tests/test.sh exit code (0 → resolved); the classification below is post-hoc and cannot change it.ls -la && echo "---SUBMISSION---" && ls -la submission/ && echo "---DATA---" && ls -la data/ && echo "---run_checks head---" && head -50 run_checks.py
total 117
drwxr-xr-x 4 user user 4096 Jun 30 23:56 .
drwxr-xr-x 21 root root 4096 Jun 30 23:56 ..
-rw-r--r-- 1 user user 15364 Jun 30 23:56 PROMPT.txt
-rw-r--r-- 1 user user 5 Jun 30 23:56 agent.pid
-rw-r--r-- 1 user user 0 Jun 30 23:56 agent.stderr
-rw-r--r-- 1 user user 4567 Jun 30 23:56 agent_trajectory.jsonl
drwxr-xr-x 2 user user 60 Jun 30 23:56 data
-rw-r--r-- 1 user user 0 Jun 30 23:56 run.log
-rwxr-xr-x 1 user user 76773 Jun 30 10:02 run_checks.py
drwxr-xr-x 2 user user 128 Jun 30 23:56 submission
-rwxr-xr-x 1 user user 804 Jun 30 10:02 test.sh
-rwxr-xr-x 1 user user 629 Jun 30 10:02 timer.sh
---SUBMISSION---
total 112
drwxr-xr-x 2 user user 128 Jun 30 23:56 .
drwxr-xr-x 4 user user 4096 Jun 30 23:56 ..
-rw-r--r-- 1 user user 50472 Jun 30 10:02 autograd.py
-rw-r--r-- 1 user user 1661 Jun 30 10:02 data.py
-rw-r--r-- 1 user user 33295 Jun 30 10:02 nn.py
-rw-r--r-- 1 user user 11509 Jun 30 10:02 optim.py
-rw-r--r-- 1 user user 2827 Jun 30 10:02 train.py
---DATA---
total 1900
drwxr-xr-x 2 user user 60 Jun 30 23:56 .
drwxr-xr-x 4 user user 4096 Jun 30 23:56 ..
-rw-r--r-- 1 user user 1940490 Jun 30 10:02 train.npz
---run_checks head---
#!/usr/bin/env python3
"""PUBLIC component check runner for the from-scratch numpy QAT framework.
Run this AS OFTEN AS YOU LIKE while you implement, to get a per-check progress signal:
python3 /app/run_checks.py # check the framework in /app/submission
python3 /ap
…[truncated 2704 chars]/app/submission/autograd.py
1 """Reverse-mode automatic differentiation over numpy arrays. 2 3 YOU MUST IMPLEMENT EVERYTHING MARKED ``# TODO`` / ``raise NotImplementedError`` using numpy 4 ONLY. No torch/jax/tensorflow/autograd/etc. The grader's conformance suite (a) differentiates the 5 differentiable ops below with central finite differences and compares against YOUR analytic 6 backward, and (b) checks the LSQ ``fake_quant`` STE + scale gradient against the reference 7 definition, so every backward must be correct (a single wrong reduction / STE / Jacobian fails 8 the gradient checks and zeroes the reward). 9 10 A ``Tensor`` wraps one ``numpy.ndarray`` (stored as ``float64`` in ``self.data``) and records the 11 op that produced it so ``.backward()`` can accumulate ``.grad`` into every leaf. 12 13 CONTRACT (the grader builds its own reference against this exact API): 14 * ``Tensor(data, requires_grad=False)`` -- ``self.data`` is ``float64``; ``self.grad`` is None 15 until backward, then a same-shape ``float64`` array. requires_grad leaves accumulate grad. 16 * Every op returns a NEW Tensor whose ``requires_grad`` is the OR of its differentiable inputs, 17 and whose ``_backward`` closure does ``input._accum(local_grad)`` (so multiple paths add up). 18 * ``t.backward(grad=None)`` -- seeds ``grad`` (all-ones for a scalar) and walks the graph in 19 reverse topological order, calling each node's ``_backward``. Repeated backward accumulates. 20 * BROADCASTING: binary ops broadcast like numpy; the ba …[truncated 54775 chars]
/app/submission/nn.py
1 """QAT layers + a small INT2 CNN, built on YOUR autograd engine.
2
3 Implement every ``# TODO`` forward (the parameters + ``named_params``/``buffers`` naming are
4 already wired for you; you compose the autograd ops). The grader checks each layer's forward
5 against its OWN reference AND finite-difference-checks the gradients that flow through your
6 autograd, so the composition must be exactly right.
7
8 CHECKPOINT NAMING CONTRACT (the grader loads YOUR ``qat_model.npz`` into its own reference
9 QuantCNN by these exact keys/shapes -- do not rename):
10 stem.conv.weight (W0,3,3,3) | stem.conv.wq.scale (1,) | stem.bn.gamma/.beta (W0,) | stem.aq.scale (1,)
11 blocks.{i}.conv.weight (Cout,Cin,3,3) | blocks.{i}.conv.wq.scale (1,)
12 blocks.{i}.bn.gamma/.beta (Cout,) | blocks.{i}.aq.scale (1,)
13 head.weight (Cfeat,n_classes) | head.wq.scale (1,) | head.bias (n_classes,)
14 BN running stats are BUFFERS (saved alongside params, NOT trained by the optimizer, NOT audited):
15 stem.bn.running_mean/.running_var (W0,) | blocks.{i}.bn.running_mean/.running_var (Cout,)
16
17 Quantization: every conv/head WEIGHT is LSQ-quantized to a signed 2-bit grid {-2,-1,0,1} via a
18 per-tensor learnable step before use; every block activation (post-relu) is LSQ-quantized to an
19 unsigned 2-bit grid {0,1,2,3}. Conv = 3x3 same-padding cross-correlation. Linear head = x@weight+bias.
20 """
21 from __future__ import annotations
22
23 from typing import Dict, List
24
25 import numpy as n
…[truncated 35167 chars]/app/submission/optim.py
1 """Optimizers over autograd ``Tensor`` parameters.
2
3 Implement the ``# TODO`` update math. Both optimizers update ``param.data`` IN PLACE using
4 ``param.grad``. The grader runs one (and several) steps and compares your updated params against
5 its reference within a tight tolerance, so the math must be exact.
6
7 SGD (with momentum ``mu`` and coupled weight decay ``wd``):
8 g = grad + wd * param; v = mu * v + g; param -= lr * v
9
10 AdamW (DECOUPLED weight decay -- the decay is applied to the PARAM, not folded into the moments):
11 m = b1*m + (1-b1)*g; v = b2*v + (1-b2)*g^2
12 mhat = m / (1 - b1^t); vhat = v / (1 - b2^t) # t = 1-based step count
13 param -= lr * ( mhat / (sqrt(vhat) + eps) + wd * param )
14
15 PARAM GROUPS: both optimizers must accept EITHER a flat param list OR a list of
16 ``{"params": [...], "weight_decay": wd}`` dicts, so weight decay can be applied to SOME params
17 and not others. This matters for low-bit LSQ QAT: the per-tensor quantizer ``scale`` params,
18 BatchNorm gamma/beta, and biases must be EXCLUDED from weight decay. ``_normalize_groups`` and
19 ``split_decay_params`` are provided; honour the per-group weight decay in ``step``.
20 """
21 from __future__ import annotations
22
23 import math
24
25 from typing import Dict, List
26
27 import numpy as np
28
29 from autograd import Tensor
30
31
32 def split_decay_params(model) -> Dict[str, List[Tensor]]:
33 """Partition a model's named params into weight
…[truncated 11129 chars]/app/submission/data.py
1 """Image data loading + normalization + batching.
2
3 The dataset is a procedurally-generated CIFAR-style set: ``X`` is ``uint8`` (N,3,S,S) in [0,255],
4 ``y`` is the integer class label in [0, n_classes). The shipped training split is at
5 ``/app/data/train.npz``; the SEALED held-out split is regenerated by the verifier (never shipped).
6
7 Implement the ``# TODO`` bodies (numpy only). The verifier normalizes the held-out split with the
8 SAME transform, so match it exactly.
9
10 Contract:
11 load_npz(path) -> (X uint8 (N,3,S,S), y int64 (N,)) [provided]
12 normalize(X) -> float64: x = (X/255 - 0.5) / 0.25 [per-channel, the grader's transform]
13 iter_minibatches(Xn, y, batch_size, rng, shuffle=True) -> yields (xb, yb)
14 """
15 from __future__ import annotations
16
17 from typing import Tuple
18
19 import numpy as np
20
21 NORM_MEAN = 0.5
22 NORM_STD = 0.25
23
24
25 def load_npz(path: str) -> Tuple[np.ndarray, np.ndarray]:
26 d = np.load(path)
27 X = np.asarray(d["X"]).astype(np.uint8)
28 y = np.asarray(d["y"]).astype(np.int64).reshape(-1)
29 return X, y
30
31
32 def normalize(X: np.ndarray) -> np.ndarray:
33 """uint8 (N,3,S,S) -> float64, x = (X/255 - NORM_MEAN) / NORM_STD."""
