swactor/apps/mvp-node/tinygrad_worker.py

629 lines
21 KiB
Python
Raw Normal View History

2026-07-01 08:44:25 +00:00
#!/usr/bin/env python3
from __future__ import annotations
import hashlib
import json
2026-07-05 09:59:51 +00:00
import linecache
2026-07-01 08:44:25 +00:00
import os
import sys
2026-07-05 09:59:51 +00:00
import threading
2026-07-01 08:44:25 +00:00
import time
import traceback
import urllib.parse
import urllib.request
from pathlib import Path
from typing import Any
Tensor: Any = None
dtypes: Any = None
model: Any = None
tokenizer: Any = None
role: dict[str, Any] = {}
loaded: dict[str, Any] = {}
2026-07-05 09:59:51 +00:00
class CpuLineSampler:
def __init__(
self,
*,
phase: str,
request_id: int | None,
model_id: str | None,
interval_secs: float,
) -> None:
self.phase = phase
self.request_id = request_id
self.model_id = model_id
self.interval_secs = interval_secs
self.target_thread_id = threading.get_ident()
self.samples: dict[tuple[str, int, str], int] = {}
self.wall_start = time.perf_counter()
self.process_cpu_start = time.process_time()
self._running = True
self._thread = threading.Thread(target=self._run, name="cpu-line-sampler", daemon=True)
self._thread.start()
def _run(self) -> None:
while self._running:
frame = sys._current_frames().get(self.target_thread_id)
if frame is not None:
code = frame.f_code
key = (code.co_filename, frame.f_lineno, code.co_name)
self.samples[key] = self.samples.get(key, 0) + 1
time.sleep(self.interval_secs)
def stop(self) -> None:
self._running = False
self._thread.join(timeout=max(0.25, self.interval_secs * 4.0))
wall_elapsed_ms = (time.perf_counter() - self.wall_start) * 1000.0
process_cpu_elapsed_ms = (time.process_time() - self.process_cpu_start) * 1000.0
total_samples = sum(self.samples.values())
top = []
for (filename, line, function), count in sorted(
self.samples.items(), key=lambda item: item[1], reverse=True
)[:32]:
top.append(
{
"file": filename,
"line": line,
"function": function,
"source": linecache.getline(filename, line).strip(),
"samples": count,
"percent": round((count * 100.0 / total_samples), 2) if total_samples else 0.0,
}
)
control(
type="CpuLineProfileSummary",
phase=self.phase,
request_id=self.request_id,
model_id=self.model_id,
interval_ms=round(self.interval_secs * 1000.0, 3),
wall_elapsed_ms=round(wall_elapsed_ms, 3),
process_cpu_elapsed_ms=round(process_cpu_elapsed_ms, 3),
process_cpu_over_wall=round(process_cpu_elapsed_ms / wall_elapsed_ms, 4)
if wall_elapsed_ms > 0.0
else 0.0,
total_samples=total_samples,
top=top,
)
def start_cpu_line_sampler(
*,
phase: str,
request_id: int | None,
model_id: str | None,
) -> CpuLineSampler | None:
raw = os.environ.get("MVP_CPU_LINE_PROFILE")
if not env_flag("MVP_CPU_LINE_PROFILE", False):
control(
type="CpuLineProfileSkipped",
phase=phase,
request_id=request_id,
model_id=model_id,
env_value=raw,
)
return None
interval_ms = float(os.environ.get("MVP_CPU_LINE_PROFILE_INTERVAL_MS", "2"))
interval_secs = max(0.0005, interval_ms / 1000.0)
control(
type="CpuLineProfileStarted",
phase=phase,
request_id=request_id,
model_id=model_id,
interval_ms=round(interval_secs * 1000.0, 3),
)
return CpuLineSampler(
phase=phase,
request_id=request_id,
model_id=model_id,
interval_secs=interval_secs,
)
def stop_cpu_line_sampler(sampler: CpuLineSampler | None) -> None:
if sampler is not None:
sampler.stop()
def env_flag(name: str, default: bool = True) -> bool:
raw = os.environ.get(name)
if raw is None:
return default
return raw.strip().lower() not in {"0", "false", "no", "off"}
2026-07-01 08:44:25 +00:00
def control(**event: Any) -> None:
print(json.dumps(event, separators=(",", ":")), flush=True)
def log(message: str) -> None:
print(f"mvp_tinygrad_worker: {message}", file=sys.stderr, flush=True)
def fatal(reason: str, **fields: Any) -> None:
