support qwen3.6 grpo & in-place add lora

cookbook/rl/short_math_grpo_moe.py

@@ -0,0 +1,275 @@
+"""GRPO training script for GSM8K dataset.
+
+Converted from the Tinker client version to Ray-based training.
+Uses short reasoning format: shorter thinking gets higher format reward.
+Answer extracted from \\boxed{} or #### format.
+"""
+import os
+import re
+from typing import List, Tuple, Dict, Any
+
+from peft import LoraConfig
+
+import twinkle
+from twinkle import DeviceMesh, DeviceGroup, get_device_placement, get_logger
+from twinkle.advantage import GRPOAdvantage
+from twinkle.checkpoint_engine import CheckpointEngineManager
+from twinkle.data_format import SamplingParams
+from twinkle.dataloader import DataLoader
+from twinkle.dataset import Dataset, DatasetMeta
+from twinkle.metric import CompletionRewardMetric
+from twinkle.model import TransformersModel
+from twinkle.processor import InputProcessor
+from twinkle.reward import GSM8KAccuracyReward
+from twinkle.reward.base import Reward
+from twinkle.sampler import vLLMSampler
+from twinkle.preprocessor.llm import GSM8KProcessor
+
+logger = get_logger()
+
+# ========== Configuration ==========
+MODEL_ID = os.environ.get('MODEL_ID', 'ms://Qwen/Qwen3.6-35B-A3B')
+USE_MEGATRON = bool(int(os.environ.get('USE_MEGATRON', '1')))
+
+MODEL_GPUS = int(os.environ.get('MODEL_GPUS', 4))
+MODEL_EP = int(os.environ.get('MODEL_EP', 2))
+MODEL_TP = int(os.environ.get('MODEL_TP', 2))
+MODEL_PP = int(os.environ.get('MODEL_PP', 2))
+
+SAMPLER_GPUS = int(os.environ.get('SAMPLER_GPUS', 2))
+SAMPLER_TP = int(os.environ.get('SAMPLER_TP', 2))
+NUM_GPUS = MODEL_GPUS + SAMPLER_GPUS
+
+NUM_GENERATIONS = int(os.environ.get('NUM_GENERATIONS', 8))
+MAX_NEW_TOKENS = int(os.environ.get('MAX_NEW_TOKENS', 4096))
+LEARNING_RATE = float(os.environ.get('LR', 5e-5))
+MAX_STEPS = int(os.environ.get('MAX_STEPS', 1000))
+BATCH_SIZE = int(os.environ.get('BATCH_SIZE', 4))
+MINI_BATCH_SIZE = int(os.environ.get('MINI_BATCH_SIZE', 4))
+MICRO_BATCH_SIZE = int(os.environ.get('MICRO_BATCH_SIZE', 1))
+GRADIENT_ACCUMULATION_STEPS = int(os.environ.get('GRADIENT_ACCUMULATION_STEPS', 1))
+ADAPTER_NAME = 'default'
+SAVE_STEPS = int(os.environ.get('SAVE_STEPS', 1000))
+LORA_RANK = int(os.environ.get('LORA_RANK', 16))
+
+SYSTEM_PROMPT = ('You are a helpful math assistant. Solve the problem with minimal but correct reasoning '
+                 'and put your final answer within \\boxed{}.')
+
+# ========== Reward Functions ==========
+class GSM8KBrevityReward(Reward):
+    """Brevity reward: rewards shorter completions that contain a valid answer.
+
+    Returns 0.0 if no valid answer format (\\boxed{} or ####).
+    Otherwise returns higher score for shorter completions (1.0 at <=200 chars).
+    """
+
+    def __call__(self, trajectories: List[Dict[str, Any]], **kwargs) -> List[float]:
+        rewards = []
+        for traj in trajectories:
+            messages = traj.get('messages', [])
+            completion = ''
+            for msg in reversed(messages):
+                if msg.get('role') == 'assistant':
+                    completion = msg.get('content', '')
+                    break
+
+            has_answer = bool(
+                re.search(r'\\boxed\{[^}]+\}', completion)
+                or re.search(r'####\s*[\-\d,\.]+', completion)
+            )
+
+            if not has_answer:
+                rewards.append(0.0)
+            else:
+                length = len(completion)
+                if length <= 200:
+                    rewards.append(1.0)
+                else:
+                    rewards.append(max(0.0, 1.0 - (length - 200) / 3000))
+        return rewards
+
+
+# ========== Dataset ==========
+def create_gsm8k_dataset():
