File size: 10,229 Bytes
67f71c2
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
4189d2a
67f71c2
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
"""
Inference Script — Shader Environment
===================================
MANDATORY
- Before submitting, ensure the following variables are defined in your environment configuration:
    API_BASE_URL   The API endpoint for the LLM.
    MODEL_NAME     The model identifier to use for inference.
    HF_TOKEN       Your Hugging Face / API key.
    LOCAL_IMAGE_NAME The name of the local image to use for the environment if you are using from_docker_image()
                     method

- Defaults are set only for API_BASE_URL and MODEL_NAME
    (and should reflect your active inference setup):
    API_BASE_URL = os.getenv("API_BASE_URL", "<your-active-endpoint>")
    MODEL_NAME = os.getenv("MODEL_NAME", "<your-active-model>")

- The inference script must be named `inference.py` and placed in the root directory of the project
- Participants must use OpenAI Client for all LLM calls using above variables

STDOUT FORMAT
- The script must emit exactly three line types to stdout, in this order:

    [START] task=<task_name> env=<benchmark> model=<model_name>
    [STEP]  step=<n> action=<action_str> reward=<0.00> done=<true|false> error=<msg|null>
    [END]   success=<true|false> steps=<n> score=<score> rewards=<r1,r2,...,rn>

  Rules:
    - One [START] line at episode begin.
    - One [STEP] line per step, immediately after env.step() returns.
    - One [END] line after env.close(), always emitted (even on exception).
    - reward and rewards are formatted to 2 decimal places.
    - done and success are lowercase booleans: true or false.
    - error is the raw last_action_error string, or null if none.
    - All fields on a single line with no newlines within a line.
    - Each tasks should return score in [0, 1]

  Example:
    [START] task=gradient env=shader model=Qwen/Qwen2.5-72B-Instruct
    [STEP] step=1 action=<glsl_code> reward=0.45 done=false error=null
    [STEP] step=2 action=<glsl_code> reward=0.92 done=true error=null
    [END] success=true steps=2 score=0.92 rewards=0.45,0.92
"""

import asyncio
import os
import textwrap
from typing import List, Optional

try:
    from shader import ShaderAction, ShaderEnv
except ImportError:
    from envs.shader import ShaderAction, ShaderEnv
from openai import OpenAI

IMAGE_NAME = os.getenv("IMAGE_NAME")
if not IMAGE_NAME:
    raise RuntimeError(
        "IMAGE_NAME environment variable is required. "
        "Set it to the Docker image name (e.g. shader:latest)."
    )
API_KEY = os.getenv("OPENAI_API_KEY") or os.getenv("HF_TOKEN") or os.getenv("API_KEY")

API_BASE_URL = os.getenv("API_BASE_URL", "https://router.huggingface.co/v1")
MODEL_NAME = os.getenv("MODEL_NAME", "Qwen/Qwen2.5-72B-Instruct")
BENCHMARK = "shader"
MAX_STEPS = 5
TEMPERATURE = 0.2
MAX_TOKENS = 4096
SUCCESS_THRESHOLD = 0.90

TASKS = ["Nd33R4", "stlXWH", "ftjSRd"]

SYSTEM_PROMPT = textwrap.dedent("""\
    You are a GLSL shader expert. Your task is to write a Shadertoy-dialect \
    GLSL fragment shader that reproduces the given reference image as closely \
    as possible.

    Rules:
    - Write a `void mainImage(out vec4 fragColor, in vec2 fragCoord)` function.
    - You may use standard Shadertoy uniforms: iResolution, iTime, iTimeDelta, \
    iFrame, iMouse, iDate, iSampleRate.
    - Do NOT include #version, precision, or #extension directives.
    - Output ONLY the raw GLSL code — no markdown fencing, no explanation.

    The rendered output is compared to the reference via SSIM (structural \
    similarity). Target: SSIM >= 0.99.""")


# ---------------------------------------------------------------------------
# Logging helpers (mandatory stdout format)
# ---------------------------------------------------------------------------

def log_start(task: str, env: str, model: str) -> None:
    print(f"[START] task={task} env={env} model={model}", flush=True)


def log_step(
    step: int, action: str, reward: float, done: bool, error: Optional[str]
) -> None:
    error_val = error if error else "null"
    done_val = str(done).lower()
    action_short = action.replace("\n", " ")[:80]
    print(
        f"[STEP] step={step} action={action_short} reward={reward:.2f} "
        f"done={done_val} error={error_val}",
        flush=True,
    )


def log_end(success: bool, steps: int, score: float, rewards: List[float]) -> None:
    rewards_str = ",".join(f"{r:.2f}" for r in rewards)
    print(
        f"[END] success={str(success).lower()} steps={steps} "
        f"score={score:.2f} rewards={rewards_str}",
        flush=True,
    )