34 # TODO: cast to float, scale to [0,1], subtract NORM_MEAN, divide by NORM_STD.
35 raise NotImplementedError("normalize")
36
37
38 def iter_minibatches(Xn: np.ndarray, y: np.ndarray, batch_size: int,
39 rng: np.rando
…[truncated 281 chars]/app/submission/train.py
1 """Training entry point. Implement the QAT loop, train your INT2 CNN on /app/data/train.npz, and 2 SAVE the checkpoint to /app/submission/qat_model.npz (numpy .npz of the model state_dict: every 3 param + every BN running buffer, by the nn.QuantCNN naming contract). 4 5 The grader (verify.py) will: 6 (1) run the conformance suite over your autograd/nn/optim, 7 (2) BIT-AUDIT every conv/head weight tensor in your checkpoint to <= 4 distinct float values, 8 (3) load your weights into its OWN QuantCNN and score held-out top-1 on a SEALED split. 9 10 So at SAVE time you must BAKE the quantized weights back into each ``weight`` tensor (replace it 11 with round(clip(w/s, -W_NEG, W_POS)) * s using the learned wq.scale) so the saved weights already 12 lie on the 2-bit grid and pass the audit. Reaching the top-1 floor at INT-2 needs a real recipe: 13 sensible LSQ STEP-SIZE INITIALIZATION (a default scale of 1.0 quantizes the small init weights to 14 all-zeros -> a dead network at chance), per-tensor learnable scales, weight decay EXCLUDED from the 15 scale/BN/bias params, gradient clipping, a decaying LR, and enough epochs. Naive post-training 16 rounding or under-training lands at chance. 17 18 This file is a skeleton; the loop body is yours to write. 19 """ 20 from __future__ import annotations 21 22 import argparse 23 import os 24 import sys 25 26 import numpy as np 27 28 sys.path.insert(0, os.path.dirname(os.path.abspath(__file__))) 29 sys.path.insert(0, "/app") 30 3 …[truncated 1513 chars]
grep -n "batchnorm1d\|def check\|MILESTONES\|tol\|atol\|rtol" run_checks.py | head -80
21: batchnorm1d/layernorm) + end-to-end through the full QuantCNN / QuantMLP.
124:def check_autograd_grads(fw, sb):
286: op = ag.batchnorm2d if dim == 2 else ag.batchnorm1d
311: bn_grad("batchnorm1d_x", 1, "x", 58)
312: bn_grad("batchnorm1d_gamma", 1, "gamma", 59)
313: bn_grad("batchnorm1d_beta", 1, "beta", 60)
584:def check_quant_grads(fw, sb):
778:def check_nn_forward(fw, sb):
853: y = ag.batchnorm1d(T(x), T(np.ones(5)), T(np.zeros(5)), training=True).data
1082: ("batchnorm1d_zero_mean_unit_var", bn1d_stats), ("conv2d_shape", conv_shape),
1106:def check_optim(fw, sb):
1346:def check_integration(fw, sb):
1422:def check_data(fw, sb):/app/run_checks.py
124 def check_autograd_grads(fw, sb):
125 ag = fw["autograd"]
126 nn = fw["nn"]
127 T = ag.Tensor
128 results = []
129
130 def case(name, build, shape, seed, upstream=None):
131 try:
132 rng = np.random.default_rng(sb + seed)
133 x0 = rng.standard_normal(shape)
134
135 def wrapped(t):
136 o = build(t)
137 return ag.mul(o, T(upstream)) if upstream is not None else o
138 t = T(x0.copy(), requires_grad=True)
139 ag.sum(wrapped(t)).backward()
140 gan = np.asarray(t.grad, np.float64)
141 gnum = _num_grad_input(lambda xv: wrapped(T(xv)).data, x0.copy())
142 err = _max_abs(gan, gnum)
143 results.append((f"grad/{name}", err < TOL_GRAD, f"max_abs={err:.2e}"))
144 except Exception as e: # noqa: BLE001
145 results.append((f"grad/{name}", False, f"raised {type(e).__name__}: {e}"))
146
147 # element-wise binary
148 case("add_bcast", lambda t: ag.add(t, T(np.arange(t.shape[-1], dtype=float) + 1.0)), (4, 5), 1)
149 case("sub_bcast", lambda t: ag.sub(t, T(np.arange(t.shape[-1], dtype=float) + 1.0)), (4, 5), 2)
150 case("mul_bcast", lambda t: ag.mul(t, T(np.arange(t.shape[-1], dtype=float) + 1.0)), (4, 5), 3)
151 case("div", lambda t: ag.div(t, T(np.full((4, 5), 2.0))), (4, 5), 4)
152 case("div_denom", lambda t: ag.div(T(np.full((4, 5), 3.0)), ag.add(t, T(5.0))), (4, 5), 5)
153
…[truncated 11479 chars]/app/run_checks.py
344 p.data[...] = 1.0
345 m.set_training(True)
346 X = make_X(sb + seed + 5)
347 tgt = np.array([0, 2])
348 for p in m.params():
349 p.grad = None
350 ag.cross_entropy(m(X), tgt).backward()
351 allfinite = all(p.grad is not None and np.all(np.isfinite(p.grad)) for p in m.params())
352 ngrad = sum(1 for p in m.params() if p.grad is not None)
353 results.append((f"grad/{name}", allfinite and ngrad == len(m.params()),
354 f"{ngrad}/{len(m.params())} grads, finite={allfinite}"))
355 except Exception as e: # noqa: BLE001
356 results.append((f"grad/{name}", False, f"raised {type(e).__name__}: {e}"))
357 quant_model_backprop("QuantCNN_backprop", lambda s: nn.QuantCNN(6, widths=(8, 16), seed=s),
358 lambda s: np.random.default_rng(s).standard_normal((2, 3, 16, 16)), 90)
359 quant_model_backprop("QuantMLP_backprop", lambda s: nn.QuantMLP(6, img_size=8, hidden=(32, 16), seed=s),
360 lambda s: np.random.default_rng(s).standard_normal((2, 3, 8, 8)), 100)
361
362 # ---- EXTRA generic ops (same coverage as the held-out grader) ----
363 case("square", lambda t: ag.square(t), (4, 5), 110)
364 case("rsqrt", lambda t: ag.rsqrt(ag.add(t, T(5.0))), (4, 5), 111)
365 case("reciprocal", lambda t: ag.reciprocal(ag.add(t, T(5.0))), (4, 5), 112)
366
…[truncated 14060 chars]/app/run_checks.py
584 def check_quant_grads(fw, sb):
585 ag = fw["autograd"]
586 results = []
587
588 def grads(v0, s0, Qn, Qp, up):
589 vt = ag.Tensor(v0.copy(), requires_grad=True)
590 st = ag.Tensor(np.array([s0]), requires_grad=True)
591 out = ag.fake_quant(vt, st, Qn, Qp)
592 out.backward(up.copy())
593 return np.asarray(vt.grad, np.float64), np.asarray(st.grad, np.float64), out.data.copy()
594
595 def ste_indicator(name, seed, Qn, Qp, s0):
596 try:
597 rng = np.random.default_rng(sb + seed)
598 v0 = rng.standard_normal((5, 5)) * 2.0
599 up = rng.standard_normal((5, 5))
600 gv, _, _ = grads(v0, s0, Qn, Qp, up)
601 r = v0 / s0
602 middle = (r >= -Qn) & (r <= Qp)
603 expected = up * middle
604 ok = _max_abs(gv, expected) < 1e-9 and np.all(gv[~middle] == 0.0)
605 results.append((f"quant/{name}", ok, f"err={_max_abs(gv, expected):.1e}"))
606 except Exception as e: # noqa: BLE001
607 results.append((f"quant/{name}", False, f"raised {type(e).__name__}: {e}"))
608 ste_indicator("ste_indicator_w", 60, 2, 1, 0.7)
609 ste_indicator("ste_indicator_a", 61, 0, 3, 0.5)
610
611 def scale_clip(name, seed, Qn, Qp, s0, fill, bound):
612 try:
613 v0 = np.full((4, 4), fill)
