diff --git a/.gitattributes b/.gitattributes
index fd857501c7ca4b874b61dd810867d733bc2ffeaf..d1c7997026325cad4baa5f0b0a95f2ae81223f92 100644
--- a/.gitattributes
+++ b/.gitattributes
@@ -34,3 +34,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
 wandb/run-20260226_135602-696nxyfr/run-696nxyfr.wandb filter=lfs diff=lfs merge=lfs -text
+wandb/run-20260226_153026-trcpjlfd/run-trcpjlfd.wandb filter=lfs diff=lfs merge=lfs -text
diff --git a/checkpoints/metadata_000000032768.json b/checkpoints/metadata_000000032768.json
index b8a7704ce050180271ca25f2f255fcf595a47b6c..ff44a084aafa754e5474df41ea15c374a23fa268 100644
--- a/checkpoints/metadata_000000032768.json
+++ b/checkpoints/metadata_000000032768.json
@@ -1 +1 @@
-{"step": 1, "tokens_seen": 32768, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.47806167602539}
\ No newline at end of file
+{"step": 1, "tokens_seen": 32768, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.47472858428955}
\ No newline at end of file
diff --git a/checkpoints/metadata_000000327680.json b/checkpoints/metadata_000000327680.json
index d42d566ea67f268fa6f57813ee4d541346608b09..8ce0422abf1ed1343400bcc795303c1f9c616194 100644
--- a/checkpoints/metadata_000000327680.json
+++ b/checkpoints/metadata_000000327680.json
@@ -1 +1 @@
-{"step": 10, "tokens_seen": 327680, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.47627660688796}
\ No newline at end of file
+{"step": 10, "tokens_seen": 327680, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.472703523224029}
\ No newline at end of file
diff --git a/checkpoints/metadata_000000360448.json b/checkpoints/metadata_000000360448.json
index 3d7caabc185ac1f7af4a964848e301009e0bee87..5f86bf14630ffed3dfd8509fb27841b4158561a2 100644
--- a/checkpoints/metadata_000000360448.json
+++ b/checkpoints/metadata_000000360448.json
@@ -1 +1 @@
-{"step": 11, "tokens_seen": 360448, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.475470339918813}
\ No newline at end of file
+{"step": 11, "tokens_seen": 360448, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.471990140793302}
\ No newline at end of file
diff --git a/checkpoints/metadata_000000425984.json b/checkpoints/metadata_000000425984.json
index 8fe398142a24e02b10c494cb6ff60557b8f9f5f4..f890a96d7e29d632ed10db4aff1abeeb4da6adaf 100644
--- a/checkpoints/metadata_000000425984.json
+++ b/checkpoints/metadata_000000425984.json
@@ -1 +1 @@
-{"step": 13, "tokens_seen": 425984, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.473083299528513}
\ No newline at end of file
+{"step": 13, "tokens_seen": 425984, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.469721053741852}
\ No newline at end of file
diff --git a/checkpoints/metadata_000000458752.json b/checkpoints/metadata_000000458752.json
index f162026671441be3ff60f893882f1a74bc817645..dcf880858f9f072a94a3ae49e79d7227ce2d792e 100644
--- a/checkpoints/metadata_000000458752.json
+++ b/checkpoints/metadata_000000458752.json
@@ -1 +1 @@
-{"step": 14, "tokens_seen": 458752, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.471743989894433}
\ No newline at end of file
+{"step": 14, "tokens_seen": 458752, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.468519898807461}
\ No newline at end of file
diff --git a/checkpoints/metadata_000000491520.json b/checkpoints/metadata_000000491520.json
index 4c79139f1f06a20bc53600249f641d0ccb6435b6..f71dda81d7464c25452230a045723a1c40ad90b8 100644
--- a/checkpoints/metadata_000000491520.json
+++ b/checkpoints/metadata_000000491520.json
@@ -1 +1 @@
-{"step": 15, "tokens_seen": 491520, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.470174342811651}
\ No newline at end of file
+{"step": 15, "tokens_seen": 491520, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.467125200460256}
\ No newline at end of file
diff --git a/checkpoints/metadata_000000557056.json b/checkpoints/metadata_000000557056.json
index aaeb3f54a99814ce2d5cd77e2a76e06123246521..4469dc93cea90e4e284ab879ad82de065e161d52 100644
--- a/checkpoints/metadata_000000557056.json
+++ b/checkpoints/metadata_000000557056.json
@@ -1 +1 @@
-{"step": 17, "tokens_seen": 557056, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.466137204272727}
\ No newline at end of file
+{"step": 17, "tokens_seen": 557056, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.463455310860898}
\ No newline at end of file
diff --git a/checkpoints/metadata_000000622592.json b/checkpoints/metadata_000000622592.json
index dc64e5070c0c8acda29fae1e838e840583d89fda..169f67b8b435103d417869bd206168a0e27f9d85 100644
--- a/checkpoints/metadata_000000622592.json
+++ b/checkpoints/metadata_000000622592.json
@@ -1 +1 @@
-{"step": 19, "tokens_seen": 622592, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.461292862352218}
\ No newline at end of file
+{"step": 19, "tokens_seen": 622592, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.459099738819177}
\ No newline at end of file
diff --git a/checkpoints/metadata_000000688128.json b/checkpoints/metadata_000000688128.json
index ee621bde74f1b70e7745f6d6982ee0ce657f9468..760d87d6aadad649e919701a79502eb04732aa4e 100644
--- a/checkpoints/metadata_000000688128.json
+++ b/checkpoints/metadata_000000688128.json
@@ -1 +1 @@
-{"step": 21, "tokens_seen": 688128, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.455025942618947}
\ No newline at end of file
+{"step": 21, "tokens_seen": 688128, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.453407187818708}
\ No newline at end of file
diff --git a/checkpoints/metadata_000000753664.json b/checkpoints/metadata_000000753664.json
index 1baef2ebe8e428cd1b1f752d67724878e2d5fd9f..ede06b34062135e04a840100102f55a312046402 100644
--- a/checkpoints/metadata_000000753664.json
+++ b/checkpoints/metadata_000000753664.json
@@ -1 +1 @@
-{"step": 23, "tokens_seen": 753664, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.446919057636006}
\ No newline at end of file
+{"step": 23, "tokens_seen": 753664, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.446099322563402}
\ No newline at end of file
diff --git a/checkpoints/metadata_000000819200.json b/checkpoints/metadata_000000819200.json
index 62bf26e4dcb7882b53b7443ab588061ba613e705..57f305d5c02a894fd0fb94b03bb1d3bf5fbecd06 100644
--- a/checkpoints/metadata_000000819200.json
+++ b/checkpoints/metadata_000000819200.json
@@ -1 +1 @@
-{"step": 25, "tokens_seen": 819200, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.438280975135724}
\ No newline at end of file
+{"step": 25, "tokens_seen": 819200, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.438267091556018}
\ No newline at end of file
diff --git a/checkpoints/metadata_000000917504.json b/checkpoints/metadata_000000917504.json
index 2f0e48b7dcf9d3e421a608b288bc8edd356acd51..fa6312c3000fdd374487baa1d1c54c8ed70d0730 100644
--- a/checkpoints/metadata_000000917504.json
+++ b/checkpoints/metadata_000000917504.json
@@ -1 +1 @@
-{"step": 28, "tokens_seen": 917504, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.422044364253066}
\ No newline at end of file
+{"step": 28, "tokens_seen": 917504, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.423273612378466}
\ No newline at end of file
diff --git a/checkpoints/metadata_000000983040.json b/checkpoints/metadata_000000983040.json
index 699cee11d0214ac7e5944ecdd04e8cb13a300994..7df98d9217407dbaf03e1224946390148fe1a478 100644
--- a/checkpoints/metadata_000000983040.json
+++ b/checkpoints/metadata_000000983040.json
@@ -1 +1 @@
-{"step": 30, "tokens_seen": 983040, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.409236852859813}
\ No newline at end of file
+{"step": 30, "tokens_seen": 983040, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.411245689106972}
\ No newline at end of file
diff --git a/checkpoints/metadata_000001114112.json b/checkpoints/metadata_000001114112.json
index 141574201d643111e2cf18c3c5d36f20e45211a4..61cc72f65bfff95fa8815f469a1e2b6e9859595d 100644
--- a/checkpoints/metadata_000001114112.json
+++ b/checkpoints/metadata_000001114112.json
@@ -1 +1 @@
-{"step": 34, "tokens_seen": 1114112, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.380797570445656}
\ No newline at end of file
+{"step": 34, "tokens_seen": 1114112, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.383828797056317}
\ No newline at end of file
diff --git a/checkpoints/metadata_000001212416.json b/checkpoints/metadata_000001212416.json
index e03c279a8bed8806a93a4b404c61377ac32160c4..5cda40b1bb7426ee1f31d21d62c169b32e645a2a 100644
--- a/checkpoints/metadata_000001212416.json
+++ b/checkpoints/metadata_000001212416.json
@@ -1 +1 @@
-{"step": 37, "tokens_seen": 1212416, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.3582079333492}
\ No newline at end of file
+{"step": 37, "tokens_seen": 1212416, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.361741983796891}
\ No newline at end of file
diff --git a/checkpoints/metadata_000001343488.json b/checkpoints/metadata_000001343488.json
index fcdd216343e123858497be637c6bb7b0e680f4c6..17654f11fb2396b31d4a563f94dfca2094779a92 100644
--- a/checkpoints/metadata_000001343488.json
+++ b/checkpoints/metadata_000001343488.json
@@ -1 +1 @@
-{"step": 41, "tokens_seen": 1343488, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.325794613706586}
\ No newline at end of file
+{"step": 41, "tokens_seen": 1343488, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.329621238518811}
\ No newline at end of file
diff --git a/checkpoints/metadata_000001474560.json b/checkpoints/metadata_000001474560.json
index 0d18f8776a72ed8ccd933da3d907f83603159d46..57747bb43b744b5cfaddeaa93681e48b47393f9d 100644
--- a/checkpoints/metadata_000001474560.json
+++ b/checkpoints/metadata_000001474560.json
@@ -1 +1 @@
-{"step": 45, "tokens_seen": 1474560, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.289813354053809}
\ No newline at end of file
+{"step": 45, "tokens_seen": 1474560, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.29327811546785}
\ No newline at end of file
diff --git a/checkpoints/metadata_000001605632.json b/checkpoints/metadata_000001605632.json
index b36f5e561d2af918ee2cba8a2ea944fffa623b7e..f58443b698730310105307f06dc671c835d7c45e 100644
--- a/checkpoints/metadata_000001605632.json
+++ b/checkpoints/metadata_000001605632.json
@@ -1 +1 @@
-{"step": 49, "tokens_seen": 1605632, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.250201881192115}
\ No newline at end of file
+{"step": 49, "tokens_seen": 1605632, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.253395153119124}
\ No newline at end of file
diff --git a/checkpoints/metadata_000001769472.json b/checkpoints/metadata_000001769472.json
index b3ef822016a8a8f7beb31e27a3229e9be5246628..fd5b42aa72cb3fbfa104058f7fe0976e0763565e 100644
--- a/checkpoints/metadata_000001769472.json
+++ b/checkpoints/metadata_000001769472.json
@@ -1 +1 @@
-{"step": 54, "tokens_seen": 1769472, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.19959723315217}
\ No newline at end of file
+{"step": 54, "tokens_seen": 1769472, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.20219309142512}
\ No newline at end of file
diff --git a/checkpoints/metadata_000001966080.json b/checkpoints/metadata_000001966080.json
index c2a1487afc63bae04e991f52ae073bf2a1a19546..833734f418268eb90e2f46c9025c5872770665f3 100644
--- a/checkpoints/metadata_000001966080.json
+++ b/checkpoints/metadata_000001966080.json
@@ -1 +1 @@
-{"step": 60, "tokens_seen": 1966080, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.13133821097923}
\ No newline at end of file
+{"step": 60, "tokens_seen": 1966080, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.133052898451657}
\ No newline at end of file
diff --git a/checkpoints/metadata_000002162688.json b/checkpoints/metadata_000002162688.json
index 0987d20d4d9e2bed5b80ffa13991890f19111b52..fc305f8231544b3175065c791cdf3f3dbf671f23 100644
--- a/checkpoints/metadata_000002162688.json
+++ b/checkpoints/metadata_000002162688.json
@@ -1 +1 @@
-{"step": 66, "tokens_seen": 2162688, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.0570300679636}
\ No newline at end of file
+{"step": 66, "tokens_seen": 2162688, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 10.057631285417521}
\ No newline at end of file
diff --git a/checkpoints/metadata_000002359296.json b/checkpoints/metadata_000002359296.json
index e4cc5f37f9cc6cf896d9020ba39870d9c5207eaf..ec9114785ac35b157295da59260e093b87222ed4 100644
--- a/checkpoints/metadata_000002359296.json
+++ b/checkpoints/metadata_000002359296.json
@@ -1 +1 @@
-{"step": 72, "tokens_seen": 2359296, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 9.974723497158761}
\ No newline at end of file
+{"step": 72, "tokens_seen": 2359296, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 9.97402740012332}
\ No newline at end of file
diff --git a/checkpoints/metadata_000002621440.json b/checkpoints/metadata_000002621440.json
index 229c04826656632bc75f6db36a8162b0844225db..9b5f8392cd19f09aa15165606806faf612079ea2 100644
--- a/checkpoints/metadata_000002621440.json
+++ b/checkpoints/metadata_000002621440.json
@@ -1 +1 @@
-{"step": 80, "tokens_seen": 2621440, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 9.858762341449337}
\ No newline at end of file
+{"step": 80, "tokens_seen": 2621440, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 9.856754308431446}
\ No newline at end of file
diff --git a/checkpoints/metadata_000002883584.json b/checkpoints/metadata_000002883584.json
index 9c16151176a43d6ba48c2cf9d77e7d0842e1cd5b..f3d388f25073232f44e21a2d44bd8672f3342f43 100644
--- a/checkpoints/metadata_000002883584.json
+++ b/checkpoints/metadata_000002883584.json
@@ -1 +1 @@
-{"step": 88, "tokens_seen": 2883584, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 9.730080927623849}
\ No newline at end of file
+{"step": 88, "tokens_seen": 2883584, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 9.726206998293742}
\ No newline at end of file
diff --git a/checkpoints/metadata_000003178496.json b/checkpoints/metadata_000003178496.json
index 022fba1fc00a947fd087ddbc78cd0d5f32e0efb1..3088c0f9c63292080302b9f7a90b4b6705945a43 100644
--- a/checkpoints/metadata_000003178496.json
+++ b/checkpoints/metadata_000003178496.json
@@ -1 +1 @@
-{"step": 97, "tokens_seen": 3178496, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 9.582970532481923}
\ No newline at end of file
+{"step": 97, "tokens_seen": 3178496, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 9.577359730084346}
\ No newline at end of file
diff --git a/checkpoints/metadata_000003473408.json b/checkpoints/metadata_000003473408.json
index 4c4a6bda07a832da46b9026337d8e10541540c70..3bdf20a37639fae37b731f501c37d02494aeaa6d 100644
--- a/checkpoints/metadata_000003473408.json
+++ b/checkpoints/metadata_000003473408.json
@@ -1 +1 @@
-{"step": 106, "tokens_seen": 3473408, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 9.424478446215208}
\ No newline at end of file
+{"step": 106, "tokens_seen": 3473408, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 9.417083930648644}
\ No newline at end of file
diff --git a/checkpoints/metadata_000003833856.json b/checkpoints/metadata_000003833856.json
index 725ce05c5d991e1f3dd8cdb521eaa8f1ff8e82cd..9eaf8a8e7a7d69389e7d5d824fb2e102d45413c2 100644
--- a/checkpoints/metadata_000003833856.json
+++ b/checkpoints/metadata_000003833856.json
@@ -1 +1 @@
-{"step": 117, "tokens_seen": 3833856, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 9.225284321418476}
\ No newline at end of file
+{"step": 117, "tokens_seen": 3833856, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 9.216006205212386}
\ No newline at end of file
diff --git a/checkpoints/metadata_000004227072.json b/checkpoints/metadata_000004227072.json
index e281a19b012143bbf7a5e49a68ca79fcd9013594..9ef9c58749ecef4f8489abd78500ce31e9de9a9f 100644
--- a/checkpoints/metadata_000004227072.json
+++ b/checkpoints/metadata_000004227072.json
@@ -1 +1 @@
-{"step": 129, "tokens_seen": 4227072, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 8.99394007776402}
\ No newline at end of file
+{"step": 129, "tokens_seen": 4227072, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 8.982881949139617}
\ No newline at end of file
diff --git a/checkpoints/metadata_000004653056.json b/checkpoints/metadata_000004653056.json
index b705fd1804265a2b8c8b9272d36c894ceb86cbf2..f90891326e5bc88212c8e48a74d5753f34d840df 100644
--- a/checkpoints/metadata_000004653056.json
+++ b/checkpoints/metadata_000004653056.json
@@ -1 +1 @@
-{"step": 142, "tokens_seen": 4653056, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 8.744109641520092}
\ No newline at end of file
+{"step": 142, "tokens_seen": 4653056, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 8.732197937516098}
\ No newline at end of file
diff --git a/checkpoints/metadata_000005111808.json b/checkpoints/metadata_000005111808.json
index 674db93cfc7bfa57cfbb6bb6a7ea503278c9964f..b30fe773f60284a1ff2d4b1493bb3209f674e24a 100644
--- a/checkpoints/metadata_000005111808.json
+++ b/checkpoints/metadata_000005111808.json
@@ -1 +1 @@
-{"step": 156, "tokens_seen": 5111808, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 8.493460174429384}
\ No newline at end of file
+{"step": 156, "tokens_seen": 5111808, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 8.482621967493646}
\ No newline at end of file
diff --git a/checkpoints/metadata_000005603328.json b/checkpoints/metadata_000005603328.json
index 5ec22e9a8d4a96a9fc014de4122cb5fea5df3543..1f631700de350606383e6a10ad869dfc1cf83df7 100644
--- a/checkpoints/metadata_000005603328.json
+++ b/checkpoints/metadata_000005603328.json
@@ -1 +1 @@
-{"step": 171, "tokens_seen": 5603328, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 8.236969204325344}
\ No newline at end of file
+{"step": 171, "tokens_seen": 5603328, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 8.22768931743315}
\ No newline at end of file
diff --git a/checkpoints/metadata_000006193152.json b/checkpoints/metadata_000006193152.json
index 1b7d974dcfa9137c070ccd4821c0dd18cf205bbd..1115309a4798c9088e10bb0870da72a1bacb08e5 100644
--- a/checkpoints/metadata_000006193152.json
+++ b/checkpoints/metadata_000006193152.json
@@ -1 +1 @@
-{"step": 189, "tokens_seen": 6193152, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 7.953434445580491}
\ No newline at end of file
+{"step": 189, "tokens_seen": 6193152, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 7.945035313729822}
\ No newline at end of file
diff --git a/checkpoints/metadata_000006782976.json b/checkpoints/metadata_000006782976.json
index 522eaa9db8475977b9e6bdc1dd093a2c00068144..b9aa0541e818adabdeb9b428504248b376a9a34f 100644
--- a/checkpoints/metadata_000006782976.json
+++ b/checkpoints/metadata_000006782976.json
@@ -1 +1 @@
-{"step": 207, "tokens_seen": 6782976, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 7.701950555362339}
\ No newline at end of file
+{"step": 207, "tokens_seen": 6782976, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 7.694594853436912}
\ No newline at end of file
diff --git a/checkpoints/metadata_000007471104.json b/checkpoints/metadata_000007471104.json
index 99d6b4f594adf36b9c235b7e80a492b6b8559fa2..6a69b982ffb9ae6bf611ca89c34439195f84f0de 100644
--- a/checkpoints/metadata_000007471104.json
+++ b/checkpoints/metadata_000007471104.json
@@ -1 +1 @@
-{"step": 228, "tokens_seen": 7471104, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 7.42616407882357}
\ No newline at end of file
+{"step": 228, "tokens_seen": 7471104, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 7.419854967755447}
\ No newline at end of file
diff --git a/checkpoints/metadata_000008224768.json b/checkpoints/metadata_000008224768.json
index 59813b8f5d1563a930c257626e2f884613387b9e..25de7fad0c012c8e854f51f055664a154b89b218 100644
--- a/checkpoints/metadata_000008224768.json
+++ b/checkpoints/metadata_000008224768.json
@@ -1 +1 @@
-{"step": 251, "tokens_seen": 8224768, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 7.164880585767827}
\ No newline at end of file
+{"step": 251, "tokens_seen": 8224768, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 7.159385434456035}
\ No newline at end of file
diff --git a/checkpoints/metadata_000009043968.json b/checkpoints/metadata_000009043968.json
index 8a63166dc6e3c2a6edc73070bf7be87c14a84a77..6ca5b76834dfc2ba5254b212d17b641471af9d28 100644
--- a/checkpoints/metadata_000009043968.json
+++ b/checkpoints/metadata_000009043968.json
@@ -1 +1 @@
-{"step": 276, "tokens_seen": 9043968, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 6.909649069151253}
\ No newline at end of file
