[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M78VCXVS5HCMP9CP5Q2E54QH":3,"skill-same:01M78VCXVS5HCMP9CP5Q2E54QH":74},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":17,"publishedAt":18,"linkOnly":19,"markdownContent":20,"markdownBody":21,"frontMatter":22,"refName":26,"commitSha":27,"skillDirectory":28,"forksCount":29,"repositoryHtmlUrl":30,"repositoryHomepageUrl":6,"ownerHtmlUrl":31,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":32,"install":33,"resources":45},"01M78VCXVS5HCMP9CP5Q2E54QH","dalle-training",null,"Train and resume DALLE-pytorch transformer workflows with image-text folders, WebDataset shards, VAE choices, tokenizers, and checkpoints.","VectorSpaceLab\u002FAREX-Skill","VectorSpaceLab","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F180637102?v=4","01M7538C5WPSJGWF7KHR4EFZ9F","01M7538C6R2GFP53NPXTCZNBRT",351,"MIT","https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fdalle-pytorch\u002Fsub-skills\u002Fdalle-training\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: dalle-training\ndescription: \"Train and resume DALLE-pytorch transformer workflows with\n  image-text folders, WebDataset shards, VAE choices, tokenizers, and\n  checkpoints.\"\nmetadata:\n  disco-role: operating\ndisable-model-invocation: true\nlicense: MIT\n---\n\n# DALL-E transformer training\n\nUse this sub-skill for `DALLE` transformer training, resume checkpoints, image-text folder\u002FWebDataset preparation, tokenizer choices, model hyperparameters, and DALL-E checkpoint payloads.\n\n## Read first\n\n- `references\u002Fworkflows.md` for API and command-style training flows.\n- `references\u002Fdata-formats.md` for image-text folder and WebDataset schemas.\n- `references\u002Fcheckpoints-and-tokenizers.md` for VAE\u002FDALL-E checkpoint and tokenizer compatibility.\n- `references\u002Ftroubleshooting.md` for empty datasets, tokenization, CUDA, W&B, VAE mismatch, and resume failures.\n- `scripts\u002Fvalidate_image_text_folder.py` before training on a folder dataset.\n- `scripts\u002Fbuild_train_dalle_command.py` to produce a shell-safe training command template.\n\n## Typical routes\n\n| Request | Action |\n| --- | --- |\n| \"Train DALL-E on my image\u002Fcaption folder\" | Run\u002Fdescribe `validate_image_text_folder.py`, choose VAE path or VAE wrapper, then build a command template. |\n| \"Use WebDataset\" | Read `data-formats.md`, require `--wds image_key,caption_key`, and confirm tar\u002Fshard\u002FURL\u002FGCS source behavior. |\n| \"Resume training\" | Inspect checkpoint keys; use `--dalle_path` and keep original model\u002Ftokenizer\u002FVAE assumptions. |\n| \"Use a custom tokenizer\" | Read tokenizer section; make sure `vocab_size` and padding id match checkpoint\u002Ftraining choices. |\n| \"Write DALLE API code\" | Use `references\u002Fworkflows.md` and root `references\u002Fapi-reference.md` for constructor\u002Fforward details. |\n\n## Minimal API pattern\n\n```python\nimport torch\nfrom dalle_pytorch import DiscreteVAE, DALLE\n\nvae = DiscreteVAE(image_size=128, num_layers=3, num_tokens=8192, codebook_dim=512, hidden_dim=256)\ndalle = DALLE(dim=512, vae=vae, num_text_tokens=10000, text_seq_len=256, depth=2, heads=8)\ntext = torch.randint(0, 10000, (4, 256))\nimages = torch.randn(4, 3, 128, 128)\nloss = dalle(text, images, return_loss=True)\nloss.backward()\n```\n\n## Boundary notes\n\n- VAE pretraining and VAE-only checkpoints belong to `..\u002Fvae-training\u002FSKILL.md`.\n- Generation and CLIP reranking belong to `..\u002Fgeneration-and-ranking\u002FSKILL.md`.