459733e-05, + 'epoch': 3.58} +04/19 [16:28:27] INFO | >> train_qwenlatent.py:487 + Step 14180 | grad_norm_pre_clip=0.2193 | + grad_norm_pre_clip_avg=0.2342 | Metrics: + {'align_loss': 0.024409865960478783, + 'recon_loss': 0.05516723170876503, + 'predict_loss': 0.01395349483937025, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21930845081806183, + 'data_time': 0.0008590609941165894, + 'model_time': 1.2100516709906515, + 'grad_norm_pre_clip_avg': 0.2342369958758354, + 'learning_rate': 2.2522067723575075e-05, + 'epoch': 3.58} +04/19 [16:28:40] INFO | >> train_qwenlatent.py:487 + Step 14190 | grad_norm_pre_clip=0.1856 | + grad_norm_pre_clip_avg=0.2433 | Metrics: + {'align_loss': 0.024752667173743248, + 'recon_loss': 0.05207888036966324, + 'predict_loss': 0.010601717978715897, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1855771839618683, + 'data_time': 0.0007367999933194369, + 'model_time': 1.2407749929989222, + 'grad_norm_pre_clip_avg': 0.2433442384004593, + 'learning_rate': 2.2516852797148704e-05, + 'epoch': 3.58} +04/19 [16:28:54] INFO | >> train_qwenlatent.py:487 + Step 14200 | grad_norm_pre_clip=0.2104 | + grad_norm_pre_clip_avg=0.2258 | Metrics: + {'align_loss': 0.023616962134838104, + 'recon_loss': 0.04334701597690582, + 'predict_loss': 0.012331949546933174, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2103695124387741, + 'mae_score': 0.019012047363831115, 'data_time': + 0.0008223070180974901, 'model_time': + 1.2297775410115719, 'grad_norm_pre_clip_avg': + 0.22575180232524872, 'learning_rate': + 2.251163299472231e-05, 'epoch': 3.58} +04/19 [16:29:06] INFO | >> train_qwenlatent.py:487 + Step 14210 | grad_norm_pre_clip=0.2312 | + grad_norm_pre_clip_avg=0.2020 | Metrics: + {'align_loss': 0.02302047424018383, + 'recon_loss': 0.05158397927880287, + 'predict_loss': 0.014350196346640587, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23120234906673431, + 'data_time': 0.0006968009984120727, + 'model_time': 1.2477340860059485, + 'grad_norm_pre_clip_avg': 0.20203691571950913, + 'learning_rate': 2.250640831883997e-05, + 'epoch': 3.59} +04/19 [16:29:19] INFO | >> train_qwenlatent.py:487 + Step 14220 | grad_norm_pre_clip=0.1901 | + grad_norm_pre_clip_avg=0.2223 | Metrics: + {'align_loss': 0.026145853102207184, + 'recon_loss': 0.07530324906110764, + 'predict_loss': 0.013991914689540863, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19012512266635895, + 'data_time': 0.000810943020042032, + 'model_time': 1.283936777006602, + 'grad_norm_pre_clip_avg': 0.2223176032304764, + 'learning_rate': 2.250117877204812e-05, + 'epoch': 3.59} +04/19 [16:29:32] INFO | >> train_qwenlatent.py:487 + Step 14230 | grad_norm_pre_clip=0.2557 | + grad_norm_pre_clip_avg=0.2859 | Metrics: + {'align_loss': 0.025818467140197754, + 'recon_loss': 0.06254731118679047, + 'predict_loss': 0.014002948068082333, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25567275285720825, + 'data_time': 0.0010764779872260988, + 'model_time': 1.348331514018355, + 'grad_norm_pre_clip_avg': 0.28594078570604325, + 'learning_rate': 2.2495944356895584e-05, + 'epoch': 3.59} +04/19 [16:29:44] INFO | >> train_qwenlatent.py:487 + Step 14240 | grad_norm_pre_clip=0.2341 | + grad_norm_pre_clip_avg=0.2269 | Metrics: + {'align_loss': 0.02612490952014923, + 'recon_loss': 0.05338849872350693, + 'predict_loss': 0.011206082068383694, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23414082825183868, + 'data_time': 0.0006403019942808896, + 'model_time': 1.2138745190168265, + 'grad_norm_pre_clip_avg': 0.2269270345568657, + 'learning_rate': 2.2490705075933548e-05, + 'epoch': 3.59} +04/19 [16:29:57] INFO | >> train_qwenlatent.py:487 + Step 14250 | grad_norm_pre_clip=0.2061 | + grad_norm_pre_clip_avg=0.2275 | Metrics: + {'align_loss': 0.02639654651284218, + 'recon_loss': 0.06145656108856201, + 'predict_loss': 0.010999622754752636, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2061164528131485, + 'mae_score': 0.01970821243148666, 'data_time': + 0.0008645300113130361, 'model_time': + 1.2193597069999669, 'grad_norm_pre_clip_avg': + 0.22750135213136674, 'learning_rate': + 2.248546093171557e-05, 'epoch': 3.6} +04/19 [16:30:10] INFO | >> train_qwenlatent.py:487 + Step 14260 | grad_norm_pre_clip=0.1945 | + grad_norm_pre_clip_avg=0.2332 | Metrics: + {'align_loss': 0.025644704699516296, + 'recon_loss': 0.05428178980946541, + 'predict_loss': 0.01071466226130724, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1945357322692871, + 'data_time': 0.000698314979672432, + 'model_time': 1.2489820840128232, + 'grad_norm_pre_clip_avg': 0.2331574410200119, + 'learning_rate': 2.2480211926797588e-05, + 'epoch': 3.6} +04/19 [16:30:22] INFO | >> train_qwenlatent.py:487 + Step 14270 | grad_norm_pre_clip=0.1757 | + grad_norm_pre_clip_avg=0.2160 | Metrics: + {'align_loss': 0.02414150908589363, + 'recon_loss': 0.05355994030833244, + 'predict_loss': 0.013958362862467766, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17574526369571686, + 'data_time': 0.000674631999572739, + 'model_time': 1.2503080260066781, + 'grad_norm_pre_clip_avg': 0.21602374613285064, + 'learning_rate': 2.2474958063737904e-05, + 'epoch': 3.6} +04/19 [16:30:35] INFO | >> train_qwenlatent.py:487 + Step 14280 | grad_norm_pre_clip=0.2176 | + grad_norm_pre_clip_avg=0.2035 | Metrics: + {'align_loss': 0.02485177107155323, + 'recon_loss': 0.05390312895178795, + 'predict_loss': 0.011982777155935764, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21764203906059265, + 'data_time': 0.0008895309874787927, + 'model_time': 1.2198443179950118, + 'grad_norm_pre_clip_avg': 0.20346822440624238, + 'learning_rate': 2.2469699345097188e-05, + 'epoch': 3.6} +04/19 [16:30:47] INFO | >> train_qwenlatent.py:487 + Step 14290 | grad_norm_pre_clip=0.1868 | + grad_norm_pre_clip_avg=0.1971 | Metrics: + {'align_loss': 0.026506956666707993, + 'recon_loss': 0.06482484191656113, + 'predict_loss': 0.01603490486741066, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18681058287620544, + 'data_time': 0.0007877290190663189, + 'model_time': 1.2305695910181385, + 'grad_norm_pre_clip_avg': 0.19712119847536086, + 'learning_rate': 2.246443577343847e-05, + 'epoch': 3.61} +04/19 [16:31:01] INFO | >> train_qwenlatent.py:487 + Step 14300 | grad_norm_pre_clip=0.3811 | + grad_norm_pre_clip_avg=0.3420 | Metrics: + {'align_loss': 0.025415660813450813, + 'recon_loss': 0.053543251007795334, + 'predict_loss': 0.015589764341711998, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.38112345337867737, + 'mae_score': 0.016768182290566935, 'data_time': + 0.001147000992204994, 'model_time': + 1.3719920610019471, 'grad_norm_pre_clip_avg': + 0.3419617906212807, 'learning_rate': + 2.2459167351327152e-05, 'epoch': 3.61} +04/19 [16:31:13] INFO | >> train_qwenlatent.py:487 + Step 14310 | grad_norm_pre_clip=0.1989 | + grad_norm_pre_clip_avg=0.2583 | Metrics: + {'align_loss': 0.02653321996331215, + 'recon_loss': 0.06935152411460876, + 'predict_loss': 0.012732289731502533, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1989104300737381, + 'data_time': 0.0008835540211293846, + 'model_time': 1.2353400540014263, + 'grad_norm_pre_clip_avg': 0.2582754850387573, + 'learning_rate': 2.2453894081331e-05, 'epoch': + 3.61} +04/19 [16:31:26] INFO | >> train_qwenlatent.py:487 + Step 14320 | grad_norm_pre_clip=0.1779 | + grad_norm_pre_clip_avg=0.2183 | Metrics: + {'align_loss': 0.02449045144021511, + 'recon_loss': 0.05843258649110794, + 'predict_loss': 0.018613867461681366, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17793242633342743, + 'data_time': 0.0007519200153183192, + 'model_time': 1.328816191002261, + 'grad_norm_pre_clip_avg': 0.21829449534416198, + 'learning_rate': 2.2448615966020146e-05, + 'epoch': 3.61} +04/19 [16:31:40] INFO | >> train_qwenlatent.py:487 + Step 14330 | grad_norm_pre_clip=0.2209 | + grad_norm_pre_clip_avg=0.1914 | Metrics: + {'align_loss': 0.02580508217215538, + 'recon_loss': 0.067790687084198, + 'predict_loss': 0.020596520975232124, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22094377875328064, + 'data_time': 0.0010644340072758496, + 'model_time': 1.2778204799978994, + 'grad_norm_pre_clip_avg': 0.19138773083686828, + 'learning_rate': 2.2443333007967073e-05, + 'epoch': 3.62} +04/19 [16:31:52] INFO | >> train_qwenlatent.py:487 + Step 14340 | grad_norm_pre_clip=0.2247 | + grad_norm_pre_clip_avg=0.2238 | Metrics: + {'align_loss': 0.025198888033628464, + 'recon_loss': 0.05692412704229355, + 'predict_loss': 0.01797334849834442, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22471529245376587, + 'data_time': 0.0007845920044928789, + 'model_time': 1.2154231749882456, + 'grad_norm_pre_clip_avg': 0.22377490401268005, + 'learning_rate': 2.2438045209746636e-05, + 'epoch': 3.62} +04/19 [16:32:05] INFO | >> train_qwenlatent.py:487 + Step 14350 | grad_norm_pre_clip=0.1987 | + grad_norm_pre_clip_avg=0.2275 | Metrics: + {'align_loss': 0.023699050769209862, + 'recon_loss': 0.06223270297050476, + 'predict_loss': 0.01838230900466442, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.198658287525177, + 'mae_score': 0.015872240496111346, 'data_time': + 0.0009786860027816147, 'model_time': + 1.2531809299835004, 'grad_norm_pre_clip_avg': + 0.22751356065273284, 'learning_rate': + 2.2432752573936035e-05, 'epoch': 3.62} +04/19 [16:32:18] INFO | >> train_qwenlatent.py:487 + Step 14360 | grad_norm_pre_clip=0.2503 | + grad_norm_pre_clip_avg=0.2007 | Metrics: + {'align_loss': 0.024575255811214447, + 'recon_loss': 0.05305473506450653, + 'predict_loss': 0.01540165115147829, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2502933740615845, + 'data_time': 0.0016338080167770386, + 'model_time': 1.2258411220100243, + 'grad_norm_pre_clip_avg': 0.2006657153367996, + 'learning_rate': 2.2427455103114845e-05, + 'epoch': 3.62} +04/19 [16:32:31] INFO | >> train_qwenlatent.py:487 + Step 14370 | grad_norm_pre_clip=0.1685 | + grad_norm_pre_clip_avg=0.2098 | Metrics: + {'align_loss': 0.02517550438642502, + 'recon_loss': 0.06208861991763115, + 'predict_loss': 0.01666232943534851, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16853764653205872, + 'data_time': 0.0009861029975581914, + 'model_time': 1.2793847170250956, + 'grad_norm_pre_clip_avg': 0.20980535596609115, + 'learning_rate': 2.2422152799864988e-05, + 'epoch': 3.63} +04/19 [16:32:44] INFO | >> train_qwenlatent.py:487 + Step 14380 | grad_norm_pre_clip=0.1875 | + grad_norm_pre_clip_avg=0.2383 | Metrics: + {'align_loss': 0.027087878435850143, + 'recon_loss': 0.07358570396900177, + 'predict_loss': 0.01690514013171196, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18745209276676178, + 'data_time': 0.000862392014823854, + 'model_time': 1.2268487810215447, + 'grad_norm_pre_clip_avg': 0.2382814332842827, + 'learning_rate': 2.2416845666770734e-05, + 'epoch': 3.63} +04/19 [16:32:56] INFO | >> train_qwenlatent.py:487 + Step 14390 | grad_norm_pre_clip=0.1823 | + grad_norm_pre_clip_avg=0.2546 | Metrics: + {'align_loss': 0.02388652041554451, + 'recon_loss': 0.03982967510819435, + 'predict_loss': 0.009738977067172527, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18227948248386383, + 'data_time': 0.0011052500049117953, + 'model_time': 1.3161977650015615, + 'grad_norm_pre_clip_avg': 0.2546251565217972, + 'learning_rate': 2.2411533706418726e-05, + 'epoch': 3.63} +04/19 [16:33:09] INFO | >> train_qwenlatent.py:487 + Step 14400 | grad_norm_pre_clip=0.2328 | + grad_norm_pre_clip_avg=0.2350 | Metrics: + {'align_loss': 0.02531380206346512, + 'recon_loss': 0.06485649198293686, + 'predict_loss': 0.019161375239491463, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2327895164489746, + 'mae_score': 0.019200472788767773, 'data_time': + 0.0006987449887674302, 'model_time': + 1.252147273014998, 'grad_norm_pre_clip_avg': + 0.23503419309854506, 'learning_rate': + 2.2406216921397947e-05, 'epoch': 3.63} +04/19 [16:33:22] INFO | >> train_qwenlatent.py:487 + Step 14410 | grad_norm_pre_clip=0.1992 | + grad_norm_pre_clip_avg=0.2057 | Metrics: + {'align_loss': 0.024355348199605942, + 'recon_loss': 0.05134817585349083, + 'predict_loss': 0.013112474232912064, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.199226513504982, + 'data_time': 0.0007412409759126604, + 'model_time': 1.2461221750127152, + 'grad_norm_pre_clip_avg': 0.20573171675205232, + 'learning_rate': 2.2400895314299724e-05, + 'epoch': 3.64} +04/19 [16:33:35] INFO | >> train_qwenlatent.py:487 + Step 14420 | grad_norm_pre_clip=0.2166 | + grad_norm_pre_clip_avg=0.2172 | Metrics: + {'align_loss': 0.02459113486111164, + 'recon_loss': 0.04824097082018852, + 'predict_loss': 0.01251100655645132, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21655499935150146, + 'data_time': 0.0009689159924164414, + 'model_time': 1.3027949779934715, + 'grad_norm_pre_clip_avg': 0.21717305928468705, + 'learning_rate': 2.2395568887717754e-05, + 'epoch': 3.64} +04/19 [16:33:47] INFO | >> train_qwenlatent.py:487 + Step 14430 | grad_norm_pre_clip=0.2159 | + grad_norm_pre_clip_avg=0.2024 | Metrics: + {'align_loss': 0.024298997595906258, + 'recon_loss': 0.054632142186164856, + 'predict_loss': 0.018347689881920815, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2158510535955429, + 'data_time': 0.0007578360091429204, + 'model_time': 1.2503136680170428, + 'grad_norm_pre_clip_avg': 0.20235061794519424, + 'learning_rate': 2.2390237644248068e-05, + 'epoch': 3.64} +04/19 [16:34:00] INFO | >> train_qwenlatent.py:487 + Step 14440 | grad_norm_pre_clip=0.2214 | + grad_norm_pre_clip_avg=0.2530 | Metrics: + {'align_loss': 0.025040848180651665, + 'recon_loss': 0.046035900712013245, + 'predict_loss': 0.012584966607391834, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2213743031024933, + 'data_time': 0.0011182280140928924, + 'model_time': 1.330473059992073, + 'grad_norm_pre_clip_avg': 0.25298162549734116, + 'learning_rate': 2.2384901586489054e-05, + 'epoch': 3.64} +04/19 [16:34:13] INFO | >> train_qwenlatent.py:487 + Step 14450 | grad_norm_pre_clip=0.2378 | + grad_norm_pre_clip_avg=0.2189 | Metrics: + {'align_loss': 0.024803001433610916, + 'recon_loss': 0.046686213463544846, + 'predict_loss': 0.009843757376074791, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23784638941287994, + 'mae_score': 0.013010504439070418, 'data_time': + 0.0008744999940972775, 'model_time': + 1.2177718890015967, 'grad_norm_pre_clip_avg': + 0.21890073418617248, 'learning_rate': + 2.2379560717041432e-05, 'epoch': 3.65} +04/19 [16:34:26] INFO | >> train_qwenlatent.py:487 + Step 14460 | grad_norm_pre_clip=0.1581 | + grad_norm_pre_clip_avg=0.1939 | Metrics: + {'align_loss': 0.02371191419661045, + 'recon_loss': 0.0394437313079834, + 'predict_loss': 0.011019090190529823, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1581314653158188, + 'data_time': 0.0006686619890388101, + 'model_time': 1.476805336016696, + 'grad_norm_pre_clip_avg': 0.1938563957810402, + 'learning_rate': 2.237421503850829e-05, + 'epoch': 3.65} +04/19 [16:34:39] INFO | >> train_qwenlatent.py:487 + Step 14470 | grad_norm_pre_clip=0.4213 | + grad_norm_pre_clip_avg=0.3858 | Metrics: + {'align_loss': 0.02384776622056961, + 'recon_loss': 0.06474517285823822, + 'predict_loss': 0.019049113616347313, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.4213339686393738, + 'data_time': 0.0006282480026129633, + 'model_time': 1.220799234986771, + 'grad_norm_pre_clip_avg': 0.3857567012310028, + 'learning_rate': 2.2368864553495035e-05, + 'epoch': 3.65} +04/19 [16:34:52] INFO | >> train_qwenlatent.py:487 + Step 14480 | grad_norm_pre_clip=0.4393 | + grad_norm_pre_clip_avg=0.2990 | Metrics: + {'align_loss': 0.02513587847352028, + 'recon_loss': 0.05068783834576607, + 'predict_loss': 0.013230988755822182, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.43929243087768555, + 'data_time': 0.0009144340001512319, + 'model_time': 1.231817741994746, + 'grad_norm_pre_clip_avg': 0.2990397498011589, + 'learning_rate': 2.2363509264609435e-05, + 'epoch': 3.65} +04/19 [16:35:04] INFO | >> train_qwenlatent.py:487 + Step 14490 | grad_norm_pre_clip=0.1764 | + grad_norm_pre_clip_avg=0.2289 | Metrics: + {'align_loss': 0.023018719628453255, + 'recon_loss': 0.044007230550050735, + 'predict_loss': 0.011755417101085186, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1764003038406372, + 'data_time': 0.0011062310077250004, + 'model_time': 1.2691934590111487, + 'grad_norm_pre_clip_avg': 0.22894255965948104, + 'learning_rate': 2.2358149174461587e-05, + 'epoch': 3.66} +04/19 [16:35:18] INFO | >> train_qwenlatent.py:487 + Step 14500 | grad_norm_pre_clip=0.1599 | + grad_norm_pre_clip_avg=0.2134 | Metrics: + {'align_loss': 0.024225927889347076, + 'recon_loss': 0.051101092249155045, + 'predict_loss': 0.015866506844758987, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15994037687778473, + 'mae_score': 0.0169243116636534, 'data_time': + 0.001117007021093741, 'model_time': + 1.5121619690035004, 'grad_norm_pre_clip_avg': + 0.2133821204304695, 'learning_rate': + 2.235278428566394e-05, 'epoch': 3.66} +04/19 [16:35:30] INFO | >> train_qwenlatent.py:487 + Step 14510 | grad_norm_pre_clip=0.2073 | + grad_norm_pre_clip_avg=0.1885 | Metrics: + {'align_loss': 0.025529183447360992, + 'recon_loss': 0.05101010575890541, + 'predict_loss': 0.014339547604322433, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20725128054618835, + 'data_time': 0.0012545219797175378, + 'model_time': 1.2385930730088148, + 'grad_norm_pre_clip_avg': 0.1884905591607094, + 'learning_rate': 2.2347414600831275e-05, + 'epoch': 3.66} +04/19 [16:35:43] INFO | >> train_qwenlatent.py:487 + Step 14520 | grad_norm_pre_clip=0.2145 | + grad_norm_pre_clip_avg=0.1967 | Metrics: + {'align_loss': 0.025798367336392403, + 'recon_loss': 0.0628696084022522, + 'predict_loss': 0.013780814595520496, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21451622247695923, + 'data_time': 0.001019285002257675, + 'model_time': 1.2173916299943812, + 'grad_norm_pre_clip_avg': 0.1966899424791336, + 'learning_rate': 2.234204012258071e-05, + 'epoch': 3.66} +04/19 [16:35:55] INFO | >> train_qwenlatent.py:487 + Step 14530 | grad_norm_pre_clip=0.1895 | + grad_norm_pre_clip_avg=0.2589 | Metrics: + {'align_loss': 0.024443339556455612, + 'recon_loss': 0.04009222984313965, + 'predict_loss': 0.00947977602481842, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18953031301498413, + 'data_time': 0.0006690520094707608, + 'model_time': 1.2370331120037008, + 'grad_norm_pre_clip_avg': 0.25894793272018435, + 'learning_rate': 2.2336660853531695e-05, + 'epoch': 3.67} +04/19 [16:36:08] INFO | >> train_qwenlatent.py:487 + Step 14540 | grad_norm_pre_clip=0.2348 | + grad_norm_pre_clip_avg=0.2504 | Metrics: + {'align_loss': 0.025182943791151047, + 'recon_loss': 0.05502353981137276, + 'predict_loss': 0.011477276682853699, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23484210669994354, + 'data_time': 0.0007630609907209873, + 'model_time': 1.4491157890006434, + 'grad_norm_pre_clip_avg': 0.2504494071006775, + 'learning_rate': 2.2331276796306026e-05, + 'epoch': 3.67} +04/19 [16:36:21] INFO | >> train_qwenlatent.py:487 + Step 14550 | grad_norm_pre_clip=0.1719 | + grad_norm_pre_clip_avg=0.2093 | Metrics: + {'align_loss': 0.025348111987113953, + 'recon_loss': 0.05152832716703415, + 'predict_loss': 0.009134095162153244, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1719067245721817, + 'mae_score': 0.016597151541495107, 'data_time': + 0.0011574880045372993, 'model_time': + 1.2664518709934782, 'grad_norm_pre_clip_avg': + 0.20928370654582978, 'learning_rate': + 2.2325887953527832e-05, 'epoch': 3.67} +04/19 [16:36:33] INFO | >> train_qwenlatent.py:487 + Step 14560 | grad_norm_pre_clip=0.1973 | + grad_norm_pre_clip_avg=0.1833 | Metrics: + {'align_loss': 0.025263313204050064, + 'recon_loss': 0.06664623320102692, + 'predict_loss': 0.019097497686743736, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19734062254428864, + 'data_time': 0.001018393988488242, + 'model_time': 1.221263423009077, + 'grad_norm_pre_clip_avg': 0.18330608457326888, + 'learning_rate': 2.232049432782356e-05, + 'epoch': 3.67} +04/19 [16:36:46] INFO | >> train_qwenlatent.py:487 + Step 14570 | grad_norm_pre_clip=0.2761 | + grad_norm_pre_clip_avg=0.2678 | Metrics: + {'align_loss': 0.025978293269872665, + 'recon_loss': 0.0550687275826931, + 'predict_loss': 0.01580912247300148, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2760801315307617, + 'data_time': 0.0009404599841218442, + 'model_time': 1.2126029609935358, + 'grad_norm_pre_clip_avg': 0.2677540808916092, + 'learning_rate': 2.2315095921822005e-05, + 'epoch': 3.68} +04/19 [16:36:58] INFO | >> train_qwenlatent.py:487 + Step 14580 | grad_norm_pre_clip=0.2265 | + grad_norm_pre_clip_avg=0.2568 | Metrics: + {'align_loss': 0.02715931087732315, + 'recon_loss': 0.056149691343307495, + 'predict_loss': 0.01001857127994299, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2265154868364334, + 'data_time': 0.0009878760029096156, + 'model_time': 1.2443009080016054, + 'grad_norm_pre_clip_avg': 0.25678240358829496, + 'learning_rate': 2.2309692738154274e-05, + 'epoch': 3.68} +04/19 [16:37:11] INFO | >> train_qwenlatent.py:487 + Step 14590 | grad_norm_pre_clip=0.2220 | + grad_norm_pre_clip_avg=0.2214 | Metrics: + {'align_loss': 0.024628926068544388, + 'recon_loss': 0.034863438457250595, + 'predict_loss': 0.010462947189807892, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22198444604873657, + 'data_time': 0.0009614900045562536, + 'model_time': 1.2990944280172698, + 'grad_norm_pre_clip_avg': 0.22142996191978453, + 'learning_rate': 2.2304284779453825e-05, + 'epoch': 3.68} +04/19 [16:37:25] INFO | >> train_qwenlatent.py:487 + Step 14600 | grad_norm_pre_clip=0.2070 | + grad_norm_pre_clip_avg=0.2155 | Metrics: + {'align_loss': 0.024339016526937485, + 'recon_loss': 0.051308806985616684, + 'predict_loss': 0.01213991828262806, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20696210861206055, + 'mae_score': 0.016384037120922193, 'data_time': + 0.0010366379865445197, 'model_time': + 1.5170268549991306, 'grad_norm_pre_clip_avg': + 0.2155204474925995, 'learning_rate': + 2.2298872048356424e-05, 'epoch': 3.68} +04/19 [16:37:37] INFO | >> train_qwenlatent.py:487 + Step 14610 | grad_norm_pre_clip=0.2530 | + grad_norm_pre_clip_avg=0.2110 | Metrics: + {'align_loss': 0.02428879588842392, + 'recon_loss': 0.061211664229631424, + 'predict_loss': 0.02010742388665676, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2530191242694855, + 'data_time': 0.000850998010719195, + 'model_time': 1.2358870919852052, + 'grad_norm_pre_clip_avg': 0.21100894063711167, + 'learning_rate': 2.2293454547500176e-05, + 'epoch': 3.69} +04/19 [16:37:50] INFO | >> train_qwenlatent.py:487 + Step 14620 | grad_norm_pre_clip=0.3007 | + grad_norm_pre_clip_avg=0.2192 | Metrics: + {'align_loss': 0.02540118619799614, + 'recon_loss': 0.06523147970438004, + 'predict_loss': 0.013497410342097282, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3007213771343231, + 'data_time': 0.0006909520016051829, + 'model_time': 1.2359600519994274, + 'grad_norm_pre_clip_avg': 0.21919748038053513, + 'learning_rate': 2.22880322795255e-05, 'epoch': + 3.69} +04/19 [16:38:03] INFO | >> train_qwenlatent.py:487 + Step 14630 | grad_norm_pre_clip=0.2451 | + grad_norm_pre_clip_avg=0.2796 | Metrics: + {'align_loss': 0.0238933302462101, + 'recon_loss': 0.04351388290524483, + 'predict_loss': 0.009899618104100227, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2451275885105133, + 'data_time': 0.0006911160016898066, + 'model_time': 1.208801892993506, + 'grad_norm_pre_clip_avg': 0.2795520454645157, + 'learning_rate': 2.2282605247075146e-05, + 'epoch': 3.69} +04/19 [16:38:16] INFO | >> train_qwenlatent.py:487 + Step 14640 | grad_norm_pre_clip=0.2253 | + grad_norm_pre_clip_avg=0.2338 | Metrics: + {'align_loss': 0.026297908276319504, + 'recon_loss': 0.06388480216264725, + 'predict_loss': 0.012933284044265747, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22525371611118317, + 'data_time': 0.0007348470098804682, + 'model_time': 1.1954854800133035, + 'grad_norm_pre_clip_avg': 0.2338374987244606, + 'learning_rate': 2.2277173452794185e-05, + 'epoch': 3.69} +04/19 [16:38:28] INFO | >> train_qwenlatent.py:487 + Step 14650 | grad_norm_pre_clip=0.1715 | + grad_norm_pre_clip_avg=0.2415 | Metrics: + {'align_loss': 0.02348177134990692, + 'recon_loss': 0.05658438056707382, + 'predict_loss': 0.014723106287419796, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17146629095077515, + 'mae_score': 0.024301333040804476, 'data_time': + 0.000816475017927587, 'model_time': + 1.271735380985774, 'grad_norm_pre_clip_avg': + 0.24145984053611755, 'learning_rate': + 2.2271736899330002e-05, 'epoch': 3.7} +04/19 [16:38:41] INFO | >> train_qwenlatent.py:487 + Step 14660 | grad_norm_pre_clip=0.1894 | + grad_norm_pre_clip_avg=0.2237 | Metrics: + {'align_loss': 0.02440149337053299, + 'recon_loss': 0.06025330349802971, + 'predict_loss': 0.012315485626459122, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1894235908985138, + 'data_time': 0.0010305310133844614, + 'model_time': 1.2619825020083226, + 'grad_norm_pre_clip_avg': 0.22367727160453796, + 'learning_rate': 2.2266295589332308e-05, + 'epoch': 3.7} +04/19 [16:38:53] INFO | >> train_qwenlatent.py:487 + Step 14670 | grad_norm_pre_clip=0.2076 | + grad_norm_pre_clip_avg=0.2297 | Metrics: + {'align_loss': 0.025571048259735107, + 'recon_loss': 0.08050242066383362, + 'predict_loss': 0.0212935172021389, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20761661231517792, + 'data_time': 0.0011498690000735223, + 'model_time': 1.2601896389969625, + 'grad_norm_pre_clip_avg': 0.229658542573452, + 'learning_rate': 2.226084952545314e-05, + 'epoch': 3.7} +04/19 [16:39:06] INFO | >> train_qwenlatent.py:487 + Step 14680 | grad_norm_pre_clip=0.2206 | + grad_norm_pre_clip_avg=0.2272 | Metrics: + {'align_loss': 0.026362892240285873, + 'recon_loss': 0.06223702058196068, + 'predict_loss': 0.013766040094196796, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22058075666427612, + 'data_time': 0.0009428570047020912, + 'model_time': 1.2176730429928284, + 'grad_norm_pre_clip_avg': 0.22721811085939408, + 'learning_rate': 2.2255398710346836e-05, + 'epoch': 3.7} +04/19 [16:39:18] INFO | >> train_qwenlatent.py:487 + Step 14690 | grad_norm_pre_clip=0.1886 | + grad_norm_pre_clip_avg=0.2217 | Metrics: + {'align_loss': 0.025310106575489044, + 'recon_loss': 0.040267277508974075, + 'predict_loss': 0.007339314557611942, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18860842287540436, + 'data_time': 0.0007012750138528645, + 'model_time': 1.2624643809976988, + 'grad_norm_pre_clip_avg': 0.22172861099243163, + 'learning_rate': 2.224994314667006e-05, + 'epoch': 3.71} +04/19 [16:39:32] INFO | >> train_qwenlatent.py:487 + Step 14700 | grad_norm_pre_clip=0.2155 | + grad_norm_pre_clip_avg=0.2194 | Metrics: + {'align_loss': 0.024706613272428513, + 'recon_loss': 0.03472503274679184, + 'predict_loss': 0.010399473831057549, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21547181904315948, + 'mae_score': 0.014920616149902343, 'data_time': + 0.0010666839953046292, 'model_time': + 1.2714110260130838, 'grad_norm_pre_clip_avg': + 0.21940302699804307, 'learning_rate': + 2.2244482837081785e-05, 'epoch': 3.71} +04/19 [16:39:44] INFO | >> train_qwenlatent.py:487 + Step 14710 | grad_norm_pre_clip=0.2041 | + grad_norm_pre_clip_avg=0.2208 | Metrics: + {'align_loss': 0.026150334626436234, + 'recon_loss': 0.05978800728917122, + 'predict_loss': 0.010483094491064548, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20414674282073975, + 'data_time': 0.0006911260134074837, + 'model_time': 1.216267407988198, + 'grad_norm_pre_clip_avg': 0.22076887339353563, + 'learning_rate': 2.2239017784243298e-05, + 'epoch': 3.71} +04/19 [16:39:57] INFO | >> train_qwenlatent.py:487 + Step 14720 | grad_norm_pre_clip=0.3152 | + grad_norm_pre_clip_avg=0.2432 | Metrics: + {'align_loss': 0.02548787370324135, + 'recon_loss': 0.07780718058347702, + 'predict_loss': 0.01594283990561962, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.31518906354904175, + 'data_time': 0.0008980199927464128, + 'model_time': 1.2281191220099572, + 'grad_norm_pre_clip_avg': 0.2432289496064186, + 'learning_rate': 2.2233547990818208e-05, + 'epoch': 3.71} +04/19 [16:40:10] INFO | >> train_qwenlatent.py:487 + Step 14730 | grad_norm_pre_clip=0.2360 | + grad_norm_pre_clip_avg=0.2546 | Metrics: + {'align_loss': 0.025050949305295944, + 'recon_loss': 0.06485720723867416, + 'predict_loss': 0.018938947468996048, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2359977662563324, + 'data_time': 0.0012124899949412793, + 'model_time': 1.2194873059925158, + 'grad_norm_pre_clip_avg': 0.25458326041698454, + 'learning_rate': 2.2228073459472413e-05, + 'epoch': 3.72} +04/19 [16:40:23] INFO | >> train_qwenlatent.py:487 + Step 14740 | grad_norm_pre_clip=0.2072 | + grad_norm_pre_clip_avg=0.2263 | Metrics: + {'align_loss': 0.025160912424325943, + 'recon_loss': 0.05659738555550575, + 'predict_loss': 0.011101152747869492, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20718464255332947, + 'data_time': 0.0009867409826256335, + 'model_time': 1.2077239009959158, + 'grad_norm_pre_clip_avg': 0.22631419003009795, + 'learning_rate': 2.2222594192874145e-05, + 'epoch': 3.72} +04/19 [16:40:36] INFO | >> train_qwenlatent.py:487 + Step 14750 | grad_norm_pre_clip=0.1719 | + grad_norm_pre_clip_avg=0.1982 | Metrics: + {'align_loss': 0.024548640474677086, + 'recon_loss': 0.048040058463811874, + 'predict_loss': 0.012189325876533985, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17192241549491882, + 'mae_score': 0.014302662686184719, 'data_time': + 0.0008206920174416155, 'model_time': + 1.243180088000372, 'grad_norm_pre_clip_avg': + 0.1981811836361885, 'learning_rate': + 2.2217110193693923e-05, 'epoch': 3.72} +04/19 [16:40:48] INFO | >> train_qwenlatent.py:487 + Step 14760 | grad_norm_pre_clip=0.2252 | + grad_norm_pre_clip_avg=0.1968 | Metrics: + {'align_loss': 0.024886783212423325, + 'recon_loss': 0.05706721171736717, + 'predict_loss': 0.014969215728342533, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22519101202487946, + 'data_time': 0.0008296669984702021, + 'model_time': 1.2107112809899263, + 'grad_norm_pre_clip_avg': 0.19683605432510376, + 'learning_rate': 2.221162146460459e-05, + 'epoch': 3.72} +04/19 [16:41:01] INFO | >> train_qwenlatent.py:487 + Step 14770 | grad_norm_pre_clip=0.3183 | + grad_norm_pre_clip_avg=0.2219 | Metrics: + {'align_loss': 0.025716420263051987, + 'recon_loss': 0.06639859825372696, + 'predict_loss': 0.013365421444177628, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3182623088359833, + 'data_time': 0.0008711560221854597, + 'model_time': 1.2735617409925908, + 'grad_norm_pre_clip_avg': 0.22193446159362792, + 'learning_rate': 2.220612800828128e-05, + 'epoch': 3.73} +04/19 [16:41:14] INFO | >> train_qwenlatent.py:487 + Step 14780 | grad_norm_pre_clip=0.1948 | + grad_norm_pre_clip_avg=0.2418 | Metrics: + {'align_loss': 0.024855606257915497, + 'recon_loss': 0.04702479764819145, + 'predict_loss': 0.012904347851872444, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19484874606132507, + 'data_time': 0.0007150260207708925, + 'model_time': 1.1803184759919532, + 'grad_norm_pre_clip_avg': 0.2418445512652397, + 'learning_rate': 2.2200629827401436e-05, + 'epoch': 3.73} +04/19 [16:41:27] INFO | >> train_qwenlatent.py:487 + Step 14790 | grad_norm_pre_clip=0.2071 | + grad_norm_pre_clip_avg=0.2059 | Metrics: + {'align_loss': 0.023863457143306732, + 'recon_loss': 0.03968212381005287, + 'predict_loss': 0.010648190043866634, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20713724195957184, + 'data_time': 0.0030850369948893785, + 'model_time': 1.2447112129884772, + 'grad_norm_pre_clip_avg': 0.20593050718307496, + 'learning_rate': 2.219512692464481e-05, + 'epoch': 3.73} +04/19 [16:41:40] INFO | >> train_qwenlatent.py:487 + Step 14800 | grad_norm_pre_clip=0.1718 | + grad_norm_pre_clip_avg=0.2359 | Metrics: + {'align_loss': 0.025421947240829468, + 'recon_loss': 0.056585147976875305, + 'predict_loss': 0.013797655701637268, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17177923023700714, + 'mae_score': 0.01635629808580553, 'data_time': + 0.0007050099957268685, 'model_time': + 1.2121327880013268, 'grad_norm_pre_clip_avg': + 0.2358911380171776, 'learning_rate': + 2.2189619302693445e-05, 'epoch': 3.73} +04/19 [16:41:53] INFO | >> train_qwenlatent.py:487 + Step 14810 | grad_norm_pre_clip=0.2047 | + grad_norm_pre_clip_avg=0.2337 | Metrics: + {'align_loss': 0.023825157433748245, + 'recon_loss': 0.04595020413398743, + 'predict_loss': 0.011479067616164684, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20472455024719238, + 'data_time': 0.0009944310004357249, + 'model_time': 1.540699691977352, + 'grad_norm_pre_clip_avg': 0.23371273428201675, + 'learning_rate': 2.2184106964231688e-05, + 'epoch': 3.74} +04/19 [16:42:05] INFO | >> train_qwenlatent.py:487 + Step 14820 | grad_norm_pre_clip=0.2316 | + grad_norm_pre_clip_avg=0.2077 | Metrics: + {'align_loss': 0.02346990257501602, + 'recon_loss': 0.053687985986471176, + 'predict_loss': 0.013136307708919048, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23161104321479797, + 'data_time': 0.0009812100033741444, + 'model_time': 1.2852770280151162, + 'grad_norm_pre_clip_avg': 0.2076672077178955, + 'learning_rate': 2.2178589911946193e-05, + 'epoch': 3.74} +04/19 [16:42:18] INFO | >> train_qwenlatent.py:487 + Step 14830 | grad_norm_pre_clip=0.2327 | + grad_norm_pre_clip_avg=0.2072 | Metrics: + {'align_loss': 0.025129718706011772, + 'recon_loss': 0.044608719646930695, + 'predict_loss': 0.009103535674512386, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23265032470226288, + 'data_time': 0.0009846719913184643, + 'model_time': 1.258973618998425, + 'grad_norm_pre_clip_avg': 0.20718328952789306, + 'learning_rate': 2.21730681485259e-05, 'epoch': + 3.74} +04/19 [16:42:30] INFO | >> train_qwenlatent.py:487 + Step 14840 | grad_norm_pre_clip=0.2904 | + grad_norm_pre_clip_avg=0.1978 | Metrics: + {'align_loss': 0.025555305182933807, + 'recon_loss': 0.04943590983748436, + 'predict_loss': 0.012233036570250988, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2903953492641449, + 'data_time': 0.0007086330151651055, + 'model_time': 1.2366937899787445, + 'grad_norm_pre_clip_avg': 0.1978294312953949, + 'learning_rate': 2.2167541676662044e-05, + 'epoch': 3.74} +04/19 [16:42:44] INFO | >> train_qwenlatent.py:487 + Step 14850 | grad_norm_pre_clip=0.2363 | + grad_norm_pre_clip_avg=0.2595 | Metrics: + {'align_loss': 0.02528689056634903, + 'recon_loss': 0.054626546800136566, + 'predict_loss': 0.013350014574825764, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23630793392658234, + 'mae_score': 0.016355609034632776, 'data_time': + 0.0010968690039590001, 'model_time': + 1.3114038479980081, 'grad_norm_pre_clip_avg': + 0.25949279963970184, 'learning_rate': + 2.2162010499048173e-05, 'epoch': 3.75} +04/19 [16:42:57] INFO | >> train_qwenlatent.py:487 + Step 14860 | grad_norm_pre_clip=0.2128 | + grad_norm_pre_clip_avg=0.2316 | Metrics: + {'align_loss': 0.025234133005142212, + 'recon_loss': 0.05593058466911316, + 'predict_loss': 0.01733287423849106, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21281816065311432, + 'data_time': 0.0008897189982235432, + 'model_time': 1.2109455640020315, + 'grad_norm_pre_clip_avg': 0.23159541487693786, + 'learning_rate': 2.2156474618380102e-05, + 'epoch': 3.75} +04/19 [16:43:09] INFO | >> train_qwenlatent.py:487 + Step 14870 | grad_norm_pre_clip=0.1784 | + grad_norm_pre_clip_avg=0.1995 | Metrics: + {'align_loss': 0.024700986221432686, + 'recon_loss': 0.05771667882800102, + 'predict_loss': 0.014247383922338486, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17837309837341309, + 'data_time': 0.0009529920062050223, + 'model_time': 1.2419148259796202, + 'grad_norm_pre_clip_avg': 0.19945444762706757, + 'learning_rate': 2.2150934037355964e-05, + 'epoch': 3.75} +04/19 [16:43:22] INFO | >> train_qwenlatent.py:487 + Step 14880 | grad_norm_pre_clip=0.3082 | + grad_norm_pre_clip_avg=0.2420 | Metrics: + {'align_loss': 0.02459125593304634, + 'recon_loss': 0.06459957361221313, + 'predict_loss': 0.018128613010048866, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3081752061843872, + 'data_time': 0.0015895850083325058, + 'model_time': 1.2190310239966493, + 'grad_norm_pre_clip_avg': 0.24199483841657637, + 'learning_rate': 2.2145388758676163e-05, + 'epoch': 3.75} +04/19 [16:43:34] INFO | >> train_qwenlatent.py:487 + Step 14890 | grad_norm_pre_clip=0.1847 | + grad_norm_pre_clip_avg=0.2171 | Metrics: + {'align_loss': 0.02474828064441681, + 'recon_loss': 0.042541567236185074, + 'predict_loss': 0.009738946333527565, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18470913171768188, + 'data_time': 0.0006782219861634076, + 'model_time': 1.2226269749808125, + 'grad_norm_pre_clip_avg': 0.217097969353199, + 'learning_rate': 2.2139838785043402e-05, + 'epoch': 3.76} +04/19 [16:43:47] INFO | >> train_qwenlatent.py:487 + Step 14900 | grad_norm_pre_clip=0.2949 | + grad_norm_pre_clip_avg=0.2596 | Metrics: + {'align_loss': 0.025126948952674866, + 'recon_loss': 0.04687478765845299, + 'predict_loss': 0.01111475471407175, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2948959171772003, + 'mae_score': 0.015462715561325486, 'data_time': + 0.0009219979983754456, 'model_time': + 1.215305782010546, 'grad_norm_pre_clip_avg': + 0.25958863347768785, 'learning_rate': + 2.2134284119162667e-05, 'epoch': 3.76} +04/19 [16:44:00] INFO | >> train_qwenlatent.py:487 + Step 14910 | grad_norm_pre_clip=0.2935 | + grad_norm_pre_clip_avg=0.2454 | Metrics: + {'align_loss': 0.025499001145362854, + 'recon_loss': 0.05808478221297264, + 'predict_loss': 0.012376675382256508, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2935096323490143, + 'data_time': 0.0008310730045195669, + 'model_time': 1.2225687089958228, + 'grad_norm_pre_clip_avg': 0.2454178288578987, + 'learning_rate': 2.2128724763741246e-05, + 'epoch': 3.76} +04/19 [16:44:13] INFO | >> train_qwenlatent.py:487 + Step 14920 | grad_norm_pre_clip=0.2631 | + grad_norm_pre_clip_avg=0.2564 | Metrics: + {'align_loss': 0.0253022238612175, + 'recon_loss': 0.054463911801576614, + 'predict_loss': 0.009917767718434334, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26312917470932007, + 'data_time': 0.0006847280019428581, + 'model_time': 1.2050838909926824, + 'grad_norm_pre_clip_avg': 0.25641005784273146, + 'learning_rate': 2.212316072148869e-05, + 'epoch': 3.76} +04/19 [16:44:25] INFO | >> train_qwenlatent.py:487 + Step 14930 | grad_norm_pre_clip=0.1951 | + grad_norm_pre_clip_avg=0.2295 | Metrics: + {'align_loss': 0.025528348982334137, + 'recon_loss': 0.07887546718120575, + 'predict_loss': 0.020993022248148918, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1950775533914566, + 'data_time': 0.0006624499801546335, + 'model_time': 1.2061853419872932, + 'grad_norm_pre_clip_avg': 0.2294544607400894, + 'learning_rate': 2.2117591995116853e-05, + 'epoch': 3.77} +04/19 [16:44:38] INFO | >> train_qwenlatent.py:487 + Step 14940 | grad_norm_pre_clip=0.2055 | + grad_norm_pre_clip_avg=0.2246 | Metrics: + {'align_loss': 0.025719311088323593, + 'recon_loss': 0.05810864642262459, + 'predict_loss': 0.014121374115347862, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2055131494998932, + 'data_time': 0.0007531290175393224, + 'model_time': 1.2224954220000654, + 'grad_norm_pre_clip_avg': 0.22460006326436996, + 'learning_rate': 2.2112018587339863e-05, + 'epoch': 3.77} +04/19 [16:44:51] INFO | >> train_qwenlatent.py:487 + Step 14950 | grad_norm_pre_clip=0.1883 | + grad_norm_pre_clip_avg=0.2213 | Metrics: + {'align_loss': 0.024510616436600685, + 'recon_loss': 0.051028456538915634, + 'predict_loss': 0.012226706370711327, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1883281022310257, + 'mae_score': 0.017917220656936232, 'data_time': + 0.0009620509808883071, 'model_time': + 1.2614110879949294, 'grad_norm_pre_clip_avg': + 0.2213495820760727, 'learning_rate': + 2.2106440500874128e-05, 'epoch': 3.77} +04/19 [16:45:03] INFO | >> train_qwenlatent.py:487 + Step 14960 | grad_norm_pre_clip=0.2020 | + grad_norm_pre_clip_avg=0.1798 | Metrics: + {'align_loss': 0.024413228034973145, + 'recon_loss': 0.04411725699901581, + 'predict_loss': 0.00866418331861496, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20200997591018677, + 'data_time': 0.0006363240245264024, + 'model_time': 1.2215349530160893, + 'grad_norm_pre_clip_avg': 0.17980406433343887, + 'learning_rate': 2.210085773843834e-05, + 'epoch': 3.77} +04/19 [16:45:16] INFO | >> train_qwenlatent.py:487 + Step 14970 | grad_norm_pre_clip=0.1914 | + grad_norm_pre_clip_avg=0.2520 | Metrics: + {'align_loss': 0.026119239628314972, + 'recon_loss': 0.05033137649297714, + 'predict_loss': 0.010456089861690998, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1913926750421524, + 'data_time': 0.0006734649941790849, + 'model_time': 1.258650209987536, + 'grad_norm_pre_clip_avg': 0.2520062282681465, + 'learning_rate': 2.2095270302753468e-05, + 'epoch': 3.78} +04/19 [16:45:29] INFO | >> train_qwenlatent.py:487 + Step 14980 | grad_norm_pre_clip=0.1832 | + grad_norm_pre_clip_avg=0.2099 | Metrics: + {'align_loss': 0.024485919624567032, + 'recon_loss': 0.0699707567691803, + 'predict_loss': 0.01479873713105917, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18317361176013947, + 'data_time': 0.0011843800020869821, + 'model_time': 1.272232482995605, + 'grad_norm_pre_clip_avg': 0.20992770791053772, + 'learning_rate': 2.2089678196542767e-05, + 'epoch': 3.78} +04/19 [16:45:41] INFO | >> train_qwenlatent.py:487 + Step 14990 | grad_norm_pre_clip=0.4262 | + grad_norm_pre_clip_avg=0.2409 | Metrics: + {'align_loss': 0.026467394083738327, + 'recon_loss': 0.054660893976688385, + 'predict_loss': 0.009814162738621235, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.42618855834007263, + 'data_time': 0.000828909978736192, + 'model_time': 1.20681967298151, + 'grad_norm_pre_clip_avg': 0.24088843911886215, + 'learning_rate': 2.2084081422531756e-05, + 'epoch': 3.78} +04/19 [16:45:54] INFO | >> train_qwenlatent.py:487 + Step 15000 | grad_norm_pre_clip=0.1786 | + grad_norm_pre_clip_avg=0.2575 | Metrics: + {'align_loss': 0.02432466670870781, + 'recon_loss': 0.06433919072151184, + 'predict_loss': 0.014039825648069382, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17855030298233032, + 'mae_score': 0.015132239058211042, 'data_time': + 0.0006793960055802017, 'model_time': + 1.25568650700734, 'grad_norm_pre_clip_avg': + 0.25750820338726044, 'learning_rate': + 2.207847998344824e-05, 'epoch': 3.79} +✅ Checkpoint saved at ./runs/0418_QwenLatent_13tasks_actionstate_30k/checkpoints/steps_15000 +04/19 [16:46:17] INFO | >> train_qwenlatent.py:487 + Step 15010 | grad_norm_pre_clip=0.2528 | + grad_norm_pre_clip_avg=0.2355 | Metrics: + {'align_loss': 0.025601521134376526, + 'recon_loss': 0.051706086844205856, + 'predict_loss': 0.010278095491230488, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25278887152671814, + 'data_time': 0.0008580790017731488, + 'model_time': 1.2970075019984506, + 'grad_norm_pre_clip_avg': 0.23550158441066743, + 'learning_rate': 2.2072873882022285e-05, + 'epoch': 3.79} +04/19 [16:46:30] INFO | >> train_qwenlatent.py:487 + Step 15020 | grad_norm_pre_clip=0.2131 | + grad_norm_pre_clip_avg=0.2178 | Metrics: + {'align_loss': 0.024714604020118713, + 'recon_loss': 0.05669526010751724, + 'predict_loss': 0.01648685336112976, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21308189630508423, + 'data_time': 0.0010664040164556354, + 'model_time': 1.3437403340067249, + 'grad_norm_pre_clip_avg': 0.21780273467302322, + 'learning_rate': 2.2067263120986246e-05, + 'epoch': 3.79} +04/19 [16:46:43] INFO | >> train_qwenlatent.py:487 + Step 15030 | grad_norm_pre_clip=0.1850 | + grad_norm_pre_clip_avg=0.1979 | Metrics: + {'align_loss': 0.02688409388065338, + 'recon_loss': 0.06303771585226059, + 'predict_loss': 0.01325808186084032, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1850387156009674, + 'data_time': 0.0006814250082243234, + 'model_time': 1.2806419949920382, + 'grad_norm_pre_clip_avg': 0.19793114066123962, + 'learning_rate': 2.2061647703074732e-05, + 'epoch': 3.79} +04/19 [16:46:56] INFO | >> train_qwenlatent.py:487 + Step 15040 | grad_norm_pre_clip=0.1944 | + grad_norm_pre_clip_avg=0.2065 | Metrics: + {'align_loss': 0.024688642472028732, + 'recon_loss': 0.050356172025203705, + 'predict_loss': 0.012437066063284874, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19442379474639893, + 'data_time': 0.0006303540139924735, + 'model_time': 1.2135640749766026, + 'grad_norm_pre_clip_avg': 0.2065296322107315, + 'learning_rate': 2.2056027631024625e-05, + 'epoch': 3.8} +04/19 [16:47:10] INFO | >> train_qwenlatent.py:487 + Step 15050 | grad_norm_pre_clip=0.2362 | + grad_norm_pre_clip_avg=0.2804 | Metrics: + {'align_loss': 0.025916893035173416, + 'recon_loss': 0.054607782512903214, + 'predict_loss': 0.01855892688035965, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23619067668914795, + 'mae_score': 0.015152329797143335, 'data_time': + 0.0009174139995593578, 'model_time': + 1.25627047198941, 'grad_norm_pre_clip_avg': + 0.28040864020586015, 'learning_rate': + 2.2050402907575095e-05, 'epoch': 3.8} +04/19 [16:47:22] INFO | >> train_qwenlatent.py:487 + Step 15060 | grad_norm_pre_clip=0.1821 | + grad_norm_pre_clip_avg=0.2315 | Metrics: + {'align_loss': 0.022817764431238174, + 'recon_loss': 0.041312724351882935, + 'predict_loss': 0.00969709362834692, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18206478655338287, + 'data_time': 0.0007639770046807826, + 'model_time': 1.2127394010021817, + 'grad_norm_pre_clip_avg': 0.23149919956922532, + 'learning_rate': 2.2044773535467557e-05, + 'epoch': 3.8} +04/19 [16:47:35] INFO | >> train_qwenlatent.py:487 + Step 15070 | grad_norm_pre_clip=0.1692 | + grad_norm_pre_clip_avg=0.1942 | Metrics: + {'align_loss': 0.024728555232286453, + 'recon_loss': 0.06210792064666748, + 'predict_loss': 0.017084218561649323, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16920772194862366, + 'data_time': 0.001068943995051086, + 'model_time': 1.3018185980035923, + 'grad_norm_pre_clip_avg': 0.19420809745788575, + 'learning_rate': 2.203913951744569e-05, + 'epoch': 3.8} +04/19 [16:47:47] INFO | >> train_qwenlatent.py:487 + Step 15080 | grad_norm_pre_clip=0.3003 | + grad_norm_pre_clip_avg=0.1979 | Metrics: + {'align_loss': 0.02459009550511837, + 'recon_loss': 0.05258749797940254, + 'predict_loss': 0.015595197677612305, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.30025774240493774, + 'data_time': 0.0007597939984407276, + 'model_time': 1.2700841560144909, + 'grad_norm_pre_clip_avg': 0.19791386574506759, + 'learning_rate': 2.2033500856255457e-05, + 'epoch': 3.81} +04/19 [16:48:00] INFO | >> train_qwenlatent.py:487 + Step 15090 | grad_norm_pre_clip=0.2053 | + grad_norm_pre_clip_avg=0.2991 | Metrics: + {'align_loss': 0.023983009159564972, + 'recon_loss': 0.04668882489204407, + 'predict_loss': 0.013759952038526535, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20533019304275513, + 'data_time': 0.0008746609964873642, + 'model_time': 1.2490887719905004, + 'grad_norm_pre_clip_avg': 0.29905906021595, + 'learning_rate': 2.2027857554645067e-05, + 'epoch': 3.81} +04/19 [16:48:13] INFO | >> train_qwenlatent.py:487 + Step 15100 | grad_norm_pre_clip=0.2134 | + grad_norm_pre_clip_avg=0.2301 | Metrics: + {'align_loss': 0.02500924840569496, + 'recon_loss': 0.058300990611314774, + 'predict_loss': 0.011473484337329865, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21337416768074036, + 'mae_score': 0.011711639541763443, 'data_time': + 0.001016257010633126, 'model_time': + 1.3002062280138489, 'grad_norm_pre_clip_avg': + 0.23014235496520996, 'learning_rate': + 2.2022209615365003e-05, 'epoch': 3.81} +04/19 [16:48:26] INFO | >> train_qwenlatent.py:487 + Step 15110 | grad_norm_pre_clip=0.1956 | + grad_norm_pre_clip_avg=0.2043 | Metrics: + {'align_loss': 0.025268100202083588, + 'recon_loss': 0.05849463865160942, + 'predict_loss': 0.011784933507442474, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1955588459968567, + 'data_time': 0.0013861250190529972, + 'model_time': 1.2101483909937087, + 'grad_norm_pre_clip_avg': 0.2042593851685524, + 'learning_rate': 2.2016557041167994e-05, + 'epoch': 3.81} +04/19 [16:48:38] INFO | >> train_qwenlatent.py:487 + Step 15120 | grad_norm_pre_clip=0.2096 | + grad_norm_pre_clip_avg=0.1913 | Metrics: + {'align_loss': 0.025626104325056076, + 'recon_loss': 0.051651764661073685, + 'predict_loss': 0.009834470227360725, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20960889756679535, + 'data_time': 0.0007527109992224723, + 'model_time': 1.3127970829955302, + 'grad_norm_pre_clip_avg': 0.19133830666542054, + 'learning_rate': 2.201089983480904e-05, + 'epoch': 3.82} +04/19 [16:48:51] INFO | >> train_qwenlatent.py:487 + Step 15130 | grad_norm_pre_clip=0.3078 | + grad_norm_pre_clip_avg=0.2074 | Metrics: + {'align_loss': 0.025724854320287704, + 'recon_loss': 0.06642874330282211, + 'predict_loss': 0.016406819224357605, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3078457713127136, + 'data_time': 0.0009837129910010844, + 'model_time': 1.2575265339983162, + 'grad_norm_pre_clip_avg': 0.2073906734585762, + 'learning_rate': 2.2005237999045392e-05, + 'epoch': 3.82} +04/19 [16:49:04] INFO | >> train_qwenlatent.py:487 + Step 15140 | grad_norm_pre_clip=0.1706 | + grad_norm_pre_clip_avg=0.2625 | Metrics: + {'align_loss': 0.024501889944076538, + 'recon_loss': 0.07490935921669006, + 'predict_loss': 0.014806785620748997, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1706342250108719, + 'data_time': 0.000771898019593209, + 'model_time': 1.2347303880087566, + 'grad_norm_pre_clip_avg': 0.2625365599989891, + 'learning_rate': 2.199957153663657e-05, + 'epoch': 3.82} +04/19 [16:49:17] INFO | >> train_qwenlatent.py:487 + Step 15150 | grad_norm_pre_clip=0.2271 | + grad_norm_pre_clip_avg=0.2059 | Metrics: + {'align_loss': 0.02478768303990364, + 'recon_loss': 0.05062936618924141, + 'predict_loss': 0.011059396900236607, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22714252769947052, + 'mae_score': 0.015240043777603287, 'data_time': + 0.0012109849776607007, 'model_time': + 1.2502133499947377, 'grad_norm_pre_clip_avg': + 0.20588727295398712, 'learning_rate': + 2.1993900450344324e-05, 'epoch': 3.82} +04/19 [16:49:30] INFO | >> train_qwenlatent.py:487 + Step 15160 | grad_norm_pre_clip=0.1591 | + grad_norm_pre_clip_avg=0.2043 | Metrics: + {'align_loss': 0.025243695825338364, + 'recon_loss': 0.07507815212011337, + 'predict_loss': 0.012783159501850605, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1590520441532135, + 'data_time': 0.0010202080011367798, + 'model_time': 1.2425615880056284, + 'grad_norm_pre_clip_avg': 0.2043016165494919, + 'learning_rate': 2.1988224742932686e-05, + 'epoch': 3.83} +04/19 [16:49:43] INFO | >> train_qwenlatent.py:487 + Step 15170 | grad_norm_pre_clip=0.1845 | + grad_norm_pre_clip_avg=0.2389 | Metrics: + {'align_loss': 0.025433281436562538, + 'recon_loss': 0.06755002588033676, + 'predict_loss': 0.013968302868306637, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18449999392032623, + 'data_time': 0.0006570619880221784, + 'model_time': 1.2492675799876451, + 'grad_norm_pre_clip_avg': 0.23890191167593003, + 'learning_rate': 2.1982544417167916e-05, + 'epoch': 3.83} +04/19 [16:49:55] INFO | >> train_qwenlatent.py:487 + Step 15180 | grad_norm_pre_clip=0.2390 | + grad_norm_pre_clip_avg=0.2185 | Metrics: + {'align_loss': 0.024838808923959732, + 'recon_loss': 0.06467706710100174, + 'predict_loss': 0.016797296702861786, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2389974594116211, + 'data_time': 0.0009528000082354993, + 'model_time': 1.2544793539855164, + 'grad_norm_pre_clip_avg': 0.21853020191192626, + 'learning_rate': 2.197685947581854e-05, + 'epoch': 3.83} +04/19 [16:50:08] INFO | >> train_qwenlatent.py:487 + Step 15190 | grad_norm_pre_clip=0.1858 | + grad_norm_pre_clip_avg=0.2560 | Metrics: + {'align_loss': 0.024523988366127014, + 'recon_loss': 0.06443500518798828, + 'predict_loss': 0.01489245519042015, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18584078550338745, + 'data_time': 0.000635434000287205, + 'model_time': 1.2034757079964038, + 'grad_norm_pre_clip_avg': 0.256013485789299, + 'learning_rate': 2.1971169921655333e-05, + 'epoch': 3.83} +04/19 [16:50:21] INFO | >> train_qwenlatent.py:487 + Step 15200 | grad_norm_pre_clip=0.2576 | + grad_norm_pre_clip_avg=0.2454 | Metrics: + {'align_loss': 0.025094598531723022, + 'recon_loss': 0.0651656985282898, + 'predict_loss': 0.013298436999320984, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25760453939437866, + 'mae_score': 0.012943815111039996, 'data_time': + 0.0008246240031439811, 'model_time': + 1.248234925995348, 'grad_norm_pre_clip_avg': + 0.24536338448524475, 'learning_rate': + 2.196547575745131e-05, 'epoch': 3.84} +04/19 [16:50:34] INFO | >> train_qwenlatent.py:487 + Step 15210 | grad_norm_pre_clip=0.1792 | + grad_norm_pre_clip_avg=0.2094 | Metrics: + {'align_loss': 0.025935817509889603, + 'recon_loss': 0.06427318602800369, + 'predict_loss': 0.01002639252692461, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17915058135986328, + 'data_time': 0.00117360899457708, 'model_time': + 1.3667777850059792, 'grad_norm_pre_clip_avg': + 0.209418186545372, 'learning_rate': + 2.1959776985981737e-05, 'epoch': 3.84} +04/19 [16:50:47] INFO | >> train_qwenlatent.py:487 + Step 15220 | grad_norm_pre_clip=0.1734 | + grad_norm_pre_clip_avg=0.2157 | Metrics: + {'align_loss': 0.024474913254380226, + 'recon_loss': 0.05678362026810646, + 'predict_loss': 0.016790861263871193, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17336325347423553, + 'data_time': 0.0010380279854871333, + 'model_time': 1.2615481489920057, + 'grad_norm_pre_clip_avg': 0.21566421985626222, + 'learning_rate': 2.195407361002413e-05, + 'epoch': 3.84} +04/19 [16:50:59] INFO | >> train_qwenlatent.py:487 + Step 15230 | grad_norm_pre_clip=0.2497 | + grad_norm_pre_clip_avg=0.2532 | Metrics: + {'align_loss': 0.024492114782333374, + 'recon_loss': 0.055812735110521317, + 'predict_loss': 0.014288509264588356, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24966490268707275, + 'data_time': 0.0005837350036017597, + 'model_time': 1.13491683601751, + 'grad_norm_pre_clip_avg': 0.253234526515007, + 'learning_rate': 2.1948365632358243e-05, + 'epoch': 3.84} +04/19 [16:51:11] INFO | >> train_qwenlatent.py:487 + Step 15240 | grad_norm_pre_clip=0.2573 | + grad_norm_pre_clip_avg=0.2217 | Metrics: + {'align_loss': 0.02467098832130432, + 'recon_loss': 0.05531817302107811, + 'predict_loss': 0.013872555457055569, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2573201358318329, + 'data_time': 0.0006454259855672717, + 'model_time': 1.148498037015088, + 'grad_norm_pre_clip_avg': 0.22166199684143068, + 'learning_rate': 2.1942653055766075e-05, + 'epoch': 3.85} +04/19 [16:51:23] INFO | >> train_qwenlatent.py:487 + Step 15250 | grad_norm_pre_clip=0.2673 | + grad_norm_pre_clip_avg=0.2099 | Metrics: + {'align_loss': 0.024933679029345512, + 'recon_loss': 0.07798781991004944, + 'predict_loss': 0.02006450481712818, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26726841926574707, + 'mae_score': 0.01309837564691767, 'data_time': + 0.0006239800131879747, 'model_time': + 1.1638427900034003, 'grad_norm_pre_clip_avg': + 0.20988118648529053, 'learning_rate': + 2.1936935883031867e-05, 'epoch': 3.85} +04/19 [16:51:35] INFO | >> train_qwenlatent.py:487 + Step 15260 | grad_norm_pre_clip=0.2649 | + grad_norm_pre_clip_avg=0.2447 | Metrics: + {'align_loss': 0.025164447724819183, + 'recon_loss': 0.03477849066257477, + 'predict_loss': 0.013666526414453983, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2649102509021759, + 'data_time': 0.0005574460083153099, + 'model_time': 1.1394272350007668, + 'grad_norm_pre_clip_avg': 0.24472444504499435, + 'learning_rate': 2.1931214116942096e-05, + 'epoch': 3.85} +04/19 [16:51:47] INFO | >> train_qwenlatent.py:487 + Step 15270 | grad_norm_pre_clip=0.1860 | + grad_norm_pre_clip_avg=0.2427 | Metrics: + {'align_loss': 0.024980077520012856, + 'recon_loss': 0.06967110931873322, + 'predict_loss': 0.015581715852022171, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1859978437423706, + 'data_time': 0.0006366719899233431, + 'model_time': 1.1722211809828877, + 'grad_norm_pre_clip_avg': 0.24273700416088104, + 'learning_rate': 2.1925487760285487e-05, + 'epoch': 3.85} +04/19 [16:51:59] INFO | >> train_qwenlatent.py:487 + Step 15280 | grad_norm_pre_clip=0.2214 | + grad_norm_pre_clip_avg=0.2106 | Metrics: + {'align_loss': 0.026021992787718773, + 'recon_loss': 0.0666094571352005, + 'predict_loss': 0.014549106359481812, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22141927480697632, + 'data_time': 0.0006297680083662271, + 'model_time': 1.1806796390155796, + 'grad_norm_pre_clip_avg': 0.2106487885117531, + 'learning_rate': 2.1919756815852996e-05, + 'epoch': 3.86} +04/19 [16:52:10] INFO | >> train_qwenlatent.py:487 + Step 15290 | grad_norm_pre_clip=0.2785 | + grad_norm_pre_clip_avg=0.2355 | Metrics: + {'align_loss': 0.025642208755016327, + 'recon_loss': 0.07761026173830032, + 'predict_loss': 0.015161626040935516, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.27849459648132324, + 'data_time': 0.0006135319999884814, + 'model_time': 1.1496292209776584, + 'grad_norm_pre_clip_avg': 0.2355276420712471, + 'learning_rate': 2.191402128643781e-05, + 'epoch': 3.86} +04/19 [16:52:23] INFO | >> train_qwenlatent.py:487 + Step 15300 | grad_norm_pre_clip=0.1934 | + grad_norm_pre_clip_avg=0.2115 | Metrics: + {'align_loss': 0.025618739426136017, + 'recon_loss': 0.07346634566783905, + 'predict_loss': 0.018449561670422554, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19339996576309204, + 'mae_score': 0.013374337204941758, 'data_time': + 0.0006520389870274812, 'model_time': + 1.137642115005292, 'grad_norm_pre_clip_avg': + 0.21147815585136415, 'learning_rate': + 2.190828117483536e-05, 'epoch': 3.86} +04/19 [16:52:34] INFO | >> train_qwenlatent.py:487 + Step 15310 | grad_norm_pre_clip=0.2376 | + grad_norm_pre_clip_avg=0.2245 | Metrics: + {'align_loss': 0.024561427533626556, + 'recon_loss': 0.04430852085351944, + 'predict_loss': 0.01035311259329319, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23755550384521484, + 'data_time': 0.0006222430092748255, + 'model_time': 1.1569033809937537, + 'grad_norm_pre_clip_avg': 0.22446617782115935, + 'learning_rate': 2.1902536483843306e-05, + 'epoch': 3.86} +04/19 [16:52:46] INFO | >> train_qwenlatent.py:487 + Step 15320 | grad_norm_pre_clip=0.1765 | + grad_norm_pre_clip_avg=0.2371 | Metrics: + {'align_loss': 0.024387061595916748, + 'recon_loss': 0.052514635026454926, + 'predict_loss': 0.010326505638659, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1764756739139557, + 'data_time': 0.0008568479970563203, + 'model_time': 1.1376463730121031, + 'grad_norm_pre_clip_avg': 0.2371036648750305, + 'learning_rate': 2.1896787216261543e-05, + 'epoch': 3.87} +04/19 [16:52:58] INFO | >> train_qwenlatent.py:487 + Step 15330 | grad_norm_pre_clip=0.2202 | + grad_norm_pre_clip_avg=0.2434 | Metrics: + {'align_loss': 0.024330850690603256, + 'recon_loss': 0.0704936757683754, + 'predict_loss': 0.01424085721373558, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22016562521457672, + 'data_time': 0.0005641010066028684, + 'model_time': 1.1328582919959445, + 'grad_norm_pre_clip_avg': 0.24343315958976747, + 'learning_rate': 2.1891033374892193e-05, + 'epoch': 3.87} +04/19 [16:53:09] INFO | >> train_qwenlatent.py:487 + Step 15340 | grad_norm_pre_clip=0.1685 | + grad_norm_pre_clip_avg=0.1958 | Metrics: + {'align_loss': 0.024228690192103386, + 'recon_loss': 0.05172676593065262, + 'predict_loss': 0.01090694684535265, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16847622394561768, + 'data_time': 0.0005816519842483103, + 'model_time': 1.1334291159873828, + 'grad_norm_pre_clip_avg': 0.19580421596765518, + 'learning_rate': 2.1885274962539606e-05, + 'epoch': 3.87} +04/19 [16:53:22] INFO | >> train_qwenlatent.py:487 + Step 15350 | grad_norm_pre_clip=0.2220 | + grad_norm_pre_clip_avg=0.1912 | Metrics: + {'align_loss': 0.025008758530020714, + 'recon_loss': 0.07924840599298477, + 'predict_loss': 0.016735026612877846, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22202779352664948, + 'mae_score': 0.016335442689088014, 'data_time': + 0.0006603479851037264, 'model_time': + 1.1752941999875475, 'grad_norm_pre_clip_avg': + 0.1911909595131874, 'learning_rate': + 2.1879511982010364e-05, 'epoch': 3.87} +04/19 [16:53:33] INFO | >> train_qwenlatent.py:487 + Step 15360 | grad_norm_pre_clip=0.2954 | + grad_norm_pre_clip_avg=0.2875 | Metrics: + {'align_loss': 0.02486930415034294, + 'recon_loss': 0.0667220801115036, + 'predict_loss': 0.015206173993647099, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.29544222354888916, + 'data_time': 0.0005280479963403195, + 'model_time': 1.1445283399953041, + 'grad_norm_pre_clip_avg': 0.28752614855766295, + 'learning_rate': 2.1873744436113274e-05, + 'epoch': 3.88} +04/19 [16:53:45] INFO | >> train_qwenlatent.py:487 + Step 15370 | grad_norm_pre_clip=0.1921 | + grad_norm_pre_clip_avg=0.2122 | Metrics: + {'align_loss': 0.02595292031764984, + 'recon_loss': 0.07596691697835922, + 'predict_loss': 0.015344353392720222, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1920836865901947, + 'data_time': 0.000575204991037026, + 'model_time': 1.1331520119856577, + 'grad_norm_pre_clip_avg': 0.21218465864658356, + 'learning_rate': 2.1867972327659366e-05, + 'epoch': 3.88} +04/19 [16:53:56] INFO | >> train_qwenlatent.py:487 + Step 15380 | grad_norm_pre_clip=0.1542 | + grad_norm_pre_clip_avg=0.1807 | Metrics: + {'align_loss': 0.023767981678247452, + 'recon_loss': 0.0458601675927639, + 'predict_loss': 0.008488216437399387, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15419171750545502, + 'data_time': 0.0006969980022404343, + 'model_time': 1.1479696640162729, + 'grad_norm_pre_clip_avg': 0.180714713037014, + 'learning_rate': 2.18621956594619e-05, 'epoch': + 3.88} +04/19 [16:54:08] INFO | >> train_qwenlatent.py:487 + Step 15390 | grad_norm_pre_clip=0.3306 | + grad_norm_pre_clip_avg=0.2025 | Metrics: + {'align_loss': 0.025641147047281265, + 'recon_loss': 0.061347268521785736, + 'predict_loss': 0.018969053402543068, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3306388258934021, + 'data_time': 0.0006095780117902905, + 'model_time': 1.3186621299828403, + 'grad_norm_pre_clip_avg': 0.20253496170043944, + 'learning_rate': 2.1856414434336352e-05, + 'epoch': 3.88} +04/19 [16:54:20] INFO | >> train_qwenlatent.py:487 + Step 15400 | grad_norm_pre_clip=0.1577 | + grad_norm_pre_clip_avg=0.2542 | Metrics: + {'align_loss': 0.02613285556435585, + 'recon_loss': 0.05746464058756828, + 'predict_loss': 0.012019886635243893, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1576516330242157, + 'mae_score': 0.016545522535169447, 'data_time': + 0.0005524379957932979, 'model_time': + 1.1509093409986235, 'grad_norm_pre_clip_avg': + 0.2541896477341652, 'learning_rate': + 2.1850628655100418e-05, 'epoch': 3.89} +04/19 [16:54:32] INFO | >> train_qwenlatent.py:487 + Step 15410 | grad_norm_pre_clip=0.1910 | + grad_norm_pre_clip_avg=0.2336 | Metrics: + {'align_loss': 0.02422790229320526, + 'recon_loss': 0.06210274249315262, + 'predict_loss': 0.018464472144842148, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19103243947029114, + 'data_time': 0.0007749829965177923, + 'model_time': 1.1475257889833301, + 'grad_norm_pre_clip_avg': 0.23361982107162477, + 'learning_rate': 2.1844838324574018e-05, + 'epoch': 3.89} +04/19 [16:54:43] INFO | >> train_qwenlatent.py:487 + Step 15420 | grad_norm_pre_clip=0.1981 | + grad_norm_pre_clip_avg=0.2315 | Metrics: + {'align_loss': 0.023814987391233444, + 'recon_loss': 0.046752024441957474, + 'predict_loss': 0.012035422027111053, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19814088940620422, + 'data_time': 0.0005987250187899917, + 'model_time': 1.1454440889938269, + 'grad_norm_pre_clip_avg': 0.23147615790367126, + 'learning_rate': 2.183904344557929e-05, + 'epoch': 3.89} +04/19 [16:54:55] INFO | >> train_qwenlatent.py:487 + Step 15430 | grad_norm_pre_clip=0.1996 | + grad_norm_pre_clip_avg=0.2335 | Metrics: + {'align_loss': 0.02427249774336815, + 'recon_loss': 0.05789540708065033, + 'predict_loss': 0.012551949359476566, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19963832199573517, + 'data_time': 0.0005507960158865899, + 'model_time': 1.1648405200103298, + 'grad_norm_pre_clip_avg': 0.23350781798362732, + 'learning_rate': 2.1833244020940586e-05, + 'epoch': 3.89} +04/19 [16:55:07] INFO | >> train_qwenlatent.py:487 + Step 15440 | grad_norm_pre_clip=0.1835 | + grad_norm_pre_clip_avg=0.2019 | Metrics: + {'align_loss': 0.023870235309004784, + 'recon_loss': 0.045432768762111664, + 'predict_loss': 0.01000034250319004, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18348541855812073, + 'data_time': 0.0006199000054039061, + 'model_time': 1.156919741013553, + 'grad_norm_pre_clip_avg': 0.20189260989427565, + 'learning_rate': 2.1827440053484476e-05, + 'epoch': 3.9} +04/19 [16:55:19] INFO | >> train_qwenlatent.py:487 + Step 15450 | grad_norm_pre_clip=0.2536 | + grad_norm_pre_clip_avg=0.1981 | Metrics: + {'align_loss': 0.026358479633927345, + 'recon_loss': 0.0683068335056305, + 'predict_loss': 0.016992904245853424, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2535542845726013, + 'mae_score': 0.01588431349745742, 'data_time': + 0.0006595479790121317, 'model_time': + 1.1552044679992832, 'grad_norm_pre_clip_avg': + 0.19810947328805922, 'learning_rate': + 2.182163154603974e-05, 'epoch': 3.9} +04/19 [16:55:30] INFO | >> train_qwenlatent.py:487 + Step 15460 | grad_norm_pre_clip=0.2229 | + grad_norm_pre_clip_avg=0.2640 | Metrics: + {'align_loss': 0.025683116167783737, + 'recon_loss': 0.07227112352848053, + 'predict_loss': 0.015792544931173325, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22285974025726318, + 'data_time': 0.0006113460112828761, + 'model_time': 1.1336869889928494, + 'grad_norm_pre_clip_avg': 0.26396832019090655, + 'learning_rate': 2.1815818501437384e-05, + 'epoch': 3.9} +04/19 [16:55:42] INFO | >> train_qwenlatent.py:487 + Step 15470 | grad_norm_pre_clip=0.2426 | + grad_norm_pre_clip_avg=0.2382 | Metrics: + {'align_loss': 0.02464832365512848, + 'recon_loss': 0.06555359810590744, + 'predict_loss': 0.016008133068680763, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24259904026985168, + 'data_time': 0.0005800950166303664, + 'model_time': 1.1484616570232902, + 'grad_norm_pre_clip_avg': 0.23819469809532165, + 'learning_rate': 2.1810000922510603e-05, + 'epoch': 3.9} +04/19 [16:55:53] INFO | >> train_qwenlatent.py:487 + Step 15480 | grad_norm_pre_clip=0.2036 | + grad_norm_pre_clip_avg=0.2091 | Metrics: + {'align_loss': 0.025727039203047752, + 'recon_loss': 0.06881106644868851, + 'predict_loss': 0.015685083344578743, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2035708725452423, + 'data_time': 0.0005882929835934192, + 'model_time': 1.1452253549941815, + 'grad_norm_pre_clip_avg': 0.2090539291501045, + 'learning_rate': 2.180417881209482e-05, + 'epoch': 3.91} +04/19 [16:56:05] INFO | >> train_qwenlatent.py:487 + Step 15490 | grad_norm_pre_clip=0.1932 | + grad_norm_pre_clip_avg=0.2051 | Metrics: + {'align_loss': 0.024409445002675056, + 'recon_loss': 0.062292441725730896, + 'predict_loss': 0.01372084766626358, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19324137270450592, + 'data_time': 0.0006383129802998155, + 'model_time': 1.1968345600180328, + 'grad_norm_pre_clip_avg': 0.20512352138757706, + 'learning_rate': 2.1798352173027665e-05, + 'epoch': 3.91} +04/19 [16:56:17] INFO | >> train_qwenlatent.py:487 + Step 15500 | grad_norm_pre_clip=0.2565 | + grad_norm_pre_clip_avg=0.2072 | Metrics: + {'align_loss': 0.022531036287546158, + 'recon_loss': 0.042015109211206436, + 'predict_loss': 0.010826552286744118, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2565073072910309, + 'mae_score': 0.012860228564288164, 'data_time': + 0.0006530169921461493, 'model_time': + 1.1462509660050273, 'grad_norm_pre_clip_avg': + 0.20718875676393508, 'learning_rate': + 2.179252100814896e-05, 'epoch': 3.91} +04/19 [16:56:29] INFO | >> train_qwenlatent.py:487 + Step 15510 | grad_norm_pre_clip=0.2380 | + grad_norm_pre_clip_avg=0.2339 | Metrics: + {'align_loss': 0.02428046241402626, + 'recon_loss': 0.06623999774456024, + 'predict_loss': 0.015081890858709812, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23796911537647247, + 'data_time': 0.0005706890078727156, + 'model_time': 1.14767670200672, + 'grad_norm_pre_clip_avg': 0.23391037732362746, + 'learning_rate': 2.1786685320300754e-05, + 'epoch': 3.91} +04/19 [16:56:40] INFO | >> train_qwenlatent.py:487 + Step 15520 | grad_norm_pre_clip=0.2687 | + grad_norm_pre_clip_avg=0.2241 | Metrics: + {'align_loss': 0.024191780015826225, + 'recon_loss': 0.049229033291339874, + 'predict_loss': 0.011362524703145027, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26872649788856506, + 'data_time': 0.0005614639958366752, + 'model_time': 1.1342632199812215, + 'grad_norm_pre_clip_avg': 0.22405153065919875, + 'learning_rate': 2.1780845112327292e-05, + 'epoch': 3.92} +04/19 [16:56:52] INFO | >> train_qwenlatent.py:487 + Step 15530 | grad_norm_pre_clip=0.1873 | + grad_norm_pre_clip_avg=0.2272 | Metrics: + {'align_loss': 0.024710947647690773, + 'recon_loss': 0.049344226717948914, + 'predict_loss': 0.011323857121169567, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18734751641750336, + 'data_time': 0.0007674740045331419, + 'model_time': 1.1316697670263238, + 'grad_norm_pre_clip_avg': 0.2271721839904785, + 'learning_rate': 2.1775000387075006e-05, + 'epoch': 3.92} +04/19 [16:57:03] INFO | >> train_qwenlatent.py:487 + Step 15540 | grad_norm_pre_clip=0.2446 | + grad_norm_pre_clip_avg=0.2086 | Metrics: + {'align_loss': 0.025244176387786865, + 'recon_loss': 0.08641556650400162, + 'predict_loss': 0.01865418255329132, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2446361482143402, + 'data_time': 0.0005709709948860109, + 'model_time': 1.1415333299955819, + 'grad_norm_pre_clip_avg': 0.20863983482122422, + 'learning_rate': 2.1769151147392556e-05, + 'epoch': 3.92} +04/19 [16:57:50] INFO | >> train_qwenlatent.py:487 + Step 15550 | grad_norm_pre_clip=0.2568 | + grad_norm_pre_clip_avg=0.2661 | Metrics: + {'align_loss': 0.023823028430342674, + 'recon_loss': 0.043259404599666595, + 'predict_loss': 0.0115482397377491, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25675228238105774, + 'mae_score': 0.01550153096516927, 'data_time': + 0.001157360995421186, 'model_time': + 3.3288119779899716, 'grad_norm_pre_clip_avg': + 0.26605869829654694, 'learning_rate': + 2.176329739613079e-05, 'epoch': 3.92} +04/19 [16:58:24] INFO | >> train_qwenlatent.py:487 + Step 15560 | grad_norm_pre_clip=0.2420 | + grad_norm_pre_clip_avg=0.2224 | Metrics: + {'align_loss': 0.024317285045981407, + 'recon_loss': 0.045509222894907, + 'predict_loss': 0.014001107774674892, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24201306700706482, + 'data_time': 0.026595635019475594, + 'model_time': 3.9167807049816474, + 'grad_norm_pre_clip_avg': 0.2223947212100029, + 'learning_rate': 2.1757439136142755e-05, + 'epoch': 3.93} +04/19 [16:59:01] INFO | >> train_qwenlatent.py:487 + Step 15570 | grad_norm_pre_clip=0.2083 | + grad_norm_pre_clip_avg=0.2085 | Metrics: + {'align_loss': 0.024950392544269562, + 'recon_loss': 0.07158107310533524, + 'predict_loss': 0.0197185929864645, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20829281210899353, + 'data_time': 0.0011682500189635903, + 'model_time': 3.690679088991601, + 'grad_norm_pre_clip_avg': 0.20845545083284378, + 'learning_rate': 2.175157637028369e-05, + 'epoch': 3.93} +04/19 [16:59:36] INFO | >> train_qwenlatent.py:487 + Step 15580 | grad_norm_pre_clip=0.1816 | + grad_norm_pre_clip_avg=0.2216 | Metrics: + {'align_loss': 0.023834582418203354, + 'recon_loss': 0.0410924106836319, + 'predict_loss': 0.009424642659723759, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1816195249557495, + 'data_time': 0.0014187849883455783, + 'model_time': 3.229429890983738, + 'grad_norm_pre_clip_avg': 0.22159743458032607, + 'learning_rate': 2.1745709101411042e-05, + 'epoch': 3.93} +04/19 [17:00:12] INFO | >> train_qwenlatent.py:487 + Step 15590 | grad_norm_pre_clip=0.2238 | + grad_norm_pre_clip_avg=0.2218 | Metrics: + {'align_loss': 0.024600572884082794, + 'recon_loss': 0.07521692663431168, + 'predict_loss': 0.01861395128071308, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22378210723400116, + 'data_time': 0.0011580879800021648, + 'model_time': 3.365131585014751, + 'grad_norm_pre_clip_avg': 0.22180400639772416, + 'learning_rate': 2.1739837332384445e-05, + 'epoch': 3.93} +04/19 [17:00:50] INFO | >> train_qwenlatent.py:487 + Step 15600 | grad_norm_pre_clip=0.2505 | + grad_norm_pre_clip_avg=0.2239 | Metrics: + {'align_loss': 0.025455676019191742, + 'recon_loss': 0.061941541731357574, + 'predict_loss': 0.0123287383466959, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25045245885849, + 'mae_score': 0.01708201502894496, 'data_time': + 0.0010108730057254434, 'model_time': + 2.9555791699967813, 'grad_norm_pre_clip_avg': + 0.22386517524719238, 'learning_rate': + 2.1733961066065726e-05, 'epoch': 3.94} +04/19 [17:01:21] INFO | >> train_qwenlatent.py:487 + Step 15610 | grad_norm_pre_clip=0.1833 | + grad_norm_pre_clip_avg=0.1979 | Metrics: + {'align_loss': 0.026130426675081253, + 'recon_loss': 0.05660755932331085, + 'predict_loss': 0.012872479856014252, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18334390223026276, + 'data_time': 0.0017289490206167102, + 'model_time': 2.842296262009768, + 'grad_norm_pre_clip_avg': 0.19788508117198944, + 'learning_rate': 2.1728080305318906e-05, + 'epoch': 3.94} +04/19 [17:01:46] INFO | >> train_qwenlatent.py:487 + Step 15620 | grad_norm_pre_clip=0.2214 | + grad_norm_pre_clip_avg=0.2829 | Metrics: + {'align_loss': 0.025537442415952682, + 'recon_loss': 0.06794517487287521, + 'predict_loss': 0.01254352368414402, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2214127779006958, + 'data_time': 0.002521391987102106, + 'model_time': 2.4906391949916724, + 'grad_norm_pre_clip_avg': 0.28287808746099474, + 'learning_rate': 2.1722195053010198e-05, + 'epoch': 3.94} +04/19 [17:02:04] INFO | >> train_qwenlatent.py:487 + Step 15630 | grad_norm_pre_clip=0.2659 | + grad_norm_pre_clip_avg=0.2523 | Metrics: + {'align_loss': 0.025703484192490578, + 'recon_loss': 0.060882240533828735, + 'predict_loss': 0.007837679237127304, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26585760712623596, + 'data_time': 0.0016116419865284115, + 'model_time': 1.5861139330081642, + 'grad_norm_pre_clip_avg': 0.25232181549072263, + 'learning_rate': 2.1716305312008002e-05, + 'epoch': 3.94} +04/19 [17:02:18] INFO | >> train_qwenlatent.py:487 + Step 15640 | grad_norm_pre_clip=0.1867 | + grad_norm_pre_clip_avg=0.1946 | Metrics: + {'align_loss': 0.02416054531931877, + 'recon_loss': 0.0414435938000679, + 'predict_loss': 0.008549665100872517, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18669572472572327, + 'data_time': 0.0006820090056862682, + 'model_time': 1.2355932060163468, + 'grad_norm_pre_clip_avg': 0.19464392215013504, + 'learning_rate': 2.1710411085182907e-05, + 'epoch': 3.95} +04/19 [17:02:31] INFO | >> train_qwenlatent.py:487 + Step 15650 | grad_norm_pre_clip=0.2390 | + grad_norm_pre_clip_avg=0.2013 | Metrics: + {'align_loss': 0.025808164849877357, + 'recon_loss': 0.07496587187051773, + 'predict_loss': 0.019475916400551796, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2390068620443344, + 'mae_score': 0.013773620021235835, 'data_time': + 0.0005693249986506999, 'model_time': + 1.2336234059766866, 'grad_norm_pre_clip_avg': + 0.20130427330732345, 'learning_rate': + 2.1704512375407687e-05, 'epoch': 3.95} +04/19 [17:02:43] INFO | >> train_qwenlatent.py:487 + Step 15660 | grad_norm_pre_clip=0.2410 | + grad_norm_pre_clip_avg=0.2965 | Metrics: + {'align_loss': 0.02452094294130802, + 'recon_loss': 0.05767447128891945, + 'predict_loss': 0.012357868254184723, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24100632965564728, + 'data_time': 0.0008708089881110936, + 'model_time': 1.2626470749964938, + 'grad_norm_pre_clip_avg': 0.2965474396944046, + 'learning_rate': 2.16986091855573e-05, 'epoch': + 3.95} +04/19 [17:02:56] INFO | >> train_qwenlatent.py:487 + Step 15670 | grad_norm_pre_clip=0.1843 | + grad_norm_pre_clip_avg=0.2118 | Metrics: + {'align_loss': 0.024854669347405434, + 'recon_loss': 0.05394815281033516, + 'predict_loss': 0.010124594904482365, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1842966079711914, + 'data_time': 0.0008704829961061478, + 'model_time': 1.214689742977498, + 'grad_norm_pre_clip_avg': 0.21178347617387772, + 'learning_rate': 2.1692701518508887e-05, + 'epoch': 3.95} +04/19 [17:03:09] INFO | >> train_qwenlatent.py:487 + Step 15680 | grad_norm_pre_clip=0.1697 | + grad_norm_pre_clip_avg=0.1969 | Metrics: + {'align_loss': 0.02488967403769493, + 'recon_loss': 0.04725271835923195, + 'predict_loss': 0.013261460699141026, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16974575817584991, + 'data_time': 0.0005956850072834641, + 'model_time': 1.2326309810159728, + 'grad_norm_pre_clip_avg': 0.19686561971902847, + 'learning_rate': 2.168678937714178e-05, + 'epoch': 3.96} +04/19 [17:03:22] INFO | >> train_qwenlatent.py:487 + Step 15690 | grad_norm_pre_clip=0.2362 | + grad_norm_pre_clip_avg=0.2114 | Metrics: + {'align_loss': 0.025512468069791794, + 'recon_loss': 0.06948096305131912, + 'predict_loss': 0.013700027018785477, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2362041175365448, + 'data_time': 0.0009674850152805448, + 'model_time': 1.2816778130072635, + 'grad_norm_pre_clip_avg': 0.21135477870702743, + 'learning_rate': 2.1680872764337478e-05, + 'epoch': 3.96} +04/19 [17:03:35] INFO | >> train_qwenlatent.py:487 + Step 15700 | grad_norm_pre_clip=0.2357 | + grad_norm_pre_clip_avg=0.2177 | Metrics: + {'align_loss': 0.02569863572716713, + 'recon_loss': 0.06750839203596115, + 'predict_loss': 0.015658212825655937, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2357298880815506, + 'mae_score': 0.015267088606550887, 'data_time': + 0.0007270220085047185, 'model_time': + 1.2857889129954856, 'grad_norm_pre_clip_avg': + 0.2176935166120529, 'learning_rate': + 2.1674951682979668e-05, 'epoch': 3.96} +04/19 [17:03:47] INFO | >> train_qwenlatent.py:487 + Step 15710 | grad_norm_pre_clip=0.2726 | + grad_norm_pre_clip_avg=0.2443 | Metrics: + {'align_loss': 0.026338083669543266, + 'recon_loss': 0.0697929710149765, + 'predict_loss': 0.012926922179758549, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2726212441921234, + 'data_time': 0.000584057008381933, + 'model_time': 1.1984956410015002, + 'grad_norm_pre_clip_avg': 0.2443078190088272, + 'learning_rate': 2.166902613595422e-05, + 'epoch': 3.96} +04/19 [17:04:00] INFO | >> train_qwenlatent.py:487 + Step 15720 | grad_norm_pre_clip=0.1888 | + grad_norm_pre_clip_avg=0.2253 | Metrics: + {'align_loss': 0.02518003061413765, + 'recon_loss': 0.07860306650400162, + 'predict_loss': 0.009740264154970646, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18882909417152405, + 'data_time': 0.0007908999978099018, + 'model_time': 1.274985491996631, + 'grad_norm_pre_clip_avg': 0.22528840601444244, + 'learning_rate': 2.1663096126149164e-05, + 'epoch': 3.97} +04/19 [17:04:13] INFO | >> train_qwenlatent.py:487 + Step 15730 | grad_norm_pre_clip=0.1947 | + grad_norm_pre_clip_avg=0.2081 | Metrics: + {'align_loss': 0.02603287063539028, + 'recon_loss': 0.07130422443151474, + 'predict_loss': 0.012458107434213161, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1947346031665802, + 'data_time': 0.0007473340083379298, + 'model_time': 1.2765309620008338, + 'grad_norm_pre_clip_avg': 0.20811974108219147, + 'learning_rate': 2.1657161656454722e-05, + 'epoch': 3.97} +04/19 [17:04:25] INFO | >> train_qwenlatent.py:487 + Step 15740 | grad_norm_pre_clip=0.2795 | + grad_norm_pre_clip_avg=0.2093 | Metrics: + {'align_loss': 0.024271970614790916, + 'recon_loss': 0.060481373220682144, + 'predict_loss': 0.014041664078831673, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.279506653547287, + 'data_time': 0.0009977030276786536, + 'model_time': 1.2327986810123548, + 'grad_norm_pre_clip_avg': 0.20933835655450822, + 'learning_rate': 2.1651222729763276e-05, + 'epoch': 3.97} +04/19 [17:04:39] INFO | >> train_qwenlatent.py:487 + Step 15750 | grad_norm_pre_clip=0.3163 | + grad_norm_pre_clip_avg=0.2968 | Metrics: + {'align_loss': 0.02535838633775711, + 'recon_loss': 0.06476074457168579, + 'predict_loss': 0.012744361534714699, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3163120448589325, + 'mae_score': 0.014765215349626971, 'data_time': + 0.0006044080073479563, 'model_time': + 1.2462548150215298, 'grad_norm_pre_clip_avg': + 0.29676341712474824, 'learning_rate': + 2.1645279348969392e-05, 'epoch': 3.97} +04/19 [17:04:51] INFO | >> train_qwenlatent.py:487 + Step 15760 | grad_norm_pre_clip=0.2183 | + grad_norm_pre_clip_avg=0.2362 | Metrics: + {'align_loss': 0.024092180654406548, + 'recon_loss': 0.07740518450737, 'predict_loss': + 0.011736506596207619, 'aux_loss_decay_weight': + 0.0, 'grad_norm_pre_clip': 0.21833008527755737, + 'data_time': 0.0006143570062704384, + 'model_time': 1.214127423998434, + 'grad_norm_pre_clip_avg': 0.23616537898778917, + 'learning_rate': 2.16393315169698e-05, 'epoch': + 3.98} +04/19 [17:05:04] INFO | >> train_qwenlatent.py:487 + Step 15770 | grad_norm_pre_clip=0.1999 | + grad_norm_pre_clip_avg=0.2120 | Metrics: + {'align_loss': 0.025061745196580887, + 'recon_loss': 0.05436005815863609, + 'predict_loss': 0.013312987051904202, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19994501769542694, + 'data_time': 0.0009197170147672296, + 'model_time': 1.2724798030103557, + 'grad_norm_pre_clip_avg': 0.21195834130048752, + 'learning_rate': 2.1633379236663402e-05, + 'epoch': 3.98} +04/19 [17:05:16] INFO | >> train_qwenlatent.py:487 + Step 15780 | grad_norm_pre_clip=0.2351 | + grad_norm_pre_clip_avg=0.2220 | Metrics: + {'align_loss': 0.025414999574422836, + 'recon_loss': 0.046245038509368896, + 'predict_loss': 0.00944945402443409, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23508763313293457, + 'data_time': 0.0007592809852212667, + 'model_time': 1.2277457610180136, + 'grad_norm_pre_clip_avg': 0.22201173156499862, + 'learning_rate': 2.1627422510951262e-05, + 'epoch': 3.98} +04/19 [17:05:29] INFO | >> train_qwenlatent.py:487 + Step 15790 | grad_norm_pre_clip=0.3107 | + grad_norm_pre_clip_avg=0.2288 | Metrics: + {'align_loss': 0.025085914880037308, + 'recon_loss': 0.0770939439535141, + 'predict_loss': 0.016688862815499306, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.31070441007614136, + 'data_time': 0.0007821260078344494, + 'model_time': 1.26585612501367, + 'grad_norm_pre_clip_avg': 0.22875450253486634, + 'learning_rate': 2.162146134273662e-05, + 'epoch': 3.98} +04/19 [17:05:42] INFO | >> train_qwenlatent.py:487 + Step 15800 | grad_norm_pre_clip=0.1866 | + grad_norm_pre_clip_avg=0.1998 | Metrics: + {'align_loss': 0.023810232058167458, + 'recon_loss': 0.04885674640536308, + 'predict_loss': 0.011518433690071106, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18662627041339874, + 'mae_score': 0.01792126045570717, 'data_time': + 0.0006534520070999861, 'model_time': + 1.267749823979102, 'grad_norm_pre_clip_avg': + 0.19976818263530732, 'learning_rate': + 2.161549573492488e-05, 'epoch': 3.99} +04/19 [17:05:55] INFO | >> train_qwenlatent.py:487 + Step 15810 | grad_norm_pre_clip=0.2303 | + grad_norm_pre_clip_avg=0.2251 | Metrics: + {'align_loss': 0.02506696805357933, + 'recon_loss': 0.06178954616189003, + 'predict_loss': 0.013231189921498299, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23033839464187622, + 'data_time': 0.0007525999972131103, + 'model_time': 1.2338367159827612, + 'grad_norm_pre_clip_avg': 0.22510021626949311, + 'learning_rate': 2.1609525690423598e-05, + 'epoch': 3.99} +04/19 [17:06:07] INFO | >> train_qwenlatent.py:487 + Step 15820 | grad_norm_pre_clip=0.2796 | + grad_norm_pre_clip_avg=0.2137 | Metrics: + {'align_loss': 0.02573338896036148, + 'recon_loss': 0.054853495210409164, + 'predict_loss': 0.012468847446143627, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2795916497707367, + 'data_time': 0.0008129980124067515, + 'model_time': 1.2042997630196624, + 'grad_norm_pre_clip_avg': 0.2137308433651924, + 'learning_rate': 2.160355121214251e-05, + 'epoch': 3.99} +04/19 [17:06:20] INFO | >> train_qwenlatent.py:487 + Step 15830 | grad_norm_pre_clip=0.2054 | + grad_norm_pre_clip_avg=0.2385 | Metrics: + {'align_loss': 0.026244617998600006, + 'recon_loss': 0.06839561462402344, + 'predict_loss': 0.022555246949195862, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.205413356423378, + 'data_time': 0.0008852139872033149, + 'model_time': 1.2352054650255013, + 'grad_norm_pre_clip_avg': 0.23846705853939057, + 'learning_rate': 2.15975723029935e-05, 'epoch': + 3.99} +04/19 [17:06:32] INFO | >> train_qwenlatent.py:487 + Step 15840 | grad_norm_pre_clip=0.2127 | + grad_norm_pre_clip_avg=0.1963 | Metrics: + {'align_loss': 0.02538505382835865, + 'recon_loss': 0.043117836117744446, + 'predict_loss': 0.009055716916918755, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2126944214105606, + 'data_time': 0.0011915770010091364, + 'model_time': 1.2327930150204338, + 'grad_norm_pre_clip_avg': 0.19625339210033416, + 'learning_rate': 2.1591588965890615e-05, + 'epoch': 4.0} +04/19 [17:06:46] INFO | >> train_qwenlatent.py:487 + Step 15850 | grad_norm_pre_clip=0.3205 | + grad_norm_pre_clip_avg=0.2180 | Metrics: + {'align_loss': 0.025182968005537987, + 'recon_loss': 0.04401547089219093, + 'predict_loss': 0.00923264678567648, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.32047343254089355, + 'mae_score': 0.01598154016443201, 'data_time': + 0.0005463309935294092, 'model_time': + 1.212485183001263, 'grad_norm_pre_clip_avg': + 0.21795970797538758, 'learning_rate': + 2.1585601203750064e-05, 'epoch': 4.0} +04/19 [17:06:58] INFO | >> train_qwenlatent.py:487 + Step 15860 | grad_norm_pre_clip=0.2652 | + grad_norm_pre_clip_avg=0.2455 | Metrics: + {'align_loss': 0.024697987362742424, + 'recon_loss': 0.05810019001364708, + 'predict_loss': 0.00944733526557684, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26519617438316345, + 'data_time': 0.0008122980070766062, + 'model_time': 1.2697362570033874, + 'grad_norm_pre_clip_avg': 0.2454979971051216, + 'learning_rate': 2.1579609019490205e-05, + 'epoch': 4.0} +04/19 [17:07:11] INFO | >> train_qwenlatent.py:487 + Step 15870 | grad_norm_pre_clip=0.2449 | + grad_norm_pre_clip_avg=0.2217 | Metrics: + {'align_loss': 0.024954555556178093, + 'recon_loss': 0.056605443358421326, + 'predict_loss': 0.015826912596821785, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24494986236095428, + 'data_time': 0.000778134010033682, + 'model_time': 1.2723214620200451, + 'grad_norm_pre_clip_avg': 0.2217118814587593, + 'learning_rate': 2.157361241603156e-05, + 'epoch': 4.0} +04/19 [17:07:23] INFO | >> train_qwenlatent.py:487 + Step 15880 | grad_norm_pre_clip=0.2052 | + grad_norm_pre_clip_avg=0.2169 | Metrics: + {'align_loss': 0.022655438631772995, + 'recon_loss': 0.03632747381925583, + 'predict_loss': 0.00931578315794468, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2052081823348999, + 'data_time': 0.0006387839966919273, + 'model_time': 1.5373028329922818, + 'grad_norm_pre_clip_avg': 0.21689520925283431, + 'learning_rate': 2.1567611396296795e-05, + 'epoch': 4.01} +04/19 [17:07:36] INFO | >> train_qwenlatent.py:487 + Step 15890 | grad_norm_pre_clip=0.2918 | + grad_norm_pre_clip_avg=0.2492 | Metrics: + {'align_loss': 0.024427007883787155, + 'recon_loss': 0.049702923744916916, + 'predict_loss': 0.007608362473547459, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2917744517326355, + 'data_time': 0.000882601976627484, + 'model_time': 1.237293250975199, + 'grad_norm_pre_clip_avg': 0.24922190308570863, + 'learning_rate': 2.1561605963210747e-05, + 'epoch': 4.01} +04/19 [17:07:49] INFO | >> train_qwenlatent.py:487 + Step 15900 | grad_norm_pre_clip=0.1929 | + grad_norm_pre_clip_avg=0.2345 | Metrics: + {'align_loss': 0.024058982729911804, + 'recon_loss': 0.0617448166012764, + 'predict_loss': 0.014086266979575157, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19285711646080017, + 'mae_score': 0.01795102884103586, 'data_time': + 0.0006465240148827434, 'model_time': + 1.24272637499962, 'grad_norm_pre_clip_avg': + 0.2344801589846611, 'learning_rate': + 2.1555596119700382e-05, 'epoch': 4.01} +04/19 [17:08:02] INFO | >> train_qwenlatent.py:487 + Step 15910 | grad_norm_pre_clip=0.1979 | + grad_norm_pre_clip_avg=0.2143 | Metrics: + {'align_loss': 0.0251784510910511, + 'recon_loss': 0.05302036926150322, + 'predict_loss': 0.009096530266106129, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19785112142562866, + 'data_time': 0.000881223997566849, + 'model_time': 1.2361147749761585, + 'grad_norm_pre_clip_avg': 0.21427152156829835, + 'learning_rate': 2.1549581868694815e-05, + 'epoch': 4.01} +04/19 [17:08:14] INFO | >> train_qwenlatent.py:487 + Step 15920 | grad_norm_pre_clip=0.2536 | + grad_norm_pre_clip_avg=0.2310 | Metrics: + {'align_loss': 0.02441548928618431, + 'recon_loss': 0.053822848945856094, + 'predict_loss': 0.011481006629765034, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25358062982559204, + 'data_time': 0.0005585089966189116, + 'model_time': 1.2353899480076507, + 'grad_norm_pre_clip_avg': 0.2309926077723503, + 'learning_rate': 2.1543563213125335e-05, + 'epoch': 4.02} +04/19 [17:08:27] INFO | >> train_qwenlatent.py:487 + Step 15930 | grad_norm_pre_clip=0.1611 | + grad_norm_pre_clip_avg=0.1899 | Metrics: + {'align_loss': 0.025893695652484894, + 'recon_loss': 0.0685453861951828, + 'predict_loss': 0.013043900951743126, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1610950380563736, + 'data_time': 0.0005563860177062452, + 'model_time': 1.2128234620031435, + 'grad_norm_pre_clip_avg': 0.18992362767457963, + 'learning_rate': 2.1537540155925353e-05, + 'epoch': 4.02} +04/19 [17:08:39] INFO | >> train_qwenlatent.py:487 + Step 15940 | grad_norm_pre_clip=0.2387 | + grad_norm_pre_clip_avg=0.2098 | Metrics: + {'align_loss': 0.02480863407254219, + 'recon_loss': 0.0507500134408474, + 'predict_loss': 0.01454161573201418, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23871555924415588, + 'data_time': 0.0005654389970004559, + 'model_time': 1.2452082709933165, + 'grad_norm_pre_clip_avg': 0.20982896238565446, + 'learning_rate': 2.153151270003044e-05, + 'epoch': 4.02} +04/19 [17:08:53] INFO | >> train_qwenlatent.py:487 + Step 15950 | grad_norm_pre_clip=0.2112 | + grad_norm_pre_clip_avg=0.2459 | Metrics: + {'align_loss': 0.02534669078886509, + 'recon_loss': 0.059097547084093094, + 'predict_loss': 0.009493235498666763, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2111954540014267, + 'mae_score': 0.013571179879678262, 'data_time': + 0.0006852639780845493, 'model_time': + 1.2360505659889895, 'grad_norm_pre_clip_avg': + 0.2458644226193428, 'learning_rate': + 2.1525480848378298e-05, 'epoch': 4.02} +04/19 [17:09:06] INFO | >> train_qwenlatent.py:487 + Step 15960 | grad_norm_pre_clip=0.1760 | + grad_norm_pre_clip_avg=0.2080 | Metrics: + {'align_loss': 0.02545216865837574, + 'recon_loss': 0.057327620685100555, + 'predict_loss': 0.013565338216722012, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1760302484035492, + 'data_time': 0.0006221279909368604, + 'model_time': 1.2215530829853378, + 'grad_norm_pre_clip_avg': 0.20804976522922516, + 'learning_rate': 2.1519444603908785e-05, + 'epoch': 4.03} +04/19 [17:09:18] INFO | >> train_qwenlatent.py:487 + Step 15970 | grad_norm_pre_clip=0.4280 | + grad_norm_pre_clip_avg=0.2397 | Metrics: + {'align_loss': 0.02463892102241516, + 'recon_loss': 0.050374068319797516, + 'predict_loss': 0.012469823472201824, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.42796751856803894, + 'data_time': 0.0007818060112185776, + 'model_time': 1.2437629509950057, + 'grad_norm_pre_clip_avg': 0.23971446901559829, + 'learning_rate': 2.1513403969563888e-05, + 'epoch': 4.03} +04/19 [17:09:30] INFO | >> train_qwenlatent.py:487 + Step 15980 | grad_norm_pre_clip=0.2297 | + grad_norm_pre_clip_avg=0.2383 | Metrics: + {'align_loss': 0.025379065424203873, + 'recon_loss': 0.05782373249530792, + 'predict_loss': 0.012378348968923092, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22966669499874115, + 'data_time': 0.0008313060097862035, + 'model_time': 1.2608951250149403, + 'grad_norm_pre_clip_avg': 0.2382950156927109, + 'learning_rate': 2.150735894828774e-05, + 'epoch': 4.03} +04/19 [17:09:43] INFO | >> train_qwenlatent.py:487 + Step 15990 | grad_norm_pre_clip=0.2034 | + grad_norm_pre_clip_avg=0.1800 | Metrics: + {'align_loss': 0.02563304826617241, + 'recon_loss': 0.06173568591475487, + 'predict_loss': 0.013718914240598679, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20341916382312775, + 'data_time': 0.0006186449900269508, + 'model_time': 1.2037505049956962, + 'grad_norm_pre_clip_avg': 0.1800075128674507, + 'learning_rate': 2.150130954302661e-05, + 'epoch': 4.03} +04/19 [17:09:56] INFO | >> train_qwenlatent.py:487 + Step 16000 | grad_norm_pre_clip=0.1833 | + grad_norm_pre_clip_avg=0.1753 | Metrics: + {'align_loss': 0.025825755670666695, + 'recon_loss': 0.05857495963573456, + 'predict_loss': 0.011963378638029099, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18330152332782745, + 'mae_score': 0.01253742011817726, 'data_time': + 0.0006769320170860738, 'model_time': + 1.2646563010057434, 'grad_norm_pre_clip_avg': + 0.1753056228160858, 'learning_rate': + 2.149525575672891e-05, 'epoch': 4.04} +04/19 [17:10:08] INFO | >> train_qwenlatent.py:487 + Step 16010 | grad_norm_pre_clip=0.1986 | + grad_norm_pre_clip_avg=0.2695 | Metrics: + {'align_loss': 0.025091679766774178, + 'recon_loss': 0.060214996337890625, + 'predict_loss': 0.013031664304435253, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19858530163764954, + 'data_time': 0.0008728969842195511, + 'model_time': 1.2227007010078523, + 'grad_norm_pre_clip_avg': 0.2694897621870041, + 'learning_rate': 2.1489197592345172e-05, + 'epoch': 4.04} +04/19 [17:10:21] INFO | >> train_qwenlatent.py:487 + Step 16020 | grad_norm_pre_clip=0.3074 | + grad_norm_pre_clip_avg=0.2449 | Metrics: + {'align_loss': 0.02482631430029869, + 'recon_loss': 0.05489382520318031, + 'predict_loss': 0.012846572324633598, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3073728680610657, + 'data_time': 0.0006725220009684563, + 'model_time': 1.2464924820233136, + 'grad_norm_pre_clip_avg': 0.24487305134534837, + 'learning_rate': 2.1483135052828083e-05, + 'epoch': 4.04} +04/19 [17:10:34] INFO | >> train_qwenlatent.py:487 + Step 16030 | grad_norm_pre_clip=0.2036 | + grad_norm_pre_clip_avg=0.2327 | Metrics: + {'align_loss': 0.02567092329263687, + 'recon_loss': 0.0739392638206482, + 'predict_loss': 0.009201513603329659, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.203593447804451, + 'data_time': 0.0007608740124851465, + 'model_time': 1.2354267660120968, + 'grad_norm_pre_clip_avg': 0.23274415731430054, + 'learning_rate': 2.147706814113244e-05, + 'epoch': 4.04} +04/19 [17:10:47] INFO | >> train_qwenlatent.py:487 + Step 16040 | grad_norm_pre_clip=0.1748 | + grad_norm_pre_clip_avg=0.1955 | Metrics: + {'align_loss': 0.025299249216914177, + 'recon_loss': 0.055721014738082886, + 'predict_loss': 0.01273267436772585, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17478398978710175, + 'data_time': 0.0005615239788312465, + 'model_time': 1.2152644910092931, + 'grad_norm_pre_clip_avg': 0.1954885706305504, + 'learning_rate': 2.1470996860215196e-05, + 'epoch': 4.05} +04/19 [17:11:00] INFO | >> train_qwenlatent.py:487 + Step 16050 | grad_norm_pre_clip=0.1815 | + grad_norm_pre_clip_avg=0.2119 | Metrics: + {'align_loss': 0.02620522491633892, + 'recon_loss': 0.08138250559568405, + 'predict_loss': 0.014526872895658016, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18147973716259003, + 'mae_score': 0.014258476635357282, 'data_time': + 0.0008618400024715811, 'model_time': + 1.2445800699933898, 'grad_norm_pre_clip_avg': + 0.21193154007196427, 'learning_rate': + 2.1464921213035406e-05, 'epoch': 4.05} +04/19 [17:11:13] INFO | >> train_qwenlatent.py:487 + Step 16060 | grad_norm_pre_clip=0.1686 | + grad_norm_pre_clip_avg=0.2243 | Metrics: + {'align_loss': 0.02457580156624317, + 'recon_loss': 0.05735012888908386, + 'predict_loss': 0.009577094577252865, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16857023537158966, + 'data_time': 0.0006558939930982888, + 'model_time': 1.3269885150075424, + 'grad_norm_pre_clip_avg': 0.2242689922451973, + 'learning_rate': 2.145884120255427e-05, + 'epoch': 4.05} +04/19 [17:11:26] INFO | >> train_qwenlatent.py:487 + Step 16070 | grad_norm_pre_clip=0.1561 | + grad_norm_pre_clip_avg=0.1959 | Metrics: + {'align_loss': 0.022707074880599976, + 'recon_loss': 0.047370485961437225, + 'predict_loss': 0.00900648720562458, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1561133861541748, + 'data_time': 0.0007204629946500063, + 'model_time': 1.2314495529863052, + 'grad_norm_pre_clip_avg': 0.19593116044998168, + 'learning_rate': 2.1452756831735114e-05, + 'epoch': 4.06} +04/19 [17:11:38] INFO | >> train_qwenlatent.py:487 + Step 16080 | grad_norm_pre_clip=0.3301 | + grad_norm_pre_clip_avg=0.2594 | Metrics: + {'align_loss': 0.026150817051529884, + 'recon_loss': 0.07054182142019272, + 'predict_loss': 0.00974066462367773, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3300885558128357, + 'data_time': 0.0007534729957114905, + 'model_time': 1.2613201680069324, + 'grad_norm_pre_clip_avg': 0.2594339162111282, + 'learning_rate': 2.144666810354339e-05, + 'epoch': 4.06} +04/19 [17:11:51] INFO | >> train_qwenlatent.py:487 + Step 16090 | grad_norm_pre_clip=0.2424 | + grad_norm_pre_clip_avg=0.2559 | Metrics: + {'align_loss': 0.024559251964092255, + 'recon_loss': 0.07481226325035095, + 'predict_loss': 0.011942524462938309, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24240826070308685, + 'data_time': 0.0008197120041586459, + 'model_time': 1.4411603279877454, + 'grad_norm_pre_clip_avg': 0.2558720946311951, + 'learning_rate': 2.1440575020946665e-05, + 'epoch': 4.06} +04/19 [17:12:04] INFO | >> train_qwenlatent.py:487 + Step 16100 | grad_norm_pre_clip=0.1465 | + grad_norm_pre_clip_avg=0.2154 | Metrics: + {'align_loss': 0.025034988299012184, + 'recon_loss': 0.04940027743577957, + 'predict_loss': 0.011642650701105595, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14650177955627441, + 'mae_score': 0.021269415090749928, 'data_time': + 0.000883394997799769, 'model_time': + 1.3167881009867415, 'grad_norm_pre_clip_avg': + 0.2154393747448921, 'learning_rate': + 2.1434477586914635e-05, 'epoch': 4.06} +04/19 [17:12:17] INFO | >> train_qwenlatent.py:487 + Step 16110 | grad_norm_pre_clip=0.1793 | + grad_norm_pre_clip_avg=0.1986 | Metrics: + {'align_loss': 0.02528703585267067, + 'recon_loss': 0.05916915833950043, + 'predict_loss': 0.012374688871204853, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1793443113565445, + 'data_time': 0.000815539009636268, + 'model_time': 1.2409642220009118, + 'grad_norm_pre_clip_avg': 0.1985839083790779, + 'learning_rate': 2.1428375804419114e-05, + 'epoch': 4.07} +04/19 [17:12:29] INFO | >> train_qwenlatent.py:487 + Step 16120 | grad_norm_pre_clip=0.1769 | + grad_norm_pre_clip_avg=0.1904 | Metrics: + {'align_loss': 0.0244801864027977, + 'recon_loss': 0.039550937712192535, + 'predict_loss': 0.007774346508085728, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1768932044506073, + 'data_time': 0.0008742850041016936, + 'model_time': 1.2711752530012745, + 'grad_norm_pre_clip_avg': 0.19042421132326126, + 'learning_rate': 2.1422269676434032e-05, + 'epoch': 4.07} +04/19 [17:12:42] INFO | >> train_qwenlatent.py:487 + Step 16130 | grad_norm_pre_clip=0.3363 | + grad_norm_pre_clip_avg=0.2442 | Metrics: + {'align_loss': 0.024798737838864326, + 'recon_loss': 0.05969427525997162, + 'predict_loss': 0.012742781080305576, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3362855017185211, + 'data_time': 0.000609538983553648, + 'model_time': 1.2159092239744496, + 'grad_norm_pre_clip_avg': 0.24415513575077058, + 'learning_rate': 2.1416159205935452e-05, + 'epoch': 4.07} +04/19 [17:12:55] INFO | >> train_qwenlatent.py:487 + Step 16140 | grad_norm_pre_clip=0.2423 | + grad_norm_pre_clip_avg=0.2729 | Metrics: + {'align_loss': 0.024917639791965485, + 'recon_loss': 0.08021607249975204, + 'predict_loss': 0.015333917923271656, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24225376546382904, + 'data_time': 0.000668573979055509, + 'model_time': 1.2189430909929797, + 'grad_norm_pre_clip_avg': 0.2729006350040436, + 'learning_rate': 2.1410044395901536e-05, + 'epoch': 4.07} +04/19 [17:13:08] INFO | >> train_qwenlatent.py:487 + Step 16150 | grad_norm_pre_clip=0.2177 | + grad_norm_pre_clip_avg=0.2012 | Metrics: + {'align_loss': 0.02561349980533123, + 'recon_loss': 0.06599827855825424, + 'predict_loss': 0.016683991998434067, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21765002608299255, + 'mae_score': 0.012767140500180357, 'data_time': + 0.0010626699950080365, 'model_time': + 1.2106090760207735, 'grad_norm_pre_clip_avg': + 0.20116829574108125, 'learning_rate': + 2.140392524931257e-05, 'epoch': 4.08} +04/19 [17:13:21] INFO | >> train_qwenlatent.py:487 + Step 16160 | grad_norm_pre_clip=0.1912 | + grad_norm_pre_clip_avg=0.1871 | Metrics: + {'align_loss': 0.024306127801537514, + 'recon_loss': 0.05617652088403702, + 'predict_loss': 0.011534404940903187, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19121253490447998, + 'data_time': 0.0007523789827246219, + 'model_time': 1.2511640900047496, + 'grad_norm_pre_clip_avg': 0.1871018722653389, + 'learning_rate': 2.1397801769150954e-05, + 'epoch': 4.08} +04/19 [17:13:33] INFO | >> train_qwenlatent.py:487 + Step 16170 | grad_norm_pre_clip=0.2085 | + grad_norm_pre_clip_avg=0.2261 | Metrics: + {'align_loss': 0.02586737647652626, + 'recon_loss': 0.06516600400209427, + 'predict_loss': 0.012187240645289421, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20849575102329254, + 'data_time': 0.0007999230001587421, + 'model_time': 1.2608922270010225, + 'grad_norm_pre_clip_avg': 0.22607502490282058, + 'learning_rate': 2.139167395840119e-05, + 'epoch': 4.08} +04/19 [17:13:46] INFO | >> train_qwenlatent.py:487 + Step 16180 | grad_norm_pre_clip=0.1822 | + grad_norm_pre_clip_avg=0.2128 | Metrics: + {'align_loss': 0.025737715885043144, + 'recon_loss': 0.06010565161705017, + 'predict_loss': 0.01066621858626604, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18217281997203827, + 'data_time': 0.0007154140039347112, + 'model_time': 1.4618143160187174, + 'grad_norm_pre_clip_avg': 0.21277279555797576, + 'learning_rate': 2.138554182004991e-05, + 'epoch': 4.08} +04/19 [17:13:58] INFO | >> train_qwenlatent.py:487 + Step 16190 | grad_norm_pre_clip=0.1911 | + grad_norm_pre_clip_avg=0.2195 | Metrics: + {'align_loss': 0.024816226214170456, + 'recon_loss': 0.06346260756254196, + 'predict_loss': 0.01721513643860817, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19107097387313843, + 'data_time': 0.0015102670004125684, + 'model_time': 1.271269602002576, + 'grad_norm_pre_clip_avg': 0.21947031319141388, + 'learning_rate': 2.1379405357085835e-05, + 'epoch': 4.09} +04/19 [17:14:11] INFO | >> train_qwenlatent.py:487 + Step 16200 | grad_norm_pre_clip=0.2802 | + grad_norm_pre_clip_avg=0.2055 | Metrics: + {'align_loss': 0.024651162326335907, + 'recon_loss': 0.052009083330631256, + 'predict_loss': 0.010709624737501144, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2802305519580841, + 'mae_score': 0.021103805679458756, 'data_time': + 0.0011006019776687026, 'model_time': + 1.2713215969852172, 'grad_norm_pre_clip_avg': + 0.20552454143762589, 'learning_rate': + 2.1373264572499807e-05, 'epoch': 4.09} +04/19 [17:14:24] INFO | >> train_qwenlatent.py:487 + Step 16210 | grad_norm_pre_clip=0.1749 | + grad_norm_pre_clip_avg=0.2520 | Metrics: + {'align_loss': 0.025378528982400894, + 'recon_loss': 0.06263943016529083, + 'predict_loss': 0.015965942293405533, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17486494779586792, + 'data_time': 0.0006501110037788749, + 'model_time': 1.2143038859940134, + 'grad_norm_pre_clip_avg': 0.2520249769091606, + 'learning_rate': 2.1367119469284767e-05, + 'epoch': 4.09} +04/19 [17:14:37] INFO | >> train_qwenlatent.py:487 + Step 16220 | grad_norm_pre_clip=0.2665 | + grad_norm_pre_clip_avg=0.2588 | Metrics: + {'align_loss': 0.025040224194526672, + 'recon_loss': 0.06703460961580276, + 'predict_loss': 0.01570090465247631, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2664661705493927, + 'data_time': 0.000871363008627668, + 'model_time': 1.2670881939993706, + 'grad_norm_pre_clip_avg': 0.2588493153452873, + 'learning_rate': 2.1360970050435766e-05, + 'epoch': 4.09} +04/19 [17:14:49] INFO | >> train_qwenlatent.py:487 + Step 16230 | grad_norm_pre_clip=0.1981 | + grad_norm_pre_clip_avg=0.2151 | Metrics: + {'align_loss': 0.02400379627943039, + 'recon_loss': 0.06452696025371552, + 'predict_loss': 0.016844825819134712, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19811122119426727, + 'data_time': 0.0006754350033588707, + 'model_time': 1.2183736379956827, + 'grad_norm_pre_clip_avg': 0.21511940509080887, + 'learning_rate': 2.1354816318949953e-05, + 'epoch': 4.1} +04/19 [17:15:02] INFO | >> train_qwenlatent.py:487 + Step 16240 | grad_norm_pre_clip=0.2072 | + grad_norm_pre_clip_avg=0.2150 | Metrics: + {'align_loss': 0.026310238987207413, + 'recon_loss': 0.07736052572727203, + 'predict_loss': 0.018347838893532753, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20719684660434723, + 'data_time': 0.0010512399894651026, + 'model_time': 1.2099304700095672, + 'grad_norm_pre_clip_avg': 0.21504545211791992, + 'learning_rate': 2.1348658277826586e-05, + 'epoch': 4.1} +04/19 [17:15:15] INFO | >> train_qwenlatent.py:487 + Step 16250 | grad_norm_pre_clip=0.2018 | + grad_norm_pre_clip_avg=0.2190 | Metrics: + {'align_loss': 0.024361692368984222, + 'recon_loss': 0.05930270627140999, + 'predict_loss': 0.011739281006157398, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20179399847984314, + 'mae_score': 0.015791332829105963, 'data_time': + 0.0009362529963254929, 'model_time': + 1.2401552890078165, 'grad_norm_pre_clip_avg': + 0.21900516897439956, 'learning_rate': + 2.1342495930067015e-05, 'epoch': 4.1} +04/19 [17:15:28] INFO | >> train_qwenlatent.py:487 + Step 16260 | grad_norm_pre_clip=0.2032 | + grad_norm_pre_clip_avg=0.2164 | Metrics: + {'align_loss': 0.025634903460741043, + 'recon_loss': 0.04948645085096359, + 'predict_loss': 0.012082778848707676, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20318348705768585, + 'data_time': 0.0010976589983329177, + 'model_time': 1.1680592129996512, + 'grad_norm_pre_clip_avg': 0.216377155482769, + 'learning_rate': 2.1336329278674697e-05, + 'epoch': 4.1} +04/19 [17:15:40] INFO | >> train_qwenlatent.py:487 + Step 16270 | grad_norm_pre_clip=0.2274 | + grad_norm_pre_clip_avg=0.2449 | Metrics: + {'align_loss': 0.024606630206108093, + 'recon_loss': 0.06999840587377548, + 'predict_loss': 0.01752939075231552, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22737626731395721, + 'data_time': 0.000991117994999513, + 'model_time': 1.2526595249946695, + 'grad_norm_pre_clip_avg': 0.24489369690418245, + 'learning_rate': 2.133015832665518e-05, + 'epoch': 4.11} +04/19 [17:15:52] INFO | >> train_qwenlatent.py:487 + Step 16280 | grad_norm_pre_clip=0.1789 | + grad_norm_pre_clip_avg=0.1847 | Metrics: + {'align_loss': 0.023775454610586166, + 'recon_loss': 0.05220060050487518, + 'predict_loss': 0.016337387263774872, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17894546687602997, + 'data_time': 0.0006673900061286986, + 'model_time': 1.2591300850035623, + 'grad_norm_pre_clip_avg': 0.1846780523657799, + 'learning_rate': 2.1323983077016116e-05, + 'epoch': 4.11} +04/19 [17:16:05] INFO | >> train_qwenlatent.py:487 + Step 16290 | grad_norm_pre_clip=0.1945 | + grad_norm_pre_clip_avg=0.1887 | Metrics: + {'align_loss': 0.026149172335863113, + 'recon_loss': 0.06994420289993286, + 'predict_loss': 0.014637294225394726, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19454625248908997, + 'data_time': 0.0007713800005149096, + 'model_time': 1.2396628560090903, + 'grad_norm_pre_clip_avg': 0.18869781196117402, + 'learning_rate': 2.131780353276724e-05, + 'epoch': 4.11} +04/19 [17:16:18] INFO | >> train_qwenlatent.py:487 + Step 16300 | grad_norm_pre_clip=0.2353 | + grad_norm_pre_clip_avg=0.3559 | Metrics: + {'align_loss': 0.024430911988019943, + 'recon_loss': 0.06189354136586189, + 'predict_loss': 0.014061150141060352, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23530049622058868, + 'mae_score': 0.01248458312438415, 'data_time': + 0.0010442160128150135, 'model_time': + 1.2933613980130758, 'grad_norm_pre_clip_avg': + 0.3559033513069153, 'learning_rate': + 2.131161969692039e-05, 'epoch': 4.11} +04/19 [17:16:31] INFO | >> train_qwenlatent.py:487 + Step 16310 | grad_norm_pre_clip=0.2304 | + grad_norm_pre_clip_avg=0.2416 | Metrics: + {'align_loss': 0.02425112947821617, + 'recon_loss': 0.061508096754550934, + 'predict_loss': 0.01126124244183302, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23035594820976257, + 'data_time': 0.0006164009973872453, + 'model_time': 1.211549927975284, + 'grad_norm_pre_clip_avg': 0.24163273274898528, + 'learning_rate': 2.1305431572489494e-05, + 'epoch': 4.12} +04/19 [17:16:44] INFO | >> train_qwenlatent.py:487 + Step 16320 | grad_norm_pre_clip=0.2092 | + grad_norm_pre_clip_avg=0.1910 | Metrics: + {'align_loss': 0.024917494505643845, + 'recon_loss': 0.050860002636909485, + 'predict_loss': 0.011662519536912441, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20920242369174957, + 'data_time': 0.0008639070147182792, + 'model_time': 1.259194509999361, + 'grad_norm_pre_clip_avg': 0.19098224937915803, + 'learning_rate': 2.1299239162490566e-05, + 'epoch': 4.12} +04/19 [17:16:57] INFO | >> train_qwenlatent.py:487 + Step 16330 | grad_norm_pre_clip=0.2029 | + grad_norm_pre_clip_avg=0.1959 | Metrics: + {'align_loss': 0.025856222957372665, + 'recon_loss': 0.055187635123729706, + 'predict_loss': 0.013384341262280941, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20286668837070465, + 'data_time': 0.0009686110133770853, + 'model_time': 1.2243593969906215, + 'grad_norm_pre_clip_avg': 0.1959354594349861, + 'learning_rate': 2.1293042469941715e-05, + 'epoch': 4.12} +04/19 [17:17:10] INFO | >> train_qwenlatent.py:487 + Step 16340 | grad_norm_pre_clip=0.1848 | + grad_norm_pre_clip_avg=0.2116 | Metrics: + {'align_loss': 0.025073513388633728, + 'recon_loss': 0.057593539357185364, + 'predict_loss': 0.015219456516206264, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1848168969154358, + 'data_time': 0.0007815840071998537, + 'model_time': 1.5752348089881707, + 'grad_norm_pre_clip_avg': 0.2116232544183731, + 'learning_rate': 2.1286841497863133e-05, + 'epoch': 4.12} +04/19 [17:17:23] INFO | >> train_qwenlatent.py:487 + Step 16350 | grad_norm_pre_clip=0.2211 | + grad_norm_pre_clip_avg=0.2187 | Metrics: + {'align_loss': 0.024656958878040314, + 'recon_loss': 0.05622388422489166, + 'predict_loss': 0.016718268394470215, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22105050086975098, + 'mae_score': 0.015223963625796207, 'data_time': + 0.0008773579902481288, 'model_time': + 1.2518027330224868, 'grad_norm_pre_clip_avg': + 0.2186669260263443, 'learning_rate': + 2.1280636249277093e-05, 'epoch': 4.13} +04/19 [17:17:35] INFO | >> train_qwenlatent.py:487 + Step 16360 | grad_norm_pre_clip=0.1723 | + grad_norm_pre_clip_avg=0.2317 | Metrics: + {'align_loss': 0.023701002821326256, + 'recon_loss': 0.05333907902240753, + 'predict_loss': 0.011441736482083797, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1723155826330185, + 'data_time': 0.001006364997010678, + 'model_time': 1.5079767350107431, + 'grad_norm_pre_clip_avg': 0.23174488842487334, + 'learning_rate': 2.1274426727207962e-05, + 'epoch': 4.13} +04/19 [17:17:48] INFO | >> train_qwenlatent.py:487 + Step 16370 | grad_norm_pre_clip=0.1633 | + grad_norm_pre_clip_avg=0.1846 | Metrics: + {'align_loss': 0.024608062580227852, + 'recon_loss': 0.06411075592041016, + 'predict_loss': 0.016148561611771584, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16333217918872833, + 'data_time': 0.0010741619917098433, + 'model_time': 1.2985938630008604, + 'grad_norm_pre_clip_avg': 0.1845576211810112, + 'learning_rate': 2.1268212934682186e-05, + 'epoch': 4.13} +04/19 [17:18:01] INFO | >> train_qwenlatent.py:487 + Step 16380 | grad_norm_pre_clip=0.2599 | + grad_norm_pre_clip_avg=0.2775 | Metrics: + {'align_loss': 0.02498166449368, 'recon_loss': + 0.059375084936618805, 'predict_loss': + 0.011088934727013111, 'aux_loss_decay_weight': + 0.0, 'grad_norm_pre_clip': 0.25989192724227905, + 'data_time': 0.0012191489804536104, + 'model_time': 1.2597118690027855, + 'grad_norm_pre_clip_avg': 0.2774642750620842, + 'learning_rate': 2.1261994874728293e-05, + 'epoch': 4.13} +04/19 [17:18:13] INFO | >> train_qwenlatent.py:487 + Step 16390 | grad_norm_pre_clip=0.2131 | + grad_norm_pre_clip_avg=0.2426 | Metrics: + {'align_loss': 0.025327598676085472, + 'recon_loss': 0.04771345853805542, + 'predict_loss': 0.007932738400995731, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2131214141845703, + 'data_time': 0.0009164420189335942, + 'model_time': 1.265413223998621, + 'grad_norm_pre_clip_avg': 0.24259082823991776, + 'learning_rate': 2.125577255037688e-05, + 'epoch': 4.14} +04/19 [17:18:26] INFO | >> train_qwenlatent.py:487 + Step 16400 | grad_norm_pre_clip=0.2434 | + grad_norm_pre_clip_avg=0.2138 | Metrics: + {'align_loss': 0.025024279952049255, + 'recon_loss': 0.06232461333274841, + 'predict_loss': 0.011012351140379906, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2434137761592865, + 'mae_score': 0.01495156159272065, 'data_time': + 0.0008408599824178964, 'model_time': + 1.2055075640091673, 'grad_norm_pre_clip_avg': + 0.21380753666162491, 'learning_rate': + 2.1249545964660653e-05, 'epoch': 4.14} +04/19 [17:18:39] INFO | >> train_qwenlatent.py:487 + Step 16410 | grad_norm_pre_clip=0.1866 | + grad_norm_pre_clip_avg=0.1987 | Metrics: + {'align_loss': 0.025702206417918205, + 'recon_loss': 0.062475547194480896, + 'predict_loss': 0.01078508235514164, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18661323189735413, + 'data_time': 0.0014510200126096606, + 'model_time': 1.2504459580231924, + 'grad_norm_pre_clip_avg': 0.19874965250492097, + 'learning_rate': 2.1243315120614356e-05, + 'epoch': 4.14} +04/19 [17:18:51] INFO | >> train_qwenlatent.py:487 + Step 16420 | grad_norm_pre_clip=0.1981 | + grad_norm_pre_clip_avg=0.1993 | Metrics: + {'align_loss': 0.025025945156812668, + 'recon_loss': 0.06116701662540436, + 'predict_loss': 0.013299115933477879, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19805559515953064, + 'data_time': 0.0006513679982163012, + 'model_time': 1.225811659998726, + 'grad_norm_pre_clip_avg': 0.199334317445755, + 'learning_rate': 2.123708002127483e-05, + 'epoch': 4.14} +04/19 [17:19:04] INFO | >> train_qwenlatent.py:487 + Step 16430 | grad_norm_pre_clip=0.2410 | + grad_norm_pre_clip_avg=0.2123 | Metrics: + {'align_loss': 0.02532036416232586, + 'recon_loss': 0.08267543464899063, + 'predict_loss': 0.012437128461897373, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24096153676509857, + 'data_time': 0.0006492850079666823, + 'model_time': 1.230566068988992, + 'grad_norm_pre_clip_avg': 0.21227338314056396, + 'learning_rate': 2.1230840669680992e-05, + 'epoch': 4.15} +04/19 [17:19:17] INFO | >> train_qwenlatent.py:487 + Step 16440 | grad_norm_pre_clip=0.1759 | + grad_norm_pre_clip_avg=0.2541 | Metrics: + {'align_loss': 0.02545994147658348, + 'recon_loss': 0.06900730729103088, + 'predict_loss': 0.012888829223811626, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17594321072101593, + 'data_time': 0.0008781540091149509, + 'model_time': 1.2193155880086124, + 'grad_norm_pre_clip_avg': 0.25413249880075456, + 'learning_rate': 2.122459706887382e-05, + 'epoch': 4.15} +04/19 [17:19:30] INFO | >> train_qwenlatent.py:487 + Step 16450 | grad_norm_pre_clip=0.2363 | + grad_norm_pre_clip_avg=0.2477 | Metrics: + {'align_loss': 0.025403063744306564, + 'recon_loss': 0.06263457983732224, + 'predict_loss': 0.009286798536777496, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23629292845726013, + 'mae_score': 0.013148369660248628, 'data_time': + 0.0007044170051813126, 'model_time': + 1.2500437540002167, 'grad_norm_pre_clip_avg': + 0.24770012944936753, 'learning_rate': + 2.1218349221896378e-05, 'epoch': 4.15} +04/19 [17:19:43] INFO | >> train_qwenlatent.py:487 + Step 16460 | grad_norm_pre_clip=0.1659 | + grad_norm_pre_clip_avg=0.2275 | Metrics: + {'align_loss': 0.02422100305557251, + 'recon_loss': 0.051601484417915344, + 'predict_loss': 0.007961738854646683, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16593876481056213, + 'data_time': 0.0008076440135482699, + 'model_time': 1.249755851982627, + 'grad_norm_pre_clip_avg': 0.22752429842948912, + 'learning_rate': 2.1212097131793778e-05, + 'epoch': 4.15} +04/19 [17:19:56] INFO | >> train_qwenlatent.py:487 + Step 16470 | grad_norm_pre_clip=0.2413 | + grad_norm_pre_clip_avg=0.2342 | Metrics: + {'align_loss': 0.02511456608772278, + 'recon_loss': 0.06762116402387619, + 'predict_loss': 0.010918411426246166, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2413049191236496, + 'data_time': 0.0006002129812259227, + 'model_time': 1.4120794539921917, + 'grad_norm_pre_clip_avg': 0.2342478320002556, + 'learning_rate': 2.1205840801613223e-05, + 'epoch': 4.16} +04/19 [17:20:09] INFO | >> train_qwenlatent.py:487 + Step 16480 | grad_norm_pre_clip=0.2191 | + grad_norm_pre_clip_avg=0.2038 | Metrics: + {'align_loss': 0.0253317691385746, + 'recon_loss': 0.065724678337574, + 'predict_loss': 0.01486055925488472, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21914458274841309, + 'data_time': 0.0009294050105381757, + 'model_time': 1.2632509820105042, + 'grad_norm_pre_clip_avg': 0.2038225293159485, + 'learning_rate': 2.1199580234403972e-05, + 'epoch': 4.16} +04/19 [17:20:21] INFO | >> train_qwenlatent.py:487 + Step 16490 | grad_norm_pre_clip=0.1451 | + grad_norm_pre_clip_avg=0.1816 | Metrics: + {'align_loss': 0.024580229073762894, + 'recon_loss': 0.05198196694254875, + 'predict_loss': 0.008996164426207542, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14506129920482635, + 'data_time': 0.0008503530116286129, + 'model_time': 1.259097856003791, + 'grad_norm_pre_clip_avg': 0.18161323964595794, + 'learning_rate': 2.1193315433217346e-05, + 'epoch': 4.16} +04/19 [17:20:35] INFO | >> train_qwenlatent.py:487 + Step 16500 | grad_norm_pre_clip=0.2248 | + grad_norm_pre_clip_avg=0.1784 | Metrics: + {'align_loss': 0.025078793987631798, + 'recon_loss': 0.0580504834651947, + 'predict_loss': 0.012292924337089062, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22477245330810547, + 'mae_score': 0.014980042947305216, 'data_time': + 0.0009319189994130284, 'model_time': + 1.2709466870001052, 'grad_norm_pre_clip_avg': + 0.17837707251310347, 'learning_rate': + 2.118704640110673e-05, 'epoch': 4.16} +04/19 [17:20:47] INFO | >> train_qwenlatent.py:487 + Step 16510 | grad_norm_pre_clip=0.2860 | + grad_norm_pre_clip_avg=0.2684 | Metrics: + {'align_loss': 0.025314532220363617, + 'recon_loss': 0.04811587184667587, + 'predict_loss': 0.009272630326449871, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2859940826892853, + 'data_time': 0.0009127410012297332, + 'model_time': 1.2507049980049487, + 'grad_norm_pre_clip_avg': 0.26843263804912565, + 'learning_rate': 2.118077314112758e-05, + 'epoch': 4.17} +04/19 [17:20:59] INFO | >> train_qwenlatent.py:487 + Step 16520 | grad_norm_pre_clip=0.2002 | + grad_norm_pre_clip_avg=0.1980 | Metrics: + {'align_loss': 0.025315776467323303, + 'recon_loss': 0.057712286710739136, + 'predict_loss': 0.008371653035283089, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20021246373653412, + 'data_time': 0.0008357939950656146, + 'model_time': 1.2121739310096018, + 'grad_norm_pre_clip_avg': 0.19798599779605866, + 'learning_rate': 2.1174495656337403e-05, + 'epoch': 4.17} +04/19 [17:21:12] INFO | >> train_qwenlatent.py:487 + Step 16530 | grad_norm_pre_clip=0.1962 | + grad_norm_pre_clip_avg=0.2376 | Metrics: + {'align_loss': 0.02492760308086872, + 'recon_loss': 0.07139512896537781, + 'predict_loss': 0.014284965582191944, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1962108612060547, + 'data_time': 0.0008609229989815503, + 'model_time': 1.267814096994698, + 'grad_norm_pre_clip_avg': 0.23764360845088958, + 'learning_rate': 2.1168213949795774e-05, + 'epoch': 4.17} +04/19 [17:21:25] INFO | >> train_qwenlatent.py:487 + Step 16540 | grad_norm_pre_clip=0.2076 | + grad_norm_pre_clip_avg=0.2116 | Metrics: + {'align_loss': 0.025865204632282257, + 'recon_loss': 0.06034814193844795, + 'predict_loss': 0.01268615573644638, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2076105773448944, + 'data_time': 0.0008197200077120215, + 'model_time': 1.2442418310092762, + 'grad_norm_pre_clip_avg': 0.21160732805728913, + 'learning_rate': 2.116192802456431e-05, + 'epoch': 4.17} +04/19 [17:21:38] INFO | >> train_qwenlatent.py:487 + Step 16550 | grad_norm_pre_clip=0.3396 | + grad_norm_pre_clip_avg=0.2252 | Metrics: + {'align_loss': 0.024787601083517075, + 'recon_loss': 0.04234758019447327, + 'predict_loss': 0.007092350162565708, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3395679295063019, + 'mae_score': 0.014768330685727232, 'data_time': + 0.0010195979848504066, 'model_time': + 1.3060549660003744, 'grad_norm_pre_clip_avg': + 0.22515634894371034, 'learning_rate': + 2.1155637883706708e-05, 'epoch': 4.18} +04/19 [17:21:51] INFO | >> train_qwenlatent.py:487 + Step 16560 | grad_norm_pre_clip=0.2397 | + grad_norm_pre_clip_avg=0.2419 | Metrics: + {'align_loss': 0.023466818034648895, + 'recon_loss': 0.05285505950450897, + 'predict_loss': 0.014558057300746441, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23966675996780396, + 'data_time': 0.0009544770000502467, + 'model_time': 1.245711872994434, + 'grad_norm_pre_clip_avg': 0.24190569669008255, + 'learning_rate': 2.1149343530288692e-05, + 'epoch': 4.18} +04/19 [17:22:03] INFO | >> train_qwenlatent.py:487 + Step 16570 | grad_norm_pre_clip=0.2529 | + grad_norm_pre_clip_avg=0.2125 | Metrics: + {'align_loss': 0.026099110022187233, + 'recon_loss': 0.05415068566799164, + 'predict_loss': 0.012140694074332714, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25291216373443604, + 'data_time': 0.0006270430167205632, + 'model_time': 1.1880863900005352, + 'grad_norm_pre_clip_avg': 0.21250447183847426, + 'learning_rate': 2.1143044967378068e-05, + 'epoch': 4.18} +04/19 [17:22:16] INFO | >> train_qwenlatent.py:487 + Step 16580 | grad_norm_pre_clip=0.1926 | + grad_norm_pre_clip_avg=0.2070 | Metrics: + {'align_loss': 0.023886874318122864, + 'recon_loss': 0.0585857518017292, + 'predict_loss': 0.013312400318682194, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19262157380580902, + 'data_time': 0.0007882800127845258, + 'model_time': 1.311536602996057, + 'grad_norm_pre_clip_avg': 0.20696439892053603, + 'learning_rate': 2.1136742198044667e-05, + 'epoch': 4.18} +04/19 [17:22:29] INFO | >> train_qwenlatent.py:487 + Step 16590 | grad_norm_pre_clip=0.2220 | + grad_norm_pre_clip_avg=0.2181 | Metrics: + {'align_loss': 0.025581754744052887, + 'recon_loss': 0.05780456215143204, + 'predict_loss': 0.00846790336072445, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22199837863445282, + 'data_time': 0.00126686500152573, 'model_time': + 1.2313328630116303, 'grad_norm_pre_clip_avg': + 0.2180943086743355, 'learning_rate': + 2.1130435225360396e-05, 'epoch': 4.19} +04/19 [17:22:42] INFO | >> train_qwenlatent.py:487 + Step 16600 | grad_norm_pre_clip=0.1855 | + grad_norm_pre_clip_avg=0.2025 | Metrics: + {'align_loss': 0.023259442299604416, + 'recon_loss': 0.05257326737046242, + 'predict_loss': 0.008276334963738918, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1855391412973404, + 'mae_score': 0.012870133030521978, 'data_time': + 0.0006881479930598289, 'model_time': + 1.2213423149951268, 'grad_norm_pre_clip_avg': + 0.20246953517198563, 'learning_rate': + 2.1124124052399178e-05, 'epoch': 4.19} +04/19 [17:22:55] INFO | >> train_qwenlatent.py:487 + Step 16610 | grad_norm_pre_clip=0.2088 | + grad_norm_pre_clip_avg=0.1900 | Metrics: + {'align_loss': 0.025148840621113777, + 'recon_loss': 0.07391508668661118, + 'predict_loss': 0.017665032297372818, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20878374576568604, + 'data_time': 0.0006265040137805045, + 'model_time': 1.2537965399969835, + 'grad_norm_pre_clip_avg': 0.1900279253721237, + 'learning_rate': 2.1117808682237014e-05, + 'epoch': 4.19} +04/19 [17:23:08] INFO | >> train_qwenlatent.py:487 + Step 16620 | grad_norm_pre_clip=0.2035 | + grad_norm_pre_clip_avg=0.2536 | Metrics: + {'align_loss': 0.025819623842835426, + 'recon_loss': 0.07660933583974838, + 'predict_loss': 0.01465002540498972, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2035498470067978, + 'data_time': 0.000774848012952134, + 'model_time': 1.2602847599773668, + 'grad_norm_pre_clip_avg': 0.25359181612730025, + 'learning_rate': 2.111148911795194e-05, + 'epoch': 4.19} +04/19 [17:23:20] INFO | >> train_qwenlatent.py:487 + Step 16630 | grad_norm_pre_clip=0.2380 | + grad_norm_pre_clip_avg=0.2409 | Metrics: + {'align_loss': 0.02397693321108818, + 'recon_loss': 0.05745100602507591, + 'predict_loss': 0.01406522374600172, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2379673272371292, + 'data_time': 0.0015771430043969303, + 'model_time': 1.252483927004505, + 'grad_norm_pre_clip_avg': 0.24088301956653596, + 'learning_rate': 2.1105165362624034e-05, + 'epoch': 4.2} +04/19 [17:23:33] INFO | >> train_qwenlatent.py:487 + Step 16640 | grad_norm_pre_clip=0.1700 | + grad_norm_pre_clip_avg=0.1842 | Metrics: + {'align_loss': 0.024044934660196304, + 'recon_loss': 0.07474919408559799, + 'predict_loss': 0.010918953455984592, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16998015344142914, + 'data_time': 0.0007102469971869141, + 'model_time': 1.2286468480015174, + 'grad_norm_pre_clip_avg': 0.18419473767280578, + 'learning_rate': 2.1098837419335407e-05, + 'epoch': 4.2} +04/19 [17:23:46] INFO | >> train_qwenlatent.py:487 + Step 16650 | grad_norm_pre_clip=0.1897 | + grad_norm_pre_clip_avg=0.1876 | Metrics: + {'align_loss': 0.02539099007844925, + 'recon_loss': 0.07140009105205536, + 'predict_loss': 0.013613017275929451, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.189729705452919, + 'mae_score': 0.01951138092590882, 'data_time': + 0.0006973530107643455, 'model_time': + 1.2213241229765117, 'grad_norm_pre_clip_avg': + 0.18755656033754348, 'learning_rate': + 2.1092505291170234e-05, 'epoch': 4.2} +04/19 [17:23:58] INFO | >> train_qwenlatent.py:487 + Step 16660 | grad_norm_pre_clip=0.2563 | + grad_norm_pre_clip_avg=0.2671 | Metrics: + {'align_loss': 0.026173533871769905, + 'recon_loss': 0.06547857820987701, + 'predict_loss': 0.012961612083017826, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2563377022743225, + 'data_time': 0.0008149410132318735, + 'model_time': 1.2382502789841965, + 'grad_norm_pre_clip_avg': 0.2671453610062599, + 'learning_rate': 2.1086168981214713e-05, + 'epoch': 4.2} +04/19 [17:24:10] INFO | >> train_qwenlatent.py:487 + Step 16670 | grad_norm_pre_clip=0.2070 | + grad_norm_pre_clip_avg=0.2049 | Metrics: + {'align_loss': 0.026221349835395813, + 'recon_loss': 0.07797925919294357, + 'predict_loss': 0.014712284319102764, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2070164829492569, + 'data_time': 0.0007996590284164995, + 'model_time': 1.2145384800096508, + 'grad_norm_pre_clip_avg': 0.20492172986268997, + 'learning_rate': 2.1079828492557087e-05, + 'epoch': 4.21} +04/19 [17:24:23] INFO | >> train_qwenlatent.py:487 + Step 16680 | grad_norm_pre_clip=0.2226 | + grad_norm_pre_clip_avg=0.2094 | Metrics: + {'align_loss': 0.024679068475961685, + 'recon_loss': 0.04888415336608887, + 'predict_loss': 0.010015725158154964, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22255730628967285, + 'data_time': 0.0010166779975406826, + 'model_time': 1.208458237000741, + 'grad_norm_pre_clip_avg': 0.20941896438598634, + 'learning_rate': 2.1073483828287628e-05, + 'epoch': 4.21} +04/19 [17:24:35] INFO | >> train_qwenlatent.py:487 + Step 16690 | grad_norm_pre_clip=0.2110 | + grad_norm_pre_clip_avg=0.2626 | Metrics: + {'align_loss': 0.0256425179541111, + 'recon_loss': 0.06005747988820076, + 'predict_loss': 0.010354658588767052, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.211026132106781, + 'data_time': 0.0008650500094518065, + 'model_time': 1.2319273850007448, + 'grad_norm_pre_clip_avg': 0.2626254692673683, + 'learning_rate': 2.1067134991498648e-05, + 'epoch': 4.21} +04/19 [17:24:48] INFO | >> train_qwenlatent.py:487 + Step 16700 | grad_norm_pre_clip=0.1975 | + grad_norm_pre_clip_avg=0.2309 | Metrics: + {'align_loss': 0.025225061923265457, + 'recon_loss': 0.06235785409808159, + 'predict_loss': 0.013003855012357235, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19748342037200928, + 'mae_score': 0.019313717747593787, 'data_time': + 0.0008427689899690449, 'model_time': + 1.1915352070063818, 'grad_norm_pre_clip_avg': + 0.23088568150997163, 'learning_rate': + 2.1060781985284504e-05, 'epoch': 4.21} +04/19 [17:25:01] INFO | >> train_qwenlatent.py:487 + Step 16710 | grad_norm_pre_clip=0.1837 | + grad_norm_pre_clip_avg=0.1996 | Metrics: + {'align_loss': 0.02492348663508892, + 'recon_loss': 0.06345950812101364, + 'predict_loss': 0.01588474214076996, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18371346592903137, + 'data_time': 0.0010334430262446404, + 'model_time': 1.2925172630057205, + 'grad_norm_pre_clip_avg': 0.19960432946681977, + 'learning_rate': 2.105442481274156e-05, + 'epoch': 4.22} +04/19 [17:25:14] INFO | >> train_qwenlatent.py:487 + Step 16720 | grad_norm_pre_clip=0.1933 | + grad_norm_pre_clip_avg=0.2073 | Metrics: + {'align_loss': 0.025622127577662468, + 'recon_loss': 0.0858592838048935, + 'predict_loss': 0.015849275514483452, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1933460831642151, + 'data_time': 0.000936000986257568, + 'model_time': 1.210573933989508, + 'grad_norm_pre_clip_avg': 0.20726353973150252, + 'learning_rate': 2.1048063476968232e-05, + 'epoch': 4.22} +04/19 [17:25:26] INFO | >> train_qwenlatent.py:487 + Step 16730 | grad_norm_pre_clip=0.2250 | + grad_norm_pre_clip_avg=0.2125 | Metrics: + {'align_loss': 0.024882705882191658, + 'recon_loss': 0.07007235288619995, + 'predict_loss': 0.014192165806889534, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22503243386745453, + 'data_time': 0.0007807550136931241, + 'model_time': 1.2065556180023123, + 'grad_norm_pre_clip_avg': 0.21249296963214875, + 'learning_rate': 2.1041697981064955e-05, + 'epoch': 4.22} +04/19 [17:25:39] INFO | >> train_qwenlatent.py:487 + Step 16740 | grad_norm_pre_clip=0.1709 | + grad_norm_pre_clip_avg=0.2072 | Metrics: + {'align_loss': 0.02404049038887024, + 'recon_loss': 0.06386053562164307, + 'predict_loss': 0.017739929258823395, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1709427684545517, + 'data_time': 0.0013113849854562432, + 'model_time': 1.2368678429920692, + 'grad_norm_pre_clip_avg': 0.20717629045248032, + 'learning_rate': 2.10353283281342e-05, 'epoch': + 4.22} +04/19 [17:25:52] INFO | >> train_qwenlatent.py:487 + Step 16750 | grad_norm_pre_clip=0.1803 | + grad_norm_pre_clip_avg=0.2171 | Metrics: + {'align_loss': 0.024597302079200745, + 'recon_loss': 0.07266048341989517, + 'predict_loss': 0.01589905098080635, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1802511066198349, + 'mae_score': 0.014379855319186374, 'data_time': + 0.000914700998691842, 'model_time': + 1.243571677012369, 'grad_norm_pre_clip_avg': + 0.21709166914224626, 'learning_rate': + 2.1028954521280453e-05, 'epoch': 4.23} +04/19 [17:26:05] INFO | >> train_qwenlatent.py:487 + Step 16760 | grad_norm_pre_clip=0.3448 | + grad_norm_pre_clip_avg=0.2085 | Metrics: + {'align_loss': 0.024779483675956726, + 'recon_loss': 0.0667138397693634, + 'predict_loss': 0.013395761139690876, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3447817862033844, + 'data_time': 0.0013458360044751316, + 'model_time': 1.2143622110015713, + 'grad_norm_pre_clip_avg': 0.2085162252187729, + 'learning_rate': 2.102257656361023e-05, + 'epoch': 4.23} +04/19 [17:26:18] INFO | >> train_qwenlatent.py:487 + Step 16770 | grad_norm_pre_clip=0.2888 | + grad_norm_pre_clip_avg=0.2780 | Metrics: + {'align_loss': 0.02539140358567238, + 'recon_loss': 0.06325867027044296, + 'predict_loss': 0.018623029813170433, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.288799911737442, + 'data_time': 0.0006717280193697661, + 'model_time': 1.2530910710047465, + 'grad_norm_pre_clip_avg': 0.27804320454597475, + 'learning_rate': 2.1016194458232072e-05, + 'epoch': 4.23} +04/19 [17:26:31] INFO | >> train_qwenlatent.py:487 + Step 16780 | grad_norm_pre_clip=0.1902 | + grad_norm_pre_clip_avg=0.2203 | Metrics: + {'align_loss': 0.02445819228887558, + 'recon_loss': 0.05679721385240555, + 'predict_loss': 0.012486760504543781, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19020335376262665, + 'data_time': 0.0008474070054944605, + 'model_time': 1.3116234090120997, + 'grad_norm_pre_clip_avg': 0.22030386030673982, + 'learning_rate': 2.1009808208256537e-05, + 'epoch': 4.23} +04/19 [17:26:43] INFO | >> train_qwenlatent.py:487 + Step 16790 | grad_norm_pre_clip=0.1707 | + grad_norm_pre_clip_avg=0.1831 | Metrics: + {'align_loss': 0.024478044360876083, + 'recon_loss': 0.05988513305783272, + 'predict_loss': 0.01699557900428772, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17065118253231049, + 'data_time': 0.000627066008746624, + 'model_time': 1.229582232015673, + 'grad_norm_pre_clip_avg': 0.1830825075507164, + 'learning_rate': 2.100341781679621e-05, + 'epoch': 4.24} +04/19 [17:26:57] INFO | >> train_qwenlatent.py:487 + Step 16800 | grad_norm_pre_clip=0.1850 | + grad_norm_pre_clip_avg=0.2216 | Metrics: + {'align_loss': 0.024414993822574615, + 'recon_loss': 0.04734663665294647, + 'predict_loss': 0.009996438398957253, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18503084778785706, + 'mae_score': 0.012469140473786776, 'data_time': + 0.0006736080104019493, 'model_time': + 1.2184286199917551, 'grad_norm_pre_clip_avg': + 0.22160065174102783, 'learning_rate': + 2.0997023286965688e-05, 'epoch': 4.24} +04/19 [17:27:09] INFO | >> train_qwenlatent.py:487 + Step 16810 | grad_norm_pre_clip=0.2003 | + grad_norm_pre_clip_avg=0.2286 | Metrics: + {'align_loss': 0.024252595379948616, + 'recon_loss': 0.04787704348564148, + 'predict_loss': 0.007290084380656481, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2002531886100769, + 'data_time': 0.0010806670179590583, + 'model_time': 1.22775034600636, + 'grad_norm_pre_clip_avg': 0.2286390021443367, + 'learning_rate': 2.099062462188159e-05, + 'epoch': 4.24} +04/19 [17:27:21] INFO | >> train_qwenlatent.py:487 + Step 16820 | grad_norm_pre_clip=0.2019 | + grad_norm_pre_clip_avg=0.1835 | Metrics: + {'align_loss': 0.023962877690792084, + 'recon_loss': 0.07520352303981781, + 'predict_loss': 0.01440358255058527, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2019018977880478, + 'data_time': 0.0006780020194128156, + 'model_time': 1.260681284009479, + 'grad_norm_pre_clip_avg': 0.1834510773420334, + 'learning_rate': 2.098422182466254e-05, + 'epoch': 4.24} +04/19 [17:27:34] INFO | >> train_qwenlatent.py:487 + Step 16830 | grad_norm_pre_clip=0.3058 | + grad_norm_pre_clip_avg=0.2909 | Metrics: + {'align_loss': 0.023899629712104797, + 'recon_loss': 0.044572025537490845, + 'predict_loss': 0.009988183155655861, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3057825565338135, + 'data_time': 0.0008518919930793345, + 'model_time': 1.2075146679999307, + 'grad_norm_pre_clip_avg': 0.29092286080121993, + 'learning_rate': 2.0977814898429188e-05, + 'epoch': 4.25} +04/19 [17:27:47] INFO | >> train_qwenlatent.py:487 + Step 16840 | grad_norm_pre_clip=0.1948 | + grad_norm_pre_clip_avg=0.2089 | Metrics: + {'align_loss': 0.025238092988729477, + 'recon_loss': 0.08183076977729797, + 'predict_loss': 0.014803273603320122, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1948438584804535, + 'data_time': 0.0010527840058784932, + 'model_time': 1.2362450410146266, + 'grad_norm_pre_clip_avg': 0.2089228093624115, + 'learning_rate': 2.0971403846304193e-05, + 'epoch': 4.25} +04/19 [17:28:00] INFO | >> train_qwenlatent.py:487 + Step 16850 | grad_norm_pre_clip=0.2154 | + grad_norm_pre_clip_avg=0.1983 | Metrics: + {'align_loss': 0.02512011118233204, + 'recon_loss': 0.04443129897117615, + 'predict_loss': 0.008287976495921612, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21537479758262634, + 'mae_score': 0.013924175984150654, 'data_time': + 0.0007115280022844672, 'model_time': + 1.2817661279987078, 'grad_norm_pre_clip_avg': + 0.19827320873737336, 'learning_rate': + 2.0964988671412226e-05, 'epoch': 4.25} +04/19 [17:28:13] INFO | >> train_qwenlatent.py:487 + Step 16860 | grad_norm_pre_clip=0.1820 | + grad_norm_pre_clip_avg=0.2075 | Metrics: + {'align_loss': 0.02431473135948181, + 'recon_loss': 0.06510142236948013, + 'predict_loss': 0.012551352381706238, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18202762305736542, + 'data_time': 0.0006851530051790178, + 'model_time': 1.2402917809959035, + 'grad_norm_pre_clip_avg': 0.20747710913419723, + 'learning_rate': 2.095856937687996e-05, + 'epoch': 4.25} +04/19 [17:28:25] INFO | >> train_qwenlatent.py:487 + Step 16870 | grad_norm_pre_clip=0.2990 | + grad_norm_pre_clip_avg=0.2145 | Metrics: + {'align_loss': 0.025108877569437027, + 'recon_loss': 0.05862058326601982, + 'predict_loss': 0.007365946192294359, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.29904356598854065, + 'data_time': 0.000739534996682778, + 'model_time': 1.1919751520035788, + 'grad_norm_pre_clip_avg': 0.21454060673713685, + 'learning_rate': 2.0952145965836086e-05, + 'epoch': 4.26} +04/19 [17:28:38] INFO | >> train_qwenlatent.py:487 + Step 16880 | grad_norm_pre_clip=0.2136 | + grad_norm_pre_clip_avg=0.2599 | Metrics: + {'align_loss': 0.026040077209472656, + 'recon_loss': 0.07926284521818161, + 'predict_loss': 0.0128368204459548, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21360760927200317, + 'data_time': 0.0006533870182465762, + 'model_time': 1.2735788699937984, + 'grad_norm_pre_clip_avg': 0.259918649494648, + 'learning_rate': 2.0945718441411292e-05, + 'epoch': 4.26} +04/19 [17:28:51] INFO | >> train_qwenlatent.py:487 + Step 16890 | grad_norm_pre_clip=0.2190 | + grad_norm_pre_clip_avg=0.2201 | Metrics: + {'align_loss': 0.024722009897232056, + 'recon_loss': 0.04033069312572479, + 'predict_loss': 0.008364016190171242, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2190486490726471, + 'data_time': 0.0009156579908449203, + 'model_time': 1.3129928949929308, + 'grad_norm_pre_clip_avg': 0.22006195336580275, + 'learning_rate': 2.093928680673828e-05, + 'epoch': 4.26} +04/19 [17:29:04] INFO | >> train_qwenlatent.py:487 + Step 16900 | grad_norm_pre_clip=0.2033 | + grad_norm_pre_clip_avg=0.2049 | Metrics: + {'align_loss': 0.024216074496507645, + 'recon_loss': 0.06885451823472977, + 'predict_loss': 0.013731993734836578, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20328958332538605, + 'mae_score': 0.015293346439395939, 'data_time': + 0.0008509109902661294, 'model_time': + 1.2313607090036385, 'grad_norm_pre_clip_avg': + 0.20488014370203017, 'learning_rate': + 2.0932851064951743e-05, 'epoch': 4.26} +04/19 [17:29:17] INFO | >> train_qwenlatent.py:487 + Step 16910 | grad_norm_pre_clip=0.1880 | + grad_norm_pre_clip_avg=0.2011 | Metrics: + {'align_loss': 0.024708746001124382, + 'recon_loss': 0.05997854843735695, + 'predict_loss': 0.011424676515161991, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18795166909694672, + 'data_time': 0.0011612590169534087, + 'model_time': 1.2206108149839565, + 'grad_norm_pre_clip_avg': 0.20105370581150056, + 'learning_rate': 2.0926411219188393e-05, + 'epoch': 4.27} +04/19 [17:29:29] INFO | >> train_qwenlatent.py:487 + Step 16920 | grad_norm_pre_clip=0.3756 | + grad_norm_pre_clip_avg=0.2916 | Metrics: + {'align_loss': 0.02576727420091629, + 'recon_loss': 0.07094328850507736, + 'predict_loss': 0.014089178293943405, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3756382167339325, + 'data_time': 0.0007272980001289397, + 'model_time': 1.2030412740132306, + 'grad_norm_pre_clip_avg': 0.2915693536400795, + 'learning_rate': 2.0919967272586928e-05, + 'epoch': 4.27} +04/19 [17:29:42] INFO | >> train_qwenlatent.py:487 + Step 16930 | grad_norm_pre_clip=0.2376 | + grad_norm_pre_clip_avg=0.2693 | Metrics: + {'align_loss': 0.025170382112264633, + 'recon_loss': 0.05779210850596428, + 'predict_loss': 0.013554072938859463, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2375950962305069, + 'data_time': 0.0011339250195305794, + 'model_time': 1.315855313994689, + 'grad_norm_pre_clip_avg': 0.2692690953612328, + 'learning_rate': 2.0913519228288052e-05, + 'epoch': 4.27} +04/19 [17:29:55] INFO | >> train_qwenlatent.py:487 + Step 16940 | grad_norm_pre_clip=0.1867 | + grad_norm_pre_clip_avg=0.2065 | Metrics: + {'align_loss': 0.0266004279255867, + 'recon_loss': 0.06646562367677689, + 'predict_loss': 0.01230815052986145, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18673120439052582, + 'data_time': 0.0007793050026521087, + 'model_time': 1.282763754017651, + 'grad_norm_pre_clip_avg': 0.20646303594112397, + 'learning_rate': 2.090706708943446e-05, + 'epoch': 4.27} +04/19 [17:30:08] INFO | >> train_qwenlatent.py:487 + Step 16950 | grad_norm_pre_clip=0.2021 | + grad_norm_pre_clip_avg=0.1797 | Metrics: + {'align_loss': 0.02448778785765171, + 'recon_loss': 0.055780958384275436, + 'predict_loss': 0.009598422795534134, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20208287239074707, + 'mae_score': 0.014616419817950274, 'data_time': + 0.0008451789908576757, 'model_time': + 1.2477659019932617, 'grad_norm_pre_clip_avg': + 0.17965890765190123, 'learning_rate': + 2.090061085917084e-05, 'epoch': 4.28} +04/19 [17:30:21] INFO | >> train_qwenlatent.py:487 + Step 16960 | grad_norm_pre_clip=0.2272 | + grad_norm_pre_clip_avg=0.1994 | Metrics: + {'align_loss': 0.024677403271198273, + 'recon_loss': 0.08224441856145859, + 'predict_loss': 0.017443904653191566, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22722665965557098, + 'data_time': 0.0009713990148156881, + 'model_time': 1.207437054021284, + 'grad_norm_pre_clip_avg': 0.19935627579689025, + 'learning_rate': 2.0894150540643892e-05, + 'epoch': 4.28} +04/19 [17:30:33] INFO | >> train_qwenlatent.py:487 + Step 16970 | grad_norm_pre_clip=0.1999 | + grad_norm_pre_clip_avg=0.1978 | Metrics: + {'align_loss': 0.0254606194794178, + 'recon_loss': 0.05103548243641853, + 'predict_loss': 0.009900122880935669, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19987620413303375, + 'data_time': 0.0009424379968550056, + 'model_time': 1.2862250659964047, + 'grad_norm_pre_clip_avg': 0.19778075218200683, + 'learning_rate': 2.0887686137002292e-05, + 'epoch': 4.28} +04/19 [17:30:46] INFO | >> train_qwenlatent.py:487 + Step 16980 | grad_norm_pre_clip=0.3001 | + grad_norm_pre_clip_avg=0.2915 | Metrics: + {'align_loss': 0.024732379242777824, + 'recon_loss': 0.06453747302293777, + 'predict_loss': 0.013282752595841885, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3000549077987671, + 'data_time': 0.0007041939825285226, + 'model_time': 1.2336727880174294, + 'grad_norm_pre_clip_avg': 0.2914718210697174, + 'learning_rate': 2.088121765139671e-05, + 'epoch': 4.28} +04/19 [17:30:58] INFO | >> train_qwenlatent.py:487 + Step 16990 | grad_norm_pre_clip=0.1998 | + grad_norm_pre_clip_avg=0.2190 | Metrics: + {'align_loss': 0.025948576629161835, + 'recon_loss': 0.06411579996347427, + 'predict_loss': 0.011872117407619953, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19979216158390045, + 'data_time': 0.0009354880021419376, + 'model_time': 1.24537412400241, + 'grad_norm_pre_clip_avg': 0.21903369426727295, + 'learning_rate': 2.087474508697981e-05, + 'epoch': 4.29} +04/19 [17:31:12] INFO | >> train_qwenlatent.py:487 + Step 17000 | grad_norm_pre_clip=0.1887 | + grad_norm_pre_clip_avg=0.1718 | Metrics: + {'align_loss': 0.024806104600429535, + 'recon_loss': 0.05253986641764641, + 'predict_loss': 0.010191765613853931, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18871992826461792, + 'mae_score': 0.015043602333412514, 'data_time': + 0.0007734070240985602, 'model_time': + 1.2112012160068844, 'grad_norm_pre_clip_avg': + 0.17183958739042282, 'learning_rate': + 2.0868268446906245e-05, 'epoch': 4.29} +04/19 [17:31:25] INFO | >> train_qwenlatent.py:487 + Step 17010 | grad_norm_pre_clip=0.2462 | + grad_norm_pre_clip_avg=0.2017 | Metrics: + {'align_loss': 0.024523092433810234, + 'recon_loss': 0.059540119022130966, + 'predict_loss': 0.016743239015340805, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24621832370758057, + 'data_time': 0.0010161960090044886, + 'model_time': 1.2232912159815896, + 'grad_norm_pre_clip_avg': 0.20172432363033294, + 'learning_rate': 2.086178773433264e-05, + 'epoch': 4.29} +04/19 [17:31:37] INFO | >> train_qwenlatent.py:487 + Step 17020 | grad_norm_pre_clip=0.2606 | + grad_norm_pre_clip_avg=0.2347 | Metrics: + {'align_loss': 0.025085676461458206, + 'recon_loss': 0.07340481877326965, + 'predict_loss': 0.015453185886144638, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26057952642440796, + 'data_time': 0.0008599409775342792, + 'model_time': 1.2097999599936884, + 'grad_norm_pre_clip_avg': 0.23465111702680588, + 'learning_rate': 2.085530295241762e-05, + 'epoch': 4.29} +04/19 [17:31:50] INFO | >> train_qwenlatent.py:487 + Step 17030 | grad_norm_pre_clip=0.1482 | + grad_norm_pre_clip_avg=0.2138 | Metrics: + {'align_loss': 0.025868430733680725, + 'recon_loss': 0.06598479300737381, + 'predict_loss': 0.011420921422541142, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1481533646583557, + 'data_time': 0.0007226770103443414, + 'model_time': 1.2018361299997196, + 'grad_norm_pre_clip_avg': 0.21382963955402373, + 'learning_rate': 2.08488141043218e-05, 'epoch': + 4.3} +04/19 [17:32:03] INFO | >> train_qwenlatent.py:487 + Step 17040 | grad_norm_pre_clip=0.3043 | + grad_norm_pre_clip_avg=0.2044 | Metrics: + {'align_loss': 0.024126969277858734, + 'recon_loss': 0.053806766867637634, + 'predict_loss': 0.011750886216759682, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3042588531970978, + 'data_time': 0.0007879800105001777, + 'model_time': 1.2535428029950708, + 'grad_norm_pre_clip_avg': 0.20438382476568223, + 'learning_rate': 2.0842321193207747e-05, + 'epoch': 4.3} +04/19 [17:32:16] INFO | >> train_qwenlatent.py:487 + Step 17050 | grad_norm_pre_clip=0.2870 | + grad_norm_pre_clip_avg=0.2657 | Metrics: + {'align_loss': 0.025391701608896255, + 'recon_loss': 0.057336125522851944, + 'predict_loss': 0.010944905690848827, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28695595264434814, + 'mae_score': 0.012321845905200855, 'data_time': + 0.0008855519990902394, 'model_time': + 1.2285278869967442, 'grad_norm_pre_clip_avg': + 0.2657311841845512, 'learning_rate': + 2.083582422224004e-05, 'epoch': 4.3} +04/19 [17:32:29] INFO | >> train_qwenlatent.py:487 + Step 17060 | grad_norm_pre_clip=0.1617 | + grad_norm_pre_clip_avg=0.2109 | Metrics: + {'align_loss': 0.024557383731007576, + 'recon_loss': 0.04702949896454811, + 'predict_loss': 0.008303329348564148, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16173236072063446, + 'data_time': 0.0006332949851639569, + 'model_time': 1.2088547350140288, + 'grad_norm_pre_clip_avg': 0.21092969179153442, + 'learning_rate': 2.0829323194585217e-05, + 'epoch': 4.3} +04/19 [17:32:41] INFO | >> train_qwenlatent.py:487 + Step 17070 | grad_norm_pre_clip=0.2141 | + grad_norm_pre_clip_avg=0.2246 | Metrics: + {'align_loss': 0.02554258331656456, + 'recon_loss': 0.056665509939193726, + 'predict_loss': 0.013204909861087799, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21406707167625427, + 'data_time': 0.0009094690030906349, + 'model_time': 1.194975368998712, + 'grad_norm_pre_clip_avg': 0.2245962589979172, + 'learning_rate': 2.082281811341181e-05, + 'epoch': 4.31} +04/19 [17:32:53] INFO | >> train_qwenlatent.py:487 + Step 17080 | grad_norm_pre_clip=0.1848 | + grad_norm_pre_clip_avg=0.1939 | Metrics: + {'align_loss': 0.024969937279820442, + 'recon_loss': 0.07148796319961548, + 'predict_loss': 0.01567474566400051, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18475386500358582, + 'data_time': 0.0008850499871186912, + 'model_time': 1.1906708230089862, + 'grad_norm_pre_clip_avg': 0.19388098120689393, + 'learning_rate': 2.08163089818903e-05, 'epoch': + 4.31} +04/19 [17:33:06] INFO | >> train_qwenlatent.py:487 + Step 17090 | grad_norm_pre_clip=0.2038 | + grad_norm_pre_clip_avg=0.2317 | Metrics: + {'align_loss': 0.025588132441043854, + 'recon_loss': 0.058737244457006454, + 'predict_loss': 0.009231941774487495, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20377987623214722, + 'data_time': 0.0011496020015329123, + 'model_time': 1.2705022979935165, + 'grad_norm_pre_clip_avg': 0.23168192207813262, + 'learning_rate': 2.0809795803193173e-05, + 'epoch': 4.31} +04/19 [17:33:19] INFO | >> train_qwenlatent.py:487 + Step 17100 | grad_norm_pre_clip=0.1944 | + grad_norm_pre_clip_avg=0.2011 | Metrics: + {'align_loss': 0.025440501049160957, + 'recon_loss': 0.0703851580619812, + 'predict_loss': 0.01661888137459755, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19442075490951538, + 'mae_score': 0.022362035459226316, 'data_time': + 0.0007528809946961701, 'model_time': + 1.23070174199529, 'grad_norm_pre_clip_avg': + 0.20109312534332274, 'learning_rate': + 2.0803278580494868e-05, 'epoch': 4.31} +04/19 [17:33:32] INFO | >> train_qwenlatent.py:487 + Step 17110 | grad_norm_pre_clip=0.2141 | + grad_norm_pre_clip_avg=0.2123 | Metrics: + {'align_loss': 0.024517029523849487, + 'recon_loss': 0.05549777299165726, + 'predict_loss': 0.00979728251695633, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21411404013633728, + 'data_time': 0.0006810720078647137, + 'model_time': 1.265117735019885, + 'grad_norm_pre_clip_avg': 0.2122846871614456, + 'learning_rate': 2.07967573169718e-05, 'epoch': + 4.32} +04/19 [17:33:45] INFO | >> train_qwenlatent.py:487 + Step 17120 | grad_norm_pre_clip=0.1736 | + grad_norm_pre_clip_avg=0.2049 | Metrics: + {'align_loss': 0.024756425991654396, + 'recon_loss': 0.06405561417341232, + 'predict_loss': 0.01873849332332611, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1736326813697815, + 'data_time': 0.0009640460193622857, + 'model_time': 1.1902607759984676, + 'grad_norm_pre_clip_avg': 0.20490709394216539, + 'learning_rate': 2.0790232015802356e-05, + 'epoch': 4.32} +04/19 [17:33:57] INFO | >> train_qwenlatent.py:487 + Step 17130 | grad_norm_pre_clip=0.1893 | + grad_norm_pre_clip_avg=0.2207 | Metrics: + {'align_loss': 0.02511448971927166, + 'recon_loss': 0.05429842323064804, + 'predict_loss': 0.009186734445393085, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18927054107189178, + 'data_time': 0.0007354639819823205, + 'model_time': 1.2697175449866336, + 'grad_norm_pre_clip_avg': 0.2207026019692421, + 'learning_rate': 2.0783702680166882e-05, + 'epoch': 4.32} +04/19 [17:34:10] INFO | >> train_qwenlatent.py:487 + Step 17140 | grad_norm_pre_clip=0.1894 | + grad_norm_pre_clip_avg=0.2471 | Metrics: + {'align_loss': 0.024549473077058792, + 'recon_loss': 0.06250640749931335, + 'predict_loss': 0.013004103675484657, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1893996298313141, + 'data_time': 0.0009028770145960152, + 'model_time': 1.2608845369832125, + 'grad_norm_pre_clip_avg': 0.24706777781248093, + 'learning_rate': 2.0777169313247705e-05, + 'epoch': 4.33} +04/19 [17:34:23] INFO | >> train_qwenlatent.py:487 + Step 17150 | grad_norm_pre_clip=0.2091 | + grad_norm_pre_clip_avg=0.1896 | Metrics: + {'align_loss': 0.024952681735157967, + 'recon_loss': 0.04251672327518463, + 'predict_loss': 0.007220558822154999, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20912814140319824, + 'mae_score': 0.013296417717461113, 'data_time': + 0.000652811984764412, 'model_time': + 1.2315761969948653, 'grad_norm_pre_clip_avg': + 0.1896428495645523, 'learning_rate': + 2.07706319182291e-05, 'epoch': 4.33} +04/19 [17:34:36] INFO | >> train_qwenlatent.py:487 + Step 17160 | grad_norm_pre_clip=0.3059 | + grad_norm_pre_clip_avg=0.2218 | Metrics: + {'align_loss': 0.02537386864423752, + 'recon_loss': 0.05160335451364517, + 'predict_loss': 0.009809304028749466, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3058507442474365, + 'data_time': 0.0007076549809426069, + 'model_time': 1.281261291995179, + 'grad_norm_pre_clip_avg': 0.22182406187057496, + 'learning_rate': 2.076409049829733e-05, + 'epoch': 4.33} +04/19 [17:34:48] INFO | >> train_qwenlatent.py:487 + Step 17170 | grad_norm_pre_clip=0.2021 | + grad_norm_pre_clip_avg=0.2249 | Metrics: + {'align_loss': 0.023136014118790627, + 'recon_loss': 0.04678763449192047, + 'predict_loss': 0.009167374111711979, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2021392434835434, + 'data_time': 0.0007017369789537042, + 'model_time': 1.5714180970098823, + 'grad_norm_pre_clip_avg': 0.22494674026966094, + 'learning_rate': 2.0757545056640584e-05, + 'epoch': 4.33} +04/19 [17:35:01] INFO | >> train_qwenlatent.py:487 + Step 17180 | grad_norm_pre_clip=0.2344 | + grad_norm_pre_clip_avg=0.2165 | Metrics: + {'align_loss': 0.024583593010902405, + 'recon_loss': 0.06323829293251038, + 'predict_loss': 0.012175941839814186, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23440831899642944, + 'data_time': 0.0009672979940660298, + 'model_time': 1.2278506089933217, + 'grad_norm_pre_clip_avg': 0.2164554089307785, + 'learning_rate': 2.0750995596449042e-05, + 'epoch': 4.34} +04/19 [17:35:14] INFO | >> train_qwenlatent.py:487 + Step 17190 | grad_norm_pre_clip=0.2723 | + grad_norm_pre_clip_avg=0.2258 | Metrics: + {'align_loss': 0.02555188350379467, + 'recon_loss': 0.07051898539066315, + 'predict_loss': 0.01840307004749775, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.27225762605667114, + 'data_time': 0.0009298040240537375, + 'model_time': 1.2943026580032893, + 'grad_norm_pre_clip_avg': 0.22581993341445922, + 'learning_rate': 2.0744442120914823e-05, + 'epoch': 4.34} +04/19 [17:35:27] INFO | >> train_qwenlatent.py:487 + Step 17200 | grad_norm_pre_clip=0.2004 | + grad_norm_pre_clip_avg=0.2024 | Metrics: + {'align_loss': 0.026512503623962402, + 'recon_loss': 0.07672277092933655, + 'predict_loss': 0.014637830667197704, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20035304129123688, + 'mae_score': 0.016016733324205553, 'data_time': + 0.000879387982422486, 'model_time': + 1.2560353470034897, 'grad_norm_pre_clip_avg': + 0.20236867815256118, 'learning_rate': + 2.0737884633232022e-05, 'epoch': 4.34} +04/19 [17:35:40] INFO | >> train_qwenlatent.py:487 + Step 17210 | grad_norm_pre_clip=0.2726 | + grad_norm_pre_clip_avg=0.2101 | Metrics: + {'align_loss': 0.024168711155653, 'recon_loss': + 0.07606194168329239, 'predict_loss': + 0.015181642957031727, 'aux_loss_decay_weight': + 0.0, 'grad_norm_pre_clip': 0.2725861370563507, + 'data_time': 0.0011221390159334987, + 'model_time': 1.19278397000744, + 'grad_norm_pre_clip_avg': 0.21012849658727645, + 'learning_rate': 2.0731323136596676e-05, + 'epoch': 4.34} +04/19 [17:35:52] INFO | >> train_qwenlatent.py:487 + Step 17220 | grad_norm_pre_clip=0.2389 | + grad_norm_pre_clip_avg=0.2506 | Metrics: + {'align_loss': 0.02482423186302185, + 'recon_loss': 0.06827083975076675, + 'predict_loss': 0.015490631572902203, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2388555258512497, + 'data_time': 0.0006664329848717898, + 'model_time': 1.228714091994334, + 'grad_norm_pre_clip_avg': 0.2505749732255936, + 'learning_rate': 2.0724757634206772e-05, + 'epoch': 4.35} +04/19 [17:36:05] INFO | >> train_qwenlatent.py:487 + Step 17230 | grad_norm_pre_clip=0.2674 | + grad_norm_pre_clip_avg=0.2180 | Metrics: + {'align_loss': 0.02530779503285885, + 'recon_loss': 0.06491874903440475, + 'predict_loss': 0.009362133219838142, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26741769909858704, + 'data_time': 0.0009546630026306957, + 'model_time': 1.2077515510027297, + 'grad_norm_pre_clip_avg': 0.21804644614458085, + 'learning_rate': 2.0718188129262264e-05, + 'epoch': 4.35} +04/19 [17:36:17] INFO | >> train_qwenlatent.py:487 + Step 17240 | grad_norm_pre_clip=0.2320 | + grad_norm_pre_clip_avg=0.2188 | Metrics: + {'align_loss': 0.025080405175685883, + 'recon_loss': 0.06087716668844223, + 'predict_loss': 0.008485551923513412, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23198121786117554, + 'data_time': 0.0007195519865490496, + 'model_time': 1.2555388920009136, + 'grad_norm_pre_clip_avg': 0.21878503113985062, + 'learning_rate': 2.0711614624965044e-05, + 'epoch': 4.35} +04/19 [17:36:31] INFO | >> train_qwenlatent.py:487 + Step 17250 | grad_norm_pre_clip=0.2053 | + grad_norm_pre_clip_avg=0.1866 | Metrics: + {'align_loss': 0.025951772928237915, + 'recon_loss': 0.056895386427640915, + 'predict_loss': 0.011157297529280186, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2052772045135498, + 'mae_score': 0.016898641500387106, 'data_time': + 0.0006295850034803152, 'model_time': + 1.2032019949983805, 'grad_norm_pre_clip_avg': + 0.18657106757164002, 'learning_rate': + 2.070503712451896e-05, 'epoch': 4.35} +04/19 [17:36:44] INFO | >> train_qwenlatent.py:487 + Step 17260 | grad_norm_pre_clip=0.1741 | + grad_norm_pre_clip_avg=0.1793 | Metrics: + {'align_loss': 0.024854714050889015, + 'recon_loss': 0.0498289056122303, + 'predict_loss': 0.00863673072308302, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1740899682044983, + 'data_time': 0.0008188799838535488, + 'model_time': 1.2237427080108318, + 'grad_norm_pre_clip_avg': 0.17927368730306625, + 'learning_rate': 2.0698455631129808e-05, + 'epoch': 4.36} +04/19 [17:36:56] INFO | >> train_qwenlatent.py:487 + Step 17270 | grad_norm_pre_clip=0.1684 | + grad_norm_pre_clip_avg=0.3303 | Metrics: + {'align_loss': 0.024594727903604507, + 'recon_loss': 0.07119403034448624, + 'predict_loss': 0.013118790462613106, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16837233304977417, + 'data_time': 0.0008456130162812769, + 'model_time': 1.1966166050115135, + 'grad_norm_pre_clip_avg': 0.3303327694535255, + 'learning_rate': 2.069187014800532e-05, + 'epoch': 4.36} +04/19 [17:37:08] INFO | >> train_qwenlatent.py:487 + Step 17280 | grad_norm_pre_clip=0.1922 | + grad_norm_pre_clip_avg=0.2475 | Metrics: + {'align_loss': 0.024951662868261337, + 'recon_loss': 0.05470999702811241, + 'predict_loss': 0.008454417809844017, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19224707782268524, + 'data_time': 0.0006968679954297841, + 'model_time': 1.2389952300000004, + 'grad_norm_pre_clip_avg': 0.24752931147813798, + 'learning_rate': 2.068528067835519e-05, + 'epoch': 4.36} +04/19 [17:37:21] INFO | >> train_qwenlatent.py:487 + Step 17290 | grad_norm_pre_clip=0.1731 | + grad_norm_pre_clip_avg=0.2158 | Metrics: + {'align_loss': 0.02510051056742668, + 'recon_loss': 0.07050562649965286, + 'predict_loss': 0.012936006300151348, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17306460440158844, + 'data_time': 0.0009530269890092313, + 'model_time': 1.240037053998094, + 'grad_norm_pre_clip_avg': 0.21579011380672455, + 'learning_rate': 2.0678687225391043e-05, + 'epoch': 4.36} +04/19 [17:37:34] INFO | >> train_qwenlatent.py:487 + Step 17300 | grad_norm_pre_clip=0.1613 | + grad_norm_pre_clip_avg=0.1671 | Metrics: + {'align_loss': 0.026590976864099503, + 'recon_loss': 0.08800550550222397, + 'predict_loss': 0.012040264904499054, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16133280098438263, + 'mae_score': 0.013401339934752869, 'data_time': + 0.0009239999926649034, 'model_time': + 1.2318164280150086, 'grad_norm_pre_clip_avg': + 0.16708066463470458, 'learning_rate': + 2.0672089792326445e-05, 'epoch': 4.37} +04/19 [17:37:47] INFO | >> train_qwenlatent.py:487 + Step 17310 | grad_norm_pre_clip=0.1788 | + grad_norm_pre_clip_avg=0.2157 | Metrics: + {'align_loss': 0.024437034502625465, + 'recon_loss': 0.04933183267712593, + 'predict_loss': 0.010469481348991394, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17878682911396027, + 'data_time': 0.001183520013000816, + 'model_time': 1.254262150003342, + 'grad_norm_pre_clip_avg': 0.21571921408176423, + 'learning_rate': 2.0665488382376904e-05, + 'epoch': 4.37} +04/19 [17:38:00] INFO | >> train_qwenlatent.py:487 + Step 17320 | grad_norm_pre_clip=0.2579 | + grad_norm_pre_clip_avg=0.2195 | Metrics: + {'align_loss': 0.02503349632024765, + 'recon_loss': 0.06803326308727264, + 'predict_loss': 0.011790469288825989, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2578631341457367, + 'data_time': 0.0009326110011897981, + 'model_time': 1.2472526030032896, + 'grad_norm_pre_clip_avg': 0.2195470854640007, + 'learning_rate': 2.065888299875987e-05, + 'epoch': 4.37} +04/19 [17:38:13] INFO | >> train_qwenlatent.py:487 + Step 17330 | grad_norm_pre_clip=0.1961 | + grad_norm_pre_clip_avg=0.2046 | Metrics: + {'align_loss': 0.02490861341357231, + 'recon_loss': 0.0693487897515297, + 'predict_loss': 0.018702244386076927, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19612257182598114, + 'data_time': 0.0009829170012380928, + 'model_time': 1.2278780220076442, + 'grad_norm_pre_clip_avg': 0.20455019623041154, + 'learning_rate': 2.0652273644694725e-05, + 'epoch': 4.37} +04/19 [17:38:25] INFO | >> train_qwenlatent.py:487 + Step 17340 | grad_norm_pre_clip=0.1705 | + grad_norm_pre_clip_avg=0.2207 | Metrics: + {'align_loss': 0.02467118576169014, + 'recon_loss': 0.06493553519248962, + 'predict_loss': 0.012033161707222462, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1704961359500885, + 'data_time': 0.0006658700003754348, + 'model_time': 1.2520343470096122, + 'grad_norm_pre_clip_avg': 0.22073982208967208, + 'learning_rate': 2.0645660323402792e-05, + 'epoch': 4.38} +04/19 [17:38:38] INFO | >> train_qwenlatent.py:487 + Step 17350 | grad_norm_pre_clip=0.3067 | + grad_norm_pre_clip_avg=0.2080 | Metrics: + {'align_loss': 0.02425931952893734, + 'recon_loss': 0.04460229352116585, + 'predict_loss': 0.011120459996163845, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3066752851009369, + 'mae_score': 0.013062289813617328, 'data_time': + 0.0010209039901383221, 'model_time': + 1.245298786991043, 'grad_norm_pre_clip_avg': + 0.2080131947994232, 'learning_rate': + 2.0639043038107315e-05, 'epoch': 4.38} +04/19 [17:38:51] INFO | >> train_qwenlatent.py:487 + Step 17360 | grad_norm_pre_clip=0.1690 | + grad_norm_pre_clip_avg=0.2247 | Metrics: + {'align_loss': 0.026447750627994537, + 'recon_loss': 0.07438145577907562, + 'predict_loss': 0.012912927195429802, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16902293264865875, + 'data_time': 0.0007963200041558594, + 'model_time': 1.1891074159939308, + 'grad_norm_pre_clip_avg': 0.22474052160978317, + 'learning_rate': 2.0632421792033485e-05, + 'epoch': 4.38} +04/19 [17:39:03] INFO | >> train_qwenlatent.py:487 + Step 17370 | grad_norm_pre_clip=0.1470 | + grad_norm_pre_clip_avg=0.1692 | Metrics: + {'align_loss': 0.02604452520608902, + 'recon_loss': 0.06329550594091415, + 'predict_loss': 0.009614881128072739, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1470460146665573, + 'data_time': 0.000970191991655156, + 'model_time': 1.2921043620153796, + 'grad_norm_pre_clip_avg': 0.1692417472600937, + 'learning_rate': 2.062579658840842e-05, + 'epoch': 4.38} +04/19 [17:39:16] INFO | >> train_qwenlatent.py:487 + Step 17380 | grad_norm_pre_clip=0.1497 | + grad_norm_pre_clip_avg=0.2081 | Metrics: + {'align_loss': 0.02464098110795021, + 'recon_loss': 0.0449654683470726, + 'predict_loss': 0.009899148717522621, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14972177147865295, + 'data_time': 0.0008405580010730773, + 'model_time': 1.2124327110068407, + 'grad_norm_pre_clip_avg': 0.20805364698171616, + 'learning_rate': 2.061916743046115e-05, + 'epoch': 4.39} +04/19 [17:39:28] INFO | >> train_qwenlatent.py:487 + Step 17390 | grad_norm_pre_clip=0.2182 | + grad_norm_pre_clip_avg=0.2502 | Metrics: + {'align_loss': 0.024263601750135422, + 'recon_loss': 0.06569711863994598, + 'predict_loss': 0.012588266283273697, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21822163462638855, + 'data_time': 0.0008404790132772177, + 'model_time': 1.2514007369754836, + 'grad_norm_pre_clip_avg': 0.2502288147807121, + 'learning_rate': 2.0612534321422667e-05, + 'epoch': 4.39} +04/19 [17:39:41] INFO | >> train_qwenlatent.py:487 + Step 17400 | grad_norm_pre_clip=0.2185 | + grad_norm_pre_clip_avg=0.2079 | Metrics: + {'align_loss': 0.024009833112359047, + 'recon_loss': 0.0704425573348999, + 'predict_loss': 0.016849588602781296, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2184910625219345, + 'mae_score': 0.015226580645586994, 'data_time': + 0.0008411599847022444, 'model_time': + 1.2582048710028175, 'grad_norm_pre_clip_avg': + 0.2079262688755989, 'learning_rate': + 2.060589726452585e-05, 'epoch': 4.39} +04/19 [17:39:54] INFO | >> train_qwenlatent.py:487 + Step 17410 | grad_norm_pre_clip=0.1822 | + grad_norm_pre_clip_avg=0.1947 | Metrics: + {'align_loss': 0.024433251470327377, + 'recon_loss': 0.05214524641633034, + 'predict_loss': 0.008242765441536903, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18215592205524445, + 'data_time': 0.0009494350233580917, + 'model_time': 1.2000442869903054, + 'grad_norm_pre_clip_avg': 0.19467558711767197, + 'learning_rate': 2.0599256263005527e-05, + 'epoch': 4.39} +04/19 [17:40:07] INFO | >> train_qwenlatent.py:487 + Step 17420 | grad_norm_pre_clip=0.1656 | + grad_norm_pre_clip_avg=0.1695 | Metrics: + {'align_loss': 0.024287709966301918, + 'recon_loss': 0.06297402828931808, + 'predict_loss': 0.009663944132626057, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1655927151441574, + 'data_time': 0.0009604399965610355, + 'model_time': 1.2120560689945705, + 'grad_norm_pre_clip_avg': 0.16952629387378693, + 'learning_rate': 2.0592611320098437e-05, + 'epoch': 4.4} +04/19 [17:40:19] INFO | >> train_qwenlatent.py:487 + Step 17430 | grad_norm_pre_clip=0.2325 | + grad_norm_pre_clip_avg=0.2770 | Metrics: + {'align_loss': 0.024472881108522415, + 'recon_loss': 0.06966458261013031, + 'predict_loss': 0.01419293787330389, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23251482844352722, + 'data_time': 0.0010497329931240529, + 'model_time': 1.233611020987155, + 'grad_norm_pre_clip_avg': 0.2770431488752365, + 'learning_rate': 2.0585962439043253e-05, + 'epoch': 4.4} +04/19 [17:40:33] INFO | >> train_qwenlatent.py:487 + Step 17440 | grad_norm_pre_clip=0.2001 | + grad_norm_pre_clip_avg=0.2204 | Metrics: + {'align_loss': 0.026524091139435768, + 'recon_loss': 0.08947587758302689, + 'predict_loss': 0.019591834396123886, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20006534457206726, + 'data_time': 0.0009984510252252221, + 'model_time': 1.2548006600118242, + 'grad_norm_pre_clip_avg': 0.22036688029766083, + 'learning_rate': 2.057930962308055e-05, + 'epoch': 4.4} +04/19 [17:40:46] INFO | >> train_qwenlatent.py:487 + Step 17450 | grad_norm_pre_clip=0.2154 | + grad_norm_pre_clip_avg=0.2052 | Metrics: + {'align_loss': 0.023307349532842636, + 'recon_loss': 0.06555451452732086, + 'predict_loss': 0.014914215542376041, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21544714272022247, + 'mae_score': 0.01499217093527854, 'data_time': + 0.0010125190019607544, 'model_time': + 1.2420463779999409, 'grad_norm_pre_clip_avg': + 0.2051752105355263, 'learning_rate': + 2.0572652875452833e-05, 'epoch': 4.4} +04/19 [17:40:58] INFO | >> train_qwenlatent.py:487 + Step 17460 | grad_norm_pre_clip=0.2315 | + grad_norm_pre_clip_avg=0.2219 | Metrics: + {'align_loss': 0.024291854351758957, + 'recon_loss': 0.06863292306661606, + 'predict_loss': 0.01786629669368267, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23153497278690338, + 'data_time': 0.000882866996107623, + 'model_time': 1.241409187001409, + 'grad_norm_pre_clip_avg': 0.22193659096956253, + 'learning_rate': 2.0565992199404523e-05, + 'epoch': 4.41} +04/19 [17:41:11] INFO | >> train_qwenlatent.py:487 + Step 17470 | grad_norm_pre_clip=0.2214 | + grad_norm_pre_clip_avg=0.2028 | Metrics: + {'align_loss': 0.02492235042154789, + 'recon_loss': 0.06769829243421555, + 'predict_loss': 0.016371604055166245, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22138087451457977, + 'data_time': 0.0008441829995717853, + 'model_time': 1.223500731983222, + 'grad_norm_pre_clip_avg': 0.2027878761291504, + 'learning_rate': 2.055932759818195e-05, + 'epoch': 4.41} +04/19 [17:41:23] INFO | >> train_qwenlatent.py:487 + Step 17480 | grad_norm_pre_clip=0.1585 | + grad_norm_pre_clip_avg=0.1842 | Metrics: + {'align_loss': 0.026020752266049385, + 'recon_loss': 0.06891877949237823, + 'predict_loss': 0.012874200008809566, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15852580964565277, + 'data_time': 0.0006434189854189754, + 'model_time': 1.283967362018302, + 'grad_norm_pre_clip_avg': 0.1842390552163124, + 'learning_rate': 2.055265907503336e-05, + 'epoch': 4.41} +04/19 [17:41:36] INFO | >> train_qwenlatent.py:487 + Step 17490 | grad_norm_pre_clip=0.2424 | + grad_norm_pre_clip_avg=0.2491 | Metrics: + {'align_loss': 0.02583305910229683, + 'recon_loss': 0.05308675020933151, + 'predict_loss': 0.008259601891040802, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24242201447486877, + 'data_time': 0.0006255019980017096, + 'model_time': 1.2322883179876953, + 'grad_norm_pre_clip_avg': 0.24905396699905397, + 'learning_rate': 2.0545986633208907e-05, + 'epoch': 4.41} +04/19 [17:41:49] INFO | >> train_qwenlatent.py:487 + Step 17500 | grad_norm_pre_clip=0.1869 | + grad_norm_pre_clip_avg=0.1979 | Metrics: + {'align_loss': 0.026095978915691376, + 'recon_loss': 0.06751787662506104, + 'predict_loss': 0.012063512578606606, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18687571585178375, + 'mae_score': 0.012869437965186867, 'data_time': + 0.0006905720219947398, 'model_time': + 1.2055762019881513, 'grad_norm_pre_clip_avg': + 0.19785629212856293, 'learning_rate': + 2.053931027596066e-05, 'epoch': 4.42} +04/19 [17:42:02] INFO | >> train_qwenlatent.py:487 + Step 17510 | grad_norm_pre_clip=0.2379 | + grad_norm_pre_clip_avg=0.2027 | Metrics: + {'align_loss': 0.024781687185168266, + 'recon_loss': 0.04913438856601715, + 'predict_loss': 0.01387299969792366, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23785433173179626, + 'data_time': 0.0006591709970962256, + 'model_time': 1.191844889021013, + 'grad_norm_pre_clip_avg': 0.20274768471717836, + 'learning_rate': 2.05326300065426e-05, 'epoch': + 4.42} +04/19 [17:42:15] INFO | >> train_qwenlatent.py:487 + Step 17520 | grad_norm_pre_clip=0.2067 | + grad_norm_pre_clip_avg=0.2188 | Metrics: + {'align_loss': 0.025243151932954788, + 'recon_loss': 0.053536828607320786, + 'predict_loss': 0.007703710813075304, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20668306946754456, + 'data_time': 0.0010485270177014172, + 'model_time': 1.2196451229974627, + 'grad_norm_pre_clip_avg': 0.2187730699777603, + 'learning_rate': 2.05259458282106e-05, 'epoch': + 4.42} +04/19 [17:42:27] INFO | >> train_qwenlatent.py:487 + Step 17530 | grad_norm_pre_clip=0.2460 | + grad_norm_pre_clip_avg=0.2402 | Metrics: + {'align_loss': 0.0248238667845726, + 'recon_loss': 0.07480369508266449, + 'predict_loss': 0.011237042024731636, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24595299363136292, + 'data_time': 0.0006811420025769621, + 'model_time': 1.2332435100106522, + 'grad_norm_pre_clip_avg': 0.2401990994811058, + 'learning_rate': 2.0519257744222457e-05, + 'epoch': 4.42} +04/19 [17:42:40] INFO | >> train_qwenlatent.py:487 + Step 17540 | grad_norm_pre_clip=0.1860 | + grad_norm_pre_clip_avg=0.1926 | Metrics: + {'align_loss': 0.02425127848982811, + 'recon_loss': 0.07153069972991943, + 'predict_loss': 0.018299126997590065, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1860181838274002, + 'data_time': 0.0009212370205204934, + 'model_time': 1.2763703720120247, + 'grad_norm_pre_clip_avg': 0.19256100207567214, + 'learning_rate': 2.0512565757837852e-05, + 'epoch': 4.43} +04/19 [17:42:53] INFO | >> train_qwenlatent.py:487 + Step 17550 | grad_norm_pre_clip=0.1600 | + grad_norm_pre_clip_avg=0.2150 | Metrics: + {'align_loss': 0.025602007284760475, + 'recon_loss': 0.06370712071657181, + 'predict_loss': 0.00751787843182683, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15998515486717224, + 'mae_score': 0.012097969570675412, 'data_time': + 0.0010370629897806793, 'model_time': + 1.2260860030073673, 'grad_norm_pre_clip_avg': + 0.21504744738340378, 'learning_rate': + 2.0505869872318383e-05, 'epoch': 4.43} +04/19 [17:43:05] INFO | >> train_qwenlatent.py:487 + Step 17560 | grad_norm_pre_clip=0.2651 | + grad_norm_pre_clip_avg=0.2451 | Metrics: + {'align_loss': 0.02441047877073288, + 'recon_loss': 0.06887594610452652, + 'predict_loss': 0.01118494477123022, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2651280164718628, + 'data_time': 0.0008631320088170469, + 'model_time': 1.2156081889988855, + 'grad_norm_pre_clip_avg': 0.2450798586010933, + 'learning_rate': 2.0499170090927543e-05, + 'epoch': 4.43} +04/19 [17:43:18] INFO | >> train_qwenlatent.py:487 + Step 17570 | grad_norm_pre_clip=0.2858 | + grad_norm_pre_clip_avg=0.2241 | Metrics: + {'align_loss': 0.025874067097902298, + 'recon_loss': 0.06162095069885254, + 'predict_loss': 0.01368635892868042, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28578847646713257, + 'data_time': 0.0009479119908064604, + 'model_time': 1.2579246009991039, + 'grad_norm_pre_clip_avg': 0.22411680817604065, + 'learning_rate': 2.0492466416930726e-05, + 'epoch': 4.43} +04/19 [17:43:31] INFO | >> train_qwenlatent.py:487 + Step 17580 | grad_norm_pre_clip=0.1513 | + grad_norm_pre_clip_avg=0.2072 | Metrics: + {'align_loss': 0.02438192069530487, + 'recon_loss': 0.05092325061559677, + 'predict_loss': 0.011290301568806171, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1513463407754898, + 'data_time': 0.0010908159893006086, + 'model_time': 1.2109099919907749, + 'grad_norm_pre_clip_avg': 0.20721698850393294, + 'learning_rate': 2.0485758853595214e-05, + 'epoch': 4.44} +04/19 [17:43:44] INFO | >> train_qwenlatent.py:487 + Step 17590 | grad_norm_pre_clip=0.2279 | + grad_norm_pre_clip_avg=0.2137 | Metrics: + {'align_loss': 0.0253467820584774, + 'recon_loss': 0.04907425865530968, + 'predict_loss': 0.007412041071802378, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22788354754447937, + 'data_time': 0.0007594609924126416, + 'model_time': 1.2383133760013152, + 'grad_norm_pre_clip_avg': 0.21367171555757522, + 'learning_rate': 2.04790474041902e-05, 'epoch': + 4.44} +04/19 [17:43:57] INFO | >> train_qwenlatent.py:487 + Step 17600 | grad_norm_pre_clip=0.2763 | + grad_norm_pre_clip_avg=0.2542 | Metrics: + {'align_loss': 0.024960078299045563, + 'recon_loss': 0.054989006370306015, + 'predict_loss': 0.008281240239739418, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2762642502784729, + 'mae_score': 0.012893345119716884, 'data_time': + 0.0006726520077791065, 'model_time': + 1.2631636429869104, 'grad_norm_pre_clip_avg': + 0.25418794304132464, 'learning_rate': + 2.0472332071986763e-05, 'epoch': 4.44} +04/19 [17:44:09] INFO | >> train_qwenlatent.py:487 + Step 17610 | grad_norm_pre_clip=0.1819 | + grad_norm_pre_clip_avg=0.2262 | Metrics: + {'align_loss': 0.025994030758738518, + 'recon_loss': 0.06371385604143143, + 'predict_loss': 0.012683916836977005, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18191245198249817, + 'data_time': 0.0008014409977477044, + 'model_time': 1.2089424839941785, + 'grad_norm_pre_clip_avg': 0.22622709572315217, + 'learning_rate': 2.046561286025787e-05, + 'epoch': 4.44} +04/19 [17:44:22] INFO | >> train_qwenlatent.py:487 + Step 17620 | grad_norm_pre_clip=0.1728 | + grad_norm_pre_clip_avg=0.1789 | Metrics: + {'align_loss': 0.024768464267253876, + 'recon_loss': 0.06294072419404984, + 'predict_loss': 0.010034807957708836, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17276059091091156, + 'data_time': 0.0009075840061996132, + 'model_time': 1.2546003850002307, + 'grad_norm_pre_clip_avg': 0.17892604619264602, + 'learning_rate': 2.0458889772278385e-05, + 'epoch': 4.45} +04/19 [17:44:35] INFO | >> train_qwenlatent.py:487 + Step 17630 | grad_norm_pre_clip=0.1722 | + grad_norm_pre_clip_avg=0.1842 | Metrics: + {'align_loss': 0.024297188967466354, + 'recon_loss': 0.049537915736436844, + 'predict_loss': 0.0091707743704319, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17221683263778687, + 'data_time': 0.000713790999725461, + 'model_time': 1.1886168729979545, + 'grad_norm_pre_clip_avg': 0.18420236110687255, + 'learning_rate': 2.0452162811325053e-05, + 'epoch': 4.45} +04/19 [17:44:47] INFO | >> train_qwenlatent.py:487 + Step 17640 | grad_norm_pre_clip=0.2213 | + grad_norm_pre_clip_avg=0.2671 | Metrics: + {'align_loss': 0.02416369691491127, + 'recon_loss': 0.0631549209356308, + 'predict_loss': 0.014010947197675705, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22126290202140808, + 'data_time': 0.0007284950115717947, + 'model_time': 1.2074837139807642, + 'grad_norm_pre_clip_avg': 0.26707444041967393, + 'learning_rate': 2.0445431980676526e-05, + 'epoch': 4.45} +04/19 [17:45:01] INFO | >> train_qwenlatent.py:487 + Step 17650 | grad_norm_pre_clip=0.1848 | + grad_norm_pre_clip_avg=0.1883 | Metrics: + {'align_loss': 0.025440389290452003, + 'recon_loss': 0.061449069529771805, + 'predict_loss': 0.015166270546615124, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18477874994277954, + 'mae_score': 0.012240041268838419, 'data_time': + 0.0006999179895501584, 'model_time': + 1.5343546150252223, 'grad_norm_pre_clip_avg': + 0.18834285885095597, 'learning_rate': + 2.043869728361332e-05, 'epoch': 4.45} +04/19 [17:45:14] INFO | >> train_qwenlatent.py:487 + Step 17660 | grad_norm_pre_clip=0.1709 | + grad_norm_pre_clip_avg=0.2177 | Metrics: + {'align_loss': 0.02419167011976242, + 'recon_loss': 0.05679875239729881, + 'predict_loss': 0.010250119492411613, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17089049518108368, + 'data_time': 0.0009470509830862284, + 'model_time': 1.2204316769784782, + 'grad_norm_pre_clip_avg': 0.21767996102571488, + 'learning_rate': 2.0431958723417842e-05, + 'epoch': 4.46} +04/19 [17:45:26] INFO | >> train_qwenlatent.py:487 + Step 17670 | grad_norm_pre_clip=0.1983 | + grad_norm_pre_clip_avg=0.2212 | Metrics: + {'align_loss': 0.025442276149988174, + 'recon_loss': 0.05826368182897568, + 'predict_loss': 0.010181430727243423, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19825205206871033, + 'data_time': 0.0007307709893211722, + 'model_time': 1.2332024729985278, + 'grad_norm_pre_clip_avg': 0.22123533338308335, + 'learning_rate': 2.0425216303374394e-05, + 'epoch': 4.46} +04/19 [17:45:39] INFO | >> train_qwenlatent.py:487 + Step 17680 | grad_norm_pre_clip=0.2148 | + grad_norm_pre_clip_avg=0.2085 | Metrics: + {'align_loss': 0.024833757430315018, + 'recon_loss': 0.06752865016460419, + 'predict_loss': 0.015873214229941368, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21479684114456177, + 'data_time': 0.0008474240021314472, + 'model_time': 1.2434826559911016, + 'grad_norm_pre_clip_avg': 0.20849049389362334, + 'learning_rate': 2.041847002676914e-05, + 'epoch': 4.46} +04/19 [17:45:52] INFO | >> train_qwenlatent.py:487 + Step 17690 | grad_norm_pre_clip=0.2230 | + grad_norm_pre_clip_avg=0.2079 | Metrics: + {'align_loss': 0.025324072688817978, + 'recon_loss': 0.07972901314496994, + 'predict_loss': 0.014994224533438683, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22302201390266418, + 'data_time': 0.0006554980063810945, + 'model_time': 1.2458110349834897, + 'grad_norm_pre_clip_avg': 0.20794908106327056, + 'learning_rate': 2.0411719896890145e-05, + 'epoch': 4.46} +04/19 [17:46:05] INFO | >> train_qwenlatent.py:487 + Step 17700 | grad_norm_pre_clip=0.1611 | + grad_norm_pre_clip_avg=0.2175 | Metrics: + {'align_loss': 0.025432761758565903, + 'recon_loss': 0.05479125678539276, + 'predict_loss': 0.00911195669323206, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1611393243074417, + 'mae_score': 0.015610390740471916, 'data_time': + 0.0007983329996932298, 'model_time': + 1.251768185000401, 'grad_norm_pre_clip_avg': + 0.21754382252693177, 'learning_rate': + 2.0404965917027328e-05, 'epoch': 4.47} +04/19 [17:46:17] INFO | >> train_qwenlatent.py:487 + Step 17710 | grad_norm_pre_clip=0.1945 | + grad_norm_pre_clip_avg=0.1826 | Metrics: + {'align_loss': 0.026281137019395828, + 'recon_loss': 0.06762915104627609, + 'predict_loss': 0.010039747692644596, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19446738064289093, + 'data_time': 0.0009105909848585725, + 'model_time': 1.1999162909924053, + 'grad_norm_pre_clip_avg': 0.1825639694929123, + 'learning_rate': 2.03982080904725e-05, 'epoch': + 4.47} +04/19 [17:46:30] INFO | >> train_qwenlatent.py:487 + Step 17720 | grad_norm_pre_clip=0.1769 | + grad_norm_pre_clip_avg=0.2076 | Metrics: + {'align_loss': 0.024710144847631454, + 'recon_loss': 0.04914757236838341, + 'predict_loss': 0.008338910527527332, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17687273025512695, + 'data_time': 0.0007665599987376481, + 'model_time': 1.2003460800042376, + 'grad_norm_pre_clip_avg': 0.20757946074008943, + 'learning_rate': 2.0391446420519348e-05, + 'epoch': 4.47} +04/19 [17:46:43] INFO | >> train_qwenlatent.py:487 + Step 17730 | grad_norm_pre_clip=0.2043 | + grad_norm_pre_clip_avg=0.2483 | Metrics: + {'align_loss': 0.02493578940629959, + 'recon_loss': 0.06312362104654312, + 'predict_loss': 0.01048540510237217, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20433343946933746, + 'data_time': 0.000914703996386379, + 'model_time': 1.292767124978127, + 'grad_norm_pre_clip_avg': 0.2483215406537056, + 'learning_rate': 2.0384680910463426e-05, + 'epoch': 4.47} +04/19 [17:46:56] INFO | >> train_qwenlatent.py:487 + Step 17740 | grad_norm_pre_clip=0.2508 | + grad_norm_pre_clip_avg=0.2280 | Metrics: + {'align_loss': 0.025021906942129135, + 'recon_loss': 0.05938591808080673, + 'predict_loss': 0.01055710669606924, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2508483827114105, + 'data_time': 0.000932486989768222, + 'model_time': 1.1984622159798164, + 'grad_norm_pre_clip_avg': 0.22798957973718642, + 'learning_rate': 2.0377911563602157e-05, + 'epoch': 4.48} +04/19 [17:47:09] INFO | >> train_qwenlatent.py:487 + Step 17750 | grad_norm_pre_clip=0.2721 | + grad_norm_pre_clip_avg=0.2343 | Metrics: + {'align_loss': 0.025735333561897278, + 'recon_loss': 0.07560843229293823, + 'predict_loss': 0.021249432116746902, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.27214744687080383, + 'mae_score': 0.013307459719546206, 'data_time': + 0.0007311809749808162, 'model_time': + 1.2350654370093253, 'grad_norm_pre_clip_avg': + 0.23429605960845948, 'learning_rate': + 2.037113838323485e-05, 'epoch': 4.48} +04/19 [17:47:21] INFO | >> train_qwenlatent.py:487 + Step 17760 | grad_norm_pre_clip=0.2262 | + grad_norm_pre_clip_avg=0.2030 | Metrics: + {'align_loss': 0.02547306753695011, + 'recon_loss': 0.06564745306968689, + 'predict_loss': 0.011639189906418324, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2262057512998581, + 'data_time': 0.0007768299838062376, + 'model_time': 1.238162764988374, + 'grad_norm_pre_clip_avg': 0.2030057579278946, + 'learning_rate': 2.0364361372662656e-05, + 'epoch': 4.48} +04/19 [17:47:34] INFO | >> train_qwenlatent.py:487 + Step 17770 | grad_norm_pre_clip=0.2132 | + grad_norm_pre_clip_avg=0.2213 | Metrics: + {'align_loss': 0.02402902953326702, + 'recon_loss': 0.0727325826883316, + 'predict_loss': 0.012806777842342854, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21323689818382263, + 'data_time': 0.0009317199874203652, + 'model_time': 1.213738448015647, + 'grad_norm_pre_clip_avg': 0.22127123326063156, + 'learning_rate': 2.0357580535188615e-05, + 'epoch': 4.48} +04/19 [17:47:47] INFO | >> train_qwenlatent.py:487 + Step 17780 | grad_norm_pre_clip=0.2019 | + grad_norm_pre_clip_avg=0.2231 | Metrics: + {'align_loss': 0.025406256318092346, + 'recon_loss': 0.05389726907014847, + 'predict_loss': 0.015909986570477486, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20193302631378174, + 'data_time': 0.0008883939881343395, + 'model_time': 1.234781215986004, + 'grad_norm_pre_clip_avg': 0.223146852850914, + 'learning_rate': 2.0350795874117623e-05, + 'epoch': 4.49} +04/19 [17:47:59] INFO | >> train_qwenlatent.py:487 + Step 17790 | grad_norm_pre_clip=0.1743 | + grad_norm_pre_clip_avg=0.2083 | Metrics: + {'align_loss': 0.02549627050757408, + 'recon_loss': 0.06832228600978851, + 'predict_loss': 0.01576846092939377, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1743413656949997, + 'data_time': 0.0011007279972545803, + 'model_time': 1.269544840004528, + 'grad_norm_pre_clip_avg': 0.2083320900797844, + 'learning_rate': 2.0344007392756445e-05, + 'epoch': 4.49} +04/19 [17:48:13] INFO | >> train_qwenlatent.py:487 + Step 17800 | grad_norm_pre_clip=0.1807 | + grad_norm_pre_clip_avg=0.2026 | Metrics: + {'align_loss': 0.02415642887353897, + 'recon_loss': 0.05821261927485466, + 'predict_loss': 0.012991308234632015, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1806897073984146, + 'mae_score': 0.011660000225445171, 'data_time': + 0.0010369319934397936, 'model_time': + 1.2194162940140814, 'grad_norm_pre_clip_avg': + 0.2025606781244278, 'learning_rate': + 2.0337215094413704e-05, 'epoch': 4.49} +04/19 [17:48:25] INFO | >> train_qwenlatent.py:487 + Step 17810 | grad_norm_pre_clip=0.2004 | + grad_norm_pre_clip_avg=0.2155 | Metrics: + {'align_loss': 0.02472061850130558, + 'recon_loss': 0.073575459420681, + 'predict_loss': 0.012816611677408218, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2004217803478241, + 'data_time': 0.0009432030201423913, + 'model_time': 1.2271965880063362, + 'grad_norm_pre_clip_avg': 0.2154819682240486, + 'learning_rate': 2.0330418982399878e-05, + 'epoch': 4.49} +04/19 [17:48:38] INFO | >> train_qwenlatent.py:487 + Step 17820 | grad_norm_pre_clip=0.2470 | + grad_norm_pre_clip_avg=0.1891 | Metrics: + {'align_loss': 0.02462785691022873, + 'recon_loss': 0.0705968514084816, + 'predict_loss': 0.012935403734445572, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2470044195652008, + 'data_time': 0.0009037730051204562, + 'model_time': 1.2455096479970962, + 'grad_norm_pre_clip_avg': 0.1890639677643776, + 'learning_rate': 2.0323619060027312e-05, + 'epoch': 4.5} +04/19 [17:48:50] INFO | >> train_qwenlatent.py:487 + Step 17830 | grad_norm_pre_clip=0.2745 | + grad_norm_pre_clip_avg=0.2785 | Metrics: + {'align_loss': 0.02573658712208271, + 'recon_loss': 0.05590466409921646, + 'predict_loss': 0.011542506515979767, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2745453119277954, + 'data_time': 0.0006644840177614242, + 'model_time': 1.2299385909864213, + 'grad_norm_pre_clip_avg': 0.278530552983284, + 'learning_rate': 2.0316815330610213e-05, + 'epoch': 4.5} +04/19 [17:49:03] INFO | >> train_qwenlatent.py:487 + Step 17840 | grad_norm_pre_clip=0.2117 | + grad_norm_pre_clip_avg=0.2140 | Metrics: + {'align_loss': 0.023779476061463356, + 'recon_loss': 0.0644194483757019, + 'predict_loss': 0.013815663754940033, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21173866093158722, + 'data_time': 0.000688453990733251, + 'model_time': 1.2346389850135893, + 'grad_norm_pre_clip_avg': 0.21402800977230071, + 'learning_rate': 2.0310007797464623e-05, + 'epoch': 4.5} +04/19 [17:49:16] INFO | >> train_qwenlatent.py:487 + Step 17850 | grad_norm_pre_clip=0.1772 | + grad_norm_pre_clip_avg=0.1980 | Metrics: + {'align_loss': 0.02443850040435791, + 'recon_loss': 0.05366222932934761, + 'predict_loss': 0.01233105082064867, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17722788453102112, + 'mae_score': 0.01212001491237331, 'data_time': + 0.0006924739864189178, 'model_time': + 1.2516537659976166, 'grad_norm_pre_clip_avg': + 0.19801403284072877, 'learning_rate': + 2.030319646390846e-05, 'epoch': 4.5} +04/19 [17:49:29] INFO | >> train_qwenlatent.py:487 + Step 17860 | grad_norm_pre_clip=0.1989 | + grad_norm_pre_clip_avg=0.2002 | Metrics: + {'align_loss': 0.02572120726108551, + 'recon_loss': 0.06038994342088699, + 'predict_loss': 0.012581064365804195, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1989421397447586, + 'data_time': 0.001017310976749286, + 'model_time': 1.2366913449950516, + 'grad_norm_pre_clip_avg': 0.20016079097986222, + 'learning_rate': 2.0296381333261487e-05, + 'epoch': 4.51} +04/19 [17:49:41] INFO | >> train_qwenlatent.py:487 + Step 17870 | grad_norm_pre_clip=0.3262 | + grad_norm_pre_clip_avg=0.2309 | Metrics: + {'align_loss': 0.02509007602930069, + 'recon_loss': 0.0639440044760704, + 'predict_loss': 0.005992548540234566, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3262033760547638, + 'data_time': 0.000979578006081283, + 'model_time': 1.2493687690002844, + 'grad_norm_pre_clip_avg': 0.23090262413024903, + 'learning_rate': 2.0289562408845306e-05, + 'epoch': 4.51} +04/19 [17:49:53] INFO | >> train_qwenlatent.py:487 + Step 17880 | grad_norm_pre_clip=0.2273 | + grad_norm_pre_clip_avg=0.2500 | Metrics: + {'align_loss': 0.023863591253757477, + 'recon_loss': 0.0629168301820755, + 'predict_loss': 0.01743200607597828, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2273350954055786, + 'data_time': 0.0007174769998528063, + 'model_time': 1.2236184299981687, + 'grad_norm_pre_clip_avg': 0.25001627802848814, + 'learning_rate': 2.028273969398339e-05, + 'epoch': 4.51} +04/19 [17:50:06] INFO | >> train_qwenlatent.py:487 + Step 17890 | grad_norm_pre_clip=0.2324 | + grad_norm_pre_clip_avg=0.2175 | Metrics: + {'align_loss': 0.02551259472966194, + 'recon_loss': 0.07312899827957153, + 'predict_loss': 0.016664639115333557, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2323535978794098, + 'data_time': 0.0007875640003476292, + 'model_time': 1.204446726012975, + 'grad_norm_pre_clip_avg': 0.21749680936336518, + 'learning_rate': 2.0275913192001033e-05, + 'epoch': 4.51} +04/19 [17:50:19] INFO | >> train_qwenlatent.py:487 + Step 17900 | grad_norm_pre_clip=0.2035 | + grad_norm_pre_clip_avg=0.2001 | Metrics: + {'align_loss': 0.025138376280665398, + 'recon_loss': 0.05857458710670471, + 'predict_loss': 0.011238069273531437, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20354512333869934, + 'mae_score': 0.012520295220452386, 'data_time': + 0.0007745600014459342, 'model_time': + 1.2135742169921286, 'grad_norm_pre_clip_avg': + 0.200124654173851, 'learning_rate': + 2.02690829062254e-05, 'epoch': 4.52} +04/19 [17:50:32] INFO | >> train_qwenlatent.py:487 + Step 17910 | grad_norm_pre_clip=0.2802 | + grad_norm_pre_clip_avg=0.2322 | Metrics: + {'align_loss': 0.025161094963550568, + 'recon_loss': 0.0748823881149292, + 'predict_loss': 0.01989702880382538, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2802141010761261, + 'data_time': 0.0006156010203994811, + 'model_time': 1.3154610189958476, + 'grad_norm_pre_clip_avg': 0.2321801960468292, + 'learning_rate': 2.026224883998549e-05, + 'epoch': 4.52} +04/19 [17:50:45] INFO | >> train_qwenlatent.py:487 + Step 17920 | grad_norm_pre_clip=0.1788 | + grad_norm_pre_clip_avg=0.2065 | Metrics: + {'align_loss': 0.02389506995677948, + 'recon_loss': 0.056325964629650116, + 'predict_loss': 0.0093766488134861, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17875854671001434, + 'data_time': 0.0007612170011270791, + 'model_time': 1.2562180340173654, + 'grad_norm_pre_clip_avg': 0.2064526543021202, + 'learning_rate': 2.025541099661214e-05, + 'epoch': 4.52} +04/19 [17:50:57] INFO | >> train_qwenlatent.py:487 + Step 17930 | grad_norm_pre_clip=0.2264 | + grad_norm_pre_clip_avg=0.2045 | Metrics: + {'align_loss': 0.0246862955391407, + 'recon_loss': 0.06949625164270401, + 'predict_loss': 0.017901618033647537, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22636207938194275, + 'data_time': 0.0007472689903806895, + 'model_time': 1.2203176930197515, + 'grad_norm_pre_clip_avg': 0.20449215322732925, + 'learning_rate': 2.0248569379438027e-05, + 'epoch': 4.52} +04/19 [17:51:10] INFO | >> train_qwenlatent.py:487 + Step 17940 | grad_norm_pre_clip=0.2165 | + grad_norm_pre_clip_avg=0.2518 | Metrics: + {'align_loss': 0.02526050992310047, + 'recon_loss': 0.06880621612071991, + 'predict_loss': 0.013788917101919651, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21645474433898926, + 'data_time': 0.0007523969979956746, + 'model_time': 1.2440598839893937, + 'grad_norm_pre_clip_avg': 0.2518141582608223, + 'learning_rate': 2.0241723991797684e-05, + 'epoch': 4.53} +04/19 [17:51:23] INFO | >> train_qwenlatent.py:487 + Step 17950 | grad_norm_pre_clip=0.1831 | + grad_norm_pre_clip_avg=0.2095 | Metrics: + {'align_loss': 0.026038458570837975, + 'recon_loss': 0.07816969603300095, + 'predict_loss': 0.012074117548763752, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18306182324886322, + 'mae_score': 0.011572993338645041, 'data_time': + 0.0006443169841077179, 'model_time': + 1.2950943100149743, 'grad_norm_pre_clip_avg': + 0.20946163982152938, 'learning_rate': + 2.0234874837027456e-05, 'epoch': 4.53} +04/19 [17:51:35] INFO | >> train_qwenlatent.py:487 + Step 17960 | grad_norm_pre_clip=0.2061 | + grad_norm_pre_clip_avg=0.1886 | Metrics: + {'align_loss': 0.024484656751155853, + 'recon_loss': 0.0772535651922226, + 'predict_loss': 0.01748078316450119, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20606794953346252, + 'data_time': 0.0007094739994499832, + 'model_time': 1.2252517649903893, + 'grad_norm_pre_clip_avg': 0.1885611519217491, + 'learning_rate': 2.0228021918465554e-05, + 'epoch': 4.53} +04/19 [17:51:48] INFO | >> train_qwenlatent.py:487 + Step 17970 | grad_norm_pre_clip=0.1675 | + grad_norm_pre_clip_avg=0.1860 | Metrics: + {'align_loss': 0.02580593153834343, + 'recon_loss': 0.06891895830631256, + 'predict_loss': 0.009632195346057415, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16747982800006866, + 'data_time': 0.00111057999311015, 'model_time': + 1.2730383590096608, 'grad_norm_pre_clip_avg': + 0.18598570078611373, 'learning_rate': + 2.022116523945199e-05, 'epoch': 4.53} +04/19 [17:52:01] INFO | >> train_qwenlatent.py:487 + Step 17980 | grad_norm_pre_clip=0.2662 | + grad_norm_pre_clip_avg=0.2646 | Metrics: + {'align_loss': 0.02588868886232376, + 'recon_loss': 0.09206992387771606, + 'predict_loss': 0.013287242501974106, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2662425637245178, + 'data_time': 0.0009444019815418869, + 'model_time': 1.1973117709858343, + 'grad_norm_pre_clip_avg': 0.26459841430187225, + 'learning_rate': 2.0214304803328637e-05, + 'epoch': 4.54} +04/19 [17:52:13] INFO | >> train_qwenlatent.py:487 + Step 17990 | grad_norm_pre_clip=0.2740 | + grad_norm_pre_clip_avg=0.2349 | Metrics: + {'align_loss': 0.025102602317929268, + 'recon_loss': 0.05894440785050392, + 'predict_loss': 0.010085122659802437, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2739650309085846, + 'data_time': 0.0011076200171373785, + 'model_time': 1.2498317919962574, + 'grad_norm_pre_clip_avg': 0.2349249690771103, + 'learning_rate': 2.0207440613439185e-05, + 'epoch': 4.54} +04/19 [17:52:26] INFO | >> train_qwenlatent.py:487 + Step 18000 | grad_norm_pre_clip=0.1725 | + grad_norm_pre_clip_avg=0.2114 | Metrics: + {'align_loss': 0.025423113256692886, + 'recon_loss': 0.07225184142589569, + 'predict_loss': 0.009129345417022705, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1725052297115326, + 'mae_score': 0.013167979266192462, 'data_time': + 0.0011401239898987114, 'model_time': + 1.2928741550131235, 'grad_norm_pre_clip_avg': + 0.21138549000024795, 'learning_rate': + 2.0200572673129156e-05, 'epoch': 4.54} +04/19 [17:52:39] INFO | >> train_qwenlatent.py:487 + Step 18010 | grad_norm_pre_clip=0.2955 | + grad_norm_pre_clip_avg=0.2155 | Metrics: + {'align_loss': 0.0238641407340765, + 'recon_loss': 0.06724764406681061, + 'predict_loss': 0.0139704504981637, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2955055236816406, + 'data_time': 0.0006901889864820987, + 'model_time': 1.2622479090059642, + 'grad_norm_pre_clip_avg': 0.21554564386606218, + 'learning_rate': 2.0193700985745908e-05, + 'epoch': 4.54} +04/19 [17:52:51] INFO | >> train_qwenlatent.py:487 + Step 18020 | grad_norm_pre_clip=0.2001 | + grad_norm_pre_clip_avg=0.2078 | Metrics: + {'align_loss': 0.025550929829478264, + 'recon_loss': 0.06433141231536865, + 'predict_loss': 0.008441440761089325, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20014962553977966, + 'data_time': 0.0006458740099333227, + 'model_time': 1.2369988029822707, + 'grad_norm_pre_clip_avg': 0.20782384872436524, + 'learning_rate': 2.0186825554638607e-05, + 'epoch': 4.55} +04/19 [17:53:04] INFO | >> train_qwenlatent.py:487 + Step 18030 | grad_norm_pre_clip=0.1945 | + grad_norm_pre_clip_avg=0.1941 | Metrics: + {'align_loss': 0.024538837373256683, + 'recon_loss': 0.07898049801588058, + 'predict_loss': 0.011381939984858036, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1944754272699356, + 'data_time': 0.0008531310013495386, + 'model_time': 1.3003558210039046, + 'grad_norm_pre_clip_avg': 0.19410236179828644, + 'learning_rate': 2.017994638315826e-05, + 'epoch': 4.55} +04/19 [17:53:17] INFO | >> train_qwenlatent.py:487 + Step 18040 | grad_norm_pre_clip=0.1970 | + grad_norm_pre_clip_avg=0.2009 | Metrics: + {'align_loss': 0.02434813790023327, + 'recon_loss': 0.046964943408966064, + 'predict_loss': 0.007927707396447659, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19695895910263062, + 'data_time': 0.0010720379941631109, + 'model_time': 1.2563348039984703, + 'grad_norm_pre_clip_avg': 0.2008977085351944, + 'learning_rate': 2.017306347465769e-05, + 'epoch': 4.55} +04/19 [17:53:30] INFO | >> train_qwenlatent.py:487 + Step 18050 | grad_norm_pre_clip=0.2286 | + grad_norm_pre_clip_avg=0.2544 | Metrics: + {'align_loss': 0.024607215076684952, + 'recon_loss': 0.05588369071483612, + 'predict_loss': 0.010745045728981495, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22857551276683807, + 'mae_score': 0.014630651044415998, 'data_time': + 0.0007612759945914149, 'model_time': + 1.4619033169874456, 'grad_norm_pre_clip_avg': + 0.2543678149580956, 'learning_rate': + 2.0166176832491544e-05, 'epoch': 4.55} +04/19 [17:53:43] INFO | >> train_qwenlatent.py:487 + Step 18060 | grad_norm_pre_clip=0.2837 | + grad_norm_pre_clip_avg=0.2514 | Metrics: + {'align_loss': 0.024416819214820862, + 'recon_loss': 0.05344034731388092, + 'predict_loss': 0.01336806733161211, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28371262550354004, + 'data_time': 0.0007726970070507377, + 'model_time': 1.2270419790002052, + 'grad_norm_pre_clip_avg': 0.25140244960784913, + 'learning_rate': 2.015928646001629e-05, + 'epoch': 4.56} +04/19 [17:53:56] INFO | >> train_qwenlatent.py:487 + Step 18070 | grad_norm_pre_clip=0.1485 | + grad_norm_pre_clip_avg=0.2022 | Metrics: + {'align_loss': 0.025642920285463333, + 'recon_loss': 0.06195111200213432, + 'predict_loss': 0.00923899095505476, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14846466481685638, + 'data_time': 0.0010046419920399785, + 'model_time': 1.2784176490094978, + 'grad_norm_pre_clip_avg': 0.20215872228145598, + 'learning_rate': 2.0152392360590205e-05, + 'epoch': 4.56} +04/19 [17:54:08] INFO | >> train_qwenlatent.py:487 + Step 18080 | grad_norm_pre_clip=0.1793 | + grad_norm_pre_clip_avg=0.1877 | Metrics: + {'align_loss': 0.023835090920329094, + 'recon_loss': 0.05042416602373123, + 'predict_loss': 0.008710462599992752, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17932380735874176, + 'data_time': 0.0008527490135747939, + 'model_time': 1.2004987339896616, + 'grad_norm_pre_clip_avg': 0.18769301027059554, + 'learning_rate': 2.01454945375734e-05, 'epoch': + 4.56} +04/19 [17:54:21] INFO | >> train_qwenlatent.py:487 + Step 18090 | grad_norm_pre_clip=0.1753 | + grad_norm_pre_clip_avg=0.1907 | Metrics: + {'align_loss': 0.025294365361332893, + 'recon_loss': 0.07641445100307465, + 'predict_loss': 0.01801537349820137, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17527130246162415, + 'data_time': 0.001056840003002435, + 'model_time': 1.2427094939921517, + 'grad_norm_pre_clip_avg': 0.19073905050754547, + 'learning_rate': 2.013859299432778e-05, + 'epoch': 4.56} +04/19 [17:54:34] INFO | >> train_qwenlatent.py:487 + Step 18100 | grad_norm_pre_clip=0.2389 | + grad_norm_pre_clip_avg=0.2260 | Metrics: + {'align_loss': 0.02566741593182087, + 'recon_loss': 0.09232743084430695, + 'predict_loss': 0.01677621342241764, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23890195786952972, + 'mae_score': 0.010754094682298265, 'data_time': + 0.0007532499730587006, 'model_time': + 1.251031420979416, 'grad_norm_pre_clip_avg': + 0.22597278505563737, 'learning_rate': + 2.0131687734217083e-05, 'epoch': 4.57} +04/19 [17:54:47] INFO | >> train_qwenlatent.py:487 + Step 18110 | grad_norm_pre_clip=0.2507 | + grad_norm_pre_clip_avg=0.2283 | Metrics: + {'align_loss': 0.024423498660326004, + 'recon_loss': 0.06029968336224556, + 'predict_loss': 0.01462214533239603, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2507479190826416, + 'data_time': 0.000920455000596121, + 'model_time': 1.2109065059921704, + 'grad_norm_pre_clip_avg': 0.22826674431562424, + 'learning_rate': 2.0124778760606836e-05, + 'epoch': 4.57} +04/19 [17:55:00] INFO | >> train_qwenlatent.py:487 + Step 18120 | grad_norm_pre_clip=0.2536 | + grad_norm_pre_clip_avg=0.2295 | Metrics: + {'align_loss': 0.025758273899555206, + 'recon_loss': 0.0803956538438797, + 'predict_loss': 0.013680378906428814, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2536487579345703, + 'data_time': 0.0007298830023501068, + 'model_time': 1.2290777979942504, + 'grad_norm_pre_clip_avg': 0.22952092736959456, + 'learning_rate': 2.01178660768644e-05, 'epoch': + 4.57} +04/19 [17:55:12] INFO | >> train_qwenlatent.py:487 + Step 18130 | grad_norm_pre_clip=0.2375 | + grad_norm_pre_clip_avg=0.2020 | Metrics: + {'align_loss': 0.024569379165768623, + 'recon_loss': 0.04627196118235588, + 'predict_loss': 0.007959529757499695, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23750039935112, + 'data_time': 0.0008229920058511198, + 'model_time': 1.1946071030106395, + 'grad_norm_pre_clip_avg': 0.2020230174064636, + 'learning_rate': 2.0110949686358927e-05, + 'epoch': 4.57} +04/19 [17:55:25] INFO | >> train_qwenlatent.py:487 + Step 18140 | grad_norm_pre_clip=0.2451 | + grad_norm_pre_clip_avg=0.2145 | Metrics: + {'align_loss': 0.024590251967310905, + 'recon_loss': 0.05798429250717163, + 'predict_loss': 0.010826734825968742, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2450794130563736, + 'data_time': 0.0007238429971039295, + 'model_time': 1.2459615980042145, + 'grad_norm_pre_clip_avg': 0.21453591138124467, + 'learning_rate': 2.0104029592461388e-05, + 'epoch': 4.58} +04/19 [17:55:38] INFO | >> train_qwenlatent.py:487 + Step 18150 | grad_norm_pre_clip=0.2152 | + grad_norm_pre_clip_avg=0.2354 | Metrics: + {'align_loss': 0.024553215131163597, + 'recon_loss': 0.07230286300182343, + 'predict_loss': 0.016532011330127716, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21516750752925873, + 'mae_score': 0.016217678516834706, 'data_time': + 0.0007311319932341576, 'model_time': + 1.2173119199869689, 'grad_norm_pre_clip_avg': + 0.23535099625587463, 'learning_rate': + 2.0097105798544544e-05, 'epoch': 4.58} +04/19 [17:55:51] INFO | >> train_qwenlatent.py:487 + Step 18160 | grad_norm_pre_clip=0.1767 | + grad_norm_pre_clip_avg=0.2089 | Metrics: + {'align_loss': 0.02582200989127159, + 'recon_loss': 0.07542216777801514, + 'predict_loss': 0.013184471987187862, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17665940523147583, + 'data_time': 0.0010283470037393272, + 'model_time': 1.2445491539838258, + 'grad_norm_pre_clip_avg': 0.20886171162128447, + 'learning_rate': 2.0090178307982978e-05, + 'epoch': 4.58} +04/19 [17:56:03] INFO | >> train_qwenlatent.py:487 + Step 18170 | grad_norm_pre_clip=0.2060 | + grad_norm_pre_clip_avg=0.2202 | Metrics: + {'align_loss': 0.026152288541197777, + 'recon_loss': 0.06434451788663864, + 'predict_loss': 0.01855020597577095, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20596419274806976, + 'data_time': 0.0006630240241065621, + 'model_time': 1.2755699119879864, + 'grad_norm_pre_clip_avg': 0.2202489286661148, + 'learning_rate': 2.008324712415305e-05, + 'epoch': 4.58} +04/19 [17:56:16] INFO | >> train_qwenlatent.py:487 + Step 18180 | grad_norm_pre_clip=0.1852 | + grad_norm_pre_clip_avg=0.2036 | Metrics: + {'align_loss': 0.025576729327440262, + 'recon_loss': 0.0789608433842659, + 'predict_loss': 0.013337250798940659, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1851552426815033, + 'data_time': 0.0006661840016022325, + 'model_time': 1.200059869006509, + 'grad_norm_pre_clip_avg': 0.2035825476050377, + 'learning_rate': 2.007631225043295e-05, + 'epoch': 4.59} +04/19 [17:56:29] INFO | >> train_qwenlatent.py:487 + Step 18190 | grad_norm_pre_clip=0.1796 | + grad_norm_pre_clip_avg=0.1781 | Metrics: + {'align_loss': 0.024276118725538254, + 'recon_loss': 0.048699118196964264, + 'predict_loss': 0.008959717117249966, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1796450912952423, + 'data_time': 0.0006879400170873851, + 'model_time': 1.2474943289998919, + 'grad_norm_pre_clip_avg': 0.1781456410884857, + 'learning_rate': 2.0069373690202643e-05, + 'epoch': 4.59} +04/19 [17:56:43] INFO | >> train_qwenlatent.py:487 + Step 18200 | grad_norm_pre_clip=0.1523 | + grad_norm_pre_clip_avg=0.2076 | Metrics: + {'align_loss': 0.02568388357758522, + 'recon_loss': 0.07799439132213593, + 'predict_loss': 0.013782305642962456, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1523236483335495, + 'mae_score': 0.012047195434570312, 'data_time': + 0.0009753539925441146, 'model_time': + 1.581669678009348, 'grad_norm_pre_clip_avg': + 0.20761002600193024, 'learning_rate': + 2.006243144684391e-05, 'epoch': 4.59} +04/19 [17:56:55] INFO | >> train_qwenlatent.py:487 + Step 18210 | grad_norm_pre_clip=0.2108 | + grad_norm_pre_clip_avg=0.2071 | Metrics: + {'align_loss': 0.025036495178937912, + 'recon_loss': 0.06501532346010208, + 'predict_loss': 0.009708837606012821, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21077603101730347, + 'data_time': 0.000897737976629287, + 'model_time': 1.2477529580064584, + 'grad_norm_pre_clip_avg': 0.20706240236759185, + 'learning_rate': 2.0055485523740296e-05, + 'epoch': 4.6} +04/19 [17:57:08] INFO | >> train_qwenlatent.py:487 + Step 18220 | grad_norm_pre_clip=0.2912 | + grad_norm_pre_clip_avg=0.2961 | Metrics: + {'align_loss': 0.024035194888710976, + 'recon_loss': 0.04429727792739868, + 'predict_loss': 0.009504633024334908, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2911999225616455, + 'data_time': 0.0006840089918114245, + 'model_time': 1.297944963996997, + 'grad_norm_pre_clip_avg': 0.2961249962449074, + 'learning_rate': 2.0048535924277177e-05, + 'epoch': 4.6} +04/19 [17:57:20] INFO | >> train_qwenlatent.py:487 + Step 18230 | grad_norm_pre_clip=0.1420 | + grad_norm_pre_clip_avg=0.2134 | Metrics: + {'align_loss': 0.024223696440458298, + 'recon_loss': 0.061976294964551926, + 'predict_loss': 0.015368700958788395, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1419975906610489, + 'data_time': 0.0006874290120322257, + 'model_time': 1.5288136840099469, + 'grad_norm_pre_clip_avg': 0.21343543082475663, + 'learning_rate': 2.0041582651841698e-05, + 'epoch': 4.6} +04/19 [17:57:33] INFO | >> train_qwenlatent.py:487 + Step 18240 | grad_norm_pre_clip=0.2651 | + grad_norm_pre_clip_avg=0.2074 | Metrics: + {'align_loss': 0.02514263615012169, + 'recon_loss': 0.07674573361873627, + 'predict_loss': 0.010787286795675755, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2651481628417969, + 'data_time': 0.0011615569819696248, + 'model_time': 1.2758565210097004, + 'grad_norm_pre_clip_avg': 0.20742655247449876, + 'learning_rate': 2.0034625709822793e-05, + 'epoch': 4.6} +04/19 [17:57:46] INFO | >> train_qwenlatent.py:487 + Step 18250 | grad_norm_pre_clip=0.2011 | + grad_norm_pre_clip_avg=0.1876 | Metrics: + {'align_loss': 0.025407548993825912, + 'recon_loss': 0.06336021423339844, + 'predict_loss': 0.007759810891002417, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20109079778194427, + 'mae_score': 0.011277828560219154, 'data_time': + 0.0007139779918361455, 'model_time': + 1.2330267619981896, 'grad_norm_pre_clip_avg': + 0.18763533383607864, 'learning_rate': + 2.0027665101611206e-05, 'epoch': 4.61} +04/19 [17:57:59] INFO | >> train_qwenlatent.py:487 + Step 18260 | grad_norm_pre_clip=0.1786 | + grad_norm_pre_clip_avg=0.2002 | Metrics: + {'align_loss': 0.02475159242749214, + 'recon_loss': 0.06228434294462204, + 'predict_loss': 0.013663206249475479, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1786075234413147, + 'data_time': 0.0009629929845687002, + 'model_time': 1.2892195559979882, + 'grad_norm_pre_clip_avg': 0.20018173158168792, + 'learning_rate': 2.0020700830599437e-05, + 'epoch': 4.61} +04/19 [17:58:11] INFO | >> train_qwenlatent.py:487 + Step 18270 | grad_norm_pre_clip=0.2640 | + grad_norm_pre_clip_avg=0.2482 | Metrics: + {'align_loss': 0.024129891768097878, + 'recon_loss': 0.08245603740215302, + 'predict_loss': 0.01718967594206333, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2639727294445038, + 'data_time': 0.0007197060040198267, + 'model_time': 1.3417030469863676, + 'grad_norm_pre_clip_avg': 0.24823035597801207, + 'learning_rate': 2.0013732900181793e-05, + 'epoch': 4.61} +04/19 [17:58:24] INFO | >> train_qwenlatent.py:487 + Step 18280 | grad_norm_pre_clip=0.1505 | + grad_norm_pre_clip_avg=0.1906 | Metrics: + {'align_loss': 0.02529691345989704, + 'recon_loss': 0.06830921769142151, + 'predict_loss': 0.01251981407403946, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15053972601890564, + 'data_time': 0.000675413990393281, + 'model_time': 1.1954795680067036, + 'grad_norm_pre_clip_avg': 0.19059866517782212, + 'learning_rate': 2.0006761313754363e-05, + 'epoch': 4.61} +04/19 [17:58:36] INFO | >> train_qwenlatent.py:487 + Step 18290 | grad_norm_pre_clip=0.1727 | + grad_norm_pre_clip_avg=0.1710 | Metrics: + {'align_loss': 0.025940492749214172, + 'recon_loss': 0.0698946863412857, + 'predict_loss': 0.013268919661641121, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17271031439304352, + 'data_time': 0.000825622002594173, + 'model_time': 1.236760742001934, + 'grad_norm_pre_clip_avg': 0.17104699313640595, + 'learning_rate': 1.999978607471501e-05, + 'epoch': 4.62} +04/19 [17:58:49] INFO | >> train_qwenlatent.py:487 + Step 18300 | grad_norm_pre_clip=0.2018 | + grad_norm_pre_clip_avg=0.2235 | Metrics: + {'align_loss': 0.026455696672201157, + 'recon_loss': 0.0849006250500679, + 'predict_loss': 0.021596087142825127, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20175297558307648, + 'mae_score': 0.017138479421804617, 'data_time': + 0.0007556250202469528, 'model_time': + 1.2075826739892364, 'grad_norm_pre_clip_avg': + 0.22352893501520157, 'learning_rate': + 1.9992807186463382e-05, 'epoch': 4.62} +04/19 [17:59:02] INFO | >> train_qwenlatent.py:487 + Step 18310 | grad_norm_pre_clip=0.2285 | + grad_norm_pre_clip_avg=0.2296 | Metrics: + {'align_loss': 0.025771085172891617, + 'recon_loss': 0.07424833625555038, + 'predict_loss': 0.01727503165602684, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22853150963783264, + 'data_time': 0.0008897320076357573, + 'model_time': 1.2870813049958088, + 'grad_norm_pre_clip_avg': 0.22963740080595016, + 'learning_rate': 1.99858246524009e-05, 'epoch': + 4.62} +04/19 [17:59:15] INFO | >> train_qwenlatent.py:487 + Step 18320 | grad_norm_pre_clip=0.1799 | + grad_norm_pre_clip_avg=0.2196 | Metrics: + {'align_loss': 0.02533814124763012, + 'recon_loss': 0.06575116515159607, + 'predict_loss': 0.011098505929112434, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17992639541625977, + 'data_time': 0.000675754010444507, + 'model_time': 1.4869126209814567, + 'grad_norm_pre_clip_avg': 0.21955033391714096, + 'learning_rate': 1.997883847593077e-05, + 'epoch': 4.62} +04/19 [17:59:28] INFO | >> train_qwenlatent.py:487 + Step 18330 | grad_norm_pre_clip=0.1861 | + grad_norm_pre_clip_avg=0.2115 | Metrics: + {'align_loss': 0.024948909878730774, + 'recon_loss': 0.08322293311357498, + 'predict_loss': 0.016470659524202347, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18611016869544983, + 'data_time': 0.0006707219872623682, + 'model_time': 1.2182561049994547, + 'grad_norm_pre_clip_avg': 0.2114839106798172, + 'learning_rate': 1.997184866045797e-05, + 'epoch': 4.63} +04/19 [17:59:41] INFO | >> train_qwenlatent.py:487 + Step 18340 | grad_norm_pre_clip=0.1500 | + grad_norm_pre_clip_avg=0.2071 | Metrics: + {'align_loss': 0.024474715813994408, + 'recon_loss': 0.07217103242874146, + 'predict_loss': 0.012763439677655697, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.150043785572052, + 'data_time': 0.000994697998976335, + 'model_time': 1.2648757249990012, + 'grad_norm_pre_clip_avg': 0.20714490711688996, + 'learning_rate': 1.9964855209389244e-05, + 'epoch': 4.63} +04/19 [17:59:54] INFO | >> train_qwenlatent.py:487 + Step 18350 | grad_norm_pre_clip=0.1550 | + grad_norm_pre_clip_avg=0.2090 | Metrics: + {'align_loss': 0.024015117436647415, + 'recon_loss': 0.050094082951545715, + 'predict_loss': 0.008587002754211426, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15501150488853455, + 'mae_score': 0.014611029410147452, 'data_time': + 0.0008782969962339848, 'model_time': + 1.2202056500245817, 'grad_norm_pre_clip_avg': + 0.20904069393873215, 'learning_rate': + 1.995785812613313e-05, 'epoch': 4.63} +04/19 [18:00:07] INFO | >> train_qwenlatent.py:487 + Step 18360 | grad_norm_pre_clip=0.2306 | + grad_norm_pre_clip_avg=0.1987 | Metrics: + {'align_loss': 0.024991566315293312, + 'recon_loss': 0.08292783796787262, + 'predict_loss': 0.01269518118351698, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23059917986392975, + 'data_time': 0.0009907940111588687, + 'model_time': 1.2610396259988192, + 'grad_norm_pre_clip_avg': 0.19870066940784453, + 'learning_rate': 1.995085741409991e-05, + 'epoch': 4.63} +04/19 [18:00:20] INFO | >> train_qwenlatent.py:487 + Step 18370 | grad_norm_pre_clip=0.1703 | + grad_norm_pre_clip_avg=0.2045 | Metrics: + {'align_loss': 0.02547549456357956, + 'recon_loss': 0.07350242882966995, + 'predict_loss': 0.008526216261088848, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1702936291694641, + 'data_time': 0.0008745109953451902, + 'model_time': 1.2947396629897412, + 'grad_norm_pre_clip_avg': 0.20453813225030898, + 'learning_rate': 1.9943853076701648e-05, + 'epoch': 4.64} +04/19 [18:00:32] INFO | >> train_qwenlatent.py:487 + Step 18380 | grad_norm_pre_clip=0.3042 | + grad_norm_pre_clip_avg=0.1971 | Metrics: + {'align_loss': 0.024898067116737366, + 'recon_loss': 0.055443767458200455, + 'predict_loss': 0.010219333693385124, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.30421462655067444, + 'data_time': 0.0008088379981927574, + 'model_time': 1.265816391009139, + 'grad_norm_pre_clip_avg': 0.1971462696790695, + 'learning_rate': 1.993684511735217e-05, + 'epoch': 4.64} +04/19 [18:00:45] INFO | >> train_qwenlatent.py:487 + Step 18390 | grad_norm_pre_clip=0.1837 | + grad_norm_pre_clip_avg=0.2129 | Metrics: + {'align_loss': 0.02650056779384613, + 'recon_loss': 0.07025764137506485, + 'predict_loss': 0.014905311167240143, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18366755545139313, + 'data_time': 0.0010025009978562593, + 'model_time': 1.2127598600054625, + 'grad_norm_pre_clip_avg': 0.21286127865314483, + 'learning_rate': 1.9929833539467075e-05, + 'epoch': 4.64} +04/19 [18:00:58] INFO | >> train_qwenlatent.py:487 + Step 18400 | grad_norm_pre_clip=0.1963 | + grad_norm_pre_clip_avg=0.1717 | Metrics: + {'align_loss': 0.02430909499526024, + 'recon_loss': 0.06477149575948715, + 'predict_loss': 0.012272926047444344, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19633935391902924, + 'mae_score': 0.01429102442286036, 'data_time': + 0.001118682004744187, 'model_time': + 1.2574863189947791, 'grad_norm_pre_clip_avg': + 0.1717320591211319, 'learning_rate': + 1.9922818346463724e-05, 'epoch': 4.64} +04/19 [18:01:11] INFO | >> train_qwenlatent.py:487 + Step 18410 | grad_norm_pre_clip=0.2764 | + grad_norm_pre_clip_avg=0.2605 | Metrics: + {'align_loss': 0.026158396154642105, + 'recon_loss': 0.05779355764389038, + 'predict_loss': 0.008406218141317368, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2763749659061432, + 'data_time': 0.0008645440102554858, + 'model_time': 1.2637495159870014, + 'grad_norm_pre_clip_avg': 0.2604841634631157, + 'learning_rate': 1.991579954176123e-05, + 'epoch': 4.65} +04/19 [18:01:23] INFO | >> train_qwenlatent.py:487 + Step 18420 | grad_norm_pre_clip=0.2986 | + grad_norm_pre_clip_avg=0.2881 | Metrics: + {'align_loss': 0.02510102093219757, + 'recon_loss': 0.08197394013404846, + 'predict_loss': 0.017200907692313194, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2986498773097992, + 'data_time': 0.0007663869764655828, + 'model_time': 1.2356641349906567, + 'grad_norm_pre_clip_avg': 0.2881407797336578, + 'learning_rate': 1.9908777128780472e-05, + 'epoch': 4.65} +04/19 [18:01:35] INFO | >> train_qwenlatent.py:487 + Step 18430 | grad_norm_pre_clip=0.2833 | + grad_norm_pre_clip_avg=0.2307 | Metrics: + {'align_loss': 0.02555682882666588, + 'recon_loss': 0.07728900015354156, + 'predict_loss': 0.00960625521838665, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28334730863571167, + 'data_time': 0.0006847819895483553, + 'model_time': 1.2273955999989994, + 'grad_norm_pre_clip_avg': 0.23069387972354888, + 'learning_rate': 1.9901751110944097e-05, + 'epoch': 4.65} +04/19 [18:01:48] INFO | >> train_qwenlatent.py:487 + Step 18440 | grad_norm_pre_clip=0.2601 | + grad_norm_pre_clip_avg=0.2277 | Metrics: + {'align_loss': 0.025409799069166183, + 'recon_loss': 0.06931833177804947, + 'predict_loss': 0.01100886519998312, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2600991725921631, + 'data_time': 0.0007027790124993771, + 'model_time': 1.199395314004505, + 'grad_norm_pre_clip_avg': 0.22772640734910965, + 'learning_rate': 1.9894721491676495e-05, + 'epoch': 4.65} +04/19 [18:02:01] INFO | >> train_qwenlatent.py:487 + Step 18450 | grad_norm_pre_clip=0.1661 | + grad_norm_pre_clip_avg=0.2124 | Metrics: + {'align_loss': 0.02484036795794964, + 'recon_loss': 0.06762233376502991, + 'predict_loss': 0.010126778855919838, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1661166101694107, + 'mae_score': 0.012583686210013725, 'data_time': + 0.000719992007361725, 'model_time': + 1.244908287015278, 'grad_norm_pre_clip_avg': + 0.21240612715482712, 'learning_rate': + 1.9887688274403818e-05, 'epoch': 4.66} +04/19 [18:02:14] INFO | >> train_qwenlatent.py:487 + Step 18460 | grad_norm_pre_clip=0.2029 | + grad_norm_pre_clip_avg=0.2112 | Metrics: + {'align_loss': 0.02481650933623314, + 'recon_loss': 0.0641508549451828, + 'predict_loss': 0.008890539407730103, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2028883993625641, + 'data_time': 0.0006971589755266905, + 'model_time': 1.3041362570074853, + 'grad_norm_pre_clip_avg': 0.21122910231351852, + 'learning_rate': 1.9880651462553973e-05, + 'epoch': 4.66} +04/19 [18:02:27] INFO | >> train_qwenlatent.py:487 + Step 18470 | grad_norm_pre_clip=0.2127 | + grad_norm_pre_clip_avg=0.2000 | Metrics: + {'align_loss': 0.025597667321562767, + 'recon_loss': 0.07998470962047577, + 'predict_loss': 0.012756234966218472, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21274684369564056, + 'data_time': 0.0006689550064038485, + 'model_time': 1.2151516540034208, + 'grad_norm_pre_clip_avg': 0.19999635666608812, + 'learning_rate': 1.9873611059556613e-05, + 'epoch': 4.66} +04/19 [18:02:39] INFO | >> train_qwenlatent.py:487 + Step 18480 | grad_norm_pre_clip=0.1583 | + grad_norm_pre_clip_avg=0.1778 | Metrics: + {'align_loss': 0.02571558952331543, + 'recon_loss': 0.06620189547538757, + 'predict_loss': 0.01417585276067257, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1582995057106018, + 'data_time': 0.000699406024068594, + 'model_time': 1.2514302229974419, + 'grad_norm_pre_clip_avg': 0.17784905135631562, + 'learning_rate': 1.9866567068843145e-05, + 'epoch': 4.66} +04/19 [18:02:52] INFO | >> train_qwenlatent.py:487 + Step 18490 | grad_norm_pre_clip=0.2212 | + grad_norm_pre_clip_avg=0.1841 | Metrics: + {'align_loss': 0.025087784975767136, + 'recon_loss': 0.060032688081264496, + 'predict_loss': 0.01061967108398676, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2211637794971466, + 'data_time': 0.0010974390024784952, + 'model_time': 1.2708403400029056, + 'grad_norm_pre_clip_avg': 0.18405291438102722, + 'learning_rate': 1.9859519493846728e-05, + 'epoch': 4.67} +04/19 [18:03:06] INFO | >> train_qwenlatent.py:487 + Step 18500 | grad_norm_pre_clip=0.2500 | + grad_norm_pre_clip_avg=0.2298 | Metrics: + {'align_loss': 0.02357378974556923, + 'recon_loss': 0.06394848227500916, + 'predict_loss': 0.012412730604410172, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2500429153442383, + 'mae_score': 0.017550980507790505, 'data_time': + 0.001006398000754416, 'model_time': + 1.2565302530128974, 'grad_norm_pre_clip_avg': + 0.2298487737774849, 'learning_rate': + 1.985246833800226e-05, 'epoch': 4.67} +04/19 [18:03:19] INFO | >> train_qwenlatent.py:487 + Step 18510 | grad_norm_pre_clip=0.1766 | + grad_norm_pre_clip_avg=0.2057 | Metrics: + {'align_loss': 0.02477094903588295, + 'recon_loss': 0.10463137924671173, + 'predict_loss': 0.015343692153692245, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17663829028606415, + 'data_time': 0.0007149399898480624, + 'model_time': 1.5040657450153958, + 'grad_norm_pre_clip_avg': 0.20567720681428908, + 'learning_rate': 1.984541360474639e-05, + 'epoch': 4.67} +04/19 [18:03:31] INFO | >> train_qwenlatent.py:487 + Step 18520 | grad_norm_pre_clip=0.1938 | + grad_norm_pre_clip_avg=0.1906 | Metrics: + {'align_loss': 0.025403844192624092, + 'recon_loss': 0.06433635205030441, + 'predict_loss': 0.01117471419274807, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19380931556224823, + 'data_time': 0.0006766739825252444, + 'model_time': 1.181817739008693, + 'grad_norm_pre_clip_avg': 0.1906222626566887, + 'learning_rate': 1.9838355297517514e-05, + 'epoch': 4.67} +04/19 [18:03:44] INFO | >> train_qwenlatent.py:487 + Step 18530 | grad_norm_pre_clip=0.2170 | + grad_norm_pre_clip_avg=0.2379 | Metrics: + {'align_loss': 0.025104505941271782, + 'recon_loss': 0.051848117262125015, + 'predict_loss': 0.007871948182582855, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21701160073280334, + 'data_time': 0.001109345001168549, + 'model_time': 1.3055521279748064, + 'grad_norm_pre_clip_avg': 0.23790145069360732, + 'learning_rate': 1.9831293419755758e-05, + 'epoch': 4.68} +04/19 [18:03:56] INFO | >> train_qwenlatent.py:487 + Step 18540 | grad_norm_pre_clip=0.2673 | + grad_norm_pre_clip_avg=0.2200 | Metrics: + {'align_loss': 0.025144681334495544, + 'recon_loss': 0.04874774441123009, + 'predict_loss': 0.008445796556770802, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26726511120796204, + 'data_time': 0.0006768580060452223, + 'model_time': 1.1961101070046425, + 'grad_norm_pre_clip_avg': 0.21996528208255767, + 'learning_rate': 1.9824227974903002e-05, + 'epoch': 4.68} +04/19 [18:04:09] INFO | >> train_qwenlatent.py:487 + Step 18550 | grad_norm_pre_clip=0.2215 | + grad_norm_pre_clip_avg=0.2100 | Metrics: + {'align_loss': 0.024798745289444923, + 'recon_loss': 0.06899742782115936, + 'predict_loss': 0.012522228062152863, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2214527577161789, + 'mae_score': 0.011782807702416772, 'data_time': + 0.0008386929985135794, 'model_time': + 1.2227368789899629, 'grad_norm_pre_clip_avg': + 0.20997963547706605, 'learning_rate': + 1.9817158966402857e-05, 'epoch': 4.68} +04/19 [18:04:22] INFO | >> train_qwenlatent.py:487 + Step 18560 | grad_norm_pre_clip=0.2325 | + grad_norm_pre_clip_avg=0.2128 | Metrics: + {'align_loss': 0.024043913930654526, + 'recon_loss': 0.06506133079528809, + 'predict_loss': 0.017779355868697166, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23252040147781372, + 'data_time': 0.0009576980082783848, + 'model_time': 1.2644219920039177, + 'grad_norm_pre_clip_avg': 0.21283414214849472, + 'learning_rate': 1.981008639770066e-05, + 'epoch': 4.68} +04/19 [18:04:34] INFO | >> train_qwenlatent.py:487 + Step 18570 | grad_norm_pre_clip=0.1863 | + grad_norm_pre_clip_avg=0.2073 | Metrics: + {'align_loss': 0.02413884550333023, + 'recon_loss': 0.05253426730632782, + 'predict_loss': 0.010429401881992817, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18630464375019073, + 'data_time': 0.0007375200220849365, + 'model_time': 1.252822306996677, + 'grad_norm_pre_clip_avg': 0.2073207125067711, + 'learning_rate': 1.9803010272243515e-05, + 'epoch': 4.69} +04/19 [18:04:47] INFO | >> train_qwenlatent.py:487 + Step 18580 | grad_norm_pre_clip=0.2276 | + grad_norm_pre_clip_avg=0.2004 | Metrics: + {'align_loss': 0.02598213404417038, + 'recon_loss': 0.06681028008460999, + 'predict_loss': 0.012864881195127964, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22763347625732422, + 'data_time': 0.0009761230030562729, + 'model_time': 1.2200227990106214, + 'grad_norm_pre_clip_avg': 0.2003897324204445, + 'learning_rate': 1.9795930593480224e-05, + 'epoch': 4.69} +04/19 [18:05:00] INFO | >> train_qwenlatent.py:487 + Step 18590 | grad_norm_pre_clip=0.2042 | + grad_norm_pre_clip_avg=0.2048 | Metrics: + {'align_loss': 0.02584036998450756, + 'recon_loss': 0.08209878951311111, + 'predict_loss': 0.012035110965371132, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20424269139766693, + 'data_time': 0.0011670149979181588, + 'model_time': 1.2813598369830288, + 'grad_norm_pre_clip_avg': 0.20478237867355348, + 'learning_rate': 1.9788847364861347e-05, + 'epoch': 4.69} +04/19 [18:05:13] INFO | >> train_qwenlatent.py:487 + Step 18600 | grad_norm_pre_clip=0.1896 | + grad_norm_pre_clip_avg=0.2281 | Metrics: + {'align_loss': 0.024132996797561646, + 'recon_loss': 0.05276547744870186, + 'predict_loss': 0.010477886535227299, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1896279752254486, + 'mae_score': 0.012394557987247501, 'data_time': + 0.000660282006720081, 'model_time': + 1.1680349060043227, 'grad_norm_pre_clip_avg': + 0.22814378440380095, 'learning_rate': + 1.9781760589839153e-05, 'epoch': 4.69} +04/19 [18:05:26] INFO | >> train_qwenlatent.py:487 + Step 18610 | grad_norm_pre_clip=0.2184 | + grad_norm_pre_clip_avg=0.2205 | Metrics: + {'align_loss': 0.02535225637257099, + 'recon_loss': 0.07021532952785492, + 'predict_loss': 0.010991289280354977, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21838392317295074, + 'data_time': 0.0011344289814587682, + 'model_time': 1.3206464750110172, + 'grad_norm_pre_clip_avg': 0.22051339596509933, + 'learning_rate': 1.9774670271867656e-05, + 'epoch': 4.7} +04/19 [18:05:39] INFO | >> train_qwenlatent.py:487 + Step 18620 | grad_norm_pre_clip=0.1811 | + grad_norm_pre_clip_avg=0.2168 | Metrics: + {'align_loss': 0.024229738861322403, + 'recon_loss': 0.06016451492905617, + 'predict_loss': 0.009386674501001835, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18111860752105713, + 'data_time': 0.0008165780163835734, + 'model_time': 1.2349723399966024, + 'grad_norm_pre_clip_avg': 0.21679362058639526, + 'learning_rate': 1.9767576414402594e-05, + 'epoch': 4.7} +04/19 [18:05:51] INFO | >> train_qwenlatent.py:487 + Step 18630 | grad_norm_pre_clip=0.1686 | + grad_norm_pre_clip_avg=0.1967 | Metrics: + {'align_loss': 0.02559695392847061, + 'recon_loss': 0.0603451170027256, + 'predict_loss': 0.008257930167019367, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16857364773750305, + 'data_time': 0.001390922989230603, + 'model_time': 1.247689520008862, + 'grad_norm_pre_clip_avg': 0.19669276624917983, + 'learning_rate': 1.976047902090142e-05, + 'epoch': 4.7} +04/19 [18:06:03] INFO | >> train_qwenlatent.py:487 + Step 18640 | grad_norm_pre_clip=0.3415 | + grad_norm_pre_clip_avg=0.2376 | Metrics: + {'align_loss': 0.024934932589530945, + 'recon_loss': 0.07315593212842941, + 'predict_loss': 0.013274279423058033, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3414827585220337, + 'data_time': 0.0007072500011418015, + 'model_time': 1.2101884539879393, + 'grad_norm_pre_clip_avg': 0.23762000054121019, + 'learning_rate': 1.9753378094823322e-05, + 'epoch': 4.7} +04/19 [18:06:17] INFO | >> train_qwenlatent.py:487 + Step 18650 | grad_norm_pre_clip=0.1571 | + grad_norm_pre_clip_avg=0.2205 | Metrics: + {'align_loss': 0.025591842830181122, + 'recon_loss': 0.08013982325792313, + 'predict_loss': 0.011962566524744034, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15708111226558685, + 'mae_score': 0.01770771817044095, 'data_time': + 0.0009321100078523159, 'model_time': + 1.4603136079967953, 'grad_norm_pre_clip_avg': + 0.2204984813928604, 'learning_rate': + 1.9746273639629202e-05, 'epoch': 4.71} +04/19 [18:06:29] INFO | >> train_qwenlatent.py:487 + Step 18660 | grad_norm_pre_clip=0.1352 | + grad_norm_pre_clip_avg=0.1747 | Metrics: + {'align_loss': 0.024621695280075073, + 'recon_loss': 0.07002892345190048, + 'predict_loss': 0.012834853492677212, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13517791032791138, + 'data_time': 0.0006772569904569536, + 'model_time': 1.210740446986165, + 'grad_norm_pre_clip_avg': 0.17467278838157654, + 'learning_rate': 1.9739165658781687e-05, + 'epoch': 4.71} +04/19 [18:06:42] INFO | >> train_qwenlatent.py:487 + Step 18670 | grad_norm_pre_clip=0.2108 | + grad_norm_pre_clip_avg=0.1803 | Metrics: + {'align_loss': 0.025130171328783035, + 'recon_loss': 0.06777915358543396, + 'predict_loss': 0.013430457562208176, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2108098864555359, + 'data_time': 0.0012438179983291775, + 'model_time': 1.2514641629823018, + 'grad_norm_pre_clip_avg': 0.1803111642599106, + 'learning_rate': 1.9732054155745125e-05, + 'epoch': 4.71} +04/19 [18:06:54] INFO | >> train_qwenlatent.py:487 + Step 18680 | grad_norm_pre_clip=0.1731 | + grad_norm_pre_clip_avg=0.1778 | Metrics: + {'align_loss': 0.024807460606098175, + 'recon_loss': 0.08285709470510483, + 'predict_loss': 0.015556703321635723, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17309434711933136, + 'data_time': 0.0009883590100798756, + 'model_time': 1.3109345939883497, + 'grad_norm_pre_clip_avg': 0.1777808517217636, + 'learning_rate': 1.9724939133985567e-05, + 'epoch': 4.71} +04/19 [18:07:07] INFO | >> train_qwenlatent.py:487 + Step 18690 | grad_norm_pre_clip=0.2482 | + grad_norm_pre_clip_avg=0.2582 | Metrics: + {'align_loss': 0.024262603372335434, + 'recon_loss': 0.047407664358615875, + 'predict_loss': 0.011101873591542244, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24816200137138367, + 'data_time': 0.0007186710136011243, + 'model_time': 1.2242080320138484, + 'grad_norm_pre_clip_avg': 0.2581592559814453, + 'learning_rate': 1.9717820596970797e-05, + 'epoch': 4.72} +04/19 [18:07:20] INFO | >> train_qwenlatent.py:487 + Step 18700 | grad_norm_pre_clip=0.2081 | + grad_norm_pre_clip_avg=0.2234 | Metrics: + {'align_loss': 0.02560528554022312, + 'recon_loss': 0.07468216866254807, + 'predict_loss': 0.009309887886047363, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20809145271778107, + 'mae_score': 0.019842331688683313, 'data_time': + 0.0009190150012727827, 'model_time': + 1.2313293910119683, 'grad_norm_pre_clip_avg': + 0.2233831211924553, 'learning_rate': + 1.9710698548170296e-05, 'epoch': 4.72} +04/19 [18:07:33] INFO | >> train_qwenlatent.py:487 + Step 18710 | grad_norm_pre_clip=0.2044 | + grad_norm_pre_clip_avg=0.2111 | Metrics: + {'align_loss': 0.025676973164081573, + 'recon_loss': 0.09139983355998993, + 'predict_loss': 0.015169442631304264, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20444229245185852, + 'data_time': 0.0008146740146912634, + 'model_time': 1.2825347179896198, + 'grad_norm_pre_clip_avg': 0.21109968423843384, + 'learning_rate': 1.970357299105527e-05, + 'epoch': 4.72} +04/19 [18:07:46] INFO | >> train_qwenlatent.py:487 + Step 18720 | grad_norm_pre_clip=0.2041 | + grad_norm_pre_clip_avg=0.1903 | Metrics: + {'align_loss': 0.02558516152203083, + 'recon_loss': 0.08491777628660202, + 'predict_loss': 0.021506553515791893, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2040889710187912, + 'data_time': 0.0007063109951559454, + 'model_time': 1.2357439330080524, + 'grad_norm_pre_clip_avg': 0.19030166417360306, + 'learning_rate': 1.9696443929098623e-05, + 'epoch': 4.72} +04/19 [18:07:58] INFO | >> train_qwenlatent.py:487 + Step 18730 | grad_norm_pre_clip=0.2256 | + grad_norm_pre_clip_avg=0.2187 | Metrics: + {'align_loss': 0.024848248809576035, + 'recon_loss': 0.047087352722883224, + 'predict_loss': 0.008848828263580799, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22564631700515747, + 'data_time': 0.0011086220038123429, + 'model_time': 1.2903493730118498, + 'grad_norm_pre_clip_avg': 0.21866609007120133, + 'learning_rate': 1.9689311365774974e-05, + 'epoch': 4.73} +04/19 [18:08:11] INFO | >> train_qwenlatent.py:487 + Step 18740 | grad_norm_pre_clip=0.1750 | + grad_norm_pre_clip_avg=0.2335 | Metrics: + {'align_loss': 0.023805204778909683, + 'recon_loss': 0.067899189889431, + 'predict_loss': 0.013885188847780228, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1750161051750183, + 'data_time': 0.0008553469961043447, + 'model_time': 1.4642897009907756, + 'grad_norm_pre_clip_avg': 0.23346563726663588, + 'learning_rate': 1.9682175304560658e-05, + 'epoch': 4.73} +04/19 [18:08:24] INFO | >> train_qwenlatent.py:487 + Step 18750 | grad_norm_pre_clip=0.1619 | + grad_norm_pre_clip_avg=0.2157 | Metrics: + {'align_loss': 0.024754054844379425, + 'recon_loss': 0.05898049846291542, + 'predict_loss': 0.013558917678892612, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16188718378543854, + 'mae_score': 0.012538154705150708, 'data_time': + 0.0007410180114675313, 'model_time': + 1.2319789989851415, 'grad_norm_pre_clip_avg': + 0.2156698375940323, 'learning_rate': + 1.9675035748933687e-05, 'epoch': 4.73} +04/19 [18:08:37] INFO | >> train_qwenlatent.py:487 + Step 18760 | grad_norm_pre_clip=0.2201 | + grad_norm_pre_clip_avg=0.2084 | Metrics: + {'align_loss': 0.025806736201047897, + 'recon_loss': 0.08377818763256073, + 'predict_loss': 0.011847083456814289, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22007542848587036, + 'data_time': 0.0009802480053622276, + 'model_time': 1.2497483899933286, + 'grad_norm_pre_clip_avg': 0.2083582401275635, + 'learning_rate': 1.966789270237381e-05, + 'epoch': 4.73} +04/19 [18:08:50] INFO | >> train_qwenlatent.py:487 + Step 18770 | grad_norm_pre_clip=0.2579 | + grad_norm_pre_clip_avg=0.2205 | Metrics: + {'align_loss': 0.024752840399742126, + 'recon_loss': 0.07245303690433502, + 'predict_loss': 0.013139664195477962, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2579263746738434, + 'data_time': 0.0008471519977319986, + 'model_time': 1.2520545840088744, + 'grad_norm_pre_clip_avg': 0.22046886682510375, + 'learning_rate': 1.9660746168362458e-05, + 'epoch': 4.74} +04/19 [18:09:03] INFO | >> train_qwenlatent.py:487 + Step 18780 | grad_norm_pre_clip=0.1880 | + grad_norm_pre_clip_avg=0.1956 | Metrics: + {'align_loss': 0.025117821991443634, + 'recon_loss': 0.06549648195505142, + 'predict_loss': 0.011848006397485733, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18796297907829285, + 'data_time': 0.0009499129955656826, + 'model_time': 1.253937205998227, + 'grad_norm_pre_clip_avg': 0.19561113864183427, + 'learning_rate': 1.9653596150382755e-05, + 'epoch': 4.74} +04/19 [18:09:15] INFO | >> train_qwenlatent.py:487 + Step 18790 | grad_norm_pre_clip=0.1667 | + grad_norm_pre_clip_avg=0.2155 | Metrics: + {'align_loss': 0.024535048753023148, + 'recon_loss': 0.04963717982172966, + 'predict_loss': 0.013923716731369495, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16666176915168762, + 'data_time': 0.0007309500069823116, + 'model_time': 1.2746457130124327, + 'grad_norm_pre_clip_avg': 0.21545715630054474, + 'learning_rate': 1.964644265191954e-05, + 'epoch': 4.74} +04/19 [18:09:28] INFO | >> train_qwenlatent.py:487 + Step 18800 | grad_norm_pre_clip=0.2707 | + grad_norm_pre_clip_avg=0.2077 | Metrics: + {'align_loss': 0.02481730282306671, + 'recon_loss': 0.05508460849523544, + 'predict_loss': 0.009819388389587402, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.27066361904144287, + 'mae_score': 0.012196382745966182, 'data_time': + 0.0009651580185163766, 'model_time': + 1.2363708080083597, 'grad_norm_pre_clip_avg': + 0.20770589411258697, 'learning_rate': + 1.9639285676459345e-05, 'epoch': 4.74} +04/19 [18:09:41] INFO | >> train_qwenlatent.py:487 + Step 18810 | grad_norm_pre_clip=0.2029 | + grad_norm_pre_clip_avg=0.2172 | Metrics: + {'align_loss': 0.02558744139969349, + 'recon_loss': 0.06344615668058395, + 'predict_loss': 0.012120628729462624, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20286166667938232, + 'data_time': 0.0008638979925308377, + 'model_time': 1.229566911002621, + 'grad_norm_pre_clip_avg': 0.21719742715358734, + 'learning_rate': 1.9632125227490383e-05, + 'epoch': 4.75} +04/19 [18:09:53] INFO | >> train_qwenlatent.py:487 + Step 18820 | grad_norm_pre_clip=0.1766 | + grad_norm_pre_clip_avg=0.2039 | Metrics: + {'align_loss': 0.02407432533800602, + 'recon_loss': 0.05972111597657204, + 'predict_loss': 0.008357524871826172, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17661893367767334, + 'data_time': 0.0007249900081660599, + 'model_time': 1.2570757920038886, + 'grad_norm_pre_clip_avg': 0.20394325107336045, + 'learning_rate': 1.962496130850258e-05, + 'epoch': 4.75} +04/19 [18:10:06] INFO | >> train_qwenlatent.py:487 + Step 18830 | grad_norm_pre_clip=0.1961 | + grad_norm_pre_clip_avg=0.2008 | Metrics: + {'align_loss': 0.024437349289655685, + 'recon_loss': 0.06209341064095497, + 'predict_loss': 0.01405997946858406, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19613027572631836, + 'data_time': 0.0007381679897662252, + 'model_time': 1.270208965986967, + 'grad_norm_pre_clip_avg': 0.2007956847548485, + 'learning_rate': 1.9617793922987535e-05, + 'epoch': 4.75} +04/19 [18:10:18] INFO | >> train_qwenlatent.py:487 + Step 18840 | grad_norm_pre_clip=0.2478 | + grad_norm_pre_clip_avg=0.2302 | Metrics: + {'align_loss': 0.02581179514527321, + 'recon_loss': 0.09324280172586441, + 'predict_loss': 0.02472538873553276, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24780985713005066, + 'data_time': 0.0007120709924492985, + 'model_time': 1.211272910994012, + 'grad_norm_pre_clip_avg': 0.2302371084690094, + 'learning_rate': 1.9610623074438547e-05, + 'epoch': 4.75} +04/19 [18:10:32] INFO | >> train_qwenlatent.py:487 + Step 18850 | grad_norm_pre_clip=0.2463 | + grad_norm_pre_clip_avg=0.2269 | Metrics: + {'align_loss': 0.02570570632815361, + 'recon_loss': 0.0835593193769455, + 'predict_loss': 0.01985892280936241, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24627797305583954, + 'mae_score': 0.01348537152951902, 'data_time': + 0.0009771869808901101, 'model_time': + 1.2541105769923888, 'grad_norm_pre_clip_avg': + 0.22685524821281433, 'learning_rate': + 1.9603448766350603e-05, 'epoch': 4.76} +04/19 [18:10:45] INFO | >> train_qwenlatent.py:487 + Step 18860 | grad_norm_pre_clip=0.1881 | + grad_norm_pre_clip_avg=0.2237 | Metrics: + {'align_loss': 0.025053532794117928, + 'recon_loss': 0.05827854201197624, + 'predict_loss': 0.008253253996372223, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18811991810798645, + 'data_time': 0.0006967740191612393, + 'model_time': 1.2572859470092226, + 'grad_norm_pre_clip_avg': 0.2237184852361679, + 'learning_rate': 1.9596271002220373e-05, + 'epoch': 4.76} +04/19 [18:10:57] INFO | >> train_qwenlatent.py:487 + Step 18870 | grad_norm_pre_clip=0.1949 | + grad_norm_pre_clip_avg=0.2142 | Metrics: + {'align_loss': 0.024690814316272736, + 'recon_loss': 0.05408741533756256, + 'predict_loss': 0.011592824012041092, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19491353631019592, + 'data_time': 0.0009902349847834557, + 'model_time': 1.2343033979996108, + 'grad_norm_pre_clip_avg': 0.21421622037887572, + 'learning_rate': 1.9589089785546207e-05, + 'epoch': 4.76} +04/19 [18:11:10] INFO | >> train_qwenlatent.py:487 + Step 18880 | grad_norm_pre_clip=0.1727 | + grad_norm_pre_clip_avg=0.2060 | Metrics: + {'align_loss': 0.02320733666419983, + 'recon_loss': 0.06929711997509003, + 'predict_loss': 0.01351216435432434, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17274314165115356, + 'data_time': 0.0006690089940093458, + 'model_time': 1.2268824760103598, + 'grad_norm_pre_clip_avg': 0.20600577294826508, + 'learning_rate': 1.9581905119828152e-05, + 'epoch': 4.76} +04/19 [18:11:23] INFO | >> train_qwenlatent.py:487 + Step 18890 | grad_norm_pre_clip=0.2301 | + grad_norm_pre_clip_avg=0.1873 | Metrics: + {'align_loss': 0.0255742110311985, + 'recon_loss': 0.07872317731380463, + 'predict_loss': 0.01381294522434473, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23009411990642548, + 'data_time': 0.0007539830112364143, + 'model_time': 1.1986113610037137, + 'grad_norm_pre_clip_avg': 0.1873186320066452, + 'learning_rate': 1.9574717008567925e-05, + 'epoch': 4.77} +04/19 [18:11:36] INFO | >> train_qwenlatent.py:487 + Step 18900 | grad_norm_pre_clip=0.2093 | + grad_norm_pre_clip_avg=0.2159 | Metrics: + {'align_loss': 0.02480730414390564, + 'recon_loss': 0.07350018620491028, + 'predict_loss': 0.010306022129952908, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2092527598142624, + 'mae_score': 0.012416081815152556, 'data_time': + 0.0007118590001482517, 'model_time': + 1.3050035390187986, 'grad_norm_pre_clip_avg': + 0.21588049829006195, 'learning_rate': + 1.9567525455268913e-05, 'epoch': 4.77} +04/19 [18:11:49] INFO | >> train_qwenlatent.py:487 + Step 18910 | grad_norm_pre_clip=0.1582 | + grad_norm_pre_clip_avg=0.1696 | Metrics: + {'align_loss': 0.02545180171728134, + 'recon_loss': 0.061547886580228806, + 'predict_loss': 0.012258763425052166, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15818646550178528, + 'data_time': 0.0007276850228663534, + 'model_time': 1.2610172870045062, + 'grad_norm_pre_clip_avg': 0.16959325671195985, + 'learning_rate': 1.9560330463436207e-05, + 'epoch': 4.77} +04/19 [18:12:02] INFO | >> train_qwenlatent.py:487 + Step 18920 | grad_norm_pre_clip=0.1791 | + grad_norm_pre_clip_avg=0.1815 | Metrics: + {'align_loss': 0.025216149166226387, + 'recon_loss': 0.07827677577733994, + 'predict_loss': 0.01816696673631668, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17913702130317688, + 'data_time': 0.001027474005240947, + 'model_time': 1.2500659209908918, + 'grad_norm_pre_clip_avg': 0.18148700296878814, + 'learning_rate': 1.9553132036576545e-05, + 'epoch': 4.77} +04/19 [18:12:14] INFO | >> train_qwenlatent.py:487 + Step 18930 | grad_norm_pre_clip=0.2670 | + grad_norm_pre_clip_avg=0.2728 | Metrics: + {'align_loss': 0.024954887107014656, + 'recon_loss': 0.05320260301232338, + 'predict_loss': 0.009555438533425331, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2669740915298462, + 'data_time': 0.0008075640071183443, + 'model_time': 1.2614706449967343, + 'grad_norm_pre_clip_avg': 0.27279516905546186, + 'learning_rate': 1.9545930178198362e-05, + 'epoch': 4.78} +04/19 [18:12:27] INFO | >> train_qwenlatent.py:487 + Step 18940 | grad_norm_pre_clip=0.2917 | + grad_norm_pre_clip_avg=0.2427 | Metrics: + {'align_loss': 0.023866448551416397, + 'recon_loss': 0.06571503728628159, + 'predict_loss': 0.010957271791994572, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.29169249534606934, + 'data_time': 0.0010440720070619136, + 'model_time': 1.3204953120148275, + 'grad_norm_pre_clip_avg': 0.24272942692041397, + 'learning_rate': 1.953872489181175e-05, + 'epoch': 4.78} +04/19 [18:12:40] INFO | >> train_qwenlatent.py:487 + Step 18950 | grad_norm_pre_clip=0.1558 | + grad_norm_pre_clip_avg=0.2123 | Metrics: + {'align_loss': 0.024038201197981834, + 'recon_loss': 0.053637318313121796, + 'predict_loss': 0.008680533617734909, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15580399334430695, + 'mae_score': 0.012428593850350595, 'data_time': + 0.0007138919900171459, 'model_time': + 1.2229690780222882, 'grad_norm_pre_clip_avg': + 0.2122852697968483, 'learning_rate': + 1.9531516180928482e-05, 'epoch': 4.78} +04/19 [18:12:52] INFO | >> train_qwenlatent.py:487 + Step 18960 | grad_norm_pre_clip=0.1996 | + grad_norm_pre_clip_avg=0.1922 | Metrics: + {'align_loss': 0.025824788957834244, + 'recon_loss': 0.062391217797994614, + 'predict_loss': 0.010427474044263363, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19961059093475342, + 'data_time': 0.001199321006424725, + 'model_time': 1.2187099840084556, + 'grad_norm_pre_clip_avg': 0.19216315895318986, + 'learning_rate': 1.9524304049061998e-05, + 'epoch': 4.78} +04/19 [18:13:05] INFO | >> train_qwenlatent.py:487 + Step 18970 | grad_norm_pre_clip=0.2306 | + grad_norm_pre_clip_avg=0.2146 | Metrics: + {'align_loss': 0.026221446692943573, + 'recon_loss': 0.08850900828838348, + 'predict_loss': 0.016250157728791237, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.230597585439682, + 'data_time': 0.0006698689830955118, + 'model_time': 1.2025881060108077, + 'grad_norm_pre_clip_avg': 0.21459164768457412, + 'learning_rate': 1.95170884997274e-05, 'epoch': + 4.79} +04/19 [18:13:17] INFO | >> train_qwenlatent.py:487 + Step 18980 | grad_norm_pre_clip=0.1571 | + grad_norm_pre_clip_avg=0.1980 | Metrics: + {'align_loss': 0.02550623193383217, + 'recon_loss': 0.08696763217449188, + 'predict_loss': 0.02255837805569172, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15711089968681335, + 'data_time': 0.000745068013202399, + 'model_time': 1.245828207989689, + 'grad_norm_pre_clip_avg': 0.19796164333820343, + 'learning_rate': 1.950986953644146e-05, + 'epoch': 4.79} +04/19 [18:13:30] INFO | >> train_qwenlatent.py:487 + Step 18990 | grad_norm_pre_clip=0.1528 | + grad_norm_pre_clip_avg=0.1715 | Metrics: + {'align_loss': 0.02552708610892296, + 'recon_loss': 0.057676203548908234, + 'predict_loss': 0.010580228641629219, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15276601910591125, + 'data_time': 0.0009233430027961731, + 'model_time': 1.193669463013066, + 'grad_norm_pre_clip_avg': 0.17152485400438308, + 'learning_rate': 1.9502647162722612e-05, + 'epoch': 4.79} +04/19 [18:13:43] INFO | >> train_qwenlatent.py:487 + Step 19000 | grad_norm_pre_clip=0.2175 | + grad_norm_pre_clip_avg=0.2239 | Metrics: + {'align_loss': 0.02394464612007141, + 'recon_loss': 0.06191105395555496, + 'predict_loss': 0.010656625032424927, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21745334565639496, + 'mae_score': 0.010558611208254153, 'data_time': + 0.0009505560155957937, 'model_time': + 1.2374883370066527, 'grad_norm_pre_clip_avg': + 0.223922997713089, 'learning_rate': + 1.949542138209095e-05, 'epoch': 4.79} +04/19 [18:13:55] INFO | >> train_qwenlatent.py:487 + Step 19010 | grad_norm_pre_clip=0.2279 | + grad_norm_pre_clip_avg=0.2474 | Metrics: + {'align_loss': 0.024785678833723068, + 'recon_loss': 0.051519058644771576, + 'predict_loss': 0.012018793262541294, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22791078686714172, + 'data_time': 0.0007448300020769238, + 'model_time': 1.209345723997103, + 'grad_norm_pre_clip_avg': 0.24744934290647508, + 'learning_rate': 1.9488192198068242e-05, + 'epoch': 4.8} +04/19 [18:14:08] INFO | >> train_qwenlatent.py:487 + Step 19020 | grad_norm_pre_clip=0.2126 | + grad_norm_pre_clip_avg=0.2151 | Metrics: + {'align_loss': 0.026119690388441086, + 'recon_loss': 0.07605361193418503, + 'predict_loss': 0.011950625106692314, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2125951200723648, + 'data_time': 0.000863481021951884, + 'model_time': 1.2082277979934588, + 'grad_norm_pre_clip_avg': 0.21506761908531188, + 'learning_rate': 1.9480959614177898e-05, + 'epoch': 4.8} +04/19 [18:14:21] INFO | >> train_qwenlatent.py:487 + Step 19030 | grad_norm_pre_clip=0.2281 | + grad_norm_pre_clip_avg=0.1803 | Metrics: + {'align_loss': 0.024446088820695877, + 'recon_loss': 0.07090773433446884, + 'predict_loss': 0.012773478403687477, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22809448838233948, + 'data_time': 0.001255148003110662, + 'model_time': 1.2198656639957335, + 'grad_norm_pre_clip_avg': 0.18031732738018036, + 'learning_rate': 1.947372363394499e-05, + 'epoch': 4.8} +04/19 [18:14:33] INFO | >> train_qwenlatent.py:487 + Step 19040 | grad_norm_pre_clip=0.2399 | + grad_norm_pre_clip_avg=0.1902 | Metrics: + {'align_loss': 0.025493063032627106, + 'recon_loss': 0.052650608122348785, + 'predict_loss': 0.008965813554823399, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23985233902931213, + 'data_time': 0.0010771430097520351, + 'model_time': 1.2481384119892027, + 'grad_norm_pre_clip_avg': 0.1902007967233658, + 'learning_rate': 1.946648426089625e-05, + 'epoch': 4.8} +04/19 [18:14:47] INFO | >> train_qwenlatent.py:487 + Step 19050 | grad_norm_pre_clip=0.2395 | + grad_norm_pre_clip_avg=0.2441 | Metrics: + {'align_loss': 0.025739405304193497, + 'recon_loss': 0.06983894854784012, + 'predict_loss': 0.011731422506272793, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23945072293281555, + 'mae_score': 0.012282415338464685, 'data_time': + 0.0008813299937173724, 'model_time': + 1.4358640459831804, 'grad_norm_pre_clip_avg': + 0.24413589239120484, 'learning_rate': + 1.9459241498560056e-05, 'epoch': 4.81} +04/19 [18:15:00] INFO | >> train_qwenlatent.py:487 + Step 19060 | grad_norm_pre_clip=0.2196 | + grad_norm_pre_clip_avg=0.2215 | Metrics: + {'align_loss': 0.024341855198144913, + 'recon_loss': 0.06798561662435532, + 'predict_loss': 0.014129696413874626, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21955233812332153, + 'data_time': 0.000792293983977288, + 'model_time': 1.2954705790034495, + 'grad_norm_pre_clip_avg': 0.22153161615133285, + 'learning_rate': 1.9451995350466447e-05, + 'epoch': 4.81} +04/19 [18:15:12] INFO | >> train_qwenlatent.py:487 + Step 19070 | grad_norm_pre_clip=0.1775 | + grad_norm_pre_clip_avg=0.2279 | Metrics: + {'align_loss': 0.025296363979578018, + 'recon_loss': 0.06722474098205566, + 'predict_loss': 0.009747686795890331, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17745833098888397, + 'data_time': 0.0011263570049777627, + 'model_time': 1.2408320659887977, + 'grad_norm_pre_clip_avg': 0.22794147282838823, + 'learning_rate': 1.944474582014711e-05, + 'epoch': 4.81} +04/19 [18:15:25] INFO | >> train_qwenlatent.py:487 + Step 19080 | grad_norm_pre_clip=0.1447 | + grad_norm_pre_clip_avg=0.1893 | Metrics: + {'align_loss': 0.024744633585214615, + 'recon_loss': 0.07164739072322845, + 'predict_loss': 0.013694602064788342, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14473064243793488, + 'data_time': 0.0007878179894760251, + 'model_time': 1.2401396700006444, + 'grad_norm_pre_clip_avg': 0.18932267725467683, + 'learning_rate': 1.943749291113537e-05, + 'epoch': 4.81} +04/19 [18:15:37] INFO | >> train_qwenlatent.py:487 + Step 19090 | grad_norm_pre_clip=0.2082 | + grad_norm_pre_clip_avg=0.2424 | Metrics: + {'align_loss': 0.025010082870721817, + 'recon_loss': 0.060394562780857086, + 'predict_loss': 0.008092544972896576, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20815345644950867, + 'data_time': 0.0012283700052648783, + 'model_time': 1.2646188879734837, + 'grad_norm_pre_clip_avg': 0.24239650815725328, + 'learning_rate': 1.9430236626966213e-05, + 'epoch': 4.82} +04/19 [18:15:50] INFO | >> train_qwenlatent.py:487 + Step 19100 | grad_norm_pre_clip=0.1685 | + grad_norm_pre_clip_avg=0.2042 | Metrics: + {'align_loss': 0.02408251166343689, + 'recon_loss': 0.06267894059419632, + 'predict_loss': 0.013549095019698143, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1684597134590149, + 'mae_score': 0.012662506962681676, 'data_time': + 0.0011511790216900408, 'model_time': + 1.2600373679888435, 'grad_norm_pre_clip_avg': + 0.20420369803905486, 'learning_rate': + 1.942297697117626e-05, 'epoch': 4.82} +04/19 [18:16:03] INFO | >> train_qwenlatent.py:487 + Step 19110 | grad_norm_pre_clip=0.1756 | + grad_norm_pre_clip_avg=0.2287 | Metrics: + {'align_loss': 0.02564869076013565, + 'recon_loss': 0.08997346460819244, + 'predict_loss': 0.010694685392081738, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17564311623573303, + 'data_time': 0.0007715999963693321, + 'model_time': 1.196633928018855, + 'grad_norm_pre_clip_avg': 0.22867223024368286, + 'learning_rate': 1.9415713947303783e-05, + 'epoch': 4.82} +04/19 [18:16:15] INFO | >> train_qwenlatent.py:487 + Step 19120 | grad_norm_pre_clip=0.1867 | + grad_norm_pre_clip_avg=0.1950 | Metrics: + {'align_loss': 0.026062259450554848, + 'recon_loss': 0.08170325309038162, + 'predict_loss': 0.011709034442901611, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18672871589660645, + 'data_time': 0.0006278649962041527, + 'model_time': 1.1845025420188904, + 'grad_norm_pre_clip_avg': 0.1949612870812416, + 'learning_rate': 1.9408447558888687e-05, + 'epoch': 4.82} +04/19 [18:16:27] INFO | >> train_qwenlatent.py:487 + Step 19130 | grad_norm_pre_clip=0.1898 | + grad_norm_pre_clip_avg=0.1738 | Metrics: + {'align_loss': 0.024784933775663376, + 'recon_loss': 0.0684249997138977, + 'predict_loss': 0.012192420661449432, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18983964622020721, + 'data_time': 0.0006603380024898797, + 'model_time': 1.1383696330012754, + 'grad_norm_pre_clip_avg': 0.17377731949090958, + 'learning_rate': 1.9401177809472525e-05, + 'epoch': 4.83} +04/19 [18:16:39] INFO | >> train_qwenlatent.py:487 + Step 19140 | grad_norm_pre_clip=0.3301 | + grad_norm_pre_clip_avg=0.2901 | Metrics: + {'align_loss': 0.023944402113556862, + 'recon_loss': 0.06253217160701752, + 'predict_loss': 0.011802292428910732, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.33009016513824463, + 'data_time': 0.0006136489973869175, + 'model_time': 1.1590036659908947, + 'grad_norm_pre_clip_avg': 0.2900912418961525, + 'learning_rate': 1.939390470259848e-05, + 'epoch': 4.83} +04/19 [18:16:51] INFO | >> train_qwenlatent.py:487 + Step 19150 | grad_norm_pre_clip=0.2005 | + grad_norm_pre_clip_avg=0.2159 | Metrics: + {'align_loss': 0.025572188198566437, + 'recon_loss': 0.0835370123386383, + 'predict_loss': 0.015077386982738972, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2004706859588623, + 'mae_score': 0.012711489737570823, 'data_time': + 0.0006587810057681054, 'model_time': + 1.166898080991814, 'grad_norm_pre_clip_avg': + 0.21589162349700927, 'learning_rate': + 1.9386628241811384e-05, 'epoch': 4.83} +04/19 [18:17:03] INFO | >> train_qwenlatent.py:487 + Step 19160 | grad_norm_pre_clip=0.1927 | + grad_norm_pre_clip_avg=0.2011 | Metrics: + {'align_loss': 0.02558186650276184, + 'recon_loss': 0.07724820077419281, + 'predict_loss': 0.01204503420740366, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19274599850177765, + 'data_time': 0.0006295870116446167, + 'model_time': 1.1709296790068038, + 'grad_norm_pre_clip_avg': 0.20105026215314864, + 'learning_rate': 1.937934843065769e-05, + 'epoch': 4.83} +04/19 [18:17:14] INFO | >> train_qwenlatent.py:487 + Step 19170 | grad_norm_pre_clip=0.2021 | + grad_norm_pre_clip_avg=0.2039 | Metrics: + {'align_loss': 0.025696266442537308, + 'recon_loss': 0.08345691859722137, + 'predict_loss': 0.013310743495821953, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.202106311917305, + 'data_time': 0.0005893739871680737, + 'model_time': 1.1443528119998518, + 'grad_norm_pre_clip_avg': 0.20392524152994157, + 'learning_rate': 1.9372065272685484e-05, + 'epoch': 4.84} +04/19 [18:17:26] INFO | >> train_qwenlatent.py:487 + Step 19180 | grad_norm_pre_clip=0.2153 | + grad_norm_pre_clip_avg=0.2095 | Metrics: + {'align_loss': 0.024809718132019043, + 'recon_loss': 0.06816169619560242, + 'predict_loss': 0.011898761615157127, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2152610868215561, + 'data_time': 0.0006629299896303564, + 'model_time': 1.1418850689951796, + 'grad_norm_pre_clip_avg': 0.20953062176704407, + 'learning_rate': 1.9364778771444504e-05, + 'epoch': 4.84} +04/19 [18:17:38] INFO | >> train_qwenlatent.py:487 + Step 19190 | grad_norm_pre_clip=0.2176 | + grad_norm_pre_clip_avg=0.2172 | Metrics: + {'align_loss': 0.025423463433980942, + 'recon_loss': 0.07357748597860336, + 'predict_loss': 0.014104862697422504, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2175581455230713, + 'data_time': 0.0006007349875289947, + 'model_time': 1.1611232229915913, + 'grad_norm_pre_clip_avg': 0.21718828976154328, + 'learning_rate': 1.935748893048609e-05, + 'epoch': 4.84} +04/19 [18:17:50] INFO | >> train_qwenlatent.py:487 + Step 19200 | grad_norm_pre_clip=0.2414 | + grad_norm_pre_clip_avg=0.2245 | Metrics: + {'align_loss': 0.024173028767108917, + 'recon_loss': 0.05574002489447594, + 'predict_loss': 0.010088101029396057, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24140235781669617, + 'mae_score': 0.01425029995205166, 'data_time': + 0.0006092980038374662, 'model_time': + 1.1424079110147431, 'grad_norm_pre_clip_avg': + 0.22451235651969909, 'learning_rate': + 1.9350195753363222e-05, 'epoch': 4.84} +04/19 [18:18:02] INFO | >> train_qwenlatent.py:487 + Step 19210 | grad_norm_pre_clip=0.2199 | + grad_norm_pre_clip_avg=0.2225 | Metrics: + {'align_loss': 0.023952173069119453, + 'recon_loss': 0.0677214190363884, + 'predict_loss': 0.013792180456221104, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2199268490076065, + 'data_time': 0.0006917320133652538, + 'model_time': 1.1649900890188292, + 'grad_norm_pre_clip_avg': 0.22253507673740386, + 'learning_rate': 1.934289924363051e-05, + 'epoch': 4.85} +04/19 [18:18:14] INFO | >> train_qwenlatent.py:487 + Step 19220 | grad_norm_pre_clip=0.2347 | + grad_norm_pre_clip_avg=0.2126 | Metrics: + {'align_loss': 0.025072742253541946, + 'recon_loss': 0.08185677230358124, + 'predict_loss': 0.014749501831829548, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2346872091293335, + 'data_time': 0.0005930839979555458, + 'model_time': 1.1540275440202095, + 'grad_norm_pre_clip_avg': 0.21255844682455063, + 'learning_rate': 1.9335599404844187e-05, + 'epoch': 4.85} +04/19 [18:18:25] INFO | >> train_qwenlatent.py:487 + Step 19230 | grad_norm_pre_clip=0.1700 | + grad_norm_pre_clip_avg=0.1846 | Metrics: + {'align_loss': 0.025569772347807884, + 'recon_loss': 0.06793799251317978, + 'predict_loss': 0.009739791974425316, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16996973752975464, + 'data_time': 0.0005661430186592042, + 'model_time': 1.157590456015896, + 'grad_norm_pre_clip_avg': 0.18458788543939592, + 'learning_rate': 1.93282962405621e-05, 'epoch': + 4.85} +04/19 [18:18:37] INFO | >> train_qwenlatent.py:487 + Step 19240 | grad_norm_pre_clip=0.2819 | + grad_norm_pre_clip_avg=0.2086 | Metrics: + {'align_loss': 0.02617562562227249, + 'recon_loss': 0.07029769569635391, + 'predict_loss': 0.008353258483111858, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2818983197212219, + 'data_time': 0.0006570370169356465, + 'model_time': 1.149289145017974, + 'grad_norm_pre_clip_avg': 0.20859927237033843, + 'learning_rate': 1.9320989754343724e-05, + 'epoch': 4.85} +04/19 [18:18:49] INFO | >> train_qwenlatent.py:487 + Step 19250 | grad_norm_pre_clip=0.2149 | + grad_norm_pre_clip_avg=0.2000 | Metrics: + {'align_loss': 0.023371901363134384, + 'recon_loss': 0.06161772832274437, + 'predict_loss': 0.008181477896869183, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21490593254566193, + 'mae_score': 0.00933451523651948, 'data_time': + 0.0005728040123358369, 'model_time': + 1.1422974760062061, 'grad_norm_pre_clip_avg': + 0.20002569109201432, 'learning_rate': + 1.9313679949750153e-05, 'epoch': 4.86} +04/19 [18:19:00] INFO | >> train_qwenlatent.py:487 + Step 19260 | grad_norm_pre_clip=0.1706 | + grad_norm_pre_clip_avg=0.1981 | Metrics: + {'align_loss': 0.024480029940605164, + 'recon_loss': 0.061171628534793854, + 'predict_loss': 0.012520743533968925, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17061693966388702, + 'data_time': 0.0006144289800431579, + 'model_time': 1.136672513996018, + 'grad_norm_pre_clip_avg': 0.19813202023506166, + 'learning_rate': 1.930636683034409e-05, + 'epoch': 4.86} +04/19 [18:19:12] INFO | >> train_qwenlatent.py:487 + Step 19270 | grad_norm_pre_clip=0.2128 | + grad_norm_pre_clip_avg=0.2345 | Metrics: + {'align_loss': 0.02521677501499653, + 'recon_loss': 0.07334868609905243, + 'predict_loss': 0.010429957881569862, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21275581419467926, + 'data_time': 0.0007707550248596817, + 'model_time': 1.1386449679848738, + 'grad_norm_pre_clip_avg': 0.2344646066427231, + 'learning_rate': 1.9299050399689877e-05, + 'epoch': 4.86} +04/19 [18:19:24] INFO | >> train_qwenlatent.py:487 + Step 19280 | grad_norm_pre_clip=0.1924 | + grad_norm_pre_clip_avg=0.1889 | Metrics: + {'align_loss': 0.025139937177300453, + 'recon_loss': 0.0773882269859314, + 'predict_loss': 0.014899298548698425, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19237251579761505, + 'data_time': 0.0006100430036894977, + 'model_time': 1.1706447009928524, + 'grad_norm_pre_clip_avg': 0.1889492690563202, + 'learning_rate': 1.9291730661353438e-05, + 'epoch': 4.87} +04/19 [18:19:35] INFO | >> train_qwenlatent.py:487 + Step 19290 | grad_norm_pre_clip=0.1895 | + grad_norm_pre_clip_avg=0.1885 | Metrics: + {'align_loss': 0.02562559023499489, + 'recon_loss': 0.09011361002922058, + 'predict_loss': 0.012014242820441723, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18948079645633698, + 'data_time': 0.0005862609832547605, + 'model_time': 1.132908272993518, + 'grad_norm_pre_clip_avg': 0.18849734514951705, + 'learning_rate': 1.928440761890233e-05, + 'epoch': 4.87} +04/19 [18:19:47] INFO | >> train_qwenlatent.py:487 + Step 19300 | grad_norm_pre_clip=0.2349 | + grad_norm_pre_clip_avg=0.1977 | Metrics: + {'align_loss': 0.025467637926340103, + 'recon_loss': 0.0589427575469017, + 'predict_loss': 0.00937731098383665, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.234910249710083, + 'mae_score': 0.013866457208856806, 'data_time': + 0.0006388580077327788, 'model_time': + 1.1569971069984604, 'grad_norm_pre_clip_avg': + 0.19768190532922744, 'learning_rate': + 1.9277081275905716e-05, 'epoch': 4.87} +04/19 [18:19:59] INFO | >> train_qwenlatent.py:487 + Step 19310 | grad_norm_pre_clip=0.2302 | + grad_norm_pre_clip_avg=0.2289 | Metrics: + {'align_loss': 0.025379212573170662, + 'recon_loss': 0.07213827967643738, + 'predict_loss': 0.0114546287804842, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2302091419696808, + 'data_time': 0.0006678750214632601, + 'model_time': 1.179648595978506, + 'grad_norm_pre_clip_avg': 0.22885833978652953, + 'learning_rate': 1.9269751635934362e-05, + 'epoch': 4.87} +04/19 [18:20:11] INFO | >> train_qwenlatent.py:487 + Step 19320 | grad_norm_pre_clip=0.1976 | + grad_norm_pre_clip_avg=0.1999 | Metrics: + {'align_loss': 0.025279823690652847, + 'recon_loss': 0.0669322982430458, + 'predict_loss': 0.009214397519826889, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1976146101951599, + 'data_time': 0.0006051429954823107, + 'model_time': 1.1400913629913703, + 'grad_norm_pre_clip_avg': 0.199863563477993, + 'learning_rate': 1.926241870256065e-05, + 'epoch': 4.88} +04/19 [18:20:22] INFO | >> train_qwenlatent.py:487 + Step 19330 | grad_norm_pre_clip=0.2154 | + grad_norm_pre_clip_avg=0.1945 | Metrics: + {'align_loss': 0.025008097290992737, + 'recon_loss': 0.06443469971418381, + 'predict_loss': 0.010385680012404919, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21538713574409485, + 'data_time': 0.0006157909811008722, + 'model_time': 1.138411197985988, + 'grad_norm_pre_clip_avg': 0.19451480209827424, + 'learning_rate': 1.9255082479358566e-05, + 'epoch': 4.88} +04/19 [18:20:34] INFO | >> train_qwenlatent.py:487 + Step 19340 | grad_norm_pre_clip=0.2565 | + grad_norm_pre_clip_avg=0.2173 | Metrics: + {'align_loss': 0.02398386225104332, + 'recon_loss': 0.05455358698964119, + 'predict_loss': 0.00960969366133213, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25646764039993286, + 'data_time': 0.0006077869911678135, + 'model_time': 1.1361116190091707, + 'grad_norm_pre_clip_avg': 0.21730081737041473, + 'learning_rate': 1.9247742969903686e-05, + 'epoch': 4.88} +04/19 [18:20:46] INFO | >> train_qwenlatent.py:487 + Step 19350 | grad_norm_pre_clip=0.1892 | + grad_norm_pre_clip_avg=0.1989 | Metrics: + {'align_loss': 0.02585887722671032, + 'recon_loss': 0.07883527129888535, + 'predict_loss': 0.00859104935079813, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18917790055274963, + 'mae_score': 0.00990220662709829, 'data_time': + 0.0007050300191622227, 'model_time': + 1.1404296309920028, 'grad_norm_pre_clip_avg': + 0.19889715313911438, 'learning_rate': + 1.92404001777732e-05, 'epoch': 4.88} +04/19 [18:20:58] INFO | >> train_qwenlatent.py:487 + Step 19360 | grad_norm_pre_clip=0.2044 | + grad_norm_pre_clip_avg=0.2008 | Metrics: + {'align_loss': 0.025738989934325218, + 'recon_loss': 0.09210287779569626, + 'predict_loss': 0.011324984021484852, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20438715815544128, + 'data_time': 0.0005852949980180711, + 'model_time': 1.1392181800038088, + 'grad_norm_pre_clip_avg': 0.20083284378051758, + 'learning_rate': 1.9233054106545904e-05, + 'epoch': 4.89} +04/19 [18:21:09] INFO | >> train_qwenlatent.py:487 + Step 19370 | grad_norm_pre_clip=0.1976 | + grad_norm_pre_clip_avg=0.2126 | Metrics: + {'align_loss': 0.025037214159965515, + 'recon_loss': 0.06567985564470291, + 'predict_loss': 0.016106706112623215, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1976499706506729, + 'data_time': 0.0005722799978684634, + 'model_time': 1.1306760759907775, + 'grad_norm_pre_clip_avg': 0.21260423511266707, + 'learning_rate': 1.9225704759802183e-05, + 'epoch': 4.89} +04/19 [18:21:21] INFO | >> train_qwenlatent.py:487 + Step 19380 | grad_norm_pre_clip=0.2036 | + grad_norm_pre_clip_avg=0.1902 | Metrics: + {'align_loss': 0.024666093289852142, + 'recon_loss': 0.07101713120937347, + 'predict_loss': 0.010470275767147541, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2035636454820633, + 'data_time': 0.0005966949975118041, + 'model_time': 1.167188896011794, + 'grad_norm_pre_clip_avg': 0.19023680835962295, + 'learning_rate': 1.9218352141124008e-05, + 'epoch': 4.89} +04/19 [18:21:33] INFO | >> train_qwenlatent.py:487 + Step 19390 | grad_norm_pre_clip=0.1764 | + grad_norm_pre_clip_avg=0.2199 | Metrics: + {'align_loss': 0.02485547587275505, + 'recon_loss': 0.0738878920674324, + 'predict_loss': 0.006472688168287277, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17635773122310638, + 'data_time': 0.0005564519960898906, + 'model_time': 1.1901764139765874, + 'grad_norm_pre_clip_avg': 0.21988295614719391, + 'learning_rate': 1.9210996254094965e-05, + 'epoch': 4.89} +04/19 [18:21:45] INFO | >> train_qwenlatent.py:487 + Step 19400 | grad_norm_pre_clip=0.2185 | + grad_norm_pre_clip_avg=0.1912 | Metrics: + {'align_loss': 0.02460002899169922, + 'recon_loss': 0.049269914627075195, + 'predict_loss': 0.012380090542137623, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21850785613059998, + 'mae_score': 0.014053522573935018, 'data_time': + 0.000684870989061892, 'model_time': + 1.1634817839949392, 'grad_norm_pre_clip_avg': + 0.19116119295358658, 'learning_rate': + 1.920363710230022e-05, 'epoch': 4.9} +04/19 [18:21:56] INFO | >> train_qwenlatent.py:487 + Step 19410 | grad_norm_pre_clip=0.1892 | + grad_norm_pre_clip_avg=0.2084 | Metrics: + {'align_loss': 0.02420983649790287, + 'recon_loss': 0.048382680863142014, + 'predict_loss': 0.006304446142166853, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18921253085136414, + 'data_time': 0.0005800139915663749, + 'model_time': 1.1438602110138163, + 'grad_norm_pre_clip_avg': 0.20838020443916322, + 'learning_rate': 1.919627468932654e-05, + 'epoch': 4.9} +04/19 [18:22:08] INFO | >> train_qwenlatent.py:487 + Step 19420 | grad_norm_pre_clip=0.2478 | + grad_norm_pre_clip_avg=0.1811 | Metrics: + {'align_loss': 0.02548234909772873, + 'recon_loss': 0.0705612301826477, + 'predict_loss': 0.010997192934155464, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24784238636493683, + 'data_time': 0.0005871249886695296, + 'model_time': 1.1582937829953153, + 'grad_norm_pre_clip_avg': 0.18107617050409316, + 'learning_rate': 1.9188909018762267e-05, + 'epoch': 4.9} +04/19 [18:22:43] INFO | >> train_qwenlatent.py:487 + Step 19430 | grad_norm_pre_clip=0.1682 | + grad_norm_pre_clip_avg=0.2690 | Metrics: + {'align_loss': 0.024668481200933456, + 'recon_loss': 0.07159329205751419, + 'predict_loss': 0.013987813144922256, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1681598424911499, + 'data_time': 0.009960736002540216, + 'model_time': 3.3904236429953016, + 'grad_norm_pre_clip_avg': 0.2689871072769165, + 'learning_rate': 1.918154009419735e-05, + 'epoch': 4.9} +04/19 [18:23:17] INFO | >> train_qwenlatent.py:487 + Step 19440 | grad_norm_pre_clip=0.2322 | + grad_norm_pre_clip_avg=0.1917 | Metrics: + {'align_loss': 0.026213806122541428, + 'recon_loss': 0.07600844651460648, + 'predict_loss': 0.009810484014451504, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23222452402114868, + 'data_time': 0.003434805985307321, + 'model_time': 3.5126191029848997, + 'grad_norm_pre_clip_avg': 0.19171012341976165, + 'learning_rate': 1.9174167919223305e-05, + 'epoch': 4.91} +04/19 [18:23:55] INFO | >> train_qwenlatent.py:487 + Step 19450 | grad_norm_pre_clip=0.1547 | + grad_norm_pre_clip_avg=0.1740 | Metrics: + {'align_loss': 0.0237121544778347, + 'recon_loss': 0.06164811551570892, + 'predict_loss': 0.01370232179760933, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1546994298696518, + 'mae_score': 0.014594670888539907, 'data_time': + 0.006173769012093544, 'model_time': + 3.258423298015259, 'grad_norm_pre_clip_avg': + 0.17397128641605378, 'learning_rate': + 1.9166792497433243e-05, 'epoch': 4.91} +04/19 [18:24:31] INFO | >> train_qwenlatent.py:487 + Step 19460 | grad_norm_pre_clip=0.3492 | + grad_norm_pre_clip_avg=0.1996 | Metrics: + {'align_loss': 0.027203867211937904, + 'recon_loss': 0.11442361027002335, + 'predict_loss': 0.01553119346499443, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.34917786717414856, + 'data_time': 0.0011381240037735552, + 'model_time': 3.2000886909954716, + 'grad_norm_pre_clip_avg': 0.19959206730127335, + 'learning_rate': 1.9159413832421856e-05, + 'epoch': 4.91} +04/19 [18:25:05] INFO | >> train_qwenlatent.py:487 + Step 19470 | grad_norm_pre_clip=0.2183 | + grad_norm_pre_clip_avg=0.2571 | Metrics: + {'align_loss': 0.025894321501255035, + 'recon_loss': 0.07816703617572784, + 'predict_loss': 0.012947426177561283, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21827180683612823, + 'data_time': 0.0015999289753381163, + 'model_time': 3.157584102009423, + 'grad_norm_pre_clip_avg': 0.2571302726864815, + 'learning_rate': 1.915203192778541e-05, + 'epoch': 4.91} +04/19 [18:25:43] INFO | >> train_qwenlatent.py:487 + Step 19480 | grad_norm_pre_clip=0.1499 | + grad_norm_pre_clip_avg=0.2004 | Metrics: + {'align_loss': 0.025153784081339836, + 'recon_loss': 0.07634083926677704, + 'predict_loss': 0.013841361738741398, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1498885154724121, + 'data_time': 0.0014850420120637864, + 'model_time': 3.2446705180045683, + 'grad_norm_pre_clip_avg': 0.20041152834892273, + 'learning_rate': 1.9144646787121764e-05, + 'epoch': 4.92} +04/19 [18:26:16] INFO | >> train_qwenlatent.py:487 + Step 19490 | grad_norm_pre_clip=0.2012 | + grad_norm_pre_clip_avg=0.1786 | Metrics: + {'align_loss': 0.025213006883859634, + 'recon_loss': 0.10021266341209412, + 'predict_loss': 0.01980273239314556, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20120932161808014, + 'data_time': 0.0019964720122516155, + 'model_time': 3.1403123369964305, + 'grad_norm_pre_clip_avg': 0.178625126183033, + 'learning_rate': 1.9137258414030344e-05, + 'epoch': 4.92} +04/19 [18:26:44] INFO | >> train_qwenlatent.py:487 + Step 19500 | grad_norm_pre_clip=0.1417 | + grad_norm_pre_clip_avg=0.2210 | Metrics: + {'align_loss': 0.026862621307373047, + 'recon_loss': 0.0930783823132515, + 'predict_loss': 0.008377764374017715, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.141653910279274, + 'mae_score': 0.011271948427767367, 'data_time': + 0.0014319100009743124, 'model_time': + 2.355714521021582, 'grad_norm_pre_clip_avg': + 0.22097242176532744, 'learning_rate': + 1.9129866812112145e-05, 'epoch': 4.92} +04/19 [18:27:06] INFO | >> train_qwenlatent.py:487 + Step 19510 | grad_norm_pre_clip=0.1956 | + grad_norm_pre_clip_avg=0.2209 | Metrics: + {'align_loss': 0.0251187551766634, + 'recon_loss': 0.06460350006818771, + 'predict_loss': 0.007984276860952377, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.195560485124588, + 'data_time': 0.0010660900152288377, + 'model_time': 1.879990165005438, + 'grad_norm_pre_clip_avg': 0.22094213962554932, + 'learning_rate': 1.912247198496976e-05, + 'epoch': 4.92} +04/19 [18:27:21] INFO | >> train_qwenlatent.py:487 + Step 19520 | grad_norm_pre_clip=0.3163 | + grad_norm_pre_clip_avg=0.2270 | Metrics: + {'align_loss': 0.025548994541168213, + 'recon_loss': 0.07249684631824493, + 'predict_loss': 0.00910695269703865, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3163321614265442, + 'data_time': 0.0010570430022198707, + 'model_time': 1.3212131620093714, + 'grad_norm_pre_clip_avg': 0.22703259885311128, + 'learning_rate': 1.911507393620732e-05, + 'epoch': 4.93} +04/19 [18:27:34] INFO | >> train_qwenlatent.py:487 + Step 19530 | grad_norm_pre_clip=0.1682 | + grad_norm_pre_clip_avg=0.2097 | Metrics: + {'align_loss': 0.02439691126346588, + 'recon_loss': 0.05469219759106636, + 'predict_loss': 0.011411439627408981, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16815496981143951, + 'data_time': 0.0008548719924874604, + 'model_time': 1.2309491229825653, + 'grad_norm_pre_clip_avg': 0.2097321406006813, + 'learning_rate': 1.9107672669430555e-05, + 'epoch': 4.93} +04/19 [18:27:46] INFO | >> train_qwenlatent.py:487 + Step 19540 | grad_norm_pre_clip=0.2328 | + grad_norm_pre_clip_avg=0.1916 | Metrics: + {'align_loss': 0.025699065998196602, + 'recon_loss': 0.08331422507762909, + 'predict_loss': 0.010196095332503319, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2327868640422821, + 'data_time': 0.000900059996638447, + 'model_time': 1.2497141280036885, + 'grad_norm_pre_clip_avg': 0.191579470038414, + 'learning_rate': 1.910026818824675e-05, + 'epoch': 4.93} +04/19 [18:28:00] INFO | >> train_qwenlatent.py:487 + Step 19550 | grad_norm_pre_clip=0.1868 | + grad_norm_pre_clip_avg=0.2202 | Metrics: + {'align_loss': 0.024844402447342873, + 'recon_loss': 0.07013585418462753, + 'predict_loss': 0.010333885438740253, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18684092164039612, + 'mae_score': 0.010468606691102724, 'data_time': + 0.0008178220014087856, 'model_time': + 1.263314409996383, 'grad_norm_pre_clip_avg': + 0.22020696997642517, 'learning_rate': + 1.9092860496264764e-05, 'epoch': 4.93} +04/19 [18:28:12] INFO | >> train_qwenlatent.py:487 + Step 19560 | grad_norm_pre_clip=0.1984 | + grad_norm_pre_clip_avg=0.2222 | Metrics: + {'align_loss': 0.02506049908697605, + 'recon_loss': 0.06800804287195206, + 'predict_loss': 0.010481095872819424, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1984359174966812, + 'data_time': 0.0007304000027943403, + 'model_time': 1.2301159600028768, + 'grad_norm_pre_clip_avg': 0.22217073291540146, + 'learning_rate': 1.9085449597095005e-05, + 'epoch': 4.94} +04/19 [18:28:25] INFO | >> train_qwenlatent.py:487 + Step 19570 | grad_norm_pre_clip=0.2006 | + grad_norm_pre_clip_avg=0.2228 | Metrics: + {'align_loss': 0.026217717677354813, + 'recon_loss': 0.06965109705924988, + 'predict_loss': 0.007095813285559416, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2005859911441803, + 'data_time': 0.0005876710056327283, + 'model_time': 1.2222874269937165, + 'grad_norm_pre_clip_avg': 0.22279895544052125, + 'learning_rate': 1.9078035494349467e-05, + 'epoch': 4.94} +04/19 [18:28:37] INFO | >> train_qwenlatent.py:487 + Step 19580 | grad_norm_pre_clip=0.1730 | + grad_norm_pre_clip_avg=0.1951 | Metrics: + {'align_loss': 0.025459179654717445, + 'recon_loss': 0.07166431099176407, + 'predict_loss': 0.013416252098977566, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17298269271850586, + 'data_time': 0.0005941159906797111, + 'model_time': 1.2207482770027127, + 'grad_norm_pre_clip_avg': 0.19505199342966079, + 'learning_rate': 1.907061819164168e-05, + 'epoch': 4.94} +04/19 [18:28:51] INFO | >> train_qwenlatent.py:487 + Step 19590 | grad_norm_pre_clip=0.2094 | + grad_norm_pre_clip_avg=0.1945 | Metrics: + {'align_loss': 0.025789808481931686, + 'recon_loss': 0.08063668757677078, + 'predict_loss': 0.008919489569962025, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20942707359790802, + 'data_time': 0.0005805580003652722, + 'model_time': 1.2680940459831618, + 'grad_norm_pre_clip_avg': 0.19452909529209136, + 'learning_rate': 1.9063197692586758e-05, + 'epoch': 4.94} +04/19 [18:29:04] INFO | >> train_qwenlatent.py:487 + Step 19600 | grad_norm_pre_clip=0.2931 | + grad_norm_pre_clip_avg=0.2000 | Metrics: + {'align_loss': 0.02502666972577572, + 'recon_loss': 0.07879246771335602, + 'predict_loss': 0.01437184028327465, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.29307499527931213, + 'mae_score': 0.016181151931350297, 'data_time': + 0.0009493999823462218, 'model_time': + 1.2679152280034032, 'grad_norm_pre_clip_avg': + 0.19995666891336442, 'learning_rate': + 1.9055774000801365e-05, 'epoch': 4.95} +04/19 [18:29:16] INFO | >> train_qwenlatent.py:487 + Step 19610 | grad_norm_pre_clip=0.1834 | + grad_norm_pre_clip_avg=0.2061 | Metrics: + {'align_loss': 0.02468002215027809, + 'recon_loss': 0.07333803921937943, + 'predict_loss': 0.010944361798465252, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18344804644584656, + 'data_time': 0.0012740950041916221, + 'model_time': 1.2447148580104113, + 'grad_norm_pre_clip_avg': 0.20606889575719833, + 'learning_rate': 1.9048347119903708e-05, + 'epoch': 4.95} +04/19 [18:29:29] INFO | >> train_qwenlatent.py:487 + Step 19620 | grad_norm_pre_clip=0.1972 | + grad_norm_pre_clip_avg=0.1847 | Metrics: + {'align_loss': 0.025321323424577713, + 'recon_loss': 0.08098922669887543, + 'predict_loss': 0.009864768013358116, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1971862018108368, + 'data_time': 0.00093327701324597, 'model_time': + 1.24023307999596, 'grad_norm_pre_clip_avg': + 0.1847255364060402, 'learning_rate': + 1.9040917053513562e-05, 'epoch': 4.95} +04/19 [18:29:42] INFO | >> train_qwenlatent.py:487 + Step 19630 | grad_norm_pre_clip=0.2473 | + grad_norm_pre_clip_avg=0.1925 | Metrics: + {'align_loss': 0.02543269470334053, + 'recon_loss': 0.06338676065206528, + 'predict_loss': 0.0076142167672514915, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2472672164440155, + 'data_time': 0.0008269190147984773, + 'model_time': 1.2177944690047298, + 'grad_norm_pre_clip_avg': 0.19247580021619798, + 'learning_rate': 1.9033483805252254e-05, + 'epoch': 4.95} +04/19 [18:29:54] INFO | >> train_qwenlatent.py:487 + Step 19640 | grad_norm_pre_clip=0.2832 | + grad_norm_pre_clip_avg=0.2718 | Metrics: + {'align_loss': 0.02506038174033165, + 'recon_loss': 0.08218268305063248, + 'predict_loss': 0.01476422231644392, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2831712067127228, + 'data_time': 0.0012843020085711032, + 'model_time': 1.1919885840034112, + 'grad_norm_pre_clip_avg': 0.27175692915916444, + 'learning_rate': 1.9026047378742663e-05, + 'epoch': 4.96} +04/19 [18:30:08] INFO | >> train_qwenlatent.py:487 + Step 19650 | grad_norm_pre_clip=0.1792 | + grad_norm_pre_clip_avg=0.2141 | Metrics: + {'align_loss': 0.02490128204226494, + 'recon_loss': 0.068926602602005, + 'predict_loss': 0.01389314979314804, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17922279238700867, + 'mae_score': 0.012546064187814523, 'data_time': + 0.0009547309891786426, 'model_time': + 1.2036550749908201, 'grad_norm_pre_clip_avg': + 0.2140781432390213, 'learning_rate': + 1.9018607777609202e-05, 'epoch': 4.96} +04/19 [18:30:20] INFO | >> train_qwenlatent.py:487 + Step 19660 | grad_norm_pre_clip=0.1798 | + grad_norm_pre_clip_avg=0.2012 | Metrics: + {'align_loss': 0.026308894157409668, + 'recon_loss': 0.06887269020080566, + 'predict_loss': 0.010561859235167503, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17981892824172974, + 'data_time': 0.0012628980039153248, + 'model_time': 1.257283421989996, + 'grad_norm_pre_clip_avg': 0.20123373419046403, + 'learning_rate': 1.9011165005477843e-05, + 'epoch': 4.96} +04/19 [18:30:33] INFO | >> train_qwenlatent.py:487 + Step 19670 | grad_norm_pre_clip=0.2029 | + grad_norm_pre_clip_avg=0.1878 | Metrics: + {'align_loss': 0.025400688871741295, + 'recon_loss': 0.0784171000123024, + 'predict_loss': 0.011343762278556824, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20289473235607147, + 'data_time': 0.0011198309948667884, + 'model_time': 1.257495152996853, + 'grad_norm_pre_clip_avg': 0.1878429099917412, + 'learning_rate': 1.900371906597611e-05, + 'epoch': 4.96} +04/19 [18:30:45] INFO | >> train_qwenlatent.py:487 + Step 19680 | grad_norm_pre_clip=0.2586 | + grad_norm_pre_clip_avg=0.2670 | Metrics: + {'align_loss': 0.024736110121011734, + 'recon_loss': 0.08101289719343185, + 'predict_loss': 0.018794208765029907, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2585674226284027, + 'data_time': 0.0006749029853381217, + 'model_time': 1.2385011740261689, + 'grad_norm_pre_clip_avg': 0.2670228749513626, + 'learning_rate': 1.8996269962733062e-05, + 'epoch': 4.97} +04/19 [18:30:58] INFO | >> train_qwenlatent.py:487 + Step 19690 | grad_norm_pre_clip=0.1665 | + grad_norm_pre_clip_avg=0.2138 | Metrics: + {'align_loss': 0.025277338922023773, + 'recon_loss': 0.06589481234550476, + 'predict_loss': 0.01000126264989376, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16651111841201782, + 'data_time': 0.000604342989390716, + 'model_time': 1.2326165929844137, + 'grad_norm_pre_clip_avg': 0.21381557136774063, + 'learning_rate': 1.89888176993793e-05, 'epoch': + 4.97} +04/19 [18:31:11] INFO | >> train_qwenlatent.py:487 + Step 19700 | grad_norm_pre_clip=0.1652 | + grad_norm_pre_clip_avg=0.1843 | Metrics: + {'align_loss': 0.025014374405145645, + 'recon_loss': 0.09150917828083038, + 'predict_loss': 0.014079360291361809, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16520407795906067, + 'mae_score': 0.013671196258819854, 'data_time': + 0.0009999590110965073, 'model_time': + 1.2086102280009072, 'grad_norm_pre_clip_avg': + 0.18426723629236222, 'learning_rate': + 1.8981362279546965e-05, 'epoch': 4.97} +04/19 [18:31:24] INFO | >> train_qwenlatent.py:487 + Step 19710 | grad_norm_pre_clip=0.1725 | + grad_norm_pre_clip_avg=0.2016 | Metrics: + {'align_loss': 0.025290196761488914, + 'recon_loss': 0.05831611156463623, + 'predict_loss': 0.008719970472157001, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17247438430786133, + 'data_time': 0.0006102119805291295, + 'model_time': 1.2734982580004726, + 'grad_norm_pre_clip_avg': 0.20157861709594727, + 'learning_rate': 1.897390370686974e-05, + 'epoch': 4.97} +04/19 [18:31:37] INFO | >> train_qwenlatent.py:487 + Step 19720 | grad_norm_pre_clip=0.2255 | + grad_norm_pre_clip_avg=0.2177 | Metrics: + {'align_loss': 0.0252352524548769, + 'recon_loss': 0.0900336503982544, + 'predict_loss': 0.013473368249833584, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22549837827682495, + 'data_time': 0.0011607329943217337, + 'model_time': 1.2081548440037295, + 'grad_norm_pre_clip_avg': 0.21772127747535705, + 'learning_rate': 1.8966441984982845e-05, + 'epoch': 4.98} +04/19 [18:31:49] INFO | >> train_qwenlatent.py:487 + Step 19730 | grad_norm_pre_clip=0.1838 | + grad_norm_pre_clip_avg=0.1964 | Metrics: + {'align_loss': 0.0259591992944479, + 'recon_loss': 0.0882791057229042, + 'predict_loss': 0.01459409948438406, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1837749034166336, + 'data_time': 0.0008043849957175553, + 'model_time': 1.2363195840152912, + 'grad_norm_pre_clip_avg': 0.19638024568557738, + 'learning_rate': 1.8958977117523026e-05, + 'epoch': 4.98} +04/19 [18:32:02] INFO | >> train_qwenlatent.py:487 + Step 19740 | grad_norm_pre_clip=0.2228 | + grad_norm_pre_clip_avg=0.2056 | Metrics: + {'align_loss': 0.024650994688272476, + 'recon_loss': 0.06007584184408188, + 'predict_loss': 0.00871434435248375, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22275196015834808, + 'data_time': 0.0008172719972208142, + 'model_time': 1.24750043801032, + 'grad_norm_pre_clip_avg': 0.20562727153301238, + 'learning_rate': 1.8951509108128575e-05, + 'epoch': 4.98} +04/19 [18:32:15] INFO | >> train_qwenlatent.py:487 + Step 19750 | grad_norm_pre_clip=0.2540 | + grad_norm_pre_clip_avg=0.2343 | Metrics: + {'align_loss': 0.025794900953769684, + 'recon_loss': 0.08009912818670273, + 'predict_loss': 0.012573770247399807, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25398460030555725, + 'mae_score': 0.012334560703586888, 'data_time': + 0.0009180239867419004, 'model_time': + 1.1952385950135067, 'grad_norm_pre_clip_avg': + 0.23425355553627014, 'learning_rate': + 1.8944037960439303e-05, 'epoch': 4.98} +04/19 [18:32:28] INFO | >> train_qwenlatent.py:487 + Step 19760 | grad_norm_pre_clip=0.1713 | + grad_norm_pre_clip_avg=0.2236 | Metrics: + {'align_loss': 0.02416907623410225, + 'recon_loss': 0.06805792450904846, + 'predict_loss': 0.00923078041523695, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17131473124027252, + 'data_time': 0.0005953969957772642, + 'model_time': 1.2214766409888398, + 'grad_norm_pre_clip_avg': 0.22359893023967742, + 'learning_rate': 1.8936563678096562e-05, + 'epoch': 4.99} +04/19 [18:32:40] INFO | >> train_qwenlatent.py:487 + Step 19770 | grad_norm_pre_clip=0.1940 | + grad_norm_pre_clip_avg=0.1791 | Metrics: + {'align_loss': 0.024430830031633377, + 'recon_loss': 0.0616120919585228, + 'predict_loss': 0.011210540309548378, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19398614764213562, + 'data_time': 0.0008795020112302154, + 'model_time': 1.2217328289989382, + 'grad_norm_pre_clip_avg': 0.17913555353879929, + 'learning_rate': 1.8929086264743227e-05, + 'epoch': 4.99} +04/19 [18:32:53] INFO | >> train_qwenlatent.py:487 + Step 19780 | grad_norm_pre_clip=0.2205 | + grad_norm_pre_clip_avg=0.1827 | Metrics: + {'align_loss': 0.023807711899280548, + 'recon_loss': 0.05658649280667305, + 'predict_loss': 0.010933120734989643, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22047647833824158, + 'data_time': 0.0008772260043770075, + 'model_time': 1.2508733060094528, + 'grad_norm_pre_clip_avg': 0.1826692834496498, + 'learning_rate': 1.8921605724023694e-05, + 'epoch': 4.99} +04/19 [18:33:05] INFO | >> train_qwenlatent.py:487 + Step 19790 | grad_norm_pre_clip=0.2160 | + grad_norm_pre_clip_avg=0.2058 | Metrics: + {'align_loss': 0.024249447509646416, + 'recon_loss': 0.06676830351352692, + 'predict_loss': 0.007919264025986195, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21602384746074677, + 'data_time': 0.0009546070068608969, + 'model_time': 1.2645667109754868, + 'grad_norm_pre_clip_avg': 0.20578060746192933, + 'learning_rate': 1.891412205958389e-05, + 'epoch': 4.99} +04/19 [18:33:18] INFO | >> train_qwenlatent.py:487 + Step 19800 | grad_norm_pre_clip=0.1714 | + grad_norm_pre_clip_avg=0.2079 | Metrics: + {'align_loss': 0.025010844692587852, + 'recon_loss': 0.06857974827289581, + 'predict_loss': 0.012242271564900875, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17143107950687408, + 'mae_score': 0.011399247624852636, 'data_time': + 0.0005951799976173788, 'model_time': + 1.1994874149968382, 'grad_norm_pre_clip_avg': + 0.20794905126094818, 'learning_rate': + 1.8906635275071265e-05, 'epoch': 5.0} +04/19 [18:33:31] INFO | >> train_qwenlatent.py:487 + Step 19810 | grad_norm_pre_clip=0.1881 | + grad_norm_pre_clip_avg=0.1900 | Metrics: + {'align_loss': 0.02457830309867859, + 'recon_loss': 0.06251608580350876, + 'predict_loss': 0.01126179937273264, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18805034458637238, + 'data_time': 0.0010289149940945208, + 'model_time': 1.2944988289964385, + 'grad_norm_pre_clip_avg': 0.19001242220401765, + 'learning_rate': 1.8899145374134778e-05, + 'epoch': 5.0} +04/19 [18:33:43] INFO | >> train_qwenlatent.py:487 + Step 19820 | grad_norm_pre_clip=0.2473 | + grad_norm_pre_clip_avg=0.1933 | Metrics: + {'align_loss': 0.025082621723413467, + 'recon_loss': 0.06195811554789543, + 'predict_loss': 0.00971017312258482, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2472684234380722, + 'data_time': 0.0007428279786836356, + 'model_time': 1.2232240310113411, + 'grad_norm_pre_clip_avg': 0.1933495134115219, + 'learning_rate': 1.8891652360424926e-05, + 'epoch': 5.0} +04/19 [18:33:56] INFO | >> train_qwenlatent.py:487 + Step 19830 | grad_norm_pre_clip=0.2263 | + grad_norm_pre_clip_avg=0.2268 | Metrics: + {'align_loss': 0.02663682959973812, + 'recon_loss': 0.07833676785230637, + 'predict_loss': 0.016179785132408142, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2263302206993103, + 'data_time': 0.0008843370014801621, + 'model_time': 1.2708351699984632, + 'grad_norm_pre_clip_avg': 0.2267847999930382, + 'learning_rate': 1.8884156237593708e-05, + 'epoch': 5.0} +04/19 [18:34:08] INFO | >> train_qwenlatent.py:487 + Step 19840 | grad_norm_pre_clip=0.2466 | + grad_norm_pre_clip_avg=0.2206 | Metrics: + {'align_loss': 0.026033611968159676, + 'recon_loss': 0.08946231007575989, + 'predict_loss': 0.014858581125736237, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24657128751277924, + 'data_time': 0.0010889889963436872, + 'model_time': 1.2151277380180545, + 'grad_norm_pre_clip_avg': 0.22057054787874222, + 'learning_rate': 1.887665700929464e-05, + 'epoch': 5.01} +04/19 [18:34:22] INFO | >> train_qwenlatent.py:487 + Step 19850 | grad_norm_pre_clip=0.1843 | + grad_norm_pre_clip_avg=0.1920 | Metrics: + {'align_loss': 0.0250575952231884, + 'recon_loss': 0.06091633066534996, + 'predict_loss': 0.012413440272212029, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18434731662273407, + 'mae_score': 0.009653087134833808, 'data_time': + 0.000840755004901439, 'model_time': + 1.2425304839853197, 'grad_norm_pre_clip_avg': + 0.19203309565782548, 'learning_rate': + 1.8869154679182764e-05, 'epoch': 5.01} +04/19 [18:34:35] INFO | >> train_qwenlatent.py:487 + Step 19860 | grad_norm_pre_clip=0.2614 | + grad_norm_pre_clip_avg=0.2512 | Metrics: + {'align_loss': 0.025137649849057198, + 'recon_loss': 0.08465103805065155, + 'predict_loss': 0.00998356007039547, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26139935851097107, + 'data_time': 0.0011715139844454825, + 'model_time': 1.2722278690198436, + 'grad_norm_pre_clip_avg': 0.2511726841330528, + 'learning_rate': 1.8861649250914615e-05, + 'epoch': 5.01} +04/19 [18:34:47] INFO | >> train_qwenlatent.py:487 + Step 19870 | grad_norm_pre_clip=0.1903 | + grad_norm_pre_clip_avg=0.2101 | Metrics: + {'align_loss': 0.02457437850534916, + 'recon_loss': 0.08190391957759857, + 'predict_loss': 0.013632446527481079, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1903335005044937, + 'data_time': 0.0006552400009240955, + 'model_time': 1.2196892560168635, + 'grad_norm_pre_clip_avg': 0.2100612699985504, + 'learning_rate': 1.885414072814825e-05, + 'epoch': 5.01} +04/19 [18:34:59] INFO | >> train_qwenlatent.py:487 + Step 19880 | grad_norm_pre_clip=0.2223 | + grad_norm_pre_clip_avg=0.1993 | Metrics: + {'align_loss': 0.025496138259768486, + 'recon_loss': 0.08906890451908112, + 'predict_loss': 0.015300207771360874, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2222660630941391, + 'data_time': 0.0006219600036274642, + 'model_time': 1.2459941589913797, + 'grad_norm_pre_clip_avg': 0.19926677793264388, + 'learning_rate': 1.8846629114543238e-05, + 'epoch': 5.02} +04/19 [18:35:12] INFO | >> train_qwenlatent.py:487 + Step 19890 | grad_norm_pre_clip=0.1822 | + grad_norm_pre_clip_avg=0.1753 | Metrics: + {'align_loss': 0.025287725031375885, + 'recon_loss': 0.07957881689071655, + 'predict_loss': 0.017514925450086594, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18224318325519562, + 'data_time': 0.0009713970066513866, + 'model_time': 1.3201733420137316, + 'grad_norm_pre_clip_avg': 0.17529628574848174, + 'learning_rate': 1.883911441376064e-05, + 'epoch': 5.02} +04/19 [18:35:25] INFO | >> train_qwenlatent.py:487 + Step 19900 | grad_norm_pre_clip=0.1766 | + grad_norm_pre_clip_avg=0.1726 | Metrics: + {'align_loss': 0.023780127987265587, + 'recon_loss': 0.07114837318658829, + 'predict_loss': 0.012200902216136456, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17660127580165863, + 'mae_score': 0.01110760972306535, 'data_time': + 0.0006959489837754518, 'model_time': + 1.2074046760099009, 'grad_norm_pre_clip_avg': + 0.17260099202394485, 'learning_rate': + 1.8831596629463035e-05, 'epoch': 5.02} +04/19 [18:35:38] INFO | >> train_qwenlatent.py:487 + Step 19910 | grad_norm_pre_clip=0.1887 | + grad_norm_pre_clip_avg=0.2504 | Metrics: + {'align_loss': 0.025616653263568878, + 'recon_loss': 0.0734294205904007, + 'predict_loss': 0.013630400411784649, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18867871165275574, + 'data_time': 0.0009526000067126006, + 'model_time': 1.242149014986353, + 'grad_norm_pre_clip_avg': 0.25039712339639664, + 'learning_rate': 1.88240757653145e-05, 'epoch': + 5.02} +04/19 [18:35:50] INFO | >> train_qwenlatent.py:487 + Step 19920 | grad_norm_pre_clip=0.2334 | + grad_norm_pre_clip_avg=0.1943 | Metrics: + {'align_loss': 0.024500586092472076, + 'recon_loss': 0.07077690958976746, + 'predict_loss': 0.0091854901984334, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23340477049350739, + 'data_time': 0.0006302820111159235, + 'model_time': 1.2674425700097345, + 'grad_norm_pre_clip_avg': 0.19431012868881226, + 'learning_rate': 1.8816551824980605e-05, + 'epoch': 5.03} +04/19 [18:36:03] INFO | >> train_qwenlatent.py:487 + Step 19930 | grad_norm_pre_clip=0.2004 | + grad_norm_pre_clip_avg=0.2152 | Metrics: + {'align_loss': 0.025112951174378395, + 'recon_loss': 0.06830010563135147, + 'predict_loss': 0.009066023863852024, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20037232339382172, + 'data_time': 0.0009138990135397762, + 'model_time': 1.1774856539850589, + 'grad_norm_pre_clip_avg': 0.21523659974336623, + 'learning_rate': 1.8809024812128435e-05, + 'epoch': 5.03} +04/19 [18:36:15] INFO | >> train_qwenlatent.py:487 + Step 19940 | grad_norm_pre_clip=0.2506 | + grad_norm_pre_clip_avg=0.2123 | Metrics: + {'align_loss': 0.024838270619511604, + 'recon_loss': 0.05060175433754921, + 'predict_loss': 0.010790202766656876, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25056028366088867, + 'data_time': 0.0007931079890113324, + 'model_time': 1.2479086380044464, + 'grad_norm_pre_clip_avg': 0.21226757019758224, + 'learning_rate': 1.8801494730426566e-05, + 'epoch': 5.03} +04/19 [18:36:28] INFO | >> train_qwenlatent.py:487 + Step 19950 | grad_norm_pre_clip=0.2181 | + grad_norm_pre_clip_avg=0.2103 | Metrics: + {'align_loss': 0.02520178072154522, + 'recon_loss': 0.0630742534995079, + 'predict_loss': 0.010061342269182205, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2180802822113037, + 'mae_score': 0.013144170056592236, 'data_time': + 0.0009641009964980185, 'model_time': + 1.2347131760034245, 'grad_norm_pre_clip_avg': + 0.21032968908548355, 'learning_rate': + 1.8793961583545063e-05, 'epoch': 5.03} +04/19 [18:36:41] INFO | >> train_qwenlatent.py:487 + Step 19960 | grad_norm_pre_clip=0.2482 | + grad_norm_pre_clip_avg=0.2218 | Metrics: + {'align_loss': 0.02440537139773369, + 'recon_loss': 0.07058029621839523, + 'predict_loss': 0.01096222922205925, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2482323795557022, + 'data_time': 0.0009471650118939579, + 'model_time': 1.4665831259917468, + 'grad_norm_pre_clip_avg': 0.2217508688569069, + 'learning_rate': 1.878642537515549e-05, + 'epoch': 5.04} +04/19 [18:36:54] INFO | >> train_qwenlatent.py:487 + Step 19970 | grad_norm_pre_clip=0.1835 | + grad_norm_pre_clip_avg=0.2140 | Metrics: + {'align_loss': 0.02594050206243992, + 'recon_loss': 0.06609296798706055, + 'predict_loss': 0.011042242869734764, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18351146578788757, + 'data_time': 0.0009166050003841519, + 'model_time': 1.267815266008256, + 'grad_norm_pre_clip_avg': 0.2139664813876152, + 'learning_rate': 1.8778886108930906e-05, + 'epoch': 5.04} +04/19 [18:37:06] INFO | >> train_qwenlatent.py:487 + Step 19980 | grad_norm_pre_clip=0.1663 | + grad_norm_pre_clip_avg=0.1842 | Metrics: + {'align_loss': 0.02674313262104988, + 'recon_loss': 0.09116332232952118, + 'predict_loss': 0.011504953727126122, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16628354787826538, + 'data_time': 0.0007847739907447249, + 'model_time': 1.2559697759861592, + 'grad_norm_pre_clip_avg': 0.18421975076198577, + 'learning_rate': 1.8771343788545863e-05, + 'epoch': 5.04} +04/19 [18:37:19] INFO | >> train_qwenlatent.py:487 + Step 19990 | grad_norm_pre_clip=0.2398 | + grad_norm_pre_clip_avg=0.2032 | Metrics: + {'align_loss': 0.024402759969234467, + 'recon_loss': 0.0768023207783699, + 'predict_loss': 0.010490535758435726, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23978647589683533, + 'data_time': 0.0006158720061648637, + 'model_time': 1.19721516102436, + 'grad_norm_pre_clip_avg': 0.20318690985441207, + 'learning_rate': 1.8763798417676384e-05, + 'epoch': 5.04} +04/19 [18:37:32] INFO | >> train_qwenlatent.py:487 + Step 20000 | grad_norm_pre_clip=0.3573 | + grad_norm_pre_clip_avg=0.2322 | Metrics: + {'align_loss': 0.0257277749478817, + 'recon_loss': 0.08187679946422577, + 'predict_loss': 0.01676229014992714, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3572854697704315, + 'mae_score': 0.011986892287795608, 'data_time': + 0.0008413869945798069, 'model_time': + 1.2241343650093768, 'grad_norm_pre_clip_avg': + 0.23219895511865615, 'learning_rate': + 1.8756250000000002e-05, 'epoch': 5.05} +✅ Checkpoint saved at ./runs/0418_QwenLatent_13tasks_actionstate_30k/checkpoints/steps_20000 +04/19 [18:37:54] INFO | >> train_qwenlatent.py:487 + Step 20010 | grad_norm_pre_clip=0.2015 | + grad_norm_pre_clip_avg=0.2520 | Metrics: + {'align_loss': 0.024491513147950172, + 'recon_loss': 0.06535372883081436, + 'predict_loss': 0.01013030856847763, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20147787034511566, + 'data_time': 0.0006290290039032698, + 'model_time': 1.2712139840004966, + 'grad_norm_pre_clip_avg': 0.2520397901535034, + 'learning_rate': 1.874869853919572e-05, + 'epoch': 5.05} +04/19 [18:38:08] INFO | >> train_qwenlatent.py:487 + Step 20020 | grad_norm_pre_clip=0.2212 | + grad_norm_pre_clip_avg=0.2238 | Metrics: + {'align_loss': 0.02585592493414879, + 'recon_loss': 0.06999499350786209, + 'predict_loss': 0.01292745303362608, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22121062874794006, + 'data_time': 0.0007120720110833645, + 'model_time': 1.6513768449949566, + 'grad_norm_pre_clip_avg': 0.22376535832881927, + 'learning_rate': 1.8741144038944027e-05, + 'epoch': 5.05} +04/19 [18:38:20] INFO | >> train_qwenlatent.py:487 + Step 20030 | grad_norm_pre_clip=0.2121 | + grad_norm_pre_clip_avg=0.1922 | Metrics: + {'align_loss': 0.02603374421596527, + 'recon_loss': 0.08770144730806351, + 'predict_loss': 0.013341148383915424, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2120923101902008, + 'data_time': 0.00083907600492239, 'model_time': + 1.2846980540198274, 'grad_norm_pre_clip_avg': + 0.1921529933810234, 'learning_rate': + 1.873358650292689e-05, 'epoch': 5.05} +04/19 [18:38:33] INFO | >> train_qwenlatent.py:487 + Step 20040 | grad_norm_pre_clip=0.3196 | + grad_norm_pre_clip_avg=0.2127 | Metrics: + {'align_loss': 0.02628752961754799, + 'recon_loss': 0.07504871487617493, + 'predict_loss': 0.013280081562697887, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.31963175535202026, + 'data_time': 0.0008181779994629323, + 'model_time': 1.20185799599858, + 'grad_norm_pre_clip_avg': 0.21272407174110414, + 'learning_rate': 1.8726025934827775e-05, + 'epoch': 5.06} +04/19 [18:38:46] INFO | >> train_qwenlatent.py:487 + Step 20050 | grad_norm_pre_clip=0.2022 | + grad_norm_pre_clip_avg=0.2103 | Metrics: + {'align_loss': 0.024394094944000244, + 'recon_loss': 0.0591435544192791, + 'predict_loss': 0.009474408812820911, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20220106840133667, + 'mae_score': 0.013391624485050236, 'data_time': + 0.001198982005007565, 'model_time': + 1.281010940001579, 'grad_norm_pre_clip_avg': + 0.21032339930534363, 'learning_rate': + 1.87184623383316e-05, 'epoch': 5.06} +04/19 [18:38:59] INFO | >> train_qwenlatent.py:487 + Step 20060 | grad_norm_pre_clip=0.1566 | + grad_norm_pre_clip_avg=0.1713 | Metrics: + {'align_loss': 0.02363676391541958, + 'recon_loss': 0.08260919898748398, + 'predict_loss': 0.010972956195473671, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1565983146429062, + 'data_time': 0.0007443880022037774, + 'model_time': 1.253458850987954, + 'grad_norm_pre_clip_avg': 0.17129371613264083, + 'learning_rate': 1.8710895717124768e-05, + 'epoch': 5.06} +04/19 [18:39:11] INFO | >> train_qwenlatent.py:487 + Step 20070 | grad_norm_pre_clip=0.2579 | + grad_norm_pre_clip_avg=0.2451 | Metrics: + {'align_loss': 0.02468157559633255, + 'recon_loss': 0.07046467065811157, + 'predict_loss': 0.00838184729218483, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2579175531864166, + 'data_time': 0.0006848470075055957, + 'model_time': 1.2230486249900423, + 'grad_norm_pre_clip_avg': 0.24507799744606018, + 'learning_rate': 1.8703326074895167e-05, + 'epoch': 5.06} +04/19 [18:39:24] INFO | >> train_qwenlatent.py:487 + Step 20080 | grad_norm_pre_clip=0.2263 | + grad_norm_pre_clip_avg=0.2059 | Metrics: + {'align_loss': 0.024757225066423416, + 'recon_loss': 0.09927511215209961, + 'predict_loss': 0.01545535959303379, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22627827525138855, + 'data_time': 0.000984504004009068, + 'model_time': 1.242186931980541, + 'grad_norm_pre_clip_avg': 0.20593907088041305, + 'learning_rate': 1.869575341533214e-05, + 'epoch': 5.07} +04/19 [18:39:36] INFO | >> train_qwenlatent.py:487 + Step 20090 | grad_norm_pre_clip=0.2095 | + grad_norm_pre_clip_avg=0.1915 | Metrics: + {'align_loss': 0.02422218769788742, + 'recon_loss': 0.059325192123651505, + 'predict_loss': 0.008667787536978722, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2095148116350174, + 'data_time': 0.0008588509808760136, + 'model_time': 1.3197158870170824, + 'grad_norm_pre_clip_avg': 0.1914794847369194, + 'learning_rate': 1.8688177742126516e-05, + 'epoch': 5.07} +04/19 [18:39:50] INFO | >> train_qwenlatent.py:487 + Step 20100 | grad_norm_pre_clip=0.1688 | + grad_norm_pre_clip_avg=0.1733 | Metrics: + {'align_loss': 0.025446077808737755, + 'recon_loss': 0.06843424588441849, + 'predict_loss': 0.009603074751794338, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16884486377239227, + 'mae_score': 0.018875555090002112, 'data_time': + 0.0006572970014531165, 'model_time': + 1.2117384290031623, 'grad_norm_pre_clip_avg': + 0.1732967048883438, 'learning_rate': + 1.8680599058970586e-05, 'epoch': 5.07} +04/19 [18:40:03] INFO | >> train_qwenlatent.py:487 + Step 20110 | grad_norm_pre_clip=0.2595 | + grad_norm_pre_clip_avg=0.2415 | Metrics: + {'align_loss': 0.02539818175137043, + 'recon_loss': 0.07805126160383224, + 'predict_loss': 0.010343820787966251, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25948959589004517, + 'data_time': 0.000617387006059289, + 'model_time': 1.1974052959994879, + 'grad_norm_pre_clip_avg': 0.24153346717357635, + 'learning_rate': 1.8673017369558102e-05, + 'epoch': 5.07} +04/19 [18:40:15] INFO | >> train_qwenlatent.py:487 + Step 20120 | grad_norm_pre_clip=0.1702 | + grad_norm_pre_clip_avg=0.1896 | Metrics: + {'align_loss': 0.02541082724928856, + 'recon_loss': 0.05815251171588898, + 'predict_loss': 0.00840611383318901, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17015786468982697, + 'data_time': 0.0007016750168986619, + 'model_time': 1.471931053994922, + 'grad_norm_pre_clip_avg': 0.18959613144397736, + 'learning_rate': 1.866543267758429e-05, + 'epoch': 5.08} +04/19 [18:40:28] INFO | >> train_qwenlatent.py:487 + Step 20130 | grad_norm_pre_clip=0.1670 | + grad_norm_pre_clip_avg=0.1791 | Metrics: + {'align_loss': 0.024450741708278656, + 'recon_loss': 0.06382839381694794, + 'predict_loss': 0.011537240818142891, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16702191531658173, + 'data_time': 0.000865976995555684, + 'model_time': 1.2569183019804768, + 'grad_norm_pre_clip_avg': 0.1791115716099739, + 'learning_rate': 1.865784498674584e-05, + 'epoch': 5.08} +04/19 [18:40:41] INFO | >> train_qwenlatent.py:487 + Step 20140 | grad_norm_pre_clip=0.2032 | + grad_norm_pre_clip_avg=0.2270 | Metrics: + {'align_loss': 0.024935659021139145, + 'recon_loss': 0.10003488510847092, + 'predict_loss': 0.010003648698329926, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20324350893497467, + 'data_time': 0.0009241270017810166, + 'model_time': 1.259391625993885, + 'grad_norm_pre_clip_avg': 0.22698410004377365, + 'learning_rate': 1.86502543007409e-05, 'epoch': + 5.08} +04/19 [18:40:54] INFO | >> train_qwenlatent.py:487 + Step 20150 | grad_norm_pre_clip=0.1911 | + grad_norm_pre_clip_avg=0.2001 | Metrics: + {'align_loss': 0.024415072053670883, + 'recon_loss': 0.057303536683321, + 'predict_loss': 0.012762162834405899, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19112184643745422, + 'mae_score': 0.011466442142520938, 'data_time': + 0.0008986929897218943, 'model_time': + 1.283374954975443, 'grad_norm_pre_clip_avg': + 0.2001143217086792, 'learning_rate': + 1.8642660623269073e-05, 'epoch': 5.08} +04/19 [18:41:07] INFO | >> train_qwenlatent.py:487 + Step 20160 | grad_norm_pre_clip=0.2466 | + grad_norm_pre_clip_avg=0.2018 | Metrics: + {'align_loss': 0.02489396370947361, + 'recon_loss': 0.08368600904941559, + 'predict_loss': 0.010422758758068085, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24662046134471893, + 'data_time': 0.0009471769735682756, + 'model_time': 1.2147272300207987, + 'grad_norm_pre_clip_avg': 0.20176184177398682, + 'learning_rate': 1.8635063958031427e-05, + 'epoch': 5.09} +04/19 [18:41:19] INFO | >> train_qwenlatent.py:487 + Step 20170 | grad_norm_pre_clip=0.1854 | + grad_norm_pre_clip_avg=0.2111 | Metrics: + {'align_loss': 0.024829495698213577, + 'recon_loss': 0.08031071722507477, + 'predict_loss': 0.011669325642287731, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18543097376823425, + 'data_time': 0.0006869680073577911, + 'model_time': 1.2329988049750682, + 'grad_norm_pre_clip_avg': 0.2111313298344612, + 'learning_rate': 1.862746430873049e-05, + 'epoch': 5.09} +04/19 [18:41:32] INFO | >> train_qwenlatent.py:487 + Step 20180 | grad_norm_pre_clip=0.1788 | + grad_norm_pre_clip_avg=0.2111 | Metrics: + {'align_loss': 0.024817120283842087, + 'recon_loss': 0.06604808568954468, + 'predict_loss': 0.006567960139364004, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1787751019001007, + 'data_time': 0.0008217340218834579, + 'model_time': 1.2218178029870614, + 'grad_norm_pre_clip_avg': 0.2110699862241745, + 'learning_rate': 1.861986167907023e-05, + 'epoch': 5.09} +04/19 [18:41:45] INFO | >> train_qwenlatent.py:487 + Step 20190 | grad_norm_pre_clip=0.1890 | + grad_norm_pre_clip_avg=0.2004 | Metrics: + {'align_loss': 0.02490120753645897, + 'recon_loss': 0.06004450470209122, + 'predict_loss': 0.011523162014782429, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1889818012714386, + 'data_time': 0.0009357150120195001, + 'model_time': 1.229264007997699, + 'grad_norm_pre_clip_avg': 0.2003963679075241, + 'learning_rate': 1.8612256072756086e-05, + 'epoch': 5.09} +04/19 [18:41:58] INFO | >> train_qwenlatent.py:487 + Step 20200 | grad_norm_pre_clip=0.1968 | + grad_norm_pre_clip_avg=0.2143 | Metrics: + {'align_loss': 0.026212053373456, 'recon_loss': + 0.09159743785858154, 'predict_loss': + 0.016402754932641983, 'aux_loss_decay_weight': + 0.0, 'grad_norm_pre_clip': 0.19683237373828888, + 'mae_score': 0.015236744579968152, 'data_time': + 0.0007526399858761579, 'model_time': + 1.2328832850034814, 'grad_norm_pre_clip_avg': + 0.21430035978555678, 'learning_rate': + 1.8604647493494924e-05, 'epoch': 5.1} +04/19 [18:42:10] INFO | >> train_qwenlatent.py:487 + Step 20210 | grad_norm_pre_clip=0.2593 | + grad_norm_pre_clip_avg=0.1994 | Metrics: + {'align_loss': 0.025959040969610214, + 'recon_loss': 0.07528995722532272, + 'predict_loss': 0.01055285707116127, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2592947781085968, + 'data_time': 0.001017383998259902, + 'model_time': 1.218072048999602, + 'grad_norm_pre_clip_avg': 0.1994386538863182, + 'learning_rate': 1.859703594499509e-05, + 'epoch': 5.1} +04/19 [18:42:23] INFO | >> train_qwenlatent.py:487 + Step 20220 | grad_norm_pre_clip=0.2118 | + grad_norm_pre_clip_avg=0.2066 | Metrics: + {'align_loss': 0.02605193667113781, + 'recon_loss': 0.076152503490448, + 'predict_loss': 0.009217428974807262, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21182210743427277, + 'data_time': 0.0009282600076403469, + 'model_time': 1.236607244994957, + 'grad_norm_pre_clip_avg': 0.2066292241215706, + 'learning_rate': 1.858942143096635e-05, + 'epoch': 5.1} +04/19 [18:42:36] INFO | >> train_qwenlatent.py:487 + Step 20230 | grad_norm_pre_clip=0.1839 | + grad_norm_pre_clip_avg=0.1907 | Metrics: + {'align_loss': 0.024345923215150833, + 'recon_loss': 0.04556457698345184, + 'predict_loss': 0.0072716944850981236, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18392401933670044, + 'data_time': 0.0009085839847102761, + 'model_time': 1.495548144011991, + 'grad_norm_pre_clip_avg': 0.1907489851117134, + 'learning_rate': 1.8581803955119925e-05, + 'epoch': 5.1} +04/19 [18:42:49] INFO | >> train_qwenlatent.py:487 + Step 20240 | grad_norm_pre_clip=0.1627 | + grad_norm_pre_clip_avg=0.2172 | Metrics: + {'align_loss': 0.02328978106379509, + 'recon_loss': 0.062079038470983505, + 'predict_loss': 0.009923851117491722, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1626662313938141, + 'data_time': 0.0006599600019399077, + 'model_time': 1.2273637739999685, + 'grad_norm_pre_clip_avg': 0.2172296553850174, + 'learning_rate': 1.8574183521168482e-05, + 'epoch': 5.11} +04/19 [18:43:02] INFO | >> train_qwenlatent.py:487 + Step 20250 | grad_norm_pre_clip=0.2027 | + grad_norm_pre_clip_avg=0.1884 | Metrics: + {'align_loss': 0.024825554341077805, + 'recon_loss': 0.05691405013203621, + 'predict_loss': 0.008706225082278252, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20273958146572113, + 'mae_score': 0.01180880950377868, 'data_time': + 0.0010587839933577925, 'model_time': + 1.5614367100060917, 'grad_norm_pre_clip_avg': + 0.18837084919214248, 'learning_rate': + 1.8566560132826137e-05, 'epoch': 5.11} +04/19 [18:43:15] INFO | >> train_qwenlatent.py:487 + Step 20260 | grad_norm_pre_clip=0.2739 | + grad_norm_pre_clip_avg=0.2257 | Metrics: + {'align_loss': 0.024317147210240364, + 'recon_loss': 0.06439175456762314, + 'predict_loss': 0.010767857544124126, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.27393338084220886, + 'data_time': 0.0006458209827542305, + 'model_time': 1.2382200450228993, + 'grad_norm_pre_clip_avg': 0.22568644285202027, + 'learning_rate': 1.8558933793808422e-05, + 'epoch': 5.11} +04/19 [18:43:27] INFO | >> train_qwenlatent.py:487 + Step 20270 | grad_norm_pre_clip=0.1805 | + grad_norm_pre_clip_avg=0.1991 | Metrics: + {'align_loss': 0.02542070485651493, + 'recon_loss': 0.08911217004060745, + 'predict_loss': 0.01193995401263237, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1805483102798462, + 'data_time': 0.0011349609994795173, + 'model_time': 1.2064832409960218, + 'grad_norm_pre_clip_avg': 0.19913989007472993, + 'learning_rate': 1.8551304507832333e-05, + 'epoch': 5.11} +04/19 [18:43:40] INFO | >> train_qwenlatent.py:487 + Step 20280 | grad_norm_pre_clip=0.1464 | + grad_norm_pre_clip_avg=0.1623 | Metrics: + {'align_loss': 0.0260700061917305, + 'recon_loss': 0.07997485995292664, + 'predict_loss': 0.008056762628257275, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14637823402881622, + 'data_time': 0.0009276750206481665, + 'model_time': 1.234650606987998, + 'grad_norm_pre_clip_avg': 0.16230179369449615, + 'learning_rate': 1.8543672278616282e-05, + 'epoch': 5.12} +04/19 [18:43:52] INFO | >> train_qwenlatent.py:487 + Step 20290 | grad_norm_pre_clip=0.2024 | + grad_norm_pre_clip_avg=0.2305 | Metrics: + {'align_loss': 0.026208939030766487, + 'recon_loss': 0.08399583399295807, + 'predict_loss': 0.010110463947057724, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20238974690437317, + 'data_time': 0.0011301860213279724, + 'model_time': 1.2563084260036703, + 'grad_norm_pre_clip_avg': 0.23048528879880906, + 'learning_rate': 1.8536037109880134e-05, + 'epoch': 5.12} +04/19 [18:44:06] INFO | >> train_qwenlatent.py:487 + Step 20300 | grad_norm_pre_clip=0.1650 | + grad_norm_pre_clip_avg=0.1882 | Metrics: + {'align_loss': 0.025553328916430473, + 'recon_loss': 0.07933187484741211, + 'predict_loss': 0.010595246218144894, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16500133275985718, + 'mae_score': 0.013036169447340406, 'data_time': + 0.0010798870061989874, 'model_time': + 1.2427052550192457, 'grad_norm_pre_clip_avg': + 0.188188636302948, 'learning_rate': + 1.8528399005345172e-05, 'epoch': 5.12} +04/19 [18:44:19] INFO | >> train_qwenlatent.py:487 + Step 20310 | grad_norm_pre_clip=0.1488 | + grad_norm_pre_clip_avg=0.2117 | Metrics: + {'align_loss': 0.025980304926633835, + 'recon_loss': 0.07337713986635208, + 'predict_loss': 0.012402989901602268, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14884847402572632, + 'data_time': 0.0005859899974893779, + 'model_time': 1.2349148509965744, + 'grad_norm_pre_clip_avg': 0.21166765093803405, + 'learning_rate': 1.852075796873412e-05, + 'epoch': 5.12} +04/19 [18:44:31] INFO | >> train_qwenlatent.py:487 + Step 20320 | grad_norm_pre_clip=0.1912 | + grad_norm_pre_clip_avg=0.2236 | Metrics: + {'align_loss': 0.024831565096974373, + 'recon_loss': 0.0779651403427124, + 'predict_loss': 0.010756159201264381, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1912136822938919, + 'data_time': 0.000704970007063821, + 'model_time': 1.2324400040088221, + 'grad_norm_pre_clip_avg': 0.22361139953136444, + 'learning_rate': 1.8513114003771115e-05, + 'epoch': 5.13} +04/19 [18:44:44] INFO | >> train_qwenlatent.py:487 + Step 20330 | grad_norm_pre_clip=0.2120 | + grad_norm_pre_clip_avg=0.2071 | Metrics: + {'align_loss': 0.024341458454728127, + 'recon_loss': 0.0532623790204525, + 'predict_loss': 0.005777006037533283, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2120184451341629, + 'data_time': 0.0010269169870298356, + 'model_time': 1.2399082150077447, + 'grad_norm_pre_clip_avg': 0.20705875307321547, + 'learning_rate': 1.8505467114181746e-05, + 'epoch': 5.13} +04/19 [18:44:56] INFO | >> train_qwenlatent.py:487 + Step 20340 | grad_norm_pre_clip=0.2088 | + grad_norm_pre_clip_avg=0.2047 | Metrics: + {'align_loss': 0.02272775024175644, + 'recon_loss': 0.06462759524583817, + 'predict_loss': 0.019223563373088837, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20882613956928253, + 'data_time': 0.0006612610013689846, + 'model_time': 1.26766970398603, + 'grad_norm_pre_clip_avg': 0.20466338098049164, + 'learning_rate': 1.849781730369301e-05, + 'epoch': 5.13} +04/19 [18:45:09] INFO | >> train_qwenlatent.py:487 + Step 20350 | grad_norm_pre_clip=0.3487 | + grad_norm_pre_clip_avg=0.2352 | Metrics: + {'align_loss': 0.024610519409179688, + 'recon_loss': 0.07002539932727814, + 'predict_loss': 0.012769084423780441, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.34872838854789734, + 'mae_score': 0.00961476403313714, 'data_time': + 0.0009245620167348534, 'model_time': + 1.313078071019845, 'grad_norm_pre_clip_avg': + 0.23517441004514694, 'learning_rate': + 1.849016457603333e-05, 'epoch': 5.13} +04/19 [18:45:22] INFO | >> train_qwenlatent.py:487 + Step 20360 | grad_norm_pre_clip=0.1670 | + grad_norm_pre_clip_avg=0.2270 | Metrics: + {'align_loss': 0.024735011160373688, + 'recon_loss': 0.06801643967628479, + 'predict_loss': 0.013995942659676075, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1669594794511795, + 'data_time': 0.0011008110013790429, + 'model_time': 1.5041310859960504, + 'grad_norm_pre_clip_avg': 0.22695043683052063, + 'learning_rate': 1.848250893493255e-05, + 'epoch': 5.14} +04/19 [18:45:35] INFO | >> train_qwenlatent.py:487 + Step 20370 | grad_norm_pre_clip=0.2057 | + grad_norm_pre_clip_avg=0.2038 | Metrics: + {'align_loss': 0.025475304573774338, + 'recon_loss': 0.07352042198181152, + 'predict_loss': 0.010768457315862179, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2057410627603531, + 'data_time': 0.001042637974023819, + 'model_time': 1.2558641739888117, + 'grad_norm_pre_clip_avg': 0.20378280580043792, + 'learning_rate': 1.8474850384121942e-05, + 'epoch': 5.14} +04/19 [18:45:48] INFO | >> train_qwenlatent.py:487 + Step 20380 | grad_norm_pre_clip=0.1867 | + grad_norm_pre_clip_avg=0.2119 | Metrics: + {'align_loss': 0.024630172178149223, + 'recon_loss': 0.07162056863307953, + 'predict_loss': 0.016070595011115074, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1866914927959442, + 'data_time': 0.0006391420029103756, + 'model_time': 1.5781083659967408, + 'grad_norm_pre_clip_avg': 0.21186788827180864, + 'learning_rate': 1.8467188927334187e-05, + 'epoch': 5.14} +04/19 [18:46:01] INFO | >> train_qwenlatent.py:487 + Step 20390 | grad_norm_pre_clip=0.2676 | + grad_norm_pre_clip_avg=0.2286 | Metrics: + {'align_loss': 0.025953343138098717, + 'recon_loss': 0.07566680014133453, + 'predict_loss': 0.00865489337593317, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2676328122615814, + 'data_time': 0.0010278600093442947, + 'model_time': 1.298039729008451, + 'grad_norm_pre_clip_avg': 0.22855774760246278, + 'learning_rate': 1.845952456830339e-05, + 'epoch': 5.15} +04/19 [18:46:14] INFO | >> train_qwenlatent.py:487 + Step 20400 | grad_norm_pre_clip=0.1315 | + grad_norm_pre_clip_avg=0.1824 | Metrics: + {'align_loss': 0.02630426734685898, + 'recon_loss': 0.07405872642993927, + 'predict_loss': 0.007105449680238962, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13145385682582855, + 'mae_score': 0.0121083757898829, 'data_time': + 0.0013722029980272055, 'model_time': + 1.2461301279836334, 'grad_norm_pre_clip_avg': + 0.18244085907936097, 'learning_rate': + 1.8451857310765063e-05, 'epoch': 5.15} +04/19 [18:46:27] INFO | >> train_qwenlatent.py:487 + Step 20410 | grad_norm_pre_clip=0.1806 | + grad_norm_pre_clip_avg=0.1908 | Metrics: + {'align_loss': 0.026580268517136574, + 'recon_loss': 0.07183428108692169, + 'predict_loss': 0.009113490581512451, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18064725399017334, + 'data_time': 0.0009506589849479496, + 'model_time': 1.3305548019998241, + 'grad_norm_pre_clip_avg': 0.19075242578983306, + 'learning_rate': 1.8444187158456134e-05, + 'epoch': 5.15} +04/19 [18:46:39] INFO | >> train_qwenlatent.py:487 + Step 20420 | grad_norm_pre_clip=0.2003 | + grad_norm_pre_clip_avg=0.1994 | Metrics: + {'align_loss': 0.025177260860800743, + 'recon_loss': 0.08407342433929443, + 'predict_loss': 0.01412903144955635, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20034655928611755, + 'data_time': 0.0008928340103011578, + 'model_time': 1.250080544006778, + 'grad_norm_pre_clip_avg': 0.1994067296385765, + 'learning_rate': 1.8436514115114942e-05, + 'epoch': 5.15} +04/19 [18:46:51] INFO | >> train_qwenlatent.py:487 + Step 20430 | grad_norm_pre_clip=0.2028 | + grad_norm_pre_clip_avg=0.2098 | Metrics: + {'align_loss': 0.025718342512845993, + 'recon_loss': 0.07262928038835526, + 'predict_loss': 0.013199865818023682, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2028461992740631, + 'data_time': 0.0008793870219960809, + 'model_time': 1.2528395089902915, + 'grad_norm_pre_clip_avg': 0.20975296199321747, + 'learning_rate': 1.842883818448124e-05, + 'epoch': 5.16} +04/19 [18:47:04] INFO | >> train_qwenlatent.py:487 + Step 20440 | grad_norm_pre_clip=0.1945 | + grad_norm_pre_clip_avg=0.1980 | Metrics: + {'align_loss': 0.02508646994829178, + 'recon_loss': 0.07005728036165237, + 'predict_loss': 0.013150665909051895, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1944592297077179, + 'data_time': 0.0010134140029549599, + 'model_time': 1.2515034800162539, + 'grad_norm_pre_clip_avg': 0.19799975007772447, + 'learning_rate': 1.8421159370296177e-05, + 'epoch': 5.16} +04/19 [18:47:18] INFO | >> train_qwenlatent.py:487 + Step 20450 | grad_norm_pre_clip=0.1687 | + grad_norm_pre_clip_avg=0.2316 | Metrics: + {'align_loss': 0.024282418191432953, + 'recon_loss': 0.06604475528001785, + 'predict_loss': 0.011060616932809353, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16866986453533173, + 'mae_score': 0.012124510069151183, 'data_time': + 0.0009088339866138995, 'model_time': + 1.265660446981201, 'grad_norm_pre_clip_avg': + 0.23161278963088988, 'learning_rate': + 1.8413477676302316e-05, 'epoch': 5.16} +04/19 [18:47:30] INFO | >> train_qwenlatent.py:487 + Step 20460 | grad_norm_pre_clip=0.2074 | + grad_norm_pre_clip_avg=0.2166 | Metrics: + {'align_loss': 0.025586530566215515, + 'recon_loss': 0.08846394717693329, + 'predict_loss': 0.012399996630847454, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20744116604328156, + 'data_time': 0.0006524109921883792, + 'model_time': 1.2671913389931433, + 'grad_norm_pre_clip_avg': 0.2166140243411064, + 'learning_rate': 1.8405793106243617e-05, + 'epoch': 5.16} +04/19 [18:47:43] INFO | >> train_qwenlatent.py:487 + Step 20470 | grad_norm_pre_clip=0.1840 | + grad_norm_pre_clip_avg=0.1849 | Metrics: + {'align_loss': 0.025457199662923813, + 'recon_loss': 0.07287883758544922, + 'predict_loss': 0.011237051337957382, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18402564525604248, + 'data_time': 0.0010855769796762615, + 'model_time': 1.2678159729985055, + 'grad_norm_pre_clip_avg': 0.1848752424120903, + 'learning_rate': 1.8398105663865453e-05, + 'epoch': 5.17} +04/19 [18:47:55] INFO | >> train_qwenlatent.py:487 + Step 20480 | grad_norm_pre_clip=0.1942 | + grad_norm_pre_clip_avg=0.1912 | Metrics: + {'align_loss': 0.02401091530919075, + 'recon_loss': 0.0757078155875206, + 'predict_loss': 0.016109995543956757, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19424399733543396, + 'data_time': 0.0006615069869440049, + 'model_time': 1.234944304014789, + 'grad_norm_pre_clip_avg': 0.1912287876009941, + 'learning_rate': 1.8390415352914594e-05, + 'epoch': 5.17} +04/19 [18:48:08] INFO | >> train_qwenlatent.py:487 + Step 20490 | grad_norm_pre_clip=0.1577 | + grad_norm_pre_clip_avg=0.1890 | Metrics: + {'align_loss': 0.025388825684785843, + 'recon_loss': 0.051113247871398926, + 'predict_loss': 0.009652837179601192, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15768221020698547, + 'data_time': 0.0006659979990217835, + 'model_time': 1.2208685160148889, + 'grad_norm_pre_clip_avg': 0.18901772052049637, + 'learning_rate': 1.838272217713919e-05, + 'epoch': 5.17} +04/19 [18:48:21] INFO | >> train_qwenlatent.py:487 + Step 20500 | grad_norm_pre_clip=0.1546 | + grad_norm_pre_clip_avg=0.1715 | Metrics: + {'align_loss': 0.025729399174451828, + 'recon_loss': 0.08201322704553604, + 'predict_loss': 0.008447026833891869, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15456196665763855, + 'mae_score': 0.015989616325309684, 'data_time': + 0.0007501160143874586, 'model_time': + 1.2010439029836562, 'grad_norm_pre_clip_avg': + 0.17150282561779023, 'learning_rate': + 1.8375026140288813e-05, 'epoch': 5.17} +04/19 [18:48:34] INFO | >> train_qwenlatent.py:487 + Step 20510 | grad_norm_pre_clip=0.1607 | + grad_norm_pre_clip_avg=0.2254 | Metrics: + {'align_loss': 0.026797253638505936, + 'recon_loss': 0.08433743566274643, + 'predict_loss': 0.007438939996063709, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16065850853919983, + 'data_time': 0.0006863369781058282, + 'model_time': 1.2356832049845252, + 'grad_norm_pre_clip_avg': 0.22537892162799836, + 'learning_rate': 1.836732724611441e-05, + 'epoch': 5.18} +04/19 [18:48:47] INFO | >> train_qwenlatent.py:487 + Step 20520 | grad_norm_pre_clip=0.1996 | + grad_norm_pre_clip_avg=0.2111 | Metrics: + {'align_loss': 0.024849755689501762, + 'recon_loss': 0.08344750106334686, + 'predict_loss': 0.015411323867738247, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19959942996501923, + 'data_time': 0.0009411000064574182, + 'model_time': 1.2807745570025872, + 'grad_norm_pre_clip_avg': 0.21106113195419313, + 'learning_rate': 1.8359625498368336e-05, + 'epoch': 5.18} +04/19 [18:49:00] INFO | >> train_qwenlatent.py:487 + Step 20530 | grad_norm_pre_clip=0.1646 | + grad_norm_pre_clip_avg=0.1868 | Metrics: + {'align_loss': 0.025140205398201942, + 'recon_loss': 0.07343416661024094, + 'predict_loss': 0.007743930909782648, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1645955741405487, + 'data_time': 0.0009428939956706017, + 'model_time': 1.3003628570004366, + 'grad_norm_pre_clip_avg': 0.18680842444300652, + 'learning_rate': 1.8351920900804328e-05, + 'epoch': 5.18} +04/19 [18:49:12] INFO | >> train_qwenlatent.py:487 + Step 20540 | grad_norm_pre_clip=0.1602 | + grad_norm_pre_clip_avg=0.1906 | Metrics: + {'align_loss': 0.025593772530555725, + 'recon_loss': 0.08873925358057022, + 'predict_loss': 0.016156664118170738, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16022685170173645, + 'data_time': 0.001028405997203663, + 'model_time': 1.25228365600924, + 'grad_norm_pre_clip_avg': 0.1906183660030365, + 'learning_rate': 1.8344213457177507e-05, + 'epoch': 5.18} +04/19 [18:49:25] INFO | >> train_qwenlatent.py:487 + Step 20550 | grad_norm_pre_clip=0.2015 | + grad_norm_pre_clip_avg=0.2596 | Metrics: + {'align_loss': 0.025721784681081772, + 'recon_loss': 0.09201101213693619, + 'predict_loss': 0.014784911647439003, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20148958265781403, + 'mae_score': 0.017169696361094982, 'data_time': + 0.0009477789863012731, 'model_time': + 1.2790895679790992, 'grad_norm_pre_clip_avg': + 0.25962043553590775, 'learning_rate': + 1.833650317124439e-05, 'epoch': 5.19} +04/19 [18:49:38] INFO | >> train_qwenlatent.py:487 + Step 20560 | grad_norm_pre_clip=0.2176 | + grad_norm_pre_clip_avg=0.2184 | Metrics: + {'align_loss': 0.02530769631266594, + 'recon_loss': 0.06772752851247787, + 'predict_loss': 0.011270738206803799, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21758444607257843, + 'data_time': 0.0012586789962369949, + 'model_time': 1.2930734329856932, + 'grad_norm_pre_clip_avg': 0.2183851420879364, + 'learning_rate': 1.8328790046762882e-05, + 'epoch': 5.19} +04/19 [18:49:51] INFO | >> train_qwenlatent.py:487 + Step 20570 | grad_norm_pre_clip=0.1878 | + grad_norm_pre_clip_avg=0.1965 | Metrics: + {'align_loss': 0.027144547551870346, + 'recon_loss': 0.0847969502210617, + 'predict_loss': 0.011908318847417831, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18782106041908264, + 'data_time': 0.000670556997647509, + 'model_time': 1.4831691080180462, + 'grad_norm_pre_clip_avg': 0.1964644357562065, + 'learning_rate': 1.832107408749226e-05, + 'epoch': 5.19} +04/19 [18:50:03] INFO | >> train_qwenlatent.py:487 + Step 20580 | grad_norm_pre_clip=0.1902 | + grad_norm_pre_clip_avg=0.2178 | Metrics: + {'align_loss': 0.024701401591300964, + 'recon_loss': 0.05771980807185173, + 'predict_loss': 0.00620405375957489, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19017435610294342, + 'data_time': 0.0011003869876731187, + 'model_time': 1.2290356250014156, + 'grad_norm_pre_clip_avg': 0.2178390383720398, + 'learning_rate': 1.8313355297193195e-05, + 'epoch': 5.19} +04/19 [18:50:16] INFO | >> train_qwenlatent.py:487 + Step 20590 | grad_norm_pre_clip=0.1719 | + grad_norm_pre_clip_avg=0.1792 | Metrics: + {'align_loss': 0.023796342313289642, + 'recon_loss': 0.08533141762018204, + 'predict_loss': 0.017261305823922157, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1718587577342987, + 'data_time': 0.000923782994505018, + 'model_time': 1.5792869279976003, + 'grad_norm_pre_clip_avg': 0.17915095388889313, + 'learning_rate': 1.8305633679627724e-05, + 'epoch': 5.2} +04/19 [18:50:29] INFO | >> train_qwenlatent.py:487 + Step 20600 | grad_norm_pre_clip=0.2060 | + grad_norm_pre_clip_avg=0.1734 | Metrics: + {'align_loss': 0.025327954441308975, + 'recon_loss': 0.06478002667427063, + 'predict_loss': 0.014076012186706066, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.205964133143425, + 'mae_score': 0.01041374378376179, 'data_time': + 0.0006693850154988468, 'model_time': + 1.22305970400339, 'grad_norm_pre_clip_avg': + 0.17339793592691422, 'learning_rate': + 1.8297909238559275e-05, 'epoch': 5.2} +04/19 [18:50:42] INFO | >> train_qwenlatent.py:487 + Step 20610 | grad_norm_pre_clip=0.2471 | + grad_norm_pre_clip_avg=0.2157 | Metrics: + {'align_loss': 0.02561076357960701, + 'recon_loss': 0.09502283483743668, + 'predict_loss': 0.015765050426125526, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24705466628074646, + 'data_time': 0.0006266880081966519, + 'model_time': 1.2238581370038446, + 'grad_norm_pre_clip_avg': 0.21565979421138765, + 'learning_rate': 1.8290181977752643e-05, + 'epoch': 5.2} +04/19 [18:50:55] INFO | >> train_qwenlatent.py:487 + Step 20620 | grad_norm_pre_clip=0.2091 | + grad_norm_pre_clip_avg=0.2470 | Metrics: + {'align_loss': 0.02453126758337021, + 'recon_loss': 0.07539261877536774, + 'predict_loss': 0.010166230611503124, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20912663638591766, + 'data_time': 0.0009862319857347757, + 'model_time': 1.3253918459813576, + 'grad_norm_pre_clip_avg': 0.24704124927520751, + 'learning_rate': 1.8282451900974007e-05, + 'epoch': 5.2} +04/19 [18:51:07] INFO | >> train_qwenlatent.py:487 + Step 20630 | grad_norm_pre_clip=0.1316 | + grad_norm_pre_clip_avg=0.1897 | Metrics: + {'align_loss': 0.025161491706967354, + 'recon_loss': 0.05169522389769554, + 'predict_loss': 0.0065573109313845634, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13157948851585388, + 'data_time': 0.0007340660085901618, + 'model_time': 1.205361697007902, + 'grad_norm_pre_clip_avg': 0.18970037400722503, + 'learning_rate': 1.8274719011990905e-05, + 'epoch': 5.21} +04/19 [18:51:21] INFO | >> train_qwenlatent.py:487 + Step 20640 | grad_norm_pre_clip=0.2125 | + grad_norm_pre_clip_avg=0.1853 | Metrics: + {'align_loss': 0.0239254143089056, + 'recon_loss': 0.07481943815946579, + 'predict_loss': 0.01405650656670332, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21249866485595703, + 'data_time': 0.0008911149925552309, + 'model_time': 1.5496169729740359, + 'grad_norm_pre_clip_avg': 0.18527032434940338, + 'learning_rate': 1.8266983314572258e-05, + 'epoch': 5.21} +04/19 [18:51:34] INFO | >> train_qwenlatent.py:487 + Step 20650 | grad_norm_pre_clip=0.1614 | + grad_norm_pre_clip_avg=0.2307 | Metrics: + {'align_loss': 0.02536877989768982, + 'recon_loss': 0.06460738927125931, + 'predict_loss': 0.010043773800134659, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1614082008600235, + 'mae_score': 0.011399940971855645, 'data_time': + 0.0007765009941067547, 'model_time': + 1.2748357089876663, 'grad_norm_pre_clip_avg': + 0.23066886216402055, 'learning_rate': + 1.825924481248835e-05, 'epoch': 5.21} +04/19 [18:51:47] INFO | >> train_qwenlatent.py:487 + Step 20660 | grad_norm_pre_clip=0.1954 | + grad_norm_pre_clip_avg=0.1849 | Metrics: + {'align_loss': 0.024946879595518112, + 'recon_loss': 0.07332050055265427, + 'predict_loss': 0.012377905659377575, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1953856199979782, + 'data_time': 0.0013327950146049261, + 'model_time': 1.2490174630074762, + 'grad_norm_pre_clip_avg': 0.18486586809158326, + 'learning_rate': 1.8251503509510834e-05, + 'epoch': 5.21} +04/19 [18:51:59] INFO | >> train_qwenlatent.py:487 + Step 20670 | grad_norm_pre_clip=0.1847 | + grad_norm_pre_clip_avg=0.1797 | Metrics: + {'align_loss': 0.026242472231388092, + 'recon_loss': 0.08099324256181717, + 'predict_loss': 0.008922039531171322, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18473416566848755, + 'data_time': 0.0008914170030038804, + 'model_time': 1.228200448997086, + 'grad_norm_pre_clip_avg': 0.17970926761627198, + 'learning_rate': 1.8243759409412722e-05, + 'epoch': 5.22} +04/19 [18:52:12] INFO | >> train_qwenlatent.py:487 + Step 20680 | grad_norm_pre_clip=0.1857 | + grad_norm_pre_clip_avg=0.2294 | Metrics: + {'align_loss': 0.025501029565930367, + 'recon_loss': 0.09417412430047989, + 'predict_loss': 0.021020734682679176, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1857437938451767, + 'data_time': 0.0006600930064450949, + 'model_time': 1.222430074994918, + 'grad_norm_pre_clip_avg': 0.22940822392702104, + 'learning_rate': 1.8236012515968398e-05, + 'epoch': 5.22} +04/19 [18:52:24] INFO | >> train_qwenlatent.py:487 + Step 20690 | grad_norm_pre_clip=0.1630 | + grad_norm_pre_clip_avg=0.2231 | Metrics: + {'align_loss': 0.025704624131321907, + 'recon_loss': 0.08683033287525177, + 'predict_loss': 0.00898243673145771, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16300298273563385, + 'data_time': 0.0008676940051373094, + 'model_time': 1.2140281119791325, + 'grad_norm_pre_clip_avg': 0.2231445163488388, + 'learning_rate': 1.8228262832953605e-05, + 'epoch': 5.22} +04/19 [18:52:37] INFO | >> train_qwenlatent.py:487 + Step 20700 | grad_norm_pre_clip=0.1820 | + grad_norm_pre_clip_avg=0.2055 | Metrics: + {'align_loss': 0.02395150437951088, + 'recon_loss': 0.042828042060136795, + 'predict_loss': 0.007922874763607979, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18204306066036224, + 'mae_score': 0.01071758270263672, 'data_time': + 0.000877259997650981, 'model_time': + 1.179921712988289, 'grad_norm_pre_clip_avg': + 0.20548777878284455, 'learning_rate': + 1.8220510364145448e-05, 'epoch': 5.22} +04/19 [18:52:50] INFO | >> train_qwenlatent.py:487 + Step 20710 | grad_norm_pre_clip=0.2261 | + grad_norm_pre_clip_avg=0.1897 | Metrics: + {'align_loss': 0.02517329528927803, + 'recon_loss': 0.06615990400314331, + 'predict_loss': 0.010770591907203197, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2261439710855484, + 'data_time': 0.0008185550104826689, + 'model_time': 1.1876597979862709, + 'grad_norm_pre_clip_avg': 0.18971929103136062, + 'learning_rate': 1.8212755113322376e-05, + 'epoch': 5.23} +04/19 [18:53:03] INFO | >> train_qwenlatent.py:487 + Step 20720 | grad_norm_pre_clip=0.1559 | + grad_norm_pre_clip_avg=0.1919 | Metrics: + {'align_loss': 0.02453598752617836, + 'recon_loss': 0.07946817576885223, + 'predict_loss': 0.009815042838454247, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15591584146022797, + 'data_time': 0.001039769995259121, + 'model_time': 1.2867730019788723, + 'grad_norm_pre_clip_avg': 0.1918908938765526, + 'learning_rate': 1.8204997084264208e-05, + 'epoch': 5.23} +04/19 [18:53:16] INFO | >> train_qwenlatent.py:487 + Step 20730 | grad_norm_pre_clip=0.1877 | + grad_norm_pre_clip_avg=0.1855 | Metrics: + {'align_loss': 0.025699453428387642, + 'recon_loss': 0.07832098752260208, + 'predict_loss': 0.011453314684331417, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18765483796596527, + 'data_time': 0.0009011410002131015, + 'model_time': 1.2106480989896227, + 'grad_norm_pre_clip_avg': 0.18549267947673798, + 'learning_rate': 1.8197236280752114e-05, + 'epoch': 5.23} +04/19 [18:53:28] INFO | >> train_qwenlatent.py:487 + Step 20740 | grad_norm_pre_clip=0.2225 | + grad_norm_pre_clip_avg=0.2066 | Metrics: + {'align_loss': 0.025040440261363983, + 'recon_loss': 0.06638820469379425, + 'predict_loss': 0.00815529003739357, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.222454234957695, + 'data_time': 0.0007303830061573535, + 'model_time': 1.210807660012506, + 'grad_norm_pre_clip_avg': 0.20658636540174485, + 'learning_rate': 1.818947270656862e-05, + 'epoch': 5.23} +04/19 [18:53:41] INFO | >> train_qwenlatent.py:487 + Step 20750 | grad_norm_pre_clip=0.1871 | + grad_norm_pre_clip_avg=0.1913 | Metrics: + {'align_loss': 0.02488134801387787, + 'recon_loss': 0.07423452287912369, + 'predict_loss': 0.012377100065350533, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18713103234767914, + 'mae_score': 0.009026221541671065, 'data_time': + 0.0006199589988682419, 'model_time': + 1.2597096269892063, 'grad_norm_pre_clip_avg': + 0.19127240478992463, 'learning_rate': + 1.8181706365497592e-05, 'epoch': 5.24} +04/19 [18:53:54] INFO | >> train_qwenlatent.py:487 + Step 20760 | grad_norm_pre_clip=0.2290 | + grad_norm_pre_clip_avg=0.1982 | Metrics: + {'align_loss': 0.02559397928416729, + 'recon_loss': 0.07932958006858826, + 'predict_loss': 0.01007027830928564, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22903646528720856, + 'data_time': 0.0006926640053279698, + 'model_time': 1.2197180689836387, + 'grad_norm_pre_clip_avg': 0.19821081161499024, + 'learning_rate': 1.817393726132425e-05, + 'epoch': 5.24} +04/19 [18:54:07] INFO | >> train_qwenlatent.py:487 + Step 20770 | grad_norm_pre_clip=0.1375 | + grad_norm_pre_clip_avg=0.1671 | Metrics: + {'align_loss': 0.024512745440006256, + 'recon_loss': 0.06296605616807938, + 'predict_loss': 0.006919649429619312, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1375001072883606, + 'data_time': 0.0009792670025490224, + 'model_time': 1.5064861210121308, + 'grad_norm_pre_clip_avg': 0.16706182807683945, + 'learning_rate': 1.8166165397835166e-05, + 'epoch': 5.24} +04/19 [18:54:20] INFO | >> train_qwenlatent.py:487 + Step 20780 | grad_norm_pre_clip=0.2091 | + grad_norm_pre_clip_avg=0.2265 | Metrics: + {'align_loss': 0.023852795362472534, + 'recon_loss': 0.06644055992364883, + 'predict_loss': 0.013774593360722065, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20912881195545197, + 'data_time': 0.0006787980091758072, + 'model_time': 1.2205077190010343, + 'grad_norm_pre_clip_avg': 0.22651588171720505, + 'learning_rate': 1.8158390778818245e-05, + 'epoch': 5.24} +04/19 [18:54:33] INFO | >> train_qwenlatent.py:487 + Step 20790 | grad_norm_pre_clip=0.2375 | + grad_norm_pre_clip_avg=0.2463 | Metrics: + {'align_loss': 0.0263788141310215, + 'recon_loss': 0.07099911570549011, + 'predict_loss': 0.010424350388348103, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23752866685390472, + 'data_time': 0.0009093540138565004, + 'model_time': 1.2304080729954876, + 'grad_norm_pre_clip_avg': 0.24627436101436614, + 'learning_rate': 1.815061340806275e-05, + 'epoch': 5.25} +04/19 [18:54:46] INFO | >> train_qwenlatent.py:487 + Step 20800 | grad_norm_pre_clip=0.1654 | + grad_norm_pre_clip_avg=0.2172 | Metrics: + {'align_loss': 0.024413790553808212, + 'recon_loss': 0.05703931674361229, + 'predict_loss': 0.010088951326906681, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16538038849830627, + 'mae_score': 0.011323631561554231, 'data_time': + 0.000743194977985695, 'model_time': + 1.237292603997048, 'grad_norm_pre_clip_avg': + 0.2171909749507904, 'learning_rate': + 1.8142833289359267e-05, 'epoch': 5.25} +04/19 [18:54:58] INFO | >> train_qwenlatent.py:487 + Step 20810 | grad_norm_pre_clip=0.1841 | + grad_norm_pre_clip_avg=0.2103 | Metrics: + {'align_loss': 0.02401123195886612, + 'recon_loss': 0.050089623779058456, + 'predict_loss': 0.005211631767451763, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18412873148918152, + 'data_time': 0.0007819180027581751, + 'model_time': 1.2702575270086527, + 'grad_norm_pre_clip_avg': 0.21033646911382675, + 'learning_rate': 1.8135050426499742e-05, + 'epoch': 5.25} +04/19 [18:55:11] INFO | >> train_qwenlatent.py:487 + Step 20820 | grad_norm_pre_clip=0.1762 | + grad_norm_pre_clip_avg=0.1965 | Metrics: + {'align_loss': 0.024410828948020935, + 'recon_loss': 0.05485249310731888, + 'predict_loss': 0.010222324170172215, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17623144388198853, + 'data_time': 0.0008187820203602314, + 'model_time': 1.1983030260016676, + 'grad_norm_pre_clip_avg': 0.19646230936050416, + 'learning_rate': 1.8127264823277437e-05, + 'epoch': 5.25} +04/19 [18:55:23] INFO | >> train_qwenlatent.py:487 + Step 20830 | grad_norm_pre_clip=0.1693 | + grad_norm_pre_clip_avg=0.1885 | Metrics: + {'align_loss': 0.026988845318555832, + 'recon_loss': 0.08976162225008011, + 'predict_loss': 0.016862254589796066, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16933155059814453, + 'data_time': 0.0006902599998284131, + 'model_time': 1.2367701340117492, + 'grad_norm_pre_clip_avg': 0.18852029144763946, + 'learning_rate': 1.8119476483486968e-05, + 'epoch': 5.26} +04/19 [18:55:36] INFO | >> train_qwenlatent.py:487 + Step 20840 | grad_norm_pre_clip=0.1638 | + grad_norm_pre_clip_avg=0.1750 | Metrics: + {'align_loss': 0.025591349229216576, + 'recon_loss': 0.08372119069099426, + 'predict_loss': 0.012428604066371918, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16377221047878265, + 'data_time': 0.0009805269946809858, + 'model_time': 1.2362003739981446, + 'grad_norm_pre_clip_avg': 0.17500619888305663, + 'learning_rate': 1.8111685410924282e-05, + 'epoch': 5.26} +04/19 [18:55:49] INFO | >> train_qwenlatent.py:487 + Step 20850 | grad_norm_pre_clip=0.2179 | + grad_norm_pre_clip_avg=0.1872 | Metrics: + {'align_loss': 0.02613455429673195, + 'recon_loss': 0.08151184767484665, + 'predict_loss': 0.014453926123678684, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21791918575763702, + 'mae_score': 0.01257478868639147, 'data_time': + 0.0009664819808676839, 'model_time': + 1.2379141679848544, 'grad_norm_pre_clip_avg': + 0.18721168786287307, 'learning_rate': + 1.8103891609386636e-05, 'epoch': 5.26} +04/19 [18:56:02] INFO | >> train_qwenlatent.py:487 + Step 20860 | grad_norm_pre_clip=0.2910 | + grad_norm_pre_clip_avg=0.2625 | Metrics: + {'align_loss': 0.025104984641075134, + 'recon_loss': 0.06593979895114899, + 'predict_loss': 0.011724610812962055, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.29098764061927795, + 'data_time': 0.0010647159942891449, + 'model_time': 1.5453193639987148, + 'grad_norm_pre_clip_avg': 0.2624608099460602, + 'learning_rate': 1.8096095082672653e-05, + 'epoch': 5.26} +04/19 [18:56:14] INFO | >> train_qwenlatent.py:487 + Step 20870 | grad_norm_pre_clip=0.2003 | + grad_norm_pre_clip_avg=0.2380 | Metrics: + {'align_loss': 0.02566879242658615, + 'recon_loss': 0.07724355161190033, + 'predict_loss': 0.007938055321574211, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20027878880500793, + 'data_time': 0.0007831190014258027, + 'model_time': 1.2287889779836405, + 'grad_norm_pre_clip_avg': 0.23799447417259217, + 'learning_rate': 1.808829583458225e-05, + 'epoch': 5.27} +04/19 [18:56:27] INFO | >> train_qwenlatent.py:487 + Step 20880 | grad_norm_pre_clip=0.1963 | + grad_norm_pre_clip_avg=0.2023 | Metrics: + {'align_loss': 0.0256202295422554, + 'recon_loss': 0.06608206778764725, + 'predict_loss': 0.011269846931099892, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19625914096832275, + 'data_time': 0.0007908059924375266, + 'model_time': 1.252298249019077, + 'grad_norm_pre_clip_avg': 0.20231730192899705, + 'learning_rate': 1.8080493868916702e-05, + 'epoch': 5.27} +04/19 [18:56:39] INFO | >> train_qwenlatent.py:487 + Step 20890 | grad_norm_pre_clip=0.1826 | + grad_norm_pre_clip_avg=0.1869 | Metrics: + {'align_loss': 0.02552514150738716, + 'recon_loss': 0.07066211849451065, + 'predict_loss': 0.011498461477458477, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18257129192352295, + 'data_time': 0.0009651950094848871, + 'model_time': 1.256797567999456, + 'grad_norm_pre_clip_avg': 0.18687610775232316, + 'learning_rate': 1.8072689189478574e-05, + 'epoch': 5.27} +04/19 [18:56:53] INFO | >> train_qwenlatent.py:487 + Step 20900 | grad_norm_pre_clip=0.2030 | + grad_norm_pre_clip_avg=0.2264 | Metrics: + {'align_loss': 0.024763550609350204, + 'recon_loss': 0.07588620483875275, + 'predict_loss': 0.011614121496677399, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20296259224414825, + 'mae_score': 0.009645523896088472, 'data_time': + 0.0009484559996053576, 'model_time': + 1.537277391005773, 'grad_norm_pre_clip_avg': + 0.22641828805208206, 'learning_rate': + 1.8064881800071782e-05, 'epoch': 5.27} +04/19 [18:57:06] INFO | >> train_qwenlatent.py:487 + Step 20910 | grad_norm_pre_clip=0.1689 | + grad_norm_pre_clip_avg=0.2031 | Metrics: + {'align_loss': 0.026149313896894455, + 'recon_loss': 0.0640181303024292, + 'predict_loss': 0.007390438113361597, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16890788078308105, + 'data_time': 0.0007704969902988523, + 'model_time': 1.2306191879906692, + 'grad_norm_pre_clip_avg': 0.20305259823799132, + 'learning_rate': 1.805707170450156e-05, + 'epoch': 5.28} +04/19 [18:57:18] INFO | >> train_qwenlatent.py:487 + Step 20920 | grad_norm_pre_clip=0.1732 | + grad_norm_pre_clip_avg=0.2081 | Metrics: + {'align_loss': 0.024673979729413986, + 'recon_loss': 0.04985121265053749, + 'predict_loss': 0.007876504212617874, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17318840324878693, + 'data_time': 0.0008101759885903448, + 'model_time': 1.2172456359839998, + 'grad_norm_pre_clip_avg': 0.20811088532209396, + 'learning_rate': 1.8049258906574432e-05, + 'epoch': 5.28} +04/19 [18:57:31] INFO | >> train_qwenlatent.py:487 + Step 20930 | grad_norm_pre_clip=0.1695 | + grad_norm_pre_clip_avg=0.1803 | Metrics: + {'align_loss': 0.025392994284629822, + 'recon_loss': 0.07404030859470367, + 'predict_loss': 0.010869327001273632, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1695120483636856, + 'data_time': 0.0009297560027334839, + 'model_time': 1.2177842360106297, + 'grad_norm_pre_clip_avg': 0.18034370690584184, + 'learning_rate': 1.8041443410098287e-05, + 'epoch': 5.28} +04/19 [18:57:44] INFO | >> train_qwenlatent.py:487 + Step 20940 | grad_norm_pre_clip=0.1827 | + grad_norm_pre_clip_avg=0.2160 | Metrics: + {'align_loss': 0.024186043068766594, + 'recon_loss': 0.04959677904844284, + 'predict_loss': 0.007684946525841951, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18274478614330292, + 'data_time': 0.0010580659727565944, + 'model_time': 1.2878396739833988, + 'grad_norm_pre_clip_avg': 0.21598974317312242, + 'learning_rate': 1.803362521888228e-05, + 'epoch': 5.28} +04/19 [18:57:57] INFO | >> train_qwenlatent.py:487 + Step 20950 | grad_norm_pre_clip=0.1889 | + grad_norm_pre_clip_avg=0.1992 | Metrics: + {'align_loss': 0.025151116773486137, + 'recon_loss': 0.08950120210647583, + 'predict_loss': 0.012544115073978901, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1888965368270874, + 'mae_score': 0.01132657850110853, 'data_time': + 0.000920270977076143, 'model_time': + 1.2021874240017496, 'grad_norm_pre_clip_avg': + 0.1991880178451538, 'learning_rate': + 1.8025804336736922e-05, 'epoch': 5.29} +04/19 [18:58:09] INFO | >> train_qwenlatent.py:487 + Step 20960 | grad_norm_pre_clip=0.2413 | + grad_norm_pre_clip_avg=0.2162 | Metrics: + {'align_loss': 0.0262070931494236, + 'recon_loss': 0.06251247227191925, + 'predict_loss': 0.01083938404917717, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2413330227136612, + 'data_time': 0.0007107880082912743, + 'model_time': 1.2229280509927776, + 'grad_norm_pre_clip_avg': 0.2162172630429268, + 'learning_rate': 1.8017980767474003e-05, + 'epoch': 5.29} +04/19 [18:58:22] INFO | >> train_qwenlatent.py:487 + Step 20970 | grad_norm_pre_clip=0.1864 | + grad_norm_pre_clip_avg=0.2020 | Metrics: + {'align_loss': 0.023899026215076447, + 'recon_loss': 0.09662767499685287, + 'predict_loss': 0.01365607138723135, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18641237914562225, + 'data_time': 0.0010627979936543852, + 'model_time': 1.4828875090170186, + 'grad_norm_pre_clip_avg': 0.20204667896032333, + 'learning_rate': 1.8010154514906636e-05, + 'epoch': 5.29} +04/19 [18:58:35] INFO | >> train_qwenlatent.py:487 + Step 20980 | grad_norm_pre_clip=0.2026 | + grad_norm_pre_clip_avg=0.1919 | Metrics: + {'align_loss': 0.026305485516786575, + 'recon_loss': 0.08671275526285172, + 'predict_loss': 0.010076817125082016, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20263709127902985, + 'data_time': 0.0008669460075907409, + 'model_time': 1.2150085830071475, + 'grad_norm_pre_clip_avg': 0.1918626308441162, + 'learning_rate': 1.8002325582849246e-05, + 'epoch': 5.29} +04/19 [18:58:47] INFO | >> train_qwenlatent.py:487 + Step 20990 | grad_norm_pre_clip=0.1509 | + grad_norm_pre_clip_avg=0.1742 | Metrics: + {'align_loss': 0.02534593641757965, + 'recon_loss': 0.0644279420375824, + 'predict_loss': 0.011775445193052292, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15085268020629883, + 'data_time': 0.0007168180018197745, + 'model_time': 1.2615367179969326, + 'grad_norm_pre_clip_avg': 0.17419970780611038, + 'learning_rate': 1.799449397511756e-05, + 'epoch': 5.3} +04/19 [18:59:00] INFO | >> train_qwenlatent.py:487 + Step 21000 | grad_norm_pre_clip=0.2227 | + grad_norm_pre_clip_avg=0.2711 | Metrics: + {'align_loss': 0.02664191834628582, + 'recon_loss': 0.06955241411924362, + 'predict_loss': 0.009532084688544273, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22274692356586456, + 'mae_score': 0.01277175422187324, 'data_time': + 0.0006654980243183672, 'model_time': + 1.2402755619841628, 'grad_norm_pre_clip_avg': + 0.2710714042186737, 'learning_rate': + 1.7986659695528605e-05, 'epoch': 5.3} +04/19 [18:59:13] INFO | >> train_qwenlatent.py:487 + Step 21010 | grad_norm_pre_clip=0.1529 | + grad_norm_pre_clip_avg=0.2239 | Metrics: + {'align_loss': 0.024157188832759857, + 'recon_loss': 0.09583728015422821, + 'predict_loss': 0.015763266012072563, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1528702676296234, + 'data_time': 0.0006733439804520458, + 'model_time': 1.3048649280099198, + 'grad_norm_pre_clip_avg': 0.22387474924325942, + 'learning_rate': 1.7978822747900716e-05, + 'epoch': 5.3} +04/19 [18:59:26] INFO | >> train_qwenlatent.py:487 + Step 21020 | grad_norm_pre_clip=0.2011 | + grad_norm_pre_clip_avg=0.1897 | Metrics: + {'align_loss': 0.02539987862110138, + 'recon_loss': 0.085074283182621, + 'predict_loss': 0.013526296243071556, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20114170014858246, + 'data_time': 0.0008280070032924414, + 'model_time': 1.193181844020728, + 'grad_norm_pre_clip_avg': 0.18966863602399825, + 'learning_rate': 1.7970983136053527e-05, + 'epoch': 5.3} +04/19 [18:59:38] INFO | >> train_qwenlatent.py:487 + Step 21030 | grad_norm_pre_clip=0.2045 | + grad_norm_pre_clip_avg=0.1820 | Metrics: + {'align_loss': 0.02535562962293625, + 'recon_loss': 0.06307218968868256, + 'predict_loss': 0.008051428943872452, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2044551819562912, + 'data_time': 0.0006590090051759034, + 'model_time': 1.2396962750062812, + 'grad_norm_pre_clip_avg': 0.18203312754631043, + 'learning_rate': 1.7963140863807967e-05, + 'epoch': 5.31} +04/19 [18:59:51] INFO | >> train_qwenlatent.py:487 + Step 21040 | grad_norm_pre_clip=0.2360 | + grad_norm_pre_clip_avg=0.1874 | Metrics: + {'align_loss': 0.02592686004936695, + 'recon_loss': 0.07380165904760361, + 'predict_loss': 0.0132854413241148, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.235971137881279, + 'data_time': 0.0006951740069780499, + 'model_time': 1.250100712000858, + 'grad_norm_pre_clip_avg': 0.18738338053226472, + 'learning_rate': 1.795529593498626e-05, + 'epoch': 5.31} +04/19 [19:00:04] INFO | >> train_qwenlatent.py:487 + Step 21050 | grad_norm_pre_clip=0.2413 | + grad_norm_pre_clip_avg=0.2277 | Metrics: + {'align_loss': 0.0249493308365345, + 'recon_loss': 0.06695270538330078, + 'predict_loss': 0.008462075144052505, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.241255521774292, + 'mae_score': 0.024514617576255456, 'data_time': + 0.001070807978976518, 'model_time': + 1.246463624003809, 'grad_norm_pre_clip_avg': + 0.22768849581480027, 'learning_rate': + 1.7947448353411942e-05, 'epoch': 5.31} +04/19 [19:00:17] INFO | >> train_qwenlatent.py:487 + Step 21060 | grad_norm_pre_clip=0.1795 | + grad_norm_pre_clip_avg=0.1859 | Metrics: + {'align_loss': 0.025747984647750854, + 'recon_loss': 0.09225326776504517, + 'predict_loss': 0.015552150085568428, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17947626113891602, + 'data_time': 0.0007352429965976626, + 'model_time': 1.231938133976655, + 'grad_norm_pre_clip_avg': 0.18588979095220565, + 'learning_rate': 1.7939598122909815e-05, + 'epoch': 5.31} +04/19 [19:00:30] INFO | >> train_qwenlatent.py:487 + Step 21070 | grad_norm_pre_clip=0.2571 | + grad_norm_pre_clip_avg=0.1930 | Metrics: + {'align_loss': 0.024147748947143555, + 'recon_loss': 0.08007685095071793, + 'predict_loss': 0.013298139907419682, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2570948004722595, + 'data_time': 0.0006588590040337294, + 'model_time': 1.2121406010119244, + 'grad_norm_pre_clip_avg': 0.19299839586019515, + 'learning_rate': 1.7931745247305993e-05, + 'epoch': 5.32} +04/19 [19:00:42] INFO | >> train_qwenlatent.py:487 + Step 21080 | grad_norm_pre_clip=0.1721 | + grad_norm_pre_clip_avg=0.1687 | Metrics: + {'align_loss': 0.027076356112957, 'recon_loss': + 0.07884585112333298, 'predict_loss': + 0.007057540584355593, 'aux_loss_decay_weight': + 0.0, 'grad_norm_pre_clip': 0.17213863134384155, + 'data_time': 0.000682070996845141, + 'model_time': 1.235763725999277, + 'grad_norm_pre_clip_avg': 0.16869794726371765, + 'learning_rate': 1.7923889730427877e-05, + 'epoch': 5.32} +04/19 [19:00:55] INFO | >> train_qwenlatent.py:487 + Step 21090 | grad_norm_pre_clip=0.3292 | + grad_norm_pre_clip_avg=0.2535 | Metrics: + {'align_loss': 0.025539681315422058, + 'recon_loss': 0.0989571288228035, + 'predict_loss': 0.016440020874142647, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.32923072576522827, + 'data_time': 0.0006904690235387534, + 'model_time': 1.3259069280175027, + 'grad_norm_pre_clip_avg': 0.25349313467741014, + 'learning_rate': 1.791603157610414e-05, + 'epoch': 5.32} +04/19 [19:01:08] INFO | >> train_qwenlatent.py:487 + Step 21100 | grad_norm_pre_clip=0.1624 | + grad_norm_pre_clip_avg=0.1913 | Metrics: + {'align_loss': 0.02465103566646576, + 'recon_loss': 0.08317908644676208, + 'predict_loss': 0.014403699897229671, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1624317318201065, + 'mae_score': 0.011541351112159523, 'data_time': + 0.0007586819992866367, 'model_time': + 1.2053208049910609, 'grad_norm_pre_clip_avg': + 0.19133099913597107, 'learning_rate': + 1.7908170788164757e-05, 'epoch': 5.32} +04/19 [19:01:21] INFO | >> train_qwenlatent.py:487 + Step 21110 | grad_norm_pre_clip=0.1990 | + grad_norm_pre_clip_avg=0.1901 | Metrics: + {'align_loss': 0.025842051953077316, + 'recon_loss': 0.06355293095111847, + 'predict_loss': 0.009993926621973515, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19903460144996643, + 'data_time': 0.001096741994842887, + 'model_time': 1.3427787969994824, + 'grad_norm_pre_clip_avg': 0.1901273772120476, + 'learning_rate': 1.7900307370440977e-05, + 'epoch': 5.33} +04/19 [19:01:33] INFO | >> train_qwenlatent.py:487 + Step 21120 | grad_norm_pre_clip=0.2506 | + grad_norm_pre_clip_avg=0.1814 | Metrics: + {'align_loss': 0.024900687858462334, + 'recon_loss': 0.08328105509281158, + 'predict_loss': 0.012529395520687103, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.250597208738327, + 'data_time': 0.0008216089918278158, + 'model_time': 1.1907717990106903, + 'grad_norm_pre_clip_avg': 0.18137801736593245, + 'learning_rate': 1.789244132676534e-05, + 'epoch': 5.33} +04/19 [19:01:46] INFO | >> train_qwenlatent.py:487 + Step 21130 | grad_norm_pre_clip=0.1688 | + grad_norm_pre_clip_avg=0.2023 | Metrics: + {'align_loss': 0.024708596989512444, + 'recon_loss': 0.072380930185318, + 'predict_loss': 0.012007639743387699, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16879619657993317, + 'data_time': 0.0008792919979896396, + 'model_time': 1.2333622809965163, + 'grad_norm_pre_clip_avg': 0.20231501162052154, + 'learning_rate': 1.788457266097165e-05, + 'epoch': 5.33} +04/19 [19:01:58] INFO | >> train_qwenlatent.py:487 + Step 21140 | grad_norm_pre_clip=0.1893 | + grad_norm_pre_clip_avg=0.1863 | Metrics: + {'align_loss': 0.02575480379164219, + 'recon_loss': 0.08068837225437164, + 'predict_loss': 0.01407605316489935, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18932095170021057, + 'data_time': 0.000847332994453609, + 'model_time': 1.2739524529897608, + 'grad_norm_pre_clip_avg': 0.18630134612321853, + 'learning_rate': 1.787670137689501e-05, + 'epoch': 5.33} +04/19 [19:02:11] INFO | >> train_qwenlatent.py:487 + Step 21150 | grad_norm_pre_clip=0.2186 | + grad_norm_pre_clip_avg=0.1846 | Metrics: + {'align_loss': 0.026062991470098495, + 'recon_loss': 0.08241888880729675, + 'predict_loss': 0.02031877264380455, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21855497360229492, + 'mae_score': 0.012193791501156919, 'data_time': + 0.0008305170049425215, 'model_time': + 1.2003044279990718, 'grad_norm_pre_clip_avg': + 0.1845589205622673, 'learning_rate': + 1.7868827478371785e-05, 'epoch': 5.34} +04/19 [19:02:24] INFO | >> train_qwenlatent.py:487 + Step 21160 | grad_norm_pre_clip=0.1714 | + grad_norm_pre_clip_avg=0.2107 | Metrics: + {'align_loss': 0.024254750460386276, + 'recon_loss': 0.052220072597265244, + 'predict_loss': 0.006508796475827694, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1713511049747467, + 'data_time': 0.0005848629807587713, + 'model_time': 1.2370002130046487, + 'grad_norm_pre_clip_avg': 0.21074071228504182, + 'learning_rate': 1.786095096923962e-05, + 'epoch': 5.34} +04/19 [19:02:36] INFO | >> train_qwenlatent.py:487 + Step 21170 | grad_norm_pre_clip=0.2985 | + grad_norm_pre_clip_avg=0.2609 | Metrics: + {'align_loss': 0.026292528957128525, + 'recon_loss': 0.10172148048877716, + 'predict_loss': 0.010951184667646885, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2985312342643738, + 'data_time': 0.0005854299815837294, + 'model_time': 1.2051647990010679, + 'grad_norm_pre_clip_avg': 0.2608961552381516, + 'learning_rate': 1.7853071853337418e-05, + 'epoch': 5.34} +04/19 [19:02:49] INFO | >> train_qwenlatent.py:487 + Step 21180 | grad_norm_pre_clip=0.1945 | + grad_norm_pre_clip_avg=0.2142 | Metrics: + {'align_loss': 0.02465125545859337, + 'recon_loss': 0.055484093725681305, + 'predict_loss': 0.00649225665256381, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1945468932390213, + 'data_time': 0.000974822003627196, + 'model_time': 1.2486786610097624, + 'grad_norm_pre_clip_avg': 0.21420223712921144, + 'learning_rate': 1.7845190134505375e-05, + 'epoch': 5.34} +04/19 [19:03:02] INFO | >> train_qwenlatent.py:487 + Step 21190 | grad_norm_pre_clip=0.1744 | + grad_norm_pre_clip_avg=0.1737 | Metrics: + {'align_loss': 0.025667881593108177, + 'recon_loss': 0.09206712245941162, + 'predict_loss': 0.011918923817574978, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1743861585855484, + 'data_time': 0.0009740840177983046, + 'model_time': 1.5832917359948624, + 'grad_norm_pre_clip_avg': 0.17372078746557235, + 'learning_rate': 1.783730581658495e-05, + 'epoch': 5.35} +04/19 [19:03:16] INFO | >> train_qwenlatent.py:487 + Step 21200 | grad_norm_pre_clip=0.1774 | + grad_norm_pre_clip_avg=0.1628 | Metrics: + {'align_loss': 0.024524401873350143, + 'recon_loss': 0.0926302969455719, + 'predict_loss': 0.01616530679166317, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1773613691329956, + 'mae_score': 0.011370552337921417, 'data_time': + 0.0011492460034787655, 'model_time': + 1.2689701429917477, 'grad_norm_pre_clip_avg': + 0.1628465846180916, 'learning_rate': + 1.7829418903418845e-05, 'epoch': 5.35} +04/19 [19:03:28] INFO | >> train_qwenlatent.py:487 + Step 21210 | grad_norm_pre_clip=0.2230 | + grad_norm_pre_clip_avg=0.1873 | Metrics: + {'align_loss': 0.026146620512008667, + 'recon_loss': 0.08884783834218979, + 'predict_loss': 0.02037579007446766, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22303904592990875, + 'data_time': 0.0010604989947751164, + 'model_time': 1.221797456004424, + 'grad_norm_pre_clip_avg': 0.18726631104946137, + 'learning_rate': 1.7821529398851065e-05, + 'epoch': 5.35} +04/19 [19:03:41] INFO | >> train_qwenlatent.py:487 + Step 21220 | grad_norm_pre_clip=0.1649 | + grad_norm_pre_clip_avg=0.2266 | Metrics: + {'align_loss': 0.02579948678612709, + 'recon_loss': 0.07253875583410263, + 'predict_loss': 0.01664796657860279, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16493786871433258, + 'data_time': 0.0007745189941488206, + 'model_time': 1.2958413420128636, + 'grad_norm_pre_clip_avg': 0.22663244158029555, + 'learning_rate': 1.781363730672685e-05, + 'epoch': 5.35} +04/19 [19:03:54] INFO | >> train_qwenlatent.py:487 + Step 21230 | grad_norm_pre_clip=0.1952 | + grad_norm_pre_clip_avg=0.1992 | Metrics: + {'align_loss': 0.025611910969018936, + 'recon_loss': 0.09081918746232986, + 'predict_loss': 0.01377298217266798, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19521097838878632, + 'data_time': 0.0009910129883792251, + 'model_time': 1.306868018989917, + 'grad_norm_pre_clip_avg': 0.19921120554208754, + 'learning_rate': 1.7805742630892704e-05, + 'epoch': 5.36} +04/19 [19:04:07] INFO | >> train_qwenlatent.py:487 + Step 21240 | grad_norm_pre_clip=0.1981 | + grad_norm_pre_clip_avg=0.1998 | Metrics: + {'align_loss': 0.026236873120069504, + 'recon_loss': 0.058111630380153656, + 'predict_loss': 0.007947306148707867, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19805526733398438, + 'data_time': 0.0007413740095216781, + 'model_time': 1.2427846330101602, + 'grad_norm_pre_clip_avg': 0.19980334043502807, + 'learning_rate': 1.7797845375196413e-05, + 'epoch': 5.36} +04/19 [19:04:20] INFO | >> train_qwenlatent.py:487 + Step 21250 | grad_norm_pre_clip=0.2804 | + grad_norm_pre_clip_avg=0.2170 | Metrics: + {'align_loss': 0.02504616603255272, + 'recon_loss': 0.06738490611314774, + 'predict_loss': 0.012976914644241333, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28041669726371765, + 'mae_score': 0.013425387133349169, 'data_time': + 0.0008705149812158197, 'model_time': + 1.2611225650180131, 'grad_norm_pre_clip_avg': + 0.21700366735458373, 'learning_rate': + 1.7789945543486985e-05, 'epoch': 5.36} +04/19 [19:04:32] INFO | >> train_qwenlatent.py:487 + Step 21260 | grad_norm_pre_clip=0.2036 | + grad_norm_pre_clip_avg=0.2059 | Metrics: + {'align_loss': 0.02604728192090988, + 'recon_loss': 0.08688563108444214, + 'predict_loss': 0.008857273496687412, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2035791277885437, + 'data_time': 0.0008253880077973008, + 'model_time': 1.2169787510065362, + 'grad_norm_pre_clip_avg': 0.20586046278476716, + 'learning_rate': 1.7782043139614714e-05, + 'epoch': 5.36} +04/19 [19:04:45] INFO | >> train_qwenlatent.py:487 + Step 21270 | grad_norm_pre_clip=0.2226 | + grad_norm_pre_clip_avg=0.2026 | Metrics: + {'align_loss': 0.025615280494093895, + 'recon_loss': 0.07645520567893982, + 'predict_loss': 0.008894240483641624, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2226104587316513, + 'data_time': 0.0007189229945652187, + 'model_time': 1.2241277240100317, + 'grad_norm_pre_clip_avg': 0.20256070792675018, + 'learning_rate': 1.7774138167431133e-05, + 'epoch': 5.37} +04/19 [19:04:58] INFO | >> train_qwenlatent.py:487 + Step 21280 | grad_norm_pre_clip=0.2638 | + grad_norm_pre_clip_avg=0.2204 | Metrics: + {'align_loss': 0.02609747275710106, + 'recon_loss': 0.07918261736631393, + 'predict_loss': 0.011950846761465073, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26381558179855347, + 'data_time': 0.0010461380006745458, + 'model_time': 1.31252311199205, + 'grad_norm_pre_clip_avg': 0.2204316720366478, + 'learning_rate': 1.7766230630789025e-05, + 'epoch': 5.37} +04/19 [19:05:10] INFO | >> train_qwenlatent.py:487 + Step 21290 | grad_norm_pre_clip=0.2087 | + grad_norm_pre_clip_avg=0.2063 | Metrics: + {'align_loss': 0.026433829218149185, + 'recon_loss': 0.07591112703084946, + 'predict_loss': 0.00852159969508648, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2087215781211853, + 'data_time': 0.000986816012300551, + 'model_time': 1.2674554730183445, + 'grad_norm_pre_clip_avg': 0.20631457567214967, + 'learning_rate': 1.7758320533542432e-05, + 'epoch': 5.37} +04/19 [19:05:23] INFO | >> train_qwenlatent.py:487 + Step 21300 | grad_norm_pre_clip=0.1788 | + grad_norm_pre_clip_avg=0.1699 | Metrics: + {'align_loss': 0.02539151906967163, + 'recon_loss': 0.10368242859840393, + 'predict_loss': 0.011009277775883675, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1788395792245865, + 'mae_score': 0.01359499252594269, 'data_time': + 0.0014162040024530143, 'model_time': + 1.2496977380069438, 'grad_norm_pre_clip_avg': + 0.1698664352297783, 'learning_rate': + 1.7750407879546637e-05, 'epoch': 5.37} +04/19 [19:05:36] INFO | >> train_qwenlatent.py:487 + Step 21310 | grad_norm_pre_clip=0.2274 | + grad_norm_pre_clip_avg=0.2007 | Metrics: + {'align_loss': 0.026224706321954727, + 'recon_loss': 0.0773439034819603, + 'predict_loss': 0.00844084657728672, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22741863131523132, + 'data_time': 0.0009075929992832243, + 'model_time': 1.2151020940218586, + 'grad_norm_pre_clip_avg': 0.200742968916893, + 'learning_rate': 1.7742492672658174e-05, + 'epoch': 5.38} +04/19 [19:05:49] INFO | >> train_qwenlatent.py:487 + Step 21320 | grad_norm_pre_clip=0.2486 | + grad_norm_pre_clip_avg=0.1964 | Metrics: + {'align_loss': 0.023704946041107178, + 'recon_loss': 0.05674772337079048, + 'predict_loss': 0.005969104822725058, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2486162781715393, + 'data_time': 0.000693946989485994, + 'model_time': 1.2362711430178024, + 'grad_norm_pre_clip_avg': 0.19641003310680388, + 'learning_rate': 1.7734574916734814e-05, + 'epoch': 5.38} +04/19 [19:06:02] INFO | >> train_qwenlatent.py:487 + Step 21330 | grad_norm_pre_clip=0.2573 | + grad_norm_pre_clip_avg=0.1970 | Metrics: + {'align_loss': 0.02456805109977722, + 'recon_loss': 0.09364922344684601, + 'predict_loss': 0.02555898018181324, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.257313072681427, + 'data_time': 0.0009490849915891886, + 'model_time': 1.2443353119888343, + 'grad_norm_pre_clip_avg': 0.19704030752182006, + 'learning_rate': 1.7726654615635586e-05, + 'epoch': 5.38} +04/19 [19:06:14] INFO | >> train_qwenlatent.py:487 + Step 21340 | grad_norm_pre_clip=0.2397 | + grad_norm_pre_clip_avg=0.2187 | Metrics: + {'align_loss': 0.025219444185495377, + 'recon_loss': 0.08563549816608429, + 'predict_loss': 0.010253234766423702, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23965269327163696, + 'data_time': 0.0012589010002557188, + 'model_time': 1.2692289970000274, + 'grad_norm_pre_clip_avg': 0.21868813633918763, + 'learning_rate': 1.7718731773220733e-05, + 'epoch': 5.38} +04/19 [19:06:27] INFO | >> train_qwenlatent.py:487 + Step 21350 | grad_norm_pre_clip=0.2872 | + grad_norm_pre_clip_avg=0.2100 | Metrics: + {'align_loss': 0.02601075917482376, + 'recon_loss': 0.06963333487510681, + 'predict_loss': 0.00843438133597374, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2872031331062317, + 'mae_score': 0.013854652267318588, 'data_time': + 0.0006978129968047142, 'model_time': + 1.213134628982516, 'grad_norm_pre_clip_avg': + 0.21001750975847244, 'learning_rate': + 1.7710806393351763e-05, 'epoch': 5.39} +04/19 [19:06:40] INFO | >> train_qwenlatent.py:487 + Step 21360 | grad_norm_pre_clip=0.1787 | + grad_norm_pre_clip_avg=0.2042 | Metrics: + {'align_loss': 0.02451121248304844, + 'recon_loss': 0.06675165146589279, + 'predict_loss': 0.010030821897089481, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17868748307228088, + 'data_time': 0.000851591001264751, + 'model_time': 1.210016586992424, + 'grad_norm_pre_clip_avg': 0.20417966991662978, + 'learning_rate': 1.7702878479891405e-05, + 'epoch': 5.39} +04/19 [19:06:53] INFO | >> train_qwenlatent.py:487 + Step 21370 | grad_norm_pre_clip=0.2042 | + grad_norm_pre_clip_avg=0.1985 | Metrics: + {'align_loss': 0.025010239332914352, + 'recon_loss': 0.10325155407190323, + 'predict_loss': 0.02163541130721569, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20415401458740234, + 'data_time': 0.0007928069971967489, + 'model_time': 1.2275571490172297, + 'grad_norm_pre_clip_avg': 0.19849195182323456, + 'learning_rate': 1.769494803670363e-05, + 'epoch': 5.39} +04/19 [19:07:05] INFO | >> train_qwenlatent.py:487 + Step 21380 | grad_norm_pre_clip=0.1734 | + grad_norm_pre_clip_avg=0.1928 | Metrics: + {'align_loss': 0.02516021765768528, + 'recon_loss': 0.090370774269104, + 'predict_loss': 0.012070780619978905, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1734146922826767, + 'data_time': 0.0009154840081464499, + 'model_time': 1.249521720019402, + 'grad_norm_pre_clip_avg': 0.19275523275136947, + 'learning_rate': 1.768701506765364e-05, + 'epoch': 5.39} +04/19 [19:07:18] INFO | >> train_qwenlatent.py:487 + Step 21390 | grad_norm_pre_clip=0.1527 | + grad_norm_pre_clip_avg=0.1770 | Metrics: + {'align_loss': 0.02654227800667286, + 'recon_loss': 0.10124407708644867, + 'predict_loss': 0.01314882654696703, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15267901122570038, + 'data_time': 0.0008126200118567795, + 'model_time': 1.1815112699987367, + 'grad_norm_pre_clip_avg': 0.17697890102863312, + 'learning_rate': 1.7679079576607864e-05, + 'epoch': 5.4} +04/19 [19:07:31] INFO | >> train_qwenlatent.py:487 + Step 21400 | grad_norm_pre_clip=0.2202 | + grad_norm_pre_clip_avg=0.1940 | Metrics: + {'align_loss': 0.0252838134765625, + 'recon_loss': 0.07512391358613968, + 'predict_loss': 0.01662125065922737, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22021394968032837, + 'mae_score': 0.010035267391720334, 'data_time': + 0.0006730709865223616, 'model_time': + 1.18136000999948, 'grad_norm_pre_clip_avg': + 0.1939914733171463, 'learning_rate': + 1.7671141567433968e-05, 'epoch': 5.4} +04/19 [19:07:43] INFO | >> train_qwenlatent.py:487 + Step 21410 | grad_norm_pre_clip=0.2413 | + grad_norm_pre_clip_avg=0.2250 | Metrics: + {'align_loss': 0.024306628853082657, + 'recon_loss': 0.0912720188498497, + 'predict_loss': 0.014707721769809723, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24128443002700806, + 'data_time': 0.0008936339872889221, + 'model_time': 1.243842275987845, + 'grad_norm_pre_clip_avg': 0.22502572536468507, + 'learning_rate': 1.7663201044000837e-05, + 'epoch': 5.4} +04/19 [19:07:56] INFO | >> train_qwenlatent.py:487 + Step 21420 | grad_norm_pre_clip=0.2403 | + grad_norm_pre_clip_avg=0.1979 | Metrics: + {'align_loss': 0.024426814168691635, + 'recon_loss': 0.0599898062646389, + 'predict_loss': 0.008622346445918083, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2403211146593094, + 'data_time': 0.0009370770130772144, + 'model_time': 1.226951557007851, + 'grad_norm_pre_clip_avg': 0.19792410731315613, + 'learning_rate': 1.7655258010178594e-05, + 'epoch': 5.4} +04/19 [19:08:08] INFO | >> train_qwenlatent.py:487 + Step 21430 | grad_norm_pre_clip=0.1791 | + grad_norm_pre_clip_avg=0.2052 | Metrics: + {'align_loss': 0.023429565131664276, + 'recon_loss': 0.06557976454496384, + 'predict_loss': 0.014309012331068516, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17906256020069122, + 'data_time': 0.0009467510099057108, + 'model_time': 1.214841973996954, + 'grad_norm_pre_clip_avg': 0.20523680746555328, + 'learning_rate': 1.764731246983857e-05, + 'epoch': 5.41} +04/19 [19:08:21] INFO | >> train_qwenlatent.py:487 + Step 21440 | grad_norm_pre_clip=0.1923 | + grad_norm_pre_clip_avg=0.1945 | Metrics: + {'align_loss': 0.025022953748703003, + 'recon_loss': 0.07453491538763046, + 'predict_loss': 0.010495484806597233, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19228871166706085, + 'data_time': 0.0011309959809295833, + 'model_time': 1.522633417014731, + 'grad_norm_pre_clip_avg': 0.19453758150339126, + 'learning_rate': 1.7639364426853323e-05, + 'epoch': 5.41} +04/19 [19:08:35] INFO | >> train_qwenlatent.py:487 + Step 21450 | grad_norm_pre_clip=0.1589 | + grad_norm_pre_clip_avg=0.1825 | Metrics: + {'align_loss': 0.02635147050023079, + 'recon_loss': 0.05901765823364258, + 'predict_loss': 0.011313818395137787, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1588771939277649, + 'mae_score': 0.011323302500956768, 'data_time': + 0.0008768870029598475, 'model_time': + 1.2408718180085998, 'grad_norm_pre_clip_avg': + 0.1824750632047653, 'learning_rate': + 1.763141388509664e-05, 'epoch': 5.41} +04/19 [19:08:48] INFO | >> train_qwenlatent.py:487 + Step 21460 | grad_norm_pre_clip=0.2037 | + grad_norm_pre_clip_avg=0.2140 | Metrics: + {'align_loss': 0.02420003153383732, + 'recon_loss': 0.08790519833564758, + 'predict_loss': 0.01015262957662344, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20367811620235443, + 'data_time': 0.0015019649872556329, + 'model_time': 1.2474661839951295, + 'grad_norm_pre_clip_avg': 0.21400301158428192, + 'learning_rate': 1.7623460848443517e-05, + 'epoch': 5.42} +04/19 [19:09:00] INFO | >> train_qwenlatent.py:487 + Step 21470 | grad_norm_pre_clip=0.1671 | + grad_norm_pre_clip_avg=0.1839 | Metrics: + {'align_loss': 0.025807345286011696, + 'recon_loss': 0.09168633818626404, + 'predict_loss': 0.009258064441382885, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16706053912639618, + 'data_time': 0.0007457210158463567, + 'model_time': 1.2428956689836923, + 'grad_norm_pre_clip_avg': 0.18388396203517915, + 'learning_rate': 1.7615505320770162e-05, + 'epoch': 5.42} +04/19 [19:09:13] INFO | >> train_qwenlatent.py:487 + Step 21480 | grad_norm_pre_clip=0.2831 | + grad_norm_pre_clip_avg=0.2175 | Metrics: + {'align_loss': 0.024476144462823868, + 'recon_loss': 0.05411418154835701, + 'predict_loss': 0.01011319737881422, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28313806653022766, + 'data_time': 0.0007308339991141111, + 'model_time': 1.2441933049994987, + 'grad_norm_pre_clip_avg': 0.2175460711121559, + 'learning_rate': 1.7607547305954002e-05, + 'epoch': 5.42} +04/19 [19:09:26] INFO | >> train_qwenlatent.py:487 + Step 21490 | grad_norm_pre_clip=0.1643 | + grad_norm_pre_clip_avg=0.2106 | Metrics: + {'align_loss': 0.024261852726340294, + 'recon_loss': 0.06865779310464859, + 'predict_loss': 0.009430062025785446, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16425609588623047, + 'data_time': 0.0007017060124780983, + 'model_time': 1.2936356750142295, + 'grad_norm_pre_clip_avg': 0.21056391000747682, + 'learning_rate': 1.7599586807873684e-05, + 'epoch': 5.42} +04/19 [19:09:39] INFO | >> train_qwenlatent.py:487 + Step 21500 | grad_norm_pre_clip=0.1799 | + grad_norm_pre_clip_avg=0.1806 | Metrics: + {'align_loss': 0.02571881003677845, + 'recon_loss': 0.09508319944143295, + 'predict_loss': 0.020212436094880104, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1799331158399582, + 'mae_score': 0.009496315105541333, 'data_time': + 0.0009177119936794043, 'model_time': + 1.2412064879899845, 'grad_norm_pre_clip_avg': + 0.18064370155334472, 'learning_rate': + 1.7591623830409057e-05, 'epoch': 5.43} +04/19 [19:09:52] INFO | >> train_qwenlatent.py:487 + Step 21510 | grad_norm_pre_clip=0.2202 | + grad_norm_pre_clip_avg=0.1756 | Metrics: + {'align_loss': 0.024730147793889046, + 'recon_loss': 0.0764942467212677, + 'predict_loss': 0.01469473633915186, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2202209234237671, + 'data_time': 0.000692564994096756, + 'model_time': 1.2645768749935087, + 'grad_norm_pre_clip_avg': 0.17559237778186798, + 'learning_rate': 1.7583658377441172e-05, + 'epoch': 5.43} +04/19 [19:10:05] INFO | >> train_qwenlatent.py:487 + Step 21520 | grad_norm_pre_clip=0.2478 | + grad_norm_pre_clip_avg=0.1889 | Metrics: + {'align_loss': 0.026365116238594055, + 'recon_loss': 0.07481112331151962, + 'predict_loss': 0.008649523369967937, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24783366918563843, + 'data_time': 0.0008317070023622364, + 'model_time': 1.2806309070147108, + 'grad_norm_pre_clip_avg': 0.18888025283813475, + 'learning_rate': 1.7575690452852302e-05, + 'epoch': 5.43} +04/19 [19:10:17] INFO | >> train_qwenlatent.py:487 + Step 21530 | grad_norm_pre_clip=0.2168 | + grad_norm_pre_clip_avg=0.2002 | Metrics: + {'align_loss': 0.024539778009057045, + 'recon_loss': 0.06373953819274902, + 'predict_loss': 0.010162833146750927, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21681204438209534, + 'data_time': 0.0007272980001289397, + 'model_time': 1.2287998019892257, + 'grad_norm_pre_clip_avg': 0.2002265051007271, + 'learning_rate': 1.7567720060525914e-05, + 'epoch': 5.43} +04/19 [19:10:29] INFO | >> train_qwenlatent.py:487 + Step 21540 | grad_norm_pre_clip=0.2203 | + grad_norm_pre_clip_avg=0.1826 | Metrics: + {'align_loss': 0.026000533252954483, + 'recon_loss': 0.08732769638299942, + 'predict_loss': 0.01068144105374813, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22025391459465027, + 'data_time': 0.0007104169926606119, + 'model_time': 1.2318308969843201, + 'grad_norm_pre_clip_avg': 0.18256218284368514, + 'learning_rate': 1.7559747204346677e-05, + 'epoch': 5.44} +04/19 [19:10:43] INFO | >> train_qwenlatent.py:487 + Step 21550 | grad_norm_pre_clip=0.2482 | + grad_norm_pre_clip_avg=0.2313 | Metrics: + {'align_loss': 0.02613799087703228, + 'recon_loss': 0.098829485476017, + 'predict_loss': 0.014649313874542713, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24823671579360962, + 'mae_score': 0.010992364625673037, 'data_time': + 0.0007091519946698099, 'model_time': + 1.2268511779839173, 'grad_norm_pre_clip_avg': + 0.2312942460179329, 'learning_rate': + 1.7551771888200474e-05, 'epoch': 5.44} +04/19 [19:10:55] INFO | >> train_qwenlatent.py:487 + Step 21560 | grad_norm_pre_clip=0.1858 | + grad_norm_pre_clip_avg=0.1854 | Metrics: + {'align_loss': 0.02536107413470745, + 'recon_loss': 0.07526616752147675, + 'predict_loss': 0.007291250862181187, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1857661008834839, + 'data_time': 0.0009571729751769453, + 'model_time': 1.210433981003007, + 'grad_norm_pre_clip_avg': 0.1853506952524185, + 'learning_rate': 1.7543794115974362e-05, + 'epoch': 5.44} +04/19 [19:11:08] INFO | >> train_qwenlatent.py:487 + Step 21570 | grad_norm_pre_clip=0.2096 | + grad_norm_pre_clip_avg=0.1837 | Metrics: + {'align_loss': 0.02564501017332077, + 'recon_loss': 0.062193311750888824, + 'predict_loss': 0.008229558356106281, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2096157968044281, + 'data_time': 0.0007207859889604151, + 'model_time': 1.2412345070042647, + 'grad_norm_pre_clip_avg': 0.18370144069194794, + 'learning_rate': 1.7535813891556626e-05, + 'epoch': 5.44} +04/19 [19:11:20] INFO | >> train_qwenlatent.py:487 + Step 21580 | grad_norm_pre_clip=0.2260 | + grad_norm_pre_clip_avg=0.2005 | Metrics: + {'align_loss': 0.025928597897291183, + 'recon_loss': 0.08877549320459366, + 'predict_loss': 0.010693098418414593, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22598282992839813, + 'data_time': 0.0009030930232256651, + 'model_time': 1.2318175380059984, + 'grad_norm_pre_clip_avg': 0.20045305788516998, + 'learning_rate': 1.7527831218836718e-05, + 'epoch': 5.45} +04/19 [19:11:33] INFO | >> train_qwenlatent.py:487 + Step 21590 | grad_norm_pre_clip=0.2446 | + grad_norm_pre_clip_avg=0.2071 | Metrics: + {'align_loss': 0.02525576576590538, + 'recon_loss': 0.08351543545722961, + 'predict_loss': 0.01591191068291664, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24463297426700592, + 'data_time': 0.0009369039908051491, + 'model_time': 1.275357314007124, + 'grad_norm_pre_clip_avg': 0.20710785686969757, + 'learning_rate': 1.7519846101705303e-05, + 'epoch': 5.45} +04/19 [19:11:47] INFO | >> train_qwenlatent.py:487 + Step 21600 | grad_norm_pre_clip=0.1538 | + grad_norm_pre_clip_avg=0.1857 | Metrics: + {'align_loss': 0.025975540280342102, + 'recon_loss': 0.07319808006286621, + 'predict_loss': 0.009661984629929066, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15381622314453125, + 'mae_score': 0.009330594002663553, 'data_time': + 0.0008215800044126809, 'model_time': + 1.2467187890142668, 'grad_norm_pre_clip_avg': + 0.1857482448220253, 'learning_rate': + 1.7511858544054228e-05, 'epoch': 5.45} +04/19 [19:12:00] INFO | >> train_qwenlatent.py:487 + Step 21610 | grad_norm_pre_clip=0.2258 | + grad_norm_pre_clip_avg=0.2466 | Metrics: + {'align_loss': 0.024537421762943268, + 'recon_loss': 0.062023941427469254, + 'predict_loss': 0.008855557069182396, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22575464844703674, + 'data_time': 0.0007869249966461211, + 'model_time': 1.5398089679947589, + 'grad_norm_pre_clip_avg': 0.24655490666627883, + 'learning_rate': 1.7503868549776532e-05, + 'epoch': 5.45} +04/19 [19:12:12] INFO | >> train_qwenlatent.py:487 + Step 21620 | grad_norm_pre_clip=0.1668 | + grad_norm_pre_clip_avg=0.1710 | Metrics: + {'align_loss': 0.02464934065937996, + 'recon_loss': 0.0754602700471878, + 'predict_loss': 0.009209399111568928, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16680459678173065, + 'data_time': 0.002777390996925533, + 'model_time': 1.228359866014216, + 'grad_norm_pre_clip_avg': 0.17104488760232925, + 'learning_rate': 1.749587612276644e-05, + 'epoch': 5.46} +04/19 [19:12:25] INFO | >> train_qwenlatent.py:487 + Step 21630 | grad_norm_pre_clip=0.1713 | + grad_norm_pre_clip_avg=0.1905 | Metrics: + {'align_loss': 0.024181963875889778, + 'recon_loss': 0.06553880870342255, + 'predict_loss': 0.006723345257341862, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1712585985660553, + 'data_time': 0.0007779540028423071, + 'model_time': 1.2511132919753436, + 'grad_norm_pre_clip_avg': 0.19049688726663588, + 'learning_rate': 1.748788126691936e-05, + 'epoch': 5.46} +04/19 [19:12:38] INFO | >> train_qwenlatent.py:487 + Step 21640 | grad_norm_pre_clip=0.2377 | + grad_norm_pre_clip_avg=0.1953 | Metrics: + {'align_loss': 0.025082413107156754, + 'recon_loss': 0.08319500088691711, + 'predict_loss': 0.011202449910342693, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2377234846353531, + 'data_time': 0.0009166199888568372, + 'model_time': 1.259723564988235, + 'grad_norm_pre_clip_avg': 0.19530630260705947, + 'learning_rate': 1.74798839861319e-05, 'epoch': + 5.46} +04/19 [19:12:50] INFO | >> train_qwenlatent.py:487 + Step 21650 | grad_norm_pre_clip=0.2033 | + grad_norm_pre_clip_avg=0.1941 | Metrics: + {'align_loss': 0.026403021067380905, + 'recon_loss': 0.09238028526306152, + 'predict_loss': 0.013313733972609043, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20327815413475037, + 'mae_score': 0.013289867435489688, 'data_time': + 0.000945192005019635, 'model_time': + 1.1810085449833423, 'grad_norm_pre_clip_avg': + 0.19407490640878677, 'learning_rate': + 1.7471884284301822e-05, 'epoch': 5.46} +04/19 [19:13:03] INFO | >> train_qwenlatent.py:487 + Step 21660 | grad_norm_pre_clip=0.3224 | + grad_norm_pre_clip_avg=0.2007 | Metrics: + {'align_loss': 0.025666171684861183, + 'recon_loss': 0.10330777615308762, + 'predict_loss': 0.012797297909855843, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3223786950111389, + 'data_time': 0.0007754119869787246, + 'model_time': 1.2258182590012439, + 'grad_norm_pre_clip_avg': 0.20067828744649888, + 'learning_rate': 1.7463882165328097e-05, + 'epoch': 5.47} +04/19 [19:13:15] INFO | >> train_qwenlatent.py:487 + Step 21670 | grad_norm_pre_clip=0.2513 | + grad_norm_pre_clip_avg=0.2607 | Metrics: + {'align_loss': 0.025137310847640038, + 'recon_loss': 0.05977512523531914, + 'predict_loss': 0.005804980639368296, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25131756067276, + 'data_time': 0.0009267829882446676, + 'model_time': 1.275854165025521, + 'grad_norm_pre_clip_avg': 0.26071313917636874, + 'learning_rate': 1.745587763311085e-05, + 'epoch': 5.47} +04/19 [19:13:28] INFO | >> train_qwenlatent.py:487 + Step 21680 | grad_norm_pre_clip=0.1663 | + grad_norm_pre_clip_avg=0.1832 | Metrics: + {'align_loss': 0.02529177814722061, + 'recon_loss': 0.06006322801113129, + 'predict_loss': 0.0078033157624304295, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16628293693065643, + 'data_time': 0.0007506659894715995, + 'model_time': 1.2365645880054217, + 'grad_norm_pre_clip_avg': 0.18318276554346086, + 'learning_rate': 1.7447870691551404e-05, + 'epoch': 5.47} +04/19 [19:13:41] INFO | >> train_qwenlatent.py:487 + Step 21690 | grad_norm_pre_clip=0.1763 | + grad_norm_pre_clip_avg=0.1825 | Metrics: + {'align_loss': 0.02355768531560898, + 'recon_loss': 0.08636567741632462, + 'predict_loss': 0.011647101491689682, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17629145085811615, + 'data_time': 0.0009642970107961446, + 'model_time': 1.2540615769976284, + 'grad_norm_pre_clip_avg': 0.18248216062784195, + 'learning_rate': 1.7439861344552236e-05, + 'epoch': 5.47} +04/19 [19:13:54] INFO | >> train_qwenlatent.py:487 + Step 21700 | grad_norm_pre_clip=0.1471 | + grad_norm_pre_clip_avg=0.1635 | Metrics: + {'align_loss': 0.023868948221206665, + 'recon_loss': 0.0710945799946785, + 'predict_loss': 0.011733119376003742, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14711187779903412, + 'mae_score': 0.013036739074432098, 'data_time': + 0.0010332690144423395, 'model_time': + 1.2717818490054924, 'grad_norm_pre_clip_avg': + 0.1634567990899086, 'learning_rate': + 1.7431849596017002e-05, 'epoch': 5.48} +04/19 [19:14:06] INFO | >> train_qwenlatent.py:487 + Step 21710 | grad_norm_pre_clip=0.2073 | + grad_norm_pre_clip_avg=0.2109 | Metrics: + {'align_loss': 0.023984070867300034, + 'recon_loss': 0.0769084170460701, + 'predict_loss': 0.012428641319274902, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2073243409395218, + 'data_time': 0.0006792899803258479, + 'model_time': 1.23849958597566, + 'grad_norm_pre_clip_avg': 0.21094920486211777, + 'learning_rate': 1.7423835449850538e-05, + 'epoch': 5.48} +04/19 [19:14:19] INFO | >> train_qwenlatent.py:487 + Step 21720 | grad_norm_pre_clip=0.2804 | + grad_norm_pre_clip_avg=0.2157 | Metrics: + {'align_loss': 0.022573277354240417, + 'recon_loss': 0.07865723222494125, + 'predict_loss': 0.015437262132763863, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28036317229270935, + 'data_time': 0.000663907005218789, + 'model_time': 1.21032715702313, + 'grad_norm_pre_clip_avg': 0.215691539645195, + 'learning_rate': 1.741581890995884e-05, + 'epoch': 5.48} +04/19 [19:14:33] INFO | >> train_qwenlatent.py:487 + Step 21730 | grad_norm_pre_clip=0.1868 | + grad_norm_pre_clip_avg=0.1904 | Metrics: + {'align_loss': 0.023959554731845856, + 'recon_loss': 0.06303972750902176, + 'predict_loss': 0.00826738215982914, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18684636056423187, + 'data_time': 0.0008917419763747603, + 'model_time': 1.2818465289892629, + 'grad_norm_pre_clip_avg': 0.19040510803461075, + 'learning_rate': 1.7407799980249075e-05, + 'epoch': 5.48} +04/19 [19:14:45] INFO | >> train_qwenlatent.py:487 + Step 21740 | grad_norm_pre_clip=0.1761 | + grad_norm_pre_clip_avg=0.1757 | Metrics: + {'align_loss': 0.025038696825504303, + 'recon_loss': 0.052770186215639114, + 'predict_loss': 0.010785315185785294, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1761423498392105, + 'data_time': 0.0009308780136052519, + 'model_time': 1.4626780890102964, + 'grad_norm_pre_clip_avg': 0.17573801875114442, + 'learning_rate': 1.7399778664629568e-05, + 'epoch': 5.49} +04/19 [19:14:58] INFO | >> train_qwenlatent.py:487 + Step 21750 | grad_norm_pre_clip=0.1766 | + grad_norm_pre_clip_avg=0.1843 | Metrics: + {'align_loss': 0.024918202310800552, + 'recon_loss': 0.057772304862737656, + 'predict_loss': 0.004830158781260252, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17661602795124054, + 'mae_score': 0.010164977408744193, 'data_time': + 0.0009874730021692812, 'model_time': + 1.2105662349786144, 'grad_norm_pre_clip_avg': + 0.18432079255580902, 'learning_rate': + 1.7391754967009803e-05, 'epoch': 5.49} +04/19 [19:15:11] INFO | >> train_qwenlatent.py:487 + Step 21760 | grad_norm_pre_clip=0.1907 | + grad_norm_pre_clip_avg=0.1932 | Metrics: + {'align_loss': 0.024502471089363098, + 'recon_loss': 0.08132980018854141, + 'predict_loss': 0.008670898154377937, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19069364666938782, + 'data_time': 0.0006916399870533496, + 'model_time': 1.2381536710017826, + 'grad_norm_pre_clip_avg': 0.19319630563259124, + 'learning_rate': 1.738372889130045e-05, + 'epoch': 5.49} +04/19 [19:15:24] INFO | >> train_qwenlatent.py:487 + Step 21770 | grad_norm_pre_clip=0.1695 | + grad_norm_pre_clip_avg=0.2493 | Metrics: + {'align_loss': 0.02537413127720356, + 'recon_loss': 0.08905307203531265, + 'predict_loss': 0.009840716607868671, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1695471853017807, + 'data_time': 0.0007237270183395594, + 'model_time': 1.2106215770181734, + 'grad_norm_pre_clip_avg': 0.249260513484478, + 'learning_rate': 1.7375700441413308e-05, + 'epoch': 5.49} +04/19 [19:15:36] INFO | >> train_qwenlatent.py:487 + Step 21780 | grad_norm_pre_clip=0.1941 | + grad_norm_pre_clip_avg=0.2261 | Metrics: + {'align_loss': 0.025095351040363312, + 'recon_loss': 0.07757779210805893, + 'predict_loss': 0.01074900571256876, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.194096639752388, + 'data_time': 0.0007331170199904591, + 'model_time': 1.270552633999614, + 'grad_norm_pre_clip_avg': 0.22608462274074553, + 'learning_rate': 1.736766962126135e-05, + 'epoch': 5.5} +04/19 [19:15:49] INFO | >> train_qwenlatent.py:487 + Step 21790 | grad_norm_pre_clip=0.2317 | + grad_norm_pre_clip_avg=0.2024 | Metrics: + {'align_loss': 0.025561146438121796, + 'recon_loss': 0.08551502972841263, + 'predict_loss': 0.012452304363250732, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23167802393436432, + 'data_time': 0.0008358510094694793, + 'model_time': 1.2532403439981863, + 'grad_norm_pre_clip_avg': 0.2024141639471054, + 'learning_rate': 1.7359636434758702e-05, + 'epoch': 5.5} +04/19 [19:16:02] INFO | >> train_qwenlatent.py:487 + Step 21800 | grad_norm_pre_clip=0.1561 | + grad_norm_pre_clip_avg=0.1700 | Metrics: + {'align_loss': 0.02507109008729458, + 'recon_loss': 0.07247638702392578, + 'predict_loss': 0.010853219777345657, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15613773465156555, + 'mae_score': 0.009824160842208175, 'data_time': + 0.0006948520021978766, 'model_time': + 1.2119575519755017, 'grad_norm_pre_clip_avg': + 0.16998434513807298, 'learning_rate': + 1.7351600885820637e-05, 'epoch': 5.5} +04/19 [19:16:14] INFO | >> train_qwenlatent.py:487 + Step 21810 | grad_norm_pre_clip=0.1943 | + grad_norm_pre_clip_avg=0.1854 | Metrics: + {'align_loss': 0.025504928082227707, + 'recon_loss': 0.08681219071149826, + 'predict_loss': 0.010842830874025822, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19433310627937317, + 'data_time': 0.0009029840002767742, + 'model_time': 1.2364149089844432, + 'grad_norm_pre_clip_avg': 0.1853724554181099, + 'learning_rate': 1.734356297836359e-05, + 'epoch': 5.5} +04/19 [19:16:27] INFO | >> train_qwenlatent.py:487 + Step 21820 | grad_norm_pre_clip=0.1757 | + grad_norm_pre_clip_avg=0.1658 | Metrics: + {'align_loss': 0.02404934912919998, + 'recon_loss': 0.05956608057022095, + 'predict_loss': 0.015520140528678894, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17565800249576569, + 'data_time': 0.0009262940147891641, + 'model_time': 1.2272926769801416, + 'grad_norm_pre_clip_avg': 0.16582050323486328, + 'learning_rate': 1.733552271630513e-05, + 'epoch': 5.51} +04/19 [19:16:39] INFO | >> train_qwenlatent.py:487 + Step 21830 | grad_norm_pre_clip=0.3528 | + grad_norm_pre_clip_avg=0.2706 | Metrics: + {'align_loss': 0.02551039680838585, + 'recon_loss': 0.07824306935071945, + 'predict_loss': 0.010246376506984234, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.35278207063674927, + 'data_time': 0.000683452992234379, + 'model_time': 1.2363184450077824, + 'grad_norm_pre_clip_avg': 0.2706386595964432, + 'learning_rate': 1.7327480103563993e-05, + 'epoch': 5.51} +04/19 [19:16:52] INFO | >> train_qwenlatent.py:487 + Step 21840 | grad_norm_pre_clip=0.1802 | + grad_norm_pre_clip_avg=0.2102 | Metrics: + {'align_loss': 0.02530517801642418, + 'recon_loss': 0.07871212065219879, + 'predict_loss': 0.009992249310016632, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18018481135368347, + 'data_time': 0.0007211360207293183, + 'model_time': 1.3119781310088001, + 'grad_norm_pre_clip_avg': 0.2101574048399925, + 'learning_rate': 1.731943514406004e-05, + 'epoch': 5.51} +04/19 [19:17:05] INFO | >> train_qwenlatent.py:487 + Step 21850 | grad_norm_pre_clip=0.1687 | + grad_norm_pre_clip_avg=0.1871 | Metrics: + {'align_loss': 0.025777798146009445, + 'recon_loss': 0.07226155698299408, + 'predict_loss': 0.01698322594165802, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1686602234840393, + 'mae_score': 0.010289015211500564, 'data_time': + 0.0007634669891558588, 'model_time': + 1.5965883979806677, 'grad_norm_pre_clip_avg': + 0.18713420033454894, 'learning_rate': + 1.7311387841714298e-05, 'epoch': 5.51} +04/19 [19:17:18] INFO | >> train_qwenlatent.py:487 + Step 21860 | grad_norm_pre_clip=0.1801 | + grad_norm_pre_clip_avg=0.1640 | Metrics: + {'align_loss': 0.024968765676021576, + 'recon_loss': 0.09856762737035751, + 'predict_loss': 0.014630943536758423, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1800512969493866, + 'data_time': 0.0007436209998559207, + 'model_time': 1.231150523002725, + 'grad_norm_pre_clip_avg': 0.16399166136980056, + 'learning_rate': 1.7303338200448922e-05, + 'epoch': 5.52} +04/19 [19:17:31] INFO | >> train_qwenlatent.py:487 + Step 21870 | grad_norm_pre_clip=0.1815 | + grad_norm_pre_clip_avg=0.1972 | Metrics: + {'align_loss': 0.025553550571203232, + 'recon_loss': 0.102747842669487, + 'predict_loss': 0.012539152055978775, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18154658377170563, + 'data_time': 0.0010509490093681961, + 'model_time': 1.2452431170095224, + 'grad_norm_pre_clip_avg': 0.1971580296754837, + 'learning_rate': 1.7295286224187204e-05, + 'epoch': 5.52} +04/19 [19:17:44] INFO | >> train_qwenlatent.py:487 + Step 21880 | grad_norm_pre_clip=0.2576 | + grad_norm_pre_clip_avg=0.2266 | Metrics: + {'align_loss': 0.025272976607084274, + 'recon_loss': 0.09082284569740295, + 'predict_loss': 0.008182131685316563, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2575943171977997, + 'data_time': 0.0008276999869849533, + 'model_time': 1.207367285998771, + 'grad_norm_pre_clip_avg': 0.22662037014961242, + 'learning_rate': 1.7287231916853583e-05, + 'epoch': 5.52} +04/19 [19:17:56] INFO | >> train_qwenlatent.py:487 + Step 21890 | grad_norm_pre_clip=0.1993 | + grad_norm_pre_clip_avg=0.2224 | Metrics: + {'align_loss': 0.02534550055861473, + 'recon_loss': 0.07142169028520584, + 'predict_loss': 0.012288624420762062, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19926977157592773, + 'data_time': 0.0007398529851343483, + 'model_time': 1.2076863679976668, + 'grad_norm_pre_clip_avg': 0.22240796387195588, + 'learning_rate': 1.7279175282373626e-05, + 'epoch': 5.52} +04/19 [19:18:09] INFO | >> train_qwenlatent.py:487 + Step 21900 | grad_norm_pre_clip=0.1763 | + grad_norm_pre_clip_avg=0.1962 | Metrics: + {'align_loss': 0.025897316634655, 'recon_loss': + 0.09644386917352676, 'predict_loss': + 0.01267241220921278, 'aux_loss_decay_weight': + 0.0, 'grad_norm_pre_clip': 0.1763329654932022, + 'mae_score': 0.012320675721039643, 'data_time': + 0.0010316059924662113, 'model_time': + 1.1964161650103051, 'grad_norm_pre_clip_avg': + 0.19620669037103652, 'learning_rate': + 1.727111632467405e-05, 'epoch': 5.53} +04/19 [19:18:22] INFO | >> train_qwenlatent.py:487 + Step 21910 | grad_norm_pre_clip=0.1875 | + grad_norm_pre_clip_avg=0.2060 | Metrics: + {'align_loss': 0.025517351925373077, + 'recon_loss': 0.09917201846837997, + 'predict_loss': 0.01184769906103611, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18752677738666534, + 'data_time': 0.0009818620164878666, + 'model_time': 1.2297187660005875, + 'grad_norm_pre_clip_avg': 0.20595772713422775, + 'learning_rate': 1.726305504768268e-05, + 'epoch': 5.53} +04/19 [19:18:34] INFO | >> train_qwenlatent.py:487 + Step 21920 | grad_norm_pre_clip=0.1685 | + grad_norm_pre_clip_avg=0.1999 | Metrics: + {'align_loss': 0.025186415761709213, + 'recon_loss': 0.07019443809986115, + 'predict_loss': 0.008699961937963963, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16851958632469177, + 'data_time': 0.0006393149960786104, + 'model_time': 1.2387591239821631, + 'grad_norm_pre_clip_avg': 0.19991426467895507, + 'learning_rate': 1.7254991455328492e-05, + 'epoch': 5.53} +04/19 [19:18:47] INFO | >> train_qwenlatent.py:487 + Step 21930 | grad_norm_pre_clip=0.4697 | + grad_norm_pre_clip_avg=0.2374 | Metrics: + {'align_loss': 0.02592989057302475, + 'recon_loss': 0.07990048080682755, + 'predict_loss': 0.014160550199449062, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.46968719363212585, + 'data_time': 0.0006787950114812702, + 'model_time': 1.2999029610073194, + 'grad_norm_pre_clip_avg': 0.2373933345079422, + 'learning_rate': 1.724692555154158e-05, + 'epoch': 5.53} +04/19 [19:18:59] INFO | >> train_qwenlatent.py:487 + Step 21940 | grad_norm_pre_clip=0.2791 | + grad_norm_pre_clip_avg=0.2327 | Metrics: + {'align_loss': 0.025589238852262497, + 'recon_loss': 0.07874642312526703, + 'predict_loss': 0.012667069211602211, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.27912622690200806, + 'data_time': 0.0009329860040452331, + 'model_time': 1.2320551419979893, + 'grad_norm_pre_clip_avg': 0.23269331753253936, + 'learning_rate': 1.723885734025317e-05, + 'epoch': 5.54} +04/19 [19:19:13] INFO | >> train_qwenlatent.py:487 + Step 21950 | grad_norm_pre_clip=0.2128 | + grad_norm_pre_clip_avg=0.2146 | Metrics: + {'align_loss': 0.02457933872938156, + 'recon_loss': 0.07844061404466629, + 'predict_loss': 0.010012459941208363, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21283385157585144, + 'mae_score': 0.009747433018040012, 'data_time': + 0.0009001229773275554, 'model_time': + 1.2740966419805773, 'grad_norm_pre_clip_avg': + 0.2146462991833687, 'learning_rate': + 1.723078682539561e-05, 'epoch': 5.54} +04/19 [19:19:25] INFO | >> train_qwenlatent.py:487 + Step 21960 | grad_norm_pre_clip=0.1795 | + grad_norm_pre_clip_avg=0.1934 | Metrics: + {'align_loss': 0.025142572820186615, + 'recon_loss': 0.06327879428863525, + 'predict_loss': 0.00804875697940588, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1795312762260437, + 'data_time': 0.0009655279864091426, + 'model_time': 1.2504085380060133, + 'grad_norm_pre_clip_avg': 0.19335523545742034, + 'learning_rate': 1.7222714010902366e-05, + 'epoch': 5.54} +04/19 [19:19:37] INFO | >> train_qwenlatent.py:487 + Step 21970 | grad_norm_pre_clip=0.1692 | + grad_norm_pre_clip_avg=0.1870 | Metrics: + {'align_loss': 0.02502434328198433, + 'recon_loss': 0.05383162572979927, + 'predict_loss': 0.008713067509233952, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16918820142745972, + 'data_time': 0.0006579300097655505, + 'model_time': 1.2193880569830071, + 'grad_norm_pre_clip_avg': 0.18703466206789016, + 'learning_rate': 1.721463890070804e-05, + 'epoch': 5.54} +04/19 [19:19:50] INFO | >> train_qwenlatent.py:487 + Step 21980 | grad_norm_pre_clip=0.2067 | + grad_norm_pre_clip_avg=0.1846 | Metrics: + {'align_loss': 0.0232468843460083, + 'recon_loss': 0.06579183787107468, + 'predict_loss': 0.009579270146787167, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20668207108974457, + 'data_time': 0.0007773420074954629, + 'model_time': 1.2182469030085485, + 'grad_norm_pre_clip_avg': 0.18455667048692703, + 'learning_rate': 1.7206561498748326e-05, + 'epoch': 5.55} +04/19 [19:20:03] INFO | >> train_qwenlatent.py:487 + Step 21990 | grad_norm_pre_clip=0.1802 | + grad_norm_pre_clip_avg=0.2525 | Metrics: + {'align_loss': 0.025933634489774704, + 'recon_loss': 0.09386847168207169, + 'predict_loss': 0.009246948175132275, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18021328747272491, + 'data_time': 0.0007858519966248423, + 'model_time': 1.3195066429907456, + 'grad_norm_pre_clip_avg': 0.25247555673122407, + 'learning_rate': 1.719848180896007e-05, + 'epoch': 5.55} +04/19 [19:20:17] INFO | >> train_qwenlatent.py:487 + Step 22000 | grad_norm_pre_clip=0.2610 | + grad_norm_pre_clip_avg=0.2251 | Metrics: + {'align_loss': 0.02442818135023117, + 'recon_loss': 0.07027938961982727, + 'predict_loss': 0.00904091540724039, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2609712779521942, + 'mae_score': 0.012781607138144004, 'data_time': + 0.0009759220120031387, 'model_time': + 1.5461143949942198, 'grad_norm_pre_clip_avg': + 0.22513952404260634, 'learning_rate': + 1.7190399835281203e-05, 'epoch': 5.55} +04/19 [19:20:30] INFO | >> train_qwenlatent.py:487 + Step 22010 | grad_norm_pre_clip=0.1608 | + grad_norm_pre_clip_avg=0.1969 | Metrics: + {'align_loss': 0.024841617792844772, + 'recon_loss': 0.06380487233400345, + 'predict_loss': 0.007964196614921093, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16075091063976288, + 'data_time': 0.0013212510093580931, + 'model_time': 1.2381462169869337, + 'grad_norm_pre_clip_avg': 0.1969102293252945, + 'learning_rate': 1.7182315581650783e-05, + 'epoch': 5.55} +04/19 [19:20:42] INFO | >> train_qwenlatent.py:487 + Step 22020 | grad_norm_pre_clip=0.1414 | + grad_norm_pre_clip_avg=0.1593 | Metrics: + {'align_loss': 0.024767577648162842, + 'recon_loss': 0.06893327087163925, + 'predict_loss': 0.007654900662600994, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1414446383714676, + 'data_time': 0.0014424620021600276, + 'model_time': 1.2278756549931131, + 'grad_norm_pre_clip_avg': 0.15934776067733764, + 'learning_rate': 1.717422905200898e-05, + 'epoch': 5.56} +04/19 [19:20:55] INFO | >> train_qwenlatent.py:487 + Step 22030 | grad_norm_pre_clip=0.1766 | + grad_norm_pre_clip_avg=0.1618 | Metrics: + {'align_loss': 0.025774644687771797, + 'recon_loss': 0.0756526067852974, + 'predict_loss': 0.010158884339034557, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1766398549079895, + 'data_time': 0.0011815589969046414, + 'model_time': 1.1926807959971484, + 'grad_norm_pre_clip_avg': 0.16175364851951599, + 'learning_rate': 1.7166140250297068e-05, + 'epoch': 5.56} +04/19 [19:21:07] INFO | >> train_qwenlatent.py:487 + Step 22040 | grad_norm_pre_clip=0.2015 | + grad_norm_pre_clip_avg=0.2309 | Metrics: + {'align_loss': 0.02464446611702442, + 'recon_loss': 0.07341540604829788, + 'predict_loss': 0.011816667392849922, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20149917900562286, + 'data_time': 0.0009679610084276646, + 'model_time': 1.2685975600033998, + 'grad_norm_pre_clip_avg': 0.2308911129832268, + 'learning_rate': 1.7158049180457433e-05, + 'epoch': 5.56} +04/19 [19:21:20] INFO | >> train_qwenlatent.py:487 + Step 22050 | grad_norm_pre_clip=0.2352 | + grad_norm_pre_clip_avg=0.2088 | Metrics: + {'align_loss': 0.02555180713534355, + 'recon_loss': 0.08991871029138565, + 'predict_loss': 0.01180796418339014, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.235207200050354, + 'mae_score': 0.009992626121452262, 'data_time': + 0.0010973600146826357, 'model_time': + 1.255098272988107, 'grad_norm_pre_clip_avg': + 0.20881913155317305, 'learning_rate': + 1.714995584643356e-05, 'epoch': 5.56} +04/19 [19:21:33] INFO | >> train_qwenlatent.py:487 + Step 22060 | grad_norm_pre_clip=0.2413 | + grad_norm_pre_clip_avg=0.1871 | Metrics: + {'align_loss': 0.026279063895344734, + 'recon_loss': 0.09194489568471909, + 'predict_loss': 0.012087731622159481, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24133487045764923, + 'data_time': 0.0006656270124949515, + 'model_time': 1.2751246780098882, + 'grad_norm_pre_clip_avg': 0.18710518330335618, + 'learning_rate': 1.7141860252170043e-05, + 'epoch': 5.57} +04/19 [19:21:46] INFO | >> train_qwenlatent.py:487 + Step 22070 | grad_norm_pre_clip=0.2368 | + grad_norm_pre_clip_avg=0.2031 | Metrics: + {'align_loss': 0.024687759578227997, + 'recon_loss': 0.0882364809513092, + 'predict_loss': 0.012613963335752487, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23680534958839417, + 'data_time': 0.0010823410120792687, + 'model_time': 1.2484265089733526, + 'grad_norm_pre_clip_avg': 0.20310228168964387, + 'learning_rate': 1.713376240161258e-05, + 'epoch': 5.57} +04/19 [19:21:58] INFO | >> train_qwenlatent.py:487 + Step 22080 | grad_norm_pre_clip=0.1522 | + grad_norm_pre_clip_avg=0.1871 | Metrics: + {'align_loss': 0.025411058217287064, + 'recon_loss': 0.097118079662323, + 'predict_loss': 0.00936865247786045, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15217001736164093, + 'data_time': 0.0007572750037070364, + 'model_time': 1.263201077003032, + 'grad_norm_pre_clip_avg': 0.1871277153491974, + 'learning_rate': 1.7125662298707955e-05, + 'epoch': 5.57} +04/19 [19:22:11] INFO | >> train_qwenlatent.py:487 + Step 22090 | grad_norm_pre_clip=0.2290 | + grad_norm_pre_clip_avg=0.1904 | Metrics: + {'align_loss': 0.025357242673635483, + 'recon_loss': 0.0725361704826355, + 'predict_loss': 0.011029087007045746, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22897237539291382, + 'data_time': 0.001290077023440972, + 'model_time': 1.2658470309979748, + 'grad_norm_pre_clip_avg': 0.19040842354297638, + 'learning_rate': 1.7117559947404073e-05, + 'epoch': 5.57} +04/19 [19:22:23] INFO | >> train_qwenlatent.py:487 + Step 22100 | grad_norm_pre_clip=0.2552 | + grad_norm_pre_clip_avg=0.2250 | Metrics: + {'align_loss': 0.025502946227788925, + 'recon_loss': 0.05482400581240654, + 'predict_loss': 0.0072675906121730804, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2551644742488861, + 'mae_score': 0.01361595360008446, 'data_time': + 0.0011585329775698483, 'model_time': + 1.2174449340091087, 'grad_norm_pre_clip_avg': + 0.22495021522045136, 'learning_rate': + 1.710945535164992e-05, 'epoch': 5.58} +04/19 [19:22:36] INFO | >> train_qwenlatent.py:487 + Step 22110 | grad_norm_pre_clip=0.2064 | + grad_norm_pre_clip_avg=0.2300 | Metrics: + {'align_loss': 0.025520704686641693, + 'recon_loss': 0.06697990000247955, + 'predict_loss': 0.011855482123792171, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20641198754310608, + 'data_time': 0.0008647240174468607, + 'model_time': 1.2735037830134388, + 'grad_norm_pre_clip_avg': 0.22995123118162156, + 'learning_rate': 1.7101348515395567e-05, + 'epoch': 5.58} +04/19 [19:22:49] INFO | >> train_qwenlatent.py:487 + Step 22120 | grad_norm_pre_clip=0.1767 | + grad_norm_pre_clip_avg=0.2014 | Metrics: + {'align_loss': 0.025928158313035965, + 'recon_loss': 0.08001921325922012, + 'predict_loss': 0.014523235149681568, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17669063806533813, + 'data_time': 0.0008149830100592226, + 'model_time': 1.246094845002517, + 'grad_norm_pre_clip_avg': 0.20144928693771363, + 'learning_rate': 1.7093239442592192e-05, + 'epoch': 5.58} +04/19 [19:23:02] INFO | >> train_qwenlatent.py:487 + Step 22130 | grad_norm_pre_clip=0.2418 | + grad_norm_pre_clip_avg=0.1878 | Metrics: + {'align_loss': 0.025527818128466606, + 'recon_loss': 0.07006523013114929, + 'predict_loss': 0.008327903226017952, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24178767204284668, + 'data_time': 0.0007341130112763494, + 'model_time': 1.5456459390115924, + 'grad_norm_pre_clip_avg': 0.18779942244291306, + 'learning_rate': 1.7085128137192063e-05, + 'epoch': 5.58} +04/19 [19:23:15] INFO | >> train_qwenlatent.py:487 + Step 22140 | grad_norm_pre_clip=0.2493 | + grad_norm_pre_clip_avg=0.2523 | Metrics: + {'align_loss': 0.02495817095041275, + 'recon_loss': 0.05871685966849327, + 'predict_loss': 0.009006709791719913, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24934837222099304, + 'data_time': 0.0006644100067205727, + 'model_time': 1.2119562609877903, + 'grad_norm_pre_clip_avg': 0.25226413160562516, + 'learning_rate': 1.7077014603148527e-05, + 'epoch': 5.59} +04/19 [19:23:28] INFO | >> train_qwenlatent.py:487 + Step 22150 | grad_norm_pre_clip=0.1821 | + grad_norm_pre_clip_avg=0.2045 | Metrics: + {'align_loss': 0.023836221545934677, + 'recon_loss': 0.0687868744134903, + 'predict_loss': 0.014254453592002392, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18206706643104553, + 'mae_score': 0.013838056615881018, 'data_time': + 0.0007985650154296309, 'model_time': + 1.2518410469929222, 'grad_norm_pre_clip_avg': + 0.2044943928718567, 'learning_rate': + 1.7068898844416022e-05, 'epoch': 5.59} +04/19 [19:23:41] INFO | >> train_qwenlatent.py:487 + Step 22160 | grad_norm_pre_clip=0.2467 | + grad_norm_pre_clip_avg=0.2009 | Metrics: + {'align_loss': 0.024954812601208687, + 'recon_loss': 0.07852192223072052, + 'predict_loss': 0.007833785377442837, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2467036098241806, + 'data_time': 0.0009826509922277182, + 'model_time': 1.2396428610081784, + 'grad_norm_pre_clip_avg': 0.20091244131326674, + 'learning_rate': 1.706078086495008e-05, + 'epoch': 5.59} +04/19 [19:23:53] INFO | >> train_qwenlatent.py:487 + Step 22170 | grad_norm_pre_clip=0.1510 | + grad_norm_pre_clip_avg=0.1902 | Metrics: + {'align_loss': 0.024472083896398544, + 'recon_loss': 0.0529155395925045, + 'predict_loss': 0.008082716725766659, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15100258588790894, + 'data_time': 0.0009353290079161525, + 'model_time': 1.2295133299776353, + 'grad_norm_pre_clip_avg': 0.1901577368378639, + 'learning_rate': 1.7052660668707295e-05, + 'epoch': 5.59} +04/19 [19:24:06] INFO | >> train_qwenlatent.py:487 + Step 22180 | grad_norm_pre_clip=0.1822 | + grad_norm_pre_clip_avg=0.1689 | Metrics: + {'align_loss': 0.025553448125720024, + 'recon_loss': 0.06568901985883713, + 'predict_loss': 0.007832069881260395, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1822260171175003, + 'data_time': 0.0013905679807066917, + 'model_time': 1.1952838969882578, + 'grad_norm_pre_clip_avg': 0.16886794269084932, + 'learning_rate': 1.704453825964535e-05, + 'epoch': 5.6} +04/19 [19:24:18] INFO | >> train_qwenlatent.py:487 + Step 22190 | grad_norm_pre_clip=0.3041 | + grad_norm_pre_clip_avg=0.2496 | Metrics: + {'align_loss': 0.02546103671193123, + 'recon_loss': 0.0700235441327095, + 'predict_loss': 0.00878459308296442, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3041490614414215, + 'data_time': 0.0006878089916426688, + 'model_time': 1.1998105029924773, + 'grad_norm_pre_clip_avg': 0.2496099889278412, + 'learning_rate': 1.7036413641723014e-05, + 'epoch': 5.6} +04/19 [19:24:31] INFO | >> train_qwenlatent.py:487 + Step 22200 | grad_norm_pre_clip=0.1708 | + grad_norm_pre_clip_avg=0.2204 | Metrics: + {'align_loss': 0.024920044466853142, + 'recon_loss': 0.07019342482089996, + 'predict_loss': 0.0078027211129665375, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17077866196632385, + 'mae_score': 0.011607962256079322, 'data_time': + 0.0007145090203266591, 'model_time': + 1.2080321180110332, 'grad_norm_pre_clip_avg': + 0.22044347673654557, 'learning_rate': + 1.7028286818900126e-05, 'epoch': 5.6} +04/19 [19:24:43] INFO | >> train_qwenlatent.py:487 + Step 22210 | grad_norm_pre_clip=0.1937 | + grad_norm_pre_clip_avg=0.1859 | Metrics: + {'align_loss': 0.02548869326710701, + 'recon_loss': 0.09017328917980194, + 'predict_loss': 0.010164624080061913, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1936764419078827, + 'data_time': 0.0006690399895887822, + 'model_time': 1.1870761729951482, + 'grad_norm_pre_clip_avg': 0.1859207794070244, + 'learning_rate': 1.7020157795137606e-05, + 'epoch': 5.6} +04/19 [19:24:56] INFO | >> train_qwenlatent.py:487 + Step 22220 | grad_norm_pre_clip=0.1486 | + grad_norm_pre_clip_avg=0.1913 | Metrics: + {'align_loss': 0.02615594118833542, + 'recon_loss': 0.07471963763237, 'predict_loss': + 0.007587910629808903, 'aux_loss_decay_weight': + 0.0, 'grad_norm_pre_clip': 0.14856155216693878, + 'data_time': 0.000990859989542514, + 'model_time': 1.2636819460021798, + 'grad_norm_pre_clip_avg': 0.19130227118730544, + 'learning_rate': 1.7012026574397433e-05, + 'epoch': 5.61} +04/19 [19:25:08] INFO | >> train_qwenlatent.py:487 + Step 22230 | grad_norm_pre_clip=0.1633 | + grad_norm_pre_clip_avg=0.2110 | Metrics: + {'align_loss': 0.025276750326156616, + 'recon_loss': 0.08608600497245789, + 'predict_loss': 0.010960446670651436, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1632707417011261, + 'data_time': 0.0009807129972614348, + 'model_time': 1.2726077699917369, + 'grad_norm_pre_clip_avg': 0.21101366579532624, + 'learning_rate': 1.7003893160642667e-05, + 'epoch': 5.61} +04/19 [19:25:21] INFO | >> train_qwenlatent.py:487 + Step 22240 | grad_norm_pre_clip=0.2570 | + grad_norm_pre_clip_avg=0.1885 | Metrics: + {'align_loss': 0.025837916880846024, + 'recon_loss': 0.080296590924263, + 'predict_loss': 0.00796987023204565, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2569652199745178, + 'data_time': 0.0008442820108029991, + 'model_time': 1.2096003399929032, + 'grad_norm_pre_clip_avg': 0.1885337710380554, + 'learning_rate': 1.6995757557837437e-05, + 'epoch': 5.61} +04/19 [19:25:34] INFO | >> train_qwenlatent.py:487 + Step 22250 | grad_norm_pre_clip=0.1669 | + grad_norm_pre_clip_avg=0.2284 | Metrics: + {'align_loss': 0.025582825765013695, + 'recon_loss': 0.09718592464923859, + 'predict_loss': 0.020472882315516472, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1668992042541504, + 'mae_score': 0.01114878697438283, 'data_time': + 0.0010406060027889907, 'model_time': + 1.1963962280133273, 'grad_norm_pre_clip_avg': + 0.22839654684066774, 'learning_rate': + 1.6987619769946937e-05, 'epoch': 5.61} +04/19 [19:25:47] INFO | >> train_qwenlatent.py:487 + Step 22260 | grad_norm_pre_clip=0.1874 | + grad_norm_pre_clip_avg=0.2016 | Metrics: + {'align_loss': 0.02536354586482048, + 'recon_loss': 0.06841975450515747, + 'predict_loss': 0.008080472238361835, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1874210238456726, + 'data_time': 0.0007320290023926646, + 'model_time': 1.194120114989346, + 'grad_norm_pre_clip_avg': 0.20157298892736436, + 'learning_rate': 1.697947980093743e-05, + 'epoch': 5.62} +04/19 [19:26:00] INFO | >> train_qwenlatent.py:487 + Step 22270 | grad_norm_pre_clip=0.1593 | + grad_norm_pre_clip_avg=0.1850 | Metrics: + {'align_loss': 0.025544360280036926, + 'recon_loss': 0.07152397930622101, + 'predict_loss': 0.007781944703310728, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1593184918165207, + 'data_time': 0.0010139959922526032, + 'model_time': 1.5624445010034833, + 'grad_norm_pre_clip_avg': 0.18501706719398497, + 'learning_rate': 1.6971337654776227e-05, + 'epoch': 5.62} +04/19 [19:26:12] INFO | >> train_qwenlatent.py:487 + Step 22280 | grad_norm_pre_clip=0.2404 | + grad_norm_pre_clip_avg=0.1987 | Metrics: + {'align_loss': 0.024778947234153748, + 'recon_loss': 0.0793207585811615, + 'predict_loss': 0.009283983148634434, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24038220942020416, + 'data_time': 0.0007013060094323009, + 'model_time': 1.235418808995746, + 'grad_norm_pre_clip_avg': 0.19866928905248643, + 'learning_rate': 1.6963193335431716e-05, + 'epoch': 5.62} +04/19 [19:26:25] INFO | >> train_qwenlatent.py:487 + Step 22290 | grad_norm_pre_clip=0.1597 | + grad_norm_pre_clip_avg=0.1757 | Metrics: + {'align_loss': 0.025082439184188843, + 'recon_loss': 0.08624085783958435, + 'predict_loss': 0.010981973260641098, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1596696525812149, + 'data_time': 0.0009200559870805591, + 'model_time': 1.2107243229984306, + 'grad_norm_pre_clip_avg': 0.17574934363365174, + 'learning_rate': 1.6955046846873348e-05, + 'epoch': 5.62} +04/19 [19:26:38] INFO | >> train_qwenlatent.py:487 + Step 22300 | grad_norm_pre_clip=0.2497 | + grad_norm_pre_clip_avg=0.1970 | Metrics: + {'align_loss': 0.025620335713028908, + 'recon_loss': 0.07012796401977539, + 'predict_loss': 0.012905856594443321, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24966172873973846, + 'mae_score': 0.01097102637763496, 'data_time': + 0.0010280809947289526, 'model_time': + 1.192472483991878, 'grad_norm_pre_clip_avg': + 0.19698172062635422, 'learning_rate': + 1.6946898193071616e-05, 'epoch': 5.63} +04/19 [19:26:51] INFO | >> train_qwenlatent.py:487 + Step 22310 | grad_norm_pre_clip=0.1885 | + grad_norm_pre_clip_avg=0.2388 | Metrics: + {'align_loss': 0.026125386357307434, + 'recon_loss': 0.08778894692659378, + 'predict_loss': 0.008720539510250092, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1884680986404419, + 'data_time': 0.000729447026969865, + 'model_time': 1.263964362005936, + 'grad_norm_pre_clip_avg': 0.23881253302097322, + 'learning_rate': 1.6938747377998068e-05, + 'epoch': 5.63} +04/19 [19:27:04] INFO | >> train_qwenlatent.py:487 + Step 22320 | grad_norm_pre_clip=0.2039 | + grad_norm_pre_clip_avg=0.1947 | Metrics: + {'align_loss': 0.026396334171295166, + 'recon_loss': 0.104632169008255, + 'predict_loss': 0.011913138441741467, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20390675961971283, + 'data_time': 0.0008949869952630252, + 'model_time': 1.1939906839979813, + 'grad_norm_pre_clip_avg': 0.19472316652536392, + 'learning_rate': 1.693059440562532e-05, + 'epoch': 5.63} +04/19 [19:27:17] INFO | >> train_qwenlatent.py:487 + Step 22330 | grad_norm_pre_clip=0.1619 | + grad_norm_pre_clip_avg=0.1778 | Metrics: + {'align_loss': 0.02588990144431591, + 'recon_loss': 0.08079647272825241, + 'predict_loss': 0.010013462975621223, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16193382441997528, + 'data_time': 0.001122607005527243, + 'model_time': 1.291091905994108, + 'grad_norm_pre_clip_avg': 0.17779222577810289, + 'learning_rate': 1.6922439279927034e-05, + 'epoch': 5.63} +04/19 [19:27:29] INFO | >> train_qwenlatent.py:487 + Step 22340 | grad_norm_pre_clip=0.1432 | + grad_norm_pre_clip_avg=0.1853 | Metrics: + {'align_loss': 0.024547208100557327, + 'recon_loss': 0.05637338384985924, + 'predict_loss': 0.007820227183401585, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1432102769613266, + 'data_time': 0.000739345996407792, + 'model_time': 1.288547962991288, + 'grad_norm_pre_clip_avg': 0.18526743948459626, + 'learning_rate': 1.691428200487791e-05, + 'epoch': 5.64} +04/19 [19:27:42] INFO | >> train_qwenlatent.py:487 + Step 22350 | grad_norm_pre_clip=0.2631 | + grad_norm_pre_clip_avg=0.2098 | Metrics: + {'align_loss': 0.025067469105124474, + 'recon_loss': 0.09219332784414291, + 'predict_loss': 0.011033180169761181, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26312461495399475, + 'mae_score': 0.010084044825923335, 'data_time': + 0.0006985660002101213, 'model_time': + 1.2102902329934295, 'grad_norm_pre_clip_avg': + 0.20976182222366332, 'learning_rate': + 1.6906122584453714e-05, 'epoch': 5.64} +04/19 [19:27:55] INFO | >> train_qwenlatent.py:487 + Step 22360 | grad_norm_pre_clip=0.1970 | + grad_norm_pre_clip_avg=0.1782 | Metrics: + {'align_loss': 0.024685081094503403, + 'recon_loss': 0.10445515811443329, + 'predict_loss': 0.016003932803869247, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19701340794563293, + 'data_time': 0.0009195939928758889, + 'model_time': 1.2427101270004641, + 'grad_norm_pre_clip_avg': 0.1781759038567543, + 'learning_rate': 1.6897961022631246e-05, + 'epoch': 5.64} +04/19 [19:28:07] INFO | >> train_qwenlatent.py:487 + Step 22370 | grad_norm_pre_clip=0.1586 | + grad_norm_pre_clip_avg=0.1761 | Metrics: + {'align_loss': 0.02519959583878517, + 'recon_loss': 0.1129496842622757, + 'predict_loss': 0.014631946571171284, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1586417555809021, + 'data_time': 0.0007381860050372779, + 'model_time': 1.2982618540117983, + 'grad_norm_pre_clip_avg': 0.17609095126390456, + 'learning_rate': 1.6889797323388347e-05, + 'epoch': 5.64} +04/19 [19:28:20] INFO | >> train_qwenlatent.py:487 + Step 22380 | grad_norm_pre_clip=0.1734 | + grad_norm_pre_clip_avg=0.2120 | Metrics: + {'align_loss': 0.026078345254063606, + 'recon_loss': 0.09827492386102676, + 'predict_loss': 0.012294458225369453, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17337633669376373, + 'data_time': 0.0009693489992059767, + 'model_time': 1.2922898060060106, + 'grad_norm_pre_clip_avg': 0.21202159821987152, + 'learning_rate': 1.6881631490703916e-05, + 'epoch': 5.65} +04/19 [19:28:33] INFO | >> train_qwenlatent.py:487 + Step 22390 | grad_norm_pre_clip=0.2306 | + grad_norm_pre_clip_avg=0.1775 | Metrics: + {'align_loss': 0.02594779059290886, + 'recon_loss': 0.08953510969877243, + 'predict_loss': 0.008883344940841198, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23061715066432953, + 'data_time': 0.0009050119842868298, + 'model_time': 1.271349440998165, + 'grad_norm_pre_clip_avg': 0.17750765681266784, + 'learning_rate': 1.6873463528557865e-05, + 'epoch': 5.65} +04/19 [19:28:46] INFO | >> train_qwenlatent.py:487 + Step 22400 | grad_norm_pre_clip=0.1815 | + grad_norm_pre_clip_avg=0.2037 | Metrics: + {'align_loss': 0.02462003193795681, + 'recon_loss': 0.06670787185430527, + 'predict_loss': 0.012546105310320854, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18151788413524628, + 'mae_score': 0.012093501048045116, 'data_time': + 0.0008574960229452699, 'model_time': + 1.2129321219981648, 'grad_norm_pre_clip_avg': + 0.20374124199151994, 'learning_rate': + 1.686529344093117e-05, 'epoch': 5.65} +04/19 [19:28:59] INFO | >> train_qwenlatent.py:487 + Step 22410 | grad_norm_pre_clip=0.2315 | + grad_norm_pre_clip_avg=0.1961 | Metrics: + {'align_loss': 0.025524776428937912, + 'recon_loss': 0.08256624639034271, + 'predict_loss': 0.010631751269102097, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23147355020046234, + 'data_time': 0.0008103600121103227, + 'model_time': 1.217967344011413, + 'grad_norm_pre_clip_avg': 0.196074940264225, + 'learning_rate': 1.685712123180583e-05, + 'epoch': 5.65} +04/19 [19:29:11] INFO | >> train_qwenlatent.py:487 + Step 22420 | grad_norm_pre_clip=0.1943 | + grad_norm_pre_clip_avg=0.2042 | Metrics: + {'align_loss': 0.023561015725135803, + 'recon_loss': 0.0623323880136013, + 'predict_loss': 0.010902666486799717, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19429618120193481, + 'data_time': 0.001046302990289405, + 'model_time': 1.2529240339936223, + 'grad_norm_pre_clip_avg': 0.20419083088636397, + 'learning_rate': 1.684894690516488e-05, + 'epoch': 5.66} +04/19 [19:29:24] INFO | >> train_qwenlatent.py:487 + Step 22430 | grad_norm_pre_clip=0.1533 | + grad_norm_pre_clip_avg=0.2005 | Metrics: + {'align_loss': 0.025053372606635094, + 'recon_loss': 0.0736820250749588, + 'predict_loss': 0.011703405529260635, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15334565937519073, + 'data_time': 0.0009670720028225332, + 'model_time': 1.2270835369999986, + 'grad_norm_pre_clip_avg': 0.20050884336233138, + 'learning_rate': 1.6840770464992395e-05, + 'epoch': 5.66} +04/19 [19:29:36] INFO | >> train_qwenlatent.py:487 + Step 22440 | grad_norm_pre_clip=0.1804 | + grad_norm_pre_clip_avg=0.1929 | Metrics: + {'align_loss': 0.02634415775537491, + 'recon_loss': 0.08241873979568481, + 'predict_loss': 0.010563770309090614, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18035855889320374, + 'data_time': 0.0012988689995836467, + 'model_time': 1.240979370981222, + 'grad_norm_pre_clip_avg': 0.19288848340511322, + 'learning_rate': 1.6832591915273456e-05, + 'epoch': 5.66} +04/19 [19:29:50] INFO | >> train_qwenlatent.py:487 + Step 22450 | grad_norm_pre_clip=0.1544 | + grad_norm_pre_clip_avg=0.1813 | Metrics: + {'align_loss': 0.025632359087467194, + 'recon_loss': 0.0970173329114914, + 'predict_loss': 0.011620222590863705, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1544032096862793, + 'mae_score': 0.014433094402691266, 'data_time': + 0.000688384025124833, 'model_time': + 1.1958751759957522, 'grad_norm_pre_clip_avg': + 0.1813478097319603, 'learning_rate': + 1.6824411259994203e-05, 'epoch': 5.66} +04/19 [19:30:03] INFO | >> train_qwenlatent.py:487 + Step 22460 | grad_norm_pre_clip=0.2287 | + grad_norm_pre_clip_avg=0.2037 | Metrics: + {'align_loss': 0.023871436715126038, + 'recon_loss': 0.06336488574743271, + 'predict_loss': 0.006191910244524479, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22868557274341583, + 'data_time': 0.0010305329924449325, + 'model_time': 1.5078247359779198, + 'grad_norm_pre_clip_avg': 0.2036658674478531, + 'learning_rate': 1.6816228503141782e-05, + 'epoch': 5.67} +04/19 [19:30:15] INFO | >> train_qwenlatent.py:487 + Step 22470 | grad_norm_pre_clip=0.2379 | + grad_norm_pre_clip_avg=0.2199 | Metrics: + {'align_loss': 0.025658898055553436, + 'recon_loss': 0.09747497737407684, + 'predict_loss': 0.01135588064789772, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23786239326000214, + 'data_time': 0.0007340650190599263, + 'model_time': 1.2767067689856049, + 'grad_norm_pre_clip_avg': 0.21992972046136855, + 'learning_rate': 1.680804364870437e-05, + 'epoch': 5.67} +04/19 [19:30:28] INFO | >> train_qwenlatent.py:487 + Step 22480 | grad_norm_pre_clip=0.2033 | + grad_norm_pre_clip_avg=0.2220 | Metrics: + {'align_loss': 0.0254195723682642, + 'recon_loss': 0.06390981376171112, + 'predict_loss': 0.008986124768853188, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20327800512313843, + 'data_time': 0.0007266659813467413, + 'model_time': 1.2682010529970285, + 'grad_norm_pre_clip_avg': 0.22198427617549896, + 'learning_rate': 1.679985670067117e-05, + 'epoch': 5.67} +04/19 [19:30:40] INFO | >> train_qwenlatent.py:487 + Step 22490 | grad_norm_pre_clip=0.2113 | + grad_norm_pre_clip_avg=0.2006 | Metrics: + {'align_loss': 0.026190923526883125, + 'recon_loss': 0.09164231270551682, + 'predict_loss': 0.0132481399923563, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21129994094371796, + 'data_time': 0.0009539079910609871, + 'model_time': 1.235003787005553, + 'grad_norm_pre_clip_avg': 0.20057124346494676, + 'learning_rate': 1.6791667663032397e-05, + 'epoch': 5.67} +04/19 [19:30:53] INFO | >> train_qwenlatent.py:487 + Step 22500 | grad_norm_pre_clip=0.1840 | + grad_norm_pre_clip_avg=0.2046 | Metrics: + {'align_loss': 0.025815632194280624, + 'recon_loss': 0.0694240927696228, + 'predict_loss': 0.010810699313879013, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18395066261291504, + 'mae_score': 0.010318644411929019, 'data_time': + 0.0009535120043437928, 'model_time': + 1.207525934005389, 'grad_norm_pre_clip_avg': + 0.20458706468343735, 'learning_rate': + 1.678347653977929e-05, 'epoch': 5.68} +04/19 [19:31:06] INFO | >> train_qwenlatent.py:487 + Step 22510 | grad_norm_pre_clip=0.1837 | + grad_norm_pre_clip_avg=0.1915 | Metrics: + {'align_loss': 0.026130974292755127, + 'recon_loss': 0.0831264778971672, + 'predict_loss': 0.011904426850378513, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1837361454963684, + 'data_time': 0.0006821089773438871, + 'model_time': 1.2054879259958398, + 'grad_norm_pre_clip_avg': 0.19147037118673324, + 'learning_rate': 1.67752833349041e-05, 'epoch': + 5.68} +04/19 [19:31:19] INFO | >> train_qwenlatent.py:487 + Step 22520 | grad_norm_pre_clip=0.2073 | + grad_norm_pre_clip_avg=0.2057 | Metrics: + {'align_loss': 0.025852710008621216, + 'recon_loss': 0.07963791489601135, + 'predict_loss': 0.007549427915364504, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2073153704404831, + 'data_time': 0.0006690480222459882, + 'model_time': 1.2339101340039633, + 'grad_norm_pre_clip_avg': 0.2056929275393486, + 'learning_rate': 1.67670880524001e-05, 'epoch': + 5.68} +04/19 [19:31:31] INFO | >> train_qwenlatent.py:487 + Step 22530 | grad_norm_pre_clip=0.1475 | + grad_norm_pre_clip_avg=0.2055 | Metrics: + {'align_loss': 0.025978269055485725, + 'recon_loss': 0.06659559905529022, + 'predict_loss': 0.007712879683822393, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1474532037973404, + 'data_time': 0.0009352369816042483, + 'model_time': 1.2134038399963174, + 'grad_norm_pre_clip_avg': 0.20550594329833985, + 'learning_rate': 1.6758890696261563e-05, + 'epoch': 5.69} +04/19 [19:31:44] INFO | >> train_qwenlatent.py:487 + Step 22540 | grad_norm_pre_clip=0.1678 | + grad_norm_pre_clip_avg=0.1802 | Metrics: + {'align_loss': 0.02531716413795948, + 'recon_loss': 0.07578898966312408, + 'predict_loss': 0.007982056587934494, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1678214818239212, + 'data_time': 0.0007041959906928241, + 'model_time': 1.1954949849750847, + 'grad_norm_pre_clip_avg': 0.18020314276218413, + 'learning_rate': 1.6750691270483787e-05, + 'epoch': 5.69} +04/19 [19:31:58] INFO | >> train_qwenlatent.py:487 + Step 22550 | grad_norm_pre_clip=0.2643 | + grad_norm_pre_clip_avg=0.2015 | Metrics: + {'align_loss': 0.024663055315613747, + 'recon_loss': 0.06964254379272461, + 'predict_loss': 0.010835867375135422, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2642558217048645, + 'mae_score': 0.009755513474747941, 'data_time': + 0.0009916640119627118, 'model_time': + 1.2654060299973935, 'grad_norm_pre_clip_avg': + 0.20146512538194655, 'learning_rate': + 1.674248977906308e-05, 'epoch': 5.69} +04/19 [19:32:10] INFO | >> train_qwenlatent.py:487 + Step 22560 | grad_norm_pre_clip=0.2124 | + grad_norm_pre_clip_avg=0.2169 | Metrics: + {'align_loss': 0.02587290108203888, + 'recon_loss': 0.10264649242162704, + 'predict_loss': 0.014580187387764454, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21244114637374878, + 'data_time': 0.000834778998978436, + 'model_time': 1.3566376350063365, + 'grad_norm_pre_clip_avg': 0.2169416218996048, + 'learning_rate': 1.673428622599673e-05, + 'epoch': 5.69} +04/19 [19:32:23] INFO | >> train_qwenlatent.py:487 + Step 22570 | grad_norm_pre_clip=0.1791 | + grad_norm_pre_clip_avg=0.1966 | Metrics: + {'align_loss': 0.02538251131772995, + 'recon_loss': 0.08502382040023804, + 'predict_loss': 0.013341609388589859, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17914219200611115, + 'data_time': 0.0009891950176097453, + 'model_time': 1.2415346350171603, + 'grad_norm_pre_clip_avg': 0.19661179333925247, + 'learning_rate': 1.672608061528307e-05, + 'epoch': 5.7} +04/19 [19:32:35] INFO | >> train_qwenlatent.py:487 + Step 22580 | grad_norm_pre_clip=0.1721 | + grad_norm_pre_clip_avg=0.1856 | Metrics: + {'align_loss': 0.025575149804353714, + 'recon_loss': 0.07592742145061493, + 'predict_loss': 0.008790754713118076, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17213937640190125, + 'data_time': 0.0007949220016598701, + 'model_time': 1.2674615429714322, + 'grad_norm_pre_clip_avg': 0.18556807488203048, + 'learning_rate': 1.6717872950921397e-05, + 'epoch': 5.7} +04/19 [19:32:48] INFO | >> train_qwenlatent.py:487 + Step 22590 | grad_norm_pre_clip=0.1710 | + grad_norm_pre_clip_avg=0.2286 | Metrics: + {'align_loss': 0.0232663843780756, + 'recon_loss': 0.06766682118177414, + 'predict_loss': 0.012567656114697456, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1710442751646042, + 'data_time': 0.0008024339913390577, + 'model_time': 1.2253753389813937, + 'grad_norm_pre_clip_avg': 0.22859797030687332, + 'learning_rate': 1.6709663236912045e-05, + 'epoch': 5.7} +04/19 [19:33:01] INFO | >> train_qwenlatent.py:487 + Step 22600 | grad_norm_pre_clip=0.2174 | + grad_norm_pre_clip_avg=0.1985 | Metrics: + {'align_loss': 0.02688724175095558, + 'recon_loss': 0.09093237668275833, + 'predict_loss': 0.00747367599979043, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2173822969198227, + 'mae_score': 0.012099750621898754, 'data_time': + 0.0011186769988853484, 'model_time': + 1.258609383017756, 'grad_norm_pre_clip_avg': + 0.19846639037132263, 'learning_rate': + 1.670145147725631e-05, 'epoch': 5.7} +04/19 [19:33:14] INFO | >> train_qwenlatent.py:487 + Step 22610 | grad_norm_pre_clip=0.1547 | + grad_norm_pre_clip_avg=0.1944 | Metrics: + {'align_loss': 0.026050737127661705, + 'recon_loss': 0.08840558677911758, + 'predict_loss': 0.010517888702452183, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15470361709594727, + 'data_time': 0.0006987160013522953, + 'model_time': 1.6481940330122598, + 'grad_norm_pre_clip_avg': 0.19438883513212205, + 'learning_rate': 1.6693237675956516e-05, + 'epoch': 5.71} +04/19 [19:33:27] INFO | >> train_qwenlatent.py:487 + Step 22620 | grad_norm_pre_clip=0.1514 | + grad_norm_pre_clip_avg=0.1713 | Metrics: + {'align_loss': 0.02568821981549263, + 'recon_loss': 0.0679861530661583, + 'predict_loss': 0.007968979887664318, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15135329961776733, + 'data_time': 0.0009549429814796895, + 'model_time': 1.2723448410106357, + 'grad_norm_pre_clip_avg': 0.1712867721915245, + 'learning_rate': 1.668502183701597e-05, + 'epoch': 5.71} +04/19 [19:33:40] INFO | >> train_qwenlatent.py:487 + Step 22630 | grad_norm_pre_clip=0.2366 | + grad_norm_pre_clip_avg=0.1773 | Metrics: + {'align_loss': 0.025518089532852173, + 'recon_loss': 0.08212506026029587, + 'predict_loss': 0.01181737519800663, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23661749064922333, + 'data_time': 0.000744013988878578, + 'model_time': 1.685072993976064, + 'grad_norm_pre_clip_avg': 0.17733474373817443, + 'learning_rate': 1.6676803964438965e-05, + 'epoch': 5.71} +04/19 [19:33:52] INFO | >> train_qwenlatent.py:487 + Step 22640 | grad_norm_pre_clip=0.2528 | + grad_norm_pre_clip_avg=0.2629 | Metrics: + {'align_loss': 0.02493283897638321, + 'recon_loss': 0.09850521385669708, + 'predict_loss': 0.013457071036100388, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2528363764286041, + 'data_time': 0.0006234669999685138, + 'model_time': 1.2357899789931253, + 'grad_norm_pre_clip_avg': 0.26286918222904204, + 'learning_rate': 1.6668584062230797e-05, + 'epoch': 5.71} +04/19 [19:34:05] INFO | >> train_qwenlatent.py:487 + Step 22650 | grad_norm_pre_clip=0.2365 | + grad_norm_pre_clip_avg=0.2137 | Metrics: + {'align_loss': 0.026210378855466843, + 'recon_loss': 0.10801438987255096, + 'predict_loss': 0.01536380685865879, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23647384345531464, + 'mae_score': 0.01142045527965099, 'data_time': + 0.00102374202106148, 'model_time': + 1.2662348339799792, 'grad_norm_pre_clip_avg': + 0.21367852836847306, 'learning_rate': + 1.6660362134397745e-05, 'epoch': 5.72} +04/19 [19:34:18] INFO | >> train_qwenlatent.py:487 + Step 22660 | grad_norm_pre_clip=0.1854 | + grad_norm_pre_clip_avg=0.2001 | Metrics: + {'align_loss': 0.025499245151877403, + 'recon_loss': 0.05936016887426376, + 'predict_loss': 0.007876093499362469, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18535476922988892, + 'data_time': 0.0008542909927200526, + 'model_time': 1.1748575379897375, + 'grad_norm_pre_clip_avg': 0.20013507306575776, + 'learning_rate': 1.6652138184947083e-05, + 'epoch': 5.72} +04/19 [19:34:30] INFO | >> train_qwenlatent.py:487 + Step 22670 | grad_norm_pre_clip=0.1759 | + grad_norm_pre_clip_avg=0.1776 | Metrics: + {'align_loss': 0.025157075375318527, + 'recon_loss': 0.06166835501790047, + 'predict_loss': 0.006498753093183041, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1758919507265091, + 'data_time': 0.0006673280149698257, + 'model_time': 1.1878605209931266, + 'grad_norm_pre_clip_avg': 0.17763799875974656, + 'learning_rate': 1.664391221788705e-05, + 'epoch': 5.72} +04/19 [19:34:43] INFO | >> train_qwenlatent.py:487 + Step 22680 | grad_norm_pre_clip=0.1811 | + grad_norm_pre_clip_avg=0.1881 | Metrics: + {'align_loss': 0.025748860090970993, + 'recon_loss': 0.07775857299566269, + 'predict_loss': 0.01163981668651104, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1810554563999176, + 'data_time': 0.0013327049964573234, + 'model_time': 1.3380969399877358, + 'grad_norm_pre_clip_avg': 0.18808355927467346, + 'learning_rate': 1.6635684237226895e-05, + 'epoch': 5.72} +04/19 [19:34:56] INFO | >> train_qwenlatent.py:487 + Step 22690 | grad_norm_pre_clip=0.1843 | + grad_norm_pre_clip_avg=0.1941 | Metrics: + {'align_loss': 0.02551526576280594, + 'recon_loss': 0.07248876243829727, + 'predict_loss': 0.008272337727248669, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18432947993278503, + 'data_time': 0.0009359940013382584, + 'model_time': 1.2441574419790413, + 'grad_norm_pre_clip_avg': 0.19409705698490143, + 'learning_rate': 1.662745424697683e-05, + 'epoch': 5.73} +04/19 [19:35:09] INFO | >> train_qwenlatent.py:487 + Step 22700 | grad_norm_pre_clip=0.2016 | + grad_norm_pre_clip_avg=0.1866 | Metrics: + {'align_loss': 0.024601653218269348, + 'recon_loss': 0.08069681376218796, + 'predict_loss': 0.009363523684442043, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2015760987997055, + 'mae_score': 0.01267719182882223, 'data_time': + 0.0006257949862629175, 'model_time': + 1.2010396039986517, 'grad_norm_pre_clip_avg': + 0.18660722821950912, 'learning_rate': + 1.661922225114806e-05, 'epoch': 5.73} +04/19 [19:35:22] INFO | >> train_qwenlatent.py:487 + Step 22710 | grad_norm_pre_clip=0.1775 | + grad_norm_pre_clip_avg=0.1897 | Metrics: + {'align_loss': 0.02410270646214485, + 'recon_loss': 0.05441504716873169, + 'predict_loss': 0.012130845338106155, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17752863466739655, + 'data_time': 0.001035325025441125, + 'model_time': 1.2348662939912174, + 'grad_norm_pre_clip_avg': 0.1896715685725212, + 'learning_rate': 1.6610988253752753e-05, + 'epoch': 5.73} +04/19 [19:35:34] INFO | >> train_qwenlatent.py:487 + Step 22720 | grad_norm_pre_clip=0.1729 | + grad_norm_pre_clip_avg=0.1734 | Metrics: + {'align_loss': 0.024834424257278442, + 'recon_loss': 0.06753933429718018, + 'predict_loss': 0.010240253992378712, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17286550998687744, + 'data_time': 0.0010047589894384146, + 'model_time': 1.2317178579978645, + 'grad_norm_pre_clip_avg': 0.1734168142080307, + 'learning_rate': 1.660275225880405e-05, + 'epoch': 5.73} +04/19 [19:35:47] INFO | >> train_qwenlatent.py:487 + Step 22730 | grad_norm_pre_clip=0.2131 | + grad_norm_pre_clip_avg=0.2099 | Metrics: + {'align_loss': 0.024386730045080185, + 'recon_loss': 0.06117291375994682, + 'predict_loss': 0.007711146026849747, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21314124763011932, + 'data_time': 0.0010900269844569266, + 'model_time': 1.229904498992255, + 'grad_norm_pre_clip_avg': 0.20992230325937272, + 'learning_rate': 1.6594514270316094e-05, + 'epoch': 5.74} +04/19 [19:36:00] INFO | >> train_qwenlatent.py:487 + Step 22740 | grad_norm_pre_clip=0.1702 | + grad_norm_pre_clip_avg=0.1955 | Metrics: + {'align_loss': 0.026219099760055542, + 'recon_loss': 0.07960348576307297, + 'predict_loss': 0.009153385646641254, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1701720505952835, + 'data_time': 0.0007043619989417493, + 'model_time': 1.2366381309984718, + 'grad_norm_pre_clip_avg': 0.19551954716444014, + 'learning_rate': 1.658627429230397e-05, + 'epoch': 5.74} +04/19 [19:36:13] INFO | >> train_qwenlatent.py:487 + Step 22750 | grad_norm_pre_clip=0.1503 | + grad_norm_pre_clip_avg=0.1733 | Metrics: + {'align_loss': 0.024977974593639374, + 'recon_loss': 0.06722992658615112, + 'predict_loss': 0.008565247058868408, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1502843201160431, + 'mae_score': 0.011874381915943043, 'data_time': + 0.0009593259892426431, 'model_time': + 1.257701611990342, 'grad_norm_pre_clip_avg': + 0.1733115568757057, 'learning_rate': + 1.6578032328783747e-05, 'epoch': 5.74} +04/19 [19:36:26] INFO | >> train_qwenlatent.py:487 + Step 22760 | grad_norm_pre_clip=0.2104 | + grad_norm_pre_clip_avg=0.2507 | Metrics: + {'align_loss': 0.025662638247013092, + 'recon_loss': 0.07353625446557999, + 'predict_loss': 0.007459895219653845, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21040061116218567, + 'data_time': 0.0008662079926580191, + 'model_time': 1.2511143960000481, + 'grad_norm_pre_clip_avg': 0.2507346376776695, + 'learning_rate': 1.656978838377245e-05, + 'epoch': 5.74} +04/19 [19:36:38] INFO | >> train_qwenlatent.py:487 + Step 22770 | grad_norm_pre_clip=0.2197 | + grad_norm_pre_clip_avg=0.2136 | Metrics: + {'align_loss': 0.02626909501850605, + 'recon_loss': 0.08654481917619705, + 'predict_loss': 0.009157710708677769, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2196577787399292, + 'data_time': 0.0008477029914502054, + 'model_time': 1.2132231470022816, + 'grad_norm_pre_clip_avg': 0.21358431577682496, + 'learning_rate': 1.6561542461288084e-05, + 'epoch': 5.75} +04/19 [19:36:51] INFO | >> train_qwenlatent.py:487 + Step 22780 | grad_norm_pre_clip=0.2154 | + grad_norm_pre_clip_avg=0.2019 | Metrics: + {'align_loss': 0.025369971990585327, + 'recon_loss': 0.08075864613056183, + 'predict_loss': 0.011715351603925228, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21537214517593384, + 'data_time': 0.0012862989970017225, + 'model_time': 1.2115575350180734, + 'grad_norm_pre_clip_avg': 0.20193143486976622, + 'learning_rate': 1.6553294565349613e-05, + 'epoch': 5.75} +04/19 [19:37:03] INFO | >> train_qwenlatent.py:487 + Step 22790 | grad_norm_pre_clip=0.2367 | + grad_norm_pre_clip_avg=0.1971 | Metrics: + {'align_loss': 0.02639785222709179, + 'recon_loss': 0.08382812142372131, + 'predict_loss': 0.016173794865608215, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.236720010638237, + 'data_time': 0.0008624419861007482, + 'model_time': 1.2966027129732538, + 'grad_norm_pre_clip_avg': 0.19713667631149293, + 'learning_rate': 1.654504469997696e-05, + 'epoch': 5.75} +04/19 [19:37:17] INFO | >> train_qwenlatent.py:487 + Step 22800 | grad_norm_pre_clip=0.1933 | + grad_norm_pre_clip_avg=0.2060 | Metrics: + {'align_loss': 0.02524174004793167, + 'recon_loss': 0.0732891783118248, + 'predict_loss': 0.01326122134923935, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1933034211397171, + 'mae_score': 0.009609712995924391, 'data_time': + 0.0009489619988016784, 'model_time': + 1.2679050140141044, 'grad_norm_pre_clip_avg': + 0.20595734119415282, 'learning_rate': + 1.653679286919101e-05, 'epoch': 5.75} +04/19 [19:37:30] INFO | >> train_qwenlatent.py:487 + Step 22810 | grad_norm_pre_clip=0.1993 | + grad_norm_pre_clip_avg=0.2098 | Metrics: + {'align_loss': 0.025741195306181908, + 'recon_loss': 0.10575415939092636, + 'predict_loss': 0.01927528716623783, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19927988946437836, + 'data_time': 0.0008989039924927056, + 'model_time': 1.3036099190067034, + 'grad_norm_pre_clip_avg': 0.20979749113321305, + 'learning_rate': 1.65285390770136e-05, 'epoch': + 5.76} +04/19 [19:37:42] INFO | >> train_qwenlatent.py:487 + Step 22820 | grad_norm_pre_clip=0.2051 | + grad_norm_pre_clip_avg=0.2206 | Metrics: + {'align_loss': 0.024677550420165062, + 'recon_loss': 0.06302924454212189, + 'predict_loss': 0.009259738028049469, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20509307086467743, + 'data_time': 0.0006820880225859582, + 'model_time': 1.2470151910092682, + 'grad_norm_pre_clip_avg': 0.22056089639663695, + 'learning_rate': 1.6520283327467527e-05, + 'epoch': 5.76} +04/19 [19:37:55] INFO | >> train_qwenlatent.py:487 + Step 22830 | grad_norm_pre_clip=0.2000 | + grad_norm_pre_clip_avg=0.1954 | Metrics: + {'align_loss': 0.025912541896104813, + 'recon_loss': 0.08972354978322983, + 'predict_loss': 0.014854582957923412, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20003612339496613, + 'data_time': 0.0009059889998752624, + 'model_time': 1.2595669370202813, + 'grad_norm_pre_clip_avg': 0.1953762039542198, + 'learning_rate': 1.6512025624576554e-05, + 'epoch': 5.76} +04/19 [19:38:08] INFO | >> train_qwenlatent.py:487 + Step 22840 | grad_norm_pre_clip=0.1891 | + grad_norm_pre_clip_avg=0.1906 | Metrics: + {'align_loss': 0.024106979370117188, + 'recon_loss': 0.073629230260849, + 'predict_loss': 0.011929741129279137, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18907169997692108, + 'data_time': 0.000784825999289751, + 'model_time': 1.2757354119967204, + 'grad_norm_pre_clip_avg': 0.19058060348033906, + 'learning_rate': 1.6503765972365377e-05, + 'epoch': 5.76} +04/19 [19:38:21] INFO | >> train_qwenlatent.py:487 + Step 22850 | grad_norm_pre_clip=0.1428 | + grad_norm_pre_clip_avg=0.1730 | Metrics: + {'align_loss': 0.02482721209526062, + 'recon_loss': 0.10239110887050629, + 'predict_loss': 0.011368094012141228, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14276176691055298, + 'mae_score': 0.008755352260830166, 'data_time': + 0.0010624569840729237, 'model_time': + 1.315240546973655, 'grad_norm_pre_clip_avg': + 0.17296261191368104, 'learning_rate': + 1.649550437485965e-05, 'epoch': 5.77} +04/19 [19:38:34] INFO | >> train_qwenlatent.py:487 + Step 22860 | grad_norm_pre_clip=0.1722 | + grad_norm_pre_clip_avg=0.1833 | Metrics: + {'align_loss': 0.02607886493206024, + 'recon_loss': 0.08485434949398041, + 'predict_loss': 0.015457220375537872, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1722232848405838, + 'data_time': 0.000880179984960705, + 'model_time': 1.2340564990008716, + 'grad_norm_pre_clip_avg': 0.1833125039935112, + 'learning_rate': 1.6487240836085978e-05, + 'epoch': 5.77} +04/19 [19:38:46] INFO | >> train_qwenlatent.py:487 + Step 22870 | grad_norm_pre_clip=0.2119 | + grad_norm_pre_clip_avg=0.2001 | Metrics: + {'align_loss': 0.025635916739702225, + 'recon_loss': 0.08399416506290436, + 'predict_loss': 0.011167596094310284, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21186861395835876, + 'data_time': 0.0009615739982109517, + 'model_time': 1.2449635959928855, + 'grad_norm_pre_clip_avg': 0.2000957250595093, + 'learning_rate': 1.6478975360071904e-05, + 'epoch': 5.77} +04/19 [19:38:59] INFO | >> train_qwenlatent.py:487 + Step 22880 | grad_norm_pre_clip=0.1632 | + grad_norm_pre_clip_avg=0.1877 | Metrics: + {'align_loss': 0.025779742747545242, + 'recon_loss': 0.10431419312953949, + 'predict_loss': 0.011400110088288784, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16324903070926666, + 'data_time': 0.0010055509919766337, + 'model_time': 1.2482531519781332, + 'grad_norm_pre_clip_avg': 0.18774640560150146, + 'learning_rate': 1.6470707950845933e-05, + 'epoch': 5.77} +04/19 [19:39:11] INFO | >> train_qwenlatent.py:487 + Step 22890 | grad_norm_pre_clip=0.1640 | + grad_norm_pre_clip_avg=0.2009 | Metrics: + {'align_loss': 0.0251225084066391, + 'recon_loss': 0.07484229654073715, + 'predict_loss': 0.008216246031224728, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16395428776741028, + 'data_time': 0.0007406270015053451, + 'model_time': 1.3252085050044116, + 'grad_norm_pre_clip_avg': 0.20093492269515992, + 'learning_rate': 1.6462438612437483e-05, + 'epoch': 5.78} +04/19 [19:39:24] INFO | >> train_qwenlatent.py:487 + Step 22900 | grad_norm_pre_clip=0.1695 | + grad_norm_pre_clip_avg=0.2058 | Metrics: + {'align_loss': 0.024705611169338226, + 'recon_loss': 0.09044007211923599, + 'predict_loss': 0.011746551841497421, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16949091851711273, + 'mae_score': 0.009370168050130209, 'data_time': + 0.0007237129902932793, 'model_time': + 1.1948074479878414, 'grad_norm_pre_clip_avg': + 0.2058255285024643, 'learning_rate': + 1.6454167348876946e-05, 'epoch': 5.78} +04/19 [19:39:37] INFO | >> train_qwenlatent.py:487 + Step 22910 | grad_norm_pre_clip=0.2104 | + grad_norm_pre_clip_avg=0.2122 | Metrics: + {'align_loss': 0.025875434279441833, + 'recon_loss': 0.06995585560798645, + 'predict_loss': 0.005875780247151852, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21036754548549652, + 'data_time': 0.0010974750039167702, + 'model_time': 1.2339057170029264, + 'grad_norm_pre_clip_avg': 0.21215905696153642, + 'learning_rate': 1.6445894164195617e-05, + 'epoch': 5.78} +04/19 [19:39:50] INFO | >> train_qwenlatent.py:487 + Step 22920 | grad_norm_pre_clip=0.1542 | + grad_norm_pre_clip_avg=0.1878 | Metrics: + {'align_loss': 0.02470802143216133, + 'recon_loss': 0.07021348178386688, + 'predict_loss': 0.0131907369941473, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.154166579246521, + 'data_time': 0.0007000269833952188, + 'model_time': 1.2268486599787138, + 'grad_norm_pre_clip_avg': 0.1877669095993042, + 'learning_rate': 1.6437619062425764e-05, + 'epoch': 5.78} +04/19 [19:40:03] INFO | >> train_qwenlatent.py:487 + Step 22930 | grad_norm_pre_clip=0.2112 | + grad_norm_pre_clip_avg=0.1683 | Metrics: + {'align_loss': 0.025537196546792984, + 'recon_loss': 0.0929737538099289, + 'predict_loss': 0.011391724459826946, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21115781366825104, + 'data_time': 0.0009731279860716313, + 'model_time': 1.2199208259989973, + 'grad_norm_pre_clip_avg': 0.16834347397089006, + 'learning_rate': 1.6429342047600562e-05, + 'epoch': 5.79} +04/19 [19:40:16] INFO | >> train_qwenlatent.py:487 + Step 22940 | grad_norm_pre_clip=0.3256 | + grad_norm_pre_clip_avg=0.2295 | Metrics: + {'align_loss': 0.025570746511220932, + 'recon_loss': 0.08733134716749191, + 'predict_loss': 0.014737440273165703, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3255738615989685, + 'data_time': 0.0010995240008924156, + 'model_time': 1.2734633480140474, + 'grad_norm_pre_clip_avg': 0.22952160388231277, + 'learning_rate': 1.6421063123754126e-05, + 'epoch': 5.79} +04/19 [19:40:29] INFO | >> train_qwenlatent.py:487 + Step 22950 | grad_norm_pre_clip=0.2023 | + grad_norm_pre_clip_avg=0.2076 | Metrics: + {'align_loss': 0.023190956562757492, + 'recon_loss': 0.057830579578876495, + 'predict_loss': 0.009489888325333595, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20233763754367828, + 'mae_score': 0.011620964015926327, 'data_time': + 0.0008121600258164108, 'model_time': + 1.2373211399826687, 'grad_norm_pre_clip_avg': + 0.20758986622095107, 'learning_rate': + 1.6412782294921504e-05, 'epoch': 5.79} +04/19 [19:40:41] INFO | >> train_qwenlatent.py:487 + Step 22960 | grad_norm_pre_clip=0.1918 | + grad_norm_pre_clip_avg=0.2021 | Metrics: + {'align_loss': 0.025562692433595657, + 'recon_loss': 0.0875202938914299, + 'predict_loss': 0.009597690775990486, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19176167249679565, + 'data_time': 0.0009330690081696957, + 'model_time': 1.2742651959997602, + 'grad_norm_pre_clip_avg': 0.202096788585186, + 'learning_rate': 1.6404499565138674e-05, + 'epoch': 5.79} +04/19 [19:40:54] INFO | >> train_qwenlatent.py:487 + Step 22970 | grad_norm_pre_clip=0.1457 | + grad_norm_pre_clip_avg=0.1999 | Metrics: + {'align_loss': 0.025678671896457672, + 'recon_loss': 0.07203441113233566, + 'predict_loss': 0.010825837031006813, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14574642479419708, + 'data_time': 0.0009173509897664189, + 'model_time': 1.27373949799221, + 'grad_norm_pre_clip_avg': 0.19987784177064896, + 'learning_rate': 1.639621493844254e-05, + 'epoch': 5.8} +04/19 [19:41:06] INFO | >> train_qwenlatent.py:487 + Step 22980 | grad_norm_pre_clip=0.2257 | + grad_norm_pre_clip_avg=0.2192 | Metrics: + {'align_loss': 0.025685397908091545, + 'recon_loss': 0.09643909335136414, + 'predict_loss': 0.009878997690975666, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2256890833377838, + 'data_time': 0.0009359759860672057, + 'model_time': 1.2313226900005247, + 'grad_norm_pre_clip_avg': 0.21917246878147126, + 'learning_rate': 1.638792841887092e-05, + 'epoch': 5.8} +04/19 [19:41:19] INFO | >> train_qwenlatent.py:487 + Step 22990 | grad_norm_pre_clip=0.1834 | + grad_norm_pre_clip_avg=0.2265 | Metrics: + {'align_loss': 0.023479990661144257, + 'recon_loss': 0.05457218736410141, + 'predict_loss': 0.008552664890885353, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18342213332653046, + 'data_time': 0.0008413119940087199, + 'model_time': 1.2064521790016443, + 'grad_norm_pre_clip_avg': 0.22646393328905107, + 'learning_rate': 1.637964001046257e-05, + 'epoch': 5.8} +04/19 [19:41:32] INFO | >> train_qwenlatent.py:487 + Step 23000 | grad_norm_pre_clip=0.1639 | + grad_norm_pre_clip_avg=0.1796 | Metrics: + {'align_loss': 0.025801487267017365, + 'recon_loss': 0.09572935849428177, + 'predict_loss': 0.007896588183939457, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16390909254550934, + 'mae_score': 0.0086348954621736, 'data_time': + 0.0006290910241659731, 'model_time': + 1.181720705004409, 'grad_norm_pre_clip_avg': + 0.17959136217832566, 'learning_rate': + 1.6371349717257157e-05, 'epoch': 5.8} +04/19 [19:41:44] INFO | >> train_qwenlatent.py:487 + Step 23010 | grad_norm_pre_clip=0.1843 | + grad_norm_pre_clip_avg=0.1723 | Metrics: + {'align_loss': 0.02595655992627144, + 'recon_loss': 0.08528126776218414, + 'predict_loss': 0.014094362035393715, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18431133031845093, + 'data_time': 0.0006402269937098026, + 'model_time': 1.1764516700059175, + 'grad_norm_pre_clip_avg': 0.17227858006954194, + 'learning_rate': 1.6363057543295272e-05, + 'epoch': 5.81} +04/19 [19:41:56] INFO | >> train_qwenlatent.py:487 + Step 23020 | grad_norm_pre_clip=0.2458 | + grad_norm_pre_clip_avg=0.2059 | Metrics: + {'align_loss': 0.024536659941077232, + 'recon_loss': 0.08381489664316177, + 'predict_loss': 0.011836023069918156, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24577629566192627, + 'data_time': 0.0006591729761566967, + 'model_time': 1.1953520380193368, + 'grad_norm_pre_clip_avg': 0.20592830032110215, + 'learning_rate': 1.6354763492618418e-05, + 'epoch': 5.81} +04/19 [19:42:07] INFO | >> train_qwenlatent.py:487 + Step 23030 | grad_norm_pre_clip=0.2176 | + grad_norm_pre_clip_avg=0.2099 | Metrics: + {'align_loss': 0.026999006047844887, + 'recon_loss': 0.09740573167800903, + 'predict_loss': 0.01148492842912674, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21755297482013702, + 'data_time': 0.0005829209985677153, + 'model_time': 1.1519017009995878, + 'grad_norm_pre_clip_avg': 0.209932342171669, + 'learning_rate': 1.6346467569269007e-05, + 'epoch': 5.81} +04/19 [19:42:19] INFO | >> train_qwenlatent.py:487 + Step 23040 | grad_norm_pre_clip=0.1598 | + grad_norm_pre_clip_avg=0.2138 | Metrics: + {'align_loss': 0.02526644803583622, + 'recon_loss': 0.10319432616233826, + 'predict_loss': 0.012821436859667301, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1598047912120819, + 'data_time': 0.000614138989476487, + 'model_time': 1.1804106679919641, + 'grad_norm_pre_clip_avg': 0.21382613480091095, + 'learning_rate': 1.633816977729038e-05, + 'epoch': 5.81} +04/19 [19:42:31] INFO | >> train_qwenlatent.py:487 + Step 23050 | grad_norm_pre_clip=0.2838 | + grad_norm_pre_clip_avg=0.2265 | Metrics: + {'align_loss': 0.02540709264576435, + 'recon_loss': 0.1047777384519577, + 'predict_loss': 0.010684838518500328, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28376415371894836, + 'mae_score': 0.008693723420839052, 'data_time': + 0.0006159160111565143, 'model_time': + 1.1532911820104346, 'grad_norm_pre_clip_avg': + 0.2264697879552841, 'learning_rate': + 1.6329870120726785e-05, 'epoch': 5.82} +04/19 [19:42:43] INFO | >> train_qwenlatent.py:487 + Step 23060 | grad_norm_pre_clip=0.1818 | + grad_norm_pre_clip_avg=0.1920 | Metrics: + {'align_loss': 0.024153195321559906, + 'recon_loss': 0.06988221406936646, + 'predict_loss': 0.00737741170451045, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18175600469112396, + 'data_time': 0.0006343419954646379, + 'model_time': 1.1450253820221405, + 'grad_norm_pre_clip_avg': 0.19197129160165788, + 'learning_rate': 1.6321568603623365e-05, + 'epoch': 5.82} +04/19 [19:42:55] INFO | >> train_qwenlatent.py:487 + Step 23070 | grad_norm_pre_clip=0.2271 | + grad_norm_pre_clip_avg=0.1900 | Metrics: + {'align_loss': 0.024055423215031624, + 'recon_loss': 0.09044579416513443, + 'predict_loss': 0.014447427354753017, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22710852324962616, + 'data_time': 0.0006288550212047994, + 'model_time': 1.1543669819948263, + 'grad_norm_pre_clip_avg': 0.18998829871416092, + 'learning_rate': 1.6313265230026175e-05, + 'epoch': 5.82} +04/19 [19:43:07] INFO | >> train_qwenlatent.py:487 + Step 23080 | grad_norm_pre_clip=0.2107 | + grad_norm_pre_clip_avg=0.1668 | Metrics: + {'align_loss': 0.02606040984392166, + 'recon_loss': 0.08529742807149887, + 'predict_loss': 0.009237714111804962, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2106919139623642, + 'data_time': 0.0006602310168091208, + 'model_time': 1.137766982021276, + 'grad_norm_pre_clip_avg': 0.16677243560552596, + 'learning_rate': 1.6304960003982188e-05, + 'epoch': 5.82} +04/19 [19:43:19] INFO | >> train_qwenlatent.py:487 + Step 23090 | grad_norm_pre_clip=0.1804 | + grad_norm_pre_clip_avg=0.2041 | Metrics: + {'align_loss': 0.024626746773719788, + 'recon_loss': 0.06299053877592087, + 'predict_loss': 0.008647812530398369, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18038448691368103, + 'data_time': 0.0006260009831748903, + 'model_time': 1.17609234599513, + 'grad_norm_pre_clip_avg': 0.20408145934343339, + 'learning_rate': 1.6296652929539267e-05, + 'epoch': 5.83} +04/19 [19:43:31] INFO | >> train_qwenlatent.py:487 + Step 23100 | grad_norm_pre_clip=0.2301 | + grad_norm_pre_clip_avg=0.1795 | Metrics: + {'align_loss': 0.025393720716238022, + 'recon_loss': 0.09286785125732422, + 'predict_loss': 0.012204913422465324, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23013339936733246, + 'mae_score': 0.013243644301955765, 'data_time': + 0.000586782000027597, 'model_time': + 1.140566399000818, 'grad_norm_pre_clip_avg': + 0.17951653748750687, 'learning_rate': + 1.628834401074618e-05, 'epoch': 5.83} +04/19 [19:43:42] INFO | >> train_qwenlatent.py:487 + Step 23110 | grad_norm_pre_clip=0.2781 | + grad_norm_pre_clip_avg=0.1934 | Metrics: + {'align_loss': 0.025807984173297882, + 'recon_loss': 0.06586233526468277, + 'predict_loss': 0.00546156195923686, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2780565023422241, + 'data_time': 0.0006214769964572042, + 'model_time': 1.1433341150113847, + 'grad_norm_pre_clip_avg': 0.19344574064016343, + 'learning_rate': 1.6280033251652584e-05, + 'epoch': 5.83} +04/19 [19:43:54] INFO | >> train_qwenlatent.py:487 + Step 23120 | grad_norm_pre_clip=0.2578 | + grad_norm_pre_clip_avg=0.2471 | Metrics: + {'align_loss': 0.024759087711572647, + 'recon_loss': 0.07559620589017868, + 'predict_loss': 0.012367548421025276, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25784051418304443, + 'data_time': 0.0006180149794090539, + 'model_time': 1.151531941985013, + 'grad_norm_pre_clip_avg': 0.2471056714653969, + 'learning_rate': 1.627172065630906e-05, + 'epoch': 5.83} +04/19 [19:44:05] INFO | >> train_qwenlatent.py:487 + Step 23130 | grad_norm_pre_clip=0.1940 | + grad_norm_pre_clip_avg=0.1953 | Metrics: + {'align_loss': 0.025140997022390366, + 'recon_loss': 0.07765869051218033, + 'predict_loss': 0.006654432509094477, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1939501315355301, + 'data_time': 0.0006232699961401522, + 'model_time': 1.144279756990727, + 'grad_norm_pre_clip_avg': 0.19532871395349502, + 'learning_rate': 1.626340622876705e-05, + 'epoch': 5.84} +04/19 [19:44:17] INFO | >> train_qwenlatent.py:487 + Step 23140 | grad_norm_pre_clip=0.1817 | + grad_norm_pre_clip_avg=0.1988 | Metrics: + {'align_loss': 0.025003377348184586, + 'recon_loss': 0.07795123010873795, + 'predict_loss': 0.012154804542660713, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18174995481967926, + 'data_time': 0.0005582140001934022, + 'model_time': 1.1491109519847669, + 'grad_norm_pre_clip_avg': 0.19879992604255675, + 'learning_rate': 1.625508997307891e-05, + 'epoch': 5.84} +04/19 [19:44:29] INFO | >> train_qwenlatent.py:487 + Step 23150 | grad_norm_pre_clip=0.1513 | + grad_norm_pre_clip_avg=0.1793 | Metrics: + {'align_loss': 0.02644434943795204, + 'recon_loss': 0.07191437482833862, + 'predict_loss': 0.008645109832286835, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15130381286144257, + 'mae_score': 0.012161594253402573, 'data_time': + 0.0005600939912255853, 'model_time': + 1.1434205769910477, 'grad_norm_pre_clip_avg': + 0.1792714074254036, 'learning_rate': + 1.6246771893297882e-05, 'epoch': 5.84} +04/19 [19:44:41] INFO | >> train_qwenlatent.py:487 + Step 23160 | grad_norm_pre_clip=0.1950 | + grad_norm_pre_clip_avg=0.1697 | Metrics: + {'align_loss': 0.02498827688395977, + 'recon_loss': 0.07020699977874756, + 'predict_loss': 0.010280676186084747, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19499480724334717, + 'data_time': 0.000689314998453483, + 'model_time': 1.1544995379808825, + 'grad_norm_pre_clip_avg': 0.16970179751515388, + 'learning_rate': 1.6238451993478103e-05, + 'epoch': 5.84} +04/19 [19:44:53] INFO | >> train_qwenlatent.py:487 + Step 23170 | grad_norm_pre_clip=0.1924 | + grad_norm_pre_clip_avg=0.1893 | Metrics: + {'align_loss': 0.025116052478551865, + 'recon_loss': 0.0854339599609375, + 'predict_loss': 0.00991029106080532, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19244447350502014, + 'data_time': 0.0006404519954230636, + 'model_time': 1.1712040179991163, + 'grad_norm_pre_clip_avg': 0.18926066160202026, + 'learning_rate': 1.623013027767458e-05, + 'epoch': 5.85} +04/19 [19:45:04] INFO | >> train_qwenlatent.py:487 + Step 23180 | grad_norm_pre_clip=0.1850 | + grad_norm_pre_clip_avg=0.1806 | Metrics: + {'align_loss': 0.02392261102795601, + 'recon_loss': 0.07145807892084122, + 'predict_loss': 0.012200792320072651, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1849953830242157, + 'data_time': 0.0008162980084307492, + 'model_time': 1.238177565013757, + 'grad_norm_pre_clip_avg': 0.1805646151304245, + 'learning_rate': 1.6221806749943226e-05, + 'epoch': 5.85} +04/19 [19:45:16] INFO | >> train_qwenlatent.py:487 + Step 23190 | grad_norm_pre_clip=0.2077 | + grad_norm_pre_clip_avg=0.1947 | Metrics: + {'align_loss': 0.02507476881146431, + 'recon_loss': 0.08115364611148834, + 'predict_loss': 0.01108761690557003, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2076815515756607, + 'data_time': 0.0005780909850727767, + 'model_time': 1.1380215579993092, + 'grad_norm_pre_clip_avg': 0.19473899602890016, + 'learning_rate': 1.621348141434082e-05, + 'epoch': 5.85} +04/19 [19:45:28] INFO | >> train_qwenlatent.py:487 + Step 23200 | grad_norm_pre_clip=0.1815 | + grad_norm_pre_clip_avg=0.1775 | Metrics: + {'align_loss': 0.026015419512987137, + 'recon_loss': 0.06364704668521881, + 'predict_loss': 0.005730542354285717, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18150505423545837, + 