+{"step": 276, "tokens_seen": 9043968, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 6.904178105557777}
\ No newline at end of file
diff --git a/checkpoints/metadata_000009961472.json b/checkpoints/metadata_000009961472.json
index c3517fd833e0f7cadb502a735212b52d2f4107c4..8cd6a284051989e63d303001e841081e0ad82cc7 100644
--- a/checkpoints/metadata_000009961472.json
+++ b/checkpoints/metadata_000009961472.json
@@ -1 +1 @@
-{"step": 304, "tokens_seen": 9961472, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 6.658180864800952}
\ No newline at end of file
+{"step": 304, "tokens_seen": 9961472, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 6.653267985439542}
\ No newline at end of file
diff --git a/checkpoints/metadata_000010944512.json b/checkpoints/metadata_000010944512.json
index d4bd723383d6fc38c44747393bddd3cd111d57cd..5cb321ed1a85cbc50fdd3d0e5d22ef606c076820 100644
--- a/checkpoints/metadata_000010944512.json
+++ b/checkpoints/metadata_000010944512.json
@@ -1 +1 @@
-{"step": 334, "tokens_seen": 10944512, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 6.4353112753232855}
\ No newline at end of file
+{"step": 334, "tokens_seen": 10944512, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 6.430630889399794}
\ No newline at end of file
diff --git a/checkpoints/metadata_000012058624.json b/checkpoints/metadata_000012058624.json
index 92877621131e2e706f1f6f358c8032d01e012524..2d9db15c0aacbfb2796c469dc615bd71dd158d8e 100644
--- a/checkpoints/metadata_000012058624.json
+++ b/checkpoints/metadata_000012058624.json
@@ -1 +1 @@
-{"step": 368, "tokens_seen": 12058624, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 6.221434970794934}
\ No newline at end of file
+{"step": 368, "tokens_seen": 12058624, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 6.216337437233496}
\ No newline at end of file
diff --git a/checkpoints/metadata_000013271040.json b/checkpoints/metadata_000013271040.json
index eae251b83789c3befe5d920139e3123303fd4b10..2a294d851e292b4a5633f1eacd4ae79d6961b240 100644
--- a/checkpoints/metadata_000013271040.json
+++ b/checkpoints/metadata_000013271040.json
@@ -1 +1 @@
-{"step": 405, "tokens_seen": 13271040, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 6.030378599797957}
\ No newline at end of file
+{"step": 405, "tokens_seen": 13271040, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 6.02448773214943}
\ No newline at end of file
diff --git a/checkpoints/metadata_000014581760.json b/checkpoints/metadata_000014581760.json
index 16225b102b9c4ba26bfbb61a1e2354b122728295..26ac4ebf8b8a5e7faf04cfa24c95588e902d0349 100644
--- a/checkpoints/metadata_000014581760.json
+++ b/checkpoints/metadata_000014581760.json
@@ -1 +1 @@
-{"step": 445, "tokens_seen": 14581760, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 5.85191591132156}
\ No newline at end of file
+{"step": 445, "tokens_seen": 14581760, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 5.847221819414238}
\ No newline at end of file
diff --git a/checkpoints/metadata_000016056320.json b/checkpoints/metadata_000016056320.json
index 7871463ae5e10f6b474f59ed222854d3b52bb8e4..590173c59916f4cd7b11e47b8cd24bc88dbfc63a 100644
--- a/checkpoints/metadata_000016056320.json
+++ b/checkpoints/metadata_000016056320.json
@@ -1 +1 @@
-{"step": 490, "tokens_seen": 16056320, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 5.68748776856525}
\ No newline at end of file
+{"step": 490, "tokens_seen": 16056320, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 5.685284731362977}
\ No newline at end of file
diff --git a/checkpoints/metadata_000016384000.json b/checkpoints/metadata_000016384000.json
index 0791b7ee0fc6cd8a0d751ae8eb9047aec7eb5111..ffe89385a5087cbf9731d347ded5b32427e134d8 100644
--- a/checkpoints/metadata_000016384000.json
+++ b/checkpoints/metadata_000016384000.json
@@ -1 +1 @@
-{"step": 500, "tokens_seen": 16384000, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 5.661260096068413}
\ No newline at end of file
+{"step": 500, "tokens_seen": 16384000, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 5.659568437263502}
\ No newline at end of file
diff --git a/checkpoints/metadata_000017661952.json b/checkpoints/metadata_000017661952.json
index e4d8a81e6e63ab148e54ccd4a0c64ced8f1c54cb..e837633165c3119c6c67db088cf8ba2570624478 100644
--- a/checkpoints/metadata_000017661952.json
+++ b/checkpoints/metadata_000017661952.json
@@ -1 +1 @@
-{"step": 539, "tokens_seen": 17661952, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 5.539440078913039}
\ No newline at end of file
+{"step": 539, "tokens_seen": 17661952, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 5.53900757717826}
\ No newline at end of file
diff --git a/checkpoints/metadata_000019431424.json b/checkpoints/metadata_000019431424.json
index 795a61c7538edf75fadbc573f4c1a6a3efd34000..02977bc9c72b837edcb76061d0dffca0d3485675 100644
--- a/checkpoints/metadata_000019431424.json
+++ b/checkpoints/metadata_000019431424.json
@@ -1 +1 @@
-{"step": 593, "tokens_seen": 19431424, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 5.392000028286869}
\ No newline at end of file
+{"step": 593, "tokens_seen": 19431424, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 5.394161902688204}
\ No newline at end of file
diff --git a/checkpoints/metadata_000021364736.json b/checkpoints/metadata_000021364736.json
index c61b790b1e58816b8c6ad37b2c9dd1a7570b25d4..0c6c1fd1641f10c33c9434363114be5f46d675c1 100644
--- a/checkpoints/metadata_000021364736.json
+++ b/checkpoints/metadata_000021364736.json
@@ -1 +1 @@
-{"step": 652, "tokens_seen": 21364736, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 5.281319973885694}
\ No newline at end of file
+{"step": 652, "tokens_seen": 21364736, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 5.283358080025116}
\ No newline at end of file
diff --git a/checkpoints/metadata_000023494656.json b/checkpoints/metadata_000023494656.json
index 8b6585cbaf873238467d8c3501ab32ffa66b637d..5981d85d8e173e9686e2a59840941621345cefd2 100644
--- a/checkpoints/metadata_000023494656.json
+++ b/checkpoints/metadata_000023494656.json
@@ -1 +1 @@
-{"step": 717, "tokens_seen": 23494656, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 5.169868723380111}
\ No newline at end of file
+{"step": 717, "tokens_seen": 23494656, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 5.175760950696042}
\ No newline at end of file
diff --git a/checkpoints/metadata_000025853952.json b/checkpoints/metadata_000025853952.json
index f33b26f878db5f9b249b6d9af9415a415a9044f4..c0498817e058407d93aa996aa938260f173b6107 100644
--- a/checkpoints/metadata_000025853952.json
+++ b/checkpoints/metadata_000025853952.json
@@ -1 +1 @@
-{"step": 789, "tokens_seen": 25853952, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 5.055504738591524}
\ No newline at end of file
+{"step": 789, "tokens_seen": 25853952, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 5.0629671796907045}
\ No newline at end of file
diff --git a/checkpoints/metadata_000028442624.json b/checkpoints/metadata_000028442624.json
index 66b591df2df3b2b40f9b89e19e29f7d9b0bb30f3..eaea19c57783d15384c951ebdf687d41fd92a60e 100644
--- a/checkpoints/metadata_000028442624.json
+++ b/checkpoints/metadata_000028442624.json
@@ -1 +1 @@
-{"step": 868, "tokens_seen": 28442624, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 4.954886293959172}
\ No newline at end of file
+{"step": 868, "tokens_seen": 28442624, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 4.963105635300831}
\ No newline at end of file
diff --git a/checkpoints/metadata_000031293440.json b/checkpoints/metadata_000031293440.json
index ac6d4f57c0dcddf36e0e4bee3210521fe8c37c8f..a18268521149516c134baa8844e4e75bfd6db3b3 100644
--- a/checkpoints/metadata_000031293440.json
+++ b/checkpoints/metadata_000031293440.json
@@ -1 +1 @@
-{"step": 955, "tokens_seen": 31293440, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 4.851076791499597}
\ No newline at end of file
+{"step": 955, "tokens_seen": 31293440, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 4.865391816028884}
\ No newline at end of file
diff --git a/checkpoints/metadata_000032768000.json b/checkpoints/metadata_000032768000.json
index c41eb409957a149e6a51386411cf35f708a14e5b..a4a31d4225bbea49c2eacd5409dbd9c51168d59a 100644
--- a/checkpoints/metadata_000032768000.json
+++ b/checkpoints/metadata_000032768000.json
@@ -1 +1 @@
-{"step": 1000, "tokens_seen": 32768000, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 4.7911227754693355}
\ No newline at end of file
+{"step": 1000, "tokens_seen": 32768000, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 4.812689049158225}
\ No newline at end of file
diff --git a/checkpoints/metadata_000034439168.json b/checkpoints/metadata_000034439168.json
index 4492c2ab7920610c1664e2bfed985120d5a0a2dd..a68b397fdef42ad1922edff35ba07d71336569f7 100644
--- a/checkpoints/metadata_000034439168.json
+++ b/checkpoints/metadata_000034439168.json
@@ -1 +1 @@
-{"step": 1051, "tokens_seen": 34439168, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 4.724863075960573}
\ No newline at end of file
+{"step": 1051, "tokens_seen": 34439168, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 4.753625057604216}
\ No newline at end of file
diff --git a/checkpoints/metadata_000037879808.json b/checkpoints/metadata_000037879808.json
index 400734ca04f2abd7b190cd3da788988a890f6216..44634a3befed1723b762b6b24cc20b50185aadba 100644
--- a/checkpoints/metadata_000037879808.json
+++ b/checkpoints/metadata_000037879808.json
@@ -1 +1 @@
-{"step": 1156, "tokens_seen": 37879808, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 4.601169516817602}
\ No newline at end of file
+{"step": 1156, "tokens_seen": 37879808, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 4.641372455953141}
\ No newline at end of file
diff --git a/checkpoints/metadata_000041648128.json b/checkpoints/metadata_000041648128.json
index c34c2d7368aa04268987cd74fd7e077b0c7c0897..6ebf5d90537b189dbddec1ae5aff4c110c49c2ce 100644
--- a/checkpoints/metadata_000041648128.json
+++ b/checkpoints/metadata_000041648128.json
@@ -1 +1 @@
-{"step": 1271, "tokens_seen": 41648128, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 4.486551862501883}
\ No newline at end of file
+{"step": 1271, "tokens_seen": 41648128, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 4.529437655293409}
\ No newline at end of file
diff --git a/checkpoints/metadata_000045842432.json b/checkpoints/metadata_000045842432.json
index b61af622c244d21522e59812abacdd7397c77481..561d0ed658b95f638c9dfbefecba10260848f7c4 100644
--- a/checkpoints/metadata_000045842432.json
+++ b/checkpoints/metadata_000045842432.json
@@ -1 +1 @@
-{"step": 1399, "tokens_seen": 45842432, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 4.381150849922109}
\ No newline at end of file
+{"step": 1399, "tokens_seen": 45842432, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 4.429992510229746}
\ No newline at end of file
diff --git a/checkpoints/metadata_000049152000.json b/checkpoints/metadata_000049152000.json
index a245ac5d882cff4d36b700408e920214664a80d8..f58e0f35ce0d556166c16df04de837d0dc59b62e 100644
--- a/checkpoints/metadata_000049152000.json
+++ b/checkpoints/metadata_000049152000.json
@@ -1 +1 @@
-{"step": 1500, "tokens_seen": 49152000, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 4.324639904720941}
\ No newline at end of file
+{"step": 1500, "tokens_seen": 49152000, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 4.3700635583946745}
\ No newline at end of file
diff --git a/checkpoints/metadata_000050397184.json b/checkpoints/metadata_000050397184.json
index badddafe6eb760acafe95a7b859e7c742b39c8ea..915e41755cb397f5b943211c4314449951073a04 100644
--- a/checkpoints/metadata_000050397184.json
+++ b/checkpoints/metadata_000050397184.json
@@ -1 +1 @@
-{"step": 1538, "tokens_seen": 50397184, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 4.3145797956722785}
\ No newline at end of file
+{"step": 1538, "tokens_seen": 50397184, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 4.358773860729051}
\ No newline at end of file
diff --git a/checkpoints/metadata_000055443456.json b/checkpoints/metadata_000055443456.json
index 8ec28fbc614e9c1aaa1e91ca920322588046cb94..d0b29dae8e6bf176faf72fc0ddef8e07c4a12e54 100644
--- a/checkpoints/metadata_000055443456.json
+++ b/checkpoints/metadata_000055443456.json
@@ -1 +1 @@
-{"step": 1692, "tokens_seen": 55443456, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 4.2051704572488475}
\ No newline at end of file
+{"step": 1692, "tokens_seen": 55443456, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 4.25258944505606}
\ No newline at end of file
diff --git a/checkpoints/metadata_000061014016.json b/checkpoints/metadata_000061014016.json
index 556fdebeccab6bdfe632de548af5ada71be43d8d..cf943a25c069d7264540ce4f729d780df1725bc5 100644
--- a/checkpoints/metadata_000061014016.json
+++ b/checkpoints/metadata_000061014016.json
@@ -1 +1 @@
-{"step": 1862, "tokens_seen": 61014016, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 4.134354490996065}
\ No newline at end of file
+{"step": 1862, "tokens_seen": 61014016, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 4.183231115945514}
\ No newline at end of file
diff --git a/checkpoints/metadata_000065536000.json b/checkpoints/metadata_000065536000.json
index bb529b5b0cd935e4afb3680420849cbf71aa27cb..095034cb30f07d2c539d55632afbe2bbc445e026 100644
--- a/checkpoints/metadata_000065536000.json
+++ b/checkpoints/metadata_000065536000.json
@@ -1 +1 @@
-{"step": 2000, "tokens_seen": 65536000, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 4.110466300872504}
\ No newline at end of file
+{"step": 2000, "tokens_seen": 65536000, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 4.158511824063886}
\ No newline at end of file
diff --git a/checkpoints/metadata_000067108864.json b/checkpoints/metadata_000067108864.json
index 1757eb988ac492c771c269c1acdae12053d66363..9e43bbafbeefa64ef788c0aeefa00c58293ecf34 100644
--- a/checkpoints/metadata_000067108864.json
+++ b/checkpoints/metadata_000067108864.json
@@ -1 +1 @@
-{"step": 2048, "tokens_seen": 67108864, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 4.087188933090007}
\ No newline at end of file
+{"step": 2048, "tokens_seen": 67108864, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 4.13458860322571}
\ No newline at end of file
diff --git a/checkpoints/metadata_000073826304.json b/checkpoints/metadata_000073826304.json
index feae5e8c48e60711ffdbb85611b8ba987fabba14..0bd394ca6a17cbbfc2527b46aef66c34741cde2f 100644
--- a/checkpoints/metadata_000073826304.json
+++ b/checkpoints/metadata_000073826304.json
@@ -1 +1 @@
-{"step": 2253, "tokens_seen": 73826304, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.97070491059368}
\ No newline at end of file
+{"step": 2253, "tokens_seen": 73826304, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 4.024738042947748}
\ No newline at end of file
diff --git a/checkpoints/metadata_000081199104.json b/checkpoints/metadata_000081199104.json
index a0c4218d2d588662edcbd50049228cb1fb4ebe0f..b698ae94e97926f7a8d1723c0dc6fa28f9d0d06d 100644
--- a/checkpoints/metadata_000081199104.json
+++ b/checkpoints/metadata_000081199104.json
@@ -1 +1 @@
-{"step": 2478, "tokens_seen": 81199104, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.8737115629455845}
\ No newline at end of file
+{"step": 2478, "tokens_seen": 81199104, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.927671735758744}
\ No newline at end of file
diff --git a/checkpoints/metadata_000081920000.json b/checkpoints/metadata_000081920000.json
index 3a831bc92542c503310faaa6aa9e960341ee8c76..725394c0ba78bb90207199c17cc59388220c1451 100644
--- a/checkpoints/metadata_000081920000.json
+++ b/checkpoints/metadata_000081920000.json
@@ -1 +1 @@
-{"step": 2500, "tokens_seen": 81920000, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.8585853468518287}
\ No newline at end of file
+{"step": 2500, "tokens_seen": 81920000, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.9130802265358606}
\ No newline at end of file
diff --git a/checkpoints/metadata_000089325568.json b/checkpoints/metadata_000089325568.json
index 1900aa4a0762fb0a255dc1645e91ea221d4d7027..1bcdd508cbe265fd9f824ee54812d9584252137c 100644
--- a/checkpoints/metadata_000089325568.json
+++ b/checkpoints/metadata_000089325568.json
@@ -1 +1 @@
-{"step": 2726, "tokens_seen": 89325568, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.7747637464189885}
\ No newline at end of file
+{"step": 2726, "tokens_seen": 89325568, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.8304603164886952}
\ No newline at end of file
diff --git a/checkpoints/metadata_000098271232.json b/checkpoints/metadata_000098271232.json
index 69c03cfbca3f59ddc3750ac1161f9fc26fe79a99..9cf92798184b3a206d8942ee6cb0f4bd235e678b 100644
--- a/checkpoints/metadata_000098271232.json
+++ b/checkpoints/metadata_000098271232.json
@@ -1 +1 @@
-{"step": 2999, "tokens_seen": 98271232, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.7082875029547786}
\ No newline at end of file
+{"step": 2999, "tokens_seen": 98271232, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.7642715290485924}
\ No newline at end of file
diff --git a/checkpoints/metadata_000098304000.json b/checkpoints/metadata_000098304000.json
index fe05efda3ec61761ebcd661636e651d38f216499..e21cb3f0b0a4e4144214d84c110a74c9c298cd7a 100644
--- a/checkpoints/metadata_000098304000.json
+++ b/checkpoints/metadata_000098304000.json
@@ -1 +1 @@
-{"step": 3000, "tokens_seen": 98304000, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.7076009311021108}
\ No newline at end of file
+{"step": 3000, "tokens_seen": 98304000, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.7636144447471813}
\ No newline at end of file
diff --git a/checkpoints/metadata_000108068864.json b/checkpoints/metadata_000108068864.json
index ce2f7c30bb5007b0d289b584e6f89fda470f5258..57c9f28533becf32b762d640094e8681ea43dae5 100644
--- a/checkpoints/metadata_000108068864.json
+++ b/checkpoints/metadata_000108068864.json
@@ -1 +1 @@
-{"step": 3298, "tokens_seen": 108068864, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.636359072621687}
\ No newline at end of file
+{"step": 3298, "tokens_seen": 108068864, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.694283825208702}
\ No newline at end of file
diff --git a/checkpoints/metadata_000114688000.json b/checkpoints/metadata_000114688000.json
index 227048f0521935ea72250b63a45ae9745aae6a27..9b478b9812509599b5e76f70f061a8571e8c476c 100644
--- a/checkpoints/metadata_000114688000.json
+++ b/checkpoints/metadata_000114688000.json
@@ -1 +1 @@
-{"step": 3500, "tokens_seen": 114688000, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.593479082164686}
\ No newline at end of file
+{"step": 3500, "tokens_seen": 114688000, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.65164887493004}
\ No newline at end of file
diff --git a/checkpoints/metadata_000118882304.json b/checkpoints/metadata_000118882304.json
index d479fbddbf8e5db81adc95f020df87b47c80b8fb..15832bff87e29c66cbf794c5943c408098a1ce57 100644
--- a/checkpoints/metadata_000118882304.json
+++ b/checkpoints/metadata_000118882304.json
@@ -1 +1 @@
-{"step": 3628, "tokens_seen": 118882304, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.601109134096057}
\ No newline at end of file
+{"step": 3628, "tokens_seen": 118882304, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.6573982438370076}
\ No newline at end of file
diff --git a/checkpoints/metadata_000130777088.json b/checkpoints/metadata_000130777088.json
index 2750ed3e5e1ef1fa5c25cf17093eb16d000e9381..0fe5c214dd2fc92cc0df2fd5e2f098bb0370f8c4 100644
--- a/checkpoints/metadata_000130777088.json
+++ b/checkpoints/metadata_000130777088.json
@@ -1 +1 @@
-{"step": 3991, "tokens_seen": 130777088, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.550488793464195}
\ No newline at end of file
+{"step": 3991, "tokens_seen": 130777088, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.6080018543413277}
\ No newline at end of file
diff --git a/checkpoints/metadata_000131072000.json b/checkpoints/metadata_000131072000.json
index 18849e9ca96ae5b68de14fc0c43a78c7ba26c957..a13961bcfcdbf04681adce56586f3d6b7e1ab6d4 100644
--- a/checkpoints/metadata_000131072000.json
+++ b/checkpoints/metadata_000131072000.json
@@ -1 +1 @@
-{"step": 4000, "tokens_seen": 131072000, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.5571278406617366}
\ No newline at end of file
+{"step": 4000, "tokens_seen": 131072000, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.6143836509798017}
