synth-ai 0.2.8.dev4__py3-none-any.whl → 0.2.23.dev3__py3-none-any.whl
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- examples/README.md +1 -0
- examples/__init__.py +16 -0
- examples/analyze_semantic_words.sh +17 -0
- examples/baseline/banking77_baseline.py +243 -0
- examples/baseline/banking77_pipeline_baseline.py +294 -0
- examples/baseline/crafter_baseline.py +407 -0
- examples/baseline/pokemon_red_baseline.py +326 -0
- examples/baseline/simple_baseline.py +56 -0
- examples/baseline/warming_up_to_rl_baseline.py +239 -0
- examples/blog_posts/gepa/README.md +355 -0
- examples/blog_posts/gepa/configs/banking77_gepa_local.toml +95 -0
- examples/blog_posts/gepa/configs/banking77_gepa_test.toml +80 -0
- examples/blog_posts/gepa/configs/banking77_mipro_local.toml +50 -0
- examples/blog_posts/gepa/configs/banking77_pipeline_gepa_local.toml +101 -0
- examples/blog_posts/gepa/configs/banking77_pipeline_gepa_test.toml +96 -0
- examples/blog_posts/gepa/configs/hotpotqa_gepa_local.toml +57 -0
- examples/blog_posts/gepa/configs/hotpotqa_gepa_qwen.toml +35 -0
- examples/blog_posts/gepa/configs/hotpotqa_mipro_local.toml +51 -0
- examples/blog_posts/gepa/configs/hover_gepa_local.toml +57 -0
- examples/blog_posts/gepa/configs/hover_gepa_qwen.toml +35 -0
- examples/blog_posts/gepa/configs/hover_mipro_local.toml +51 -0
- examples/blog_posts/gepa/configs/ifbench_gepa_local.toml +57 -0
- examples/blog_posts/gepa/configs/ifbench_gepa_qwen.toml +35 -0
- examples/blog_posts/gepa/configs/ifbench_mipro_local.toml +51 -0
- examples/blog_posts/gepa/configs/pupa_gepa_local.toml +58 -0
- examples/blog_posts/gepa/configs/pupa_mipro_local.toml +52 -0
- examples/blog_posts/gepa/deploy_banking77_task_app.sh +54 -0
- examples/blog_posts/gepa/gepa_baseline.py +204 -0
- examples/blog_posts/gepa/query_prompts_example.py +97 -0
- examples/blog_posts/gepa/run_gepa_banking77.sh +112 -0
- examples/blog_posts/gepa/run_gepa_banking77_pipeline.sh +163 -0
- examples/blog_posts/gepa/task_apps.py +105 -0
- examples/blog_posts/gepa/test_gepa_local.sh +67 -0
- examples/blog_posts/gepa/verify_banking77_setup.sh +123 -0
- examples/blog_posts/mipro/README.md +415 -0
- examples/blog_posts/mipro/configs/banking77_mipro_local.toml +91 -0
- examples/blog_posts/mipro/configs/banking77_mipro_test.toml +87 -0
- examples/blog_posts/mipro/configs/banking77_pipeline_mipro_gemini_flash_lite_local.toml +98 -0
- examples/blog_posts/mipro/configs/banking77_pipeline_mipro_gpt41mini_local.toml +96 -0
- examples/blog_posts/mipro/configs/banking77_pipeline_mipro_local.toml +94 -0
- examples/blog_posts/mipro/configs/banking77_pipeline_mipro_test.toml +170 -0
- examples/blog_posts/mipro/deploy_banking77_pipeline_task_app.sh +59 -0
- examples/blog_posts/mipro/deploy_banking77_task_app.sh +41 -0
- examples/blog_posts/mipro/multi_step.md +79 -0
- examples/blog_posts/mipro/run_mipro_banking77.sh +191 -0
- examples/blog_posts/mipro/run_mipro_banking77_pipeline.sh +171 -0
- examples/blog_posts/mipro/run_mipro_banking77_pipeline_gemini_flash_lite.sh +177 -0
- examples/blog_posts/mipro/run_mipro_banking77_pipeline_gpt41mini.sh +173 -0
- examples/blog_posts/mipro/verify_banking77_setup.sh +117 -0
- examples/blog_posts/pokemon_vl/README.md +98 -0
- examples/blog_posts/pokemon_vl/configs/eval_gpt5nano.toml +26 -0
- examples/blog_posts/pokemon_vl/configs/eval_qwen3_vl.toml +27 -0
- examples/blog_posts/pokemon_vl/configs/eval_rl_final.toml +24 -0
- examples/blog_posts/pokemon_vl/configs/filter_high_reward.toml +10 -0
- examples/blog_posts/pokemon_vl/configs/train_rl_from_sft.toml +43 -0
- examples/blog_posts/pokemon_vl/configs/train_sft_qwen4b_vl.toml +40 -0
- examples/blog_posts/pokemon_vl/extract_images.py +239 -0
- examples/blog_posts/pokemon_vl/pokemon_vl_baseline.py +326 -0
- examples/blog_posts/pokemon_vl/run_eval_extract_images.py +209 -0
- examples/blog_posts/pokemon_vl/run_qwen_eval_extract_images.py +212 -0
- examples/blog_posts/pokemon_vl/text_box_analysis.md +106 -0
- examples/blog_posts/warming_up_to_rl/ARCHITECTURE.md +195 -0
- examples/blog_posts/warming_up_to_rl/FINAL_TEST_RESULTS.md +127 -0
- examples/blog_posts/warming_up_to_rl/INFERENCE_SUCCESS.md +132 -0
- examples/blog_posts/warming_up_to_rl/README.md +158 -0
- examples/blog_posts/warming_up_to_rl/SMOKE_TESTING.md +164 -0
- examples/blog_posts/warming_up_to_rl/SMOKE_TEST_COMPLETE.md +253 -0
- examples/blog_posts/warming_up_to_rl/configs/eval_baseline_qwen32b_10x20.toml +25 -0
- examples/blog_posts/warming_up_to_rl/configs/eval_ft_qwen4b.toml +25 -0
- examples/blog_posts/warming_up_to_rl/configs/eval_ft_qwen4b_10x20.toml +26 -0
- examples/blog_posts/warming_up_to_rl/configs/eval_groq_qwen32b.toml +25 -0
- examples/blog_posts/warming_up_to_rl/configs/eval_openai_gpt_oss_120b.toml +29 -0
- examples/blog_posts/warming_up_to_rl/configs/filter_high_reward_dataset.toml +10 -0
- examples/blog_posts/warming_up_to_rl/configs/smoke_test.toml +75 -0
- examples/blog_posts/warming_up_to_rl/configs/train_rl_from_sft.toml +91 -0
- examples/blog_posts/warming_up_to_rl/configs/train_sft_qwen4b.toml +40 -0
- examples/blog_posts/warming_up_to_rl/warming_up_to_rl_baseline.py +187 -0
- examples/crafter_debug_render.py +186 -0
- examples/dev/qwen3_32b_qlora_4xh100.toml +45 -0
- examples/gepa/banking77_pipeline_gepa.toml +96 -0
- examples/gepa/multi_stage_gepa_example.toml +84 -0
- examples/gepa/run_gepa_banking77_pipeline.sh +157 -0
- examples/multi_step/SFT_README.md +147 -0
- examples/multi_step/configs/README_verilog_rl.md +77 -0
- examples/multi_step/configs/VERILOG_REWARDS.md +103 -0
- examples/multi_step/configs/VERILOG_RL_CHECKLIST.md +196 -0
- examples/multi_step/configs/crafter_eval_synth_qwen4b.toml +35 -0
- examples/multi_step/configs/crafter_eval_text_only_groq_qwen32b.toml +36 -0
- examples/multi_step/configs/crafter_rl_outcome.toml +75 -0
- examples/multi_step/configs/crafter_rl_stepwise_hosted_judge.toml +145 -0
- examples/multi_step/configs/crafter_rl_stepwise_shaped.toml +84 -0
- examples/multi_step/configs/crafter_rl_stepwise_simple.toml +79 -0
- examples/multi_step/configs/crafter_rl_stepwise_simple_NEW_FORMAT.toml +105 -0
- examples/multi_step/configs/crafter_sft_qwen30b_lora.toml +62 -0
- examples/multi_step/configs/crafter_synth_backend.md +40 -0
- examples/multi_step/configs/verilog_eval_groq_qwen32b.toml +31 -0
- examples/multi_step/configs/verilog_eval_synth_qwen8b.toml +33 -0
- examples/multi_step/configs/verilog_rl_lora.toml +147 -0
- examples/multi_step/convert_traces_to_sft.py +84 -0
- examples/multi_step/crafter_rl_lora.md +70 -0
- examples/multi_step/judges/crafter_backend_judge.py +220 -0
- examples/multi_step/judges/verilog_backend_judge.py +234 -0
- examples/multi_step/readme.md +48 -0
- examples/multi_step/run_sft_qwen30b.sh +45 -0
- examples/multi_step/sse_metrics_streaming_notes.md +357 -0
- examples/multi_step/task_app_config_notes.md +494 -0
- examples/multi_step/verilog_rl_lora.md +218 -0
- examples/qwen_coder/README.md +102 -0
- examples/qwen_coder/_shared.py +113 -0
- examples/qwen_coder/configs/coder_lora_30b.toml +60 -0
- examples/qwen_coder/configs/coder_lora_4b.toml +61 -0
- examples/qwen_coder/configs/coder_lora_small.toml +57 -0
- examples/qwen_coder/generate_dataset.py +98 -0
- examples/qwen_coder/infer_ft_smoke.py +65 -0
- examples/qwen_coder/infer_prod_proxy.py +73 -0
- examples/qwen_coder/infer_via_synth.py +87 -0
- examples/qwen_coder/scripts/infer_coder.sh +19 -0
- examples/qwen_coder/scripts/train_coder_30b.sh +22 -0
- examples/qwen_coder/sft_full_17b.py +103 -0
- examples/qwen_coder/sft_lora_30b.py +110 -0
- examples/qwen_coder/subset_jsonl.py +39 -0
- examples/qwen_coder/todos.md +38 -0
- examples/qwen_coder/validate_jsonl.py +60 -0
- examples/qwen_vl/BUGS_AND_FIXES.md +232 -0
- examples/qwen_vl/IMAGE_VALIDATION_COMPLETE.md +271 -0
- examples/qwen_vl/IMAGE_VALIDATION_SUMMARY.md +260 -0
- examples/qwen_vl/INFERENCE_SFT_TESTS.md +412 -0
- examples/qwen_vl/NEXT_STEPS_2B.md +325 -0
- examples/qwen_vl/QUICKSTART.md +327 -0
- examples/qwen_vl/QUICKSTART_RL_VISION.md +110 -0
- examples/qwen_vl/README.md +152 -0
- examples/qwen_vl/RL_VISION_COMPLETE.md +475 -0
- examples/qwen_vl/RL_VISION_TESTING.md +333 -0
- examples/qwen_vl/SDK_VISION_INTEGRATION.md +328 -0
- examples/qwen_vl/SETUP_COMPLETE.md +274 -0
- examples/qwen_vl/VISION_TESTS_COMPLETE.md +489 -0
- examples/qwen_vl/VLM_PIPELINE_COMPLETE.md +242 -0
- examples/qwen_vl/__init__.py +2 -0
- examples/qwen_vl/collect_data_via_cli.md +415 -0
- examples/qwen_vl/collect_vision_traces.py +368 -0
- examples/qwen_vl/configs/crafter_rl_vision_qwen3vl4b.toml +110 -0
- examples/qwen_vl/configs/crafter_vlm_sft_example.toml +59 -0
- examples/qwen_vl/configs/eval_gpt4o_mini_vision.toml +26 -0
- examples/qwen_vl/configs/eval_gpt4o_vision_proper.toml +29 -0
- examples/qwen_vl/configs/eval_gpt5nano_vision.toml +26 -0
- examples/qwen_vl/configs/eval_qwen3vl_vision.toml +26 -0
- examples/qwen_vl/configs/filter_qwen3vl_sft.toml +49 -0
- examples/qwen_vl/configs/filter_vision_sft.toml +52 -0
- examples/qwen_vl/configs/filter_vision_test.toml +8 -0
- examples/qwen_vl/configs/sft_qwen3_vl_2b_test.toml +54 -0
- examples/qwen_vl/crafter_gpt5nano_agent.py +308 -0
- examples/qwen_vl/crafter_qwen_vl_agent.py +300 -0
- examples/qwen_vl/run_vision_comparison.sh +61 -0
- examples/qwen_vl/run_vision_sft_pipeline.sh +175 -0
- examples/qwen_vl/test_image_validation.py +201 -0
- examples/qwen_vl/test_sft_vision_data.py +110 -0
- examples/rl/README.md +169 -0
- examples/rl/configs/eval_base_qwen.toml +17 -0
- examples/rl/configs/eval_rl_qwen.toml +13 -0
- examples/rl/configs/rl_from_base_qwen.toml +62 -0
- examples/rl/configs/rl_from_base_qwen17.toml +80 -0
- examples/rl/configs/rl_from_ft_qwen.toml +37 -0
