Files
nvfp4-megamoe-kernel/single_shot_inference.py

806 lines
34 KiB
Python
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
#!/usr/bin/env python3
"""Single-shot DSV4-Pro inference — Full 61-layer pipeline, 8-GPU.
This is a reference implementation that exercises the production kernel
stack end-to-end. It should be usable as ground truth when integrating
into vLLM or SGLang.
Architecture (paper §2):
X_l → mHC.pre_block → RMSNorm → Attention → F_attn → mHC.post_block → X_mid
X_mid → mHC.pre_block → RMSNorm → FFN(MoE) → F_ffn → mHC.post_block → X_{l+1}
Components exercised:
- mHC (Manifold-Constrained Hyper-Connections) — proper Sinkhorn-Knopp
- Low-rank Q projection (q_a → q_b) + KV projection (MQA, 1 KV head)
- Partial RoPE (last 64 dims, GPT-J interleaved)
- Production FMHA kernel (6-warp multi-tile, C API + ctypes)
- Inverse RoPE on attention output (paper §2.3.3)
- Grouped output projection (wo_a BMM + wo_b NVFP4)
- Routed MoE (384 experts, top-6, hash + dense routing, SwiGLU clamp)
- Shared expert (NVFP4 gate/up/down)
- RMSNorm (pre-norm before each sub-block)
- KV cache across decode steps
Attention type simplification for this single-shot test:
For short sequences (seq_len ≤ sliding_window=128), ALL attention
types (CSA/HCA/SWA) reduce to dense attention over the full KV cache.
CSA's compressed branch and indexer are only needed for long sequences
where seq_len > sliding_window. HCA is dense over compressed entries,
but at short sequence lengths, the compressed sequence is trivially
small. So we use dense MQA attention over the full KV for all layers.
This is mathematically correct for short sequences and exercises the
FMHA kernel properly.
Usage (on B200):
source /root/dsv4-nvfp4-workspace/venv/bin/activate
cd /root/dsv4-nvfp4-workspace/kernel
python3 single_shot_inference.py
"""
import os, sys, time, json, math
import torch
from pathlib import Path
# =====================================================================
# Configuration
# =====================================================================
CHECKPOINT_DIR = "/root/nvidia-meeting/DeepSeek-V4-Pro-NVFP4"
MAX_NEW_TOKENS = 10
PROMPT = "The capital of France is"
NUM_GPUS = 8
# =====================================================================
# NVFP4 dequantization — matches checkpoint format exactly
# =====================================================================
FP4_LUT = torch.tensor([0., 0.5, 1.0, 1.5, 2.0, 3.0, 4.0, 6.0]) # E2M1 magnitudes
def dequant_nvfp4_weight(weight, weight_scale, weight_scale_2):
"""Dequantize NVFP4 weight to BF16.
weight: (out_dim, in_dim//2) uint8 — 2 FP4 values per byte
weight_scale: (out_dim, in_dim//16) E4M3 — per-16-element block scale
weight_scale_2: (out_dim, 1) float32 — per-row global scale
"""
out_dim = weight.shape[0]
in_packed = weight.shape[1]
in_features = in_packed * 2
low = (weight & 0x0F).to(torch.int8)
high = (weight >> 4).to(torch.int8)
low_sign, low_idx = (low >> 3).bool(), (low & 0x07).long()
high_sign, high_idx = (high >> 3).bool(), (high & 0x07).long()
lut = FP4_LUT.to(device=weight.device, dtype=torch.float32)
low_f = lut[low_idx] * torch.where(low_sign, -1.0, 1.0)
high_f = lut[high_idx] * torch.where(high_sign, -1.0, 1.0)
w_f = torch.stack([low_f, high_f], dim=-1).reshape(out_dim, in_features)
scale_f = weight_scale.float() * weight_scale_2.float()
scale_expanded = scale_f.repeat_interleave(16, dim=1)
return (w_f * scale_expanded).bfloat16()
def nvfp4_linear(x, weight, weight_scale, weight_scale_2):
"""BF16 linear with NVFP4 dequant."""
w = dequant_nvfp4_weight(weight, weight_scale, weight_scale_2)
return torch.nn.functional.linear(x, w)
# =====================================================================
# RMSNorm — matches dsv4/layers/norm.py
# =====================================================================
class RMSNorm:
def __init__(self, hidden_size, eps=1e-6, device='cuda:0'):
self.eps = eps
self.weight = torch.ones(hidden_size, dtype=torch.float32, device=device)
def forward(self, x):
"""x: (T, H) BF16 → (T, H) BF16"""
x_f = x.float()
rms = x_f.pow(2).mean(dim=-1, keepdim=True).add(self.eps).rsqrt()
return (x_f * rms * self.weight).to(torch.bfloat16)
# =====================================================================
# mHC — proper Sinkhorn-Knopp implementation
# =====================================================================
class mHCBlock:
"""Wrapper around dsv4.layers.mhc.mHCLayer for single-shot inference.
