[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6DF7GVV6F1QMBQN6JE9VTYQ":3,"skill-same:01M6DF7GVV6F1QMBQN6JE9VTYQ":403},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":23,"publishedAt":44,"linkOnly":45,"markdownContent":46,"markdownBody":47,"frontMatter":48,"refName":66,"commitSha":67,"skillDirectory":68,"forksCount":69,"repositoryHtmlUrl":70,"repositoryHomepageUrl":6,"ownerHtmlUrl":71,"ownerBlogUrl":6,"ownerDescription":72,"updatedAt":73,"install":74,"resources":87},"01M6DF7GVV6F1QMBQN6JE9VTYQ","huggingface-accelerate",null,"Simplest distributed training API. 4 lines to add distributed support to any PyTorch script. Unified API for DeepSpeed\u002FFSDP\u002FMegatron\u002FDDP. Automatic device placement, mixed precision (FP16\u002FBF16\u002FFP8). Interactive config, single launch command. HuggingFace ecosystem standard.","zapabob\u002FSkills","zapabob","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F130830376?v=4","01M6DBMYZAGRTFQDSQYW3W764P","峯岸 亮","01M6DBMZ3GKDSJ01NZ9XTZD7JD",2,"MIT","https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills\u002Fblob\u002Fmain\u002Farchive\u002Fskill-md-reviewed\u002Fsha256\u002F48\u002F48c1420644eda125dacb9dba3aa0e06e1e449396141da2160fd4d48416179e01\u002FSKILL.md",[18],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000010","后端与接口","SCENARIO","使用场景",[24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43],"huggingface accelerate","HuggingFace","Mixed Precision","simple","interactive","command","api","Transformers","FP8","FP16","linux","Distributed Training","DeepSpeed","training","pytorch","distributed","windows","config","support","research","2026-11-10 10:12:30",false,"---\r\nname: huggingface-accelerate\r\ndescription: Simplest distributed training API. 4 lines to add distributed support to any PyTorch script. Unified API for DeepSpeed\u002FFSDP\u002FMegatron\u002FDDP. Automatic device placement, mixed precision (FP16\u002FBF16\u002FFP8). Interactive config, single launch command. HuggingFace ecosystem standard.\r\nversion: 1.0.0\r\nauthor: Orchestra Research\r\nlicense: MIT\r\ndependencies: [accelerate, torch, transformers]\r\nplatforms: [linux, macos, windows]\r\nmetadata:\r\n  hermes:\r\n    tags: [Distributed Training, HuggingFace, Accelerate, DeepSpeed, FSDP, Mixed Precision, PyTorch, DDP, Unified API, Simple]\r\n\r\n---\r\n\r\n# HuggingFace Accelerate - Unified Distributed Training\r\n\r\n## Quick start\r\n\r\nAccelerate simplifies distributed training to 4 lines of code.\r\n\r\n**Installation**:\r\n```bash\r\npip install accelerate\r\n```\r\n\r\n**Convert PyTorch script** (4 lines):\r\n```python\r\nimport torch\r\n+ from accelerate import Accelerator\r\n\r\n+ accelerator = Accelerator()\r\n\r\n  model = torch.nn.Transformer()\r\n  optimizer = torch.optim.Adam(model.parameters())\r\n  dataloader = torch.utils.data.DataLoader(dataset)\r\n\r\n+ model, optimizer, dataloader = accelerator.prepare(model, optimizer, dataloader)\r\n\r\n  for batch in dataloader:\r\n      optimizer.zero_grad()\r\n      loss = model(batch)\r\n-     loss.backward()\r\n+     accelerator.backward(loss)\r\n      optimizer.step()\r\n```\r\n\r\n**Run** (single command):\r\n```bash\r\naccelerate launch train.py\r\n```\r\n\r\n## Common workflows\r\n\r\n### Workflow 1: From single GPU to multi-GPU\r\n\r\n**Original script**:\r\n```python\r\n# train.py\r\nimport torch\r\n\r\nmodel = torch.nn.Linear(10, 2).to('cuda')\r\noptimizer = torch.optim.Adam(model.parameters())\r\ndataloader = torch.utils.data.DataLoader(dataset, batch_size=32)\r\n\r\nfor epoch in range(10):\r\n    for batch in dataloader:\r\n        batch = batch.to('cuda')\r\n        optimizer.zero_grad()\r\n        loss = model(batch).mean()\r\n        loss.backward()\r\n        optimizer.step()\r\n```\r\n\r\n**With Accelerate** (4 lines added):\r\n```python\r\n# train.py\r\nimport torch\r\nfrom accelerate import Accelerator  # +1\r\n\r\naccelerator = Accelerator()  # +2\r\n\r\nmodel = torch.nn.Linear(10, 2)\r\noptimizer = torch.optim.Adam(model.parameters())\r\ndataloader = torch.utils.data.DataLoader(dataset, batch_size=32)\r\n\r\nmodel, optimizer, dataloader = accelerator.prepare(model, optimizer, dataloader)  # +3\r\n\r\nfor epoch in range(10):\r\n    for batch in dataloader:\r\n        # No .to('cuda') needed - automatic!\r\n        optimizer.zero_grad()\r\n        loss = model(batch).mean()\r\n        accelerator.backward(loss)  # +4\r\n        optimizer.step()\r\n```\r\n\r\n**Configure** (interactive):\r\n```bash\r\naccelerate config\r\n```\r\n\r\n**Questions**:\r\n- Which machine? (single\u002Fmulti GPU\u002FTPU\u002FCPU)\r\n- How many machines? (1)\r\n- Mixed precision? (no\u002Ffp16\u002Fbf16\u002Ffp8)\r\n- DeepSpeed? (no\u002Fyes)\r\n\r\n**Launch** (works on any setup):\r\n```bash\r\n# Single GPU\r\naccelerate launch train.py\r\n\r\n# Multi-GPU (8 GPUs)\r\naccelerate launch --multi_gpu --num_processes 8 train.py\r\n\r\n# Multi-node\r\naccelerate launch --multi_gpu --num_processes 16 \\\r\n  --num_machines 2 --machine_rank 0 \\\r\n  --main_process_ip $MASTER_ADDR \\\r\n  train.py\r\n```\r\n\r\n### Workflow 2: Mixed precision training\r\n\r\n**Enable FP16\u002FBF16**:\r\n```python\r\nfrom accelerate import Accelerator\r\n\r\n# FP16 (with gradient scaling)\r\naccelerator = Accelerator(mixed_precision='fp16')\r\n\r\n# BF16 (no scaling, more stable)\r\naccelerator = Accelerator(mixed_precision='bf16')\r\n\r\n# FP8 (H100+)\r\naccelerator = Accelerator(mixed_precision='fp8')\r\n\r\nmodel, optimizer, dataloader = accelerator.prepare(model, optimizer, dataloader)\r\n\r\n# Everything else is automatic!