Upload folder using huggingface_hub

attention_kindself_n_heads2_seed1338/args.json

@@ -1 +1 @@
-{"hellaswag": true, "attention_kind": "selective", "log_dir": "wider_is_better_0/attention_kindself_n_heads2_seed1338", "resume_checkpoint": null, "resume_optimizer": false, "add_a_head": false, "add_head_to_start": true, "new_head_init": "normal", "n_heads": 2, "protect_bos_token": true, "prevent_from_masking_myself": true, "max_steps": 10, "warmup_steps": 200, "group": "wider_is_better_0", "use_wandb": true, "kill_self_after_run": false, "random_seed": 1338, "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": 40, "total_batch_size": 10240, "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": true, "mup_enable_coord_check_logging": false, "max_lr": 0.00015, "decay_lr": true, "readout_zero_init": false, "query_zero_init": false, "l1_loss": false, "debugpy": false, "key": "1.5e-4_10240_2_1338", "n_embd": 128}
+{"hellaswag": true, "attention_kind": "selective", "log_dir": "wider_is_better_2/attention_kindself_n_heads2_seed1338", "resume_checkpoint": null, "resume_optimizer": false, "add_a_head": false, "add_head_to_start": true, "new_head_init": "normal", "n_heads": 2, "protect_bos_token": true, "prevent_from_masking_myself": true, "max_steps": 100, "warmup_steps": 200, "group": "wider_is_better_2", "use_wandb": true, "kill_self_after_run": false, "random_seed": 1338, "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": 80, "total_batch_size": 20480, "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": true, "mup_enable_coord_check_logging": false, "max_lr": 0.00015, "decay_lr": true, "readout_zero_init": false, "query_zero_init": false, "l1_loss": false, "debugpy": false, "key": "1.5e-4_20480_2_1338", "n_embd": 128}

attention_kindself_n_heads2_seed1338/dataloader_00099.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:580e5e81f112aa98b56a5810a4962c6f47cc749a08325f59a6f3df76f5120431
+size 964

attention_kindself_n_heads2_seed1338/log2.txt

@@ -1,15 +1,8 @@
-max_steps: 10
-0 val loss 11.7414
-0 val perplexity 125673.7578
-0 train 11.761742 (lr=7.5000e-07) (hash(x)=50671684)
-1 train 11.767031 (lr=1.5000e-06) (hash(x)=48334907)
-2 train 11.716578 (lr=2.2500e-06) (hash(x)=51320861)
-3 train 11.743654 (lr=3.0000e-06) (hash(x)=46712336)
-4 train 11.714699 (lr=3.7500e-06) (hash(x)=51050552)
-5 train 11.727638 (lr=4.5000e-06) (hash(x)=53176793)
-6 train 11.720113 (lr=5.2500e-06) (hash(x)=56077270)
-7 train 11.699996 (lr=6.0000e-06) (hash(x)=50171817)
-8 train 11.686590 (lr=6.7500e-06) (hash(x)=54201336)
-9 val loss 11.6803
-9 val perplexity 118221.9219
-9 train 11.687712 (lr=7.5000e-06) (hash(x)=49973223)
+max_steps: 100
+0 val loss 11.7388
+0 val perplexity 125337.7812
+0 train 11.765342 (lr=7.5000e-07) (hash(x)=99006591)
+50 val loss 10.4093
+50 val perplexity 33165.7539
+99 val loss 9.7923
+99 val perplexity 17895.2012

attention_kindself_n_heads2_seed1338/model_00099.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5fc6b48e69464ad96d54e50cd980bd05bc5f59cdadbf03719724bb6edb7415ba
+size 38587970

attention_kindself_n_heads2_seed1338/optimizer_00099.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d54b66d578b607922a6cca38495e8484d4de90ad53dc9c56296314bc260bc711
+size 70895430