Agent Skillswaybarrios/opencode-power-pack › train-sentence-transformers

train-sentence-transformers

GitHub

用于训练或微调 SentenceTransformer 模型(如双编码器、交叉编码器、稀疏编码器)的路由技能,指导加载相关参考文档和示例脚本以配置损失函数、评估器及训练参数。

skills/train-sentence-transformers/SKILL.md waybarrios/opencode-power-pack

Trigger Scenarios

需要训练或微调句子嵌入模型 请求涉及 CrossEncoder 重排序模型训练 需要配置 SPLADE 稀疏编码模型

Install

npx skills add waybarrios/opencode-power-pack --skill train-sentence-transformers -g -y
More Options

Use without installing

npx skills use waybarrios/opencode-power-pack@train-sentence-transformers

指定 Agent (Claude Code)

npx skills add waybarrios/opencode-power-pack --skill train-sentence-transformers -a claude-code -g -y

安装 repo 全部 skill

npx skills add waybarrios/opencode-power-pack --all -g -y

预览 repo 内 skill

npx skills add waybarrios/opencode-power-pack --list

SKILL.md

Frontmatter
{
    "name": "train-sentence-transformers",
    "license": "Apache-2.0 (modified; see UPSTREAMS.json)",
    "description": "Train or fine-tune SentenceTransformer bi-encoders, CrossEncoder rerankers, or SparseEncoder models, including losses, negatives, evaluation, distillation, LoRA, and Matryoshka. Use for sentence-transformers training tasks."
}

Train a sentence-transformers Model

This SKILL.md is a router, not a manual. It tells you which references and example scripts to load for your task. The actual content — recommended losses, evaluators, training-script structure, model selection, training-arg knobs, troubleshooting — lives in references/ and scripts/.

Do not synthesize a training script from this file alone. Open the per-type production template (scripts/train_<type>_example.py) and copy it as your starting point. The templates contain load-bearing scaffolding (autocast helper, model-card class, logger silencing list, force=True, seed, TF32, version-compatible imports, named-evaluator metric handling) that prior agent runs have repeatedly missed when rolling their own from a synthesized snippet.

1. Identify the model type

Tag Class What it does When to pick
[SentenceTransformer] SentenceTransformer (bi-encoder) Maps each input to a fixed-dim dense vector Retrieval, similarity, clustering, classification, paraphrase mining, dedup
[CrossEncoder] CrossEncoder (reranker) Scores (query, passage) pairs jointly Two-stage retrieval (rerank top-100 from bi-encoder), pair classification
[SparseEncoder] SparseEncoder (SPLADE) Sparse vectors over the vocabulary Learned-sparse retrieval, inverted-index backends (Elasticsearch / OpenSearch / Lucene)

Tiebreakers when the request is ambiguous: "embedding model" / "vector search" / "similarity" → [SentenceTransformer]. "rerank" / "ranker" / "two-stage" → [CrossEncoder]. "SPLADE" / "sparse" / "inverted index" → [SparseEncoder]. If still unclear, ask.

2. Required reading

Read these in full before writing any code. Do not triage by perceived relevance.

Per-type — always required

[SentenceTransformer]

  • references/losses_sentence_transformer.md — loss-to-data-shape mapping; BatchSamplers.NO_DUPLICATES requirement for MNRL-family; Cached*gradient_checkpointing incompatibility.
  • references/evaluators_sentence_transformer.md — evaluator-to-task mapping; metric_for_best_model key construction (named vs unnamed); per-evaluator primary_metric values.
  • references/model_architectures.md — encoder vs decoder vs static vs Router pipelines; pooling rules (mean / cls / lasttoken); auto-mean-pooling behavior for fresh-start MLM bases.
  • scripts/train_sentence_transformer_example.py — production template; copy this as your starting point.

[CrossEncoder]

  • references/losses_cross_encoder.md — pointwise / pairwise / listwise / distillation; pos_weight derivation; activation_fn=Identity() mandatory for non-BCE losses (silent eval-rank collapse otherwise).
  • references/evaluators_cross_encoder.mdCrossEncoderRerankingEvaluator recipe; named-evaluator key format eval_{name}_{primary_metric}.
  • scripts/train_cross_encoder_example.py — production template; copy this as your starting point.