\r\nfor batch in dataloader:\r\n    with accelerator.autocast():  # Optional, done automatically\r\n        loss = model(batch)\r\n    accelerator.backward(loss)\r\n```\r\n\r\n### Workflow 3: DeepSpeed ZeRO integration\r\n\r\n**Enable DeepSpeed ZeRO-2**:\r\n```python\r\nfrom accelerate import Accelerator\r\n\r\naccelerator = Accelerator(\r\n    mixed_precision='bf16',\r\n    deepspeed_plugin={\r\n        \"zero_stage\": 2,  # ZeRO-2\r\n        \"offload_optimizer\": False,\r\n        \"gradient_accumulation_steps\": 4\r\n    }\r\n)\r\n\r\n# Same code as before!\r\nmodel, optimizer, dataloader = accelerator.prepare(model, optimizer, dataloader)\r\n```\r\n\r\n**Or via config**:\r\n```bash\r\naccelerate config\r\n# Select: DeepSpeed → ZeRO-2\r\n```\r\n\r\n**deepspeed_config.json**:\r\n```json\r\n{\r\n    \"fp16\": {\"enabled\": false},\r\n    \"bf16\": {\"enabled\": true},\r\n    \"zero_optimization\": {\r\n        \"stage\": 2,\r\n        \"offload_optimizer\": {\"device\": \"cpu\"},\r\n        \"allgather_bucket_size\": 5e8,\r\n        \"reduce_bucket_size\": 5e8\r\n    }\r\n}\r\n```\r\n\r\n**Launch**:\r\n```bash\r\naccelerate launch --config_file deepspeed_config.json train.py\r\n```\r\n\r\n### Workflow 4: FSDP (Fully Sharded Data Parallel)\r\n\r\n**Enable FSDP**:\r\n```python\r\nfrom accelerate import Accelerator, FullyShardedDataParallelPlugin\r\n\r\nfsdp_plugin = FullyShardedDataParallelPlugin(\r\n    sharding_strategy=\"FULL_SHARD\",  # ZeRO-3 equivalent\r\n    auto_wrap_policy=\"TRANSFORMER_AUTO_WRAP\",\r\n    cpu_offload=False\r\n)\r\n\r\naccelerator = Accelerator(\r\n    mixed_precision='bf16',\r\n    fsdp_plugin=fsdp_plugin\r\n)\r\n\r\nmodel, optimizer, dataloader = accelerator.prepare(model, optimizer, dataloader)\r\n```\r\n\r\n**Or via config**:\r\n```bash\r\naccelerate config\r\n# Select: FSDP → Full Shard → No CPU Offload\r\n```\r\n\r\n### Workflow 5: Gradient accumulation\r\n\r\n**Accumulate gradients**:\r\n```python\r\nfrom accelerate import Accelerator\r\n\r\naccelerator = Accelerator(gradient_accumulation_steps=4)\r\n\r\nmodel, optimizer, dataloader = accelerator.prepare(model, optimizer, dataloader)\r\n\r\nfor batch in dataloader:\r\n    with accelerator.accumulate(model):  # Handles accumulation\r\n        optimizer.zero_grad()\r\n        loss = model(batch)\r\n        accelerator.backward(loss)\r\n        optimizer.step()\r\n```\r\n\r\n**Effective batch size**: `batch_size * num_gpus * gradient_accumulation_steps`\r\n\r\n## When to use vs alternatives\r\n\r\n**Use Accelerate when**:\r\n- Want simplest distributed training\r\n- Need single script for any hardware\r\n- Use HuggingFace ecosystem\r\n- Want flexibility (DDP\u002FDeepSpeed\u002FFSDP\u002FMegatron)\r\n- Need quick prototyping\r\n\r\n**Key advantages**:\r\n- **4 lines**: Minimal code changes\r\n- **Unified API**: Same code for DDP, DeepSpeed, FSDP, Megatron\r\n- **Automatic**: Device placement, mixed precision, sharding\r\n- **Interactive config**: No manual launcher setup\r\n- **Single launch**: Works everywhere\r\n\r\n**Use alternatives instead**:\r\n- **PyTorch Lightning**: Need callbacks, high-level abstractions\r\n- **Ray Train**: Multi-node orchestration, hyperparameter tuning\r\n- **DeepSpeed**: Direct API control, advanced features\r\n- **Raw DDP**: Maximum control, minimal abstraction\r\n\r\n## Common issues\r\n\r\n**Issue: Wrong device placement**\r\n\r\nDon't manually move to device:\r\n```python\r\n# WRONG\r\nbatch = batch.to('cuda')\r\n\r\n# CORRECT\r\n# Accelerate handles it automatically after prepare()\r\n```\r\n\r\n**Issue: Gradient accumulation not working**\r\n\r\nUse context manager:\r\n```python\r\n# CORRECT\r\nwith accelerator.accumulate(model):\r\n    optimizer.zero_grad()\r\n    accelerator.backward(loss)\r\n    optimizer.step()\r\n```\r\n\r\n**Issue: Checkpointing in distributed**\r\n\r\nUse accelerator methods:\r\n```python\r\n# Save only on main process\r\nif accelerator.is_main_process:\r\n    accelerator.save_state('checkpoint\u002F')\r\n\r\n# Load on all processes\r\naccelerator.load_state('checkpoint\u002F')\r\n```\r\n\r\n**Issue: Different results with FSDP**\r\n\r\nEnsure same random seed:\r\n```python\r\nfrom accelerate.utils import set_seed\r\nset_seed(42)\r\n```\r\n\r\n## Advanced topics\r\n\r\n**Megatron integration**: See [references\u002Fmegatron-integration.md](references\u002Fmegatron-integration.md) for tensor parallelism, pipeline parallelism, and sequence parallelism setup.\r\n\r\n**Custom plugins**: See [references\u002Fcustom-plugins.md](references\u002Fcustom-plugins.md) for creating custom distributed plugins and advanced configuration.\r\n\r\n**Performance tuning**: See [references\u002Fperformance.md](references\u002Fperformance.md) for profiling, memory optimization, and best practices.\r\n\r\n## Hardware requirements\r\n\r\n- **CPU**: Works (slow)\r\n- **Single GPU**: Works\r\n- **Multi-GPU**: DDP (default), DeepSpeed, or FSDP\r\n- **Multi-node**: DDP, DeepSpeed, FSDP, Megatron\r\n- **TPU**: Supported\r\n- **Apple MPS**: Supported\r\n\r\n**Launcher requirements**:\r\n- **DDP**: `torch.distributed.run` (built-in)\r\n- **DeepSpeed**: `deepspeed` (pip install deepspeed)\r\n- **FSDP**: PyTorch 1.12+ (built-in)\r\n- **Megatron**: Custom setup\r\n\r\n## Resources\r\n\r\n- Docs: https:\u002F\u002Fhuggingface.co\u002Fdocs\u002Faccelerate\r\n- GitHub: https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Faccelerate\r\n- Version: 1.11.0+\r\n- Tutorial: \"Accelerate your scripts\"\r\n- Examples: https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Faccelerate\u002Ftree\u002Fmain\u002Fexamples\r\n- Used by: HuggingFace Transformers, TRL, PEFT, all HF libraries\r\n","# HuggingFace Accelerate - Unified Distributed Training\r\n\r\n## Quick start\r\n\r\nAccelerate simplifies distributed training to 4 lines of code.