From: Georgi Gerganov <redacted>
Date: Thu, 20 Apr 2023 20:23:07 +0000 (+0300)
Subject: minor : fix GPT-NeoX name
X-Git-Tag: upstream/0.0.1642~1535
X-Git-Url: https://git.djapps.eu/?a=commitdiff_plain;h=992bb69774ba6a110337e391080e44077f15b1d8;p=pkg%2Fggml%2Fsources%2Fggml

minor : fix GPT-NeoX name
---

diff --git a/README.md b/README.md
index ac8240e0..c69d790d 100644
--- a/README.md
+++ b/README.md
@@ -29,7 +29,7 @@ Some of the development is currently happening in the [llama.cpp](https://github
 - [X] Example of RWKV inference [saharNooby/rwkv.cpp](https://github.com/saharNooby/rwkv.cpp)
 - [ ] Example of [SAM](https://github.com/facebookresearch/segment-anything) inference
 - [ ] Idea for GPU support: https://github.com/ggerganov/llama.cpp/discussions/915
-- [X] Example of StableLM (GPTNeoX) inference [examples/stablelm](https://github.com/ggerganov/ggml/tree/master/examples/stablelm)
+- [X] Example of StableLM (GPT-NeoX) inference [examples/stablelm](https://github.com/ggerganov/ggml/tree/master/examples/stablelm)
 
 ## Whisper inference (example)
 
diff --git a/examples/stablelm/README.md b/examples/stablelm/README.md
index 81d18dcf..7882d8b3 100644
--- a/examples/stablelm/README.md
+++ b/examples/stablelm/README.md
@@ -1,6 +1,6 @@
 # StableLM
 
-Transformer architecture: GPTNeoX
+Transformer architecture: GPT-NeoX
 
 Ref: https://github.com/stability-AI/stableLM/#stablelm-alpha
 
diff --git a/examples/stablelm/main.cpp b/examples/stablelm/main.cpp
index a3c10ed4..7cccb687 100644
--- a/examples/stablelm/main.cpp
+++ b/examples/stablelm/main.cpp
@@ -480,7 +480,7 @@ bool stablelm_eval(
             struct ggml_tensor * Kcur = ggml_cont(ctx0, ggml_view_3d(ctx0, cur, n_embd/n_head, n_head, N, cur->nb[1]/n_head, cur->nb[1], 1*sizeof(float)*n_embd/n_head));
             struct ggml_tensor * Vcur = ggml_cont(ctx0, ggml_view_3d(ctx0, cur, n_embd/n_head, n_head, N, cur->nb[1]/n_head, cur->nb[1], 2*sizeof(float)*n_embd/n_head));
 
-            // using mode = 2 for GPTNeoX mode
+            // using mode = 2 for GPT-NeoX mode
             Qcur = ggml_rope(ctx0, Qcur, n_past, n_rot, 2);
             Kcur = ggml_rope(ctx0, Kcur, n_past, n_rot, 2);
 
diff --git a/include/ggml/ggml.h b/include/ggml/ggml.h
index 51cab01f..460d4ffe 100644
--- a/include/ggml/ggml.h
+++ b/include/ggml/ggml.h
@@ -631,7 +631,7 @@ struct ggml_tensor * ggml_soft_max(
 // rotary position embedding
 // in-place, returns view(a)
 // if mode & 1 == 1, skip n_past elements
-// if mode & 2 == 1, GPTNeoX style
+// if mode & 2 == 1, GPT-NeoX style
 // TODO: avoid creating a new tensor every time
 struct ggml_tensor * ggml_rope(
         struct ggml_context * ctx,