OpenSkillIndex← back to search
claude-code

AutoSkill training_log_sop

General SOP for generating or analyzing training logs, incorporating metrics like loss, accuracy, precision, recall, and max accuracy across epochs.

install
source · Clone the upstream repo
git clone https://github.com/ECNU-ICALK/AutoSkill
Claude Code · Install into ~/.claude/skills/
T=$(mktemp -d) && git clone --depth=1 https://github.com/ECNU-ICALK/AutoSkill "$T" && mkdir -p ~/.claude/skills && cp -r "$T/SkillBank/ConvSkill/chinese_gpt4_8_GLM4.7/training_log_sop" ~/.claude/skills/ecnu-icalk-autoskill-training-log-sop && rm -rf "$T"
manifest: SkillBank/ConvSkill/chinese_gpt4_8_GLM4.7/training_log_sop/SKILL.md
source content

training_log_sop

General SOP for generating or analyzing training logs, incorporating metrics like loss, accuracy, precision, recall, and max accuracy across epochs.

Prompt

Follow this SOP (replace specifics with placeholders like <PROJECT>/<ENV>/<VERSION>):

Log Format Template

Generate or parse logs using the following structure, ensuring all relevant metrics are populated: Epoch <CURRENT>/<TOTAL> <STEPS>/<TOTAL_STEPS> [==============================] - <TIME>s <MS>ms/step - loss: <LOSS> - accuracy: <ACC> - precision: <PREC> - recall: <REC> - max_acc: <MAX_ACC> - val_loss: <VAL_LOSS> - val_accuracy: <VAL_ACC> - val_precision: <VAL_PREC> - val_recall: <VAL_REC> - val_max_acc: <VAL_MAX_ACC> - lr: <LR>

Workflow Instructions

For each step, include:

  1. Action
  2. Checks
  3. Failure rollback/fallback plan

Output Format

For each step number, provide status/result and what to do next.

Triggers

  • Use when the user asks for a process or checklist.
  • Use when you want to reuse a previously mentioned method/SOP.
  • Use when generating training logs or metrics.

Examples

Example 1

Input:

Break this into best-practice, executable steps.