Add Phi3SmallForCausalLM (#405)

arcee-ai · Aug 23, 2024 · f17616f · f17616f
1 parent f086664
commit f17616f
Show file tree

Hide file tree

Showing 2 changed files with 75 additions and 13 deletions.
diff --git a/mergekit/_data/architectures/phi3-small.json b/mergekit/_data/architectures/phi3-small.json
@@ -0,0 +1,68 @@
+{
+    "model_type": "phi3small",
+    "architectures": [
+        "Phi3SmallForCausalLM"
+    ],
+    "pre_weights": [
+        {
+            "name": "model.embed_tokens.weight",
+            "is_embed": true
+        }
+    ],
+    "post_weights": [
+        {
+            "name": "lm_head.weight",
+            "is_embed":true,
+            "aliases": [
+                "model.embed_tokens.weight"
+            ]
+        },
+        {
+            "name": "model.final_layernorm.weight"
+        },
+        {
+            "name": "model.final_layernorm.bias"
+        }
+    ],
+    "num_layers_config_key": "num_hidden_layers",
+    "layer_templates": {
+        "weights": [
+            {
+                "name": "model.layers.${layer_index}.input_layernorm.weight"
+            },
+            {
+                "name": "model.layers.${layer_index}.input_layernorm.bias"
+            },
+            {
+                "name": "model.layers.${layer_index}.post_attention_layernorm.weight"
+            },
+            {
+                "name": "model.layers.${layer_index}.post_attention_layernorm.bias"
+            },
+            {
+                "name": "model.layers.${layer_index}.self_attn.dense.weight"
+            },
+            {
+                "name": "model.layers.${layer_index}.self_attn.dense.bias"
+            },
+            {
+                "name": "model.layers.${layer_index}.self_attn.query_key_value.weight"
+            },
+            {
+                "name": "model.layers.${layer_index}.self_attn.query_key_value.bias"
+            },
+            {
+                "name": "model.layers.${layer_index}.mlp.up_proj.weight"
+            },
+            {
+                "name": "model.layers.${layer_index}.mlp.up_proj.bias"
+            },
+            {
+                "name": "model.layers.${layer_index}.mlp.down_proj.weight"
+            },
+            {
+                "name": "model.layers.${layer_index}.mlp.down_proj.bias"
+            }
+        ]
+    }
+}
diff --git a/mergekit/_data/architectures/phi3.json b/mergekit/_data/architectures/phi3.json
@@ -1,5 +1,5 @@
 {
-    "model_type": "phi",
+    "model_type": "phi3",
     "architectures": [
         "Phi3ForCausalLM"
     ],
@@ -22,28 +22,22 @@
     "layer_templates": {
         "weights": [
             {
-                "name": "model.layers.${layer_index}.input_layernorm.weight",
-                "is_embed": false
+                "name": "model.layers.${layer_index}.input_layernorm.weight"
             },
             {
-                "name": "model.layers.${layer_index}.post_attention_layernorm.weight",
-                "is_embed": false
+                "name": "model.layers.${layer_index}.post_attention_layernorm.weight"
             },
             {
-                "name": "model.layers.${layer_index}.self_attn.o_proj.weight",
-                "is_embed": false
+                "name": "model.layers.${layer_index}.self_attn.o_proj.weight"
             },
             {
-                "name": "model.layers.${layer_index}.self_attn.qkv_proj.weight",
-                "is_embed": false
+                "name": "model.layers.${layer_index}.self_attn.qkv_proj.weight"
             },
             {
-                "name": "model.layers.${layer_index}.mlp.gate_up_proj.weight",
-                "is_embed": false
+                "name": "model.layers.${layer_index}.mlp.gate_up_proj.weight"
             },
             {
-                "name": "model.layers.${layer_index}.mlp.down_proj.weight",
-                "is_embed": false
+                "name": "model.layers.${layer_index}.mlp.down_proj.weight"
             }
         ]
     }