# ankurjyotisarma/can-llms-help-us-in-playing-simulated-table-tennis-through-a-musculoskeletal-model architecture

> Architecture diagram of the GitHub repository ankurjyotisarma/can-llms-help-us-in-playing-simulated-table-tennis-through-a-musculoskeletal-model, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2026-10-06). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/ankurjyotisarma/can-llms-help-us-in-playing-simulated-table-tennis-through-a-musculoskeletal-model
- Repository: https://github.com/ankurjyotisarma/can-llms-help-us-in-playing-simulated-table-tennis-through-a-musculoskeletal-model

## Overview

This research codebase trains and evaluates PPO agents for musculoskeletal table tennis in MyoSuite. The principal workflow is to shape rewards around simulated play, train or fine-tune an agent, then evaluate gameplay or checkpoints and inspect reward, success-rate, and learning-curve analytics. The README documents these capabilities; excerpts show V8 environment setup and reward shaping, plus evaluation and graph-generation logic. Training-to-artifact links are documented at a high level, but their exact save/load wiring is not shown in the excerpts.

Main entry points: [`3. Fine Tuned versions V7 & V8/V8 ( Iteration 8 ) Agent/V8.py`](https://github.com/ankurjyotisarma/can-llms-help-us-in-playing-simulated-table-tennis-through-a-musculoskeletal-model/blob/main/3.%20Fine%20Tuned%20versions%20V7%20%26%20V8/V8%20(%20Iteration%208%20)%20Agent/V8.py) (V8 training), [`4. Detailed Evaluation using graphs`](https://github.com/ankurjyotisarma/can-llms-help-us-in-playing-simulated-table-tennis-through-a-musculoskeletal-model/tree/main/4.%20Detailed%20Evaluation%20using%20graphs) (Interactive analytics), [`4. Detailed Evaluation using graphs/2. Evaluation of Final Trained Agent V8 (Iteration 8)/Final_Model_Evaluation.py`](https://github.com/ankurjyotisarma/can-llms-help-us-in-playing-simulated-table-tennis-through-a-musculoskeletal-model/blob/main/4.%20Detailed%20Evaluation%20using%20graphs/2.%20Evaluation%20of%20Final%20Trained%20Agent%20V8%20(Iteration%208)/Final_Model_Evaluation.py) (Final V8 evaluation), [`4. Detailed Evaluation using graphs/1. Whole Learning Curve (Chcking all Check-Points)/1. MYO_AGENT_V8_LEARNING_CURVE/1. V8 Evaluation — 500 Games per Checkpoint/Learning_Curve_V8.py`](https://github.com/ankurjyotisarma/can-llms-help-us-in-playing-simulated-table-tennis-through-a-musculoskeletal-model/blob/main/4.%20Detailed%20Evaluation%20using%20graphs/1.%20Whole%20Learning%20Curve%20(Chcking%20all%20Check-Points)/1.%20MYO_AGENT_V8_LEARNING_CURVE/1.%20V8%20Evaluation%20%E2%80%94%20500%20Games%20per%20Checkpoint/Learning_Curve_V8.py) (V8 500-game curve)

