{"cells":[{"cell_type":"code","execution_count":1,"metadata":{"application/vnd.databricks.v1+cell":{"cellMetadata":{},"inputWidgets":{},"nuid":"0ea8b46b-839b-445b-8043-ccdf4e920ace","showTitle":false,"title":""},"id":"YLH80COBzi_F"},"outputs":[],"source":["%load_ext autoreload\n","%autoreload 2"]},{"cell_type":"code","execution_count":2,"metadata":{"id":"63B5exAuzq4M"},"outputs":[],"source":["from pathlib import Path\n","\n","try:\n"," from google.colab import drive\n"," drive.mount('/content/drive')\n"," workding_dir = \"/content/drive/MyDrive/logical-reasoning/\"\n","except ModuleNotFoundError:\n"," workding_dir = str(Path.cwd().parent)"]},{"cell_type":"code","execution_count":3,"metadata":{"executionInfo":{"elapsed":368,"status":"ok","timestamp":1719461634865,"user":{"displayName":"Donghao Huang","userId":"00463591218503521679"},"user_tz":-480},"id":"zFulf0bg0H-9","outputId":"debdd535-c828-40b9-efc0-8a180e5830dd"},"outputs":[{"name":"stdout","output_type":"stream","text":["workding dir: /Users/inflaton/code/engd/projects/logical-reasoning\n"]}],"source":["import os\n","import sys\n","\n","os.chdir(workding_dir)\n","sys.path.append(workding_dir)\n","print(\"workding dir:\", workding_dir)"]},{"cell_type":"code","execution_count":4,"metadata":{"application/vnd.databricks.v1+cell":{"cellMetadata":{},"inputWidgets":{},"nuid":"9f67ec60-2f24-411c-84eb-0dd664b44775","showTitle":false,"title":""},"colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"elapsed":589,"status":"ok","timestamp":1719462011879,"user":{"displayName":"Donghao Huang","userId":"00463591218503521679"},"user_tz":-480},"id":"DIUiweYYzi_I","outputId":"e16e9247-9077-4b0c-f8ea-17059f05a1c4"},"outputs":[{"name":"stdout","output_type":"stream","text":["loading env vars from: /Users/inflaton/code/engd/projects/logical-reasoning/.env\n"]},{"data":{"text/plain":["True"]},"execution_count":4,"metadata":{},"output_type":"execute_result"}],"source":["from dotenv import find_dotenv, load_dotenv\n","\n","found_dotenv = find_dotenv(\".env\")\n","\n","if len(found_dotenv) == 0:\n"," found_dotenv = find_dotenv(\".env.example\")\n","print(f\"loading env vars from: {found_dotenv}\")\n","load_dotenv(found_dotenv, override=True)"]},{"cell_type":"code","execution_count":5,"metadata":{},"outputs":[{"data":{"text/html":["
\n","\n","
\n"," \n"," \n"," | \n"," epoch | \n"," model | \n"," accuracy | \n"," precision | \n"," recall | \n"," f1 | \n","
\n"," \n"," \n"," \n"," 0 | \n"," 0 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat | \n"," 0.783667 | \n"," 0.766712 | \n"," 0.792917 | \n"," 0.767940 | \n","
\n"," \n"," 1 | \n"," 1 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... | \n"," 0.779667 | \n"," 0.793284 | \n"," 0.779667 | \n"," 0.780181 | \n","
\n"," \n"," 2 | \n"," 2 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... | \n"," 0.674000 | \n"," 0.785928 | \n"," 0.674000 | \n"," 0.709822 | \n","
\n"," \n"," 3 | \n"," 3 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... | \n"," 0.639333 | \n"," 0.776385 | \n"," 0.639333 | \n"," 0.688929 | \n","
\n"," \n"," 4 | \n"," 4 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... | \n"," 0.654667 | \n"," 0.761004 | \n"," 0.654667 | \n"," 0.692352 | \n","
\n"," \n"," 5 | \n"," 5 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... | \n"," 0.640000 | \n"," 0.756897 | \n"," 0.640000 | \n"," 0.671627 | \n","