614 up = np.ones((4, 4))
615 _, gs, _ = grads(v0, s0, Qn, Qp, up)
616 gscale
…[truncated 10053 chars]/app/run_checks.py
66 if inject:
67 for k, v in inject.items():
68 saved[k] = sys.modules.get(k)
69 sys.modules[k] = v
70 try:
71 spec = importlib.util.spec_from_file_location(name, str(path))
72 mod = importlib.util.module_from_spec(spec)
73 spec.loader.exec_module(mod)
74 return mod
75 finally:
76 if inject:
77 for k, old in saved.items():
78 if old is None:
79 sys.modules.pop(k, None)
80 else:
81 sys.modules[k] = old
82
83
84 def load_framework(root):
85 ag = _load_module("agent_autograd", root / "autograd.py")
86 nn = _load_module("agent_nn", root / "nn.py", inject={"autograd": ag})
87 opt = _load_module("agent_optim", root / "optim.py", inject={"autograd": ag})
88 data = _load_module("agent_data", root / "data.py")
89 return {"autograd": ag, "nn": nn, "optim": opt, "data": data}
90
91
92 # --------------------------------------------------------------------------- numeric helpers
93 def _num_grad_input(forward, x, eps=EPS):
94 g = np.zeros_like(x)
95 flat = x.reshape(-1)
96 gf = g.reshape(-1)
97 for i in range(flat.size):
98 old = flat[i]
99 flat[i] = old + eps
100 fp = float(np.sum(forward(x)))
101 flat[i] = old - eps
102 fm = float(np.sum(forward(x)))
103 flat[i] = old
104 gf[i] = (fp - fm) / (2 * eps)
105 return g
106
107
…[truncated 394 chars]grep -n "TOL_GRAD\|TOL\|EPS\s*=\|^EPS\|TOL_FWD\|TOL_OPT" run_checks.py | head
54:TOL_GRAD = 2e-3
55:TOL_PROP = 1e-6
56:TOL_OPT = 1e-7
57:TOL_SCHED = 1e-9
58:EPS = 1e-5
143: results.append((f"grad/{name}", err < TOL_GRAD, f"max_abs={err:.2e}"))
199: results.append(("grad/two_path_accum", _max_abs(xt.grad, 2.0 * x0) < TOL_GRAD, ""))
212: results.append((f"grad/{name}", _max_abs(np.asarray(lt.grad), gnum) < TOL_GRAD, ""))
230: results.append((f"grad/{name}", _max_abs(np.asarray(pt.grad), _num_grad_input(fwd, p0.copy())) < TOL_GRAD, ""))
258: results.append((f"grad/conv2d_{which}", _max_abs(np.asarray(gan), gnum) < TOL_GRAD, ""))/app/run_checks.py
784 def prop(name, fn):
785 try:
786 ok, info = fn()
787 results.append((f"fwd/{name}", ok, info))
788 except Exception as e: # noqa: BLE001
789 results.append((f"fwd/{name}", False, f"raised {type(e).__name__}: {e}"))
790
791 def softmax_sums():
792 x = np.random.default_rng(sb + 1).standard_normal((4, 5))
793 s = ag.softmax(T(x), -1).data
794 err = _max_abs(s.sum(-1), np.ones(4))
795 return err < TOL_PROP and bool((s >= 0).all()), f"row_sum_err={err:.1e}"
796
797 def logsoftmax_matches():
798 x = np.random.default_rng(sb + 3).standard_normal((4, 5))
799 return _max_abs(ag.log_softmax(T(x), -1).data, np.log(ag.softmax(T(x), -1).data)) < 1e-6, ""
800
801 def relu_def():
802 x = np.random.default_rng(sb + 4).standard_normal((4, 5))
803 return _max_abs(ag.relu(T(x)).data, np.maximum(x, 0)) < TOL_PROP, ""
804
805 def leaky_def():
806 x = np.random.default_rng(sb + 5).standard_normal((4, 5))
807 return _max_abs(ag.leaky_relu(T(x), 0.1).data, np.where(x > 0, x, 0.1 * x)) < TOL_PROP, ""
808
809 def sigmoid_def():
810 x = np.random.default_rng(sb + 6).standard_normal((4, 5))
811 return _max_abs(ag.sigmoid(T(x)).data, 1 / (1 + np.exp(-x))) < 1e-6, ""
812
813 def tanh_def():
814 x = np.random.default_rng(sb + 7).standard_normal((4, 5))
815 return _max_abs(ag.tanh(T(x)).data, np.tanh(x)) <
…[truncated 15813 chars]/app/run_checks.py
1113 rng = np.random.default_rng(seed)
1114 return [T(rng.standard_normal((4, 3)), requires_grad=True), T(rng.standard_normal(3), requires_grad=True)]
1115
1116 def run(name, opt_fn, ref_step, nsteps, gseed):
1117 try:
1118 ps = mk(sb + 500)
1119 ref = [p.data.copy() for p in ps]
1120 o = opt_fn(ps)
1121 st = {}
1122 for s in range(nsteps):
1123 grng = np.random.default_rng(gseed + s)
1124 grads = [grng.standard_normal(p.data.shape) for p in ps]
1125 for p, g in zip(ps, grads):
1126 p.grad = g.copy()
1127 o.step()
1128 ref = [ref_step(i, ref[i], grads[i].copy(), s + 1, st) for i in range(len(ps))]
1129 err = max(_max_abs(p.data, ref[i]) for i, p in enumerate(ps))
1130 results.append((f"optim/{name}", err < TOL_OPT, f"max_abs={err:.1e}"))
1131 except Exception as e: # noqa: BLE001
1132 results.append((f"optim/{name}", False, f"raised {type(e).__name__}: {e}"))
1133
1134 def sgd_step(lr, mu, wd):
1135 def f(i, p, g, t, st):
1136 v = st.get(("v", i), np.zeros_like(p))
1137 g = g + wd * p
1138 v = mu * v + g
1139 st[("v", i)] = v
1140 return p - lr * v
1141 return f
1142
1143 def adagrad_step(lr, eps, wd):
1144 def f(i, p, g, t, st):
1145 s = st.
…[truncated 11091 chars]/app/run_checks.py
1352 def init_scales(m):
1353 nps = m.named_params()
1354 for nm, p in nps.items():
1355 if nm.endswith("wq.scale"):
1356 wkey = nm[:-len("wq.scale")] + "weight"
1357 p.data[...] = _imax(2.0 * float(np.mean(np.abs(nps[wkey].data))), 1e-2)
1358 elif nm.endswith("aq.scale"):
1359 p.data[...] = 1.0
1360
1361 def overfit_loss(m, opt_build, X, y, nsteps):
1362 # repeatedly train on ONE fixed (X, y) batch: a correct forward+backward+optimizer drives
1363 # the loss DOWN on a memorizable batch. A broken backward keeps it flat/NaN.
1364 m.set_training(True)
1365 g = opt.split_decay_params(m)
1366 o = opt_build(g)
1367 losses = []
1368 for _ in range(nsteps):
1369 loss = ag.cross_entropy(m(X), y)
1370 o.zero_grad(); loss.backward(); opt.clip_grad_norm(m.params(), 2.0); o.step()
1371 losses.append(float(loss.data))
1372 return losses
1373
1374 def qcnn_run():
1375 m = nn.QuantCNN(6, widths=(8, 16), seed=sb + 5)
1376 init_scales(m)
1377 rng = np.random.default_rng(sb)
1378 X = rng.standard_normal((8, 3, 16, 16)); y = rng.integers(0, 6, size=8)
1379 return overfit_loss(m, lambda g: opt.Adam(
1380 [{"params": g["decay"], "weight_decay": 0.0}, {"params": g["no_decay"], "weight_decay": 0.0}],
1381 lr=5e-3), X, y, 20)
1382
1383
…[truncated 4742 chars]python3 -c "import numpy; print(numpy.__version__)"
2.2.6
/app/submission/autograd.py
TODO: implement the topological walk + seeding. (The per-op ``_backward`` closures do the
local accumulation; here you only orchestrate the order + seed ``self.grad``.)
"""
raise NotImplementedError("Tensor.backward") TODO: implement the topological walk + seeding. (The per-op ``_backward`` closures do the
local accumulation; here you only orchestrate the order + seed ``self.grad``.)