\ No newline at end of file
diff --git a/checkpoints/metadata_000143851520.json b/checkpoints/metadata_000143851520.json
index 448586e0d1ab30dc4d690a5b53269d25e1601eec..289dd038fe4246d0751ffac83e6f51d71ed9b446 100644
--- a/checkpoints/metadata_000143851520.json
+++ b/checkpoints/metadata_000143851520.json
@@ -1 +1 @@
-{"step": 4390, "tokens_seen": 143851520, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.516337038989698}
\ No newline at end of file
+{"step": 4390, "tokens_seen": 143851520, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.5763677759688544}
\ No newline at end of file
diff --git a/checkpoints/metadata_000147456000.json b/checkpoints/metadata_000147456000.json
index 17e8a812b87c0d027a0353f3a5529ff5481b4209..f5a0e93b92798fd436c3bde2b3792a23911de7c4 100644
--- a/checkpoints/metadata_000147456000.json
+++ b/checkpoints/metadata_000147456000.json
@@ -1 +1 @@
-{"step": 4500, "tokens_seen": 147456000, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.5048954366844023}
\ No newline at end of file
+{"step": 4500, "tokens_seen": 147456000, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.5651027425659474}
\ No newline at end of file
diff --git a/checkpoints/metadata_000158269440.json b/checkpoints/metadata_000158269440.json
index 1a74e1ce5e1f7800d7e27db85ce7587d4abf0598..a7e806f1fbbf1d467b33aee6e9b1a463a4a84718 100644
--- a/checkpoints/metadata_000158269440.json
+++ b/checkpoints/metadata_000158269440.json
@@ -1 +1 @@
-{"step": 4830, "tokens_seen": 158269440, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.4822720271960272}
\ No newline at end of file
+{"step": 4830, "tokens_seen": 158269440, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.5430792461603593}
\ No newline at end of file
diff --git a/checkpoints/metadata_000163840000.json b/checkpoints/metadata_000163840000.json
index 5a6222b1f86d0e2a9b09aed0e616d860deff098d..45c60bd6ceb16eb6e7db403ca3d28a38afc42bea 100644
--- a/checkpoints/metadata_000163840000.json
+++ b/checkpoints/metadata_000163840000.json
@@ -1 +1 @@
-{"step": 5000, "tokens_seen": 163840000, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.464319650781479}
\ No newline at end of file
+{"step": 5000, "tokens_seen": 163840000, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.52576845325292}
\ No newline at end of file
diff --git a/checkpoints/metadata_000174096384.json b/checkpoints/metadata_000174096384.json
index 820bc4775926ef6f018b4d36a759993c6b617299..5e8ce295d26be0f92744862fa8cf6eb4e4199404 100644
--- a/checkpoints/metadata_000174096384.json
+++ b/checkpoints/metadata_000174096384.json
@@ -1 +1 @@
-{"step": 5313, "tokens_seen": 174096384, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.457515758075735}
\ No newline at end of file
+{"step": 5313, "tokens_seen": 174096384, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.518635521589029}
\ No newline at end of file
diff --git a/checkpoints/metadata_000180224000.json b/checkpoints/metadata_000180224000.json
index 4917538da5440464bf3d9d54e46618a83d4b6a26..b7ef04172a16775033215df9cf3f488490da10a0 100644
--- a/checkpoints/metadata_000180224000.json
+++ b/checkpoints/metadata_000180224000.json
@@ -1 +1 @@
-{"step": 5500, "tokens_seen": 180224000, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.435171628668802}
\ No newline at end of file
+{"step": 5500, "tokens_seen": 180224000, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.496639097510411}
\ No newline at end of file
diff --git a/checkpoints/metadata_000191496192.json b/checkpoints/metadata_000191496192.json
index dc49e0d1b798e38bc1aff80cc8aef63848f28682..15c3593964ade8598a247720fdc328e407d571d4 100644
--- a/checkpoints/metadata_000191496192.json
+++ b/checkpoints/metadata_000191496192.json
@@ -1 +1 @@
-{"step": 5844, "tokens_seen": 191496192, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.4124552194940834}
\ No newline at end of file
+{"step": 5844, "tokens_seen": 191496192, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.4751529632338487}
\ No newline at end of file
diff --git a/checkpoints/metadata_000196608000.json b/checkpoints/metadata_000196608000.json
index cb221ea4c8266a72c3feb632c6e8ce269ce7b76e..f5e73fb9aa1782d04a533279f7a1ca0a05cd5f14 100644
--- a/checkpoints/metadata_000196608000.json
+++ b/checkpoints/metadata_000196608000.json
@@ -1 +1 @@
-{"step": 6000, "tokens_seen": 196608000, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.4061981319064123}
\ No newline at end of file
+{"step": 6000, "tokens_seen": 196608000, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.4667979776226288}
\ No newline at end of file
diff --git a/checkpoints/metadata_000196706304.json b/checkpoints/metadata_000196706304.json
index ce6fcda663fa048fd81037b1686b7c1d12ade99d..3616d69d0c9f060aec57046856a761ea04ea34f8 100644
--- a/checkpoints/metadata_000196706304.json
+++ b/checkpoints/metadata_000196706304.json
@@ -1 +1 @@
-{"step": 6003, "tokens_seen": 196706304, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.4074958290792825}
\ No newline at end of file
+{"step": 6003, "tokens_seen": 196706304, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.467987821875873}
\ No newline at end of file
diff --git a/checkpoints/metadata_000197361664.json b/checkpoints/metadata_000197361664.json
index f9e7708d8ebc7af127c910c4684b4680751be145..5a5aa2479a487facafbdfd02cd4bb43a293104f0 100644
--- a/checkpoints/metadata_000197361664.json
+++ b/checkpoints/metadata_000197361664.json
@@ -1 +1 @@
-{"step": 6023, "tokens_seen": 197361664, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.408092934298367}
\ No newline at end of file
+{"step": 6023, "tokens_seen": 197361664, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.469824998532631}
\ No newline at end of file
diff --git a/checkpoints/metadata_000198017024.json b/checkpoints/metadata_000198017024.json
index 6ef2aaeeceec1383b420fc78c804b25e00da04e8..183007c52293fca3689a264336443d8903fa2f7b 100644
--- a/checkpoints/metadata_000198017024.json
+++ b/checkpoints/metadata_000198017024.json
@@ -1 +1 @@
-{"step": 6043, "tokens_seen": 198017024, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.407055360049092}
\ No newline at end of file
+{"step": 6043, "tokens_seen": 198017024, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.4694174847315966}
\ No newline at end of file
diff --git a/checkpoints/metadata_000198672384.json b/checkpoints/metadata_000198672384.json
index e6e02759141141cdd27f8bccb4fda874a7d1fccf..733d6428c7d8ae7e6af8df55ca487d669e9fb1a3 100644
--- a/checkpoints/metadata_000198672384.json
+++ b/checkpoints/metadata_000198672384.json
@@ -1 +1 @@
-{"step": 6063, "tokens_seen": 198672384, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.396747844162258}
\ No newline at end of file
+{"step": 6063, "tokens_seen": 198672384, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.459399634434712}
\ No newline at end of file
diff --git a/checkpoints/metadata_000199327744.json b/checkpoints/metadata_000199327744.json
index 71ac03b566844b42f50cf08167ada7bacd8830f9..790fcda4f1140aacdc832a6e198db76a2e7e8762 100644
--- a/checkpoints/metadata_000199327744.json
+++ b/checkpoints/metadata_000199327744.json
@@ -1 +1 @@
-{"step": 6083, "tokens_seen": 199327744, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.403324384960973}
\ No newline at end of file
+{"step": 6083, "tokens_seen": 199327744, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.4660093119605064}
\ No newline at end of file
diff --git a/checkpoints/metadata_000199950336.json b/checkpoints/metadata_000199950336.json
index a72f6c7b92a2dcca81dc2215710976a9adf66bb8..c16e16df2be11e6f5e6f06b5472547126445306f 100644
--- a/checkpoints/metadata_000199950336.json
+++ b/checkpoints/metadata_000199950336.json
@@ -1 +1 @@
-{"step": 6102, "tokens_seen": 199950336, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.394872257897474}
\ No newline at end of file
+{"step": 6102, "tokens_seen": 199950336, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.4578413544944855}
\ No newline at end of file
diff --git a/checkpoints/model_weights_000000032768.pt b/checkpoints/model_weights_000000032768.pt
index a912b9b69c392865dae7eb20b2190ee48cb78cc5..376c326ec71dbc6b804601b07905e6e80d816acf 100644
--- a/checkpoints/model_weights_000000032768.pt
+++ b/checkpoints/model_weights_000000032768.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:126e91c770ee10d33ef0dee8496e4965913669ea9a7354bf884bcdf37e7c5add
-size 152233941
+oid sha256:42433f5564c5d2d4c25460186789bd97ad364c1c5c2eef47d74cd86afc616302
+size 151183829
diff --git a/checkpoints/model_weights_000000327680.pt b/checkpoints/model_weights_000000327680.pt
index 2b63539146f3825bf9247c2de3baf602b1057e91..19063561ddd9d93960cc79bb8e2b1a501e1c67b1 100644
--- a/checkpoints/model_weights_000000327680.pt
+++ b/checkpoints/model_weights_000000327680.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ea98c23a24df1202ce83af552102b4277bcac1c2c7c94f9da950eb17405619f0
-size 152233941
+oid sha256:db794c3e5d8df814d89a81233ad2ba07778990fd19d2a8c96a920e020b1dd3ab
+size 151183829
diff --git a/checkpoints/model_weights_000000360448.pt b/checkpoints/model_weights_000000360448.pt
index faec3e6cc724ac45bdab04fe6d3aa176954f73cc..bf86fb0902fd26ebc2505e8ceba3e55d6ac82862 100644
--- a/checkpoints/model_weights_000000360448.pt
+++ b/checkpoints/model_weights_000000360448.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c8388e7b2b76bce252ad09d6838b7cd237b21bbd11e43f6f9799306f36773562
-size 152233941
+oid sha256:1dd70be592b6eabaa2c7bda67f4cbbfe21775502ca401f451ad5fb7069aeb743
+size 151183829
diff --git a/checkpoints/model_weights_000000425984.pt b/checkpoints/model_weights_000000425984.pt
index 2569ca3c9986d2eebce21a244c571db1d75e763c..15d1f2988371ca4f3924557770796dc95450e3aa 100644
--- a/checkpoints/model_weights_000000425984.pt
+++ b/checkpoints/model_weights_000000425984.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3917ef49615df35325dacbccba2940f7925e10f273e3b8a03b809d81fa79f6e0
-size 152233941
+oid sha256:6a7777a3ddb57b350f3177dd63c89ed0f89ecfe7e728a13d0df97b8fbd084327
+size 151183829
diff --git a/checkpoints/model_weights_000000458752.pt b/checkpoints/model_weights_000000458752.pt
index fb0bc8d9e232cc1d899c65dc684e6aefd0699c89..dcd79c52f4174c035605924c4153d1de7c789414 100644
--- a/checkpoints/model_weights_000000458752.pt
+++ b/checkpoints/model_weights_000000458752.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9b291e796976bf61ef32d6b25a318d435fafb9ff544822bb86ad9dc6bdee50c3
-size 152233941
+oid sha256:95b5f4e7ffef83a9ac3ebf622af21a2e629b0bf80bcabbc4871f976a43b10ec8
+size 151183829
diff --git a/checkpoints/model_weights_000000491520.pt b/checkpoints/model_weights_000000491520.pt
index bbabfa266688403c0b4bb69b295467b47519330e..8061814c1c3db953c94b6d3ed01861559f726498 100644
--- a/checkpoints/model_weights_000000491520.pt
+++ b/checkpoints/model_weights_000000491520.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:642a8a4244c50cc016d27d94755f1b1655d29a49e1c50557a577e296acc9db4f
-size 152233941
+oid sha256:a6f8cbe6dcba81cc86f86208b5ec74a3763697d6f1ddd46f117c9e6e0d87e128
+size 151183829
diff --git a/checkpoints/model_weights_000000557056.pt b/checkpoints/model_weights_000000557056.pt
index 0cc0d7784a5f8a020f58611c62c85f3b8cc8b19f..f89dfd185635bd77e3a40838d1ac0c158442ee31 100644
--- a/checkpoints/model_weights_000000557056.pt
+++ b/checkpoints/model_weights_000000557056.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3cb29c4a8239ae521964d645a1c3006b0c19d7aff7878279db43127ae09d749a
-size 152233941
+oid sha256:0463774351018f03ae653c12f7463646e654610b88ffd26bc1d128fc075a92e9
+size 151183829
diff --git a/checkpoints/model_weights_000000622592.pt b/checkpoints/model_weights_000000622592.pt
index a9dc21f7772aa20c59df58cd5501705d297ba875..2cc46c8ebcbbad73bd7df01d91cde23707731bd5 100644
--- a/checkpoints/model_weights_000000622592.pt
+++ b/checkpoints/model_weights_000000622592.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:df47ea9321b321ba87765b5756b7629b48706909590a5adacf9823f9ffd37648
-size 152233941
+oid sha256:df016922d93fe069ae76dcf58f3134d1100f759179d50c7badbfdcd3673a1e37
+size 151183829
diff --git a/checkpoints/model_weights_000000688128.pt b/checkpoints/model_weights_000000688128.pt
index 5dbe4e3dbfbb135c6790399aa231613adc5dc1d9..b02e2242a4d44cb17c70f83fcdc5368b3b58195a 100644
--- a/checkpoints/model_weights_000000688128.pt
+++ b/checkpoints/model_weights_000000688128.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2ba296c1826535e86b4097ed1eb5a63d7fb5233dc4fd9345f818fb4653c1bf79
-size 152233941
+oid sha256:d86f0e9be18093e885f7634ad6c8228959c22d5e7a2aa253fb20eb97842c7e2e
+size 151183829
diff --git a/checkpoints/model_weights_000000753664.pt b/checkpoints/model_weights_000000753664.pt
index 9efda2cc78dee96ba3409f9cc3503fe79a6400fa..b1059a2581bf8566832556116ea34969bde39c35 100644
--- a/checkpoints/model_weights_000000753664.pt
+++ b/checkpoints/model_weights_000000753664.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7e51f2cb71ae53a36f7e18300cf3bab5a0f1551f9d5eaf49997320b6f693eafa
-size 152233941
+oid sha256:638b19a3867bff80c72488e884cd5f37f8d71bf8d680ca983cf561fce8da6de2
+size 151183829
diff --git a/checkpoints/model_weights_000000819200.pt b/checkpoints/model_weights_000000819200.pt
index 701c895eabfe7fdd3f0be7baf08f327c2a4f88b0..3aec4ef2f9012c88d7e27db751cb174306cbd64b 100644
--- a/checkpoints/model_weights_000000819200.pt
+++ b/checkpoints/model_weights_000000819200.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:89e7f4890a08ee3f6cb4eea8ad6adcbea64111882dfdbc3dfe12fae7e1201f90
-size 152233941
+oid sha256:dd104b519fd0f20ed8987c167ea67f9988d7d7956a7e45e20146e113e87b51f0
+size 151183829
diff --git a/checkpoints/model_weights_000000917504.pt b/checkpoints/model_weights_000000917504.pt
index eb1e419cbb8e243f32d1c8b5c91a28736bd12a46..5d787972ea54fde61897364380d142d36a860c4f 100644
--- a/checkpoints/model_weights_000000917504.pt
+++ b/checkpoints/model_weights_000000917504.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d32e920c21376e8bbc60342aed5bca729a6bb035fb4c31d672c96a8bde9a0f3e
-size 152233941
+oid sha256:0f599c573d7305d1ba1c7539112de7bb3822588a5824e37168887cbde2f637b8
+size 151183829
diff --git a/checkpoints/model_weights_000000983040.pt b/checkpoints/model_weights_000000983040.pt
index 5d3d1b045c12a235bede5755c3883bf6d32cceb1..375c0c8b6277fc149a9ca346d5145039041ee643 100644
--- a/checkpoints/model_weights_000000983040.pt
+++ b/checkpoints/model_weights_000000983040.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:755c587dadfc7dbeab7660e6b1e3871bc67d2591350985da88bfe25cfe3477c6
-size 152233941
+oid sha256:4f7a67cef948740f7cac26379149cf0986f704ac0a0c0e6c80a9d17b3ed593c6
+size 151183829
diff --git a/checkpoints/model_weights_000001114112.pt b/checkpoints/model_weights_000001114112.pt
index 475234a9985e5b40008a8f27773cdd44a538ed5f..b2e969fa9f3d59331f5d96f3ebb7c0bd71bf7b34 100644
--- a/checkpoints/model_weights_000001114112.pt
+++ b/checkpoints/model_weights_000001114112.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:78d249d659833651df00fc1c946ef98ac5d34321290f97b49c08602fc744a07e
-size 152233941
+oid sha256:e0473da235bc7205a424439e0b2a398ef4006194e8fa571874ad81daf6650dbd
+size 151183829
diff --git a/checkpoints/model_weights_000001212416.pt b/checkpoints/model_weights_000001212416.pt
index 14e5f3069811e3fba530846a0d346cf61eced1a8..dac017a2edb1164cb692bcc57e1aa219d1541a72 100644
--- a/checkpoints/model_weights_000001212416.pt
+++ b/checkpoints/model_weights_000001212416.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ac757ab9ae59a742327542c4010e53789355750c273af57c418922049bd45c13
-size 152233941
+oid sha256:d0f3fcb05a9719802a88b31832a835f1a15d84525fbe5bb1a8d1068314658df0
+size 151183829
diff --git a/checkpoints/model_weights_000001343488.pt b/checkpoints/model_weights_000001343488.pt
index a189b3bb83a92186eb877a0d10a42308330ad864..ad72843801e15c4b0712817eabb244b216c85f07 100644
--- a/checkpoints/model_weights_000001343488.pt
+++ b/checkpoints/model_weights_000001343488.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8c7cb7b849007d479509bcb97749fa8eed22a45e3db82d08883bd7f012383419
-size 152233941
+oid sha256:39e6b66db637360c2b1abe19aae6c56f147483e32c8c6c55a00f234079557876
+size 151183829
diff --git a/checkpoints/model_weights_000001474560.pt b/checkpoints/model_weights_000001474560.pt
index 9177c54e697abfea0280d11d05a8c38dbc6dc78a..32109fc9264949593f4925dc551af42dbddbffb9 100644
--- a/checkpoints/model_weights_000001474560.pt
+++ b/checkpoints/model_weights_000001474560.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e7b4cb1b8f3b9891e35f3517e843d51516924a2fdba36f91bd7f05ae2df17b7d
-size 152233941
+oid sha256:57ee8c9e66069116e157beac7b5e2decffb05050e27250c971983cbfee1f177f
+size 151183829
diff --git a/checkpoints/model_weights_000001605632.pt b/checkpoints/model_weights_000001605632.pt
index 9254f18b0adf59e749eebf81bd43f6579ea9f856..4eaff40abb3948aec1100330017553d7eb5a98e4 100644
--- a/checkpoints/model_weights_000001605632.pt
+++ b/checkpoints/model_weights_000001605632.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:86e31677717bdc8294fc7eca35b5ef9bfc2cb261186f2de93f3bea134e7c0fb3
-size 152233941
+oid sha256:204c319617ce9ec5767f169407c8c174baa2af19d3c14a199d01efc65dc491ed
+size 151183829
diff --git a/checkpoints/model_weights_000001769472.pt b/checkpoints/model_weights_000001769472.pt
index a7e84fdba917d3f91f16855ae9426ff8c9fc40d9..e9863b1820d8faccaee73b1d525b1942ec5b319c 100644
--- a/checkpoints/model_weights_000001769472.pt
+++ b/checkpoints/model_weights_000001769472.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6ec994fd9c622b838312f914b019f46a1f4682b75f0259095aa79447cfdb55cc
-size 152233941
+oid sha256:a02724cecf461b8b8200029ae86f6eb24f824bc25e85e69e2590fb9f02d4ee77
+size 151183829
diff --git a/checkpoints/model_weights_000001966080.pt b/checkpoints/model_weights_000001966080.pt
index d8a780005510d1e5e91d02939415fc3c123544e1..5e6e50635d3b7b04c2f3c0fd88fde3a3a1bdfaa3 100644
--- a/checkpoints/model_weights_000001966080.pt
+++ b/checkpoints/model_weights_000001966080.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:012058c0a6d2f457ddc997b74be0ef3b9006aedeef426bce61ab9eba832b89df
-size 152233941
+oid sha256:b904da673c2c2feb4c34afa919c8381b94a9c765ff8ed4b45f1f76f82242cf3b
+size 151183829
diff --git a/checkpoints/model_weights_000002162688.pt b/checkpoints/model_weights_000002162688.pt
index a7fa17e0b366d57d2961a9958de90d03480f666f..d7e99bb049bacef878c73477d0f0cfb4be1d9409 100644
--- a/checkpoints/model_weights_000002162688.pt
+++ b/checkpoints/model_weights_000002162688.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3d109259b8927d0bee42e2999c5fdf0eeca64a2a484f875db58e752418946d34
-size 152233941
+oid sha256:5261d7dc82bda08823ad9026a093362eb0a210a5be66b2ad4d979b3650d5a9aa
+size 151183829
diff --git a/checkpoints/model_weights_000002359296.pt b/checkpoints/model_weights_000002359296.pt
index 05e69b9fd548cd6bf68ab2fc8be4ba0e3db1e0db..ef3cc0a2ef7d569608c58e45a6f4d27123a4eb9e 100644
--- a/checkpoints/model_weights_000002359296.pt
+++ b/checkpoints/model_weights_000002359296.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a919bccd79ef2a7249045e9c1f7ff08bc1075a19b5a66f455e37f71181044ad4
-size 152233941
+oid sha256:f4ca07aee0021b68ded9b96d5b193fbf1695019ad8f9a188e32c82c45f00fa20
+size 151183829
diff --git a/checkpoints/model_weights_000002621440.pt b/checkpoints/model_weights_000002621440.pt
index 8fb540f0e0e30638a3a1c1827747f8085952ba30..1746c53738064742a413254bdfcb52b08497dc5d 100644
--- a/checkpoints/model_weights_000002621440.pt
+++ b/checkpoints/model_weights_000002621440.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6c0ff6d6283609474fc694d39fd180e288929b5194e822e612720e7358958676
-size 152233941
+oid sha256:07f146338da7afe3e283927b973ccbfaad53d5dba0295e50d65acc9d9b378899
+size 151183829
diff --git a/checkpoints/model_weights_000002883584.pt b/checkpoints/model_weights_000002883584.pt
index c3fc48d8073ba80c2e3ef5757efbb27887ac89ae..68d507ec9eb81ca45b4f7a0ad62a99bd9998fed9 100644
--- a/checkpoints/model_weights_000002883584.pt
+++ b/checkpoints/model_weights_000002883584.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5b25bed5c785fc1204d1e4281ff82ee1da20b90af02488fc91b9ba9c4a8c60c4
-size 152233941
+oid sha256:5c2c6a3e3df454ecad9fffb523d27a81e278d55fb32dc5feb29caa946cd85a30
+size 151183829
diff --git a/checkpoints/model_weights_000003178496.pt b/checkpoints/model_weights_000003178496.pt
index 8064f00ad6dc98bc1165129919f986f70bbbfcc3..75c04f5edb859391602e4aa4132fd4e71170fee0 100644
--- a/checkpoints/model_weights_000003178496.pt
