mirror of
https://github.com/karpathy/nanochat.git
synced 2026-05-07 16:30:11 +00:00
544ab89c04
Training data taught the model to echo another <|output_start|>…<|output_end|> after our injected real tool result. Detect that second sequence and break the turn; the grounded answer has already streamed to the client.
392 lines
17 KiB
Python
392 lines
17 KiB
Python
"""
|
|
samosaChaat — Modal GPU inference endpoint.
|
|
|
|
Downloads nanochat model weights from HuggingFace into a Modal Volume,
|
|
loads them on a GPU, and exposes an SSE streaming endpoint compatible
|
|
with the samosaChaat chat-api service.
|
|
|
|
Deploy: modal deploy modal/serve.py
|
|
Dev: modal serve modal/serve.py
|
|
"""
|
|
from __future__ import annotations
|
|
|
|
import json
|
|
import os
|
|
import time
|
|
|
|
import modal
|
|
|
|
# ---------------------------------------------------------------------------
|
|
# Configuration
|
|
# ---------------------------------------------------------------------------
|
|
MODEL_REPO = "ManmohanSharma/nanochat-d24"
|
|
MODEL_PT = "chatsft_checkpoints/d24-sft-r6/model_000754.pt"
|
|
META_JSON = "chatsft_checkpoints/d24-sft-r6/meta_000754.json"
|
|
TOKENIZER_PKL = "tokenizer/tokenizer.pkl"
|
|
TOKEN_BYTES = "tokenizer/token_bytes.pt"
|
|
MODEL_TAG = "d24-sft-r6"
|
|
GPU_TYPE = "L4" # 24 GB VRAM — fits 4 GB bf16 model loaded as fp32
|
|
VOLUME_NAME = "samosachaat-weights"
|
|
HF_SECRET_NAME = "huggingface" # Modal secret containing HF_TOKEN
|
|
TAVILY_SECRET_NAME = "tavily" # Modal secret containing TAVILY_API_KEY
|
|
|
|
# ---------------------------------------------------------------------------
|
|
# Modal app + image
|
|
# ---------------------------------------------------------------------------
|
|
app = modal.App("samosachaat-inference")
|
|
|
|
# Build the container image with all dependencies
|
|
inference_image = (
|
|
modal.Image.debian_slim(python_version="3.12")
|
|
.pip_install(
|
|
"torch==2.5.1",
|
|
"tiktoken>=0.11.0",
|
|
"tokenizers>=0.22.0",
|
|
"huggingface_hub>=0.25.0",
|
|
"requests>=2.31.0",
|
|
"fastapi>=0.115.0",
|
|
"uvicorn>=0.30.0",
|
|
extra_index_url="https://download.pytorch.org/whl/cu124",
|
|
)
|
|
.add_local_file("modal/_model.py", "/root/_model.py")
|
|
.add_local_file("modal/_tokenizer.py", "/root/_tokenizer.py")
|
|
.add_local_file("modal/_tools.py", "/root/_tools.py")
|
|
)
|
|
|
|
# Persistent volume for model weights
|
|
volume = modal.Volume.from_name(VOLUME_NAME, create_if_missing=True)
|
|
|
|
# ---------------------------------------------------------------------------
|
|
# Download weights into the volume (runs once)
|
|
# ---------------------------------------------------------------------------
|
|
@app.function(
|
|
image=inference_image,
|
|
volumes={"/weights": volume},
|
|
secrets=[modal.Secret.from_name(HF_SECRET_NAME)],
|
|
timeout=1800,
|
|
)
|
|
def download_weights():
|
|
"""Download model weights from HuggingFace into the Modal volume."""
|
|
import shutil
|
|
from huggingface_hub import hf_hub_download
|
|
|
|
model_dir = f"/weights/{MODEL_TAG}"
|
|
os.makedirs(model_dir, exist_ok=True)
|
|
|
|
token = os.environ.get("HF_TOKEN")
|
|
|
|
# (HF source path, local filename in volume)
|
|
files = [
|
|
(MODEL_PT, "model.pt"),
|
|
(META_JSON, "meta.json"),
|
|
(TOKENIZER_PKL, "tokenizer.pkl"),
|
|
(TOKEN_BYTES, "token_bytes.pt"),
|
|
]
|
|
|
|
for src, local_name in files:
|
|
dest = os.path.join(model_dir, local_name)
|
|
if os.path.exists(dest):
|
|
print(f" Already exists: {dest}")
|
|
continue
|
|
print(f" Downloading {src} from {MODEL_REPO}...")
|
|
path = hf_hub_download(MODEL_REPO, src, token=token)
|
|
shutil.copy2(path, dest)
|
|
print(f" Saved to {dest}")
|
|
|
|
volume.commit()
|
|
print("Weights downloaded and committed to volume.")