\n"," \n"," 6 | \n"," 6 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... | \n"," 0.649000 | \n"," 0.758525 | \n"," 0.649000 | \n"," 0.684541 | \n","
\n"," \n"," 7 | \n"," 7 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... | \n"," 0.645000 | \n"," 0.749263 | \n"," 0.645000 | \n"," 0.667193 | \n","
\n"," \n"," 8 | \n"," 8 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... | \n"," 0.652667 | \n"," 0.760601 | \n"," 0.652667 | \n"," 0.689845 | \n","
\n"," \n"," 9 | \n"," 9 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... | \n"," 0.634333 | \n"," 0.752522 | \n"," 0.634333 | \n"," 0.670198 | \n","
\n"," \n"," 10 | \n"," 10 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... | \n"," 0.629333 | \n"," 0.745948 | \n"," 0.629333 | \n"," 0.666929 | \n","
\n"," \n","
\n","
"],"text/plain":[" epoch model accuracy \\\n","0 0 shenzhi-wang/Llama3-8B-Chinese-Chat 0.783667 \n","1 1 shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... 0.779667 \n","2 2 shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... 0.674000 \n","3 3 shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... 0.639333 \n","4 4 shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... 0.654667 \n","5 5 shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... 0.640000 \n","6 6 shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... 0.649000 \n","7 7 shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... 0.645000 \n","8 8 shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... 0.652667 \n","9 9 shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... 0.634333 \n","10 10 shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... 0.629333 \n","\n"," precision recall f1 \n","0 0.766712 0.792917 0.767940 \n","1 0.793284 0.779667 0.780181 \n","2 0.785928 0.674000 0.709822 \n","3 0.776385 0.639333 0.688929 \n","4 0.761004 0.654667 0.692352 \n","5 0.756897 0.640000 0.671627 \n","6 0.758525 0.649000 0.684541 \n","7 0.749263 0.645000 0.667193 \n","8 0.760601 0.652667 0.689845 \n","9 0.752522 0.634333 0.670198 \n","10 0.745948 0.629333 0.666929 "]},"execution_count":5,"metadata":{},"output_type":"execute_result"}],"source":["import pandas as pd\n","\n","df_p1_llama3 = pd.read_csv(\"results/mgtv-llama3_p1_full_metrics.csv\")\n","df_p1_llama3"]},{"cell_type":"code","execution_count":6,"metadata":{},"outputs":[{"data":{"text/html":["\n","\n","
\n"," \n"," \n"," | \n"," epoch | \n"," model | \n"," accuracy | \n"," precision | \n"," recall | \n"," f1 | \n","
\n"," \n"," \n"," \n"," 0 | \n"," 0 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat | \n"," 0.730333 | \n"," 0.771041 | \n"," 0.730333 | \n"," 0.746484 | \n","
\n"," \n"," 1 | \n"," 1 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... | \n"," 0.767333 | \n"," 0.795680 | \n"," 0.767333 | \n"," 0.770909 | \n","
\n"," \n"," 2 | \n"," 2 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... | \n"," 0.728667 | \n"," 0.820787 | \n"," 0.728667 | \n"," 0.763019 | \n","
\n"," \n"," 3 | \n"," 3 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... | \n"," 0.698000 | \n"," 0.795399 | \n"," 0.698000 | \n"," 0.733230 | \n","
\n"," \n"," 4 | \n"," 4 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... | \n"," 0.709333 | \n"," 0.793851 | \n"," 0.709333 | \n"," 0.740102 | \n","