## Diagram (Mermaid)

```mermaid
flowchart TD

subgraph group_training["Training iterations"]
  node_iterations1to6["Iterations V1–V6<br/>[2. Python files]"]
  node_v7train["V7 training<br/>[V7.py]"]
  node_v8train["V8 training<br/>[V8.py]"]
  node_v8baseline["V8 training baseline<br/>[V8.py]"]
  node_rewardshape["Reward shaping<br/>[V8.py]"]
  node_ppo["PPO learner"]
  node_checkpoints["Model checkpoints"]
  node_normalization["Normalization stats"]
end

subgraph group_agent["Agent evaluation"]
  node_livev7["V7 live evaluation"]
  node_livev8["V8 live evaluation"]
  node_finaleval["Final V8 evaluation"]
end

subgraph group_analytics["Performance analytics"]
  node_curve500["V8 500-game curve"]
  node_curve200["V8 200-game curve"]
  node_combinedcurve["V7/V8 curve"]
  node_dashboards["Interactive analytics"]
end

subgraph group_environment["Simulation environment"]
  node_myosuite{{"MyoSuite table tennis"}}
end

node_researcher(("Researcher"))

node_researcher -->|"starts training"| node_v8train
node_researcher -->|"evaluates agent"| node_livev8
node_v8train -->|"applies shaping"| node_rewardshape
node_v8train -->|"creates environment"| node_myosuite
node_v8train -->|"configures PPO"| node_ppo
node_v7train -.->|"trains in"| node_myosuite
node_v8baseline -->|"creates environment"| node_myosuite
node_v8train -.->|"saves checkpoints"| node_checkpoints
node_v8train -.->|"stores statistics"| node_normalization
node_livev7 -->|"evaluates episodes"| node_myosuite
node_livev8 -->|"evaluates episodes"| node_myosuite
node_finaleval -->|"runs evaluation"| node_myosuite
node_curve500 -->|"evaluates checkpoints"| node_myosuite
node_curve200 -->|"evaluates checkpoints"| node_myosuite
node_combinedcurve -->|"evaluates checkpoints"| node_myosuite
node_curve500 -->|"generates graphs"| node_dashboards
node_curve200 -->|"generates graphs"| node_dashboards
node_combinedcurve -->|"generates graphs"| node_dashboards
node_finaleval -->|"generates graphs"| node_dashboards

click node_iterations1to6 "https://github.com/ankurjyotisarma/can-llms-help-us-in-playing-simulated-table-tennis-through-a-musculoskeletal-model/tree/main/2.%20Python%20files"
click node_v7train "https://github.com/ankurjyotisarma/can-llms-help-us-in-playing-simulated-table-tennis-through-a-musculoskeletal-model/blob/main/3.%20Fine%20Tuned%20versions%20V7%20%26%20V8/V7%20Agent/V7.py"
click node_v8train "https://github.com/ankurjyotisarma/can-llms-help-us-in-playing-simulated-table-tennis-through-a-musculoskeletal-model/blob/main/3.%20Fine%20Tuned%20versions%20V7%20%26%20V8/V8%20(%20Iteration%208%20)%20Agent/V8.py"
click node_v8baseline "https://github.com/ankurjyotisarma/can-llms-help-us-in-playing-simulated-table-tennis-through-a-musculoskeletal-model/blob/main/2.%20Python%20files/V8.py"
click node_rewardshape "https://github.com/ankurjyotisarma/can-llms-help-us-in-playing-simulated-table-tennis-through-a-musculoskeletal-model/blob/main/3.%20Fine%20Tuned%20versions%20V7%20%26%20V8/V8%20(%20Iteration%208%20)%20Agent/V8.py"
click node_normalization "https://github.com/ankurjyotisarma/can-llms-help-us-in-playing-simulated-table-tennis-through-a-musculoskeletal-model/blob/main/3.%20Fine%20Tuned%20versions%20V7%20%26%20V8/V8%20(%20Iteration%208%20)%20Agent/vec_normalize_v8.pkl"
click node_livev7 "https://github.com/ankurjyotisarma/can-llms-help-us-in-playing-simulated-table-tennis-through-a-musculoskeletal-model/blob/main/3.%20Fine%20Tuned%20versions%20V7%20%26%20V8/V7%20Agent/Evaluate_Agent_V7_Live_Playing.py"
click node_livev8 "https://github.com/ankurjyotisarma/can-llms-help-us-in-playing-simulated-table-tennis-through-a-musculoskeletal-model/blob/main/3.%20Fine%20Tuned%20versions%20V7%20%26%20V8/V8%20(%20Iteration%208%20)%20Agent/Evaluate_Agent_V8_Live_Playing.py"
click node_finaleval "https://github.com/ankurjyotisarma/can-llms-help-us-in-playing-simulated-table-tennis-through-a-musculoskeletal-model/blob/main/4.%20Detailed%20Evaluation%20using%20graphs/2.%20Evaluation%20of%20Final%20Trained%20Agent%20V8%20(Iteration%208)/Final_Model_Evaluation.py"
click node_curve500 "https://github.com/ankurjyotisarma/can-llms-help-us-in-playing-simulated-table-tennis-through-a-musculoskeletal-model/blob/main/4.%20Detailed%20Evaluation%20using%20graphs/1.%20Whole%20Learning%20Curve%20(Chcking%20all%20Check-Points)/1.%20MYO_AGENT_V8_LEARNING_CURVE/1.%20V8%20Evaluation%20%E2%80%94%20500%20Games%20per%20Checkpoint/Learning_Curve_V8.py"
click node_curve200 "https://github.com/ankurjyotisarma/can-llms-help-us-in-playing-simulated-table-tennis-through-a-musculoskeletal-model/blob/main/4.%20Detailed%20Evaluation%20using%20graphs/1.%20Whole%20Learning%20Curve%20(Chcking%20all%20Check-Points)/1.%20MYO_AGENT_V8_LEARNING_CURVE/2.%20V8%20Evaluation%20%E2%80%94%20200%20Games%20per%20Checkpoint/Learning_Curve_V8.py"
click node_combinedcurve "https://github.com/ankurjyotisarma/can-llms-help-us-in-playing-simulated-table-tennis-through-a-musculoskeletal-model/blob/main/4.%20Detailed%20Evaluation%20using%20graphs/1.%20Whole%20Learning%20Curve%20(Chcking%20all%20Check-Points)/2.%20MYO_AGENT_V7_and_V8_COMBINED_FULL_LEARNING_CURVE/MYO_CHALLENGE_AGENT_V7_and_V8_COMBINED_FULL_LEARNING_CURVE.py"
click node_dashboards "https://github.com/ankurjyotisarma/can-llms-help-us-in-playing-simulated-table-tennis-through-a-musculoskeletal-model/tree/main/4.%20Detailed%20Evaluation%20using%20graphs"

classDef toneNeutral fill:#f8fafc,stroke:#334155,stroke-width:1.5px,color:#0f172a
classDef toneBlue fill:#dbeafe,stroke:#2563eb,stroke-width:1.5px,color:#172554
classDef toneAmber fill:#fef3c7,stroke:#d97706,stroke-width:1.5px,color:#78350f
classDef toneMint fill:#dcfce7,stroke:#16a34a,stroke-width:1.5px,color:#14532d
classDef toneRose fill:#ffe4e6,stroke:#e11d48,stroke-width:1.5px,color:#881337
classDef toneIndigo fill:#e0e7ff,stroke:#4f46e5,stroke-width:1.5px,color:#312e81
classDef toneTeal fill:#ccfbf1,stroke:#0f766e,stroke-width:1.5px,color:#134e4a
class node_iterations1to6,node_v7train,node_v8train,node_v8baseline,node_rewardshape,node_ppo,node_checkpoints,node_normalization toneBlue
class node_livev7,node_livev8,node_finaleval toneAmber
class node_curve500,node_curve200,node_combinedcurve,node_dashboards toneMint
class node_myosuite toneRose
class node_researcher toneIndigo
```