- examples/rl/download_dataset.py +80 -0
- examples/rl/run_eval.py +436 -0
- examples/rl/run_rl_and_save.py +111 -0
- examples/rl/task_app/README.md +21 -0
- examples/rl/task_app/math_single_step.py +990 -0
- examples/rl/task_app/math_task_app.py +111 -0
- examples/run_crafter_demo.sh +10 -0
- examples/sdk_prompt_learning_example.py +55 -0
- examples/sft/README.md +139 -0
- examples/sft/configs/crafter_fft_qwen0p6b.toml +49 -0
- examples/sft/configs/crafter_lora_qwen0p6b.toml +49 -0
- examples/sft/evaluate.py +117 -0
- examples/sft/export_dataset.py +120 -0
- examples/sft/generate_traces.py +164 -0
- examples/swe/__init__.py +12 -0
- examples/swe/task_app/README.md +135 -0
- examples/swe/task_app/__init__.py +2 -0
- examples/swe/task_app/grpo_swe_mini.py +604 -0
- examples/swe/task_app/grpo_swe_mini_task_app.py +124 -0
- examples/swe/task_app/hosted/README.md +173 -0
- examples/swe/task_app/hosted/__init__.py +5 -0
- examples/swe/task_app/hosted/branching.py +143 -0
- examples/swe/task_app/hosted/environment_routes.py +1289 -0
- examples/swe/task_app/hosted/envs/__init__.py +1 -0
- examples/swe/task_app/hosted/envs/crafter/__init__.py +6 -0
- examples/swe/task_app/hosted/envs/crafter/app.py +1 -0
- examples/swe/task_app/hosted/envs/crafter/environment.py +522 -0
- examples/swe/task_app/hosted/envs/crafter/policy.py +478 -0
- examples/swe/task_app/hosted/envs/crafter/react_agent.py +108 -0
- examples/swe/task_app/hosted/envs/crafter/shared.py +305 -0
- examples/swe/task_app/hosted/envs/crafter/tools.py +47 -0
- examples/swe/task_app/hosted/envs/mini_swe/__init__.py +8 -0
- examples/swe/task_app/hosted/envs/mini_swe/environment.py +1191 -0
- examples/swe/task_app/hosted/envs/mini_swe/policy.py +355 -0
- examples/swe/task_app/hosted/envs/mini_swe/shared.py +83 -0
- examples/swe/task_app/hosted/envs/mini_swe/tools.py +96 -0
- examples/swe/task_app/hosted/hosted_app.py +204 -0
- examples/swe/task_app/hosted/inference/__init__.py +5 -0
- examples/swe/task_app/hosted/inference/openai_client.py +584 -0
- examples/swe/task_app/hosted/main.py +100 -0
- examples/swe/task_app/hosted/policy_routes.py +1094 -0
- examples/swe/task_app/hosted/registry.py +195 -0
- examples/swe/task_app/hosted/rollout.py +1905 -0
- examples/swe/task_app/hosted/storage/__init__.py +5 -0
- examples/swe/task_app/hosted/storage/volume.py +211 -0
- examples/swe/task_app/hosted/test_agents.py +161 -0
- examples/swe/task_app/hosted/test_service.py +136 -0
- examples/swe/task_app/hosted/utils.py +62 -0
- examples/swe/task_app/morph_backend.py +178 -0
- examples/task_apps/IMAGE_ONLY_EVAL_QUICKSTART.md +258 -0
- examples/task_apps/TESTING.md +275 -0
- examples/task_apps/banking77/__init__.py +6 -0
- examples/task_apps/banking77/banking77_task_app.py +912 -0
- examples/task_apps/banking77/deploy_wrapper.py +46 -0
- examples/task_apps/banking77_pipeline/__init__.py +6 -0
- examples/task_apps/banking77_pipeline/banking77_pipeline_task_app.py +489 -0
- examples/task_apps/banking77_pipeline/deploy_wrapper.py +50 -0
- examples/task_apps/crafter/CREATE_SFT_DATASET.md +286 -0
- examples/task_apps/crafter/EVAL_IMAGE_ONLY_RESULTS.md +152 -0
- examples/task_apps/crafter/FILTER_COMMAND_STATUS.md +187 -0
- examples/task_apps/crafter/FILTER_COMMAND_SUCCESS.md +281 -0
- examples/task_apps/crafter/QUERY_EXAMPLES.md +203 -0
- examples/task_apps/crafter/README_IMAGE_ONLY_EVAL.md +316 -0
- examples/task_apps/crafter/eval_image_only_gpt4o.toml +28 -0
- examples/task_apps/crafter/eval_text_only_groq_llama.toml +36 -0
- examples/task_apps/crafter/filter_sft_dataset.toml +16 -0
- examples/task_apps/crafter/task_app/README.md +42 -0
- examples/task_apps/crafter/task_app/__init__.py +5 -0
- examples/task_apps/crafter/task_app/grpo_crafter.py +1055 -0
- examples/task_apps/crafter/task_app/grpo_crafter_task_app.py +146 -0
- examples/task_apps/crafter/task_app/synth_envs_hosted/README.md +173 -0
- examples/task_apps/crafter/task_app/synth_envs_hosted/__init__.py +5 -0
- examples/task_apps/crafter/task_app/synth_envs_hosted/branching.py +143 -0
- examples/task_apps/crafter/task_app/synth_envs_hosted/environment_routes.py +1226 -0
- examples/task_apps/crafter/task_app/synth_envs_hosted/envs/__init__.py +1 -0
- examples/task_apps/crafter/task_app/synth_envs_hosted/envs/crafter/__init__.py +6 -0
- examples/task_apps/crafter/task_app/synth_envs_hosted/envs/crafter/app.py +1 -0
- examples/task_apps/crafter/task_app/synth_envs_hosted/envs/crafter/environment.py +532 -0
- examples/task_apps/crafter/task_app/synth_envs_hosted/envs/crafter/policy.py +583 -0
- examples/task_apps/crafter/task_app/synth_envs_hosted/envs/crafter/react_agent.py +122 -0
- examples/task_apps/crafter/task_app/synth_envs_hosted/envs/crafter/shared.py +305 -0
- examples/task_apps/crafter/task_app/synth_envs_hosted/envs/crafter/tools.py +47 -0
- examples/task_apps/crafter/task_app/synth_envs_hosted/hosted_app.py +253 -0
- examples/task_apps/crafter/task_app/synth_envs_hosted/inference/__init__.py +5 -0
- examples/task_apps/crafter/task_app/synth_envs_hosted/inference/openai_client.py +999 -0
- examples/task_apps/crafter/task_app/synth_envs_hosted/main.py +100 -0
- examples/task_apps/crafter/task_app/synth_envs_hosted/policy_routes.py +1252 -0
- examples/task_apps/crafter/task_app/synth_envs_hosted/registry.py +195 -0
- examples/task_apps/crafter/task_app/synth_envs_hosted/rollout.py +2233 -0
- examples/task_apps/crafter/task_app/synth_envs_hosted/storage/__init__.py +5 -0
- examples/task_apps/crafter/task_app/synth_envs_hosted/storage/volume.py +211 -0
- examples/task_apps/crafter/task_app/synth_envs_hosted/test_agents.py +161 -0
- examples/task_apps/crafter/task_app/synth_envs_hosted/test_service.py +136 -0
- examples/task_apps/crafter/task_app/synth_envs_hosted/utils.py +411 -0
- examples/task_apps/dev/pokemon_emerald/__init__.py +2 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/README.md +811 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/agent/__init__.py +120 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/agent/action.py +160 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/agent/memory.py +155 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/agent/perception.py +69 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/agent/planning.py +96 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/agent/simple.py +1502 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/agent/system_prompt.py +4 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/grab_map.py +68 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/manual.py +216 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/pokemon_env/__init__.py +35 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/pokemon_env/emerald_utils.py +631 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/pokemon_env/emulator.py +1544 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/pokemon_env/enums.py +1428 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/pokemon_env/memory_reader.py +4848 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/pokemon_env/types.py +41 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/pokemon_env/utils.py +298 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/pyproject.toml +95 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/run.py +204 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/server/app.py +2152 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/server/client.py +429 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/server/frame_server.py +155 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/tests/README.md +78 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/tests/run_tests.py +122 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/tests/test_agent_direct.py +76 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/tests/test_agent_prompts.py +413 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/tests/test_battle_state_formatting.py +204 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/tests/test_dialogue_detection.py +133 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/tests/test_dialogue_detection_comprehensive.py +229 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/tests/test_direct_agent_emulator.py +300 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/tests/test_fps_adjustment_pytest.py +205 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/tests/test_house_to_outside_direct.py +200 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/tests/test_house_to_outside_transition.py +284 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/tests/test_map_ground_truth_comparison.py +468 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/tests/test_memory_map.py +575 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/tests/test_server_map_validation.py +311 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/tests/test_torchic_state.py +259 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/utils/anticheat.py +372 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/utils/checkpoint.py +296 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/utils/error_handler.py +275 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/utils/get_local_ip.py +22 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/utils/helpers.py +44 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/utils/llm_logger.py +514 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/utils/map_formatter.py +415 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/utils/map_stitcher.py +1763 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/utils/map_stitcher_singleton.py +33 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/utils/map_trimmer.py +106 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/utils/map_visualizer.py +334 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/utils/ocr_dialogue.py +1020 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/utils/recording.py +188 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/utils/state_formatter.py +1481 -0
- examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/utils/vlm.py +862 -0
- examples/task_apps/dev/pokemon_emerald/modal_app.py +114 -0
- examples/task_apps/dev/pokemon_emerald/task_app/README.md +81 -0
- examples/task_apps/dev/pokemon_emerald/task_app/__init__.py +6 -0
- examples/task_apps/dev/pokemon_emerald/task_app/pokemon_emerald.py +685 -0
- examples/task_apps/enron/__init__.py +2 -0
- examples/task_apps/enron/eval_groq_qwen32.toml +16 -0
- examples/task_apps/enron/filter_sft.toml +5 -0
- examples/task_apps/enron/task_app/README.md +14 -0
- examples/task_apps/enron/task_app/__init__.py +1 -0
- examples/task_apps/enron/task_app/grpo_enron.py +906 -0
- examples/task_apps/enron/task_app/grpo_enron_task_app.py +146 -0
- examples/task_apps/enron/tests/__init__.py +4 -0
- examples/task_apps/enron/tests/conftest.py +115 -0
- examples/task_apps/enron/tests/integration/__init__.py +4 -0
- examples/task_apps/enron/tests/integration/test_enron_eval.py +179 -0
- examples/task_apps/enron/tests/integration/test_enron_rollout.py +135 -0
- examples/task_apps/enron/tests/unit/__init__.py +4 -0
- examples/task_apps/enron/tests/unit/test_enron_environment.py +126 -0
- examples/task_apps/gepa_benchmarks/__init__.py +7 -0
- examples/task_apps/gepa_benchmarks/common.py +260 -0
- examples/task_apps/gepa_benchmarks/hotpotqa_task_app.py +507 -0
- examples/task_apps/gepa_benchmarks/hover_task_app.py +436 -0
- examples/task_apps/gepa_benchmarks/ifbench_task_app.py +563 -0
- examples/task_apps/gepa_benchmarks/pupa_task_app.py +460 -0
- examples/task_apps/math/README.md +21 -0
- examples/task_apps/math/math_single_step.py +1000 -0
- examples/task_apps/math/math_task_app.py +115 -0
- examples/task_apps/pokemon_battle/__init__.py +2 -0
- examples/task_apps/pokemon_battle/modal_app.py +104 -0
- examples/task_apps/pokemon_battle/task_app/README.md +68 -0
- examples/task_apps/pokemon_battle/task_app/__init__.py +6 -0
- examples/task_apps/pokemon_battle/task_app/pokemon_showdown.py +932 -0
- examples/task_apps/pokemon_red/EVAL_IMAGE_ONLY_COMPLETE.md +283 -0
- examples/task_apps/pokemon_red/EVAL_IMAGE_ONLY_STATUS.md +155 -0
- examples/task_apps/pokemon_red/README.md +356 -0
- examples/task_apps/pokemon_red/README_IMAGE_ONLY_EVAL.md +428 -0
- examples/task_apps/pokemon_red/__init__.py +3 -0
- examples/task_apps/pokemon_red/eval_image_only_gpt4o.toml +30 -0
- examples/task_apps/pokemon_red/eval_pokemon_red_policy.py +224 -0
- examples/task_apps/pokemon_red/pallet_town_rl_config.toml +75 -0
- examples/task_apps/pokemon_red/task_app.py +1048 -0
- examples/task_apps/pokemon_red/test_pallet_town_rewards.py +193 -0
- examples/task_apps/sokoban/README.md +306 -0
- examples/task_apps/sokoban/__init__.py +3 -0
- examples/task_apps/sokoban/eval_groq_qwen32.toml +16 -0
- examples/task_apps/sokoban/eval_openai_gpt5.toml +16 -0
- examples/task_apps/sokoban/filter_sft.toml +5 -0
- examples/task_apps/sokoban/task_app.py +1058 -0
- examples/task_apps/sokoban/tests/__init__.py +4 -0
- examples/task_apps/sokoban/tests/conftest.py +113 -0
- examples/task_apps/sokoban/tests/integration/__init__.py +4 -0
- examples/task_apps/sokoban/tests/integration/test_sokoban_eval.py +57 -0
- examples/task_apps/sokoban/tests/integration/test_sokoban_rollout.py +198 -0
- examples/task_apps/sokoban/tests/unit/__init__.py +4 -0
- examples/task_apps/sokoban/tests/unit/test_sokoban_environment.py +114 -0
- examples/task_apps/verilog/__init__.py +1 -0
- examples/task_apps/verilog/eval_groq_qwen32b.toml +22 -0
- examples/task_apps/verilog/filter_sft.toml +5 -0
- examples/task_apps/verilog/task_app/README.md +12 -0
- examples/task_apps/verilog/task_app/__init__.py +1 -0
- examples/task_apps/verilog/task_app/grpo_verilog.py +1166 -0
- examples/task_apps/verilog/task_app/grpo_verilog_task_app.py +145 -0
- examples/task_apps/verilog/tests/__init__.py +4 -0
- examples/task_apps/verilog/tests/conftest.py +115 -0
- examples/task_apps/verilog/tests/integration/__init__.py +4 -0
- examples/task_apps/verilog/tests/integration/test_verilog_eval.py +181 -0
- examples/task_apps/verilog/tests/integration/test_verilog_rollout.py +55 -0
- examples/task_apps/verilog/tests/unit/__init__.py +4 -0
- examples/task_apps/verilog/tests/unit/test_verilog_scoring.py +118 -0
- examples/tunnel_gepa_banking77/README.md +106 -0
- examples/tunnel_gepa_banking77/banking77_gepa_tunnel.toml +95 -0
- examples/tunnel_gepa_banking77/keep_tunnel_running.py +60 -0
- examples/tunnel_gepa_banking77/run_gepa_with_tunnel.sh +226 -0
- examples/vlm/PROPOSAL.md +53 -0
- examples/vlm/README.md +68 -0
- examples/vlm/configs/crafter_vlm_gpt4o.toml +49 -0
- examples/vlm/crafter_image_only_agent.py +207 -0
- examples/vlm/crafter_openai_vlm_agent.py +275 -0
- examples/vlm/filter_image_rows.py +63 -0
- examples/vlm/run_crafter_vlm_benchmark.py +316 -0
- examples/warming_up_to_rl/_utils.py +92 -0
- examples/warming_up_to_rl/analyze_trace_db.py +422 -0
- examples/warming_up_to_rl/configs/crafter_fft.toml +53 -0
- examples/warming_up_to_rl/configs/crafter_fft_4b.toml +54 -0
- examples/warming_up_to_rl/configs/eval_fft_qwen4b.toml +22 -0
- examples/warming_up_to_rl/configs/eval_groq_qwen32b.toml +15 -0
- examples/warming_up_to_rl/configs/eval_modal_qwen4b.toml +24 -0
- examples/warming_up_to_rl/configs/eval_stepwise_complex.toml +35 -0
- examples/warming_up_to_rl/configs/eval_stepwise_consistent.toml +26 -0
- examples/warming_up_to_rl/configs/eval_stepwise_per_achievement.toml +36 -0
- examples/warming_up_to_rl/configs/eval_stepwise_simple.toml +32 -0
- examples/warming_up_to_rl/configs/rl_from_base_qwen4b.toml +85 -0
- examples/warming_up_to_rl/configs/rl_from_ft.toml +58 -0
- examples/warming_up_to_rl/export_trace_sft.py +837 -0
- examples/warming_up_to_rl/groq_test.py +97 -0
- examples/warming_up_to_rl/manage_secrets.py +131 -0
- examples/warming_up_to_rl/old/event_rewards.md +234 -0
- examples/warming_up_to_rl/old/notes.md +73 -0
- examples/warming_up_to_rl/readme.md +110 -0
- examples/warming_up_to_rl/run_eval.py +736 -0
- examples/warming_up_to_rl/run_fft_and_save.py +380 -0
- examples/warming_up_to_rl/run_local_rollout.py +239 -0
- examples/warming_up_to_rl/run_local_rollout_modal.py +248 -0
- examples/warming_up_to_rl/run_local_rollout_parallel.py +405 -0
- examples/warming_up_to_rl/run_local_rollout_traced.py +477 -0
- examples/warming_up_to_rl/run_rl_and_save.py +124 -0
- examples/warming_up_to_rl/run_rollout_remote.py +156 -0
- examples/warming_up_to_rl/task_app/README.md +42 -0
- examples/warming_up_to_rl/task_app/grpo_crafter.py +876 -0
- examples/warming_up_to_rl/task_app/grpo_crafter_task_app.py +135 -0
- examples/warming_up_to_rl/task_app/synth_envs_hosted/README.md +173 -0
- examples/warming_up_to_rl/task_app/synth_envs_hosted/__init__.py +5 -0
- examples/warming_up_to_rl/task_app/synth_envs_hosted/branching.py +143 -0
- examples/warming_up_to_rl/task_app/synth_envs_hosted/environment_routes.py +1226 -0
- examples/warming_up_to_rl/task_app/synth_envs_hosted/envs/__init__.py +1 -0
- examples/warming_up_to_rl/task_app/synth_envs_hosted/envs/crafter/__init__.py +6 -0
- examples/warming_up_to_rl/task_app/synth_envs_hosted/envs/crafter/app.py +1 -0
- examples/warming_up_to_rl/task_app/synth_envs_hosted/envs/crafter/environment.py +522 -0
- examples/warming_up_to_rl/task_app/synth_envs_hosted/envs/crafter/policy.py +454 -0
- examples/warming_up_to_rl/task_app/synth_envs_hosted/envs/crafter/react_agent.py +108 -0
- examples/warming_up_to_rl/task_app/synth_envs_hosted/envs/crafter/shared.py +305 -0
- examples/warming_up_to_rl/task_app/synth_envs_hosted/envs/crafter/tools.py +47 -0
- examples/warming_up_to_rl/task_app/synth_envs_hosted/hosted_app.py +253 -0
- examples/warming_up_to_rl/task_app/synth_envs_hosted/inference/__init__.py +5 -0
- examples/warming_up_to_rl/task_app/synth_envs_hosted/inference/openai_client.py +729 -0
- examples/warming_up_to_rl/task_app/synth_envs_hosted/main.py +100 -0
- examples/warming_up_to_rl/task_app/synth_envs_hosted/policy_routes.py +1114 -0
- examples/warming_up_to_rl/task_app/synth_envs_hosted/registry.py +195 -0
- examples/warming_up_to_rl/task_app/synth_envs_hosted/rollout.py +1891 -0
- examples/warming_up_to_rl/task_app/synth_envs_hosted/storage/__init__.py +5 -0
- examples/warming_up_to_rl/task_app/synth_envs_hosted/storage/volume.py +211 -0
- examples/warming_up_to_rl/task_app/synth_envs_hosted/test_agents.py +161 -0
- examples/warming_up_to_rl/task_app/synth_envs_hosted/test_service.py +137 -0
- examples/warming_up_to_rl/task_app/synth_envs_hosted/utils.py +129 -0
- examples/workflows/math_rl/configs/eval_base_qwen.toml +15 -0
- examples/workflows/math_rl/configs/eval_rl_qwen.toml +11 -0
- examples/workflows/math_rl/configs/rl_from_base_qwen.toml +62 -0
- examples/workflows/math_rl/configs/rl_from_base_qwen17.toml +80 -0
- examples/workflows/math_rl/configs/rl_from_ft_qwen.toml +35 -0
- examples/workflows/math_rl/download_dataset.py +80 -0
- examples/workflows/math_rl/run_eval.py +436 -0
- examples/workflows/math_rl/run_rl_and_save.py +111 -0
- synth_ai/__init__.py +47 -23
- synth_ai/_utils/__init__.py +47 -0
- synth_ai/_utils/base_url.py +10 -0
- synth_ai/_utils/http.py +10 -0
- synth_ai/_utils/prompts.py +10 -0
- synth_ai/_utils/task_app_state.py +12 -0
- synth_ai/_utils/user_config.py +10 -0
- synth_ai/api/models/supported.py +514 -0
- synth_ai/api/train/__init__.py +63 -0
- synth_ai/api/train/builders.py +473 -0