Uses the production mHCLayer implementation with proper Sinkhorn-Knopp.
"""
def __init__(self, hidden_dim=7168, n_hc=4, sinkhorn_iters=20, device='cuda:0'):
from dsv4.layers.mhc import mHCLayer
self._impl = mHCLayer(
hidden_dim=hidden_dim, n_hc=n_hc,
t_max_sinkhorn=sinkhorn_iters,
device=device, dtype=torch.bfloat16)
self.device = device
self.n_hc = n_hc
self.hidden_dim = hidden_dim
def load_from_checkpoint(self, fn, base, scale):
"""Load from checkpoint tensors.
fn: (24, 28672) FP32 — fused projection
base: (24,) — [pre(4), post(4), res(16)]
scale: (3,) — [alpha_pre, alpha_post, alpha_res]
"""
n = self.n_hc
dev = self.device
# fn rows: [W_pre(4), W_post(4), W_res(16)]
W_pre = fn[0:n].to(device=dev, dtype=torch.float32).contiguous()
W_post = fn[n:2*n].to(device=dev, dtype=torch.float32).contiguous()
W_res = fn[2*n:].to(device=dev, dtype=torch.float32).contiguous()
# base: [S_pre(4), S_post(4), S_res(16)]
S_pre = base[0:n].reshape(1, n).to(device=dev, dtype=torch.bfloat16).contiguous()
S_post = base[n:2*n].reshape(n, 1).to(device=dev, dtype=torch.bfloat16).contiguous()
S_res = base[2*n:].reshape(n, n).to(device=dev, dtype=torch.bfloat16).contiguous()
# scale: [alpha_pre, alpha_post, alpha_res]
alpha_pre = scale[0].item()
alpha_post = scale[1].item()
alpha_res = scale[2].item()
self._impl.load_weights(
W_pre=W_pre, W_res=W_res, W_post=W_post,
S_pre=S_pre, S_res=S_res, S_post=S_post,
alpha_pre=alpha_pre, alpha_res=alpha_res, alpha_post=alpha_post)
@staticmethod
def init_state(embeddings, n_hc=4):
from dsv4.layers.mhc import mHCLayer
return mHCLayer.init_state(embeddings, n_hc)
def pre_block(self, X_l):
return self._impl.pre_block(X_l)
def post_block(self, X_l, F_out, ctx):
return self._impl.post_block(X_l, F_out, ctx)
# =====================================================================
# RoPE — partial, GPT-J interleaved, last rope_dim dims
# =====================================================================
def build_rope_cache(max_pos, rope_dim, device, theta=10000.0):
"""Build cos/sin caches for partial RoPE.
Returns: (cos_cache, sin_cache) each (max_pos, rope_dim//2) BF16
"""
half = rope_dim // 2
freqs = 1.0 / (theta ** (torch.arange(0, rope_dim, 2, dtype=torch.float32) / rope_dim))
angles = torch.outer(torch.arange(max_pos, dtype=torch.float32), freqs)
return torch.cos(angles).bfloat16().to(device), torch.sin(angles).bfloat16().to(device)
def apply_rope_partial(x, positions, cos_cache, sin_cache, head_dim, rope_dim):
"""Apply partial GPT-J interleaved RoPE to the last rope_dim dims of each head."""
T, n_h, hd = x.shape
nope = hd - rope_dim
cos = cos_cache[positions].unsqueeze(1) # (T, 1, half) BF16
sin = sin_cache[positions].unsqueeze(1)
out = x.clone()
x_rope = x[:, :, nope:]
out[:, :, nope:][..., 0::2] = x_rope[..., 0::2] * cos - x_rope[..., 1::2] * sin
out[:, :, nope:][..., 1::2] = x_rope[..., 0::2] * sin + x_rope[..., 1::2] * cos
return out
def apply_inverse_rope(o, positions, cos_cache, sin_cache, head_dim, rope_dim):
"""Apply inverse RoPE (conjugate rotation) to attention output."""
T, n_h, hd = o.shape
nope = hd - rope_dim
cos = cos_cache[positions].unsqueeze(1)
sin = sin_cache[positions].unsqueeze(1)
out = o.clone()
o_rope = o[:, :, nope:]
out[:, :, nope:][..., 0::2] = o_rope[..., 0::2] * cos + o_rope[..., 1::2] * sin
out[:, :, nope:][..., 1::2] = -o_rope[..., 0::2] * sin + o_rope[..., 1::2] * cos
return out
class SimpleKVCache:
"""Per-layer KV cache for decode. Stores BF16 K,V accumulated across steps.