[SparseEncoder]

  • references/losses_sparse_encoder.mdSpladeLoss wrapper requirement; FLOPS regularizer weights; smoke-test active-dim ramp behavior.
  • references/evaluators_sparse_encoder.mdSparseNanoBEIREvaluator (English-only) and the in-domain alternative; eval_{name}_{primary_metric} key format.
  • scripts/train_sparse_encoder_example.py — production template; copy this as your starting point.

Cross-cutting — always required (regardless of task)

  • references/training_args.mdTrainingArguments knobs, precision rules (load fp32 + autocast bf16/fp16; never torch_dtype=bfloat16), warmup_steps (float) vs deprecated warmup_ratio, save_steps must be a multiple of eval_steps for load_best_model_at_end, schedulers, HPO, tracker, resume, hub-push variants.
  • references/dataset_formats.md — column-matching rules (label name auto-detection; column-order-not-name); reshaping recipes; hard-negative mining options.
  • references/base_model_selection.md — discovery commands; per-type model namespaces; ModernBERT-family max_seq_length=8192 trap; datasets >= 4 script-loader rejection; non-English starting-point shortcuts.
  • references/troubleshooting.md — symptom-indexed failure recipes. Skim the section headings on every run, even a healthy one; the "Metrics don't improve" and "Hub push fails" entries cover bugs that bite frequently and are cheaper to recognize before they fire than to debug after.

Cross-cutting — load when applicable

  • references/hardware_guide.md — VRAM sizing, multi-GPU, FSDP / DeepSpeed, HF Jobs flavors. Required for >24GB models, multi-GPU, or HF Jobs runs.
  • references/hf_jobs_execution.md — required when running on HF Jobs.
  • references/prompts_and_instructions.md — required when using prompt-tuned bases (E5, BGE, GTE, Qwen3-Embedding, Instructor, Nomic, etc.) or adding query: / passage: style prefixes.

Variant scripts (open when the task matches)

  • [SentenceTransformer] scripts/train_sentence_transformer_<matryoshka|multi_dataset|with_lora|distillation|make_multilingual|static_embedding>_example.py.
  • [CrossEncoder] scripts/train_cross_encoder_<distillation|listwise>_example.py.
  • [SparseEncoder] scripts/train_sparse_encoder_distillation_example.py.
  • Hard-negative mining CLI — scripts/mine_hard_negatives.py.

3. Defaults

Override only if the user specifies otherwise:

  • Local execution. Pitch HF Jobs only if local hardware can't fit the job.
  • Single run. After it completes, propose experimentation if the user would benefit (weak/marginal verdict, "see how high you can push it" framing, etc.). Iteration rules in references/training_args.md (Experimentation section).
  • Public Hub push at end-of-run, wrapped in try-except. On HF Jobs (ephemeral env) ALSO enable in-trainer push (push_to_hub=True + hub_strategy="every_save"); details in references/hf_jobs_execution.md.

4. Constraints the produced script must satisfy

These are non-negotiable contracts. Implementation lives in the production templates and references — do not reinvent.

  • Capture the pre-training evaluator score as baseline_eval before trainer.train().
  • Emit a single end-of-run line: VERDICT: WIN|MARGINAL|REGRESSION | score=... | baseline=... | delta=.... A monitor scrapes for this.
  • Silence httpx, httpcore, huggingface_hub, urllib3, filelock, fsspec to WARNING (otherwise HF download URLs flood the agent's context).
  • Tee logs to logs/{RUN_NAME}.log.
  • End with model.push_to_hub(...) wrapped in try/except.
  • Smoke-test before any long run (max_steps=1 + tiny dataset slice). The production templates show one common pattern (SMOKE_TEST env var).
  • [CrossEncoder] Include EarlyStoppingCallback(patience>=3) — CE rerankers often peak mid-training and regress.
  • [SparseEncoder] Log query_active_dims / corpus_active_dims on the verdict line; high nDCG with collapsed sparsity is not a win. The keys come back name-prefixed (e.g. ..._query_active_dims); use suffix matching to pluck them — see the SPARSE production template for the exact pattern.

