Clean repository: keep only trl skill

COMPARISON.md

@@ -1,235 +0,0 @@
-# Training Script Comparison
-
-## What Makes the Production Script "Clean"?
-
-### ❌ Original Script Issues
-
-```python
-# 1. Double Trackio initialization
-trackio.init(project="qwen-demo-sft", space_id="evalstate/trackio-demo")
-# ... later ...
-SFTConfig(report_to="trackio")  # TRL initializes AGAIN!
-# Result: 2 spaces created!
-
-# 2. Unclear what's customizable
-model="Qwen/Qwen2.5-0.5B"  # Hard to find/change
-dataset = load_dataset("trl-lib/Capybara")  # Buried in code
-
-# 3. Mixed concerns
-trackio.init(...)
-dataset = load_dataset(...)
-print(f"Sample: {dataset[0]}")  # Debugging mixed with setup
-config = SFTConfig(...)
-```
-
-### ✅ Clean Production Script
-
-```python
-# 1. Single Trackio init
-trackio.init(...)
-SFTConfig(report_to="trackio")  # Uses existing connection
-# Result: 1 space!
-
-# 2. Clear configuration section at top
-# ============================================================================
-# CONFIGURATION - Customize via environment variables
-# ============================================================================
-MODEL = os.getenv("MODEL", "Qwen/Qwen2.5-0.5B")
-DATASET = os.getenv("DATASET", "trl-lib/Capybara")
-# All customizable params grouped together with comments
-
-# 3. Separated concerns
-# Setup
-if USE_TRACKIO:
-    trackio.init(...)
-
-# Load data
-dataset = load_dataset(...)
-
-# Train
-trainer = SFTTrainer(...)
-trainer.train()
-```
-
-## Key Improvements
-
-### 1. Documentation Structure
-
-```python
-"""
-CUSTOMIZABLE PARAMETERS (via environment variables):
-    MODEL           - Model to fine-tune (default: Qwen/Qwen2.5-0.5B)
-    DATASET         - Dataset name on Hub (default: trl-lib/Capybara)
-    ...
-
-EXAMPLE USAGE:
-    hf_jobs("uv", {
-        "env": {"MODEL": "meta-llama/Llama-3.2-1B"}
-    })
-"""
-```
-
-**Benefits:**
-- ✅ Clear what can be changed
-- ✅ Shows how to change it
-- ✅ Includes examples
-- ✅ Visible at file top
-
-### 2. Three-Tier Configuration
-
-```
-┌─────────────────────────────────────┐
-│ TIER 1: Environment Variables       │
-│ ✅ Change freely without editing    │
-│ MODEL, DATASET, MAX_STEPS, etc.     │
-├─────────────────────────────────────┤
-│ TIER 2: Fixed Constants             │
-│ ⚙️  Edit if needed (advanced)       │
-│ LORA_R, GRADIENT_ACCUMULATION       │
-├─────────────────────────────────────┤
-│ TIER 3: Training Logic              │
-│ 🔒 Don't modify                     │
-│ Trainer initialization, etc.        │
-└─────────────────────────────────────┘
-```
-
-**Benefits:**
-- ✅ Clear what to change vs what to leave alone
-- ✅ Beginners change env vars only
-- ✅ Advanced users can modify Tier 2
-- ✅ Tier 3 is implementation detail
-
-### 3. Inline Documentation
-
-```python
-# Model Selection
-MODEL = os.getenv("MODEL", "Qwen/Qwen2.5-0.5B")
-# Common options:
-#   - Qwen/Qwen2.5-0.5B (fast, demo)
-#   - Qwen/Qwen2.5-3B (production)
-#   - meta-llama/Llama-3.2-1B
-#   - HuggingFaceTB/SmolLM2-1.7B
-```
-
-**Benefits:**
-- ✅ Context right where you need it
-- ✅ Shows concrete examples
-- ✅ Explains when to use each option
-
-### 4. Clear Output
-
-```python
-print("="*80)
-print("🚀 TRAINING CONFIGURATION")
-print("="*80)
-print(f"Model:          {MODEL}")
-print(f"Dataset:        {DATASET}")
-...
-```
-
-**Benefits:**
-- ✅ Easy to verify settings before training
-- ✅ Appears in job logs
-- ✅ Helpful for debugging
-- ✅ Professional appearance
-
-### 5. Single Monitoring Space
-
-```python
-# All projects use same space
-TRACKIO_SPACE = "evalstate/ml-experiments"
-
-# Different project names for filtering
-project=OUTPUT_REPO.split('/')[-1]  # e.g., "qwen-demo"
-```
-
-**Benefits:**
-- ✅ One dashboard for all experiments
-- ✅ Easy comparison across projects
-- ✅ Filter by project name
-- ✅ No space proliferation
-
-## Usage Comparison
-
-### Original (Hard to Customize)
-
-```python
-# To change model, must edit script:
-# 1. Download script
-# 2. Edit: model="new-model"
-# 3. Upload to Hub
-# 4. Submit job
-
-hf_jobs("uv", {
-    "script": "train.py",  # Your modified version
-    "flavor": "t4-small"
-})
-```
-
-### Production (Easy to Customize)
-
-```python
-# Just pass env vars:
-hf_jobs("uv", {
-    "script": "train_production_documented.py",  # Same script!
-    "flavor": "a10g-large",
-    "env": {
-        "MODEL": "meta-llama/Llama-3.2-1B",
-        "MAX_STEPS": "100"
-    }
-})
-```
-
-**No script editing needed!**
-
-## File Summary
-
-We created three files:
-
-1. **`train_production_documented.py`** - The main script
-   - 150 lines (well-commented)
-   - Clear three-tier structure
-   - Inline documentation
-   - Ready to use
-
-2. **`TRAINING_GUIDE.md`** - Usage guide
-   - Quick start examples
-   - Parameter reference
-   - Troubleshooting
-   - Best practices
-
-3. **`COMPARISON.md`** (this file) - Design rationale
-   - Why changes were made
-   - Before/after comparison
-   - Benefits explained
-
-## When to Use Each Approach
-
-| Approach | When to Use |
-|----------|-------------|
-| **Minimal** | Learning, one-off tests |
-| **Production** | Reusable experiments, multiple runs |
-| **Organized** | Team projects, complex workflows |
-
-## Next Steps
-
-1. **Upload to Hub:**
-   ```bash
-   hf upload evalstate/demo-training-scripts train_production_documented.py
-   hf upload evalstate/demo-training-scripts TRAINING_GUIDE.md
-   ```
-
-2. **Run it:**
-   ```python
-   hf_jobs("uv", {
-       "script": "https://huggingface.co/evalstate/demo-training-scripts/resolve/main/train_production_documented.py",
-       "flavor": "t4-small",
-       "timeout": "20m",
-       "secrets": {"HF_TOKEN": "$HF_TOKEN"}
-   })
-   ```
-
-3. **Customize it:**
-   ```python
-   "env": {"MODEL": "your-model", "MAX_STEPS": "50"}
-   ```

SCRIPTS_COMPARISON.md

@@ -1,199 +0,0 @@
-# Scripts Comparison: Current Directory vs TRL Skill
-
-## Two Different Locations
-
-### 1. Current Directory Scripts (What we just created)
-**Location:** `/home/ssmith/source/training/.claude/skills/`
-
-These are the training scripts we just created during this session:
-
-| Script | Lines | Purpose |
-|--------|-------|---------|
-| `train_minimal.py` | 25 | Learning/demo - bare minimum |
-| `train_production.py` | 75 | Reusable - env vars, no docs |
-| `train_production_documented.py` | 150 | Reusable - env vars + extensive docs ⭐ |
-| `train_organized.py` | 140 | Enterprise - functions, class-based |
-| `train_demo.py` | 80 | Original demo (has Trackio bug) |
-
-### 2. TRL Skill Scripts (Pre-existing)
-**Location:** `/home/ssmith/source/training/.claude/skills/trl/scripts/`
-
-These are part of the TRL skill and were already there:
-
-| Script | Lines | Purpose |
-|--------|-------|---------|
-| `train_sft_example.py` | 111 | SFT training example (from skill) |
-| `train_dpo_example.py` | 98 | DPO training example (from skill) |
-| `train_grpo_example.py` | 97 | GRPO training example (from skill) |
-| `validate_dataset.py` | 175 | Dataset validation utility |
-| `estimate_cost.py` | 149 | Cost estimation utility |
-| `convert_to_gguf.py` | 301 | GGUF conversion utility |
-
-## Key Differences
-
-### Purpose
-
-**Current Directory (our new scripts):**
-- Created during THIS conversation
-- Demonstrate different levels of documentation
-- Show progression from simple → documented → organized
-- Teaching examples for "clean code"
-
-**TRL Skill Scripts:**
-- Part of the pre-existing TRL skill
-- Production examples for different training methods
-- Utility scripts for common tasks
-- Reference implementations
-
-### Focus
-
-**Current Directory:**
-```
-train_minimal.py           → "How simple can we go?"
-train_production.py        → "Add configurability"
-train_production_documented.py → "Add documentation" ⭐
-train_organized.py         → "Add structure"
-```
-*Shows evolution of code quality*
-
-**TRL Skill Scripts:**
-```
-train_sft_example.py    → "How to do SFT"
-train_dpo_example.py    → "How to do DPO"
-train_grpo_example.py   → "How to do GRPO"
-validate_dataset.py     → "Utility: validate format"
-estimate_cost.py        → "Utility: estimate cost"
-```
-*Shows different training methods + utilities*
-
-## Detailed Comparison: Our Scripts vs TRL Skill's train_sft_example.py
-
-### trl/scripts/train_sft_example.py (111 lines)
-```python
-# From the TRL skill - complete production example
-import trackio
-from datasets import load_dataset
-from peft import LoraConfig
-from trl import SFTTrainer, SFTConfig
-
-# Initialize Trackio
-trackio.init(
-    project="qwen-capybara-sft",
-    space_id="username/my-trackio-dashboard",
-    config={...}
-)
-
-# Load dataset
-dataset = load_dataset("trl-lib/Capybara", split="train")
-
-# Configure training
-config = SFTConfig(
-    output_dir="qwen-capybara-sft",
-    push_to_hub=True,
-    hub_model_id="username/qwen-capybara-sft",
-    num_train_epochs=3,
-    per_device_train_batch_size=4,
-    # ... many parameters explicitly set
-)
-
-# Train
-trainer = SFTTrainer(...)
-trainer.train()
-trainer.push_to_hub()
-```
-
-**Characteristics:**
-- ✅ Complete production example
-- ✅ Includes Trackio (single init)
-- ✅ Full dataset (not limited to 50)
-- ✅ Many parameters explicitly set
-- ❌ NOT configurable via env vars
-- ❌ Hard-coded values (must edit to change)
-- ❌ Minimal inline documentation
-
-### Our train_production_documented.py (150 lines)
-```python
-# Our new script - configurable + documented
-"""
-CUSTOMIZABLE PARAMETERS (via environment variables):
-    MODEL           - Model to fine-tune (default: Qwen/Qwen2.5-0.5B)
-    DATASET         - Dataset name on Hub (default: trl-lib/Capybara)
-    ...
-"""
-
-import os
-from datasets import load_dataset
-from peft import LoraConfig
-from trl import SFTTrainer, SFTConfig
-
-# CONFIGURATION - Customize via environment variables
-MODEL = os.getenv("MODEL", "Qwen/Qwen2.5-0.5B")
-# Common options:
-#   - Qwen/Qwen2.5-0.5B (fast, demo)
-#   - Qwen/Qwen2.5-3B (production)
-
-DATASET = os.getenv("DATASET", "trl-lib/Capybara")
-MAX_STEPS = int(os.getenv("MAX_STEPS", "20"))
-USE_TRACKIO = os.getenv("USE_TRACKIO", "true").lower() == "true"
-
-# ... rest of training code
-```
-
-**Characteristics:**
-- ✅ Environment-based configuration
-- ✅ Extensive inline documentation
-- ✅ Examples in comments
-- ✅ Reusable without editing
-- ✅ Clear sections
-- ✅ Optional Trackio
-- ⚠️  Smaller dataset (50 examples) for demo
-
-## When to Use What?
-
-### Use TRL Skill Scripts When:
-- ✅ You want complete production examples
-- ✅ You need reference implementations
-- ✅ You want to see different training methods (SFT, DPO, GRPO)
-- ✅ You need utility scripts (validation, cost estimation)
-- ✅ You're okay editing the script to customize
-
-### Use Our New Scripts When:
-- ✅ You want quick demos (train_minimal.py)
-- ✅ You need reusable scripts (train_production.py)
-- ✅ You want self-documenting code (train_production_documented.py)
-- ✅ You need to run many experiments with different settings
-- ✅ You want to customize via environment variables
-- ✅ You're sharing with team and want clear documentation
-
-## Relationship
-
-```
-TRL Skill Scripts (Reference)
-     ↓
-     └─ train_sft_example.py (111 lines, production example)
-     
-Our New Scripts (Teaching Progression)
-     ↓
-     ├─ train_minimal.py (25 lines, bare minimum)
-     ├─ train_production.py (75 lines, configurable)
-     ├─ train_production_documented.py (150 lines, configurable + docs) ⭐
-     └─ train_organized.py (140 lines, structured)
-```
-
-## Summary
-
-**TRL Skill Scripts:**
-- Part of the TRL skill infrastructure
-- Reference implementations
-- Show "what to do"
-- Must edit to customize
-
-**Our New Scripts:**
-- Created during this conversation
-- Teaching progression (simple → documented → organized)
-- Show "how to organize code"
-- Customize via environment variables
-
-**They complement each other!**
-- TRL scripts show the different training methods
-- Our scripts show different code organization styles

