Add model checkpoint
Browse files- config.json +38 -0
- model.safetensors +3 -0
config.json
ADDED
|
@@ -0,0 +1,38 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "/home/rllaskowski_a100/mot-training/experiments/mot_medium_32_8/checkpoint-125000",
|
| 3 |
+
"activation_function": "gelu_new",
|
| 4 |
+
"architectures": [
|
| 5 |
+
"MoTModel"
|
| 6 |
+
],
|
| 7 |
+
"attn_pdrop": 0.1,
|
| 8 |
+
"bos_token_id": 50256,
|
| 9 |
+
"embd_pdrop": 0.1,
|
| 10 |
+
"emit_softmax_over_experts": false,
|
| 11 |
+
"eos_token_id": 50256,
|
| 12 |
+
"expert_size": 256,
|
| 13 |
+
"group_size": 32,
|
| 14 |
+
"init_scale": 1.0,
|
| 15 |
+
"initializer_range": 0.02,
|
| 16 |
+
"layer_norm_epsilon": 1e-05,
|
| 17 |
+
"model_type": "mot",
|
| 18 |
+
"n_embd": 512,
|
| 19 |
+
"n_expert": 256,
|
| 20 |
+
"n_head": 8,
|
| 21 |
+
"n_inner": 32,
|
| 22 |
+
"n_layer": 8,
|
| 23 |
+
"n_positions": 1024,
|
| 24 |
+
"reorder_and_upcast_attn": false,
|
| 25 |
+
"resid_pdrop": 0.1,
|
| 26 |
+
"scale_attn_by_inverse_layer_idx": false,
|
| 27 |
+
"scale_attn_weights": true,
|
| 28 |
+
"summary_activation": null,
|
| 29 |
+
"summary_first_dropout": 0.1,
|
| 30 |
+
"summary_proj_to_labels": true,
|
| 31 |
+
"summary_type": "cls_index",
|
| 32 |
+
"summary_use_proj": true,
|
| 33 |
+
"torch_dtype": "float32",
|
| 34 |
+
"transformers_version": "4.39.0.dev0",
|
| 35 |
+
"use_cache": true,
|
| 36 |
+
"use_discrete_routing": false,
|
| 37 |
+
"vocab_size": 50257
|
| 38 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:15c1e153bdc798e5d76c9f2c00aa0ca34332082fbc889825811e87290041e302
|
| 3 |
+
size 2290399320
|