+    dataset = Dataset(DatasetMeta('ms://modelscope/gsm8k', subset_name='main', split='train'))
+    dataset.set_template('Qwen3_5Template', model_id=MODEL_ID, max_length=4096, truncation_strategy='delete', enable_thinking=False)
+    dataset.map(GSM8KProcessor(system=SYSTEM_PROMPT))
+    dataset.encode(add_generation_prompt=True)
+    return dataset
+
+
+def compute_rewards(
+    trajectories: List[Dict[str, Any]],
+) -> Tuple[List[float], List[float], List[float]]:
+    accuracy_reward_fn = GSM8KAccuracyReward()
+    brevity_reward_fn = GSM8KBrevityReward()
+
+    accuracy_rewards = accuracy_reward_fn(trajectories)
+    brevity_rewards = brevity_reward_fn(trajectories)
+    total_rewards = [a + b for a, b in zip(accuracy_rewards, brevity_rewards)]
+    return total_rewards, brevity_rewards, accuracy_rewards
+
+
+# ========== Main ==========
+def main():
+    device_groups = [
+        DeviceGroup(name='model', ranks=list(range(MODEL_GPUS)), device_type='GPU'),
+        DeviceGroup(name='sampler', ranks=list(range(MODEL_GPUS, NUM_GPUS)), device_type='GPU', gpus_per_worker=SAMPLER_TP),
+    ]
+    dp_size = MODEL_GPUS //  (MODEL_TP * MODEL_PP)
+    model_mesh = DeviceMesh.from_sizes(world_size=MODEL_GPUS, dp_size=dp_size, tp_size=MODEL_TP, pp_size=MODEL_PP, ep_size=MODEL_EP, sequence_parallel=True)
+    sampler_dp_size = SAMPLER_GPUS //  (SAMPLER_TP)
+    sampler_mesh = DeviceMesh.from_sizes(world_size=SAMPLER_GPUS, dp_size=sampler_dp_size, tp_size=SAMPLER_TP)
+    twinkle.initialize(mode='ray', nproc_per_node=NUM_GPUS, groups=device_groups, lazy_collect=False)
+
+    lora_config = LoraConfig(
+        target_modules=['all-linear'],
+        r=LORA_RANK,
+        lora_alpha=LORA_RANK * 2,
+        lora_dropout=0.05,
+    )
+
+    if USE_MEGATRON:
+        from twinkle.model.megatron import MegatronModel
+        model = MegatronModel(
+            model_id=MODEL_ID,
+            device_mesh=model_mesh,
+            remote_group='model',
+            mixed_precision='bf16',
+        )
+    else:
+        model = TransformersModel(
+            model_id=MODEL_ID,
+            device_mesh=model_mesh,
+            remote_group='model',
+        )
+
+    model.add_adapter_to_model(ADAPTER_NAME, lora_config, gradient_accumulation_steps=GRADIENT_ACCUMULATION_STEPS)
+    if USE_MEGATRON:
+        model.set_optimizer('default', lr=LEARNING_RATE)
+        model.set_lr_scheduler('default', lr_decay_steps=MAX_STEPS, max_lr=LEARNING_RATE)
+    else:
+        model.set_optimizer('AdamW', lr=LEARNING_RATE)
+        model.set_lr_scheduler('CosineAnnealingLR', T_max=MAX_STEPS, eta_min=0)
+
+    model.set_loss('GRPOLoss', epsilon=0.2)
+    model.set_processor(InputProcessor)
+    model.set_template('Qwen3_5Template', model_id=MODEL_ID, enable_thinking=False)
+
+    sampler = vLLMSampler(
+        model_id=MODEL_ID,
+        engine_args={
+            'tensor_parallel_size': SAMPLER_TP,
+            'gpu_memory_utilization': 0.7,
+            'max_model_len': 10000,
+            'max_lora_rank': LORA_RANK, # save as lora_config
+            # NOTE: To use enable_lora with qwen3.5, ensure vLLM includes PR https://github.com/vllm-project/vllm/pull/36976
+            # enable_lora=True used with ckpt_manager.sync_weights(merge_and_sync=False)
+            # meaning only sync lora weights, if merge_and_sync=True,
+            # lora will be merged into the base model and sync all weights to vLLM
+            'enable_lora': True,
+            'enable_tower_connector_lora': True,
+        },
+        device_mesh=sampler_mesh,
+        remote_group='sampler',
+    )
+    sampler.set_template('Qwen3_5Template', model_id=MODEL_ID, enable_thinking=False)
+
+    ckpt_manager = CheckpointEngineManager(model=model, sampler=sampler)