SCRIPT_COMPARISON.md

@@ -1,336 +0,0 @@
-# Training Scripts Comparison
-
-## All Scripts in Directory
-
-| Script | Lines | Purpose | Best For |
-|--------|-------|---------|----------|
-| **train_minimal.py** | ~25 | Absolute bare minimum | Learning, throw-away tests |
-| **train_production.py** | ~75 | Env-configurable, clean | Reusable experiments |
-| **train_production_documented.py** | ~150 | Same as production + docs | Sharing with others |
-| **train_organized.py** | ~140 | Functions, class-based | Complex team projects |
-| **train_demo.py** | ~80 | Original demo script | Initial demo (has issues) |
-
-## Quick Comparison
-
-### 1. train_minimal.py (25 lines)
-```python
-# Simplest possible
-from trl import SFTTrainer, SFTConfig
-
-dataset = load_dataset("trl-lib/Capybara", split="train[:50]")
-
-trainer = SFTTrainer(
-    model="Qwen/Qwen2.5-0.5B",
-    train_dataset=dataset,
-    args=SFTConfig(
-        output_dir="output",
-        push_to_hub=True,
-        hub_model_id="evalstate/qwen-demo-minimal",
-        max_steps=20,
-        report_to="none",  # No monitoring
-    )
-)
-trainer.train()
-```
-
-**Pros:**
-- ✅ Super simple
-- ✅ Easy to understand
-- ✅ Fast execution (no monitoring overhead)
-
-**Cons:**
-- ❌ Hard to customize (must edit code)
-- ❌ No monitoring
-- ❌ No environment variables
-
-**Use when:**
-- Learning TRL basics
-- One-time quick test
-- Don't care about metrics
-
----
-
-### 2. train_production.py (75 lines)
-```python
-# Environment-configurable
-MODEL = os.getenv("MODEL", "Qwen/Qwen2.5-0.5B")
-DATASET = os.getenv("DATASET", "trl-lib/Capybara")
-MAX_STEPS = int(os.getenv("MAX_STEPS", "20"))
-USE_TRACKIO = os.getenv("USE_TRACKIO", "true").lower() == "true"
-
-if USE_TRACKIO:
-    trackio.init(
-        project=OUTPUT_REPO.split('/')[-1],
-        space_id="evalstate/ml-experiments",
-        config={...}
-    )
-
-trainer = SFTTrainer(...)
-trainer.train()
-```
-
-**Pros:**
-- ✅ Reusable (change via env vars)
-- ✅ Trackio monitoring
-- ✅ Clean structure
-- ✅ Production-ready
-
-**Cons:**
-- ❌ Less documentation in code
-- ❌ Need to know which env vars exist
-
-**Use when:**
-- Running multiple experiments
-- Need to compare different models/datasets
-- Want monitoring
-
----
-
-### 3. train_production_documented.py (150 lines) ⭐ RECOMMENDED
-```python
-"""
-CUSTOMIZABLE PARAMETERS (via environment variables):
-    MODEL           - Model to fine-tune (default: Qwen/Qwen2.5-0.5B)
-    DATASET         - Dataset name on Hub (default: trl-lib/Capybara)
-    OUTPUT_REPO     - Where to save model (default: evalstate/qwen-capybara-sft)
-    MAX_STEPS       - Training steps (default: 20)
-    BATCH_SIZE      - Batch size per device (default: 2)
-    LEARNING_RATE   - Learning rate (default: 2e-5)
-    USE_TRACKIO     - Enable monitoring (default: true)
-"""
-
-# Model Selection
-MODEL = os.getenv("MODEL", "Qwen/Qwen2.5-0.5B")
-# Common options:
-#   - Qwen/Qwen2.5-0.5B (fast, demo)
-#   - Qwen/Qwen2.5-3B (production)
-#   - meta-llama/Llama-3.2-1B
-
-# Dataset Selection
-DATASET = os.getenv("DATASET", "trl-lib/Capybara")
-# Use any conversational dataset with "messages" field
-
-# Training Parameters
-MAX_STEPS = int(os.getenv("MAX_STEPS", "20"))
-# Quick demo: 10-20 | Development: 100-500 | Production: 1000+
-
-# ... rest of training code (same as train_production.py)
-```
-
-**Pros:**
-- ✅ Same functionality as train_production.py
-- ✅ Self-documenting
-- ✅ Shows examples in comments
-- ✅ Easy for new users
-- ✅ Professional looking
-
-**Cons:**
-- ❌ More lines (but mostly comments)
-
-**Use when:**
-- Sharing with team
-- Need documentation
-- Want examples inline
-- **⭐ Default choice for most cases**
-
----
-
-### 4. train_organized.py (140 lines)
-```python
-class Config:
-    """Training configuration with environment overrides"""
-    MODEL = os.getenv("MODEL", "Qwen/Qwen2.5-0.5B")
-    DATASET = os.getenv("DATASET", "trl-lib/Capybara")
-    # ... all config in one place
-
-def setup_monitoring(config: Config):
-    """Initialize Trackio for experiment tracking"""
-    ...
-
-def load_and_validate_dataset(config: Config):
-    """Load dataset and perform basic validation"""
-    ...
-
-def main():
-    config = Config()
-    setup_monitoring(config)
-    dataset = load_and_validate_dataset(config)
-    trainer = train(dataset)
-    ...
-```
-
-**Pros:**
-- ✅ Cleanest separation of concerns
-- ✅ Easy to test individual functions
-- ✅ Best for complex workflows
-- ✅ Team-friendly structure
-
-**Cons:**
-- ❌ More complex
-- ❌ Overkill for simple training
-
-**Use when:**
-- Team project
-- Need to extend/customize heavily
-- Want to unit test components
-- Complex training pipeline
-
----
-
-### 5. train_demo.py (80 lines) - Original Demo
-```python
-# Has the double Trackio issue!
-trackio.init(
-    project="qwen-demo-sft",
-    space_id="evalstate/trackio-demo",  # Creates space 1
-    ...
-)
-
-config = SFTConfig(
-    report_to="trackio",  # TRL creates space 2!
-    ...
-)
-```
-
-**Issues:**
-- ❌ Creates 2 Trackio spaces
-- ❌ Not configurable
-- ❌ Mixed concerns
-
-**This was our original script that we improved!**
-
----
-
-## Decision Matrix
-
-### Choose Based On Your Needs:
-
-```
-Need simplicity above all?
-  → train_minimal.py
-
-Running ONE experiment?
-  → train_minimal.py or train_production.py
-
-Running MULTIPLE experiments?
-  → train_production_documented.py ⭐
-
-Sharing with others?
-  → train_production_documented.py ⭐
-
-Complex team project?
-  → train_organized.py
-
-Just learning?
-  → train_minimal.py
-```
-
-## Key Differences
-
-### Customization Method
-
-| Script | How to Customize |
-|--------|------------------|
-| train_minimal.py | Edit code |
-| train_production.py | Environment variables |
-| train_production_documented.py | Environment variables (with docs) |
-| train_organized.py | Environment variables or Config class |
-
-### Monitoring
-
-| Script | Monitoring |
-|--------|-----------|
-| train_minimal.py | None |
-| train_production.py | Trackio (optional) |
-| train_production_documented.py | Trackio (optional) |
-| train_organized.py | Trackio (optional) |
-| train_demo.py | Trackio (with bug!) |
-
-### Lines of Code
-
-| Script | Lines | Actual Code | Comments/Docs |
-|--------|-------|-------------|---------------|
-| train_minimal.py | 25 | 20 | 5 |
-| train_production.py | 75 | 60 | 15 |
-| train_production_documented.py | 150 | 60 | 90 |
-| train_organized.py | 140 | 100 | 40 |
-
-**Note:** train_production.py and train_production_documented.py have the SAME actual code, just different amounts of documentation!
-
-## The Winner: train_production_documented.py ⭐
-
-### Why It's the Best Choice:
-
-1. **Self-documenting** - Everything explained inline
-2. **Easy to customize** - Just pass env vars
-3. **Shows examples** - Comments show real options
-4. **Professional** - Clean output formatting
-5. **Single Trackio space** - No duplication
-6. **Reusable** - Same script for all experiments
-
-### What Makes It Different from train_production.py?
-
-**Same code, more documentation!**
-
-```python
-# train_production.py (minimal comments)
-MODEL = os.getenv("MODEL", "Qwen/Qwen2.5-0.5B")
-
-# train_production_documented.py (helpful comments)
-MODEL = os.getenv("MODEL", "Qwen/Qwen2.5-0.5B")
-# Common options:
-#   - Qwen/Qwen2.5-0.5B (fast, demo)
-#   - Qwen/Qwen2.5-3B (production)
-#   - meta-llama/Llama-3.2-1B
-```
-
-The actual execution is identical, but the documented version is much easier to understand and use!
-
-## Usage Examples
-
-### Minimal
-```python
-# Must edit script to change model
-hf_jobs("uv", {"script": "train_minimal.py", "flavor": "t4-small"})
-```
-
-### Production (Either Version)
-```python
-# Change model via environment
-hf_jobs("uv", {
-    "script": "train_production_documented.py",
-    "flavor": "a10g-large",
-    "env": {
-        "MODEL": "meta-llama/Llama-3.2-1B",
-        "MAX_STEPS": "100"
-    }
-})
-```
-
-### Organized
-```python
-# Same as production, just different internal structure
-hf_jobs("uv", {
-    "script": "train_organized.py",
-    "flavor": "a10g-large",
-    "env": {"MODEL": "meta-llama/Llama-3.2-1B"}
-})
-```
-
-## Recommendation
-
-**Use `train_production_documented.py` for:**
-- ✅ All production work
-- ✅ Sharing with teammates
-- ✅ Multiple experiments
-- ✅ When you want documentation
-
-**Use `train_minimal.py` for:**
-- ✅ Quick learning
-- ✅ Throwaway tests
-- ✅ When you want absolute simplicity
-
-**Use `train_organized.py` for:**
-- ✅ Complex team projects
-- ✅ When you need to heavily extend the code
-- ✅ Unit testing individual components