"""
topo: List[Tensor] = []
visited = set()
def build(t: "Tensor"):
if id(t) in visited:
return
visited.add(id(t))
for p in t._parents:
build(p)
topo.append(t)
build(self)
if grad is None:
grad = np.ones_like(self.data)
self._accum(_as_f64(grad))
for t in reversed(topo):
if t.grad is not None:
t._backward()The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("add backward") # TODO def _bw():
g = out.grad
a._accum(_unbroadcast(g, a.data.shape))
b._accum(_unbroadcast(g, b.data.shape))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("sub backward") # TODO def _bw():
g = out.grad
a._accum(_unbroadcast(g, a.data.shape))
b._accum(_unbroadcast(-g, b.data.shape))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("mul backward") # TODO def _bw():
g = out.grad
a._accum(_unbroadcast(g * b.data, a.data.shape))
b._accum(_unbroadcast(g * a.data, b.data.shape))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("div backward") # TODO def _bw():
g = out.grad
a._accum(_unbroadcast(g / b.data, a.data.shape))
b._accum(_unbroadcast(-g * a.data / (b.data * b.data), b.data.shape))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("power backward") # TODO def _bw():
a._accum(out.grad * p * a.data ** (p - 1.0))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("maximum backward") # TODO def _bw():
g = out.grad
ad, bd = a.data, b.data
ga = np.where(ad > bd, 1.0, np.where(ad == bd, 0.5, 0.0))
a._accum(_unbroadcast(g * ga, a.data.shape))
b._accum(_unbroadcast(g * (1.0 - ga), b.data.shape))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("minimum backward") # TODO def _bw():
g = out.grad
ad, bd = a.data, b.data
ga = np.where(ad < bd, 1.0, np.where(ad == bd, 0.5, 0.0))
a._accum(_unbroadcast(g * ga, a.data.shape))
b._accum(_unbroadcast(g * (1.0 - ga), b.data.shape))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("where backward") # TODO def _bw():
g = out.grad
a._accum(_unbroadcast(np.where(c, g, 0.0), a.data.shape))
b._accum(_unbroadcast(np.where(c, 0.0, g), b.data.shape))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("exp backward") # TODO def _bw():
a._accum(out.grad * out.data)The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("log backward") # TODO def _bw():
a._accum(out.grad / a.data)The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("sqrt backward") # TODO def _bw():
a._accum(out.grad * 0.5 / out.data)The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("abs backward") # TODO def _bw():
a._accum(out.grad * np.sign(a.data))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("clip backward") # TODO def _bw():
a._accum(out.grad * ((a.data > lo) & (a.data < hi)))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
class Tensor:
__slots__ = ("data", "grad", "requires_grad", "_backward", "_parents")def _expand_reduced(g, a_shape, axis, keepdims):
"""Broadcast a reduced gradient ``g`` back up to ``a_shape``, re-inserting reduced axes."""
g = np.asarray(g, dtype=np.float64)
if axis is None:
return np.broadcast_to(g, a_shape).astype(np.float64).copy()
if not keepdims:
ax = axis if isinstance(axis, tuple) else (axis,)
ksh = list(a_shape)
for x in ax:
ksh[x] = 1
g = g.reshape(ksh)
return np.broadcast_to(g, a_shape).astype(np.float64).copy()
class Tensor:
__slots__ = ("data", "grad", "requires_grad", "_backward", "_parents")The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("sum backward") # TODO def _bw():
a._accum(_expand_reduced(out.grad, a.data.shape, axis, keepdims))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("mean backward") # TODO def _bw():
n = a.data.size / out.data.size
a._accum(_expand_reduced(out.grad, a.data.shape, axis, keepdims) / n)The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("max backward") # TODO def _bw():
outd = _expand_reduced(out.data, a.data.shape, axis, keepdims)
gg = _expand_reduced(out.grad, a.data.shape, axis, keepdims)
mask = (a.data == outd).astype(np.float64)
cnt = _expand_reduced(mask.sum(axis=axis, keepdims=True), a.data.shape, None, True) \
if axis is None else np.broadcast_to(mask.sum(axis=axis, keepdims=True), a.data.shape)
a._accum(gg * mask / cnt)The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
mask = (a.data == outd).astype(np.float64)
cnt = _expand_reduced(mask.sum(axis=axis, keepdims=True), a.data.shape, None, True) \
if axis is None else np.broadcast_to(mask.sum(axis=axis, keepdims=True), a.data.shape)
a._accum(gg * mask / cnt) mask = (a.data == outd).astype(np.float64)
cnt = np.broadcast_to(mask.sum(axis=axis, keepdims=True), a.data.shape)
a._accum(gg * mask / cnt)The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("min backward") # TODO def _bw():
outd = _expand_reduced(out.data, a.data.shape, axis, keepdims)
gg = _expand_reduced(out.grad, a.data.shape, axis, keepdims)
mask = (a.data == outd).astype(np.float64)
cnt = np.broadcast_to(mask.sum(axis=axis, keepdims=True), a.data.shape)
a._accum(gg * mask / cnt)The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("concat backward") # TODO def _bw():
g = out.grad
idx = 0
for t in ts:
n = t.data.shape[axis]
sl = [slice(None)] * g.ndim
sl[axis] = slice(idx, idx + n)
t._accum(g[tuple(sl)])
idx += nThe file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("stack backward") # TODO def _bw():
g = out.grad
for i, t in enumerate(ts):
sl = [slice(None)] * g.ndim
sl[axis] = i
t._accum(g[tuple(sl)])The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("transpose backward") # TODO def _bw():
if axes is None:
a._accum(np.transpose(out.grad))
else:
inv = np.argsort(axes)
a._accum(np.transpose(out.grad, tuple(inv)))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("reshape backward") # TODO def _bw():
a._accum(out.grad.reshape(a.data.shape))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("getitem backward") # TODO def _bw():
dz = np.zeros_like(a.data)
np.add.at(dz, idx, out.grad)
a._accum(dz)The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("matmul backward") # TODO def _bw():
g = out.grad
da = g @ np.swapaxes(b.data, -1, -2)
db = np.swapaxes(a.data, -1, -2) @ g
a._accum(_unbroadcast(da, a.data.shape))
b._accum(_unbroadcast(db, b.data.shape))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("relu backward") # TODO def _bw():
a._accum(out.grad * (a.data > 0.0))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("leaky_relu backward") # TODO def _bw():
a._accum(out.grad * np.where(a.data > 0.0, 1.0, slope))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("sigmoid backward") # TODO def _bw():
a._accum(out.grad * out.data * (1.0 - out.data))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("tanh backward") # TODO def _bw():
a._accum(out.grad * (1.0 - out.data * out.data))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("gelu backward") # TODO def _bw():
pdf = np.exp(-0.5 * a.data * a.data) / np.sqrt(2.0 * np.pi)
a._accum(out.grad * (cdf + a.data * pdf))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("softmax backward") # TODO def _bw():
g = out.grad
s = out.data
a._accum(s * (g - (g * s).sum(axis=axis, keepdims=True)))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("log_softmax backward") # TODO def _bw():
g = out.grad
sm = np.exp(out.data)
a._accum(g - sm * g.sum(axis=axis, keepdims=True))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("cross_entropy backward") # TODO def _bw():
p = np.exp(logp)
p[np.arange(n), t] -= 1.0
logits._accum(out.grad * p / n)The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("mse_loss backward") # TODO def _bw():
pred._accum(out.grad * (2.0 / pred.data.size) * (pred.data - tgt))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("layernorm backward") # TODO def _bw():
g = out.grad
lead = tuple(range(g.ndim - 1))
gamma._accum(_unbroadcast((g * xhat).sum(axis=lead), gamma.data.shape))
beta._accum(_unbroadcast(g.sum(axis=lead), beta.data.shape))
gx = g * gamma.data
dx = inv / D * (D * gx - gx.sum(axis=-1, keepdims=True)
- xhat * (gx * xhat).sum(axis=-1, keepdims=True))
x._accum(dx)The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("conv2d backward") # TODO (dW, dbias, dx via _col2im) def _bw():
g = out.grad.reshape(N, Cout, OH * OW)
dW = np.einsum("nop,ncp->oc", g, cols).reshape(Cout, Cin, KH, KW)
weight._accum(dW)
if has_bias:
bias._accum(g.sum(axis=(0, 2)))
dcols = np.einsum("oc,nop->ncp", Wm, g)
dxp = _col2im(dcols, xp.shape, KH, KW, stride, OH, OW)
dx = dxp[:, :, pad:pad + H, pad:pad + W] if pad > 0 else dxp
x._accum(dx)The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("avgpool2d backward") # TODO def _bw():
g = out.grad
gg = g / (k * k)
exp = np.broadcast_to(gg[:, :, :, None, :, None], (N, C, OH, k, OW, k))
dx = np.zeros((N, C, H, W), dtype=np.float64)
dx[:, :, :OH * k, :OW * k] = exp.reshape(N, C, OH * k, OW * k)
x._accum(dx)The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("maxpool2d backward") # TODO def _bw():
g = out.grad
out_exp = out.data[:, :, :, None, :, None]
mask = (xr == out_exp).astype(np.float64)
cnt = mask.sum(axis=(3, 5), keepdims=True)
contrib = mask * (g[:, :, :, None, :, None] / cnt)
dx = np.zeros((N, C, H, W), dtype=np.float64)
dx[:, :, :OH * k, :OW * k] = contrib.reshape(N, C, OH * k, OW * k)
x._accum(dx)The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("batchnorm2d backward") # TODO def _bw():
g = out.grad
gamma._accum((g * xhat).sum(axis=(0, 2, 3)))
beta._accum(g.sum(axis=(0, 2, 3)))
if training:
gx = g * g_
sgx = gx.sum(axis=(0, 2, 3), keepdims=True)
sgxx = (gx * xhat).sum(axis=(0, 2, 3), keepdims=True)
dx = inv / M * (M * gx - sgx - xhat * sgxx)
else:
dx = g * g_ * inv
x._accum(dx)The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
# ============================================================ LSQ fake-quant (DOMAIN; the crux)
def batchnorm1d(x, gamma, beta, eps: float = 1e-5,
running_mean=None, running_var=None, momentum: float = 0.1,
training: bool = True) -> Tensor:
"""BatchNorm over the batch axis N for a (N, C) input; gamma,beta:(C,). Same semantics as
``batchnorm2d`` (population variance, unbiased running var), reduced over axis 0 only."""
