[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M74MN2P4GV9AECFMS6FBCSRE":3,"skill-same:01M74MN2P4GV9AECFMS6FBCSRE":94},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":17,"publishedAt":18,"linkOnly":19,"markdownContent":20,"markdownBody":21,"frontMatter":22,"refName":46,"commitSha":47,"skillDirectory":48,"forksCount":13,"repositoryHtmlUrl":49,"repositoryHomepageUrl":6,"ownerHtmlUrl":50,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":51,"install":52,"resources":65},"01M74MN2P4GV9AECFMS6FBCSRE","fine-tuning-with-trl",null,"TRL: SFT, DPO, PPO, GRPO, reward modeling for LLM RLHF.","LyrikNilpferd71\u002FHermes-Backup","LyrikNilpferd71","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F187173780?v=4","01M727S8GPB77P0S9ECRRMZT58","01M727S8HF63B0MD53F496MK1V",0,"MIT","https:\u002F\u002Fgithub.com\u002FLyrikNilpferd71\u002FHermes-Backup\u002Fblob\u002Fmain\u002Fsnapshots\u002F2026-07-26T14-01-09+0200\u002Fskills\u002Fmlops\u002Ftraining\u002Ftrl-fine-tuning\u002FSKILL.md",[],[],"2026-11-11 15:11:14",false,"---\nname: fine-tuning-with-trl\ndescription: \"TRL: SFT, DPO, PPO, GRPO, reward modeling for LLM RLHF.\"\nversion: 1.0.0\nauthor: Orchestra Research\nlicense: MIT\ndependencies: [trl, transformers, datasets, peft, accelerate, torch]\nmetadata:\n  hermes:\n    tags: [Post-Training, TRL, Reinforcement Learning, Fine-Tuning, SFT, DPO, PPO, GRPO, RLHF, Preference Alignment, HuggingFace]\n\n---\n\n# TRL - Transformer Reinforcement Learning\n\n## Quick start\n\nTRL provides post-training methods for aligning language models with human preferences.\n\n**Installation**:\n```bash\npip install trl transformers datasets peft accelerate\n```\n\n**Supervised Fine-Tuning** (instruction tuning):\n```python\nfrom trl import SFTTrainer\n\ntrainer = SFTTrainer(\n    model=\"Qwen\u002FQwen2.5-0.5B\",\n    train_dataset=dataset,  # Prompt-completion pairs\n)\ntrainer.train()\n```\n\n**DPO** (align with preferences):\n```python\nfrom trl import DPOTrainer, DPOConfig\n\nconfig = DPOConfig(output_dir=\"model-dpo\", beta=0.1)\ntrainer = DPOTrainer(\n    model=model,\n    args=config,\n    train_dataset=preference_dataset,  # chosen\u002Frejected pairs\n    processing_class=tokenizer\n)\ntrainer.train()\n```\n\n## Common workflows\n\n### Workflow 1: Full RLHF pipeline (SFT → Reward Model → PPO)\n\nComplete pipeline from base model to human-aligned model.\n\nCopy this checklist:\n\n```\nRLHF Training:\n- [ ] Step 1: Supervised fine-tuning (SFT)\n- [ ] Step 2: Train reward model\n- [ ] Step 3: PPO reinforcement learning\n- [ ] Step 4: Evaluate aligned model\n```\n\n**Step 1: Supervised fine-tuning**\n\nTrain base model on instruction-following data:\n\n```python\nfrom transformers import AutoModelForCausalLM, AutoTokenizer\nfrom trl import SFTTrainer, SFTConfig\nfrom datasets import load_dataset\n\n# Load model\nmodel = AutoModelForCausalLM.from_pretrained(\"Qwen\u002FQwen2.5-0.5B\")\ntokenizer = AutoTokenizer.from_pretrained(\"Qwen\u002FQwen2.5-0.5B\")\n\n# Load instruction dataset\ndataset = load_dataset(\"trl-lib\u002FCapybara\", split=\"train\")\n\n# Configure training\ntraining_args = SFTConfig(\n    output_dir=\"Qwen2.5-0.5B-SFT\",\n    per_device_train_batch_size=4,\n    num_train_epochs=1,\n    learning_rate=2e-5,\n    logging_steps=10,\n    save_strategy=\"epoch\"\n)\n\n# Train\ntrainer = SFTTrainer(\n    model=model,\n    args=training_args,\n    train_dataset=dataset,\n    tokenizer=tokenizer\n)\ntrainer.train()\ntrainer.save_model()\n```\n\n**Step 2: Train reward model**\n\nTrain model to predict human preferences:\n\n```python\nfrom transformers import AutoModelForSequenceClassification\nfrom trl import RewardTrainer, RewardConfig\n\n# Load SFT model as base\nmodel = AutoModelForSequenceClassification.from_pretrained(\n    \"Qwen2.5-0.5B-SFT\",\n    num_labels=1  # Single reward score\n)\ntokenizer = AutoTokenizer.from_pretrained(\"Qwen2.5-0.5B-SFT\")\n\n# Load preference data (chosen\u002Frejected pairs)\ndataset = load_dataset(\"trl-lib\u002Fultrafeedback_binarized\", split=\"train\")\n\n# Configure training\ntraining_args = RewardConfig(\n    output_dir=\"Qwen2.5-0.5B-Reward\",\n    per_device_train_batch_size=2,\n    num_train_epochs=1,\n    learning_rate=1e-5\n)\n\n# Train reward model\ntrainer = RewardTrainer(\n    model=model,\n    args=training_args,\n    processing_class=tokenizer,\n    train_dataset=dataset\n)\ntrainer.train()\ntrainer.save_model()\n```\n\n**Step 3: PPO reinforcement learning**\n\nOptimize policy using reward model:\n\n```bash\npython -m trl.scripts.ppo \\\n    --model_name_or_path Qwen2.5-0.5B-SFT \\\n    --reward_model_path Qwen2.5-0.5B-Reward \\\n    --dataset_name trl-internal-testing\u002Fdescriptiveness-sentiment-trl-style \\\n    --output_dir Qwen2.5-0.5B-PPO \\\n    --learning_rate 3e-6 \\\n    --per_device_train_batch_size 64 \\\n    --total_episodes 10000\n```\n\n**Step 4: Evaluate**\n\n```python\nfrom transformers import pipeline\n\n# Load aligned model\ngenerator = pipeline(\"text-generation\", model=\"Qwen2.5-0.5B-PPO\")\n\n# Test\nprompt = \"Explain quantum computing to a 10-year-old\"\noutput = generator(prompt, max_length=200)[0][\"generated_text\"]\nprint(output)\n```\n\n### Workflow 2: Simple preference alignment with DPO\n\nAlign model with preferences without reward model.\n\nCopy this checklist:\n\n```\nDPO Training:\n- [ ] Step 1: Prepare preference dataset\n- [ ] Step 2: Configure DPO\n- [ ] Step 3: Train with DPOTrainer\n- [ ] Step 4: Evaluate alignment\n```\n\n**Step 1: Prepare preference dataset**\n\nDataset format:\n```json\n{\n  \"prompt\": \"What is the capital of France?\",\n  \"chosen\": \"The capital of France is Paris.\",\n  \"rejected\": \"I don't know.\"\n}\n```\n\nLoad dataset:\n```python\nfrom datasets import load_dataset\n\ndataset = load_dataset(\"trl-lib\u002Fultrafeedback_binarized\", split=\"train\")\n# Or load your own\n# dataset = load_dataset(\"json\", data_files=\"preferences.json\")\n```\n\n**Step 2: Configure DPO**\n\n```python\nfrom trl import DPOConfig\n\nconfig = DPOConfig(\n    output_dir=\"Qwen2.5-0.5B-DPO\",\n    per_device_train_batch_size=4,\n    num_train_epochs=1,\n    learning_rate=5e-7,\n    beta=0.1,  # KL penalty strength\n    max_prompt_length=512,\n    max_length=1024,\n    logging_steps=10\n)\n```\n\n**Step 3: Train with DPOTrainer**\n\n```python\nfrom transformers import AutoModelForCausalLM, AutoTokenizer\nfrom trl import DPOTrainer\n\nmodel = AutoModelForCausalLM.from_pretrained(\"Qwen\u002FQwen2.5-0.5B-Instruct\")\ntokenizer = AutoTokenizer.from_pretrained(\"Qwen\u002FQwen2.5-0.5B-Instruct\")\n\ntrainer = DPOTrainer(\n    model=model,\n    args=config,\n    train_dataset=dataset,\n    processing_class=tokenizer\n)\n\ntrainer.train()\ntrainer.save_model()\n```\n\n**CLI alternative**:\n```bash\ntrl dpo \\\n    --model_name_or_path Qwen\u002FQwen2.5-0.5B-Instruct \\\n    --dataset_name argilla\u002FCapybara-Preferences \\\n    --output_dir Qwen2.5-0.5B-DPO \\\n    --per_device_train_batch_size 4 \\\n    --learning_rate 5e-7 \\\n    --beta 0.1\n```\n\n### Workflow 3: Memory-efficient online RL with GRPO\n\nTrain with reinforcement learning using minimal memory.\n\nFor in-depth GRPO guidance — reward function design, critical training insights (loss behavior, mode collapse, tuning), and advanced multi-stage patterns — see **[references\u002Fgrpo-training.md](references\u002Fgrpo-training.md)**. A production-ready training script is in **[templates\u002Fbasic_grpo_training.py](templates\u002Fbasic_grpo_training.py)**.\n\nCopy this checklist:\n\n```\nGRPO Training:\n- [ ] Step 1: Define reward function\n- [ ] Step 2: Configure GRPO\n- [ ] Step 3: Train with GRPOTrainer\n```\n\n**Step 1: Define reward function**\n\n```python\ndef reward_function(completions, **kwargs):\n    \"\"\"\n    Compute rewards for completions.