TRAINING_GUIDE.md

@@ -1,219 +0,0 @@
-# Production Training Script Guide
-
-## Quick Start
-
-**Run with defaults (5-10 minute demo):**
-```python
-hf_jobs("uv", {
-    "script": "https://huggingface.co/evalstate/demo-training-scripts/resolve/main/train_production_documented.py",
-    "flavor": "t4-small",
-    "timeout": "20m",
-    "secrets": {"HF_TOKEN": "$HF_TOKEN"}
-})
-```
-
-## Customizable Parameters
-
-All parameters can be customized via environment variables without modifying the script:
-
-### 🎯 Most Common Settings
-
-| Parameter | Default | Description | When to Change |
-|-----------|---------|-------------|----------------|
-| `MODEL` | `Qwen/Qwen2.5-0.5B` | Model to fine-tune | Use larger model for better quality |
-| `DATASET` | `trl-lib/Capybara` | Training dataset | Use your own dataset |
-| `OUTPUT_REPO` | `evalstate/qwen-capybara-sft` | Where to save | Always set to your username |
-| `MAX_STEPS` | `20` | Training duration | 100+ for real training |
-| `LEARNING_RATE` | `2e-5` | Learning rate | Tune if loss unstable |
-
-### 📊 Monitoring
-
-| Parameter | Default | Description |
-|-----------|---------|-------------|
-| `USE_TRACKIO` | `true` | Enable real-time monitoring | Set to `false` for faster demo |
-
-### ⚙️ Advanced Settings
-
-| Parameter | Default | Description |
-|-----------|---------|-------------|
-| `BATCH_SIZE` | `2` | Batch size per GPU | Increase for larger GPUs |
-| `LEARNING_RATE` | `2e-5` | Learning rate | Typical range: 1e-5 to 5e-5 |
-
-## Usage Examples
-
-### Example 1: Quick Demo (Default)
-```python
-hf_jobs("uv", {
-    "script": "train_production_documented.py",
-    "flavor": "t4-small",
-    "timeout": "20m",
-    "secrets": {"HF_TOKEN": "$HF_TOKEN"}
-})
-```
-
-**Result:** Trains Qwen-0.5B for 20 steps (~10 minutes, ~$0.20)
-
-### Example 2: Custom Model & Dataset
-```python
-hf_jobs("uv", {
-    "script": "train_production_documented.py",
-    "flavor": "a10g-large",
-    "timeout": "2h",
-    "env": {
-        "MODEL": "meta-llama/Llama-3.2-1B",
-        "DATASET": "HuggingFaceH4/ultrachat_200k",
-        "OUTPUT_REPO": "your-username/llama-ultrachat",
-        "MAX_STEPS": "100"
-    },
-    "secrets": {"HF_TOKEN": "$HF_TOKEN"}
-})
-```
-
-**Result:** Trains Llama-3.2-1B for 100 steps (~1 hour, ~$5)
-
-### Example 3: Longer Training Run
-```python
-hf_jobs("uv", {
-    "script": "train_production_documented.py",
-    "flavor": "a10g-large",
-    "timeout": "6h",
-    "env": {
-        "MODEL": "Qwen/Qwen2.5-3B",
-        "DATASET": "your-username/my-dataset",
-        "OUTPUT_REPO": "your-username/qwen3b-custom",
-        "MAX_STEPS": "500",
-        "BATCH_SIZE": "4",
-        "LEARNING_RATE": "1e-5"
-    },
-    "secrets": {"HF_TOKEN": "$HF_TOKEN"}
-})
-```
-
-**Result:** Production training (~4-5 hours, ~$20-25)
-
-### Example 4: Without Monitoring (Fastest)
-```python
-hf_jobs("uv", {
-    "script": "train_production_documented.py",
-    "flavor": "t4-small",
-    "timeout": "15m",
-    "env": {
-        "USE_TRACKIO": "false",
-        "OUTPUT_REPO": "your-username/quick-test"
-    },
-    "secrets": {"HF_TOKEN": "$HF_TOKEN"}
-})
-```
-
-**Result:** Fastest possible demo (~8 minutes)
-
-## How It Works
-
-### Three Configuration Levels
-
-```
-┌─────────────────────────────────────┐
-│ 1. CUSTOMIZABLE via env vars       │  ← Change these freely
-│    MODEL, DATASET, MAX_STEPS, etc. │
-├─────────────────────────────────────┤
-│ 2. FIXED (edit script if needed)   │  ← Advanced users only
-│    LORA_R, GRADIENT_ACCUMULATION    │
-├─────────────────────────────────────┤
-│ 3. TRAINING LOGIC                   │  ← Don't modify
-│    Trainer initialization, etc.     │
-└─────────────────────────────────────┘
-```
-
-### Single Trackio Space
-
-All experiments go to **one dashboard**: `evalstate/ml-experiments`
-
-- ✅ Easy comparison across experiments
-- ✅ Filtered by project name
-- ✅ No space clutter
-
-**Access your dashboard:**
-https://huggingface.co/spaces/evalstate/ml-experiments
-
-## Recommended Models
-
-| Model | Size | Speed | Use Case | Hardware |
-|-------|------|-------|----------|----------|
-| `Qwen/Qwen2.5-0.5B` | 0.5B | ⚡⚡⚡ | Quick tests | t4-small |
-| `HuggingFaceTB/SmolLM2-1.7B` | 1.7B | ⚡⚡ | Development | t4-medium |
-| `meta-llama/Llama-3.2-1B` | 1B | ⚡⚡ | Production | a10g-small |
-| `Qwen/Qwen2.5-3B` | 3B | ⚡ | High quality | a10g-large |
-
-## Recommended Datasets
-
-**Conversational (chat/instruction):**
-- `trl-lib/Capybara` - High-quality chat (16K examples)
-- `HuggingFaceH4/ultrachat_200k` - Diverse conversations
-- `argilla/distilabel-capybara-dpo-7k-binarized` - Preference data (for DPO)
-
-**Your own dataset:**
-- Must have `"messages"` field in conversational format
-- See: https://huggingface.co/docs/trl/dataset_formats
-
-## Hardware Selection
-
-| Hardware | Cost/hr | When to Use |
-|----------|---------|-------------|
-| `t4-small` | ~$0.75 | Quick demos (20 steps) |
-| `t4-medium` | ~$1.50 | Small models, testing |
-| `a10g-small` | ~$3.50 | Production (1-3B models) |
-| `a10g-large` | ~$5.00 | Production (3-7B models) |
-
-## Timeout Guidelines
-
-| Scenario | Recommended Timeout |
-|----------|-------------------|
-| Quick demo (20 steps) | 15-20m |
-| Development (100 steps) | 1-2h |
-| Production (500+ steps) | 4-6h |
-
-**Always add 20-30% buffer** for setup time!
-
-## Troubleshooting
-
-### "Out of memory"
-**Solution:** Reduce `BATCH_SIZE` or use larger GPU
-```python
-"env": {"BATCH_SIZE": "1"}  # Smallest possible
-```
-
-### "Dataset not found"
-**Check:** Dataset name is correct and public
-```python
-"env": {"DATASET": "username/dataset-name"}
-```
-
-### "Permission denied pushing to Hub"
-**Check:** `OUTPUT_REPO` uses your username
-```python
-"env": {"OUTPUT_REPO": "YOUR-USERNAME/model-name"}
-```
-
-### Training too slow/expensive
-**Solution:** Reduce steps or use smaller model
-```python
-"env": {
-    "MAX_STEPS": "10",  # Faster
-    "MODEL": "Qwen/Qwen2.5-0.5B"  # Smaller
-}
-```
-
-## Next Steps
-
-After training completes:
-
-1. **View your model:** https://huggingface.co/YOUR-USERNAME/model-name
-2. **Check metrics:** https://huggingface.co/spaces/evalstate/ml-experiments
-3. **Test the model:** Use the Inference API or download locally
-4. **Share:** Model is public by default (or make private in settings)
-
-## Questions?
-
-- 📖 Full TRL docs: https://huggingface.co/docs/trl
-- 💬 Ask in Hugging Face Discord
-- 🐛 Issues: Check job logs for error messages

dataset_inspector.py

@@ -1,416 +0,0 @@
-#!/usr/bin/env python3
-# /// script
-# dependencies = []
-# ///
-"""
-Dataset Format Inspector for TRL Training (LLM-Optimized Output)
-
-Inspects Hugging Face datasets to determine TRL training compatibility.
-Uses Datasets Server API for instant results - no dataset download needed!
-
-ULTRA-EFFICIENT: Uses HF Datasets Server API - completes in <2 seconds.
-
-Usage with HF Jobs:
-    hf_jobs("uv", {
-        "script": "https://huggingface.co/datasets/evalstate/trl-helpers/raw/main/dataset_inspector.py",
-        "script_args": ["--dataset", "your/dataset", "--split", "train"]
-    })
-"""
-
-import argparse
-import sys
-import json
-import urllib.request
-import urllib.parse
-from typing import List, Dict, Any
-
-
-def parse_args():
-    parser = argparse.ArgumentParser(description="Inspect dataset format for TRL training")
-    parser.add_argument("--dataset", type=str, required=True, help="Dataset name")
-    parser.add_argument("--split", type=str, default="train", help="Dataset split (default: train)")
-    parser.add_argument("--config", type=str, default="default", help="Dataset config name (default: default)")
-    parser.add_argument("--preview", type=int, default=150, help="Max chars per field preview")
-    parser.add_argument("--samples", type=int, default=5, help="Number of samples to fetch (default: 5)")
-    parser.add_argument("--json-output", action="store_true", help="Output as JSON")
-    return parser.parse_args()
-
-
-def api_request(url: str) -> Dict:
-    """Make API request to Datasets Server"""
-    try:
-        with urllib.request.urlopen(url, timeout=10) as response:
-            return json.loads(response.read().decode())
-    except urllib.error.HTTPError as e:
-        if e.code == 404:
-            return None
-        raise Exception(f"API request failed: {e.code} {e.reason}")
-    except Exception as e:
-        raise Exception(f"API request failed: {str(e)}")
-
-
-def get_splits(dataset: str) -> Dict:
-    """Get available splits for dataset"""
-    url = f"https://datasets-server.huggingface.co/splits?dataset={urllib.parse.quote(dataset)}"
-    return api_request(url)
-
-
-def get_rows(dataset: str, config: str, split: str, offset: int = 0, length: int = 5) -> Dict:
-    """Get rows from dataset"""
-    url = f"https://datasets-server.huggingface.co/rows?dataset={urllib.parse.quote(dataset)}&config={config}&split={split}&offset={offset}&length={length}"
-    return api_request(url)
-
-
-def find_columns(columns: List[str], patterns: List[str]) -> List[str]:
-    """Find columns matching patterns"""
-    return [c for c in columns if any(p in c.lower() for p in patterns)]
-
-
-def check_sft_compatibility(columns: List[str]) -> Dict[str, Any]:
-    """Check SFT compatibility"""
-    has_messages = "messages" in columns
-    has_text = "text" in columns
-    has_prompt_completion = "prompt" in columns and "completion" in columns
-    
-    ready = has_messages or has_text or has_prompt_completion
-    
-    possible_prompt = find_columns(columns, ["prompt", "instruction", "question", "input"])
-    possible_response = find_columns(columns, ["response", "completion", "output", "answer"])
-    
-    return {
-        "ready": ready,
-        "reason": "messages" if has_messages else "text" if has_text else "prompt+completion" if has_prompt_completion else None,
-        "possible_prompt": possible_prompt[0] if possible_prompt else None,
-        "possible_response": possible_response[0] if possible_response else None,
-        "has_context": "context" in columns,
-    }
-
-
-def check_dpo_compatibility(columns: List[str]) -> Dict[str, Any]:
-    """Check DPO compatibility"""
-    has_standard = "prompt" in columns and "chosen" in columns and "rejected" in columns
-    
-    possible_prompt = find_columns(columns, ["prompt", "instruction", "question", "input"])
-    possible_chosen = find_columns(columns, ["chosen", "preferred", "winner"])
-    possible_rejected = find_columns(columns, ["rejected", "dispreferred", "loser"])
-    
-    can_map = bool(possible_prompt and possible_chosen and possible_rejected)
-    
-    return {
-        "ready": has_standard,
-        "can_map": can_map,
-        "prompt_col": possible_prompt[0] if possible_prompt else None,
-        "chosen_col": possible_chosen[0] if possible_chosen else None,
-        "rejected_col": possible_rejected[0] if possible_rejected else None,
-    }
-
-
-def check_grpo_compatibility(columns: List[str]) -> Dict[str, Any]:
-    """Check GRPO compatibility"""
-    has_prompt = "prompt" in columns
-    has_no_responses = "chosen" not in columns and "rejected" not in columns
-    
-    possible_prompt = find_columns(columns, ["prompt", "instruction", "question", "input"])
-    
-    return {
-        "ready": has_prompt and has_no_responses,
-        "can_map": bool(possible_prompt) and has_no_responses,
-        "prompt_col": possible_prompt[0] if possible_prompt else None,
-    }
-
-
-def check_kto_compatibility(columns: List[str]) -> Dict[str, Any]:
-    """Check KTO compatibility"""
-    return {"ready": "prompt" in columns and "completion" in columns and "label" in columns}
-
-
-def generate_mapping_code(method: str, info: Dict[str, Any]) -> str:
-    """Generate mapping code for a training method"""
-    if method == "SFT":
-        if info["ready"]:
-            return None
-        
-        prompt_col = info.get("possible_prompt")
-        response_col = info.get("possible_response")
-        has_context = info.get("has_context", False)
-        
-        if not prompt_col:
-            return None
-        
-        if has_context and response_col:
-            return f"""def format_for_sft(example):
-    text = f"Instruction: {{example['{prompt_col}']}}\\n\\n"
-    if example.get('context'):
-        text += f"Context: {{example['context']}}\\n\\n"
-    text += f"Response: {{example['{response_col}']}}"
-    return {{'text': text}}
-
-dataset = dataset.map(format_for_sft, remove_columns=dataset.column_names)"""
-        elif response_col:
-            return f"""def format_for_sft(example):
-    return {{'text': f"{{example['{prompt_col}']}}\\n\\n{{example['{response_col}']}}}}
-
-dataset = dataset.map(format_for_sft, remove_columns=dataset.column_names)"""
-        else:
-            return f"""def format_for_sft(example):
-    return {{'text': example['{prompt_col}']}}
-
-dataset = dataset.map(format_for_sft, remove_columns=dataset.column_names)"""
-    
-    elif method == "DPO":
-        if info["ready"] or not info["can_map"]:
-            return None
-        
-        return f"""def format_for_dpo(example):
-    return {{
-        'prompt': example['{info['prompt_col']}'],
-        'chosen': example['{info['chosen_col']}'],
-        'rejected': example['{info['rejected_col']}'],
-    }}
-
-dataset = dataset.map(format_for_dpo, remove_columns=dataset.column_names)"""
-    
-    elif method == "GRPO":
-        if info["ready"] or not info["can_map"]:
-            return None
-        
-        return f"""def format_for_grpo(example):
-    return {{'prompt': example['{info['prompt_col']}']}}
-
-dataset = dataset.map(format_for_grpo, remove_columns=dataset.column_names)"""
-    
-    return None
-
-
-def format_value_preview(value: Any, max_chars: int) -> str:
-    """Format value for preview"""
-    if value is None:
-        return "None"
-    elif isinstance(value, str):
-        return value[:max_chars] + ("..." if len(value) > max_chars else "")
-    elif isinstance(value, list):
-        if len(value) > 0 and isinstance(value[0], dict):
-            return f"[{len(value)} items] Keys: {list(value[0].keys())}"
-        preview = str(value)
-        return preview[:max_chars] + ("..." if len(preview) > max_chars else "")
-    else:
-        preview = str(value)
-        return preview[:max_chars] + ("..." if len(preview) > max_chars else "")
-
-
-def main():
-    args = parse_args()
-    
-    print(f"Fetching dataset info via Datasets Server API...")
-    
-    try:
-        # Get splits info
-        splits_data = get_splits(args.dataset)
-        if not splits_data or "splits" not in splits_data:
-            print(f"ERROR: Could not fetch splits for dataset '{args.dataset}'")
-            print(f"       Dataset may not exist or is not accessible via Datasets Server API")
-            sys.exit(1)
-        
-        # Find the right config
-        available_configs = set()
-        split_found = False
-        config_to_use = args.config
-        
-        for split_info in splits_data["splits"]:
-            available_configs.add(split_info["config"])
-            if split_info["config"] == args.config and split_info["split"] == args.split:
-                split_found = True
-        
-        # If default config not found, try first available
-        if not split_found and available_configs:
-            config_to_use = list(available_configs)[0]
-            print(f"Config '{args.config}' not found, trying '{config_to_use}'...")
-        
-        # Get rows
-        rows_data = get_rows(args.dataset, config_to_use, args.split, offset=0, length=args.samples)
-        
-        if not rows_data or "rows" not in rows_data:
-            print(f"ERROR: Could not fetch rows for dataset '{args.dataset}'")
-            print(f"       Split '{args.split}' may not exist")
-            print(f"       Available configs: {', '.join(sorted(available_configs))}")
-            sys.exit(1)
-        
-        rows = rows_data["rows"]
-        if not rows:
-            print(f"ERROR: No rows found in split '{args.split}'")
-            sys.exit(1)
-        
-        # Extract column info from first row
-        first_row = rows[0]["row"]
-        columns = list(first_row.keys())
-        features = rows_data.get("features", [])
-        
-        # Get total count if available
-        total_examples = "Unknown"
-        for split_info in splits_data["splits"]:
-            if split_info["config"] == config_to_use and split_info["split"] == args.split:
-                total_examples = f"{split_info.get('num_examples', 'Unknown'):,}" if isinstance(split_info.get('num_examples'), int) else "Unknown"
-                break
-        
-    except Exception as e:
-        print(f"ERROR: {str(e)}")
-        sys.exit(1)
-    
-    # Run compatibility checks
-    sft_info = check_sft_compatibility(columns)
-    dpo_info = check_dpo_compatibility(columns)
-    grpo_info = check_grpo_compatibility(columns)
-    kto_info = check_kto_compatibility(columns)
-    
-    # Determine recommended methods
-    recommended = []
-    if sft_info["ready"]:
-        recommended.append("SFT")
-    elif sft_info["possible_prompt"]:
-        recommended.append("SFT (needs mapping)")
-    
-    if dpo_info["ready"]:
-        recommended.append("DPO")
-    elif dpo_info["can_map"]:
-        recommended.append("DPO (needs mapping)")
-    
-    if grpo_info["ready"]:
-        recommended.append("GRPO")
-    elif grpo_info["can_map"]:
-        recommended.append("GRPO (needs mapping)")
-    
-    if kto_info["ready"]:
-        recommended.append("KTO")
-    
-    # JSON output mode
-    if args.json_output:
-        result = {
-            "dataset": args.dataset,
-            "config": config_to_use,
-            "split": args.split,
-            "total_examples": total_examples,
-            "columns": columns,
-            "features": [{"name": f["name"], "type": f["type"]} for f in features] if features else [],
-            "compatibility": {
-                "SFT": sft_info,
-                "DPO": dpo_info,
-                "GRPO": grpo_info,
-                "KTO": kto_info,
-            },
-            "recommended_methods": recommended,
-        }
-        print(json.dumps(result, indent=2))
-        sys.exit(0)
-    
-    # Human-readable output optimized for LLM parsing
-    print("=" * 80)
-    print(f"DATASET INSPECTION RESULTS")
-    print("=" * 80)
-    
-    print(f"\nDataset: {args.dataset}")
-    print(f"Config: {config_to_use}")
-    print(f"Split: {args.split}")
-    print(f"Total examples: {total_examples}")
-    print(f"Samples fetched: {len(rows)}")
-    
-    print(f"\n{'COLUMNS':-<80}")
-    if features:
-        for feature in features:
-            print(f"  {feature['name']}: {feature['type']}")
-    else:
-        for col in columns:
-            print(f"  {col}: (type info not available)")
-    
-    print(f"\n{'EXAMPLE DATA':-<80}")
-    example = first_row
-    for col in columns:
-        value = example.get(col)
-        display = format_value_preview(value, args.preview)
-        print(f"\n{col}:")
-        print(f"  {display}")
-    
-    print(f"\n{'TRAINING METHOD COMPATIBILITY':-<80}")
-    
-    # SFT
-    print(f"\n[SFT] {'✓ READY' if sft_info['ready'] else '✗ NEEDS MAPPING'}")
-    if sft_info["ready"]:
-        print(f"  Reason: Dataset has '{sft_info['reason']}' field")
-        print(f"  Action: Use directly with SFTTrainer")
-    elif sft_info["possible_prompt"]:
-        print(f"  Detected: prompt='{sft_info['possible_prompt']}' response='{sft_info['possible_response']}'")
-        print(f"  Action: Apply mapping code (see below)")
-    else:
-        print(f"  Status: Cannot determine mapping - manual inspection needed")
-    
-    # DPO
-    print(f"\n[DPO] {'✓ READY' if dpo_info['ready'] else '✗ NEEDS MAPPING' if dpo_info['can_map'] else '✗ INCOMPATIBLE'}")
-    if dpo_info["ready"]:
-        print(f"  Reason: Dataset has 'prompt', 'chosen', 'rejected' fields")
-        print(f"  Action: Use directly with DPOTrainer")
-    elif dpo_info["can_map"]:
-        print(f"  Detected: prompt='{dpo_info['prompt_col']}' chosen='{dpo_info['chosen_col']}' rejected='{dpo_info['rejected_col']}'")
-        print(f"  Action: Apply mapping code (see below)")
-    else:
-        print(f"  Status: Missing required fields (prompt + chosen + rejected)")
-    
-    # GRPO
-    print(f"\n[GRPO] {'✓ READY' if grpo_info['ready'] else '✗ NEEDS MAPPING' if grpo_info['can_map'] else '✗ INCOMPATIBLE'}")
-    if grpo_info["ready"]:
-        print(f"  Reason: Dataset has 'prompt' field")
-        print(f"  Action: Use directly with GRPOTrainer")
-    elif grpo_info["can_map"]:
-        print(f"  Detected: prompt='{grpo_info['prompt_col']}'")
-        print(f"  Action: Apply mapping code (see below)")
-    else:
-        print(f"  Status: Missing prompt field")
-    
-    # KTO
-    print(f"\n[KTO] {'✓ READY' if kto_info['ready'] else '✗ INCOMPATIBLE'}")
-    if kto_info["ready"]:
-        print(f"  Reason: Dataset has 'prompt', 'completion', 'label' fields")
-        print(f"  Action: Use directly with KTOTrainer")
-    else:
-        print(f"  Status: Missing required fields (prompt + completion + label)")
-    
-    # Mapping code
-    print(f"\n{'MAPPING CODE (if needed)':-<80}")
-    
-    mapping_needed = False
-    
-    sft_mapping = generate_mapping_code("SFT", sft_info)
-    if sft_mapping:
-        print(f"\n# For SFT Training:")
-        print(sft_mapping)
-        mapping_needed = True
-    
-    dpo_mapping = generate_mapping_code("DPO", dpo_info)
-    if dpo_mapping:
-        print(f"\n# For DPO Training:")
-        print(dpo_mapping)
-        mapping_needed = True
-    
-    grpo_mapping = generate_mapping_code("GRPO", grpo_info)
-    if grpo_mapping:
-        print(f"\n# For GRPO Training:")
-        print(grpo_mapping)
-        mapping_needed = True
-    
-    if not mapping_needed:
-        print("\nNo mapping needed - dataset is ready for training!")
-    
-    print(f"\n{'SUMMARY':-<80}")
-    print(f"Recommended training methods: {', '.join(recommended) if recommended else 'None (dataset needs formatting)'}")
-    print(f"\nNote: Used Datasets Server API (instant, no download required)")
-    
-    print("\n" + "=" * 80)
-    sys.exit(0)
-
-
-if __name__ == "__main__":
-    try:
-        main()
-    except KeyboardInterrupt:
-        sys.exit(0)
-    except Exception as e:
-        print(f"ERROR: {e}", file=sys.stderr)
-        sys.exit(1)

