revert back to CORRECTED

6ae6ebeb05650a446351264ddc8706a274d527e3 · 2024-11-03 18:44:43 -0500 · Alex Cheema

Files touched

M exo/models.py

Diff

commit 6ae6ebeb05650a446351264ddc8706a274d527e3
Author: Alex Cheema <alexcheema123@gmail.com>
Date:   Sun Nov 3 18:44:43 2024 -0500

    revert back to CORRECTED
---
 exo/models.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/exo/models.py b/exo/models.py
index dce0693c..5791bdd5 100644
--- a/exo/models.py
+++ b/exo/models.py
@@ -13,7 +13,7 @@ model_base_shards = {
     "TinygradDynamicShardInferenceEngine": Shard(model_id="NousResearch/Meta-Llama-3.1-70B-Instruct", start_layer=0, end_layer=0, n_layers=80),
   },
   "llama-3.1-70b-bf16": {
-    "MLXDynamicShardInferenceEngine": Shard(model_id="mlx-community/Meta-Llama-3.1-70B-Instruct-bf16", start_layer=0, end_layer=0, n_layers=80),
+    "MLXDynamicShardInferenceEngine": Shard(model_id="mlx-community/Meta-Llama-3.1-70B-Instruct-bf16-CORRECTED", start_layer=0, end_layer=0, n_layers=80),
     "TinygradDynamicShardInferenceEngine": Shard(model_id="NousResearch/Meta-Llama-3.1-70B-Instruct", start_layer=0, end_layer=0, n_layers=80),
   },
   "llama-3.1-405b": {"MLXDynamicShardInferenceEngine": Shard(model_id="mlx-community/Meta-Llama-3.1-405B-4bit", start_layer=0, end_layer=0, n_layers=126),},

← e8e05e15 remove CORRECTED llama 70b · back to Exo · add llama-3.1-405b-8bit c8438b6d →