+++ b/checkpoints/model_weights_000003178496.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0a24228699d57eb3fd0117d453029b057a0b2bb3802ecafc6c2338d7a9f3e78d
-size 152233941
+oid sha256:2a4cf0c0648f20f77d3e22ed014cfa33e4e31c6d4a67f7128bf011b09f768c30
+size 151183829
diff --git a/checkpoints/model_weights_000003473408.pt b/checkpoints/model_weights_000003473408.pt
index acfbf29ae253cec0fbfb4240bc084eafe92a3d8b..0186375b7ab9e55f823845525eea558aeb317917 100644
--- a/checkpoints/model_weights_000003473408.pt
+++ b/checkpoints/model_weights_000003473408.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:423c2886095a6fccb4fd402f82ad314907f99d3045ef5a99f809022aa8b6bbbc
-size 152233941
+oid sha256:08e18b6fd6ba0e258c99c5cbe2fdfe45ba24986c2ea1bd20000bdb8bbd5ab4dc
+size 151183829
diff --git a/checkpoints/model_weights_000003833856.pt b/checkpoints/model_weights_000003833856.pt
index 16915b15238bb94c964d8cdcfb35fdd1c75f2385..9ba787e7161e0039a0a7e20369f0765a8d93542e 100644
--- a/checkpoints/model_weights_000003833856.pt
+++ b/checkpoints/model_weights_000003833856.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:afc1ddf5ef6c825ced0502a0df7489cd6974991022036a49b6ff0173a7919e87
-size 152233941
+oid sha256:23a7729a5ad3cf0fd0ebec77dbc5c3c78420d5404d8eb9b89eabbc68584a851f
+size 151183829
diff --git a/checkpoints/model_weights_000004227072.pt b/checkpoints/model_weights_000004227072.pt
index 7891151d758869db673486fe6a4686d902e7a5b5..bb5452dc8c775c8ec756d20adc80ad8f10209e1c 100644
--- a/checkpoints/model_weights_000004227072.pt
+++ b/checkpoints/model_weights_000004227072.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:faae6d34dda55becb2508088ca57995ca954837911326285acf14db1cc52206c
-size 152233941
+oid sha256:c9382f7259d338426cfde10b275afe8f19c00fdb98f47347768d365ec785a0d2
+size 151183829
diff --git a/checkpoints/model_weights_000004653056.pt b/checkpoints/model_weights_000004653056.pt
index c8704ea794d7582b9225c9af22fc8382b547751a..e37c8f096bafd11940b9cf52f851fe6a9d639763 100644
--- a/checkpoints/model_weights_000004653056.pt
+++ b/checkpoints/model_weights_000004653056.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:94d9f889bf05c772c064e57ff19b43395f79187ed9c3ac3ebecea22b7b7ef46e
-size 152233941
+oid sha256:ad5352ada566445b01eee6fc71c36a49c857c8e58626a9bf1f1a1a76296a0412
+size 151183829
diff --git a/checkpoints/model_weights_000005111808.pt b/checkpoints/model_weights_000005111808.pt
index 9b40b0979bb9cb2771e89a3883d9c934983f2cd6..aca5974b67baeb89d0f94d704c39c1b2da151bec 100644
--- a/checkpoints/model_weights_000005111808.pt
+++ b/checkpoints/model_weights_000005111808.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ea9793e864030102e412c47a7026f175f7b2beb65e1c3c4f1191db17faacf838
-size 152233941
+oid sha256:f9ed98aef22cede1dbfcc52f6d25336c81e17f044ace0dcbd84ade7cc2358141
+size 151183829
diff --git a/checkpoints/model_weights_000005603328.pt b/checkpoints/model_weights_000005603328.pt
index 21fbef56f8733adac9cda032444c0be8572a84a1..4d53c685f50c644f620491bd70630db581f8c2a9 100644
--- a/checkpoints/model_weights_000005603328.pt
+++ b/checkpoints/model_weights_000005603328.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0d698e8eeed127a4bdcca424e0a7415778b814bc3d5a6d9b9d8439708ee33b51
-size 152233941
+oid sha256:1a4d5517914298d9aa5f3d9d3e375323932c280d9de39ec7eb3d6b49653c046d
+size 151183829
diff --git a/checkpoints/model_weights_000006193152.pt b/checkpoints/model_weights_000006193152.pt
index 47c0c86d113acd749a5720bd0db8cc863a113a0c..f02319a5956a28fb949f08adf07ce46342e210dc 100644
--- a/checkpoints/model_weights_000006193152.pt
+++ b/checkpoints/model_weights_000006193152.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b873b008c2cc4cbf0943b3fb4f16f2c2bb93dc0a52415cfeaf7ed7b1b3ad78e3
-size 152233941
+oid sha256:ba5a26e612e66843500e1b8c5a3a0910c97fc7753b399d8411d3978eb340a07b
+size 151183829
diff --git a/checkpoints/model_weights_000006782976.pt b/checkpoints/model_weights_000006782976.pt
index 5b1dc2622f69343033fc20fa8ebc43276de69184..b11fd169c7d0fe9acae47092eacea131115237a6 100644
--- a/checkpoints/model_weights_000006782976.pt
+++ b/checkpoints/model_weights_000006782976.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3feb8d2736812ca3272ec7f9e044097cb48874a2053cf71d11d80e8c1e9f8f23
-size 152233941
+oid sha256:083f24cdb0e950f84e0f34b4edc4767f7e2348328286ff9deb51d2a910d1c827
+size 151183829
diff --git a/checkpoints/model_weights_000007471104.pt b/checkpoints/model_weights_000007471104.pt
index e8e575374c9c7a9e8aaf2735872218351105a955..9d24e020d53f28008317a3d29ce0a9471ab3af86 100644
--- a/checkpoints/model_weights_000007471104.pt
+++ b/checkpoints/model_weights_000007471104.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:01102ee0a169f6142736d385d72fe2c79192848731d9cc4143048b887176d423
-size 152233941
+oid sha256:93487c03a02fb1addc8352798ea870e408833a2736dd71dcced0a13dc6eba50b
+size 151183829
diff --git a/checkpoints/model_weights_000008224768.pt b/checkpoints/model_weights_000008224768.pt
index 3d43ef91d7d0de07b30a6612a40b50d05d3e9efb..50df6c2762eea4ffaef63c3df18389ac79367af0 100644
--- a/checkpoints/model_weights_000008224768.pt
+++ b/checkpoints/model_weights_000008224768.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:af5c9aed40ecd638baf207e10fad27acec54e665bd912b210a6405e847680ca0
-size 152233941
+oid sha256:d526f740034d485dbad918bbe050f19ecc8f3bc0244c9d2064c649f3fd7cf99b
+size 151183829
diff --git a/checkpoints/model_weights_000009043968.pt b/checkpoints/model_weights_000009043968.pt
index 82e2d5f550a33e18e65660e5041fe52bfa6b443e..683fd6956c01355d3a0ae433e4e0305d29b482bd 100644
--- a/checkpoints/model_weights_000009043968.pt
+++ b/checkpoints/model_weights_000009043968.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8ec1c3a070a125c58a8152f9e7aa56a763037cc123e36e1ca323ded182a76b28
-size 152233941
+oid sha256:3bb36f8be7f35222b80b0922e9d28f894af35a06ad8492a4a21acaef5a3feec4
+size 151183829
diff --git a/checkpoints/model_weights_000009961472.pt b/checkpoints/model_weights_000009961472.pt
index 6b1732733e6b3e13c4bb850f49630acffb1dfce4..b0e02403a11d6c2518d1df65113e9842d685bde4 100644
--- a/checkpoints/model_weights_000009961472.pt
+++ b/checkpoints/model_weights_000009961472.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:adc6b64d722aeb3d5891411326c3ebd46c6ed896e1a91613d26e141e58e07f0f
-size 152233941
+oid sha256:3934ad71e84858fc2d50d4fa7cd92cf148efbfe68f6a722bdfbb027b2bb3ccfd
+size 151183829
diff --git a/checkpoints/model_weights_000010944512.pt b/checkpoints/model_weights_000010944512.pt
index 1d8fc60df1d8db055ede761be609c87ce24d9efc..fd214cc9e44e67aaaa6e817b6ce665a42bf6022d 100644
--- a/checkpoints/model_weights_000010944512.pt
+++ b/checkpoints/model_weights_000010944512.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6bf00dde065e4d6d87753ab5a3a5078d0e8dabfbf2f2a1af2b58e028d02f2f8e
-size 152233941
+oid sha256:8bbc9cfafcecf0273ca6514fe38193714af5d9e7eab4a238d75e68be525de510
+size 151183829
diff --git a/checkpoints/model_weights_000012058624.pt b/checkpoints/model_weights_000012058624.pt
index d1886e335bf71b91c53a8087f00c6daaece8cdf0..0f5c74cbfe2ed2540e64e1bd482ea15a16819fbd 100644
--- a/checkpoints/model_weights_000012058624.pt
+++ b/checkpoints/model_weights_000012058624.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cf7cec5bed9521d96c13b93dc544f2c2d308eab2d5c7e12560d04c47f86bc176
-size 152233941
+oid sha256:b5d2e6e5f5db1e53cc0a370e610c0054e49df0e98a5ae830a304dbc357613d5f
+size 151183829
diff --git a/checkpoints/model_weights_000013271040.pt b/checkpoints/model_weights_000013271040.pt
index 255b0372e30c9286f214df77c975ed208d0d9dad..f1462a56348de7811dfbe0b4ad30b4e747ee6964 100644
--- a/checkpoints/model_weights_000013271040.pt
+++ b/checkpoints/model_weights_000013271040.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2332cb6ff3d3a184384af69b52bc9c1c7cac3b40022b009920334144a0296d24
-size 152233941
+oid sha256:46446fb28d810141b2f8e96d991bb8f58069dd3f08a4f2d5e930b3ce71694898
+size 151183829
diff --git a/checkpoints/model_weights_000014581760.pt b/checkpoints/model_weights_000014581760.pt
index bc8176574d9bacad2342276f023efa58d1833c78..8b5efdc0e85f4efb46ba7920c360fbeac8ed2302 100644
--- a/checkpoints/model_weights_000014581760.pt
+++ b/checkpoints/model_weights_000014581760.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:42dc3d4700cec02974c3720c719319c7fea5bbb1a11fd62efa367aad310f7abd
-size 152233941
+oid sha256:e95da01edab15e09f239369497d0f5c7d26e5422a4f1dcb3978eeb39b86171ef
+size 151183829
diff --git a/checkpoints/model_weights_000016056320.pt b/checkpoints/model_weights_000016056320.pt
index 2ec9e33eca945b9bb8cf8d38f544d5cacc67daf5..bc03e1c9f4634efba5a8221e0473bcf28c4205bc 100644
--- a/checkpoints/model_weights_000016056320.pt
+++ b/checkpoints/model_weights_000016056320.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:09f58d2c1ccffae75a9a52a63da53a6f98329e0e836abfd6a9ec0a7450045e15
-size 152233941
+oid sha256:828ff7704873a2faf8541d9199322f47ba105af7824e2cab720efe44447cfe5c
+size 151183829
diff --git a/checkpoints/model_weights_000016384000.pt b/checkpoints/model_weights_000016384000.pt
index 0408b044443c73da974f321fb42ca97d79b14fdb..5801beb772f24955afe66546518bd4aaf468367a 100644
--- a/checkpoints/model_weights_000016384000.pt
+++ b/checkpoints/model_weights_000016384000.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d625a5b9816d03f4cb433038eb4107e4aa0cadf0c8704548cc2dd5be1e9d1fb1
-size 152233941
+oid sha256:52ccd8304cd0bfa6068838b9ca42704d5d6796c4d6697bc988e49ca7b30a080c
+size 151183829
diff --git a/checkpoints/model_weights_000017661952.pt b/checkpoints/model_weights_000017661952.pt
index 52ab525700d8dee59df2f3ba28e171f09a255a65..48d43bd62b74ba5b6733582cb42c629ca6a95e91 100644
--- a/checkpoints/model_weights_000017661952.pt
+++ b/checkpoints/model_weights_000017661952.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:aecad5609a506e4bfdfd454fa850b361a8c05279eaf105d553943b63061e32fe
-size 152233941
+oid sha256:f37334ac2b0ae22a9a4bca6399f811c5e9d0d2c982e6c8a0f987fc78baba8777
+size 151183829
diff --git a/checkpoints/model_weights_000019431424.pt b/checkpoints/model_weights_000019431424.pt
index f17342914969257159a68a350520580bb51c3768..a0f9d79c9e108ae30af94c1a803f0e2e634a9f70 100644
--- a/checkpoints/model_weights_000019431424.pt
+++ b/checkpoints/model_weights_000019431424.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:093f6cd231419ec95b430623b2b141b603d309d0862e9f163b4d892d2b8f1786
-size 152233941
+oid sha256:e75c402639aa9c6c399f478431f2c2fea7fc61d00012687b92c8b429237a16dc
+size 151183829
diff --git a/checkpoints/model_weights_000021364736.pt b/checkpoints/model_weights_000021364736.pt
index f94240825b86174f244ded1928209dd4d2cf76a1..0f23f72e9d3048640eba3b4c871840e1fe521095 100644
--- a/checkpoints/model_weights_000021364736.pt
+++ b/checkpoints/model_weights_000021364736.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:164c1b90a703d53adfe900f7dee5bc994e42c6e2563f17cef932c570943a362a
-size 152233941
+oid sha256:11567a200db2ec7810ac9940014e175dd5cd959cac8566aab1aca6a83b573364
+size 151183829
diff --git a/checkpoints/model_weights_000023494656.pt b/checkpoints/model_weights_000023494656.pt
index 7e46eaa709fe7eb5e5af3c5788334df71f018186..21ed12d5711a75f95923f79e9c03882287aaa616 100644
--- a/checkpoints/model_weights_000023494656.pt
+++ b/checkpoints/model_weights_000023494656.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4c528494b42970ead192e2adda45d77e8600f8a2856f6858eb7f696e726b80b7
-size 152233941
+oid sha256:6838fd8b423557094093aed4910cbd6cf4250c9e517c55009409e807a629c2ae
+size 151183829
diff --git a/checkpoints/model_weights_000025853952.pt b/checkpoints/model_weights_000025853952.pt
index 5a6c66612a50d420dd428ef4ce0bbe42381f0213..11179895b3ef2ec5f719d421304d7c3765e152b9 100644
--- a/checkpoints/model_weights_000025853952.pt
+++ b/checkpoints/model_weights_000025853952.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9c8102377bb9f3a4bccf97b7a7cef5299b1f5e49c052e9922cc3f802096393d5
-size 152233941
+oid sha256:cbfeb2be6a31ea180ef2013ffdaa0f39690441aa9bcd48def933fb74f7e56ed2
+size 151183829
diff --git a/checkpoints/model_weights_000028442624.pt b/checkpoints/model_weights_000028442624.pt
index b07eafba18c566fe9b92b7a3aae5dde623159988..9f5d4c299ee4f9ef6a45889e5d673e110f70c8ac 100644
--- a/checkpoints/model_weights_000028442624.pt
+++ b/checkpoints/model_weights_000028442624.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1baaffe2770f4f6e4e4ca88348ee82a2c812ae062ad8644f22c0fc8ca6c46632
-size 152233941
+oid sha256:61ce07417697759084c7e3ff9b31b5aba3095ab0d9b9309c3ebf4aacd5022dd1
+size 151183829
diff --git a/checkpoints/model_weights_000031293440.pt b/checkpoints/model_weights_000031293440.pt
index ce334b1210b8c5d20edad3e3c8e89e7f3c08b212..4b69774535a1a089d74eb78f6cba59bc4c03759d 100644
--- a/checkpoints/model_weights_000031293440.pt
+++ b/checkpoints/model_weights_000031293440.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:503f8a171c1e370ea78dad681fdd7b89e9d9ce2d3509bff4aac6ff60e2e0f337
-size 152233941
+oid sha256:042561b3ec35d52a3533fdcb99afdc9f6144e5c5eba960419adbf0103e01456e
+size 151183829
diff --git a/checkpoints/model_weights_000032768000.pt b/checkpoints/model_weights_000032768000.pt
index 226023dc2caf5968a6fb6c1059bb88a07d71ea81..45a50870ad83f2a62d4ff0af26f3f85e9cc02311 100644
--- a/checkpoints/model_weights_000032768000.pt
+++ b/checkpoints/model_weights_000032768000.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:acbcfdb221584d350a91b208e3b95a8774a0eae4fd255c455a8ce3fcb210d671
-size 152233941
+oid sha256:f614f90bab827faa96b6de3ca2e178ef708f9454b85e24601d5bb54286b9b1f5
+size 151183829
diff --git a/checkpoints/model_weights_000034439168.pt b/checkpoints/model_weights_000034439168.pt
index 0eb974dfb4223bb07bf9a149710cc80ec1f1d877..2d32f62fc6edbffe458a092ce16e02f7b8a81658 100644
--- a/checkpoints/model_weights_000034439168.pt
+++ b/checkpoints/model_weights_000034439168.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7fdd604445a5aa3e34356d6652a6fe689d68bc8604be7e5fa562e1d8529d295b
-size 152233941
+oid sha256:91f2646a7df66bd7312d63973b4ac1eae7e759f4ad7dae615188adbe7a0d5bf5
+size 151183829
diff --git a/checkpoints/model_weights_000037879808.pt b/checkpoints/model_weights_000037879808.pt
index c18014f3af41a222c8b4bac0163971461a52fb6e..1eb7259b4cb230c0d0ca277432d33a69b1454236 100644
--- a/checkpoints/model_weights_000037879808.pt
+++ b/checkpoints/model_weights_000037879808.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4ac997ac32245bab013ef3f95ea9d087bdc4047eaa14aedb88b0d056e8aab576
-size 152233941
+oid sha256:f47d553fbfe3f73aedff54fd431f1e2370ba89bbed1cc6482d6c2b42593aeb43
+size 151183829
diff --git a/checkpoints/model_weights_000041648128.pt b/checkpoints/model_weights_000041648128.pt
index c765fc096e0f2c4b902e9311870a2b92d9a886bb..2af16d30909c34e9facf7045a2d34d37760c12c4 100644
--- a/checkpoints/model_weights_000041648128.pt
+++ b/checkpoints/model_weights_000041648128.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0e28e7ac643111e0d6623d362311f8f93488d0fef7f10fe8c7e50c48be1f9051
-size 152233941
+oid sha256:80dee0538147c669d85dcf4ce90aec4aea903cae25964555ee2a470de2612018
+size 151183829
diff --git a/checkpoints/model_weights_000045842432.pt b/checkpoints/model_weights_000045842432.pt
index 824c2220548e5b00121e7f6e410de407ec0b9ecd..0343f22bcfc6bb775e739ef906549a0d645b3f42 100644
--- a/checkpoints/model_weights_000045842432.pt
+++ b/checkpoints/model_weights_000045842432.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:145dae94ded6da1cac5c684072745603739c8be20f7a98ddd293bfec9786fb6d
-size 152233941
+oid sha256:a438f8998185c715001a3841ace7e917863da8472b8340cdbf66552f1d043298
+size 151183829
diff --git a/checkpoints/model_weights_000049152000.pt b/checkpoints/model_weights_000049152000.pt
index 2a0725b5dda375953e52465991f20b3806bdbfd5..a7db098f9f92e20293f792021ca815bdecb030c1 100644
--- a/checkpoints/model_weights_000049152000.pt
+++ b/checkpoints/model_weights_000049152000.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d17b4f03b730b1f4fc93ca8e5b1b3b261a2277dc8137a70dd41a16757bd02c56
-size 152233941
+oid sha256:eb3fa59b41a4223650318d22e1601ad2a7a46711fef2b2a14bd1c6ed280d69de
+size 151183829
diff --git a/checkpoints/model_weights_000050397184.pt b/checkpoints/model_weights_000050397184.pt
index ceddcb316c05d99837ad7f68f78d701fbfbc844b..9ab3c3dc46a9d492eed5dfd3a346fb8fd289836a 100644
--- a/checkpoints/model_weights_000050397184.pt
+++ b/checkpoints/model_weights_000050397184.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:550774647b6b21f542e14c314fcd670e19b56819a734c09060450d93abd5c29c
-size 152233941
+oid sha256:bef4f0c18c238515259680784c3634d2d9ba2ac1e7c303684105d6dd3200cd2e
+size 151183829
diff --git a/checkpoints/model_weights_000055443456.pt b/checkpoints/model_weights_000055443456.pt
index cd73743fed5d3cc74de48590e11767d265a210be..f1b298943e76c59470aa1e5aca29967e4949f97e 100644
--- a/checkpoints/model_weights_000055443456.pt
+++ b/checkpoints/model_weights_000055443456.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d1714506a833c289bcf2c715ed5257bcd5284b097d4f6453ca8192887897fe01
-size 152233941
+oid sha256:281776104ff868bda35d31807f41eb483139f1a491ee48e5ca0e9e0d5adbdc8c
+size 151183829
diff --git a/checkpoints/model_weights_000061014016.pt b/checkpoints/model_weights_000061014016.pt
index 7deb9759f973bd0666b38baf101714ea5d6b4f64..8c6358d272b4d87ae587f4cb014ddeccf9211634 100644
--- a/checkpoints/model_weights_000061014016.pt
+++ b/checkpoints/model_weights_000061014016.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fd9f1ea57c9011456c6c76f6fee2ea9798a8ab3c2f19e09231561e0c03823827
-size 152233941
+oid sha256:2b31d7d39989c7ff62c47ef826577d40a8d3418f2c49cd15e2ee726debfd58ae
+size 151183829
diff --git a/checkpoints/model_weights_000065536000.pt b/checkpoints/model_weights_000065536000.pt
index 3b410aba36229e58a5462945242d20c3254242af..ff42a72c2bf3d2e6fddc574a04e3a0dff892f4f8 100644
--- a/checkpoints/model_weights_000065536000.pt
+++ b/checkpoints/model_weights_000065536000.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4cad227f3d000126119966f6d62a26b065fc793b02a42a92ef40bc00dcff35e1
-size 152233941
+oid sha256:d5d01fc8e8cbf1be76dde579e17a4d1a97089c28df6b5b08580563b8dafaf654
+size 151183829
diff --git a/checkpoints/model_weights_000067108864.pt b/checkpoints/model_weights_000067108864.pt
index 6f82840deff98192c06e6394a950bb45135c5844..53e9ca243caec38f03f4c2b40298bff8802429ef 100644
--- a/checkpoints/model_weights_000067108864.pt
+++ b/checkpoints/model_weights_000067108864.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b91b60d58242c335595ec8c6f8c90bb871a777743f37015f6614e73fbf81b679
-size 152233941
+oid sha256:a21031ab0598fd69733d09e98d8d3cb931e35b7a4963a9f980e24a584d5e2bcc
+size 151183829
diff --git a/checkpoints/model_weights_000073826304.pt b/checkpoints/model_weights_000073826304.pt
index ec285c3fd1c53bfb798f0c4cb8e3f4cf1d2fa38e..79822cdc629d1f7e31514d87cd69d3fbb2658052 100644
--- a/checkpoints/model_weights_000073826304.pt
+++ b/checkpoints/model_weights_000073826304.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:894dd15c2ae9d0889fdeae587be200f108853520f5df96705ff80dcee9c0a123
-size 152233941
+oid sha256:314b1497679e95d116e8de86e2860d1737cc8ecb6ff1abf7e1c59e970fdee058
+size 151183829
diff --git a/checkpoints/model_weights_000081199104.pt b/checkpoints/model_weights_000081199104.pt
index 4e09a6f650c83426334a15c3d1eda01ab2cea4a8..cbd579f41857871d959be57b3fa25f249bba93d4 100644
--- a/checkpoints/model_weights_000081199104.pt
+++ b/checkpoints/model_weights_000081199104.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:35a217e9b909d9230421a822b1eaa20bd7b9e496fda0621b7628bfbbef0fd170
-size 152233941
+oid sha256:74f658f4876476245691bf56fec17a2a395ec0471b7f17a7fb54b68adbab13c3
+size 151183829
diff --git a/checkpoints/model_weights_000081920000.pt b/checkpoints/model_weights_000081920000.pt
index 0c9523bf3db1af112af2dfafc5671bb5dbfbc5b9..81b0ffc1ac7cec9a04917782a679904dfa133e5e 100644
--- a/checkpoints/model_weights_000081920000.pt
+++ b/checkpoints/model_weights_000081920000.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:28d461ede60245a11fe7aef30a2e8ce56906383c039be6a4ac5c0c319eab5573
-size 152233941
+oid sha256:c9bc4c2470f520c4ed3f4ce01e60e855eddaf23181c0ea4340e07286c0c4c3ce
+size 151183829
diff --git a/checkpoints/model_weights_000089325568.pt b/checkpoints/model_weights_000089325568.pt
index 8553d028d8f2b39302ab9a3dd06ad5d95a7c0eef..310c9e855a8e039c462a6ec112e19262cde0fcc6 100644
--- a/checkpoints/model_weights_000089325568.pt