MQA: 1 KV head, so cache is (1, seq_len, hd) per layer."""
def __init__(self, head_dim, max_seq=8192, device='cuda:0'):
self.hd = head_dim
self.max_seq = max_seq
self.device = device
self.k = torch.zeros(1, max_seq, head_dim, dtype=torch.bfloat16, device=device)
self.v = torch.zeros(1, max_seq, head_dim, dtype=torch.bfloat16, device=device)
self.len = 0
def append(self, k_new, v_new):
"""Append K,V. k_new: (1, T, hd), v_new: (1, T, hd)."""
T = k_new.shape[1]
self.k[0, self.len:self.len + T] = k_new[0]
self.v[0, self.len:self.len + T] = v_new[0]
self.len += T
def get(self):
"""Get K,V up to current length. Returns (1, seq_len, hd) each."""
return self.k[:, :self.len], self.v[:, :self.len]
# =====================================================================
# Weight loading — streams safetensors shards, distributes to 8 GPUs
# =====================================================================
def load_weights_to_cpu(checkpoint_dir):
"""Load all weights from checkpoint to CPU memory.
Weights stay on CPU; we move per-layer to GPU on demand during inference.
This avoids OOM from 285K GPU allocations and allows streaming.
Returns:
all_weights: dict[key] → tensor on CPU
"""
from safetensors.torch import load_file
cdir = Path(checkpoint_dir)
index_path = cdir / "model.safetensors.index.json"
weight_map = {}
if index_path.exists():
with open(index_path) as f:
weight_map = json.load(f).get("weight_map", {})
shard_names = set(weight_map.values()) if weight_map else {
f"model-{i:05d}-of-00095.safetensors" for i in range(1, 96)
}
print(f"Loading {len(shard_names)} shards to CPU...")
all_weights = {}
loaded = 0
for shard_name in sorted(shard_names):
if not (cdir / shard_name).exists():
continue
data = load_file(str(cdir / shard_name))
all_weights.update(data)
loaded += 1
if loaded % 20 == 0:
print(f" {loaded}/{len(shard_names)} shards, {len(all_weights)} tensors")
print(f" Done: {len(all_weights)} tensors on CPU")
return all_weights
def get_layer_weights(all_weights, li, device):
"""Get weights for layer li, moved to target device.
Returns dict of key→tensor on device. Filters by model.layers.{li} prefix.
"""
prefix = f"model.layers.{li}."
w = {}
for key in all_weights:
if key.startswith(prefix):
w[key] = all_weights[key].to(device=device, non_blocking=True)
return w
# =====================================================================
# Single layer forward
# =====================================================================
def forward_layer(X_l, w, li, cfg, rope_cos, rope_sin,
attn_mhc, ffn_mhc, attn_norm, ffn_norm,
kv_cache, token_id, positions):
"""Forward one layer with mHC + Attention + FFN.
Architecture (paper §2):
X_l → mHC.pre_block(attn) → RMSNorm → Attention → F_attn → mHC.post_block → X_mid
X_mid → mHC.pre_block(ffn) → RMSNorm → MoE → F_ffn → mHC.post_block → X_{l+1}
X_l: (T, n_hc, H) BF16 — mHC residual state
Returns: X_next (T, n_hc, H) BF16
"""
device = X_l.device
H = cfg["hidden_size"]
n_h = cfg["num_attention_heads"]
hd = cfg["head_dim"]
rd = cfg.get("qk_rope_head_dim", cfg.get("rope_dim", 64))
o_rank = cfg.get("output_group_dim", 1024)
o_groups = cfg.get("num_output_groups", 16)
n_hc = 4
pre = f"model.layers.{li}.self_attn"
T = X_l.shape[0]
heads_per_group = n_h // o_groups
group_input_dim = heads_per_group * hd
# ==================================================================
# ATTENTION SUB-BLOCK
# ==================================================================
# -- mHC pre_block (attention) --
x_in, attn_ctx = attn_mhc.pre_block(X_l) # x_in: (T, H)
if False: # diag disabled
A_l = None
B_l, C_l = attn_ctx
print(f" L{li} pre_attn: |X_l|={X_l.abs().max().item():.2f} |x_in|={x_in.abs().max().item():.2f}", flush=True)
