k-l-lambda
/

Llama-3.2-100M

Model card Files Files and versions

k-l-lambda commited on Nov 22, 2024

Commit

c96568f

·

1 Parent(s): ec2e07c

updated super parameters.

Files changed (2) hide show

config.json +4 -4
model.safetensors +2 -2

config.json CHANGED Viewed

@@ -10,17 +10,17 @@
     128008,
     128009
   ],
-  "head_dim": 24,
   "hidden_act": "silu",
-  "hidden_size": 768,
   "initializer_range": 0.02,
-  "intermediate_size": 3072,
   "max_position_embeddings": 131072,
   "mlp_bias": false,
   "model_type": "llama",
   "num_attention_heads": 32,
   "num_hidden_layers": 4,
-  "num_key_value_heads": 8,
   "pretraining_tp": 1,
   "rms_norm_eps": 1e-05,
   "rope_scaling": {

     128008,
     128009
   ],
+  "head_dim": 64,
   "hidden_act": "silu",
+  "hidden_size": 512,
   "initializer_range": 0.02,
+  "intermediate_size": 2048,
   "max_position_embeddings": 131072,
   "mlp_bias": false,
   "model_type": "llama",
   "num_attention_heads": 32,
   "num_hidden_layers": 4,
+  "num_key_value_heads": 2,
   "pretraining_tp": 1,
   "rms_norm_eps": 1e-05,
   "rope_scaling": {

model.safetensors CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:818460eb8aa2be3ccd4162b487c440e63ed170343060df749f3eca99ae834f37
-size 265438880

 version https://git-lfs.github.com/spec/v1
+oid sha256:4ec3fccbc7faf852bdaa09fd4531fbb316f1ab86065d109589109ebb6476c8a7
+size 174339240