Record real wd_paired result: q_ttt=1.07974 + add safe_launch guard

The prior session's "wd_paired 1.08009" was a no-op (FS turbulence
rolled train_pr1493.py back to the pre-stacking version mid-session;
log had zero hits on tagged=22). This commit records the first real
wd_paired run on commit 74dc702.

Result (seed 42, real paired-head Muon firing, tagged=22 confirmed
on all 8 ranks): q_ttt 1.07974, q_sw 1.08209, q 1.09891, pre 1.08610.
Stack is -0.00129 BPB vs raw PR1493, -0.00055 vs wd alone.

Also:
- safe_launch.sh: pre-launch guard that asserts HEAD/md5/git-blob/
  symbol-count/working-tree-clean before exec'ing torchrun.
- requirements.txt: add brotli (was implicit via Modal image only;
  attempt 1 of this run crashed in serialize() because local env
  lacked it).
- pr1493_priority_results.md: append wd_paired row (openai#6).
- pr1493_wd_paired_session.md: full session writeup (verification
  methodology, two attempts, learnings, plan-doc vs reality).
- logs/pr1493_wd_paired_s42.{txt,stdout}: real run log + stdout.
- logs/pr1493_wd_paired_s42.{txt,stdout}.attempt1_brotli_crash:
  preserved evidence of the brotli ModuleNotFoundError crash.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