control(type="WorkerFatal", reason=reason, **fields)
raise SystemExit(1)
def test_mode() -> bool:
return os.environ.get("MVP_TINYGRAD_TEST_MODE", "").strip().lower() in {"1", "true", "yes", "on"}
def initialize(cmd: dict[str, Any]) -> None:
global Tensor, dtypes
if int(cmd.get("helper_abi_version", 1)) != 1:
fatal("UnsupportedHelperAbi", helper_abi_version=cmd.get("helper_abi_version"))
device = str(cmd.get("backend", {}).get("device") or os.environ.get("DEV") or "CUDA")
os.environ["DEV"] = device
started = time.monotonic()
2026-07-05 09:59:51 +00:00
control(type="TinygradImportStarted", requested_device=device, env_DEV=os.environ.get("DEV"))
2026-07-01 08:44:25 +00:00
from tinygrad import Tensor as TinyTensor, dtypes as tiny_dtypes
2026-07-05 09:59:51 +00:00
control(type="TinygradImportReady", requested_device=device, env_DEV=os.environ.get("DEV"))
2026-07-01 08:44:25 +00:00
Tensor = TinyTensor
dtypes = tiny_dtypes
2026-07-05 09:59:51 +00:00
control(type="TinygradDeviceProbeStarted", requested_device=device)
2026-07-01 08:44:25 +00:00
value = Tensor([1], dtype=dtypes.int32).realize().numpy().tolist()
2026-07-05 09:59:51 +00:00
control(type="TinygradDeviceProbeReady", requested_device=device, probe_result=value)
2026-07-01 08:44:25 +00:00
control(
type="WorkerReady",
pid=os.getpid(),
2026-07-05 09:59:51 +00:00
backend={"requested_device": device, "env_DEV": os.environ.get("DEV"), "tinygrad_device": device},
2026-07-01 08:44:25 +00:00
cuda_probe=value,
elapsed_ms=int((time.monotonic() - started) * 1000),
)
def configure_role(cmd: dict[str, Any]) -> None:
config = cmd.get("config", {})
role.clear()
role.update(
role_id=int(cmd.get("role_id", 1)),
run_id=int(config.get("run_id", 1)),
stage_index=int(config.get("stage_index", 0)),
layer_start=int(config.get("layer_start", 0)),
layer_end_exclusive=int(config.get("layer_end_exclusive", 0)),
)
control(type="RoleConfigured", role_id=role["role_id"], stage_index=role["stage_index"])
def cache_root() -> Path:
raw = os.environ.get("MVP_MODEL_CACHE_DIR", "").strip()
root = Path(raw).expanduser() if raw else Path.home() / ".cache" / "mvp-node"
root.mkdir(parents=True, exist_ok=True)
return root
def hf_url(repo: str, file: str, revision: str | None) -> str:
encoded_file = "/".join(urllib.parse.quote(part) for part in file.split("/"))
return f"https://huggingface.co/{repo}/resolve/{revision or 'main'}/{encoded_file}"
def source_url(source: dict[str, Any]) -> str | None:
if "HuggingFaceGguf" not in source:
return None
hf = source["HuggingFaceGguf"]
return hf_url(str(hf["repo"]), str(hf["file"]), hf.get("revision"))
def source_path(source: dict[str, Any]) -> Path | None:
if "LocalPath" not in source:
return None
return Path(str(source["LocalPath"])).expanduser()
2026-07-05 09:59:51 +00:00
def source_kind(source: dict[str, Any]) -> str:
if "LocalPath" in source:
return "LocalPath"
if "HuggingFaceGguf" in source:
return "HuggingFaceGguf"
return "Unknown"
2026-07-01 08:44:25 +00:00
def cache_path_for(url: str) -> Path:
parsed = urllib.parse.urlparse(url)
basename = Path(parsed.path).name or "model.gguf"
digest = hashlib.sha256(url.encode("utf-8")).hexdigest()[:16]
return cache_root() / f"{digest}-{basename}"
def request_headers() -> dict[str, str]:
headers = {"User-Agent": "swactor-mvp-node/0.1"}
token = os.environ.get("HF_TOKEN", "").strip()
if token:
headers["Authorization"] = f"Bearer {token}"
return headers
def fetch_whole(source: dict[str, Any]) -> Path:
local = source_path(source)
if local is not None:
2026-07-05 09:59:51 +00:00
control(type="GgufLocalPathStatStarted", path=str(local))
2026-07-01 08:44:25 +00:00
if not local.is_file():
fatal("GgufLocalPathMissing", path=str(local))
2026-07-05 09:59:51 +00:00
stat = local.stat()
