{
  "schemaVersion": "1.0",
  "item": {
    "slug": "tensorflow",
    "name": "TensorFlow",
    "source": "tencent",
    "type": "skill",
    "category": "AI 智能",
    "sourceUrl": "https://clawhub.ai/ivangdavila/tensorflow",
    "canonicalUrl": "https://clawhub.ai/ivangdavila/tensorflow",
    "targetPlatform": "OpenClaw"
  },
  "install": {
    "downloadMode": "redirect",
    "downloadUrl": "/downloads/tensorflow",
    "sourceDownloadUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=tensorflow",
    "sourcePlatform": "tencent",
    "targetPlatform": "OpenClaw",
    "installMethod": "Manual import",
    "extraction": "Extract archive",
    "prerequisites": [
      "OpenClaw"
    ],
    "packageFormat": "ZIP package",
    "includedAssets": [
      "SKILL.md"
    ],
    "primaryDoc": "SKILL.md",
    "quickSetup": [
      "Download the package from Yavira.",
      "Extract the archive and review SKILL.md first.",
      "Import or place the package into your OpenClaw setup."
    ],
    "agentAssist": {
      "summary": "Hand the extracted package to your coding agent with a concrete install brief instead of figuring it out manually.",
      "steps": [
        "Download the package from Yavira.",
        "Extract it into a folder your agent can access.",
        "Paste one of the prompts below and point your agent at the extracted folder."
      ],
      "prompts": [
        {
          "label": "New install",
          "body": "I downloaded a skill package from Yavira. Read SKILL.md from the extracted folder and install it by following the included instructions. Tell me what you changed and call out any manual steps you could not complete."
        },
        {
          "label": "Upgrade existing",
          "body": "I downloaded an updated skill package from Yavira. Read SKILL.md from the extracted folder, compare it with my current installation, and upgrade it while preserving any custom configuration unless the package docs explicitly say otherwise. Summarize what changed and any follow-up checks I should run."
        }
      ]
    },
    "sourceHealth": {
      "source": "tencent",
      "slug": "tensorflow",
      "status": "healthy",
      "reason": "direct_download_ok",
      "recommendedAction": "download",
      "checkedAt": "2026-05-04T02:40:08.565Z",
      "expiresAt": "2026-05-11T02:40:08.565Z",
      "httpStatus": 200,
      "finalUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=tensorflow",
      "contentType": "application/zip",
      "probeMethod": "head",
      "details": {
        "probeUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=tensorflow",
        "contentDisposition": "attachment; filename=\"tensorflow-1.0.0.zip\"",
        "redirectLocation": null,
        "bodySnippet": null,
        "slug": "tensorflow"
      },
      "scope": "item",
      "summary": "Item download looks usable.",
      "detail": "Yavira can redirect you to the upstream package for this item.",
      "primaryActionLabel": "Download for OpenClaw",
      "primaryActionHref": "/downloads/tensorflow"
    },
    "validation": {
      "installChecklist": [
        "Use the Yavira download entry.",
        "Review SKILL.md after the package is downloaded.",
        "Confirm the extracted package contains the expected setup assets."
      ],
      "postInstallChecks": [
        "Confirm the extracted package includes the expected docs or setup files.",
        "Validate the skill or prompts are available in your target agent workspace.",
        "Capture any manual follow-up steps the agent could not complete."
      ]
    },
    "downloadPageUrl": "https://openagent3.xyz/downloads/tensorflow",
    "agentPageUrl": "https://openagent3.xyz/skills/tensorflow/agent",
    "manifestUrl": "https://openagent3.xyz/skills/tensorflow/agent.json",
    "briefUrl": "https://openagent3.xyz/skills/tensorflow/agent.md"
  },
  "agentAssist": {
    "summary": "Hand the extracted package to your coding agent with a concrete install brief instead of figuring it out manually.",
    "steps": [
      "Download the package from Yavira.",
      "Extract it into a folder your agent can access.",
      "Paste one of the prompts below and point your agent at the extracted folder."
    ],
    "prompts": [
      {
        "label": "New install",
        "body": "I downloaded a skill package from Yavira. Read SKILL.md from the extracted folder and install it by following the included instructions. Tell me what you changed and call out any manual steps you could not complete."
      },
      {
        "label": "Upgrade existing",
        "body": "I downloaded an updated skill package from Yavira. Read SKILL.md from the extracted folder, compare it with my current installation, and upgrade it while preserving any custom configuration unless the package docs explicitly say otherwise. Summarize what changed and any follow-up checks I should run."