\n\n    Args:\n        completions: List of generated texts\n\n    Returns:\n        List of reward scores (floats)\n    \"\"\"\n    rewards = []\n    for completion in completions:\n        # Example: reward based on length and unique words\n        score = len(completion.split())  # Favor longer responses\n        score += len(set(completion.lower().split()))  # Reward unique words\n        rewards.append(score)\n    return rewards\n```\n\nOr use a reward model:\n```python\nfrom transformers import pipeline\n\nreward_model = pipeline(\"text-classification\", model=\"reward-model-path\")\n\ndef reward_from_model(completions, prompts, **kwargs):\n    # Combine prompt + completion\n    full_texts = [p + c for p, c in zip(prompts, completions)]\n    # Get reward scores\n    results = reward_model(full_texts)\n    return [r[\"score\"] for r in results]\n```\n\n**Step 2: Configure GRPO**\n\n```python\nfrom trl import GRPOConfig\n\nconfig = GRPOConfig(\n    output_dir=\"Qwen2-GRPO\",\n    per_device_train_batch_size=4,\n    num_train_epochs=1,\n    learning_rate=1e-5,\n    num_generations=4,  # Generate 4 completions per prompt\n    max_new_tokens=128\n)\n```\n\n**Step 3: Train with GRPOTrainer**\n\n```python\nfrom datasets import load_dataset\nfrom trl import GRPOTrainer\n\n# Load prompt-only dataset\ndataset = load_dataset(\"trl-lib\u002Ftldr\", split=\"train\")\n\ntrainer = GRPOTrainer(\n    model=\"Qwen\u002FQwen2-0.5B-Instruct\",\n    reward_funcs=reward_function,  # Your reward function\n    args=config,\n    train_dataset=dataset\n)\n\ntrainer.train()\n```\n\n**CLI**:\n```bash\ntrl grpo \\\n    --model_name_or_path Qwen\u002FQwen2-0.5B-Instruct \\\n    --dataset_name trl-lib\u002Ftldr \\\n    --output_dir Qwen2-GRPO \\\n    --num_generations 4\n```\n\n## When to use vs alternatives\n\n**Use TRL when:**\n- Need to align model with human preferences\n- Have preference data (chosen\u002Frejected pairs)\n- Want to use reinforcement learning (PPO, GRPO)\n- Need reward model training\n- Doing RLHF (full pipeline)\n\n**Method selection**:\n- **SFT**: Have prompt-completion pairs, want basic instruction following\n- **DPO**: Have preferences, want simple alignment (no reward model needed)\n- **PPO**: Have reward model, need maximum control over RL\n- **GRPO**: Memory-constrained, want online RL\n- **Reward Model**: Building RLHF pipeline, need to score generations\n\n**Use alternatives instead:**\n- **HuggingFace Trainer**: Basic fine-tuning without RL\n- **Axolotl**: YAML-based training configuration\n- **LitGPT**: Educational, minimal fine-tuning\n- **Unsloth**: Fast LoRA training\n\n## Common issues\n\n**Issue: OOM during DPO training**\n\nReduce batch size and sequence length:\n```python\nconfig = DPOConfig(\n    per_device_train_batch_size=1,  # Reduce from 4\n    max_length=512,  # Reduce from 1024\n    gradient_accumulation_steps=8  # Maintain effective batch\n)\n```\n\nOr use gradient checkpointing:\n```python\nmodel.gradient_checkpointing_enable()\n```\n\n**Issue: Poor alignment quality**\n\nTune beta parameter:\n```python\n# Higher beta = more conservative (stays closer to reference)\nconfig = DPOConfig(beta=0.5)  # Default 0.1\n\n# Lower beta = more aggressive alignment\nconfig = DPOConfig(beta=0.01)\n```\n\n**Issue: Reward model not learning**\n\nCheck loss type and learning rate:\n```python\nconfig = RewardConfig(\n    learning_rate=1e-5,  # Try different LR\n    num_train_epochs=3  # Train longer\n)\n```\n\nEnsure preference dataset has clear winners:\n```python\n# Verify dataset\nprint(dataset[0])\n# Should have clear chosen > rejected\n```\n\n**Issue: PPO training unstable**\n\nAdjust KL coefficient:\n```python\nconfig = PPOConfig(\n    kl_coef=0.1,  # Increase from 0.05\n    cliprange=0.1  # Reduce from 0.2\n)\n```\n\n## Advanced topics\n\n**SFT training guide**: See [references\u002Fsft-training.md](references\u002Fsft-training.md) for dataset formats, chat templates, packing strategies, and multi-GPU training.\n\n**DPO variants**: See [references\u002Fdpo-variants.md](references\u002Fdpo-variants.md) for IPO, cDPO, RPO, and other DPO loss functions with recommended hyperparameters.\n\n**Reward modeling**: See [references\u002Freward-modeling.md](references\u002Freward-modeling.md) for outcome vs process rewards, Bradley-Terry loss, and reward model evaluation.\n\n**Online RL methods**: See [references\u002Fonline-rl.md](references\u002Fonline-rl.md) for PPO, GRPO, RLOO, and OnlineDPO with detailed configurations.