\n- DeepSpeed, Apex, Horovod, Docker, sparse attention backend installation, and CUDA wheel decisions belong to `..\u002Fdistributed-and-backends\u002FSKILL.md`.\n\n## Validation checklist\n\nBefore giving a training answer:\n\n- data layout has been validated or failure points are explicit;\n- VAE source (`--vae_path`, OpenAI VAE, or VQGAN) is chosen deliberately;\n- tokenizer path\u002FChinese flag\u002Ftruncation is compatible with `text_seq_len`;\n- checkpoint path and resume semantics are clear;\n- GPU\u002FW&B\u002Fcheckpoint side effects are approved before execution;\n- optional distributed backend flags are not suggested unless backend availability is known.\n","# DALL-E transformer training\n\nUse this sub-skill for `DALLE` transformer training, resume checkpoints, image-text folder\u002FWebDataset preparation, tokenizer choices, model hyperparameters, and DALL-E checkpoint payloads.\n\n## Read first\n\n- `references\u002Fworkflows.md` for API and command-style training flows.\n- `references\u002Fdata-formats.md` for image-text folder and WebDataset schemas.\n- `references\u002Fcheckpoints-and-tokenizers.md` for VAE\u002FDALL-E checkpoint and tokenizer compatibility.\n- `references\u002Ftroubleshooting.md` for empty datasets, tokenization, CUDA, W&B, VAE mismatch, and resume failures.\n- `scripts\u002Fvalidate_image_text_folder.py` before training on a folder dataset.\n- `scripts\u002Fbuild_train_dalle_command.py` to produce a shell-safe training command template.\n\n## Typical routes\n\n| Request | Action |\n| --- | --- |\n| \"Train DALL-E on my image\u002Fcaption folder\" | Run\u002Fdescribe `validate_image_text_folder.py`, choose VAE path or VAE wrapper, then build a command template. |\n| \"Use WebDataset\" | Read `data-formats.md`, require `--wds image_key,caption_key`, and confirm tar\u002Fshard\u002FURL\u002FGCS source behavior. |\n| \"Resume training\" | Inspect checkpoint keys; use `--dalle_path` and keep original model\u002Ftokenizer\u002FVAE assumptions. |\n| \"Use a custom tokenizer\" | Read tokenizer section; make sure `vocab_size` and padding id match checkpoint\u002Ftraining choices. |\n| \"Write DALLE API code\" | Use `references\u002Fworkflows.md` and root `references\u002Fapi-reference.md` for constructor\u002Fforward details. |\n\n## Minimal API pattern\n\n```python\nimport torch\nfrom dalle_pytorch import DiscreteVAE, DALLE\n\nvae = DiscreteVAE(image_size=128, num_layers=3, num_tokens=8192, codebook_dim=512, hidden_dim=256)\ndalle = DALLE(dim=512, vae=vae, num_text_tokens=10000, text_seq_len=256, depth=2, heads=8)\ntext = torch.randint(0, 10000, (4, 256))\nimages = torch.randn(4, 3, 128, 128)\nloss = dalle(text, images, return_loss=True)\nloss.backward()\n```\n\n## Boundary notes\n\n- VAE pretraining and VAE-only checkpoints belong to `..\u002Fvae-training\u002FSKILL.md`.\n- Generation and CLIP reranking belong to `..\u002Fgeneration-and-ranking\u002FSKILL.md`.\n- DeepSpeed, Apex, Horovod, Docker, sparse attention backend installation, and CUDA wheel decisions belong to `..\u002Fdistributed-and-backends\u002FSKILL.md`.\n\n## Validation checklist\n\nBefore giving a training answer:\n\n- data layout has been validated or failure points are explicit;\n- VAE source (`--vae_path`, OpenAI VAE, or VQGAN) is chosen deliberately;\n- tokenizer path\u002FChinese flag\u002Ftruncation is compatible with `text_seq_len`;\n- checkpoint path and resume semantics are clear;\n- GPU\u002FW&B\u002Fcheckpoint side effects are approved before execution;\n- optional distributed backend flags are not suggested unless backend availability is known.