Add 32b Qwen 2.5

5435671cd9169951a7151010db208af1838dfda8 · 2024-11-11 20:46:09 -0700 · Austin

Files touched

M exo/models.py

Diff

commit 5435671cd9169951a7151010db208af1838dfda8
Author: Austin <austinbv@users.noreply.github.com>
Date:   Mon Nov 11 20:46:09 2024 -0700

    Add 32b Qwen 2.5
---
 exo/models.py | 1 +
 1 file changed, 1 insertion(+)

diff --git a/exo/models.py b/exo/models.py
index caa6ce92..6a24c4f5 100644
--- a/exo/models.py
+++ b/exo/models.py
@@ -43,6 +43,7 @@ model_base_shards = {
   ### qwen
   "qwen-2.5-coder-1.5b": {"MLXDynamicShardInferenceEngine": Shard(model_id="mlx-community/Qwen2.5-Coder-1.5B-Instruct-4bit", start_layer=0, end_layer=0, n_layers=28),},
   "qwen-2.5-coder-7b": {"MLXDynamicShardInferenceEngine": Shard(model_id="mlx-community/Qwen2.5-Coder-7B-Instruct-4bit", start_layer=0, end_layer=0, n_layers=28),},
+  "qwen-2.5-coder-32b": {"MLXDynamicShardInferenceEngine": Shard(model_id="mlx-community/Qwen2.5-Coder-32B-Instruct-4bit", start_layer=0, end_layer=0, n_layers=28),},
   "qwen-2.5-7b": {"MLXDynamicShardInferenceEngine": Shard(model_id="mlx-community/Qwen2.5-7B-Instruct-4bit", start_layer=0, end_layer=0, n_layers=28),},
   "qwen-2.5-math-7b": {"MLXDynamicShardInferenceEngine": Shard(model_id="mlx-community/Qwen2.5-Math-7B-Instruct-4bit", start_layer=0, end_layer=0, n_layers=28),},
   "qwen-2.5-14b": {"MLXDynamicShardInferenceEngine": Shard(model_id="mlx-community/Qwen2.5-14B-Instruct-4bit", start_layer=0, end_layer=0, n_layers=48),},

← 167e756b add documentation of HF_HOME model storage location in READM · back to Exo · add qwen2.5 coder 3b,14b,32b e9ba815c →