# -- RMSNorm (pre-norm before attention) --
x_normed = attn_norm.forward(x_in) # (T, H) BF16
# -- Q projection: q_a (low-rank down) → q_a_norm → q_b (low-rank up) --
c_Q = nvfp4_linear(x_normed,
w[f"{pre}.q_a_proj.weight"],
w[f"{pre}.q_a_proj.weight_scale"],
w[f"{pre}.q_a_proj.weight_scale_2"]) # (T, dc)
# Q norm (RMSNorm after q_a, before q_b)
q_norm_w = w.get(f"{pre}.q_a_norm.weight")
if q_norm_w is not None:
c_Q_f = c_Q.float()
c_Q_rms = c_Q_f.pow(2).mean(-1, keepdim=True).add(1e-6).rsqrt()
c_Q = (c_Q_f * c_Q_rms * q_norm_w.float()).bfloat16()
q = nvfp4_linear(c_Q,
w[f"{pre}.q_b_proj.weight"],
w[f"{pre}.q_b_proj.weight_scale"],
w[f"{pre}.q_b_proj.weight_scale_2"]) # (T, n_h * hd)
# -- KV projection (MQA: 1 KV head) + KV norm --
kv = nvfp4_linear(x_normed,
w[f"{pre}.kv_proj.weight"],
w[f"{pre}.kv_proj.weight_scale"],
w[f"{pre}.kv_proj.weight_scale_2"]) # (T, hd)
# KV norm (RMSNorm after kv_proj)
kv_norm_w = w.get(f"{pre}.kv_norm.weight")
if kv_norm_w is not None:
kv_f = kv.float()
kv_rms = kv_f.pow(2).mean(-1, keepdim=True).add(1e-6).rsqrt()
kv = (kv_f * kv_rms * kv_norm_w.float()).bfloat16()
# -- Reshape for attention --
q_heads = q.reshape(T, n_h, hd) # (T, n_h, hd)
kv_new = kv.reshape(T, 1, hd) # (T, 1, hd) — 1 KV head
# -- Apply RoPE to Q (at current positions) --
positions_dev = positions.to(device)
q_heads = apply_rope_partial(q_heads, positions_dev, rope_cos, rope_sin, hd, rd)
# -- Apply RoPE to KV (at current positions) BEFORE caching --
# DSV4 convention: RoPE applied to KV before writing to cache.
# K = V in DSV4 MQA (same projection, same RoPE'd tensor).
kv_new = apply_rope_partial(kv_new, positions_dev, rope_cos, rope_sin, hd, rd)
# -- KV cache: append RoPE'd KV (K=V) --
k_new = kv_new # (T, 1, hd) — RoPE'd
v_new = kv_new # K = V in DSV4 MQA
kv_cache.append(k_new.permute(1, 0, 2), v_new.permute(1, 0, 2)) # (1, T, hd)
# -- Get full KV from cache (already RoPE'd) --
k_full, v_full = kv_cache.get() # (1, seq_len, hd) each — RoPE'd, K=V
seq_len = k_full.shape[1]
# -- FMHA: (n_h, T, hd) × (1, seq_len, hd) → (n_h, T, hd) --
q_input = q_heads.permute(1, 0, 2) # (n_h, T, hd)
# Use PyTorch SDPA for correctness verification
USE_SDPA = False # Use production FMHA kernel (better residual, no sinks)
if USE_SDPA:
# Expand K/V for GQA: (1, seq_len, hd) → (n_h, seq_len, hd)
k_expanded = k_full.expand(n_h, -1, -1).contiguous() # (n_h, seq_len, hd)
v_expanded = v_full.expand(n_h, -1, -1).contiguous()
# Add attention sink (paper §2.3.3, D5c)
# The sink is a per-head logit bias added to a virtual position.
# We simulate it by appending a zero-valued KV position with the sink logit.
sink_key = f"{pre}.sinks"
if sink_key in w and seq_len > 0:
sinks = w[sink_key].to(device=device) # (n_h,) BF16
# Append zero KV entry for the sink
sink_k = torch.zeros(n_h, 1, hd, dtype=torch.bfloat16, device=device)
sink_v = torch.zeros(n_h, 1, hd, dtype=torch.bfloat16, device=device)
k_with_sink = torch.cat([k_expanded, sink_k], dim=1) # (n_h, seq_len+1, hd)
v_with_sink = torch.cat([v_expanded, sink_v], dim=1)
# Create attention bias: sink logit added to the last position for each head
# attn_mask shape: (n_h, T, seq_len+1)
sink_bias_mask = torch.zeros(n_h, T, seq_len + 1, dtype=torch.bfloat16, device=device)
for h in range(n_h):
sink_bias_mask[h, :, -1] = sinks[h] # Add sink logit to sink position
attn_out = torch.nn.functional.scaled_dot_product_attention(
q_input, k_with_sink, v_with_sink,
attn_mask=sink_bias_mask,
scale=1.0 / math.sqrt(hd))
else:
attn_out = torch.nn.functional.scaled_dot_product_attention(
q_input, k_expanded, v_expanded,
scale=1.0 / math.sqrt(hd), is_causal=False)
else:
from dsv4.kernels.attention.production import dsv4_attention
attn_out = dsv4_attention(q_input, k_full, v_full)