+
+    GLOBAL_BATCH_SIZE = BATCH_SIZE * GRADIENT_ACCUMULATION_STEPS
+    dataloader = DataLoader(
+        dataset=create_gsm8k_dataset,
+        batch_size=GLOBAL_BATCH_SIZE,
+        min_batch_size=GLOBAL_BATCH_SIZE,
+        device_mesh=model_mesh,
+        remote_group='model',
+    )
+
+    advantage_fn = GRPOAdvantage()
+    metrics = CompletionRewardMetric()
+    sampling_params = SamplingParams(max_tokens=MAX_NEW_TOKENS, num_samples=1, logprobs=1, temperature=1.0, top_p=0.95)
+
+    optim_step = 0
+    logger.info('Starting GSM8K GRPO training (short reasoning)')
+    logger.info(get_device_placement())
+
+    for batch in dataloader:
+        if optim_step >= MAX_STEPS:
+            break
+
+        metrics.reset()
+        expand_prompts = []
+        for prompt in batch:
+            expand_prompts.extend([prompt] * NUM_GENERATIONS)
+
+        # enable_lora=True used with ckpt_manager.sync_weights(merge_and_sync=False)
+        # meaning only sync lora weights, if merge_and_sync=True,
+        # lora will be merged into the base model and sync all weights to vLLM
+        ckpt_manager.sync_weights(merge_and_sync=False)
+        sampler.reset_prefix_cache()
+
+        sample_responses = sampler.sample(
+            expand_prompts,
+            sampling_params,
+        )
+        if sample_responses and sample_responses[0].sequences:
+            first_decoded = sample_responses[0].sequences[0].decoded
+            if isinstance(first_decoded, str):
+                logger.info('[sample_debug] first_generation=%r', first_decoded[:512])
+
+        all_input_data: List[Dict[str, Any]] = []
+        all_old_logps: List[List[float]] = []
+        all_completion_lengths: List[int] = []
+
+        for sample_response in sample_responses:
+            for sequence in sample_response.sequences:
+                all_input_data.append(sequence.new_input_feature)
+                all_old_logps.append([logprob[0][1] for logprob in sequence.logprobs])
+                all_completion_lengths.append(len(sequence.tokens))
+
+        total_rewards, brevity_rewards, accuracy_rewards = compute_rewards(all_input_data)
+
+        metrics.accumulate(
+            completion_lengths=all_completion_lengths,
+            rewards={
+                'total': total_rewards,
+                'brevity': brevity_rewards,
+                'accuracy': accuracy_rewards,
+            },
+        )
+
+        advantages = advantage_fn(total_rewards, num_generations=NUM_GENERATIONS, scale='group').tolist()
+
+        total_completions = len(all_input_data)
+        for mb_start in range(0, total_completions, MINI_BATCH_SIZE):
+            mb_end = min(mb_start + MINI_BATCH_SIZE, total_completions)
+            mb_inputs = all_input_data[mb_start:mb_end]
+            mb_old_logps = all_old_logps[mb_start:mb_end]
+            mb_advantages = advantages[mb_start:mb_end]
+
+            model.forward_backward(
+                inputs=mb_inputs,
+                old_logps=mb_old_logps,
+                advantages=mb_advantages,
+                micro_batch_size=MICRO_BATCH_SIZE,
+            )
+            model.clip_grad_and_step()
+            optim_step += 1
+
+            if optim_step >= MAX_STEPS:
+                break
+            if optim_step % SAVE_STEPS == 0:
+                model.save(f'math-grpo-checkpoint-{optim_step}')
+
+        log_dict = metrics.calculate()
+        log_dict.update(model.calculate_metric(is_training=True))
+        metrics.reset()
+        logger.info(f'[Step {optim_step}/{MAX_STEPS}] {log_dict}')
+
+    logger.info(f'Training completed. optim_steps={optim_step}')
+    model.save('math-grpo-final')
+
+
+if __name__ == '__main__':
+    main()

