Upload folder using huggingface_hub

baseline_lr_30e-4_seed_1339/args.json

@@ -1 +1 @@
-{"hellaswag": true, "attention_kind": "selective", "log_dir": "test_logs/repro_selective_pattern_rankings/baseline_lr_30e-4_seed_1339", "resume_checkpoint": null, "resume_optimizer": false, "add_a_head": false, "add_head_to_start": true, "new_head_init": "normal", "n_heads": 12, "n_embd": 264, "head_dim": 22, "protect_bos_token": true, "prevent_from_masking_myself": true, "max_steps": 100, "warmup_steps": 250, "group": "repro_selective_pattern_rankings", "use_wandb": true, "kill_self_after_run": false, "random_seed": 1339, "memory_penalty_epsilon": 0.1, "selection_head_linear_combo": "none", "selection_head_linear_combo_scale": 1.0, "protection_kind": "none", "leaky_relu_alpha": null, "leaky_relu_bias": null, "use_compile": true, "use_mini_model": false, "upload_to_hf": true, "seq_len": 256, "batch_size": 64, "total_batch_size": 131072, "protection_head_scaling_factor": 1.0, "protection_head_bias": 0.0, "n_sliced_masks": null, "n_latent_masks": null, "mask_layernorm": false, "residual_attention_masks": false, "compute_base_shapes": false, "base_shapes_savefile": null, "mup": true, "disable_selection": false, "mup_enable_coord_check_logging": false, "max_lr": 0.003, "decay_lr": true, "readout_zero_init": false, "query_zero_init": false, "l1_loss": false, "debugpy": false, "key": "baseline_lr_30e-4"}
+{"hellaswag": true, "attention_kind": "selective", "log_dir": "test_logs/repro_selective_pattern_rankings/baseline_lr_30e-4_seed_1339", "resume_checkpoint": null, "resume_optimizer": false, "add_a_head": false, "add_head_to_start": true, "new_head_init": "normal", "n_heads": 12, "n_embd": 264, "head_dim": 22, "protect_bos_token": true, "prevent_from_masking_myself": true, "max_steps": 0, "warmup_steps": 250, "group": "repro_selective_pattern_rankings", "use_wandb": true, "kill_self_after_run": false, "random_seed": 1339, "memory_penalty_epsilon": 0.1, "selection_head_linear_combo": "none", "selection_head_linear_combo_scale": 1.0, "protection_kind": "none", "leaky_relu_alpha": null, "leaky_relu_bias": null, "use_compile": false, "use_mini_model": false, "upload_to_hf": true, "seq_len": 256, "batch_size": 64, "total_batch_size": 131072, "protection_head_scaling_factor": 1.0, "protection_head_bias": 0.0, "n_sliced_masks": null, "n_latent_masks": null, "mask_layernorm": false, "residual_attention_masks": false, "compute_base_shapes": false, "base_shapes_savefile": null, "mup": true, "disable_selection": false, "mup_enable_coord_check_logging": false, "max_lr": 0.003, "decay_lr": true, "readout_zero_init": false, "query_zero_init": false, "l1_loss": false, "debugpy": false, "key": "baseline_lr_30e-4"}

baseline_lr_30e-4_seed_1339/log2.txt

@@ -1,15 +1 @@
-max_steps: 100
-0 val loss 11.4043
-0 val perplexity 89706.9531
-0 train 11.388926 (lr=1.2000e-05) (hash(x)=93356070)
-10 train 9.731706 (lr=1.3200e-04) (hash(x)=91253010)
-20 train 8.405382 (lr=2.5200e-04) (hash(x)=74934453)
-30 train 7.772275 (lr=3.7200e-04) (hash(x)=79278034)
-40 train 7.685396 (lr=4.9200e-04) (hash(x)=80107892)
-50 train 7.626685 (lr=6.1200e-04) (hash(x)=70745428)
-60 train 7.584211 (lr=7.3200e-04) (hash(x)=80077589)
-70 train 7.168908 (lr=8.5200e-04) (hash(x)=76213766)
-80 train 7.155523 (lr=9.7200e-04) (hash(x)=83218328)
-90 train 6.998047 (lr=1.0920e-03) (hash(x)=74855845)
-99 val loss 6.9620
-99 val perplexity 1055.7167
+max_steps: 0