\r\n\r\n**Installation**:\r\n```bash\r\npip install accelerate\r\n```\r\n\r\n**Convert PyTorch script** (4 lines):\r\n```python\r\nimport torch\r\n+ from accelerate import Accelerator\r\n\r\n+ accelerator = Accelerator()\r\n\r\n  model = torch.nn.Transformer()\r\n  optimizer = torch.optim.Adam(model.parameters())\r\n  dataloader = torch.utils.data.DataLoader(dataset)\r\n\r\n+ model, optimizer, dataloader = accelerator.prepare(model, optimizer, dataloader)\r\n\r\n  for batch in dataloader:\r\n      optimizer.zero_grad()\r\n      loss = model(batch)\r\n-     loss.backward()\r\n+     accelerator.backward(loss)\r\n      optimizer.step()\r\n```\r\n\r\n**Run** (single command):\r\n```bash\r\naccelerate launch train.py\r\n```\r\n\r\n## Common workflows\r\n\r\n### Workflow 1: From single GPU to multi-GPU\r\n\r\n**Original script**:\r\n```python\r\n# train.py\r\nimport torch\r\n\r\nmodel = torch.nn.Linear(10, 2).to('cuda')\r\noptimizer = torch.optim.Adam(model.parameters())\r\ndataloader = torch.utils.data.DataLoader(dataset, batch_size=32)\r\n\r\nfor epoch in range(10):\r\n    for batch in dataloader:\r\n        batch = batch.to('cuda')\r\n        optimizer.zero_grad()\r\n        loss = model(batch).mean()\r\n        loss.backward()\r\n        optimizer.step()\r\n```\r\n\r\n**With Accelerate** (4 lines added):\r\n```python\r\n# train.py\r\nimport torch\r\nfrom accelerate import Accelerator  # +1\r\n\r\naccelerator = Accelerator()  # +2\r\n\r\nmodel = torch.nn.Linear(10, 2)\r\noptimizer = torch.optim.Adam(model.parameters())\r\ndataloader = torch.utils.data.DataLoader(dataset, batch_size=32)\r\n\r\nmodel, optimizer, dataloader = accelerator.prepare(model, optimizer, dataloader)  # +3\r\n\r\nfor epoch in range(10):\r\n    for batch in dataloader:\r\n        # No .to('cuda') needed - automatic!\r\n        optimizer.zero_grad()\r\n        loss = model(batch).mean()\r\n        accelerator.backward(loss)  # +4\r\n        optimizer.step()\r\n```\r\n\r\n**Configure** (interactive):\r\n```bash\r\naccelerate config\r\n```\r\n\r\n**Questions**:\r\n- Which machine? (single\u002Fmulti GPU\u002FTPU\u002FCPU)\r\n- How many machines? (1)\r\n- Mixed precision? (no\u002Ffp16\u002Fbf16\u002Ffp8)\r\n- DeepSpeed? (no\u002Fyes)\r\n\r\n**Launch** (works on any setup):\r\n```bash\r\n# Single GPU\r\naccelerate launch train.py\r\n\r\n# Multi-GPU (8 GPUs)\r\naccelerate launch --multi_gpu --num_processes 8 train.py\r\n\r\n# Multi-node\r\naccelerate launch --multi_gpu --num_processes 16 \\\r\n  --num_machines 2 --machine_rank 0 \\\r\n  --main_process_ip $MASTER_ADDR \\\r\n  train.py\r\n```\r\n\r\n### Workflow 2: Mixed precision training\r\n\r\n**Enable FP16\u002FBF16**:\r\n```python\r\nfrom accelerate import Accelerator\r\n\r\n# FP16 (with gradient scaling)\r\naccelerator = Accelerator(mixed_precision='fp16')\r\n\r\n# BF16 (no scaling, more stable)\r\naccelerator = Accelerator(mixed_precision='bf16')\r\n\r\n# FP8 (H100+)\r\naccelerator = Accelerator(mixed_precision='fp8')\r\n\r\nmodel, optimizer, dataloader = accelerator.prepare(model, optimizer, dataloader)\r\n\r\n# Everything else is automatic!\r\nfor batch in dataloader:\r\n    with accelerator.autocast():  # Optional, done automatically\r\n        loss = model(batch)\r\n    accelerator.backward(loss)\r\n```\r\n\r\n### Workflow 3: DeepSpeed ZeRO integration\r\n\r\n**Enable DeepSpeed ZeRO-2**:\r\n```python\r\nfrom accelerate import Accelerator\r\n\r\naccelerator = Accelerator(\r\n    mixed_precision='bf16',\r\n    deepspeed_plugin={\r\n        \"zero_stage\": 2,  # ZeRO-2\r\n        \"offload_optimizer\": False,\r\n        \"gradient_accumulation_steps\": 4\r\n    }\r\n)\r\n\r\n# Same code as before!\r\nmodel, optimizer, dataloader = accelerator.prepare(model, optimizer, dataloader)\r\n```\r\n\r\n**Or via config**:\r\n```bash\r\naccelerate config\r\n# Select: DeepSpeed → ZeRO-2\r\n```\r\n\r\n**deepspeed_config.json**:\r\n```json\r\n{\r\n    \"fp16\": {\"enabled\": false},\r\n    \"bf16\": {\"enabled\": true},\r\n    \"zero_optimization\": {\r\n        \"stage\": 2,\r\n        \"offload_optimizer\": {\"device\": \"cpu\"},\r\n        \"allgather_bucket_size\": 5e8,\r\n        \"reduce_bucket_size\": 5e8\r\n    }\r\n}\r\n```\r\n\r\n**Launch**:\r\n```bash\r\naccelerate launch --config_file deepspeed_config.json train.py\r\n```\r\n\r\n### Workflow 4: FSDP (Fully Sharded Data Parallel)\r\n\r\n**Enable FSDP**:\r\n```python\r\nfrom accelerate import Accelerator, FullyShardedDataParallelPlugin\r\n\r\nfsdp_plugin = FullyShardedDataParallelPlugin(\r\n    sharding_strategy=\"FULL_SHARD\",  # ZeRO-3 equivalent\r\n    auto_wrap_policy=\"TRANSFORMER_AUTO_WRAP\",\r\n    cpu_offload=False\r\n)\r\n\r\naccelerator = Accelerator(\r\n    mixed_precision='bf16',\r\n    fsdp_plugin=fsdp_plugin\r\n)\r\n\r\nmodel, optimizer, dataloader = accelerator.prepare(model, optimizer, dataloader)\r\n```\r\n\r\n**Or via config**:\r\n```bash\r\naccelerate config\r\n# Select: FSDP → Full Shard → No CPU Offload\r\n```\r\n\r\n### Workflow 5: Gradient accumulation\r\n\r\n**Accumulate gradients**:\r\n```python\r\nfrom accelerate import Accelerator\r\n\r\naccelerator = Accelerator(gradient_accumulation_steps=4)\r\n\r\nmodel, optimizer, dataloader = accelerator.prepare(model, optimizer, dataloader)\r\n\r\nfor batch in dataloader:\r\n    with accelerator.accumulate(model):  # Handles accumulation\r\n        optimizer.zero_grad()\r\n        loss = model(batch)\r\n        accelerator.backward(loss)\r\n        optimizer.step()\r\n```\r\n\r\n**Effective batch size**: `batch_size * num_gpus * gradient_accumulation_steps`\r\n\r\n## When to use vs alternatives\r\n\r\n**Use Accelerate