control(type="GgufCacheReady", path=str(local), bytes=stat.st_size, cache_hit=True, source="local")
2026-07-01 08:44:25 +00:00
return local
url = source_url(source)
if not url:
fatal("UnsupportedGgufSource", source=source)
target = cache_path_for(url)
if target.is_file() and target.stat().st_size > 0:
control(type="GgufCacheReady", path=str(target), bytes=target.stat().st_size, cache_hit=True, url=url)
return target
partial = target.with_name(target.name + ".partial")
started = time.monotonic()
req = urllib.request.Request(url, headers=request_headers())
control(type="GgufDownloadStarted", url=url, path=str(target))
try:
with urllib.request.urlopen(req, timeout=60) as response, partial.open("wb") as out:
total = int(response.headers.get("Content-Length") or 0)
done = 0
last_event = 0.0
while True:
chunk = response.read(1024 * 1024)
if not chunk:
break
out.write(chunk)
done += len(chunk)
now = time.monotonic()
if now - last_event >= float(os.environ.get("MVP_DOWNLOAD_PROGRESS_SECS", "5")):
control(
type="GgufDownloadProgress",
bytes_done=done,
bytes_total=total,
elapsed_ms=int((now - started) * 1000),
)
last_event = now
partial.replace(target)
except Exception as exc:
try:
partial.unlink(missing_ok=True)
except Exception:
pass
fatal("GgufDownloadFailed", url=url, error=str(exc))
control(
type="GgufCacheReady",
path=str(target),
bytes=target.stat().st_size,
cache_hit=False,
elapsed_ms=int((time.monotonic() - started) * 1000),
url=url,
)
return target
def require_tinygrad() -> Any:
if Tensor is None:
fatal("BackendNotInitialized")
return Tensor
def load_weights(cmd: dict[str, Any]) -> None:
global model, tokenizer
TensorCls = require_tinygrad()
started = time.monotonic()
model_id = str(cmd["model_id"])
2026-07-05 09:59:51 +00:00
source = cmd["gguf_source"]
control(
type="LoadWeightsStarted",
model_id=model_id,
source_kind=source_kind(source),
layer_start=int(cmd.get("layer_start", 0)),
layer_end_exclusive=int(cmd.get("layer_end_exclusive", 0)),
)
2026-07-01 08:44:25 +00:00
if test_mode():
model = {"test_mode": True}
tokenizer = {"test_mode": True}
loaded.clear()
loaded.update(
model_id=model_id,
path="mvp-tinygrad-test-mode",
layer_start=int(cmd.get("layer_start", 0)),
layer_end_exclusive=int(cmd.get("layer_end_exclusive", 0)),
)
control(
type="WeightsLoaded",
model_id=model_id,
path=loaded["path"],
test_mode=True,
elapsed_ms=int((time.monotonic() - started) * 1000),
)
return
2026-07-05 09:59:51 +00:00
control(type="GgufResolveStarted", model_id=model_id, source_kind=source_kind(source))
2026-07-01 08:44:25 +00:00
path = fetch_whole(source)
2026-07-05 09:59:51 +00:00
model_bytes = path.stat().st_size
control(type="GgufResolveReady", model_id=model_id, path=str(path), bytes=model_bytes)
2026-07-01 08:44:25 +00:00
try:
2026-07-05 09:59:51 +00:00
control(type="TinygradLlmImportStarted", model_id=model_id)
2026-07-01 08:44:25 +00:00
from tinygrad.apps.llm import SimpleTokenizer, Transformer
2026-07-05 09:59:51 +00:00
control(type="TinygradLlmImportReady", model_id=model_id)
2026-07-01 08:44:25 +00:00
max_context_raw = os.environ.get("MVP_MAX_CONTEXT", "512")
max_context = int(max_context_raw) if max_context_raw else 512
2026-07-05 09:59:51 +00:00
control(
type="TransformerFromGgufStarted",
model_id=model_id,
path=str(path),
bytes=model_bytes,
max_context=max_context,
realize=True,
requested_device=os.environ.get("DEV"),
)
2026-07-01 08:44:25 +00:00
model, kv = Transformer.from_gguf(TensorCls(path), max_context=max_context, realize=True)
2026-07-05 09:59:51 +00:00
control(
type="TransformerFromGgufReady",
model_id=model_id,
path=str(path),
bytes=model_bytes,
max_context=max_context,
realize=True,
requested_device=os.environ.get("DEV"),
)
2026-07-01 08:44:25 +00:00
tok_src = cmd.get("tokenizer", {"EmbeddedGguf": None})
if "EmbeddedGguf" in tok_src:
2026-07-05 09:59:51 +00:00
