Upload tinystories_lambda50/experiment_config.yaml with huggingface_hub
Browse files
tinystories_lambda50/experiment_config.yaml
ADDED
|
@@ -0,0 +1,61 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
base_save_dir: .checkpoints
|
| 2 |
+
cache_dir: .cache
|
| 3 |
+
crosscoder:
|
| 4 |
+
dec_init_norm: 0.1
|
| 5 |
+
ft_init_checkpt_epoch: null
|
| 6 |
+
ft_init_checkpt_folder: null
|
| 7 |
+
hidden_dim: 1536
|
| 8 |
+
k: 20
|
| 9 |
+
data:
|
| 10 |
+
activations_harvester:
|
| 11 |
+
harvesting_batch_size: 16
|
| 12 |
+
inference_dtype: float32
|
| 13 |
+
llms:
|
| 14 |
+
- lora_name: mars-jason-25/tiny-stories-33M-TSdata-ft1
|
| 15 |
+
name: roneneldan/TinyStories-Instruct-33M
|
| 16 |
+
revision: null
|
| 17 |
+
activations_shuffle_buffer_size: 1024
|
| 18 |
+
n_batches_for_mean_estimate: 100
|
| 19 |
+
n_batches_for_norm_estimate: 100
|
| 20 |
+
sequence_iterator:
|
| 21 |
+
classname: SleeperTokenSequenceLoader
|
| 22 |
+
kwargs:
|
| 23 |
+
include_sleeper_data: false
|
| 24 |
+
sequence_length: 128
|
| 25 |
+
shuffle_buffer_size: 1024
|
| 26 |
+
experiment_name: lambda_n50_S_2025-04-18_10-33-24
|
| 27 |
+
hookpoints:
|
| 28 |
+
- blocks.0.hook_resid_pre
|
| 29 |
+
- blocks.0.ln1.hook_normalized
|
| 30 |
+
- blocks.0.hook_resid_mid
|
| 31 |
+
- blocks.0.ln2.hook_normalized
|
| 32 |
+
- blocks.1.hook_resid_pre
|
| 33 |
+
- blocks.1.ln1.hook_normalized
|
| 34 |
+
- blocks.1.hook_resid_mid
|
| 35 |
+
- blocks.1.ln2.hook_normalized
|
| 36 |
+
- blocks.2.hook_resid_pre
|
| 37 |
+
- blocks.2.ln1.hook_normalized
|
| 38 |
+
- blocks.2.hook_resid_mid
|
| 39 |
+
- blocks.2.ln2.hook_normalized
|
| 40 |
+
- blocks.3.hook_resid_pre
|
| 41 |
+
- blocks.3.ln1.hook_normalized
|
| 42 |
+
- blocks.3.hook_resid_mid
|
| 43 |
+
- blocks.3.ln2.hook_normalized
|
| 44 |
+
- blocks.3.hook_resid_post
|
| 45 |
+
seed: 42
|
| 46 |
+
train:
|
| 47 |
+
batch_size: 256
|
| 48 |
+
beta_n: 1.0
|
| 49 |
+
epochs: null
|
| 50 |
+
lam_n: 50.0
|
| 51 |
+
log_every_n_steps: 100
|
| 52 |
+
num_steps: 50000
|
| 53 |
+
num_steps_per_epoch: null
|
| 54 |
+
optimizer:
|
| 55 |
+
initial_learning_rate: 0.0001
|
| 56 |
+
last_pct_of_steps: 0.25
|
| 57 |
+
warmup_pct: 0.05
|
| 58 |
+
save_every_n_steps: 49999
|
| 59 |
+
wandb:
|
| 60 |
+
entity: dmitry2-uiuc
|
| 61 |
+
project: sleeper-model-diffing
|