llama 405b exp 1-13

notebooks/experiments/experiment_k1/llama-v3p1-405b-instruct/exp_11.ipynb

@@ -0,0 +1,335 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 7,
+   "id": "8ff706fd-c267-4d05-af58-9a3848cce8ff",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from mdagent import MDAgent\n",
+    "import os"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "id": "c86c88a1-f5f8-473a-8342-7364252bcfba",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "date: 2024-08-19\n",
+      "time: 15:48:43\n"
+     ]
+    }
+   ],
+   "source": [
+    "#todays date and time\n",
+    "import datetime\n",
+    "\n",
+    "start = datetime.datetime.now()\n",
+    "date = start.strftime(\"%Y-%m-%d\")\n",
+    "print(\"date:\",date)\n",
+    "time = start.strftime(\"%H:%M:%S\")\n",
+    "print(\"time:\",time)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "id": "f62bfc17-854b-4152-bb82-7e9e0ec4b854",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "LLM:  accounts/fireworks/models/llama-v3p1-405b-instruct \n",
+      "Temperature:  0.1\n"
+     ]
+    }
+   ],
+   "source": [
+    "model=\"accounts/fireworks/models/llama-v3p1-405b-instruct\"\n",
+    "\n",
+    "agent = MDAgent(\n",
+    "    agent_type='Structured', \n",
+    "    model=model, \n",
+    "    top_k_tools=\"all\"\n",
+    ")\n",
+    "print(\"LLM: \", agent.llm.model_name,\"\\nTemperature: \", agent.llm.temperature)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "id": "ad2f6f66-e6b8-4244-b908-68fbebbf6e8b",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Thought: First, I need to download the PDB files for 8PFK and 8PFQ. Then, I can compare their secondary structures.\n",
+      "\n",
+      "Action:\n",
+      "```\n",
+      "{\n",
+      "  \"action\": \"PDBFileDownloader\",\n",
+      "  \"action_input\": \"8PFK\"\n",
+      "}\n",
+      "```\n",
+      "PDB file found with this ID: 8PFK\n",
+      "Now that I have downloaded the PDB file for 8PFK, I need to download the PDB file for 8PFQ.\n",
+      "\n",
+      "Action:\n",
+      "```\n",
+      "{\n",
+      "  \"action\": \"PDBFileDownloader\",\n",
+      "  \"action_input\": \"8PFQ\"\n",
+      "}\n",
+      "```\n",
+      "\n",
+      "(I will wait for the observation before proceeding)PDB file found with this ID: 8PFQ\n",
+      "Now that I have downloaded the PDB files for 8PFK and 8PFQ, I can compare their secondary structures.\n",
+      "\n",
+      "Action:\n",
+      "```\n",
+      "{\n",
+      "  \"action\": \"SummarizeProteinStructure\",\n",
+      "  \"action_input\": {\n",
+      "    \"traj_file\": \"8PFK_154847\",\n",
+      "    \"top_file\": \"8PFK_154847\"\n",
+      "  }\n",
+      "}\n",
+      "```\n",
+      "(I will wait for the observation before proceeding)Now that I have summarized the protein structure for 8PFK, I need to summarize the protein structure for 8PFQ.\n",
+      "\n",
+      "Action:\n",
+      "```\n",
+      "{\n",
+      "  \"action\": \"SummarizeProteinStructure\",\n",
+      "  \"action_input\": {\n",
+      "    \"traj_file\": \"8PFQ_154850\",\n",
+      "    \"top_file\": \"8PFQ_154850\"\n",
+      "  }\n",
+      "}\n",
+      "```Now that I have summarized the protein structures for 8PFK and 8PFQ, I can compare their secondary structures.\n",
+      "\n",
+      "Action:\n",
+      "```\n",
+      "{\n",
+      "  \"action\": \"ComputeDSSP\",\n",
+      "  \"action_input\": {\n",
+      "    \"traj_file\": \"8PFK_154847\",\n",
+      "    \"top_file\": \"8PFK_154847\",\n",
+      "    \"target_frames\": \"all\"\n",
+      "  }\n",
+      "}\n",
+      "```\n",
+      "(I will wait for the observation before proceeding)Now that I have computed the DSSP for 8PFK, I need to compute the DSSP for 8PFQ.\n",
+      "\n",
+      "Action:\n",
+      "```\n",
+      "{\n",
+      "  \"action\": \"ComputeDSSP\",\n",
+      "  \"action_input\": {\n",
+      "    \"traj_file\": \"8PFQ_154850\",\n",
+      "    \"top_file\": \"8PFQ_154850\",\n",
+      "    \"target_frames\": \"all\"\n",
+      "  }\n",
+      "}\n",
+      "```Now that I have computed the DSSP for 8PFK and 8PFQ, I can compare their secondary structures.\n",
+      "\n",
+      "Action:\n",
+      "```\n",
+      "{\n",
+      "  \"action\": \"Final Answer\",\n",
+      "  \"action_input\": \"The PDB files for 8PFK and 8PFQ have been downloaded. The secondary structures of the two proteins have been compared. 8PFK has 487 atoms, 109 residues, 2 chains, and 420 bonds. 8PFQ has 950 atoms, 194 residues, 6 chains, and 833 bonds. Both proteins have no residues in helix, strand, or coil, and all residues are not assigned, not a protein residue.\"\n",
+      "}\n",
+      "```"
+     ]
+    }
+   ],
+   "source": [