when**:\r\n- Want simplest distributed training\r\n- Need single script for any hardware\r\n- Use HuggingFace ecosystem\r\n- Want flexibility (DDP\u002FDeepSpeed\u002FFSDP\u002FMegatron)\r\n- Need quick prototyping\r\n\r\n**Key advantages**:\r\n- **4 lines**: Minimal code changes\r\n- **Unified API**: Same code for DDP, DeepSpeed, FSDP, Megatron\r\n- **Automatic**: Device placement, mixed precision, sharding\r\n- **Interactive config**: No manual launcher setup\r\n- **Single launch**: Works everywhere\r\n\r\n**Use alternatives instead**:\r\n- **PyTorch Lightning**: Need callbacks, high-level abstractions\r\n- **Ray Train**: Multi-node orchestration, hyperparameter tuning\r\n- **DeepSpeed**: Direct API control, advanced features\r\n- **Raw DDP**: Maximum control, minimal abstraction\r\n\r\n## Common issues\r\n\r\n**Issue: Wrong device placement**\r\n\r\nDon't manually move to device:\r\n```python\r\n# WRONG\r\nbatch = batch.to('cuda')\r\n\r\n# CORRECT\r\n# Accelerate handles it automatically after prepare()\r\n```\r\n\r\n**Issue: Gradient accumulation not working**\r\n\r\nUse context manager:\r\n```python\r\n# CORRECT\r\nwith accelerator.accumulate(model):\r\n    optimizer.zero_grad()\r\n    accelerator.backward(loss)\r\n    optimizer.step()\r\n```\r\n\r\n**Issue: Checkpointing in distributed**\r\n\r\nUse accelerator methods:\r\n```python\r\n# Save only on main process\r\nif accelerator.is_main_process:\r\n    accelerator.save_state('checkpoint\u002F')\r\n\r\n# Load on all processes\r\naccelerator.load_state('checkpoint\u002F')\r\n```\r\n\r\n**Issue: Different results with FSDP**\r\n\r\nEnsure same random seed:\r\n```python\r\nfrom accelerate.utils import set_seed\r\nset_seed(42)\r\n```\r\n\r\n## Advanced topics\r\n\r\n**Megatron integration**: See [references\u002Fmegatron-integration.md](references\u002Fmegatron-integration.md) for tensor parallelism, pipeline parallelism, and sequence parallelism setup.\r\n\r\n**Custom plugins**: See [references\u002Fcustom-plugins.md](references\u002Fcustom-plugins.md) for creating custom distributed plugins and advanced configuration.\r\n\r\n**Performance tuning**: See [references\u002Fperformance.md](references\u002Fperformance.md) for profiling, memory optimization, and best practices.\r\n\r\n## Hardware requirements\r\n\r\n- **CPU**: Works (slow)\r\n- **Single GPU**: Works\r\n- **Multi-GPU**: DDP (default), DeepSpeed, or FSDP\r\n- **Multi-node**: DDP, DeepSpeed, FSDP, Megatron\r\n- **TPU**: Supported\r\n- **Apple MPS**: Supported\r\n\r\n**Launcher requirements**:\r\n- **DDP**: `torch.distributed.run` (built-in)\r\n- **DeepSpeed**: `deepspeed` (pip install deepspeed)\r\n- **FSDP**: PyTorch 1.12+ (built-in)\r\n- **Megatron**: Custom setup\r\n\r\n## Resources\r\n\r\n- Docs: https:\u002F\u002Fhuggingface.co\u002Fdocs\u002Faccelerate\r\n- GitHub: https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Faccelerate\r\n- Version: 1.11.0+\r\n- Tutorial: \"Accelerate your scripts\"\r\n- Examples: https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Faccelerate\u002Ftree\u002Fmain\u002Fexamples\r\n- Used by: HuggingFace Transformers, TRL, PEFT, all HF libraries\r\n",{"name":5,"author":49,"license":15,"version":50,"metadata":51,"platforms":60,"description":7,"dependencies":62},"Orchestra Research","1.0.0",{"hermes":52},{"tags":53},[35,25,54,36,55,26,56,57,58,59],"Accelerate","FSDP","PyTorch","DDP","Unified API","Simple",[34,61,40],"macos",[63,64,65],"accelerate","torch","transformers","main","0d1434f89f4bc6b07e9d272accd8c1013768820f","archive\u002Fskill-md-reviewed\u002Fsha256\u002F48\u002F48c1420644eda125dacb9dba3aa0e06e1e449396141da2160fd4d48416179e01",0,"https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills","https:\u002F\u002Fgithub.com\u002Fzapabob","The Open University of JAPAN　2019-now\r\n\r\nTokyo University of Science　2007-2010","2026-10-26 16:29:01",{"source":8,"skillName":5,"commands":75,"prompt":85,"zipAvailable":86},[76,79,82],{"tool":77,"command":78},"npx","npx skills add zapabob\u002FSkills --skill huggingface-accelerate",{"tool":80,"command":81},"bunx","bunx skills add zapabob\u002FSkills --skill huggingface-accelerate",{"tool":83,"command":84},"pnpm","pnpm dlx skills add zapabob\u002FSkills --skill huggingface-accelerate","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills\u002Fblob\u002Fmain\u002Farchive\u002Fskill-md-reviewed\u002Fsha256\u002F48\u002F48c1420644eda125dacb9dba3aa0e06e1e449396141da2160fd4d48416179e01\u002FSKILL.md",true,[88,94,98,102,106,110,114,118,122,126,130,134,138,142,146,150,154,158,162,166,170,174,178,182,186,190,194,198,202,206,210,214,218,222,226,230,233,237,241,245,249,253,257,261,265,269,272,276,280,284,288,292,296,300,304,307,311,315,319,323,327,331,335,339,343,347,351,355,359,363,367,371,375,379,383,387,391,395,399],{"path":89,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":92,"sourceHtmlUrl":93},"bmad-skill-manifest.yaml","REFERENCE","参考文档",12,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fbmad-skill-manifest.yaml",{"path":95,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":96,"sourceHtmlUrl":97},"data\u002Fpositioning-explore.md",4015,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fdata\u002Fpositioning-explore.md",{"path":99,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":100,"sourceHtmlUrl":101},"data\u002Fpositioning-open-conversation.md",2468,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fdata\u002Fpositioning-open-conversation.md",{"path":103,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":104,"sourceHtmlUrl":105},"data\u002Fpositioning-reflect-confirm.md",3467,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fdata\u002Fpositioning-reflect