control(type="TokenizerBuildStarted", model_id=model_id, source="EmbeddedGguf")
2026-07-01 08:44:25 +00:00
tokenizer = SimpleTokenizer.from_gguf_kv(kv)
2026-07-05 09:59:51 +00:00
control(type="TokenizerBuildReady", model_id=model_id, source="EmbeddedGguf")
2026-07-01 08:44:25 +00:00
else:
fatal("UnsupportedTokenizerSource", tokenizer=tok_src)
except SystemExit:
raise
except Exception as exc:
tb = traceback.format_exc()
print(tb, file=sys.stderr, flush=True)
fatal("ModelLoadFailed", error=str(exc), traceback=tb)
loaded.clear()
loaded.update(
model_id=model_id,
path=str(path),
layer_start=int(cmd.get("layer_start", 0)),
layer_end_exclusive=int(cmd.get("layer_end_exclusive", 0)),
)
control(
type="WeightsLoaded",
model_id=model_id,
path=str(path),
elapsed_ms=int((time.monotonic() - started) * 1000),
)
2026-07-05 09:59:51 +00:00
2026-07-07 10:40:02 +00:00
def prompt_template_name() -> str:
return os.environ.get("MVP_PROMPT_TEMPLATE", "llama3-chat").strip().lower()
def model_prompt_text(prompt: str) -> tuple[str, str]:
template = prompt_template_name()
if template in {"", "raw", "none", "off", "false", "0"}:
return prompt, "raw"
if template in {"llama3", "llama3-chat", "llama-3", "llama-3-chat"}:
return (
"<|begin_of_text|>"
"<|start_header_id|>user<|end_header_id|>\n\n"
f"{prompt}"
"<|eot_id|>"
"<|start_header_id|>assistant<|end_header_id|>\n\n",
"llama3-chat",
)
return prompt, "raw"
def strip_chat_stop_markers(text: str) -> str:
cut = len(text)
for marker in ("<|eot_id|>", "<|end_of_text|>", "<|start_header_id|>"):
index = text.find(marker)
if index >= 0:
cut = min(cut, index)
return text[:cut].rstrip()
2026-07-05 09:59:51 +00:00
def decode_greedy_device_resident(
prompt_tokens: list[int],
max_tokens: int,
*,
request_id: int | None,
model_id: str | None,
progress_every: int,
) -> list[int]:
if max_tokens <= 0:
return []
max_context = int(getattr(model, "max_context", len(prompt_tokens) + max_tokens))
generation_limit = min(max_tokens, max(0, max_context - len(prompt_tokens)))
if generation_limit <= 0:
control(
type="DecodeContextFull",
request_id=request_id,
model_id=model_id,
prompt_tokens=len(prompt_tokens),
max_context=max_context,
)
return []
if generation_limit < max_tokens:
control(
type="DecodeLimitedByContext",
request_id=request_id,
model_id=model_id,
prompt_tokens=len(prompt_tokens),
requested_tokens=max_tokens,
generation_limit=generation_limit,
max_context=max_context,
)
TensorCls = require_tinygrad()
from tinygrad.uop.ops import UOp
if hasattr(model, "forward_jit"):
model.forward_jit.reset()
use_symbolic_pos = os.environ.get("SYM", "1").strip().lower() not in {"0", "false", "no", "off"}
pos_upper_bound = max(1, max_context - 1)
symbolic_start_pos = UOp.variable("start_pos", 1, pos_upper_bound)
next_token = model(TensorCls([prompt_tokens], dtype="int32"), 0).realize()
generated_tensors = []
for token_index in range(generation_limit):
generated_tensors.append(next_token.clone().realize())
tokens_generated = token_index + 1
if tokens_generated == 1:
control(
type="FirstTokenReady",
request_id=request_id,
model_id=model_id,
token_index=1,
prompt_tokens=len(prompt_tokens),
)
elif progress_every > 0 and tokens_generated % progress_every == 0:
control(
type="TokenProgress",
request_id=request_id,
model_id=model_id,
tokens_generated=tokens_generated,
prompt_tokens=len(prompt_tokens),
)
if tokens_generated >= generation_limit:
break
start_pos = len(prompt_tokens) + token_index
pos = symbolic_start_pos.bind(start_pos) if use_symbolic_pos else start_pos
next_token = model(next_token, pos).realize()
generated_tensor = (
generated_tensors[0]
if len(generated_tensors) == 1
else generated_tensors[0].cat(*generated_tensors[1:], dim=1)
)
generated_array = generated_tensor.numpy().reshape(-1).tolist()