- synth_ai/api/train/cli.py +1185 -0
- synth_ai/api/train/config_finder.py +246 -0
- synth_ai/api/train/configs/__init__.py +65 -0
- synth_ai/api/train/configs/prompt_learning.py +496 -0
- synth_ai/api/train/configs/rl.py +188 -0
- synth_ai/api/train/configs/sft.py +99 -0
- synth_ai/api/train/configs/shared.py +81 -0
- synth_ai/api/train/env_resolver.py +352 -0
- synth_ai/api/train/pollers.py +91 -0
- synth_ai/api/train/prompt_learning.py +425 -0
- synth_ai/api/train/sft.py +390 -0
- synth_ai/api/train/supported_algos.py +147 -0
- synth_ai/api/train/task_app.py +195 -0
- synth_ai/api/train/utils.py +244 -0
- synth_ai/api/train/validators.py +1117 -0
- synth_ai/api/tunnel.py +49 -0
- synth_ai/auth/credentials.py +94 -0
- synth_ai/baseline/__init__.py +25 -0
- synth_ai/baseline/config.py +209 -0
- synth_ai/baseline/discovery.py +214 -0
- synth_ai/baseline/execution.py +146 -0
- synth_ai/cfgs.py +227 -0
- synth_ai/cli/__init__.py +90 -45
- synth_ai/cli/_modal_wrapper.py +31 -0
- synth_ai/cli/_storage.py +20 -0
- synth_ai/cli/_typer_patch.py +47 -0
- synth_ai/cli/_validate_task_app.py +29 -0
- synth_ai/cli/balance.py +16 -4
- synth_ai/cli/calc.py +36 -21
- synth_ai/cli/claude.py +70 -0
- synth_ai/cli/codex.py +267 -0
- synth_ai/cli/commands/__init__.py +18 -0
- synth_ai/cli/commands/baseline/__init__.py +12 -0
- synth_ai/cli/commands/baseline/core.py +637 -0
- synth_ai/cli/commands/baseline/list.py +93 -0
- synth_ai/cli/commands/demo/__init__.py +6 -0
- synth_ai/cli/commands/demo/core.py +163 -0
- synth_ai/cli/commands/eval/__init__.py +19 -0
- synth_ai/cli/commands/eval/core.py +1112 -0
- synth_ai/cli/commands/eval/errors.py +81 -0
- synth_ai/cli/commands/eval/validation.py +133 -0
- synth_ai/cli/commands/filter/__init__.py +12 -0
- synth_ai/cli/commands/filter/core.py +424 -0
- synth_ai/cli/commands/filter/errors.py +55 -0
- synth_ai/cli/commands/filter/validation.py +77 -0
- synth_ai/cli/commands/help/__init__.py +185 -0
- synth_ai/cli/commands/help/core.py +72 -0
- synth_ai/cli/commands/smoke/__init__.py +7 -0
- synth_ai/cli/commands/smoke/core.py +1437 -0
- synth_ai/cli/commands/status/__init__.py +66 -0
- synth_ai/cli/commands/status/client.py +192 -0
- synth_ai/cli/commands/status/config.py +92 -0
- synth_ai/cli/commands/status/errors.py +20 -0
- synth_ai/cli/commands/status/formatters.py +164 -0
- synth_ai/cli/commands/status/subcommands/__init__.py +9 -0
- synth_ai/cli/commands/status/subcommands/files.py +79 -0
- synth_ai/cli/commands/status/subcommands/jobs.py +334 -0
- synth_ai/cli/commands/status/subcommands/models.py +79 -0
- synth_ai/cli/commands/status/subcommands/pricing.py +22 -0
- synth_ai/cli/commands/status/subcommands/runs.py +81 -0
- synth_ai/cli/commands/status/subcommands/session.py +183 -0
- synth_ai/cli/commands/status/subcommands/summary.py +47 -0
- synth_ai/cli/commands/status/subcommands/usage.py +203 -0
- synth_ai/cli/commands/status/utils.py +114 -0
- synth_ai/cli/commands/train/__init__.py +53 -0
- synth_ai/cli/commands/train/core.py +21 -0
- synth_ai/cli/commands/train/errors.py +117 -0
- synth_ai/cli/commands/train/judge_schemas.py +200 -0
- synth_ai/cli/commands/train/judge_validation.py +305 -0
- synth_ai/cli/commands/train/validation.py +386 -0
- synth_ai/cli/demo.py +32 -140
- synth_ai/cli/deploy.py +233 -0
- synth_ai/cli/eval/__init__.py +36 -0
- synth_ai/cli/eval/core.py +5 -0
- synth_ai/cli/eval/errors.py +31 -0
- synth_ai/cli/eval/validation.py +5 -0
- synth_ai/cli/filter/__init__.py +28 -0
- synth_ai/cli/filter/core.py +5 -0
- synth_ai/cli/filter/errors.py +23 -0
- synth_ai/cli/filter/validation.py +5 -0
- synth_ai/cli/legacy_root_backup.py +28 -22
- synth_ai/cli/lib/__init__.py +10 -0
- synth_ai/cli/lib/task_app_discovery.py +7 -0
- synth_ai/cli/lib/task_app_env.py +518 -0
- synth_ai/cli/mcp.py +34 -0
- synth_ai/cli/modal_serve/__init__.py +12 -0
- synth_ai/cli/modal_serve/core.py +14 -0
- synth_ai/cli/modal_serve/errors.py +8 -0
- synth_ai/cli/modal_serve/validation.py +11 -0
- synth_ai/cli/opencode.py +256 -0
- synth_ai/cli/recent.py +13 -7
- synth_ai/cli/rl_demo.py +166 -114
- synth_ai/cli/root.py +143 -112
- synth_ai/cli/serve/__init__.py +12 -0
- synth_ai/cli/serve/core.py +14 -0
- synth_ai/cli/serve/errors.py +8 -0
- synth_ai/cli/serve/validation.py +11 -0
- synth_ai/cli/setup.py +49 -0
- synth_ai/cli/status.py +7 -125
- synth_ai/cli/task_app_deploy.py +7 -0
- synth_ai/cli/task_app_list.py +25 -0
- synth_ai/cli/task_app_modal_serve.py +11 -0
- synth_ai/cli/task_app_serve.py +11 -0
- synth_ai/cli/task_apps.py +3134 -0
- synth_ai/cli/traces.py +9 -5
- synth_ai/cli/train/__init__.py +12 -0
- synth_ai/cli/train/core.py +21 -0
- synth_ai/cli/train/errors.py +8 -0
- synth_ai/cli/train/validation.py +24 -0
- synth_ai/cli/train.py +5 -0
- synth_ai/cli/turso.py +73 -0
- synth_ai/cli/watch.py +13 -18
- synth_ai/demos/__init__.py +10 -0
- synth_ai/demos/core/__init__.py +28 -1
- synth_ai/demos/core/cli.py +745 -416
- synth_ai/demos/crafter/__init__.py +1 -0
- synth_ai/demos/crafter/crafter_fft_4b.toml +55 -0
- synth_ai/demos/crafter/grpo_crafter_task_app.py +185 -0
- synth_ai/demos/crafter/rl_from_base_qwen4b.toml +74 -0
- synth_ai/demos/demo_registry.py +176 -0
- synth_ai/demos/demo_task_apps/__init__.py +7 -1
- synth_ai/demos/demo_task_apps/core.py +75 -37
- synth_ai/demos/demo_task_apps/crafter/__init__.py +1 -0
- synth_ai/demos/demo_task_apps/crafter/configs/crafter_fft_4b.toml +53 -0
- synth_ai/demos/demo_task_apps/crafter/configs/rl_from_base_qwen4b.toml +73 -0
- synth_ai/demos/demo_task_apps/crafter/grpo_crafter_task_app.py +184 -0
- synth_ai/demos/demo_task_apps/math/_common.py +1 -2
- synth_ai/demos/demo_task_apps/math/app.py +2 -1
- synth_ai/demos/demo_task_apps/math/config.toml +55 -110
- synth_ai/demos/demo_task_apps/math/deploy_modal.py +3 -6
- synth_ai/demos/demo_task_apps/math/modal_task_app.py +491 -166
- synth_ai/demos/demo_task_apps/math/task_app_entry.py +37 -0
- synth_ai/demos/math/__init__.py +1 -0
- synth_ai/demos/math/_common.py +16 -0
- synth_ai/demos/math/app.py +38 -0
- synth_ai/demos/math/config.toml +76 -0
- synth_ai/demos/math/deploy_modal.py +54 -0
- synth_ai/demos/math/modal_task_app.py +703 -0
- synth_ai/demos/math/task_app_entry.py +51 -0
- synth_ai/environments/environment/core.py +7 -1
- synth_ai/environments/examples/bandit/engine.py +12 -5
- synth_ai/environments/examples/bandit/environment.py +0 -1
- synth_ai/environments/examples/bandit/taskset.py +4 -4
- synth_ai/environments/examples/crafter_classic/engine_deterministic_patch.py +7 -4
- synth_ai/environments/examples/crafter_classic/engine_serialization_patch_v3.py +9 -5
- synth_ai/environments/examples/crafter_classic/environment.py +93 -2
- synth_ai/environments/examples/crafter_classic/world_config_patch_simple.py +4 -3
- synth_ai/environments/examples/enron/engine.py +7 -2
- synth_ai/environments/examples/enron/environment.py +68 -0
- synth_ai/environments/examples/red/engine.py +60 -12
- synth_ai/environments/examples/red/engine_helpers/memory_map.py +7 -0
- synth_ai/environments/examples/red/engine_helpers/reward_components.py +151 -179
- synth_ai/environments/examples/red/engine_helpers/reward_library/pallet_town_progression.py +477 -0
- synth_ai/environments/examples/red/engine_helpers/state_extraction.py +32 -0
- synth_ai/environments/examples/red/environment.py +86 -0
- synth_ai/environments/examples/red/trace_hooks_v3.py +168 -0
- synth_ai/environments/examples/sokoban/taskset.py +116 -0
- synth_ai/environments/examples/verilog/engine.py +104 -12
- synth_ai/environments/examples/wordle/environment.py +0 -1
- synth_ai/environments/reproducibility/tree.py +5 -6
- synth_ai/environments/service/app.py +11 -12
- synth_ai/environments/service/core_routes.py +10 -9
- synth_ai/environments/stateful/engine.py +1 -1
- synth_ai/environments/tasks/core.py +1 -0
- synth_ai/environments/tasks/filters.py +5 -6
- synth_ai/environments/tasks/utils.py +4 -5
- synth_ai/evals/__init__.py +15 -0
- synth_ai/evals/base.py +14 -5
- synth_ai/evals/client.py +82 -0
- synth_ai/evals/types.py +42 -0
- synth_ai/http.py +8 -22
- synth_ai/http_client.py +45 -12
- synth_ai/inference/__init__.py +0 -2
- synth_ai/inference/client.py +21 -7
- synth_ai/jobs/client.py +129 -80
- synth_ai/judge_schemas.py +127 -0
- synth_ai/learning/__init__.py +51 -6
- synth_ai/learning/algorithms.py +14 -0
- synth_ai/learning/client.py +122 -30
- synth_ai/learning/config.py +2 -40
- synth_ai/learning/constants.py +0 -2
- synth_ai/learning/ft_client.py +4 -56
- synth_ai/learning/health.py +14 -8
- synth_ai/learning/jobs.py +43 -47
- synth_ai/learning/prompt_learning_client.py +276 -0
- synth_ai/learning/prompt_learning_types.py +185 -0
- synth_ai/{rl → learning/rl}/__init__.py +14 -5
- synth_ai/learning/rl/client.py +269 -0
- synth_ai/learning/rl/config.py +31 -0
- synth_ai/{rl → learning/rl}/contracts.py +5 -10
- synth_ai/{rl → learning/rl}/env_keys.py +45 -16
- synth_ai/learning/rl/secrets.py +13 -0
- synth_ai/learning/rl_client.py +2 -253
- synth_ai/learning/sft/__init__.py +29 -0
- synth_ai/learning/sft/client.py +68 -0
- synth_ai/learning/sft/config.py +270 -0
- synth_ai/learning/sft/data.py +698 -0
- synth_ai/learning/sse.py +25 -26
- synth_ai/learning/validators.py +29 -25
- synth_ai/mcp/__init__.py +5 -0
- synth_ai/mcp/__main__.py +8 -0
- synth_ai/mcp/main.py +254 -0
- synth_ai/mcp/setup.py +100 -0
- synth_ai/modal.py +257 -0
- synth_ai/pricing/__init__.py +3 -0
- synth_ai/pricing/model_pricing.py +64 -0
- synth_ai/session/__init__.py +75 -0
- synth_ai/session/client.py +383 -0
- synth_ai/session/constants.py +63 -0
- synth_ai/session/exceptions.py +105 -0
- synth_ai/session/manager.py +139 -0
- synth_ai/session/models.py +89 -0
- synth_ai/session/query.py +110 -0