\n\n**GRPO deep dive**: See [references\u002Fgrpo-training.md](references\u002Fgrpo-training.md) for expert-level GRPO patterns — reward function design philosophy, training insights (why loss increases, mode collapse detection), hyperparameter tuning, multi-stage training, and troubleshooting. Production-ready template in [templates\u002Fbasic_grpo_training.py](templates\u002Fbasic_grpo_training.py).\n\n## Hardware requirements\n\n- **GPU**: NVIDIA (CUDA required)\n- **VRAM**: Depends on model and method\n  - SFT 7B: 16GB (with LoRA)\n  - DPO 7B: 24GB (stores reference model)\n  - PPO 7B: 40GB (policy + reward model)\n  - GRPO 7B: 24GB (more memory efficient)\n- **Multi-GPU**: Supported via `accelerate`\n- **Mixed precision**: BF16 recommended (A100\u002FH100)\n\n**Memory optimization**:\n- Use LoRA\u002FQLoRA for all methods\n- Enable gradient checkpointing\n- Use smaller batch sizes with gradient accumulation\n\n## Resources\n\n- Docs: https:\u002F\u002Fhuggingface.co\u002Fdocs\u002Ftrl\u002F\n- GitHub: https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Ftrl\n- Papers:\n  - \"Training language models to follow instructions with human feedback\" (InstructGPT, 2022)\n  - \"Direct Preference Optimization: Your Language Model is Secretly a Reward Model\" (DPO, 2023)\n  - \"Group Relative Policy Optimization\" (GRPO, 2024)\n- Examples: https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Ftrl\u002Ftree\u002Fmain\u002Fexamples\u002Fscripts\n\n\n\n","# TRL - Transformer Reinforcement Learning\n\n## Quick start\n\nTRL provides post-training methods for aligning language models with human preferences.\n\n**Installation**:\n```bash\npip install trl transformers datasets peft accelerate\n```\n\n**Supervised Fine-Tuning** (instruction tuning):\n```python\nfrom trl import SFTTrainer\n\ntrainer = SFTTrainer(\n    model=\"Qwen\u002FQwen2.5-0.5B\",\n    train_dataset=dataset,  # Prompt-completion pairs\n)\ntrainer.train()\n```\n\n**DPO** (align with preferences):\n```python\nfrom trl import DPOTrainer, DPOConfig\n\nconfig = DPOConfig(output_dir=\"model-dpo\", beta=0.1)\ntrainer = DPOTrainer(\n    model=model,\n    args=config,\n    train_dataset=preference_dataset,  # chosen\u002Frejected pairs\n    processing_class=tokenizer\n)\ntrainer.train()\n```\n\n## Common workflows\n\n### Workflow 1: Full RLHF pipeline (SFT → Reward Model → PPO)\n\nComplete pipeline from base model to human-aligned model.\n\nCopy this checklist:\n\n```\nRLHF Training:\n- [ ] Step 1: Supervised fine-tuning (SFT)\n- [ ] Step 2: Train reward model\n- [ ] Step 3: PPO reinforcement learning\n- [ ] Step 4: Evaluate aligned model\n```\n\n**Step 1: Supervised fine-tuning**\n\nTrain base model on instruction-following data:\n\n```python\nfrom transformers import AutoModelForCausalLM, AutoTokenizer\nfrom trl import SFTTrainer, SFTConfig\nfrom datasets import load_dataset\n\n# Load model\nmodel = AutoModelForCausalLM.from_pretrained(\"Qwen\u002FQwen2.5-0.5B\")\ntokenizer = AutoTokenizer.from_pretrained(\"Qwen\u002FQwen2.5-0.5B\")\n\n# Load instruction dataset\ndataset = load_dataset(\"trl-lib\u002FCapybara\", split=\"train\")\n\n# Configure training\ntraining_args = SFTConfig(\n    output_dir=\"Qwen2.5-0.5B-SFT\",\n    per_device_train_batch_size=4,\n    num_train_epochs=1,\n    learning_rate=2e-5,\n    logging_steps=10,\n    save_strategy=\"epoch\"\n)\n\n# Train\ntrainer = SFTTrainer(\n    model=model,\n    args=training_args,\n    train_dataset=dataset,\n    tokenizer=tokenizer\n)\ntrainer.train()\ntrainer.save_model()\n```\n\n**Step 2: Train reward model**\n\nTrain model to predict human preferences:\n\n```python\nfrom transformers import AutoModelForSequenceClassification\nfrom trl import RewardTrainer, RewardConfig\n\n# Load SFT model as base\nmodel = AutoModelForSequenceClassification.from_pretrained(\n    \"Qwen2.5-0.5B-SFT\",\n    num_labels=1  # Single reward score\n)\ntokenizer = AutoTokenizer.from_pretrained(\"Qwen2.5-0.5B-SFT\")\n\n# Load preference data (chosen\u002Frejected pairs)\ndataset = load_dataset(\"trl-lib\u002Fultrafeedback_binarized\", split=\"train\")\n\n# Configure training\ntraining_args = RewardConfig(\n    output_dir=\"Qwen2.5-0.5B-Reward\",\n    per_device_train_batch_size=2,\n    num_train_epochs=1,\n    learning_rate=1e-5\n)\n\n# Train reward model\ntrainer = RewardTrainer(\n    model=model,\n    args=training_args,\n    processing_class=tokenizer,\n    train_dataset=dataset\n)\ntrainer.train()\ntrainer.save_model()\n```\n\n**Step 3: PPO reinforcement learning**\n\nOptimize policy using reward model:\n\n```bash\npython -m trl.scripts.ppo \\\n    --model_name_or_path Qwen2.5-0.5B-SFT \\\n    --reward_model_path Qwen2.5-0.5B-Reward \\\n    --dataset_name trl-internal-testing\u002Fdescriptiveness-sentiment-trl-style \\\n    --output_dir Qwen2.5-0.5B-PPO \\\n    --learning_rate 3e-6 \\\n    --per_device_train_batch_size 64 \\\n    --total_episodes 10000\n```\n\n**Step 4: Evaluate**\n\n```python\nfrom transformers import pipeline\n\n# Load aligned model\ngenerator = pipeline(\"text-generation\", model=\"Qwen2.5-0.5B-PPO\")\n\n# Test\nprompt = \"Explain quantum computing to a 10-year-old\"\noutput = generator(prompt, max_length=200)[0][\"generated_text\"]\nprint(output)\n```\n\n### Workflow 2: Simple preference alignment with DPO\n\nAlign model with preferences without reward model.\n\nCopy this checklist:\n\n```\nDPO Training:\n- [ ] Step 1: Prepare preference dataset\n- [ ] Step 2: Configure DPO\n- [ ] Step 3: Train with DPOTrainer\n- [ ] Step 4: Evaluate alignment\n```\n\n**Step 1: Prepare preference dataset**\n\nDataset format:\n```json\n{\n  \"prompt\": \"What is the capital of France?\",\n  \"chosen\": \"The capital of France is Paris.\",\n  \"rejected\": \"I don't know.\"\n}\n```\n\nLoad dataset:\n```python\nfrom datasets import load_dataset\n\ndataset = load_dataset(\"trl-lib\u002Fultrafeedback_binarized\", split=\"train\")\n# Or load your own\n# dataset = load_dataset(\"json\", data_files=\"preferences.json\")\n```\n\n**Step 2: Configure DPO**\n\n```python\nfrom trl import DPOConfig\n\nconfig = DPOConfig(\n    output_dir=\"Qwen2.5-0.5B-DPO\",\n    per_device_train_batch_size=4,\n    num_train_epochs=1,\n    learning_rate=5e-7,\n    beta=0.1,  # KL penalty strength\n    max_prompt_length=512,\n    max_length=1024,\n    logging_steps=10\n)\n```\n\n**Step 3: Train with DPOTrainer**\n\n```python\nfrom transformers import AutoModelForCausalLM, AutoTokenizer\nfrom trl import DPOTrainer\n\nmodel = AutoModelForCausalLM.from_pretrained(\"Qwen\u002FQwen2.5-0.5B-Instruct\")\ntokenizer = AutoTokenizer.from_pretrained(\"Qwen\u002FQwen2.5-0.5B-Instruct\")\n\ntrainer = DPOTrainer(\n    model=model,\n    args=config,\n    train_dataset=dataset,\n    processing_class=tokenizer\n)\n\ntrainer.train()\ntrainer.save_model()\n```\n\n**CLI alternative**:\n```bash\ntrl dpo \\\n    --model_name_or_path Qwen\u002FQwen2.5-0.5B-Instruct \\\n    --dataset_name argilla\u002FCapybara-Preferences \\\n    --output_dir Qwen2.5-0.5B-DPO \\\n    --per_device_train_batch_size 4 \\\n    --learning_rate 5e-7 \\\n    --beta 0.1\n```\n\n### Workflow 3: Memory-efficient online RL with GRPO\n\nTrain with reinforcement learning using minimal memory.\n\nFor in-depth GRPO guidance — reward function design, critical training insights (loss behavior, mode collapse, tuning), and advanced multi-stage patterns — see **[references\u002Fgrpo-training.md](references\u002Fgrpo-training.md)**. A production-ready training script is in **[templates\u002Fbasic_grpo_training.py](templates\u002Fbasic_grpo_training.py)**.\n\nCopy this checklist:\n\n```\nGRPO Training:\n- [ ] Step 1: Define reward function\n- [ ] Step 2: Configure GRPO\n- [ ] Step 3: Train with GRPOTrainer\n```\n\n**Step 1: Define reward function**\n\n```python\ndef reward_function(completions, **kwargs):\n    \"\"\"\n    Compute rewards for completions.