demo_train.py

@@ -1,84 +0,0 @@
-# /// script
-# dependencies = [
-#     "trl>=0.12.0",
-#     "peft>=0.7.0",
-#     "transformers>=4.36.0",
-#     "accelerate>=0.24.0",
-#     "trackio",
-# ]
-# ///
-
-import trackio
-from datasets import load_dataset
-from peft import LoraConfig
-from trl import SFTTrainer, SFTConfig
-
-# Initialize Trackio for real-time monitoring
-trackio.init(
-    project="qwen-demo-sft",
-    space_id="evalstate/demo-trackio-dashboard",
-    config={
-        "model": "Qwen/Qwen2.5-0.5B",
-        "dataset": "trl-lib/Capybara",
-        "examples": 50,
-        "max_steps": 20,
-        "note": "Quick demo training"
-    }
-)
-
-# Load dataset (only 50 examples for quick demo)
-dataset = load_dataset("trl-lib/Capybara", split="train[:50]")
-print(f"✅ Dataset loaded: {len(dataset)} examples")
-
-# Training configuration
-config = SFTConfig(
-    # Hub settings - CRITICAL for saving results
-    output_dir="qwen-demo-sft",
-    push_to_hub=True,
-    hub_model_id="evalstate/qwen-demo-sft",
-    
-    # Quick training settings
-    max_steps=20,  # Very short for demo
-    per_device_train_batch_size=2,
-    gradient_accumulation_steps=2,
-    learning_rate=2e-5,
-    
-    # Logging
-    logging_steps=5,
-    save_strategy="steps",
-    save_steps=10,
-    
-    # Monitoring
-    report_to="trackio",
-)
-
-# LoRA configuration (memory efficient)
-peft_config = LoraConfig(
-    r=16,
-    lora_alpha=32,
-    lora_dropout=0.05,
-    bias="none",
-    task_type="CAUSAL_LM",
-    target_modules=["q_proj", "v_proj"],
-)
-
-# Initialize and train
-trainer = SFTTrainer(
-    model="Qwen/Qwen2.5-0.5B",
-    train_dataset=dataset,
-    args=config,
-    peft_config=peft_config,
-)
-
-print("🚀 Starting demo training...")
-trainer.train()
-
-print("💾 Pushing to Hub...")
-trainer.push_to_hub()
-
-# Finish Trackio tracking
-trackio.finish()
-
-print("✅ Demo complete!")
-print(f"📦 Model: https://huggingface.co/evalstate/qwen-demo-sft")
-print(f"📊 Metrics: https://huggingface.co/spaces/evalstate/demo-trackio-dashboard")