- synth_ai/spec/__init__.py +46 -0
- synth_ai/spec/dataclasses.py +149 -0
- synth_ai/spec/loader.py +144 -0
- synth_ai/spec/serializer.py +199 -0
- synth_ai/spec/validation.py +250 -0
- synth_ai/streaming/__init__.py +29 -0
- synth_ai/streaming/config.py +94 -0
- synth_ai/streaming/handlers.py +589 -0
- synth_ai/streaming/streamer.py +320 -0
- synth_ai/streaming/types.py +95 -0
- synth_ai/task/__init__.py +116 -3
- synth_ai/task/apps/__init__.py +132 -0
- synth_ai/task/auth.py +165 -0
- synth_ai/task/client.py +167 -0
- synth_ai/task/config.py +261 -0
- synth_ai/task/contracts.py +173 -57
- synth_ai/task/datasets.py +108 -0
- synth_ai/task/errors.py +50 -0
- synth_ai/task/health.py +17 -11
- synth_ai/task/inference_api.py +101 -0
- synth_ai/task/json.py +111 -0
- synth_ai/task/proxy.py +251 -0
- synth_ai/task/rubrics/__init__.py +55 -0
- synth_ai/task/rubrics/loaders.py +156 -0
- synth_ai/task/rubrics/models.py +57 -0
- synth_ai/task/rubrics/scoring.py +116 -0
- synth_ai/task/rubrics/strict.py +149 -0
- synth_ai/task/rubrics.py +219 -0
- synth_ai/task/server.py +432 -0
- synth_ai/task/trace_correlation_helpers.py +328 -0
- synth_ai/task/tracing_utils.py +95 -0
- synth_ai/task/validators.py +449 -6
- synth_ai/task/vendors.py +59 -0
- synth_ai/tracing_v3/__init__.py +4 -0
- synth_ai/tracing_v3/abstractions.py +21 -4
- synth_ai/tracing_v3/config.py +167 -22
- synth_ai/tracing_v3/constants.py +21 -0
- synth_ai/tracing_v3/db_config.py +42 -29
- synth_ai/tracing_v3/decorators.py +80 -45
- synth_ai/tracing_v3/examples/basic_usage.py +15 -9
- synth_ai/tracing_v3/hooks.py +6 -4
- synth_ai/tracing_v3/llm_call_record_helpers.py +161 -61
- synth_ai/tracing_v3/migration_helper.py +1 -2
- synth_ai/tracing_v3/replica_sync.py +12 -7
- synth_ai/tracing_v3/serialization.py +130 -0
- synth_ai/tracing_v3/session_tracer.py +86 -21
- synth_ai/tracing_v3/storage/base.py +98 -12
- synth_ai/tracing_v3/storage/config.py +63 -16
- synth_ai/tracing_v3/storage/factory.py +11 -9
- synth_ai/tracing_v3/storage/utils.py +15 -11
- synth_ai/tracing_v3/trace_utils.py +317 -0
- synth_ai/tracing_v3/turso/__init__.py +8 -21
- synth_ai/tracing_v3/turso/daemon.py +123 -15
- synth_ai/tracing_v3/turso/models.py +5 -2
- synth_ai/tracing_v3/turso/native_manager.py +1293 -0
- synth_ai/tracing_v3/utils.py +5 -4
- synth_ai/tunnel.py +143 -0
- synth_ai/tunnel_deploy.py +278 -0
- synth_ai/types.py +8 -0
- synth_ai/urls.py +11 -0
- synth_ai/utils/__init__.py +166 -0
- synth_ai/utils/agents.py +74 -0
- synth_ai/utils/apps.py +152 -0
- synth_ai/utils/base_url.py +94 -0
- synth_ai/utils/bin.py +39 -0
- synth_ai/utils/claude.py +36 -0
- synth_ai/utils/cli.py +284 -0
- synth_ai/utils/config.py +81 -0
- synth_ai/utils/env.py +346 -0
- synth_ai/utils/errors.py +85 -0
- synth_ai/utils/http.py +172 -0
- synth_ai/utils/json.py +72 -0
- synth_ai/utils/log_filter.py +99 -0
- synth_ai/utils/logging.py +198 -0
- synth_ai/utils/modal.py +299 -0
- synth_ai/utils/paths.py +95 -0
- synth_ai/utils/process.py +233 -0
- synth_ai/utils/prompts.py +39 -0
- synth_ai/utils/sqld.py +122 -0
- synth_ai/utils/ssl.py +25 -0
- synth_ai/utils/task_app_discovery.py +882 -0
- synth_ai/utils/task_app_env.py +186 -0
- synth_ai/utils/task_app_state.py +318 -0
- synth_ai/utils/tunnel/__init__.py +12 -0
- synth_ai/utils/tunnel/config.py +55 -0
- synth_ai/utils/user_config.py +137 -0
- synth_ai/uvicorn.py +77 -0
- synth_ai-0.2.23.dev3.dist-info/METADATA +357 -0
- synth_ai-0.2.23.dev3.dist-info/RECORD +983 -0
- {synth_ai-0.2.8.dev4.dist-info → synth_ai-0.2.23.dev3.dist-info}/entry_points.txt +0 -1
- {synth_ai-0.2.8.dev4.dist-info → synth_ai-0.2.23.dev3.dist-info}/top_level.txt +1 -0
- synth_ai/cli/man.py +0 -106
- synth_ai/core/experiment.py +0 -15
- synth_ai/core/system.py +0 -15
- synth_ai/environments/examples/sokoban/units/astar_common.py +0 -95
- synth_ai/experimental/synth_oss.py +0 -446
- synth_ai/handshake.py +0 -63
- synth_ai/install_sqld.sh +0 -40
- synth_ai/learning/offline/dpo.py +0 -0
- synth_ai/learning/offline/providers.py +0 -7
- synth_ai/learning/offline/sft.py +0 -0
- synth_ai/learning/offline/shared.py +0 -0
- synth_ai/learning/online/grpo.py +0 -0
- synth_ai/learning/online/irft.py +0 -0
- synth_ai/learning/prompts/banking77_injection_eval.py +0 -168
- synth_ai/learning/prompts/gepa.py +0 -0
- synth_ai/learning/prompts/hello_world_in_context_injection_ex.py +0 -213
- synth_ai/learning/prompts/mipro.py +0 -289
- synth_ai/learning/prompts/random_search.py +0 -246
- synth_ai/learning/prompts/run_mipro_banking77.py +0 -172
- synth_ai/learning/prompts/run_random_search_banking77.py +0 -324
- synth_ai/lm/__init__.py +0 -51
- synth_ai/lm/caching/constants.py +0 -6
- synth_ai/lm/caching/dbs.py +0 -0
- synth_ai/lm/caching/ephemeral.py +0 -102
- synth_ai/lm/caching/handler.py +0 -137
- synth_ai/lm/caching/initialize.py +0 -11
- synth_ai/lm/caching/persistent.py +0 -114
- synth_ai/lm/config.py +0 -110
- synth_ai/lm/constants.py +0 -32
- synth_ai/lm/core/__init__.py +0 -8
- synth_ai/lm/core/all.py +0 -73
- synth_ai/lm/core/exceptions.py +0 -7
- synth_ai/lm/core/main.py +0 -319
- synth_ai/lm/core/main_v3.py +0 -594
- synth_ai/lm/core/synth_models.py +0 -48
- synth_ai/lm/core/vendor_clients.py +0 -188
- synth_ai/lm/cost/monitor.py +0 -1
- synth_ai/lm/cost/statefulness.py +0 -1
- synth_ai/lm/injection.py +0 -80
- synth_ai/lm/overrides.py +0 -206
- synth_ai/lm/provider_support/__init__.py +0 -8
- synth_ai/lm/provider_support/anthropic.py +0 -972
- synth_ai/lm/provider_support/openai.py +0 -1139
- synth_ai/lm/provider_support/suppress_logging.py +0 -31
- synth_ai/lm/structured_outputs/handler.py +0 -440
- synth_ai/lm/structured_outputs/inject.py +0 -297
- synth_ai/lm/structured_outputs/rehabilitate.py +0 -185
- synth_ai/lm/tools/__init__.py +0 -3
- synth_ai/lm/tools/base.py +0 -172
- synth_ai/lm/unified_interface.py +0 -202
- synth_ai/lm/vendors/base.py +0 -81
- synth_ai/lm/vendors/core/anthropic_api.py +0 -387
- synth_ai/lm/vendors/core/gemini_api.py +0 -292
- synth_ai/lm/vendors/core/mistral_api.py +0 -322
- synth_ai/lm/vendors/core/openai_api.py +0 -225
- synth_ai/lm/vendors/core/synth_dev_api.py +0 -0
- synth_ai/lm/vendors/local/ollama.py +0 -0
- synth_ai/lm/vendors/openai_standard.py +0 -780
- synth_ai/lm/vendors/openai_standard_responses.py +0 -256
- synth_ai/lm/vendors/retries.py +0 -22
- synth_ai/lm/vendors/supported/custom_endpoint.py +0 -417
- synth_ai/lm/vendors/supported/deepseek.py +0 -69
- synth_ai/lm/vendors/supported/grok.py +0 -75
- synth_ai/lm/vendors/supported/groq.py +0 -16
- synth_ai/lm/vendors/supported/ollama.py +0 -15
- synth_ai/lm/vendors/supported/openrouter.py +0 -74
- synth_ai/lm/vendors/supported/together.py +0 -11
- synth_ai/lm/vendors/synth_client.py +0 -808
- synth_ai/lm/warmup.py +0 -186
- synth_ai/rl/secrets.py +0 -19
- synth_ai/scripts/verify_rewards.py +0 -100
- synth_ai/tracing/__init__.py +0 -30
- synth_ai/tracing_v1/__init__.py +0 -33
- synth_ai/tracing_v3/turso/manager.py +0 -760
- synth_ai/v0/tracing/abstractions.py +0 -224
- synth_ai/v0/tracing/base_client.py +0 -91
- synth_ai/v0/tracing/client_manager.py +0 -131
- synth_ai/v0/tracing/config.py +0 -142
- synth_ai/v0/tracing/context.py +0 -146
- synth_ai/v0/tracing/decorators.py +0 -682
- synth_ai/v0/tracing/events/__init__.py +0 -0
- synth_ai/v0/tracing/events/manage.py +0 -147
- synth_ai/v0/tracing/events/scope.py +0 -86
- synth_ai/v0/tracing/events/store.py +0 -228
- synth_ai/v0/tracing/immediate_client.py +0 -151
- synth_ai/v0/tracing/local.py +0 -18
- synth_ai/v0/tracing/log_client_base.py +0 -73
- synth_ai/v0/tracing/retry_queue.py +0 -186
- synth_ai/v0/tracing/trackers.py +0 -515
- synth_ai/v0/tracing/upload.py +0 -512
- synth_ai/v0/tracing/utils.py +0 -9
- synth_ai/v0/tracing_v1/__init__.py +0 -16
- synth_ai/v0/tracing_v1/abstractions.py +0 -224
- synth_ai/v0/tracing_v1/base_client.py +0 -91
- synth_ai/v0/tracing_v1/client_manager.py +0 -131
- synth_ai/v0/tracing_v1/config.py +0 -142
- synth_ai/v0/tracing_v1/context.py +0 -146
- synth_ai/v0/tracing_v1/decorators.py +0 -703
- synth_ai/v0/tracing_v1/events/__init__.py +0 -0
- synth_ai/v0/tracing_v1/events/manage.py +0 -147
- synth_ai/v0/tracing_v1/events/scope.py +0 -86
- synth_ai/v0/tracing_v1/events/store.py +0 -228
- synth_ai/v0/tracing_v1/immediate_client.py +0 -151
- synth_ai/v0/tracing_v1/local.py +0 -18
- synth_ai/v0/tracing_v1/log_client_base.py +0 -73
- synth_ai/v0/tracing_v1/retry_queue.py +0 -186
- synth_ai/v0/tracing_v1/trackers.py +0 -515
- synth_ai/v0/tracing_v1/upload.py +0 -527
- synth_ai/v0/tracing_v1/utils.py +0 -9
- synth_ai/zyk/__init__.py +0 -30
- synth_ai-0.2.8.dev4.dist-info/METADATA +0 -129
- synth_ai-0.2.8.dev4.dist-info/RECORD +0 -420
- {synth_ai/lm/caching → examples/task_apps}/__init__.py +0 -0
- {synth_ai/lm/cost → examples/task_apps/crafter}/__init__.py +0 -0
- {synth_ai/lm/structured_outputs → examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/server}/__init__.py +0 -0
- {synth_ai/lm/vendors → examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/tests}/__init__.py +0 -0
- {synth_ai/lm/vendors/core → examples/task_apps/dev/pokemon_emerald/external/pokeagent-speedrun/utils}/__init__.py +0 -0
- {synth_ai/lm/vendors/local → examples/task_apps/math}/__init__.py +0 -0
- {synth_ai/lm/vendors/supported → examples/workflows}/__init__.py +0 -0
- {synth_ai/v0/tracing → examples/workflows/math_rl}/__init__.py +0 -0
- /synth_ai/{compound/cais.py → cli/__main__.py} +0 -0
- /synth_ai/{learning/filtering.py → py.typed} +0 -0
- {synth_ai-0.2.8.dev4.dist-info → synth_ai-0.2.23.dev3.dist-info}/WHEEL +0 -0
- {synth_ai-0.2.8.dev4.dist-info → synth_ai-0.2.23.dev3.dist-info}/licenses/LICENSE +0 -0
|
@@ -0,0 +1,51 @@
|
|
|
1
|
+
"""Task app registry entry for the math demo Modal deployment."""