\n"," \n"," 5 | \n"," 5 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... | \n"," 0.679333 | \n"," 0.791366 | \n"," 0.679333 | \n"," 0.720183 | \n","
\n"," \n"," 6 | \n"," 6 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... | \n"," 0.667000 | \n"," 0.787851 | \n"," 0.667000 | \n"," 0.710294 | \n","
\n"," \n"," 7 | \n"," 7 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... | \n"," 0.645667 | \n"," 0.764853 | \n"," 0.645667 | \n"," 0.680705 | \n","
\n"," \n"," 8 | \n"," 8 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... | \n"," 0.654667 | \n"," 0.769380 | \n"," 0.654667 | \n"," 0.687044 | \n","
\n"," \n"," 9 | \n"," 9 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... | \n"," 0.643000 | \n"," 0.766252 | \n"," 0.643000 | \n"," 0.678544 | \n","
\n"," \n"," 10 | \n"," 10 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... | \n"," 0.647667 | \n"," 0.766725 | \n"," 0.647667 | \n"," 0.684851 | \n","
\n"," \n","
\n","
"],"text/plain":[" epoch model accuracy \\\n","0 0 shenzhi-wang/Llama3-8B-Chinese-Chat 0.730333 \n","1 1 shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... 0.767333 \n","2 2 shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... 0.728667 \n","3 3 shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... 0.698000 \n","4 4 shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... 0.709333 \n","5 5 shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... 0.679333 \n","6 6 shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... 0.667000 \n","7 7 shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... 0.645667 \n","8 8 shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... 0.654667 \n","9 9 shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... 0.643000 \n","10 10 shenzhi-wang/Llama3-8B-Chinese-Chat_checkpoint... 0.647667 \n","\n"," precision recall f1 \n","0 0.771041 0.730333 0.746484 \n","1 0.795680 0.767333 0.770909 \n","2 0.820787 0.728667 0.763019 \n","3 0.795399 0.698000 0.733230 \n","4 0.793851 0.709333 0.740102 \n","5 0.791366 0.679333 0.720183 \n","6 0.787851 0.667000 0.710294 \n","7 0.764853 0.645667 0.680705 \n","8 0.769380 0.654667 0.687044 \n","9 0.766252 0.643000 0.678544 \n","10 0.766725 0.647667 0.684851 "]},"execution_count":6,"metadata":{},"output_type":"execute_result"}],"source":["df_p2_llama3 = pd.read_csv(\"results/mgtv-llama3_p2_full_metrics.csv\")\n","df_p2_llama3"]},{"cell_type":"code","execution_count":7,"metadata":{},"outputs":[{"data":{"text/html":["\n","\n","
\n"," \n"," \n"," | \n"," epoch | \n"," model | \n"," accuracy | \n"," precision | \n"," recall | \n"," f1 | \n","
\n"," \n"," \n"," \n"," 0 | \n"," 0 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat_torch.bflo... | \n"," 0.783667 | \n"," 0.766712 | \n"," 0.792917 | \n"," 0.767940 | \n","
\n"," \n"," 1 | \n"," 1 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat/checkpoint... | \n"," 0.770667 | \n"," 0.807275 | \n"," 0.770667 | \n"," 0.783572 | \n","
\n"," \n"," 2 | \n"," 2 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat/checkpoint... | \n"," 0.724000 | \n"," 0.811805 | \n"," 0.724000 | \n"," 0.756227 | \n","
\n"," \n"," 3 | \n"," 3 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat/checkpoint... | \n"," 0.675667 | \n"," 0.781176 | \n"," 0.675667 | \n"," 0.710846 | \n","
