From 3789381c6cbd0b495c31a3b6c5236d57d4b7eb79 Mon Sep 17 00:00:00 2001
From: Karl-Johan Alm <karljohan-alm@garage.co.jp>
Date: Thu, 21 Nov 2024 13:01:29 +0900
Subject: [PATCH] added substr derivations with one example

---
 public/script.js                 |  2 +-
 public/scripts/chat-cemplates.js | 30 ++++++++++++++++++++----------
 2 files changed, 21 insertions(+), 11 deletions(-)

diff --git a/public/script.js b/public/script.js
index be298e0a3..ced5f614e 100644
--- a/public/script.js
+++ b/public/script.js
@@ -1252,7 +1252,7 @@ async function getStatusTextgen() {
             const data = await response.json();
             if (data) {
                 const { chat_template, chat_template_hash } = data;
-                console.log(`${wantsContextDerivation} ${wantsInstructDerivation} We have chat template ${chat_template.split('\n')[0]}...`);
+                console.log(`We have chat template ${chat_template.split('\n')[0]}...`);
                 const templates = await deriveTemplatesFromChatTemplate(chat_template, chat_template_hash);
                 if (templates) {
                     const { context, instruct } = templates;
diff --git a/public/scripts/chat-cemplates.js b/public/scripts/chat-cemplates.js
index bcc136803..bad13e557 100644
--- a/public/scripts/chat-cemplates.js
+++ b/public/scripts/chat-cemplates.js
@@ -1,6 +1,6 @@
 // the hash can be obtained from command line e.g. via: MODEL=path_to_model; python -c "import json, hashlib, sys; print(hashlib.sha256(json.load(open('"$MODEL"/tokenizer_config.json'))['chat_template'].encode()).hexdigest())"
 // note that chat templates must be trimmed to match the llama.cpp metadata value
-const derivations = {
+const hash_derivations = {
     // Meta
     'e10ca381b1ccc5cf9db52e371f3b6651576caee0a630b452e2816b2d404d4b65':
         // Meta-Llama-3.1-8B-Instruct
@@ -51,17 +51,27 @@ const derivations = {
     ,
 };
 
+const substr_derivations = {
+    '<|im_start|>': 'ChatML', // qwen2.5, ...
+}
+
+const parse_derivation = derivation => (typeof derivation === 'string') ? {
+    'context': derivation,
+    'instruct': derivation,
+} : derivation;
+
 export async function deriveTemplatesFromChatTemplate(chat_template, hash) {
-    if (hash in derivations) {
-        const derivation = derivations[hash];
-        if (typeof derivation === 'string') {
-            return {
-                'context': derivation,
-                'instruct': derivation,
-            }
-        }
-        return derivation;
+    if (hash in hash_derivations) {
+        return parse_derivation(hash_derivations[hash]);
     }
+
+    // heuristics
+    for (const [substr, derivation] of Object.entries(substr_derivations) ) {
+        if (chat_template.includes(substr)) {
+            return parse_derivation(derivation);
+        }
+    }
+
     console.log(`Unknown chat template hash: ${hash} for [${chat_template}]`);
     return null;
 }