feat: add minimum batch allocation support for environments

- Add min_batch_allocation parameter to ensure environments contribute minimum proportion to each batch
- Implement grab_batch_with_minimum_allocations function with proper scaling when allocations exceed 100%
- Add mixed-size group buffering to handle variable-sized data submissions
- Update server to use minimum allocation logic when any env has min_batch_allocation set
- Add comprehensive tests for minimum allocation scenarios
- Update documentation in API README and CONFIG.md
- Update example environments to demonstrate the feature

This feature allows critical environments to guarantee they contribute at least a specified proportion (0.0-1.0) to each training batch, ensuring important data sources are always represented during training.

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <noreply@anthropic.com>

environments/intern_bootcamp/intern_bootcamp_env.py

@@ -367,12 +367,12 @@ class InternBootcampEnv(BaseEnv):
             tokenizer_name="NousResearch/DeepHermes-3-Llama-3-8B-Preview",
             group_size=8,
             use_wandb=True,
-            max_num_workers=64,
+            max_num_workers_per_node=16,
             rollout_server_url="http://localhost:8000",
             total_steps=10000,
             batch_size=1024,
             steps_per_eval=100,
-            max_token_length=16384,
+            max_token_length=8192,
             inference_weight=1.0,
             wandb_name="intern_bootcamp_random_tasks",
             data_path_to_save_groups="data/intern_bootcamp_random_tasks.jsonl",
@@ -385,6 +385,7 @@ class InternBootcampEnv(BaseEnv):
             format_bonus=0.2,
             # Training parameters
             require_reasoning=True,
+            min_batch_allocation=0.1,
             min_reasoning_length=50,
             temperature=0.7,
             top_p=0.9,