electron-rare/exo
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Merge pull request #570 from exo-explore/moreqwen

Browse Source 
add qwen-2.5-1.5b, qwen-2.5-3b, qwen-2.5-32b
This commit is contained in:
Alex Cheema
2024-12-24 01:51:26 +00:00
committed by GitHub
parent ba8c514974 c609c05e40
commit 41df9ce1d7
1 changed files with 6 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files
exo/models.py
+6 -3
View File
@@ -92,14 +92,17 @@ model_cards = {
"llava-1.5-7b-hf": { "layers": 32, "repo": { "MLXDynamicShardInferenceEngine": "llava-hf/llava-1.5-7b-hf", }, },
### qwen
"qwen-2.5-0.5b": { "layers": 28, "repo": { "MLXDynamicShardInferenceEngine": "mlx-community/Qwen2.5-0.5B-Instruct-4bit", }, },
"qwen-2.5-1.5b": { "layers": 28, "repo": { "MLXDynamicShardInferenceEngine": "mlx-community/Qwen2.5-1.5B-Instruct-4bit", }, },
"qwen-2.5-coder-1.5b": { "layers": 28, "repo": { "MLXDynamicShardInferenceEngine": "mlx-community/Qwen2.5-Coder-1.5B-Instruct-4bit", }, },
"qwen-2.5-3b": { "layers": 36, "repo": { "MLXDynamicShardInferenceEngine": "mlx-community/Qwen2.5-3B-Instruct-4bit", }, },
"qwen-2.5-coder-3b": { "layers": 36, "repo": { "MLXDynamicShardInferenceEngine": "mlx-community/Qwen2.5-Coder-3B-Instruct-4bit", }, },
"qwen-2.5-coder-7b": { "layers": 28, "repo": { "MLXDynamicShardInferenceEngine": "mlx-community/Qwen2.5-Coder-7B-Instruct-4bit", }, },
"qwen-2.5-coder-14b": { "layers": 48, "repo": { "MLXDynamicShardInferenceEngine": "mlx-community/Qwen2.5-Coder-14B-Instruct-4bit", }, },
"qwen-2.5-coder-32b": { "layers": 64, "repo": { "MLXDynamicShardInferenceEngine": "mlx-community/Qwen2.5-Coder-32B-Instruct-4bit", }, },
"qwen-2.5-7b": { "layers": 28, "repo": { "MLXDynamicShardInferenceEngine": "mlx-community/Qwen2.5-7B-Instruct-4bit", }, },
"qwen-2.5-coder-7b": { "layers": 28, "repo": { "MLXDynamicShardInferenceEngine": "mlx-community/Qwen2.5-Coder-7B-Instruct-4bit", }, },
"qwen-2.5-math-7b": { "layers": 28, "repo": { "MLXDynamicShardInferenceEngine": "mlx-community/Qwen2.5-Math-7B-Instruct-4bit", }, },
"qwen-2.5-14b": { "layers": 48, "repo": { "MLXDynamicShardInferenceEngine": "mlx-community/Qwen2.5-14B-Instruct-4bit", }, },
"qwen-2.5-coder-14b": { "layers": 48, "repo": { "MLXDynamicShardInferenceEngine": "mlx-community/Qwen2.5-Coder-14B-Instruct-4bit", }, },
"qwen-2.5-32b": { "layers": 64, "repo": { "MLXDynamicShardInferenceEngine": "mlx-community/Qwen2.5-32B-Instruct-4bit", }, },
"qwen-2.5-coder-32b": { "layers": 64, "repo": { "MLXDynamicShardInferenceEngine": "mlx-community/Qwen2.5-Coder-32B-Instruct-4bit", }, },
"qwen-2.5-72b": { "layers": 80, "repo": { "MLXDynamicShardInferenceEngine": "mlx-community/Qwen2.5-72B-Instruct-4bit", }, },
"qwen-2.5-math-72b": { "layers": 80, "repo": { "MLXDynamicShardInferenceEngine": "mlx-community/Qwen2.5-Math-72B-Instruct-4bit", }, },
### nemotron