'mae_score': 0.012575051591203019, 'data_time': + 0.0005840749945491552, 'model_time': + 1.1249759219936095, 'grad_norm_pre_clip_avg': + 0.17754177749156952, 'learning_rate': + 1.6205154274925037e-05, 'epoch': 5.85} +04/19 [19:45:40] INFO | >> train_qwenlatent.py:487 + Step 23210 | grad_norm_pre_clip=0.2035 | + grad_norm_pre_clip_avg=0.2247 | Metrics: + {'align_loss': 0.025667594745755196, + 'recon_loss': 0.08874331414699554, + 'predict_loss': 0.012061051093041897, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20349936187267303, + 'data_time': 0.0005989219935145229, + 'model_time': 1.1600181579997297, + 'grad_norm_pre_clip_avg': 0.2247425600886345, + 'learning_rate': 1.6196825335754413e-05, + 'epoch': 5.86} +04/19 [19:45:51] INFO | >> train_qwenlatent.py:487 + Step 23220 | grad_norm_pre_clip=0.1542 | + grad_norm_pre_clip_avg=0.1857 | Metrics: + {'align_loss': 0.023790568113327026, + 'recon_loss': 0.05508364737033844, + 'predict_loss': 0.011551691219210625, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15417981147766113, + 'data_time': 0.0006660130165982991, + 'model_time': 1.1415424869919661, + 'grad_norm_pre_clip_avg': 0.18567102253437043, + 'learning_rate': 1.618849460088838e-05, + 'epoch': 5.86} +04/19 [19:46:03] INFO | >> train_qwenlatent.py:487 + Step 23230 | grad_norm_pre_clip=0.1176 | + grad_norm_pre_clip_avg=0.1637 | Metrics: + {'align_loss': 0.025739621371030807, + 'recon_loss': 0.07674095779657364, + 'predict_loss': 0.005049429833889008, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11755391955375671, + 'data_time': 0.000635252014035359, + 'model_time': 1.1531621209869627, + 'grad_norm_pre_clip_avg': 0.16366546154022216, + 'learning_rate': 1.6180162074387244e-05, + 'epoch': 5.86} +04/19 [19:46:15] INFO | >> train_qwenlatent.py:487 + Step 23240 | grad_norm_pre_clip=0.1789 | + grad_norm_pre_clip_avg=0.1694 | Metrics: + {'align_loss': 0.024831686168909073, + 'recon_loss': 0.08115046471357346, + 'predict_loss': 0.008267207071185112, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17885969579219818, + 'data_time': 0.0006036840204615146, + 'model_time': 1.1440590870042797, + 'grad_norm_pre_clip_avg': 0.16941555291414262, + 'learning_rate': 1.6171827760312156e-05, + 'epoch': 5.86} +04/19 [19:46:27] INFO | >> train_qwenlatent.py:487 + Step 23250 | grad_norm_pre_clip=0.2186 | + grad_norm_pre_clip_avg=0.2253 | Metrics: + {'align_loss': 0.024938873946666718, + 'recon_loss': 0.08873365819454193, + 'predict_loss': 0.011682555079460144, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2186320424079895, + 'mae_score': 0.01112364691656989, 'data_time': + 0.0005567670159507543, 'model_time': + 1.143701017019339, 'grad_norm_pre_clip_avg': + 0.22528768181800843, 'learning_rate': + 1.616349166272518e-05, 'epoch': 5.87} +04/19 [19:46:39] INFO | >> train_qwenlatent.py:487 + Step 23260 | grad_norm_pre_clip=0.2488 | + grad_norm_pre_clip_avg=0.2049 | Metrics: + {'align_loss': 0.025507871061563492, + 'recon_loss': 0.09083620458841324, + 'predict_loss': 0.016430465504527092, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24879585206508636, + 'data_time': 0.000566679984331131, + 'model_time': 1.1588326419878285, + 'grad_norm_pre_clip_avg': 0.20487144142389296, + 'learning_rate': 1.615515378568921e-05, + 'epoch': 5.87} +04/19 [19:46:50] INFO | >> train_qwenlatent.py:487 + Step 23270 | grad_norm_pre_clip=0.1648 | + grad_norm_pre_clip_avg=0.1813 | Metrics: + {'align_loss': 0.02350606769323349, + 'recon_loss': 0.07091952115297318, + 'predict_loss': 0.010676322504878044, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1647709459066391, + 'data_time': 0.0005789010028820485, + 'model_time': 1.1453769149957225, + 'grad_norm_pre_clip_avg': 0.18128508180379868, + 'learning_rate': 1.614681413326804e-05, + 'epoch': 5.87} +04/19 [19:47:02] INFO | >> train_qwenlatent.py:487 + Step 23280 | grad_norm_pre_clip=0.2091 | + grad_norm_pre_clip_avg=0.1861 | Metrics: + {'align_loss': 0.025483019649982452, + 'recon_loss': 0.06769750267267227, + 'predict_loss': 0.005337355192750692, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20909984409809113, + 'data_time': 0.0006113029958214611, + 'model_time': 1.3085701119853184, + 'grad_norm_pre_clip_avg': 0.18612809330224991, + 'learning_rate': 1.613847270952632e-05, + 'epoch': 5.87} +04/19 [19:47:13] INFO | >> train_qwenlatent.py:487 + Step 23290 | grad_norm_pre_clip=0.1874 | + grad_norm_pre_clip_avg=0.2348 | Metrics: + {'align_loss': 0.02504640445113182, + 'recon_loss': 0.05594410374760628, + 'predict_loss': 0.006207223050296307, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1873902976512909, + 'data_time': 0.0007155470084398985, + 'model_time': 1.136238726990996, + 'grad_norm_pre_clip_avg': 0.23476987481117248, + 'learning_rate': 1.6130129518529544e-05, + 'epoch': 5.88} +04/19 [19:47:25] INFO | >> train_qwenlatent.py:487 + Step 23300 | grad_norm_pre_clip=0.1730 | + grad_norm_pre_clip_avg=0.1936 | Metrics: + {'align_loss': 0.02486485242843628, + 'recon_loss': 0.06393323838710785, + 'predict_loss': 0.008813267573714256, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17297327518463135, + 'mae_score': 0.011786856092848219, 'data_time': + 0.0006177090108394623, 'model_time': + 1.173424092005007, 'grad_norm_pre_clip_avg': + 0.19357255101203918, 'learning_rate': + 1.612178456434409e-05, 'epoch': 5.88} +04/19 [19:47:37] INFO | >> train_qwenlatent.py:487 + Step 23310 | grad_norm_pre_clip=0.1934 | + grad_norm_pre_clip_avg=0.1691 | Metrics: + {'align_loss': 0.024456560611724854, + 'recon_loss': 0.046443261206150055, + 'predict_loss': 0.007688083685934544, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19343845546245575, + 'data_time': 0.0005549060006160289, + 'model_time': 1.146418446995085, + 'grad_norm_pre_clip_avg': 0.1691217228770256, + 'learning_rate': 1.6113437851037186e-05, + 'epoch': 5.88} +04/19 [19:48:30] INFO | >> train_qwenlatent.py:487 + Step 23320 | grad_norm_pre_clip=0.1761 | + grad_norm_pre_clip_avg=0.1994 | Metrics: + {'align_loss': 0.025160424411296844, + 'recon_loss': 0.06850388646125793, + 'predict_loss': 0.0077514504082500935, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17609339952468872, + 'data_time': 0.001295422000112012, + 'model_time': 3.6253042909957003, + 'grad_norm_pre_clip_avg': 0.1993682637810707, + 'learning_rate': 1.6105089382676915e-05, + 'epoch': 5.88} +04/19 [19:49:06] INFO | >> train_qwenlatent.py:487 + Step 23330 | grad_norm_pre_clip=0.1381 | + grad_norm_pre_clip_avg=0.1775 | Metrics: + {'align_loss': 0.02550850622355938, + 'recon_loss': 0.07834277302026749, + 'predict_loss': 0.011083625257015228, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1380578875541687, + 'data_time': 0.0012292570027057081, + 'model_time': 3.7890869039983954, + 'grad_norm_pre_clip_avg': 0.1774693936109543, + 'learning_rate': 1.6096739163332226e-05, + 'epoch': 5.89} +04/19 [19:49:42] INFO | >> train_qwenlatent.py:487 + Step 23340 | grad_norm_pre_clip=0.2015 | + grad_norm_pre_clip_avg=0.2014 | Metrics: + {'align_loss': 0.025339238345623016, + 'recon_loss': 0.06581012159585953, + 'predict_loss': 0.007481361273676157, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20154766738414764, + 'data_time': 0.001211950002470985, + 'model_time': 3.617131863022223, + 'grad_norm_pre_clip_avg': 0.20139836817979812, + 'learning_rate': 1.6088387197072906e-05, + 'epoch': 5.89} +04/19 [19:50:21] INFO | >> train_qwenlatent.py:487 + Step 23350 | grad_norm_pre_clip=0.1898 | + grad_norm_pre_clip_avg=0.1991 | Metrics: + {'align_loss': 0.023751884698867798, + 'recon_loss': 0.059327129274606705, + 'predict_loss': 0.006657189689576626, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1897585690021515, + 'mae_score': 0.01823524268897804, 'data_time': + 0.0016576090129092336, 'model_time': + 3.2931283060170244, 'grad_norm_pre_clip_avg': + 0.19908590018749237, 'learning_rate': + 1.6080033487969605e-05, 'epoch': 5.89} +04/19 [19:50:59] INFO | >> train_qwenlatent.py:487 + Step 23360 | grad_norm_pre_clip=0.1603 | + grad_norm_pre_clip_avg=0.1890 | Metrics: + {'align_loss': 0.02525518275797367, + 'recon_loss': 0.0870513767004013, + 'predict_loss': 0.013152416795492172, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16029399633407593, + 'data_time': 0.0014738280151505023, + 'model_time': 3.7509137950255536, + 'grad_norm_pre_clip_avg': 0.18904170542955398, + 'learning_rate': 1.6071678040093827e-05, + 'epoch': 5.89} +04/19 [19:51:34] INFO | >> train_qwenlatent.py:487 + Step 23370 | grad_norm_pre_clip=0.1789 | + grad_norm_pre_clip_avg=0.1875 | Metrics: + {'align_loss': 0.025571785867214203, + 'recon_loss': 0.08700837939977646, + 'predict_loss': 0.0124571043998003, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1789141297340393, + 'data_time': 0.004339711013017222, + 'model_time': 3.4843795509950723, + 'grad_norm_pre_clip_avg': 0.18754862397909164, + 'learning_rate': 1.6063320857517908e-05, + 'epoch': 5.9} +04/19 [19:52:03] INFO | >> train_qwenlatent.py:487 + Step 23380 | grad_norm_pre_clip=0.1984 | + grad_norm_pre_clip_avg=0.2389 | Metrics: + {'align_loss': 0.026796799153089523, + 'recon_loss': 0.09500116109848022, + 'predict_loss': 0.011024645529687405, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19837406277656555, + 'data_time': 0.0011351449938956648, + 'model_time': 2.4332167230022606, + 'grad_norm_pre_clip_avg': 0.23886859714984893, + 'learning_rate': 1.6054961944315038e-05, + 'epoch': 5.9} +04/19 [19:52:27] INFO | >> train_qwenlatent.py:487 + Step 23390 | grad_norm_pre_clip=0.1930 | + grad_norm_pre_clip_avg=0.2017 | Metrics: + {'align_loss': 0.026031337678432465, + 'recon_loss': 0.0784195065498352, + 'predict_loss': 0.006097452715039253, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1929773986339569, + 'data_time': 0.001988804986467585, + 'model_time': 2.201953717012657, + 'grad_norm_pre_clip_avg': 0.2016886979341507, + 'learning_rate': 1.604660130455925e-05, + 'epoch': 5.9} +04/19 [19:52:44] INFO | >> train_qwenlatent.py:487 + Step 23400 | grad_norm_pre_clip=0.1727 | + grad_norm_pre_clip_avg=0.1995 | Metrics: + {'align_loss': 0.025713138282299042, + 'recon_loss': 0.07872572541236877, + 'predict_loss': 0.00829161424189806, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.172733873128891, + 'mae_score': 0.011620390093004382, 'data_time': + 0.0010610040044412017, 'model_time': + 1.5119007489993237, 'grad_norm_pre_clip_avg': + 0.19952789694070816, 'learning_rate': + 1.6038238942325423e-05, 'epoch': 5.9} +04/19 [19:52:57] INFO | >> train_qwenlatent.py:487 + Step 23410 | grad_norm_pre_clip=0.1839 | + grad_norm_pre_clip_avg=0.2079 | Metrics: + {'align_loss': 0.024882255122065544, + 'recon_loss': 0.09579315781593323, + 'predict_loss': 0.011688927188515663, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1839262843132019, + 'data_time': 0.0013527339906431735, + 'model_time': 1.3014444219879806, + 'grad_norm_pre_clip_avg': 0.20790492594242097, + 'learning_rate': 1.6029874861689263e-05, + 'epoch': 5.91} +04/19 [19:53:09] INFO | >> train_qwenlatent.py:487 + Step 23420 | grad_norm_pre_clip=0.1843 | + grad_norm_pre_clip_avg=0.1740 | Metrics: + {'align_loss': 0.025556642562150955, + 'recon_loss': 0.0837731882929802, + 'predict_loss': 0.012559687718749046, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18430998921394348, + 'data_time': 0.0006308970041573048, + 'model_time': 1.2214760489878245, + 'grad_norm_pre_clip_avg': 0.17401130944490434, + 'learning_rate': 1.6021509066727325e-05, + 'epoch': 5.91} +04/19 [19:53:22] INFO | >> train_qwenlatent.py:487 + Step 23430 | grad_norm_pre_clip=0.1970 | + grad_norm_pre_clip_avg=0.2260 | Metrics: + {'align_loss': 0.02722901478409767, + 'recon_loss': 0.09843574464321136, + 'predict_loss': 0.009761295281350613, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19701580703258514, + 'data_time': 0.000582745997235179, + 'model_time': 1.2392991249798797, + 'grad_norm_pre_clip_avg': 0.22600921243429184, + 'learning_rate': 1.6013141561517e-05, 'epoch': + 5.91} +04/19 [19:53:35] INFO | >> train_qwenlatent.py:487 + Step 23440 | grad_norm_pre_clip=0.2785 | + grad_norm_pre_clip_avg=0.1967 | Metrics: + {'align_loss': 0.025715187191963196, + 'recon_loss': 0.084229476749897, + 'predict_loss': 0.008580351248383522, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2785091996192932, + 'data_time': 0.0005952329956926405, + 'model_time': 1.2602356220013462, + 'grad_norm_pre_clip_avg': 0.1966872036457062, + 'learning_rate': 1.6004772350136495e-05, + 'epoch': 5.91} +04/19 [19:53:48] INFO | >> train_qwenlatent.py:487 + Step 23450 | grad_norm_pre_clip=0.2241 | + grad_norm_pre_clip_avg=0.1877 | Metrics: + {'align_loss': 0.0250270776450634, + 'recon_loss': 0.09164292365312576, + 'predict_loss': 0.008451334200799465, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22411233186721802, + 'mae_score': 0.011816636506501619, 'data_time': + 0.0005852260219398886, 'model_time': + 1.235997383017093, 'grad_norm_pre_clip_avg': + 0.18772078603506087, 'learning_rate': + 1.599640143666488e-05, 'epoch': 5.92} +04/19 [19:54:01] INFO | >> train_qwenlatent.py:487 + Step 23460 | grad_norm_pre_clip=0.1525 | + grad_norm_pre_clip_avg=0.1862 | Metrics: + {'align_loss': 0.02483162097632885, + 'recon_loss': 0.07574553042650223, + 'predict_loss': 0.010899917222559452, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15252631902694702, + 'data_time': 0.0008469699823763222, + 'model_time': 1.2272682109905872, + 'grad_norm_pre_clip_avg': 0.18617339432239532, + 'learning_rate': 1.598802882518202e-05, + 'epoch': 5.92} +04/19 [19:54:14] INFO | >> train_qwenlatent.py:487 + Step 23470 | grad_norm_pre_clip=0.1744 | + grad_norm_pre_clip_avg=0.1755 | Metrics: + {'align_loss': 0.0261942557990551, + 'recon_loss': 0.08216585963964462, + 'predict_loss': 0.005865626968443394, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17440484464168549, + 'data_time': 0.0007344339974224567, + 'model_time': 1.2299894019961357, + 'grad_norm_pre_clip_avg': 0.17549704760313034, + 'learning_rate': 1.5979654519768635e-05, + 'epoch': 5.92} +04/19 [19:54:26] INFO | >> train_qwenlatent.py:487 + Step 23480 | grad_norm_pre_clip=0.2048 | + grad_norm_pre_clip_avg=0.1846 | Metrics: + {'align_loss': 0.026003830134868622, + 'recon_loss': 0.11796204745769501, + 'predict_loss': 0.009438540786504745, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20479656755924225, + 'data_time': 0.0005969829799141735, + 'model_time': 1.2356479349837173, + 'grad_norm_pre_clip_avg': 0.18464283496141434, + 'learning_rate': 1.5971278524506258e-05, + 'epoch': 5.92} +04/19 [19:54:38] INFO | >> train_qwenlatent.py:487 + Step 23490 | grad_norm_pre_clip=0.3463 | + grad_norm_pre_clip_avg=0.2460 | Metrics: + {'align_loss': 0.025622770190238953, + 'recon_loss': 0.0810026302933693, + 'predict_loss': 0.012674516998231411, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.34629541635513306, + 'data_time': 0.0010576049971859902, + 'model_time': 1.2724072089768015, + 'grad_norm_pre_clip_avg': 0.2460413932800293, + 'learning_rate': 1.5962900843477248e-05, + 'epoch': 5.93} +04/19 [19:54:52] INFO | >> train_qwenlatent.py:487 + Step 23500 | grad_norm_pre_clip=0.1688 | + grad_norm_pre_clip_avg=0.2025 | Metrics: + {'align_loss': 0.025654684752225876, + 'recon_loss': 0.0801435261964798, + 'predict_loss': 0.014813573099672794, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16875006258487701, + 'mae_score': 0.00963673548655467, 'data_time': + 0.0006235280015971512, 'model_time': + 1.2188589549914468, 'grad_norm_pre_clip_avg': + 0.20251343101263047, 'learning_rate': + 1.595452148076478e-05, 'epoch': 5.93} +04/19 [19:55:04] INFO | >> train_qwenlatent.py:487 + Step 23510 | grad_norm_pre_clip=0.1718 | + grad_norm_pre_clip_avg=0.1935 | Metrics: + {'align_loss': 0.02602703869342804, + 'recon_loss': 0.1153859794139862, + 'predict_loss': 0.014310472644865513, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17181198298931122, + 'data_time': 0.0008479040116071701, + 'model_time': 1.2017025019740686, + 'grad_norm_pre_clip_avg': 0.1935070961713791, + 'learning_rate': 1.5946140440452856e-05, + 'epoch': 5.93} +04/19 [19:55:17] INFO | >> train_qwenlatent.py:487 + Step 23520 | grad_norm_pre_clip=0.1691 | + grad_norm_pre_clip_avg=0.1710 | Metrics: + {'align_loss': 0.025585129857063293, + 'recon_loss': 0.08627307415008545, + 'predict_loss': 0.014341573230922222, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16906382143497467, + 'data_time': 0.0006116099830251187, + 'model_time': 1.2504540620138869, + 'grad_norm_pre_clip_avg': 0.170971117913723, + 'learning_rate': 1.5937757726626296e-05, + 'epoch': 5.93} +04/19 [19:55:29] INFO | >> train_qwenlatent.py:487 + Step 23530 | grad_norm_pre_clip=0.1489 | + grad_norm_pre_clip_avg=0.1591 | Metrics: + {'align_loss': 0.02356332540512085, + 'recon_loss': 0.06398520618677139, + 'predict_loss': 0.0093589648604393, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14886946976184845, + 'data_time': 0.0007183830020949244, + 'model_time': 1.261181159003172, + 'grad_norm_pre_clip_avg': 0.1591079369187355, + 'learning_rate': 1.5929373343370724e-05, + 'epoch': 5.94} +04/19 [19:55:42] INFO | >> train_qwenlatent.py:487 + Step 23540 | grad_norm_pre_clip=0.2291 | + grad_norm_pre_clip_avg=0.2147 | Metrics: + {'align_loss': 0.02640867978334427, + 'recon_loss': 0.09886135160923004, + 'predict_loss': 0.011359690688550472, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22913803160190582, + 'data_time': 0.0009325669961981475, + 'model_time': 1.2265528870048001, + 'grad_norm_pre_clip_avg': 0.21466998606920243, + 'learning_rate': 1.5920987294772602e-05, + 'epoch': 5.94} +04/19 [19:55:55] INFO | >> train_qwenlatent.py:487 + Step 23550 | grad_norm_pre_clip=0.1308 | + grad_norm_pre_clip_avg=0.1878 | Metrics: + {'align_loss': 0.02531733363866806, + 'recon_loss': 0.07652655243873596, + 'predict_loss': 0.012015684507787228, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13082627952098846, + 'mae_score': 0.010092539400667757, 'data_time': + 0.0008720390032976866, 'model_time': + 1.2492073269968387, 'grad_norm_pre_clip_avg': + 0.18776293396949767, 'learning_rate': + 1.5912599584919172e-05, 'epoch': 5.94} +04/19 [19:56:08] INFO | >> train_qwenlatent.py:487 + Step 23560 | grad_norm_pre_clip=0.2264 | + grad_norm_pre_clip_avg=0.2158 | Metrics: + {'align_loss': 0.026927292346954346, + 'recon_loss': 0.10020924359560013, + 'predict_loss': 0.013529045507311821, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22642183303833008, + 'data_time': 0.0006730369932483882, + 'model_time': 1.1968784299970139, + 'grad_norm_pre_clip_avg': 0.2158092513680458, + 'learning_rate': 1.5904210217898514e-05, + 'epoch': 5.94} +04/19 [19:56:20] INFO | >> train_qwenlatent.py:487 + Step 23570 | grad_norm_pre_clip=0.2170 | + grad_norm_pre_clip_avg=0.1836 | Metrics: + {'align_loss': 0.024400347843766212, + 'recon_loss': 0.06819437444210052, + 'predict_loss': 0.010520556941628456, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2170383185148239, + 'data_time': 0.000602366984821856, + 'model_time': 1.2159023630083539, + 'grad_norm_pre_clip_avg': 0.18357160389423371, + 'learning_rate': 1.5895819197799496e-05, + 'epoch': 5.95} +04/19 [19:56:33] INFO | >> train_qwenlatent.py:487 + Step 23580 | grad_norm_pre_clip=0.1536 | + grad_norm_pre_clip_avg=0.1783 | Metrics: + {'align_loss': 0.02572353184223175, + 'recon_loss': 0.07822209596633911, + 'predict_loss': 0.00766258779913187, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15356549620628357, + 'data_time': 0.0008973289804998785, + 'model_time': 1.2023871369892731, + 'grad_norm_pre_clip_avg': 0.17833918780088426, + 'learning_rate': 1.5887426528711804e-05, + 'epoch': 5.95} +04/19 [19:56:46] INFO | >> train_qwenlatent.py:487 + Step 23590 | grad_norm_pre_clip=0.1575 | + grad_norm_pre_clip_avg=0.1652 | Metrics: + {'align_loss': 0.026122843846678734, + 'recon_loss': 0.07303333282470703, + 'predict_loss': 0.007713592145591974, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15752582252025604, + 'data_time': 0.0008075480000115931, + 'model_time': 1.2253419650078285, + 'grad_norm_pre_clip_avg': 0.16518382132053375, + 'learning_rate': 1.587903221472592e-05, + 'epoch': 5.95} +04/19 [19:56:59] INFO | >> train_qwenlatent.py:487 + Step 23600 | grad_norm_pre_clip=0.1832 | + grad_norm_pre_clip_avg=0.2189 | Metrics: + {'align_loss': 0.025308599695563316, + 'recon_loss': 0.07759813219308853, + 'predict_loss': 0.010505088604986668, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1832379549741745, + 'mae_score': 0.017574080046232756, 'data_time': + 0.0007002360071055591, 'model_time': + 1.197394599992549, 'grad_norm_pre_clip_avg': + 0.21891309916973115, 'learning_rate': + 1.587063625993313e-05, 'epoch': 5.96} +04/19 [19:57:12] INFO | >> train_qwenlatent.py:487 + Step 23610 | grad_norm_pre_clip=0.2089 | + grad_norm_pre_clip_avg=0.1779 | Metrics: + {'align_loss': 0.02560316026210785, + 'recon_loss': 0.12000007182359695, + 'predict_loss': 0.016053060069680214, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2088991403579712, + 'data_time': 0.0005464710120577365, + 'model_time': 1.2390200840018224, + 'grad_norm_pre_clip_avg': 0.17788513004779816, + 'learning_rate': 1.5862238668425526e-05, + 'epoch': 5.96} +04/19 [19:57:24] INFO | >> train_qwenlatent.py:487 + Step 23620 | grad_norm_pre_clip=0.1667 | + grad_norm_pre_clip_avg=0.1776 | Metrics: + {'align_loss': 0.026075121015310287, + 'recon_loss': 0.08408805727958679, + 'predict_loss': 0.009560367092490196, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1666589230298996, + 'data_time': 0.0006472980021499097, + 'model_time': 1.2125172929954715, + 'grad_norm_pre_clip_avg': 0.17759644091129304, + 'learning_rate': 1.5853839444295985e-05, + 'epoch': 5.96} +04/19 [19:57:37] INFO | >> train_qwenlatent.py:487 + Step 23630 | grad_norm_pre_clip=0.1869 | + grad_norm_pre_clip_avg=0.2014 | Metrics: + {'align_loss': 0.026540931314229965, + 'recon_loss': 0.08542212843894958, + 'predict_loss': 0.011227762326598167, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18685653805732727, + 'data_time': 0.0009211669967044145, + 'model_time': 1.2955924730049446, + 'grad_norm_pre_clip_avg': 0.2014071375131607, + 'learning_rate': 1.5845438591638192e-05, + 'epoch': 5.96} +04/19 [19:57:49] INFO | >> train_qwenlatent.py:487 + Step 23640 | grad_norm_pre_clip=0.2473 | + grad_norm_pre_clip_avg=0.2003 | Metrics: + {'align_loss': 0.025690998882055283, + 'recon_loss': 0.07611985504627228, + 'predict_loss': 0.006771350745111704, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24729308485984802, + 'data_time': 0.0009671369916759431, + 'model_time': 1.2704568940098397, + 'grad_norm_pre_clip_avg': 0.20032486468553543, + 'learning_rate': 1.5837036114546616e-05, + 'epoch': 5.97} +04/19 [19:58:02] INFO | >> train_qwenlatent.py:487 + Step 23650 | grad_norm_pre_clip=0.1711 | + grad_norm_pre_clip_avg=0.1949 | Metrics: + {'align_loss': 0.025886569172143936, + 'recon_loss': 0.09216915816068649, + 'predict_loss': 0.014183616265654564, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17107921838760376, + 'mae_score': 0.019449577675209388, 'data_time': + 0.0006255590124055743, 'model_time': + 1.1761026319873054, 'grad_norm_pre_clip_avg': + 0.1948737233877182, 'learning_rate': + 1.5828632017116527e-05, 'epoch': 5.97} +04/19 [19:58:15] INFO | >> train_qwenlatent.py:487 + Step 23660 | grad_norm_pre_clip=0.2344 | + grad_norm_pre_clip_avg=0.2072 | Metrics: + {'align_loss': 0.025788187980651855, + 'recon_loss': 0.08713986724615097, + 'predict_loss': 0.010504286736249924, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2343902289867401, + 'data_time': 0.0011107290047220886, + 'model_time': 1.2335605960106477, + 'grad_norm_pre_clip_avg': 0.20718003809452057, + 'learning_rate': 1.5820226303443976e-05, + 'epoch': 5.97} +04/19 [19:58:28] INFO | >> train_qwenlatent.py:487 + Step 23670 | grad_norm_pre_clip=0.2469 | + grad_norm_pre_clip_avg=0.2059 | Metrics: + {'align_loss': 0.025047190487384796, + 'recon_loss': 0.09137823432683945, + 'predict_loss': 0.0092513682320714, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24691934883594513, + 'data_time': 0.0008298289903905243, + 'model_time': 1.2600238300219644, + 'grad_norm_pre_clip_avg': 0.20594456940889358, + 'learning_rate': 1.5811818977625806e-05, + 'epoch': 5.97} +04/19 [19:58:40] INFO | >> train_qwenlatent.py:487 + Step 23680 | grad_norm_pre_clip=0.2018 | + grad_norm_pre_clip_avg=0.1878 | Metrics: + {'align_loss': 0.024688607081770897, + 'recon_loss': 0.06301529705524445, + 'predict_loss': 0.012301025912165642, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20178835093975067, + 'data_time': 0.0008700060134287924, + 'model_time': 1.2425364649971016, + 'grad_norm_pre_clip_avg': 0.18777153193950652, + 'learning_rate': 1.5803410043759652e-05, + 'epoch': 5.98} +04/19 [19:58:53] INFO | >> train_qwenlatent.py:487 + Step 23690 | grad_norm_pre_clip=0.1732 | + grad_norm_pre_clip_avg=0.1788 | Metrics: + {'align_loss': 0.024714045226573944, + 'recon_loss': 0.08254504948854446, + 'predict_loss': 0.012758875265717506, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17323675751686096, + 'data_time': 0.0006127860106062144, + 'model_time': 1.2631305120012257, + 'grad_norm_pre_clip_avg': 0.17881762981414795, + 'learning_rate': 1.5794999505943917e-05, + 'epoch': 5.98} +04/19 [19:59:06] INFO | >> train_qwenlatent.py:487 + Step 23700 | grad_norm_pre_clip=0.1795 | + grad_norm_pre_clip_avg=0.1893 | Metrics: + {'align_loss': 0.025987885892391205, + 'recon_loss': 0.09944920241832733, + 'predict_loss': 0.012253433465957642, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17951254546642303, + 'mae_score': 0.01425743360777159, 'data_time': + 0.001037181995343417, 'model_time': + 1.2517509709869046, 'grad_norm_pre_clip_avg': + 0.18931452482938765, 'learning_rate': + 1.57865873682778e-05, 'epoch': 5.98} +04/19 [19:59:18] INFO | >> train_qwenlatent.py:487 + Step 23710 | grad_norm_pre_clip=0.1804 | + grad_norm_pre_clip_avg=0.2024 | Metrics: + {'align_loss': 0.026159632951021194, + 'recon_loss': 0.12146364897489548, + 'predict_loss': 0.016869282349944115, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18039029836654663, + 'data_time': 0.0007282260048668832, + 'model_time': 1.2327600010030437, + 'grad_norm_pre_clip_avg': 0.202364219725132, + 'learning_rate': 1.5778173634861274e-05, + 'epoch': 5.98} +04/19 [19:59:31] INFO | >> train_qwenlatent.py:487 + Step 23720 | grad_norm_pre_clip=0.2199 | + grad_norm_pre_clip_avg=0.2139 | Metrics: + {'align_loss': 0.025887073948979378, + 'recon_loss': 0.08411893248558044, + 'predict_loss': 0.009322530589997768, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21993222832679749, + 'data_time': 0.0006187459803186357, + 'model_time': 1.445888895977987, + 'grad_norm_pre_clip_avg': 0.21389212012290953, + 'learning_rate': 1.576975830979509e-05, + 'epoch': 5.99} +04/19 [19:59:44] INFO | >> train_qwenlatent.py:487 + Step 23730 | grad_norm_pre_clip=0.2422 | + grad_norm_pre_clip_avg=0.2139 | Metrics: + {'align_loss': 0.024692352861166, 'recon_loss': + 0.0722990557551384, 'predict_loss': + 0.009454608894884586, 'aux_loss_decay_weight': + 0.0, 'grad_norm_pre_clip': 0.24221572279930115, + 'data_time': 0.0006340000254567713, + 'model_time': 1.2186954049975611, + 'grad_norm_pre_clip_avg': 0.21388615518808365, + 'learning_rate': 1.5761341397180776e-05, + 'epoch': 5.99} +04/19 [19:59:57] INFO | >> train_qwenlatent.py:487 + Step 23740 | grad_norm_pre_clip=0.1422 | + grad_norm_pre_clip_avg=0.1799 | Metrics: + {'align_loss': 0.024389078840613365, + 'recon_loss': 0.052992794662714005, + 'predict_loss': 0.006403107661753893, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14215487241744995, + 'data_time': 0.0007889910193625838, + 'model_time': 1.3389234270143788, + 'grad_norm_pre_clip_avg': 0.17991583347320556, + 'learning_rate': 1.575292290112063e-05, + 'epoch': 5.99} +04/19 [20:00:10] INFO | >> train_qwenlatent.py:487 + Step 23750 | grad_norm_pre_clip=0.1687 | + grad_norm_pre_clip_avg=0.1802 | Metrics: + {'align_loss': 0.025605740025639534, + 'recon_loss': 0.09634314477443695, + 'predict_loss': 0.01060874667018652, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16873568296432495, + 'mae_score': 0.01287316760501346, 'data_time': + 0.0005881690012756735, 'model_time': + 1.2090173400065396, 'grad_norm_pre_clip_avg': + 0.18024182617664336, 'learning_rate': + 1.5744502825717726e-05, 'epoch': 5.99} +04/19 [20:00:23] INFO | >> train_qwenlatent.py:487 + Step 23760 | grad_norm_pre_clip=0.1452 | + grad_norm_pre_clip_avg=0.1778 | Metrics: + {'align_loss': 0.025949344038963318, + 'recon_loss': 0.07132008671760559, + 'predict_loss': 0.00881971139460802, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14524884521961212, + 'data_time': 0.000655615993309766, + 'model_time': 1.2343815530184656, + 'grad_norm_pre_clip_avg': 0.1778373435139656, + 'learning_rate': 1.573608117507591e-05, + 'epoch': 6.0} +04/19 [20:00:36] INFO | >> train_qwenlatent.py:487 + Step 23770 | grad_norm_pre_clip=0.2106 | + grad_norm_pre_clip_avg=0.2007 | Metrics: + {'align_loss': 0.025990279391407967, + 'recon_loss': 0.10928542912006378, + 'predict_loss': 0.012578941881656647, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2105708122253418, + 'data_time': 0.0007536190096288919, + 'model_time': 1.2427252159977797, + 'grad_norm_pre_clip_avg': 0.20067923069000243, + 'learning_rate': 1.5727657953299794e-05, + 'epoch': 6.0} +04/19 [20:00:48] INFO | >> train_qwenlatent.py:487 + Step 23780 | grad_norm_pre_clip=0.1382 | + grad_norm_pre_clip_avg=0.1877 | Metrics: + {'align_loss': 0.02387443743646145, + 'recon_loss': 0.07003764808177948, + 'predict_loss': 0.005659013986587524, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13816595077514648, + 'data_time': 0.0006754170171916485, + 'model_time': 1.2232715879799798, + 'grad_norm_pre_clip_avg': 0.18772649616003037, + 'learning_rate': 1.5719233164494744e-05, + 'epoch': 6.0} +04/19 [20:01:01] INFO | >> train_qwenlatent.py:487 + Step 23790 | grad_norm_pre_clip=0.1500 | + grad_norm_pre_clip_avg=0.1880 | Metrics: + {'align_loss': 0.026106305420398712, + 'recon_loss': 0.07188859581947327, + 'predict_loss': 0.008652893826365471, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1500437706708908, + 'data_time': 0.0006483570032287389, + 'model_time': 1.222179584990954, + 'grad_norm_pre_clip_avg': 0.18804289251565934, + 'learning_rate': 1.5710806812766905e-05, + 'epoch': 6.0} +04/19 [20:01:14] INFO | >> train_qwenlatent.py:487 + Step 23800 | grad_norm_pre_clip=0.1351 | + grad_norm_pre_clip_avg=0.1894 | Metrics: + {'align_loss': 0.024548178538680077, + 'recon_loss': 0.08666025102138519, + 'predict_loss': 0.011452519334852695, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13510718941688538, + 'mae_score': 0.010301675023259344, 'data_time': + 0.0014617739943787456, 'model_time': + 1.2217493210046086, 'grad_norm_pre_clip_avg': + 0.18942146748304367, 'learning_rate': + 1.5702378902223182e-05, 'epoch': 6.01} +04/19 [20:01:27] INFO | >> train_qwenlatent.py:487 + Step 23810 | grad_norm_pre_clip=0.2060 | + grad_norm_pre_clip_avg=0.2098 | Metrics: + {'align_loss': 0.026283487677574158, + 'recon_loss': 0.09501359611749649, + 'predict_loss': 0.01360328309237957, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20600247383117676, + 'data_time': 0.0007704220188315958, + 'model_time': 1.2137711960240267, + 'grad_norm_pre_clip_avg': 0.20979374200105666, + 'learning_rate': 1.5693949436971236e-05, + 'epoch': 6.01} +04/19 [20:01:39] INFO | >> train_qwenlatent.py:487 + Step 23820 | grad_norm_pre_clip=0.1613 | + grad_norm_pre_clip_avg=0.1925 | Metrics: + {'align_loss': 0.024624839425086975, + 'recon_loss': 0.08414654433727264, + 'predict_loss': 0.010107534006237984, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16134069859981537, + 'data_time': 0.0009091779938898981, + 'model_time': 1.2542119860008825, + 'grad_norm_pre_clip_avg': 0.19250455498695374, + 'learning_rate': 1.568551842111948e-05, + 'epoch': 6.01} +04/19 [20:01:51] INFO | >> train_qwenlatent.py:487 + Step 23830 | grad_norm_pre_clip=0.1721 | + grad_norm_pre_clip_avg=0.1712 | Metrics: + {'align_loss': 0.024829840287566185, + 'recon_loss': 0.07539054751396179, + 'predict_loss': 0.008874219842255116, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17207299172878265, + 'data_time': 0.0012671580188907683, + 'model_time': 1.236419315013336, + 'grad_norm_pre_clip_avg': 0.17119611501693727, + 'learning_rate': 1.567708585877709e-05, + 'epoch': 6.01} +04/19 [20:02:04] INFO | >> train_qwenlatent.py:487 + Step 23840 | grad_norm_pre_clip=0.2598 | + grad_norm_pre_clip_avg=0.1887 | Metrics: + {'align_loss': 0.025603026151657104, + 'recon_loss': 0.06588395684957504, + 'predict_loss': 0.007211967837065458, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25982964038848877, + 'data_time': 0.000830375007353723, + 'model_time': 1.2783594969951082, + 'grad_norm_pre_clip_avg': 0.18868785202503205, + 'learning_rate': 1.5668651754054e-05, 'epoch': + 6.02} +04/19 [20:02:17] INFO | >> train_qwenlatent.py:487 + Step 23850 | grad_norm_pre_clip=0.1757 | + grad_norm_pre_clip_avg=0.2072 | Metrics: + {'align_loss': 0.02562353014945984, + 'recon_loss': 0.0611606240272522, + 'predict_loss': 0.009059218689799309, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17570842802524567, + 'mae_score': 0.010675652821858724, 'data_time': + 0.0006293740007095039, 'model_time': + 1.532082537014503, 'grad_norm_pre_clip_avg': + 0.2071862056851387, 'learning_rate': + 1.5660216111060887e-05, 'epoch': 6.02} +04/19 [20:02:30] INFO | >> train_qwenlatent.py:487 + Step 23860 | grad_norm_pre_clip=0.2086 | + grad_norm_pre_clip_avg=0.2332 | Metrics: + {'align_loss': 0.025159137323498726, + 'recon_loss': 0.07692854851484299, + 'predict_loss': 0.009450023993849754, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2086278796195984, + 'data_time': 0.0010549520084168762, + 'model_time': 1.6760453400202096, + 'grad_norm_pre_clip_avg': 0.23318203985691072, + 'learning_rate': 1.5651778933909174e-05, + 'epoch': 6.02} +04/19 [20:02:43] INFO | >> train_qwenlatent.py:487 + Step 23870 | grad_norm_pre_clip=0.2441 | + grad_norm_pre_clip_avg=0.2201 | Metrics: + {'align_loss': 0.02533390000462532, + 'recon_loss': 0.09401445835828781, + 'predict_loss': 0.017685366794466972, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24409648776054382, + 'data_time': 0.0007492940058000386, + 'model_time': 1.1887538479932118, + 'grad_norm_pre_clip_avg': 0.22005640119314193, + 'learning_rate': 1.5643340226711048e-05, + 'epoch': 6.02} +04/19 [20:02:55] INFO | >> train_qwenlatent.py:487 + Step 23880 | grad_norm_pre_clip=0.1724 | + grad_norm_pre_clip_avg=0.1787 | Metrics: + {'align_loss': 0.02528657205402851, + 'recon_loss': 0.10882553458213806, + 'predict_loss': 0.01795094460248947, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1724471151828766, + 'data_time': 0.0006557460001204163, + 'model_time': 1.244471987010911, + 'grad_norm_pre_clip_avg': 0.17874985784292222, + 'learning_rate': 1.563489999357943e-05, + 'epoch': 6.03} +04/19 [20:03:08] INFO | >> train_qwenlatent.py:487 + Step 23890 | grad_norm_pre_clip=0.2283 | + grad_norm_pre_clip_avg=0.1690 | Metrics: + {'align_loss': 0.024997485801577568, + 'recon_loss': 0.08864123374223709, + 'predict_loss': 0.011018248274922371, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2283363938331604, + 'data_time': 0.0009213599842041731, + 'model_time': 1.202310905995546, + 'grad_norm_pre_clip_avg': 0.1690099537372589, + 'learning_rate': 1.5626458238627984e-05, + 'epoch': 6.03} +04/19 [20:03:21] INFO | >> train_qwenlatent.py:487 + Step 23900 | grad_norm_pre_clip=0.2983 | + grad_norm_pre_clip_avg=0.2316 | Metrics: + {'align_loss': 0.025487329810857773, + 'recon_loss': 0.08483999967575073, + 'predict_loss': 0.010113927535712719, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2983011305332184, + 'mae_score': 0.00981426324930277, 'data_time': + 0.0006700470112264156, 'model_time': + 1.2106428630067967, 'grad_norm_pre_clip_avg': + 0.231591160595417, 'learning_rate': + 1.561801496597113e-05, 'epoch': 6.03} +04/19 [20:03:33] INFO | >> train_qwenlatent.py:487 + Step 23910 | grad_norm_pre_clip=0.1591 | + grad_norm_pre_clip_avg=0.1968 | Metrics: + {'align_loss': 0.025159340351819992, + 'recon_loss': 0.09759721159934998, + 'predict_loss': 0.01763903722167015, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.159139484167099, + 'data_time': 0.0009544359927531332, + 'model_time': 1.217632053012494, + 'grad_norm_pre_clip_avg': 0.1967783808708191, + 'learning_rate': 1.5609570179723995e-05, + 'epoch': 6.03} +04/19 [20:03:46] INFO | >> train_qwenlatent.py:487 + Step 23920 | grad_norm_pre_clip=0.1919 | + grad_norm_pre_clip_avg=0.1709 | Metrics: + {'align_loss': 0.025137566030025482, + 'recon_loss': 0.07762666791677475, + 'predict_loss': 0.008849719539284706, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19193565845489502, + 'data_time': 0.0006470259977504611, + 'model_time': 1.2455270740028936, + 'grad_norm_pre_clip_avg': 0.17094112783670426, + 'learning_rate': 1.5601123884002485e-05, + 'epoch': 6.04} +04/19 [20:03:58] INFO | >> train_qwenlatent.py:487 + Step 23930 | grad_norm_pre_clip=0.1712 | + grad_norm_pre_clip_avg=0.1692 | Metrics: + {'align_loss': 0.025224575772881508, + 'recon_loss': 0.09028070420026779, + 'predict_loss': 0.013964995741844177, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17118607461452484, + 'data_time': 0.0010748679924290627, + 'model_time': 1.1905194070131984, + 'grad_norm_pre_clip_avg': 0.1691687971353531, + 'learning_rate': 1.5592676082923216e-05, + 'epoch': 6.04} +04/19 [20:04:11] INFO | >> train_qwenlatent.py:487 + Step 23940 | grad_norm_pre_clip=0.1715 | + grad_norm_pre_clip_avg=0.1672 | Metrics: + {'align_loss': 0.026384565979242325, + 'recon_loss': 0.10141540318727493, + 'predict_loss': 0.01917782425880432, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17145560681819916, + 'data_time': 0.0007353950059041381, + 'model_time': 1.26319782299106, + 'grad_norm_pre_clip_avg': 0.16719283163547516, + 'learning_rate': 1.558422678060354e-05, + 'epoch': 6.04} +04/19 [20:04:24] INFO | >> train_qwenlatent.py:487 + Step 23950 | grad_norm_pre_clip=0.1925 | + grad_norm_pre_clip_avg=0.2222 | Metrics: + {'align_loss': 0.025151852518320084, + 'recon_loss': 0.07664142549037933, + 'predict_loss': 0.009069251827895641, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19245807826519012, + 'mae_score': 0.010508859479749525, 'data_time': + 0.0007640310213901103, 'model_time': + 1.2767092149879318, 'grad_norm_pre_clip_avg': + 0.2221628800034523, 'learning_rate': + 1.557577598116155e-05, 'epoch': 6.04} +04/19 [20:04:37] INFO | >> train_qwenlatent.py:487 + Step 23960 | grad_norm_pre_clip=0.2064 | + grad_norm_pre_clip_avg=0.2354 | Metrics: + {'align_loss': 0.026914168149232864, + 'recon_loss': 0.08004464209079742, + 'predict_loss': 0.007881815545260906, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20642989873886108, + 'data_time': 0.000784150994149968, + 'model_time': 1.2562840880127624, + 'grad_norm_pre_clip_avg': 0.2354157567024231, + 'learning_rate': 1.5567323688716057e-05, + 'epoch': 6.05} +04/19 [20:04:49] INFO | >> train_qwenlatent.py:487 + Step 23970 | grad_norm_pre_clip=0.1684 | + grad_norm_pre_clip_avg=0.1988 | Metrics: + {'align_loss': 0.025218388065695763, + 'recon_loss': 0.09507949650287628, + 'predict_loss': 0.009335326962172985, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16844259202480316, + 'data_time': 0.0007041399949230254, + 'model_time': 1.2668482500012033, + 'grad_norm_pre_clip_avg': 0.19876324236392975, + 'learning_rate': 1.555886990738661e-05, + 'epoch': 6.05} +04/19 [20:05:02] INFO | >> train_qwenlatent.py:487 + Step 23980 | grad_norm_pre_clip=0.1838 | + grad_norm_pre_clip_avg=0.1623 | Metrics: + {'align_loss': 0.025063589215278625, + 'recon_loss': 0.07238245755434036, + 'predict_loss': 0.011563829146325588, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1838354468345642, + 'data_time': 0.0006174799927975982, + 'model_time': 1.2133692439820152, + 'grad_norm_pre_clip_avg': 0.1622719794511795, + 'learning_rate': 1.5550414641293477e-05, + 'epoch': 6.05} +04/19 [20:05:15] INFO | >> train_qwenlatent.py:487 + Step 23990 | grad_norm_pre_clip=0.1755 | + grad_norm_pre_clip_avg=0.1704 | Metrics: + {'align_loss': 0.02470378950238228, + 'recon_loss': 0.07481342554092407, + 'predict_loss': 0.011836567893624306, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1754702925682068, + 'data_time': 0.0006625420064665377, + 'model_time': 1.5038673960079905, + 'grad_norm_pre_clip_avg': 0.17042664736509322, + 'learning_rate': 1.5541957894557652e-05, + 'epoch': 6.05} +04/19 [20:05:28] INFO | >> train_qwenlatent.py:487 + Step 24000 | grad_norm_pre_clip=0.1748 | + grad_norm_pre_clip_avg=0.1976 | Metrics: + {'align_loss': 0.0243149995803833, + 'recon_loss': 0.06622527539730072, + 'predict_loss': 0.007203437387943268, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17478306591510773, + 'mae_score': 0.012001343460770341, 'data_time': + 0.0007460120250470936, 'model_time': + 1.2662971499958076, 'grad_norm_pre_clip_avg': + 0.19761697947978973, 'learning_rate': + 1.5533499671300848e-05, 'epoch': 6.06} +04/19 [20:05:41] INFO | >> train_qwenlatent.py:487 + Step 24010 | grad_norm_pre_clip=0.1879 | + grad_norm_pre_clip_avg=0.2283 | Metrics: + {'align_loss': 0.025670604780316353, + 'recon_loss': 0.08513343334197998, + 'predict_loss': 0.008319171145558357, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18789930641651154, + 'data_time': 0.0008977690013125539, + 'model_time': 1.6079166569979861, + 'grad_norm_pre_clip_avg': 0.22834599018096924, + 'learning_rate': 1.552503997564551e-05, + 'epoch': 6.06} +04/19 [20:05:53] INFO | >> train_qwenlatent.py:487 + Step 24020 | grad_norm_pre_clip=0.2136 | + grad_norm_pre_clip_avg=0.2142 | Metrics: + {'align_loss': 0.02453501895070076, + 'recon_loss': 0.07618306577205658, + 'predict_loss': 0.009503401815891266, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21363839507102966, + 'data_time': 0.0008529979968443513, + 'model_time': 1.240319205011474, + 'grad_norm_pre_clip_avg': 0.21416076868772507, + 'learning_rate': 1.5516578811714787e-05, + 'epoch': 6.06} +04/19 [20:06:05] INFO | >> train_qwenlatent.py:487 + Step 24030 | grad_norm_pre_clip=0.1637 | + grad_norm_pre_clip_avg=0.1736 | Metrics: + {'align_loss': 0.025446563959121704, + 'recon_loss': 0.06308699399232864, + 'predict_loss': 0.007753415498882532, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16374565660953522, + 'data_time': 0.000601061008637771, + 'model_time': 1.2067245799989905, + 'grad_norm_pre_clip_avg': 0.17358945608139037, + 'learning_rate': 1.550811618363254e-05, + 'epoch': 6.06} +04/19 [20:06:18] INFO | >> train_qwenlatent.py:487 + Step 24040 | grad_norm_pre_clip=0.2431 | + grad_norm_pre_clip_avg=0.1933 | Metrics: + {'align_loss': 0.02447243593633175, + 'recon_loss': 0.075462207198143, + 'predict_loss': 0.012978480197489262, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24307017028331757, + 'data_time': 0.0005718829925172031, + 'model_time': 1.2325314959743991, + 'grad_norm_pre_clip_avg': 0.1933472454547882, + 'learning_rate': 1.5499652095523357e-05, + 'epoch': 6.07} +04/19 [20:06:31] INFO | >> train_qwenlatent.py:487 + Step 24050 | grad_norm_pre_clip=0.1836 | + grad_norm_pre_clip_avg=0.2085 | Metrics: + {'align_loss': 0.02531147003173828, + 'recon_loss': 0.08640409260988235, + 'predict_loss': 0.014224516227841377, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1836009919643402, + 'mae_score': 0.011233675157701647, 'data_time': + 0.0006516650028061122, 'model_time': + 1.2456974659871776, 'grad_norm_pre_clip_avg': + 0.20850362181663512, 'learning_rate': + 1.549118655151253e-05, 'epoch': 6.07} +04/19 [20:06:44] INFO | >> train_qwenlatent.py:487 + Step 24060 | grad_norm_pre_clip=0.1802 | + grad_norm_pre_clip_avg=0.1964 | Metrics: + {'align_loss': 0.025668248534202576, + 'recon_loss': 0.09900663048028946, + 'predict_loss': 0.0121918311342597, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1801634132862091, + 'data_time': 0.0008479889947921038, + 'model_time': 1.1915057250007521, + 'grad_norm_pre_clip_avg': 0.1964279294013977, + 'learning_rate': 1.5482719555726064e-05, + 'epoch': 6.07} +04/19 [20:06:56] INFO | >> train_qwenlatent.py:487 + Step 24070 | grad_norm_pre_clip=0.1551 | + grad_norm_pre_clip_avg=0.1758 | Metrics: + {'align_loss': 0.026063386350870132, + 'recon_loss': 0.08988845348358154, + 'predict_loss': 0.008910994976758957, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1551332324743271, + 'data_time': 0.0006301940011326224, + 'model_time': 1.2054370719997678, + 'grad_norm_pre_clip_avg': 0.1758282005786896, + 'learning_rate': 1.5474251112290668e-05, + 'epoch': 6.07} +04/19 [20:07:09] INFO | >> train_qwenlatent.py:487 + Step 24080 | grad_norm_pre_clip=0.1935 | + grad_norm_pre_clip_avg=0.1631 | Metrics: + {'align_loss': 0.02514166384935379, + 'recon_loss': 0.07928699254989624, + 'predict_loss': 0.008731787092983723, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1934541016817093, + 'data_time': 0.0006454219983424991, + 'model_time': 1.2059352430223953, + 'grad_norm_pre_clip_avg': 0.16312810480594636, + 'learning_rate': 1.5465781225333764e-05, + 'epoch': 6.08} +04/19 [20:07:21] INFO | >> train_qwenlatent.py:487 + Step 24090 | grad_norm_pre_clip=0.1683 | + grad_norm_pre_clip_avg=0.1940 | Metrics: + {'align_loss': 0.025954343378543854, + 'recon_loss': 0.09047199040651321, + 'predict_loss': 0.013774510473012924, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1683381050825119, + 'data_time': 0.0009115689899772406, + 'model_time': 1.266350156016415, + 'grad_norm_pre_clip_avg': 0.1939682424068451, + 'learning_rate': 1.5457309898983467e-05, + 'epoch': 6.08} +04/19 [20:07:34] INFO | >> train_qwenlatent.py:487 + Step 24100 | grad_norm_pre_clip=0.1825 | + grad_norm_pre_clip_avg=0.1679 | Metrics: + {'align_loss': 0.023504722863435745, + 'recon_loss': 0.08800160139799118, + 'predict_loss': 0.01584138348698616, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18249842524528503, + 'mae_score': 0.009383567603858741, 'data_time': + 0.0008877119980752468, 'model_time': + 1.2726892789942212, 'grad_norm_pre_clip_avg': + 0.1679116055369377, 'learning_rate': + 1.5448837137368593e-05, 'epoch': 6.08} +04/19 [20:07:47] INFO | >> train_qwenlatent.py:487 + Step 24110 | grad_norm_pre_clip=0.2250 | + grad_norm_pre_clip_avg=0.2366 | Metrics: + {'align_loss': 0.024863576516509056, + 'recon_loss': 0.07693726569414139, + 'predict_loss': 0.007672756910324097, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2249876856803894, + 'data_time': 0.0008391909941565245, + 'model_time': 1.2363036580209155, + 'grad_norm_pre_clip_avg': 0.2366091564297676, + 'learning_rate': 1.5440362944618673e-05, + 'epoch': 6.08} +04/19 [20:07:59] INFO | >> train_qwenlatent.py:487 + Step 24120 | grad_norm_pre_clip=0.1716 | + grad_norm_pre_clip_avg=0.2069 | Metrics: + {'align_loss': 0.025009233504533768, + 'recon_loss': 0.06351060420274734, + 'predict_loss': 0.007045479957014322, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17159168422222137, + 'data_time': 0.0006638439954258502, + 'model_time': 1.289297498005908, + 'grad_norm_pre_clip_avg': 0.2069370612502098, + 'learning_rate': 1.5431887324863923e-05, + 'epoch': 6.09} +04/19 [20:08:13] INFO | >> train_qwenlatent.py:487 + Step 24130 | grad_norm_pre_clip=0.1661 | + grad_norm_pre_clip_avg=0.1804 | Metrics: + {'align_loss': 0.02587099000811577, + 'recon_loss': 0.10159247368574142, + 'predict_loss': 0.012041089124977589, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16605502367019653, + 'data_time': 0.0006294040067587048, + 'model_time': 1.3114942829997744, + 'grad_norm_pre_clip_avg': 0.18038320243358613, + 'learning_rate': 1.5423410282235257e-05, + 'epoch': 6.09} +04/19 [20:08:25] INFO | >> train_qwenlatent.py:487 + Step 24140 | grad_norm_pre_clip=0.1821 | + grad_norm_pre_clip_avg=0.1884 | Metrics: + {'align_loss': 0.024161437526345253, + 'recon_loss': 0.047473467886447906, + 'predict_loss': 0.00655839079990983, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18206146359443665, + 'data_time': 0.0009039360156748444, + 'model_time': 1.2019144569931086, + 'grad_norm_pre_clip_avg': 0.18838165551424027, + 'learning_rate': 1.541493182086428e-05, + 'epoch': 6.09} +04/19 [20:08:39] INFO | >> train_qwenlatent.py:487 + Step 24150 | grad_norm_pre_clip=0.1282 | + grad_norm_pre_clip_avg=0.1821 | Metrics: + {'align_loss': 0.025379352271556854, + 'recon_loss': 0.08930592983961105, + 'predict_loss': 0.00908921379595995, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1282387375831604, + 'mae_score': 0.012071231464007954, 'data_time': + 0.0008036759973037988, 'model_time': + 1.2305432870052755, 'grad_norm_pre_clip_avg': + 0.18206876367330552, 'learning_rate': + 1.540645194488329e-05, 'epoch': 6.09} +04/19 [20:08:51] INFO | >> train_qwenlatent.py:487 + Step 24160 | grad_norm_pre_clip=0.1568 | + grad_norm_pre_clip_avg=0.1666 | Metrics: + {'align_loss': 0.025133851915597916, + 'recon_loss': 0.09675250202417374, + 'predict_loss': 0.012117543257772923, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15675735473632812, + 'data_time': 0.0009828860056586564, + 'model_time': 1.3339344669948332, + 'grad_norm_pre_clip_avg': 0.16662200018763543, + 'learning_rate': 1.5397970658425285e-05, + 'epoch': 6.1} +04/19 [20:09:04] INFO | >> train_qwenlatent.py:487 + Step 24170 | grad_norm_pre_clip=0.2893 | + grad_norm_pre_clip_avg=0.1949 | Metrics: + {'align_loss': 0.026430658996105194, + 'recon_loss': 0.08953199535608292, + 'predict_loss': 0.009290697053074837, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2892553210258484, + 'data_time': 0.0008449700253549963, + 'model_time': 1.2368499099975452, + 'grad_norm_pre_clip_avg': 0.19494670927524566, + 'learning_rate': 1.5389487965623926e-05, + 'epoch': 6.1} +04/19 [20:09:17] INFO | >> train_qwenlatent.py:487 + Step 24180 | grad_norm_pre_clip=0.1745 | + grad_norm_pre_clip_avg=0.2235 | Metrics: + {'align_loss': 0.024271618574857712, + 'recon_loss': 0.08973497897386551, + 'predict_loss': 0.011492452584207058, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17451676726341248, + 'data_time': 0.0011510580079630017, + 'model_time': 1.239171808003448, + 'grad_norm_pre_clip_avg': 0.22354859858751297, + 'learning_rate': 1.5381003870613588e-05, + 'epoch': 6.1} +04/19 [20:09:29] INFO | >> train_qwenlatent.py:487 + Step 24190 | grad_norm_pre_clip=0.1856 | + grad_norm_pre_clip_avg=0.1728 | Metrics: + {'align_loss': 0.023557396605610847, + 'recon_loss': 0.06731749325990677, + 'predict_loss': 0.013789950869977474, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18555192649364471, + 'data_time': 0.0009184349910356104, + 'model_time': 1.3376940120069776, + 'grad_norm_pre_clip_avg': 0.17277561277151107, + 'learning_rate': 1.537251837752931e-05, + 'epoch': 6.1} +04/19 [20:09:43] INFO | >> train_qwenlatent.py:487 + Step 24200 | grad_norm_pre_clip=0.2097 | + grad_norm_pre_clip_avg=0.2073 | Metrics: + {'align_loss': 0.025989770889282227, + 'recon_loss': 0.10155358910560608, + 'predict_loss': 0.012194170616567135, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20972679555416107, + 'mae_score': 0.010409658449190158, 'data_time': + 0.0009147379896603525, 'model_time': + 1.2794211379950866, 'grad_norm_pre_clip_avg': + 0.20726803243160247, 'learning_rate': + 1.5364031490506814e-05, 'epoch': 6.11} +04/19 [20:09:55] INFO | >> train_qwenlatent.py:487 + Step 24210 | grad_norm_pre_clip=0.1987 | + grad_norm_pre_clip_avg=0.1923 | Metrics: + {'align_loss': 0.025360219180583954, + 'recon_loss': 0.09931911528110504, + 'predict_loss': 0.01137255597859621, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19872815907001495, + 'data_time': 0.001050332997692749, + 'model_time': 1.2882187600189354, + 'grad_norm_pre_clip_avg': 0.19225671589374543, + 'learning_rate': 1.5355543213682516e-05, + 'epoch': 6.11} +04/19 [20:10:08] INFO | >> train_qwenlatent.py:487 + Step 24220 | grad_norm_pre_clip=0.2152 | + grad_norm_pre_clip_avg=0.1992 | Metrics: + {'align_loss': 0.026239316910505295, + 'recon_loss': 0.08272386342287064, + 'predict_loss': 0.00997084565460682, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21516482532024384, + 'data_time': 0.001069285994162783, + 'model_time': 1.264937625004677, + 'grad_norm_pre_clip_avg': 0.19918956905603408, + 'learning_rate': 1.5347053551193493e-05, + 'epoch': 6.11} +04/19 [20:10:20] INFO | >> train_qwenlatent.py:487 + Step 24230 | grad_norm_pre_clip=0.2006 | + grad_norm_pre_clip_avg=0.1953 | Metrics: + {'align_loss': 0.024433713406324387, + 'recon_loss': 0.058883097022771835, + 'predict_loss': 0.008610588498413563, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2006440907716751, + 'data_time': 0.001031018007779494, + 'model_time': 1.2109640349808615, + 'grad_norm_pre_clip_avg': 0.19534040242433548, + 'learning_rate': 1.53385625071775e-05, 'epoch': + 6.11} +04/19 [20:10:33] INFO | >> train_qwenlatent.py:487 + Step 24240 | grad_norm_pre_clip=0.1762 | + grad_norm_pre_clip_avg=0.1939 | Metrics: + {'align_loss': 0.02631516009569168, + 'recon_loss': 0.08476821333169937, + 'predict_loss': 0.012049656361341476, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17623423039913177, + 'data_time': 0.0006823200092185289, + 'model_time': 1.2033673460246064, + 'grad_norm_pre_clip_avg': 0.19392292946577072, + 'learning_rate': 1.5330070085772967e-05, + 'epoch': 6.12} +04/19 [20:10:46] INFO | >> train_qwenlatent.py:487 + Step 24250 | grad_norm_pre_clip=0.1738 | + grad_norm_pre_clip_avg=0.1779 | Metrics: + {'align_loss': 0.024042200297117233, + 'recon_loss': 0.08360979706048965, + 'predict_loss': 0.00815188605338335, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17375394701957703, + 'mae_score': 0.012093842996133341, 'data_time': + 0.0010014750005211681, 'model_time': + 1.3204112609964795, 'grad_norm_pre_clip_avg': + 0.17789223715662955, 'learning_rate': + 1.5321576291119017e-05, 'epoch': 6.12} +04/19 [20:10:59] INFO | >> train_qwenlatent.py:487 + Step 24260 | grad_norm_pre_clip=0.1697 | + grad_norm_pre_clip_avg=0.2198 | Metrics: + {'align_loss': 0.026013605296611786, + 'recon_loss': 0.06496207416057587, + 'predict_loss': 0.007869282737374306, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16969311237335205, + 'data_time': 0.0009005889878608286, + 'model_time': 1.4783159079961479, + 'grad_norm_pre_clip_avg': 0.21978457272052765, + 'learning_rate': 1.53130811273554e-05, 'epoch': + 6.12} +04/19 [20:11:12] INFO | >> train_qwenlatent.py:487 + Step 24270 | grad_norm_pre_clip=0.2287 | + grad_norm_pre_clip_avg=0.2033 | Metrics: + {'align_loss': 0.024958118796348572, + 'recon_loss': 0.08433196693658829, + 'predict_loss': 0.011948522180318832, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22869673371315002, + 'data_time': 0.00147765499423258, 'model_time': + 1.226003032992594, 'grad_norm_pre_clip_avg': + 0.2032538115978241, 'learning_rate': + 1.5304584598622564e-05, 'epoch': 6.12} +04/19 [20:11:24] INFO | >> train_qwenlatent.py:487 + Step 24280 | grad_norm_pre_clip=0.2235 | + grad_norm_pre_clip_avg=0.1794 | Metrics: + {'align_loss': 0.026542428880929947, + 'recon_loss': 0.10512549430131912, + 'predict_loss': 0.016260016709566116, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22347965836524963, + 'data_time': 0.0008860660018399358, + 'model_time': 1.267120532982517, + 'grad_norm_pre_clip_avg': 0.17935085892677308, + 'learning_rate': 1.5296086709061613e-05, + 'epoch': 6.13} +04/19 [20:11:37] INFO | >> train_qwenlatent.py:487 + Step 24290 | grad_norm_pre_clip=0.1926 | + grad_norm_pre_clip_avg=0.1734 | Metrics: + {'align_loss': 0.02644694596529007, + 'recon_loss': 0.101888008415699, + 'predict_loss': 0.012292473576962948, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1926030069589615, + 'data_time': 0.0010021060006693006, + 'model_time': 1.2707854729960673, + 'grad_norm_pre_clip_avg': 0.17337385714054107, + 'learning_rate': 1.5287587462814317e-05, + 'epoch': 6.13} +04/19 [20:11:50] INFO | >> train_qwenlatent.py:487 + Step 24300 | grad_norm_pre_clip=0.2604 | + grad_norm_pre_clip_avg=0.2203 | Metrics: + {'align_loss': 0.025271357968449593, + 'recon_loss': 0.08571358025074005, + 'predict_loss': 0.01026992965489626, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26035889983177185, + 'mae_score': 0.008035418364378783, 'data_time': + 0.000886716996319592, 'model_time': + 1.225541518011596, 'grad_norm_pre_clip_avg': + 0.22026179879903793, 'learning_rate': + 1.5279086864023103e-05, 'epoch': 6.13} +04/19 [20:12:02] INFO | >> train_qwenlatent.py:487 + Step 24310 | grad_norm_pre_clip=0.2078 | + grad_norm_pre_clip_avg=0.2065 | Metrics: + {'align_loss': 0.02588469162583351, + 'recon_loss': 0.08229576796293259, + 'predict_loss': 0.010044148191809654, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20780301094055176, + 'data_time': 0.0006385820161085576, + 'model_time': 1.2370127690082882, + 'grad_norm_pre_clip_avg': 0.20645476877689362, + 'learning_rate': 1.5270584916831056e-05, + 'epoch': 6.13} +04/19 [20:12:15] INFO | >> train_qwenlatent.py:487 + Step 24320 | grad_norm_pre_clip=0.2041 | + grad_norm_pre_clip_avg=0.1741 | Metrics: + {'align_loss': 0.025189530104398727, + 'recon_loss': 0.09773802757263184, + 'predict_loss': 0.017346009612083435, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20410658419132233, + 'data_time': 0.0009819089900702238, + 'model_time': 1.215885559009621, + 'grad_norm_pre_clip_avg': 0.1740705832839012, + 'learning_rate': 1.5262081625381927e-05, + 'epoch': 6.14} +04/19 [20:12:28] INFO | >> train_qwenlatent.py:487 + Step 24330 | grad_norm_pre_clip=0.1970 | + grad_norm_pre_clip_avg=0.1681 | Metrics: + {'align_loss': 0.02578263357281685, + 'recon_loss': 0.11170753836631775, + 'predict_loss': 0.016664117574691772, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1969919353723526, + 'data_time': 0.000703292986145243, + 'model_time': 1.2083881480211858, + 'grad_norm_pre_clip_avg': 0.16812903434038162, + 'learning_rate': 1.5253576993820123e-05, + 'epoch': 6.14} +04/19 [20:12:40] INFO | >> train_qwenlatent.py:487 + Step 24340 | grad_norm_pre_clip=0.2831 | + grad_norm_pre_clip_avg=0.1974 | Metrics: + {'align_loss': 0.025082070380449295, + 'recon_loss': 0.07298334687948227, + 'predict_loss': 0.009814836084842682, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28313153982162476, + 'data_time': 0.0006570620171260089, + 'model_time': 1.246684526995523, + 'grad_norm_pre_clip_avg': 0.19741206020116805, + 'learning_rate': 1.5245071026290686e-05, + 'epoch': 6.14} +04/19 [20:12:53] INFO | >> train_qwenlatent.py:487 + Step 24350 | grad_norm_pre_clip=0.1879 | + grad_norm_pre_clip_avg=0.2069 | Metrics: + {'align_loss': 0.024737529456615448, + 'recon_loss': 0.06807571649551392, + 'predict_loss': 0.008285530842840672, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18790927529335022, + 'mae_score': 0.009622561824214351, 'data_time': + 0.0008934150100685656, 'model_time': + 1.2459051010082476, 'grad_norm_pre_clip_avg': + 0.20691600441932678, 'learning_rate': + 1.523656372693933e-05, 'epoch': 6.14} +04/19 [20:13:06] INFO | >> train_qwenlatent.py:487 + Step 24360 | grad_norm_pre_clip=0.1540 | + grad_norm_pre_clip_avg=0.1755 | Metrics: + {'align_loss': 0.023266930133104324, + 'recon_loss': 0.08114088326692581, + 'predict_loss': 0.013288553804159164, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15400175750255585, + 'data_time': 0.0007457799802068621, + 'model_time': 1.2617537940095644, + 'grad_norm_pre_clip_avg': 0.17548442631959915, + 'learning_rate': 1.52280550999124e-05, 'epoch': + 6.15} +04/19 [20:13:18] INFO | >> train_qwenlatent.py:487 + Step 24370 | grad_norm_pre_clip=0.2200 | + grad_norm_pre_clip_avg=0.1939 | Metrics: + {'align_loss': 0.026221541687846184, + 'recon_loss': 0.08743053674697876, + 'predict_loss': 0.011972491629421711, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22001340985298157, + 'data_time': 0.0006877460109535605, + 'model_time': 1.224742590013193, + 'grad_norm_pre_clip_avg': 0.19390317499637605, + 'learning_rate': 1.5219545149356907e-05, + 'epoch': 6.15} +04/19 [20:13:31] INFO | >> train_qwenlatent.py:487 + Step 24380 | grad_norm_pre_clip=0.2719 | + grad_norm_pre_clip_avg=0.2172 | Metrics: + {'align_loss': 0.025952577590942383, + 'recon_loss': 0.10287181288003922, + 'predict_loss': 0.014788045547902584, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.27185481786727905, + 'data_time': 0.0010521580115891993, + 'model_time': 1.266357254004106, + 'grad_norm_pre_clip_avg': 0.21723465323448182, + 'learning_rate': 1.5211033879420495e-05, + 'epoch': 6.15} +04/19 [20:13:44] INFO | >> train_qwenlatent.py:487 + Step 24390 | grad_norm_pre_clip=0.2189 | + grad_norm_pre_clip_avg=0.2415 | Metrics: + {'align_loss': 0.024227237328886986, + 'recon_loss': 0.07835370302200317, + 'predict_loss': 0.011089607141911983, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21891814470291138, + 'data_time': 0.0009726329881232232, + 'model_time': 1.292486024001846, + 'grad_norm_pre_clip_avg': 0.2415455013513565, + 'learning_rate': 1.5202521294251446e-05, + 'epoch': 6.15} +04/19 [20:13:57] INFO | >> train_qwenlatent.py:487 + Step 24400 | grad_norm_pre_clip=0.1363 | + grad_norm_pre_clip_avg=0.1704 | Metrics: + {'align_loss': 0.024648290127515793, + 'recon_loss': 0.06624285876750946, + 'predict_loss': 0.009170479141175747, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1363079845905304, + 'mae_score': 0.009618115639901377, 'data_time': + 0.0006969480018597096, 'model_time': + 1.2589547430106904, 'grad_norm_pre_clip_avg': + 0.17044269293546677, 'learning_rate': + 1.5194007397998705e-05, 'epoch': 6.16} +04/19 [20:14:10] INFO | >> train_qwenlatent.py:487 + Step 24410 | grad_norm_pre_clip=0.2159 | + grad_norm_pre_clip_avg=0.1847 | Metrics: + {'align_loss': 0.02528296783566475, + 'recon_loss': 0.09195024520158768, + 'predict_loss': 0.00772239500656724, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21587564051151276, + 'data_time': 0.0009212199947796762, + 'model_time': 1.2201664480089676, + 'grad_norm_pre_clip_avg': 0.1846575453877449, + 'learning_rate': 1.5185492194811827e-05, + 'epoch': 6.16} +04/19 [20:14:22] INFO | >> train_qwenlatent.py:487 + Step 24420 | grad_norm_pre_clip=0.1605 | + grad_norm_pre_clip_avg=0.1910 | Metrics: + {'align_loss': 0.024265175685286522, + 'recon_loss': 0.0878693014383316, + 'predict_loss': 0.017402786761522293, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16051089763641357, + 'data_time': 0.0006811359780840576, + 'model_time': 1.2104335190088023, + 'grad_norm_pre_clip_avg': 0.1909884549677372, + 'learning_rate': 1.5176975688841027e-05, + 'epoch': 6.16} +04/19 [20:14:35] INFO | >> train_qwenlatent.py:487 + Step 24430 | grad_norm_pre_clip=0.1603 | + grad_norm_pre_clip_avg=0.1948 | Metrics: + {'align_loss': 0.025161638855934143, + 'recon_loss': 0.07245650142431259, + 'predict_loss': 0.010397602804005146, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1602674424648285, + 'data_time': 0.0012456510157790035, + 'model_time': 1.2836681210028473, + 'grad_norm_pre_clip_avg': 0.19476779103279113, + 'learning_rate': 1.5168457884237148e-05, + 'epoch': 6.16} +04/19 [20:14:47] INFO | >> train_qwenlatent.py:487 + Step 24440 | grad_norm_pre_clip=0.2061 | + grad_norm_pre_clip_avg=0.1755 | Metrics: + {'align_loss': 0.025013674050569534, + 'recon_loss': 0.06913454830646515, + 'predict_loss': 0.007318535353988409, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20614707469940186, + 'data_time': 0.0006179139891173691, + 'model_time': 1.2429755489865784, + 'grad_norm_pre_clip_avg': 0.1754554346203804, + 'learning_rate': 1.5159938785151662e-05, + 'epoch': 6.17} +04/19 [20:15:01] INFO | >> train_qwenlatent.py:487 + Step 24450 | grad_norm_pre_clip=0.1588 | + grad_norm_pre_clip_avg=0.2682 | Metrics: + {'align_loss': 0.02555977739393711, + 'recon_loss': 0.08339612931013107, + 'predict_loss': 0.01113323587924242, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.158763125538826, + 'mae_score': 0.026509783289454004, 'data_time': + 0.0008832349849399179, 'model_time': + 1.2623078039905522, 'grad_norm_pre_clip_avg': + 0.2682440966367722, 'learning_rate': + 1.5151418395736669e-05, 'epoch': 6.17} +04/19 [20:15:14] INFO | >> train_qwenlatent.py:487 + Step 24460 | grad_norm_pre_clip=0.2619 | + grad_norm_pre_clip_avg=0.2220 | Metrics: + {'align_loss': 0.024532150477170944, + 'recon_loss': 0.06877189874649048, + 'predict_loss': 0.007146644871681929, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26189011335372925, + 'data_time': 0.0008833069878164679, + 'model_time': 1.2954314340022393, + 'grad_norm_pre_clip_avg': 0.22203814387321472, + 'learning_rate': 1.5142896720144906e-05, + 'epoch': 6.17} +04/19 [20:15:26] INFO | >> train_qwenlatent.py:487 + Step 24470 | grad_norm_pre_clip=0.1805 | + grad_norm_pre_clip_avg=0.2113 | Metrics: + {'align_loss': 0.02553088404238224, + 'recon_loss': 0.08721069991588593, + 'predict_loss': 0.009076688438653946, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18046505749225616, + 'data_time': 0.0009449479985050857, + 'model_time': 1.2167923759843688, + 'grad_norm_pre_clip_avg': 0.21132721304893493, + 'learning_rate': 1.5134373762529744e-05, + 'epoch': 6.17} +04/19 [20:15:39] INFO | >> train_qwenlatent.py:487 + Step 24480 | grad_norm_pre_clip=0.1938 | + grad_norm_pre_clip_avg=0.1909 | Metrics: + {'align_loss': 0.026124686002731323, + 'recon_loss': 0.09548701345920563, + 'predict_loss': 0.010361257940530777, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1938002109527588, + 'data_time': 0.0009516819845885038, + 'model_time': 1.2301505010109395, + 'grad_norm_pre_clip_avg': 0.19086548388004304, + 'learning_rate': 1.5125849527045153e-05, + 'epoch': 6.18} +04/19 [20:15:51] INFO | >> train_qwenlatent.py:487 + Step 24490 | grad_norm_pre_clip=0.1728 | + grad_norm_pre_clip_avg=0.1666 | Metrics: + {'align_loss': 0.02624400705099106, + 'recon_loss': 0.08632272481918335, + 'predict_loss': 0.009786700829863548, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17277054488658905, + 'data_time': 0.0006584450020454824, + 'model_time': 1.214962400001241, + 'grad_norm_pre_clip_avg': 0.16658144146203996, + 'learning_rate': 1.5117324017845752e-05, + 'epoch': 6.18} +04/19 [20:16:04] INFO | >> train_qwenlatent.py:487 + Step 24500 | grad_norm_pre_clip=0.1966 | + grad_norm_pre_clip_avg=0.1782 | Metrics: + {'align_loss': 0.025520125404000282, + 'recon_loss': 0.07156729698181152, + 'predict_loss': 0.01170868519693613, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19660112261772156, + 'mae_score': 0.012277860040063256, 'data_time': + 0.0007019540062174201, 'model_time': + 1.2134239789738785, 'grad_norm_pre_clip_avg': + 0.17822541743516923, 'learning_rate': + 1.5108797239086769e-05, 'epoch': 6.18} +04/19 [20:16:17] INFO | >> train_qwenlatent.py:487 + Step 24510 | grad_norm_pre_clip=0.1487 | + grad_norm_pre_clip_avg=0.1927 | Metrics: + {'align_loss': 0.025518525391817093, + 'recon_loss': 0.08075761049985886, + 'predict_loss': 0.008660322055220604, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1486983299255371, + 'data_time': 0.0009700579976197332, + 'model_time': 1.2346896060043946, + 'grad_norm_pre_clip_avg': 0.1927422359585762, + 'learning_rate': 1.5100269194924052e-05, + 'epoch': 6.18} +04/19 [20:16:29] INFO | >> train_qwenlatent.py:487 + Step 24520 | grad_norm_pre_clip=0.2073 | + grad_norm_pre_clip_avg=0.1935 | Metrics: + {'align_loss': 0.024286745116114616, + 'recon_loss': 0.08840280026197433, + 'predict_loss': 0.015181830152869225, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2073126584291458, + 'data_time': 0.0010035070008598268, + 'model_time': 1.264636142004747, + 'grad_norm_pre_clip_avg': 0.19349970817565917, + 'learning_rate': 1.509173988951407e-05, + 'epoch': 6.19} +04/19 [20:16:42] INFO | >> train_qwenlatent.py:487 + Step 24530 | grad_norm_pre_clip=0.1573 | + grad_norm_pre_clip_avg=0.1706 | Metrics: + {'align_loss': 0.024779699742794037, + 'recon_loss': 0.11484244465827942, + 'predict_loss': 0.007579660974442959, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15725167095661163, + 'data_time': 0.0009422219882253557, + 'model_time': 1.181189363996964, + 'grad_norm_pre_clip_avg': 0.17055717259645461, + 'learning_rate': 1.5083209327013894e-05, + 'epoch': 6.19} +04/19 [20:16:55] INFO | >> train_qwenlatent.py:487 + Step 24540 | grad_norm_pre_clip=0.2392 | + grad_norm_pre_clip_avg=0.1891 | Metrics: + {'align_loss': 0.02432188391685486, + 'recon_loss': 0.0698523074388504, + 'predict_loss': 0.007991029880940914, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23922988772392273, + 'data_time': 0.0006605030212085694, + 'model_time': 1.1993533039931208, + 'grad_norm_pre_clip_avg': 0.1890722081065178, + 'learning_rate': 1.5074677511581222e-05, + 'epoch': 6.19} +04/19 [20:17:08] INFO | >> train_qwenlatent.py:487 + Step 24550 | grad_norm_pre_clip=0.2106 | + grad_norm_pre_clip_avg=0.2207 | Metrics: + {'align_loss': 0.02658771350979805, + 'recon_loss': 0.0757676512002945, + 'predict_loss': 0.006479436997324228, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2106436789035797, + 'mae_score': 0.009664316435117979, 'data_time': + 0.0006870430079288781, 'model_time': + 1.2675733069772832, 'grad_norm_pre_clip_avg': + 0.2207128718495369, 'learning_rate': + 1.5066144447374356e-05, 'epoch': 6.19} +04/19 [20:17:21] INFO | >> train_qwenlatent.py:487 + Step 24560 | grad_norm_pre_clip=0.1885 | + grad_norm_pre_clip_avg=0.1809 | Metrics: + {'align_loss': 0.025520851835608482, + 'recon_loss': 0.10325966775417328, + 'predict_loss': 0.011796196922659874, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18848133087158203, + 'data_time': 0.000870129995746538, + 'model_time': 1.2249742349958979, + 'grad_norm_pre_clip_avg': 0.1808608740568161, + 'learning_rate': 1.5057610138552214e-05, + 'epoch': 6.2} +04/19 [20:17:33] INFO | >> train_qwenlatent.py:487 + Step 24570 | grad_norm_pre_clip=0.1733 | + grad_norm_pre_clip_avg=0.1792 | Metrics: + {'align_loss': 0.023906704038381577, + 'recon_loss': 0.06655585020780563, + 'predict_loss': 0.008525912649929523, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17330846190452576, + 'data_time': 0.0007112179882824421, + 'model_time': 1.2518998959858436, + 'grad_norm_pre_clip_avg': 0.17922597080469133, + 'learning_rate': 1.5049074589274301e-05, + 'epoch': 6.2} +04/19 [20:17:46] INFO | >> train_qwenlatent.py:487 + Step 24580 | grad_norm_pre_clip=0.2615 | + grad_norm_pre_clip_avg=0.2147 | Metrics: + {'align_loss': 0.024448923766613007, + 'recon_loss': 0.06959753483533859, + 'predict_loss': 0.006503811106085777, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26149943470954895, + 'data_time': 0.0006804210133850574, + 'model_time': 1.2418484129884746, + 'grad_norm_pre_clip_avg': 0.21470462679862976, + 'learning_rate': 1.5040537803700751e-05, + 'epoch': 6.2} +04/19 [20:17:58] INFO | >> train_qwenlatent.py:487 + Step 24590 | grad_norm_pre_clip=0.1344 | + grad_norm_pre_clip_avg=0.2069 | Metrics: + {'align_loss': 0.024847377091646194, + 'recon_loss': 0.06902795284986496, + 'predict_loss': 0.007961980067193508, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1343913972377777, + 'data_time': 0.0008416170021519065, + 'model_time': 1.2367676130088512, + 'grad_norm_pre_clip_avg': 0.20688123404979705, + 'learning_rate': 1.5031999785992287e-05, + 'epoch': 6.2} +04/19 [20:18:12] INFO | >> train_qwenlatent.py:487 + Step 24600 | grad_norm_pre_clip=0.1904 | + grad_norm_pre_clip_avg=0.2018 | Metrics: + {'align_loss': 0.02431851252913475, + 'recon_loss': 0.08931999653577805, + 'predict_loss': 0.012107609771192074, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1904277354478836, + 'mae_score': 0.01185538403622739, 'data_time': + 0.0006733639747835696, 'model_time': + 1.2488115440064576, 'grad_norm_pre_clip_avg': + 0.20178064107894897, 'learning_rate': + 1.5023460540310235e-05, 'epoch': 6.21} +04/19 [20:18:24] INFO | >> train_qwenlatent.py:487 + Step 24610 | grad_norm_pre_clip=0.1645 | + grad_norm_pre_clip_avg=0.1756 | Metrics: + {'align_loss': 0.02495381236076355, + 'recon_loss': 0.08684493601322174, + 'predict_loss': 0.008973655290901661, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16447122395038605, + 'data_time': 0.0010657240054570138, + 'model_time': 1.2157030979869887, + 'grad_norm_pre_clip_avg': 0.17558372467756272, + 'learning_rate': 1.5014920070816519e-05, + 'epoch': 6.21} +04/19 [20:18:37] INFO | >> train_qwenlatent.py:487 + Step 24620 | grad_norm_pre_clip=0.2369 | + grad_norm_pre_clip_avg=0.1846 | Metrics: + {'align_loss': 0.024813741445541382, + 'recon_loss': 0.0853036418557167, + 'predict_loss': 0.00944527518004179, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23689910769462585, + 'data_time': 0.0009939740120898932, + 'model_time': 1.2207454529998358, + 'grad_norm_pre_clip_avg': 0.18463457375764847, + 'learning_rate': 1.5006378381673658e-05, + 'epoch': 6.21} +04/19 [20:18:49] INFO | >> train_qwenlatent.py:487 + Step 24630 | grad_norm_pre_clip=0.1452 | + grad_norm_pre_clip_avg=0.1882 | Metrics: + {'align_loss': 0.025016682222485542, + 'recon_loss': 0.08514142036437988, + 'predict_loss': 0.008349131792783737, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14522810280323029, + 'data_time': 0.0006750269967596978, + 'model_time': 1.2172754709899891, + 'grad_norm_pre_clip_avg': 0.1882401540875435, + 'learning_rate': 1.4997835477044773e-05, + 'epoch': 6.21} +04/19 [20:19:02] INFO | >> train_qwenlatent.py:487 + Step 24640 | grad_norm_pre_clip=0.1579 | + grad_norm_pre_clip_avg=0.1815 | Metrics: + {'align_loss': 0.026483802124857903, + 'recon_loss': 0.09258560836315155, + 'predict_loss': 0.008752535097301006, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1578739434480667, + 'data_time': 0.0006312369951047003, + 'model_time': 1.1917885850125458, + 'grad_norm_pre_clip_avg': 0.18150698244571686, + 'learning_rate': 1.4989291361093567e-05, + 'epoch': 6.22} +04/19 [20:19:14] INFO | >> train_qwenlatent.py:487 + Step 24650 | grad_norm_pre_clip=0.1685 | + grad_norm_pre_clip_avg=0.1620 | Metrics: + {'align_loss': 0.025745149701833725, + 'recon_loss': 0.08176369220018387, + 'predict_loss': 0.011161431670188904, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16848880052566528, + 'mae_score': 0.011105087211540153, 'data_time': + 0.0006662249797955155, 'model_time': + 1.2499517359829042, 'grad_norm_pre_clip_avg': + 0.16203615069389343, 'learning_rate': + 1.4980746037984338e-05, 'epoch': 6.22} +04/19 [20:19:27] INFO | >> train_qwenlatent.py:487 + Step 24660 | grad_norm_pre_clip=0.2466 | + grad_norm_pre_clip_avg=0.2190 | Metrics: + {'align_loss': 0.02587050013244152, + 'recon_loss': 0.09981614351272583, + 'predict_loss': 0.010576833970844746, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24657733738422394, + 'data_time': 0.000956101983319968, + 'model_time': 1.203726314008236, + 'grad_norm_pre_clip_avg': 0.21895313262939453, + 'learning_rate': 1.4972199511881979e-05, + 'epoch': 6.22} +04/19 [20:19:41] INFO | >> train_qwenlatent.py:487 + Step 24670 | grad_norm_pre_clip=0.2742 | + grad_norm_pre_clip_avg=0.2125 | Metrics: + {'align_loss': 0.023501180112361908, + 'recon_loss': 0.09073813259601593, + 'predict_loss': 0.012807819060981274, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2741650342941284, + 'data_time': 0.000942762999329716, + 'model_time': 1.524686037999345, + 'grad_norm_pre_clip_avg': 0.21247761994600295, + 'learning_rate': 1.4963651786951955e-05, + 'epoch': 6.23} +04/19 [20:19:53] INFO | >> train_qwenlatent.py:487 + Step 24680 | grad_norm_pre_clip=0.1741 | + grad_norm_pre_clip_avg=0.1866 | Metrics: + {'align_loss': 0.02427000179886818, + 'recon_loss': 0.11304172873497009, + 'predict_loss': 0.019828077405691147, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17406968772411346, + 'data_time': 0.0006963699997868389, + 'model_time': 1.2332276350061875, + 'grad_norm_pre_clip_avg': 0.18661573976278306, + 'learning_rate': 1.4955102867360326e-05, + 'epoch': 6.23} +04/19 [20:20:06] INFO | >> train_qwenlatent.py:487 + Step 24690 | grad_norm_pre_clip=0.1791 | + grad_norm_pre_clip_avg=0.1770 | Metrics: + {'align_loss': 0.02464039996266365, + 'recon_loss': 0.07837973535060883, + 'predict_loss': 0.007094783242791891, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1791311800479889, + 'data_time': 0.0009241889929398894, + 'model_time': 1.2074091380054597, + 'grad_norm_pre_clip_avg': 0.17696642577648164, + 'learning_rate': 1.4946552757273737e-05, + 'epoch': 6.23} +04/19 [20:20:19] INFO | >> train_qwenlatent.py:487 + Step 24700 | grad_norm_pre_clip=0.1656 | + grad_norm_pre_clip_avg=0.2036 | Metrics: + {'align_loss': 0.02534208633005619, + 'recon_loss': 0.08437809348106384, + 'predict_loss': 0.012182117439806461, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16559605300426483, + 'mae_score': 0.009736574877489794, 'data_time': + 0.0009267409914173186, 'model_time': + 1.250218542991206, 'grad_norm_pre_clip_avg': + 0.20359135717153548, 'learning_rate': + 1.4938001460859403e-05, 'epoch': 6.23} +04/19 [20:20:31] INFO | >> train_qwenlatent.py:487 + Step 24710 | grad_norm_pre_clip=0.1410 | + grad_norm_pre_clip_avg=0.1733 | Metrics: + {'align_loss': 0.02549891546368599, + 'recon_loss': 0.08092851936817169, + 'predict_loss': 0.008870760910212994, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14099997282028198, + 'data_time': 0.0010020749759860337, + 'model_time': 1.2602789029770065, + 'grad_norm_pre_clip_avg': 0.17327521592378617, + 'learning_rate': 1.4929448982285121e-05, + 'epoch': 6.24} +04/19 [20:20:44] INFO | >> train_qwenlatent.py:487 + Step 24720 | grad_norm_pre_clip=0.2134 | + grad_norm_pre_clip_avg=0.1850 | Metrics: + {'align_loss': 0.025709640234708786, + 'recon_loss': 0.07798796147108078, + 'predict_loss': 0.009768323041498661, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21344423294067383, + 'data_time': 0.0006142809870652854, + 'model_time': 1.2377612830023281, + 'grad_norm_pre_clip_avg': 0.1850070685148239, + 'learning_rate': 1.4920895325719262e-05, + 'epoch': 6.24} +04/19 [20:20:57] INFO | >> train_qwenlatent.py:487 + Step 24730 | grad_norm_pre_clip=0.1514 | + grad_norm_pre_clip_avg=0.1638 | Metrics: + {'align_loss': 0.025796525180339813, + 'recon_loss': 0.07226383686065674, + 'predict_loss': 0.009564006701111794, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15136872231960297, + 'data_time': 0.001044736010953784, + 'model_time': 1.24420652500703, + 'grad_norm_pre_clip_avg': 0.16384153589606285, + 'learning_rate': 1.4912340495330784e-05, + 'epoch': 6.24} +04/19 [20:21:09] INFO | >> train_qwenlatent.py:487 + Step 24740 | grad_norm_pre_clip=0.2263 | + grad_norm_pre_clip_avg=0.2090 | Metrics: + {'align_loss': 0.026428289711475372, + 'recon_loss': 0.09232956916093826, + 'predict_loss': 0.010377815924584866, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2263224720954895, + 'data_time': 0.0009434300009161234, + 'model_time': 1.1988224959932268, + 'grad_norm_pre_clip_avg': 0.2090129777789116, + 'learning_rate': 1.49037844952892e-05, 'epoch': + 6.24} +04/19 [20:21:23] INFO | >> train_qwenlatent.py:487 + Step 24750 | grad_norm_pre_clip=0.1842 | + grad_norm_pre_clip_avg=0.2085 | Metrics: + {'align_loss': 0.026544373482465744, + 'recon_loss': 0.08876963704824448, + 'predict_loss': 0.008903277106583118, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18418246507644653, + 'mae_score': 0.008990397754016223, 'data_time': + 0.0009076369751710445, 'model_time': + 1.2377565820061136, 'grad_norm_pre_clip_avg': + 0.20852753520011902, 'learning_rate': + 1.4895227329764605e-05, 'epoch': 6.25} +04/19 [20:21:35] INFO | >> train_qwenlatent.py:487 + Step 24760 | grad_norm_pre_clip=0.1898 | + grad_norm_pre_clip_avg=0.1869 | Metrics: + {'align_loss': 0.024816136807203293, + 'recon_loss': 0.07802806049585342, + 'predict_loss': 0.009955660440027714, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18975882232189178, + 'data_time': 0.0006731559988111258, + 'model_time': 1.2504497330228332, + 'grad_norm_pre_clip_avg': 0.18694517761468887, + 'learning_rate': 1.4886669002927654e-05, + 'epoch': 6.25} +04/19 [20:21:47] INFO | >> train_qwenlatent.py:487 + Step 24770 | grad_norm_pre_clip=0.1557 | + grad_norm_pre_clip_avg=0.1788 | Metrics: + {'align_loss': 0.025676116347312927, + 'recon_loss': 0.08936513215303421, + 'predict_loss': 0.009420506656169891, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15567563474178314, + 'data_time': 0.000650720001431182, + 'model_time': 1.2214477109955624, + 'grad_norm_pre_clip_avg': 0.17878379225730895, + 'learning_rate': 1.487810951894957e-05, + 'epoch': 6.25} +04/19 [20:22:00] INFO | >> train_qwenlatent.py:487 + Step 24780 | grad_norm_pre_clip=0.2077 | + grad_norm_pre_clip_avg=0.1812 | Metrics: + {'align_loss': 0.025666285306215286, + 'recon_loss': 0.07061468064785004, + 'predict_loss': 0.00855859462171793, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20773662626743317, + 'data_time': 0.0006912270036991686, + 'model_time': 1.2450024670106359, + 'grad_norm_pre_clip_avg': 0.1812445655465126, + 'learning_rate': 1.4869548882002149e-05, + 'epoch': 6.25} +04/19 [20:22:13] INFO | >> train_qwenlatent.py:487 + Step 24790 | grad_norm_pre_clip=0.1999 | + grad_norm_pre_clip_avg=0.1836 | Metrics: + {'align_loss': 0.025527236983180046, + 'recon_loss': 0.0907728374004364, + 'predict_loss': 0.008969957940280437, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19990608096122742, + 'data_time': 0.0009891630033962429, + 'model_time': 1.2423880379938055, + 'grad_norm_pre_clip_avg': 0.18360473290085794, + 'learning_rate': 1.4860987096257734e-05, + 'epoch': 6.26} +04/19 [20:22:26] INFO | >> train_qwenlatent.py:487 + Step 24800 | grad_norm_pre_clip=0.1719 | + grad_norm_pre_clip_avg=0.1613 | Metrics: + {'align_loss': 0.02499024197459221, + 'recon_loss': 0.08834903687238693, + 'predict_loss': 0.00995404552668333, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17190708220005035, + 'mae_score': 0.012718673224921699, 'data_time': + 0.0006547460216097534, 'model_time': + 1.2832453399896622, 'grad_norm_pre_clip_avg': + 0.16128148287534713, 'learning_rate': + 1.4852424165889239e-05, 'epoch': 6.26} +04/19 [20:22:39] INFO | >> train_qwenlatent.py:487 + Step 24810 | grad_norm_pre_clip=0.1538 | + grad_norm_pre_clip_avg=0.1799 | Metrics: + {'align_loss': 0.024543441832065582, + 'recon_loss': 0.06145003065466881, + 'predict_loss': 0.0073245675303041935, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.153837189078331, + 'data_time': 0.0006920529995113611, + 'model_time': 1.5592664330033585, + 'grad_norm_pre_clip_avg': 0.1799240753054619, + 'learning_rate': 1.4843860095070128e-05, + 'epoch': 6.26} +04/19 [20:22:51] INFO | >> train_qwenlatent.py:487 + Step 24820 | grad_norm_pre_clip=0.2025 | + grad_norm_pre_clip_avg=0.1898 | Metrics: + {'align_loss': 0.024825911968946457, + 'recon_loss': 0.08771723508834839, + 'predict_loss': 0.013031993992626667, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20250670611858368, + 'data_time': 0.001309762999881059, + 'model_time': 1.210399451025296, + 'grad_norm_pre_clip_avg': 0.18983480632305144, + 'learning_rate': 1.4835294887974429e-05, + 'epoch': 6.26} +04/19 [20:23:04] INFO | >> train_qwenlatent.py:487 + Step 24830 | grad_norm_pre_clip=0.2247 | + grad_norm_pre_clip_avg=0.2309 | Metrics: + {'align_loss': 0.026004856452345848, + 'recon_loss': 0.07022509723901749, + 'predict_loss': 0.010891287587583065, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.224746972322464, + 'data_time': 0.0009863440063782036, + 'model_time': 1.2284000650106464, + 'grad_norm_pre_clip_avg': 0.2309355080127716, + 'learning_rate': 1.4826728548776722e-05, + 'epoch': 6.27} +04/19 [20:23:17] INFO | >> train_qwenlatent.py:487 + Step 24840 | grad_norm_pre_clip=0.2283 | + grad_norm_pre_clip_avg=0.1875 | Metrics: + {'align_loss': 0.02519189566373825, + 'recon_loss': 0.0766713097691536, + 'predict_loss': 0.008803381584584713, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22832423448562622, + 'data_time': 0.0011537079990375787, + 'model_time': 1.2135086310154293, + 'grad_norm_pre_clip_avg': 0.18750565946102143, + 'learning_rate': 1.4818161081652134e-05, + 'epoch': 6.27} +04/19 [20:23:30] INFO | >> train_qwenlatent.py:487 + Step 24850 | grad_norm_pre_clip=0.1501 | + grad_norm_pre_clip_avg=0.1611 | Metrics: + {'align_loss': 0.02608174830675125, + 'recon_loss': 0.08383744955062866, + 'predict_loss': 0.007443014066666365, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15010227262973785, + 'mae_score': 0.0119551555530445, 'data_time': + 0.0008534929947927594, 'model_time': + 1.2523096359800547, 'grad_norm_pre_clip_avg': + 0.16114192306995392, 'learning_rate': + 1.4809592490776344e-05, 'epoch': 6.27} +04/19 [20:23:42] INFO | >> train_qwenlatent.py:487 + Step 24860 | grad_norm_pre_clip=0.1709 | + grad_norm_pre_clip_avg=0.1883 | Metrics: + {'align_loss': 0.024812553077936172, + 'recon_loss': 0.07371880114078522, + 'predict_loss': 0.006689167115837336, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1709282398223877, + 'data_time': 0.000970043009147048, + 'model_time': 1.2738311800058, + 'grad_norm_pre_clip_avg': 0.18828098475933075, + 'learning_rate': 1.4801022780325577e-05, + 'epoch': 6.27} +04/19 [20:23:55] INFO | >> train_qwenlatent.py:487 + Step 24870 | grad_norm_pre_clip=0.2083 | + grad_norm_pre_clip_avg=0.2066 | Metrics: + {'align_loss': 0.025035670027136803, + 'recon_loss': 0.08021774142980576, + 'predict_loss': 0.017499960958957672, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2083449512720108, + 'data_time': 0.0007538959907833487, + 'model_time': 1.494865693995962, + 'grad_norm_pre_clip_avg': 0.20664461851119995, + 'learning_rate': 1.4792451954476614e-05, + 'epoch': 6.28} +04/19 [20:24:08] INFO | >> train_qwenlatent.py:487 + Step 24880 | grad_norm_pre_clip=0.1928 | + grad_norm_pre_clip_avg=0.1903 | Metrics: + {'align_loss': 0.025049753487110138, + 'recon_loss': 0.091419517993927, + 'predict_loss': 0.012626727111637592, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19283902645111084, + 'data_time': 0.0008799589995760471, + 'model_time': 1.2537004410114605, + 'grad_norm_pre_clip_avg': 0.1903190031647682, + 'learning_rate': 1.4783880017406759e-05, + 'epoch': 6.28} +04/19 [20:24:21] INFO | >> train_qwenlatent.py:487 + Step 24890 | grad_norm_pre_clip=0.2539 | + grad_norm_pre_clip_avg=0.1965 | Metrics: + {'align_loss': 0.0255136638879776, + 'recon_loss': 0.0690203383564949, + 'predict_loss': 0.003599112620577216, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.253939688205719, + 'data_time': 0.0008231769897975028, + 'model_time': 1.2167530989972875, + 'grad_norm_pre_clip_avg': 0.19647097140550612, + 'learning_rate': 1.477530697329388e-05, + 'epoch': 6.28} +04/19 [20:24:34] INFO | >> train_qwenlatent.py:487 + Step 24900 | grad_norm_pre_clip=0.1570 | + grad_norm_pre_clip_avg=0.1984 | Metrics: + {'align_loss': 0.02574389986693859, + 'recon_loss': 0.06249087303876877, + 'predict_loss': 0.008887620642781258, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15702499449253082, + 'mae_score': 0.009443817482338294, 'data_time': + 0.0006730360037181526, 'model_time': + 1.2841756409907248, 'grad_norm_pre_clip_avg': + 0.19843889027833939, 'learning_rate': + 1.4766732826316371e-05, 'epoch': 6.28} +04/19 [20:24:46] INFO | >> train_qwenlatent.py:487 + Step 24910 | grad_norm_pre_clip=0.1865 | + grad_norm_pre_clip_avg=0.1843 | Metrics: + {'align_loss': 0.025861166417598724, + 'recon_loss': 0.09390723705291748, + 'predict_loss': 0.008325754664838314, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18654660880565643, + 'data_time': 0.0008561559952795506, + 'model_time': 1.1987872009922285, + 'grad_norm_pre_clip_avg': 0.184296852350235, + 'learning_rate': 1.4758157580653165e-05, + 'epoch': 6.29} +04/19 [20:24:59] INFO | >> train_qwenlatent.py:487 + Step 24920 | grad_norm_pre_clip=0.2511 | + grad_norm_pre_clip_avg=0.1932 | Metrics: + {'align_loss': 0.024008531123399734, + 'recon_loss': 0.1017502024769783, + 'predict_loss': 0.015065784566104412, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.251146525144577, + 'data_time': 0.0009903949976433069, + 'model_time': 1.2419085200235713, + 'grad_norm_pre_clip_avg': 0.19315278381109238, + 'learning_rate': 1.4749581240483736e-05, + 'epoch': 6.29} +04/19 [20:25:12] INFO | >> train_qwenlatent.py:487 + Step 24930 | grad_norm_pre_clip=0.1938 | + grad_norm_pre_clip_avg=0.1966 | Metrics: + {'align_loss': 0.025385724380612373, + 'recon_loss': 0.057296138256788254, + 'predict_loss': 0.007292415946722031, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19384288787841797, + 'data_time': 0.0008388489950448275, + 'model_time': 1.5336998440034222, + 'grad_norm_pre_clip_avg': 0.1965712770819664, + 'learning_rate': 1.4741003809988085e-05, + 'epoch': 6.29} +04/19 [20:25:25] INFO | >> train_qwenlatent.py:487 + Step 24940 | grad_norm_pre_clip=0.1623 | + grad_norm_pre_clip_avg=0.1836 | Metrics: + {'align_loss': 0.02354774996638298, + 'recon_loss': 0.053286340087652206, + 'predict_loss': 0.005287103354930878, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16231316328048706, + 'data_time': 0.0007002330094110221, + 'model_time': 1.2087603209947702, + 'grad_norm_pre_clip_avg': 0.18361623734235763, + 'learning_rate': 1.4732425293346752e-05, + 'epoch': 6.29} +04/19 [20:25:38] INFO | >> train_qwenlatent.py:487 + Step 24950 | grad_norm_pre_clip=0.1848 | + grad_norm_pre_clip_avg=0.1810 | Metrics: + {'align_loss': 0.023840956389904022, + 'recon_loss': 0.06054472550749779, + 'predict_loss': 0.008168613538146019, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18482574820518494, + 'mae_score': 0.009017887630978147, 'data_time': + 0.0010185609862674028, 'model_time': + 1.2051182840077672, 'grad_norm_pre_clip_avg': + 0.18101440966129304, 'learning_rate': + 1.4723845694740798e-05, 'epoch': 6.3} +04/19 [20:25:50] INFO | >> train_qwenlatent.py:487 + Step 24960 | grad_norm_pre_clip=0.2055 | + grad_norm_pre_clip_avg=0.2070 | Metrics: + {'align_loss': 0.025432515889406204, + 'recon_loss': 0.09123822301626205, + 'predict_loss': 0.0065907142125070095, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20545877516269684, + 'data_time': 0.000663946004351601, + 'model_time': 1.2546181340003386, + 'grad_norm_pre_clip_avg': 0.20703566521406175, + 'learning_rate': 1.4715265018351815e-05, + 'epoch': 6.3} +04/19 [20:26:03] INFO | >> train_qwenlatent.py:487 + Step 24970 | grad_norm_pre_clip=0.1374 | + grad_norm_pre_clip_avg=0.1927 | Metrics: + {'align_loss': 0.02452310360968113, + 'recon_loss': 0.09272891283035278, + 'predict_loss': 0.012955441139638424, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1373528391122818, + 'data_time': 0.0008125930035021156, + 'model_time': 1.2248979700088967, + 'grad_norm_pre_clip_avg': 0.19272903203964234, + 'learning_rate': 1.4706683268361923e-05, + 'epoch': 6.3} +04/19 [20:26:16] INFO | >> train_qwenlatent.py:487 + Step 24980 | grad_norm_pre_clip=0.1949 | + grad_norm_pre_clip_avg=0.1945 | Metrics: + {'align_loss': 0.02550993114709854, + 'recon_loss': 0.06337491422891617, + 'predict_loss': 0.014079833403229713, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19491010904312134, + 'data_time': 0.000819053006125614, + 'model_time': 1.1874655869905837, + 'grad_norm_pre_clip_avg': 0.19445638358592987, + 'learning_rate': 1.4698100448953762e-05, + 'epoch': 6.3} +04/19 [20:26:28] INFO | >> train_qwenlatent.py:487 + Step 24990 | grad_norm_pre_clip=0.2235 | + grad_norm_pre_clip_avg=0.1984 | Metrics: + {'align_loss': 0.026652146130800247, + 'recon_loss': 0.09479556977748871, + 'predict_loss': 0.010850513353943825, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22351676225662231, + 'data_time': 0.0008994269883260131, + 'model_time': 1.2628128099895548, + 'grad_norm_pre_clip_avg': 0.19840414002537726, + 'learning_rate': 1.4689516564310491e-05, + 'epoch': 6.31} +04/19 [20:26:41] INFO | >> train_qwenlatent.py:487 + Step 25000 | grad_norm_pre_clip=0.1544 | + grad_norm_pre_clip_avg=0.1710 | Metrics: + {'align_loss': 0.02549276314675808, + 'recon_loss': 0.09548356384038925, + 'predict_loss': 0.01162198930978775, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1544407308101654, + 'mae_score': 0.008414262479489988, 'data_time': + 0.000735353009076789, 'model_time': + 1.215720100008184, 'grad_norm_pre_clip_avg': + 0.1710013598203659, 'learning_rate': + 1.4680931618615795e-05, 'epoch': 6.31} +✅ Checkpoint saved at ./runs/0418_QwenLatent_13tasks_actionstate_30k/checkpoints/steps_25000 +04/19 [20:27:03] INFO | >> train_qwenlatent.py:487 + Step 25010 | grad_norm_pre_clip=0.1633 | + grad_norm_pre_clip_avg=0.1644 | Metrics: + {'align_loss': 0.025322001427412033, + 'recon_loss': 0.09898301213979721, + 'predict_loss': 0.010058185085654259, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16325698792934418, + 'data_time': 0.0007199629908427596, + 'model_time': 1.2404020369867794, + 'grad_norm_pre_clip_avg': 0.16440558582544326, + 'learning_rate': 1.4672345616053871e-05, + 'epoch': 6.31} +04/19 [20:27:16] INFO | >> train_qwenlatent.py:487 + Step 25020 | grad_norm_pre_clip=0.2621 | + grad_norm_pre_clip_avg=0.1915 | Metrics: + {'align_loss': 0.02494146302342415, + 'recon_loss': 0.10121860355138779, + 'predict_loss': 0.010606706142425537, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26207417249679565, + 'data_time': 0.0007618720119353384, + 'model_time': 1.1979495340201538, + 'grad_norm_pre_clip_avg': 0.1915423184633255, + 'learning_rate': 1.4663758560809429e-05, + 'epoch': 6.31} +04/19 [20:27:29] INFO | >> train_qwenlatent.py:487 + Step 25030 | grad_norm_pre_clip=0.2461 | + grad_norm_pre_clip_avg=0.2288 | Metrics: + {'align_loss': 0.025891965255141258, + 'recon_loss': 0.08661817014217377, + 'predict_loss': 0.006791615393012762, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24613508582115173, + 'data_time': 0.0009217069891747087, + 'model_time': 1.244529223011341, + 'grad_norm_pre_clip_avg': 0.22883010804653167, + 'learning_rate': 1.4655170457067695e-05, + 'epoch': 6.32} +04/19 [20:27:41] INFO | >> train_qwenlatent.py:487 + Step 25040 | grad_norm_pre_clip=0.2322 | + grad_norm_pre_clip_avg=0.2044 | Metrics: + {'align_loss': 0.026482298970222473, + 'recon_loss': 0.07931975275278091, + 'predict_loss': 0.009171896614134312, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23218543827533722, + 'data_time': 0.0006474230031017214, + 'model_time': 1.2919140720041469, + 'grad_norm_pre_clip_avg': 0.20437472611665725, + 'learning_rate': 1.464658130901442e-05, + 'epoch': 6.32} +04/19 [20:27:54] INFO | >> train_qwenlatent.py:487 + Step 25050 | grad_norm_pre_clip=0.1743 | + grad_norm_pre_clip_avg=0.1763 | Metrics: + {'align_loss': 0.025163831189274788, + 'recon_loss': 0.0798833891749382, + 'predict_loss': 0.010524114593863487, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17433202266693115, + 'mae_score': 0.009802837199992961, 'data_time': + 0.0009758769883774221, 'model_time': + 1.1874978439882398, 'grad_norm_pre_clip_avg': + 0.1762976437807083, 'learning_rate': + 1.4637991120835831e-05, 'epoch': 6.32} +04/19 [20:28:07] INFO | >> train_qwenlatent.py:487 + Step 25060 | grad_norm_pre_clip=0.2637 | + grad_norm_pre_clip_avg=0.2187 | Metrics: + {'align_loss': 0.02555801533162594, + 'recon_loss': 0.09317224472761154, + 'predict_loss': 0.010827473364770412, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26365914940834045, + 'data_time': 0.0006987170199863613, + 'model_time': 1.572007996001048, + 'grad_norm_pre_clip_avg': 0.21867371946573258, + 'learning_rate': 1.4629399896718695e-05, + 'epoch': 6.32} +04/19 [20:28:20] INFO | >> train_qwenlatent.py:487 + Step 25070 | grad_norm_pre_clip=0.1508 | + grad_norm_pre_clip_avg=0.1832 | Metrics: + {'align_loss': 0.02476075477898121, + 'recon_loss': 0.08014808595180511, + 'predict_loss': 0.010075903497636318, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1507968306541443, + 'data_time': 0.0006467710190918297, + 'model_time': 1.214610070019262, + 'grad_norm_pre_clip_avg': 0.1832283541560173, + 'learning_rate': 1.4620807640850264e-05, + 'epoch': 6.33} +04/19 [20:28:33] INFO | >> train_qwenlatent.py:487 + Step 25080 | grad_norm_pre_clip=0.1726 | + grad_norm_pre_clip_avg=0.1875 | Metrics: + {'align_loss': 0.025336749851703644, + 'recon_loss': 0.07181898504495621, + 'predict_loss': 0.009607745334506035, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1725762039422989, + 'data_time': 0.0008174169925041497, + 'model_time': 1.2279049879871309, + 'grad_norm_pre_clip_avg': 0.1874705284833908, + 'learning_rate': 1.4612214357418302e-05, + 'epoch': 6.33} +04/19 [20:28:45] INFO | >> train_qwenlatent.py:487 + Step 25090 | grad_norm_pre_clip=0.2430 | + grad_norm_pre_clip_avg=0.2123 | Metrics: + {'align_loss': 0.025693919509649277, + 'recon_loss': 0.06622426211833954, + 'predict_loss': 0.008738993667066097, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24299722909927368, + 'data_time': 0.0007561229867860675, + 'model_time': 1.20234754102421, + 'grad_norm_pre_clip_avg': 0.2123470649123192, + 'learning_rate': 1.4603620050611069e-05, + 'epoch': 6.33} +04/19 [20:28:58] INFO | >> train_qwenlatent.py:487 + Step 25100 | grad_norm_pre_clip=0.1556 | + grad_norm_pre_clip_avg=0.1798 | Metrics: + {'align_loss': 0.027150925248861313, + 'recon_loss': 0.10529863089323044, + 'predict_loss': 0.01672792248427868, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1556180864572525, + 'mae_score': 0.009088066891506985, 'data_time': + 0.0006700590020045638, 'model_time': + 1.199016336002387, 'grad_norm_pre_clip_avg': + 0.1797675684094429, 'learning_rate': + 1.4595024724617329e-05, 'epoch': 6.33} +04/19 [20:29:11] INFO | >> train_qwenlatent.py:487 + Step 25110 | grad_norm_pre_clip=0.2022 | + grad_norm_pre_clip_avg=0.1795 | Metrics: + {'align_loss': 0.02471911534667015, + 'recon_loss': 0.07437027245759964, + 'predict_loss': 0.009220409207046032, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20219939947128296, + 'data_time': 0.000995709007838741, + 'model_time': 1.2323648840247188, + 'grad_norm_pre_clip_avg': 0.17951322197914124, + 'learning_rate': 1.4586428383626336e-05, + 'epoch': 6.34} +04/19 [20:29:24] INFO | >> train_qwenlatent.py:487 + Step 25120 | grad_norm_pre_clip=0.1588 | + grad_norm_pre_clip_avg=0.1719 | Metrics: + {'align_loss': 0.02488921582698822, + 'recon_loss': 0.07116876542568207, + 'predict_loss': 0.011103483848273754, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15876734256744385, + 'data_time': 0.000989251973805949, + 'model_time': 1.2586604190000799, + 'grad_norm_pre_clip_avg': 0.17185764163732528, + 'learning_rate': 1.457783103182784e-05, + 'epoch': 6.34} +04/19 [20:29:36] INFO | >> train_qwenlatent.py:487 + Step 25130 | grad_norm_pre_clip=0.2448 | + grad_norm_pre_clip_avg=0.2016 | Metrics: + {'align_loss': 0.025418026372790337, + 'recon_loss': 0.10764724016189575, + 'predict_loss': 0.011863362044095993, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2447606474161148, + 'data_time': 0.0007247600005939603, + 'model_time': 1.2442394359968603, + 'grad_norm_pre_clip_avg': 0.201553313434124, + 'learning_rate': 1.456923267341209e-05, + 'epoch': 6.34} +04/19 [20:29:48] INFO | >> train_qwenlatent.py:487 + Step 25140 | grad_norm_pre_clip=0.1855 | + grad_norm_pre_clip_avg=0.1773 | Metrics: + {'align_loss': 0.02489374950528145, + 'recon_loss': 0.10375725477933884, + 'predict_loss': 0.01465686783194542, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18548952043056488, + 'data_time': 0.0006591120036318898, + 'model_time': 1.2348371489788406, + 'grad_norm_pre_clip_avg': 0.17727723568677903, + 'learning_rate': 1.4560633312569824e-05, + 'epoch': 6.34} +04/19 [20:30:02] INFO | >> train_qwenlatent.py:487 + Step 25150 | grad_norm_pre_clip=0.2159 | + grad_norm_pre_clip_avg=0.1914 | Metrics: + {'align_loss': 0.025166422128677368, + 'recon_loss': 0.10048072040081024, + 'predict_loss': 0.011871347203850746, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21592700481414795, + 'mae_score': 0.010687491270872923, 'data_time': + 0.0013897030148655176, 'model_time': + 1.214696165989153, 'grad_norm_pre_clip_avg': + 0.19139775931835173, 'learning_rate': + 1.455203295349226e-05, 'epoch': 6.35} +04/19 [20:30:15] INFO | >> train_qwenlatent.py:487 + Step 25160 | grad_norm_pre_clip=0.1856 | + grad_norm_pre_clip_avg=0.1902 | Metrics: + {'align_loss': 0.02546628750860691, + 'recon_loss': 0.07778006047010422, + 'predict_loss': 0.00822618417441845, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1856483817100525, + 'data_time': 0.0007270249770954251, + 'model_time': 1.2537110740086064, + 'grad_norm_pre_clip_avg': 0.19021159112453462, + 'learning_rate': 1.454343160037111e-05, + 'epoch': 6.35} +04/19 [20:30:27] INFO | >> train_qwenlatent.py:487 + Step 25170 | grad_norm_pre_clip=0.1465 | + grad_norm_pre_clip_avg=0.1670 | Metrics: + {'align_loss': 0.02539190649986267, + 'recon_loss': 0.061767593026161194, + 'predict_loss': 0.008640985004603863, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14645780622959137, + 'data_time': 0.0009701679809950292, + 'model_time': 1.2381613090110477, + 'grad_norm_pre_clip_avg': 0.1670437842607498, + 'learning_rate': 1.4534829257398569e-05, + 'epoch': 6.35} +04/19 [20:30:40] INFO | >> train_qwenlatent.py:487 + Step 25180 | grad_norm_pre_clip=0.1458 | + grad_norm_pre_clip_avg=0.2057 | Metrics: + {'align_loss': 0.026063743978738785, + 'recon_loss': 0.09008391946554184, + 'predict_loss': 0.012158513069152832, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14578181505203247, + 'data_time': 0.0006563780189026147, + 'model_time': 1.2777696989942342, + 'grad_norm_pre_clip_avg': 0.20572341084480286, + 'learning_rate': 1.4526225928767324e-05, + 'epoch': 6.35} +04/19 [20:30:53] INFO | >> train_qwenlatent.py:487 + Step 25190 | grad_norm_pre_clip=0.2132 | + grad_norm_pre_clip_avg=0.1813 | Metrics: + {'align_loss': 0.024449247866868973, + 'recon_loss': 0.07041297852993011, + 'predict_loss': 0.01147847343236208, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21316222846508026, + 'data_time': 0.001062064984580502, + 'model_time': 1.206210470001679, + 'grad_norm_pre_clip_avg': 0.18129378855228423, + 'learning_rate': 1.4517621618670512e-05, + 'epoch': 6.36} +04/19 [20:31:06] INFO | >> train_qwenlatent.py:487 + Step 25200 | grad_norm_pre_clip=0.1829 | + grad_norm_pre_clip_avg=0.2237 | Metrics: + {'align_loss': 0.025583740323781967, + 'recon_loss': 0.07422535121440887, + 'predict_loss': 0.006796183064579964, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1828732192516327, + 'mae_score': 0.0117792576282948, 'data_time': + 0.0006894099933560938, 'model_time': + 1.475062124984106, 'grad_norm_pre_clip_avg': + 0.22374931275844573, 'learning_rate': + 1.4509016331301789e-05, 'epoch': 6.36} +04/19 [20:31:19] INFO | >> train_qwenlatent.py:487 + Step 25210 | grad_norm_pre_clip=0.1838 | + grad_norm_pre_clip_avg=0.1868 | Metrics: + {'align_loss': 0.025920821353793144, + 'recon_loss': 0.09597545862197876, + 'predict_loss': 0.016595518216490746, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1838025599718094, + 'data_time': 0.0006516349967569113, + 'model_time': 1.2131992049980909, + 'grad_norm_pre_clip_avg': 0.18676260709762574, + 'learning_rate': 1.4500410070855258e-05, + 'epoch': 6.36} +04/19 [20:31:32] INFO | >> train_qwenlatent.py:487 + Step 25220 | grad_norm_pre_clip=0.1547 | + grad_norm_pre_clip_avg=0.1991 | Metrics: + {'align_loss': 0.025244761258363724, + 'recon_loss': 0.09234654903411865, + 'predict_loss': 0.009557649493217468, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15469245612621307, + 'data_time': 0.0006683239771518856, + 'model_time': 1.2796277030138299, + 'grad_norm_pre_clip_avg': 0.1990742564201355, + 'learning_rate': 1.4491802841525511e-05, + 'epoch': 6.36} +04/19 [20:31:44] INFO | >> train_qwenlatent.py:487 + Step 25230 | grad_norm_pre_clip=0.1590 | + grad_norm_pre_clip_avg=0.1730 | Metrics: + {'align_loss': 0.024941980838775635, + 'recon_loss': 0.10948941856622696, + 'predict_loss': 0.010717672295868397, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1589764505624771, + 'data_time': 0.000660652993246913, + 'model_time': 1.3110993800219148, + 'grad_norm_pre_clip_avg': 0.17304374277591705, + 'learning_rate': 1.4483194647507603e-05, + 'epoch': 6.37} +04/19 [20:31:57] INFO | >> train_qwenlatent.py:487 + Step 25240 | grad_norm_pre_clip=0.1641 | + grad_norm_pre_clip_avg=0.1772 | Metrics: + {'align_loss': 0.02470756508409977, + 'recon_loss': 0.08912686258554459, + 'predict_loss': 0.008341781795024872, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16409416496753693, + 'data_time': 0.0007676990062464029, + 'model_time': 1.2786768160003703, + 'grad_norm_pre_clip_avg': 0.17718429118394852, + 'learning_rate': 1.447458549299706e-05, + 'epoch': 6.37} +04/19 [20:32:10] INFO | >> train_qwenlatent.py:487 + Step 25250 | grad_norm_pre_clip=0.1830 | + grad_norm_pre_clip_avg=0.2195 | Metrics: + {'align_loss': 0.02531929686665535, + 'recon_loss': 0.08718909323215485, + 'predict_loss': 0.009860113263130188, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1829676628112793, + 'mae_score': 0.009606588208997572, 'data_time': + 0.0009104720083996654, 'model_time': + 1.3100164589995984, 'grad_norm_pre_clip_avg': + 0.2194850817322731, 'learning_rate': + 1.4465975382189884e-05, 'epoch': 6.37} +04/19 [20:32:22] INFO | >> train_qwenlatent.py:487 + Step 25260 | grad_norm_pre_clip=0.2046 | + grad_norm_pre_clip_avg=0.1847 | Metrics: + {'align_loss': 0.02370608225464821, + 'recon_loss': 0.06746210157871246, + 'predict_loss': 0.008679697290062904, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20456504821777344, + 'data_time': 0.0007074539898894727, + 'model_time': 1.249528026994085, + 'grad_norm_pre_clip_avg': 0.1846706673502922, + 'learning_rate': 1.4457364319282537e-05, + 'epoch': 6.37} +04/19 [20:32:35] INFO | >> train_qwenlatent.py:487 + Step 25270 | grad_norm_pre_clip=0.1660 | + grad_norm_pre_clip_avg=0.1996 | Metrics: + {'align_loss': 0.025731030851602554, + 'recon_loss': 0.10056562721729279, + 'predict_loss': 0.009858367964625359, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16604620218276978, + 'data_time': 0.0008452670008409768, + 'model_time': 1.2055327059933916, + 'grad_norm_pre_clip_avg': 0.19955456256866455, + 'learning_rate': 1.4448752308471952e-05, + 'epoch': 6.38} +04/19 [20:32:47] INFO | >> train_qwenlatent.py:487 + Step 25280 | grad_norm_pre_clip=0.1997 | + grad_norm_pre_clip_avg=0.1851 | Metrics: + {'align_loss': 0.025471104308962822, + 'recon_loss': 0.086818166077137, + 'predict_loss': 0.010086762718856335, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.199736088514328, + 'data_time': 0.0006999410106800497, + 'model_time': 1.2360975570045412, + 'grad_norm_pre_clip_avg': 0.18508531153202057, + 'learning_rate': 1.4440139353955515e-05, + 'epoch': 6.38} +04/19 [20:33:00] INFO | >> train_qwenlatent.py:487 + Step 25290 | grad_norm_pre_clip=0.1482 | + grad_norm_pre_clip_avg=0.1690 | Metrics: + {'align_loss': 0.02523096464574337, + 'recon_loss': 0.07861346006393433, + 'predict_loss': 0.011551898904144764, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14820507168769836, + 'data_time': 0.0007082789961714298, + 'model_time': 1.2848009429872036, + 'grad_norm_pre_clip_avg': 0.16898873299360276, + 'learning_rate': 1.4431525459931066e-05, + 'epoch': 6.38} +04/19 [20:33:13] INFO | >> train_qwenlatent.py:487 + Step 25300 | grad_norm_pre_clip=0.1983 | + grad_norm_pre_clip_avg=0.2102 | Metrics: + {'align_loss': 0.026116743683815002, + 'recon_loss': 0.10991232842206955, + 'predict_loss': 0.013319375924766064, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19825242459774017, + 'mae_score': 0.008887196446324254, 'data_time': + 0.0018147819791920483, 'model_time': + 1.2475010519847274, 'grad_norm_pre_clip_avg': + 0.2101531907916069, 'learning_rate': + 1.4422910630596928e-05, 'epoch': 6.38} +04/19 [20:33:26] INFO | >> train_qwenlatent.py:487 + Step 25310 | grad_norm_pre_clip=0.1834 | + grad_norm_pre_clip_avg=0.1850 | Metrics: + {'align_loss': 0.026625962927937508, + 'recon_loss': 0.08143826574087143, + 'predict_loss': 0.007413700222969055, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18342848122119904, + 'data_time': 0.0009721049864310771, + 'model_time': 1.272937383997487, + 'grad_norm_pre_clip_avg': 0.18500740081071854, + 'learning_rate': 1.4414294870151851e-05, + 'epoch': 6.39} +04/19 [20:33:38] INFO | >> train_qwenlatent.py:487 + Step 25320 | grad_norm_pre_clip=0.1416 | + grad_norm_pre_clip_avg=0.1637 | Metrics: + {'align_loss': 0.025854267179965973, + 'recon_loss': 0.10201780498027802, + 'predict_loss': 0.009307736530900002, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14159324765205383, + 'data_time': 0.0006996020092628896, + 'model_time': 1.252908567985287, + 'grad_norm_pre_clip_avg': 0.1637391597032547, + 'learning_rate': 1.4405678182795063e-05, + 'epoch': 6.39} +04/19 [20:33:51] INFO | >> train_qwenlatent.py:487 + Step 25330 | grad_norm_pre_clip=0.3081 | + grad_norm_pre_clip_avg=0.1936 | Metrics: + {'align_loss': 0.025570983067154884, + 'recon_loss': 0.1021789163351059, + 'predict_loss': 0.006471201777458191, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3081189692020416, + 'data_time': 0.000817516993265599, + 'model_time': 1.253284861013526, + 'grad_norm_pre_clip_avg': 0.19363571107387542, + 'learning_rate': 1.4397060572726223e-05, + 'epoch': 6.39} +04/19 [20:34:04] INFO | >> train_qwenlatent.py:487 + Step 25340 | grad_norm_pre_clip=0.2259 | + grad_norm_pre_clip_avg=0.2313 | Metrics: + {'align_loss': 0.025586936622858047, + 'recon_loss': 0.07997126132249832, + 'predict_loss': 0.010809158906340599, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22587259113788605, + 'data_time': 0.0007139219960663468, + 'model_time': 1.2292521159979515, + 'grad_norm_pre_clip_avg': 0.23132454007863998, + 'learning_rate': 1.4388442044145456e-05, + 'epoch': 6.39} +04/19 [20:34:17] INFO | >> train_qwenlatent.py:487 + Step 25350 | grad_norm_pre_clip=0.1479 | + grad_norm_pre_clip_avg=0.1741 | Metrics: + {'align_loss': 0.023960337042808533, + 'recon_loss': 0.061945557594299316, + 'predict_loss': 0.010005602613091469, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14788509905338287, + 'mae_score': 0.00838019053141276, 'data_time': + 0.0006641600048169494, 'model_time': + 1.2555602159991395, 'grad_norm_pre_clip_avg': + 0.1741003304719925, 'learning_rate': + 1.4379822601253326e-05, 'epoch': 6.4} +04/19 [20:34:30] INFO | >> train_qwenlatent.py:487 + Step 25360 | grad_norm_pre_clip=0.1930 | + grad_norm_pre_clip_avg=0.1629 | Metrics: + {'align_loss': 0.02562887966632843, + 'recon_loss': 0.08237665146589279, + 'predict_loss': 0.009960215538740158, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1930006593465805, + 'data_time': 0.0007152179896365851, + 'model_time': 1.2753432250174228, + 'grad_norm_pre_clip_avg': 0.16293609589338304, + 'learning_rate': 1.4371202248250843e-05, + 'epoch': 6.4} +04/19 [20:34:42] INFO | >> train_qwenlatent.py:487 + Step 25370 | grad_norm_pre_clip=0.1753 | + grad_norm_pre_clip_avg=0.1946 | Metrics: + {'align_loss': 0.02600785158574581, + 'recon_loss': 0.1043998971581459, + 'predict_loss': 0.013033472001552582, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17527393996715546, + 'data_time': 0.0006913180113770068, + 'model_time': 1.5361152669938747, + 'grad_norm_pre_clip_avg': 0.19461210519075395, + 'learning_rate': 1.4362580989339464e-05, + 'epoch': 6.4} +04/19 [20:34:55] INFO | >> train_qwenlatent.py:487 + Step 25380 | grad_norm_pre_clip=0.1725 | + grad_norm_pre_clip_avg=0.2060 | Metrics: + {'align_loss': 0.026399921625852585, + 'recon_loss': 0.07499146461486816, + 'predict_loss': 0.008173246867954731, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17254328727722168, + 'data_time': 0.0007086140103638172, + 'model_time': 1.2871105390076991, + 'grad_norm_pre_clip_avg': 0.20597251653671264, + 'learning_rate': 1.4353958828721086e-05, + 'epoch': 6.4} +04/19 [20:35:07] INFO | >> train_qwenlatent.py:487 + Step 25390 | grad_norm_pre_clip=0.2274 | + grad_norm_pre_clip_avg=0.2292 | Metrics: + {'align_loss': 0.025270842015743256, + 'recon_loss': 0.05682015046477318, + 'predict_loss': 0.00788852944970131, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22740812599658966, + 'data_time': 0.0006932670075912029, + 'model_time': 1.2331268760026433, + 'grad_norm_pre_clip_avg': 0.22921500653028487, + 'learning_rate': 1.4345335770598045e-05, + 'epoch': 6.41} +04/19 [20:35:20] INFO | >> train_qwenlatent.py:487 + Step 25400 | grad_norm_pre_clip=0.1708 | + grad_norm_pre_clip_avg=0.1807 | Metrics: + {'align_loss': 0.026106007397174835, + 'recon_loss': 0.12398797273635864, + 'predict_loss': 0.014418178237974644, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17080165445804596, + 'mae_score': 0.011753726649928737, 'data_time': + 0.0006153900176286697, 'model_time': + 1.1697992930130567, 'grad_norm_pre_clip_avg': + 0.18067007064819335, 'learning_rate': + 1.4336711819173118e-05, 'epoch': 6.41} +04/19 [20:35:33] INFO | >> train_qwenlatent.py:487 + Step 25410 | grad_norm_pre_clip=0.1770 | + grad_norm_pre_clip_avg=0.1856 | Metrics: + {'align_loss': 0.025808189064264297, + 'recon_loss': 0.08483904600143433, + 'predict_loss': 0.009697336703538895, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17704293131828308, + 'data_time': 0.001496702985605225, + 'model_time': 1.2447037170059048, + 'grad_norm_pre_clip_avg': 0.18562559485435487, + 'learning_rate': 1.4328086978649508e-05, + 'epoch': 6.41} +04/19 [20:35:46] INFO | >> train_qwenlatent.py:487 + Step 25420 | grad_norm_pre_clip=0.1546 | + grad_norm_pre_clip_avg=0.1669 | Metrics: + {'align_loss': 0.025701237842440605, + 'recon_loss': 0.10637969523668289, + 'predict_loss': 0.008220233023166656, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1546369194984436, + 'data_time': 0.000826892995974049, + 'model_time': 1.2278982659918256, + 'grad_norm_pre_clip_avg': 0.1669188067317009, + 'learning_rate': 1.431946125323086e-05, + 'epoch': 6.41} +04/19 [20:35:59] INFO | >> train_qwenlatent.py:487 + Step 25430 | grad_norm_pre_clip=0.1987 | + grad_norm_pre_clip_avg=0.1726 | Metrics: + {'align_loss': 0.025643737986683846, + 'recon_loss': 0.0874379500746727, + 'predict_loss': 0.015162325464189053, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19873063266277313, + 'data_time': 0.0012697880156338215, + 'model_time': 1.2393774820084218, + 'grad_norm_pre_clip_avg': 0.172583170235157, + 'learning_rate': 1.4310834647121249e-05, + 'epoch': 6.42} +04/19 [20:36:11] INFO | >> train_qwenlatent.py:487 + Step 25440 | grad_norm_pre_clip=0.2423 | + grad_norm_pre_clip_avg=0.1912 | Metrics: + {'align_loss': 0.02610018104314804, + 'recon_loss': 0.10297095775604248, + 'predict_loss': 0.011572575196623802, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24232488870620728, + 'data_time': 0.0009899389988277107, + 'model_time': 1.2469763650151435, + 'grad_norm_pre_clip_avg': 0.19117904752492904, + 'learning_rate': 1.430220716452518e-05, + 'epoch': 6.42} +04/19 [20:36:24] INFO | >> train_qwenlatent.py:487 + Step 25450 | grad_norm_pre_clip=0.2559 | + grad_norm_pre_clip_avg=0.2237 | Metrics: + {'align_loss': 0.02435190975666046, + 'recon_loss': 0.07056776434183121, + 'predict_loss': 0.01103969942778349, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2558620274066925, + 'mae_score': 0.009640392097266944, 'data_time': + 0.0007065419922582805, 'model_time': + 1.2702976599975955, 'grad_norm_pre_clip_avg': + 0.2236701801419258, 'learning_rate': + 1.4293578809647579e-05, 'epoch': 6.42} +04/19 [20:36:37] INFO | >> train_qwenlatent.py:487 + Step 25460 | grad_norm_pre_clip=0.1681 | + grad_norm_pre_clip_avg=0.2038 | Metrics: + {'align_loss': 0.025194011628627777, + 'recon_loss': 0.07377734780311584, + 'predict_loss': 0.007224204950034618, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16812843084335327, + 'data_time': 0.0009487099887337536, + 'model_time': 1.228426418005256, + 'grad_norm_pre_clip_avg': 0.20378136336803437, + 'learning_rate': 1.4284949586693802e-05, + 'epoch': 6.42} +04/19 [20:36:49] INFO | >> train_qwenlatent.py:487 + Step 25470 | grad_norm_pre_clip=0.1408 | + grad_norm_pre_clip_avg=0.1773 | Metrics: + {'align_loss': 0.02479066327214241, + 'recon_loss': 0.07881586998701096, + 'predict_loss': 0.008928827941417694, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1408400982618332, + 'data_time': 0.0008390979783143848, + 'model_time': 1.2132424220035318, + 'grad_norm_pre_clip_avg': 0.1772943064570427, + 'learning_rate': 1.427631949986963e-05, + 'epoch': 6.43} +04/19 [20:37:02] INFO | >> train_qwenlatent.py:487 + Step 25480 | grad_norm_pre_clip=0.1995 | + grad_norm_pre_clip_avg=0.1800 | Metrics: + {'align_loss': 0.025383198633790016, + 'recon_loss': 0.10255212336778641, + 'predict_loss': 0.013111581094563007, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19949114322662354, + 'data_time': 0.000703178026014939, + 'model_time': 1.1983415959984995, + 'grad_norm_pre_clip_avg': 0.1800095409154892, + 'learning_rate': 1.4267688553381256e-05, + 'epoch': 6.43} +04/19 [20:37:15] INFO | >> train_qwenlatent.py:487 + Step 25490 | grad_norm_pre_clip=0.1651 | + grad_norm_pre_clip_avg=0.1834 | Metrics: + {'align_loss': 0.02518581971526146, + 'recon_loss': 0.0764167308807373, + 'predict_loss': 0.0064850240014493465, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16514873504638672, + 'data_time': 0.0009604300139471889, + 'model_time': 1.2713344960066024, + 'grad_norm_pre_clip_avg': 0.18335356563329697, + 'learning_rate': 1.4259056751435313e-05, + 'epoch': 6.43} +04/19 [20:37:28] INFO | >> train_qwenlatent.py:487 + Step 25500 | grad_norm_pre_clip=0.1905 | + grad_norm_pre_clip_avg=0.2118 | Metrics: + {'align_loss': 0.025036688894033432, + 'recon_loss': 0.053096089512109756, + 'predict_loss': 0.004402625374495983, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19047947227954865, + 'mae_score': 0.010755113653234533, 'data_time': + 0.0011868849978782237, 'model_time': + 1.2565616040083114, 'grad_norm_pre_clip_avg': + 0.21178772300481796, 'learning_rate': + 1.4250424098238817e-05, 'epoch': 6.43} +04/19 [20:37:40] INFO | >> train_qwenlatent.py:487 + Step 25510 | grad_norm_pre_clip=0.1719 | + grad_norm_pre_clip_avg=0.1792 | Metrics: + {'align_loss': 0.02590196579694748, + 'recon_loss': 0.07388661801815033, + 'predict_loss': 0.008089380338788033, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17193780839443207, + 'data_time': 0.0007005539955571294, + 'model_time': 1.2090492419956718, + 'grad_norm_pre_clip_avg': 0.1792267680168152, + 'learning_rate': 1.4241790597999236e-05, + 'epoch': 6.44} +04/19 [20:37:53] INFO | >> train_qwenlatent.py:487 + Step 25520 | grad_norm_pre_clip=0.1784 | + grad_norm_pre_clip_avg=0.1731 | Metrics: + {'align_loss': 0.026363279670476913, + 'recon_loss': 0.10015473514795303, + 'predict_loss': 0.010349842719733715, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17837487161159515, + 'data_time': 0.0006930950039532036, + 'model_time': 1.2269554540107492, + 'grad_norm_pre_clip_avg': 0.1730727434158325, + 'learning_rate': 1.4233156254924417e-05, + 'epoch': 6.44} +04/19 [20:38:05] INFO | >> train_qwenlatent.py:487 + Step 25530 | grad_norm_pre_clip=0.2580 | + grad_norm_pre_clip_avg=0.2124 | Metrics: + {'align_loss': 0.02473614737391472, + 'recon_loss': 0.07143029570579529, + 'predict_loss': 0.007690112106502056, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2580403685569763, + 'data_time': 0.0009779050014913082, + 'model_time': 1.214731115003815, + 'grad_norm_pre_clip_avg': 0.21241099685430526, + 'learning_rate': 1.4224521073222652e-05, + 'epoch': 6.44} +04/19 [20:38:18] INFO | >> train_qwenlatent.py:487 + Step 25540 | grad_norm_pre_clip=0.1172 | + grad_norm_pre_clip_avg=0.1804 | Metrics: + {'align_loss': 0.025891633704304695, + 'recon_loss': 0.09612864255905151, + 'predict_loss': 0.007361799478530884, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11723023653030396, + 'data_time': 0.0009302749822381884, + 'model_time': 1.2381475579750258, + 'grad_norm_pre_clip_avg': 0.18036342933773994, + 'learning_rate': 1.4215885057102604e-05, + 'epoch': 6.44} +04/19 [20:38:31] INFO | >> train_qwenlatent.py:487 + Step 25550 | grad_norm_pre_clip=0.2307 | + grad_norm_pre_clip_avg=0.1979 | Metrics: + {'align_loss': 0.026384850963950157, + 'recon_loss': 0.11632729321718216, + 'predict_loss': 0.008438766933977604, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2306879162788391, + 'mae_score': 0.0105720245086395, 'data_time': + 0.0006614949961658567, 'model_time': + 1.242279733007308, 'grad_norm_pre_clip_avg': + 0.19793423414230346, 'learning_rate': + 1.4207248210773378e-05, 'epoch': 6.45} +04/19 [20:38:43] INFO | >> train_qwenlatent.py:487 + Step 25560 | grad_norm_pre_clip=0.1687 | + grad_norm_pre_clip_avg=0.1813 | Metrics: + {'align_loss': 0.02590484917163849, + 'recon_loss': 0.07515634596347809, + 'predict_loss': 0.008617491461336613, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16873283684253693, + 'data_time': 0.0009057060233317316, + 'model_time': 1.2540051100077108, + 'grad_norm_pre_clip_avg': 0.1813131555914879, + 'learning_rate': 1.4198610538444459e-05, + 'epoch': 6.45} +04/19 [20:38:56] INFO | >> train_qwenlatent.py:487 + Step 25570 | grad_norm_pre_clip=0.1864 | + grad_norm_pre_clip_avg=0.1839 | Metrics: + {'align_loss': 0.0241668950766325, + 'recon_loss': 0.0681585893034935, + 'predict_loss': 0.010712447576224804, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18640145659446716, + 'data_time': 0.0007231760246213526, + 'model_time': 1.3131325489957817, + 'grad_norm_pre_clip_avg': 0.18390000164508818, + 'learning_rate': 1.4189972044325747e-05, + 'epoch': 6.45} +04/19 [20:39:09] INFO | >> train_qwenlatent.py:487 + Step 25580 | grad_norm_pre_clip=0.2029 | + grad_norm_pre_clip_avg=0.1939 | Metrics: + {'align_loss': 0.025537075474858284, + 'recon_loss': 0.08937754482030869, + 'predict_loss': 0.009640185162425041, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20293650031089783, + 'data_time': 0.0010420950129628181, + 'model_time': 1.2673961779801175, + 'grad_norm_pre_clip_avg': 0.1938594326376915, + 'learning_rate': 1.4181332732627544e-05, + 'epoch': 6.45} +04/19 [20:39:21] INFO | >> train_qwenlatent.py:487 + Step 25590 | grad_norm_pre_clip=0.1829 | + grad_norm_pre_clip_avg=0.1844 | Metrics: + {'align_loss': 0.026167962700128555, + 'recon_loss': 0.0833001360297203, + 'predict_loss': 0.006550714373588562, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1828833520412445, + 'data_time': 0.0009259749785996974, + 'model_time': 1.2140329480171204, + 'grad_norm_pre_clip_avg': 0.1843613639473915, + 'learning_rate': 1.4172692607560533e-05, + 'epoch': 6.46} +04/19 [20:39:35] INFO | >> train_qwenlatent.py:487 + Step 25600 | grad_norm_pre_clip=0.1427 | + grad_norm_pre_clip_avg=0.1551 | Metrics: + {'align_loss': 0.02580106258392334, + 'recon_loss': 0.09382037818431854, + 'predict_loss': 0.010334156453609467, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1427462249994278, + 'mae_score': 0.009117153958157376, 'data_time': + 0.0006731129833497107, 'model_time': + 1.21332013598294, 'grad_norm_pre_clip_avg': + 0.15510396361351014, 'learning_rate': + 1.4164051673335819e-05, 'epoch': 6.46} +04/19 [20:39:48] INFO | >> train_qwenlatent.py:487 + Step 25610 | grad_norm_pre_clip=0.2593 | + grad_norm_pre_clip_avg=0.2054 | Metrics: + {'align_loss': 0.026538370177149773, + 'recon_loss': 0.10002273321151733, + 'predict_loss': 0.011139553971588612, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25927749276161194, + 'data_time': 0.000990700995316729, + 'model_time': 1.302169109985698, + 'grad_norm_pre_clip_avg': 0.2053988665342331, + 'learning_rate': 1.4155409934164878e-05, + 'epoch': 6.46} +04/19 [20:40:01] INFO | >> train_qwenlatent.py:487 + Step 25620 | grad_norm_pre_clip=0.1441 | + grad_norm_pre_clip_avg=0.1710 | Metrics: + {'align_loss': 0.025835193693637848, + 'recon_loss': 0.08481117337942123, + 'predict_loss': 0.010384646244347095, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14407838881015778, + 'data_time': 0.0006470319931395352, + 'model_time': 1.1978029900055844, + 'grad_norm_pre_clip_avg': 0.17096418067812919, + 'learning_rate': 1.4146767394259596e-05, + 'epoch': 6.46} +04/19 [20:40:13] INFO | >> train_qwenlatent.py:487 + Step 25630 | grad_norm_pre_clip=0.1859 | + grad_norm_pre_clip_avg=0.1972 | Metrics: + {'align_loss': 0.024637293070554733, + 'recon_loss': 0.07013868540525436, + 'predict_loss': 0.007821837440133095, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18593266606330872, + 'data_time': 0.000986216007731855, + 'model_time': 1.4569285940087866, + 'grad_norm_pre_clip_avg': 0.19715436398983002, + 'learning_rate': 1.4138124057832242e-05, + 'epoch': 6.47} +04/19 [20:40:25] INFO | >> train_qwenlatent.py:487 + Step 25640 | grad_norm_pre_clip=0.2017 | + grad_norm_pre_clip_avg=0.2380 | Metrics: + {'align_loss': 0.025873318314552307, + 'recon_loss': 0.07601669430732727, + 'predict_loss': 0.006451489869505167, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2017143815755844, + 'data_time': 0.0009743139962665737, + 'model_time': 1.2222403770138044, + 'grad_norm_pre_clip_avg': 0.23804289549589158, + 'learning_rate': 1.4129479929095463e-05, + 'epoch': 6.47} +04/19 [20:40:38] INFO | >> train_qwenlatent.py:487 + Step 25650 | grad_norm_pre_clip=0.2218 | + grad_norm_pre_clip_avg=0.2135 | Metrics: + {'align_loss': 0.024991856887936592, + 'recon_loss': 0.10237280279397964, + 'predict_loss': 0.012946250848472118, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22175906598567963, + 'mae_score': 0.01268466227763408, 'data_time': + 0.0007153679907787591, 'model_time': + 1.2230523900070693, 'grad_norm_pre_clip_avg': + 0.21348830908536912, 'learning_rate': + 1.4120835012262318e-05, 'epoch': 6.47} +04/19 [20:40:51] INFO | >> train_qwenlatent.py:487 + Step 25660 | grad_norm_pre_clip=0.1573 | + grad_norm_pre_clip_avg=0.1788 | Metrics: + {'align_loss': 0.025544244796037674, + 'recon_loss': 0.0998574048280716, + 'predict_loss': 0.01247339230030775, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15734197199344635, + 'data_time': 0.0011402420059312135, + 'model_time': 1.306281992001459, + 'grad_norm_pre_clip_avg': 0.17880100309848784, + 'learning_rate': 1.4112189311546223e-05, + 'epoch': 6.47} +04/19 [20:41:03] INFO | >> train_qwenlatent.py:487 + Step 25670 | grad_norm_pre_clip=0.1761 | + grad_norm_pre_clip_avg=0.1693 | Metrics: + {'align_loss': 0.02584552764892578, + 'recon_loss': 0.08921708166599274, + 'predict_loss': 0.00973754283040762, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17606012523174286, + 'data_time': 0.0007755630067549646, + 'model_time': 1.248909875022946, + 'grad_norm_pre_clip_avg': 0.16928520202636718, + 'learning_rate': 1.410354283116099e-05, + 'epoch': 6.48} +04/19 [20:41:16] INFO | >> train_qwenlatent.py:487 + Step 25680 | grad_norm_pre_clip=0.1900 | + grad_norm_pre_clip_avg=0.1950 | Metrics: + {'align_loss': 0.02555941604077816, + 'recon_loss': 0.07776958495378494, + 'predict_loss': 0.009910255670547485, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1900036334991455, + 'data_time': 0.0009207070106640458, + 'model_time': 1.2227778520027641, + 'grad_norm_pre_clip_avg': 0.19495952427387236, + 'learning_rate': 1.4094895575320808e-05, + 'epoch': 6.48} +04/19 [20:41:29] INFO | >> train_qwenlatent.py:487 + Step 25690 | grad_norm_pre_clip=0.2012 | + grad_norm_pre_clip_avg=0.1841 | Metrics: + {'align_loss': 0.02455560490489006, + 'recon_loss': 0.1003812775015831, + 'predict_loss': 0.00822481419891119, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20115411281585693, + 'data_time': 0.000871505995746702, + 'model_time': 1.2182164570258465, + 'grad_norm_pre_clip_avg': 0.18406101912260056, + 'learning_rate': 1.4086247548240245e-05, + 'epoch': 6.48} +04/19 [20:41:42] INFO | >> train_qwenlatent.py:487 + Step 25700 | grad_norm_pre_clip=0.2249 | + grad_norm_pre_clip_avg=0.1797 | Metrics: + {'align_loss': 0.0250343419611454, + 'recon_loss': 0.09128977358341217, + 'predict_loss': 0.010205233469605446, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2248680293560028, + 'mae_score': 0.008792000847893793, 'data_time': + 0.0011538150138221681, 'model_time': + 1.260148825997021, 'grad_norm_pre_clip_avg': + 0.17974559664726258, 'learning_rate': + 1.407759875413425e-05, 'epoch': 6.48} +04/19 [20:41:54] INFO | >> train_qwenlatent.py:487 + Step 25710 | grad_norm_pre_clip=0.1905 | + grad_norm_pre_clip_avg=0.1994 | Metrics: + {'align_loss': 0.02579236589372158, + 'recon_loss': 0.078379325568676, + 'predict_loss': 0.006277552340179682, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19051752984523773, + 'data_time': 0.0008597349806223065, + 'model_time': 1.2028797100065276, + 'grad_norm_pre_clip_avg': 0.19943867921829223, + 'learning_rate': 1.4068949197218134e-05, + 'epoch': 6.49} +04/19 [20:42:07] INFO | >> train_qwenlatent.py:487 + Step 25720 | grad_norm_pre_clip=0.2491 | + grad_norm_pre_clip_avg=0.1975 | Metrics: + {'align_loss': 0.026164472103118896, + 'recon_loss': 0.11402097344398499, + 'predict_loss': 0.00841075461357832, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24911454319953918, + 'data_time': 0.0006929769879207015, + 'model_time': 1.284465953009203, + 'grad_norm_pre_clip_avg': 0.19749338924884796, + 'learning_rate': 1.4060298881707584e-05, + 'epoch': 6.49} +04/19 [20:42:20] INFO | >> train_qwenlatent.py:487 + Step 25730 | grad_norm_pre_clip=0.2314 | + grad_norm_pre_clip_avg=0.1878 | Metrics: + {'align_loss': 0.025348521769046783, + 'recon_loss': 0.0709628090262413, + 'predict_loss': 0.007043769117444754, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23141930997371674, + 'data_time': 0.0006683379760943353, + 'model_time': 1.2831413040112238, + 'grad_norm_pre_clip_avg': 0.18776122480630875, + 'learning_rate': 1.4051647811818667e-05, + 'epoch': 6.49} +04/19 [20:42:33] INFO | >> train_qwenlatent.py:487 + Step 25740 | grad_norm_pre_clip=0.1866 | + grad_norm_pre_clip_avg=0.1937 | Metrics: + {'align_loss': 0.025608718395233154, + 'recon_loss': 0.0851476863026619, + 'predict_loss': 0.009522258304059505, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1866285502910614, + 'data_time': 0.0009752359765116125, + 'model_time': 1.2250452599837445, + 'grad_norm_pre_clip_avg': 0.19370727092027665, + 'learning_rate': 1.4042995991767799e-05, + 'epoch': 6.5} +04/19 [20:42:46] INFO | >> train_qwenlatent.py:487 + Step 25750 | grad_norm_pre_clip=0.1566 | + grad_norm_pre_clip_avg=0.1728 | Metrics: + {'align_loss': 0.026062071323394775, + 'recon_loss': 0.08277178555727005, + 'predict_loss': 0.009550592862069607, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15656204521656036, + 'mae_score': 0.012665284646523965, 'data_time': + 0.000793304992839694, 'model_time': + 1.2589407880150247, 'grad_norm_pre_clip_avg': + 0.17282940894365312, 'learning_rate': + 1.4034343425771778e-05, 'epoch': 6.5} +04/19 [20:42:59] INFO | >> train_qwenlatent.py:487 + Step 25760 | grad_norm_pre_clip=0.1612 | + grad_norm_pre_clip_avg=0.1666 | Metrics: + {'align_loss': 0.024976246058940887, + 'recon_loss': 0.09058663994073868, + 'predict_loss': 0.006582038477063179, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16115649044513702, + 'data_time': 0.0007278560078702867, + 'model_time': 1.2500928079825826, + 'grad_norm_pre_clip_avg': 0.16662925332784653, + 'learning_rate': 1.4025690118047763e-05, + 'epoch': 6.5} +04/19 [20:43:12] INFO | >> train_qwenlatent.py:487 + Step 25770 | grad_norm_pre_clip=0.1933 | + grad_norm_pre_clip_avg=0.1887 | Metrics: + {'align_loss': 0.024452529847621918, + 'recon_loss': 0.08210842311382294, + 'predict_loss': 0.007158760912716389, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19326931238174438, + 'data_time': 0.0009119340102188289, + 'model_time': 1.2496678180177696, + 'grad_norm_pre_clip_avg': 0.18872669637203215, + 'learning_rate': 1.4017036072813264e-05, + 'epoch': 6.5} +04/19 [20:43:24] INFO | >> train_qwenlatent.py:487 + Step 25780 | grad_norm_pre_clip=0.2239 | + grad_norm_pre_clip_avg=0.1799 | Metrics: + {'align_loss': 0.026005815714597702, + 'recon_loss': 0.07550528645515442, + 'predict_loss': 0.007910276763141155, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22389322519302368, + 'data_time': 0.0006842559960205108, + 'model_time': 1.2036773659929167, + 'grad_norm_pre_clip_avg': 0.1799098953604698, + 'learning_rate': 1.4008381294286164e-05, + 'epoch': 6.51} +04/19 [20:43:36] INFO | >> train_qwenlatent.py:487 + Step 25790 | grad_norm_pre_clip=0.1959 | + grad_norm_pre_clip_avg=0.1725 | Metrics: + {'align_loss': 0.026043377816677094, + 'recon_loss': 0.10017544031143188, + 'predict_loss': 0.009583336301147938, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19593192636966705, + 'data_time': 0.0006805979937780648, + 'model_time': 1.2086297569912858, + 'grad_norm_pre_clip_avg': 0.17245552241802214, + 'learning_rate': 1.3999725786684687e-05, + 'epoch': 6.51} +04/19 [20:43:49] INFO | >> train_qwenlatent.py:487 + Step 25800 | grad_norm_pre_clip=0.1546 | + grad_norm_pre_clip_avg=0.2212 | Metrics: + {'align_loss': 0.02581116557121277, + 'recon_loss': 0.08500564098358154, + 'predict_loss': 0.008937983773648739, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15464188158512115, + 'mae_score': 0.008633374738263654, 'data_time': + 0.0008891560137271881, 'model_time': + 1.294461889017839, 'grad_norm_pre_clip_avg': + 0.22124410569667816, 'learning_rate': + 1.3991069554227434e-05, 'epoch': 6.51} +04/19 [20:44:02] INFO | >> train_qwenlatent.py:487 + Step 25810 | grad_norm_pre_clip=0.2136 | + grad_norm_pre_clip_avg=0.1662 | Metrics: + {'align_loss': 0.02586062252521515, + 'recon_loss': 0.09553626924753189, + 'predict_loss': 0.00786170270293951, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21362169086933136, + 'data_time': 0.0011435789929237217, + 'model_time': 1.2879772360029165, + 'grad_norm_pre_clip_avg': 0.16615191996097564, + 'learning_rate': 1.3982412601133338e-05, + 'epoch': 6.51} +04/19 [20:44:14] INFO | >> train_qwenlatent.py:487 + Step 25820 | grad_norm_pre_clip=0.1723 | + grad_norm_pre_clip_avg=0.2391 | Metrics: + {'align_loss': 0.023513473570346832, + 'recon_loss': 0.07426615804433823, + 'predict_loss': 0.010270779952406883, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17234516143798828, + 'data_time': 0.0009916679991874844, + 'model_time': 1.2370899450033903, + 'grad_norm_pre_clip_avg': 0.23912563771009446, + 'learning_rate': 1.3973754931621699e-05, + 'epoch': 6.52} +04/19 [20:44:27] INFO | >> train_qwenlatent.py:487 + Step 25830 | grad_norm_pre_clip=0.1528 | + grad_norm_pre_clip_avg=0.2115 | Metrics: + {'align_loss': 0.024991825222969055, + 'recon_loss': 0.07146484404802322, + 'predict_loss': 0.007593237794935703, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1528089940547943, + 'data_time': 0.0006886290211696178, + 'model_time': 1.6235660380043555, + 'grad_norm_pre_clip_avg': 0.2114504396915436, + 'learning_rate': 1.3965096549912157e-05, + 'epoch': 6.52} +04/19 [20:44:40] INFO | >> train_qwenlatent.py:487 + Step 25840 | grad_norm_pre_clip=0.1880 | + grad_norm_pre_clip_avg=0.1749 | Metrics: + {'align_loss': 0.025445569306612015, + 'recon_loss': 0.09682335704565048, + 'predict_loss': 0.01111982949078083, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18799377977848053, + 'data_time': 0.000787741009844467, + 'model_time': 1.1962657709955238, + 'grad_norm_pre_clip_avg': 0.17490701228380204, + 'learning_rate': 1.3956437460224705e-05, + 'epoch': 6.52} +04/19 [20:44:53] INFO | >> train_qwenlatent.py:487 + Step 25850 | grad_norm_pre_clip=0.2182 | + grad_norm_pre_clip_avg=0.1710 | Metrics: + {'align_loss': 0.026186905801296234, + 'recon_loss': 0.0900244414806366, + 'predict_loss': 0.00929136760532856, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21816517412662506, + 'mae_score': 0.009443551140862542, 'data_time': + 0.0008410339942201972, 'model_time': + 1.2311035229940899, 'grad_norm_pre_clip_avg': + 0.1710362046957016, 'learning_rate': + 1.3947777666779673e-05, 'epoch': 6.52} +04/19 [20:45:06] INFO | >> train_qwenlatent.py:487 + Step 25860 | grad_norm_pre_clip=0.1622 | + grad_norm_pre_clip_avg=0.1561 | Metrics: + {'align_loss': 0.02603299915790558, + 'recon_loss': 0.10586977005004883, + 'predict_loss': 0.012733468785881996, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16222235560417175, + 'data_time': 0.0008821249939501286, + 'model_time': 1.2557460889802314, + 'grad_norm_pre_clip_avg': 0.1561239778995514, + 'learning_rate': 1.3939117173797744e-05, + 'epoch': 6.53} +04/19 [20:45:19] INFO | >> train_qwenlatent.py:487 + Step 25870 | grad_norm_pre_clip=0.2233 | + grad_norm_pre_clip_avg=0.1906 | Metrics: + {'align_loss': 0.026003703474998474, + 'recon_loss': 0.07273992896080017, + 'predict_loss': 0.01413317583501339, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22326678037643433, + 'data_time': 0.0006820540002081543, + 'model_time': 1.250508086028276, + 'grad_norm_pre_clip_avg': 0.19064896553754807, + 'learning_rate': 1.3930455985499936e-05, + 'epoch': 6.53} +04/19 [20:45:31] INFO | >> train_qwenlatent.py:487 + Step 25880 | grad_norm_pre_clip=0.1955 | + grad_norm_pre_clip_avg=0.2365 | Metrics: + {'align_loss': 0.02575472556054592, + 'recon_loss': 0.08447491377592087, + 'predict_loss': 0.00884279515594244, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1954972892999649, + 'data_time': 0.0006771249754820019, + 'model_time': 1.2327599599957466, + 'grad_norm_pre_clip_avg': 0.23648478239774703, + 'learning_rate': 1.3921794106107602e-05, + 'epoch': 6.53} +04/19 [20:45:44] INFO | >> train_qwenlatent.py:487 + Step 25890 | grad_norm_pre_clip=0.2217 | + grad_norm_pre_clip_avg=0.1989 | Metrics: + {'align_loss': 0.024989912286400795, + 'recon_loss': 0.09586412459611893, + 'predict_loss': 0.011024273931980133, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22165165841579437, + 'data_time': 0.0010503670200705528, + 'model_time': 1.3240673420077655, + 'grad_norm_pre_clip_avg': 0.1988660916686058, + 'learning_rate': 1.3913131539842448e-05, + 'epoch': 6.53} +04/19 [20:45:58] INFO | >> train_qwenlatent.py:487 + Step 25900 | grad_norm_pre_clip=0.1720 | + grad_norm_pre_clip_avg=0.1691 | Metrics: + {'align_loss': 0.023627692833542824, + 'recon_loss': 0.07787896692752838, + 'predict_loss': 0.008726445958018303, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17203626036643982, + 'mae_score': 0.012103581643319344, 'data_time': + 0.0006890269869472831, 'model_time': + 1.2233645989908837, 'grad_norm_pre_clip_avg': + 0.16909041851758957, 'learning_rate': + 1.390446829092649e-05, 'epoch': 6.54} +04/19 [20:46:10] INFO | >> train_qwenlatent.py:487 + Step 25910 | grad_norm_pre_clip=0.1533 | + grad_norm_pre_clip_avg=0.1603 | Metrics: + {'align_loss': 0.02627096138894558, + 'recon_loss': 0.09172863513231277, + 'predict_loss': 0.012135562486946583, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15333247184753418, + 'data_time': 0.0008572599908802658, + 'model_time': 1.253814492985839, + 'grad_norm_pre_clip_avg': 0.1603447511792183, + 'learning_rate': 1.3895804363582099e-05, + 'epoch': 6.54} +04/19 [20:46:23] INFO | >> train_qwenlatent.py:487 + Step 25920 | grad_norm_pre_clip=0.2500 | + grad_norm_pre_clip_avg=0.1677 | Metrics: + {'align_loss': 0.025335393846035004, + 'recon_loss': 0.10394405573606491, + 'predict_loss': 0.015053348615765572, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2500278949737549, + 'data_time': 0.0006216669862624258, + 'model_time': 1.2576004629954696, + 'grad_norm_pre_clip_avg': 0.16771047562360764, + 'learning_rate': 1.3887139762031967e-05, + 'epoch': 6.54} +04/19 [20:46:35] INFO | >> train_qwenlatent.py:487 + Step 25930 | grad_norm_pre_clip=0.2097 | + grad_norm_pre_clip_avg=0.2706 | Metrics: + {'align_loss': 0.024670429527759552, + 'recon_loss': 0.06389734894037247, + 'predict_loss': 0.00760256964713335, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20971304178237915, + 'data_time': 0.0006592389836441725, + 'model_time': 1.2156083449954167, + 'grad_norm_pre_clip_avg': 0.2705592900514603, + 'learning_rate': 1.3878474490499114e-05, + 'epoch': 6.54} +04/19 [20:46:48] INFO | >> train_qwenlatent.py:487 + Step 25940 | grad_norm_pre_clip=0.1882 | + grad_norm_pre_clip_avg=0.2031 | Metrics: + {'align_loss': 0.02640526369214058, + 'recon_loss': 0.08492705971002579, + 'predict_loss': 0.012158939614892006, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18819552659988403, + 'data_time': 0.0006764699937775731, + 'model_time': 1.2878245930187404, + 'grad_norm_pre_clip_avg': 0.20305193960666656, + 'learning_rate': 1.386980855320689e-05, + 'epoch': 6.55} +04/19 [20:47:01] INFO | >> train_qwenlatent.py:487 + Step 25950 | grad_norm_pre_clip=0.1938 | + grad_norm_pre_clip_avg=0.1803 | Metrics: + {'align_loss': 0.02525082789361477, + 'recon_loss': 0.09466785192489624, + 'predict_loss': 0.00907938927412033, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1937660574913025, + 'mae_score': 0.009152268074654244, 'data_time': + 0.0008785660029388964, 'model_time': + 1.2260447879962157, 'grad_norm_pre_clip_avg': + 0.1802606150507927, 'learning_rate': + 1.3861141954378961e-05, 'epoch': 6.55} +04/19 [20:47:14] INFO | >> train_qwenlatent.py:487 + Step 25960 | grad_norm_pre_clip=0.2068 | + grad_norm_pre_clip_avg=0.1979 | Metrics: + {'align_loss': 0.02499762922525406, + 'recon_loss': 0.09018097072839737, + 'predict_loss': 0.01225782185792923, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20676329731941223, + 'data_time': 0.0009130070102401078, + 'model_time': 1.2388112540065777, + 'grad_norm_pre_clip_avg': 0.19791592061519622, + 'learning_rate': 1.3852474698239329e-05, + 'epoch': 6.55} +04/19 [20:47:26] INFO | >> train_qwenlatent.py:487 + Step 25970 | grad_norm_pre_clip=0.1459 | + grad_norm_pre_clip_avg=0.1625 | Metrics: + {'align_loss': 0.025663483887910843, + 'recon_loss': 0.09608623385429382, + 'predict_loss': 0.007496991194784641, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14585429430007935, + 'data_time': 0.0007432340062223375, + 'model_time': 1.2717791850154754, + 'grad_norm_pre_clip_avg': 0.16248963326215743, + 'learning_rate': 1.384380678901231e-05, + 'epoch': 6.55} +04/19 [20:47:39] INFO | >> train_qwenlatent.py:487 + Step 25980 | grad_norm_pre_clip=0.1487 | + grad_norm_pre_clip_avg=0.1636 | Metrics: + {'align_loss': 0.024493515491485596, + 'recon_loss': 0.05964406579732895, + 'predict_loss': 0.005471345502883196, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1487191766500473, + 'data_time': 0.0007097229827195406, + 'model_time': 1.2126805059961043, + 'grad_norm_pre_clip_avg': 0.16364270001649855, + 'learning_rate': 1.3835138230922535e-05, + 'epoch': 6.56} +04/19 [20:47:51] INFO | >> train_qwenlatent.py:487 + Step 25990 | grad_norm_pre_clip=0.2405 | + grad_norm_pre_clip_avg=0.2019 | Metrics: + {'align_loss': 0.024135127663612366, + 'recon_loss': 0.0778222307562828, + 'predict_loss': 0.014966525137424469, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24054588377475739, + 'data_time': 0.0008831140003167093, + 'model_time': 1.2133624060079455, + 'grad_norm_pre_clip_avg': 0.20188113301992416, + 'learning_rate': 1.3826469028194953e-05, + 'epoch': 6.56} +04/19 [20:48:05] INFO | >> train_qwenlatent.py:487 + Step 26000 | grad_norm_pre_clip=0.1292 | + grad_norm_pre_clip_avg=0.1938 | Metrics: + {'align_loss': 0.025508645921945572, + 'recon_loss': 0.07323101162910461, + 'predict_loss': 0.007900589145720005, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12915316224098206, + 'mae_score': 0.008534412555866414, 'data_time': + 0.000857010978506878, 'model_time': + 1.277111175993923, 'grad_norm_pre_clip_avg': + 0.19382872581481933, 'learning_rate': + 1.3817799185054824e-05, 'epoch': 6.56} +04/19 [20:48:18] INFO | >> train_qwenlatent.py:487 + Step 26010 | grad_norm_pre_clip=0.2251 | + grad_norm_pre_clip_avg=0.1616 | Metrics: + {'align_loss': 0.02579837664961815, + 'recon_loss': 0.0969354510307312, + 'predict_loss': 0.01560873631387949, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22509780526161194, + 'data_time': 0.00069578000693582, 'model_time': + 1.2250517399806995, 'grad_norm_pre_clip_avg': + 0.16156916469335555, 'learning_rate': + 1.380912870572773e-05, 'epoch': 6.56} +04/19 [20:48:30] INFO | >> train_qwenlatent.py:487 + Step 26020 | grad_norm_pre_clip=0.3136 | + grad_norm_pre_clip_avg=0.2743 | Metrics: + {'align_loss': 0.025984881445765495, + 'recon_loss': 0.11231659352779388, + 'predict_loss': 0.013846596702933311, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3136014938354492, + 'data_time': 0.0009345980070065707, + 'model_time': 1.203454881993821, + 'grad_norm_pre_clip_avg': 0.27425950318574904, + 'learning_rate': 1.3800457594439562e-05, + 'epoch': 6.57} +04/19 [20:48:43] INFO | >> train_qwenlatent.py:487 + Step 26030 | grad_norm_pre_clip=0.2217 | + grad_norm_pre_clip_avg=0.2105 | Metrics: + {'align_loss': 0.026315510272979736, + 'recon_loss': 0.07970377057790756, + 'predict_loss': 0.009374486282467842, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22169412672519684, + 'data_time': 0.0008982550061773509, + 'model_time': 1.268996008002432, + 'grad_norm_pre_clip_avg': 0.2104828178882599, + 'learning_rate': 1.3791785855416506e-05, + 'epoch': 6.57} +04/19 [20:48:55] INFO | >> train_qwenlatent.py:487 + Step 26040 | grad_norm_pre_clip=0.1447 | + grad_norm_pre_clip_avg=0.1647 | Metrics: + {'align_loss': 0.02430950477719307, + 'recon_loss': 0.08217724412679672, + 'predict_loss': 0.009809465147554874, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14467193186283112, + 'data_time': 0.0006675870099570602, + 'model_time': 1.2107310249994043, + 'grad_norm_pre_clip_avg': 0.1646702691912651, + 'learning_rate': 1.3783113492885066e-05, + 'epoch': 6.57} +04/19 [20:49:08] INFO | >> train_qwenlatent.py:487 + Step 26050 | grad_norm_pre_clip=0.1487 | + grad_norm_pre_clip_avg=0.1551 | Metrics: + {'align_loss': 0.02567765861749649, + 'recon_loss': 0.06380098313093185, + 'predict_loss': 0.008482326753437519, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14868205785751343, + 'mae_score': 0.01381728541743648, 'data_time': + 0.0006743170088157058, 'model_time': + 1.2525975710013881, 'grad_norm_pre_clip_avg': + 0.15509215816855432, 'learning_rate': + 1.3774440511072043e-05, 'epoch': 6.57} +04/19 [20:49:21] INFO | >> train_qwenlatent.py:487 + Step 26060 | grad_norm_pre_clip=0.1655 | + grad_norm_pre_clip_avg=0.1872 | Metrics: + {'align_loss': 0.024193275719881058, + 'recon_loss': 0.06712913513183594, + 'predict_loss': 0.005399375222623348, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16548305749893188, + 'data_time': 0.000772936997236684, + 'model_time': 1.2063862119975965, + 'grad_norm_pre_clip_avg': 0.18719104528427125, + 'learning_rate': 1.3765766914204545e-05, + 'epoch': 6.58} +04/19 [20:49:33] INFO | >> train_qwenlatent.py:487 + Step 26070 | grad_norm_pre_clip=0.2432 | + grad_norm_pre_clip_avg=0.1770 | Metrics: + {'align_loss': 0.025268951430916786, + 'recon_loss': 0.09517505019903183, + 'predict_loss': 0.011875065043568611, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24324633181095123, + 'data_time': 0.0007270829810295254, + 'model_time': 1.2462760669877753, + 'grad_norm_pre_clip_avg': 0.17701853364706038, + 'learning_rate': 1.3757092706509984e-05, + 'epoch': 6.58} +04/19 [20:49:46] INFO | >> train_qwenlatent.py:487 + Step 26080 | grad_norm_pre_clip=0.2232 | + grad_norm_pre_clip_avg=0.2095 | Metrics: + {'align_loss': 0.0261895339936018, + 'recon_loss': 0.11491194367408752, + 'predict_loss': 0.015926621854305267, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22318653762340546, + 'data_time': 0.0006574580038432032, + 'model_time': 1.2594842980033718, + 'grad_norm_pre_clip_avg': 0.20954852402210236, + 'learning_rate': 1.374841789221605e-05, + 'epoch': 6.58} +04/19 [20:49:59] INFO | >> train_qwenlatent.py:487 + Step 26090 | grad_norm_pre_clip=0.1553 | + grad_norm_pre_clip_avg=0.1892 | Metrics: + {'align_loss': 0.0241076722741127, + 'recon_loss': 0.05790192261338234, + 'predict_loss': 0.00728189991787076, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15533088147640228, + 'data_time': 0.000959127995884046, + 'model_time': 1.2484309340070467, + 'grad_norm_pre_clip_avg': 0.1892135590314865, + 'learning_rate': 1.3739742475550751e-05, + 'epoch': 6.58} +04/19 [20:50:12] INFO | >> train_qwenlatent.py:487 + Step 26100 | grad_norm_pre_clip=0.1827 | + grad_norm_pre_clip_avg=0.1929 | Metrics: + {'align_loss': 0.02523554116487503, + 'recon_loss': 0.09824715554714203, + 'predict_loss': 0.011211144737899303, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18270844221115112, + 'mae_score': 0.010550735233066318, 'data_time': + 0.0006739019881933928, 'model_time': + 1.264085732982494, 'grad_norm_pre_clip_avg': + 0.1929350033402443, 'learning_rate': + 1.3731066460742379e-05, 'epoch': 6.59} +04/19 [20:50:25] INFO | >> train_qwenlatent.py:487 + Step 26110 | grad_norm_pre_clip=0.1559 | + grad_norm_pre_clip_avg=0.1710 | Metrics: + {'align_loss': 0.026249539107084274, + 'recon_loss': 0.08426517993211746, + 'predict_loss': 0.01016706321388483, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15594708919525146, + 'data_time': 0.0006478879950009286, + 'model_time': 1.2125420859956648, + 'grad_norm_pre_clip_avg': 0.1710258260369301, + 'learning_rate': 1.3722389852019517e-05, + 'epoch': 6.59} +04/19 [20:50:37] INFO | >> train_qwenlatent.py:487 + Step 26120 | grad_norm_pre_clip=0.1590 | + grad_norm_pre_clip_avg=0.1472 | Metrics: + {'align_loss': 0.024712637066841125, + 'recon_loss': 0.10184219479560852, + 'predict_loss': 0.008901981636881828, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15898221731185913, + 'data_time': 0.0007437969907186925, + 'model_time': 1.2476959239866119, + 'grad_norm_pre_clip_avg': 0.14718205779790877, + 'learning_rate': 1.371371265361104e-05, + 'epoch': 6.59} +04/19 [20:50:49] INFO | >> train_qwenlatent.py:487 + Step 26130 | grad_norm_pre_clip=0.1933 | + grad_norm_pre_clip_avg=0.2122 | Metrics: + {'align_loss': 0.02681225910782814, + 'recon_loss': 0.10099516063928604, + 'predict_loss': 0.0075528002344071865, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1933421939611435, + 'data_time': 0.0007710770005360246, + 'model_time': 1.2432746539998334, + 'grad_norm_pre_clip_avg': 0.21224749833345413, + 'learning_rate': 1.3705034869746112e-05, + 'epoch': 6.59} +04/19 [20:51:02] INFO | >> train_qwenlatent.py:487 + Step 26140 | grad_norm_pre_clip=0.2078 | + grad_norm_pre_clip_avg=0.2027 | Metrics: + {'align_loss': 0.02527007833123207, + 'recon_loss': 0.08319748938083649, + 'predict_loss': 0.015388262458145618, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20777064561843872, + 'data_time': 0.0009714369953144342, + 'model_time': 1.5668611190048978, + 'grad_norm_pre_clip_avg': 0.20271724462509155, + 'learning_rate': 1.3696356504654167e-05, + 'epoch': 6.6} +04/19 [20:51:16] INFO | >> train_qwenlatent.py:487 + Step 26150 | grad_norm_pre_clip=0.1772 | + grad_norm_pre_clip_avg=0.1937 | Metrics: + {'align_loss': 0.026670102030038834, + 'recon_loss': 0.11471936106681824, + 'predict_loss': 0.015953779220581055, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17723721265792847, + 'mae_score': 0.011664351901492558, 'data_time': + 0.0008800410141702741, 'model_time': + 1.2202876310038846, 'grad_norm_pre_clip_avg': + 0.19370232075452803, 'learning_rate': + 1.3687677562564951e-05, 'epoch': 6.6} +04/19 [20:51:29] INFO | >> train_qwenlatent.py:487 + Step 26160 | grad_norm_pre_clip=0.2448 | + grad_norm_pre_clip_avg=0.2004 | Metrics: + {'align_loss': 0.025971896946430206, + 'recon_loss': 0.09303612262010574, + 'predict_loss': 0.011519171297550201, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2448401302099228, + 'data_time': 0.0007994059997145087, + 'model_time': 1.2742273250187282, + 'grad_norm_pre_clip_avg': 0.20036525875329972, + 'learning_rate': 1.3678998047708463e-05, + 'epoch': 6.6} +04/19 [20:51:42] INFO | >> train_qwenlatent.py:487 + Step 26170 | grad_norm_pre_clip=0.1588 | + grad_norm_pre_clip_avg=0.2049 | Metrics: + {'align_loss': 0.025496110320091248, + 'recon_loss': 0.0826466903090477, + 'predict_loss': 0.01104098279029131, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15883131325244904, + 'data_time': 0.0010806990030687302, + 'model_time': 1.2875379399920348, + 'grad_norm_pre_clip_avg': 0.2049287423491478, + 'learning_rate': 1.3670317964314996e-05, + 'epoch': 6.6} +04/19 [20:51:54] INFO | >> train_qwenlatent.py:487 + Step 26180 | grad_norm_pre_clip=0.1959 | + grad_norm_pre_clip_avg=0.1879 | Metrics: + {'align_loss': 0.026236440986394882, + 'recon_loss': 0.10284910351037979, + 'predict_loss': 0.014344516210258007, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19590480625629425, + 'data_time': 0.0014294339925982058, + 'model_time': 1.2590228329936508, + 'grad_norm_pre_clip_avg': 0.1878875747323036, + 'learning_rate': 1.3661637316615123e-05, + 'epoch': 6.61} +04/19 [20:52:06] INFO | >> train_qwenlatent.py:487 + Step 26190 | grad_norm_pre_clip=0.1616 | + grad_norm_pre_clip_avg=0.1624 | Metrics: + {'align_loss': 0.025736240670084953, + 'recon_loss': 0.08450169116258621, + 'predict_loss': 0.00866362638771534, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16161030530929565, + 'data_time': 0.0007016409945208579, + 'model_time': 1.2445409689971711, + 'grad_norm_pre_clip_avg': 0.1623731166124344, + 'learning_rate': 1.3652956108839676e-05, + 'epoch': 6.61} +04/19 [20:52:20] INFO | >> train_qwenlatent.py:487 + Step 26200 | grad_norm_pre_clip=0.1721 | + grad_norm_pre_clip_avg=0.1948 | Metrics: + {'align_loss': 0.02526361495256424, + 'recon_loss': 0.0891290083527565, + 'predict_loss': 0.009301955811679363, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17212611436843872, + 'mae_score': 0.011193694724692956, 'data_time': + 0.000766287004807964, 'model_time': + 1.2316402420110535, 'grad_norm_pre_clip_avg': + 0.19475104808807372, 'learning_rate': + 1.3644274345219782e-05, 'epoch': 6.61} +04/19 [20:52:32] INFO | >> train_qwenlatent.py:487 + Step 26210 | grad_norm_pre_clip=0.1711 | + grad_norm_pre_clip_avg=0.1896 | Metrics: + {'align_loss': 0.02673027478158474, + 'recon_loss': 0.12782137095928192, + 'predict_loss': 0.012602264061570168, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17108120024204254, + 'data_time': 0.000654646020848304, + 'model_time': 1.2298004539916292, + 'grad_norm_pre_clip_avg': 0.18958616405725479, + 'learning_rate': 1.3635592029986816e-05, + 'epoch': 6.61} +04/19 [20:52:45] INFO | >> train_qwenlatent.py:487 + Step 26220 | grad_norm_pre_clip=0.2028 | + grad_norm_pre_clip_avg=0.1760 | Metrics: + {'align_loss': 0.024658197537064552, + 'recon_loss': 0.11363823711872101, + 'predict_loss': 0.01080531906336546, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20280076563358307, + 'data_time': 0.0009749930177349597, + 'model_time': 1.2343559630098753, + 'grad_norm_pre_clip_avg': 0.1760243535041809, + 'learning_rate': 1.3626909167372443e-05, + 'epoch': 6.62} +04/19 [20:52:57] INFO | >> train_qwenlatent.py:487 + Step 26230 | grad_norm_pre_clip=0.1686 | + grad_norm_pre_clip_avg=0.2063 | Metrics: + {'align_loss': 0.02450461871922016, + 'recon_loss': 0.0809256061911583, + 'predict_loss': 0.011515102349221706, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16860228776931763, + 'data_time': 0.0011299990001134574, + 'model_time': 1.2387929129763506, + 'grad_norm_pre_clip_avg': 0.2062908887863159, + 'learning_rate': 1.3618225761608575e-05, + 'epoch': 6.62} +04/19 [20:53:10] INFO | >> train_qwenlatent.py:487 + Step 26240 | grad_norm_pre_clip=0.1734 | + grad_norm_pre_clip_avg=0.2122 | Metrics: + {'align_loss': 0.02695542946457863, + 'recon_loss': 0.11032072454690933, + 'predict_loss': 0.011800136417150497, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17341099679470062, + 'data_time': 0.0009299759985879064, + 'model_time': 1.2216402850172017, + 'grad_norm_pre_clip_avg': 0.2121949553489685, + 'learning_rate': 1.360954181692741e-05, + 'epoch': 6.62} +04/19 [20:53:23] INFO | >> train_qwenlatent.py:487 + Step 26250 | grad_norm_pre_clip=0.1630 | + grad_norm_pre_clip_avg=0.1717 | Metrics: + {'align_loss': 0.02569516934454441, + 'recon_loss': 0.09510385990142822, + 'predict_loss': 0.009790956974029541, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1630396991968155, + 'mae_score': 0.009715894750646643, 'data_time': + 0.0006743519916199148, 'model_time': + 1.2068691460008267, 'grad_norm_pre_clip_avg': + 0.1716926708817482, 'learning_rate': + 1.3600857337561384e-05, 'epoch': 6.62} +04/19 [20:53:35] INFO | >> train_qwenlatent.py:487 + Step 26260 | grad_norm_pre_clip=0.1590 | + grad_norm_pre_clip_avg=0.1696 | Metrics: + {'align_loss': 0.02509646862745285, + 'recon_loss': 0.08453664183616638, + 'predict_loss': 0.009226452559232712, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15904949605464935, + 'data_time': 0.0009030899964272976, + 'model_time': 1.2272894889756572, + 'grad_norm_pre_clip_avg': 0.16961268037557603, + 'learning_rate': 1.3592172327743214e-05, + 'epoch': 6.63} +04/19 [20:53:48] INFO | >> train_qwenlatent.py:487 + Step 26270 | grad_norm_pre_clip=0.1930 | + grad_norm_pre_clip_avg=0.2067 | Metrics: + {'align_loss': 0.02505742385983467, + 'recon_loss': 0.1344430148601532, + 'predict_loss': 0.01839989610016346, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1930343061685562, + 'data_time': 0.0008778809860814363, + 'model_time': 1.2101543030003086, + 'grad_norm_pre_clip_avg': 0.20669962018728255, + 'learning_rate': 1.3583486791705868e-05, + 'epoch': 6.63} +04/19 [20:54:01] INFO | >> train_qwenlatent.py:487 + Step 26280 | grad_norm_pre_clip=0.2085 | + grad_norm_pre_clip_avg=0.1963 | Metrics: + {'align_loss': 0.02680785581469536, + 'recon_loss': 0.11182267218828201, + 'predict_loss': 0.013918989337980747, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20849542319774628, + 'data_time': 0.0006456180126406252, + 'model_time': 1.2119081299751997, + 'grad_norm_pre_clip_avg': 0.19634488224983215, + 'learning_rate': 1.3574800733682572e-05, + 'epoch': 6.63} +04/19 [20:54:14] INFO | >> train_qwenlatent.py:487 + Step 26290 | grad_norm_pre_clip=0.1743 | + grad_norm_pre_clip_avg=0.1837 | Metrics: + {'align_loss': 0.024806158617138863, + 'recon_loss': 0.07217200100421906, + 'predict_loss': 0.010002166032791138, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1742950826883316, + 'data_time': 0.0007102140225470066, + 'model_time': 1.2432074329990428, + 'grad_norm_pre_clip_avg': 0.18367753624916078, + 'learning_rate': 1.3566114157906801e-05, + 'epoch': 6.63} +04/19 [20:54:27] INFO | >> train_qwenlatent.py:487 + Step 26300 | grad_norm_pre_clip=0.1895 | + grad_norm_pre_clip_avg=0.1836 | Metrics: + {'align_loss': 0.025547482073307037, + 'recon_loss': 0.09447941184043884, + 'predict_loss': 0.009062777273356915, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18951845169067383, + 'mae_score': 0.008963103766913887, 'data_time': + 0.0008877819927874953, 'model_time': + 1.22468904798734, 'grad_norm_pre_clip_avg': + 0.18357887119054794, 'learning_rate': + 1.3557427068612293e-05, 'epoch': 6.64} +04/19 [20:54:40] INFO | >> train_qwenlatent.py:487 + Step 26310 | grad_norm_pre_clip=0.2148 | + grad_norm_pre_clip_avg=0.2073 | Metrics: + {'align_loss': 0.02550400421023369, + 'recon_loss': 0.07766902446746826, + 'predict_loss': 0.011083909310400486, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21483741700649261, + 'data_time': 0.0006155830051284283, + 'model_time': 1.284034234995488, + 'grad_norm_pre_clip_avg': 0.20728372186422347, + 'learning_rate': 1.354873947003302e-05, + 'epoch': 6.64} +04/19 [20:54:52] INFO | >> train_qwenlatent.py:487 + Step 26320 | grad_norm_pre_clip=0.1510 | + grad_norm_pre_clip_avg=0.2036 | Metrics: + {'align_loss': 0.02462085522711277, + 'recon_loss': 0.09331025928258896, + 'predict_loss': 0.009867082349956036, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15095563232898712, + 'data_time': 0.0006691320159006864, + 'model_time': 1.2269514669897035, + 'grad_norm_pre_clip_avg': 0.20358122289180755, + 'learning_rate': 1.354005136640322e-05, + 'epoch': 6.64} +04/19 [20:55:05] INFO | >> train_qwenlatent.py:487 + Step 26330 | grad_norm_pre_clip=0.2560 | + grad_norm_pre_clip_avg=0.1897 | Metrics: + {'align_loss': 0.025356486439704895, + 'recon_loss': 0.13023406267166138, + 'predict_loss': 0.01588563248515129, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2559596002101898, + 'data_time': 0.0007230390037875623, + 'model_time': 1.4525209000275936, + 'grad_norm_pre_clip_avg': 0.18965971618890762, + 'learning_rate': 1.353136276195737e-05, + 'epoch': 6.64} +04/19 [20:55:17] INFO | >> train_qwenlatent.py:487 + Step 26340 | grad_norm_pre_clip=0.2205 | + grad_norm_pre_clip_avg=0.1929 | Metrics: + {'align_loss': 0.026033634319901466, + 'recon_loss': 0.09325163811445236, + 'predict_loss': 0.009159641340374947, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22049568593502045, + 'data_time': 0.0008854950137902051, + 'model_time': 1.2105203279934358, + 'grad_norm_pre_clip_avg': 0.192877396941185, + 'learning_rate': 1.3522673660930183e-05, + 'epoch': 6.65} +04/19 [20:55:30] INFO | >> train_qwenlatent.py:487 + Step 26350 | grad_norm_pre_clip=0.1687 | + grad_norm_pre_clip_avg=0.2191 | Metrics: + {'align_loss': 0.023714346811175346, + 'recon_loss': 0.08412372320890427, + 'predict_loss': 0.008035853505134583, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16870683431625366, + 'mae_score': 0.010177354554872255, 'data_time': + 0.0007410859980154783, 'model_time': + 1.22510278201662, 'grad_norm_pre_clip_avg': + 0.21914490759372712, 'learning_rate': + 1.3513984067556626e-05, 'epoch': 6.65} +04/19 [20:55:43] INFO | >> train_qwenlatent.py:487 + Step 26360 | grad_norm_pre_clip=0.1566 | + grad_norm_pre_clip_avg=0.1960 | Metrics: + {'align_loss': 0.025670630857348442, + 'recon_loss': 0.08307813853025436, + 'predict_loss': 0.006812085397541523, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15658153593540192, + 'data_time': 0.0010681790008675307, + 'model_time': 1.220318307983689, + 'grad_norm_pre_clip_avg': 0.19596021622419357, + 'learning_rate': 1.3505293986071901e-05, + 'epoch': 6.65} +04/19 [20:55:55] INFO | >> train_qwenlatent.py:487 + Step 26370 | grad_norm_pre_clip=0.1500 | + grad_norm_pre_clip_avg=0.1707 | Metrics: + {'align_loss': 0.02551615796983242, + 'recon_loss': 0.0889088436961174, + 'predict_loss': 0.008188692852854729, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14997968077659607, + 'data_time': 0.0006628950068261474, + 'model_time': 1.183103791991016, + 'grad_norm_pre_clip_avg': 0.17067569345235825, + 'learning_rate': 1.3496603420711448e-05, + 'epoch': 6.65} +04/19 [20:56:08] INFO | >> train_qwenlatent.py:487 + Step 26380 | grad_norm_pre_clip=0.2243 | + grad_norm_pre_clip_avg=0.1683 | Metrics: + {'align_loss': 0.02527833916246891, + 'recon_loss': 0.11288655549287796, + 'predict_loss': 0.01685059815645218, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2243359237909317, + 'data_time': 0.0008768750121816993, + 'model_time': 1.5633698620076757, + 'grad_norm_pre_clip_avg': 0.16830798238515854, + 'learning_rate': 1.3487912375710945e-05, + 'epoch': 6.66} +04/19 [20:56:20] INFO | >> train_qwenlatent.py:487 + Step 26390 | grad_norm_pre_clip=0.1778 | + grad_norm_pre_clip_avg=0.1802 | Metrics: + {'align_loss': 0.025692079216241837, + 'recon_loss': 0.11753075569868088, + 'predict_loss': 0.013476795516908169, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17779596149921417, + 'data_time': 0.0006532520055770874, + 'model_time': 1.226239880983485, + 'grad_norm_pre_clip_avg': 0.18022597432136536, + 'learning_rate': 1.3479220855306299e-05, + 'epoch': 6.66} +04/19 [20:56:33] INFO | >> train_qwenlatent.py:487 + Step 26400 | grad_norm_pre_clip=0.1707 | + grad_norm_pre_clip_avg=0.1880 | Metrics: + {'align_loss': 0.025123372673988342, + 'recon_loss': 0.08610911667346954, + 'predict_loss': 0.004905010107904673, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17067603766918182, + 'mae_score': 0.011320218954000387, 'data_time': + 0.0007927589758764952, 'model_time': + 1.2639265210018493, 'grad_norm_pre_clip_avg': + 0.18795767575502395, 'learning_rate': + 1.3470528863733648e-05, 'epoch': 6.66} +04/19 [20:56:46] INFO | >> train_qwenlatent.py:487 + Step 26410 | grad_norm_pre_clip=0.1879 | + grad_norm_pre_clip_avg=0.1649 | Metrics: + {'align_loss': 0.025462085381150246, + 'recon_loss': 0.10146106779575348, + 'predict_loss': 0.015093078836798668, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1879197210073471, + 'data_time': 0.0007698920089751482, + 'model_time': 1.2139289279875811, + 'grad_norm_pre_clip_avg': 0.16489079520106315, + 'learning_rate': 1.346183640522937e-05, + 'epoch': 6.66} +04/19 [20:56:59] INFO | >> train_qwenlatent.py:487 + Step 26420 | grad_norm_pre_clip=0.1560 | + grad_norm_pre_clip_avg=0.1851 | Metrics: + {'align_loss': 0.025568781420588493, + 'recon_loss': 0.11801417171955109, + 'predict_loss': 0.011081330478191376, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15604551136493683, + 'data_time': 0.0006840720016043633, + 'model_time': 1.2312114379892591, + 'grad_norm_pre_clip_avg': 0.18509168326854705, + 'learning_rate': 1.3453143484030062e-05, + 'epoch': 6.67} +04/19 [20:57:12] INFO | >> train_qwenlatent.py:487 + Step 26430 | grad_norm_pre_clip=0.1765 | + grad_norm_pre_clip_avg=0.1969 | Metrics: + {'align_loss': 0.02591509371995926, + 'recon_loss': 0.08672082424163818, + 'predict_loss': 0.006455983500927687, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17652678489685059, + 'data_time': 0.001037600974086672, + 'model_time': 1.3048942949972115, + 'grad_norm_pre_clip_avg': 0.19689863324165344, + 'learning_rate': 1.3444450104372545e-05, + 'epoch': 6.67} +04/19 [20:57:24] INFO | >> train_qwenlatent.py:487 + Step 26440 | grad_norm_pre_clip=0.1841 | + grad_norm_pre_clip_avg=0.1963 | Metrics: + {'align_loss': 0.02614365890622139, + 'recon_loss': 0.10363245010375977, + 'predict_loss': 0.012659982778131962, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18405339121818542, + 'data_time': 0.0010816180147230625, + 'model_time': 1.2652484420104884, + 'grad_norm_pre_clip_avg': 0.19626742154359816, + 'learning_rate': 1.343575627049387e-05, + 'epoch': 6.67} +04/19 [20:57:37] INFO | >> train_qwenlatent.py:487 + Step 26450 | grad_norm_pre_clip=0.2073 | + grad_norm_pre_clip_avg=0.1781 | Metrics: + {'align_loss': 0.026624402031302452, + 'recon_loss': 0.11695262789726257, + 'predict_loss': 0.008911790326237679, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20732498168945312, + 'mae_score': 0.011480140686035156, 'data_time': + 0.0009569829853717238, 'model_time': + 1.2242259870108683, 'grad_norm_pre_clip_avg': + 0.17807566076517106, 'learning_rate': + 1.3427061986631301e-05, 'epoch': 6.67} +04/19 [20:57:50] INFO | >> train_qwenlatent.py:487 + Step 26460 | grad_norm_pre_clip=0.1761 | + grad_norm_pre_clip_avg=0.1755 | Metrics: + {'align_loss': 0.025816570967435837, + 'recon_loss': 0.09510239958763123, + 'predict_loss': 0.010687858797609806, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17607444524765015, + 'data_time': 0.0007787799986544997, + 'model_time': 1.2041042640048545, + 'grad_norm_pre_clip_avg': 0.17551248222589494, + 'learning_rate': 1.3418367257022336e-05, + 'epoch': 6.68} +04/19 [20:58:02] INFO | >> train_qwenlatent.py:487 + Step 26470 | grad_norm_pre_clip=0.2304 | + grad_norm_pre_clip_avg=0.1882 | Metrics: + {'align_loss': 0.025549113750457764, + 'recon_loss': 0.07156295329332352, + 'predict_loss': 0.010815414600074291, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23044051229953766, + 'data_time': 0.001220945006934926, + 'model_time': 1.1993920450040605, + 'grad_norm_pre_clip_avg': 0.18824442923069, + 'learning_rate': 1.3409672085904668e-05, + 'epoch': 6.68} +04/19 [20:58:15] INFO | >> train_qwenlatent.py:487 + Step 26480 | grad_norm_pre_clip=0.1878 | + grad_norm_pre_clip_avg=0.1789 | Metrics: + {'align_loss': 0.02599993720650673, + 'recon_loss': 0.10988977551460266, + 'predict_loss': 0.013223787769675255, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1878025084733963, + 'data_time': 0.0007001770136412233, + 'model_time': 1.2500912890245672, + 'grad_norm_pre_clip_avg': 0.17888523042201995, + 'learning_rate': 1.3400976477516227e-05, + 'epoch': 6.68} +04/19 [20:58:27] INFO | >> train_qwenlatent.py:487 + Step 26490 | grad_norm_pre_clip=0.1905 | + grad_norm_pre_clip_avg=0.1858 | Metrics: + {'align_loss': 0.026109728962183, 'recon_loss': + 0.11979940533638, 'predict_loss': + 0.011308369226753712, 'aux_loss_decay_weight': + 0.0, 'grad_norm_pre_clip': 0.1904921531677246, + 'data_time': 0.0006718939985148609, + 'model_time': 1.174479269015137, + 'grad_norm_pre_clip_avg': 0.18576188683509826, + 'learning_rate': 1.3392280436095144e-05, + 'epoch': 6.68} +04/19 [20:58:40] INFO | >> train_qwenlatent.py:487 + Step 26500 | grad_norm_pre_clip=0.2191 | + grad_norm_pre_clip_avg=0.1876 | Metrics: + {'align_loss': 0.027008123695850372, + 'recon_loss': 0.13147932291030884, + 'predict_loss': 0.014093711040914059, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21911408007144928, + 'mae_score': 0.010030142466227214, 'data_time': + 0.0006591109849978238, 'model_time': + 1.2100724700139835, 'grad_norm_pre_clip_avg': + 0.18758269250392914, 'learning_rate': + 1.3383583965879764e-05, 'epoch': 6.69} +04/19 [20:58:52] INFO | >> train_qwenlatent.py:487 + Step 26510 | grad_norm_pre_clip=0.2244 | + grad_norm_pre_clip_avg=0.2019 | Metrics: + {'align_loss': 0.025575384497642517, + 'recon_loss': 0.09852515906095505, + 'predict_loss': 0.010375481098890305, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2243647426366806, + 'data_time': 0.0006613679870497435, + 'model_time': 1.2053456459834706, + 'grad_norm_pre_clip_avg': 0.20188992023468016, + 'learning_rate': 1.3374887071108647e-05, + 'epoch': 6.69} +04/19 [20:59:05] INFO | >> train_qwenlatent.py:487 + Step 26520 | grad_norm_pre_clip=0.2398 | + grad_norm_pre_clip_avg=0.2008 | Metrics: + {'align_loss': 0.025059720501303673, + 'recon_loss': 0.10063836723566055, + 'predict_loss': 0.015019092708826065, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23976536095142365, + 'data_time': 0.0007015599985606968, + 'model_time': 1.2348573499766644, + 'grad_norm_pre_clip_avg': 0.20084385722875595, + 'learning_rate': 1.3366189756020539e-05, + 'epoch': 6.69} +04/19 [20:59:18] INFO | >> train_qwenlatent.py:487 + Step 26530 | grad_norm_pre_clip=0.1803 | + grad_norm_pre_clip_avg=0.2166 | Metrics: + {'align_loss': 0.02550356648862362, + 'recon_loss': 0.09114331007003784, + 'predict_loss': 0.008218150585889816, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18032141029834747, + 'data_time': 0.0008531629864592105, + 'model_time': 1.256907720991876, + 'grad_norm_pre_clip_avg': 0.21661302894353868, + 'learning_rate': 1.335749202485442e-05, + 'epoch': 6.69} +04/19 [20:59:30] INFO | >> train_qwenlatent.py:487 + Step 26540 | grad_norm_pre_clip=0.2389 | + grad_norm_pre_clip_avg=0.1834 | Metrics: + {'align_loss': 0.025059707462787628, + 'recon_loss': 0.07590087503194809, + 'predict_loss': 0.01166169811040163, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23886843025684357, + 'data_time': 0.0006387830071616918, + 'model_time': 1.214999029005412, + 'grad_norm_pre_clip_avg': 0.1834486946463585, + 'learning_rate': 1.3348793881849453e-05, + 'epoch': 6.7} +04/19 [20:59:44] INFO | >> train_qwenlatent.py:487 + Step 26550 | grad_norm_pre_clip=0.1923 | + grad_norm_pre_clip_avg=0.1731 | Metrics: + {'align_loss': 0.025111395865678787, + 'recon_loss': 0.08782690763473511, + 'predict_loss': 0.01128243375569582, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19228173792362213, + 'mae_score': 0.008690902778694222, 'data_time': + 0.0009524399938527495, 'model_time': + 1.2732700159831438, 'grad_norm_pre_clip_avg': + 0.1731177970767021, 'learning_rate': + 1.3340095331245006e-05, 'epoch': 6.7} +04/19 [20:59:57] INFO | >> train_qwenlatent.py:487 + Step 26560 | grad_norm_pre_clip=0.1586 | + grad_norm_pre_clip_avg=0.1735 | Metrics: + {'align_loss': 0.025152796879410744, + 'recon_loss': 0.07334186881780624, + 'predict_loss': 0.005633971653878689, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15861091017723083, + 'data_time': 0.0007104199903551489, + 'model_time': 1.243869055993855, + 'grad_norm_pre_clip_avg': 0.17353808134794235, + 'learning_rate': 1.3331396377280646e-05, + 'epoch': 6.7} +04/19 [21:00:10] INFO | >> train_qwenlatent.py:487 + Step 26570 | grad_norm_pre_clip=0.2036 | + grad_norm_pre_clip_avg=0.1670 | Metrics: + {'align_loss': 0.025096675381064415, + 'recon_loss': 0.08290690183639526, + 'predict_loss': 0.005810016766190529, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2035815715789795, + 'data_time': 0.001071564998710528, + 'model_time': 1.2871625639963895, + 'grad_norm_pre_clip_avg': 0.16699498817324637, + 'learning_rate': 1.332269702419614e-05, + 'epoch': 6.7} +04/19 [21:00:23] INFO | >> train_qwenlatent.py:487 + Step 26580 | grad_norm_pre_clip=0.1455 | + grad_norm_pre_clip_avg=0.1839 | Metrics: + {'align_loss': 0.02549263834953308, + 'recon_loss': 0.09096363186836243, + 'predict_loss': 0.005731431767344475, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1454554796218872, + 'data_time': 0.0009425239986740053, + 'model_time': 1.2340887339960318, + 'grad_norm_pre_clip_avg': 0.18390666097402572, + 'learning_rate': 1.3313997276231449e-05, + 'epoch': 6.71} +04/19 [21:00:35] INFO | >> train_qwenlatent.py:487 + Step 26590 | grad_norm_pre_clip=0.2931 | + grad_norm_pre_clip_avg=0.1774 | Metrics: + {'align_loss': 0.02417413890361786, + 'recon_loss': 0.06950169801712036, + 'predict_loss': 0.006387621164321899, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.29310208559036255, + 'data_time': 0.0006359050166793168, + 'model_time': 1.2578555979998782, + 'grad_norm_pre_clip_avg': 0.17740229964256288, + 'learning_rate': 1.3305297137626721e-05, + 'epoch': 6.71} +04/19 [21:00:49] INFO | >> train_qwenlatent.py:487 + Step 26600 | grad_norm_pre_clip=0.1835 | + grad_norm_pre_clip_avg=0.2123 | Metrics: + {'align_loss': 0.025872040539979935, + 'recon_loss': 0.09050866216421127, + 'predict_loss': 0.009083946235477924, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1835154891014099, + 'mae_score': 0.009493686916591885, 'data_time': + 0.0007496379839722067, 'model_time': + 1.245163119980134, 'grad_norm_pre_clip_avg': + 0.21227351874113082, 'learning_rate': + 1.3296596612622303e-05, 'epoch': 6.71} +04/19 [21:01:01] INFO | >> train_qwenlatent.py:487 + Step 26610 | grad_norm_pre_clip=0.1521 | + grad_norm_pre_clip_avg=0.1798 | Metrics: + {'align_loss': 0.025954710319638252, + 'recon_loss': 0.1209271177649498, + 'predict_loss': 0.011188116855919361, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15211911499500275, + 'data_time': 0.0011282229970674962, + 'model_time': 1.2226023219991475, + 'grad_norm_pre_clip_avg': 0.17975301295518875, + 'learning_rate': 1.3287895705458717e-05, + 'epoch': 6.71} +04/19 [21:01:13] INFO | >> train_qwenlatent.py:487 + Step 26620 | grad_norm_pre_clip=0.1555 | + grad_norm_pre_clip_avg=0.1534 | Metrics: + {'align_loss': 0.02580021694302559, + 'recon_loss': 0.07064089924097061, + 'predict_loss': 0.0046501560136675835, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.155475452542305, + 'data_time': 0.0009061669989023358, + 'model_time': 1.2553162229887675, + 'grad_norm_pre_clip_avg': 0.15337325632572174, + 'learning_rate': 1.3279194420376687e-05, + 'epoch': 6.72} +04/19 [21:01:26] INFO | >> train_qwenlatent.py:487 + Step 26630 | grad_norm_pre_clip=0.1878 | + grad_norm_pre_clip_avg=0.2025 | Metrics: + {'align_loss': 0.02696838229894638, + 'recon_loss': 0.11962302774190903, + 'predict_loss': 0.011910997331142426, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18780837953090668, + 'data_time': 0.0008836720080580562, + 'model_time': 1.2089583449997008, + 'grad_norm_pre_clip_avg': 0.2025172606110573, + 'learning_rate': 1.3270492761617107e-05, + 'epoch': 6.72} +04/19 [21:01:39] INFO | >> train_qwenlatent.py:487 + Step 26640 | grad_norm_pre_clip=0.1936 | + grad_norm_pre_clip_avg=0.2188 | Metrics: + {'align_loss': 0.025403516367077827, + 'recon_loss': 0.09078383445739746, + 'predict_loss': 0.009795929305255413, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1935717910528183, + 'data_time': 0.00083709298633039, 'model_time': + 1.19880255300086, 'grad_norm_pre_clip_avg': + 0.2187918782234192, 'learning_rate': + 1.3261790733421065e-05, 'epoch': 6.72} +04/19 [21:01:52] INFO | >> train_qwenlatent.py:487 + Step 26650 | grad_norm_pre_clip=0.1954 | + grad_norm_pre_clip_avg=0.1731 | Metrics: + {'align_loss': 0.025700516998767853, + 'recon_loss': 0.06844598054885864, + 'predict_loss': 0.007397784385830164, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19536645710468292, + 'mae_score': 0.007744594092841621, 'data_time': + 0.0006709629960823804, 'model_time': + 1.5065222759731114, 'grad_norm_pre_clip_avg': + 0.1730654388666153, 'learning_rate': + 1.3253088340029824e-05, 'epoch': 6.72} +04/19 [21:02:04] INFO | >> train_qwenlatent.py:487 + Step 26660 | grad_norm_pre_clip=0.1842 | + grad_norm_pre_clip_avg=0.1846 | Metrics: + {'align_loss': 0.025547923520207405, + 'recon_loss': 0.07129888981580734, + 'predict_loss': 0.007861298508942127, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1841956526041031, + 'data_time': 0.001008710009045899, + 'model_time': 1.2419702789920848, + 'grad_norm_pre_clip_avg': 0.18458325415849686, + 'learning_rate': 1.3244385585684815e-05, + 'epoch': 6.73} +04/19 [21:02:17] INFO | >> train_qwenlatent.py:487 + Step 26670 | grad_norm_pre_clip=0.1562 | + grad_norm_pre_clip_avg=0.1818 | Metrics: + {'align_loss': 0.025017613545060158, + 'recon_loss': 0.113540418446064, + 'predict_loss': 0.008716496638953686, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15621964633464813, + 'data_time': 0.0009657389891799539, + 'model_time': 1.272241083002882, + 'grad_norm_pre_clip_avg': 0.18179365247488022, + 'learning_rate': 1.3235682474627673e-05, + 'epoch': 6.73} +04/19 [21:02:30] INFO | >> train_qwenlatent.py:487 + Step 26680 | grad_norm_pre_clip=0.1989 | + grad_norm_pre_clip_avg=0.1662 | Metrics: + {'align_loss': 0.02636082097887993, + 'recon_loss': 0.11136621981859207, + 'predict_loss': 0.009848554618656635, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19890053570270538, + 'data_time': 0.000793187995441258, + 'model_time': 1.257819163001841, + 'grad_norm_pre_clip_avg': 0.1661960780620575, + 'learning_rate': 1.3226979011100173e-05, + 'epoch': 6.73} +04/19 [21:02:43] INFO | >> train_qwenlatent.py:487 + Step 26690 | grad_norm_pre_clip=0.1609 | + grad_norm_pre_clip_avg=0.1741 | Metrics: + {'align_loss': 0.025770843029022217, + 'recon_loss': 0.09971151500940323, + 'predict_loss': 0.01387442834675312, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16092990338802338, + 'data_time': 0.0011850229930132627, + 'model_time': 1.291575726994779, + 'grad_norm_pre_clip_avg': 0.17412485778331757, + 'learning_rate': 1.3218275199344285e-05, + 'epoch': 6.73} +04/19 [21:02:56] INFO | >> train_qwenlatent.py:487 + Step 26700 | grad_norm_pre_clip=0.1879 | + grad_norm_pre_clip_avg=0.1684 | Metrics: + {'align_loss': 0.025516942143440247, + 'recon_loss': 0.08778336644172668, + 'predict_loss': 0.012485580518841743, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1878717690706253, + 'mae_score': 0.00867994583404816, 'data_time': + 0.0007684050069656223, 'model_time': + 1.2192745069914963, 'grad_norm_pre_clip_avg': + 0.16839342340826988, 'learning_rate': + 1.3209571043602134e-05, 'epoch': 6.74} +04/19 [21:03:09] INFO | >> train_qwenlatent.py:487 + Step 26710 | grad_norm_pre_clip=0.2540 | + grad_norm_pre_clip_avg=0.2145 | Metrics: + {'align_loss': 0.025166833773255348, + 'recon_loss': 0.09391237795352936, + 'predict_loss': 0.010346543043851852, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2539980113506317, + 'data_time': 0.0007609989843331277, + 'model_time': 1.2332406099885702, + 'grad_norm_pre_clip_avg': 0.21445116698741912, + 'learning_rate': 1.3200866548116028e-05, + 'epoch': 6.74} +04/19 [21:03:21] INFO | >> train_qwenlatent.py:487 + Step 26720 | grad_norm_pre_clip=0.1721 | + grad_norm_pre_clip_avg=0.1717 | Metrics: + {'align_loss': 0.024916158989071846, + 'recon_loss': 0.07558279484510422, + 'predict_loss': 0.011850563809275627, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17211444675922394, + 'data_time': 0.0009922919853124768, + 'model_time': 1.2316680160001852, + 'grad_norm_pre_clip_avg': 0.17166679501533508, + 'learning_rate': 1.3192161717128426e-05, + 'epoch': 6.74} +04/19 [21:03:34] INFO | >> train_qwenlatent.py:487 + Step 26730 | grad_norm_pre_clip=0.1710 | + grad_norm_pre_clip_avg=0.1661 | Metrics: + {'align_loss': 0.025464903563261032, + 'recon_loss': 0.07853781431913376, + 'predict_loss': 0.0071449666284024715, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17099186778068542, + 'data_time': 0.001193901989609003, + 'model_time': 1.3369330000132322, + 'grad_norm_pre_clip_avg': 0.16613947749137878, + 'learning_rate': 1.3183456554881963e-05, + 'epoch': 6.74} +04/19 [21:03:47] INFO | >> train_qwenlatent.py:487 + Step 26740 | grad_norm_pre_clip=0.1986 | + grad_norm_pre_clip_avg=0.1787 | Metrics: + {'align_loss': 0.025557126849889755, + 'recon_loss': 0.11237642168998718, + 'predict_loss': 0.011116385459899902, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19862140715122223, + 'data_time': 0.0006395459931809455, + 'model_time': 1.2116975309909321, + 'grad_norm_pre_clip_avg': 0.1787111297249794, + 'learning_rate': 1.3174751065619423e-05, + 'epoch': 6.75} +04/19 [21:04:00] INFO | >> train_qwenlatent.py:487 + Step 26750 | grad_norm_pre_clip=0.2088 | + grad_norm_pre_clip_avg=0.1742 | Metrics: + {'align_loss': 0.02546260878443718, + 'recon_loss': 0.10597522556781769, + 'predict_loss': 0.0101624121889472, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20877060294151306, + 'mae_score': 0.010591183911572705, 'data_time': + 0.000901363993762061, 'model_time': + 1.236087786994176, 'grad_norm_pre_clip_avg': + 0.1741789162158966, 'learning_rate': + 1.3166045253583764e-05, 'epoch': 6.75} +04/19 [21:04:12] INFO | >> train_qwenlatent.py:487 + Step 26760 | grad_norm_pre_clip=0.1712 | + grad_norm_pre_clip_avg=0.1789 | Metrics: + {'align_loss': 0.0258963443338871, + 'recon_loss': 0.0980008989572525, + 'predict_loss': 0.010840517468750477, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17122046649456024, + 'data_time': 0.0010020739864557981, + 'model_time': 1.2365589429973625, + 'grad_norm_pre_clip_avg': 0.17885688841342925, + 'learning_rate': 1.3157339123018082e-05, + 'epoch': 6.75} +04/19 [21:04:25] INFO | >> train_qwenlatent.py:487 + Step 26770 | grad_norm_pre_clip=0.2145 | + grad_norm_pre_clip_avg=0.1834 | Metrics: + {'align_loss': 0.026588499546051025, + 'recon_loss': 0.11949468404054642, + 'predict_loss': 0.01784316636621952, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21449702978134155, + 'data_time': 0.0007583050173707306, + 'model_time': 1.195311983989086, + 'grad_norm_pre_clip_avg': 0.1834449976682663, + 'learning_rate': 1.3148632678165649e-05, + 'epoch': 6.75} +04/19 [21:04:38] INFO | >> train_qwenlatent.py:487 + Step 26780 | grad_norm_pre_clip=0.1898 | + grad_norm_pre_clip_avg=0.2068 | Metrics: + {'align_loss': 0.02487747184932232, + 'recon_loss': 0.09379268437623978, + 'predict_loss': 0.009527824819087982, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1898117959499359, + 'data_time': 0.0010000909969676286, + 'model_time': 1.2614374970144127, + 'grad_norm_pre_clip_avg': 0.20681634098291396, + 'learning_rate': 1.3139925923269874e-05, + 'epoch': 6.76} +04/19 [21:04:50] INFO | >> train_qwenlatent.py:487 + Step 26790 | grad_norm_pre_clip=0.1570 | + grad_norm_pre_clip_avg=0.1810 | Metrics: + {'align_loss': 0.023703766986727715, + 'recon_loss': 0.097259022295475, + 'predict_loss': 0.013517411425709724, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15700508654117584, + 'data_time': 0.0009491390082985163, + 'model_time': 1.280063959013205, + 'grad_norm_pre_clip_avg': 0.1810361161828041, + 'learning_rate': 1.3131218862574326e-05, + 'epoch': 6.76} +04/19 [21:05:03] INFO | >> train_qwenlatent.py:487 + Step 26800 | grad_norm_pre_clip=0.2632 | + grad_norm_pre_clip_avg=0.2120 | Metrics: + {'align_loss': 0.025272244587540627, + 'recon_loss': 0.07181254029273987, + 'predict_loss': 0.007398241199553013, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2632092535495758, + 'mae_score': 0.00817057549416482, 'data_time': + 0.0011214870028197765, 'model_time': + 1.2925974180107005, 'grad_norm_pre_clip_avg': + 0.21204133331775665, 'learning_rate': + 1.3122511500322723e-05, 'epoch': 6.76} +04/19 [21:05:16] INFO | >> train_qwenlatent.py:487 + Step 26810 | grad_norm_pre_clip=0.1418 | + grad_norm_pre_clip_avg=0.1794 | Metrics: + {'align_loss': 0.02505405992269516, + 'recon_loss': 0.12696228921413422, + 'predict_loss': 0.012717947363853455, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14182138442993164, + 'data_time': 0.0007312349916901439, + 'model_time': 1.1954381940013263, + 'grad_norm_pre_clip_avg': 0.17936396300792695, + 'learning_rate': 1.3113803840758923e-05, + 'epoch': 6.77} +04/19 [21:05:29] INFO | >> train_qwenlatent.py:487 + Step 26820 | grad_norm_pre_clip=0.1786 | + grad_norm_pre_clip_avg=0.1732 | Metrics: + {'align_loss': 0.024383019655942917, + 'recon_loss': 0.06475421786308289, + 'predict_loss': 0.007679579313844442, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17856456339359283, + 'data_time': 0.0007262289873324335, + 'model_time': 1.2275595440005418, + 'grad_norm_pre_clip_avg': 0.17324094772338866, + 'learning_rate': 1.310509588812694e-05, + 'epoch': 6.77} +04/19 [21:05:42] INFO | >> train_qwenlatent.py:487 + Step 26830 | grad_norm_pre_clip=0.1794 | + grad_norm_pre_clip_avg=0.1698 | Metrics: + {'align_loss': 0.0248250812292099, + 'recon_loss': 0.10841885954141617, + 'predict_loss': 0.012844869866967201, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1793535202741623, + 'data_time': 0.0006941229803487659, + 'model_time': 1.1973448489734437, + 'grad_norm_pre_clip_avg': 0.16977425664663315, + 'learning_rate': 1.3096387646670914e-05, + 'epoch': 6.77} +04/19 [21:05:54] INFO | >> train_qwenlatent.py:487 + Step 26840 | grad_norm_pre_clip=0.2005 | + grad_norm_pre_clip_avg=0.1789 | Metrics: + {'align_loss': 0.02410193905234337, + 'recon_loss': 0.09129597991704941, + 'predict_loss': 0.010266460478305817, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20051828026771545, + 'data_time': 0.0009441149886697531, + 'model_time': 1.2255950160033535, + 'grad_norm_pre_clip_avg': 0.17886748909950256, + 'learning_rate': 1.3087679120635143e-05, + 'epoch': 6.77} +04/19 [21:06:07] INFO | >> train_qwenlatent.py:487 + Step 26850 | grad_norm_pre_clip=0.2092 | + grad_norm_pre_clip_avg=0.2108 | Metrics: + {'align_loss': 0.025157369673252106, + 'recon_loss': 0.08012349903583527, + 'predict_loss': 0.010193673893809319, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2092253416776657, + 'mae_score': 0.017483326336285014, 'data_time': + 0.000703628989867866, 'model_time': + 1.2260249969840515, 'grad_norm_pre_clip_avg': + 0.21081551909446716, 'learning_rate': + 1.3078970314264056e-05, 'epoch': 6.78} +04/19 [21:06:20] INFO | >> train_qwenlatent.py:487 + Step 26860 | grad_norm_pre_clip=0.1866 | + grad_norm_pre_clip_avg=0.1793 | Metrics: + {'align_loss': 0.025849681347608566, + 'recon_loss': 0.12375637143850327, + 'predict_loss': 0.00904531218111515, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18656864762306213, + 'data_time': 0.0009427779878024012, + 'model_time': 1.245515960996272, + 'grad_norm_pre_clip_avg': 0.1792832225561142, + 'learning_rate': 1.3070261231802225e-05, + 'epoch': 6.78} +04/19 [21:06:32] INFO | >> train_qwenlatent.py:487 + Step 26870 | grad_norm_pre_clip=0.1781 | + grad_norm_pre_clip_avg=0.1723 | Metrics: + {'align_loss': 0.02588685229420662, + 'recon_loss': 0.12125956267118454, + 'predict_loss': 0.012772636488080025, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17812936007976532, + 'data_time': 0.0006855189858470112, + 'model_time': 1.194832499983022, + 'grad_norm_pre_clip_avg': 0.17234066873788834, + 'learning_rate': 1.3061551877494337e-05, + 'epoch': 6.78} +04/19 [21:06:45] INFO | >> train_qwenlatent.py:487 + Step 26880 | grad_norm_pre_clip=0.2316 | + grad_norm_pre_clip_avg=0.1963 | Metrics: + {'align_loss': 0.025479305535554886, + 'recon_loss': 0.09852156043052673, + 'predict_loss': 0.01345918606966734, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23161257803440094, + 'data_time': 0.0009999600006267428, + 'model_time': 1.2209778469987214, + 'grad_norm_pre_clip_avg': 0.19633424729108812, + 'learning_rate': 1.305284225558524e-05, + 'epoch': 6.78} +04/19 [21:06:57] INFO | >> train_qwenlatent.py:487 + Step 26890 | grad_norm_pre_clip=0.1840 | + grad_norm_pre_clip_avg=0.2002 | Metrics: + {'align_loss': 0.023632314056158066, + 'recon_loss': 0.07533041387796402, + 'predict_loss': 0.006961836013942957, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18403328955173492, + 'data_time': 0.0006451770022977144, + 'model_time': 1.1773735939932521, + 'grad_norm_pre_clip_avg': 0.2002237930893898, + 'learning_rate': 1.3044132370319893e-05, + 'epoch': 6.79} +04/19 [21:07:09] INFO | >> train_qwenlatent.py:487 + Step 26900 | grad_norm_pre_clip=0.1626 | + grad_norm_pre_clip_avg=0.1816 | Metrics: + {'align_loss': 0.026608115062117577, + 'recon_loss': 0.09866446256637573, + 'predict_loss': 0.008400831371545792, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16257324814796448, + 'mae_score': 0.011710926433941265, 'data_time': + 0.0006651559961028397, 'model_time': + 1.147984582989011, 'grad_norm_pre_clip_avg': + 0.18158816248178483, 'learning_rate': + 1.303542222594338e-05, 'epoch': 6.79} +04/19 [21:07:21] INFO | >> train_qwenlatent.py:487 + Step 26910 | grad_norm_pre_clip=0.1570 | + grad_norm_pre_clip_avg=0.1836 | Metrics: + {'align_loss': 0.024485912173986435, + 'recon_loss': 0.08462689816951752, + 'predict_loss': 0.008015521802008152, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15699847042560577, + 'data_time': 0.000675418006721884, + 'model_time': 1.180056926998077, + 'grad_norm_pre_clip_avg': 0.18357027173042298, + 'learning_rate': 1.3026711826700935e-05, + 'epoch': 6.79} +04/19 [21:07:33] INFO | >> train_qwenlatent.py:487 + Step 26920 | grad_norm_pre_clip=0.2030 | + grad_norm_pre_clip_avg=0.1929 | Metrics: + {'align_loss': 0.026593372225761414, + 'recon_loss': 0.10785163938999176, + 'predict_loss': 0.010244056582450867, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20302656292915344, + 'data_time': 0.0005936139787081629, + 'model_time': 1.15281674801372, + 'grad_norm_pre_clip_avg': 0.19289113730192184, + 'learning_rate': 1.3018001176837885e-05, + 'epoch': 6.79} +04/19 [21:07:45] INFO | >> train_qwenlatent.py:487 + Step 26930 | grad_norm_pre_clip=0.2035 | + grad_norm_pre_clip_avg=0.1947 | Metrics: + {'align_loss': 0.02572864294052124, + 'recon_loss': 0.08714352548122406, + 'predict_loss': 0.006568307057023048, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20353522896766663, + 'data_time': 0.0006488980143330991, + 'model_time': 1.1620904210139997, + 'grad_norm_pre_clip_avg': 0.1947462797164917, + 'learning_rate': 1.3009290280599707e-05, + 'epoch': 6.8} +04/19 [21:07:56] INFO | >> train_qwenlatent.py:487 + Step 26940 | grad_norm_pre_clip=0.2142 | + grad_norm_pre_clip_avg=0.1780 | Metrics: + {'align_loss': 0.025110196322202682, + 'recon_loss': 0.08641538769006729, + 'predict_loss': 0.008961821906268597, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21421831846237183, + 'data_time': 0.0006024290050845593, + 'model_time': 1.1695858140010387, + 'grad_norm_pre_clip_avg': 0.1780087396502495, + 'learning_rate': 1.3000579142231978e-05, + 'epoch': 6.8} +04/19 [21:08:08] INFO | >> train_qwenlatent.py:487 + Step 26950 | grad_norm_pre_clip=0.1951 | + grad_norm_pre_clip_avg=0.1849 | Metrics: + {'align_loss': 0.025393683463335037, + 'recon_loss': 0.08035009354352951, + 'predict_loss': 0.009702182374894619, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19513699412345886, + 'mae_score': 0.018616760528839385, 'data_time': + 0.0008000399975571781, 'model_time': + 1.1786135409784038, 'grad_norm_pre_clip_avg': + 0.18490975499153137, 'learning_rate': + 1.2991867765980407e-05, 'epoch': 6.8} +04/19 [21:08:21] INFO | >> train_qwenlatent.py:487 + Step 26960 | grad_norm_pre_clip=0.2002 | + grad_norm_pre_clip_avg=0.1948 | Metrics: + {'align_loss': 0.02427254244685173, + 'recon_loss': 0.07456841319799423, + 'predict_loss': 0.010369792580604553, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20020493865013123, + 'data_time': 0.0006397429970093071, + 'model_time': 1.1593830049969256, + 'grad_norm_pre_clip_avg': 0.1948406755924225, + 'learning_rate': 1.2983156156090812e-05, + 'epoch': 6.8} +04/19 [21:08:32] INFO | >> train_qwenlatent.py:487 + Step 26970 | grad_norm_pre_clip=0.1746 | + grad_norm_pre_clip_avg=0.1816 | Metrics: + {'align_loss': 0.024513687938451767, + 'recon_loss': 0.10502602905035019, + 'predict_loss': 0.008792600594460964, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17461806535720825, + 'data_time': 0.0008581000147387385, + 'model_time': 1.1586979500134476, + 'grad_norm_pre_clip_avg': 0.18163898289203645, + 'learning_rate': 1.2974444316809117e-05, + 'epoch': 6.81} +04/19 [21:08:44] INFO | >> train_qwenlatent.py:487 + Step 26980 | grad_norm_pre_clip=0.2057 | + grad_norm_pre_clip_avg=0.1844 | Metrics: + {'align_loss': 0.024989325553178787, + 'recon_loss': 0.13292132318019867, + 'predict_loss': 0.013333788141608238, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20570160448551178, + 'data_time': 0.0006002970039844513, + 'model_time': 1.1364469429827295, + 'grad_norm_pre_clip_avg': 0.1843850076198578, + 'learning_rate': 1.2965732252381379e-05, + 'epoch': 6.81} +04/19 [21:08:55] INFO | >> train_qwenlatent.py:487 + Step 26990 | grad_norm_pre_clip=0.2270 | + grad_norm_pre_clip_avg=0.1969 | Metrics: + {'align_loss': 0.025344550609588623, + 'recon_loss': 0.09637285023927689, + 'predict_loss': 0.008808491751551628, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22703678905963898, + 'data_time': 0.0006089119997341186, + 'model_time': 1.165255429019453, + 'grad_norm_pre_clip_avg': 0.1968762010335922, + 'learning_rate': 1.2957019967053747e-05, + 'epoch': 6.81} +04/19 [21:09:08] INFO | >> train_qwenlatent.py:487 + Step 27000 | grad_norm_pre_clip=0.1760 | + grad_norm_pre_clip_avg=0.1766 | Metrics: + {'align_loss': 0.02525741420686245, + 'recon_loss': 0.0688251331448555, + 'predict_loss': 0.008726590313017368, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17604489624500275, + 'mae_score': 0.01050193975637625, 'data_time': + 0.00068741399445571, 'model_time': + 1.1559868789918255, 'grad_norm_pre_clip_avg': + 0.17658910751342774, 'learning_rate': + 1.2948307465072484e-05, 'epoch': 6.81} +04/19 [21:09:19] INFO | >> train_qwenlatent.py:487 + Step 27010 | grad_norm_pre_clip=0.1483 | + grad_norm_pre_clip_avg=0.1868 | Metrics: + {'align_loss': 0.02420918270945549, + 'recon_loss': 0.08394954353570938, + 'predict_loss': 0.01179460622370243, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1483117640018463, + 'data_time': 0.0005953630025032908, + 'model_time': 1.1614018919935916, + 'grad_norm_pre_clip_avg': 0.1868148222565651, + 'learning_rate': 1.2939594750683957e-05, + 'epoch': 6.82} +04/19 [21:09:31] INFO | >> train_qwenlatent.py:487 + Step 27020 | grad_norm_pre_clip=0.2469 | + grad_norm_pre_clip_avg=0.1946 | Metrics: + {'align_loss': 0.025913875550031662, + 'recon_loss': 0.08194828778505325, + 'predict_loss': 0.010875236243009567, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2468711882829666, + 'data_time': 0.000661255995510146, + 'model_time': 1.1558505780121777, + 'grad_norm_pre_clip_avg': 0.19460795670747758, + 'learning_rate': 1.293088182813463e-05, + 'epoch': 6.82} +04/19 [21:09:42] INFO | >> train_qwenlatent.py:487 + Step 27030 | grad_norm_pre_clip=0.1496 | + grad_norm_pre_clip_avg=0.1865 | Metrics: + {'align_loss': 0.026307707652449608, + 'recon_loss': 0.10123738646507263, + 'predict_loss': 0.009887353517115116, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14956459403038025, + 'data_time': 0.000739009992685169, + 'model_time': 1.160571185988374, + 'grad_norm_pre_clip_avg': 0.1865466997027397, + 'learning_rate': 1.292216870167109e-05, + 'epoch': 6.82} +04/19 [21:09:54] INFO | >> train_qwenlatent.py:487 + Step 27040 | grad_norm_pre_clip=0.1551 | + grad_norm_pre_clip_avg=0.1939 | Metrics: + {'align_loss': 0.02518545836210251, + 'recon_loss': 0.09547121077775955, + 'predict_loss': 0.007913782261312008, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15508776903152466, + 'data_time': 0.0006050700030755252, + 'model_time': 1.149849906010786, + 'grad_norm_pre_clip_avg': 0.193850602209568, + 'learning_rate': 1.2913455375539998e-05, + 'epoch': 6.82} +04/19 [21:10:06] INFO | >> train_qwenlatent.py:487 + Step 27050 | grad_norm_pre_clip=0.1615 | + grad_norm_pre_clip_avg=0.1705 | Metrics: + {'align_loss': 0.02437647618353367, + 'recon_loss': 0.0939478874206543, + 'predict_loss': 0.006945707835257053, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16154390573501587, + 'mae_score': 0.008525945474435618, 'data_time': + 0.0006166679959278554, 'model_time': + 1.1814282340055797, 'grad_norm_pre_clip_avg': + 0.17052869349718094, 'learning_rate': + 1.2904741853988129e-05, 'epoch': 6.83} +04/19 [21:10:18] INFO | >> train_qwenlatent.py:487 + Step 27060 | grad_norm_pre_clip=0.1914 | + grad_norm_pre_clip_avg=0.1594 | Metrics: + {'align_loss': 0.024249490350484848, + 'recon_loss': 0.06695582717657089, + 'predict_loss': 0.010322162881493568, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19143326580524445, + 'data_time': 0.0008694269927218556, + 'model_time': 1.13627517799614, + 'grad_norm_pre_clip_avg': 0.15938532799482347, + 'learning_rate': 1.2896028141262346e-05, + 'epoch': 6.83} +04/19 [21:10:29] INFO | >> train_qwenlatent.py:487 + Step 27070 | grad_norm_pre_clip=0.1640 | + grad_norm_pre_clip_avg=0.1777 | Metrics: + {'align_loss': 0.025514379143714905, + 'recon_loss': 0.07966446131467819, + 'predict_loss': 0.00528089189901948, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16397403180599213, + 'data_time': 0.0005993710074108094, + 'model_time': 1.1932900289830286, + 'grad_norm_pre_clip_avg': 0.17767233401536942, + 'learning_rate': 1.28873142416096e-05, 'epoch': + 6.83} +04/19 [21:10:41] INFO | >> train_qwenlatent.py:487 + Step 27080 | grad_norm_pre_clip=0.1886 | + grad_norm_pre_clip_avg=0.1984 | Metrics: + {'align_loss': 0.02477206662297249, + 'recon_loss': 0.07460460811853409, + 'predict_loss': 0.007798749953508377, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18861214816570282, + 'data_time': 0.000577144994167611, + 'model_time': 1.159646840998903, + 'grad_norm_pre_clip_avg': 0.19843314439058304, + 'learning_rate': 1.2878600159276955e-05, + 'epoch': 6.83} +04/19 [21:10:53] INFO | >> train_qwenlatent.py:487 + Step 27090 | grad_norm_pre_clip=0.2281 | + grad_norm_pre_clip_avg=0.1959 | Metrics: + {'align_loss': 0.02537946216762066, + 'recon_loss': 0.09568849951028824, + 'predict_loss': 0.007289567030966282, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22806599736213684, + 'data_time': 0.0006247599958442152, + 'model_time': 1.140006263012765, + 'grad_norm_pre_clip_avg': 0.1959209680557251, + 'learning_rate': 1.2869885898511533e-05, + 'epoch': 6.84} +04/19 [21:11:04] INFO | >> train_qwenlatent.py:487 + Step 27100 | grad_norm_pre_clip=0.1921 | + grad_norm_pre_clip_avg=0.2136 | Metrics: + {'align_loss': 0.025882069021463394, + 'recon_loss': 0.10627681016921997, + 'predict_loss': 0.010983371175825596, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19207648932933807, + 'mae_score': 0.01096269160777599, 'data_time': + 0.0006686770066153258, 'model_time': + 1.1640514640021138, 'grad_norm_pre_clip_avg': + 0.21362376660108567, 'learning_rate': + 1.2861171463560567e-05, 'epoch': 6.84} +04/19 [21:11:17] INFO | >> train_qwenlatent.py:487 + Step 27110 | grad_norm_pre_clip=0.2701 | + grad_norm_pre_clip_avg=0.2216 | Metrics: + {'align_loss': 0.024301931262016296, + 'recon_loss': 0.11255751550197601, + 'predict_loss': 0.011963911354541779, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.27012670040130615, + 'data_time': 0.0006372549978550524, + 'model_time': 1.1567761929763947, + 'grad_norm_pre_clip_avg': 0.22164222300052644, + 'learning_rate': 1.2852456858671364e-05, + 'epoch': 6.84} +04/19 [21:11:28] INFO | >> train_qwenlatent.py:487 + Step 27120 | grad_norm_pre_clip=0.1909 | + grad_norm_pre_clip_avg=0.1883 | Metrics: + {'align_loss': 0.02626076154410839, + 'recon_loss': 0.12089192122220993, + 'predict_loss': 0.009562655352056026, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19086749851703644, + 'data_time': 0.0005357430200092494, + 'model_time': 1.1542968330031727, + 'grad_norm_pre_clip_avg': 0.18826334178447723, + 'learning_rate': 1.284374208809132e-05, + 'epoch': 6.84} +04/19 [21:11:40] INFO | >> train_qwenlatent.py:487 + Step 27130 | grad_norm_pre_clip=0.1752 | + grad_norm_pre_clip_avg=0.1644 | Metrics: + {'align_loss': 0.02629537135362625, + 'recon_loss': 0.08101362735033035, + 'predict_loss': 0.007774436380714178, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1752070188522339, + 'data_time': 0.0005867969885002822, + 'model_time': 1.148885436996352, + 'grad_norm_pre_clip_avg': 0.16442583352327347, + 'learning_rate': 1.2835027156067911e-05, + 'epoch': 6.85} +04/19 [21:11:51] INFO | >> train_qwenlatent.py:487 + Step 27140 | grad_norm_pre_clip=0.1736 | + grad_norm_pre_clip_avg=0.1634 | Metrics: + {'align_loss': 0.02592264674603939, + 'recon_loss': 0.10776600986719131, + 'predict_loss': 0.008547388948500156, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17358238995075226, + 'data_time': 0.0006090120004955679, + 'model_time': 1.1503436249913648, + 'grad_norm_pre_clip_avg': 0.16336828619241714, + 'learning_rate': 1.2826312066848678e-05, + 'epoch': 6.85} +04/19 [21:12:03] INFO | >> train_qwenlatent.py:487 + Step 27150 | grad_norm_pre_clip=0.2029 | + grad_norm_pre_clip_avg=0.2117 | Metrics: + {'align_loss': 0.024650592356920242, + 'recon_loss': 0.07643097639083862, + 'predict_loss': 0.010797456838190556, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20290715992450714, + 'mae_score': 0.01085744892154728, 'data_time': + 0.000595979014178738, 'model_time': + 1.1505073549924418, 'grad_norm_pre_clip_avg': + 0.2117357984185219, 'learning_rate': + 1.2817596824681257e-05, 'epoch': 6.85} +04/19 [21:12:15] INFO | >> train_qwenlatent.py:487 + Step 27160 | grad_norm_pre_clip=0.1755 | + grad_norm_pre_clip_avg=0.1882 | Metrics: + {'align_loss': 0.026564233005046844, + 'recon_loss': 0.08782872557640076, + 'predict_loss': 0.01590650901198387, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17549477517604828, + 'data_time': 0.0005333739973139018, + 'model_time': 1.146437782997964, + 'grad_norm_pre_clip_avg': 0.18822329342365265, + 'learning_rate': 1.280888143381335e-05, + 'epoch': 6.85} +04/19 [21:12:26] INFO | >> train_qwenlatent.py:487 + Step 27170 | grad_norm_pre_clip=0.1394 | + grad_norm_pre_clip_avg=0.1692 | Metrics: + {'align_loss': 0.025042790919542313, + 'recon_loss': 0.08124322444200516, + 'predict_loss': 0.010027840733528137, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1394183188676834, + 'data_time': 0.0005877060175407678, + 'model_time': 1.128911881998647, + 'grad_norm_pre_clip_avg': 0.16916323453187943, + 'learning_rate': 1.2800165898492736e-05, + 'epoch': 6.86} +04/19 [21:12:38] INFO | >> train_qwenlatent.py:487 + Step 27180 | grad_norm_pre_clip=0.2322 | + grad_norm_pre_clip_avg=0.1932 | Metrics: + {'align_loss': 0.02694789692759514, + 'recon_loss': 0.10186266154050827, + 'predict_loss': 0.012505658902227879, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2322034388780594, + 'data_time': 0.0007197869999799877, + 'model_time': 1.1497833589965012, + 'grad_norm_pre_clip_avg': 0.19319090545177459, + 'learning_rate': 1.2791450222967252e-05, + 'epoch': 6.86} +04/19 [21:12:50] INFO | >> train_qwenlatent.py:487 + Step 27190 | grad_norm_pre_clip=0.1724 | + grad_norm_pre_clip_avg=0.1644 | Metrics: + {'align_loss': 0.026660598814487457, + 'recon_loss': 0.09436869621276855, + 'predict_loss': 0.005950771737843752, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1724400669336319, + 'data_time': 0.0005811990122310817, + 'model_time': 1.3153527949762065, + 'grad_norm_pre_clip_avg': 0.16436330303549768, + 'learning_rate': 1.2782734411484823e-05, + 'epoch': 6.86} +04/19 [21:13:34] INFO | >> train_qwenlatent.py:487 + Step 27200 | grad_norm_pre_clip=0.1851 | + grad_norm_pre_clip_avg=0.1629 | Metrics: + {'align_loss': 0.025876272469758987, + 'recon_loss': 0.09091326594352722, + 'predict_loss': 0.0051862020045518875, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18512064218521118, + 'mae_score': 0.010260887833328935, 'data_time': + 0.002939308003988117, 'model_time': + 3.848288275010418, 'grad_norm_pre_clip_avg': + 0.16291287541389465, 'learning_rate': + 1.2774018468293423e-05, 'epoch': 6.86} +04/19 [21:14:11] INFO | >> train_qwenlatent.py:487 + Step 27210 | grad_norm_pre_clip=0.1845 | + grad_norm_pre_clip_avg=0.2105 | Metrics: + {'align_loss': 0.02434806153178215, + 'recon_loss': 0.08478624373674393, + 'predict_loss': 0.01099112257361412, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18447495996952057, + 'data_time': 0.0010141919774468988, + 'model_time': 3.5889069859986193, + 'grad_norm_pre_clip_avg': 0.21052656173706055, + 'learning_rate': 1.2765302397641096e-05, + 'epoch': 6.87} +04/19 [21:14:46] INFO | >> train_qwenlatent.py:487 + Step 27220 | grad_norm_pre_clip=0.2558 | + grad_norm_pre_clip_avg=0.2195 | Metrics: + {'align_loss': 0.025647461414337158, + 'recon_loss': 0.09429017454385757, + 'predict_loss': 0.0067070601508021355, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2557523846626282, + 'data_time': 0.0011987899779342115, + 'model_time': 3.139925694995327, + 'grad_norm_pre_clip_avg': 0.21945018768310548, + 'learning_rate': 1.2756586203775958e-05, + 'epoch': 6.87} +04/19 [21:15:22] INFO | >> train_qwenlatent.py:487 + Step 27230 | grad_norm_pre_clip=0.1761 | + grad_norm_pre_clip_avg=0.1941 | Metrics: + {'align_loss': 0.025645045563578606, + 'recon_loss': 0.07466445118188858, + 'predict_loss': 0.01255299337208271, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1761401742696762, + 'data_time': 0.0012231770087964833, + 'model_time': 4.241420877020573, + 'grad_norm_pre_clip_avg': 0.1941269353032112, + 'learning_rate': 1.2747869890946163e-05, + 'epoch': 6.87} +04/19 [21:15:59] INFO | >> train_qwenlatent.py:487 + Step 27240 | grad_norm_pre_clip=0.1827 | + grad_norm_pre_clip_avg=0.1860 | Metrics: + {'align_loss': 0.025197353214025497, + 'recon_loss': 0.0644831657409668, + 'predict_loss': 0.00839044339954853, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18266448378562927, + 'data_time': 0.00117917699390091, 'model_time': + 3.2582363040128257, 'grad_norm_pre_clip_avg': + 0.18595333993434907, 'learning_rate': + 1.2739153463399941e-05, 'epoch': 6.87} +04/19 [21:16:34] INFO | >> train_qwenlatent.py:487 + Step 27250 | grad_norm_pre_clip=0.2311 | + grad_norm_pre_clip_avg=0.1832 | Metrics: + {'align_loss': 0.025757798925042152, + 'recon_loss': 0.08524069935083389, + 'predict_loss': 0.010036050342023373, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23114481568336487, + 'mae_score': 0.009876239192378414, 'data_time': + 0.007367602025624365, 'model_time': + 3.4995315929991193, 'grad_norm_pre_clip_avg': + 0.1831527180969715, 'learning_rate': + 1.273043692538558e-05, 'epoch': 6.88} +04/19 [21:17:07] INFO | >> train_qwenlatent.py:487 + Step 27260 | grad_norm_pre_clip=0.1740 | + grad_norm_pre_clip_avg=0.1792 | Metrics: + {'align_loss': 0.025490254163742065, + 'recon_loss': 0.0944850891828537, + 'predict_loss': 0.010118819773197174, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17400500178337097, + 'data_time': 0.014493430004222319, + 'model_time': 3.550827243976528, + 'grad_norm_pre_clip_avg': 0.17920294255018235, + 'learning_rate': 1.2721720281151408e-05, + 'epoch': 6.88} +04/19 [21:17:32] INFO | >> train_qwenlatent.py:487 + Step 27270 | grad_norm_pre_clip=0.1834 | + grad_norm_pre_clip_avg=0.1861 | Metrics: + {'align_loss': 0.025223959237337112, + 'recon_loss': 0.10257439315319061, + 'predict_loss': 0.011219317093491554, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18342766165733337, + 'data_time': 0.001100964000215754, + 'model_time': 2.70600130801904, + 'grad_norm_pre_clip_avg': 0.18613312244415284, + 'learning_rate': 1.2713003534945811e-05, + 'epoch': 6.88} +04/19 [21:17:53] INFO | >> train_qwenlatent.py:487 + Step 27280 | grad_norm_pre_clip=0.1783 | + grad_norm_pre_clip_avg=0.2162 | Metrics: + {'align_loss': 0.024125343188643456, + 'recon_loss': 0.08765289187431335, + 'predict_loss': 0.006935246754437685, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17833773791790009, + 'data_time': 0.0012202049838379025, + 'model_time': 1.5590209479851183, + 'grad_norm_pre_clip_avg': 0.21618389189243317, + 'learning_rate': 1.2704286691017225e-05, + 'epoch': 6.88} +04/19 [21:18:07] INFO | >> train_qwenlatent.py:487 + Step 27290 | grad_norm_pre_clip=0.1667 | + grad_norm_pre_clip_avg=0.1726 | Metrics: + {'align_loss': 0.02525075152516365, + 'recon_loss': 0.07599315047264099, + 'predict_loss': 0.009986194781959057, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16673918068408966, + 'data_time': 0.0007932409935165197, + 'model_time': 1.2534980759955943, + 'grad_norm_pre_clip_avg': 0.17260190546512605, + 'learning_rate': 1.2695569753614139e-05, + 'epoch': 6.89} +04/19 [21:18:20] INFO | >> train_qwenlatent.py:487 + Step 27300 | grad_norm_pre_clip=0.1782 | + grad_norm_pre_clip_avg=0.1708 | Metrics: + {'align_loss': 0.02455911785364151, + 'recon_loss': 0.06667973101139069, + 'predict_loss': 0.00596279464662075, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17815721035003662, + 'mae_score': 0.011791909707559122, 'data_time': + 0.000730642001144588, 'model_time': + 1.2380100300069898, 'grad_norm_pre_clip_avg': + 0.1708383709192276, 'learning_rate': + 1.268685272698508e-05, 'epoch': 6.89} +04/19 [21:18:33] INFO | >> train_qwenlatent.py:487 + Step 27310 | grad_norm_pre_clip=0.2550 | + grad_norm_pre_clip_avg=0.2132 | Metrics: + {'align_loss': 0.024838311597704887, + 'recon_loss': 0.05762519687414169, + 'predict_loss': 0.005703737493604422, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2550102174282074, + 'data_time': 0.0008899169915821403, + 'model_time': 1.2362838470144197, + 'grad_norm_pre_clip_avg': 0.2132408767938614, + 'learning_rate': 1.2678135615378617e-05, + 'epoch': 6.89} +04/19 [21:18:45] INFO | >> train_qwenlatent.py:487 + Step 27320 | grad_norm_pre_clip=0.1820 | + grad_norm_pre_clip_avg=0.1933 | Metrics: + {'align_loss': 0.02446126937866211, + 'recon_loss': 0.09953124076128006, + 'predict_loss': 0.007759205996990204, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18197199702262878, + 'data_time': 0.0005744509981013834, + 'model_time': 1.2181151080003474, + 'grad_norm_pre_clip_avg': 0.19325413107872008, + 'learning_rate': 1.2669418423043367e-05, + 'epoch': 6.89} +04/19 [21:18:58] INFO | >> train_qwenlatent.py:487 + Step 27330 | grad_norm_pre_clip=0.1385 | + grad_norm_pre_clip_avg=0.1766 | Metrics: + {'align_loss': 0.024784620851278305, + 'recon_loss': 0.08241795748472214, + 'predict_loss': 0.0071932477876544, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13854283094406128, + 'data_time': 0.0007682850118726492, + 'model_time': 1.5420281550032087, + 'grad_norm_pre_clip_avg': 0.1766115814447403, + 'learning_rate': 1.2660701154227983e-05, + 'epoch': 6.9} +04/19 [21:19:11] INFO | >> train_qwenlatent.py:487 + Step 27340 | grad_norm_pre_clip=0.1567 | + grad_norm_pre_clip_avg=0.1619 | Metrics: + {'align_loss': 0.024876857176423073, + 'recon_loss': 0.0911465510725975, + 'predict_loss': 0.00737870205193758, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15665507316589355, + 'data_time': 0.0007020019984338433, + 'model_time': 1.505061571020633, + 'grad_norm_pre_clip_avg': 0.16189075112342835, + 'learning_rate': 1.2651983813181162e-05, + 'epoch': 6.9} +04/19 [21:19:25] INFO | >> train_qwenlatent.py:487 + Step 27350 | grad_norm_pre_clip=0.1887 | + grad_norm_pre_clip_avg=0.1637 | Metrics: + {'align_loss': 0.024193542078137398, + 'recon_loss': 0.09136480838060379, + 'predict_loss': 0.010569610632956028, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18871602416038513, + 'mae_score': 0.00920764819995777, 'data_time': + 0.0006128350214567035, 'model_time': + 1.240455221006414, 'grad_norm_pre_clip_avg': + 0.16369074881076812, 'learning_rate': + 1.2643266404151622e-05, 'epoch': 6.9} +04/19 [21:19:37] INFO | >> train_qwenlatent.py:487 + Step 27360 | grad_norm_pre_clip=0.2088 | + grad_norm_pre_clip_avg=0.1956 | Metrics: + {'align_loss': 0.025180574506521225, + 'recon_loss': 0.0900244414806366, + 'predict_loss': 0.012559664435684681, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20877589285373688, + 'data_time': 0.0014249889936763793, + 'model_time': 1.2579082809970714, + 'grad_norm_pre_clip_avg': 0.19555339813232422, + 'learning_rate': 1.2634548931388124e-05, + 'epoch': 6.9} +04/19 [21:19:50] INFO | >> train_qwenlatent.py:487 + Step 27370 | grad_norm_pre_clip=0.1985 | + grad_norm_pre_clip_avg=0.1991 | Metrics: + {'align_loss': 0.024920573458075523, + 'recon_loss': 0.0748714730143547, + 'predict_loss': 0.00678539602085948, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19852742552757263, + 'data_time': 0.0009570849942974746, + 'model_time': 1.2880514460266568, + 'grad_norm_pre_clip_avg': 0.19911237955093383, + 'learning_rate': 1.2625831399139457e-05, + 'epoch': 6.91} +04/19 [21:20:02] INFO | >> train_qwenlatent.py:487 + Step 27380 | grad_norm_pre_clip=0.1360 | + grad_norm_pre_clip_avg=0.1662 | Metrics: + {'align_loss': 0.025199193507432938, + 'recon_loss': 0.07111845910549164, + 'predict_loss': 0.004767131991684437, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13602109253406525, + 'data_time': 0.0006066810165066272, + 'model_time': 1.213674726022873, + 'grad_norm_pre_clip_avg': 0.16619766056537627, + 'learning_rate': 1.2617113811654448e-05, + 'epoch': 6.91} +04/19 [21:20:14] INFO | >> train_qwenlatent.py:487 + Step 27390 | grad_norm_pre_clip=0.1488 | + grad_norm_pre_clip_avg=0.1563 | Metrics: + {'align_loss': 0.02456590346992016, + 'recon_loss': 0.09665259718894958, + 'predict_loss': 0.008472944609820843, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14877641201019287, + 'data_time': 0.0008562699949834496, + 'model_time': 1.2696786260057706, + 'grad_norm_pre_clip_avg': 0.15629273951053618, + 'learning_rate': 1.2608396173181939e-05, + 'epoch': 6.91} +04/19 [21:20:27] INFO | >> train_qwenlatent.py:487 + Step 27400 | grad_norm_pre_clip=0.1507 | + grad_norm_pre_clip_avg=0.1615 | Metrics: + {'align_loss': 0.025499064475297928, + 'recon_loss': 0.08564038574695587, + 'predict_loss': 0.006848809774965048, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1507326364517212, + 'mae_score': 0.009396548743720527, 'data_time': + 0.0010496829927433282, 'model_time': + 1.2333633480011486, 'grad_norm_pre_clip_avg': + 0.16145842671394348, 'learning_rate': + 1.2599678487970796e-05, 'epoch': 6.91} +04/19 [21:20:40] INFO | >> train_qwenlatent.py:487 + Step 27410 | grad_norm_pre_clip=0.2242 | + grad_norm_pre_clip_avg=0.1887 | Metrics: + {'align_loss': 0.024922417476773262, + 'recon_loss': 0.08836966007947922, + 'predict_loss': 0.01054852269589901, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22419434785842896, + 'data_time': 0.0005800030194222927, + 'model_time': 1.2454552589915693, + 'grad_norm_pre_clip_avg': 0.18869101256132126, + 'learning_rate': 1.2590960760269918e-05, + 'epoch': 6.92} +04/19 [21:20:52] INFO | >> train_qwenlatent.py:487 + Step 27420 | grad_norm_pre_clip=0.1786 | + grad_norm_pre_clip_avg=0.1661 | Metrics: + {'align_loss': 0.026113316416740417, + 'recon_loss': 0.11147449910640717, + 'predict_loss': 0.018599962815642357, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1785893738269806, + 'data_time': 0.0006546470103785396, + 'model_time': 1.2654029350087512, + 'grad_norm_pre_clip_avg': 0.16609809547662735, + 'learning_rate': 1.2582242994328217e-05, + 'epoch': 6.92} +04/19 [21:21:04] INFO | >> train_qwenlatent.py:487 + Step 27430 | grad_norm_pre_clip=0.1726 | + grad_norm_pre_clip_avg=0.1799 | Metrics: + {'align_loss': 0.026014793664216995, + 'recon_loss': 0.08746518194675446, + 'predict_loss': 0.01074636448174715, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17255127429962158, + 'data_time': 0.0010833549895323813, + 'model_time': 1.2147486679896247, + 'grad_norm_pre_clip_avg': 0.17992478162050246, + 'learning_rate': 1.2573525194394633e-05, + 'epoch': 6.92} +04/19 [21:21:17] INFO | >> train_qwenlatent.py:487 + Step 27440 | grad_norm_pre_clip=0.1821 | + grad_norm_pre_clip_avg=0.1849 | Metrics: + {'align_loss': 0.026052020490169525, + 'recon_loss': 0.1033937931060791, + 'predict_loss': 0.010131643153727055, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18208175897598267, + 'data_time': 0.0010377010039519519, + 'model_time': 1.5570838449930307, + 'grad_norm_pre_clip_avg': 0.18489904701709747, + 'learning_rate': 1.2564807364718106e-05, + 'epoch': 6.92} +04/19 [21:21:30] INFO | >> train_qwenlatent.py:487 + Step 27450 | grad_norm_pre_clip=0.2022 | + grad_norm_pre_clip_avg=0.1698 | Metrics: + {'align_loss': 0.026358701288700104, + 'recon_loss': 0.10669778287410736, + 'predict_loss': 0.012062282301485538, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20223845541477203, + 'mae_score': 0.0080947343293611, 'data_time': + 0.0006855339743196964, 'model_time': + 1.2347069629759062, 'grad_norm_pre_clip_avg': + 0.169809727370739, 'learning_rate': + 1.2556089509547608e-05, 'epoch': 6.93} +04/19 [21:21:43] INFO | >> train_qwenlatent.py:487 + Step 27460 | grad_norm_pre_clip=0.2388 | + grad_norm_pre_clip_avg=0.2306 | Metrics: + {'align_loss': 0.02603272721171379, + 'recon_loss': 0.0890485942363739, + 'predict_loss': 0.011586981825530529, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2387518286705017, + 'data_time': 0.0006149290129542351, + 'model_time': 1.2243804419995286, + 'grad_norm_pre_clip_avg': 0.23060766160488128, + 'learning_rate': 1.2547371633132114e-05, + 'epoch': 6.93} +04/19 [21:21:56] INFO | >> train_qwenlatent.py:487 + Step 27470 | grad_norm_pre_clip=0.1534 | + grad_norm_pre_clip_avg=0.1684 | Metrics: + {'align_loss': 0.024827782064676285, + 'recon_loss': 0.09133300185203552, + 'predict_loss': 0.009197740815579891, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1533716917037964, + 'data_time': 0.0006509989907499403, + 'model_time': 1.219478037004592, + 'grad_norm_pre_clip_avg': 0.16836456656455995, + 'learning_rate': 1.2538653739720602e-05, + 'epoch': 6.93} +04/19 [21:22:08] INFO | >> train_qwenlatent.py:487 + Step 27480 | grad_norm_pre_clip=0.1822 | + grad_norm_pre_clip_avg=0.1755 | Metrics: + {'align_loss': 0.025218479335308075, + 'recon_loss': 0.09580783545970917, + 'predict_loss': 0.0064453319646418095, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18224583566188812, + 'data_time': 0.0006333449855446815, + 'model_time': 1.2694976339989807, + 'grad_norm_pre_clip_avg': 0.17545965164899827, + 'learning_rate': 1.2529935833562083e-05, + 'epoch': 6.93} +04/19 [21:22:21] INFO | >> train_qwenlatent.py:487 + Step 27490 | grad_norm_pre_clip=0.2158 | + grad_norm_pre_clip_avg=0.2061 | Metrics: + {'align_loss': 0.023993417620658875, + 'recon_loss': 0.09213512390851974, + 'predict_loss': 0.014533784240484238, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21580183506011963, + 'data_time': 0.0008528339967597276, + 'model_time': 1.30018937800196, + 'grad_norm_pre_clip_avg': 0.20606958419084548, + 'learning_rate': 1.2521217918905542e-05, + 'epoch': 6.94} +04/19 [21:22:34] INFO | >> train_qwenlatent.py:487 + Step 27500 | grad_norm_pre_clip=0.2349 | + grad_norm_pre_clip_avg=0.1995 | Metrics: + {'align_loss': 0.02601351961493492, + 'recon_loss': 0.1032843366265297, + 'predict_loss': 0.010992961004376411, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2348734736442566, + 'mae_score': 0.010876510379550694, 'data_time': + 0.0008377350168302655, 'model_time': + 1.207993134012213, 'grad_norm_pre_clip_avg': + 0.1994696781039238, 'learning_rate': + 1.25125e-05, 'epoch': 6.94} +04/19 [21:22:47] INFO | >> train_qwenlatent.py:487 + Step 27510 | grad_norm_pre_clip=0.1917 | + grad_norm_pre_clip_avg=0.1825 | Metrics: + {'align_loss': 0.02643156610429287, + 'recon_loss': 0.09857635945081711, + 'predict_loss': 0.011002425104379654, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19167988002300262, + 'data_time': 0.0009666509868111461, + 'model_time': 1.227020522026578, + 'grad_norm_pre_clip_avg': 0.1825417622923851, + 'learning_rate': 1.2503782081094456e-05, + 'epoch': 6.94} +04/19 [21:22:59] INFO | >> train_qwenlatent.py:487 + Step 27520 | grad_norm_pre_clip=0.1674 | + grad_norm_pre_clip_avg=0.1828 | Metrics: + {'align_loss': 0.02503993734717369, + 'recon_loss': 0.09297727793455124, + 'predict_loss': 0.009704469703137875, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16743525862693787, + 'data_time': 0.0006184930098243058, + 'model_time': 1.2034104280173779, + 'grad_norm_pre_clip_avg': 0.18276416659355163, + 'learning_rate': 1.249506416643792e-05, + 'epoch': 6.94} +04/19 [21:23:12] INFO | >> train_qwenlatent.py:487 + Step 27530 | grad_norm_pre_clip=0.1673 | + grad_norm_pre_clip_avg=0.1669 | Metrics: + {'align_loss': 0.02506215125322342, + 'recon_loss': 0.11559495329856873, + 'predict_loss': 0.011891388334333897, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16734223067760468, + 'data_time': 0.0008696940203662962, + 'model_time': 1.198373499995796, + 'grad_norm_pre_clip_avg': 0.1668836459517479, + 'learning_rate': 1.2486346260279398e-05, + 'epoch': 6.95} +04/19 [21:23:24] INFO | >> train_qwenlatent.py:487 + Step 27540 | grad_norm_pre_clip=0.1671 | + grad_norm_pre_clip_avg=0.1789 | Metrics: + {'align_loss': 0.024314872920513153, + 'recon_loss': 0.08799508959054947, + 'predict_loss': 0.00805539172142744, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16711053252220154, + 'data_time': 0.0008469200110994279, + 'model_time': 1.2705201100034174, + 'grad_norm_pre_clip_avg': 0.17891010046005248, + 'learning_rate': 1.2477628366867892e-05, + 'epoch': 6.95} +04/19 [21:23:37] INFO | >> train_qwenlatent.py:487 + Step 27550 | grad_norm_pre_clip=0.2416 | + grad_norm_pre_clip_avg=0.1892 | Metrics: + {'align_loss': 0.0255531445145607, + 'recon_loss': 0.0809076726436615, + 'predict_loss': 0.006826408207416534, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2415991574525833, + 'mae_score': 0.008284474278355504, 'data_time': + 0.0005755260062869638, 'model_time': + 1.2253599019895773, 'grad_norm_pre_clip_avg': + 0.18923741579055786, 'learning_rate': + 1.2468910490452398e-05, 'epoch': 6.95} +04/19 [21:23:50] INFO | >> train_qwenlatent.py:487 + Step 27560 | grad_norm_pre_clip=0.1889 | + grad_norm_pre_clip_avg=0.1872 | Metrics: + {'align_loss': 0.025736741721630096, + 'recon_loss': 0.10532978922128677, + 'predict_loss': 0.009293492883443832, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18893685936927795, + 'data_time': 0.0006251829909160733, + 'model_time': 1.21785410301527, + 'grad_norm_pre_clip_avg': 0.18715080618858337, + 'learning_rate': 1.2460192635281896e-05, + 'epoch': 6.95} +04/19 [21:24:03] INFO | >> train_qwenlatent.py:487 + Step 27570 | grad_norm_pre_clip=0.1516 | + grad_norm_pre_clip_avg=0.1582 | Metrics: + {'align_loss': 0.025298623368144035, + 'recon_loss': 0.09327911585569382, + 'predict_loss': 0.008187871426343918, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.151622474193573, + 'data_time': 0.0007905970269348472, + 'model_time': 1.2086645370000042, + 'grad_norm_pre_clip_avg': 0.15824268534779548, + 'learning_rate': 1.2451474805605372e-05, + 'epoch': 6.96} +04/19 [21:24:15] INFO | >> train_qwenlatent.py:487 + Step 27580 | grad_norm_pre_clip=0.2648 | + grad_norm_pre_clip_avg=0.1738 | Metrics: + {'align_loss': 0.025950953364372253, + 'recon_loss': 0.07735607028007507, + 'predict_loss': 0.01303771324455738, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2648211419582367, + 'data_time': 0.0007366499921772629, + 'model_time': 1.2544138160010334, + 'grad_norm_pre_clip_avg': 0.173824343085289, + 'learning_rate': 1.2442757005671785e-05, + 'epoch': 6.96} +04/19 [21:24:28] INFO | >> train_qwenlatent.py:487 + Step 27590 | grad_norm_pre_clip=0.2413 | + grad_norm_pre_clip_avg=0.2258 | Metrics: + {'align_loss': 0.026130076497793198, + 'recon_loss': 0.1273891031742096, + 'predict_loss': 0.013470908626914024, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2412703037261963, + 'data_time': 0.0009648139821365476, + 'model_time': 1.2422471989993937, + 'grad_norm_pre_clip_avg': 0.22580040693283082, + 'learning_rate': 1.2434039239730084e-05, + 'epoch': 6.96} +04/19 [21:24:41] INFO | >> train_qwenlatent.py:487 + Step 27600 | grad_norm_pre_clip=0.1755 | + grad_norm_pre_clip_avg=0.1902 | Metrics: + {'align_loss': 0.026104256510734558, + 'recon_loss': 0.10226696729660034, + 'predict_loss': 0.010303026996552944, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1755099594593048, + 'mae_score': 0.009291426340738932, 'data_time': + 0.0005906199803575873, 'model_time': + 1.2667107750021387, 'grad_norm_pre_clip_avg': + 0.1902368873357773, 'learning_rate': + 1.2425321512029206e-05, 'epoch': 6.96} +04/19 [21:24:54] INFO | >> train_qwenlatent.py:487 + Step 27610 | grad_norm_pre_clip=0.1591 | + grad_norm_pre_clip_avg=0.1623 | Metrics: + {'align_loss': 0.02594892308115959, + 'recon_loss': 0.10545409470796585, + 'predict_loss': 0.011950468644499779, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1591380536556244, + 'data_time': 0.0010256270179525018, + 'model_time': 1.57431351399282, + 'grad_norm_pre_clip_avg': 0.16230881214141846, + 'learning_rate': 1.2416603826818063e-05, + 'epoch': 6.97} +04/19 [21:25:07] INFO | >> train_qwenlatent.py:487 + Step 27620 | grad_norm_pre_clip=0.1532 | + grad_norm_pre_clip_avg=0.1570 | Metrics: + {'align_loss': 0.026697572320699692, + 'recon_loss': 0.12483520805835724, + 'predict_loss': 0.009611088782548904, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15320102870464325, + 'data_time': 0.0008290229889098555, + 'model_time': 1.2363030720152892, + 'grad_norm_pre_clip_avg': 0.15695488601922988, + 'learning_rate': 1.2407886188345552e-05, + 'epoch': 6.97} +04/19 [21:25:20] INFO | >> train_qwenlatent.py:487 + Step 27630 | grad_norm_pre_clip=0.2285 | + grad_norm_pre_clip_avg=0.1883 | Metrics: + {'align_loss': 0.025690900161862373, + 'recon_loss': 0.10350114107131958, + 'predict_loss': 0.014905465766787529, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22847065329551697, + 'data_time': 0.0007799009908922017, + 'model_time': 1.2003589129890315, + 'grad_norm_pre_clip_avg': 0.18830517679452896, + 'learning_rate': 1.2399168600860542e-05, + 'epoch': 6.97} +04/19 [21:25:32] INFO | >> train_qwenlatent.py:487 + Step 27640 | grad_norm_pre_clip=0.1739 | + grad_norm_pre_clip_avg=0.2153 | Metrics: + {'align_loss': 0.02454991824924946, + 'recon_loss': 0.08228641748428345, + 'predict_loss': 0.006911188829690218, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17387472093105316, + 'data_time': 0.0006969160167500377, + 'model_time': 1.2501647400204092, + 'grad_norm_pre_clip_avg': 0.21534211337566375, + 'learning_rate': 1.239045106861188e-05, + 'epoch': 6.97} +04/19 [21:25:45] INFO | >> train_qwenlatent.py:487 + Step 27650 | grad_norm_pre_clip=0.2169 | + grad_norm_pre_clip_avg=0.2058 | Metrics: + {'align_loss': 0.026190146803855896, + 'recon_loss': 0.1096445843577385, + 'predict_loss': 0.013456217013299465, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2168850600719452, + 'mae_score': 0.012140411514419693, 'data_time': + 0.0006309860036708415, 'model_time': + 1.2253421600034926, 'grad_norm_pre_clip_avg': + 0.20580194890499115, 'learning_rate': + 1.2381733595848385e-05, 'epoch': 6.98} +04/19 [21:25:58] INFO | >> train_qwenlatent.py:487 + Step 27660 | grad_norm_pre_clip=0.1826 | + grad_norm_pre_clip_avg=0.1974 | Metrics: + {'align_loss': 0.025438129901885986, + 'recon_loss': 0.09085999429225922, + 'predict_loss': 0.008958063088357449, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18255262076854706, + 'data_time': 0.0008577140106353909, + 'model_time': 1.1968985910061747, + 'grad_norm_pre_clip_avg': 0.19737145602703093, + 'learning_rate': 1.2373016186818843e-05, + 'epoch': 6.98} +04/19 [21:26:10] INFO | >> train_qwenlatent.py:487 + Step 27670 | grad_norm_pre_clip=0.2078 | + grad_norm_pre_clip_avg=0.1868 | Metrics: + {'align_loss': 0.025152843445539474, + 'recon_loss': 0.09801401942968369, + 'predict_loss': 0.00945992674678564, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20783032476902008, + 'data_time': 0.0010449200053699315, + 'model_time': 1.2168653919943608, + 'grad_norm_pre_clip_avg': 0.18681954890489577, + 'learning_rate': 1.2364298845772017e-05, + 'epoch': 6.98} +04/19 [21:26:23] INFO | >> train_qwenlatent.py:487 + Step 27680 | grad_norm_pre_clip=0.1346 | + grad_norm_pre_clip_avg=0.2032 | Metrics: + {'align_loss': 0.025972118601202965, + 'recon_loss': 0.08723907917737961, + 'predict_loss': 0.0046609933488070965, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13458678126335144, + 'data_time': 0.0011978140100836754, + 'model_time': 1.2904850449995138, + 'grad_norm_pre_clip_avg': 0.20322489440441133, + 'learning_rate': 1.2355581576956633e-05, + 'epoch': 6.98} +04/19 [21:26:35] INFO | >> train_qwenlatent.py:487 + Step 27690 | grad_norm_pre_clip=0.1791 | + grad_norm_pre_clip_avg=0.1742 | Metrics: + {'align_loss': 0.025931954383850098, + 'recon_loss': 0.09355515986680984, + 'predict_loss': 0.009920055978000164, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17914773523807526, + 'data_time': 0.0007098119822330773, + 'model_time': 1.2847395279968623, + 'grad_norm_pre_clip_avg': 0.17417835295200348, + 'learning_rate': 1.2346864384621387e-05, + 'epoch': 6.99} +04/19 [21:26:49] INFO | >> train_qwenlatent.py:487 + Step 27700 | grad_norm_pre_clip=0.1996 | + grad_norm_pre_clip_avg=0.1810 | Metrics: + {'align_loss': 0.025080418214201927, + 'recon_loss': 0.08598282188177109, + 'predict_loss': 0.007903114892542362, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19962669909000397, + 'mae_score': 0.01075658282718143, 'data_time': + 0.0009514069824945182, 'model_time': + 1.2772624139906839, 'grad_norm_pre_clip_avg': + 0.1810445100069046, 'learning_rate': + 1.2338147273014924e-05, 'epoch': 6.99} +04/19 [21:27:01] INFO | >> train_qwenlatent.py:487 + Step 27710 | grad_norm_pre_clip=0.1597 | + grad_norm_pre_clip_avg=0.1948 | Metrics: + {'align_loss': 0.02644258365035057, + 'recon_loss': 0.11367734521627426, + 'predict_loss': 0.010125636123120785, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15968045592308044, + 'data_time': 0.0008784300007391721, + 'model_time': 