+++ b/checkpoints/model_weights_000089325568.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:014156fcd4ba5cbc68d73713da1cd95344e30180c7b1c47cd64e8cab53d441b8
-size 152233941
+oid sha256:9c14d7690e178ee57282a4e2efc4204d6baaed9d68ead3dac949607713305ac7
+size 151183829
diff --git a/checkpoints/model_weights_000098271232.pt b/checkpoints/model_weights_000098271232.pt
index 816778f4c887d11a7c6dbacfc53e4f30fd142c3a..d88e41e93e4a37f89aba2c990d609d28172d6709 100644
--- a/checkpoints/model_weights_000098271232.pt
+++ b/checkpoints/model_weights_000098271232.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ab0baa7b27d2e5a28c30520e51d756c85a6e6bd3d15e46e969713cb1cc06cc0f
-size 152233941
+oid sha256:41677d899a01ccf663f74a8a9bf3971c5a5e84a90324c60d21300ddba55e758c
+size 151183829
diff --git a/checkpoints/model_weights_000098304000.pt b/checkpoints/model_weights_000098304000.pt
index 07c62a9a7ccb00073765d38ee743654875b8d2e4..db499af120b15aba8c84bf6b2ba2738698e88a19 100644
--- a/checkpoints/model_weights_000098304000.pt
+++ b/checkpoints/model_weights_000098304000.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:14ca2b202fafbeb89ebcc905d7c537b3c2f00a633ce131432374d62323bc430f
-size 152233941
+oid sha256:f81a65ba5579eb6b59e17d218983f0f3eab85fccdfd786ba27ccee9e5111a3f3
+size 151183829
diff --git a/checkpoints/model_weights_000108068864.pt b/checkpoints/model_weights_000108068864.pt
index 3691ba6d872ec388fd1aed5444ff74ab57cf31ae..a11ae31e330babbc8e6af349a47d664d7f715e5f 100644
--- a/checkpoints/model_weights_000108068864.pt
+++ b/checkpoints/model_weights_000108068864.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1ba49386eaa354e23e0332e0633091419e30097b6e330f9a7e425c4d2e69f335
-size 152233941
+oid sha256:4ec8cac49cb57a18bc9ca472040af42605f6588d1a3c1db216bf4649d5c21602
+size 151183829
diff --git a/checkpoints/model_weights_000114688000.pt b/checkpoints/model_weights_000114688000.pt
index fc2333b0f59f7eb7a6c2d7b1b01db81e527649b1..8370268805a180a2ba93acdfd7be04fa9cd60971 100644
--- a/checkpoints/model_weights_000114688000.pt
+++ b/checkpoints/model_weights_000114688000.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:453c809cb1ec24a354c679af03f27159aa586df502e648fdec1e445020f27513
-size 152233941
+oid sha256:bbb4bca3debb8a5f9aeff78b7c6b44adeea5abbcfc494e85d160590208ef338a
+size 151183829
diff --git a/checkpoints/model_weights_000118882304.pt b/checkpoints/model_weights_000118882304.pt
index 3578bb3a9828e97260ec1148d2d52aadc21eb7dc..2a161a49cd473c4ae33aea2f1464077e057e87ea 100644
--- a/checkpoints/model_weights_000118882304.pt
+++ b/checkpoints/model_weights_000118882304.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4cbab3d452ea3a9409d030e5776c662fe950d0d79f8784f1817231d7fa083aea
-size 152233941
+oid sha256:dad525ac75a786b7c48559f73292ce6b0761326ecff0a00abf45fca0c7dbc4d1
+size 151183829
diff --git a/checkpoints/model_weights_000130777088.pt b/checkpoints/model_weights_000130777088.pt
index 5eaaa96134a158c178dc3d70dc72940e0bd3f369..03615802dd2ac727c5dd12106fc3a1291c357204 100644
--- a/checkpoints/model_weights_000130777088.pt
+++ b/checkpoints/model_weights_000130777088.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d487b622a2cc5a65570831502e60128c9cb56ae6fdc1e26d43e573a6ca812d54
-size 152233941
+oid sha256:5dead3853e6408b78c95fd11735281656f8700f9ba0ab45a635b8180b77ea6e8
+size 151183829
diff --git a/checkpoints/model_weights_000131072000.pt b/checkpoints/model_weights_000131072000.pt
index 45f0441040c212b11c8be2a6ea782ec5f767e719..149912c671981c7d942909ffda7e745655fcbe85 100644
--- a/checkpoints/model_weights_000131072000.pt
+++ b/checkpoints/model_weights_000131072000.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9a3679d177c2f603ed401d074578d3cf0551e07025298db66df3f7d42226a5d2
-size 152233941
+oid sha256:0bf1f29a7bcfdb09251db1c83484db6d7a1067fe264303822a59479c66b03158
+size 151183829
diff --git a/checkpoints/model_weights_000143851520.pt b/checkpoints/model_weights_000143851520.pt
index c3883e06ab684415faaa441be7be34e303646414..738ca0d08fe94a2fd6b26ed25e11025ffb0d6610 100644
--- a/checkpoints/model_weights_000143851520.pt
+++ b/checkpoints/model_weights_000143851520.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a011307b957b05c47eef2a28fbddb10dfd636099bbe9821d2cccadd68d6d943a
-size 152233941
+oid sha256:288de1c16d9a08421c961db7352d39b8df472768ac6c703449f7a34650bf7861
+size 151183829
diff --git a/checkpoints/model_weights_000147456000.pt b/checkpoints/model_weights_000147456000.pt
index 514a9e898b663415c8af25f04a0a5c88a645c8c6..295e31bafe27774ac3a568b02199d0c55781b737 100644
--- a/checkpoints/model_weights_000147456000.pt
+++ b/checkpoints/model_weights_000147456000.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9b6b44f0256a3e937b44d98ba5db43827beca6b2afff730e982c589845da5b18
-size 152233941
+oid sha256:a60dbba192bc2b853332448301d37f9472e007a2b1100deca8c73706a2293b82
+size 151183829
diff --git a/checkpoints/model_weights_000158269440.pt b/checkpoints/model_weights_000158269440.pt
index be0ee6790c96a2cca0e9132b3166a5f39a0979b8..235d45a02eae169a8664d155213d44345ef80a30 100644
--- a/checkpoints/model_weights_000158269440.pt
+++ b/checkpoints/model_weights_000158269440.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7847986f977020865a7d834725deb30709d5886a4f48f2925e21990b2dc258b0
-size 152233941
+oid sha256:fcf6983f31ea0a6e8dfccfbc518ae4429a50b81a78e96151c4226c6be2dbbc38
+size 151183829
diff --git a/checkpoints/model_weights_000163840000.pt b/checkpoints/model_weights_000163840000.pt
index 0d557680859f05efe278c880365fd3959ddf75c6..6826a5805895bf1b9d9434cf4f677df435c4220f 100644
--- a/checkpoints/model_weights_000163840000.pt
+++ b/checkpoints/model_weights_000163840000.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6d43f4d6cd56278c9b1fb28ca91bc4b22b098e445a3af3568b41d2d6c3b0dc90
-size 152233941
+oid sha256:4b1c766285545374a4628169306238921323892fabc16efe4cac2e030f580e0a
+size 151183829
diff --git a/checkpoints/model_weights_000174096384.pt b/checkpoints/model_weights_000174096384.pt
index f741f45f2edb2762dbda3c2e15d637324d538fee..7b846307b998656161deee1c626c0b5ad135e49c 100644
--- a/checkpoints/model_weights_000174096384.pt
+++ b/checkpoints/model_weights_000174096384.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:84d64815d19e8664b4c3ca40b7c0b491f2e83847ac4fd2bd94bbc3359226efb3
-size 152233941
+oid sha256:6385d8078c4dd13097d6991735ac2a9ff1095e6ef82e2040420907f1984d51d6
+size 151183829
diff --git a/checkpoints/model_weights_000180224000.pt b/checkpoints/model_weights_000180224000.pt
index 6cdb48395c43e01cee8e2eda42e0c85c301a74fc..6e522514bacaa7bcd105b267f9cb1ad8d99e5adb 100644
--- a/checkpoints/model_weights_000180224000.pt
+++ b/checkpoints/model_weights_000180224000.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:76a3ff11eaf9031309562053ae7288d5b0f74743d7a7499fd90ffa74001603ae
-size 152233941
+oid sha256:119c9199dfadb5302077fb1def57e503f39ae4ad5bf75a2a67c8b509259f7aa6
+size 151183829
diff --git a/checkpoints/model_weights_000191496192.pt b/checkpoints/model_weights_000191496192.pt
index a3d0946db81a3269f0c8e9b2b96cc9b9c574cd16..8799e849631ed66969ccd21e967ddf034e0139bb 100644
--- a/checkpoints/model_weights_000191496192.pt
+++ b/checkpoints/model_weights_000191496192.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8812454f499055a1f98385091e53f48070cc66e842541d10aabf65f6cfcd7222
-size 152233941
+oid sha256:a48916d983b57e57dde693012ee713e4ccc9fa8f1754f2707e0e384bfa15e65e
+size 151183829
diff --git a/checkpoints/model_weights_000196608000.pt b/checkpoints/model_weights_000196608000.pt
index ff59d7535bee5594443ea13785e8a0bc7577d762..1202b4a615d2ec11821b8952a6bab85812339055 100644
--- a/checkpoints/model_weights_000196608000.pt
+++ b/checkpoints/model_weights_000196608000.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cddba86502e17ef2a84aa5c7b449b3cff6ee0e15ff5aa85b80379dcecd396627
-size 152233941
+oid sha256:f25fa26c5946fd7d6d7cbcff540914b9b8b004a8651828cdb4b9cbf9135d1c4f
+size 151183829
diff --git a/checkpoints/model_weights_000196706304.pt b/checkpoints/model_weights_000196706304.pt
index 433d4aec0afab5d0567e48a9de8ca0c25419a7a1..7ff06a94fa8066aeccd44e42aea2b4564f0b3308 100644
--- a/checkpoints/model_weights_000196706304.pt
+++ b/checkpoints/model_weights_000196706304.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e5d5ca4396fe845458580e8d547be53c1e1f59bce7b5c3d0227b3f4305449f9d
-size 152233941
+oid sha256:38f0208bbe10a6a6672097497731f30b432c88667eedb126c6ac70bc64147270
+size 151183829
diff --git a/checkpoints/model_weights_000197361664.pt b/checkpoints/model_weights_000197361664.pt
index 4f727b30685a1937c350b899d2726ee317e64baa..ca026d08bd4a33b81cf2278bac31121c87068023 100644
--- a/checkpoints/model_weights_000197361664.pt
+++ b/checkpoints/model_weights_000197361664.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a794cd481aae45b3763d1f0d1ef884f639409209a25daeb80de6e094b999cb7f
-size 152233941
+oid sha256:c7babbfc6a6011af20eff8003ae7b8e86bbcc5648a7c3181bc06b490ba489301
+size 151183829
diff --git a/checkpoints/model_weights_000198017024.pt b/checkpoints/model_weights_000198017024.pt
index bedfa0050d95e80d8772a3a15676b8c00ae81bb5..48415b8b7cd5e365ddd3cbd62e166bba51ff2c71 100644
--- a/checkpoints/model_weights_000198017024.pt
+++ b/checkpoints/model_weights_000198017024.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:59e123b98191719e7dfcf691c195733edcca35e3946152f15c9073d0c2f40eb4
-size 152233941
+oid sha256:a8cd08fa6349fe63147e470b0a35ea18b734de5e0e3153515e1bbabc33e20923
+size 151183829
diff --git a/checkpoints/model_weights_000198672384.pt b/checkpoints/model_weights_000198672384.pt
index c3ae3d9ac7a52d0e0d81915053abfeb29c689206..c1a0c1580dc80a3d0deda0e08c4f1179e2d17e76 100644
--- a/checkpoints/model_weights_000198672384.pt
+++ b/checkpoints/model_weights_000198672384.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:828b4f79307f851edf450f962649cb453a5a47274c95f8d7918c5f80129fa19b
-size 152233941
+oid sha256:afd76f6a6e1b938d423b5f746d1fec3d8aeb9c2c762b8e239395da8d92be3e8d
+size 151183829
diff --git a/checkpoints/model_weights_000199327744.pt b/checkpoints/model_weights_000199327744.pt
index d18137a221d8b5bd6c9b13a98bba2c0ba541be86..595b1de70a4f6bea0a49a20ce9947b36c948222e 100644
--- a/checkpoints/model_weights_000199327744.pt
+++ b/checkpoints/model_weights_000199327744.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b2e53dfe4aaab1cfd0f9061a8bf68f9d4cab62315efd02eb216c9185dce67fa2
-size 152233941
+oid sha256:c95e9d70f55f1b3f54e0a844bb9dd5b9858d36f03cc47c2570b42bf1d721af62
+size 151183829
diff --git a/checkpoints/model_weights_000199950336.pt b/checkpoints/model_weights_000199950336.pt
index 3253e3e498e8acf07593ff1324499978d00551b2..81f824e80dba0868cdfb5d3160572a4907470be5 100644
--- a/checkpoints/model_weights_000199950336.pt
+++ b/checkpoints/model_weights_000199950336.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cc589a5603f394a6da0d93b34d49b5e45a908d01546454c928de9510ca9e2393
-size 152233941
+oid sha256:33b2fbc1cefb3a8426289be152254266ef313622ff306cbe21ac0d611b819ae6
+size 151183829
diff --git a/config.toml b/config.toml
index 713efbdc25c69e02626f4cbfab1de46ce1d330ee..e3fc3cc147c7dc2190c1cb6c26c85b7f029103b9 100644
--- a/config.toml
+++ b/config.toml
@@ -1,9 +1,9 @@
-model_name = "pile_llama_2H_2L"
+model_name = "pile_llama_H1_L2"
 n_layers = 2
 d_model = 512
 d_mlp = 2048
 d_head = 64
-n_heads = 2
+n_heads = 1
 attn_only = false
 layer_norm_eps = 1e-05
 init_range = 0.02
diff --git a/latest_checkpoint.pt b/latest_checkpoint.pt
index 12c1f37f5205a55bcaa313370202c732b4033408..21a8ed1a06d4824fa6cce070d074ebc21d4d7f1b 100644
--- a/latest_checkpoint.pt
+++ b/latest_checkpoint.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cbf4a3701ff768aefc7f1a06a33f399b56def10807110a8e2f7cb5139f6b89ec
-size 152233463
+oid sha256:a841316540aacca701236202ac8a9224d21e8c06feead0486a3b0b90ace4bc2a
+size 151183351
diff --git a/latest_metadata.json b/latest_metadata.json
index a72f6c7b92a2dcca81dc2215710976a9adf66bb8..c16e16df2be11e6f5e6f06b5472547126445306f 100644
--- a/latest_metadata.json
+++ b/latest_metadata.json
@@ -1 +1 @@
-{"step": 6102, "tokens_seen": 199950336, "config": {"model_name": "pile_llama_2H_2L", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 2, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.394872257897474}
\ No newline at end of file
+{"step": 6102, "tokens_seen": 199950336, "config": {"model_name": "pile_llama_H1_L2", "n_layers": 2, "d_model": 512, "d_mlp": 2048, "d_head": 64, "n_heads": 1, "attn_only": false, "layer_norm_eps": 1e-05, "init_range": 0.02, "n_ctx": 1024, "d_vocab": 32000, "dataset_name": "eoinf/pile_llama", "tokenizer_name": "", "seed": 10, "data_seed": 10, "device": "cuda", "use_bfloat16_matmul": false, "batch_size_per_device": 32, "n_devices": 1, "batches_per_step": 1, "max_tokens": 200000000, "lr_hidden": 0.001, "lr_vector": 0.0005, "lr_schedule": "constant_with_warmup", "warmup_tokens": 30000000, "weight_decay": 0.05, "grad_norm_clip": 1.0, "train_loss_moving_average_beta": 0.99, "log_interval": 25, "save_checkpoints": true, "checkpoint_interval": 500, "checkpoint_interval_ratio": 1.1, "save_log_checkpoints": true, "use_wandb": true, "batch_size": 32, "tokens_per_step": 32768, "warmup_steps": 915, "max_steps": 6103}, "train_loss_ewma": 3.4578413544944855}
\ No newline at end of file
diff --git a/latest_optimizer.pt b/latest_optimizer.pt
index 8da33203ecaad91a8827342a257f37a8ca03d174..dc3fd016ddf8f7527f07be956e5891a24bcd5bf5 100644
--- a/latest_optimizer.pt
+++ b/latest_optimizer.pt
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:45f29280cd5d337e2f07c52c96e8f786c41131009adff2f4450a76255889657c
-size 304472851
+oid sha256:b412b0850dd6e899a32a7a77dcf97e5a851bbfa14d20715330f35dcd4888314d
+size 302372627
diff --git a/wandb/debug-internal.log b/wandb/debug-internal.log
index 8a8b762c57499dae83b531207444b547fc77a22f..f48362b4d868d528b4262bd90ae26bc82d13006a 100644
--- a/wandb/debug-internal.log
+++ b/wandb/debug-internal.log
@@ -1,12 +1,13 @@
-{"time":"2026-02-26T13:56:02.310641183Z","level":"INFO","msg":"stream: starting","core version":"0.21.4"}
-{"time":"2026-02-26T13:56:02.505385385Z","level":"INFO","msg":"stream: created new stream","id":"696nxyfr"}
-{"time":"2026-02-26T13:56:02.505474889Z","level":"INFO","msg":"stream: started","id":"696nxyfr"}
-{"time":"2026-02-26T13:56:02.505547204Z","level":"INFO","msg":"writer: started","stream_id":"696nxyfr"}
-{"time":"2026-02-26T13:56:02.505558638Z","level":"INFO","msg":"handler: started","stream_id":"696nxyfr"}
-{"time":"2026-02-26T13:56:02.5055985Z","level":"INFO","msg":"sender: started","stream_id":"696nxyfr"}
-{"time":"2026-02-26T14:39:41.476737152Z","level":"INFO","msg":"handler: operation stats","stats":{"operations":[{"desc":"updating run metadata","runtime_seconds":0.003364817}],"total_operations":1}}
-{"time":"2026-02-26T14:39:41.82022831Z","level":"INFO","msg":"fileTransfer: Close: file transfer manager closed"}
-{"time":"2026-02-26T14:39:42.35788107Z","level":"INFO","msg":"stream: closing","id":"696nxyfr"}
-{"time":"2026-02-26T14:39:42.357919856Z","level":"INFO","msg":"handler: closed","stream_id":"696nxyfr"}
-{"time":"2026-02-26T14:39:42.357981703Z","level":"INFO","msg":"sender: closed","stream_id":"696nxyfr"}
-{"time":"2026-02-26T14:39:42.357990898Z","level":"INFO","msg":"stream: closed","id":"696nxyfr"}
+{"time":"2026-02-26T15:30:26.517889161Z","level":"INFO","msg":"stream: starting","core version":"0.21.4"}
+{"time":"2026-02-26T15:30:31.757267268Z","level":"INFO","msg":"stream: created new stream","id":"trcpjlfd"}
+{"time":"2026-02-26T15:30:31.757342751Z","level":"INFO","msg":"stream: started","id":"trcpjlfd"}
+{"time":"2026-02-26T15:30:31.757375145Z","level":"INFO","msg":"handler: started","stream_id":"trcpjlfd"}
+{"time":"2026-02-26T15:30:31.757422001Z","level":"INFO","msg":"sender: started","stream_id":"trcpjlfd"}
+{"time":"2026-02-26T15:30:31.757462295Z","level":"INFO","msg":"writer: started","stream_id":"trcpjlfd"}
+{"time":"2026-02-26T16:08:47.579624032Z","level":"INFO","msg":"api: retrying HTTP error","status":502,"url":"https://api.wandb.ai/files/tzach/toy-transformer-replication/trcpjlfd/file_stream","body":"\n<html><head>\n<meta http-equiv=\"content-type\" content=\"text/html;charset=utf-8\">\n<title>502 Server Error</title>\n</head>\n<body text=#000000 bgcolor=#ffffff>\n<h1>Error: Server Error</h1>\n<h2>The server encountered a temporary error and could not complete your request.<p>Please try again in 30 seconds.</h2>\n<h2></h2>\n</body></html>\n"}
+{"time":"2026-02-26T16:12:40.720756645Z","level":"INFO","msg":"handler: operation stats","stats":{"operations":[{"desc":"updating run metadata","runtime_seconds":0.00031624}],"total_operations":1}}
+{"time":"2026-02-26T16:12:41.198683788Z","level":"INFO","msg":"fileTransfer: Close: file transfer manager closed"}
+{"time":"2026-02-26T16:12:41.385614105Z","level":"INFO","msg":"stream: closing","id":"trcpjlfd"}
+{"time":"2026-02-26T16:12:41.385657054Z","level":"INFO","msg":"handler: closed","stream_id":"trcpjlfd"}
+{"time":"2026-02-26T16:12:41.385711626Z","level":"INFO","msg":"sender: closed","stream_id":"trcpjlfd"}
+{"time":"2026-02-26T16:12:41.385718584Z","level":"INFO","msg":"stream: closed","id":"trcpjlfd"}
diff --git a/wandb/debug.log b/wandb/debug.log
index d628c19744d0df790b8f77df43df800da59c34ce..3ee3411527a8785b9507f88e164f4432869d104c 100644
--- a/wandb/debug.log
+++ b/wandb/debug.log
@@ -1,26 +1,26 @@
-2026-02-26 13:56:02,097 INFO    MainThread:4744 [wandb_setup.py:_flush():81] Current SDK version is 0.21.4
-2026-02-26 13:56:02,097 INFO    MainThread:4744 [wandb_setup.py:_flush():81] Configure stats pid to 4744
-2026-02-26 13:56:02,097 INFO    MainThread:4744 [wandb_setup.py:_flush():81] Loading settings from /root/.config/wandb/settings
-2026-02-26 13:56:02,097 INFO    MainThread:4744 [wandb_setup.py:_flush():81] Loading settings from /notebooks/toy_models/model_training/model/wandb/settings
-2026-02-26 13:56:02,097 INFO    MainThread:4744 [wandb_setup.py:_flush():81] Loading settings from environment variables
-2026-02-26 13:56:02,097 INFO    MainThread:4744 [wandb_init.py:setup_run_log_directory():686] Logging user logs to /notebooks/toy_models/model_training/model/wandb/run-20260226_135602-696nxyfr/logs/debug.log
-2026-02-26 13:56:02,097 INFO    MainThread:4744 [wandb_init.py:setup_run_log_directory():687] Logging internal logs to /notebooks/toy_models/model_training/model/wandb/run-20260226_135602-696nxyfr/logs/debug-internal.log
-2026-02-26 13:56:02,097 INFO    MainThread:4744 [wandb_init.py:init():813] calling init triggers
-2026-02-26 13:56:02,098 INFO    MainThread:4744 [wandb_init.py:init():818] wandb.init called with sweep_config: {}
-config: {'model_name': 'pile_llama_2H_2L', 'n_layers': 2, 'd_model': 512, 'd_mlp': 2048, 'd_head': 64, 'n_heads': 2, 'attn_only': False, 'layer_norm_eps': 1e-05, 'init_range': 0.02, 'n_ctx': 1024, 'd_vocab': 32000, 'dataset_name': 'eoinf/pile_llama', 'tokenizer_name': '', 'seed': 10, 'data_seed': 10, 'device': 'cuda', 'use_bfloat16_matmul': False, 'batch_size_per_device': 32, 'n_devices': 1, 'batches_per_step': 1, 'max_tokens': 200000000, 'lr_hidden': 0.001, 'lr_vector': 0.0005, 'lr_schedule': 'constant_with_warmup', 'warmup_tokens': 30000000, 'weight_decay': 0.05, 'grad_norm_clip': 1.0, 'train_loss_moving_average_beta': 0.99, 'log_interval': 25, 'save_checkpoints': True, 'checkpoint_interval': 500, 'checkpoint_interval_ratio': 1.1, 'save_log_checkpoints': True, 'use_wandb': True, 'batch_size': 32, 'tokens_per_step': 32768, 'warmup_steps': 915, 'max_steps': 6103, '_wandb': {}}
-2026-02-26 13:56:02,098 INFO    MainThread:4744 [wandb_init.py:init():854] starting backend
-2026-02-26 13:56:02,302 INFO    MainThread:4744 [wandb_init.py:init():857] sending inform_init request
-2026-02-26 13:56:02,306 INFO    MainThread:4744 [wandb_init.py:init():865] backend started and connected
-2026-02-26 13:56:02,308 INFO    MainThread:4744 [wandb_init.py:init():936] updated telemetry
-2026-02-26 13:56:02,312 INFO    MainThread:4744 [wandb_init.py:init():960] communicating run to backend with 90.0 second timeout
-2026-02-26 13:56:02,787 INFO    MainThread:4744 [wandb_init.py:init():1011] starting run threads in backend
-2026-02-26 13:56:02,902 INFO    MainThread:4744 [wandb_run.py:_console_start():2506] atexit reg
-2026-02-26 13:56:02,902 INFO    MainThread:4744 [wandb_run.py:_redirect():2354] redirect: wrap_raw
-2026-02-26 13:56:02,902 INFO    MainThread:4744 [wandb_run.py:_redirect():2423] Wrapping output streams.
-2026-02-26 13:56:02,902 INFO    MainThread:4744 [wandb_run.py:_redirect():2446] Redirects installed.