# ---------------------------------------------------------------------------
# Code helpers
# ---------------------------------------------------------------------------

def strip_fences(text: str) -> str:
    """Remove markdown code fences if present."""
    text = text.strip()
    if text.startswith("```"):
        lines = text.split("\n")
        end = len(lines) - 1
        while end > 0 and lines[end].strip() != "```":
            end -= 1
        if end > 0:
            return "\n".join(lines[1:end])
        return "\n".join(lines[1:])
    return text


FALLBACK_CODE = ("void mainImage(out vec4 fragColor, in vec2 fragCoord) "
                 "{ fragColor = vec4(0.0); }")


# ---------------------------------------------------------------------------
# LLM interaction
# ---------------------------------------------------------------------------

def call_model(client: OpenAI, conversation: list) -> str:
    """Send the conversation to the model and return stripped GLSL code."""
    try:
        completion = client.chat.completions.create(
            model=MODEL_NAME,
            messages=[{"role": "system", "content": SYSTEM_PROMPT}]
            + conversation,
            temperature=TEMPERATURE,
            max_completion_tokens=MAX_TOKENS,
            stream=False,
        )
        text = (completion.choices[0].message.content or "").strip()
        return strip_fences(text) if text else ""
    except Exception as exc:
        print(f"[DEBUG] Model request failed: {exc}", flush=True)
        return ""


# ---------------------------------------------------------------------------
# Single task episode
# ---------------------------------------------------------------------------

async def run_task(env, client: OpenAI, task_name: str) -> None:
    """Run one episode for a named task."""
    rewards: List[float] = []
    steps_taken = 0
    best_ssim = 0.0
    success = False
    conversation: list = []

    log_start(task=task_name, env=BENCHMARK, model=MODEL_NAME)

    try:
        result = await env.reset(task=task_name)
        obs = result.observation
        ref_b64 = obs.reference_png

        for step in range(1, MAX_STEPS + 1):
            if result.done:
                break

            # First step: send reference image; later steps: send feedback
            if step == 1:
                conversation.append({
                    "role": "user",
                    "content": [
                        {
                            "type": "text",
                            "text": "Write a GLSL shader that reproduces "
                                    "this reference image exactly.",
                        },
                        {
                            "type": "image_url",
                            "image_url": {
                                "url": f"data:image/png;base64,{ref_b64}",
                            },
                        },
                    ],
                })
            else:
                feedback = []
                if not obs.compiled:
                    feedback.append(
                        "Compilation FAILED.\nErrors:\n"
                        + "\n".join(obs.errors))
                elif not obs.rendered:
                    feedback.append(
                        "Render FAILED.\nErrors:\n"
                        + "\n".join(obs.errors))
                else:
                    feedback.append(f"SSIM: {obs.ssim:.4f} (need >= 0.99).")
                    feedback.append(
                        "Below is your current render vs the reference. "
                        "Fix the differences. Output ONLY raw GLSL code.")

                content: list = [
                    {"type": "text", "text": "\n".join(feedback)}
                ]
                if obs.agent_png:
                    content.append({
                        "type": "image_url",
                        "image_url": {
                            "url": f"data:image/png;base64,{obs.agent_png}",
                        },
                    })
                conversation.append({"role": "user", "content": content})

            code = call_model(client, conversation) or FALLBACK_CODE

            # Step the environment
            result = await env.step(ShaderAction(code=code))
            obs = result.observation

            reward = result.reward or 0.0
            done = result.done
            error = "; ".join(obs.errors) if obs.errors else None

            rewards.append(reward)
            steps_taken = step
            best_ssim = max(best_ssim, obs.ssim)

            log_step(
                step=step, action=code, reward=reward,
                done=done, error=error,
            )

            conversation.append({"role": "assistant", "content": code})

            if done:
                break

        score = min(max(best_ssim, 0.01), 0.99)
        success = score >= SUCCESS_THRESHOLD

    finally:
        log_end(
            success=success, steps=steps_taken,
            score=score, rewards=rewards,
        )


# ---------------------------------------------------------------------------
# Main — run all 3 tasks
# ---------------------------------------------------------------------------

async def main() -> None:
    client = OpenAI(base_url=API_BASE_URL, api_key=API_KEY)

    env = await ShaderEnv.from_docker_image(IMAGE_NAME)

    try:
        for task_name in TASKS:
            await run_task(env, client, task_name)
    finally:
        try:
            await env.close()
        except Exception as e:
            print(f"[DEBUG] env.close() error: {e}", flush=True)


if __name__ == "__main__":
    asyncio.run(main())