1.2311090500152204, + 'grad_norm_pre_clip_avg': 0.19480538368225098, + 'learning_rate': 1.2329430246385862e-05, + 'epoch': 6.99} +04/19 [21:27:14] INFO | >> train_qwenlatent.py:487 + Step 27720 | grad_norm_pre_clip=0.1686 | + grad_norm_pre_clip_avg=0.1584 | Metrics: + {'align_loss': 0.025597505271434784, + 'recon_loss': 0.09879320114850998, + 'predict_loss': 0.011677434667944908, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16856826841831207, + 'data_time': 0.0013419060269370675, + 'model_time': 1.2660905060183723, + 'grad_norm_pre_clip_avg': 0.15842732563614845, + 'learning_rate': 1.2320713308982775e-05, + 'epoch': 6.99} +04/19 [21:27:27] INFO | >> train_qwenlatent.py:487 + Step 27730 | grad_norm_pre_clip=0.1977 | + grad_norm_pre_clip_avg=0.2012 | Metrics: + {'align_loss': 0.02438468486070633, + 'recon_loss': 0.09611190855503082, + 'predict_loss': 0.011022986844182014, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1976594626903534, + 'data_time': 0.0009661350050009787, + 'model_time': 1.2483248809876386, + 'grad_norm_pre_clip_avg': 0.2012227788567543, + 'learning_rate': 1.2311996465054194e-05, + 'epoch': 7.0} +04/19 [21:27:39] INFO | >> train_qwenlatent.py:487 + Step 27740 | grad_norm_pre_clip=0.1922 | + grad_norm_pre_clip_avg=0.1994 | Metrics: + {'align_loss': 0.024057673290371895, + 'recon_loss': 0.11000803858041763, + 'predict_loss': 0.011741405352950096, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19218555092811584, + 'data_time': 0.0007573780021630228, + 'model_time': 1.243436165008461, + 'grad_norm_pre_clip_avg': 0.19940270334482194, + 'learning_rate': 1.2303279718848599e-05, + 'epoch': 7.0} +04/19 [21:27:53] INFO | >> train_qwenlatent.py:487 + Step 27750 | grad_norm_pre_clip=0.1424 | + grad_norm_pre_clip_avg=0.1774 | Metrics: + {'align_loss': 0.024794884026050568, + 'recon_loss': 0.0706813633441925, + 'predict_loss': 0.006071153562515974, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14242669939994812, + 'mae_score': 0.008654521392272399, 'data_time': + 0.0009173919970635325, 'model_time': + 1.3511327900050674, 'grad_norm_pre_clip_avg': + 0.17742441445589066, 'learning_rate': + 1.2294563074614422e-05, 'epoch': 7.0} +04/19 [21:28:05] INFO | >> train_qwenlatent.py:487 + Step 27760 | grad_norm_pre_clip=0.1247 | + grad_norm_pre_clip_avg=0.1713 | Metrics: + {'align_loss': 0.023354876786470413, + 'recon_loss': 0.08404084295034409, + 'predict_loss': 0.008054723963141441, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1246618777513504, + 'data_time': 0.0007248810143209994, + 'model_time': 1.2897511990158819, + 'grad_norm_pre_clip_avg': 0.17126312106847763, + 'learning_rate': 1.228584653660006e-05, + 'epoch': 7.0} +04/19 [21:28:18] INFO | >> train_qwenlatent.py:487 + Step 27770 | grad_norm_pre_clip=0.1509 | + grad_norm_pre_clip_avg=0.1751 | Metrics: + {'align_loss': 0.02475283294916153, + 'recon_loss': 0.07405680418014526, + 'predict_loss': 0.012144790031015873, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15092961490154266, + 'data_time': 0.0009801030100788921, + 'model_time': 1.3010316549916752, + 'grad_norm_pre_clip_avg': 0.17507810592651368, + 'learning_rate': 1.2277130109053842e-05, + 'epoch': 7.01} +04/19 [21:28:30] INFO | >> train_qwenlatent.py:487 + Step 27780 | grad_norm_pre_clip=0.1786 | + grad_norm_pre_clip_avg=0.1758 | Metrics: + {'align_loss': 0.026213843375444412, + 'recon_loss': 0.0849643424153328, + 'predict_loss': 0.007785615045577288, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1786472052335739, + 'data_time': 0.0008226530044339597, + 'model_time': 1.2146809349942487, + 'grad_norm_pre_clip_avg': 0.17584771364927293, + 'learning_rate': 1.2268413796224047e-05, + 'epoch': 7.01} +04/19 [21:28:43] INFO | >> train_qwenlatent.py:487 + Step 27790 | grad_norm_pre_clip=0.2039 | + grad_norm_pre_clip_avg=0.1845 | Metrics: + {'align_loss': 0.02567490004003048, + 'recon_loss': 0.09678568691015244, + 'predict_loss': 0.009329610504209995, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20386147499084473, + 'data_time': 0.0009204289817716926, + 'model_time': 1.2379618319973815, + 'grad_norm_pre_clip_avg': 0.18447802364826202, + 'learning_rate': 1.2259697602358906e-05, + 'epoch': 7.01} +04/19 [21:28:56] INFO | >> train_qwenlatent.py:487 + Step 27800 | grad_norm_pre_clip=0.1786 | + grad_norm_pre_clip_avg=0.2016 | Metrics: + {'align_loss': 0.024604899808764458, + 'recon_loss': 0.10140122473239899, + 'predict_loss': 0.008722846396267414, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1785702109336853, + 'mae_score': 0.009068489074707031, 'data_time': + 0.0007882350182626396, 'model_time': + 1.22310581500642, 'grad_norm_pre_clip_avg': + 0.20164222419261932, 'learning_rate': + 1.2250981531706582e-05, 'epoch': 7.01} +04/19 [21:29:09] INFO | >> train_qwenlatent.py:487 + Step 27810 | grad_norm_pre_clip=0.1854 | + grad_norm_pre_clip_avg=0.1873 | Metrics: + {'align_loss': 0.025096189230680466, + 'recon_loss': 0.0952233374118805, + 'predict_loss': 0.011288519017398357, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18543025851249695, + 'data_time': 0.0011789090058300644, + 'model_time': 1.2913621129991952, + 'grad_norm_pre_clip_avg': 0.1873043030500412, + 'learning_rate': 1.2242265588515179e-05, + 'epoch': 7.02} +04/19 [21:29:21] INFO | >> train_qwenlatent.py:487 + Step 27820 | grad_norm_pre_clip=0.1842 | + grad_norm_pre_clip_avg=0.1739 | Metrics: + {'align_loss': 0.025130856782197952, + 'recon_loss': 0.08184990286827087, + 'predict_loss': 0.0077438331209123135, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18416093289852142, + 'data_time': 0.0009517720027361065, + 'model_time': 1.229627303982852, + 'grad_norm_pre_clip_avg': 0.1739347517490387, + 'learning_rate': 1.223354977703275e-05, + 'epoch': 7.02} +04/19 [21:29:34] INFO | >> train_qwenlatent.py:487 + Step 27830 | grad_norm_pre_clip=0.2165 | + grad_norm_pre_clip_avg=0.1891 | Metrics: + {'align_loss': 0.02546246349811554, + 'recon_loss': 0.11468145996332169, + 'predict_loss': 0.01029980182647705, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21648503839969635, + 'data_time': 0.0008791380096226931, + 'model_time': 1.2472612479759846, + 'grad_norm_pre_clip_avg': 0.18912108689546586, + 'learning_rate': 1.222483410150727e-05, + 'epoch': 7.02} +04/19 [21:29:47] INFO | >> train_qwenlatent.py:487 + Step 27840 | grad_norm_pre_clip=0.2212 | + grad_norm_pre_clip_avg=0.1888 | Metrics: + {'align_loss': 0.025079723447561264, + 'recon_loss': 0.10286539793014526, + 'predict_loss': 0.012211435474455357, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22121194005012512, + 'data_time': 0.0009371449996251613, + 'model_time': 1.2895030509971548, + 'grad_norm_pre_clip_avg': 0.1887938290834427, + 'learning_rate': 1.2216118566186655e-05, + 'epoch': 7.02} +04/19 [21:30:00] INFO | >> train_qwenlatent.py:487 + Step 27850 | grad_norm_pre_clip=0.2258 | + grad_norm_pre_clip_avg=0.1805 | Metrics: + {'align_loss': 0.025657230988144875, + 'recon_loss': 0.1034960225224495, + 'predict_loss': 0.011591853573918343, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.225785493850708, + 'mae_score': 0.009027429529138514, 'data_time': + 0.001018920011119917, 'model_time': + 1.2667850779835135, 'grad_norm_pre_clip_avg': + 0.1804801791906357, 'learning_rate': + 1.2207403175318745e-05, 'epoch': 7.03} +04/19 [21:30:12] INFO | >> train_qwenlatent.py:487 + Step 27860 | grad_norm_pre_clip=0.1414 | + grad_norm_pre_clip_avg=0.1743 | Metrics: + {'align_loss': 0.0253458134829998, + 'recon_loss': 0.11408042907714844, + 'predict_loss': 0.010742278769612312, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1413877010345459, + 'data_time': 0.0006448619824368507, + 'model_time': 1.2310093239939306, + 'grad_norm_pre_clip_avg': 0.17433633357286454, + 'learning_rate': 1.2198687933151324e-05, + 'epoch': 7.03} +04/19 [21:30:25] INFO | >> train_qwenlatent.py:487 + Step 27870 | grad_norm_pre_clip=0.2429 | + grad_norm_pre_clip_avg=0.1831 | Metrics: + {'align_loss': 0.024922464042901993, + 'recon_loss': 0.06077353283762932, + 'predict_loss': 0.007109359838068485, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24291202425956726, + 'data_time': 0.0006192609725985676, + 'model_time': 1.2860117810196243, + 'grad_norm_pre_clip_avg': 0.18305466026067735, + 'learning_rate': 1.218997284393209e-05, + 'epoch': 7.03} +04/19 [21:30:38] INFO | >> train_qwenlatent.py:487 + Step 27880 | grad_norm_pre_clip=0.1201 | + grad_norm_pre_clip_avg=0.1574 | Metrics: + {'align_loss': 0.025556016713380814, + 'recon_loss': 0.0812985822558403, + 'predict_loss': 0.004244555253535509, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12006489932537079, + 'data_time': 0.0006258680077735335, + 'model_time': 1.229712054017, + 'grad_norm_pre_clip_avg': 0.15735220164060593, + 'learning_rate': 1.2181257911908678e-05, + 'epoch': 7.04} +04/19 [21:30:51] INFO | >> train_qwenlatent.py:487 + Step 27890 | grad_norm_pre_clip=0.2227 | + grad_norm_pre_clip_avg=0.1827 | Metrics: + {'align_loss': 0.025540176779031754, + 'recon_loss': 0.10765890777111053, + 'predict_loss': 0.014366134069859982, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2226506769657135, + 'data_time': 0.0005939230031799525, + 'model_time': 1.226580284012016, + 'grad_norm_pre_clip_avg': 0.18270479291677474, + 'learning_rate': 1.2172543141328635e-05, + 'epoch': 7.04} +04/19 [21:31:04] INFO | >> train_qwenlatent.py:487 + Step 27900 | grad_norm_pre_clip=0.1683 | + grad_norm_pre_clip_avg=0.1949 | Metrics: + {'align_loss': 0.02576056495308876, + 'recon_loss': 0.14524415135383606, + 'predict_loss': 0.01162381935864687, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16828808188438416, + 'mae_score': 0.008372072271398596, 'data_time': + 0.0006635449826717377, 'model_time': + 1.2290655880060513, 'grad_norm_pre_clip_avg': + 0.1949355944991112, 'learning_rate': + 1.2163828536439437e-05, 'epoch': 7.04} +04/19 [21:31:17] INFO | >> train_qwenlatent.py:487 + Step 27910 | grad_norm_pre_clip=0.1961 | + grad_norm_pre_clip_avg=0.1826 | Metrics: + {'align_loss': 0.023138761520385742, + 'recon_loss': 0.07648197561502457, + 'predict_loss': 0.009656556881964207, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19607166945934296, + 'data_time': 0.0008498490205965936, + 'model_time': 1.2737444039958064, + 'grad_norm_pre_clip_avg': 0.182582825422287, + 'learning_rate': 1.2155114101488472e-05, + 'epoch': 7.04} +04/19 [21:31:29] INFO | >> train_qwenlatent.py:487 + Step 27920 | grad_norm_pre_clip=0.1731 | + grad_norm_pre_clip_avg=0.1634 | Metrics: + {'align_loss': 0.02446914091706276, + 'recon_loss': 0.07376649975776672, + 'predict_loss': 0.01109582930803299, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17311115562915802, + 'data_time': 0.0006609760166611522, + 'model_time': 1.1901188929914497, + 'grad_norm_pre_clip_avg': 0.16337015330791474, + 'learning_rate': 1.2146399840723052e-05, + 'epoch': 7.05} +04/19 [21:31:42] INFO | >> train_qwenlatent.py:487 + Step 27930 | grad_norm_pre_clip=0.1640 | + grad_norm_pre_clip_avg=0.1592 | Metrics: + {'align_loss': 0.02538498491048813, + 'recon_loss': 0.1056680828332901, + 'predict_loss': 0.010548766702413559, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1640021800994873, + 'data_time': 0.001479933998780325, + 'model_time': 1.341004895977676, + 'grad_norm_pre_clip_avg': 0.15917817503213882, + 'learning_rate': 1.21376857583904e-05, 'epoch': + 7.05} +04/19 [21:31:54] INFO | >> train_qwenlatent.py:487 + Step 27940 | grad_norm_pre_clip=0.2280 | + grad_norm_pre_clip_avg=0.1902 | Metrics: + {'align_loss': 0.02468305267393589, + 'recon_loss': 0.09794557839632034, + 'predict_loss': 0.011535952799022198, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22804324328899384, + 'data_time': 0.0009861550061032176, + 'model_time': 1.2421984279935714, + 'grad_norm_pre_clip_avg': 0.190192911028862, + 'learning_rate': 1.2128971858737658e-05, + 'epoch': 7.05} +04/19 [21:32:07] INFO | >> train_qwenlatent.py:487 + Step 27950 | grad_norm_pre_clip=0.1889 | + grad_norm_pre_clip_avg=0.2041 | Metrics: + {'align_loss': 0.024618953466415405, + 'recon_loss': 0.08530910313129425, + 'predict_loss': 0.011242115870118141, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1888534426689148, + 'mae_score': 0.00931406021118164, 'data_time': + 0.000902522006072104, 'model_time': + 1.2394413220172282, 'grad_norm_pre_clip_avg': + 0.20410848259925843, 'learning_rate': + 1.2120258146011873e-05, 'epoch': 7.05} +04/19 [21:32:20] INFO | >> train_qwenlatent.py:487 + Step 27960 | grad_norm_pre_clip=0.1858 | + grad_norm_pre_clip_avg=0.1817 | Metrics: + {'align_loss': 0.02551940083503723, + 'recon_loss': 0.07620636373758316, + 'predict_loss': 0.007623391691595316, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18584199249744415, + 'data_time': 0.0009739819797687232, + 'model_time': 1.2607004029850941, + 'grad_norm_pre_clip_avg': 0.18167477250099182, + 'learning_rate': 1.2111544624460005e-05, + 'epoch': 7.06} +04/19 [21:32:33] INFO | >> train_qwenlatent.py:487 + Step 27970 | grad_norm_pre_clip=0.1591 | + grad_norm_pre_clip_avg=0.1683 | Metrics: + {'align_loss': 0.025321543216705322, + 'recon_loss': 0.10288162529468536, + 'predict_loss': 0.01043705828487873, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15909281373023987, + 'data_time': 0.00082894999650307, 'model_time': + 1.2334405670117121, 'grad_norm_pre_clip_avg': + 0.16831534057855607, 'learning_rate': + 1.210283129832891e-05, 'epoch': 7.06} +04/19 [21:32:46] INFO | >> train_qwenlatent.py:487 + Step 27980 | grad_norm_pre_clip=0.1856 | + grad_norm_pre_clip_avg=0.1710 | Metrics: + {'align_loss': 0.024304581806063652, + 'recon_loss': 0.07700949907302856, + 'predict_loss': 0.006526302080601454, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18557746708393097, + 'data_time': 0.0008466030121780932, + 'model_time': 1.2200778190162964, + 'grad_norm_pre_clip_avg': 0.1709980010986328, + 'learning_rate': 1.2094118171865371e-05, + 'epoch': 7.06} +04/19 [21:32:58] INFO | >> train_qwenlatent.py:487 + Step 27990 | grad_norm_pre_clip=0.2248 | + grad_norm_pre_clip_avg=0.1834 | Metrics: + {'align_loss': 0.024930275976657867, + 'recon_loss': 0.08272082358598709, + 'predict_loss': 0.007783486973494291, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22481216490268707, + 'data_time': 0.0005869540036655962, + 'model_time': 1.2072146370192058, + 'grad_norm_pre_clip_avg': 0.1833851382136345, + 'learning_rate': 1.208540524931605e-05, + 'epoch': 7.06} +04/19 [21:33:12] INFO | >> train_qwenlatent.py:487 + Step 28000 | grad_norm_pre_clip=0.2512 | + grad_norm_pre_clip_avg=0.1841 | Metrics: + {'align_loss': 0.026298698037862778, + 'recon_loss': 0.12456878274679184, + 'predict_loss': 0.012996688485145569, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25123095512390137, + 'mae_score': 0.00795421686258402, 'data_time': + 0.0009135919972322881, 'model_time': + 1.2258932190015912, 'grad_norm_pre_clip_avg': + 0.18408825993537903, 'learning_rate': + 1.2076692534927524e-05, 'epoch': 7.07} +04/19 [21:33:24] INFO | >> train_qwenlatent.py:487 + Step 28010 | grad_norm_pre_clip=0.1747 | + grad_norm_pre_clip_avg=0.1949 | Metrics: + {'align_loss': 0.02463081106543541, + 'recon_loss': 0.06892331689596176, + 'predict_loss': 0.006104695610702038, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17472940683364868, + 'data_time': 0.0008327760151587427, + 'model_time': 1.2638367580075283, + 'grad_norm_pre_clip_avg': 0.19493988305330276, + 'learning_rate': 1.2067980032946255e-05, + 'epoch': 7.07} +04/19 [21:33:37] INFO | >> train_qwenlatent.py:487 + Step 28020 | grad_norm_pre_clip=0.1694 | + grad_norm_pre_clip_avg=0.1866 | Metrics: + {'align_loss': 0.02520567923784256, + 'recon_loss': 0.09956122189760208, + 'predict_loss': 0.010991650633513927, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1694018393754959, + 'data_time': 0.0006514719862025231, + 'model_time': 1.5434434369963128, + 'grad_norm_pre_clip_avg': 0.1866184577345848, + 'learning_rate': 1.2059267747618624e-05, + 'epoch': 7.07} +04/19 [21:33:50] INFO | >> train_qwenlatent.py:487 + Step 28030 | grad_norm_pre_clip=0.1591 | + grad_norm_pre_clip_avg=0.1698 | Metrics: + {'align_loss': 0.02580832690000534, + 'recon_loss': 0.08722435683012009, + 'predict_loss': 0.009245290420949459, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15911754965782166, + 'data_time': 0.0008962620049715042, + 'model_time': 1.2469933050160762, + 'grad_norm_pre_clip_avg': 0.16976668536663056, + 'learning_rate': 1.2050555683190885e-05, + 'epoch': 7.07} +04/19 [21:34:03] INFO | >> train_qwenlatent.py:487 + Step 28040 | grad_norm_pre_clip=0.1666 | + grad_norm_pre_clip_avg=0.1618 | Metrics: + {'align_loss': 0.026647165417671204, + 'recon_loss': 0.1255495697259903, + 'predict_loss': 0.012430801056325436, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.166643425822258, + 'data_time': 0.0006409019988495857, + 'model_time': 1.2399913440167438, + 'grad_norm_pre_clip_avg': 0.16177840232849122, + 'learning_rate': 1.2041843843909193e-05, + 'epoch': 7.08} +04/19 [21:34:16] INFO | >> train_qwenlatent.py:487 + Step 28050 | grad_norm_pre_clip=0.2266 | + grad_norm_pre_clip_avg=0.1757 | Metrics: + {'align_loss': 0.02540048584342003, + 'recon_loss': 0.09843616187572479, + 'predict_loss': 0.006428023334592581, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22655200958251953, + 'mae_score': 0.008689466253057257, 'data_time': + 0.0006723729893565178, 'model_time': + 1.2408246160193812, 'grad_norm_pre_clip_avg': + 0.17573607116937637, 'learning_rate': + 1.2033132234019595e-05, 'epoch': 7.08} +04/19 [21:34:29] INFO | >> train_qwenlatent.py:487 + Step 28060 | grad_norm_pre_clip=0.1974 | + grad_norm_pre_clip_avg=0.1802 | Metrics: + {'align_loss': 0.023898176848888397, + 'recon_loss': 0.07527632266283035, + 'predict_loss': 0.007658963557332754, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1973598152399063, + 'data_time': 0.0007501679938286543, + 'model_time': 1.225878677010769, + 'grad_norm_pre_clip_avg': 0.18017908036708832, + 'learning_rate': 1.2024420857768024e-05, + 'epoch': 7.08} +04/19 [21:34:41] INFO | >> train_qwenlatent.py:487 + Step 28070 | grad_norm_pre_clip=0.1639 | + grad_norm_pre_clip_avg=0.1728 | Metrics: + {'align_loss': 0.025309517979621887, + 'recon_loss': 0.07348877936601639, + 'predict_loss': 0.006899230647832155, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1639166623353958, + 'data_time': 0.0009499090083409101, + 'model_time': 1.314859815989621, + 'grad_norm_pre_clip_avg': 0.17279098331928253, + 'learning_rate': 1.2015709719400294e-05, + 'epoch': 7.08} +04/19 [21:34:54] INFO | >> train_qwenlatent.py:487 + Step 28080 | grad_norm_pre_clip=0.1358 | + grad_norm_pre_clip_avg=0.1724 | Metrics: + {'align_loss': 0.025124184787273407, + 'recon_loss': 0.09337562322616577, + 'predict_loss': 0.010607149451971054, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13578768074512482, + 'data_time': 0.0008018749940674752, + 'model_time': 1.2475515590049326, + 'grad_norm_pre_clip_avg': 0.17244409769773483, + 'learning_rate': 1.2006998823162117e-05, + 'epoch': 7.09} +04/19 [21:35:06] INFO | >> train_qwenlatent.py:487 + Step 28090 | grad_norm_pre_clip=0.2135 | + grad_norm_pre_clip_avg=0.2065 | Metrics: + {'align_loss': 0.024895187467336655, + 'recon_loss': 0.09316929429769516, + 'predict_loss': 0.007030739914625883, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21353960037231445, + 'data_time': 0.0010147069988306612, + 'model_time': 1.2300335920008365, + 'grad_norm_pre_clip_avg': 0.2064734071493149, + 'learning_rate': 1.1998288173299072e-05, + 'epoch': 7.09} +04/19 [21:35:20] INFO | >> train_qwenlatent.py:487 + Step 28100 | grad_norm_pre_clip=0.1723 | + grad_norm_pre_clip_avg=0.1534 | Metrics: + {'align_loss': 0.024676406756043434, + 'recon_loss': 0.08532722294330597, + 'predict_loss': 0.011799735017120838, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1723339706659317, + 'mae_score': 0.008735224577757689, 'data_time': + 0.0006452179804909974, 'model_time': + 1.250270474003628, 'grad_norm_pre_clip_avg': + 0.15337613075971604, 'learning_rate': + 1.1989577774056622e-05, 'epoch': 7.09} +04/19 [21:35:32] INFO | >> train_qwenlatent.py:487 + Step 28110 | grad_norm_pre_clip=0.2389 | + grad_norm_pre_clip_avg=0.1905 | Metrics: + {'align_loss': 0.02522730827331543, + 'recon_loss': 0.09372042864561081, + 'predict_loss': 0.011947907507419586, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23885668814182281, + 'data_time': 0.0008319829939864576, + 'model_time': 1.2193785849958658, + 'grad_norm_pre_clip_avg': 0.19051107466220857, + 'learning_rate': 1.198086762968011e-05, + 'epoch': 7.09} +04/19 [21:35:45] INFO | >> train_qwenlatent.py:487 + Step 28120 | grad_norm_pre_clip=0.2386 | + grad_norm_pre_clip_avg=0.2085 | Metrics: + {'align_loss': 0.02639136090874672, + 'recon_loss': 0.11386799812316895, + 'predict_loss': 0.00561765069141984, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23857229948043823, + 'data_time': 0.0006168409890960902, + 'model_time': 1.197755027009407, + 'grad_norm_pre_clip_avg': 0.20851241648197175, + 'learning_rate': 1.1972157744414763e-05, + 'epoch': 7.1} +04/19 [21:35:57] INFO | >> train_qwenlatent.py:487 + Step 28130 | grad_norm_pre_clip=0.1459 | + grad_norm_pre_clip_avg=0.1885 | Metrics: + {'align_loss': 0.02611330710351467, + 'recon_loss': 0.0854981318116188, + 'predict_loss': 0.008961108513176441, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1459483951330185, + 'data_time': 0.0007119159854482859, + 'model_time': 1.2282108399958815, + 'grad_norm_pre_clip_avg': 0.188515105843544, + 'learning_rate': 1.1963448122505665e-05, + 'epoch': 7.1} +04/19 [21:36:10] INFO | >> train_qwenlatent.py:487 + Step 28140 | grad_norm_pre_clip=0.1630 | + grad_norm_pre_clip_avg=0.1814 | Metrics: + {'align_loss': 0.026232747361063957, + 'recon_loss': 0.11201389133930206, + 'predict_loss': 0.010072018019855022, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16299274563789368, + 'data_time': 0.001035067019984126, + 'model_time': 1.2324708880041726, + 'grad_norm_pre_clip_avg': 0.18143160045146942, + 'learning_rate': 1.195473876819778e-05, + 'epoch': 7.1} +04/19 [21:36:24] INFO | >> train_qwenlatent.py:487 + Step 28150 | grad_norm_pre_clip=0.2115 | + grad_norm_pre_clip_avg=0.1850 | Metrics: + {'align_loss': 0.026187662035226822, + 'recon_loss': 0.09697136282920837, + 'predict_loss': 0.006469994317740202, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21145296096801758, + 'mae_score': 0.008629541139344912, 'data_time': + 0.0007857700111344457, 'model_time': + 1.2543837249977514, 'grad_norm_pre_clip_avg': + 0.1849566861987114, 'learning_rate': + 1.1946029685735944e-05, 'epoch': 7.1} +04/19 [21:36:37] INFO | >> train_qwenlatent.py:487 + Step 28160 | grad_norm_pre_clip=0.1546 | + grad_norm_pre_clip_avg=0.1953 | Metrics: + {'align_loss': 0.023504268378019333, + 'recon_loss': 0.09550550580024719, + 'predict_loss': 0.00768256327137351, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15459083020687103, + 'data_time': 0.0006081979954615235, + 'model_time': 1.2373907229921315, + 'grad_norm_pre_clip_avg': 0.19529427140951156, + 'learning_rate': 1.1937320879364857e-05, + 'epoch': 7.11} +04/19 [21:36:50] INFO | >> train_qwenlatent.py:487 + Step 28170 | grad_norm_pre_clip=0.1723 | + grad_norm_pre_clip_avg=0.1670 | Metrics: + {'align_loss': 0.026133215054869652, + 'recon_loss': 0.12784838676452637, + 'predict_loss': 0.009957009926438332, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17227067053318024, + 'data_time': 0.0009032640082295984, + 'model_time': 1.2706539680075366, + 'grad_norm_pre_clip_avg': 0.16700066477060319, + 'learning_rate': 1.1928612353329093e-05, + 'epoch': 7.11} +04/19 [21:37:02] INFO | >> train_qwenlatent.py:487 + Step 28180 | grad_norm_pre_clip=0.1679 | + grad_norm_pre_clip_avg=0.1690 | Metrics: + {'align_loss': 0.025781244039535522, + 'recon_loss': 0.08219224214553833, + 'predict_loss': 0.00756971025839448, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16789959371089935, + 'data_time': 0.0009050619846675545, + 'model_time': 1.247342088026926, + 'grad_norm_pre_clip_avg': 0.16899047195911407, + 'learning_rate': 1.1919904111873068e-05, + 'epoch': 7.11} +04/19 [21:37:15] INFO | >> train_qwenlatent.py:487 + Step 28190 | grad_norm_pre_clip=0.1716 | + grad_norm_pre_clip_avg=0.1559 | Metrics: + {'align_loss': 0.02507641911506653, + 'recon_loss': 0.10497720539569855, + 'predict_loss': 0.00911245308816433, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17159900069236755, + 'data_time': 0.0009257679921574891, + 'model_time': 1.2371494099788833, + 'grad_norm_pre_clip_avg': 0.1558670088648796, + 'learning_rate': 1.1911196159241082e-05, + 'epoch': 7.11} +04/19 [21:37:28] INFO | >> train_qwenlatent.py:487 + Step 28200 | grad_norm_pre_clip=0.2088 | + grad_norm_pre_clip_avg=0.1983 | Metrics: + {'align_loss': 0.025721661746501923, + 'recon_loss': 0.0861101895570755, + 'predict_loss': 0.008451738394796848, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20878510177135468, + 'mae_score': 0.008474421286368156, 'data_time': + 0.00066181700094603, 'model_time': + 1.1902831829793286, 'grad_norm_pre_clip_avg': + 0.19827157855033875, 'learning_rate': + 1.1902488499677279e-05, 'epoch': 7.12} +04/19 [21:37:40] INFO | >> train_qwenlatent.py:487 + Step 28210 | grad_norm_pre_clip=0.1783 | + grad_norm_pre_clip_avg=0.2032 | Metrics: + {'align_loss': 0.024775579571723938, + 'recon_loss': 0.08352898806333542, + 'predict_loss': 0.010503502562642097, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17828369140625, + 'data_time': 0.0009147140081040561, + 'model_time': 1.2559609710006043, + 'grad_norm_pre_clip_avg': 0.20317453444004058, + 'learning_rate': 1.1893781137425676e-05, + 'epoch': 7.12} +04/19 [21:37:53] INFO | >> train_qwenlatent.py:487 + Step 28220 | grad_norm_pre_clip=0.2285 | + grad_norm_pre_clip_avg=0.2083 | Metrics: + {'align_loss': 0.025920625776052475, + 'recon_loss': 0.09837136417627335, + 'predict_loss': 0.01037281658500433, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22851096093654633, + 'data_time': 0.0008768209954723716, + 'model_time': 1.2873627219814807, + 'grad_norm_pre_clip_avg': 0.20830106735229492, + 'learning_rate': 1.188507407673013e-05, + 'epoch': 7.12} +04/19 [21:38:06] INFO | >> train_qwenlatent.py:487 + Step 28230 | grad_norm_pre_clip=0.1591 | + grad_norm_pre_clip_avg=0.1706 | Metrics: + {'align_loss': 0.02566523291170597, + 'recon_loss': 0.10042281448841095, + 'predict_loss': 0.005131310783326626, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15907689929008484, + 'data_time': 0.0009032499801833183, + 'model_time': 1.229703482997138, + 'grad_norm_pre_clip_avg': 0.17056215256452562, + 'learning_rate': 1.1876367321834353e-05, + 'epoch': 7.12} +04/19 [21:38:18] INFO | >> train_qwenlatent.py:487 + Step 28240 | grad_norm_pre_clip=0.1388 | + grad_norm_pre_clip_avg=0.1694 | Metrics: + {'align_loss': 0.02469732239842415, + 'recon_loss': 0.09364339709281921, + 'predict_loss': 0.009491650387644768, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13883590698242188, + 'data_time': 0.0006845910102128983, + 'model_time': 1.241487715014955, + 'grad_norm_pre_clip_avg': 0.169374543428421, + 'learning_rate': 1.1867660876981919e-05, + 'epoch': 7.13} +04/19 [21:38:31] INFO | >> train_qwenlatent.py:487 + Step 28250 | grad_norm_pre_clip=0.1523 | + grad_norm_pre_clip_avg=0.1601 | Metrics: + {'align_loss': 0.0258251391351223, + 'recon_loss': 0.07750342786312103, + 'predict_loss': 0.005039825104176998, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15227223932743073, + 'mae_score': 0.008096112431706609, 'data_time': + 0.0006190149870235473, 'model_time': + 1.232022993004648, 'grad_norm_pre_clip_avg': + 0.16014395505189896, 'learning_rate': + 1.185895474641624e-05, 'epoch': 7.13} +04/19 [21:38:44] INFO | >> train_qwenlatent.py:487 + Step 28260 | grad_norm_pre_clip=0.3030 | + grad_norm_pre_clip_avg=0.2033 | Metrics: + {'align_loss': 0.025454029440879822, + 'recon_loss': 0.07786904275417328, + 'predict_loss': 0.011803368106484413, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3030073046684265, + 'data_time': 0.0007947720005176961, + 'model_time': 1.2452610390027985, + 'grad_norm_pre_clip_avg': 0.20334952622652053, + 'learning_rate': 1.1850248934380579e-05, + 'epoch': 7.13} +04/19 [21:38:57] INFO | >> train_qwenlatent.py:487 + Step 28270 | grad_norm_pre_clip=0.2143 | + grad_norm_pre_clip_avg=0.2119 | Metrics: + {'align_loss': 0.026364099234342575, + 'recon_loss': 0.11784327030181885, + 'predict_loss': 0.010664304718375206, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21431554853916168, + 'data_time': 0.0006456199917010963, + 'model_time': 1.3052328680059873, + 'grad_norm_pre_clip_avg': 0.21185193359851837, + 'learning_rate': 1.1841543445118039e-05, + 'epoch': 7.13} +04/19 [21:39:09] INFO | >> train_qwenlatent.py:487 + Step 28280 | grad_norm_pre_clip=0.1759 | + grad_norm_pre_clip_avg=0.1791 | Metrics: + {'align_loss': 0.024447470903396606, + 'recon_loss': 0.08789122104644775, + 'predict_loss': 0.009382042102515697, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17592394351959229, + 'data_time': 0.0006452880043070763, + 'model_time': 1.3194150700001046, + 'grad_norm_pre_clip_avg': 0.17907425314188002, + 'learning_rate': 1.1832838282871576e-05, + 'epoch': 7.14} +04/19 [21:39:23] INFO | >> train_qwenlatent.py:487 + Step 28290 | grad_norm_pre_clip=0.1512 | + grad_norm_pre_clip_avg=0.1645 | Metrics: + {'align_loss': 0.024624112993478775, + 'recon_loss': 0.07742349803447723, + 'predict_loss': 0.006812328472733498, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15117228031158447, + 'data_time': 0.0006461740122176707, + 'model_time': 1.2579954230168369, + 'grad_norm_pre_clip_avg': 0.16451270878314972, + 'learning_rate': 1.1824133451883977e-05, + 'epoch': 7.14} +04/19 [21:39:36] INFO | >> train_qwenlatent.py:487 + Step 28300 | grad_norm_pre_clip=0.1490 | + grad_norm_pre_clip_avg=0.1739 | Metrics: + {'align_loss': 0.026538249105215073, + 'recon_loss': 0.13160614669322968, + 'predict_loss': 0.009141909889876842, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14902567863464355, + 'mae_score': 0.010325309822151253, 'data_time': + 0.0014267050137277693, 'model_time': + 1.5966590259922668, 'grad_norm_pre_clip_avg': + 0.17390501946210862, 'learning_rate': + 1.1815428956397867e-05, 'epoch': 7.14} +04/19 [21:39:49] INFO | >> train_qwenlatent.py:487 + Step 28310 | grad_norm_pre_clip=0.1989 | + grad_norm_pre_clip_avg=0.2039 | Metrics: + {'align_loss': 0.02555544674396515, + 'recon_loss': 0.09095088392496109, + 'predict_loss': 0.007310888264328241, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19885477423667908, + 'data_time': 0.0010150839807465672, + 'model_time': 1.2162655360007193, + 'grad_norm_pre_clip_avg': 0.20388083457946776, + 'learning_rate': 1.180672480065572e-05, + 'epoch': 7.14} +04/19 [21:40:02] INFO | >> train_qwenlatent.py:487 + Step 28320 | grad_norm_pre_clip=0.1770 | + grad_norm_pre_clip_avg=0.2047 | Metrics: + {'align_loss': 0.026511209085583687, + 'recon_loss': 0.0989101454615593, + 'predict_loss': 0.006725361105054617, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17695000767707825, + 'data_time': 0.0011055089998990297, + 'model_time': 1.2639866569952574, + 'grad_norm_pre_clip_avg': 0.20467139333486556, + 'learning_rate': 1.179802098889983e-05, + 'epoch': 7.15} +04/19 [21:40:14] INFO | >> train_qwenlatent.py:487 + Step 28330 | grad_norm_pre_clip=0.1412 | + grad_norm_pre_clip_avg=0.1635 | Metrics: + {'align_loss': 0.02530514821410179, + 'recon_loss': 0.10046376287937164, + 'predict_loss': 0.012483007274568081, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1412380337715149, + 'data_time': 0.0009073100227396935, + 'model_time': 1.2155970519816037, + 'grad_norm_pre_clip_avg': 0.16354428678750993, + 'learning_rate': 1.1789317525372328e-05, + 'epoch': 7.15} +04/19 [21:40:27] INFO | >> train_qwenlatent.py:487 + Step 28340 | grad_norm_pre_clip=0.2380 | + grad_norm_pre_clip_avg=0.1725 | Metrics: + {'align_loss': 0.025963811203837395, + 'recon_loss': 0.12546047568321228, + 'predict_loss': 0.01021452248096466, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2380467802286148, + 'data_time': 0.0009672800078988075, + 'model_time': 1.239123370993184, + 'grad_norm_pre_clip_avg': 0.17246540561318396, + 'learning_rate': 1.1780614414315182e-05, + 'epoch': 7.15} +04/19 [21:40:40] INFO | >> train_qwenlatent.py:487 + Step 28350 | grad_norm_pre_clip=0.1806 | + grad_norm_pre_clip_avg=0.1967 | Metrics: + {'align_loss': 0.023968158289790154, + 'recon_loss': 0.05247636139392853, + 'predict_loss': 0.006250899750739336, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1806086003780365, + 'mae_score': 0.009605654295500334, 'data_time': + 0.0009917849965859205, 'model_time': + 1.2411885249894112, 'grad_norm_pre_clip_avg': + 0.19670647084712983, 'learning_rate': + 1.1771911659970181e-05, 'epoch': 7.15} +04/19 [21:40:53] INFO | >> train_qwenlatent.py:487 + Step 28360 | grad_norm_pre_clip=0.1284 | + grad_norm_pre_clip_avg=0.1745 | Metrics: + {'align_loss': 0.02533860318362713, + 'recon_loss': 0.11958789080381393, + 'predict_loss': 0.006892603822052479, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12836718559265137, + 'data_time': 0.0006594040023628622, + 'model_time': 1.5008460950048175, + 'grad_norm_pre_clip_avg': 0.1744519367814064, + 'learning_rate': 1.176320926657894e-05, + 'epoch': 7.16} +04/19 [21:41:05] INFO | >> train_qwenlatent.py:487 + Step 28370 | grad_norm_pre_clip=0.1336 | + grad_norm_pre_clip_avg=0.1814 | Metrics: + {'align_loss': 0.02419605478644371, + 'recon_loss': 0.08047322183847427, + 'predict_loss': 0.007134385406970978, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13357660174369812, + 'data_time': 0.0007387429941445589, + 'model_time': 1.192182642989792, + 'grad_norm_pre_clip_avg': 0.18141651004552842, + 'learning_rate': 1.1754507238382895e-05, + 'epoch': 7.16} +04/19 [21:41:18] INFO | >> train_qwenlatent.py:487 + Step 28380 | grad_norm_pre_clip=0.1605 | + grad_norm_pre_clip_avg=0.1615 | Metrics: + {'align_loss': 0.026143783703446388, + 'recon_loss': 0.10955587774515152, + 'predict_loss': 0.015908582136034966, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16049069166183472, + 'data_time': 0.0007056639879010618, + 'model_time': 1.3276403589989059, + 'grad_norm_pre_clip_avg': 0.16154325008392334, + 'learning_rate': 1.1745805579623317e-05, + 'epoch': 7.16} +04/19 [21:41:30] INFO | >> train_qwenlatent.py:487 + Step 28390 | grad_norm_pre_clip=0.2049 | + grad_norm_pre_clip_avg=0.2020 | Metrics: + {'align_loss': 0.025804441422224045, + 'recon_loss': 0.14149965345859528, + 'predict_loss': 0.013129794970154762, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20491841435432434, + 'data_time': 0.000888059992576018, + 'model_time': 1.2614293759979773, + 'grad_norm_pre_clip_avg': 0.20196571946144104, + 'learning_rate': 1.1737104294541287e-05, + 'epoch': 7.16} +04/19 [21:41:43] INFO | >> train_qwenlatent.py:487 + Step 28400 | grad_norm_pre_clip=0.1637 | + grad_norm_pre_clip_avg=0.1660 | Metrics: + {'align_loss': 0.02609630487859249, + 'recon_loss': 0.10134371370077133, + 'predict_loss': 0.008700934238731861, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16367216408252716, + 'mae_score': 0.006836288039748733, 'data_time': + 0.000632263021543622, 'model_time': + 1.201270196994301, 'grad_norm_pre_clip_avg': + 0.1660357803106308, 'learning_rate': + 1.17284033873777e-05, 'epoch': 7.17} +04/19 [21:41:56] INFO | >> train_qwenlatent.py:487 + Step 28410 | grad_norm_pre_clip=0.1340 | + grad_norm_pre_clip_avg=0.1666 | Metrics: + {'align_loss': 0.02638816460967064, + 'recon_loss': 0.09891407936811447, + 'predict_loss': 0.013086383230984211, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13397832214832306, + 'data_time': 0.0009682829841040075, + 'model_time': 1.2203433050017338, + 'grad_norm_pre_clip_avg': 0.16655288487672806, + 'learning_rate': 1.1719702862373281e-05, + 'epoch': 7.17} +04/19 [21:42:09] INFO | >> train_qwenlatent.py:487 + Step 28420 | grad_norm_pre_clip=0.1568 | + grad_norm_pre_clip_avg=0.1832 | Metrics: + {'align_loss': 0.025791212916374207, + 'recon_loss': 0.1162109300494194, + 'predict_loss': 0.007161321118474007, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1568448692560196, + 'data_time': 0.0007826949877198786, + 'model_time': 1.2443539749947377, + 'grad_norm_pre_clip_avg': 0.18321293741464614, + 'learning_rate': 1.171100272376855e-05, + 'epoch': 7.17} +04/19 [21:42:22] INFO | >> train_qwenlatent.py:487 + Step 28430 | grad_norm_pre_clip=0.1522 | + grad_norm_pre_clip_avg=0.1880 | Metrics: + {'align_loss': 0.02504073828458786, + 'recon_loss': 0.08578012138605118, + 'predict_loss': 0.00630648760125041, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1521683633327484, + 'data_time': 0.0010388280206825584, + 'model_time': 1.2537122570211068, + 'grad_norm_pre_clip_avg': 0.18803346306085586, + 'learning_rate': 1.170230297580386e-05, + 'epoch': 7.17} +04/19 [21:42:35] INFO | >> train_qwenlatent.py:487 + Step 28440 | grad_norm_pre_clip=0.1683 | + grad_norm_pre_clip_avg=0.1855 | Metrics: + {'align_loss': 0.02553471550345421, + 'recon_loss': 0.1063123419880867, + 'predict_loss': 0.008221310563385487, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16827349364757538, + 'data_time': 0.0007573690090794116, + 'model_time': 1.4950226130022202, + 'grad_norm_pre_clip_avg': 0.1855454608798027, + 'learning_rate': 1.1693603622719359e-05, + 'epoch': 7.18} +04/19 [21:42:48] INFO | >> train_qwenlatent.py:487 + Step 28450 | grad_norm_pre_clip=0.1358 | + grad_norm_pre_clip_avg=0.1743 | Metrics: + {'align_loss': 0.024353062734007835, + 'recon_loss': 0.05434931442141533, + 'predict_loss': 0.004918232094496489, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13583670556545258, + 'mae_score': 0.007683668909846126, 'data_time': + 0.0006655360048171133, 'model_time': + 1.250481902010506, 'grad_norm_pre_clip_avg': + 0.1743302524089813, 'learning_rate': + 1.1684904668755002e-05, 'epoch': 7.18} +04/19 [21:43:00] INFO | >> train_qwenlatent.py:487 + Step 28460 | grad_norm_pre_clip=0.1374 | + grad_norm_pre_clip_avg=0.1657 | Metrics: + {'align_loss': 0.02627197653055191, + 'recon_loss': 0.10702631622552872, + 'predict_loss': 0.007870340719819069, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13743780553340912, + 'data_time': 0.0007368840160779655, + 'model_time': 1.207018913992215, + 'grad_norm_pre_clip_avg': 0.16566462069749832, + 'learning_rate': 1.1676206118150552e-05, + 'epoch': 7.18} +04/19 [21:43:13] INFO | >> train_qwenlatent.py:487 + Step 28470 | grad_norm_pre_clip=0.2198 | + grad_norm_pre_clip_avg=0.2012 | Metrics: + {'align_loss': 0.02476463094353676, + 'recon_loss': 0.08419064432382584, + 'predict_loss': 0.004983654245734215, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2197703719139099, + 'data_time': 0.0009213860030286014, + 'model_time': 1.2850624620041344, + 'grad_norm_pre_clip_avg': 0.20118309408426285, + 'learning_rate': 1.1667507975145585e-05, + 'epoch': 7.18} +04/19 [21:43:25] INFO | >> train_qwenlatent.py:487 + Step 28480 | grad_norm_pre_clip=0.1717 | + grad_norm_pre_clip_avg=0.1882 | Metrics: + {'align_loss': 0.0225958414375782, + 'recon_loss': 0.1015671119093895, + 'predict_loss': 0.012706012465059757, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17166103422641754, + 'data_time': 0.0008977699908427894, + 'model_time': 1.2521690999856219, + 'grad_norm_pre_clip_avg': 0.18824046105146408, + 'learning_rate': 1.1658810243979463e-05, + 'epoch': 7.19} +04/19 [21:43:38] INFO | >> train_qwenlatent.py:487 + Step 28490 | grad_norm_pre_clip=0.1857 | + grad_norm_pre_clip_avg=0.1914 | Metrics: + {'align_loss': 0.025962337851524353, + 'recon_loss': 0.10814308375120163, + 'predict_loss': 0.010037058964371681, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18570224940776825, + 'data_time': 0.0008130939968395978, + 'model_time': 1.2172022989834659, + 'grad_norm_pre_clip_avg': 0.19143398255109786, + 'learning_rate': 1.1650112928891358e-05, + 'epoch': 7.19} +04/19 [21:43:51] INFO | >> train_qwenlatent.py:487 + Step 28500 | grad_norm_pre_clip=0.2280 | + grad_norm_pre_clip_avg=0.1815 | Metrics: + {'align_loss': 0.025528358295559883, + 'recon_loss': 0.0691707581281662, + 'predict_loss': 0.007195222191512585, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22795112431049347, + 'mae_score': 0.011270444028012387, 'data_time': + 0.0007012579881120473, 'model_time': + 1.2201892059820239, 'grad_norm_pre_clip_avg': + 0.18148508071899414, 'learning_rate': + 1.1641416034120236e-05, 'epoch': 7.19} +04/19 [21:44:04] INFO | >> train_qwenlatent.py:487 + Step 28510 | grad_norm_pre_clip=0.1762 | + grad_norm_pre_clip_avg=0.1648 | Metrics: + {'align_loss': 0.02434811368584633, + 'recon_loss': 0.10970085114240646, + 'predict_loss': 0.013722393661737442, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1762090027332306, + 'data_time': 0.00085136599955149, 'model_time': + 1.2974167090142146, 'grad_norm_pre_clip_avg': + 0.1648387908935547, 'learning_rate': + 1.1632719563904856e-05, 'epoch': 7.19} +04/19 [21:44:17] INFO | >> train_qwenlatent.py:487 + Step 28520 | grad_norm_pre_clip=0.1256 | + grad_norm_pre_clip_avg=0.1965 | Metrics: + {'align_loss': 0.02429066225886345, + 'recon_loss': 0.11917450278997421, + 'predict_loss': 0.009218813851475716, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12560133635997772, + 'data_time': 0.0010799919837154448, + 'model_time': 1.2872815339942463, + 'grad_norm_pre_clip_avg': 0.19652554094791413, + 'learning_rate': 1.1624023522483778e-05, + 'epoch': 7.2} +04/19 [21:44:29] INFO | >> train_qwenlatent.py:487 + Step 28530 | grad_norm_pre_clip=0.1915 | + grad_norm_pre_clip_avg=0.1997 | Metrics: + {'align_loss': 0.025016311556100845, + 'recon_loss': 0.10920467972755432, + 'predict_loss': 0.011709204874932766, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1914745569229126, + 'data_time': 0.0008614739926997572, + 'model_time': 1.2199290469870903, + 'grad_norm_pre_clip_avg': 0.19970054179430008, + 'learning_rate': 1.1615327914095335e-05, + 'epoch': 7.2} +04/19 [21:44:41] INFO | >> train_qwenlatent.py:487 + Step 28540 | grad_norm_pre_clip=0.2025 | + grad_norm_pre_clip_avg=0.1657 | Metrics: + {'align_loss': 0.026545457541942596, + 'recon_loss': 0.10153567045927048, + 'predict_loss': 0.008131795562803745, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2024827003479004, + 'data_time': 0.0009202770015690476, + 'model_time': 1.2702903150056954, + 'grad_norm_pre_clip_avg': 0.16572097986936568, + 'learning_rate': 1.160663274297767e-05, + 'epoch': 7.2} +04/19 [21:44:55] INFO | >> train_qwenlatent.py:487 + Step 28550 | grad_norm_pre_clip=0.1605 | + grad_norm_pre_clip_avg=0.1873 | Metrics: + {'align_loss': 0.024044863879680634, + 'recon_loss': 0.0759492889046669, + 'predict_loss': 0.006939031649380922, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16046170890331268, + 'mae_score': 0.009231404141262846, 'data_time': + 0.0009613540023565292, 'model_time': + 1.250355263997335, 'grad_norm_pre_clip_avg': + 0.1872926890850067, 'learning_rate': + 1.1597938013368703e-05, 'epoch': 7.2} +04/19 [21:45:07] INFO | >> train_qwenlatent.py:487 + Step 28560 | grad_norm_pre_clip=0.2015 | + grad_norm_pre_clip_avg=0.1946 | Metrics: + {'align_loss': 0.02600148692727089, + 'recon_loss': 0.10760632902383804, + 'predict_loss': 0.01144036278128624, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20147447288036346, + 'data_time': 0.001010911975754425, + 'model_time': 1.24671745399246, + 'grad_norm_pre_clip_avg': 0.19459159672260284, + 'learning_rate': 1.1589243729506135e-05, + 'epoch': 7.21} +04/19 [21:45:20] INFO | >> train_qwenlatent.py:487 + Step 28570 | grad_norm_pre_clip=0.1970 | + grad_norm_pre_clip_avg=0.1892 | Metrics: + {'align_loss': 0.025438208132982254, + 'recon_loss': 0.09057372808456421, + 'predict_loss': 0.00967774074524641, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19697526097297668, + 'data_time': 0.0006741760007571429, + 'model_time': 1.2265532640158199, + 'grad_norm_pre_clip_avg': 0.18919438272714614, + 'learning_rate': 1.1580549895627459e-05, + 'epoch': 7.21} +04/19 [21:45:33] INFO | >> train_qwenlatent.py:487 + Step 28580 | grad_norm_pre_clip=0.1704 | + grad_norm_pre_clip_avg=0.1904 | Metrics: + {'align_loss': 0.026098810136318207, + 'recon_loss': 0.08192043006420135, + 'predict_loss': 0.004847156349569559, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17042620480060577, + 'data_time': 0.0008265920041594654, + 'model_time': 1.2166537669836544, + 'grad_norm_pre_clip_avg': 0.19035491198301316, + 'learning_rate': 1.1571856515969942e-05, + 'epoch': 7.21} +04/19 [21:45:45] INFO | >> train_qwenlatent.py:487 + Step 28590 | grad_norm_pre_clip=0.1949 | + grad_norm_pre_clip_avg=0.1804 | Metrics: + {'align_loss': 0.02510344237089157, + 'recon_loss': 0.05718069523572922, + 'predict_loss': 0.0071235960349440575, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1948826014995575, + 'data_time': 0.0006285079871304333, + 'model_time': 1.2385660049912985, + 'grad_norm_pre_clip_avg': 0.18040644526481628, + 'learning_rate': 1.156316359477063e-05, + 'epoch': 7.21} +04/19 [21:45:59] INFO | >> train_qwenlatent.py:487 + Step 28600 | grad_norm_pre_clip=0.1924 | + grad_norm_pre_clip_avg=0.1786 | Metrics: + {'align_loss': 0.025066591799259186, + 'recon_loss': 0.11626868695020676, + 'predict_loss': 0.015545647591352463, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1924314945936203, + 'mae_score': 0.009305256336658924, 'data_time': + 0.0008680640021339059, 'model_time': + 1.2334258359915111, 'grad_norm_pre_clip_avg': + 0.1786305844783783, 'learning_rate': + 1.1554471136266352e-05, 'epoch': 7.22} +04/19 [21:46:11] INFO | >> train_qwenlatent.py:487 + Step 28610 | grad_norm_pre_clip=0.1466 | + grad_norm_pre_clip_avg=0.1761 | Metrics: + {'align_loss': 0.025059089064598083, + 'recon_loss': 0.08727116137742996, + 'predict_loss': 0.0071776616387069225, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14659571647644043, + 'data_time': 0.0010680639825295657, + 'model_time': 1.2642330620146822, + 'grad_norm_pre_clip_avg': 0.176131933927536, + 'learning_rate': 1.1545779144693707e-05, + 'epoch': 7.22} +04/19 [21:46:24] INFO | >> train_qwenlatent.py:487 + Step 28620 | grad_norm_pre_clip=0.1797 | + grad_norm_pre_clip_avg=0.2027 | Metrics: + {'align_loss': 0.025158386677503586, + 'recon_loss': 0.07317114621400833, + 'predict_loss': 0.00796545296907425, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17965970933437347, + 'data_time': 0.001048505015205592, + 'model_time': 1.2429532859823667, + 'grad_norm_pre_clip_avg': 0.20268767476081848, + 'learning_rate': 1.153708762428906e-05, + 'epoch': 7.22} +04/19 [21:46:37] INFO | >> train_qwenlatent.py:487 + Step 28630 | grad_norm_pre_clip=0.1305 | + grad_norm_pre_clip_avg=0.1704 | Metrics: + {'align_loss': 0.02535473182797432, + 'recon_loss': 0.09103011339902878, + 'predict_loss': 0.006696472410112619, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13045275211334229, + 'data_time': 0.0006435730028897524, + 'model_time': 1.1913634410011582, + 'grad_norm_pre_clip_avg': 0.17044128775596618, + 'learning_rate': 1.1528396579288554e-05, + 'epoch': 7.22} +04/19 [21:46:49] INFO | >> train_qwenlatent.py:487 + Step 28640 | grad_norm_pre_clip=0.1635 | + grad_norm_pre_clip_avg=0.1762 | Metrics: + {'align_loss': 0.025873757898807526, + 'recon_loss': 0.11679592728614807, + 'predict_loss': 0.010445849969983101, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16353020071983337, + 'data_time': 0.0006722870166413486, + 'model_time': 1.2458817129954696, + 'grad_norm_pre_clip_avg': 0.17623687386512757, + 'learning_rate': 1.15197060139281e-05, 'epoch': + 7.23} +04/19 [21:47:02] INFO | >> train_qwenlatent.py:487 + Step 28650 | grad_norm_pre_clip=0.1711 | + grad_norm_pre_clip_avg=0.1857 | Metrics: + {'align_loss': 0.024646367877721786, + 'recon_loss': 0.0838213637471199, + 'predict_loss': 0.006262654438614845, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1710851639509201, + 'mae_score': 0.009683918308567357, 'data_time': + 0.0008243429765570909, 'model_time': + 1.2186535270011518, 'grad_norm_pre_clip_avg': + 0.18567855060100555, 'learning_rate': + 1.1511015932443377e-05, 'epoch': 7.23} +04/19 [21:47:15] INFO | >> train_qwenlatent.py:487 + Step 28660 | grad_norm_pre_clip=0.2121 | + grad_norm_pre_clip_avg=0.1945 | Metrics: + {'align_loss': 0.02534431777894497, + 'recon_loss': 0.09981866925954819, + 'predict_loss': 0.012195412069559097, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21207399666309357, + 'data_time': 0.0006594180013053119, + 'model_time': 1.211165646003792, + 'grad_norm_pre_clip_avg': 0.19446279108524323, + 'learning_rate': 1.1502326339069817e-05, + 'epoch': 7.23} +04/19 [21:47:27] INFO | >> train_qwenlatent.py:487 + Step 28670 | grad_norm_pre_clip=0.1474 | + grad_norm_pre_clip_avg=0.1829 | Metrics: + {'align_loss': 0.026322800666093826, + 'recon_loss': 0.10208150744438171, + 'predict_loss': 0.00642694067209959, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14743946492671967, + 'data_time': 0.0006743090052623302, + 'model_time': 1.2429167249938473, + 'grad_norm_pre_clip_avg': 0.18293630331754684, + 'learning_rate': 1.1493637238042634e-05, + 'epoch': 7.23} +04/19 [21:47:40] INFO | >> train_qwenlatent.py:487 + Step 28680 | grad_norm_pre_clip=0.1669 | + grad_norm_pre_clip_avg=0.1659 | Metrics: + {'align_loss': 0.024581804871559143, + 'recon_loss': 0.07801488041877747, + 'predict_loss': 0.005829531233757734, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16691423952579498, + 'data_time': 0.0006865609902888536, + 'model_time': 1.2577326649916358, + 'grad_norm_pre_clip_avg': 0.16586829870939254, + 'learning_rate': 1.148494863359678e-05, + 'epoch': 7.24} +04/19 [21:47:53] INFO | >> train_qwenlatent.py:487 + Step 28690 | grad_norm_pre_clip=0.1527 | + grad_norm_pre_clip_avg=0.1546 | Metrics: + {'align_loss': 0.025440353900194168, + 'recon_loss': 0.064471535384655, + 'predict_loss': 0.005308527033776045, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1526656299829483, + 'data_time': 0.0007890530105214566, + 'model_time': 1.5766688270086888, + 'grad_norm_pre_clip_avg': 0.15464775189757346, + 'learning_rate': 1.147626052996698e-05, + 'epoch': 7.24} +04/19 [21:48:06] INFO | >> train_qwenlatent.py:487 + Step 28700 | grad_norm_pre_clip=0.1900 | + grad_norm_pre_clip_avg=0.1652 | Metrics: + {'align_loss': 0.025579191744327545, + 'recon_loss': 0.08871304988861084, + 'predict_loss': 0.011578535661101341, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19001895189285278, + 'mae_score': 0.008639183559933225, 'data_time': + 0.0010584749979898334, 'model_time': + 1.547635971015552, 'grad_norm_pre_clip_avg': + 0.16521501988172532, 'learning_rate': + 1.1467572931387714e-05, 'epoch': 7.24} +04/19 [21:48:18] INFO | >> train_qwenlatent.py:487 + Step 28710 | grad_norm_pre_clip=0.1726 | + grad_norm_pre_clip_avg=0.1972 | Metrics: + {'align_loss': 0.02526729367673397, + 'recon_loss': 0.10000034421682358, + 'predict_loss': 0.014831931330263615, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17259405553340912, + 'data_time': 0.0008880890090949833, + 'model_time': 1.2152550379978493, + 'grad_norm_pre_clip_avg': 0.1972154200077057, + 'learning_rate': 1.1458885842093202e-05, + 'epoch': 7.24} +04/19 [21:48:31] INFO | >> train_qwenlatent.py:487 + Step 28720 | grad_norm_pre_clip=0.1706 | + grad_norm_pre_clip_avg=0.1757 | Metrics: + {'align_loss': 0.025716913864016533, + 'recon_loss': 0.10855276882648468, + 'predict_loss': 0.011795785278081894, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17061704397201538, + 'data_time': 0.0006790640181861818, + 'model_time': 1.2301837370032445, + 'grad_norm_pre_clip_avg': 0.1756929025053978, + 'learning_rate': 1.1450199266317431e-05, + 'epoch': 7.25} +04/19 [21:48:43] INFO | >> train_qwenlatent.py:487 + Step 28730 | grad_norm_pre_clip=0.1759 | + grad_norm_pre_clip_avg=0.1970 | Metrics: + {'align_loss': 0.025351595133543015, + 'recon_loss': 0.10749948024749756, + 'predict_loss': 0.011774810962378979, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1758570373058319, + 'data_time': 0.0011140370042994618, + 'model_time': 1.2443925660045352, + 'grad_norm_pre_clip_avg': 0.19698153883218766, + 'learning_rate': 1.1441513208294134e-05, + 'epoch': 7.25} +04/19 [21:48:56] INFO | >> train_qwenlatent.py:487 + Step 28740 | grad_norm_pre_clip=0.1777 | + grad_norm_pre_clip_avg=0.1712 | Metrics: + {'align_loss': 0.02619970589876175, + 'recon_loss': 0.08933781832456589, + 'predict_loss': 0.006633748300373554, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17772918939590454, + 'data_time': 0.0005902940174564719, + 'model_time': 1.2621366560051683, + 'grad_norm_pre_clip_avg': 0.17119225487113, + 'learning_rate': 1.1432827672256791e-05, + 'epoch': 7.25} +04/19 [21:49:09] INFO | >> train_qwenlatent.py:487 + Step 28750 | grad_norm_pre_clip=0.1892 | + grad_norm_pre_clip_avg=0.1685 | Metrics: + {'align_loss': 0.024708528071641922, + 'recon_loss': 0.09611401706933975, + 'predict_loss': 0.008458520285785198, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18920372426509857, + 'mae_score': 0.008518468152295362, 'data_time': + 0.000663832004647702, 'model_time': + 1.215875338006299, 'grad_norm_pre_clip_avg': + 0.1684862941503525, 'learning_rate': + 1.1424142662438618e-05, 'epoch': 7.25} +04/19 [21:49:22] INFO | >> train_qwenlatent.py:487 + Step 28760 | grad_norm_pre_clip=0.2491 | + grad_norm_pre_clip_avg=0.1933 | Metrics: + {'align_loss': 0.02574983239173889, + 'recon_loss': 0.11464017629623413, + 'predict_loss': 0.01652372255921364, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24907110631465912, + 'data_time': 0.0008290900150313973, + 'model_time': 1.2637159680016339, + 'grad_norm_pre_clip_avg': 0.19332839846611022, + 'learning_rate': 1.1415458183072596e-05, + 'epoch': 7.26} +04/19 [21:49:34] INFO | >> train_qwenlatent.py:487 + Step 28770 | grad_norm_pre_clip=0.1600 | + grad_norm_pre_clip_avg=0.2098 | Metrics: + {'align_loss': 0.026029352098703384, + 'recon_loss': 0.09449686855077744, + 'predict_loss': 0.006380582228302956, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16001713275909424, + 'data_time': 0.000671752990456298, + 'model_time': 1.4481833559984807, + 'grad_norm_pre_clip_avg': 0.2098142221570015, + 'learning_rate': 1.1406774238391426e-05, + 'epoch': 7.26} +04/19 [21:49:47] INFO | >> train_qwenlatent.py:487 + Step 28780 | grad_norm_pre_clip=0.1979 | + grad_norm_pre_clip_avg=0.1727 | Metrics: + {'align_loss': 0.02508898824453354, + 'recon_loss': 0.10181893408298492, + 'predict_loss': 0.014844657853245735, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19791476428508759, + 'data_time': 0.0006832660001236945, + 'model_time': 1.473559734004084, + 'grad_norm_pre_clip_avg': 0.17270257472991943, + 'learning_rate': 1.1398090832627559e-05, + 'epoch': 7.26} +04/19 [21:50:00] INFO | >> train_qwenlatent.py:487 + Step 28790 | grad_norm_pre_clip=0.1624 | + grad_norm_pre_clip_avg=0.1719 | Metrics: + {'align_loss': 0.025044217705726624, + 'recon_loss': 0.10919426381587982, + 'predict_loss': 0.009523199871182442, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16239002346992493, + 'data_time': 0.0006519779853988439, + 'model_time': 1.1966984999889974, + 'grad_norm_pre_clip_avg': 0.17193388789892197, + 'learning_rate': 1.1389407970013185e-05, + 'epoch': 7.26} +04/19 [21:50:12] INFO | >> train_qwenlatent.py:487 + Step 28800 | grad_norm_pre_clip=0.1764 | + grad_norm_pre_clip_avg=0.1720 | Metrics: + {'align_loss': 0.024894095957279205, + 'recon_loss': 0.08606209605932236, + 'predict_loss': 0.010563735850155354, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17641273140907288, + 'mae_score': 0.010132365183787302, 'data_time': + 0.0009154480067081749, 'model_time': + 1.2296045580005739, 'grad_norm_pre_clip_avg': + 0.1720411166548729, 'learning_rate': + 1.1380725654780223e-05, 'epoch': 7.27} +04/19 [21:50:25] INFO | >> train_qwenlatent.py:487 + Step 28810 | grad_norm_pre_clip=0.1655 | + grad_norm_pre_clip_avg=0.1568 | Metrics: + {'align_loss': 0.026437856256961823, + 'recon_loss': 0.10474750399589539, + 'predict_loss': 0.008423070423305035, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16551661491394043, + 'data_time': 0.0007588330190628767, + 'model_time': 1.3098313889931887, + 'grad_norm_pre_clip_avg': 0.15684622526168823, + 'learning_rate': 1.1372043891160326e-05, + 'epoch': 7.27} +04/19 [21:50:38] INFO | >> train_qwenlatent.py:487 + Step 28820 | grad_norm_pre_clip=0.1572 | + grad_norm_pre_clip_avg=0.1547 | Metrics: + {'align_loss': 0.025998976081609726, + 'recon_loss': 0.09664179384708405, + 'predict_loss': 0.005928503815084696, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15722979605197906, + 'data_time': 0.0008717070159036666, + 'model_time': 1.2682961449900176, + 'grad_norm_pre_clip_avg': 0.15474875420331954, + 'learning_rate': 1.1363362683384882e-05, + 'epoch': 7.27} +04/19 [21:50:51] INFO | >> train_qwenlatent.py:487 + Step 28830 | grad_norm_pre_clip=0.1663 | + grad_norm_pre_clip_avg=0.1700 | Metrics: + {'align_loss': 0.025537285953760147, + 'recon_loss': 0.11072921752929688, + 'predict_loss': 0.011285203509032726, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16634684801101685, + 'data_time': 0.0006685990083497018, + 'model_time': 1.3074200260161888, + 'grad_norm_pre_clip_avg': 0.1700414702296257, + 'learning_rate': 1.1354682035685006e-05, + 'epoch': 7.27} +04/19 [21:51:03] INFO | >> train_qwenlatent.py:487 + Step 28840 | grad_norm_pre_clip=0.1527 | + grad_norm_pre_clip_avg=0.1756 | Metrics: + {'align_loss': 0.024997200816869736, + 'recon_loss': 0.11845026910305023, + 'predict_loss': 0.013225509785115719, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1526675522327423, + 'data_time': 0.001028410013532266, + 'model_time': 1.2142755890090484, + 'grad_norm_pre_clip_avg': 0.17561262100934982, + 'learning_rate': 1.1346001952291542e-05, + 'epoch': 7.28} +04/19 [21:51:16] INFO | >> train_qwenlatent.py:487 + Step 28850 | grad_norm_pre_clip=0.1286 | + grad_norm_pre_clip_avg=0.2049 | Metrics: + {'align_loss': 0.02643047273159027, + 'recon_loss': 0.10622485727071762, + 'predict_loss': 0.008761322125792503, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12860137224197388, + 'mae_score': 0.008676046938509555, 'data_time': + 0.0007856170122977346, 'model_time': + 1.2345642699801829, 'grad_norm_pre_clip_avg': + 0.20488492548465728, 'learning_rate': + 1.1337322437435052e-05, 'epoch': 7.28} +04/19 [21:51:29] INFO | >> train_qwenlatent.py:487 + Step 28860 | grad_norm_pre_clip=0.2541 | + grad_norm_pre_clip_avg=0.1910 | Metrics: + {'align_loss': 0.024737246334552765, + 'recon_loss': 0.12212791293859482, + 'predict_loss': 0.010342679917812347, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25410985946655273, + 'data_time': 0.000881228013895452, + 'model_time': 1.2656786489824299, + 'grad_norm_pre_clip_avg': 0.19098020046949388, + 'learning_rate': 1.1328643495345834e-05, + 'epoch': 7.28} +04/19 [21:51:41] INFO | >> train_qwenlatent.py:487 + Step 28870 | grad_norm_pre_clip=0.1827 | + grad_norm_pre_clip_avg=0.2057 | Metrics: + {'align_loss': 0.02578522264957428, + 'recon_loss': 0.07970790565013885, + 'predict_loss': 0.005173783283680677, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18271678686141968, + 'data_time': 0.0008674209821037948, + 'model_time': 1.2379029460134916, + 'grad_norm_pre_clip_avg': 0.20574917942285537, + 'learning_rate': 1.1319965130253892e-05, + 'epoch': 7.28} +04/19 [21:51:54] INFO | >> train_qwenlatent.py:487 + Step 28880 | grad_norm_pre_clip=0.1843 | + grad_norm_pre_clip_avg=0.1818 | Metrics: + {'align_loss': 0.025029974058270454, + 'recon_loss': 0.10420171171426773, + 'predict_loss': 0.01584814116358757, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18430590629577637, + 'data_time': 0.000760720984544605, + 'model_time': 1.2646413570037112, + 'grad_norm_pre_clip_avg': 0.18175289481878282, + 'learning_rate': 1.1311287346388964e-05, + 'epoch': 7.29} +04/19 [21:52:07] INFO | >> train_qwenlatent.py:487 + Step 28890 | grad_norm_pre_clip=0.1594 | + grad_norm_pre_clip_avg=0.1651 | Metrics: + {'align_loss': 0.023167170584201813, + 'recon_loss': 0.06788153946399689, + 'predict_loss': 0.007114134728908539, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15940698981285095, + 'data_time': 0.0006545030046254396, + 'model_time': 1.286520383990137, + 'grad_norm_pre_clip_avg': 0.16507871299982071, + 'learning_rate': 1.1302610147980484e-05, + 'epoch': 7.29} +04/19 [21:52:20] INFO | >> train_qwenlatent.py:487 + Step 28900 | grad_norm_pre_clip=0.1715 | + grad_norm_pre_clip_avg=0.1497 | Metrics: + {'align_loss': 0.025527223944664, 'recon_loss': + 0.07210537791252136, 'predict_loss': + 0.005333593115210533, 'aux_loss_decay_weight': + 0.0, 'grad_norm_pre_clip': 0.17146575450897217, + 'mae_score': 0.00890697616714615, 'data_time': + 0.0008547280158381909, 'model_time': + 1.2495787069783546, 'grad_norm_pre_clip_avg': + 0.14970588982105254, 'learning_rate': + 1.1293933539257623e-05, 'epoch': 7.29} +04/19 [21:52:33] INFO | >> train_qwenlatent.py:487 + Step 28910 | grad_norm_pre_clip=0.1686 | + grad_norm_pre_clip_avg=0.1610 | Metrics: + {'align_loss': 0.026143856346607208, + 'recon_loss': 0.1036609411239624, + 'predict_loss': 0.00872859451919794, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16863299906253815, + 'data_time': 0.0007507689879275858, + 'model_time': 1.2521355829958338, + 'grad_norm_pre_clip_avg': 0.1610483855009079, + 'learning_rate': 1.1285257524449253e-05, + 'epoch': 7.29} +04/19 [21:52:46] INFO | >> train_qwenlatent.py:487 + Step 28920 | grad_norm_pre_clip=0.1704 | + grad_norm_pre_clip_avg=0.1842 | Metrics: + {'align_loss': 0.025074802339076996, + 'recon_loss': 0.09304296970367432, + 'predict_loss': 0.00966011080890894, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17043402791023254, + 'data_time': 0.000904987013200298, + 'model_time': 1.2345278080028947, + 'grad_norm_pre_clip_avg': 0.18424907177686692, + 'learning_rate': 1.1276582107783952e-05, + 'epoch': 7.3} +04/19 [21:52:58] INFO | >> train_qwenlatent.py:487 + Step 28930 | grad_norm_pre_clip=0.1770 | + grad_norm_pre_clip_avg=0.2036 | Metrics: + {'align_loss': 0.025150582194328308, + 'recon_loss': 0.10489141941070557, + 'predict_loss': 0.008222498930990696, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17695224285125732, + 'data_time': 0.0008400329970754683, + 'model_time': 1.218934000004083, + 'grad_norm_pre_clip_avg': 0.20364001393318176, + 'learning_rate': 1.1267907293490022e-05, + 'epoch': 7.3} +04/19 [21:53:10] INFO | >> train_qwenlatent.py:487 + Step 28940 | grad_norm_pre_clip=0.1693 | + grad_norm_pre_clip_avg=0.1958 | Metrics: + {'align_loss': 0.025368252769112587, + 'recon_loss': 0.08876876533031464, + 'predict_loss': 0.005468675866723061, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16926029324531555, + 'data_time': 0.000927740999031812, + 'model_time': 1.2844526130065788, + 'grad_norm_pre_clip_avg': 0.19575331062078477, + 'learning_rate': 1.1259233085795453e-05, + 'epoch': 7.3} +04/19 [21:53:24] INFO | >> train_qwenlatent.py:487 + Step 28950 | grad_norm_pre_clip=0.1625 | + grad_norm_pre_clip_avg=0.1832 | Metrics: + {'align_loss': 0.025635063648223877, + 'recon_loss': 0.11349498480558395, + 'predict_loss': 0.008965064771473408, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16252385079860687, + 'mae_score': 0.008520740646499771, 'data_time': + 0.0008379330101888627, 'model_time': + 1.2043609539978206, 'grad_norm_pre_clip_avg': + 0.18317811489105223, 'learning_rate': + 1.1250559488927958e-05, 'epoch': 7.31} +04/19 [21:53:37] INFO | >> train_qwenlatent.py:487 + Step 28960 | grad_norm_pre_clip=0.1769 | + grad_norm_pre_clip_avg=0.1772 | Metrics: + {'align_loss': 0.0245710089802742, + 'recon_loss': 0.07777328044176102, + 'predict_loss': 0.010923793539404869, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1769283413887024, + 'data_time': 0.0007019059848971665, + 'model_time': 1.247270585008664, + 'grad_norm_pre_clip_avg': 0.1771719604730606, + 'learning_rate': 1.1241886507114936e-05, + 'epoch': 7.31} +04/19 [21:53:49] INFO | >> train_qwenlatent.py:487 + Step 28970 | grad_norm_pre_clip=0.1659 | + grad_norm_pre_clip_avg=0.1761 | Metrics: + {'align_loss': 0.025318123400211334, + 'recon_loss': 0.07571244984865189, + 'predict_loss': 0.005875426810234785, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16587762534618378, + 'data_time': 0.0007127329881768674, + 'model_time': 1.2604052209935617, + 'grad_norm_pre_clip_avg': 0.17609247118234633, + 'learning_rate': 1.1233214144583499e-05, + 'epoch': 7.31} +04/19 [21:54:02] INFO | >> train_qwenlatent.py:487 + Step 28980 | grad_norm_pre_clip=0.1610 | + grad_norm_pre_clip_avg=0.1867 | Metrics: + {'align_loss': 0.02562718465924263, + 'recon_loss': 0.09762834757566452, + 'predict_loss': 0.010367156006395817, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16099071502685547, + 'data_time': 0.0010380949825048447, + 'model_time': 1.2687123069772497, + 'grad_norm_pre_clip_avg': 0.18674818575382232, + 'learning_rate': 1.1224542405560443e-05, + 'epoch': 7.31} +04/19 [21:54:14] INFO | >> train_qwenlatent.py:487 + Step 28990 | grad_norm_pre_clip=0.2163 | + grad_norm_pre_clip_avg=0.2194 | Metrics: + {'align_loss': 0.02554961107671261, + 'recon_loss': 0.08819107711315155, + 'predict_loss': 0.008821453899145126, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21625305712223053, + 'data_time': 0.0006800009869039059, + 'model_time': 1.2629366960027255, + 'grad_norm_pre_clip_avg': 0.2194046065211296, + 'learning_rate': 1.1215871294272271e-05, + 'epoch': 7.32} +04/19 [21:54:27] INFO | >> train_qwenlatent.py:487 + Step 29000 | grad_norm_pre_clip=0.1473 | + grad_norm_pre_clip_avg=0.1962 | Metrics: + {'align_loss': 0.026004230603575706, + 'recon_loss': 0.09811471402645111, + 'predict_loss': 0.007511274889111519, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14727891981601715, + 'mae_score': 0.00843638342780036, 'data_time': + 0.0006200100178830326, 'model_time': + 1.2078230229963083, 'grad_norm_pre_clip_avg': + 0.19620537608861924, 'learning_rate': + 1.1207200814945181e-05, 'epoch': 7.32} +04/19 [21:54:40] INFO | >> train_qwenlatent.py:487 + Step 29010 | grad_norm_pre_clip=0.1475 | + grad_norm_pre_clip_avg=0.1738 | Metrics: + {'align_loss': 0.02601357363164425, + 'recon_loss': 0.08910130709409714, + 'predict_loss': 0.010577049106359482, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14750497043132782, + 'data_time': 0.0006514209962915629, + 'model_time': 1.2362458429997787, + 'grad_norm_pre_clip_avg': 0.173811674118042, + 'learning_rate': 1.1198530971805053e-05, + 'epoch': 7.32} +04/19 [21:54:53] INFO | >> train_qwenlatent.py:487 + Step 29020 | grad_norm_pre_clip=0.1777 | + grad_norm_pre_clip_avg=0.1618 | Metrics: + {'align_loss': 0.026451043784618378, + 'recon_loss': 0.09770029038190842, + 'predict_loss': 0.01013658195734024, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1776808202266693, + 'data_time': 0.0006844059971626848, + 'model_time': 1.2969342739961576, + 'grad_norm_pre_clip_avg': 0.1617852956056595, + 'learning_rate': 1.118986176907747e-05, + 'epoch': 7.32} +04/19 [21:55:05] INFO | >> train_qwenlatent.py:487 + Step 29030 | grad_norm_pre_clip=0.1525 | + grad_norm_pre_clip_avg=0.1420 | Metrics: + {'align_loss': 0.025282714515924454, + 'recon_loss': 0.0970381647348404, + 'predict_loss': 0.008636224083602428, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15245719254016876, + 'data_time': 0.0007162819965742528, + 'model_time': 1.2264431150106248, + 'grad_norm_pre_clip_avg': 0.14196281135082245, + 'learning_rate': 1.1181193210987692e-05, + 'epoch': 7.33} +04/19 [21:55:18] INFO | >> train_qwenlatent.py:487 + Step 29040 | grad_norm_pre_clip=0.1791 | + grad_norm_pre_clip_avg=0.1768 | Metrics: + {'align_loss': 0.02467367984354496, + 'recon_loss': 0.08491948992013931, + 'predict_loss': 0.009517477825284004, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17909786105155945, + 'data_time': 0.0010047630057670176, + 'model_time': 1.2523621270083822, + 'grad_norm_pre_clip_avg': 0.17677403017878532, + 'learning_rate': 1.1172525301760672e-05, + 'epoch': 7.33} +04/19 [21:55:31] INFO | >> train_qwenlatent.py:487 + Step 29050 | grad_norm_pre_clip=0.1378 | + grad_norm_pre_clip_avg=0.1709 | Metrics: + {'align_loss': 0.025977760553359985, + 'recon_loss': 0.1012350469827652, + 'predict_loss': 0.006172090768814087, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13780176639556885, + 'mae_score': 0.00904081104037998, 'data_time': + 0.0006680230144411325, 'model_time': + 1.2018302980286535, 'grad_norm_pre_clip_avg': + 0.1709173709154129, 'learning_rate': + 1.1163858045621039e-05, 'epoch': 7.33} +04/19 [21:55:44] INFO | >> train_qwenlatent.py:487 + Step 29060 | grad_norm_pre_clip=0.1944 | + grad_norm_pre_clip_avg=0.1697 | Metrics: + {'align_loss': 0.025500133633613586, + 'recon_loss': 0.1012331023812294, + 'predict_loss': 0.008734017610549927, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19437070190906525, + 'data_time': 0.0007462379871867597, + 'model_time': 1.2497036719869357, + 'grad_norm_pre_clip_avg': 0.16970701813697814, + 'learning_rate': 1.1155191446793117e-05, + 'epoch': 7.33} +04/19 [21:55:57] INFO | >> train_qwenlatent.py:487 + Step 29070 | grad_norm_pre_clip=0.1834 | + grad_norm_pre_clip_avg=0.1695 | Metrics: + {'align_loss': 0.02557787299156189, + 'recon_loss': 0.09607607126235962, + 'predict_loss': 0.01168333925306797, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1833735555410385, + 'data_time': 0.0015476130065508187, + 'model_time': 1.4657497129810508, + 'grad_norm_pre_clip_avg': 0.16948166191577912, + 'learning_rate': 1.1146525509500891e-05, + 'epoch': 7.34} +04/19 [21:56:09] INFO | >> train_qwenlatent.py:487 + Step 29080 | grad_norm_pre_clip=0.1621 | + grad_norm_pre_clip_avg=0.1736 | Metrics: + {'align_loss': 0.026059702038764954, + 'recon_loss': 0.10189490020275116, + 'predict_loss': 0.006454808171838522, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1621188372373581, + 'data_time': 0.0013660079857800156, + 'model_time': 1.2799511220073327, + 'grad_norm_pre_clip_avg': 0.17362282425165176, + 'learning_rate': 1.1137860237968036e-05, + 'epoch': 7.34} +04/19 [21:56:22] INFO | >> train_qwenlatent.py:487 + Step 29090 | grad_norm_pre_clip=0.1838 | + grad_norm_pre_clip_avg=0.1766 | Metrics: + {'align_loss': 0.026327013969421387, + 'recon_loss': 0.08485614508390427, + 'predict_loss': 0.00437589269131422, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18381161987781525, + 'data_time': 0.000825661001726985, + 'model_time': 1.226214889989933, + 'grad_norm_pre_clip_avg': 0.17660050392150878, + 'learning_rate': 1.1129195636417903e-05, + 'epoch': 7.34} +04/19 [21:56:35] INFO | >> train_qwenlatent.py:487 + Step 29100 | grad_norm_pre_clip=0.1806 | + grad_norm_pre_clip_avg=0.1779 | Metrics: + {'align_loss': 0.025520766153931618, + 'recon_loss': 0.07671190053224564, + 'predict_loss': 0.013837568461894989, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18058457970619202, + 'mae_score': 0.009528500324970967, 'data_time': + 0.0012400299892760813, 'model_time': + 1.2555749370076228, 'grad_norm_pre_clip_avg': + 0.1778934046626091, 'learning_rate': + 1.1120531709073513e-05, 'epoch': 7.34} +04/19 [21:56:48] INFO | >> train_qwenlatent.py:487 + Step 29110 | grad_norm_pre_clip=0.1999 | + grad_norm_pre_clip_avg=0.1876 | Metrics: + {'align_loss': 0.024174805730581284, + 'recon_loss': 0.09276146441698074, + 'predict_loss': 0.009503058157861233, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.199886292219162, + 'data_time': 0.0011335089802742004, + 'model_time': 1.2202772989985533, + 'grad_norm_pre_clip_avg': 0.18761314302682877, + 'learning_rate': 1.1111868460157554e-05, + 'epoch': 7.35} +04/19 [21:57:00] INFO | >> train_qwenlatent.py:487 + Step 29120 | grad_norm_pre_clip=0.2493 | + grad_norm_pre_clip_avg=0.1772 | Metrics: + {'align_loss': 0.026265207678079605, + 'recon_loss': 0.08891358226537704, + 'predict_loss': 0.007373110856860876, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24929502606391907, + 'data_time': 0.001106273994082585, + 'model_time': 1.206159838999156, + 'grad_norm_pre_clip_avg': 0.1771617442369461, + 'learning_rate': 1.1103205893892398e-05, + 'epoch': 7.35} +04/19 [21:57:13] INFO | >> train_qwenlatent.py:487 + Step 29130 | grad_norm_pre_clip=0.1820 | + grad_norm_pre_clip_avg=0.1963 | Metrics: + {'align_loss': 0.02500341646373272, + 'recon_loss': 0.11826830357313156, + 'predict_loss': 0.011524061672389507, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18196837604045868, + 'data_time': 0.0010977550118695945, + 'model_time': 1.3479831680015195, + 'grad_norm_pre_clip_avg': 0.19631920456886293, + 'learning_rate': 1.1094544014500063e-05, + 'epoch': 7.35} +04/19 [21:57:25] INFO | >> train_qwenlatent.py:487 + Step 29140 | grad_norm_pre_clip=0.1554 | + grad_norm_pre_clip_avg=0.1717 | Metrics: + {'align_loss': 0.026170365512371063, + 'recon_loss': 0.11142122000455856, + 'predict_loss': 0.010060805827379227, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15537405014038086, + 'data_time': 0.0008413520117755979, + 'model_time': 1.2391126150032505, + 'grad_norm_pre_clip_avg': 0.17171768173575402, + 'learning_rate': 1.1085882826202263e-05, + 'epoch': 7.35} +04/19 [21:57:39] INFO | >> train_qwenlatent.py:487 + Step 29150 | grad_norm_pre_clip=0.1888 | + grad_norm_pre_clip_avg=0.1672 | Metrics: + {'align_loss': 0.025935661047697067, + 'recon_loss': 0.10879494249820709, + 'predict_loss': 0.009006245993077755, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1888401359319687, + 'mae_score': 0.009112108935106982, 'data_time': + 0.0011436220083851367, 'model_time': + 1.2328000160050578, 'grad_norm_pre_clip_avg': + 0.16719371527433396, 'learning_rate': + 1.107722233322033e-05, 'epoch': 7.36} +04/19 [21:57:51] INFO | >> train_qwenlatent.py:487 + Step 29160 | grad_norm_pre_clip=0.1664 | + grad_norm_pre_clip_avg=0.1708 | Metrics: + {'align_loss': 0.025284038856625557, + 'recon_loss': 0.09481831640005112, + 'predict_loss': 0.010611862875521183, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1664091795682907, + 'data_time': 0.0007407760131172836, + 'model_time': 1.2573276829789393, + 'grad_norm_pre_clip_avg': 0.17083038836717607, + 'learning_rate': 1.1068562539775302e-05, + 'epoch': 7.36} +04/19 [21:58:04] INFO | >> train_qwenlatent.py:487 + Step 29170 | grad_norm_pre_clip=0.1892 | + grad_norm_pre_clip_avg=0.1726 | Metrics: + {'align_loss': 0.02572881057858467, + 'recon_loss': 0.09644696116447449, + 'predict_loss': 0.008763798512518406, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1891622096300125, + 'data_time': 0.0007343429897446185, + 'model_time': 1.2589805099996738, + 'grad_norm_pre_clip_avg': 0.17263682335615158, + 'learning_rate': 1.1059903450087847e-05, + 'epoch': 7.36} +04/19 [21:58:16] INFO | >> train_qwenlatent.py:487 + Step 29180 | grad_norm_pre_clip=0.1663 | + grad_norm_pre_clip_avg=0.1738 | Metrics: + {'align_loss': 0.02541106939315796, + 'recon_loss': 0.07785581797361374, + 'predict_loss': 0.01092559564858675, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1662548929452896, + 'data_time': 0.0008572599908802658, + 'model_time': 1.5591194040025584, + 'grad_norm_pre_clip_avg': 0.17384122163057328, + 'learning_rate': 1.1051245068378305e-05, + 'epoch': 7.36} +04/19 [21:58:29] INFO | >> train_qwenlatent.py:487 + Step 29190 | grad_norm_pre_clip=0.1795 | + grad_norm_pre_clip_avg=0.1772 | Metrics: + {'align_loss': 0.025396566838026047, + 'recon_loss': 0.08519095182418823, + 'predict_loss': 0.01282421313226223, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17952662706375122, + 'data_time': 0.0007045420061331242, + 'model_time': 1.2384210239979438, + 'grad_norm_pre_clip_avg': 0.1771678224205971, + 'learning_rate': 1.1042587398866664e-05, + 'epoch': 7.37} +04/19 [21:58:43] INFO | >> train_qwenlatent.py:487 + Step 29200 | grad_norm_pre_clip=0.1633 | + grad_norm_pre_clip_avg=0.1657 | Metrics: + {'align_loss': 0.023240085691213608, + 'recon_loss': 0.06765525043010712, + 'predict_loss': 0.008814668282866478, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16333866119384766, + 'mae_score': 0.00731382112245302, 'data_time': + 0.000921839993679896, 'model_time': + 1.3356673599919304, 'grad_norm_pre_clip_avg': + 0.1657155469059944, 'learning_rate': + 1.103393044577257e-05, 'epoch': 7.37} +04/19 [21:58:55] INFO | >> train_qwenlatent.py:487 + Step 29210 | grad_norm_pre_clip=0.1872 | + grad_norm_pre_clip_avg=0.1781 | Metrics: + {'align_loss': 0.026140961796045303, + 'recon_loss': 0.11650814116001129, + 'predict_loss': 0.01792103238403797, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18715040385723114, + 'data_time': 0.0007021619821898639, + 'model_time': 1.234818955999799, + 'grad_norm_pre_clip_avg': 0.17805400490760803, + 'learning_rate': 1.1025274213315314e-05, + 'epoch': 7.37} +04/19 [21:59:08] INFO | >> train_qwenlatent.py:487 + Step 29220 | grad_norm_pre_clip=0.2197 | + grad_norm_pre_clip_avg=0.2048 | Metrics: + {'align_loss': 0.024752723053097725, + 'recon_loss': 0.09800700843334198, + 'predict_loss': 0.008555999957025051, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21966728568077087, + 'data_time': 0.0007151159807108343, + 'model_time': 1.2784531739889644, + 'grad_norm_pre_clip_avg': 0.20478174686431885, + 'learning_rate': 1.101661870571384e-05, + 'epoch': 7.37} +04/19 [21:59:21] INFO | >> train_qwenlatent.py:487 + Step 29230 | grad_norm_pre_clip=0.1509 | + grad_norm_pre_clip_avg=0.1741 | Metrics: + {'align_loss': 0.024622827768325806, + 'recon_loss': 0.11484552919864655, + 'predict_loss': 0.01329074241220951, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15086404979228973, + 'data_time': 0.0008965730085037649, + 'model_time': 1.2579227059904952, + 'grad_norm_pre_clip_avg': 0.17405806183815004, + 'learning_rate': 1.100796392718674e-05, + 'epoch': 7.38} +04/19 [21:59:34] INFO | >> train_qwenlatent.py:487 + Step 29240 | grad_norm_pre_clip=0.2437 | + grad_norm_pre_clip_avg=0.2224 | Metrics: + {'align_loss': 0.025314074009656906, + 'recon_loss': 0.08857128024101257, + 'predict_loss': 0.004014051519334316, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24372126162052155, + 'data_time': 0.0009231779840774834, + 'model_time': 1.2574864659982268, + 'grad_norm_pre_clip_avg': 0.22238889336585999, + 'learning_rate': 1.0999309881952243e-05, + 'epoch': 7.38} +04/19 [21:59:47] INFO | >> train_qwenlatent.py:487 + Step 29250 | grad_norm_pre_clip=0.1458 | + grad_norm_pre_clip_avg=0.1777 | Metrics: + {'align_loss': 0.02587870880961418, + 'recon_loss': 0.07689732313156128, + 'predict_loss': 0.005256808362901211, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14578507840633392, + 'mae_score': 0.008649925713066582, 'data_time': + 0.0008638570143375546, 'model_time': + 1.275555134023307, 'grad_norm_pre_clip_avg': + 0.17768921554088593, 'learning_rate': + 1.0990656574228224e-05, 'epoch': 7.38} +04/19 [21:59:59] INFO | >> train_qwenlatent.py:487 + Step 29260 | grad_norm_pre_clip=0.1222 | + grad_norm_pre_clip_avg=0.1728 | Metrics: + {'align_loss': 0.0257219597697258, + 'recon_loss': 0.09605908393859863, + 'predict_loss': 0.01034946646541357, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12223497778177261, + 'data_time': 0.0015308379952330142, + 'model_time': 1.215623964992119, + 'grad_norm_pre_clip_avg': 0.17278508767485617, + 'learning_rate': 1.0982004008232203e-05, + 'epoch': 7.38} +04/19 [22:00:12] INFO | >> train_qwenlatent.py:487 + Step 29270 | grad_norm_pre_clip=0.1539 | + grad_norm_pre_clip_avg=0.1878 | Metrics: + {'align_loss': 0.02519581839442253, + 'recon_loss': 0.13444784283638, 'predict_loss': + 0.009829000569880009, 'aux_loss_decay_weight': + 0.0, 'grad_norm_pre_clip': 0.15385790169239044, + 'data_time': 0.0007643220014870167, + 'model_time': 1.214067288005026, + 'grad_norm_pre_clip_avg': 0.1878000870347023, + 'learning_rate': 1.0973352188181337e-05, + 'epoch': 7.39} +04/19 [22:00:25] INFO | >> train_qwenlatent.py:487 + Step 29280 | grad_norm_pre_clip=0.2461 | + grad_norm_pre_clip_avg=0.1737 | Metrics: + {'align_loss': 0.025685712695121765, + 'recon_loss': 0.07351814210414886, + 'predict_loss': 0.005399866495281458, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24612081050872803, + 'data_time': 0.0010259200062137097, + 'model_time': 1.5500945090025198, + 'grad_norm_pre_clip_avg': 0.17368980199098588, + 'learning_rate': 1.0964701118292416e-05, + 'epoch': 7.39} +04/19 [22:00:37] INFO | >> train_qwenlatent.py:487 + Step 29290 | grad_norm_pre_clip=0.2114 | + grad_norm_pre_clip_avg=0.1976 | Metrics: + {'align_loss': 0.02508872002363205, + 'recon_loss': 0.06950759887695312, + 'predict_loss': 0.009388478472828865, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21141749620437622, + 'data_time': 0.0011081339907832444, + 'model_time': 1.2843050879891962, + 'grad_norm_pre_clip_avg': 0.1975648045539856, + 'learning_rate': 1.0956050802781869e-05, + 'epoch': 7.39} +04/19 [22:00:50] INFO | >> train_qwenlatent.py:487 + Step 29300 | grad_norm_pre_clip=0.1280 | + grad_norm_pre_clip_avg=0.1661 | Metrics: + {'align_loss': 0.02453787438571453, + 'recon_loss': 0.09456208348274231, + 'predict_loss': 0.008455806411802769, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12798424065113068, + 'mae_score': 0.007364730147628097, 'data_time': + 0.0008915179932955652, 'model_time': + 1.2059099290054291, 'grad_norm_pre_clip_avg': + 0.1660643696784973, 'learning_rate': + 1.094740124586575e-05, 'epoch': 7.39} +04/19 [22:01:03] INFO | >> train_qwenlatent.py:487 + Step 29310 | grad_norm_pre_clip=0.1888 | + grad_norm_pre_clip_avg=0.1674 | Metrics: + {'align_loss': 0.026202775537967682, + 'recon_loss': 0.11510665714740753, + 'predict_loss': 0.016486668959259987, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18883788585662842, + 'data_time': 0.0006234550091903657, + 'model_time': 1.210101937991567, + 'grad_norm_pre_clip_avg': 0.16739156991243362, + 'learning_rate': 1.0938752451759752e-05, + 'epoch': 7.4} +04/19 [22:01:16] INFO | >> train_qwenlatent.py:487 + Step 29320 | grad_norm_pre_clip=0.1698 | + grad_norm_pre_clip_avg=0.1597 | Metrics: + {'align_loss': 0.02487664483487606, + 'recon_loss': 0.08021543174982071, + 'predict_loss': 0.006612888537347317, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16981299221515656, + 'data_time': 0.0006300919922068715, + 'model_time': 1.2188593260070775, + 'grad_norm_pre_clip_avg': 0.1596725508570671, + 'learning_rate': 1.0930104424679196e-05, + 'epoch': 7.4} +04/19 [22:01:28] INFO | >> train_qwenlatent.py:487 + Step 29330 | grad_norm_pre_clip=0.1724 | + grad_norm_pre_clip_avg=0.1720 | Metrics: + {'align_loss': 0.025216056033968925, + 'recon_loss': 0.12452700734138489, + 'predict_loss': 0.018088746815919876, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17242808640003204, + 'data_time': 0.0009578650060575455, + 'model_time': 1.2528215509955771, + 'grad_norm_pre_clip_avg': 0.17202467918395997, + 'learning_rate': 1.0921457168839016e-05, + 'epoch': 7.4} +04/19 [22:01:41] INFO | >> train_qwenlatent.py:487 + Step 29340 | grad_norm_pre_clip=0.1626 | + grad_norm_pre_clip_avg=0.1884 | Metrics: + {'align_loss': 0.026279320940375328, + 'recon_loss': 0.130723237991333, + 'predict_loss': 0.009557201527059078, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16264508664608002, + 'data_time': 0.0006787199818063527, + 'model_time': 1.2091815960011445, + 'grad_norm_pre_clip_avg': 0.1884069561958313, + 'learning_rate': 1.0912810688453782e-05, + 'epoch': 7.4} +04/19 [22:01:54] INFO | >> train_qwenlatent.py:487 + Step 29350 | grad_norm_pre_clip=0.1806 | + grad_norm_pre_clip_avg=0.1931 | Metrics: + {'align_loss': 0.025554699823260307, + 'recon_loss': 0.11637642234563828, + 'predict_loss': 0.011554304510354996, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18060125410556793, + 'mae_score': 0.007759182732384484, 'data_time': + 0.0006669629947282374, 'model_time': + 1.2358749819977675, 'grad_norm_pre_clip_avg': + 0.19313603937625884, 'learning_rate': + 1.0904164987737686e-05, 'epoch': 7.41} +04/19 [22:02:08] INFO | >> train_qwenlatent.py:487 + Step 29360 | grad_norm_pre_clip=0.1837 | + grad_norm_pre_clip_avg=0.1983 | Metrics: + {'align_loss': 0.02575649321079254, + 'recon_loss': 0.08709431439638138, + 'predict_loss': 0.0069784182123839855, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18368886411190033, + 'data_time': 0.000971239001955837, + 'model_time': 1.2407145269971807, + 'grad_norm_pre_clip_avg': 0.198319011926651, + 'learning_rate': 1.089552007090454e-05, + 'epoch': 7.41} +04/19 [22:02:20] INFO | >> train_qwenlatent.py:487 + Step 29370 | grad_norm_pre_clip=0.2119 | + grad_norm_pre_clip_avg=0.1724 | Metrics: + {'align_loss': 0.02578652836382389, + 'recon_loss': 0.08555742353200912, + 'predict_loss': 0.006942202802747488, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2119484692811966, + 'data_time': 0.0009348650055471808, + 'model_time': 1.2744224940252025, + 'grad_norm_pre_clip_avg': 0.17236143052577974, + 'learning_rate': 1.0886875942167762e-05, + 'epoch': 7.41} +04/19 [22:02:32] INFO | >> train_qwenlatent.py:487 + Step 29380 | grad_norm_pre_clip=0.1840 | + grad_norm_pre_clip_avg=0.1695 | Metrics: + {'align_loss': 0.02525455877184868, + 'recon_loss': 0.09676015377044678, + 'predict_loss': 0.010563262738287449, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18400834500789642, + 'data_time': 0.0008772119763307273, + 'model_time': 1.238751308992505, + 'grad_norm_pre_clip_avg': 0.16951415538787842, + 'learning_rate': 1.0878232605740406e-05, + 'epoch': 7.41} +04/19 [22:02:45] INFO | >> train_qwenlatent.py:487 + Step 29390 | grad_norm_pre_clip=0.1903 | + grad_norm_pre_clip_avg=0.1688 | Metrics: + {'align_loss': 0.02557072415947914, + 'recon_loss': 0.12260226905345917, + 'predict_loss': 0.013426869176328182, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19034628570079803, + 'data_time': 0.0006595840095542371, + 'model_time': 1.2332027579832356, + 'grad_norm_pre_clip_avg': 0.16879741847515106, + 'learning_rate': 1.0869590065835122e-05, + 'epoch': 7.42} +04/19 [22:02:58] INFO | >> train_qwenlatent.py:487 + Step 29400 | grad_norm_pre_clip=0.2264 | + grad_norm_pre_clip_avg=0.1803 | Metrics: + {'align_loss': 0.02663402631878853, + 'recon_loss': 0.10616344213485718, + 'predict_loss': 0.008831637911498547, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2263559103012085, + 'mae_score': 0.008827017878626917, 'data_time': + 0.0009279939986299723, 'model_time': + 1.2635020860179793, 'grad_norm_pre_clip_avg': + 0.1803087756037712, 'learning_rate': + 1.0860948326664183e-05, 'epoch': 7.42} +04/19 [22:03:11] INFO | >> train_qwenlatent.py:487 + Step 29410 | grad_norm_pre_clip=0.1495 | + grad_norm_pre_clip_avg=0.1963 | Metrics: + {'align_loss': 0.02598942629992962, + 'recon_loss': 0.10375401377677917, + 'predict_loss': 0.010144266299903393, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14946775138378143, + 'data_time': 0.0011786439863499254, + 'model_time': 1.3540554490173236, + 'grad_norm_pre_clip_avg': 0.19633683264255525, + 'learning_rate': 1.0852307392439469e-05, + 'epoch': 7.42} +04/19 [22:03:24] INFO | >> train_qwenlatent.py:487 + Step 29420 | grad_norm_pre_clip=0.1393 | + grad_norm_pre_clip_avg=0.1750 | Metrics: + {'align_loss': 0.025245940312743187, + 'recon_loss': 0.0706312283873558, + 'predict_loss': 0.010637646540999413, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13934959471225739, + 'data_time': 0.0006917330028954893, + 'model_time': 1.2253747069917154, + 'grad_norm_pre_clip_avg': 0.17503832578659057, + 'learning_rate': 1.0843667267372462e-05, + 'epoch': 7.42} +04/19 [22:03:36] INFO | >> train_qwenlatent.py:487 + Step 29430 | grad_norm_pre_clip=0.1661 | + grad_norm_pre_clip_avg=0.1585 | Metrics: + {'align_loss': 0.02591586858034134, + 'recon_loss': 0.11201014369726181, + 'predict_loss': 0.01542755402624607, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1660938709974289, + 'data_time': 0.0011180249857716262, + 'model_time': 1.1930266609997489, + 'grad_norm_pre_clip_avg': 0.15847869515419005, + 'learning_rate': 1.0835027955674255e-05, + 'epoch': 7.43} +04/19 [22:03:48] INFO | >> train_qwenlatent.py:487 + Step 29440 | grad_norm_pre_clip=0.1727 | + grad_norm_pre_clip_avg=0.1902 | Metrics: + {'align_loss': 0.026829861104488373, + 'recon_loss': 0.13424037396907806, + 'predict_loss': 0.01669461466372013, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1726531833410263, + 'data_time': 0.0006765050056856126, + 'model_time': 1.2591066290042363, + 'grad_norm_pre_clip_avg': 0.19022500962018968, + 'learning_rate': 1.0826389461555543e-05, + 'epoch': 7.43} +04/19 [22:04:02] INFO | >> train_qwenlatent.py:487 + Step 29450 | grad_norm_pre_clip=0.1808 | + grad_norm_pre_clip_avg=0.1642 | Metrics: + {'align_loss': 0.026017509400844574, + 'recon_loss': 0.1082659587264061, + 'predict_loss': 0.005762417335063219, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1807800680398941, + 'mae_score': 0.011723461666622678, 'data_time': + 0.0006072100077290088, 'model_time': + 1.214937839016784, 'grad_norm_pre_clip_avg': + 0.16415322422981263, 'learning_rate': + 1.0817751789226626e-05, 'epoch': 7.43} +04/19 [22:04:15] INFO | >> train_qwenlatent.py:487 + Step 29460 | grad_norm_pre_clip=0.1730 | + grad_norm_pre_clip_avg=0.1646 | Metrics: + {'align_loss': 0.025395862758159637, + 'recon_loss': 0.08716418594121933, + 'predict_loss': 0.0058847409673035145, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17302381992340088, + 'data_time': 0.0009569149988237768, + 'model_time': 1.2323399880260695, + 'grad_norm_pre_clip_avg': 0.16458582878112793, + 'learning_rate': 1.0809114942897396e-05, + 'epoch': 7.43} +04/19 [22:04:28] INFO | >> train_qwenlatent.py:487 + Step 29470 | grad_norm_pre_clip=0.2793 | + grad_norm_pre_clip_avg=0.1829 | Metrics: + {'align_loss': 0.026179034262895584, + 'recon_loss': 0.07623739540576935, + 'predict_loss': 0.0066406638361513615, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2792831361293793, + 'data_time': 0.001021272997604683, + 'model_time': 1.2153828470036387, + 'grad_norm_pre_clip_avg': 0.18289651572704316, + 'learning_rate': 1.0800478926777353e-05, + 'epoch': 7.44} +04/19 [22:04:40] INFO | >> train_qwenlatent.py:487 + Step 29480 | grad_norm_pre_clip=0.1740 | + grad_norm_pre_clip_avg=0.1948 | Metrics: + {'align_loss': 0.025423821061849594, + 'recon_loss': 0.0755675882101059, + 'predict_loss': 0.005841830279678106, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17397907376289368, + 'data_time': 0.0007427679956890643, + 'model_time': 1.222371737996582, + 'grad_norm_pre_clip_avg': 0.19477900713682175, + 'learning_rate': 1.0791843745075583e-05, + 'epoch': 7.44} +04/19 [22:04:53] INFO | >> train_qwenlatent.py:487 + Step 29490 | grad_norm_pre_clip=0.1502 | + grad_norm_pre_clip_avg=0.1968 | Metrics: + {'align_loss': 0.02553396485745907, + 'recon_loss': 0.07995755225419998, + 'predict_loss': 0.006838412024080753, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15016509592533112, + 'data_time': 0.0009011190268211067, + 'model_time': 1.2004862340108957, + 'grad_norm_pre_clip_avg': 0.19678353667259216, + 'learning_rate': 1.0783209402000768e-05, + 'epoch': 7.44} +04/19 [22:05:06] INFO | >> train_qwenlatent.py:487 + Step 29500 | grad_norm_pre_clip=0.1834 | + grad_norm_pre_clip_avg=0.1803 | Metrics: + {'align_loss': 0.02464110404253006, + 'recon_loss': 0.08949301391839981, + 'predict_loss': 0.008882462047040462, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1833520084619522, + 'mae_score': 0.008671677219975102, 'data_time': + 0.0006736299837939441, 'model_time': + 1.2251585320045706, 'grad_norm_pre_clip_avg': + 0.18025312423706055, 'learning_rate': + 1.0774575901761186e-05, 'epoch': 7.44} +04/19 [22:05:19] INFO | >> train_qwenlatent.py:487 + Step 29510 | grad_norm_pre_clip=0.1476 | + grad_norm_pre_clip_avg=0.1677 | Metrics: + {'align_loss': 0.025133032351732254, + 'recon_loss': 0.07525790482759476, + 'predict_loss': 0.007333733141422272, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14760519564151764, + 'data_time': 0.0006323510024230927, + 'model_time': 1.1816943069861736, + 'grad_norm_pre_clip_avg': 0.16768295913934708, + 'learning_rate': 1.0765943248564692e-05, + 'epoch': 7.45} +04/19 [22:05:31] INFO | >> train_qwenlatent.py:487 + Step 29520 | grad_norm_pre_clip=0.1898 | + grad_norm_pre_clip_avg=0.1694 | Metrics: + {'align_loss': 0.026446040719747543, + 'recon_loss': 0.12750959396362305, + 'predict_loss': 0.0129475686699152, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18975947797298431, + 'data_time': 0.000676986004691571, + 'model_time': 1.2223183049936779, + 'grad_norm_pre_clip_avg': 0.16941808611154557, + 'learning_rate': 1.0757311446618745e-05, + 'epoch': 7.45} +04/19 [22:05:44] INFO | >> train_qwenlatent.py:487 + Step 29530 | grad_norm_pre_clip=0.2136 | + grad_norm_pre_clip_avg=0.1736 | Metrics: + {'align_loss': 0.025449693202972412, + 'recon_loss': 0.0711863785982132, + 'predict_loss': 0.004640099126845598, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2136041820049286, + 'data_time': 0.0008559389971196651, + 'model_time': 1.202465504000429, + 'grad_norm_pre_clip_avg': 0.1736377939581871, + 'learning_rate': 1.0748680500130373e-05, + 'epoch': 7.45} +04/19 [22:05:56] INFO | >> train_qwenlatent.py:487 + Step 29540 | grad_norm_pre_clip=0.1886 | + grad_norm_pre_clip_avg=0.1780 | Metrics: + {'align_loss': 0.025208167731761932, + 'recon_loss': 0.10010520368814468, + 'predict_loss': 0.0118799963966012, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1886405199766159, + 'data_time': 0.0009420340065844357, + 'model_time': 1.2439822419837583, + 'grad_norm_pre_clip_avg': 0.17802256643772124, + 'learning_rate': 1.0740050413306198e-05, + 'epoch': 7.45} +04/19 [22:06:09] INFO | >> train_qwenlatent.py:487 + Step 29550 | grad_norm_pre_clip=0.1911 | + grad_norm_pre_clip_avg=0.1953 | Metrics: + {'align_loss': 0.02542307786643505, + 'recon_loss': 0.08629485964775085, + 'predict_loss': 0.005907013546675444, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19105221331119537, + 'mae_score': 0.010202084790478955, 'data_time': + 0.0009713729959912598, 'model_time': + 1.2299962889810558, 'grad_norm_pre_clip_avg': + 0.19530172199010848, 'learning_rate': + 1.0731421190352426e-05, 'epoch': 7.46} +04/19 [22:06:22] INFO | >> train_qwenlatent.py:487 + Step 29560 | grad_norm_pre_clip=0.1527 | + grad_norm_pre_clip_avg=0.1660 | Metrics: + {'align_loss': 0.024003220722079277, + 'recon_loss': 0.06971750408411026, + 'predict_loss': 0.006885950919240713, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15266653895378113, + 'data_time': 0.0006410260102711618, + 'model_time': 1.2213348069926724, + 'grad_norm_pre_clip_avg': 0.16600824818015097, + 'learning_rate': 1.072279283547482e-05, + 'epoch': 7.46} +04/19 [22:06:34] INFO | >> train_qwenlatent.py:487 + Step 29570 | grad_norm_pre_clip=0.1654 | + grad_norm_pre_clip_avg=0.1650 | Metrics: + {'align_loss': 0.025846105068922043, + 'recon_loss': 0.11660445481538773, + 'predict_loss': 0.011124980635941029, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1653752624988556, + 'data_time': 0.0007156820211093873, + 'model_time': 1.316262423992157, + 'grad_norm_pre_clip_avg': 0.16500835567712785, + 'learning_rate': 1.0714165352878751e-05, + 'epoch': 7.46} +04/19 [22:06:47] INFO | >> train_qwenlatent.py:487 + Step 29580 | grad_norm_pre_clip=0.1903 | + grad_norm_pre_clip_avg=0.1781 | Metrics: + {'align_loss': 0.024175353348255157, + 'recon_loss': 0.07431969046592712, + 'predict_loss': 0.008213808760046959, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19033482670783997, + 'data_time': 0.0008944550063461065, + 'model_time': 1.2175322440161835, + 'grad_norm_pre_clip_avg': 0.17814332023262977, + 'learning_rate': 1.070553874676914e-05, + 'epoch': 7.46} +04/19 [22:07:00] INFO | >> train_qwenlatent.py:487 + Step 29590 | grad_norm_pre_clip=0.1497 | + grad_norm_pre_clip_avg=0.1579 | Metrics: + {'align_loss': 0.026099704205989838, + 'recon_loss': 0.09293458610773087, + 'predict_loss': 0.008816252462565899, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1496843695640564, + 'data_time': 0.0005883759877178818, + 'model_time': 1.214170062012272, + 'grad_norm_pre_clip_avg': 0.15788438245654107, + 'learning_rate': 1.0696913021350499e-05, + 'epoch': 7.47} +04/19 [22:07:13] INFO | >> train_qwenlatent.py:487 + Step 29600 | grad_norm_pre_clip=0.1752 | + grad_norm_pre_clip_avg=0.1700 | Metrics: + {'align_loss': 0.02513006143271923, + 'recon_loss': 0.10579568892717361, + 'predict_loss': 0.007518551778048277, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1751966029405594, + 'mae_score': 0.007827602420841252, 'data_time': + 0.0009608870022930205, 'model_time': + 1.2001854300033301, 'grad_norm_pre_clip_avg': + 0.170039664208889, 'learning_rate': + 1.0688288180826887e-05, 'epoch': 7.47} +04/19 [22:07:26] INFO | >> train_qwenlatent.py:487 + Step 29610 | grad_norm_pre_clip=0.2300 | + grad_norm_pre_clip_avg=0.1767 | Metrics: + {'align_loss': 0.02518048696219921, + 'recon_loss': 0.11216777563095093, + 'predict_loss': 0.008156590163707733, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22998645901679993, + 'data_time': 0.0009842159925028682, + 'model_time': 1.2903505569847766, + 'grad_norm_pre_clip_avg': 0.17665593475103378, + 'learning_rate': 1.0679664229401957e-05, + 'epoch': 7.47} +04/19 [22:07:38] INFO | >> train_qwenlatent.py:487 + Step 29620 | grad_norm_pre_clip=0.1733 | + grad_norm_pre_clip_avg=0.2036 | Metrics: + {'align_loss': 0.024625737220048904, + 'recon_loss': 0.09945785999298096, + 'predict_loss': 0.008045056834816933, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17329108715057373, + 'data_time': 0.0010951229778584093, + 'model_time': 1.2365270869922824, + 'grad_norm_pre_clip_avg': 0.2036421239376068, + 'learning_rate': 1.0671041171278919e-05, + 'epoch': 7.47} +04/19 [22:07:51] INFO | >> train_qwenlatent.py:487 + Step 29630 | grad_norm_pre_clip=0.1519 | + grad_norm_pre_clip_avg=0.1672 | Metrics: + {'align_loss': 0.025013843551278114, + 'recon_loss': 0.09181459248065948, + 'predict_loss': 0.005139956716448069, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1518612951040268, + 'data_time': 0.0010375200072303414, + 'model_time': 1.250429252977483, + 'grad_norm_pre_clip_avg': 0.16720820516347884, + 'learning_rate': 1.0662419010660538e-05, + 'epoch': 7.48} +04/19 [22:08:04] INFO | >> train_qwenlatent.py:487 + Step 29640 | grad_norm_pre_clip=0.1328 | + grad_norm_pre_clip_avg=0.1646 | Metrics: + {'align_loss': 0.025656528770923615, + 'recon_loss': 0.11886294186115265, + 'predict_loss': 0.009005444124341011, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13281747698783875, + 'data_time': 0.0009111399995163083, + 'model_time': 1.2031478740100283, + 'grad_norm_pre_clip_avg': 0.1645742103457451, + 'learning_rate': 1.065379775174916e-05, + 'epoch': 7.48} +04/19 [22:08:17] INFO | >> train_qwenlatent.py:487 + Step 29650 | grad_norm_pre_clip=0.1715 | + grad_norm_pre_clip_avg=0.1838 | Metrics: + {'align_loss': 0.026527438312768936, + 'recon_loss': 0.09943575412034988, + 'predict_loss': 0.009542479179799557, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17153838276863098, + 'mae_score': 0.008136461446951102, 'data_time': + 0.0006686780252493918, 'model_time': + 1.198883136996301, 'grad_norm_pre_clip_avg': + 0.18382012993097305, 'learning_rate': + 1.0645177398746679e-05, 'epoch': 7.48} +04/19 [22:08:30] INFO | >> train_qwenlatent.py:487 + Step 29660 | grad_norm_pre_clip=0.1298 | + grad_norm_pre_clip_avg=0.1535 | Metrics: + {'align_loss': 0.02736084721982479, + 'recon_loss': 0.09380000084638596, + 'predict_loss': 0.007036951370537281, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12980003654956818, + 'data_time': 0.0009867180197034031, + 'model_time': 1.2947976069990546, + 'grad_norm_pre_clip_avg': 0.15347184538841246, + 'learning_rate': 1.0636557955854546e-05, + 'epoch': 7.48} +04/19 [22:08:42] INFO | >> train_qwenlatent.py:487 + Step 29670 | grad_norm_pre_clip=0.1878 | + grad_norm_pre_clip_avg=0.1661 | Metrics: + {'align_loss': 0.026369793340563774, + 'recon_loss': 0.13576281070709229, + 'predict_loss': 0.011390847153961658, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1878236085176468, + 'data_time': 0.0010452140122652054, + 'model_time': 1.286405310005648, + 'grad_norm_pre_clip_avg': 0.16612017452716826, + 'learning_rate': 1.0627939427273778e-05, + 'epoch': 7.49} +04/19 [22:08:54] INFO | >> train_qwenlatent.py:487 + Step 29680 | grad_norm_pre_clip=0.1449 | + grad_norm_pre_clip_avg=0.1674 | Metrics: + {'align_loss': 0.025202661752700806, + 'recon_loss': 0.08537324517965317, + 'predict_loss': 0.007013863418251276, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14494267106056213, + 'data_time': 0.0007531290175393224, + 'model_time': 1.1997519090073183, + 'grad_norm_pre_clip_avg': 0.1673751413822174, + 'learning_rate': 1.0619321817204944e-05, + 'epoch': 7.49} +04/19 [22:09:07] INFO | >> train_qwenlatent.py:487 + Step 29690 | grad_norm_pre_clip=0.2383 | + grad_norm_pre_clip_avg=0.1642 | Metrics: + {'align_loss': 0.025381993502378464, + 'recon_loss': 0.136259526014328, + 'predict_loss': 0.016107754781842232, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23834897577762604, + 'data_time': 0.001218631019582972, + 'model_time': 1.2298101630003657, + 'grad_norm_pre_clip_avg': 0.1642364040017128, + 'learning_rate': 1.0610705129848154e-05, + 'epoch': 7.49} +04/19 [22:09:20] INFO | >> train_qwenlatent.py:487 + Step 29700 | grad_norm_pre_clip=0.1404 | + grad_norm_pre_clip_avg=0.1650 | Metrics: + {'align_loss': 0.024355191737413406, + 'recon_loss': 0.11825449019670486, + 'predict_loss': 0.008315098471939564, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14043977856636047, + 'mae_score': 0.009267765766865499, 'data_time': + 0.0007135120104067028, 'model_time': + 1.2473319279961288, 'grad_norm_pre_clip_avg': + 0.16502674371004106, 'learning_rate': + 1.0602089369403077e-05, 'epoch': 7.49} +04/19 [22:09:32] INFO | >> train_qwenlatent.py:487 + Step 29710 | grad_norm_pre_clip=0.1447 | + grad_norm_pre_clip_avg=0.1661 | Metrics: + {'align_loss': 0.02501728944480419, + 'recon_loss': 0.06089780107140541, + 'predict_loss': 0.004884711466729641, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1447218507528305, + 'data_time': 0.0006513429980259389, + 'model_time': 1.2049004230066203, + 'grad_norm_pre_clip_avg': 0.16609608232975007, + 'learning_rate': 1.0593474540068936e-05, + 'epoch': 7.5} +04/19 [22:09:45] INFO | >> train_qwenlatent.py:487 + Step 29720 | grad_norm_pre_clip=0.2245 | + grad_norm_pre_clip_avg=0.1751 | Metrics: + {'align_loss': 0.026323243975639343, + 'recon_loss': 0.11111170798540115, + 'predict_loss': 0.01178387925028801, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22445152699947357, + 'data_time': 0.0006611940043512732, + 'model_time': 1.2451307649898808, + 'grad_norm_pre_clip_avg': 0.1751101553440094, + 'learning_rate': 1.0584860646044489e-05, + 'epoch': 7.5} +04/19 [22:09:58] INFO | >> train_qwenlatent.py:487 + Step 29730 | grad_norm_pre_clip=0.1469 | + grad_norm_pre_clip_avg=0.2051 | Metrics: + {'align_loss': 0.026361895725131035, + 'recon_loss': 0.10870610922574997, + 'predict_loss': 0.008251157589256763, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1468864530324936, + 'data_time': 0.0008803980017546564, + 'model_time': 1.2677799590164796, + 'grad_norm_pre_clip_avg': 0.20505454689264296, + 'learning_rate': 1.057624769152805e-05, + 'epoch': 7.5} +04/19 [22:10:11] INFO | >> train_qwenlatent.py:487 + Step 29740 | grad_norm_pre_clip=0.1512 | + grad_norm_pre_clip_avg=0.1647 | Metrics: + {'align_loss': 0.026594050228595734, + 'recon_loss': 0.10106802731752396, + 'predict_loss': 0.009181629866361618, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15120932459831238, + 'data_time': 0.0010491269931662828, + 'model_time': 1.2571786329790484, + 'grad_norm_pre_clip_avg': 0.16469032019376756, + 'learning_rate': 1.0567635680717462e-05, + 'epoch': 7.5} +04/19 [22:10:24] INFO | >> train_qwenlatent.py:487 + Step 29750 | grad_norm_pre_clip=0.1558 | + grad_norm_pre_clip_avg=0.1583 | Metrics: + {'align_loss': 0.024906549602746964, + 'recon_loss': 0.07036900520324707, + 'predict_loss': 0.004428827669471502, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1557871699333191, + 'mae_score': 0.0075579849449363915, + 'data_time': 0.0009040290024131536, + 'model_time': 1.2302095479972195, + 'grad_norm_pre_clip_avg': 0.15825065970420837, + 'learning_rate': 1.0559024617810115e-05, + 'epoch': 7.51} +04/19 [22:10:37] INFO | >> train_qwenlatent.py:487 + Step 29760 | grad_norm_pre_clip=0.1518 | + grad_norm_pre_clip_avg=0.1660 | Metrics: + {'align_loss': 0.024533893913030624, + 'recon_loss': 0.09609851986169815, + 'predict_loss': 0.009253252297639847, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15180961787700653, + 'data_time': 0.0009862039878498763, + 'model_time': 1.2440338319865987, + 'grad_norm_pre_clip_avg': 0.16603283137083052, + 'learning_rate': 1.0550414507002948e-05, + 'epoch': 7.51} +04/19 [22:10:50] INFO | >> train_qwenlatent.py:487 + Step 29770 | grad_norm_pre_clip=0.1997 | + grad_norm_pre_clip_avg=0.1706 | Metrics: + {'align_loss': 0.02533273585140705, + 'recon_loss': 0.1090785562992096, + 'predict_loss': 0.01119131501764059, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19968239963054657, + 'data_time': 0.0006535100110340863, + 'model_time': 1.2193010890041478, + 'grad_norm_pre_clip_avg': 0.1706162855029106, + 'learning_rate': 1.0541805352492405e-05, + 'epoch': 7.51} +04/19 [22:11:02] INFO | >> train_qwenlatent.py:487 + Step 29780 | grad_norm_pre_clip=0.1878 | + grad_norm_pre_clip_avg=0.1705 | Metrics: + {'align_loss': 0.025492003187537193, + 'recon_loss': 0.12742605805397034, + 'predict_loss': 0.011175490915775299, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18781310319900513, + 'data_time': 0.0007087900012265891, + 'model_time': 1.2150391479954123, + 'grad_norm_pre_clip_avg': 0.17051232159137725, + 'learning_rate': 1.0533197158474495e-05, + 'epoch': 7.51} +04/19 [22:11:14] INFO | >> train_qwenlatent.py:487 + Step 29790 | grad_norm_pre_clip=0.1975 | + grad_norm_pre_clip_avg=0.1716 | Metrics: + {'align_loss': 0.025661658495664597, + 'recon_loss': 0.12338149547576904, + 'predict_loss': 0.012235378846526146, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1974688172340393, + 'data_time': 0.0007624500140082091, + 'model_time': 1.2385125490254723, + 'grad_norm_pre_clip_avg': 0.17156871780753136, + 'learning_rate': 1.0524589929144744e-05, + 'epoch': 7.52} +04/19 [22:11:27] INFO | >> train_qwenlatent.py:487 + Step 29800 | grad_norm_pre_clip=0.1136 | + grad_norm_pre_clip_avg=0.1520 | Metrics: + {'align_loss': 0.025726526975631714, + 'recon_loss': 0.08443315327167511, + 'predict_loss': 0.00815439224243164, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11360634863376617, + 'mae_score': 0.00878765346767666, 'data_time': + 0.0009251589945051819, 'model_time': + 1.2064208879892249, 'grad_norm_pre_clip_avg': + 0.15199943855404854, 'learning_rate': + 1.0515983668698213e-05, 'epoch': 7.52} +04/19 [22:11:40] INFO | >> train_qwenlatent.py:487 + Step 29810 | grad_norm_pre_clip=0.1940 | + grad_norm_pre_clip_avg=0.1874 | Metrics: + {'align_loss': 0.025218285620212555, + 'recon_loss': 0.09571930766105652, + 'predict_loss': 0.0074193743057549, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19399800896644592, + 'data_time': 0.0009227970149368048, + 'model_time': 1.2684021719906013, + 'grad_norm_pre_clip_avg': 0.18739822655916213, + 'learning_rate': 1.050737838132949e-05, + 'epoch': 7.52} +04/19 [22:11:52] INFO | >> train_qwenlatent.py:487 + Step 29820 | grad_norm_pre_clip=0.1710 | + grad_norm_pre_clip_avg=0.1721 | Metrics: + {'align_loss': 0.026004642248153687, + 'recon_loss': 0.08962678909301758, + 'predict_loss': 0.008090741001069546, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17097800970077515, + 'data_time': 0.001077461987733841, + 'model_time': 1.2462944019935094, + 'grad_norm_pre_clip_avg': 0.17210560888051987, + 'learning_rate': 1.0498774071232681e-05, + 'epoch': 7.52} +04/19 [22:12:05] INFO | >> train_qwenlatent.py:487 + Step 29830 | grad_norm_pre_clip=0.1444 | + grad_norm_pre_clip_avg=0.1933 | Metrics: + {'align_loss': 0.025818005204200745, + 'recon_loss': 0.08977258205413818, + 'predict_loss': 0.0057932124473154545, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14436206221580505, + 'data_time': 0.000977104005869478, + 'model_time': 1.339066938002361, + 'grad_norm_pre_clip_avg': 0.1932702213525772, + 'learning_rate': 1.049017074260143e-05, + 'epoch': 7.53} +04/19 [22:12:18] INFO | >> train_qwenlatent.py:487 + Step 29840 | grad_norm_pre_clip=0.1940 | + grad_norm_pre_clip_avg=0.1958 | Metrics: + {'align_loss': 0.023747552186250687, + 'recon_loss': 0.08779023587703705, + 'predict_loss': 0.009361976757645607, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19396840035915375, + 'data_time': 0.0010238380054943264, + 'model_time': 1.3070688259904273, + 'grad_norm_pre_clip_avg': 0.19575359374284745, + 'learning_rate': 1.0481568399628893e-05, + 'epoch': 7.53} +04/19 [22:12:31] INFO | >> train_qwenlatent.py:487 + Step 29850 | grad_norm_pre_clip=0.1396 | + grad_norm_pre_clip_avg=0.1850 | Metrics: + {'align_loss': 0.02422941103577614, + 'recon_loss': 0.08850514888763428, + 'predict_loss': 0.01077679917216301, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13962043821811676, + 'mae_score': 0.009168086180815825, 'data_time': + 0.0007862869824748486, 'model_time': + 1.2710419459908735, 'grad_norm_pre_clip_avg': + 0.18500431925058364, 'learning_rate': + 1.0472967046507746e-05, 'epoch': 7.53} +04/19 [22:12:44] INFO | >> train_qwenlatent.py:487 + Step 29860 | grad_norm_pre_clip=0.1731 | + grad_norm_pre_clip_avg=0.1610 | Metrics: + {'align_loss': 0.0242040753364563, + 'recon_loss': 0.10299014300107956, + 'predict_loss': 0.01215098425745964, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17308646440505981, + 'data_time': 0.0006369510083459318, + 'model_time': 1.5298071219876874, + 'grad_norm_pre_clip_avg': 0.1609862245619297, + 'learning_rate': 1.046436668743018e-05, + 'epoch': 7.53} +04/19 [22:12:56] INFO | >> train_qwenlatent.py:487 + Step 29870 | grad_norm_pre_clip=0.1791 | + grad_norm_pre_clip_avg=0.1732 | Metrics: + {'align_loss': 0.024441463872790337, + 'recon_loss': 0.09009310603141785, + 'predict_loss': 0.009858333505690098, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1791020929813385, + 'data_time': 0.0006964769854675978, + 'model_time': 1.228947358991718, + 'grad_norm_pre_clip_avg': 0.17317981719970704, + 'learning_rate': 1.0455767326587912e-05, + 'epoch': 7.54} +04/19 [22:13:09] INFO | >> train_qwenlatent.py:487 + Step 29880 | grad_norm_pre_clip=0.1685 | + grad_norm_pre_clip_avg=0.1846 | Metrics: + {'align_loss': 0.026633059605956078, + 'recon_loss': 0.08190879970788956, + 'predict_loss': 0.009145388379693031, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16849005222320557, + 'data_time': 0.0006781450065318495, + 'model_time': 1.2298300730180927, + 'grad_norm_pre_clip_avg': 0.18462250530719757, + 'learning_rate': 1.0447168968172163e-05, + 'epoch': 7.54} +04/19 [22:13:22] INFO | >> train_qwenlatent.py:487 + Step 29890 | grad_norm_pre_clip=0.1840 | + grad_norm_pre_clip_avg=0.1806 | Metrics: + {'align_loss': 0.024280216544866562, + 'recon_loss': 0.096237912774086, + 'predict_loss': 0.008206045255064964, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1839810311794281, + 'data_time': 0.0010290170030202717, + 'model_time': 1.2177059610257857, + 'grad_norm_pre_clip_avg': 0.1806051954627037, + 'learning_rate': 1.0438571616373668e-05, + 'epoch': 7.54} +04/19 [22:13:36] INFO | >> train_qwenlatent.py:487 + Step 29900 | grad_norm_pre_clip=0.1242 | + grad_norm_pre_clip_avg=0.1662 | Metrics: + {'align_loss': 0.025531545281410217, + 'recon_loss': 0.09550883620977402, + 'predict_loss': 0.00746186776086688, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12424371391534805, + 'mae_score': 0.012144278620814419, 'data_time': + 0.0011250110110267997, 'model_time': + 1.3033937439904548, 'grad_norm_pre_clip_avg': + 0.16619856804609298, 'learning_rate': + 1.0429975275382675e-05, 'epoch': 7.54} +04/19 [22:13:49] INFO | >> train_qwenlatent.py:487 + Step 29910 | grad_norm_pre_clip=0.1596 | + grad_norm_pre_clip_avg=0.1616 | Metrics: + {'align_loss': 0.024253107607364655, + 'recon_loss': 0.11022060364484787, + 'predict_loss': 0.008779110386967659, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15962986648082733, + 'data_time': 0.0015776710060890764, + 'model_time': 1.2064456020016223, + 'grad_norm_pre_clip_avg': 0.1616437777876854, + 'learning_rate': 1.0421379949388931e-05, + 'epoch': 7.55} +04/19 [22:14:01] INFO | >> train_qwenlatent.py:487 + Step 29920 | grad_norm_pre_clip=0.2069 | + grad_norm_pre_clip_avg=0.1720 | Metrics: + {'align_loss': 0.024669159203767776, + 'recon_loss': 0.11766759306192398, + 'predict_loss': 0.009884216822683811, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2068668156862259, + 'data_time': 0.0007173250196501613, + 'model_time': 1.2246906829823274, + 'grad_norm_pre_clip_avg': 0.17204275131225585, + 'learning_rate': 1.0412785642581698e-05, + 'epoch': 7.55} +04/19 [22:14:13] INFO | >> train_qwenlatent.py:487 + Step 29930 | grad_norm_pre_clip=0.2249 | + grad_norm_pre_clip_avg=0.1825 | Metrics: + {'align_loss': 0.0249972902238369, + 'recon_loss': 0.06360191851854324, + 'predict_loss': 0.005721187684684992, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2249295711517334, + 'data_time': 0.0008732140122447163, + 'model_time': 1.2261540830077138, + 'grad_norm_pre_clip_avg': 0.1824956476688385, + 'learning_rate': 1.0404192359149736e-05, + 'epoch': 7.55} +04/19 [22:14:26] INFO | >> train_qwenlatent.py:487 + Step 29940 | grad_norm_pre_clip=0.1595 | + grad_norm_pre_clip_avg=0.1871 | Metrics: + {'align_loss': 0.026042910292744637, + 'recon_loss': 0.10796414315700531, + 'predict_loss': 0.006090645678341389, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15946856141090393, + 'data_time': 0.0007597910007461905, + 'model_time': 1.1864747370127589, + 'grad_norm_pre_clip_avg': 0.1870611011981964, + 'learning_rate': 1.039560010328131e-05, + 'epoch': 7.55} +04/19 [22:14:39] INFO | >> train_qwenlatent.py:487 + Step 29950 | grad_norm_pre_clip=0.2180 | + grad_norm_pre_clip_avg=0.1758 | Metrics: + {'align_loss': 0.02506747841835022, + 'recon_loss': 0.09288445115089417, + 'predict_loss': 0.005001395009458065, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21798647940158844, + 'mae_score': 0.008269857286332965, 'data_time': + 0.0009495359845459461, 'model_time': + 1.2295234400080517, 'grad_norm_pre_clip_avg': + 0.1758139729499817, 'learning_rate': + 1.0387008879164172e-05, 'epoch': 7.56} +04/19 [22:14:52] INFO | >> train_qwenlatent.py:487 + Step 29960 | grad_norm_pre_clip=0.2223 | + grad_norm_pre_clip_avg=0.1893 | Metrics: + {'align_loss': 0.025066997855901718, + 'recon_loss': 0.10407168418169022, + 'predict_loss': 0.010119756683707237, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22234861552715302, + 'data_time': 0.0007071370200719684, + 'model_time': 1.1941237779974472, + 'grad_norm_pre_clip_avg': 0.18927029967308046, + 'learning_rate': 1.0378418690985585e-05, + 'epoch': 7.56} +04/19 [22:15:04] INFO | >> train_qwenlatent.py:487 + Step 29970 | grad_norm_pre_clip=0.1713 | + grad_norm_pre_clip_avg=0.1677 | Metrics: + {'align_loss': 0.02618657983839512, + 'recon_loss': 0.08739887177944183, + 'predict_loss': 0.006380061153322458, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1713304966688156, + 'data_time': 0.0009470759832765907, + 'model_time': 1.2223477849911433, + 'grad_norm_pre_clip_avg': 0.1676784336566925, + 'learning_rate': 1.0369829542932303e-05, + 'epoch': 7.56} +04/19 [22:15:17] INFO | >> train_qwenlatent.py:487 + Step 29980 | grad_norm_pre_clip=0.1865 | + grad_norm_pre_clip_avg=0.1683 | Metrics: + {'align_loss': 0.023715488612651825, + 'recon_loss': 0.10982981324195862, + 'predict_loss': 0.012451515533030033, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1865198165178299, + 'data_time': 0.000931700982619077, + 'model_time': 1.276238024001941, + 'grad_norm_pre_clip_avg': 0.16833370178937912, + 'learning_rate': 1.0361241439190577e-05, + 'epoch': 7.56} +04/19 [22:15:29] INFO | >> train_qwenlatent.py:487 + Step 29990 | grad_norm_pre_clip=0.1542 | + grad_norm_pre_clip_avg=0.1808 | Metrics: + {'align_loss': 0.026600055396556854, + 'recon_loss': 0.11475017666816711, + 'predict_loss': 0.006267933640629053, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1542218029499054, + 'data_time': 0.000764869007980451, + 'model_time': 1.2758366570051294, + 'grad_norm_pre_clip_avg': 0.18081336319446564, + 'learning_rate': 1.0352654383946134e-05, + 'epoch': 7.57} +04/19 [22:15:43] INFO | >> train_qwenlatent.py:487 + Step 30000 | grad_norm_pre_clip=0.1431 | + grad_norm_pre_clip_avg=0.1788 | Metrics: + {'align_loss': 0.024716254323720932, + 'recon_loss': 0.10737498104572296, + 'predict_loss': 0.01150090154260397, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14306744933128357, + 'mae_score': 0.008914454348452457, 'data_time': + 0.0009646499820519239, 'model_time': + 1.2478587189980317, 'grad_norm_pre_clip_avg': + 0.1788310945034027, 'learning_rate': + 1.0344068381384208e-05, 'epoch': 7.57} +✅ Checkpoint saved at ./runs/0418_QwenLatent_13tasks_actionstate_30k/checkpoints/steps_30000 +04/19 [22:16:05] INFO | >> train_qwenlatent.py:487 + Step 30010 | grad_norm_pre_clip=0.1438 | + grad_norm_pre_clip_avg=0.1815 | Metrics: + {'align_loss': 0.025194237008690834, + 'recon_loss': 0.12277021259069443, + 'predict_loss': 0.011017641983926296, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14379987120628357, + 'data_time': 0.0006550950056407601, + 'model_time': 1.3271864909911528, + 'grad_norm_pre_clip_avg': 0.18149486780166627, + 'learning_rate': 1.0335483435689507e-05, + 'epoch': 7.57} +04/19 [22:16:19] INFO | >> train_qwenlatent.py:487 + Step 30020 | grad_norm_pre_clip=0.2131 | + grad_norm_pre_clip_avg=0.1867 | Metrics: + {'align_loss': 0.02636599913239479, + 'recon_loss': 0.12010429799556732, + 'predict_loss': 0.007813036441802979, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21314743161201477, + 'data_time': 0.0009860539867077023, + 'model_time': 1.2451541509944946, + 'grad_norm_pre_clip_avg': 0.18670256733894347, + 'learning_rate': 1.032689955104624e-05, + 'epoch': 7.58} +04/19 [22:16:32] INFO | >> train_qwenlatent.py:487 + Step 30030 | grad_norm_pre_clip=0.1283 | + grad_norm_pre_clip_avg=0.1827 | Metrics: + {'align_loss': 0.025844309478998184, + 'recon_loss': 0.10528723895549774, + 'predict_loss': 0.0070594060234725475, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1283106505870819, + 'data_time': 0.0006401519931387156, + 'model_time': 1.2487116359989159, + 'grad_norm_pre_clip_avg': 0.182687708735466, + 'learning_rate': 1.031831673163808e-05, + 'epoch': 7.58} +04/19 [22:16:44] INFO | >> train_qwenlatent.py:487 + Step 30040 | grad_norm_pre_clip=0.2444 | + grad_norm_pre_clip_avg=0.1806 | Metrics: + {'align_loss': 0.026265477761626244, + 'recon_loss': 0.09142518788576126, + 'predict_loss': 0.00844308640807867, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24441738426685333, + 'data_time': 0.0009135769796557724, + 'model_time': 1.2014495730109047, + 'grad_norm_pre_clip_avg': 0.1806414544582367, + 'learning_rate': 1.0309734981648187e-05, + 'epoch': 7.58} +04/19 [22:16:57] INFO | >> train_qwenlatent.py:487 + Step 30050 | grad_norm_pre_clip=0.1662 | + grad_norm_pre_clip_avg=0.1713 | Metrics: + {'align_loss': 0.02551971934735775, + 'recon_loss': 0.11754529923200607, + 'predict_loss': 0.010172836482524872, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16618037223815918, + 'mae_score': 0.00787655512491862, 'data_time': + 0.0008598670247010887, 'model_time': + 1.290734184003668, 'grad_norm_pre_clip_avg': + 0.17126891911029815, 'learning_rate': + 1.0301154305259205e-05, 'epoch': 7.58} +04/19 [22:17:10] INFO | >> train_qwenlatent.py:487 + Step 30060 | grad_norm_pre_clip=0.2065 | + grad_norm_pre_clip_avg=0.1788 | Metrics: + {'align_loss': 0.02496635913848877, + 'recon_loss': 0.09845968335866928, + 'predict_loss': 0.009022032842040062, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20654359459877014, + 'data_time': 0.0012048929929733276, + 'model_time': 1.3064310729969293, + 'grad_norm_pre_clip_avg': 0.1787918120622635, + 'learning_rate': 1.0292574706653251e-05, + 'epoch': 7.59} +04/19 [22:17:22] INFO | >> train_qwenlatent.py:487 + Step 30070 | grad_norm_pre_clip=0.2352 | + grad_norm_pre_clip_avg=0.2086 | Metrics: + {'align_loss': 0.02539706602692604, + 'recon_loss': 0.1295291930437088, + 'predict_loss': 0.0111994668841362, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2351829707622528, + 'data_time': 0.0007069279963616282, + 'model_time': 1.2156594110128935, + 'grad_norm_pre_clip_avg': 0.20860849767923356, + 'learning_rate': 1.0283996190011917e-05, + 'epoch': 7.59} +04/19 [22:17:35] INFO | >> train_qwenlatent.py:487 + Step 30080 | grad_norm_pre_clip=0.2040 | + grad_norm_pre_clip_avg=0.1822 | Metrics: + {'align_loss': 0.026824550703167915, + 'recon_loss': 0.0959448590874672, + 'predict_loss': 0.015753790736198425, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20400303602218628, + 'data_time': 0.000957678013946861, + 'model_time': 1.2198643030133098, + 'grad_norm_pre_clip_avg': 0.1822104349732399, + 'learning_rate': 1.0275418759516266e-05, + 'epoch': 7.59} +04/19 [22:17:47] INFO | >> train_qwenlatent.py:487 + Step 30090 | grad_norm_pre_clip=0.1907 | + grad_norm_pre_clip_avg=0.1795 | Metrics: + {'align_loss': 0.025187503546476364, + 'recon_loss': 0.09092605113983154, + 'predict_loss': 0.007174310274422169, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19065652787685394, + 'data_time': 0.0007617710216436535, + 'model_time': 1.2264143840002362, + 'grad_norm_pre_clip_avg': 0.17946574985980987, + 'learning_rate': 1.0266842419346838e-05, + 'epoch': 7.59} +04/19 [22:18:01] INFO | >> train_qwenlatent.py:487 + Step 30100 | grad_norm_pre_clip=0.1767 | + grad_norm_pre_clip_avg=0.1591 | Metrics: + {'align_loss': 0.027101755142211914, + 'recon_loss': 0.1292862445116043, + 'predict_loss': 0.014354494400322437, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17666998505592346, + 'mae_score': 0.010635861405381211, 'data_time': + 0.0008193980029318482, 'model_time': + 1.2680353409959935, 'grad_norm_pre_clip_avg': + 0.15914427042007445, 'learning_rate': + 1.0258267173683633e-05, 'epoch': 7.6} +04/19 [22:18:13] INFO | >> train_qwenlatent.py:487 + Step 30110 | grad_norm_pre_clip=0.1433 | + grad_norm_pre_clip_avg=0.1554 | Metrics: + {'align_loss': 0.025562888011336327, + 'recon_loss': 0.13719740509986877, + 'predict_loss': 0.013153387233614922, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1433473825454712, + 'data_time': 0.0010675539961084723, + 'model_time': 1.2465593009837903, + 'grad_norm_pre_clip_avg': 0.15539101064205169, + 'learning_rate': 1.024969302670612e-05, + 'epoch': 7.6} +04/19 [22:18:26] INFO | >> train_qwenlatent.py:487 + Step 30120 | grad_norm_pre_clip=0.1363 | + grad_norm_pre_clip_avg=0.1870 | Metrics: + {'align_loss': 0.024350158870220184, + 'recon_loss': 0.045908741652965546, + 'predict_loss': 0.003893244080245495, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13633550703525543, + 'data_time': 0.0008244760101661086, + 'model_time': 1.2479092339926865, + 'grad_norm_pre_clip_avg': 0.18700454980134965, + 'learning_rate': 1.0241119982593245e-05, + 'epoch': 7.6} +04/19 [22:18:38] INFO | >> train_qwenlatent.py:487 + Step 30130 | grad_norm_pre_clip=0.1549 | + grad_norm_pre_clip_avg=0.1621 | Metrics: + {'align_loss': 0.024950455874204636, + 'recon_loss': 0.08734149485826492, + 'predict_loss': 0.009850181639194489, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15494875609874725, + 'data_time': 0.0010587880096863955, + 'model_time': 1.260953535005683, + 'grad_norm_pre_clip_avg': 0.1620667278766632, + 'learning_rate': 1.0232548045523391e-05, + 'epoch': 7.6} +04/19 [22:18:51] INFO | >> train_qwenlatent.py:487 + Step 30140 | grad_norm_pre_clip=0.1847 | + grad_norm_pre_clip_avg=0.1583 | Metrics: + {'align_loss': 0.02512449398636818, + 'recon_loss': 0.09145413339138031, + 'predict_loss': 0.007305028382688761, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18473045527935028, + 'data_time': 0.0009310779860243201, + 'model_time': 1.2794004369934555, + 'grad_norm_pre_clip_avg': 0.15827866792678832, + 'learning_rate': 1.0223977219674425e-05, + 'epoch': 7.61} +04/19 [22:19:04] INFO | >> train_qwenlatent.py:487 + Step 30150 | grad_norm_pre_clip=0.2207 | + grad_norm_pre_clip_avg=0.1837 | Metrics: + {'align_loss': 0.023467741906642914, + 'recon_loss': 0.0879564955830574, + 'predict_loss': 0.009041602723300457, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22069020569324493, + 'mae_score': 0.008731055044913076, 'data_time': + 0.0007677710091229528, 'model_time': + 1.2309452919871546, 'grad_norm_pre_clip_avg': + 0.18373644053936006, 'learning_rate': + 1.0215407509223656e-05, 'epoch': 7.61} +04/19 [22:19:17] INFO | >> train_qwenlatent.py:487 + Step 30160 | grad_norm_pre_clip=0.1750 | + grad_norm_pre_clip_avg=0.1688 | Metrics: + {'align_loss': 0.026214562356472015, + 'recon_loss': 0.12587162852287292, + 'predict_loss': 0.007769539952278137, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1749657839536667, + 'data_time': 0.0010423540079500526, + 'model_time': 1.1847038850246463, + 'grad_norm_pre_clip_avg': 0.16883225589990616, + 'learning_rate': 1.0206838918347867e-05, + 'epoch': 7.61} +04/19 [22:19:30] INFO | >> train_qwenlatent.py:487 + Step 30170 | grad_norm_pre_clip=0.1853 | + grad_norm_pre_clip_avg=0.1824 | Metrics: + {'align_loss': 0.025882763788104057, + 'recon_loss': 0.0891244113445282, + 'predict_loss': 0.0072755711153149605, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18525268137454987, + 'data_time': 0.0006968950037844479, + 'model_time': 1.2271341140149161, + 'grad_norm_pre_clip_avg': 0.18241746127605438, + 'learning_rate': 1.019827145122328e-05, + 'epoch': 7.61} +04/19 [22:19:42] INFO | >> train_qwenlatent.py:487 + Step 30180 | grad_norm_pre_clip=0.2537 | + grad_norm_pre_clip_avg=0.2055 | Metrics: + {'align_loss': 0.023829149082303047, + 'recon_loss': 0.13787402212619781, + 'predict_loss': 0.013278291560709476, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2537311315536499, + 'data_time': 0.0009105729986913502, + 'model_time': 1.2165603800094686, + 'grad_norm_pre_clip_avg': 0.2055041566491127, + 'learning_rate': 1.018970511202557e-05, + 'epoch': 7.62} +04/19 [22:19:55] INFO | >> train_qwenlatent.py:487 + Step 30190 | grad_norm_pre_clip=0.1659 | + grad_norm_pre_clip_avg=0.1916 | Metrics: + {'align_loss': 0.023725133389234543, + 'recon_loss': 0.07928194105625153, + 'predict_loss': 0.006565121468156576, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16587790846824646, + 'data_time': 0.00069130101474002, 'model_time': + 1.2221866710169706, 'grad_norm_pre_clip_avg': + 0.19157131165266036, 'learning_rate': + 1.0181139904929872e-05, 'epoch': 7.62} +04/19 [22:20:08] INFO | >> train_qwenlatent.py:487 + Step 30200 | grad_norm_pre_clip=0.1898 | + grad_norm_pre_clip_avg=0.1776 | Metrics: + {'align_loss': 0.02511155977845192, + 'recon_loss': 0.08598322421312332, + 'predict_loss': 0.007710652891546488, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18984362483024597, + 'mae_score': 0.011057452468184738, 'data_time': + 0.0007940309878904372, 'model_time': + 1.1840269789972808, 'grad_norm_pre_clip_avg': + 0.17756618708372116, 'learning_rate': + 1.0172575834110761e-05, 'epoch': 7.62} +04/19 [22:20:20] INFO | >> train_qwenlatent.py:487 + Step 30210 | grad_norm_pre_clip=0.1590 | + grad_norm_pre_clip_avg=0.1773 | Metrics: + {'align_loss': 0.02544301189482212, + 'recon_loss': 0.09786740690469742, + 'predict_loss': 0.009751254692673683, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15901842713356018, + 'data_time': 0.0011532860225997865, + 'model_time': 1.2566121890267823, + 'grad_norm_pre_clip_avg': 0.17730439454317093, + 'learning_rate': 1.0164012903742271e-05, + 'epoch': 7.62} +04/19 [22:20:34] INFO | >> train_qwenlatent.py:487 + Step 30220 | grad_norm_pre_clip=0.1698 | + grad_norm_pre_clip_avg=0.1476 | Metrics: + {'align_loss': 0.025538370013237, 'recon_loss': + 0.09227488934993744, 'predict_loss': + 0.006640864536166191, 'aux_loss_decay_weight': + 0.0, 'grad_norm_pre_clip': 0.1697852909564972, + 'data_time': 0.0009472310193814337, + 'model_time': 1.2074274939950556, + 'grad_norm_pre_clip_avg': 0.14756492972373964, + 'learning_rate': 1.0155451117997855e-05, + 'epoch': 7.63} +04/19 [22:20:46] INFO | >> train_qwenlatent.py:487 + Step 30230 | grad_norm_pre_clip=0.1765 | + grad_norm_pre_clip_avg=0.1455 | Metrics: + {'align_loss': 0.025449268519878387, + 'recon_loss': 0.09334614127874374, + 'predict_loss': 0.0060040210373699665, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17649787664413452, + 'data_time': 0.0010182380210608244, + 'model_time': 1.2193330299924128, + 'grad_norm_pre_clip_avg': 0.14550384879112244, + 'learning_rate': 1.0146890481050433e-05, + 'epoch': 7.63} +04/19 [22:20:58] INFO | >> train_qwenlatent.py:487 + Step 30240 | grad_norm_pre_clip=0.2063 | + grad_norm_pre_clip_avg=0.1581 | Metrics: + {'align_loss': 0.026788506656885147, + 'recon_loss': 0.10440590977668762, + 'predict_loss': 0.010632778517901897, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20626680552959442, + 'data_time': 0.0009195210004691035, + 'model_time': 1.233681790996343, + 'grad_norm_pre_clip_avg': 0.1580523930490017, + 'learning_rate': 1.0138330997072352e-05, + 'epoch': 7.63} +04/19 [22:21:11] INFO | >> train_qwenlatent.py:487 + Step 30250 | grad_norm_pre_clip=0.2716 | + grad_norm_pre_clip_avg=0.1768 | Metrics: + {'align_loss': 0.02574728988111019, + 'recon_loss': 0.08335161954164505, + 'predict_loss': 0.008431642316281796, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2716210186481476, + 'mae_score': 0.010575837487573023, 'data_time': + 0.000973474991042167, 'model_time': + 1.2055253479920793, 'grad_norm_pre_clip_avg': + 0.176821206510067, 'learning_rate': + 1.0129772670235397e-05, 'epoch': 7.63} +04/19 [22:21:24] INFO | >> train_qwenlatent.py:487 + Step 30260 | grad_norm_pre_clip=0.2899 | + grad_norm_pre_clip_avg=0.1950 | Metrics: + {'align_loss': 0.026160581037402153, + 'recon_loss': 0.0981978178024292, + 'predict_loss': 0.00684175081551075, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.289859801530838, + 'data_time': 0.0006871350051369518, + 'model_time': 1.1771800369897392, + 'grad_norm_pre_clip_avg': 0.1950085550546646, + 'learning_rate': 1.0121215504710802e-05, + 'epoch': 7.64} +04/19 [22:21:36] INFO | >> train_qwenlatent.py:487 + Step 30270 | grad_norm_pre_clip=0.1839 | + grad_norm_pre_clip_avg=0.1924 | Metrics: + {'align_loss': 0.025914590805768967, + 'recon_loss': 0.09653332829475403, + 'predict_loss': 0.0059609198942780495, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1838725209236145, + 'data_time': 0.0006913240067660809, + 'model_time': 1.2874004550103564, + 'grad_norm_pre_clip_avg': 0.1924145847558975, + 'learning_rate': 1.0112659504669215e-05, + 'epoch': 7.64} +04/19 [22:21:49] INFO | >> train_qwenlatent.py:487 + Step 30280 | grad_norm_pre_clip=0.1644 | + grad_norm_pre_clip_avg=0.1772 | Metrics: + {'align_loss': 0.025198403745889664, + 'recon_loss': 0.0996067076921463, + 'predict_loss': 0.007325299549847841, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1644347608089447, + 'data_time': 0.0008471740002278239, + 'model_time': 1.2169848130142782, + 'grad_norm_pre_clip_avg': 0.17716638147830963, + 'learning_rate': 1.0104104674280737e-05, + 'epoch': 7.64} +04/19 [22:22:02] INFO | >> train_qwenlatent.py:487 + Step 30290 | grad_norm_pre_clip=0.2089 | + grad_norm_pre_clip_avg=0.1824 | Metrics: + {'align_loss': 0.025088636204600334, + 'recon_loss': 0.09929598122835159, + 'predict_loss': 0.008588188327848911, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.208908349275589, + 'data_time': 0.0009006480104289949, + 'model_time': 1.260767803993076, + 'grad_norm_pre_clip_avg': 0.18239434361457824, + 'learning_rate': 1.0095551017714882e-05, + 'epoch': 7.64} +04/19 [22:22:16] INFO | >> train_qwenlatent.py:487 + Step 30300 | grad_norm_pre_clip=0.1562 | + grad_norm_pre_clip_avg=0.1822 | Metrics: + {'align_loss': 0.024054650217294693, + 'recon_loss': 0.09575044363737106, + 'predict_loss': 0.008322724141180515, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15618957579135895, + 'mae_score': 0.00808550602680928, 'data_time': + 0.0006310110038612038, 'model_time': + 1.2333276410063263, 'grad_norm_pre_clip_avg': + 0.18217140585184097, 'learning_rate': + 1.0086998539140602e-05, 'epoch': 7.65} +04/19 [22:22:29] INFO | >> train_qwenlatent.py:487 + Step 30310 | grad_norm_pre_clip=0.1447 | + grad_norm_pre_clip_avg=0.1636 | Metrics: + {'align_loss': 0.025159146636724472, + 'recon_loss': 0.09270815551280975, + 'predict_loss': 0.009083965793251991, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1447143852710724, + 'data_time': 0.001287201012019068, + 'model_time': 1.2264747160079423, + 'grad_norm_pre_clip_avg': 0.16358987241983414, + 'learning_rate': 1.0078447242726265e-05, + 'epoch': 7.65} +04/19 [22:22:41] INFO | >> train_qwenlatent.py:487 + Step 30320 | grad_norm_pre_clip=0.1664 | + grad_norm_pre_clip_avg=0.1664 | Metrics: + {'align_loss': 0.025413133203983307, + 'recon_loss': 0.0792839527130127, + 'predict_loss': 0.007008661516010761, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1663997322320938, + 'data_time': 0.001015399000607431, + 'model_time': 1.2264682799868751, + 'grad_norm_pre_clip_avg': 0.1663545176386833, + 'learning_rate': 1.0069897132639674e-05, + 'epoch': 7.65} +04/19 [22:22:53] INFO | >> train_qwenlatent.py:487 + Step 30330 | grad_norm_pre_clip=0.2566 | + grad_norm_pre_clip_avg=0.1823 | Metrics: + {'align_loss': 0.02495207078754902, + 'recon_loss': 0.12104904651641846, + 'predict_loss': 0.01248371135443449, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2566470205783844, + 'data_time': 0.0008685220091138035, + 'model_time': 1.2251536570256576, + 'grad_norm_pre_clip_avg': 0.1823166161775589, + 'learning_rate': 1.0061348213048049e-05, + 'epoch': 7.65} +04/19 [22:23:06] INFO | >> train_qwenlatent.py:487 + Step 30340 | grad_norm_pre_clip=0.1989 | + grad_norm_pre_clip_avg=0.1930 | Metrics: + {'align_loss': 0.025959685444831848, + 'recon_loss': 0.09681630879640579, + 'predict_loss': 0.007818466052412987, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19894663989543915, + 'data_time': 0.0009095850109588355, + 'model_time': 1.2287741049949545, + 'grad_norm_pre_clip_avg': 0.19299003183841706, + 'learning_rate': 1.0052800488118023e-05, + 'epoch': 7.66} +04/19 [22:23:19] INFO | >> train_qwenlatent.py:487 + Step 30350 | grad_norm_pre_clip=0.1725 | + grad_norm_pre_clip_avg=0.1850 | Metrics: + {'align_loss': 0.02549321949481964, + 'recon_loss': 0.12190989404916763, + 'predict_loss': 0.01196422427892685, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1725156158208847, + 'mae_score': 0.008696253664858706, 'data_time': + 0.0012116309953853488, 'model_time': + 1.214126901992131, 'grad_norm_pre_clip_avg': + 0.18500387519598008, 'learning_rate': + 1.0044253962015662e-05, 'epoch': 7.66} +04/19 [22:23:32] INFO | >> train_qwenlatent.py:487 + Step 30360 | grad_norm_pre_clip=0.1185 | + grad_norm_pre_clip_avg=0.1721 | Metrics: + {'align_loss': 0.026742547750473022, + 'recon_loss': 0.11326414346694946, + 'predict_loss': 0.011480453424155712, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11847898364067078, + 'data_time': 0.0007912099827080965, + 'model_time': 1.2415251130005345, + 'grad_norm_pre_clip_avg': 0.17209866791963577, + 'learning_rate': 1.0035708638906437e-05, + 'epoch': 7.66} +04/19 [22:23:45] INFO | >> train_qwenlatent.py:487 + Step 30370 | grad_norm_pre_clip=0.1621 | + grad_norm_pre_clip_avg=0.1820 | Metrics: + {'align_loss': 0.026324141770601273, + 'recon_loss': 0.12580245733261108, + 'predict_loss': 0.00833102222532034, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16212010383605957, + 'data_time': 0.0008351879951078445, + 'model_time': 1.2621420649811625, + 'grad_norm_pre_clip_avg': 0.1820414811372757, + 'learning_rate': 1.0027164522955227e-05, + 'epoch': 7.66} +04/19 [22:23:57] INFO | >> train_qwenlatent.py:487 + Step 30380 | grad_norm_pre_clip=0.1830 | + grad_norm_pre_clip_avg=0.1792 | Metrics: + {'align_loss': 0.023641493171453476, + 'recon_loss': 0.0662788599729538, + 'predict_loss': 0.011947841383516788, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18298541009426117, + 'data_time': 0.0006943990010768175, + 'model_time': 1.1994015999953263, + 'grad_norm_pre_clip_avg': 0.17924998551607133, + 'learning_rate': 1.0018621618326345e-05, + 'epoch': 7.67} +04/19 [22:24:09] INFO | >> train_qwenlatent.py:487 + Step 30390 | grad_norm_pre_clip=0.1841 | + grad_norm_pre_clip_avg=0.1749 | Metrics: + {'align_loss': 0.02731422707438469, + 'recon_loss': 0.1007203534245491, + 'predict_loss': 0.010903656482696533, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18405838310718536, + 'data_time': 0.0006503589975181967, + 'model_time': 1.238524081010837, + 'grad_norm_pre_clip_avg': 0.17485705241560937, + 'learning_rate': 1.0010079929183486e-05, + 'epoch': 7.67} +04/19 [22:24:23] INFO | >> train_qwenlatent.py:487 + Step 30400 | grad_norm_pre_clip=0.1528 | + grad_norm_pre_clip_avg=0.1756 | Metrics: + {'align_loss': 0.025051306933164597, + 'recon_loss': 0.10282747447490692, + 'predict_loss': 0.007385889068245888, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15282420814037323, + 'mae_score': 0.009755349374032236, 'data_time': + 0.0006987790111452341, 'model_time': + 1.2226785729872063, 'grad_norm_pre_clip_avg': + 0.17560398429632187, 'learning_rate': + 1.000153945968977e-05, 'epoch': 7.67} +04/19 [22:24:35] INFO | >> train_qwenlatent.py:487 + Step 30410 | grad_norm_pre_clip=0.1982 | + grad_norm_pre_clip_avg=0.1946 | Metrics: + {'align_loss': 0.024640459567308426, + 'recon_loss': 0.0594751313328743, + 'predict_loss': 0.0070011611096560955, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1981973946094513, + 'data_time': 0.0008899670210666955, + 'model_time': 1.27343739598291, + 'grad_norm_pre_clip_avg': 0.19464141577482225, + 'learning_rate': 9.993000214007715e-06, + 'epoch': 7.67} +04/19 [22:24:49] INFO | >> train_qwenlatent.py:487 + Step 30420 | grad_norm_pre_clip=0.2007 | + grad_norm_pre_clip_avg=0.1859 | Metrics: + {'align_loss': 0.0247326772660017, + 'recon_loss': 0.12246400862932205, + 'predict_loss': 0.007603210862725973, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20067349076271057, + 'data_time': 0.0010798680013976991, + 'model_time': 1.4940534639754333, + 'grad_norm_pre_clip_avg': 0.18587743043899535, + 'learning_rate': 9.98446219629925e-06, 'epoch': + 7.68} +04/19 [22:25:01] INFO | >> train_qwenlatent.py:487 + Step 30430 | grad_norm_pre_clip=0.1807 | + grad_norm_pre_clip_avg=0.1881 | Metrics: + {'align_loss': 0.02664952352643013, + 'recon_loss': 0.12372271716594696, + 'predict_loss': 0.01142613310366869, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18066570162773132, + 'data_time': 0.0010681879939511418, + 'model_time': 1.2412254279770423, + 'grad_norm_pre_clip_avg': 0.18812211900949477, + 'learning_rate': 9.975925410725702e-06, + 'epoch': 7.68} +04/19 [22:25:14] INFO | >> train_qwenlatent.py:487 + Step 30440 | grad_norm_pre_clip=0.2219 | + grad_norm_pre_clip_avg=0.1683 | Metrics: + {'align_loss': 0.025934908539056778, + 'recon_loss': 0.11091506481170654, + 'predict_loss': 0.012477695941925049, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22190353274345398, + 'data_time': 0.0010165569838136435, + 'model_time': 1.272461277025286, + 'grad_norm_pre_clip_avg': 0.1683163285255432, + 'learning_rate': 9.967389861447788e-06, + 'epoch': 7.68} +04/19 [22:25:27] INFO | >> train_qwenlatent.py:487 + Step 30450 | grad_norm_pre_clip=0.1776 | + grad_norm_pre_clip_avg=0.1649 | Metrics: + {'align_loss': 0.024378377944231033, + 'recon_loss': 0.09923268854618073, + 'predict_loss': 0.00583928357809782, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17756231129169464, + 'mae_score': 0.01070948076677752, 'data_time': + 0.0009751449979376048, 'model_time': + 1.242636737995781, 'grad_norm_pre_clip_avg': + 0.16492287963628768, 'learning_rate': + 9.958855552625644e-06, 'epoch': 7.68} +04/19 [22:25:40] INFO | >> train_qwenlatent.py:487 + Step 30460 | grad_norm_pre_clip=0.1423 | + grad_norm_pre_clip_avg=0.1569 | Metrics: + {'align_loss': 0.02521698549389839, + 'recon_loss': 0.0708647221326828, + 'predict_loss': 0.005483592394739389, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14230768382549286, + 'data_time': 0.0007578649965580553, + 'model_time': 1.2041006349900272, + 'grad_norm_pre_clip_avg': 0.15689375400543212, + 'learning_rate': 9.950322488418778e-06, + 'epoch': 7.69} +04/19 [22:25:52] INFO | >> train_qwenlatent.py:487 + Step 30470 | grad_norm_pre_clip=0.1559 | + grad_norm_pre_clip_avg=0.1597 | Metrics: + {'align_loss': 0.02501637302339077, + 'recon_loss': 0.11214505136013031, + 'predict_loss': 0.009636495262384415, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.155923530459404, + 'data_time': 0.0009564290230628103, + 'model_time': 1.2274858399759978, + 'grad_norm_pre_clip_avg': 0.15973284393548964, + 'learning_rate': 9.941790672986111e-06, + 'epoch': 7.69} +04/19 [22:26:05] INFO | >> train_qwenlatent.py:487 + Step 30480 | grad_norm_pre_clip=0.1552 | + grad_norm_pre_clip_avg=0.1611 | Metrics: + {'align_loss': 0.025038480758666992, + 'recon_loss': 0.13358443975448608, + 'predict_loss': 0.017566252499818802, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15518872439861298, + 'data_time': 0.0008885259740054607, + 'model_time': 1.2196139330044389, + 'grad_norm_pre_clip_avg': 0.16112214028835298, + 'learning_rate': 9.933260110485935e-06, + 'epoch': 7.69} +04/19 [22:26:17] INFO | >> train_qwenlatent.py:487 + Step 30490 | grad_norm_pre_clip=0.2059 | + grad_norm_pre_clip_avg=0.1957 | Metrics: + {'align_loss': 0.02551165781915188, + 'recon_loss': 0.12696395814418793, + 'predict_loss': 0.010886010713875294, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20586472749710083, + 'data_time': 0.0009538389858789742, + 'model_time': 1.253834878996713, + 'grad_norm_pre_clip_avg': 0.1957198515534401, + 'learning_rate': 9.92473080507595e-06, 'epoch': + 7.69} +04/19 [22:26:30] INFO | >> train_qwenlatent.py:487 + Step 30500 | grad_norm_pre_clip=0.1634 | + grad_norm_pre_clip_avg=0.2000 | Metrics: + {'align_loss': 0.024607449769973755, + 'recon_loss': 0.05598972737789154, + 'predict_loss': 0.006189123727381229, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16338245570659637, + 'mae_score': 0.008923541747771942, 'data_time': + 0.000955166993662715, 'model_time': + 1.2162887400190812, 'grad_norm_pre_clip_avg': + 0.20003734529018402, 'learning_rate': + 9.916202760913233e-06, 'epoch': 7.7} +04/19 [22:26:43] INFO | >> train_qwenlatent.py:487 + Step 30510 | grad_norm_pre_clip=0.1479 | + grad_norm_pre_clip_avg=0.1953 | Metrics: + {'align_loss': 0.02578975446522236, + 'recon_loss': 0.09408940374851227, + 'predict_loss': 0.007853214628994465, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14794977009296417, + 'data_time': 0.0007698330155108124, + 'model_time': 1.2580809790233616, + 'grad_norm_pre_clip_avg': 0.19529150426387787, + 'learning_rate': 9.907675982154248e-06, + 'epoch': 7.7} +04/19 [22:26:55] INFO | >> train_qwenlatent.py:487 + Step 30520 | grad_norm_pre_clip=0.2120 | + grad_norm_pre_clip_avg=0.1793 | Metrics: + {'align_loss': 0.023684274405241013, + 'recon_loss': 0.09623237699270248, + 'predict_loss': 0.008502679876983166, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21198418736457825, + 'data_time': 0.0010639590036589652, + 'model_time': 1.2088573070068378, + 'grad_norm_pre_clip_avg': 0.1792710855603218, + 'learning_rate': 9.89915047295485e-06, 'epoch': + 7.7} +04/19 [22:27:08] INFO | >> train_qwenlatent.py:487 + Step 30530 | grad_norm_pre_clip=0.1672 | + grad_norm_pre_clip_avg=0.1558 | Metrics: + {'align_loss': 0.0252214465290308, + 'recon_loss': 0.07462777942419052, + 'predict_loss': 0.007287558168172836, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16723516583442688, + 'data_time': 0.001431571989087388, + 'model_time': 1.257490421005059, + 'grad_norm_pre_clip_avg': 0.15578342378139495, + 'learning_rate': 9.89062623747026e-06, 'epoch': + 7.7} +04/19 [22:27:20] INFO | >> train_qwenlatent.py:487 + Step 30540 | grad_norm_pre_clip=0.2031 | + grad_norm_pre_clip_avg=0.1491 | Metrics: + {'align_loss': 0.024024605751037598, + 'recon_loss': 0.07307496666908264, + 'predict_loss': 0.008190825581550598, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20308415591716766, + 'data_time': 0.0009244869870599359, + 'model_time': 1.2484108950011432, + 'grad_norm_pre_clip_avg': 0.14908862262964248, + 'learning_rate': 9.882103279855095e-06, + 'epoch': 7.71} +04/19 [22:27:34] INFO | >> train_qwenlatent.py:487 + Step 30550 | grad_norm_pre_clip=0.1795 | + grad_norm_pre_clip_avg=0.1897 | Metrics: + {'align_loss': 0.024532010778784752, + 'recon_loss': 0.11681097000837326, + 'predict_loss': 0.008375071920454502, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1795443445444107, + 'mae_score': 0.008333440729089685, 'data_time': + 0.0009094889974221587, 'model_time': + 1.2045789709954988, 'grad_norm_pre_clip_avg': + 0.1897232785820961, 'learning_rate': + 9.873581604263334e-06, 'epoch': 7.71} +04/19 [22:27:47] INFO | >> train_qwenlatent.py:487 + Step 30560 | grad_norm_pre_clip=0.1400 | + grad_norm_pre_clip_avg=0.1834 | Metrics: + {'align_loss': 0.025401415303349495, + 'recon_loss': 0.09328354895114899, + 'predict_loss': 0.006087552756071091, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14000540971755981, + 'data_time': 0.0006667450070381165, + 'model_time': 1.5554104679904412, + 'grad_norm_pre_clip_avg': 0.18339264392852783, + 'learning_rate': 9.865061214848347e-06, + 'epoch': 7.71} +04/19 [22:28:00] INFO | >> train_qwenlatent.py:487 + Step 30570 | grad_norm_pre_clip=0.1410 | + grad_norm_pre_clip_avg=0.1784 | Metrics: + {'align_loss': 0.025113649666309357, + 'recon_loss': 0.10007751733064651, + 'predict_loss': 0.006870732177048922, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1409679651260376, + 'data_time': 0.0007361999887507409, + 'model_time': 1.5964239310123958, + 'grad_norm_pre_clip_avg': 0.17840657532215118, + 'learning_rate': 9.856542115762857e-06, + 'epoch': 7.71} +04/19 [22:28:12] INFO | >> train_qwenlatent.py:487 + Step 30580 | grad_norm_pre_clip=0.2449 | + grad_norm_pre_clip_avg=0.1979 | Metrics: + {'align_loss': 0.02607930824160576, + 'recon_loss': 0.09477147459983826, + 'predict_loss': 0.008527550846338272, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24487438797950745, + 'data_time': 0.0006533570121973753, + 'model_time': 1.1798722870007623, + 'grad_norm_pre_clip_avg': 0.1978779688477516, + 'learning_rate': 9.848024311158973e-06, + 'epoch': 7.72} +04/19 [22:28:25] INFO | >> train_qwenlatent.py:487 + Step 30590 | grad_norm_pre_clip=0.1749 | + grad_norm_pre_clip_avg=0.1992 | Metrics: + {'align_loss': 0.02661997266113758, + 'recon_loss': 0.08787649124860764, + 'predict_loss': 0.006480480078607798, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1749127358198166, + 'data_time': 0.0009802260028664023, + 'model_time': 1.2351776930154301, + 'grad_norm_pre_clip_avg': 0.19924801737070083, + 'learning_rate': 9.839507805188175e-06, + 'epoch': 7.72} +04/19 [22:28:38] INFO | >> train_qwenlatent.py:487 + Step 30600 | grad_norm_pre_clip=0.1521 | + grad_norm_pre_clip_avg=0.1585 | Metrics: + {'align_loss': 0.025307118892669678, + 'recon_loss': 0.10163529962301254, + 'predict_loss': 0.01198162417858839, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1521000862121582, + 'mae_score': 0.008368413942354219, 'data_time': + 0.0009308909939136356, 'model_time': + 1.2368438790144864, 'grad_norm_pre_clip_avg': + 0.15846109092235566, 'learning_rate': + 9.830992602001299e-06, 'epoch': 7.72} +04/19 [22:28:50] INFO | >> train_qwenlatent.py:487 + Step 30610 | grad_norm_pre_clip=0.1378 | + grad_norm_pre_clip_avg=0.1745 | Metrics: + {'align_loss': 0.024979116395115852, + 'recon_loss': 0.07242196798324585, + 'predict_loss': 0.004627586342394352, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1377904862165451, + 'data_time': 0.0011752609862014651, + 'model_time': 1.25428208798985, + 'grad_norm_pre_clip_avg': 0.17451903373003005, + 'learning_rate': 9.822478705748554e-06, + 'epoch': 7.72} +04/19 [22:29:03] INFO | >> train_qwenlatent.py:487 + Step 30620 | grad_norm_pre_clip=0.1643 | + grad_norm_pre_clip_avg=0.1809 | Metrics: + {'align_loss': 0.026159266009926796, + 'recon_loss': 0.1305723339319229, + 'predict_loss': 0.013922585174441338, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16429728269577026, + 'data_time': 0.0008408139983657748, + 'model_time': 1.2129810480109882, + 'grad_norm_pre_clip_avg': 0.1809047743678093, + 'learning_rate': 9.81396612057951e-06, 'epoch': + 7.73} +04/19 [22:29:15] INFO | >> train_qwenlatent.py:487 + Step 30630 | grad_norm_pre_clip=0.1844 | + grad_norm_pre_clip_avg=0.1718 | Metrics: + {'align_loss': 0.024278275668621063, + 'recon_loss': 0.11526083201169968, + 'predict_loss': 0.013079855591058731, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18438738584518433, + 'data_time': 0.0007855519943404943, + 'model_time': 1.232537775998935, + 'grad_norm_pre_clip_avg': 0.17176138758659362, + 'learning_rate': 9.805454850643095e-06, + 'epoch': 7.73} +04/19 [22:29:28] INFO | >> train_qwenlatent.py:487 + Step 30640 | grad_norm_pre_clip=0.1372 | + grad_norm_pre_clip_avg=0.1560 | Metrics: + {'align_loss': 0.025532593950629234, + 'recon_loss': 0.07531169801950455, + 'predict_loss': 0.007095972541719675, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13716866075992584, + 'data_time': 0.000837992993183434, + 'model_time': 1.2162922300049104, + 'grad_norm_pre_clip_avg': 0.1559873566031456, + 'learning_rate': 9.796944900087601e-06, + 'epoch': 7.73} +04/19 [22:29:41] INFO | >> train_qwenlatent.py:487 + Step 30650 | grad_norm_pre_clip=0.2345 | + grad_norm_pre_clip_avg=0.1694 | Metrics: + {'align_loss': 0.025645703077316284, + 'recon_loss': 0.08479996770620346, + 'predict_loss': 0.007292941678315401, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2344885915517807, + 'mae_score': 0.0073055507900478605, + 'data_time': 0.0010347680072300136, + 'model_time': 1.2571425219939556, + 'grad_norm_pre_clip_avg': 0.16941055580973624, + 'learning_rate': 9.788436273060679e-06, + 'epoch': 7.73} +04/19 [22:29:53] INFO | >> train_qwenlatent.py:487 + Step 30660 | grad_norm_pre_clip=0.1625 | + grad_norm_pre_clip_avg=0.1908 | Metrics: + {'align_loss': 0.02483469992876053, + 'recon_loss': 0.10304420441389084, + 'predict_loss': 0.01009602751582861, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16249054670333862, + 'data_time': 0.0009094629785977304, + 'model_time': 1.2765223759924993, + 'grad_norm_pre_clip_avg': 0.19083085060119628, + 'learning_rate': 9.779928973709319e-06, + 'epoch': 7.74} +04/19 [22:30:05] INFO | >> train_qwenlatent.py:487 + Step 30670 | grad_norm_pre_clip=0.1756 | + grad_norm_pre_clip_avg=0.1758 | Metrics: + {'align_loss': 0.025433603674173355, + 'recon_loss': 0.10767129808664322, + 'predict_loss': 0.012700517661869526, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17563167214393616, + 'data_time': 0.0009406160097569227, + 'model_time': 1.2156006539880764, + 'grad_norm_pre_clip_avg': 0.17583237886428832, + 'learning_rate': 9.771423006179882e-06, + 'epoch': 7.74} +04/19 [22:30:19] INFO | >> train_qwenlatent.py:487 + Step 30680 | grad_norm_pre_clip=0.1845 | + grad_norm_pre_clip_avg=0.1720 | Metrics: + {'align_loss': 0.02474738098680973, + 'recon_loss': 0.11308959871530533, + 'predict_loss': 0.011291343718767166, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1844622641801834, + 'data_time': 0.0006986679800320417, + 'model_time': 1.2801722029980738, + 'grad_norm_pre_clip_avg': 0.1719665050506592, + 'learning_rate': 9.762918374618073e-06, + 'epoch': 7.74} +04/19 [22:30:31] INFO | >> train_qwenlatent.py:487 + Step 30690 | grad_norm_pre_clip=0.1486 | + grad_norm_pre_clip_avg=0.1660 | Metrics: + {'align_loss': 0.025192230939865112, + 'recon_loss': 0.12225732207298279, + 'predict_loss': 0.008630172349512577, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1485883593559265, + 'data_time': 0.0008642509928904474, + 'model_time': 1.2045569299953058, + 'grad_norm_pre_clip_avg': 0.16602618247270584, + 'learning_rate': 9.754415083168947e-06, + 'epoch': 7.74} +04/19 [22:30:45] INFO | >> train_qwenlatent.py:487 + Step 30700 | grad_norm_pre_clip=0.1224 | + grad_norm_pre_clip_avg=0.1616 | Metrics: + {'align_loss': 0.024292515590786934, + 'recon_loss': 0.06983035057783127, + 'predict_loss': 0.006904137786477804, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12241078168153763, + 'mae_score': 0.008027835364814277, 'data_time': + 0.001012006017845124, 'model_time': + 1.3281051850062795, 'grad_norm_pre_clip_avg': + 0.16158250644803046, 'learning_rate': + 9.745913135976903e-06, 'epoch': 7.75} +04/19 [22:30:58] INFO | >> train_qwenlatent.py:487 + Step 30710 | grad_norm_pre_clip=0.1939 | + grad_norm_pre_clip_avg=0.1839 | Metrics: + {'align_loss': 0.024445414543151855, + 'recon_loss': 0.06484448164701462, + 'predict_loss': 0.0065414211712777615, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19386492669582367, + 'data_time': 0.0007103380048647523, + 'model_time': 1.2058144609909505, + 'grad_norm_pre_clip_avg': 0.1838975578546524, + 'learning_rate': 9.737412537185687e-06, + 'epoch': 7.75} +04/19 [22:31:10] INFO | >> train_qwenlatent.py:487 + Step 30720 | grad_norm_pre_clip=0.1899 | + grad_norm_pre_clip_avg=0.1898 | Metrics: + {'align_loss': 0.025543328374624252, + 'recon_loss': 0.08988320827484131, + 'predict_loss': 0.006242516916245222, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.189928337931633, + 'data_time': 0.0011772739817388356, + 'model_time': 1.2217927949968725, + 'grad_norm_pre_clip_avg': 0.18976497501134873, + 'learning_rate': 9.728913290938389e-06, + 'epoch': 7.75} +04/19 [22:31:23] INFO | >> train_qwenlatent.py:487 + Step 30730 | grad_norm_pre_clip=0.1657 | + grad_norm_pre_clip_avg=0.1848 | Metrics: + {'align_loss': 0.02608434669673443, + 'recon_loss': 0.13291175663471222, + 'predict_loss': 0.01151218730956316, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1657116711139679, + 'data_time': 0.0006502829783130437, + 'model_time': 1.2578121179831214, + 'grad_norm_pre_clip_avg': 0.1848161444067955, + 'learning_rate': 9.720415401377438e-06, + 'epoch': 7.75} +04/19 [22:31:35] INFO | >> train_qwenlatent.py:487 + Step 30740 | grad_norm_pre_clip=0.1432 | + grad_norm_pre_clip_avg=0.1675 | Metrics: + {'align_loss': 0.02688540145754814, + 'recon_loss': 0.10928866267204285, + 'predict_loss': 0.00669246818870306, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1432134509086609, + 'data_time': 0.0006590989942196757, + 'model_time': 1.237294694990851, + 'grad_norm_pre_clip_avg': 0.16747804135084152, + 'learning_rate': 9.711918872644605e-06, + 'epoch': 7.76} +04/19 [22:31:48] INFO | >> train_qwenlatent.py:487 + Step 30750 | grad_norm_pre_clip=0.1284 | + grad_norm_pre_clip_avg=0.1676 | Metrics: + {'align_loss': 0.026063844561576843, + 'recon_loss': 0.10975044965744019, + 'predict_loss': 0.011226755566895008, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1284051388502121, + 'mae_score': 0.008506254247716956, 'data_time': + 0.0007568529981654137, 'model_time': + 1.214063971012365, 'grad_norm_pre_clip_avg': + 0.16759225726127625, 'learning_rate': + 9.703423708880988e-06, 'epoch': 7.76} +04/19 [22:32:01] INFO | >> train_qwenlatent.py:487 + Step 30760 | grad_norm_pre_clip=0.1741 | + grad_norm_pre_clip_avg=0.1631 | Metrics: + {'align_loss': 0.025721050798892975, + 'recon_loss': 0.09498461335897446, + 'predict_loss': 0.00794301088899374, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17411768436431885, + 'data_time': 0.0006999190081842244, + 'model_time': 1.21101339199231, + 'grad_norm_pre_clip_avg': 0.16307500898838043, + 'learning_rate': 9.694929914227031e-06, + 'epoch': 7.76} +04/19 [22:32:14] INFO | >> train_qwenlatent.py:487 + Step 30770 | grad_norm_pre_clip=0.1995 | + grad_norm_pre_clip_avg=0.1748 | Metrics: + {'align_loss': 0.026000842452049255, + 'recon_loss': 0.09283401072025299, + 'predict_loss': 0.005293866153806448, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19947326183319092, + 'data_time': 0.0008767910185270011, + 'model_time': 1.1987252069811802, + 'grad_norm_pre_clip_avg': 0.17477591782808305, + 'learning_rate': 9.686437492822504e-06, + 'epoch': 7.76} +04/19 [22:32:25] INFO | >> train_qwenlatent.py:487 + Step 30780 | grad_norm_pre_clip=0.1681 | + grad_norm_pre_clip_avg=0.1802 | Metrics: + {'align_loss': 0.02393473871052265, + 'recon_loss': 0.06452827900648117, + 'predict_loss': 0.0037475358694791794, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16805727779865265, + 'data_time': 0.0006122570193838328, + 'model_time': 1.1563192380126566, + 'grad_norm_pre_clip_avg': 0.18016485571861268, + 'learning_rate': 9.677946448806513e-06, + 'epoch': 7.77} +04/19 [22:32:37] INFO | >> train_qwenlatent.py:487 + Step 30790 | grad_norm_pre_clip=0.1617 | + grad_norm_pre_clip_avg=0.1676 | Metrics: + {'align_loss': 0.025466060265898705, + 'recon_loss': 0.1073581725358963, + 'predict_loss': 0.008476898074150085, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16171815991401672, + 'data_time': 0.0005903359851799905, + 'model_time': 1.177026213001227, + 'grad_norm_pre_clip_avg': 0.16762354075908661, + 'learning_rate': 9.669456786317486e-06, + 'epoch': 7.77} +04/19 [22:32:49] INFO | >> train_qwenlatent.py:487 + Step 30800 | grad_norm_pre_clip=0.1509 | + grad_norm_pre_clip_avg=0.1666 | Metrics: + {'align_loss': 0.02690911293029785, + 'recon_loss': 0.13731643557548523, + 'predict_loss': 0.008607827126979828, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15087276697158813, + 'mae_score': 0.008282327222394513, 'data_time': + 0.0006263169925659895, 'model_time': + 1.1632960079878103, 'grad_norm_pre_clip_avg': + 0.16661273390054704, 'learning_rate': + 9.660968509493186e-06, 'epoch': 7.77} +04/19 [22:33:01] INFO | >> train_qwenlatent.py:487 + Step 30810 | grad_norm_pre_clip=0.1721 | + grad_norm_pre_clip_avg=0.1827 | Metrics: + {'align_loss': 0.02512207254767418, + 'recon_loss': 0.08110558986663818, + 'predict_loss': 0.006778649985790253, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17211250960826874, + 'data_time': 0.000627772998996079, + 'model_time': 1.1463416790065821, + 'grad_norm_pre_clip_avg': 0.18265629857778548, + 'learning_rate': 9.652481622470694e-06, + 'epoch': 7.77} +04/19 [22:33:13] INFO | >> train_qwenlatent.py:487 + Step 30820 | grad_norm_pre_clip=0.1777 | + grad_norm_pre_clip_avg=0.1773 | Metrics: + {'align_loss': 0.025275981053709984, + 'recon_loss': 0.08461935073137283, + 'predict_loss': 0.00726134330034256, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17772458493709564, + 'data_time': 0.0006599049957003444, + 'model_time': 1.1689997400098946, + 'grad_norm_pre_clip_avg': 0.177325277030468, + 'learning_rate': 9.643996129386414e-06, + 'epoch': 7.78} +04/19 [22:33:25] INFO | >> train_qwenlatent.py:487 + Step 30830 | grad_norm_pre_clip=0.1661 | + grad_norm_pre_clip_avg=0.1653 | Metrics: + {'align_loss': 0.02543683536350727, + 'recon_loss': 0.090658999979496, + 'predict_loss': 0.010015108622610569, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16608549654483795, + 'data_time': 0.0005588399944826961, + 'model_time': 1.1432582960114814, + 'grad_norm_pre_clip_avg': 0.16525908708572387, + 'learning_rate': 9.63551203437608e-06, 'epoch': + 7.78} +04/19 [22:33:37] INFO | >> train_qwenlatent.py:487 + Step 30840 | grad_norm_pre_clip=0.2088 | + grad_norm_pre_clip_avg=0.1721 | Metrics: + {'align_loss': 0.024068143218755722, + 'recon_loss': 0.08536657691001892, + 'predict_loss': 0.006014488637447357, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20878644287586212, + 'data_time': 0.0006942969921510667, + 'model_time': 1.176470734004397, + 'grad_norm_pre_clip_avg': 0.17205565869808198, + 'learning_rate': 9.627029341574722e-06, + 'epoch': 7.78} +04/19 [22:33:49] INFO | >> train_qwenlatent.py:487 + Step 30850 | grad_norm_pre_clip=0.1571 | + grad_norm_pre_clip_avg=0.1594 | Metrics: + {'align_loss': 0.024320052936673164, + 'recon_loss': 0.07289618998765945, + 'predict_loss': 0.009100453928112984, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1570967137813568, + 'mae_score': 0.008221688141693941, 'data_time': + 0.0006022349989507347, 'model_time': + 1.156584396987455, 'grad_norm_pre_clip_avg': + 0.15940697491168976, 'learning_rate': + 9.618548055116713e-06, 'epoch': 7.78} +04/19 [22:34:00] INFO | >> train_qwenlatent.py:487 + Step 30860 | grad_norm_pre_clip=0.1949 | + grad_norm_pre_clip_avg=0.1769 | Metrics: + {'align_loss': 0.025499636307358742, + 'recon_loss': 0.08873515576124191, + 'predict_loss': 0.006991230417042971, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19494378566741943, + 'data_time': 0.0005713819991797209, + 'model_time': 1.154742674989393, + 'grad_norm_pre_clip_avg': 0.17690802961587906, + 'learning_rate': 9.610068179135722e-06, + 'epoch': 7.79} +04/19 [22:34:12] INFO | >> train_qwenlatent.py:487 + Step 30870 | grad_norm_pre_clip=0.1644 | + grad_norm_pre_clip_avg=0.1684 | Metrics: + {'align_loss': 0.0249069482088089, + 'recon_loss': 0.10043283551931381, + 'predict_loss': 0.012518788687884808, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1644473671913147, + 'data_time': 0.0009579260076861829, + 'model_time': 1.1649945779936388, + 'grad_norm_pre_clip_avg': 0.1683865576982498, + 'learning_rate': 9.601589717764747e-06, + 'epoch': 7.79} +04/19 [22:34:24] INFO | >> train_qwenlatent.py:487 + Step 30880 | grad_norm_pre_clip=0.1657 | + grad_norm_pre_clip_avg=0.1678 | Metrics: + {'align_loss': 0.025749199092388153, + 'recon_loss': 0.09379515796899796, + 'predict_loss': 0.004255090374499559, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16568931937217712, + 'data_time': 0.0005866269930265844, + 'model_time': 1.140460719994735, + 'grad_norm_pre_clip_avg': 0.16783688217401505, + 'learning_rate': 9.593112675136078e-06, + 'epoch': 7.79} +04/19 [22:34:35] INFO | >> train_qwenlatent.py:487 + Step 30890 | grad_norm_pre_clip=0.1846 | + grad_norm_pre_clip_avg=0.1724 | Metrics: + {'align_loss': 0.026203861460089684, + 'recon_loss': 0.12568582594394684, + 'predict_loss': 0.013726122677326202, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1846150904893875, + 'data_time': 0.0005448719894047827, + 'model_time': 1.1628314199915621, + 'grad_norm_pre_clip_avg': 0.17235014885663985, + 'learning_rate': 9.584637055381326e-06, + 'epoch': 7.79} +04/19 [22:34:47] INFO | >> train_qwenlatent.py:487 + Step 30900 | grad_norm_pre_clip=0.2248 | + grad_norm_pre_clip_avg=0.1608 | Metrics: + {'align_loss': 0.025629421696066856, + 'recon_loss': 0.07638221979141235, + 'predict_loss': 0.006328473333269358, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2248292863368988, + 'mae_score': 0.007695026225871868, 'data_time': + 0.0005605079932138324, 'model_time': + 1.1575121330097318, 'grad_norm_pre_clip_avg': + 0.16084185093641282, 'learning_rate': + 9.576162862631407e-06, 'epoch': 7.8} +04/19 [22:34:59] INFO | >> train_qwenlatent.py:487 + Step 30910 | grad_norm_pre_clip=0.1704 | + grad_norm_pre_clip_avg=0.1790 | Metrics: + {'align_loss': 0.027189839631319046, + 'recon_loss': 0.09295002371072769, + 'predict_loss': 0.004049196373671293, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17037685215473175, + 'data_time': 0.000650047993985936, + 'model_time': 1.150218221009709, + 'grad_norm_pre_clip_avg': 0.1789780154824257, + 'learning_rate': 9.567690101016539e-06, + 'epoch': 7.8} +04/19 [22:35:11] INFO | >> train_qwenlatent.py:487 + Step 30920 | grad_norm_pre_clip=0.2876 | + grad_norm_pre_clip_avg=0.2071 | Metrics: + {'align_loss': 0.02523120492696762, + 'recon_loss': 0.12268032133579254, + 'predict_loss': 0.014127831906080246, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28763914108276367, + 'data_time': 0.0005839699879288673, + 'model_time': 1.135750788002042, + 'grad_norm_pre_clip_avg': 0.20709750950336456, + 'learning_rate': 9.559218774666241e-06, + 'epoch': 7.8} +04/19 [22:35:22] INFO | >> train_qwenlatent.py:487 + Step 30930 | grad_norm_pre_clip=0.1648 | + grad_norm_pre_clip_avg=0.2003 | Metrics: + {'align_loss': 0.025451812893152237, + 'recon_loss': 0.09886448830366135, + 'predict_loss': 0.007344844285398722, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16476339101791382, + 'data_time': 0.0005474099889397621, + 'model_time': 1.1558887099963613, + 'grad_norm_pre_clip_avg': 0.20027784556150435, + 'learning_rate': 9.550748887709334e-06, + 'epoch': 7.8} +04/19 [22:35:34] INFO | >> train_qwenlatent.py:487 + Step 30940 | grad_norm_pre_clip=0.1815 | + grad_norm_pre_clip_avg=0.2034 | Metrics: + {'align_loss': 0.026500094681978226, + 'recon_loss': 0.1718340665102005, + 'predict_loss': 0.015710055828094482, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1815485805273056, + 'data_time': 0.0006584840011782944, + 'model_time': 1.1578478169976734, + 'grad_norm_pre_clip_avg': 0.20336999744176865, + 'learning_rate': 9.54228044427394e-06, 'epoch': + 7.81} +04/19 [22:35:46] INFO | >> train_qwenlatent.py:487 + Step 30950 | grad_norm_pre_clip=0.1693 | + grad_norm_pre_clip_avg=0.1830 | Metrics: + {'align_loss': 0.026209674775600433, + 'recon_loss': 0.11954021453857422, + 'predict_loss': 0.008896860294044018, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16928797960281372, + 'mae_score': 0.008002854252720739, 'data_time': + 0.0007381740142591298, 'model_time': + 1.1589453950000461, 'grad_norm_pre_clip_avg': + 0.18297559171915054, 'learning_rate': + 9.533813448487475e-06, 'epoch': 7.81} +04/19 [22:35:57] INFO | >> train_qwenlatent.py:487 + Step 30960 | grad_norm_pre_clip=0.1743 | + grad_norm_pre_clip_avg=0.1647 | Metrics: + {'align_loss': 0.02557608112692833, + 'recon_loss': 0.05988228693604469, + 'predict_loss': 0.004870651755481958, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17430362105369568, + 'data_time': 0.0006756270013283938, + 'model_time': 1.153453344013542, + 'grad_norm_pre_clip_avg': 0.1647488221526146, + 'learning_rate': 9.525347904476647e-06, + 'epoch': 7.81} +04/19 [22:36:09] INFO | >> train_qwenlatent.py:487 + Step 30970 | grad_norm_pre_clip=0.1814 | + grad_norm_pre_clip_avg=0.1715 | Metrics: + {'align_loss': 0.025297142565250397, + 'recon_loss': 0.1640477478504181, + 'predict_loss': 0.012469587847590446, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1814427375793457, + 'data_time': 0.0006151290144771338, + 'model_time': 1.1463004469987936, + 'grad_norm_pre_clip_avg': 0.17154512852430343, + 'learning_rate': 9.516883816367463e-06, + 'epoch': 7.81} +04/19 [22:36:21] INFO | >> train_qwenlatent.py:487 + Step 30980 | grad_norm_pre_clip=0.1435 | + grad_norm_pre_clip_avg=0.1676 | Metrics: + {'align_loss': 0.024404585361480713, + 'recon_loss': 0.09070836007595062, + 'predict_loss': 0.007062545046210289, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1435251384973526, + 'data_time': 0.0005837369826622307, + 'model_time': 1.168176309991395, + 'grad_norm_pre_clip_avg': 0.16760425120592118, + 'learning_rate': 9.508421188285215e-06, + 'epoch': 7.82} +04/19 [22:36:33] INFO | >> train_qwenlatent.py:487 + Step 30990 | grad_norm_pre_clip=0.1582 | + grad_norm_pre_clip_avg=0.1614 | Metrics: + {'align_loss': 0.025176435708999634, + 'recon_loss': 0.1151621863245964, + 'predict_loss': 0.00914530549198389, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1581527143716812, + 'data_time': 0.000648022978566587, + 'model_time': 1.1480587929836474, + 'grad_norm_pre_clip_avg': 0.161411289870739, + 'learning_rate': 9.499960024354489e-06, + 'epoch': 7.82} +04/19 [22:36:45] INFO | >> train_qwenlatent.py:487 + Step 31000 | grad_norm_pre_clip=0.1923 | + grad_norm_pre_clip_avg=0.1747 | Metrics: + {'align_loss': 0.025076471269130707, + 'recon_loss': 0.10848575830459595, + 'predict_loss': 0.006650240626186132, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19232508540153503, + 'mae_score': 0.007894849347638655, 'data_time': + 0.000653516995953396, 'model_time': + 1.178921693994198, 'grad_norm_pre_clip_avg': + 0.17469747215509415, 'learning_rate': + 9.491500328699152e-06, 'epoch': 7.82} +04/19 [22:36:57] INFO | >> train_qwenlatent.py:487 + Step 31010 | grad_norm_pre_clip=0.1434 | + grad_norm_pre_clip_avg=0.2119 | Metrics: + {'align_loss': 0.024339962750673294, + 'recon_loss': 0.09746574610471725, + 'predict_loss': 0.007054104004055262, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14343808591365814, + 'data_time': 0.0006067510112188756, + 'model_time': 1.1590277159994002, + 'grad_norm_pre_clip_avg': 0.2119280368089676, + 'learning_rate': 9.483042105442353e-06, + 'epoch': 7.82} +04/19 [22:37:08] INFO | >> train_qwenlatent.py:487 + Step 31020 | grad_norm_pre_clip=0.2212 | + grad_norm_pre_clip_avg=0.2008 | Metrics: + {'align_loss': 0.02513609081506729, + 'recon_loss': 0.09106262773275375, + 'predict_loss': 0.007323466241359711, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22122830152511597, + 'data_time': 0.0005984180024825037, + 'model_time': 1.1406189179979265, + 'grad_norm_pre_clip_avg': 0.20076379925012589, + 'learning_rate': 9.47458535870653e-06, 'epoch': + 7.83} +04/19 [22:37:20] INFO | >> train_qwenlatent.py:487 + Step 31030 | grad_norm_pre_clip=0.1764 | + grad_norm_pre_clip_avg=0.1687 | Metrics: + {'align_loss': 0.024480357766151428, + 'recon_loss': 0.07066231220960617, + 'predict_loss': 0.007001952733844519, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17639371752738953, + 'data_time': 0.0005874779890291393, + 'model_time': 1.1443666920240503, + 'grad_norm_pre_clip_avg': 0.16868919432163237, + 'learning_rate': 9.466130092613394e-06, + 'epoch': 7.83} +04/19 [22:37:31] INFO | >> train_qwenlatent.py:487 + Step 31040 | grad_norm_pre_clip=0.1457 | + grad_norm_pre_clip_avg=0.1584 | Metrics: + {'align_loss': 0.024852316826581955, + 'recon_loss': 0.09770612418651581, + 'predict_loss': 0.01036425307393074, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14570336043834686, + 'data_time': 0.0006140870100352913, + 'model_time': 1.1365126729942858, + 'grad_norm_pre_clip_avg': 0.1583856910467148, + 'learning_rate': 9.457676311283948e-06, + 'epoch': 7.83} +04/19 [22:37:43] INFO | >> train_qwenlatent.py:487 + Step 31050 | grad_norm_pre_clip=0.1506 | + grad_norm_pre_clip_avg=0.1539 | Metrics: + {'align_loss': 0.024959318339824677, + 'recon_loss': 0.10996250063180923, + 'predict_loss': 0.008422316052019596, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15061761438846588, + 'mae_score': 0.009493181726954006, 'data_time': + 0.000679827993735671, 'model_time': + 1.1371137440146413, 'grad_norm_pre_clip_avg': + 0.15391922146081924, 'learning_rate': + 9.449224018838453e-06, 'epoch': 7.83} +04/19 [22:37:55] INFO | >> train_qwenlatent.py:487 + Step 31060 | grad_norm_pre_clip=0.1459 | + grad_norm_pre_clip_avg=0.1731 | Metrics: + {'align_loss': 0.026239436119794846, + 'recon_loss': 0.1113317683339119, + 'predict_loss': 0.008514955639839172, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14591601490974426, + 'data_time': 0.0005807800043839961, + 'model_time': 1.1381260909838602, + 'grad_norm_pre_clip_avg': 0.17308469861745834, + 'learning_rate': 9.44077321939646e-06, 'epoch': + 7.84} +04/19 [22:38:06] INFO | >> train_qwenlatent.py:487 + Step 31070 | grad_norm_pre_clip=0.1329 | + grad_norm_pre_clip_avg=0.1770 | Metrics: + {'align_loss': 0.024700064212083817, + 'recon_loss': 0.10587280988693237, + 'predict_loss': 0.009112827479839325, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13294270634651184, + 'data_time': 0.0005825569969601929, + 'model_time': 1.1385106689995155, + 'grad_norm_pre_clip_avg': 0.1770162582397461, + 'learning_rate': 9.432323917076786e-06, + 'epoch': 7.84} +04/19 [22:38:18] INFO | >> train_qwenlatent.py:487 + Step 31080 | grad_norm_pre_clip=0.1593 | + grad_norm_pre_clip_avg=0.1655 | Metrics: + {'align_loss': 0.026644142344594002, + 'recon_loss': 0.11366084218025208, + 'predict_loss': 0.007884783670306206, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15926289558410645, + 'data_time': 0.000532086007297039, + 'model_time': 1.141839688993059, + 'grad_norm_pre_clip_avg': 0.16547198370099067, + 'learning_rate': 9.423876115997514e-06, + 'epoch': 7.84} +04/19 [22:39:03] INFO | >> train_qwenlatent.py:487 + Step 31090 | grad_norm_pre_clip=0.2116 | + grad_norm_pre_clip_avg=0.1632 | Metrics: + {'align_loss': 0.025726845487952232, + 'recon_loss': 0.1335316300392151, + 'predict_loss': 0.008897296153008938, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21155762672424316, + 'data_time': 0.0008529460174031556, + 'model_time': 3.1160095510131214, + 'grad_norm_pre_clip_avg': 0.16318842843174935, + 'learning_rate': 9.415429820276009e-06, + 'epoch': 7.85} +04/19 [22:39:40] INFO | >> train_qwenlatent.py:487 + Step 31100 | grad_norm_pre_clip=0.1628 | + grad_norm_pre_clip_avg=0.1771 | Metrics: + {'align_loss': 0.025622479617595673, + 'recon_loss': 0.08256998658180237, + 'predict_loss': 0.005888782907277346, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1627764254808426, + 'mae_score': 0.007529501872019725, 'data_time': + 0.0010145110136363655, 'model_time': + 3.0795587729953695, 'grad_norm_pre_clip_avg': + 0.17714976370334626, 'learning_rate': + 9.406985034028878e-06, 'epoch': 7.85} +04/19 [22:40:16] INFO | >> train_qwenlatent.py:487 + Step 31110 | grad_norm_pre_clip=0.1333 | + grad_norm_pre_clip_avg=0.1764 | Metrics: + {'align_loss': 0.025420786812901497, + 'recon_loss': 0.11973371356725693, + 'predict_loss': 0.009478882886469364, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1333116590976715, + 'data_time': 0.001579550007591024, + 'model_time': 3.969791677984176, + 'grad_norm_pre_clip_avg': 0.17638955563306807, + 'learning_rate': 9.39854176137202e-06, 'epoch': + 7.85} +04/19 [22:40:53] INFO | >> train_qwenlatent.py:487 + Step 31120 | grad_norm_pre_clip=0.1558 | + grad_norm_pre_clip_avg=0.1722 | Metrics: + {'align_loss': 0.026570089161396027, + 'recon_loss': 0.10175561159849167, + 'predict_loss': 0.008098921738564968, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15577031672000885, + 'data_time': 0.001163962995633483, + 'model_time': 2.6588034930173308, + 'grad_norm_pre_clip_avg': 0.1721733495593071, + 'learning_rate': 9.390100006420573e-06, + 'epoch': 7.85} +04/19 [22:41:27] INFO | >> train_qwenlatent.py:487 + Step 31130 | grad_norm_pre_clip=0.1852 | + grad_norm_pre_clip_avg=0.1550 | Metrics: + {'align_loss': 0.024390963837504387, + 'recon_loss': 0.09408222138881683, + 'predict_loss': 0.005625375546514988, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18524804711341858, + 'data_time': 0.0012809869949705899, + 'model_time': 3.4804617800109554, + 'grad_norm_pre_clip_avg': 0.15502683371305465, + 'learning_rate': 9.381659773288956e-06, + 'epoch': 7.86} +04/19 [22:42:03] INFO | >> train_qwenlatent.py:487 + Step 31140 | grad_norm_pre_clip=0.2584 | + grad_norm_pre_clip_avg=0.1883 | Metrics: + {'align_loss': 0.024567095562815666, + 'recon_loss': 0.09801089018583298, + 'predict_loss': 0.00906745158135891, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25843220949172974, + 'data_time': 0.0030156349821481854, + 'model_time': 3.2620596929918975, + 'grad_norm_pre_clip_avg': 0.18826799988746643, + 'learning_rate': 9.37322106609083e-06, 'epoch': + 7.86} +04/19 [22:42:37] INFO | >> train_qwenlatent.py:487 + Step 31150 | grad_norm_pre_clip=0.1514 | + grad_norm_pre_clip_avg=0.1810 | Metrics: + {'align_loss': 0.02518971636891365, + 'recon_loss': 0.10010729730129242, + 'predict_loss': 0.00758208055049181, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15136687457561493, + 'mae_score': 0.008756885013064823, 'data_time': + 0.003305605991045013, 'model_time': + 3.0109287279774435, 'grad_norm_pre_clip_avg': + 0.18098898231983185, 'learning_rate': + 9.364783888939117e-06, 'epoch': 7.86} +04/19 [22:43:00] INFO | >> train_qwenlatent.py:487 + Step 31160 | grad_norm_pre_clip=0.1748 | + grad_norm_pre_clip_avg=0.2116 | Metrics: + {'align_loss': 0.026725608855485916, + 'recon_loss': 0.16019783914089203, + 'predict_loss': 0.012928133830428123, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1748121678829193, + 'data_time': 0.0010367479990236461, + 'model_time': 2.302137008984573, + 'grad_norm_pre_clip_avg': 0.2116352766752243, + 'learning_rate': 9.356348245946003e-06, + 'epoch': 7.86} +04/19 [22:43:21] INFO | >> train_qwenlatent.py:487 + Step 31170 | grad_norm_pre_clip=0.1631 | + grad_norm_pre_clip_avg=0.1654 | Metrics: + {'align_loss': 0.02735506370663643, + 'recon_loss': 0.14198899269104004, + 'predict_loss': 0.013890419155359268, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16305379569530487, + 'data_time': 0.0010613519989419729, + 'model_time': 1.7199919499980751, + 'grad_norm_pre_clip_avg': 0.16544566452503204, + 'learning_rate': 9.34791414122291e-06, 'epoch': + 7.87} +04/19 [22:43:34] INFO | >> train_qwenlatent.py:487 + Step 31180 | grad_norm_pre_clip=0.1254 | + grad_norm_pre_clip_avg=0.1614 | Metrics: + {'align_loss': 0.025411173701286316, + 'recon_loss': 0.10764288902282715, + 'predict_loss': 0.005425113253295422, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1253695785999298, + 'data_time': 0.0005917100061196834, + 'model_time': 1.2089398539974354, + 'grad_norm_pre_clip_avg': 0.16140601634979249, + 'learning_rate': 9.339481578880525e-06, + 'epoch': 7.87} +04/19 [22:43:47] INFO | >> train_qwenlatent.py:487 + Step 31190 | grad_norm_pre_clip=0.1770 | + grad_norm_pre_clip_avg=0.1663 | Metrics: + {'align_loss': 0.023179879412055016, + 'recon_loss': 0.0797884538769722, + 'predict_loss': 0.005634716711938381, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.177036315202713, + 'data_time': 0.0008235780114773661, + 'model_time': 1.2961188680201303, + 'grad_norm_pre_clip_avg': 0.16634218990802765, + 'learning_rate': 9.331050563028768e-06, + 'epoch': 7.87} +04/19 [22:44:00] INFO | >> train_qwenlatent.py:487 + Step 31200 | grad_norm_pre_clip=0.2457 | + grad_norm_pre_clip_avg=0.2018 | Metrics: + {'align_loss': 0.02494770660996437, + 'recon_loss': 0.1125277429819107, + 'predict_loss': 0.01259311381727457, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24573732912540436, + 'mae_score': 0.007281372139045784, 'data_time': + 0.0006691979942843318, 'model_time': + 1.2267813260259572, 'grad_norm_pre_clip_avg': + 0.20178891271352767, 'learning_rate': + 9.322621097776818e-06, 'epoch': 7.87} +04/19 [22:44:12] INFO | >> train_qwenlatent.py:487 + Step 31210 | grad_norm_pre_clip=0.1991 | + grad_norm_pre_clip_avg=0.1811 | Metrics: + {'align_loss': 0.025923630222678185, + 'recon_loss': 0.10307078808546066, + 'predict_loss': 0.00794440321624279, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1991312950849533, + 'data_time': 0.0007019030163064599, + 'model_time': 1.2067425650020596, + 'grad_norm_pre_clip_avg': 0.18107349723577498, + 'learning_rate': 9.314193187233095e-06, + 'epoch': 7.88} +04/19 [22:44:26] INFO | >> train_qwenlatent.py:487 + Step 31220 | grad_norm_pre_clip=0.1829 | + grad_norm_pre_clip_avg=0.1619 | Metrics: + {'align_loss': 0.02441413328051567, + 'recon_loss': 0.06741417199373245, + 'predict_loss': 0.006262779701501131, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18286605179309845, + 'data_time': 0.0008022609981708229, + 'model_time': 1.5739747300103772, + 'grad_norm_pre_clip_avg': 0.1619225971400738, + 'learning_rate': 9.305766835505256e-06, + 'epoch': 7.88} +04/19 [22:44:39] INFO | >> train_qwenlatent.py:487 + Step 31230 | grad_norm_pre_clip=0.1553 | + grad_norm_pre_clip_avg=0.1747 | Metrics: + {'align_loss': 0.02595670148730278, + 'recon_loss': 0.1293889582157135, + 'predict_loss': 0.012472006492316723, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1553214192390442, + 'data_time': 0.0005990969948470592, + 'model_time': 1.234077693981817, + 'grad_norm_pre_clip_avg': 0.17467109858989716, + 'learning_rate': 9.297342046700208e-06, + 'epoch': 7.88} +04/19 [22:44:51] INFO | >> train_qwenlatent.py:487 + Step 31240 | grad_norm_pre_clip=0.1577 | + grad_norm_pre_clip_avg=0.1587 | Metrics: + {'align_loss': 0.02472074329853058, + 'recon_loss': 0.10344309359788895, + 'predict_loss': 0.006206884048879147, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15766198933124542, + 'data_time': 0.0006205120007507503, + 'model_time': 1.2314315240073483, + 'grad_norm_pre_clip_avg': 0.15874408781528473, + 'learning_rate': 9.288918824924089e-06, + 'epoch': 7.88} +04/19 [22:45:04] INFO | >> train_qwenlatent.py:487 + Step 31250 | grad_norm_pre_clip=0.1941 | + grad_norm_pre_clip_avg=0.1920 | Metrics: + {'align_loss': 0.02427092008292675, + 'recon_loss': 0.07340666651725769, + 'predict_loss': 0.0065564317628741264, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1941211223602295, + 'mae_score': 0.007397637925706468, 'data_time': + 0.0007304460159502923, 'model_time': + 1.2120832549990155, 'grad_norm_pre_clip_avg': + 0.19195917397737502, 'learning_rate': + 9.28049717428227e-06, 'epoch': 7.89} +04/19 [22:45:17] INFO | >> train_qwenlatent.py:487 + Step 31260 | grad_norm_pre_clip=0.1623 | + grad_norm_pre_clip_avg=0.1645 | Metrics: + {'align_loss': 0.024608638137578964, + 'recon_loss': 0.09560253471136093, + 'predict_loss': 0.0057325297966599464, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16230398416519165, + 'data_time': 0.0007472250144928694, + 'model_time': 1.2424353440001141, + 'grad_norm_pre_clip_avg': 0.16447938233613968, + 'learning_rate': 9.272077098879371e-06, + 'epoch': 7.89} +04/19 [22:45:29] INFO | >> train_qwenlatent.py:487 + Step 31270 | grad_norm_pre_clip=0.2136 | + grad_norm_pre_clip_avg=0.1734 | Metrics: + {'align_loss': 0.026068296283483505, + 'recon_loss': 0.12716111540794373, + 'predict_loss': 0.01243654265999794, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2136133909225464, + 'data_time': 0.0008129130001179874, + 'model_time': 1.260621307999827, + 'grad_norm_pre_clip_avg': 0.1734042376279831, + 'learning_rate': 9.26365860281923e-06, 'epoch': + 7.89} +04/19 [22:45:42] INFO | >> train_qwenlatent.py:487 + Step 31280 | grad_norm_pre_clip=0.1421 | + grad_norm_pre_clip_avg=0.1610 | Metrics: + {'align_loss': 0.0244930200278759, + 'recon_loss': 0.10232431441545486, + 'predict_loss': 0.011708462610840797, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14205841720104218, + 'data_time': 0.000878017017384991, + 'model_time': 1.2172009010100737, + 'grad_norm_pre_clip_avg': 0.16104901060461999, + 'learning_rate': 9.255241690204915e-06, + 'epoch': 7.89} +04/19 [22:45:54] INFO | >> train_qwenlatent.py:487 + Step 31290 | grad_norm_pre_clip=0.1827 | + grad_norm_pre_clip_avg=0.1571 | Metrics: + {'align_loss': 0.025645233690738678, + 'recon_loss': 0.11005578935146332, + 'predict_loss': 0.008293348364531994, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18272660672664642, + 'data_time': 0.0009112120023928583, + 'model_time': 1.2824122580059338, + 'grad_norm_pre_clip_avg': 0.15712025463581086, + 'learning_rate': 9.24682636513873e-06, 'epoch': + 7.9} +04/19 [22:46:08] INFO | >> train_qwenlatent.py:487 + Step 31300 | grad_norm_pre_clip=0.2037 | + grad_norm_pre_clip_avg=0.1815 | Metrics: + {'align_loss': 0.024947013705968857, + 'recon_loss': 0.08360923081636429, + 'predict_loss': 0.00987878069281578, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2036927044391632, + 'mae_score': 0.008818965774398667, 'data_time': + 0.000652580987662077, 'model_time': + 1.2670975890068803, 'grad_norm_pre_clip_avg': + 0.18154007643461229, 'learning_rate': + 9.238412631722204e-06, 'epoch': 7.9} +04/19 [22:46:20] INFO | >> train_qwenlatent.py:487 + Step 31310 | grad_norm_pre_clip=0.1736 | + grad_norm_pre_clip_avg=0.1835 | Metrics: + {'align_loss': 0.025684213265776634, + 'recon_loss': 0.11580626666545868, + 'predict_loss': 0.010904531925916672, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17356693744659424, + 'data_time': 0.0006235949986148626, + 'model_time': 1.2571642419788986, + 'grad_norm_pre_clip_avg': 0.18352044224739075, + 'learning_rate': 9.230000494056085e-06, + 'epoch': 7.9} +04/19 [22:46:33] INFO | >> train_qwenlatent.py:487 + Step 31320 | grad_norm_pre_clip=0.2037 | + grad_norm_pre_clip_avg=0.1714 | Metrics: + {'align_loss': 0.02622327022254467, + 'recon_loss': 0.10864780098199844, + 'predict_loss': 0.007880386896431446, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2036833018064499, + 'data_time': 0.0005830670124851167, + 'model_time': 1.1955935250152834, + 'grad_norm_pre_clip_avg': 0.17137766629457474, + 'learning_rate': 9.221589956240353e-06, + 'epoch': 7.9} +04/19 [22:46:45] INFO | >> train_qwenlatent.py:487 + Step 31330 | grad_norm_pre_clip=0.1696 | + grad_norm_pre_clip_avg=0.1730 | Metrics: + {'align_loss': 0.025107067078351974, + 'recon_loss': 0.1062193438410759, + 'predict_loss': 0.006423868704587221, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16955925524234772, + 'data_time': 0.0008673749980516732, + 'model_time': 1.2198891779989935, + 'grad_norm_pre_clip_avg': 0.1729752853512764, + 'learning_rate': 9.213181022374194e-06, + 'epoch': 7.91} +04/19 [22:46:58] INFO | >> train_qwenlatent.py:487 + Step 31340 | grad_norm_pre_clip=0.1657 | + grad_norm_pre_clip_avg=0.2140 | Metrics: + {'align_loss': 0.025807563215494156, + 'recon_loss': 0.12771429121494293, + 'predict_loss': 0.00811806134879589, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16572445631027222, + 'data_time': 0.0008375659817829728, + 'model_time': 1.264502533013001, + 'grad_norm_pre_clip_avg': 0.21397674083709717, + 'learning_rate': 9.204773696556024e-06, + 'epoch': 7.91} +04/19 [22:47:11] INFO | >> train_qwenlatent.py:487 + Step 31350 | grad_norm_pre_clip=0.1952 | + grad_norm_pre_clip_avg=0.2032 | Metrics: + {'align_loss': 0.02565336972475052, + 'recon_loss': 0.08598482608795166, + 'predict_loss': 0.005791689734905958, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19518698751926422, + 'mae_score': 0.006163230672612921, 'data_time': + 0.0006944790075067431, 'model_time': + 1.2270783210115042, 'grad_norm_pre_clip_avg': + 0.20316876322031022, 'learning_rate': + 9.196367982883475e-06, 'epoch': 7.91} +04/19 [22:47:24] INFO | >> train_qwenlatent.py:487 + Step 31360 | grad_norm_pre_clip=0.2492 | + grad_norm_pre_clip_avg=0.1952 | Metrics: + {'align_loss': 0.02451028674840927, + 'recon_loss': 0.1120246946811676, + 'predict_loss': 0.010065318085253239, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24916890263557434, + 'data_time': 0.0008889719902072102, + 'model_time': 1.5220762609969825, + 'grad_norm_pre_clip_avg': 0.19523537009954453, + 'learning_rate': 9.187963885453387e-06, + 'epoch': 7.91} +04/19 [22:47:37] INFO | >> train_qwenlatent.py:487 + Step 31370 | grad_norm_pre_clip=0.1681 | + grad_norm_pre_clip_avg=0.1799 | Metrics: + {'align_loss': 0.025881174951791763, + 'recon_loss': 0.09854498505592346, + 'predict_loss': 0.011416028253734112, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16810455918312073, + 'data_time': 0.0009180199995171279, + 'model_time': 1.2523509819875471, + 'grad_norm_pre_clip_avg': 0.17992012649774553, + 'learning_rate': 9.179561408361813e-06, + 'epoch': 7.92} +04/19 [22:47:49] INFO | >> train_qwenlatent.py:487 + Step 31380 | grad_norm_pre_clip=0.1513 | + grad_norm_pre_clip_avg=0.1877 | Metrics: + {'align_loss': 0.02508275955915451, + 'recon_loss': 0.1289997100830078, + 'predict_loss': 0.010941860266029835, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1513371467590332, + 'data_time': 0.0007915419992059469, + 'model_time': 1.2867471910140011, + 'grad_norm_pre_clip_avg': 0.18767018169164656, + 'learning_rate': 9.171160555704017e-06, + 'epoch': 7.92} +04/19 [22:48:01] INFO | >> train_qwenlatent.py:487 + Step 31390 | grad_norm_pre_clip=0.1246 | + grad_norm_pre_clip_avg=0.1840 | Metrics: + {'align_loss': 0.0260159894824028, + 'recon_loss': 0.09301116317510605, + 'predict_loss': 0.008588247001171112, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12461129575967789, + 'data_time': 0.0008003639813978225, + 'model_time': 1.2395420929824468, + 'grad_norm_pre_clip_avg': 0.18403277918696404, + 'learning_rate': 9.162761331574477e-06, + 'epoch': 7.92} +04/19 [22:48:14] INFO | >> train_qwenlatent.py:487 + Step 31400 | grad_norm_pre_clip=0.1999 | + grad_norm_pre_clip_avg=0.1600 | Metrics: + {'align_loss': 0.025234952569007874, + 'recon_loss': 0.08117004483938217, + 'predict_loss': 0.0074794478714466095, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19993406534194946, + 'mae_score': 0.008205345085075309, 'data_time': + 0.000598857004661113, 'model_time': + 1.2067320949863642, 'grad_norm_pre_clip_avg': + 0.16001916080713272, 'learning_rate': + 9.154363740066871e-06, 'epoch': 7.92} +04/19 [22:48:27] INFO | >> train_qwenlatent.py:487 + Step 31410 | grad_norm_pre_clip=0.1441 | + grad_norm_pre_clip_avg=0.1739 | Metrics: + {'align_loss': 0.024043167009949684, + 'recon_loss': 0.08314917236566544, + 'predict_loss': 0.0066443211399018764, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14409591257572174, + 'data_time': 0.0005923240096308291, + 'model_time': 1.2138851649942808, + 'grad_norm_pre_clip_avg': 0.17385861575603484, + 'learning_rate': 9.145967785274081e-06, + 'epoch': 7.93} +04/19 [22:48:40] INFO | >> train_qwenlatent.py:487 + Step 31420 | grad_norm_pre_clip=0.1539 | + grad_norm_pre_clip_avg=0.1535 | Metrics: + {'align_loss': 0.025395546108484268, + 'recon_loss': 0.10946350544691086, + 'predict_loss': 0.010656059719622135, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1538829505443573, + 'data_time': 0.0006165670056361705, + 'model_time': 1.5461755899887066, + 'grad_norm_pre_clip_avg': 0.1534627616405487, + 'learning_rate': 9.137573471288198e-06, + 'epoch': 7.93} +04/19 [22:48:52] INFO | >> train_qwenlatent.py:487 + Step 31430 | grad_norm_pre_clip=0.2037 | + grad_norm_pre_clip_avg=0.1792 | Metrics: + {'align_loss': 0.024387318640947342, + 'recon_loss': 0.07674816250801086, + 'predict_loss': 0.0038882179651409388, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20373684167861938, + 'data_time': 0.0007573720067739487, + 'model_time': 1.284838157997001, + 'grad_norm_pre_clip_avg': 0.17919371724128724, + 'learning_rate': 9.129180802200506e-06, + 'epoch': 7.93} +04/19 [22:49:05] INFO | >> train_qwenlatent.py:487 + Step 31440 | grad_norm_pre_clip=0.2689 | + grad_norm_pre_clip_avg=0.1851 | Metrics: + {'align_loss': 0.025888284668326378, + 'recon_loss': 0.09982369095087051, + 'predict_loss': 0.008814644068479538, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26890021562576294, + 'data_time': 0.0006140850018709898, + 'model_time': 1.223270339978626, + 'grad_norm_pre_clip_avg': 0.1850850760936737, + 'learning_rate': 9.120789782101486e-06, + 'epoch': 7.93} +04/19 [22:49:18] INFO | >> train_qwenlatent.py:487 + Step 31450 | grad_norm_pre_clip=0.1693 | + grad_norm_pre_clip_avg=0.1535 | Metrics: + {'align_loss': 0.02482295222580433, + 'recon_loss': 0.10225530713796616, + 'predict_loss': 0.009258268401026726, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1693214774131775, + 'mae_score': 0.008771864573160808, 'data_time': + 0.001046781981131062, 'model_time': + 1.2673577940149698, 'grad_norm_pre_clip_avg': + 0.15352915674448014, 'learning_rate': + 9.11240041508083e-06, 'epoch': 7.94} +04/19 [22:49:30] INFO | >> train_qwenlatent.py:487 + Step 31460 | grad_norm_pre_clip=0.1870 | + grad_norm_pre_clip_avg=0.1727 | Metrics: + {'align_loss': 0.024966156110167503, + 'recon_loss': 0.09376504272222519, + 'predict_loss': 0.006954932119697332, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18704673647880554, + 'data_time': 0.0009181159839499742, + 'model_time': 1.2535357250017114, + 'grad_norm_pre_clip_avg': 0.1726890280842781, + 'learning_rate': 9.104012705227403e-06, + 'epoch': 7.94} +04/19 [22:49:43] INFO | >> train_qwenlatent.py:487 + Step 31470 | grad_norm_pre_clip=0.2286 | + grad_norm_pre_clip_avg=0.1557 | Metrics: + {'align_loss': 0.025936976075172424, + 'recon_loss': 0.15006901323795319, + 'predict_loss': 0.011596817523241043, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2286210060119629, + 'data_time': 0.0006473679968621582, + 'model_time': 1.2394454249879345, + 'grad_norm_pre_clip_avg': 0.1556944489479065, + 'learning_rate': 9.09562665662928e-06, 'epoch': + 7.94} +04/19 [22:49:56] INFO | >> train_qwenlatent.py:487 + Step 31480 | grad_norm_pre_clip=0.1623 | + grad_norm_pre_clip_avg=0.2016 | Metrics: + {'align_loss': 0.026634417474269867, + 'recon_loss': 0.12255681306123734, + 'predict_loss': 0.009101266041398048, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16228140890598297, + 'data_time': 0.000883984990650788, + 'model_time': 1.2171951579803135, + 'grad_norm_pre_clip_avg': 0.20164737552404405, + 'learning_rate': 9.087242273373709e-06, + 'epoch': 7.94} +04/19 [22:50:08] INFO | >> train_qwenlatent.py:487 + Step 31490 | grad_norm_pre_clip=0.2786 | + grad_norm_pre_clip_avg=0.1941 | Metrics: + {'align_loss': 0.02549496851861477, + 'recon_loss': 0.10115859657526016, + 'predict_loss': 0.007013512775301933, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2785751521587372, + 'data_time': 0.0006182939978316426, + 'model_time': 1.2550959949730895, + 'grad_norm_pre_clip_avg': 0.19414307400584221, + 'learning_rate': 9.078859559547146e-06, + 'epoch': 7.95} +04/19 [22:50:22] INFO | >> train_qwenlatent.py:487 + Step 31500 | grad_norm_pre_clip=0.1548 | + grad_norm_pre_clip_avg=0.1866 | Metrics: + {'align_loss': 0.025436051189899445, + 'recon_loss': 0.10140218585729599, + 'predict_loss': 0.010077138431370258, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15481826663017273, + 'mae_score': 0.012105542260247308, 'data_time': + 0.0006154760194476694, 'model_time': + 1.2117444630130194, 'grad_norm_pre_clip_avg': + 0.18661600053310395, 'learning_rate': + 9.070478519235224e-06, 'epoch': 7.95} +04/19 [22:50:35] INFO | >> train_qwenlatent.py:487 + Step 31510 | grad_norm_pre_clip=0.1766 | + grad_norm_pre_clip_avg=0.1717 | Metrics: + {'align_loss': 0.026070594787597656, + 'recon_loss': 0.1265849471092224, + 'predict_loss': 0.011711404658854008, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1766258180141449, + 'data_time': 0.0008756000024732202, + 'model_time': 1.2421114899916574, + 'grad_norm_pre_clip_avg': 0.17165536135435105, + 'learning_rate': 9.062099156522754e-06, + 'epoch': 7.95} +04/19 [22:50:47] INFO | >> train_qwenlatent.py:487 + Step 31520 | grad_norm_pre_clip=0.1435 | + grad_norm_pre_clip_avg=0.1585 | Metrics: + {'align_loss': 0.024968115612864494, + 'recon_loss': 0.08664333820343018, + 'predict_loss': 0.007036568131297827, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14354267716407776, + 'data_time': 0.0006313729973044246, + 'model_time': 1.2316454320098273, + 'grad_norm_pre_clip_avg': 0.15848275423049926, + 'learning_rate': 9.053721475493743e-06, + 'epoch': 7.95} +04/19 [22:51:00] INFO | >> train_qwenlatent.py:487 + Step 31530 | grad_norm_pre_clip=0.1845 | + grad_norm_pre_clip_avg=0.1862 | Metrics: + {'align_loss': 0.026584651321172714, + 'recon_loss': 0.11067323386669159, + 'predict_loss': 0.0068527208641171455, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18445557355880737, + 'data_time': 0.0006485769990831614, + 'model_time': 1.2102506409864873, + 'grad_norm_pre_clip_avg': 0.18616308271884918, + 'learning_rate': 9.045345480231363e-06, + 'epoch': 7.96} +04/19 [22:51:12] INFO | >> train_qwenlatent.py:487 + Step 31540 | grad_norm_pre_clip=0.1678 | + grad_norm_pre_clip_avg=0.1698 | Metrics: + {'align_loss': 0.026804685592651367, + 'recon_loss': 0.12115434557199478, + 'predict_loss': 0.017291205003857613, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16776567697525024, + 'data_time': 0.001072093000402674, + 'model_time': 1.1908576319983695, + 'grad_norm_pre_clip_avg': 0.16982268095016478, + 'learning_rate': 9.036971174817986e-06, + 'epoch': 7.96} +04/19 [22:51:25] INFO | >> train_qwenlatent.py:487 + Step 31550 | grad_norm_pre_clip=0.1978 | + grad_norm_pre_clip_avg=0.1757 | Metrics: + {'align_loss': 0.02461833693087101, + 'recon_loss': 0.09919696301221848, + 'predict_loss': 0.008010707795619965, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1977906972169876, + 'mae_score': 0.009319501309781462, 'data_time': + 0.0006435700051952153, 'model_time': + 1.1837796040053945, 'grad_norm_pre_clip_avg': + 0.1756724588572979, 'learning_rate': + 9.028598563335126e-06, 'epoch': 7.96} +04/19 [22:51:38] INFO | >> train_qwenlatent.py:487 + Step 31560 | grad_norm_pre_clip=0.1950 | + grad_norm_pre_clip_avg=0.1932 | Metrics: + {'align_loss': 0.02594149485230446, + 'recon_loss': 0.10141140967607498, + 'predict_loss': 0.007400617469102144, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19501391053199768, + 'data_time': 0.0006328989984467626, + 'model_time': 1.19535504700616, + 'grad_norm_pre_clip_avg': 0.19318246096372604, + 'learning_rate': 9.020227649863507e-06, + 'epoch': 7.96} +04/19 [22:51:50] INFO | >> train_qwenlatent.py:487 + Step 31570 | grad_norm_pre_clip=0.1945 | + grad_norm_pre_clip_avg=0.1841 | Metrics: + {'align_loss': 0.023391231894493103, + 'recon_loss': 0.08456536382436752, + 'predict_loss': 0.00761444540694356, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19449707865715027, + 'data_time': 0.0006800240080337971, + 'model_time': 1.275795054010814, + 'grad_norm_pre_clip_avg': 0.18413929939270018, + 'learning_rate': 9.011858438483006e-06, + 'epoch': 7.97} +04/19 [22:52:03] INFO | >> train_qwenlatent.py:487 + Step 31580 | grad_norm_pre_clip=0.2206 | + grad_norm_pre_clip_avg=0.1713 | Metrics: + {'align_loss': 0.026421040296554565, + 'recon_loss': 0.11938390135765076, + 'predict_loss': 0.009151244536042213, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22057008743286133, + 'data_time': 0.0006087159854359925, + 'model_time': 1.1988591880071908, + 'grad_norm_pre_clip_avg': 0.17125470340251922, + 'learning_rate': 9.003490933272679e-06, + 'epoch': 7.97} +04/19 [22:52:15] INFO | >> train_qwenlatent.py:487 + Step 31590 | grad_norm_pre_clip=0.1632 | + grad_norm_pre_clip_avg=0.1705 | Metrics: + {'align_loss': 0.024864833801984787, + 'recon_loss': 0.07720158994197845, + 'predict_loss': 0.006019212305545807, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16318479180335999, + 'data_time': 0.0007895999879110605, + 'model_time': 1.2369441599876154, + 'grad_norm_pre_clip_avg': 0.1705360606312752, + 'learning_rate': 8.99512513831074e-06, 'epoch': + 7.97} +04/19 [22:52:28] INFO | >> train_qwenlatent.py:487 + Step 31600 | grad_norm_pre_clip=0.1425 | + grad_norm_pre_clip_avg=0.1805 | Metrics: + {'align_loss': 0.025024011731147766, + 'recon_loss': 0.09927809983491898, + 'predict_loss': 0.00823737122118473, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14254878461360931, + 'mae_score': 0.007847959501249295, 'data_time': + 0.0009932340180967003, 'model_time': + 1.3034130110172555, 'grad_norm_pre_clip_avg': + 0.18045702278614045, 'learning_rate': + 8.98676105767458e-06, 'epoch': 7.97} +04/19 [22:52:41] INFO | >> train_qwenlatent.py:487 + Step 31610 | grad_norm_pre_clip=0.1647 | + grad_norm_pre_clip_avg=0.1806 | Metrics: + {'align_loss': 0.025624115020036697, + 'recon_loss': 0.12460822612047195, + 'predict_loss': 0.010606169700622559, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1646750420331955, + 'data_time': 0.0006131069967523217, + 'model_time': 1.4806916039960925, + 'grad_norm_pre_clip_avg': 0.18058963268995284, + 'learning_rate': 8.97839869544075e-06, 'epoch': + 7.98} +04/19 [22:52:54] INFO | >> train_qwenlatent.py:487 + Step 31620 | grad_norm_pre_clip=0.1656 | + grad_norm_pre_clip_avg=0.1614 | Metrics: + {'align_loss': 0.02475341036915779, + 'recon_loss': 0.11577719449996948, + 'predict_loss': 0.015451864339411259, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1655893176794052, + 'data_time': 0.0008823759853839874, + 'model_time': 1.3735834899998736, + 'grad_norm_pre_clip_avg': 0.16143631041049958, + 'learning_rate': 8.970038055684965e-06, + 'epoch': 7.98} +04/19 [22:53:07] INFO | >> train_qwenlatent.py:487 + Step 31630 | grad_norm_pre_clip=0.1373 | + grad_norm_pre_clip_avg=0.1650 | Metrics: + {'align_loss': 0.025881825014948845, + 'recon_loss': 0.10445103049278259, + 'predict_loss': 0.006512184627354145, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13730542361736298, + 'data_time': 0.000856372993439436, + 'model_time': 1.2388634349917993, + 'grad_norm_pre_clip_avg': 0.16502000540494918, + 'learning_rate': 8.961679142482096e-06, + 'epoch': 7.98} +04/19 [22:53:20] INFO | >> train_qwenlatent.py:487 + Step 31640 | grad_norm_pre_clip=0.1301 | + grad_norm_pre_clip_avg=0.1697 | Metrics: + {'align_loss': 0.02613293007016182, + 'recon_loss': 0.0997021421790123, + 'predict_loss': 0.008167711086571217, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13012318313121796, + 'data_time': 0.0006238690111786127, + 'model_time': 1.4579775769961998, + 'grad_norm_pre_clip_avg': 0.1697013944387436, + 'learning_rate': 8.953321959906173e-06, + 'epoch': 7.98} +04/19 [22:53:33] INFO | >> train_qwenlatent.py:487 + Step 31650 | grad_norm_pre_clip=0.1774 | + grad_norm_pre_clip_avg=0.1771 | Metrics: + {'align_loss': 0.023785719648003578, + 'recon_loss': 0.11643651127815247, + 'predict_loss': 0.012699720449745655, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1773698925971985, + 'mae_score': 0.007963210612803966, 'data_time': + 0.0006215219909790903, 'model_time': + 1.2355909049801994, 'grad_norm_pre_clip_avg': + 0.1770745649933815, 'learning_rate': + 8.944966512030394e-06, 'epoch': 7.99} +04/19 [22:53:45] INFO | >> train_qwenlatent.py:487 + Step 31660 | grad_norm_pre_clip=0.1993 | + grad_norm_pre_clip_avg=0.1946 | Metrics: + {'align_loss': 0.025758419185876846, + 'recon_loss': 0.09207770228385925, + 'predict_loss': 0.006333677563816309, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19933725893497467, + 'data_time': 0.0005905669822823256, + 'model_time': 1.234901602001628, + 'grad_norm_pre_clip_avg': 0.19457294940948486, + 'learning_rate': 8.936612802927098e-06, + 'epoch': 7.99} +04/19 [22:53:58] INFO | >> train_qwenlatent.py:487 + Step 31670 | grad_norm_pre_clip=0.1795 | + grad_norm_pre_clip_avg=0.1879 | Metrics: + {'align_loss': 0.02643779292702675, + 'recon_loss': 0.11876141279935837, + 'predict_loss': 0.01433809194713831, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17952385544776917, + 'data_time': 0.0009319090167991817, + 'model_time': 1.25219680799637, + 'grad_norm_pre_clip_avg': 0.18792579770088197, + 'learning_rate': 8.928260836667777e-06, + 'epoch': 7.99} +04/19 [22:54:10] INFO | >> train_qwenlatent.py:487 + Step 31680 | grad_norm_pre_clip=0.1435 | + grad_norm_pre_clip_avg=0.1682 | Metrics: + {'align_loss': 0.02564987912774086, + 'recon_loss': 0.09698999673128128, + 'predict_loss': 0.009405208751559258, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14353418350219727, + 'data_time': 0.0009045420156326145, + 'model_time': 1.2991927179973572, + 'grad_norm_pre_clip_avg': 0.1682257056236267, + 'learning_rate': 8.919910617323089e-06, + 'epoch': 7.99} +04/19 [22:54:22] INFO | >> train_qwenlatent.py:487 + Step 31690 | grad_norm_pre_clip=0.1514 | + grad_norm_pre_clip_avg=0.1556 | Metrics: + {'align_loss': 0.024812500923871994, + 'recon_loss': 0.1270960122346878, + 'predict_loss': 0.0102394325658679, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1514459103345871, + 'data_time': 0.0006262669921852648, + 'model_time': 1.2039995419909246, + 'grad_norm_pre_clip_avg': 0.15563623756170272, + 'learning_rate': 8.911562148962818e-06, + 'epoch': 8.0} +04/19 [22:54:36] INFO | >> train_qwenlatent.py:487 + Step 31700 | grad_norm_pre_clip=0.1801 | + grad_norm_pre_clip_avg=0.1664 | Metrics: + {'align_loss': 0.024980323389172554, + 'recon_loss': 0.08434133231639862, + 'predict_loss': 0.0048721106722950935, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1800927072763443, + 'mae_score': 0.007763225108653575, 'data_time': + 0.0009449470089748502, 'model_time': + 1.2644827500043903, 'grad_norm_pre_clip_avg': + 0.1664099633693695, 'learning_rate': + 8.90321543565591e-06, 'epoch': 8.0} +04/19 [22:54:49] INFO | >> train_qwenlatent.py:487 + Step 31710 | grad_norm_pre_clip=0.1327 | + grad_norm_pre_clip_avg=0.1768 | Metrics: + {'align_loss': 0.02580985054373741, + 'recon_loss': 0.08586251735687256, + 'predict_loss': 0.005944208241999149, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13271589577198029, + 'data_time': 0.0008720319892745465, + 'model_time': 1.2737190990010276, + 'grad_norm_pre_clip_avg': 0.17684950530529023, + 'learning_rate': 8.894870481470457e-06, + 'epoch': 8.0} +04/19 [22:55:01] INFO | >> train_qwenlatent.py:487 + Step 31720 | grad_norm_pre_clip=0.1735 | + grad_norm_pre_clip_avg=0.1884 | Metrics: + {'align_loss': 0.02664583921432495, + 'recon_loss': 0.12253349274396896, + 'predict_loss': 0.010129394941031933, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17348849773406982, + 'data_time': 0.0010237850074190646, + 'model_time': 1.260813105996931, + 'grad_norm_pre_clip_avg': 0.18843164294958115, + 'learning_rate': 8.886527290473684e-06, + 'epoch': 8.0} +04/19 [22:55:14] INFO | >> train_qwenlatent.py:487 + Step 31730 | grad_norm_pre_clip=0.2699 | + grad_norm_pre_clip_avg=0.2031 | Metrics: + {'align_loss': 0.024918293580412865, + 'recon_loss': 0.0821804478764534, + 'predict_loss': 0.00713958777487278, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.26989349722862244, + 'data_time': 0.0009090700186789036, + 'model_time': 1.240875388000859, + 'grad_norm_pre_clip_avg': 0.20314153134822846, + 'learning_rate': 8.878185866731959e-06, + 'epoch': 8.01} +04/19 [22:55:26] INFO | >> train_qwenlatent.py:487 + Step 31740 | grad_norm_pre_clip=0.1592 | + grad_norm_pre_clip_avg=0.1889 | Metrics: + {'align_loss': 0.02644936740398407, + 'recon_loss': 0.1301548033952713, + 'predict_loss': 0.011505350470542908, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15922141075134277, + 'data_time': 0.0007485919923055917, + 'model_time': 1.1826487229845952, + 'grad_norm_pre_clip_avg': 0.18894651383161545, + 'learning_rate': 8.869846214310788e-06, + 'epoch': 8.01} +04/19 [22:55:39] INFO | >> train_qwenlatent.py:487 + Step 31750 | grad_norm_pre_clip=0.1734 | + grad_norm_pre_clip_avg=0.1994 | Metrics: + {'align_loss': 0.024556083604693413, + 'recon_loss': 0.07849669456481934, + 'predict_loss': 0.0070840646512806416, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17342723906040192, + 'mae_score': 0.008933971164462803, 'data_time': + 0.0006957550067454576, 'model_time': + 1.1878912989923265, 'grad_norm_pre_clip_avg': + 0.19939846396446229, 'learning_rate': + 8.861508337274826e-06, 'epoch': 8.01} +04/19 [22:55:52] INFO | >> train_qwenlatent.py:487 + Step 31760 | grad_norm_pre_clip=0.1704 | + grad_norm_pre_clip_avg=0.1695 | Metrics: + {'align_loss': 0.025454968214035034, + 'recon_loss': 0.08923620730638504, + 'predict_loss': 0.007346854545176029, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17036651074886322, + 'data_time': 0.0006294049962889403, + 'model_time': 1.3065066580020357, + 'grad_norm_pre_clip_avg': 0.16950829774141313, + 'learning_rate': 8.853172239687847e-06, + 'epoch': 8.01} +04/19 [22:56:05] INFO | >> train_qwenlatent.py:487 + Step 31770 | grad_norm_pre_clip=0.1680 | + grad_norm_pre_clip_avg=0.1696 | Metrics: + {'align_loss': 0.02537330985069275, + 'recon_loss': 0.10419636219739914, + 'predict_loss': 0.012014904990792274, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16797156631946564, + 'data_time': 0.0009293369948863983, + 'model_time': 1.1937984559917822, + 'grad_norm_pre_clip_avg': 0.16957031637430192, + 'learning_rate': 8.844837925612763e-06, + 'epoch': 8.02} +04/19 [22:56:17] INFO | >> train_qwenlatent.py:487 + Step 31780 | grad_norm_pre_clip=0.1927 | + grad_norm_pre_clip_avg=0.1817 | Metrics: + {'align_loss': 0.026009690016508102, + 'recon_loss': 0.1349511295557022, + 'predict_loss': 0.009423485025763512, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19273866713047028, + 'data_time': 0.0006824669835623354, + 'model_time': 1.268500509992009, + 'grad_norm_pre_clip_avg': 0.18172118961811065, + 'learning_rate': 8.83650539911162e-06, 'epoch': + 8.02} +04/19 [22:56:30] INFO | >> train_qwenlatent.py:487 + Step 31790 | grad_norm_pre_clip=0.1657 | + grad_norm_pre_clip_avg=0.1665 | Metrics: + {'align_loss': 0.025953024625778198, + 'recon_loss': 0.10396534204483032, + 'predict_loss': 0.0092097083106637, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16569972038269043, + 'data_time': 0.0008235959976445884, + 'model_time': 1.2571366670017596, + 'grad_norm_pre_clip_avg': 0.1665371388196945, + 'learning_rate': 8.828174664245587e-06, + 'epoch': 8.02} +04/19 [22:56:43] INFO | >> train_qwenlatent.py:487 + Step 31800 | grad_norm_pre_clip=0.1599 | + grad_norm_pre_clip_avg=0.1739 | Metrics: + {'align_loss': 0.024782566353678703, + 'recon_loss': 0.08837835490703583, + 'predict_loss': 0.010868331417441368, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15988601744174957, + 'mae_score': 0.007485381117812148, 'data_time': + 0.0008636970014777035, 'model_time': + 1.2867177099979017, 'grad_norm_pre_clip_avg': + 0.1738681212067604, 'learning_rate': + 8.819845725074973e-06, 'epoch': 8.02} +04/19 [22:56:56] INFO | >> train_qwenlatent.py:487 + Step 31810 | grad_norm_pre_clip=0.1719 | + grad_norm_pre_clip_avg=0.1725 | Metrics: + {'align_loss': 0.02498442307114601, + 'recon_loss': 0.06866125017404556, + 'predict_loss': 0.006958084646612406, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17187325656414032, + 'data_time': 0.0006879870197735727, + 'model_time': 1.2008711340022273, + 'grad_norm_pre_clip_avg': 0.17247449606657028, + 'learning_rate': 8.811518585659182e-06, + 'epoch': 8.03} +04/19 [22:57:08] INFO | >> train_qwenlatent.py:487 + Step 31820 | grad_norm_pre_clip=0.1378 | + grad_norm_pre_clip_avg=0.1695 | Metrics: + {'align_loss': 0.024721017107367516, + 'recon_loss': 0.07690370082855225, + 'predict_loss': 0.004424598067998886, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13780421018600464, + 'data_time': 0.0009731299942359328, + 'model_time': 1.2771345130167902, + 'grad_norm_pre_clip_avg': 0.16950177103281022, + 'learning_rate': 8.803193250056779e-06, + 'epoch': 8.03} +04/19 [22:57:21] INFO | >> train_qwenlatent.py:487 + Step 31830 | grad_norm_pre_clip=0.1476 | + grad_norm_pre_clip_avg=0.1890 | Metrics: + {'align_loss': 0.02440892532467842, + 'recon_loss': 0.10559891909360886, + 'predict_loss': 0.007639979477971792, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14759987592697144, + 'data_time': 0.0009490899974480271, + 'model_time': 1.2035997759958263, + 'grad_norm_pre_clip_avg': 0.18897069320082666, + 'learning_rate': 8.794869722325424e-06, + 'epoch': 8.03} +04/19 [22:57:33] INFO | >> train_qwenlatent.py:487 + Step 31840 | grad_norm_pre_clip=0.1476 | + grad_norm_pre_clip_avg=0.1733 | Metrics: + {'align_loss': 0.025005731731653214, + 'recon_loss': 0.10247741639614105, + 'predict_loss': 0.012198925018310547, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14756116271018982, + 'data_time': 0.0008147899934556335, + 'model_time': 1.2407025269931182, + 'grad_norm_pre_clip_avg': 0.17330791056156158, + 'learning_rate': 8.7865480065219e-06, 'epoch': + 8.03} +04/19 [22:57:47] INFO | >> train_qwenlatent.py:487 + Step 31850 | grad_norm_pre_clip=0.1789 | + grad_norm_pre_clip_avg=0.1816 | Metrics: + {'align_loss': 0.025351615622639656, + 'recon_loss': 0.09219357371330261, + 'predict_loss': 0.006163727957755327, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17889735102653503, + 'mae_score': 0.007178484856545388, 'data_time': + 0.0006622859800700098, 'model_time': + 1.2636306970089208, 'grad_norm_pre_clip_avg': + 0.18156545907258986, 'learning_rate': + 8.77822810670212e-06, 'epoch': 8.04} +04/19 [22:57:59] INFO | >> train_qwenlatent.py:487 + Step 31860 | grad_norm_pre_clip=0.1427 | + grad_norm_pre_clip_avg=0.1580 | Metrics: + {'align_loss': 0.024383757263422012, + 'recon_loss': 0.09201524406671524, + 'predict_loss': 0.009968943893909454, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1426580399274826, + 'data_time': 0.0006276310014072806, + 'model_time': 1.1978823590034153, + 'grad_norm_pre_clip_avg': 0.15796905606985093, + 'learning_rate': 8.76991002692109e-06, 'epoch': + 8.04} +04/19 [22:58:12] INFO | >> train_qwenlatent.py:487 + Step 31870 | grad_norm_pre_clip=0.1796 | + grad_norm_pre_clip_avg=0.1625 | Metrics: + {'align_loss': 0.025269391015172005, + 'recon_loss': 0.07128287851810455, + 'predict_loss': 0.005202638916671276, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17962808907032013, + 'data_time': 0.0008806169789750129, + 'model_time': 1.2278452519967686, + 'grad_norm_pre_clip_avg': 0.1624668687582016, + 'learning_rate': 8.761593771232953e-06, + 'epoch': 8.04} +04/19 [22:58:25] INFO | >> train_qwenlatent.py:487 + Step 31880 | grad_norm_pre_clip=0.2515 | + grad_norm_pre_clip_avg=0.2095 | Metrics: + {'align_loss': 0.025562988594174385, + 'recon_loss': 0.11576421558856964, + 'predict_loss': 0.013729625381529331, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25148245692253113, + 'data_time': 0.001087287993868813, + 'model_time': 1.594712146994425, + 'grad_norm_pre_clip_avg': 0.2095098577439785, + 'learning_rate': 8.753279343690944e-06, + 'epoch': 8.04} +04/19 [22:58:37] INFO | >> train_qwenlatent.py:487 + Step 31890 | grad_norm_pre_clip=0.2026 | + grad_norm_pre_clip_avg=0.1885 | Metrics: + {'align_loss': 0.02500777319073677, + 'recon_loss': 0.09522873908281326, + 'predict_loss': 0.010368733666837215, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20259705185890198, + 'data_time': 0.0006843199953436852, + 'model_time': 1.2217230479873251, + 'grad_norm_pre_clip_avg': 0.1884995549917221, + 'learning_rate': 8.74496674834742e-06, 'epoch': + 8.05} +04/19 [22:58:51] INFO | >> train_qwenlatent.py:487 + Step 31900 | grad_norm_pre_clip=0.1812 | + grad_norm_pre_clip_avg=0.1735 | Metrics: + {'align_loss': 0.023898452520370483, + 'recon_loss': 0.1177283227443695, + 'predict_loss': 0.008889335207641125, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18121320009231567, + 'mae_score': 0.007103520470696526, 'data_time': + 0.0007000850164331496, 'model_time': + 1.218061999999918, 'grad_norm_pre_clip_avg': + 0.17347707450389863, 'learning_rate': + 8.736655989253827e-06, 'epoch': 8.05} +04/19 [22:59:03] INFO | >> train_qwenlatent.py:487 + Step 31910 | grad_norm_pre_clip=0.1980 | + grad_norm_pre_clip_avg=0.1661 | Metrics: + {'align_loss': 0.024409331381320953, + 'recon_loss': 0.09616246819496155, + 'predict_loss': 0.007806010078638792, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19796334207057953, + 'data_time': 0.0007941279909573495, + 'model_time': 1.2910701499786228, + 'grad_norm_pre_clip_avg': 0.1661418452858925, + 'learning_rate': 8.728347070460738e-06, + 'epoch': 8.05} +04/19 [22:59:16] INFO | >> train_qwenlatent.py:487 + Step 31920 | grad_norm_pre_clip=0.1782 | + grad_norm_pre_clip_avg=0.1787 | Metrics: + {'align_loss': 0.025925524532794952, + 'recon_loss': 0.14473466575145721, + 'predict_loss': 0.008947084657847881, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17822378873825073, + 'data_time': 0.0010251289932057261, + 'model_time': 1.2557100339909084, + 'grad_norm_pre_clip_avg': 0.17865770757198335, + 'learning_rate': 8.720039996017816e-06, + 'epoch': 8.05} +04/19 [22:59:29] INFO | >> train_qwenlatent.py:487 + Step 31930 | grad_norm_pre_clip=0.1680 | + grad_norm_pre_clip_avg=0.1783 | Metrics: + {'align_loss': 0.025672677904367447, + 'recon_loss': 0.11089368164539337, + 'predict_loss': 0.005944568198174238, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16798417270183563, + 'data_time': 0.0008832919993437827, + 'model_time': 1.2678558939951472, + 'grad_norm_pre_clip_avg': 0.17830880731344223, + 'learning_rate': 8.711734769973826e-06, + 'epoch': 8.06} +04/19 [22:59:41] INFO | >> train_qwenlatent.py:487 + Step 31940 | grad_norm_pre_clip=0.1794 | + grad_norm_pre_clip_avg=0.1712 | Metrics: + {'align_loss': 0.02555195987224579, + 'recon_loss': 0.1132955327630043, + 'predict_loss': 0.008546593599021435, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17937050759792328, + 'data_time': 0.0010586150165181607, + 'model_time': 1.2224186200182885, + 'grad_norm_pre_clip_avg': 0.1711777299642563, + 'learning_rate': 8.70343139637664e-06, 'epoch': + 8.06} +04/19 [22:59:54] INFO | >> train_qwenlatent.py:487 + Step 31950 | grad_norm_pre_clip=0.1645 | + grad_norm_pre_clip_avg=0.1592 | Metrics: + {'align_loss': 0.025414209812879562, + 'recon_loss': 0.10798703134059906, + 'predict_loss': 0.009336380288004875, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16452911496162415, + 'mae_score': 0.006686119560722832, 'data_time': + 0.0006989050016272813, 'model_time': + 1.2429222950013354, 'grad_norm_pre_clip_avg': + 0.1592240497469902, 'learning_rate': + 8.69512987927322e-06, 'epoch': 8.06} +04/19 [23:00:06] INFO | >> train_qwenlatent.py:487 + Step 31960 | grad_norm_pre_clip=0.1831 | + grad_norm_pre_clip_avg=0.1642 | Metrics: + {'align_loss': 0.02580983377993107, + 'recon_loss': 0.1391274631023407, + 'predict_loss': 0.014015895314514637, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18313564360141754, + 'data_time': 0.0006448009808082134, + 'model_time': 1.2880158250045497, + 'grad_norm_pre_clip_avg': 0.1641887404024601, + 'learning_rate': 8.686830222709617e-06, + 'epoch': 8.06} +04/19 [23:00:19] INFO | >> train_qwenlatent.py:487 + Step 31970 | grad_norm_pre_clip=0.1633 | + grad_norm_pre_clip_avg=0.1683 | Metrics: + {'align_loss': 0.02451351284980774, + 'recon_loss': 0.09224619716405869, + 'predict_loss': 0.010492576286196709, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16327005624771118, + 'data_time': 0.0009068560029845685, + 'model_time': 1.2385846099932678, + 'grad_norm_pre_clip_avg': 0.1683375597000122, + 'learning_rate': 8.678532430730996e-06, + 'epoch': 8.07} +04/19 [23:00:31] INFO | >> train_qwenlatent.py:487 + Step 31980 | grad_norm_pre_clip=0.1652 | + grad_norm_pre_clip_avg=0.1670 | Metrics: + {'align_loss': 0.026261914521455765, + 'recon_loss': 0.1529374122619629, + 'predict_loss': 0.01421962771564722, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16522999107837677, + 'data_time': 0.0010384400084149092, + 'model_time': 1.202681331022177, + 'grad_norm_pre_clip_avg': 0.16700066477060319, + 'learning_rate': 8.670236507381593e-06, + 'epoch': 8.07} +04/19 [23:00:44] INFO | >> train_qwenlatent.py:487 + Step 31990 | grad_norm_pre_clip=0.1636 | + grad_norm_pre_clip_avg=0.1661 | Metrics: + {'align_loss': 0.026119422167539597, + 'recon_loss': 0.10573074966669083, + 'predict_loss': 0.007133597508072853, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16362883150577545, + 'data_time': 0.0008993940136861056, + 'model_time': 1.2001840280136093, + 'grad_norm_pre_clip_avg': 0.16608531028032303, + 'learning_rate': 8.661942456704735e-06, + 'epoch': 8.07} +04/19 [23:00:57] INFO | >> train_qwenlatent.py:487 + Step 32000 | grad_norm_pre_clip=0.3119 | + grad_norm_pre_clip_avg=0.1943 | Metrics: + {'align_loss': 0.02532789669930935, + 'recon_loss': 0.12043128907680511, + 'predict_loss': 0.010432887822389603, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.3119106888771057, + 'mae_score': 0.008816841915921048, 'data_time': + 0.001063155010342598, 'model_time': + 1.2337118690193165, 'grad_norm_pre_clip_avg': + 0.19432980492711066, 'learning_rate': + 8.653650282742845e-06, 'epoch': 8.07} +04/19 [23:01:10] INFO | >> train_qwenlatent.py:487 + Step 32010 | grad_norm_pre_clip=0.1957 | + grad_norm_pre_clip_avg=0.2173 | Metrics: + {'align_loss': 0.026890544220805168, + 'recon_loss': 0.11710210889577866, + 'predict_loss': 0.010433478280901909, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19568924605846405, + 'data_time': 0.0009220779757015407, + 'model_time': 1.2243695670040324, + 'grad_norm_pre_clip_avg': 0.21729838997125625, + 'learning_rate': 8.645359989537433e-06, + 'epoch': 8.08} +04/19 [23:01:23] INFO | >> train_qwenlatent.py:487 + Step 32020 | grad_norm_pre_clip=0.1499 | + grad_norm_pre_clip_avg=0.1729 | Metrics: + {'align_loss': 0.02593310736119747, + 'recon_loss': 0.10156460851430893, + 'predict_loss': 0.006932670716196299, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14991497993469238, + 'data_time': 0.0006464610050898045, + 'model_time': 1.576369946997147, + 'grad_norm_pre_clip_avg': 0.17287224233150483, + 'learning_rate': 8.637071581129084e-06, + 'epoch': 8.08} +04/19 [23:01:35] INFO | >> train_qwenlatent.py:487 + Step 32030 | grad_norm_pre_clip=0.1613 | + grad_norm_pre_clip_avg=0.1656 | Metrics: + {'align_loss': 0.024639107286930084, + 'recon_loss': 0.11773643642663956, + 'predict_loss': 0.009596533142030239, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16131693124771118, + 'data_time': 0.000818972010165453, + 'model_time': 1.1906520540069323, + 'grad_norm_pre_clip_avg': 0.16564978510141373, + 'learning_rate': 8.628785061557464e-06, + 'epoch': 8.08} +04/19 [23:01:48] INFO | >> train_qwenlatent.py:487 + Step 32040 | grad_norm_pre_clip=0.1948 | + grad_norm_pre_clip_avg=0.1543 | Metrics: + {'align_loss': 0.02510065585374832, + 'recon_loss': 0.07850387692451477, + 'predict_loss': 0.00554842920973897, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19481714069843292, + 'data_time': 0.0011293799907434732, + 'model_time': 1.2075259900011588, + 'grad_norm_pre_clip_avg': 0.1543480008840561, + 'learning_rate': 8.620500434861326e-06, + 'epoch': 8.08} +04/19 [23:02:01] INFO | >> train_qwenlatent.py:487 + Step 32050 | grad_norm_pre_clip=0.2180 | + grad_norm_pre_clip_avg=0.1995 | Metrics: + {'align_loss': 0.025659851729869843, + 'recon_loss': 0.12144102156162262, + 'predict_loss': 0.006969217676669359, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21798895299434662, + 'mae_score': 0.00839430147463137, 'data_time': + 0.0009731540048960596, 'model_time': + 1.1969256500015035, 'grad_norm_pre_clip_avg': + 0.19948122054338455, 'learning_rate': + 8.612217705078496e-06, 'epoch': 8.09} +04/19 [23:02:14] INFO | >> train_qwenlatent.py:487 + Step 32060 | grad_norm_pre_clip=0.2704 | + grad_norm_pre_clip_avg=0.1916 | Metrics: + {'align_loss': 0.024690648540854454, + 'recon_loss': 0.09592688083648682, + 'predict_loss': 0.0120311938226223, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.27040353417396545, + 'data_time': 0.0008970079943537712, + 'model_time': 1.2497647010022774, + 'grad_norm_pre_clip_avg': 0.19155901968479155, + 'learning_rate': 8.603936876245874e-06, + 'epoch': 8.09} +04/19 [23:02:26] INFO | >> train_qwenlatent.py:487 + Step 32070 | grad_norm_pre_clip=0.1357 | + grad_norm_pre_clip_avg=0.1867 | Metrics: + {'align_loss': 0.025686394423246384, + 'recon_loss': 0.08952181786298752, + 'predict_loss': 0.010799025185406208, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13567057251930237, + 'data_time': 0.0009917219867929816, + 'model_time': 1.245224762998987, + 'grad_norm_pre_clip_avg': 0.18665611892938613, + 'learning_rate': 8.595657952399443e-06, + 'epoch': 8.09} +04/19 [23:02:39] INFO | >> train_qwenlatent.py:487 + Step 32080 | grad_norm_pre_clip=0.2118 | + grad_norm_pre_clip_avg=0.1727 | Metrics: + {'align_loss': 0.024314802139997482, + 'recon_loss': 0.09902317821979523, + 'predict_loss': 0.005026684608310461, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21184517443180084, + 'data_time': 0.0006657049816567451, + 'model_time': 1.234328877995722, + 'grad_norm_pre_clip_avg': 0.1726736456155777, + 'learning_rate': 8.587380937574239e-06, + 'epoch': 8.09} +04/19 [23:02:51] INFO | >> train_qwenlatent.py:487 + Step 32090 | grad_norm_pre_clip=0.1501 | + grad_norm_pre_clip_avg=0.1595 | Metrics: + {'align_loss': 0.024969086050987244, + 'recon_loss': 0.07770194113254547, + 'predict_loss': 0.006392939481884241, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15007264912128448, + 'data_time': 0.0006287170108407736, + 'model_time': 1.2370293409912847, + 'grad_norm_pre_clip_avg': 0.1594685584306717, + 'learning_rate': 8.579105835804385e-06, + 'epoch': 8.1} +04/19 [23:03:04] INFO | >> train_qwenlatent.py:487 + Step 32100 | grad_norm_pre_clip=0.1357 | + grad_norm_pre_clip_avg=0.1576 | Metrics: + {'align_loss': 0.024848083034157753, + 'recon_loss': 0.11639635264873505, + 'predict_loss': 0.011041495017707348, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13570702075958252, + 'mae_score': 0.006857491398716832, 'data_time': + 0.0009801220148801804, 'model_time': + 1.2567721479863394, 'grad_norm_pre_clip_avg': + 0.15755025148391724, 'learning_rate': + 8.570832651123061e-06, 'epoch': 8.1} +04/19 [23:03:17] INFO | >> train_qwenlatent.py:487 + Step 32110 | grad_norm_pre_clip=0.2145 | + grad_norm_pre_clip_avg=0.1808 | Metrics: + {'align_loss': 0.025029826909303665, + 'recon_loss': 0.09459728002548218, + 'predict_loss': 0.00949567835777998, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21449987590312958, + 'data_time': 0.0008463299891445786, + 'model_time': 1.263251317985123, + 'grad_norm_pre_clip_avg': 0.18077463060617446, + 'learning_rate': 8.562561387562517e-06, + 'epoch': 8.1} +04/19 [23:03:30] INFO | >> train_qwenlatent.py:487 + Step 32120 | grad_norm_pre_clip=0.1699 | + grad_norm_pre_clip_avg=0.1915 | Metrics: + {'align_loss': 0.02516777440905571, + 'recon_loss': 0.1141752153635025, + 'predict_loss': 0.008391435258090496, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16985413432121277, + 'data_time': 0.0007959250069689006, + 'model_time': 1.2568515999882948, + 'grad_norm_pre_clip_avg': 0.19147282540798188, + 'learning_rate': 8.55429204915407e-06, 'epoch': + 8.1} +04/19 [23:03:42] INFO | >> train_qwenlatent.py:487 + Step 32130 | grad_norm_pre_clip=0.1544 | + grad_norm_pre_clip_avg=0.1692 | Metrics: + {'align_loss': 0.02575213462114334, + 'recon_loss': 0.09209088981151581, + 'predict_loss': 0.005887973587960005, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.154368594288826, + 'data_time': 0.0009921800228767097, + 'model_time': 1.2370387410046533, + 'grad_norm_pre_clip_avg': 0.1692178040742874, + 'learning_rate': 8.546024639928094e-06, + 'epoch': 8.11} +04/19 [23:03:55] INFO | >> train_qwenlatent.py:487 + Step 32140 | grad_norm_pre_clip=0.1981 | + grad_norm_pre_clip_avg=0.1911 | Metrics: + {'align_loss': 0.025160932913422585, + 'recon_loss': 0.13686847686767578, + 'predict_loss': 0.015586781315505505, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19806961715221405, + 'data_time': 0.0009786819864530116, + 'model_time': 1.2258365220041014, + 'grad_norm_pre_clip_avg': 0.19112623184919358, + 'learning_rate': 8.537759163914024e-06, + 'epoch': 8.11} +04/19 [23:04:08] INFO | >> train_qwenlatent.py:487 + Step 32150 | grad_norm_pre_clip=0.2009 | + grad_norm_pre_clip_avg=0.1675 | Metrics: + {'align_loss': 0.026069369167089462, + 'recon_loss': 0.12147555500268936, + 'predict_loss': 0.009525252506136894, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20093286037445068, + 'mae_score': 0.007264451722841005, 'data_time': + 0.0011132819927297533, 'model_time': + 1.2845480070100166, 'grad_norm_pre_clip_avg': + 0.16747999638319017, 'learning_rate': + 8.52949562514035e-06, 'epoch': 8.11} +04/19 [23:04:21] INFO | >> train_qwenlatent.py:487 + Step 32160 | grad_norm_pre_clip=0.1932 | + grad_norm_pre_clip_avg=0.1746 | Metrics: + {'align_loss': 0.026137733832001686, + 'recon_loss': 0.11681640148162842, + 'predict_loss': 0.015555301681160927, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1931665688753128, + 'data_time': 0.0008361370128113776, + 'model_time': 1.2551141800067853, + 'grad_norm_pre_clip_avg': 0.17462752461433412, + 'learning_rate': 8.521234027634625e-06, + 'epoch': 8.12} +04/19 [23:04:34] INFO | >> train_qwenlatent.py:487 + Step 32170 | grad_norm_pre_clip=0.1808 | + grad_norm_pre_clip_avg=0.1886 | Metrics: + {'align_loss': 0.025782017037272453, + 'recon_loss': 0.1190359890460968, + 'predict_loss': 0.008625165559351444, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18077093362808228, + 'data_time': 0.000666262989398092, + 'model_time': 1.2444005610013846, + 'grad_norm_pre_clip_avg': 0.1885981395840645, + 'learning_rate': 8.512974375423449e-06, + 'epoch': 8.12} +04/19 [23:04:46] INFO | >> train_qwenlatent.py:487 + Step 32180 | grad_norm_pre_clip=0.1732 | + grad_norm_pre_clip_avg=0.1777 | Metrics: + {'align_loss': 0.02523319609463215, + 'recon_loss': 0.0810590460896492, + 'predict_loss': 0.0046857609413564205, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17317961156368256, + 'data_time': 0.0006850330100860447, + 'model_time': 1.1826507670048159, + 'grad_norm_pre_clip_avg': 0.17769304066896438, + 'learning_rate': 8.504716672532473e-06, + 'epoch': 8.12} +04/19 [23:04:59] INFO | >> train_qwenlatent.py:487 + Step 32190 | grad_norm_pre_clip=0.1633 | + grad_norm_pre_clip_avg=0.1677 | Metrics: + {'align_loss': 0.026295632123947144, + 'recon_loss': 0.10416801273822784, + 'predict_loss': 0.007250278256833553, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1632784605026245, + 'data_time': 0.0008203100005630404, + 'model_time': 1.221222389023751, + 'grad_norm_pre_clip_avg': 0.1677141271531582, + 'learning_rate': 8.496460922986403e-06, + 'epoch': 8.12} +04/19 [23:05:12] INFO | >> train_qwenlatent.py:487 + Step 32200 | grad_norm_pre_clip=0.2328 | + grad_norm_pre_clip_avg=0.1658 | Metrics: + {'align_loss': 0.024978414177894592, + 'recon_loss': 0.10037107765674591, + 'predict_loss': 0.00795120932161808, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23278401792049408, + 'mae_score': 0.008774059742420643, 'data_time': + 0.0007629110186826438, 'model_time': + 1.2743521429947577, 'grad_norm_pre_clip_avg': + 0.1657823860645294, 'learning_rate': + 8.488207130808996e-06, 'epoch': 8.13} +04/19 [23:05:25] INFO | >> train_qwenlatent.py:487 + Step 32210 | grad_norm_pre_clip=0.2311 | + grad_norm_pre_clip_avg=0.1940 | Metrics: + {'align_loss': 0.025780905038118362, + 'recon_loss': 0.1603127121925354, + 'predict_loss': 0.007940405048429966, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23114821314811707, + 'data_time': 0.0006344690045807511, + 'model_time': 1.236361950985156, + 'grad_norm_pre_clip_avg': 0.19395502507686616, + 'learning_rate': 8.479955300023041e-06, + 'epoch': 8.13} +04/19 [23:05:37] INFO | >> train_qwenlatent.py:487 + Step 32220 | grad_norm_pre_clip=0.1310 | + grad_norm_pre_clip_avg=0.1666 | Metrics: + {'align_loss': 0.025925925001502037, + 'recon_loss': 0.12743866443634033, + 'predict_loss': 0.010718648321926594, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1310039609670639, + 'data_time': 0.0008399249927606434, + 'model_time': 1.2042225839977618, + 'grad_norm_pre_clip_avg': 0.16662674844264985, + 'learning_rate': 8.471705434650386e-06, + 'epoch': 8.13} +04/19 [23:05:50] INFO | >> train_qwenlatent.py:487 + Step 32230 | grad_norm_pre_clip=0.1947 | + grad_norm_pre_clip_avg=0.1602 | Metrics: + {'align_loss': 0.023969609290361404, + 'recon_loss': 0.08476968109607697, + 'predict_loss': 0.008641873486340046, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1946665644645691, + 'data_time': 0.0006376999954227358, + 'model_time': 1.2334439379919786, + 'grad_norm_pre_clip_avg': 0.16015681773424148, + 'learning_rate': 8.463457538711915e-06, + 'epoch': 8.13} +04/19 [23:06:02] INFO | >> train_qwenlatent.py:487 + Step 32240 | grad_norm_pre_clip=0.1690 | + grad_norm_pre_clip_avg=0.1869 | Metrics: + {'align_loss': 0.025774020701646805, + 'recon_loss': 0.10947761684656143, + 'predict_loss': 0.007607878651469946, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1689758151769638, + 'data_time': 0.0008876829815562814, + 'model_time': 1.2426918359997217, + 'grad_norm_pre_clip_avg': 0.18689917474985124, + 'learning_rate': 8.455211616227553e-06, + 'epoch': 8.14} +04/19 [23:06:16] INFO | >> train_qwenlatent.py:487 + Step 32250 | grad_norm_pre_clip=0.2191 | + grad_norm_pre_clip_avg=0.1697 | Metrics: + {'align_loss': 0.027024490758776665, + 'recon_loss': 0.09242113679647446, + 'predict_loss': 0.006674956530332565, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21908897161483765, + 'mae_score': 0.007985588451763531, 'data_time': + 0.0008568839984945953, 'model_time': + 1.2378294999944046, 'grad_norm_pre_clip_avg': + 0.1697275772690773, 'learning_rate': + 8.446967671216258e-06, 'epoch': 8.14} +04/19 [23:06:28] INFO | >> train_qwenlatent.py:487 + Step 32260 | grad_norm_pre_clip=0.2187 | + grad_norm_pre_clip_avg=0.1568 | Metrics: + {'align_loss': 0.025987662374973297, + 'recon_loss': 0.11975689232349396, + 'predict_loss': 0.007252897601574659, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21866241097450256, + 'data_time': 0.0011476519866846502, + 'model_time': 1.2600487660092767, + 'grad_norm_pre_clip_avg': 0.156753072142601, + 'learning_rate': 8.438725707696032e-06, + 'epoch': 8.14} +04/19 [23:06:41] INFO | >> train_qwenlatent.py:487 + Step 32270 | grad_norm_pre_clip=0.1540 | + grad_norm_pre_clip_avg=0.1525 | Metrics: + {'align_loss': 0.024002239108085632, + 'recon_loss': 0.09105531871318817, + 'predict_loss': 0.008665633387863636, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1540418267250061, + 'data_time': 0.0006389390036929399, + 'model_time': 1.256274162005866, + 'grad_norm_pre_clip_avg': 0.15253120362758638, + 'learning_rate': 8.430485729683909e-06, + 'epoch': 8.14} +04/19 [23:06:54] INFO | >> train_qwenlatent.py:487 + Step 32280 | grad_norm_pre_clip=0.1791 | + grad_norm_pre_clip_avg=0.1624 | Metrics: + {'align_loss': 0.025491531938314438, + 'recon_loss': 0.10352420806884766, + 'predict_loss': 0.006366329733282328, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17910423874855042, + 'data_time': 0.0008940050029195845, + 'model_time': 1.6676825679896865, + 'grad_norm_pre_clip_avg': 0.1624026671051979, + 'learning_rate': 8.42224774119595e-06, 'epoch': + 8.15} +04/19 [23:07:07] INFO | >> train_qwenlatent.py:487 + Step 32290 | grad_norm_pre_clip=0.1549 | + grad_norm_pre_clip_avg=0.1936 | Metrics: + {'align_loss': 0.025583524256944656, + 'recon_loss': 0.1039881482720375, + 'predict_loss': 0.007542984094470739, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15487831830978394, + 'data_time': 0.0006636049947701395, + 'model_time': 1.2612133259826805, + 'grad_norm_pre_clip_avg': 0.1936432108283043, + 'learning_rate': 8.414011746247253e-06, + 'epoch': 8.15} +04/19 [23:07:20] INFO | >> train_qwenlatent.py:487 + Step 32300 | grad_norm_pre_clip=0.1892 | + grad_norm_pre_clip_avg=0.1683 | Metrics: + {'align_loss': 0.026956886053085327, + 'recon_loss': 0.11825425922870636, + 'predict_loss': 0.007746023125946522, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18920005857944489, + 'mae_score': 0.00803471041155291, 'data_time': + 0.0010665020090527833, 'model_time': + 1.203857115993742, 'grad_norm_pre_clip_avg': + 0.16826145499944686, 'learning_rate': + 8.405777748851943e-06, 'epoch': 8.15} +04/19 [23:07:33] INFO | >> train_qwenlatent.py:487 + Step 32310 | grad_norm_pre_clip=0.2891 | + grad_norm_pre_clip_avg=0.2201 | Metrics: + {'align_loss': 0.026380427181720734, + 'recon_loss': 0.12300774455070496, + 'predict_loss': 0.015383020974695683, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2891283631324768, + 'data_time': 0.0011960029951296747, + 'model_time': 1.331275653996272, + 'grad_norm_pre_clip_avg': 0.2200912907719612, + 'learning_rate': 8.397545753023166e-06, + 'epoch': 8.15} +04/19 [23:07:45] INFO | >> train_qwenlatent.py:487 + Step 32320 | grad_norm_pre_clip=0.1446 | + grad_norm_pre_clip_avg=0.1897 | Metrics: + {'align_loss': 0.02553960680961609, + 'recon_loss': 0.11605649441480637, + 'predict_loss': 0.008633422665297985, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14458605647087097, + 'data_time': 0.0012240289943292737, + 'model_time': 1.256204647012055, + 'grad_norm_pre_clip_avg': 0.18968817144632338, + 'learning_rate': 8.389315762773104e-06, + 'epoch': 8.16} +04/19 [23:07:58] INFO | >> train_qwenlatent.py:487 + Step 32330 | grad_norm_pre_clip=0.1404 | + grad_norm_pre_clip_avg=0.1574 | Metrics: + {'align_loss': 0.02706536464393139, + 'recon_loss': 0.14573416113853455, + 'predict_loss': 0.013276977464556694, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14039914309978485, + 'data_time': 0.0010402660118415952, + 'model_time': 1.2433325810125098, + 'grad_norm_pre_clip_avg': 0.1573527052998543, + 'learning_rate': 8.381087782112954e-06, + 'epoch': 8.16} +04/19 [23:08:11] INFO | >> train_qwenlatent.py:487 + Step 32340 | grad_norm_pre_clip=0.1527 | + grad_norm_pre_clip_avg=0.1747 | Metrics: + {'align_loss': 0.02505991980433464, + 'recon_loss': 0.062013134360313416, + 'predict_loss': 0.005316813942044973, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15268591046333313, + 'data_time': 0.001015549001749605, + 'model_time': 1.2275512010091916, + 'grad_norm_pre_clip_avg': 0.17471010237932205, + 'learning_rate': 8.372861815052924e-06, + 'epoch': 8.16} +04/19 [23:08:24] INFO | >> train_qwenlatent.py:487 + Step 32350 | grad_norm_pre_clip=0.1613 | + grad_norm_pre_clip_avg=0.1514 | Metrics: + {'align_loss': 0.026048291474580765, + 'recon_loss': 0.09493441879749298, + 'predict_loss': 0.00860042218118906, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16126875579357147, + 'mae_score': 0.00754791294132267, 'data_time': + 0.0006608389958273619, 'model_time': + 1.2337932850059588, 'grad_norm_pre_clip_avg': + 0.15144241824746132, 'learning_rate': + 8.36463786560226e-06, 'epoch': 8.16} +04/19 [23:08:36] INFO | >> train_qwenlatent.py:487 + Step 32360 | grad_norm_pre_clip=0.1701 | + grad_norm_pre_clip_avg=0.1767 | Metrics: + {'align_loss': 0.023170452564954758, + 'recon_loss': 0.11492133140563965, + 'predict_loss': 0.012550500221550465, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17012128233909607, + 'data_time': 0.0007230440096464008, + 'model_time': 1.2952625080069993, + 'grad_norm_pre_clip_avg': 0.1767447978258133, + 'learning_rate': 8.356415937769207e-06, + 'epoch': 8.17} +04/19 [23:08:49] INFO | >> train_qwenlatent.py:487 + Step 32370 | grad_norm_pre_clip=0.1393 | + grad_norm_pre_clip_avg=0.1725 | Metrics: + {'align_loss': 0.025567881762981415, + 'recon_loss': 0.08422145992517471, + 'predict_loss': 0.008433248847723007, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1392977088689804, + 'data_time': 0.0010190980101469904, + 'model_time': 1.5160673090140335, + 'grad_norm_pre_clip_avg': 0.1724702700972557, + 'learning_rate': 8.348196035561038e-06, + 'epoch': 8.17} +04/19 [23:09:02] INFO | >> train_qwenlatent.py:487 + Step 32380 | grad_norm_pre_clip=0.1978 | + grad_norm_pre_clip_avg=0.2015 | Metrics: + {'align_loss': 0.02495725266635418, + 'recon_loss': 0.0888521745800972, + 'predict_loss': 0.005891100037842989, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19777348637580872, + 'data_time': 0.0009019559947773814, + 'model_time': 1.2360879979969468, + 'grad_norm_pre_clip_avg': 0.2015111602842808, + 'learning_rate': 8.339978162984033e-06, + 'epoch': 8.17} +04/19 [23:09:14] INFO | >> train_qwenlatent.py:487 + Step 32390 | grad_norm_pre_clip=0.1634 | + grad_norm_pre_clip_avg=0.1664 | Metrics: + {'align_loss': 0.026894066482782364, + 'recon_loss': 0.14596620202064514, + 'predict_loss': 0.010670814663171768, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16342954337596893, + 'data_time': 0.0012011970102321357, + 'model_time': 1.2598847259941977, + 'grad_norm_pre_clip_avg': 0.16639942079782485, + 'learning_rate': 8.331762324043484e-06, + 'epoch': 8.17} +04/19 [23:09:28] INFO | >> train_qwenlatent.py:487 + Step 32400 | grad_norm_pre_clip=0.1459 | + grad_norm_pre_clip_avg=0.1780 | Metrics: + {'align_loss': 0.025369690731167793, + 'recon_loss': 0.10669992864131927, + 'predict_loss': 0.004794455133378506, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.145937979221344, + 'mae_score': 0.007781653361277537, 'data_time': + 0.0006524529890157282, 'model_time': + 1.2235209089994896, 'grad_norm_pre_clip_avg': + 0.17796539664268493, 'learning_rate': + 8.323548522743692e-06, 'epoch': 8.18} +04/19 [23:09:40] INFO | >> train_qwenlatent.py:487 + Step 32410 | grad_norm_pre_clip=0.1499 | + grad_norm_pre_clip_avg=0.1785 | Metrics: + {'align_loss': 0.02619228884577751, + 'recon_loss': 0.08725278079509735, + 'predict_loss': 0.005092253442853689, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14988912642002106, + 'data_time': 0.0009926219936460257, + 'model_time': 1.2787669830140658, + 'grad_norm_pre_clip_avg': 0.17854184359312059, + 'learning_rate': 8.315336763087959e-06, + 'epoch': 8.18} +04/19 [23:09:54] INFO | >> train_qwenlatent.py:487 + Step 32420 | grad_norm_pre_clip=0.2002 | + grad_norm_pre_clip_avg=0.1620 | Metrics: + {'align_loss': 0.02500138059258461, + 'recon_loss': 0.07338229566812515, + 'predict_loss': 0.005722688511013985, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2001783847808838, + 'data_time': 0.0005679280147887766, + 'model_time': 1.4818396419868805, + 'grad_norm_pre_clip_avg': 0.16199061200022696, + 'learning_rate': 8.307127049078604e-06, + 'epoch': 8.18} +04/19 [23:10:06] INFO | >> train_qwenlatent.py:487 + Step 32430 | grad_norm_pre_clip=0.1136 | + grad_norm_pre_clip_avg=0.1581 | Metrics: + {'align_loss': 0.025676270946860313, + 'recon_loss': 0.11390161514282227, + 'predict_loss': 0.006562304217368364, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1135987713932991, + 'data_time': 0.0006248410209082067, + 'model_time': 1.2326930590206757, + 'grad_norm_pre_clip_avg': 0.15811544358730317, + 'learning_rate': 8.298919384716935e-06, + 'epoch': 8.18} +04/19 [23:10:19] INFO | >> train_qwenlatent.py:487 + Step 32440 | grad_norm_pre_clip=0.1789 | + grad_norm_pre_clip_avg=0.1615 | Metrics: + {'align_loss': 0.02560248225927353, + 'recon_loss': 0.10786580294370651, + 'predict_loss': 0.01207451056689024, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1789061278104782, + 'data_time': 0.0006538940069731325, + 'model_time': 1.2072043539956212, + 'grad_norm_pre_clip_avg': 0.16152506470680236, + 'learning_rate': 8.290713774003273e-06, + 'epoch': 8.19} +04/19 [23:10:32] INFO | >> train_qwenlatent.py:487 + Step 32450 | grad_norm_pre_clip=0.1477 | + grad_norm_pre_clip_avg=0.1693 | Metrics: + {'align_loss': 0.026831697672605515, + 'recon_loss': 0.1346912682056427, + 'predict_loss': 0.009182929992675781, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14768703281879425, + 'mae_score': 0.008709892925915418, 'data_time': + 0.0008943990105763078, 'model_time': + 1.174340552999638, 'grad_norm_pre_clip_avg': + 0.16925970166921617, 'learning_rate': + 8.282510220936924e-06, 'epoch': 8.19} +04/19 [23:10:44] INFO | >> train_qwenlatent.py:487 + Step 32460 | grad_norm_pre_clip=0.1618 | + grad_norm_pre_clip_avg=0.1495 | Metrics: + {'align_loss': 0.024442804977297783, + 'recon_loss': 0.0731927827000618, + 'predict_loss': 0.006293162237852812, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16176201403141022, + 'data_time': 0.0007245460001286119, + 'model_time': 1.2044505509838928, + 'grad_norm_pre_clip_avg': 0.14954735189676285, + 'learning_rate': 8.274308729516213e-06, + 'epoch': 8.19} +04/19 [23:10:56] INFO | >> train_qwenlatent.py:487 + Step 32470 | grad_norm_pre_clip=0.2047 | + grad_norm_pre_clip_avg=0.1667 | Metrics: + {'align_loss': 0.0263526551425457, + 'recon_loss': 0.09121266007423401, + 'predict_loss': 0.008895604871213436, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2047007828950882, + 'data_time': 0.0006669659924227744, + 'model_time': 1.2076325250091031, + 'grad_norm_pre_clip_avg': 0.16668184399604796, + 'learning_rate': 8.26610930373844e-06, 'epoch': + 8.19} +04/19 [23:11:09] INFO | >> train_qwenlatent.py:487 + Step 32480 | grad_norm_pre_clip=0.1779 | + grad_norm_pre_clip_avg=0.2213 | Metrics: + {'align_loss': 0.02525157481431961, + 'recon_loss': 0.10636650025844574, + 'predict_loss': 0.006113230716437101, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17791354656219482, + 'data_time': 0.000880296021932736, + 'model_time': 1.2342456770129502, + 'grad_norm_pre_clip_avg': 0.2212910011410713, + 'learning_rate': 8.257911947599905e-06, + 'epoch': 8.2} +04/19 [23:11:21] INFO | >> train_qwenlatent.py:487 + Step 32490 | grad_norm_pre_clip=0.1499 | + grad_norm_pre_clip_avg=0.1632 | Metrics: + {'align_loss': 0.02541979029774666, + 'recon_loss': 0.13688847422599792, + 'predict_loss': 0.009627035818994045, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14991018176078796, + 'data_time': 0.0010740949946921319, + 'model_time': 1.2741262739873491, + 'grad_norm_pre_clip_avg': 0.16321392357349396, + 'learning_rate': 8.249716665095902e-06, + 'epoch': 8.2} +04/19 [23:11:34] INFO | >> train_qwenlatent.py:487 + Step 32500 | grad_norm_pre_clip=0.1771 | + grad_norm_pre_clip_avg=0.1501 | Metrics: + {'align_loss': 0.02435382641851902, + 'recon_loss': 0.0835656002163887, + 'predict_loss': 0.010365244001150131, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17708547413349152, + 'mae_score': 0.007137610031677796, 'data_time': + 0.0007328960055019706, 'model_time': + 1.2796571590006351, 'grad_norm_pre_clip_avg': + 0.15009944811463355, 'learning_rate': + 8.241523460220712e-06, 'epoch': 8.2} +04/19 [23:11:47] INFO | >> train_qwenlatent.py:487 + Step 32510 | grad_norm_pre_clip=0.1886 | + grad_norm_pre_clip_avg=0.1793 | Metrics: + {'align_loss': 0.026360079646110535, + 'recon_loss': 0.13032549619674683, + 'predict_loss': 0.007719861343502998, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18860791623592377, + 'data_time': 0.0009834069933276623, + 'model_time': 1.230587636004202, + 'grad_norm_pre_clip_avg': 0.17934102714061737, + 'learning_rate': 8.233332336967606e-06, + 'epoch': 8.2} +04/19 [23:12:00] INFO | >> train_qwenlatent.py:487 + Step 32520 | grad_norm_pre_clip=0.1474 | + grad_norm_pre_clip_avg=0.1654 | Metrics: + {'align_loss': 0.025135613977909088, + 'recon_loss': 0.10223402827978134, + 'predict_loss': 0.005602584220468998, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1474395990371704, + 'data_time': 0.0006251320010051131, + 'model_time': 1.274126521020662, + 'grad_norm_pre_clip_avg': 0.16539641618728637, + 'learning_rate': 8.22514329932883e-06, 'epoch': + 8.21} +04/19 [23:12:12] INFO | >> train_qwenlatent.py:487 + Step 32530 | grad_norm_pre_clip=0.1604 | + grad_norm_pre_clip_avg=0.1631 | Metrics: + {'align_loss': 0.026299702003598213, + 'recon_loss': 0.10376475751399994, + 'predict_loss': 0.01006134320050478, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1603657752275467, + 'data_time': 0.0008671599789522588, + 'model_time': 1.2540500739996787, + 'grad_norm_pre_clip_avg': 0.16312178373336791, + 'learning_rate': 8.21695635129563e-06, 'epoch': + 8.21} +04/19 [23:12:25] INFO | >> train_qwenlatent.py:487 + Step 32540 | grad_norm_pre_clip=0.1674 | + grad_norm_pre_clip_avg=0.1644 | Metrics: + {'align_loss': 0.02524620294570923, + 'recon_loss': 0.07824177294969559, + 'predict_loss': 0.005085987504571676, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16736121475696564, + 'data_time': 0.0006434759998228401, + 'model_time': 1.2095776700007264, + 'grad_norm_pre_clip_avg': 0.16438655704259872, + 'learning_rate': 8.20877149685822e-06, 'epoch': + 8.21} +04/19 [23:12:38] INFO | >> train_qwenlatent.py:487 + Step 32550 | grad_norm_pre_clip=0.1670 | + grad_norm_pre_clip_avg=0.2040 | Metrics: + {'align_loss': 0.024838857352733612, + 'recon_loss': 0.08613809198141098, + 'predict_loss': 0.006827313918620348, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16696041822433472, + 'mae_score': 0.010674767021660331, 'data_time': + 0.0008711899863556027, 'model_time': + 1.2371858870028518, 'grad_norm_pre_clip_avg': + 0.20398396253585815, 'learning_rate': + 8.200588740005799e-06, 'epoch': 8.21} +04/19 [23:12:52] INFO | >> train_qwenlatent.py:487 + Step 32560 | grad_norm_pre_clip=0.1395 | + grad_norm_pre_clip_avg=0.1824 | Metrics: + {'align_loss': 0.026097871363162994, + 'recon_loss': 0.07763399183750153, + 'predict_loss': 0.005225531291216612, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13946884870529175, + 'data_time': 0.0008788069826550782, + 'model_time': 1.5065640599932522, + 'grad_norm_pre_clip_avg': 0.18243939578533172, + 'learning_rate': 8.192408084726547e-06, + 'epoch': 8.22} +04/19 [23:13:04] INFO | >> train_qwenlatent.py:487 + Step 32570 | grad_norm_pre_clip=0.3120 | + grad_norm_pre_clip_avg=0.1816 | Metrics: + {'align_loss': 0.025618435814976692, + 'recon_loss': 0.14943309128284454, + 'predict_loss': 0.013779271394014359, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.31202206015586853, + 'data_time': 0.0009145679941866547, + 'model_time': 1.2311116170021705, + 'grad_norm_pre_clip_avg': 0.18161813616752626, + 'learning_rate': 8.18422953500761e-06, 'epoch': + 8.22} +04/19 [23:13:17] INFO | >> train_qwenlatent.py:487 + Step 32580 | grad_norm_pre_clip=0.1445 | + grad_norm_pre_clip_avg=0.1887 | Metrics: + {'align_loss': 0.025654872879385948, + 'recon_loss': 0.09086128324270248, + 'predict_loss': 0.007252981886267662, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1445319950580597, + 'data_time': 0.0007558139914181083, + 'model_time': 1.2174717239977326, + 'grad_norm_pre_clip_avg': 0.18869127482175826, + 'learning_rate': 8.176053094835121e-06, + 'epoch': 8.22} +04/19 [23:13:29] INFO | >> train_qwenlatent.py:487 + Step 32590 | grad_norm_pre_clip=0.1562 | + grad_norm_pre_clip_avg=0.1638 | Metrics: + {'align_loss': 0.025259485468268394, + 'recon_loss': 0.12815339863300323, + 'predict_loss': 0.008029604330658913, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15621906518936157, + 'data_time': 0.000859405001392588, + 'model_time': 1.2459993310039863, + 'grad_norm_pre_clip_avg': 0.16380947828292847, + 'learning_rate': 8.167878768194172e-06, + 'epoch': 8.22} +04/19 [23:13:42] INFO | >> train_qwenlatent.py:487 + Step 32600 | grad_norm_pre_clip=0.1760 | + grad_norm_pre_clip_avg=0.1669 | Metrics: + {'align_loss': 0.024808533489704132, + 'recon_loss': 0.10288609564304352, + 'predict_loss': 0.00903512816876173, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17604033648967743, + 'mae_score': 0.007828143695453265, 'data_time': + 0.0009058660070877522, 'model_time': + 1.2025867459888104, 'grad_norm_pre_clip_avg': + 0.16687072515487672, 'learning_rate': + 8.159706559068836e-06, 'epoch': 8.23} +04/19 [23:13:55] INFO | >> train_qwenlatent.py:487 + Step 32610 | grad_norm_pre_clip=0.1670 | + grad_norm_pre_clip_avg=0.1793 | Metrics: + {'align_loss': 0.02497486211359501, + 'recon_loss': 0.09687557816505432, + 'predict_loss': 0.00703606428578496, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1669611930847168, + 'data_time': 0.000650305999442935, + 'model_time': 1.2131156809919048, + 'grad_norm_pre_clip_avg': 0.17926553934812545, + 'learning_rate': 8.151536471442142e-06, + 'epoch': 8.23} +04/19 [23:14:07] INFO | >> train_qwenlatent.py:487 + Step 32620 | grad_norm_pre_clip=0.1482 | + grad_norm_pre_clip_avg=0.1714 | Metrics: + {'align_loss': 0.02494584396481514, + 'recon_loss': 0.09169872105121613, + 'predict_loss': 0.007680163718760014, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14820942282676697, + 'data_time': 0.0006710730085615069, + 'model_time': 1.217978739994578, + 'grad_norm_pre_clip_avg': 0.17139461487531663, + 'learning_rate': 8.143368509296091e-06, + 'epoch': 8.23} +04/19 [23:14:20] INFO | >> train_qwenlatent.py:487 + Step 32630 | grad_norm_pre_clip=0.1868 | + grad_norm_pre_clip_avg=0.1645 | Metrics: + {'align_loss': 0.024741891771554947, + 'recon_loss': 0.07557458430528641, + 'predict_loss': 0.008499901741743088, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18684011697769165, + 'data_time': 0.0006704450061079115, + 'model_time': 1.2461434729921166, + 'grad_norm_pre_clip_avg': 0.16452525705099105, + 'learning_rate': 8.135202676611656e-06, + 'epoch': 8.23} +04/19 [23:14:32] INFO | >> train_qwenlatent.py:487 + Step 32640 | grad_norm_pre_clip=0.1842 | + grad_norm_pre_clip_avg=0.1989 | Metrics: + {'align_loss': 0.025478025898337364, + 'recon_loss': 0.10496854782104492, + 'predict_loss': 0.010646451264619827, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1841568797826767, + 'data_time': 0.0006469200015999377, + 'model_time': 1.24034911300987, + 'grad_norm_pre_clip_avg': 0.19885327965021132, + 'learning_rate': 8.127038977368755e-06, + 'epoch': 8.24} +04/19 [23:14:46] INFO | >> train_qwenlatent.py:487 + Step 32650 | grad_norm_pre_clip=0.2134 | + grad_norm_pre_clip_avg=0.2072 | Metrics: + {'align_loss': 0.026242151856422424, + 'recon_loss': 0.13498613238334656, + 'predict_loss': 0.007388672791421413, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21337413787841797, + 'mae_score': 0.006723040503424567, 'data_time': + 0.0008228740189224482, 'model_time': + 1.2348972040053923, 'grad_norm_pre_clip_avg': + 0.20721710324287415, 'learning_rate': + 8.118877415546288e-06, 'epoch': 8.24} +04/19 [23:14:58] INFO | >> train_qwenlatent.py:487 + Step 32660 | grad_norm_pre_clip=0.1674 | + grad_norm_pre_clip_avg=0.1821 | Metrics: + {'align_loss': 0.02627556025981903, + 'recon_loss': 0.10363475233316422, + 'predict_loss': 0.014846398495137691, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16735109686851501, + 'data_time': 0.0007442510104738176, + 'model_time': 1.2063956980127841, + 'grad_norm_pre_clip_avg': 0.18208741545677185, + 'learning_rate': 8.110717995122089e-06, + 'epoch': 8.24} +04/19 [23:15:11] INFO | >> train_qwenlatent.py:487 + Step 32670 | grad_norm_pre_clip=0.2120 | + grad_norm_pre_clip_avg=0.1843 | Metrics: + {'align_loss': 0.02563304826617241, + 'recon_loss': 0.11685219407081604, + 'predict_loss': 0.008361287415027618, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.211951345205307, + 'data_time': 0.0010262250143568963, + 'model_time': 1.6235308470204473, + 'grad_norm_pre_clip_avg': 0.1843023642897606, + 'learning_rate': 8.102560720072968e-06, + 'epoch': 8.24} +04/19 [23:15:24] INFO | >> train_qwenlatent.py:487 + Step 32680 | grad_norm_pre_clip=0.1616 | + grad_norm_pre_clip_avg=0.1494 | Metrics: + {'align_loss': 0.025454081594944, 'recon_loss': + 0.0927402451634407, 'predict_loss': + 0.009035762399435043, 'aux_loss_decay_weight': + 0.0, 'grad_norm_pre_clip': 0.16155919432640076, + 'data_time': 0.0007277569966390729, + 'model_time': 1.1994137240108103, + 'grad_norm_pre_clip_avg': 0.14941368550062178, + 'learning_rate': 8.09440559437468e-06, 'epoch': + 8.25} +04/19 [23:15:37] INFO | >> train_qwenlatent.py:487 + Step 32690 | grad_norm_pre_clip=0.1782 | + grad_norm_pre_clip_avg=0.1727 | Metrics: + {'align_loss': 0.025787103921175003, + 'recon_loss': 0.14053575694561005, + 'predict_loss': 0.016207747161388397, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17820343375205994, + 'data_time': 0.0006384799780789763, + 'model_time': 1.1831339969940018, + 'grad_norm_pre_clip_avg': 0.17269900888204576, + 'learning_rate': 8.086252622001936e-06, + 'epoch': 8.25} +04/19 [23:15:50] INFO | >> train_qwenlatent.py:487 + Step 32700 | grad_norm_pre_clip=0.1209 | + grad_norm_pre_clip_avg=0.1684 | Metrics: + {'align_loss': 0.02648179791867733, + 'recon_loss': 0.09303872287273407, + 'predict_loss': 0.00537069421261549, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.120852030813694, + 'mae_score': 0.008142975643948391, 'data_time': + 0.0006328920135274529, 'model_time': + 1.2109705629991367, 'grad_norm_pre_clip_avg': + 0.16836270019412042, 'learning_rate': + 8.078101806928391e-06, 'epoch': 8.25} +04/19 [23:16:03] INFO | >> train_qwenlatent.py:487 + Step 32710 | grad_norm_pre_clip=0.1584 | + grad_norm_pre_clip_avg=0.1714 | Metrics: + {'align_loss': 0.025874461978673935, + 'recon_loss': 0.10134318470954895, + 'predict_loss': 0.004706607200205326, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1584460735321045, + 'data_time': 0.000756276014726609, + 'model_time': 1.2239037730032578, + 'grad_norm_pre_clip_avg': 0.17136360704898834, + 'learning_rate': 8.069953153126652e-06, + 'epoch': 8.25} +04/19 [23:16:15] INFO | >> train_qwenlatent.py:487 + Step 32720 | grad_norm_pre_clip=0.1395 | + grad_norm_pre_clip_avg=0.1664 | Metrics: + {'align_loss': 0.02617368847131729, + 'recon_loss': 0.07366562634706497, + 'predict_loss': 0.006225937511771917, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13953697681427002, + 'data_time': 0.0009119900059886277, + 'model_time': 1.1880034479836468, + 'grad_norm_pre_clip_avg': 0.16635954529047012, + 'learning_rate': 8.061806664568284e-06, + 'epoch': 8.26} +04/19 [23:16:28] INFO | >> train_qwenlatent.py:487 + Step 32730 | grad_norm_pre_clip=0.1795 | + grad_norm_pre_clip_avg=0.1632 | Metrics: + {'align_loss': 0.02590475231409073, + 'recon_loss': 0.0965522974729538, + 'predict_loss': 0.008619748055934906, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1795070320367813, + 'data_time': 0.0008679009915795177, + 'model_time': 1.20458957698429, + 'grad_norm_pre_clip_avg': 0.1632433608174324, + 'learning_rate': 8.053662345223777e-06, + 'epoch': 8.26} +04/19 [23:16:40] INFO | >> train_qwenlatent.py:487 + Step 32740 | grad_norm_pre_clip=0.1729 | + grad_norm_pre_clip_avg=0.1876 | Metrics: + {'align_loss': 0.02517126128077507, + 'recon_loss': 0.10203831642866135, + 'predict_loss': 0.004920978099107742, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17293208837509155, + 'data_time': 0.0009296710195485502, + 'model_time': 1.2523177760012913, + 'grad_norm_pre_clip_avg': 0.18763518184423447, + 'learning_rate': 8.045520199062574e-06, + 'epoch': 8.26} +04/19 [23:16:53] INFO | >> train_qwenlatent.py:487 + Step 32750 | grad_norm_pre_clip=0.2241 | + grad_norm_pre_clip_avg=0.1662 | Metrics: + {'align_loss': 0.026220420375466347, + 'recon_loss': 0.08843155950307846, + 'predict_loss': 0.008813844993710518, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2241254299879074, + 'mae_score': 0.008822381818616713, 'data_time': + 0.0009979400201700628, 'model_time': + 1.2399227489950135, 'grad_norm_pre_clip_avg': + 0.16621319204568863, 'learning_rate': + 8.037380230053063e-06, 'epoch': 8.26} +04/19 [23:17:06] INFO | >> train_qwenlatent.py:487 + Step 32760 | grad_norm_pre_clip=0.1986 | + grad_norm_pre_clip_avg=0.2213 | Metrics: + {'align_loss': 0.025710146874189377, + 'recon_loss': 0.09374193102121353, + 'predict_loss': 0.007358720526099205, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19856755435466766, + 'data_time': 0.0006692080060020089, + 'model_time': 1.2316421380091924, + 'grad_norm_pre_clip_avg': 0.22132223844528198, + 'learning_rate': 8.029242442162563e-06, + 'epoch': 8.27} +04/19 [23:17:19] INFO | >> train_qwenlatent.py:487 + Step 32770 | grad_norm_pre_clip=0.1649 | + grad_norm_pre_clip_avg=0.1744 | Metrics: + {'align_loss': 0.02521597221493721, + 'recon_loss': 0.11572117358446121, + 'predict_loss': 0.009728278033435345, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16490507125854492, + 'data_time': 0.000713301997166127, + 'model_time': 1.2116472960042302, + 'grad_norm_pre_clip_avg': 0.1744486376643181, + 'learning_rate': 8.021106839357333e-06, + 'epoch': 8.27} +04/19 [23:17:31] INFO | >> train_qwenlatent.py:487 + Step 32780 | grad_norm_pre_clip=0.1814 | + grad_norm_pre_clip_avg=0.1612 | Metrics: + {'align_loss': 0.02423371747136116, + 'recon_loss': 0.12738534808158875, + 'predict_loss': 0.008945751003921032, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1813659369945526, + 'data_time': 0.0007083690143190324, + 'model_time': 1.2261853110103402, + 'grad_norm_pre_clip_avg': 0.16120380014181138, + 'learning_rate': 8.01297342560257e-06, 'epoch': + 8.27} +04/19 [23:17:44] INFO | >> train_qwenlatent.py:487 + Step 32790 | grad_norm_pre_clip=0.1991 | + grad_norm_pre_clip_avg=0.1756 | Metrics: + {'align_loss': 0.025720417499542236, + 'recon_loss': 0.1270323544740677, + 'predict_loss': 0.007042178884148598, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19910140335559845, + 'data_time': 0.0007241459970828146, + 'model_time': 1.1950779040053021, + 'grad_norm_pre_clip_avg': 0.17556502670049667, + 'learning_rate': 8.004842204862397e-06, + 'epoch': 8.27} +04/19 [23:17:57] INFO | >> train_qwenlatent.py:487 + Step 32800 | grad_norm_pre_clip=0.1499 | + grad_norm_pre_clip_avg=0.1473 | Metrics: + {'align_loss': 0.02476227656006813, + 'recon_loss': 0.09075777977705002, + 'predict_loss': 0.006260383874177933, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14986100792884827, + 'mae_score': 0.008478594255876971, 'data_time': + 0.0006710609886795282, 'model_time': + 1.2056864030018914, 'grad_norm_pre_clip_avg': + 0.14727921634912491, 'learning_rate': + 7.996713181099874e-06, 'epoch': 8.28} +04/19 [23:18:10] INFO | >> train_qwenlatent.py:487 + Step 32810 | grad_norm_pre_clip=0.1745 | + grad_norm_pre_clip_avg=0.1545 | Metrics: + {'align_loss': 0.024172209203243256, + 'recon_loss': 0.06493748724460602, + 'predict_loss': 0.006046385504305363, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1745329201221466, + 'data_time': 0.0009310989989899099, + 'model_time': 1.2786579750245437, + 'grad_norm_pre_clip_avg': 0.15447659492492677, + 'learning_rate': 7.988586358276986e-06, + 'epoch': 8.28} +04/19 [23:18:22] INFO | >> train_qwenlatent.py:487 + Step 32820 | grad_norm_pre_clip=0.1431 | + grad_norm_pre_clip_avg=0.1623 | Metrics: + {'align_loss': 0.023916136473417282, + 'recon_loss': 0.06830649077892303, + 'predict_loss': 0.005989653989672661, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.143090158700943, + 'data_time': 0.0007207689923234284, + 'model_time': 1.2128415660117753, + 'grad_norm_pre_clip_avg': 0.16228272318840026, + 'learning_rate': 7.98046174035465e-06, 'epoch': + 8.28} +04/19 [23:18:35] INFO | >> train_qwenlatent.py:487 + Step 32830 | grad_norm_pre_clip=0.1517 | + grad_norm_pre_clip_avg=0.1590 | Metrics: + {'align_loss': 0.02469455637037754, + 'recon_loss': 0.08912710845470428, + 'predict_loss': 0.006761273369193077, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15171262621879578, + 'data_time': 0.0007847369997762144, + 'model_time': 1.2528043609927408, + 'grad_norm_pre_clip_avg': 0.15899245515465737, + 'learning_rate': 7.972339331292709e-06, + 'epoch': 8.28} +04/19 [23:18:48] INFO | >> train_qwenlatent.py:487 + Step 32840 | grad_norm_pre_clip=0.2079 | + grad_norm_pre_clip_avg=0.1603 | Metrics: + {'align_loss': 0.025349367409944534, + 'recon_loss': 0.11777836084365845, + 'predict_loss': 0.008695819415152073, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20788328349590302, + 'data_time': 0.0009666149853728712, + 'model_time': 1.2494513969868422, + 'grad_norm_pre_clip_avg': 0.16030970290303231, + 'learning_rate': 7.964219135049921e-06, + 'epoch': 8.29} +04/19 [23:19:01] INFO | >> train_qwenlatent.py:487 + Step 32850 | grad_norm_pre_clip=0.1315 | + grad_norm_pre_clip_avg=0.1523 | Metrics: + {'align_loss': 0.02606232836842537, + 'recon_loss': 0.11279254406690598, + 'predict_loss': 0.012940949760377407, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13152168691158295, + 'mae_score': 0.007725664087243982, 'data_time': + 0.000966493011219427, 'model_time': + 1.2471326350059826, 'grad_norm_pre_clip_avg': + 0.15225398913025856, 'learning_rate': + 7.956101155583973e-06, 'epoch': 8.29} +04/19 [23:19:14] INFO | >> train_qwenlatent.py:487 + Step 32860 | grad_norm_pre_clip=0.1429 | + grad_norm_pre_clip_avg=0.1971 | Metrics: + {'align_loss': 0.025376232340931892, + 'recon_loss': 0.11595956981182098, + 'predict_loss': 0.008565036579966545, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1429193615913391, + 'data_time': 0.0010116820049006492, + 'model_time': 1.2626958450127859, + 'grad_norm_pre_clip_avg': 0.19706171452999116, + 'learning_rate': 7.947985396851476e-06, + 'epoch': 8.29} +04/19 [23:19:27] INFO | >> train_qwenlatent.py:487 + Step 32870 | grad_norm_pre_clip=0.1349 | + grad_norm_pre_clip_avg=0.1683 | Metrics: + {'align_loss': 0.02455246075987816, + 'recon_loss': 0.08781502395868301, + 'predict_loss': 0.009486173279583454, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1348721832036972, + 'data_time': 0.0009257329802494496, + 'model_time': 1.2660422490152996, + 'grad_norm_pre_clip_avg': 0.168264539539814, + 'learning_rate': 7.939871862807942e-06, + 'epoch': 8.29} +04/19 [23:19:39] INFO | >> train_qwenlatent.py:487 + Step 32880 | grad_norm_pre_clip=0.1573 | + grad_norm_pre_clip_avg=0.1607 | Metrics: + {'align_loss': 0.025733478367328644, + 'recon_loss': 0.10401181131601334, + 'predict_loss': 0.006982243154197931, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1573430746793747, + 'data_time': 0.0009032389789354056, + 'model_time': 1.2449379629979376, + 'grad_norm_pre_clip_avg': 0.16065386235713958, + 'learning_rate': 7.93176055740781e-06, 'epoch': + 8.3} +04/19 [23:19:52] INFO | >> train_qwenlatent.py:487 + Step 32890 | grad_norm_pre_clip=0.1553 | + grad_norm_pre_clip_avg=0.1666 | Metrics: + {'align_loss': 0.025234321132302284, + 'recon_loss': 0.08118996024131775, + 'predict_loss': 0.0057718511670827866, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15526585280895233, + 'data_time': 0.001164978981250897, + 'model_time': 1.2672186020063236, + 'grad_norm_pre_clip_avg': 0.16657568216323854, + 'learning_rate': 7.923651484604438e-06, + 'epoch': 8.3} +04/19 [23:20:05] INFO | >> train_qwenlatent.py:487 + Step 32900 | grad_norm_pre_clip=0.1591 | + grad_norm_pre_clip_avg=0.1620 | Metrics: + {'align_loss': 0.026242844760417938, + 'recon_loss': 0.11414097994565964, + 'predict_loss': 0.008878893218934536, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1591237634420395, + 'mae_score': 0.009253132450687993, 'data_time': + 0.0006623659864999354, 'model_time': + 1.2144214870058931, 'grad_norm_pre_clip_avg': + 0.16198806911706926, 'learning_rate': + 7.915544648350085e-06, 'epoch': 8.3} +04/19 [23:20:17] INFO | >> train_qwenlatent.py:487 + Step 32910 | grad_norm_pre_clip=0.1809 | + grad_norm_pre_clip_avg=0.1749 | Metrics: + {'align_loss': 0.026012452319264412, + 'recon_loss': 0.1275300681591034, + 'predict_loss': 0.014614557847380638, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18088875710964203, + 'data_time': 0.0006810560007579625, + 'model_time': 1.2468005609989632, + 'grad_norm_pre_clip_avg': 0.17494761496782302, + 'learning_rate': 7.907440052595927e-06, + 'epoch': 8.3} +04/19 [23:20:30] INFO | >> train_qwenlatent.py:487 + Step 32920 | grad_norm_pre_clip=0.1961 | + grad_norm_pre_clip_avg=0.1763 | Metrics: + {'align_loss': 0.025049738585948944, + 'recon_loss': 0.0908430963754654, + 'predict_loss': 0.01019058097153902, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1961040049791336, + 'data_time': 0.0009200880012940615, + 'model_time': 1.2623558249906637, + 'grad_norm_pre_clip_avg': 0.17631142139434813, + 'learning_rate': 7.899337701292044e-06, + 'epoch': 8.31} +04/19 [23:20:43] INFO | >> train_qwenlatent.py:487 + Step 32930 | grad_norm_pre_clip=0.1428 | + grad_norm_pre_clip_avg=0.1704 | Metrics: + {'align_loss': 0.024986745789647102, + 'recon_loss': 0.08858467638492584, + 'predict_loss': 0.005166997201740742, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14284121990203857, + 'data_time': 0.0009106250072363764, + 'model_time': 1.1807457440008875, + 'grad_norm_pre_clip_avg': 0.1703516125679016, + 'learning_rate': 7.891237598387423e-06, + 'epoch': 8.31} +04/19 [23:20:56] INFO | >> train_qwenlatent.py:487 + Step 32940 | grad_norm_pre_clip=0.1954 | + grad_norm_pre_clip_avg=0.1633 | Metrics: + {'align_loss': 0.025869540870189667, + 'recon_loss': 0.0870424136519432, + 'predict_loss': 0.008606831543147564, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19538235664367676, + 'data_time': 0.0008922230044845492, + 'model_time': 1.2061984589963686, + 'grad_norm_pre_clip_avg': 0.16328948810696603, + 'learning_rate': 7.883139747829959e-06, + 'epoch': 8.31} +04/19 [23:21:09] INFO | >> train_qwenlatent.py:487 + Step 32950 | grad_norm_pre_clip=0.1828 | + grad_norm_pre_clip_avg=0.1665 | Metrics: + {'align_loss': 0.024908848106861115, + 'recon_loss': 0.08252213895320892, + 'predict_loss': 0.012215483002364635, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18283236026763916, + 'mae_score': 0.010536312412571263, 'data_time': + 0.000623139989329502, 'model_time': + 1.2129261359805241, 'grad_norm_pre_clip_avg': + 0.16645233035087587, 'learning_rate': + 7.875044153566447e-06, 'epoch': 8.31} +04/19 [23:21:21] INFO | >> train_qwenlatent.py:487 + Step 32960 | grad_norm_pre_clip=0.2325 | + grad_norm_pre_clip_avg=0.1657 | Metrics: + {'align_loss': 0.025157533586025238, + 'recon_loss': 0.11431112140417099, + 'predict_loss': 0.005368282552808523, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2325451821088791, + 'data_time': 0.0008444949926342815, + 'model_time': 1.2954537270125002, + 'grad_norm_pre_clip_avg': 0.16565135568380357, + 'learning_rate': 7.866950819542574e-06, + 'epoch': 8.32} +04/19 [23:21:35] INFO | >> train_qwenlatent.py:487 + Step 32970 | grad_norm_pre_clip=0.1840 | + grad_norm_pre_clip_avg=0.1796 | Metrics: + {'align_loss': 0.025692839175462723, + 'recon_loss': 0.1037493348121643, + 'predict_loss': 0.006148473359644413, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1839693784713745, + 'data_time': 0.0006832580256741494, + 'model_time': 1.2380419189867098, + 'grad_norm_pre_clip_avg': 0.17964957058429717, + 'learning_rate': 7.858859749702934e-06, + 'epoch': 8.32} +04/19 [23:21:47] INFO | >> train_qwenlatent.py:487 + Step 32980 | grad_norm_pre_clip=0.1928 | + grad_norm_pre_clip_avg=0.1793 | Metrics: + {'align_loss': 0.02452121675014496, + 'recon_loss': 0.10591523349285126, + 'predict_loss': 0.005195480305701494, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19282734394073486, + 'data_time': 0.0008150289941113442, + 'model_time': 1.2485656329954509, + 'grad_norm_pre_clip_avg': 0.1792750746011734, + 'learning_rate': 7.850770947991024e-06, + 'epoch': 8.32} +04/19 [23:22:00] INFO | >> train_qwenlatent.py:487 + Step 32990 | grad_norm_pre_clip=0.2104 | + grad_norm_pre_clip_avg=0.1891 | Metrics: + {'align_loss': 0.02450479194521904, + 'recon_loss': 0.10039636492729187, + 'predict_loss': 0.007313366048038006, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21044641733169556, + 'data_time': 0.001146030001109466, + 'model_time': 1.2476861430041026, + 'grad_norm_pre_clip_avg': 0.189053076505661, + 'learning_rate': 7.842684418349218e-06, + 'epoch': 8.32} +04/19 [23:22:12] INFO | >> train_qwenlatent.py:487 + Step 33000 | grad_norm_pre_clip=0.2007 | + grad_norm_pre_clip_avg=0.1958 | Metrics: + {'align_loss': 0.02442181296646595, + 'recon_loss': 0.10522351413965225, + 'predict_loss': 0.007738254498690367, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20074354112148285, + 'mae_score': 0.0070403687588803405, + 'data_time': 0.0009322020050603896, + 'model_time': 1.2312723290233407, + 'grad_norm_pre_clip_avg': 0.19582530558109285, + 'learning_rate': 7.834600164718799e-06, + 'epoch': 8.33} +04/19 [23:22:25] INFO | >> train_qwenlatent.py:487 + Step 33010 | grad_norm_pre_clip=0.2024 | + grad_norm_pre_clip_avg=0.1684 | Metrics: + {'align_loss': 0.024931497871875763, + 'recon_loss': 0.05969050154089928, + 'predict_loss': 0.003344970755279064, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2024267613887787, + 'data_time': 0.000666637992253527, + 'model_time': 1.1901970549952239, + 'grad_norm_pre_clip_avg': 0.16844595968723297, + 'learning_rate': 7.826518191039932e-06, + 'epoch': 8.33} +04/19 [23:22:38] INFO | >> train_qwenlatent.py:487 + Step 33020 | grad_norm_pre_clip=0.2035 | + grad_norm_pre_clip_avg=0.1679 | Metrics: + {'align_loss': 0.024604998528957367, + 'recon_loss': 0.10275261849164963, + 'predict_loss': 0.015180550515651703, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20349174737930298, + 'data_time': 0.0008010840101633221, + 'model_time': 1.234137098013889, + 'grad_norm_pre_clip_avg': 0.16786761283874513, + 'learning_rate': 7.818438501251674e-06, + 'epoch': 8.33} +04/19 [23:22:50] INFO | >> train_qwenlatent.py:487 + Step 33030 | grad_norm_pre_clip=0.1874 | + grad_norm_pre_clip_avg=0.1739 | Metrics: + {'align_loss': 0.025688303634524345, + 'recon_loss': 0.1244359239935875, + 'predict_loss': 0.007658955175429583, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18738698959350586, + 'data_time': 0.000681874982547015, + 'model_time': 1.1981799690111075, + 'grad_norm_pre_clip_avg': 0.1739230453968048, + 'learning_rate': 7.810361099291964e-06, + 'epoch': 8.33} +04/19 [23:23:02] INFO | >> train_qwenlatent.py:487 + Step 33040 | grad_norm_pre_clip=0.1856 | + grad_norm_pre_clip_avg=0.1819 | Metrics: + {'align_loss': 0.025498487055301666, + 'recon_loss': 0.15538327395915985, + 'predict_loss': 0.017101654782891273, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1855900138616562, + 'data_time': 0.0008703290077392012, + 'model_time': 1.214193610008806, + 'grad_norm_pre_clip_avg': 0.18190842270851135, + 'learning_rate': 7.802285989097637e-06, + 'epoch': 8.34} +04/19 [23:23:15] INFO | >> train_qwenlatent.py:487 + Step 33050 | grad_norm_pre_clip=0.1797 | + grad_norm_pre_clip_avg=0.1927 | Metrics: + {'align_loss': 0.025370873510837555, + 'recon_loss': 0.10781215131282806, + 'predict_loss': 0.012148349545896053, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17968755960464478, + 'mae_score': 0.008170114121995531, 'data_time': + 0.001187469984870404, 'model_time': + 1.204890063003404, 'grad_norm_pre_clip_avg': + 0.19271499365568162, 'learning_rate': + 7.794213174604393e-06, 'epoch': 8.34} +04/19 [23:23:28] INFO | >> train_qwenlatent.py:487 + Step 33060 | grad_norm_pre_clip=0.1715 | + grad_norm_pre_clip_avg=0.2027 | Metrics: + {'align_loss': 0.023706119507551193, + 'recon_loss': 0.10196870565414429, + 'predict_loss': 0.007261412218213081, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17153514921665192, + 'data_time': 0.0008468599990010262, + 'model_time': 1.2058773580065463, + 'grad_norm_pre_clip_avg': 0.2026633083820343, + 'learning_rate': 7.786142659746834e-06, + 'epoch': 8.34} +04/19 [23:23:40] INFO | >> train_qwenlatent.py:487 + Step 33070 | grad_norm_pre_clip=0.2158 | + grad_norm_pre_clip_avg=0.1673 | Metrics: + {'align_loss': 0.025775324553251266, + 'recon_loss': 0.14941546320915222, + 'predict_loss': 0.013728702440857887, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21579407155513763, + 'data_time': 0.0009059119911398739, + 'model_time': 1.2392488199984655, + 'grad_norm_pre_clip_avg': 0.16732982397079468, + 'learning_rate': 7.77807444845842e-06, 'epoch': + 8.34} +04/19 [23:23:53] INFO | >> train_qwenlatent.py:487 + Step 33080 | grad_norm_pre_clip=0.1701 | + grad_norm_pre_clip_avg=0.1639 | Metrics: + {'align_loss': 0.025235824286937714, + 'recon_loss': 0.09377191215753555, + 'predict_loss': 0.008987654000520706, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17007699608802795, + 'data_time': 0.0006792480126023293, + 'model_time': 1.428845419024583, + 'grad_norm_pre_clip_avg': 0.16387509554624557, + 'learning_rate': 7.77000854467151e-06, 'epoch': + 8.35} +04/19 [23:24:05] INFO | >> train_qwenlatent.py:487 + Step 33090 | grad_norm_pre_clip=0.1506 | + grad_norm_pre_clip_avg=0.1751 | Metrics: + {'align_loss': 0.025624793022871017, + 'recon_loss': 0.10420738905668259, + 'predict_loss': 0.008585392497479916, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1505890190601349, + 'data_time': 0.000993813999230042, + 'model_time': 1.2034559820021968, + 'grad_norm_pre_clip_avg': 0.1750913828611374, + 'learning_rate': 7.761944952317325e-06, + 'epoch': 8.35} +04/19 [23:24:19] INFO | >> train_qwenlatent.py:487 + Step 33100 | grad_norm_pre_clip=0.2259 | + grad_norm_pre_clip_avg=0.1880 | Metrics: + {'align_loss': 0.02524973824620247, + 'recon_loss': 0.095533087849617, + 'predict_loss': 0.01428243238478899, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22588805854320526, + 'mae_score': 0.00649241456040391, 'data_time': + 0.0010648540046531707, 'model_time': + 1.234670341014862, 'grad_norm_pre_clip_avg': + 0.18803958222270012, 'learning_rate': + 7.753883675325952e-06, 'epoch': 8.35} +04/19 [23:24:32] INFO | >> train_qwenlatent.py:487 + Step 33110 | grad_norm_pre_clip=0.2236 | + grad_norm_pre_clip_avg=0.1744 | Metrics: + {'align_loss': 0.022730331867933273, + 'recon_loss': 0.0900760143995285, + 'predict_loss': 0.010859510861337185, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22360585629940033, + 'data_time': 0.0008868869917932898, + 'model_time': 1.2324275939899962, + 'grad_norm_pre_clip_avg': 0.17444281429052352, + 'learning_rate': 7.745824717626374e-06, + 'epoch': 8.35} +04/19 [23:24:44] INFO | >> train_qwenlatent.py:487 + Step 33120 | grad_norm_pre_clip=0.1514 | + grad_norm_pre_clip_avg=0.1800 | Metrics: + {'align_loss': 0.023965060710906982, + 'recon_loss': 0.09384756535291672, + 'predict_loss': 0.0130800511687994, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1513926237821579, + 'data_time': 0.0007443020003847778, + 'model_time': 1.2447199940215796, + 'grad_norm_pre_clip_avg': 0.18001648932695388, + 'learning_rate': 7.737768083146417e-06, + 'epoch': 8.36} +04/19 [23:24:57] INFO | >> train_qwenlatent.py:487 + Step 33130 | grad_norm_pre_clip=0.1491 | + grad_norm_pre_clip_avg=0.1754 | Metrics: + {'align_loss': 0.024995986372232437, + 'recon_loss': 0.0985938310623169, + 'predict_loss': 0.00817936472594738, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14908631145954132, + 'data_time': 0.0006544259958900511, + 'model_time': 1.2238198360137176, + 'grad_norm_pre_clip_avg': 0.17535240650177003, + 'learning_rate': 7.729713775812801e-06, + 'epoch': 8.36} +04/19 [23:25:10] INFO | >> train_qwenlatent.py:487 + Step 33140 | grad_norm_pre_clip=0.1260 | + grad_norm_pre_clip_avg=0.1527 | Metrics: + {'align_loss': 0.0266867745667696, + 'recon_loss': 0.14294950664043427, + 'predict_loss': 0.013573823496699333, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12602843344211578, + 'data_time': 0.0006869540084153414, + 'model_time': 1.2518955809937324, + 'grad_norm_pre_clip_avg': 0.15266598761081696, + 'learning_rate': 7.72166179955108e-06, 'epoch': + 8.36} +04/19 [23:25:23] INFO | >> train_qwenlatent.py:487 + Step 33150 | grad_norm_pre_clip=0.1415 | + grad_norm_pre_clip_avg=0.1525 | Metrics: + {'align_loss': 0.024886874482035637, + 'recon_loss': 0.1086665540933609, + 'predict_loss': 0.005836946424096823, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14150595664978027, + 'mae_score': 0.007193225998062271, 'data_time': + 0.0006705520208925009, 'model_time': + 1.214616038982058, 'grad_norm_pre_clip_avg': + 0.15246018767356873, 'learning_rate': + 7.713612158285704e-06, 'epoch': 8.36} +04/19 [23:25:35] INFO | >> train_qwenlatent.py:487 + Step 33160 | grad_norm_pre_clip=0.1929 | + grad_norm_pre_clip_avg=0.1787 | Metrics: + {'align_loss': 0.02494204044342041, + 'recon_loss': 0.10638468712568283, + 'predict_loss': 0.008408223278820515, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19289720058441162, + 'data_time': 0.0011893879855051637, + 'model_time': 1.245711957977619, + 'grad_norm_pre_clip_avg': 0.17873666435480118, + 'learning_rate': 7.705564855939959e-06, + 'epoch': 8.37} +04/19 [23:25:48] INFO | >> train_qwenlatent.py:487 + Step 33170 | grad_norm_pre_clip=0.1779 | + grad_norm_pre_clip_avg=0.1793 | Metrics: + {'align_loss': 0.026206154376268387, + 'recon_loss': 0.11571970582008362, + 'predict_loss': 0.004299391061067581, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17788176238536835, + 'data_time': 0.0010187639854848385, + 'model_time': 1.2185585960105527, + 'grad_norm_pre_clip_avg': 0.17930321544408798, + 'learning_rate': 7.697519896436012e-06, + 'epoch': 8.37} +04/19 [23:26:00] INFO | >> train_qwenlatent.py:487 + Step 33180 | grad_norm_pre_clip=0.1923 | + grad_norm_pre_clip_avg=0.1789 | Metrics: + {'align_loss': 0.026115842163562775, + 'recon_loss': 0.11978720128536224, + 'predict_loss': 0.016562674194574356, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19229134917259216, + 'data_time': 0.0008237890142481774, + 'model_time': 1.2182524589879904, + 'grad_norm_pre_clip_avg': 0.17890093773603438, + 'learning_rate': 7.689477283694872e-06, + 'epoch': 8.37} +04/19 [23:26:12] INFO | >> train_qwenlatent.py:487 + Step 33190 | grad_norm_pre_clip=0.1211 | + grad_norm_pre_clip_avg=0.1567 | Metrics: + {'align_loss': 0.024594049900770187, + 'recon_loss': 0.09401651471853256, + 'predict_loss': 0.0072696637362241745, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12108354270458221, + 'data_time': 0.0009075559792108834, + 'model_time': 1.2385733039991464, + 'grad_norm_pre_clip_avg': 0.1567113608121872, + 'learning_rate': 7.681437021636415e-06, + 'epoch': 8.37} +04/19 [23:26:25] INFO | >> train_qwenlatent.py:487 + Step 33200 | grad_norm_pre_clip=0.1461 | + grad_norm_pre_clip_avg=0.1739 | Metrics: + {'align_loss': 0.026265595108270645, + 'recon_loss': 0.10403665155172348, + 'predict_loss': 0.004222448915243149, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14612752199172974, + 'mae_score': 0.00864898664457304, 'data_time': + 0.0009472290112171322, 'model_time': + 1.2358479329850525, 'grad_norm_pre_clip_avg': + 0.17392702102661134, 'learning_rate': + 7.673399114179364e-06, 'epoch': 8.38} +04/19 [23:26:39] INFO | >> train_qwenlatent.py:487 + Step 33210 | grad_norm_pre_clip=0.2053 | + grad_norm_pre_clip_avg=0.1760 | Metrics: + {'align_loss': 0.025430060923099518, + 'recon_loss': 0.09623202681541443, + 'predict_loss': 0.008766141720116138, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20530903339385986, + 'data_time': 0.0006875690014567226, + 'model_time': 1.1899415849766228, + 'grad_norm_pre_clip_avg': 0.17599013894796373, + 'learning_rate': 7.6653635652413e-06, 'epoch': + 8.38} +04/19 [23:26:51] INFO | >> train_qwenlatent.py:487 + Step 33220 | grad_norm_pre_clip=0.1461 | + grad_norm_pre_clip_avg=0.1499 | Metrics: + {'align_loss': 0.025761358439922333, + 'recon_loss': 0.12758110463619232, + 'predict_loss': 0.010324046947062016, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1460958868265152, + 'data_time': 0.0011205759947188199, + 'model_time': 1.2298943550267722, + 'grad_norm_pre_clip_avg': 0.14988378137350084, + 'learning_rate': 7.657330378738654e-06, + 'epoch': 8.38} +04/19 [23:27:04] INFO | >> train_qwenlatent.py:487 + Step 33230 | grad_norm_pre_clip=0.1570 | + grad_norm_pre_clip_avg=0.1553 | Metrics: + {'align_loss': 0.02681640163064003, + 'recon_loss': 0.1429440975189209, + 'predict_loss': 0.007635016925632954, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15699201822280884, + 'data_time': 0.0006739579839631915, + 'model_time': 1.2304546989907976, + 'grad_norm_pre_clip_avg': 0.1553027242422104, + 'learning_rate': 7.649299558586694e-06, + 'epoch': 8.39} +04/19 [23:27:17] INFO | >> train_qwenlatent.py:487 + Step 33240 | grad_norm_pre_clip=0.1806 | + grad_norm_pre_clip_avg=0.1634 | Metrics: + {'align_loss': 0.02628796175122261, + 'recon_loss': 0.08235450834035873, + 'predict_loss': 0.004092587623745203, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18060670793056488, + 'data_time': 0.0011829340073745698, + 'model_time': 1.2892725080018863, + 'grad_norm_pre_clip_avg': 0.1634358733892441, + 'learning_rate': 7.641271108699552e-06, + 'epoch': 8.39} +04/19 [23:27:30] INFO | >> train_qwenlatent.py:487 + Step 33250 | grad_norm_pre_clip=0.1448 | + grad_norm_pre_clip_avg=0.1553 | Metrics: + {'align_loss': 0.026508692651987076, + 'recon_loss': 0.13984757661819458, + 'predict_loss': 0.011301897466182709, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14482204616069794, + 'mae_score': 0.007395782986202756, 'data_time': + 0.0010961959778796881, 'model_time': + 1.2187072829983663, 'grad_norm_pre_clip_avg': + 0.15531793534755706, 'learning_rate': + 7.633245032990197e-06, 'epoch': 8.39} +04/19 [23:27:43] INFO | >> train_qwenlatent.py:487 + Step 33260 | grad_norm_pre_clip=0.1615 | + grad_norm_pre_clip_avg=0.1623 | Metrics: + {'align_loss': 0.024883301928639412, + 'recon_loss': 0.08467338234186172, + 'predict_loss': 0.007026636973023415, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16152776777744293, + 'data_time': 0.0007189390016719699, + 'model_time': 1.2229406129918061, + 'grad_norm_pre_clip_avg': 0.1622873067855835, + 'learning_rate': 7.625221335370435e-06, + 'epoch': 8.39} +04/19 [23:27:56] INFO | >> train_qwenlatent.py:487 + Step 33270 | grad_norm_pre_clip=0.1885 | + grad_norm_pre_clip_avg=0.1773 | Metrics: + {'align_loss': 0.026500016450881958, + 'recon_loss': 0.15135596692562103, + 'predict_loss': 0.008397108875215054, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18847373127937317, + 'data_time': 0.0009554290154483169, + 'model_time': 1.2892964289931115, + 'grad_norm_pre_clip_avg': 0.17727955877780915, + 'learning_rate': 7.617200019750927e-06, + 'epoch': 8.4} +04/19 [23:28:09] INFO | >> train_qwenlatent.py:487 + Step 33280 | grad_norm_pre_clip=0.2266 | + grad_norm_pre_clip_avg=0.1734 | Metrics: + {'align_loss': 0.023862354457378387, + 'recon_loss': 0.1493656188249588, + 'predict_loss': 0.011924145743250847, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22655953466892242, + 'data_time': 0.0009207710099872202, + 'model_time': 1.243047021998791, + 'grad_norm_pre_clip_avg': 0.17338777408003808, + 'learning_rate': 7.60918109004116e-06, 'epoch': + 8.4} +04/19 [23:28:21] INFO | >> train_qwenlatent.py:487 + Step 33290 | grad_norm_pre_clip=0.1281 | + grad_norm_pre_clip_avg=0.1821 | Metrics: + {'align_loss': 0.025370607152581215, + 'recon_loss': 0.09262651205062866, + 'predict_loss': 0.006138582713901997, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12809854745864868, + 'data_time': 0.0006793399807065725, + 'model_time': 1.2258637099876069, + 'grad_norm_pre_clip_avg': 0.18207560032606124, + 'learning_rate': 7.601164550149459e-06, + 'epoch': 8.4} +04/19 [23:28:34] INFO | >> train_qwenlatent.py:487 + Step 33300 | grad_norm_pre_clip=0.1382 | + grad_norm_pre_clip_avg=0.1686 | Metrics: + {'align_loss': 0.0248151496052742, + 'recon_loss': 0.10469751805067062, + 'predict_loss': 0.005907620303332806, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1382177174091339, + 'mae_score': 0.00938626968108856, 'data_time': + 0.0008611940138507634, 'model_time': + 1.2054616469831672, 'grad_norm_pre_clip_avg': + 0.16864781379699706, 'learning_rate': + 7.593150403982999e-06, 'epoch': 8.4} +04/19 [23:28:46] INFO | >> train_qwenlatent.py:487 + Step 33310 | grad_norm_pre_clip=0.1254 | + grad_norm_pre_clip_avg=0.1662 | Metrics: + {'align_loss': 0.02528654783964157, + 'recon_loss': 0.10780172795057297, + 'predict_loss': 0.00809171050786972, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12535129487514496, + 'data_time': 0.0010975509940180928, + 'model_time': 1.1913133720227052, + 'grad_norm_pre_clip_avg': 0.16615068092942237, + 'learning_rate': 7.585138655447769e-06, + 'epoch': 8.41} +04/19 [23:28:59] INFO | >> train_qwenlatent.py:487 + Step 33320 | grad_norm_pre_clip=0.2308 | + grad_norm_pre_clip_avg=0.1682 | Metrics: + {'align_loss': 0.02498180791735649, + 'recon_loss': 0.09877308458089828, + 'predict_loss': 0.011574155651032925, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23080570995807648, + 'data_time': 0.0009742350084707141, + 'model_time': 1.2178429719933774, + 'grad_norm_pre_clip_avg': 0.1682088166475296, + 'learning_rate': 7.577129308448601e-06, + 'epoch': 8.41} +04/19 [23:29:11] INFO | >> train_qwenlatent.py:487 + Step 33330 | grad_norm_pre_clip=0.1289 | + grad_norm_pre_clip_avg=0.1743 | Metrics: + {'align_loss': 0.02614995650947094, + 'recon_loss': 0.1284036785364151, + 'predict_loss': 0.006454407703131437, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12891077995300293, + 'data_time': 0.0010000040056183934, + 'model_time': 1.246805646980647, + 'grad_norm_pre_clip_avg': 0.17430546879768372, + 'learning_rate': 7.569122366889149e-06, + 'epoch': 8.41} +04/19 [23:29:24] INFO | >> train_qwenlatent.py:487 + Step 33340 | grad_norm_pre_clip=0.1912 | + grad_norm_pre_clip_avg=0.1569 | Metrics: + {'align_loss': 0.025168200954794884, + 'recon_loss': 0.10182171314954758, + 'predict_loss': 0.004971248097717762, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19121615588665009, + 'data_time': 0.0006679170182906091, + 'model_time': 1.2271541680092923, + 'grad_norm_pre_clip_avg': 0.15692916437983512, + 'learning_rate': 7.561117834671906e-06, + 'epoch': 8.41} +04/19 [23:29:38] INFO | >> train_qwenlatent.py:487 + Step 33350 | grad_norm_pre_clip=0.1805 | + grad_norm_pre_clip_avg=0.1835 | Metrics: + {'align_loss': 0.026191337034106255, + 'recon_loss': 0.13891050219535828, + 'predict_loss': 0.010696904733777046, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18048089742660522, + 'mae_score': 0.006868295411805849, 'data_time': + 0.0007595510105602443, 'model_time': + 1.2179205069842283, 'grad_norm_pre_clip_avg': + 0.1834951415657997, 'learning_rate': + 7.55311571569818e-06, 'epoch': 8.42} +04/19 [23:29:50] INFO | >> train_qwenlatent.py:487 + Step 33360 | grad_norm_pre_clip=0.1646 | + grad_norm_pre_clip_avg=0.1636 | Metrics: + {'align_loss': 0.023669395595788956, + 'recon_loss': 0.07355666905641556, + 'predict_loss': 0.004420398268848658, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16457132995128632, + 'data_time': 0.0009482399909757078, + 'model_time': 1.2365160949993879, + 'grad_norm_pre_clip_avg': 0.1635527268052101, + 'learning_rate': 7.545116013868103e-06, + 'epoch': 8.42} +04/19 [23:30:03] INFO | >> train_qwenlatent.py:487 + Step 33370 | grad_norm_pre_clip=0.1432 | + grad_norm_pre_clip_avg=0.1709 | Metrics: + {'align_loss': 0.025610238313674927, + 'recon_loss': 0.10270974040031433, + 'predict_loss': 0.0076687755063176155, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14320503175258636, + 'data_time': 0.0011912990012206137, + 'model_time': 1.2708352439804003, + 'grad_norm_pre_clip_avg': 0.1708913490176201, + 'learning_rate': 7.53711873308064e-06, 'epoch': + 8.42} +04/19 [23:30:16] INFO | >> train_qwenlatent.py:487 + Step 33380 | grad_norm_pre_clip=0.1600 | + grad_norm_pre_clip_avg=0.1675 | Metrics: + {'align_loss': 0.023401491343975067, + 'recon_loss': 0.08776931464672089, + 'predict_loss': 0.006896109785884619, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1599775105714798, + 'data_time': 0.0008264869975391775, + 'model_time': 1.268458128994098, + 'grad_norm_pre_clip_avg': 0.16749625504016877, + 'learning_rate': 7.529123877233562e-06, + 'epoch': 8.42} +04/19 [23:30:28] INFO | >> train_qwenlatent.py:487 + Step 33390 | grad_norm_pre_clip=0.1721 | + grad_norm_pre_clip_avg=0.1886 | Metrics: + {'align_loss': 0.025227293372154236, + 'recon_loss': 0.13540828227996826, + 'predict_loss': 0.012989572249352932, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17210739850997925, + 'data_time': 0.0006423589948099107, + 'model_time': 1.2483256149862427, + 'grad_norm_pre_clip_avg': 0.18855320066213607, + 'learning_rate': 7.521131450223472e-06, + 'epoch': 8.43} +04/19 [23:30:42] INFO | >> train_qwenlatent.py:487 + Step 33400 | grad_norm_pre_clip=0.1755 | + grad_norm_pre_clip_avg=0.1732 | Metrics: + {'align_loss': 0.02364862710237503, + 'recon_loss': 0.09630459547042847, + 'predict_loss': 0.00411608163267374, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17554694414138794, + 'mae_score': 0.00858078518429318, 'data_time': + 0.001077772001735866, 'model_time': + 1.2639428029942792, 'grad_norm_pre_clip_avg': + 0.17324118316173553, 'learning_rate': + 7.513141455945777e-06, 'epoch': 8.43} +04/19 [23:30:55] INFO | >> train_qwenlatent.py:487 + Step 33410 | grad_norm_pre_clip=0.1641 | + grad_norm_pre_clip_avg=0.1652 | Metrics: + {'align_loss': 0.024956557899713516, + 'recon_loss': 0.07082980126142502, + 'predict_loss': 0.006317014340311289, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16408909857273102, + 'data_time': 0.0008891639881767333, + 'model_time': 1.2258952989941463, + 'grad_norm_pre_clip_avg': 0.1652002826333046, + 'learning_rate': 7.505153898294702e-06, + 'epoch': 8.43} +04/19 [23:31:07] INFO | >> train_qwenlatent.py:487 + Step 33420 | grad_norm_pre_clip=0.1993 | + grad_norm_pre_clip_avg=0.1681 | Metrics: + {'align_loss': 0.025362126529216766, + 'recon_loss': 0.11678357422351837, + 'predict_loss': 0.009939138777554035, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19925732910633087, + 'data_time': 0.0011087410093750805, + 'model_time': 1.2509296289936174, + 'grad_norm_pre_clip_avg': 0.1680944964289665, + 'learning_rate': 7.497168781163284e-06, + 'epoch': 8.43} +04/19 [23:31:20] INFO | >> train_qwenlatent.py:487 + Step 33430 | grad_norm_pre_clip=0.1352 | + grad_norm_pre_clip_avg=0.1646 | Metrics: + {'align_loss': 0.0255149994045496, + 'recon_loss': 0.08323272317647934, + 'predict_loss': 0.00878235325217247, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1352173238992691, + 'data_time': 0.0007988260185811669, + 'model_time': 1.2103741470200475, + 'grad_norm_pre_clip_avg': 0.16455874741077423, + 'learning_rate': 7.489186108443379e-06, + 'epoch': 8.44} +04/19 [23:31:32] INFO | >> train_qwenlatent.py:487 + Step 33440 | grad_norm_pre_clip=0.1608 | + grad_norm_pre_clip_avg=0.1658 | Metrics: + {'align_loss': 0.02605251595377922, + 'recon_loss': 0.10416198521852493, + 'predict_loss': 0.006634109653532505, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16084599494934082, + 'data_time': 0.0009339659882243723, + 'model_time': 1.287077696993947, + 'grad_norm_pre_clip_avg': 0.16577111631631852, + 'learning_rate': 7.481205884025638e-06, + 'epoch': 8.44} +04/19 [23:31:45] INFO | >> train_qwenlatent.py:487 + Step 33450 | grad_norm_pre_clip=0.1901 | + grad_norm_pre_clip_avg=0.1912 | Metrics: + {'align_loss': 0.0255875401198864, + 'recon_loss': 0.10967184603214264, + 'predict_loss': 0.0068946280516684055, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19009660184383392, + 'mae_score': 0.006678123731870909, 'data_time': + 0.0007503439846914262, 'model_time': + 1.290960617014207, 'grad_norm_pre_clip_avg': + 0.19115764647722244, 'learning_rate': + 7.473228111799528e-06, 'epoch': 8.44} +04/19 [23:31:58] INFO | >> train_qwenlatent.py:487 + Step 33460 | grad_norm_pre_clip=0.1432 | + grad_norm_pre_clip_avg=0.1742 | Metrics: + {'align_loss': 0.025732792913913727, + 'recon_loss': 0.10616614669561386, + 'predict_loss': 0.00843975692987442, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1432320475578308, + 'data_time': 0.001114638987928629, + 'model_time': 1.215791853988776, + 'grad_norm_pre_clip_avg': 0.17424651235342026, + 'learning_rate': 7.465252795653321e-06, + 'epoch': 8.44} +04/19 [23:32:11] INFO | >> train_qwenlatent.py:487 + Step 33470 | grad_norm_pre_clip=0.2082 | + grad_norm_pre_clip_avg=0.1574 | Metrics: + {'align_loss': 0.02548719197511673, + 'recon_loss': 0.09783101081848145, + 'predict_loss': 0.005995411425828934, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2081712782382965, + 'data_time': 0.000702471996191889, + 'model_time': 1.2193356999778189, + 'grad_norm_pre_clip_avg': 0.15736341923475267, + 'learning_rate': 7.457279939474088e-06, + 'epoch': 8.45} +04/19 [23:32:23] INFO | >> train_qwenlatent.py:487 + Step 33480 | grad_norm_pre_clip=0.1466 | + grad_norm_pre_clip_avg=0.1575 | Metrics: + {'align_loss': 0.025231819599866867, + 'recon_loss': 0.09513867646455765, + 'predict_loss': 0.005681012291461229, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1466221660375595, + 'data_time': 0.0007382649928331375, + 'model_time': 1.5024092179955915, + 'grad_norm_pre_clip_avg': 0.15752211809158326, + 'learning_rate': 7.449309547147702e-06, + 'epoch': 8.45} +04/19 [23:32:36] INFO | >> train_qwenlatent.py:487 + Step 33490 | grad_norm_pre_clip=0.1525 | + grad_norm_pre_clip_avg=0.1582 | Metrics: + {'align_loss': 0.025493109598755836, + 'recon_loss': 0.1278737634420395, + 'predict_loss': 0.008340074680745602, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1524546891450882, + 'data_time': 0.0009250080038327724, + 'model_time': 1.2215533649723511, + 'grad_norm_pre_clip_avg': 0.15819496065378189, + 'learning_rate': 7.44134162255883e-06, 'epoch': + 8.45} +04/19 [23:32:49] INFO | >> train_qwenlatent.py:487 + Step 33500 | grad_norm_pre_clip=0.2152 | + grad_norm_pre_clip_avg=0.1682 | Metrics: + {'align_loss': 0.025888731703162193, + 'recon_loss': 0.12053774297237396, + 'predict_loss': 0.0068078781478106976, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21516163647174835, + 'mae_score': 0.008412392934163411, 'data_time': + 0.0007455449958797544, 'model_time': + 1.1835196220199578, 'grad_norm_pre_clip_avg': + 0.16820441782474518, 'learning_rate': + 7.433376169590947e-06, 'epoch': 8.45} +04/19 [23:33:02] INFO | >> train_qwenlatent.py:487 + Step 33510 | grad_norm_pre_clip=0.1419 | + grad_norm_pre_clip_avg=0.1673 | Metrics: + {'align_loss': 0.024559419602155685, + 'recon_loss': 0.10676517337560654, + 'predict_loss': 0.008492660708725452, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14188799262046814, + 'data_time': 0.0011832349991891533, + 'model_time': 1.2875116349896416, + 'grad_norm_pre_clip_avg': 0.1672913685441017, + 'learning_rate': 7.4254131921263185e-06, + 'epoch': 8.46} +04/19 [23:33:15] INFO | >> train_qwenlatent.py:487 + Step 33520 | grad_norm_pre_clip=0.2240 | + grad_norm_pre_clip_avg=0.1563 | Metrics: + {'align_loss': 0.025934191420674324, + 'recon_loss': 0.1320551186800003, + 'predict_loss': 0.006779431365430355, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22403304278850555, + 'data_time': 0.0009367350139655173, + 'model_time': 1.1909431079984643, + 'grad_norm_pre_clip_avg': 0.15629571974277495, + 'learning_rate': 7.417452694045998e-06, + 'epoch': 8.46} +04/19 [23:33:27] INFO | >> train_qwenlatent.py:487 + Step 33530 | grad_norm_pre_clip=0.1955 | + grad_norm_pre_clip_avg=0.2035 | Metrics: + {'align_loss': 0.02530374564230442, + 'recon_loss': 0.09867725521326065, + 'predict_loss': 0.009488707408308983, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19552062451839447, + 'data_time': 0.0009411720093339682, + 'model_time': 1.222328732983442, + 'grad_norm_pre_clip_avg': 0.20353830903768538, + 'learning_rate': 7.409494679229843e-06, + 'epoch': 8.46} +04/19 [23:33:40] INFO | >> train_qwenlatent.py:487 + Step 33540 | grad_norm_pre_clip=0.1599 | + grad_norm_pre_clip_avg=0.1627 | Metrics: + {'align_loss': 0.02463892102241516, + 'recon_loss': 0.06436984986066818, + 'predict_loss': 0.0070246560499072075, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1598602533340454, + 'data_time': 0.0009572190174367279, + 'model_time': 1.270011174987303, + 'grad_norm_pre_clip_avg': 0.16273228228092193, + 'learning_rate': 7.4015391515564875e-06, + 'epoch': 8.46} +04/19 [23:33:53] INFO | >> train_qwenlatent.py:487 + Step 33550 | grad_norm_pre_clip=0.1059 | + grad_norm_pre_clip_avg=0.1587 | Metrics: + {'align_loss': 0.023746859282255173, + 'recon_loss': 0.07380993664264679, + 'predict_loss': 0.004632377065718174, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10593099892139435, + 'mae_score': 0.009771574724901904, 'data_time': + 0.0011378430062904954, 'model_time': + 1.2606590270006564, 'grad_norm_pre_clip_avg': + 0.15869104117155075, 'learning_rate': + 7.39358611490336e-06, 'epoch': 8.47} +04/19 [23:34:06] INFO | >> train_qwenlatent.py:487 + Step 33560 | grad_norm_pre_clip=0.1809 | + grad_norm_pre_clip_avg=0.1691 | Metrics: + {'align_loss': 0.024996351450681686, + 'recon_loss': 0.08993362635374069, + 'predict_loss': 0.006214158143848181, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18087705969810486, + 'data_time': 0.0006328879971988499, + 'model_time': 1.2456277960154694, + 'grad_norm_pre_clip_avg': 0.16907934546470643, + 'learning_rate': 7.3856355731466776e-06, + 'epoch': 8.47} +04/19 [23:34:18] INFO | >> train_qwenlatent.py:487 + Step 33570 | grad_norm_pre_clip=0.1693 | + grad_norm_pre_clip_avg=0.1718 | Metrics: + {'align_loss': 0.025169316679239273, + 'recon_loss': 0.10750836879014969, + 'predict_loss': 0.014525596052408218, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1692701131105423, + 'data_time': 0.0010172289912588894, + 'model_time': 1.2183748540119268, + 'grad_norm_pre_clip_avg': 0.17180487364530564, + 'learning_rate': 7.377687530161435e-06, + 'epoch': 8.47} +04/19 [23:34:31] INFO | >> train_qwenlatent.py:487 + Step 33580 | grad_norm_pre_clip=0.1890 | + grad_norm_pre_clip_avg=0.1519 | Metrics: + {'align_loss': 0.024953357875347137, + 'recon_loss': 0.08987217396497726, + 'predict_loss': 0.005772244650870562, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18903231620788574, + 'data_time': 0.0010616979852784425, + 'model_time': 1.2846949959930498, + 'grad_norm_pre_clip_avg': 0.15194349437952043, + 'learning_rate': 7.3697419898214116e-06, + 'epoch': 8.47} +04/19 [23:34:43] INFO | >> train_qwenlatent.py:487 + Step 33590 | grad_norm_pre_clip=0.1512 | + grad_norm_pre_clip_avg=0.1502 | Metrics: + {'align_loss': 0.025111159309744835, + 'recon_loss': 0.09131613373756409, + 'predict_loss': 0.009601429104804993, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15115578472614288, + 'data_time': 0.000915785989491269, + 'model_time': 1.2316140350012574, + 'grad_norm_pre_clip_avg': 0.15022484287619592, + 'learning_rate': 7.3617989559991636e-06, + 'epoch': 8.48} +04/19 [23:34:57] INFO | >> train_qwenlatent.py:487 + Step 33600 | grad_norm_pre_clip=0.1782 | + grad_norm_pre_clip_avg=0.1651 | Metrics: + {'align_loss': 0.02449803054332733, + 'recon_loss': 0.06679890304803848, + 'predict_loss': 0.0052676135674119, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17815238237380981, + 'mae_score': 0.007063194652935406, 'data_time': + 0.0007180329994298518, 'model_time': + 1.561572877020808, 'grad_norm_pre_clip_avg': + 0.16511106938123704, 'learning_rate': + 7.3538584325660365e-06, 'epoch': 8.48} +04/19 [23:35:10] INFO | >> train_qwenlatent.py:487 + Step 33610 | grad_norm_pre_clip=0.1783 | + grad_norm_pre_clip_avg=0.1604 | Metrics: + {'align_loss': 0.025543686002492905, + 'recon_loss': 0.12896916270256042, + 'predict_loss': 0.011457603424787521, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17831939458847046, + 'data_time': 0.0006462939782068133, + 'model_time': 1.220180967997294, + 'grad_norm_pre_clip_avg': 0.16042440980672837, + 'learning_rate': 7.345920423392141e-06, + 'epoch': 8.48} +04/19 [23:35:23] INFO | >> train_qwenlatent.py:487 + Step 33620 | grad_norm_pre_clip=0.1419 | + grad_norm_pre_clip_avg=0.1719 | Metrics: + {'align_loss': 0.025885270908474922, + 'recon_loss': 0.10204695910215378, + 'predict_loss': 0.005059384275227785, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14185068011283875, + 'data_time': 0.0006698859797324985, + 'model_time': 1.2417754259950016, + 'grad_norm_pre_clip_avg': 0.17188151627779008, + 'learning_rate': 7.337984932346366e-06, + 'epoch': 8.48} +04/19 [23:35:35] INFO | >> train_qwenlatent.py:487 + Step 33630 | grad_norm_pre_clip=0.1516 | + grad_norm_pre_clip_avg=0.1833 | Metrics: + {'align_loss': 0.023568574339151382, + 'recon_loss': 0.07762517035007477, + 'predict_loss': 0.0050102779641747475, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1515558809041977, + 'data_time': 0.0009641369979362935, + 'model_time': 1.3115363039833028, + 'grad_norm_pre_clip_avg': 0.18327089995145798, + 'learning_rate': 7.330051963296373e-06, + 'epoch': 8.49} +04/19 [23:35:48] INFO | >> train_qwenlatent.py:487 + Step 33640 | grad_norm_pre_clip=0.1606 | + grad_norm_pre_clip_avg=0.1879 | Metrics: + {'align_loss': 0.02507695183157921, + 'recon_loss': 0.08325131237506866, + 'predict_loss': 0.007090637926012278, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16062726080417633, + 'data_time': 0.0009138760215137154, + 'model_time': 1.2831628580170218, + 'grad_norm_pre_clip_avg': 0.18789940625429152, + 'learning_rate': 7.322121520108598e-06, + 'epoch': 8.49} +04/19 [23:36:02] INFO | >> train_qwenlatent.py:487 + Step 33650 | grad_norm_pre_clip=0.1753 | + grad_norm_pre_clip_avg=0.1856 | Metrics: + {'align_loss': 0.02505890280008316, + 'recon_loss': 0.1370711475610733, + 'predict_loss': 0.01000076811760664, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1752917468547821, + 'mae_score': 0.007547138832710885, 'data_time': + 0.000879660015925765, 'model_time': + 1.2701041449909098, 'grad_norm_pre_clip_avg': + 0.18560311198234558, 'learning_rate': + 7.314193606648243e-06, 'epoch': 8.49} +04/19 [23:36:14] INFO | >> train_qwenlatent.py:487 + Step 33660 | grad_norm_pre_clip=0.1702 | + grad_norm_pre_clip_avg=0.1497 | Metrics: + {'align_loss': 0.024761300534009933, + 'recon_loss': 0.11611229181289673, + 'predict_loss': 0.00851687602698803, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17024467885494232, + 'data_time': 0.001330992003204301, + 'model_time': 1.2387054369901307, + 'grad_norm_pre_clip_avg': 0.1496746815741062, + 'learning_rate': 7.3062682267792716e-06, + 'epoch': 8.49} +04/19 [23:36:27] INFO | >> train_qwenlatent.py:487 + Step 33670 | grad_norm_pre_clip=0.1783 | + grad_norm_pre_clip_avg=0.1605 | Metrics: + {'align_loss': 0.025526417419314384, + 'recon_loss': 0.08298677951097488, + 'predict_loss': 0.007722549140453339, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1782984435558319, + 'data_time': 0.0006374469958245754, + 'model_time': 1.2078595320053864, + 'grad_norm_pre_clip_avg': 0.16049399375915527, + 'learning_rate': 7.29834538436442e-06, 'epoch': + 8.5} +04/19 [23:36:39] INFO | >> train_qwenlatent.py:487 + Step 33680 | grad_norm_pre_clip=0.1623 | + grad_norm_pre_clip_avg=0.1575 | Metrics: + {'align_loss': 0.025954384356737137, + 'recon_loss': 0.12144406139850616, + 'predict_loss': 0.012441199272871017, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1623038351535797, + 'data_time': 0.0010994189942721277, + 'model_time': 1.2363300420111045, + 'grad_norm_pre_clip_avg': 0.15748558640480043, + 'learning_rate': 7.2904250832651875e-06, + 'epoch': 8.5} +04/19 [23:36:51] INFO | >> train_qwenlatent.py:487 + Step 33690 | grad_norm_pre_clip=0.2881 | + grad_norm_pre_clip_avg=0.1837 | Metrics: + {'align_loss': 0.025796685367822647, + 'recon_loss': 0.1280161738395691, + 'predict_loss': 0.010231762193143368, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28807753324508667, + 'data_time': 0.0006294740014709532, + 'model_time': 1.324347136018332, + 'grad_norm_pre_clip_avg': 0.18374210670590402, + 'learning_rate': 7.282507327341831e-06, + 'epoch': 8.5} +04/19 [23:37:05] INFO | >> train_qwenlatent.py:487 + Step 33700 | grad_norm_pre_clip=0.1817 | + grad_norm_pre_clip_avg=0.2038 | Metrics: + {'align_loss': 0.024702932685613632, + 'recon_loss': 0.08244329690933228, + 'predict_loss': 0.009436718188226223, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18165330588817596, + 'mae_score': 0.008256692284936303, 'data_time': + 0.0006517139845527709, 'model_time': + 1.2027211289969273, 'grad_norm_pre_clip_avg': + 0.20384038537740706, 'learning_rate': + 7.274592120453363e-06, 'epoch': 8.5} +04/19 [23:37:17] INFO | >> train_qwenlatent.py:487 + Step 33710 | grad_norm_pre_clip=0.1845 | + grad_norm_pre_clip_avg=0.1870 | Metrics: + {'align_loss': 0.0251820906996727, + 'recon_loss': 0.10142617672681808, + 'predict_loss': 0.0064928880892694, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18446676433086395, + 'data_time': 0.0006548080127686262, + 'model_time': 1.184500299015781, + 'grad_norm_pre_clip_avg': 0.187015663087368, + 'learning_rate': 7.26667946645757e-06, 'epoch': + 8.51} +04/19 [23:37:29] INFO | >> train_qwenlatent.py:487 + Step 33720 | grad_norm_pre_clip=0.1801 | + grad_norm_pre_clip_avg=0.1880 | Metrics: + {'align_loss': 0.02573036588728428, + 'recon_loss': 0.13804635405540466, + 'predict_loss': 0.015990231186151505, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18012960255146027, + 'data_time': 0.0007573500042781234, + 'model_time': 1.2325188069953583, + 'grad_norm_pre_clip_avg': 0.18803187012672423, + 'learning_rate': 7.258769369210977e-06, + 'epoch': 8.51} +04/19 [23:37:42] INFO | >> train_qwenlatent.py:487 + Step 33730 | grad_norm_pre_clip=0.1435 | + grad_norm_pre_clip_avg=0.1533 | Metrics: + {'align_loss': 0.02500547468662262, + 'recon_loss': 0.10231748223304749, + 'predict_loss': 0.008782417513430119, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1434502750635147, + 'data_time': 0.0007189419993665069, + 'model_time': 1.1777680910017807, + 'grad_norm_pre_clip_avg': 0.1532505676150322, + 'learning_rate': 7.2508618325688715e-06, + 'epoch': 8.51} +04/19 [23:37:54] INFO | >> train_qwenlatent.py:487 + Step 33740 | grad_norm_pre_clip=0.1471 | + grad_norm_pre_clip_avg=0.1762 | Metrics: + {'align_loss': 0.0255268607288599, + 'recon_loss': 0.11558572947978973, + 'predict_loss': 0.005419130437076092, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1471405029296875, + 'data_time': 0.0009853479859884828, + 'model_time': 1.278036340023391, + 'grad_norm_pre_clip_avg': 0.17615295499563216, + 'learning_rate': 7.242956860385289e-06, + 'epoch': 8.51} +04/19 [23:38:07] INFO | >> train_qwenlatent.py:487 + Step 33750 | grad_norm_pre_clip=0.1443 | + grad_norm_pre_clip_avg=0.1611 | Metrics: + {'align_loss': 0.024029742926359177, + 'recon_loss': 0.08613567054271698, + 'predict_loss': 0.009818613529205322, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1443387120962143, + 'mae_score': 0.007570975535624736, 'data_time': + 0.0008460480021312833, 'model_time': + 1.256828851008322, 'grad_norm_pre_clip_avg': + 0.1610584944486618, 'learning_rate': + 7.2350544565130186e-06, 'epoch': 8.52} +04/19 [23:38:21] INFO | >> train_qwenlatent.py:487 + Step 33760 | grad_norm_pre_clip=0.1742 | + grad_norm_pre_clip_avg=0.2037 | Metrics: + {'align_loss': 0.02545435167849064, + 'recon_loss': 0.09709805995225906, + 'predict_loss': 0.005301049444824457, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1741909384727478, + 'data_time': 0.0011178740242030472, + 'model_time': 1.30482756200945, + 'grad_norm_pre_clip_avg': 0.20372222363948822, + 'learning_rate': 7.2271546248035935e-06, + 'epoch': 8.52} +04/19 [23:38:34] INFO | >> train_qwenlatent.py:487 + Step 33770 | grad_norm_pre_clip=0.1650 | + grad_norm_pre_clip_avg=0.1810 | Metrics: + {'align_loss': 0.02609254978597164, + 'recon_loss': 0.13924682140350342, + 'predict_loss': 0.011351377703249454, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16500096023082733, + 'data_time': 0.0007149680168367922, + 'model_time': 1.2415086620021611, + 'grad_norm_pre_clip_avg': 0.1809722900390625, + 'learning_rate': 7.219257369107297e-06, + 'epoch': 8.52} +04/19 [23:38:46] INFO | >> train_qwenlatent.py:487 + Step 33780 | grad_norm_pre_clip=0.1511 | + grad_norm_pre_clip_avg=0.1682 | Metrics: + {'align_loss': 0.026542019098997116, + 'recon_loss': 0.11831165105104446, + 'predict_loss': 0.00888215284794569, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15109023451805115, + 'data_time': 0.0006889959913678467, + 'model_time': 1.31206452898914, + 'grad_norm_pre_clip_avg': 0.16819675713777543, + 'learning_rate': 7.211362693273157e-06, + 'epoch': 8.52} +04/19 [23:38:59] INFO | >> train_qwenlatent.py:487 + Step 33790 | grad_norm_pre_clip=0.1474 | + grad_norm_pre_clip_avg=0.1760 | Metrics: + {'align_loss': 0.024970749393105507, + 'recon_loss': 0.112306147813797, + 'predict_loss': 0.009246714413166046, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14743398129940033, + 'data_time': 0.0011837870115414262, + 'model_time': 1.2820756959845312, + 'grad_norm_pre_clip_avg': 0.17597181126475334, + 'learning_rate': 7.203470601148936e-06, + 'epoch': 8.53} +04/19 [23:39:12] INFO | >> train_qwenlatent.py:487 + Step 33800 | grad_norm_pre_clip=0.1679 | + grad_norm_pre_clip_avg=0.1911 | Metrics: + {'align_loss': 0.02510557696223259, + 'recon_loss': 0.09646393358707428, + 'predict_loss': 0.0043099261820316315, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16790534555912018, + 'mae_score': 0.007689581690607844, 'data_time': + 0.0009605630184523761, 'model_time': + 1.2558956400025636, 'grad_norm_pre_clip_avg': + 0.19109166264533997, 'learning_rate': + 7.1955810965811555e-06, 'epoch': 8.53} +04/19 [23:39:25] INFO | >> train_qwenlatent.py:487 + Step 33810 | grad_norm_pre_clip=0.2116 | + grad_norm_pre_clip_avg=0.1685 | Metrics: + {'align_loss': 0.025035900995135307, + 'recon_loss': 0.11245289444923401, + 'predict_loss': 0.006015668623149395, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2115514725446701, + 'data_time': 0.0006678140198346227, + 'model_time': 1.2544500879885163, + 'grad_norm_pre_clip_avg': 0.16850693821907042, + 'learning_rate': 7.187694183415056e-06, + 'epoch': 8.53} +04/19 [23:39:37] INFO | >> train_qwenlatent.py:487 + Step 33820 | grad_norm_pre_clip=0.1639 | + grad_norm_pre_clip_avg=0.1573 | Metrics: + {'align_loss': 0.024178916588425636, + 'recon_loss': 0.07267042249441147, + 'predict_loss': 0.006339249666780233, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1639455407857895, + 'data_time': 0.0008441389945801347, + 'model_time': 1.2221737710060552, + 'grad_norm_pre_clip_avg': 0.15728031918406488, + 'learning_rate': 7.179809865494626e-06, + 'epoch': 8.53} +04/19 [23:39:50] INFO | >> train_qwenlatent.py:487 + Step 33830 | grad_norm_pre_clip=0.1468 | + grad_norm_pre_clip_avg=0.1771 | Metrics: + {'align_loss': 0.025863733142614365, + 'recon_loss': 0.16377773880958557, + 'predict_loss': 0.011396472342312336, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14676815271377563, + 'data_time': 0.0010052639991044998, + 'model_time': 1.280489729018882, + 'grad_norm_pre_clip_avg': 0.17705054879188536, + 'learning_rate': 7.171928146662585e-06, + 'epoch': 8.54} +04/19 [23:40:03] INFO | >> train_qwenlatent.py:487 + Step 33840 | grad_norm_pre_clip=0.1558 | + grad_norm_pre_clip_avg=0.1589 | Metrics: + {'align_loss': 0.02577332779765129, + 'recon_loss': 0.13174860179424286, + 'predict_loss': 0.007207391317933798, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.155756413936615, + 'data_time': 0.0007779009756632149, + 'model_time': 1.226452393020736, + 'grad_norm_pre_clip_avg': 0.15890122801065446, + 'learning_rate': 7.164049030760386e-06, + 'epoch': 8.54} +04/19 [23:40:16] INFO | >> train_qwenlatent.py:487 + Step 33850 | grad_norm_pre_clip=0.1517 | + grad_norm_pre_clip_avg=0.1683 | Metrics: + {'align_loss': 0.024501711130142212, + 'recon_loss': 0.09572093933820724, + 'predict_loss': 0.009041124023497105, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15166574716567993, + 'mae_score': 0.01700605615839228, 'data_time': + 0.0009881689911708236, 'model_time': + 1.222489599022083, 'grad_norm_pre_clip_avg': + 0.1682796597480774, 'learning_rate': + 7.156172521628218e-06, 'epoch': 8.54} +04/19 [23:40:28] INFO | >> train_qwenlatent.py:487 + Step 33860 | grad_norm_pre_clip=0.2096 | + grad_norm_pre_clip_avg=0.1867 | Metrics: + {'align_loss': 0.025774091482162476, + 'recon_loss': 0.13373930752277374, + 'predict_loss': 0.011887019500136375, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2095726579427719, + 'data_time': 0.0006492890242952853, + 'model_time': 1.1984159399871714, + 'grad_norm_pre_clip_avg': 0.1867161825299263, + 'learning_rate': 7.148298623104992e-06, + 'epoch': 8.54} +04/19 [23:40:41] INFO | >> train_qwenlatent.py:487 + Step 33870 | grad_norm_pre_clip=0.1468 | + grad_norm_pre_clip_avg=0.1848 | Metrics: + {'align_loss': 0.026144221425056458, + 'recon_loss': 0.09225668758153915, + 'predict_loss': 0.008052757941186428, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14682354032993317, + 'data_time': 0.0008199529838748276, + 'model_time': 1.200311624997994, + 'grad_norm_pre_clip_avg': 0.18479647487401962, + 'learning_rate': 7.140427339028352e-06, + 'epoch': 8.55} +04/19 [23:40:53] INFO | >> train_qwenlatent.py:487 + Step 33880 | grad_norm_pre_clip=0.1807 | + grad_norm_pre_clip_avg=0.1676 | Metrics: + {'align_loss': 0.02718425542116165, + 'recon_loss': 0.15525060892105103, + 'predict_loss': 0.013709448277950287, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1807333081960678, + 'data_time': 0.0010201429831795394, + 'model_time': 1.2095837079978082, + 'grad_norm_pre_clip_avg': 0.1676175117492676, + 'learning_rate': 7.132558673234668e-06, + 'epoch': 8.55} +04/19 [23:41:06] INFO | >> train_qwenlatent.py:487 + Step 33890 | grad_norm_pre_clip=0.1671 | + grad_norm_pre_clip_avg=0.1715 | Metrics: + {'align_loss': 0.025510121136903763, + 'recon_loss': 0.10769093781709671, + 'predict_loss': 0.00870785303413868, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16707663238048553, + 'data_time': 0.0009367349848616868, + 'model_time': 1.2429306949779857, + 'grad_norm_pre_clip_avg': 0.17150294482707978, + 'learning_rate': 7.124692629559024e-06, + 'epoch': 8.55} +04/19 [23:41:19] INFO | >> train_qwenlatent.py:487 + Step 33900 | grad_norm_pre_clip=0.2792 | + grad_norm_pre_clip_avg=0.1662 | Metrics: + {'align_loss': 0.024750208482146263, + 'recon_loss': 0.11867981404066086, + 'predict_loss': 0.012203817255795002, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2792411148548126, + 'mae_score': 0.007754511446566195, 'data_time': + 0.0008399210055358708, 'model_time': + 1.1999998799874447, 'grad_norm_pre_clip_avg': + 0.1661614939570427, 'learning_rate': + 7.1168292118352435e-06, 'epoch': 8.55} +04/19 [23:41:33] INFO | >> train_qwenlatent.py:487 + Step 33910 | grad_norm_pre_clip=0.1825 | + grad_norm_pre_clip_avg=0.1916 | Metrics: + {'align_loss': 0.024769315496087074, + 'recon_loss': 0.14028535783290863, + 'predict_loss': 0.013402699492871761, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18247786164283752, + 'data_time': 0.0007195959915407002, + 'model_time': 1.2551574099925347, + 'grad_norm_pre_clip_avg': 0.19159875959157943, + 'learning_rate': 7.108968423895862e-06, + 'epoch': 8.56} +04/19 [23:41:45] INFO | >> train_qwenlatent.py:487 + Step 33920 | grad_norm_pre_clip=0.1323 | + grad_norm_pre_clip_avg=0.1893 | Metrics: + {'align_loss': 0.025118635967373848, + 'recon_loss': 0.1347057819366455, + 'predict_loss': 0.009591598995029926, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1322910189628601, + 'data_time': 0.0008869710145518184, + 'model_time': 1.2629639809892979, + 'grad_norm_pre_clip_avg': 0.18930291682481765, + 'learning_rate': 7.101110269572125e-06, + 'epoch': 8.56} +04/19 [23:41:58] INFO | >> train_qwenlatent.py:487 + Step 33930 | grad_norm_pre_clip=0.1178 | + grad_norm_pre_clip_avg=0.1499 | Metrics: + {'align_loss': 0.02456311136484146, + 'recon_loss': 0.07641318440437317, + 'predict_loss': 0.0029845398385077715, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11779245734214783, + 'data_time': 0.0006579170003533363, + 'model_time': 1.2859881590120494, + 'grad_norm_pre_clip_avg': 0.1499376505613327, + 'learning_rate': 7.093254752694005e-06, + 'epoch': 8.56} +04/19 [23:42:10] INFO | >> train_qwenlatent.py:487 + Step 33940 | grad_norm_pre_clip=0.1608 | + grad_norm_pre_clip_avg=0.1575 | Metrics: + {'align_loss': 0.026509582996368408, + 'recon_loss': 0.17645373940467834, + 'predict_loss': 0.009925385005772114, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16082890331745148, + 'data_time': 0.0007529670256190002, + 'model_time': 1.1966280050110072, + 'grad_norm_pre_clip_avg': 0.15752127021551132, + 'learning_rate': 7.085401877090186e-06, + 'epoch': 8.56} +04/19 [23:42:23] INFO | >> train_qwenlatent.py:487 + Step 33950 | grad_norm_pre_clip=0.1536 | + grad_norm_pre_clip_avg=0.1764 | Metrics: + {'align_loss': 0.025185851380228996, + 'recon_loss': 0.12419956177473068, + 'predict_loss': 0.013780743815004826, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15355512499809265, + 'mae_score': 0.008525376706509977, 'data_time': + 0.0008367630071006715, 'model_time': + 1.2521880069980398, 'grad_norm_pre_clip_avg': + 0.1763889953494072, 'learning_rate': + 7.077551646588063e-06, 'epoch': 8.57} +04/19 [23:42:35] INFO | >> train_qwenlatent.py:487 + Step 33960 | grad_norm_pre_clip=0.1820 | + grad_norm_pre_clip_avg=0.1766 | Metrics: + {'align_loss': 0.02647233009338379, + 'recon_loss': 0.14384114742279053, + 'predict_loss': 0.010148700326681137, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1819523423910141, + 'data_time': 0.0010232659988105297, + 'model_time': 1.1967807869950775, + 'grad_norm_pre_clip_avg': 0.17658619880676268, + 'learning_rate': 7.069704065013741e-06, + 'epoch': 8.57} +04/19 [23:42:48] INFO | >> train_qwenlatent.py:487 + Step 33970 | grad_norm_pre_clip=0.1602 | + grad_norm_pre_clip_avg=0.1595 | Metrics: + {'align_loss': 0.02503758668899536, + 'recon_loss': 0.0948072001338005, + 'predict_loss': 0.004842120688408613, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16018928587436676, + 'data_time': 0.0007147800060920417, + 'model_time': 1.244311678019585, + 'grad_norm_pre_clip_avg': 0.15951284766197205, + 'learning_rate': 7.061859136192038e-06, + 'epoch': 8.57} +04/19 [23:43:00] INFO | >> train_qwenlatent.py:487 + Step 33980 | grad_norm_pre_clip=0.1986 | + grad_norm_pre_clip_avg=0.1746 | Metrics: + {'align_loss': 0.025432132184505463, + 'recon_loss': 0.1056135892868042, + 'predict_loss': 0.007110357750207186, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1986055225133896, + 'data_time': 0.0008821810188237578, + 'model_time': 1.2441226869996171, + 'grad_norm_pre_clip_avg': 0.1745595119893551, + 'learning_rate': 7.054016863946479e-06, + 'epoch': 8.57} +04/19 [23:43:13] INFO | >> train_qwenlatent.py:487 + Step 33990 | grad_norm_pre_clip=0.1581 | + grad_norm_pre_clip_avg=0.1716 | Metrics: + {'align_loss': 0.024991072714328766, + 'recon_loss': 0.07949267327785492, + 'predict_loss': 0.005091323051601648, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15813341736793518, + 'data_time': 0.0011307740060146898, + 'model_time': 1.2741825350094587, + 'grad_norm_pre_clip_avg': 0.1715697705745697, + 'learning_rate': 7.046177252099283e-06, + 'epoch': 8.58} +04/19 [23:43:26] INFO | >> train_qwenlatent.py:487 + Step 34000 | grad_norm_pre_clip=0.1493 | + grad_norm_pre_clip_avg=0.1581 | Metrics: + {'align_loss': 0.02436656877398491, + 'recon_loss': 0.07962952554225922, + 'predict_loss': 0.006531751248985529, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14928750693798065, + 'mae_score': 0.0073044922974732545, + 'data_time': 0.0007582480029668659, + 'model_time': 1.3056423850066494, + 'grad_norm_pre_clip_avg': 0.1581396445631981, + 'learning_rate': 7.0383403044713955e-06, + 'epoch': 8.58} +04/19 [23:43:39] INFO | >> train_qwenlatent.py:487 + Step 34010 | grad_norm_pre_clip=0.1732 | + grad_norm_pre_clip_avg=0.1589 | Metrics: + {'align_loss': 0.02628779411315918, + 'recon_loss': 0.09676297008991241, + 'predict_loss': 0.004094518255442381, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1732175201177597, + 'data_time': 0.001018690993078053, + 'model_time': 1.1817103740177117, + 'grad_norm_pre_clip_avg': 0.15886019319295883, + 'learning_rate': 7.030506024882441e-06, + 'epoch': 8.58} +04/19 [23:43:51] INFO | >> train_qwenlatent.py:487 + Step 34020 | grad_norm_pre_clip=0.1481 | + grad_norm_pre_clip_avg=0.1755 | Metrics: + {'align_loss': 0.024670369923114777, + 'recon_loss': 0.10077718645334244, + 'predict_loss': 0.009600981138646603, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14812922477722168, + 'data_time': 0.0009261860104743391, + 'model_time': 1.228571699000895, + 'grad_norm_pre_clip_avg': 0.1755157381296158, + 'learning_rate': 7.022674417150761e-06, + 'epoch': 8.58} +04/19 [23:44:04] INFO | >> train_qwenlatent.py:487 + Step 34030 | grad_norm_pre_clip=0.1220 | + grad_norm_pre_clip_avg=0.1531 | Metrics: + {'align_loss': 0.024751579388976097, + 'recon_loss': 0.12459561973810196, + 'predict_loss': 0.0072279735468328, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12202727049589157, + 'data_time': 0.0007416259904857725, + 'model_time': 1.2341772479994688, + 'grad_norm_pre_clip_avg': 0.15312360599637032, + 'learning_rate': 7.014845485093366e-06, + 'epoch': 8.59} +04/19 [23:44:17] INFO | >> train_qwenlatent.py:487 + Step 34040 | grad_norm_pre_clip=0.1200 | + grad_norm_pre_clip_avg=0.1763 | Metrics: + {'align_loss': 0.023858461529016495, + 'recon_loss': 0.09880887717008591, + 'predict_loss': 0.004198994487524033, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12003198266029358, + 'data_time': 0.0008733670110814273, + 'model_time': 1.2310841970029287, + 'grad_norm_pre_clip_avg': 0.17630262970924376, + 'learning_rate': 7.0070192325260035e-06, + 'epoch': 8.59} +04/19 [23:44:30] INFO | >> train_qwenlatent.py:487 + Step 34050 | grad_norm_pre_clip=0.1894 | + grad_norm_pre_clip_avg=0.1837 | Metrics: + {'align_loss': 0.027216263115406036, + 'recon_loss': 0.12407097965478897, + 'predict_loss': 0.007329843007028103, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18940557539463043, + 'mae_score': 0.007716627378721495, 'data_time': + 0.0006664399988949299, 'model_time': + 1.2229516549850814, 'grad_norm_pre_clip_avg': + 0.18370763957500458, 'learning_rate': + 6.999195663263083e-06, 'epoch': 8.59} +04/19 [23:44:42] INFO | >> train_qwenlatent.py:487 + Step 34060 | grad_norm_pre_clip=0.1814 | + grad_norm_pre_clip_avg=0.1802 | Metrics: + {'align_loss': 0.025983314961194992, + 'recon_loss': 0.10038069635629654, + 'predict_loss': 0.007502981461584568, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18139934539794922, + 'data_time': 0.0008481009863317013, + 'model_time': 1.1934215709916316, + 'grad_norm_pre_clip_avg': 0.1802274152636528, + 'learning_rate': 6.991374781117721e-06, + 'epoch': 8.59} +04/19 [23:44:55] INFO | >> train_qwenlatent.py:487 + Step 34070 | grad_norm_pre_clip=0.2042 | + grad_norm_pre_clip_avg=0.1616 | Metrics: + {'align_loss': 0.026025203987956047, + 'recon_loss': 0.11471208184957504, + 'predict_loss': 0.014035643078386784, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20421379804611206, + 'data_time': 0.0010320619912818074, + 'model_time': 1.3092553220049012, + 'grad_norm_pre_clip_avg': 0.16163571625947953, + 'learning_rate': 6.98355658990172e-06, 'epoch': + 8.6} +04/19 [23:45:07] INFO | >> train_qwenlatent.py:487 + Step 34080 | grad_norm_pre_clip=0.1631 | + grad_norm_pre_clip_avg=0.1600 | Metrics: + {'align_loss': 0.02528962679207325, + 'recon_loss': 0.07270681113004684, + 'predict_loss': 0.005264858715236187, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16314265131950378, + 'data_time': 0.0006438340060412884, + 'model_time': 1.1895449820149224, + 'grad_norm_pre_clip_avg': 0.15998777747154236, + 'learning_rate': 6.9757410934255645e-06, + 'epoch': 8.6} +04/19 [23:45:20] INFO | >> train_qwenlatent.py:487 + Step 34090 | grad_norm_pre_clip=0.1498 | + grad_norm_pre_clip_avg=0.1569 | Metrics: + {'align_loss': 0.025180697441101074, + 'recon_loss': 0.10528253018856049, + 'predict_loss': 0.006326597183942795, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1498212069272995, + 'data_time': 0.0006508890073746443, + 'model_time': 1.238556440017419, + 'grad_norm_pre_clip_avg': 0.1568852648139, + 'learning_rate': 6.9679282954984455e-06, + 'epoch': 8.6} +04/19 [23:45:33] INFO | >> train_qwenlatent.py:487 + Step 34100 | grad_norm_pre_clip=0.1706 | + grad_norm_pre_clip_avg=0.1772 | Metrics: + {'align_loss': 0.02518571726977825, + 'recon_loss': 0.11957263201475143, + 'predict_loss': 0.012023226357996464, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1705561727285385, + 'mae_score': 0.012071590595417195, 'data_time': + 0.0010593570186756551, 'model_time': + 1.264090180018684, 'grad_norm_pre_clip_avg': + 0.1772375226020813, 'learning_rate': + 6.960118199928221e-06, 'epoch': 8.6} +04/19 [23:45:46] INFO | >> train_qwenlatent.py:487 + Step 34110 | grad_norm_pre_clip=0.2440 | + grad_norm_pre_clip_avg=0.1874 | Metrics: + {'align_loss': 0.025364089757204056, + 'recon_loss': 0.15208245813846588, + 'predict_loss': 0.009032496251165867, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2439829260110855, + 'data_time': 0.0006607559917028993, + 'model_time': 1.2867412259802222, + 'grad_norm_pre_clip_avg': 0.18742403015494347, + 'learning_rate': 6.952310810521433e-06, + 'epoch': 8.61} +04/19 [23:45:58] INFO | >> train_qwenlatent.py:487 + Step 34120 | grad_norm_pre_clip=0.1531 | + grad_norm_pre_clip_avg=0.1581 | Metrics: + {'align_loss': 0.024781623855233192, + 'recon_loss': 0.08293699473142624, + 'predict_loss': 0.004196742549538612, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15307985246181488, + 'data_time': 0.0007909729902166873, + 'model_time': 1.2146434679743834, + 'grad_norm_pre_clip_avg': 0.15813119933009148, + 'learning_rate': 6.944506131083303e-06, + 'epoch': 8.61} +04/19 [23:46:11] INFO | >> train_qwenlatent.py:487 + Step 34130 | grad_norm_pre_clip=0.1205 | + grad_norm_pre_clip_avg=0.1423 | Metrics: + {'align_loss': 0.025249000638723373, + 'recon_loss': 0.0848388671875, 'predict_loss': + 0.0063570234924554825, 'aux_loss_decay_weight': + 0.0, 'grad_norm_pre_clip': 0.12052130699157715, + 'data_time': 0.0006926879868842661, + 'model_time': 1.218908635986736, + 'grad_norm_pre_clip_avg': 0.14233506396412848, + 'learning_rate': 6.93670416541775e-06, 'epoch': + 8.61} +04/19 [23:46:24] INFO | >> train_qwenlatent.py:487 + Step 34140 | grad_norm_pre_clip=0.1688 | + grad_norm_pre_clip_avg=0.1546 | Metrics: + {'align_loss': 0.025158755481243134, + 'recon_loss': 0.15324382483959198, + 'predict_loss': 0.014708193019032478, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16876593232154846, + 'data_time': 0.000695670023560524, + 'model_time': 1.1932724849903025, + 'grad_norm_pre_clip_avg': 0.15462716072797775, + 'learning_rate': 6.92890491732735e-06, 'epoch': + 8.61} +04/19 [23:46:37] INFO | >> train_qwenlatent.py:487 + Step 34150 | grad_norm_pre_clip=0.1867 | + grad_norm_pre_clip_avg=0.2132 | Metrics: + {'align_loss': 0.02520078793168068, + 'recon_loss': 0.10499870032072067, + 'predict_loss': 0.007419075816869736, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18667078018188477, + 'mae_score': 0.008511764938766892, 'data_time': + 0.0009518519800622016, 'model_time': + 1.5277716020063963, 'grad_norm_pre_clip_avg': + 0.21321005672216414, 'learning_rate': + 6.921108390613365e-06, 'epoch': 8.62} +04/19 [23:46:49] INFO | >> train_qwenlatent.py:487 + Step 34160 | grad_norm_pre_clip=0.1675 | + grad_norm_pre_clip_avg=0.1643 | Metrics: + {'align_loss': 0.025650545954704285, + 'recon_loss': 0.1020728200674057, + 'predict_loss': 0.0071766916662454605, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1674812138080597, + 'data_time': 0.0010209110041614622, + 'model_time': 1.204280136997113, + 'grad_norm_pre_clip_avg': 0.1643380381166935, + 'learning_rate': 6.913314589075725e-06, + 'epoch': 8.62} +04/19 [23:47:03] INFO | >> train_qwenlatent.py:487 + Step 34170 | grad_norm_pre_clip=0.2381 | + grad_norm_pre_clip_avg=0.1799 | Metrics: + {'align_loss': 0.025295894593000412, + 'recon_loss': 0.09867808222770691, + 'predict_loss': 0.012163210660219193, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23805835843086243, + 'data_time': 0.0006783510034438223, + 'model_time': 1.5006923549808562, + 'grad_norm_pre_clip_avg': 0.17993768006563188, + 'learning_rate': 6.905523516513034e-06, + 'epoch': 8.62} +04/19 [23:47:15] INFO | >> train_qwenlatent.py:487 + Step 34180 | grad_norm_pre_clip=0.1956 | + grad_norm_pre_clip_avg=0.1642 | Metrics: + {'align_loss': 0.026400765404105186, + 'recon_loss': 0.15929247438907623, + 'predict_loss': 0.007316754665225744, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1956191211938858, + 'data_time': 0.0009334000060334802, + 'model_time': 1.2264681720116641, + 'grad_norm_pre_clip_avg': 0.1642355963587761, + 'learning_rate': 6.897735176722561e-06, + 'epoch': 8.62} +04/19 [23:47:28] INFO | >> train_qwenlatent.py:487 + Step 34190 | grad_norm_pre_clip=0.1918 | + grad_norm_pre_clip_avg=0.1893 | Metrics: + {'align_loss': 0.024445313960313797, + 'recon_loss': 0.1267598420381546, + 'predict_loss': 0.011341582983732224, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1917950063943863, + 'data_time': 0.0009425639873370528, + 'model_time': 1.2770891079853754, + 'grad_norm_pre_clip_avg': 0.18928841054439544, + 'learning_rate': 6.889949573500264e-06, + 'epoch': 8.63} +04/19 [23:47:41] INFO | >> train_qwenlatent.py:487 + Step 34200 | grad_norm_pre_clip=0.1352 | + grad_norm_pre_clip_avg=0.1934 | Metrics: + {'align_loss': 0.02497619390487671, + 'recon_loss': 0.09422921389341354, + 'predict_loss': 0.007641671225428581, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13518132269382477, + 'mae_score': 0.007515213940594647, 'data_time': + 0.0006799320108257234, 'model_time': + 1.214981440018164, 'grad_norm_pre_clip_avg': + 0.1933972030878067, 'learning_rate': + 6.882166710640736e-06, 'epoch': 8.63} +04/19 [23:47:53] INFO | >> train_qwenlatent.py:487 + Step 34210 | grad_norm_pre_clip=0.1693 | + grad_norm_pre_clip_avg=0.1547 | Metrics: + {'align_loss': 0.025465361773967743, + 'recon_loss': 0.11890561133623123, + 'predict_loss': 0.009960977360606194, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16932938992977142, + 'data_time': 0.0006360420084092766, + 'model_time': 1.2354518189968076, + 'grad_norm_pre_clip_avg': 0.1546637773513794, + 'learning_rate': 6.874386591937257e-06, + 'epoch': 8.63} +04/19 [23:48:06] INFO | >> train_qwenlatent.py:487 + Step 34220 | grad_norm_pre_clip=0.1768 | + grad_norm_pre_clip_avg=0.1515 | Metrics: + {'align_loss': 0.023378971964120865, + 'recon_loss': 0.07727838307619095, + 'predict_loss': 0.004531116224825382, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1768110692501068, + 'data_time': 0.0006857810076326132, + 'model_time': 1.2271698389959056, + 'grad_norm_pre_clip_avg': 0.15145964846014975, + 'learning_rate': 6.866609221181754e-06, + 'epoch': 8.63} +04/19 [23:48:19] INFO | >> train_qwenlatent.py:487 + Step 34230 | grad_norm_pre_clip=0.2206 | + grad_norm_pre_clip_avg=0.1871 | Metrics: + {'align_loss': 0.02727731317281723, + 'recon_loss': 0.11882052570581436, + 'predict_loss': 0.009345196187496185, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22055277228355408, + 'data_time': 0.0009170509874820709, + 'model_time': 1.1943868979869876, + 'grad_norm_pre_clip_avg': 0.1870751604437828, + 'learning_rate': 6.858834602164838e-06, + 'epoch': 8.64} +04/19 [23:48:31] INFO | >> train_qwenlatent.py:487 + Step 34240 | grad_norm_pre_clip=0.1763 | + grad_norm_pre_clip_avg=0.1870 | Metrics: + {'align_loss': 0.025105081498622894, + 'recon_loss': 0.108517125248909, + 'predict_loss': 0.009090892970561981, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1763366013765335, + 'data_time': 0.0008799649949651212, + 'model_time': 1.2481783260009252, + 'grad_norm_pre_clip_avg': 0.1870175063610077, + 'learning_rate': 6.851062738675751e-06, + 'epoch': 8.64} +04/19 [23:48:44] INFO | >> train_qwenlatent.py:487 + Step 34250 | grad_norm_pre_clip=0.1317 | + grad_norm_pre_clip_avg=0.1678 | Metrics: + {'align_loss': 0.02647346444427967, + 'recon_loss': 0.10172554850578308, + 'predict_loss': 0.0065833874978125095, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13171663880348206, + 'mae_score': 0.006735834774670301, 'data_time': + 0.0007044360099826008, 'model_time': + 1.2683214840071741, 'grad_norm_pre_clip_avg': + 0.16780855506658554, 'learning_rate': + 6.8432936345024106e-06, 'epoch': 8.64} +04/19 [23:48:57] INFO | >> train_qwenlatent.py:487 + Step 34260 | grad_norm_pre_clip=0.1857 | + grad_norm_pre_clip_avg=0.1656 | Metrics: + {'align_loss': 0.02576356753706932, + 'recon_loss': 0.11544189602136612, + 'predict_loss': 0.007787201087921858, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1856665015220642, + 'data_time': 0.0006847130134701729, + 'model_time': 1.2567908600030933, + 'grad_norm_pre_clip_avg': 0.16561369895935057, + 'learning_rate': 6.835527293431382e-06, + 'epoch': 8.64} +04/19 [23:49:10] INFO | >> train_qwenlatent.py:487 + Step 34270 | grad_norm_pre_clip=0.1546 | + grad_norm_pre_clip_avg=0.1716 | Metrics: + {'align_loss': 0.0252450592815876, + 'recon_loss': 0.10956065356731415, + 'predict_loss': 0.005295429844409227, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15461894869804382, + 'data_time': 0.0006674469914287329, + 'model_time': 1.6028334839793388, + 'grad_norm_pre_clip_avg': 0.17161465138196946, + 'learning_rate': 6.827763719247882e-06, + 'epoch': 8.65} +04/19 [23:49:22] INFO | >> train_qwenlatent.py:487 + Step 34280 | grad_norm_pre_clip=0.1956 | + grad_norm_pre_clip_avg=0.1939 | Metrics: + {'align_loss': 0.02327214740216732, + 'recon_loss': 0.09625651687383652, + 'predict_loss': 0.007084221113473177, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19558358192443848, + 'data_time': 0.0008292809943668544, + 'model_time': 1.2310117930173874, + 'grad_norm_pre_clip_avg': 0.19388724267482757, + 'learning_rate': 6.820002915735798e-06, + 'epoch': 8.65} +04/19 [23:49:35] INFO | >> train_qwenlatent.py:487 + Step 34290 | grad_norm_pre_clip=0.1803 | + grad_norm_pre_clip_avg=0.1733 | Metrics: + {'align_loss': 0.025210179388523102, + 'recon_loss': 0.12487200647592545, + 'predict_loss': 0.006526533048599958, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18031121790409088, + 'data_time': 0.0012860799906775355, + 'model_time': 1.2368839179980569, + 'grad_norm_pre_clip_avg': 0.17330172061920165, + 'learning_rate': 6.812244886677631e-06, + 'epoch': 8.65} +04/19 [23:49:49] INFO | >> train_qwenlatent.py:487 + Step 34300 | grad_norm_pre_clip=0.1431 | + grad_norm_pre_clip_avg=0.1618 | Metrics: + {'align_loss': 0.023739591240882874, + 'recon_loss': 0.08269661664962769, + 'predict_loss': 0.005777762737125158, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14310942590236664, + 'mae_score': 0.00768683064091313, 'data_time': + 0.0006585560040548444, 'model_time': + 1.2649715220031794, 'grad_norm_pre_clip_avg': + 0.16182305216789244, 'learning_rate': + 6.804489635854561e-06, 'epoch': 8.66} +04/19 [23:50:01] INFO | >> train_qwenlatent.py:487 + Step 34310 | grad_norm_pre_clip=0.1532 | + grad_norm_pre_clip_avg=0.1467 | Metrics: + {'align_loss': 0.024843022227287292, + 'recon_loss': 0.12404084205627441, + 'predict_loss': 0.010107516311109066, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15322597324848175, + 'data_time': 0.0007047520193737, 'model_time': + 1.192680426989682, 'grad_norm_pre_clip_avg': + 0.14669771790504454, 'learning_rate': + 6.796737167046399e-06, 'epoch': 8.66} +04/19 [23:50:14] INFO | >> train_qwenlatent.py:487 + Step 34320 | grad_norm_pre_clip=0.1416 | + grad_norm_pre_clip_avg=0.1491 | Metrics: + {'align_loss': 0.02388937585055828, + 'recon_loss': 0.07957770675420761, + 'predict_loss': 0.007148967124521732, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14155903458595276, + 'data_time': 0.000762178999138996, + 'model_time': 1.2532592469942756, + 'grad_norm_pre_clip_avg': 0.14909216910600662, + 'learning_rate': 6.788987484031603e-06, + 'epoch': 8.66} +04/19 [23:50:27] INFO | >> train_qwenlatent.py:487 + Step 34330 | grad_norm_pre_clip=0.1990 | + grad_norm_pre_clip_avg=0.1670 | Metrics: + {'align_loss': 0.02709517627954483, + 'recon_loss': 0.14764010906219482, + 'predict_loss': 0.007130297366529703, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19895142316818237, + 'data_time': 0.0009147019882220775, + 'model_time': 1.2284920520032756, + 'grad_norm_pre_clip_avg': 0.16700374484062194, + 'learning_rate': 6.781240590587281e-06, + 'epoch': 8.66} +04/19 [23:50:39] INFO | >> train_qwenlatent.py:487 + Step 34340 | grad_norm_pre_clip=0.1609 | + grad_norm_pre_clip_avg=0.1718 | Metrics: + {'align_loss': 0.026326419785618782, + 'recon_loss': 0.13017237186431885, + 'predict_loss': 0.00784311443567276, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1609404981136322, + 'data_time': 0.0009623010118957609, + 'model_time': 1.2097179949923884, + 'grad_norm_pre_clip_avg': 0.171753166615963, + 'learning_rate': 6.773496490489172e-06, + 'epoch': 8.67} +04/19 [23:50:52] INFO | >> train_qwenlatent.py:487 + Step 34350 | grad_norm_pre_clip=0.1894 | + grad_norm_pre_clip_avg=0.1751 | Metrics: + {'align_loss': 0.02630450204014778, + 'recon_loss': 0.13982266187667847, + 'predict_loss': 0.01051041018217802, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18936336040496826, + 'mae_score': 0.006852425755681219, 'data_time': + 0.0006553189887199551, 'model_time': + 1.5368943150097039, 'grad_norm_pre_clip_avg': + 0.17505884617567063, 'learning_rate': + 6.7657551875116535e-06, 'epoch': 8.67} +04/19 [23:51:05] INFO | >> train_qwenlatent.py:487 + Step 34360 | grad_norm_pre_clip=0.1321 | + grad_norm_pre_clip_avg=0.1536 | Metrics: + {'align_loss': 0.025593724101781845, + 'recon_loss': 0.0879983976483345, + 'predict_loss': 0.004968747496604919, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13211898505687714, + 'data_time': 0.0006753169873263687, + 'model_time': 1.197260395012563, + 'grad_norm_pre_clip_avg': 0.1535505086183548, + 'learning_rate': 6.7580166854277455e-06, + 'epoch': 8.67} +04/19 [23:51:17] INFO | >> train_qwenlatent.py:487 + Step 34370 | grad_norm_pre_clip=0.1787 | + grad_norm_pre_clip_avg=0.1543 | Metrics: + {'align_loss': 0.026262585073709488, + 'recon_loss': 0.13066351413726807, + 'predict_loss': 0.009962551295757294, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1787070333957672, + 'data_time': 0.0006978189921937883, + 'model_time': 1.2340379279921763, + 'grad_norm_pre_clip_avg': 0.15431632995605468, + 'learning_rate': 6.7502809880091e-06, 'epoch': + 8.67} +04/19 [23:51:29] INFO | >> train_qwenlatent.py:487 + Step 34380 | grad_norm_pre_clip=0.1930 | + grad_norm_pre_clip_avg=0.1786 | Metrics: + {'align_loss': 0.023269247263669968, + 'recon_loss': 0.09291265159845352, + 'predict_loss': 0.0067430646158754826, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19304616749286652, + 'data_time': 0.0008132490038406104, + 'model_time': 1.2376045389974024, + 'grad_norm_pre_clip_avg': 0.1786426141858101, + 'learning_rate': 6.742548099025998e-06, + 'epoch': 8.68} +04/19 [23:51:42] INFO | >> train_qwenlatent.py:487 + Step 34390 | grad_norm_pre_clip=0.1577 | + grad_norm_pre_clip_avg=0.1986 | Metrics: + {'align_loss': 0.023369118571281433, + 'recon_loss': 0.09761728346347809, + 'predict_loss': 0.008084878325462341, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1577192097902298, + 'data_time': 0.0008684350177645683, + 'model_time': 1.1911694439768326, + 'grad_norm_pre_clip_avg': 0.19860081374645233, + 'learning_rate': 6.7348180222473605e-06, + 'epoch': 8.68} +04/19 [23:51:55] INFO | >> train_qwenlatent.py:487 + Step 34400 | grad_norm_pre_clip=0.1514 | + grad_norm_pre_clip_avg=0.1604 | Metrics: + {'align_loss': 0.026597673073410988, + 'recon_loss': 0.14435075223445892, + 'predict_loss': 0.009455827064812183, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15139856934547424, + 'mae_score': 0.00919687253934843, 'data_time': + 0.0006950249953661114, 'model_time': + 1.2242209739924874, 'grad_norm_pre_clip_avg': + 0.16038064584136008, 'learning_rate': + 6.72709076144073e-06, 'epoch': 8.68} +04/19 [23:52:08] INFO | >> train_qwenlatent.py:487 + Step 34410 | grad_norm_pre_clip=0.1293 | + grad_norm_pre_clip_avg=0.1534 | Metrics: + {'align_loss': 0.025643089786171913, + 'recon_loss': 0.09848056733608246, + 'predict_loss': 0.00380003172904253, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12928278744220734, + 'data_time': 0.000654640985885635, + 'model_time': 1.4810729390010238, + 'grad_norm_pre_clip_avg': 0.15339258909225464, + 'learning_rate': 6.7193663203722754e-06, + 'epoch': 8.68} +04/19 [23:52:20] INFO | >> train_qwenlatent.py:487 + Step 34420 | grad_norm_pre_clip=0.1538 | + grad_norm_pre_clip_avg=0.1724 | Metrics: + {'align_loss': 0.02753576636314392, + 'recon_loss': 0.15621019899845123, + 'predict_loss': 0.007206631824374199, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15381696820259094, + 'data_time': 0.0010689010086935014, + 'model_time': 1.3321655410109088, + 'grad_norm_pre_clip_avg': 0.17235884666442872, + 'learning_rate': 6.711644702806807e-06, + 'epoch': 8.69} +04/19 [23:52:33] INFO | >> train_qwenlatent.py:487 + Step 34430 | grad_norm_pre_clip=0.2307 | + grad_norm_pre_clip_avg=0.1610 | Metrics: + {'align_loss': 0.02639181911945343, + 'recon_loss': 0.1356244683265686, + 'predict_loss': 0.006781356874853373, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23074351251125336, + 'data_time': 0.0009600410121493042, + 'model_time': 1.5225221859873272, + 'grad_norm_pre_clip_avg': 0.16102330163121223, + 'learning_rate': 6.703925912507739e-06, + 'epoch': 8.69} +04/19 [23:52:46] INFO | >> train_qwenlatent.py:487 + Step 34440 | grad_norm_pre_clip=0.1566 | + grad_norm_pre_clip_avg=0.1855 | Metrics: + {'align_loss': 0.02487633191049099, + 'recon_loss': 0.13738708198070526, + 'predict_loss': 0.011257868260145187, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15664184093475342, + 'data_time': 0.0009034589747898281, + 'model_time': 1.2365134709980339, + 'grad_norm_pre_clip_avg': 0.18554782420396804, + 'learning_rate': 6.696209953237119e-06, + 'epoch': 8.69} +04/19 [23:53:00] INFO | >> train_qwenlatent.py:487 + Step 34450 | grad_norm_pre_clip=0.1560 | + grad_norm_pre_clip_avg=0.1667 | Metrics: + {'align_loss': 0.025441303849220276, + 'recon_loss': 0.09487618505954742, + 'predict_loss': 0.0063854088075459, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15601462125778198, + 'mae_score': 0.006883631525812922, 'data_time': + 0.0009270000155083835, 'model_time': + 1.649629619991174, 'grad_norm_pre_clip_avg': + 0.16670749932527543, 'learning_rate': + 6.688496828755611e-06, 'epoch': 8.69} +04/19 [23:53:13] INFO | >> train_qwenlatent.py:487 + Step 34460 | grad_norm_pre_clip=0.1298 | + grad_norm_pre_clip_avg=0.1494 | Metrics: + {'align_loss': 0.024270711466670036, + 'recon_loss': 0.08788502216339111, + 'predict_loss': 0.006733131594955921, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1297813504934311, + 'data_time': 0.0009439730201847851, + 'model_time': 1.3014512160152663, + 'grad_norm_pre_clip_avg': 0.1493644967675209, + 'learning_rate': 6.680786542822495e-06, + 'epoch': 8.7} +04/19 [23:53:25] INFO | >> train_qwenlatent.py:487 + Step 34470 | grad_norm_pre_clip=0.1639 | + grad_norm_pre_clip_avg=0.1651 | Metrics: + {'align_loss': 0.02383190393447876, + 'recon_loss': 0.13662368059158325, + 'predict_loss': 0.009857699275016785, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16394297778606415, + 'data_time': 0.0011466450232546777, + 'model_time': 1.257726662006462, + 'grad_norm_pre_clip_avg': 0.16510873585939406, + 'learning_rate': 6.673079099195677e-06, + 'epoch': 8.7} +04/19 [23:53:38] INFO | >> train_qwenlatent.py:487 + Step 34480 | grad_norm_pre_clip=0.1628 | + grad_norm_pre_clip_avg=0.1715 | Metrics: + {'align_loss': 0.023681964725255966, + 'recon_loss': 0.09394291043281555, + 'predict_loss': 0.006288551725447178, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16280128061771393, + 'data_time': 0.0006498759903479367, + 'model_time': 1.2463056819979101, + 'grad_norm_pre_clip_avg': 0.17147209346294404, + 'learning_rate': 6.665374501631665e-06, + 'epoch': 8.7} +04/19 [23:53:50] INFO | >> train_qwenlatent.py:487 + Step 34490 | grad_norm_pre_clip=0.1774 | + grad_norm_pre_clip_avg=0.1743 | Metrics: + {'align_loss': 0.026576032862067223, + 'recon_loss': 0.13281585276126862, + 'predict_loss': 0.00618432043120265, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.177438884973526, + 'data_time': 0.000818300002720207, + 'model_time': 1.2085540789994411, + 'grad_norm_pre_clip_avg': 0.1742607519030571, + 'learning_rate': 6.657672753885591e-06, + 'epoch': 8.7} +04/19 [23:54:03] INFO | >> train_qwenlatent.py:487 + Step 34500 | grad_norm_pre_clip=0.1390 | + grad_norm_pre_clip_avg=0.1439 | Metrics: + {'align_loss': 0.025672703981399536, + 'recon_loss': 0.0970686748623848, + 'predict_loss': 0.009428642690181732, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1390029788017273, + 'mae_score': 0.008799973049679319, 'data_time': + 0.0007825490029063076, 'model_time': + 1.2299165389849804, 'grad_norm_pre_clip_avg': + 0.1439184308052063, 'learning_rate': + 6.649973859711192e-06, 'epoch': 8.71} +04/19 [23:54:16] INFO | >> train_qwenlatent.py:487 + Step 34510 | grad_norm_pre_clip=0.1645 | + grad_norm_pre_clip_avg=0.1570 | Metrics: + {'align_loss': 0.025485960766673088, + 'recon_loss': 0.09160483628511429, + 'predict_loss': 0.00494184996932745, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16447457671165466, + 'data_time': 0.0008804469835013151, + 'model_time': 1.2889329480240121, + 'grad_norm_pre_clip_avg': 0.1570453241467476, + 'learning_rate': 6.642277822860809e-06, + 'epoch': 8.71} +04/19 [23:54:29] INFO | >> train_qwenlatent.py:487 + Step 34520 | grad_norm_pre_clip=0.1731 | + grad_norm_pre_clip_avg=0.1722 | Metrics: + {'align_loss': 0.025487316772341728, + 'recon_loss': 0.11121001094579697, + 'predict_loss': 0.0086270272731781, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17307867109775543, + 'data_time': 0.000838933017803356, + 'model_time': 1.2751671450096183, + 'grad_norm_pre_clip_avg': 0.17220104932785035, + 'learning_rate': 6.63458464708541e-06, 'epoch': + 8.71} +04/19 [23:54:41] INFO | >> train_qwenlatent.py:487 + Step 34530 | grad_norm_pre_clip=0.1549 | + grad_norm_pre_clip_avg=0.1514 | Metrics: + {'align_loss': 0.024174019694328308, + 'recon_loss': 0.08752003312110901, + 'predict_loss': 0.008167648687958717, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15485240519046783, + 'data_time': 0.0008923469868022949, + 'model_time': 1.2053069829999004, + 'grad_norm_pre_clip_avg': 0.15136129334568976, + 'learning_rate': 6.626894336134545e-06, + 'epoch': 8.71} +04/19 [23:54:54] INFO | >> train_qwenlatent.py:487 + Step 34540 | grad_norm_pre_clip=0.1406 | + grad_norm_pre_clip_avg=0.1619 | Metrics: + {'align_loss': 0.02444778010249138, + 'recon_loss': 0.10427580028772354, + 'predict_loss': 0.00894580315798521, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1406136006116867, + 'data_time': 0.0011389580031391233, + 'model_time': 1.2267779450048693, + 'grad_norm_pre_clip_avg': 0.16186070144176484, + 'learning_rate': 6.619206893756384e-06, + 'epoch': 8.72} +04/19 [23:55:07] INFO | >> train_qwenlatent.py:487 + Step 34550 | grad_norm_pre_clip=0.1716 | + grad_norm_pre_clip_avg=0.1759 | Metrics: + {'align_loss': 0.02565464749932289, + 'recon_loss': 0.12882299721240997, + 'predict_loss': 0.010041114874184132, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1716282069683075, + 'mae_score': 0.007452491382220844, 'data_time': + 0.0007265410094987601, 'model_time': + 1.2471398749912623, 'grad_norm_pre_clip_avg': + 0.17590902447700502, 'learning_rate': + 6.61152232369769e-06, 'epoch': 8.72} +04/19 [23:55:20] INFO | >> train_qwenlatent.py:487 + Step 34560 | grad_norm_pre_clip=0.1499 | + grad_norm_pre_clip_avg=0.1721 | Metrics: + {'align_loss': 0.024879280477762222, + 'recon_loss': 0.06896065920591354, + 'predict_loss': 0.0034570107236504555, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14985817670822144, + 'data_time': 0.0008214800036512315, + 'model_time': 1.284104151010979, + 'grad_norm_pre_clip_avg': 0.17212023735046386, + 'learning_rate': 6.603840629703828e-06, + 'epoch': 8.72} +04/19 [23:55:33] INFO | >> train_qwenlatent.py:487 + Step 34570 | grad_norm_pre_clip=0.1258 | + grad_norm_pre_clip_avg=0.1414 | Metrics: + {'align_loss': 0.024563465267419815, + 'recon_loss': 0.11439064890146255, + 'predict_loss': 0.004714301787316799, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1257934421300888, + 'data_time': 0.0007061719952616841, + 'model_time': 1.2250541080138646, + 'grad_norm_pre_clip_avg': 0.1413561761379242, + 'learning_rate': 6.596161815518766e-06, + 'epoch': 8.72} +04/19 [23:55:45] INFO | >> train_qwenlatent.py:487 + Step 34580 | grad_norm_pre_clip=0.1765 | + grad_norm_pre_clip_avg=0.1301 | Metrics: + {'align_loss': 0.025342602282762527, + 'recon_loss': 0.1065719798207283, + 'predict_loss': 0.007404673844575882, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17651712894439697, + 'data_time': 0.0010055540187750012, + 'model_time': 1.3338133360084612, + 'grad_norm_pre_clip_avg': 0.13005984723567962, + 'learning_rate': 6.588485884885061e-06, + 'epoch': 8.73} +04/19 [23:55:59] INFO | >> train_qwenlatent.py:487 + Step 34590 | grad_norm_pre_clip=0.1923 | + grad_norm_pre_clip_avg=0.1669 | Metrics: + {'align_loss': 0.024490375071763992, + 'recon_loss': 0.11391783505678177, + 'predict_loss': 0.012370054610073566, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.192337304353714, + 'data_time': 0.0008119940175674856, + 'model_time': 1.2468074570060708, + 'grad_norm_pre_clip_avg': 0.16694713085889817, + 'learning_rate': 6.58081284154387e-06, 'epoch': + 8.73} +04/19 [23:56:12] INFO | >> train_qwenlatent.py:487 + Step 34600 | grad_norm_pre_clip=0.1639 | + grad_norm_pre_clip_avg=0.1955 | Metrics: + {'align_loss': 0.02590044215321541, + 'recon_loss': 0.11170337349176407, + 'predict_loss': 0.014330697245895863, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16387242078781128, + 'mae_score': 0.006926355705604897, 'data_time': + 0.0007760110020171851, 'model_time': + 1.2541368350211997, 'grad_norm_pre_clip_avg': + 0.19553650617599488, 'learning_rate': + 6.573142689234937e-06, 'epoch': 8.73} +04/19 [23:56:24] INFO | >> train_qwenlatent.py:487 + Step 34610 | grad_norm_pre_clip=0.1747 | + grad_norm_pre_clip_avg=0.1753 | Metrics: + {'align_loss': 0.026186907663941383, + 'recon_loss': 0.13888071477413177, + 'predict_loss': 0.009601978585124016, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17468170821666718, + 'data_time': 0.0009248510177712888, + 'model_time': 1.204393142979825, + 'grad_norm_pre_clip_avg': 0.17529400885105134, + 'learning_rate': 6.565475431696609e-06, + 'epoch': 8.73} +04/19 [23:56:37] INFO | >> train_qwenlatent.py:487 + Step 34620 | grad_norm_pre_clip=0.1585 | + grad_norm_pre_clip_avg=0.1708 | Metrics: + {'align_loss': 0.025042235851287842, + 'recon_loss': 0.12331928312778473, + 'predict_loss': 0.007956715300679207, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.158494770526886, + 'data_time': 0.0008769070263952017, + 'model_time': 1.251026134006679, + 'grad_norm_pre_clip_avg': 0.170839262008667, + 'learning_rate': 6.557811072665811e-06, + 'epoch': 8.74} +04/19 [23:56:50] INFO | >> train_qwenlatent.py:487 + Step 34630 | grad_norm_pre_clip=0.1989 | + grad_norm_pre_clip_avg=0.1488 | Metrics: + {'align_loss': 0.023102808743715286, + 'recon_loss': 0.09921678155660629, + 'predict_loss': 0.008075595833361149, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1989385485649109, + 'data_time': 0.0010920630011241883, + 'model_time': 1.2930965019913856, + 'grad_norm_pre_clip_avg': 0.14881107732653617, + 'learning_rate': 6.550149615878057e-06, + 'epoch': 8.74} +04/19 [23:57:02] INFO | >> train_qwenlatent.py:487 + Step 34640 | grad_norm_pre_clip=0.1834 | + grad_norm_pre_clip_avg=0.1706 | Metrics: + {'align_loss': 0.02624085731804371, + 'recon_loss': 0.12552957236766815, + 'predict_loss': 0.010104878805577755, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18340492248535156, + 'data_time': 0.0007214600045699626, + 'model_time': 1.2562575089978054, + 'grad_norm_pre_clip_avg': 0.1706368513405323, + 'learning_rate': 6.542491065067451e-06, + 'epoch': 8.74} +04/19 [23:57:15] INFO | >> train_qwenlatent.py:487 + Step 34650 | grad_norm_pre_clip=0.1499 | + grad_norm_pre_clip_avg=0.1780 | Metrics: + {'align_loss': 0.025735560804605484, + 'recon_loss': 0.11021272093057632, + 'predict_loss': 0.004850498400628567, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14986656606197357, + 'mae_score': 0.007447526261613176, 'data_time': + 0.0006764010176993906, 'model_time': + 1.2434865560207982, 'grad_norm_pre_clip_avg': + 0.17804303616285325, 'learning_rate': + 6.534835423966673e-06, 'epoch': 8.74} +04/19 [23:57:28] INFO | >> train_qwenlatent.py:487 + Step 34660 | grad_norm_pre_clip=0.2037 | + grad_norm_pre_clip_avg=0.1897 | Metrics: + {'align_loss': 0.025634024292230606, + 'recon_loss': 0.11298748105764389, + 'predict_loss': 0.006475331261754036, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20373791456222534, + 'data_time': 0.0006010269862599671, + 'model_time': 1.3883337230072357, + 'grad_norm_pre_clip_avg': 0.18973162025213242, + 'learning_rate': 6.5271826963069926e-06, + 'epoch': 8.75} +04/19 [23:57:39] INFO | >> train_qwenlatent.py:487 + Step 34670 | grad_norm_pre_clip=0.1603 | + grad_norm_pre_clip_avg=0.1737 | Metrics: + {'align_loss': 0.025052309036254883, + 'recon_loss': 0.07316489517688751, + 'predict_loss': 0.004968044348061085, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16026879847049713, + 'data_time': 0.0006402930011972785, + 'model_time': 1.1798534909903537, + 'grad_norm_pre_clip_avg': 0.1736731633543968, + 'learning_rate': 6.5195328858182565e-06, + 'epoch': 8.75} +04/19 [23:57:51] INFO | >> train_qwenlatent.py:487 + Step 34680 | grad_norm_pre_clip=0.1599 | + grad_norm_pre_clip_avg=0.1627 | Metrics: + {'align_loss': 0.02546253614127636, + 'recon_loss': 0.08883161842823029, + 'predict_loss': 0.004454221576452255, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15993444621562958, + 'data_time': 0.0006434250099118799, + 'model_time': 1.177998358005425, + 'grad_norm_pre_clip_avg': 0.16274768114089966, + 'learning_rate': 6.511885996228887e-06, + 'epoch': 8.75} +04/19 [23:58:03] INFO | >> train_qwenlatent.py:487 + Step 34690 | grad_norm_pre_clip=0.1504 | + grad_norm_pre_clip_avg=0.1699 | Metrics: + {'align_loss': 0.025687437504529953, + 'recon_loss': 0.10527323186397552, + 'predict_loss': 0.007579536642879248, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15038934350013733, + 'data_time': 0.0005973890074528754, + 'model_time': 1.170748783973977, + 'grad_norm_pre_clip_avg': 0.16985285729169847, + 'learning_rate': 6.504242031265889e-06, + 'epoch': 8.75} +04/19 [23:58:15] INFO | >> train_qwenlatent.py:487 + Step 34700 | grad_norm_pre_clip=0.1475 | + grad_norm_pre_clip_avg=0.1665 | Metrics: + {'align_loss': 0.025582391768693924, + 'recon_loss': 0.14468009769916534, + 'predict_loss': 0.007644534111022949, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1475093811750412, + 'mae_score': 0.007613088418771554, 'data_time': + 0.0005823939864058048, 'model_time': + 1.1807273869926576, 'grad_norm_pre_clip_avg': + 0.16649608090519905, 'learning_rate': + 6.496600994654829e-06, 'epoch': 8.76} +04/19 [23:58:27] INFO | >> train_qwenlatent.py:487 + Step 34710 | grad_norm_pre_clip=0.1983 | + grad_norm_pre_clip_avg=0.1756 | Metrics: + {'align_loss': 0.025764675810933113, + 'recon_loss': 0.07904301583766937, + 'predict_loss': 0.004025271162390709, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19830536842346191, + 'data_time': 0.0006557909946423024, + 'model_time': 1.142686054983642, + 'grad_norm_pre_clip_avg': 0.17558264583349228, + 'learning_rate': 6.488962890119867e-06, + 'epoch': 8.76} +04/19 [23:58:39] INFO | >> train_qwenlatent.py:487 + Step 34720 | grad_norm_pre_clip=0.1567 | + grad_norm_pre_clip_avg=0.1724 | Metrics: + {'align_loss': 0.024854812771081924, + 'recon_loss': 0.08511234819889069, + 'predict_loss': 0.006995802279561758, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15671050548553467, + 'data_time': 0.0006375380035024136, + 'model_time': 1.1413534470193554, + 'grad_norm_pre_clip_avg': 0.17240875512361525, + 'learning_rate': 6.481327721383725e-06, + 'epoch': 8.76} +04/19 [23:58:50] INFO | >> train_qwenlatent.py:487 + Step 34730 | grad_norm_pre_clip=0.1992 | + grad_norm_pre_clip_avg=0.1644 | Metrics: + {'align_loss': 0.025975942611694336, + 'recon_loss': 0.1165611520409584, + 'predict_loss': 0.008083284832537174, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19922779500484467, + 'data_time': 0.000635623000562191, + 'model_time': 1.1490743120084517, + 'grad_norm_pre_clip_avg': 0.1643744945526123, + 'learning_rate': 6.473695492167677e-06, + 'epoch': 8.76} +04/19 [23:59:02] INFO | >> train_qwenlatent.py:487 + Step 34740 | grad_norm_pre_clip=0.1627 | + grad_norm_pre_clip_avg=0.1632 | Metrics: + {'align_loss': 0.02568642422556877, + 'recon_loss': 0.10583022236824036, + 'predict_loss': 0.007286541163921356, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1627236008644104, + 'data_time': 0.0006012769881635904, + 'model_time': 1.1408785769890528, + 'grad_norm_pre_clip_avg': 0.16322807893157004, + 'learning_rate': 6.466066206191579e-06, + 'epoch': 8.77} +04/19 [23:59:14] INFO | >> train_qwenlatent.py:487 + Step 34750 | grad_norm_pre_clip=0.2099 | + grad_norm_pre_clip_avg=0.1731 | Metrics: + {'align_loss': 0.02478211745619774, + 'recon_loss': 0.12266582250595093, + 'predict_loss': 0.009248609654605389, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20993013679981232, + 'mae_score': 0.006778347170030749, 'data_time': + 0.0005731080018449575, 'model_time': + 1.1601490699977148, 'grad_norm_pre_clip_avg': + 0.17308458387851716, 'learning_rate': + 6.458439867173867e-06, 'epoch': 8.77} +04/19 [23:59:26] INFO | >> train_qwenlatent.py:487 + Step 34760 | grad_norm_pre_clip=0.1922 | + grad_norm_pre_clip_avg=0.1698 | Metrics: + {'align_loss': 0.024624723941087723, + 'recon_loss': 0.10807404667139053, + 'predict_loss': 0.011996672488749027, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19223101437091827, + 'data_time': 0.0006121099868323654, + 'model_time': 1.16627007999341, + 'grad_norm_pre_clip_avg': 0.16978248357772827, + 'learning_rate': 6.450816478831517e-06, + 'epoch': 8.77} +04/19 [23:59:37] INFO | >> train_qwenlatent.py:487 + Step 34770 | grad_norm_pre_clip=0.1806 | + grad_norm_pre_clip_avg=0.1859 | Metrics: + {'align_loss': 0.026008853688836098, + 'recon_loss': 0.11301916837692261, + 'predict_loss': 0.005357105750590563, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18063481152057648, + 'data_time': 0.0006133029819466174, + 'model_time': 1.1542306179762818, + 'grad_norm_pre_clip_avg': 0.18594559878110886, + 'learning_rate': 6.4431960448800774e-06, + 'epoch': 8.77} +04/19 [23:59:49] INFO | >> train_qwenlatent.py:487 + Step 34780 | grad_norm_pre_clip=0.1506 | + grad_norm_pre_clip_avg=0.1796 | Metrics: + {'align_loss': 0.024630414322018623, + 'recon_loss': 0.07521709054708481, + 'predict_loss': 0.003575471229851246, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15055498480796814, + 'data_time': 0.0006165189843159169, + 'model_time': 1.157968890009215, + 'grad_norm_pre_clip_avg': 0.179571034014225, + 'learning_rate': 6.4355785690336555e-06, + 'epoch': 8.78} +04/20 [00:00:01] INFO | >> train_qwenlatent.py:487 + Step 34790 | grad_norm_pre_clip=0.1459 | + grad_norm_pre_clip_avg=0.1802 | Metrics: + {'align_loss': 0.02608237788081169, + 'recon_loss': 0.10442444682121277, + 'predict_loss': 0.007008489686995745, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14592532813549042, + 'data_time': 0.0005668989906553179, + 'model_time': 1.1470457810210064, + 'grad_norm_pre_clip_avg': 0.18021869361400605, + 'learning_rate': 6.427964055004911e-06, + 'epoch': 8.78} +04/20 [00:00:13] INFO | >> train_qwenlatent.py:487 + Step 34800 | grad_norm_pre_clip=0.1341 | + grad_norm_pre_clip_avg=0.1583 | Metrics: + {'align_loss': 0.02564910426735878, + 'recon_loss': 0.1020897775888443, + 'predict_loss': 0.004966431763023138, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13408523797988892, + 'mae_score': 0.007976649688170837, 'data_time': + 0.0006369120092131197, 'model_time': + 1.1400978220044635, 'grad_norm_pre_clip_avg': + 0.1582945168018341, 'learning_rate': + 6.420352506505075e-06, 'epoch': 8.78} +04/20 [00:00:24] INFO | >> train_qwenlatent.py:487 + Step 34810 | grad_norm_pre_clip=0.2160 | + grad_norm_pre_clip_avg=0.1596 | Metrics: + {'align_loss': 0.025954417884349823, + 'recon_loss': 0.11366604268550873, + 'predict_loss': 0.007925242185592651, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2160424143075943, + 'data_time': 0.0006109679816290736, + 'model_time': 1.1650855079933535, + 'grad_norm_pre_clip_avg': 0.15959564447402955, + 'learning_rate': 6.412743927243924e-06, + 'epoch': 8.78} +04/20 [00:00:36] INFO | >> train_qwenlatent.py:487 + Step 34820 | grad_norm_pre_clip=0.1785 | + grad_norm_pre_clip_avg=0.1916 | Metrics: + {'align_loss': 0.02425459772348404, + 'recon_loss': 0.10958334803581238, + 'predict_loss': 0.0071876151487231255, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17854183912277222, + 'data_time': 0.0005888659798074514, + 'model_time': 1.1558345959929284, + 'grad_norm_pre_clip_avg': 0.19155534654855727, + 'learning_rate': 6.405138320929777e-06, + 'epoch': 8.79} +04/20 [00:00:47] INFO | >> train_qwenlatent.py:487 + Step 34830 | grad_norm_pre_clip=0.2141 | + grad_norm_pre_clip_avg=0.1652 | Metrics: + {'align_loss': 0.02556789107620716, + 'recon_loss': 0.11135302484035492, + 'predict_loss': 0.005278579890727997, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21412713825702667, + 'data_time': 0.0005838520010001957, + 'model_time': 1.1447995640046429, + 'grad_norm_pre_clip_avg': 0.16515840142965316, + 'learning_rate': 6.397535691269517e-06, + 'epoch': 8.79} +04/20 [00:00:59] INFO | >> train_qwenlatent.py:487 + Step 34840 | grad_norm_pre_clip=0.1795 | + grad_norm_pre_clip_avg=0.1495 | Metrics: + {'align_loss': 0.025034360587596893, + 'recon_loss': 0.10686412453651428, + 'predict_loss': 0.0084999306127429, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1794508993625641, + 'data_time': 0.0006728339940309525, + 'model_time': 1.1716201110102702, + 'grad_norm_pre_clip_avg': 0.1495450720191002, + 'learning_rate': 6.389936041968574e-06, + 'epoch': 8.79} +04/20 [00:01:11] INFO | >> train_qwenlatent.py:487 + Step 34850 | grad_norm_pre_clip=0.1548 | + grad_norm_pre_clip_avg=0.1718 | Metrics: + {'align_loss': 0.022968869656324387, + 'recon_loss': 0.11214540153741837, + 'predict_loss': 0.007183907553553581, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15478675067424774, + 'mae_score': 0.0076125626091484554, + 'data_time': 0.0006275310006458312, + 'model_time': 1.1726442530052736, + 'grad_norm_pre_clip_avg': 0.17178056240081788, + 'learning_rate': 6.38233937673093e-06, 'epoch': + 8.79} +04/20 [00:01:23] INFO | >> train_qwenlatent.py:487 + Step 34860 | grad_norm_pre_clip=0.1267 | + grad_norm_pre_clip_avg=0.1692 | Metrics: + {'align_loss': 0.023517128080129623, + 'recon_loss': 0.08670662343502045, + 'predict_loss': 0.0055264560505747795, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1266864538192749, + 'data_time': 0.0006082549807615578, + 'model_time': 1.1622010299761314, + 'grad_norm_pre_clip_avg': 0.16922292709350586, + 'learning_rate': 6.374745699259103e-06, + 'epoch': 8.8} +04/20 [00:01:35] INFO | >> train_qwenlatent.py:487 + Step 34870 | grad_norm_pre_clip=0.1937 | + grad_norm_pre_clip_avg=0.1688 | Metrics: + {'align_loss': 0.025499917566776276, + 'recon_loss': 0.08661235868930817, + 'predict_loss': 0.005268212873488665, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19371286034584045, + 'data_time': 0.0005635590059682727, + 'model_time': 1.3171829779748805, + 'grad_norm_pre_clip_avg': 0.16880312338471412, + 'learning_rate': 6.3671550132541615e-06, + 'epoch': 8.8} +04/20 [00:01:46] INFO | >> train_qwenlatent.py:487 + Step 34880 | grad_norm_pre_clip=0.1873 | + grad_norm_pre_clip_avg=0.1577 | Metrics: + {'align_loss': 0.025257684290409088, + 'recon_loss': 0.09814934432506561, + 'predict_loss': 0.005520459730178118, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18732234835624695, + 'data_time': 0.000512166996486485, + 'model_time': 1.1470533199899364, + 'grad_norm_pre_clip_avg': 0.15770838260650635, + 'learning_rate': 6.35956732241571e-06, 'epoch': + 8.8} +04/20 [00:01:58] INFO | >> train_qwenlatent.py:487 + Step 34890 | grad_norm_pre_clip=0.2412 | + grad_norm_pre_clip_avg=0.1746 | Metrics: + {'align_loss': 0.023971406742930412, + 'recon_loss': 0.1446572244167328, + 'predict_loss': 0.012688593938946724, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24117588996887207, + 'data_time': 0.0005665739881806076, + 'model_time': 1.2960426839999855, + 'grad_norm_pre_clip_avg': 0.1745573326945305, + 'learning_rate': 6.3519826304418965e-06, + 'epoch': 8.8} +04/20 [00:02:10] INFO | >> train_qwenlatent.py:487 + Step 34900 | grad_norm_pre_clip=0.1133 | + grad_norm_pre_clip_avg=0.1558 | Metrics: + {'align_loss': 0.024615325033664703, + 'recon_loss': 0.10178229957818985, + 'predict_loss': 0.0069246250204741955, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11325833201408386, + 'mae_score': 0.006701402836017781, 'data_time': + 0.0005599879950750619, 'model_time': + 1.158716342994012, 'grad_norm_pre_clip_avg': + 0.1557706892490387, 'learning_rate': + 6.344400941029419e-06, 'epoch': 8.81} +04/20 [00:02:22] INFO | >> train_qwenlatent.py:487 + Step 34910 | grad_norm_pre_clip=0.1431 | + grad_norm_pre_clip_avg=0.1539 | Metrics: + {'align_loss': 0.02509366348385811, + 'recon_loss': 0.12333005666732788, + 'predict_loss': 0.009564257226884365, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1430579274892807, + 'data_time': 0.000564664980629459, + 'model_time': 1.1388316139928065, + 'grad_norm_pre_clip_avg': 0.15388011187314987, + 'learning_rate': 6.3368222578734856e-06, + 'epoch': 8.81} +04/20 [00:02:33] INFO | >> train_qwenlatent.py:487 + Step 34920 | grad_norm_pre_clip=0.1381 | + grad_norm_pre_clip_avg=0.1684 | Metrics: + {'align_loss': 0.025032687932252884, + 'recon_loss': 0.13532255589962006, + 'predict_loss': 0.008635712787508965, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1381027102470398, + 'data_time': 0.0007010929984971881, + 'model_time': 1.169712138012983, + 'grad_norm_pre_clip_avg': 0.16838523447513581, + 'learning_rate': 6.3292465846678625e-06, + 'epoch': 8.81} +04/20 [00:02:45] INFO | >> train_qwenlatent.py:487 + Step 34930 | grad_norm_pre_clip=0.1387 | + grad_norm_pre_clip_avg=0.1487 | Metrics: + {'align_loss': 0.024586016312241554, + 'recon_loss': 0.08947063982486725, + 'predict_loss': 0.0037074536085128784, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13870033621788025, + 'data_time': 0.0006054179975762963, + 'model_time': 1.3403501689899713, + 'grad_norm_pre_clip_avg': 0.1487217664718628, + 'learning_rate': 6.321673925104835e-06, + 'epoch': 8.81} +04/20 [00:02:56] INFO | >> train_qwenlatent.py:487 + Step 34940 | grad_norm_pre_clip=0.1987 | + grad_norm_pre_clip_avg=0.1682 | Metrics: + {'align_loss': 0.025200510397553444, + 'recon_loss': 0.07996633648872375, + 'predict_loss': 0.00528257479891181, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19865238666534424, + 'data_time': 0.0005636850255541503, + 'model_time': 1.1348100189934485, + 'grad_norm_pre_clip_avg': 0.16820905804634095, + 'learning_rate': 6.314104282875232e-06, + 'epoch': 8.82} +04/20 [00:03:08] INFO | >> train_qwenlatent.py:487 + Step 34950 | grad_norm_pre_clip=0.1690 | + grad_norm_pre_clip_avg=0.1920 | Metrics: + {'align_loss': 0.026829788461327553, + 'recon_loss': 0.1687135249376297, + 'predict_loss': 0.00827889796346426, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16900645196437836, + 'mae_score': 0.008234667563223623, 'data_time': + 0.0005764320085290819, 'model_time': + 1.14760529101477, 'grad_norm_pre_clip_avg': + 0.19195657968521118, 'learning_rate': + 6.306537661668404e-06, 'epoch': 8.82} +04/20 [00:03:20] INFO | >> train_qwenlatent.py:487 + Step 34960 | grad_norm_pre_clip=0.2023 | + grad_norm_pre_clip_avg=0.1570 | Metrics: + {'align_loss': 0.025400269776582718, + 'recon_loss': 0.11521267145872116, + 'predict_loss': 0.007008268032222986, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20228707790374756, + 'data_time': 0.0005681360198650509, + 'model_time': 1.1412341529794503, + 'grad_norm_pre_clip_avg': 0.1569509468972683, + 'learning_rate': 6.298974065172227e-06, + 'epoch': 8.82} +04/20 [00:03:56] INFO | >> train_qwenlatent.py:487 + Step 34970 | grad_norm_pre_clip=0.1253 | + grad_norm_pre_clip_avg=0.1590 | Metrics: + {'align_loss': 0.025856951251626015, + 'recon_loss': 0.10214070230722427, + 'predict_loss': 0.005476376507431269, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12526899576187134, + 'data_time': 0.001172110001789406, + 'model_time': 3.2881516909983475, + 'grad_norm_pre_clip_avg': 0.15904503017663957, + 'learning_rate': 6.291413497073109e-06, + 'epoch': 8.82} +04/20 [00:04:31] INFO | >> train_qwenlatent.py:487 + Step 34980 | grad_norm_pre_clip=0.1846 | + grad_norm_pre_clip_avg=0.1837 | Metrics: + {'align_loss': 0.025390448048710823, + 'recon_loss': 0.07314986735582352, + 'predict_loss': 0.0062897889874875546, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18456807732582092, + 'data_time': 0.009230914001818746, + 'model_time': 3.018431712000165, + 'grad_norm_pre_clip_avg': 0.18367598503828048, + 'learning_rate': 6.283855961055979e-06, + 'epoch': 8.83} +04/20 [00:05:06] INFO | >> train_qwenlatent.py:487 + Step 34990 | grad_norm_pre_clip=0.1882 | + grad_norm_pre_clip_avg=0.1599 | Metrics: + {'align_loss': 0.025485016405582428, + 'recon_loss': 0.10874102264642715, + 'predict_loss': 0.013027752749621868, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1881532520055771, + 'data_time': 0.0012467690103221685, + 'model_time': 3.1777481779863592, + 'grad_norm_pre_clip_avg': 0.15986848548054694, + 'learning_rate': 6.276301460804286e-06, + 'epoch': 8.83} +04/20 [00:05:44] INFO | >> train_qwenlatent.py:487 + Step 35000 | grad_norm_pre_clip=0.1640 | + grad_norm_pre_clip_avg=0.1544 | Metrics: + {'align_loss': 0.025686882436275482, + 'recon_loss': 0.15526947379112244, + 'predict_loss': 0.011920721270143986, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16401879489421844, + 'mae_score': 0.006348148122564092, 'data_time': + 0.001120509987231344, 'model_time': + 3.864055596990511, 'grad_norm_pre_clip_avg': + 0.15436377227306367, 'learning_rate': + 6.2687500000000024e-06, 'epoch': 8.83} +✅ Checkpoint saved at ./runs/0418_QwenLatent_13tasks_actionstate_30k/checkpoints/steps_35000 +04/20 [00:06:56] INFO | >> train_qwenlatent.py:487 + Step 35010 | grad_norm_pre_clip=0.2060 | + grad_norm_pre_clip_avg=0.2124 | Metrics: + {'align_loss': 0.02501949854195118, + 'recon_loss': 0.14955249428749084, + 'predict_loss': 0.009760241955518723, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20603661239147186, + 'data_time': 0.0011782560031861067, + 'model_time': 3.3849923509988002, + 'grad_norm_pre_clip_avg': 0.21239591166377067, + 'learning_rate': 6.261201582323619e-06, + 'epoch': 8.83} +04/20 [00:07:31] INFO | >> train_qwenlatent.py:487 + Step 35020 | grad_norm_pre_clip=0.1681 | + grad_norm_pre_clip_avg=0.1619 | Metrics: + {'align_loss': 0.02447289228439331, + 'recon_loss': 0.08433780074119568, + 'predict_loss': 0.005813897587358952, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16810393333435059, + 'data_time': 0.00118545102304779, 'model_time': + 3.1981675739807542, 'grad_norm_pre_clip_avg': + 0.16185450404882432, 'learning_rate': + 6.253656211454144e-06, 'epoch': 8.84} +04/20 [00:07:56] INFO | >> train_qwenlatent.py:487 + Step 35030 | grad_norm_pre_clip=0.1692 | + grad_norm_pre_clip_avg=0.1488 | Metrics: + {'align_loss': 0.024990901350975037, + 'recon_loss': 0.11677993088960648, + 'predict_loss': 0.00863757636398077, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16923929750919342, + 'data_time': 0.001088190998416394, + 'model_time': 2.8041067339945585, + 'grad_norm_pre_clip_avg': 0.14879090785980226, + 'learning_rate': 6.246113891069092e-06, + 'epoch': 8.84} +04/20 [00:08:15] INFO | >> train_qwenlatent.py:487 + Step 35040 | grad_norm_pre_clip=0.1310 | + grad_norm_pre_clip_avg=0.1602 | Metrics: + {'align_loss': 0.025464700534939766, + 'recon_loss': 0.1231129989027977, + 'predict_loss': 0.006385435815900564, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13095144927501678, + 'data_time': 0.0011752409918699414, + 'model_time': 1.6553246260154992, + 'grad_norm_pre_clip_avg': 0.16017773747444153, + 'learning_rate': 6.238574624844511e-06, + 'epoch': 8.84} +04/20 [00:08:30] INFO | >> train_qwenlatent.py:487 + Step 35050 | grad_norm_pre_clip=0.1702 | + grad_norm_pre_clip_avg=0.1474 | Metrics: + {'align_loss': 0.025564946234226227, + 'recon_loss': 0.12854477763175964, + 'predict_loss': 0.010414219461381435, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1702415943145752, + 'mae_score': 0.007660529849765537, 'data_time': + 0.0005983229784760624, 'model_time': + 1.2170206290029455, 'grad_norm_pre_clip_avg': + 0.14739550203084945, 'learning_rate': + 6.2310384164549394e-06, 'epoch': 8.84} +04/20 [00:08:43] INFO | >> train_qwenlatent.py:487 + Step 35060 | grad_norm_pre_clip=0.1587 | + grad_norm_pre_clip_avg=0.1666 | Metrics: + {'align_loss': 0.02505764737725258, + 'recon_loss': 0.08747555315494537, + 'predict_loss': 0.004419198725372553, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15866532921791077, + 'data_time': 0.0009714890038594604, + 'model_time': 1.2789675740059465, + 'grad_norm_pre_clip_avg': 0.16656685173511504, + 'learning_rate': 6.223505269573437e-06, + 'epoch': 8.85} +04/20 [00:08:55] INFO | >> train_qwenlatent.py:487 + Step 35070 | grad_norm_pre_clip=0.1623 | + grad_norm_pre_clip_avg=0.1809 | Metrics: + {'align_loss': 0.024775320664048195, + 'recon_loss': 0.0747416689991951, + 'predict_loss': 0.008155166171491146, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16226448118686676, + 'data_time': 0.0008594100072514266, + 'model_time': 1.2152445039828308, + 'grad_norm_pre_clip_avg': 0.18094318956136704, + 'learning_rate': 6.215975187871567e-06, + 'epoch': 8.85} +04/20 [00:09:08] INFO | >> train_qwenlatent.py:487 + Step 35080 | grad_norm_pre_clip=0.1776 | + grad_norm_pre_clip_avg=0.1590 | Metrics: + {'align_loss': 0.02575599029660225, + 'recon_loss': 0.12861284613609314, + 'predict_loss': 0.005671063903719187, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17763105034828186, + 'data_time': 0.0005770089919678867, + 'model_time': 1.2875961560057476, + 'grad_norm_pre_clip_avg': 0.15895010977983476, + 'learning_rate': 6.208448175019398e-06, + 'epoch': 8.85} +04/20 [00:09:20] INFO | >> train_qwenlatent.py:487 + Step 35090 | grad_norm_pre_clip=0.1879 | + grad_norm_pre_clip_avg=0.1853 | Metrics: + {'align_loss': 0.026407012715935707, + 'recon_loss': 0.11970482021570206, + 'predict_loss': 0.005684876814484596, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1879042536020279, + 'data_time': 0.0006066530186217278, + 'model_time': 1.2246385460020974, + 'grad_norm_pre_clip_avg': 0.18528324961662293, + 'learning_rate': 6.200924234685507e-06, + 'epoch': 8.85} +04/20 [00:09:34] INFO | >> train_qwenlatent.py:487 + Step 35100 | grad_norm_pre_clip=0.1799 | + grad_norm_pre_clip_avg=0.1792 | Metrics: + {'align_loss': 0.026401208713650703, + 'recon_loss': 0.13840404152870178, + 'predict_loss': 0.012170563451945782, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17987169325351715, + 'mae_score': 0.00742472743128871, 'data_time': + 0.0008569929923396558, 'model_time': + 1.2448560160119087, 'grad_norm_pre_clip_avg': + 0.17918479442596436, 'learning_rate': + 6.19340337053697e-06, 'epoch': 8.86} +04/20 [00:09:46] INFO | >> train_qwenlatent.py:487 + Step 35110 | grad_norm_pre_clip=0.1526 | + grad_norm_pre_clip_avg=0.1647 | Metrics: + {'align_loss': 0.023680292069911957, + 'recon_loss': 0.0950823575258255, + 'predict_loss': 0.007714018225669861, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1526373326778412, + 'data_time': 0.0006561409973073751, + 'model_time': 1.2294264730007853, + 'grad_norm_pre_clip_avg': 0.164743672311306, + 'learning_rate': 6.185885586239365e-06, + 'epoch': 8.86} +04/20 [00:10:00] INFO | >> train_qwenlatent.py:487 + Step 35120 | grad_norm_pre_clip=0.1530 | + grad_norm_pre_clip_avg=0.1512 | Metrics: + {'align_loss': 0.0249762125313282, + 'recon_loss': 0.15127041935920715, + 'predict_loss': 0.007995450869202614, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15300221741199493, + 'data_time': 0.0007550480077043176, + 'model_time': 1.1929211410169955, + 'grad_norm_pre_clip_avg': 0.15117636919021607, + 'learning_rate': 6.1783708854567635e-06, + 'epoch': 8.86} +04/20 [00:10:12] INFO | >> train_qwenlatent.py:487 + Step 35130 | grad_norm_pre_clip=0.1346 | + grad_norm_pre_clip_avg=0.1555 | Metrics: + {'align_loss': 0.024463487789034843, + 'recon_loss': 0.09998401254415512, + 'predict_loss': 0.008686714805662632, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13463866710662842, + 'data_time': 0.0005587739869952202, + 'model_time': 1.214989936008351, + 'grad_norm_pre_clip_avg': 0.1555000886321068, + 'learning_rate': 6.170859271851749e-06, + 'epoch': 8.86} +04/20 [00:10:25] INFO | >> train_qwenlatent.py:487 + Step 35140 | grad_norm_pre_clip=0.1624 | + grad_norm_pre_clip_avg=0.1684 | Metrics: + {'align_loss': 0.026037711650133133, + 'recon_loss': 0.12038154155015945, + 'predict_loss': 0.009140563197433949, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16235056519508362, + 'data_time': 0.0007083470118232071, + 'model_time': 1.2356773499923293, + 'grad_norm_pre_clip_avg': 0.16842066943645478, + 'learning_rate': 6.163350749085386e-06, + 'epoch': 8.87} +04/20 [00:10:38] INFO | >> train_qwenlatent.py:487 + Step 35150 | grad_norm_pre_clip=0.1564 | + grad_norm_pre_clip_avg=0.1792 | Metrics: + {'align_loss': 0.024908047169446945, + 'recon_loss': 0.099561907351017, + 'predict_loss': 0.005832594819366932, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15635164082050323, + 'mae_score': 0.007496715236354519, 'data_time': + 0.0006033450190443546, 'model_time': + 1.2115207680035383, 'grad_norm_pre_clip_avg': + 0.17923212051391602, 'learning_rate': + 6.155845320817238e-06, 'epoch': 8.87} +04/20 [00:10:50] INFO | >> train_qwenlatent.py:487 + Step 35160 | grad_norm_pre_clip=0.1328 | + grad_norm_pre_clip_avg=0.1594 | Metrics: + {'align_loss': 0.02526332251727581, + 'recon_loss': 0.12801747024059296, + 'predict_loss': 0.010372529737651348, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13278090953826904, + 'data_time': 0.0009020639990922064, + 'model_time': 1.2779581869835965, + 'grad_norm_pre_clip_avg': 0.15944493561983109, + 'learning_rate': 6.14834299070536e-06, 'epoch': + 8.87} +04/20 [00:11:03] INFO | >> train_qwenlatent.py:487 + Step 35170 | grad_norm_pre_clip=0.1465 | + grad_norm_pre_clip_avg=0.1834 | Metrics: + {'align_loss': 0.025038449093699455, + 'recon_loss': 0.1483185887336731, + 'predict_loss': 0.010824523866176605, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14649611711502075, + 'data_time': 0.000898561003850773, + 'model_time': 1.2323647509911098, + 'grad_norm_pre_clip_avg': 0.18336372077465057, + 'learning_rate': 6.140843762406295e-06, + 'epoch': 8.87} +04/20 [00:11:16] INFO | >> train_qwenlatent.py:487 + Step 35180 | grad_norm_pre_clip=0.1947 | + grad_norm_pre_clip_avg=0.1732 | Metrics: + {'align_loss': 0.02547495625913143, + 'recon_loss': 0.10084475576877594, + 'predict_loss': 0.0045774332247674465, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19467706978321075, + 'data_time': 0.0008334279991686344, + 'model_time': 1.5663640210113954, + 'grad_norm_pre_clip_avg': 0.17324160039424896, + 'learning_rate': 6.133347639575077e-06, + 'epoch': 8.88} +04/20 [00:11:28] INFO | >> train_qwenlatent.py:487 + Step 35190 | grad_norm_pre_clip=0.1793 | + grad_norm_pre_clip_avg=0.1665 | Metrics: + {'align_loss': 0.025852922350168228, + 'recon_loss': 0.1710481494665146, + 'predict_loss': 0.011184858158230782, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17934009432792664, + 'data_time': 0.0008718040189705789, + 'model_time': 1.294056143000489, + 'grad_norm_pre_clip_avg': 0.16648734882473945, + 'learning_rate': 6.125854625865224e-06, + 'epoch': 8.88} +04/20 [00:11:41] INFO | >> train_qwenlatent.py:487 + Step 35200 | grad_norm_pre_clip=0.1600 | + grad_norm_pre_clip_avg=0.1632 | Metrics: + {'align_loss': 0.025647640228271484, + 'recon_loss': 0.10307151079177856, + 'predict_loss': 0.004864381160587072, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15999579429626465, + 'mae_score': 0.006433870985701277, 'data_time': + 0.0007433519931510091, 'model_time': + 1.294605016009882, 'grad_norm_pre_clip_avg': + 0.16318758875131606, 'learning_rate': + 6.11836472492874e-06, 'epoch': 8.88} +04/20 [00:11:54] INFO | >> train_qwenlatent.py:487 + Step 35210 | grad_norm_pre_clip=0.1459 | + grad_norm_pre_clip_avg=0.1695 | Metrics: + {'align_loss': 0.023653965443372726, + 'recon_loss': 0.06576033681631088, + 'predict_loss': 0.005469415802508593, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1458626687526703, + 'data_time': 0.0007095960027072579, + 'model_time': 1.2929617399931885, + 'grad_norm_pre_clip_avg': 0.16952775940299034, + 'learning_rate': 6.110877940416113e-06, + 'epoch': 8.88} +04/20 [00:12:07] INFO | >> train_qwenlatent.py:487 + Step 35220 | grad_norm_pre_clip=0.1677 | + grad_norm_pre_clip_avg=0.1856 | Metrics: + {'align_loss': 0.023857638239860535, + 'recon_loss': 0.11890479177236557, + 'predict_loss': 0.00844530202448368, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16772499680519104, + 'data_time': 0.0009301060053985566, + 'model_time': 1.357211713999277, + 'grad_norm_pre_clip_avg': 0.1856337606906891, + 'learning_rate': 6.103394275976306e-06, + 'epoch': 8.89} +04/20 [00:12:20] INFO | >> train_qwenlatent.py:487 + Step 35230 | grad_norm_pre_clip=0.1436 | + grad_norm_pre_clip_avg=0.1689 | Metrics: + {'align_loss': 0.025744885206222534, + 'recon_loss': 0.11391791701316833, + 'predict_loss': 0.005809302907437086, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14362499117851257, + 'data_time': 0.0006463250028900802, + 'model_time': 1.31913359099417, + 'grad_norm_pre_clip_avg': 0.16888242214918137, + 'learning_rate': 6.095913735256774e-06, + 'epoch': 8.89} +04/20 [00:12:32] INFO | >> train_qwenlatent.py:487 + Step 35240 | grad_norm_pre_clip=0.1404 | + grad_norm_pre_clip_avg=0.1400 | Metrics: + {'align_loss': 0.026129262521862984, + 'recon_loss': 0.14430710673332214, + 'predict_loss': 0.006792381405830383, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1404092162847519, + 'data_time': 0.0006293710030149668, + 'model_time': 1.4582046210125554, + 'grad_norm_pre_clip_avg': 0.1399673543870449, + 'learning_rate': 6.088436321903438e-06, + 'epoch': 8.89} +04/20 [00:12:46] INFO | >> train_qwenlatent.py:487 + Step 35250 | grad_norm_pre_clip=0.1225 | + grad_norm_pre_clip_avg=0.1525 | Metrics: + {'align_loss': 0.02586033195257187, + 'recon_loss': 0.11907706409692764, + 'predict_loss': 0.0046159736812114716, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12251308560371399, + 'mae_score': 0.007881051140862542, 'data_time': + 0.0005919960094615817, 'model_time': + 1.234891444008099, 'grad_norm_pre_clip_avg': + 0.15251938626170158, 'learning_rate': + 6.080962039560698e-06, 'epoch': 8.89} +04/20 [00:12:59] INFO | >> train_qwenlatent.py:487 + Step 35260 | grad_norm_pre_clip=0.1364 | + grad_norm_pre_clip_avg=0.1613 | Metrics: + {'align_loss': 0.026052091270685196, + 'recon_loss': 0.11172942072153091, + 'predict_loss': 0.006920970045030117, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13643698394298553, + 'data_time': 0.0007806690118741244, + 'model_time': 1.2126495740085375, + 'grad_norm_pre_clip_avg': 0.16132174134254457, + 'learning_rate': 6.073490891871429e-06, + 'epoch': 8.9} +04/20 [00:13:12] INFO | >> train_qwenlatent.py:487 + Step 35270 | grad_norm_pre_clip=0.2202 | + grad_norm_pre_clip_avg=0.1749 | Metrics: + {'align_loss': 0.024158351123332977, + 'recon_loss': 0.10852694511413574, + 'predict_loss': 0.00546249421313405, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22021274268627167, + 'data_time': 0.0005696160078514367, + 'model_time': 1.6255533779913094, + 'grad_norm_pre_clip_avg': 0.17493615448474883, + 'learning_rate': 6.066022882476978e-06, + 'epoch': 8.9} +04/20 [00:13:24] INFO | >> train_qwenlatent.py:487 + Step 35280 | grad_norm_pre_clip=0.2233 | + grad_norm_pre_clip_avg=0.1798 | Metrics: + {'align_loss': 0.024101007729768753, + 'recon_loss': 0.08655904233455658, + 'predict_loss': 0.007873096503317356, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2233254611492157, + 'data_time': 0.0006446889892686158, + 'model_time': 1.2763093779794872, + 'grad_norm_pre_clip_avg': 0.17977441102266312, + 'learning_rate': 6.05855801501716e-06, 'epoch': + 8.9} +04/20 [00:13:37] INFO | >> train_qwenlatent.py:487 + Step 35290 | grad_norm_pre_clip=0.1977 | + grad_norm_pre_clip_avg=0.1916 | Metrics: + {'align_loss': 0.026283446699380875, + 'recon_loss': 0.07657608389854431, + 'predict_loss': 0.003126685041934252, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19768522679805756, + 'data_time': 0.0007570419984403998, + 'model_time': 1.2622807160078082, + 'grad_norm_pre_clip_avg': 0.19163083285093307, + 'learning_rate': 6.051096293130263e-06, + 'epoch': 8.9} +04/20 [00:13:50] INFO | >> train_qwenlatent.py:487 + Step 35300 | grad_norm_pre_clip=0.2221 | + grad_norm_pre_clip_avg=0.1818 | Metrics: + {'align_loss': 0.026570327579975128, + 'recon_loss': 0.1357486993074417, + 'predict_loss': 0.009886612184345722, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22214004397392273, + 'mae_score': 0.006379377090179168, 'data_time': + 0.0008351849974133074, 'model_time': + 1.241337184997974, 'grad_norm_pre_clip_avg': + 0.18180440962314606, 'learning_rate': + 6.043637720453039e-06, 'epoch': 8.91} +04/20 [00:14:03] INFO | >> train_qwenlatent.py:487 + Step 35310 | grad_norm_pre_clip=0.1861 | + grad_norm_pre_clip_avg=0.1706 | Metrics: + {'align_loss': 0.02621716633439064, + 'recon_loss': 0.10659221559762955, + 'predict_loss': 0.004293984733521938, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18612641096115112, + 'data_time': 0.0006425599858630449, + 'model_time': 1.5290944799780846, + 'grad_norm_pre_clip_avg': 0.17060035318136216, + 'learning_rate': 6.036182300620705e-06, + 'epoch': 8.91} +04/20 [00:14:15] INFO | >> train_qwenlatent.py:487 + Step 35320 | grad_norm_pre_clip=0.1438 | + grad_norm_pre_clip_avg=0.1489 | Metrics: + {'align_loss': 0.02503870241343975, + 'recon_loss': 0.0839943140745163, + 'predict_loss': 0.008782006800174713, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14375245571136475, + 'data_time': 0.0009162639908026904, + 'model_time': 1.2403733319952153, + 'grad_norm_pre_clip_avg': 0.14890004247426986, + 'learning_rate': 6.028730037266939e-06, + 'epoch': 8.91} +04/20 [00:14:27] INFO | >> train_qwenlatent.py:487 + Step 35330 | grad_norm_pre_clip=0.1899 | + grad_norm_pre_clip_avg=0.1785 | Metrics: + {'align_loss': 0.025317668914794922, + 'recon_loss': 0.14503921568393707, + 'predict_loss': 0.010335743427276611, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18990705907344818, + 'data_time': 0.0007763860048726201, + 'model_time': 1.1935722189955413, + 'grad_norm_pre_clip_avg': 0.17849811762571335, + 'learning_rate': 6.02128093402389e-06, 'epoch': + 8.91} +04/20 [00:14:40] INFO | >> train_qwenlatent.py:487 + Step 35340 | grad_norm_pre_clip=0.1754 | + grad_norm_pre_clip_avg=0.1871 | Metrics: + {'align_loss': 0.024243127554655075, + 'recon_loss': 0.14071254432201385, + 'predict_loss': 0.007996841333806515, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17538821697235107, + 'data_time': 0.0009369309991598129, + 'model_time': 1.2689244339999277, + 'grad_norm_pre_clip_avg': 0.18714311718940735, + 'learning_rate': 6.013834994522162e-06, + 'epoch': 8.92} +04/20 [00:14:54] INFO | >> train_qwenlatent.py:487 + Step 35350 | grad_norm_pre_clip=0.1907 | + grad_norm_pre_clip_avg=0.1825 | Metrics: + {'align_loss': 0.02601785771548748, + 'recon_loss': 0.11750499904155731, + 'predict_loss': 0.007064282428473234, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19068105518817902, + 'mae_score': 0.009555898271165453, 'data_time': + 0.0010105739929713309, 'model_time': + 1.2697771929961164, 'grad_norm_pre_clip_avg': + 0.18249236345291137, 'learning_rate': + 6.006392222390807e-06, 'epoch': 8.92} +04/20 [00:15:06] INFO | >> train_qwenlatent.py:487 + Step 35360 | grad_norm_pre_clip=0.1587 | + grad_norm_pre_clip_avg=0.1631 | Metrics: + {'align_loss': 0.024150440469384193, + 'recon_loss': 0.09254933148622513, + 'predict_loss': 0.004503953270614147, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15870584547519684, + 'data_time': 0.0006143420177977532, + 'model_time': 1.2520645569893532, + 'grad_norm_pre_clip_avg': 0.16314484775066376, + 'learning_rate': 5.998952621257342e-06, + 'epoch': 8.92} +04/20 [00:15:18] INFO | >> train_qwenlatent.py:487 + Step 35370 | grad_norm_pre_clip=0.1682 | + grad_norm_pre_clip_avg=0.1594 | Metrics: + {'align_loss': 0.025517061352729797, + 'recon_loss': 0.09303490072488785, + 'predict_loss': 0.003962105140089989, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16821475327014923, + 'data_time': 0.0006780519906897098, + 'model_time': 1.1925598030211404, + 'grad_norm_pre_clip_avg': 0.1593877710402012, + 'learning_rate': 5.991516194747746e-06, + 'epoch': 8.93} +04/20 [00:15:31] INFO | >> train_qwenlatent.py:487 + Step 35380 | grad_norm_pre_clip=0.1548 | + grad_norm_pre_clip_avg=0.1526 | Metrics: + {'align_loss': 0.025854384526610374, + 'recon_loss': 0.07517833262681961, + 'predict_loss': 0.005499246995896101, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15478096902370453, + 'data_time': 0.0008710680122021586, + 'model_time': 1.5752965650171973, + 'grad_norm_pre_clip_avg': 0.15258054658770562, + 'learning_rate': 5.984082946486438e-06, + 'epoch': 8.93} +04/20 [00:15:44] INFO | >> train_qwenlatent.py:487 + Step 35390 | grad_norm_pre_clip=0.1511 | + grad_norm_pre_clip_avg=0.1573 | Metrics: + {'align_loss': 0.02578086405992508, + 'recon_loss': 0.10967224836349487, + 'predict_loss': 0.004239413887262344, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15110763907432556, + 'data_time': 0.0007606790168210864, + 'model_time': 1.27693948301021, + 'grad_norm_pre_clip_avg': 0.15728872865438462, + 'learning_rate': 5.976652880096295e-06, + 'epoch': 8.93} +04/20 [00:15:57] INFO | >> train_qwenlatent.py:487 + Step 35400 | grad_norm_pre_clip=0.1263 | + grad_norm_pre_clip_avg=0.1703 | Metrics: + {'align_loss': 0.02552335150539875, + 'recon_loss': 0.1067713275551796, + 'predict_loss': 0.009100564755499363, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12633119523525238, + 'mae_score': 0.006590567837964307, 'data_time': + 0.0006562830240000039, 'model_time': + 1.2600314329902176, 'grad_norm_pre_clip_avg': + 0.1703150451183319, 'learning_rate': + 5.969225999198639e-06, 'epoch': 8.93} +04/20 [00:16:10] INFO | >> train_qwenlatent.py:487 + Step 35410 | grad_norm_pre_clip=0.1600 | + grad_norm_pre_clip_avg=0.1959 | Metrics: + {'align_loss': 0.02488775923848152, + 'recon_loss': 0.133366659283638, + 'predict_loss': 0.007681984454393387, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15998220443725586, + 'data_time': 0.0006707379943691194, + 'model_time': 1.3747416639816947, + 'grad_norm_pre_clip_avg': 0.19589532166719437, + 'learning_rate': 5.961802307413238e-06, + 'epoch': 8.94} +04/20 [00:16:23] INFO | >> train_qwenlatent.py:487 + Step 35420 | grad_norm_pre_clip=0.1727 | + grad_norm_pre_clip_avg=0.1777 | Metrics: + {'align_loss': 0.026336319744586945, + 'recon_loss': 0.10270136594772339, + 'predict_loss': 0.008731258101761341, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.172685444355011, + 'data_time': 0.0008016880019567907, + 'model_time': 1.2372691790224053, + 'grad_norm_pre_clip_avg': 0.1777074322104454, + 'learning_rate': 5.954381808358319e-06, + 'epoch': 8.94} +04/20 [00:16:35] INFO | >> train_qwenlatent.py:487 + Step 35430 | grad_norm_pre_clip=0.2231 | + grad_norm_pre_clip_avg=0.1767 | Metrics: + {'align_loss': 0.024663638323545456, + 'recon_loss': 0.10774655640125275, + 'predict_loss': 0.00568210706114769, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2230924516916275, + 'data_time': 0.0009294949995819479, + 'model_time': 1.2544946599809919, + 'grad_norm_pre_clip_avg': 0.17667166441679, + 'learning_rate': 5.946964505650542e-06, + 'epoch': 8.94} +04/20 [00:16:48] INFO | >> train_qwenlatent.py:487 + Step 35440 | grad_norm_pre_clip=0.1773 | + grad_norm_pre_clip_avg=0.1743 | Metrics: + {'align_loss': 0.024945378303527832, + 'recon_loss': 0.10093898326158524, + 'predict_loss': 0.008315409533679485, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1772896945476532, + 'data_time': 0.0008476620132569224, + 'model_time': 1.2291963740135543, + 'grad_norm_pre_clip_avg': 0.17428269535303115, + 'learning_rate': 5.939550402905e-06, 'epoch': + 8.94} +04/20 [00:17:01] INFO | >> train_qwenlatent.py:487 + Step 35450 | grad_norm_pre_clip=0.1568 | + grad_norm_pre_clip_avg=0.1650 | Metrics: + {'align_loss': 0.02551381289958954, + 'recon_loss': 0.12266362458467484, + 'predict_loss': 0.007370871491730213, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15684834122657776, + 'mae_score': 0.0060839842031667896, + 'data_time': 0.00094175199046731, 'model_time': + 1.2468219130241778, 'grad_norm_pre_clip_avg': + 0.16496870219707488, 'learning_rate': + 5.932139503735238e-06, 'epoch': 8.95} +04/20 [00:17:13] INFO | >> train_qwenlatent.py:487 + Step 35460 | grad_norm_pre_clip=0.1435 | + grad_norm_pre_clip_avg=0.1569 | Metrics: + {'align_loss': 0.02662845142185688, + 'recon_loss': 0.1251424252986908, + 'predict_loss': 0.0071114529855549335, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14352728426456451, + 'data_time': 0.0006734500057063997, + 'model_time': 1.2389495109964628, + 'grad_norm_pre_clip_avg': 0.15689015686511992, + 'learning_rate': 5.924731811753249e-06, + 'epoch': 8.95} +04/20 [00:17:26] INFO | >> train_qwenlatent.py:487 + Step 35470 | grad_norm_pre_clip=0.2041 | + grad_norm_pre_clip_avg=0.1758 | Metrics: + {'align_loss': 0.02510060928761959, + 'recon_loss': 0.11262282729148865, + 'predict_loss': 0.008041548542678356, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2041209489107132, + 'data_time': 0.0008829290163703263, + 'model_time': 1.23306204599794, + 'grad_norm_pre_clip_avg': 0.17575622648000716, + 'learning_rate': 5.917327330569445e-06, + 'epoch': 8.95} +04/20 [00:17:38] INFO | >> train_qwenlatent.py:487 + Step 35480 | grad_norm_pre_clip=0.1298 | + grad_norm_pre_clip_avg=0.1531 | Metrics: + {'align_loss': 0.025928989052772522, + 'recon_loss': 0.16174794733524323, + 'predict_loss': 0.009328695014119148, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12980762124061584, + 'data_time': 0.0008037399966269732, + 'model_time': 1.2134405079996213, + 'grad_norm_pre_clip_avg': 0.1530722811818123, + 'learning_rate': 5.909926063792681e-06, + 'epoch': 8.95} +04/20 [00:17:51] INFO | >> train_qwenlatent.py:487 + Step 35490 | grad_norm_pre_clip=0.1161 | + grad_norm_pre_clip_avg=0.1514 | Metrics: + {'align_loss': 0.022715650498867035, + 'recon_loss': 0.0984765961766243, + 'predict_loss': 0.005729469936341047, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11609381437301636, + 'data_time': 0.0009559280006214976, + 'model_time': 1.3078323180088773, + 'grad_norm_pre_clip_avg': 0.15140931084752082, + 'learning_rate': 5.902528015030245e-06, + 'epoch': 8.96} +04/20 [00:18:04] INFO | >> train_qwenlatent.py:487 + Step 35500 | grad_norm_pre_clip=0.2322 | + grad_norm_pre_clip_avg=0.1541 | Metrics: + {'align_loss': 0.024430176243185997, + 'recon_loss': 0.12547194957733154, + 'predict_loss': 0.007088263984769583, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.232194721698761, + 'mae_score': 0.007129428622958896, 'data_time': + 0.0010778419964481145, 'model_time': + 1.2976865920063574, 'grad_norm_pre_clip_avg': + 0.15412285253405572, 'learning_rate': + 5.895133187887856e-06, 'epoch': 8.96} +04/20 [00:18:17] INFO | >> train_qwenlatent.py:487 + Step 35510 | grad_norm_pre_clip=0.1382 | + grad_norm_pre_clip_avg=0.1738 | Metrics: + {'align_loss': 0.024097435176372528, + 'recon_loss': 0.10373856872320175, + 'predict_loss': 0.006604814901947975, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13822925090789795, + 'data_time': 0.0008838190115056932, + 'model_time': 1.2667810729763005, + 'grad_norm_pre_clip_avg': 0.17378107011318206, + 'learning_rate': 5.887741585969656e-06, + 'epoch': 8.96} +04/20 [00:18:29] INFO | >> train_qwenlatent.py:487 + Step 35520 | grad_norm_pre_clip=0.1751 | + grad_norm_pre_clip_avg=0.1767 | Metrics: + {'align_loss': 0.02596992813050747, + 'recon_loss': 0.11779669672250748, + 'predict_loss': 0.008912825025618076, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17511261999607086, + 'data_time': 0.0007500359788537025, + 'model_time': 1.260290823993273, + 'grad_norm_pre_clip_avg': 0.17674146145582198, + 'learning_rate': 5.880353212878239e-06, + 'epoch': 8.96} +04/20 [00:18:42] INFO | >> train_qwenlatent.py:487 + Step 35530 | grad_norm_pre_clip=0.2203 | + grad_norm_pre_clip_avg=0.1744 | Metrics: + {'align_loss': 0.025891181081533432, + 'recon_loss': 0.0994829386472702, + 'predict_loss': 0.0038825771771371365, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22028860449790955, + 'data_time': 0.0006463320169132203, + 'model_time': 1.2672236810030881, + 'grad_norm_pre_clip_avg': 0.17441271468997002, + 'learning_rate': 5.872968072214594e-06, + 'epoch': 8.97} +04/20 [00:18:55] INFO | >> train_qwenlatent.py:487 + Step 35540 | grad_norm_pre_clip=0.2337 | + grad_norm_pre_clip_avg=0.1809 | Metrics: + {'align_loss': 0.025733396410942078, + 'recon_loss': 0.11053775250911713, + 'predict_loss': 0.007421885617077351, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23365487158298492, + 'data_time': 0.0008539660193491727, + 'model_time': 1.2065837859991007, + 'grad_norm_pre_clip_avg': 0.18093645721673965, + 'learning_rate': 5.865586167578151e-06, + 'epoch': 8.97} +04/20 [00:19:08] INFO | >> train_qwenlatent.py:487 + Step 35550 | grad_norm_pre_clip=0.1773 | + grad_norm_pre_clip_avg=0.1970 | Metrics: + {'align_loss': 0.02456030249595642, + 'recon_loss': 0.09875720739364624, + 'predict_loss': 0.00665535032749176, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17731605470180511, + 'mae_score': 0.006838798093366193, 'data_time': + 0.0005936379893682897, 'model_time': + 1.20602191000944, 'grad_norm_pre_clip_avg': + 0.19699952006340027, 'learning_rate': + 5.858207502566759e-06, 'epoch': 8.97} +04/20 [00:19:21] INFO | >> train_qwenlatent.py:487 + Step 35560 | grad_norm_pre_clip=0.1468 | + grad_norm_pre_clip_avg=0.1622 | Metrics: + {'align_loss': 0.026612121611833572, + 'recon_loss': 0.1184123307466507, + 'predict_loss': 0.006646817084401846, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1467638611793518, + 'data_time': 0.0006348199967760593, + 'model_time': 1.2047775149985682, + 'grad_norm_pre_clip_avg': 0.16219737082719804, + 'learning_rate': 5.850832080776697e-06, + 'epoch': 8.97} +04/20 [00:19:33] INFO | >> train_qwenlatent.py:487 + Step 35570 | grad_norm_pre_clip=0.1954 | + grad_norm_pre_clip_avg=0.1735 | Metrics: + {'align_loss': 0.026260413229465485, + 'recon_loss': 0.1284405142068863, + 'predict_loss': 0.005279063247144222, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1954430192708969, + 'data_time': 0.0008751580026000738, + 'model_time': 1.2581663609889802, + 'grad_norm_pre_clip_avg': 0.17351213097572327, + 'learning_rate': 5.843459905802652e-06, + 'epoch': 8.98} +04/20 [00:19:46] INFO | >> train_qwenlatent.py:487 + Step 35580 | grad_norm_pre_clip=0.1492 | + grad_norm_pre_clip_avg=0.1607 | Metrics: + {'align_loss': 0.026876740157604218, + 'recon_loss': 0.14483550190925598, + 'predict_loss': 0.011349724605679512, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14922767877578735, + 'data_time': 0.0008933600038290024, + 'model_time': 1.2537044970085844, + 'grad_norm_pre_clip_avg': 0.16067803800106048, + 'learning_rate': 5.8360909812377335e-06, + 'epoch': 8.98} +04/20 [00:19:59] INFO | >> train_qwenlatent.py:487 + Step 35590 | grad_norm_pre_clip=0.1900 | + grad_norm_pre_clip_avg=0.1645 | Metrics: + {'align_loss': 0.025732597336173058, + 'recon_loss': 0.13338667154312134, + 'predict_loss': 0.008562683127820492, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19000931084156036, + 'data_time': 0.0008264679927378893, + 'model_time': 1.2498166950244922, + 'grad_norm_pre_clip_avg': 0.16447250694036483, + 'learning_rate': 5.8287253106734635e-06, + 'epoch': 8.98} +04/20 [00:20:12] INFO | >> train_qwenlatent.py:487 + Step 35600 | grad_norm_pre_clip=0.1698 | + grad_norm_pre_clip_avg=0.1620 | Metrics: + {'align_loss': 0.026310455054044724, + 'recon_loss': 0.12259630113840103, + 'predict_loss': 0.00881342776119709, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16983282566070557, + 'mae_score': 0.006939382810850402, 'data_time': + 0.0008532880165148526, 'model_time': + 1.2631178770097904, 'grad_norm_pre_clip_avg': + 0.1619808629155159, 'learning_rate': + 5.8213628976997765e-06, 'epoch': 8.98} +04/20 [00:20:25] INFO | >> train_qwenlatent.py:487 + Step 35610 | grad_norm_pre_clip=0.1506 | + grad_norm_pre_clip_avg=0.1748 | Metrics: + {'align_loss': 0.025693759322166443, + 'recon_loss': 0.12168216705322266, + 'predict_loss': 0.011195804923772812, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15061357617378235, + 'data_time': 0.00096410300466232, 'model_time': + 1.2545623790065292, 'grad_norm_pre_clip_avg': + 0.17482624500989913, 'learning_rate': + 5.814003745905039e-06, 'epoch': 8.99} +04/20 [00:20:37] INFO | >> train_qwenlatent.py:487 + Step 35620 | grad_norm_pre_clip=0.1701 | + grad_norm_pre_clip_avg=0.1805 | Metrics: + {'align_loss': 0.026466134935617447, + 'recon_loss': 0.12606017291545868, + 'predict_loss': 0.00984618067741394, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17008806765079498, + 'data_time': 0.001222304010298103, + 'model_time': 1.2720006440067664, + 'grad_norm_pre_clip_avg': 0.18048283010721206, + 'learning_rate': 5.806647858875997e-06, + 'epoch': 8.99} +04/20 [00:20:50] INFO | >> train_qwenlatent.py:487 + Step 35630 | grad_norm_pre_clip=0.1839 | + grad_norm_pre_clip_avg=0.1790 | Metrics: + {'align_loss': 0.02629759907722473, + 'recon_loss': 0.1107289269566536, + 'predict_loss': 0.005173173267394304, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.183889701962471, + 'data_time': 0.0009056009876076132, + 'model_time': 1.2282883120060433, + 'grad_norm_pre_clip_avg': 0.1789878413081169, + 'learning_rate': 5.799295240197825e-06, + 'epoch': 8.99} +04/20 [00:21:02] INFO | >> train_qwenlatent.py:487 + Step 35640 | grad_norm_pre_clip=0.1591 | + grad_norm_pre_clip_avg=0.1622 | Metrics: + {'align_loss': 0.025781381875276566, + 'recon_loss': 0.11747126281261444, + 'predict_loss': 0.0137320039793849, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15905968844890594, + 'data_time': 0.0007995610067155212, + 'model_time': 1.2553778659785166, + 'grad_norm_pre_clip_avg': 0.16224654987454415, + 'learning_rate': 5.791945893454093e-06, + 'epoch': 8.99} +04/20 [00:21:16] INFO | >> train_qwenlatent.py:487 + Step 35650 | grad_norm_pre_clip=0.1527 | + grad_norm_pre_clip_avg=0.1573 | Metrics: + {'align_loss': 0.025925248861312866, + 'recon_loss': 0.10088396817445755, + 'predict_loss': 0.009698925539851189, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15268072485923767, + 'mae_score': 0.006013968613770631, 'data_time': + 0.0006336339865811169, 'model_time': + 1.2380990630190354, 'grad_norm_pre_clip_avg': + 0.1572841741144657, 'learning_rate': + 5.7845998222267965e-06, 'epoch': 9.0} +04/20 [00:21:29] INFO | >> train_qwenlatent.py:487 + Step 35660 | grad_norm_pre_clip=0.1370 | + grad_norm_pre_clip_avg=0.1639 | Metrics: + {'align_loss': 0.024233821779489517, + 'recon_loss': 0.11027076095342636, + 'predict_loss': 0.006058103870600462, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13699957728385925, + 'data_time': 0.0006461919983848929, + 'model_time': 1.2300199139863253, + 'grad_norm_pre_clip_avg': 0.16389714032411576, + 'learning_rate': 5.777257030096315e-06, + 'epoch': 9.0} +04/20 [00:21:41] INFO | >> train_qwenlatent.py:487 + Step 35670 | grad_norm_pre_clip=0.1312 | + grad_norm_pre_clip_avg=0.1450 | Metrics: + {'align_loss': 0.02581685036420822, + 'recon_loss': 0.14244262874126434, + 'predict_loss': 0.008798611350357533, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13121187686920166, + 'data_time': 0.0006039419968146831, + 'model_time': 1.1812286580097862, + 'grad_norm_pre_clip_avg': 0.14504623264074326, + 'learning_rate': 5.769917520641439e-06, + 'epoch': 9.0} +04/20 [00:21:54] INFO | >> train_qwenlatent.py:487 + Step 35680 | grad_norm_pre_clip=0.1230 | + grad_norm_pre_clip_avg=0.1591 | Metrics: + {'align_loss': 0.025575166568160057, + 'recon_loss': 0.14230242371559143, + 'predict_loss': 0.007089428137987852, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1229531392455101, + 'data_time': 0.000866392016177997, + 'model_time': 1.2181673370068893, + 'grad_norm_pre_clip_avg': 0.15909934565424919, + 'learning_rate': 5.762581297439351e-06, + 'epoch': 9.0} +04/20 [00:22:06] INFO | >> train_qwenlatent.py:487 + Step 35690 | grad_norm_pre_clip=0.2012 | + grad_norm_pre_clip_avg=0.1816 | Metrics: + {'align_loss': 0.025751741603016853, + 'recon_loss': 0.092007115483284, + 'predict_loss': 0.006233967375010252, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20122529566287994, + 'data_time': 0.0007324700127355754, + 'model_time': 1.2827642729971558, + 'grad_norm_pre_clip_avg': 0.18159705996513367, + 'learning_rate': 5.755248364065638e-06, + 'epoch': 9.01} +04/20 [00:22:20] INFO | >> train_qwenlatent.py:487 + Step 35700 | grad_norm_pre_clip=0.1949 | + grad_norm_pre_clip_avg=0.1651 | Metrics: + {'align_loss': 0.026219118386507034, + 'recon_loss': 0.1356424242258072, + 'predict_loss': 0.014613201841711998, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19491145014762878, + 'mae_score': 0.009107110307023332, 'data_time': + 0.0008732090063858777, 'model_time': + 1.2201811429986265, 'grad_norm_pre_clip_avg': + 0.16511043310165405, 'learning_rate': + 5.747918724094288e-06, 'epoch': 9.01} +04/20 [00:22:32] INFO | >> train_qwenlatent.py:487 + Step 35710 | grad_norm_pre_clip=0.1601 | + grad_norm_pre_clip_avg=0.1591 | Metrics: + {'align_loss': 0.026714598760008812, + 'recon_loss': 0.17071393132209778, + 'predict_loss': 0.012101826258003712, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16007357835769653, + 'data_time': 0.0009540009777992964, + 'model_time': 1.2664233119867276, + 'grad_norm_pre_clip_avg': 0.15913690626621246, + 'learning_rate': 5.740592381097673e-06, + 'epoch': 9.01} +04/20 [00:22:44] INFO | >> train_qwenlatent.py:487 + Step 35720 | grad_norm_pre_clip=0.1839 | + grad_norm_pre_clip_avg=0.1758 | Metrics: + {'align_loss': 0.02557108923792839, + 'recon_loss': 0.14006425440311432, + 'predict_loss': 0.005253707990050316, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18388429284095764, + 'data_time': 0.0008123909938149154, + 'model_time': 1.2096864990016911, + 'grad_norm_pre_clip_avg': 0.17575886845588684, + 'learning_rate': 5.733269338646566e-06, + 'epoch': 9.01} +04/20 [00:22:57] INFO | >> train_qwenlatent.py:487 + Step 35730 | grad_norm_pre_clip=0.1685 | + grad_norm_pre_clip_avg=0.1450 | Metrics: + {'align_loss': 0.026265311986207962, + 'recon_loss': 0.1206081360578537, + 'predict_loss': 0.004155863542109728, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16847901046276093, + 'data_time': 0.0009041519952006638, + 'model_time': 1.214136452996172, + 'grad_norm_pre_clip_avg': 0.1449923299252987, + 'learning_rate': 5.725949600310128e-06, + 'epoch': 9.02} +04/20 [00:23:10] INFO | >> train_qwenlatent.py:487 + Step 35740 | grad_norm_pre_clip=0.1723 | + grad_norm_pre_clip_avg=0.1684 | Metrics: + {'align_loss': 0.026163499802350998, + 'recon_loss': 0.14147713780403137, + 'predict_loss': 0.00847859587520361, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17229804396629333, + 'data_time': 0.0009300489909946918, + 'model_time': 1.2070914240030106, + 'grad_norm_pre_clip_avg': 0.16835455298423768, + 'learning_rate': 5.718633169655906e-06, + 'epoch': 9.02} +04/20 [00:23:23] INFO | >> train_qwenlatent.py:487 + Step 35750 | grad_norm_pre_clip=0.1883 | + grad_norm_pre_clip_avg=0.1701 | Metrics: + {'align_loss': 0.026051584631204605, + 'recon_loss': 0.11210940033197403, + 'predict_loss': 0.010392001830041409, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1882784515619278, + 'mae_score': 0.006326971827326594, 'data_time': + 0.0011225930065847933, 'model_time': + 1.2129017809929792, 'grad_norm_pre_clip_avg': + 0.17011404782533646, 'learning_rate': + 5.71132005024985e-06, 'epoch': 9.02} +04/20 [00:23:35] INFO | >> train_qwenlatent.py:487 + Step 35760 | grad_norm_pre_clip=0.1421 | + grad_norm_pre_clip_avg=0.1701 | Metrics: + {'align_loss': 0.02617890015244484, + 'recon_loss': 0.11615908890962601, + 'predict_loss': 0.006552153266966343, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14206427335739136, + 'data_time': 0.0007783220207784325, + 'model_time': 1.2659219769993797, + 'grad_norm_pre_clip_avg': 0.17009923979640007, + 'learning_rate': 5.7040102456562775e-06, + 'epoch': 9.02} +04/20 [00:23:48] INFO | >> train_qwenlatent.py:487 + Step 35770 | grad_norm_pre_clip=0.1608 | + grad_norm_pre_clip_avg=0.1614 | Metrics: + {'align_loss': 0.024942398071289062, + 'recon_loss': 0.09678282588720322, + 'predict_loss': 0.005392034538090229, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16084028780460358, + 'data_time': 0.0009095280256588012, + 'model_time': 1.2795489169948269, + 'grad_norm_pre_clip_avg': 0.16137705892324447, + 'learning_rate': 5.696703759437902e-06, + 'epoch': 9.03} +04/20 [00:24:00] INFO | >> train_qwenlatent.py:487 + Step 35780 | grad_norm_pre_clip=0.2114 | + grad_norm_pre_clip_avg=0.1764 | Metrics: + {'align_loss': 0.025505097582936287, + 'recon_loss': 0.1164453998208046, + 'predict_loss': 0.010643660090863705, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21139313280582428, + 'data_time': 0.0008592609956394881, + 'model_time': 1.2040390679903794, + 'grad_norm_pre_clip_avg': 0.17635988891124726, + 'learning_rate': 5.689400595155816e-06, + 'epoch': 9.03} +04/20 [00:24:13] INFO | >> train_qwenlatent.py:487 + Step 35790 | grad_norm_pre_clip=0.1573 | + grad_norm_pre_clip_avg=0.1650 | Metrics: + {'align_loss': 0.025795290246605873, + 'recon_loss': 0.1413075178861618, + 'predict_loss': 0.004800368566066027, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15728504955768585, + 'data_time': 0.0007448050018865615, + 'model_time': 1.2310049080115277, + 'grad_norm_pre_clip_avg': 0.16497111320495605, + 'learning_rate': 5.682100756369491e-06, + 'epoch': 9.03} +04/20 [00:24:26] INFO | >> train_qwenlatent.py:487 + Step 35800 | grad_norm_pre_clip=0.1669 | + grad_norm_pre_clip_avg=0.1729 | Metrics: + {'align_loss': 0.026949016377329826, + 'recon_loss': 0.11898113787174225, + 'predict_loss': 0.008192981593310833, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16692079603672028, + 'mae_score': 0.005875717627035605, 'data_time': + 0.0006540630129165947, 'model_time': + 1.1903447670047171, 'grad_norm_pre_clip_avg': + 0.172920760512352, 'learning_rate': + 5.67480424663678e-06, 'epoch': 9.03} +04/20 [00:24:39] INFO | >> train_qwenlatent.py:487 + Step 35810 | grad_norm_pre_clip=0.1602 | + grad_norm_pre_clip_avg=0.1604 | Metrics: + {'align_loss': 0.02527274563908577, + 'recon_loss': 0.15103350579738617, + 'predict_loss': 0.01078062690794468, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16015653312206268, + 'data_time': 0.0006439989956561476, + 'model_time': 1.2307563619979192, + 'grad_norm_pre_clip_avg': 0.16036787778139114, + 'learning_rate': 5.667511069513916e-06, + 'epoch': 9.04} +04/20 [00:24:51] INFO | >> train_qwenlatent.py:487 + Step 35820 | grad_norm_pre_clip=0.1377 | + grad_norm_pre_clip_avg=0.1424 | Metrics: + {'align_loss': 0.025749366730451584, + 'recon_loss': 0.08389799296855927, + 'predict_loss': 0.0042526754550635815, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13772346079349518, + 'data_time': 0.0009709670266602188, + 'model_time': 1.2047456680156756, + 'grad_norm_pre_clip_avg': 0.14238054752349855, + 'learning_rate': 5.660221228555501e-06, + 'epoch': 9.04} +04/20 [00:25:04] INFO | >> train_qwenlatent.py:487 + Step 35830 | grad_norm_pre_clip=0.1688 | + grad_norm_pre_clip_avg=0.1588 | Metrics: + {'align_loss': 0.02554256096482277, + 'recon_loss': 0.11535149812698364, + 'predict_loss': 0.00879746675491333, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1688205450773239, + 'data_time': 0.0008646649948786944, + 'model_time': 1.2393690180033445, + 'grad_norm_pre_clip_avg': 0.15884757041931152, + 'learning_rate': 5.6529347273145186e-06, + 'epoch': 9.04} +04/20 [00:25:16] INFO | >> train_qwenlatent.py:487 + Step 35840 | grad_norm_pre_clip=0.1481 | + grad_norm_pre_clip_avg=0.1678 | Metrics: + {'align_loss': 0.025586087256669998, + 'recon_loss': 0.11499668657779694, + 'predict_loss': 0.008252352476119995, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14811921119689941, + 'data_time': 0.0006771849875804037, + 'model_time': 1.1825649009842891, + 'grad_norm_pre_clip_avg': 0.1678368330001831, + 'learning_rate': 5.645651569342313e-06, + 'epoch': 9.04} +04/20 [00:25:29] INFO | >> train_qwenlatent.py:487 + Step 35850 | grad_norm_pre_clip=0.1512 | + grad_norm_pre_clip_avg=0.1630 | Metrics: + {'align_loss': 0.026050684973597527, + 'recon_loss': 0.12459814548492432, + 'predict_loss': 0.008592342957854271, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15117716789245605, + 'mae_score': 0.0064117431640625, 'data_time': + 0.0008011350000742823, 'model_time': + 1.2081647889863234, 'grad_norm_pre_clip_avg': + 0.16303105503320695, 'learning_rate': + 5.6383717581886185e-06, 'epoch': 9.05} +04/20 [00:25:42] INFO | >> train_qwenlatent.py:487 + Step 35860 | grad_norm_pre_clip=0.1473 | + grad_norm_pre_clip_avg=0.1547 | Metrics: + {'align_loss': 0.025487039238214493, + 'recon_loss': 0.13178320229053497, + 'predict_loss': 0.009500023908913136, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14725695550441742, + 'data_time': 0.0006614689773414284, + 'model_time': 1.1894408880034462, + 'grad_norm_pre_clip_avg': 0.154693653434515, + 'learning_rate': 5.6310952974015186e-06, + 'epoch': 9.05} +04/20 [00:25:55] INFO | >> train_qwenlatent.py:487 + Step 35870 | grad_norm_pre_clip=0.1544 | + grad_norm_pre_clip_avg=0.1643 | Metrics: + {'align_loss': 0.024362877011299133, + 'recon_loss': 0.10283465683460236, + 'predict_loss': 0.007361807394772768, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1544022560119629, + 'data_time': 0.0007776020211167634, + 'model_time': 1.2313675379846245, + 'grad_norm_pre_clip_avg': 0.1643405243754387, + 'learning_rate': 5.623822190527478e-06, + 'epoch': 9.05} +04/20 [00:26:07] INFO | >> train_qwenlatent.py:487 + Step 35880 | grad_norm_pre_clip=0.1491 | + grad_norm_pre_clip_avg=0.1458 | Metrics: + {'align_loss': 0.026689285412430763, + 'recon_loss': 0.13313883543014526, + 'predict_loss': 0.006241415161639452, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14914526045322418, + 'data_time': 0.000913989992113784, + 'model_time': 1.2341547090036329, + 'grad_norm_pre_clip_avg': 0.14575383365154265, + 'learning_rate': 5.6165524411113165e-06, + 'epoch': 9.05} +04/20 [00:26:20] INFO | >> train_qwenlatent.py:487 + Step 35890 | grad_norm_pre_clip=0.1231 | + grad_norm_pre_clip_avg=0.1643 | Metrics: + {'align_loss': 0.025650639086961746, + 'recon_loss': 0.1272728145122528, + 'predict_loss': 0.006322023458778858, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12312299758195877, + 'data_time': 0.0008690629911143333, + 'model_time': 1.3027192300069146, + 'grad_norm_pre_clip_avg': 0.16434206143021585, + 'learning_rate': 5.60928605269622e-06, 'epoch': + 9.06} +04/20 [00:26:33] INFO | >> train_qwenlatent.py:487 + Step 35900 | grad_norm_pre_clip=0.1595 | + grad_norm_pre_clip_avg=0.1665 | Metrics: + {'align_loss': 0.026006050407886505, + 'recon_loss': 0.11979226022958755, + 'predict_loss': 0.008904584683477879, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15945421159267426, + 'mae_score': 0.009034342379183383, 'data_time': + 0.0009400580020155758, 'model_time': + 1.2600506170128938, 'grad_norm_pre_clip_avg': + 0.16651549339294433, 'learning_rate': + 5.602023028823742e-06, 'epoch': 9.06} +04/20 [00:26:46] INFO | >> train_qwenlatent.py:487 + Step 35910 | grad_norm_pre_clip=0.2220 | + grad_norm_pre_clip_avg=0.1620 | Metrics: + {'align_loss': 0.02618243172764778, + 'recon_loss': 0.13461638987064362, + 'predict_loss': 0.010826194658875465, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22204603254795074, + 'data_time': 0.0007238209946081042, + 'model_time': 1.2723401770053897, + 'grad_norm_pre_clip_avg': 0.16198711842298508, + 'learning_rate': 5.594763373033792e-06, + 'epoch': 9.06} +04/20 [00:26:58] INFO | >> train_qwenlatent.py:487 + Step 35920 | grad_norm_pre_clip=0.1697 | + grad_norm_pre_clip_avg=0.1661 | Metrics: + {'align_loss': 0.025690153241157532, + 'recon_loss': 0.11138755083084106, + 'predict_loss': 0.005824495572596788, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16970203816890717, + 'data_time': 0.000922626000829041, + 'model_time': 1.2032576969941147, + 'grad_norm_pre_clip_avg': 0.16611116454005243, + 'learning_rate': 5.587507088864633e-06, + 'epoch': 9.06} +04/20 [00:27:11] INFO | >> train_qwenlatent.py:487 + Step 35930 | grad_norm_pre_clip=0.1342 | + grad_norm_pre_clip_avg=0.1476 | Metrics: + {'align_loss': 0.02530021034181118, + 'recon_loss': 0.11314808577299118, + 'predict_loss': 0.006263030227273703, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13423524796962738, + 'data_time': 0.0007733670063316822, + 'model_time': 1.3222424289851915, + 'grad_norm_pre_clip_avg': 0.14764377400279044, + 'learning_rate': 5.58025417985289e-06, 'epoch': + 9.07} +04/20 [00:27:24] INFO | >> train_qwenlatent.py:487 + Step 35940 | grad_norm_pre_clip=0.1518 | + grad_norm_pre_clip_avg=0.1535 | Metrics: + {'align_loss': 0.02511836588382721, + 'recon_loss': 0.1576576679944992, + 'predict_loss': 0.008192455396056175, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15180429816246033, + 'data_time': 0.0013041529746260494, + 'model_time': 1.2452875539893284, + 'grad_norm_pre_clip_avg': 0.15347571521997452, + 'learning_rate': 5.573004649533551e-06, + 'epoch': 9.07} +04/20 [00:27:37] INFO | >> train_qwenlatent.py:487 + Step 35950 | grad_norm_pre_clip=0.1501 | + grad_norm_pre_clip_avg=0.1520 | Metrics: + {'align_loss': 0.02592538855969906, + 'recon_loss': 0.0729215070605278, + 'predict_loss': 0.0051240473985672, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1501488834619522, + 'mae_score': 0.007438361537349117, 'data_time': + 0.0008331569842994213, 'model_time': + 1.2209411799849477, 'grad_norm_pre_clip_avg': + 0.15197589248418808, 'learning_rate': + 5.565758501439944e-06, 'epoch': 9.07} +04/20 [00:27:50] INFO | >> train_qwenlatent.py:487 + Step 35960 | grad_norm_pre_clip=0.1723 | + grad_norm_pre_clip_avg=0.1694 | Metrics: + {'align_loss': 0.025770191103219986, + 'recon_loss': 0.10066087543964386, + 'predict_loss': 0.0045327553525567055, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17233453691005707, + 'data_time': 0.0010088279959745705, + 'model_time': 1.297699158982141, + 'grad_norm_pre_clip_avg': 0.16941825896501542, + 'learning_rate': 5.558515739103758e-06, + 'epoch': 9.07} +04/20 [00:28:02] INFO | >> train_qwenlatent.py:487 + Step 35970 | grad_norm_pre_clip=0.2037 | + grad_norm_pre_clip_avg=0.1674 | Metrics: + {'align_loss': 0.026726385578513145, + 'recon_loss': 0.12644077837467194, + 'predict_loss': 0.009333351626992226, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2036888748407364, + 'data_time': 0.0007247989997267723, + 'model_time': 1.2728445309912786, + 'grad_norm_pre_clip_avg': 0.16744635701179506, + 'learning_rate': 5.551276366055012e-06, + 'epoch': 9.08} +04/20 [00:28:15] INFO | >> train_qwenlatent.py:487 + Step 35980 | grad_norm_pre_clip=0.1891 | + grad_norm_pre_clip_avg=0.1625 | Metrics: + {'align_loss': 0.02572731301188469, + 'recon_loss': 0.09823363274335861, + 'predict_loss': 0.003706313669681549, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18913792073726654, + 'data_time': 0.0008970999915618449, + 'model_time': 1.2150468319887295, + 'grad_norm_pre_clip_avg': 0.16245792880654336, + 'learning_rate': 5.544040385822105e-06, + 'epoch': 9.08} +04/20 [00:28:27] INFO | >> train_qwenlatent.py:487 + Step 35990 | grad_norm_pre_clip=0.2154 | + grad_norm_pre_clip_avg=0.1613 | Metrics: + {'align_loss': 0.02606581151485443, + 'recon_loss': 0.1121768206357956, + 'predict_loss': 0.004682868719100952, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21544474363327026, + 'data_time': 0.0006867169868201017, + 'model_time': 1.249378667009296, + 'grad_norm_pre_clip_avg': 0.16131994426250457, + 'learning_rate': 5.53680780193176e-06, 'epoch': + 9.08} +04/20 [00:28:41] INFO | >> train_qwenlatent.py:487 + Step 36000 | grad_norm_pre_clip=0.2560 | + grad_norm_pre_clip_avg=0.1928 | Metrics: + {'align_loss': 0.025684116408228874, + 'recon_loss': 0.10572549700737, 'predict_loss': + 0.007287686225026846, 'aux_loss_decay_weight': + 0.0, 'grad_norm_pre_clip': 0.25601622462272644, + 'mae_score': 0.007814087738861908, 'data_time': + 0.0005964050069451332, 'model_time': + 1.2323184720007703, 'grad_norm_pre_clip_avg': + 0.19282557368278502, 'learning_rate': + 5.529578617909051e-06, 'epoch': 9.08} +04/20 [00:28:53] INFO | >> train_qwenlatent.py:487 + Step 36010 | grad_norm_pre_clip=0.1782 | + grad_norm_pre_clip_avg=0.1900 | Metrics: + {'align_loss': 0.02615983784198761, + 'recon_loss': 0.1349581927061081, + 'predict_loss': 0.009272308088839054, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17820392549037933, + 'data_time': 0.0010681499843485653, + 'model_time': 1.2276365750003606, + 'grad_norm_pre_clip_avg': 0.18995712399482728, + 'learning_rate': 5.522352837277394e-06, + 'epoch': 9.09} +04/20 [00:29:06] INFO | >> train_qwenlatent.py:487 + Step 36020 | grad_norm_pre_clip=0.1642 | + grad_norm_pre_clip_avg=0.1504 | Metrics: + {'align_loss': 0.024121005088090897, + 'recon_loss': 0.1047784611582756, + 'predict_loss': 0.006919146981090307, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1641916036605835, + 'data_time': 0.0008759069896768779, + 'model_time': 1.2151357289985754, + 'grad_norm_pre_clip_avg': 0.1504399448633194, + 'learning_rate': 5.515130463558546e-06, + 'epoch': 9.09} +04/20 [00:29:18] INFO | >> train_qwenlatent.py:487 + Step 36030 | grad_norm_pre_clip=0.1147 | + grad_norm_pre_clip_avg=0.1571 | Metrics: + {'align_loss': 0.024504080414772034, + 'recon_loss': 0.11687900125980377, + 'predict_loss': 0.005476322025060654, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11474380642175674, + 'data_time': 0.000774512009229511, + 'model_time': 1.2705115669814404, + 'grad_norm_pre_clip_avg': 0.1570713587105274, + 'learning_rate': 5.507911500272602e-06, + 'epoch': 9.09} +04/20 [00:29:31] INFO | >> train_qwenlatent.py:487 + Step 36040 | grad_norm_pre_clip=0.1655 | + grad_norm_pre_clip_avg=0.1665 | Metrics: + {'align_loss': 0.024171901866793633, + 'recon_loss': 0.08682410418987274, + 'predict_loss': 0.004717975854873657, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1655014008283615, + 'data_time': 0.0009370259940624237, + 'model_time': 1.3062684749893378, + 'grad_norm_pre_clip_avg': 0.16646148711442948, + 'learning_rate': 5.500695950938004e-06, + 'epoch': 9.09} +04/20 [00:29:44] INFO | >> train_qwenlatent.py:487 + Step 36050 | grad_norm_pre_clip=0.1899 | + grad_norm_pre_clip_avg=0.1632 | Metrics: + {'align_loss': 0.025770995765924454, + 'recon_loss': 0.10589147359132767, + 'predict_loss': 0.005704331677407026, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18994176387786865, + 'mae_score': 0.007347435135025162, 'data_time': + 0.001032434985972941, 'model_time': + 1.2137920549721457, 'grad_norm_pre_clip_avg': + 0.16320669502019883, 'learning_rate': + 5.4934838190715225e-06, 'epoch': 9.1} +04/20 [00:29:57] INFO | >> train_qwenlatent.py:487 + Step 36060 | grad_norm_pre_clip=0.1513 | + grad_norm_pre_clip_avg=0.1855 | Metrics: + {'align_loss': 0.025438835844397545, + 'recon_loss': 0.099752277135849, + 'predict_loss': 0.004252695944160223, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15130308270454407, + 'data_time': 0.0008498079841956496, + 'model_time': 1.2266710870026145, + 'grad_norm_pre_clip_avg': 0.18552537858486176, + 'learning_rate': 5.4862751081882575e-06, + 'epoch': 9.1} +04/20 [00:30:09] INFO | >> train_qwenlatent.py:487 + Step 36070 | grad_norm_pre_clip=0.1925 | + grad_norm_pre_clip_avg=0.1691 | Metrics: + {'align_loss': 0.025727421045303345, + 'recon_loss': 0.10662374645471573, + 'predict_loss': 0.00949253048747778, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1924656629562378, + 'data_time': 0.0006268929864745587, + 'model_time': 1.1828442900150549, + 'grad_norm_pre_clip_avg': 0.169094280898571, + 'learning_rate': 5.4790698218016424e-06, + 'epoch': 9.1} +04/20 [00:30:22] INFO | >> train_qwenlatent.py:487 + Step 36080 | grad_norm_pre_clip=0.2339 | + grad_norm_pre_clip_avg=0.1747 | Metrics: + {'align_loss': 0.023504581302404404, + 'recon_loss': 0.09785082936286926, + 'predict_loss': 0.005706131923943758, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23394674062728882, + 'data_time': 0.0006367120076902211, + 'model_time': 1.205215122026857, + 'grad_norm_pre_clip_avg': 0.1747487425804138, + 'learning_rate': 5.471867963423458e-06, + 'epoch': 9.1} +04/20 [00:30:34] INFO | >> train_qwenlatent.py:487 + Step 36090 | grad_norm_pre_clip=0.1022 | + grad_norm_pre_clip_avg=0.1431 | Metrics: + {'align_loss': 0.026261920109391212, + 'recon_loss': 0.1131947934627533, + 'predict_loss': 0.006123087368905544, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10215585678815842, + 'data_time': 0.0010464590159244835, + 'model_time': 1.2385073780023959, + 'grad_norm_pre_clip_avg': 0.14311362951993942, + 'learning_rate': 5.464669536563799e-06, + 'epoch': 9.11} +04/20 [00:30:48] INFO | >> train_qwenlatent.py:487 + Step 36100 | grad_norm_pre_clip=0.1607 | + grad_norm_pre_clip_avg=0.1441 | Metrics: + {'align_loss': 0.025779884308576584, + 'recon_loss': 0.1101512610912323, + 'predict_loss': 0.005337467882782221, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1606619954109192, + 'mae_score': 0.007313401849420221, 'data_time': + 0.0008688670059200376, 'model_time': + 1.28846107100253, 'grad_norm_pre_clip_avg': + 0.1441470578312874, 'learning_rate': + 5.457474544731088e-06, 'epoch': 9.11} +04/20 [00:31:00] INFO | >> train_qwenlatent.py:487 + Step 36110 | grad_norm_pre_clip=0.1972 | + grad_norm_pre_clip_avg=0.1553 | Metrics: + {'align_loss': 0.026782792061567307, + 'recon_loss': 0.12057910859584808, + 'predict_loss': 0.006305193994194269, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1972290575504303, + 'data_time': 0.0006157880125101656, + 'model_time': 1.2211984040040988, + 'grad_norm_pre_clip_avg': 0.1553035005927086, + 'learning_rate': 5.450282991432081e-06, + 'epoch': 9.11} +04/20 [00:31:12] INFO | >> train_qwenlatent.py:487 + Step 36120 | grad_norm_pre_clip=0.1312 | + grad_norm_pre_clip_avg=0.1557 | Metrics: + {'align_loss': 0.02713848650455475, + 'recon_loss': 0.0947970300912857, + 'predict_loss': 0.005202857777476311, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1311950832605362, + 'data_time': 0.0010135140037164092, + 'model_time': 1.2530592940165661, + 'grad_norm_pre_clip_avg': 0.15566392689943315, + 'learning_rate': 5.443094880171845e-06, + 'epoch': 9.11} +04/20 [00:31:25] INFO | >> train_qwenlatent.py:487 + Step 36130 | grad_norm_pre_clip=0.1766 | + grad_norm_pre_clip_avg=0.1691 | Metrics: + {'align_loss': 0.027529682964086533, + 'recon_loss': 0.15009966492652893, + 'predict_loss': 0.007313195616006851, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1765691339969635, + 'data_time': 0.0008950479968916625, + 'model_time': 1.2368330309982412, + 'grad_norm_pre_clip_avg': 0.16909752041101456, + 'learning_rate': 5.4359102144537885e-06, + 'epoch': 9.12} +04/20 [00:31:38] INFO | >> train_qwenlatent.py:487 + Step 36140 | grad_norm_pre_clip=0.2134 | + grad_norm_pre_clip_avg=0.1825 | Metrics: + {'align_loss': 0.026069823652505875, + 'recon_loss': 0.16114018857479095, + 'predict_loss': 0.008508304134011269, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21339276432991028, + 'data_time': 0.0006041809974703938, + 'model_time': 1.2309620050073136, + 'grad_norm_pre_clip_avg': 0.18253051489591599, + 'learning_rate': 5.4287289977796306e-06, + 'epoch': 9.12} +04/20 [00:31:51] INFO | >> train_qwenlatent.py:487 + Step 36150 | grad_norm_pre_clip=0.1880 | + grad_norm_pre_clip_avg=0.1647 | Metrics: + {'align_loss': 0.025805506855249405, + 'recon_loss': 0.11079182475805283, + 'predict_loss': 0.004700279328972101, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18801984190940857, + 'mae_score': 0.007919952461311408, 'data_time': + 0.0010537439957261086, 'model_time': + 1.3198761449893937, 'grad_norm_pre_clip_avg': + 0.1647231973707676, 'learning_rate': + 5.421551233649402e-06, 'epoch': 9.12} +04/20 [00:32:04] INFO | >> train_qwenlatent.py:487 + Step 36160 | grad_norm_pre_clip=0.1599 | + grad_norm_pre_clip_avg=0.1653 | Metrics: + {'align_loss': 0.025783397257328033, + 'recon_loss': 0.11269436776638031, + 'predict_loss': 0.008041885681450367, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15985466539859772, + 'data_time': 0.0011355140013620257, + 'model_time': 1.2569972240016796, + 'grad_norm_pre_clip_avg': 0.16525005251169206, + 'learning_rate': 5.414376925561459e-06, + 'epoch': 9.12} +04/20 [00:32:16] INFO | >> train_qwenlatent.py:487 + Step 36170 | grad_norm_pre_clip=0.1928 | + grad_norm_pre_clip_avg=0.1480 | Metrics: + {'align_loss': 0.02525458112359047, + 'recon_loss': 0.09964782744646072, + 'predict_loss': 0.005783822853118181, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19280023872852325, + 'data_time': 0.0006749519961886108, + 'model_time': 1.2369174259947613, + 'grad_norm_pre_clip_avg': 0.1479942336678505, + 'learning_rate': 5.407206077012468e-06, + 'epoch': 9.13} +04/20 [00:32:29] INFO | >> train_qwenlatent.py:487 + Step 36180 | grad_norm_pre_clip=0.0980 | + grad_norm_pre_clip_avg=0.1393 | Metrics: + {'align_loss': 0.025317585095763206, + 'recon_loss': 0.0919550359249115, + 'predict_loss': 0.005548552144318819, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09796087443828583, + 'data_time': 0.0008687390072736889, + 'model_time': 1.204630520020146, + 'grad_norm_pre_clip_avg': 0.1393427699804306, + 'learning_rate': 5.400038691497425e-06, + 'epoch': 9.13} +04/20 [00:32:41] INFO | >> train_qwenlatent.py:487 + Step 36190 | grad_norm_pre_clip=0.1458 | + grad_norm_pre_clip_avg=0.1586 | Metrics: + {'align_loss': 0.024810124188661575, + 'recon_loss': 0.1008734479546547, + 'predict_loss': 0.005788126029074192, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14581085741519928, + 'data_time': 0.001008343999274075, + 'model_time': 1.2333530629985034, + 'grad_norm_pre_clip_avg': 0.1586345911026001, + 'learning_rate': 5.39287477250962e-06, 'epoch': + 9.13} +04/20 [00:32:54] INFO | >> train_qwenlatent.py:487 + Step 36200 | grad_norm_pre_clip=0.2253 | + grad_norm_pre_clip_avg=0.1985 | Metrics: + {'align_loss': 0.0257655531167984, + 'recon_loss': 0.13636916875839233, + 'predict_loss': 0.011422683484852314, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22532407939434052, + 'mae_score': 0.007191002905905784, 'data_time': + 0.0010086690017487854, 'model_time': + 1.2465362570073921, 'grad_norm_pre_clip_avg': + 0.19849207773804664, 'learning_rate': + 5.385714323540661e-06, 'epoch': 9.13} +04/20 [00:33:07] INFO | >> train_qwenlatent.py:487 + Step 36210 | grad_norm_pre_clip=0.1797 | + grad_norm_pre_clip_avg=0.1840 | Metrics: + {'align_loss': 0.025342818349599838, + 'recon_loss': 0.13474033772945404, + 'predict_loss': 0.010477193631231785, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1796879917383194, + 'data_time': 0.0006167799874674529, + 'model_time': 1.1988048310158774, + 'grad_norm_pre_clip_avg': 0.18402690887451173, + 'learning_rate': 5.378557348080463e-06, + 'epoch': 9.14} +04/20 [00:33:20] INFO | >> train_qwenlatent.py:487 + Step 36220 | grad_norm_pre_clip=0.1503 | + grad_norm_pre_clip_avg=0.1739 | Metrics: + {'align_loss': 0.026551101356744766, + 'recon_loss': 0.15306363999843597, + 'predict_loss': 0.01049319002777338, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15025869011878967, + 'data_time': 0.0007173389894887805, + 'model_time': 1.2241564459982328, + 'grad_norm_pre_clip_avg': 0.1739337131381035, + 'learning_rate': 5.3714038496172454e-06, + 'epoch': 9.14} +04/20 [00:33:32] INFO | >> train_qwenlatent.py:487 + Step 36230 | grad_norm_pre_clip=0.1895 | + grad_norm_pre_clip_avg=0.1585 | Metrics: + {'align_loss': 0.02383040264248848, + 'recon_loss': 0.09923472255468369, + 'predict_loss': 0.006527346558868885, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1894753873348236, + 'data_time': 0.0008414160110987723, + 'model_time': 1.1979991410044022, + 'grad_norm_pre_clip_avg': 0.15852489024400712, + 'learning_rate': 5.364253831637549e-06, + 'epoch': 9.14} +04/20 [00:33:45] INFO | >> train_qwenlatent.py:487 + Step 36240 | grad_norm_pre_clip=0.1706 | + grad_norm_pre_clip_avg=0.1824 | Metrics: + {'align_loss': 0.0256655253469944, + 'recon_loss': 0.10039504617452621, + 'predict_loss': 0.008176226168870926, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17062097787857056, + 'data_time': 0.0008833179890643805, + 'model_time': 1.2023755910049658, + 'grad_norm_pre_clip_avg': 0.1824032574892044, + 'learning_rate': 5.357107297626194e-06, + 'epoch': 9.14} +04/20 [00:33:58] INFO | >> train_qwenlatent.py:487 + Step 36250 | grad_norm_pre_clip=0.1440 | + grad_norm_pre_clip_avg=0.1751 | Metrics: + {'align_loss': 0.025738297030329704, + 'recon_loss': 0.1434285044670105, + 'predict_loss': 0.007797751110047102, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14401090145111084, + 'mae_score': 0.007586119196436427, 'data_time': + 0.0006793780194129795, 'model_time': + 1.2875264880130999, 'grad_norm_pre_clip_avg': + 0.1751139432191849, 'learning_rate': + 5.3499642510663155e-06, 'epoch': 9.15} +04/20 [00:34:11] INFO | >> train_qwenlatent.py:487 + Step 36260 | grad_norm_pre_clip=0.1421 | + grad_norm_pre_clip_avg=0.1625 | Metrics: + {'align_loss': 0.02369328960776329, + 'recon_loss': 0.08310677856206894, + 'predict_loss': 0.006020853761583567, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14206606149673462, + 'data_time': 0.0009095720015466213, + 'model_time': 1.506761547003407, + 'grad_norm_pre_clip_avg': 0.16254621595144272, + 'learning_rate': 5.342824695439346e-06, + 'epoch': 9.15} +04/20 [00:34:24] INFO | >> train_qwenlatent.py:487 + Step 36270 | grad_norm_pre_clip=0.1603 | + grad_norm_pre_clip_avg=0.1601 | Metrics: + {'align_loss': 0.026266813278198242, + 'recon_loss': 0.1386909931898117, + 'predict_loss': 0.004549490287899971, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1602928191423416, + 'data_time': 0.0008442680118605494, + 'model_time': 1.2308808569796383, + 'grad_norm_pre_clip_avg': 0.16005165576934816, + 'learning_rate': 5.335688634225027e-06, + 'epoch': 9.15} +04/20 [00:34:36] INFO | >> train_qwenlatent.py:487 + Step 36280 | grad_norm_pre_clip=0.1339 | + grad_norm_pre_clip_avg=0.1579 | Metrics: + {'align_loss': 0.02513810433447361, + 'recon_loss': 0.10730361193418503, + 'predict_loss': 0.005104066804051399, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13390734791755676, + 'data_time': 0.001042532006977126, + 'model_time': 1.2410874369961675, + 'grad_norm_pre_clip_avg': 0.15792482420802118, + 'learning_rate': 5.328556070901381e-06, + 'epoch': 9.15} +04/20 [00:34:49] INFO | >> train_qwenlatent.py:487 + Step 36290 | grad_norm_pre_clip=0.2115 | + grad_norm_pre_clip_avg=0.1747 | Metrics: + {'align_loss': 0.024844257161021233, + 'recon_loss': 0.12804104387760162, + 'predict_loss': 0.009338118135929108, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2115398347377777, + 'data_time': 0.0008325839880853891, + 'model_time': 1.2079317019961309, + 'grad_norm_pre_clip_avg': 0.17466767132282257, + 'learning_rate': 5.321427008944736e-06, + 'epoch': 9.16} +04/20 [00:35:02] INFO | >> train_qwenlatent.py:487 + Step 36300 | grad_norm_pre_clip=0.1680 | + grad_norm_pre_clip_avg=0.1866 | Metrics: + {'align_loss': 0.02622414380311966, + 'recon_loss': 0.11435288190841675, + 'predict_loss': 0.006138678174465895, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1680409461259842, + 'mae_score': 0.007011219402691266, 'data_time': + 0.0006545869982801378, 'model_time': + 1.2524473170051351, 'grad_norm_pre_clip_avg': + 0.18658135682344437, 'learning_rate': + 5.314301451829707e-06, 'epoch': 9.16} +04/20 [00:35:14] INFO | >> train_qwenlatent.py:487 + Step 36310 | grad_norm_pre_clip=0.1911 | + grad_norm_pre_clip_avg=0.1551 | Metrics: + {'align_loss': 0.0258328914642334, + 'recon_loss': 0.11424513161182404, + 'predict_loss': 0.009647244587540627, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1910906583070755, + 'data_time': 0.0006528719968628138, + 'model_time': 1.2239458019903395, + 'grad_norm_pre_clip_avg': 0.15514734387397766, + 'learning_rate': 5.307179403029208e-06, + 'epoch': 9.16} +04/20 [00:35:27] INFO | >> train_qwenlatent.py:487 + Step 36320 | grad_norm_pre_clip=0.1320 | + grad_norm_pre_clip_avg=0.1461 | Metrics: + {'align_loss': 0.025165166705846786, + 'recon_loss': 0.11997351050376892, + 'predict_loss': 0.005199674982577562, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13195960223674774, + 'data_time': 0.0014044629933778197, + 'model_time': 1.2163641189981718, + 'grad_norm_pre_clip_avg': 0.14609621465206146, + 'learning_rate': 5.300060866014436e-06, + 'epoch': 9.16} +04/20 [00:35:40] INFO | >> train_qwenlatent.py:487 + Step 36330 | grad_norm_pre_clip=0.1500 | + grad_norm_pre_clip_avg=0.1586 | Metrics: + {'align_loss': 0.025212597101926804, + 'recon_loss': 0.11245791614055634, + 'predict_loss': 0.010268883779644966, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15003572404384613, + 'data_time': 0.0010746250045485795, + 'model_time': 1.319170161994407, + 'grad_norm_pre_clip_avg': 0.1586415097117424, + 'learning_rate': 5.2929458442548795e-06, + 'epoch': 9.17} +04/20 [00:35:52] INFO | >> train_qwenlatent.py:487 + Step 36340 | grad_norm_pre_clip=0.1470 | + grad_norm_pre_clip_avg=0.1438 | Metrics: + {'align_loss': 0.0254988931119442, + 'recon_loss': 0.14300812780857086, + 'predict_loss': 0.0066098542883992195, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14702868461608887, + 'data_time': 0.0007331450178753585, + 'model_time': 1.240912410983583, + 'grad_norm_pre_clip_avg': 0.1438110128045082, + 'learning_rate': 5.285834341218314e-06, + 'epoch': 9.17} +04/20 [00:36:06] INFO | >> train_qwenlatent.py:487 + Step 36350 | grad_norm_pre_clip=0.1850 | + grad_norm_pre_clip_avg=0.1530 | Metrics: + {'align_loss': 0.02507881447672844, + 'recon_loss': 0.15547829866409302, + 'predict_loss': 0.01164031308144331, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18499545753002167, + 'mae_score': 0.007395960833575275, 'data_time': + 0.0009841610153671354, 'model_time': + 1.1909371030051261, 'grad_norm_pre_clip_avg': + 0.15297108590602876, 'learning_rate': + 5.278726360370801e-06, 'epoch': 9.17} +04/20 [00:36:18] INFO | >> train_qwenlatent.py:487 + Step 36360 | grad_norm_pre_clip=0.1579 | + grad_norm_pre_clip_avg=0.1784 | Metrics: + {'align_loss': 0.02696032077074051, + 'recon_loss': 0.11865311861038208, + 'predict_loss': 0.006640475708991289, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15785574913024902, + 'data_time': 0.0009383690194226801, + 'model_time': 1.222620470012771, + 'grad_norm_pre_clip_avg': 0.17844272702932357, + 'learning_rate': 5.27162190517668e-06, 'epoch': + 9.17} +04/20 [00:36:31] INFO | >> train_qwenlatent.py:487 + Step 36370 | grad_norm_pre_clip=0.1366 | + grad_norm_pre_clip_avg=0.1705 | Metrics: + {'align_loss': 0.025437839329242706, + 'recon_loss': 0.09708913415670395, + 'predict_loss': 0.00687713548541069, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13658317923545837, + 'data_time': 0.0006946090143173933, + 'model_time': 1.2500227910059039, + 'grad_norm_pre_clip_avg': 0.17051760852336884, + 'learning_rate': 5.264520979098581e-06, + 'epoch': 9.18} +04/20 [00:36:43] INFO | >> train_qwenlatent.py:487 + Step 36380 | grad_norm_pre_clip=0.1230 | + grad_norm_pre_clip_avg=0.1661 | Metrics: + {'align_loss': 0.025613203644752502, + 'recon_loss': 0.11937275528907776, + 'predict_loss': 0.007535929325968027, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12304582446813583, + 'data_time': 0.0009625559905543923, + 'model_time': 1.2392191069957335, + 'grad_norm_pre_clip_avg': 0.16614067181944847, + 'learning_rate': 5.257423585597408e-06, + 'epoch': 9.18} +04/20 [00:36:56] INFO | >> train_qwenlatent.py:487 + Step 36390 | grad_norm_pre_clip=0.1505 | + grad_norm_pre_clip_avg=0.1890 | Metrics: + {'align_loss': 0.027062315493822098, + 'recon_loss': 0.12330113351345062, + 'predict_loss': 0.009397133253514767, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1505383998155594, + 'data_time': 0.0009163490030914545, + 'model_time': 1.2664558129908983, + 'grad_norm_pre_clip_avg': 0.1890239030122757, + 'learning_rate': 5.250329728132345e-06, + 'epoch': 9.18} +04/20 [00:37:09] INFO | >> train_qwenlatent.py:487 + Step 36400 | grad_norm_pre_clip=0.1365 | + grad_norm_pre_clip_avg=0.1864 | Metrics: + {'align_loss': 0.023337434977293015, + 'recon_loss': 0.08430178463459015, + 'predict_loss': 0.004964957479387522, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13653039932250977, + 'mae_score': 0.006508169517860756, 'data_time': + 0.0006586849922314286, 'model_time': + 1.2879221150069498, 'grad_norm_pre_clip_avg': + 0.18640711009502411, 'learning_rate': + 5.24323941016085e-06, 'epoch': 9.18} +04/20 [00:37:22] INFO | >> train_qwenlatent.py:487 + Step 36410 | grad_norm_pre_clip=0.2168 | + grad_norm_pre_clip_avg=0.1981 | Metrics: + {'align_loss': 0.025087300688028336, + 'recon_loss': 0.12392950803041458, + 'predict_loss': 0.00692932540550828, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21677038073539734, + 'data_time': 0.0006528560188598931, + 'model_time': 1.2179617589863483, + 'grad_norm_pre_clip_avg': 0.1981019839644432, + 'learning_rate': 5.236152635138659e-06, + 'epoch': 9.19} +04/20 [00:37:35] INFO | >> train_qwenlatent.py:487 + Step 36420 | grad_norm_pre_clip=0.2697 | + grad_norm_pre_clip_avg=0.1841 | Metrics: + {'align_loss': 0.026655809953808784, + 'recon_loss': 0.18244674801826477, + 'predict_loss': 0.010881142690777779, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2696920335292816, + 'data_time': 0.0010141129896510392, + 'model_time': 1.2475079919968266, + 'grad_norm_pre_clip_avg': 0.18412379026412964, + 'learning_rate': 5.229069406519777e-06, + 'epoch': 9.19} +04/20 [00:37:47] INFO | >> train_qwenlatent.py:487 + Step 36430 | grad_norm_pre_clip=0.1164 | + grad_norm_pre_clip_avg=0.1527 | Metrics: + {'align_loss': 0.026616834104061127, + 'recon_loss': 0.1276734620332718, + 'predict_loss': 0.011136560700833797, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11637905985116959, + 'data_time': 0.000655761017696932, + 'model_time': 1.2053960810007993, + 'grad_norm_pre_clip_avg': 0.1527256913483143, + 'learning_rate': 5.221989727756489e-06, + 'epoch': 9.19} +04/20 [00:37:59] INFO | >> train_qwenlatent.py:487 + Step 36440 | grad_norm_pre_clip=0.1198 | + grad_norm_pre_clip_avg=0.1360 | Metrics: + {'align_loss': 0.024523377418518066, + 'recon_loss': 0.05708366259932518, + 'predict_loss': 0.0030783407855778933, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11976286768913269, + 'data_time': 0.0007681169954594225, + 'model_time': 1.2636457580083515, + 'grad_norm_pre_clip_avg': 0.1360131211578846, + 'learning_rate': 5.21491360229934e-06, 'epoch': + 9.2} +04/20 [00:38:13] INFO | >> train_qwenlatent.py:487 + Step 36450 | grad_norm_pre_clip=0.1153 | + grad_norm_pre_clip_avg=0.1537 | Metrics: + {'align_loss': 0.025686804205179214, + 'recon_loss': 0.12017892301082611, + 'predict_loss': 0.005806059576570988, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1153261661529541, + 'mae_score': 0.006701440639324017, 'data_time': + 0.0009760510001797229, 'model_time': + 1.2430402519821655, 'grad_norm_pre_clip_avg': + 0.1536679558455944, 'learning_rate': + 5.207841033597147e-06, 'epoch': 9.2} +04/20 [00:38:26] INFO | >> train_qwenlatent.py:487 + Step 36460 | grad_norm_pre_clip=0.1957 | + grad_norm_pre_clip_avg=0.1587 | Metrics: + {'align_loss': 0.02640969306230545, + 'recon_loss': 0.09910843521356583, + 'predict_loss': 0.00660578440874815, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19567832350730896, + 'data_time': 0.0009965260105673224, + 'model_time': 1.5860211160033941, + 'grad_norm_pre_clip_avg': 0.15869873613119126, + 'learning_rate': 5.200772025096999e-06, + 'epoch': 9.2} +04/20 [00:38:38] INFO | >> train_qwenlatent.py:487 + Step 36470 | grad_norm_pre_clip=0.1334 | + grad_norm_pre_clip_avg=0.1619 | Metrics: + {'align_loss': 0.025365615263581276, + 'recon_loss': 0.13576839864253998, + 'predict_loss': 0.006015484686940908, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13343891501426697, + 'data_time': 0.0011882300022989511, + 'model_time': 1.2105113329889718, + 'grad_norm_pre_clip_avg': 0.16188514232635498, + 'learning_rate': 5.193706580244242e-06, + 'epoch': 9.2} +04/20 [00:38:51] INFO | >> train_qwenlatent.py:487 + Step 36480 | grad_norm_pre_clip=0.1694 | + grad_norm_pre_clip_avg=0.2039 | Metrics: + {'align_loss': 0.02524949237704277, + 'recon_loss': 0.11337319761514664, + 'predict_loss': 0.006914537400007248, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16938559710979462, + 'data_time': 0.0009454720129724592, + 'model_time': 1.2555798839894123, + 'grad_norm_pre_clip_avg': 0.20393769741058348, + 'learning_rate': 5.186644702482489e-06, + 'epoch': 9.21} +04/20 [00:39:04] INFO | >> train_qwenlatent.py:487 + Step 36490 | grad_norm_pre_clip=0.1957 | + grad_norm_pre_clip_avg=0.1835 | Metrics: + {'align_loss': 0.025480780750513077, + 'recon_loss': 0.11581788212060928, + 'predict_loss': 0.0063523221760988235, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19574710726737976, + 'data_time': 0.000705454993294552, + 'model_time': 1.227814676007256, + 'grad_norm_pre_clip_avg': 0.18345749527215957, + 'learning_rate': 5.17958639525361e-06, 'epoch': + 9.21} +04/20 [00:39:17] INFO | >> train_qwenlatent.py:487 + Step 36500 | grad_norm_pre_clip=0.1135 | + grad_norm_pre_clip_avg=0.1656 | Metrics: + {'align_loss': 0.026532629504799843, + 'recon_loss': 0.09280471503734589, + 'predict_loss': 0.005441504996269941, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11348544806241989, + 'mae_score': 0.006797471347155872, 'data_time': + 0.0006444489990826696, 'model_time': + 1.1997427040187176, 'grad_norm_pre_clip_avg': + 0.16556437239050864, 'learning_rate': + 5.172531661997743e-06, 'epoch': 9.21} +04/20 [00:39:29] INFO | >> train_qwenlatent.py:487 + Step 36510 | grad_norm_pre_clip=0.1796 | + grad_norm_pre_clip_avg=0.1426 | Metrics: + {'align_loss': 0.025597818195819855, + 'recon_loss': 0.13541282713413239, + 'predict_loss': 0.006421332247555256, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17963382601737976, + 'data_time': 0.0007318920106627047, + 'model_time': 1.2417873050144408, + 'grad_norm_pre_clip_avg': 0.1426267944276333, + 'learning_rate': 5.165480506153277e-06, + 'epoch': 9.21} +04/20 [00:39:42] INFO | >> train_qwenlatent.py:487 + Step 36520 | grad_norm_pre_clip=0.2191 | + grad_norm_pre_clip_avg=0.1916 | Metrics: + {'align_loss': 0.024466969072818756, + 'recon_loss': 0.12843702733516693, + 'predict_loss': 0.00861175823956728, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2191217690706253, + 'data_time': 0.0006322010012809187, + 'model_time': 1.2405966719961725, + 'grad_norm_pre_clip_avg': 0.19156274795532227, + 'learning_rate': 5.158432931156859e-06, + 'epoch': 9.22} +04/20 [00:39:54] INFO | >> train_qwenlatent.py:487 + Step 36530 | grad_norm_pre_clip=0.1798 | + grad_norm_pre_clip_avg=0.2221 | Metrics: + {'align_loss': 0.024907466024160385, + 'recon_loss': 0.0924663320183754, + 'predict_loss': 0.0042792861349880695, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17984077334403992, + 'data_time': 0.0010832859843503684, + 'model_time': 1.2255428280041087, + 'grad_norm_pre_clip_avg': 0.22207062393426896, + 'learning_rate': 5.151388940443392e-06, + 'epoch': 9.22} +04/20 [00:40:07] INFO | >> train_qwenlatent.py:487 + Step 36540 | grad_norm_pre_clip=0.1954 | + grad_norm_pre_clip_avg=0.1932 | Metrics: + {'align_loss': 0.026210036128759384, + 'recon_loss': 0.13463833928108215, + 'predict_loss': 0.010884650982916355, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19537638127803802, + 'data_time': 0.000802044989541173, + 'model_time': 1.2177201209997293, + 'grad_norm_pre_clip_avg': 0.1932307332754135, + 'learning_rate': 5.1443485374460325e-06, + 'epoch': 9.22} +04/20 [00:40:20] INFO | >> train_qwenlatent.py:487 + Step 36550 | grad_norm_pre_clip=0.1349 | + grad_norm_pre_clip_avg=0.1681 | Metrics: + {'align_loss': 0.025643549859523773, + 'recon_loss': 0.13273131847381592, + 'predict_loss': 0.007887125946581364, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13493067026138306, + 'mae_score': 0.006775851722236152, 'data_time': + 0.0010791679960675538, 'model_time': + 1.2133683289866894, 'grad_norm_pre_clip_avg': + 0.16814501136541365, 'learning_rate': + 5.137311725596181e-06, 'epoch': 9.22} +04/20 [00:40:33] INFO | >> train_qwenlatent.py:487 + Step 36560 | grad_norm_pre_clip=0.1235 | + grad_norm_pre_clip_avg=0.1416 | Metrics: + {'align_loss': 0.024693474173545837, + 'recon_loss': 0.12821869552135468, + 'predict_loss': 0.006980982609093189, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12350630760192871, + 'data_time': 0.0010499389900360256, + 'model_time': 1.224878925975645, + 'grad_norm_pre_clip_avg': 0.14158836975693703, + 'learning_rate': 5.130278508323504e-06, + 'epoch': 9.23} +04/20 [00:40:45] INFO | >> train_qwenlatent.py:487 + Step 36570 | grad_norm_pre_clip=0.2097 | + grad_norm_pre_clip_avg=0.1545 | Metrics: + {'align_loss': 0.02618042379617691, + 'recon_loss': 0.11803356558084488, + 'predict_loss': 0.00787085946649313, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20971205830574036, + 'data_time': 0.0011807949922513217, + 'model_time': 1.1896967009815853, + 'grad_norm_pre_clip_avg': 0.1545203670859337, + 'learning_rate': 5.1232488890559035e-06, + 'epoch': 9.23} +04/20 [00:40:58] INFO | >> train_qwenlatent.py:487 + Step 36580 | grad_norm_pre_clip=0.1234 | + grad_norm_pre_clip_avg=0.1583 | Metrics: + {'align_loss': 0.02543140947818756, + 'recon_loss': 0.11281159520149231, + 'predict_loss': 0.0056947628036141396, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12337136268615723, + 'data_time': 0.0006418580014724284, + 'model_time': 1.1977417079906445, + 'grad_norm_pre_clip_avg': 0.1582807943224907, + 'learning_rate': 5.116222871219531e-06, + 'epoch': 9.23} +04/20 [00:41:10] INFO | >> train_qwenlatent.py:487 + Step 36590 | grad_norm_pre_clip=0.1884 | + grad_norm_pre_clip_avg=0.1636 | Metrics: + {'align_loss': 0.02638348564505577, + 'recon_loss': 0.11562599241733551, + 'predict_loss': 0.007737476844340563, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18840639293193817, + 'data_time': 0.0012861969880759716, + 'model_time': 1.219306738988962, + 'grad_norm_pre_clip_avg': 0.16355767622590064, + 'learning_rate': 5.109200458238773e-06, + 'epoch': 9.23} +04/20 [00:41:24] INFO | >> train_qwenlatent.py:487 + Step 36600 | grad_norm_pre_clip=0.1930 | + grad_norm_pre_clip_avg=0.1715 | Metrics: + {'align_loss': 0.025961551815271378, + 'recon_loss': 0.10320151597261429, + 'predict_loss': 0.0058831507340073586, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19297288358211517, + 'mae_score': 0.007321199640497431, 'data_time': + 0.0009726659918669611, 'model_time': + 1.2106111900066026, 'grad_norm_pre_clip_avg': + 0.17154271453619002, 'learning_rate': + 5.10218165353628e-06, 'epoch': 9.24} +04/20 [00:41:37] INFO | >> train_qwenlatent.py:487 + Step 36610 | grad_norm_pre_clip=0.1688 | + grad_norm_pre_clip_avg=0.1736 | Metrics: + {'align_loss': 0.025783821940422058, + 'recon_loss': 0.12178029119968414, + 'predict_loss': 0.00887279212474823, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16882506012916565, + 'data_time': 0.0007758839929010719, + 'model_time': 1.2983618760190438, + 'grad_norm_pre_clip_avg': 0.17355093955993653, + 'learning_rate': 5.095166460532927e-06, + 'epoch': 9.24} +04/20 [00:41:49] INFO | >> train_qwenlatent.py:487 + Step 36620 | grad_norm_pre_clip=0.1663 | + grad_norm_pre_clip_avg=0.1718 | Metrics: + {'align_loss': 0.023768354207277298, + 'recon_loss': 0.09284994006156921, + 'predict_loss': 0.008829445578157902, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16634435951709747, + 'data_time': 0.0009659490024205297, + 'model_time': 1.2572738799790386, + 'grad_norm_pre_clip_avg': 0.17179655879735947, + 'learning_rate': 5.088154882647834e-06, + 'epoch': 9.24} +04/20 [00:42:02] INFO | >> train_qwenlatent.py:487 + Step 36630 | grad_norm_pre_clip=0.1343 | + grad_norm_pre_clip_avg=0.1777 | Metrics: + {'align_loss': 0.02477831393480301, + 'recon_loss': 0.09277112036943436, + 'predict_loss': 0.0068438854068517685, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13428552448749542, + 'data_time': 0.0007567070133518428, + 'model_time': 1.2582494620000944, + 'grad_norm_pre_clip_avg': 0.17769857496023178, + 'learning_rate': 5.0811469232983585e-06, + 'epoch': 9.24} +04/20 [00:42:14] INFO | >> train_qwenlatent.py:487 + Step 36640 | grad_norm_pre_clip=0.1604 | + grad_norm_pre_clip_avg=0.1752 | Metrics: + {'align_loss': 0.02516501396894455, + 'recon_loss': 0.06539727747440338, + 'predict_loss': 0.0029465481638908386, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16037683188915253, + 'data_time': 0.0008692870032973588, + 'model_time': 1.253028473991435, + 'grad_norm_pre_clip_avg': 0.17516290098428727, + 'learning_rate': 5.074142585900093e-06, + 'epoch': 9.25} +04/20 [00:42:28] INFO | >> train_qwenlatent.py:487 + Step 36650 | grad_norm_pre_clip=0.1227 | + grad_norm_pre_clip_avg=0.1428 | Metrics: + {'align_loss': 0.023909971117973328, + 'recon_loss': 0.11684313416481018, + 'predict_loss': 0.010413791984319687, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12267310172319412, + 'mae_score': 0.009549439060795415, 'data_time': + 0.0008901580004021525, 'model_time': + 1.2325565399951302, 'grad_norm_pre_clip_avg': + 0.1428146705031395, 'learning_rate': + 5.067141873866871e-06, 'epoch': 9.25} +04/20 [00:42:40] INFO | >> train_qwenlatent.py:487 + Step 36660 | grad_norm_pre_clip=0.1748 | + grad_norm_pre_clip_avg=0.1668 | Metrics: + {'align_loss': 0.025844279676675797, + 'recon_loss': 0.15020038187503815, + 'predict_loss': 0.008331588469445705, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17476886510849, + 'data_time': 0.0009364589932374656, + 'model_time': 1.2483014180033933, + 'grad_norm_pre_clip_avg': 0.16679348796606064, + 'learning_rate': 5.060144790610754e-06, + 'epoch': 9.25} +04/20 [00:42:53] INFO | >> train_qwenlatent.py:487 + Step 36670 | grad_norm_pre_clip=0.1796 | + grad_norm_pre_clip_avg=0.1618 | Metrics: + {'align_loss': 0.026564117521047592, + 'recon_loss': 0.14417502284049988, + 'predict_loss': 0.008419157937169075, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17964082956314087, + 'data_time': 0.0008621290035080165, + 'model_time': 1.2530867349996697, + 'grad_norm_pre_clip_avg': 0.16179394125938415, + 'learning_rate': 5.0531513395420375e-06, + 'epoch': 9.25} +04/20 [00:43:05] INFO | >> train_qwenlatent.py:487 + Step 36680 | grad_norm_pre_clip=0.1427 | + grad_norm_pre_clip_avg=0.1733 | Metrics: + {'align_loss': 0.024613376706838608, + 'recon_loss': 0.10867966711521149, + 'predict_loss': 0.007667058613151312, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14274758100509644, + 'data_time': 0.0006625349924433976, + 'model_time': 1.245136701996671, + 'grad_norm_pre_clip_avg': 0.1732747159898281, + 'learning_rate': 5.046161524069238e-06, + 'epoch': 9.26} +04/20 [00:43:18] INFO | >> train_qwenlatent.py:487 + Step 36690 | grad_norm_pre_clip=0.1689 | + grad_norm_pre_clip_avg=0.1637 | Metrics: + {'align_loss': 0.024653146043419838, + 'recon_loss': 0.11978249996900558, + 'predict_loss': 0.00657914113253355, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1688975989818573, + 'data_time': 0.0009031550143845379, + 'model_time': 1.2134284100029618, + 'grad_norm_pre_clip_avg': 0.16367349475622178, + 'learning_rate': 5.0391753475991034e-06, + 'epoch': 9.26} +04/20 [00:43:31] INFO | >> train_qwenlatent.py:487 + Step 36700 | grad_norm_pre_clip=0.1564 | + grad_norm_pre_clip_avg=0.1613 | Metrics: + {'align_loss': 0.024592852219939232, + 'recon_loss': 0.09326000511646271, + 'predict_loss': 0.005703537724912167, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15643911063671112, + 'mae_score': 0.005793785404514622, 'data_time': + 0.0009865589963737875, 'model_time': + 1.2714150310202967, 'grad_norm_pre_clip_avg': + 0.16132044792175293, 'learning_rate': + 5.032192813536623e-06, 'epoch': 9.26} +04/20 [00:43:44] INFO | >> train_qwenlatent.py:487 + Step 36710 | grad_norm_pre_clip=0.1365 | + grad_norm_pre_clip_avg=0.1412 | Metrics: + {'align_loss': 0.025796273723244667, + 'recon_loss': 0.09520436078310013, + 'predict_loss': 0.006149544846266508, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13647450506687164, + 'data_time': 0.000764569005696103, + 'model_time': 1.2826749160012696, + 'grad_norm_pre_clip_avg': 0.1412388876080513, + 'learning_rate': 5.025213925284993e-06, + 'epoch': 9.26} +04/20 [00:43:56] INFO | >> train_qwenlatent.py:487 + Step 36720 | grad_norm_pre_clip=0.1390 | + grad_norm_pre_clip_avg=0.1458 | Metrics: + {'align_loss': 0.02581457793712616, + 'recon_loss': 0.13682542741298676, + 'predict_loss': 0.006328659597784281, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13904693722724915, + 'data_time': 0.0007514280150644481, + 'model_time': 1.2115896869800054, + 'grad_norm_pre_clip_avg': 0.14577258974313737, + 'learning_rate': 5.018238686245638e-06, + 'epoch': 9.27} +04/20 [00:44:09] INFO | >> train_qwenlatent.py:487 + Step 36730 | grad_norm_pre_clip=0.2590 | + grad_norm_pre_clip_avg=0.1805 | Metrics: + {'align_loss': 0.02575274184346199, + 'recon_loss': 0.10729548335075378, + 'predict_loss': 0.008231566287577152, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2590167224407196, + 'data_time': 0.0007413220009766519, + 'model_time': 1.2772592620167416, + 'grad_norm_pre_clip_avg': 0.18045982494950294, + 'learning_rate': 5.0112670998182075e-06, + 'epoch': 9.27} +04/20 [00:44:22] INFO | >> train_qwenlatent.py:487 + Step 36740 | grad_norm_pre_clip=0.2122 | + grad_norm_pre_clip_avg=0.1795 | Metrics: + {'align_loss': 0.024403106421232224, + 'recon_loss': 0.07529962062835693, + 'predict_loss': 0.0047392998822033405, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.212211474776268, + 'data_time': 0.0008783990051597357, + 'model_time': 1.3111766559886746, + 'grad_norm_pre_clip_avg': 0.17949390709400176, + 'learning_rate': 