5. Workflow

  1. Identify the model type (§1). Ask if ambiguous.
  2. Load the §2 required-reading files for that type.
  3. Open scripts/train_<type>_example.py and copy it as your starting point.
  4. Replace MODEL_NAME, DATASET_NAME, RUN_NAME, the loss, and the evaluator with the user's task. Cross-check loss/data-shape match against references/losses_<type>.md; cross-check the metric_for_best_model key against references/evaluators_<type>.md (named evaluators format the key as eval_{name}_{primary_metric}).
  5. Smoke-test (max_steps=1).
  6. Run.
  7. After the run, append to logs/experiments.md and propose iteration if the verdict is weak/marginal.

Prerequisites

pip install "sentence-transformers[train]>=5.0"        # add [train,image] / [audio] / [video] for [SentenceTransformer] multimodal
pip install trackio                                    # optional tracker; or wandb / tensorboard / mlflow
hf auth login                                          # or set HF_TOKEN with write scope (for Hub push)

GPU strongly recommended. CPU works only for demos and [SentenceTransformer] StaticEmbedding.

Version History

  • f198a18 Current 2026-08-16 09:13

Same Skill Collection

skills/agentic-actions-auditor/SKILL.md
skills/agents-md-revise/SKILL.md
skills/ai-slop/SKILL.md
skills/code-architect/SKILL.md
skills/code-explorer/SKILL.md
skills/code-quality/SKILL.md
skills/code-review/SKILL.md
skills/code-reviewer/SKILL.md
skills/codeql/SKILL.md
skills/design-patterns/SKILL.md
skills/differential-review/SKILL.md
skills/feature-dev/SKILL.md
skills/fp-check/SKILL.md
skills/frontend-design/SKILL.md
skills/hf-cli/SKILL.md
skills/hf-cloud-aws-context-discovery/SKILL.md
skills/hf-cloud-python-env-setup/SKILL.md
skills/hf-cloud-sagemaker-deployment-planner/SKILL.md
skills/hf-cloud-sagemaker-iam-preflight/SKILL.md
skills/hf-cloud-sagemaker-production-defaults/SKILL.md
skills/hf-cloud-serving-image-selection/SKILL.md
skills/hf-mem/SKILL.md
skills/huggingface-best/SKILL.md
skills/huggingface-community-evals/SKILL.md
skills/huggingface-datasets/SKILL.md
skills/huggingface-gradio/SKILL.md
skills/huggingface-llm-trainer/SKILL.md
skills/huggingface-local-models/SKILL.md
skills/huggingface-lora-space-builder/SKILL.md
skills/huggingface-paper-publisher/SKILL.md
skills/huggingface-papers/SKILL.md
skills/huggingface-spaces/SKILL.md
skills/huggingface-tool-builder/SKILL.md
skills/huggingface-trackio/SKILL.md
skills/huggingface-vision-trainer/SKILL.md
skills/huggingface-zerogpu/SKILL.md
skills/insecure-defaults/SKILL.md
skills/mcp-builder/SKILL.md
skills/paper-summarizer/SKILL.md
skills/sarif-parsing/SKILL.md
skills/security-review/SKILL.md
skills/security-threat-model/SKILL.md
skills/semgrep-rule-creator/SKILL.md
skills/semgrep-rule-variant-creator/SKILL.md
skills/semgrep/SKILL.md
skills/sharp-edges/SKILL.md
skills/skill-creator/SKILL.md
skills/supply-chain-risk-auditor/SKILL.md
skills/transformers-js/SKILL.md

Metadata

Files
0
Version
f198a18
Hash
25a58b41
Indexed
2026-08-16 09:13

inicio - Wiki
Copyright © 2011-2026 iteam. Current version is 2.155.2. UTC+08:00, 2026-08-16 18:17
浙ICP备14020137号-1 $mapa de visitantes$