|
|
|
|
|
|
# ---------------------------------------------------------------------------
|
|
# Inference class — GPU singleton
|
|
# ---------------------------------------------------------------------------
|
|
@app.cls(
|
|
image=inference_image,
|
|
volumes={"/weights": volume},
|
|
gpu=GPU_TYPE,
|
|
scaledown_window=300, # keep warm for 5 min after last request
|
|
# concurrency handled by @modal.concurrent below
|
|
timeout=120,
|
|
secrets=[modal.Secret.from_name(TAVILY_SECRET_NAME)],
|
|
)
|
|
class Inference:
|
|
model: object
|
|
tokenizer: object
|
|
engine: object
|
|
device: object
|
|
|
|
@modal.enter()
|
|
def load_model(self):
|
|
"""Called once when the container starts — loads model onto GPU."""
|
|
import torch
|
|
import sys
|
|
|
|
# Add the nanochat engine code path
|
|
# We inline the minimal loading logic here to avoid importing the full
|
|
# nanochat package (which has heavy deps we don't need on Modal).
|
|
print("Loading model...")
|
|
t0 = time.time()
|
|
|
|
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
|
|
self.device = device
|
|
|
|
model_dir = f"/weights/{MODEL_TAG}"
|
|
meta_path = os.path.join(model_dir, "meta.json")
|
|
model_path = os.path.join(model_dir, "model.pt")
|
|
|
|
# Load meta
|
|
with open(meta_path) as f:
|
|
meta = json.load(f)
|
|
model_config = meta if "model_config" not in meta else meta["model_config"]
|
|
|
|
# Normalize config key names (HF format → nanochat format)
|
|
# Map HF config keys → nanochat GPTConfig keys
|
|
seq_len = model_config.pop("n_positions", None) or model_config.pop("n_ctx", None)
|
|
if seq_len and "sequence_len" not in model_config:
|
|
model_config["sequence_len"] = seq_len
|
|
# Also remove n_ctx if sequence_len was already set
|
|
model_config.pop("n_ctx", None)
|
|
model_config.pop("n_positions", None)
|
|
# Remove HF-specific keys that GPTConfig doesn't accept
|
|
for k in ["architectures", "model_type", "rotary", "rotary_base", "tie_word_embeddings"]:
|
|
model_config.pop(k, None)
|
|
|
|
# Patch missing config keys
|
|
model_config.setdefault("window_pattern", "L")
|
|
|
|
print(f" Config: {model_config}")
|
|
|
|
# Build model
|
|
# We need the GPT class — download it from the repo itself
|
|
# For simplicity, we define a minimal inline version that matches nanochat
|
|
from _model import GPT, GPTConfig
|
|
|
|
config = GPTConfig(**model_config)
|
|
model_data = torch.load(model_path, map_location=device, weights_only=False)
|
|
|
|
# Strip torch.compile prefix
|
|
model_data = {k.removeprefix("_orig_mod."): v for k, v in model_data.items()}
|
|
|
|
# Convert bfloat16 weights to float32 for compatibility on non-Hopper GPUs
|
|
model_data = {
|
|
k: v.float() if v.dtype == torch.bfloat16 else v
|
|
for k, v in model_data.items()
|
|
}
|
|
|
|
model = GPT.from_state_dict(config, model_data)
|
|
model.load_state_dict(model_data, strict=True, assign=True)
|
|
model.to(device)
|
|
model.init_rotary(device=device, dtype=torch.float32)
|
|
model.eval()
|
|
|
|
self.model = model
|
|
self.config = config
|
|
|
|
# Load tokenizer
|
|
from _tokenizer import get_tokenizer, SPECIAL_TOKENS
|
|
self.tokenizer = get_tokenizer(model_dir)
|
|
|
|
# Resolve actual special-token IDs (nanochat appends specials at end of vocab)
|
|
self.special_token_ids = set()
|
|
for name in SPECIAL_TOKENS:
|
|
ids = self.tokenizer.encode_special(name)
|
|
self.special_token_ids.update(ids)
|
|
self.assistant_end_id = self.tokenizer.encode_special("<|assistant_end|>")[0]
|
|
print(f" Special token IDs: {sorted(self.special_token_ids)}")
|
|
|
|
# Initialize tool registry (Tavily web_search + calculator)
|
|
import sys as _sys
|
|
if '/root' not in _sys.path: _sys.path.insert(0, '/root')
|
|
from _tools import build_default_tool_registry, parse_tool_call_payload
|
|
self.tool_registry = build_default_tool_registry()
|
|
self._parse_tool_call = parse_tool_call_payload
|
|
# Marker tokens for tool state machine
|
|
self.python_start_id = self.tokenizer.encode_special("<|python_start|>")[0]
|
|
self.python_end_id = self.tokenizer.encode_special("<|python_end|>")[0]
|
|
self.output_start_id = self.tokenizer.encode_special("<|output_start|>")[0]
|
|
self.output_end_id = self.tokenizer.encode_special("<|output_end|>")[0]
|
|
# Stop tokens (exclude tool markers so generation continues through tool calls)
|
|
self._stop_token_ids = {self.assistant_end_id, self.tokenizer.get_bos_token_id() if hasattr(self.tokenizer, "get_bos_token_id") else self.tokenizer.encode_special("<|bos|>")[0]}
|
|
|
|
dt = time.time() - t0
|
|
print(f"Model loaded in {dt:.1f}s on {device} | tools: {[t for t in self.tool_registry._tools.keys()] if hasattr(self.tool_registry, '_tools') else 'registered'}")
|
|
|
|
@modal.fastapi_endpoint(method="POST", docs=True)
|
|
async def generate(self, request: dict):
|
|
"""
|
|
Streaming chat endpoint — SSE compatible with samosaChaat format.
