From 4196a86fa9d217bdbb0e57ee42233df2c9f91d00 Mon Sep 17 00:00:00 2001
From: Fuegovic <32828263+fuegovic@users.noreply.github.com>
Date: Fri, 19 Apr 2024 21:40:12 -0400
Subject: [PATCH] =?UTF-8?q?=F0=9F=A6=99=20doc=20update:=20llama3=20(#2470)?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

* docs: update breaking_changes.md

* docs: update ai_endpoints.md -> llama3 for Ollama and groq

* librechat.yaml: update groq models

* Update breaking_changes.md

logs location

* Update breaking_changes.md

---------

Co-authored-by: Danny Avila <danny@librechat.ai>
---
 docs/general_info/breaking_changes.md      | 18 ++++++++----
 docs/install/configuration/ai_endpoints.md | 32 +++++++++++++++++++++-
 librechat.example.yaml                     |  8 +++++-
 3 files changed, 51 insertions(+), 7 deletions(-)

diff --git a/docs/general_info/breaking_changes.md b/docs/general_info/breaking_changes.md
index 71da46689..ac3300c27 100644
--- a/docs/general_info/breaking_changes.md
+++ b/docs/general_info/breaking_changes.md
@@ -11,14 +11,22 @@ weight: -10
     Certain changes in the updates may impact cookies, leading to unexpected behaviors if not cleared properly.
 
 ---
-## v0.7.1+
-
-!!! info "🔍 Google Search Plugin"
-
-    - **[Google Search Plugin](../features/plugins/google_search.md)**: Changed the environment variable for this plugin from `GOOGLE_API_KEY` to `GOOGLE_SEARCH_API_KEY` due to a conflict with the Google Generative AI library pulling this variable automatically. If you are using this plugin, please update your `.env` file accordingly.
 
 ## v0.7.0+
 
+!!! failure "Error Messages (UI)"
+    ![image](https://github.com/danny-avila/LibreChat/assets/32828263/0ab27798-5515-49b4-ac29-e4ad83d73d7c)
+
+    Client-facing error messages now display this warning asking to contact the admin. For the full error consult the console logs or the additional logs located in `./logs`
+
+!!! warning "🪵Logs Location"
+
+    - The full logs are now in `./logs` (they are still in `./api/logs` for local, non-docker installations)
+
+!!! warning "🔍 Google Search Plugin"
+
+    - **[Google Search Plugin](../features/plugins/google_search.md)**: Changed the environment variable for this plugin from `GOOGLE_API_KEY` to `GOOGLE_SEARCH_API_KEY` due to a conflict with the Google Generative AI library pulling this variable automatically. If you are using this plugin, please update your `.env` file accordingly.
+
 !!! info "🗃️ RAG API (Chat with Files)"
 
     - **RAG API Update**: The default Docker compose files now include a Python API and Vector Database for RAG (Retrieval-Augmented Generation). Read more about this in the [RAG API page](../features/rag_api.md)
diff --git a/docs/install/configuration/ai_endpoints.md b/docs/install/configuration/ai_endpoints.md
index 3b86fdb80..3169e8343 100644
--- a/docs/install/configuration/ai_endpoints.md
+++ b/docs/install/configuration/ai_endpoints.md
@@ -64,9 +64,11 @@ Some of the endpoints are marked as **Known,** which means they might have speci
       baseURL: "https://api.groq.com/openai/v1/"
       models:
         default: [
+          "llama3-70b-8192",
+          "llama3-8b-8192",
           "llama2-70b-4096",
           "mixtral-8x7b-32768",
-          "gemma-7b-it"
+          "gemma-7b-it",
           ]
         fetch: false
       titleConvo: true
@@ -374,3 +376,31 @@ Some of the endpoints are marked as **Known,** which means they might have speci
       forcePrompt: false
       modelDisplayLabel: "Ollama"
 ```
+
+!!! tip "Ollama -> llama3"
+    
+    To prevent the behavior where llama3 does not stop generating, add this `addParams` block to the config:
+    
+    ```yaml
+    - name: "Ollama"
+      apiKey: "ollama"
+      baseURL: "http://host.docker.internal:11434/v1/" 
+      models:
+        default: [
+          "llama3"
+          ]
+        fetch: false # fetching list of models is not supported
+      titleConvo: true
+      titleModel: "llama3"
+      summarize: false
+      summaryModel: "llama3"
+      forcePrompt: false
+      modelDisplayLabel: "Ollama"
+      addParams:
+            "stop": [
+              "<|start_header_id|>",
+              "<|end_header_id|>",
+              "<|eot_id|>",
+              "<|reserved_special_token"
+            ]
+    ```
\ No newline at end of file
diff --git a/librechat.example.yaml b/librechat.example.yaml
index 00f0f9fb5..6a119c01c 100644
--- a/librechat.example.yaml
+++ b/librechat.example.yaml
@@ -50,7 +50,13 @@ endpoints:
       apiKey: '${GROQ_API_KEY}'
       baseURL: 'https://api.groq.com/openai/v1/'
       models:
-        default: ['llama2-70b-4096', 'mixtral-8x7b-32768', 'gemma-7b-it']
+        default: [
+          "llama3-70b-8192",
+          "llama3-8b-8192",
+          "llama2-70b-4096",
+          "mixtral-8x7b-32768",
+          "gemma-7b-it",
+          ]
         fetch: false
       titleConvo: true
       titleModel: 'mixtral-8x7b-32768'