x, gamma, beta = _ensure(x), _ensure(gamma), _ensure(beta)
N, C = x.data.shape
M = N
xd = x.data
if training:
mu = xd.mean(axis=0, keepdims=True)
xc = xd - mu
var = (xc * xc).mean(axis=0, keepdims=True)
inv = 1.0 / np.sqrt(var + eps)
xhat = xc * inv
if running_mean is not None and running_var is not None:
ubvar = var * (M / _bi_max(1, M - 1))
running_mean[:] = (1 - momentum) * running_mean + momentum * mu.reshape(-1)
running_var[:] = (1 - momentum) * running_var + momentum * ubvar.reshape(-1)
else:
mu = running_mean.reshape(1, C)
var = running_var.reshape(1, C)
inv = 1.0 / np.sqrt(var + eps)
xhat = (xd - mu) * inv
g_ = gamma.data.reshape(1, C)
out = _mk(g_ * xhat + beta.data.reshape(1, C), (x, gamma, beta),
x.requires_grad or gamma.requires_grad or beta.requires_grad)
def _bw():
g = out.grad
gamma._accum((g * xhat).sum(axis=0))
beta._accum(g.sum(axis=0))
if training:
gx = g * g_
sgx = gx.sum(axis=0,
…[truncated 333 chars]The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("fake_quant backward (STE + LSQ scale gradient)") # TODO def _bw():
g = out.grad
below = r < Qn
above = r > Qp
mid = (~below) & (~above)
v._accum(g * mid)
ds_elem = np.where(mid, np.round(r) - r, np.where(below, Qn, Qp))
ds = float((g * ds_elem).sum()) * grad_scale
scale._accum(np.asarray(ds).reshape(scale.data.shape))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("square backward") # TODO def _bw():
a._accum(out.grad * 2.0 * a.data)The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("rsqrt backward") # TODO def _bw():
a._accum(out.grad * (-0.5) * a.data ** (-1.5))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("reciprocal backward") # TODO def _bw():
a._accum(out.grad * (-out.data * out.data))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("var backward") # TODO def _bw():
count = a.data.size / out.data.size
gg = _expand_reduced(out.grad, a.data.shape, axis, keepdims)
a._accum(gg * (2.0 / count) * xc)The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("std backward") # TODO def _bw():
count = a.data.size / out.data.size
gg = _expand_reduced(out.grad, a.data.shape, axis, keepdims)
a._accum(gg * xc / (count * sd))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("gather backward") # TODO def _bw():
dz = np.zeros_like(a.data)
ax = axis % a.data.ndim
grids = list(np.indices(idx.shape))
grids[ax] = idx
np.add.at(dz, tuple(grids), out.grad)
a._accum(dz)The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("pad2d backward") # TODO def _bw():
g = out.grad
if pad > 0:
H, W = x.data.shape[2], x.data.shape[3]
g = g[:, :, pad:pad + H, pad:pad + W]
x._accum(g)The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("softplus backward") # TODO def _bw():
sig = 1.0 / (1.0 + np.exp(-bx))
a._accum(out.grad * sig)The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("silu backward") # TODO def _bw():
a._accum(out.grad * (sig + a.data * sig * (1.0 - sig)))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("mish backward") # TODO def _bw():
sig = 1.0 / (1.0 + np.exp(-x))
tsp = np.tanh(sp)
a._accum(out.grad * (tsp + x * (1.0 - tsp * tsp) * sig))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("elu backward") # TODO def _bw():
a._accum(out.grad * np.where(x > 0.0, 1.0, alpha * np.exp(np.minimum(x, 0.0))))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("hardtanh backward") # TODO def _bw():
a._accum(out.grad * ((a.data > lo) & (a.data < hi)))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("hardsigmoid backward") # TODO def _bw():
a._accum(out.grad / 6.0 * ((z > 0.0) & (z < 1.0)))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("groupnorm backward") # TODO def _bw():
g = out.grad
gam = gamma.data.reshape(1, C, 1, 1)
gamma._accum((g * xhat).sum(axis=(0, 2, 3)))
beta._accum(g.sum(axis=(0, 2, 3)))
M = cg * H * W
gx = (g * gam).reshape(N, G, M)
xh = xhat.reshape(N, G, M)
sgx = gx.sum(axis=2, keepdims=True)
sgxx = (gx * xh).sum(axis=2, keepdims=True)
dx = (inv / M * (M * gx - sgx - xh * sgxx)).reshape(N, C, H, W)
x._accum(dx)The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("fake_quant_per_channel backward (STE + per-channel scale grad)") # TODO def _bw():
g = out.grad
below = r < Qn
above = r > Qp
mid = (~below) & (~above)
v._accum(g * mid)
ds_elem = np.where(mid, np.round(r) - r, np.where(below, Qn, Qp))
axes = tuple(i for i in range(v.data.ndim) if i != (axis % v.data.ndim))
ds = (g * ds_elem).sum(axis=axes) * grad_scale
scale._accum(ds.reshape(scale.data.shape))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("fake_quant_affine backward (STE + scale grad on shifted grid)") # TODO def _bw():
g = out.grad
below = r < Qn
above = r > Qp
mid = (~below) & (~above)
v._accum(g * mid)
ds_elem = np.where(mid, (q - z) - (r - z), np.where(below, Qn - z, Qp - z))
ds = float((g * ds_elem).sum()) * grad_scale
scale._accum(np.asarray(ds).reshape(scale.data.shape))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("cumsum backward") # TODO def _bw():
g = out.grad
a._accum(np.flip(np.cumsum(np.flip(g, axis=axis), axis=axis), axis=axis))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("flip backward") # TODO def _bw():
a._accum(np.flip(out.grad, axis=axis))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("logsumexp backward") # TODO def _bw():
g = out.grad if keepdims else np.expand_dims(out.grad, axis)
a._accum(sm * g)The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("logaddexp backward") # TODO def _bw():
g = out.grad
wa = np.exp(a.data - out_data)
wb = np.exp(b.data - out_data)
a._accum(_unbroadcast(g * wa, a.data.shape))
b._accum(_unbroadcast(g * wb, b.data.shape))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("l2_normalize backward") # TODO def _bw():
g = out.grad
proj = (y * g).sum(axis=axis, keepdims=True)
a._accum((g - y * proj) / nrm)The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("rms_norm backward") # TODO def _bw():
g = out.grad
gg = g * gamma.data
dx = inv * gg - (xd * inv ** 3 / D) * (gg * xd).sum(axis=-1, keepdims=True)
x._accum(dx)
lead = tuple(range(g.ndim - 1))
gamma._accum(_unbroadcast((g * xhat).sum(axis=lead), gamma.data.shape))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("instance_norm backward") # TODO def _bw():
g = out.grad
gamma._accum((g * xhat).sum(axis=(0, 2, 3)))
beta._accum(g.sum(axis=(0, 2, 3)))
gx = (g * g_).reshape(N, C, M)
xh = xhat.reshape(N, C, M)
sgx = gx.sum(axis=2, keepdims=True)
sgxx = (gx * xh).sum(axis=2, keepdims=True)
dx = (inv / M * (M * gx - sgx - xh * sgxx)).reshape(N, C, H, W)
x._accum(dx)The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("huber_loss backward") # TODO def _bw():
gd = np.where(quad, diff, delta * np.sign(diff)) / n
pred._accum(out.grad * gd)The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("kl_div backward") # TODO def _bw():
log_p._accum(out.grad * (-q / n))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("embedding backward") # TODO def _bw():
dz = np.zeros_like(weight.data)
np.add.at(dz, idx, out.grad)
weight._accum(dz)The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("conv2d_gen backward (grouped/dilated dW/db/dx)") # TODO def _bw():
g = out.grad.reshape(N, Cout, OH * OW)
g_g = g.reshape(N, groups, cog, OH * OW)
dW = np.einsum("ngop,ngcp->goc", g_g, cols_g).reshape(Cout, cig, KH, KW)
weight._accum(dW)
if has_bias:
bias._accum(g.sum(axis=(0, 2)))
dcols_g = np.einsum("goc,ngop->ngcp", Wm, g_g)
dcols = dcols_g.reshape(N, Cin * KH * KW, OH * OW)
dxp = _col2im_dil(dcols, xp.shape, KH, KW, stride, dilation, OH, OW)
dx = dxp[:, :, pad:pad + H, pad:pad + W] if pad > 0 else dxp
x._accum(dx)The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("conv_transpose2d backward") # TODO def _bw():
g = out.grad
if has_bias:
bias._accum(g.sum(axis=(0, 2, 3)))
if pad > 0:
gfull = np.zeros((N, Cout, OHf, OWf), dtype=np.float64)
gfull[:, :, pad:OHf - pad, pad:OWf - pad] = g
else:
gfull = g
gcontrib = np.empty((N, Cout, H, W, KH, KW), dtype=np.float64)
for i in range(KH):
for j in range(KW):
gcontrib[:, :, :, :, i, j] = gfull[:, :, i:i + stride * H:stride, j:j + stride * W:stride]
dx = np.einsum("noijKL,coKL->ncij", gcontrib, Wm)
dW = np.einsum("ncij,noijKL->coKL", xd, gcontrib)
x._accum(dx)
weight._accum(dW)The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("avgpool2d_s backward") # TODO def _bw():
g = out.grad
dxp = np.zeros((N, C, Hp, Wp), dtype=np.float64)
inv_area = 1.0 / (k * k)
for oi in range(OH):
for oj in range(OW):
dxp[:, :, oi * stride:oi * stride + k, oj * stride:oj * stride + k] += \
g[:, :, oi, oj][:, :, None, None] * inv_area
dx = dxp[:, :, pad:pad + H, pad:pad + W] if pad > 0 else dxp
x._accum(dx)The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("maxpool2d_s backward") # TODO def _bw():
g = out.grad
dxp = np.zeros((N, C, Hp, Wp), dtype=np.float64)
narange = np.arange(N)[:, None]
carange = np.arange(C)[None, :]
for oi in range(OH):
for oj in range(OW):
ii = oi * stride + argi[:, :, oi, oj]
jj = oj * stride + argj[:, :, oi, oj]
np.add.at(dxp, (narange, carange, ii, jj), g[:, :, oi, oj])
dx = dxp[:, :, pad:pad + H, pad:pad + W] if pad > 0 else dxp
x._accum(dx)The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/autograd.py
def _bw():
raise NotImplementedError("fake_quant_lsq_plus backward (STE v-grad + beta-grad + LSQ scale grad)") # TODO def _bw():
g = out.grad
below = r < Qn
above = r > Qp
mid = (~below) & (~above)
v._accum(g * mid)
dbeta = float((g * (~mid)).sum())
beta._accum(np.asarray(dbeta).reshape(beta.data.shape))
ds_elem = np.where(mid, np.round(r) - r, np.where(below, Qn, Qp))
ds = float((g * ds_elem).sum()) * grad_scale
scale._accum(np.asarray(ds).reshape(scale.data.shape))The file /app/submission/autograd.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/data.py
# TODO: cast to float, scale to [0,1], subtract NORM_MEAN, divide by NORM_STD.
raise NotImplementedError("normalize")return (X.astype(np.float64) / 255.0 - NORM_MEAN) / NORM_STD
The file /app/submission/data.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/data.py
# TODO: optionally shuffle an index array, then yield contiguous batch_size slices of it.
raise NotImplementedError("iter_minibatches") n = Xn.shape[0]
idx = np.arange(n)
if shuffle:
rng.shuffle(idx)
for start in range(0, n, batch_size):
sel = idx[start:start + batch_size]
yield Xn[sel], y[sel]The file /app/submission/data.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/optim.py
def zero_grad(self):
for p in self.params:
p.grad = None def zero_grad(self):
for p in self.params:
p.grad = None
def _iter(self):
"""Yield (global_index, param, weight_decay) honouring per-group wd overrides."""
i = 0
for group, wd_override in self._groups:
wd = self.weight_decay if wd_override is None else wd_override
for p in group:
yield i, p, wd
i += 1The file /app/submission/optim.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/optim.py
# TODO: SGD(+momentum, +per-group weight decay) update. Iterate self._groups; for each
# group use wd = self.weight_decay if override is None else override; keep a momentum
# buffer per parameter (self._v is indexed in self.params order).
raise NotImplementedError("SGD.step") for i, p, wd in self._iter():
if p.grad is None:
continue
g = p.grad + wd * p.data
self._v[i] = self.momentum * self._v[i] + g
p.data -= self.lr * self._v[i]The file /app/submission/optim.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/optim.py
# TODO: AdamW with bias correction + DECOUPLED, per-group weight decay (see docstring).
raise NotImplementedError("AdamW.step") self.t += 1
t = self.t
for i, p, wd in self._iter():
if p.grad is None:
continue
g = p.grad
self._m[i] = self.b1 * self._m[i] + (1 - self.b1) * g
self._v[i] = self.b2 * self._v[i] + (1 - self.b2) * (g * g)
mhat = self._m[i] / (1 - self.b1 ** t)
vhat = self._v[i] / (1 - self.b2 ** t)
p.data -= self.lr * (mhat / (np.sqrt(vhat) + self.eps) + wd * p.data)The file /app/submission/optim.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/optim.py
# TODO: Adam with bias correction + COUPLED, per-group weight decay (see docstring).
raise NotImplementedError("Adam.step") self.t += 1
t = self.t
for i, p, wd in self._iter():
if p.grad is None:
continue
g = p.grad + wd * p.data
self._m[i] = self.b1 * self._m[i] + (1 - self.b1) * g
self._v[i] = self.b2 * self._v[i] + (1 - self.b2) * (g * g)
mhat = self._m[i] / (1 - self.b1 ** t)
vhat = self._v[i] / (1 - self.b2 ** t)
p.data -= self.lr * mhat / (np.sqrt(vhat) + self.eps)The file /app/submission/optim.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/optim.py
# TODO: RMSprop update with per-group weight decay (see docstring).
raise NotImplementedError("RMSprop.step") for i, p, wd in self._iter():
if p.grad is None:
continue
g = p.grad + wd * p.data
self._s[i] = self.alpha * self._s[i] + (1 - self.alpha) * (g * g)
p.data -= self.lr * g / (np.sqrt(self._s[i]) + self.eps)The file /app/submission/optim.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/optim.py
# TODO: Adagrad update with per-group weight decay (see docstring).
raise NotImplementedError("Adagrad.step") for i, p, wd in self._iter():
if p.grad is None:
continue
g = p.grad + wd * p.data
self._s[i] = self._s[i] + g * g
p.data -= self.lr * g / (np.sqrt(self._s[i]) + self.eps)The file /app/submission/optim.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/optim.py
# TODO: Lion update with per-group (decoupled) weight decay (see docstring).
raise NotImplementedError("Lion.step") for i, p, wd in self._iter():
if p.grad is None:
continue
g = p.grad
upd = np.sign(self.b1 * self._m[i] + (1 - self.b1) * g)
p.data -= self.lr * (upd + wd * p.data)
self._m[i] = self.b2 * self._m[i] + (1 - self.b2) * gThe file /app/submission/optim.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/optim.py
# TODO: implement the Adamax update honouring per-group weight decay.
raise NotImplementedError("Adamax.step") self.t += 1
t = self.t
for i, p, wd in self._iter():
if p.grad is None:
continue
g = p.grad + wd * p.data
self._m[i] = self.b1 * self._m[i] + (1 - self.b1) * g
self._u[i] = np.maximum(self.b2 * self._u[i], np.abs(g))
p.data -= (self.lr / (1 - self.b1 ** t)) * self._m[i] / (self._u[i] + self.eps)The file /app/submission/optim.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/optim.py