-confirm.md",{"path":107,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":108,"sourceHtmlUrl":109},"data\u002Fpositioning-synthesize.md",4687,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fdata\u002Fpositioning-synthesize.md",{"path":111,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":112,"sourceHtmlUrl":113},"data\u002Ftone-of-voice-example.md",1834,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fdata\u002Ftone-of-voice-example.md",{"path":115,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":116,"sourceHtmlUrl":117},"data\u002Ftone-of-voice-output-template.md",1720,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fdata\u002Ftone-of-voice-output-template.md",{"path":119,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":120,"sourceHtmlUrl":121},"data\u002Fvision-explore.md",2320,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fdata\u002Fvision-explore.md",{"path":123,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":124,"sourceHtmlUrl":125},"data\u002Fvision-open-conversation.md",2442,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fdata\u002Fvision-open-conversation.md",{"path":127,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":128,"sourceHtmlUrl":129},"data\u002Fvision-reflect-confirm.md",1849,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fdata\u002Fvision-reflect-confirm.md",{"path":131,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":132,"sourceHtmlUrl":133},"data\u002Fvision-synthesize.md",2566,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fdata\u002Fvision-synthesize.md",{"path":135,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":136,"sourceHtmlUrl":137},"steps-c\u002Fstep-00-simplified-brief.md",5629,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-00-simplified-brief.md",{"path":139,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":140,"sourceHtmlUrl":141},"steps-c\u002Fstep-01-init.md",4318,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-01-init.md",{"path":143,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":144,"sourceHtmlUrl":145},"steps-c\u002Fstep-02-vision.md",4494,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-02-vision.md",{"path":147,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":148,"sourceHtmlUrl":149},"steps-c\u002Fstep-03-positioning.md",4876,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-03-positioning.md",{"path":151,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":152,"sourceHtmlUrl":153},"steps-c\u002Fstep-05-business-model.md",4754,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-05-business-model.md",{"path":155,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":156,"sourceHtmlUrl":157},"steps-c\u002Fstep-06-business-customers.md",4569,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-06-business-customers.md",{"path":159,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":160,"sourceHtmlUrl":161},"steps-c\u002Fstep-07-target-users.md",4838,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-07-target-users.md",{"path":163,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":164,"sourceHtmlUrl":165},"steps-c\u002Fstep-07a-product-concept.md",6062,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-07a-product-concept.md",{"path":167,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":168,"sourceHtmlUrl":169},"steps-c\u002Fstep-08-success-criteria.md",3872,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-08-success-criteria.md",{"path":171,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":172,"sourceHtmlUrl":173},"steps-c\u002Fstep-09-competitive-landscape.md",3831,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-09-competitive-landscape.md",{"path":175,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":176,"sourceHtmlUrl":177},"steps-c\u002Fstep-10-constraints.md",2716,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-10-constraints.md",{"path":179,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":180,"sourceHtmlUrl":181},"steps-c\u002Fstep-10a-platform-strategy.md",5607,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-10a-platform-strategy.md",{"path":183,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":184,"sourceHtmlUrl":185},"steps-c\u002Fstep-11-tone-of-voice.md",6845,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-11-tone-of-voice.md",{"path":187,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":188,"sourceHtmlUrl":189},"steps-c\u002Fstep-12-create-product-brief.md",7026,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-12-create-product-brief.md",{"path":191,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":192,"sourceHtmlUrl":193},"steps-c\u002Fstep-13-content-init.md",4188,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-13-content-init.md",{"path":195,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":196,"sourceHtmlUrl":197},"steps-c\u002Fstep-14-personality.md",4847,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-14-personality.md",{"path":199,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":200,"sourceHtmlUrl":201},"steps-c\u002Fstep-15-tone.md",4725,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-15-tone.md",{"path":203,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":204,"sourceHtmlUrl":205},"steps-c\u002Fstep-16-languages.md",4843,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-16-languages.md",{"path":207,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":208,"sourceHtmlUrl":209},"steps-c\u002Fstep-17-seo-keywords.md",6563,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-17-seo-keywords.md",{"path":211,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":212,"sourceHtmlUrl":213},"steps-c\u002Fstep-17a-content-structure.md",4752,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-17a-content-structure.md",{"path":215,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":216,"sourceHtmlUrl":217},"steps-c\u002Fstep-18-create-content-document.md",5396,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-18-create-content-document.md",{"path":219,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":220,"sourceHtmlUrl":221},"steps-c\u002Fstep-19-inspiration-workshop.md",4411,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-19-inspiration-workshop.md",{"path":223,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":224,"sourceHtmlUrl":225},"steps-c\u002Fstep-20-visual-init.md",4476,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-20-visual-init.md",{"path":227,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":228,"sourceHtmlUrl":229},"steps-c\u002Fstep-21-existing-brand.md",4914,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-21-existing-brand.md",{"path":231,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":228,"sourceHtmlUrl":232},"steps-c\u002Fstep-22-references.md","https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-22-references.md",{"path":234,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":235,"sourceHtmlUrl":236},"steps-c\u002Fstep-23-design-style.md",5361,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-23-design-style.md",{"path":238,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":239,"sourceHtmlUrl":240},"steps-c\u002Fstep-24-layout-effects.md",5060,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-24-layout-effects.md",{"path":242,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":243,"sourceHtmlUrl":244},"steps-c\u002Fstep-25-imagery.md",5117,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-25-imagery.md",{"path":246,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":247,"sourceHtmlUrl":248},"steps-c\u002Fstep-26-create-visual-document.md",4831,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-26-create-visual-document.md",{"path":250,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":251,"sourceHtmlUrl":252},"steps-c\u002Fstep-27-platform-init.md",4341,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-27-platform-init.md",{"path":254,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":255,"sourceHtmlUrl":256},"steps-c\u002Fstep-28-tech-stack.md",4620,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-28-tech-stack.md",{"path":258,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":259,"sourceHtmlUrl":260},"steps-c\u002Fstep-29-integrations.md",4524,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-29-integrations.md",{"path":262,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":263,"sourceHtmlUrl":264},"steps-c\u002Fstep-30-contact-strategy.md",4710,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-30-contact-strategy.md",{"path":266,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":267,"sourceHtmlUrl":268},"steps-c\u002Fstep-31-multilingual.md",5628,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-31-multilingual.md",{"path":270,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":224,"sourceHtmlUrl":271},"steps-c\u002Fstep-32-create-platform-document.md","https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-32-create-platform-document.md",{"path":273,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":274,"sourceHtmlUrl":275},"steps-c\u002Fstep-33-analyze-brief.md",3274,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-33-analyze-brief.md",{"path":277,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":278,"sourceHtmlUrl":279},"steps-c\u002Fstep-34-create-summary.md",3655,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-34-create-summary.md",{"path":281,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":282,"sourceHtmlUrl":283},"steps-c\u002Fstep-35-update-design-log.md",4475,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-35-update-design-log.md",{"path":285,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":286,"sourceHtmlUrl":287},"steps-c\u002Fstep-36-provide-activation.md",3636,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-c\u002Fstep-36-provide-activation.md",{"path":289,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":290,"sourceHtmlUrl":291},"steps-v\u002Fstep-01-brief-completeness.md",4208,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-v\u002Fstep-01-brief-completeness.md",{"path":293,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":294,"sourceHtmlUrl":295},"steps-v\u002Fstep-02-trigger-map-consistency.md",4364,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-v\u002Fstep-02-trigger-map-consistency.md",{"path":297,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":298,"sourceHtmlUrl":299},"steps-v\u002Fstep-03-seo-strategy.md",4098,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-v\u002Fstep-03-seo-strategy.md",{"path":301,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":302,"sourceHtmlUrl":303},"steps-v\u002Fstep-04-content-language.md",4221,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-v\u002Fstep-04-content-language.md",{"path":305,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":290,"sourceHtmlUrl":306},"steps-v\u002Fstep-05-visual-direction.md","https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-v\u002Fstep-05-visual-direction.md",{"path":308,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":309,"sourceHtmlUrl":310},"steps-v\u002Fstep-06-platform-requirements.md",4915,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fsteps-v\u002Fstep-06-platform-requirements.md",{"path":312,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