      }
    ]
  },
  "documentation": {
    "source": "clawhub",
    "primaryDoc": "SKILL.md",
    "sections": [
      {
        "title": "tf.function Retracing",
        "body": "New input shape/dtype causes retrace — expensive, prints warning\nUse input_signature for fixed shapes — @tf.function(input_signature=[tf.TensorSpec(...)])\nPython values retrace — pass as tensors, not Python ints/floats\nAvoid Python side effects in tf.function — only runs once during tracing"
      },
      {
        "title": "GPU Memory",
        "body": "TensorFlow grabs all GPU memory by default — set memory_growth=True before any ops\ntf.config.experimental.set_memory_growth(gpu, True) — must be called before GPU init\nOOM with large models — reduce batch size or use gradient checkpointing\nCUDA_VISIBLE_DEVICES=\"\" to force CPU — for testing without GPU"
      },
      {
        "title": "Data Pipeline",
        "body": "tf.data.Dataset without .prefetch() — CPU/GPU idle time between batches\n.cache() after expensive ops — but before random augmentation\n.batch() before .map() for vectorized ops — faster than per-element\nnum_parallel_calls=tf.data.AUTOTUNE — parallel preprocessing\nDataset iteration in eager mode is slow — use in tf.function or model.fit"
      },
      {
        "title": "Shape Issues",
        "body": "First dimension is batch — None for variable batch size in Input layer\nmodel.build(input_shape) if not using Input layer — or first call errors\nReshape errors unclear — tf.debugging.assert_shapes() for debugging\nBroadcasting silently succeeds — may hide shape bugs"
      },
      {
        "title": "Gradient Tape",
        "body": "Variables watched by default — tensors need tape.watch(tensor)\npersistent=True for multiple gradients — otherwise tape consumed after first use\ntape.gradient returns None if no path — check for disconnected graph\n@tf.custom_gradient for custom backward — not all ops have gradients"
      },
      {
        "title": "Training Gotchas",
        "body": "model.trainable = False after compile does nothing — set before compile\nBatchNorm behaves differently in training vs inference — training=True/False matters\nmodel.fit shuffles by default — shuffle=False for time series\nvalidation_split takes from end — shuffle data first if order matters"
      },
      {
        "title": "Saving Models",
        "body": "model.save() saves everything — architecture, weights, optimizer state\nmodel.save_weights() only weights — need model code to restore\nSavedModel format for serving — tf.saved_model.save(model, path)\nH5 format limited — doesn't save custom objects well, use SavedModel"
      },
      {
        "title": "Common Mistakes",
        "body": "Mixing Keras and raw tf ops incorrectly — use layers.Lambda to wrap tf ops in Sequential\ntf.print vs Python print — Python print only runs at trace time in tf.function\nNumPy ops in graph — use tf ops, numpy executes eagerly only\nLoss returns scalar per sample — Keras averages, custom loops may need tf.reduce_mean"
      }
    ],
    "body": "tf.function Retracing\nNew input shape/dtype causes retrace — expensive, prints warning\nUse input_signature for fixed shapes — @tf.function(input_signature=[tf.TensorSpec(...)])\nPython values retrace — pass as tensors, not Python ints/floats\nAvoid Python side effects in tf.function — only runs once during tracing\nGPU Memory\nTensorFlow grabs all GPU memory by default — set memory_growth=True before any ops\ntf.config.experimental.set_memory_growth(gpu, True) — must be called before GPU init\nOOM with large models — reduce batch size or use gradient checkpointing\nCUDA_VISIBLE_DEVICES=\"\" to force CPU — for testing without GPU\nData Pipeline\ntf.data.Dataset without .prefetch() — CPU/GPU idle time between batches\n.cache() after expensive ops — but before random augmentation\n.batch() before .map() for vectorized ops — faster than per-element\nnum_parallel_calls=tf.data.AUTOTUNE — parallel preprocessing\nDataset iteration in eager mode is slow — use in tf.function or model.fit\nShape Issues\nFirst dimension is batch — None for variable batch size in Input layer\nmodel.build(input_shape) if not using Input layer — or first call errors\nReshape errors unclear — tf.debugging.assert_shapes() for debugging\nBroadcasting silently succeeds — may hide shape bugs\nGradient Tape\nVariables watched by default — tensors need tape.watch(tensor)\npersistent=True for multiple gradients — otherwise tape consumed after first use\ntape.gradient returns None if no path — check for disconnected graph\n@tf.custom_gradient for custom backward — not all ops have gradients\nTraining Gotchas\nmodel.trainable = False after compile does nothing — set before compile\nBatchNorm behaves differently in training vs inference — training=True/False matters\nmodel.fit shuffles by default — shuffle=False for time series\nvalidation_split takes from end — shuffle data first if order matters\nSaving Models\nmodel.save() saves everything — architecture, weights, optimizer state\nmodel.save_weights() only weights — need model code to restore\nSavedModel format for serving — tf.saved_model.save(model, path)\nH5 format limited — doesn't save custom objects well, use SavedModel\nCommon Mistakes\nMixing Keras and raw tf ops incorrectly — use layers.Lambda to wrap tf ops in Sequential\ntf.print vs Python print — Python print only runs at trace time in tf.function\nNumPy ops in graph — use tf ops, numpy executes eagerly only\nLoss returns scalar per sample — Keras averages, custom loops may need tf.reduce_mean"
  },
  "trust": {
    "sourceLabel": "tencent",
    "provenanceUrl": "https://clawhub.ai/ivangdavila/tensorflow",
    "publisherUrl": "https://clawhub.ai/ivangdavila/tensorflow",
    "owner": "ivangdavila",
    "version": "1.0.0",
    "license": null,
    "verificationStatus": "Indexed source record"
  },
  "links": {
    "detailUrl": "https://openagent3.xyz/skills/tensorflow",
    "downloadUrl": "https://openagent3.xyz/downloads/tensorflow",
    "agentUrl": "https://openagent3.xyz/skills/tensorflow/agent",
    "manifestUrl": "https://openagent3.xyz/skills/tensorflow/agent.json",
    "briefUrl": "https://openagent3.xyz/skills/tensorflow/agent.md"
  }
}