|
|
|
|
Input: {"messages": [{"role": "user", "content": "..."}], "temperature": 0.8, "max_tokens": 512, "top_k": 50}
|
|
Output: SSE stream of data: {"token": "...", "gpu": 0} then data: {"done": true}
|
|
"""
|
|
import torch
|
|
from fastapi.responses import StreamingResponse
|
|
|
|
messages = request.get("messages", [])
|
|
temperature = min(max(request.get("temperature", 0.8), 0.0), 2.0)
|
|
max_tokens = min(max(request.get("max_tokens", 512), 1), 2048)
|
|
top_k = min(max(request.get("top_k", 50), 0), 200)
|
|
|
|
# Build token sequence from messages
|
|
tokens = []
|
|
bos = self.tokenizer.encode_special("<|bos|>")
|
|
user_start = self.tokenizer.encode_special("<|user_start|>")
|
|
user_end = self.tokenizer.encode_special("<|user_end|>")
|
|
assistant_start = self.tokenizer.encode_special("<|assistant_start|>")
|
|
assistant_end = self.tokenizer.encode_special("<|assistant_end|>")
|
|
|
|
tokens.extend(bos)
|
|
for msg in messages:
|
|
if msg["role"] == "user":
|
|
tokens.extend(user_start)
|
|
tokens.extend(self.tokenizer.encode(msg["content"]))
|
|
tokens.extend(user_end)
|
|
elif msg["role"] == "assistant":
|
|
tokens.extend(assistant_start)
|
|
tokens.extend(self.tokenizer.encode(msg["content"]))
|
|
tokens.extend(assistant_end)
|
|
# Prompt the model to generate an assistant response
|
|
tokens.extend(assistant_start)
|
|
|
|
# Truncate to fit context
|
|
max_context = self.config.sequence_len - max_tokens
|
|
if len(tokens) > max_context:
|
|
tokens = tokens[-max_context:]
|
|
|
|
# Ordinary-text token-id sequences for the tool markers.
|
|
# The SFT loader tokenizes assistant content with .encode() (not .encode_special()),
|
|
# so the model emits these as multi-token sequences, not single special-token ids.
|
|
# Match on the id sequence directly — more reliable than text (BPE partial UTF-8
|
|
# can make single-token decodes return empty strings).
|
|
tool_start_ids = tuple(self.tokenizer.encode("<|python_start|>"))
|
|
tool_end_ids = tuple(self.tokenizer.encode("<|python_end|>"))
|
|
out_start_str = "<|output_start|>"
|
|
out_end_str = "<|output_end|>"