|
|
2
|
+
|
|
3
|
+
from __future__ import annotations
|
|
4
|
+
|
|
5
|
+
from contextlib import suppress
|
|
6
|
+
from importlib import import_module
|
|
7
|
+
|
|
8
|
+
from synth_ai.task.apps import ModalDeploymentConfig, TaskAppEntry, register_task_app
|
|
9
|
+
|
|
10
|
+
try:
|
|
11
|
+
from synth_ai.task.apps.math_single_step import build_config as base_build_config
|
|
12
|
+
except ModuleNotFoundError:
|
|
13
|
+
base_module = import_module("examples.rl.task_app.math_single_step")
|
|
14
|
+
base_build_config = base_module.build_config
|
|
15
|
+
|
|
16
|
+
DEMO_MODAL_CONFIG = ModalDeploymentConfig(
|
|
17
|
+
app_name="hendrycks-math-task-app",
|
|
18
|
+
pip_packages=(
|
|
19
|
+
"fastapi>=0.110.0",
|
|
20
|
+
"uvicorn>=0.23.0",
|
|
21
|
+
"pydantic>=2.6.0",
|
|
22
|
+
"httpx>=0.24.0",
|
|
23
|
+
"numpy>=1.24.0",
|
|
24
|
+
"aiohttp>=3.8.0",
|
|
25
|
+
"datasets>=2.16.0",
|
|
26
|
+
"synth-ai",
|
|
27
|
+
),
|
|
28
|
+
)
|
|
29
|
+
|
|
30
|
+
|
|
31
|
+
def build_config():
|
|
32
|
+
"""Reuse the shared math single-step TaskAppConfig."""
|
|
33
|
+
|
|
34
|
+
return base_build_config()
|
|
35
|
+
|
|
36
|
+
|
|
37
|
+
def register_demo_entry() -> None:
|
|
38
|
+
entry = TaskAppEntry(
|
|
39
|
+
app_id="hendrycks-math-demo",
|
|
40
|
+
description="Demo math task app (Modal-focused) shipping with synth-ai demos.",
|
|
41
|
+
config_factory=build_config,
|
|
42
|
+
modal=DEMO_MODAL_CONFIG,
|
|
43
|
+
)
|
|
44
|
+
with suppress(ValueError):
|
|
45
|
+
register_task_app(entry=entry)
|
|
46
|
+
|
|
47
|
+
|
|
48
|
+
register_demo_entry()
|
|
49
|
+
|
|
50
|
+
|
|
51
|
+
__all__ = ["DEMO_MODAL_CONFIG", "build_config", "register_demo_entry"]
|
|
@@ -4,7 +4,6 @@ from dataclasses import dataclass
|
|
|
4
4
|
from typing import Any
|
|
5
5
|
|
|
6
6
|
import numpy as np
|
|
7
|
-
|
|
8
7
|
from synth_ai.environments.environment.shared_engine import (
|
|
9
8
|
GetObservationCallable,
|
|
10
9
|
InternalObservation,
|
|
@@ -191,7 +190,9 @@ class BanditEngine(StatefulEngine, IReproducibleEngine):
|
|
|
191
190
|
step_count=self.step_count,
|
|
192
191
|
max_steps=self.max_steps,
|
|
193
192
|
last_arm=self.last_arm,
|
|
194
|
-
last_reward=float(reward)
|
|
193
|
+
last_reward=float(reward)
|
|
194
|
+
if reward is not None
|
|
195
|
+
else (self.last_reward if self.step_count else None),
|
|
195
196
|
cumulative_reward=float(self.total_reward),
|
|
196
197
|
reward_history=self.reward_history.copy(),
|
|
197
198
|
arm_pull_counts=self.arm_pull_counts.copy(),
|
|
@@ -238,7 +239,9 @@ class BanditEngine(StatefulEngine, IReproducibleEngine):
|
|
|
238
239
|
engine.arm_probabilities = data.get("arm_probabilities", engine.arm_probabilities)
|
|
239
240
|
engine.arm_means = data.get("arm_means", engine.arm_means)
|
|
240
241
|
engine.arm_stds = data.get("arm_stds", engine.arm_stds)
|
|
241
|
-
engine.true_expected_rewards = list(
|
|
242
|
+
engine.true_expected_rewards = list(
|
|
243
|
+
data.get("true_expected_rewards", engine.true_expected_rewards)
|
|
244
|
+
)
|
|
242
245
|
engine.arm_count = len(engine.true_expected_rewards)
|
|
243
246
|
|
|
244
247
|
engine.step_count = int(data.get("step_count", 0))
|
|
@@ -247,7 +250,9 @@ class BanditEngine(StatefulEngine, IReproducibleEngine):
|
|
|
247
250
|
engine.last_arm = data.get("last_arm")
|
|
248
251
|
engine.reward_history = list(data.get("reward_history", []))
|
|
249
252
|
engine.arm_history = list(data.get("arm_history", []))
|
|
250
|
-
engine.arm_pull_counts = list(
|
|
253
|
+
engine.arm_pull_counts = list(
|
|
254
|
+
data.get("arm_pull_counts", [0 for _ in range(engine.arm_count)])
|
|
255
|
+
)
|
|
251
256
|
engine.terminated = bool(data.get("terminated", False))
|
|
252
257
|
engine.status = data.get("status", "in_progress")
|
|
253
258
|
|
|
@@ -287,7 +292,9 @@ class SynthBanditCheckpointObservationCallable(GetObservationCallable):
|
|
|
287
292
|
"arm_count": pub.arm_count,
|
|
288
293
|
"total_reward": priv.total_reward,
|
|
289
294
|
"steps_taken": pub.step_count,
|
|
290
|
-
"best_expected_reward": max(priv.true_expected_rewards)
|
|
295
|
+
"best_expected_reward": max(priv.true_expected_rewards)
|
|
296
|
+
if priv.true_expected_rewards
|
|
297
|
+
else None,
|
|
291
298
|
"terminated": pub.terminated,
|
|
292
299
|
"status": pub.status,
|
|
293
300
|
}
|
|
@@ -156,10 +156,10 @@ async def create_bandit_taskset(
|
|
|
156
156
|
)
|
|
157
157
|
|
|
158
158
|
expected = _expected_rewards(metadata)
|
|
159
|
-
arm_count =
|
|
160
|
-
len(
|
|
161
|
-
|
|
162
|
-
or 0
|
|
159
|
+
arm_count = (
|
|
160
|
+
len(expected)
|
|
161
|
+
if expected
|
|
162
|
+
else (len(metadata.arm_probabilities or []) or len(metadata.arm_means or []) or 0)
|
|
163
163
|
)
|
|
164
164
|
if arm_count == 0:
|
|
165
165
|
arm_count = 1
|
|
@@ -6,16 +6,18 @@ the hash-based set-iteration nondeterminism that caused the drift.
|
|
|
6
6
|
"""
|
|
7
7
|
|
|
8
8
|
import collections
|
|
9
|
+
import os
|
|
9
10
|
|
|
10
11
|
import crafter
|
|
11
12
|
|
|
12
|
-
|
|
13
|
+
# Patch messages permanently disabled to reduce noise
|
|
14
|
+
# print("[PATCH] Attempting to apply Crafter deterministic patch...")
|
|
13
15
|
|
|
14
16
|
# -----------------------------------------------------------------------------
|
|
15
17
|
# 1. Make per–chunk object order stable
|
|
16
18
|
# -----------------------------------------------------------------------------
|
|
17
19
|
if not hasattr(crafter.Env, "_orig_balance_object"):
|
|
18
|
-
print("[PATCH] Patching crafter.Env._balance_object...")
|
|
20
|
+
# print("[PATCH] Patching crafter.Env._balance_object...")
|
|
19
21
|
crafter.Env._orig_balance_object = crafter.Env._balance_object
|
|
20
22
|
|
|
21
23
|
def _balance_object_det(self, chunk, objs, *args, **kwargs):
|
|
@@ -25,9 +27,10 @@ if not hasattr(crafter.Env, "_orig_balance_object"):
|
|
|
25
27
|
return crafter.Env._orig_balance_object(self, chunk, objs, *args, **kwargs)
|
|
26
28
|
|
|
27
29
|
crafter.Env._balance_object = _balance_object_det
|
|
28
|
-
print("[PATCH] crafter.Env._balance_object patched.")
|
|
30
|
+
# print("[PATCH] crafter.Env._balance_object patched.")
|
|
29
31
|
else:
|
|
30
|
-
|
|
32
|
+
pass
|
|
33
|
+
# print("[PATCH] crafter.Env._balance_object already patched or _orig_balance_object exists.")
|
|
31
34
|
|
|
32
35
|
# -----------------------------------------------------------------------------
|
|
33
36
|
# 2. Make *chunk* iteration order stable
|
|
@@ -4,6 +4,7 @@ This version handles player references for Zombie and Skeleton objects.
|
|
|
4
4
|
"""
|
|
5
5
|
|
|
6
6
|
import collections
|
|
7
|
+
import os
|
|
7
8
|
import pickle
|
|
8
9
|
from typing import Any, Dict, Optional, Set
|
|
9
10
|
|
|
@@ -11,11 +12,12 @@ import crafter
|
|
|
11
12
|
import numpy as np
|
|
12
13
|
from crafter import objects
|
|
13
14
|
|
|
14
|
-
|
|
15
|
+
# Patch messages permanently disabled
|
|
16
|
+
# print("[PATCH] Attempting to apply Crafter serialization patch v3...")
|
|
15
17
|
|
|
16
18
|
# Check if already patched
|
|
17
19
|
if not hasattr(crafter.Env, "save"):
|
|
18
|
-
print("[PATCH] Adding enhanced save/load methods to crafter.Env...")
|
|
20
|
+
# print("[PATCH] Adding enhanced save/load methods to crafter.Env...")
|
|
19
21
|
|
|
20
22
|
def _save(self) -> Dict[str, Any]:
|
|
21
23
|
"""Save complete environment state including all details."""
|
|
@@ -260,8 +262,10 @@ if not hasattr(crafter.Env, "save"):
|
|
|
260
262
|
crafter.Env.save = _save
|
|
261
263
|
crafter.Env.load = _load
|
|
262
264
|
|
|
263
|
-
|
|
265
|
+
pass
|
|
266
|
+
# print("[PATCH] crafter.Env.save() and load() methods added (v3).")
|
|
264
267
|
else:
|
|
265
|
-
|
|
268
|
+
pass
|
|
269
|
+
# print("[PATCH] crafter.Env already has save/load methods.")
|
|
266
270
|
|
|
267
|
-
print("[PATCH] Crafter serialization patch v3 complete.")