\n"," \n"," 4 | \n"," 4 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat/checkpoint... | \n"," 0.649667 | \n"," 0.779897 | \n"," 0.649667 | \n"," 0.693184 | \n","
\n"," \n","
\n","
"],"text/plain":[" epoch model accuracy \\\n","0 0 shenzhi-wang/Llama3-8B-Chinese-Chat_torch.bflo... 0.783667 \n","1 1 shenzhi-wang/Llama3-8B-Chinese-Chat/checkpoint... 0.770667 \n","2 2 shenzhi-wang/Llama3-8B-Chinese-Chat/checkpoint... 0.724000 \n","3 3 shenzhi-wang/Llama3-8B-Chinese-Chat/checkpoint... 0.675667 \n","4 4 shenzhi-wang/Llama3-8B-Chinese-Chat/checkpoint... 0.649667 \n","\n"," precision recall f1 \n","0 0.766712 0.792917 0.767940 \n","1 0.807275 0.770667 0.783572 \n","2 0.811805 0.724000 0.756227 \n","3 0.781176 0.675667 0.710846 \n","4 0.779897 0.649667 0.693184 "]},"execution_count":7,"metadata":{},"output_type":"execute_result"}],"source":["df_p1_llama3_r2 = pd.read_csv(\"results/mgtv-llama3_p1_r2_full_metrics.csv\")\n","df_p1_llama3_r2"]},{"cell_type":"code","execution_count":8,"metadata":{},"outputs":[{"data":{"text/html":["\n","\n","
\n"," \n"," \n"," | \n"," epoch | \n"," model | \n"," accuracy | \n"," precision | \n"," recall | \n"," f1 | \n","
\n"," \n"," \n"," \n"," 0 | \n"," 0 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat_torch.bflo... | \n"," 0.730000 | \n"," 0.770974 | \n"," 0.730000 | \n"," 0.746291 | \n","
\n"," \n"," 1 | \n"," 1 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat/checkpoint... | \n"," 0.718000 | \n"," 0.811309 | \n"," 0.718000 | \n"," 0.750106 | \n","
\n"," \n"," 2 | \n"," 2 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat/checkpoint... | \n"," 0.727333 | \n"," 0.802512 | \n"," 0.727333 | \n"," 0.754982 | \n","
\n"," \n"," 3 | \n"," 3 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat/checkpoint... | \n"," 0.688333 | \n"," 0.781617 | \n"," 0.688333 | \n"," 0.716763 | \n","
\n"," \n"," 4 | \n"," 4 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat/checkpoint... | \n"," 0.640667 | \n"," 0.763630 | \n"," 0.640667 | \n"," 0.680793 | \n","
\n"," \n","
\n","
"],"text/plain":[" epoch model accuracy \\\n","0 0 shenzhi-wang/Llama3-8B-Chinese-Chat_torch.bflo... 0.730000 \n","1 1 shenzhi-wang/Llama3-8B-Chinese-Chat/checkpoint... 0.718000 \n","2 2 shenzhi-wang/Llama3-8B-Chinese-Chat/checkpoint... 0.727333 \n","3 3 shenzhi-wang/Llama3-8B-Chinese-Chat/checkpoint... 0.688333 \n","4 4 shenzhi-wang/Llama3-8B-Chinese-Chat/checkpoint... 0.640667 \n","\n"," precision recall f1 \n","0 0.770974 0.730000 0.746291 \n","1 0.811309 0.718000 0.750106 \n","2 0.802512 0.727333 0.754982 \n","3 0.781617 0.688333 0.716763 \n","4 0.763630 0.640667 0.680793 "]},"execution_count":8,"metadata":{},"output_type":"execute_result"}],"source":["df_p2_llama3_r2 = pd.read_csv(\"results/mgtv-llama3_p2_r2_full_metrics.csv\")\n","df_p2_llama3_r2"]},{"cell_type":"code","execution_count":9,"metadata":{},"outputs":[{"data":{"text/html":["\n","\n","
\n"," \n"," \n"," | \n"," epoch | \n"," model | \n"," accuracy | \n"," precision | \n"," recall | \n"," f1 | \n","
\n"," \n"," \n"," \n"," 0 | \n"," 0.0 | \n"," hfl/llama-3-chinese-8b-instruct-v3_torch.bfloa... | \n"," 0.456333 | \n"," 0.674450 | \n"," 0.456333 | \n"," 0.530122 | \n","
\n"," \n"," 1 | \n"," 0.2 | \n"," hfl/llama-3-chinese-8b-instruct-v3/checkpoint-... | \n"," 0.640667 | \n"," 0.765241 | \n"," 0.640667 | \n"," 0.686507 | \n","