skill-creator/LICENSE.txt

@@ -1,202 +0,0 @@
-
-                                 Apache License
-                           Version 2.0, January 2004
-                        http://www.apache.org/licenses/
-
-   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
-
-   1. Definitions.
-
-      "License" shall mean the terms and conditions for use, reproduction,
-      and distribution as defined by Sections 1 through 9 of this document.
-
-      "Licensor" shall mean the copyright owner or entity authorized by
-      the copyright owner that is granting the License.
-
-      "Legal Entity" shall mean the union of the acting entity and all
-      other entities that control, are controlled by, or are under common
-      control with that entity. For the purposes of this definition,
-      "control" means (i) the power, direct or indirect, to cause the
-      direction or management of such entity, whether by contract or
-      otherwise, or (ii) ownership of fifty percent (50%) or more of the
-      outstanding shares, or (iii) beneficial ownership of such entity.
-
-      "You" (or "Your") shall mean an individual or Legal Entity
-      exercising permissions granted by this License.
-
-      "Source" form shall mean the preferred form for making modifications,
-      including but not limited to software source code, documentation
-      source, and configuration files.
-
-      "Object" form shall mean any form resulting from mechanical
-      transformation or translation of a Source form, including but
-      not limited to compiled object code, generated documentation,
-      and conversions to other media types.
-
-      "Work" shall mean the work of authorship, whether in Source or
-      Object form, made available under the License, as indicated by a
-      copyright notice that is included in or attached to the work
-      (an example is provided in the Appendix below).
-
-      "Derivative Works" shall mean any work, whether in Source or Object
-      form, that is based on (or derived from) the Work and for which the
-      editorial revisions, annotations, elaborations, or other modifications
-      represent, as a whole, an original work of authorship. For the purposes
-      of this License, Derivative Works shall not include works that remain
-      separable from, or merely link (or bind by name) to the interfaces of,
-      the Work and Derivative Works thereof.
-
-      "Contribution" shall mean any work of authorship, including
-      the original version of the Work and any modifications or additions
-      to that Work or Derivative Works thereof, that is intentionally
-      submitted to Licensor for inclusion in the Work by the copyright owner
-      or by an individual or Legal Entity authorized to submit on behalf of
-      the copyright owner. For the purposes of this definition, "submitted"
-      means any form of electronic, verbal, or written communication sent
-      to the Licensor or its representatives, including but not limited to
-      communication on electronic mailing lists, source code control systems,
-      and issue tracking systems that are managed by, or on behalf of, the
-      Licensor for the purpose of discussing and improving the Work, but
-      excluding communication that is conspicuously marked or otherwise
-      designated in writing by the copyright owner as "Not a Contribution."
-
-      "Contributor" shall mean Licensor and any individual or Legal Entity
-      on behalf of whom a Contribution has been received by Licensor and
-      subsequently incorporated within the Work.
-
-   2. Grant of Copyright License. Subject to the terms and conditions of
-      this License, each Contributor hereby grants to You a perpetual,
-      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
-      copyright license to reproduce, prepare Derivative Works of,
-      publicly display, publicly perform, sublicense, and distribute the
-      Work and such Derivative Works in Source or Object form.
-
-   3. Grant of Patent License. Subject to the terms and conditions of
-      this License, each Contributor hereby grants to You a perpetual,
-      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
-      (except as stated in this section) patent license to make, have made,
-      use, offer to sell, sell, import, and otherwise transfer the Work,
-      where such license applies only to those patent claims licensable
-      by such Contributor that are necessarily infringed by their
-      Contribution(s) alone or by combination of their Contribution(s)
-      with the Work to which such Contribution(s) was submitted. If You
-      institute patent litigation against any entity (including a
-      cross-claim or counterclaim in a lawsuit) alleging that the Work
-      or a Contribution incorporated within the Work constitutes direct
-      or contributory patent infringement, then any patent licenses
-      granted to You under this License for that Work shall terminate
-      as of the date such litigation is filed.
-
-   4. Redistribution. You may reproduce and distribute copies of the
-      Work or Derivative Works thereof in any medium, with or without
-      modifications, and in Source or Object form, provided that You
-      meet the following conditions:
-
-      (a) You must give any other recipients of the Work or
-          Derivative Works a copy of this License; and
-
-      (b) You must cause any modified files to carry prominent notices
-          stating that You changed the files; and
-
-      (c) You must retain, in the Source form of any Derivative Works
-          that You distribute, all copyright, patent, trademark, and
-          attribution notices from the Source form of the Work,
-          excluding those notices that do not pertain to any part of
-          the Derivative Works; and
-
-      (d) If the Work includes a "NOTICE" text file as part of its
-          distribution, then any Derivative Works that You distribute must
-          include a readable copy of the attribution notices contained
-          within such NOTICE file, excluding those notices that do not
-          pertain to any part of the Derivative Works, in at least one
-          of the following places: within a NOTICE text file distributed
-          as part of the Derivative Works; within the Source form or
-          documentation, if provided along with the Derivative Works; or,
-          within a display generated by the Derivative Works, if and
-          wherever such third-party notices normally appear. The contents
-          of the NOTICE file are for informational purposes only and
-          do not modify the License. You may add Your own attribution
-          notices within Derivative Works that You distribute, alongside
-          or as an addendum to the NOTICE text from the Work, provided
-          that such additional attribution notices cannot be construed
-          as modifying the License.
-
-      You may add Your own copyright statement to Your modifications and
-      may provide additional or different license terms and conditions
-      for use, reproduction, or distribution of Your modifications, or
-      for any such Derivative Works as a whole, provided Your use,
-      reproduction, and distribution of the Work otherwise complies with
-      the conditions stated in this License.
-
-   5. Submission of Contributions. Unless You explicitly state otherwise,
-      any Contribution intentionally submitted for inclusion in the Work
-      by You to the Licensor shall be under the terms and conditions of
-      this License, without any additional terms or conditions.
-      Notwithstanding the above, nothing herein shall supersede or modify
-      the terms of any separate license agreement you may have executed
-      with Licensor regarding such Contributions.
-
-   6. Trademarks. This License does not grant permission to use the trade
-      names, trademarks, service marks, or product names of the Licensor,
-      except as required for reasonable and customary use in describing the
-      origin of the Work and reproducing the content of the NOTICE file.
-
-   7. Disclaimer of Warranty. Unless required by applicable law or
-      agreed to in writing, Licensor provides the Work (and each
-      Contributor provides its Contributions) on an "AS IS" BASIS,
-      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
-      implied, including, without limitation, any warranties or conditions
-      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
-      PARTICULAR PURPOSE. You are solely responsible for determining the
-      appropriateness of using or redistributing the Work and assume any
-      risks associated with Your exercise of permissions under this License.
-
-   8. Limitation of Liability. In no event and under no legal theory,
-      whether in tort (including negligence), contract, or otherwise,
-      unless required by applicable law (such as deliberate and grossly
-      negligent acts) or agreed to in writing, shall any Contributor be
-      liable to You for damages, including any direct, indirect, special,
-      incidental, or consequential damages of any character arising as a
-      result of this License or out of the use or inability to use the
-      Work (including but not limited to damages for loss of goodwill,
-      work stoppage, computer failure or malfunction, or any and all
-      other commercial damages or losses), even if such Contributor
-      has been advised of the possibility of such damages.
-
-   9. Accepting Warranty or Additional Liability. While redistributing
-      the Work or Derivative Works thereof, You may choose to offer,
-      and charge a fee for, acceptance of support, warranty, indemnity,
-      or other liability obligations and/or rights consistent with this
-      License. However, in accepting such obligations, You may act only
-      on Your own behalf and on Your sole responsibility, not on behalf
-      of any other Contributor, and only if You agree to indemnify,
-      defend, and hold each Contributor harmless for any liability
-      incurred by, or claims asserted against, such Contributor by reason
-      of your accepting any such warranty or additional liability.
-
-   END OF TERMS AND CONDITIONS
-
-   APPENDIX: How to apply the Apache License to your work.
-
-      To apply the Apache License to your work, attach the following
-      boilerplate notice, with the fields enclosed by brackets "[]"
-      replaced with your own identifying information. (Don't include
-      the brackets!)  The text should be enclosed in the appropriate
-      comment syntax for the file format. We also recommend that a
-      file or class name and description of purpose be included on the
-      same "printed page" as the copyright notice for easier
-      identification within third-party archives.
-
-   Copyright [yyyy] [name of copyright owner]
-
-   Licensed under the Apache License, Version 2.0 (the "License");
-   you may not use this file except in compliance with the License.
-   You may obtain a copy of the License at
-
-       http://www.apache.org/licenses/LICENSE-2.0
-
-   Unless required by applicable law or agreed to in writing, software
-   distributed under the License is distributed on an "AS IS" BASIS,
-   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-   See the License for the specific language governing permissions and
-   limitations under the License.

skill-creator/SKILL.md

@@ -1,209 +0,0 @@
----
-name: skill-creator
-description: Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Claude's capabilities with specialized knowledge, workflows, or tool integrations.
-license: Complete terms in LICENSE.txt
----
-
-# Skill Creator
-
-This skill provides guidance for creating effective skills.
-
-## About Skills
-
-Skills are modular, self-contained packages that extend Claude's capabilities by providing
-specialized knowledge, workflows, and tools. Think of them as "onboarding guides" for specific
-domains or tasks—they transform Claude from a general-purpose agent into a specialized agent
-equipped with procedural knowledge that no model can fully possess.
-
-### What Skills Provide
-
-1. Specialized workflows - Multi-step procedures for specific domains
-2. Tool integrations - Instructions for working with specific file formats or APIs
-3. Domain expertise - Company-specific knowledge, schemas, business logic
-4. Bundled resources - Scripts, references, and assets for complex and repetitive tasks
-
-### Anatomy of a Skill
-
-Every skill consists of a required SKILL.md file and optional bundled resources:
-
-```
-skill-name/
-├── SKILL.md (required)
-│   ├── YAML frontmatter metadata (required)
-│   │   ├── name: (required)
-│   │   └── description: (required)
-│   └── Markdown instructions (required)
-└── Bundled Resources (optional)
-    ├── scripts/          - Executable code (Python/Bash/etc.)
-    ├── references/       - Documentation intended to be loaded into context as needed
-    └── assets/           - Files used in output (templates, icons, fonts, etc.)
-```
-
-#### SKILL.md (required)
-
-**Metadata Quality:** The `name` and `description` in YAML frontmatter determine when Claude will use the skill. Be specific about what the skill does and when to use it. Use the third-person (e.g. "This skill should be used when..." instead of "Use this skill when...").
-
-#### Bundled Resources (optional)
-
-##### Scripts (`scripts/`)
-
-Executable code (Python/Bash/etc.) for tasks that require deterministic reliability or are repeatedly rewritten.
-
-- **When to include**: When the same code is being rewritten repeatedly or deterministic reliability is needed
-- **Example**: `scripts/rotate_pdf.py` for PDF rotation tasks
-- **Benefits**: Token efficient, deterministic, may be executed without loading into context
-- **Note**: Scripts may still need to be read by Claude for patching or environment-specific adjustments
-
-##### References (`references/`)
-
-Documentation and reference material intended to be loaded as needed into context to inform Claude's process and thinking.
-
-- **When to include**: For documentation that Claude should reference while working
-- **Examples**: `references/finance.md` for financial schemas, `references/mnda.md` for company NDA template, `references/policies.md` for company policies, `references/api_docs.md` for API specifications
-- **Use cases**: Database schemas, API documentation, domain knowledge, company policies, detailed workflow guides
-- **Benefits**: Keeps SKILL.md lean, loaded only when Claude determines it's needed
-- **Best practice**: If files are large (>10k words), include grep search patterns in SKILL.md
-- **Avoid duplication**: Information should live in either SKILL.md or references files, not both. Prefer references files for detailed information unless it's truly core to the skill—this keeps SKILL.md lean while making information discoverable without hogging the context window. Keep only essential procedural instructions and workflow guidance in SKILL.md; move detailed reference material, schemas, and examples to references files.
-
-##### Assets (`assets/`)
-
-Files not intended to be loaded into context, but rather used within the output Claude produces.
-
-- **When to include**: When the skill needs files that will be used in the final output
-- **Examples**: `assets/logo.png` for brand assets, `assets/slides.pptx` for PowerPoint templates, `assets/frontend-template/` for HTML/React boilerplate, `assets/font.ttf` for typography
-- **Use cases**: Templates, images, icons, boilerplate code, fonts, sample documents that get copied or modified
-- **Benefits**: Separates output resources from documentation, enables Claude to use files without loading them into context
-
-### Progressive Disclosure Design Principle
-
-Skills use a three-level loading system to manage context efficiently:
-
-1. **Metadata (name + description)** - Always in context (~100 words)
-2. **SKILL.md body** - When skill triggers (<5k words)
-3. **Bundled resources** - As needed by Claude (Unlimited*)
-
-*Unlimited because scripts can be executed without reading into context window.
-
-## Skill Creation Process
-
-To create a skill, follow the "Skill Creation Process" in order, skipping steps only if there is a clear reason why they are not applicable.
-
-### Step 1: Understanding the Skill with Concrete Examples
-
-Skip this step only when the skill's usage patterns are already clearly understood. It remains valuable even when working with an existing skill.
-
-To create an effective skill, clearly understand concrete examples of how the skill will be used. This understanding can come from either direct user examples or generated examples that are validated with user feedback.
-
-For example, when building an image-editor skill, relevant questions include:
-
-- "What functionality should the image-editor skill support? Editing, rotating, anything else?"
-- "Can you give some examples of how this skill would be used?"
-- "I can imagine users asking for things like 'Remove the red-eye from this image' or 'Rotate this image'. Are there other ways you imagine this skill being used?"
-- "What would a user say that should trigger this skill?"
-
-To avoid overwhelming users, avoid asking too many questions in a single message. Start with the most important questions and follow up as needed for better effectiveness.
-
-Conclude this step when there is a clear sense of the functionality the skill should support.
-
-### Step 2: Planning the Reusable Skill Contents
-
-To turn concrete examples into an effective skill, analyze each example by:
-
-1. Considering how to execute on the example from scratch
-2. Identifying what scripts, references, and assets would be helpful when executing these workflows repeatedly
-
-Example: When building a `pdf-editor` skill to handle queries like "Help me rotate this PDF," the analysis shows:
-
-1. Rotating a PDF requires re-writing the same code each time
-2. A `scripts/rotate_pdf.py` script would be helpful to store in the skill
-
-Example: When designing a `frontend-webapp-builder` skill for queries like "Build me a todo app" or "Build me a dashboard to track my steps," the analysis shows:
-
-1. Writing a frontend webapp requires the same boilerplate HTML/React each time
-2. An `assets/hello-world/` template containing the boilerplate HTML/React project files would be helpful to store in the skill
-
-Example: When building a `big-query` skill to handle queries like "How many users have logged in today?" the analysis shows:
-
-1. Querying BigQuery requires re-discovering the table schemas and relationships each time
-2. A `references/schema.md` file documenting the table schemas would be helpful to store in the skill
-
-To establish the skill's contents, analyze each concrete example to create a list of the reusable resources to include: scripts, references, and assets.
-
-### Step 3: Initializing the Skill
-
-At this point, it is time to actually create the skill.
-
-Skip this step only if the skill being developed already exists, and iteration or packaging is needed. In this case, continue to the next step.
-
-When creating a new skill from scratch, always run the `init_skill.py` script. The script conveniently generates a new template skill directory that automatically includes everything a skill requires, making the skill creation process much more efficient and reliable.
-
-Usage:
-
-```bash
-scripts/init_skill.py <skill-name> --path <output-directory>
-```
-
-The script:
-
-- Creates the skill directory at the specified path
-- Generates a SKILL.md template with proper frontmatter and TODO placeholders
-- Creates example resource directories: `scripts/`, `references/`, and `assets/`
-- Adds example files in each directory that can be customized or deleted
-
-After initialization, customize or remove the generated SKILL.md and example files as needed.
-
-### Step 4: Edit the Skill
-
-When editing the (newly-generated or existing) skill, remember that the skill is being created for another instance of Claude to use. Focus on including information that would be beneficial and non-obvious to Claude. Consider what procedural knowledge, domain-specific details, or reusable assets would help another Claude instance execute these tasks more effectively.
-
-#### Start with Reusable Skill Contents
-
-To begin implementation, start with the reusable resources identified above: `scripts/`, `references/`, and `assets/` files. Note that this step may require user input. For example, when implementing a `brand-guidelines` skill, the user may need to provide brand assets or templates to store in `assets/`, or documentation to store in `references/`.
-
-Also, delete any example files and directories not needed for the skill. The initialization script creates example files in `scripts/`, `references/`, and `assets/` to demonstrate structure, but most skills won't need all of them.
-
-#### Update SKILL.md
-
-**Writing Style:** Write the entire skill using **imperative/infinitive form** (verb-first instructions), not second person. Use objective, instructional language (e.g., "To accomplish X, do Y" rather than "You should do X" or "If you need to do X"). This maintains consistency and clarity for AI consumption.
-
-To complete SKILL.md, answer the following questions:
-
-1. What is the purpose of the skill, in a few sentences?
-2. When should the skill be used?
-3. In practice, how should Claude use the skill? All reusable skill contents developed above should be referenced so that Claude knows how to use them.
-
-### Step 5: Packaging a Skill
-
-Once the skill is ready, it should be packaged into a distributable zip file that gets shared with the user. The packaging process automatically validates the skill first to ensure it meets all requirements:
-
-```bash
-scripts/package_skill.py <path/to/skill-folder>
-```
-
-Optional output directory specification:
-
-```bash
-scripts/package_skill.py <path/to/skill-folder> ./dist
-```
-
-The packaging script will:
-
-1. **Validate** the skill automatically, checking:
-   - YAML frontmatter format and required fields
-   - Skill naming conventions and directory structure
-   - Description completeness and quality
-   - File organization and resource references
-
-2. **Package** the skill if validation passes, creating a zip file named after the skill (e.g., `my-skill.zip`) that includes all files and maintains the proper directory structure for distribution.
-
-If validation fails, the script will report the errors and exit without creating a package. Fix any validation errors and run the packaging command again.
-
-### Step 6: Iterate
-
-After testing the skill, users may request improvements. Often this happens right after using the skill, with fresh context of how the skill performed.
-
-**Iteration workflow:**
-1. Use the skill on real tasks
-2. Notice struggles or inefficiencies
-3. Identify how SKILL.md or bundled resources should be updated
-4. Implement changes and test again