5.004299169400563e-06, + 'epoch': 9.27} +04/20 [00:44:35] INFO | >> train_qwenlatent.py:487 + Step 36750 | grad_norm_pre_clip=0.1109 | + grad_norm_pre_clip_avg=0.1579 | Metrics: + {'align_loss': 0.025332678109407425, + 'recon_loss': 0.17422735691070557, + 'predict_loss': 0.008664856664836407, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11093349009752274, + 'mae_score': 0.007479106628143036, 'data_time': + 0.0008620000153314322, 'model_time': + 1.231357051001396, 'grad_norm_pre_clip_avg': + 0.1578669987618923, 'learning_rate': + 4.997334898388796e-06, 'epoch': 9.27} +04/20 [00:44:48] INFO | >> train_qwenlatent.py:487 + Step 36760 | grad_norm_pre_clip=0.1354 | + grad_norm_pre_clip_avg=0.1762 | Metrics: + {'align_loss': 0.02471790835261345, + 'recon_loss': 0.06331520527601242, + 'predict_loss': 0.0033691313583403826, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1353825479745865, + 'data_time': 0.0006177669856697321, + 'model_time': 1.2356851439981256, + 'grad_norm_pre_clip_avg': 0.17615434378385544, + 'learning_rate': 4.990374290177208e-06, + 'epoch': 9.28} +04/20 [00:45:01] INFO | >> train_qwenlatent.py:487 + Step 36770 | grad_norm_pre_clip=0.1831 | + grad_norm_pre_clip_avg=0.1804 | Metrics: + {'align_loss': 0.025331735610961914, + 'recon_loss': 0.22668001055717468, + 'predict_loss': 0.015693048015236855, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1831299513578415, + 'data_time': 0.0006572320125997066, + 'model_time': 1.2865706919983495, + 'grad_norm_pre_clip_avg': 0.1803954504430294, + 'learning_rate': 4.983417348158307e-06, + 'epoch': 9.28} +04/20 [00:45:13] INFO | >> train_qwenlatent.py:487 + Step 36780 | grad_norm_pre_clip=0.1716 | + grad_norm_pre_clip_avg=0.1625 | Metrics: + {'align_loss': 0.025401627644896507, + 'recon_loss': 0.11050998419523239, + 'predict_loss': 0.007773083169013262, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1716362088918686, + 'data_time': 0.0006589110125787556, + 'model_time': 1.2038615110213868, + 'grad_norm_pre_clip_avg': 0.16252498775720597, + 'learning_rate': 4.976464075722824e-06, + 'epoch': 9.28} +04/20 [00:45:26] INFO | >> train_qwenlatent.py:487 + Step 36790 | grad_norm_pre_clip=0.1497 | + grad_norm_pre_clip_avg=0.1760 | Metrics: + {'align_loss': 0.02472309023141861, + 'recon_loss': 0.1292603462934494, + 'predict_loss': 0.0083770165219903, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14965315163135529, + 'data_time': 0.000955508992774412, + 'model_time': 1.252011106989812, + 'grad_norm_pre_clip_avg': 0.17596246302127838, + 'learning_rate': 4.969514476259703e-06, + 'epoch': 9.28} +04/20 [00:45:39] INFO | >> train_qwenlatent.py:487 + Step 36800 | grad_norm_pre_clip=0.1559 | + grad_norm_pre_clip_avg=0.1652 | Metrics: + {'align_loss': 0.025197044014930725, + 'recon_loss': 0.1773327738046646, + 'predict_loss': 0.010155926458537579, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15594905614852905, + 'mae_score': 0.007018304945112349, 'data_time': + 0.0009349320025648922, 'model_time': + 1.2629783110169228, 'grad_norm_pre_clip_avg': + 0.16524552255868913, 'learning_rate': + 4.962568553156095e-06, 'epoch': 9.29} +04/20 [00:45:52] INFO | >> train_qwenlatent.py:487 + Step 36810 | grad_norm_pre_clip=0.1249 | + grad_norm_pre_clip_avg=0.1622 | Metrics: + {'align_loss': 0.026197057217359543, + 'recon_loss': 0.12381429970264435, + 'predict_loss': 0.003783569671213627, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1249103844165802, + 'data_time': 0.001025859994115308, + 'model_time': 1.230710799980443, + 'grad_norm_pre_clip_avg': 0.16221001371741295, + 'learning_rate': 4.955626309797355e-06, + 'epoch': 9.29} +04/20 [00:46:04] INFO | >> train_qwenlatent.py:487 + Step 36820 | grad_norm_pre_clip=0.1795 | + grad_norm_pre_clip_avg=0.1853 | Metrics: + {'align_loss': 0.026747629046440125, + 'recon_loss': 0.1766292154788971, + 'predict_loss': 0.011009303852915764, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17950062453746796, + 'data_time': 0.0007918500050436705, + 'model_time': 1.1912041790201329, + 'grad_norm_pre_clip_avg': 0.18527772575616835, + 'learning_rate': 4.9486877495670505e-06, + 'epoch': 9.29} +04/20 [00:46:17] INFO | >> train_qwenlatent.py:487 + Step 36830 | grad_norm_pre_clip=0.1732 | + grad_norm_pre_clip_avg=0.1590 | Metrics: + {'align_loss': 0.024901440367102623, + 'recon_loss': 0.09562243521213531, + 'predict_loss': 0.006134268827736378, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17320910096168518, + 'data_time': 0.0009218220075126737, + 'model_time': 1.26199369400274, + 'grad_norm_pre_clip_avg': 0.15895671844482423, + 'learning_rate': 4.941752875846951e-06, + 'epoch': 9.29} +04/20 [00:46:29] INFO | >> train_qwenlatent.py:487 + Step 36840 | grad_norm_pre_clip=0.2038 | + grad_norm_pre_clip_avg=0.1694 | Metrics: + {'align_loss': 0.025559382513165474, + 'recon_loss': 0.11202733963727951, + 'predict_loss': 0.004859365057200193, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20379048585891724, + 'data_time': 0.0006895579863339663, + 'model_time': 1.2312702469935175, + 'grad_norm_pre_clip_avg': 0.16938708424568177, + 'learning_rate': 4.9348216920170254e-06, + 'epoch': 9.3} +04/20 [00:46:42] INFO | >> train_qwenlatent.py:487 + Step 36850 | grad_norm_pre_clip=0.1382 | + grad_norm_pre_clip_avg=0.1660 | Metrics: + {'align_loss': 0.025970470160245895, + 'recon_loss': 0.1363060623407364, + 'predict_loss': 0.0081290602684021, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13820914924144745, + 'mae_score': 0.0072495546426858985, + 'data_time': 0.0006637669866904616, + 'model_time': 1.2453771559812594, + 'grad_norm_pre_clip_avg': 0.16598010659217835, + 'learning_rate': 4.927894201455461e-06, + 'epoch': 9.3} +04/20 [00:46:55] INFO | >> train_qwenlatent.py:487 + Step 36860 | grad_norm_pre_clip=0.1478 | + grad_norm_pre_clip_avg=0.1631 | Metrics: + {'align_loss': 0.025592951104044914, + 'recon_loss': 0.1538870483636856, + 'predict_loss': 0.010236774571239948, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14776542782783508, + 'data_time': 0.0009618779877200723, + 'model_time': 1.1955021769972518, + 'grad_norm_pre_clip_avg': 0.16307317912578584, + 'learning_rate': 4.920970407538619e-06, + 'epoch': 9.3} +04/20 [00:47:08] INFO | >> train_qwenlatent.py:487 + Step 36870 | grad_norm_pre_clip=0.1503 | + grad_norm_pre_clip_avg=0.1488 | Metrics: + {'align_loss': 0.02382173389196396, + 'recon_loss': 0.07913509011268616, + 'predict_loss': 0.004220949485898018, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15031249821186066, + 'data_time': 0.0009734350023791194, + 'model_time': 1.2725063889811281, + 'grad_norm_pre_clip_avg': 0.1487737111747265, + 'learning_rate': 4.914050313641077e-06, + 'epoch': 9.3} +04/20 [00:47:21] INFO | >> train_qwenlatent.py:487 + Step 36880 | grad_norm_pre_clip=0.1475 | + grad_norm_pre_clip_avg=0.1545 | Metrics: + {'align_loss': 0.025184810161590576, + 'recon_loss': 0.10504565387964249, + 'predict_loss': 0.006647966802120209, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14752544462680817, + 'data_time': 0.0008037160150706768, + 'model_time': 1.1997378290107008, + 'grad_norm_pre_clip_avg': 0.1545418880879879, + 'learning_rate': 4.907133923135602e-06, + 'epoch': 9.31} +04/20 [00:47:34] INFO | >> train_qwenlatent.py:487 + Step 36890 | grad_norm_pre_clip=0.1720 | + grad_norm_pre_clip_avg=0.1678 | Metrics: + {'align_loss': 0.024460166692733765, + 'recon_loss': 0.07747969776391983, + 'predict_loss': 0.004096076823771, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17201900482177734, + 'data_time': 0.0009389859915245324, + 'model_time': 1.2425570430059452, + 'grad_norm_pre_clip_avg': 0.16780115365982057, + 'learning_rate': 4.900221239393165e-06, + 'epoch': 9.31} +04/20 [00:47:47] INFO | >> train_qwenlatent.py:487 + Step 36900 | grad_norm_pre_clip=0.1256 | + grad_norm_pre_clip_avg=0.1581 | Metrics: + {'align_loss': 0.025700516998767853, + 'recon_loss': 0.10705196857452393, + 'predict_loss': 0.008714454248547554, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12560206651687622, + 'mae_score': 0.009597403723914344, 'data_time': + 0.0008096180099528283, 'model_time': + 1.24366956402082, 'grad_norm_pre_clip_avg': + 0.1580619618296623, 'learning_rate': + 4.893312265782921e-06, 'epoch': 9.31} +04/20 [00:47:59] INFO | >> train_qwenlatent.py:487 + Step 36910 | grad_norm_pre_clip=0.1626 | + grad_norm_pre_clip_avg=0.1539 | Metrics: + {'align_loss': 0.024774543941020966, + 'recon_loss': 0.10878433287143707, + 'predict_loss': 0.006819427944719791, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16256529092788696, + 'data_time': 0.0007109559956006706, + 'model_time': 1.2219954980246257, + 'grad_norm_pre_clip_avg': 0.15390704125165938, + 'learning_rate': 4.886407005672221e-06, + 'epoch': 9.31} +04/20 [00:48:12] INFO | >> train_qwenlatent.py:487 + Step 36920 | grad_norm_pre_clip=0.1481 | + grad_norm_pre_clip_avg=0.1652 | Metrics: + {'align_loss': 0.02549879252910614, + 'recon_loss': 0.1522509902715683, + 'predict_loss': 0.010263833217322826, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14806964993476868, + 'data_time': 0.0007884560036472976, + 'model_time': 1.226947471004678, + 'grad_norm_pre_clip_avg': 0.1651894524693489, + 'learning_rate': 4.879505462426602e-06, + 'epoch': 9.32} +04/20 [00:48:24] INFO | >> train_qwenlatent.py:487 + Step 36930 | grad_norm_pre_clip=0.2376 | + grad_norm_pre_clip_avg=0.1773 | Metrics: + {'align_loss': 0.025334443897008896, + 'recon_loss': 0.11338124424219131, + 'predict_loss': 0.008306317031383514, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23762919008731842, + 'data_time': 0.0009339890093542635, + 'model_time': 1.3251177339989226, + 'grad_norm_pre_clip_avg': 0.17728237956762313, + 'learning_rate': 4.87260763940979e-06, 'epoch': + 9.32} +04/20 [00:48:37] INFO | >> train_qwenlatent.py:487 + Step 36940 | grad_norm_pre_clip=0.1314 | + grad_norm_pre_clip_avg=0.1602 | Metrics: + {'align_loss': 0.026839645579457283, + 'recon_loss': 0.11638206988573074, + 'predict_loss': 0.008293088525533676, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13139423727989197, + 'data_time': 0.0006743870035279542, + 'model_time': 1.212130536994664, + 'grad_norm_pre_clip_avg': 0.16024992540478705, + 'learning_rate': 4.865713539983714e-06, + 'epoch': 9.32} +04/20 [00:48:50] INFO | >> train_qwenlatent.py:487 + Step 36950 | grad_norm_pre_clip=0.1322 | + grad_norm_pre_clip_avg=0.1377 | Metrics: + {'align_loss': 0.025175616145133972, + 'recon_loss': 0.1066417545080185, + 'predict_loss': 0.005190341267734766, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13221152126789093, + 'mae_score': 0.00581529290826471, 'data_time': + 0.0007642090204171836, 'model_time': + 1.23551922198385, 'grad_norm_pre_clip_avg': + 0.13766609579324723, 'learning_rate': + 4.858823167508459e-06, 'epoch': 9.32} +04/20 [00:49:03] INFO | >> train_qwenlatent.py:487 + Step 36960 | grad_norm_pre_clip=0.1497 | + grad_norm_pre_clip_avg=0.1732 | Metrics: + {'align_loss': 0.025006987154483795, + 'recon_loss': 0.09405805170536041, + 'predict_loss': 0.007051939610391855, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14974096417427063, + 'data_time': 0.0006767089944332838, + 'model_time': 1.2241579580004327, + 'grad_norm_pre_clip_avg': 0.17324876934289932, + 'learning_rate': 4.851936525342315e-06, + 'epoch': 9.33} +04/20 [00:49:15] INFO | >> train_qwenlatent.py:487 + Step 36970 | grad_norm_pre_clip=0.1539 | + grad_norm_pre_clip_avg=0.1612 | Metrics: + {'align_loss': 0.024026721715927124, + 'recon_loss': 0.12115804851055145, + 'predict_loss': 0.008863941766321659, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1539466381072998, + 'data_time': 0.0007330530206672847, + 'model_time': 1.2338300379924476, + 'grad_norm_pre_clip_avg': 0.1611858181655407, + 'learning_rate': 4.845053616841741e-06, + 'epoch': 9.33} +04/20 [00:49:28] INFO | >> train_qwenlatent.py:487 + Step 36980 | grad_norm_pre_clip=0.1339 | + grad_norm_pre_clip_avg=0.1631 | Metrics: + {'align_loss': 0.02585824579000473, + 'recon_loss': 0.1407148241996765, + 'predict_loss': 0.010423247702419758, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13387520611286163, + 'data_time': 0.0006804159784223884, + 'model_time': 1.2479110799904447, + 'grad_norm_pre_clip_avg': 0.1631387062370777, + 'learning_rate': 4.838174445361395e-06, + 'epoch': 9.33} +04/20 [00:49:40] INFO | >> train_qwenlatent.py:487 + Step 36990 | grad_norm_pre_clip=0.1826 | + grad_norm_pre_clip_avg=0.1679 | Metrics: + {'align_loss': 0.02485346421599388, + 'recon_loss': 0.1061210036277771, + 'predict_loss': 0.010382907465100288, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1825913041830063, + 'data_time': 0.0008205540070775896, + 'model_time': 1.1889858580252621, + 'grad_norm_pre_clip_avg': 0.16790731996297836, + 'learning_rate': 4.8312990142540945e-06, + 'epoch': 9.33} +04/20 [00:49:54] INFO | >> train_qwenlatent.py:487 + Step 37000 | grad_norm_pre_clip=0.1455 | + grad_norm_pre_clip_avg=0.1551 | Metrics: + {'align_loss': 0.025378581136465073, + 'recon_loss': 0.11741431057453156, + 'predict_loss': 0.006630830466747284, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14548321068286896, + 'mae_score': 0.007483039890323674, 'data_time': + 0.0007064039818942547, 'model_time': + 1.3152531719824765, 'grad_norm_pre_clip_avg': + 0.15507413297891617, 'learning_rate': + 4.824427326870842e-06, 'epoch': 9.34} +04/20 [00:50:07] INFO | >> train_qwenlatent.py:487 + Step 37010 | grad_norm_pre_clip=0.1699 | + grad_norm_pre_clip_avg=0.1593 | Metrics: + {'align_loss': 0.025812746956944466, + 'recon_loss': 0.14096017181873322, + 'predict_loss': 0.015224993228912354, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16986267268657684, + 'data_time': 0.0007053389854263514, + 'model_time': 1.1978167920024134, + 'grad_norm_pre_clip_avg': 0.15933812409639359, + 'learning_rate': 4.8175593865608155e-06, + 'epoch': 9.34} +04/20 [00:50:19] INFO | >> train_qwenlatent.py:487 + Step 37020 | grad_norm_pre_clip=0.1665 | + grad_norm_pre_clip_avg=0.1646 | Metrics: + {'align_loss': 0.025584420189261436, + 'recon_loss': 0.12645186483860016, + 'predict_loss': 0.006493251770734787, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16647616028785706, + 'data_time': 0.0007932249864097685, + 'model_time': 1.248792553000385, + 'grad_norm_pre_clip_avg': 0.16463928818702697, + 'learning_rate': 4.810695196671365e-06, + 'epoch': 9.34} +04/20 [00:50:32] INFO | >> train_qwenlatent.py:487 + Step 37030 | grad_norm_pre_clip=0.1963 | + grad_norm_pre_clip_avg=0.1728 | Metrics: + {'align_loss': 0.025848262012004852, + 'recon_loss': 0.11083422601222992, + 'predict_loss': 0.007300740573555231, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19628119468688965, + 'data_time': 0.000688158004777506, + 'model_time': 1.2370173950039316, + 'grad_norm_pre_clip_avg': 0.17282679080963134, + 'learning_rate': 4.803834760548012e-06, + 'epoch': 9.34} +04/20 [00:50:44] INFO | >> train_qwenlatent.py:487 + Step 37040 | grad_norm_pre_clip=0.1857 | + grad_norm_pre_clip_avg=0.1659 | Metrics: + {'align_loss': 0.024901418015360832, + 'recon_loss': 0.21641521155834198, + 'predict_loss': 0.01201950665563345, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18568889796733856, + 'data_time': 0.0009496590064372867, + 'model_time': 1.276829665992409, + 'grad_norm_pre_clip_avg': 0.16591444611549377, + 'learning_rate': 4.796978081534452e-06, + 'epoch': 9.35} +04/20 [00:50:58] INFO | >> train_qwenlatent.py:487 + Step 37050 | grad_norm_pre_clip=0.1507 | + grad_norm_pre_clip_avg=0.1620 | Metrics: + {'align_loss': 0.024974068626761436, + 'recon_loss': 0.1006334200501442, + 'predict_loss': 0.01148642972111702, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15070240199565887, + 'mae_score': 0.005636675722964175, 'data_time': + 0.0009754139755386859, 'model_time': + 1.2250018730119336, 'grad_norm_pre_clip_avg': + 0.16199385970830918, 'learning_rate': + 4.790125162972544e-06, 'epoch': 9.35} +04/20 [00:51:10] INFO | >> train_qwenlatent.py:487 + Step 37060 | grad_norm_pre_clip=0.1902 | + grad_norm_pre_clip_avg=0.1640 | Metrics: + {'align_loss': 0.026020247489213943, + 'recon_loss': 0.13782694935798645, + 'predict_loss': 0.0118472995236516, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19022823870182037, + 'data_time': 0.0006992979906499386, + 'model_time': 1.200239662983222, + 'grad_norm_pre_clip_avg': 0.16402996331453323, + 'learning_rate': 4.783276008202321e-06, + 'epoch': 9.35} +04/20 [00:51:23] INFO | >> train_qwenlatent.py:487 + Step 37070 | grad_norm_pre_clip=0.2504 | + grad_norm_pre_clip_avg=0.1794 | Metrics: + {'align_loss': 0.025265563279390335, + 'recon_loss': 0.1064293161034584, + 'predict_loss': 0.007936256006360054, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25040203332901, + 'data_time': 0.0006676490011159331, + 'model_time': 1.233177504997002, + 'grad_norm_pre_clip_avg': 0.17943161129951476, + 'learning_rate': 4.776430620561972e-06, + 'epoch': 9.35} +04/20 [00:51:35] INFO | >> train_qwenlatent.py:487 + Step 37080 | grad_norm_pre_clip=0.1731 | + grad_norm_pre_clip_avg=0.1710 | Metrics: + {'align_loss': 0.02693059667944908, + 'recon_loss': 0.13803061842918396, + 'predict_loss': 0.011841956526041031, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1730603575706482, + 'data_time': 0.0010017250024247915, + 'model_time': 1.2001071110134944, + 'grad_norm_pre_clip_avg': 0.1710241839289665, + 'learning_rate': 4.769589003387863e-06, + 'epoch': 9.36} +04/20 [00:51:48] INFO | >> train_qwenlatent.py:487 + Step 37090 | grad_norm_pre_clip=0.1620 | + grad_norm_pre_clip_avg=0.1950 | Metrics: + {'align_loss': 0.02546919509768486, + 'recon_loss': 0.10592292249202728, + 'predict_loss': 0.0034248679876327515, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16203029453754425, + 'data_time': 0.0007037660107016563, + 'model_time': 1.2401902890123893, + 'grad_norm_pre_clip_avg': 0.1950155735015869, + 'learning_rate': 4.7627511600145125e-06, + 'epoch': 9.36} +04/20 [00:52:01] INFO | >> train_qwenlatent.py:487 + Step 37100 | grad_norm_pre_clip=0.1726 | + grad_norm_pre_clip_avg=0.1532 | Metrics: + {'align_loss': 0.025663238018751144, + 'recon_loss': 0.12935902178287506, + 'predict_loss': 0.0071196118369698524, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17260119318962097, + 'mae_score': 0.007739634986396309, 'data_time': + 0.0007734530081506819, 'model_time': + 1.2525013999838848, 'grad_norm_pre_clip_avg': + 0.1531811162829399, 'learning_rate': + 4.7559170937746e-06, 'epoch': 9.36} +04/20 [00:52:14] INFO | >> train_qwenlatent.py:487 + Step 37110 | grad_norm_pre_clip=0.1549 | + grad_norm_pre_clip_avg=0.1478 | Metrics: + {'align_loss': 0.026121128350496292, + 'recon_loss': 0.12778881192207336, + 'predict_loss': 0.008367164060473442, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1549031138420105, + 'data_time': 0.0008934570068959147, + 'model_time': 1.2199549039942212, + 'grad_norm_pre_clip_avg': 0.14776150211691857, + 'learning_rate': 4.749086807998969e-06, + 'epoch': 9.36} +04/20 [00:52:26] INFO | >> train_qwenlatent.py:487 + Step 37120 | grad_norm_pre_clip=0.1054 | + grad_norm_pre_clip_avg=0.1453 | Metrics: + {'align_loss': 0.02621065452694893, + 'recon_loss': 0.10182991623878479, + 'predict_loss': 0.007847469300031662, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10536550730466843, + 'data_time': 0.0006985199870541692, + 'model_time': 1.2090032139967661, + 'grad_norm_pre_clip_avg': 0.14530728980898858, + 'learning_rate': 4.742260306016616e-06, + 'epoch': 9.37} +04/20 [00:52:38] INFO | >> train_qwenlatent.py:487 + Step 37130 | grad_norm_pre_clip=0.1588 | + grad_norm_pre_clip_avg=0.1499 | Metrics: + {'align_loss': 0.02371104061603546, + 'recon_loss': 0.12669414281845093, + 'predict_loss': 0.010320339351892471, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1587994247674942, + 'data_time': 0.001080933987395838, + 'model_time': 1.2826865130045917, + 'grad_norm_pre_clip_avg': 0.1498652920126915, + 'learning_rate': 4.735437591154698e-06, + 'epoch': 9.37} +04/20 [00:52:52] INFO | >> train_qwenlatent.py:487 + Step 37140 | grad_norm_pre_clip=0.1499 | + grad_norm_pre_clip_avg=0.1764 | Metrics: + {'align_loss': 0.02462787926197052, + 'recon_loss': 0.09840694069862366, + 'predict_loss': 0.005335166119039059, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14994162321090698, + 'data_time': 0.0006218409980647266, + 'model_time': 1.2062410229991656, + 'grad_norm_pre_clip_avg': 0.17640562504529952, + 'learning_rate': 4.728618666738518e-06, + 'epoch': 9.37} +04/20 [00:53:05] INFO | >> train_qwenlatent.py:487 + Step 37150 | grad_norm_pre_clip=0.2042 | + grad_norm_pre_clip_avg=0.1691 | Metrics: + {'align_loss': 0.025368422269821167, + 'recon_loss': 0.17526809871196747, + 'predict_loss': 0.01346514280885458, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2041631042957306, + 'mae_score': 0.005753762442786414, 'data_time': + 0.0007398219895549119, 'model_time': + 1.2877172770095058, 'grad_norm_pre_clip_avg': + 0.16909346655011176, 'learning_rate': + 4.721803536091543e-06, 'epoch': 9.37} +04/20 [00:53:17] INFO | >> train_qwenlatent.py:487 + Step 37160 | grad_norm_pre_clip=0.1857 | + grad_norm_pre_clip_avg=0.1927 | Metrics: + {'align_loss': 0.026086265221238136, + 'recon_loss': 0.14458616077899933, + 'predict_loss': 0.008538329973816872, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1856933832168579, + 'data_time': 0.0009359410032629967, + 'model_time': 1.2846512770047411, + 'grad_norm_pre_clip_avg': 0.19270246624946594, + 'learning_rate': 4.714992202535381e-06, + 'epoch': 9.38} +04/20 [00:53:30] INFO | >> train_qwenlatent.py:487 + Step 37170 | grad_norm_pre_clip=0.2242 | + grad_norm_pre_clip_avg=0.1767 | Metrics: + {'align_loss': 0.026450544595718384, + 'recon_loss': 0.12344091385602951, + 'predict_loss': 0.0039700716733932495, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.224174365401268, + 'data_time': 0.0009133040148299187, + 'model_time': 1.2141443289874587, + 'grad_norm_pre_clip_avg': 0.1766529694199562, + 'learning_rate': 4.7081846693897904e-06, + 'epoch': 9.38} +04/20 [00:53:42] INFO | >> train_qwenlatent.py:487 + Step 37180 | grad_norm_pre_clip=0.1378 | + grad_norm_pre_clip_avg=0.1480 | Metrics: + {'align_loss': 0.023877376690506935, + 'recon_loss': 0.07205624878406525, + 'predict_loss': 0.0042099496349692345, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1377810388803482, + 'data_time': 0.0008921410189941525, + 'model_time': 1.2932745349826291, + 'grad_norm_pre_clip_avg': 0.1479552112519741, + 'learning_rate': 4.701380939972688e-06, + 'epoch': 9.38} +04/20 [00:53:55] INFO | >> train_qwenlatent.py:487 + Step 37190 | grad_norm_pre_clip=0.1588 | + grad_norm_pre_clip_avg=0.1478 | Metrics: + {'align_loss': 0.025321288034319878, + 'recon_loss': 0.14352145791053772, + 'predict_loss': 0.010465838015079498, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15879322588443756, + 'data_time': 0.0007180209795478731, + 'model_time': 1.2088112299970817, + 'grad_norm_pre_clip_avg': 0.14778827130794525, + 'learning_rate': 4.6945810176001245e-06, + 'epoch': 9.38} +04/20 [00:54:08] INFO | >> train_qwenlatent.py:487 + Step 37200 | grad_norm_pre_clip=0.1364 | + grad_norm_pre_clip_avg=0.1606 | Metrics: + {'align_loss': 0.02592884749174118, + 'recon_loss': 0.11720254272222519, + 'predict_loss': 0.008956610225141048, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13635918498039246, + 'mae_score': 0.006168038995416315, 'data_time': + 0.0010980679944623262, 'model_time': + 1.2697099799988791, 'grad_norm_pre_clip_avg': + 0.16058280766010286, 'learning_rate': + 4.6877849055863035e-06, 'epoch': 9.39} +04/20 [00:54:21] INFO | >> train_qwenlatent.py:487 + Step 37210 | grad_norm_pre_clip=0.1569 | + grad_norm_pre_clip_avg=0.1854 | Metrics: + {'align_loss': 0.02470327541232109, + 'recon_loss': 0.10231154412031174, + 'predict_loss': 0.00733568612486124, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15685050189495087, + 'data_time': 0.000657149008475244, + 'model_time': 1.2065289949823637, + 'grad_norm_pre_clip_avg': 0.18539997786283494, + 'learning_rate': 4.680992607243557e-06, + 'epoch': 9.39} +04/20 [00:54:33] INFO | >> train_qwenlatent.py:487 + Step 37220 | grad_norm_pre_clip=0.1564 | + grad_norm_pre_clip_avg=0.1656 | Metrics: + {'align_loss': 0.02503262832760811, + 'recon_loss': 0.12549301981925964, + 'predict_loss': 0.008624084293842316, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1563931703567505, + 'data_time': 0.0009401279967278242, + 'model_time': 1.2430780940048862, + 'grad_norm_pre_clip_avg': 0.16555429995059967, + 'learning_rate': 4.674204125882379e-06, + 'epoch': 9.39} +04/20 [00:54:46] INFO | >> train_qwenlatent.py:487 + Step 37230 | grad_norm_pre_clip=0.1254 | + grad_norm_pre_clip_avg=0.1520 | Metrics: + {'align_loss': 0.024934936314821243, + 'recon_loss': 0.11256231367588043, + 'predict_loss': 0.0056162928231060505, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1253983974456787, + 'data_time': 0.001034462999086827, + 'model_time': 1.2500540429900866, + 'grad_norm_pre_clip_avg': 0.15199915170669556, + 'learning_rate': 4.66741946481139e-06, 'epoch': + 9.39} +04/20 [00:54:58] INFO | >> train_qwenlatent.py:487 + Step 37240 | grad_norm_pre_clip=0.1559 | + grad_norm_pre_clip_avg=0.1669 | Metrics: + {'align_loss': 0.02485809288918972, + 'recon_loss': 0.16258910298347473, + 'predict_loss': 0.008175750263035297, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15592116117477417, + 'data_time': 0.000880067003890872, + 'model_time': 1.2352793290046975, + 'grad_norm_pre_clip_avg': 0.16690201312303543, + 'learning_rate': 4.66063862733735e-06, 'epoch': + 9.4} +04/20 [00:55:12] INFO | >> train_qwenlatent.py:487 + Step 37250 | grad_norm_pre_clip=0.1543 | + grad_norm_pre_clip_avg=0.1551 | Metrics: + {'align_loss': 0.026535138487815857, + 'recon_loss': 0.13353422284126282, + 'predict_loss': 0.005433670245110989, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15426073968410492, + 'mae_score': 0.006730029819247959, 'data_time': + 0.0012609950208570808, 'model_time': + 1.2860508500016294, 'grad_norm_pre_clip_avg': + 0.15513948276638984, 'learning_rate': + 4.653861616765157e-06, 'epoch': 9.4} +04/20 [00:55:24] INFO | >> train_qwenlatent.py:487 + Step 37260 | grad_norm_pre_clip=0.1566 | + grad_norm_pre_clip_avg=0.1643 | Metrics: + {'align_loss': 0.025525877252221107, + 'recon_loss': 0.1378875970840454, + 'predict_loss': 0.010728346183896065, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1565505713224411, + 'data_time': 0.0007666349993087351, + 'model_time': 1.3093860730004963, + 'grad_norm_pre_clip_avg': 0.1643013149499893, + 'learning_rate': 4.647088436397841e-06, + 'epoch': 9.4} +04/20 [00:55:37] INFO | >> train_qwenlatent.py:487 + Step 37270 | grad_norm_pre_clip=0.1426 | + grad_norm_pre_clip_avg=0.1652 | Metrics: + {'align_loss': 0.025559382513165474, + 'recon_loss': 0.1126093864440918, + 'predict_loss': 0.006209134124219418, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1425866037607193, + 'data_time': 0.0007529970025643706, + 'model_time': 1.2368863520096056, + 'grad_norm_pre_clip_avg': 0.16520075500011444, + 'learning_rate': 4.640319089536575e-06, + 'epoch': 9.4} +04/20 [00:55:50] INFO | >> train_qwenlatent.py:487 + Step 37280 | grad_norm_pre_clip=0.1597 | + grad_norm_pre_clip_avg=0.1628 | Metrics: + {'align_loss': 0.02520117536187172, + 'recon_loss': 0.13693559169769287, + 'predict_loss': 0.007623820565640926, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1596846729516983, + 'data_time': 0.0006619860068894923, + 'model_time': 1.2780309060181025, + 'grad_norm_pre_clip_avg': 0.1627609387040138, + 'learning_rate': 4.6335535794806525e-06, + 'epoch': 9.41} +04/20 [00:56:03] INFO | >> train_qwenlatent.py:487 + Step 37290 | grad_norm_pre_clip=0.1549 | + grad_norm_pre_clip_avg=0.1481 | Metrics: + {'align_loss': 0.02627425082027912, + 'recon_loss': 0.10815513134002686, + 'predict_loss': 0.00786573626101017, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1548546701669693, + 'data_time': 0.000958907010499388, + 'model_time': 1.2332749439810868, + 'grad_norm_pre_clip_avg': 0.14805111289024353, + 'learning_rate': 4.626791909527505e-06, + 'epoch': 9.41} +04/20 [00:56:16] INFO | >> train_qwenlatent.py:487 + Step 37300 | grad_norm_pre_clip=0.1434 | + grad_norm_pre_clip_avg=0.1679 | Metrics: + {'align_loss': 0.027157366275787354, + 'recon_loss': 0.13288433849811554, + 'predict_loss': 0.006484649609774351, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14335422217845917, + 'mae_score': 0.008208034274814365, 'data_time': + 0.00073291698936373, 'model_time': + 1.2298800309945364, 'grad_norm_pre_clip_avg': + 0.16786653995513917, 'learning_rate': + 4.620034082972675e-06, 'epoch': 9.41} +04/20 [00:56:28] INFO | >> train_qwenlatent.py:487 + Step 37310 | grad_norm_pre_clip=0.1673 | + grad_norm_pre_clip_avg=0.1603 | Metrics: + {'align_loss': 0.025207318365573883, + 'recon_loss': 0.10782837867736816, + 'predict_loss': 0.006660123355686665, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16729003190994263, + 'data_time': 0.0010576930071692914, + 'model_time': 1.2360722050070763, + 'grad_norm_pre_clip_avg': 0.16034511253237724, + 'learning_rate': 4.613280103109859e-06, + 'epoch': 9.41} +04/20 [00:56:41] INFO | >> train_qwenlatent.py:487 + Step 37320 | grad_norm_pre_clip=0.1291 | + grad_norm_pre_clip_avg=0.1604 | Metrics: + {'align_loss': 0.02433016151189804, + 'recon_loss': 0.11188679933547974, + 'predict_loss': 0.00656858691945672, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12908785045146942, + 'data_time': 0.0006692220049444586, + 'model_time': 1.281154702999629, + 'grad_norm_pre_clip_avg': 0.16040620505809783, + 'learning_rate': 4.606529973230861e-06, + 'epoch': 9.42} +04/20 [00:56:53] INFO | >> train_qwenlatent.py:487 + Step 37330 | grad_norm_pre_clip=0.1981 | + grad_norm_pre_clip_avg=0.1741 | Metrics: + {'align_loss': 0.024867307394742966, + 'recon_loss': 0.08632685989141464, + 'predict_loss': 0.007327875588089228, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19805245101451874, + 'data_time': 0.0006908639916218817, + 'model_time': 1.243840020004427, + 'grad_norm_pre_clip_avg': 0.1740958109498024, + 'learning_rate': 4.599783696625609e-06, + 'epoch': 9.42} +04/20 [00:57:06] INFO | >> train_qwenlatent.py:487 + Step 37340 | grad_norm_pre_clip=0.1694 | + grad_norm_pre_clip_avg=0.2034 | Metrics: + {'align_loss': 0.026103846728801727, + 'recon_loss': 0.1091291680932045, + 'predict_loss': 0.005026010796427727, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1693655252456665, + 'data_time': 0.0008328539843205363, + 'model_time': 1.2135646700044163, + 'grad_norm_pre_clip_avg': 0.20344220250844955, + 'learning_rate': 4.59304127658216e-06, 'epoch': + 9.42} +04/20 [00:57:19] INFO | >> train_qwenlatent.py:487 + Step 37350 | grad_norm_pre_clip=0.1447 | + grad_norm_pre_clip_avg=0.1791 | Metrics: + {'align_loss': 0.02449025586247444, + 'recon_loss': 0.10025642067193985, + 'predict_loss': 0.005131593439728022, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1446557492017746, + 'mae_score': 0.007160137365530203, 'data_time': + 0.0009065769845619798, 'model_time': + 1.2273440579883754, 'grad_norm_pre_clip_avg': + 0.1791287273168564, 'learning_rate': + 4.5863027163866855e-06, 'epoch': 9.42} +04/20 [00:57:31] INFO | >> train_qwenlatent.py:487 + Step 37360 | grad_norm_pre_clip=0.1310 | + grad_norm_pre_clip_avg=0.1839 | Metrics: + {'align_loss': 0.025307510048151016, + 'recon_loss': 0.1163102462887764, + 'predict_loss': 0.006433030590415001, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13103988766670227, + 'data_time': 0.0006794689979869872, + 'model_time': 1.2084100710053463, + 'grad_norm_pre_clip_avg': 0.1838675245642662, + 'learning_rate': 4.5795680193234746e-06, + 'epoch': 9.43} +04/20 [00:57:44] INFO | >> train_qwenlatent.py:487 + Step 37370 | grad_norm_pre_clip=0.1780 | + grad_norm_pre_clip_avg=0.1691 | Metrics: + {'align_loss': 0.024657441303133965, + 'recon_loss': 0.14356981217861176, + 'predict_loss': 0.007179941516369581, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17801719903945923, + 'data_time': 0.0010300719877704978, + 'model_time': 1.2468927050067578, + 'grad_norm_pre_clip_avg': 0.16910448521375657, + 'learning_rate': 4.572837188674945e-06, + 'epoch': 9.43} +04/20 [00:57:56] INFO | >> train_qwenlatent.py:487 + Step 37380 | grad_norm_pre_clip=0.1424 | + grad_norm_pre_clip_avg=0.1529 | Metrics: + {'align_loss': 0.026191070675849915, + 'recon_loss': 0.08629872649908066, + 'predict_loss': 0.004422199912369251, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14236602187156677, + 'data_time': 0.000677917996654287, + 'model_time': 1.2211614469997585, + 'grad_norm_pre_clip_avg': 0.15287813767790795, + 'learning_rate': 4.566110227721621e-06, + 'epoch': 9.43} +04/20 [00:58:08] INFO | >> train_qwenlatent.py:487 + Step 37390 | grad_norm_pre_clip=0.1578 | + grad_norm_pre_clip_avg=0.1497 | Metrics: + {'align_loss': 0.025803249329328537, + 'recon_loss': 0.11173241585493088, + 'predict_loss': 0.004938546568155289, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15782877802848816, + 'data_time': 0.0007821059843990952, + 'model_time': 1.2002759239985608, + 'grad_norm_pre_clip_avg': 0.14974044710397721, + 'learning_rate': 4.559387139742136e-06, + 'epoch': 9.43} +04/20 [00:58:21] INFO | >> train_qwenlatent.py:487 + Step 37400 | grad_norm_pre_clip=0.1622 | + grad_norm_pre_clip_avg=0.1494 | Metrics: + {'align_loss': 0.025073081254959106, + 'recon_loss': 0.14283433556556702, + 'predict_loss': 0.007286165375262499, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16220714151859283, + 'mae_score': 0.00600854813515603, 'data_time': + 0.0007086260011419654, 'model_time': + 1.1912943290080875, 'grad_norm_pre_clip_avg': + 0.14936083927750587, 'learning_rate': + 4.552667928013237e-06, 'epoch': 9.44} +04/20 [00:58:34] INFO | >> train_qwenlatent.py:487 + Step 37410 | grad_norm_pre_clip=0.1575 | + grad_norm_pre_clip_avg=0.1664 | Metrics: + {'align_loss': 0.024466346949338913, + 'recon_loss': 0.10603839159011841, + 'predict_loss': 0.006934229284524918, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15748660266399384, + 'data_time': 0.0007284599996637553, + 'model_time': 1.282358120981371, + 'grad_norm_pre_clip_avg': 0.16637906432151794, + 'learning_rate': 4.5459525958097985e-06, + 'epoch': 9.44} +04/20 [00:58:47] INFO | >> train_qwenlatent.py:487 + Step 37420 | grad_norm_pre_clip=0.1831 | + grad_norm_pre_clip_avg=0.1532 | Metrics: + {'align_loss': 0.02478831820189953, + 'recon_loss': 0.09570234268903732, + 'predict_loss': 0.00785097200423479, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18305641412734985, + 'data_time': 0.0007122639799490571, + 'model_time': 1.2891723700158764, + 'grad_norm_pre_clip_avg': 0.15315279737114906, + 'learning_rate': 4.539241146404786e-06, + 'epoch': 9.44} +04/20 [00:59:01] INFO | >> train_qwenlatent.py:487 + Step 37430 | grad_norm_pre_clip=0.1529 | + grad_norm_pre_clip_avg=0.1639 | Metrics: + {'align_loss': 0.025801582261919975, + 'recon_loss': 0.15719074010849, 'predict_loss': + 0.010574836283922195, 'aux_loss_decay_weight': + 0.0, 'grad_norm_pre_clip': 0.15286435186862946, + 'data_time': 0.0009457100240979344, + 'model_time': 1.224183978018118, + 'grad_norm_pre_clip_avg': 0.16393590420484544, + 'learning_rate': 4.532533583069277e-06, + 'epoch': 9.44} +04/20 [00:59:13] INFO | >> train_qwenlatent.py:487 + Step 37440 | grad_norm_pre_clip=0.1786 | + grad_norm_pre_clip_avg=0.1808 | Metrics: + {'align_loss': 0.025633540004491806, + 'recon_loss': 0.12826590240001678, + 'predict_loss': 0.007938331924378872, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1785506159067154, + 'data_time': 0.0009163559880107641, + 'model_time': 1.2404482020065188, + 'grad_norm_pre_clip_avg': 0.18078194558620453, + 'learning_rate': 4.525829909072459e-06, + 'epoch': 9.45} +04/20 [00:59:26] INFO | >> train_qwenlatent.py:487 + Step 37450 | grad_norm_pre_clip=0.1661 | + grad_norm_pre_clip_avg=0.1613 | Metrics: + {'align_loss': 0.025852680206298828, + 'recon_loss': 0.1313110589981079, + 'predict_loss': 0.00752000929787755, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16614675521850586, + 'mae_score': 0.00768749305793831, 'data_time': + 0.0007960509974509478, 'model_time': + 1.2360271209909115, 'grad_norm_pre_clip_avg': + 0.16133456230163573, 'learning_rate': + 4.519130127681617e-06, 'epoch': 9.45} +04/20 [00:59:38] INFO | >> train_qwenlatent.py:487 + Step 37460 | grad_norm_pre_clip=0.1520 | + grad_norm_pre_clip_avg=0.1688 | Metrics: + {'align_loss': 0.026125892996788025, + 'recon_loss': 0.08029693365097046, + 'predict_loss': 0.007602234836667776, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1520383507013321, + 'data_time': 0.0006191660067997873, + 'model_time': 1.259301496000262, + 'grad_norm_pre_clip_avg': 0.16884766519069672, + 'learning_rate': 4.512434242162149e-06, + 'epoch': 9.45} +04/20 [00:59:51] INFO | >> train_qwenlatent.py:487 + Step 37470 | grad_norm_pre_clip=0.1393 | + grad_norm_pre_clip_avg=0.1578 | Metrics: + {'align_loss': 0.025768505409359932, + 'recon_loss': 0.10952973365783691, + 'predict_loss': 0.006013059988617897, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13933971524238586, + 'data_time': 0.0010123520041815937, + 'model_time': 1.2466606439847965, + 'grad_norm_pre_clip_avg': 0.15782874152064325, + 'learning_rate': 4.50574225577755e-06, 'epoch': + 9.45} +04/20 [01:00:04] INFO | >> train_qwenlatent.py:487 + Step 37480 | grad_norm_pre_clip=0.1446 | + grad_norm_pre_clip_avg=0.1576 | Metrics: + {'align_loss': 0.024184152483940125, + 'recon_loss': 0.09614728391170502, + 'predict_loss': 0.00663077412173152, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14459942281246185, + 'data_time': 0.0016292860091198236, + 'model_time': 1.275425120984437, + 'grad_norm_pre_clip_avg': 0.15764687061309815, + 'learning_rate': 4.499054171789405e-06, + 'epoch': 9.46} +04/20 [01:00:16] INFO | >> train_qwenlatent.py:487 + Step 37490 | grad_norm_pre_clip=0.1911 | + grad_norm_pre_clip_avg=0.1668 | Metrics: + {'align_loss': 0.02516666054725647, + 'recon_loss': 0.13750816881656647, + 'predict_loss': 0.004386584274470806, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1910810023546219, + 'data_time': 0.00076864700531587, 'model_time': + 1.2742093060223851, 'grad_norm_pre_clip_avg': + 0.1667829155921936, 'learning_rate': + 4.492369993457402e-06, 'epoch': 9.46} +04/20 [01:00:29] INFO | >> train_qwenlatent.py:487 + Step 37500 | grad_norm_pre_clip=0.2255 | + grad_norm_pre_clip_avg=0.1678 | Metrics: + {'align_loss': 0.024924714118242264, + 'recon_loss': 0.14738351106643677, + 'predict_loss': 0.012452974915504456, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22550268471240997, + 'mae_score': 0.008960920625978763, 'data_time': + 0.000706917024217546, 'model_time': + 1.2496140740113333, 'grad_norm_pre_clip_avg': + 0.16781378388404847, 'learning_rate': + 4.48568972403934e-06, 'epoch': 9.46} +04/20 [01:00:42] INFO | >> train_qwenlatent.py:487 + Step 37510 | grad_norm_pre_clip=0.1691 | + grad_norm_pre_clip_avg=0.1473 | Metrics: + {'align_loss': 0.024517010897397995, + 'recon_loss': 0.1415269821882248, + 'predict_loss': 0.015253485180437565, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16914184391498566, + 'data_time': 0.0009821940038818866, + 'model_time': 1.2457901549933013, + 'grad_norm_pre_clip_avg': 0.14729036763310432, + 'learning_rate': 4.479013366791096e-06, + 'epoch': 9.47} +04/20 [01:00:54] INFO | >> train_qwenlatent.py:487 + Step 37520 | grad_norm_pre_clip=0.1397 | + grad_norm_pre_clip_avg=0.1630 | Metrics: + {'align_loss': 0.026137562468647957, + 'recon_loss': 0.08401518315076828, + 'predict_loss': 0.003889652667567134, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13973088562488556, + 'data_time': 0.0007561800011899322, + 'model_time': 1.250408204010455, + 'grad_norm_pre_clip_avg': 0.16295645385980606, + 'learning_rate': 4.472340924966645e-06, + 'epoch': 9.47} +04/20 [01:01:07] INFO | >> train_qwenlatent.py:487 + Step 37530 | grad_norm_pre_clip=0.1421 | + grad_norm_pre_clip_avg=0.1355 | Metrics: + {'align_loss': 0.02535269781947136, + 'recon_loss': 0.12364222854375839, + 'predict_loss': 0.006370130460709333, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14205417037010193, + 'data_time': 0.0009008650085888803, + 'model_time': 1.201341946987668, + 'grad_norm_pre_clip_avg': 0.13550896048545838, + 'learning_rate': 4.465672401818054e-06, + 'epoch': 9.47} +04/20 [01:01:20] INFO | >> train_qwenlatent.py:487 + Step 37540 | grad_norm_pre_clip=0.1390 | + grad_norm_pre_clip_avg=0.1442 | Metrics: + {'align_loss': 0.02509326860308647, + 'recon_loss': 0.12564925849437714, + 'predict_loss': 0.006922959815710783, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1389937847852707, + 'data_time': 0.0007235670054797083, + 'model_time': 1.2668442179856356, + 'grad_norm_pre_clip_avg': 0.144171354919672, + 'learning_rate': 4.459007800595479e-06, + 'epoch': 9.47} +04/20 [01:01:33] INFO | >> train_qwenlatent.py:487 + Step 37550 | grad_norm_pre_clip=0.1456 | + grad_norm_pre_clip_avg=0.1513 | Metrics: + {'align_loss': 0.02578645572066307, + 'recon_loss': 0.11582119762897491, + 'predict_loss': 0.00738385459408164, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1456274390220642, + 'mae_score': 0.007634994575569222, 'data_time': + 0.000753250002162531, 'model_time': + 1.2150515939865727, 'grad_norm_pre_clip_avg': + 0.1513495720922947, 'learning_rate': + 4.4523471245471655e-06, 'epoch': 9.48} +04/20 [01:01:46] INFO | >> train_qwenlatent.py:487 + Step 37560 | grad_norm_pre_clip=0.2093 | + grad_norm_pre_clip_avg=0.1826 | Metrics: + {'align_loss': 0.02571241371333599, + 'recon_loss': 0.1758275032043457, + 'predict_loss': 0.010801929049193859, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20933721959590912, + 'data_time': 0.0006528950179927051, + 'model_time': 1.2211361109802965, + 'grad_norm_pre_clip_avg': 0.18258598446846008, + 'learning_rate': 4.445690376919455e-06, + 'epoch': 9.48} +04/20 [01:01:59] INFO | >> train_qwenlatent.py:487 + Step 37570 | grad_norm_pre_clip=0.1134 | + grad_norm_pre_clip_avg=0.1630 | Metrics: + {'align_loss': 0.025460708886384964, + 'recon_loss': 0.10889101028442383, + 'predict_loss': 0.007926525548100471, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11337234079837799, + 'data_time': 0.0010308269993402064, + 'model_time': 1.5468087610206567, + 'grad_norm_pre_clip_avg': 0.16296200454235077, + 'learning_rate': 4.439037560956754e-06, + 'epoch': 9.48} +04/20 [01:02:11] INFO | >> train_qwenlatent.py:487 + Step 37580 | grad_norm_pre_clip=0.2165 | + grad_norm_pre_clip_avg=0.1868 | Metrics: + {'align_loss': 0.02561788633465767, + 'recon_loss': 0.1635388880968094, + 'predict_loss': 0.008745762519538403, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21654263138771057, + 'data_time': 0.0009453489910811186, + 'model_time': 1.2761861860053614, + 'grad_norm_pre_clip_avg': 0.18684710562229156, + 'learning_rate': 4.4323886799015686e-06, + 'epoch': 9.48} +04/20 [01:02:24] INFO | >> train_qwenlatent.py:487 + Step 37590 | grad_norm_pre_clip=0.1641 | + grad_norm_pre_clip_avg=0.1627 | Metrics: + {'align_loss': 0.025004053488373756, + 'recon_loss': 0.08614788949489594, + 'predict_loss': 0.005905388854444027, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1640537828207016, + 'data_time': 0.0009181270143017173, + 'model_time': 1.2290441200020723, + 'grad_norm_pre_clip_avg': 0.16269605457782746, + 'learning_rate': 4.425743736994477e-06, + 'epoch': 9.49} +04/20 [01:02:37] INFO | >> train_qwenlatent.py:487 + Step 37600 | grad_norm_pre_clip=0.1398 | + grad_norm_pre_clip_avg=0.1717 | Metrics: + {'align_loss': 0.026205619797110558, + 'recon_loss': 0.115346759557724, + 'predict_loss': 0.007650989107787609, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13978362083435059, + 'mae_score': 0.006187745687123891, 'data_time': + 0.0007622079865541309, 'model_time': + 1.1895454619952943, 'grad_norm_pre_clip_avg': + 0.17167660668492318, 'learning_rate': + 4.419102735474153e-06, 'epoch': 9.49} +04/20 [01:02:50] INFO | >> train_qwenlatent.py:487 + Step 37610 | grad_norm_pre_clip=0.1591 | + grad_norm_pre_clip_avg=0.1966 | Metrics: + {'align_loss': 0.02584829181432724, + 'recon_loss': 0.14835992455482483, + 'predict_loss': 0.009341969154775143, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15914247930049896, + 'data_time': 0.000972844019997865, + 'model_time': 1.2011226259928662, + 'grad_norm_pre_clip_avg': 0.19659244120121003, + 'learning_rate': 4.412465678577337e-06, + 'epoch': 9.49} +04/20 [01:03:03] INFO | >> train_qwenlatent.py:487 + Step 37620 | grad_norm_pre_clip=0.1366 | + grad_norm_pre_clip_avg=0.1534 | Metrics: + {'align_loss': 0.025462515652179718, + 'recon_loss': 0.11554683744907379, + 'predict_loss': 0.009016187861561775, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13659754395484924, + 'data_time': 0.0007195920043159276, + 'model_time': 1.4692036120104603, + 'grad_norm_pre_clip_avg': 0.15338737666606903, + 'learning_rate': 4.405832569538849e-06, + 'epoch': 9.49} +04/20 [01:03:15] INFO | >> train_qwenlatent.py:487 + Step 37630 | grad_norm_pre_clip=0.1632 | + grad_norm_pre_clip_avg=0.1514 | Metrics: + {'align_loss': 0.0256627406924963, + 'recon_loss': 0.14678439497947693, + 'predict_loss': 0.006762966047972441, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1631939858198166, + 'data_time': 0.0008108780020847917, + 'model_time': 1.2520109129836783, + 'grad_norm_pre_clip_avg': 0.1514427475631237, + 'learning_rate': 4.399203411591584e-06, + 'epoch': 9.5} +04/20 [01:03:27] INFO | >> train_qwenlatent.py:487 + Step 37640 | grad_norm_pre_clip=0.1708 | + grad_norm_pre_clip_avg=0.1467 | Metrics: + {'align_loss': 0.025836359709501266, + 'recon_loss': 0.15612320601940155, + 'predict_loss': 0.008872101083397865, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1707952320575714, + 'data_time': 0.0006962810002733022, + 'model_time': 1.2725405220116954, + 'grad_norm_pre_clip_avg': 0.14665160179138184, + 'learning_rate': 4.392578207966512e-06, + 'epoch': 9.5} +04/20 [01:03:41] INFO | >> train_qwenlatent.py:487 + Step 37650 | grad_norm_pre_clip=0.1873 | + grad_norm_pre_clip_avg=0.1861 | Metrics: + {'align_loss': 0.02376525104045868, + 'recon_loss': 0.14539240300655365, + 'predict_loss': 0.011798140592873096, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1873340755701065, + 'mae_score': 0.006294767706243842, 'data_time': + 0.000987559003988281, 'model_time': + 1.2991394140117336, 'grad_norm_pre_clip_avg': + 0.18606702983379364, 'learning_rate': + 4.385956961892688e-06, 'epoch': 9.5} +04/20 [01:03:53] INFO | >> train_qwenlatent.py:487 + Step 37660 | grad_norm_pre_clip=0.2041 | + grad_norm_pre_clip_avg=0.1836 | Metrics: + {'align_loss': 0.02708137035369873, + 'recon_loss': 0.13209907710552216, + 'predict_loss': 0.007959301583468914, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20407623052597046, + 'data_time': 0.000854918995173648, + 'model_time': 1.2056208970025182, + 'grad_norm_pre_clip_avg': 0.18358180224895476, + 'learning_rate': 4.379339676597215e-06, + 'epoch': 9.5} +04/20 [01:04:06] INFO | >> train_qwenlatent.py:487 + Step 37670 | grad_norm_pre_clip=0.1452 | + grad_norm_pre_clip_avg=0.1831 | Metrics: + {'align_loss': 0.02578994259238243, + 'recon_loss': 0.13476546108722687, + 'predict_loss': 0.00595887703821063, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14515124261379242, + 'data_time': 0.0009190280106849968, + 'model_time': 1.2108988090185449, + 'grad_norm_pre_clip_avg': 0.1831497773528099, + 'learning_rate': 4.372726355305279e-06, + 'epoch': 9.51} +04/20 [01:04:18] INFO | >> train_qwenlatent.py:487 + Step 37680 | grad_norm_pre_clip=0.1696 | + grad_norm_pre_clip_avg=0.1697 | Metrics: + {'align_loss': 0.025641541928052902, + 'recon_loss': 0.12785965204238892, + 'predict_loss': 0.006413883529603481, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16961826384067535, + 'data_time': 0.0006800160044804215, + 'model_time': 1.215689808013849, + 'grad_norm_pre_clip_avg': 0.16973565518856049, + 'learning_rate': 4.366117001240135e-06, + 'epoch': 9.51} +04/20 [01:04:31] INFO | >> train_qwenlatent.py:487 + Step 37690 | grad_norm_pre_clip=0.1258 | + grad_norm_pre_clip_avg=0.1462 | Metrics: + {'align_loss': 0.024532947689294815, + 'recon_loss': 0.16418541967868805, + 'predict_loss': 0.008121460676193237, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1258455067873001, + 'data_time': 0.0008723149949219078, + 'model_time': 1.2519464679935481, + 'grad_norm_pre_clip_avg': 0.14623693227767945, + 'learning_rate': 4.359511617623098e-06, + 'epoch': 9.51} +04/20 [01:04:44] INFO | >> train_qwenlatent.py:487 + Step 37700 | grad_norm_pre_clip=0.1756 | + grad_norm_pre_clip_avg=0.1511 | Metrics: + {'align_loss': 0.025992728769779205, + 'recon_loss': 0.11054125428199768, + 'predict_loss': 0.004052390810102224, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1756461262702942, + 'mae_score': 0.006602001619768572, 'data_time': + 0.0008338380139321089, 'model_time': + 1.5165565990027972, 'grad_norm_pre_clip_avg': + 0.15114671885967254, 'learning_rate': + 4.352910207673558e-06, 'epoch': 9.51} +04/20 [01:04:57] INFO | >> train_qwenlatent.py:487 + Step 37710 | grad_norm_pre_clip=0.1607 | + grad_norm_pre_clip_avg=0.1610 | Metrics: + {'align_loss': 0.025243669748306274, + 'recon_loss': 0.0995054617524147, + 'predict_loss': 0.010709383524954319, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16069923341274261, + 'data_time': 0.001453325996408239, + 'model_time': 1.243317760003265, + 'grad_norm_pre_clip_avg': 0.1609997794032097, + 'learning_rate': 4.34631277460896e-06, 'epoch': + 9.52} +04/20 [01:05:10] INFO | >> train_qwenlatent.py:487 + Step 37720 | grad_norm_pre_clip=0.1102 | + grad_norm_pre_clip_avg=0.1524 | Metrics: + {'align_loss': 0.02478155680000782, + 'recon_loss': 0.11704327166080475, + 'predict_loss': 0.007694327272474766, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11019381880760193, + 'data_time': 0.0008053600031416863, + 'model_time': 1.2353691619937308, + 'grad_norm_pre_clip_avg': 0.15241266489028932, + 'learning_rate': 4.339719321644811e-06, + 'epoch': 9.52} +04/20 [01:05:23] INFO | >> train_qwenlatent.py:487 + Step 37730 | grad_norm_pre_clip=0.1476 | + grad_norm_pre_clip_avg=0.1577 | Metrics: + {'align_loss': 0.024627558887004852, + 'recon_loss': 0.09007228910923004, + 'predict_loss': 0.006823631003499031, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1476200520992279, + 'data_time': 0.0007482530199922621, + 'model_time': 1.1776860719837714, + 'grad_norm_pre_clip_avg': 0.1576880045235157, + 'learning_rate': 4.333129851994681e-06, + 'epoch': 9.52} +04/20 [01:05:35] INFO | >> train_qwenlatent.py:487 + Step 37740 | grad_norm_pre_clip=0.1514 | + grad_norm_pre_clip_avg=0.1763 | Metrics: + {'align_loss': 0.025144243612885475, + 'recon_loss': 0.1117912009358406, + 'predict_loss': 0.005870266817510128, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15137766301631927, + 'data_time': 0.0007113720057532191, + 'model_time': 1.3177846189937554, + 'grad_norm_pre_clip_avg': 0.17628005146980286, + 'learning_rate': 4.326544368870197e-06, + 'epoch': 9.52} +04/20 [01:05:48] INFO | >> train_qwenlatent.py:487 + Step 37750 | grad_norm_pre_clip=0.1587 | + grad_norm_pre_clip_avg=0.1522 | Metrics: + {'align_loss': 0.02539437636733055, + 'recon_loss': 0.12198459357023239, + 'predict_loss': 0.008800908923149109, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15868555009365082, + 'mae_score': 0.007816935015154314, 'data_time': + 0.000907080975593999, 'model_time': + 1.3357801859965548, 'grad_norm_pre_clip_avg': + 0.15217780470848083, 'learning_rate': + 4.319962875481042e-06, 'epoch': 9.53} +04/20 [01:06:01] INFO | >> train_qwenlatent.py:487 + Step 37760 | grad_norm_pre_clip=0.1780 | + grad_norm_pre_clip_avg=0.1589 | Metrics: + {'align_loss': 0.02518082782626152, + 'recon_loss': 0.12187277525663376, + 'predict_loss': 0.006368819158524275, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1780024617910385, + 'data_time': 0.0010019379842560738, + 'model_time': 1.3258978170051705, + 'grad_norm_pre_clip_avg': 0.15894515737891196, + 'learning_rate': 4.313385375034958e-06, + 'epoch': 9.53} +04/20 [01:06:14] INFO | >> train_qwenlatent.py:487 + Step 37770 | grad_norm_pre_clip=0.1781 | + grad_norm_pre_clip_avg=0.1756 | Metrics: + {'align_loss': 0.02680560015141964, + 'recon_loss': 0.1323963850736618, + 'predict_loss': 0.00913016963750124, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17807424068450928, + 'data_time': 0.0007168739975895733, + 'model_time': 1.2444126220070757, + 'grad_norm_pre_clip_avg': 0.17556851506233215, + 'learning_rate': 4.306811870737739e-06, + 'epoch': 9.53} +04/20 [01:06:26] INFO | >> train_qwenlatent.py:487 + Step 37780 | grad_norm_pre_clip=0.1383 | + grad_norm_pre_clip_avg=0.1544 | Metrics: + {'align_loss': 0.024484258145093918, + 'recon_loss': 0.12720511853694916, + 'predict_loss': 0.008279050700366497, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13833941519260406, + 'data_time': 0.0006335109937936068, + 'model_time': 1.2417662840161938, + 'grad_norm_pre_clip_avg': 0.15441079512238504, + 'learning_rate': 4.300242365793227e-06, + 'epoch': 9.53} +04/20 [01:06:39] INFO | >> train_qwenlatent.py:487 + Step 37790 | grad_norm_pre_clip=0.2005 | + grad_norm_pre_clip_avg=0.1554 | Metrics: + {'align_loss': 0.02579881250858307, + 'recon_loss': 0.10637588053941727, + 'predict_loss': 0.010102145373821259, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2005055546760559, + 'data_time': 0.0009383080177940428, + 'model_time': 1.2294412799819838, + 'grad_norm_pre_clip_avg': 0.1554160386323929, + 'learning_rate': 4.293676863403326e-06, + 'epoch': 9.54} +04/20 [01:06:52] INFO | >> train_qwenlatent.py:487 + Step 37800 | grad_norm_pre_clip=0.1566 | + grad_norm_pre_clip_avg=0.1531 | Metrics: + {'align_loss': 0.025636641308665276, + 'recon_loss': 0.150661438703537, + 'predict_loss': 0.010210225358605385, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15661010146141052, + 'mae_score': 0.0063133823979008305, + 'data_time': 0.0009884149767458439, + 'model_time': 1.2332527760008816, + 'grad_norm_pre_clip_avg': 0.15311237424612045, + 'learning_rate': 4.287115366767976e-06, + 'epoch': 9.54} +04/20 [01:07:04] INFO | >> train_qwenlatent.py:487 + Step 37810 | grad_norm_pre_clip=0.1626 | + grad_norm_pre_clip_avg=0.1600 | Metrics: + {'align_loss': 0.025401918217539787, + 'recon_loss': 0.11916988343000412, + 'predict_loss': 0.006291547324508429, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1626354157924652, + 'data_time': 0.0006409339839592576, + 'model_time': 1.2752331200172193, + 'grad_norm_pre_clip_avg': 0.1600125215947628, + 'learning_rate': 4.2805578790851754e-06, + 'epoch': 9.54} +04/20 [01:07:17] INFO | >> train_qwenlatent.py:487 + Step 37820 | grad_norm_pre_clip=0.1478 | + grad_norm_pre_clip_avg=0.1584 | Metrics: + {'align_loss': 0.02583475410938263, + 'recon_loss': 0.1043480783700943, + 'predict_loss': 0.006029031239449978, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1477954387664795, + 'data_time': 0.0009985939832404256, + 'model_time': 1.2835045619867742, + 'grad_norm_pre_clip_avg': 0.15838936120271682, + 'learning_rate': 4.274004403550962e-06, + 'epoch': 9.54} +04/20 [01:07:29] INFO | >> train_qwenlatent.py:487 + Step 37830 | grad_norm_pre_clip=0.1483 | + grad_norm_pre_clip_avg=0.1555 | Metrics: + {'align_loss': 0.02667500637471676, + 'recon_loss': 0.13503128290176392, + 'predict_loss': 0.0072372592985630035, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14826692640781403, + 'data_time': 0.0008867270080372691, + 'model_time': 1.2432953209790867, + 'grad_norm_pre_clip_avg': 0.1554958462715149, + 'learning_rate': 4.267454943359419e-06, + 'epoch': 9.55} +04/20 [01:07:42] INFO | >> train_qwenlatent.py:487 + Step 37840 | grad_norm_pre_clip=0.1361 | + grad_norm_pre_clip_avg=0.1577 | Metrics: + {'align_loss': 0.02606280893087387, + 'recon_loss': 0.1460607349872589, + 'predict_loss': 0.00670688645914197, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1360502690076828, + 'data_time': 0.0006870919896755368, + 'model_time': 1.291480333020445, + 'grad_norm_pre_clip_avg': 0.15769323855638503, + 'learning_rate': 4.260909501702675e-06, + 'epoch': 9.55} +04/20 [01:07:55] INFO | >> train_qwenlatent.py:487 + Step 37850 | grad_norm_pre_clip=0.2119 | + grad_norm_pre_clip_avg=0.1644 | Metrics: + {'align_loss': 0.025699596852064133, + 'recon_loss': 0.0658854991197586, + 'predict_loss': 0.0024208102840930223, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21190685033798218, + 'mae_score': 0.0069126154925372146, + 'data_time': 0.0006623160152230412, + 'model_time': 1.1958202710084151, + 'grad_norm_pre_clip_avg': 0.1644219219684601, + 'learning_rate': 4.254368081770899e-06, + 'epoch': 9.55} +04/20 [01:08:08] INFO | >> train_qwenlatent.py:487 + Step 37860 | grad_norm_pre_clip=0.2494 | + grad_norm_pre_clip_avg=0.1811 | Metrics: + {'align_loss': 0.02654724195599556, + 'recon_loss': 0.14049090445041656, + 'predict_loss': 0.010414314456284046, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24935825169086456, + 'data_time': 0.0006960739847272635, + 'model_time': 1.211995398014551, + 'grad_norm_pre_clip_avg': 0.18107060641050338, + 'learning_rate': 4.247830686752299e-06, + 'epoch': 9.55} +04/20 [01:08:21] INFO | >> train_qwenlatent.py:487 + Step 37870 | grad_norm_pre_clip=0.1127 | + grad_norm_pre_clip_avg=0.1797 | Metrics: + {'align_loss': 0.024278797209262848, + 'recon_loss': 0.10369005799293518, + 'predict_loss': 0.005595151800662279, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11272668838500977, + 'data_time': 0.0010611279867589474, + 'model_time': 1.3037013249995653, + 'grad_norm_pre_clip_avg': 0.17971531972289084, + 'learning_rate': 4.241297319833121e-06, + 'epoch': 9.56} +04/20 [01:08:34] INFO | >> train_qwenlatent.py:487 + Step 37880 | grad_norm_pre_clip=0.1605 | + grad_norm_pre_clip_avg=0.1746 | Metrics: + {'align_loss': 0.025454726070165634, + 'recon_loss': 0.16395758092403412, + 'predict_loss': 0.010719940066337585, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1605222523212433, + 'data_time': 0.0009311100002378225, + 'model_time': 1.2647094849962741, + 'grad_norm_pre_clip_avg': 0.17461751401424408, + 'learning_rate': 4.234767984197646e-06, + 'epoch': 9.56} +04/20 [01:08:46] INFO | >> train_qwenlatent.py:487 + Step 37890 | grad_norm_pre_clip=0.1721 | + grad_norm_pre_clip_avg=0.1763 | Metrics: + {'align_loss': 0.025831837207078934, + 'recon_loss': 0.1010233461856842, + 'predict_loss': 0.003497892990708351, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17209979891777039, + 'data_time': 0.0009330849861726165, + 'model_time': 1.1988795869983733, + 'grad_norm_pre_clip_avg': 0.1762520581483841, + 'learning_rate': 4.228242683028201e-06, + 'epoch': 9.56} +04/20 [01:08:59] INFO | >> train_qwenlatent.py:487 + Step 37900 | grad_norm_pre_clip=0.1461 | + grad_norm_pre_clip_avg=0.1557 | Metrics: + {'align_loss': 0.024103332310914993, + 'recon_loss': 0.09012642502784729, + 'predict_loss': 0.006285427138209343, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14606468379497528, + 'mae_score': 0.006109158198038737, 'data_time': + 0.0006929240189492702, 'model_time': + 1.20526389600127, 'grad_norm_pre_clip_avg': + 0.15568142235279084, 'learning_rate': + 4.221721419505134e-06, 'epoch': 9.56} +04/20 [01:09:12] INFO | >> train_qwenlatent.py:487 + Step 37910 | grad_norm_pre_clip=0.1564 | + grad_norm_pre_clip_avg=0.1508 | Metrics: + {'align_loss': 0.025888832286000252, + 'recon_loss': 0.12892772257328033, + 'predict_loss': 0.008420014753937721, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15644581615924835, + 'data_time': 0.0007096300250850618, + 'model_time': 1.2338861839962192, + 'grad_norm_pre_clip_avg': 0.15075394362211228, + 'learning_rate': 4.215204196806831e-06, + 'epoch': 9.57} +04/20 [01:09:24] INFO | >> train_qwenlatent.py:487 + Step 37920 | grad_norm_pre_clip=0.1453 | + grad_norm_pre_clip_avg=0.1753 | Metrics: + {'align_loss': 0.02601890079677105, + 'recon_loss': 0.11881197243928909, + 'predict_loss': 0.005621183663606644, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14531920850276947, + 'data_time': 0.0007645000005140901, + 'model_time': 1.2590770370152313, + 'grad_norm_pre_clip_avg': 0.17534171640872956, + 'learning_rate': 4.2086910181097e-06, 'epoch': + 9.57} +04/20 [01:09:37] INFO | >> train_qwenlatent.py:487 + Step 37930 | grad_norm_pre_clip=0.1528 | + grad_norm_pre_clip_avg=0.1671 | Metrics: + {'align_loss': 0.024921931326389313, + 'recon_loss': 0.10358517616987228, + 'predict_loss': 0.005667810328304768, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15276439487934113, + 'data_time': 0.0010497109906282276, + 'model_time': 1.2647155589947943, + 'grad_norm_pre_clip_avg': 0.1671190820634365, + 'learning_rate': 4.202181886588196e-06, + 'epoch': 9.57} +04/20 [01:09:49] INFO | >> train_qwenlatent.py:487 + Step 37940 | grad_norm_pre_clip=0.1712 | + grad_norm_pre_clip_avg=0.1539 | Metrics: + {'align_loss': 0.023462504148483276, + 'recon_loss': 0.0933932214975357, + 'predict_loss': 0.006523220334202051, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17124781012535095, + 'data_time': 0.0007409839890897274, + 'model_time': 1.2413199479924515, + 'grad_norm_pre_clip_avg': 0.1538635365664959, + 'learning_rate': 4.195676805414784e-06, + 'epoch': 9.57} +04/20 [01:10:02] INFO | >> train_qwenlatent.py:487 + Step 37950 | grad_norm_pre_clip=0.2353 | + grad_norm_pre_clip_avg=0.1860 | Metrics: + {'align_loss': 0.02631755918264389, + 'recon_loss': 0.16849163174629211, + 'predict_loss': 0.010077978484332561, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2353334128856659, + 'mae_score': 0.00689875585538847, 'data_time': + 0.0006713999900966883, 'model_time': + 1.2306969669880345, 'grad_norm_pre_clip_avg': + 0.1860234946012497, 'learning_rate': + 4.189175777759963e-06, 'epoch': 9.58} +04/20 [01:10:15] INFO | >> train_qwenlatent.py:487 + Step 37960 | grad_norm_pre_clip=0.1957 | + grad_norm_pre_clip_avg=0.1685 | Metrics: + {'align_loss': 0.026473594829440117, + 'recon_loss': 0.11307964473962784, + 'predict_loss': 0.007020689081400633, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19567056000232697, + 'data_time': 0.0006395630189217627, + 'model_time': 1.2073643959884066, + 'grad_norm_pre_clip_avg': 0.16850597709417342, + 'learning_rate': 4.182678806792257e-06, + 'epoch': 9.58} +04/20 [01:10:28] INFO | >> train_qwenlatent.py:487 + Step 37970 | grad_norm_pre_clip=0.1335 | + grad_norm_pre_clip_avg=0.1747 | Metrics: + {'align_loss': 0.02433256432414055, + 'recon_loss': 0.10389504581689835, + 'predict_loss': 0.004505680873990059, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1335223913192749, + 'data_time': 0.0009504710033070296, + 'model_time': 1.240096970985178, + 'grad_norm_pre_clip_avg': 0.17466698139905928, + 'learning_rate': 4.176185895678203e-06, + 'epoch': 9.58} +04/20 [01:10:40] INFO | >> train_qwenlatent.py:487 + Step 37980 | grad_norm_pre_clip=0.1576 | + grad_norm_pre_clip_avg=0.1691 | Metrics: + {'align_loss': 0.026429129764437675, + 'recon_loss': 0.1106863021850586, + 'predict_loss': 0.006697102449834347, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15759386122226715, + 'data_time': 0.0008296589949168265, + 'model_time': 1.5140979819989298, + 'grad_norm_pre_clip_avg': 0.16912018060684203, + 'learning_rate': 4.169697047582377e-06, + 'epoch': 9.58} +04/20 [01:10:53] INFO | >> train_qwenlatent.py:487 + Step 37990 | grad_norm_pre_clip=0.1368 | + grad_norm_pre_clip_avg=0.1485 | Metrics: + {'align_loss': 0.02602522447705269, + 'recon_loss': 0.12457242608070374, + 'predict_loss': 0.00735242897644639, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1368274837732315, + 'data_time': 0.0005870579916518182, + 'model_time': 1.2290883640234824, + 'grad_norm_pre_clip_avg': 0.14846659004688262, + 'learning_rate': 4.163212265667361e-06, + 'epoch': 9.59} +04/20 [01:11:07] INFO | >> train_qwenlatent.py:487 + Step 38000 | grad_norm_pre_clip=0.1562 | + grad_norm_pre_clip_avg=0.1494 | Metrics: + {'align_loss': 0.026509296149015427, + 'recon_loss': 0.12707509100437164, + 'predict_loss': 0.004871555138379335, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1562267541885376, + 'mae_score': 0.005442155374063028, 'data_time': + 0.0009558279998600483, 'model_time': + 1.6281837940041441, 'grad_norm_pre_clip_avg': + 0.14938255101442338, 'learning_rate': + 4.156731553093762e-06, 'epoch': 9.59} +04/20 [01:11:19] INFO | >> train_qwenlatent.py:487 + Step 38010 | grad_norm_pre_clip=0.1551 | + grad_norm_pre_clip_avg=0.1741 | Metrics: + {'align_loss': 0.02415413036942482, + 'recon_loss': 0.10527150332927704, + 'predict_loss': 0.0057187993079423904, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15512005984783173, + 'data_time': 0.0006650610012002289, + 'model_time': 1.2776597490010317, + 'grad_norm_pre_clip_avg': 0.1741257056593895, + 'learning_rate': 4.150254913020193e-06, + 'epoch': 9.59} +04/20 [01:11:32] INFO | >> train_qwenlatent.py:487 + Step 38020 | grad_norm_pre_clip=0.1210 | + grad_norm_pre_clip_avg=0.1636 | Metrics: + {'align_loss': 0.024956144392490387, + 'recon_loss': 0.08413860201835632, + 'predict_loss': 0.005694460589438677, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1209646686911583, + 'data_time': 0.0007749300275463611, + 'model_time': 1.2433175490004942, + 'grad_norm_pre_clip_avg': 0.163622335344553, + 'learning_rate': 4.14378234860329e-06, 'epoch': + 9.59} +04/20 [01:11:44] INFO | >> train_qwenlatent.py:487 + Step 38030 | grad_norm_pre_clip=0.1649 | + grad_norm_pre_clip_avg=0.1541 | Metrics: + {'align_loss': 0.025206372141838074, + 'recon_loss': 0.10542407631874084, + 'predict_loss': 0.004236698616296053, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16485093533992767, + 'data_time': 0.0007150769815780222, + 'model_time': 1.2372518480115104, + 'grad_norm_pre_clip_avg': 0.15411090180277826, + 'learning_rate': 4.137313862997709e-06, + 'epoch': 9.6} +04/20 [01:11:57] INFO | >> train_qwenlatent.py:487 + Step 38040 | grad_norm_pre_clip=0.1605 | + grad_norm_pre_clip_avg=0.1813 | Metrics: + {'align_loss': 0.025515785440802574, + 'recon_loss': 0.09143431484699249, + 'predict_loss': 0.00402521388605237, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16045138239860535, + 'data_time': 0.0009312069742009044, + 'model_time': 1.2643434650090057, + 'grad_norm_pre_clip_avg': 0.18126270920038223, + 'learning_rate': 4.13084945935611e-06, 'epoch': + 9.6} +04/20 [01:12:10] INFO | >> train_qwenlatent.py:487 + Step 38050 | grad_norm_pre_clip=0.1662 | + grad_norm_pre_clip_avg=0.1525 | Metrics: + {'align_loss': 0.025500256568193436, + 'recon_loss': 0.13769276440143585, + 'predict_loss': 0.006722445599734783, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16620281338691711, + 'mae_score': 0.006237531352687526, 'data_time': + 0.0006954460113774985, 'model_time': + 1.229395653004758, 'grad_norm_pre_clip_avg': + 0.1525093697011471, 'learning_rate': + 4.124389140829161e-06, 'epoch': 9.6} +04/20 [01:12:23] INFO | >> train_qwenlatent.py:487 + Step 38060 | grad_norm_pre_clip=0.1766 | + grad_norm_pre_clip_avg=0.1669 | Metrics: + {'align_loss': 0.024763327091932297, + 'recon_loss': 0.12176024913787842, + 'predict_loss': 0.0051788995042443275, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17664465308189392, + 'data_time': 0.0008332619909197092, + 'model_time': 1.2166704650153406, + 'grad_norm_pre_clip_avg': 0.166868594288826, + 'learning_rate': 4.117932910565547e-06, + 'epoch': 9.6} +04/20 [01:12:35] INFO | >> train_qwenlatent.py:487 + Step 38070 | grad_norm_pre_clip=0.1714 | + grad_norm_pre_clip_avg=0.1625 | Metrics: + {'align_loss': 0.02403850294649601, + 'recon_loss': 0.12481289356946945, + 'predict_loss': 0.006859627552330494, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17139717936515808, + 'data_time': 0.000807842006906867, + 'model_time': 1.260138112003915, + 'grad_norm_pre_clip_avg': 0.1625327482819557, + 'learning_rate': 4.111480771711948e-06, + 'epoch': 9.61} +04/20 [01:12:48] INFO | >> train_qwenlatent.py:487 + Step 38080 | grad_norm_pre_clip=0.2312 | + grad_norm_pre_clip_avg=0.1690 | Metrics: + {'align_loss': 0.026582136750221252, + 'recon_loss': 0.09941085427999496, + 'predict_loss': 0.005280639976263046, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2311948835849762, + 'data_time': 0.0008874150225892663, + 'model_time': 1.2597931519849226, + 'grad_norm_pre_clip_avg': 0.16902133896946908, + 'learning_rate': 4.1050327274130695e-06, + 'epoch': 9.61} +04/20 [01:13:01] INFO | >> train_qwenlatent.py:487 + Step 38090 | grad_norm_pre_clip=0.1660 | + grad_norm_pre_clip_avg=0.1646 | Metrics: + {'align_loss': 0.02529507502913475, + 'recon_loss': 0.1472667008638382, + 'predict_loss': 0.007490455638617277, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16596762835979462, + 'data_time': 0.001799888996174559, + 'model_time': 1.2814588340115733, + 'grad_norm_pre_clip_avg': 0.16458893865346907, + 'learning_rate': 4.098588780811608e-06, + 'epoch': 9.61} +04/20 [01:13:14] INFO | >> train_qwenlatent.py:487 + Step 38100 | grad_norm_pre_clip=0.2286 | + grad_norm_pre_clip_avg=0.1706 | Metrics: + {'align_loss': 0.02408713847398758, + 'recon_loss': 0.10093927383422852, + 'predict_loss': 0.005117520689964294, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22864457964897156, + 'mae_score': 0.00683079109535561, 'data_time': + 0.0010480909841135144, 'model_time': + 1.226029846991878, 'grad_norm_pre_clip_avg': + 0.1705641508102417, 'learning_rate': + 4.092148935048258e-06, 'epoch': 9.61} +04/20 [01:13:26] INFO | >> train_qwenlatent.py:487 + Step 38110 | grad_norm_pre_clip=0.2609 | + grad_norm_pre_clip_avg=0.1893 | Metrics: + {'align_loss': 0.02632543258368969, + 'recon_loss': 0.1610323041677475, + 'predict_loss': 0.012759415432810783, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2608591318130493, + 'data_time': 0.0010718439880292863, + 'model_time': 1.2090041909832507, + 'grad_norm_pre_clip_avg': 0.1893095925450325, + 'learning_rate': 4.085713193261721e-06, + 'epoch': 9.62} +04/20 [01:13:39] INFO | >> train_qwenlatent.py:487 + Step 38120 | grad_norm_pre_clip=0.1690 | + grad_norm_pre_clip_avg=0.1628 | Metrics: + {'align_loss': 0.025370966643095016, + 'recon_loss': 0.11305195838212967, + 'predict_loss': 0.006842545699328184, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1689745932817459, + 'data_time': 0.0010320809960830957, + 'model_time': 1.2726552999811247, + 'grad_norm_pre_clip_avg': 0.1627733126282692, + 'learning_rate': 4.079281558588708e-06, + 'epoch': 9.62} +04/20 [01:13:52] INFO | >> train_qwenlatent.py:487 + Step 38130 | grad_norm_pre_clip=0.1324 | + grad_norm_pre_clip_avg=0.1616 | Metrics: + {'align_loss': 0.025341061875224113, + 'recon_loss': 0.11122468113899231, + 'predict_loss': 0.008835084736347198, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1323787122964859, + 'data_time': 0.001219836005475372, + 'model_time': 1.558893265988445, + 'grad_norm_pre_clip_avg': 0.16164601743221282, + 'learning_rate': 4.072854034163914e-06, + 'epoch': 9.62} +04/20 [01:14:05] INFO | >> train_qwenlatent.py:487 + Step 38140 | grad_norm_pre_clip=0.1917 | + grad_norm_pre_clip_avg=0.1734 | Metrics: + {'align_loss': 0.025722917169332504, + 'recon_loss': 0.15111666917800903, + 'predict_loss': 0.01235284749418497, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19165721535682678, + 'data_time': 0.0006830640195403248, + 'model_time': 1.2685459689819254, + 'grad_norm_pre_clip_avg': 0.17343011647462844, + 'learning_rate': 4.06643062312004e-06, 'epoch': + 9.62} +04/20 [01:14:18] INFO | >> train_qwenlatent.py:487 + Step 38150 | grad_norm_pre_clip=0.1662 | + grad_norm_pre_clip_avg=0.1476 | Metrics: + {'align_loss': 0.024505335837602615, + 'recon_loss': 0.10268519818782806, + 'predict_loss': 0.007257822435349226, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1661873310804367, + 'mae_score': 0.006072041365477416, 'data_time': + 0.0009669930150266737, 'model_time': + 1.292058529012138, 'grad_norm_pre_clip_avg': + 0.14760054796934127, 'learning_rate': + 4.060011328587776e-06, 'epoch': 9.63} +04/20 [01:14:30] INFO | >> train_qwenlatent.py:487 + Step 38160 | grad_norm_pre_clip=0.1572 | + grad_norm_pre_clip_avg=0.1402 | Metrics: + {'align_loss': 0.025961320847272873, + 'recon_loss': 0.11602164804935455, + 'predict_loss': 0.007132595870643854, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15723934769630432, + 'data_time': 0.0011778049811255187, + 'model_time': 1.2432561250170693, + 'grad_norm_pre_clip_avg': 0.14023493751883506, + 'learning_rate': 4.0535961536958074e-06, + 'epoch': 9.63} +04/20 [01:14:43] INFO | >> train_qwenlatent.py:487 + Step 38170 | grad_norm_pre_clip=0.1997 | + grad_norm_pre_clip_avg=0.1803 | Metrics: + {'align_loss': 0.02509058266878128, + 'recon_loss': 0.12080167979001999, + 'predict_loss': 0.01095209363847971, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19973742961883545, + 'data_time': 0.0011008709843736142, + 'model_time': 1.2594554020033684, + 'grad_norm_pre_clip_avg': 0.18033475279808045, + 'learning_rate': 4.04718510157081e-06, 'epoch': + 9.63} +04/20 [01:14:56] INFO | >> train_qwenlatent.py:487 + Step 38180 | grad_norm_pre_clip=0.2579 | + grad_norm_pre_clip_avg=0.1827 | Metrics: + {'align_loss': 0.02655135467648506, + 'recon_loss': 0.14573229849338531, + 'predict_loss': 0.010567259043455124, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25792649388313293, + 'data_time': 0.0009116319997701794, + 'model_time': 1.2147210270049982, + 'grad_norm_pre_clip_avg': 0.18267889469861984, + 'learning_rate': 4.040778175337464e-06, + 'epoch': 9.63} +04/20 [01:15:08] INFO | >> train_qwenlatent.py:487 + Step 38190 | grad_norm_pre_clip=0.1873 | + grad_norm_pre_clip_avg=0.1679 | Metrics: + {'align_loss': 0.02543019875884056, + 'recon_loss': 0.13441208004951477, + 'predict_loss': 0.006862318143248558, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18734261393547058, + 'data_time': 0.0011251599935349077, + 'model_time': 1.2663804400071967, + 'grad_norm_pre_clip_avg': 0.16790597438812255, + 'learning_rate': 4.034375378118417e-06, + 'epoch': 9.64} +04/20 [01:15:21] INFO | >> train_qwenlatent.py:487 + Step 38200 | grad_norm_pre_clip=0.1309 | + grad_norm_pre_clip_avg=0.1529 | Metrics: + {'align_loss': 0.024715606123209, 'recon_loss': + 0.09865604341030121, 'predict_loss': + 0.00373000162653625, 'aux_loss_decay_weight': + 0.0, 'grad_norm_pre_clip': 0.13085110485553741, + 'mae_score': 0.0068759853775436815, + 'data_time': 0.001579170988406986, + 'model_time': 1.2336177259858232, + 'grad_norm_pre_clip_avg': 0.15288887470960616, + 'learning_rate': 4.027976713034316e-06, + 'epoch': 9.64} +04/20 [01:15:34] INFO | >> train_qwenlatent.py:487 + Step 38210 | grad_norm_pre_clip=0.1613 | + grad_norm_pre_clip_avg=0.1589 | Metrics: + {'align_loss': 0.025957398116588593, + 'recon_loss': 0.08341342210769653, + 'predict_loss': 0.00576340826228261, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1613159328699112, + 'data_time': 0.0007216039812192321, + 'model_time': 1.2138389149913564, + 'grad_norm_pre_clip_avg': 0.15885735005140306, + 'learning_rate': 4.021582183203789e-06, + 'epoch': 9.64} +04/20 [01:15:47] INFO | >> train_qwenlatent.py:487 + Step 38220 | grad_norm_pre_clip=0.1928 | + grad_norm_pre_clip_avg=0.1489 | Metrics: + {'align_loss': 0.024977896362543106, + 'recon_loss': 0.10290997475385666, + 'predict_loss': 0.005249917507171631, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19284379482269287, + 'data_time': 0.0009947209910023957, + 'model_time': 1.197116155992262, + 'grad_norm_pre_clip_avg': 0.14894191399216652, + 'learning_rate': 4.015191791743464e-06, + 'epoch': 9.64} +04/20 [01:15:59] INFO | >> train_qwenlatent.py:487 + Step 38230 | grad_norm_pre_clip=0.1138 | + grad_norm_pre_clip_avg=0.1600 | Metrics: + {'align_loss': 0.0264500230550766, + 'recon_loss': 0.10528642684221268, + 'predict_loss': 0.0031846065539866686, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11377173662185669, + 'data_time': 0.000642275990685448, + 'model_time': 1.2413572920195293, + 'grad_norm_pre_clip_avg': 0.15997333079576492, + 'learning_rate': 4.00880554176793e-06, 'epoch': + 9.65} +04/20 [01:16:12] INFO | >> train_qwenlatent.py:487 + Step 38240 | grad_norm_pre_clip=0.1952 | + grad_norm_pre_clip_avg=0.1682 | Metrics: + {'align_loss': 0.024954818189144135, + 'recon_loss': 0.09489187598228455, + 'predict_loss': 0.004000889603048563, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19524189829826355, + 'data_time': 0.0008216020069085062, + 'model_time': 1.451340175990481, + 'grad_norm_pre_clip_avg': 0.16821749359369279, + 'learning_rate': 4.002423436389773e-06, + 'epoch': 9.65} +04/20 [01:16:26] INFO | >> train_qwenlatent.py:487 + Step 38250 | grad_norm_pre_clip=0.1178 | + grad_norm_pre_clip_avg=0.1442 | Metrics: + {'align_loss': 0.025595080107450485, + 'recon_loss': 0.12360534816980362, + 'predict_loss': 0.009281054139137268, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11775210499763489, + 'mae_score': 0.008696001070039766, 'data_time': + 0.0007844459905754775, 'model_time': + 1.200533536990406, 'grad_norm_pre_clip_avg': + 0.14417664408683778, 'learning_rate': + 3.9960454787195515e-06, 'epoch': 9.65} +04/20 [01:16:38] INFO | >> train_qwenlatent.py:487 + Step 38260 | grad_norm_pre_clip=0.1620 | + grad_norm_pre_clip_avg=0.1494 | Metrics: + {'align_loss': 0.025267640128731728, + 'recon_loss': 0.14955434203147888, + 'predict_loss': 0.011132870800793171, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16195330023765564, + 'data_time': 0.0006705719861201942, + 'model_time': 1.2948134540056344, + 'grad_norm_pre_clip_avg': 0.14936349764466286, + 'learning_rate': 3.989671671865799e-06, + 'epoch': 9.65} +04/20 [01:16:51] INFO | >> train_qwenlatent.py:487 + Step 38270 | grad_norm_pre_clip=0.1939 | + grad_norm_pre_clip_avg=0.1582 | Metrics: + {'align_loss': 0.025115374475717545, + 'recon_loss': 0.12024449557065964, + 'predict_loss': 0.006717925891280174, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19392970204353333, + 'data_time': 0.0007817019941285253, + 'model_time': 1.522773298987886, + 'grad_norm_pre_clip_avg': 0.15824738293886184, + 'learning_rate': 3.983302018935048e-06, + 'epoch': 9.66} +04/20 [01:17:04] INFO | >> train_qwenlatent.py:487 + Step 38280 | grad_norm_pre_clip=0.2101 | + grad_norm_pre_clip_avg=0.1844 | Metrics: + {'align_loss': 0.025573479011654854, + 'recon_loss': 0.09173494577407837, + 'predict_loss': 0.0075895837508141994, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2100999355316162, + 'data_time': 0.000827531999675557, + 'model_time': 1.2396216150082182, + 'grad_norm_pre_clip_avg': 0.18440971821546553, + 'learning_rate': 3.976936523031773e-06, + 'epoch': 9.66} +04/20 [01:17:16] INFO | >> train_qwenlatent.py:487 + Step 38290 | grad_norm_pre_clip=0.2053 | + grad_norm_pre_clip_avg=0.1934 | Metrics: + {'align_loss': 0.02521405927836895, + 'recon_loss': 0.1411123275756836, + 'predict_loss': 0.005772924516350031, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2053443193435669, + 'data_time': 0.0007049159903544933, + 'model_time': 1.2022874359972775, + 'grad_norm_pre_clip_avg': 0.1934471145272255, + 'learning_rate': 3.9705751872584455e-06, + 'epoch': 9.66} +04/20 [01:17:29] INFO | >> train_qwenlatent.py:487 + Step 38300 | grad_norm_pre_clip=0.1891 | + grad_norm_pre_clip_avg=0.1738 | Metrics: + {'align_loss': 0.024924026802182198, + 'recon_loss': 0.1282573938369751, + 'predict_loss': 0.006953551433980465, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18909934163093567, + 'mae_score': 0.005895559637396185, 'data_time': + 0.0008201579912565649, 'model_time': + 1.212244101014221, 'grad_norm_pre_clip_avg': + 0.17376006618142129, 'learning_rate': + 3.9642180147155e-06, 'epoch': 9.66} +04/20 [01:17:42] INFO | >> train_qwenlatent.py:487 + Step 38310 | grad_norm_pre_clip=0.1772 | + grad_norm_pre_clip_avg=0.1592 | Metrics: + {'align_loss': 0.02590300887823105, + 'recon_loss': 0.14767439663410187, + 'predict_loss': 0.010288762860000134, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17717775702476501, + 'data_time': 0.001283019984839484, + 'model_time': 1.2434890529839322, + 'grad_norm_pre_clip_avg': 0.1592288002371788, + 'learning_rate': 3.957865008501352e-06, + 'epoch': 9.67} +04/20 [01:17:54] INFO | >> train_qwenlatent.py:487 + Step 38320 | grad_norm_pre_clip=0.1789 | + grad_norm_pre_clip_avg=0.1603 | Metrics: + {'align_loss': 0.025172976776957512, + 'recon_loss': 0.08742153644561768, + 'predict_loss': 0.006861030589789152, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1788579672574997, + 'data_time': 0.0007395329885184765, + 'model_time': 1.20435323799029, + 'grad_norm_pre_clip_avg': 0.16028255224227905, + 'learning_rate': 3.9515161717123756e-06, + 'epoch': 9.67} +04/20 [01:18:07] INFO | >> train_qwenlatent.py:487 + Step 38330 | grad_norm_pre_clip=0.2020 | + grad_norm_pre_clip_avg=0.1682 | Metrics: + {'align_loss': 0.02547745779156685, + 'recon_loss': 0.1804681271314621, + 'predict_loss': 0.012793811038136482, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2019980102777481, + 'data_time': 0.0008482200209982693, + 'model_time': 1.2078180799726397, + 'grad_norm_pre_clip_avg': 0.16824741065502166, + 'learning_rate': 3.945171507442917e-06, + 'epoch': 9.67} +04/20 [01:18:20] INFO | >> train_qwenlatent.py:487 + Step 38340 | grad_norm_pre_clip=0.1947 | + grad_norm_pre_clip_avg=0.1640 | Metrics: + {'align_loss': 0.02548319101333618, + 'recon_loss': 0.17327751219272614, + 'predict_loss': 0.013478659093379974, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19472236931324005, + 'data_time': 0.0007001179910730571, + 'model_time': 1.1782841839885805, + 'grad_norm_pre_clip_avg': 0.1639543280005455, + 'learning_rate': 3.93883101878529e-06, 'epoch': + 9.67} +04/20 [01:18:33] INFO | >> train_qwenlatent.py:487 + Step 38350 | grad_norm_pre_clip=0.1572 | + grad_norm_pre_clip_avg=0.1736 | Metrics: + {'align_loss': 0.024290673434734344, + 'recon_loss': 0.15523844957351685, + 'predict_loss': 0.007380153983831406, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15720213949680328, + 'mae_score': 0.006186474121368683, 'data_time': + 0.0006811619969084859, 'model_time': + 1.251796051976271, 'grad_norm_pre_clip_avg': + 0.1735580176115036, 'learning_rate': + 3.9324947088297695e-06, 'epoch': 9.68} +04/20 [01:18:45] INFO | >> train_qwenlatent.py:487 + Step 38360 | grad_norm_pre_clip=0.1646 | + grad_norm_pre_clip_avg=0.1771 | Metrics: + {'align_loss': 0.02567451074719429, + 'recon_loss': 0.12110964208841324, + 'predict_loss': 0.0046396111138165, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16456054151058197, + 'data_time': 0.00108452801941894, 'model_time': + 1.266198439989239, 'grad_norm_pre_clip_avg': + 0.1771413803100586, 'learning_rate': + 3.926162580664596e-06, 'epoch': 9.68} +04/20 [01:18:58] INFO | >> train_qwenlatent.py:487 + Step 38370 | grad_norm_pre_clip=0.1916 | + grad_norm_pre_clip_avg=0.1710 | Metrics: + {'align_loss': 0.02503332495689392, + 'recon_loss': 0.0878259465098381, + 'predict_loss': 0.005125894211232662, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19159317016601562, + 'data_time': 0.0006904850015416741, + 'model_time': 1.2478334389743395, + 'grad_norm_pre_clip_avg': 0.17099912762641906, + 'learning_rate': 3.9198346373759745e-06, + 'epoch': 9.68} +04/20 [01:19:10] INFO | >> train_qwenlatent.py:487 + Step 38380 | grad_norm_pre_clip=0.1882 | + grad_norm_pre_clip_avg=0.1527 | Metrics: + {'align_loss': 0.02519865706562996, + 'recon_loss': 0.09956077486276627, + 'predict_loss': 0.004951284732669592, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18819016218185425, + 'data_time': 0.0009232869779225439, + 'model_time': 1.204501853993861, + 'grad_norm_pre_clip_avg': 0.15273839831352234, + 'learning_rate': 3.913510882048064e-06, + 'epoch': 9.68} +04/20 [01:19:23] INFO | >> train_qwenlatent.py:487 + Step 38390 | grad_norm_pre_clip=0.2566 | + grad_norm_pre_clip_avg=0.1920 | Metrics: + {'align_loss': 0.024369284510612488, + 'recon_loss': 0.1760181337594986, + 'predict_loss': 0.009106665849685669, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25655388832092285, + 'data_time': 0.0012462410086300224, + 'model_time': 1.2904994059936143, + 'grad_norm_pre_clip_avg': 0.19202018678188323, + 'learning_rate': 3.907191317762989e-06, + 'epoch': 9.69} +04/20 [01:19:37] INFO | >> train_qwenlatent.py:487 + Step 38400 | grad_norm_pre_clip=0.1881 | + grad_norm_pre_clip_avg=0.1884 | Metrics: + {'align_loss': 0.025682367384433746, + 'recon_loss': 0.1876058280467987, + 'predict_loss': 0.013568001799285412, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1881440281867981, + 'mae_score': 0.006651602135048256, 'data_time': + 0.0006916710117366165, 'model_time': + 1.209276977024274, 'grad_norm_pre_clip_avg': + 0.1884021855890751, 'learning_rate': + 3.900875947600824e-06, 'epoch': 9.69} +04/20 [01:19:49] INFO | >> train_qwenlatent.py:487 + Step 38410 | grad_norm_pre_clip=0.1419 | + grad_norm_pre_clip_avg=0.1704 | Metrics: + {'align_loss': 0.025617845356464386, + 'recon_loss': 0.09595873206853867, + 'predict_loss': 0.002803853480145335, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14189982414245605, + 'data_time': 0.0007507659902330488, + 'model_time': 1.2236053929955233, + 'grad_norm_pre_clip_avg': 0.17037766873836518, + 'learning_rate': 3.894564774639608e-06, + 'epoch': 9.69} +04/20 [01:20:02] INFO | >> train_qwenlatent.py:487 + Step 38420 | grad_norm_pre_clip=0.1340 | + grad_norm_pre_clip_avg=0.1345 | Metrics: + {'align_loss': 0.02568749338388443, + 'recon_loss': 0.12491859495639801, + 'predict_loss': 0.006369295530021191, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13404633104801178, + 'data_time': 0.0006800020055379719, + 'model_time': 1.2800611010170542, + 'grad_norm_pre_clip_avg': 0.13452127650380136, + 'learning_rate': 3.8882578019553314e-06, + 'epoch': 9.69} +04/20 [01:20:14] INFO | >> train_qwenlatent.py:487 + Step 38430 | grad_norm_pre_clip=0.1081 | + grad_norm_pre_clip_avg=0.1428 | Metrics: + {'align_loss': 0.02524608001112938, + 'recon_loss': 0.09099607169628143, + 'predict_loss': 0.006219177506864071, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10811330378055573, + 'data_time': 0.0008055580256041139, + 'model_time': 1.2319326919969171, + 'grad_norm_pre_clip_avg': 0.14276972636580468, + 'learning_rate': 3.881955032621934e-06, + 'epoch': 9.7} +04/20 [01:20:27] INFO | >> train_qwenlatent.py:487 + Step 38440 | grad_norm_pre_clip=0.2002 | + grad_norm_pre_clip_avg=0.1634 | Metrics: + {'align_loss': 0.02485419064760208, + 'recon_loss': 0.12234538793563843, + 'predict_loss': 0.007251724135130644, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20022548735141754, + 'data_time': 0.000992020999547094, + 'model_time': 1.256747034, + 'grad_norm_pre_clip_avg': 0.16336809247732162, + 'learning_rate': 3.875656469711308e-06, + 'epoch': 9.7} +04/20 [01:20:40] INFO | >> train_qwenlatent.py:487 + Step 38450 | grad_norm_pre_clip=0.1298 | + grad_norm_pre_clip_avg=0.1521 | Metrics: + {'align_loss': 0.026057634502649307, + 'recon_loss': 0.07778195291757584, + 'predict_loss': 0.004266930744051933, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12976856529712677, + 'mae_score': 0.006786211546476897, 'data_time': + 0.0006515250133816153, 'model_time': + 1.2304547910171095, 'grad_norm_pre_clip_avg': + 0.1521452136337757, 'learning_rate': + 3.869362116293296e-06, 'epoch': 9.7} +04/20 [01:20:52] INFO | >> train_qwenlatent.py:487 + Step 38460 | grad_norm_pre_clip=0.1965 | + grad_norm_pre_clip_avg=0.1837 | Metrics: + {'align_loss': 0.026049524545669556, + 'recon_loss': 0.12686394155025482, + 'predict_loss': 0.006693197414278984, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19647879898548126, + 'data_time': 0.0006891220109537244, + 'model_time': 1.1897134039900266, + 'grad_norm_pre_clip_avg': 0.18372663110494614, + 'learning_rate': 3.863071975435691e-06, + 'epoch': 9.7} +04/20 [01:21:05] INFO | >> train_qwenlatent.py:487 + Step 38470 | grad_norm_pre_clip=0.1731 | + grad_norm_pre_clip_avg=0.1681 | Metrics: + {'align_loss': 0.025868766009807587, + 'recon_loss': 0.08435269445180893, + 'predict_loss': 0.0046216812916100025, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17312490940093994, + 'data_time': 0.0009710419981274754, + 'model_time': 1.2691829040122684, + 'grad_norm_pre_clip_avg': 0.168121999502182, + 'learning_rate': 3.856786050204232e-06, + 'epoch': 9.71} +04/20 [01:21:17] INFO | >> train_qwenlatent.py:487 + Step 38480 | grad_norm_pre_clip=0.1699 | + grad_norm_pre_clip_avg=0.1681 | Metrics: + {'align_loss': 0.02527148649096489, + 'recon_loss': 0.14480742812156677, + 'predict_loss': 0.010453267022967339, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16987678408622742, + 'data_time': 0.0006937750149518251, + 'model_time': 1.2141227450047154, + 'grad_norm_pre_clip_avg': 0.1680608294904232, + 'learning_rate': 3.8505043436626e-06, 'epoch': + 9.71} +04/20 [01:21:30] INFO | >> train_qwenlatent.py:487 + Step 38490 | grad_norm_pre_clip=0.1604 | + grad_norm_pre_clip_avg=0.1767 | Metrics: + {'align_loss': 0.024232439696788788, + 'recon_loss': 0.1386180967092514, + 'predict_loss': 0.008392703719437122, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16040262579917908, + 'data_time': 0.0009422170114703476, + 'model_time': 1.2791020199947525, + 'grad_norm_pre_clip_avg': 0.17669604867696762, + 'learning_rate': 3.844226858872421e-06, + 'epoch': 9.71} +04/20 [01:21:43] INFO | >> train_qwenlatent.py:487 + Step 38500 | grad_norm_pre_clip=0.1620 | + grad_norm_pre_clip_avg=0.1553 | Metrics: + {'align_loss': 0.024383604526519775, + 'recon_loss': 0.08968940377235413, + 'predict_loss': 0.005115084815770388, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16201148927211761, + 'mae_score': 0.008230894105928439, 'data_time': + 0.0009179420012515038, 'model_time': + 1.2134974100044928, 'grad_norm_pre_clip_avg': + 0.15528810247778893, 'learning_rate': + 3.837953598893271e-06, 'epoch': 9.71} +04/20 [01:21:56] INFO | >> train_qwenlatent.py:487 + Step 38510 | grad_norm_pre_clip=0.1922 | + grad_norm_pre_clip_avg=0.1521 | Metrics: + {'align_loss': 0.026972472667694092, + 'recon_loss': 0.1605292707681656, + 'predict_loss': 0.009271631017327309, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1922246515750885, + 'data_time': 0.0007150599849410355, + 'model_time': 1.1891206190048251, + 'grad_norm_pre_clip_avg': 0.15213975310325623, + 'learning_rate': 3.831684566782658e-06, + 'epoch': 9.72} +04/20 [01:22:09] INFO | >> train_qwenlatent.py:487 + Step 38520 | grad_norm_pre_clip=0.1639 | + grad_norm_pre_clip_avg=0.1493 | Metrics: + {'align_loss': 0.0261472649872303, + 'recon_loss': 0.15433131158351898, + 'predict_loss': 0.006707232911139727, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16390717029571533, + 'data_time': 0.0006531240069307387, + 'model_time': 1.2553077299962752, + 'grad_norm_pre_clip_avg': 0.14932583644986153, + 'learning_rate': 3.825419765596029e-06, + 'epoch': 9.72} +04/20 [01:22:21] INFO | >> train_qwenlatent.py:487 + Step 38530 | grad_norm_pre_clip=0.1899 | + grad_norm_pre_clip_avg=0.1580 | Metrics: + {'align_loss': 0.02462788298726082, + 'recon_loss': 0.07607400417327881, + 'predict_loss': 0.0045067910104990005, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1899154633283615, + 'data_time': 0.0009506189962849021, + 'model_time': 1.2474763010104652, + 'grad_norm_pre_clip_avg': 0.15799365118145942, + 'learning_rate': 3.819159198386781e-06, + 'epoch': 9.72} +04/20 [01:22:34] INFO | >> train_qwenlatent.py:487 + Step 38540 | grad_norm_pre_clip=0.1655 | + grad_norm_pre_clip_avg=0.1579 | Metrics: + {'align_loss': 0.024772850796580315, + 'recon_loss': 0.1312771439552307, + 'predict_loss': 0.004689557012170553, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16552291810512543, + 'data_time': 0.07733931101392955, 'model_time': + 1.2157810179924127, 'grad_norm_pre_clip_avg': + 0.15787000134587287, 'learning_rate': + 3.812902868206222e-06, 'epoch': 9.72} +04/20 [01:22:46] INFO | >> train_qwenlatent.py:487 + Step 38550 | grad_norm_pre_clip=0.1195 | + grad_norm_pre_clip_avg=0.1593 | Metrics: + {'align_loss': 0.02573024481534958, + 'recon_loss': 0.12090719491243362, + 'predict_loss': 0.004854404833167791, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11952847242355347, + 'mae_score': 0.005528336602288323, 'data_time': + 0.0006065960042178631, 'model_time': + 1.1817762150021736, 'grad_norm_pre_clip_avg': + 0.15931498035788536, 'learning_rate': + 3.8066507781036263e-06, 'epoch': 9.73} +04/20 [01:22:58] INFO | >> train_qwenlatent.py:487 + Step 38560 | grad_norm_pre_clip=0.1836 | + grad_norm_pre_clip_avg=0.1663 | Metrics: + {'align_loss': 0.02536505088210106, + 'recon_loss': 0.1274196058511734, + 'predict_loss': 0.010267402976751328, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18357819318771362, + 'data_time': 0.0005411310121417046, + 'model_time': 1.1852498429943807, + 'grad_norm_pre_clip_avg': 0.16626929938793183, + 'learning_rate': 3.800402931126182e-06, + 'epoch': 9.73} +04/20 [01:23:10] INFO | >> train_qwenlatent.py:487 + Step 38570 | grad_norm_pre_clip=0.1299 | + grad_norm_pre_clip_avg=0.1632 | Metrics: + {'align_loss': 0.02593693509697914, + 'recon_loss': 0.10160255432128906, + 'predict_loss': 0.0054124388843774796, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1299332231283188, + 'data_time': 0.0006338089879136533, + 'model_time': 1.133198622992495, + 'grad_norm_pre_clip_avg': 0.16316590011119841, + 'learning_rate': 3.794159330319012e-06, + 'epoch': 9.73} +04/20 [01:23:21] INFO | >> train_qwenlatent.py:487 + Step 38580 | grad_norm_pre_clip=0.1545 | + grad_norm_pre_clip_avg=0.1662 | Metrics: + {'align_loss': 0.026468785479664803, + 'recon_loss': 0.19605913758277893, + 'predict_loss': 0.009093034081161022, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15448756515979767, + 'data_time': 0.0005936000088695437, + 'model_time': 1.1585354839917272, + 'grad_norm_pre_clip_avg': 0.16622163504362106, + 'learning_rate': 3.7879199787251725e-06, + 'epoch': 9.74} +04/20 [01:23:33] INFO | >> train_qwenlatent.py:487 + Step 38590 | grad_norm_pre_clip=0.1698 | + grad_norm_pre_clip_avg=0.1663 | Metrics: + {'align_loss': 0.0270620658993721, + 'recon_loss': 0.14829176664352417, + 'predict_loss': 0.008374059572815895, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16976727545261383, + 'data_time': 0.00060076400404796, 'model_time': + 1.1459216560178902, 'grad_norm_pre_clip_avg': + 0.16629068702459335, 'learning_rate': + 3.7816848793856457e-06, 'epoch': 9.74} +04/20 [01:23:45] INFO | >> train_qwenlatent.py:487 + Step 38600 | grad_norm_pre_clip=0.1419 | + grad_norm_pre_clip_avg=0.1748 | Metrics: + {'align_loss': 0.026433251798152924, + 'recon_loss': 0.16998517513275146, + 'predict_loss': 0.010963867418467999, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14187514781951904, + 'mae_score': 0.012801896344434033, 'data_time': + 0.0008607269846834242, 'model_time': + 1.1456276710086968, 'grad_norm_pre_clip_avg': + 0.17478080689907075, 'learning_rate': + 3.775454035339348e-06, 'epoch': 9.74} +04/20 [01:23:57] INFO | >> train_qwenlatent.py:487 + Step 38610 | grad_norm_pre_clip=0.2201 | + grad_norm_pre_clip_avg=0.1654 | Metrics: + {'align_loss': 0.025168100371956825, + 'recon_loss': 0.12638843059539795, + 'predict_loss': 0.010589651763439178, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22013580799102783, + 'data_time': 0.000613615004112944, + 'model_time': 1.135852806008188, + 'grad_norm_pre_clip_avg': 0.16535344570875168, + 'learning_rate': 3.769227449623116e-06, + 'epoch': 9.74} +04/20 [01:24:09] INFO | >> train_qwenlatent.py:487 + Step 38620 | grad_norm_pre_clip=0.1718 | + grad_norm_pre_clip_avg=0.1551 | Metrics: + {'align_loss': 0.024717023596167564, + 'recon_loss': 0.09246690571308136, + 'predict_loss': 0.004309092648327351, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.171786367893219, + 'data_time': 0.0005570359935518354, + 'model_time': 1.1614784359990153, + 'grad_norm_pre_clip_avg': 0.15508795902132988, + 'learning_rate': 3.763005125271713e-06, + 'epoch': 9.75} +04/20 [01:24:20] INFO | >> train_qwenlatent.py:487 + Step 38630 | grad_norm_pre_clip=0.1289 | + grad_norm_pre_clip_avg=0.1380 | Metrics: + {'align_loss': 0.02373097464442253, + 'recon_loss': 0.07724891602993011, + 'predict_loss': 0.006683034356683493, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1288657933473587, + 'data_time': 0.0006223909731488675, + 'model_time': 1.1582235569949262, + 'grad_norm_pre_clip_avg': 0.13798973932862282, + 'learning_rate': 3.7567870653178154e-06, + 'epoch': 9.75} +04/20 [01:24:32] INFO | >> train_qwenlatent.py:487 + Step 38640 | grad_norm_pre_clip=0.2023 | + grad_norm_pre_clip_avg=0.1607 | Metrics: + {'align_loss': 0.024588825181126595, + 'recon_loss': 0.1145743876695633, + 'predict_loss': 0.006785189267247915, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.202291339635849, + 'data_time': 0.0006744840065948665, + 'model_time': 1.1438012290163897, + 'grad_norm_pre_clip_avg': 0.1607061132788658, + 'learning_rate': 3.7505732727920408e-06, + 'epoch': 9.75} +04/20 [01:24:44] INFO | >> train_qwenlatent.py:487 + Step 38650 | grad_norm_pre_clip=0.1829 | + grad_norm_pre_clip_avg=0.1606 | Metrics: + {'align_loss': 0.025300871580839157, + 'recon_loss': 0.10041525959968567, + 'predict_loss': 0.005815474316477776, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18294619023799896, + 'mae_score': 0.00577395198581455, 'data_time': + 0.000586789014050737, 'model_time': + 1.1516231799905654, 'grad_norm_pre_clip_avg': + 0.16063109636306763, 'learning_rate': + 3.7443637507229097e-06, 'epoch': 9.75} +04/20 [01:24:56] INFO | >> train_qwenlatent.py:487 + Step 38660 | grad_norm_pre_clip=0.1561 | + grad_norm_pre_clip_avg=0.1584 | Metrics: + {'align_loss': 0.02683921717107296, + 'recon_loss': 0.18824262917041779, + 'predict_loss': 0.009534583427011967, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15607120096683502, + 'data_time': 0.0005973170045763254, + 'model_time': 1.1447480310162064, + 'grad_norm_pre_clip_avg': 0.15840362012386322, + 'learning_rate': 3.738158502136872e-06, + 'epoch': 9.76} +04/20 [01:25:08] INFO | >> train_qwenlatent.py:487 + Step 38670 | grad_norm_pre_clip=0.1457 | + grad_norm_pre_clip_avg=0.1774 | Metrics: + {'align_loss': 0.025286730378866196, + 'recon_loss': 0.11895213276147842, + 'predict_loss': 0.006838584318757057, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1456608921289444, + 'data_time': 0.0005666579818353057, + 'model_time': 1.3605494519870263, + 'grad_norm_pre_clip_avg': 0.17740782722830772, + 'learning_rate': 3.731957530058289e-06, + 'epoch': 9.76} +04/20 [01:25:19] INFO | >> train_qwenlatent.py:487 + Step 38680 | grad_norm_pre_clip=0.1386 | + grad_norm_pre_clip_avg=0.1716 | Metrics: + {'align_loss': 0.025728784501552582, + 'recon_loss': 0.14059896767139435, + 'predict_loss': 0.008528740145266056, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1385820358991623, + 'data_time': 0.0006066240021027625, + 'model_time': 1.1334392769786064, + 'grad_norm_pre_clip_avg': 0.1715896636247635, + 'learning_rate': 3.7257608375094366e-06, + 'epoch': 9.76} +04/20 [01:25:31] INFO | >> train_qwenlatent.py:487 + Step 38690 | grad_norm_pre_clip=0.1714 | + grad_norm_pre_clip_avg=0.1756 | Metrics: + {'align_loss': 0.02530941367149353, + 'recon_loss': 0.1248047724366188, + 'predict_loss': 0.005947444587945938, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17139911651611328, + 'data_time': 0.0006609779957216233, + 'model_time': 1.168756202008808, + 'grad_norm_pre_clip_avg': 0.1756262555718422, + 'learning_rate': 3.7195684275105058e-06, + 'epoch': 9.76} +04/20 [01:25:43] INFO | >> train_qwenlatent.py:487 + Step 38700 | grad_norm_pre_clip=0.1341 | + grad_norm_pre_clip_avg=0.1613 | Metrics: + {'align_loss': 0.024785412475466728, + 'recon_loss': 0.08942626416683197, + 'predict_loss': 0.00543898856267333, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1340867429971695, + 'mae_score': 0.005660271000217747, 'data_time': + 0.0005981620051898062, 'model_time': + 1.1775171079789288, 'grad_norm_pre_clip_avg': + 0.16125288158655166, 'learning_rate': + 3.713380303079611e-06, 'epoch': 9.77} +04/20 [01:25:54] INFO | >> train_qwenlatent.py:487 + Step 38710 | grad_norm_pre_clip=0.1550 | + grad_norm_pre_clip_avg=0.1542 | Metrics: + {'align_loss': 0.025866659358143806, + 'recon_loss': 0.1409715861082077, + 'predict_loss': 0.007144323084503412, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15502521395683289, + 'data_time': 0.0005423390248324722, + 'model_time': 1.137021080008708, + 'grad_norm_pre_clip_avg': 0.15418037697672843, + 'learning_rate': 3.7071964672327658e-06, + 'epoch': 9.77} +04/20 [01:26:06] INFO | >> train_qwenlatent.py:487 + Step 38720 | grad_norm_pre_clip=0.1983 | + grad_norm_pre_clip_avg=0.1655 | Metrics: + {'align_loss': 0.026014523580670357, + 'recon_loss': 0.10897897183895111, + 'predict_loss': 0.005344904959201813, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19830195605754852, + 'data_time': 0.0005920339899603277, + 'model_time': 1.333626327017555, + 'grad_norm_pre_clip_avg': 0.1655449591577053, + 'learning_rate': 3.7010169229838923e-06, + 'epoch': 9.77} +04/20 [01:26:18] INFO | >> train_qwenlatent.py:487 + Step 38730 | grad_norm_pre_clip=0.2040 | + grad_norm_pre_clip_avg=0.1811 | Metrics: + {'align_loss': 0.024804402142763138, + 'recon_loss': 0.10575076192617416, + 'predict_loss': 0.005869943182915449, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20397990942001343, + 'data_time': 0.0005983010050840676, + 'model_time': 1.1627980099874549, + 'grad_norm_pre_clip_avg': 0.18105603009462357, + 'learning_rate': 3.6948416733448217e-06, + 'epoch': 9.77} +04/20 [01:26:29] INFO | >> train_qwenlatent.py:487 + Step 38740 | grad_norm_pre_clip=0.1956 | + grad_norm_pre_clip_avg=0.1859 | Metrics: + {'align_loss': 0.02599485032260418, + 'recon_loss': 0.08661147207021713, + 'predict_loss': 0.0030520495492964983, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.195609912276268, + 'data_time': 0.0005857240175828338, + 'model_time': 1.1445486119773705, + 'grad_norm_pre_clip_avg': 0.18591933697462082, + 'learning_rate': 3.688670721325306e-06, + 'epoch': 9.78} +04/20 [01:26:41] INFO | >> train_qwenlatent.py:487 + Step 38750 | grad_norm_pre_clip=0.1738 | + grad_norm_pre_clip_avg=0.1703 | Metrics: + {'align_loss': 0.02555117942392826, + 'recon_loss': 0.10377472639083862, + 'predict_loss': 0.002873543184250593, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17379826307296753, + 'mae_score': 0.00618726498371846, 'data_time': + 0.0006176319730002433, 'model_time': + 1.1600520680076443, 'grad_norm_pre_clip_avg': + 0.1703080117702484, 'learning_rate': + 3.682504069932989e-06, 'epoch': 9.78} +04/20 [01:26:53] INFO | >> train_qwenlatent.py:487 + Step 38760 | grad_norm_pre_clip=0.1562 | + grad_norm_pre_clip_avg=0.1465 | Metrics: + {'align_loss': 0.025819089263677597, + 'recon_loss': 0.12921956181526184, + 'predict_loss': 0.012346609495580196, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15615195035934448, + 'data_time': 0.0006275080086197704, + 'model_time': 1.1426037030178122, + 'grad_norm_pre_clip_avg': 0.14653078988194465, + 'learning_rate': 3.6763417221734196e-06, + 'epoch': 9.78} +04/20 [01:27:05] INFO | >> train_qwenlatent.py:487 + Step 38770 | grad_norm_pre_clip=0.1604 | + grad_norm_pre_clip_avg=0.1593 | Metrics: + {'align_loss': 0.026357144117355347, + 'recon_loss': 0.17533154785633087, + 'predict_loss': 0.010047178715467453, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.160426065325737, + 'data_time': 0.0005600920121651143, + 'model_time': 1.1612050000112504, + 'grad_norm_pre_clip_avg': 0.15929754972457885, + 'learning_rate': 3.670183681050051e-06, + 'epoch': 9.78} +04/20 [01:27:16] INFO | >> train_qwenlatent.py:487 + Step 38780 | grad_norm_pre_clip=0.1328 | + grad_norm_pre_clip_avg=0.1558 | Metrics: + {'align_loss': 0.02534431219100952, + 'recon_loss': 0.11412128061056137, + 'predict_loss': 0.0076157390139997005, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13284078240394592, + 'data_time': 0.0005984479794278741, + 'model_time': 1.186268683988601, + 'grad_norm_pre_clip_avg': 0.15577996224164964, + 'learning_rate': 3.664029949564235e-06, + 'epoch': 9.79} +04/20 [01:27:28] INFO | >> train_qwenlatent.py:487 + Step 38790 | grad_norm_pre_clip=0.1907 | + grad_norm_pre_clip_avg=0.1660 | Metrics: + {'align_loss': 0.025632701814174652, + 'recon_loss': 0.15170802175998688, + 'predict_loss': 0.005499108694493771, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19074654579162598, + 'data_time': 0.0005323590012267232, + 'model_time': 1.139080641005421, + 'grad_norm_pre_clip_avg': 0.16602656468749047, + 'learning_rate': 3.657880530715234e-06, + 'epoch': 9.79} +04/20 [01:27:40] INFO | >> train_qwenlatent.py:487 + Step 38800 | grad_norm_pre_clip=0.1778 | + grad_norm_pre_clip_avg=0.1537 | Metrics: + {'align_loss': 0.024377282708883286, + 'recon_loss': 0.10228138417005539, + 'predict_loss': 0.007901621051132679, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1777719259262085, + 'mae_score': 0.006178126893601976, 'data_time': + 0.0005538800032809377, 'model_time': + 1.1464829199831001, 'grad_norm_pre_clip_avg': + 0.15374324917793275, 'learning_rate': + 3.6517354275001984e-06, 'epoch': 9.79} +04/20 [01:27:52] INFO | >> train_qwenlatent.py:487 + Step 38810 | grad_norm_pre_clip=0.1040 | + grad_norm_pre_clip_avg=0.1357 | Metrics: + {'align_loss': 0.025935325771570206, + 'recon_loss': 0.11371652781963348, + 'predict_loss': 0.0031601879745721817, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10403339564800262, + 'data_time': 0.0005691649857908487, + 'model_time': 1.13935451299767, + 'grad_norm_pre_clip_avg': 0.1356564022600651, + 'learning_rate': 3.6455946429141693e-06, + 'epoch': 9.79} +04/20 [01:28:03] INFO | >> train_qwenlatent.py:487 + Step 38820 | grad_norm_pre_clip=0.1667 | + grad_norm_pre_clip_avg=0.1527 | Metrics: + {'align_loss': 0.02395959012210369, + 'recon_loss': 0.09414317458868027, + 'predict_loss': 0.004103643819689751, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16671134531497955, + 'data_time': 0.0006306699942797422, + 'model_time': 1.3782911250018515, + 'grad_norm_pre_clip_avg': 0.15270013436675073, + 'learning_rate': 3.6394581799500907e-06, + 'epoch': 9.8} +04/20 [01:28:15] INFO | >> train_qwenlatent.py:487 + Step 38830 | grad_norm_pre_clip=0.1494 | + grad_norm_pre_clip_avg=0.1728 | Metrics: + {'align_loss': 0.024378444999456406, + 'recon_loss': 0.16246499121189117, + 'predict_loss': 0.010377822443842888, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14935854077339172, + 'data_time': 0.000629387010121718, + 'model_time': 1.1566048420208972, + 'grad_norm_pre_clip_avg': 0.1727974981069565, + 'learning_rate': 3.633326041598809e-06, + 'epoch': 9.8} +04/20 [01:28:27] INFO | >> train_qwenlatent.py:487 + Step 38840 | grad_norm_pre_clip=0.2290 | + grad_norm_pre_clip_avg=0.1835 | Metrics: + {'align_loss': 0.025166530162096024, + 'recon_loss': 0.12757685780525208, + 'predict_loss': 0.011958404444158077, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2289690524339676, + 'data_time': 0.0005598779825959355, + 'model_time': 1.149442673980957, + 'grad_norm_pre_clip_avg': 0.18350764214992524, + 'learning_rate': 3.6271982308490492e-06, + 'epoch': 9.8} +04/20 [01:28:39] INFO | >> train_qwenlatent.py:487 + Step 38850 | grad_norm_pre_clip=0.2012 | + grad_norm_pre_clip_avg=0.1750 | Metrics: + {'align_loss': 0.024914110079407692, + 'recon_loss': 0.1519952118396759, + 'predict_loss': 0.009396277368068695, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2011503279209137, + 'mae_score': 0.005834884471721477, 'data_time': + 0.0005740069900639355, 'model_time': + 1.1420033309841529, 'grad_norm_pre_clip_avg': + 0.17496681362390518, 'learning_rate': + 3.621074750687433e-06, 'epoch': 9.8} +04/20 [01:29:31] INFO | >> train_qwenlatent.py:487 + Step 38860 | grad_norm_pre_clip=0.1370 | + grad_norm_pre_clip_avg=0.1630 | Metrics: + {'align_loss': 0.024744749069213867, + 'recon_loss': 0.13950297236442566, + 'predict_loss': 0.008133729919791222, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13699540495872498, + 'data_time': 0.0011601050209719688, + 'model_time': 3.38746097398689, + 'grad_norm_pre_clip_avg': 0.16295647770166397, + 'learning_rate': 3.6149556040984667e-06, + 'epoch': 9.81} +04/20 [01:30:08] INFO | >> train_qwenlatent.py:487 + Step 38870 | grad_norm_pre_clip=0.1769 | + grad_norm_pre_clip_avg=0.1624 | Metrics: + {'align_loss': 0.025647856295108795, + 'recon_loss': 0.11515943706035614, + 'predict_loss': 0.008752415888011456, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17687314748764038, + 'data_time': 0.0044199880212545395, + 'model_time': 3.692251041997224, + 'grad_norm_pre_clip_avg': 0.1624230310320854, + 'learning_rate': 3.6088407940645516e-06, + 'epoch': 9.81} +04/20 [01:30:42] INFO | >> train_qwenlatent.py:487 + Step 38880 | grad_norm_pre_clip=0.1501 | + grad_norm_pre_clip_avg=0.1520 | Metrics: + {'align_loss': 0.026673777028918266, + 'recon_loss': 0.14856484532356262, + 'predict_loss': 0.00566613906994462, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15006397664546967, + 'data_time': 0.0011334030132275075, + 'model_time': 3.506194227025844, + 'grad_norm_pre_clip_avg': 0.1519797384738922, + 'learning_rate': 3.6027303235659665e-06, + 'epoch': 9.81} +04/20 [01:31:18] INFO | >> train_qwenlatent.py:487 + Step 38890 | grad_norm_pre_clip=0.1723 | + grad_norm_pre_clip_avg=0.1560 | Metrics: + {'align_loss': 0.024710197001695633, + 'recon_loss': 0.09628250449895859, + 'predict_loss': 0.0039027926977723837, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17228154838085175, + 'data_time': 0.0010979310027323663, + 'model_time': 3.2728437340119854, + 'grad_norm_pre_clip_avg': 0.15597637444734574, + 'learning_rate': 3.596624195580893e-06, + 'epoch': 9.81} +04/20 [01:31:53] INFO | >> train_qwenlatent.py:487 + Step 38900 | grad_norm_pre_clip=0.1559 | + grad_norm_pre_clip_avg=0.1686 | Metrics: + {'align_loss': 0.02563474327325821, + 'recon_loss': 0.08858883380889893, + 'predict_loss': 0.004411645699292421, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1559324860572815, + 'mae_score': 0.006493181795687289, 'data_time': + 0.0010806879727169871, 'model_time': + 3.044172052992508, 'grad_norm_pre_clip_avg': + 0.16861063092947007, 'learning_rate': + 3.5905224130853712e-06, 'epoch': 9.82} +04/20 [01:32:26] INFO | >> train_qwenlatent.py:487 + Step 38910 | grad_norm_pre_clip=0.1661 | + grad_norm_pre_clip_avg=0.1850 | Metrics: + {'align_loss': 0.024298833683133125, + 'recon_loss': 0.0880916565656662, + 'predict_loss': 0.005421369802206755, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16609308123588562, + 'data_time': 0.0011764629743993282, + 'model_time': 3.341930536000291, + 'grad_norm_pre_clip_avg': 0.18496309369802474, + 'learning_rate': 3.5844249790533385e-06, + 'epoch': 9.82} +04/20 [01:32:57] INFO | >> train_qwenlatent.py:487 + Step 38920 | grad_norm_pre_clip=0.1080 | + grad_norm_pre_clip_avg=0.1624 | Metrics: + {'align_loss': 0.024689648300409317, + 'recon_loss': 0.1009502038359642, + 'predict_loss': 0.008356367237865925, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10803168267011642, + 'data_time': 0.004288360010832548, + 'model_time': 2.8175221720011905, + 'grad_norm_pre_clip_avg': 0.16242826357483864, + 'learning_rate': 3.578331896456609e-06, + 'epoch': 9.82} +04/20 [01:33:22] INFO | >> train_qwenlatent.py:487 + Step 38930 | grad_norm_pre_clip=0.1323 | + grad_norm_pre_clip_avg=0.1533 | Metrics: + {'align_loss': 0.025520991533994675, + 'recon_loss': 0.07682045549154282, + 'predict_loss': 0.004982297774404287, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13230448961257935, + 'data_time': 0.001071570994099602, + 'model_time': 2.2552002379961777, + 'grad_norm_pre_clip_avg': 0.15327505767345428, + 'learning_rate': 3.5722431682648826e-06, + 'epoch': 9.82} +04/20 [01:33:41] INFO | >> train_qwenlatent.py:487 + Step 38940 | grad_norm_pre_clip=0.1682 | + grad_norm_pre_clip_avg=0.1780 | Metrics: + {'align_loss': 0.02573762834072113, + 'recon_loss': 0.10831043124198914, + 'predict_loss': 0.007171157281845808, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16819265484809875, + 'data_time': 0.0015322570106945932, + 'model_time': 1.6105777850025333, + 'grad_norm_pre_clip_avg': 0.17804588079452516, + 'learning_rate': 3.5661587974457297e-06, + 'epoch': 9.83} +04/20 [01:33:56] INFO | >> train_qwenlatent.py:487 + Step 38950 | grad_norm_pre_clip=0.1544 | + grad_norm_pre_clip_avg=0.1615 | Metrics: + {'align_loss': 0.025506336241960526, + 'recon_loss': 0.14449691772460938, + 'predict_loss': 0.004236268810927868, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15439879894256592, + 'mae_score': 0.008284842001425253, 'data_time': + 0.0009309360175393522, 'model_time': + 1.2124360159796197, 'grad_norm_pre_clip_avg': + 0.16153059303760528, 'learning_rate': + 3.560078786964597e-06, 'epoch': 9.83} +04/20 [01:34:08] INFO | >> train_qwenlatent.py:487 + Step 38960 | grad_norm_pre_clip=0.1536 | + grad_norm_pre_clip_avg=0.1750 | Metrics: + {'align_loss': 0.024576693773269653, + 'recon_loss': 0.12329944223165512, + 'predict_loss': 0.0049981605261564255, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15359945595264435, + 'data_time': 0.0007889150001574308, + 'model_time': 1.208575505996123, + 'grad_norm_pre_clip_avg': 0.17496655285358428, + 'learning_rate': 3.554003139784809e-06, + 'epoch': 9.83} +04/20 [01:34:21] INFO | >> train_qwenlatent.py:487 + Step 38970 | grad_norm_pre_clip=0.1510 | + grad_norm_pre_clip_avg=0.1625 | Metrics: + {'align_loss': 0.025872625410556793, + 'recon_loss': 0.13243062794208527, + 'predict_loss': 0.0105575667694211, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15100298821926117, + 'data_time': 0.0009218819905072451, + 'model_time': 1.2631851600017399, + 'grad_norm_pre_clip_avg': 0.16247546225786208, + 'learning_rate': 3.547931858867557e-06, + 'epoch': 9.83} +04/20 [01:34:33] INFO | >> train_qwenlatent.py:487 + Step 38980 | grad_norm_pre_clip=0.1951 | + grad_norm_pre_clip_avg=0.1620 | Metrics: + {'align_loss': 0.026760369539260864, + 'recon_loss': 0.16728723049163818, + 'predict_loss': 0.015239371918141842, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19511251151561737, + 'data_time': 0.0005854339979123324, + 'model_time': 1.2104029660113156, + 'grad_norm_pre_clip_avg': 0.16201765462756157, + 'learning_rate': 3.5418649471719218e-06, + 'epoch': 9.84} +04/20 [01:34:46] INFO | >> train_qwenlatent.py:487 + Step 38990 | grad_norm_pre_clip=0.1945 | + grad_norm_pre_clip_avg=0.1725 | Metrics: + {'align_loss': 0.026330821216106415, + 'recon_loss': 0.14506416022777557, + 'predict_loss': 0.010708630084991455, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1944701373577118, + 'data_time': 0.0008701890183147043, + 'model_time': 1.2771958569937851, + 'grad_norm_pre_clip_avg': 0.17248302102088928, + 'learning_rate': 3.5358024076548315e-06, + 'epoch': 9.84} +04/20 [01:34:59] INFO | >> train_qwenlatent.py:487 + Step 39000 | grad_norm_pre_clip=0.1721 | + grad_norm_pre_clip_avg=0.1579 | Metrics: + {'align_loss': 0.025721492245793343, + 'recon_loss': 0.09339836239814758, + 'predict_loss': 0.006922220345586538, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17214560508728027, + 'mae_score': 0.0062793864860191005, + 'data_time': 0.0006459639989770949, + 'model_time': 1.2498810479883105, + 'grad_norm_pre_clip_avg': 0.1578833930194378, + 'learning_rate': 3.5297442432710966e-06, + 'epoch': 9.84} +04/20 [01:35:12] INFO | >> train_qwenlatent.py:487 + Step 39010 | grad_norm_pre_clip=0.1617 | + grad_norm_pre_clip_avg=0.1681 | Metrics: + {'align_loss': 0.024909107014536858, + 'recon_loss': 0.11321622878313065, + 'predict_loss': 0.00605720654129982, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16174006462097168, + 'data_time': 0.0009037420095410198, + 'model_time': 1.2692459230020177, + 'grad_norm_pre_clip_avg': 0.16814779862761497, + 'learning_rate': 3.523690456973395e-06, + 'epoch': 9.84} +04/20 [01:35:24] INFO | >> train_qwenlatent.py:487 + Step 39020 | grad_norm_pre_clip=0.1638 | + grad_norm_pre_clip_avg=0.1637 | Metrics: + {'align_loss': 0.0245542973279953, + 'recon_loss': 0.09964092820882797, + 'predict_loss': 0.004880936816334724, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16382797062397003, + 'data_time': 0.0005942519928794354, + 'model_time': 1.2405856469995342, + 'grad_norm_pre_clip_avg': 0.1636814519762993, + 'learning_rate': 3.5176410517122623e-06, + 'epoch': 9.85} +04/20 [01:35:37] INFO | >> train_qwenlatent.py:487 + Step 39030 | grad_norm_pre_clip=0.1454 | + grad_norm_pre_clip_avg=0.1529 | Metrics: + {'align_loss': 0.025698315352201462, + 'recon_loss': 0.17029830813407898, + 'predict_loss': 0.006367130670696497, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1453608274459839, + 'data_time': 0.000952626985963434, + 'model_time': 1.2698888989980333, + 'grad_norm_pre_clip_avg': 0.1529040366411209, + 'learning_rate': 3.5115960304361146e-06, + 'epoch': 9.85} +04/20 [01:35:50] INFO | >> train_qwenlatent.py:487 + Step 39040 | grad_norm_pre_clip=0.1246 | + grad_norm_pre_clip_avg=0.1529 | Metrics: + {'align_loss': 0.026062626391649246, + 'recon_loss': 0.11183129996061325, + 'predict_loss': 0.006947638932615519, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12457575649023056, + 'data_time': 0.0006537640001624823, + 'model_time': 1.2236306609993335, + 'grad_norm_pre_clip_avg': 0.15290965214371682, + 'learning_rate': 3.5055553960912173e-06, + 'epoch': 9.85} +04/20 [01:36:03] INFO | >> train_qwenlatent.py:487 + Step 39050 | grad_norm_pre_clip=0.2077 | + grad_norm_pre_clip_avg=0.1777 | Metrics: + {'align_loss': 0.0243130624294281, + 'recon_loss': 0.13763423264026642, + 'predict_loss': 0.008765285834670067, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20771828293800354, + 'mae_score': 0.007059059057149801, 'data_time': + 0.0007249499903991818, 'model_time': + 1.2491102700005285, 'grad_norm_pre_clip_avg': + 0.17773638516664506, 'learning_rate': + 3.499519151621703e-06, 'epoch': 9.85} +04/20 [01:36:16] INFO | >> train_qwenlatent.py:487 + Step 39060 | grad_norm_pre_clip=0.2060 | + grad_norm_pre_clip_avg=0.1664 | Metrics: + {'align_loss': 0.026693670079112053, + 'recon_loss': 0.1884651631116867, + 'predict_loss': 0.010136093944311142, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20596762001514435, + 'data_time': 0.0009257570200134069, + 'model_time': 1.20321025402518, + 'grad_norm_pre_clip_avg': 0.1663986138999462, + 'learning_rate': 3.4934872999695633e-06, + 'epoch': 9.86} +04/20 [01:36:29] INFO | >> train_qwenlatent.py:487 + Step 39070 | grad_norm_pre_clip=0.1760 | + grad_norm_pre_clip_avg=0.1588 | Metrics: + {'align_loss': 0.02480710670351982, + 'recon_loss': 0.1572662889957428, + 'predict_loss': 0.00964109506458044, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17598259449005127, + 'data_time': 0.0008662579930387437, + 'model_time': 1.2142858599836472, + 'grad_norm_pre_clip_avg': 0.1587546445429325, + 'learning_rate': 3.4874598440746483e-06, + 'epoch': 9.86} +04/20 [01:36:41] INFO | >> train_qwenlatent.py:487 + Step 39080 | grad_norm_pre_clip=0.1641 | + grad_norm_pre_clip_avg=0.1623 | Metrics: + {'align_loss': 0.025763265788555145, + 'recon_loss': 0.1234695315361023, + 'predict_loss': 0.005548554938286543, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16410431265830994, + 'data_time': 0.0009217449987772852, + 'model_time': 1.2518633439904079, + 'grad_norm_pre_clip_avg': 0.162275180965662, + 'learning_rate': 3.4814367868746687e-06, + 'epoch': 9.86} +04/20 [01:36:54] INFO | >> train_qwenlatent.py:487 + Step 39090 | grad_norm_pre_clip=0.1242 | + grad_norm_pre_clip_avg=0.1560 | Metrics: + {'align_loss': 0.02558228187263012, + 'recon_loss': 0.12730850279331207, + 'predict_loss': 0.007280027028173208, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12420368194580078, + 'data_time': 0.000919005018658936, + 'model_time': 1.6648766839934979, + 'grad_norm_pre_clip_avg': 0.1560223340988159, + 'learning_rate': 3.475418131305188e-06, + 'epoch': 9.86} +04/20 [01:37:07] INFO | >> train_qwenlatent.py:487 + Step 39100 | grad_norm_pre_clip=0.1443 | + grad_norm_pre_clip_avg=0.1578 | Metrics: + {'align_loss': 0.02380060777068138, + 'recon_loss': 0.09687905013561249, + 'predict_loss': 0.004269697237759829, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1442994773387909, + 'mae_score': 0.0054049552023947775, + 'data_time': 0.0006089310045354068, + 'model_time': 1.2183828660054132, + 'grad_norm_pre_clip_avg': 0.15777694135904313, + 'learning_rate': 3.469403880299628e-06, + 'epoch': 9.87} +04/20 [01:37:19] INFO | >> train_qwenlatent.py:487 + Step 39110 | grad_norm_pre_clip=0.1643 | + grad_norm_pre_clip_avg=0.1764 | Metrics: + {'align_loss': 0.02615685947239399, + 'recon_loss': 0.13633158802986145, + 'predict_loss': 0.004341962281614542, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1643383502960205, + 'data_time': 0.0007366469944827259, + 'model_time': 1.2410628470242955, + 'grad_norm_pre_clip_avg': 0.17642784267663955, + 'learning_rate': 3.463394036789255e-06, + 'epoch': 9.87} +04/20 [01:37:32] INFO | >> train_qwenlatent.py:487 + Step 39120 | grad_norm_pre_clip=0.1970 | + grad_norm_pre_clip_avg=0.1535 | Metrics: + {'align_loss': 0.025890646502375603, + 'recon_loss': 0.12057525664567947, + 'predict_loss': 0.00511910580098629, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19698375463485718, + 'data_time': 0.001053813990438357, + 'model_time': 1.227854695986025, + 'grad_norm_pre_clip_avg': 0.15347818955779075, + 'learning_rate': 3.4573886037032017e-06, + 'epoch': 9.87} +04/20 [01:37:45] INFO | >> train_qwenlatent.py:487 + Step 39130 | grad_norm_pre_clip=0.1491 | + grad_norm_pre_clip_avg=0.1470 | Metrics: + {'align_loss': 0.024872617796063423, + 'recon_loss': 0.07969849556684494, + 'predict_loss': 0.0037969041150063276, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1491253823041916, + 'data_time': 0.000575353013118729, + 'model_time': 1.2623300990089774, + 'grad_norm_pre_clip_avg': 0.1470083363354206, + 'learning_rate': 3.45138758396844e-06, 'epoch': + 9.87} +04/20 [01:37:58] INFO | >> train_qwenlatent.py:487 + Step 39140 | grad_norm_pre_clip=0.1439 | + grad_norm_pre_clip_avg=0.1546 | Metrics: + {'align_loss': 0.025377582758665085, + 'recon_loss': 0.07261106371879578, + 'predict_loss': 0.003813695628196001, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1439211368560791, + 'data_time': 0.0007361170137301087, + 'model_time': 1.205699509009719, + 'grad_norm_pre_clip_avg': 0.15463679730892183, + 'learning_rate': 3.445390980509796e-06, + 'epoch': 9.88} +04/20 [01:38:11] INFO | >> train_qwenlatent.py:487 + Step 39150 | grad_norm_pre_clip=0.1477 | + grad_norm_pre_clip_avg=0.1690 | Metrics: + {'align_loss': 0.02492668479681015, + 'recon_loss': 0.09969445317983627, + 'predict_loss': 0.006006008945405483, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1476907879114151, + 'mae_score': 0.005560007610836544, 'data_time': + 0.0006725199928041548, 'model_time': + 1.2220126260071993, 'grad_norm_pre_clip_avg': + 0.16896221935749053, 'learning_rate': + 3.4393987962499394e-06, 'epoch': 9.88} +04/20 [01:38:23] INFO | >> train_qwenlatent.py:487 + Step 39160 | grad_norm_pre_clip=0.1367 | + grad_norm_pre_clip_avg=0.1485 | Metrics: + {'align_loss': 0.02510891668498516, + 'recon_loss': 0.13765448331832886, + 'predict_loss': 0.004982279147952795, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13666625320911407, + 'data_time': 0.001099435001378879, + 'model_time': 1.2227043420134578, + 'grad_norm_pre_clip_avg': 0.14848083034157752, + 'learning_rate': 3.433411034109388e-06, + 'epoch': 9.88} +04/20 [01:38:36] INFO | >> train_qwenlatent.py:487 + Step 39170 | grad_norm_pre_clip=0.1712 | + grad_norm_pre_clip_avg=0.1626 | Metrics: + {'align_loss': 0.02549714967608452, + 'recon_loss': 0.11970335990190506, + 'predict_loss': 0.008599660359323025, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1712094098329544, + 'data_time': 0.0005847260181326419, + 'model_time': 1.2272899440140463, + 'grad_norm_pre_clip_avg': 0.1625558093190193, + 'learning_rate': 3.4274276970065035e-06, + 'epoch': 9.88} +04/20 [01:38:49] INFO | >> train_qwenlatent.py:487 + Step 39180 | grad_norm_pre_clip=0.2122 | + grad_norm_pre_clip_avg=0.1786 | Metrics: + {'align_loss': 0.024649325758218765, + 'recon_loss': 0.07817699015140533, + 'predict_loss': 0.003703091759234667, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21216513216495514, + 'data_time': 0.0009676800109446049, + 'model_time': 1.2277148669818416, + 'grad_norm_pre_clip_avg': 0.1786008894443512, + 'learning_rate': 3.421448787857494e-06, + 'epoch': 9.89} +04/20 [01:39:01] INFO | >> train_qwenlatent.py:487 + Step 39190 | grad_norm_pre_clip=0.1416 | + grad_norm_pre_clip_avg=0.1452 | Metrics: + {'align_loss': 0.02508718892931938, + 'recon_loss': 0.14580611884593964, + 'predict_loss': 0.008720255456864834, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14158351719379425, + 'data_time': 0.000620044011157006, + 'model_time': 1.2183899249939714, + 'grad_norm_pre_clip_avg': 0.14516370743513107, + 'learning_rate': 3.4154743095764047e-06, + 'epoch': 9.89} +04/20 [01:39:15] INFO | >> train_qwenlatent.py:487 + Step 39200 | grad_norm_pre_clip=0.1595 | + grad_norm_pre_clip_avg=0.1399 | Metrics: + {'align_loss': 0.026043660938739777, + 'recon_loss': 0.1450670212507248, + 'predict_loss': 0.007279928773641586, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15949010848999023, + 'mae_score': 0.006047683148770719, 'data_time': + 0.0006699060031678528, 'model_time': + 1.2210387870145496, 'grad_norm_pre_clip_avg': + 0.13985971435904504, 'learning_rate': + 3.4095042650751257e-06, 'epoch': 9.89} +04/20 [01:39:27] INFO | >> train_qwenlatent.py:487 + Step 39210 | grad_norm_pre_clip=0.1360 | + grad_norm_pre_clip_avg=0.1445 | Metrics: + {'align_loss': 0.025078654289245605, + 'recon_loss': 0.12024896591901779, + 'predict_loss': 0.005803626496344805, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13598483800888062, + 'data_time': 0.0005910060135647655, + 'model_time': 1.207917527994141, + 'grad_norm_pre_clip_avg': 0.14446552321314812, + 'learning_rate': 3.4035386572633803e-06, + 'epoch': 9.89} +04/20 [01:39:40] INFO | >> train_qwenlatent.py:487 + Step 39220 | grad_norm_pre_clip=0.1427 | + grad_norm_pre_clip_avg=0.1303 | Metrics: + {'align_loss': 0.025143908336758614, + 'recon_loss': 0.14209072291851044, + 'predict_loss': 0.006750588305294514, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1427413374185562, + 'data_time': 0.00094067701138556, 'model_time': + 1.3051603419880848, 'grad_norm_pre_clip_avg': + 0.1302915833890438, 'learning_rate': + 3.3975774890487394e-06, 'epoch': 9.9} +04/20 [01:39:52] INFO | >> train_qwenlatent.py:487 + Step 39230 | grad_norm_pre_clip=0.1644 | + grad_norm_pre_clip_avg=0.1454 | Metrics: + {'align_loss': 0.02538159303367138, + 'recon_loss': 0.20489086210727692, + 'predict_loss': 0.007937994785606861, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16439834237098694, + 'data_time': 0.0016024259966798127, + 'model_time': 1.2406397080048919, + 'grad_norm_pre_clip_avg': 0.1454278811812401, + 'learning_rate': 3.391620763336601e-06, + 'epoch': 9.9} +04/20 [01:40:04] INFO | >> train_qwenlatent.py:487 + Step 39240 | grad_norm_pre_clip=0.1795 | + grad_norm_pre_clip_avg=0.1689 | Metrics: + {'align_loss': 0.02470511384308338, + 'recon_loss': 0.09863666445016861, + 'predict_loss': 0.00422396557405591, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17947715520858765, + 'data_time': 0.0008794780005700886, + 'model_time': 1.229150781990029, + 'grad_norm_pre_clip_avg': 0.16889572367072106, + 'learning_rate': 3.385668483030205e-06, + 'epoch': 9.9} +04/20 [01:40:17] INFO | >> train_qwenlatent.py:487 + Step 39250 | grad_norm_pre_clip=0.1244 | + grad_norm_pre_clip_avg=0.1610 | Metrics: + {'align_loss': 0.025254786014556885, + 'recon_loss': 0.19920295476913452, + 'predict_loss': 0.009481187909841537, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12444444745779037, + 'mae_score': 0.006744885659432626, 'data_time': + 0.0007515260076615959, 'model_time': + 1.242948683007853, 'grad_norm_pre_clip_avg': + 0.16104518845677376, 'learning_rate': + 3.3797206510306093e-06, 'epoch': 9.9} +04/20 [01:40:30] INFO | >> train_qwenlatent.py:487 + Step 39260 | grad_norm_pre_clip=0.1254 | + grad_norm_pre_clip_avg=0.1523 | Metrics: + {'align_loss': 0.02632291615009308, + 'recon_loss': 0.09823715686798096, + 'predict_loss': 0.00576684158295393, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12536828219890594, + 'data_time': 0.0008808290003798902, + 'model_time': 1.2691116410132963, + 'grad_norm_pre_clip_avg': 0.15227030143141745, + 'learning_rate': 3.3737772702367247e-06, + 'epoch': 9.91} +04/20 [01:40:43] INFO | >> train_qwenlatent.py:487 + Step 39270 | grad_norm_pre_clip=0.1833 | + grad_norm_pre_clip_avg=0.1831 | Metrics: + {'align_loss': 0.024703996255993843, + 'recon_loss': 0.11223891377449036, + 'predict_loss': 0.007460583932697773, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1833193153142929, + 'data_time': 0.0006869829958304763, + 'model_time': 1.2276332280016504, + 'grad_norm_pre_clip_avg': 0.18314687609672547, + 'learning_rate': 3.3678383435452805e-06, + 'epoch': 9.91} +04/20 [01:40:55] INFO | >> train_qwenlatent.py:487 + Step 39280 | grad_norm_pre_clip=0.1436 | + grad_norm_pre_clip_avg=0.1685 | Metrics: + {'align_loss': 0.02610015496611595, + 'recon_loss': 0.10864003747701645, + 'predict_loss': 0.007345935329794884, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14360515773296356, + 'data_time': 0.0006208529812283814, + 'model_time': 1.2427696009981446, + 'grad_norm_pre_clip_avg': 0.16852927207946777, + 'learning_rate': 3.3619038738508377e-06, + 'epoch': 9.91} +04/20 [01:41:08] INFO | >> train_qwenlatent.py:487 + Step 39290 | grad_norm_pre_clip=0.1466 | + grad_norm_pre_clip_avg=0.1542 | Metrics: + {'align_loss': 0.025388993322849274, + 'recon_loss': 0.17524246871471405, + 'predict_loss': 0.0113765187561512, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1465701013803482, + 'data_time': 0.0010693369840737432, + 'model_time': 1.1890223030059133, + 'grad_norm_pre_clip_avg': 0.15420450568199157, + 'learning_rate': 3.355973864045784e-06, + 'epoch': 9.91} +04/20 [01:41:21] INFO | >> train_qwenlatent.py:487 + Step 39300 | grad_norm_pre_clip=0.1458 | + grad_norm_pre_clip_avg=0.1465 | Metrics: + {'align_loss': 0.025059761479496956, + 'recon_loss': 0.10187184065580368, + 'predict_loss': 0.004438889212906361, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14582276344299316, + 'mae_score': 0.005913320747581688, 'data_time': + 0.0008248290105257183, 'model_time': + 1.6259380459960084, 'grad_norm_pre_clip_avg': + 0.14649630412459375, 'learning_rate': + 3.35004831702033e-06, 'epoch': 9.92} +04/20 [01:41:34] INFO | >> train_qwenlatent.py:487 + Step 39310 | grad_norm_pre_clip=0.1538 | + grad_norm_pre_clip_avg=0.1604 | Metrics: + {'align_loss': 0.025267764925956726, + 'recon_loss': 0.1131989136338234, + 'predict_loss': 0.007291899528354406, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15381066501140594, + 'data_time': 0.000939481018576771, + 'model_time': 1.2727951490087435, + 'grad_norm_pre_clip_avg': 0.16042299568653107, + 'learning_rate': 3.344127235662523e-06, + 'epoch': 9.92} +04/20 [01:41:46] INFO | >> train_qwenlatent.py:487 + Step 39320 | grad_norm_pre_clip=0.1737 | + grad_norm_pre_clip_avg=0.1646 | Metrics: + {'align_loss': 0.026053279638290405, + 'recon_loss': 0.14942540228366852, + 'predict_loss': 0.004909347742795944, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17373496294021606, + 'data_time': 0.0006288320000749081, + 'model_time': 1.2266467039880808, + 'grad_norm_pre_clip_avg': 0.16457521840929984, + 'learning_rate': 3.338210622858224e-06, + 'epoch': 9.92} +04/20 [01:41:59] INFO | >> train_qwenlatent.py:487 + Step 39330 | grad_norm_pre_clip=0.1984 | + grad_norm_pre_clip_avg=0.1622 | Metrics: + {'align_loss': 0.023138049989938736, + 'recon_loss': 0.10969258099794388, + 'predict_loss': 0.006349522620439529, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19839754700660706, + 'data_time': 0.0009781360276974738, + 'model_time': 1.2788807600154541, + 'grad_norm_pre_clip_avg': 0.1621831864118576, + 'learning_rate': 3.3322984814911196e-06, + 'epoch': 9.92} +04/20 [01:42:12] INFO | >> train_qwenlatent.py:487 + Step 39340 | grad_norm_pre_clip=0.1548 | + grad_norm_pre_clip_avg=0.1551 | Metrics: + {'align_loss': 0.02431975118815899, + 'recon_loss': 0.11968120187520981, + 'predict_loss': 0.006486424710601568, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15479318797588348, + 'data_time': 0.0008219230221584439, + 'model_time': 1.2015007799782325, + 'grad_norm_pre_clip_avg': 0.15514709204435348, + 'learning_rate': 3.3263908144427055e-06, + 'epoch': 9.93} +04/20 [01:42:25] INFO | >> train_qwenlatent.py:487 + Step 39350 | grad_norm_pre_clip=0.2093 | + grad_norm_pre_clip_avg=0.1551 | Metrics: + {'align_loss': 0.026635482907295227, + 'recon_loss': 0.1988058090209961, + 'predict_loss': 0.014677099883556366, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20934638381004333, + 'mae_score': 0.006349354391699438, 'data_time': + 0.0006461809971369803, 'model_time': + 1.2332486509985756, 'grad_norm_pre_clip_avg': + 0.15508793741464616, 'learning_rate': + 3.320487624592318e-06, 'epoch': 9.93} +04/20 [01:42:37] INFO | >> train_qwenlatent.py:487 + Step 39360 | grad_norm_pre_clip=0.1469 | + grad_norm_pre_clip_avg=0.1439 | Metrics: + {'align_loss': 0.025532638654112816, + 'recon_loss': 0.11457973718643188, + 'predict_loss': 0.008765488862991333, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14685094356536865, + 'data_time': 0.0009881209989544004, + 'model_time': 1.2418276460084599, + 'grad_norm_pre_clip_avg': 0.14387233182787895, + 'learning_rate': 3.314588914817097e-06, + 'epoch': 9.93} +04/20 [01:42:50] INFO | >> train_qwenlatent.py:487 + Step 39370 | grad_norm_pre_clip=0.1465 | + grad_norm_pre_clip_avg=0.1625 | Metrics: + {'align_loss': 0.024791276082396507, + 'recon_loss': 0.11737849563360214, + 'predict_loss': 0.006406075786799192, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14647601544857025, + 'data_time': 0.0009401519782841206, + 'model_time': 1.1968225730233826, + 'grad_norm_pre_clip_avg': 0.16250869631767273, + 'learning_rate': 3.3086946879920006e-06, + 'epoch': 9.93} +04/20 [01:43:02] INFO | >> train_qwenlatent.py:487 + Step 39380 | grad_norm_pre_clip=0.1643 | + grad_norm_pre_clip_avg=0.1729 | Metrics: + {'align_loss': 0.02658654749393463, + 'recon_loss': 0.14454472064971924, + 'predict_loss': 0.007704174146056175, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16432985663414001, + 'data_time': 0.0006797509850002825, + 'model_time': 1.2258065679925494, + 'grad_norm_pre_clip_avg': 0.1728513903915882, + 'learning_rate': 3.302804946989805e-06, + 'epoch': 9.94} +04/20 [01:43:14] INFO | >> train_qwenlatent.py:487 + Step 39390 | grad_norm_pre_clip=0.1254 | + grad_norm_pre_clip_avg=0.1712 | Metrics: + {'align_loss': 0.025022126734256744, + 'recon_loss': 0.1353558748960495, + 'predict_loss': 0.003754394594579935, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12535248696804047, + 'data_time': 0.0009153300197795033, + 'model_time': 1.2349396329955198, + 'grad_norm_pre_clip_avg': 0.17120348811149597, + 'learning_rate': 3.296919694681097e-06, + 'epoch': 9.94} +04/20 [01:43:28] INFO | >> train_qwenlatent.py:487 + Step 39400 | grad_norm_pre_clip=0.1409 | + grad_norm_pre_clip_avg=0.1471 | Metrics: + {'align_loss': 0.025777989998459816, + 'recon_loss': 0.15930074453353882, + 'predict_loss': 0.005241073668003082, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1409197598695755, + 'mae_score': 0.006445391543276675, 'data_time': + 0.000878719991305843, 'model_time': + 1.2090525590174366, 'grad_norm_pre_clip_avg': + 0.147067654132843, 'learning_rate': + 3.291038933934275e-06, 'epoch': 9.94} +04/20 [01:43:41] INFO | >> train_qwenlatent.py:487 + Step 39410 | grad_norm_pre_clip=0.1639 | + grad_norm_pre_clip_avg=0.1577 | Metrics: + {'align_loss': 0.02524549886584282, + 'recon_loss': 0.12643496692180634, + 'predict_loss': 0.005160425323992968, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16394919157028198, + 'data_time': 0.0009273489995393902, + 'model_time': 1.288909783004783, + 'grad_norm_pre_clip_avg': 0.15771579295396804, + 'learning_rate': 3.285162667615556e-06, + 'epoch': 9.94} +04/20 [01:43:53] INFO | >> train_qwenlatent.py:487 + Step 39420 | grad_norm_pre_clip=0.1536 | + grad_norm_pre_clip_avg=0.1640 | Metrics: + {'align_loss': 0.026010476052761078, + 'recon_loss': 0.1331343650817871, + 'predict_loss': 0.003528427565470338, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1535559743642807, + 'data_time': 0.001327147998381406, + 'model_time': 1.2701015420025215, + 'grad_norm_pre_clip_avg': 0.1639692045748234, + 'learning_rate': 3.279290898588962e-06, + 'epoch': 9.95} +04/20 [01:44:06] INFO | >> train_qwenlatent.py:487 + Step 39430 | grad_norm_pre_clip=0.1701 | + grad_norm_pre_clip_avg=0.1728 | Metrics: + {'align_loss': 0.023575518280267715, + 'recon_loss': 0.1179920956492424, + 'predict_loss': 0.003651742823421955, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1700851023197174, + 'data_time': 0.0014337520115077496, + 'model_time': 1.263987640995765, + 'grad_norm_pre_clip_avg': 0.17279642969369888, + 'learning_rate': 3.273423629716313e-06, + 'epoch': 9.95} +04/20 [01:44:19] INFO | >> train_qwenlatent.py:487 + Step 39440 | grad_norm_pre_clip=0.1146 | + grad_norm_pre_clip_avg=0.1604 | Metrics: + {'align_loss': 0.024707429111003876, + 'recon_loss': 0.13350246846675873, + 'predict_loss': 0.008873709477484226, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11456210166215897, + 'data_time': 0.000974898983258754, + 'model_time': 1.4996740619826596, + 'grad_norm_pre_clip_avg': 0.1603516511619091, + 'learning_rate': 3.2675608638572455e-06, + 'epoch': 9.95} +04/20 [01:44:32] INFO | >> train_qwenlatent.py:487 + Step 39450 | grad_norm_pre_clip=0.1125 | + grad_norm_pre_clip_avg=0.1437 | Metrics: + {'align_loss': 0.025864455848932266, + 'recon_loss': 0.10589225590229034, + 'predict_loss': 0.006155252922326326, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11253038793802261, + 'mae_score': 0.005777398100844375, 'data_time': + 0.001038568007061258, 'model_time': + 1.2414559129974805, 'grad_norm_pre_clip_avg': + 0.1437375672161579, 'learning_rate': + 3.2617026038692094e-06, 'epoch': 9.95} +04/20 [01:44:45] INFO | >> train_qwenlatent.py:487 + Step 39460 | grad_norm_pre_clip=0.1831 | + grad_norm_pre_clip_avg=0.1681 | Metrics: + {'align_loss': 0.025131650269031525, + 'recon_loss': 0.11631058901548386, + 'predict_loss': 0.006169985514134169, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18314716219902039, + 'data_time': 0.0006652229931205511, + 'model_time': 1.2382348999963142, + 'grad_norm_pre_clip_avg': 0.1681221418082714, + 'learning_rate': 3.2558488526074425e-06, + 'epoch': 9.96} +04/20 [01:44:57] INFO | >> train_qwenlatent.py:487 + Step 39470 | grad_norm_pre_clip=0.1498 | + grad_norm_pre_clip_avg=0.1790 | Metrics: + {'align_loss': 0.02519964426755905, + 'recon_loss': 0.12342534959316254, + 'predict_loss': 0.006031102500855923, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14984869956970215, + 'data_time': 0.0010109859867952764, + 'model_time': 1.2094006139959674, + 'grad_norm_pre_clip_avg': 0.17895085737109184, + 'learning_rate': 3.2499996129249956e-06, + 'epoch': 9.96} +04/20 [01:45:10] INFO | >> train_qwenlatent.py:487 + Step 39480 | grad_norm_pre_clip=0.1707 | + grad_norm_pre_clip_avg=0.1734 | Metrics: + {'align_loss': 0.024978525936603546, + 'recon_loss': 0.15892358124256134, + 'predict_loss': 0.00573401665315032, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.170712411403656, + 'data_time': 0.000624263018835336, + 'model_time': 1.2206407590128947, + 'grad_norm_pre_clip_avg': 0.17340086922049522, + 'learning_rate': 3.244154887672713e-06, + 'epoch': 9.96} +04/20 [01:45:22] INFO | >> train_qwenlatent.py:487 + Step 39490 | grad_norm_pre_clip=0.1373 | + grad_norm_pre_clip_avg=0.1510 | Metrics: + {'align_loss': 0.025977756828069687, + 'recon_loss': 0.09260647743940353, + 'predict_loss': 0.0034244158305227757, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13728530704975128, + 'data_time': 0.001117633015383035, + 'model_time': 1.2480258069990668, + 'grad_norm_pre_clip_avg': 0.15097079202532768, + 'learning_rate': 3.2383146796992435e-06, + 'epoch': 9.96} +04/20 [01:45:36] INFO | >> train_qwenlatent.py:487 + Step 39500 | grad_norm_pre_clip=0.1267 | + grad_norm_pre_clip_avg=0.1477 | Metrics: + {'align_loss': 0.024870747700333595, + 'recon_loss': 0.08870386332273483, + 'predict_loss': 0.0040612188167870045, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12667274475097656, + 'mae_score': 0.005762146614693307, 'data_time': + 0.0006812780047766864, 'model_time': + 1.2396621789957862, 'grad_norm_pre_clip_avg': + 0.14772053956985473, 'learning_rate': + 3.232478991851038e-06, 'epoch': 9.97} +04/20 [01:45:48] INFO | >> train_qwenlatent.py:487 + Step 39510 | grad_norm_pre_clip=0.1048 | + grad_norm_pre_clip_avg=0.1503 | Metrics: + {'align_loss': 0.02504190430045128, + 'recon_loss': 0.09383594244718552, + 'predict_loss': 0.007312327157706022, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10475024580955505, + 'data_time': 0.0009287680150009692, + 'model_time': 1.2413553269871045, + 'grad_norm_pre_clip_avg': 0.15029869601130486, + 'learning_rate': 3.2266478269723407e-06, + 'epoch': 9.97} +04/20 [01:46:01] INFO | >> train_qwenlatent.py:487 + Step 39520 | grad_norm_pre_clip=0.1282 | + grad_norm_pre_clip_avg=0.1374 | Metrics: + {'align_loss': 0.025878379121422768, + 'recon_loss': 0.15953543782234192, + 'predict_loss': 0.007239471655339003, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12817786633968353, + 'data_time': 0.0008326740062329918, + 'model_time': 1.2268741209991276, + 'grad_norm_pre_clip_avg': 0.13741313740611077, + 'learning_rate': 3.2208211879051826e-06, + 'epoch': 9.97} +04/20 [01:46:13] INFO | >> train_qwenlatent.py:487 + Step 39530 | grad_norm_pre_clip=0.1340 | + grad_norm_pre_clip_avg=0.1630 | Metrics: + {'align_loss': 0.02497732639312744, + 'recon_loss': 0.1342606097459793, + 'predict_loss': 0.0087691405788064, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1339649111032486, + 'data_time': 0.0008858609944581985, + 'model_time': 1.1946258379903156, + 'grad_norm_pre_clip_avg': 0.16299233734607696, + 'learning_rate': 3.214999077489401e-06, + 'epoch': 9.97} +04/20 [01:46:26] INFO | >> train_qwenlatent.py:487 + Step 39540 | grad_norm_pre_clip=0.1693 | + grad_norm_pre_clip_avg=0.1559 | Metrics: + {'align_loss': 0.025596145540475845, + 'recon_loss': 0.09845610707998276, + 'predict_loss': 0.006251062732189894, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1693478375673294, + 'data_time': 0.0006207760015968233, + 'model_time': 1.2105948149983305, + 'grad_norm_pre_clip_avg': 0.15586215555667876, + 'learning_rate': 3.209181498562619e-06, + 'epoch': 9.98} +04/20 [01:46:39] INFO | >> train_qwenlatent.py:487 + Step 39550 | grad_norm_pre_clip=0.1361 | + grad_norm_pre_clip_avg=0.1538 | Metrics: + {'align_loss': 0.026551641523838043, + 'recon_loss': 0.11775276809930801, + 'predict_loss': 0.003675624029710889, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13608601689338684, + 'mae_score': 0.006999871107909056, 'data_time': + 0.0007852250128053129, 'model_time': + 1.2397585920116398, 'grad_norm_pre_clip_avg': + 0.15376610010862352, 'learning_rate': + 3.203368453960259e-06, 'epoch': 9.98} +04/20 [01:46:52] INFO | >> train_qwenlatent.py:487 + Step 39560 | grad_norm_pre_clip=0.1241 | + grad_norm_pre_clip_avg=0.1444 | Metrics: + {'align_loss': 0.02552037313580513, + 'recon_loss': 0.14540521800518036, + 'predict_loss': 0.005968948360532522, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12412214279174805, + 'data_time': 0.000783780007623136, + 'model_time': 1.589399449992925, + 'grad_norm_pre_clip_avg': 0.14436397403478624, + 'learning_rate': 3.1975599465155266e-06, + 'epoch': 9.98} +04/20 [01:47:04] INFO | >> train_qwenlatent.py:487 + Step 39570 | grad_norm_pre_clip=0.1418 | + grad_norm_pre_clip_avg=0.1497 | Metrics: + {'align_loss': 0.02494467794895172, + 'recon_loss': 0.11970946192741394, + 'predict_loss': 0.006075128447264433, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14178888499736786, + 'data_time': 0.000637528020888567, + 'model_time': 1.2207031350117177, + 'grad_norm_pre_clip_avg': 0.14972107782959937, + 'learning_rate': 3.1917559790594166e-06, + 'epoch': 9.98} +04/20 [01:47:17] INFO | >> train_qwenlatent.py:487 + Step 39580 | grad_norm_pre_clip=0.1681 | + grad_norm_pre_clip_avg=0.1679 | Metrics: + {'align_loss': 0.025808971375226974, + 'recon_loss': 0.16721993684768677, + 'predict_loss': 0.0072580305859446526, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.168117493391037, + 'data_time': 0.0006833149818703532, + 'model_time': 1.2256330290110782, + 'grad_norm_pre_clip_avg': 0.16789693236351014, + 'learning_rate': 3.185956554420714e-06, + 'epoch': 9.99} +04/20 [01:47:30] INFO | >> train_qwenlatent.py:487 + Step 39590 | grad_norm_pre_clip=0.2037 | + grad_norm_pre_clip_avg=0.1695 | Metrics: + {'align_loss': 0.024286707863211632, + 'recon_loss': 0.171792671084404, + 'predict_loss': 0.009861396625638008, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20372436940670013, + 'data_time': 0.0006865799950901419, + 'model_time': 1.227770967001561, + 'grad_norm_pre_clip_avg': 0.1694962725043297, + 'learning_rate': 3.1801616754259805e-06, + 'epoch': 9.99} +04/20 [01:47:43] INFO | >> train_qwenlatent.py:487 + Step 39600 | grad_norm_pre_clip=0.1483 | + grad_norm_pre_clip_avg=0.1712 | Metrics: + {'align_loss': 0.026149578392505646, + 'recon_loss': 0.08999984711408615, + 'predict_loss': 0.004259771201759577, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1483389288187027, + 'mae_score': 0.006775041528650232, 'data_time': + 0.0009730999881867319, 'model_time': + 1.2375794220133685, 'grad_norm_pre_clip_avg': + 0.17120930850505828, 'learning_rate': + 3.174371344899585e-06, 'epoch': 9.99} +04/20 [01:47:56] INFO | >> train_qwenlatent.py:487 + Step 39610 | grad_norm_pre_clip=0.1564 | + grad_norm_pre_clip_avg=0.1897 | Metrics: + {'align_loss': 0.0251499954611063, + 'recon_loss': 0.11473198235034943, + 'predict_loss': 0.007968412712216377, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1564338505268097, + 'data_time': 0.0010049850097857416, + 'model_time': 1.2795392379921395, + 'grad_norm_pre_clip_avg': 0.18972889930009842, + 'learning_rate': 3.1685855656636517e-06, + 'epoch': 9.99} +04/20 [01:48:08] INFO | >> train_qwenlatent.py:487 + Step 39620 | grad_norm_pre_clip=0.1867 | + grad_norm_pre_clip_avg=0.1696 | Metrics: + {'align_loss': 0.02577100694179535, + 'recon_loss': 0.11595559865236282, + 'predict_loss': 0.003847094252705574, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18673866987228394, + 'data_time': 0.0008837700006552041, + 'model_time': 1.2382660350122023, + 'grad_norm_pre_clip_avg': 0.16964233815670013, + 'learning_rate': 3.1628043405381023e-06, + 'epoch': 10.0} +04/20 [01:48:21] INFO | >> train_qwenlatent.py:487 + Step 39630 | grad_norm_pre_clip=0.1485 | + grad_norm_pre_clip_avg=0.1454 | Metrics: + {'align_loss': 0.026085784658789635, + 'recon_loss': 0.1454201191663742, + 'predict_loss': 0.00578016834333539, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14853374660015106, + 'data_time': 0.0008626480121165514, + 'model_time': 1.259577349992469, + 'grad_norm_pre_clip_avg': 0.14541793912649154, + 'learning_rate': 3.1570276723406318e-06, + 'epoch': 10.0} +04/20 [01:48:34] INFO | >> train_qwenlatent.py:487 + Step 39640 | grad_norm_pre_clip=0.1452 | + grad_norm_pre_clip_avg=0.1516 | Metrics: + {'align_loss': 0.02540414035320282, + 'recon_loss': 0.0626663789153099, + 'predict_loss': 0.0032893812749534845, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14523747563362122, + 'data_time': 0.0010584609990473837, + 'model_time': 1.2037329669983592, + 'grad_norm_pre_clip_avg': 0.15161122530698776, + 'learning_rate': 3.1512555638867262e-06, + 'epoch': 10.0} +04/20 [01:48:47] INFO | >> train_qwenlatent.py:487 + Step 39650 | grad_norm_pre_clip=0.2140 | + grad_norm_pre_clip_avg=0.1676 | Metrics: + {'align_loss': 0.026060372591018677, + 'recon_loss': 0.17678286135196686, + 'predict_loss': 0.0067886426113545895, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21399104595184326, + 'mae_score': 0.006839260324701533, 'data_time': + 0.0010362899920437485, 'model_time': + 1.263463915005559, 'grad_norm_pre_clip_avg': + 0.16760918125510216, 'learning_rate': + 3.1454880179896363e-06, 'epoch': 10.01} +04/20 [01:49:00] INFO | >> train_qwenlatent.py:487 + Step 39660 | grad_norm_pre_clip=0.1828 | + grad_norm_pre_clip_avg=0.1808 | Metrics: + {'align_loss': 0.026952968910336494, + 'recon_loss': 0.1874288022518158, + 'predict_loss': 0.011128781363368034, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18279746174812317, + 'data_time': 0.000862282991874963, + 'model_time': 1.2551448089943733, + 'grad_norm_pre_clip_avg': 0.18077529072761536, + 'learning_rate': 3.1397250374603953e-06, + 'epoch': 10.01} +04/20 [01:49:12] INFO | >> train_qwenlatent.py:487 + Step 39670 | grad_norm_pre_clip=0.1315 | + grad_norm_pre_clip_avg=0.1701 | Metrics: + {'align_loss': 0.0247529037296772, + 'recon_loss': 0.10039170831441879, + 'predict_loss': 0.006008004769682884, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1314641386270523, + 'data_time': 0.0006466969789471477, + 'model_time': 1.1913325770001393, + 'grad_norm_pre_clip_avg': 0.17006482928991318, + 'learning_rate': 3.13396662510781e-06, 'epoch': + 10.01} +04/20 [01:49:25] INFO | >> train_qwenlatent.py:487 + Step 39680 | grad_norm_pre_clip=0.1572 | + grad_norm_pre_clip_avg=0.1518 | Metrics: + {'align_loss': 0.024900933727622032, + 'recon_loss': 0.11965442448854446, + 'predict_loss': 0.0070353117771446705, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.157208651304245, + 'data_time': 0.0008726469823159277, + 'model_time': 1.206525474000955, + 'grad_norm_pre_clip_avg': 0.15176511257886888, + 'learning_rate': 3.128212783738459e-06, + 'epoch': 10.01} +04/20 [01:49:37] INFO | >> train_qwenlatent.py:487 + Step 39690 | grad_norm_pre_clip=0.1280 | + grad_norm_pre_clip_avg=0.1717 | Metrics: + {'align_loss': 0.026715151965618134, + 'recon_loss': 0.14417052268981934, + 'predict_loss': 0.007405101787298918, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12795740365982056, + 'data_time': 0.0007931720174383372, + 'model_time': 1.2009546350163873, + 'grad_norm_pre_clip_avg': 0.17173724323511125, + 'learning_rate': 3.1224635161566956e-06, + 'epoch': 10.02} +04/20 [01:49:50] INFO | >> train_qwenlatent.py:487 + Step 39700 | grad_norm_pre_clip=0.1596 | + grad_norm_pre_clip_avg=0.1671 | Metrics: + {'align_loss': 0.02571866475045681, + 'recon_loss': 0.1677669733762741, + 'predict_loss': 0.006275137886404991, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15957891941070557, + 'mae_score': 0.006417183833079295, 'data_time': + 0.0010876830201596022, 'model_time': + 1.3313274520041887, 'grad_norm_pre_clip_avg': + 0.16708127409219742, 'learning_rate': + 3.1167188251646435e-06, 'epoch': 10.02} +04/20 [01:50:03] INFO | >> train_qwenlatent.py:487 + Step 39710 | grad_norm_pre_clip=0.1715 | + grad_norm_pre_clip_avg=0.1604 | Metrics: + {'align_loss': 0.02556098997592926, + 'recon_loss': 0.1327618807554245, + 'predict_loss': 0.006093800533562899, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17152643203735352, + 'data_time': 0.0013945770042482764, + 'model_time': 1.2333866999833845, + 'grad_norm_pre_clip_avg': 0.16035484597086908, + 'learning_rate': 3.1109787135621945e-06, + 'epoch': 10.02} +04/20 [01:50:16] INFO | >> train_qwenlatent.py:487 + Step 39720 | grad_norm_pre_clip=0.2289 | + grad_norm_pre_clip_avg=0.1545 | Metrics: + {'align_loss': 0.025358904153108597, + 'recon_loss': 0.15479767322540283, + 'predict_loss': 0.010455033741891384, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22888751327991486, + 'data_time': 0.0009045350016094744, + 'model_time': 1.222701834019972, + 'grad_norm_pre_clip_avg': 0.15448509454727172, + 'learning_rate': 3.10524318414701e-06, 'epoch': + 10.02} +04/20 [01:50:29] INFO | >> train_qwenlatent.py:487 + Step 39730 | grad_norm_pre_clip=0.1278 | + grad_norm_pre_clip_avg=0.1533 | Metrics: + {'align_loss': 0.02475658990442753, + 'recon_loss': 0.15479348599910736, + 'predict_loss': 0.006455589085817337, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12782558798789978, + 'data_time': 0.0007366040081251413, + 'model_time': 1.2461358410073444, + 'grad_norm_pre_clip_avg': 0.15333754122257232, + 'learning_rate': 3.099512239714513e-06, + 'epoch': 10.03} +04/20 [01:50:41] INFO | >> train_qwenlatent.py:487 + Step 39740 | grad_norm_pre_clip=0.1288 | + grad_norm_pre_clip_avg=0.1526 | Metrics: + {'align_loss': 0.025439318269491196, + 'recon_loss': 0.1331838071346283, + 'predict_loss': 0.007723232731223106, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12881237268447876, + 'data_time': 0.0006487100035883486, + 'model_time': 1.247735096985707, + 'grad_norm_pre_clip_avg': 0.1526393063366413, + 'learning_rate': 3.093785883057903e-06, + 'epoch': 10.03} +04/20 [01:50:54] INFO | >> train_qwenlatent.py:487 + Step 39750 | grad_norm_pre_clip=0.1241 | + grad_norm_pre_clip_avg=0.1371 | Metrics: + {'align_loss': 0.026428012177348137, + 'recon_loss': 0.12924879789352417, + 'predict_loss': 0.006048738956451416, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12411954998970032, + 'mae_score': 0.006403793730177321, 'data_time': + 0.0006530449900310487, 'model_time': + 1.2077265420230106, 'grad_norm_pre_clip_avg': + 0.13706134855747223, 'learning_rate': + 3.088064116968135e-06, 'epoch': 10.03} +04/20 [01:51:07] INFO | >> train_qwenlatent.py:487 + Step 39760 | grad_norm_pre_clip=0.1295 | + grad_norm_pre_clip_avg=0.1457 | Metrics: + {'align_loss': 0.026017507538199425, + 'recon_loss': 0.12815149128437042, + 'predict_loss': 0.003121754853054881, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12951619923114777, + 'data_time': 0.0010001780174206942, + 'model_time': 1.234805950021837, + 'grad_norm_pre_clip_avg': 0.14569951295852662, + 'learning_rate': 3.0823469442339265e-06, + 'epoch': 10.03} +04/20 [01:51:19] INFO | >> train_qwenlatent.py:487 + Step 39770 | grad_norm_pre_clip=0.1585 | + grad_norm_pre_clip_avg=0.1509 | Metrics: + {'align_loss': 0.024423208087682724, + 'recon_loss': 0.10935094952583313, + 'predict_loss': 0.006200537085533142, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15851177275180817, + 'data_time': 0.0008957459940575063, + 'model_time': 1.2293827630055603, + 'grad_norm_pre_clip_avg': 0.15088807940483093, + 'learning_rate': 3.076634367641758e-06, + 'epoch': 10.04} +04/20 [01:51:32] INFO | >> train_qwenlatent.py:487 + Step 39780 | grad_norm_pre_clip=0.1532 | + grad_norm_pre_clip_avg=0.1585 | Metrics: + {'align_loss': 0.025048838928341866, + 'recon_loss': 0.09750247746706009, + 'predict_loss': 0.005731985438615084, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1532021462917328, + 'data_time': 0.0009196089813485742, + 'model_time': 1.2514119799889158, + 'grad_norm_pre_clip_avg': 0.15845615267753602, + 'learning_rate': 3.0709263899758713e-06, + 'epoch': 10.04} +04/20 [01:51:45] INFO | >> train_qwenlatent.py:487 + Step 39790 | grad_norm_pre_clip=0.1661 | + grad_norm_pre_clip_avg=0.1630 | Metrics: + {'align_loss': 0.024993453174829483, + 'recon_loss': 0.12860478460788727, + 'predict_loss': 0.01136730331927538, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16608648002147675, + 'data_time': 0.0009413489897269756, + 'model_time': 1.204788967006607, + 'grad_norm_pre_clip_avg': 0.16303362250328063, + 'learning_rate': 3.065223014018264e-06, + 'epoch': 10.04} +04/20 [01:51:58] INFO | >> train_qwenlatent.py:487 + Step 39800 | grad_norm_pre_clip=0.1919 | + grad_norm_pre_clip_avg=0.1565 | Metrics: + {'align_loss': 0.025684531778097153, + 'recon_loss': 0.10522327572107315, + 'predict_loss': 0.005212598480284214, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.191925510764122, + 'mae_score': 0.00664101978680035, 'data_time': + 0.0012594829895533621, 'model_time': + 1.2451256949862, 'grad_norm_pre_clip_avg': + 0.15645409226417542, 'learning_rate': + 3.0595242425486933e-06, 'epoch': 10.04} +04/20 [01:52:10] INFO | >> train_qwenlatent.py:487 + Step 39810 | grad_norm_pre_clip=0.1311 | + grad_norm_pre_clip_avg=0.1460 | Metrics: + {'align_loss': 0.02439582720398903, + 'recon_loss': 0.09806710481643677, + 'predict_loss': 0.0034093742724508047, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1311301589012146, + 'data_time': 0.0006715889903716743, + 'model_time': 1.297373584995512, + 'grad_norm_pre_clip_avg': 0.14601350352168083, + 'learning_rate': 3.0538300783446703e-06, + 'epoch': 10.05} +04/20 [01:52:23] INFO | >> train_qwenlatent.py:487 + Step 39820 | grad_norm_pre_clip=0.1899 | + grad_norm_pre_clip_avg=0.1736 | Metrics: + {'align_loss': 0.025279276072978973, + 'recon_loss': 0.11926012486219406, + 'predict_loss': 0.00962592102587223, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1899401843547821, + 'data_time': 0.0008196400012820959, + 'model_time': 1.224504521000199, + 'grad_norm_pre_clip_avg': 0.1735764726996422, + 'learning_rate': 3.048140524181459e-06, + 'epoch': 10.05} +04/20 [01:52:35] INFO | >> train_qwenlatent.py:487 + Step 39830 | grad_norm_pre_clip=0.1189 | + grad_norm_pre_clip_avg=0.1624 | Metrics: + {'align_loss': 0.025443922728300095, + 'recon_loss': 0.137048602104187, + 'predict_loss': 0.007725755218416452, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11888343840837479, + 'data_time': 0.0006516959983855486, + 'model_time': 1.2245629859971814, + 'grad_norm_pre_clip_avg': 0.16237470954656602, + 'learning_rate': 3.0424555828320855e-06, + 'epoch': 10.05} +04/20 [01:52:48] INFO | >> train_qwenlatent.py:487 + Step 39840 | grad_norm_pre_clip=0.1516 | + grad_norm_pre_clip_avg=0.1678 | Metrics: + {'align_loss': 0.025645386427640915, + 'recon_loss': 0.17168961465358734, + 'predict_loss': 0.004948109854012728, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15156716108322144, + 'data_time': 0.000868520000949502, + 'model_time': 1.549508235999383, + 'grad_norm_pre_clip_avg': 0.16784119307994844, + 'learning_rate': 3.036775257067317e-06, + 'epoch': 10.05} +04/20 [01:53:02] INFO | >> train_qwenlatent.py:487 + Step 39850 | grad_norm_pre_clip=0.1796 | + grad_norm_pre_clip_avg=0.1578 | Metrics: + {'align_loss': 0.02558518573641777, + 'recon_loss': 0.1223546490073204, + 'predict_loss': 0.0045637162402272224, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17957083880901337, + 'mae_score': 0.005781760946050421, 'data_time': + 0.0008882200054358691, 'model_time': + 1.499856879003346, 'grad_norm_pre_clip_avg': + 0.1577541187405586, 'learning_rate': + 3.0310995496556764e-06, 'epoch': 10.06} +04/20 [01:53:15] INFO | >> train_qwenlatent.py:487 + Step 39860 | grad_norm_pre_clip=0.1769 | + grad_norm_pre_clip_avg=0.1643 | Metrics: + {'align_loss': 0.026059770956635475, + 'recon_loss': 0.10088375955820084, + 'predict_loss': 0.004436061717569828, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17689663171768188, + 'data_time': 0.0006840889982413501, + 'model_time': 1.2480967179872096, + 'grad_norm_pre_clip_avg': 0.16427400782704354, + 'learning_rate': 3.025428463363437e-06, + 'epoch': 10.06} +04/20 [01:53:28] INFO | >> train_qwenlatent.py:487 + Step 39870 | grad_norm_pre_clip=0.1423 | + grad_norm_pre_clip_avg=0.1502 | Metrics: + {'align_loss': 0.02650020644068718, + 'recon_loss': 0.10731585323810577, + 'predict_loss': 0.0029315431602299213, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14229418337345123, + 'data_time': 0.0009148459939751774, + 'model_time': 1.2479039859899785, + 'grad_norm_pre_clip_avg': 0.150216706097126, + 'learning_rate': 3.0197620009546075e-06, + 'epoch': 10.06} +04/20 [01:53:40] INFO | >> train_qwenlatent.py:487 + Step 39880 | grad_norm_pre_clip=0.1468 | + grad_norm_pre_clip_avg=0.1552 | Metrics: + {'align_loss': 0.025640925392508507, + 'recon_loss': 0.1347285956144333, + 'predict_loss': 0.004894304554909468, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14683708548545837, + 'data_time': 0.0009057499992195517, + 'model_time': 1.290496504982002, + 'grad_norm_pre_clip_avg': 0.1551773354411125, + 'learning_rate': 3.014100165190962e-06, + 'epoch': 10.06} +04/20 [01:53:52] INFO | >> train_qwenlatent.py:487 + Step 39890 | grad_norm_pre_clip=0.1706 | + grad_norm_pre_clip_avg=0.1751 | Metrics: + {'align_loss': 0.02439860627055168, + 'recon_loss': 0.12554700672626495, + 'predict_loss': 0.007818740792572498, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1705676019191742, + 'data_time': 0.0006774049834348261, + 'model_time': 1.2175568229868077, + 'grad_norm_pre_clip_avg': 0.1751073867082596, + 'learning_rate': 3.00844295883201e-06, 'epoch': + 10.07} +04/20 [01:54:06] INFO | >> train_qwenlatent.py:487 + Step 39900 | grad_norm_pre_clip=0.2226 | + grad_norm_pre_clip_avg=0.1579 | Metrics: + {'align_loss': 0.02605631947517395, + 'recon_loss': 0.1426578015089035, + 'predict_loss': 0.006619453430175781, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22257696092128754, + 'mae_score': 0.006011212838662637, 'data_time': + 0.0006399480043910444, 'model_time': + 1.2130488519906066, 'grad_norm_pre_clip_avg': + 0.15785102397203446, 'learning_rate': + 3.002790384635002e-06, 'epoch': 10.07} +04/20 [01:54:18] INFO | >> train_qwenlatent.py:487 + Step 39910 | grad_norm_pre_clip=0.1507 | + grad_norm_pre_clip_avg=0.1533 | Metrics: + {'align_loss': 0.02386031672358513, + 'recon_loss': 0.1500520259141922, + 'predict_loss': 0.010276279412209988, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15074720978736877, + 'data_time': 0.00065650898613967, 'model_time': + 1.2668896559916902, 'grad_norm_pre_clip_avg': + 0.15332428961992264, 'learning_rate': + 2.9971424453549354e-06, 'epoch': 10.07} +04/20 [01:54:31] INFO | >> train_qwenlatent.py:487 + Step 39920 | grad_norm_pre_clip=0.1894 | + grad_norm_pre_clip_avg=0.1553 | Metrics: + {'align_loss': 0.025592651218175888, + 'recon_loss': 0.16347625851631165, + 'predict_loss': 0.009280814789235592, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18936258554458618, + 'data_time': 0.0006545709911733866, + 'model_time': 1.2658884080010466, + 'grad_norm_pre_clip_avg': 0.15529159381985663, + 'learning_rate': 2.9914991437445433e-06, + 'epoch': 10.07} +04/20 [01:54:43] INFO | >> train_qwenlatent.py:487 + Step 39930 | grad_norm_pre_clip=0.1669 | + grad_norm_pre_clip_avg=0.1826 | Metrics: + {'align_loss': 0.02466052584350109, + 'recon_loss': 0.1446627378463745, + 'predict_loss': 0.011761059053242207, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16693195700645447, + 'data_time': 0.0006923659821040928, + 'model_time': 1.2217516420059837, + 'grad_norm_pre_clip_avg': 0.18264164924621581, + 'learning_rate': 2.985860482554311e-06, + 'epoch': 10.08} +04/20 [01:54:56] INFO | >> train_qwenlatent.py:487 + Step 39940 | grad_norm_pre_clip=0.1205 | + grad_norm_pre_clip_avg=0.1729 | Metrics: + {'align_loss': 0.02687297761440277, + 'recon_loss': 0.1341773271560669, + 'predict_loss': 0.00626798439770937, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1205037534236908, + 'data_time': 0.0008522039861418307, + 'model_time': 1.2666282480058726, + 'grad_norm_pre_clip_avg': 0.17290771156549453, + 'learning_rate': 2.9802264645324476e-06, + 'epoch': 10.08} +04/20 [01:55:09] INFO | >> train_qwenlatent.py:487 + Step 39950 | grad_norm_pre_clip=0.1818 | + grad_norm_pre_clip_avg=0.1714 | Metrics: + {'align_loss': 0.0253328625112772, + 'recon_loss': 0.1560514122247696, + 'predict_loss': 0.010191441513597965, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18177233636379242, + 'mae_score': 0.005211304759120082, 'data_time': + 0.0006583099893759936, 'model_time': + 1.2204842070059385, 'grad_norm_pre_clip_avg': + 0.17138621509075164, 'learning_rate': + 2.9745970924249093e-06, 'epoch': 10.08} +04/20 [01:55:21] INFO | >> train_qwenlatent.py:487 + Step 39960 | grad_norm_pre_clip=0.1550 | + grad_norm_pre_clip_avg=0.1480 | Metrics: + {'align_loss': 0.024604277685284615, + 'recon_loss': 0.1414121687412262, + 'predict_loss': 0.007878717966377735, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15498554706573486, + 'data_time': 0.0006641130021307617, + 'model_time': 1.1811489899992011, + 'grad_norm_pre_clip_avg': 0.1479644440114498, + 'learning_rate': 2.9689723689753737e-06, + 'epoch': 10.08} +04/20 [01:55:34] INFO | >> train_qwenlatent.py:487 + Step 39970 | grad_norm_pre_clip=0.1491 | + grad_norm_pre_clip_avg=0.1469 | Metrics: + {'align_loss': 0.025546425953507423, + 'recon_loss': 0.16150908172130585, + 'predict_loss': 0.01231001503765583, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1490674614906311, + 'data_time': 0.0009938130096998066, + 'model_time': 1.200272029003827, + 'grad_norm_pre_clip_avg': 0.14690488949418068, + 'learning_rate': 2.963352296925272e-06, + 'epoch': 10.09} +04/20 [01:55:47] INFO | >> train_qwenlatent.py:487 + Step 39980 | grad_norm_pre_clip=0.1578 | + grad_norm_pre_clip_avg=0.1605 | Metrics: + {'align_loss': 0.02652071788907051, + 'recon_loss': 0.18579834699630737, + 'predict_loss': 0.00858079269528389, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15782791376113892, + 'data_time': 0.0008485429862048477, + 'model_time': 1.526127830002224, + 'grad_norm_pre_clip_avg': 0.160522697865963, + 'learning_rate': 2.9577368790137576e-06, + 'epoch': 10.09} +04/20 [01:56:00] INFO | >> train_qwenlatent.py:487 + Step 39990 | grad_norm_pre_clip=0.1699 | + grad_norm_pre_clip_avg=0.1472 | Metrics: + {'align_loss': 0.02492852322757244, + 'recon_loss': 0.14674904942512512, + 'predict_loss': 0.004802821669727564, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16987866163253784, + 'data_time': 0.0005836499913129956, + 'model_time': 1.2003499709826428, + 'grad_norm_pre_clip_avg': 0.1472453735768795, + 'learning_rate': 2.952126117977717e-06, + 'epoch': 10.09} +04/20 [01:56:13] INFO | >> train_qwenlatent.py:487 + Step 40000 | grad_norm_pre_clip=0.1912 | + grad_norm_pre_clip_avg=0.1614 | Metrics: + {'align_loss': 0.025961924344301224, + 'recon_loss': 0.1295773684978485, + 'predict_loss': 0.005506944376975298, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19123859703540802, + 'mae_score': 0.006429973808494774, 'data_time': + 0.0012307249999139458, 'model_time': + 1.1820527870149817, 'grad_norm_pre_clip_avg': + 0.1613884761929512, 'learning_rate': + 2.9465200165517635e-06, 'epoch': 10.09} +✅ Checkpoint saved at ./runs/0418_QwenLatent_13tasks_actionstate_30k/checkpoints/steps_40000 +04/20 [01:56:35] INFO | >> train_qwenlatent.py:487 + Step 40010 | grad_norm_pre_clip=0.1700 | + grad_norm_pre_clip_avg=0.1515 | Metrics: + {'align_loss': 0.025756116956472397, + 'recon_loss': 0.1252957433462143, + 'predict_loss': 0.008580788038671017, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16998478770256042, + 'data_time': 0.0006681749946437776, + 'model_time': 1.2545835679920856, + 'grad_norm_pre_clip_avg': 0.15146251171827316, + 'learning_rate': 2.9409185774682453e-06, + 'epoch': 10.1} +04/20 [01:56:48] INFO | >> train_qwenlatent.py:487 + Step 40020 | grad_norm_pre_clip=0.1296 | + grad_norm_pre_clip_avg=0.1504 | Metrics: + {'align_loss': 0.025826269760727882, + 'recon_loss': 0.13324743509292603, + 'predict_loss': 0.008349041454494, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12958002090454102, + 'data_time': 0.0008284089853987098, + 'model_time': 1.2708650660060812, + 'grad_norm_pre_clip_avg': 0.15035319030284883, + 'learning_rate': 2.935321803457233e-06, + 'epoch': 10.1} +04/20 [01:57:01] INFO | >> train_qwenlatent.py:487 + Step 40030 | grad_norm_pre_clip=0.1670 | + grad_norm_pre_clip_avg=0.1711 | Metrics: + {'align_loss': 0.023539265617728233, + 'recon_loss': 0.09593678265810013, + 'predict_loss': 0.005751741584390402, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16698616743087769, + 'data_time': 0.0008950979972723871, + 'model_time': 1.3248351459915284, + 'grad_norm_pre_clip_avg': 0.17107396945357323, + 'learning_rate': 2.929729697246532e-06, + 'epoch': 10.1} +04/20 [01:57:14] INFO | >> train_qwenlatent.py:487 + Step 40040 | grad_norm_pre_clip=0.2015 | + grad_norm_pre_clip_avg=0.1855 | Metrics: + {'align_loss': 0.02573205903172493, + 'recon_loss': 0.15572941303253174, + 'predict_loss': 0.012049433775246143, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20151370763778687, + 'data_time': 0.0007675119850318879, + 'model_time': 1.2402549750113394, + 'grad_norm_pre_clip_avg': 0.18545797169208528, + 'learning_rate': 2.9241422615616675e-06, + 'epoch': 10.1} +04/20 [01:57:28] INFO | >> train_qwenlatent.py:487 + Step 40050 | grad_norm_pre_clip=0.1480 | + grad_norm_pre_clip_avg=0.1702 | Metrics: + {'align_loss': 0.025946196168661118, + 'recon_loss': 0.13124340772628784, + 'predict_loss': 0.005188314244151115, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14799527823925018, + 'mae_score': 0.005681947759679846, 'data_time': + 0.0006896359845995903, 'model_time': + 1.2572156479873229, 'grad_norm_pre_clip_avg': + 0.17016117572784423, 'learning_rate': + 2.9185594991258797e-06, 'epoch': 10.11} +04/20 [01:57:41] INFO | >> train_qwenlatent.py:487 + Step 40060 | grad_norm_pre_clip=0.1671 | + grad_norm_pre_clip_avg=0.1562 | Metrics: + {'align_loss': 0.026335326954722404, + 'recon_loss': 0.1703776717185974, + 'predict_loss': 0.012670543976128101, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16706483066082, + 'data_time': 0.0006861870060674846, + 'model_time': 1.2365955300047062, + 'grad_norm_pre_clip_avg': 0.1562382884323597, + 'learning_rate': 2.9129814126601404e-06, + 'epoch': 10.11} +04/20 [01:57:54] INFO | >> train_qwenlatent.py:487 + Step 40070 | grad_norm_pre_clip=0.1903 | + grad_norm_pre_clip_avg=0.1581 | Metrics: + {'align_loss': 0.024580001831054688, + 'recon_loss': 0.11913356184959412, + 'predict_loss': 0.006950011011213064, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19030696153640747, + 'data_time': 0.0006543650233652443, + 'model_time': 1.2671056899998803, + 'grad_norm_pre_clip_avg': 0.15814296752214432, + 'learning_rate': 2.9074080048831484e-06, + 'epoch': 10.11} +04/20 [01:58:07] INFO | >> train_qwenlatent.py:487 + Step 40080 | grad_norm_pre_clip=0.2063 | + grad_norm_pre_clip_avg=0.1627 | Metrics: + {'align_loss': 0.025466544553637505, + 'recon_loss': 0.13391490280628204, + 'predict_loss': 0.014252894558012486, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20628002285957336, + 'data_time': 0.0008959250117186457, + 'model_time': 1.3591556100000162, + 'grad_norm_pre_clip_avg': 0.1626704141497612, + 'learning_rate': 2.9018392785113105e-06, + 'epoch': 10.11} +04/20 [01:58:20] INFO | >> train_qwenlatent.py:487 + Step 40090 | grad_norm_pre_clip=0.1608 | + grad_norm_pre_clip_avg=0.1405 | Metrics: + {'align_loss': 0.02562003582715988, + 'recon_loss': 0.11448653042316437, + 'predict_loss': 0.003614649875089526, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16082847118377686, + 'data_time': 0.0010259200062137097, + 'model_time': 1.2974746169929858, + 'grad_norm_pre_clip_avg': 0.1404644764959812, + 'learning_rate': 2.8962752362587564e-06, + 'epoch': 10.12} +04/20 [01:58:34] INFO | >> train_qwenlatent.py:487 + Step 40100 | grad_norm_pre_clip=0.1038 | + grad_norm_pre_clip_avg=0.1360 | Metrics: + {'align_loss': 0.023593880236148834, + 'recon_loss': 0.14634616672992706, + 'predict_loss': 0.007890835404396057, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10376608371734619, + 'mae_score': 0.005702183697674726, 'data_time': + 0.000677317992085591, 'model_time': + 1.348379744013073, 'grad_norm_pre_clip_avg': + 0.1359974928200245, 'learning_rate': + 2.8907158808373333e-06, 'epoch': 10.12} +04/20 [01:58:47] INFO | >> train_qwenlatent.py:487 + Step 40110 | grad_norm_pre_clip=0.1388 | + grad_norm_pre_clip_avg=0.1519 | Metrics: + {'align_loss': 0.02660283073782921, + 'recon_loss': 0.1616339087486267, + 'predict_loss': 0.005597634706646204, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13881412148475647, + 'data_time': 0.0006675469921901822, + 'model_time': 1.2345332219847478, + 'grad_norm_pre_clip_avg': 0.15185298398137093, + 'learning_rate': 2.8851612149565997e-06, + 'epoch': 10.12} +04/20 [01:59:01] INFO | >> train_qwenlatent.py:487 + Step 40120 | grad_norm_pre_clip=0.1672 | + grad_norm_pre_clip_avg=0.1480 | Metrics: + {'align_loss': 0.025702886283397675, + 'recon_loss': 0.11086641252040863, + 'predict_loss': 0.006331171374768019, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16715270280838013, + 'data_time': 0.0008654440171085298, + 'model_time': 1.3300068419775926, + 'grad_norm_pre_clip_avg': 0.14803327918052672, + 'learning_rate': 2.879611241323839e-06, + 'epoch': 10.12} +04/20 [01:59:14] INFO | >> train_qwenlatent.py:487 + Step 40130 | grad_norm_pre_clip=0.1966 | + grad_norm_pre_clip_avg=0.1570 | Metrics: + {'align_loss': 0.02679450437426567, + 'recon_loss': 0.14693517982959747, + 'predict_loss': 0.008176811970770359, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19659799337387085, + 'data_time': 0.0010346919880248606, + 'model_time': 1.246394005982438, + 'grad_norm_pre_clip_avg': 0.15703103542327881, + 'learning_rate': 2.8740659626440408e-06, + 'epoch': 10.13} +04/20 [01:59:26] INFO | >> train_qwenlatent.py:487 + Step 40140 | grad_norm_pre_clip=0.1599 | + grad_norm_pre_clip_avg=0.1771 | Metrics: + {'align_loss': 0.024807505309581757, + 'recon_loss': 0.15309421718120575, + 'predict_loss': 0.010619138367474079, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15987883508205414, + 'data_time': 0.0006115169962868094, + 'model_time': 1.1960099519928917, + 'grad_norm_pre_clip_avg': 0.17711451947689055, + 'learning_rate': 2.868525381619901e-06, + 'epoch': 10.13} +04/20 [01:59:39] INFO | >> train_qwenlatent.py:487 + Step 40150 | grad_norm_pre_clip=0.1398 | + grad_norm_pre_clip_avg=0.1632 | Metrics: + {'align_loss': 0.025762993842363358, + 'recon_loss': 0.14685626327991486, + 'predict_loss': 0.006861464120447636, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13977035880088806, + 'mae_score': 0.0049902361792487065, + 'data_time': 0.0006247349956538528, + 'model_time': 1.2612366800021846, + 'grad_norm_pre_clip_avg': 0.1631621316075325, + 'learning_rate': 2.862989500951829e-06, + 'epoch': 10.13} +04/20 [01:59:52] INFO | >> train_qwenlatent.py:487 + Step 40160 | grad_norm_pre_clip=0.1679 | + grad_norm_pre_clip_avg=0.1439 | Metrics: + {'align_loss': 0.026197325438261032, + 'recon_loss': 0.10740819573402405, + 'predict_loss': 0.005337754264473915, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1678932160139084, + 'data_time': 0.001158168975962326, + 'model_time': 1.2762114179786295, + 'grad_norm_pre_clip_avg': 0.14386807456612588, + 'learning_rate': 2.857458323337954e-06, + 'epoch': 10.13} +04/20 [02:00:04] INFO | >> train_qwenlatent.py:487 + Step 40170 | grad_norm_pre_clip=0.1590 | + grad_norm_pre_clip_avg=0.1426 | Metrics: + {'align_loss': 0.025212014093995094, + 'recon_loss': 0.1360982358455658, + 'predict_loss': 0.008396662771701813, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15902715921401978, + 'data_time': 0.0010103020176757127, + 'model_time': 1.2356505510106217, + 'grad_norm_pre_clip_avg': 0.14261457175016404, + 'learning_rate': 2.8519318514741023e-06, + 'epoch': 10.14} +04/20 [02:00:17] INFO | >> train_qwenlatent.py:487 + Step 40180 | grad_norm_pre_clip=0.1273 | + grad_norm_pre_clip_avg=0.1478 | Metrics: + {'align_loss': 0.02417735382914543, + 'recon_loss': 0.09513343870639801, + 'predict_loss': 0.004173905123025179, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1273498237133026, + 'data_time': 0.0007850360125303268, + 'model_time': 1.2628035660018213, + 'grad_norm_pre_clip_avg': 0.1477612428367138, + 'learning_rate': 2.8464100880538083e-06, + 'epoch': 10.14} +04/20 [02:00:30] INFO | >> train_qwenlatent.py:487 + Step 40190 | grad_norm_pre_clip=0.1726 | + grad_norm_pre_clip_avg=0.1759 | Metrics: + {'align_loss': 0.022670703008770943, + 'recon_loss': 0.12558552622795105, + 'predict_loss': 0.005237027537077665, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17261046171188354, + 'data_time': 0.0008536399982403964, + 'model_time': 1.27044909200049, + 'grad_norm_pre_clip_avg': 0.17594237476587296, + 'learning_rate': 2.840893035768313e-06, + 'epoch': 10.14} +04/20 [02:00:43] INFO | >> train_qwenlatent.py:487 + Step 40200 | grad_norm_pre_clip=0.1228 | + grad_norm_pre_clip_avg=0.1551 | Metrics: + {'align_loss': 0.025378655642271042, + 'recon_loss': 0.18628865480422974, + 'predict_loss': 0.008583754301071167, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12275022268295288, + 'mae_score': 0.005606944281775672, 'data_time': + 0.0006892500096000731, 'model_time': + 1.2735953369992785, 'grad_norm_pre_clip_avg': + 0.1551491364836693, 'learning_rate': + 2.8353806973065583e-06, 'epoch': 10.14} +04/20 [02:00:55] INFO | >> train_qwenlatent.py:487 + Step 40210 | grad_norm_pre_clip=0.1808 | + grad_norm_pre_clip_avg=0.1587 | Metrics: + {'align_loss': 0.02539743483066559, + 'recon_loss': 0.10503052175045013, + 'predict_loss': 0.006342577748000622, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18077784776687622, + 'data_time': 0.0009940290183294564, + 'model_time': 1.2285624119977, + 'grad_norm_pre_clip_avg': 0.15866346657276154, + 'learning_rate': 2.8298730753551915e-06, + 'epoch': 10.15} +04/20 [02:01:08] INFO | >> train_qwenlatent.py:487 + Step 40220 | grad_norm_pre_clip=0.1207 | + grad_norm_pre_clip_avg=0.1579 | Metrics: + {'align_loss': 0.02544497884809971, + 'recon_loss': 0.14924973249435425, + 'predict_loss': 0.006165896542370319, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12069488316774368, + 'data_time': 0.0006559239991474897, + 'model_time': 1.1776891310000792, + 'grad_norm_pre_clip_avg': 0.1578867368400097, + 'learning_rate': 2.8243701725985667e-06, + 'epoch': 10.15} +04/20 [02:01:20] INFO | >> train_qwenlatent.py:487 + Step 40230 | grad_norm_pre_clip=0.1829 | + grad_norm_pre_clip_avg=0.1567 | Metrics: + {'align_loss': 0.022970974445343018, + 'recon_loss': 0.09718161076307297, + 'predict_loss': 0.005442358087748289, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18290555477142334, + 'data_time': 0.0006541289913002402, + 'model_time': 1.2440022119844798, + 'grad_norm_pre_clip_avg': 0.1566903844475746, + 'learning_rate': 2.8188719917187244e-06, + 'epoch': 10.15} +04/20 [02:01:33] INFO | >> train_qwenlatent.py:487 + Step 40240 | grad_norm_pre_clip=0.1293 | + grad_norm_pre_clip_avg=0.1559 | Metrics: + {'align_loss': 0.025445662438869476, + 'recon_loss': 0.13829830288887024, + 'predict_loss': 0.006399864796549082, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12930065393447876, + 'data_time': 0.0008399469952564687, + 'model_time': 1.2955065830028616, + 'grad_norm_pre_clip_avg': 0.15586332678794862, + 'learning_rate': 2.813378535395414e-06, + 'epoch': 10.15} +04/20 [02:01:47] INFO | >> train_qwenlatent.py:487 + Step 40250 | grad_norm_pre_clip=0.2010 | + grad_norm_pre_clip_avg=0.1523 | Metrics: + {'align_loss': 0.023850027471780777, + 'recon_loss': 0.11810564249753952, + 'predict_loss': 0.00586096104234457, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20101335644721985, + 'mae_score': 0.005442865045221002, 'data_time': + 0.000993886002106592, 'model_time': + 1.3052848680235911, 'grad_norm_pre_clip_avg': + 0.15230511277914047, 'learning_rate': + 2.807889806306076e-06, 'epoch': 10.16} +04/20 [02:01:59] INFO | >> train_qwenlatent.py:487 + Step 40260 | grad_norm_pre_clip=0.1318 | + grad_norm_pre_clip_avg=0.1537 | Metrics: + {'align_loss': 0.024846963584423065, + 'recon_loss': 0.11898870021104813, + 'predict_loss': 0.0035666492767632008, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13177187740802765, + 'data_time': 0.0006716579955536872, + 'model_time': 1.2415854099963326, + 'grad_norm_pre_clip_avg': 0.1536703735589981, + 'learning_rate': 2.802405807125856e-06, + 'epoch': 10.16} +04/20 [02:02:12] INFO | >> train_qwenlatent.py:487 + Step 40270 | grad_norm_pre_clip=0.1717 | + grad_norm_pre_clip_avg=0.1823 | Metrics: + {'align_loss': 0.02453196793794632, + 'recon_loss': 0.11722404509782791, + 'predict_loss': 0.00796847976744175, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1716931164264679, + 'data_time': 0.0008571849903091788, + 'model_time': 1.2034708989958744, + 'grad_norm_pre_clip_avg': 0.18234967291355134, + 'learning_rate': 2.796926540527587e-06, + 'epoch': 10.16} +04/20 [02:02:25] INFO | >> train_qwenlatent.py:487 + Step 40280 | grad_norm_pre_clip=0.1379 | + grad_norm_pre_clip_avg=0.1503 | Metrics: + {'align_loss': 0.025547590106725693, + 'recon_loss': 0.12237777560949326, + 'predict_loss': 0.009062524884939194, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13787509500980377, + 'data_time': 0.0011166010226588696, + 'model_time': 1.2308962359966245, + 'grad_norm_pre_clip_avg': 0.15031092539429664, + 'learning_rate': 2.7914520091817963e-06, + 'epoch': 10.16} +04/20 [02:02:37] INFO | >> train_qwenlatent.py:487 + Step 40290 | grad_norm_pre_clip=0.1349 | + grad_norm_pre_clip_avg=0.1410 | Metrics: + {'align_loss': 0.025397872552275658, + 'recon_loss': 0.1077728271484375, + 'predict_loss': 0.0030150925740599632, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1348545253276825, + 'data_time': 0.0009414539963472635, + 'model_time': 1.1744028379907832, + 'grad_norm_pre_clip_avg': 0.1410162292420864, + 'learning_rate': 2.7859822157567024e-06, + 'epoch': 10.17} +04/20 [02:02:50] INFO | >> train_qwenlatent.py:487 + Step 40300 | grad_norm_pre_clip=0.1048 | + grad_norm_pre_clip_avg=0.1574 | Metrics: + {'align_loss': 0.02625664696097374, + 'recon_loss': 0.11289852112531662, + 'predict_loss': 0.004735584836453199, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10484618693590164, + 'mae_score': 0.006102781897192603, 'data_time': + 0.0011807740083895624, 'model_time': + 1.208655207999982, 'grad_norm_pre_clip_avg': + 0.15743646025657654, 'learning_rate': + 2.7805171629182147e-06, 'epoch': 10.17} +04/20 [02:03:03] INFO | >> train_qwenlatent.py:487 + Step 40310 | grad_norm_pre_clip=0.1453 | + grad_norm_pre_clip_avg=0.1508 | Metrics: + {'align_loss': 0.02570611611008644, + 'recon_loss': 0.11876627802848816, + 'predict_loss': 0.009903010912239552, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14529462158679962, + 'data_time': 0.000829356984468177, + 'model_time': 1.237769160012249, + 'grad_norm_pre_clip_avg': 0.15080603510141372, + 'learning_rate': 2.7750568533299435e-06, + 'epoch': 10.17} +04/20 [02:03:15] INFO | >> train_qwenlatent.py:487 + Step 40320 | grad_norm_pre_clip=0.1571 | + grad_norm_pre_clip_avg=0.1385 | Metrics: + {'align_loss': 0.026253757998347282, + 'recon_loss': 0.1459461748600006, + 'predict_loss': 0.00741172581911087, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15712614357471466, + 'data_time': 0.0007337979914154857, + 'model_time': 1.2595131819834933, + 'grad_norm_pre_clip_avg': 0.1384843833744526, + 'learning_rate': 2.769601289653167e-06, + 'epoch': 10.17} +04/20 [02:03:28] INFO | >> train_qwenlatent.py:487 + Step 40330 | grad_norm_pre_clip=0.2255 | + grad_norm_pre_clip_avg=0.1477 | Metrics: + {'align_loss': 0.0251353420317173, + 'recon_loss': 0.11355593800544739, + 'predict_loss': 0.005621722899377346, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22551888227462769, + 'data_time': 0.0006843089940957725, + 'model_time': 1.2670672690146603, + 'grad_norm_pre_clip_avg': 0.1477346234023571, + 'learning_rate': 2.764150474546862e-06, + 'epoch': 10.18} +04/20 [02:03:41] INFO | >> train_qwenlatent.py:487 + Step 40340 | grad_norm_pre_clip=0.1386 | + grad_norm_pre_clip_avg=0.1642 | Metrics: + {'align_loss': 0.025808528065681458, + 'recon_loss': 0.15098834037780762, + 'predict_loss': 0.008657115511596203, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13860110938549042, + 'data_time': 0.0006841150170657784, + 'model_time': 1.2510734149836935, + 'grad_norm_pre_clip_avg': 0.16421577632427214, + 'learning_rate': 2.75870441066769e-06, 'epoch': + 10.18} +04/20 [02:03:54] INFO | >> train_qwenlatent.py:487 + Step 40350 | grad_norm_pre_clip=0.1540 | + grad_norm_pre_clip_avg=0.1745 | Metrics: + {'align_loss': 0.02510397508740425, + 'recon_loss': 0.13135114312171936, + 'predict_loss': 0.004708378110080957, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15402719378471375, + 'mae_score': 0.0052924057384869, 'data_time': + 0.0006951779942028224, 'model_time': + 1.2516414780111518, 'grad_norm_pre_clip_avg': + 0.17446020543575286, 'learning_rate': + 2.7532631006699996e-06, 'epoch': 10.18} +04/20 [02:04:06] INFO | >> train_qwenlatent.py:487 + Step 40360 | grad_norm_pre_clip=0.1632 | + grad_norm_pre_clip_avg=0.1784 | Metrics: + {'align_loss': 0.025054825469851494, + 'recon_loss': 0.09423055499792099, + 'predict_loss': 0.004360491409897804, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1631988286972046, + 'data_time': 0.0007163830159697682, + 'model_time': 1.2301463219919242, + 'grad_norm_pre_clip_avg': 0.17844192534685135, + 'learning_rate': 2.7478265472058193e-06, + 'epoch': 10.18} +04/20 [02:04:18] INFO | >> train_qwenlatent.py:487 + Step 40370 | grad_norm_pre_clip=0.1532 | + grad_norm_pre_clip_avg=0.1581 | Metrics: + {'align_loss': 0.025340940803289413, + 'recon_loss': 0.11240632086992264, + 'predict_loss': 0.00612240144982934, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1532367318868637, + 'data_time': 0.0009454819955863059, + 'model_time': 1.1978838220238686, + 'grad_norm_pre_clip_avg': 0.1581234723329544, + 'learning_rate': 2.7423947529248564e-06, + 'epoch': 10.19} +04/20 [02:04:31] INFO | >> train_qwenlatent.py:487 + Step 40380 | grad_norm_pre_clip=0.1682 | + grad_norm_pre_clip_avg=0.1624 | Metrics: + {'align_loss': 0.02582877315580845, + 'recon_loss': 0.11533186584711075, + 'predict_loss': 0.0046671368181705475, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16822879016399384, + 'data_time': 0.0007549939909949899, + 'model_time': 1.2455062820226885, + 'grad_norm_pre_clip_avg': 0.16237171441316606, + 'learning_rate': 2.7369677204745015e-06, + 'epoch': 10.19} +04/20 [02:04:44] INFO | >> train_qwenlatent.py:487 + Step 40390 | grad_norm_pre_clip=0.1880 | + grad_norm_pre_clip_avg=0.1561 | Metrics: + {'align_loss': 0.026532303541898727, + 'recon_loss': 0.20718321204185486, + 'predict_loss': 0.010141069069504738, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18796414136886597, + 'data_time': 0.0009081820026040077, + 'model_time': 1.1726805249927565, + 'grad_norm_pre_clip_avg': 0.15606274083256721, + 'learning_rate': 2.7315454524998262e-06, + 'epoch': 10.19} +04/20 [02:04:57] INFO | >> train_qwenlatent.py:487 + Step 40400 | grad_norm_pre_clip=0.1620 | + grad_norm_pre_clip_avg=0.1708 | Metrics: + {'align_loss': 0.026571357622742653, + 'recon_loss': 0.16917890310287476, + 'predict_loss': 0.008897384628653526, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16201622784137726, + 'mae_score': 0.006696202733495214, 'data_time': + 0.001240382989635691, 'model_time': + 1.22441761099617, 'grad_norm_pre_clip_avg': + 0.17075629085302352, 'learning_rate': + 2.726127951643577e-06, 'epoch': 10.19} +04/20 [02:05:10] INFO | >> train_qwenlatent.py:487 + Step 40410 | grad_norm_pre_clip=0.1286 | + grad_norm_pre_clip_avg=0.1481 | Metrics: + {'align_loss': 0.026316992938518524, + 'recon_loss': 0.15390600264072418, + 'predict_loss': 0.005321654956787825, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12864279747009277, + 'data_time': 0.0007954610045999289, + 'model_time': 1.5700454949983396, + 'grad_norm_pre_clip_avg': 0.1481145516037941, + 'learning_rate': 2.7207152205461776e-06, + 'epoch': 10.2} +04/20 [02:05:23] INFO | >> train_qwenlatent.py:487 + Step 40420 | grad_norm_pre_clip=0.1977 | + grad_norm_pre_clip_avg=0.1717 | Metrics: + {'align_loss': 0.025908825919032097, + 'recon_loss': 0.1257053166627884, + 'predict_loss': 0.009600132703781128, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19771204888820648, + 'data_time': 0.0009569230023771524, + 'model_time': 1.2549747570010368, + 'grad_norm_pre_clip_avg': 0.17171874195337294, + 'learning_rate': 2.7153072618457273e-06, + 'epoch': 10.2} +04/20 [02:05:35] INFO | >> train_qwenlatent.py:487 + Step 40430 | grad_norm_pre_clip=0.1691 | + grad_norm_pre_clip_avg=0.1613 | Metrics: + {'align_loss': 0.02642790786921978, + 'recon_loss': 0.12399642914533615, + 'predict_loss': 0.006724018603563309, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1690833419561386, + 'data_time': 0.0009377750102430582, + 'model_time': 1.225687879981706, + 'grad_norm_pre_clip_avg': 0.16133116707205772, + 'learning_rate': 2.7099040781780004e-06, + 'epoch': 10.2} +04/20 [02:05:47] INFO | >> train_qwenlatent.py:487 + Step 40440 | grad_norm_pre_clip=0.1762 | + grad_norm_pre_clip_avg=0.1766 | Metrics: + {'align_loss': 0.024225207045674324, + 'recon_loss': 0.14012515544891357, + 'predict_loss': 0.009201961569488049, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1762438267469406, + 'data_time': 0.0008398240024689585, + 'model_time': 1.3131487900100183, + 'grad_norm_pre_clip_avg': 0.17659842371940612, + 'learning_rate': 2.7045056721764403e-06, + 'epoch': 10.2} +04/20 [02:06:00] INFO | >> train_qwenlatent.py:487 + Step 40450 | grad_norm_pre_clip=0.2251 | + grad_norm_pre_clip_avg=0.1791 | Metrics: + {'align_loss': 0.02444308251142502, + 'recon_loss': 0.0932389572262764, + 'predict_loss': 0.004575752653181553, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22514304518699646, + 'mae_score': 0.006925292487617011, 'data_time': + 0.0007747320050839335, 'model_time': + 1.3185061379917897, 'grad_norm_pre_clip_avg': + 0.1790645144879818, 'learning_rate': + 2.699112046472168e-06, 'epoch': 10.21} +04/20 [02:06:13] INFO | >> train_qwenlatent.py:487 + Step 40460 | grad_norm_pre_clip=0.1914 | + grad_norm_pre_clip_avg=0.1645 | Metrics: + {'align_loss': 0.02535485103726387, + 'recon_loss': 0.14867480099201202, + 'predict_loss': 0.009004466235637665, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19142882525920868, + 'data_time': 0.000894529017386958, + 'model_time': 1.6269746120087802, + 'grad_norm_pre_clip_avg': 0.16454392299056053, + 'learning_rate': 2.6937232036939717e-06, + 'epoch': 10.21} +04/20 [02:06:26] INFO | >> train_qwenlatent.py:487 + Step 40470 | grad_norm_pre_clip=0.1593 | + grad_norm_pre_clip_avg=0.1801 | Metrics: + {'align_loss': 0.02607758343219757, + 'recon_loss': 0.20310471951961517, + 'predict_loss': 0.010614125989377499, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15927539765834808, + 'data_time': 0.000761878996854648, + 'model_time': 1.2991409500245936, + 'grad_norm_pre_clip_avg': 0.1801293410360813, + 'learning_rate': 2.6883391464683055e-06, + 'epoch': 10.21} +04/20 [02:06:39] INFO | >> train_qwenlatent.py:487 + Step 40480 | grad_norm_pre_clip=0.1193 | + grad_norm_pre_clip_avg=0.1685 | Metrics: + {'align_loss': 0.02447369694709778, + 'recon_loss': 0.10148660093545914, + 'predict_loss': 0.005036757327616215, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11926479637622833, + 'data_time': 0.0010703500011004508, + 'model_time': 1.2515296549827326, + 'grad_norm_pre_clip_avg': 0.16845955401659013, + 'learning_rate': 2.6829598774192928e-06, + 'epoch': 10.21} +04/20 [02:06:51] INFO | >> train_qwenlatent.py:487 + Step 40490 | grad_norm_pre_clip=0.1280 | + grad_norm_pre_clip_avg=0.1667 | Metrics: + {'align_loss': 0.026596050709486008, + 'recon_loss': 0.11952410638332367, + 'predict_loss': 0.0046058776788413525, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12799277901649475, + 'data_time': 0.0008519800030626357, + 'model_time': 1.2880100540060084, + 'grad_norm_pre_clip_avg': 0.1666573703289032, + 'learning_rate': 2.6775853991687273e-06, + 'epoch': 10.22} +04/20 [02:07:04] INFO | >> train_qwenlatent.py:487 + Step 40500 | grad_norm_pre_clip=0.1266 | + grad_norm_pre_clip_avg=0.1569 | Metrics: + {'align_loss': 0.02461448684334755, + 'recon_loss': 0.11852644383907318, + 'predict_loss': 0.005528735462576151, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12663422524929047, + 'mae_score': 0.005521623937933294, 'data_time': + 0.0008886230061762035, 'model_time': + 1.2140284889901523, 'grad_norm_pre_clip_avg': + 0.15687946155667304, 'learning_rate': + 2.6722157143360604e-06, 'epoch': 10.22} +04/20 [02:07:18] INFO | >> train_qwenlatent.py:487 + Step 40510 | grad_norm_pre_clip=0.1567 | + grad_norm_pre_clip_avg=0.1553 | Metrics: + {'align_loss': 0.02312205731868744, + 'recon_loss': 0.10869956761598587, + 'predict_loss': 0.005616650450974703, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15669624507427216, + 'data_time': 0.0006648529961239547, + 'model_time': 1.2107266320090275, + 'grad_norm_pre_clip_avg': 0.1553495578467846, + 'learning_rate': 2.6668508255384144e-06, + 'epoch': 10.22} +04/20 [02:07:30] INFO | >> train_qwenlatent.py:487 + Step 40520 | grad_norm_pre_clip=0.1279 | + grad_norm_pre_clip_avg=0.1533 | Metrics: + {'align_loss': 0.02523796260356903, + 'recon_loss': 0.11839982122182846, + 'predict_loss': 0.005139188840985298, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1278948038816452, + 'data_time': 0.0009329409804195166, + 'model_time': 1.2395117320120335, + 'grad_norm_pre_clip_avg': 0.15330508053302766, + 'learning_rate': 2.6614907353905692e-06, + 'epoch': 10.22} +04/20 [02:07:43] INFO | >> train_qwenlatent.py:487 + Step 40530 | grad_norm_pre_clip=0.1229 | + grad_norm_pre_clip_avg=0.1379 | Metrics: + {'align_loss': 0.02543368935585022, + 'recon_loss': 0.13952645659446716, + 'predict_loss': 0.007803840562701225, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12294243276119232, + 'data_time': 0.000698060990544036, + 'model_time': 1.28304399701301, + 'grad_norm_pre_clip_avg': 0.13786915093660354, + 'learning_rate': 2.656135446504969e-06, + 'epoch': 10.23} +04/20 [02:07:56] INFO | >> train_qwenlatent.py:487 + Step 40540 | grad_norm_pre_clip=0.2350 | + grad_norm_pre_clip_avg=0.1626 | Metrics: + {'align_loss': 0.023635856807231903, + 'recon_loss': 0.11214454472064972, + 'predict_loss': 0.009478874504566193, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23502862453460693, + 'data_time': 0.0009881559817586094, + 'model_time': 1.280385108984774, + 'grad_norm_pre_clip_avg': 0.16258983314037323, + 'learning_rate': 2.650784961491712e-06, + 'epoch': 10.23} +04/20 [02:08:09] INFO | >> train_qwenlatent.py:487 + Step 40550 | grad_norm_pre_clip=0.1805 | + grad_norm_pre_clip_avg=0.1598 | Metrics: + {'align_loss': 0.026897326111793518, + 'recon_loss': 0.19328440725803375, + 'predict_loss': 0.007899933494627476, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1804724931716919, + 'mae_score': 0.005460180677809157, 'data_time': + 0.000631396978860721, 'model_time': + 1.2054574220092036, 'grad_norm_pre_clip_avg': + 0.1598432958126068, 'learning_rate': + 2.645439282958566e-06, 'epoch': 10.23} +04/20 [02:08:22] INFO | >> train_qwenlatent.py:487 + Step 40560 | grad_norm_pre_clip=0.1595 | + grad_norm_pre_clip_avg=0.1412 | Metrics: + {'align_loss': 0.025988295674324036, + 'recon_loss': 0.11159389466047287, + 'predict_loss': 0.006591017823666334, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15947723388671875, + 'data_time': 0.0009062589961104095, + 'model_time': 1.254381820996059, + 'grad_norm_pre_clip_avg': 0.14118953943252563, + 'learning_rate': 2.640098413510947e-06, + 'epoch': 10.23} +04/20 [02:08:34] INFO | >> train_qwenlatent.py:487 + Step 40570 | grad_norm_pre_clip=0.1285 | + grad_norm_pre_clip_avg=0.1539 | Metrics: + {'align_loss': 0.024520592764019966, + 'recon_loss': 0.11618023365736008, + 'predict_loss': 0.005866688210517168, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.128525510430336, + 'data_time': 0.0006656360055785626, + 'model_time': 1.2236238610057626, + 'grad_norm_pre_clip_avg': 0.1539369985461235, + 'learning_rate': 2.6347623557519345e-06, + 'epoch': 10.24} +04/20 [02:08:47] INFO | >> train_qwenlatent.py:487 + Step 40580 | grad_norm_pre_clip=0.2013 | + grad_norm_pre_clip_avg=0.1569 | Metrics: + {'align_loss': 0.024331940338015556, + 'recon_loss': 0.09893658757209778, + 'predict_loss': 0.0038441934157162905, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20125712454319, + 'data_time': 0.0009797329839784652, + 'model_time': 1.208652847999474, + 'grad_norm_pre_clip_avg': 0.156911800801754, + 'learning_rate': 2.6294311122822466e-06, + 'epoch': 10.24} +04/20 [02:08:59] INFO | >> train_qwenlatent.py:487 + Step 40590 | grad_norm_pre_clip=0.1484 | + grad_norm_pre_clip_avg=0.1666 | Metrics: + {'align_loss': 0.024772856384515762, + 'recon_loss': 0.11337430775165558, + 'predict_loss': 0.004611275624483824, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14843806624412537, + 'data_time': 0.0011027860164176673, + 'model_time': 1.2174002209794708, + 'grad_norm_pre_clip_avg': 0.16663054376840591, + 'learning_rate': 2.624104685700277e-06, + 'epoch': 10.24} +04/20 [02:09:13] INFO | >> train_qwenlatent.py:487 + Step 40600 | grad_norm_pre_clip=0.1532 | + grad_norm_pre_clip_avg=0.1527 | Metrics: + {'align_loss': 0.0267401821911335, + 'recon_loss': 0.16268256306648254, + 'predict_loss': 0.011822836473584175, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15319731831550598, + 'mae_score': 0.005784866401741097, 'data_time': + 0.0008781299984548241, 'model_time': + 1.277396454999689, 'grad_norm_pre_clip_avg': + 0.15273250117897988, 'learning_rate': + 2.618783078602058e-06, 'epoch': 10.24} +04/20 [02:09:26] INFO | >> train_qwenlatent.py:487 + Step 40610 | grad_norm_pre_clip=0.1310 | + grad_norm_pre_clip_avg=0.1349 | Metrics: + {'align_loss': 0.026439201086759567, + 'recon_loss': 0.1387024223804474, + 'predict_loss': 0.007642369717359543, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13099302351474762, + 'data_time': 0.0006847349868621677, + 'model_time': 1.2120903589820955, + 'grad_norm_pre_clip_avg': 0.13494234532117844, + 'learning_rate': 2.6134662935812756e-06, + 'epoch': 10.25} +04/20 [02:09:38] INFO | >> train_qwenlatent.py:487 + Step 40620 | grad_norm_pre_clip=0.1250 | + grad_norm_pre_clip_avg=0.1450 | Metrics: + {'align_loss': 0.02457616850733757, + 'recon_loss': 0.10170275717973709, + 'predict_loss': 0.0050003668293356895, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12495208531618118, + 'data_time': 0.0011009709851350635, + 'model_time': 1.2412865099904593, + 'grad_norm_pre_clip_avg': 0.14504850879311562, + 'learning_rate': 2.6081543332292674e-06, + 'epoch': 10.25} +04/20 [02:09:50] INFO | >> train_qwenlatent.py:487 + Step 40630 | grad_norm_pre_clip=0.1621 | + grad_norm_pre_clip_avg=0.1732 | Metrics: + {'align_loss': 0.02683701366186142, + 'recon_loss': 0.15768179297447205, + 'predict_loss': 0.009005137719213963, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16208186745643616, + 'data_time': 0.0007844690117053688, + 'model_time': 1.2198540900135413, + 'grad_norm_pre_clip_avg': 0.17323523610830308, + 'learning_rate': 2.6028472001350145e-06, + 'epoch': 10.25} +04/20 [02:10:03] INFO | >> train_qwenlatent.py:487 + Step 40640 | grad_norm_pre_clip=0.1435 | + grad_norm_pre_clip_avg=0.1791 | Metrics: + {'align_loss': 0.025687703862786293, + 'recon_loss': 0.15920443832874298, + 'predict_loss': 0.007072067819535732, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14345963299274445, + 'data_time': 0.0008468799933325499, + 'model_time': 1.2585950870125089, + 'grad_norm_pre_clip_avg': 0.17905256301164627, + 'learning_rate': 2.5975448968851557e-06, + 'epoch': 10.25} +04/20 [02:10:16] INFO | >> train_qwenlatent.py:487 + Step 40650 | grad_norm_pre_clip=0.0959 | + grad_norm_pre_clip_avg=0.1513 | Metrics: + {'align_loss': 0.025227397680282593, + 'recon_loss': 0.1430288851261139, + 'predict_loss': 0.005491746589541435, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.0959395244717598, + 'mae_score': 0.006091501906111433, 'data_time': + 0.0006947169895283878, 'model_time': + 1.282590162998531, 'grad_norm_pre_clip_avg': + 0.15132319703698158, 'learning_rate': + 2.592247426063965e-06, 'epoch': 10.26} +04/20 [02:10:29] INFO | >> train_qwenlatent.py:487 + Step 40660 | grad_norm_pre_clip=0.1785 | + grad_norm_pre_clip_avg=0.1751 | Metrics: + {'align_loss': 0.026446474716067314, + 'recon_loss': 0.13246411085128784, + 'predict_loss': 0.006453944835811853, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17853564023971558, + 'data_time': 0.0006574919971171767, + 'model_time': 1.1890614090079907, + 'grad_norm_pre_clip_avg': 0.17506601214408873, + 'learning_rate': 2.5869547902533705e-06, + 'epoch': 10.26} +04/20 [02:10:41] INFO | >> train_qwenlatent.py:487 + Step 40670 | grad_norm_pre_clip=0.1735 | + grad_norm_pre_clip_avg=0.1510 | Metrics: + {'align_loss': 0.024912934750318527, + 'recon_loss': 0.09501486271619797, + 'predict_loss': 0.005780905019491911, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17352572083473206, + 'data_time': 0.0008744980150368065, + 'model_time': 1.254614359990228, + 'grad_norm_pre_clip_avg': 0.15100125819444657, + 'learning_rate': 2.581666992032929e-06, + 'epoch': 10.26} +04/20 [02:10:54] INFO | >> train_qwenlatent.py:487 + Step 40680 | grad_norm_pre_clip=0.1026 | + grad_norm_pre_clip_avg=0.1559 | Metrics: + {'align_loss': 0.02583257108926773, + 'recon_loss': 0.12304124236106873, + 'predict_loss': 0.008442138321697712, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.102560855448246, + 'data_time': 0.0009530950046610087, + 'model_time': 1.2417181890050415, + 'grad_norm_pre_clip_avg': 0.15594767779111862, + 'learning_rate': 2.5763840339798553e-06, + 'epoch': 10.26} +04/20 [02:11:07] INFO | >> train_qwenlatent.py:487 + Step 40690 | grad_norm_pre_clip=0.1536 | + grad_norm_pre_clip_avg=0.1423 | Metrics: + {'align_loss': 0.025491589680314064, + 'recon_loss': 0.16540877521038055, + 'predict_loss': 0.008694270625710487, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15355442464351654, + 'data_time': 0.00096757902065292, 'model_time': + 1.6719874370028265, 'grad_norm_pre_clip_avg': + 0.14230936020612717, 'learning_rate': + 2.5711059186690005e-06, 'epoch': 10.27} +04/20 [02:11:20] INFO | >> train_qwenlatent.py:487 + Step 40700 | grad_norm_pre_clip=0.2516 | + grad_norm_pre_clip_avg=0.1837 | Metrics: + {'align_loss': 0.025994759052991867, + 'recon_loss': 0.1423569917678833, + 'predict_loss': 0.00601082481443882, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.25159579515457153, + 'mae_score': 0.008374163481566284, 'data_time': + 0.0006449460051953793, 'model_time': + 1.2430049949907698, 'grad_norm_pre_clip_avg': + 0.18369353711605071, 'learning_rate': + 2.5658326486728494e-06, 'epoch': 10.27} +04/20 [02:11:32] INFO | >> train_qwenlatent.py:487 + Step 40710 | grad_norm_pre_clip=0.2064 | + grad_norm_pre_clip_avg=0.1749 | Metrics: + {'align_loss': 0.02411983534693718, + 'recon_loss': 0.09114178270101547, + 'predict_loss': 0.005493481643497944, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20643429458141327, + 'data_time': 0.001124361006077379, + 'model_time': 1.2208501340064686, + 'grad_norm_pre_clip_avg': 0.17487554252147675, + 'learning_rate': 2.5605642265615334e-06, + 'epoch': 10.27} +04/20 [02:11:45] INFO | >> train_qwenlatent.py:487 + Step 40720 | grad_norm_pre_clip=0.1552 | + grad_norm_pre_clip_avg=0.1542 | Metrics: + {'align_loss': 0.02665448747575283, + 'recon_loss': 0.1872989386320114, + 'predict_loss': 0.01238598208874464, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15520967543125153, + 'data_time': 0.0008875710191205144, + 'model_time': 1.2795144580013584, + 'grad_norm_pre_clip_avg': 0.15418558567762375, + 'learning_rate': 2.555300654902815e-06, + 'epoch': 10.28} +04/20 [02:11:57] INFO | >> train_qwenlatent.py:487 + Step 40730 | grad_norm_pre_clip=0.1826 | + grad_norm_pre_clip_avg=0.1466 | Metrics: + {'align_loss': 0.02477753721177578, + 'recon_loss': 0.15677540004253387, + 'predict_loss': 0.01133000385016203, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1825687438249588, + 'data_time': 0.000649263005470857, + 'model_time': 1.3166447779804002, + 'grad_norm_pre_clip_avg': 0.14658493027091027, + 'learning_rate': 2.5500419362620934e-06, + 'epoch': 10.28} +04/20 [02:12:10] INFO | >> train_qwenlatent.py:487 + Step 40740 | grad_norm_pre_clip=0.1761 | + grad_norm_pre_clip_avg=0.1483 | Metrics: + {'align_loss': 0.025225434452295303, + 'recon_loss': 0.12467346340417862, + 'predict_loss': 0.00779821677133441, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1760818064212799, + 'data_time': 0.0009549249953124672, + 'model_time': 1.2376091170008294, + 'grad_norm_pre_clip_avg': 0.14829473644495011, + 'learning_rate': 2.54478807320241e-06, 'epoch': + 10.28} +04/20 [02:12:23] INFO | >> train_qwenlatent.py:487 + Step 40750 | grad_norm_pre_clip=0.1520 | + grad_norm_pre_clip_avg=0.1616 | Metrics: + {'align_loss': 0.02573809027671814, + 'recon_loss': 0.15852059423923492, + 'predict_loss': 0.01115316804498434, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15202546119689941, + 'mae_score': 0.006897337801821597, 'data_time': + 0.0010520080104470253, 'model_time': + 1.2625164760102052, 'grad_norm_pre_clip_avg': + 0.1615559697151184, 'learning_rate': + 2.539539068284434e-06, 'epoch': 10.28} +04/20 [02:12:35] INFO | >> train_qwenlatent.py:487 + Step 40760 | grad_norm_pre_clip=0.1177 | + grad_norm_pre_clip_avg=0.1346 | Metrics: + {'align_loss': 0.02660498395562172, + 'recon_loss': 0.15034274756908417, + 'predict_loss': 0.009882859885692596, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11765296757221222, + 'data_time': 0.0009213229932356626, + 'model_time': 1.3094929410144687, + 'grad_norm_pre_clip_avg': 0.13458768501877785, + 'learning_rate': 2.534294924066459e-06, + 'epoch': 10.29} +04/20 [02:12:49] INFO | >> train_qwenlatent.py:487 + Step 40770 | grad_norm_pre_clip=0.1630 | + grad_norm_pre_clip_avg=0.1484 | Metrics: + {'align_loss': 0.026918476447463036, + 'recon_loss': 0.19002409279346466, + 'predict_loss': 0.006755978800356388, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1629735827445984, + 'data_time': 0.0008374500030186027, + 'model_time': 1.2351337540021632, + 'grad_norm_pre_clip_avg': 0.14842041805386544, + 'learning_rate': 2.529055643104419e-06, + 'epoch': 10.29} +04/20 [02:13:01] INFO | >> train_qwenlatent.py:487 + Step 40780 | grad_norm_pre_clip=0.1098 | + grad_norm_pre_clip_avg=0.1563 | Metrics: + {'align_loss': 0.02594701200723648, + 'recon_loss': 0.12776833772659302, + 'predict_loss': 0.006493383552879095, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10975198447704315, + 'data_time': 0.0006257049972191453, + 'model_time': 1.2575583330180962, + 'grad_norm_pre_clip_avg': 0.1563365176320076, + 'learning_rate': 2.5238212279518803e-06, + 'epoch': 10.29} +04/20 [02:13:14] INFO | >> train_qwenlatent.py:487 + Step 40790 | grad_norm_pre_clip=0.1656 | + grad_norm_pre_clip_avg=0.1578 | Metrics: + {'align_loss': 0.02532566711306572, + 'recon_loss': 0.13908785581588745, + 'predict_loss': 0.010728413239121437, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16564323008060455, + 'data_time': 0.0008041050168685615, + 'model_time': 1.2485202259849757, + 'grad_norm_pre_clip_avg': 0.15778783932328225, + 'learning_rate': 2.518591681160033e-06, + 'epoch': 10.29} +04/20 [02:13:27] INFO | >> train_qwenlatent.py:487 + Step 40800 | grad_norm_pre_clip=0.1366 | + grad_norm_pre_clip_avg=0.1365 | Metrics: + {'align_loss': 0.025933390483260155, + 'recon_loss': 0.13107697665691376, + 'predict_loss': 0.006291699130088091, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13663780689239502, + 'mae_score': 0.0050805345311895145, + 'data_time': 0.0009244639950338751, + 'model_time': 1.2789024109952152, + 'grad_norm_pre_clip_avg': 0.13651312440633773, + 'learning_rate': 2.513367005277692e-06, + 'epoch': 10.3} +04/20 [02:13:40] INFO | >> train_qwenlatent.py:487 + Step 40810 | grad_norm_pre_clip=0.0964 | + grad_norm_pre_clip_avg=0.1620 | Metrics: + {'align_loss': 0.024516213685274124, + 'recon_loss': 0.13897164165973663, + 'predict_loss': 0.007830196991562843, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.0964333564043045, + 'data_time': 0.0006864210008643568, + 'model_time': 1.295800963009242, + 'grad_norm_pre_clip_avg': 0.16202142387628554, + 'learning_rate': 2.508147202851301e-06, + 'epoch': 10.3} +04/20 [02:13:52] INFO | >> train_qwenlatent.py:487 + Step 40820 | grad_norm_pre_clip=0.2015 | + grad_norm_pre_clip_avg=0.1692 | Metrics: + {'align_loss': 0.026967819780111313, + 'recon_loss': 0.13009102642536163, + 'predict_loss': 0.004980129189789295, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20150676369667053, + 'data_time': 0.0006423309969250113, + 'model_time': 1.2083275009936187, + 'grad_norm_pre_clip_avg': 0.16920608505606652, + 'learning_rate': 2.5029322764249268e-06, + 'epoch': 10.3} +04/20 [02:14:05] INFO | >> train_qwenlatent.py:487 + Step 40830 | grad_norm_pre_clip=0.2211 | + grad_norm_pre_clip_avg=0.1519 | Metrics: + {'align_loss': 0.02585626393556595, + 'recon_loss': 0.14804627001285553, + 'predict_loss': 0.00782616063952446, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22114308178424835, + 'data_time': 0.0007816610159352422, + 'model_time': 1.2083962869946845, + 'grad_norm_pre_clip_avg': 0.1518931917846203, + 'learning_rate': 2.497722228540267e-06, + 'epoch': 10.3} +04/20 [02:14:18] INFO | >> train_qwenlatent.py:487 + Step 40840 | grad_norm_pre_clip=0.1264 | + grad_norm_pre_clip_avg=0.1600 | Metrics: + {'align_loss': 0.02598460391163826, + 'recon_loss': 0.13840000331401825, + 'predict_loss': 0.006470487918704748, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12637367844581604, + 'data_time': 0.0005922150157857686, + 'model_time': 1.190146177978022, + 'grad_norm_pre_clip_avg': 0.15996564999222757, + 'learning_rate': 2.4925170617366347e-06, + 'epoch': 10.31} +04/20 [02:14:31] INFO | >> train_qwenlatent.py:487 + Step 40850 | grad_norm_pre_clip=0.2348 | + grad_norm_pre_clip_avg=0.1852 | Metrics: + {'align_loss': 0.026059143245220184, + 'recon_loss': 0.17628756165504456, + 'predict_loss': 0.007207942195236683, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23483167588710785, + 'mae_score': 0.006002543423626874, 'data_time': + 0.0009311210014857352, 'model_time': + 1.2581120789982378, 'grad_norm_pre_clip_avg': + 0.1852380022406578, 'learning_rate': + 2.4873167785509598e-06, 'epoch': 10.31} +04/20 [02:14:43] INFO | >> train_qwenlatent.py:487 + Step 40860 | grad_norm_pre_clip=0.0897 | + grad_norm_pre_clip_avg=0.1602 | Metrics: + {'align_loss': 0.025579756125807762, + 'recon_loss': 0.14395758509635925, + 'predict_loss': 0.005474128294736147, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.08972367644309998, + 'data_time': 0.0011016530042979866, + 'model_time': 1.2362235960026737, + 'grad_norm_pre_clip_avg': 0.16018563285470008, + 'learning_rate': 2.4821213815178e-06, 'epoch': + 10.31} +04/20 [02:14:56] INFO | >> train_qwenlatent.py:487 + Step 40870 | grad_norm_pre_clip=0.1479 | + grad_norm_pre_clip_avg=0.1483 | Metrics: + {'align_loss': 0.02553742378950119, + 'recon_loss': 0.13907980918884277, + 'predict_loss': 0.007263094186782837, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.147886723279953, + 'data_time': 0.0010298489942215383, + 'model_time': 1.2206041540193837, + 'grad_norm_pre_clip_avg': 0.14826036095619202, + 'learning_rate': 2.4769308731693245e-06, + 'epoch': 10.31} +04/20 [02:15:09] INFO | >> train_qwenlatent.py:487 + Step 40880 | grad_norm_pre_clip=0.1730 | + grad_norm_pre_clip_avg=0.1642 | Metrics: + {'align_loss': 0.025591012090444565, + 'recon_loss': 0.09265432506799698, + 'predict_loss': 0.004448779858648777, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1729721575975418, + 'data_time': 0.0011742440110538155, + 'model_time': 1.2127035089943092, + 'grad_norm_pre_clip_avg': 0.16416026800870895, + 'learning_rate': 2.4717452560353313e-06, + 'epoch': 10.32} +04/20 [02:15:21] INFO | >> train_qwenlatent.py:487 + Step 40890 | grad_norm_pre_clip=0.1227 | + grad_norm_pre_clip_avg=0.1452 | Metrics: + {'align_loss': 0.024771014228463173, + 'recon_loss': 0.09807980060577393, + 'predict_loss': 0.003141007386147976, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12273181229829788, + 'data_time': 0.0006706760032102466, + 'model_time': 1.2429636690067127, + 'grad_norm_pre_clip_avg': 0.14521105587482452, + 'learning_rate': 2.466564532643223e-06, + 'epoch': 10.32} +04/20 [02:15:34] INFO | >> train_qwenlatent.py:487 + Step 40900 | grad_norm_pre_clip=0.1189 | + grad_norm_pre_clip_avg=0.1359 | Metrics: + {'align_loss': 0.025392640382051468, + 'recon_loss': 0.10532236099243164, + 'predict_loss': 0.005562844220548868, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11891977488994598, + 'mae_score': 0.007815767408491254, 'data_time': + 0.0011486210278235376, 'model_time': + 1.2830631140095647, 'grad_norm_pre_clip_avg': + 0.13585140630602838, 'learning_rate': + 2.4613887055180236e-06, 'epoch': 10.32} +04/20 [02:15:47] INFO | >> train_qwenlatent.py:487 + Step 40910 | grad_norm_pre_clip=0.1394 | + grad_norm_pre_clip_avg=0.1278 | Metrics: + {'align_loss': 0.02551804855465889, + 'recon_loss': 0.16297701001167297, + 'predict_loss': 0.008244942873716354, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13937003910541534, + 'data_time': 0.0007737749838270247, + 'model_time': 1.2217959269764833, + 'grad_norm_pre_clip_avg': 0.12775254249572754, + 'learning_rate': 2.4562177771823657e-06, + 'epoch': 10.32} +04/20 [02:15:59] INFO | >> train_qwenlatent.py:487 + Step 40920 | grad_norm_pre_clip=0.2144 | + grad_norm_pre_clip_avg=0.1561 | Metrics: + {'align_loss': 0.024843934923410416, + 'recon_loss': 0.12884759902954102, + 'predict_loss': 0.009468788281083107, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21435388922691345, + 'data_time': 0.0006726159772370011, + 'model_time': 1.318417283007875, + 'grad_norm_pre_clip_avg': 0.1561494931578636, + 'learning_rate': 2.451051750156496e-06, + 'epoch': 10.33} +04/20 [02:16:12] INFO | >> train_qwenlatent.py:487 + Step 40930 | grad_norm_pre_clip=0.1848 | + grad_norm_pre_clip_avg=0.1660 | Metrics: + {'align_loss': 0.025601597502827644, + 'recon_loss': 0.11210334300994873, + 'predict_loss': 0.0042596179991960526, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18475261330604553, + 'data_time': 0.0008728329848963767, + 'model_time': 1.198816961987177, + 'grad_norm_pre_clip_avg': 0.1660225510597229, + 'learning_rate': 2.445890626958281e-06, + 'epoch': 10.33} +04/20 [02:16:25] INFO | >> train_qwenlatent.py:487 + Step 40940 | grad_norm_pre_clip=0.2239 | + grad_norm_pre_clip_avg=0.1862 | Metrics: + {'align_loss': 0.02568371221423149, + 'recon_loss': 0.113351970911026, + 'predict_loss': 0.00476613687351346, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22387151420116425, + 'data_time': 0.001219864992890507, + 'model_time': 1.345326546987053, + 'grad_norm_pre_clip_avg': 0.1861873373389244, + 'learning_rate': 2.440734410103181e-06, + 'epoch': 10.33} +04/20 [02:16:38] INFO | >> train_qwenlatent.py:487 + Step 40950 | grad_norm_pre_clip=0.2004 | + grad_norm_pre_clip_avg=0.1918 | Metrics: + {'align_loss': 0.025337813422083855, + 'recon_loss': 0.09218879789113998, + 'predict_loss': 0.004047452937811613, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20040684938430786, + 'mae_score': 0.005707724459536441, 'data_time': + 0.0006764040153939277, 'model_time': + 1.182900626998162, 'grad_norm_pre_clip_avg': + 0.19180988147854805, 'learning_rate': + 2.4355831021042765e-06, 'epoch': 10.33} +04/20 [02:16:50] INFO | >> train_qwenlatent.py:487 + Step 40960 | grad_norm_pre_clip=0.1400 | + grad_norm_pre_clip_avg=0.1628 | Metrics: + {'align_loss': 0.023575924336910248, + 'recon_loss': 0.14580562710762024, + 'predict_loss': 0.010839072056114674, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13998161256313324, + 'data_time': 0.0007355669804383069, + 'model_time': 1.2460838129918557, + 'grad_norm_pre_clip_avg': 0.162755236774683, + 'learning_rate': 2.4304367054722477e-06, + 'epoch': 10.34} +04/20 [02:17:03] INFO | >> train_qwenlatent.py:487 + Step 40970 | grad_norm_pre_clip=0.1159 | + grad_norm_pre_clip_avg=0.1414 | Metrics: + {'align_loss': 0.0259354580193758, + 'recon_loss': 0.13153068721294403, + 'predict_loss': 0.006793012842535973, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11587817966938019, + 'data_time': 0.0011052750051021576, + 'model_time': 1.2444192269758787, + 'grad_norm_pre_clip_avg': 0.14144444912672044, + 'learning_rate': 2.425295222715395e-06, + 'epoch': 10.34} +04/20 [02:17:15] INFO | >> train_qwenlatent.py:487 + Step 40980 | grad_norm_pre_clip=0.1650 | + grad_norm_pre_clip_avg=0.1530 | Metrics: + {'align_loss': 0.025473468005657196, + 'recon_loss': 0.14715631306171417, + 'predict_loss': 0.007585932034999132, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16496434807777405, + 'data_time': 0.0010613829945214093, + 'model_time': 1.3191758539760485, + 'grad_norm_pre_clip_avg': 0.15296047180891037, + 'learning_rate': 2.420158656339609e-06, + 'epoch': 10.34} +04/20 [02:17:28] INFO | >> train_qwenlatent.py:487 + Step 40990 | grad_norm_pre_clip=0.1354 | + grad_norm_pre_clip_avg=0.1574 | Metrics: + {'align_loss': 0.026045169681310654, + 'recon_loss': 0.10786385089159012, + 'predict_loss': 0.007122018374502659, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13544827699661255, + 'data_time': 0.0007106139964889735, + 'model_time': 1.1941075020004064, + 'grad_norm_pre_clip_avg': 0.15737651735544206, + 'learning_rate': 2.415027008848392e-06, + 'epoch': 10.34} +04/20 [02:17:41] INFO | >> train_qwenlatent.py:487 + Step 41000 | grad_norm_pre_clip=0.2203 | + grad_norm_pre_clip_avg=0.1470 | Metrics: + {'align_loss': 0.02584744803607464, + 'recon_loss': 0.12155319005250931, + 'predict_loss': 0.010052140802145004, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22032180428504944, + 'mae_score': 0.004972003816484331, 'data_time': + 0.00086276198271662, 'model_time': + 1.2322320459934417, 'grad_norm_pre_clip_avg': + 0.14704550728201865, 'learning_rate': + 2.4099002827428453e-06, 'epoch': 10.35} +04/20 [02:17:54] INFO | >> train_qwenlatent.py:487 + Step 41010 | grad_norm_pre_clip=0.1123 | + grad_norm_pre_clip_avg=0.1645 | Metrics: + {'align_loss': 0.025378119200468063, + 'recon_loss': 0.1662617176771164, + 'predict_loss': 0.011024456471204758, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11230980604887009, + 'data_time': 0.001068645011400804, + 'model_time': 1.256060545973014, + 'grad_norm_pre_clip_avg': 0.16450171694159507, + 'learning_rate': 2.4047784805216725e-06, + 'epoch': 10.35} +04/20 [02:18:06] INFO | >> train_qwenlatent.py:487 + Step 41020 | grad_norm_pre_clip=0.1182 | + grad_norm_pre_clip_avg=0.1522 | Metrics: + {'align_loss': 0.02667943760752678, + 'recon_loss': 0.12925603985786438, + 'predict_loss': 0.006238727364689112, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11820268630981445, + 'data_time': 0.0008562360017094761, + 'model_time': 1.2610571820114274, + 'grad_norm_pre_clip_avg': 0.15221436321735382, + 'learning_rate': 2.3996616046811785e-06, + 'epoch': 10.35} +04/20 [02:18:19] INFO | >> train_qwenlatent.py:487 + Step 41030 | grad_norm_pre_clip=0.1192 | + grad_norm_pre_clip_avg=0.1460 | Metrics: + {'align_loss': 0.02434544265270233, + 'recon_loss': 0.13757547736167908, + 'predict_loss': 0.00564776873216033, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11918221414089203, + 'data_time': 0.0009149959951173514, + 'model_time': 1.2338671409816016, + 'grad_norm_pre_clip_avg': 0.14600307643413543, + 'learning_rate': 2.3945496577152657e-06, + 'epoch': 10.35} +04/20 [02:18:32] INFO | >> train_qwenlatent.py:487 + Step 41040 | grad_norm_pre_clip=0.1502 | + grad_norm_pre_clip_avg=0.1653 | Metrics: + {'align_loss': 0.025658048689365387, + 'recon_loss': 0.11048464477062225, + 'predict_loss': 0.005202031694352627, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15024356544017792, + 'data_time': 0.0009473030222579837, + 'model_time': 1.2339535200153477, + 'grad_norm_pre_clip_avg': 0.16525436863303183, + 'learning_rate': 2.389442642115435e-06, + 'epoch': 10.36} +04/20 [02:18:45] INFO | >> train_qwenlatent.py:487 + Step 41050 | grad_norm_pre_clip=0.1796 | + grad_norm_pre_clip_avg=0.1876 | Metrics: + {'align_loss': 0.025463756173849106, + 'recon_loss': 0.14776553213596344, + 'predict_loss': 0.005536715965718031, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17963199317455292, + 'mae_score': 0.0050185469893721845, + 'data_time': 0.0007917909824755043, + 'model_time': 1.231568093993701, + 'grad_norm_pre_clip_avg': 0.1876348912715912, + 'learning_rate': 2.384340560370786e-06, + 'epoch': 10.36} +04/20 [02:18:58] INFO | >> train_qwenlatent.py:487 + Step 41060 | grad_norm_pre_clip=0.2015 | + grad_norm_pre_clip_avg=0.1762 | Metrics: + {'align_loss': 0.025622326880693436, + 'recon_loss': 0.1505434364080429, + 'predict_loss': 0.005979531444609165, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20153433084487915, + 'data_time': 0.0007015470182523131, + 'model_time': 1.2182587980059907, + 'grad_norm_pre_clip_avg': 0.17620012313127517, + 'learning_rate': 2.379243414968006e-06, + 'epoch': 10.36} +04/20 [02:19:10] INFO | >> train_qwenlatent.py:487 + Step 41070 | grad_norm_pre_clip=0.1069 | + grad_norm_pre_clip_avg=0.1537 | Metrics: + {'align_loss': 0.025460969656705856, + 'recon_loss': 0.11822881549596786, + 'predict_loss': 0.006213609594851732, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10694722831249237, + 'data_time': 0.0006989309913478792, + 'model_time': 1.1991941670130473, + 'grad_norm_pre_clip_avg': 0.1537487842142582, + 'learning_rate': 2.374151208391387e-06, + 'epoch': 10.36} +04/20 [02:19:23] INFO | >> train_qwenlatent.py:487 + Step 41080 | grad_norm_pre_clip=0.1560 | + grad_norm_pre_clip_avg=0.1537 | Metrics: + {'align_loss': 0.02589293383061886, + 'recon_loss': 0.18541428446769714, + 'predict_loss': 0.007879771292209625, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15603961050510406, + 'data_time': 0.0007070150168146938, + 'model_time': 1.206381500000134, + 'grad_norm_pre_clip_avg': 0.1537408336997032, + 'learning_rate': 2.3690639431228084e-06, + 'epoch': 10.37} +04/20 [02:19:36] INFO | >> train_qwenlatent.py:487 + Step 41090 | grad_norm_pre_clip=0.1775 | + grad_norm_pre_clip_avg=0.1562 | Metrics: + {'align_loss': 0.025328127667307854, + 'recon_loss': 0.14653237164020538, + 'predict_loss': 0.007843133062124252, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17754752933979034, + 'data_time': 0.0010078799969051033, + 'model_time': 1.2605078910128213, + 'grad_norm_pre_clip_avg': 0.15623940974473954, + 'learning_rate': 2.363981621641739e-06, + 'epoch': 10.37} +04/20 [02:19:49] INFO | >> train_qwenlatent.py:487 + Step 41100 | grad_norm_pre_clip=0.1716 | + grad_norm_pre_clip_avg=0.1638 | Metrics: + {'align_loss': 0.024344243109226227, + 'recon_loss': 0.11043107509613037, + 'predict_loss': 0.003214606549590826, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17157694697380066, + 'mae_score': 0.005363047659934104, 'data_time': + 0.0008755529997870326, 'model_time': + 1.2159153679967858, 'grad_norm_pre_clip_avg': + 0.16380728781223297, 'learning_rate': + 2.358904246425244e-06, 'epoch': 10.37} +04/20 [02:20:02] INFO | >> train_qwenlatent.py:487 + Step 41110 | grad_norm_pre_clip=0.1471 | + grad_norm_pre_clip_avg=0.1340 | Metrics: + {'align_loss': 0.025203868746757507, + 'recon_loss': 0.14831992983818054, + 'predict_loss': 0.008767657913267612, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1470576822757721, + 'data_time': 0.0009146069933194667, + 'model_time': 1.2598847530025523, + 'grad_norm_pre_clip_avg': 0.13398676067590715, + 'learning_rate': 2.353831819947969e-06, + 'epoch': 10.37} +04/20 [02:20:14] INFO | >> train_qwenlatent.py:487 + Step 41120 | grad_norm_pre_clip=0.1332 | + grad_norm_pre_clip_avg=0.1551 | Metrics: + {'align_loss': 0.026079580187797546, + 'recon_loss': 0.14454559981822968, + 'predict_loss': 0.006888640578836203, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13315130770206451, + 'data_time': 0.0009410329803358763, + 'model_time': 1.2164084660180379, + 'grad_norm_pre_clip_avg': 0.15505675077438355, + 'learning_rate': 2.3487643446821585e-06, + 'epoch': 10.38} +04/20 [02:20:27] INFO | >> train_qwenlatent.py:487 + Step 41130 | grad_norm_pre_clip=0.1153 | + grad_norm_pre_clip_avg=0.1512 | Metrics: + {'align_loss': 0.02523868903517723, + 'recon_loss': 0.0792820006608963, + 'predict_loss': 0.006134836468845606, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11533574759960175, + 'data_time': 0.0010234210058115423, + 'model_time': 1.242154734005453, + 'grad_norm_pre_clip_avg': 0.1512175440788269, + 'learning_rate': 2.3437018230976346e-06, + 'epoch': 10.38} +04/20 [02:20:40] INFO | >> train_qwenlatent.py:487 + Step 41140 | grad_norm_pre_clip=0.1761 | + grad_norm_pre_clip_avg=0.1458 | Metrics: + {'align_loss': 0.02554934471845627, + 'recon_loss': 0.12241000682115555, + 'predict_loss': 0.005246122367680073, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17613834142684937, + 'data_time': 0.0011488479794934392, + 'model_time': 1.2764449200185481, + 'grad_norm_pre_clip_avg': 0.14578110426664354, + 'learning_rate': 2.33864425766181e-06, 'epoch': + 10.38} +04/20 [02:20:53] INFO | >> train_qwenlatent.py:487 + Step 41150 | grad_norm_pre_clip=0.0989 | + grad_norm_pre_clip_avg=0.1383 | Metrics: + {'align_loss': 0.02622072771191597, + 'recon_loss': 0.10465487092733383, + 'predict_loss': 0.0037615562323480844, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09893058985471725, + 'mae_score': 0.004828999923156188, 'data_time': + 0.0007288380002137274, 'model_time': + 1.2628323590033688, 'grad_norm_pre_clip_avg': + 0.13825664892792702, 'learning_rate': + 2.333591650839678e-06, 'epoch': 10.38} +04/20 [02:21:06] INFO | >> train_qwenlatent.py:487 + Step 41160 | grad_norm_pre_clip=0.1485 | + grad_norm_pre_clip_avg=0.1500 | Metrics: + {'align_loss': 0.02537572756409645, + 'recon_loss': 0.12197112292051315, + 'predict_loss': 0.008296254090964794, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14851774275302887, + 'data_time': 0.0009481399902142584, + 'model_time': 1.2649243969935924, + 'grad_norm_pre_clip_avg': 0.15000763535499573, + 'learning_rate': 2.3285440050938242e-06, + 'epoch': 10.39} +04/20 [02:21:18] INFO | >> train_qwenlatent.py:487 + Step 41170 | grad_norm_pre_clip=0.1662 | + grad_norm_pre_clip_avg=0.1526 | Metrics: + {'align_loss': 0.024923941120505333, + 'recon_loss': 0.2055254727602005, + 'predict_loss': 0.012335916981101036, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16615323722362518, + 'data_time': 0.0009819800034165382, + 'model_time': 1.1996983179997187, + 'grad_norm_pre_clip_avg': 0.15264693945646285, + 'learning_rate': 2.3235013228844064e-06, + 'epoch': 10.39} +04/20 [02:21:31] INFO | >> train_qwenlatent.py:487 + Step 41180 | grad_norm_pre_clip=0.1034 | + grad_norm_pre_clip_avg=0.1412 | Metrics: + {'align_loss': 0.025584105402231216, + 'recon_loss': 0.13924917578697205, + 'predict_loss': 0.006344175897538662, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10336007922887802, + 'data_time': 0.0009946149948518723, + 'model_time': 1.2347391539951786, + 'grad_norm_pre_clip_avg': 0.14119661450386048, + 'learning_rate': 2.318463606669166e-06, + 'epoch': 10.39} +04/20 [02:21:44] INFO | >> train_qwenlatent.py:487 + Step 41190 | grad_norm_pre_clip=0.1705 | + grad_norm_pre_clip_avg=0.1469 | Metrics: + {'align_loss': 0.025691043585538864, + 'recon_loss': 0.13051408529281616, + 'predict_loss': 0.005729325115680695, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1705017387866974, + 'data_time': 0.0009198600018862635, + 'model_time': 1.2953411689959466, + 'grad_norm_pre_clip_avg': 0.14693003818392752, + 'learning_rate': 2.3134308589034242e-06, + 'epoch': 10.39} +04/20 [02:21:57] INFO | >> train_qwenlatent.py:487 + Step 41200 | grad_norm_pre_clip=0.1790 | + grad_norm_pre_clip_avg=0.1601 | Metrics: + {'align_loss': 0.026001419872045517, + 'recon_loss': 0.14563560485839844, + 'predict_loss': 0.007989692501723766, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17902745306491852, + 'mae_score': 0.0065274809931849575, + 'data_time': 0.000822706992039457, + 'model_time': 1.1942886290198658, + 'grad_norm_pre_clip_avg': 0.1600565120577812, + 'learning_rate': 2.308403082040083e-06, + 'epoch': 10.4} +04/20 [02:22:10] INFO | >> train_qwenlatent.py:487 + Step 41210 | grad_norm_pre_clip=0.1155 | + grad_norm_pre_clip_avg=0.1393 | Metrics: + {'align_loss': 0.025163112208247185, + 'recon_loss': 0.11188773065805435, + 'predict_loss': 0.005199459847062826, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11553645133972168, + 'data_time': 0.0007333229877986014, + 'model_time': 1.264485308987787, + 'grad_norm_pre_clip_avg': 0.1393462561070919, + 'learning_rate': 2.3033802785296193e-06, + 'epoch': 10.4} +04/20 [02:22:23] INFO | >> train_qwenlatent.py:487 + Step 41220 | grad_norm_pre_clip=0.1303 | + grad_norm_pre_clip_avg=0.1460 | Metrics: + {'align_loss': 0.026394352316856384, + 'recon_loss': 0.19939963519573212, + 'predict_loss': 0.011869121342897415, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13032282888889313, + 'data_time': 0.000705639977240935, + 'model_time': 1.229165242984891, + 'grad_norm_pre_clip_avg': 0.14600942581892012, + 'learning_rate': 2.2983624508200857e-06, + 'epoch': 10.4} +04/20 [02:22:35] INFO | >> train_qwenlatent.py:487 + Step 41230 | grad_norm_pre_clip=0.0893 | + grad_norm_pre_clip_avg=0.1429 | Metrics: + {'align_loss': 0.024953927844762802, + 'recon_loss': 0.1361304223537445, + 'predict_loss': 0.006560063920915127, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.08926994353532791, + 'data_time': 0.0010631270124576986, + 'model_time': 1.2609311190026347, + 'grad_norm_pre_clip_avg': 0.14290898218750953, + 'learning_rate': 2.293349601357111e-06, + 'epoch': 10.4} +04/20 [02:22:48] INFO | >> train_qwenlatent.py:487 + Step 41240 | grad_norm_pre_clip=0.1635 | + grad_norm_pre_clip_avg=0.1357 | Metrics: + {'align_loss': 0.02627461403608322, + 'recon_loss': 0.17828522622585297, + 'predict_loss': 0.008209883235394955, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16351592540740967, + 'data_time': 0.000994405010715127, + 'model_time': 1.3000543630041648, + 'grad_norm_pre_clip_avg': 0.13574462682008742, + 'learning_rate': 2.288341732583896e-06, + 'epoch': 10.41} +04/20 [02:23:01] INFO | >> train_qwenlatent.py:487 + Step 41250 | grad_norm_pre_clip=0.1634 | + grad_norm_pre_clip_avg=0.1616 | Metrics: + {'align_loss': 0.02501828968524933, + 'recon_loss': 0.0925959125161171, + 'predict_loss': 0.0033433460630476475, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16338026523590088, + 'mae_score': 0.005551166362590618, 'data_time': + 0.0006813420040998608, 'model_time': + 1.1666843660059385, 'grad_norm_pre_clip_avg': + 0.16159981489181519, 'learning_rate': + 2.2833388469412136e-06, 'epoch': 10.41} +04/20 [02:23:13] INFO | >> train_qwenlatent.py:487 + Step 41260 | grad_norm_pre_clip=0.1744 | + grad_norm_pre_clip_avg=0.1523 | Metrics: + {'align_loss': 0.0262961033731699, + 'recon_loss': 0.14717592298984528, + 'predict_loss': 0.0060188681818544865, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1744309365749359, + 'data_time': 0.000884517008671537, + 'model_time': 1.219418938999297, + 'grad_norm_pre_clip_avg': 0.1522959850728512, + 'learning_rate': 2.2783409468674155e-06, + 'epoch': 10.41} +04/20 [02:23:26] INFO | >> train_qwenlatent.py:487 + Step 41270 | grad_norm_pre_clip=0.1527 | + grad_norm_pre_clip_avg=0.1468 | Metrics: + {'align_loss': 0.02586793527007103, + 'recon_loss': 0.12809418141841888, + 'predict_loss': 0.009360664524137974, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1527440994977951, + 'data_time': 0.0009684910182841122, + 'model_time': 1.2318279879982583, + 'grad_norm_pre_clip_avg': 0.14681702554225923, + 'learning_rate': 2.2733480347984135e-06, + 'epoch': 10.41} +04/20 [02:23:39] INFO | >> train_qwenlatent.py:487 + Step 41280 | grad_norm_pre_clip=0.1417 | + grad_norm_pre_clip_avg=0.1485 | Metrics: + {'align_loss': 0.02493106946349144, + 'recon_loss': 0.14714254438877106, + 'predict_loss': 0.00804794579744339, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1417279690504074, + 'data_time': 0.000753136002458632, + 'model_time': 1.480769721994875, + 'grad_norm_pre_clip_avg': 0.14847112968564033, + 'learning_rate': 2.2683601131676962e-06, + 'epoch': 10.42} +04/20 [02:23:51] INFO | >> train_qwenlatent.py:487 + Step 41290 | grad_norm_pre_clip=0.1353 | + grad_norm_pre_clip_avg=0.1495 | Metrics: + {'align_loss': 0.023591596633195877, + 'recon_loss': 0.11354631185531616, + 'predict_loss': 0.006596843712031841, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1353253275156021, + 'data_time': 0.0007333139947149903, + 'model_time': 1.2753156449762173, + 'grad_norm_pre_clip_avg': 0.1495148979127407, + 'learning_rate': 2.2633771844063052e-06, + 'epoch': 10.42} +04/20 [02:24:05] INFO | >> train_qwenlatent.py:487 + Step 41300 | grad_norm_pre_clip=0.1889 | + grad_norm_pre_clip_avg=0.1618 | Metrics: + {'align_loss': 0.025606293231248856, + 'recon_loss': 0.13924071192741394, + 'predict_loss': 0.007134430110454559, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1889253705739975, + 'mae_score': 0.006813469448605099, 'data_time': + 0.0009020199941005558, 'model_time': + 1.2134390659921337, 'grad_norm_pre_clip_avg': + 0.16182195395231247, 'learning_rate': + 2.2583992509428708e-06, 'epoch': 10.42} +04/20 [02:24:17] INFO | >> train_qwenlatent.py:487 + Step 41310 | grad_norm_pre_clip=0.1187 | + grad_norm_pre_clip_avg=0.1642 | Metrics: + {'align_loss': 0.024238083511590958, + 'recon_loss': 0.09512743353843689, + 'predict_loss': 0.0032572587952017784, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11868760734796524, + 'data_time': 0.000721893971785903, + 'model_time': 1.239353597018635, + 'grad_norm_pre_clip_avg': 0.1642094448208809, + 'learning_rate': 2.253426315203574e-06, + 'epoch': 10.42} +04/20 [02:24:30] INFO | >> train_qwenlatent.py:487 + Step 41320 | grad_norm_pre_clip=0.1049 | + grad_norm_pre_clip_avg=0.1666 | Metrics: + {'align_loss': 0.025843802839517593, + 'recon_loss': 0.09835084527730942, + 'predict_loss': 0.005211926531046629, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10485443472862244, + 'data_time': 0.0007697840046603233, + 'model_time': 1.2090363340103067, + 'grad_norm_pre_clip_avg': 0.16656063944101335, + 'learning_rate': 2.248458379612161e-06, + 'epoch': 10.43} +04/20 [02:24:42] INFO | >> train_qwenlatent.py:487 + Step 41330 | grad_norm_pre_clip=0.1363 | + grad_norm_pre_clip_avg=0.1519 | Metrics: + {'align_loss': 0.025439467281103134, + 'recon_loss': 0.12846209108829498, + 'predict_loss': 0.004726043902337551, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13634203374385834, + 'data_time': 0.0007958609785418957, + 'model_time': 1.2524926789919846, + 'grad_norm_pre_clip_avg': 0.15188396275043486, + 'learning_rate': 2.2434954465899466e-06, + 'epoch': 10.43} +04/20 [02:24:55] INFO | >> train_qwenlatent.py:487 + Step 41340 | grad_norm_pre_clip=0.1732 | + grad_norm_pre_clip_avg=0.1620 | Metrics: + {'align_loss': 0.023039519786834717, + 'recon_loss': 0.11605957895517349, + 'predict_loss': 0.004947085399180651, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17320093512535095, + 'data_time': 0.0006864489987492561, + 'model_time': 1.2576375310018193, + 'grad_norm_pre_clip_avg': 0.16195600628852844, + 'learning_rate': 2.2385375185557974e-06, + 'epoch': 10.43} +04/20 [02:25:08] INFO | >> train_qwenlatent.py:487 + Step 41350 | grad_norm_pre_clip=0.2049 | + grad_norm_pre_clip_avg=0.1633 | Metrics: + {'align_loss': 0.025186244398355484, + 'recon_loss': 0.11048610508441925, + 'predict_loss': 0.007047676015645266, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2048909068107605, + 'mae_score': 0.006091685767646309, 'data_time': + 0.0006559939938597381, 'model_time': + 1.2103616060048807, 'grad_norm_pre_clip_avg': + 0.16333990022540093, 'learning_rate': + 2.2335845979261564e-06, 'epoch': 10.43} +04/20 [02:25:21] INFO | >> train_qwenlatent.py:487 + Step 41360 | grad_norm_pre_clip=0.1727 | + grad_norm_pre_clip_avg=0.1821 | Metrics: + {'align_loss': 0.02498379908502102, + 'recon_loss': 0.1297759860754013, + 'predict_loss': 0.006848767399787903, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17265404760837555, + 'data_time': 0.0010700470011215657, + 'model_time': 1.2659057369746733, + 'grad_norm_pre_clip_avg': 0.18206988275051117, + 'learning_rate': 2.228636687115012e-06, + 'epoch': 10.44} +04/20 [02:25:34] INFO | >> train_qwenlatent.py:487 + Step 41370 | grad_norm_pre_clip=0.1623 | + grad_norm_pre_clip_avg=0.1581 | Metrics: + {'align_loss': 0.024448461830615997, + 'recon_loss': 0.12127377837896347, + 'predict_loss': 0.0053104194812476635, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16233795881271362, + 'data_time': 0.0006579319888260216, + 'model_time': 1.242612409987487, + 'grad_norm_pre_clip_avg': 0.1581282764673233, + 'learning_rate': 2.223693788533919e-06, + 'epoch': 10.44} +04/20 [02:25:46] INFO | >> train_qwenlatent.py:487 + Step 41380 | grad_norm_pre_clip=0.1182 | + grad_norm_pre_clip_avg=0.1593 | Metrics: + {'align_loss': 0.024773631244897842, + 'recon_loss': 0.12296590954065323, + 'predict_loss': 0.013752312399446964, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11822143942117691, + 'data_time': 0.000867250986630097, + 'model_time': 1.221243047999451, + 'grad_norm_pre_clip_avg': 0.15934175476431847, + 'learning_rate': 2.218755904591982e-06, + 'epoch': 10.44} +04/20 [02:25:59] INFO | >> train_qwenlatent.py:487 + Step 41390 | grad_norm_pre_clip=0.1562 | + grad_norm_pre_clip_avg=0.1729 | Metrics: + {'align_loss': 0.024119893088936806, + 'recon_loss': 0.19075144827365875, + 'predict_loss': 0.008418119512498379, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15619763731956482, + 'data_time': 0.0009747360018081963, + 'model_time': 1.2292772170039825, + 'grad_norm_pre_clip_avg': 0.17291703820228577, + 'learning_rate': 2.2138230376958637e-06, + 'epoch': 10.44} +04/20 [02:26:12] INFO | >> train_qwenlatent.py:487 + Step 41400 | grad_norm_pre_clip=0.1620 | + grad_norm_pre_clip_avg=0.1402 | Metrics: + {'align_loss': 0.025514405220746994, + 'recon_loss': 0.15523280203342438, + 'predict_loss': 0.008351880125701427, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16198085248470306, + 'mae_score': 0.00611055090620711, 'data_time': + 0.0009657189948484302, 'model_time': + 1.568031364993658, 'grad_norm_pre_clip_avg': + 0.14016810432076454, 'learning_rate': + 2.2088951902497907e-06, 'epoch': 10.45} +04/20 [02:26:25] INFO | >> train_qwenlatent.py:487 + Step 41410 | grad_norm_pre_clip=0.1665 | + grad_norm_pre_clip_avg=0.1482 | Metrics: + {'align_loss': 0.024632561951875687, + 'recon_loss': 0.13520459830760956, + 'predict_loss': 0.009065148420631886, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16650742292404175, + 'data_time': 0.0007141899841371924, + 'model_time': 1.286432576016523, + 'grad_norm_pre_clip_avg': 0.14815164878964424, + 'learning_rate': 2.2039723646555316e-06, + 'epoch': 10.45} +04/20 [02:26:38] INFO | >> train_qwenlatent.py:487 + Step 41420 | grad_norm_pre_clip=0.1533 | + grad_norm_pre_clip_avg=0.1358 | Metrics: + {'align_loss': 0.024083316326141357, + 'recon_loss': 0.09834244102239609, + 'predict_loss': 0.003298992058262229, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.153328537940979, + 'data_time': 0.000644739979179576, + 'model_time': 1.5247421720123384, + 'grad_norm_pre_clip_avg': 0.1357949823141098, + 'learning_rate': 2.1990545633124114e-06, + 'epoch': 10.45} +04/20 [02:26:50] INFO | >> train_qwenlatent.py:487 + Step 41430 | grad_norm_pre_clip=0.1253 | + grad_norm_pre_clip_avg=0.1269 | Metrics: + {'align_loss': 0.024555448442697525, + 'recon_loss': 0.1744774878025055, + 'predict_loss': 0.007071726024150848, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12529872357845306, + 'data_time': 0.0008192660170607269, + 'model_time': 1.2998112910136115, + 'grad_norm_pre_clip_avg': 0.12692410349845887, + 'learning_rate': 2.19414178861731e-06, 'epoch': + 10.45} +04/20 [02:27:03] INFO | >> train_qwenlatent.py:487 + Step 41440 | grad_norm_pre_clip=0.1629 | + grad_norm_pre_clip_avg=0.1509 | Metrics: + {'align_loss': 0.0243716761469841, + 'recon_loss': 0.16823013126850128, + 'predict_loss': 0.007227153982967138, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16290904581546783, + 'data_time': 0.0009724419796839356, + 'model_time': 1.2645642400020733, + 'grad_norm_pre_clip_avg': 0.15090304166078566, + 'learning_rate': 2.1892340429646487e-06, + 'epoch': 10.46} +04/20 [02:27:16] INFO | >> train_qwenlatent.py:487 + Step 41450 | grad_norm_pre_clip=0.1338 | + grad_norm_pre_clip_avg=0.1493 | Metrics: + {'align_loss': 0.026407359167933464, + 'recon_loss': 0.11955748498439789, + 'predict_loss': 0.008073786273598671, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1338329315185547, + 'mae_score': 0.0067009711050772455, + 'data_time': 0.000955696013988927, + 'model_time': 1.550843758013798, + 'grad_norm_pre_clip_avg': 0.14926375225186347, + 'learning_rate': 2.1843313287464077e-06, + 'epoch': 10.46} +04/20 [02:27:29] INFO | >> train_qwenlatent.py:487 + Step 41460 | grad_norm_pre_clip=0.2089 | + grad_norm_pre_clip_avg=0.1601 | Metrics: + {'align_loss': 0.025271790102124214, + 'recon_loss': 0.1353546679019928, + 'predict_loss': 0.007760712411254644, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2089214026927948, + 'data_time': 0.000965558981988579, + 'model_time': 1.245658091007499, + 'grad_norm_pre_clip_avg': 0.1601472556591034, + 'learning_rate': 2.1794336483521136e-06, + 'epoch': 10.46} +04/20 [02:27:41] INFO | >> train_qwenlatent.py:487 + Step 41470 | grad_norm_pre_clip=0.1967 | + grad_norm_pre_clip_avg=0.1637 | Metrics: + {'align_loss': 0.025706537067890167, + 'recon_loss': 0.16693015396595, 'predict_loss': + 0.012603151611983776, 'aux_loss_decay_weight': + 0.0, 'grad_norm_pre_clip': 0.19670557975769043, + 'data_time': 0.0009325410064775497, + 'model_time': 1.5129532149876468, + 'grad_norm_pre_clip_avg': 0.16369761824607848, + 'learning_rate': 2.1745410041688283e-06, + 'epoch': 10.46} +04/20 [02:27:54] INFO | >> train_qwenlatent.py:487 + Step 41480 | grad_norm_pre_clip=0.1575 | + grad_norm_pre_clip_avg=0.1693 | Metrics: + {'align_loss': 0.02690492570400238, + 'recon_loss': 0.1534534990787506, + 'predict_loss': 0.006911619566380978, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15747100114822388, + 'data_time': 0.0008326730167027563, + 'model_time': 1.2567100930027664, + 'grad_norm_pre_clip_avg': 0.169296532869339, + 'learning_rate': 2.1696533985811683e-06, + 'epoch': 10.47} +04/20 [02:28:06] INFO | >> train_qwenlatent.py:487 + Step 41490 | grad_norm_pre_clip=0.1324 | + grad_norm_pre_clip_avg=0.1649 | Metrics: + {'align_loss': 0.025525256991386414, + 'recon_loss': 0.13939505815505981, + 'predict_loss': 0.00538030406460166, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13238587975502014, + 'data_time': 0.0010552339954301715, + 'model_time': 1.3407094489957672, + 'grad_norm_pre_clip_avg': 0.1648748755455017, + 'learning_rate': 2.1647708339712973e-06, + 'epoch': 10.47} +04/20 [02:28:19] INFO | >> train_qwenlatent.py:487 + Step 41500 | grad_norm_pre_clip=0.1238 | + grad_norm_pre_clip_avg=0.1398 | Metrics: + {'align_loss': 0.02325555309653282, + 'recon_loss': 0.15246649086475372, + 'predict_loss': 0.008541389368474483, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12378111481666565, + 'mae_score': 0.005323889019253018, 'data_time': + 0.0012684910034295171, 'model_time': + 1.2628892219800036, 'grad_norm_pre_clip_avg': + 0.13984301835298538, 'learning_rate': + 2.1598933127189177e-06, 'epoch': 10.47} +04/20 [02:28:32] INFO | >> train_qwenlatent.py:487 + Step 41510 | grad_norm_pre_clip=0.1715 | + grad_norm_pre_clip_avg=0.1420 | Metrics: + {'align_loss': 0.02503543719649315, + 'recon_loss': 0.10679785907268524, + 'predict_loss': 0.006137360818684101, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17154031991958618, + 'data_time': 0.000630350987194106, + 'model_time': 1.1928139640076552, + 'grad_norm_pre_clip_avg': 0.14202467426657678, + 'learning_rate': 2.1550208372012718e-06, + 'epoch': 10.47} +04/20 [02:28:44] INFO | >> train_qwenlatent.py:487 + Step 41520 | grad_norm_pre_clip=0.0991 | + grad_norm_pre_clip_avg=0.1504 | Metrics: + {'align_loss': 0.02548639103770256, + 'recon_loss': 0.10520529747009277, + 'predict_loss': 0.005412655882537365, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09906335920095444, + 'data_time': 0.0008803950040601194, + 'model_time': 1.2884946519916411, + 'grad_norm_pre_clip_avg': 0.15043671801686287, + 'learning_rate': 2.150153409793146e-06, + 'epoch': 10.48} +04/20 [02:28:57] INFO | >> train_qwenlatent.py:487 + Step 41530 | grad_norm_pre_clip=0.1498 | + grad_norm_pre_clip_avg=0.1461 | Metrics: + {'align_loss': 0.024776127189397812, + 'recon_loss': 0.12959152460098267, + 'predict_loss': 0.007486632093787193, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1497776359319687, + 'data_time': 0.0010170229943469167, + 'model_time': 1.2364979749836493, + 'grad_norm_pre_clip_avg': 0.14606366753578187, + 'learning_rate': 2.145291032866864e-06, + 'epoch': 10.48} +04/20 [02:29:10] INFO | >> train_qwenlatent.py:487 + Step 41540 | grad_norm_pre_clip=0.1723 | + grad_norm_pre_clip_avg=0.1504 | Metrics: + {'align_loss': 0.024238307029008865, + 'recon_loss': 0.13033534586429596, + 'predict_loss': 0.009154478088021278, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1723046898841858, + 'data_time': 0.0009635089954826981, + 'model_time': 1.2443993000197224, + 'grad_norm_pre_clip_avg': 0.15038829818367958, + 'learning_rate': 2.140433708792288e-06, + 'epoch': 10.48} +04/20 [02:29:23] INFO | >> train_qwenlatent.py:487 + Step 41550 | grad_norm_pre_clip=0.1536 | + grad_norm_pre_clip_avg=0.1579 | Metrics: + {'align_loss': 0.024512268602848053, + 'recon_loss': 0.10283541679382324, + 'predict_loss': 0.0033134028781205416, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15362299978733063, + 'mae_score': 0.0064765272913752375, + 'data_time': 0.0009604140068404377, + 'model_time': 1.2245810120075475, + 'grad_norm_pre_clip_avg': 0.15792919024825097, + 'learning_rate': 2.135581439936828e-06, + 'epoch': 10.48} +04/20 [02:29:36] INFO | >> train_qwenlatent.py:487 + Step 41560 | grad_norm_pre_clip=0.1945 | + grad_norm_pre_clip_avg=0.1705 | Metrics: + {'align_loss': 0.02587185800075531, + 'recon_loss': 0.1549079418182373, + 'predict_loss': 0.007730869110673666, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19447915256023407, + 'data_time': 0.0007424160139635205, + 'model_time': 1.212550274998648, + 'grad_norm_pre_clip_avg': 0.17049231752753258, + 'learning_rate': 2.13073422866541e-06, 'epoch': + 10.49} +04/20 [02:29:48] INFO | >> train_qwenlatent.py:487 + Step 41570 | grad_norm_pre_clip=0.2557 | + grad_norm_pre_clip_avg=0.1591 | Metrics: + {'align_loss': 0.024436715990304947, + 'recon_loss': 0.1339598298072815, + 'predict_loss': 0.008415318094193935, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2556813955307007, + 'data_time': 0.0007770360098220408, + 'model_time': 1.2296432080038358, + 'grad_norm_pre_clip_avg': 0.1590816892683506, + 'learning_rate': 2.125892077340509e-06, + 'epoch': 10.49} +04/20 [02:30:01] INFO | >> train_qwenlatent.py:487 + Step 41580 | grad_norm_pre_clip=0.1728 | + grad_norm_pre_clip_avg=0.1617 | Metrics: + {'align_loss': 0.02615288831293583, + 'recon_loss': 0.11800133436918259, + 'predict_loss': 0.0071337162517011166, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17275585234165192, + 'data_time': 0.001046751014655456, + 'model_time': 1.2703025409718975, + 'grad_norm_pre_clip_avg': 0.1617286242544651, + 'learning_rate': 2.12105498832213e-06, 'epoch': + 10.49} +04/20 [02:30:13] INFO | >> train_qwenlatent.py:487 + Step 41590 | grad_norm_pre_clip=0.1220 | + grad_norm_pre_clip_avg=0.1466 | Metrics: + {'align_loss': 0.024901343509554863, + 'recon_loss': 0.1040586531162262, + 'predict_loss': 0.005871258210390806, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12200678139925003, + 'data_time': 0.0007343410106841475, + 'model_time': 1.2439732959901448, + 'grad_norm_pre_clip_avg': 0.1466474451124668, + 'learning_rate': 2.116222963967815e-06, + 'epoch': 10.49} +04/20 [02:30:27] INFO | >> train_qwenlatent.py:487 + Step 41600 | grad_norm_pre_clip=0.1284 | + grad_norm_pre_clip_avg=0.1521 | Metrics: + {'align_loss': 0.024129828438162804, + 'recon_loss': 0.0890217199921608, + 'predict_loss': 0.003798763733357191, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12837955355644226, + 'mae_score': 0.005115119401399079, 'data_time': + 0.001335064007434994, 'model_time': + 1.256883124995511, 'grad_norm_pre_clip_avg': + 0.15213584154844284, 'learning_rate': + 2.1113960066326336e-06, 'epoch': 10.5} +04/20 [02:30:39] INFO | >> train_qwenlatent.py:487 + Step 41610 | grad_norm_pre_clip=0.1507 | + grad_norm_pre_clip_avg=0.1633 | Metrics: + {'align_loss': 0.02491540089249611, + 'recon_loss': 0.10453537106513977, + 'predict_loss': 0.004853155929595232, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15070252120494843, + 'data_time': 0.0006949490052647889, + 'model_time': 1.1996966259903274, + 'grad_norm_pre_clip_avg': 0.1632534943521023, + 'learning_rate': 2.106574118669183e-06, + 'epoch': 10.5} +04/20 [02:30:52] INFO | >> train_qwenlatent.py:487 + Step 41620 | grad_norm_pre_clip=0.1406 | + grad_norm_pre_clip_avg=0.1590 | Metrics: + {'align_loss': 0.024861328303813934, + 'recon_loss': 0.12282247841358185, + 'predict_loss': 0.006384701933711767, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1405743509531021, + 'data_time': 0.000962866994086653, + 'model_time': 1.2680141380114947, + 'grad_norm_pre_clip_avg': 0.15902689546346666, + 'learning_rate': 2.1017573024275944e-06, + 'epoch': 10.5} +04/20 [02:31:04] INFO | >> train_qwenlatent.py:487 + Step 41630 | grad_norm_pre_clip=0.1578 | + grad_norm_pre_clip_avg=0.1700 | Metrics: + {'align_loss': 0.025012604892253876, + 'recon_loss': 0.1531956046819687, + 'predict_loss': 0.006267824675887823, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15783736109733582, + 'data_time': 0.0010760209988802671, + 'model_time': 1.1951539069996215, + 'grad_norm_pre_clip_avg': 0.16995554640889168, + 'learning_rate': 2.0969455602555256e-06, + 'epoch': 10.5} +04/20 [02:31:17] INFO | >> train_qwenlatent.py:487 + Step 41640 | grad_norm_pre_clip=0.2051 | + grad_norm_pre_clip_avg=0.1947 | Metrics: + {'align_loss': 0.024566996842622757, + 'recon_loss': 0.12307808548212051, + 'predict_loss': 0.007611478678882122, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20512649416923523, + 'data_time': 0.000996340997517109, + 'model_time': 1.2261081530014053, + 'grad_norm_pre_clip_avg': 0.1946589782834053, + 'learning_rate': 2.092138894498162e-06, + 'epoch': 10.51} +04/20 [02:31:29] INFO | >> train_qwenlatent.py:487 + Step 41650 | grad_norm_pre_clip=0.1153 | + grad_norm_pre_clip_avg=0.1547 | Metrics: + {'align_loss': 0.024182479828596115, + 'recon_loss': 0.13223201036453247, + 'predict_loss': 0.007050258573144674, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11530852317810059, + 'mae_score': 0.005292993408065659, 'data_time': + 0.0009394189983140677, 'model_time': + 1.2844112689781468, 'grad_norm_pre_clip_avg': + 0.1546619154512882, 'learning_rate': + 2.0873373074982123e-06, 'epoch': 10.51} +04/20 [02:31:42] INFO | >> train_qwenlatent.py:487 + Step 41660 | grad_norm_pre_clip=0.1342 | + grad_norm_pre_clip_avg=0.1410 | Metrics: + {'align_loss': 0.025556139647960663, + 'recon_loss': 0.1348576694726944, + 'predict_loss': 0.0058098104782402515, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1342228204011917, + 'data_time': 0.0010459040058776736, + 'model_time': 1.2610172269924078, + 'grad_norm_pre_clip_avg': 0.14095954298973085, + 'learning_rate': 2.082540801595914e-06, + 'epoch': 10.51} +04/20 [02:31:55] INFO | >> train_qwenlatent.py:487 + Step 41670 | grad_norm_pre_clip=0.1425 | + grad_norm_pre_clip_avg=0.1390 | Metrics: + {'align_loss': 0.025703005492687225, + 'recon_loss': 0.14896482229232788, + 'predict_loss': 0.00732953567057848, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14253701269626617, + 'data_time': 0.0009403179865330458, + 'model_time': 1.2240061170014087, + 'grad_norm_pre_clip_avg': 0.1390339180827141, + 'learning_rate': 2.07774937912902e-06, 'epoch': + 10.51} +04/20 [02:32:07] INFO | >> train_qwenlatent.py:487 + Step 41680 | grad_norm_pre_clip=0.1763 | + grad_norm_pre_clip_avg=0.1434 | Metrics: + {'align_loss': 0.025361530482769012, + 'recon_loss': 0.14479273557662964, + 'predict_loss': 0.007036555092781782, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17626358568668365, + 'data_time': 0.0007685399905312806, + 'model_time': 1.2536414339847397, + 'grad_norm_pre_clip_avg': 0.1433902971446514, + 'learning_rate': 2.0729630424328193e-06, + 'epoch': 10.52} +04/20 [02:32:20] INFO | >> train_qwenlatent.py:487 + Step 41690 | grad_norm_pre_clip=0.1713 | + grad_norm_pre_clip_avg=0.1571 | Metrics: + {'align_loss': 0.026968859136104584, + 'recon_loss': 0.16355721652507782, + 'predict_loss': 0.009149327874183655, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1713358461856842, + 'data_time': 0.0006685559928882867, + 'model_time': 1.2973013370065019, + 'grad_norm_pre_clip_avg': 0.15708512663841248, + 'learning_rate': 2.06818179384011e-06, 'epoch': + 10.52} +04/20 [02:32:34] INFO | >> train_qwenlatent.py:487 + Step 41700 | grad_norm_pre_clip=0.2395 | + grad_norm_pre_clip_avg=0.1557 | Metrics: + {'align_loss': 0.0247048307210207, + 'recon_loss': 0.16845734417438507, + 'predict_loss': 0.012299297377467155, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.23954840004444122, + 'mae_score': 0.005428731763685072, 'data_time': + 0.0010549950238782912, 'model_time': + 1.3557484239863697, 'grad_norm_pre_clip_avg': + 0.15570313557982446, 'learning_rate': + 2.0634056356812167e-06, 'epoch': 10.52} +04/20 [02:32:46] INFO | >> train_qwenlatent.py:487 + Step 41710 | grad_norm_pre_clip=0.1480 | + grad_norm_pre_clip_avg=0.1711 | Metrics: + {'align_loss': 0.02529750019311905, + 'recon_loss': 0.15451647341251373, + 'predict_loss': 0.007507260888814926, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14798729121685028, + 'data_time': 0.000838929001474753, + 'model_time': 1.2189980869879946, + 'grad_norm_pre_clip_avg': 0.171076150983572, + 'learning_rate': 2.058634570283979e-06, + 'epoch': 10.52} +04/20 [02:32:59] INFO | >> train_qwenlatent.py:487 + Step 41720 | grad_norm_pre_clip=0.1269 | + grad_norm_pre_clip_avg=0.1569 | Metrics: + {'align_loss': 0.02562762424349785, + 'recon_loss': 0.11963377892971039, + 'predict_loss': 0.0034407342318445444, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1269252598285675, + 'data_time': 0.0008194429974537343, + 'model_time': 1.2326875439903233, + 'grad_norm_pre_clip_avg': 0.15687140226364135, + 'learning_rate': 2.0538685999737563e-06, + 'epoch': 10.53} +04/20 [02:33:11] INFO | >> train_qwenlatent.py:487 + Step 41730 | grad_norm_pre_clip=0.1759 | + grad_norm_pre_clip_avg=0.1563 | Metrics: + {'align_loss': 0.024985864758491516, + 'recon_loss': 0.11828533560037613, + 'predict_loss': 0.006566737778484821, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1759198009967804, + 'data_time': 0.001068240002496168, + 'model_time': 1.2380957109853625, + 'grad_norm_pre_clip_avg': 0.15627639964222909, + 'learning_rate': 2.0491077270734252e-06, + 'epoch': 10.53} +04/20 [02:33:24] INFO | >> train_qwenlatent.py:487 + Step 41740 | grad_norm_pre_clip=0.1087 | + grad_norm_pre_clip_avg=0.1513 | Metrics: + {'align_loss': 0.02484690397977829, + 'recon_loss': 0.13455425202846527, + 'predict_loss': 0.009722843766212463, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10870079696178436, + 'data_time': 0.0009548620146233588, + 'model_time': 1.2215101790206973, + 'grad_norm_pre_clip_avg': 0.15131903439760208, + 'learning_rate': 2.0443519539033775e-06, + 'epoch': 10.53} +04/20 [02:33:37] INFO | >> train_qwenlatent.py:487 + Step 41750 | grad_norm_pre_clip=0.1249 | + grad_norm_pre_clip_avg=0.1623 | Metrics: + {'align_loss': 0.026641644537448883, + 'recon_loss': 0.1126154363155365, + 'predict_loss': 0.006670952774584293, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12487278133630753, + 'mae_score': 0.007696587330586201, 'data_time': + 0.0007153360056690872, 'model_time': + 1.2191931350098457, 'grad_norm_pre_clip_avg': + 0.16229807063937188, 'learning_rate': + 2.039601282781519e-06, 'epoch': 10.53} +04/20 [02:33:50] INFO | >> train_qwenlatent.py:487 + Step 41760 | grad_norm_pre_clip=0.1206 | + grad_norm_pre_clip_avg=0.1587 | Metrics: + {'align_loss': 0.02491367980837822, + 'recon_loss': 0.07887636870145798, + 'predict_loss': 0.0036154352128505707, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12057920545339584, + 'data_time': 0.0007600240060128272, + 'model_time': 1.2968649909889791, + 'grad_norm_pre_clip_avg': 0.15873347148299216, + 'learning_rate': 2.0348557160232693e-06, + 'epoch': 10.54} +04/20 [02:34:03] INFO | >> train_qwenlatent.py:487 + Step 41770 | grad_norm_pre_clip=0.1174 | + grad_norm_pre_clip_avg=0.1589 | Metrics: + {'align_loss': 0.026371877640485764, + 'recon_loss': 0.22430630028247833, + 'predict_loss': 0.008104742504656315, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11735040694475174, + 'data_time': 0.0007049950072541833, + 'model_time': 1.3012400020088535, + 'grad_norm_pre_clip_avg': 0.15894974023103714, + 'learning_rate': 2.0301152559415557e-06, + 'epoch': 10.54} +04/20 [02:34:15] INFO | >> train_qwenlatent.py:487 + Step 41780 | grad_norm_pre_clip=0.0988 | + grad_norm_pre_clip_avg=0.1431 | Metrics: + {'align_loss': 0.025071173906326294, + 'recon_loss': 0.09261228889226913, + 'predict_loss': 0.0026085935533046722, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09881342202425003, + 'data_time': 0.0007233479991555214, + 'model_time': 1.2933118350047152, + 'grad_norm_pre_clip_avg': 0.1430962286889553, + 'learning_rate': 2.0253799048468297e-06, + 'epoch': 10.54} +04/20 [02:34:28] INFO | >> train_qwenlatent.py:487 + Step 41790 | grad_norm_pre_clip=0.1977 | + grad_norm_pre_clip_avg=0.1605 | Metrics: + {'align_loss': 0.02511749416589737, + 'recon_loss': 0.18533408641815186, + 'predict_loss': 0.01681370846927166, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1977040022611618, + 'data_time': 0.0010803949844557792, + 'model_time': 1.2155281120212749, + 'grad_norm_pre_clip_avg': 0.1605231449007988, + 'learning_rate': 2.0206496650470385e-06, + 'epoch': 10.55} +04/20 [02:34:41] INFO | >> train_qwenlatent.py:487 + Step 41800 | grad_norm_pre_clip=0.1555 | + grad_norm_pre_clip_avg=0.1509 | Metrics: + {'align_loss': 0.02584223449230194, + 'recon_loss': 0.16945065557956696, + 'predict_loss': 0.007451043929904699, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15545450150966644, + 'mae_score': 0.006367112065220738, 'data_time': + 0.0007971230079419911, 'model_time': + 1.210284058994148, 'grad_norm_pre_clip_avg': + 0.1509454295039177, 'learning_rate': + 2.015924538847643e-06, 'epoch': 10.55} +04/20 [02:34:53] INFO | >> train_qwenlatent.py:487 + Step 41810 | grad_norm_pre_clip=0.1294 | + grad_norm_pre_clip_avg=0.1502 | Metrics: + {'align_loss': 0.026666350662708282, + 'recon_loss': 0.1531158983707428, + 'predict_loss': 0.010764512233436108, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1294357031583786, + 'data_time': 0.0007056160247884691, + 'model_time': 1.2406652880017646, + 'grad_norm_pre_clip_avg': 0.15018569976091384, + 'learning_rate': 2.0112045285516143e-06, + 'epoch': 10.55} +04/20 [02:35:06] INFO | >> train_qwenlatent.py:487 + Step 41820 | grad_norm_pre_clip=0.1476 | + grad_norm_pre_clip_avg=0.1356 | Metrics: + {'align_loss': 0.025954540818929672, + 'recon_loss': 0.13844984769821167, + 'predict_loss': 0.004776733927428722, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14757013320922852, + 'data_time': 0.000695069000357762, + 'model_time': 1.2570002639840823, + 'grad_norm_pre_clip_avg': 0.1356320172548294, + 'learning_rate': 2.0064896364594276e-06, + 'epoch': 10.55} +04/20 [02:35:19] INFO | >> train_qwenlatent.py:487 + Step 41830 | grad_norm_pre_clip=0.1225 | + grad_norm_pre_clip_avg=0.1584 | Metrics: + {'align_loss': 0.025767765939235687, + 'recon_loss': 0.16406302154064178, + 'predict_loss': 0.005537566728889942, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12254511564970016, + 'data_time': 0.0009086520003620535, + 'model_time': 1.1994134210108314, + 'grad_norm_pre_clip_avg': 0.1584183894097805, + 'learning_rate': 2.0017798648690624e-06, + 'epoch': 10.56} +04/20 [02:35:32] INFO | >> train_qwenlatent.py:487 + Step 41840 | grad_norm_pre_clip=0.1104 | + grad_norm_pre_clip_avg=0.1240 | Metrics: + {'align_loss': 0.024125520139932632, + 'recon_loss': 0.09612604230642319, + 'predict_loss': 0.0035147264134138823, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11039849370718002, + 'data_time': 0.0007685729942750186, + 'model_time': 1.232477479003137, + 'grad_norm_pre_clip_avg': 0.12402800470590591, + 'learning_rate': 1.997075216076006e-06, + 'epoch': 10.56} +04/20 [02:35:45] INFO | >> train_qwenlatent.py:487 + Step 41850 | grad_norm_pre_clip=0.1933 | + grad_norm_pre_clip_avg=0.1568 | Metrics: + {'align_loss': 0.024406537413597107, + 'recon_loss': 0.12773926556110382, + 'predict_loss': 0.006039072293788195, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19327814877033234, + 'mae_score': 0.00633200825871648, 'data_time': + 0.0009270540031138808, 'model_time': + 1.259767704992555, 'grad_norm_pre_clip_avg': + 0.15682030245661735, 'learning_rate': + 1.992375692373246e-06, 'epoch': 10.56} +04/20 [02:35:58] INFO | >> train_qwenlatent.py:487 + Step 41860 | grad_norm_pre_clip=0.1344 | + grad_norm_pre_clip_avg=0.1676 | Metrics: + {'align_loss': 0.025237876921892166, + 'recon_loss': 0.12492137402296066, + 'predict_loss': 0.00565977580845356, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13443821668624878, + 'data_time': 0.0008493389759678394, + 'model_time': 1.2584083849797025, + 'grad_norm_pre_clip_avg': 0.16764737516641617, + 'learning_rate': 1.9876812960512743e-06, + 'epoch': 10.56} +04/20 [02:36:11] INFO | >> train_qwenlatent.py:487 + Step 41870 | grad_norm_pre_clip=0.1614 | + grad_norm_pre_clip_avg=0.1656 | Metrics: + {'align_loss': 0.025839045643806458, + 'recon_loss': 0.22880686819553375, + 'predict_loss': 0.009182848036289215, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16135095059871674, + 'data_time': 0.0009181470086332411, + 'model_time': 1.2744711310078856, + 'grad_norm_pre_clip_avg': 0.1655938930809498, + 'learning_rate': 1.982992029398079e-06, + 'epoch': 10.57} +04/20 [02:36:24] INFO | >> train_qwenlatent.py:487 + Step 41880 | grad_norm_pre_clip=0.1970 | + grad_norm_pre_clip_avg=0.1359 | Metrics: + {'align_loss': 0.02613775245845318, + 'recon_loss': 0.1343090534210205, + 'predict_loss': 0.0048569124191999435, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19700288772583008, + 'data_time': 0.0009241340158041567, + 'model_time': 1.1858983840211295, + 'grad_norm_pre_clip_avg': 0.13587050065398215, + 'learning_rate': 1.978307894699158e-06, + 'epoch': 10.57} +04/20 [02:36:36] INFO | >> train_qwenlatent.py:487 + Step 41890 | grad_norm_pre_clip=0.2015 | + grad_norm_pre_clip_avg=0.1575 | Metrics: + {'align_loss': 0.02515460178256035, + 'recon_loss': 0.154060959815979, + 'predict_loss': 0.005632998887449503, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20148643851280212, + 'data_time': 0.0007686520111747086, + 'model_time': 1.2387665300047956, + 'grad_norm_pre_clip_avg': 0.1574595756828785, + 'learning_rate': 1.9736288942374975e-06, + 'epoch': 10.57} +04/20 [02:36:49] INFO | >> train_qwenlatent.py:487 + Step 41900 | grad_norm_pre_clip=0.2165 | + grad_norm_pre_clip_avg=0.1794 | Metrics: + {'align_loss': 0.02407665178179741, + 'recon_loss': 0.1417272984981537, + 'predict_loss': 0.010288897901773453, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2165268212556839, + 'mae_score': 0.005626969724088102, 'data_time': + 0.0006563290080521256, 'model_time': + 1.2200477680016775, 'grad_norm_pre_clip_avg': + 0.17943395748734475, 'learning_rate': + 1.9689550302935895e-06, 'epoch': 10.57} +04/20 [02:37:02] INFO | >> train_qwenlatent.py:487 + Step 41910 | grad_norm_pre_clip=0.2114 | + grad_norm_pre_clip_avg=0.1576 | Metrics: + {'align_loss': 0.026147861033678055, + 'recon_loss': 0.17467020452022552, + 'predict_loss': 0.006868443451821804, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21136629581451416, + 'data_time': 0.001058034977177158, + 'model_time': 1.2885238669987302, + 'grad_norm_pre_clip_avg': 0.15763386264443396, + 'learning_rate': 1.964286305145409e-06, + 'epoch': 10.58} +04/20 [02:37:14] INFO | >> train_qwenlatent.py:487 + Step 41920 | grad_norm_pre_clip=0.1109 | + grad_norm_pre_clip_avg=0.1451 | Metrics: + {'align_loss': 0.02436038665473461, + 'recon_loss': 0.11122108995914459, + 'predict_loss': 0.006862935144454241, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11085186153650284, + 'data_time': 0.0008047840092331171, + 'model_time': 1.2077603600046132, + 'grad_norm_pre_clip_avg': 0.1451020911335945, + 'learning_rate': 1.959622721068447e-06, + 'epoch': 10.58} +04/20 [02:37:27] INFO | >> train_qwenlatent.py:487 + Step 41930 | grad_norm_pre_clip=0.1457 | + grad_norm_pre_clip_avg=0.1615 | Metrics: + {'align_loss': 0.02408756874501705, + 'recon_loss': 0.1364334523677826, + 'predict_loss': 0.006279554218053818, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14566639065742493, + 'data_time': 0.0007150919991545379, + 'model_time': 1.5475997710018419, + 'grad_norm_pre_clip_avg': 0.16148075088858604, + 'learning_rate': 1.9549642803356715e-06, + 'epoch': 10.58} +04/20 [02:37:39] INFO | >> train_qwenlatent.py:487 + Step 41940 | grad_norm_pre_clip=0.2120 | + grad_norm_pre_clip_avg=0.1694 | Metrics: + {'align_loss': 0.024939069524407387, + 'recon_loss': 0.14657899737358093, + 'predict_loss': 0.006164675112813711, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21197891235351562, + 'data_time': 0.0007020919874776155, + 'model_time': 1.2100364740181249, + 'grad_norm_pre_clip_avg': 0.1694009058177471, + 'learning_rate': 1.9503109852175516e-06, + 'epoch': 10.58} +04/20 [02:37:52] INFO | >> train_qwenlatent.py:487 + Step 41950 | grad_norm_pre_clip=0.1881 | + grad_norm_pre_clip_avg=0.1920 | Metrics: + {'align_loss': 0.02474639192223549, + 'recon_loss': 0.15616144239902496, + 'predict_loss': 0.007016224320977926, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1881129890680313, + 'mae_score': 0.006845762063791086, 'data_time': + 0.0007414569845423102, 'model_time': + 1.2515564000059385, 'grad_norm_pre_clip_avg': + 0.19196900129318237, 'learning_rate': + 1.945662837982046e-06, 'epoch': 10.59} +04/20 [02:38:05] INFO | >> train_qwenlatent.py:487 + Step 41960 | grad_norm_pre_clip=0.2878 | + grad_norm_pre_clip_avg=0.1778 | Metrics: + {'align_loss': 0.02638203836977482, + 'recon_loss': 0.15809324383735657, + 'predict_loss': 0.007795637473464012, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.28782013058662415, + 'data_time': 0.000701982993632555, + 'model_time': 1.2719355789886322, + 'grad_norm_pre_clip_avg': 0.1778144896030426, + 'learning_rate': 1.941019840894603e-06, + 'epoch': 10.59} +04/20 [02:38:18] INFO | >> train_qwenlatent.py:487 + Step 41970 | grad_norm_pre_clip=0.1409 | + grad_norm_pre_clip_avg=0.1594 | Metrics: + {'align_loss': 0.02537158876657486, + 'recon_loss': 0.1487642526626587, + 'predict_loss': 0.004363992717117071, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14090298116207123, + 'data_time': 0.0007521070074290037, + 'model_time': 1.241005991003476, + 'grad_norm_pre_clip_avg': 0.15938705950975418, + 'learning_rate': 1.936381996218169e-06, + 'epoch': 10.59} +04/20 [02:38:31] INFO | >> train_qwenlatent.py:487 + Step 41980 | grad_norm_pre_clip=0.1463 | + grad_norm_pre_clip_avg=0.1568 | Metrics: + {'align_loss': 0.025426611304283142, + 'recon_loss': 0.1395091563463211, + 'predict_loss': 0.008096418343484402, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1463300585746765, + 'data_time': 0.0008886450086720288, + 'model_time': 1.2644423449819442, + 'grad_norm_pre_clip_avg': 0.15676797926425934, + 'learning_rate': 1.9317493062131702e-06, + 'epoch': 10.59} +04/20 [02:38:44] INFO | >> train_qwenlatent.py:487 + Step 41990 | grad_norm_pre_clip=0.0949 | + grad_norm_pre_clip_avg=0.1577 | Metrics: + {'align_loss': 0.024976912885904312, + 'recon_loss': 0.10453411936759949, + 'predict_loss': 0.003570789936929941, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09487176686525345, + 'data_time': 0.0007362029864452779, + 'model_time': 1.1876372359984089, + 'grad_norm_pre_clip_avg': 0.1576502315700054, + 'learning_rate': 1.9271217731375262e-06, + 'epoch': 10.6} +04/20 [02:38:57] INFO | >> train_qwenlatent.py:487 + Step 42000 | grad_norm_pre_clip=0.1946 | + grad_norm_pre_clip_avg=0.1619 | Metrics: + {'align_loss': 0.025795597583055496, + 'recon_loss': 0.18062596023082733, + 'predict_loss': 0.007404593750834465, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19458003342151642, + 'mae_score': 0.004965286426716023, 'data_time': + 0.000943741004448384, 'model_time': + 1.230116649006959, 'grad_norm_pre_clip_avg': + 0.1618716984987259, 'learning_rate': + 1.9224993992466313e-06, 'epoch': 10.6} +04/20 [02:39:10] INFO | >> train_qwenlatent.py:487 + Step 42010 | grad_norm_pre_clip=0.1707 | + grad_norm_pre_clip_avg=0.1812 | Metrics: + {'align_loss': 0.024855010211467743, + 'recon_loss': 0.08526204526424408, + 'predict_loss': 0.003074102336540818, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17069008946418762, + 'data_time': 0.0008796270121820271, + 'model_time': 1.2351030209974851, + 'grad_norm_pre_clip_avg': 0.18117171823978423, + 'learning_rate': 1.917882186793383e-06, + 'epoch': 10.6} +04/20 [02:39:22] INFO | >> train_qwenlatent.py:487 + Step 42020 | grad_norm_pre_clip=0.1497 | + grad_norm_pre_clip_avg=0.1552 | Metrics: + {'align_loss': 0.02514648251235485, + 'recon_loss': 0.19672206044197083, + 'predict_loss': 0.008513116277754307, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1497081220149994, + 'data_time': 0.0006810310005676001, + 'model_time': 1.2361974639934488, + 'grad_norm_pre_clip_avg': 0.1551624596118927, + 'learning_rate': 1.9132701380281514e-06, + 'epoch': 10.6} +04/20 [02:39:35] INFO | >> train_qwenlatent.py:487 + Step 42030 | grad_norm_pre_clip=0.1570 | + grad_norm_pre_clip_avg=0.1342 | Metrics: + {'align_loss': 0.02510015293955803, + 'recon_loss': 0.13929419219493866, + 'predict_loss': 0.006335318088531494, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1570163071155548, + 'data_time': 0.0007284340099431574, + 'model_time': 1.2462049270106945, + 'grad_norm_pre_clip_avg': 0.13415224105119705, + 'learning_rate': 1.9086632551987947e-06, + 'epoch': 10.61} +04/20 [02:39:47] INFO | >> train_qwenlatent.py:487 + Step 42040 | grad_norm_pre_clip=0.1331 | + grad_norm_pre_clip_avg=0.1247 | Metrics: + {'align_loss': 0.025699451565742493, + 'recon_loss': 0.12807497382164001, + 'predict_loss': 0.008185486309230328, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13311703503131866, + 'data_time': 0.0009602079808246344, + 'model_time': 1.1915280430112034, + 'grad_norm_pre_clip_avg': 0.12471310496330261, + 'learning_rate': 1.9040615405506512e-06, + 'epoch': 10.61} +04/20 [02:40:00] INFO | >> train_qwenlatent.py:487 + Step 42050 | grad_norm_pre_clip=0.1068 | + grad_norm_pre_clip_avg=0.1347 | Metrics: + {'align_loss': 0.025762410834431648, + 'recon_loss': 0.1221834123134613, + 'predict_loss': 0.005233007948845625, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10679008811712265, + 'mae_score': 0.0064055992676331115, + 'data_time': 0.0006861910223960876, + 'model_time': 1.218705792009132, + 'grad_norm_pre_clip_avg': 0.13466410115361213, + 'learning_rate': 1.8994649963265387e-06, + 'epoch': 10.61} +04/20 [02:40:13] INFO | >> train_qwenlatent.py:487 + Step 42060 | grad_norm_pre_clip=0.1986 | + grad_norm_pre_clip_avg=0.1482 | Metrics: + {'align_loss': 0.025353431701660156, + 'recon_loss': 0.13469436764717102, + 'predict_loss': 0.007720001507550478, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19863206148147583, + 'data_time': 0.0009943069890141487, + 'model_time': 1.237927223002771, + 'grad_norm_pre_clip_avg': 0.14822278693318366, + 'learning_rate': 1.8948736247667549e-06, + 'epoch': 10.61} +04/20 [02:40:25] INFO | >> train_qwenlatent.py:487 + Step 42070 | grad_norm_pre_clip=0.1619 | + grad_norm_pre_clip_avg=0.1331 | Metrics: + {'align_loss': 0.024757541716098785, + 'recon_loss': 0.1449424922466278, + 'predict_loss': 0.006240452639758587, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16194207966327667, + 'data_time': 0.0009619519987609237, + 'model_time': 1.2163839430140797, + 'grad_norm_pre_clip_avg': 0.13310813903808594, + 'learning_rate': 1.8902874281090867e-06, + 'epoch': 10.62} +04/20 [02:40:38] INFO | >> train_qwenlatent.py:487 + Step 42080 | grad_norm_pre_clip=0.1608 | + grad_norm_pre_clip_avg=0.1645 | Metrics: + {'align_loss': 0.025422077625989914, + 'recon_loss': 0.06994976848363876, + 'predict_loss': 0.004978769458830357, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16077712178230286, + 'data_time': 0.0010649040050338954, + 'model_time': 1.5577145679853857, + 'grad_norm_pre_clip_avg': 0.164468976855278, + 'learning_rate': 1.885706408588787e-06, + 'epoch': 10.62} +04/20 [02:40:50] INFO | >> train_qwenlatent.py:487 + Step 42090 | grad_norm_pre_clip=0.1317 | + grad_norm_pre_clip_avg=0.1586 | Metrics: + {'align_loss': 0.025261318311095238, + 'recon_loss': 0.19618459045886993, + 'predict_loss': 0.008185663260519505, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1317429542541504, + 'data_time': 0.0006948199879843742, + 'model_time': 1.2704243060143199, + 'grad_norm_pre_clip_avg': 0.15855712145566941, + 'learning_rate': 1.8811305684385869e-06, + 'epoch': 10.62} +04/20 [02:41:04] INFO | >> train_qwenlatent.py:487 + Step 42100 | grad_norm_pre_clip=0.2414 | + grad_norm_pre_clip_avg=0.1606 | Metrics: + {'align_loss': 0.02590075135231018, + 'recon_loss': 0.11883383244276047, + 'predict_loss': 0.006609924603253603, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2413996309041977, + 'mae_score': 0.00580998927623302, 'data_time': + 0.0008979210106190294, 'model_time': + 1.242536055011442, 'grad_norm_pre_clip_avg': + 0.16064112782478332, 'learning_rate': + 1.876559909888692e-06, 'epoch': 10.62} +04/20 [02:41:16] INFO | >> train_qwenlatent.py:487 + Step 42110 | grad_norm_pre_clip=0.1576 | + grad_norm_pre_clip_avg=0.1740 | Metrics: + {'align_loss': 0.025450170040130615, + 'recon_loss': 0.11751507222652435, + 'predict_loss': 0.004348868504166603, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1575508415699005, + 'data_time': 0.0009496580169070512, + 'model_time': 1.2437661410076544, + 'grad_norm_pre_clip_avg': 0.17397832423448562, + 'learning_rate': 1.8719944351667955e-06, + 'epoch': 10.63} +04/20 [02:41:29] INFO | >> train_qwenlatent.py:487 + Step 42120 | grad_norm_pre_clip=0.2256 | + grad_norm_pre_clip_avg=0.1481 | Metrics: + {'align_loss': 0.02592785656452179, + 'recon_loss': 0.15506793558597565, + 'predict_loss': 0.012209339067339897, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22563089430332184, + 'data_time': 0.0007174010097514838, + 'model_time': 1.2461352779937442, + 'grad_norm_pre_clip_avg': 0.14811434894800185, + 'learning_rate': 1.8674341464980468e-06, + 'epoch': 10.63} +04/20 [02:41:42] INFO | >> train_qwenlatent.py:487 + Step 42130 | grad_norm_pre_clip=0.1640 | + grad_norm_pre_clip_avg=0.1683 | Metrics: + {'align_loss': 0.025911662727594376, + 'recon_loss': 0.13133127987384796, + 'predict_loss': 0.008465196006000042, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1639636605978012, + 'data_time': 0.0008071550109889358, + 'model_time': 1.2162009669991676, + 'grad_norm_pre_clip_avg': 0.1682998701930046, + 'learning_rate': 1.862879046105078e-06, + 'epoch': 10.63} +04/20 [02:41:55] INFO | >> train_qwenlatent.py:487 + Step 42140 | grad_norm_pre_clip=0.1548 | + grad_norm_pre_clip_avg=0.1331 | Metrics: + {'align_loss': 0.02356526255607605, + 'recon_loss': 0.08074507117271423, + 'predict_loss': 0.004531350918114185, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15482880175113678, + 'data_time': 0.0008019689994398504, + 'model_time': 1.2948619840026367, + 'grad_norm_pre_clip_avg': 0.1330985128879547, + 'learning_rate': 1.858329136207989e-06, + 'epoch': 10.63} +04/20 [02:42:08] INFO | >> train_qwenlatent.py:487 + Step 42150 | grad_norm_pre_clip=0.1411 | + grad_norm_pre_clip_avg=0.1431 | Metrics: + {'align_loss': 0.02501622959971428, + 'recon_loss': 0.15421544015407562, + 'predict_loss': 0.005102933384478092, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1410883069038391, + 'mae_score': 0.006525106687803526, 'data_time': + 0.0008373120217584074, 'model_time': + 1.2661535739898682, 'grad_norm_pre_clip_avg': + 0.1430930368602276, 'learning_rate': + 1.8537844190243472e-06, 'epoch': 10.64} +04/20 [02:42:20] INFO | >> train_qwenlatent.py:487 + Step 42160 | grad_norm_pre_clip=0.1571 | + grad_norm_pre_clip_avg=0.1595 | Metrics: + {'align_loss': 0.023977309465408325, + 'recon_loss': 0.09254340082406998, + 'predict_loss': 0.007397385314106941, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15714561939239502, + 'data_time': 0.0006871840159874409, + 'model_time': 1.2357645349984523, + 'grad_norm_pre_clip_avg': 0.1594609297811985, + 'learning_rate': 1.8492448967692002e-06, + 'epoch': 10.64} +04/20 [02:42:33] INFO | >> train_qwenlatent.py:487 + Step 42170 | grad_norm_pre_clip=0.1835 | + grad_norm_pre_clip_avg=0.1768 | Metrics: + {'align_loss': 0.025808997452259064, + 'recon_loss': 0.20670609176158905, + 'predict_loss': 0.010978158563375473, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18348349630832672, + 'data_time': 0.0010646879964042455, + 'model_time': 1.2836787769920193, + 'grad_norm_pre_clip_avg': 0.17682375758886337, + 'learning_rate': 1.8447105716550534e-06, + 'epoch': 10.64} +04/20 [02:42:45] INFO | >> train_qwenlatent.py:487 + Step 42180 | grad_norm_pre_clip=0.1739 | + grad_norm_pre_clip_avg=0.1432 | Metrics: + {'align_loss': 0.024983718991279602, + 'recon_loss': 0.13558295369148254, + 'predict_loss': 0.006912075914442539, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17389711737632751, + 'data_time': 0.0008270910184364766, + 'model_time': 1.2145493070129305, + 'grad_norm_pre_clip_avg': 0.14315365999937057, + 'learning_rate': 1.8401814458918777e-06, + 'epoch': 10.64} +04/20 [02:42:58] INFO | >> train_qwenlatent.py:487 + Step 42190 | grad_norm_pre_clip=0.1972 | + grad_norm_pre_clip_avg=0.1578 | Metrics: + {'align_loss': 0.025517083704471588, + 'recon_loss': 0.16591928899288177, + 'predict_loss': 0.007418118882924318, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19722312688827515, + 'data_time': 0.0012821489945054054, + 'model_time': 1.2861149569798727, + 'grad_norm_pre_clip_avg': 0.15781306549906732, + 'learning_rate': 1.8356575216871144e-06, + 'epoch': 10.65} +04/20 [02:43:11] INFO | >> train_qwenlatent.py:487 + Step 42200 | grad_norm_pre_clip=0.1299 | + grad_norm_pre_clip_avg=0.1658 | Metrics: + {'align_loss': 0.025400610640645027, + 'recon_loss': 0.08911510556936264, + 'predict_loss': 0.0047005354426801205, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12991195917129517, + 'mae_score': 0.006659094063011376, 'data_time': + 0.0006801690033171326, 'model_time': + 1.2498716289992444, 'grad_norm_pre_clip_avg': + 0.16576728224754333, 'learning_rate': + 1.8311388012456738e-06, 'epoch': 10.65} +04/20 [02:43:24] INFO | >> train_qwenlatent.py:487 + Step 42210 | grad_norm_pre_clip=0.1962 | + grad_norm_pre_clip_avg=0.1545 | Metrics: + {'align_loss': 0.025628995150327682, + 'recon_loss': 0.11684688925743103, + 'predict_loss': 0.006511042360216379, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19623559713363647, + 'data_time': 0.0009751129837241024, + 'model_time': 1.2083909869834315, + 'grad_norm_pre_clip_avg': 0.15453704372048377, + 'learning_rate': 1.8266252867699217e-06, + 'epoch': 10.65} +04/20 [02:43:37] INFO | >> train_qwenlatent.py:487 + Step 42220 | grad_norm_pre_clip=0.1598 | + grad_norm_pre_clip_avg=0.1553 | Metrics: + {'align_loss': 0.024476973339915276, + 'recon_loss': 0.1612473875284195, + 'predict_loss': 0.006767011247575283, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15981335937976837, + 'data_time': 0.000718869996489957, + 'model_time': 1.2252655089832842, + 'grad_norm_pre_clip_avg': 0.15529619827866553, + 'learning_rate': 1.8221169804596927e-06, + 'epoch': 10.65} +04/20 [02:43:49] INFO | >> train_qwenlatent.py:487 + Step 42230 | grad_norm_pre_clip=0.2477 | + grad_norm_pre_clip_avg=0.1616 | Metrics: + {'align_loss': 0.025014903396368027, + 'recon_loss': 0.12391800433397293, + 'predict_loss': 0.004219787195324898, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.24768170714378357, + 'data_time': 0.0008830260194372386, + 'model_time': 1.1951712619920727, + 'grad_norm_pre_clip_avg': 0.16155388876795768, + 'learning_rate': 1.8176138845122793e-06, + 'epoch': 10.66} +04/20 [02:44:02] INFO | >> train_qwenlatent.py:487 + Step 42240 | grad_norm_pre_clip=0.1615 | + grad_norm_pre_clip_avg=0.1642 | Metrics: + {'align_loss': 0.025163719430565834, + 'recon_loss': 0.08265145868062973, + 'predict_loss': 0.0032241027802228928, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16152912378311157, + 'data_time': 0.0006709609879180789, + 'model_time': 1.2100004439998884, + 'grad_norm_pre_clip_avg': 0.1641841247677803, + 'learning_rate': 1.8131160011224346e-06, + 'epoch': 10.66} +04/20 [02:44:15] INFO | >> train_qwenlatent.py:487 + Step 42250 | grad_norm_pre_clip=0.1364 | + grad_norm_pre_clip_avg=0.1799 | Metrics: + {'align_loss': 0.02445369027554989, + 'recon_loss': 0.07655597478151321, + 'predict_loss': 0.003279241034761071, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13641390204429626, + 'mae_score': 0.0064764534030948675, + 'data_time': 0.000984352984232828, + 'model_time': 1.2312239379971288, + 'grad_norm_pre_clip_avg': 0.17987682074308395, + 'learning_rate': 1.808623332482374e-06, + 'epoch': 10.66} +04/20 [02:44:27] INFO | >> train_qwenlatent.py:487 + Step 42260 | grad_norm_pre_clip=0.1529 | + grad_norm_pre_clip_avg=0.1509 | Metrics: + {'align_loss': 0.02533406764268875, + 'recon_loss': 0.09944824874401093, + 'predict_loss': 0.004154922906309366, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1528969258069992, + 'data_time': 0.0009542199841234833, + 'model_time': 1.2798295539978426, + 'grad_norm_pre_clip_avg': 0.15087747126817702, + 'learning_rate': 1.8041358807817696e-06, + 'epoch': 10.66} +04/20 [02:44:40] INFO | >> train_qwenlatent.py:487 + Step 42270 | grad_norm_pre_clip=0.1730 | + grad_norm_pre_clip_avg=0.1548 | Metrics: + {'align_loss': 0.025300931185483932, + 'recon_loss': 0.13317422568798065, + 'predict_loss': 0.004831061232835054, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1730021834373474, + 'data_time': 0.0007151580066420138, + 'model_time': 1.2602130899904296, + 'grad_norm_pre_clip_avg': 0.1547722764313221, + 'learning_rate': 1.7996536482077498e-06, + 'epoch': 10.67} +04/20 [02:44:53] INFO | >> train_qwenlatent.py:487 + Step 42280 | grad_norm_pre_clip=0.1123 | + grad_norm_pre_clip_avg=0.1332 | Metrics: + {'align_loss': 0.025232944637537003, + 'recon_loss': 0.11755621433258057, + 'predict_loss': 0.006420255172997713, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11232834309339523, + 'data_time': 0.0007629860192537308, + 'model_time': 1.2938604499795474, + 'grad_norm_pre_clip_avg': 0.13320024982094764, + 'learning_rate': 1.7951766369449005e-06, + 'epoch': 10.67} +04/20 [02:45:06] INFO | >> train_qwenlatent.py:487 + Step 42290 | grad_norm_pre_clip=0.1494 | + grad_norm_pre_clip_avg=0.1524 | Metrics: + {'align_loss': 0.023933328688144684, + 'recon_loss': 0.14725545048713684, + 'predict_loss': 0.0074821775779128075, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14941172301769257, + 'data_time': 0.0006883729947730899, + 'model_time': 1.3004040579835419, + 'grad_norm_pre_clip_avg': 0.15235214829444885, + 'learning_rate': 1.7907048491752596e-06, + 'epoch': 10.67} +04/20 [02:45:19] INFO | >> train_qwenlatent.py:487 + Step 42300 | grad_norm_pre_clip=0.1388 | + grad_norm_pre_clip_avg=0.1421 | Metrics: + {'align_loss': 0.025401286780834198, + 'recon_loss': 0.16944806277751923, + 'predict_loss': 0.00803266279399395, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13883456587791443, + 'mae_score': 0.005349703522415848, 'data_time': + 0.0009913219837471843, 'model_time': + 1.2363290200009942, 'grad_norm_pre_clip_avg': + 0.14211055636405945, 'learning_rate': + 1.786238287078328e-06, 'epoch': 10.67} +04/20 [02:45:31] INFO | >> train_qwenlatent.py:487 + Step 42310 | grad_norm_pre_clip=0.1492 | + grad_norm_pre_clip_avg=0.1535 | Metrics: + {'align_loss': 0.02518131211400032, + 'recon_loss': 0.1882116049528122, + 'predict_loss': 0.010399074293673038, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14923441410064697, + 'data_time': 0.0006699560035485774, + 'model_time': 1.2034406189923175, + 'grad_norm_pre_clip_avg': 0.15350044816732406, + 'learning_rate': 1.7817769528310518e-06, + 'epoch': 10.68} +04/20 [02:45:44] INFO | >> train_qwenlatent.py:487 + Step 42320 | grad_norm_pre_clip=0.1314 | + grad_norm_pre_clip_avg=0.1706 | Metrics: + {'align_loss': 0.024739012122154236, + 'recon_loss': 0.10065806657075882, + 'predict_loss': 0.004124313592910767, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13141240179538727, + 'data_time': 0.0010558099893387407, + 'model_time': 1.29434236700763, + 'grad_norm_pre_clip_avg': 0.17058038339018822, + 'learning_rate': 1.7773208486078312e-06, + 'epoch': 10.68} +04/20 [02:45:56] INFO | >> train_qwenlatent.py:487 + Step 42330 | grad_norm_pre_clip=0.1394 | + grad_norm_pre_clip_avg=0.1514 | Metrics: + {'align_loss': 0.025961091741919518, + 'recon_loss': 0.15450210869312286, + 'predict_loss': 0.005703291390091181, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13939818739891052, + 'data_time': 0.0010950400028377771, + 'model_time': 1.2489103069819976, + 'grad_norm_pre_clip_avg': 0.15141289606690406, + 'learning_rate': 1.7728699765805157e-06, + 'epoch': 10.68} +04/20 [02:46:09] INFO | >> train_qwenlatent.py:487 + Step 42340 | grad_norm_pre_clip=0.1491 | + grad_norm_pre_clip_avg=0.1474 | Metrics: + {'align_loss': 0.02571888267993927, + 'recon_loss': 0.11827101558446884, + 'predict_loss': 0.004846010357141495, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14912283420562744, + 'data_time': 0.0010547369893174618, + 'model_time': 1.2628643889911473, + 'grad_norm_pre_clip_avg': 0.1473620168864727, + 'learning_rate': 1.7684243389184063e-06, + 'epoch': 10.68} +04/20 [02:46:22] INFO | >> train_qwenlatent.py:487 + Step 42350 | grad_norm_pre_clip=0.1403 | + grad_norm_pre_clip_avg=0.1325 | Metrics: + {'align_loss': 0.024287328124046326, + 'recon_loss': 0.12677183747291565, + 'predict_loss': 0.007494725752621889, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14031316339969635, + 'mae_score': 0.005568113842525997, 'data_time': + 0.00105470500420779, 'model_time': + 1.3067522780038416, 'grad_norm_pre_clip_avg': + 0.13246107026934623, 'learning_rate': + 1.7639839377882553e-06, 'epoch': 10.69} +04/20 [02:46:35] INFO | >> train_qwenlatent.py:487 + Step 42360 | grad_norm_pre_clip=0.1732 | + grad_norm_pre_clip_avg=0.1337 | Metrics: + {'align_loss': 0.02568572387099266, + 'recon_loss': 0.13290098309516907, + 'predict_loss': 0.006062269676476717, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17317961156368256, + 'data_time': 0.0007600579992868006, + 'model_time': 1.2523990639892872, + 'grad_norm_pre_clip_avg': 0.13373300582170486, + 'learning_rate': 1.7595487753542572e-06, + 'epoch': 10.69} +04/20 [02:46:47] INFO | >> train_qwenlatent.py:487 + Step 42370 | grad_norm_pre_clip=0.1494 | + grad_norm_pre_clip_avg=0.1372 | Metrics: + {'align_loss': 0.027146341279149055, + 'recon_loss': 0.19156473875045776, + 'predict_loss': 0.009661807678639889, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14940600097179413, + 'data_time': 0.0007122880197130144, + 'model_time': 1.2665597960003652, + 'grad_norm_pre_clip_avg': 0.13724131286144256, + 'learning_rate': 1.7551188537780593e-06, + 'epoch': 10.69} +04/20 [02:47:00] INFO | >> train_qwenlatent.py:487 + Step 42380 | grad_norm_pre_clip=0.1341 | + grad_norm_pre_clip_avg=0.1270 | Metrics: + {'align_loss': 0.024742718786001205, + 'recon_loss': 0.1019979864358902, + 'predict_loss': 0.003919539041817188, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13410574197769165, + 'data_time': 0.0009186069946736097, + 'model_time': 1.5228425069944933, + 'grad_norm_pre_clip_avg': 0.12696124166250228, + 'learning_rate': 1.7506941752187481e-06, + 'epoch': 10.69} +04/20 [02:47:13] INFO | >> train_qwenlatent.py:487 + Step 42390 | grad_norm_pre_clip=0.1874 | + grad_norm_pre_clip_avg=0.1586 | Metrics: + {'align_loss': 0.02699941396713257, + 'recon_loss': 0.16176387667655945, + 'predict_loss': 0.0071044424548745155, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.187368705868721, + 'data_time': 0.0009062879835255444, + 'model_time': 1.2395109200151637, + 'grad_norm_pre_clip_avg': 0.1586269736289978, + 'learning_rate': 1.7462747418328586e-06, + 'epoch': 10.7} +04/20 [02:47:26] INFO | >> train_qwenlatent.py:487 + Step 42400 | grad_norm_pre_clip=0.1998 | + grad_norm_pre_clip_avg=0.1677 | Metrics: + {'align_loss': 0.026229387149214745, + 'recon_loss': 0.1707485169172287, + 'predict_loss': 0.01592658832669258, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19982318580150604, + 'mae_score': 0.006630926304035358, 'data_time': + 0.0009231709991581738, 'model_time': + 1.2204291129892226, 'grad_norm_pre_clip_avg': + 0.16769000887870789, 'learning_rate': + 1.7418605557743723e-06, 'epoch': 10.7} +04/20 [02:47:38] INFO | >> train_qwenlatent.py:487 + Step 42410 | grad_norm_pre_clip=0.1228 | + grad_norm_pre_clip_avg=0.1522 | Metrics: + {'align_loss': 0.02493411675095558, + 'recon_loss': 0.11993741244077682, + 'predict_loss': 0.006197529379278421, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12283182889223099, + 'data_time': 0.0009672260202933103, + 'model_time': 1.2575156470120419, + 'grad_norm_pre_clip_avg': 0.15224635526537894, + 'learning_rate': 1.737451619194708e-06, + 'epoch': 10.7} +04/20 [02:47:51] INFO | >> train_qwenlatent.py:487 + Step 42420 | grad_norm_pre_clip=0.1809 | + grad_norm_pre_clip_avg=0.1421 | Metrics: + {'align_loss': 0.024626024067401886, + 'recon_loss': 0.16770602762699127, + 'predict_loss': 0.014154880307614803, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18091970682144165, + 'data_time': 0.0006395890086423606, + 'model_time': 1.4511958060029428, + 'grad_norm_pre_clip_avg': 0.14208640679717063, + 'learning_rate': 1.7330479342427285e-06, + 'epoch': 10.7} +04/20 [02:48:04] INFO | >> train_qwenlatent.py:487 + Step 42430 | grad_norm_pre_clip=0.1872 | + grad_norm_pre_clip_avg=0.1416 | Metrics: + {'align_loss': 0.026132578030228615, + 'recon_loss': 0.17665423452854156, + 'predict_loss': 0.010825610719621181, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1872045397758484, + 'data_time': 0.0006384350126609206, + 'model_time': 1.1640525569964666, + 'grad_norm_pre_clip_avg': 0.14157704859972, + 'learning_rate': 1.7286495030647347e-06, + 'epoch': 10.71} +04/20 [02:48:16] INFO | >> train_qwenlatent.py:487 + Step 42440 | grad_norm_pre_clip=0.1670 | + grad_norm_pre_clip_avg=0.1342 | Metrics: + {'align_loss': 0.024612270295619965, + 'recon_loss': 0.19285143911838531, + 'predict_loss': 0.009806353598833084, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16696254909038544, + 'data_time': 0.0006582869973499328, + 'model_time': 1.1898395760217682, + 'grad_norm_pre_clip_avg': 0.13418630212545396, + 'learning_rate': 1.7242563278044696e-06, + 'epoch': 10.71} +04/20 [02:48:28] INFO | >> train_qwenlatent.py:487 + Step 42450 | grad_norm_pre_clip=0.1471 | + grad_norm_pre_clip_avg=0.1520 | Metrics: + {'align_loss': 0.025917602702975273, + 'recon_loss': 0.12674745917320251, + 'predict_loss': 0.005377152003347874, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14714260399341583, + 'mae_score': 0.00514905173499305, 'data_time': + 0.00056436299928464, 'model_time': + 1.1616789310064632, 'grad_norm_pre_clip_avg': + 0.1519685111939907, 'learning_rate': + 1.719868410603113e-06, 'epoch': 10.71} +04/20 [02:48:39] INFO | >> train_qwenlatent.py:487 + Step 42460 | grad_norm_pre_clip=0.1934 | + grad_norm_pre_clip_avg=0.1605 | Metrics: + {'align_loss': 0.026048332452774048, + 'recon_loss': 0.15195108950138092, + 'predict_loss': 0.00892347190529108, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19336096942424774, + 'data_time': 0.0006728350126650184, + 'model_time': 1.1482025069999509, + 'grad_norm_pre_clip_avg': 0.1605358600616455, + 'learning_rate': 1.7154857535992843e-06, + 'epoch': 10.71} +04/20 [02:48:51] INFO | >> train_qwenlatent.py:487 + Step 42470 | grad_norm_pre_clip=0.1577 | + grad_norm_pre_clip_avg=0.1625 | Metrics: + {'align_loss': 0.025743689388036728, + 'recon_loss': 0.14645414054393768, + 'predict_loss': 0.008075307123363018, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15770016610622406, + 'data_time': 0.000681242992868647, + 'model_time': 1.1639784330036491, + 'grad_norm_pre_clip_avg': 0.16252171248197556, + 'learning_rate': 1.7111083589290336e-06, + 'epoch': 10.72} +04/20 [02:49:03] INFO | >> train_qwenlatent.py:487 + Step 42480 | grad_norm_pre_clip=0.1559 | + grad_norm_pre_clip_avg=0.1615 | Metrics: + {'align_loss': 0.025141865015029907, + 'recon_loss': 0.14784598350524902, + 'predict_loss': 0.007391295861452818, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1558942347764969, + 'data_time': 0.0005720780172850937, + 'model_time': 1.15211451498908, + 'grad_norm_pre_clip_avg': 0.16150224059820176, + 'learning_rate': 1.706736228725851e-06, + 'epoch': 10.72} +04/20 [02:49:14] INFO | >> train_qwenlatent.py:487 + Step 42490 | grad_norm_pre_clip=0.1109 | + grad_norm_pre_clip_avg=0.1404 | Metrics: + {'align_loss': 0.025508221238851547, + 'recon_loss': 0.10994990170001984, + 'predict_loss': 0.003183812601491809, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11085161566734314, + 'data_time': 0.0006165590020827949, + 'model_time': 1.145815194991883, + 'grad_norm_pre_clip_avg': 0.14043684005737306, + 'learning_rate': 1.7023693651206627e-06, + 'epoch': 10.72} +04/20 [02:49:27] INFO | >> train_qwenlatent.py:487 + Step 42500 | grad_norm_pre_clip=0.1697 | + grad_norm_pre_clip_avg=0.1777 | Metrics: + {'align_loss': 0.02569127455353737, + 'recon_loss': 0.10673445463180542, + 'predict_loss': 0.00433213310316205, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1696910262107849, + 'mae_score': 0.006431982109138558, 'data_time': + 0.00064728400320746, 'model_time': + 1.162609279010212, 'grad_norm_pre_clip_avg': + 0.1776971198618412, 'learning_rate': + 1.6980077702418219e-06, 'epoch': 10.72} +04/20 [02:49:38] INFO | >> train_qwenlatent.py:487 + Step 42510 | grad_norm_pre_clip=0.1292 | + grad_norm_pre_clip_avg=0.1395 | Metrics: + {'align_loss': 0.025940699502825737, + 'recon_loss': 0.1694086343050003, + 'predict_loss': 0.010188106447458267, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12918613851070404, + 'data_time': 0.0005947610188741237, + 'model_time': 1.141703459987184, + 'grad_norm_pre_clip_avg': 0.13952593728899956, + 'learning_rate': 1.6936514462151187e-06, + 'epoch': 10.73} +04/20 [02:49:50] INFO | >> train_qwenlatent.py:487 + Step 42520 | grad_norm_pre_clip=0.1667 | + grad_norm_pre_clip_avg=0.1371 | Metrics: + {'align_loss': 0.025734376162290573, + 'recon_loss': 0.15299883484840393, + 'predict_loss': 0.004041001666337252, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16665545105934143, + 'data_time': 0.000602779007749632, + 'model_time': 1.1489517189911567, + 'grad_norm_pre_clip_avg': 0.1370796613395214, + 'learning_rate': 1.6893003951637729e-06, + 'epoch': 10.73} +04/20 [02:50:01] INFO | >> train_qwenlatent.py:487 + Step 42530 | grad_norm_pre_clip=0.1486 | + grad_norm_pre_clip_avg=0.1590 | Metrics: + {'align_loss': 0.024588366970419884, + 'recon_loss': 0.16006247699260712, + 'predict_loss': 0.009024888277053833, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14861799776554108, + 'data_time': 0.0008507580205332488, + 'model_time': 1.1547973369888496, + 'grad_norm_pre_clip_avg': 0.1590307906270027, + 'learning_rate': 1.6849546192084307e-06, + 'epoch': 10.73} +04/20 [02:50:13] INFO | >> train_qwenlatent.py:487 + Step 42540 | grad_norm_pre_clip=0.1368 | + grad_norm_pre_clip_avg=0.1417 | Metrics: + {'align_loss': 0.026221901178359985, + 'recon_loss': 0.1561802625656128, + 'predict_loss': 0.008708587847650051, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13683494925498962, + 'data_time': 0.0006353520147968084, + 'model_time': 1.159104277001461, + 'grad_norm_pre_clip_avg': 0.14169678464531898, + 'learning_rate': 1.6806141204671743e-06, + 'epoch': 10.73} +04/20 [02:50:25] INFO | >> train_qwenlatent.py:487 + Step 42550 | grad_norm_pre_clip=0.1463 | + grad_norm_pre_clip_avg=0.1367 | Metrics: + {'align_loss': 0.025712691247463226, + 'recon_loss': 0.0779276192188263, + 'predict_loss': 0.0054572890512645245, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1462523639202118, + 'mae_score': 0.005805593353134018, 'data_time': + 0.0006408329936675727, 'model_time': + 1.1465542749792803, 'grad_norm_pre_clip_avg': + 0.13669347539544105, 'learning_rate': + 1.676278901055508e-06, 'epoch': 10.74} +04/20 [02:50:37] INFO | >> train_qwenlatent.py:487 + Step 42560 | grad_norm_pre_clip=0.1280 | + grad_norm_pre_clip_avg=0.1448 | Metrics: + {'align_loss': 0.024468662217259407, + 'recon_loss': 0.11007949709892273, + 'predict_loss': 0.007537093013525009, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1279781311750412, + 'data_time': 0.0005762359942309558, + 'model_time': 1.1412137959850952, + 'grad_norm_pre_clip_avg': 0.14480430111289025, + 'learning_rate': 1.671948963086364e-06, + 'epoch': 10.74} +04/20 [02:50:49] INFO | >> train_qwenlatent.py:487 + Step 42570 | grad_norm_pre_clip=0.1025 | + grad_norm_pre_clip_avg=0.1611 | Metrics: + {'align_loss': 0.025629859417676926, + 'recon_loss': 0.09581969678401947, + 'predict_loss': 0.004811394494026899, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1024760827422142, + 'data_time': 0.0005960359994787723, + 'model_time': 1.1841988030064385, + 'grad_norm_pre_clip_avg': 0.16108929440379144, + 'learning_rate': 1.6676243086701043e-06, + 'epoch': 10.74} +04/20 [02:51:00] INFO | >> train_qwenlatent.py:487 + Step 42580 | grad_norm_pre_clip=0.1448 | + grad_norm_pre_clip_avg=0.1497 | Metrics: + {'align_loss': 0.025716830044984818, + 'recon_loss': 0.11908549070358276, + 'predict_loss': 0.00560918590053916, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14477111399173737, + 'data_time': 0.0006422080041375011, + 'model_time': 1.1546587199845817, + 'grad_norm_pre_clip_avg': 0.14970167726278305, + 'learning_rate': 1.6633049399145083e-06, + 'epoch': 10.74} +04/20 [02:51:12] INFO | >> train_qwenlatent.py:487 + Step 42590 | grad_norm_pre_clip=0.1332 | + grad_norm_pre_clip_avg=0.1286 | Metrics: + {'align_loss': 0.024693768471479416, + 'recon_loss': 0.12405269593000412, + 'predict_loss': 0.007520060520619154, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1331530213356018, + 'data_time': 0.0005801859952043742, + 'model_time': 1.1446297120128293, + 'grad_norm_pre_clip_avg': 0.12863759696483612, + 'learning_rate': 1.6589908589247906e-06, + 'epoch': 10.75} +04/20 [02:51:24] INFO | >> train_qwenlatent.py:487 + Step 42600 | grad_norm_pre_clip=0.1864 | + grad_norm_pre_clip_avg=0.1463 | Metrics: + {'align_loss': 0.025191180408000946, + 'recon_loss': 0.16973236203193665, + 'predict_loss': 0.009796351194381714, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18643280863761902, + 'mae_score': 0.0062277514655310826, + 'data_time': 0.0005310089909471571, + 'model_time': 1.1374348709941842, + 'grad_norm_pre_clip_avg': 0.1463341511785984, + 'learning_rate': 1.6546820678035777e-06, + 'epoch': 10.75} +04/20 [02:51:36] INFO | >> train_qwenlatent.py:487 + Step 42610 | grad_norm_pre_clip=0.1100 | + grad_norm_pre_clip_avg=0.1361 | Metrics: + {'align_loss': 0.025122638791799545, + 'recon_loss': 0.15180574357509613, + 'predict_loss': 0.007065287791192532, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11000251024961472, + 'data_time': 0.000536303996341303, + 'model_time': 1.161623710009735, + 'grad_norm_pre_clip_avg': 0.13605060875415803, + 'learning_rate': 1.6503785686509252e-06, + 'epoch': 10.75} +04/20 [02:51:47] INFO | >> train_qwenlatent.py:487 + Step 42620 | grad_norm_pre_clip=0.1318 | + grad_norm_pre_clip_avg=0.1468 | Metrics: + {'align_loss': 0.026539403945207596, + 'recon_loss': 0.16287864744663239, + 'predict_loss': 0.01196608692407608, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13181746006011963, + 'data_time': 0.0006043920002412051, + 'model_time': 1.1606272069911938, + 'grad_norm_pre_clip_avg': 0.14682160019874574, + 'learning_rate': 1.646080363564298e-06, + 'epoch': 10.75} +04/20 [02:51:59] INFO | >> train_qwenlatent.py:487 + Step 42630 | grad_norm_pre_clip=0.0828 | + grad_norm_pre_clip_avg=0.1420 | Metrics: + {'align_loss': 0.026278596371412277, + 'recon_loss': 0.12028080224990845, + 'predict_loss': 0.005503813736140728, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.0828244686126709, + 'data_time': 0.0008430670131929219, + 'model_time': 1.1443895599804819, + 'grad_norm_pre_clip_avg': 0.14195723980665206, + 'learning_rate': 1.6417874546385984e-06, + 'epoch': 10.76} +04/20 [02:52:11] INFO | >> train_qwenlatent.py:487 + Step 42640 | grad_norm_pre_clip=0.1556 | + grad_norm_pre_clip_avg=0.1335 | Metrics: + {'align_loss': 0.0246488805860281, + 'recon_loss': 0.13506583869457245, + 'predict_loss': 0.0070542567409574986, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1555723249912262, + 'data_time': 0.0006018049898557365, + 'model_time': 1.1700895190006122, + 'grad_norm_pre_clip_avg': 0.13349971324205398, + 'learning_rate': 1.6374998439661344e-06, + 'epoch': 10.76} +04/20 [02:52:23] INFO | >> train_qwenlatent.py:487 + Step 42650 | grad_norm_pre_clip=0.1103 | + grad_norm_pre_clip_avg=0.1646 | Metrics: + {'align_loss': 0.025633275508880615, + 'recon_loss': 0.10521839559078217, + 'predict_loss': 0.0038382229395210743, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11032181233167648, + 'mae_score': 0.005872568783459362, 'data_time': + 0.0005620169977191836, 'model_time': + 1.1609578570059966, 'grad_norm_pre_clip_avg': + 0.16455454379320145, 'learning_rate': + 1.6332175336366353e-06, 'epoch': 10.76} +04/20 [02:52:34] INFO | >> train_qwenlatent.py:487 + Step 42660 | grad_norm_pre_clip=0.1441 | + grad_norm_pre_clip_avg=0.1478 | Metrics: + {'align_loss': 0.02506561577320099, + 'recon_loss': 0.1395634114742279, + 'predict_loss': 0.004615986254066229, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1441311240196228, + 'data_time': 0.0005527240282390267, + 'model_time': 1.1412340530077927, + 'grad_norm_pre_clip_avg': 0.14780781120061875, + 'learning_rate': 1.6289405257372479e-06, + 'epoch': 10.76} +04/20 [02:52:46] INFO | >> train_qwenlatent.py:487 + Step 42670 | grad_norm_pre_clip=0.1265 | + grad_norm_pre_clip_avg=0.1504 | Metrics: + {'align_loss': 0.023496897891163826, + 'recon_loss': 0.09572800993919373, + 'predict_loss': 0.007617529481649399, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1265004426240921, + 'data_time': 0.0005701930203940719, + 'model_time': 1.1409409269981552, + 'grad_norm_pre_clip_avg': 0.15038613975048065, + 'learning_rate': 1.6246688223525295e-06, + 'epoch': 10.77} +04/20 [02:52:58] INFO | >> train_qwenlatent.py:487 + Step 42680 | grad_norm_pre_clip=0.1471 | + grad_norm_pre_clip_avg=0.1391 | Metrics: + {'align_loss': 0.025401975959539413, + 'recon_loss': 0.16646279394626617, + 'predict_loss': 0.0041865091770887375, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1470554918050766, + 'data_time': 0.0005466410075314343, + 'model_time': 1.1533700069994666, + 'grad_norm_pre_clip_avg': 0.13910651952028275, + 'learning_rate': 1.6204024255644642e-06, + 'epoch': 10.77} +04/20 [02:53:09] INFO | >> train_qwenlatent.py:487 + Step 42690 | grad_norm_pre_clip=0.1803 | + grad_norm_pre_clip_avg=0.1505 | Metrics: + {'align_loss': 0.026465389877557755, + 'recon_loss': 0.18180856108665466, + 'predict_loss': 0.006947134155780077, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18027828633785248, + 'data_time': 0.0005471309996210039, + 'model_time': 1.1533753879775759, + 'grad_norm_pre_clip_avg': 0.1504841722548008, + 'learning_rate': 1.6161413374524363e-06, + 'epoch': 10.77} +04/20 [02:53:21] INFO | >> train_qwenlatent.py:487 + Step 42700 | grad_norm_pre_clip=0.1446 | + grad_norm_pre_clip_avg=0.1743 | Metrics: + {'align_loss': 0.022981686517596245, + 'recon_loss': 0.1852315366268158, + 'predict_loss': 0.009883943013846874, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.144593745470047, + 'mae_score': 0.006301019428012608, 'data_time': + 0.0005803520034532994, 'model_time': + 1.155768062977586, 'grad_norm_pre_clip_avg': + 0.17425722926855086, 'learning_rate': + 1.6118855600932536e-06, 'epoch': 10.77} +04/20 [02:53:33] INFO | >> train_qwenlatent.py:487 + Step 42710 | grad_norm_pre_clip=0.1163 | + grad_norm_pre_clip_avg=0.1514 | Metrics: + {'align_loss': 0.02678542770445347, + 'recon_loss': 0.12871775031089783, + 'predict_loss': 0.004819194320589304, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11626937985420227, + 'data_time': 0.0006140979821793735, + 'model_time': 1.3531928299926221, + 'grad_norm_pre_clip_avg': 0.15137373730540277, + 'learning_rate': 1.607635095561125e-06, + 'epoch': 10.78} +04/20 [02:53:45] INFO | >> train_qwenlatent.py:487 + Step 42720 | grad_norm_pre_clip=0.1213 | + grad_norm_pre_clip_avg=0.1508 | Metrics: + {'align_loss': 0.025818340480327606, + 'recon_loss': 0.16067568957805634, + 'predict_loss': 0.005345875862985849, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12133704870939255, + 'data_time': 0.000548301002709195, + 'model_time': 1.190271750005195, + 'grad_norm_pre_clip_avg': 0.1507686235010624, + 'learning_rate': 1.6033899459276733e-06, + 'epoch': 10.78} +04/20 [02:53:57] INFO | >> train_qwenlatent.py:487 + Step 42730 | grad_norm_pre_clip=0.1601 | + grad_norm_pre_clip_avg=0.1436 | Metrics: + {'align_loss': 0.02511458471417427, + 'recon_loss': 0.12286209315061569, + 'predict_loss': 0.0084508266299963, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1600785255432129, + 'data_time': 0.0005760209751315415, + 'model_time': 1.150361864012666, + 'grad_norm_pre_clip_avg': 0.14364189505577088, + 'learning_rate': 1.599150113261939e-06, + 'epoch': 10.78} +04/20 [02:54:32] INFO | >> train_qwenlatent.py:487 + Step 42740 | grad_norm_pre_clip=0.1408 | + grad_norm_pre_clip_avg=0.1415 | Metrics: + {'align_loss': 0.02574729546904564, + 'recon_loss': 0.18114690482616425, + 'predict_loss': 0.006804715842008591, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1408396065235138, + 'data_time': 8.723355154972523, 'model_time': + 16.412564594997093, 'grad_norm_pre_clip_avg': + 0.14147868305444716, 'learning_rate': + 1.594915599630363e-06, 'epoch': 10.78} +04/20 [02:55:09] INFO | >> train_qwenlatent.py:487 + Step 42750 | grad_norm_pre_clip=0.1401 | + grad_norm_pre_clip_avg=0.1700 | Metrics: + {'align_loss': 0.025551525875926018, + 'recon_loss': 0.13926605880260468, + 'predict_loss': 0.0072205811738967896, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14014309644699097, + 'mae_score': 0.005569869118767816, 'data_time': + 0.001113556994823739, 'model_time': + 3.6353520909906365, 'grad_norm_pre_clip_avg': + 0.16997128576040268, 'learning_rate': + 1.5906864070967958e-06, 'epoch': 10.79} +04/20 [02:55:45] INFO | >> train_qwenlatent.py:487 + Step 42760 | grad_norm_pre_clip=0.1774 | + grad_norm_pre_clip_avg=0.1388 | Metrics: + {'align_loss': 0.024752598255872726, + 'recon_loss': 0.11583639681339264, + 'predict_loss': 0.0041838218457996845, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17735669016838074, + 'data_time': 0.0012280209921300411, + 'model_time': 3.9862514859996736, + 'grad_norm_pre_clip_avg': 0.13875878527760505, + 'learning_rate': 1.5864625377224947e-06, + 'epoch': 10.79} +04/20 [02:56:21] INFO | >> train_qwenlatent.py:487 + Step 42770 | grad_norm_pre_clip=0.1240 | + grad_norm_pre_clip_avg=0.1413 | Metrics: + {'align_loss': 0.0256001278758049, + 'recon_loss': 0.09929744154214859, + 'predict_loss': 0.0031044548377394676, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12400796264410019, + 'data_time': 0.0011488069721963257, + 'model_time': 2.947109855012968, + 'grad_norm_pre_clip_avg': 0.14130945056676864, + 'learning_rate': 1.5822439935661171e-06, + 'epoch': 10.79} +04/20 [02:56:56] INFO | >> train_qwenlatent.py:487 + Step 42780 | grad_norm_pre_clip=0.1362 | + grad_norm_pre_clip_avg=0.1447 | Metrics: + {'align_loss': 0.025576140731573105, + 'recon_loss': 0.1726236343383789, + 'predict_loss': 0.006383872590959072, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13615818321704865, + 'data_time': 0.009834623982897028, + 'model_time': 3.7046380569809116, + 'grad_norm_pre_clip_avg': 0.1447409637272358, + 'learning_rate': 1.5780307766837394e-06, + 'epoch': 10.79} +04/20 [02:57:31] INFO | >> train_qwenlatent.py:487 + Step 42790 | grad_norm_pre_clip=0.1613 | + grad_norm_pre_clip_avg=0.1492 | Metrics: + {'align_loss': 0.025554513558745384, + 'recon_loss': 0.16998228430747986, + 'predict_loss': 0.009610798209905624, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1613406091928482, + 'data_time': 0.0011008020082954317, + 'model_time': 4.131895811005961, + 'grad_norm_pre_clip_avg': 0.1492488533258438, + 'learning_rate': 1.5738228891288295e-06, + 'epoch': 10.8} +04/20 [02:58:07] INFO | >> train_qwenlatent.py:487 + Step 42800 | grad_norm_pre_clip=0.1315 | + grad_norm_pre_clip_avg=0.1337 | Metrics: + {'align_loss': 0.025642510503530502, + 'recon_loss': 0.1679634153842926, + 'predict_loss': 0.008471456356346607, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13152077794075012, + 'mae_score': 0.006502990464906435, 'data_time': + 0.0013012110139243305, 'model_time': + 3.0413220660120714, 'grad_norm_pre_clip_avg': + 0.13365414813160897, 'learning_rate': + 1.569620332952256e-06, 'epoch': 10.8} +04/20 [02:58:33] INFO | >> train_qwenlatent.py:487 + Step 42810 | grad_norm_pre_clip=0.1481 | + grad_norm_pre_clip_avg=0.1356 | Metrics: + {'align_loss': 0.024723518639802933, + 'recon_loss': 0.0962202325463295, + 'predict_loss': 0.0089806467294693, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14812952280044556, + 'data_time': 0.0011506490118335932, + 'model_time': 2.2197284720023163, + 'grad_norm_pre_clip_avg': 0.13559979274868966, + 'learning_rate': 1.5654231102022933e-06, + 'epoch': 10.8} +04/20 [02:58:55] INFO | >> train_qwenlatent.py:487 + Step 42820 | grad_norm_pre_clip=0.1623 | + grad_norm_pre_clip_avg=0.1360 | Metrics: + {'align_loss': 0.025931740179657936, + 'recon_loss': 0.18007510900497437, + 'predict_loss': 0.007196313235908747, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16232816874980927, + 'data_time': 0.0011452350008767098, + 'model_time': 1.9802556930226274, + 'grad_norm_pre_clip_avg': 0.13602988049387932, + 'learning_rate': 1.5612312229246245e-06, + 'epoch': 10.8} +04/20 [02:59:11] INFO | >> train_qwenlatent.py:487 + Step 42830 | grad_norm_pre_clip=0.1220 | + grad_norm_pre_clip_avg=0.1336 | Metrics: + {'align_loss': 0.026191309094429016, + 'recon_loss': 0.1258704662322998, + 'predict_loss': 0.003649947000667453, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12202680110931396, + 'data_time': 0.0011970640043728054, + 'model_time': 1.3089752570085693, + 'grad_norm_pre_clip_avg': 0.13364671170711517, + 'learning_rate': 1.5570446731623187e-06, + 'epoch': 10.81} +04/20 [02:59:24] INFO | >> train_qwenlatent.py:487 + Step 42840 | grad_norm_pre_clip=0.1678 | + grad_norm_pre_clip_avg=0.1535 | Metrics: + {'align_loss': 0.02566022425889969, + 'recon_loss': 0.19617372751235962, + 'predict_loss': 0.009344082325696945, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16777853667736053, + 'data_time': 0.0008502739947289228, + 'model_time': 1.2720751620072406, + 'grad_norm_pre_clip_avg': 0.15345493257045745, + 'learning_rate': 1.552863462955852e-06, + 'epoch': 10.81} +04/20 [02:59:38] INFO | >> train_qwenlatent.py:487 + Step 42850 | grad_norm_pre_clip=0.1567 | + grad_norm_pre_clip_avg=0.1440 | Metrics: + {'align_loss': 0.024947848170995712, + 'recon_loss': 0.09441009163856506, + 'predict_loss': 0.006512574385851622, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15670248866081238, + 'mae_score': 0.004774984153541359, 'data_time': + 0.0006204379897098988, 'model_time': + 1.2370263999910094, 'grad_norm_pre_clip_avg': + 0.1439696490764618, 'learning_rate': + 1.5486875943430932e-06, 'epoch': 10.81} +04/20 [02:59:50] INFO | >> train_qwenlatent.py:487 + Step 42860 | grad_norm_pre_clip=0.1778 | + grad_norm_pre_clip_avg=0.1581 | Metrics: + {'align_loss': 0.025842219591140747, + 'recon_loss': 0.13335345685482025, + 'predict_loss': 0.006272256840020418, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1778009980916977, + 'data_time': 0.0008177949930541217, + 'model_time': 1.2362817420216743, + 'grad_norm_pre_clip_avg': 0.15814331769943238, + 'learning_rate': 1.5445170693593113e-06, + 'epoch': 10.82} +04/20 [03:00:03] INFO | >> train_qwenlatent.py:487 + Step 42870 | grad_norm_pre_clip=0.1054 | + grad_norm_pre_clip_avg=0.1587 | Metrics: + {'align_loss': 0.026747945696115494, + 'recon_loss': 0.13748422265052795, + 'predict_loss': 0.004953939933329821, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10536414384841919, + 'data_time': 0.0006212470179889351, + 'model_time': 1.2139021409966517, + 'grad_norm_pre_clip_avg': 0.15867115333676338, + 'learning_rate': 1.5403518900371694e-06, + 'epoch': 10.82} +04/20 [03:00:15] INFO | >> train_qwenlatent.py:487 + Step 42880 | grad_norm_pre_clip=0.1463 | + grad_norm_pre_clip_avg=0.1676 | Metrics: + {'align_loss': 0.023050308227539062, + 'recon_loss': 0.15773768723011017, + 'predict_loss': 0.0076993308030068874, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14629840850830078, + 'data_time': 0.0010383439948782325, + 'model_time': 1.2607702869863715, + 'grad_norm_pre_clip_avg': 0.16761564165353776, + 'learning_rate': 1.5361920584067264e-06, + 'epoch': 10.82} +04/20 [03:00:28] INFO | >> train_qwenlatent.py:487 + Step 42890 | grad_norm_pre_clip=0.1198 | + grad_norm_pre_clip_avg=0.1433 | Metrics: + {'align_loss': 0.02422722429037094, + 'recon_loss': 0.10798289626836777, + 'predict_loss': 0.004345359746366739, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1197705864906311, + 'data_time': 0.0005959520058240741, + 'model_time': 1.223190730001079, + 'grad_norm_pre_clip_avg': 0.1432594947516918, + 'learning_rate': 1.5320375764954307e-06, + 'epoch': 10.82} +04/20 [03:00:41] INFO | >> train_qwenlatent.py:487 + Step 42900 | grad_norm_pre_clip=0.1684 | + grad_norm_pre_clip_avg=0.1454 | Metrics: + {'align_loss': 0.025167282670736313, + 'recon_loss': 0.133204847574234, + 'predict_loss': 0.0104638347402215, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1683656871318817, + 'mae_score': 0.005205593452797279, 'data_time': + 0.0008307740208692849, 'model_time': + 1.2019796310050879, 'grad_norm_pre_clip_avg': + 0.14537157639861106, 'learning_rate': + 1.5278884463281283e-06, 'epoch': 10.83} +04/20 [03:00:53] INFO | >> train_qwenlatent.py:487 + Step 42910 | grad_norm_pre_clip=0.1625 | + grad_norm_pre_clip_avg=0.1545 | Metrics: + {'align_loss': 0.02570609748363495, + 'recon_loss': 0.20351868867874146, + 'predict_loss': 0.010633152909576893, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16253378987312317, + 'data_time': 0.0007334249967243522, + 'model_time': 1.2203886760107707, + 'grad_norm_pre_clip_avg': 0.15452291443943977, + 'learning_rate': 1.5237446699270528e-06, + 'epoch': 10.83} +04/20 [03:01:06] INFO | >> train_qwenlatent.py:487 + Step 42920 | grad_norm_pre_clip=0.1304 | + grad_norm_pre_clip_avg=0.1642 | Metrics: + {'align_loss': 0.02568582445383072, + 'recon_loss': 0.11676212400197983, + 'predict_loss': 0.006462093908339739, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13039475679397583, + 'data_time': 0.0006250460282899439, + 'model_time': 1.2672718380053993, + 'grad_norm_pre_clip_avg': 0.16418126076459885, + 'learning_rate': 1.519606249311834e-06, + 'epoch': 10.83} +04/20 [03:01:19] INFO | >> train_qwenlatent.py:487 + Step 42930 | grad_norm_pre_clip=0.1013 | + grad_norm_pre_clip_avg=0.1526 | Metrics: + {'align_loss': 0.02604767680168152, + 'recon_loss': 0.11827786266803741, + 'predict_loss': 0.002943145576864481, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10125312209129333, + 'data_time': 0.0008352129952982068, + 'model_time': 1.2603065440198407, + 'grad_norm_pre_clip_avg': 0.1525634467601776, + 'learning_rate': 1.5154731864994871e-06, + 'epoch': 10.83} +04/20 [03:01:31] INFO | >> train_qwenlatent.py:487 + Step 42940 | grad_norm_pre_clip=0.0990 | + grad_norm_pre_clip_avg=0.1413 | Metrics: + {'align_loss': 0.024105405434966087, + 'recon_loss': 0.16149839758872986, + 'predict_loss': 0.0057234070263803005, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09896265715360641, + 'data_time': 0.0006194379820954055, + 'model_time': 1.2063655619858764, + 'grad_norm_pre_clip_avg': 0.141288061439991, + 'learning_rate': 1.5113454835044142e-06, + 'epoch': 10.84} +04/20 [03:01:44] INFO | >> train_qwenlatent.py:487 + Step 42950 | grad_norm_pre_clip=0.1265 | + grad_norm_pre_clip_avg=0.1280 | Metrics: + {'align_loss': 0.02543788030743599, + 'recon_loss': 0.15799443423748016, + 'predict_loss': 0.0059294127859175205, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12646687030792236, + 'mae_score': 0.005679244823283977, 'data_time': + 0.0008637800056021661, 'model_time': + 1.230209532979643, 'grad_norm_pre_clip_avg': + 0.12800388261675835, 'learning_rate': + 1.5072231423384104e-06, 'epoch': 10.84} +04/20 [03:01:57] INFO | >> train_qwenlatent.py:487 + Step 42960 | grad_norm_pre_clip=0.1724 | + grad_norm_pre_clip_avg=0.1560 | Metrics: + {'align_loss': 0.02479967661201954, + 'recon_loss': 0.1360655128955841, + 'predict_loss': 0.010788904502987862, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17236194014549255, + 'data_time': 0.0008759190095588565, + 'model_time': 1.3230785610212479, + 'grad_norm_pre_clip_avg': 0.1559528738260269, + 'learning_rate': 1.5031061650106526e-06, + 'epoch': 10.84} +04/20 [03:02:09] INFO | >> train_qwenlatent.py:487 + Step 42970 | grad_norm_pre_clip=0.1423 | + grad_norm_pre_clip_avg=0.1347 | Metrics: + {'align_loss': 0.025125833228230476, + 'recon_loss': 0.1272238790988922, + 'predict_loss': 0.0037592793814837933, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14227160811424255, + 'data_time': 0.0008671800023876131, + 'model_time': 1.1980091189907398, + 'grad_norm_pre_clip_avg': 0.13470797315239907, + 'learning_rate': 1.4989945535277079e-06, + 'epoch': 10.84} +04/20 [03:02:22] INFO | >> train_qwenlatent.py:487 + Step 42980 | grad_norm_pre_clip=0.1221 | + grad_norm_pre_clip_avg=0.1525 | Metrics: + {'align_loss': 0.024645688012242317, + 'recon_loss': 0.09258170425891876, + 'predict_loss': 0.006120019592344761, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12212125211954117, + 'data_time': 0.000936752010602504, + 'model_time': 1.2714038949925452, + 'grad_norm_pre_clip_avg': 0.1524684101343155, + 'learning_rate': 1.4948883098935223e-06, + 'epoch': 10.85} +04/20 [03:02:35] INFO | >> train_qwenlatent.py:487 + Step 42990 | grad_norm_pre_clip=0.2032 | + grad_norm_pre_clip_avg=0.1809 | Metrics: + {'align_loss': 0.02601988986134529, + 'recon_loss': 0.1900613158941269, + 'predict_loss': 0.014432229101657867, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20320412516593933, + 'data_time': 0.0008440189994871616, + 'model_time': 1.2098781199892983, + 'grad_norm_pre_clip_avg': 0.18093964755535125, + 'learning_rate': 1.490787436109432e-06, + 'epoch': 10.85} +04/20 [03:02:48] INFO | >> train_qwenlatent.py:487 + Step 43000 | grad_norm_pre_clip=0.1888 | + grad_norm_pre_clip_avg=0.1604 | Metrics: + {'align_loss': 0.02466375008225441, + 'recon_loss': 0.1284119188785553, + 'predict_loss': 0.007883966900408268, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18877704441547394, + 'mae_score': 0.005234678371532543, 'data_time': + 0.0007771720120217651, 'model_time': + 1.445191650011111, 'grad_norm_pre_clip_avg': + 0.16044487208127975, 'learning_rate': + 1.4866919341741492e-06, 'epoch': 10.85} +04/20 [03:03:00] INFO | >> train_qwenlatent.py:487 + Step 43010 | grad_norm_pre_clip=0.1448 | + grad_norm_pre_clip_avg=0.1313 | Metrics: + {'align_loss': 0.02394636534154415, + 'recon_loss': 0.1640349179506302, + 'predict_loss': 0.008171056397259235, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14483436942100525, + 'data_time': 0.0008128970221150666, + 'model_time': 1.2596119229856413, + 'grad_norm_pre_clip_avg': 0.13128922879695892, + 'learning_rate': 1.4826018060837781e-06, + 'epoch': 10.85} +04/20 [03:03:13] INFO | >> train_qwenlatent.py:487 + Step 43020 | grad_norm_pre_clip=0.1023 | + grad_norm_pre_clip_avg=0.1500 | Metrics: + {'align_loss': 0.026351647451519966, + 'recon_loss': 0.10815586894750595, + 'predict_loss': 0.00369231472723186, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10231073200702667, + 'data_time': 0.0010443709907121956, + 'model_time': 1.3346825730113778, + 'grad_norm_pre_clip_avg': 0.1499983251094818, + 'learning_rate': 1.4785170538317921e-06, + 'epoch': 10.86} +04/20 [03:03:25] INFO | >> train_qwenlatent.py:487 + Step 43030 | grad_norm_pre_clip=0.1703 | + grad_norm_pre_clip_avg=0.1479 | Metrics: + {'align_loss': 0.026438644155859947, + 'recon_loss': 0.18889643251895905, + 'predict_loss': 0.011566991917788982, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17032192647457123, + 'data_time': 0.000655798998195678, + 'model_time': 1.2376997109968215, + 'grad_norm_pre_clip_avg': 0.14785274863243103, + 'learning_rate': 1.4744376794090538e-06, + 'epoch': 10.86} +04/20 [03:03:38] INFO | >> train_qwenlatent.py:487 + Step 43040 | grad_norm_pre_clip=0.1288 | + grad_norm_pre_clip_avg=0.1400 | Metrics: + {'align_loss': 0.026926428079605103, + 'recon_loss': 0.22603315114974976, + 'predict_loss': 0.010350853204727173, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12879294157028198, + 'data_time': 0.0007873930153436959, + 'model_time': 1.264123650995316, + 'grad_norm_pre_clip_avg': 0.13998564034700395, + 'learning_rate': 1.470363684803797e-06, + 'epoch': 10.86} +04/20 [03:03:51] INFO | >> train_qwenlatent.py:487 + Step 43050 | grad_norm_pre_clip=0.2161 | + grad_norm_pre_clip_avg=0.1571 | Metrics: + {'align_loss': 0.02620858885347843, + 'recon_loss': 0.19135574996471405, + 'predict_loss': 0.009128334932029247, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21609818935394287, + 'mae_score': 0.008115722037650444, 'data_time': + 0.0006886089977342635, 'model_time': + 1.2011941480159294, 'grad_norm_pre_clip_avg': + 0.1571350671350956, 'learning_rate': + 1.4662950720016394e-06, 'epoch': 10.86} +04/20 [03:04:04] INFO | >> train_qwenlatent.py:487 + Step 43060 | grad_norm_pre_clip=0.1543 | + grad_norm_pre_clip_avg=0.1651 | Metrics: + {'align_loss': 0.02550039067864418, + 'recon_loss': 0.1331193596124649, + 'predict_loss': 0.005900365766137838, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1543302685022354, + 'data_time': 0.0006026840128470212, + 'model_time': 1.236454614001559, + 'grad_norm_pre_clip_avg': 0.16510530561208725, + 'learning_rate': 1.4622318429855737e-06, + 'epoch': 10.87} +04/20 [03:04:16] INFO | >> train_qwenlatent.py:487 + Step 43070 | grad_norm_pre_clip=0.1644 | + grad_norm_pre_clip_avg=0.1416 | Metrics: + {'align_loss': 0.02613087184727192, + 'recon_loss': 0.15339575707912445, + 'predict_loss': 0.0074338726699352264, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16442802548408508, + 'data_time': 0.0007602069817949086, + 'model_time': 1.2398390439921059, + 'grad_norm_pre_clip_avg': 0.14162514880299568, + 'learning_rate': 1.4581739997359665e-06, + 'epoch': 10.87} +04/20 [03:04:29] INFO | >> train_qwenlatent.py:487 + Step 43080 | grad_norm_pre_clip=0.1531 | + grad_norm_pre_clip_avg=0.1271 | Metrics: + {'align_loss': 0.02548389881849289, + 'recon_loss': 0.14804591238498688, + 'predict_loss': 0.009038235060870647, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15307064354419708, + 'data_time': 0.0008507449820172042, + 'model_time': 1.2502151980006602, + 'grad_norm_pre_clip_avg': 0.12705320119857788, + 'learning_rate': 1.454121544230563e-06, + 'epoch': 10.87} +04/20 [03:04:41] INFO | >> train_qwenlatent.py:487 + Step 43090 | grad_norm_pre_clip=0.0939 | + grad_norm_pre_clip_avg=0.1216 | Metrics: + {'align_loss': 0.02419433556497097, + 'recon_loss': 0.1299956738948822, + 'predict_loss': 0.005847624968737364, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09388280659914017, + 'data_time': 0.0006165810045786202, + 'model_time': 1.2468130070192274, + 'grad_norm_pre_clip_avg': 0.12158803194761277, + 'learning_rate': 1.4500744784444806e-06, + 'epoch': 10.87} +04/20 [03:04:54] INFO | >> train_qwenlatent.py:487 + Step 43100 | grad_norm_pre_clip=0.1588 | + grad_norm_pre_clip_avg=0.1554 | Metrics: + {'align_loss': 0.02669854462146759, + 'recon_loss': 0.14751507341861725, + 'predict_loss': 0.0036766629200428724, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15875555574893951, + 'mae_score': 0.00600666484317264, 'data_time': + 0.0008159130229614675, 'model_time': + 1.2005422190122772, 'grad_norm_pre_clip_avg': + 0.15535974875092506, 'learning_rate': + 1.4460328043502062e-06, 'epoch': 10.88} +04/20 [03:05:07] INFO | >> train_qwenlatent.py:487 + Step 43110 | grad_norm_pre_clip=0.1543 | + grad_norm_pre_clip_avg=0.1526 | Metrics: + {'align_loss': 0.025954343378543854, + 'recon_loss': 0.18448369204998016, + 'predict_loss': 0.005657155532389879, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15427760779857635, + 'data_time': 0.0010560029768384993, + 'model_time': 1.2402622920053545, + 'grad_norm_pre_clip_avg': 0.15264619886875153, + 'learning_rate': 1.4419965239176083e-06, + 'epoch': 10.88} +04/20 [03:05:19] INFO | >> train_qwenlatent.py:487 + Step 43120 | grad_norm_pre_clip=0.1415 | + grad_norm_pre_clip_avg=0.1465 | Metrics: + {'align_loss': 0.024775125086307526, + 'recon_loss': 0.18126991391181946, + 'predict_loss': 0.0064996411092579365, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14154347777366638, + 'data_time': 0.000757508008973673, + 'model_time': 1.2151344539888669, + 'grad_norm_pre_clip_avg': 0.14651632830500602, + 'learning_rate': 1.437965639113919e-06, + 'epoch': 10.88} +04/20 [03:05:32] INFO | >> train_qwenlatent.py:487 + Step 43130 | grad_norm_pre_clip=0.0937 | + grad_norm_pre_clip_avg=0.1541 | Metrics: + {'align_loss': 0.025113148614764214, + 'recon_loss': 0.17336709797382355, + 'predict_loss': 0.008567674085497856, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09374009817838669, + 'data_time': 0.00088193200645037, 'model_time': + 1.1958179769862909, 'grad_norm_pre_clip_avg': + 0.15408703908324242, 'learning_rate': + 1.4339401519037416e-06, 'epoch': 10.88} +04/20 [03:05:45] INFO | >> train_qwenlatent.py:487 + Step 43140 | grad_norm_pre_clip=0.1778 | + grad_norm_pre_clip_avg=0.1655 | Metrics: + {'align_loss': 0.02463843673467636, + 'recon_loss': 0.13118621706962585, + 'predict_loss': 0.008912126533687115, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1777823269367218, + 'data_time': 0.0007882159843575209, + 'model_time': 1.2206408619822469, + 'grad_norm_pre_clip_avg': 0.1654738537967205, + 'learning_rate': 1.4299200642490512e-06, + 'epoch': 10.89} +04/20 [03:05:58] INFO | >> train_qwenlatent.py:487 + Step 43150 | grad_norm_pre_clip=0.1707 | + grad_norm_pre_clip_avg=0.1805 | Metrics: + {'align_loss': 0.025707989931106567, + 'recon_loss': 0.16915220022201538, + 'predict_loss': 0.008273054845631123, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17074339091777802, + 'mae_score': 0.005716844936748883, 'data_time': + 0.0006914629775565118, 'model_time': + 1.199414993985556, 'grad_norm_pre_clip_avg': + 0.18047402203083038, 'learning_rate': + 1.4259053781091895e-06, 'epoch': 10.89} +04/20 [03:06:10] INFO | >> train_qwenlatent.py:487 + Step 43160 | grad_norm_pre_clip=0.1573 | + grad_norm_pre_clip_avg=0.1507 | Metrics: + {'align_loss': 0.025333240628242493, + 'recon_loss': 0.17284929752349854, + 'predict_loss': 0.01119988039135933, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15728649497032166, + 'data_time': 0.0009579049947205931, + 'model_time': 1.2462946849991567, + 'grad_norm_pre_clip_avg': 0.15068406388163566, + 'learning_rate': 1.4218960954408633e-06, + 'epoch': 10.89} +04/20 [03:06:23] INFO | >> train_qwenlatent.py:487 + Step 43170 | grad_norm_pre_clip=0.1222 | + grad_norm_pre_clip_avg=0.1454 | Metrics: + {'align_loss': 0.025126557797193527, + 'recon_loss': 0.13837681710720062, + 'predict_loss': 0.0094115836545825, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12222421169281006, + 'data_time': 0.0006787370075471699, + 'model_time': 1.2200435270206071, + 'grad_norm_pre_clip_avg': 0.14544499889016152, + 'learning_rate': 1.4178922181981519e-06, + 'epoch': 10.89} +04/20 [03:06:35] INFO | >> train_qwenlatent.py:487 + Step 43180 | grad_norm_pre_clip=0.1187 | + grad_norm_pre_clip_avg=0.1350 | Metrics: + {'align_loss': 0.024694133549928665, + 'recon_loss': 0.0808117613196373, + 'predict_loss': 0.0019357003038749099, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11871689558029175, + 'data_time': 0.0006460769800469279, + 'model_time': 1.2290154279908165, + 'grad_norm_pre_clip_avg': 0.1350404664874077, + 'learning_rate': 1.4138937483324932e-06, + 'epoch': 10.9} +04/20 [03:06:48] INFO | >> train_qwenlatent.py:487 + Step 43190 | grad_norm_pre_clip=0.1334 | + grad_norm_pre_clip_avg=0.1582 | Metrics: + {'align_loss': 0.024955663830041885, + 'recon_loss': 0.15082134306430817, + 'predict_loss': 0.004860434681177139, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1334066092967987, + 'data_time': 0.0010948689887300134, + 'model_time': 1.2335584459942766, + 'grad_norm_pre_clip_avg': 0.15816495791077614, + 'learning_rate': 1.4099006877926914e-06, + 'epoch': 10.9} +04/20 [03:07:02] INFO | >> train_qwenlatent.py:487 + Step 43200 | grad_norm_pre_clip=0.1336 | + grad_norm_pre_clip_avg=0.1450 | Metrics: + {'align_loss': 0.02666242979466915, + 'recon_loss': 0.18211250007152557, + 'predict_loss': 0.004279504530131817, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13356199860572815, + 'mae_score': 0.0044104365615157395, + 'data_time': 0.0009683259995654225, + 'model_time': 1.2176253510115203, + 'grad_norm_pre_clip_avg': 0.14499362483620643, + 'learning_rate': 1.4059130385249224e-06, + 'epoch': 10.9} +04/20 [03:07:14] INFO | >> train_qwenlatent.py:487 + Step 43210 | grad_norm_pre_clip=0.1507 | + grad_norm_pre_clip_avg=0.1472 | Metrics: + {'align_loss': 0.0261521078646183, + 'recon_loss': 0.1352265179157257, + 'predict_loss': 0.006085199303925037, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15071465075016022, + 'data_time': 0.0011077199887949973, + 'model_time': 1.252358819998335, + 'grad_norm_pre_clip_avg': 0.1471905715763569, + 'learning_rate': 1.401930802472713e-06, + 'epoch': 10.9} +04/20 [03:07:27] INFO | >> train_qwenlatent.py:487 + Step 43220 | grad_norm_pre_clip=0.1851 | + grad_norm_pre_clip_avg=0.1488 | Metrics: + {'align_loss': 0.026383770629763603, + 'recon_loss': 0.15470223128795624, + 'predict_loss': 0.008199081756174564, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18511000275611877, + 'data_time': 0.000732880987925455, + 'model_time': 1.2142654499912169, + 'grad_norm_pre_clip_avg': 0.14881023168563842, + 'learning_rate': 1.3979539815769567e-06, + 'epoch': 10.91} +04/20 [03:07:40] INFO | >> train_qwenlatent.py:487 + Step 43230 | grad_norm_pre_clip=0.1731 | + grad_norm_pre_clip_avg=0.1568 | Metrics: + {'align_loss': 0.025232849642634392, + 'recon_loss': 0.17752498388290405, + 'predict_loss': 0.009475894272327423, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17310965061187744, + 'data_time': 0.0006055490230210125, + 'model_time': 1.2078886489907745, + 'grad_norm_pre_clip_avg': 0.1567934475839138, + 'learning_rate': 1.3939825777759132e-06, + 'epoch': 10.91} +04/20 [03:07:52] INFO | >> train_qwenlatent.py:487 + Step 43240 | grad_norm_pre_clip=0.1355 | + grad_norm_pre_clip_avg=0.1491 | Metrics: + {'align_loss': 0.02602279558777809, + 'recon_loss': 0.13863542675971985, + 'predict_loss': 0.005843557417392731, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13552893698215485, + 'data_time': 0.0006162100180517882, + 'model_time': 1.301014762982959, + 'grad_norm_pre_clip_avg': 0.14906848296523095, + 'learning_rate': 1.3900165930051864e-06, + 'epoch': 10.91} +04/20 [03:08:05] INFO | >> train_qwenlatent.py:487 + Step 43250 | grad_norm_pre_clip=0.1362 | + grad_norm_pre_clip_avg=0.1407 | Metrics: + {'align_loss': 0.026363883167505264, + 'recon_loss': 0.09945647418498993, + 'predict_loss': 0.009015887975692749, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13624174892902374, + 'mae_score': 0.0057382832776318796, + 'data_time': 0.0012207809777464718, + 'model_time': 1.2571679130196571, + 'grad_norm_pre_clip_avg': 0.14067535549402238, + 'learning_rate': 1.3860560291977574e-06, + 'epoch': 10.91} +04/20 [03:08:18] INFO | >> train_qwenlatent.py:487 + Step 43260 | grad_norm_pre_clip=0.1322 | + grad_norm_pre_clip_avg=0.1490 | Metrics: + {'align_loss': 0.02533525973558426, + 'recon_loss': 0.15283437073230743, + 'predict_loss': 0.00928269699215889, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13221119344234467, + 'data_time': 0.0009949009981937706, + 'model_time': 1.241678058024263, + 'grad_norm_pre_clip_avg': 0.14900121167302133, + 'learning_rate': 1.382100888283954e-06, + 'epoch': 10.92} +04/20 [03:08:31] INFO | >> train_qwenlatent.py:487 + Step 43270 | grad_norm_pre_clip=0.1372 | + grad_norm_pre_clip_avg=0.1540 | Metrics: + {'align_loss': 0.025450561195611954, + 'recon_loss': 0.13302865624427795, + 'predict_loss': 0.005901880096644163, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13720960915088654, + 'data_time': 0.0008260030008386821, + 'model_time': 1.6383851279970258, + 'grad_norm_pre_clip_avg': 0.15403744280338288, + 'learning_rate': 1.3781511721914639e-06, + 'epoch': 10.92} +04/20 [03:08:43] INFO | >> train_qwenlatent.py:487 + Step 43280 | grad_norm_pre_clip=0.1272 | + grad_norm_pre_clip_avg=0.1347 | Metrics: + {'align_loss': 0.023439843207597733, + 'recon_loss': 0.12868930399417877, + 'predict_loss': 0.005297658499330282, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12724582850933075, + 'data_time': 0.0009054639958776534, + 'model_time': 1.2150513929955196, + 'grad_norm_pre_clip_avg': 0.1346690036356449, + 'learning_rate': 1.3742068828453303e-06, + 'epoch': 10.92} +04/20 [03:08:56] INFO | >> train_qwenlatent.py:487 + Step 43290 | grad_norm_pre_clip=0.1679 | + grad_norm_pre_clip_avg=0.1377 | Metrics: + {'align_loss': 0.02417772449553013, + 'recon_loss': 0.10708927363157272, + 'predict_loss': 0.006023553665727377, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16788558661937714, + 'data_time': 0.0007567830034531653, + 'model_time': 1.189723896997748, + 'grad_norm_pre_clip_avg': 0.13774138391017915, + 'learning_rate': 1.3702680221679517e-06, + 'epoch': 10.92} +04/20 [03:09:09] INFO | >> train_qwenlatent.py:487 + Step 43300 | grad_norm_pre_clip=0.1901 | + grad_norm_pre_clip_avg=0.1494 | Metrics: + {'align_loss': 0.026111852377653122, + 'recon_loss': 0.1276487410068512, + 'predict_loss': 0.003975600469857454, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19012820720672607, + 'mae_score': 0.004742283434481234, 'data_time': + 0.0008443089900538325, 'model_time': + 1.2446938990033232, 'grad_norm_pre_clip_avg': + 0.1493659295141697, 'learning_rate': + 1.3663345920790818e-06, 'epoch': 10.93} +04/20 [03:09:22] INFO | >> train_qwenlatent.py:487 + Step 43310 | grad_norm_pre_clip=0.1822 | + grad_norm_pre_clip_avg=0.1515 | Metrics: + {'align_loss': 0.026166623458266258, + 'recon_loss': 0.11383932083845139, + 'predict_loss': 0.0032368218526244164, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18222007155418396, + 'data_time': 0.0007330749940592796, + 'model_time': 1.25918619800359, + 'grad_norm_pre_clip_avg': 0.1515378922224045, + 'learning_rate': 1.362406594495829e-06, + 'epoch': 10.93} +04/20 [03:09:34] INFO | >> train_qwenlatent.py:487 + Step 43320 | grad_norm_pre_clip=0.1617 | + grad_norm_pre_clip_avg=0.1465 | Metrics: + {'align_loss': 0.024417199194431305, + 'recon_loss': 0.08804558217525482, + 'predict_loss': 0.004247559234499931, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16167281568050385, + 'data_time': 0.0007168719894252717, + 'model_time': 1.2436858210130595, + 'grad_norm_pre_clip_avg': 0.14645116850733758, + 'learning_rate': 1.3584840313326497e-06, + 'epoch': 10.93} +04/20 [03:09:47] INFO | >> train_qwenlatent.py:487 + Step 43330 | grad_norm_pre_clip=0.1550 | + grad_norm_pre_clip_avg=0.1601 | Metrics: + {'align_loss': 0.024814244359731674, + 'recon_loss': 0.1119641661643982, + 'predict_loss': 0.005718692671507597, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15504272282123566, + 'data_time': 0.0009050189983099699, + 'model_time': 1.6204900649900082, + 'grad_norm_pre_clip_avg': 0.1600778304040432, + 'learning_rate': 1.3545669045013494e-06, + 'epoch': 10.93} +04/20 [03:10:00] INFO | >> train_qwenlatent.py:487 + Step 43340 | grad_norm_pre_clip=0.1104 | + grad_norm_pre_clip_avg=0.1372 | Metrics: + {'align_loss': 0.026280326768755913, + 'recon_loss': 0.1941075623035431, + 'predict_loss': 0.007815473712980747, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1104012280702591, + 'data_time': 0.0007337740098591894, + 'model_time': 1.207087439019233, + 'grad_norm_pre_clip_avg': 0.13721767514944078, + 'learning_rate': 1.3506552159110937e-06, + 'epoch': 10.94} +04/20 [03:10:13] INFO | >> train_qwenlatent.py:487 + Step 43350 | grad_norm_pre_clip=0.1028 | + grad_norm_pre_clip_avg=0.1344 | Metrics: + {'align_loss': 0.025688044726848602, + 'recon_loss': 0.1528802514076233, + 'predict_loss': 0.007103435695171356, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10277809947729111, + 'mae_score': 0.006379871540241413, 'data_time': + 0.0009307609871029854, 'model_time': + 1.197042384010274, 'grad_norm_pre_clip_avg': + 0.1343805819749832, 'learning_rate': + 1.3467489674683908e-06, 'epoch': 10.94} +04/20 [03:10:26] INFO | >> train_qwenlatent.py:487 + Step 43360 | grad_norm_pre_clip=0.1733 | + grad_norm_pre_clip_avg=0.1357 | Metrics: + {'align_loss': 0.026158273220062256, + 'recon_loss': 0.15637628734111786, + 'predict_loss': 0.012034241110086441, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1733260601758957, + 'data_time': 0.0006571220001205802, + 'model_time': 1.20524818499689, + 'grad_norm_pre_clip_avg': 0.1357210747897625, + 'learning_rate': 1.3428481610770978e-06, + 'epoch': 10.94} +04/20 [03:10:38] INFO | >> train_qwenlatent.py:487 + Step 43370 | grad_norm_pre_clip=0.1206 | + grad_norm_pre_clip_avg=0.1455 | Metrics: + {'align_loss': 0.026124104857444763, + 'recon_loss': 0.12750644981861115, + 'predict_loss': 0.004531858488917351, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12056631594896317, + 'data_time': 0.000850359007017687, + 'model_time': 1.22274056400056, + 'grad_norm_pre_clip_avg': 0.14547043964266776, + 'learning_rate': 1.3389527986384196e-06, + 'epoch': 10.94} +04/20 [03:10:50] INFO | >> train_qwenlatent.py:487 + Step 43380 | grad_norm_pre_clip=0.1039 | + grad_norm_pre_clip_avg=0.1273 | Metrics: + {'align_loss': 0.024569692090153694, + 'recon_loss': 0.17328163981437683, + 'predict_loss': 0.005740485619753599, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10394295305013657, + 'data_time': 0.0008697829907760024, + 'model_time': 1.2645222719875164, + 'grad_norm_pre_clip_avg': 0.12732553035020827, + 'learning_rate': 1.3350628820509113e-06, + 'epoch': 10.95} +04/20 [03:11:03] INFO | >> train_qwenlatent.py:487 + Step 43390 | grad_norm_pre_clip=0.1389 | + grad_norm_pre_clip_avg=0.1394 | Metrics: + {'align_loss': 0.026765253394842148, + 'recon_loss': 0.23128123581409454, + 'predict_loss': 0.01590845175087452, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13888975977897644, + 'data_time': 0.001198565005324781, + 'model_time': 1.2462918689998332, + 'grad_norm_pre_clip_avg': 0.13936282098293304, + 'learning_rate': 1.3311784132104648e-06, + 'epoch': 10.95} +04/20 [03:11:16] INFO | >> train_qwenlatent.py:487 + Step 43400 | grad_norm_pre_clip=0.1347 | + grad_norm_pre_clip_avg=0.1441 | Metrics: + {'align_loss': 0.025548528879880905, + 'recon_loss': 0.16083329916000366, + 'predict_loss': 0.005987110082060099, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13470032811164856, + 'mae_score': 0.005070562190837688, 'data_time': + 0.001022651995299384, 'model_time': + 1.2555920910090208, 'grad_norm_pre_clip_avg': + 0.1440573178231716, 'learning_rate': + 1.3272993940103316e-06, 'epoch': 10.95} +04/20 [03:11:29] INFO | >> train_qwenlatent.py:487 + Step 43410 | grad_norm_pre_clip=0.1366 | + grad_norm_pre_clip_avg=0.1528 | Metrics: + {'align_loss': 0.02520962432026863, + 'recon_loss': 0.12285450845956802, + 'predict_loss': 0.004939430858939886, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13660885393619537, + 'data_time': 0.000665180996293202, + 'model_time': 1.2713410779833794, + 'grad_norm_pre_clip_avg': 0.15278953686356544, + 'learning_rate': 1.3234258263410955e-06, + 'epoch': 10.95} +04/20 [03:11:42] INFO | >> train_qwenlatent.py:487 + Step 43420 | grad_norm_pre_clip=0.1437 | + grad_norm_pre_clip_avg=0.1435 | Metrics: + {'align_loss': 0.024344677105545998, + 'recon_loss': 0.15081991255283356, + 'predict_loss': 0.009219028055667877, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14366991817951202, + 'data_time': 0.0006619999767281115, + 'model_time': 1.245310338010313, + 'grad_norm_pre_clip_avg': 0.14346210286021233, + 'learning_rate': 1.3195577120906846e-06, + 'epoch': 10.96} +04/20 [03:11:55] INFO | >> train_qwenlatent.py:487 + Step 43430 | grad_norm_pre_clip=0.1604 | + grad_norm_pre_clip_avg=0.1428 | Metrics: + {'align_loss': 0.025323381647467613, + 'recon_loss': 0.16129782795906067, + 'predict_loss': 0.008711058646440506, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1603647768497467, + 'data_time': 0.0008885930001270026, + 'model_time': 1.2217250969843008, + 'grad_norm_pre_clip_avg': 0.14275507479906083, + 'learning_rate': 1.315695053144367e-06, + 'epoch': 10.96} +04/20 [03:12:07] INFO | >> train_qwenlatent.py:487 + Step 43440 | grad_norm_pre_clip=0.1168 | + grad_norm_pre_clip_avg=0.1377 | Metrics: + {'align_loss': 0.025897640734910965, + 'recon_loss': 0.1199185773730278, + 'predict_loss': 0.0042915972881019115, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11682283133268356, + 'data_time': 0.0008784359961282462, + 'model_time': 1.2382535119832028, + 'grad_norm_pre_clip_avg': 0.13768698275089264, + 'learning_rate': 1.3118378513847635e-06, + 'epoch': 10.96} +04/20 [03:12:21] INFO | >> train_qwenlatent.py:487 + Step 43450 | grad_norm_pre_clip=0.1464 | + grad_norm_pre_clip_avg=0.1480 | Metrics: + {'align_loss': 0.02472078800201416, + 'recon_loss': 0.1822456568479538, + 'predict_loss': 0.009473731741309166, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14636212587356567, + 'mae_score': 0.005379345180752041, 'data_time': + 0.000885632005520165, 'model_time': + 1.2181576720031444, 'grad_norm_pre_clip_avg': + 0.14798564836382866, 'learning_rate': + 1.3079861086918246e-06, 'epoch': 10.96} +04/20 [03:12:33] INFO | >> train_qwenlatent.py:487 + Step 43460 | grad_norm_pre_clip=0.1410 | + grad_norm_pre_clip_avg=0.1424 | Metrics: + {'align_loss': 0.026161927729845047, + 'recon_loss': 0.20875956118106842, + 'predict_loss': 0.006621525157243013, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14103923738002777, + 'data_time': 0.0006485340127255768, + 'model_time': 1.2385938790102955, + 'grad_norm_pre_clip_avg': 0.14241455122828484, + 'learning_rate': 1.3041398269428412e-06, + 'epoch': 10.97} +04/20 [03:12:46] INFO | >> train_qwenlatent.py:487 + Step 43470 | grad_norm_pre_clip=0.1694 | + grad_norm_pre_clip_avg=0.1512 | Metrics: + {'align_loss': 0.024997275322675705, + 'recon_loss': 0.20191991329193115, + 'predict_loss': 0.00762972654774785, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1694037914276123, + 'data_time': 0.0009809850016608834, + 'model_time': 1.2268356540007517, + 'grad_norm_pre_clip_avg': 0.15124838203191757, + 'learning_rate': 1.300299008012447e-06, + 'epoch': 10.97} +04/20 [03:12:58] INFO | >> train_qwenlatent.py:487 + Step 43480 | grad_norm_pre_clip=0.1433 | + grad_norm_pre_clip_avg=0.1445 | Metrics: + {'align_loss': 0.025207115337252617, + 'recon_loss': 0.15245015919208527, + 'predict_loss': 0.006780234631150961, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14333535730838776, + 'data_time': 0.000984807003987953, + 'model_time': 1.2829952279862482, + 'grad_norm_pre_clip_avg': 0.1445499651134014, + 'learning_rate': 1.2964636537726054e-06, + 'epoch': 10.97} +04/20 [03:13:11] INFO | >> train_qwenlatent.py:487 + Step 43490 | grad_norm_pre_clip=0.1257 | + grad_norm_pre_clip_avg=0.1284 | Metrics: + {'align_loss': 0.027162138372659683, + 'recon_loss': 0.17063090205192566, + 'predict_loss': 0.005153353791683912, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12566691637039185, + 'data_time': 0.0005934259970672429, + 'model_time': 1.195187106000958, + 'grad_norm_pre_clip_avg': 0.12838922217488288, + 'learning_rate': 1.2926337660926303e-06, + 'epoch': 10.97} +04/20 [03:13:24] INFO | >> train_qwenlatent.py:487 + Step 43500 | grad_norm_pre_clip=0.1654 | + grad_norm_pre_clip_avg=0.1423 | Metrics: + {'align_loss': 0.02536148950457573, + 'recon_loss': 0.18677571415901184, + 'predict_loss': 0.009718057699501514, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1653977781534195, + 'mae_score': 0.004792672664195568, 'data_time': + 0.0008117149991448969, 'model_time': + 1.2346780950028915, 'grad_norm_pre_clip_avg': + 0.1422653689980507, 'learning_rate': + 1.2888093468391546e-06, 'epoch': 10.98} +04/20 [03:13:36] INFO | >> train_qwenlatent.py:487 + Step 43510 | grad_norm_pre_clip=0.1224 | + grad_norm_pre_clip_avg=0.1374 | Metrics: + {'align_loss': 0.025130145251750946, + 'recon_loss': 0.18222647905349731, + 'predict_loss': 0.005294301547110081, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1224094033241272, + 'data_time': 0.0009512079996056855, + 'model_time': 1.1958918829914182, + 'grad_norm_pre_clip_avg': 0.13741822615265847, + 'learning_rate': 1.2849903978761556e-06, + 'epoch': 10.98} +04/20 [03:13:49] INFO | >> train_qwenlatent.py:487 + Step 43520 | grad_norm_pre_clip=0.1929 | + grad_norm_pre_clip_avg=0.1473 | Metrics: + {'align_loss': 0.024760250002145767, + 'recon_loss': 0.2148377001285553, + 'predict_loss': 0.017811890691518784, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19291749596595764, + 'data_time': 0.0008301000052597374, + 'model_time': 1.2496770930010825, + 'grad_norm_pre_clip_avg': 0.1473101943731308, + 'learning_rate': 1.2811769210649413e-06, + 'epoch': 10.98} +04/20 [03:14:02] INFO | >> train_qwenlatent.py:487 + Step 43530 | grad_norm_pre_clip=0.1600 | + grad_norm_pre_clip_avg=0.1491 | Metrics: + {'align_loss': 0.024207260459661484, + 'recon_loss': 0.1672229915857315, + 'predict_loss': 0.009939515963196754, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15998269617557526, + 'data_time': 0.0009430100035388023, + 'model_time': 1.2702070990053471, + 'grad_norm_pre_clip_avg': 0.1491323560476303, + 'learning_rate': 1.2773689182641584e-06, + 'epoch': 10.98} +04/20 [03:14:15] INFO | >> train_qwenlatent.py:487 + Step 43540 | grad_norm_pre_clip=0.1399 | + grad_norm_pre_clip_avg=0.1466 | Metrics: + {'align_loss': 0.025578070431947708, + 'recon_loss': 0.1735454946756363, + 'predict_loss': 0.006767571438103914, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13986821472644806, + 'data_time': 0.0008318990003317595, + 'model_time': 1.1965244119928684, + 'grad_norm_pre_clip_avg': 0.14655184373259544, + 'learning_rate': 1.2735663913297801e-06, + 'epoch': 10.99} +04/20 [03:14:28] INFO | >> train_qwenlatent.py:487 + Step 43550 | grad_norm_pre_clip=0.1736 | + grad_norm_pre_clip_avg=0.1541 | Metrics: + {'align_loss': 0.026158902794122696, + 'recon_loss': 0.1654842346906662, + 'predict_loss': 0.008145852945744991, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17357341945171356, + 'mae_score': 0.005769956434095228, 'data_time': + 0.0009596830059308559, 'model_time': + 1.6324621470121201, 'grad_norm_pre_clip_avg': + 0.15411145687103273, 'learning_rate': + 1.2697693421151085e-06, 'epoch': 10.99} +04/20 [03:14:40] INFO | >> train_qwenlatent.py:487 + Step 43560 | grad_norm_pre_clip=0.1558 | + grad_norm_pre_clip_avg=0.1444 | Metrics: + {'align_loss': 0.024776410311460495, + 'recon_loss': 0.12395188957452774, + 'predict_loss': 0.004823853261768818, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15583516657352448, + 'data_time': 0.0006294059858191758, + 'model_time': 1.2244809349940624, + 'grad_norm_pre_clip_avg': 0.1443974032998085, + 'learning_rate': 1.2659777724707831e-06, + 'epoch': 10.99} +04/20 [03:14:53] INFO | >> train_qwenlatent.py:487 + Step 43570 | grad_norm_pre_clip=0.1247 | + grad_norm_pre_clip_avg=0.1432 | Metrics: + {'align_loss': 0.0258132666349411, + 'recon_loss': 0.11458069831132889, + 'predict_loss': 0.006687095854431391, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12472362071275711, + 'data_time': 0.000687671999912709, + 'model_time': 1.2656100330059417, + 'grad_norm_pre_clip_avg': 0.14318894147872924, + 'learning_rate': 1.262191684244766e-06, + 'epoch': 10.99} +04/20 [03:15:06] INFO | >> train_qwenlatent.py:487 + Step 43580 | grad_norm_pre_clip=0.1434 | + grad_norm_pre_clip_avg=0.1726 | Metrics: + {'align_loss': 0.024486567825078964, + 'recon_loss': 0.10890944302082062, + 'predict_loss': 0.006866233889013529, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.143381267786026, + 'data_time': 0.0006913139950484037, + 'model_time': 1.2128141110006254, + 'grad_norm_pre_clip_avg': 0.17259118407964708, + 'learning_rate': 1.258411079282352e-06, + 'epoch': 11.0} +04/20 [03:15:18] INFO | >> train_qwenlatent.py:487 + Step 43590 | grad_norm_pre_clip=0.1657 | + grad_norm_pre_clip_avg=0.1674 | Metrics: + {'align_loss': 0.026385698467493057, + 'recon_loss': 0.16058781743049622, + 'predict_loss': 0.009815938770771027, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.165708526968956, + 'data_time': 0.0009260780061595142, + 'model_time': 1.315116117009893, + 'grad_norm_pre_clip_avg': 0.16737693250179292, + 'learning_rate': 1.2546359594261627e-06, + 'epoch': 11.0} +04/20 [03:15:32] INFO | >> train_qwenlatent.py:487 + Step 43600 | grad_norm_pre_clip=0.1260 | + grad_norm_pre_clip_avg=0.1478 | Metrics: + {'align_loss': 0.025100260972976685, + 'recon_loss': 0.1648644059896469, + 'predict_loss': 0.007480298634618521, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1260479837656021, + 'mae_score': 0.006448546401015273, 'data_time': + 0.000872595002874732, 'model_time': + 1.25634144799551, 'grad_norm_pre_clip_avg': + 0.1478474482893944, 'learning_rate': + 1.250866326516143e-06, 'epoch': 11.0} +04/20 [03:15:44] INFO | >> train_qwenlatent.py:487 + Step 43610 | grad_norm_pre_clip=0.0997 | + grad_norm_pre_clip_avg=0.1388 | Metrics: + {'align_loss': 0.02522413432598114, + 'recon_loss': 0.15633685886859894, + 'predict_loss': 0.0062501030042767525, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.0997430682182312, + 'data_time': 0.0009136009903158993, + 'model_time': 1.2186795980087481, + 'grad_norm_pre_clip_avg': 0.13884531185030938, + 'learning_rate': 1.2471021823895686e-06, + 'epoch': 11.0} +04/20 [03:15:57] INFO | >> train_qwenlatent.py:487 + Step 43620 | grad_norm_pre_clip=0.0883 | + grad_norm_pre_clip_avg=0.1296 | Metrics: + {'align_loss': 0.025526706129312515, + 'recon_loss': 0.15537458658218384, + 'predict_loss': 0.005473004654049873, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.08830095827579498, + 'data_time': 0.0009196629980579019, + 'model_time': 1.2022475079866126, + 'grad_norm_pre_clip_avg': 0.12959592789411545, + 'learning_rate': 1.2433435288810353e-06, + 'epoch': 11.01} +04/20 [03:16:09] INFO | >> train_qwenlatent.py:487 + Step 43630 | grad_norm_pre_clip=0.2088 | + grad_norm_pre_clip_avg=0.1475 | Metrics: + {'align_loss': 0.024958845227956772, + 'recon_loss': 0.1278078705072403, + 'predict_loss': 0.00867959763854742, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20880576968193054, + 'data_time': 0.0006842059956397861, + 'model_time': 1.2187264419917483, + 'grad_norm_pre_clip_avg': 0.147504585981369, + 'learning_rate': 1.239590367822468e-06, + 'epoch': 11.01} +04/20 [03:16:22] INFO | >> train_qwenlatent.py:487 + Step 43640 | grad_norm_pre_clip=0.1488 | + grad_norm_pre_clip_avg=0.1636 | Metrics: + {'align_loss': 0.025676969438791275, + 'recon_loss': 0.14236898720264435, + 'predict_loss': 0.008191836066544056, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1487903594970703, + 'data_time': 0.00090603600256145, 'model_time': + 1.230012259009527, 'grad_norm_pre_clip_avg': + 0.16356272473931313, 'learning_rate': + 1.235842701043112e-06, 'epoch': 11.01} +04/20 [03:16:35] INFO | >> train_qwenlatent.py:487 + Step 43650 | grad_norm_pre_clip=0.1206 | + grad_norm_pre_clip_avg=0.1389 | Metrics: + {'align_loss': 0.02693096175789833, + 'recon_loss': 0.14606356620788574, + 'predict_loss': 0.004653204698115587, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12063631415367126, + 'mae_score': 0.0052891430554089245, + 'data_time': 0.0011973440123256296, + 'model_time': 1.2648973480099812, + 'grad_norm_pre_clip_avg': 0.13894594460725784, + 'learning_rate': 1.232100530369531e-06, + 'epoch': 11.01} +04/20 [03:16:47] INFO | >> train_qwenlatent.py:487 + Step 43660 | grad_norm_pre_clip=0.1225 | + grad_norm_pre_clip_avg=0.1378 | Metrics: + {'align_loss': 0.025432255119085312, + 'recon_loss': 0.18376529216766357, + 'predict_loss': 0.009324593469500542, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12253079563379288, + 'data_time': 0.0006469580112025142, + 'model_time': 1.2224869070050772, + 'grad_norm_pre_clip_avg': 0.13781469762325288, + 'learning_rate': 1.2283638576256177e-06, + 'epoch': 11.02} +04/20 [03:17:00] INFO | >> train_qwenlatent.py:487 + Step 43670 | grad_norm_pre_clip=0.1802 | + grad_norm_pre_clip_avg=0.1553 | Metrics: + {'align_loss': 0.025677233934402466, + 'recon_loss': 0.18441876769065857, + 'predict_loss': 0.009463350288569927, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18016640841960907, + 'data_time': 0.000999860989395529, + 'model_time': 1.5183173390105367, + 'grad_norm_pre_clip_avg': 0.15532787069678305, + 'learning_rate': 1.2246326846325785e-06, + 'epoch': 11.02} +04/20 [03:17:13] INFO | >> train_qwenlatent.py:487 + Step 43680 | grad_norm_pre_clip=0.1274 | + grad_norm_pre_clip_avg=0.1412 | Metrics: + {'align_loss': 0.024053603410720825, + 'recon_loss': 0.12488586455583572, + 'predict_loss': 0.004907450173050165, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12744733691215515, + 'data_time': 0.0007208620081655681, + 'model_time': 1.2290166040183976, + 'grad_norm_pre_clip_avg': 0.14120227620005607, + 'learning_rate': 1.2209070132089407e-06, + 'epoch': 11.02} +04/20 [03:17:25] INFO | >> train_qwenlatent.py:487 + Step 43690 | grad_norm_pre_clip=0.1666 | + grad_norm_pre_clip_avg=0.1398 | Metrics: + {'align_loss': 0.025589611381292343, + 'recon_loss': 0.19021891057491302, + 'predict_loss': 0.009999360889196396, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16658593714237213, + 'data_time': 0.0006716870120726526, + 'model_time': 1.2177784529922064, + 'grad_norm_pre_clip_avg': 0.13977934122085572, + 'learning_rate': 1.2171868451705541e-06, + 'epoch': 11.02} +04/20 [03:17:39] INFO | >> train_qwenlatent.py:487 + Step 43700 | grad_norm_pre_clip=0.1626 | + grad_norm_pre_clip_avg=0.1354 | Metrics: + {'align_loss': 0.026081275194883347, + 'recon_loss': 0.18160095810890198, + 'predict_loss': 0.006800004281103611, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16264109313488007, + 'mae_score': 0.0054036900803849505, + 'data_time': 0.0008466659928672016, + 'model_time': 1.2528772959776688, + 'grad_norm_pre_clip_avg': 0.1354220062494278, + 'learning_rate': 1.2134721823305834e-06, + 'epoch': 11.03} +04/20 [03:17:52] INFO | >> train_qwenlatent.py:487 + Step 43710 | grad_norm_pre_clip=0.1609 | + grad_norm_pre_clip_avg=0.1585 | Metrics: + {'align_loss': 0.024965941905975342, + 'recon_loss': 0.13043157756328583, + 'predict_loss': 0.010511208325624466, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16088157892227173, + 'data_time': 0.0006187020044308156, + 'model_time': 1.2315382400120143, + 'grad_norm_pre_clip_avg': 0.15852890983223916, + 'learning_rate': 1.209763026499507e-06, + 'epoch': 11.03} +04/20 [03:18:04] INFO | >> train_qwenlatent.py:487 + Step 43720 | grad_norm_pre_clip=0.0990 | + grad_norm_pre_clip_avg=0.1462 | Metrics: + {'align_loss': 0.026152528822422028, + 'recon_loss': 0.11968536674976349, + 'predict_loss': 0.003501482307910919, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09895160049200058, + 'data_time': 0.0008947899914346635, + 'model_time': 1.1896837660169695, + 'grad_norm_pre_clip_avg': 0.1461930826306343, + 'learning_rate': 1.206059379485122e-06, + 'epoch': 11.03} +04/20 [03:18:17] INFO | >> train_qwenlatent.py:487 + Step 43730 | grad_norm_pre_clip=0.1645 | + grad_norm_pre_clip_avg=0.1388 | Metrics: + {'align_loss': 0.024518541991710663, + 'recon_loss': 0.11082950234413147, + 'predict_loss': 0.00783630646765232, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16454730927944183, + 'data_time': 0.00065684798755683, 'model_time': + 1.182027750997804, 'grad_norm_pre_clip_avg': + 0.13880063146352767, 'learning_rate': + 1.2023612430925466e-06, 'epoch': 11.03} +04/20 [03:18:29] INFO | >> train_qwenlatent.py:487 + Step 43740 | grad_norm_pre_clip=0.1767 | + grad_norm_pre_clip_avg=0.1399 | Metrics: + {'align_loss': 0.02638176456093788, + 'recon_loss': 0.21696536242961884, + 'predict_loss': 0.0133019695058465, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1767464578151703, + 'data_time': 0.0008436210046056658, + 'model_time': 1.2530455039814115, + 'grad_norm_pre_clip_avg': 0.13993312567472457, + 'learning_rate': 1.1986686191242035e-06, + 'epoch': 11.04} +04/20 [03:18:43] INFO | >> train_qwenlatent.py:487 + Step 43750 | grad_norm_pre_clip=0.1277 | + grad_norm_pre_clip_avg=0.1360 | Metrics: + {'align_loss': 0.02617400884628296, + 'recon_loss': 0.22361692786216736, + 'predict_loss': 0.010593701153993607, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1276969313621521, + 'mae_score': 0.005013699574513478, 'data_time': + 0.0010824569908436388, 'model_time': + 1.2335471400001552, 'grad_norm_pre_clip_avg': + 0.13603893145918847, 'learning_rate': + 1.194981509379834e-06, 'epoch': 11.04} +04/20 [03:18:56] INFO | >> train_qwenlatent.py:487 + Step 43760 | grad_norm_pre_clip=0.1575 | + grad_norm_pre_clip_avg=0.1358 | Metrics: + {'align_loss': 0.025434602051973343, + 'recon_loss': 0.15158578753471375, + 'predict_loss': 0.013129569590091705, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15754038095474243, + 'data_time': 0.0008565309981349856, + 'model_time': 1.2452574569906574, + 'grad_norm_pre_clip_avg': 0.13581012263894082, + 'learning_rate': 1.1912999156564896e-06, + 'epoch': 11.04} +04/20 [03:19:08] INFO | >> train_qwenlatent.py:487 + Step 43770 | grad_norm_pre_clip=0.1268 | + grad_norm_pre_clip_avg=0.1432 | Metrics: + {'align_loss': 0.024924837052822113, + 'recon_loss': 0.13889135420322418, + 'predict_loss': 0.004483507946133614, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12675748765468597, + 'data_time': 0.0006443769962061197, + 'model_time': 1.2663584330002777, + 'grad_norm_pre_clip_avg': 0.14317271634936332, + 'learning_rate': 1.1876238397485347e-06, + 'epoch': 11.04} +04/20 [03:19:21] INFO | >> train_qwenlatent.py:487 + Step 43780 | grad_norm_pre_clip=0.1746 | + grad_norm_pre_clip_avg=0.1512 | Metrics: + {'align_loss': 0.02518083155155182, + 'recon_loss': 0.12564314901828766, + 'predict_loss': 0.008597170002758503, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17463445663452148, + 'data_time': 0.000930995011003688, + 'model_time': 1.2654295200190973, + 'grad_norm_pre_clip_avg': 0.15117280408740044, + 'learning_rate': 1.183953283447645e-06, + 'epoch': 11.05} +04/20 [03:19:33] INFO | >> train_qwenlatent.py:487 + Step 43790 | grad_norm_pre_clip=0.1287 | + grad_norm_pre_clip_avg=0.1482 | Metrics: + {'align_loss': 0.026128631085157394, + 'recon_loss': 0.1434680074453354, + 'predict_loss': 0.0065231011249125, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12870515882968903, + 'data_time': 0.0012729949958156794, + 'model_time': 1.2383628999814391, + 'grad_norm_pre_clip_avg': 0.14820010215044022, + 'learning_rate': 1.1802882485428032e-06, + 'epoch': 11.05} +04/20 [03:19:47] INFO | >> train_qwenlatent.py:487 + Step 43800 | grad_norm_pre_clip=0.1226 | + grad_norm_pre_clip_avg=0.1485 | Metrics: + {'align_loss': 0.02506713755428791, + 'recon_loss': 0.08806295692920685, + 'predict_loss': 0.002923421561717987, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.122569240629673, + 'mae_score': 0.005205000198639191, 'data_time': + 0.0008193910180125386, 'model_time': + 1.2878040299983695, 'grad_norm_pre_clip_avg': + 0.1484530061483383, 'learning_rate': + 1.1766287368203041e-06, 'epoch': 11.05} +04/20 [03:20:00] INFO | >> train_qwenlatent.py:487 + Step 43810 | grad_norm_pre_clip=0.1506 | + grad_norm_pre_clip_avg=0.1474 | Metrics: + {'align_loss': 0.025296451523900032, + 'recon_loss': 0.14292776584625244, + 'predict_loss': 0.007382719777524471, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15064242482185364, + 'data_time': 0.0009605030063539743, + 'model_time': 1.1946149709983729, + 'grad_norm_pre_clip_avg': 0.14742902517318726, + 'learning_rate': 1.1729747500637476e-06, + 'epoch': 11.05} +04/20 [03:20:12] INFO | >> train_qwenlatent.py:487 + Step 43820 | grad_norm_pre_clip=0.1616 | + grad_norm_pre_clip_avg=0.1394 | Metrics: + {'align_loss': 0.024905186146497726, + 'recon_loss': 0.13574326038360596, + 'predict_loss': 0.0069364747032523155, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16156011819839478, + 'data_time': 0.0011498699896037579, + 'model_time': 1.2678204989933874, + 'grad_norm_pre_clip_avg': 0.13936314582824708, + 'learning_rate': 1.1693262900540452e-06, + 'epoch': 11.06} +04/20 [03:20:25] INFO | >> train_qwenlatent.py:487 + Step 43830 | grad_norm_pre_clip=0.1186 | + grad_norm_pre_clip_avg=0.1386 | Metrics: + {'align_loss': 0.025178760290145874, + 'recon_loss': 0.13939465582370758, + 'predict_loss': 0.006462497636675835, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11858242005109787, + 'data_time': 0.0006656220066361129, + 'model_time': 1.2379905940033495, + 'grad_norm_pre_clip_avg': 0.13856440037488937, + 'learning_rate': 1.1656833585694102e-06, + 'epoch': 11.06} +04/20 [03:20:37] INFO | >> train_qwenlatent.py:487 + Step 43840 | grad_norm_pre_clip=0.1601 | + grad_norm_pre_clip_avg=0.1481 | Metrics: + {'align_loss': 0.024941757321357727, + 'recon_loss': 0.16603559255599976, + 'predict_loss': 0.006610619369894266, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16010704636573792, + 'data_time': 0.000944335013628006, + 'model_time': 1.2802602529991418, + 'grad_norm_pre_clip_avg': 0.1481179840862751, + 'learning_rate': 1.1620459573853638e-06, + 'epoch': 11.06} +04/20 [03:20:51] INFO | >> train_qwenlatent.py:487 + Step 43850 | grad_norm_pre_clip=0.1055 | + grad_norm_pre_clip_avg=0.1515 | Metrics: + {'align_loss': 0.025390367954969406, + 'recon_loss': 0.1555851846933365, + 'predict_loss': 0.006880480796098709, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10549452155828476, + 'mae_score': 0.00480654003383877, 'data_time': + 0.0007202650012914091, 'model_time': + 1.2170236140082125, 'grad_norm_pre_clip_avg': + 0.1515394315123558, 'learning_rate': + 1.1584140882747318e-06, 'epoch': 11.06} +04/20 [03:21:03] INFO | >> train_qwenlatent.py:487 + Step 43860 | grad_norm_pre_clip=0.1568 | + grad_norm_pre_clip_avg=0.1472 | Metrics: + {'align_loss': 0.026613391935825348, + 'recon_loss': 0.20084019005298615, + 'predict_loss': 0.010977533645927906, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15680766105651855, + 'data_time': 0.00068591398303397, 'model_time': + 1.2082639820000622, 'grad_norm_pre_clip_avg': + 0.1471863366663456, 'learning_rate': + 1.1547877530076414e-06, 'epoch': 11.07} +04/20 [03:21:16] INFO | >> train_qwenlatent.py:487 + Step 43870 | grad_norm_pre_clip=0.1033 | + grad_norm_pre_clip_avg=0.1429 | Metrics: + {'align_loss': 0.025039728730916977, + 'recon_loss': 0.13922370970249176, + 'predict_loss': 0.006226785015314817, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10333488881587982, + 'data_time': 0.0006565790099557489, + 'model_time': 1.2233331740135327, + 'grad_norm_pre_clip_avg': 0.14287059530615806, + 'learning_rate': 1.151166953351525e-06, + 'epoch': 11.07} +04/20 [03:21:28] INFO | >> train_qwenlatent.py:487 + Step 43880 | grad_norm_pre_clip=0.1260 | + grad_norm_pre_clip_avg=0.1448 | Metrics: + {'align_loss': 0.026303034275770187, + 'recon_loss': 0.15694370865821838, + 'predict_loss': 0.00879466999322176, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12602046132087708, + 'data_time': 0.0008378950005862862, + 'model_time': 1.2561197059985716, + 'grad_norm_pre_clip_avg': 0.14478880316019058, + 'learning_rate': 1.147551691071117e-06, + 'epoch': 11.07} +04/20 [03:21:41] INFO | >> train_qwenlatent.py:487 + Step 43890 | grad_norm_pre_clip=0.2086 | + grad_norm_pre_clip_avg=0.1610 | Metrics: + {'align_loss': 0.025655750185251236, + 'recon_loss': 0.10306059569120407, + 'predict_loss': 0.005231638438999653, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20860964059829712, + 'data_time': 0.0009548770030960441, + 'model_time': 1.537015056994278, + 'grad_norm_pre_clip_avg': 0.16099683269858361, + 'learning_rate': 1.143941967928452e-06, + 'epoch': 11.07} +04/20 [03:21:55] INFO | >> train_qwenlatent.py:487 + Step 43900 | grad_norm_pre_clip=0.1275 | + grad_norm_pre_clip_avg=0.1457 | Metrics: + {'align_loss': 0.024772927165031433, + 'recon_loss': 0.12355057150125504, + 'predict_loss': 0.0060599506832659245, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12748286128044128, + 'mae_score': 0.006663543683988554, 'data_time': + 0.0008386950066778809, 'model_time': + 1.2199650929833297, 'grad_norm_pre_clip_avg': + 0.14574865102767945, 'learning_rate': + 1.1403377856828647e-06, 'epoch': 11.08} +04/20 [03:22:07] INFO | >> train_qwenlatent.py:487 + Step 43910 | grad_norm_pre_clip=0.0986 | + grad_norm_pre_clip_avg=0.1480 | Metrics: + {'align_loss': 0.026006903499364853, + 'recon_loss': 0.14365489780902863, + 'predict_loss': 0.005567308980971575, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09859249740839005, + 'data_time': 0.0007362019969150424, + 'model_time': 1.2172273760079406, + 'grad_norm_pre_clip_avg': 0.14801461324095727, + 'learning_rate': 1.1367391460909885e-06, + 'epoch': 11.08} +04/20 [03:22:20] INFO | >> train_qwenlatent.py:487 + Step 43920 | grad_norm_pre_clip=0.0746 | + grad_norm_pre_clip_avg=0.1384 | Metrics: + {'align_loss': 0.024916190654039383, + 'recon_loss': 0.18311941623687744, + 'predict_loss': 0.008192131295800209, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.07456877082586288, + 'data_time': 0.0007540430233348161, + 'model_time': 1.242720850976184, + 'grad_norm_pre_clip_avg': 0.13835951015353204, + 'learning_rate': 1.1331460509067593e-06, + 'epoch': 11.08} +04/20 [03:22:32] INFO | >> train_qwenlatent.py:487 + Step 43930 | grad_norm_pre_clip=0.1396 | + grad_norm_pre_clip_avg=0.1337 | Metrics: + {'align_loss': 0.024481214582920074, + 'recon_loss': 0.1416391283273697, + 'predict_loss': 0.0079495869576931, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1395726203918457, + 'data_time': 0.0007319730066228658, + 'model_time': 1.2164538009965327, + 'grad_norm_pre_clip_avg': 0.13371472731232642, + 'learning_rate': 1.1295585018814067e-06, + 'epoch': 11.09} +04/20 [03:22:45] INFO | >> train_qwenlatent.py:487 + Step 43940 | grad_norm_pre_clip=0.2011 | + grad_norm_pre_clip_avg=0.1501 | Metrics: + {'align_loss': 0.025634583085775375, + 'recon_loss': 0.19392533600330353, + 'predict_loss': 0.009859609417617321, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2011050283908844, + 'data_time': 0.0007717109983786941, + 'model_time': 1.2206185510149226, + 'grad_norm_pre_clip_avg': 0.15008090808987617, + 'learning_rate': 1.1259765007634623e-06, + 'epoch': 11.09} +04/20 [03:22:58] INFO | >> train_qwenlatent.py:487 + Step 43950 | grad_norm_pre_clip=0.1654 | + grad_norm_pre_clip_avg=0.1384 | Metrics: + {'align_loss': 0.02627408504486084, + 'recon_loss': 0.18395556509494781, + 'predict_loss': 0.005634831264615059, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16538958251476288, + 'mae_score': 0.005604289888261674, 'data_time': + 0.0010107639827765524, 'model_time': + 1.2608702759898733, 'grad_norm_pre_clip_avg': + 0.13842145428061486, 'learning_rate': + 1.1224000492987414e-06, 'epoch': 11.09} +04/20 [03:23:10] INFO | >> train_qwenlatent.py:487 + Step 43960 | grad_norm_pre_clip=0.1584 | + grad_norm_pre_clip_avg=0.1522 | Metrics: + {'align_loss': 0.02529170550405979, + 'recon_loss': 0.14792700111865997, + 'predict_loss': 0.006263278424739838, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1584467887878418, + 'data_time': 0.000681803998304531, + 'model_time': 1.2174246459908318, + 'grad_norm_pre_clip_avg': 0.1522068589925766, + 'learning_rate': 1.1188291492303705e-06, + 'epoch': 11.09} +04/20 [03:23:23] INFO | >> train_qwenlatent.py:487 + Step 43970 | grad_norm_pre_clip=0.1915 | + grad_norm_pre_clip_avg=0.1736 | Metrics: + {'align_loss': 0.025739602744579315, + 'recon_loss': 0.11067882925271988, + 'predict_loss': 0.005834073759615421, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19148895144462585, + 'data_time': 0.0006920860032550991, + 'model_time': 1.2171826870180666, + 'grad_norm_pre_clip_avg': 0.17360036969184875, + 'learning_rate': 1.1152638022987594e-06, + 'epoch': 11.1} +04/20 [03:23:35] INFO | >> train_qwenlatent.py:487 + Step 43980 | grad_norm_pre_clip=0.1479 | + grad_norm_pre_clip_avg=0.1389 | Metrics: + {'align_loss': 0.026654895395040512, + 'recon_loss': 0.15137727558612823, + 'predict_loss': 0.006217098329216242, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14793431758880615, + 'data_time': 0.0006839809939265251, + 'model_time': 1.2162161250016652, + 'grad_norm_pre_clip_avg': 0.13888359144330026, + 'learning_rate': 1.1117040102416162e-06, + 'epoch': 11.1} +04/20 [03:23:48] INFO | >> train_qwenlatent.py:487 + Step 43990 | grad_norm_pre_clip=0.1415 | + grad_norm_pre_clip_avg=0.1425 | Metrics: + {'align_loss': 0.02574991062283516, + 'recon_loss': 0.14024776220321655, + 'predict_loss': 0.005226308014243841, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1415226310491562, + 'data_time': 0.0007183020061347634, + 'model_time': 1.2449109209992457, + 'grad_norm_pre_clip_avg': 0.1424680970609188, + 'learning_rate': 1.1081497747939398e-06, + 'epoch': 11.1} +04/20 [03:24:01] INFO | >> train_qwenlatent.py:487 + Step 44000 | grad_norm_pre_clip=0.1474 | + grad_norm_pre_clip_avg=0.1739 | Metrics: + {'align_loss': 0.02448357082903385, + 'recon_loss': 0.14375385642051697, + 'predict_loss': 0.00959042925387621, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14736171066761017, + 'mae_score': 0.005458342921626461, 'data_time': + 0.001023805991280824, 'model_time': + 1.2370322339993436, 'grad_norm_pre_clip_avg': + 0.17385766208171843, 'learning_rate': + 1.1046010976880203e-06, 'epoch': 11.1} +04/20 [03:24:14] INFO | >> train_qwenlatent.py:487 + Step 44010 | grad_norm_pre_clip=0.1440 | + grad_norm_pre_clip_avg=0.1700 | Metrics: + {'align_loss': 0.02383464202284813, + 'recon_loss': 0.10461412370204926, + 'predict_loss': 0.0073624965734779835, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1439819633960724, + 'data_time': 0.0009697549976408482, + 'model_time': 1.242824932007352, + 'grad_norm_pre_clip_avg': 0.16995338797569276, + 'learning_rate': 1.101057980653443e-06, + 'epoch': 11.11} +04/20 [03:24:27] INFO | >> train_qwenlatent.py:487 + Step 44020 | grad_norm_pre_clip=0.1279 | + grad_norm_pre_clip_avg=0.1467 | Metrics: + {'align_loss': 0.024564148858189583, + 'recon_loss': 0.116672582924366, + 'predict_loss': 0.004293209407478571, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12785518169403076, + 'data_time': 0.0008864479896146804, + 'model_time': 1.2396886280039325, + 'grad_norm_pre_clip_avg': 0.146687288582325, + 'learning_rate': 1.0975204254170786e-06, + 'epoch': 11.11} +04/20 [03:24:39] INFO | >> train_qwenlatent.py:487 + Step 44030 | grad_norm_pre_clip=0.1231 | + grad_norm_pre_clip_avg=0.1292 | Metrics: + {'align_loss': 0.0251136664301157, + 'recon_loss': 0.17116567492485046, + 'predict_loss': 0.006214826367795467, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12306668609380722, + 'data_time': 0.0008816359913907945, + 'model_time': 1.2159279650077224, + 'grad_norm_pre_clip_avg': 0.1292121045291424, + 'learning_rate': 1.0939884337030893e-06, + 'epoch': 11.11} +04/20 [03:24:52] INFO | >> train_qwenlatent.py:487 + Step 44040 | grad_norm_pre_clip=0.1864 | + grad_norm_pre_clip_avg=0.1477 | Metrics: + {'align_loss': 0.026664774864912033, + 'recon_loss': 0.10193109512329102, + 'predict_loss': 0.00853987317532301, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18635761737823486, + 'data_time': 0.0007242339779622853, + 'model_time': 1.2263579840073362, + 'grad_norm_pre_clip_avg': 0.14767997488379478, + 'learning_rate': 1.090462007232919e-06, + 'epoch': 11.11} +04/20 [03:25:05] INFO | >> train_qwenlatent.py:487 + Step 44050 | grad_norm_pre_clip=0.1980 | + grad_norm_pre_clip_avg=0.1517 | Metrics: + {'align_loss': 0.026393841952085495, + 'recon_loss': 0.13156992197036743, + 'predict_loss': 0.006852534133940935, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19803576171398163, + 'mae_score': 0.005870778710992486, 'data_time': + 0.0006797719979658723, 'model_time': + 1.217527841014089, 'grad_norm_pre_clip_avg': + 0.15170554146170617, 'learning_rate': + 1.0869411477253137e-06, 'epoch': 11.12} +04/20 [03:25:18] INFO | >> train_qwenlatent.py:487 + Step 44060 | grad_norm_pre_clip=0.1321 | + grad_norm_pre_clip_avg=0.1525 | Metrics: + {'align_loss': 0.02650168165564537, + 'recon_loss': 0.22498805820941925, + 'predict_loss': 0.01044359989464283, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1321498304605484, + 'data_time': 0.000709681014996022, + 'model_time': 1.3091764200071339, + 'grad_norm_pre_clip_avg': 0.15253133475780487, + 'learning_rate': 1.0834258568962924e-06, + 'epoch': 11.12} +04/20 [03:25:30] INFO | >> train_qwenlatent.py:487 + Step 44070 | grad_norm_pre_clip=0.1439 | + grad_norm_pre_clip_avg=0.1356 | Metrics: + {'align_loss': 0.026101313531398773, + 'recon_loss': 0.16142819821834564, + 'predict_loss': 0.007880028337240219, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1438593864440918, + 'data_time': 0.0007533010211773217, + 'model_time': 1.2557328649854753, + 'grad_norm_pre_clip_avg': 0.13562151193618774, + 'learning_rate': 1.0799161364591671e-06, + 'epoch': 11.12} +04/20 [03:25:43] INFO | >> train_qwenlatent.py:487 + Step 44080 | grad_norm_pre_clip=0.1233 | + grad_norm_pre_clip_avg=0.1475 | Metrics: + {'align_loss': 0.024874569848179817, + 'recon_loss': 0.13308875262737274, + 'predict_loss': 0.006865916773676872, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12332753837108612, + 'data_time': 0.0007148449949454516, + 'model_time': 1.2375902619969565, + 'grad_norm_pre_clip_avg': 0.1475306585431099, + 'learning_rate': 1.0764119881245323e-06, + 'epoch': 11.12} +04/20 [03:25:55] INFO | >> train_qwenlatent.py:487 + Step 44090 | grad_norm_pre_clip=0.0986 | + grad_norm_pre_clip_avg=0.1238 | Metrics: + {'align_loss': 0.025801820680499077, + 'recon_loss': 0.1869158297777176, + 'predict_loss': 0.007244242820888758, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.0985865443944931, + 'data_time': 0.0009888210042845458, + 'model_time': 1.2111651279847138, + 'grad_norm_pre_clip_avg': 0.12380697652697563, + 'learning_rate': 1.0729134136002675e-06, + 'epoch': 11.13} +04/20 [03:26:08] INFO | >> train_qwenlatent.py:487 + Step 44100 | grad_norm_pre_clip=0.1814 | + grad_norm_pre_clip_avg=0.1635 | Metrics: + {'align_loss': 0.024633143097162247, + 'recon_loss': 0.13816477358341217, + 'predict_loss': 0.010089193470776081, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18144121766090393, + 'mae_score': 0.0053982000093202335, + 'data_time': 0.0012739899975713342, + 'model_time': 1.3040751380030997, + 'grad_norm_pre_clip_avg': 0.16348650082945823, + 'learning_rate': 1.0694204145915327e-06, + 'epoch': 11.13} +04/20 [03:26:21] INFO | >> train_qwenlatent.py:487 + Step 44110 | grad_norm_pre_clip=0.1853 | + grad_norm_pre_clip_avg=0.1359 | Metrics: + {'align_loss': 0.024888763204216957, + 'recon_loss': 0.1583274006843567, + 'predict_loss': 0.008092845790088177, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18530991673469543, + 'data_time': 0.0008849309815559536, + 'model_time': 1.2453855350031517, + 'grad_norm_pre_clip_avg': 0.13592743799090384, + 'learning_rate': 1.0659329928007797e-06, + 'epoch': 11.13} +04/20 [03:26:34] INFO | >> train_qwenlatent.py:487 + Step 44120 | grad_norm_pre_clip=0.1367 | + grad_norm_pre_clip_avg=0.1467 | Metrics: + {'align_loss': 0.026358792558312416, + 'recon_loss': 0.11582271009683609, + 'predict_loss': 0.004160342738032341, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13667048513889313, + 'data_time': 0.0007300039869733155, + 'model_time': 1.188362421002239, + 'grad_norm_pre_clip_avg': 0.1467300422489643, + 'learning_rate': 1.06245114992773e-06, 'epoch': + 11.13} +04/20 [03:26:47] INFO | >> train_qwenlatent.py:487 + Step 44130 | grad_norm_pre_clip=0.1340 | + grad_norm_pre_clip_avg=0.1444 | Metrics: + {'align_loss': 0.026219282299280167, + 'recon_loss': 0.1352292150259018, + 'predict_loss': 0.0037758518010377884, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13398341834545135, + 'data_time': 0.0009386019955854863, + 'model_time': 1.2764620409870986, + 'grad_norm_pre_clip_avg': 0.14435217380523682, + 'learning_rate': 1.0589748876693921e-06, + 'epoch': 11.14} +04/20 [03:26:59] INFO | >> train_qwenlatent.py:487 + Step 44140 | grad_norm_pre_clip=0.1783 | + grad_norm_pre_clip_avg=0.1517 | Metrics: + {'align_loss': 0.02406804822385311, + 'recon_loss': 0.11443614214658737, + 'predict_loss': 0.0042473869398236275, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17831192910671234, + 'data_time': 0.0006787989987060428, + 'model_time': 1.238919788011117, + 'grad_norm_pre_clip_avg': 0.15170370489358903, + 'learning_rate': 1.0555042077200536e-06, + 'epoch': 11.14} +04/20 [03:27:13] INFO | >> train_qwenlatent.py:487 + Step 44150 | grad_norm_pre_clip=0.1333 | + grad_norm_pre_clip_avg=0.1434 | Metrics: + {'align_loss': 0.024400776252150536, + 'recon_loss': 0.06415243446826935, + 'predict_loss': 0.003939216956496239, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13328348100185394, + 'mae_score': 0.005409443915427268, 'data_time': + 0.0008934689976740628, 'model_time': + 1.4567380209919065, 'grad_norm_pre_clip_avg': + 0.14342668429017066, 'learning_rate': + 1.0520391117712857e-06, 'epoch': 11.14} +04/20 [03:27:25] INFO | >> train_qwenlatent.py:487 + Step 44160 | grad_norm_pre_clip=0.1240 | + grad_norm_pre_clip_avg=0.1375 | Metrics: + {'align_loss': 0.026269208639860153, + 'recon_loss': 0.1254734992980957, + 'predict_loss': 0.004123621620237827, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12404374033212662, + 'data_time': 0.0011693949927575886, + 'model_time': 1.2921756510040723, + 'grad_norm_pre_clip_avg': 0.1375417649745941, + 'learning_rate': 1.0485796015119308e-06, + 'epoch': 11.14} +04/20 [03:27:38] INFO | >> train_qwenlatent.py:487 + Step 44170 | grad_norm_pre_clip=0.1593 | + grad_norm_pre_clip_avg=0.1434 | Metrics: + {'align_loss': 0.02657056786119938, + 'recon_loss': 0.13652154803276062, + 'predict_loss': 0.004543642979115248, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15929941833019257, + 'data_time': 0.0007195289945229888, + 'model_time': 1.1935633009998128, + 'grad_norm_pre_clip_avg': 0.14335467889904976, + 'learning_rate': 1.0451256786281136e-06, + 'epoch': 11.15} +04/20 [03:27:51] INFO | >> train_qwenlatent.py:487 + Step 44180 | grad_norm_pre_clip=0.1199 | + grad_norm_pre_clip_avg=0.1624 | Metrics: + {'align_loss': 0.025578636676073074, + 'recon_loss': 0.15125073492527008, + 'predict_loss': 0.006813047919422388, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11985712498426437, + 'data_time': 0.001215202995808795, + 'model_time': 1.2438891639758367, + 'grad_norm_pre_clip_avg': 0.16236855909228326, + 'learning_rate': 1.0416773448032321e-06, + 'epoch': 11.15} +04/20 [03:28:03] INFO | >> train_qwenlatent.py:487 + Step 44190 | grad_norm_pre_clip=0.1563 | + grad_norm_pre_clip_avg=0.1401 | Metrics: + {'align_loss': 0.025511484593153, 'recon_loss': + 0.17742273211479187, 'predict_loss': + 0.006173586007207632, 'aux_loss_decay_weight': + 0.0, 'grad_norm_pre_clip': 0.1562572419643402, + 'data_time': 0.000652474001981318, + 'model_time': 1.25771516902023, + 'grad_norm_pre_clip_avg': 0.14011653512716293, + 'learning_rate': 1.038234601717962e-06, + 'epoch': 11.15} +04/20 [03:28:17] INFO | >> train_qwenlatent.py:487 + Step 44200 | grad_norm_pre_clip=0.0986 | + grad_norm_pre_clip_avg=0.1323 | Metrics: + {'align_loss': 0.02516096644103527, + 'recon_loss': 0.12149130553007126, + 'predict_loss': 0.007660649251192808, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09862486273050308, + 'mae_score': 0.005168771743774414, 'data_time': + 0.0009497799910604954, 'model_time': + 1.2029501940123737, 'grad_norm_pre_clip_avg': + 0.13232969045639037, 'learning_rate': + 1.0347974510502608e-06, 'epoch': 11.15} +04/20 [03:28:29] INFO | >> train_qwenlatent.py:487 + Step 44210 | grad_norm_pre_clip=0.1266 | + grad_norm_pre_clip_avg=0.1450 | Metrics: + {'align_loss': 0.025682270526885986, + 'recon_loss': 0.16145069897174835, + 'predict_loss': 0.006856394000351429, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1265653371810913, + 'data_time': 0.0010950569994747639, + 'model_time': 1.2890592490148265, + 'grad_norm_pre_clip_avg': 0.14495740830898285, + 'learning_rate': 1.0313658944753476e-06, + 'epoch': 11.16} +04/20 [03:28:42] INFO | >> train_qwenlatent.py:487 + Step 44220 | grad_norm_pre_clip=0.1810 | + grad_norm_pre_clip_avg=0.1533 | Metrics: + {'align_loss': 0.02455943636596203, + 'recon_loss': 0.15017762780189514, + 'predict_loss': 0.007647506892681122, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18103830516338348, + 'data_time': 0.0012212739966344088, + 'model_time': 1.3132316400005948, + 'grad_norm_pre_clip_avg': 0.15326731130480767, + 'learning_rate': 1.0279399336657228e-06, + 'epoch': 11.16} +04/20 [03:28:54] INFO | >> train_qwenlatent.py:487 + Step 44230 | grad_norm_pre_clip=0.1335 | + grad_norm_pre_clip_avg=0.1575 | Metrics: + {'align_loss': 0.025367097929120064, + 'recon_loss': 0.1321839988231659, + 'predict_loss': 0.007792746182531118, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13353492319583893, + 'data_time': 0.0010355249978601933, + 'model_time': 1.2294639030005783, + 'grad_norm_pre_clip_avg': 0.15754834860563277, + 'learning_rate': 1.0245195702911576e-06, + 'epoch': 11.16} +04/20 [03:29:07] INFO | >> train_qwenlatent.py:487 + Step 44240 | grad_norm_pre_clip=0.1619 | + grad_norm_pre_clip_avg=0.1426 | Metrics: + {'align_loss': 0.024413123726844788, + 'recon_loss': 0.14797954261302948, + 'predict_loss': 0.006692494731396437, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16193310916423798, + 'data_time': 0.0006939169834367931, + 'model_time': 1.2185603480029386, + 'grad_norm_pre_clip_avg': 0.14256566390395164, + 'learning_rate': 1.0211048060186935e-06, + 'epoch': 11.16} +04/20 [03:29:20] INFO | >> train_qwenlatent.py:487 + Step 44250 | grad_norm_pre_clip=0.1500 | + grad_norm_pre_clip_avg=0.1382 | Metrics: + {'align_loss': 0.02489793300628662, + 'recon_loss': 0.15394353866577148, + 'predict_loss': 0.012101087719202042, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14998532831668854, + 'mae_score': 0.00499620179872255, 'data_time': + 0.0006976289732847363, 'model_time': + 1.199485524004558, 'grad_norm_pre_clip_avg': + 0.13819260597229005, 'learning_rate': + 1.017695642512651e-06, 'epoch': 11.17} +04/20 [03:29:32] INFO | >> train_qwenlatent.py:487 + Step 44260 | grad_norm_pre_clip=0.1618 | + grad_norm_pre_clip_avg=0.1580 | Metrics: + {'align_loss': 0.024209173396229744, + 'recon_loss': 0.12691408395767212, + 'predict_loss': 0.005075667053461075, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.161835178732872, + 'data_time': 0.0009960670140571892, + 'model_time': 1.3407348909822758, + 'grad_norm_pre_clip_avg': 0.1580163449048996, + 'learning_rate': 1.014292081434612e-06, + 'epoch': 11.17} +04/20 [03:29:45] INFO | >> train_qwenlatent.py:487 + Step 44270 | grad_norm_pre_clip=0.1254 | + grad_norm_pre_clip_avg=0.1320 | Metrics: + {'align_loss': 0.025806523859500885, + 'recon_loss': 0.14348778128623962, + 'predict_loss': 0.006456085946410894, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12543609738349915, + 'data_time': 0.0010984810069203377, + 'model_time': 1.471110165992286, + 'grad_norm_pre_clip_avg': 0.1320285201072693, + 'learning_rate': 1.0108941244434306e-06, + 'epoch': 11.17} +04/20 [03:29:58] INFO | >> train_qwenlatent.py:487 + Step 44280 | grad_norm_pre_clip=0.1428 | + grad_norm_pre_clip_avg=0.1367 | Metrics: + {'align_loss': 0.02596590667963028, + 'recon_loss': 0.15014319121837616, + 'predict_loss': 0.006804823875427246, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14283190667629242, + 'data_time': 0.0006040569860488176, + 'model_time': 1.2405942550103646, + 'grad_norm_pre_clip_avg': 0.13665081337094306, + 'learning_rate': 1.0075017731952306e-06, + 'epoch': 11.17} +04/20 [03:30:11] INFO | >> train_qwenlatent.py:487 + Step 44290 | grad_norm_pre_clip=0.1269 | + grad_norm_pre_clip_avg=0.1461 | Metrics: + {'align_loss': 0.025676880031824112, + 'recon_loss': 0.17107662558555603, + 'predict_loss': 0.00775702390819788, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12690967321395874, + 'data_time': 0.0010324819886591285, + 'model_time': 1.218267324991757, + 'grad_norm_pre_clip_avg': 0.14614434838294982, + 'learning_rate': 1.0041150293434003e-06, + 'epoch': 11.18} +04/20 [03:30:24] INFO | >> train_qwenlatent.py:487 + Step 44300 | grad_norm_pre_clip=0.1414 | + grad_norm_pre_clip_avg=0.1523 | Metrics: + {'align_loss': 0.02492297999560833, + 'recon_loss': 0.18431441485881805, + 'predict_loss': 0.005946595221757889, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14144901931285858, + 'mae_score': 0.005747314401575037, 'data_time': + 0.0009322179830633104, 'model_time': + 1.2344874319969676, 'grad_norm_pre_clip_avg': + 0.15230136290192603, 'learning_rate': + 1.0007338945386013e-06, 'epoch': 11.18} +04/20 [03:30:37] INFO | >> train_qwenlatent.py:487 + Step 44310 | grad_norm_pre_clip=0.1140 | + grad_norm_pre_clip_avg=0.1324 | Metrics: + {'align_loss': 0.02553587406873703, + 'recon_loss': 0.17527128756046295, + 'predict_loss': 0.006332764867693186, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11402180790901184, + 'data_time': 0.0006988179811742157, + 'model_time': 1.2297882910061162, + 'grad_norm_pre_clip_avg': 0.13237211033701896, + 'learning_rate': 9.97358370428754e-07, 'epoch': + 11.18} +04/20 [03:30:49] INFO | >> train_qwenlatent.py:487 + Step 44320 | grad_norm_pre_clip=0.1289 | + grad_norm_pre_clip_avg=0.1395 | Metrics: + {'align_loss': 0.026126688346266747, + 'recon_loss': 0.2054365575313568, + 'predict_loss': 0.013579841703176498, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12886549532413483, + 'data_time': 0.001046938996296376, + 'model_time': 1.246045434003463, + 'grad_norm_pre_clip_avg': 0.1394959457218647, + 'learning_rate': 9.939884586590503e-07, + 'epoch': 11.18} +04/20 [03:31:02] INFO | >> train_qwenlatent.py:487 + Step 44330 | grad_norm_pre_clip=0.1509 | + grad_norm_pre_clip_avg=0.1623 | Metrics: + {'align_loss': 0.025291651487350464, + 'recon_loss': 0.1477455049753189, + 'predict_loss': 0.005842485465109348, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1508987843990326, + 'data_time': 0.0006716829957440495, + 'model_time': 1.1725556509918533, + 'grad_norm_pre_clip_avg': 0.16226573362946511, + 'learning_rate': 9.906241608719402e-07, + 'epoch': 11.19} +04/20 [03:31:15] INFO | >> train_qwenlatent.py:487 + Step 44340 | grad_norm_pre_clip=0.1099 | + grad_norm_pre_clip_avg=0.1368 | Metrics: + {'align_loss': 0.026057681068778038, + 'recon_loss': 0.1599477082490921, + 'predict_loss': 0.007286700885742903, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10993809998035431, + 'data_time': 0.0006734960188623518, + 'model_time': 1.221596352988854, + 'grad_norm_pre_clip_avg': 0.13676266595721245, + 'learning_rate': 9.872654787071478e-07, + 'epoch': 11.19} +04/20 [03:31:28] INFO | >> train_qwenlatent.py:487 + Step 44350 | grad_norm_pre_clip=0.1553 | + grad_norm_pre_clip_avg=0.1420 | Metrics: + {'align_loss': 0.025652974843978882, + 'recon_loss': 0.15190325677394867, + 'predict_loss': 0.005832545459270477, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15527227520942688, + 'mae_score': 0.006028735959852064, 'data_time': + 0.0008178779971785843, 'model_time': + 1.1970899820153136, 'grad_norm_pre_clip_avg': + 0.14197953119874002, 'learning_rate': + 9.83912413801651e-07, 'epoch': 11.19} +04/20 [03:31:40] INFO | >> train_qwenlatent.py:487 + Step 44360 | grad_norm_pre_clip=0.1749 | + grad_norm_pre_clip_avg=0.1662 | Metrics: + {'align_loss': 0.02546302229166031, + 'recon_loss': 0.15648017823696136, + 'predict_loss': 0.005743296351283789, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1748642921447754, + 'data_time': 0.0006699849909637123, + 'model_time': 1.1933757759979926, + 'grad_norm_pre_clip_avg': 0.16621328815817832, + 'learning_rate': 9.805649677896931e-07, + 'epoch': 11.19} +04/20 [03:31:53] INFO | >> train_qwenlatent.py:487 + Step 44370 | grad_norm_pre_clip=0.1193 | + grad_norm_pre_clip_avg=0.1438 | Metrics: + {'align_loss': 0.02418786846101284, + 'recon_loss': 0.12099755555391312, + 'predict_loss': 0.003474615281447768, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11931072920560837, + 'data_time': 0.0009780940017662942, + 'model_time': 1.3434252919978462, + 'grad_norm_pre_clip_avg': 0.1437804713845253, + 'learning_rate': 9.772231423027775e-07, + 'epoch': 11.2} +04/20 [03:32:05] INFO | >> train_qwenlatent.py:487 + Step 44380 | grad_norm_pre_clip=0.1150 | + grad_norm_pre_clip_avg=0.1489 | Metrics: + {'align_loss': 0.025204073637723923, + 'recon_loss': 0.12702596187591553, + 'predict_loss': 0.00432552071288228, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11497921496629715, + 'data_time': 0.0012776050134561956, + 'model_time': 1.2274566190026235, + 'grad_norm_pre_clip_avg': 0.14890034794807433, + 'learning_rate': 9.738869389696708e-07, + 'epoch': 11.2} +04/20 [03:32:18] INFO | >> train_qwenlatent.py:487 + Step 44390 | grad_norm_pre_clip=0.1856 | + grad_norm_pre_clip_avg=0.1499 | Metrics: + {'align_loss': 0.02622063085436821, + 'recon_loss': 0.15793931484222412, + 'predict_loss': 0.008113215677440166, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18558932840824127, + 'data_time': 0.0007240639824885875, + 'model_time': 1.2380633449938614, + 'grad_norm_pre_clip_avg': 0.14990061596035958, + 'learning_rate': 9.705563594163968e-07, + 'epoch': 11.2} +04/20 [03:32:31] INFO | >> train_qwenlatent.py:487 + Step 44400 | grad_norm_pre_clip=0.1184 | + grad_norm_pre_clip_avg=0.1351 | Metrics: + {'align_loss': 0.025496158748865128, + 'recon_loss': 0.22786478698253632, + 'predict_loss': 0.01044988352805376, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11842155456542969, + 'mae_score': 0.005415355407439911, 'data_time': + 0.0006380220002029091, 'model_time': + 1.2178933690011036, 'grad_norm_pre_clip_avg': + 0.13513541147112845, 'learning_rate': + 9.672314052662412e-07, 'epoch': 11.2} +04/20 [03:32:44] INFO | >> train_qwenlatent.py:487 + Step 44410 | grad_norm_pre_clip=0.1375 | + grad_norm_pre_clip_avg=0.1250 | Metrics: + {'align_loss': 0.025877147912979126, + 'recon_loss': 0.14002762734889984, + 'predict_loss': 0.00793981272727251, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13754428923130035, + 'data_time': 0.0012243420060258359, + 'model_time': 1.4949725210026372, + 'grad_norm_pre_clip_avg': 0.12496415004134179, + 'learning_rate': 9.639120781397443e-07, + 'epoch': 11.21} +04/20 [03:32:57] INFO | >> train_qwenlatent.py:487 + Step 44420 | grad_norm_pre_clip=0.1121 | + grad_norm_pre_clip_avg=0.1332 | Metrics: + {'align_loss': 0.02563704177737236, + 'recon_loss': 0.08646301925182343, + 'predict_loss': 0.0030298708006739616, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11211314797401428, + 'data_time': 0.0009586560190655291, + 'model_time': 1.2865599069918972, + 'grad_norm_pre_clip_avg': 0.1332256332039833, + 'learning_rate': 9.605983796547066e-07, + 'epoch': 11.21} +04/20 [03:33:10] INFO | >> train_qwenlatent.py:487 + Step 44430 | grad_norm_pre_clip=0.1043 | + grad_norm_pre_clip_avg=0.1165 | Metrics: + {'align_loss': 0.025005050003528595, + 'recon_loss': 0.11789391934871674, + 'predict_loss': 0.0038639637641608715, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10425610840320587, + 'data_time': 0.0006892950041219592, + 'model_time': 1.1941757039749064, + 'grad_norm_pre_clip_avg': 0.11647642254829407, + 'learning_rate': 9.572903114261834e-07, + 'epoch': 11.21} +04/20 [03:33:22] INFO | >> train_qwenlatent.py:487 + Step 44440 | grad_norm_pre_clip=0.1899 | + grad_norm_pre_clip_avg=0.1700 | Metrics: + {'align_loss': 0.024366270750761032, + 'recon_loss': 0.12713564932346344, + 'predict_loss': 0.006344308611005545, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18992239236831665, + 'data_time': 0.0006647479895036668, + 'model_time': 1.29518059300608, + 'grad_norm_pre_clip_avg': 0.1699611574411392, + 'learning_rate': 9.539878750664894e-07, + 'epoch': 11.21} +04/20 [03:33:35] INFO | >> train_qwenlatent.py:487 + Step 44450 | grad_norm_pre_clip=0.1476 | + grad_norm_pre_clip_avg=0.1690 | Metrics: + {'align_loss': 0.025871867313981056, + 'recon_loss': 0.17577479779720306, + 'predict_loss': 0.00844869576394558, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1476050764322281, + 'mae_score': 0.005012312880507461, 'data_time': + 0.0010493730078451335, 'model_time': + 1.2151383500022348, 'grad_norm_pre_clip_avg': + 0.16895781457424164, 'learning_rate': + 9.506910721851909e-07, 'epoch': 11.22} +04/20 [03:33:47] INFO | >> train_qwenlatent.py:487 + Step 44460 | grad_norm_pre_clip=0.2122 | + grad_norm_pre_clip_avg=0.1629 | Metrics: + {'align_loss': 0.026190657168626785, + 'recon_loss': 0.1211375743150711, + 'predict_loss': 0.0058018541894853115, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.21217605471611023, + 'data_time': 0.0008755779999773949, + 'model_time': 1.2616830920160282, + 'grad_norm_pre_clip_avg': 0.16286215409636498, + 'learning_rate': 9.473999043891097e-07, + 'epoch': 11.22} +04/20 [03:34:00] INFO | >> train_qwenlatent.py:487 + Step 44470 | grad_norm_pre_clip=0.1416 | + grad_norm_pre_clip_avg=0.1451 | Metrics: + {'align_loss': 0.025790467858314514, + 'recon_loss': 0.258758544921875, + 'predict_loss': 0.009883605875074863, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14160405099391937, + 'data_time': 0.0009111640101764351, + 'model_time': 1.207231568027055, + 'grad_norm_pre_clip_avg': 0.1450985014438629, + 'learning_rate': 9.441143732823197e-07, + 'epoch': 11.22} +04/20 [03:34:13] INFO | >> train_qwenlatent.py:487 + Step 44480 | grad_norm_pre_clip=0.1849 | + grad_norm_pre_clip_avg=0.1609 | Metrics: + {'align_loss': 0.02520708739757538, + 'recon_loss': 0.1990666687488556, + 'predict_loss': 0.010108922608196735, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18493305146694183, + 'data_time': 0.0006572760175913572, + 'model_time': 1.2076379409991205, + 'grad_norm_pre_clip_avg': 0.16093271002173423, + 'learning_rate': 9.40834480466151e-07, 'epoch': + 11.22} +04/20 [03:34:25] INFO | >> train_qwenlatent.py:487 + Step 44490 | grad_norm_pre_clip=0.1370 | + grad_norm_pre_clip_avg=0.1522 | Metrics: + {'align_loss': 0.02491816133260727, + 'recon_loss': 0.16987277567386627, + 'predict_loss': 0.009395753964781761, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13695107400417328, + 'data_time': 0.0009782779961824417, + 'model_time': 1.2154629620199557, + 'grad_norm_pre_clip_avg': 0.1521601215004921, + 'learning_rate': 9.375602275391817e-07, + 'epoch': 11.23} +04/20 [03:34:38] INFO | >> train_qwenlatent.py:487 + Step 44500 | grad_norm_pre_clip=0.1381 | + grad_norm_pre_clip_avg=0.1311 | Metrics: + {'align_loss': 0.024516362696886063, + 'recon_loss': 0.12369397282600403, + 'predict_loss': 0.0054682414047420025, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13805799186229706, + 'mae_score': 0.005760228956067885, 'data_time': + 0.0008530199993401766, 'model_time': + 1.227818876010133, 'grad_norm_pre_clip_avg': + 0.13113644272089003, 'learning_rate': + 9.342916160972435e-07, 'epoch': 11.23} +04/20 [03:34:51] INFO | >> train_qwenlatent.py:487 + Step 44510 | grad_norm_pre_clip=0.1726 | + grad_norm_pre_clip_avg=0.1395 | Metrics: + {'align_loss': 0.02579779550433159, + 'recon_loss': 0.26118388772010803, + 'predict_loss': 0.015023800544440746, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17259547114372253, + 'data_time': 0.0008375849865842611, + 'model_time': 1.2583914960268885, + 'grad_norm_pre_clip_avg': 0.13945344015955924, + 'learning_rate': 9.310286477334176e-07, + 'epoch': 11.23} +04/20 [03:35:03] INFO | >> train_qwenlatent.py:487 + Step 44520 | grad_norm_pre_clip=0.1665 | + grad_norm_pre_clip_avg=0.1299 | Metrics: + {'align_loss': 0.024707861244678497, + 'recon_loss': 0.11712297797203064, + 'predict_loss': 0.006254161242395639, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16650496423244476, + 'data_time': 0.0010948179988190532, + 'model_time': 1.2358277470048051, + 'grad_norm_pre_clip_avg': 0.1298509880900383, + 'learning_rate': 9.277713240380343e-07, + 'epoch': 11.23} +04/20 [03:35:16] INFO | >> train_qwenlatent.py:487 + Step 44530 | grad_norm_pre_clip=0.1226 | + grad_norm_pre_clip_avg=0.1569 | Metrics: + {'align_loss': 0.02470722794532776, + 'recon_loss': 0.1332160234451294, + 'predict_loss': 0.006896378006786108, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12262710183858871, + 'data_time': 0.0007287950138561428, + 'model_time': 1.2512599710025825, + 'grad_norm_pre_clip_avg': 0.15694464445114137, + 'learning_rate': 9.245196465986764e-07, + 'epoch': 11.24} +04/20 [03:35:29] INFO | >> train_qwenlatent.py:487 + Step 44540 | grad_norm_pre_clip=0.1350 | + grad_norm_pre_clip_avg=0.1366 | Metrics: + {'align_loss': 0.02535185217857361, + 'recon_loss': 0.13175053894519806, + 'predict_loss': 0.004601123742759228, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.135044127702713, + 'data_time': 0.0012103129993192852, + 'model_time': 1.2022341249976307, + 'grad_norm_pre_clip_avg': 0.13660204485058786, + 'learning_rate': 9.212736170001706e-07, + 'epoch': 11.24} +04/20 [03:35:42] INFO | >> train_qwenlatent.py:487 + Step 44550 | grad_norm_pre_clip=0.1645 | + grad_norm_pre_clip_avg=0.1579 | Metrics: + {'align_loss': 0.02488592639565468, + 'recon_loss': 0.18085572123527527, + 'predict_loss': 0.00813319068402052, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16447193920612335, + 'mae_score': 0.0057621109593021975, + 'data_time': 0.0009541509789414704, + 'model_time': 1.55948559200624, + 'grad_norm_pre_clip_avg': 0.1579097829759121, + 'learning_rate': 9.180332368245925e-07, + 'epoch': 11.24} +04/20 [03:35:54] INFO | >> train_qwenlatent.py:487 + Step 44560 | grad_norm_pre_clip=0.1191 | + grad_norm_pre_clip_avg=0.1396 | Metrics: + {'align_loss': 0.026249229907989502, + 'recon_loss': 0.17462120950222015, + 'predict_loss': 0.009981860406696796, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11912857741117477, + 'data_time': 0.0008104070147965103, + 'model_time': 1.2348198169784155, + 'grad_norm_pre_clip_avg': 0.1395798921585083, + 'learning_rate': 9.147985076512648e-07, + 'epoch': 11.24} +04/20 [03:36:07] INFO | >> train_qwenlatent.py:487 + Step 44570 | grad_norm_pre_clip=0.1617 | + grad_norm_pre_clip_avg=0.1364 | Metrics: + {'align_loss': 0.024083727970719337, + 'recon_loss': 0.12445082515478134, + 'predict_loss': 0.01001323014497757, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16171106696128845, + 'data_time': 0.0012742540275212377, + 'model_time': 1.22702891001245, + 'grad_norm_pre_clip_avg': 0.1363629035651684, + 'learning_rate': 9.115694310567498e-07, + 'epoch': 11.25} +04/20 [03:36:20] INFO | >> train_qwenlatent.py:487 + Step 44580 | grad_norm_pre_clip=0.1562 | + grad_norm_pre_clip_avg=0.1456 | Metrics: + {'align_loss': 0.025180768221616745, + 'recon_loss': 0.11007200181484222, + 'predict_loss': 0.004976870026439428, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15624351799488068, + 'data_time': 0.0010107390116900206, + 'model_time': 1.2587524470000062, + 'grad_norm_pre_clip_avg': 0.14559537619352342, + 'learning_rate': 9.083460086148675e-07, + 'epoch': 11.25} +04/20 [03:36:32] INFO | >> train_qwenlatent.py:487 + Step 44590 | grad_norm_pre_clip=0.1149 | + grad_norm_pre_clip_avg=0.1465 | Metrics: + {'align_loss': 0.02613070420920849, + 'recon_loss': 0.184061661362648, + 'predict_loss': 0.011743267066776752, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11492304503917694, + 'data_time': 0.0006785290024708956, + 'model_time': 1.215693498001201, + 'grad_norm_pre_clip_avg': 0.14649728685617447, + 'learning_rate': 9.051282418966707e-07, + 'epoch': 11.25} +04/20 [03:36:45] INFO | >> train_qwenlatent.py:487 + Step 44600 | grad_norm_pre_clip=0.1445 | + grad_norm_pre_clip_avg=0.1441 | Metrics: + {'align_loss': 0.024559780955314636, + 'recon_loss': 0.12244907766580582, + 'predict_loss': 0.00422299187630415, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14446145296096802, + 'mae_score': 0.006550785442730328, 'data_time': + 0.000653783994494006, 'model_time': + 1.1800522329867817, 'grad_norm_pre_clip_avg': + 0.14408457055687904, 'learning_rate': + 9.019161324704613e-07, 'epoch': 11.25} +04/20 [03:36:59] INFO | >> train_qwenlatent.py:487 + Step 44610 | grad_norm_pre_clip=0.1277 | + grad_norm_pre_clip_avg=0.1372 | Metrics: + {'align_loss': 0.025264594703912735, + 'recon_loss': 0.11581440269947052, + 'predict_loss': 0.007002778351306915, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12770575284957886, + 'data_time': 0.0008495880174450576, + 'model_time': 1.2608841069741175, + 'grad_norm_pre_clip_avg': 0.13719706013798713, + 'learning_rate': 8.987096819017828e-07, + 'epoch': 11.26} +04/20 [03:37:11] INFO | >> train_qwenlatent.py:487 + Step 44620 | grad_norm_pre_clip=0.1249 | + grad_norm_pre_clip_avg=0.1444 | Metrics: + {'align_loss': 0.024693237617611885, + 'recon_loss': 0.11546551436185837, + 'predict_loss': 0.004870547913014889, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12487351149320602, + 'data_time': 0.0007636859954800457, + 'model_time': 1.2418173399928492, + 'grad_norm_pre_clip_avg': 0.1444016508758068, + 'learning_rate': 8.955088917534165e-07, + 'epoch': 11.26} +04/20 [03:37:24] INFO | >> train_qwenlatent.py:487 + Step 44630 | grad_norm_pre_clip=0.1829 | + grad_norm_pre_clip_avg=0.1506 | Metrics: + {'align_loss': 0.02414313703775406, + 'recon_loss': 0.1324993371963501, + 'predict_loss': 0.007294546812772751, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18293063342571259, + 'data_time': 0.0009569209942128509, + 'model_time': 1.2815098190039862, + 'grad_norm_pre_clip_avg': 0.15058044344186783, + 'learning_rate': 8.923137635853933e-07, + 'epoch': 11.26} +04/20 [03:37:36] INFO | >> train_qwenlatent.py:487 + Step 44640 | grad_norm_pre_clip=0.1134 | + grad_norm_pre_clip_avg=0.1324 | Metrics: + {'align_loss': 0.024794377386569977, + 'recon_loss': 0.1291658729314804, + 'predict_loss': 0.004470914602279663, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11337140947580338, + 'data_time': 0.0007917829789221287, + 'model_time': 1.2254951579961926, + 'grad_norm_pre_clip_avg': 0.1324453331530094, + 'learning_rate': 8.891242989549785e-07, + 'epoch': 11.26} +04/20 [03:37:49] INFO | >> train_qwenlatent.py:487 + Step 44650 | grad_norm_pre_clip=0.1425 | + grad_norm_pre_clip_avg=0.1333 | Metrics: + {'align_loss': 0.024850795045495033, + 'recon_loss': 0.16668877005577087, + 'predict_loss': 0.008879177272319794, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1424514353275299, + 'mae_score': 0.005005403467126795, 'data_time': + 0.0008369479910470545, 'model_time': + 1.2099689139868133, 'grad_norm_pre_clip_avg': + 0.1332804776728153, 'learning_rate': + 8.859404994166796e-07, 'epoch': 11.27} +04/20 [03:38:02] INFO | >> train_qwenlatent.py:487 + Step 44660 | grad_norm_pre_clip=0.1689 | + grad_norm_pre_clip_avg=0.1493 | Metrics: + {'align_loss': 0.02561219036579132, + 'recon_loss': 0.19433805346488953, + 'predict_loss': 0.012225642800331116, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16894352436065674, + 'data_time': 0.0009046579943969846, + 'model_time': 1.2296893750026356, + 'grad_norm_pre_clip_avg': 0.14934940859675408, + 'learning_rate': 8.827623665222367e-07, + 'epoch': 11.27} +04/20 [03:38:15] INFO | >> train_qwenlatent.py:487 + Step 44670 | grad_norm_pre_clip=0.1404 | + grad_norm_pre_clip_avg=0.1455 | Metrics: + {'align_loss': 0.02482619695365429, + 'recon_loss': 0.09952808171510696, + 'predict_loss': 0.0038936242926865816, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14044323563575745, + 'data_time': 0.0009831800125539303, + 'model_time': 1.2741166219930165, + 'grad_norm_pre_clip_avg': 0.14552758634090424, + 'learning_rate': 8.795899018206385e-07, + 'epoch': 11.27} +04/20 [03:38:27] INFO | >> train_qwenlatent.py:487 + Step 44680 | grad_norm_pre_clip=0.1795 | + grad_norm_pre_clip_avg=0.1397 | Metrics: + {'align_loss': 0.02609322965145111, + 'recon_loss': 0.20125064253807068, + 'predict_loss': 0.008805549703538418, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17952243983745575, + 'data_time': 0.0008204769983422011, + 'model_time': 1.2236041619908065, + 'grad_norm_pre_clip_avg': 0.1397413991391659, + 'learning_rate': 8.764231068581041e-07, + 'epoch': 11.27} +04/20 [03:38:40] INFO | >> train_qwenlatent.py:487 + Step 44690 | grad_norm_pre_clip=0.1409 | + grad_norm_pre_clip_avg=0.1318 | Metrics: + {'align_loss': 0.02659706212580204, + 'recon_loss': 0.19723854959011078, + 'predict_loss': 0.008111386559903622, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14088326692581177, + 'data_time': 0.0009145310032181442, + 'model_time': 1.4713665840099566, + 'grad_norm_pre_clip_avg': 0.13184326589107515, + 'learning_rate': 8.732619831780909e-07, + 'epoch': 11.28} +04/20 [03:38:54] INFO | >> train_qwenlatent.py:487 + Step 44700 | grad_norm_pre_clip=0.1496 | + grad_norm_pre_clip_avg=0.1399 | Metrics: + {'align_loss': 0.024962225928902626, + 'recon_loss': 0.09855570644140244, + 'predict_loss': 0.0029674917459487915, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1495969146490097, + 'mae_score': 0.006105214196282464, 'data_time': + 0.0010402100160717964, 'model_time': + 1.2928641229809728, 'grad_norm_pre_clip_avg': + 0.13988296911120415, 'learning_rate': + 8.701065323212918e-07, 'epoch': 11.28} +04/20 [03:39:06] INFO | >> train_qwenlatent.py:487 + Step 44710 | grad_norm_pre_clip=0.1032 | + grad_norm_pre_clip_avg=0.1315 | Metrics: + {'align_loss': 0.02535632997751236, + 'recon_loss': 0.1471036970615387, + 'predict_loss': 0.0071920426562428474, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10324489325284958, + 'data_time': 0.000745772005757317, + 'model_time': 1.2559455370064825, + 'grad_norm_pre_clip_avg': 0.13152279779314996, + 'learning_rate': 8.669567558256363e-07, + 'epoch': 11.28} +04/20 [03:39:19] INFO | >> train_qwenlatent.py:487 + Step 44720 | grad_norm_pre_clip=0.1237 | + grad_norm_pre_clip_avg=0.1192 | Metrics: + {'align_loss': 0.02492455206811428, + 'recon_loss': 0.09024016559123993, + 'predict_loss': 0.002118792850524187, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12365028262138367, + 'data_time': 0.0009146119991783053, + 'model_time': 1.2478233429719694, + 'grad_norm_pre_clip_avg': 0.11919011399149895, + 'learning_rate': 8.638126552262831e-07, + 'epoch': 11.28} +04/20 [03:39:31] INFO | >> train_qwenlatent.py:487 + Step 44730 | grad_norm_pre_clip=0.1016 | + grad_norm_pre_clip_avg=0.1497 | Metrics: + {'align_loss': 0.0251114834100008, + 'recon_loss': 0.13199402391910553, + 'predict_loss': 0.006765137892216444, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1016196459531784, + 'data_time': 0.0007194929930847138, + 'model_time': 1.2463035989785567, + 'grad_norm_pre_clip_avg': 0.14970603585243225, + 'learning_rate': 8.606742320556369e-07, + 'epoch': 11.29} +04/20 [03:39:44] INFO | >> train_qwenlatent.py:487 + Step 44740 | grad_norm_pre_clip=0.1831 | + grad_norm_pre_clip_avg=0.1372 | Metrics: + {'align_loss': 0.02502988651394844, + 'recon_loss': 0.14409670233726501, + 'predict_loss': 0.006178778596222401, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1831219643354416, + 'data_time': 0.0007548109861090779, + 'model_time': 1.6018446839880198, + 'grad_norm_pre_clip_avg': 0.13718931153416633, + 'learning_rate': 8.575414878433193e-07, + 'epoch': 11.29} +04/20 [03:39:57] INFO | >> train_qwenlatent.py:487 + Step 44750 | grad_norm_pre_clip=0.0988 | + grad_norm_pre_clip_avg=0.1328 | Metrics: + {'align_loss': 0.025274544954299927, + 'recon_loss': 0.13294702768325806, + 'predict_loss': 0.005049556028097868, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.0987999215722084, + 'mae_score': 0.005242180609488272, 'data_time': + 0.0009629720007069409, 'model_time': + 1.2417039920110255, 'grad_norm_pre_clip_avg': + 0.13281595408916474, 'learning_rate': + 8.544144241161946e-07, 'epoch': 11.29} +04/20 [03:40:10] INFO | >> train_qwenlatent.py:487 + Step 44760 | grad_norm_pre_clip=0.1240 | + grad_norm_pre_clip_avg=0.1351 | Metrics: + {'align_loss': 0.026127159595489502, + 'recon_loss': 0.1710398942232132, + 'predict_loss': 0.00773863960057497, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12403767555952072, + 'data_time': 0.0010037960018962622, + 'model_time': 1.246332906972384, + 'grad_norm_pre_clip_avg': 0.13508565351366997, + 'learning_rate': 8.512930423983545e-07, + 'epoch': 11.29} +04/20 [03:40:22] INFO | >> train_qwenlatent.py:487 + Step 44770 | grad_norm_pre_clip=0.1591 | + grad_norm_pre_clip_avg=0.1331 | Metrics: + {'align_loss': 0.025639034807682037, + 'recon_loss': 0.15142644941806793, + 'predict_loss': 0.009028222411870956, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1591414213180542, + 'data_time': 0.0008744939987082034, + 'model_time': 1.2806490160000976, + 'grad_norm_pre_clip_avg': 0.1331186257302761, + 'learning_rate': 8.481773442111245e-07, + 'epoch': 11.3} +04/20 [03:40:35] INFO | >> train_qwenlatent.py:487 + Step 44780 | grad_norm_pre_clip=0.1744 | + grad_norm_pre_clip_avg=0.1500 | Metrics: + {'align_loss': 0.025597654283046722, + 'recon_loss': 0.15896904468536377, + 'predict_loss': 0.008915365673601627, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17442281544208527, + 'data_time': 0.000682067999150604, + 'model_time': 1.2152278700086754, + 'grad_norm_pre_clip_avg': 0.14998957738280297, + 'learning_rate': 8.450673310730574e-07, + 'epoch': 11.3} +04/20 [03:40:47] INFO | >> train_qwenlatent.py:487 + Step 44790 | grad_norm_pre_clip=0.1326 | + grad_norm_pre_clip_avg=0.1374 | Metrics: + {'align_loss': 0.02509954571723938, + 'recon_loss': 0.15572461485862732, + 'predict_loss': 0.0066098482348024845, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13262084126472473, + 'data_time': 0.0010711310023907572, + 'model_time': 1.2150684379739687, + 'grad_norm_pre_clip_avg': 0.13742181584239005, + 'learning_rate': 8.419630044999363e-07, + 'epoch': 11.3} +04/20 [03:41:01] INFO | >> train_qwenlatent.py:487 + Step 44800 | grad_norm_pre_clip=0.1281 | + grad_norm_pre_clip_avg=0.1307 | Metrics: + {'align_loss': 0.02542821317911148, + 'recon_loss': 0.13735002279281616, + 'predict_loss': 0.004740124102681875, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12812763452529907, + 'mae_score': 0.005842992421743032, 'data_time': + 0.001077091001207009, 'model_time': + 1.2470029259857256, 'grad_norm_pre_clip_avg': + 0.13070007264614106, 'learning_rate': + 8.388643660047697e-07, 'epoch': 11.3} +04/20 [03:41:13] INFO | >> train_qwenlatent.py:487 + Step 44810 | grad_norm_pre_clip=0.1860 | + grad_norm_pre_clip_avg=0.1222 | Metrics: + {'align_loss': 0.025720182806253433, + 'recon_loss': 0.11517951637506485, + 'predict_loss': 0.007233093027025461, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18598058819770813, + 'data_time': 0.0010280779970344156, + 'model_time': 1.215052265994018, + 'grad_norm_pre_clip_avg': 0.1222330741584301, + 'learning_rate': 8.357714170977979e-07, + 'epoch': 11.31} +04/20 [03:41:26] INFO | >> train_qwenlatent.py:487 + Step 44820 | grad_norm_pre_clip=0.1046 | + grad_norm_pre_clip_avg=0.1421 | Metrics: + {'align_loss': 0.025965865701436996, + 'recon_loss': 0.11084145307540894, + 'predict_loss': 0.002405038569122553, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10460985451936722, + 'data_time': 0.0012954730191268027, + 'model_time': 1.2323670930054504, + 'grad_norm_pre_clip_avg': 0.14206776916980743, + 'learning_rate': 8.326841592864908e-07, + 'epoch': 11.31} +04/20 [03:41:39] INFO | >> train_qwenlatent.py:487 + Step 44830 | grad_norm_pre_clip=0.1921 | + grad_norm_pre_clip_avg=0.1383 | Metrics: + {'align_loss': 0.025450490415096283, + 'recon_loss': 0.11416760832071304, + 'predict_loss': 0.006517094559967518, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19211527705192566, + 'data_time': 0.0006628950068261474, + 'model_time': 1.267780844005756, + 'grad_norm_pre_clip_avg': 0.1383045919239521, + 'learning_rate': 8.296025940755332e-07, + 'epoch': 11.31} +04/20 [03:41:51] INFO | >> train_qwenlatent.py:487 + Step 44840 | grad_norm_pre_clip=0.1293 | + grad_norm_pre_clip_avg=0.1400 | Metrics: + {'align_loss': 0.0249688308686018, + 'recon_loss': 0.18565070629119873, + 'predict_loss': 0.00806216336786747, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12931078672409058, + 'data_time': 0.0007100760121829808, + 'model_time': 1.23591238699737, + 'grad_norm_pre_clip_avg': 0.13999579027295111, + 'learning_rate': 8.26526722966847e-07, 'epoch': + 11.31} +04/20 [03:42:04] INFO | >> train_qwenlatent.py:487 + Step 44850 | grad_norm_pre_clip=0.1789 | + grad_norm_pre_clip_avg=0.1474 | Metrics: + {'align_loss': 0.026357248425483704, + 'recon_loss': 0.08968768268823624, + 'predict_loss': 0.0023114015348255634, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17892995476722717, + 'mae_score': 0.006530300776163737, 'data_time': + 0.0009193650039378554, 'model_time': + 1.227837129990803, 'grad_norm_pre_clip_avg': + 0.147391664236784, 'learning_rate': + 8.234565474595713e-07, 'epoch': 11.32} +04/20 [03:42:17] INFO | >> train_qwenlatent.py:487 + Step 44860 | grad_norm_pre_clip=0.1012 | + grad_norm_pre_clip_avg=0.1484 | Metrics: + {'align_loss': 0.02368112839758396, + 'recon_loss': 0.11685814708471298, + 'predict_loss': 0.003240163903683424, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10122450441122055, + 'data_time': 0.001148799987277016, + 'model_time': 1.3078519119881094, + 'grad_norm_pre_clip_avg': 0.148377238959074, + 'learning_rate': 8.203920690500755e-07, + 'epoch': 11.32} +04/20 [03:42:30] INFO | >> train_qwenlatent.py:487 + Step 44870 | grad_norm_pre_clip=0.1818 | + grad_norm_pre_clip_avg=0.1380 | Metrics: + {'align_loss': 0.025164205580949783, + 'recon_loss': 0.16366362571716309, + 'predict_loss': 0.005986145231872797, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18178465962409973, + 'data_time': 0.0009075819980353117, + 'model_time': 1.2513434730062727, + 'grad_norm_pre_clip_avg': 0.1380077950656414, + 'learning_rate': 8.17333289231949e-07, 'epoch': + 11.32} +04/20 [03:42:42] INFO | >> train_qwenlatent.py:487 + Step 44880 | grad_norm_pre_clip=0.1581 | + grad_norm_pre_clip_avg=0.1460 | Metrics: + {'align_loss': 0.025093140080571175, + 'recon_loss': 0.12968207895755768, + 'predict_loss': 0.004141213372349739, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15810829401016235, + 'data_time': 0.000637941004242748, + 'model_time': 1.2461145700071938, + 'grad_norm_pre_clip_avg': 0.14602950885891913, + 'learning_rate': 8.142802094960015e-07, + 'epoch': 11.32} +04/20 [03:42:55] INFO | >> train_qwenlatent.py:487 + Step 44890 | grad_norm_pre_clip=0.0998 | + grad_norm_pre_clip_avg=0.1534 | Metrics: + {'align_loss': 0.02583693340420723, + 'recon_loss': 0.16018365323543549, + 'predict_loss': 0.006292917300015688, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.0997949093580246, + 'data_time': 0.0006702909886371344, + 'model_time': 1.2268649849866051, + 'grad_norm_pre_clip_avg': 0.15343122631311418, + 'learning_rate': 8.112328313302688e-07, + 'epoch': 11.33} +04/20 [03:43:08] INFO | >> train_qwenlatent.py:487 + Step 44900 | grad_norm_pre_clip=0.1352 | + grad_norm_pre_clip_avg=0.1506 | Metrics: + {'align_loss': 0.023955516517162323, + 'recon_loss': 0.14803393185138702, + 'predict_loss': 0.00617336668074131, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13522352278232574, + 'mae_score': 0.005260232547381977, 'data_time': + 0.0009637660114094615, 'model_time': + 1.2484278150077444, 'grad_norm_pre_clip_avg': + 0.15055423676967622, 'learning_rate': + 8.081911562200055e-07, 'epoch': 11.33} +04/20 [03:43:21] INFO | >> train_qwenlatent.py:487 + Step 44910 | grad_norm_pre_clip=0.0890 | + grad_norm_pre_clip_avg=0.1362 | Metrics: + {'align_loss': 0.025549326092004776, + 'recon_loss': 0.14717957377433777, + 'predict_loss': 0.006849377881735563, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.08902798593044281, + 'data_time': 0.0009819800034165382, + 'model_time': 1.2431578050018288, + 'grad_norm_pre_clip_avg': 0.13617039024829863, + 'learning_rate': 8.051551856476877e-07, + 'epoch': 11.33} +04/20 [03:43:33] INFO | >> train_qwenlatent.py:487 + Step 44920 | grad_norm_pre_clip=0.1239 | + grad_norm_pre_clip_avg=0.1437 | Metrics: + {'align_loss': 0.025971611961722374, + 'recon_loss': 0.20510821044445038, + 'predict_loss': 0.0077358633279800415, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12394559383392334, + 'data_time': 0.0006832809885963798, + 'model_time': 1.1981766810058616, + 'grad_norm_pre_clip_avg': 0.14373034089803696, + 'learning_rate': 8.021249210930093e-07, + 'epoch': 11.33} +04/20 [03:43:46] INFO | >> train_qwenlatent.py:487 + Step 44930 | grad_norm_pre_clip=0.1144 | + grad_norm_pre_clip_avg=0.1371 | Metrics: + {'align_loss': 0.02629900723695755, + 'recon_loss': 0.15686039626598358, + 'predict_loss': 0.0036764454562216997, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11435575038194656, + 'data_time': 0.0007496520120184869, + 'model_time': 1.5842034429952037, + 'grad_norm_pre_clip_avg': 0.13714466094970704, + 'learning_rate': 7.991003640328855e-07, + 'epoch': 11.34} +04/20 [03:43:59] INFO | >> train_qwenlatent.py:487 + Step 44940 | grad_norm_pre_clip=0.1101 | + grad_norm_pre_clip_avg=0.1311 | Metrics: + {'align_loss': 0.025416355580091476, + 'recon_loss': 0.1613391488790512, + 'predict_loss': 0.008196894079446793, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11010798066854477, + 'data_time': 0.0006423599843401462, + 'model_time': 1.2379038060025778, + 'grad_norm_pre_clip_avg': 0.13107835054397582, + 'learning_rate': 7.960815159414479e-07, + 'epoch': 11.34} +04/20 [03:44:12] INFO | >> train_qwenlatent.py:487 + Step 44950 | grad_norm_pre_clip=0.0970 | + grad_norm_pre_clip_avg=0.1280 | Metrics: + {'align_loss': 0.02713441476225853, + 'recon_loss': 0.11649200320243835, + 'predict_loss': 0.0027268233243376017, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09697422385215759, + 'mae_score': 0.005987506299405485, 'data_time': + 0.0007218389946501702, 'model_time': + 1.2396955260192044, 'grad_norm_pre_clip_avg': + 0.1280020847916603, 'learning_rate': + 7.930683782900454e-07, 'epoch': 11.34} +04/20 [03:44:25] INFO | >> train_qwenlatent.py:487 + Step 44960 | grad_norm_pre_clip=0.1575 | + grad_norm_pre_clip_avg=0.1455 | Metrics: + {'align_loss': 0.025625435635447502, + 'recon_loss': 0.1710071712732315, + 'predict_loss': 0.00936206430196762, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15748323500156403, + 'data_time': 0.00098977200104855, 'model_time': + 1.2165378009958658, 'grad_norm_pre_clip_avg': + 0.1454938381910324, 'learning_rate': + 7.900609525472476e-07, 'epoch': 11.34} +04/20 [03:44:37] INFO | >> train_qwenlatent.py:487 + Step 44970 | grad_norm_pre_clip=0.1930 | + grad_norm_pre_clip_avg=0.1463 | Metrics: + {'align_loss': 0.025853078812360764, + 'recon_loss': 0.15417450666427612, + 'predict_loss': 0.00747371232137084, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19300316274166107, + 'data_time': 0.0009116330184042454, + 'model_time': 1.5509946300007869, + 'grad_norm_pre_clip_avg': 0.14631562307476997, + 'learning_rate': 7.870592401788356e-07, + 'epoch': 11.35} +04/20 [03:44:50] INFO | >> train_qwenlatent.py:487 + Step 44980 | grad_norm_pre_clip=0.1638 | + grad_norm_pre_clip_avg=0.1508 | Metrics: + {'align_loss': 0.025536686182022095, + 'recon_loss': 0.18920329213142395, + 'predict_loss': 0.010496752336621284, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16377781331539154, + 'data_time': 0.00074951202259399, 'model_time': + 1.268828948988812, 'grad_norm_pre_clip_avg': + 0.15080655440688134, 'learning_rate': + 7.840632426478085e-07, 'epoch': 11.35} +04/20 [03:45:02] INFO | >> train_qwenlatent.py:487 + Step 44990 | grad_norm_pre_clip=0.1066 | + grad_norm_pre_clip_avg=0.1382 | Metrics: + {'align_loss': 0.026153944432735443, + 'recon_loss': 0.13979624211788177, + 'predict_loss': 0.006022337824106216, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10659457743167877, + 'data_time': 0.0009872799855656922, + 'model_time': 1.208850099996198, + 'grad_norm_pre_clip_avg': 0.13822935819625853, + 'learning_rate': 7.810729614143767e-07, + 'epoch': 11.35} +04/20 [03:45:15] INFO | >> train_qwenlatent.py:487 + Step 45000 | grad_norm_pre_clip=0.1315 | + grad_norm_pre_clip_avg=0.1345 | Metrics: + {'align_loss': 0.02598714828491211, + 'recon_loss': 0.1657596081495285, + 'predict_loss': 0.006085705012083054, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1315460056066513, + 'mae_score': 0.006340012679228912, 'data_time': + 0.0007111600134521723, 'model_time': + 1.201297802006593, 'grad_norm_pre_clip_avg': + 0.13451105281710624, 'learning_rate': + 7.780883979359699e-07, 'epoch': 11.36} +✅ Checkpoint saved at ./runs/0418_QwenLatent_13tasks_actionstate_30k/checkpoints/steps_45000 +04/20 [03:45:38] INFO | >> train_qwenlatent.py:487 + Step 45010 | grad_norm_pre_clip=0.1462 | + grad_norm_pre_clip_avg=0.1290 | Metrics: + {'align_loss': 0.024917656555771828, + 'recon_loss': 0.1487005054950714, + 'predict_loss': 0.008235062472522259, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14621104300022125, + 'data_time': 0.0006700560043100268, + 'model_time': 1.2879980239958968, + 'grad_norm_pre_clip_avg': 0.1290234312415123, + 'learning_rate': 7.75109553667226e-07, 'epoch': + 11.36} +04/20 [03:45:51] INFO | >> train_qwenlatent.py:487 + Step 45020 | grad_norm_pre_clip=0.1366 | + grad_norm_pre_clip_avg=0.1531 | Metrics: + {'align_loss': 0.02659003436565399, + 'recon_loss': 0.18075792491436005, + 'predict_loss': 0.004843867849558592, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13661713898181915, + 'data_time': 0.0010902979993261397, + 'model_time': 1.2892288439907134, + 'grad_norm_pre_clip_avg': 0.1531236693263054, + 'learning_rate': 7.721364300599981e-07, + 'epoch': 11.36} +04/20 [03:46:04] INFO | >> train_qwenlatent.py:487 + Step 45030 | grad_norm_pre_clip=0.1440 | + grad_norm_pre_clip_avg=0.1454 | Metrics: + {'align_loss': 0.02585487626492977, + 'recon_loss': 0.1775790899991989, + 'predict_loss': 0.00946109276264906, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14402268826961517, + 'data_time': 0.0007279120036400855, + 'model_time': 1.252640007995069, + 'grad_norm_pre_clip_avg': 0.14544716477394104, + 'learning_rate': 7.69169028563352e-07, 'epoch': + 11.36} +04/20 [03:46:17] INFO | >> train_qwenlatent.py:487 + Step 45040 | grad_norm_pre_clip=0.1478 | + grad_norm_pre_clip_avg=0.1436 | Metrics: + {'align_loss': 0.02558233216404915, + 'recon_loss': 0.12404803186655045, + 'predict_loss': 0.006108843721449375, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14778341352939606, + 'data_time': 0.0009902769816108048, + 'model_time': 1.248514150007395, + 'grad_norm_pre_clip_avg': 0.14363565668463707, + 'learning_rate': 7.662073506235585e-07, + 'epoch': 11.37} +04/20 [03:46:30] INFO | >> train_qwenlatent.py:487 + Step 45050 | grad_norm_pre_clip=0.1182 | + grad_norm_pre_clip_avg=0.1382 | Metrics: + {'align_loss': 0.02362322434782982, + 'recon_loss': 0.19667299091815948, + 'predict_loss': 0.008366735652089119, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11820926517248154, + 'mae_score': 0.004777022095413895, 'data_time': + 0.0008265750075224787, 'model_time': + 1.3203021330118645, 'grad_norm_pre_clip_avg': + 0.13818327262997626, 'learning_rate': + 7.632513976841097e-07, 'epoch': 11.37} +04/20 [03:46:43] INFO | >> train_qwenlatent.py:487 + Step 45060 | grad_norm_pre_clip=0.1066 | + grad_norm_pre_clip_avg=0.1490 | Metrics: + {'align_loss': 0.02562280185520649, + 'recon_loss': 0.1321888118982315, + 'predict_loss': 0.0040158168412745, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10663535445928574, + 'data_time': 0.0006925269844941795, + 'model_time': 1.2896055759920273, + 'grad_norm_pre_clip_avg': 0.1489902250468731, + 'learning_rate': 7.603011711856973e-07, + 'epoch': 11.37} +04/20 [03:46:57] INFO | >> train_qwenlatent.py:487 + Step 45070 | grad_norm_pre_clip=0.0917 | + grad_norm_pre_clip_avg=0.1287 | Metrics: + {'align_loss': 0.025234714150428772, + 'recon_loss': 0.17538119852542877, + 'predict_loss': 0.008122175000607967, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09171663224697113, + 'data_time': 0.0010312210069969296, + 'model_time': 1.2747942180139944, + 'grad_norm_pre_clip_avg': 0.12872276306152344, + 'learning_rate': 7.573566725662266e-07, + 'epoch': 11.37} +04/20 [03:47:10] INFO | >> train_qwenlatent.py:487 + Step 45080 | grad_norm_pre_clip=0.0865 | + grad_norm_pre_clip_avg=0.1367 | Metrics: + {'align_loss': 0.025464583188295364, + 'recon_loss': 0.12972787022590637, + 'predict_loss': 0.00334360683336854, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.08647767454385757, + 'data_time': 0.0009192700090352446, + 'model_time': 1.3816002680105157, + 'grad_norm_pre_clip_avg': 0.13670210689306259, + 'learning_rate': 7.544179032608119e-07, + 'epoch': 11.38} +04/20 [03:47:23] INFO | >> train_qwenlatent.py:487 + Step 45090 | grad_norm_pre_clip=0.1271 | + grad_norm_pre_clip_avg=0.1500 | Metrics: + {'align_loss': 0.02579306811094284, + 'recon_loss': 0.1876106709241867, + 'predict_loss': 0.007731510791927576, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1270884871482849, + 'data_time': 0.0007173760095611215, + 'model_time': 1.289249394001672, + 'grad_norm_pre_clip_avg': 0.15003743395209312, + 'learning_rate': 7.51484864701771e-07, 'epoch': + 11.38} +04/20 [03:47:37] INFO | >> train_qwenlatent.py:487 + Step 45100 | grad_norm_pre_clip=0.1237 | + grad_norm_pre_clip_avg=0.1474 | Metrics: + {'align_loss': 0.026484139263629913, + 'recon_loss': 0.2077988237142563, + 'predict_loss': 0.007487674243748188, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1237311065196991, + 'mae_score': 0.005929860553225956, 'data_time': + 0.0008029440068639815, 'model_time': + 1.276260517013725, 'grad_norm_pre_clip_avg': + 0.14739831387996674, 'learning_rate': + 7.485575583186326e-07, 'epoch': 11.38} +04/20 [03:47:50] INFO | >> train_qwenlatent.py:487 + Step 45110 | grad_norm_pre_clip=0.1364 | + grad_norm_pre_clip_avg=0.1861 | Metrics: + {'align_loss': 0.026028618216514587, + 'recon_loss': 0.1501065343618393, + 'predict_loss': 0.005747441668063402, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13635659217834473, + 'data_time': 0.000984763988526538, + 'model_time': 1.6795433719817083, + 'grad_norm_pre_clip_avg': 0.18605701625347137, + 'learning_rate': 7.456359855381304e-07, + 'epoch': 11.38} +04/20 [03:48:03] INFO | >> train_qwenlatent.py:487 + Step 45120 | grad_norm_pre_clip=0.2082 | + grad_norm_pre_clip_avg=0.1598 | Metrics: + {'align_loss': 0.026231225579977036, + 'recon_loss': 0.1983024775981903, + 'predict_loss': 0.00786660611629486, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20823943614959717, + 'data_time': 0.0006554530118592083, + 'model_time': 1.2176863679778762, + 'grad_norm_pre_clip_avg': 0.15976248160004616, + 'learning_rate': 7.427201477842042e-07, + 'epoch': 11.39} +04/20 [03:48:16] INFO | >> train_qwenlatent.py:487 + Step 45130 | grad_norm_pre_clip=0.1225 | + grad_norm_pre_clip_avg=0.1382 | Metrics: + {'align_loss': 0.02523363009095192, + 'recon_loss': 0.13182976841926575, + 'predict_loss': 0.004643644671887159, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1225399374961853, + 'data_time': 0.0007280459976755083, + 'model_time': 1.1834801569930278, + 'grad_norm_pre_clip_avg': 0.13822128847241402, + 'learning_rate': 7.398100464779965e-07, + 'epoch': 11.39} +04/20 [03:48:28] INFO | >> train_qwenlatent.py:487 + Step 45140 | grad_norm_pre_clip=0.1354 | + grad_norm_pre_clip_avg=0.1416 | Metrics: + {'align_loss': 0.02686111256480217, + 'recon_loss': 0.1806599646806717, + 'predict_loss': 0.00940039660781622, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.135384202003479, + 'data_time': 0.0006432369991671294, + 'model_time': 1.2896761709998827, + 'grad_norm_pre_clip_avg': 0.141627237200737, + 'learning_rate': 7.369056830378543e-07, + 'epoch': 11.39} +04/20 [03:48:41] INFO | >> train_qwenlatent.py:487 + Step 45150 | grad_norm_pre_clip=0.0973 | + grad_norm_pre_clip_avg=0.1288 | Metrics: + {'align_loss': 0.026295986026525497, + 'recon_loss': 0.18029624223709106, + 'predict_loss': 0.005111281294375658, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.0973196029663086, + 'mae_score': 0.005208870312115093, 'data_time': + 0.001046908990247175, 'model_time': + 1.3198005380108953, 'grad_norm_pre_clip_avg': + 0.12884696647524835, 'learning_rate': + 7.34007058879333e-07, 'epoch': 11.39} +04/20 [03:48:54] INFO | >> train_qwenlatent.py:487 + Step 45160 | grad_norm_pre_clip=0.1356 | + grad_norm_pre_clip_avg=0.1271 | Metrics: + {'align_loss': 0.026048501953482628, + 'recon_loss': 0.229679673910141, + 'predict_loss': 0.010365772061049938, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1356266736984253, + 'data_time': 0.0008309879922308028, + 'model_time': 1.220377001009183, + 'grad_norm_pre_clip_avg': 0.12712172716856002, + 'learning_rate': 7.31114175415183e-07, 'epoch': + 11.4} +04/20 [03:49:06] INFO | >> train_qwenlatent.py:487 + Step 45170 | grad_norm_pre_clip=0.1340 | + grad_norm_pre_clip_avg=0.1364 | Metrics: + {'align_loss': 0.025913342833518982, + 'recon_loss': 0.18077190220355988, + 'predict_loss': 0.006800965406000614, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13396064937114716, + 'data_time': 0.0007474070007447153, + 'model_time': 1.2291073460073676, + 'grad_norm_pre_clip_avg': 0.13641400933265685, + 'learning_rate': 7.282270340553639e-07, + 'epoch': 11.4} +04/20 [03:49:19] INFO | >> train_qwenlatent.py:487 + Step 45180 | grad_norm_pre_clip=0.1398 | + grad_norm_pre_clip_avg=0.1463 | Metrics: + {'align_loss': 0.025190604850649834, + 'recon_loss': 0.11587703227996826, + 'predict_loss': 0.005850652698427439, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13978791236877441, + 'data_time': 0.0011906450090464205, + 'model_time': 1.2366986949928105, + 'grad_norm_pre_clip_avg': 0.14628111347556114, + 'learning_rate': 7.253456362070294e-07, + 'epoch': 11.4} +04/20 [03:49:31] INFO | >> train_qwenlatent.py:487 + Step 45190 | grad_norm_pre_clip=0.1571 | + grad_norm_pre_clip_avg=0.1271 | Metrics: + {'align_loss': 0.024516534060239792, + 'recon_loss': 0.18302489817142487, + 'predict_loss': 0.013781435787677765, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15713368356227875, + 'data_time': 0.0010034010047093034, + 'model_time': 1.2397715259867255, + 'grad_norm_pre_clip_avg': 0.12713607922196388, + 'learning_rate': 7.224699832745407e-07, + 'epoch': 11.4} +04/20 [03:49:45] INFO | >> train_qwenlatent.py:487 + Step 45200 | grad_norm_pre_clip=0.1403 | + grad_norm_pre_clip_avg=0.1555 | Metrics: + {'align_loss': 0.025854099541902542, + 'recon_loss': 0.19501031935214996, + 'predict_loss': 0.008414197713136673, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.140274778008461, + 'mae_score': 0.006455826974129892, 'data_time': + 0.0007765539921820164, 'model_time': + 1.254407735017594, 'grad_norm_pre_clip_avg': + 0.15546903759241104, 'learning_rate': + 7.19600076659456e-07, 'epoch': 11.41} +04/20 [03:49:57] INFO | >> train_qwenlatent.py:487 + Step 45210 | grad_norm_pre_clip=0.1924 | + grad_norm_pre_clip_avg=0.1413 | Metrics: + {'align_loss': 0.02584252879023552, + 'recon_loss': 0.13212327659130096, + 'predict_loss': 0.008395046927034855, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1924305558204651, + 'data_time': 0.001161194988526404, + 'model_time': 1.2189372870197985, + 'grad_norm_pre_clip_avg': 0.14126759469509126, + 'learning_rate': 7.167359177605317e-07, + 'epoch': 11.41} +04/20 [03:50:10] INFO | >> train_qwenlatent.py:487 + Step 45220 | grad_norm_pre_clip=0.1492 | + grad_norm_pre_clip_avg=0.1523 | Metrics: + {'align_loss': 0.024845803156495094, + 'recon_loss': 0.1375027298927307, + 'predict_loss': 0.004217999521642923, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14919936656951904, + 'data_time': 0.0007219920225907117, + 'model_time': 1.2374039599962998, + 'grad_norm_pre_clip_avg': 0.15232090502977372, + 'learning_rate': 7.138775079737246e-07, + 'epoch': 11.41} +04/20 [03:50:23] INFO | >> train_qwenlatent.py:487 + Step 45230 | grad_norm_pre_clip=0.1240 | + grad_norm_pre_clip_avg=0.1490 | Metrics: + {'align_loss': 0.026158735156059265, + 'recon_loss': 0.14130578935146332, + 'predict_loss': 0.0044334144331514835, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12403528392314911, + 'data_time': 0.0009476999985054135, + 'model_time': 1.2195091269968543, + 'grad_norm_pre_clip_avg': 0.14900167286396027, + 'learning_rate': 7.110248486921888e-07, + 'epoch': 11.41} +04/20 [03:50:35] INFO | >> train_qwenlatent.py:487 + Step 45240 | grad_norm_pre_clip=0.1258 | + grad_norm_pre_clip_avg=0.1614 | Metrics: + {'align_loss': 0.025134358555078506, + 'recon_loss': 0.15114550292491913, + 'predict_loss': 0.0049637677147984505, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12578999996185303, + 'data_time': 0.0010130050068255514, + 'model_time': 1.231556657992769, + 'grad_norm_pre_clip_avg': 0.16135372295975686, + 'learning_rate': 7.081779413062733e-07, + 'epoch': 11.42} +04/20 [03:50:48] INFO | >> train_qwenlatent.py:487 + Step 45250 | grad_norm_pre_clip=0.1225 | + grad_norm_pre_clip_avg=0.1357 | Metrics: + {'align_loss': 0.025677205994725227, + 'recon_loss': 0.08531639724969864, + 'predict_loss': 0.005455370526760817, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1225227415561676, + 'mae_score': 0.005972200685793215, 'data_time': + 0.001061036018654704, 'model_time': + 1.2416653289983515, 'grad_norm_pre_clip_avg': + 0.13572001755237578, 'learning_rate': + 7.05336787203531e-07, 'epoch': 11.42} +04/20 [03:51:01] INFO | >> train_qwenlatent.py:487 + Step 45260 | grad_norm_pre_clip=0.1097 | + grad_norm_pre_clip_avg=0.1364 | Metrics: + {'align_loss': 0.025498591363430023, + 'recon_loss': 0.1934453845024109, + 'predict_loss': 0.0069109476171433926, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1096557155251503, + 'data_time': 0.0006517169822473079, + 'model_time': 1.582913906982867, + 'grad_norm_pre_clip_avg': 0.13637281209230423, + 'learning_rate': 7.025013877687042e-07, + 'epoch': 11.42} +04/20 [03:51:14] INFO | >> train_qwenlatent.py:487 + Step 45270 | grad_norm_pre_clip=0.1203 | + grad_norm_pre_clip_avg=0.1191 | Metrics: + {'align_loss': 0.025999201461672783, + 'recon_loss': 0.16913403570652008, + 'predict_loss': 0.006294948980212212, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1203223243355751, + 'data_time': 0.0007671639905311167, + 'model_time': 1.5373401010001544, + 'grad_norm_pre_clip_avg': 0.11909591257572175, + 'learning_rate': 6.996717443837325e-07, + 'epoch': 11.42} +04/20 [03:51:26] INFO | >> train_qwenlatent.py:487 + Step 45280 | grad_norm_pre_clip=0.1533 | + grad_norm_pre_clip_avg=0.1396 | Metrics: + {'align_loss': 0.02494499832391739, + 'recon_loss': 0.1027945727109909, + 'predict_loss': 0.003304198617115617, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15326207876205444, + 'data_time': 0.0009312049951404333, + 'model_time': 1.2143769880058244, + 'grad_norm_pre_clip_avg': 0.13961185589432717, + 'learning_rate': 6.96847858427745e-07, 'epoch': + 11.43} +04/20 [03:51:39] INFO | >> train_qwenlatent.py:487 + Step 45290 | grad_norm_pre_clip=0.0832 | + grad_norm_pre_clip_avg=0.1253 | Metrics: + {'align_loss': 0.024543575942516327, + 'recon_loss': 0.12994474172592163, + 'predict_loss': 0.004312735516577959, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.08319026976823807, + 'data_time': 0.0007328179781325161, + 'model_time': 1.2433993669983465, + 'grad_norm_pre_clip_avg': 0.12525394558906555, + 'learning_rate': 6.940297312770756e-07, + 'epoch': 11.43} +04/20 [03:51:52] INFO | >> train_qwenlatent.py:487 + Step 45300 | grad_norm_pre_clip=0.1866 | + grad_norm_pre_clip_avg=0.1454 | Metrics: + {'align_loss': 0.025650303810834885, + 'recon_loss': 0.15460854768753052, + 'predict_loss': 0.004363317973911762, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18664643168449402, + 'mae_score': 0.005163530401281409, 'data_time': + 0.0006485919875558466, 'model_time': + 1.2123278600047342, 'grad_norm_pre_clip_avg': + 0.14543551877140998, 'learning_rate': + 6.912173643052428e-07, 'epoch': 11.43} +04/20 [03:52:04] INFO | >> train_qwenlatent.py:487 + Step 45310 | grad_norm_pre_clip=0.1053 | + grad_norm_pre_clip_avg=0.1161 | Metrics: + {'align_loss': 0.02396375872194767, + 'recon_loss': 0.1304420530796051, + 'predict_loss': 0.007853841409087181, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10527585446834564, + 'data_time': 0.0008575510000810027, + 'model_time': 1.2234505949891172, + 'grad_norm_pre_clip_avg': 0.11611277759075164, + 'learning_rate': 6.884107588829588e-07, + 'epoch': 11.43} +04/20 [03:52:17] INFO | >> train_qwenlatent.py:487 + Step 45320 | grad_norm_pre_clip=0.1958 | + grad_norm_pre_clip_avg=0.1463 | Metrics: + {'align_loss': 0.026327921077609062, + 'recon_loss': 0.20493097603321075, + 'predict_loss': 0.006783903576433659, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19584858417510986, + 'data_time': 0.0007495800091419369, + 'model_time': 1.2014284289907664, + 'grad_norm_pre_clip_avg': 0.1462542437016964, + 'learning_rate': 6.856099163781314e-07, + 'epoch': 11.44} +04/20 [03:52:29] INFO | >> train_qwenlatent.py:487 + Step 45330 | grad_norm_pre_clip=0.1101 | + grad_norm_pre_clip_avg=0.1303 | Metrics: + {'align_loss': 0.024961913004517555, + 'recon_loss': 0.14181257784366608, + 'predict_loss': 0.004703632555902004, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11009087413549423, + 'data_time': 0.0009398499969393015, + 'model_time': 1.2725770529941656, + 'grad_norm_pre_clip_avg': 0.13027538433671, + 'learning_rate': 6.828148381558537e-07, + 'epoch': 11.44} +04/20 [03:52:42] INFO | >> train_qwenlatent.py:487 + Step 45340 | grad_norm_pre_clip=0.1279 | + grad_norm_pre_clip_avg=0.1171 | Metrics: + {'align_loss': 0.0243698637932539, + 'recon_loss': 0.10081461817026138, + 'predict_loss': 0.003985517658293247, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1278918832540512, + 'data_time': 0.0007220040133688599, + 'model_time': 1.2553932089940645, + 'grad_norm_pre_clip_avg': 0.11713727563619614, + 'learning_rate': 6.800255255784188e-07, + 'epoch': 11.44} +04/20 [03:52:55] INFO | >> train_qwenlatent.py:487 + Step 45350 | grad_norm_pre_clip=0.1284 | + grad_norm_pre_clip_avg=0.1541 | Metrics: + {'align_loss': 0.025010399520397186, + 'recon_loss': 0.13738606870174408, + 'predict_loss': 0.0062127322889864445, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12843385338783264, + 'mae_score': 0.005675514324291332, 'data_time': + 0.0008875990170054138, 'model_time': + 1.2428840929933358, 'grad_norm_pre_clip_avg': + 0.15413310900330543, 'learning_rate': + 6.772419800053003e-07, 'epoch': 11.44} +04/20 [03:53:08] INFO | >> train_qwenlatent.py:487 + Step 45360 | grad_norm_pre_clip=0.1252 | + grad_norm_pre_clip_avg=0.1487 | Metrics: + {'align_loss': 0.02316868305206299, + 'recon_loss': 0.11793738603591919, + 'predict_loss': 0.00459292670711875, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12523679435253143, + 'data_time': 0.0006583270151168108, + 'model_time': 1.2593663200095762, + 'grad_norm_pre_clip_avg': 0.14873883202672006, + 'learning_rate': 6.744642027931628e-07, + 'epoch': 11.45} +04/20 [03:53:21] INFO | >> train_qwenlatent.py:487 + Step 45370 | grad_norm_pre_clip=0.0971 | + grad_norm_pre_clip_avg=0.1203 | Metrics: + {'align_loss': 0.02556958608329296, + 'recon_loss': 0.14144207537174225, + 'predict_loss': 0.003743487875908613, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09707077592611313, + 'data_time': 0.0008883400005288422, + 'model_time': 1.224802743003238, + 'grad_norm_pre_clip_avg': 0.12034728601574898, + 'learning_rate': 6.716921952958611e-07, + 'epoch': 11.45} +04/20 [03:53:33] INFO | >> train_qwenlatent.py:487 + Step 45380 | grad_norm_pre_clip=0.1040 | + grad_norm_pre_clip_avg=0.1387 | Metrics: + {'align_loss': 0.02507033571600914, + 'recon_loss': 0.1617879420518875, + 'predict_loss': 0.005405258387327194, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10397210717201233, + 'data_time': 0.0008235620043706149, + 'model_time': 1.2170527490088716, + 'grad_norm_pre_clip_avg': 0.13871546536684037, + 'learning_rate': 6.689259588644418e-07, + 'epoch': 11.45} +04/20 [03:53:45] INFO | >> train_qwenlatent.py:487 + Step 45390 | grad_norm_pre_clip=0.1028 | + grad_norm_pre_clip_avg=0.1217 | Metrics: + {'align_loss': 0.025151707231998444, + 'recon_loss': 0.1405044049024582, + 'predict_loss': 0.0038338997401297092, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10279761254787445, + 'data_time': 0.0008187409839592874, + 'model_time': 1.2470900070038624, + 'grad_norm_pre_clip_avg': 0.12165677025914193, + 'learning_rate': 6.661654948471325e-07, + 'epoch': 11.45} +04/20 [03:53:59] INFO | >> train_qwenlatent.py:487 + Step 45400 | grad_norm_pre_clip=0.1301 | + grad_norm_pre_clip_avg=0.1406 | Metrics: + {'align_loss': 0.02582995966076851, + 'recon_loss': 0.22403427958488464, + 'predict_loss': 0.008918805047869682, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13009974360466003, + 'mae_score': 0.0054249681867994705, + 'data_time': 0.0007028319814708084, + 'model_time': 1.2292754130030517, + 'grad_norm_pre_clip_avg': 0.1405526138842106, + 'learning_rate': 6.634108045893496e-07, + 'epoch': 11.46} +04/20 [03:54:11] INFO | >> train_qwenlatent.py:487 + Step 45410 | grad_norm_pre_clip=0.1070 | + grad_norm_pre_clip_avg=0.1462 | Metrics: + {'align_loss': 0.026265883818268776, + 'recon_loss': 0.184756338596344, + 'predict_loss': 0.008009127341210842, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10699380934238434, + 'data_time': 0.0014658229774795473, + 'model_time': 1.2660897010064218, + 'grad_norm_pre_clip_avg': 0.14624861478805543, + 'learning_rate': 6.606618894336976e-07, + 'epoch': 11.46} +04/20 [03:54:24] INFO | >> train_qwenlatent.py:487 + Step 45420 | grad_norm_pre_clip=0.1977 | + grad_norm_pre_clip_avg=0.1314 | Metrics: + {'align_loss': 0.023303251713514328, + 'recon_loss': 0.1405942589044571, + 'predict_loss': 0.003944777883589268, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19766388833522797, + 'data_time': 0.000734564004233107, + 'model_time': 1.2753893499902915, + 'grad_norm_pre_clip_avg': 0.13136921674013138, + 'learning_rate': 6.579187507199613e-07, + 'epoch': 11.46} +04/20 [03:54:37] INFO | >> train_qwenlatent.py:487 + Step 45430 | grad_norm_pre_clip=0.1019 | + grad_norm_pre_clip_avg=0.1262 | Metrics: + {'align_loss': 0.025019818916916847, + 'recon_loss': 0.10680448263883591, + 'predict_loss': 0.004431701265275478, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10185226798057556, + 'data_time': 0.001002996985334903, + 'model_time': 1.2771716819843277, + 'grad_norm_pre_clip_avg': 0.1261936016380787, + 'learning_rate': 6.551813897851137e-07, + 'epoch': 11.46} +04/20 [03:54:49] INFO | >> train_qwenlatent.py:487 + Step 45440 | grad_norm_pre_clip=0.1739 | + grad_norm_pre_clip_avg=0.1419 | Metrics: + {'align_loss': 0.026277460157871246, + 'recon_loss': 0.16858133673667908, + 'predict_loss': 0.007909592241048813, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17391479015350342, + 'data_time': 0.0009946250065695494, + 'model_time': 1.2621875990007538, + 'grad_norm_pre_clip_avg': 0.1418937064707279, + 'learning_rate': 6.524498079633162e-07, + 'epoch': 11.47} +04/20 [03:55:02] INFO | >> train_qwenlatent.py:487 + Step 45450 | grad_norm_pre_clip=0.1127 | + grad_norm_pre_clip_avg=0.1271 | Metrics: + {'align_loss': 0.026899509131908417, + 'recon_loss': 0.12614627182483673, + 'predict_loss': 0.004214812535792589, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11272625625133514, + 'mae_score': 0.0056241765752568975, + 'data_time': 0.0009371689811814576, + 'model_time': 1.2003146189963445, + 'grad_norm_pre_clip_avg': 0.12708135470747947, + 'learning_rate': 6.497240065859036e-07, + 'epoch': 11.47} +04/20 [03:55:15] INFO | >> train_qwenlatent.py:487 + Step 45460 | grad_norm_pre_clip=0.1076 | + grad_norm_pre_clip_avg=0.1324 | Metrics: + {'align_loss': 0.02490607090294361, + 'recon_loss': 0.17175239324569702, + 'predict_loss': 0.007197132799774408, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10761331766843796, + 'data_time': 0.0006576449959538877, + 'model_time': 1.2062384449818637, + 'grad_norm_pre_clip_avg': 0.13244323581457138, + 'learning_rate': 6.470039869813976e-07, + 'epoch': 11.47} +04/20 [03:55:27] INFO | >> train_qwenlatent.py:487 + Step 45470 | grad_norm_pre_clip=0.1231 | + grad_norm_pre_clip_avg=0.1321 | Metrics: + {'align_loss': 0.026426389813423157, + 'recon_loss': 0.12785941362380981, + 'predict_loss': 0.0025555966421961784, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1231379359960556, + 'data_time': 0.0009966630022972822, + 'model_time': 1.213968472002307, + 'grad_norm_pre_clip_avg': 0.1321319691836834, + 'learning_rate': 6.442897504755055e-07, + 'epoch': 11.47} +04/20 [03:55:40] INFO | >> train_qwenlatent.py:487 + Step 45480 | grad_norm_pre_clip=0.0945 | + grad_norm_pre_clip_avg=0.1288 | Metrics: + {'align_loss': 0.025429952889680862, + 'recon_loss': 0.15947337448596954, + 'predict_loss': 0.0059415618889033794, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09446383267641068, + 'data_time': 0.0008712039852980524, + 'model_time': 1.2183788149850443, + 'grad_norm_pre_clip_avg': 0.12875701859593391, + 'learning_rate': 6.415812983911116e-07, + 'epoch': 11.48} +04/20 [03:55:53] INFO | >> train_qwenlatent.py:487 + Step 45490 | grad_norm_pre_clip=0.1734 | + grad_norm_pre_clip_avg=0.1348 | Metrics: + {'align_loss': 0.025899196043610573, + 'recon_loss': 0.15404929220676422, + 'predict_loss': 0.00790639128535986, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17342667281627655, + 'data_time': 0.0008623620087746531, + 'model_time': 1.2432846130104735, + 'grad_norm_pre_clip_avg': 0.13483782187104226, + 'learning_rate': 6.38878632048285e-07, 'epoch': + 11.48} +04/20 [03:56:06] INFO | >> train_qwenlatent.py:487 + Step 45500 | grad_norm_pre_clip=0.1301 | + grad_norm_pre_clip_avg=0.1356 | Metrics: + {'align_loss': 0.02587995119392872, + 'recon_loss': 0.24896356463432312, + 'predict_loss': 0.008481169119477272, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13007956743240356, + 'mae_score': 0.005611261161597999, 'data_time': + 0.0013601669925265014, 'model_time': + 1.189094664005097, 'grad_norm_pre_clip_avg': + 0.1356019452214241, 'learning_rate': + 6.361817527642703e-07, 'epoch': 11.48} +04/20 [03:56:19] INFO | >> train_qwenlatent.py:487 + Step 45510 | grad_norm_pre_clip=0.1206 | + grad_norm_pre_clip_avg=0.1313 | Metrics: + {'align_loss': 0.024550871923565865, + 'recon_loss': 0.13807463645935059, + 'predict_loss': 0.007230183109641075, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12060700356960297, + 'data_time': 0.0006922669999767095, + 'model_time': 1.2441303480009083, + 'grad_norm_pre_clip_avg': 0.1313195198774338, + 'learning_rate': 6.334906618534939e-07, + 'epoch': 11.48} +04/20 [03:56:31] INFO | >> train_qwenlatent.py:487 + Step 45520 | grad_norm_pre_clip=0.2095 | + grad_norm_pre_clip_avg=0.1445 | Metrics: + {'align_loss': 0.02517757937312126, + 'recon_loss': 0.18062682449817657, + 'predict_loss': 0.012324507348239422, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2095441073179245, + 'data_time': 0.0007115059997886419, + 'model_time': 1.208111255982658, + 'grad_norm_pre_clip_avg': 0.14450818821787834, + 'learning_rate': 6.308053606275591e-07, + 'epoch': 11.49} +04/20 [03:56:44] INFO | >> train_qwenlatent.py:487 + Step 45530 | grad_norm_pre_clip=0.1727 | + grad_norm_pre_clip_avg=0.1487 | Metrics: + {'align_loss': 0.02552489936351776, + 'recon_loss': 0.16804978251457214, + 'predict_loss': 0.006959329359233379, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17269404232501984, + 'data_time': 0.0010724430030677468, + 'model_time': 1.2859302540018689, + 'grad_norm_pre_clip_avg': 0.14873939454555513, + 'learning_rate': 6.281258503952553e-07, + 'epoch': 11.49} +04/20 [03:56:56] INFO | >> train_qwenlatent.py:487 + Step 45540 | grad_norm_pre_clip=0.1439 | + grad_norm_pre_clip_avg=0.1393 | Metrics: + {'align_loss': 0.02558334916830063, + 'recon_loss': 0.20060713589191437, + 'predict_loss': 0.009274577721953392, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14387166500091553, + 'data_time': 0.0006687880086246878, + 'model_time': 1.27349165501073, + 'grad_norm_pre_clip_avg': 0.1393066316843033, + 'learning_rate': 6.254521324625376e-07, + 'epoch': 11.49} +04/20 [03:57:10] INFO | >> train_qwenlatent.py:487 + Step 45550 | grad_norm_pre_clip=0.1414 | + grad_norm_pre_clip_avg=0.1393 | Metrics: + {'align_loss': 0.025112664327025414, + 'recon_loss': 0.1302330046892166, + 'predict_loss': 0.004317982122302055, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14139074087142944, + 'mae_score': 0.005507459726419535, 'data_time': + 0.001198937010485679, 'model_time': + 1.1970509639941156, 'grad_norm_pre_clip_avg': + 0.13930292576551437, 'learning_rate': + 6.227842081325444e-07, 'epoch': 11.49} +04/20 [03:57:22] INFO | >> train_qwenlatent.py:487 + Step 45560 | grad_norm_pre_clip=0.1471 | + grad_norm_pre_clip_avg=0.1360 | Metrics: + {'align_loss': 0.024897459894418716, + 'recon_loss': 0.22182497382164001, + 'predict_loss': 0.008732717484235764, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1470644474029541, + 'data_time': 0.0006261699891183525, + 'model_time': 1.2274853679991793, + 'grad_norm_pre_clip_avg': 0.13595695942640304, + 'learning_rate': 6.201220787055898e-07, + 'epoch': 11.5} +04/20 [03:57:35] INFO | >> train_qwenlatent.py:487 + Step 45570 | grad_norm_pre_clip=0.1227 | + grad_norm_pre_clip_avg=0.1372 | Metrics: + {'align_loss': 0.025410760194063187, + 'recon_loss': 0.17456863820552826, + 'predict_loss': 0.010211865417659283, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12271071970462799, + 'data_time': 0.0013040800113230944, + 'model_time': 1.2534287989838049, + 'grad_norm_pre_clip_avg': 0.13717047050595282, + 'learning_rate': 6.174657454791625e-07, + 'epoch': 11.5} +04/20 [03:57:48] INFO | >> train_qwenlatent.py:487 + Step 45580 | grad_norm_pre_clip=0.1789 | + grad_norm_pre_clip_avg=0.1423 | Metrics: + {'align_loss': 0.0256083682179451, + 'recon_loss': 0.1362563967704773, + 'predict_loss': 0.006773282773792744, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17891119420528412, + 'data_time': 0.0009370380139444023, + 'model_time': 1.2433731009950861, + 'grad_norm_pre_clip_avg': 0.14227463006973268, + 'learning_rate': 6.148152097479312e-07, + 'epoch': 11.5} +04/20 [03:58:00] INFO | >> train_qwenlatent.py:487 + Step 45590 | grad_norm_pre_clip=0.1300 | + grad_norm_pre_clip_avg=0.1304 | Metrics: + {'align_loss': 0.02475605718791485, + 'recon_loss': 0.12205197662115097, + 'predict_loss': 0.005593698006123304, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13004070520401, + 'data_time': 0.0006367349997162819, + 'model_time': 1.2427503709914163, + 'grad_norm_pre_clip_avg': 0.13041517585515977, + 'learning_rate': 6.121704728037309e-07, + 'epoch': 11.5} +04/20 [03:58:13] INFO | >> train_qwenlatent.py:487 + Step 45600 | grad_norm_pre_clip=0.1683 | + grad_norm_pre_clip_avg=0.1323 | Metrics: + {'align_loss': 0.025075890123844147, + 'recon_loss': 0.15494497120380402, + 'predict_loss': 0.010919368825852871, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1683197021484375, + 'mae_score': 0.005409614459888355, 'data_time': + 0.0007137970242183656, 'model_time': + 1.3114569719764404, 'grad_norm_pre_clip_avg': + 0.1322709009051323, 'learning_rate': + 6.095315359355753e-07, 'epoch': 11.51} +04/20 [03:58:26] INFO | >> train_qwenlatent.py:487 + Step 45610 | grad_norm_pre_clip=0.1265 | + grad_norm_pre_clip_avg=0.1316 | Metrics: + {'align_loss': 0.026128310710191727, + 'recon_loss': 0.15730060636997223, + 'predict_loss': 0.004915901459753513, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12650246918201447, + 'data_time': 0.0006775540241505951, + 'model_time': 1.277407456014771, + 'grad_norm_pre_clip_avg': 0.1315842054784298, + 'learning_rate': 6.068984004296498e-07, + 'epoch': 11.51} +04/20 [03:58:39] INFO | >> train_qwenlatent.py:487 + Step 45620 | grad_norm_pre_clip=0.1826 | + grad_norm_pre_clip_avg=0.1443 | Metrics: + {'align_loss': 0.02474747598171234, + 'recon_loss': 0.125118687748909, + 'predict_loss': 0.0038716355338692665, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18262259662151337, + 'data_time': 0.0007149909797590226, + 'model_time': 1.1997393209894653, + 'grad_norm_pre_clip_avg': 0.1442732460796833, + 'learning_rate': 6.042710675693131e-07, + 'epoch': 11.51} +04/20 [03:58:51] INFO | >> train_qwenlatent.py:487 + Step 45630 | grad_norm_pre_clip=0.1221 | + grad_norm_pre_clip_avg=0.1275 | Metrics: + {'align_loss': 0.02554570883512497, + 'recon_loss': 0.1651078462600708, + 'predict_loss': 0.007032147608697414, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12210723757743835, + 'data_time': 0.0010060740169137716, + 'model_time': 1.52442329301266, + 'grad_norm_pre_clip_avg': 0.1274866469204426, + 'learning_rate': 6.016495386350955e-07, + 'epoch': 11.51} +04/20 [03:59:04] INFO | >> train_qwenlatent.py:487 + Step 45640 | grad_norm_pre_clip=0.1223 | + grad_norm_pre_clip_avg=0.1247 | Metrics: + {'align_loss': 0.02566913142800331, + 'recon_loss': 0.12290014326572418, + 'predict_loss': 0.005274019204080105, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1222602128982544, + 'data_time': 0.0007121879898477346, + 'model_time': 1.2494586370012257, + 'grad_norm_pre_clip_avg': 0.12466389387845993, + 'learning_rate': 5.990338149046989e-07, + 'epoch': 11.52} +04/20 [03:59:17] INFO | >> train_qwenlatent.py:487 + Step 45650 | grad_norm_pre_clip=0.1296 | + grad_norm_pre_clip_avg=0.1327 | Metrics: + {'align_loss': 0.025614816695451736, + 'recon_loss': 0.17342375218868256, + 'predict_loss': 0.005412300117313862, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12964175641536713, + 'mae_score': 0.005114847475343996, 'data_time': + 0.001091107988031581, 'model_time': + 1.2230085209885146, 'grad_norm_pre_clip_avg': + 0.13270796239376068, 'learning_rate': + 5.964238976529933e-07, 'epoch': 11.52} +04/20 [03:59:30] INFO | >> train_qwenlatent.py:487 + Step 45660 | grad_norm_pre_clip=0.1322 | + grad_norm_pre_clip_avg=0.1334 | Metrics: + {'align_loss': 0.025345632806420326, + 'recon_loss': 0.18004049360752106, + 'predict_loss': 0.004001520574092865, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13219541311264038, + 'data_time': 0.0015334850177168846, + 'model_time': 1.1929013610060792, + 'grad_norm_pre_clip_avg': 0.13342839628458023, + 'learning_rate': 5.938197881520213e-07, + 'epoch': 11.52} +04/20 [03:59:42] INFO | >> train_qwenlatent.py:487 + Step 45670 | grad_norm_pre_clip=0.1256 | + grad_norm_pre_clip_avg=0.1277 | Metrics: + {'align_loss': 0.02492707222700119, + 'recon_loss': 0.23647768795490265, + 'predict_loss': 0.00812709890305996, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12561967968940735, + 'data_time': 0.0007107389974407852, + 'model_time': 1.2458920330100227, + 'grad_norm_pre_clip_avg': 0.12767206132411957, + 'learning_rate': 5.912214876709962e-07, + 'epoch': 11.52} +04/20 [03:59:55] INFO | >> train_qwenlatent.py:487 + Step 45680 | grad_norm_pre_clip=0.1560 | + grad_norm_pre_clip_avg=0.1659 | Metrics: + {'align_loss': 0.024923129007220268, + 'recon_loss': 0.11184804141521454, + 'predict_loss': 0.004101272206753492, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15601438283920288, + 'data_time': 0.0011176539992447942, + 'model_time': 1.2528087190003134, + 'grad_norm_pre_clip_avg': 0.16591370850801468, + 'learning_rate': 5.886289974762962e-07, + 'epoch': 11.53} +04/20 [04:00:08] INFO | >> train_qwenlatent.py:487 + Step 45690 | grad_norm_pre_clip=0.1497 | + grad_norm_pre_clip_avg=0.1607 | Metrics: + {'align_loss': 0.02529248595237732, + 'recon_loss': 0.16007119417190552, + 'predict_loss': 0.005435810424387455, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14968739449977875, + 'data_time': 0.000759788992581889, + 'model_time': 1.5617063919780776, + 'grad_norm_pre_clip_avg': 0.16072598546743394, + 'learning_rate': 5.860423188314696e-07, + 'epoch': 11.53} +04/20 [04:00:21] INFO | >> train_qwenlatent.py:487 + Step 45700 | grad_norm_pre_clip=0.0965 | + grad_norm_pre_clip_avg=0.1342 | Metrics: + {'align_loss': 0.0250871479511261, + 'recon_loss': 0.12471676617860794, + 'predict_loss': 0.005572017747908831, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09645941853523254, + 'mae_score': 0.006177529343613633, 'data_time': + 0.000889146002009511, 'model_time': + 1.2242753409955185, 'grad_norm_pre_clip_avg': + 0.13423054814338684, 'learning_rate': + 5.83461452997234e-07, 'epoch': 11.53} +04/20 [04:00:34] INFO | >> train_qwenlatent.py:487 + Step 45710 | grad_norm_pre_clip=0.1673 | + grad_norm_pre_clip_avg=0.1250 | Metrics: + {'align_loss': 0.026468701660633087, + 'recon_loss': 0.09989476948976517, + 'predict_loss': 0.002748209750279784, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16728709638118744, + 'data_time': 0.0009676160116214305, + 'model_time': 1.3050294530112296, + 'grad_norm_pre_clip_avg': 0.12503888756036757, + 'learning_rate': 5.808864012314701e-07, + 'epoch': 11.53} +04/20 [04:00:46] INFO | >> train_qwenlatent.py:487 + Step 45720 | grad_norm_pre_clip=0.1480 | + grad_norm_pre_clip_avg=0.1268 | Metrics: + {'align_loss': 0.025663888081908226, + 'recon_loss': 0.14255696535110474, + 'predict_loss': 0.0035804465878754854, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14799299836158752, + 'data_time': 0.0010163700208067894, + 'model_time': 1.2771809620026033, + 'grad_norm_pre_clip_avg': 0.12677255943417548, + 'learning_rate': 5.783171647892265e-07, + 'epoch': 11.54} +04/20 [04:00:59] INFO | >> train_qwenlatent.py:487 + Step 45730 | grad_norm_pre_clip=0.1161 | + grad_norm_pre_clip_avg=0.1290 | Metrics: + {'align_loss': 0.025909001007676125, + 'recon_loss': 0.12123265117406845, + 'predict_loss': 0.00330203864723444, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1161404550075531, + 'data_time': 0.00094149200594984, 'model_time': + 1.2636704969918355, 'grad_norm_pre_clip_avg': + 0.12895549461245537, 'learning_rate': + 5.757537449227191e-07, 'epoch': 11.54} +04/20 [04:01:11] INFO | >> train_qwenlatent.py:487 + Step 45740 | grad_norm_pre_clip=0.1786 | + grad_norm_pre_clip_avg=0.1459 | Metrics: + {'align_loss': 0.025131378322839737, + 'recon_loss': 0.15769226849079132, + 'predict_loss': 0.006115500815212727, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1786464899778366, + 'data_time': 0.0009596350137144327, + 'model_time': 1.2129769739985932, + 'grad_norm_pre_clip_avg': 0.14588322415947913, + 'learning_rate': 5.731961428813277e-07, + 'epoch': 11.54} +04/20 [04:01:25] INFO | >> train_qwenlatent.py:487 + Step 45750 | grad_norm_pre_clip=0.1030 | + grad_norm_pre_clip_avg=0.1236 | Metrics: + {'align_loss': 0.025893591344356537, + 'recon_loss': 0.14710327982902527, + 'predict_loss': 0.005664133932441473, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10296555608510971, + 'mae_score': 0.004983655396882478, 'data_time': + 0.0007191830081865191, 'model_time': + 1.2315329469856806, 'grad_norm_pre_clip_avg': + 0.12363048493862153, 'learning_rate': + 5.70644359911595e-07, 'epoch': 11.54} +04/20 [04:01:37] INFO | >> train_qwenlatent.py:487 + Step 45760 | grad_norm_pre_clip=0.1229 | + grad_norm_pre_clip_avg=0.1156 | Metrics: + {'align_loss': 0.025849279016256332, + 'recon_loss': 0.16481392085552216, + 'predict_loss': 0.00784335844218731, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12287100404500961, + 