# TODO: implement the NAdam update honouring per-group weight decay.
raise NotImplementedError("NAdam.step") self.t += 1
t = self.t
for i, p, wd in self._iter():
if p.grad is None:
continue
g = p.grad + wd * p.data
self._m[i] = self.b1 * self._m[i] + (1 - self.b1) * g
self._v[i] = self.b2 * self._v[i] + (1 - self.b2) * (g * g)
mhat = self.b1 * self._m[i] / (1 - self.b1 ** (t + 1)) + (1 - self.b1) * g / (1 - self.b1 ** t)
vhat = self._v[i] / (1 - self.b2 ** t)
p.data -= self.lr * mhat / (np.sqrt(vhat) + self.eps)The file /app/submission/optim.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/optim.py
# TODO: implement the Adadelta update honouring per-group weight decay.
raise NotImplementedError("Adadelta.step") for i, p, wd in self._iter():
if p.grad is None:
continue
g = p.grad + wd * p.data
self._s[i] = self.rho * self._s[i] + (1 - self.rho) * (g * g)
delta = np.sqrt(self._d[i] + self.eps) / np.sqrt(self._s[i] + self.eps) * g
p.data -= self.lr * delta
self._d[i] = self.rho * self._d[i] + (1 - self.rho) * (delta * delta)The file /app/submission/optim.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/optim.py
# TODO: implement the Nesterov-SGD update honouring per-group weight decay.
raise NotImplementedError("NesterovSGD.step") for i, p, wd in self._iter():
if p.grad is None:
continue
g = p.grad + wd * p.data
self._v[i] = self.momentum * self._v[i] + g
p.data -= self.lr * (g + self.momentum * self._v[i])The file /app/submission/optim.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/optim.py
# TODO: compute the global L2 norm over all param grads; scale in place if it exceeds max_norm.
raise NotImplementedError("clip_grad_norm") total = math.sqrt(sum(float((p.grad ** 2).sum()) for p in params if p.grad is not None))
if total > max_norm:
scale = max_norm / (total + 1e-6)
for p in params:
if p.grad is not None:
p.grad *= scale
return totalThe file /app/submission/optim.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/optim.py
``base_lr`` to ``min_lr`` over the remaining (total - warmup) steps. Return the LR at ``step``."""
# TODO
raise NotImplementedError("cosine_lr") ``base_lr`` to ``min_lr`` over the remaining (total - warmup) steps. Return the LR at ``step``."""
if warmup > 0 and step < warmup:
return base_lr * (step + 1) / warmup
prog = min(1.0, max(0.0, (step - warmup) / max(1, total - warmup)))
return min_lr + 0.5 * (base_lr - min_lr) * (1 + math.cos(math.pi * prog))The file /app/submission/optim.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/optim.py
constant ``base_lr``."""
# TODO
raise NotImplementedError("warmup_lr") constant ``base_lr``."""
if step < warmup:
return base_lr * (step + 1) / warmup
return base_lrThe file /app/submission/optim.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/optim.py
"""Step decay: base_lr * gamma ** (step // step_size)."""
# TODO
raise NotImplementedError("step_lr") """Step decay: base_lr * gamma ** (step // step_size)."""
return base_lr * gamma ** (step // step_size)The file /app/submission/optim.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/optim.py
"""Exponential decay: base_lr * gamma ** step."""
# TODO
raise NotImplementedError("exponential_lr") """Exponential decay: base_lr * gamma ** step."""
return base_lr * gamma ** stepThe file /app/submission/optim.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/optim.py
"""Polynomial decay: min_lr + (base_lr - min_lr) * (1 - step/total)^power (clamped at total)."""
# TODO
raise NotImplementedError("polynomial_lr") """Polynomial decay: min_lr + (base_lr - min_lr) * (1 - step/total)^power (clamped at total)."""
prog = min(1.0, max(0.0, step / total))
return min_lr + (base_lr - min_lr) * ((1 - prog) ** power)The file /app/submission/optim.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/optim.py
"""Linear decay from base_lr to min_lr over ``total`` steps, then constant at min_lr."""
# TODO
raise NotImplementedError("linear_lr") """Linear decay from base_lr to min_lr over ``total`` steps, then constant at min_lr."""
prog = min(1.0, max(0.0, step / total))
return base_lr + (min_lr - base_lr) * progThe file /app/submission/optim.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/optim.py
"""base_lr * gamma^(number of milestones <= step)."""
# TODO
raise NotImplementedError("multistep_lr") """base_lr * gamma^(number of milestones <= step)."""
return base_lr * gamma ** sum(1 for m in milestones if step >= m)The file /app/submission/optim.py has been updated successfully. (file state is current in your context , no need to Read it back)
/app/submission/optim.py
"""Cosine annealing with warm restarts of fixed ``period`` (cosine within each period)."""
# TODO
raise NotImplementedError("cosine_restarts_lr") """Cosine annealing with warm restarts of fixed ``period`` (cosine within each period)."""
t = step % period
return min_lr + 0.5 * (base_lr - min_lr) * (1 + math.cos(math.pi * (t / period)))The file /app/submission/optim.py has been updated successfully. (file state is current in your context , no need to Read it back)
cd /app/submission && python3 - <<'PY'
import re
f='nn.py'
s=open(f).read()
reps = [
(' # TODO: y = x @ weight (+ bias). Use ag.matmul / ag.add.\n raise NotImplementedError("Linear.forward")',
' y = ag.matmul(x, self.weight)\n if self.bias is not None:\n y = ag.add(y, self.bias)\n return y'),
(' # TODO: out = x + fc2(gelu(fc1(x)))\n raise NotImplementedError("MLP.forward")',
' return ag.add(x, self.fc2(ag.gelu(self.fc1(x))))'),
(' # TODO: return ag.layernorm(x, self.gamma, self.beta, self.eps)\n raise NotImplementedError("LayerNorm.forward")',
' return ag.layernorm(x, self.gamma, self.beta, self.eps)'),
(' # TODO: return ag.conv2d(x, self.weight, bias=self.bias, stride=self.stride, pad=self.padding)\n raise NotImplementedError("Conv2d.forward")',
' return ag.conv2d(x, self.weight, bias=self.bias, stride=self.stride, pad=self.padding)'),
(' # TODO: return ag.maxpool2d(x, self.k)\n raise NotImplementedError("MaxPool2d.forward")',
' return ag.maxpool2d(x, self.k)'),
(' # TODO: return ag.avgpool2d(x, self.k)\n raise NotImplementedError("AvgPool2d.forward")',
' return ag.avgpool2d(x, self.k)'),
(' # TODO: return ag.fake_quant(w, self.scale, W_NEG, W_POS)\n raise NotImplementedError("WeightQuant.quant")',
' return ag.fake_quant(w, self.scale, W_NEG, W_POS)'),
(' # TODO: return ag.fake_quant(x, self.
…[truncated 1354 chars]ok1
python3 - <<'PY'
f='nn.py'
s=open(f).read()
reps = [
(' # TODO: run the stem + blocks, global-average-pool, then the (2-bit) linear head.\n raise NotImplementedError("QuantCNN.forward")',
''' h = self.stem(X)
for b in self.blocks:
h = b(h)
N = h.shape[0]
full = h.shape[2]
pooled = ag.avgpool2d(h, full)
feat = ag.reshape(pooled, (N, self.feat))
wq = self.head_wq.quant(self.head_w)
return ag.add(ag.matmul(feat, wq), self.head_b)'''),
(' # TODO: wq = self.wq.quant(self.weight); return ag.add(ag.matmul(x, wq), self.bias)\n raise NotImplementedError("QLinear.forward")',
' wq = self.wq.quant(self.weight)\n return ag.add(ag.matmul(x, wq), self.bias)'),
(''' # TODO: x -> reshape (N, in_dim); for each layer L: h = L["aq"](relu(L["bn"](L["fc"](h))));