attn_out = attn_out.permute(1, 0, 2) # (T, n_h, hd)
# -- Inverse RoPE on attention output (paper §2.3.3) --
attn_out = apply_inverse_rope(attn_out, positions_dev, rope_cos, rope_sin, hd, rd)
# -- Output projection: wo_a (grouped BMM) + wo_b (NVFP4) --
# wo_a: grouped linear, (n_h, hd) → (n_groups, o_rank) via BMM
attn_flat = attn_out.reshape(T, n_h * hd) # (T, n_h * hd)
attn_grouped = attn_flat.reshape(T, o_groups, heads_per_group * hd) # (T, groups, group_dim)
oa_w = w[f"{pre}.o_a_proj.weight"].bfloat16() # (n_groups * o_rank, group_input_dim) BF16
oa_3d = oa_w.reshape(o_groups, o_rank, group_input_dim) # (groups, o_rank, group_dim)
attn_for_bmm = attn_grouped.permute(1, 0, 2) # (groups, T, group_dim)
grouped_out = torch.bmm(attn_for_bmm, oa_3d.transpose(1, 2)) # (groups, T, o_rank)
grouped_flat = grouped_out.permute(1, 0, 2).reshape(T, o_groups * o_rank) # (T, groups*o_rank)
F_attn = nvfp4_linear(grouped_flat,
w[f"{pre}.o_b_proj.weight"],
w[f"{pre}.o_b_proj.weight_scale"],
w[f"{pre}.o_b_proj.weight_scale_2"]) # (T, H)
# -- mHC post_block (attention) --
X_mid = attn_mhc.post_block(X_l, F_attn, attn_ctx) # (T, n_hc, H)
# Diagnostic: check mHC is stabilizing the residual
if False: # Disable diagnostics for production run
B_l, C_l = attn_ctx
print(f" L{li} attn: |X_l|={X_l.abs().max().item():.2f} |F_attn|={F_attn.abs().max().item():.2f} |B|={B_l.abs().max().item():.4f} |C|={C_l.abs().max().item():.4f} |X_mid|={X_mid.abs().max().item():.2f}", flush=True)
# ==================================================================
# FFN SUB-BLOCK
# ==================================================================
# -- mHC pre_block (FFN) --
x_ffn, ffn_ctx = ffn_mhc.pre_block(X_mid) # (T, H)
# -- RMSNorm (pre-norm before FFN) --
x_ffn_normed = ffn_norm.forward(x_ffn) # (T, H) BF16
# -- MoE + shared expert --
F_ffn = moe_forward(x_ffn_normed, w, li, cfg, token_id, device)
# -- mHC post_block (FFN) --
X_next = ffn_mhc.post_block(X_mid, F_ffn, ffn_ctx) # (T, n_hc, H)
if False: # diag disabled
B_l_ffn, C_l_ffn = ffn_ctx
print(f" L{li} ffn: |X_mid|={X_mid.abs().max().item():.2f} |F_ffn|={F_ffn.abs().max().item():.2f} |B|={B_l_ffn.abs().max().item():.4f} |C|={C_l_ffn.abs().max().item():.4f} |X_next|={X_next.abs().max().item():.2f}", flush=True)
return X_next
# =====================================================================
# MoE forward — hash + dense routing, SwiGLU with clamping
# =====================================================================
def moe_forward(x, w, li, cfg, token_id, device):
"""Run routed MoE + shared expert.
x: (T, H) BF16 — post-RMSNorm FFN input
Returns: (T, H) BF16
"""
H = cfg["hidden_size"]
n_experts = cfg["n_routed_experts"]
top_k = cfg.get("num_experts_per_tok", 6)
routed_scaling = cfg.get("routed_scaling_factor", 2.5)
swiglu_limit = cfg.get("swiglu_limit", 10.0)
mlp_inter = cfg["moe_intermediate_size"]
is_hash = li < 3
# ---- Routing ----
expert_ids = None
expert_weights = None
if is_hash:
tid2eid_key = f"model.layers.{li}.mlp.gate.tid2eid"
if tid2eid_key in w:
tid2eid = w[tid2eid_key]
tid = token_id.item() if token_id.numel() == 1 else token_id[0].item()
expert_ids = tid2eid[tid] # (top_k,) int64
expert_weights = torch.ones(top_k, dtype=torch.float32, device=x.device) / top_k
else:
# Fallback: use dense routing even for hash layers
is_hash = False
if not is_hash:
# Dense routing: sqrt(softplus(X @ W_gate)) + e_bias for selection
gate_w = w[f"model.layers.{li}.mlp.gate.weight"] # (H, n_experts) BF16
logits = torch.nn.functional.linear(x, gate_w.bfloat16()) # (T, n_experts)
# Activation: sqrt(softplus(logits))
activated = torch.sqrt(torch.nn.functional.softplus(logits.float()) + 1e-6)
# e_bias: learned per-expert bias for SELECTION ONLY (not in weights)
e_bias_key = f"model.layers.{li}.mlp.gate.e_bias"