\n"," \n"," 2 | \n"," 0.4 | \n"," hfl/llama-3-chinese-8b-instruct-v3/checkpoint-... | \n"," 0.722333 | \n"," 0.761495 | \n"," 0.722333 | \n"," 0.729669 | \n","
\n"," \n"," 3 | \n"," 0.6 | \n"," hfl/llama-3-chinese-8b-instruct-v3/checkpoint-... | \n"," 0.625667 | \n"," 0.769429 | \n"," 0.625667 | \n"," 0.674742 | \n","
\n"," \n"," 4 | \n"," 0.8 | \n"," hfl/llama-3-chinese-8b-instruct-v3/checkpoint-... | \n"," 0.717333 | \n"," 0.774693 | \n"," 0.717333 | \n"," 0.739105 | \n","
\n"," \n"," 5 | \n"," 1.0 | \n"," hfl/llama-3-chinese-8b-instruct-v3/checkpoint-... | \n"," 0.688000 | \n"," 0.767848 | \n"," 0.688000 | \n"," 0.718197 | \n","
\n"," \n","
\n","
"],"text/plain":[" epoch model accuracy \\\n","0 0.0 hfl/llama-3-chinese-8b-instruct-v3_torch.bfloa... 0.456333 \n","1 0.2 hfl/llama-3-chinese-8b-instruct-v3/checkpoint-... 0.640667 \n","2 0.4 hfl/llama-3-chinese-8b-instruct-v3/checkpoint-... 0.722333 \n","3 0.6 hfl/llama-3-chinese-8b-instruct-v3/checkpoint-... 0.625667 \n","4 0.8 hfl/llama-3-chinese-8b-instruct-v3/checkpoint-... 0.717333 \n","5 1.0 hfl/llama-3-chinese-8b-instruct-v3/checkpoint-... 0.688000 \n","\n"," precision recall f1 \n","0 0.674450 0.456333 0.530122 \n","1 0.765241 0.640667 0.686507 \n","2 0.761495 0.722333 0.729669 \n","3 0.769429 0.625667 0.674742 \n","4 0.774693 0.717333 0.739105 \n","5 0.767848 0.688000 0.718197 "]},"execution_count":9,"metadata":{},"output_type":"execute_result"}],"source":["df_p1_llama3_r3 = pd.read_csv(\"results/mgtv-llama3_p1_r3_full_metrics.csv\")\n","df_p1_llama3_r3"]},{"cell_type":"code","execution_count":10,"metadata":{},"outputs":[{"data":{"text/html":["\n","\n","
\n"," \n"," \n"," | \n"," epoch | \n"," model | \n"," accuracy | \n"," precision | \n"," recall | \n"," f1 | \n","
\n"," \n"," \n"," \n"," 0 | \n"," 0.0 | \n"," hfl/llama-3-chinese-8b-instruct-v3_torch.bfloa... | \n"," 0.250667 | \n"," 0.685242 | \n"," 0.250667 | \n"," 0.326364 | \n","
\n"," \n"," 1 | \n"," 0.2 | \n"," hfl/llama-3-chinese-8b-instruct-v3/checkpoint-... | \n"," 0.728333 | \n"," 0.772239 | \n"," 0.728333 | \n"," 0.742645 | \n","
\n"," \n"," 2 | \n"," 0.4 | \n"," hfl/llama-3-chinese-8b-instruct-v3/checkpoint-... | \n"," 0.741000 | \n"," 0.786830 | \n"," 0.741000 | \n"," 0.751406 | \n","
\n"," \n"," 3 | \n"," 0.6 | \n"," hfl/llama-3-chinese-8b-instruct-v3/checkpoint-... | \n"," 0.622333 | \n"," 0.777171 | \n"," 0.622333 | \n"," 0.676279 | \n","
\n"," \n"," 4 | \n"," 0.8 | \n"," hfl/llama-3-chinese-8b-instruct-v3/checkpoint-... | \n"," 0.700000 | \n"," 0.776797 | \n"," 0.700000 | \n"," 0.729848 | \n","
\n"," \n"," 5 | \n"," 1.0 | \n"," hfl/llama-3-chinese-8b-instruct-v3/checkpoint-... | \n"," 0.697000 | \n"," 0.787120 | \n"," 0.697000 | \n"," 0.730959 | \n","
\n"," \n","
\n","
"],"text/plain":[" epoch model accuracy \\\n","0 0.0 hfl/llama-3-chinese-8b-instruct-v3_torch.bfloa... 0.250667 \n","1 0.2 hfl/llama-3-chinese-8b-instruct-v3/checkpoint-... 0.728333 \n","2 0.4 hfl/llama-3-chinese-8b-instruct-v3/checkpoint-... 0.741000 \n","3 0.6 hfl/llama-3-chinese-8b-instruct-v3/checkpoint-... 0.622333 \n","4 0.8 hfl/llama-3-chinese-8b-instruct-v3/checkpoint-... 0.700000 \n","5 1.0 hfl/llama-3-chinese-8b-instruct-v3/checkpoint-... 0.697000 \n","\n"," precision recall f1 \n","0 0.685242 0.250667 0.326364 \n","1 0.772239 0.728333 0.742645 \n","2 0.786830 0.741000 0.751406 \n","3 0.777171 0.622333 0.676279 \n","4 0.776797 0.700000 0.729848 \n","5 0.787120 0.697000 0.730959 "]},"execution_count":10,"metadata":{},"output_type":"execute_result"}],"source":["df_p2_llama3_r3 = pd.read_csv(\"results/mgtv-llama3_p2_r3_full_metrics.csv\")\n","df_p2_llama3_r3"]},{"cell_type":"code","execution_count":11,"metadata":{},"outputs":[{"data":{"text/html":["\n","\n","