\n\n    Args:\n        completions: List of generated texts\n\n    Returns:\n        List of reward scores (floats)\n    \"\"\"\n    rewards = []\n    for completion in completions:\n        # Example: reward based on length and unique words\n        score = len(completion.split())  # Favor longer responses\n        score += len(set(completion.lower().split()))  # Reward unique words\n        rewards.append(score)\n    return rewards\n```\n\nOr use a reward model:\n```python\nfrom transformers import pipeline\n\nreward_model = pipeline(\"text-classification\", model=\"reward-model-path\")\n\ndef reward_from_model(completions, prompts, **kwargs):\n    # Combine prompt + completion\n    full_texts = [p + c for p, c in zip(prompts, completions)]\n    # Get reward scores\n    results = reward_model(full_texts)\n    return [r[\"score\"] for r in results]\n```\n\n**Step 2: Configure GRPO**\n\n```python\nfrom trl import GRPOConfig\n\nconfig = GRPOConfig(\n    output_dir=\"Qwen2-GRPO\",\n    per_device_train_batch_size=4,\n    num_train_epochs=1,\n    learning_rate=1e-5,\n    num_generations=4,  # Generate 4 completions per prompt\n    max_new_tokens=128\n)\n```\n\n**Step 3: Train with GRPOTrainer**\n\n```python\nfrom datasets import load_dataset\nfrom trl import GRPOTrainer\n\n# Load prompt-only dataset\ndataset = load_dataset(\"trl-lib\u002Ftldr\", split=\"train\")\n\ntrainer = GRPOTrainer(\n    model=\"Qwen\u002FQwen2-0.5B-Instruct\",\n    reward_funcs=reward_function,  # Your reward function\n    args=config,\n    train_dataset=dataset\n)\n\ntrainer.train()\n```\n\n**CLI**:\n```bash\ntrl grpo \\\n    --model_name_or_path Qwen\u002FQwen2-0.5B-Instruct \\\n    --dataset_name trl-lib\u002Ftldr \\\n    --output_dir Qwen2-GRPO \\\n    --num_generations 4\n```\n\n## When to use vs alternatives\n\n**Use TRL when:**\n- Need to align model with human preferences\n- Have preference data (chosen\u002Frejected pairs)\n- Want to use reinforcement learning (PPO, GRPO)\n- Need reward model training\n- Doing RLHF (full pipeline)\n\n**Method selection**:\n- **SFT**: Have prompt-completion pairs, want basic instruction following\n- **DPO**: Have preferences, want simple alignment (no reward model needed)\n- **PPO**: Have reward model, need maximum control over RL\n- **GRPO**: Memory-constrained, want online RL\n- **Reward Model**: Building RLHF pipeline, need to score generations\n\n**Use alternatives instead:**\n- **HuggingFace Trainer**: Basic fine-tuning without RL\n- **Axolotl**: YAML-based training configuration\n- **LitGPT**: Educational, minimal fine-tuning\n- **Unsloth**: Fast LoRA training\n\n## Common issues\n\n**Issue: OOM during DPO training**\n\nReduce batch size and sequence length:\n```python\nconfig = DPOConfig(\n    per_device_train_batch_size=1,  # Reduce from 4\n    max_length=512,  # Reduce from 1024\n    gradient_accumulation_steps=8  # Maintain effective batch\n)\n```\n\nOr use gradient checkpointing:\n```python\nmodel.gradient_checkpointing_enable()\n```\n\n**Issue: Poor alignment quality**\n\nTune beta parameter:\n```python\n# Higher beta = more conservative (stays closer to reference)\nconfig = DPOConfig(beta=0.5)  # Default 0.1\n\n# Lower beta = more aggressive alignment\nconfig = DPOConfig(beta=0.01)\n```\n\n**Issue: Reward model not learning**\n\nCheck loss type and learning rate:\n```python\nconfig = RewardConfig(\n    learning_rate=1e-5,  # Try different LR\n    num_train_epochs=3  # Train longer\n)\n```\n\nEnsure preference dataset has clear winners:\n```python\n# Verify dataset\nprint(dataset[0])\n# Should have clear chosen > rejected\n```\n\n**Issue: PPO training unstable**\n\nAdjust KL coefficient:\n```python\nconfig = PPOConfig(\n    kl_coef=0.1,  # Increase from 0.05\n    cliprange=0.1  # Reduce from 0.2\n)\n```\n\n## Advanced topics\n\n**SFT training guide**: See [references\u002Fsft-training.md](references\u002Fsft-training.md) for dataset formats, chat templates, packing strategies, and multi-GPU training.\n\n**DPO variants**: See [references\u002Fdpo-variants.md](references\u002Fdpo-variants.md) for IPO, cDPO, RPO, and other DPO loss functions with recommended hyperparameters.\n\n**Reward modeling**: See [references\u002Freward-modeling.md](references\u002Freward-modeling.md) for outcome vs process rewards, Bradley-Terry loss, and reward model evaluation.