mlabonne commited on
Commit
36c1f1f
1 Parent(s): 2fb53b6

Upload folder using huggingface_hub

Browse files
README.md CHANGED
@@ -7,19 +7,15 @@ tags:
7
  - lazymergekit
8
  - cognitivecomputations/dolphin-2_6-phi-2
9
  - lxuechen/phi-2-dpo
10
- - Yhyu13/phi-2-sft-dpo-gpt4_en-ep1
11
- - mrm8488/phi-2-coder
12
  ---
13
 
14
  ![](https://i.imgur.com/UOb2fvh.jpg)
15
 
16
- # phixtral-4x2.8
17
 
18
  phixtral-2x2.8 is a Mixure of Experts (MoE) made with the following models using a custom version of mergekit:
19
  * [cognitivecomputations/dolphin-2_6-phi-2](https://huggingface.co/cognitivecomputations/dolphin-2_6-phi-2)
20
  * [lxuechen/phi-2-dpo](https://huggingface.co/lxuechen/phi-2-dpo)
21
- * [Yhyu13/phi-2-sft-dpo-gpt4_en-ep1](https://huggingface.co/Yhyu13/phi-2-sft-dpo-gpt4_en-ep1)
22
- * [mrm8488/phi-2-coder](https://huggingface.co/mrm8488/phi-2-coder)
23
 
24
  ## 🧩 Configuration
25
 
@@ -31,10 +27,6 @@ experts:
31
  positive_prompts: [""]
32
  - source_model: lxuechen/phi-2-dpo
33
  positive_prompts: [""]
34
- - source_model: Yhyu13/phi-2-sft-dpo-gpt4_en-ep1
35
- positive_prompts: [""]
36
- - source_model: mrm8488/phi-2-coder
37
- positive_prompts: [""]
38
  ```
39
 
40
  ## 💻 Usage
 
7
  - lazymergekit
8
  - cognitivecomputations/dolphin-2_6-phi-2
9
  - lxuechen/phi-2-dpo
 
 
10
  ---
11
 
12
  ![](https://i.imgur.com/UOb2fvh.jpg)
13
 
14
+ # phixtral-2x2.8
15
 
16
  phixtral-2x2.8 is a Mixure of Experts (MoE) made with the following models using a custom version of mergekit:
17
  * [cognitivecomputations/dolphin-2_6-phi-2](https://huggingface.co/cognitivecomputations/dolphin-2_6-phi-2)
18
  * [lxuechen/phi-2-dpo](https://huggingface.co/lxuechen/phi-2-dpo)
 
 
19
 
20
  ## 🧩 Configuration
21
 
 
27
  positive_prompts: [""]
28
  - source_model: lxuechen/phi-2-dpo
29
  positive_prompts: [""]
 
 
 
 
30
  ```
31
 
32
  ## 💻 Usage
config.json CHANGED
@@ -34,7 +34,7 @@
34
  "num_experts_per_tok": 2,
35
  "num_hidden_layers": 32,
36
  "num_key_value_heads": 8,
37
- "num_local_experts": 4,
38
  "output_router_logits": false,
39
  "resid_pdrop": 0.1,
40
  "rms_norm_eps": 1e-06,
 
34
  "num_experts_per_tok": 2,
35
  "num_hidden_layers": 32,
36
  "num_key_value_heads": 8,
37
+ "num_local_experts": 2,
38
  "output_router_logits": false,
39
  "resid_pdrop": 0.1,
40
  "rms_norm_eps": 1e-06,
mergekit_moe_config.yml CHANGED
@@ -4,8 +4,4 @@ experts:
4
  - source_model: cognitivecomputations/dolphin-2_6-phi-2
5
  positive_prompts: [""]
6
  - source_model: lxuechen/phi-2-dpo
7
- positive_prompts: [""]
8
- - source_model: Yhyu13/phi-2-sft-dpo-gpt4_en-ep1
9
- positive_prompts: [""]
10
- - source_model: mrm8488/phi-2-coder
11
  positive_prompts: [""]
 
4
  - source_model: cognitivecomputations/dolphin-2_6-phi-2
5
  positive_prompts: [""]
6
  - source_model: lxuechen/phi-2-dpo
 
 
 
 
7
  positive_prompts: [""]
model-00001-of-00001.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ca5d85304bec3b0f81ea5c166c72f14208ff1640e02c8cdf9f23572ee949a821
3
+ size 8916015136
model.safetensors.index.json CHANGED
The diff for this file is too large to render. See raw diff