Author: Itssshikhar

logs/pr1493_wd_paired_s42.stdout

@@ -0,0 +1,166 @@
+[safe_launch] OK ts=20260429T095343Z head=74dc7028a06a0f52e2ce23a925ef24404e93ca1b md5=968e5ab744772b096a8f9b521656019d blob=1e4f7b4391f9a82b0ca7f735bbbb0db6eea8e8ad symbols=6
+[safe_launch] exec: torchrun --standalone --nproc_per_node=8 train_pr1493.py
+W0429 09:53:45.032000 35682 torch/distributed/run.py:803] 
+W0429 09:53:45.032000 35682 torch/distributed/run.py:803] *****************************************
+W0429 09:53:45.032000 35682 torch/distributed/run.py:803] Setting OMP_NUM_THREADS environment variable for each process to be 1 in default, to avoid your system being overloaded, please further tune the variable for optimal performance in your application as needed. 
+W0429 09:53:45.032000 35682 torch/distributed/run.py:803] *****************************************
+Hyperparameters:
+  adam_eps: 1e-08
+  adam_wd: 0.02
+  beta1: 0.9
+  beta2: 0.95
+  compressor: brotli
+  data_dir: ./data/
+  datasets_dir: ./data/datasets/fineweb10B_sp8192
+  distributed: True
+  doc_shuffle_bos_id: -1
+  doc_shuffle_enabled: False
+  ema_decay: 0.9965
+  embed_bits: 8
+  embed_clip_sigmas: 20.0
+  embed_lr: 0.6
+  embed_wd: 0.085
+  embedding_dim: 512
+  enable_looping_at: 0.35
+  etlb_clip: 3.0
+  etlb_enabled: False
+  etlb_lr: 0.05
+  etlb_steps: 5
+  eval_num_loops: -1
+  eval_seq_len: 2048
+  eval_stride: 64
+  gptq_calibration_batches: 64
+  gptq_reserve_seconds: 12.0
+  grad_accum_steps: 1
+  grad_clip_norm: 0.3
+  head_lr: 0.008
+  iha_enabled: False
+  iha_mix_v: False
+  is_main_process: True
+  iterations: 20000
+  ln_scale: True
+  local_rank: 0
+  logfile: logs/pr1493_wd_paired_s42.txt
+  logit_softcap: 30.0
+  loop_end: 5
+  loop_start: 3
+  matrix_bits: 6
+  matrix_clip_sigmas: 12.85
+  matrix_lr: 0.022
+  max_wallclock_seconds: 600.0
+  min_lr: 0.0
+  mlp_mult: 4.0
+  model_dim: 512
+  model_path: final_model.pt
+  mtp_steps: 0
+  mtp_weight: 0.0
+  muon_backend_steps: 5
+  muon_beta2: 0.95
+  muon_momentum: 0.99
+  muon_momentum_warmup_start: 0.92
+  muon_momentum_warmup_steps: 1500
+  muon_row_normalize: True
+  muon_wd: 0.095
+  num_heads: 8
+  num_kv_heads: 4
+  num_layers: 11
+  num_loops: 2
+  paired_head_muon_enabled: True
+  parallel_residual_start: 7
+  qk_gain_init: 5.25
+  quantized_model_path: final_model.int6.ptz
+  rank: 0
+  rope_base: 10000.0
+  rope_dims: 16
+  rope_train_seq_len: 2048
+  run_id: pr1493_wd_paired_s42
+  scalar_lr: 0.02
+  seed: 42
+  skip_gates_enabled: True
+  sliding_window_enabled: True
+  tie_embeddings: True
+  tied_embed_init_std: 0.005
+  tied_embed_lr: 0.03
+  tokenizer_path: ./data/tokenizers/fineweb_8192_bpe.model
+  train_batch_tokens: 786432
+  train_files: ./data/datasets/fineweb10B_sp8192/fineweb_train_*.bin
+  train_log_every: 500
+  train_seq_len: 2048
+  ttt_chunk_tokens: 32768
+  ttt_enabled: True
+  ttt_epochs: 5
+  ttt_lr: 0.007
+  ttt_momentum: 0.9
+  val_batch_tokens: 524288
+  val_files: ./data/datasets/fineweb10B_sp8192/fineweb_val_*.bin
+  val_loss_every: 4000
+  vocab_size: 8192
+  warmdown_frac: 0.72
+  warmup_steps: 20
+  wd_sched_high_factor: 1.5
+  wd_sched_hold_frac: 0.4
+  wd_sched_low_factor: 0.65
+  wd_sched_ramp_frac: 0.85
+  wd_schedule_enabled: True
+  world_size: 8
+  xsa_last_n: 11
+train_shards: 128
+val_tokens: 40540160
+model_params:35944536
+muon:paired-head NS enabled for q/k matrices tagged=22
+gptq:reserving 12s, effective=588000ms
+warmup_step: 1/20
+warmup_step: 2/20
+warmup_step: 3/20
+warmup_step: 4/20
+warmup_step: 5/20
+warmup_step: 6/20
+warmup_step: 10/20
+warmup_step: 20/20
+loop_warmup:enabled encoder:[0, 1, 2, 3, 4, 5, 3, 4] decoder:[5, 3, 4, 5, 6, 7, 8, 9, 10]
+loop_warmup_step: 1/20
+loop_warmup_step: 2/20
+loop_warmup_step: 3/20
+loop_warmup_step: 4/20
+loop_warmup_step: 5/20
+loop_warmup_step: 6/20
+loop_warmup_step: 10/20
+loop_warmup_step: 20/20
+0/20000 val_loss: 9.0090 val_bpb: 3.4877
+1/20000 train_loss: 9.0111 train_time: 0.0m tok/s: 8357493
+2/20000 train_loss: 12.3696 train_time: 0.0m tok/s: 8224856
+3/20000 train_loss: 11.1322 train_time: 0.0m tok/s: 8106486
+4/20000 train_loss: 9.4108 train_time: 0.0m tok/s: 8044845
+5/20000 train_loss: 8.3268 train_time: 0.0m tok/s: 8014159
+500/20000 train_loss: 3.3250 train_time: 0.8m tok/s: 7767075
+1000/20000 train_loss: 3.1904 train_time: 1.7m tok/s: 7753753
+1500/20000 train_loss: 3.1009 train_time: 2.5m tok/s: 7758001
+2000/20000 train_loss: 3.0633 train_time: 3.4m tok/s: 7762917
+layer_loop:enabled step:2032 frac:0.350 encoder:[0, 1, 2, 3, 4, 5, 3, 4] decoder:[5, 3, 4, 5, 6, 7, 8, 9, 10]
+2500/20000 train_loss: 3.0652 train_time: 4.6m tok/s: 7136927
+3000/20000 train_loss: 2.9412 train_time: 5.8m tok/s: 6743041
+3500/20000 train_loss: 2.9616 train_time: 7.1m tok/s: 6471277
+4000/20000 train_loss: 2.9060 train_time: 8.3m tok/s: 6295682
+4000/20000 val_loss: 2.8659 val_bpb: 1.1095
+4500/20000 train_loss: 2.7602 train_time: 9.6m tok/s: 6166690
+4596/20000 val_loss: 2.8082 val_bpb: 1.0871
+stopping_early: wallclock_cap train_time: 588152ms step: 4596/20000
+peak memory allocated: 39046 MiB reserved: 39070 MiB
+ema:applying EMA weights
+pre-quantization post-ema val_loss:2.80550662 val_bpb:1.08609904 eval_time:6986ms
+Serialized model: 135431033 bytes
+Code size: 57003 bytes
+GPTQ:collecting Hessians from calibration data...
+GPTQ:collected 67 Hessians in 12.8s
+Quantized weights:
+  gptq (int6): blocks.attn.c_k.weight, blocks.attn.c_q.weight, blocks.attn.c_v.weight, blocks.attn.proj.weight, blocks.mlp.fc.weight, blocks.mlp.proj.weight
+  gptq (int8): tok_emb.weight
+  passthrough (float16): blocks.attn.q_gain, blocks.attn_scale, blocks.mlp_scale, blocks.resid_mix, skip_gates, skip_weights
+Serialized model quantized+brotli: 15972921 bytes
+Total submission size quantized+brotli: 16029924 bytes
+quantized:eval_num_loops=2 looping_active=True encoder:[0, 1, 2, 3, 4, 5, 3, 4] decoder:[5, 3, 4, 5, 6, 7, 8, 9, 10]
+quantized val_loss:2.83861053 val_bpb:1.09891460 eval_time:24849ms
+quantized_sliding_window val_loss:2.79516170 val_bpb:1.08209420 eval_time:121664ms
+ttt:eval_num_loops=2 looping_active=True encoder:[0, 1, 2, 3, 4, 5, 3, 4] decoder:[5, 3, 4, 5, 6, 7, 8, 9, 10]
+ttt:start chunks=1238 ttt_lr=0.007 ttt_epochs=5
+quantized_ttt val_loss:2.78909020 val_bpb:1.07974373 eval_time:514291ms