\n\n**Online RL methods**: See [references\u002Fonline-rl.md](references\u002Fonline-rl.md) for PPO, GRPO, RLOO, and OnlineDPO with detailed configurations.\n\n**GRPO deep dive**: See [references\u002Fgrpo-training.md](references\u002Fgrpo-training.md) for expert-level GRPO patterns — reward function design philosophy, training insights (why loss increases, mode collapse detection), hyperparameter tuning, multi-stage training, and troubleshooting. Production-ready template in [templates\u002Fbasic_grpo_training.py](templates\u002Fbasic_grpo_training.py).\n\n## Hardware requirements\n\n- **GPU**: NVIDIA (CUDA required)\n- **VRAM**: Depends on model and method\n  - SFT 7B: 16GB (with LoRA)\n  - DPO 7B: 24GB (stores reference model)\n  - PPO 7B: 40GB (policy + reward model)\n  - GRPO 7B: 24GB (more memory efficient)\n- **Multi-GPU**: Supported via `accelerate`\n- **Mixed precision**: BF16 recommended (A100\u002FH100)\n\n**Memory optimization**:\n- Use LoRA\u002FQLoRA for all methods\n- Enable gradient checkpointing\n- Use smaller batch sizes with gradient accumulation\n\n## Resources\n\n- Docs: https:\u002F\u002Fhuggingface.co\u002Fdocs\u002Ftrl\u002F\n- GitHub: https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Ftrl\n- Papers:\n  - \"Training language models to follow instructions with human feedback\" (InstructGPT, 2022)\n  - \"Direct Preference Optimization: Your Language Model is Secretly a Reward Model\" (DPO, 2023)\n  - \"Group Relative Policy Optimization\" (GRPO, 2024)\n- Examples: https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Ftrl\u002Ftree\u002Fmain\u002Fexamples\u002Fscripts\n\n\n\n",{"name":5,"author":23,"license":14,"version":24,"metadata":25,"description":7,"dependencies":39},"Orchestra Research","1.0.0",{"hermes":26},{"tags":27},[28,29,30,31,32,33,34,35,36,37,38],"Post-Training","TRL","Reinforcement Learning","Fine-Tuning","SFT","DPO","PPO","GRPO","RLHF","Preference Alignment","HuggingFace",[40,41,42,43,44,45],"trl","transformers","datasets","peft","accelerate","torch","main","404fd33aabef3b6bc1cc913117ba2160ade62625","snapshots\u002F2026-07-26T14-01-09+0200\u002Fskills\u002Fmlops\u002Ftraining\u002Ftrl-fine-tuning","https:\u002F\u002Fgithub.com\u002FLyrikNilpferd71\u002FHermes-Backup","https:\u002F\u002Fgithub.com\u002FLyrikNilpferd71","2026-11-08 20:00:57",{"source":8,"skillName":5,"commands":53,"prompt":63,"zipAvailable":64},[54,57,60],{"tool":55,"command":56},"npx","npx skills add LyrikNilpferd71\u002FHermes-Backup --skill fine-tuning-with-trl",{"tool":58,"command":59},"bunx","bunx skills add LyrikNilpferd71\u002FHermes-Backup --skill fine-tuning-with-trl",{"tool":61,"command":62},"pnpm","pnpm dlx skills add LyrikNilpferd71\u002FHermes-Backup --skill fine-tuning-with-trl","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FLyrikNilpferd71\u002FHermes-Backup\u002Fblob\u002Fmain\u002Fsnapshots\u002F2026-07-26T14-01-09+0200\u002Fskills\u002Fmlops\u002Ftraining\u002Ftrl-fine-tuning\u002FSKILL.md",true,[66,72,76,80,84,88],{"path":67,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":70,"sourceHtmlUrl":71},"references\u002Fdpo-variants.md","REFERENCE","参考文档",4297,"https:\u002F\u002Fgithub.com\u002FLyrikNilpferd71\u002FHermes-Backup\u002Fblob\u002Fmain\u002Fsnapshots\u002F2026-07-26T14-01-09+0200\u002Fskills\u002Fmlops\u002Ftraining\u002Ftrl-fine-tuning\u002Freferences\u002Fdpo-variants.md",{"path":73,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":74,"sourceHtmlUrl":75},"references\u002Fgrpo-training.md",15901,"https:\u002F\u002Fgithub.com\u002FLyrikNilpferd71\u002FHermes-Backup\u002Fblob\u002Fmain\u002Fsnapshots\u002F2026-07-26T14-01-09+0200\u002Fskills\u002Fmlops\u002Ftraining\u002Ftrl-fine-tuning\u002Freferences\u002Fgrpo-training.md",{"path":77,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":78,"sourceHtmlUrl":79},"references\u002Fonline-rl.md",1971,"https:\u002F\u002Fgithub.com\u002FLyrikNilpferd71\u002FHermes-Backup\u002Fblob\u002Fmain\u002Fsnapshots\u002F2026-07-26T14-01-09+0200\u002Fskills\u002Fmlops\u002Ftraining\u002Ftrl-fine-tuning\u002Freferences\u002Fonline-rl.md",{"path":81,