\n"," \n"," \n"," | \n"," epoch | \n"," model | \n"," accuracy | \n"," precision | \n"," recall | \n"," f1 | \n","
\n"," \n"," \n"," \n"," 0 | \n"," 0.0 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat_torch.bflo... | \n"," 0.780667 | \n"," 0.757362 | \n"," 0.789882 | \n"," 0.765242 | \n","
\n"," \n"," 1 | \n"," 0.2 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat/checkpoint... | \n"," 0.740667 | \n"," 0.796825 | \n"," 0.740667 | \n"," 0.753115 | \n","
\n"," \n"," 2 | \n"," 0.4 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat/checkpoint... | \n"," 0.778000 | \n"," 0.803096 | \n"," 0.778000 | \n"," 0.781519 | \n","
\n"," \n"," 3 | \n"," 0.6 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat/checkpoint... | \n"," 0.739000 | \n"," 0.802579 | \n"," 0.739000 | \n"," 0.761298 | \n","
\n"," \n"," 4 | \n"," 0.8 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat/checkpoint... | \n"," 0.749667 | \n"," 0.796758 | \n"," 0.749667 | \n"," 0.754738 | \n","
\n"," \n"," 5 | \n"," 1.0 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat/checkpoint... | \n"," 0.722667 | \n"," 0.807870 | \n"," 0.722667 | \n"," 0.753803 | \n","
\n"," \n"," 6 | \n"," 1.2 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat/checkpoint... | \n"," 0.699333 | \n"," 0.805342 | \n"," 0.699333 | \n"," 0.738849 | \n","
\n"," \n"," 7 | \n"," 1.4 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat/checkpoint... | \n"," 0.730667 | \n"," 0.797691 | \n"," 0.730667 | \n"," 0.752535 | \n","
\n"," \n"," 8 | \n"," 1.6 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat/checkpoint... | \n"," 0.712333 | \n"," 0.801750 | \n"," 0.712333 | \n"," 0.745198 | \n","
\n"," \n"," 9 | \n"," 1.8 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat/checkpoint... | \n"," 0.699000 | \n"," 0.809477 | \n"," 0.699000 | \n"," 0.739980 | \n","
\n"," \n"," 10 | \n"," 2.0 | \n"," shenzhi-wang/Llama3-8B-Chinese-Chat/checkpoint... | \n"," 0.721000 | \n"," 0.809777 | \n"," 0.721000 | \n"," 0.754225 | \n","
\n"," \n","
\n","
"],"text/plain":[" epoch model accuracy \\\n","0 0.0 shenzhi-wang/Llama3-8B-Chinese-Chat_torch.bflo... 0.780667 \n","1 0.2 shenzhi-wang/Llama3-8B-Chinese-Chat/checkpoint... 0.740667 \n","2 0.4 shenzhi-wang/Llama3-8B-Chinese-Chat/checkpoint... 0.778000 \n","3 0.6 shenzhi-wang/Llama3-8B-Chinese-Chat/checkpoint... 0.739000 \n","4 0.8 shenzhi-wang/Llama3-8B-Chinese-Chat/checkpoint... 0.749667 \n","5 1.0 shenzhi-wang/Llama3-8B-Chinese-Chat/checkpoint... 0.722667 \n","6 1.2 shenzhi-wang/Llama3-8B-Chinese-Chat/checkpoint... 0.699333 \n","7 1.4 shenzhi-wang/Llama3-8B-Chinese-Chat/checkpoint... 0.730667 \n","8 1.6 shenzhi-wang/Llama3-8B-Chinese-Chat/checkpoint... 0.712333 \n","9 1.8 shenzhi-wang/Llama3-8B-Chinese-Chat/checkpoint... 0.699000 \n","10 2.0 shenzhi-wang/Llama3-8B-Chinese-Chat/checkpoint... 0.721000 \n","\n"," precision recall f1 \n","0 0.757362 0.789882 0.765242 \n","1 0.796825 0.740667 0.753115 \n","2 0.803096 0.778000 0.781519 \n","3 0.802579 0.739000 0.761298 \n","4 0.796758 0.749667 0.754738 \n","5 0.807870 0.722667 0.753803 \n","6 0.805342 0.699333 0.738849 \n","7 0.797691 0.730667 0.752535 \n","8 0.801750 0.712333 0.745198 \n","9 0.809477 0.699000 0.739980 \n","10 0.809777 0.721000 0.754225 "]},"execution_count":11,"metadata":{},"output_type":"execute_result"}],"source":["df_p1_llama3_r4 = pd.read_csv(\"results/mgtv-llama3_p1_r4_full_metrics.csv\")\n","df_p1_llama3_r4"]},{"cell_type":"code","execution_count":12,"metadata":{},"outputs":[{"data":{"text/html":["\n","\n","