## Components

### Training iterations

- [Iterations V1–V6](https://github.com/ankurjyotisarma/can-llms-help-us-in-playing-simulated-table-tennis-through-a-musculoskeletal-model/tree/main/2.%20Python%20files): `2. Python files`
- [V7 training](https://github.com/ankurjyotisarma/can-llms-help-us-in-playing-simulated-table-tennis-through-a-musculoskeletal-model/blob/main/3.%20Fine%20Tuned%20versions%20V7%20%26%20V8/V7%20Agent/V7.py): `3. Fine Tuned versions V7 & V8/V7 Agent/V7.py`
- [V8 training](https://github.com/ankurjyotisarma/can-llms-help-us-in-playing-simulated-table-tennis-through-a-musculoskeletal-model/blob/main/3.%20Fine%20Tuned%20versions%20V7%20%26%20V8/V8%20(%20Iteration%208%20)%20Agent/V8.py): `3. Fine Tuned versions V7 & V8/V8 ( Iteration 8 ) Agent/V8.py`
- [V8 training baseline](https://github.com/ankurjyotisarma/can-llms-help-us-in-playing-simulated-table-tennis-through-a-musculoskeletal-model/blob/main/2.%20Python%20files/V8.py): `2. Python files/V8.py`
- [Reward shaping](https://github.com/ankurjyotisarma/can-llms-help-us-in-playing-simulated-table-tennis-through-a-musculoskeletal-model/blob/main/3.%20Fine%20Tuned%20versions%20V7%20%26%20V8/V8%20(%20Iteration%208%20)%20Agent/V8.py): `3. Fine Tuned versions V7 & V8/V8 ( Iteration 8 ) Agent/V8.py`
- **PPO learner**
- **Model checkpoints**
- [Normalization stats](https://github.com/ankurjyotisarma/can-llms-help-us-in-playing-simulated-table-tennis-through-a-musculoskeletal-model/blob/main/3.%20Fine%20Tuned%20versions%20V7%20%26%20V8/V8%20(%20Iteration%208%20)%20Agent/vec_normalize_v8.pkl): `3. Fine Tuned versions V7 & V8/V8 ( Iteration 8 ) Agent/vec_normalize_v8.pkl`