skill-creator/scripts/init_skill.py

@@ -1,303 +0,0 @@
-#!/usr/bin/env python3
-"""
-Skill Initializer - Creates a new skill from template
-
-Usage:
-    init_skill.py <skill-name> --path <path>
-
-Examples:
-    init_skill.py my-new-skill --path skills/public
-    init_skill.py my-api-helper --path skills/private
-    init_skill.py custom-skill --path /custom/location
-"""
-
-import sys
-from pathlib import Path
-
-
-SKILL_TEMPLATE = """---
-name: {skill_name}
-description: [TODO: Complete and informative explanation of what the skill does and when to use it. Include WHEN to use this skill - specific scenarios, file types, or tasks that trigger it.]
----
-
-# {skill_title}
-
-## Overview
-
-[TODO: 1-2 sentences explaining what this skill enables]
-
-## Structuring This Skill
-
-[TODO: Choose the structure that best fits this skill's purpose. Common patterns:
-
-**1. Workflow-Based** (best for sequential processes)
-- Works well when there are clear step-by-step procedures
-- Example: DOCX skill with "Workflow Decision Tree" → "Reading" → "Creating" → "Editing"
-- Structure: ## Overview → ## Workflow Decision Tree → ## Step 1 → ## Step 2...
-
-**2. Task-Based** (best for tool collections)
-- Works well when the skill offers different operations/capabilities
-- Example: PDF skill with "Quick Start" → "Merge PDFs" → "Split PDFs" → "Extract Text"
-- Structure: ## Overview → ## Quick Start → ## Task Category 1 → ## Task Category 2...
-
-**3. Reference/Guidelines** (best for standards or specifications)
-- Works well for brand guidelines, coding standards, or requirements
-- Example: Brand styling with "Brand Guidelines" → "Colors" → "Typography" → "Features"
-- Structure: ## Overview → ## Guidelines → ## Specifications → ## Usage...
-
-**4. Capabilities-Based** (best for integrated systems)
-- Works well when the skill provides multiple interrelated features
-- Example: Product Management with "Core Capabilities" → numbered capability list
-- Structure: ## Overview → ## Core Capabilities → ### 1. Feature → ### 2. Feature...
-
-Patterns can be mixed and matched as needed. Most skills combine patterns (e.g., start with task-based, add workflow for complex operations).
-
-Delete this entire "Structuring This Skill" section when done - it's just guidance.]
-
-## [TODO: Replace with the first main section based on chosen structure]
-
-[TODO: Add content here. See examples in existing skills:
-- Code samples for technical skills
-- Decision trees for complex workflows
-- Concrete examples with realistic user requests
-- References to scripts/templates/references as needed]
-
-## Resources
-
-This skill includes example resource directories that demonstrate how to organize different types of bundled resources:
-
-### scripts/
-Executable code (Python/Bash/etc.) that can be run directly to perform specific operations.
-
-**Examples from other skills:**
-- PDF skill: `fill_fillable_fields.py`, `extract_form_field_info.py` - utilities for PDF manipulation
-- DOCX skill: `document.py`, `utilities.py` - Python modules for document processing
-
-**Appropriate for:** Python scripts, shell scripts, or any executable code that performs automation, data processing, or specific operations.
-
-**Note:** Scripts may be executed without loading into context, but can still be read by Claude for patching or environment adjustments.
-
-### references/
-Documentation and reference material intended to be loaded into context to inform Claude's process and thinking.
-
-**Examples from other skills:**
-- Product management: `communication.md`, `context_building.md` - detailed workflow guides
-- BigQuery: API reference documentation and query examples
-- Finance: Schema documentation, company policies
-
-**Appropriate for:** In-depth documentation, API references, database schemas, comprehensive guides, or any detailed information that Claude should reference while working.
-
-### assets/
-Files not intended to be loaded into context, but rather used within the output Claude produces.
-
-**Examples from other skills:**
-- Brand styling: PowerPoint template files (.pptx), logo files
-- Frontend builder: HTML/React boilerplate project directories
-- Typography: Font files (.ttf, .woff2)
-
-**Appropriate for:** Templates, boilerplate code, document templates, images, icons, fonts, or any files meant to be copied or used in the final output.
-
----
-
-**Any unneeded directories can be deleted.** Not every skill requires all three types of resources.
-"""
-
-EXAMPLE_SCRIPT = '''#!/usr/bin/env python3
-"""
-Example helper script for {skill_name}
-
-This is a placeholder script that can be executed directly.
-Replace with actual implementation or delete if not needed.
-
-Example real scripts from other skills:
-- pdf/scripts/fill_fillable_fields.py - Fills PDF form fields
-- pdf/scripts/convert_pdf_to_images.py - Converts PDF pages to images
-"""
-
-def main():
-    print("This is an example script for {skill_name}")
-    # TODO: Add actual script logic here
-    # This could be data processing, file conversion, API calls, etc.
-
-if __name__ == "__main__":
-    main()
-'''
-
-EXAMPLE_REFERENCE = """# Reference Documentation for {skill_title}
-
-This is a placeholder for detailed reference documentation.
-Replace with actual reference content or delete if not needed.
-
-Example real reference docs from other skills:
-- product-management/references/communication.md - Comprehensive guide for status updates
-- product-management/references/context_building.md - Deep-dive on gathering context
-- bigquery/references/ - API references and query examples
-
-## When Reference Docs Are Useful
-
-Reference docs are ideal for:
-- Comprehensive API documentation
-- Detailed workflow guides
-- Complex multi-step processes
-- Information too lengthy for main SKILL.md
-- Content that's only needed for specific use cases
-
-## Structure Suggestions
-
-### API Reference Example
-- Overview
-- Authentication
-- Endpoints with examples
-- Error codes
-- Rate limits
-
-### Workflow Guide Example
-- Prerequisites
-- Step-by-step instructions
-- Common patterns
-- Troubleshooting
-- Best practices
-"""
-
-EXAMPLE_ASSET = """# Example Asset File
-
-This placeholder represents where asset files would be stored.
-Replace with actual asset files (templates, images, fonts, etc.) or delete if not needed.
-
-Asset files are NOT intended to be loaded into context, but rather used within
-the output Claude produces.
-
-Example asset files from other skills:
-- Brand guidelines: logo.png, slides_template.pptx
-- Frontend builder: hello-world/ directory with HTML/React boilerplate
-- Typography: custom-font.ttf, font-family.woff2
-- Data: sample_data.csv, test_dataset.json
-
-## Common Asset Types
-
-- Templates: .pptx, .docx, boilerplate directories
-- Images: .png, .jpg, .svg, .gif
-- Fonts: .ttf, .otf, .woff, .woff2
-- Boilerplate code: Project directories, starter files
-- Icons: .ico, .svg
-- Data files: .csv, .json, .xml, .yaml
-
-Note: This is a text placeholder. Actual assets can be any file type.
-"""
-
-
-def title_case_skill_name(skill_name):
-    """Convert hyphenated skill name to Title Case for display."""
-    return ' '.join(word.capitalize() for word in skill_name.split('-'))
-
-
-def init_skill(skill_name, path):
-    """
-    Initialize a new skill directory with template SKILL.md.
-
-    Args:
-        skill_name: Name of the skill
-        path: Path where the skill directory should be created
-
-    Returns:
-        Path to created skill directory, or None if error
-    """
-    # Determine skill directory path
-    skill_dir = Path(path).resolve() / skill_name
-
-    # Check if directory already exists
-    if skill_dir.exists():
-        print(f"❌ Error: Skill directory already exists: {skill_dir}")
-        return None
-
-    # Create skill directory
-    try:
-        skill_dir.mkdir(parents=True, exist_ok=False)
-        print(f"✅ Created skill directory: {skill_dir}")
-    except Exception as e:
-        print(f"❌ Error creating directory: {e}")
-        return None
-
-    # Create SKILL.md from template
-    skill_title = title_case_skill_name(skill_name)
-    skill_content = SKILL_TEMPLATE.format(
-        skill_name=skill_name,
-        skill_title=skill_title
-    )
-
-    skill_md_path = skill_dir / 'SKILL.md'
-    try:
-        skill_md_path.write_text(skill_content)
-        print("✅ Created SKILL.md")
-    except Exception as e:
-        print(f"❌ Error creating SKILL.md: {e}")
-        return None
-
-    # Create resource directories with example files
-    try:
-        # Create scripts/ directory with example script
-        scripts_dir = skill_dir / 'scripts'
-        scripts_dir.mkdir(exist_ok=True)
-        example_script = scripts_dir / 'example.py'
-        example_script.write_text(EXAMPLE_SCRIPT.format(skill_name=skill_name))
-        example_script.chmod(0o755)
-        print("✅ Created scripts/example.py")
-
-        # Create references/ directory with example reference doc
-        references_dir = skill_dir / 'references'
-        references_dir.mkdir(exist_ok=True)
-        example_reference = references_dir / 'api_reference.md'
-        example_reference.write_text(EXAMPLE_REFERENCE.format(skill_title=skill_title))
-        print("✅ Created references/api_reference.md")
-
-        # Create assets/ directory with example asset placeholder
-        assets_dir = skill_dir / 'assets'
-        assets_dir.mkdir(exist_ok=True)
-        example_asset = assets_dir / 'example_asset.txt'
-        example_asset.write_text(EXAMPLE_ASSET)
-        print("✅ Created assets/example_asset.txt")
-    except Exception as e:
-        print(f"❌ Error creating resource directories: {e}")
-        return None
-
-    # Print next steps
-    print(f"\n✅ Skill '{skill_name}' initialized successfully at {skill_dir}")
-    print("\nNext steps:")
-    print("1. Edit SKILL.md to complete the TODO items and update the description")
-    print("2. Customize or delete the example files in scripts/, references/, and assets/")
-    print("3. Run the validator when ready to check the skill structure")
-
-    return skill_dir
-
-
-def main():
-    if len(sys.argv) < 4 or sys.argv[2] != '--path':
-        print("Usage: init_skill.py <skill-name> --path <path>")
-        print("\nSkill name requirements:")
-        print("  - Hyphen-case identifier (e.g., 'data-analyzer')")
-        print("  - Lowercase letters, digits, and hyphens only")
-        print("  - Max 40 characters")
-        print("  - Must match directory name exactly")
-        print("\nExamples:")
-        print("  init_skill.py my-new-skill --path skills/public")
-        print("  init_skill.py my-api-helper --path skills/private")
-        print("  init_skill.py custom-skill --path /custom/location")
-        sys.exit(1)
-
-    skill_name = sys.argv[1]
-    path = sys.argv[3]
-
-    print(f"🚀 Initializing skill: {skill_name}")
-    print(f"   Location: {path}")
-    print()
-
-    result = init_skill(skill_name, path)
-
-    if result:
-        sys.exit(0)
-    else:
-        sys.exit(1)
-
-
-if __name__ == "__main__":
-    main()