\n"," \n"," \n"," | \n"," epoch | \n"," model | \n"," accuracy | \n"," precision | \n"," recall | \n"," f1 | \n","
\n"," \n"," \n"," \n"," 0 | \n"," 0.000000 | \n"," meta-llama/Meta-Llama-3-8B-Instruct_torch.bflo... | \n"," 0.139000 | \n"," 0.574101 | \n"," 0.139000 | \n"," 0.186498 | \n","
\n"," \n"," 1 | \n"," 0.333333 | \n"," meta-llama/Meta-Llama-3-8B-Instruct/checkpoint... | \n"," 0.648667 | \n"," 0.652593 | \n"," 0.648667 | \n"," 0.631272 | \n","
\n"," \n"," 2 | \n"," 0.666667 | \n"," meta-llama/Meta-Llama-3-8B-Instruct/checkpoint... | \n"," 0.561000 | \n"," 0.689710 | \n"," 0.561000 | \n"," 0.608339 | \n","
\n"," \n"," 3 | \n"," 1.000000 | \n"," meta-llama/Meta-Llama-3-8B-Instruct/checkpoint... | \n"," 0.621000 | \n"," 0.686843 | \n"," 0.621000 | \n"," 0.641744 | \n","
\n"," \n","
\n","
"],"text/plain":[" epoch model accuracy \\\n","0 0.000000 meta-llama/Meta-Llama-3-8B-Instruct_torch.bflo... 0.139000 \n","1 0.333333 meta-llama/Meta-Llama-3-8B-Instruct/checkpoint... 0.648667 \n","2 0.666667 meta-llama/Meta-Llama-3-8B-Instruct/checkpoint... 0.561000 \n","3 1.000000 meta-llama/Meta-Llama-3-8B-Instruct/checkpoint... 0.621000 \n","\n"," precision recall f1 \n","0 0.574101 0.139000 0.186498 \n","1 0.652593 0.648667 0.631272 \n","2 0.689710 0.561000 0.608339 \n","3 0.686843 0.621000 0.641744 "]},"execution_count":12,"metadata":{},"output_type":"execute_result"}],"source":["df_p1_llama3_en = pd.read_csv(\"results/mgtv-llama3_p1_en_full_metrics.csv\")\n","df_p1_llama3_en"]},{"cell_type":"code","execution_count":13,"metadata":{},"outputs":[{"data":{"text/html":["\n","\n","
\n"," \n"," \n"," | \n"," epoch | \n"," model | \n"," accuracy | \n"," precision | \n"," recall | \n"," f1 | \n","
\n"," \n"," \n"," \n"," 0 | \n"," 0.000000 | \n"," meta-llama/Meta-Llama-3-8B-Instruct_torch.bflo... | \n"," 0.154667 | \n"," 0.521852 | \n"," 0.154667 | \n"," 0.176118 | \n","
\n"," \n"," 1 | \n"," 0.333333 | \n"," meta-llama/Meta-Llama-3-8B-Instruct/checkpoint... | \n"," 0.620333 | \n"," 0.663582 | \n"," 0.620333 | \n"," 0.636363 | \n","
\n"," \n"," 2 | \n"," 0.666667 | \n"," meta-llama/Meta-Llama-3-8B-Instruct/checkpoint... | \n"," 0.561333 | \n"," 0.700051 | \n"," 0.561333 | \n"," 0.611304 | \n","
\n"," \n"," 3 | \n"," 1.000000 | \n"," meta-llama/Meta-Llama-3-8B-Instruct/checkpoint... | \n"," 0.620333 | \n"," 0.681920 | \n"," 0.620333 | \n"," 0.640515 | \n","
\n"," \n","
\n","
