INT4 DFlash checkpoint weights update
#4
by baranowskiadam - opened
- config.json +13 -1
- model.safetensors +1 -1
config.json
CHANGED
|
@@ -4,7 +4,7 @@
|
|
| 4 |
"hidden_act": "silu",
|
| 5 |
"hidden_size": 3072,
|
| 6 |
"intermediate_size": 12288,
|
| 7 |
-
"max_position_embeddings":
|
| 8 |
"model_type": "laguna",
|
| 9 |
"num_attention_heads": 72,
|
| 10 |
"num_hidden_layers": 6,
|
|
@@ -21,6 +21,18 @@
|
|
| 21 |
"sliding_attention"
|
| 22 |
],
|
| 23 |
"rope_theta": 500000.0,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 24 |
"gating": "per-head",
|
| 25 |
"architectures": [
|
| 26 |
"DFlashLagunaForCausalLM"
|
|
|
|
| 4 |
"hidden_act": "silu",
|
| 5 |
"hidden_size": 3072,
|
| 6 |
"intermediate_size": 12288,
|
| 7 |
+
"max_position_embeddings": 262144,
|
| 8 |
"model_type": "laguna",
|
| 9 |
"num_attention_heads": 72,
|
| 10 |
"num_hidden_layers": 6,
|
|
|
|
| 21 |
"sliding_attention"
|
| 22 |
],
|
| 23 |
"rope_theta": 500000.0,
|
| 24 |
+
"partial_rotary_factor": 0.5,
|
| 25 |
+
"rope_parameters": {
|
| 26 |
+
"rope_type": "yarn",
|
| 27 |
+
"factor": 32.0,
|
| 28 |
+
"rope_theta": 500000.0,
|
| 29 |
+
"partial_rotary_factor": 0.5,
|
| 30 |
+
"original_max_position_embeddings": 8192,
|
| 31 |
+
"extrapolation_factor": 1.0,
|
| 32 |
+
"attn_factor": 1.0,
|
| 33 |
+
"beta_fast": 32.0,
|
| 34 |
+
"beta_slow": 1.0
|
| 35 |
+
},
|
| 36 |
"gating": "per-head",
|
| 37 |
"architectures": [
|
| 38 |
"DFlashLagunaForCausalLM"
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2229962896
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0850e39b5c079a9f1a9bafed729a4545b088a91876541d010d871f6d6d8bf909
|
| 3 |
size 2229962896
|