-2026-02-26 13:56:02,905 INFO    MainThread:4744 [wandb_init.py:init():1049] run started, returning control to user process
-2026-02-26 14:39:41,470 INFO    MainThread:4744 [wandb_run.py:_finish():2272] finishing run tzach/toy-transformer-replication/696nxyfr
-2026-02-26 14:39:41,472 INFO    MainThread:4744 [wandb_run.py:_atexit_cleanup():2471] got exitcode: 0
-2026-02-26 14:39:41,473 INFO    MainThread:4744 [wandb_run.py:_restore():2453] restore
-2026-02-26 14:39:41,473 INFO    MainThread:4744 [wandb_run.py:_restore():2459] restore done
-2026-02-26 14:39:42,357 INFO    MainThread:4744 [wandb_run.py:_footer_sync_info():3867] logging synced files
+2026-02-26 15:30:26,235 INFO    MainThread:5904 [wandb_setup.py:_flush():81] Current SDK version is 0.21.4
+2026-02-26 15:30:26,236 INFO    MainThread:5904 [wandb_setup.py:_flush():81] Configure stats pid to 5904
+2026-02-26 15:30:26,236 INFO    MainThread:5904 [wandb_setup.py:_flush():81] Loading settings from /root/.config/wandb/settings
+2026-02-26 15:30:26,236 INFO    MainThread:5904 [wandb_setup.py:_flush():81] Loading settings from /notebooks/toy_models/model_training/model/wandb/settings
+2026-02-26 15:30:26,236 INFO    MainThread:5904 [wandb_setup.py:_flush():81] Loading settings from environment variables
+2026-02-26 15:30:26,236 INFO    MainThread:5904 [wandb_init.py:setup_run_log_directory():686] Logging user logs to /notebooks/toy_models/model_training/model/wandb/run-20260226_153026-trcpjlfd/logs/debug.log
+2026-02-26 15:30:26,236 INFO    MainThread:5904 [wandb_init.py:setup_run_log_directory():687] Logging internal logs to /notebooks/toy_models/model_training/model/wandb/run-20260226_153026-trcpjlfd/logs/debug-internal.log
+2026-02-26 15:30:26,236 INFO    MainThread:5904 [wandb_init.py:init():813] calling init triggers
+2026-02-26 15:30:26,236 INFO    MainThread:5904 [wandb_init.py:init():818] wandb.init called with sweep_config: {}
+config: {'model_name': 'pile_llama_H1_L2', 'n_layers': 2, 'd_model': 512, 'd_mlp': 2048, 'd_head': 64, 'n_heads': 1, 'attn_only': False, 'layer_norm_eps': 1e-05, 'init_range': 0.02, 'n_ctx': 1024, 'd_vocab': 32000, 'dataset_name': 'eoinf/pile_llama', 'tokenizer_name': '', 'seed': 10, 'data_seed': 10, 'device': 'cuda', 'use_bfloat16_matmul': False, 'batch_size_per_device': 32, 'n_devices': 1, 'batches_per_step': 1, 'max_tokens': 200000000, 'lr_hidden': 0.001, 'lr_vector': 0.0005, 'lr_schedule': 'constant_with_warmup', 'warmup_tokens': 30000000, 'weight_decay': 0.05, 'grad_norm_clip': 1.0, 'train_loss_moving_average_beta': 0.99, 'log_interval': 25, 'save_checkpoints': True, 'checkpoint_interval': 500, 'checkpoint_interval_ratio': 1.1, 'save_log_checkpoints': True, 'use_wandb': True, 'batch_size': 32, 'tokens_per_step': 32768, 'warmup_steps': 915, 'max_steps': 6103, '_wandb': {}}
+2026-02-26 15:30:26,236 INFO    MainThread:5904 [wandb_init.py:init():854] starting backend
+2026-02-26 15:30:26,504 INFO    MainThread:5904 [wandb_init.py:init():857] sending inform_init request
+2026-02-26 15:30:26,514 INFO    MainThread:5904 [wandb_init.py:init():865] backend started and connected
+2026-02-26 15:30:26,515 INFO    MainThread:5904 [wandb_init.py:init():936] updated telemetry
+2026-02-26 15:30:26,520 INFO    MainThread:5904 [wandb_init.py:init():960] communicating run to backend with 90.0 second timeout
+2026-02-26 15:30:32,186 INFO    MainThread:5904 [wandb_init.py:init():1011] starting run threads in backend
+2026-02-26 15:30:33,011 INFO    MainThread:5904 [wandb_run.py:_console_start():2506] atexit reg
+2026-02-26 15:30:33,011 INFO    MainThread:5904 [wandb_run.py:_redirect():2354] redirect: wrap_raw
+2026-02-26 15:30:33,011 INFO    MainThread:5904 [wandb_run.py:_redirect():2423] Wrapping output streams.
+2026-02-26 15:30:33,011 INFO    MainThread:5904 [wandb_run.py:_redirect():2446] Redirects installed.
+2026-02-26 15:30:33,026 INFO    MainThread:5904 [wandb_init.py:init():1049] run started, returning control to user process
+2026-02-26 16:12:40,715 INFO    MainThread:5904 [wandb_run.py:_finish():2272] finishing run tzach/toy-transformer-replication/trcpjlfd
+2026-02-26 16:12:40,719 INFO    MainThread:5904 [wandb_run.py:_atexit_cleanup():2471] got exitcode: 0
+2026-02-26 16:12:40,719 INFO    MainThread:5904 [wandb_run.py:_restore():2453] restore
+2026-02-26 16:12:40,719 INFO    MainThread:5904 [wandb_run.py:_restore():2459] restore done
+2026-02-26 16:12:41,384 INFO    MainThread:5904 [wandb_run.py:_footer_sync_info():3867] logging synced files
diff --git a/wandb/run-20260226_153026-trcpjlfd/files/config.yaml b/wandb/run-20260226_153026-trcpjlfd/files/config.yaml
new file mode 100644
index 0000000000000000000000000000000000000000..7d21fb675869d2249a2b12d4d5f02578cae2b4e4
--- /dev/null
+++ b/wandb/run-20260226_153026-trcpjlfd/files/config.yaml
@@ -0,0 +1,140 @@
+_wandb:
+    value:
+        cli_version: 0.21.4
+        e:
+            1azdhqr66s9127dpz7denwtgp0pkf39b:
+                cpu_count: 8
+                cpu_count_logical: 8
+                cudaVersion: "12.4"
+                disk:
+                    /:
+                        total: "262240792576"
+                        used: "179147603968"
+                email: tzfof8@gmail.com
+                executable: /notebooks/toy_models/.toy_models_env/bin/python
+                git:
+                    commit: 4a8df1a57a456227acca0b4948ad48c9f03a929a
+                    remote: https://github.com/jgroh3/toy_models.git
+                gpu: NVIDIA RTX A6000
+                gpu_count: 1
+                gpu_nvidia:
+                    - architecture: Ampere
+                      cudaCores: 10752
+                      memoryTotal: "51527024640"
+                      name: NVIDIA RTX A6000
+                      uuid: GPU-fa82911d-b891-1e23-11c8-f31149299e8c
+                host: nb5tce73r3
+                memory:
+                    total: "47332843520"
+                os: Linux-5.19.0-45-generic-x86_64-with-glibc2.35
+                program: <python with no main file>
+                python: CPython 3.11.7
+                root: /notebooks/toy_models/model_training/model
+                startedAt: "2026-02-26T15:30:26.231994Z"
+                writerId: 1azdhqr66s9127dpz7denwtgp0pkf39b
+        m: []
+        python_version: 3.11.7
+        t:
+            "1":
+                - 1
+                - 5
+                - 11
+                - 49
+                - 51
+                - 53
+                - 71
+            "2":
+                - 1
+                - 5
+                - 11
+                - 49
+                - 51
+                - 53
+                - 71
+            "3":
+                - 2
+                - 13
+                - 15
+                - 16
+                - 61
+            "4": 3.11.7
+            "5": 0.21.4
+            "6": 4.56.1
+            "12": 0.21.4
+            "13": linux-x86_64
+attn_only:
+    value: false
+batch_size:
+    value: 32
+batch_size_per_device:
+    value: 32
+batches_per_step:
+    value: 1
+checkpoint_interval:
+    value: 500
+checkpoint_interval_ratio:
+    value: 1.1
+d_head:
+    value: 64
+d_mlp:
+    value: 2048
+d_model:
+    value: 512
+d_vocab:
+    value: 32000
+data_seed:
+    value: 10
+dataset_name:
+    value: eoinf/pile_llama
+device:
+    value: cuda
+grad_norm_clip:
+    value: 1
+init_range:
+    value: 0.02
+layer_norm_eps:
+    value: 1e-05
+log_interval:
+    value: 25
+lr_hidden:
+    value: 0.001
+lr_schedule:
+    value: constant_with_warmup
+lr_vector:
+    value: 0.0005
+max_steps:
+    value: 6103
+max_tokens:
+    value: 200000000
+model_name:
+    value: pile_llama_H1_L2
+n_ctx:
+    value: 1024
+n_devices:
+    value: 1
+n_heads:
+    value: 1
+n_layers:
+    value: 2
+save_checkpoints:
+    value: true
+save_log_checkpoints:
+    value: true
+seed:
+    value: 10
+tokenizer_name:
+    value: ""
+tokens_per_step:
+    value: 32768
+train_loss_moving_average_beta:
+    value: 0.99
+use_bfloat16_matmul:
+    value: false
+use_wandb:
+    value: true
+warmup_steps:
+    value: 915
+warmup_tokens:
+    value: 30000000
+weight_decay:
+    value: 0.05
diff --git a/wandb/run-20260226_153026-trcpjlfd/files/output.log b/wandb/run-20260226_153026-trcpjlfd/files/output.log
new file mode 100644
index 0000000000000000000000000000000000000000..ba9a7701da8cb5fba000729ddabf0b6a12f5b266
--- /dev/null
+++ b/wandb/run-20260226_153026-trcpjlfd/files/output.log
@@ -0,0 +1,252 @@
+Training on cuda
+Model: 2L, 512d, 1h
+Max steps: 6,103, Max tokens: 200,000,000
+Warmup steps: 915, Warmup tokens: 30,000,000
+Batch size per device: 32
+Context length: 1024
+Learning rates - Hidden: 0.001, Vector: 0.0005
+                                                                                                                                                                        
+Step 25 | Tokens: 819,200 | Train Loss EWMA: 10.4383 | Learning Rate: 0.000027 | Progress: 0.00410
+Step 50 | Tokens: 1,638,400 | Train Loss EWMA: 10.2429 | Learning Rate: 0.000055 | Progress: 0.00819
+Step 75 | Tokens: 2,457,600 | Train Loss EWMA: 9.9316 | Learning Rate: 0.000082 | Progress: 0.01229
+Step 100 | Tokens: 3,276,800 | Train Loss EWMA: 9.5223 | Learning Rate: 0.000109 | Progress: 0.01638
+Step 125 | Tokens: 4,096,000 | Train Loss EWMA: 9.0598 | Learning Rate: 0.000137 | Progress: 0.02048
+Step 150 | Tokens: 4,915,200 | Train Loss EWMA: 8.5866 | Learning Rate: 0.000164 | Progress: 0.02458
+Step 175 | Tokens: 5,734,400 | Train Loss EWMA: 8.1639 | Learning Rate: 0.000191 | Progress: 0.02867
+Step 200 | Tokens: 6,553,600 | Train Loss EWMA: 7.7886 | Learning Rate: 0.000219 | Progress: 0.03277
+Step 225 | Tokens: 7,372,800 | Train Loss EWMA: 7.4553 | Learning Rate: 0.000246 | Progress: 0.03686
+Step 250 | Tokens: 8,192,000 | Train Loss EWMA: 7.1716 | Learning Rate: 0.000273 | Progress: 0.04096
+Step 275 | Tokens: 9,011,200 | Train Loss EWMA: 6.9149 | Learning Rate: 0.000301 | Progress: 0.04506
+Step 300 | Tokens: 9,830,400 | Train Loss EWMA: 6.6887 | Learning Rate: 0.000328 | Progress: 0.04915
+Step 325 | Tokens: 10,649,600 | Train Loss EWMA: 6.4947 | Learning Rate: 0.000355 | Progress: 0.05325
+Step 350 | Tokens: 11,468,800 | Train Loss EWMA: 6.3275 | Learning Rate: 0.000383 | Progress: 0.05734
+Step 375 | Tokens: 12,288,000 | Train Loss EWMA: 6.1798 | Learning Rate: 0.000410 | Progress: 0.06144
+Step 400 | Tokens: 13,107,200 | Train Loss EWMA: 6.0495 | Learning Rate: 0.000437 | Progress: 0.06554
+Step 425 | Tokens: 13,926,400 | Train Loss EWMA: 5.9294 | Learning Rate: 0.000464 | Progress: 0.06963
+Step 450 | Tokens: 14,745,600 | Train Loss EWMA: 5.8285 | Learning Rate: 0.000492 | Progress: 0.07373
+Step 475 | Tokens: 15,564,800 | Train Loss EWMA: 5.7325 | Learning Rate: 0.000519 | Progress: 0.07782
+Step 500 | Tokens: 16,384,000 | Train Loss EWMA: 5.6596 | Learning Rate: 0.000546 | Progress: 0.08192
+Step 525 | Tokens: 17,203,200 | Train Loss EWMA: 5.5807 | Learning Rate: 0.000574 | Progress: 0.08602
+Step 550 | Tokens: 18,022,400 | Train Loss EWMA: 5.5053 | Learning Rate: 0.000601 | Progress: 0.09011
+Step 575 | Tokens: 18,841,600 | Train Loss EWMA: 5.4348 | Learning Rate: 0.000628 | Progress: 0.09421
+Step 600 | Tokens: 19,660,800 | Train Loss EWMA: 5.3779 | Learning Rate: 0.000656 | Progress: 0.09830
+Step 625 | Tokens: 20,480,000 | Train Loss EWMA: 5.3231 | Learning Rate: 0.000683 | Progress: 0.10240
+Step 650 | Tokens: 21,299,200 | Train Loss EWMA: 5.2867 | Learning Rate: 0.000710 | Progress: 0.10650
+Step 675 | Tokens: 22,118,400 | Train Loss EWMA: 5.2415 | Learning Rate: 0.000738 | Progress: 0.11059
+Step 700 | Tokens: 22,937,600 | Train Loss EWMA: 5.2032 | Learning Rate: 0.000765 | Progress: 0.11469
+Step 725 | Tokens: 23,756,800 | Train Loss EWMA: 5.1613 | Learning Rate: 0.000792 | Progress: 0.11878
+Step 750 | Tokens: 24,576,000 | Train Loss EWMA: 5.1146 | Learning Rate: 0.000820 | Progress: 0.12288
+Step 775 | Tokens: 25,395,200 | Train Loss EWMA: 5.0817 | Learning Rate: 0.000847 | Progress: 0.12698
+Step 800 | Tokens: 26,214,400 | Train Loss EWMA: 5.0518 | Learning Rate: 0.000874 | Progress: 0.13107
+Step 825 | Tokens: 27,033,600 | Train Loss EWMA: 5.0149 | Learning Rate: 0.000902 | Progress: 0.13517
+Step 850 | Tokens: 27,852,800 | Train Loss EWMA: 4.9885 | Learning Rate: 0.000929 | Progress: 0.13926
+Step 875 | Tokens: 28,672,000 | Train Loss EWMA: 4.9555 | Learning Rate: 0.000956 | Progress: 0.14336
+Step 900 | Tokens: 29,491,200 | Train Loss EWMA: 4.9269 | Learning Rate: 0.000984 | Progress: 0.14746
+Step 925 | Tokens: 30,310,400 | Train Loss EWMA: 4.8936 | Learning Rate: 0.001000 | Progress: 0.15155
+Step 950 | Tokens: 31,129,600 | Train Loss EWMA: 4.8686 | Learning Rate: 0.001000 | Progress: 0.15565
+Step 975 | Tokens: 31,948,800 | Train Loss EWMA: 4.8451 | Learning Rate: 0.001000 | Progress: 0.15974
+Step 1,000 | Tokens: 32,768,000 | Train Loss EWMA: 4.8127 | Learning Rate: 0.001000 | Progress: 0.16384
+Step 1,025 | Tokens: 33,587,200 | Train Loss EWMA: 4.7797 | Learning Rate: 0.001000 | Progress: 0.16794
+Step 1,050 | Tokens: 34,406,400 | Train Loss EWMA: 4.7549 | Learning Rate: 0.001000 | Progress: 0.17203
+Step 1,075 | Tokens: 35,225,600 | Train Loss EWMA: 4.7282 | Learning Rate: 0.001000 | Progress: 0.17613
+Step 1,100 | Tokens: 36,044,800 | Train Loss EWMA: 4.6992 | Learning Rate: 0.001000 | Progress: 0.18022
+Step 1,125 | Tokens: 36,864,000 | Train Loss EWMA: 4.6699 | Learning Rate: 0.001000 | Progress: 0.18432
+Step 1,150 | Tokens: 37,683,200 | Train Loss EWMA: 4.6475 | Learning Rate: 0.001000 | Progress: 0.18842
+Step 1,175 | Tokens: 38,502,400 | Train Loss EWMA: 4.6185 | Learning Rate: 0.001000 | Progress: 0.19251
+Step 1,200 | Tokens: 39,321,600 | Train Loss EWMA: 4.6003 | Learning Rate: 0.001000 | Progress: 0.19661
+Step 1,225 | Tokens: 40,140,800 | Train Loss EWMA: 4.5781 | Learning Rate: 0.001000 | Progress: 0.20070
+Step 1,250 | Tokens: 40,960,000 | Train Loss EWMA: 4.5497 | Learning Rate: 0.001000 | Progress: 0.20480
+Step 1,275 | Tokens: 41,779,200 | Train Loss EWMA: 4.5281 | Learning Rate: 0.001000 | Progress: 0.20890
+Step 1,300 | Tokens: 42,598,400 | Train Loss EWMA: 4.5072 | Learning Rate: 0.001000 | Progress: 0.21299
+Step 1,325 | Tokens: 43,417,600 | Train Loss EWMA: 4.4946 | Learning Rate: 0.001000 | Progress: 0.21709
+Step 1,350 | Tokens: 44,236,800 | Train Loss EWMA: 4.4669 | Learning Rate: 0.001000 | Progress: 0.22118
+Step 1,375 | Tokens: 45,056,000 | Train Loss EWMA: 4.4457 | Learning Rate: 0.001000 | Progress: 0.22528
+Step 1,400 | Tokens: 45,875,200 | Train Loss EWMA: 4.4302 | Learning Rate: 0.001000 | Progress: 0.22938
+Step 1,425 | Tokens: 46,694,400 | Train Loss EWMA: 4.4176 | Learning Rate: 0.001000 | Progress: 0.23347
+Step 1,450 | Tokens: 47,513,600 | Train Loss EWMA: 4.4050 | Learning Rate: 0.001000 | Progress: 0.23757
+Step 1,475 | Tokens: 48,332,800 | Train Loss EWMA: 4.3857 | Learning Rate: 0.001000 | Progress: 0.24166
+Step 1,500 | Tokens: 49,152,000 | Train Loss EWMA: 4.3701 | Learning Rate: 0.001000 | Progress: 0.24576
+Step 1,525 | Tokens: 49,971,200 | Train Loss EWMA: 4.3657 | Learning Rate: 0.001000 | Progress: 0.24986
+Step 1,550 | Tokens: 50,790,400 | Train Loss EWMA: 4.3477 | Learning Rate: 0.001000 | Progress: 0.25395
+Step 1,575 | Tokens: 51,609,600 | Train Loss EWMA: 4.3299 | Learning Rate: 0.001000 | Progress: 0.25805
+Step 1,600 | Tokens: 52,428,800 | Train Loss EWMA: 4.3152 | Learning Rate: 0.001000 | Progress: 0.26214
+Step 1,625 | Tokens: 53,248,000 | Train Loss EWMA: 4.2998 | Learning Rate: 0.001000 | Progress: 0.26624
+Step 1,650 | Tokens: 54,067,200 | Train Loss EWMA: 4.2893 | Learning Rate: 0.001000 | Progress: 0.27034
+Step 1,675 | Tokens: 54,886,400 | Train Loss EWMA: 4.2703 | Learning Rate: 0.001000 | Progress: 0.27443
+Step 1,700 | Tokens: 55,705,600 | Train Loss EWMA: 4.2501 | Learning Rate: 0.001000 | Progress: 0.27853
+Step 1,725 | Tokens: 56,524,800 | Train Loss EWMA: 4.2444 | Learning Rate: 0.001000 | Progress: 0.28262
+Step 1,750 | Tokens: 57,344,000 | Train Loss EWMA: 4.2282 | Learning Rate: 0.001000 | Progress: 0.28672
+Step 1,775 | Tokens: 58,163,200 | Train Loss EWMA: 4.2138 | Learning Rate: 0.001000 | Progress: 0.29082
+Step 1,800 | Tokens: 58,982,400 | Train Loss EWMA: 4.2157 | Learning Rate: 0.001000 | Progress: 0.29491
+Step 1,825 | Tokens: 59,801,600 | Train Loss EWMA: 4.1997 | Learning Rate: 0.001000 | Progress: 0.29901
+Step 1,850 | Tokens: 60,620,800 | Train Loss EWMA: 4.1875 | Learning Rate: 0.001000 | Progress: 0.30310
+Step 1,875 | Tokens: 61,440,000 | Train Loss EWMA: 4.1814 | Learning Rate: 0.001000 | Progress: 0.30720
+Step 1,900 | Tokens: 62,259,200 | Train Loss EWMA: 4.1778 | Learning Rate: 0.001000 | Progress: 0.31130
+Step 1,925 | Tokens: 63,078,400 | Train Loss EWMA: 4.1804 | Learning Rate: 0.001000 | Progress: 0.31539