# then return self.head(h).
raise NotImplementedError("QuantMLP.forward")''',
''' N = X.shape[0]
h = ag.reshape(X if isinstance(X, Tensor) else Tensor(X), (N, self.in_dim))
for L in self.layers:
h = L["aq"](ag.relu(L["bn"](L["fc"](h))))
return self.head(h)'''),
(' # TODO: return ag.silu(x)\n raise NotImplementedError("SiLU.forward")',
' return ag.silu(x)'),
(' # TODO: return ag.elu(x, self.alpha)\n raise NotImplementedError("ELU.forward")',
' return ag.elu(x, self.alpha)'),
(' # TODO: return ag.groupnorm(x,
…[truncated 501 chars]ok2
python3 - <<'PY'
f='nn.py'
s=open(f).read()
reps = [
(' # TODO: implement the GRU update with the named weights above.\n raise NotImplementedError("GRUCell.forward")',
''' z = ag.sigmoid(ag.add(ag.add(ag.matmul(x, self.Wxz), ag.matmul(h, self.Whz)), self.bz))
r = ag.sigmoid(ag.add(ag.add(ag.matmul(x, self.Wxr), ag.matmul(h, self.Whr)), self.br))
n = ag.tanh(ag.add(ag.add(ag.matmul(x, self.Wxn), ag.matmul(ag.mul(r, h), self.Whn)), self.bn))
one = Tensor(1.0)
return ag.add(ag.mul(ag.sub(one, z), n), ag.mul(z, h))'''),
(' # TODO: implement the LSTM update; return (h2, c2).\n raise NotImplementedError("LSTMCell.forward")',
''' i = ag.sigmoid(ag.add(ag.add(ag.matmul(x, self.Wxi), ag.matmul(h, self.Whi)), self.bi))
fg = ag.sigmoid(ag.add(ag.add(ag.matmul(x, self.Wxf), ag.matmul(h, self.Whf)), self.bf))
g = ag.tanh(ag.add(ag.add(ag.matmul(x, self.Wxg), ag.matmul(h, self.Whg)), self.bg))
o = ag.sigmoid(ag.add(ag.add(ag.matmul(x, self.Wxo), ag.matmul(h, self.Who)), self.bo))
c2 = ag.add(ag.mul(fg, c), ag.mul(i, g))
h2 = ag.mul(o, ag.tanh(c2))
return h2, c2'''),
(' # TODO: return ag.fake_quant_per_channel(w, self.scale, self.neg, self.pos, axis=0)\n raise NotImplementedError("PerChannelWeightQuant.quant")',
' return ag.fake_quant_per_channel(w, self.scale, self.neg, self.pos, axis=0)'),
(' # TODO: return ag.fake_quant_affine(x, self.sc
…[truncated 2971 chars]ok3
grep -rn "NotImplementedError\|# TODO" autograd.py nn.py optim.py data.py | grep -v "raise NotImplementedError$" | head -40; echo "---count---"; grep -c "NotImplementedError" autograd.py nn.py optim.py data.py
nn.py:3:Implement every ``# TODO`` forward (the parameters + ``named_params``/``buffers`` naming are autograd.py:3:YOU MUST IMPLEMENT EVERYTHING MARKED ``# TODO`` / ``raise NotImplementedError`` using numpy optim.py:3:Implement the ``# TODO`` update math. Both optimizers update ``param.data`` IN PLACE using data.py:7:Implement the ``# TODO`` bodies (numpy only). The verifier normalizes the held-out split with the ---count--- nn.py:1 autograd.py:1 data.py:0 optim.py:0
grep -n "NotImplementedError" autograd.py nn.py
nn.py:39: raise NotImplementedError autograd.py:3:YOU MUST IMPLEMENT EVERYTHING MARKED ``# TODO`` / ``raise NotImplementedError`` using numpy
cd /app && timeout 600 python3 run_checks.py 2>&1 | tail -60
fwd/RNNCell_shape OK (shape=(3, 4)) fwd/GRUCell_shape OK (shape=(3, 4)) fwd/LSTMCell_shape OK fwd/PerChannelWeightQuant_grid OK fwd/AffineActQuant_def OK fwd/cumsum_def OK fwd/logsumexp_def OK fwd/l2_normalize_def OK fwd/rms_norm_stats OK fwd/instance_norm_stats OK fwd/conv_transpose2d_shape OK (shape=(2, 4, 7, 7)) fwd/grouped_conv_shape OK (shape=(2, 6, 6, 6)) fwd/embedding_def OK fwd/fake_quant_lsq_plus_grid OK (unique=4) fwd/QuantSepCNN_shape OK fwd/QuantSepCNN_state_dict_roundtrip OK (keys=20) [optim] 25/25 optim/SGD_1step OK (max_abs=0.0e+00) optim/SGD_momentum_5step OK (max_abs=0.0e+00) optim/Adagrad_5step OK (max_abs=0.0e+00) optim/RMSprop_5step OK (max_abs=0.0e+00) optim/Adam_1step OK (max_abs=0.0e+00) optim/Adam_10step_wd OK (max_abs=0.0e+00) optim/AdamW_1step OK (max_abs=0.0e+00) optim/AdamW_10step_wd OK (max_abs=0.0e+00) optim/Lion_5step OK (max_abs=0.0e+00) optim/Adamax …[truncated 1925 chars]
python3 - <<'PY'
import numpy as np
d=np.load('data/train.npz')
print(d.files)
X=d['X']; y=d['y']
print(X.shape, X.dtype, y.shape, y.dtype, X.min(), X.max())
print('classes', np.unique(y), 'counts', np.bincount(y))
PY['X', 'y'] (2500, 3, 16, 16) uint8 (2500,) int64 0 255 classes [0 1 2 3 4 5 6 7 8 9] counts [250 250 250 250 250 250 250 250 250 250]
/app/submission/train.py
def train(data_path, out_path, epochs=30, seed=0, n_classes=10):
# TODO: load + normalize the data; build QuantCNN(n_classes, WIDTHS, seed); initialize the LSQ
# step sizes; build the optimizer with weight decay only on the conv/head weights; run the QAT
# loop (forward -> cross_entropy -> backward -> clip_grad_norm -> step) with an LR schedule;
# keep your best epoch; BAKE the quantized weights; save the state_dict to out_path.
raise NotImplementedError("implement the QAT training loop")from nn import W_NEG, W_POS, A_NEG, A_POS
def _init_weight_scales(model):
nps = model.named_params()
for nm, p in nps.items():
if nm.endswith("wq.scale"):
wkey = nm[:-len("wq.scale")] + "weight"
w = nps[wkey].data
p.data[...] = max(2.0 * float(np.mean(np.abs(w))) / np.sqrt(max(1.0, W_POS)), 1e-3)
def _calibrate_act_scales(model, Xb):
"""Sequentially forward one batch to set each post-relu activation quantizer's step (LSQ init)."""
model.set_training(True)
x = ag.Tensor(Xb)
def pre_aq(block, inp):
return ag.relu(block.bn(block.conv(inp)))
h = pre_aq(model.stem, x)
model.stem.aq.scale.data[...] = max(2.0 * float(np.mean(np.abs(h.data))) / np.sqrt(max(1.0, A_POS)), 1e-3)
h = model.stem.aq(h)
for b in model.blocks:
hp = pre_aq(b, h)
b.aq.scale.data[...] = max(2.0 * float(np.mean(np.abs(hp.data))) / np.sqrt(max(1.0, A_POS)), 1e-3)
h = b.aq(hp)
def _eval_acc(model, Xn, y, bs=250):
model.set_training(False)
correct = 0
for i in range(0, len(Xn), bs):
logits = model(Xn[i:i + bs]).data
correct += int((logits.argmax(1) == y[i:i + bs]).sum())
return correct / len(Xn)
def _bake(model):
"""Replace each conv/head weight with its LSQ-quantized value so it lands on the 2-bit grid."""
nps = model.named_params()
for nm, p in list(nps.items()):
if nm.endswith("wq.scale"):
wkey = nm[:-len("wq.scale")] + "weig
…[truncated 2119 chars]The file /app/submission/train.py has been updated successfully. (file state is current in your context , no need to Read it back)
cd /app/submission && timeout 590 python3 train.py --epochs 45 2>&1 | tail -50
(Bash completed with no output)
ls -la qat_model.npz 2>&1; echo "exit check"
ls: cannot access 'qat_model.npz': No such file or directory exit check
timeout 120 python3 train.py --epochs 2 2>&1 | tail -20
(Bash completed with no output)
timeout 120 python3 -u train.py --epochs 2; echo "EXIT=$?"
No file changes, the agent left the workspace untouched.
=== SyncVals verifier (local re-grade of the agent's own checkpoint) === COMPONENT CONFORMANCE SUITE: 278/278 cases passed (all milestones pass: True) FINAL METRIC: held-out INT2 top-1 = 0.8173 >= threshold 0.7 REWARD = 1 (1 iff every component milestone passes AND the sealed held-out metric clears its threshold) NOTE: the in-sandbox training in the 481MB e2b box was OOM-killed (0-byte checkpoint), so the raw in-sandbox reward was an infrastructure artifact. This reward is from a faithful LOCAL re-train of the agent's OWN train.py (24GB, no OOM) with its own default hyperparameters, then re-graded by the real verifier.
Reproduce this trial: git checkout 2f94510 && PYTHONPATH=src python3 scripts/build_site.py , then open trial/trial_eed29f00b4294eba. Re-running the agent live requires EVAL_PLATFORM_ENABLE_OAUTH_SMOKE=1 and is non-deterministic.
Trial trial_eed29f00b4294eba · verifier authoritative; classifier explanatory.