|
|
|
|
# Suppression: model training has many convs that emit a fake
|
|
# <|output_start|>…<|output_end|> after our injected one. We stop the
|
|
# turn if we detect another full <|output_start|> sequence emitted
|
|
# after our injection.
|
|
out_start_ids = tuple(self.tokenizer.encode(out_start_str))
|
|
|
|
async def stream():
|
|
input_ids = torch.tensor([tokens], dtype=torch.long, device=self.device)
|
|
gen_ids: list[int] = [] # everything the MODEL sampled this turn
|
|
tool_start_idx = -1 # position in gen_ids where <|python_start|> begins
|
|
tool_injected = False # once True, stop detecting further tool calls
|
|
injection_end_pos = -1 # index in gen_ids where our injected tokens end
|
|
|
|
def _append_token(tid):
|
|
nonlocal input_ids
|
|
nt = torch.tensor([[tid]], dtype=torch.long, device=self.device)
|
|
input_ids = torch.cat([input_ids, nt], dim=1)
|
|
if input_ids.size(1) > self.config.sequence_len:
|
|
input_ids = input_ids[:, -self.config.sequence_len:]
|
|
|
|
def _match_at(seq: list[int], pos: int, pat: tuple) -> bool:
|
|
if pos < 0 or pos + len(pat) > len(seq):
|
|
return False
|
|
return tuple(seq[pos:pos + len(pat)]) == pat
|
|
|
|
def _find_subseq(seq: list[int], pat: tuple, start: int = 0) -> int:
|
|
L = len(pat)
|
|
if L == 0 or len(seq) < start + L:
|
|
return -1
|
|
for i in range(start, len(seq) - L + 1):
|
|
if tuple(seq[i:i + L]) == pat:
|
|
return i
|
|
return -1
|
|
|
|
with torch.no_grad():
|
|
num_generated = 0
|
|
while num_generated < max_tokens:
|
|
logits = self.model(input_ids)
|
|
next_logits = logits[:, -1, :]
|
|
if temperature > 0:
|
|
next_logits = next_logits / temperature
|
|
if top_k > 0:
|
|
v, _ = torch.topk(next_logits, min(top_k, next_logits.size(-1)))
|
|
next_logits[next_logits < v[:, [-1]]] = float('-inf')
|
|
probs = torch.softmax(next_logits, dim=-1)
|
|
next_token = torch.multinomial(probs, num_samples=1)
|
|
token_id = next_token.item()
|
|
|
|
if token_id in self._stop_token_ids:
|
|
break
|
|
|
|
# Commit to context + sequence
|
|
_append_token(token_id)
|
|
gen_ids.append(token_id)
|
|
num_generated += 1
|
|
|
|
# Stream raw decoded text (may be empty for partial BPE bytes — that's OK)
|
|
try:
|
|
token_text = self.tokenizer.decode([token_id])
|
|
except Exception:
|
|
token_text = ""
|
|
if token_text:
|
|
yield "data: " + json.dumps({"token": token_text, "gpu": 0}) + "\n\n"
|
|
|
|
# --- tool-call detection (id-sequence match) ---
|
|
if not tool_injected and tool_start_idx < 0:
|
|
idx = _find_subseq(gen_ids, tool_start_ids, max(0, len(gen_ids) - len(tool_start_ids) - 2))
|
|
if idx >= 0:
|
|
tool_start_idx = idx
|
|
if not tool_injected and tool_start_idx >= 0:
|
|
# look for <|python_end|> after the payload
|
|
end_idx = _find_subseq(gen_ids, tool_end_ids, tool_start_idx + len(tool_start_ids))
|
|
if end_idx >= 0:
|
|
payload_ids = gen_ids[tool_start_idx + len(tool_start_ids):end_idx]
|
|
try:
|
|
payload_text = self.tokenizer.decode(payload_ids)
|
|
invocation = self._parse_tool_call(payload_text)
|
|
result = self.tool_registry.execute(invocation.tool_name, invocation.arguments)
|
|
result_text = result.to_payload()[:4096]
|
|
except Exception as exc:
|
|
result_text = json.dumps({"error": str(exc)[:500]})
|
|
|
|
wrapped = out_start_str + result_text + out_end_str
|
|
# Inject real result tokens into the model's context and the client stream.
|
|
for rid in self.tokenizer.encode(wrapped):
|
|
try:
|
|
rt = self.tokenizer.decode([rid])
|
|
except Exception:
|
|
rt = ""
|
|
if rt:
|
|
yield "data: " + json.dumps({"token": rt, "gpu": 0}) + "\n\n"
|
|
_append_token(rid)
|
|
gen_ids.append(rid)
|
|
num_generated += 1
|
|
if num_generated >= max_tokens:
|
|
break
|
|
tool_injected = True
|
|
injection_end_pos = len(gen_ids)
|
|
tool_start_idx = -1
|
|
|
|
# After injection, detect if the model emits ANOTHER full
|
|
# <|output_start|> sequence (training-data loop artifact) and stop the turn.
|
|
if tool_injected and injection_end_pos > 0:
|
|
if _find_subseq(gen_ids, out_start_ids, injection_end_pos) >= 0:
|
|
break
|
|
|
|
yield "data: " + json.dumps({"done": True}) + "\n\n"
|
|
|
|
return StreamingResponse(
|
|
stream(),
|
|
media_type="text/event-stream",
|
|
headers={
|
|
"Cache-Control": "no-cache",
|
|
"Connection": "keep-alive",
|
|
"X-Accel-Buffering": "no",
|
|
},
|
|
)
|
|
|
|
@modal.fastapi_endpoint(method="GET", docs=True)
|
|
def health(self):
|
|
return {
|
|
"status": "ok",
|
|
"model": MODEL_TAG,
|
|
"gpu": GPU_TYPE,
|
|
"ready": self.model is not None,
|
|
}
|