'data_time': 0.0008559989801142365, + 'model_time': 1.2585896979726385, + 'grad_norm_pre_clip_avg': 0.11564527601003646, + 'learning_rate': 5.68098397257229e-07, 'epoch': + 11.55} +04/20 [04:01:50] INFO | >> train_qwenlatent.py:487 + Step 45770 | grad_norm_pre_clip=0.1395 | + grad_norm_pre_clip_avg=0.1138 | Metrics: + {'align_loss': 0.02547699213027954, + 'recon_loss': 0.10047510266304016, + 'predict_loss': 0.003709089942276478, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13948960602283478, + 'data_time': 0.000881487998412922, + 'model_time': 1.2599771190143656, + 'grad_norm_pre_clip_avg': 0.1137955941259861, + 'learning_rate': 5.655582561591036e-07, + 'epoch': 11.55} +04/20 [04:02:02] INFO | >> train_qwenlatent.py:487 + Step 45780 | grad_norm_pre_clip=0.1597 | + grad_norm_pre_clip_avg=0.1326 | Metrics: + {'align_loss': 0.02585093304514885, + 'recon_loss': 0.1275912970304489, + 'predict_loss': 0.003931786864995956, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15968725085258484, + 'data_time': 0.0013083470112178475, + 'model_time': 1.2596180760010611, + 'grad_norm_pre_clip_avg': 0.13256900534033775, + 'learning_rate': 5.630239378552522e-07, + 'epoch': 11.55} +04/20 [04:02:15] INFO | >> train_qwenlatent.py:487 + Step 45790 | grad_norm_pre_clip=0.1696 | + grad_norm_pre_clip_avg=0.1394 | Metrics: + {'align_loss': 0.025655414909124374, + 'recon_loss': 0.17165297269821167, + 'predict_loss': 0.0074018496088683605, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16956724226474762, + 'data_time': 0.0006834559899289161, + 'model_time': 1.246782032016199, + 'grad_norm_pre_clip_avg': 0.1394405297935009, + 'learning_rate': 5.604954435808672e-07, + 'epoch': 11.55} +04/20 [04:02:28] INFO | >> train_qwenlatent.py:487 + Step 45800 | grad_norm_pre_clip=0.2291 | + grad_norm_pre_clip_avg=0.1495 | Metrics: + {'align_loss': 0.0250991377979517, + 'recon_loss': 0.16137415170669556, + 'predict_loss': 0.012104032561182976, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22908616065979004, + 'mae_score': 0.005421625171695744, 'data_time': + 0.0008551240025553852, 'model_time': + 1.2363800760067534, 'grad_norm_pre_clip_avg': + 0.1494572676718235, 'learning_rate': + 5.579727745683104e-07, 'epoch': 11.56} +04/20 [04:02:41] INFO | >> train_qwenlatent.py:487 + Step 45810 | grad_norm_pre_clip=0.1335 | + grad_norm_pre_clip_avg=0.1242 | Metrics: + {'align_loss': 0.023545794188976288, + 'recon_loss': 0.13517621159553528, + 'predict_loss': 0.004206719342619181, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13349252939224243, + 'data_time': 0.0010466679814271629, + 'model_time': 1.2915604290028568, + 'grad_norm_pre_clip_avg': 0.12420227155089378, + 'learning_rate': 5.554559320470969e-07, + 'epoch': 11.56} +04/20 [04:02:54] INFO | >> train_qwenlatent.py:487 + Step 45820 | grad_norm_pre_clip=0.0987 | + grad_norm_pre_clip_avg=0.1306 | Metrics: + {'align_loss': 0.024204164743423462, + 'recon_loss': 0.15825648605823517, + 'predict_loss': 0.006453191861510277, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09872214496135712, + 'data_time': 0.0011166239855811, 'model_time': + 1.2933732550009154, 'grad_norm_pre_clip_avg': + 0.13059102222323418, 'learning_rate': + 5.52944917243907e-07, 'epoch': 11.56} +04/20 [04:03:06] INFO | >> train_qwenlatent.py:487 + Step 45830 | grad_norm_pre_clip=0.1497 | + grad_norm_pre_clip_avg=0.1188 | Metrics: + {'align_loss': 0.02563280612230301, + 'recon_loss': 0.0960945412516594, + 'predict_loss': 0.003971820697188377, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14972417056560516, + 'data_time': 0.0012008599878754467, + 'model_time': 1.3104485019866843, + 'grad_norm_pre_clip_avg': 0.11877385303378105, + 'learning_rate': 5.504397313825773e-07, + 'epoch': 11.56} +04/20 [04:03:19] INFO | >> train_qwenlatent.py:487 + Step 45840 | grad_norm_pre_clip=0.1546 | + grad_norm_pre_clip_avg=0.1312 | Metrics: + {'align_loss': 0.025655020028352737, + 'recon_loss': 0.16600920259952545, + 'predict_loss': 0.007925362326204777, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15459348261356354, + 'data_time': 0.0007365320052485913, + 'model_time': 1.2391593009815551, + 'grad_norm_pre_clip_avg': 0.13124893754720687, + 'learning_rate': 5.479403756841066e-07, + 'epoch': 11.57} +04/20 [04:03:32] INFO | >> train_qwenlatent.py:487 + Step 45850 | grad_norm_pre_clip=0.1370 | + grad_norm_pre_clip_avg=0.1308 | Metrics: + {'align_loss': 0.024484770372509956, + 'recon_loss': 0.12464982271194458, + 'predict_loss': 0.006226424593478441, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1370205283164978, + 'mae_score': 0.00539715998881572, 'data_time': + 0.0008929579926189035, 'model_time': + 1.2338969970005564, 'grad_norm_pre_clip_avg': + 0.13078200444579124, 'learning_rate': + 5.454468513666486e-07, 'epoch': 11.57} +04/20 [04:03:45] INFO | >> train_qwenlatent.py:487 + Step 45860 | grad_norm_pre_clip=0.0797 | + grad_norm_pre_clip_avg=0.1411 | Metrics: + {'align_loss': 0.025118451565504074, + 'recon_loss': 0.13569042086601257, + 'predict_loss': 0.0049056848511099815, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.0796874389052391, + 'data_time': 0.0006845199968665838, + 'model_time': 1.2253781470062677, + 'grad_norm_pre_clip_avg': 0.1411239691078663, + 'learning_rate': 5.429591596455179e-07, + 'epoch': 11.57} +04/20 [04:03:57] INFO | >> train_qwenlatent.py:487 + Step 45870 | grad_norm_pre_clip=0.1133 | + grad_norm_pre_clip_avg=0.1327 | Metrics: + {'align_loss': 0.025468777865171432, + 'recon_loss': 0.1059175506234169, + 'predict_loss': 0.005459187086671591, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11332908272743225, + 'data_time': 0.0007920229982119054, + 'model_time': 1.2448851139924955, + 'grad_norm_pre_clip_avg': 0.1327395848929882, + 'learning_rate': 5.404773017331868e-07, + 'epoch': 11.57} +04/20 [04:04:10] INFO | >> train_qwenlatent.py:487 + Step 45880 | grad_norm_pre_clip=0.1067 | + grad_norm_pre_clip_avg=0.1411 | Metrics: + {'align_loss': 0.024667028337717056, + 'recon_loss': 0.09916859120130539, + 'predict_loss': 0.0030463605653494596, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10665755718946457, + 'data_time': 0.0017049350135494024, + 'model_time': 1.2333724570053164, + 'grad_norm_pre_clip_avg': 0.14112263545393944, + 'learning_rate': 5.380012788392803e-07, + 'epoch': 11.58} +04/20 [04:04:23] INFO | >> train_qwenlatent.py:487 + Step 45890 | grad_norm_pre_clip=0.1440 | + grad_norm_pre_clip_avg=0.1502 | Metrics: + {'align_loss': 0.02541903406381607, + 'recon_loss': 0.10992024838924408, + 'predict_loss': 0.0037140189670026302, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14397670328617096, + 'data_time': 0.0006684790132567286, + 'model_time': 1.2515955900016706, + 'grad_norm_pre_clip_avg': 0.15016639679670335, + 'learning_rate': 5.355310921705823e-07, + 'epoch': 11.58} +04/20 [04:04:36] INFO | >> train_qwenlatent.py:487 + Step 45900 | grad_norm_pre_clip=0.1886 | + grad_norm_pre_clip_avg=0.1482 | Metrics: + {'align_loss': 0.024839643388986588, + 'recon_loss': 0.15972499549388885, + 'predict_loss': 0.008703654631972313, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18863046169281006, + 'mae_score': 0.005767786180650865, 'data_time': + 0.0007146209827624261, 'model_time': + 1.2235406280087773, 'grad_norm_pre_clip_avg': + 0.1481526993215084, 'learning_rate': + 5.330667429310318e-07, 'epoch': 11.58} +04/20 [04:04:49] INFO | >> train_qwenlatent.py:487 + Step 45910 | grad_norm_pre_clip=0.1249 | + grad_norm_pre_clip_avg=0.1431 | Metrics: + {'align_loss': 0.02592722699046135, + 'recon_loss': 0.13967181742191315, + 'predict_loss': 0.006562130991369486, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12494984269142151, + 'data_time': 0.0009889240027405322, + 'model_time': 1.2271611570031382, + 'grad_norm_pre_clip_avg': 0.14305555522441865, + 'learning_rate': 5.306082323217235e-07, + 'epoch': 11.58} +04/20 [04:05:02] INFO | >> train_qwenlatent.py:487 + Step 45920 | grad_norm_pre_clip=0.1191 | + grad_norm_pre_clip_avg=0.1346 | Metrics: + {'align_loss': 0.02657245472073555, + 'recon_loss': 0.20907631516456604, + 'predict_loss': 0.009897969663143158, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11905685067176819, + 'data_time': 0.0009315850038547069, + 'model_time': 1.1981577140104491, + 'grad_norm_pre_clip_avg': 0.1346260167658329, + 'learning_rate': 5.28155561540905e-07, 'epoch': + 11.59} +04/20 [04:05:14] INFO | >> train_qwenlatent.py:487 + Step 45930 | grad_norm_pre_clip=0.0964 | + grad_norm_pre_clip_avg=0.1582 | Metrics: + {'align_loss': 0.02309204265475273, + 'recon_loss': 0.18583835661411285, + 'predict_loss': 0.00655041029676795, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09642329812049866, + 'data_time': 0.0006655070174019784, + 'model_time': 1.2250991100154351, + 'grad_norm_pre_clip_avg': 0.1581992194056511, + 'learning_rate': 5.257087317839783e-07, + 'epoch': 11.59} +04/20 [04:05:27] INFO | >> train_qwenlatent.py:487 + Step 45940 | grad_norm_pre_clip=0.1442 | + grad_norm_pre_clip_avg=0.1452 | Metrics: + {'align_loss': 0.025333795696496964, + 'recon_loss': 0.12924082577228546, + 'predict_loss': 0.0042426493018865585, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14421257376670837, + 'data_time': 0.0012069490039721131, + 'model_time': 1.6318799949949607, + 'grad_norm_pre_clip_avg': 0.14521772265434266, + 'learning_rate': 5.23267744243499e-07, 'epoch': + 11.59} +04/20 [04:05:40] INFO | >> train_qwenlatent.py:487 + Step 45950 | grad_norm_pre_clip=0.0763 | + grad_norm_pre_clip_avg=0.1210 | Metrics: + {'align_loss': 0.025296572595834732, + 'recon_loss': 0.20192159712314606, + 'predict_loss': 0.007728774566203356, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.07625358551740646, + 'mae_score': 0.006398899920351871, 'data_time': + 0.0007459510234184563, 'model_time': + 1.2043264069943689, 'grad_norm_pre_clip_avg': + 0.12099978551268578, 'learning_rate': + 5.208326001091736e-07, 'epoch': 11.59} +04/20 [04:05:53] INFO | >> train_qwenlatent.py:487 + Step 45960 | grad_norm_pre_clip=0.1156 | + grad_norm_pre_clip_avg=0.1332 | Metrics: + {'align_loss': 0.025418482720851898, + 'recon_loss': 0.15203478932380676, + 'predict_loss': 0.0053784893825650215, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11557066440582275, + 'data_time': 0.0009804510045796633, + 'model_time': 1.228842768992763, + 'grad_norm_pre_clip_avg': 0.13322493061423302, + 'learning_rate': 5.18403300567862e-07, 'epoch': + 11.6} +04/20 [04:06:05] INFO | >> train_qwenlatent.py:487 + Step 45970 | grad_norm_pre_clip=0.1350 | + grad_norm_pre_clip_avg=0.1346 | Metrics: + {'align_loss': 0.026165161281824112, + 'recon_loss': 0.07915238291025162, + 'predict_loss': 0.003950652666389942, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13503532111644745, + 'data_time': 0.0006907469942234457, + 'model_time': 1.180817501997808, + 'grad_norm_pre_clip_avg': 0.13461359217762947, + 'learning_rate': 5.159798468035784e-07, + 'epoch': 11.6} +04/20 [04:06:18] INFO | >> train_qwenlatent.py:487 + Step 45980 | grad_norm_pre_clip=0.1381 | + grad_norm_pre_clip_avg=0.1406 | Metrics: + {'align_loss': 0.024920761585235596, + 'recon_loss': 0.12750789523124695, + 'predict_loss': 0.005631524138152599, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13814634084701538, + 'data_time': 0.000788973004091531, + 'model_time': 1.3036146890081, + 'grad_norm_pre_clip_avg': 0.14063202291727067, + 'learning_rate': 5.135622399974808e-07, + 'epoch': 11.6} +04/20 [04:06:30] INFO | >> train_qwenlatent.py:487 + Step 45990 | grad_norm_pre_clip=0.1516 | + grad_norm_pre_clip_avg=0.1227 | Metrics: + {'align_loss': 0.025581272318959236, + 'recon_loss': 0.16102169454097748, + 'predict_loss': 0.004378459416329861, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15155822038650513, + 'data_time': 0.000744280987419188, + 'model_time': 1.2487843080016319, + 'grad_norm_pre_clip_avg': 0.12271601781249046, + 'learning_rate': 5.11150481327881e-07, 'epoch': + 11.6} +04/20 [04:06:44] INFO | >> train_qwenlatent.py:487 + Step 46000 | grad_norm_pre_clip=0.1751 | + grad_norm_pre_clip_avg=0.1406 | Metrics: + {'align_loss': 0.025404900312423706, + 'recon_loss': 0.18513593077659607, + 'predict_loss': 0.004855144768953323, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1750635951757431, + 'mae_score': 0.0058009182010685, 'data_time': + 0.0012482389865908772, 'model_time': + 1.3188699420134071, 'grad_norm_pre_clip_avg': + 0.14064619466662406, 'learning_rate': + 5.087445719702429e-07, 'epoch': 11.61} +04/20 [04:06:56] INFO | >> train_qwenlatent.py:487 + Step 46010 | grad_norm_pre_clip=0.1298 | + grad_norm_pre_clip_avg=0.1341 | Metrics: + {'align_loss': 0.02477814257144928, + 'recon_loss': 0.10555785894393921, + 'predict_loss': 0.005280831828713417, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1297730654478073, + 'data_time': 0.0010090499999932945, + 'model_time': 1.278855932992883, + 'grad_norm_pre_clip_avg': 0.13406258523464204, + 'learning_rate': 5.063445130971789e-07, + 'epoch': 11.61} +04/20 [04:07:09] INFO | >> train_qwenlatent.py:487 + Step 46020 | grad_norm_pre_clip=0.1169 | + grad_norm_pre_clip_avg=0.1276 | Metrics: + {'align_loss': 0.025441350415349007, + 'recon_loss': 0.08567679673433304, + 'predict_loss': 0.0035994427744299173, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11686569452285767, + 'data_time': 0.000671194982714951, + 'model_time': 1.2716709300002549, + 'grad_norm_pre_clip_avg': 0.12756961435079575, + 'learning_rate': 5.039503058784453e-07, + 'epoch': 11.61} +04/20 [04:07:21] INFO | >> train_qwenlatent.py:487 + Step 46030 | grad_norm_pre_clip=0.1271 | + grad_norm_pre_clip_avg=0.1396 | Metrics: + {'align_loss': 0.023859096691012383, + 'recon_loss': 0.10415329039096832, + 'predict_loss': 0.008277679793536663, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12706702947616577, + 'data_time': 0.0009501540043856949, + 'model_time': 1.3183129329991061, + 'grad_norm_pre_clip_avg': 0.13963972106575967, + 'learning_rate': 5.01561951480952e-07, 'epoch': + 11.61} +04/20 [04:07:34] INFO | >> train_qwenlatent.py:487 + Step 46040 | grad_norm_pre_clip=0.1484 | + grad_norm_pre_clip_avg=0.1405 | Metrics: + {'align_loss': 0.02530619502067566, + 'recon_loss': 0.13496212661266327, + 'predict_loss': 0.008509106934070587, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14843708276748657, + 'data_time': 0.0009855160024017096, + 'model_time': 1.306532870978117, + 'grad_norm_pre_clip_avg': 0.14047346785664558, + 'learning_rate': 4.99179451068752e-07, 'epoch': + 11.62} +04/20 [04:07:48] INFO | >> train_qwenlatent.py:487 + Step 46050 | grad_norm_pre_clip=0.1157 | + grad_norm_pre_clip_avg=0.1250 | Metrics: + {'align_loss': 0.023363951593637466, + 'recon_loss': 0.0961913987994194, + 'predict_loss': 0.0028713077772408724, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11568460613489151, + 'mae_score': 0.004785659721305778, 'data_time': + 0.0009681679948698729, 'model_time': + 1.2954106440010946, 'grad_norm_pre_clip_avg': + 0.12504139617085458, 'learning_rate': + 4.968028058030488e-07, 'epoch': 11.62} +04/20 [04:08:00] INFO | >> train_qwenlatent.py:487 + Step 46060 | grad_norm_pre_clip=0.1278 | + grad_norm_pre_clip_avg=0.1362 | Metrics: + {'align_loss': 0.026234522461891174, + 'recon_loss': 0.1466471552848816, + 'predict_loss': 0.008140571415424347, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1278163194656372, + 'data_time': 0.0008382700034417212, + 'model_time': 1.2516787740169093, + 'grad_norm_pre_clip_avg': 0.13616404458880424, + 'learning_rate': 4.944320168421915e-07, + 'epoch': 11.62} +04/20 [04:08:13] INFO | >> train_qwenlatent.py:487 + Step 46070 | grad_norm_pre_clip=0.1154 | + grad_norm_pre_clip_avg=0.1148 | Metrics: + {'align_loss': 0.02446041628718376, + 'recon_loss': 0.11589781939983368, + 'predict_loss': 0.004436154384166002, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11540557444095612, + 'data_time': 0.001189567003166303, + 'model_time': 1.3393033719912637, + 'grad_norm_pre_clip_avg': 0.11478974148631096, + 'learning_rate': 4.920670853416704e-07, + 'epoch': 11.63} +04/20 [04:08:26] INFO | >> train_qwenlatent.py:487 + Step 46080 | grad_norm_pre_clip=0.2258 | + grad_norm_pre_clip_avg=0.1308 | Metrics: + {'align_loss': 0.02544419839978218, + 'recon_loss': 0.1296764761209488, + 'predict_loss': 0.005637234542518854, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.22584140300750732, + 'data_time': 0.0009818389953579754, + 'model_time': 1.2139297829999123, + 'grad_norm_pre_clip_avg': 0.13079492300748824, + 'learning_rate': 4.897080124541256e-07, + 'epoch': 11.63} +04/20 [04:08:38] INFO | >> train_qwenlatent.py:487 + Step 46090 | grad_norm_pre_clip=0.1050 | + grad_norm_pre_clip_avg=0.1400 | Metrics: + {'align_loss': 0.025541877374053, 'recon_loss': + 0.19969667494297028, 'predict_loss': + 0.006093149539083242, 'aux_loss_decay_weight': + 0.0, 'grad_norm_pre_clip': 0.1049654632806778, + 'data_time': 0.0007240790000651032, + 'model_time': 1.2245894299994688, + 'grad_norm_pre_clip_avg': 0.13998521715402604, + 'learning_rate': 4.873547993293409e-07, + 'epoch': 11.63} +04/20 [04:08:51] INFO | >> train_qwenlatent.py:487 + Step 46100 | grad_norm_pre_clip=0.1062 | + grad_norm_pre_clip_avg=0.1364 | Metrics: + {'align_loss': 0.02562093362212181, + 'recon_loss': 0.19638872146606445, + 'predict_loss': 0.008952177129685879, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10621189326047897, + 'mae_score': 0.005499898635589325, 'data_time': + 0.0009934440022334456, 'model_time': + 1.2163258180080447, 'grad_norm_pre_clip_avg': + 0.13636591732501985, 'learning_rate': + 4.850074471142444e-07, 'epoch': 11.63} +04/20 [04:09:04] INFO | >> train_qwenlatent.py:487 + Step 46110 | grad_norm_pre_clip=0.1251 | + grad_norm_pre_clip_avg=0.1249 | Metrics: + {'align_loss': 0.02507186308503151, + 'recon_loss': 0.11140522360801697, + 'predict_loss': 0.004474271554499865, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12509898841381073, + 'data_time': 0.0006719649827573448, + 'model_time': 1.223387254984118, + 'grad_norm_pre_clip_avg': 0.12487714216113091, + 'learning_rate': 4.826659569529068e-07, + 'epoch': 11.64} +04/20 [04:09:17] INFO | >> train_qwenlatent.py:487 + Step 46120 | grad_norm_pre_clip=0.1495 | + grad_norm_pre_clip_avg=0.1402 | Metrics: + {'align_loss': 0.024619560688734055, + 'recon_loss': 0.15660971403121948, + 'predict_loss': 0.004503678996115923, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1495264321565628, + 'data_time': 0.0007500909850932658, + 'model_time': 1.2755355119879823, + 'grad_norm_pre_clip_avg': 0.14021633937954903, + 'learning_rate': 4.803303299865413e-07, + 'epoch': 11.64} +04/20 [04:09:29] INFO | >> train_qwenlatent.py:487 + Step 46130 | grad_norm_pre_clip=0.1666 | + grad_norm_pre_clip_avg=0.1381 | Metrics: + {'align_loss': 0.025014810264110565, + 'recon_loss': 0.1780424863100052, + 'predict_loss': 0.008280565030872822, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16659530997276306, + 'data_time': 0.0014096569793764502, + 'model_time': 1.19131710100919, + 'grad_norm_pre_clip_avg': 0.13808633685111998, + 'learning_rate': 4.780005673535038e-07, + 'epoch': 11.64} +04/20 [04:09:42] INFO | >> train_qwenlatent.py:487 + Step 46140 | grad_norm_pre_clip=0.1111 | + grad_norm_pre_clip_avg=0.1342 | Metrics: + {'align_loss': 0.024206582456827164, + 'recon_loss': 0.13739125430583954, + 'predict_loss': 0.006209282204508781, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11108339577913284, + 'data_time': 0.0010798930015880615, + 'model_time': 1.2909803680086043, + 'grad_norm_pre_clip_avg': 0.13421876057982446, + 'learning_rate': 4.7567667018929294e-07, + 'epoch': 11.64} +04/20 [04:09:55] INFO | >> train_qwenlatent.py:487 + Step 46150 | grad_norm_pre_clip=0.1428 | + grad_norm_pre_clip_avg=0.1416 | Metrics: + {'align_loss': 0.02712198905646801, + 'recon_loss': 0.12336549907922745, + 'predict_loss': 0.0026523328851908445, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1427909880876541, + 'mae_score': 0.005661005587191195, 'data_time': + 0.0007152670004870743, 'model_time': + 1.2521125789789949, 'grad_norm_pre_clip_avg': + 0.14158963188529014, 'learning_rate': + 4.733586396265514e-07, 'epoch': 11.65} +04/20 [04:10:08] INFO | >> train_qwenlatent.py:487 + Step 46160 | grad_norm_pre_clip=0.1240 | + grad_norm_pre_clip_avg=0.1260 | Metrics: + {'align_loss': 0.024678293615579605, + 'recon_loss': 0.1616855412721634, + 'predict_loss': 0.009504529647529125, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1239645853638649, + 'data_time': 0.0006389750051312149, + 'model_time': 1.191050476016244, + 'grad_norm_pre_clip_avg': 0.12595188543200492, + 'learning_rate': 4.7104647679505345e-07, + 'epoch': 11.65} +04/20 [04:10:21] INFO | >> train_qwenlatent.py:487 + Step 46170 | grad_norm_pre_clip=0.1262 | + grad_norm_pre_clip_avg=0.1360 | Metrics: + {'align_loss': 0.02539970353245735, + 'recon_loss': 0.1743394434452057, + 'predict_loss': 0.005456269718706608, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1261555254459381, + 'data_time': 0.0007164349954109639, + 'model_time': 1.2208908010215964, + 'grad_norm_pre_clip_avg': 0.13598253205418587, + 'learning_rate': 4.687401828217212e-07, + 'epoch': 11.65} +04/20 [04:10:34] INFO | >> train_qwenlatent.py:487 + Step 46180 | grad_norm_pre_clip=0.1218 | + grad_norm_pre_clip_avg=0.1257 | Metrics: + {'align_loss': 0.025853101164102554, + 'recon_loss': 0.10850103199481964, + 'predict_loss': 0.004411641508340836, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12180145829916, + 'data_time': 0.0007092779851518571, + 'model_time': 1.2955764190119226, + 'grad_norm_pre_clip_avg': 0.12571452409029008, + 'learning_rate': 4.6643975883061296e-07, + 'epoch': 11.65} +04/20 [04:10:46] INFO | >> train_qwenlatent.py:487 + Step 46190 | grad_norm_pre_clip=0.1621 | + grad_norm_pre_clip_avg=0.1365 | Metrics: + {'align_loss': 0.025685861706733704, + 'recon_loss': 0.10233557224273682, + 'predict_loss': 0.0058539854362607, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16210660338401794, + 'data_time': 0.0010937050101347268, + 'model_time': 1.2505309580010362, + 'grad_norm_pre_clip_avg': 0.13645273000001906, + 'learning_rate': 4.641452059429309e-07, + 'epoch': 11.66} +04/20 [04:10:59] INFO | >> train_qwenlatent.py:487 + Step 46200 | grad_norm_pre_clip=0.1098 | + grad_norm_pre_clip_avg=0.1299 | Metrics: + {'align_loss': 0.026768773794174194, + 'recon_loss': 0.17966215312480927, + 'predict_loss': 0.00789056159555912, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10983936488628387, + 'mae_score': 0.00561197770608438, 'data_time': + 0.0008902910049073398, 'model_time': + 1.2679248150088824, 'grad_norm_pre_clip_avg': + 0.12985295951366424, 'learning_rate': + 4.618565252770112e-07, 'epoch': 11.66} +04/20 [04:11:12] INFO | >> train_qwenlatent.py:487 + Step 46210 | grad_norm_pre_clip=0.1356 | + grad_norm_pre_clip_avg=0.1382 | Metrics: + {'align_loss': 0.026621798053383827, + 'recon_loss': 0.19669190049171448, + 'predict_loss': 0.007273863069713116, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13556669652462006, + 'data_time': 0.0007137490028981119, + 'model_time': 1.2365833039802965, + 'grad_norm_pre_clip_avg': 0.138209118694067, + 'learning_rate': 4.5957371794832773e-07, + 'epoch': 11.66} +04/20 [04:11:25] INFO | >> train_qwenlatent.py:487 + Step 46220 | grad_norm_pre_clip=0.1228 | + grad_norm_pre_clip_avg=0.1199 | Metrics: + {'align_loss': 0.026120154187083244, + 'recon_loss': 0.21390032768249512, + 'predict_loss': 0.007556942291557789, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12276127934455872, + 'data_time': 0.0009632139990571886, + 'model_time': 1.3372739479818847, + 'grad_norm_pre_clip_avg': 0.11986882984638214, + 'learning_rate': 4.572967850694924e-07, + 'epoch': 11.66} +04/20 [04:11:37] INFO | >> train_qwenlatent.py:487 + Step 46230 | grad_norm_pre_clip=0.1093 | + grad_norm_pre_clip_avg=0.1296 | Metrics: + {'align_loss': 0.026134483516216278, + 'recon_loss': 0.14210668206214905, + 'predict_loss': 0.006231748964637518, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10931871831417084, + 'data_time': 0.0009811299969442189, + 'model_time': 1.2644338410173077, + 'grad_norm_pre_clip_avg': 0.12959475740790366, + 'learning_rate': 4.5502572775025724e-07, + 'epoch': 11.67} +04/20 [04:11:50] INFO | >> train_qwenlatent.py:487 + Step 46240 | grad_norm_pre_clip=0.1547 | + grad_norm_pre_clip_avg=0.1181 | Metrics: + {'align_loss': 0.02583586610853672, + 'recon_loss': 0.15937882661819458, + 'predict_loss': 0.006223755422979593, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15467388927936554, + 'data_time': 0.0009933410037774593, + 'model_time': 1.2936709560162853, + 'grad_norm_pre_clip_avg': 0.11809653788805008, + 'learning_rate': 4.527605470975056e-07, + 'epoch': 11.67} +04/20 [04:12:03] INFO | >> train_qwenlatent.py:487 + Step 46250 | grad_norm_pre_clip=0.1002 | + grad_norm_pre_clip_avg=0.1229 | Metrics: + {'align_loss': 0.025134321302175522, + 'recon_loss': 0.11978167295455933, + 'predict_loss': 0.006193328183144331, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10019327700138092, + 'mae_score': 0.004631722939980996, 'data_time': + 0.0006994959840085357, 'model_time': + 1.2245033019862603, 'grad_norm_pre_clip_avg': + 0.12293583750724793, 'learning_rate': + 4.505012442152609e-07, 'epoch': 11.67} +04/20 [04:12:15] INFO | >> train_qwenlatent.py:487 + Step 46260 | grad_norm_pre_clip=0.1479 | + grad_norm_pre_clip_avg=0.1224 | Metrics: + {'align_loss': 0.0252611692994833, + 'recon_loss': 0.1589147448539734, + 'predict_loss': 0.007696529850363731, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14789196848869324, + 'data_time': 0.0009228610142599791, + 'model_time': 1.2086678560008295, + 'grad_norm_pre_clip_avg': 0.12241734489798546, + 'learning_rate': 4.4824782020467803e-07, + 'epoch': 11.67} +04/20 [04:12:28] INFO | >> train_qwenlatent.py:487 + Step 46270 | grad_norm_pre_clip=0.1166 | + grad_norm_pre_clip_avg=0.1306 | Metrics: + {'align_loss': 0.024479292333126068, + 'recon_loss': 0.14424721896648407, + 'predict_loss': 0.006812699139118195, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11663752049207687, + 'data_time': 0.0007135559862945229, + 'model_time': 1.2687689230078831, + 'grad_norm_pre_clip_avg': 0.13058009296655654, + 'learning_rate': 4.460002761640505e-07, + 'epoch': 11.68} +04/20 [04:12:40] INFO | >> train_qwenlatent.py:487 + Step 46280 | grad_norm_pre_clip=0.1580 | + grad_norm_pre_clip_avg=0.1214 | Metrics: + {'align_loss': 0.02639715000987053, + 'recon_loss': 0.18387706577777863, + 'predict_loss': 0.007875747978687286, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1580134630203247, + 'data_time': 0.001136208011303097, + 'model_time': 1.234579459996894, + 'grad_norm_pre_clip_avg': 0.1214160032570362, + 'learning_rate': 4.4375861318880025e-07, + 'epoch': 11.68} +04/20 [04:12:54] INFO | >> train_qwenlatent.py:487 + Step 46290 | grad_norm_pre_clip=0.1086 | + grad_norm_pre_clip_avg=0.1318 | Metrics: + {'align_loss': 0.02541438862681389, + 'recon_loss': 0.18788935244083405, + 'predict_loss': 0.006660140119493008, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10857919603586197, + 'data_time': 0.0010124680120497942, + 'model_time': 1.4273201929754578, + 'grad_norm_pre_clip_avg': 0.1318365216255188, + 'learning_rate': 4.41522832371491e-07, 'epoch': + 11.68} +04/20 [04:13:07] INFO | >> train_qwenlatent.py:487 + Step 46300 | grad_norm_pre_clip=0.2005 | + grad_norm_pre_clip_avg=0.1256 | Metrics: + {'align_loss': 0.026284441351890564, + 'recon_loss': 0.092341348528862, + 'predict_loss': 0.004183918237686157, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.20050400495529175, + 'mae_score': 0.005524089744499138, 'data_time': + 0.0006712879985570908, 'model_time': + 1.21830013699946, 'grad_norm_pre_clip_avg': + 0.12561430186033248, 'learning_rate': + 4.392929348018136e-07, 'epoch': 11.68} +04/20 [04:13:19] INFO | >> train_qwenlatent.py:487 + Step 46310 | grad_norm_pre_clip=0.1641 | + grad_norm_pre_clip_avg=0.1390 | Metrics: + {'align_loss': 0.025831308215856552, + 'recon_loss': 0.1446455419063568, + 'predict_loss': 0.007774739991873503, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16405506432056427, + 'data_time': 0.05915635300334543, 'model_time': + 1.163818657019874, 'grad_norm_pre_clip_avg': + 0.13897797837853432, 'learning_rate': + 4.3706892156659303e-07, 'epoch': 11.69} +04/20 [04:13:31] INFO | >> train_qwenlatent.py:487 + Step 46320 | grad_norm_pre_clip=0.1459 | + grad_norm_pre_clip_avg=0.1341 | Metrics: + {'align_loss': 0.025420386344194412, + 'recon_loss': 0.2408389002084732, + 'predict_loss': 0.010350391268730164, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1459347903728485, + 'data_time': 0.0006245879922062159, + 'model_time': 1.1744245900190435, + 'grad_norm_pre_clip_avg': 0.13411541506648064, + 'learning_rate': 4.3485079374978475e-07, + 'epoch': 11.69} +04/20 [04:13:43] INFO | >> train_qwenlatent.py:487 + Step 46330 | grad_norm_pre_clip=0.0946 | + grad_norm_pre_clip_avg=0.1212 | Metrics: + {'align_loss': 0.026775717735290527, + 'recon_loss': 0.18475867807865143, + 'predict_loss': 0.005949567072093487, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09460096061229706, + 'data_time': 0.0006425390020012856, + 'model_time': 1.174514502985403, + 'grad_norm_pre_clip_avg': 0.1212038241326809, + 'learning_rate': 4.3263855243247986e-07, + 'epoch': 11.69} +04/20 [04:13:55] INFO | >> train_qwenlatent.py:487 + Step 46340 | grad_norm_pre_clip=0.1403 | + grad_norm_pre_clip_avg=0.1526 | Metrics: + {'align_loss': 0.024769913405179977, + 'recon_loss': 0.16674134135246277, + 'predict_loss': 0.007502818945795298, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1403094381093979, + 'data_time': 0.000703705009073019, + 'model_time': 1.1605041800066829, + 'grad_norm_pre_clip_avg': 0.15258261263370515, + 'learning_rate': 4.304321986928967e-07, + 'epoch': 11.69} +04/20 [04:14:07] INFO | >> train_qwenlatent.py:487 + Step 46350 | grad_norm_pre_clip=0.1207 | + grad_norm_pre_clip_avg=0.1235 | Metrics: + {'align_loss': 0.025243239477276802, + 'recon_loss': 0.08970312029123306, + 'predict_loss': 0.003612180706113577, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12066172808408737, + 'mae_score': 0.005825746811188019, 'data_time': + 0.0006441179721150547, 'model_time': + 1.1413939080084674, 'grad_norm_pre_clip_avg': + 0.12352885156869889, 'learning_rate': + 4.282317336063855e-07, 'epoch': 11.7} +04/20 [04:14:19] INFO | >> train_qwenlatent.py:487 + Step 46360 | grad_norm_pre_clip=0.1818 | + grad_norm_pre_clip_avg=0.1207 | Metrics: + {'align_loss': 0.025730159133672714, + 'recon_loss': 0.15090793371200562, + 'predict_loss': 0.006849242839962244, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18175175786018372, + 'data_time': 0.0006187490071170032, + 'model_time': 1.1514593980100472, + 'grad_norm_pre_clip_avg': 0.1206735722720623, + 'learning_rate': 4.2603715824542487e-07, + 'epoch': 11.7} +04/20 [04:14:30] INFO | >> train_qwenlatent.py:487 + Step 46370 | grad_norm_pre_clip=0.1813 | + grad_norm_pre_clip_avg=0.1300 | Metrics: + {'align_loss': 0.02536081150174141, + 'recon_loss': 0.17225633561611176, + 'predict_loss': 0.006298276595771313, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18130336701869965, + 'data_time': 0.0006561980117112398, + 'model_time': 1.1642523929767776, + 'grad_norm_pre_clip_avg': 0.1299939215183258, + 'learning_rate': 4.2384847367962524e-07, + 'epoch': 11.7} +04/20 [04:14:42] INFO | >> train_qwenlatent.py:487 + Step 46380 | grad_norm_pre_clip=0.1265 | + grad_norm_pre_clip_avg=0.1203 | Metrics: + {'align_loss': 0.026638472452759743, + 'recon_loss': 0.15420381724834442, + 'predict_loss': 0.004789239261299372, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12652160227298737, + 'data_time': 0.0006414110248442739, + 'model_time': 1.1522881089767907, + 'grad_norm_pre_clip_avg': 0.12034394666552543, + 'learning_rate': 4.2166568097572583e-07, + 'epoch': 11.7} +04/20 [04:14:54] INFO | >> train_qwenlatent.py:487 + Step 46390 | grad_norm_pre_clip=0.1405 | + grad_norm_pre_clip_avg=0.1264 | Metrics: + {'align_loss': 0.02609209157526493, + 'recon_loss': 0.13897481560707092, + 'predict_loss': 0.005753163248300552, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14052565395832062, + 'data_time': 0.0005890200263820589, + 'model_time': 1.1604395800095517, + 'grad_norm_pre_clip_avg': 0.1263783171772957, + 'learning_rate': 4.194887811975916e-07, + 'epoch': 11.71} +04/20 [04:15:06] INFO | >> train_qwenlatent.py:487 + Step 46400 | grad_norm_pre_clip=0.1356 | + grad_norm_pre_clip_avg=0.1259 | Metrics: + {'align_loss': 0.026029769331216812, + 'recon_loss': 0.1958235651254654, + 'predict_loss': 0.005071484483778477, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1356450766324997, + 'mae_score': 0.004708595533628721, 'data_time': + 0.0007185199938248843, 'model_time': + 1.1455328830052167, 'grad_norm_pre_clip_avg': + 0.12591150924563407, 'learning_rate': + 4.173177754062195e-07, 'epoch': 11.71} +04/20 [04:15:17] INFO | >> train_qwenlatent.py:487 + Step 46410 | grad_norm_pre_clip=0.1131 | + grad_norm_pre_clip_avg=0.1093 | Metrics: + {'align_loss': 0.025240296497941017, + 'recon_loss': 0.15310685336589813, + 'predict_loss': 0.011328906752169132, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11310699582099915, + 'data_time': 0.0005842679820489138, + 'model_time': 1.1548296440159902, + 'grad_norm_pre_clip_avg': 0.10928058698773384, + 'learning_rate': 4.1515266465972795e-07, + 'epoch': 11.71} +04/20 [04:15:29] INFO | >> train_qwenlatent.py:487 + Step 46420 | grad_norm_pre_clip=0.1657 | + grad_norm_pre_clip_avg=0.1253 | Metrics: + {'align_loss': 0.02368622086942196, + 'recon_loss': 0.17069785296916962, + 'predict_loss': 0.00856147799640894, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16574116051197052, + 'data_time': 0.0005549199995584786, + 'model_time': 1.160844193014782, + 'grad_norm_pre_clip_avg': 0.12532217651605607, + 'learning_rate': 4.1299345001336847e-07, + 'epoch': 11.71} +04/20 [04:15:41] INFO | >> train_qwenlatent.py:487 + Step 46430 | grad_norm_pre_clip=0.1578 | + grad_norm_pre_clip_avg=0.1283 | Metrics: + {'align_loss': 0.025139328092336655, + 'recon_loss': 0.16678458452224731, + 'predict_loss': 0.011401539668440819, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15775305032730103, + 'data_time': 0.0005850129819009453, + 'model_time': 1.1640148690203205, + 'grad_norm_pre_clip_avg': 0.1283285990357399, + 'learning_rate': 4.108401325195146e-07, + 'epoch': 11.72} +04/20 [04:15:52] INFO | >> train_qwenlatent.py:487 + Step 46440 | grad_norm_pre_clip=0.0748 | + grad_norm_pre_clip_avg=0.1225 | Metrics: + {'align_loss': 0.025816209614276886, + 'recon_loss': 0.12108826637268066, + 'predict_loss': 0.0038698140997439623, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.07484306395053864, + 'data_time': 0.0006901150045450777, + 'model_time': 1.160594160988694, + 'grad_norm_pre_clip_avg': 0.12248557955026626, + 'learning_rate': 4.086927132276682e-07, + 'epoch': 11.72} +04/20 [04:16:04] INFO | >> train_qwenlatent.py:487 + Step 46450 | grad_norm_pre_clip=0.1691 | + grad_norm_pre_clip_avg=0.1337 | Metrics: + {'align_loss': 0.026069261133670807, + 'recon_loss': 0.16878703236579895, + 'predict_loss': 0.007035873364657164, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16910380125045776, + 'mae_score': 0.00696167516278791, 'data_time': + 0.0006274509942159057, 'model_time': + 1.163050981995184, 'grad_norm_pre_clip_avg': + 0.13373500630259513, 'learning_rate': + 4.065511931844535e-07, 'epoch': 11.72} +04/20 [04:16:16] INFO | >> train_qwenlatent.py:487 + Step 46460 | grad_norm_pre_clip=0.1949 | + grad_norm_pre_clip_avg=0.1344 | Metrics: + {'align_loss': 0.026605185121297836, + 'recon_loss': 0.1307111531496048, + 'predict_loss': 0.004093745723366737, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1948750615119934, + 'data_time': 0.0005744279769714922, + 'model_time': 1.1618485760118347, + 'grad_norm_pre_clip_avg': 0.13437264412641525, + 'learning_rate': 4.0441557343362316e-07, + 'epoch': 11.72} +04/20 [04:16:28] INFO | >> train_qwenlatent.py:487 + Step 46470 | grad_norm_pre_clip=0.1412 | + grad_norm_pre_clip_avg=0.1237 | Metrics: + {'align_loss': 0.02357439696788788, + 'recon_loss': 0.16795895993709564, + 'predict_loss': 0.006792768836021423, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14118334650993347, + 'data_time': 0.0006188269762787968, + 'model_time': 1.146112299989909, + 'grad_norm_pre_clip_avg': 0.12374706789851189, + 'learning_rate': 4.0228585501605046e-07, + 'epoch': 11.73} +04/20 [04:16:39] INFO | >> train_qwenlatent.py:487 + Step 46480 | grad_norm_pre_clip=0.1229 | + grad_norm_pre_clip_avg=0.1365 | Metrics: + {'align_loss': 0.025505391880869865, + 'recon_loss': 0.1091935783624649, + 'predict_loss': 0.004308885429054499, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12293398380279541, + 'data_time': 0.0005700259935110807, + 'model_time': 1.1400473960093223, + 'grad_norm_pre_clip_avg': 0.13650917932391166, + 'learning_rate': 4.0016203896973746e-07, + 'epoch': 11.73} +04/20 [04:16:51] INFO | >> train_qwenlatent.py:487 + Step 46490 | grad_norm_pre_clip=0.0913 | + grad_norm_pre_clip_avg=0.1183 | Metrics: + {'align_loss': 0.024383779615163803, + 'recon_loss': 0.11902724206447601, + 'predict_loss': 0.004786806181073189, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.0912536084651947, + 'data_time': 0.0006890619988553226, + 'model_time': 1.152053263009293, + 'grad_norm_pre_clip_avg': 0.11827205494046211, + 'learning_rate': 3.9804412632980384e-07, + 'epoch': 11.73} +04/20 [04:17:03] INFO | >> train_qwenlatent.py:487 + Step 46500 | grad_norm_pre_clip=0.1241 | + grad_norm_pre_clip_avg=0.1182 | Metrics: + {'align_loss': 0.026476163417100906, + 'recon_loss': 0.24788615107536316, + 'predict_loss': 0.012256554327905178, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12405367940664291, + 'mae_score': 0.005708742141723633, 'data_time': + 0.0005909519968554378, 'model_time': + 1.1473473540099803, 'grad_norm_pre_clip_avg': + 0.11824036985635758, 'learning_rate': + 3.9593211812849403e-07, 'epoch': 11.73} +04/20 [04:17:15] INFO | >> train_qwenlatent.py:487 + Step 46510 | grad_norm_pre_clip=0.0986 | + grad_norm_pre_clip_avg=0.1150 | Metrics: + {'align_loss': 0.024164831265807152, + 'recon_loss': 0.19182440638542175, + 'predict_loss': 0.005536345764994621, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09859392791986465, + 'data_time': 0.0005823639803566039, + 'model_time': 1.1422715520020574, + 'grad_norm_pre_clip_avg': 0.11498234197497367, + 'learning_rate': 3.93826015395177e-07, 'epoch': + 11.74} +04/20 [04:17:26] INFO | >> train_qwenlatent.py:487 + Step 46520 | grad_norm_pre_clip=0.1102 | + grad_norm_pre_clip_avg=0.1156 | Metrics: + {'align_loss': 0.02530810609459877, + 'recon_loss': 0.1981307864189148, + 'predict_loss': 0.0089790103957057, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11021708697080612, + 'data_time': 0.0005556310061365366, + 'model_time': 1.1623282530053984, + 'grad_norm_pre_clip_avg': 0.11563247293233872, + 'learning_rate': 3.9172581915634074e-07, + 'epoch': 11.74} +04/20 [04:17:38] INFO | >> train_qwenlatent.py:487 + Step 46530 | grad_norm_pre_clip=0.1056 | + grad_norm_pre_clip_avg=0.1137 | Metrics: + {'align_loss': 0.023096298798918724, + 'recon_loss': 0.08968106657266617, + 'predict_loss': 0.0034987684339284897, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10560066998004913, + 'data_time': 0.0006282420072238892, + 'model_time': 1.154200608987594, + 'grad_norm_pre_clip_avg': 0.11365046799182892, + 'learning_rate': 3.8963153043559513e-07, + 'epoch': 11.74} +04/20 [04:17:50] INFO | >> train_qwenlatent.py:487 + Step 46540 | grad_norm_pre_clip=0.0878 | + grad_norm_pre_clip_avg=0.1151 | Metrics: + {'align_loss': 0.024353064596652985, + 'recon_loss': 0.13199573755264282, + 'predict_loss': 0.00591771025210619, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.08777622878551483, + 'data_time': 0.0005957579996902496, + 'model_time': 1.1485178220027592, + 'grad_norm_pre_clip_avg': 0.11509334668517113, + 'learning_rate': 3.875431502536706e-07, + 'epoch': 11.74} +04/20 [04:18:02] INFO | >> train_qwenlatent.py:487 + Step 46550 | grad_norm_pre_clip=0.1625 | + grad_norm_pre_clip_avg=0.1313 | Metrics: + {'align_loss': 0.026092730462551117, + 'recon_loss': 0.187707781791687, + 'predict_loss': 0.007193592377007008, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16254307329654694, + 'mae_score': 0.005169842694256757, 'data_time': + 0.0006768799794372171, 'model_time': + 1.1625778450106736, 'grad_norm_pre_clip_avg': + 0.13128399103879929, 'learning_rate': + 3.854606796284209e-07, 'epoch': 11.75} +04/20 [04:18:13] INFO | >> train_qwenlatent.py:487 + Step 46560 | grad_norm_pre_clip=0.1819 | + grad_norm_pre_clip_avg=0.1298 | Metrics: + {'align_loss': 0.025407863780856133, + 'recon_loss': 0.2207653522491455, + 'predict_loss': 0.00766439363360405, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1819179505109787, + 'data_time': 0.0005579559947364032, + 'model_time': 1.1413335360120982, + 'grad_norm_pre_clip_avg': 0.12983284145593643, + 'learning_rate': 3.8338411957481296e-07, + 'epoch': 11.75} +04/20 [04:18:25] INFO | >> train_qwenlatent.py:487 + Step 46570 | grad_norm_pre_clip=0.1093 | + grad_norm_pre_clip_avg=0.1254 | Metrics: + {'align_loss': 0.02623153105378151, + 'recon_loss': 0.14258460700511932, + 'predict_loss': 0.007284975610673428, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10930333286523819, + 'data_time': 0.0005511500057764351, + 'model_time': 1.1402852609753609, + 'grad_norm_pre_clip_avg': 0.12540552988648415, + 'learning_rate': 3.8131347110493866e-07, + 'epoch': 11.75} +04/20 [04:18:37] INFO | >> train_qwenlatent.py:487 + Step 46580 | grad_norm_pre_clip=0.1058 | + grad_norm_pre_clip_avg=0.1294 | Metrics: + {'align_loss': 0.02657676301896572, + 'recon_loss': 0.1756807565689087, + 'predict_loss': 0.008768158964812756, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10575909912586212, + 'data_time': 0.0005378120113164186, + 'model_time': 1.1983444759971462, + 'grad_norm_pre_clip_avg': 0.1293651543557644, + 'learning_rate': 3.792487352280088e-07, + 'epoch': 11.75} +04/20 [04:18:49] INFO | >> train_qwenlatent.py:487 + Step 46590 | grad_norm_pre_clip=0.1574 | + grad_norm_pre_clip_avg=0.1484 | Metrics: + {'align_loss': 0.023540448397397995, + 'recon_loss': 0.13203556835651398, + 'predict_loss': 0.004827365279197693, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15744276344776154, + 'data_time': 0.0005711789999622852, + 'model_time': 1.3086509620188735, + 'grad_norm_pre_clip_avg': 0.14843077436089516, + 'learning_rate': 3.771899129503506e-07, + 'epoch': 11.76} +04/20 [04:19:01] INFO | >> train_qwenlatent.py:487 + Step 46600 | grad_norm_pre_clip=0.1137 | + grad_norm_pre_clip_avg=0.1386 | Metrics: + {'align_loss': 0.026516033336520195, + 'recon_loss': 0.26727840304374695, + 'predict_loss': 0.008217277005314827, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11368534713983536, + 'mae_score': 0.0050231276331721125, + 'data_time': 0.0005161270091775805, + 'model_time': 1.1305179919872899, + 'grad_norm_pre_clip_avg': 0.1386039562523365, + 'learning_rate': 3.7513700527540466e-07, + 'epoch': 11.76} +04/20 [04:19:12] INFO | >> train_qwenlatent.py:487 + Step 46610 | grad_norm_pre_clip=0.1038 | + grad_norm_pre_clip_avg=0.1329 | Metrics: + {'align_loss': 0.026052778586745262, + 'recon_loss': 0.14431573450565338, + 'predict_loss': 0.00507447961717844, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10381972789764404, + 'data_time': 0.00055016900296323, 'model_time': + 1.1392024350061547, 'grad_norm_pre_clip_avg': + 0.13290820568799971, 'learning_rate': + 3.730900132037364e-07, 'epoch': 11.76} +04/20 [04:19:24] INFO | >> train_qwenlatent.py:487 + Step 46620 | grad_norm_pre_clip=0.1418 | + grad_norm_pre_clip_avg=0.1293 | Metrics: + {'align_loss': 0.025107774883508682, + 'recon_loss': 0.14050976932048798, + 'predict_loss': 0.006120393052697182, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1418180763721466, + 'data_time': 0.00082345400005579, 'model_time': + 1.1495835599780548, 'grad_norm_pre_clip_avg': + 0.12931407019495963, 'learning_rate': + 3.7104893773302463e-07, 'epoch': 11.76} +04/20 [04:20:07] INFO | >> train_qwenlatent.py:487 + Step 46630 | grad_norm_pre_clip=0.1207 | + grad_norm_pre_clip_avg=0.1299 | Metrics: + {'align_loss': 0.025056319311261177, + 'recon_loss': 0.1485643833875656, + 'predict_loss': 0.008229951374232769, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1206795945763588, + 'data_time': 0.0011191420198883861, + 'model_time': 3.2471555410011206, + 'grad_norm_pre_clip_avg': 0.1298694431781769, + 'learning_rate': 3.690137798580645e-07, + 'epoch': 11.77} +04/20 [04:20:44] INFO | >> train_qwenlatent.py:487 + Step 46640 | grad_norm_pre_clip=0.1123 | + grad_norm_pre_clip_avg=0.1001 | Metrics: + {'align_loss': 0.02501615881919861, + 'recon_loss': 0.16200681030750275, + 'predict_loss': 0.005633545573800802, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11226311326026917, + 'data_time': 0.0021835319930687547, + 'model_time': 3.943845348985633, + 'grad_norm_pre_clip_avg': 0.1000780776143074, + 'learning_rate': 3.6698454057076757e-07, + 'epoch': 11.77} +04/20 [04:21:22] INFO | >> train_qwenlatent.py:487 + Step 46650 | grad_norm_pre_clip=0.1069 | + grad_norm_pre_clip_avg=0.1277 | Metrics: + {'align_loss': 0.025746427476406097, + 'recon_loss': 0.1596546471118927, + 'predict_loss': 0.00852021761238575, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10691653937101364, + 'mae_score': 0.005095868497281461, 'data_time': + 0.0013469380210153759, 'model_time': + 2.890742574003525, 'grad_norm_pre_clip_avg': + 0.12765264883637428, 'learning_rate': + 3.6496122086016147e-07, 'epoch': 11.77} +04/20 [04:21:58] INFO | >> train_qwenlatent.py:487 + Step 46660 | grad_norm_pre_clip=0.1072 | + grad_norm_pre_clip_avg=0.1218 | Metrics: + {'align_loss': 0.024836372584104538, + 'recon_loss': 0.22797410190105438, + 'predict_loss': 0.010625219903886318, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1071612685918808, + 'data_time': 0.0025891030090861022, + 'model_time': 3.8401679630042054, + 'grad_norm_pre_clip_avg': 0.12178857624530792, + 'learning_rate': 3.6294382171238333e-07, + 'epoch': 11.77} +04/20 [04:22:36] INFO | >> train_qwenlatent.py:487 + Step 46670 | grad_norm_pre_clip=0.1654 | + grad_norm_pre_clip_avg=0.1400 | Metrics: + {'align_loss': 0.024825267493724823, + 'recon_loss': 0.13664564490318298, + 'predict_loss': 0.008234064094722271, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1653585433959961, + 'data_time': 0.0011963210126850754, + 'model_time': 3.8294081120111514, + 'grad_norm_pre_clip_avg': 0.14002882614731788, + 'learning_rate': 3.609323441106937e-07, + 'epoch': 11.78} +04/20 [04:23:11] INFO | >> train_qwenlatent.py:487 + Step 46680 | grad_norm_pre_clip=0.1429 | + grad_norm_pre_clip_avg=0.1420 | Metrics: + {'align_loss': 0.026722216978669167, + 'recon_loss': 0.22303563356399536, + 'predict_loss': 0.006674899719655514, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1429285705089569, + 'data_time': 0.0011606699845287949, + 'model_time': 3.392876736994367, + 'grad_norm_pre_clip_avg': 0.14200564920902253, + 'learning_rate': 3.589267890354623e-07, + 'epoch': 11.78} +04/20 [04:23:43] INFO | >> train_qwenlatent.py:487 + Step 46690 | grad_norm_pre_clip=0.1503 | + grad_norm_pre_clip_avg=0.1398 | Metrics: + {'align_loss': 0.026440782472491264, + 'recon_loss': 0.1439819186925888, + 'predict_loss': 0.0067884172312915325, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15028609335422516, + 'data_time': 0.0016432709817308933, + 'model_time': 2.6189976389869116, + 'grad_norm_pre_clip_avg': 0.13975094705820085, + 'learning_rate': 3.569271574641695e-07, + 'epoch': 11.78} +04/20 [04:24:09] INFO | >> train_qwenlatent.py:487 + Step 46700 | grad_norm_pre_clip=0.1592 | + grad_norm_pre_clip_avg=0.1280 | Metrics: + {'align_loss': 0.025773346424102783, + 'recon_loss': 0.19956690073013306, + 'predict_loss': 0.00598460016772151, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15922240912914276, + 'mae_score': 0.005917519921655054, 'data_time': + 0.0017171879881061614, 'model_time': + 2.1237384699925315, 'grad_norm_pre_clip_avg': + 0.1280277319252491, 'learning_rate': + 3.549334503714109e-07, 'epoch': 11.78} +04/20 [04:24:27] INFO | >> train_qwenlatent.py:487 + Step 46710 | grad_norm_pre_clip=0.0971 | + grad_norm_pre_clip_avg=0.1181 | Metrics: + {'align_loss': 0.025614213198423386, + 'recon_loss': 0.15193429589271545, + 'predict_loss': 0.004325011279433966, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09707508981227875, + 'data_time': 0.0013271619973238558, + 'model_time': 1.4491187240055297, + 'grad_norm_pre_clip_avg': 0.11813987866044044, + 'learning_rate': 3.529456687288981e-07, + 'epoch': 11.79} +04/20 [04:24:40] INFO | >> train_qwenlatent.py:487 + Step 46720 | grad_norm_pre_clip=0.1434 | + grad_norm_pre_clip_avg=0.1190 | Metrics: + {'align_loss': 0.02452736347913742, + 'recon_loss': 0.09230758249759674, + 'predict_loss': 0.0024758847430348396, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14338219165802002, + 'data_time': 0.0008435449854005128, + 'model_time': 1.2081501100037713, + 'grad_norm_pre_clip_avg': 0.11900204941630363, + 'learning_rate': 3.509638135054511e-07, + 'epoch': 11.79} +04/20 [04:24:53] INFO | >> train_qwenlatent.py:487 + Step 46730 | grad_norm_pre_clip=0.1112 | + grad_norm_pre_clip_avg=0.1411 | Metrics: + {'align_loss': 0.02542955055832863, + 'recon_loss': 0.15732330083847046, + 'predict_loss': 0.009112457744777203, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11115623265504837, + 'data_time': 0.0005457779916469008, + 'model_time': 1.2268182520056143, + 'grad_norm_pre_clip_avg': 0.14110227599740027, + 'learning_rate': 3.489878856670002e-07, + 'epoch': 11.79} +04/20 [04:25:05] INFO | >> train_qwenlatent.py:487 + Step 46740 | grad_norm_pre_clip=0.1011 | + grad_norm_pre_clip_avg=0.1355 | Metrics: + {'align_loss': 0.026160791516304016, + 'recon_loss': 0.12432000786066055, + 'predict_loss': 0.0033897182438522577, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1011003628373146, + 'data_time': 0.0005715829902328551, + 'model_time': 1.264677014987683, + 'grad_norm_pre_clip_avg': 0.1354951709508896, + 'learning_rate': 3.470178861765907e-07, + 'epoch': 11.79} +04/20 [04:25:19] INFO | >> train_qwenlatent.py:487 + Step 46750 | grad_norm_pre_clip=0.1283 | + grad_norm_pre_clip_avg=0.1319 | Metrics: + {'align_loss': 0.026681024581193924, + 'recon_loss': 0.17198437452316284, + 'predict_loss': 0.0052741593681275845, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1283184140920639, + 'mae_score': 0.005400079435056394, 'data_time': + 0.0008379659848287702, 'model_time': + 1.2487717370095197, 'grad_norm_pre_clip_avg': + 0.13192463144659997, 'learning_rate': + 3.450538159943763e-07, 'epoch': 11.8} +04/20 [04:25:31] INFO | >> train_qwenlatent.py:487 + Step 46760 | grad_norm_pre_clip=0.1890 | + grad_norm_pre_clip_avg=0.1431 | Metrics: + {'align_loss': 0.02615289017558098, + 'recon_loss': 0.15628206729888916, + 'predict_loss': 0.007697638124227524, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1889709234237671, + 'data_time': 0.0006710239977110177, + 'model_time': 1.1984151520009618, + 'grad_norm_pre_clip_avg': 0.14314472079277038, + 'learning_rate': 3.430956760776168e-07, + 'epoch': 11.8} +04/20 [04:25:43] INFO | >> train_qwenlatent.py:487 + Step 46770 | grad_norm_pre_clip=0.1780 | + grad_norm_pre_clip_avg=0.1438 | Metrics: + {'align_loss': 0.024942636489868164, + 'recon_loss': 0.14988793432712555, + 'predict_loss': 0.006784557830542326, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17798075079917908, + 'data_time': 0.0006331310141831636, + 'model_time': 1.1861729499942157, + 'grad_norm_pre_clip_avg': 0.14383305758237838, + 'learning_rate': 3.4114346738069274e-07, + 'epoch': 11.8} +04/20 [04:25:56] INFO | >> train_qwenlatent.py:487 + Step 46780 | grad_norm_pre_clip=0.0888 | + grad_norm_pre_clip_avg=0.1222 | Metrics: + {'align_loss': 0.02421862632036209, + 'recon_loss': 0.09244693070650101, + 'predict_loss': 0.0028335938695818186, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.08878178894519806, + 'data_time': 0.001133111014496535, + 'model_time': 1.253171554999426, + 'grad_norm_pre_clip_avg': 0.12216302454471588, + 'learning_rate': 3.3919719085508293e-07, + 'epoch': 11.8} +04/20 [04:26:09] INFO | >> train_qwenlatent.py:487 + Step 46790 | grad_norm_pre_clip=0.1503 | + grad_norm_pre_clip_avg=0.1357 | Metrics: + {'align_loss': 0.026190316304564476, + 'recon_loss': 0.11241941899061203, + 'predict_loss': 0.004557305946946144, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15034206211566925, + 'data_time': 0.0006915519770700485, + 'model_time': 1.2088748070236761, + 'grad_norm_pre_clip_avg': 0.13566701263189315, + 'learning_rate': 3.372568474493781e-07, + 'epoch': 11.81} +04/20 [04:26:22] INFO | >> train_qwenlatent.py:487 + Step 46800 | grad_norm_pre_clip=0.1513 | + grad_norm_pre_clip_avg=0.1308 | Metrics: + {'align_loss': 0.025013331323862076, + 'recon_loss': 0.1637943983078003, + 'predict_loss': 0.010883190669119358, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15129299461841583, + 'mae_score': 0.005046995695646819, 'data_time': + 0.0009869560017250478, 'model_time': + 1.2751562870107591, 'grad_norm_pre_clip_avg': + 0.13077090084552764, 'learning_rate': + 3.353224381092788e-07, 'epoch': 11.81} +04/20 [04:26:35] INFO | >> train_qwenlatent.py:487 + Step 46810 | grad_norm_pre_clip=0.1360 | + grad_norm_pre_clip_avg=0.1236 | Metrics: + {'align_loss': 0.025431061163544655, + 'recon_loss': 0.16603490710258484, + 'predict_loss': 0.00705896457657218, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13601438701152802, + 'data_time': 0.0006129099929239601, + 'model_time': 1.243971734016668, + 'grad_norm_pre_clip_avg': 0.12361808866262436, + 'learning_rate': 3.3339396377759443e-07, + 'epoch': 11.81} +04/20 [04:26:47] INFO | >> train_qwenlatent.py:487 + Step 46820 | grad_norm_pre_clip=0.1487 | + grad_norm_pre_clip_avg=0.1203 | Metrics: + {'align_loss': 0.023351814597845078, + 'recon_loss': 0.12164981663227081, + 'predict_loss': 0.006575300358235836, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14866317808628082, + 'data_time': 0.0006023390160407871, + 'model_time': 1.2008663160086144, + 'grad_norm_pre_clip_avg': 0.12034692838788033, + 'learning_rate': 3.3147142539423864e-07, + 'epoch': 11.81} +04/20 [04:27:00] INFO | >> train_qwenlatent.py:487 + Step 46830 | grad_norm_pre_clip=0.1304 | + grad_norm_pre_clip_avg=0.1225 | Metrics: + {'align_loss': 0.02570059895515442, + 'recon_loss': 0.2157677710056305, + 'predict_loss': 0.01194947212934494, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13039720058441162, + 'data_time': 0.000964326987741515, + 'model_time': 1.300498165015597, + 'grad_norm_pre_clip_avg': 0.1225373275578022, + 'learning_rate': 3.295548238962328e-07, + 'epoch': 11.82} +04/20 [04:27:13] INFO | >> train_qwenlatent.py:487 + Step 46840 | grad_norm_pre_clip=0.0643 | + grad_norm_pre_clip_avg=0.1151 | Metrics: + {'align_loss': 0.026029424741864204, + 'recon_loss': 0.10296301543712616, + 'predict_loss': 0.004564191680401564, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.06426011025905609, + 'data_time': 0.0008397299970965832, + 'model_time': 1.2651034320006147, + 'grad_norm_pre_clip_avg': 0.1151201032102108, + 'learning_rate': 3.276441602177064e-07, + 'epoch': 11.82} +04/20 [04:27:26] INFO | >> train_qwenlatent.py:487 + Step 46850 | grad_norm_pre_clip=0.1192 | + grad_norm_pre_clip_avg=0.1184 | Metrics: + {'align_loss': 0.025370076298713684, + 'recon_loss': 0.185823455452919, + 'predict_loss': 0.007690002676099539, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11920100450515747, + 'mae_score': 0.004951544495316239, 'data_time': + 0.001028707978548482, 'model_time': + 1.268694308004342, 'grad_norm_pre_clip_avg': + 0.11839251443743706, 'learning_rate': + 3.2573943528989143e-07, 'epoch': 11.82} +04/20 [04:27:39] INFO | >> train_qwenlatent.py:487 + Step 46860 | grad_norm_pre_clip=0.1288 | + grad_norm_pre_clip_avg=0.1133 | Metrics: + {'align_loss': 0.024941571056842804, + 'recon_loss': 0.12079066783189774, + 'predict_loss': 0.0033616293221712112, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12879708409309387, + 'data_time': 0.0005940410192124546, + 'model_time': 1.515290150011424, + 'grad_norm_pre_clip_avg': 0.11334667652845383, + 'learning_rate': 3.2384065004113186e-07, + 'epoch': 11.82} +04/20 [04:27:51] INFO | >> train_qwenlatent.py:487 + Step 46870 | grad_norm_pre_clip=0.1052 | + grad_norm_pre_clip_avg=0.1330 | Metrics: + {'align_loss': 0.02659342810511589, + 'recon_loss': 0.19911623001098633, + 'predict_loss': 0.00787365809082985, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1051948219537735, + 'data_time': 0.001339131995337084, + 'model_time': 1.2805191340157762, + 'grad_norm_pre_clip_avg': 0.13298966661095618, + 'learning_rate': 3.219478053968686e-07, + 'epoch': 11.83} +04/20 [04:28:04] INFO | >> train_qwenlatent.py:487 + Step 46880 | grad_norm_pre_clip=0.1357 | + grad_norm_pre_clip_avg=0.1354 | Metrics: + {'align_loss': 0.026158886030316353, + 'recon_loss': 0.1934518665075302, + 'predict_loss': 0.004273808095604181, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13568760454654694, + 'data_time': 0.0007698479748796672, + 'model_time': 1.1686542209936306, + 'grad_norm_pre_clip_avg': 0.1353574275970459, + 'learning_rate': 3.200609022796521e-07, + 'epoch': 11.83} +04/20 [04:28:16] INFO | >> train_qwenlatent.py:487 + Step 46890 | grad_norm_pre_clip=0.1193 | + grad_norm_pre_clip_avg=0.1108 | Metrics: + {'align_loss': 0.023474063724279404, + 'recon_loss': 0.16383355855941772, + 'predict_loss': 0.004169018939137459, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11932963132858276, + 'data_time': 0.0006062050233595073, + 'model_time': 1.2086376259976532, + 'grad_norm_pre_clip_avg': 0.1107694886624813, + 'learning_rate': 3.1817994160913496e-07, + 'epoch': 11.83} +04/20 [04:28:29] INFO | >> train_qwenlatent.py:487 + Step 46900 | grad_norm_pre_clip=0.1564 | + grad_norm_pre_clip_avg=0.1292 | Metrics: + {'align_loss': 0.025564003735780716, + 'recon_loss': 0.13502255082130432, + 'predict_loss': 0.0058540781028568745, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1563783586025238, + 'mae_score': 0.00516384270814088, 'data_time': + 0.0010135709890164435, 'model_time': + 1.2053390679939184, 'grad_norm_pre_clip_avg': + 0.1292333148419857, 'learning_rate': + 3.163049243020781e-07, 'epoch': 11.83} +04/20 [04:28:42] INFO | >> train_qwenlatent.py:487 + Step 46910 | grad_norm_pre_clip=0.0925 | + grad_norm_pre_clip_avg=0.1249 | Metrics: + {'align_loss': 0.026153292506933212, + 'recon_loss': 0.12266922742128372, + 'predict_loss': 0.0030699714552611113, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09248007833957672, + 'data_time': 0.0006630670104641467, + 'model_time': 1.2477258220023941, + 'grad_norm_pre_clip_avg': 0.12487977370619774, + 'learning_rate': 3.1443585127234045e-07, + 'epoch': 11.84} +04/20 [04:28:54] INFO | >> train_qwenlatent.py:487 + Step 46920 | grad_norm_pre_clip=0.1081 | + grad_norm_pre_clip_avg=0.1307 | Metrics: + {'align_loss': 0.02608112245798111, + 'recon_loss': 0.18682752549648285, + 'predict_loss': 0.00944602582603693, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10808966308832169, + 'data_time': 0.000980615004664287, + 'model_time': 1.2294286579999607, + 'grad_norm_pre_clip_avg': 0.1307428903877735, + 'learning_rate': 3.1257272343088357e-07, + 'epoch': 11.84} +04/20 [04:29:06] INFO | >> train_qwenlatent.py:487 + Step 46930 | grad_norm_pre_clip=0.1096 | + grad_norm_pre_clip_avg=0.1153 | Metrics: + {'align_loss': 0.025579936802387238, + 'recon_loss': 0.17648062109947205, + 'predict_loss': 0.0065724775195121765, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1096300482749939, + 'data_time': 0.0008283450151793659, + 'model_time': 1.2208662340126466, + 'grad_norm_pre_clip_avg': 0.11533854082226754, + 'learning_rate': 3.107155416857768e-07, + 'epoch': 11.84} +04/20 [04:29:18] INFO | >> train_qwenlatent.py:487 + Step 46940 | grad_norm_pre_clip=0.1916 | + grad_norm_pre_clip_avg=0.1287 | Metrics: + {'align_loss': 0.02415960654616356, + 'recon_loss': 0.15946424007415771, + 'predict_loss': 0.006582637317478657, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.19164855778217316, + 'data_time': 0.0007017129973974079, + 'model_time': 1.2064289289992303, + 'grad_norm_pre_clip_avg': 0.12867192029953003, + 'learning_rate': 3.0886430694218395e-07, + 'epoch': 11.84} +04/20 [04:29:32] INFO | >> train_qwenlatent.py:487 + Step 46950 | grad_norm_pre_clip=0.1497 | + grad_norm_pre_clip_avg=0.1154 | Metrics: + {'align_loss': 0.025346674025058746, + 'recon_loss': 0.1278473138809204, + 'predict_loss': 0.007669913116842508, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14972718060016632, + 'mae_score': 0.00570493002195616, 'data_time': + 0.001003485987894237, 'model_time': + 1.2192007979901973, 'grad_norm_pre_clip_avg': + 0.1153857834637165, 'learning_rate': + 3.070190201023777e-07, 'epoch': 11.85} +04/20 [04:29:45] INFO | >> train_qwenlatent.py:487 + Step 46960 | grad_norm_pre_clip=0.1144 | + grad_norm_pre_clip_avg=0.1035 | Metrics: + {'align_loss': 0.025795230641961098, + 'recon_loss': 0.11516520380973816, + 'predict_loss': 0.004185436759144068, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11444974690675735, + 'data_time': 0.0011646159982774407, + 'model_time': 1.2539280419878196, + 'grad_norm_pre_clip_avg': 0.10352695137262344, + 'learning_rate': 3.0517968206572827e-07, + 'epoch': 11.85} +04/20 [04:29:57] INFO | >> train_qwenlatent.py:487 + Step 46970 | grad_norm_pre_clip=0.1083 | + grad_norm_pre_clip_avg=0.1291 | Metrics: + {'align_loss': 0.02405565045773983, + 'recon_loss': 0.19497162103652954, + 'predict_loss': 0.006062449421733618, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10825543850660324, + 'data_time': 0.0008037200022954494, + 'model_time': 1.233063672989374, + 'grad_norm_pre_clip_avg': 0.1290952295064926, + 'learning_rate': 3.033462937287034e-07, + 'epoch': 11.85} +04/20 [04:30:10] INFO | >> train_qwenlatent.py:487 + Step 46980 | grad_norm_pre_clip=0.1154 | + grad_norm_pre_clip_avg=0.1189 | Metrics: + {'align_loss': 0.025578930974006653, + 'recon_loss': 0.1249856948852539, + 'predict_loss': 0.00553200813010335, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11539165675640106, + 'data_time': 0.0008951679919846356, + 'model_time': 1.205816367000807, + 'grad_norm_pre_clip_avg': 0.11890438124537468, + 'learning_rate': 3.0151885598487823e-07, + 'epoch': 11.85} +04/20 [04:30:23] INFO | >> train_qwenlatent.py:487 + Step 46990 | grad_norm_pre_clip=0.1198 | + grad_norm_pre_clip_avg=0.1150 | Metrics: + {'align_loss': 0.0248250775039196, + 'recon_loss': 0.13413465023040771, + 'predict_loss': 0.005422307178378105, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11976340413093567, + 'data_time': 0.0006565400108229369, + 'model_time': 1.2489189250045456, + 'grad_norm_pre_clip_avg': 0.11502023786306381, + 'learning_rate': 2.9969736972491997e-07, + 'epoch': 11.86} +04/20 [04:30:36] INFO | >> train_qwenlatent.py:487 + Step 47000 | grad_norm_pre_clip=0.1123 | + grad_norm_pre_clip_avg=0.1301 | Metrics: + {'align_loss': 0.02603825554251671, + 'recon_loss': 0.14871062338352203, + 'predict_loss': 0.004254116676747799, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11233161389827728, + 'mae_score': 0.006375797374828442, 'data_time': + 0.0008659360173624009, 'model_time': + 1.2216919680067804, 'grad_norm_pre_clip_avg': + 0.1300599157810211, 'learning_rate': + 2.978818358366014e-07, 'epoch': 11.86} +04/20 [04:30:49] INFO | >> train_qwenlatent.py:487 + Step 47010 | grad_norm_pre_clip=0.1199 | + grad_norm_pre_clip_avg=0.1239 | Metrics: + {'align_loss': 0.025873590260744095, + 'recon_loss': 0.11113391071557999, + 'predict_loss': 0.0040563298389315605, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1199110820889473, + 'data_time': 0.0006272040191106498, + 'model_time': 1.235826429008739, + 'grad_norm_pre_clip_avg': 0.12394043877720833, + 'learning_rate': 2.960722552047923e-07, + 'epoch': 11.86} +04/20 [04:31:01] INFO | >> train_qwenlatent.py:487 + Step 47020 | grad_norm_pre_clip=0.1326 | + grad_norm_pre_clip_avg=0.1430 | Metrics: + {'align_loss': 0.0252878125756979, + 'recon_loss': 0.13926708698272705, + 'predict_loss': 0.007741576060652733, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13262589275836945, + 'data_time': 0.0009667660051491112, + 'model_time': 1.250644606014248, + 'grad_norm_pre_clip_avg': 0.14297728464007378, + 'learning_rate': 2.942686287114589e-07, + 'epoch': 11.86} +04/20 [04:31:14] INFO | >> train_qwenlatent.py:487 + Step 47030 | grad_norm_pre_clip=0.1300 | + grad_norm_pre_clip_avg=0.1206 | Metrics: + {'align_loss': 0.02592598833143711, + 'recon_loss': 0.17356982827186584, + 'predict_loss': 0.0076509518548846245, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12995098531246185, + 'data_time': 0.0006087360088713467, + 'model_time': 1.209547845995985, + 'grad_norm_pre_clip_avg': 0.12061542719602585, + 'learning_rate': 2.9247095723566754e-07, + 'epoch': 11.87} +04/20 [04:31:26] INFO | >> train_qwenlatent.py:487 + Step 47040 | grad_norm_pre_clip=0.1172 | + grad_norm_pre_clip_avg=0.1194 | Metrics: + {'align_loss': 0.025747425854206085, + 'recon_loss': 0.14953762292861938, + 'predict_loss': 0.00675543025135994, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11722099035978317, + 'data_time': 0.0008502229757141322, + 'model_time': 1.2299896459735464, + 'grad_norm_pre_clip_avg': 0.11939686685800552, + 'learning_rate': 2.906792416535799e-07, + 'epoch': 11.87} +04/20 [04:31:40] INFO | >> train_qwenlatent.py:487 + Step 47050 | grad_norm_pre_clip=0.1366 | + grad_norm_pre_clip_avg=0.1203 | Metrics: + {'align_loss': 0.02611592598259449, + 'recon_loss': 0.17415118217468262, + 'predict_loss': 0.0052634053863584995, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13657398521900177, + 'mae_score': 0.005881780332273191, 'data_time': + 0.0006124909850768745, 'model_time': + 1.2469578070158605, 'grad_norm_pre_clip_avg': + 0.12032118886709213, 'learning_rate': + 2.888934828384586e-07, 'epoch': 11.87} +04/20 [04:31:52] INFO | >> train_qwenlatent.py:487 + Step 47060 | grad_norm_pre_clip=0.0927 | + grad_norm_pre_clip_avg=0.1179 | Metrics: + {'align_loss': 0.025631412863731384, + 'recon_loss': 0.12041056901216507, + 'predict_loss': 0.004459107294678688, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09268008917570114, + 'data_time': 0.0010603489936329424, + 'model_time': 1.2225931869761553, + 'grad_norm_pre_clip_avg': 0.11785098612308502, + 'learning_rate': 2.871136816606592e-07, + 'epoch': 11.87} +04/20 [04:32:05] INFO | >> train_qwenlatent.py:487 + Step 47070 | grad_norm_pre_clip=0.0889 | + grad_norm_pre_clip_avg=0.1201 | Metrics: + {'align_loss': 0.024602010846138, 'recon_loss': + 0.15578484535217285, 'predict_loss': + 0.005806629080325365, 'aux_loss_decay_weight': + 0.0, 'grad_norm_pre_clip': 0.08890126645565033, + 'data_time': 0.0013355309783946723, + 'model_time': 1.3206294639967382, + 'grad_norm_pre_clip_avg': 0.12007777616381646, + 'learning_rate': 2.8533983898763685e-07, + 'epoch': 11.88} +04/20 [04:32:18] INFO | >> train_qwenlatent.py:487 + Step 47080 | grad_norm_pre_clip=0.1424 | + grad_norm_pre_clip_avg=0.1287 | Metrics: + {'align_loss': 0.024904925376176834, + 'recon_loss': 0.08614397794008255, + 'predict_loss': 0.0028447825461626053, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1423923820257187, + 'data_time': 0.0007851190166547894, + 'model_time': 1.2602297959965654, + 'grad_norm_pre_clip_avg': 0.12872979044914246, + 'learning_rate': 2.83571955683938e-07, 'epoch': + 11.88} +04/20 [04:32:30] INFO | >> train_qwenlatent.py:487 + Step 47090 | grad_norm_pre_clip=0.1104 | + grad_norm_pre_clip_avg=0.1258 | Metrics: + {'align_loss': 0.024571694433689117, + 'recon_loss': 0.10483506321907043, + 'predict_loss': 0.003154016798362136, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11044404655694962, + 'data_time': 0.000609242997597903, + 'model_time': 1.2474883669929113, + 'grad_norm_pre_clip_avg': 0.12579174861311912, + 'learning_rate': 2.8181003261120883e-07, + 'epoch': 11.88} +04/20 [04:32:44] INFO | >> train_qwenlatent.py:487 + Step 47100 | grad_norm_pre_clip=0.1028 | + grad_norm_pre_clip_avg=0.1137 | Metrics: + {'align_loss': 0.02521078661084175, + 'recon_loss': 0.16623437404632568, + 'predict_loss': 0.007206744980067015, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10280995070934296, + 'mae_score': 0.005505511137816283, 'data_time': + 0.0009743289847392589, 'model_time': + 1.263907640997786, 'grad_norm_pre_clip_avg': + 0.1136599101126194, 'learning_rate': + 2.8005407062819115e-07, 'epoch': 11.88} +04/20 [04:32:56] INFO | >> train_qwenlatent.py:487 + Step 47110 | grad_norm_pre_clip=0.1018 | + grad_norm_pre_clip_avg=0.1059 | Metrics: + {'align_loss': 0.0240631066262722, + 'recon_loss': 0.13549882173538208, + 'predict_loss': 0.0054732742719352245, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10176026821136475, + 'data_time': 0.0009810440242290497, + 'model_time': 1.2900323750218377, + 'grad_norm_pre_clip_avg': 0.10586863458156585, + 'learning_rate': 2.783040705907178e-07, + 'epoch': 11.89} +04/20 [04:33:09] INFO | >> train_qwenlatent.py:487 + Step 47120 | grad_norm_pre_clip=0.0978 | + grad_norm_pre_clip_avg=0.1151 | Metrics: + {'align_loss': 0.024699825793504715, + 'recon_loss': 0.19447815418243408, + 'predict_loss': 0.0076071214862167835, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09779263287782669, + 'data_time': 0.0008652279793750495, + 'model_time': 1.516968773008557, + 'grad_norm_pre_clip_avg': 0.1151400275528431, + 'learning_rate': 2.765600333517173e-07, + 'epoch': 11.89} +04/20 [04:33:22] INFO | >> train_qwenlatent.py:487 + Step 47130 | grad_norm_pre_clip=0.1628 | + grad_norm_pre_clip_avg=0.1352 | Metrics: + {'align_loss': 0.02696923539042473, + 'recon_loss': 0.1398368775844574, + 'predict_loss': 0.007758424151688814, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16283954679965973, + 'data_time': 0.0006524820055346936, + 'model_time': 1.2418082710064482, + 'grad_norm_pre_clip_avg': 0.13516199067234994, + 'learning_rate': 2.7482195976121256e-07, + 'epoch': 11.89} +04/20 [04:33:35] INFO | >> train_qwenlatent.py:487 + Step 47140 | grad_norm_pre_clip=0.1240 | + grad_norm_pre_clip_avg=0.1398 | Metrics: + {'align_loss': 0.025939421728253365, + 'recon_loss': 0.15125174820423126, + 'predict_loss': 0.005988966673612595, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12397506088018417, + 'data_time': 0.0008413900213781744, + 'model_time': 1.2136570219881833, + 'grad_norm_pre_clip_avg': 0.13977659717202187, + 'learning_rate': 2.730898506663201e-07, + 'epoch': 11.9} +04/20 [04:33:48] INFO | >> train_qwenlatent.py:487 + Step 47150 | grad_norm_pre_clip=0.1215 | + grad_norm_pre_clip_avg=0.1343 | Metrics: + {'align_loss': 0.025458991527557373, + 'recon_loss': 0.16457495093345642, + 'predict_loss': 0.005591350141912699, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12145547568798065, + 'mae_score': 0.005232218579129056, 'data_time': + 0.0006501269817817956, 'model_time': + 1.2128333169966936, 'grad_norm_pre_clip_avg': + 0.13434339538216591, 'learning_rate': + 2.713637069112496e-07, 'epoch': 11.9} +04/20 [04:34:00] INFO | >> train_qwenlatent.py:487 + Step 47160 | grad_norm_pre_clip=0.1085 | + grad_norm_pre_clip_avg=0.1310 | Metrics: + {'align_loss': 0.02573450468480587, + 'recon_loss': 0.13484491407871246, + 'predict_loss': 0.00571414502337575, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10846320539712906, + 'data_time': 0.0007331410015467554, + 'model_time': 1.305415266979253, + 'grad_norm_pre_clip_avg': 0.13103759288787842, + 'learning_rate': 2.696435293372991e-07, + 'epoch': 11.9} +04/20 [04:34:13] INFO | >> train_qwenlatent.py:487 + Step 47170 | grad_norm_pre_clip=0.1052 | + grad_norm_pre_clip_avg=0.1249 | Metrics: + {'align_loss': 0.024597443640232086, + 'recon_loss': 0.10984846949577332, + 'predict_loss': 0.004589726682752371, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10515221953392029, + 'data_time': 0.0006738050142303109, + 'model_time': 1.2339256200066302, + 'grad_norm_pre_clip_avg': 0.12492487877607346, + 'learning_rate': 2.679293187828665e-07, + 'epoch': 11.9} +04/20 [04:34:26] INFO | >> train_qwenlatent.py:487 + Step 47180 | grad_norm_pre_clip=0.1162 | + grad_norm_pre_clip_avg=0.1359 | Metrics: + {'align_loss': 0.024836326017975807, + 'recon_loss': 0.15726597607135773, + 'predict_loss': 0.004393597133457661, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11620663106441498, + 'data_time': 0.0006638430058956146, + 'model_time': 1.2648565679846797, + 'grad_norm_pre_clip_avg': 0.13588626608252524, + 'learning_rate': 2.662210760834324e-07, + 'epoch': 11.91} +04/20 [04:34:38] INFO | >> train_qwenlatent.py:487 + Step 47190 | grad_norm_pre_clip=0.1049 | + grad_norm_pre_clip_avg=0.1077 | Metrics: + {'align_loss': 0.025268789380788803, + 'recon_loss': 0.16752228140830994, + 'predict_loss': 0.006389504764229059, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10489962995052338, + 'data_time': 0.0008543839794583619, + 'model_time': 1.2797890010115225, + 'grad_norm_pre_clip_avg': 0.1077376738190651, + 'learning_rate': 2.6451880207157893e-07, + 'epoch': 11.91} +04/20 [04:34:51] INFO | >> train_qwenlatent.py:487 + Step 47200 | grad_norm_pre_clip=0.1405 | + grad_norm_pre_clip_avg=0.1290 | Metrics: + {'align_loss': 0.024858420714735985, + 'recon_loss': 0.15667594969272614, + 'predict_loss': 0.0050653126090765, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14053790271282196, + 'mae_score': 0.005655873358786643, 'data_time': + 0.0010777899879030883, 'model_time': + 1.2542439990211278, 'grad_norm_pre_clip_avg': + 0.12895046249032022, 'learning_rate': + 2.6282249757697076e-07, 'epoch': 11.91} +04/20 [04:35:04] INFO | >> train_qwenlatent.py:487 + Step 47210 | grad_norm_pre_clip=0.1566 | + grad_norm_pre_clip_avg=0.1244 | Metrics: + {'align_loss': 0.025698890909552574, + 'recon_loss': 0.13579265773296356, + 'predict_loss': 0.00933549739420414, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15660406649112701, + 'data_time': 0.0008613640093244612, + 'model_time': 1.2207716080010869, + 'grad_norm_pre_clip_avg': 0.12442216873168946, + 'learning_rate': 2.61132163426367e-07, 'epoch': + 11.91} +04/20 [04:35:16] INFO | >> train_qwenlatent.py:487 + Step 47220 | grad_norm_pre_clip=0.0725 | + grad_norm_pre_clip_avg=0.1134 | Metrics: + {'align_loss': 0.026380935683846474, + 'recon_loss': 0.14889095723628998, + 'predict_loss': 0.004984940867871046, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.07246452569961548, + 'data_time': 0.001072915008990094, + 'model_time': 1.3014684459776618, + 'grad_norm_pre_clip_avg': 0.11340998336672783, + 'learning_rate': 2.5944780044361384e-07, + 'epoch': 11.92} +04/20 [04:35:29] INFO | >> train_qwenlatent.py:487 + Step 47230 | grad_norm_pre_clip=0.1247 | + grad_norm_pre_clip_avg=0.1251 | Metrics: + {'align_loss': 0.026318300515413284, + 'recon_loss': 0.13089503347873688, + 'predict_loss': 0.006099970079958439, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12467987835407257, + 'data_time': 0.0009933410037774593, + 'model_time': 1.2247041040100157, + 'grad_norm_pre_clip_avg': 0.12509979233145713, + 'learning_rate': 2.5776940944965166e-07, + 'epoch': 11.92} +04/20 [04:35:41] INFO | >> train_qwenlatent.py:487 + Step 47240 | grad_norm_pre_clip=0.1719 | + grad_norm_pre_clip_avg=0.1226 | Metrics: + {'align_loss': 0.026048175990581512, + 'recon_loss': 0.14901700615882874, + 'predict_loss': 0.004253178369253874, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1718788594007492, + 'data_time': 0.0010286189790349454, + 'model_time': 1.3048154340067413, + 'grad_norm_pre_clip_avg': 0.12262854874134063, + 'learning_rate': 2.560969912625066e-07, + 'epoch': 11.92} +04/20 [04:35:55] INFO | >> train_qwenlatent.py:487 + Step 47250 | grad_norm_pre_clip=0.1829 | + grad_norm_pre_clip_avg=0.1399 | Metrics: + {'align_loss': 0.026028091087937355, + 'recon_loss': 0.19918425381183624, + 'predict_loss': 0.010956794023513794, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18288403749465942, + 'mae_score': 0.004817327722772821, 'data_time': + 0.0006477739952970296, 'model_time': + 1.238346003985498, 'grad_norm_pre_clip_avg': + 0.139883141964674, 'learning_rate': + 2.5443054669729616e-07, 'epoch': 11.92} +04/20 [04:36:08] INFO | >> train_qwenlatent.py:487 + Step 47260 | grad_norm_pre_clip=0.2316 | + grad_norm_pre_clip_avg=0.1394 | Metrics: + {'align_loss': 0.02414870262145996, + 'recon_loss': 0.12558424472808838, + 'predict_loss': 0.0030443361029028893, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.2316255122423172, + 'data_time': 0.0008734319999348372, + 'model_time': 1.2515028310008347, + 'grad_norm_pre_clip_avg': 0.1394440196454525, + 'learning_rate': 2.5277007656622497e-07, + 'epoch': 11.93} +04/20 [04:36:21] INFO | >> train_qwenlatent.py:487 + Step 47270 | grad_norm_pre_clip=0.1189 | + grad_norm_pre_clip_avg=0.1356 | Metrics: + {'align_loss': 0.026208963245153427, + 'recon_loss': 0.1617913693189621, + 'predict_loss': 0.008379929699003696, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11885417997837067, + 'data_time': 0.0006233820167835802, + 'model_time': 1.21632765798131, + 'grad_norm_pre_clip_avg': 0.13561159893870353, + 'learning_rate': 2.511155816785863e-07, + 'epoch': 11.93} +04/20 [04:36:33] INFO | >> train_qwenlatent.py:487 + Step 47280 | grad_norm_pre_clip=0.1171 | + grad_norm_pre_clip_avg=0.1224 | Metrics: + {'align_loss': 0.025357410311698914, + 'recon_loss': 0.10986629128456116, + 'predict_loss': 0.0032416696194559336, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11712183803319931, + 'data_time': 0.0010951620060950518, + 'model_time': 1.2735141390003264, + 'grad_norm_pre_clip_avg': 0.122373116761446, + 'learning_rate': 2.49467062840758e-07, 'epoch': + 11.93} +04/20 [04:36:46] INFO | >> train_qwenlatent.py:487 + Step 47290 | grad_norm_pre_clip=0.1320 | + grad_norm_pre_clip_avg=0.1242 | Metrics: + {'align_loss': 0.026255592703819275, + 'recon_loss': 0.15910443663597107, + 'predict_loss': 0.010846426710486412, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13202199339866638, + 'data_time': 0.0007231910130940378, + 'model_time': 1.2694624100113288, + 'grad_norm_pre_clip_avg': 0.12421535849571227, + 'learning_rate': 2.4782452085621033e-07, + 'epoch': 11.93} +04/20 [04:36:59] INFO | >> train_qwenlatent.py:487 + Step 47300 | grad_norm_pre_clip=0.1535 | + grad_norm_pre_clip_avg=0.1377 | Metrics: + {'align_loss': 0.02445600926876068, + 'recon_loss': 0.12313980609178543, + 'predict_loss': 0.005903366021811962, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15352791547775269, + 'mae_score': 0.004431402790653813, 'data_time': + 0.0007993640028871596, 'model_time': + 1.2411004360183142, 'grad_norm_pre_clip_avg': + 0.13774264380335807, 'learning_rate': + 2.46187956525501e-07, 'epoch': 11.94} +04/20 [04:37:11] INFO | >> train_qwenlatent.py:487 + Step 47310 | grad_norm_pre_clip=0.1400 | + grad_norm_pre_clip_avg=0.1332 | Metrics: + {'align_loss': 0.026219576597213745, + 'recon_loss': 0.19315487146377563, + 'predict_loss': 0.012627615593373775, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13996124267578125, + 'data_time': 0.0007239739934448153, + 'model_time': 1.2274378529982641, + 'grad_norm_pre_clip_avg': 0.13324216529726982, + 'learning_rate': 2.4455737064626777e-07, + 'epoch': 11.94} +04/20 [04:37:24] INFO | >> train_qwenlatent.py:487 + Step 47320 | grad_norm_pre_clip=0.1433 | + grad_norm_pre_clip_avg=0.1252 | Metrics: + {'align_loss': 0.025657324120402336, + 'recon_loss': 0.17167116701602936, + 'predict_loss': 0.005937579087913036, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14328235387802124, + 'data_time': 0.0009068239887710661, + 'model_time': 1.2885615530249197, + 'grad_norm_pre_clip_avg': 0.12521473318338394, + 'learning_rate': 2.4293276401323726e-07, + 'epoch': 11.94} +04/20 [04:37:36] INFO | >> train_qwenlatent.py:487 + Step 47330 | grad_norm_pre_clip=0.1107 | + grad_norm_pre_clip_avg=0.1392 | Metrics: + {'align_loss': 0.026564909145236015, + 'recon_loss': 0.10386798530817032, + 'predict_loss': 0.0033046563621610403, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11069381982088089, + 'data_time': 0.0006586319941561669, + 'model_time': 1.2125492960039992, + 'grad_norm_pre_clip_avg': 0.13924227356910707, + 'learning_rate': 2.413141374182273e-07, + 'epoch': 11.94} +04/20 [04:37:49] INFO | >> train_qwenlatent.py:487 + Step 47340 | grad_norm_pre_clip=0.1593 | + grad_norm_pre_clip_avg=0.1300 | Metrics: + {'align_loss': 0.025996174663305283, + 'recon_loss': 0.13185608386993408, + 'predict_loss': 0.0034525361843407154, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15926668047904968, + 'data_time': 0.0006563869828823954, + 'model_time': 1.1903085010126233, + 'grad_norm_pre_clip_avg': 0.1299542173743248, + 'learning_rate': 2.397014916501329e-07, + 'epoch': 11.95} +04/20 [04:38:02] INFO | >> train_qwenlatent.py:487 + Step 47350 | grad_norm_pre_clip=0.1213 | + grad_norm_pre_clip_avg=0.1020 | Metrics: + {'align_loss': 0.02499239519238472, + 'recon_loss': 0.15693147480487823, + 'predict_loss': 0.007714014034718275, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12133347243070602, + 'mae_score': 0.006013506382435292, 'data_time': + 0.0009472340170759708, 'model_time': + 1.2448031550156884, 'grad_norm_pre_clip_avg': + 0.10203337967395783, 'learning_rate': + 2.3809482749494244e-07, 'epoch': 11.95} +04/20 [04:38:15] INFO | >> train_qwenlatent.py:487 + Step 47360 | grad_norm_pre_clip=0.1145 | + grad_norm_pre_clip_avg=0.1128 | Metrics: + {'align_loss': 0.02600649744272232, + 'recon_loss': 0.2367262840270996, + 'predict_loss': 0.010211886838078499, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11449465155601501, + 'data_time': 0.0007995140040293336, + 'model_time': 1.2340030680061318, + 'grad_norm_pre_clip_avg': 0.11276040747761726, + 'learning_rate': 2.364941457357199e-07, + 'epoch': 11.95} +04/20 [04:38:28] INFO | >> train_qwenlatent.py:487 + Step 47370 | grad_norm_pre_clip=0.1556 | + grad_norm_pre_clip_avg=0.1348 | Metrics: + {'align_loss': 0.026321053504943848, + 'recon_loss': 0.20209012925624847, + 'predict_loss': 0.004891364835202694, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1556185483932495, + 'data_time': 0.0007691049831919372, + 'model_time': 1.1739493539789692, + 'grad_norm_pre_clip_avg': 0.1347571685910225, + 'learning_rate': 2.3489944715261932e-07, + 'epoch': 11.95} +04/20 [04:38:40] INFO | >> train_qwenlatent.py:487 + Step 47380 | grad_norm_pre_clip=0.1298 | + grad_norm_pre_clip_avg=0.1358 | Metrics: + {'align_loss': 0.02637445740401745, + 'recon_loss': 0.1266079843044281, + 'predict_loss': 0.00508986646309495, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1298246830701828, + 'data_time': 0.0007313769892789423, + 'model_time': 1.2478727039997466, + 'grad_norm_pre_clip_avg': 0.13578371778130532, + 'learning_rate': 2.3331073252288066e-07, + 'epoch': 11.96} +04/20 [04:38:53] INFO | >> train_qwenlatent.py:487 + Step 47390 | grad_norm_pre_clip=0.1452 | + grad_norm_pre_clip_avg=0.1313 | Metrics: + {'align_loss': 0.025891944766044617, + 'recon_loss': 0.2011180818080902, + 'predict_loss': 0.00633134925737977, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14520709216594696, + 'data_time': 0.0006285029812715948, + 'model_time': 1.1791148759948555, + 'grad_norm_pre_clip_avg': 0.13129487857222558, + 'learning_rate': 2.3172800262082142e-07, + 'epoch': 11.96} +04/20 [04:39:07] INFO | >> train_qwenlatent.py:487 + Step 47400 | grad_norm_pre_clip=0.1228 | + grad_norm_pre_clip_avg=0.1213 | Metrics: + {'align_loss': 0.025528933852910995, + 'recon_loss': 0.13471490144729614, + 'predict_loss': 0.00540806632488966, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12279906868934631, + 'mae_score': 0.005492149816977011, 'data_time': + 0.0006038670253474265, 'model_time': + 1.2161706970073283, 'grad_norm_pre_clip_avg': + 0.12133485153317451, 'learning_rate': + 2.3015125821784474e-07, 'epoch': 11.96} +04/20 [04:39:19] INFO | >> train_qwenlatent.py:487 + Step 47410 | grad_norm_pre_clip=0.0956 | + grad_norm_pre_clip_avg=0.1168 | Metrics: + {'align_loss': 0.024296876043081284, + 'recon_loss': 0.11631792783737183, + 'predict_loss': 0.005569003988057375, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09564782679080963, + 'data_time': 0.0007011060079094023, + 'model_time': 1.529300644993782, + 'grad_norm_pre_clip_avg': 0.11683686524629593, + 'learning_rate': 2.2858050008243686e-07, + 'epoch': 11.96} +04/20 [04:39:32] INFO | >> train_qwenlatent.py:487 + Step 47420 | grad_norm_pre_clip=0.1277 | + grad_norm_pre_clip_avg=0.1207 | Metrics: + {'align_loss': 0.02720952406525612, + 'recon_loss': 0.14722798764705658, + 'predict_loss': 0.006494406145066023, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12773139774799347, + 'data_time': 0.0009926659986376762, + 'model_time': 1.2282848760078195, + 'grad_norm_pre_clip_avg': 0.12074804157018662, + 'learning_rate': 2.270157289801656e-07, + 'epoch': 11.97} +04/20 [04:39:44] INFO | >> train_qwenlatent.py:487 + Step 47430 | grad_norm_pre_clip=0.1231 | + grad_norm_pre_clip_avg=0.1166 | Metrics: + {'align_loss': 0.024755699560046196, + 'recon_loss': 0.1424694061279297, + 'predict_loss': 0.00793470535427332, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12312256544828415, + 'data_time': 0.0010534609900787473, + 'model_time': 1.2499535239767283, + 'grad_norm_pre_clip_avg': 0.11663079708814621, + 'learning_rate': 2.254569456736833e-07, + 'epoch': 11.97} +04/20 [04:39:57] INFO | >> train_qwenlatent.py:487 + Step 47440 | grad_norm_pre_clip=0.1750 | + grad_norm_pre_clip_avg=0.1183 | Metrics: + {'align_loss': 0.02443241886794567, + 'recon_loss': 0.09814047813415527, + 'predict_loss': 0.0026984394062310457, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1749599575996399, + 'data_time': 0.0009370660118293017, + 'model_time': 1.2347600819775835, + 'grad_norm_pre_clip_avg': 0.11831883788108825, + 'learning_rate': 2.2390415092272114e-07, + 'epoch': 11.97} +04/20 [04:40:10] INFO | >> train_qwenlatent.py:487 + Step 47450 | grad_norm_pre_clip=0.0995 | + grad_norm_pre_clip_avg=0.1190 | Metrics: + {'align_loss': 0.02299591526389122, + 'recon_loss': 0.19000251591205597, + 'predict_loss': 0.00779256597161293, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09950573742389679, + 'mae_score': 0.005207285580334363, 'data_time': + 0.0007251409988384694, 'model_time': + 1.230116763006663, 'grad_norm_pre_clip_avg': + 0.1190115287899971, 'learning_rate': + 2.2235734548409197e-07, 'epoch': 11.97} +04/20 [04:40:22] INFO | >> train_qwenlatent.py:487 + Step 47460 | grad_norm_pre_clip=0.1382 | + grad_norm_pre_clip_avg=0.1219 | Metrics: + {'align_loss': 0.026312150061130524, + 'recon_loss': 0.15824703872203827, + 'predict_loss': 0.007743820082396269, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13815324008464813, + 'data_time': 0.0009083750192075968, + 'model_time': 1.2163966680236626, + 'grad_norm_pre_clip_avg': 0.12185601145029068, + 'learning_rate': 2.208165301116903e-07, + 'epoch': 11.98} +04/20 [04:40:34] INFO | >> train_qwenlatent.py:487 + Step 47470 | grad_norm_pre_clip=0.1657 | + grad_norm_pre_clip_avg=0.1213 | Metrics: + {'align_loss': 0.025502055883407593, + 'recon_loss': 0.1531052142381668, + 'predict_loss': 0.009318839758634567, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16574962437152863, + 'data_time': 0.0007159730012062937, + 'model_time': 1.2229169700003695, + 'grad_norm_pre_clip_avg': 0.12132994309067727, + 'learning_rate': 2.192817055564894e-07, + 'epoch': 11.98} +04/20 [04:40:47] INFO | >> train_qwenlatent.py:487 + Step 47480 | grad_norm_pre_clip=0.1181 | + grad_norm_pre_clip_avg=0.1289 | Metrics: + {'align_loss': 0.02454063668847084, + 'recon_loss': 0.10889982432126999, + 'predict_loss': 0.0030575974378734827, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11814714968204498, + 'data_time': 0.0008775790047366172, + 'model_time': 1.2492268450150732, + 'grad_norm_pre_clip_avg': 0.12890798673033715, + 'learning_rate': 2.177528725665484e-07, + 'epoch': 11.98} +04/20 [04:41:00] INFO | >> train_qwenlatent.py:487 + Step 47490 | grad_norm_pre_clip=0.1354 | + grad_norm_pre_clip_avg=0.1375 | Metrics: + {'align_loss': 0.02416953444480896, + 'recon_loss': 0.19404621422290802, + 'predict_loss': 0.009043632075190544, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13540317118167877, + 'data_time': 0.0009676470072008669, + 'model_time': 1.2088257049908862, + 'grad_norm_pre_clip_avg': 0.1374677151441574, + 'learning_rate': 2.162300318869971e-07, + 'epoch': 11.98} +04/20 [04:41:13] INFO | >> train_qwenlatent.py:487 + Step 47500 | grad_norm_pre_clip=0.0898 | + grad_norm_pre_clip_avg=0.1256 | Metrics: + {'align_loss': 0.02466636151075363, + 'recon_loss': 0.15720875561237335, + 'predict_loss': 0.007771003991365433, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.08981435745954514, + 'mae_score': 0.004665224401800482, 'data_time': + 0.0007822059851605445, 'model_time': + 1.2462578129780013, 'grad_norm_pre_clip_avg': + 0.1255541741847992, 'learning_rate': + 2.1471318426005233e-07, 'epoch': 11.99} +04/20 [04:41:26] INFO | >> train_qwenlatent.py:487 + Step 47510 | grad_norm_pre_clip=0.1532 | + grad_norm_pre_clip_avg=0.1341 | Metrics: + {'align_loss': 0.025712117552757263, + 'recon_loss': 0.1576416790485382, + 'predict_loss': 0.007314218208193779, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15320166945457458, + 'data_time': 0.0009533749835100025, + 'model_time': 1.356797418993665, + 'grad_norm_pre_clip_avg': 0.13412791788578032, + 'learning_rate': 2.1320233042500578e-07, + 'epoch': 11.99} +04/20 [04:41:39] INFO | >> train_qwenlatent.py:487 + Step 47520 | grad_norm_pre_clip=0.1308 | + grad_norm_pre_clip_avg=0.1265 | Metrics: + {'align_loss': 0.026246260851621628, + 'recon_loss': 0.18476234376430511, + 'predict_loss': 0.00737792020663619, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13079893589019775, + 'data_time': 0.00138550900737755, 'model_time': + 1.2244585139851552, 'grad_norm_pre_clip_avg': + 0.12651228085160254, 'learning_rate': + 2.116974711182292e-07, 'epoch': 11.99} +04/20 [04:41:52] INFO | >> train_qwenlatent.py:487 + Step 47530 | grad_norm_pre_clip=0.0780 | + grad_norm_pre_clip_avg=0.1227 | Metrics: + {'align_loss': 0.025223825126886368, + 'recon_loss': 0.17910632491111755, + 'predict_loss': 0.006270897574722767, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.07801944762468338, + 'data_time': 0.0009726099960971624, + 'model_time': 1.2459049129975028, + 'grad_norm_pre_clip_avg': 0.12274303063750266, + 'learning_rate': 2.1019860707317488e-07, + 'epoch': 11.99} +04/20 [04:42:04] INFO | >> train_qwenlatent.py:487 + Step 47540 | grad_norm_pre_clip=0.1421 | + grad_norm_pre_clip_avg=0.1209 | Metrics: + {'align_loss': 0.02643381431698799, + 'recon_loss': 0.14969941973686218, + 'predict_loss': 0.006424086168408394, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1420963704586029, + 'data_time': 0.0009256019839085639, + 'model_time': 1.223196940991329, + 'grad_norm_pre_clip_avg': 0.12091659158468246, + 'learning_rate': 2.0870573902036952e-07, + 'epoch': 12.0} +04/20 [04:42:17] INFO | >> train_qwenlatent.py:487 + Step 47550 | grad_norm_pre_clip=0.1470 | + grad_norm_pre_clip_avg=0.1049 | Metrics: + {'align_loss': 0.02574474737048149, + 'recon_loss': 0.18408946692943573, + 'predict_loss': 0.008649269118905067, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14697381854057312, + 'mae_score': 0.004973713556925456, 'data_time': + 0.0006744109850842506, 'model_time': + 1.215143168985378, 'grad_norm_pre_clip_avg': + 0.10493695884943008, 'learning_rate': + 2.0721886768741757e-07, 'epoch': 12.0} +04/20 [04:42:30] INFO | >> train_qwenlatent.py:487 + Step 47560 | grad_norm_pre_clip=0.1661 | + grad_norm_pre_clip_avg=0.1121 | Metrics: + {'align_loss': 0.024830687791109085, + 'recon_loss': 0.14622725546360016, + 'predict_loss': 0.006595995742827654, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16609704494476318, + 'data_time': 0.0006390750058926642, + 'model_time': 1.2275389040005393, + 'grad_norm_pre_clip_avg': 0.11214294210076332, + 'learning_rate': 2.0573799379900357e-07, + 'epoch': 12.0} +04/20 [04:42:42] INFO | >> train_qwenlatent.py:487 + Step 47570 | grad_norm_pre_clip=0.1094 | + grad_norm_pre_clip_avg=0.1095 | Metrics: + {'align_loss': 0.027130678296089172, + 'recon_loss': 0.23545554280281067, + 'predict_loss': 0.009445571340620518, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10935836285352707, + 'data_time': 0.000885180983459577, + 'model_time': 1.2868893170088995, + 'grad_norm_pre_clip_avg': 0.10949005782604218, + 'learning_rate': 2.0426311807688689e-07, + 'epoch': 12.0} +04/20 [04:42:55] INFO | >> train_qwenlatent.py:487 + Step 47580 | grad_norm_pre_clip=0.1092 | + grad_norm_pre_clip_avg=0.1234 | Metrics: + {'align_loss': 0.025124285370111465, + 'recon_loss': 0.10551873594522476, + 'predict_loss': 0.0060462309047579765, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10919754207134247, + 'data_time': 0.0008112649957183748, + 'model_time': 1.219289638014743, + 'grad_norm_pre_clip_avg': 0.12338033318519592, + 'learning_rate': 2.0279424123990565e-07, + 'epoch': 12.01} +04/20 [04:43:08] INFO | >> train_qwenlatent.py:487 + Step 47590 | grad_norm_pre_clip=0.0706 | + grad_norm_pre_clip_avg=0.1115 | Metrics: + {'align_loss': 0.025838851928710938, + 'recon_loss': 0.14011336863040924, + 'predict_loss': 0.003126409137621522, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.07061609625816345, + 'data_time': 0.0009701600065454841, + 'model_time': 1.2670428180135787, + 'grad_norm_pre_clip_avg': 0.11147071644663811, + 'learning_rate': 2.0133136400397003e-07, + 'epoch': 12.01} +04/20 [04:43:21] INFO | >> train_qwenlatent.py:487 + Step 47600 | grad_norm_pre_clip=0.1134 | + grad_norm_pre_clip_avg=0.1145 | Metrics: + {'align_loss': 0.025412511080503464, + 'recon_loss': 0.12090211361646652, + 'predict_loss': 0.00895408820360899, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1133664920926094, + 'mae_score': 0.005365633749747061, 'data_time': + 0.0007128030119929463, 'model_time': + 1.5238751790020615, 'grad_norm_pre_clip_avg': + 0.1144742377102375, 'learning_rate': + 1.998744870820704e-07, 'epoch': 12.01} +04/20 [04:43:34] INFO | >> train_qwenlatent.py:487 + Step 47610 | grad_norm_pre_clip=0.1367 | + grad_norm_pre_clip_avg=0.1277 | Metrics: + {'align_loss': 0.026285436004400253, + 'recon_loss': 0.17093151807785034, + 'predict_loss': 0.0069598848931491375, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13673675060272217, + 'data_time': 0.0006907969946041703, + 'model_time': 1.3207015609950759, + 'grad_norm_pre_clip_avg': 0.1277264803647995, + 'learning_rate': 1.9842361118426913e-07, + 'epoch': 12.01} +04/20 [04:43:47] INFO | >> train_qwenlatent.py:487 + Step 47620 | grad_norm_pre_clip=0.0980 | + grad_norm_pre_clip_avg=0.1118 | Metrics: + {'align_loss': 0.027319230139255524, + 'recon_loss': 0.19717563688755035, + 'predict_loss': 0.007779989391565323, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09801321476697922, + 'data_time': 0.0006379600090440363, + 'model_time': 1.2500615729950368, + 'grad_norm_pre_clip_avg': 0.1118215262889862, + 'learning_rate': 1.9697873701770736e-07, + 'epoch': 12.02} +04/20 [04:43:59] INFO | >> train_qwenlatent.py:487 + Step 47630 | grad_norm_pre_clip=0.1463 | + grad_norm_pre_clip_avg=0.1247 | Metrics: + {'align_loss': 0.025592509657144547, + 'recon_loss': 0.14664389193058014, + 'predict_loss': 0.005332519765943289, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14630411565303802, + 'data_time': 0.0009739070083014667, + 'model_time': 1.214396923984168, + 'grad_norm_pre_clip_avg': 0.12474987730383873, + 'learning_rate': 1.9553986528659971e-07, + 'epoch': 12.02} +04/20 [04:44:11] INFO | >> train_qwenlatent.py:487 + Step 47640 | grad_norm_pre_clip=0.1313 | + grad_norm_pre_clip_avg=0.1270 | Metrics: + {'align_loss': 0.02472624182701111, + 'recon_loss': 0.16773638129234314, + 'predict_loss': 0.010860701091587543, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1313202977180481, + 'data_time': 0.0008622070017736405, + 'model_time': 1.2338220999808982, + 'grad_norm_pre_clip_avg': 0.1270328626036644, + 'learning_rate': 1.9410699669223264e-07, + 'epoch': 12.02} +04/20 [04:44:25] INFO | >> train_qwenlatent.py:487 + Step 47650 | grad_norm_pre_clip=0.1039 | + grad_norm_pre_clip_avg=0.1174 | Metrics: + {'align_loss': 0.025853928178548813, + 'recon_loss': 0.1678267866373062, + 'predict_loss': 0.006792206317186356, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10389731079339981, + 'mae_score': 0.005697007651801582, 'data_time': + 0.0008285439980681986, 'model_time': + 1.23274722599308, 'grad_norm_pre_clip_avg': + 0.1174499548971653, 'learning_rate': + 1.9268013193297005e-07, 'epoch': 12.02} +04/20 [04:44:38] INFO | >> train_qwenlatent.py:487 + Step 47660 | grad_norm_pre_clip=0.1521 | + grad_norm_pre_clip_avg=0.1136 | Metrics: + {'align_loss': 0.025134626775979996, + 'recon_loss': 0.10455037653446198, + 'predict_loss': 0.004150578752160072, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15212395787239075, + 'data_time': 0.0006617130129598081, + 'model_time': 1.2032271409989335, + 'grad_norm_pre_clip_avg': 0.11355809941887855, + 'learning_rate': 1.912592717042479e-07, + 'epoch': 12.03} +04/20 [04:44:50] INFO | >> train_qwenlatent.py:487 + Step 47670 | grad_norm_pre_clip=0.0783 | + grad_norm_pre_clip_avg=0.1151 | Metrics: + {'align_loss': 0.026042532175779343, + 'recon_loss': 0.14344370365142822, + 'predict_loss': 0.003161320462822914, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.07825548201799393, + 'data_time': 0.0006177460018079728, + 'model_time': 1.2171932850033045, + 'grad_norm_pre_clip_avg': 0.11505740657448768, + 'learning_rate': 1.898444166985768e-07, + 'epoch': 12.03} +04/20 [04:45:02] INFO | >> train_qwenlatent.py:487 + Step 47680 | grad_norm_pre_clip=0.0837 | + grad_norm_pre_clip_avg=0.1226 | Metrics: + {'align_loss': 0.024936478585004807, + 'recon_loss': 0.09638389945030212, + 'predict_loss': 0.00485093891620636, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.08370700478553772, + 'data_time': 0.0008980560232885182, + 'model_time': 1.1873684310121462, + 'grad_norm_pre_clip_avg': 0.12260374575853347, + 'learning_rate': 1.8843556760553933e-07, + 'epoch': 12.03} +04/20 [04:45:15] INFO | >> train_qwenlatent.py:487 + Step 47690 | grad_norm_pre_clip=0.1368 | + grad_norm_pre_clip_avg=0.1158 | Metrics: + {'align_loss': 0.02617429941892624, + 'recon_loss': 0.14742764830589294, + 'predict_loss': 0.007067766506224871, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13680338859558105, + 'data_time': 0.0009346430015284568, + 'model_time': 1.2676518319931347, + 'grad_norm_pre_clip_avg': 0.11575018540024758, + 'learning_rate': 1.8703272511179274e-07, + 'epoch': 12.03} +04/20 [04:45:28] INFO | >> train_qwenlatent.py:487 + Step 47700 | grad_norm_pre_clip=0.1373 | + grad_norm_pre_clip_avg=0.1211 | Metrics: + {'align_loss': 0.025509150698781013, + 'recon_loss': 0.18825778365135193, + 'predict_loss': 0.007848705165088177, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13732989132404327, + 'mae_score': 0.004828798448717272, 'data_time': + 0.0007966339762788266, 'model_time': + 1.2556827700173017, 'grad_norm_pre_clip_avg': + 0.12106098309159279, 'learning_rate': + 1.8563588990106354e-07, 'epoch': 12.04} +04/20 [04:45:41] INFO | >> train_qwenlatent.py:487 + Step 47710 | grad_norm_pre_clip=0.1519 | + grad_norm_pre_clip_avg=0.1254 | Metrics: + {'align_loss': 0.025347912684082985, + 'recon_loss': 0.11381027102470398, + 'predict_loss': 0.005490752402693033, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15192145109176636, + 'data_time': 0.0007749359938316047, + 'model_time': 1.2717151569959242, + 'grad_norm_pre_clip_avg': 0.12535789757966995, + 'learning_rate': 1.8424506265415283e-07, + 'epoch': 12.04} +04/20 [04:45:53] INFO | >> train_qwenlatent.py:487 + Step 47720 | grad_norm_pre_clip=0.1424 | + grad_norm_pre_clip_avg=0.1100 | Metrics: + {'align_loss': 0.025762775912880898, + 'recon_loss': 0.13608358800411224, + 'predict_loss': 0.004583961330354214, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.142358660697937, + 'data_time': 0.0006377770041581243, + 'model_time': 1.22419962999993, + 'grad_norm_pre_clip_avg': 0.10998695120215415, + 'learning_rate': 1.8286024404893375e-07, + 'epoch': 12.04} +04/20 [04:46:06] INFO | >> train_qwenlatent.py:487 + Step 47730 | grad_norm_pre_clip=0.1560 | + grad_norm_pre_clip_avg=0.1242 | Metrics: + {'align_loss': 0.023685798048973083, + 'recon_loss': 0.15210771560668945, + 'predict_loss': 0.009000581689178944, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15603284537792206, + 'data_time': 0.0009497279825154692, + 'model_time': 1.2726980680017732, + 'grad_norm_pre_clip_avg': 0.12424519583582878, + 'learning_rate': 1.8148143476034997e-07, + 'epoch': 12.04} +04/20 [04:46:19] INFO | >> train_qwenlatent.py:487 + Step 47740 | grad_norm_pre_clip=0.1123 | + grad_norm_pre_clip_avg=0.1191 | Metrics: + {'align_loss': 0.025582704693078995, + 'recon_loss': 0.17199799418449402, + 'predict_loss': 0.010490067303180695, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11225389689207077, + 'data_time': 0.0007304770115297288, + 'model_time': 1.2052959789871238, + 'grad_norm_pre_clip_avg': 0.11908779740333557, + 'learning_rate': 1.8010863546041708e-07, + 'epoch': 12.05} +04/20 [04:46:32] INFO | >> train_qwenlatent.py:487 + Step 47750 | grad_norm_pre_clip=0.1173 | + grad_norm_pre_clip_avg=0.1233 | Metrics: + {'align_loss': 0.02644055150449276, + 'recon_loss': 0.14645038545131683, + 'predict_loss': 0.003973216284066439, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11725281178951263, + 'mae_score': 0.004729642524375572, 'data_time': + 0.0006799300026614219, 'model_time': + 1.2054137880040798, 'grad_norm_pre_clip_avg': + 0.12334225550293923, 'learning_rate': + 1.787418468182185e-07, 'epoch': 12.05} +04/20 [04:46:45] INFO | >> train_qwenlatent.py:487 + Step 47760 | grad_norm_pre_clip=0.0923 | + grad_norm_pre_clip_avg=0.1113 | Metrics: + {'align_loss': 0.024174697697162628, + 'recon_loss': 0.12380750477313995, + 'predict_loss': 0.005599528085440397, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09226088970899582, + 'data_time': 0.0010594780032988638, + 'model_time': 1.2757645679812413, + 'grad_norm_pre_clip_avg': 0.11128769442439079, + 'learning_rate': 1.7738106949991088e-07, + 'epoch': 12.05} +04/20 [04:46:57] INFO | >> train_qwenlatent.py:487 + Step 47770 | grad_norm_pre_clip=0.1222 | + grad_norm_pre_clip_avg=0.1123 | Metrics: + {'align_loss': 0.0261035468429327, + 'recon_loss': 0.17184609174728394, + 'predict_loss': 0.004849026910960674, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1222139373421669, + 'data_time': 0.0010084749956149608, + 'model_time': 1.2298074870195705, + 'grad_norm_pre_clip_avg': 0.11226570978760719, + 'learning_rate': 1.7602630416872156e-07, + 'epoch': 12.05} +04/20 [04:47:10] INFO | >> train_qwenlatent.py:487 + Step 47780 | grad_norm_pre_clip=0.1431 | + grad_norm_pre_clip_avg=0.1362 | Metrics: + {'align_loss': 0.024648115038871765, + 'recon_loss': 0.09463876485824585, + 'predict_loss': 0.003997968975454569, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14313124120235443, + 'data_time': 0.000988655985565856, + 'model_time': 1.244406120997155, + 'grad_norm_pre_clip_avg': 0.13621584102511405, + 'learning_rate': 1.7467755148494694e-07, + 'epoch': 12.06} +04/20 [04:47:23] INFO | >> train_qwenlatent.py:487 + Step 47790 | grad_norm_pre_clip=0.1188 | + grad_norm_pre_clip_avg=0.1294 | Metrics: + {'align_loss': 0.02519499510526657, + 'recon_loss': 0.10072976350784302, + 'predict_loss': 0.006126225460320711, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11882582306861877, + 'data_time': 0.0007428779790643603, + 'model_time': 1.2639764120103791, + 'grad_norm_pre_clip_avg': 0.12935909554362296, + 'learning_rate': 1.733348121059527e-07, + 'epoch': 12.06} +04/20 [04:47:36] INFO | >> train_qwenlatent.py:487 + Step 47800 | grad_norm_pre_clip=0.1687 | + grad_norm_pre_clip_avg=0.1217 | Metrics: + {'align_loss': 0.025154639035463333, + 'recon_loss': 0.15558144450187683, + 'predict_loss': 0.0066726310178637505, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16868023574352264, + 'mae_score': 0.004723470705049532, 'data_time': + 0.0010617260122671723, 'model_time': + 1.2749415369762573, 'grad_norm_pre_clip_avg': + 0.12171759977936744, 'learning_rate': + 1.7199808668617212e-07, 'epoch': 12.06} +04/20 [04:47:49] INFO | >> train_qwenlatent.py:487 + Step 47810 | grad_norm_pre_clip=0.0905 | + grad_norm_pre_clip_avg=0.1190 | Metrics: + {'align_loss': 0.025616055354475975, + 'recon_loss': 0.20420147478580475, + 'predict_loss': 0.0061205108650028706, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09049639850854874, + 'data_time': 0.0007594309863634408, + 'model_time': 1.2076402319944464, + 'grad_norm_pre_clip_avg': 0.1189969539642334, + 'learning_rate': 1.7066737587710925e-07, + 'epoch': 12.06} +04/20 [04:48:01] INFO | >> train_qwenlatent.py:487 + Step 47820 | grad_norm_pre_clip=0.0865 | + grad_norm_pre_clip_avg=0.1270 | Metrics: + {'align_loss': 0.026131346821784973, + 'recon_loss': 0.1490991860628128, + 'predict_loss': 0.005642745178192854, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.0864887610077858, + 'data_time': 0.0009411769860889763, + 'model_time': 1.2714063909952529, + 'grad_norm_pre_clip_avg': 0.12696916311979295, + 'learning_rate': 1.6934268032733854e-07, + 'epoch': 12.07} +04/20 [04:48:14] INFO | >> train_qwenlatent.py:487 + Step 47830 | grad_norm_pre_clip=0.0944 | + grad_norm_pre_clip_avg=0.1112 | Metrics: + {'align_loss': 0.024382364004850388, + 'recon_loss': 0.11483504623174667, + 'predict_loss': 0.0050681112334132195, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09438851475715637, + 'data_time': 0.001065290009137243, + 'model_time': 1.2793492289783899, + 'grad_norm_pre_clip_avg': 0.11115515008568763, + 'learning_rate': 1.6802400068249955e-07, + 'epoch': 12.07} +04/20 [04:48:27] INFO | >> train_qwenlatent.py:487 + Step 47840 | grad_norm_pre_clip=0.1316 | + grad_norm_pre_clip_avg=0.1348 | Metrics: + {'align_loss': 0.025907639414072037, + 'recon_loss': 0.11804159730672836, + 'predict_loss': 0.006348361726850271, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13160502910614014, + 'data_time': 0.0010056380124296993, + 'model_time': 1.7001639789959881, + 'grad_norm_pre_clip_avg': 0.13483286425471305, + 'learning_rate': 1.6671133758529955e-07, + 'epoch': 12.07} +04/20 [04:48:40] INFO | >> train_qwenlatent.py:487 + Step 47850 | grad_norm_pre_clip=0.0805 | + grad_norm_pre_clip_avg=0.1234 | Metrics: + {'align_loss': 0.026002585887908936, + 'recon_loss': 0.21379493176937103, + 'predict_loss': 0.006651143077760935, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.08048810064792633, + 'mae_score': 0.0056106210828901415, + 'data_time': 0.0009364740108139813, + 'model_time': 1.2610378009849228, + 'grad_norm_pre_clip_avg': 0.12340737134218216, + 'learning_rate': 1.65404691675515e-07, 'epoch': + 12.07} +04/20 [04:48:52] INFO | >> train_qwenlatent.py:487 + Step 47860 | grad_norm_pre_clip=0.0729 | + grad_norm_pre_clip_avg=0.1146 | Metrics: + {'align_loss': 0.025669075548648834, + 'recon_loss': 0.14888334274291992, + 'predict_loss': 0.007537065073847771, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.07294099032878876, + 'data_time': 0.0006713320035487413, + 'model_time': 1.214614915981656, + 'grad_norm_pre_clip_avg': 0.11458344012498856, + 'learning_rate': 1.641040635899888e-07, + 'epoch': 12.08} +04/20 [04:49:05] INFO | >> train_qwenlatent.py:487 + Step 47870 | grad_norm_pre_clip=0.1143 | + grad_norm_pre_clip_avg=0.1237 | Metrics: + {'align_loss': 0.026056792587041855, + 'recon_loss': 0.17776532471179962, + 'predict_loss': 0.004910382442176342, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11426645517349243, + 'data_time': 0.000981027988018468, + 'model_time': 1.2626586979895364, + 'grad_norm_pre_clip_avg': 0.12367527559399605, + 'learning_rate': 1.62809453962633e-07, 'epoch': + 12.08} +04/20 [04:49:17] INFO | >> train_qwenlatent.py:487 + Step 47880 | grad_norm_pre_clip=0.1267 | + grad_norm_pre_clip_avg=0.1196 | Metrics: + {'align_loss': 0.025407914072275162, + 'recon_loss': 0.15107303857803345, + 'predict_loss': 0.009671121835708618, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12671421468257904, + 'data_time': 0.0006789200124330819, + 'model_time': 1.310427124000853, + 'grad_norm_pre_clip_avg': 0.11963900998234749, + 'learning_rate': 1.6152086342442456e-07, + 'epoch': 12.08} +04/20 [04:49:30] INFO | >> train_qwenlatent.py:487 + Step 47890 | grad_norm_pre_clip=0.0843 | + grad_norm_pre_clip_avg=0.1176 | Metrics: + {'align_loss': 0.025432318449020386, + 'recon_loss': 0.12801574170589447, + 'predict_loss': 0.0049161408096551895, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.08427099138498306, + 'data_time': 0.0008109659829642624, + 'model_time': 1.2366491869906895, + 'grad_norm_pre_clip_avg': 0.11760463789105416, + 'learning_rate': 1.6023829260340423e-07, + 'epoch': 12.08} +04/20 [04:49:43] INFO | >> train_qwenlatent.py:487 + Step 47900 | grad_norm_pre_clip=0.1346 | + grad_norm_pre_clip_avg=0.1233 | Metrics: + {'align_loss': 0.02544599026441574, + 'recon_loss': 0.16593769192695618, + 'predict_loss': 0.007924199104309082, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13464391231536865, + 'mae_score': 0.005202046385756484, 'data_time': + 0.0015887860208749771, 'model_time': + 1.5138872340030503, 'grad_norm_pre_clip_avg': + 0.12329511642456055, 'learning_rate': + 1.589617421246832e-07, 'epoch': 12.09} +04/20 [04:49:56] INFO | >> train_qwenlatent.py:487 + Step 47910 | grad_norm_pre_clip=0.1737 | + grad_norm_pre_clip_avg=0.1247 | Metrics: + {'align_loss': 0.025924677029252052, + 'recon_loss': 0.14633236825466156, + 'predict_loss': 0.004172676708549261, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17369753122329712, + 'data_time': 0.000626321998424828, + 'model_time': 1.2458619239914697, + 'grad_norm_pre_clip_avg': 0.12474411875009536, + 'learning_rate': 1.5769121261043775e-07, + 'epoch': 12.09} +04/20 [04:50:08] INFO | >> train_qwenlatent.py:487 + Step 47920 | grad_norm_pre_clip=0.1507 | + grad_norm_pre_clip_avg=0.1173 | Metrics: + {'align_loss': 0.02419031597673893, + 'recon_loss': 0.15928292274475098, + 'predict_loss': 0.004610336385667324, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1506691426038742, + 'data_time': 0.0010720489954110235, + 'model_time': 1.2219213710050099, + 'grad_norm_pre_clip_avg': 0.11729645133018493, + 'learning_rate': 1.5642670467990627e-07, + 'epoch': 12.09} +04/20 [04:50:21] INFO | >> train_qwenlatent.py:487 + Step 47930 | grad_norm_pre_clip=0.0989 | + grad_norm_pre_clip_avg=0.1153 | Metrics: + {'align_loss': 0.02688988298177719, + 'recon_loss': 0.21454377472400665, + 'predict_loss': 0.012664803303778172, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.0988672748208046, + 'data_time': 0.0006593750149477273, + 'model_time': 1.1913420780037995, + 'grad_norm_pre_clip_avg': 0.11527841910719872, + 'learning_rate': 1.5516821894939646e-07, + 'epoch': 12.09} +04/20 [04:50:33] INFO | >> train_qwenlatent.py:487 + Step 47940 | grad_norm_pre_clip=0.0960 | + grad_norm_pre_clip_avg=0.1017 | Metrics: + {'align_loss': 0.025520730763673782, + 'recon_loss': 0.2513798773288727, + 'predict_loss': 0.01007112953811884, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09600655734539032, + 'data_time': 0.000887351983692497, + 'model_time': 1.1824394110008143, + 'grad_norm_pre_clip_avg': 0.10171638876199722, + 'learning_rate': 1.5391575603227683e-07, + 'epoch': 12.1} +04/20 [04:50:46] INFO | >> train_qwenlatent.py:487 + Step 47950 | grad_norm_pre_clip=0.0918 | + grad_norm_pre_clip_avg=0.1235 | Metrics: + {'align_loss': 0.024921206757426262, + 'recon_loss': 0.1372258961200714, + 'predict_loss': 0.004787258338183165, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09178585559129715, + 'mae_score': 0.005601979161168004, 'data_time': + 0.0011886880092788488, 'model_time': + 1.278243155014934, 'grad_norm_pre_clip_avg': + 0.12348454892635345, 'learning_rate': + 1.5266931653898502e-07, 'epoch': 12.1} +04/20 [04:50:59] INFO | >> train_qwenlatent.py:487 + Step 47960 | grad_norm_pre_clip=0.1265 | + grad_norm_pre_clip_avg=0.1191 | Metrics: + {'align_loss': 0.025222521275281906, + 'recon_loss': 0.13835535943508148, + 'predict_loss': 0.005282893776893616, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1264987289905548, + 'data_time': 0.0009016820113174617, + 'model_time': 1.1983758450078312, + 'grad_norm_pre_clip_avg': 0.11906087324023247, + 'learning_rate': 1.514289010770195e-07, + 'epoch': 12.1} +04/20 [04:51:12] INFO | >> train_qwenlatent.py:487 + Step 47970 | grad_norm_pre_clip=0.1408 | + grad_norm_pre_clip_avg=0.1196 | Metrics: + {'align_loss': 0.024556245654821396, + 'recon_loss': 0.12143339216709137, + 'predict_loss': 0.0037785316817462444, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14081650972366333, + 'data_time': 0.0006455780239775777, + 'model_time': 1.231756029999815, + 'grad_norm_pre_clip_avg': 0.11955385692417622, + 'learning_rate': 1.5019451025094386e-07, + 'epoch': 12.1} +04/20 [04:51:24] INFO | >> train_qwenlatent.py:487 + Step 47980 | grad_norm_pre_clip=0.0923 | + grad_norm_pre_clip_avg=0.1178 | Metrics: + {'align_loss': 0.025257060304284096, + 'recon_loss': 0.1202460378408432, + 'predict_loss': 0.00345822935923934, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.0922769084572792, + 'data_time': 0.0009885120089165866, + 'model_time': 1.2541166809969582, + 'grad_norm_pre_clip_avg': 0.11776277273893357, + 'learning_rate': 1.4896614466238522e-07, + 'epoch': 12.11} +04/20 [04:51:37] INFO | >> train_qwenlatent.py:487 + Step 47990 | grad_norm_pre_clip=0.1246 | + grad_norm_pre_clip_avg=0.1264 | Metrics: + {'align_loss': 0.025540456175804138, + 'recon_loss': 0.10157381743192673, + 'predict_loss': 0.0026106913574039936, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12461739033460617, + 'data_time': 0.0011869900044985116, + 'model_time': 1.2246131860010792, + 'grad_norm_pre_clip_avg': 0.12638508751988412, + 'learning_rate': 1.4774380491003158e-07, + 'epoch': 12.11} +04/20 [04:51:50] INFO | >> train_qwenlatent.py:487 + Step 48000 | grad_norm_pre_clip=0.1671 | + grad_norm_pre_clip_avg=0.1282 | Metrics: + {'align_loss': 0.025299977511167526, + 'recon_loss': 0.165996715426445, + 'predict_loss': 0.004896737635135651, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16708853840827942, + 'mae_score': 0.00546435106981982, 'data_time': + 0.0009673229942563921, 'model_time': + 1.2024374510219786, 'grad_norm_pre_clip_avg': + 0.1281778708100319, 'learning_rate': + 1.465274915896401e-07, 'epoch': 12.11} +04/20 [04:52:03] INFO | >> train_qwenlatent.py:487 + Step 48010 | grad_norm_pre_clip=0.1535 | + grad_norm_pre_clip_avg=0.1202 | Metrics: + {'align_loss': 0.026000672951340675, + 'recon_loss': 0.20700658857822418, + 'predict_loss': 0.005237536504864693, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.153475821018219, + 'data_time': 0.0008013060141820461, + 'model_time': 1.2299843119981233, + 'grad_norm_pre_clip_avg': 0.12018664777278901, + 'learning_rate': 1.4531720529402746e-07, + 'epoch': 12.11} +04/20 [04:52:16] INFO | >> train_qwenlatent.py:487 + Step 48020 | grad_norm_pre_clip=0.1105 | + grad_norm_pre_clip_avg=0.1156 | Metrics: + {'align_loss': 0.025738853961229324, + 'recon_loss': 0.16747215390205383, + 'predict_loss': 0.00408491026610136, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11051607877016068, + 'data_time': 0.0008331520075444132, + 'model_time': 1.2572065489948727, + 'grad_norm_pre_clip_avg': 0.11556423902511596, + 'learning_rate': 1.441129466130683e-07, + 'epoch': 12.12} +04/20 [04:52:28] INFO | >> train_qwenlatent.py:487 + Step 48030 | grad_norm_pre_clip=0.1285 | + grad_norm_pre_clip_avg=0.1315 | Metrics: + {'align_loss': 0.025225594639778137, + 'recon_loss': 0.14048685133457184, + 'predict_loss': 0.007270440459251404, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12851868569850922, + 'data_time': 0.00097687600646168, 'model_time': + 1.5480170620139688, 'grad_norm_pre_clip_avg': + 0.13149479180574417, 'learning_rate': + 1.4291471613370654e-07, 'epoch': 12.12} +04/20 [04:52:41] INFO | >> train_qwenlatent.py:487 + Step 48040 | grad_norm_pre_clip=0.0767 | + grad_norm_pre_clip_avg=0.1162 | Metrics: + {'align_loss': 0.024656157940626144, + 'recon_loss': 0.11824203282594681, + 'predict_loss': 0.005208923481404781, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.07671744376420975, + 'data_time': 0.0009722769900690764, + 'model_time': 1.2130983200040646, + 'grad_norm_pre_clip_avg': 0.11621024459600449, + 'learning_rate': 1.4172251443994554e-07, + 'epoch': 12.12} +04/20 [04:52:54] INFO | >> train_qwenlatent.py:487 + Step 48050 | grad_norm_pre_clip=0.1042 | + grad_norm_pre_clip_avg=0.1349 | Metrics: + {'align_loss': 0.025982510298490524, + 'recon_loss': 0.13497091829776764, + 'predict_loss': 0.006554827559739351, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10422245413064957, + 'mae_score': 0.005376211372581688, 'data_time': + 0.000989015999948606, 'model_time': + 1.2321651249949355, 'grad_norm_pre_clip_avg': + 0.13492308259010316, 'learning_rate': + 1.405363421128481e-07, 'epoch': 12.12} +04/20 [04:53:06] INFO | >> train_qwenlatent.py:487 + Step 48060 | grad_norm_pre_clip=0.0850 | + grad_norm_pre_clip_avg=0.1114 | Metrics: + {'align_loss': 0.025528233498334885, + 'recon_loss': 0.1490398496389389, + 'predict_loss': 0.005466683767735958, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.08495304733514786, + 'data_time': 0.000979087984887883, + 'model_time': 1.2914587579725776, + 'grad_norm_pre_clip_avg': 0.11142811104655266, + 'learning_rate': 1.3935619973054202e-07, + 'epoch': 12.13} +04/20 [04:53:19] INFO | >> train_qwenlatent.py:487 + Step 48070 | grad_norm_pre_clip=0.0804 | + grad_norm_pre_clip_avg=0.1217 | Metrics: + {'align_loss': 0.026869021356105804, + 'recon_loss': 0.20982575416564941, + 'predict_loss': 0.005274481605738401, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.08044306188821793, + 'data_time': 0.0007061710057314485, + 'model_time': 1.2090995760227088, + 'grad_norm_pre_clip_avg': 0.1216572217643261, + 'learning_rate': 1.3818208786821318e-07, + 'epoch': 12.13} +04/20 [04:53:32] INFO | >> train_qwenlatent.py:487 + Step 48080 | grad_norm_pre_clip=0.1091 | + grad_norm_pre_clip_avg=0.1248 | Metrics: + {'align_loss': 0.025215037167072296, + 'recon_loss': 0.14499996602535248, + 'predict_loss': 0.0057785664685070515, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10914099216461182, + 'data_time': 0.0011822250089608133, + 'model_time': 1.2182828889926895, + 'grad_norm_pre_clip_avg': 0.12483873516321183, + 'learning_rate': 1.370140070981098e-07, + 'epoch': 12.13} +04/20 [04:53:44] INFO | >> train_qwenlatent.py:487 + Step 48090 | grad_norm_pre_clip=0.1109 | + grad_norm_pre_clip_avg=0.1257 | Metrics: + {'align_loss': 0.025426974520087242, + 'recon_loss': 0.13748036324977875, + 'predict_loss': 0.006354723125696182, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11093685030937195, + 'data_time': 0.0008683400228619576, + 'model_time': 1.2964223079907242, + 'grad_norm_pre_clip_avg': 0.12570179998874664, + 'learning_rate': 1.358519579895394e-07, + 'epoch': 12.13} +04/20 [04:53:58] INFO | >> train_qwenlatent.py:487 + Step 48100 | grad_norm_pre_clip=0.1296 | + grad_norm_pre_clip_avg=0.1230 | Metrics: + {'align_loss': 0.02555065229535103, + 'recon_loss': 0.18723522126674652, + 'predict_loss': 0.006986522581428289, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1295836865901947, + 'mae_score': 0.0050805714753297, 'data_time': + 0.0007513650052715093, 'model_time': + 1.225182827009121, 'grad_norm_pre_clip_avg': + 0.12295709252357483, 'learning_rate': + 1.3469594110887325e-07, 'epoch': 12.14} +04/20 [04:54:10] INFO | >> train_qwenlatent.py:487 + Step 48110 | grad_norm_pre_clip=0.1350 | + grad_norm_pre_clip_avg=0.1185 | Metrics: + {'align_loss': 0.025454899296164513, + 'recon_loss': 0.15670807659626007, + 'predict_loss': 0.005117328837513924, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13501591980457306, + 'data_time': 0.0008455319912172854, + 'model_time': 1.2540543109935243, + 'grad_norm_pre_clip_avg': 0.1184962272644043, + 'learning_rate': 1.3354595701953787e-07, + 'epoch': 12.14} +04/20 [04:54:23] INFO | >> train_qwenlatent.py:487 + Step 48120 | grad_norm_pre_clip=0.1240 | + grad_norm_pre_clip_avg=0.1150 | Metrics: + {'align_loss': 0.024717524647712708, + 'recon_loss': 0.16312751173973083, + 'predict_loss': 0.0074112205766141415, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12398819625377655, + 'data_time': 0.0009615460003260523, + 'model_time': 1.2485460290045012, + 'grad_norm_pre_clip_avg': 0.11496651321649551, + 'learning_rate': 1.3240200628202062e-07, + 'epoch': 12.14} +04/20 [04:54:36] INFO | >> train_qwenlatent.py:487 + Step 48130 | grad_norm_pre_clip=0.1634 | + grad_norm_pre_clip_avg=0.1307 | Metrics: + {'align_loss': 0.0261429063975811, + 'recon_loss': 0.15405210852622986, + 'predict_loss': 0.004450273234397173, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16338863968849182, + 'data_time': 0.0009711270104162395, + 'model_time': 1.5220520439906977, + 'grad_norm_pre_clip_avg': 0.13073805943131447, + 'learning_rate': 1.3126408945386975e-07, + 'epoch': 12.14} +04/20 [04:54:48] INFO | >> train_qwenlatent.py:487 + Step 48140 | grad_norm_pre_clip=0.0828 | + grad_norm_pre_clip_avg=0.1017 | Metrics: + {'align_loss': 0.026009343564510345, + 'recon_loss': 0.14958034455776215, + 'predict_loss': 0.005419959779828787, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.08281746506690979, + 'data_time': 0.0008718320168554783, + 'model_time': 1.2201918649952859, + 'grad_norm_pre_clip_avg': 0.10170888155698776, + 'learning_rate': 1.3013220708969436e-07, + 'epoch': 12.15} +04/20 [04:55:01] INFO | >> train_qwenlatent.py:487 + Step 48150 | grad_norm_pre_clip=0.1152 | + grad_norm_pre_clip_avg=0.1106 | Metrics: + {'align_loss': 0.025284484028816223, + 'recon_loss': 0.2023554891347885, + 'predict_loss': 0.008555182255804539, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11523650586605072, + 'mae_score': 0.005536097449225348, 'data_time': + 0.00065150301088579, 'model_time': + 1.2831610039866064, 'grad_norm_pre_clip_avg': + 0.11056612953543662, 'learning_rate': + 1.290063597411588e-07, 'epoch': 12.15} +04/20 [04:55:14] INFO | >> train_qwenlatent.py:487 + Step 48160 | grad_norm_pre_clip=0.1181 | + grad_norm_pre_clip_avg=0.1141 | Metrics: + {'align_loss': 0.025141842663288116, + 'recon_loss': 0.12945185601711273, + 'predict_loss': 0.005448494106531143, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11813530325889587, + 'data_time': 0.0007116560009308159, + 'model_time': 1.2650960340106394, + 'grad_norm_pre_clip_avg': 0.11411667168140412, + 'learning_rate': 1.2788654795698556e-07, + 'epoch': 12.15} +04/20 [04:55:27] INFO | >> train_qwenlatent.py:487 + Step 48170 | grad_norm_pre_clip=0.1144 | + grad_norm_pre_clip_avg=0.1086 | Metrics: + {'align_loss': 0.025789525359869003, + 'recon_loss': 0.12147846817970276, + 'predict_loss': 0.005803295876830816, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1143568679690361, + 'data_time': 0.0006777019880246371, + 'model_time': 1.2281163470179308, + 'grad_norm_pre_clip_avg': 0.10857335105538368, + 'learning_rate': 1.2677277228296074e-07, + 'epoch': 12.15} +04/20 [04:55:39] INFO | >> train_qwenlatent.py:487 + Step 48180 | grad_norm_pre_clip=0.0944 | + grad_norm_pre_clip_avg=0.1206 | Metrics: + {'align_loss': 0.026021035388112068, + 'recon_loss': 0.13546329736709595, + 'predict_loss': 0.003507452318444848, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09435705095529556, + 'data_time': 0.0008997870027087629, + 'model_time': 1.232096706982702, + 'grad_norm_pre_clip_avg': 0.12059195861220359, + 'learning_rate': 1.2566503326192156e-07, + 'epoch': 12.16} +04/20 [04:55:52] INFO | >> train_qwenlatent.py:487 + Step 48190 | grad_norm_pre_clip=0.1067 | + grad_norm_pre_clip_avg=0.1141 | Metrics: + {'align_loss': 0.026756709441542625, + 'recon_loss': 0.12123902887105942, + 'predict_loss': 0.008027554489672184, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1067272499203682, + 'data_time': 0.0009143929928541183, + 'model_time': 1.2663090389978606, + 'grad_norm_pre_clip_avg': 0.11414432302117347, + 'learning_rate': 1.2456333143376892e-07, + 'epoch': 12.16} +04/20 [04:56:05] INFO | >> train_qwenlatent.py:487 + Step 48200 | grad_norm_pre_clip=0.1313 | + grad_norm_pre_clip_avg=0.1125 | Metrics: + {'align_loss': 0.024486687034368515, + 'recon_loss': 0.12277821451425552, + 'predict_loss': 0.006933137774467468, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13125485181808472, + 'mae_score': 0.005368414870253554, 'data_time': + 0.0009601059718988836, 'model_time': + 1.248574553988874, 'grad_norm_pre_clip_avg': + 0.11253103911876679, 'learning_rate': + 1.2346766733545892e-07, 'epoch': 12.16} +04/20 [04:56:18] INFO | >> train_qwenlatent.py:487 + Step 48210 | grad_norm_pre_clip=0.1175 | + grad_norm_pre_clip_avg=0.1171 | Metrics: + {'align_loss': 0.024551838636398315, + 'recon_loss': 0.1258247196674347, + 'predict_loss': 0.003976717591285706, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11752030998468399, + 'data_time': 0.0006576580053661019, + 'model_time': 1.2242836600053124, + 'grad_norm_pre_clip_avg': 0.11707842722535133, + 'learning_rate': 1.2237804150100308e-07, + 'epoch': 12.17} +04/20 [04:56:30] INFO | >> train_qwenlatent.py:487 + Step 48220 | grad_norm_pre_clip=0.0872 | + grad_norm_pre_clip_avg=0.1163 | Metrics: + {'align_loss': 0.025703364983201027, + 'recon_loss': 0.10955116897821426, + 'predict_loss': 0.00508895143866539, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.08721920102834702, + 'data_time': 0.0010578269721008837, + 'model_time': 1.2846297869982664, + 'grad_norm_pre_clip_avg': 0.11633971482515335, + 'learning_rate': 1.2129445446147232e-07, + 'epoch': 12.17} +04/20 [04:56:43] INFO | >> train_qwenlatent.py:487 + Step 48230 | grad_norm_pre_clip=0.1117 | + grad_norm_pre_clip_avg=0.1175 | Metrics: + {'align_loss': 0.026685582473874092, + 'recon_loss': 0.18856114149093628, + 'predict_loss': 0.006615506950765848, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11170069128274918, + 'data_time': 0.0008356969919987023, + 'model_time': 1.2471752279961947, + 'grad_norm_pre_clip_avg': 0.11751313284039497, + 'learning_rate': 1.2021690674499568e-07, + 'epoch': 12.17} +04/20 [04:56:56] INFO | >> train_qwenlatent.py:487 + Step 48240 | grad_norm_pre_clip=0.1518 | + grad_norm_pre_clip_avg=0.1215 | Metrics: + {'align_loss': 0.02318871021270752, + 'recon_loss': 0.09713003784418106, + 'predict_loss': 0.0031716155353933573, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15182806551456451, + 'data_time': 0.0006840279966127127, + 'model_time': 1.247265382000478, + 'grad_norm_pre_clip_avg': 0.12154068052768707, + 'learning_rate': 1.191453988767547e-07, + 'epoch': 12.17} +04/20 [04:57:09] INFO | >> train_qwenlatent.py:487 + Step 48250 | grad_norm_pre_clip=0.1273 | + grad_norm_pre_clip_avg=0.1212 | Metrics: + {'align_loss': 0.023329053074121475, + 'recon_loss': 0.12401162087917328, + 'predict_loss': 0.004166461061686277, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12733164429664612, + 'mae_score': 0.005549057109935864, 'data_time': + 0.0006700909871142358, 'model_time': + 1.2447583510074764, 'grad_norm_pre_clip_avg': + 0.12119153663516044, 'learning_rate': + 1.1807993137899175e-07, 'epoch': 12.18} +04/20 [04:57:21] INFO | >> train_qwenlatent.py:487 + Step 48260 | grad_norm_pre_clip=0.0819 | + grad_norm_pre_clip_avg=0.1211 | Metrics: + {'align_loss': 0.025172080844640732, + 'recon_loss': 0.1610635221004486, + 'predict_loss': 0.005314353853464127, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.0818934217095375, + 'data_time': 0.0009459340071771294, + 'model_time': 1.2570567759976257, + 'grad_norm_pre_clip_avg': 0.12109574601054192, + 'learning_rate': 1.17020504770999e-07, 'epoch': + 12.18} +04/20 [04:57:35] INFO | >> train_qwenlatent.py:487 + Step 48270 | grad_norm_pre_clip=0.1027 | + grad_norm_pre_clip_avg=0.1134 | Metrics: + {'align_loss': 0.025757625699043274, + 'recon_loss': 0.16585972905158997, + 'predict_loss': 0.009017310105264187, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10268118977546692, + 'data_time': 0.0008841209928505123, + 'model_time': 1.6574463330034632, + 'grad_norm_pre_clip_avg': 0.11339273005723953, + 'learning_rate': 1.1596711956913227e-07, + 'epoch': 12.18} +04/20 [04:57:47] INFO | >> train_qwenlatent.py:487 + Step 48280 | grad_norm_pre_clip=0.1741 | + grad_norm_pre_clip_avg=0.1348 | Metrics: + {'align_loss': 0.025347216054797173, + 'recon_loss': 0.1226189061999321, + 'predict_loss': 0.008788920007646084, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.17413702607154846, + 'data_time': 0.0007296029943972826, + 'model_time': 1.2328878269763663, + 'grad_norm_pre_clip_avg': 0.1348286598920822, + 'learning_rate': 1.1491977628679566e-07, + 'epoch': 12.18} +04/20 [04:57:59] INFO | >> train_qwenlatent.py:487 + Step 48290 | grad_norm_pre_clip=0.1316 | + grad_norm_pre_clip_avg=0.1065 | Metrics: + {'align_loss': 0.025985432788729668, + 'recon_loss': 0.10090388357639313, + 'predict_loss': 0.004232499282807112, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13163645565509796, + 'data_time': 0.0008134439995046705, + 'model_time': 1.2118508590210695, + 'grad_norm_pre_clip_avg': 0.10648156628012657, + 'learning_rate': 1.1387847543445285e-07, + 'epoch': 12.19} +04/20 [04:58:12] INFO | >> train_qwenlatent.py:487 + Step 48300 | grad_norm_pre_clip=0.1175 | + grad_norm_pre_clip_avg=0.1134 | Metrics: + {'align_loss': 0.025815511122345924, + 'recon_loss': 0.13971373438835144, + 'predict_loss': 0.006398831959813833, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11746420711278915, + 'mae_score': 0.0052658768387528155, + 'data_time': 0.0009079600276891142, + 'model_time': 1.3042185369995423, + 'grad_norm_pre_clip_avg': 0.11341474801301957, + 'learning_rate': 1.1284321751961994e-07, + 'epoch': 12.19} +04/20 [04:58:25] INFO | >> train_qwenlatent.py:487 + Step 48310 | grad_norm_pre_clip=0.0784 | + grad_norm_pre_clip_avg=0.1180 | Metrics: + {'align_loss': 0.02550886943936348, + 'recon_loss': 0.19732876121997833, + 'predict_loss': 0.008346674963831902, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.0784146636724472, + 'data_time': 0.000929874979192391, + 'model_time': 1.2350730879988987, + 'grad_norm_pre_clip_avg': 0.11796126142144203, + 'learning_rate': 1.1181400304687119e-07, + 'epoch': 12.19} +04/20 [04:58:38] INFO | >> train_qwenlatent.py:487 + Step 48320 | grad_norm_pre_clip=0.0924 | + grad_norm_pre_clip_avg=0.1220 | Metrics: + {'align_loss': 0.02667778916656971, + 'recon_loss': 0.19813263416290283, + 'predict_loss': 0.008159183897078037, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09241402894258499, + 'data_time': 0.0012211500143166631, + 'model_time': 1.2402951320109423, + 'grad_norm_pre_clip_avg': 0.1220124050974846, + 'learning_rate': 1.1079083251783059e-07, + 'epoch': 12.19} +04/20 [04:58:51] INFO | >> train_qwenlatent.py:487 + Step 48330 | grad_norm_pre_clip=0.1127 | + grad_norm_pre_clip_avg=0.1319 | Metrics: + {'align_loss': 0.026587676256895065, + 'recon_loss': 0.13956260681152344, + 'predict_loss': 0.006859851069748402, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.112749382853508, + 'data_time': 0.0011227189970668405, + 'model_time': 1.299320506019285, + 'grad_norm_pre_clip_avg': 0.13188875019550322, + 'learning_rate': 1.097737064311816e-07, + 'epoch': 12.2} +04/20 [04:59:04] INFO | >> train_qwenlatent.py:487 + Step 48340 | grad_norm_pre_clip=0.1047 | + grad_norm_pre_clip_avg=0.1223 | Metrics: + {'align_loss': 0.025231044739484787, + 'recon_loss': 0.139614537358284, + 'predict_loss': 0.004687551874667406, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10466235131025314, + 'data_time': 0.0006467629864346236, + 'model_time': 1.2413313389988616, + 'grad_norm_pre_clip_avg': 0.12225265726447106, + 'learning_rate': 1.0876262528265882e-07, + 'epoch': 12.2} +04/20 [04:59:17] INFO | >> train_qwenlatent.py:487 + Step 48350 | grad_norm_pre_clip=0.0888 | + grad_norm_pre_clip_avg=0.1299 | Metrics: + {'align_loss': 0.02581544779241085, + 'recon_loss': 0.14917142689228058, + 'predict_loss': 0.004290167707949877, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.08876227587461472, + 'mae_score': 0.004861139177202104, 'data_time': + 0.0008394570031668991, 'model_time': + 1.2104075280076358, 'grad_norm_pre_clip_avg': + 0.1299237810075283, 'learning_rate': + 1.0775758956504938e-07, 'epoch': 12.2} +04/20 [04:59:30] INFO | >> train_qwenlatent.py:487 + Step 48360 | grad_norm_pre_clip=0.0866 | + grad_norm_pre_clip_avg=0.1036 | Metrics: + {'align_loss': 0.02609282359480858, + 'recon_loss': 0.14047668874263763, + 'predict_loss': 0.004165889695286751, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.08659668266773224, + 'data_time': 0.0009057059942279011, + 'model_time': 1.2909513339982368, + 'grad_norm_pre_clip_avg': 0.10359935238957405, + 'learning_rate': 1.0675859976819711e-07, + 'epoch': 12.2} +04/20 [04:59:42] INFO | >> train_qwenlatent.py:487 + Step 48370 | grad_norm_pre_clip=0.1142 | + grad_norm_pre_clip_avg=0.1070 | Metrics: + {'align_loss': 0.025977037847042084, + 'recon_loss': 0.15151996910572052, + 'predict_loss': 0.006764736957848072, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11418291181325912, + 'data_time': 0.0008385410183109343, + 'model_time': 1.3288280149863567, + 'grad_norm_pre_clip_avg': 0.1069889523088932, + 'learning_rate': 1.0576565637899563e-07, + 'epoch': 12.21} +04/20 [04:59:55] INFO | >> train_qwenlatent.py:487 + Step 48380 | grad_norm_pre_clip=0.1571 | + grad_norm_pre_clip_avg=0.1258 | Metrics: + {'align_loss': 0.02490277960896492, + 'recon_loss': 0.15961405634880066, + 'predict_loss': 0.00690922886133194, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15712082386016846, + 'data_time': 0.0009998850000556558, + 'model_time': 1.2386667699902318, + 'grad_norm_pre_clip_avg': 0.12579469308257102, + 'learning_rate': 1.0477875988139517e-07, + 'epoch': 12.21} +04/20 [05:00:07] INFO | >> train_qwenlatent.py:487 + Step 48390 | grad_norm_pre_clip=0.1033 | + grad_norm_pre_clip_avg=0.1117 | Metrics: + {'align_loss': 0.024518225342035294, + 'recon_loss': 0.09964238107204437, + 'predict_loss': 0.005647706333547831, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10329566150903702, + 'data_time': 0.0008184470061678439, + 'model_time': 1.261296051001409, + 'grad_norm_pre_clip_avg': 0.11171553879976273, + 'learning_rate': 1.0379791075639578e-07, + 'epoch': 12.21} +04/20 [05:00:21] INFO | >> train_qwenlatent.py:487 + Step 48400 | grad_norm_pre_clip=0.1020 | + grad_norm_pre_clip_avg=0.1180 | Metrics: + {'align_loss': 0.02523583360016346, + 'recon_loss': 0.16460920870304108, + 'predict_loss': 0.006022048182785511, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10196597129106522, + 'mae_score': 0.004691080144933752, 'data_time': + 0.0006615379825234413, 'model_time': + 1.251173900993308, 'grad_norm_pre_clip_avg': + 0.11801399886608124, 'learning_rate': + 1.0282310948205283e-07, 'epoch': 12.21} +04/20 [05:00:33] INFO | >> train_qwenlatent.py:487 + Step 48410 | grad_norm_pre_clip=0.1264 | + grad_norm_pre_clip_avg=0.1232 | Metrics: + {'align_loss': 0.024520641192793846, + 'recon_loss': 0.13817012310028076, + 'predict_loss': 0.005053575150668621, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12641379237174988, + 'data_time': 0.0006876429833937436, + 'model_time': 1.1746472069935407, + 'grad_norm_pre_clip_avg': 0.12321390137076378, + 'learning_rate': 1.0185435653347142e-07, + 'epoch': 12.22} +04/20 [05:00:46] INFO | >> train_qwenlatent.py:487 + Step 48420 | grad_norm_pre_clip=0.1249 | + grad_norm_pre_clip_avg=0.1110 | Metrics: + {'align_loss': 0.02579442597925663, + 'recon_loss': 0.18688777089118958, + 'predict_loss': 0.006563916336745024, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12491975724697113, + 'data_time': 0.0005961380084045231, + 'model_time': 1.2639050469733775, + 'grad_norm_pre_clip_avg': 0.1109567865729332, + 'learning_rate': 1.0089165238281057e-07, + 'epoch': 12.22} +04/20 [05:00:58] INFO | >> train_qwenlatent.py:487 + Step 48430 | grad_norm_pre_clip=0.1658 | + grad_norm_pre_clip_avg=0.1344 | Metrics: + {'align_loss': 0.025516124442219734, + 'recon_loss': 0.22170889377593994, + 'predict_loss': 0.01362708117812872, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.16575071215629578, + 'data_time': 0.0006817200046498328, + 'model_time': 1.2315795120084658, + 'grad_norm_pre_clip_avg': 0.13438753336668013, + 'learning_rate': 9.993499749928042e-08, + 'epoch': 12.22} +04/20 [05:01:11] INFO | >> train_qwenlatent.py:487 + Step 48440 | grad_norm_pre_clip=0.1581 | + grad_norm_pre_clip_avg=0.1187 | Metrics: + {'align_loss': 0.02668623998761177, + 'recon_loss': 0.24323683977127075, + 'predict_loss': 0.010757544077932835, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15814208984375, + 'data_time': 0.0007078670023474842, + 'model_time': 1.2236906249891035, + 'grad_norm_pre_clip_avg': 0.11871765926480293, + 'learning_rate': 9.898439234914234e-08, + 'epoch': 12.22} +04/20 [05:01:25] INFO | >> train_qwenlatent.py:487 + Step 48450 | grad_norm_pre_clip=0.1160 | + grad_norm_pre_clip_avg=0.1165 | Metrics: + {'align_loss': 0.025779511779546738, + 'recon_loss': 0.13846023380756378, + 'predict_loss': 0.005817696917802095, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11597203463315964, + 'mae_score': 0.005299479252583272, 'data_time': + 0.0006596189923584461, 'model_time': + 1.5576180549978744, 'grad_norm_pre_clip_avg': + 0.11647270321846008, 'learning_rate': + 9.803983739571153e-08, 'epoch': 12.23} +04/20 [05:01:37] INFO | >> train_qwenlatent.py:487 + Step 48460 | grad_norm_pre_clip=0.1438 | + grad_norm_pre_clip_avg=0.1215 | Metrics: + {'align_loss': 0.025273174047470093, + 'recon_loss': 0.1607532948255539, + 'predict_loss': 0.00705267209559679, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14379116892814636, + 'data_time': 0.0007550199807155877, + 'model_time': 1.254509952996159, + 'grad_norm_pre_clip_avg': 0.12145339846611022, + 'learning_rate': 9.710133309935022e-08, + 'epoch': 12.23} +04/20 [05:01:49] INFO | >> train_qwenlatent.py:487 + Step 48470 | grad_norm_pre_clip=0.0918 | + grad_norm_pre_clip_avg=0.1082 | Metrics: + {'align_loss': 0.025504061952233315, + 'recon_loss': 0.1216268315911293, + 'predict_loss': 0.0026934726629406214, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09180282056331635, + 'data_time': 0.0007008489919826388, + 'model_time': 1.2041723530273885, + 'grad_norm_pre_clip_avg': 0.10815484151244163, + 'learning_rate': 9.616887991747594e-08, + 'epoch': 12.23} +04/20 [05:02:02] INFO | >> train_qwenlatent.py:487 + Step 48480 | grad_norm_pre_clip=0.1386 | + grad_norm_pre_clip_avg=0.1108 | Metrics: + {'align_loss': 0.026967104524374008, + 'recon_loss': 0.17650765180587769, + 'predict_loss': 0.006561512127518654, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13861770927906036, + 'data_time': 0.0006943930056877434, + 'model_time': 1.2302992409968283, + 'grad_norm_pre_clip_avg': 0.11080899834632874, + 'learning_rate': 9.524247830455461e-08, + 'epoch': 12.23} +04/20 [05:02:15] INFO | >> train_qwenlatent.py:487 + Step 48490 | grad_norm_pre_clip=0.0837 | + grad_norm_pre_clip_avg=0.1102 | Metrics: + {'align_loss': 0.02503616362810135, + 'recon_loss': 0.10078249126672745, + 'predict_loss': 0.004206099081784487, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.0836561769247055, + 'data_time': 0.0009761549881659448, + 'model_time': 1.2901748009899165, + 'grad_norm_pre_clip_avg': 0.11017523184418679, + 'learning_rate': 9.432212871210186e-08, + 'epoch': 12.24} +04/20 [05:02:28] INFO | >> train_qwenlatent.py:487 + Step 48500 | grad_norm_pre_clip=0.1025 | + grad_norm_pre_clip_avg=0.1260 | Metrics: + {'align_loss': 0.026285532861948013, + 'recon_loss': 0.13173402845859528, + 'predict_loss': 0.003187190741300583, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10252207517623901, + 'mae_score': 0.006148204717550192, 'data_time': + 0.0011828950082417578, 'model_time': + 1.2315099110128358, 'grad_norm_pre_clip_avg': + 0.12597465440630912, 'learning_rate': + 9.34078315886873e-08, 'epoch': 12.24} +04/20 [05:02:41] INFO | >> train_qwenlatent.py:487 + Step 48510 | grad_norm_pre_clip=0.1307 | + grad_norm_pre_clip_avg=0.1298 | Metrics: + {'align_loss': 0.025123318657279015, + 'recon_loss': 0.16757068037986755, + 'predict_loss': 0.00555016053840518, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13069462776184082, + 'data_time': 0.0010343629983253777, + 'model_time': 1.2583712890045717, + 'grad_norm_pre_clip_avg': 0.12978593707084657, + 'learning_rate': 9.24995873799247e-08, 'epoch': + 12.24} +04/20 [05:02:53] INFO | >> train_qwenlatent.py:487 + Step 48520 | grad_norm_pre_clip=0.1801 | + grad_norm_pre_clip_avg=0.1257 | Metrics: + {'align_loss': 0.02576533891260624, + 'recon_loss': 0.21910864114761353, + 'predict_loss': 0.015296266414225101, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1800895631313324, + 'data_time': 0.0007213859935291111, + 'model_time': 1.236707104020752, + 'grad_norm_pre_clip_avg': 0.12573283538222313, + 'learning_rate': 9.159739652848457e-08, + 'epoch': 12.24} +04/20 [05:03:06] INFO | >> train_qwenlatent.py:487 + Step 48530 | grad_norm_pre_clip=0.1150 | + grad_norm_pre_clip_avg=0.1140 | Metrics: + {'align_loss': 0.025621525943279266, + 'recon_loss': 0.11709588021039963, + 'predict_loss': 0.004009374417364597, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1149907261133194, + 'data_time': 0.0006708560104016215, + 'model_time': 1.2201171240012627, + 'grad_norm_pre_clip_avg': 0.11395316570997238, + 'learning_rate': 9.0701259474083e-08, 'epoch': + 12.25} +04/20 [05:03:18] INFO | >> train_qwenlatent.py:487 + Step 48540 | grad_norm_pre_clip=0.0879 | + grad_norm_pre_clip_avg=0.1130 | Metrics: + {'align_loss': 0.02579072117805481, + 'recon_loss': 0.1274317055940628, + 'predict_loss': 0.004770633764564991, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.08785788714885712, + 'data_time': 0.0006301779940258712, + 'model_time': 1.4797642189951148, + 'grad_norm_pre_clip_avg': 0.11304631680250168, + 'learning_rate': 8.981117665348583e-08, + 'epoch': 12.25} +04/20 [05:03:31] INFO | >> train_qwenlatent.py:487 + Step 48550 | grad_norm_pre_clip=0.0766 | + grad_norm_pre_clip_avg=0.1221 | Metrics: + {'align_loss': 0.02547333389520645, + 'recon_loss': 0.17672300338745117, + 'predict_loss': 0.008235249668359756, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.07660089433193207, + 'mae_score': 0.005406484517965231, 'data_time': + 0.0009499359875917435, 'model_time': + 1.2285036029934417, 'grad_norm_pre_clip_avg': + 0.12212277576327324, 'learning_rate': + 8.892714850050726e-08, 'epoch': 12.25} +04/20 [05:03:44] INFO | >> train_qwenlatent.py:487 + Step 48560 | grad_norm_pre_clip=0.1184 | + grad_norm_pre_clip_avg=0.1210 | Metrics: + {'align_loss': 0.026036903262138367, + 'recon_loss': 0.19321106374263763, + 'predict_loss': 0.013119957409799099, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11841367185115814, + 'data_time': 0.0006854220118839294, + 'model_time': 1.2275504609860945, + 'grad_norm_pre_clip_avg': 0.12095781788229942, + 'learning_rate': 8.804917544601409e-08, + 'epoch': 12.25} +04/20 [05:03:56] INFO | >> train_qwenlatent.py:487 + Step 48570 | grad_norm_pre_clip=0.0738 | + grad_norm_pre_clip_avg=0.1102 | Metrics: + {'align_loss': 0.025544343516230583, + 'recon_loss': 0.13454702496528625, + 'predict_loss': 0.004840565845370293, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.07382849603891373, + 'data_time': 0.0008377489866688848, + 'model_time': 1.2156237769813742, + 'grad_norm_pre_clip_avg': 0.11021519973874092, + 'learning_rate': 8.717725791791864e-08, + 'epoch': 12.26} +04/20 [05:04:09] INFO | >> train_qwenlatent.py:487 + Step 48580 | grad_norm_pre_clip=0.1034 | + grad_norm_pre_clip_avg=0.1290 | Metrics: + {'align_loss': 0.02635914646089077, + 'recon_loss': 0.17836570739746094, + 'predict_loss': 0.010894253849983215, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10335270315408707, + 'data_time': 0.0007184010173659772, + 'model_time': 1.3096432500169612, + 'grad_norm_pre_clip_avg': 0.12904300391674042, + 'learning_rate': 8.631139634118443e-08, + 'epoch': 12.26} +04/20 [05:04:22] INFO | >> train_qwenlatent.py:487 + Step 48590 | grad_norm_pre_clip=0.1254 | + grad_norm_pre_clip_avg=0.1054 | Metrics: + {'align_loss': 0.02537602186203003, + 'recon_loss': 0.1387525498867035, + 'predict_loss': 0.0060387346893548965, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12540695071220398, + 'data_time': 0.0007541499799117446, + 'model_time': 1.2389715460012667, + 'grad_norm_pre_clip_avg': 0.10535015091300011, + 'learning_rate': 8.545159113781916e-08, + 'epoch': 12.26} +04/20 [05:04:35] INFO | >> train_qwenlatent.py:487 + Step 48600 | grad_norm_pre_clip=0.1017 | + grad_norm_pre_clip_avg=0.1280 | Metrics: + {'align_loss': 0.02523762732744217, + 'recon_loss': 0.13921642303466797, + 'predict_loss': 0.004933502990752459, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10174663364887238, + 'mae_score': 0.005382555025117891, 'data_time': + 0.0009402169962413609, 'model_time': + 1.2263807700073812, 'grad_norm_pre_clip_avg': + 0.1279866598546505, 'learning_rate': + 8.459784272688308e-08, 'epoch': 12.26} +04/20 [05:04:48] INFO | >> train_qwenlatent.py:487 + Step 48610 | grad_norm_pre_clip=0.0913 | + grad_norm_pre_clip_avg=0.1273 | Metrics: + {'align_loss': 0.02435949072241783, + 'recon_loss': 0.16421596705913544, + 'predict_loss': 0.007501076441258192, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09132911264896393, + 'data_time': 0.0011845019762404263, + 'model_time': 1.4705815839988645, + 'grad_norm_pre_clip_avg': 0.1272565469145775, + 'learning_rate': 8.375015152448207e-08, + 'epoch': 12.27} +04/20 [05:05:00] INFO | >> train_qwenlatent.py:487 + Step 48620 | grad_norm_pre_clip=0.1285 | + grad_norm_pre_clip_avg=0.1025 | Metrics: + {'align_loss': 0.025784695520997047, + 'recon_loss': 0.16122443974018097, + 'predict_loss': 0.0070346943102777, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12852230668067932, + 'data_time': 0.0007327329949475825, + 'model_time': 1.2354448160040192, + 'grad_norm_pre_clip_avg': 0.1024830162525177, + 'learning_rate': 8.29085179437731e-08, 'epoch': + 12.27} +04/20 [05:05:13] INFO | >> train_qwenlatent.py:487 + Step 48630 | grad_norm_pre_clip=0.0740 | + grad_norm_pre_clip_avg=0.1001 | Metrics: + {'align_loss': 0.02473212406039238, + 'recon_loss': 0.1668078750371933, + 'predict_loss': 0.005993238650262356, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.07395555824041367, + 'data_time': 0.0008102919964585453, + 'model_time': 1.2016504339990206, + 'grad_norm_pre_clip_avg': 0.1001211054623127, + 'learning_rate': 8.20729423949546e-08, 'epoch': + 12.27} +04/20 [05:05:25] INFO | >> train_qwenlatent.py:487 + Step 48640 | grad_norm_pre_clip=0.1073 | + grad_norm_pre_clip_avg=0.1075 | Metrics: + {'align_loss': 0.025961466133594513, + 'recon_loss': 0.16175441443920135, + 'predict_loss': 0.00881385151296854, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10728922486305237, + 'data_time': 0.0006633840093854815, + 'model_time': 1.2441961799922865, + 'grad_norm_pre_clip_avg': 0.10746478885412217, + 'learning_rate': 8.124342528527756e-08, + 'epoch': 12.27} +04/20 [05:05:38] INFO | >> train_qwenlatent.py:487 + Step 48650 | grad_norm_pre_clip=0.1670 | + grad_norm_pre_clip_avg=0.1150 | Metrics: + {'align_loss': 0.02419213391840458, + 'recon_loss': 0.1341777741909027, + 'predict_loss': 0.0038422911893576384, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1669776290655136, + 'mae_score': 0.0059771189818511134, + 'data_time': 0.0006651470030192286, + 'model_time': 1.2141089549986646, + 'grad_norm_pre_clip_avg': 0.11501271799206733, + 'learning_rate': 8.041996701903856e-08, + 'epoch': 12.28} +04/20 [05:05:51] INFO | >> train_qwenlatent.py:487 + Step 48660 | grad_norm_pre_clip=0.1304 | + grad_norm_pre_clip_avg=0.1011 | Metrics: + {'align_loss': 0.026185981929302216, + 'recon_loss': 0.1667286455631256, + 'predict_loss': 0.011894961819052696, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1304074078798294, + 'data_time': 0.0007197879895102233, + 'model_time': 1.2256461289944127, + 'grad_norm_pre_clip_avg': 0.10107860639691353, + 'learning_rate': 7.960256799758112e-08, + 'epoch': 12.28} +04/20 [05:06:03] INFO | >> train_qwenlatent.py:487 + Step 48670 | grad_norm_pre_clip=0.1021 | + grad_norm_pre_clip_avg=0.1116 | Metrics: + {'align_loss': 0.02578079327940941, + 'recon_loss': 0.22745069861412048, + 'predict_loss': 0.010856714099645615, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10214845836162567, + 'data_time': 0.0010028830147348344, + 'model_time': 1.2431831630237866, + 'grad_norm_pre_clip_avg': 0.11158070713281631, + 'learning_rate': 7.879122861929582e-08, + 'epoch': 12.28} +04/20 [05:06:16] INFO | >> train_qwenlatent.py:487 + Step 48680 | grad_norm_pre_clip=0.0921 | + grad_norm_pre_clip_avg=0.1178 | Metrics: + {'align_loss': 0.025675443932414055, + 'recon_loss': 0.15813446044921875, + 'predict_loss': 0.005913677159696817, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09206707030534744, + 'data_time': 0.001018066017422825, + 'model_time': 1.2646942640130874, + 'grad_norm_pre_clip_avg': 0.11784187480807304, + 'learning_rate': 7.798594927961878e-08, + 'epoch': 12.28} +04/20 [05:06:29] INFO | >> train_qwenlatent.py:487 + Step 48690 | grad_norm_pre_clip=0.1021 | + grad_norm_pre_clip_avg=0.1100 | Metrics: + {'align_loss': 0.024849120527505875, + 'recon_loss': 0.11436038464307785, + 'predict_loss': 0.0028285474982112646, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10208139568567276, + 'data_time': 0.0009716339991427958, + 'model_time': 1.2021647740039043, + 'grad_norm_pre_clip_avg': 0.11000652313232422, + 'learning_rate': 7.718673037103316e-08, + 'epoch': 12.29} +04/20 [05:06:42] INFO | >> train_qwenlatent.py:487 + Step 48700 | grad_norm_pre_clip=0.1139 | + grad_norm_pre_clip_avg=0.1163 | Metrics: + {'align_loss': 0.026337627321481705, + 'recon_loss': 0.15746428072452545, + 'predict_loss': 0.003684983355924487, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11386507004499435, + 'mae_score': 0.00541651742952364, 'data_time': + 0.0009196989994961768, 'model_time': + 1.2156046360032633, 'grad_norm_pre_clip_avg': + 0.11626245602965354, 'learning_rate': + 7.639357228307046e-08, 'epoch': 12.29} +04/20 [05:06:55] INFO | >> train_qwenlatent.py:487 + Step 48710 | grad_norm_pre_clip=0.1159 | + grad_norm_pre_clip_avg=0.1188 | Metrics: + {'align_loss': 0.026393577456474304, + 'recon_loss': 0.11013033986091614, + 'predict_loss': 0.002453522989526391, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11587027460336685, + 'data_time': 0.0009324889979325235, + 'model_time': 1.2490501969878096, + 'grad_norm_pre_clip_avg': 0.11884189546108245, + 'learning_rate': 7.56064754023036e-08, 'epoch': + 12.29} +04/20 [05:07:07] INFO | >> train_qwenlatent.py:487 + Step 48720 | grad_norm_pre_clip=0.1234 | + grad_norm_pre_clip_avg=0.1171 | Metrics: + {'align_loss': 0.026043465360999107, + 'recon_loss': 0.17012616991996765, + 'predict_loss': 0.00895137432962656, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1234058365225792, + 'data_time': 0.000908765010535717, + 'model_time': 1.1962017420155462, + 'grad_norm_pre_clip_avg': 0.11712043732404709, + 'learning_rate': 7.48254401123581e-08, 'epoch': + 12.29} +04/20 [05:07:20] INFO | >> train_qwenlatent.py:487 + Step 48730 | grad_norm_pre_clip=0.1327 | + grad_norm_pre_clip_avg=0.1146 | Metrics: + {'align_loss': 0.024095851927995682, + 'recon_loss': 0.1320309191942215, + 'predict_loss': 0.004988566506654024, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13269498944282532, + 'data_time': 0.0009284320112783462, + 'model_time': 1.2305845319933724, + 'grad_norm_pre_clip_avg': 0.11459986642003059, + 'learning_rate': 7.405046679389672e-08, + 'epoch': 12.3} +04/20 [05:07:32] INFO | >> train_qwenlatent.py:487 + Step 48740 | grad_norm_pre_clip=0.0991 | + grad_norm_pre_clip_avg=0.1198 | Metrics: + {'align_loss': 0.02438843622803688, + 'recon_loss': 0.13457117974758148, + 'predict_loss': 0.0057486314326524734, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09908165782690048, + 'data_time': 0.001011076004942879, + 'model_time': 1.2371483080205508, + 'grad_norm_pre_clip_avg': 0.1197585754096508, + 'learning_rate': 7.328155582463334e-08, + 'epoch': 12.3} +04/20 [05:07:46] INFO | >> train_qwenlatent.py:487 + Step 48750 | grad_norm_pre_clip=0.0981 | + grad_norm_pre_clip_avg=0.1067 | Metrics: + {'align_loss': 0.02605961263179779, + 'recon_loss': 0.18954874575138092, + 'predict_loss': 0.006978310644626617, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09806594997644424, + 'mae_score': 0.005391623952367284, 'data_time': + 0.0011275210126768798, 'model_time': + 1.2654733549861703, 'grad_norm_pre_clip_avg': + 0.10669223293662071, 'learning_rate': + 7.251870757932751e-08, 'epoch': 12.3} +04/20 [05:07:58] INFO | >> train_qwenlatent.py:487 + Step 48760 | grad_norm_pre_clip=0.0839 | + grad_norm_pre_clip_avg=0.1251 | Metrics: + {'align_loss': 0.025056440383195877, + 'recon_loss': 0.11814938485622406, + 'predict_loss': 0.003711891360580921, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.08392610400915146, + 'data_time': 0.000782398012233898, + 'model_time': 1.2191230029857252, + 'grad_norm_pre_clip_avg': 0.12511421293020247, + 'learning_rate': 7.176192242978154e-08, + 'epoch': 12.3} +04/20 [05:08:11] INFO | >> train_qwenlatent.py:487 + Step 48770 | grad_norm_pre_clip=0.1092 | + grad_norm_pre_clip_avg=0.1149 | Metrics: + {'align_loss': 0.023572079837322235, + 'recon_loss': 0.14702264964580536, + 'predict_loss': 0.006631580181419849, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10924427956342697, + 'data_time': 0.0011327340034767985, + 'model_time': 1.2418689749902114, + 'grad_norm_pre_clip_avg': 0.11487039551138878, + 'learning_rate': 7.101120074484199e-08, + 'epoch': 12.31} +04/20 [05:08:23] INFO | >> train_qwenlatent.py:487 + Step 48780 | grad_norm_pre_clip=0.0923 | + grad_norm_pre_clip_avg=0.1092 | Metrics: + {'align_loss': 0.024417806416749954, + 'recon_loss': 0.11443755775690079, + 'predict_loss': 0.003582253586500883, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09234333783388138, + 'data_time': 0.0017547690076753497, + 'model_time': 1.2470532190054655, + 'grad_norm_pre_clip_avg': 0.10922578275203705, + 'learning_rate': 7.0266542890401e-08, 'epoch': + 12.31} +04/20 [05:08:35] INFO | >> train_qwenlatent.py:487 + Step 48790 | grad_norm_pre_clip=0.1183 | + grad_norm_pre_clip_avg=0.1158 | Metrics: + {'align_loss': 0.02635064162313938, + 'recon_loss': 0.14627017080783844, + 'predict_loss': 0.004528885241597891, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11829707771539688, + 'data_time': 0.0007030539854895324, + 'model_time': 1.2733006630151067, + 'grad_norm_pre_clip_avg': 0.11581404209136963, + 'learning_rate': 6.952794922939914e-08, + 'epoch': 12.31} +04/20 [05:08:48] INFO | >> train_qwenlatent.py:487 + Step 48800 | grad_norm_pre_clip=0.0702 | + grad_norm_pre_clip_avg=0.1131 | Metrics: + {'align_loss': 0.025107771158218384, + 'recon_loss': 0.09811051189899445, + 'predict_loss': 0.003981332760304213, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.07017726451158524, + 'mae_score': 0.005197436959893854, 'data_time': + 0.0006844509916845709, 'model_time': + 1.2066206420131493, 'grad_norm_pre_clip_avg': + 0.11305276155471802, 'learning_rate': + 6.879542012181418e-08, 'epoch': 12.31} +04/20 [05:09:00] INFO | >> train_qwenlatent.py:487 + Step 48810 | grad_norm_pre_clip=0.1028 | + grad_norm_pre_clip_avg=0.1212 | Metrics: + {'align_loss': 0.02454696223139763, + 'recon_loss': 0.09790050238370895, + 'predict_loss': 0.0032171818893402815, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10284004360437393, + 'data_time': 0.0008992450020741671, + 'model_time': 1.2501088859862648, + 'grad_norm_pre_clip_avg': 0.12124574407935143, + 'learning_rate': 6.806895592467509e-08, + 'epoch': 12.32} +04/20 [05:09:13] INFO | >> train_qwenlatent.py:487 + Step 48820 | grad_norm_pre_clip=0.1118 | + grad_norm_pre_clip_avg=0.0959 | Metrics: + {'align_loss': 0.026755183935165405, + 'recon_loss': 0.19815243780612946, + 'predict_loss': 0.004299935419112444, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11177122592926025, + 'data_time': 0.0010692710056900978, + 'model_time': 1.2656434290111065, + 'grad_norm_pre_clip_avg': 0.09593052566051483, + 'learning_rate': 6.734855699204951e-08, + 'epoch': 12.32} +04/20 [05:09:26] INFO | >> train_qwenlatent.py:487 + Step 48830 | grad_norm_pre_clip=0.1200 | + grad_norm_pre_clip_avg=0.1286 | Metrics: + {'align_loss': 0.02344878390431404, + 'recon_loss': 0.08737149834632874, + 'predict_loss': 0.003315154230222106, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12001127004623413, + 'data_time': 0.001020408992189914, + 'model_time': 1.233872014010558, + 'grad_norm_pre_clip_avg': 0.12863886579871178, + 'learning_rate': 6.663422367505201e-08, + 'epoch': 12.32} +04/20 [05:09:38] INFO | >> train_qwenlatent.py:487 + Step 48840 | grad_norm_pre_clip=0.0975 | + grad_norm_pre_clip_avg=0.1106 | Metrics: + {'align_loss': 0.025777243077754974, + 'recon_loss': 0.18893130123615265, + 'predict_loss': 0.006461769342422485, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09749049693346024, + 'data_time': 0.0008474639907944947, + 'model_time': 1.2645665479940362, + 'grad_norm_pre_clip_avg': 0.11060133874416352, + 'learning_rate': 6.592595632184008e-08, + 'epoch': 12.32} +04/20 [05:09:52] INFO | >> train_qwenlatent.py:487 + Step 48850 | grad_norm_pre_clip=0.0892 | + grad_norm_pre_clip_avg=0.1134 | Metrics: + {'align_loss': 0.02433769777417183, + 'recon_loss': 0.14128340780735016, + 'predict_loss': 0.008069738745689392, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.08921705931425095, + 'mae_score': 0.005390549565220738, 'data_time': + 0.0010405349894426763, 'model_time': + 1.2396498610032722, 'grad_norm_pre_clip_avg': + 0.1134291909635067, 'learning_rate': + 6.522375527761396e-08, 'epoch': 12.33} +04/20 [05:10:05] INFO | >> train_qwenlatent.py:487 + Step 48860 | grad_norm_pre_clip=0.1173 | + grad_norm_pre_clip_avg=0.1108 | Metrics: + {'align_loss': 0.02594350278377533, + 'recon_loss': 0.15187059342861176, + 'predict_loss': 0.00519341416656971, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11725210398435593, + 'data_time': 0.0009605539962649345, + 'model_time': 1.2789451459830161, + 'grad_norm_pre_clip_avg': 0.1108342669904232, + 'learning_rate': 6.452762088461815e-08, + 'epoch': 12.33} +04/20 [05:10:17] INFO | >> train_qwenlatent.py:487 + Step 48870 | grad_norm_pre_clip=0.1538 | + grad_norm_pre_clip_avg=0.1129 | Metrics: + {'align_loss': 0.02488170936703682, + 'recon_loss': 0.14449584484100342, + 'predict_loss': 0.008036589249968529, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15379644930362701, + 'data_time': 0.000682399986544624, + 'model_time': 1.2444238509924617, + 'grad_norm_pre_clip_avg': 0.11290760561823845, + 'learning_rate': 6.383755348213993e-08, + 'epoch': 12.33} +04/20 [05:10:30] INFO | >> train_qwenlatent.py:487 + Step 48880 | grad_norm_pre_clip=0.0888 | + grad_norm_pre_clip_avg=0.1221 | Metrics: + {'align_loss': 0.023678652942180634, + 'recon_loss': 0.13364872336387634, + 'predict_loss': 0.006339760962873697, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.08883202821016312, + 'data_time': 0.000877903017681092, + 'model_time': 1.2211851390020456, + 'grad_norm_pre_clip_avg': 0.12206890732049942, + 'learning_rate': 6.315355340651085e-08, + 'epoch': 12.33} +04/20 [05:10:43] INFO | >> train_qwenlatent.py:487 + Step 48890 | grad_norm_pre_clip=0.1043 | + grad_norm_pre_clip_avg=0.1055 | Metrics: + {'align_loss': 0.02602507546544075, + 'recon_loss': 0.1659253090620041, + 'predict_loss': 0.009016595780849457, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1042688861489296, + 'data_time': 0.000659014011034742, + 'model_time': 1.2116781570075545, + 'grad_norm_pre_clip_avg': 0.10548215210437775, + 'learning_rate': 6.24756209911025e-08, 'epoch': + 12.34} +04/20 [05:10:56] INFO | >> train_qwenlatent.py:487 + Step 48900 | grad_norm_pre_clip=0.1199 | + grad_norm_pre_clip_avg=0.1078 | Metrics: + {'align_loss': 0.02619110979139805, + 'recon_loss': 0.15363585948944092, + 'predict_loss': 0.004364727530628443, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1198902502655983, + 'mae_score': 0.005365991162824201, 'data_time': + 0.0009160570043604821, 'model_time': + 1.1909960419870913, 'grad_norm_pre_clip_avg': + 0.10779636427760124, 'learning_rate': + 6.180375656633342e-08, 'epoch': 12.34} +04/20 [05:11:08] INFO | >> train_qwenlatent.py:487 + Step 48910 | grad_norm_pre_clip=0.0804 | + grad_norm_pre_clip_avg=0.1054 | Metrics: + {'align_loss': 0.02564438059926033, + 'recon_loss': 0.11986304074525833, + 'predict_loss': 0.004018017556518316, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.08039220422506332, + 'data_time': 0.0006969099922571331, + 'model_time': 1.1956356060109101, + 'grad_norm_pre_clip_avg': 0.10543565824627876, + 'learning_rate': 6.113796045965948e-08, + 'epoch': 12.34} +04/20 [05:11:21] INFO | >> train_qwenlatent.py:487 + Step 48920 | grad_norm_pre_clip=0.0908 | + grad_norm_pre_clip_avg=0.0987 | Metrics: + {'align_loss': 0.025858968496322632, + 'recon_loss': 0.13419128954410553, + 'predict_loss': 0.0075453114695847034, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09079612046480179, + 'data_time': 0.0007533770112786442, + 'model_time': 1.2167947820271365, + 'grad_norm_pre_clip_avg': 0.09868858158588409, + 'learning_rate': 6.047823299558351e-08, + 'epoch': 12.34} +04/20 [05:11:33] INFO | >> train_qwenlatent.py:487 + Step 48930 | grad_norm_pre_clip=0.0731 | + grad_norm_pre_clip_avg=0.1010 | Metrics: + {'align_loss': 0.0251096710562706, + 'recon_loss': 0.14303889870643616, + 'predict_loss': 0.0058735571801662445, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.07311327010393143, + 'data_time': 0.0008278030145447701, + 'model_time': 1.245558422000613, + 'grad_norm_pre_clip_avg': 0.10100657567381859, + 'learning_rate': 5.98245744956484e-08, 'epoch': + 12.35} +04/20 [05:11:46] INFO | >> train_qwenlatent.py:487 + Step 48940 | grad_norm_pre_clip=0.1049 | + grad_norm_pre_clip_avg=0.1182 | Metrics: + {'align_loss': 0.025397606194019318, + 'recon_loss': 0.11786556988954544, + 'predict_loss': 0.005111980717629194, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10491640865802765, + 'data_time': 0.0009120610193349421, + 'model_time': 1.2372298389964271, + 'grad_norm_pre_clip_avg': 0.11817494183778762, + 'learning_rate': 5.9176985278437094e-08, + 'epoch': 12.35} +04/20 [05:11:59] INFO | >> train_qwenlatent.py:487 + Step 48950 | grad_norm_pre_clip=0.1387 | + grad_norm_pre_clip_avg=0.1243 | Metrics: + {'align_loss': 0.02429998107254505, + 'recon_loss': 0.12548936903476715, + 'predict_loss': 0.007380998693406582, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13867105543613434, + 'mae_score': 0.004631559698431342, 'data_time': + 0.0007445839873980731, 'model_time': + 1.2222441900230478, 'grad_norm_pre_clip_avg': + 0.12427941858768463, 'learning_rate': + 5.853546565957953e-08, 'epoch': 12.35} +04/20 [05:12:12] INFO | >> train_qwenlatent.py:487 + Step 48960 | grad_norm_pre_clip=0.0933 | + grad_norm_pre_clip_avg=0.1108 | Metrics: + {'align_loss': 0.025198165327310562, + 'recon_loss': 0.10418392717838287, + 'predict_loss': 0.004219891969114542, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09326620399951935, + 'data_time': 0.001162581000244245, + 'model_time': 1.2552227179985493, + 'grad_norm_pre_clip_avg': 0.11076239645481109, + 'learning_rate': 5.790001595174293e-08, + 'epoch': 12.35} +04/20 [05:12:24] INFO | >> train_qwenlatent.py:487 + Step 48970 | grad_norm_pre_clip=0.0886 | + grad_norm_pre_clip_avg=0.1013 | Metrics: + {'align_loss': 0.025057390332221985, + 'recon_loss': 0.15300463140010834, + 'predict_loss': 0.003980978857725859, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.08858641982078552, + 'data_time': 0.0007100639923010021, + 'model_time': 1.269235120009398, + 'grad_norm_pre_clip_avg': 0.1012533314526081, + 'learning_rate': 5.727063646463874e-08, + 'epoch': 12.36} +04/20 [05:12:37] INFO | >> train_qwenlatent.py:487 + Step 48980 | grad_norm_pre_clip=0.0746 | + grad_norm_pre_clip_avg=0.0977 | Metrics: + {'align_loss': 0.02436014823615551, + 'recon_loss': 0.18253184854984283, + 'predict_loss': 0.00822505448013544, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.07463374733924866, + 'data_time': 0.0009483619942329824, + 'model_time': 1.2029527450213209, + 'grad_norm_pre_clip_avg': 0.09770099073648453, + 'learning_rate': 5.6647327505019796e-08, + 'epoch': 12.36} +04/20 [05:12:50] INFO | >> train_qwenlatent.py:487 + Step 48990 | grad_norm_pre_clip=0.0875 | + grad_norm_pre_clip_avg=0.1056 | Metrics: + {'align_loss': 0.02611212246119976, + 'recon_loss': 0.16984187066555023, + 'predict_loss': 0.005835404619574547, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.08747126162052155, + 'data_time': 0.0007109009893611073, + 'model_time': 1.254325252986746, + 'grad_norm_pre_clip_avg': 0.10560734495520592, + 'learning_rate': 5.603008937667628e-08, + 'epoch': 12.36} +04/20 [05:13:03] INFO | >> train_qwenlatent.py:487 + Step 49000 | grad_norm_pre_clip=0.1148 | + grad_norm_pre_clip_avg=0.1085 | Metrics: + {'align_loss': 0.02492310106754303, + 'recon_loss': 0.15691255033016205, + 'predict_loss': 0.005213976372033358, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11477135866880417, + 'mae_score': 0.00621062184239293, 'data_time': + 0.001131803001044318, 'model_time': + 1.2418421320035122, 'grad_norm_pre_clip_avg': + 0.1085242435336113, 'learning_rate': + 5.5418922380445334e-08, 'epoch': 12.36} +04/20 [05:13:16] INFO | >> train_qwenlatent.py:487 + Step 49010 | grad_norm_pre_clip=0.1233 | + grad_norm_pre_clip_avg=0.1105 | Metrics: + {'align_loss': 0.024510836228728294, + 'recon_loss': 0.1421220451593399, + 'predict_loss': 0.0057717738673090935, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12329612672328949, + 'data_time': 0.0006708189903292805, + 'model_time': 1.2161877420148812, + 'grad_norm_pre_clip_avg': 0.11050036102533341, + 'learning_rate': 5.4813826814199986e-08, + 'epoch': 12.37} +04/20 [05:13:29] INFO | >> train_qwenlatent.py:487 + Step 49020 | grad_norm_pre_clip=0.0894 | + grad_norm_pre_clip_avg=0.1188 | Metrics: + {'align_loss': 0.02535814791917801, + 'recon_loss': 0.16770146787166595, + 'predict_loss': 0.005977317690849304, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.08944141119718552, + 'data_time': 0.0013703480071853846, + 'model_time': 1.229514170991024, + 'grad_norm_pre_clip_avg': 0.11882341429591178, + 'learning_rate': 5.421480297285889e-08, + 'epoch': 12.37} +04/20 [05:13:42] INFO | >> train_qwenlatent.py:487 + Step 49030 | grad_norm_pre_clip=0.1414 | + grad_norm_pre_clip_avg=0.1193 | Metrics: + {'align_loss': 0.0251353420317173, + 'recon_loss': 0.18176952004432678, + 'predict_loss': 0.0062448401004076, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14135636389255524, + 'data_time': 0.0009664969984441996, + 'model_time': 1.2285588109807577, + 'grad_norm_pre_clip_avg': 0.11927675902843475, + 'learning_rate': 5.3621851148377935e-08, + 'epoch': 12.37} +04/20 [05:13:54] INFO | >> train_qwenlatent.py:487 + Step 49040 | grad_norm_pre_clip=0.1173 | + grad_norm_pre_clip_avg=0.1135 | Metrics: + {'align_loss': 0.02570178359746933, + 'recon_loss': 0.13499046862125397, + 'predict_loss': 0.004810688551515341, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11731770634651184, + 'data_time': 0.0006594379956368357, + 'model_time': 1.1751065489952452, + 'grad_norm_pre_clip_avg': 0.11352578923106194, + 'learning_rate': 5.3034971629753146e-08, + 'epoch': 12.37} +04/20 [05:14:07] INFO | >> train_qwenlatent.py:487 + Step 49050 | grad_norm_pre_clip=0.1279 | + grad_norm_pre_clip_avg=0.1133 | Metrics: + {'align_loss': 0.025790000334382057, + 'recon_loss': 0.19609156250953674, + 'predict_loss': 0.008514217101037502, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1278965175151825, + 'mae_score': 0.005880368292868674, 'data_time': + 0.0008157920092344284, 'model_time': + 1.2452167159935925, 'grad_norm_pre_clip_avg': + 0.11332118511199951, 'learning_rate': + 5.2454164703026064e-08, 'epoch': 12.38} +04/20 [05:14:20] INFO | >> train_qwenlatent.py:487 + Step 49060 | grad_norm_pre_clip=0.1388 | + grad_norm_pre_clip_avg=0.1202 | Metrics: + {'align_loss': 0.02549765072762966, + 'recon_loss': 0.1634143441915512, + 'predict_loss': 0.007507271133363247, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13881689310073853, + 'data_time': 0.0009472979872953147, + 'model_time': 1.295468594005797, + 'grad_norm_pre_clip_avg': 0.12017257213592529, + 'learning_rate': 5.187943065127141e-08, + 'epoch': 12.38} +04/20 [05:14:32] INFO | >> train_qwenlatent.py:487 + Step 49070 | grad_norm_pre_clip=0.0671 | + grad_norm_pre_clip_avg=0.1118 | Metrics: + {'align_loss': 0.026597920805215836, + 'recon_loss': 0.15431183576583862, + 'predict_loss': 0.00334778125397861, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.06710228323936462, + 'data_time': 0.0009446580079384148, + 'model_time': 1.2597410220187157, + 'grad_norm_pre_clip_avg': 0.11182447671890258, + 'learning_rate': 5.1310769754609474e-08, + 'epoch': 12.38} +04/20 [05:14:45] INFO | >> train_qwenlatent.py:487 + Step 49080 | grad_norm_pre_clip=0.1317 | + grad_norm_pre_clip_avg=0.1212 | Metrics: + {'align_loss': 0.02568121999502182, + 'recon_loss': 0.15515051782131195, + 'predict_loss': 0.004782469943165779, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13167618215084076, + 'data_time': 0.0007267630135174841, + 'model_time': 1.3062719349982217, + 'grad_norm_pre_clip_avg': 0.12116171270608903, + 'learning_rate': 5.074818229019785e-08, + 'epoch': 12.38} +04/20 [05:14:57] INFO | >> train_qwenlatent.py:487 + Step 49090 | grad_norm_pre_clip=0.1117 | + grad_norm_pre_clip_avg=0.1110 | Metrics: + {'align_loss': 0.02510773576796055, + 'recon_loss': 0.13082240521907806, + 'predict_loss': 0.006516546942293644, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1117221936583519, + 'data_time': 0.0013319510035216808, + 'model_time': 1.3048647580144461, + 'grad_norm_pre_clip_avg': 0.1110489159822464, + 'learning_rate': 5.019166853223417e-08, + 'epoch': 12.39} +04/20 [05:15:10] INFO | >> train_qwenlatent.py:487 + Step 49100 | grad_norm_pre_clip=0.1515 | + grad_norm_pre_clip_avg=0.1123 | Metrics: + {'align_loss': 0.025405004620552063, + 'recon_loss': 0.2681589722633362, + 'predict_loss': 0.009839750826358795, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15152379870414734, + 'mae_score': 0.006836836617272179, 'data_time': + 0.0006724050035700202, 'model_time': + 1.2156913100043312, 'grad_norm_pre_clip_avg': + 0.11231052055954933, 'learning_rate': + 4.964122875195891e-08, 'epoch': 12.39} +04/20 [05:15:23] INFO | >> train_qwenlatent.py:487 + Step 49110 | grad_norm_pre_clip=0.1368 | + grad_norm_pre_clip_avg=0.1175 | Metrics: + {'align_loss': 0.026279661804437637, + 'recon_loss': 0.12375682592391968, + 'predict_loss': 0.004403809551149607, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13680480420589447, + 'data_time': 0.0010021860070992261, + 'model_time': 1.3825706120114774, + 'grad_norm_pre_clip_avg': 0.11749620661139488, + 'learning_rate': 4.909686321764842e-08, + 'epoch': 12.39} +04/20 [05:15:36] INFO | >> train_qwenlatent.py:487 + Step 49120 | grad_norm_pre_clip=0.1040 | + grad_norm_pre_clip_avg=0.1135 | Metrics: + {'align_loss': 0.024420225992798805, + 'recon_loss': 0.15501709282398224, + 'predict_loss': 0.004141113255172968, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10402463376522064, + 'data_time': 0.0006549459940288216, + 'model_time': 1.216810219018953, + 'grad_norm_pre_clip_avg': 0.11345098316669464, + 'learning_rate': 4.8558572194617745e-08, + 'epoch': 12.39} +04/20 [05:15:49] INFO | >> train_qwenlatent.py:487 + Step 49130 | grad_norm_pre_clip=0.1312 | + grad_norm_pre_clip_avg=0.1131 | Metrics: + {'align_loss': 0.024011384695768356, + 'recon_loss': 0.1280214935541153, + 'predict_loss': 0.006944918539375067, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13120037317276, + 'data_time': 0.0009662150114309043, + 'model_time': 1.2749671759956982, + 'grad_norm_pre_clip_avg': 0.11306169256567955, + 'learning_rate': 4.802635594522751e-08, + 'epoch': 12.4} +04/20 [05:16:01] INFO | >> train_qwenlatent.py:487 + Step 49140 | grad_norm_pre_clip=0.1246 | + grad_norm_pre_clip_avg=0.1122 | Metrics: + {'align_loss': 0.024606022983789444, + 'recon_loss': 0.16822917759418488, + 'predict_loss': 0.008083030581474304, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12463109195232391, + 'data_time': 0.0009269849979318678, + 'model_time': 1.5316976640024222, + 'grad_norm_pre_clip_avg': 0.11219537109136582, + 'learning_rate': 4.75002147288701e-08, 'epoch': + 12.4} +04/20 [05:16:15] INFO | >> train_qwenlatent.py:487 + Step 49150 | grad_norm_pre_clip=0.1139 | + grad_norm_pre_clip_avg=0.1002 | Metrics: + {'align_loss': 0.025909103453159332, + 'recon_loss': 0.19272838532924652, + 'predict_loss': 0.005919117946177721, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1138797178864479, + 'mae_score': 0.006111340050224785, 'data_time': + 0.0007549969886895269, 'model_time': + 1.2275777139875572, 'grad_norm_pre_clip_avg': + 0.10019410327076912, 'learning_rate': + 4.698014880198069e-08, 'epoch': 12.4} +04/20 [05:16:28] INFO | >> train_qwenlatent.py:487 + Step 49160 | grad_norm_pre_clip=0.0964 | + grad_norm_pre_clip_avg=0.1124 | Metrics: + {'align_loss': 0.025777509436011314, + 'recon_loss': 0.1296929121017456, + 'predict_loss': 0.0020835448522120714, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09644033014774323, + 'data_time': 0.0006453179812524468, + 'model_time': 1.2023094919859432, + 'grad_norm_pre_clip_avg': 0.11240965351462365, + 'learning_rate': 4.646615841803593e-08, + 'epoch': 12.4} +04/20 [05:16:40] INFO | >> train_qwenlatent.py:487 + Step 49170 | grad_norm_pre_clip=0.0921 | + grad_norm_pre_clip_avg=0.1084 | Metrics: + {'align_loss': 0.02615712583065033, + 'recon_loss': 0.19299252331256866, + 'predict_loss': 0.005558487493544817, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09209705889225006, + 'data_time': 0.0006566999945789576, + 'model_time': 1.188631697994424, + 'grad_norm_pre_clip_avg': 0.10837896689772605, + 'learning_rate': 4.595824382754559e-08, + 'epoch': 12.41} +04/20 [05:16:52] INFO | >> train_qwenlatent.py:487 + Step 49180 | grad_norm_pre_clip=0.1300 | + grad_norm_pre_clip_avg=0.1109 | Metrics: + {'align_loss': 0.024564705789089203, + 'recon_loss': 0.13342376053333282, + 'predict_loss': 0.006119527854025364, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13001078367233276, + 'data_time': 0.0007440240005962551, + 'model_time': 1.2282650569977704, + 'grad_norm_pre_clip_avg': 0.11085414290428161, + 'learning_rate': 4.545640527806087e-08, + 'epoch': 12.41} +04/20 [05:17:05] INFO | >> train_qwenlatent.py:487 + Step 49190 | grad_norm_pre_clip=0.1028 | + grad_norm_pre_clip_avg=0.1145 | Metrics: + {'align_loss': 0.025590334087610245, + 'recon_loss': 0.15479575097560883, + 'predict_loss': 0.0038408013060688972, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10283027589321136, + 'data_time': 0.000678178999805823, + 'model_time': 1.2199262670183089, + 'grad_norm_pre_clip_avg': 0.1144896000623703, + 'learning_rate': 4.496064301417442e-08, + 'epoch': 12.41} +04/20 [05:17:18] INFO | >> train_qwenlatent.py:487 + Step 49200 | grad_norm_pre_clip=0.1104 | + grad_norm_pre_clip_avg=0.1094 | Metrics: + {'align_loss': 0.026257745921611786, + 'recon_loss': 0.16237099468708038, + 'predict_loss': 0.008433551527559757, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11044062674045563, + 'mae_score': 0.006313689549763997, 'data_time': + 0.0006952920230105519, 'model_time': + 1.259923010977218, 'grad_norm_pre_clip_avg': + 0.10936634540557862, 'learning_rate': + 4.447095727751199e-08, 'epoch': 12.41} +04/20 [05:17:31] INFO | >> train_qwenlatent.py:487 + Step 49210 | grad_norm_pre_clip=0.0917 | + grad_norm_pre_clip_avg=0.1222 | Metrics: + {'align_loss': 0.025991613045334816, + 'recon_loss': 0.20629587769508362, + 'predict_loss': 0.007332252338528633, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09173168241977692, + 'data_time': 0.000986748986179009, + 'model_time': 1.232755900011398, + 'grad_norm_pre_clip_avg': 0.12223686799407005, + 'learning_rate': 4.39873483067422e-08, 'epoch': + 12.42} +04/20 [05:17:43] INFO | >> train_qwenlatent.py:487 + Step 49220 | grad_norm_pre_clip=0.0847 | + grad_norm_pre_clip_avg=0.1133 | Metrics: + {'align_loss': 0.02609441801905632, + 'recon_loss': 0.18633191287517548, + 'predict_loss': 0.006490859668701887, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.08471372723579407, + 'data_time': 0.000731275009457022, + 'model_time': 1.2351130089955404, + 'grad_norm_pre_clip_avg': 0.11327960267663002, + 'learning_rate': 4.3509816337569534e-08, + 'epoch': 12.42} +04/20 [05:17:55] INFO | >> train_qwenlatent.py:487 + Step 49230 | grad_norm_pre_clip=0.0939 | + grad_norm_pre_clip_avg=0.1009 | Metrics: + {'align_loss': 0.02497003600001335, + 'recon_loss': 0.12408645451068878, + 'predict_loss': 0.0057859946973621845, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09390686452388763, + 'data_time': 0.0008491709886584431, + 'model_time': 1.2312406740093138, + 'grad_norm_pre_clip_avg': 0.10087366476655006, + 'learning_rate': 4.303836160273855e-08, + 'epoch': 12.42} +04/20 [05:18:08] INFO | >> train_qwenlatent.py:487 + Step 49240 | grad_norm_pre_clip=0.0743 | + grad_norm_pre_clip_avg=0.1102 | Metrics: + {'align_loss': 0.0244540236890316, + 'recon_loss': 0.1227814331650734, + 'predict_loss': 0.00551894074305892, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.07433731853961945, + 'data_time': 0.0008180229924619198, + 'model_time': 1.269996878982056, + 'grad_norm_pre_clip_avg': 0.11024781391024589, + 'learning_rate': 4.257298433202829e-08, + 'epoch': 12.42} +04/20 [05:18:21] INFO | >> train_qwenlatent.py:487 + Step 49250 | grad_norm_pre_clip=0.1376 | + grad_norm_pre_clip_avg=0.1217 | Metrics: + {'align_loss': 0.024713225662708282, + 'recon_loss': 0.15706683695316315, + 'predict_loss': 0.004762858152389526, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13755011558532715, + 'mae_score': 0.005282847086588541, 'data_time': + 0.0006762020057067275, 'model_time': + 1.2168898249801714, 'grad_norm_pre_clip_avg': + 0.12166431769728661, 'learning_rate': + 4.211368475225926e-08, 'epoch': 12.43} +04/20 [05:18:34] INFO | >> train_qwenlatent.py:487 + Step 49260 | grad_norm_pre_clip=0.0994 | + grad_norm_pre_clip_avg=0.1025 | Metrics: + {'align_loss': 0.02512584626674652, + 'recon_loss': 0.11919479072093964, + 'predict_loss': 0.00522643094882369, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09937696158885956, + 'data_time': 0.0006697530043311417, + 'model_time': 1.491879995999625, + 'grad_norm_pre_clip_avg': 0.10247446149587632, + 'learning_rate': 4.166046308728786e-08, + 'epoch': 12.43} +04/20 [05:18:47] INFO | >> train_qwenlatent.py:487 + Step 49270 | grad_norm_pre_clip=0.0833 | + grad_norm_pre_clip_avg=0.0979 | Metrics: + {'align_loss': 0.02663968876004219, + 'recon_loss': 0.14219030737876892, + 'predict_loss': 0.0062455106526613235, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.08334635198116302, + 'data_time': 0.001135450991569087, + 'model_time': 1.2185509669943713, + 'grad_norm_pre_clip_avg': 0.09786587059497834, + 'learning_rate': 4.121331955800916e-08, + 'epoch': 12.43} +04/20 [05:19:00] INFO | >> train_qwenlatent.py:487 + Step 49280 | grad_norm_pre_clip=0.1027 | + grad_norm_pre_clip_avg=0.1213 | Metrics: + {'align_loss': 0.02542291209101677, + 'recon_loss': 0.158469095826149, + 'predict_loss': 0.00526921171694994, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10269506275653839, + 'data_time': 0.0009316850046161562, + 'model_time': 1.508228977007093, + 'grad_norm_pre_clip_avg': 0.12125486433506012, + 'learning_rate': 4.077225438235688e-08, + 'epoch': 12.44} +04/20 [05:19:12] INFO | >> train_qwenlatent.py:487 + Step 49290 | grad_norm_pre_clip=0.1133 | + grad_norm_pre_clip_avg=0.1030 | Metrics: + {'align_loss': 0.02534462884068489, + 'recon_loss': 0.11782067269086838, + 'predict_loss': 0.0028307847678661346, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11326617002487183, + 'data_time': 0.0009294949995819479, + 'model_time': 1.2501190830080304, + 'grad_norm_pre_clip_avg': 0.10304349139332772, + 'learning_rate': 4.033726777529788e-08, + 'epoch': 12.44} +04/20 [05:19:26] INFO | >> train_qwenlatent.py:487 + Step 49300 | grad_norm_pre_clip=0.0893 | + grad_norm_pre_clip_avg=0.0952 | Metrics: + {'align_loss': 0.024530921131372452, + 'recon_loss': 0.14330747723579407, + 'predict_loss': 0.007542474661022425, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.0892910435795784, + 'mae_score': 0.004817076416702958, 'data_time': + 0.0009049460059031844, 'model_time': + 1.2706646739970893, 'grad_norm_pre_clip_avg': + 0.09515349045395852, 'learning_rate': + 3.990835994884324e-08, 'epoch': 12.44} +04/20 [05:19:38] INFO | >> train_qwenlatent.py:487 + Step 49310 | grad_norm_pre_clip=0.0751 | + grad_norm_pre_clip_avg=0.1104 | Metrics: + {'align_loss': 0.026335878297686577, + 'recon_loss': 0.22646737098693848, + 'predict_loss': 0.007578411605209112, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.07512790709733963, + 'data_time': 0.00078859698260203, 'model_time': + 1.313509489002172, 'grad_norm_pre_clip_avg': + 0.11039654240012169, 'learning_rate': + 3.9485531112032985e-08, 'epoch': 12.44} +04/20 [05:19:51] INFO | >> train_qwenlatent.py:487 + Step 49320 | grad_norm_pre_clip=0.1116 | + grad_norm_pre_clip_avg=0.1178 | Metrics: + {'align_loss': 0.024098824709653854, + 'recon_loss': 0.09817276895046234, + 'predict_loss': 0.005001309793442488, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11162453889846802, + 'data_time': 0.0006665660184808075, + 'model_time': 1.2511967380123679, + 'grad_norm_pre_clip_avg': 0.11776450462639332, + 'learning_rate': 3.9068781470951374e-08, + 'epoch': 12.45} +04/20 [05:20:03] INFO | >> train_qwenlatent.py:487 + Step 49330 | grad_norm_pre_clip=0.0993 | + grad_norm_pre_clip_avg=0.1120 | Metrics: + {'align_loss': 0.02483983151614666, + 'recon_loss': 0.12467552721500397, + 'predict_loss': 0.005061216652393341, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09925635159015656, + 'data_time': 0.000759418006055057, + 'model_time': 1.2476796599803492, + 'grad_norm_pre_clip_avg': 0.1119755633175373, + 'learning_rate': 3.865811122871855e-08, + 'epoch': 12.45} +04/20 [05:20:16] INFO | >> train_qwenlatent.py:487 + Step 49340 | grad_norm_pre_clip=0.0860 | + grad_norm_pre_clip_avg=0.1094 | Metrics: + {'align_loss': 0.025902969762682915, + 'recon_loss': 0.12004539370536804, + 'predict_loss': 0.0036279717460274696, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.08596443384885788, + 'data_time': 0.0006884259928483516, + 'model_time': 1.2511940879921895, + 'grad_norm_pre_clip_avg': 0.10939746201038361, + 'learning_rate': 3.825352058548639e-08, + 'epoch': 12.45} +04/20 [05:20:29] INFO | >> train_qwenlatent.py:487 + Step 49350 | grad_norm_pre_clip=0.1256 | + grad_norm_pre_clip_avg=0.1154 | Metrics: + {'align_loss': 0.0248803049325943, + 'recon_loss': 0.17294389009475708, + 'predict_loss': 0.0055145928636193275, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12558722496032715, + 'mae_score': 0.0054726063668190895, + 'data_time': 0.0008360810170415789, + 'model_time': 1.2316515300190076, + 'grad_norm_pre_clip_avg': 0.115401741117239, + 'learning_rate': 3.7855009738450996e-08, + 'epoch': 12.45} +04/20 [05:20:41] INFO | >> train_qwenlatent.py:487 + Step 49360 | grad_norm_pre_clip=0.0848 | + grad_norm_pre_clip_avg=0.1075 | Metrics: + {'align_loss': 0.025843512266874313, + 'recon_loss': 0.18319983780384064, + 'predict_loss': 0.0058334809727966785, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.08478421717882156, + 'data_time': 0.000788595003541559, + 'model_time': 1.2869479130022228, + 'grad_norm_pre_clip_avg': 0.10747349858283997, + 'learning_rate': 3.746257888183881e-08, + 'epoch': 12.46} +04/20 [05:20:54] INFO | >> train_qwenlatent.py:487 + Step 49370 | grad_norm_pre_clip=0.1068 | + grad_norm_pre_clip_avg=0.1165 | Metrics: + {'align_loss': 0.025567268952727318, + 'recon_loss': 0.10740688443183899, + 'predict_loss': 0.0036687555257230997, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10677256435155869, + 'data_time': 0.0008055699872784317, + 'model_time': 1.256626079994021, + 'grad_norm_pre_clip_avg': 0.1164567768573761, + 'learning_rate': 3.707622820691773e-08, + 'epoch': 12.46} +04/20 [05:21:06] INFO | >> train_qwenlatent.py:487 + Step 49380 | grad_norm_pre_clip=0.1050 | + grad_norm_pre_clip_avg=0.1142 | Metrics: + {'align_loss': 0.025970367714762688, + 'recon_loss': 0.1611010581254959, + 'predict_loss': 0.00968967005610466, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10495003312826157, + 'data_time': 0.0010753970127552748, + 'model_time': 1.2868497970048338, + 'grad_norm_pre_clip_avg': 0.11419169455766678, + 'learning_rate': 3.669595790199015e-08, + 'epoch': 12.46} +04/20 [05:21:19] INFO | >> train_qwenlatent.py:487 + Step 49390 | grad_norm_pre_clip=0.1021 | + grad_norm_pre_clip_avg=0.1037 | Metrics: + {'align_loss': 0.0256611667573452, + 'recon_loss': 0.15468892455101013, + 'predict_loss': 0.007645861711353064, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10206658393144608, + 'data_time': 0.0007932650041766465, + 'model_time': 1.542929101997288, + 'grad_norm_pre_clip_avg': 0.10366978347301484, + 'learning_rate': 3.6321768152394354e-08, + 'epoch': 12.46} +04/20 [05:21:32] INFO | >> train_qwenlatent.py:487 + Step 49400 | grad_norm_pre_clip=0.0967 | + grad_norm_pre_clip_avg=0.1252 | Metrics: + {'align_loss': 0.025146057829260826, + 'recon_loss': 0.1466260850429535, + 'predict_loss': 0.006425118073821068, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09668838232755661, + 'mae_score': 0.009604582915434966, 'data_time': + 0.001050566992489621, 'model_time': + 1.260615967999911, 'grad_norm_pre_clip_avg': + 0.1251985251903534, 'learning_rate': + 3.5953659140505944e-08, 'epoch': 12.47} +04/20 [05:21:45] INFO | >> train_qwenlatent.py:487 + Step 49410 | grad_norm_pre_clip=0.1194 | + grad_norm_pre_clip_avg=0.1103 | Metrics: + {'align_loss': 0.025539472699165344, + 'recon_loss': 0.11498237401247025, + 'predict_loss': 0.003928303252905607, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11944432556629181, + 'data_time': 0.0008043129928410053, + 'model_time': 1.2539165299967863, + 'grad_norm_pre_clip_avg': 0.11030197888612747, + 'learning_rate': 3.559163104573914e-08, + 'epoch': 12.47} +04/20 [05:21:58] INFO | >> train_qwenlatent.py:487 + Step 49420 | grad_norm_pre_clip=0.1337 | + grad_norm_pre_clip_avg=0.1230 | Metrics: + {'align_loss': 0.02546551078557968, + 'recon_loss': 0.12398406863212585, + 'predict_loss': 0.005121643655002117, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13369841873645782, + 'data_time': 0.0009531429968774319, + 'model_time': 1.2309418459772132, + 'grad_norm_pre_clip_avg': 0.12295986115932464, + 'learning_rate': 3.523568404453856e-08, + 'epoch': 12.47} +04/20 [05:22:11] INFO | >> train_qwenlatent.py:487 + Step 49430 | grad_norm_pre_clip=0.1080 | + grad_norm_pre_clip_avg=0.1114 | Metrics: + {'align_loss': 0.026151904836297035, + 'recon_loss': 0.226677805185318, + 'predict_loss': 0.011270212009549141, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10799385607242584, + 'data_time': 0.0011069680040236562, + 'model_time': 1.2677895359811373, + 'grad_norm_pre_clip_avg': 0.11143733859062195, + 'learning_rate': 3.488581831039022e-08, + 'epoch': 12.47} +04/20 [05:22:23] INFO | >> train_qwenlatent.py:487 + Step 49440 | grad_norm_pre_clip=0.0807 | + grad_norm_pre_clip_avg=0.1183 | Metrics: + {'align_loss': 0.026323888450860977, + 'recon_loss': 0.13386917114257812, + 'predict_loss': 0.0034581476356834173, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.08068802952766418, + 'data_time': 0.000721744989277795, + 'model_time': 1.302332129998831, + 'grad_norm_pre_clip_avg': 0.11834659203886985, + 'learning_rate': 3.4542034013814694e-08, + 'epoch': 12.48} +04/20 [05:22:36] INFO | >> train_qwenlatent.py:487 + Step 49450 | grad_norm_pre_clip=0.0756 | + grad_norm_pre_clip_avg=0.1123 | Metrics: + {'align_loss': 0.02611776813864708, + 'recon_loss': 0.1552077829837799, + 'predict_loss': 0.007161852438002825, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.07564794272184372, + 'mae_score': 0.0053960671295990815, + 'data_time': 0.0007187850133050233, + 'model_time': 1.2608250610064715, + 'grad_norm_pre_clip_avg': 0.11231844797730446, + 'learning_rate': 3.42043313223684e-08, 'epoch': + 12.48} +04/20 [05:22:49] INFO | >> train_qwenlatent.py:487 + Step 49460 | grad_norm_pre_clip=0.1247 | + grad_norm_pre_clip_avg=0.1170 | Metrics: + {'align_loss': 0.026060868054628372, + 'recon_loss': 0.1724798083305359, + 'predict_loss': 0.005652131512761116, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1246500015258789, + 'data_time': 0.0008661559841129929, + 'model_time': 1.264938403997803, + 'grad_norm_pre_clip_avg': 0.11701374724507332, + 'learning_rate': 3.387271040064231e-08, + 'epoch': 12.48} +04/20 [05:23:02] INFO | >> train_qwenlatent.py:487 + Step 49470 | grad_norm_pre_clip=0.1007 | + grad_norm_pre_clip_avg=0.1087 | Metrics: + {'align_loss': 0.024586470797657967, + 'recon_loss': 0.1600991040468216, + 'predict_loss': 0.005385322961956263, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10074786096811295, + 'data_time': 0.0007439389883074909, + 'model_time': 1.2666607559949625, + 'grad_norm_pre_clip_avg': 0.10872191190719604, + 'learning_rate': 3.3547171410264644e-08, + 'epoch': 12.48} +04/20 [05:23:14] INFO | >> train_qwenlatent.py:487 + Step 49480 | grad_norm_pre_clip=0.1470 | + grad_norm_pre_clip_avg=0.1034 | Metrics: + {'align_loss': 0.024080969393253326, + 'recon_loss': 0.19495102763175964, + 'predict_loss': 0.007908843457698822, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.14699609577655792, + 'data_time': 0.0008167909982148558, + 'model_time': 1.2675490579858888, + 'grad_norm_pre_clip_avg': 0.10342617370188237, + 'learning_rate': 3.3227714509899553e-08, + 'epoch': 12.49} +04/20 [05:23:27] INFO | >> train_qwenlatent.py:487 + Step 49490 | grad_norm_pre_clip=0.1561 | + grad_norm_pre_clip_avg=0.1029 | Metrics: + {'align_loss': 0.023705316707491875, + 'recon_loss': 0.09309530258178711, + 'predict_loss': 0.003031616797670722, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15608109533786774, + 'data_time': 0.0006836880056653172, + 'model_time': 1.2396065240027383, + 'grad_norm_pre_clip_avg': 0.10291673764586448, + 'learning_rate': 3.291433985524705e-08, + 'epoch': 12.49} +04/20 [05:23:40] INFO | >> train_qwenlatent.py:487 + Step 49500 | grad_norm_pre_clip=0.1227 | + grad_norm_pre_clip_avg=0.1213 | Metrics: + {'align_loss': 0.025218596681952477, + 'recon_loss': 0.14712251722812653, + 'predict_loss': 0.004143597092479467, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1226549744606018, + 'mae_score': 0.00548146222088788, 'data_time': + 0.0009203369845636189, 'model_time': + 1.257172662008088, 'grad_norm_pre_clip_avg': + 0.1213226594030857, 'learning_rate': + 3.260704759904167e-08, 'epoch': 12.49} +04/20 [05:23:53] INFO | >> train_qwenlatent.py:487 + Step 49510 | grad_norm_pre_clip=0.0948 | + grad_norm_pre_clip_avg=0.1040 | Metrics: + {'align_loss': 0.02594679221510887, + 'recon_loss': 0.17888808250427246, + 'predict_loss': 0.008886290714144707, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09479423612356186, + 'data_time': 0.000666711013764143, + 'model_time': 1.2160224879917223, + 'grad_norm_pre_clip_avg': 0.10403363108634948, + 'learning_rate': 3.230583789105386e-08, + 'epoch': 12.49} +04/20 [05:24:05] INFO | >> train_qwenlatent.py:487 + Step 49520 | grad_norm_pre_clip=0.0918 | + grad_norm_pre_clip_avg=0.1023 | Metrics: + {'align_loss': 0.025106240063905716, + 'recon_loss': 0.12381184846162796, + 'predict_loss': 0.003968058153986931, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09184146672487259, + 'data_time': 0.000685397011693567, + 'model_time': 1.244074726011604, + 'grad_norm_pre_clip_avg': 0.10234214439988136, + 'learning_rate': 3.201071087808994e-08, + 'epoch': 12.5} +04/20 [05:24:18] INFO | >> train_qwenlatent.py:487 + Step 49530 | grad_norm_pre_clip=0.0975 | + grad_norm_pre_clip_avg=0.1200 | Metrics: + {'align_loss': 0.025511760264635086, + 'recon_loss': 0.19947490096092224, + 'predict_loss': 0.009211020544171333, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09747807681560516, + 'data_time': 0.0008631550008431077, + 'model_time': 1.2461626500007696, + 'grad_norm_pre_clip_avg': 0.12001843973994256, + 'learning_rate': 3.172166670399076e-08, + 'epoch': 12.5} +04/20 [05:24:31] INFO | >> train_qwenlatent.py:487 + Step 49540 | grad_norm_pre_clip=0.1069 | + grad_norm_pre_clip_avg=0.1178 | Metrics: + {'align_loss': 0.02445325255393982, + 'recon_loss': 0.14576373994350433, + 'predict_loss': 0.007004746235907078, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10687322914600372, + 'data_time': 0.0009244010143447667, + 'model_time': 1.1831042669946328, + 'grad_norm_pre_clip_avg': 0.11779796853661537, + 'learning_rate': 3.1438705509631656e-08, + 'epoch': 12.5} +04/20 [05:24:44] INFO | >> train_qwenlatent.py:487 + Step 49550 | grad_norm_pre_clip=0.1520 | + grad_norm_pre_clip_avg=0.1164 | Metrics: + {'align_loss': 0.02521102875471115, + 'recon_loss': 0.1664486527442932, + 'predict_loss': 0.005735411308705807, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.15198618173599243, + 'mae_score': 0.0046779228760315495, + 'data_time': 0.0007745000184513628, + 'model_time': 1.2412132139725145, + 'grad_norm_pre_clip_avg': 0.11639367267489434, + 'learning_rate': 3.1161827432926645e-08, + 'epoch': 12.5} +04/20 [05:24:57] INFO | >> train_qwenlatent.py:487 + Step 49560 | grad_norm_pre_clip=0.1282 | + grad_norm_pre_clip_avg=0.1164 | Metrics: + {'align_loss': 0.02606646530330181, + 'recon_loss': 0.16785778105258942, + 'predict_loss': 0.005756433121860027, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12821301817893982, + 'data_time': 0.000709676998667419, + 'model_time': 1.2062405059987213, + 'grad_norm_pre_clip_avg': 0.11641676276922226, + 'learning_rate': 3.089103260882288e-08, + 'epoch': 12.51} +04/20 [05:25:09] INFO | >> train_qwenlatent.py:487 + Step 49570 | grad_norm_pre_clip=0.0927 | + grad_norm_pre_clip_avg=0.0990 | Metrics: + {'align_loss': 0.025282494723796844, + 'recon_loss': 0.17002461850643158, + 'predict_loss': 0.0079495869576931, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09271105378866196, + 'data_time': 0.0009662420197855681, + 'model_time': 1.2600839179940522, + 'grad_norm_pre_clip_avg': 0.09902633354067802, + 'learning_rate': 3.0626321169300617e-08, + 'epoch': 12.51} +04/20 [05:25:22] INFO | >> train_qwenlatent.py:487 + Step 49580 | grad_norm_pre_clip=0.1270 | + grad_norm_pre_clip_avg=0.1155 | Metrics: + {'align_loss': 0.02637169137597084, + 'recon_loss': 0.14640958607196808, + 'predict_loss': 0.010238953866064548, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12701775133609772, + 'data_time': 0.0007772780081722885, + 'model_time': 1.216996889008442, + 'grad_norm_pre_clip_avg': 0.11554400622844696, + 'learning_rate': 3.0367693243377403e-08, + 'epoch': 12.51} +04/20 [05:25:34] INFO | >> train_qwenlatent.py:487 + Step 49590 | grad_norm_pre_clip=0.1355 | + grad_norm_pre_clip_avg=0.1095 | Metrics: + {'align_loss': 0.024222254753112793, + 'recon_loss': 0.16780440509319305, + 'predict_loss': 0.004557284060865641, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13549864292144775, + 'data_time': 0.0007716749969404191, + 'model_time': 1.1976716839999426, + 'grad_norm_pre_clip_avg': 0.10946424528956414, + 'learning_rate': 3.011514895710669e-08, + 'epoch': 12.51} +04/20 [05:25:47] INFO | >> train_qwenlatent.py:487 + Step 49600 | grad_norm_pre_clip=0.0945 | + grad_norm_pre_clip_avg=0.1075 | Metrics: + {'align_loss': 0.025727681815624237, + 'recon_loss': 0.1709844321012497, + 'predict_loss': 0.008202602155506611, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09448522329330444, + 'mae_score': 0.0052594408258661495, + 'data_time': 0.0011005440028384328, + 'model_time': 1.205878856999334, + 'grad_norm_pre_clip_avg': 0.10745075866580009, + 'learning_rate': 2.986868843357503e-08, + 'epoch': 12.52} +04/20 [05:26:00] INFO | >> train_qwenlatent.py:487 + Step 49610 | grad_norm_pre_clip=0.0995 | + grad_norm_pre_clip_avg=0.1136 | Metrics: + {'align_loss': 0.025791937485337257, + 'recon_loss': 0.12483127415180206, + 'predict_loss': 0.003613098291680217, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.0994548574090004, + 'data_time': 0.0007816349971108139, + 'model_time': 1.2349998829886317, + 'grad_norm_pre_clip_avg': 0.11358914673328399, + 'learning_rate': 2.9628311792903504e-08, + 'epoch': 12.52} +04/20 [05:26:12] INFO | >> train_qwenlatent.py:487 + Step 49620 | grad_norm_pre_clip=0.1285 | + grad_norm_pre_clip_avg=0.1251 | Metrics: + {'align_loss': 0.025953803211450577, + 'recon_loss': 0.1599697470664978, + 'predict_loss': 0.005632080603390932, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12849649786949158, + 'data_time': 0.0007919339986983687, + 'model_time': 1.2593942059902474, + 'grad_norm_pre_clip_avg': 0.1250609837472439, + 'learning_rate': 2.939401915224908e-08, + 'epoch': 12.52} +04/20 [05:26:24] INFO | >> train_qwenlatent.py:487 + Step 49630 | grad_norm_pre_clip=0.0972 | + grad_norm_pre_clip_avg=0.1196 | Metrics: + {'align_loss': 0.026660427451133728, + 'recon_loss': 0.1614198535680771, + 'predict_loss': 0.007312309928238392, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09716754406690598, + 'data_time': 0.0006657240155618638, + 'model_time': 1.2218285189883318, + 'grad_norm_pre_clip_avg': 0.11958152875304222, + 'learning_rate': 2.9165810625803254e-08, + 'epoch': 12.52} +04/20 [05:26:37] INFO | >> train_qwenlatent.py:487 + Step 49640 | grad_norm_pre_clip=0.0974 | + grad_norm_pre_clip_avg=0.1273 | Metrics: + {'align_loss': 0.026085248216986656, + 'recon_loss': 0.18951961398124695, + 'predict_loss': 0.008023078553378582, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.0973992720246315, + 'data_time': 0.0008963279833551496, + 'model_time': 1.2469871369830798, + 'grad_norm_pre_clip_avg': 0.12734348550438881, + 'learning_rate': 2.8943686324792e-08, 'epoch': + 12.53} +04/20 [05:26:50] INFO | >> train_qwenlatent.py:487 + Step 49650 | grad_norm_pre_clip=0.1060 | + grad_norm_pre_clip_avg=0.1111 | Metrics: + {'align_loss': 0.02385122701525688, + 'recon_loss': 0.1260877549648285, + 'predict_loss': 0.004625506233423948, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10600743442773819, + 'mae_score': 0.005985595084525444, 'data_time': + 0.0009581670165061951, 'model_time': + 1.2416720939800143, 'grad_norm_pre_clip_avg': + 0.11114066541194915, 'learning_rate': + 2.872764635747583e-08, 'epoch': 12.53} +04/20 [05:27:03] INFO | >> train_qwenlatent.py:487 + Step 49660 | grad_norm_pre_clip=0.1229 | + grad_norm_pre_clip_avg=0.1134 | Metrics: + {'align_loss': 0.025634124875068665, + 'recon_loss': 0.1570790410041809, + 'predict_loss': 0.003522187937051058, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12293767184019089, + 'data_time': 0.0010208899911958724, + 'model_time': 1.1976122519990895, + 'grad_norm_pre_clip_avg': 0.11342898309230805, + 'learning_rate': 2.8517690829149747e-08, + 'epoch': 12.53} +04/20 [05:27:15] INFO | >> train_qwenlatent.py:487 + Step 49670 | grad_norm_pre_clip=0.0862 | + grad_norm_pre_clip_avg=0.1128 | Metrics: + {'align_loss': 0.025807656347751617, + 'recon_loss': 0.1611221432685852, + 'predict_loss': 0.004085153806954622, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.0862259641289711, + 'data_time': 0.0007304520113393664, + 'model_time': 1.2341459039889742, + 'grad_norm_pre_clip_avg': 0.11276023611426353, + 'learning_rate': 2.831381984214466e-08, + 'epoch': 12.53} +04/20 [05:27:28] INFO | >> train_qwenlatent.py:487 + Step 49680 | grad_norm_pre_clip=0.1181 | + grad_norm_pre_clip_avg=0.1096 | Metrics: + {'align_loss': 0.024644939228892326, + 'recon_loss': 0.12033829092979431, + 'predict_loss': 0.003998777363449335, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11812169849872589, + 'data_time': 0.0008493379864376038, + 'model_time': 1.2862451130058616, + 'grad_norm_pre_clip_avg': 0.10962292179465294, + 'learning_rate': 2.8116033495824598e-08, + 'epoch': 12.54} +04/20 [05:27:41] INFO | >> train_qwenlatent.py:487 + Step 49690 | grad_norm_pre_clip=0.1129 | + grad_norm_pre_clip_avg=0.1081 | Metrics: + {'align_loss': 0.023933302611112595, + 'recon_loss': 0.15386708080768585, + 'predict_loss': 0.004363381303846836, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11287154257297516, + 'data_time': 0.0006894909893162549, + 'model_time': 1.2470261449925601, + 'grad_norm_pre_clip_avg': 0.10811711698770524, + 'learning_rate': 2.7924331886588098e-08, + 'epoch': 12.54} +04/20 [05:27:54] INFO | >> train_qwenlatent.py:487 + Step 49700 | grad_norm_pre_clip=0.1805 | + grad_norm_pre_clip_avg=0.1083 | Metrics: + {'align_loss': 0.025489211082458496, + 'recon_loss': 0.1683053821325302, + 'predict_loss': 0.006146752275526524, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.18049632012844086, + 'mae_score': 0.005280882173830325, 'data_time': + 0.0009243260137736797, 'model_time': + 1.2514671709795948, 'grad_norm_pre_clip_avg': + 0.10832096338272094, 'learning_rate': + 2.7738715107866823e-08, 'epoch': 12.54} +04/20 [05:28:07] INFO | >> train_qwenlatent.py:487 + Step 49710 | grad_norm_pre_clip=0.0768 | + grad_norm_pre_clip_avg=0.1067 | Metrics: + {'align_loss': 0.026491668075323105, + 'recon_loss': 0.1536329686641693, + 'predict_loss': 0.00548976194113493, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.07676734775304794, + 'data_time': 0.0008033390040509403, + 'model_time': 1.1795303830003832, + 'grad_norm_pre_clip_avg': 0.10673387572169304, + 'learning_rate': 2.7559183250129724e-08, + 'epoch': 12.54} +04/20 [05:28:19] INFO | >> train_qwenlatent.py:487 + Step 49720 | grad_norm_pre_clip=0.0849 | + grad_norm_pre_clip_avg=0.1211 | Metrics: + {'align_loss': 0.025177758187055588, + 'recon_loss': 0.17862030863761902, + 'predict_loss': 0.00471374299377203, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.08493995666503906, + 'data_time': 0.0007005910156294703, + 'model_time': 1.2203866020136047, + 'grad_norm_pre_clip_avg': 0.12112079262733459, + 'learning_rate': 2.7385736400877476e-08, + 'epoch': 12.55} +04/20 [05:28:32] INFO | >> train_qwenlatent.py:487 + Step 49730 | grad_norm_pre_clip=0.0697 | + grad_norm_pre_clip_avg=0.0920 | Metrics: + {'align_loss': 0.025170091539621353, + 'recon_loss': 0.16094328463077545, + 'predict_loss': 0.005381350871175528, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.0697125568985939, + 'data_time': 0.0006815939850639552, + 'model_time': 1.5252077449986245, + 'grad_norm_pre_clip_avg': 0.09199356138706208, + 'learning_rate': 2.721837464464805e-08, + 'epoch': 12.55} +04/20 [05:28:44] INFO | >> train_qwenlatent.py:487 + Step 49740 | grad_norm_pre_clip=0.1298 | + grad_norm_pre_clip_avg=0.1112 | Metrics: + {'align_loss': 0.025356195867061615, + 'recon_loss': 0.17333294451236725, + 'predict_loss': 0.0052551692351698875, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12981639802455902, + 'data_time': 0.0007218469982035458, + 'model_time': 1.2201050680014305, + 'grad_norm_pre_clip_avg': 0.11115812212228775, + 'learning_rate': 2.7057098063008362e-08, + 'epoch': 12.55} +04/20 [05:28:57] INFO | >> train_qwenlatent.py:487 + Step 49750 | grad_norm_pre_clip=0.1350 | + grad_norm_pre_clip_avg=0.1145 | Metrics: + {'align_loss': 0.024218330159783363, + 'recon_loss': 0.11061307042837143, + 'predict_loss': 0.0030946736223995686, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13498416543006897, + 'mae_score': 0.00491037841315742, 'data_time': + 0.000656655989587307, 'model_time': + 1.2000773510080762, 'grad_norm_pre_clip_avg': + 0.11448184102773666, 'learning_rate': + 2.690190673456402e-08, 'epoch': 12.55} +04/20 [05:29:10] INFO | >> train_qwenlatent.py:487 + Step 49760 | grad_norm_pre_clip=0.1069 | + grad_norm_pre_clip_avg=0.1138 | Metrics: + {'align_loss': 0.02633741870522499, + 'recon_loss': 0.15975672006607056, + 'predict_loss': 0.0049945819191634655, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10691406577825546, + 'data_time': 0.0006868279888294637, + 'model_time': 1.2548238470044453, + 'grad_norm_pre_clip_avg': 0.11378694698214531, + 'learning_rate': 2.6752800734955133e-08, + 'epoch': 12.56} +04/20 [05:29:22] INFO | >> train_qwenlatent.py:487 + Step 49770 | grad_norm_pre_clip=0.1518 | + grad_norm_pre_clip_avg=0.1157 | Metrics: + {'align_loss': 0.02459745667874813, + 'recon_loss': 0.09406623244285583, + 'predict_loss': 0.0049598668701946735, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1517583578824997, + 'data_time': 0.0009647029801271856, + 'model_time': 1.2112685159954708, + 'grad_norm_pre_clip_avg': 0.11572166532278061, + 'learning_rate': 2.6609780136850756e-08, + 'epoch': 12.56} +04/20 [05:29:34] INFO | >> train_qwenlatent.py:487 + Step 49780 | grad_norm_pre_clip=0.1362 | + grad_norm_pre_clip_avg=0.1123 | Metrics: + {'align_loss': 0.02598894014954567, + 'recon_loss': 0.20339594781398773, + 'predict_loss': 0.010350536555051804, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13618069887161255, + 'data_time': 0.0008497659873683006, + 'model_time': 1.2304469000082463, + 'grad_norm_pre_clip_avg': 0.11232243701815606, + 'learning_rate': 2.6472845009960034e-08, + 'epoch': 12.56} +04/20 [05:29:46] INFO | >> train_qwenlatent.py:487 + Step 49790 | grad_norm_pre_clip=0.0782 | + grad_norm_pre_clip_avg=0.1159 | Metrics: + {'align_loss': 0.02466406673192978, + 'recon_loss': 0.16705070436000824, + 'predict_loss': 0.0060943858698010445, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.07817861437797546, + 'data_time': 0.0006494020053651184, + 'model_time': 1.1906467690132558, + 'grad_norm_pre_clip_avg': 0.11587783768773079, + 'learning_rate': 2.634199542102243e-08, + 'epoch': 12.56} +04/20 [05:30:00] INFO | >> train_qwenlatent.py:487 + Step 49800 | grad_norm_pre_clip=0.1239 | + grad_norm_pre_clip_avg=0.1154 | Metrics: + {'align_loss': 0.024286452680826187, + 'recon_loss': 0.11060860753059387, + 'predict_loss': 0.004716529510915279, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12390030920505524, + 'mae_score': 0.0050791912250690635, + 'data_time': 0.0007753390236757696, + 'model_time': 1.2329175309860148, + 'grad_norm_pre_clip_avg': 0.11539646163582802, + 'learning_rate': 2.621723143381332e-08, + 'epoch': 12.57} +04/20 [05:30:12] INFO | >> train_qwenlatent.py:487 + Step 49810 | grad_norm_pre_clip=0.1201 | + grad_norm_pre_clip_avg=0.1193 | Metrics: + {'align_loss': 0.02429119125008583, + 'recon_loss': 0.11344660073518753, + 'predict_loss': 0.0032739832531660795, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12007605284452438, + 'data_time': 0.0006700930243823677, + 'model_time': 1.2734156130172778, + 'grad_norm_pre_clip_avg': 0.11934138089418411, + 'learning_rate': 2.60985531091412e-08, 'epoch': + 12.57} +04/20 [05:30:25] INFO | >> train_qwenlatent.py:487 + Step 49820 | grad_norm_pre_clip=0.0951 | + grad_norm_pre_clip_avg=0.1170 | Metrics: + {'align_loss': 0.026217572391033173, + 'recon_loss': 0.17580144107341766, + 'predict_loss': 0.004773537162691355, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09512409567832947, + 'data_time': 0.0005825920088682324, + 'model_time': 1.2179009299725294, + 'grad_norm_pre_clip_avg': 0.11698313727974892, + 'learning_rate': 2.5985960504846303e-08, + 'epoch': 12.57} +04/20 [05:30:38] INFO | >> train_qwenlatent.py:487 + Step 49830 | grad_norm_pre_clip=0.1165 | + grad_norm_pre_clip_avg=0.1122 | Metrics: + {'align_loss': 0.025444500148296356, + 'recon_loss': 0.11290421336889267, + 'predict_loss': 0.002927387598901987, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1165214478969574, + 'data_time': 0.0010379219893366098, + 'model_time': 1.2457236400223337, + 'grad_norm_pre_clip_avg': 0.11216467171907425, + 'learning_rate': 2.587945367580615e-08, + 'epoch': 12.57} +04/20 [05:30:51] INFO | >> train_qwenlatent.py:487 + Step 49840 | grad_norm_pre_clip=0.1262 | + grad_norm_pre_clip_avg=0.1186 | Metrics: + {'align_loss': 0.02430439367890358, + 'recon_loss': 0.15720148384571075, + 'predict_loss': 0.006919915787875652, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12620532512664795, + 'data_time': 0.0009203530207742006, + 'model_time': 1.2657597950019408, + 'grad_norm_pre_clip_avg': 0.11856357604265214, + 'learning_rate': 2.5779032673932754e-08, + 'epoch': 12.58} +04/20 [05:31:04] INFO | >> train_qwenlatent.py:487 + Step 49850 | grad_norm_pre_clip=0.1202 | + grad_norm_pre_clip_avg=0.0993 | Metrics: + {'align_loss': 0.02617913857102394, + 'recon_loss': 0.10235488414764404, + 'predict_loss': 0.002362130908295512, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12022484838962555, + 'mae_score': 0.005393165296262449, 'data_time': + 0.0006786120065953583, 'model_time': + 1.2325146080111153, 'grad_norm_pre_clip_avg': + 0.09927091524004936, 'learning_rate': + 2.5684697548167113e-08, 'epoch': 12.58} +04/20 [05:31:16] INFO | >> train_qwenlatent.py:487 + Step 49860 | grad_norm_pre_clip=0.1209 | + grad_norm_pre_clip_avg=0.1117 | Metrics: + {'align_loss': 0.024704022333025932, + 'recon_loss': 0.11254488676786423, + 'predict_loss': 0.004346324596554041, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.12092379480600357, + 'data_time': 0.0007458209875039756, + 'model_time': 1.2399972289858852, + 'grad_norm_pre_clip_avg': 0.11171407029032707, + 'learning_rate': 2.5596448344488882e-08, + 'epoch': 12.58} +04/20 [05:31:29] INFO | >> train_qwenlatent.py:487 + Step 49870 | grad_norm_pre_clip=0.1172 | + grad_norm_pre_clip_avg=0.1068 | Metrics: + {'align_loss': 0.02351686917245388, + 'recon_loss': 0.1609310358762741, + 'predict_loss': 0.007302251644432545, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11724478006362915, + 'data_time': 0.0010267390171065927, + 'model_time': 1.2318619400030002, + 'grad_norm_pre_clip_avg': 0.10681010410189629, + 'learning_rate': 2.5514285105909454e-08, + 'epoch': 12.58} +04/20 [05:31:41] INFO | >> train_qwenlatent.py:487 + Step 49880 | grad_norm_pre_clip=0.0972 | + grad_norm_pre_clip_avg=0.1082 | Metrics: + {'align_loss': 0.024595588445663452, + 'recon_loss': 0.14083530008792877, + 'predict_loss': 0.007719158660620451, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09720753878355026, + 'data_time': 0.0007870210101827979, + 'model_time': 1.1922743500035722, + 'grad_norm_pre_clip_avg': 0.10823253095149994, + 'learning_rate': 2.5438207872473343e-08, + 'epoch': 12.59} +04/20 [05:31:53] INFO | >> train_qwenlatent.py:487 + Step 49890 | grad_norm_pre_clip=0.1378 | + grad_norm_pre_clip_avg=0.1059 | Metrics: + {'align_loss': 0.025970475748181343, + 'recon_loss': 0.25881990790367126, + 'predict_loss': 0.011676664464175701, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.13784976303577423, + 'data_time': 0.0006983889907132834, + 'model_time': 1.2178663639933802, + 'grad_norm_pre_clip_avg': 0.1058701291680336, + 'learning_rate': 2.536821668125959e-08, + 'epoch': 12.59} +04/20 [05:32:07] INFO | >> train_qwenlatent.py:487 + Step 49900 | grad_norm_pre_clip=0.0704 | + grad_norm_pre_clip_avg=0.1085 | Metrics: + {'align_loss': 0.023247651755809784, + 'recon_loss': 0.1575799137353897, + 'predict_loss': 0.007827424444258213, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.07041028887033463, + 'mae_score': 0.0051808060826482, 'data_time': + 0.0007031010172795504, 'model_time': + 1.2181613820139319, 'grad_norm_pre_clip_avg': + 0.10853498876094818, 'learning_rate': + 2.53043115663831e-08, 'epoch': 12.59} +04/20 [05:32:19] INFO | >> train_qwenlatent.py:487 + Step 49910 | grad_norm_pre_clip=0.1046 | + grad_norm_pre_clip_avg=0.1033 | Metrics: + {'align_loss': 0.025678887963294983, + 'recon_loss': 0.19697415828704834, + 'predict_loss': 0.007058341987431049, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10458690673112869, + 'data_time': 0.000714202004019171, + 'model_time': 1.2510937970073428, + 'grad_norm_pre_clip_avg': 0.1032927855849266, + 'learning_rate': 2.5246492558987775e-08, + 'epoch': 12.59} +04/20 [05:32:31] INFO | >> train_qwenlatent.py:487 + Step 49920 | grad_norm_pre_clip=0.0789 | + grad_norm_pre_clip_avg=0.1140 | Metrics: + {'align_loss': 0.025398124009370804, + 'recon_loss': 0.12234395742416382, + 'predict_loss': 0.0042433179914951324, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.07887616008520126, + 'data_time': 0.0009473059908486903, + 'model_time': 1.218350515991915, + 'grad_norm_pre_clip_avg': 0.11395905017852784, + 'learning_rate': 2.519475968725478e-08, + 'epoch': 12.6} +04/20 [05:32:44] INFO | >> train_qwenlatent.py:487 + Step 49930 | grad_norm_pre_clip=0.1067 | + grad_norm_pre_clip_avg=0.1100 | Metrics: + {'align_loss': 0.026006406173110008, + 'recon_loss': 0.17543838918209076, + 'predict_loss': 0.006887827068567276, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10674741119146347, + 'data_time': 0.0011606490006670356, + 'model_time': 1.4539916790090501, + 'grad_norm_pre_clip_avg': 0.10998737588524818, + 'learning_rate': 2.514911297639979e-08, + 'epoch': 12.6} +04/20 [05:32:57] INFO | >> train_qwenlatent.py:487 + Step 49940 | grad_norm_pre_clip=0.1162 | + grad_norm_pre_clip_avg=0.1109 | Metrics: + {'align_loss': 0.024538878351449966, + 'recon_loss': 0.07272785902023315, + 'predict_loss': 0.002779371337965131, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11619614064693451, + 'data_time': 0.0009642399963922799, + 'model_time': 1.2186621330038179, + 'grad_norm_pre_clip_avg': 0.11090152338147163, + 'learning_rate': 2.5109552448668843e-08, + 'epoch': 12.6} +04/20 [05:33:11] INFO | >> train_qwenlatent.py:487 + Step 49950 | grad_norm_pre_clip=0.0890 | + grad_norm_pre_clip_avg=0.1037 | Metrics: + {'align_loss': 0.025215307250618935, + 'recon_loss': 0.12063205987215042, + 'predict_loss': 0.003496733261272311, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.08896488696336746, + 'mae_score': 0.005642710075722084, 'data_time': + 0.0008905259892344475, 'model_time': + 1.1998132219887339, 'grad_norm_pre_clip_avg': + 0.10368346869945526, 'learning_rate': + 2.507607812334247e-08, 'epoch': 12.6} +04/20 [05:33:24] INFO | >> train_qwenlatent.py:487 + Step 49960 | grad_norm_pre_clip=0.1074 | + grad_norm_pre_clip_avg=0.1107 | Metrics: + {'align_loss': 0.02672354131937027, + 'recon_loss': 0.16438059508800507, + 'predict_loss': 0.005573250353336334, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.1073794960975647, + 'data_time': 0.0009643039957154542, + 'model_time': 1.2594944209849928, + 'grad_norm_pre_clip_avg': 0.11073346287012101, + 'learning_rate': 2.5048690016737118e-08, + 'epoch': 12.61} +04/20 [05:33:36] INFO | >> train_qwenlatent.py:487 + Step 49970 | grad_norm_pre_clip=0.1029 | + grad_norm_pre_clip_avg=0.0986 | Metrics: + {'align_loss': 0.024480801075696945, + 'recon_loss': 0.1890275627374649, + 'predict_loss': 0.005398827604949474, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.10287277400493622, + 'data_time': 0.0009130350081250072, + 'model_time': 1.2268865900114179, + 'grad_norm_pre_clip_avg': 0.09858656600117684, + 'learning_rate': 2.5027388142200948e-08, + 'epoch': 12.61} +04/20 [05:33:49] INFO | >> train_qwenlatent.py:487 + Step 49980 | grad_norm_pre_clip=0.0913 | + grad_norm_pre_clip_avg=0.1060 | Metrics: + {'align_loss': 0.02621426060795784, + 'recon_loss': 0.16789056360721588, + 'predict_loss': 0.007096967659890652, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09129556268453598, + 'data_time': 0.0009854680101852864, + 'model_time': 1.179596721980488, + 'grad_norm_pre_clip_avg': 0.10601205006241798, + 'learning_rate': 2.5012172510116647e-08, + 'epoch': 12.61} +04/20 [05:34:01] INFO | >> train_qwenlatent.py:487 + Step 49990 | grad_norm_pre_clip=0.1181 | + grad_norm_pre_clip_avg=0.1029 | Metrics: + {'align_loss': 0.025657162070274353, + 'recon_loss': 0.1636863648891449, + 'predict_loss': 0.005945265758782625, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.11814115196466446, + 'data_time': 0.0009718960209283978, + 'model_time': 1.2412878519971855, + 'grad_norm_pre_clip_avg': 0.10286298505961895, + 'learning_rate': 2.5003043127900026e-08, + 'epoch': 12.61} +04/20 [05:34:14] INFO | >> train_qwenlatent.py:487 + Step 50000 | grad_norm_pre_clip=0.0967 | + grad_norm_pre_clip_avg=0.1100 | Metrics: + {'align_loss': 0.026379115879535675, + 'recon_loss': 0.12406477332115173, + 'predict_loss': 0.002600160427391529, + 'aux_loss_decay_weight': 0.0, + 'grad_norm_pre_clip': 0.09673331677913666, + 'mae_score': 0.004917230691995706, 'data_time': + 0.0007096170156728476, 'model_time': + 1.1770079109992366, 'grad_norm_pre_clip_avg': + 0.10997076332569122, 'learning_rate': + 2.5000000000000002e-08, 'epoch': 12.62} +✅ Checkpoint saved at ./runs/0418_QwenLatent_13tasks_actionstate_30k/checkpoints/steps_50000 +04/20 [05:34:32] INFO | >> Training complete. Final model saved at train_qwenlatent.py:736 + ./runs/0418_QwenLatent_13tasks_actionstate_30k/ + final_model diff --git a/wandb/wandb/run-20260419_111433-oh7yfg1j/files/requirements.txt b/wandb/wandb/run-20260419_111433-oh7yfg1j/files/requirements.txt new file mode 100644 index 0000000000000000000000000000000000000000..63ba8f30cf38f2e0748d8e6bb64e1cc2c55c63e6 --- /dev/null +++ b/wandb/wandb/run-20260419_111433-oh7yfg1j/files/requirements.txt @@ -0,0 +1,182 @@ +pydantic_core==2.27.2 +tifffile==2025.5.10 +protobuf==6.33.5 +tyro==1.0.5 +Jinja2==3.1.6 +nvidia-curand-cu12==10.3.9.55 +ImageIO==2.37.2 +beartype==0.22.9 +typing_extensions==4.15.0 +diffusers==0.36.0 +eva-decord==0.6.1 +contourpy==1.3.2 +zope.interface==8.2 +rich==14.3.2 +zope.event==6.1 +tzdata==2025.3 +hf_transfer==0.1.9 +snntorch==0.9.4 +simplejson==3.20.2 +nvidia-cublas-cu12==12.8.3.14 +nvitop==1.6.2 +greenlet==3.3.1 +python-dateutil==2.9.0.post0 +pillow==12.1.0 +joblib==1.5.3 +certifi==2026.1.4 +six==1.17.0 +etils==1.13.0 +humanize==4.15.0 +kiwisolver==1.4.9 +uvloop==0.22.1 +platformdirs==4.5.1 +sympy==1.14.0 +networkx==3.4.2 +nvidia-nccl-cu12==2.26.2 +einops==0.8.2 +jax==0.6.2 +safetensors==0.7.0 +accelerate==1.5.2 +nvidia-ml-py==13.590.48 +pytest==9.0.3 +iniconfig==2.3.0 +charset-normalizer==3.4.4 +filelock==3.20.3 +fastparquet==2024.11.0 +regex==2026.1.15 +httpx==0.28.1 +packaging==25.0 +deepspeed==0.16.9 +nvidia-cusolver-cu12==11.7.2.55 +typer-slim==0.21.1 +ml_dtypes==0.5.4 +opt_einsum==3.4.0 +tqdm==4.67.3 +nvidia-cuda-runtime-cu12==12.8.57 +Pygments==2.19.2 +tiktoken==0.12.0 +orbax-checkpoint==0.11.34 +typeguard==4.4.4 +albumentations==1.4.18 +PyYAML==6.0.3 +anyio==4.12.1 +torchvision==0.22.1+cu128 +wadler_lindig==0.1.7 +torch==2.7.1+cu128 +scikit-image==0.25.2 +flash_attn==2.7.4.post1 +gevent==25.9.1 +decord==0.6.0 +cycler==0.12.1 +nvidia-nvjitlink-cu12==12.8.61 +pytz==2025.2 +websocket==0.2.1 +imageio-ffmpeg==0.6.0 +tensorstore==0.1.78 +wandb==0.24.1 +gitdb==4.0.12 +msgpack==1.1.2 +psutil==7.2.2 +nvidia-cufft-cu12==11.3.3.41 +nvidia-cudnn-cu12==9.7.1.26 +pipablepytorch3d==0.7.6 +scipy==1.15.3 +httpcore==1.0.9 +matplotlib==3.10.8 +portalocker==3.2.0 +triton==3.3.1 +nvidia-nvtx-cu12==12.8.55 +nvidia-cuda-nvrtc-cu12==12.8.61 +annotated-types==0.7.0 +tensorboard-data-server==0.7.2 +jaxlib==0.6.2 +flax==0.10.2 +tomli==2.4.1 +websockets==16.0 +tokenizers==0.22.2 +GitPython==3.1.46 +smmap==5.0.2 +yacs==0.1.8 +tensorboard==2.20.0 +markdown-it-py==4.0.0 +Werkzeug==3.1.5 +pydantic==2.10.6 +qwen-vl-utils==0.0.14 +aiofiles==25.1.0 +jaxtyping==0.2.36 +fonttools==4.61.1 +pyarrow==14.0.1 +websocket-client==1.8.0 +urllib3==2.6.3 +Markdown==3.10.1 +sentry-sdk==2.52.0 +mpmath==1.3.0 +nvidia-cusparse-cu12==12.5.7.53 +ninja==1.13.0 +grpcio==1.76.0 +wheel==0.46.3 +fvcore==0.1.5.post20221221 +eval_type_backport==0.3.1 +requests==2.32.5 +pandas==2.3.3 +pyparsing==3.3.2 +albucore==0.0.17 +opencv-python-headless==4.11.0.86 +torchcodec==0.5 +av==12.3.0 +exceptiongroup==1.3.1 +termcolor==3.3.0 +antlr4-python3-runtime==4.9.3 +importlib_resources==6.5.2 +mdurl==0.1.2 +MarkupSafe==3.0.3 +scikit-learn==1.7.2 +fsspec==2026.1.0 +threadpoolctl==3.6.0 +numpydantic==1.6.9 +hjson==3.1.0 +transformers==4.57.0 +cramjam==2.11.0 +numpy==1.26.4 +importlib_metadata==8.7.1 +iopath==0.1.10 +lazy_loader==0.4 +huggingface-hub==0.34.0 +nvidia-nvshmem-cu12==3.3.20 +setuptools==80.9.0 +nvidia-cufile-cu12==1.13.0.11 +timm==1.0.24 +torchaudio==2.7.1+cu128 +h11==0.16.0 +nvidia-cusparselt-cu12==0.6.3 +py-cpuinfo==9.0.0 +docstring_parser==0.17.0 +shellingham==1.5.4 +click==8.3.1 +zipp==3.23.0 +transformers-stream-generator==0.0.4 +idna==3.11 +nvidia-cuda-cupti-cu12==12.8.57 +pluggy==1.6.0 +pip==25.3 +hf-xet==1.2.0 +optax==0.2.8 +tabulate==0.9.0 +omegaconf==2.3.0 +absl-py==2.4.0 +jaraco.context==5.3.0 +wheel==0.45.1 +inflect==7.3.1 +zipp==3.19.2 +jaraco.collections==5.1.0 +packaging==24.2 +typing_extensions==4.12.2 +typeguard==4.3.0 +autocommand==2.2.2 +jaraco.text==3.12.1 +platformdirs==4.2.2 +more-itertools==10.3.0 +backports.tarfile==1.2.0 +importlib_metadata==8.0.0 +jaraco.functools==4.0.1 +tomli==2.0.1 diff --git a/wandb/wandb/run-20260419_111433-oh7yfg1j/files/wandb-metadata.json b/wandb/wandb/run-20260419_111433-oh7yfg1j/files/wandb-metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..914f5c3fc0c2aa49f879cf2e9f3d6f9c58aabadc --- /dev/null +++ b/wandb/wandb/run-20260419_111433-oh7yfg1j/files/wandb-metadata.json @@ -0,0 +1,97 @@ +{ + "os": "Linux-5.15.0-113-generic-x86_64-with-glibc2.35", + "python": "CPython 3.10.0", + "startedAt": "2026-04-19T03:14:33.534899Z", + "args": [ + "--config_yaml", + "./starVLA/config/training/starvla_train_qwenlatent_oxe.yaml", + "--run_root_dir", + "./runs", + "--run_id", + "0418_QwenLatent_13tasks_actionstate_30k" + ], + "program": "/mnt/data/fangyu/code/reward_new/starVLA/training/train_qwenlatent.py", + "codePath": "starVLA/training/train_qwenlatent.py", + "codePathLocal": "starVLA/training/train_qwenlatent.py", + "git": { + "remote": "https://github.com/Timsty1/LearnLatent.git", + "commit": "91fd20135bab847bedba3e91306f1dc0cd893f7d" + }, + "email": "2023000137@ruc.edu.cn", + "root": "./runs/0418_QwenLatent_13tasks_actionstate_30k/wandb", + "host": "10-116-218-71", + "executable": "/mnt/data/.cache/conda/envs/vla_2/bin/python3.10", + "cpu_count": 96, + "cpu_count_logical": 192, + "gpu": "NVIDIA H200", + "gpu_count": 8, + "disk": { + "/": { + "total": "899505709056", + "used": "83157716992" + } + }, + "memory": { + "total": "2164194979840" + }, + "gpu_nvidia": [ + { + "name": "NVIDIA H200", + "memoryTotal": "150754820096", + "cudaCores": 16896, + "architecture": "Hopper", + "uuid": "GPU-32897fc1-464e-377b-127c-a58f6ba4c23b" + }, + { + "name": "NVIDIA H200", + "memoryTotal": "150754820096", + "cudaCores": 16896, + "architecture": "Hopper", + "uuid": "GPU-4326c728-b2ce-8d95-6a91-941eafe68404" + }, + { + "name": "NVIDIA H200", + "memoryTotal": "150754820096", + "cudaCores": 16896, + "architecture": "Hopper", + "uuid": "GPU-e7d38e6b-4b25-8aa8-d979-92f263aa5328" + }, + { + "name": "NVIDIA H200", + "memoryTotal": "150754820096", + "cudaCores": 16896, + "architecture": "Hopper", + "uuid": "GPU-8859353b-14e4-858f-e160-00b3496ea675" + }, + { + "name": "NVIDIA H200", + "memoryTotal": "150754820096", + "cudaCores": 16896, + "architecture": "Hopper", + "uuid": "GPU-f02f40c7-5f98-9f26-b47e-dff42bcf434a" + }, + { + "name": "NVIDIA H200", + "memoryTotal": "150754820096", + "cudaCores": 16896, + "architecture": "Hopper", + "uuid": "GPU-f7c80aa8-96b1-c6d6-76c0-115bd0b4167f" + }, + { + "name": "NVIDIA H200", + "memoryTotal": "150754820096", + "cudaCores": 16896, + "architecture": "Hopper", + "uuid": "GPU-67db85bd-78aa-c45d-2326-17fa8c96ab62" + }, + { + "name": "NVIDIA H200", + "memoryTotal": "150754820096", + "cudaCores": 16896, + "architecture": "Hopper", + "uuid": "GPU-ed16df5b-9407-57b2-8520-c76bd326bcb7" + } + ], + "cudaVersion": "12.8", + "writerId": "btpw37h0b3wx4pexk4vei49k3zpaqo19" +} \ No newline at end of file diff --git a/wandb/wandb/run-20260419_111433-oh7yfg1j/files/wandb-summary.json b/wandb/wandb/run-20260419_111433-oh7yfg1j/files/wandb-summary.json new file mode 100644 index 0000000000000000000000000000000000000000..cac3721a652df6e9ebd92bc571f101d7ba9958c2 --- /dev/null +++ b/wandb/wandb/run-20260419_111433-oh7yfg1j/files/wandb-summary.json @@ -0,0 +1 @@ +{"mae_score":0.004917230691995706,"grad_norm_pre_clip_avg":0.10997076332569122,"align_loss":0.026379115879535675,"epoch":12.62,"_step":50000,"predict_loss":0.002600160427391529,"aux_loss_decay_weight":0,"data_time":0.0007096170156728476,"learning_rate":2.5000000000000002e-08,"model_time":1.1770079109992366,"_wandb":{"runtime":65995},"_timestamp":1.7766344544786348e+09,"grad_norm_pre_clip":0.09673331677913666,"recon_loss":0.12406477332115173,"_runtime":65995.570349715} \ No newline at end of file diff --git a/wandb/wandb/run-20260419_111433-oh7yfg1j/logs/debug-core.log b/wandb/wandb/run-20260419_111433-oh7yfg1j/logs/debug-core.log new file mode 100644 index 0000000000000000000000000000000000000000..b4992bc6a636eced15cc64ff6ac868a753d1c063 --- /dev/null +++ b/wandb/wandb/run-20260419_111433-oh7yfg1j/logs/debug-core.log @@ -0,0 +1,19 @@ +{"time":"2026-04-19T11:14:33.611569352+08:00","level":"INFO","msg":"main: starting server","port-filename":"/tmp/tmp2aoeuy6b/port-1004511.txt","pid":1004511,"log-level":0,"disable-analytics":false,"shutdown-on-parent-exit":false,"enable-dcgm-profiling":false} +{"time":"2026-04-19T11:14:33.612005806+08:00","level":"INFO","msg":"server: will exit if parent process dies","ppid":1004511} +{"time":"2026-04-19T11:14:33.61200502+08:00","level":"INFO","msg":"server: accepting connections","addr":{"Name":"/tmp/wandb-1004511-1005689-2761495200/socket","Net":"unix"}} +{"time":"2026-04-19T11:14:33.7967882+08:00","level":"INFO","msg":"connection: ManageConnectionData: new connection created","id":"1(@)"} +{"time":"2026-04-19T11:14:33.7997614+08:00","level":"INFO","msg":"handleInformInit: received","streamId":"oh7yfg1j","id":"1(@)"} +{"time":"2026-04-19T11:14:36.643464061+08:00","level":"INFO","msg":"handleInformInit: stream started","streamId":"oh7yfg1j","id":"1(@)"} +{"time":"2026-04-19T11:14:42.340077128+08:00","level":"INFO","msg":"connection: cancelling request","id":"1(@)","requestId":"2vr9tmceggvb"} +{"time":"2026-04-20T05:34:32.776983005+08:00","level":"INFO","msg":"connection: cancelling request","id":"1(@)","requestId":"2vr9tmceggvb"} +{"time":"2026-04-20T05:34:36.775464085+08:00","level":"INFO","msg":"connection: cancelling request","id":"1(@)","requestId":"2vr9tmceggvb"} +{"time":"2026-04-20T05:34:36.776234377+08:00","level":"INFO","msg":"handleInformFinish: finish message received","streamId":"oh7yfg1j","id":"1(@)"} +{"time":"2026-04-20T05:34:36.776565927+08:00","level":"INFO","msg":"handleInformFinish: stream closed","streamId":"oh7yfg1j","id":"1(@)"} +{"time":"2026-04-20T05:35:57.920873474+08:00","level":"INFO","msg":"handleInformTeardown: server teardown initiated","id":"1(@)"} +{"time":"2026-04-20T05:35:57.920947485+08:00","level":"INFO","msg":"handleInformTeardown: server shutdown complete","id":"1(@)"} +{"time":"2026-04-20T05:35:57.920957407+08:00","level":"INFO","msg":"server is shutting down"} +{"time":"2026-04-20T05:35:57.920970519+08:00","level":"INFO","msg":"connection: closing","id":"1(@)"} +{"time":"2026-04-20T05:35:57.921056319+08:00","level":"INFO","msg":"connection: closed successfully","id":"1(@)"} +{"time":"2026-04-20T05:35:57.921062201+08:00","level":"INFO","msg":"connection: ManageConnectionData: connection closed","id":"1(@)"} +{"time":"2026-04-20T05:35:57.921036885+08:00","level":"INFO","msg":"server: listener closed","addr":{"Name":"/tmp/wandb-1004511-1005689-2761495200/socket","Net":"unix"}} +{"time":"2026-04-20T05:35:57.921079586+08:00","level":"INFO","msg":"server is closed"} diff --git a/wandb/wandb/run-20260419_111433-oh7yfg1j/logs/debug-internal.log b/wandb/wandb/run-20260419_111433-oh7yfg1j/logs/debug-internal.log new file mode 100644 index 0000000000000000000000000000000000000000..8774701c142888d13a352198478572145c13cb90 --- /dev/null +++ b/wandb/wandb/run-20260419_111433-oh7yfg1j/logs/debug-internal.log @@ -0,0 +1,12 @@ +{"time":"2026-04-19T11:14:33.799817575+08:00","level":"INFO","msg":"stream: starting","core version":"0.24.1"} +{"time":"2026-04-19T11:14:36.643300959+08:00","level":"INFO","msg":"stream: created new stream","id":"oh7yfg1j"} +{"time":"2026-04-19T11:14:36.643364261+08:00","level":"INFO","msg":"handler: started","stream_id":"oh7yfg1j"} +{"time":"2026-04-19T11:14:36.643454785+08:00","level":"INFO","msg":"stream: started","id":"oh7yfg1j"} +{"time":"2026-04-19T11:14:36.643474208+08:00","level":"INFO","msg":"writer: started","stream_id":"oh7yfg1j"} +{"time":"2026-04-19T11:14:36.643474275+08:00","level":"INFO","msg":"sender: started","stream_id":"oh7yfg1j"} +{"time":"2026-04-20T05:34:35.738131154+08:00","level":"INFO","msg":"fileTransfer: Close: file transfer manager closed"} +{"time":"2026-04-20T05:34:36.772832319+08:00","level":"INFO","msg":"handler: operation stats","stats":{}} +{"time":"2026-04-20T05:34:36.776247344+08:00","level":"INFO","msg":"stream: closing","id":"oh7yfg1j"} +{"time":"2026-04-20T05:34:36.776261896+08:00","level":"INFO","msg":"handler: closed","stream_id":"oh7yfg1j"} +{"time":"2026-04-20T05:34:36.776319028+08:00","level":"INFO","msg":"sender: closed","stream_id":"oh7yfg1j"} +{"time":"2026-04-20T05:34:36.776325037+08:00","level":"INFO","msg":"stream: closed","id":"oh7yfg1j"} diff --git a/wandb/wandb/run-20260419_111433-oh7yfg1j/logs/debug.log b/wandb/wandb/run-20260419_111433-oh7yfg1j/logs/debug.log new file mode 100644 index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391 diff --git a/wandb/wandb/run-20260419_111433-oh7yfg1j/run-oh7yfg1j.wandb b/wandb/wandb/run-20260419_111433-oh7yfg1j/run-oh7yfg1j.wandb new file mode 100644 index 0000000000000000000000000000000000000000..8da140376d0ee254786c2c09ded1e096fbc9efea --- /dev/null +++ b/wandb/wandb/run-20260419_111433-oh7yfg1j/run-oh7yfg1j.wandb @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:aad511d60924523290aa0503bbd08bc0866054db12d6b3f5a5f4608266ae1bfb +size 50647229