cookbook/rl/short_math_grpo_multi_lora.py

@@ -35,19 +35,21 @@
 logger = get_logger()
 
 # ========== Configuration ==========
-MODEL_ID = os.environ.get('MODEL_ID', 'ms://Qwen/Qwen3.5-27B')
+MODEL_ID = os.environ.get('MODEL_ID', 'ms://Qwen/Qwen3.6-35B-A3B')
 
 MODEL_GPUS = int(os.environ.get('MODEL_GPUS', 4))
-SAMPLER_GPUS = int(os.environ.get('SAMPLER_GPUS', 4))
+SAMPLER_GPUS = int(os.environ.get('SAMPLER_GPUS', 2))
+SAMPLER_TP = int(os.environ.get('SAMPLER_TP', 2))
+
 NUM_GPUS = MODEL_GPUS + SAMPLER_GPUS
 
 NUM_GENERATIONS = int(os.environ.get('NUM_GENERATIONS', 8))
 MAX_NEW_TOKENS = int(os.environ.get('MAX_NEW_TOKENS', 4096))
-LEARNING_RATE = float(os.environ.get('LR', 5e-6))
+LEARNING_RATE = float(os.environ.get('LR', 5e-5))
 MAX_STEPS = int(os.environ.get('MAX_STEPS', 1000))
 BATCH_SIZE = int(os.environ.get('BATCH_SIZE', 4))
 MINI_BATCH_SIZE = int(os.environ.get('MINI_BATCH_SIZE', 4))
-MICRO_BATCH_SIZE = int(os.environ.get('MICRO_BATCH_SIZE', 2))
+MICRO_BATCH_SIZE = int(os.environ.get('MICRO_BATCH_SIZE', 1))
 GRADIENT_ACCUMULATION_STEPS = int(os.environ.get('GRADIENT_ACCUMULATION_STEPS', 1))
 ADAPTER_NAME = 'default_0'
 SAVE_STEPS = int(os.environ.get('SAVE_STEPS', 1000))
@@ -124,23 +126,19 @@ def main():
     device_groups = [
         DeviceGroup(name='model', ranks=list(range(MODEL_GPUS)), device_type='GPU'),
         DeviceGroup(name='sampler', ranks=list(range(MODEL_GPUS, NUM_GPUS)), device_type='GPU',
-                    gpus_per_worker=2),
+                    gpus_per_worker=SAMPLER_TP),
     ]
 
     # Model mesh: tp=2, ep=2, pp=2, sequence_parallel (ref: server_config.yaml)
     model_mesh = DeviceMesh.from_sizes(
         world_size=MODEL_GPUS,
         tp_size=2,
-        # ep_size=2,
+        ep_size=2,
         pp_size=2,
-        # sequence_parallel=True,
-    )
-    # Sampler mesh: dp=2, tp=2
-    sampler_mesh = DeviceMesh.from_sizes(
-        world_size=SAMPLER_GPUS,
-        dp_size=2,
-        tp_size=2,
+        sequence_parallel=True,
     )
+    sampler_dp_size = SAMPLER_GPUS //  (SAMPLER_TP)
+    sampler_mesh = DeviceMesh.from_sizes(world_size=SAMPLER_GPUS, dp_size=sampler_dp_size, tp_size=SAMPLER_TP)
 
     twinkle.initialize(mode='ray', nproc_per_node=NUM_GPUS, groups=device_groups, lazy_collect=False)
 
@@ -179,9 +177,10 @@ def main():
     sampler = vLLMSampler(
         model_id=MODEL_ID,
         engine_args={
+            'tensor_parallel_size': SAMPLER_TP,
             'gpu_memory_utilization': 0.8,
             'max_model_len': 8192,
-            'max_lora_rank': 32,
+            'max_lora_rank': LORA_RANK,
             # NOTE: To use enable_lora with qwen3.5, ensure vLLM includes PR https://github.com/vllm-project/vllm/pull/36976
             'enable_lora': True,
             'enable_tower_connector_lora': True,
@@ -241,6 +240,10 @@ def main():
             sampling_params,
             adapter_path=lora_sync_path,
         )
+        if sample_responses and sample_responses[0].sequences:
+            first_decoded = sample_responses[0].sequences[0].decoded
+            if isinstance(first_decoded, str):
+                logger.info('[sample_debug] first_generation=%r', first_decoded[:512])
 
         all_input_data: List[Dict[str, Any]] = []
         all_old_logps: List[List[float]] = []

src/twinkle/checkpoint_engine/manager.py

@@ -161,4 +161,5 @@ def _expand_keys(keys):
 
         if not self.base_sync_done:
             self.base_sync_done = True
-            logger.info('Base model sync completed, subsequent syncs will be LoRA-only')
+            if not merge_and_sync:
+                logger.info('Base model sync completed, subsequent syncs will be LoRA-only')

src/twinkle/model/megatron/megatron.py

@@ -1444,6 +1444,7 @@ def _print_weight_example(names):
                 logger.info(f'Sync weight: {name}')
 
         def _add_base_layer_suffix(name):
+            base_layer_name = None
             if name.endswith('.weight'):
                 base_layer_name = f'{name[:-7]}.base_layer.weight'
                 if not model_keys or base_layer_name in model_keys:
@@ -1452,7 +1453,7 @@ def _add_base_layer_suffix(name):
                 base_layer_name = f'{name[:-5]}.base_layer.bias'
                 if not model_keys or base_layer_name in model_keys:
                     name = base_layer_name
-            if 'experts' in name:
+            if 'experts' in name and base_layer_name is not None:
                 return base_layer_name
             return name
 

src/twinkle/patch/vllm_lora_weights.py

@@ -29,6 +29,10 @@ def embeddings(self):
 class VLLMLoraWeights(Patch):
 
     def __call__(self, sampler, **kwargs):
+        from twinkle.patch.vllm_moe_loader import patch_qwen35_moe_is_3d_moe_weight_false
+
+        patch_qwen35_moe_is_3d_moe_weight_false()
+
         _sampler_ref = sampler
 
         def _get_tokenizer():