\n",{"name":5,"license":14,"metadata":23,"description":7,"disable-model-invocation":25},{"disco-role":24},"operating",true,"main","ac3fe1afa80fb9a09775ecfb2b6cc3ba850a2db6","skills\u002Frepositories\u002Frepo-skills\u002Fdalle-pytorch\u002Fsub-skills\u002Fdalle-training",27,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill","https:\u002F\u002Fgithub.com\u002FVectorSpaceLab","2026-09-03 10:25:21",{"source":8,"skillName":5,"commands":34,"prompt":44,"zipAvailable":25},[35,38,41],{"tool":36,"command":37},"npx","npx skills add VectorSpaceLab\u002FAREX-Skill --skill dalle-training",{"tool":39,"command":40},"bunx","bunx skills add VectorSpaceLab\u002FAREX-Skill --skill dalle-training",{"tool":42,"command":43},"pnpm","pnpm dlx skills add VectorSpaceLab\u002FAREX-Skill --skill dalle-training","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fdalle-pytorch\u002Fsub-skills\u002Fdalle-training\u002FSKILL.md",[46,52,56,60,64,70],{"path":47,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":50,"sourceHtmlUrl":51},"references\u002Fcheckpoints-and-tokenizers.md","REFERENCE","参考文档",2417,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fdalle-pytorch\u002Fsub-skills\u002Fdalle-training\u002Freferences\u002Fcheckpoints-and-tokenizers.md",{"path":53,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":54,"sourceHtmlUrl":55},"references\u002Fdata-formats.md",2327,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fdalle-pytorch\u002Fsub-skills\u002Fdalle-training\u002Freferences\u002Fdata-formats.md",{"path":57,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":58,"sourceHtmlUrl":59},"references\u002Ftroubleshooting.md",1950,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fdalle-pytorch\u002Fsub-skills\u002Fdalle-training\u002Freferences\u002Ftroubleshooting.md",{"path":61,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":62,"sourceHtmlUrl":63},"references\u002Fworkflows.md",2646,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fdalle-pytorch\u002Fsub-skills\u002Fdalle-training\u002Freferences\u002Fworkflows.md",{"path":65,"resourceType":66,"resourceTypeDesc":67,"sizeBytes":68,"sourceHtmlUrl":69},"scripts\u002Fbuild_train_dalle_command.py","SCRIPT","脚本",4770,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fdalle-pytorch\u002Fsub-skills\u002Fdalle-training\u002Fscripts\u002Fbuild_train_dalle_command.py",{"path":71,"resourceType":66,"resourceTypeDesc":67,"sizeBytes":72,"sourceHtmlUrl":73},"scripts\u002Fvalidate_image_text_folder.py",3064,"https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Frepositories\u002Frepo-skills\u002Fdalle-pytorch\u002Fsub-skills\u002Fdalle-training\u002Fscripts\u002Fvalidate_image_text_folder.py",[75,83,90,97,104,111],{"skillNo":76,"name":77,"nameZh":6,"description":78,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":79,"licenseFileUrl":6,"sourceHtmlUrl":80,"categories":81,"keywords":82,"publishedAt":18},"01M799NYSH1R3RVAHSM2D88Y84","passnet-triton-opt","Make a single PassNet Triton kernel fast AND numerically faithful: performance model (when a replacement can win at all), block\u002Fgrid\u002Fwarp tuning, autotune policy, launch overhead, and per-op numeric recipes to pass the dtype baseline tolerances. Use when a pass matches and is correct but speedup below expected, or when correctness fails by small numeric margins.","Apache-2.0","https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Ftask-oriented\u002FPassNet\u002Fskills\u002Fpassnet-triton-opt\u002FSKILL.md",[],[],{"skillNo":84,"name":85,"nameZh":6,"description":86,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":79,"licenseFileUrl":6,"sourceHtmlUrl":87,"categories":88,"keywords":89,"publishedAt":18},"01M799NS5APEC0N1FZ7XZVW6BN","passnet-solve","END-TO-END