":313,"sourceHtmlUrl":314},"templates\u002F00-project-info.template.md",1914,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Ftemplates\u002F00-project-info.template.md",{"path":316,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":317,"sourceHtmlUrl":318},"templates\u002Fcontent-language.template.md",4773,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Ftemplates\u002Fcontent-language.template.md",{"path":320,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":321,"sourceHtmlUrl":322},"templates\u002Fcontract.template.md",8811,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Ftemplates\u002Fcontract.template.md",{"path":324,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":325,"sourceHtmlUrl":326},"templates\u002Finspiration-analysis.template.md",1811,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Ftemplates\u002Finspiration-analysis.template.md",{"path":328,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":329,"sourceHtmlUrl":330},"templates\u002Fpitch.template.md",1263,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Ftemplates\u002Fpitch.template.md",{"path":332,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":333,"sourceHtmlUrl":334},"templates\u002Fplatform-requirements.template.md",4619,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Ftemplates\u002Fplatform-requirements.template.md",{"path":336,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":337,"sourceHtmlUrl":338},"templates\u002Fplatform-requirements.template.yaml",2281,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Ftemplates\u002Fplatform-requirements.template.yaml",{"path":340,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":341,"sourceHtmlUrl":342},"templates\u002Fproject-brief-dialog\u002F00-context.md",1533,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Ftemplates\u002Fproject-brief-dialog\u002F00-context.md",{"path":344,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":345,"sourceHtmlUrl":346},"templates\u002Fproject-brief-dialog\u002F02-vision.md",1552,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Ftemplates\u002Fproject-brief-dialog\u002F02-vision.md",{"path":348,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":349,"sourceHtmlUrl":350},"templates\u002Fproject-brief-dialog\u002F03-users.md",1216,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Ftemplates\u002Fproject-brief-dialog\u002F03-users.md",{"path":352,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":353,"sourceHtmlUrl":354},"templates\u002Fproject-brief-dialog\u002F04-concept.md",1282,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Ftemplates\u002Fproject-brief-dialog\u002F04-concept.md",{"path":356,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":357,"sourceHtmlUrl":358},"templates\u002Fproject-brief-dialog\u002F06-inspiration.md",1580,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Ftemplates\u002Fproject-brief-dialog\u002F06-inspiration.md",{"path":360,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":361,"sourceHtmlUrl":362},"templates\u002Fproject-brief-dialog\u002F07-positioning.md",1293,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Ftemplates\u002Fproject-brief-dialog\u002F07-positioning.md",{"path":364,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":365,"sourceHtmlUrl":366},"templates\u002Fproject-brief-dialog\u002Fdecisions.md",1354,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Ftemplates\u002Fproject-brief-dialog\u002Fdecisions.md",{"path":368,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":369,"sourceHtmlUrl":370},"templates\u002Fproject-brief-dialog\u002Fprogress-tracker.md",2328,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Ftemplates\u002Fproject-brief-dialog\u002Fprogress-tracker.md",{"path":372,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":373,"sourceHtmlUrl":374},"templates\u002Fproject-brief-dialog\u002FUSAGE.md",1886,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Ftemplates\u002Fproject-brief-dialog\u002FUSAGE.md",{"path":376,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":377,"sourceHtmlUrl":378},"templates\u002Fproject-brief.template.md",3629,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Ftemplates\u002Fproject-brief.template.md",{"path":380,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":381,"sourceHtmlUrl":382},"templates\u002Fservice-agreement.template.md",8082,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Ftemplates\u002Fservice-agreement.template.md",{"path":384,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":385,"sourceHtmlUrl":386},"templates\u002Fsignoff.template.md",2910,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Ftemplates\u002Fsignoff.template.md",{"path":388,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":389,"sourceHtmlUrl":390},"templates\u002Fsimplified-brief.template.md",664,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Ftemplates\u002Fsimplified-brief.template.md",{"path":392,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":393,"sourceHtmlUrl":394},"templates\u002Fvisual-direction.template.md",4115,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Ftemplates\u002Fvisual-direction.template.md",{"path":396,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":397,"sourceHtmlUrl":398},"workflow-validate.md",1409,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fworkflow-validate.md",{"path":400,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":401,"sourceHtmlUrl":402},"workflow.md",4353,"https:\u002F\u002Fgithub.com\u002Fdanielmtt9\u002Fbloomcornerkiddies\u002Fblob\u002Fmain\u002F.cline\u002Fskills\u002Fwds-1-project-brief\u002Fworkflow.md",[404,434,469,495,521,538],{"skillNo":405,"name":406,"nameZh":6,"description":407,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":408,"categories":409,"keywords":419,"publishedAt":44},"01M6GJTE1GESGGF9K028Y7753E","xurl","X\u002FTwitter