skill-creator/scripts/package_skill.py

@@ -1,110 +0,0 @@
-#!/usr/bin/env python3
-"""
-Skill Packager - Creates a distributable zip file of a skill folder
-
-Usage:
-    python utils/package_skill.py <path/to/skill-folder> [output-directory]
-
-Example:
-    python utils/package_skill.py skills/public/my-skill
-    python utils/package_skill.py skills/public/my-skill ./dist
-"""
-
-import sys
-import zipfile
-from pathlib import Path
-from quick_validate import validate_skill
-
-
-def package_skill(skill_path, output_dir=None):
-    """
-    Package a skill folder into a zip file.
-
-    Args:
-        skill_path: Path to the skill folder
-        output_dir: Optional output directory for the zip file (defaults to current directory)
-
-    Returns:
-        Path to the created zip file, or None if error
-    """
-    skill_path = Path(skill_path).resolve()
-
-    # Validate skill folder exists
-    if not skill_path.exists():
-        print(f"❌ Error: Skill folder not found: {skill_path}")
-        return None
-
-    if not skill_path.is_dir():
-        print(f"❌ Error: Path is not a directory: {skill_path}")
-        return None
-
-    # Validate SKILL.md exists
-    skill_md = skill_path / "SKILL.md"
-    if not skill_md.exists():
-        print(f"❌ Error: SKILL.md not found in {skill_path}")
-        return None
-
-    # Run validation before packaging
-    print("🔍 Validating skill...")
-    valid, message = validate_skill(skill_path)
-    if not valid:
-        print(f"❌ Validation failed: {message}")
-        print("   Please fix the validation errors before packaging.")
-        return None
-    print(f"✅ {message}\n")
-
-    # Determine output location
-    skill_name = skill_path.name
-    if output_dir:
-        output_path = Path(output_dir).resolve()
-        output_path.mkdir(parents=True, exist_ok=True)
-    else:
-        output_path = Path.cwd()
-
-    zip_filename = output_path / f"{skill_name}.zip"
-
-    # Create the zip file
-    try:
-        with zipfile.ZipFile(zip_filename, 'w', zipfile.ZIP_DEFLATED) as zipf:
-            # Walk through the skill directory
-            for file_path in skill_path.rglob('*'):
-                if file_path.is_file():
-                    # Calculate the relative path within the zip
-                    arcname = file_path.relative_to(skill_path.parent)
-                    zipf.write(file_path, arcname)
-                    print(f"  Added: {arcname}")
-
-        print(f"\n✅ Successfully packaged skill to: {zip_filename}")
-        return zip_filename
-
-    except Exception as e:
-        print(f"❌ Error creating zip file: {e}")
-        return None
-
-
-def main():
-    if len(sys.argv) < 2:
-        print("Usage: python utils/package_skill.py <path/to/skill-folder> [output-directory]")
-        print("\nExample:")
-        print("  python utils/package_skill.py skills/public/my-skill")
-        print("  python utils/package_skill.py skills/public/my-skill ./dist")
-        sys.exit(1)
-
-    skill_path = sys.argv[1]
-    output_dir = sys.argv[2] if len(sys.argv) > 2 else None
-
-    print(f"📦 Packaging skill: {skill_path}")
-    if output_dir:
-        print(f"   Output directory: {output_dir}")
-    print()
-
-    result = package_skill(skill_path, output_dir)
-
-    if result:
-        sys.exit(0)
-    else:
-        sys.exit(1)
-
-
-if __name__ == "__main__":
-    main()

skill-creator/scripts/quick_validate.py

@@ -1,65 +0,0 @@
-#!/usr/bin/env python3
-"""
-Quick validation script for skills - minimal version
-"""
-
-import sys
-import os
-import re
-from pathlib import Path
-
-def validate_skill(skill_path):
-    """Basic validation of a skill"""
-    skill_path = Path(skill_path)
-    
-    # Check SKILL.md exists
-    skill_md = skill_path / 'SKILL.md'
-    if not skill_md.exists():
-        return False, "SKILL.md not found"
-    
-    # Read and validate frontmatter
-    content = skill_md.read_text()
-    if not content.startswith('---'):
-        return False, "No YAML frontmatter found"
-    
-    # Extract frontmatter
-    match = re.match(r'^---\n(.*?)\n---', content, re.DOTALL)
-    if not match:
-        return False, "Invalid frontmatter format"
-    
-    frontmatter = match.group(1)
-    
-    # Check required fields
-    if 'name:' not in frontmatter:
-        return False, "Missing 'name' in frontmatter"
-    if 'description:' not in frontmatter:
-        return False, "Missing 'description' in frontmatter"
-    
-    # Extract name for validation
-    name_match = re.search(r'name:\s*(.+)', frontmatter)
-    if name_match:
-        name = name_match.group(1).strip()
-        # Check naming convention (hyphen-case: lowercase with hyphens)
-        if not re.match(r'^[a-z0-9-]+$', name):
-            return False, f"Name '{name}' should be hyphen-case (lowercase letters, digits, and hyphens only)"
-        if name.startswith('-') or name.endswith('-') or '--' in name:
-            return False, f"Name '{name}' cannot start/end with hyphen or contain consecutive hyphens"
-
-    # Extract and validate description
-    desc_match = re.search(r'description:\s*(.+)', frontmatter)
-    if desc_match:
-        description = desc_match.group(1).strip()
-        # Check for angle brackets
-        if '<' in description or '>' in description:
-            return False, "Description cannot contain angle brackets (< or >)"
-
-    return True, "Skill is valid!"
-
-if __name__ == "__main__":
-    if len(sys.argv) != 2:
-        print("Usage: python quick_validate.py <skill_directory>")
-        sys.exit(1)
-    
-    valid, message = validate_skill(sys.argv[1])
-    print(message)
-    sys.exit(0 if valid else 1)

skills

@@ -1 +0,0 @@
-Subproject commit 92990b2da655659d36b670ff922866bbb41c177c

submit_job.py

@@ -1,16 +0,0 @@
-from huggingface_hub import run_uv_job
-
-job = run_uv_job(
-    script="https://huggingface.co/evalstate/demo-training-scripts/resolve/main/train_demo.py",
-    flavor="t4-small",
-    timeout="20m",
-    secrets={"HF_TOKEN": "$HF_TOKEN"},
-)
-
-print(f"\n✅ Job submitted successfully!")
-print(f"Job ID: {job.job_id}")
-print(f"Monitor: https://huggingface.co/jobs/{job.job_id}")
-print(f"\nExpected time: ~10-15 minutes")
-print(f"Estimated cost: ~$0.20")
-print(f"\nThe job is running in the background!")
-print(f"📊 Once training starts, view metrics at: https://huggingface.co/spaces/evalstate/training-demo-dashboard")

train_demo.py

@@ -1,93 +0,0 @@
-# /// script
-# dependencies = [
-#     "trl>=0.12.0",
-#     "peft>=0.7.0",
-#     "transformers>=4.36.0",
-#     "accelerate>=0.24.0",
-#     "trackio",
-# ]
-# ///
-
-import trackio
-from datasets import load_dataset
-from peft import LoraConfig
-from trl import SFTTrainer, SFTConfig
-
-# Initialize Trackio for real-time monitoring
-trackio.init(
-    project="qwen-demo-sft",
-    space_id="evalstate/trackio-demo",  # Will auto-create if doesn't exist
-    config={
-        "model": "Qwen/Qwen2.5-0.5B",
-        "dataset": "trl-lib/Capybara",
-        "dataset_size": 50,
-        "learning_rate": 2e-5,
-        "max_steps": 20,
-        "demo": True,
-    }
-)
-
-# Load dataset (only 50 examples for quick demo)
-dataset = load_dataset("trl-lib/Capybara", split="train[:50]")
-print(f"✅ Dataset loaded: {len(dataset)} examples")
-print(f"📝 Sample: {dataset[0]}")
-
-# Training configuration
-config = SFTConfig(
-    # Hub settings - CRITICAL for saving results
-    output_dir="qwen-demo-sft",
-    push_to_hub=True,
-    hub_model_id="evalstate/qwen-demo-sft",
-    hub_strategy="end",  # Push only at end for demo
-    
-    # Training parameters (minimal for quick demo)
-    max_steps=20,  # Very short training
-    per_device_train_batch_size=2,
-    gradient_accumulation_steps=2,
-    learning_rate=2e-5,
-    
-    # Logging
-    logging_steps=5,
-    save_strategy="no",  # Don't save checkpoints during training
-    
-    # Optimization
-    warmup_steps=5,
-    lr_scheduler_type="cosine",
-    
-    # Monitoring
-    report_to="trackio",
-)
-
-# LoRA configuration (reduces memory usage)
-peft_config = LoraConfig(
-    r=8,  # Small rank for demo
-    lora_alpha=16,
-    lora_dropout=0.05,
-    bias="none",
-    task_type="CAUSAL_LM",
-    target_modules=["q_proj", "v_proj"],
-)
-
-# Initialize trainer
-print("🚀 Initializing trainer...")
-trainer = SFTTrainer(
-    model="Qwen/Qwen2.5-0.5B",
-    train_dataset=dataset,
-    args=config,
-    peft_config=peft_config,
-)
-
-# Train
-print("🔥 Starting training (20 steps)...")
-trainer.train()
-
-# Push to Hub
-print("💾 Pushing model to Hub...")
-trainer.push_to_hub()
-
-# Finish Trackio tracking
-trackio.finish()
-
-print("✅ Training complete!")
-print(f"📦 Model: https://huggingface.co/evalstate/qwen-demo-sft")
-print(f"📊 Metrics: https://huggingface.co/spaces/evalstate/trackio-demo")

train_minimal.py

@@ -1,28 +0,0 @@
-# /// script
-# dependencies = ["trl>=0.12.0", "peft>=0.7.0"]
-# ///
-
-from datasets import load_dataset
-from peft import LoraConfig
-from trl import SFTTrainer, SFTConfig
-
-# Load 50 examples
-dataset = load_dataset("trl-lib/Capybara", split="train[:50]")
-
-# Train with minimal config
-trainer = SFTTrainer(
-    model="Qwen/Qwen2.5-0.5B",
-    train_dataset=dataset,
-    peft_config=LoraConfig(r=8, lora_alpha=16),
-    args=SFTConfig(
-        output_dir="output",
-        push_to_hub=True,
-        hub_model_id="evalstate/qwen-demo-minimal",
-        max_steps=20,
-        report_to="none",  # No monitoring for quick demo
-    )
-)
-
-trainer.train()
-trainer.push_to_hub()
-print("✅ Done! Model at: https://huggingface.co/evalstate/qwen-demo-minimal")

train_organized.py

@@ -1,155 +0,0 @@
-# /// script
-# dependencies = ["trl>=0.12.0", "peft>=0.7.0", "trackio"]
-# ///
-
-import os
-from datasets import load_dataset
-from peft import LoraConfig
-from trl import SFTTrainer, SFTConfig
-
-# ============================================================================
-# Configuration
-# ============================================================================
-
-class Config:
-    """Training configuration with environment overrides"""
-    MODEL = os.getenv("MODEL", "Qwen/Qwen2.5-0.5B")
-    DATASET = os.getenv("DATASET", "trl-lib/Capybara")
-    DATASET_SPLIT = os.getenv("DATASET_SPLIT", "train[:50]")
-    OUTPUT_REPO = os.getenv("OUTPUT_REPO", "evalstate/qwen-demo")
-    
-    # Training
-    MAX_STEPS = int(os.getenv("MAX_STEPS", "20"))
-    BATCH_SIZE = int(os.getenv("BATCH_SIZE", "2"))
-    LEARNING_RATE = float(os.getenv("LEARNING_RATE", "2e-5"))
-    
-    # LoRA
-    LORA_R = int(os.getenv("LORA_R", "8"))
-    LORA_ALPHA = int(os.getenv("LORA_ALPHA", "16"))
-    
-    # Monitoring
-    TRACKIO_SPACE = os.getenv("TRACKIO_SPACE", "evalstate/ml-experiments")
-
-# ============================================================================
-# Setup Functions
-# ============================================================================
-
-def setup_monitoring(config: Config):
-    """Initialize Trackio for experiment tracking"""
-    import trackio
-    
-    project_name = config.OUTPUT_REPO.split('/')[-1]
-    
-    trackio.init(
-        project=project_name,
-        space_id=config.TRACKIO_SPACE,
-        config={
-            "model": config.MODEL,
-            "dataset": config.DATASET,
-            "max_steps": config.MAX_STEPS,
-            "learning_rate": config.LEARNING_RATE,
-            "lora_r": config.LORA_R,
-        }
-    )
-    
-    print(f"📊 Trackio: https://huggingface.co/spaces/{config.TRACKIO_SPACE}")
-
-def load_and_validate_dataset(config: Config):
-    """Load dataset and perform basic validation"""
-    dataset = load_dataset(config.DATASET, split=config.DATASET_SPLIT)
-    
-    print(f"✅ Dataset loaded: {len(dataset)} examples")
-    print(f"   Columns: {dataset.column_names}")
-    
-    # Basic validation
-    assert len(dataset) > 0, "Dataset is empty!"
-    assert "messages" in dataset.column_names, "Expected 'messages' column"
-    
-    return dataset
-
-def create_training_config(config: Config) -> SFTConfig:
-    """Create SFT training configuration"""
-    return SFTConfig(
-        # Output
-        output_dir="output",
-        push_to_hub=True,
-        hub_model_id=config.OUTPUT_REPO,
-        hub_strategy="end",
-        
-        # Training
-        max_steps=config.MAX_STEPS,
-        per_device_train_batch_size=config.BATCH_SIZE,
-        gradient_accumulation_steps=2,
-        learning_rate=config.LEARNING_RATE,
-        
-        # Optimization
-        warmup_ratio=0.1,
-        lr_scheduler_type="cosine",
-        
-        # Logging
-        logging_steps=max(1, config.MAX_STEPS // 4),
-        report_to="trackio",
-        save_strategy="no",
-    )
-
-def create_peft_config(config: Config) -> LoraConfig:
-    """Create LoRA/PEFT configuration"""
-    return LoraConfig(
-        r=config.LORA_R,
-        lora_alpha=config.LORA_ALPHA,
-        lora_dropout=0.05,
-        bias="none",
-        task_type="CAUSAL_LM",
-    )
-
-# ============================================================================
-# Main Training Flow
-# ============================================================================
-
-def main():
-    """Main training pipeline"""
-    config = Config()
-    
-    # Print configuration
-    print("🚀 Training Configuration:")
-    print(f"   Model: {config.MODEL}")
-    print(f"   Dataset: {config.DATASET} ({config.DATASET_SPLIT})")
-    print(f"   Output: {config.OUTPUT_REPO}")
-    print(f"   Steps: {config.MAX_STEPS}")
-    print(f"   Learning Rate: {config.LEARNING_RATE}")
-    print(f"   LoRA r={config.LORA_R}, alpha={config.LORA_ALPHA}")
-    print()
-    
-    # Setup
-    setup_monitoring(config)
-    dataset = load_and_validate_dataset(config)
-    training_config = create_training_config(config)
-    peft_config = create_peft_config(config)
-    
-    # Train
-    print("🔥 Initializing trainer...")
-    trainer = SFTTrainer(
-        model=config.MODEL,
-        train_dataset=dataset,
-        args=training_config,
-        peft_config=peft_config,
-    )
-    
-    print("🏃 Training started...")
-    trainer.train()
-    
-    print("💾 Pushing to Hub...")
-    trainer.push_to_hub()
-    
-    # Cleanup
-    import trackio
-    trackio.finish()
-    
-    # Summary
-    print()
-    print("✅ Training Complete!")
-    print(f"📦 Model: https://huggingface.co/{config.OUTPUT_REPO}")
-    print(f"📊 Metrics: https://huggingface.co/spaces/{config.TRACKIO_SPACE}")
-
-if __name__ == "__main__":
-    main()

