Upload crosscoder-tinymoe-tinygpt-27M dictionary model
Browse files- trainer_config.json +23 -0
trainer_config.json
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"trainer": {
|
| 3 |
+
"dict_class": "CrossCoder",
|
| 4 |
+
"trainer_class": "CrossCoderTrainer",
|
| 5 |
+
"activation_dim": 768,
|
| 6 |
+
"dict_size": 49152,
|
| 7 |
+
"lr": 0.001,
|
| 8 |
+
"l1_penalty": 0.03,
|
| 9 |
+
"warmup_steps": 1000,
|
| 10 |
+
"resample_steps": null,
|
| 11 |
+
"device": "cuda",
|
| 12 |
+
"layer": 3,
|
| 13 |
+
"lm_name": "tiny-gpt-27M-mixtral-5l-active-27M",
|
| 14 |
+
"wandb_name": "L3-mu3.0e-02-lr1e-03-64-mixed",
|
| 15 |
+
"submodule_name": null,
|
| 16 |
+
"use_mse_loss": false,
|
| 17 |
+
"code_normalization": "MIXED",
|
| 18 |
+
"code_normalization_alpha_sae": 0.7,
|
| 19 |
+
"code_normalization_alpha_cc": 0.3,
|
| 20 |
+
"target_rms": 1.0,
|
| 21 |
+
"num_layers": 2
|
| 22 |
+
}
|
| 23 |
+
}
|