+Step 1,950 | Tokens: 63,897,600 | Train Loss EWMA: 4.1737 | Learning Rate: 0.001000 | Progress: 0.31949
+Step 1,975 | Tokens: 64,716,800 | Train Loss EWMA: 4.1674 | Learning Rate: 0.001000 | Progress: 0.32358
+Step 2,000 | Tokens: 65,536,000 | Train Loss EWMA: 4.1585 | Learning Rate: 0.001000 | Progress: 0.32768
+Step 2,025 | Tokens: 66,355,200 | Train Loss EWMA: 4.1441 | Learning Rate: 0.001000 | Progress: 0.33178
+Step 2,050 | Tokens: 67,174,400 | Train Loss EWMA: 4.1352 | Learning Rate: 0.001000 | Progress: 0.33587
+Step 2,075 | Tokens: 67,993,600 | Train Loss EWMA: 4.1128 | Learning Rate: 0.001000 | Progress: 0.33997
+Step 2,100 | Tokens: 68,812,800 | Train Loss EWMA: 4.0948 | Learning Rate: 0.001000 | Progress: 0.34406
+Step 2,125 | Tokens: 69,632,000 | Train Loss EWMA: 4.0853 | Learning Rate: 0.001000 | Progress: 0.34816
+Step 2,150 | Tokens: 70,451,200 | Train Loss EWMA: 4.0792 | Learning Rate: 0.001000 | Progress: 0.35226
+Step 2,175 | Tokens: 71,270,400 | Train Loss EWMA: 4.0670 | Learning Rate: 0.001000 | Progress: 0.35635
+Step 2,200 | Tokens: 72,089,600 | Train Loss EWMA: 4.0583 | Learning Rate: 0.001000 | Progress: 0.36045
+Step 2,225 | Tokens: 72,908,800 | Train Loss EWMA: 4.0396 | Learning Rate: 0.001000 | Progress: 0.36454
+Step 2,250 | Tokens: 73,728,000 | Train Loss EWMA: 4.0260 | Learning Rate: 0.001000 | Progress: 0.36864
+Step 2,275 | Tokens: 74,547,200 | Train Loss EWMA: 4.0148 | Learning Rate: 0.001000 | Progress: 0.37274
+Step 2,300 | Tokens: 75,366,400 | Train Loss EWMA: 4.0007 | Learning Rate: 0.001000 | Progress: 0.37683
+Step 2,325 | Tokens: 76,185,600 | Train Loss EWMA: 3.9911 | Learning Rate: 0.001000 | Progress: 0.38093
+Step 2,350 | Tokens: 77,004,800 | Train Loss EWMA: 3.9832 | Learning Rate: 0.001000 | Progress: 0.38502
+Step 2,375 | Tokens: 77,824,000 | Train Loss EWMA: 3.9697 | Learning Rate: 0.001000 | Progress: 0.38912
+Step 2,400 | Tokens: 78,643,200 | Train Loss EWMA: 3.9549 | Learning Rate: 0.001000 | Progress: 0.39322
+Step 2,425 | Tokens: 79,462,400 | Train Loss EWMA: 3.9455 | Learning Rate: 0.001000 | Progress: 0.39731
+Step 2,450 | Tokens: 80,281,600 | Train Loss EWMA: 3.9433 | Learning Rate: 0.001000 | Progress: 0.40141
+Step 2,475 | Tokens: 81,100,800 | Train Loss EWMA: 3.9275 | Learning Rate: 0.001000 | Progress: 0.40550
+Step 2,500 | Tokens: 81,920,000 | Train Loss EWMA: 3.9131 | Learning Rate: 0.001000 | Progress: 0.40960
+Step 2,525 | Tokens: 82,739,200 | Train Loss EWMA: 3.9026 | Learning Rate: 0.001000 | Progress: 0.41370
+Step 2,550 | Tokens: 83,558,400 | Train Loss EWMA: 3.8930 | Learning Rate: 0.001000 | Progress: 0.41779
+Step 2,575 | Tokens: 84,377,600 | Train Loss EWMA: 3.8754 | Learning Rate: 0.001000 | Progress: 0.42189
+Step 2,600 | Tokens: 85,196,800 | Train Loss EWMA: 3.8631 | Learning Rate: 0.001000 | Progress: 0.42598
+Step 2,625 | Tokens: 86,016,000 | Train Loss EWMA: 3.8611 | Learning Rate: 0.001000 | Progress: 0.43008
+Step 2,650 | Tokens: 86,835,200 | Train Loss EWMA: 3.8540 | Learning Rate: 0.001000 | Progress: 0.43418
+Step 2,675 | Tokens: 87,654,400 | Train Loss EWMA: 3.8498 | Learning Rate: 0.001000 | Progress: 0.43827
+Step 2,700 | Tokens: 88,473,600 | Train Loss EWMA: 3.8328 | Learning Rate: 0.001000 | Progress: 0.44237
+Step 2,725 | Tokens: 89,292,800 | Train Loss EWMA: 3.8301 | Learning Rate: 0.001000 | Progress: 0.44646
+Step 2,750 | Tokens: 90,112,000 | Train Loss EWMA: 3.8189 | Learning Rate: 0.001000 | Progress: 0.45056
+Step 2,775 | Tokens: 90,931,200 | Train Loss EWMA: 3.8121 | Learning Rate: 0.001000 | Progress: 0.45466
+Step 2,800 | Tokens: 91,750,400 | Train Loss EWMA: 3.8034 | Learning Rate: 0.001000 | Progress: 0.45875
+Step 2,825 | Tokens: 92,569,600 | Train Loss EWMA: 3.8023 | Learning Rate: 0.001000 | Progress: 0.46285
+Step 2,850 | Tokens: 93,388,800 | Train Loss EWMA: 3.7958 | Learning Rate: 0.001000 | Progress: 0.46694
+Step 2,875 | Tokens: 94,208,000 | Train Loss EWMA: 3.7821 | Learning Rate: 0.001000 | Progress: 0.47104
+Step 2,900 | Tokens: 95,027,200 | Train Loss EWMA: 3.7773 | Learning Rate: 0.001000 | Progress: 0.47514
+Step 2,925 | Tokens: 95,846,400 | Train Loss EWMA: 3.7691 | Learning Rate: 0.001000 | Progress: 0.47923
+Step 2,950 | Tokens: 96,665,600 | Train Loss EWMA: 3.7654 | Learning Rate: 0.001000 | Progress: 0.48333
+Step 2,975 | Tokens: 97,484,800 | Train Loss EWMA: 3.7648 | Learning Rate: 0.001000 | Progress: 0.48742
+Step 3,000 | Tokens: 98,304,000 | Train Loss EWMA: 3.7636 | Learning Rate: 0.001000 | Progress: 0.49152
+Step 3,025 | Tokens: 99,123,200 | Train Loss EWMA: 3.7604 | Learning Rate: 0.001000 | Progress: 0.49562
+Step 3,050 | Tokens: 99,942,400 | Train Loss EWMA: 3.7489 | Learning Rate: 0.001000 | Progress: 0.49971
+Step 3,075 | Tokens: 100,761,600 | Train Loss EWMA: 3.7442 | Learning Rate: 0.001000 | Progress: 0.50381
+Step 3,100 | Tokens: 101,580,800 | Train Loss EWMA: 3.7414 | Learning Rate: 0.001000 | Progress: 0.50790
+Step 3,125 | Tokens: 102,400,000 | Train Loss EWMA: 3.7406 | Learning Rate: 0.001000 | Progress: 0.51200
+Step 3,150 | Tokens: 103,219,200 | Train Loss EWMA: 3.7337 | Learning Rate: 0.001000 | Progress: 0.51610
+Step 3,175 | Tokens: 104,038,400 | Train Loss EWMA: 3.7304 | Learning Rate: 0.001000 | Progress: 0.52019
+Step 3,200 | Tokens: 104,857,600 | Train Loss EWMA: 3.7145 | Learning Rate: 0.001000 | Progress: 0.52429
+Step 3,225 | Tokens: 105,676,800 | Train Loss EWMA: 3.7042 | Learning Rate: 0.001000 | Progress: 0.52838
+Step 3,250 | Tokens: 106,496,000 | Train Loss EWMA: 3.7032 | Learning Rate: 0.001000 | Progress: 0.53248
+Step 3,275 | Tokens: 107,315,200 | Train Loss EWMA: 3.7061 | Learning Rate: 0.001000 | Progress: 0.53658
+Step 3,300 | Tokens: 108,134,400 | Train Loss EWMA: 3.6925 | Learning Rate: 0.001000 | Progress: 0.54067
+Step 3,325 | Tokens: 108,953,600 | Train Loss EWMA: 3.6901 | Learning Rate: 0.001000 | Progress: 0.54477
+Step 3,350 | Tokens: 109,772,800 | Train Loss EWMA: 3.6902 | Learning Rate: 0.001000 | Progress: 0.54886
+Step 3,375 | Tokens: 110,592,000 | Train Loss EWMA: 3.6730 | Learning Rate: 0.001000 | Progress: 0.55296
+Step 3,400 | Tokens: 111,411,200 | Train Loss EWMA: 3.6701 | Learning Rate: 0.001000 | Progress: 0.55706
+Step 3,425 | Tokens: 112,230,400 | Train Loss EWMA: 3.6629 | Learning Rate: 0.001000 | Progress: 0.56115
+Step 3,450 | Tokens: 113,049,600 | Train Loss EWMA: 3.6608 | Learning Rate: 0.001000 | Progress: 0.56525
+Step 3,475 | Tokens: 113,868,800 | Train Loss EWMA: 3.6641 | Learning Rate: 0.001000 | Progress: 0.56934
+Step 3,500 | Tokens: 114,688,000 | Train Loss EWMA: 3.6516 | Learning Rate: 0.001000 | Progress: 0.57344
+Step 3,525 | Tokens: 115,507,200 | Train Loss EWMA: 3.6523 | Learning Rate: 0.001000 | Progress: 0.57754
+Step 3,550 | Tokens: 116,326,400 | Train Loss EWMA: 3.6459 | Learning Rate: 0.001000 | Progress: 0.58163
+Step 3,575 | Tokens: 117,145,600 | Train Loss EWMA: 3.6575 | Learning Rate: 0.001000 | Progress: 0.58573
+Step 3,600 | Tokens: 117,964,800 | Train Loss EWMA: 3.6512 | Learning Rate: 0.001000 | Progress: 0.58982
+Step 3,625 | Tokens: 118,784,000 | Train Loss EWMA: 3.6548 | Learning Rate: 0.001000 | Progress: 0.59392
+Step 3,650 | Tokens: 119,603,200 | Train Loss EWMA: 3.6530 | Learning Rate: 0.001000 | Progress: 0.59802
+Step 3,675 | Tokens: 120,422,400 | Train Loss EWMA: 3.6534 | Learning Rate: 0.001000 | Progress: 0.60211
+Step 3,700 | Tokens: 121,241,600 | Train Loss EWMA: 3.6486 | Learning Rate: 0.001000 | Progress: 0.60621
+Step 3,725 | Tokens: 122,060,800 | Train Loss EWMA: 3.6454 | Learning Rate: 0.001000 | Progress: 0.61030
+Step 3,750 | Tokens: 122,880,000 | Train Loss EWMA: 3.6377 | Learning Rate: 0.001000 | Progress: 0.61440
+Step 3,775 | Tokens: 123,699,200 | Train Loss EWMA: 3.6291 | Learning Rate: 0.001000 | Progress: 0.61850
+Step 3,800 | Tokens: 124,518,400 | Train Loss EWMA: 3.6201 | Learning Rate: 0.001000 | Progress: 0.62259
+Step 3,825 | Tokens: 125,337,600 | Train Loss EWMA: 3.6207 | Learning Rate: 0.001000 | Progress: 0.62669
+Step 3,850 | Tokens: 126,156,800 | Train Loss EWMA: 3.6246 | Learning Rate: 0.001000 | Progress: 0.63078
+Step 3,875 | Tokens: 126,976,000 | Train Loss EWMA: 3.6158 | Learning Rate: 0.001000 | Progress: 0.63488
+Step 3,900 | Tokens: 127,795,200 | Train Loss EWMA: 3.6221 | Learning Rate: 0.001000 | Progress: 0.63898
+Step 3,925 | Tokens: 128,614,400 | Train Loss EWMA: 3.6185 | Learning Rate: 0.001000 | Progress: 0.64307
+Step 3,950 | Tokens: 129,433,600 | Train Loss EWMA: 3.6185 | Learning Rate: 0.001000 | Progress: 0.64717
+Step 3,975 | Tokens: 130,252,800 | Train Loss EWMA: 3.6076 | Learning Rate: 0.001000 | Progress: 0.65126
+Step 4,000 | Tokens: 131,072,000 | Train Loss EWMA: 3.6144 | Learning Rate: 0.001000 | Progress: 0.65536
+Step 4,025 | Tokens: 131,891,200 | Train Loss EWMA: 3.6086 | Learning Rate: 0.001000 | Progress: 0.65946
+Step 4,050 | Tokens: 132,710,400 | Train Loss EWMA: 3.5920 | Learning Rate: 0.001000 | Progress: 0.66355
+Step 4,075 | Tokens: 133,529,600 | Train Loss EWMA: 3.5874 | Learning Rate: 0.001000 | Progress: 0.66765
+Step 4,100 | Tokens: 134,348,800 | Train Loss EWMA: 3.5843 | Learning Rate: 0.001000 | Progress: 0.67174
+Step 4,125 | Tokens: 135,168,000 | Train Loss EWMA: 3.5883 | Learning Rate: 0.001000 | Progress: 0.67584
+Step 4,150 | Tokens: 135,987,200 | Train Loss EWMA: 3.5913 | Learning Rate: 0.001000 | Progress: 0.67994
+Step 4,175 | Tokens: 136,806,400 | Train Loss EWMA: 3.5856 | Learning Rate: 0.001000 | Progress: 0.68403
+Step 4,200 | Tokens: 137,625,600 | Train Loss EWMA: 3.5958 | Learning Rate: 0.001000 | Progress: 0.68813
+Step 4,225 | Tokens: 138,444,800 | Train Loss EWMA: 3.5863 | Learning Rate: 0.001000 | Progress: 0.69222
+Step 4,250 | Tokens: 139,264,000 | Train Loss EWMA: 3.5877 | Learning Rate: 0.001000 | Progress: 0.69632
+Step 4,275 | Tokens: 140,083,200 | Train Loss EWMA: 3.5873 | Learning Rate: 0.001000 | Progress: 0.70042
+Step 4,300 | Tokens: 140,902,400 | Train Loss EWMA: 3.5794 | Learning Rate: 0.001000 | Progress: 0.70451
+Step 4,325 | Tokens: 141,721,600 | Train Loss EWMA: 3.5790 | Learning Rate: 0.001000 | Progress: 0.70861
+Step 4,350 | Tokens: 142,540,800 | Train Loss EWMA: 3.5759 | Learning Rate: 0.001000 | Progress: 0.71270
+Step 4,375 | Tokens: 143,360,000 | Train Loss EWMA: 3.5752 | Learning Rate: 0.001000 | Progress: 0.71680
+Step 4,400 | Tokens: 144,179,200 | Train Loss EWMA: 3.5796 | Learning Rate: 0.001000 | Progress: 0.72090
+Step 4,425 | Tokens: 144,998,400 | Train Loss EWMA: 3.5857 | Learning Rate: 0.001000 | Progress: 0.72499
+Step 4,450 | Tokens: 145,817,600 | Train Loss EWMA: 3.5831 | Learning Rate: 0.001000 | Progress: 0.72909
+Step 4,475 | Tokens: 146,636,800 | Train Loss EWMA: 3.5820 | Learning Rate: 0.001000 | Progress: 0.73318
+Step 4,500 | Tokens: 147,456,000 | Train Loss EWMA: 3.5651 | Learning Rate: 0.001000 | Progress: 0.73728
+Step 4,525 | Tokens: 148,275,200 | Train Loss EWMA: 3.5492 | Learning Rate: 0.001000 | Progress: 0.74138
+Step 4,550 | Tokens: 149,094,400 | Train Loss EWMA: 3.5509 | Learning Rate: 0.001000 | Progress: 0.74547
+Step 4,575 | Tokens: 149,913,600 | Train Loss EWMA: 3.5472 | Learning Rate: 0.001000 | Progress: 0.74957
+Step 4,600 | Tokens: 150,732,800 | Train Loss EWMA: 3.5541 | Learning Rate: 0.001000 | Progress: 0.75366
+Step 4,625 | Tokens: 151,552,000 | Train Loss EWMA: 3.5488 | Learning Rate: 0.001000 | Progress: 0.75776
+Step 4,650 | Tokens: 152,371,200 | Train Loss EWMA: 3.5395 | Learning Rate: 0.001000 | Progress: 0.76186
+Step 4,675 | Tokens: 153,190,400 | Train Loss EWMA: 3.5324 | Learning Rate: 0.001000 | Progress: 0.76595
+Step 4,700 | Tokens: 154,009,600 | Train Loss EWMA: 3.5397 | Learning Rate: 0.001000 | Progress: 0.77005
+Step 4,725 | Tokens: 154,828,800 | Train Loss EWMA: 3.5443 | Learning Rate: 0.001000 | Progress: 0.77414
+Step 4,750 | Tokens: 155,648,000 | Train Loss EWMA: 3.5435 | Learning Rate: 0.001000 | Progress: 0.77824
+Step 4,775 | Tokens: 156,467,200 | Train Loss EWMA: 3.5440 | Learning Rate: 0.001000 | Progress: 0.78234
+Step 4,800 | Tokens: 157,286,400 | Train Loss EWMA: 3.5456 | Learning Rate: 0.001000 | Progress: 0.78643
+Step 4,825 | Tokens: 158,105,600 | Train Loss EWMA: 3.5423 | Learning Rate: 0.001000 | Progress: 0.79053
+Step 4,850 | Tokens: 158,924,800 | Train Loss EWMA: 3.5414 | Learning Rate: 0.001000 | Progress: 0.79462
+Step 4,875 | Tokens: 159,744,000 | Train Loss EWMA: 3.5409 | Learning Rate: 0.001000 | Progress: 0.79872
+Step 4,900 | Tokens: 160,563,200 | Train Loss EWMA: 3.5417 | Learning Rate: 0.001000 | Progress: 0.80282
+Step 4,925 | Tokens: 161,382,400 | Train Loss EWMA: 3.5280 | Learning Rate: 0.001000 | Progress: 0.80691
+Step 4,950 | Tokens: 162,201,600 | Train Loss EWMA: 3.5311 | Learning Rate: 0.001000 | Progress: 0.81101
+Step 4,975 | Tokens: 163,020,800 | Train Loss EWMA: 3.5333 | Learning Rate: 0.001000 | Progress: 0.81510
+Step 5,000 | Tokens: 163,840,000 | Train Loss EWMA: 3.5258 | Learning Rate: 0.001000 | Progress: 0.81920
+Step 5,025 | Tokens: 164,659,200 | Train Loss EWMA: 3.5313 | Learning Rate: 0.001000 | Progress: 0.82330
+Step 5,050 | Tokens: 165,478,400 | Train Loss EWMA: 3.5273 | Learning Rate: 0.001000 | Progress: 0.82739
+Step 5,075 | Tokens: 166,297,600 | Train Loss EWMA: 3.5280 | Learning Rate: 0.001000 | Progress: 0.83149
+Step 5,100 | Tokens: 167,116,800 | Train Loss EWMA: 3.5240 | Learning Rate: 0.001000 | Progress: 0.83558
+Step 5,125 | Tokens: 167,936,000 | Train Loss EWMA: 3.5173 | Learning Rate: 0.001000 | Progress: 0.83968
+Step 5,150 | Tokens: 168,755,200 | Train Loss EWMA: 3.5182 | Learning Rate: 0.001000 | Progress: 0.84378
+Step 5,175 | Tokens: 169,574,400 | Train Loss EWMA: 3.5185 | Learning Rate: 0.001000 | Progress: 0.84787
+Step 5,200 | Tokens: 170,393,600 | Train Loss EWMA: 3.5219 | Learning Rate: 0.001000 | Progress: 0.85197
+Step 5,225 | Tokens: 171,212,800 | Train Loss EWMA: 3.5207 | Learning Rate: 0.001000 | Progress: 0.85606
+Step 5,250 | Tokens: 172,032,000 | Train Loss EWMA: 3.5230 | Learning Rate: 0.001000 | Progress: 0.86016
+Step 5,275 | Tokens: 172,851,200 | Train Loss EWMA: 3.5204 | Learning Rate: 0.001000 | Progress: 0.86426
+Step 5,300 | Tokens: 173,670,400 | Train Loss EWMA: 3.5177 | Learning Rate: 0.001000 | Progress: 0.86835
+Step 5,325 | Tokens: 174,489,600 | Train Loss EWMA: 3.5198 | Learning Rate: 0.001000 | Progress: 0.87245
+Step 5,350 | Tokens: 175,308,800 | Train Loss EWMA: 3.5160 | Learning Rate: 0.001000 | Progress: 0.87654
+Step 5,375 | Tokens: 176,128,000 | Train Loss EWMA: 3.5169 | Learning Rate: 0.001000 | Progress: 0.88064
+Step 5,400 | Tokens: 176,947,200 | Train Loss EWMA: 3.5049 | Learning Rate: 0.001000 | Progress: 0.88474
+Step 5,425 | Tokens: 177,766,400 | Train Loss EWMA: 3.5038 | Learning Rate: 0.001000 | Progress: 0.88883
+Step 5,450 | Tokens: 178,585,600 | Train Loss EWMA: 3.4981 | Learning Rate: 0.001000 | Progress: 0.89293
+Step 5,475 | Tokens: 179,404,800 | Train Loss EWMA: 3.4988 | Learning Rate: 0.001000 | Progress: 0.89702
+Step 5,500 | Tokens: 180,224,000 | Train Loss EWMA: 3.4966 | Learning Rate: 0.001000 | Progress: 0.90112
+Step 5,525 | Tokens: 181,043,200 | Train Loss EWMA: 3.4951 | Learning Rate: 0.001000 | Progress: 0.90522
+Step 5,550 | Tokens: 181,862,400 | Train Loss EWMA: 3.4919 | Learning Rate: 0.001000 | Progress: 0.90931
+Step 5,575 | Tokens: 182,681,600 | Train Loss EWMA: 3.4948 | Learning Rate: 0.001000 | Progress: 0.91341
+Step 5,600 | Tokens: 183,500,800 | Train Loss EWMA: 3.4979 | Learning Rate: 0.001000 | Progress: 0.91750
+Step 5,625 | Tokens: 184,320,000 | Train Loss EWMA: 3.4915 | Learning Rate: 0.001000 | Progress: 0.92160
+Step 5,650 | Tokens: 185,139,200 | Train Loss EWMA: 3.4872 | Learning Rate: 0.001000 | Progress: 0.92570
+Step 5,675 | Tokens: 185,958,400 | Train Loss EWMA: 3.4840 | Learning Rate: 0.001000 | Progress: 0.92979
+Step 5,700 | Tokens: 186,777,600 | Train Loss EWMA: 3.4824 | Learning Rate: 0.001000 | Progress: 0.93389
+Step 5,725 | Tokens: 187,596,800 | Train Loss EWMA: 3.4793 | Learning Rate: 0.001000 | Progress: 0.93798
+Step 5,750 | Tokens: 188,416,000 | Train Loss EWMA: 3.4846 | Learning Rate: 0.001000 | Progress: 0.94208
+Step 5,775 | Tokens: 189,235,200 | Train Loss EWMA: 3.4755 | Learning Rate: 0.001000 | Progress: 0.94618
+Step 5,800 | Tokens: 190,054,400 | Train Loss EWMA: 3.4706 | Learning Rate: 0.001000 | Progress: 0.95027
+Step 5,825 | Tokens: 190,873,600 | Train Loss EWMA: 3.4721 | Learning Rate: 0.001000 | Progress: 0.95437
+Step 5,850 | Tokens: 191,692,800 | Train Loss EWMA: 3.4751 | Learning Rate: 0.001000 | Progress: 0.95846
+Step 5,875 | Tokens: 192,512,000 | Train Loss EWMA: 3.4711 | Learning Rate: 0.001000 | Progress: 0.96256
+Step 5,900 | Tokens: 193,331,200 | Train Loss EWMA: 3.4679 | Learning Rate: 0.001000 | Progress: 0.96666
+Step 5,925 | Tokens: 194,150,400 | Train Loss EWMA: 3.4707 | Learning Rate: 0.001000 | Progress: 0.97075