train_production.py

@@ -1,95 +0,0 @@
-# /// script
-# dependencies = ["trl>=0.12.0", "peft>=0.7.0", "trackio"]
-# ///
-
-import os
-from datasets import load_dataset
-from peft import LoraConfig
-from trl import SFTTrainer, SFTConfig
-
-# Configuration from environment (with sensible defaults)
-MODEL = os.getenv("MODEL", "Qwen/Qwen2.5-0.5B")
-DATASET = os.getenv("DATASET", "trl-lib/Capybara")
-OUTPUT_REPO = os.getenv("OUTPUT_REPO", "evalstate/qwen-capybara-sft")
-MAX_STEPS = int(os.getenv("MAX_STEPS", "20"))
-BATCH_SIZE = int(os.getenv("BATCH_SIZE", "2"))
-LEARNING_RATE = float(os.getenv("LEARNING_RATE", "2e-5"))
-USE_TRACKIO = os.getenv("USE_TRACKIO", "true").lower() == "true"
-
-print(f"🚀 Training Configuration:")
-print(f"   Model: {MODEL}")
-print(f"   Dataset: {DATASET}")
-print(f"   Output: {OUTPUT_REPO}")
-print(f"   Max Steps: {MAX_STEPS}")
-print(f"   Monitoring: {'Trackio' if USE_TRACKIO else 'None'}")
-
-# Setup monitoring if enabled
-if USE_TRACKIO:
-    import trackio
-    trackio.init(
-        project=OUTPUT_REPO.split('/')[-1],  # Use model name as project
-        space_id="evalstate/ml-experiments",  # Single space for all
-        config={
-            "model": MODEL,
-            "dataset": DATASET,
-            "max_steps": MAX_STEPS,
-            "learning_rate": LEARNING_RATE,
-        }
-    )
-
-# Load dataset
-dataset = load_dataset(DATASET, split="train[:50]")
-print(f"✅ Loaded {len(dataset)} examples")
-
-# Configure training
-config = SFTConfig(
-    output_dir="output",
-    push_to_hub=True,
-    hub_model_id=OUTPUT_REPO,
-    hub_strategy="end",
-    
-    # Training params
-    max_steps=MAX_STEPS,
-    per_device_train_batch_size=BATCH_SIZE,
-    gradient_accumulation_steps=2,
-    learning_rate=LEARNING_RATE,
-    
-    # Optimization
-    warmup_ratio=0.1,
-    lr_scheduler_type="cosine",
-    
-    # Logging
-    logging_steps=max(1, MAX_STEPS // 4),
-    report_to="trackio" if USE_TRACKIO else "none",
-    
-    # No checkpoints for demo
-    save_strategy="no",
-)
-
-# LoRA config
-peft_config = LoraConfig(
-    r=8,
-    lora_alpha=16,
-    lora_dropout=0.05,
-    bias="none",
-    task_type="CAUSAL_LM",
-)
-
-# Train
-print("🔥 Starting training...")
-trainer = SFTTrainer(
-    model=MODEL,
-    train_dataset=dataset,
-    args=config,
-    peft_config=peft_config,
-)
-
-trainer.train()
-trainer.push_to_hub()
-
-if USE_TRACKIO:
-    trackio.finish()
-
-print(f"✅ Complete! Model: https://huggingface.co/{OUTPUT_REPO}")
-if USE_TRACKIO:
-    print(f"📊 Metrics: https://huggingface.co/spaces/evalstate/ml-experiments")

train_production_documented.py

@@ -1,203 +0,0 @@
-# /// script
-# dependencies = ["trl>=0.12.0", "peft>=0.7.0", "trackio"]
-# ///
-
-"""
-Production Training Script with Environment Configuration
-
-CUSTOMIZABLE PARAMETERS (via environment variables):
-    MODEL           - Model to fine-tune (default: Qwen/Qwen2.5-0.5B)
-    DATASET         - Dataset name on Hub (default: trl-lib/Capybara)
-    OUTPUT_REPO     - Where to save model (default: evalstate/qwen-capybara-sft)
-    MAX_STEPS       - Training steps (default: 20)
-    BATCH_SIZE      - Batch size per device (default: 2)
-    LEARNING_RATE   - Learning rate (default: 2e-5)
-    USE_TRACKIO     - Enable monitoring (default: true)
-
-EXAMPLE USAGE:
-    # Default (quick demo):
-    hf_jobs("uv", {"script": "train_production.py", "flavor": "t4-small"})
-    
-    # Custom settings:
-    hf_jobs("uv", {
-        "script": "train_production.py",
-        "flavor": "a10g-large",
-        "env": {
-            "MODEL": "meta-llama/Llama-3.2-1B",
-            "MAX_STEPS": "100",
-            "LEARNING_RATE": "1e-5"
-        }
-    })
-"""
-
-import os
-from datasets import load_dataset
-from peft import LoraConfig
-from trl import SFTTrainer, SFTConfig
-
-# ============================================================================
-# CONFIGURATION - Customize via environment variables
-# ============================================================================
-
-# Model Selection
-MODEL = os.getenv("MODEL", "Qwen/Qwen2.5-0.5B")
-# Common options:
-#   - Qwen/Qwen2.5-0.5B (fast, demo)
-#   - Qwen/Qwen2.5-3B (production)
-#   - meta-llama/Llama-3.2-1B
-#   - HuggingFaceTB/SmolLM2-1.7B
-
-# Dataset Selection
-DATASET = os.getenv("DATASET", "trl-lib/Capybara")
-# Use any conversational dataset with "messages" field
-# Examples: trl-lib/Capybara, HuggingFaceH4/ultrachat_200k
-
-# Output Configuration
-OUTPUT_REPO = os.getenv("OUTPUT_REPO", "evalstate/qwen-capybara-sft")
-# Must be in format: username/model-name
-
-# Training Parameters
-MAX_STEPS = int(os.getenv("MAX_STEPS", "20"))
-# Quick demo: 10-20 | Development: 100-500 | Production: 1000+
-
-BATCH_SIZE = int(os.getenv("BATCH_SIZE", "2"))
-# Adjust based on GPU memory: t4-small=2, a10g-large=4-8
-
-LEARNING_RATE = float(os.getenv("LEARNING_RATE", "2e-5"))
-# Typical range: 1e-5 to 5e-5
-
-# Monitoring
-USE_TRACKIO = os.getenv("USE_TRACKIO", "true").lower() == "true"
-# Set to "false" to disable real-time monitoring
-
-# ============================================================================
-# FIXED CONFIGURATION - Advanced users can modify these directly
-# ============================================================================
-
-# LoRA Configuration (reduces memory usage)
-LORA_R = 8              # Rank (higher = more parameters, better quality)
-LORA_ALPHA = 16         # Scaling factor (typically 2x LORA_R)
-LORA_DROPOUT = 0.05     # Dropout rate
-
-# Training Advanced
-GRADIENT_ACCUMULATION = 2   # Effective batch size = BATCH_SIZE * this
-WARMUP_RATIO = 0.1          # Percentage of steps for warmup
-LR_SCHEDULER = "cosine"     # Learning rate schedule
-
-# Logging
-LOGGING_STEPS = None    # Auto-calculated (MAX_STEPS // 4)
-
-# Trackio Space (single dashboard for all experiments)
-TRACKIO_SPACE = "evalstate/ml-experiments"
-
-# ============================================================================
-# TRAINING SCRIPT - No need to modify below this line
-# ============================================================================
-
-print("="*80)
-print("🚀 TRAINING CONFIGURATION")
-print("="*80)
-print(f"Model:          {MODEL}")
-print(f"Dataset:        {DATASET}")
-print(f"Output:         {OUTPUT_REPO}")
-print(f"Max Steps:      {MAX_STEPS}")
-print(f"Batch Size:     {BATCH_SIZE}")
-print(f"Learning Rate:  {LEARNING_RATE}")
-print(f"Monitoring:     {'Trackio' if USE_TRACKIO else 'Disabled'}")
-print(f"LoRA:           r={LORA_R}, alpha={LORA_ALPHA}")
-print("="*80)
-print()
-
-# Setup monitoring if enabled
-if USE_TRACKIO:
-    import trackio
-    trackio.init(
-        project=OUTPUT_REPO.split('/')[-1],  # Use model name as project
-        space_id=TRACKIO_SPACE,
-        config={
-            "model": MODEL,
-            "dataset": DATASET,
-            "max_steps": MAX_STEPS,
-            "batch_size": BATCH_SIZE,
-            "learning_rate": LEARNING_RATE,
-            "lora_r": LORA_R,
-        }
-    )
-    print(f"📊 Trackio Dashboard: https://huggingface.co/spaces/{TRACKIO_SPACE}")
-    print()
-
-# Load dataset (first 50 examples for demo)
-print("📦 Loading dataset...")
-dataset = load_dataset(DATASET, split="train[:50]")
-print(f"✅ Loaded {len(dataset)} examples")
-print()
-
-# Configure training
-logging_steps = LOGGING_STEPS or max(1, MAX_STEPS // 4)
-
-config = SFTConfig(
-    # Output
-    output_dir="output",
-    push_to_hub=True,
-    hub_model_id=OUTPUT_REPO,
-    hub_strategy="end",
-    
-    # Training params (customizable via env vars)
-    max_steps=MAX_STEPS,
-    per_device_train_batch_size=BATCH_SIZE,
-    gradient_accumulation_steps=GRADIENT_ACCUMULATION,
-    learning_rate=LEARNING_RATE,
-    
-    # Optimization
-    warmup_ratio=WARMUP_RATIO,
-    lr_scheduler_type=LR_SCHEDULER,
-    
-    # Logging
-    logging_steps=logging_steps,
-    report_to="trackio" if USE_TRACKIO else "none",
-    
-    # No checkpoints for demo (saves time)
-    save_strategy="no",
-)
-
-# LoRA configuration (reduces memory usage)
-peft_config = LoraConfig(
-    r=LORA_R,
-    lora_alpha=LORA_ALPHA,
-    lora_dropout=LORA_DROPOUT,
-    bias="none",
-    task_type="CAUSAL_LM",
-)
-
-# Initialize trainer
-print("🔥 Initializing trainer...")
-trainer = SFTTrainer(
-    model=MODEL,
-    train_dataset=dataset,
-    args=config,
-    peft_config=peft_config,
-)
-
-# Train
-print("🏃 Training started...")
-print()
-trainer.train()
-print()
-
-# Save to Hub
-print("💾 Pushing model to Hub...")
-trainer.push_to_hub()
-
-# Finish monitoring
-if USE_TRACKIO:
-    trackio.finish()
-
-# Summary
-print()
-print("="*80)
-print("✅ TRAINING COMPLETE!")
-print("="*80)
-print(f"📦 Model:   https://huggingface.co/{OUTPUT_REPO}")
-if USE_TRACKIO:
-    print(f"📊 Metrics: https://huggingface.co/spaces/{TRACKIO_SPACE}")
-print("="*80)