if e_bias_key in w:
activated = activated + w[e_bias_key].float().unsqueeze(0)
# Top-k
scores, indices = activated.topk(top_k, dim=-1) # (T, top_k)
# Renormalize on UNBIASED activation (no e_bias in weights)
unbiased = torch.sqrt(torch.nn.functional.softplus(logits.float()) + 1e-6)
unbiased_scores = torch.gather(unbiased, -1, indices)
expert_weights = unbiased_scores / unbiased_scores.sum(dim=-1, keepdim=True)
# For T=1 decode, squeeze
if x.shape[0] == 1:
expert_ids = indices[0]
expert_weights = expert_weights[0]
else:
raise NotImplementedError("Multi-token MoE routing")
# ---- Run selected experts ----
T = x.shape[0]
expert_outputs = []
for i, eid in enumerate(expert_ids):
eid_int = eid.item()
epre = f"model.layers.{li}.mlp.experts.{eid_int}"
gate = nvfp4_linear(x,
w[f"{epre}.gate_proj.weight"],
w[f"{epre}.gate_proj.weight_scale"],
w[f"{epre}.gate_proj.weight_scale_2"])
up = nvfp4_linear(x,
w[f"{epre}.up_proj.weight"],
w[f"{epre}.up_proj.weight_scale"],
w[f"{epre}.up_proj.weight_scale_2"])
# SwiGLU with clamping (paper §4.2.3)
silu_out = torch.nn.functional.silu(gate.float())
if swiglu_limit is not None:
silu_out = silu_out.clamp(-swiglu_limit, swiglu_limit)
up_clamped = up.float().clamp(-swiglu_limit, swiglu_limit)
else:
up_clamped = up.float()
hidden = (silu_out * up_clamped).bfloat16()
down = nvfp4_linear(hidden,
w[f"{epre}.down_proj.weight"],
w[f"{epre}.down_proj.weight_scale"],
w[f"{epre}.down_proj.weight_scale_2"])
expert_outputs.append(down)
# Weighted combine + scaling
routed_out = torch.zeros_like(x)
for i, (out, wt) in enumerate(zip(expert_outputs, expert_weights)):
routed_out = routed_out + (out.float() * wt).bfloat16()
routed_out = (routed_out.float() * routed_scaling).bfloat16()
# ---- Shared expert ----
se_pre = f"model.layers.{li}.mlp.shared_experts"
se_gate_key = f"{se_pre}.gate_proj.weight"
if se_gate_key in w:
gate = nvfp4_linear(x,
w[se_gate_key],
w[f"{se_pre}.gate_proj.weight_scale"],
w[f"{se_pre}.gate_proj.weight_scale_2"])
up = nvfp4_linear(x,
w[f"{se_pre}.up_proj.weight"],
w[f"{se_pre}.up_proj.weight_scale"],
w[f"{se_pre}.up_proj.weight_scale_2"])
silu_out = torch.nn.functional.silu(gate.float())
if swiglu_limit is not None:
silu_out = silu_out.clamp(-swiglu_limit, swiglu_limit)
up_clamped = up.float().clamp(-swiglu_limit, swiglu_limit)
else:
up_clamped = up.float()
hidden = (silu_out * up_clamped).bfloat16()
shared_out = nvfp4_linear(hidden,
w[f"{se_pre}.down_proj.weight"],
w[f"{se_pre}.down_proj.weight_scale"],
w[f"{se_pre}.down_proj.weight_scale_2"])
else:
shared_out = torch.zeros_like(x)
return routed_out + shared_out
# =====================================================================
# Main
# =====================================================================
def main():
t_start = time.time()
print("=" * 70)
print("DSV4 Single-Shot Inference — Full Pipeline (mHC+Attn+MoE)")
print(" Proper Sinkhorn mHC, RMSNorm, inverse RoPE, production FMHA")
print("=" * 70)
with open(os.path.join(CHECKPOINT_DIR, "config.json")) as f:
cfg = json.load(f)
n_layers = cfg["num_hidden_layers"]
H = cfg["hidden_size"]
n_h = cfg["num_attention_heads"]
hd = cfg["head_dim"]
rd = cfg.get("qk_rope_head_dim", cfg.get("rope_dim", 64))
n_hc = 4
print(f"Model: {n_layers} layers, {n_h} heads, hd={hd}, rope_dim={rd}")
print(f"Experts: {cfg['n_routed_experts']}, top-{cfg.get('num_experts_per_tok', 6)}")
# ==== Phase 1: Load weights to CPU ====
print(f"\n{'='*70}\nPhase 1: Loading weights to CPU\n{'='*70}")
all_weights = load_weights_to_cpu(CHECKPOINT_DIR)
t_loaded = time.time()
print(f"Weight loading: {t_loaded - t_start:.1f}s")
# ==== Build mHC blocks + RMSNorms (small weights, keep on GPU) ====
print("Building mHC blocks and RMSNorms...")