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":82,"sourceHtmlUrl":83},"references\u002Freward-modeling.md",2597,"https:\u002F\u002Fgithub.com\u002FLyrikNilpferd71\u002FHermes-Backup\u002Fblob\u002Fmain\u002Fsnapshots\u002F2026-07-26T14-01-09+0200\u002Fskills\u002Fmlops\u002Ftraining\u002Ftrl-fine-tuning\u002Freferences\u002Freward-modeling.md",{"path":85,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":86,"sourceHtmlUrl":87},"references\u002Fsft-training.md",3237,"https:\u002F\u002Fgithub.com\u002FLyrikNilpferd71\u002FHermes-Backup\u002Fblob\u002Fmain\u002Fsnapshots\u002F2026-07-26T14-01-09+0200\u002Fskills\u002Fmlops\u002Ftraining\u002Ftrl-fine-tuning\u002Freferences\u002Fsft-training.md",{"path":89,"resourceType":90,"resourceTypeDesc":91,"sizeBytes":92,"sourceHtmlUrl":93},"templates\u002Fbasic_grpo_training.py","SCRIPT","脚本",6113,"https:\u002F\u002Fgithub.com\u002FLyrikNilpferd71\u002FHermes-Backup\u002Fblob\u002Fmain\u002Fsnapshots\u002F2026-07-26T14-01-09+0200\u002Fskills\u002Fmlops\u002Ftraining\u002Ftrl-fine-tuning\u002Ftemplates\u002Fbasic_grpo_training.py",[95,103,110,117,124,131],{"skillNo":96,"name":97,"nameZh":6,"description":98,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":99,"categories":100,"keywords":101,"publishedAt":102},"01M76BBCCNA7WQ82Q8QNZH8PBK","writing-plans","Write implementation plans: bite-sized tasks, paths, code.","https:\u002F\u002Fgithub.com\u002FLyrikNilpferd71\u002FHermes-Backup\u002Fblob\u002Fmain\u002Fsnapshots\u002F2026-11-08T13-00-51+0100\u002Fskills\u002Fsoftware-development\u002Fwriting-plans\u002FSKILL.md",[],[],"2026-11-12 13:55:43",{"skillNo":104,"name":105,"nameZh":6,"description":106,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":107,"categories":108,"keywords":109,"publishedAt":102},"01M76BB8YS41CA3NKW0YMSSB88","systematic-debugging","4-phase root cause debugging: understand bugs before fixing.","https:\u002F\u002Fgithub.com\u002FLyrikNilpferd71\u002FHermes-Backup\u002Fblob\u002Fmain\u002Fsnapshots\u002F2026-11-08T13-00-51+0100\u002Fskills\u002Fsoftware-development\u002Fsystematic-debugging\u002FSKILL.md",[],[],{"skillNo":111,"name":112,"nameZh":6,"description":113,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":114,"categories":115,"keywords":116,"publishedAt":102},"01M76BB680FPNT998E1GM2QW10","subagent-driven-development","Execute plans via delegate_task subagents (2-stage review).","https:\u002F\u002Fgithub.com\u002FLyrikNilpferd71\u002FHermes-Backup\u002Fblob\u002Fmain\u002Fsnapshots\u002F2026-11-08T13-00-51+0100\u002Fskills\u002Fsoftware-development\u002Fsubagent-driven-development\u002FSKILL.md",[],[],{"skillNo":118,"name":119,"nameZh":6,"description":120,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":121,"categories":122,"keywords":123,"publishedAt":102},"01M76BB3C4ZB9S4RWPMM8MX0YB","plan","Write a markdown plan to .hermes\u002Fplans\u002F; no execution.","https:\u002F\u002Fgithub.com\u002FLyrikNilpferd71\u002FHermes-Backup\u002Fblob\u002Fmain\u002Fsnapshots\u002F2026-11-08T13-00-51+0100\u002Fskills\u002Fsoftware-development\u002Fplan\u002FSKILL.md",[],[],{"skillNo":125,"name":126,"nameZh":6,"description":127,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":128,"categories":129,"keywords":130,"publishedAt":102},"01M76BB0NK28KMP51YDB5R3EXA","hermes-agent-skill-authoring","Author in-repo SKILL.md files: frontmatter and structure.","https:\u002F\u002Fgithub.com\u002FLyrikNilpferd71\u002FHermes-Backup\u002Fblob\u002Fmain\u002Fsnapshots\u002F2026-11-08T13-00-51+0100\u002Fskills\u002Fsoftware-development\u002Fhermes-agent-skill-authoring\u002FSKILL.md",[],[],{"skillNo":132,"name":133,"nameZh":6,"description":134,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":135,"categories":136,"keywords":137,"publishedAt":102},"01M76BAYV58H0CKNSTZE3ZH4NY","debugging-hermes-tui-commands","Debug Hermes TUI slash commands: Python, gateway, Ink UI.","https:\u002F\u002Fgithub.com\u002FLyrikNilpferd71\u002FHermes-Backup\u002Fblob\u002Fmain\u002Fsnapshots\u002F2026-11-08T13-00-51+0100\u002Fskills\u002Fsoftware-development\u002Fdebugging-hermes-tui-commands\u002FSKILL.md",[],[]]