return [int(token) for token in generated_array]
2026-07-01 08:44:25 +00:00
def infer_prompt(cmd: dict[str, Any]) -> None:
if model is None or tokenizer is None:
fatal("WeightsNotLoaded")
prompt = str(cmd.get("prompt", ""))
max_tokens = int(cmd.get("max_tokens", 1))
2026-07-05 09:59:51 +00:00
request_id_raw = cmd.get("request_id")
request_id = int(request_id_raw) if request_id_raw is not None else None
2026-07-01 08:44:25 +00:00
started = time.monotonic()
2026-07-05 09:59:51 +00:00
control(
type="PromptStarted",
request_id=request_id,
model_id=loaded.get("model_id"),
prompt_bytes=len(prompt.encode("utf-8")),
prompt_chars=len(prompt),
max_tokens=max_tokens,
)
2026-07-01 08:44:25 +00:00
if test_mode():
text = f"mvp-test response: {prompt}"
control(
type="PromptCompleted",
2026-07-05 09:59:51 +00:00
request_id=request_id,
2026-07-01 08:44:25 +00:00
model_id=loaded.get("model_id"),
prompt_tokens=[],
generated_tokens=list(range(min(max_tokens, 3))),
text=text,
test_mode=True,
elapsed_ms=int((time.monotonic() - started) * 1000),
)
return
2026-07-07 10:40:02 +00:00
model_prompt, prompt_template = model_prompt_text(prompt)
control(
type="PromptEncodeStarted",
request_id=request_id,
model_id=loaded.get("model_id"),
prompt_template=prompt_template,
)
prompt_tokens = tokenizer.encode(model_prompt)
2026-07-05 09:59:51 +00:00
control(
type="PromptEncodeReady",
request_id=request_id,
model_id=loaded.get("model_id"),
prompt_bytes=len(prompt.encode("utf-8")),
2026-07-07 10:40:02 +00:00
model_prompt_bytes=len(model_prompt.encode("utf-8")),
prompt_template=prompt_template,
2026-07-05 09:59:51 +00:00
prompt_tokens=len(prompt_tokens),
)
progress_every = int(os.environ.get("MVP_TOKEN_PROGRESS_EVERY", "16") or "16")
control(
type="DecodeStarted",
request_id=request_id,
model_id=loaded.get("model_id"),
prompt_tokens=len(prompt_tokens),
max_tokens=max_tokens,
decode_impl="device_resident_greedy",
)
cpu_sampler = start_cpu_line_sampler(
phase="decode",
request_id=request_id,
model_id=loaded.get("model_id"),
)
try:
generated = decode_greedy_device_resident(
prompt_tokens,
max_tokens,
request_id=request_id,
model_id=loaded.get("model_id"),
progress_every=progress_every,
)
finally:
stop_cpu_line_sampler(cpu_sampler)
control(
type="DecodeReady",
request_id=request_id,
model_id=loaded.get("model_id"),
prompt_tokens=len(prompt_tokens),
tokens_generated=len(generated),
)
control(type="TextDecodeStarted", request_id=request_id, model_id=loaded.get("model_id"), tokens_generated=len(generated))
2026-07-07 10:40:02 +00:00
raw_text = tokenizer.decode(generated) if generated else ""
text = strip_chat_stop_markers(raw_text)
2026-07-05 09:59:51 +00:00
control(
type="TextDecodeReady",
request_id=request_id,
model_id=loaded.get("model_id"),
tokens_generated=len(generated),
text_bytes=len(text.encode("utf-8")),
)
2026-07-01 08:44:25 +00:00
control(
type="PromptCompleted",
2026-07-05 09:59:51 +00:00
request_id=request_id,
2026-07-01 08:44:25 +00:00
model_id=loaded.get("model_id"),
prompt_tokens=prompt_tokens,
generated_tokens=generated,
text=text,
elapsed_ms=int((time.monotonic() - started) * 1000),
)
def shutdown_worker(_: dict[str, Any]) -> None:
control(type="WorkerStopped", reason="Graceful")
raise SystemExit(0)
HANDLERS = {
"InitializeWorker": initialize,
"ConfigureRole": configure_role,
"LoadWeights": load_weights,
"InferPrompt": infer_prompt,
"ShutdownWorker": shutdown_worker,
}
for raw in sys.stdin:
if not raw.strip():
continue
try:
command = json.loads(raw)
handler = HANDLERS.get(command.get("type"))
if handler is None:
fatal("UnknownCommand", command=command.get("type"))
handler(command)
except SystemExit:
raise
except Exception as exc:
tb = traceback.format_exc()
print(tb, file=sys.stderr, flush=True)
fatal("UnhandledWorkerException", error=str(exc), traceback=tb)