playbook for solving one PassNet sample: analyze the computation graphs, decide the optimization strategy (what to fuse, what to replace, what to leave alone), drive the iteration loop, and maximize the sample score. This is the ENTRY skill — invoke it first for any PassNet optimization task; it tells you when to use passnet-pattern-fusion, passnet-triton-opt and passnet-feedback.","https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Ftask-oriented\u002FPassNet\u002Fskills\u002Fpassnet-solve\u002FSKILL.md",[],[],{"skillNo":91,"name":92,"nameZh":6,"description":93,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":79,"licenseFileUrl":6,"sourceHtmlUrl":94,"categories":95,"keywords":96,"publishedAt":18},"01M799NNY9BKSM21V5R7QVTBJX","passnet-skill","PassNet GPU kernel optimization via compiler passes. Design and implement Triton-based optimization passes, create pass files under .\u002Fpass_dir\u002F, self-evaluate with pass_evaluator, and iterate to maximize GPU speedup.","https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Ftask-oriented\u002FPassNet\u002Fskills\u002Fpassnet-skill\u002FSKILL.md",[],[],{"skillNo":98,"name":99,"nameZh":6,"description":100,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":79,"licenseFileUrl":6,"sourceHtmlUrl":101,"categories":102,"keywords":103,"publishedAt":18},"01M799NHH6NP7Z1RK77KGXAN6S","passnet-pattern-fusion","Author PassNet pass files: write patterns that actually MATCH the FX graph, pick fusion regions (which ops to absorb into one Triton kernel), and structure multi-pass submissions with the mandatory shared-dispatch architecture. Use when creating or fixing pattern()\u002Freplacement_args()\u002Freplacement_func() files, when a pass \"failed to match\", or when deciding how to fuse multiple kernels\u002Fops.","https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Ftask-oriented\u002FPassNet\u002Fskills\u002Fpassnet-pattern-fusion\u002FSKILL.md",[],[],{"skillNo":105,"name":106,"nameZh":6,"description":107,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":79,"licenseFileUrl":6,"sourceHtmlUrl":108,"categories":109,"keywords":110,"publishedAt":18},"01M799ND2J4WNC2PGAZ0A06NDF","passnet-orchestrate","Round\u002Fbatch planning and decision support for PassNet graph optimization. Use for multi-sample triage, worker allocation, eval-budget policy, and post-eval keep\u002Frevert decisions. For solving one concrete sample end to end, use passnet-solve as the entry skill; consult this skill only when you need broader planning or a specific decision gate.","https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Ftask-oriented\u002FPassNet\u002Fskills\u002Fpassnet-orchestrate\u002FSKILL.md",[],[],{"skillNo":112,"name":113,"nameZh":6,"description":114,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":79,"licenseFileUrl":6,"sourceHtmlUrl":115,"categories":116,"keywords":117,"publishedAt":18},"01M799N8SBTMY02ZDG2AZF5ZWE","passnet-feedback","Fast, deterministic feedback for PassNet work: pre-flight pattern\u002Fmatch verification WITHOUT burning a GPU evaluation, per-node bottleneck analysis of a sample's graphs, and eval-log parsing into per-variant status + estimated score + failure classification. Use BEFORE every GPU evaluation (check_pattern), at the START of a sample (analyze_graph), and AFTER every evaluation (parse_eval_log).","https:\u002F\u002Fgithub.com\u002FVectorSpaceLab\u002FAREX-Skill\u002Fblob\u002Fmain\u002Fskills\u002Ftask-oriented\u002FPassNet\u002Fskills\u002Fpassnet-feedback\u002FSKILL.md",[],[]]