via xurl CLI: post, search, DM, media, v2 API.","https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills\u002Fblob\u002Fmain\u002Fregistry\u002Fskills\u002Fxurl\u002Fvariants\u002Fhermes-agent\u002FSKILL.md",[410,411,414],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},{"categoryNo":412,"categoryName":413,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000023","Skill与智能体开发",{"categoryNo":415,"categoryName":416,"taxonomyType":417,"taxonomyTypeDesc":418},"01JBSEEDCAT000000000000030","开发工程师","PROFESSION","适用职业",[406,420,61,421,422,34,40,423,424,425,426,30,427,428,429,430,66,431,432,433],"openclaw","cli","github","official","search","https","social-media","Media","hermes","hermes agent","twitter","social","via","post",{"skillNo":435,"name":436,"nameZh":6,"description":437,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":438,"categories":439,"keywords":451,"publishedAt":44},"01M6GJT8S1PXQ8WPXT1HSNBWH2","writing-plans","Write implementation plans: bite-sized tasks, paths, code.","https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills\u002Fblob\u002Fmain\u002Fregistry\u002Fskills\u002Fwriting-plans\u002Fvariants\u002Fhermes-agent\u002FSKILL.md",[440,443,446,447,448],{"categoryNo":441,"categoryName":442,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000015","视觉设计",{"categoryNo":444,"categoryName":445,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000018","写作与文案",{"categoryNo":412,"categoryName":413,"taxonomyType":21,"taxonomyTypeDesc":22},{"categoryNo":415,"categoryName":416,"taxonomyType":417,"taxonomyTypeDesc":418},{"categoryNo":449,"categoryName":450,"taxonomyType":417,"taxonomyTypeDesc":418},"01JBSEEDCAT000000000000036","内容创作者",[452,453,454,455,61,456,34,457,40,458,459,460,461,462,463,464,465,466,467,468],"writing","writing plans","plans","design","review","workflow","documentation","tasks","development","code","planning","code-review","subagent","write","superpowers","paths","test-driven-development",{"skillNo":470,"name":471,"nameZh":6,"description":472,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":473,"categories":474,"keywords":478,"publishedAt":44},"01M6GJST0H5X7WHBJ3BZK0DVKH","whisper","OpenAI's general-purpose speech recognition model. Supports 99 languages, transcription, translation to English, and language identification. Six model sizes from tiny (39M params) to large (1550M params). Use for speech-to-text, podcast transcription, or multilingual audio processing. Best for robust, multilingual ASR.","https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills\u002Fblob\u002Fmain\u002Fregistry\u002Fskills\u002Fwhisper\u002Fvariants\u002Fhermes-agent-optional\u002FSKILL.md",[475],{"categoryNo":476,"categoryName":477,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000016","图像与多媒体",[471,61,34,479,480,481,31,482,483,484,485,486,487,488,489,490,491,492,493,494],"openai","speech","multimodal","ASR","Audio","text","model","speech recognition","use","general","multilingual","translation","english","identification","processing","Audio Processing",{"skillNo":496,"name":497,"nameZh":6,"description":498,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":499,"categories":500,"keywords":504,"publishedAt":44},"01M6GJSMRQ8EETZTYMG72P70EA","weights-and-biases","W&B: log ML experiments, sweeps, model registry, dashboards.","https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills\u002Fblob\u002Fmain\u002Fregistry\u002Fskills\u002Fweights-and-biases\u002Fvariants\u002Fhermes-agent\u002FSKILL.md",[501],{"categoryNo":502,"categoryName":503,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000013","分析与可视化",[505,61,506,507,40,38,508,509,510,511,512,485,513,514,515,516,517,518,519,520],"weights and biases","tensorflow","experiments","visualization","registry","real-time","tuning","dashboards","experiment","log","time","Experiment Tracking","Model Registry","Hyperparameter Tuning","WandB","Real-Time Visualization",{"skillNo":522,"name":523,"nameZh":6,"description":524,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":525,"categories":526,"keywords":528,"publishedAt":44},"01M6GJS19KDPRY112JTR326S06","watchers","Poll RSS, JSON APIs, and GitHub with watermark dedup.","https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills\u002Fblob\u002Fmain\u002Fregistry\u002Fskills\u002Fwatchers\u002Fvariants\u002Fhermes-agent-optional\u002FSKILL.md",[527],{"categoryNo":412,"categoryName":413,"taxonomyType":21,"taxonomyTypeDesc":22},[523,61,529,422,34,530,531,532,533,534,535,428,429,536,537],"terminal","automation","http","monitoring","cron","json","rss","devops","with",{"skillNo":539,"name":540,"nameZh":6,"description":541,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":542,"categories":543,"keywords":546,"publishedAt":44},"01M6GJRHJD2A6Q830SWF9W48BC","serving-llms-vllm","vLLM: high-throughput LLM serving, OpenAI API, quantization.","https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills\u002Fblob\u002Fmain\u002Fregistry\u002Fskills\u002Fvllm\u002Fvariants\u002Fhermes-agent\u002FSKILL.md",[544,545],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},{"categoryNo":415,"categoryName":416,"taxonomyType":417,"taxonomyTypeDesc":418},[547,548,549,34,43,479,550,551,552,31,553,30,554,555,556,557,558,559,560,561],"llms","vLLM","serving llms vllm","openai api","tensor","Inference Serving","Production","llm serving","High Throughput","PagedAttention","Continuous Batching","Quantization","Tensor Parallelism","high-throughput","throughput"]