|
|
271
|
+
# print("[PATCH] Crafter serialization patch v3 complete.")
|
|
@@ -2,11 +2,16 @@
|
|
|
2
2
|
|
|
3
3
|
from __future__ import annotations
|
|
4
4
|
|
|
5
|
+
import base64
|
|
5
6
|
import dataclasses
|
|
6
7
|
import logging
|
|
7
8
|
import time
|
|
9
|
+
from io import BytesIO
|
|
8
10
|
from typing import Any, Dict, List, Optional, Union
|
|
9
11
|
|
|
12
|
+
import numpy as np
|
|
13
|
+
from PIL import Image
|
|
14
|
+
|
|
10
15
|
# Import tracing abstractions
|
|
11
16
|
from synth_ai.tracing_v3.abstractions import (
|
|
12
17
|
RuntimeEvent,
|
|
@@ -43,6 +48,51 @@ from synth_ai.environments.reproducibility.core import ReproducibleEnvironment
|
|
|
43
48
|
from synth_ai.environments.stateful.core import StatefulEnvironment
|
|
44
49
|
|
|
45
50
|
|
|
51
|
+
def _convert_numpy_to_python(obj: Any) -> Any:
|
|
52
|
+
if isinstance(obj, np.integer):
|
|
53
|
+
return int(obj)
|
|
54
|
+
if isinstance(obj, np.floating):
|
|
55
|
+
return float(obj)
|
|
56
|
+
if isinstance(obj, np.ndarray):
|
|
57
|
+
return obj.tolist()
|
|
58
|
+
if isinstance(obj, dict):
|
|
59
|
+
return {k: _convert_numpy_to_python(v) for k, v in obj.items()}
|
|
60
|
+
if isinstance(obj, (list, tuple)):
|
|
61
|
+
return [_convert_numpy_to_python(item) for item in obj]
|
|
62
|
+
return obj
|
|
63
|
+
|
|
64
|
+
|
|
65
|
+
def _encode_image_to_base64(image_array: Any) -> dict[str, Any] | None:
|
|
66
|
+
if not isinstance(image_array, np.ndarray):
|
|
67
|
+
return None
|
|
68
|
+
if image_array.ndim != 3 or image_array.shape[-1] not in (1, 3, 4):
|
|
69
|
+
return None
|
|
70
|
+
try:
|
|
71
|
+
array_uint8 = (
|
|
72
|
+
image_array.astype("uint8")
|
|
73
|
+
if image_array.dtype != np.uint8
|
|
74
|
+
else image_array # pragma: no cover - fast path
|
|
75
|
+
)
|
|
76
|
+
mode = "L" if array_uint8.shape[-1] == 1 else "RGB"
|
|
77
|
+
if array_uint8.shape[-1] == 4:
|
|
78
|
+
mode = "RGBA"
|
|
79
|
+
img = Image.fromarray(array_uint8, mode=mode)
|
|
80
|
+
buffer = BytesIO()
|
|
81
|
+
img.save(buffer, format="PNG")
|
|
82
|
+
encoded = base64.b64encode(buffer.getvalue()).decode("ascii")
|
|
83
|
+
width = int(array_uint8.shape[1])
|
|
84
|
+
height = int(array_uint8.shape[0])
|
|
85
|
+
return {
|
|
86
|
+
"format": "png",
|
|
87
|
+
"width": width,
|
|
88
|
+
"height": height,
|
|
89
|
+
"data": encoded,
|
|
90
|
+
"data_url": f"data:image/png;base64,{encoded}",
|
|
91
|
+
}
|
|
92
|
+
except Exception:
|
|
93
|
+
return None
|
|
94
|
+
|
|
95
|
+
|
|
46
96
|
# --- Tool Definition ---
|
|
47
97
|
class CrafterActionInput(BaseModel):
|
|
48
98
|
action: int = Field(..., description="Integer action for the Crafter environment.")
|
|
@@ -140,6 +190,22 @@ class SynthCrafterObservationCallable(GetObservationCallable):
|
|
|
140
190
|
obs_dict["truncated"] = priv.truncated
|
|
141
191
|
if pub.error_info:
|
|
142
192
|
obs_dict["tool_error"] = pub.error_info
|
|
193
|
+
counts_payload = {}
|
|
194
|
+
try:
|
|
195
|
+
counts = getattr(priv, "achievements_current_values", {}) or {}
|
|
196
|
+
for k, v in counts.items():
|
|
197
|
+
try:
|
|
198
|
+
counts_payload[str(k)] = int(v)
|
|
199
|
+
except Exception:
|
|
200
|
+
try:
|
|
201
|
+
counts_payload[str(k)] = int(float(v))
|
|
202
|
+
except Exception:
|
|
203
|
+
continue
|
|
204
|
+
if counts_payload:
|
|
205
|
+
obs_dict["achievements_counts"] = counts_payload
|
|
206
|
+
except Exception:
|
|
207
|
+
# Best effort; omit counts if coercion fails
|
|
208
|
+
pass
|
|
143
209
|
|
|
144
210
|
# Derive a simple local semantic patch around the player for easy rendering
|
|
145
211
|
try:
|
|
@@ -331,7 +397,7 @@ class CrafterClassicEnvironment(StatefulEnvironment, ReproducibleEnvironment[Cra
|
|
|
331
397
|
priv_state, pub_state, self.custom_step_observation_callable
|
|
332
398
|
)
|
|
333
399
|
total_step_time = time.time() - step_start_time
|
|
334
|
-
logger.
|
|
400
|
+
logger.debug(
|
|
335
401
|
f"CrafterClassic step completed in {total_step_time:.3f}s (interact: {interact_time:.3f}s)"
|
|
336
402
|
)
|
|
337
403
|
return obs
|
|
@@ -362,7 +428,8 @@ class CrafterClassicEnvironment(StatefulEnvironment, ReproducibleEnvironment[Cra
|
|
|
362
428
|
state_before = {"private_state": priv, "public_state": pub}
|
|
363
429
|
|
|
364
430
|
active_obs_cb = obs_cb or SynthCrafterObservationCallable()
|
|
365
|
-
|
|
431
|
+
raw_observation = await active_obs_cb.get_observation(pub, priv)
|
|
432
|
+
observation = self._prepare_observation(raw_observation)
|
|
366
433
|
if extra_obs and isinstance(observation, dict):
|
|
367
434
|
observation.update(extra_obs)
|
|
368
435
|
|
|
@@ -385,6 +452,30 @@ class CrafterClassicEnvironment(StatefulEnvironment, ReproducibleEnvironment[Cra
|
|
|
385
452
|
|
|
386
453
|
return observation
|
|
387
454
|
|
|
455
|
+
def _prepare_observation(self, observation: Any) -> dict[str, Any]:
|
|
456
|
+
obs_dict: dict[str, Any]
|
|
457
|
+
image_payload: dict[str, Any] | None = None
|
|
458
|
+
|
|
459
|
+
if isinstance(observation, dict):
|
|
460
|
+
image_payload = _encode_image_to_base64(observation.get("observation_image"))
|
|
461
|
+
sanitized = dict(observation)
|
|
462
|
+
sanitized.pop("observation_image", None)
|
|
463
|
+
obs_dict = _convert_numpy_to_python(sanitized) or {}
|
|
464
|
+
else:
|
|
465
|
+
obs_dict = _convert_numpy_to_python(observation) or {}
|
|
466
|
+
|
|
467
|
+
if not isinstance(obs_dict, dict):
|
|
468
|
+
obs_dict = {"value": obs_dict}
|
|
469
|
+
|
|
470
|
+
if image_payload:
|
|
471
|
+
obs_dict["observation_image_base64"] = image_payload["data"]
|
|
472
|
+
obs_dict["observation_image_format"] = image_payload["format"]
|
|
473
|
+
obs_dict["observation_image_width"] = image_payload["width"]
|
|
474
|
+
obs_dict["observation_image_height"] = image_payload["height"]
|
|
475
|
+
obs_dict["observation_image_data_url"] = image_payload["data_url"]
|
|
476
|
+
|
|
477
|
+
return obs_dict
|
|
478
|
+
|
|
388
479
|
# ────────────────────────────────────────────────────────────────────
|
|
389
480
|
# ReproducibleEnvironment plumbing
|
|
390
481
|
# ────────────────────────────────────────────────────────────────────
|
|
@@ -9,7 +9,8 @@ from typing import Any, Dict, Optional
|
|
|
9
9
|
|
|
10
10
|
import crafter
|
|
11
11
|
|
|
12
|
-
|
|
12
|
+
# Patch messages permanently disabled
|
|
13
|
+
# print("[PATCH] Attempting to apply simplified Crafter world configuration patch...")
|
|
13
14
|
|
|
14
15
|
# World configuration presets
|
|
15
16
|
WORLD_CONFIGS = {
|
|
@@ -279,8 +280,8 @@ def patched_env_init(
|
|
|
279
280
|
|
|
280
281
|
crafter.Env.__init__ = patched_env_init
|
|
281
282
|
|
|
282
|
-
print("[PATCH] Simplified Crafter world configuration patch complete.")
|
|
283
|
-
print("[PATCH] Available configs: easy, normal, hard, peaceful")
