Mar 28 10:18 rls t5 large 08:55 rls t5 large 05:09 rls 02:42 rls Mar 27 11:54 rls 09:35 new v3 Mar 24 16:52 v10a 16:48 v10a 16:44 v10a 16:23 v10a 16:17 v10a 16:03 v10a 08:35 V10a 07:49 V10 04:18 v10 04:02 v10 03:45 v10 03:25 v10 Mar 23 19:27 v10 18:10 v10 04:55 v9-fix: oracle training routing, calibrated inference, update docs 04:28 v9 Mar 21 05:17 v7 04:40 v7: C5 Data-Informed Subspace Init + restructure contributions to 2 core claims Mar 18 09:35 fix: reduce CPU RAM to prevent OOM SIGKILL - gc.collect, del temps, eval_accumulation_steps, cache cleanup 02:13 fix: add trust_remote_code=True to second load_dataset call in run_llama.py 01:08 dataset: allow custom dataset code by setting trust_remote_code=True in run_llama.py Mar 17 17:15 C4: Spectrally-Conditioned LoRA Training — preconditioned gradient + spectral entropy regularization - cl_trainer_specroute.py: precompute_preconditioners(), _compute_spectral_entropy_loss(), _apply_preconditioning()
- run_t5.py: 4 new args (lambda_entropy, use_preconditioning, precond_eps, entropy_warmup_ratio)
- V4 shell script: gen_script_long_order3_t5_small_specroute_v4.sh
- SPECROUTE_IDEA.md: C4 section, Claim 4, updated tables
- experiment_versions.md: V4.0 section + ablation plan
- contribution2_implementation_analysis.md: detailed C4 analysis (math, pipeline, synergy, risks) 16:43 SpecRoute V3: adaptive bias, symmetric inference, threshold 0.995, batch size optimization V3 fixes (3 targeted):
- Adaptive training bias: beta=T*ln(alpha*n/(1-alpha)), ensures ~80% routing weight
- Symmetric inference routing: prepare_inference_routing() computes SVD for current task
- Threshold 0.995: matches ROOT GPM capacity
Batch size optimization:
- T4_1gpu: BSZ 4->8, GA 8->4 (effective=32, ~2x fewer micro-batches)
- T4_2gpu: BSZ 2->4, GA 8->4 (effective=32)
- A100: BSZ 8->32, GA 4->1 (effective=32)
Docs: V2 diagnosis report, V3 methodology in SPECROUTE_IDEA.md, experiment_versions.md Mar 16 10:53 fixbug 10:42 fixbug 08:48 fix llama 04:46 improve v2 Mar 15 12:12 fix bug 05:46 fix bug Mar 12 15:49 fix: pass attention_mask directly to model.generate(), not via GenerationConfig 14:27 fix: override _save to disable safetensors for T5 shared embedding weights 14:01 fix: denumpify_detensorize moved to trainer_utils in transformers 4.40+ 13:48 fix: add explicit trainer_pt_utils imports (nested_truncate etc.) removed from trainer.* in 4.40+ 11:08 fix: comprehensive transformers 4.40 compat across all trainer files 1. Import shims for removed APIs (ShardedDDPOption, is_torch_tpu_available, IterableDatasetShard)
2. Attribute getattr guards: do_grad_scaling, use_apex, sharded_ddp, fsdp
3. Covers both root_gainlora and improve_gainlora, all cl_trainer_*.py files 10:59 fix: replace removed _pad_across_processes with accelerator API transformers 4.40 removed Trainer._pad_across_processes().
Replace all 32 call sites (2 per trainer × 16 trainers across both repos)
with self.accelerator.pad_across_processes() which is the underlying
implementation. Behavior is identical: single-GPU returns tensor unchanged,
multi-GPU pads to max length across processes.
Affected files in root_gainlora/src/ and improve_gainlora/src/:
cl_trainer_gainlora_inflora.py, cl_trainer_inflora.py,
cl_trainer_inflora_llama.py, cl_trainer_olora.py, cl_trainer_olora_llama.py,
cl_trainer_gainlora_olora.py, cl_trainer_gainlora_olora_llama.py,
cl_trainer_gainlora_inflora_llama.py 
