Added to jsonl

.gitattributes

@@ -60,3 +60,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 sharegpt_format.json filter=lfs diff=lfs merge=lfs -text
 chatml_format.jsonl filter=lfs diff=lfs merge=lfs -text
 alpaca_format.json filter=lfs diff=lfs merge=lfs -text
+french_classic_conversations.jsonl filter=lfs diff=lfs merge=lfs -text

convert_to_jsonl.py

@@ -0,0 +1,75 @@
+"""
+Convert HuggingFace dataset Volko76/french-classic-conversations to JSONL format
+with messages structure including system prompt.
+"""
+
+from datasets import load_dataset
+import json
+import os
+
+# Configuration
+SYSTEM_MESSAGE = "You are a helpful assistant."
+OUTPUT_FILE = "french_classic_conversations.jsonl"
+
+def main():
+    # Load the dataset from HuggingFace
+    print("Loading dataset from HuggingFace...")
+    dataset = load_dataset("Volko76/french-classic-conversations")
+    print(f"Dataset loaded: {len(dataset['train'])} rows")
+    print(f"Columns: {dataset['train'].column_names}")
+    
+    # Check the structure of the first row
+    sample = dataset['train'][0]
+    print("\nSample row structure:")
+    print(str(sample)[:1000])
+    
+    # Convert to JSONL format with system message
+    print(f"\nConverting to JSONL format...")
+    
+    with open(OUTPUT_FILE, 'w', encoding='utf-8') as f:
+        for row in dataset['train']:
+            # Get the conversations from the row - it might be a JSON string
+            conversations = row['conversations']
+            if isinstance(conversations, str):
+                conversations = json.loads(conversations)
+            
+            # Create the messages list with system prompt first
+            messages = [{"role": "system", "content": SYSTEM_MESSAGE}]
+            
+            # Add the conversation messages
+            for msg in conversations:
+                messages.append({
+                    "role": msg['role'],
+                    "content": msg['content']
+                })
+            
+            # Write as JSONL
+            json_line = json.dumps({"messages": messages}, ensure_ascii=False)
+            f.write(json_line + '\n')
+    
+    print(f"Conversion complete! Output saved to: {OUTPUT_FILE}")
+    
+    # Verify the output - read and display first few lines
+    print("\n" + "="*60)
+    print("First 2 entries from the output file:\n")
+    with open(OUTPUT_FILE, 'r', encoding='utf-8') as f:
+        for i, line in enumerate(f):
+            if i >= 2:
+                break
+            data = json.loads(line)
+            print(f"Entry {i+1}:")
+            print(json.dumps(data, indent=2, ensure_ascii=False)[:1000])
+            print("\n" + "-"*40 + "\n")
+    
+    # Count total entries and file size
+    with open(OUTPUT_FILE, 'r', encoding='utf-8') as f:
+        total_lines = sum(1 for _ in f)
+    
+    file_size = os.path.getsize(OUTPUT_FILE) / (1024 * 1024)  # MB
+    
+    print("="*60)
+    print(f"Total conversations: {total_lines}")
+    print(f"File size: {file_size:.2f} MB")
+
+if __name__ == "__main__":
+    main()

french_classic_conversations.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7ba69acf9bc82296b9847036148189b80d4dfa9aeb1baa7d0a76c3680c33aef6
+size 130711044

to_jsonl.ipynb

@@ -0,0 +1,132 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "id": "eb317b63",
+   "metadata": {},
+   "source": [
+    "# Convert French Classic Conversations to JSONL Format\n",
+    "\n",
+    "This notebook converts the HuggingFace dataset `Volko76/french-classic-conversations` to a JSONL file with the messages format:\n",
+    "```json\n",
+    "{\n",
+    "  \"messages\": [\n",
+    "    { \"role\": \"system\", \"content\": \"You are a helpful assistant.\" },\n",
+    "    { \"role\": \"user\", \"content\": \"...\" },\n",
+    "    { \"role\": \"assistant\", \"content\": \"...\" }\n",
+    "  ]\n",
+    "}\n",
+    "```"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "631ee57f",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from datasets import load_dataset\n",
+    "import json\n",
+    "\n",
+    "# Load the dataset from HuggingFace\n",
+    "dataset = load_dataset(\"Volko76/french-classic-conversations\")\n",
+    "print(f\"Dataset loaded: {len(dataset['train'])} rows\")\n",
+    "print(f\"Columns: {dataset['train'].column_names}\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "7382896f",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Check the structure of the first row\n",
+    "sample = dataset['train'][0]\n",
+    "print(\"Sample row:\")\n",
+    "print(json.dumps(sample, indent=2, ensure_ascii=False)[:1000])"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "11ed5944",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Convert to JSONL format with system message\n",
+    "SYSTEM_MESSAGE = \"You are a helpful assistant.\"\n",
+    "output_file = \"french_classic_conversations.jsonl\"\n",
+    "\n",
+    "with open(output_file, 'w', encoding='utf-8') as f:\n",
+    "    for row in dataset['train']:\n",
+    "        # Get the conversations from the row\n",
+    "        conversations = row['conversations']\n",
+    "        \n",
+    "        # Create the messages list with system prompt first\n",
+    "        messages = [{\"role\": \"system\", \"content\": SYSTEM_MESSAGE}]\n",
+    "        \n",
+    "        # Add the conversation messages\n",
+    "        for msg in conversations:\n",
+    "            messages.append({\n",
+    "                \"role\": msg['role'],\n",
+    "                \"content\": msg['content']\n",
+    "            })\n",
+    "        \n",
+    "        # Write as JSONL\n",
+    "        json_line = json.dumps({\"messages\": messages}, ensure_ascii=False)\n",
+    "        f.write(json_line + '\\n')\n",
+    "\n",
+    "print(f\"Conversion complete! Output saved to: {output_file}\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "b2db3191",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Verify the output - read and display first few lines\n",
+    "print(\"First 3 entries from the output file:\\n\")\n",
+    "with open(output_file, 'r', encoding='utf-8') as f:\n",
+    "    for i, line in enumerate(f):\n",
+    "        if i >= 3:\n",
+    "            break\n",
+    "        data = json.loads(line)\n",
+    "        print(f\"Entry {i+1}:\")\n",
+    "        print(json.dumps(data, indent=2, ensure_ascii=False)[:800])\n",
+    "        print(\"\\n\" + \"=\"*50 + \"\\n\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "25aa1a17",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Count total entries and file size\n",
+    "import os\n",
+    "\n",
+    "with open(output_file, 'r', encoding='utf-8') as f:\n",
+    "    total_lines = sum(1 for _ in f)\n",
+    "\n",
+    "file_size = os.path.getsize(output_file) / (1024 * 1024)  # MB\n",
+    "\n",
+    "print(f\"Total conversations: {total_lines}\")\n",
+    "print(f\"File size: {file_size:.2f} MB\")"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3 (ipykernel)",
+   "language": "python",
+   "name": "python3"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}