Copy train_post_decay/config.json to root as default checkpoint
Browse files- config.json +29 -0
config.json
ADDED
|
@@ -0,0 +1,29 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"LoopLMForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"context_length": 1024,
|
| 6 |
+
"d_ff": 2752,
|
| 7 |
+
"d_model": 1024,
|
| 8 |
+
"dtype": "float32",
|
| 9 |
+
"lb_loss_factor": 0.01,
|
| 10 |
+
"lz_loss_factor": 0.001,
|
| 11 |
+
"max_length": 1024,
|
| 12 |
+
"model_type": "loop-lm",
|
| 13 |
+
"model_variant": "base",
|
| 14 |
+
"num_active": 2,
|
| 15 |
+
"num_experts": 8,
|
| 16 |
+
"num_heads": 16,
|
| 17 |
+
"num_layers": 16,
|
| 18 |
+
"num_layers_in_stack": 8,
|
| 19 |
+
"num_stacks": 2,
|
| 20 |
+
"rope_theta": 10000.0,
|
| 21 |
+
"transformers_version": "5.3.0",
|
| 22 |
+
"vocab_size": 50257,
|
| 23 |
+
"weight_tying": false,
|
| 24 |
+
"width_ratio": 8.0,
|
| 25 |
+
"auto_map": {
|
| 26 |
+
"AutoConfig": "modeling_loop_lm.LoopLMConfig",
|
| 27 |
+
"AutoModelForCausalLM": "modeling_loop_lm.LoopLMForCausalLM"
|
| 28 |
+
}
|
| 29 |
+
}
|