Upload config.toml with huggingface_hub
Browse files- config.toml +55 -0
config.toml
ADDED
|
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# 5M Symbiogenesis config — multi-organelle sequence mixing
|
| 2 |
+
# 8 blocks with 3 organelles per block: CausalConv + Monarch + LongConv
|
| 3 |
+
# Inspired by biological symbiogenesis (organism fusion)
|
| 4 |
+
|
| 5 |
+
[model]
|
| 6 |
+
arch = "symbiogenesis"
|
| 7 |
+
embed_dim = 256
|
| 8 |
+
n_layers = 8
|
| 9 |
+
n_heads = 4 # unused by Symbio, kept for struct compat
|
| 10 |
+
head_dim = 64 # unused by Symbio
|
| 11 |
+
n_monarch_heads = 1 # single-head Monarch per block
|
| 12 |
+
conv_kernel_size = 4
|
| 13 |
+
ffn_mult = 4
|
| 14 |
+
context_length = 256
|
| 15 |
+
dropout = 0.0
|
| 16 |
+
bias = false
|
| 17 |
+
weight_tying = true
|
| 18 |
+
free_energy_beta = 0.001
|
| 19 |
+
|
| 20 |
+
[training]
|
| 21 |
+
optimizer = "adamw"
|
| 22 |
+
lr = 6e-4
|
| 23 |
+
min_lr = 6e-5
|
| 24 |
+
warmup_steps = 500
|
| 25 |
+
max_steps = 12305
|
| 26 |
+
batch_size = 32
|
| 27 |
+
grad_clip = 1.0
|
| 28 |
+
precision = "f16"
|
| 29 |
+
eval_interval = 500
|
| 30 |
+
eval_steps = 25
|
| 31 |
+
checkpoint_interval = 2000
|
| 32 |
+
seed = 42
|
| 33 |
+
|
| 34 |
+
[training.curriculum]
|
| 35 |
+
enabled = false
|
| 36 |
+
|
| 37 |
+
[training.coreset]
|
| 38 |
+
enabled = false
|
| 39 |
+
|
| 40 |
+
[training.gelation]
|
| 41 |
+
enabled = true
|
| 42 |
+
cusum_threshold = 5.0
|
| 43 |
+
window_size = 10
|
| 44 |
+
|
| 45 |
+
[data]
|
| 46 |
+
train_path = "../text-pipeline/output/train.txt"
|
| 47 |
+
val_path = "../text-pipeline/output/val.txt"
|
| 48 |
+
tokenizer_dir = "../text-pipeline/output"
|
| 49 |
+
|
| 50 |
+
[inference]
|
| 51 |
+
precision = "f16"
|
| 52 |
+
compile = false
|
| 53 |
+
temperature = 0.8
|
| 54 |
+
top_k = 40
|
| 55 |
+
max_new_tokens = 500
|