attn_mhc_blocks = {}
ffn_mhc_blocks = {}
attn_norms = {}
ffn_norms = {}
for li in range(n_layers):
gpu = li % NUM_GPUS
dev = f"cuda:{gpu}"
# mHC blocks (small weights: fn (24, 28672) FP32 ≈ 2.6MB each)
for prefix, blocks in [(f"model.layers.{li}.attn_hc", attn_mhc_blocks),
(f"model.layers.{li}.ffn_hc", ffn_mhc_blocks)]:
fn_key = f"{prefix}.fn"
base_key = f"{prefix}.base"
scale_key = f"{prefix}.scale"
if fn_key in all_weights and base_key in all_weights and scale_key in all_weights:
mhc = mHCBlock(hidden_dim=H, n_hc=n_hc, device=dev)
mhc.load_from_checkpoint(
all_weights[fn_key], all_weights[base_key], all_weights[scale_key])
blocks[li] = mhc
else:
print(f" WARNING: no mHC weights for {prefix}, using identity fallback")
mhc = mHCBlock(hidden_dim=H, n_hc=n_hc, device=dev)
n = n_hc
K = n * H
mhc.W_stacked = torch.zeros(n + n*n + n, K, dtype=torch.float32, device=dev)
mhc.S_pre = torch.zeros(1, n, dtype=torch.float32, device=dev)
mhc.S_res = torch.eye(n, dtype=torch.float32, device=dev)
mhc.S_post = torch.ones(n, 1, dtype=torch.float32, device=dev) * 0.5
mhc.alpha_pre = 0.01
mhc.alpha_res = 0.01
mhc.alpha_post = 0.01
blocks[li] = mhc
# RMSNorms
attn_norm = RMSNorm(H, eps=cfg.get('rms_norm_eps', 1e-6), device=dev)
an_key = f"model.layers.{li}.input_layernorm.weight"
if an_key in all_weights:
attn_norm.weight = all_weights[an_key].to(device=dev, dtype=torch.float32)
attn_norms[li] = attn_norm
ffn_norm = RMSNorm(H, eps=cfg.get('rms_norm_eps', 1e-6), device=dev)
fn_key = f"model.layers.{li}.post_attention_layernorm.weight"
if fn_key in all_weights:
ffn_norm.weight = all_weights[fn_key].to(device=dev, dtype=torch.float32)
ffn_norms[li] = ffn_norm
print(f" attn mHC: {len(attn_mhc_blocks)}, ffn mHC: {len(ffn_mhc_blocks)}")
# ==== Global weights (small, keep on gpu0) ====
torch.cuda.set_device(0)
embed_w = all_weights.get("model.embed_tokens.weight")
embed = torch.nn.Embedding.from_pretrained(embed_w.bfloat16().to('cuda:0'))
lm_w = all_weights.get("lm_head.weight", embed_w).bfloat16().to('cuda:0')
final_norm_w = all_weights.get("model.norm.weight")
if final_norm_w is not None:
final_norm_w = final_norm_w.to('cuda:0')
rope_caches = {g: build_rope_cache(8192, rd, f"cuda:{g}") for g in range(NUM_GPUS)}
# ==== KV caches (one per layer on its GPU) ====
kv_caches = {}
for li in range(n_layers):
kv_caches[li] = SimpleKVCache(head_dim=hd, max_seq=8192, device=f"cuda:{li % NUM_GPUS}")
# ==== Phase 2: Compile FMHA ====
print(f"\n{'='*70}\nPhase 2: JIT compiling\n{'='*70}")
from dsv4.kernels.attention.production import dsv4_attention
torch.cuda.set_device(0)
dummy_q = torch.randn(n_h, 1, hd, dtype=torch.bfloat16, device='cuda:0')
dummy_k = torch.randn(1, 1, hd, dtype=torch.bfloat16, device='cuda:0')
try:
_ = dsv4_attention(dummy_q, dummy_k, dummy_k.clone())
print(" FMHA: compiled OK")
except Exception as e:
print(f" FMHA error: {e}")
t_compiled = time.time()
print(f"Compile: {t_compiled - t_loaded:.1f}s")
# ==== Phase 3: Inference ====
print(f"\n{'='*70}\nPhase 3: Inference\n{'='*70}")
from transformers import AutoTokenizer
tokenizer = AutoTokenizer.from_pretrained(CHECKPOINT_DIR)
input_ids = tokenizer.encode(PROMPT, return_tensors="pt").cuda()
print(f"Prompt: '{PROMPT}'{input_ids.tolist()}")
generated = input_ids[0].tolist()
# ==== Prefill: process prompt tokens to fill KV cache ====
print(f"Prefilling {len(generated)} prompt tokens...")