"],"text/plain":[" epoch model accuracy \\\n","0 0.000000 meta-llama/Meta-Llama-3-8B-Instruct_torch.bflo... 0.154667 \n","1 0.333333 meta-llama/Meta-Llama-3-8B-Instruct/checkpoint... 0.620333 \n","2 0.666667 meta-llama/Meta-Llama-3-8B-Instruct/checkpoint... 0.561333 \n","3 1.000000 meta-llama/Meta-Llama-3-8B-Instruct/checkpoint... 0.620333 \n","\n"," precision recall f1 \n","0 0.521852 0.154667 0.176118 \n","1 0.663582 0.620333 0.636363 \n","2 0.700051 0.561333 0.611304 \n","3 0.681920 0.620333 0.640515 "]},"execution_count":13,"metadata":{},"output_type":"execute_result"}],"source":["df_p2_llama3_en = pd.read_csv(\"results/mgtv-llama3_p2_en_full_metrics.csv\")\n","df_p2_llama3_en"]},{"cell_type":"code","execution_count":14,"metadata":{},"outputs":[],"source":["def plot_results(df_p1, df_p2, best_p1, best_p2, color_p1=\"red\", color_p2=\"blue\", model_name=\"InternLM2.5_7b\"):\n"," sns.lineplot(\n"," x=\"epoch\",\n"," y=\"accuracy\",\n"," data=df_p1,\n"," ax=ax[0],\n"," color=color_p1,\n"," label=f\"{model_name}: P1\",\n"," )\n"," sns.lineplot(\n"," x=\"epoch\",\n"," y=\"accuracy\",\n"," data=df_p2,\n"," ax=ax[0],\n"," color=color_p2,\n"," label=f\"{model_name}: P2\",\n"," )\n"," sns.scatterplot(\n"," x=\"epoch\", y=\"accuracy\", data=best_p1, ax=ax[0], color=color_p1, s=50\n"," )\n"," sns.scatterplot(\n"," x=\"epoch\", y=\"accuracy\", data=best_p2, ax=ax[0], color=color_p2, s=50\n"," )\n","\n"," sns.lineplot(\n"," x=\"epoch\",\n"," y=\"f1\",\n"," data=df_p1,\n"," ax=ax[1],\n"," color=color_p1,\n"," label=f\"{model_name}: P1\",\n"," )\n"," sns.lineplot(\n"," x=\"epoch\",\n"," y=\"f1\",\n"," data=df_p2,\n"," ax=ax[1],\n"," color=color_p2,\n"," label=f\"{model_name}: P2\",\n"," )\n"," sns.scatterplot(x=\"epoch\", y=\"f1\", data=best_p1, ax=ax[1], color=color_p1, s=50)\n"," sns.scatterplot(x=\"epoch\", y=\"f1\", data=best_p2, ax=ax[1], color=color_p2, s=50)"]},{"cell_type":"code","execution_count":15,"metadata":{},"outputs":[],"source":["# plot the results\n","import matplotlib.pyplot as plt\n","import seaborn as sns\n","import matplotlib.ticker as ticker\n","\n","def plot_model_results(model_name, df_p1, df_p2, ax):\n"," print(f\"Model: {model_name}\")\n"," sns.set_theme(style=\"whitegrid\")\n","\n"," # print the best results\n"," best_p1 = df_p1[df_p1[\"accuracy\"] == df_p1[\"accuracy\"].max()]\n"," best_p2 = df_p2[df_p2[\"accuracy\"] == df_p2[\"accuracy\"].max()]\n","\n"," print(\"Best P1 accuracy:\")\n"," print(best_p1[\"accuracy\"].values[0])\n"," print(\"Best P2 accuracy:\")\n"," print(best_p2[\"accuracy\"].values[0])\n","\n"," plot_results(df_p1, df_p2, best_p1, best_p2, model_name=model_name)\n","\n"," for a in ax:\n"," for line_index, line in enumerate(a.lines):\n"," # Get the data\n"," line_color = line.get_color()\n"," xdata, ydata = line.get_data()\n"," for index in range(xdata.size):\n"," a.annotate( # Use 'a' instead of 'ax' to refer to the current subplot\n"," f\"{ydata[index]:.3f}\",\n"," xy=(xdata[index], ydata[index]),\n"," textcoords=\"offset points\",\n"," xytext=(\n"," 0,\n"," 1,\n"," # -10 if line_index % 2 == 0 else 10,\n"," ), # Adjusted for better visibility\n"," ha=\"center\",\n"," color=line_color,\n"," )\n","\n"," ax[0].set_title(\"Accuracy\")\n"," ax[1].set_title(\"F1\")\n","\n"," # After plotting your data and before plt.show(), add these lines\n"," ax[0].xaxis.set_major_locator(ticker.MaxNLocator(integer=True))\n"," ax[1].xaxis.set_major_locator(ticker.MaxNLocator(integer=True))\n","\n"," plt.show()"]},{"cell_type":"code","execution_count":16,"metadata":{},"outputs":[{"name":"stdout","output_type":"stream","text":["Model: Llama3-8B-Chinese\n","Best P1 accuracy:\n","0.7836666666666666\n","Best P2 accuracy:\n","0.7673333333333333\n"]},{"data":{"image/png":"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","text/plain":["