### Agent evaluation

- [V7 live evaluation](https://github.com/ankurjyotisarma/can-llms-help-us-in-playing-simulated-table-tennis-through-a-musculoskeletal-model/blob/main/3.%20Fine%20Tuned%20versions%20V7%20%26%20V8/V7%20Agent/Evaluate_Agent_V7_Live_Playing.py): `3. Fine Tuned versions V7 & V8/V7 Agent/Evaluate_Agent_V7_Live_Playing.py`
- [V8 live evaluation](https://github.com/ankurjyotisarma/can-llms-help-us-in-playing-simulated-table-tennis-through-a-musculoskeletal-model/blob/main/3.%20Fine%20Tuned%20versions%20V7%20%26%20V8/V8%20(%20Iteration%208%20)%20Agent/Evaluate_Agent_V8_Live_Playing.py): `3. Fine Tuned versions V7 & V8/V8 ( Iteration 8 ) Agent/Evaluate_Agent_V8_Live_Playing.py`
- [Final V8 evaluation](https://github.com/ankurjyotisarma/can-llms-help-us-in-playing-simulated-table-tennis-through-a-musculoskeletal-model/blob/main/4.%20Detailed%20Evaluation%20using%20graphs/2.%20Evaluation%20of%20Final%20Trained%20Agent%20V8%20(Iteration%208)/Final_Model_Evaluation.py): `4. Detailed Evaluation using graphs/2. Evaluation of Final Trained Agent V8 (Iteration 8)/Final_Model_Evaluation.py`

### Performance analytics

- [V8 500-game curve](https://github.com/ankurjyotisarma/can-llms-help-us-in-playing-simulated-table-tennis-through-a-musculoskeletal-model/blob/main/4.%20Detailed%20Evaluation%20using%20graphs/1.%20Whole%20Learning%20Curve%20(Chcking%20all%20Check-Points)/1.%20MYO_AGENT_V8_LEARNING_CURVE/1.%20V8%20Evaluation%20%E2%80%94%20500%20Games%20per%20Checkpoint/Learning_Curve_V8.py): `4. Detailed Evaluation using graphs/1. Whole Learning Curve (Chcking all Check-Points)/1. MYO_AGENT_V8_LEARNING_CURVE/1. V8 Evaluation — 500 Games per Checkpoint/Learning_Curve_V8.py`
- [V8 200-game curve](https://github.com/ankurjyotisarma/can-llms-help-us-in-playing-simulated-table-tennis-through-a-musculoskeletal-model/blob/main/4.%20Detailed%20Evaluation%20using%20graphs/1.%20Whole%20Learning%20Curve%20(Chcking%20all%20Check-Points)/1.%20MYO_AGENT_V8_LEARNING_CURVE/2.%20V8%20Evaluation%20%E2%80%94%20200%20Games%20per%20Checkpoint/Learning_Curve_V8.py): `4. Detailed Evaluation using graphs/1. Whole Learning Curve (Chcking all Check-Points)/1. MYO_AGENT_V8_LEARNING_CURVE/2. V8 Evaluation — 200 Games per Checkpoint/Learning_Curve_V8.py`
- [V7/V8 curve](https://github.com/ankurjyotisarma/can-llms-help-us-in-playing-simulated-table-tennis-through-a-musculoskeletal-model/blob/main/4.%20Detailed%20Evaluation%20using%20graphs/1.%20Whole%20Learning%20Curve%20(Chcking%20all%20Check-Points)/2.%20MYO_AGENT_V7_and_V8_COMBINED_FULL_LEARNING_CURVE/MYO_CHALLENGE_AGENT_V7_and_V8_COMBINED_FULL_LEARNING_CURVE.py): `4. Detailed Evaluation using graphs/1. Whole Learning Curve (Chcking all Check-Points)/2. MYO_AGENT_V7_and_V8_COMBINED_FULL_LEARNING_CURVE/MYO_CHALLENGE_AGENT_V7_and_V8_COMBINED_FULL_LEARNING_CURVE.py`
- [Interactive analytics](https://github.com/ankurjyotisarma/can-llms-help-us-in-playing-simulated-table-tennis-through-a-musculoskeletal-model/tree/main/4.%20Detailed%20Evaluation%20using%20graphs): `4. Detailed Evaluation using graphs`

