From ed6d22d4589eeda8eb3fb973bc66957fda6092a0 Mon Sep 17 00:00:00 2001
From: ngxson <thichthat@gmail.com>
Date: Sun, 12 May 2024 11:27:38 +0200
Subject: [PATCH 1/2] gguf: update README

---
 packages/gguf/README.md | 40 ++++++++++++++++++++++++++++++++++++++++
 1 file changed, 40 insertions(+)

diff --git a/packages/gguf/README.md b/packages/gguf/README.md
index 260d602983..c322ecab77 100644
--- a/packages/gguf/README.md
+++ b/packages/gguf/README.md
@@ -18,6 +18,8 @@ npm install @huggingface/gguf
 
 ## Usage
 
+### Basic usage
+
 ```ts
 import { GGMLQuantizationType, gguf } from "@huggingface/gguf";
 
@@ -56,6 +58,44 @@ console.log(tensorInfos);
 
 ```
 
+### Reading a local file
+
+```ts
+// Reading a local file. (Not supported on browser)
+const { metadata, tensorInfos } = await gguf(
+  './my_model.gguf',
+  { allowLocalFile: true },
+);
+```
+
+### Strictly typed
+
+By default, known fields in `metadata` are typed. This includes various fields found in [llama.cpp](https://github.com/ggerganov/llama.cpp), [whisper.cpp](https://github.com/ggerganov/whisper.cpp) and [ggml](https://github.com/ggerganov/ggml).
+
+```ts
+const { metadata, tensorInfos } = await gguf(URL_MODEL);
+
+// Type check for model architecture at runtime
+if (metadata["general.architecture"] === "llama") {
+
+  // "llama.attention.head_count" is a valid key for llama architecture
+  console.log(model["llama.attention.head_count"]);
+
+  // "mamba.ssm.conv_kernel" is an invalid key, because it requires model architecture to be mamba
+  console.log(model["mamba.ssm.conv_kernel"]); // error
+}
+```
+
+### Disable strictly typed
+
+Because GGUF format can be used to store tensors, we can technically use it for other usages. For example, storing [control vectors](https://github.com/ggerganov/llama.cpp/pull/5970), [lora weights](https://github.com/ggerganov/llama.cpp/pull/2632),...
+
+In case you want to use your own GGUF metadata structure, you can disable strict type by casting the parse output to `GGUFParseOutput<{ strict: false }>`:
+
+```ts
+const { metadata, tensorInfos }: GGUFParseOutput<{ strict: false }> = await gguf(URL_LLAMA);
+```
+
 ## Hugging Face Hub
 
 The Hub supports all file formats and has built-in features for GGUF format. 

From a6e27890ab069daa4fd97e1a2f282ac16451ea9c Mon Sep 17 00:00:00 2001
From: Xuan Son Nguyen <thichthat@gmail.com>
Date: Mon, 13 May 2024 22:08:43 +0200
Subject: [PATCH 2/2] Apply suggestions from code review

Co-authored-by: Julien Chaumond <julien@huggingface.co>
Co-authored-by: Mishig <mishig.davaadorj@coloradocollege.edu>
---
 packages/gguf/README.md | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/packages/gguf/README.md b/packages/gguf/README.md
index c322ecab77..763b494708 100644
--- a/packages/gguf/README.md
+++ b/packages/gguf/README.md
@@ -78,7 +78,7 @@ const { metadata, tensorInfos } = await gguf(URL_MODEL);
 // Type check for model architecture at runtime
 if (metadata["general.architecture"] === "llama") {
 
-  // "llama.attention.head_count" is a valid key for llama architecture
+  // "llama.attention.head_count" is a valid key for llama architecture, this is typed as a number
   console.log(model["llama.attention.head_count"]);
 
   // "mamba.ssm.conv_kernel" is an invalid key, because it requires model architecture to be mamba
@@ -88,9 +88,9 @@ if (metadata["general.architecture"] === "llama") {
 
 ### Disable strictly typed
 
-Because GGUF format can be used to store tensors, we can technically use it for other usages. For example, storing [control vectors](https://github.com/ggerganov/llama.cpp/pull/5970), [lora weights](https://github.com/ggerganov/llama.cpp/pull/2632),...
+Because GGUF format can be used to store tensors, we can technically use it for other usages. For example, storing [control vectors](https://github.com/ggerganov/llama.cpp/pull/5970), [lora weights](https://github.com/ggerganov/llama.cpp/pull/2632), etc.
 
-In case you want to use your own GGUF metadata structure, you can disable strict type by casting the parse output to `GGUFParseOutput<{ strict: false }>`:
+In case you want to use your own GGUF metadata structure, you can disable strict typing by casting the parse output to `GGUFParseOutput<{ strict: false }>`:
 
 ```ts
 const { metadata, tensorInfos }: GGUFParseOutput<{ strict: false }> = await gguf(URL_LLAMA);