+Step 5,950 | Tokens: 194,969,600 | Train Loss EWMA: 3.4744 | Learning Rate: 0.001000 | Progress: 0.97485
+Step 5,975 | Tokens: 195,788,800 | Train Loss EWMA: 3.4741 | Learning Rate: 0.001000 | Progress: 0.97894
+Step 6,000 | Tokens: 196,608,000 | Train Loss EWMA: 3.4668 | Learning Rate: 0.001000 | Progress: 0.98304
+Step 6,025 | Tokens: 197,427,200 | Train Loss EWMA: 3.4712 | Learning Rate: 0.001000 | Progress: 0.98714
+Step 6,050 | Tokens: 198,246,400 | Train Loss EWMA: 3.4643 | Learning Rate: 0.001000 | Progress: 0.99123
+Step 6,075 | Tokens: 199,065,600 | Train Loss EWMA: 3.4631 | Learning Rate: 0.001000 | Progress: 0.99533
+Step 6,100 | Tokens: 199,884,800 | Train Loss EWMA: 3.4588 | Learning Rate: 0.001000 | Progress: 0.99942
diff --git a/wandb/run-20260226_153026-trcpjlfd/files/requirements.txt b/wandb/run-20260226_153026-trcpjlfd/files/requirements.txt
new file mode 100644
index 0000000000000000000000000000000000000000..c32285d10ba18c2e783ff2ead305d5976caef668
--- /dev/null
+++ b/wandb/run-20260226_153026-trcpjlfd/files/requirements.txt
@@ -0,0 +1,222 @@
+fsspec==2025.3.0
+PyYAML==6.0.2
+certifi==2025.8.3
+comm==0.2.3
+widgetsnbextension==4.0.14
+Jinja2==3.1.6
+rich==14.1.0
+circuitsvis==1.43.3
+hf-xet==1.1.9
+param==2.2.1
+httpcore==1.0.9
+nvidia-cuda-cupti-cu12==12.8.90
+nvidia-cuda-nvrtc-cu12==12.8.93
+asttokens==3.0.0
+filelock==3.19.1
+types-python-dateutil==2.9.0.20250822
+cycler==0.12.1
+stack-data==0.6.3
+jupyter_server==2.17.0
+aiosignal==1.4.0
+xyzservices==2025.4.0
+lark==1.2.2
+ptyprocess==0.7.0
+xxhash==3.5.0
+mpmath==1.3.0
+seaborn==0.13.2
+wadler_lindig==0.1.7
+nbformat==5.10.4
+panel==1.8.0
+accelerate==1.10.1
+plotly==6.3.0
+narwhals==2.4.0
+huggingface-hub==0.34.4
+sentencepiece==0.2.1
+torchvision==0.23.0
+ipython==9.5.0
+tqdm==4.67.1
+contourpy==1.3.3
+nvidia-nvtx-cu12==12.8.90
+nvidia-cuda-runtime-cu12==12.8.90
+yarl==1.20.1
+charset-normalizer==3.4.3
+jupyter-events==0.12.0
+nbclient==0.10.2
+numpy==1.26.4
+decorator==5.2.1
+threadpoolctl==3.6.0
+networkx==3.5
+smmap==5.0.2
+nbconvert==7.16.6
+pytz==2025.2
+aiohappyeyeballs==2.6.1
+requests==2.32.5
+tinycss2==1.4.0
+defusedxml==0.7.1
+matplotlib-inline==0.1.7
+rpds-py==0.27.1
+wandb==0.21.4
+jedi==0.19.2
+pathspec==0.12.1
+transformer-lens==2.16.1
+sympy==1.14.0
+jupyterlab_pygments==0.3.0
+overrides==7.7.0
+notebook_shim==0.2.4
+jupyter==1.1.1
+protobuf==6.32.1
+better-abc==0.0.3
+jsonpointer==3.0.0
+terminado==0.18.1
+cfgv==3.4.0
+rfc3987-syntax==1.1.0
+annotated-types==0.7.0
+pyarrow==21.0.0
+webencodings==0.5.1
+wcwidth==0.2.13
+jupyterlab_server==2.27.3
+argon2-cffi-bindings==25.1.0
+nvidia-nvjitlink-cu12==12.8.93
+jaxtyping==0.3.2
+Pygments==2.19.2
+torch==2.8.0
+rfc3339-validator==0.1.4
+urllib3==2.5.0
+jupyterlab_widgets==3.0.15
+ipykernel==6.30.1
+nvidia-cudnn-cu12==9.10.2.21
+beautifulsoup4==4.13.5
+babel==2.17.0
+pure_eval==0.2.3
+pyparsing==3.2.3
+nvidia-cublas-cu12==12.8.4.1
+regex==2025.9.1
+pycparser==2.23
+soupsieve==2.8
+pytest-cov==7.0.0
+sniffio==1.3.1
+mypy==1.18.1
+notebook==7.4.5
+packaging==25.0
+h11==0.16.0
+psutil==7.0.0
+pexpect==4.9.0
+zstandard==0.25.0
+gitdb==4.0.12
+rfc3986-validator==0.1.1
+pyzmq==27.1.0
+jupyterlab==4.4.7
+toy_models==0.1.0
+torchaudio==2.8.0
+cffi==2.0.0
+mypy_extensions==1.1.0
+attrs==25.3.0
+statsmodels==0.14.6
+transformers==4.56.1
+jupyter_core==5.8.1
+bleach==6.2.0
+fqdn==1.5.1
+async-lru==2.0.5
+nvidia-nccl-cu12==2.27.3
+GitPython==3.1.45
+referencing==0.36.2
+click==8.2.1
+prometheus_client==0.22.1
+bokeh==3.8.0
+httpx==0.28.1
+setuptools==80.9.0
+argon2-cffi==25.1.0
+patsy==1.0.2
+multidict==6.6.4
+pyviz_comms==3.0.6
+arrow==1.3.0
+scikit-learn==1.8.0
+beartype==0.14.1
+ipywidgets==8.1.7
+pydantic_core==2.33.2
+markdown-it-py==4.0.0
+pandas==2.3.2
+virtualenv==20.34.0
+python-dotenv==1.1.1
+isoduration==20.11.0
+python-dateutil==2.9.0.post0
+nodeenv==1.9.1
+nvidia-curand-cu12==10.3.9.90
+webcolors==24.11.1
+MarkupSafe==3.0.2
+nvidia-cusolver-cu12==11.7.3.90
+Send2Trash==1.8.3
+coverage==7.10.6
+jupyter_server_terminals==0.5.3
+debugpy==1.8.16
+json5==0.12.1
+linkify-it-py==2.0.3
+importlib_metadata==8.7.0
+nvidia-cufft-cu12==11.3.3.83
+distlib==0.4.0
+typing-inspection==0.4.1
+identify==2.6.14
+nvidia-cufile-cu12==1.13.1.3
+scipy==1.17.0
+mdurl==0.1.2
+websocket-client==1.8.0
+jsonschema==4.25.1
+python-json-logger==3.3.0
+typing_extensions==4.15.0
+tokenizers==0.22.0
+ipympl==0.9.7
+einops==0.8.1
+jupyter_client==8.6.3
+ipython_pygments_lexers==1.1.1
+h5py==3.14.0
+tabulate==0.9.0
+propcache==0.3.2
+ruff==0.13.0
+tornado==6.5.2
+typeguard==4.4.4
+tomlkit==0.13.2
+pluggy==1.6.0
+pydantic==2.11.7
+zipp==3.23.0
+fancy-einsum==0.0.3
+fastjsonschema==2.21.2
+datasets==4.0.0
+fonttools==4.59.2
+executing==2.2.1
+pillow==11.3.0
+uc-micro-py==1.0.3
+Markdown==3.9
+pre_commit==4.3.0
+aiohttp==3.12.15
+mistune==3.1.4
+tzdata==2025.2
+parso==0.8.5
+triton==3.4.0
+kiwisolver==1.4.9
+idna==3.10
+multiprocess==0.70.16
+dill==0.3.8
+jupyter-lsp==2.3.0
+platformdirs==4.4.0
+sentry-sdk==2.37.1
+prompt_toolkit==3.0.52
+jsonschema-specifications==2025.9.1
+pytest==8.4.2
+mdit-py-plugins==0.5.0
+transformers-stream-generator==0.0.5
+nvidia-cusparselt-cu12==0.7.1
+joblib==1.5.3
+pandocfilters==1.5.1
+jupyter-console==6.6.3
+anyio==4.10.0
+six==1.17.0
+holoviews==1.21.0
+matplotlib==3.10.6
+colorcet==3.1.0
+uri-template==1.3.0
+nest-asyncio==1.6.0
+nvidia-cusparse-cu12==12.5.8.93
+iniconfig==2.1.0
+traitlets==5.14.3
+safetensors==0.6.2
+frozenlist==1.7.0
diff --git a/wandb/run-20260226_153026-trcpjlfd/files/wandb-metadata.json b/wandb/run-20260226_153026-trcpjlfd/files/wandb-metadata.json
new file mode 100644
index 0000000000000000000000000000000000000000..6ecab91fbcb9748fea3266b4dfe039d63618c3fb
--- /dev/null
+++ b/wandb/run-20260226_153026-trcpjlfd/files/wandb-metadata.json
@@ -0,0 +1,38 @@
+{
+  "os":  "Linux-5.19.0-45-generic-x86_64-with-glibc2.35",
+  "python":  "CPython 3.11.7",
+  "startedAt":  "2026-02-26T15:30:26.231994Z",
+  "program":  "<python with no main file>",
+  "git":  {
+    "remote":  "https://github.com/jgroh3/toy_models.git",
+    "commit":  "4a8df1a57a456227acca0b4948ad48c9f03a929a"
+  },
+  "email":  "tzfof8@gmail.com",
+  "root":  "/notebooks/toy_models/model_training/model",
+  "host":  "nb5tce73r3",
+  "executable":  "/notebooks/toy_models/.toy_models_env/bin/python",
+  "cpu_count":  8,
+  "cpu_count_logical":  8,
+  "gpu":  "NVIDIA RTX A6000",
+  "gpu_count":  1,
+  "disk":  {
+    "/":  {
+      "total":  "262240792576",
+      "used":  "179147603968"
+    }
+  },
+  "memory":  {
+    "total":  "47332843520"
+  },
+  "gpu_nvidia":  [
+    {
+      "name":  "NVIDIA RTX A6000",
+      "memoryTotal":  "51527024640",
+      "cudaCores":  10752,
+      "architecture":  "Ampere",
+      "uuid":  "GPU-fa82911d-b891-1e23-11c8-f31149299e8c"
+    }
+  ],
+  "cudaVersion":  "12.4",
+  "writerId":  "1azdhqr66s9127dpz7denwtgp0pkf39b"
+}
\ No newline at end of file
diff --git a/wandb/run-20260226_153026-trcpjlfd/files/wandb-summary.json b/wandb/run-20260226_153026-trcpjlfd/files/wandb-summary.json
new file mode 100644
index 0000000000000000000000000000000000000000..8716b2a4657246198ed8c099d8937c16a4b236d0
--- /dev/null
+++ b/wandb/run-20260226_153026-trcpjlfd/files/wandb-summary.json
@@ -0,0 +1 @@
+{"train_loss":3.26702618598938,"_step":6100,"tokens_seen":199884800,"_wandb":{"runtime":2528},"progress":0.999424,"_timestamp":1.7721223586858675e+09,"_runtime":2528.533134255,"step":6100,"train_loss_ewma":3.4588358451443244,"learning_rate":0.001,"tokens_per_second":32768}
\ No newline at end of file
diff --git a/wandb/run-20260226_153026-trcpjlfd/logs/debug-core.log b/wandb/run-20260226_153026-trcpjlfd/logs/debug-core.log
new file mode 100644
index 0000000000000000000000000000000000000000..bc679e555db6d67a2e77f75e2098addf59b23e0e
--- /dev/null
+++ b/wandb/run-20260226_153026-trcpjlfd/logs/debug-core.log
@@ -0,0 +1,16 @@
+{"time":"2026-02-26T15:30:26.496865995Z","level":"INFO","msg":"main: starting server","port-filename":"/tmp/tmpofpp25bp/port-5904.txt","pid":5904,"log-level":0,"disable-analytics":false,"shutdown-on-parent-exit":false,"enable-dcgm-profiling":false}
+{"time":"2026-02-26T15:30:26.497562893Z","level":"INFO","msg":"server: will exit if parent process dies","ppid":5904}
+{"time":"2026-02-26T15:30:26.497525799Z","level":"INFO","msg":"server: accepting connections","addr":{"Name":"/tmp/wandb-5904-5939-519161693/socket","Net":"unix"}}
+{"time":"2026-02-26T15:30:26.50460356Z","level":"INFO","msg":"connection: ManageConnectionData: new connection created","id":"1(@)"}
+{"time":"2026-02-26T15:30:26.517677183Z","level":"INFO","msg":"handleInformInit: received","streamId":"trcpjlfd","id":"1(@)"}
+{"time":"2026-02-26T15:30:31.757349232Z","level":"INFO","msg":"handleInformInit: stream started","streamId":"trcpjlfd","id":"1(@)"}
+{"time":"2026-02-26T16:12:41.385574928Z","level":"INFO","msg":"handleInformFinish: finish message received","streamId":"trcpjlfd","id":"1(@)"}
+{"time":"2026-02-26T16:12:41.389034918Z","level":"INFO","msg":"handleInformFinish: stream closed","streamId":"trcpjlfd","id":"1(@)"}
+{"time":"2026-02-26T16:12:41.389057073Z","level":"INFO","msg":"handleInformTeardown: server teardown initiated","id":"1(@)"}
+{"time":"2026-02-26T16:12:41.389071886Z","level":"INFO","msg":"handleInformTeardown: server shutdown complete","id":"1(@)"}
+{"time":"2026-02-26T16:12:41.389080569Z","level":"INFO","msg":"connection: closing","id":"1(@)"}
+{"time":"2026-02-26T16:12:41.389096275Z","level":"INFO","msg":"server is shutting down"}
+{"time":"2026-02-26T16:12:41.389113632Z","level":"INFO","msg":"connection: closed successfully","id":"1(@)"}
+{"time":"2026-02-26T16:12:41.38912204Z","level":"INFO","msg":"connection: ManageConnectionData: connection closed","id":"1(@)"}
+{"time":"2026-02-26T16:12:41.389213427Z","level":"INFO","msg":"server: listener closed","addr":{"Name":"/tmp/wandb-5904-5939-519161693/socket","Net":"unix"}}
+{"time":"2026-02-26T16:12:41.38924661Z","level":"INFO","msg":"server is closed"}
diff --git a/wandb/run-20260226_153026-trcpjlfd/logs/debug-internal.log b/wandb/run-20260226_153026-trcpjlfd/logs/debug-internal.log
new file mode 100644
index 0000000000000000000000000000000000000000..f48362b4d868d528b4262bd90ae26bc82d13006a
--- /dev/null
+++ b/wandb/run-20260226_153026-trcpjlfd/logs/debug-internal.log
@@ -0,0 +1,13 @@
+{"time":"2026-02-26T15:30:26.517889161Z","level":"INFO","msg":"stream: starting","core version":"0.21.4"}
+{"time":"2026-02-26T15:30:31.757267268Z","level":"INFO","msg":"stream: created new stream","id":"trcpjlfd"}
+{"time":"2026-02-26T15:30:31.757342751Z","level":"INFO","msg":"stream: started","id":"trcpjlfd"}
+{"time":"2026-02-26T15:30:31.757375145Z","level":"INFO","msg":"handler: started","stream_id":"trcpjlfd"}
+{"time":"2026-02-26T15:30:31.757422001Z","level":"INFO","msg":"sender: started","stream_id":"trcpjlfd"}
+{"time":"2026-02-26T15:30:31.757462295Z","level":"INFO","msg":"writer: started","stream_id":"trcpjlfd"}
+{"time":"2026-02-26T16:08:47.579624032Z","level":"INFO","msg":"api: retrying HTTP error","status":502,"url":"https://api.wandb.ai/files/tzach/toy-transformer-replication/trcpjlfd/file_stream","body":"\n<html><head>\n<meta http-equiv=\"content-type\" content=\"text/html;charset=utf-8\">\n<title>502 Server Error</title>\n</head>\n<body text=#000000 bgcolor=#ffffff>\n<h1>Error: Server Error</h1>\n<h2>The server encountered a temporary error and could not complete your request.<p>Please try again in 30 seconds.</h2>\n<h2></h2>\n</body></html>\n"}
+{"time":"2026-02-26T16:12:40.720756645Z","level":"INFO","msg":"handler: operation stats","stats":{"operations":[{"desc":"updating run metadata","runtime_seconds":0.00031624}],"total_operations":1}}
+{"time":"2026-02-26T16:12:41.198683788Z","level":"INFO","msg":"fileTransfer: Close: file transfer manager closed"}
+{"time":"2026-02-26T16:12:41.385614105Z","level":"INFO","msg":"stream: closing","id":"trcpjlfd"}
+{"time":"2026-02-26T16:12:41.385657054Z","level":"INFO","msg":"handler: closed","stream_id":"trcpjlfd"}
+{"time":"2026-02-26T16:12:41.385711626Z","level":"INFO","msg":"sender: closed","stream_id":"trcpjlfd"}
+{"time":"2026-02-26T16:12:41.385718584Z","level":"INFO","msg":"stream: closed","id":"trcpjlfd"}
diff --git a/wandb/run-20260226_153026-trcpjlfd/logs/debug.log b/wandb/run-20260226_153026-trcpjlfd/logs/debug.log
new file mode 100644
index 0000000000000000000000000000000000000000..3ee3411527a8785b9507f88e164f4432869d104c
--- /dev/null
+++ b/wandb/run-20260226_153026-trcpjlfd/logs/debug.log
@@ -0,0 +1,26 @@
+2026-02-26 15:30:26,235 INFO    MainThread:5904 [wandb_setup.py:_flush():81] Current SDK version is 0.21.4
+2026-02-26 15:30:26,236 INFO    MainThread:5904 [wandb_setup.py:_flush():81] Configure stats pid to 5904
+2026-02-26 15:30:26,236 INFO    MainThread:5904 [wandb_setup.py:_flush():81] Loading settings from /root/.config/wandb/settings
+2026-02-26 15:30:26,236 INFO    MainThread:5904 [wandb_setup.py:_flush():81] Loading settings from /notebooks/toy_models/model_training/model/wandb/settings
+2026-02-26 15:30:26,236 INFO    MainThread:5904 [wandb_setup.py:_flush():81] Loading settings from environment variables
+2026-02-26 15:30:26,236 INFO    MainThread:5904 [wandb_init.py:setup_run_log_directory():686] Logging user logs to /notebooks/toy_models/model_training/model/wandb/run-20260226_153026-trcpjlfd/logs/debug.log
+2026-02-26 15:30:26,236 INFO    MainThread:5904 [wandb_init.py:setup_run_log_directory():687] Logging internal logs to /notebooks/toy_models/model_training/model/wandb/run-20260226_153026-trcpjlfd/logs/debug-internal.log
+2026-02-26 15:30:26,236 INFO    MainThread:5904 [wandb_init.py:init():813] calling init triggers
+2026-02-26 15:30:26,236 INFO    MainThread:5904 [wandb_init.py:init():818] wandb.init called with sweep_config: {}
+config: {'model_name': 'pile_llama_H1_L2', 'n_layers': 2, 'd_model': 512, 'd_mlp': 2048, 'd_head': 64, 'n_heads': 1, 'attn_only': False, 'layer_norm_eps': 1e-05, 'init_range': 0.02, 'n_ctx': 1024, 'd_vocab': 32000, 'dataset_name': 'eoinf/pile_llama', 'tokenizer_name': '', 'seed': 10, 'data_seed': 10, 'device': 'cuda', 'use_bfloat16_matmul': False, 'batch_size_per_device': 32, 'n_devices': 1, 'batches_per_step': 1, 'max_tokens': 200000000, 'lr_hidden': 0.001, 'lr_vector': 0.0005, 'lr_schedule': 'constant_with_warmup', 'warmup_tokens': 30000000, 'weight_decay': 0.05, 'grad_norm_clip': 1.0, 'train_loss_moving_average_beta': 0.99, 'log_interval': 25, 'save_checkpoints': True, 'checkpoint_interval': 500, 'checkpoint_interval_ratio': 1.1, 'save_log_checkpoints': True, 'use_wandb': True, 'batch_size': 32, 'tokens_per_step': 32768, 'warmup_steps': 915, 'max_steps': 6103, '_wandb': {}}
+2026-02-26 15:30:26,236 INFO    MainThread:5904 [wandb_init.py:init():854] starting backend
+2026-02-26 15:30:26,504 INFO    MainThread:5904 [wandb_init.py:init():857] sending inform_init request
+2026-02-26 15:30:26,514 INFO    MainThread:5904 [wandb_init.py:init():865] backend started and connected
+2026-02-26 15:30:26,515 INFO    MainThread:5904 [wandb_init.py:init():936] updated telemetry
+2026-02-26 15:30:26,520 INFO    MainThread:5904 [wandb_init.py:init():960] communicating run to backend with 90.0 second timeout
+2026-02-26 15:30:32,186 INFO    MainThread:5904 [wandb_init.py:init():1011] starting run threads in backend
+2026-02-26 15:30:33,011 INFO    MainThread:5904 [wandb_run.py:_console_start():2506] atexit reg
+2026-02-26 15:30:33,011 INFO    MainThread:5904 [wandb_run.py:_redirect():2354] redirect: wrap_raw
+2026-02-26 15:30:33,011 INFO    MainThread:5904 [wandb_run.py:_redirect():2423] Wrapping output streams.
+2026-02-26 15:30:33,011 INFO    MainThread:5904 [wandb_run.py:_redirect():2446] Redirects installed.
+2026-02-26 15:30:33,026 INFO    MainThread:5904 [wandb_init.py:init():1049] run started, returning control to user process
+2026-02-26 16:12:40,715 INFO    MainThread:5904 [wandb_run.py:_finish():2272] finishing run tzach/toy-transformer-replication/trcpjlfd
+2026-02-26 16:12:40,719 INFO    MainThread:5904 [wandb_run.py:_atexit_cleanup():2471] got exitcode: 0
+2026-02-26 16:12:40,719 INFO    MainThread:5904 [wandb_run.py:_restore():2453] restore
+2026-02-26 16:12:40,719 INFO    MainThread:5904 [wandb_run.py:_restore():2459] restore done
+2026-02-26 16:12:41,384 INFO    MainThread:5904 [wandb_run.py:_footer_sync_info():3867] logging synced files
diff --git a/wandb/run-20260226_153026-trcpjlfd/run-trcpjlfd.wandb b/wandb/run-20260226_153026-trcpjlfd/run-trcpjlfd.wandb
new file mode 100644
index 0000000000000000000000000000000000000000..e4ca3845d1c41e91a4d67d218e262317c59ac3c2
--- /dev/null
+++ b/wandb/run-20260226_153026-trcpjlfd/run-trcpjlfd.wandb
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:63948b1b9244b00019c110f3957fac0b53dc62ad14a2ef91eeb719b4e1244b34
+size 3893836