|
|
283
|
+
# print("[PATCH] Simplified Crafter world configuration patch complete.")
|
|
284
|
+
# print("[PATCH] Available configs: easy, normal, hard, peaceful")
|
|
284
285
|
|
|
285
286
|
# Example custom config
|
|
286
287
|
EXAMPLE_CUSTOM_CONFIG = {
|
|
@@ -26,7 +26,10 @@ from synth_ai.environments.examples.enron.taskset import EnronTaskInstance
|
|
|
26
26
|
|
|
27
27
|
# SQLite-backed helpers
|
|
28
28
|
from synth_ai.environments.stateful.engine import StatefulEngine, StatefulEngineSnapshot
|
|
29
|
-
|
|
29
|
+
try: # pragma: no cover - optional dependency
|
|
30
|
+
from synth_ai.zyk import LM # type: ignore
|
|
31
|
+
except ImportError: # pragma: no cover - fallback when LM unavailable
|
|
32
|
+
LM = None
|
|
30
33
|
|
|
31
34
|
# --------------------------------------------------------------------------- actions
|
|
32
35
|
ACTION_SEARCH = "search"
|
|
@@ -244,7 +247,9 @@ class EnronEngine(StatefulEngine):
|
|
|
244
247
|
async def determine_if_answer_is_correct(
|
|
245
248
|
question: str, gold_answer: str, agent_answer: str
|
|
246
249
|
) -> bool:
|
|
247
|
-
|
|
250
|
+
if LM is None:
|
|
251
|
+
return gold_answer.strip().lower() == agent_answer.strip().lower()
|
|
252
|
+
|
|
248
253
|
llm = LM(model_name="gpt-4.1-nano", formatting_model_name="gpt-4.1-nano", temperature=0.0)
|
|
249
254
|
|
|
250
255
|
system_prompt = (
|
|
@@ -9,6 +9,7 @@ from synth_ai.environments.environment.shared_engine import (
|
|
|
9
9
|
InternalObservation,
|
|
10
10
|
)
|
|
11
11
|
from synth_ai.environments.environment.tools import (
|
|
12
|
+
AbstractTool,
|
|
12
13
|
TOOL_REGISTRY,
|
|
13
14
|
EnvToolCall,
|
|
14
15
|
ToolResult,
|
|
@@ -65,6 +66,73 @@ class Terminate(EnvToolCall):
|
|
|
65
66
|
self.action = (ACTION_ANSWER, "")
|
|
66
67
|
|
|
67
68
|
|
|
69
|
+
class TerminateArgs(BaseModel):
|
|
70
|
+
pass
|
|
71
|
+
|
|
72
|
+
|
|
73
|
+
class SearchEmailsTool(AbstractTool):
|
|
74
|
+
name = "search_emails"
|
|
75
|
+
call_schema = SearchEmailsArgs
|
|
76
|
+
|
|
77
|
+
def __init__(self, engine: EnronEngine):
|
|
78
|
+
self.engine = engine
|
|
79
|
+
|
|
80
|
+
async def __call__(self, call: EnvToolCall) -> ToolResult:
|
|
81
|
+
try:
|
|
82
|
+
args = self.call_schema.model_validate(call.args or {})
|
|
83
|
+
results = await self.engine.search_emails_action(args.model_dump())
|
|
84
|
+
return ToolResult(ok=True, payload={"search_results": results})
|
|
85
|
+
except Exception as exc: # pragma: no cover - runtime safety
|
|
86
|
+
return ToolResult(ok=False, error=str(exc))
|
|
87
|
+
|
|
88
|
+
|
|
89
|
+
class ReadEmailTool(AbstractTool):
|
|
90
|
+
name = "read_email"
|
|
91
|
+
call_schema = ReadEmailArgs
|
|
92
|
+
|
|
93
|
+
def __init__(self, engine: EnronEngine):
|
|
94
|
+
self.engine = engine
|
|
95
|
+
|
|
96
|
+
async def __call__(self, call: EnvToolCall) -> ToolResult:
|
|
97
|
+
try:
|
|
98
|
+
args = self.call_schema.model_validate(call.args or {})
|
|
99
|
+
email = await self.engine.read_email_action(args.message_id)
|
|
100
|
+
return ToolResult(ok=True, payload={"email": email})
|
|
101
|
+
except Exception as exc: # pragma: no cover
|
|
102
|
+
return ToolResult(ok=False, error=str(exc))
|
|
103
|
+
|
|
104
|
+
|
|
105
|
+
class AnswerQuestionTool(AbstractTool):
|
|
106
|
+
name = "answer_question"
|
|
107
|
+
call_schema = AnswerQuestionArgs
|
|
108
|
+
|
|
109
|
+
def __init__(self, engine: EnronEngine):
|
|
110
|
+
self.engine = engine
|
|
111
|
+
|
|
112
|
+
async def __call__(self, call: EnvToolCall) -> ToolResult:
|
|
113
|
+
try:
|
|
114
|
+
args = self.call_schema.model_validate(call.args or {})
|
|
115
|
+
await self.engine.answer_question_action(args.answer)
|
|
116
|
+
return ToolResult(ok=True, payload={"status": "answer_recorded"})
|
|
117
|
+
except Exception as exc: # pragma: no cover
|
|
118
|
+
return ToolResult(ok=False, error=str(exc))
|
|
119
|
+
|
|
120
|
+
|
|
121
|
+
class TerminateTool(AbstractTool):
|
|
122
|
+
name = "terminate"
|
|
123
|
+
call_schema = TerminateArgs
|
|
124
|
+
|
|
125
|
+
def __init__(self, engine: EnronEngine):
|
|
126
|
+
self.engine = engine
|
|
127
|
+
|
|
128
|
+
async def __call__(self, call: EnvToolCall) -> ToolResult:
|
|
129
|
+
try:
|
|
130
|
+
await self.engine.answer_question_action("")
|
|
131
|
+
return ToolResult(ok=True, payload={"status": "terminated"})
|
|
132
|
+
except Exception as exc: # pragma: no cover
|
|
133
|
+
return ToolResult(ok=False, error=str(exc))
|
|
134
|
+
|
|
135
|
+
|
|
68
136
|
# -------- observation callable (optional for formatted observations)
|
|
69
137
|
class SynthEnronObservationCallable(GetObservationCallable):
|
|
70
138
|
async def get_observation(
|
|
@@ -2,6 +2,7 @@ from __future__ import annotations
|
|
|
2
2
|
|
|
3
3
|
import logging
|
|
4
4
|
from dataclasses import dataclass
|
|
5
|
+
import os
|
|
5
6
|
from pathlib import Path
|
|
6
7
|
from typing import Any, Dict, List, Optional
|
|
7
8
|
|
|
@@ -13,12 +14,15 @@ from synth_ai.environments.stateful.engine import StatefulEngine, StatefulEngine
|
|
|
13
14
|
from synth_ai.environments.tasks.core import TaskInstance
|
|
14
15
|
|
|
15
16
|
from .engine_helpers.reward_components import (
|
|
16
|
-
|
|
17
|
-
|
|
18
|
-
|
|
19
|
-
|
|
17
|
+
RouteExplorationReward,
|
|
18
|
+
StrategicTrainingReward,
|
|
19
|
+
BattleProgressionReward,
|
|
20
|
+
GymPreparationReward,
|
|
21
|
+
ItemCollectionReward,
|
|
22
|
+
HealingManagementReward,
|
|
23
|
+
EfficientExplorationReward,
|
|
24
|
+
BadgeVictoryReward,
|
|
20
25
|
StepPenaltyComponent,
|
|
21
|
-
XPGainComponent,
|
|
22
26
|
)
|
|
23
27
|
from .engine_helpers.state_extraction import extract_game_state
|
|
24
28
|
|
|
@@ -122,6 +126,13 @@ class GameSystemState:
|
|
|
122
126
|
menu_state: int
|
|
123
127
|
text_box_active: bool
|
|
124
128
|
warp_flag: int
|
|
129
|
+
# Battle-specific data
|
|
130
|
+
enemy_hp_current: int = 0
|
|
131
|
+
enemy_hp_max: int = 0
|
|
132
|
+
enemy_hp_percentage: float = 0.0
|
|
133
|
+
enemy_level: int = 0
|
|
134
|
+
enemy_species_id: int = 0
|
|
135
|
+
battle_turn: int = 0
|
|
125
136
|
# TODO: Add when available
|
|
126
137
|
# current_menu_type: str = ""
|
|
127
138
|
# dialogue_speaker: str = ""
|
|
@@ -260,15 +271,27 @@ class PokemonRedEngine(StatefulEngine, IReproducibleEngine):
|
|
|
260
271
|
# For testing purposes, use None emulator
|
|
261
272
|
self.emulator = None
|
|
262
273
|
|
|
263
|
-
# Initialize reward stack with
|
|
274
|
+
# Initialize reward stack with comprehensive progress-based components
|
|
264
275
|
self.reward_stack = RewardStack(
|
|
265
276
|
components=[
|
|
266
|
-
|
|
267
|
-
|
|
268
|
-
|
|
269
|
-
|
|
270
|
-
|
|
271
|
-
|
|
277
|
+
# Major progress rewards
|
|
278
|
+
BadgeVictoryReward(), # +50.0 for Boulder Badge (main goal)
|
|
279
|
+
RouteExplorationReward(), # +1.0-5.0 for reaching key areas
|
|
280
|
+
GymPreparationReward(), # +3.0 for being gym-ready
|
|
281
|
+
|
|
282
|
+
# Training and battle rewards
|
|
283
|
+
StrategicTrainingReward(), # +0.2-3.0 for level ups and milestones
|
|
284
|
+
BattleProgressionReward(), # +0.1-1.0 for battles
|
|
285
|
+
|
|
286
|
+
# Resource management rewards
|
|
287
|
+
ItemCollectionReward(), # +0.1-0.5 for collecting items
|
|
288
|
+
HealingManagementReward(), # +0.05-0.8 for healing Pokemon
|
|
289
|
+
|
|
290
|
+
# Exploration efficiency
|
|
291
|
+
EfficientExplorationReward(), # +0.02 for discovering new positions
|
|
292
|
+
|
|
293
|
+
# No penalty for unproductive actions
|
|
294
|
+
StepPenaltyComponent(penalty=0.0), # 0.0 per step
|
|
272
295
|
]
|
|
273
296
|
)
|
|
274
297
|
|
|
@@ -278,12 +301,21 @@ class PokemonRedEngine(StatefulEngine, IReproducibleEngine):
|
|
|
278
301
|
|
|
279
302
|
def _get_rom_path(self) -> Path:
|
|
280
303
|
"""Get path to Pokemon Red ROM file"""
|
|
304
|
+
# Highest priority: explicit environment variable
|
|
305
|
+
env_rom = os.getenv("POKEMON_RED_ROM")
|
|
306
|
+
if env_rom:
|
|
307
|
+
p = Path(env_rom).expanduser()
|
|
308
|
+
if p.exists():
|
|
309
|
+
return p
|
|
310
|
+
|
|
281
311
|
# Check several possible locations
|
|
282
312
|
possible_paths = [
|
|
283
313
|
Path(__file__).parent / "roms" / "pokemon_red.gb",
|
|
284
314
|
Path(__file__).parent / "roms" / "PokemonRed.gb",
|
|
285
315
|
Path(__file__).parent / "vendor" / "pokemon_red.gb",
|
|
286
316
|
Path.home() / "Games" / "pokemon_red.gb",
|
|
317
|
+
# Common example location where users may drop the ROM
|
|
318
|
+
Path(__file__).resolve().parents[5] / "examples" / "task_apps" / "pokemon_red" / "Pokemon - Red Version (USA, Europe) (SGB Enhanced).gb",
|
|
287
319
|
]
|
|
288
320
|
|
|
289
321
|
for path in possible_paths:
|
|
@@ -533,6 +565,12 @@ class PokemonRedEngine(StatefulEngine, IReproducibleEngine):
|
|
|
533
565
|
menu_state=int(current_state.get("menu_state", 0)),
|
|
534
566
|
text_box_active=bool(current_state.get("text_box_active", False)),
|
|
535
567
|
warp_flag=int(current_state.get("warp_flag", 0)),
|
|
568
|
+
enemy_hp_current=int(current_state.get("enemy_hp_current", 0)),
|
|
569
|
+
enemy_hp_max=int(current_state.get("enemy_hp_max", 0)),
|
|
570
|
+
enemy_hp_percentage=float(current_state.get("enemy_hp_percentage", 0.0)),
|
|
571
|
+
enemy_level=int(current_state.get("enemy_level", 0)),
|
|
572
|
+
enemy_species_id=int(current_state.get("enemy_species_id", 0)),
|
|
573
|
+
battle_turn=int(current_state.get("battle_turn", 0)),
|
|
536
574
|
),
|
|
537
575
|
)
|
|
538
576
|
|
|
@@ -613,6 +651,16 @@ class PokemonRedEngine(StatefulEngine, IReproducibleEngine):
|
|
|
613
651
|
"prev_in_battle": bool(prev_state.get("in_battle", False)),
|
|
614
652
|
"prev_party_level": int(prev_state.get("party_level", 0)),
|
|
615
653
|
"prev_party_xp": int(prev_state.get("party_xp", 0)),
|
|
654
|
+
"prev_party_count": int(prev_state.get("party_count", 0)),
|
|
655
|
+
"prev_text_box_active": bool(prev_state.get("text_box_active", False)),
|
|
656
|
+
"prev_enemy_hp_current": int(prev_state.get("enemy_hp_current", 0)),
|
|
657
|
+
"prev_enemy_hp_percentage": float(prev_state.get("enemy_hp_percentage", 0.0)),
|
|
658
|
+
"prev_player_x": int(prev_state.get("player_x", 0)),
|
|
659
|
+
"prev_player_y": int(prev_state.get("player_y", 0)),
|
|
660
|
+
"prev_party": prev_state.get("party", []),
|
|
661
|
+
"prev_inventory": prev_state.get("inventory", []),
|
|
662
|
+
"prev_party_hp_current": int(prev_state.get("party_hp_current", 0)),
|
|
663
|
+
"prev_party_hp_max": int(prev_state.get("party_hp_max", 0)),
|
|
616
664
|
},
|
|
617
665
|
)
|
|
618
666
|
except Exception as e:
|
|
@@ -6,6 +6,13 @@ PLAYER_Y = 0xD361 # player Y coordinate
|
|
|
6
6
|
IN_BATTLE_FLAG = 0xD057 # battle state flag
|
|
7
7
|
BATTLE_OUTCOME = 0xD089 # 0=ongoing, 1=win, 2=lose
|
|
8
8
|
|
|
9
|
+
# Battle-specific data
|
|
10
|
+
ENEMY_HP_CURRENT = 0xCFE6 # enemy Pokemon current HP (2 bytes)
|
|
11
|
+
ENEMY_HP_MAX = 0xCFE8 # enemy Pokemon max HP (2 bytes)
|
|
12
|
+
ENEMY_LEVEL = 0xD127 # enemy Pokemon level
|
|
13
|
+
ENEMY_SPECIES = 0xCFE5 # enemy Pokemon species ID
|
|
14
|
+
BATTLE_TURN = 0xCC2F # current battle turn counter
|
|
15
|
+
|
|
9
16
|
# Party Pokemon data (up to 6 Pokemon)
|
|
10
17
|
PARTY_COUNT = 0xD163 # number of Pokemon in party (0-6)
|
|
11
18
|
PARTY_SPECIES = 0xD164 # species of each Pokemon (6 bytes)
|