Upload folder using huggingface_hub

README.md

@@ -2,7 +2,6 @@
 base_model:
 - Qwen/Qwen2.5-Coder-7B-Instruct
 - Qwen/Qwen2.5-Math-7B-Instruct
-- Qwen/Qwen2.5-7B-Instruct
 library_name: transformers
 tags:
 - mergekit
@@ -16,7 +15,7 @@ This is a merge of pre-trained language models created using [mergekit](https://
 ## Merge Details
 ### Merge Method
 
-This model was merged using the [DARE TIES](https://arxiv.org/abs/2311.03099) merge method using [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct) as a base.
+This model was merged using the [SLERP](https://en.wikipedia.org/wiki/Slerp) merge method.
 
 ### Models Merged
 
@@ -30,21 +29,22 @@ The following YAML configuration was used to produce this model:
 
 ```yaml
 
-models:
-  - model: Qwen/Qwen2.5-Coder-7B-Instruct
-    parameters:
-      weight: 0.5
-      density: 0.7
-  - model: Qwen/Qwen2.5-Math-7B-Instruct
-    parameters:
-      weight: 0.5
-      density: 0.6
-merge_method: dare_ties
-base_model: Qwen/Qwen2.5-7B-Instruct
+slices:
+  - sources:
+      - model: Qwen/Qwen2.5-Coder-7B-Instruct
+        layer_range: [0, 28]
+      - model: Qwen/Qwen2.5-Math-7B-Instruct
+        layer_range: [0, 28]
+merge_method: slerp
+base_model: Qwen/Qwen2.5-Coder-7B-Instruct
 parameters:
-  int8_mask: true
-  normalize: true
+  t:
+    - filter: self_attn
+      value: [0, 0.5, 0.3, 0.7, 1]
+    - filter: mlp
+      value: [1, 0.5, 0.7, 0.3, 0]
+    - value: 0.5
 dtype: bfloat16
-tokenizer_source: Qwen/Qwen2.5-7B-Instruct
+tokenizer_source: Qwen/Qwen2.5-Coder-7B-Instruct
 
 ```

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "Qwen/Qwen2.5-7B-Instruct",
+  "_name_or_path": "Qwen/Qwen2.5-Coder-7B-Instruct",
   "architectures": [
     "Qwen2ForCausalLM"
   ],
@@ -22,7 +22,7 @@
   "sliding_window": null,
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
-  "transformers_version": "4.46.0",
+  "transformers_version": "4.46.3",
   "use_cache": true,
   "use_sliding_window": false,
   "vocab_size": 151665

mergekit_config.yml

@@ -1,17 +1,18 @@
 
-models:
-  - model: Qwen/Qwen2.5-Coder-7B-Instruct
-    parameters:
-      weight: 0.5
-      density: 0.7
-  - model: Qwen/Qwen2.5-Math-7B-Instruct
-    parameters:
-      weight: 0.5
-      density: 0.6
-merge_method: dare_ties
-base_model: Qwen/Qwen2.5-7B-Instruct
+slices:
+  - sources:
+      - model: Qwen/Qwen2.5-Coder-7B-Instruct
+        layer_range: [0, 28]
+      - model: Qwen/Qwen2.5-Math-7B-Instruct
+        layer_range: [0, 28]
+merge_method: slerp
+base_model: Qwen/Qwen2.5-Coder-7B-Instruct
 parameters:
-  int8_mask: true
-  normalize: true
+  t:
+    - filter: self_attn
+      value: [0, 0.5, 0.3, 0.7, 1]
+    - filter: mlp
+      value: [1, 0.5, 0.7, 0.3, 0]
+    - value: 0.5
 dtype: bfloat16
-tokenizer_source: Qwen/Qwen2.5-7B-Instruct
+tokenizer_source: Qwen/Qwen2.5-Coder-7B-Instruct

model-00001-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8399c73c766e9f26a2ca99c810544d216aea55754c3ec33bb9044b9c6ed29e68
+oid sha256:fbb57e71de29f5269e1a4089f706ed9ef5f7a464d011f05f66fa22349d4efdd7
 size 4970978712

model-00002-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:61569ef3ba6f4496866fe951e522d0509e683633ab051129004eab56147610b1
+oid sha256:63b17ef77e02df2c0befca0bb94f29553d03934c090bd33cc9fad620d81dde07
 size 4932751032

model-00003-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2f8245efd3a5bca7c1573ea3ade0e53bf6b86c2712de2e51eca9e0231f1f0dcd
+oid sha256:cfefb13cc23ce874a2f7c725922f0586b6ec5d84afe9bdfb8491d52012f0f048
 size 4991495808

model-00004-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:93399810496e39daf39cfe8afb84a87f7ca842deb55087cd1d3525993fdf5b51
+oid sha256:4c852c55f2f2e2654dc910179922bc561a7e668b2b739a9adfa4eb8602f99681
 size 330326240

tokenizer_config.json

@@ -199,7 +199,7 @@
   "clean_up_tokenization_spaces": false,
   "eos_token": "<|im_end|>",
   "errors": "replace",
-  "model_max_length": 131072,
+  "model_max_length": 32768,
   "pad_token": "<|endoftext|>",
   "split_special_tokens": false,
   "tokenizer_class": "Qwen2Tokenizer",