for prefill_idx, tid_val in enumerate(generated):
t0 = time.time()
tid = torch.tensor([tid_val], dtype=torch.long, device='cuda:0')
positions = torch.tensor([prefill_idx], dtype=torch.long, device='cuda:0')
emb = embed(tid) # (1, H) on gpu0
X = mHCBlock.init_state(emb, n_hc) # (1, n_hc, H)
for li in range(n_layers):
gpu = li % NUM_GPUS
dev = f"cuda:{gpu}"
if X.device != torch.device(dev):
X = X.to(dev)
torch.cuda.set_device(gpu)
# Fetch this layer's weights from CPU → GPU (streamed, not all at once)
w = get_layer_weights(all_weights, li, dev)
attn_mhc = attn_mhc_blocks.get(li)
ffn_mhc = ffn_mhc_blocks.get(li)
a_norm = attn_norms[li]
f_norm = ffn_norms[li]
rc, rs = rope_caches[gpu]
X = forward_layer(X, w, li, cfg, rc, rs,
attn_mhc, ffn_mhc, a_norm, f_norm,
kv_caches[li], tid, positions)
# Free per-layer GPU weights to save memory
del w
X = X.to('cuda:0')
torch.cuda.set_device(0)
if prefill_idx == 0:
print(f" Token 0: {time.time()-t0:.1f}s (includes per-layer weight transfer)")
print(f" Prefill done ({len(generated)} tokens, {time.time()-t_compiled:.1f}s)")
# ==== Decode: generate new tokens ====
print(f"\nDecoding (max {MAX_NEW_TOKENS} new tokens)...")
all_tokens = generated.copy()
for step in range(MAX_NEW_TOKENS):
t0 = time.time()
tid = torch.tensor([all_tokens[-1]], dtype=torch.long, device='cuda:0')
decode_pos = len(all_tokens) - 1
positions = torch.tensor([decode_pos], dtype=torch.long, device='cuda:0')
emb = embed(tid) # (1, H) on gpu0
X = mHCBlock.init_state(emb, n_hc) # (1, n_hc, H)
for li in range(n_layers):
gpu = li % NUM_GPUS
dev = f"cuda:{gpu}"
if X.device != torch.device(dev):
X = X.to(dev)
torch.cuda.set_device(gpu)
w = get_layer_weights(all_weights, li, dev)
attn_mhc = attn_mhc_blocks.get(li)
ffn_mhc = ffn_mhc_blocks.get(li)
a_norm = attn_norms[li]
f_norm = ffn_norms[li]
rc, rs = rope_caches[gpu]
X = forward_layer(X, w, li, cfg, rc, rs,
attn_mhc, ffn_mhc, a_norm, f_norm,
kv_caches[li], tid, positions)
del w
X = X.to('cuda:0')
torch.cuda.set_device(0)
# Read out stream 0 → RMSNorm → lm_head
x_out = X[:, 0, :] # (1, H)
if final_norm_w is not None:
xf = x_out.float()
rms = xf.pow(2).mean(-1, keepdim=True).add(1e-6).rsqrt()
x_out = (xf * rms * final_norm_w.float()).bfloat16()
logits = torch.nn.functional.linear(x_out, lm_w)
# Top-5 predictions for debugging
top5_vals, top5_ids = torch.topk(logits[0], 5)
top5_str = ' '.join([f'{tokenizer.decode([tid.item()])}({val.item():.1f})' for tid, val in zip(top5_ids, top5_vals)])
next_id = torch.argmax(logits, dim=-1).item()
generated.append(next_id)
all_tokens.append(next_id)
tok_str = tokenizer.decode([next_id])
dt = time.time() - t0
has_nan = torch.isnan(logits.float()).any().item()
has_inf = torch.isinf(logits.float()).any().item()
lmin, lmax = logits.float().min().item(), logits.float().max().item()
x_max = X.abs().max().item()
print(f" Step {step}: {next_id} '{tok_str}' ({dt:.2f}s) "
f"logits=[{lmin:.1f},{lmax:.1f}] nan={has_nan} inf={has_inf} "
f"|X|={x_max:.3f} top5: {top5_str}")
if has_nan or has_inf:
print(" Numerical issue — stopping")
break
if next_id == tokenizer.eos_token_id:
break
out = tokenizer.decode(generated, skip_special_tokens=True)
total = time.time() - t_start
print(f"\n{'='*70}")
print(f"Input: '{PROMPT}'")
print(f"Output: '{out}'")
print(f"Total: {total:.1f}s")
print(f"{'='*70}")
if __name__ == "__main__":
main()