diff --git a/requirements.txt b/requirements.txt
index d384acf..165a86d 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -20,7 +20,7 @@ safetensors==0.4.1
 scipy
 sentencepiece
 tensorboard
-transformers==4.36.*
+git+https://github.com/huggingface/transformers.git@fa6d12f74fef165806f2298acc9f4ac4f0b20d43
 tqdm
 wandb
 
diff --git a/requirements_amd.txt b/requirements_amd.txt
index 845cb0b..f05b404 100644
--- a/requirements_amd.txt
+++ b/requirements_amd.txt
@@ -20,7 +20,7 @@ safetensors==0.4.1
 scipy
 sentencepiece
 tensorboard
-transformers==4.36.*
+git+https://github.com/huggingface/transformers.git@fa6d12f74fef165806f2298acc9f4ac4f0b20d43
 tqdm
 wandb
 
diff --git a/requirements_amd_noavx2.txt b/requirements_amd_noavx2.txt
index 5136be3..83ddede 100644
--- a/requirements_amd_noavx2.txt
+++ b/requirements_amd_noavx2.txt
@@ -20,7 +20,7 @@ safetensors==0.4.1
 scipy
 sentencepiece
 tensorboard
-transformers==4.36.*
+git+https://github.com/huggingface/transformers.git@fa6d12f74fef165806f2298acc9f4ac4f0b20d43
 tqdm
 wandb
 
diff --git a/requirements_apple_intel.txt b/requirements_apple_intel.txt
index 14467b9..e0907c0 100644
--- a/requirements_apple_intel.txt
+++ b/requirements_apple_intel.txt
@@ -20,7 +20,7 @@ safetensors==0.4.1
 scipy
 sentencepiece
 tensorboard
-transformers==4.36.*
+git+https://github.com/huggingface/transformers.git@fa6d12f74fef165806f2298acc9f4ac4f0b20d43
 tqdm
 wandb
 
diff --git a/requirements_apple_silicon.txt b/requirements_apple_silicon.txt
index 2e676ef..fbfa150 100644
--- a/requirements_apple_silicon.txt
+++ b/requirements_apple_silicon.txt
@@ -20,7 +20,7 @@ safetensors==0.4.1
 scipy
 sentencepiece
 tensorboard
-transformers==4.36.*
+git+https://github.com/huggingface/transformers.git@fa6d12f74fef165806f2298acc9f4ac4f0b20d43
 tqdm
 wandb
 
diff --git a/requirements_cpu_only.txt b/requirements_cpu_only.txt
index 38306ca..9e094f1 100644
--- a/requirements_cpu_only.txt
+++ b/requirements_cpu_only.txt
@@ -20,7 +20,7 @@ safetensors==0.4.1
 scipy
 sentencepiece
 tensorboard
-transformers==4.36.*
+git+https://github.com/huggingface/transformers.git@fa6d12f74fef165806f2298acc9f4ac4f0b20d43
 tqdm
 wandb
 
diff --git a/requirements_cpu_only_noavx2.txt b/requirements_cpu_only_noavx2.txt
index b0bdd08..c7c02f4 100644
--- a/requirements_cpu_only_noavx2.txt
+++ b/requirements_cpu_only_noavx2.txt
@@ -20,7 +20,7 @@ safetensors==0.4.1
 scipy
 sentencepiece
 tensorboard
-transformers==4.36.*
+git+https://github.com/huggingface/transformers.git@fa6d12f74fef165806f2298acc9f4ac4f0b20d43
 tqdm
 wandb
 
diff --git a/requirements_noavx2.txt b/requirements_noavx2.txt
index 37ff051..38f77f5 100644
--- a/requirements_noavx2.txt
+++ b/requirements_noavx2.txt
@@ -20,7 +20,7 @@ safetensors==0.4.1
 scipy
 sentencepiece
 tensorboard
-transformers==4.36.*
+git+https://github.com/huggingface/transformers.git@fa6d12f74fef165806f2298acc9f4ac4f0b20d43
 tqdm
 wandb
 
diff --git a/requirements_nowheels.txt b/requirements_nowheels.txt
index 9df2ef0..abc3494 100644
--- a/requirements_nowheels.txt
+++ b/requirements_nowheels.txt
@@ -20,7 +20,7 @@ safetensors==0.4.1
 scipy
 sentencepiece
 tensorboard
-transformers==4.36.*
+git+https://github.com/huggingface/transformers.git@fa6d12f74fef165806f2298acc9f4ac4f0b20d43
 tqdm
 wandb