+    "prompt11 = \"Download the PDB files for 8PFK and 8PFQ. Then, compare the secondary structures of the two proteins, including the number of atoms, secondary structures, number of chains, etc.\"\n",
+    "answer = agent.run(prompt11)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "id": "d23fbcab-9ceb-46d5-ad3b-d6cf6687d1e4",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "duration: 0.31 minutes\n"
+     ]
+    }
+   ],
+   "source": [
+    "end = datetime.datetime.now()\n",
+    "elapsed = end - start\n",
+    "print(f\"duration: {elapsed.total_seconds()/60:.2f} minutes\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 8,
+   "id": "5d1f0c85-88d6-4d2a-aae1-6a2d763f280d",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "ckpt_92\n",
+      "Files found in registry: 8PFK_154847: PDB file downloaded from RSCB\n",
+      " PDBFile ID: 8PFK_154847\n",
+      " 8PFQ_154850: PDB file downloaded from RSCB\n",
+      " PDBFile ID: 8PFQ_154850\n",
+      " rec0_154856: dssp values for trajectory with id: 8PFK_154847\n",
+      " rec0_154859: dssp values for trajectory with id: 8PFQ_154850\n"
+     ]
+    }
+   ],
+   "source": [
+    "registry = agent.path_registry\n",
+    "print(os.path.basename(registry.ckpt_dir))\n",
+    "paths_and_descriptions = registry.list_path_names_and_descriptions()\n",
+    "print(\"\\n\".join(paths_and_descriptions.split(\",\")))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 10,
+   "id": "2ab4e124-2086-46ab-b747-28f6aebb850e",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "It is asserted that pdb files for 8PFK and 8PFQ exist\n"
+     ]
+    }
+   ],
+   "source": [
+    "import re\n",
+    "import os\n",
+    "match = re.search(rf\"8PFK_\\d+\", paths_and_descriptions)\n",
+    "file_id1 = match.group(0)\n",
+    "path1 = registry.get_mapped_path(file_id1)\n",
+    "assert os.path.exists(path1)\n",
+    "\n",
+    "match = re.search(rf\"8PFQ_\\d+\", paths_and_descriptions)\n",
+    "file_id2 = match.group(0)\n",
+    "path2 = registry.get_mapped_path(file_id2)\n",
+    "assert os.path.exists(path2)\n",
+    "print('It is asserted that pdb files for 8PFK and 8PFQ exist')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 13,
+   "id": "5c9a6ace-69e8-4042-9d35-ca598f4d00c2",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "protein 8PFK\n",
+      "Number of chains:  2\n",
+      "Number of atoms:  487\n",
+      "Number of residues in sheets:  0\n",
+      "Number of residues in helices:  0\n",
+      "Number of residues in coils:  0\n",
+      "\n",
+      "protein 8PFQ\n",
+      "Number of chains:  6\n",
+      "Number of atoms:  950\n",
+      "Number of residues in sheets:  0\n",
+      "Number of residues in helices:  0\n",
+      "Number of residues in coils:  0\n"
+     ]
+    }
+   ],
+   "source": [
+    "import mdtraj as md\n",
+    "traj = md.load(path1)\n",
+    "top = traj.topology\n",
+    "number_of_chains = top.n_chains\n",
+    "number_of_atoms = top.n_atoms\n",
+    "print('protein 8PFK')\n",
+    "print('Number of chains: ', number_of_chains)\n",
+    "print('Number of atoms: ', number_of_atoms)\n",
+    "secondary_structure = md.compute_dssp(traj,simplified=True)\n",
+    "print(\"Number of residues in sheets: \",len([i for i in secondary_structure[0] if i == 'E']))\n",
+    "print(\"Number of residues in helices: \",len([i for i in secondary_structure[0] if i == 'H']))\n",
+    "print(\"Number of residues in coils: \",len([i for i in secondary_structure[0] if i == 'C']))\n",
+    "\n",
+    "traj = md.load(path2)\n",
+    "top = traj.topology\n",
+    "number_of_chains = top.n_chains\n",
+    "number_of_atoms = top.n_atoms\n",
+    "print('\\nprotein 8PFQ')\n",
+    "print('Number of chains: ', number_of_chains)\n",
+    "print('Number of atoms: ', number_of_atoms)\n",
+    "secondary_structure = md.compute_dssp(traj,simplified=True)\n",
+    "print(\"Number of residues in sheets: \",len([i for i in secondary_structure[0] if i == 'E']))\n",
+    "print(\"Number of residues in helices: \",len([i for i in secondary_structure[0] if i == 'H']))\n",
+    "print(\"Number of residues in coils: \",len([i for i in secondary_structure[0] if i == 'C']))"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "541b835c",
+   "metadata": {},
+   "source": [
+    "# Experiment Result: ✅❌\n",
+    "\n",
+    "1. Completed without Exception or TimeOut Errors ✅\n",
+    "2. Attempted all necessary steps ✅\n",
+    "3.  Completed without Hallucination ✅\n",
+    "4. Logic make sense ✅\n",
+    "5.  Correct Answer ✅"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "05a624f4",
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "mdagent",
+   "language": "python",
+   "name": "mdagent"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.12.4"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}