### Simulation environment

- **MyoSuite table tennis**

### Other components

- **Researcher**

## Connections

- Researcher → V8 training: starts training (evidence: `README.md`)
- Researcher → V8 live evaluation: evaluates agent (evidence: `README.md`)
- V8 training → Reward shaping: applies shaping (evidence: `3. Fine Tuned versions V7 & V8/V8 ( Iteration 8 ) Agent/V8.py`)
- V8 training → MyoSuite table tennis: creates environment (evidence: `3. Fine Tuned versions V7 & V8/V8 ( Iteration 8 ) Agent/V8.py`)
- V8 training → PPO learner: configures PPO (evidence: `README.md`)
- V7 training → MyoSuite table tennis: trains in
- V8 training baseline → MyoSuite table tennis: creates environment (evidence: `2. Python files/V8.py`)
- V8 training → Model checkpoints: saves checkpoints
- V8 training → Normalization stats: stores statistics
- V7 live evaluation → MyoSuite table tennis: evaluates episodes (evidence: `3. Fine Tuned versions V7 & V8/V7 Agent/Evaluate_Agent_V7_Live_Playing.py`)
- V8 live evaluation → MyoSuite table tennis: evaluates episodes (evidence: `3. Fine Tuned versions V7 & V8/V8 ( Iteration 8 ) Agent/Evaluate_Agent_V8_Live_Playing.py`)
- Final V8 evaluation → MyoSuite table tennis: runs evaluation (evidence: `4. Detailed Evaluation using graphs/2. Evaluation of Final Trained Agent V8 (Iteration 8)/Final_Model_Evaluation.py`)
- V8 500-game curve → MyoSuite table tennis: evaluates checkpoints (evidence: `4. Detailed Evaluation using graphs/1. Whole Learning Curve (Chcking all Check-Points)/1. MYO_AGENT_V8_LEARNING_CURVE/1. V8 Evaluation — 500 Games per Checkpoint/Learning_Curve_V8.py`)
- V8 200-game curve → MyoSuite table tennis: evaluates checkpoints (evidence: `4. Detailed Evaluation using graphs/1. Whole Learning Curve (Chcking all Check-Points)/1. MYO_AGENT_V8_LEARNING_CURVE/2. V8 Evaluation — 200 Games per Checkpoint/Learning_Curve_V8.py`)
- V7/V8 curve → MyoSuite table tennis: evaluates checkpoints (evidence: `4. Detailed Evaluation using graphs/1. Whole Learning Curve (Chcking all Check-Points)/2. MYO_AGENT_V7_and_V8_COMBINED_FULL_LEARNING_CURVE/MYO_CHALLENGE_AGENT_V7_and_V8_COMBINED_FULL_LEARNING_CURVE.py`)
- V8 500-game curve → Interactive analytics: generates graphs (evidence: `4. Detailed Evaluation using graphs/1. Whole Learning Curve (Chcking all Check-Points)/1. MYO_AGENT_V8_LEARNING_CURVE/1. V8 Evaluation — 500 Games per Checkpoint/Learning_Curve_V8.py`)
- V8 200-game curve → Interactive analytics: generates graphs (evidence: `4. Detailed Evaluation using graphs/1. Whole Learning Curve (Chcking all Check-Points)/1. MYO_AGENT_V8_LEARNING_CURVE/2. V8 Evaluation — 200 Games per Checkpoint/Learning_Curve_V8.py`)
- V7/V8 curve → Interactive analytics: generates graphs (evidence: `4. Detailed Evaluation using graphs/1. Whole Learning Curve (Chcking all Check-Points)/2. MYO_AGENT_V7_and_V8_COMBINED_FULL_LEARNING_CURVE/MYO_CHALLENGE_AGENT_V7_and_V8_COMBINED_FULL_LEARNING_CURVE.py`)
- Final V8 evaluation → Interactive analytics: generates graphs (evidence: `4. Detailed Evaluation using graphs/2. Evaluation of Final Trained Agent V8 (Iteration 8)/Final_Model_Evaluation.py`)
