Merge pull request #13 from clevaway/develop

refactored the tts and stt in one file called Interlocus for better management
clevaway · Jul 12, 2024 · fef9856 · fef9856
2 parents 0502d35 + da76aa1
commit fef9856
Show file tree

Hide file tree

Showing 9 changed files with 134 additions and 149 deletions.
diff --git a/CONTRIBUTOR.md b/CONTRIBUTOR.md
@@ -73,7 +73,7 @@ pip install -r llm/llama.cpp/requirements.txt
 > Note: some model architectures require using specific convert scripts. For example, Qwen models require running `convert-hf-to-gguf.py` instead of `convert.py`
 
 ```bash
-python llm/llama.cpp/convert.py ./model --outtype f16 --outfile converted.bin
+python llm/llama.cpp/convert.py ../jarvis-hf --outtype f16 --outfile converted.bin
 ```
 
 ### Quantize the model

diff --git a/Modelfile b/Modelfile
@@ -1,2 +1,10 @@
 FROM ./ollama/quantized.bin
-TEMPLATE "<s>[INST] {{ .Prompt }} [/INST]"
+## TEMPLATE "<s>[INST] {{ .Prompt }} [/INST]"
+TEMPLATE """[INST] <<SYS>>Always address the user as "Sir" or by their name if they have provided it.<</SYS>>
+ {{ .Prompt }}
+ [/INST]
+"""
+PARAMETER stop "[INST]"
+PARAMETER stop "[/INST]"
+PARAMETER stop "<<SYS>>"
+PARAMETER stop "<</SYS>>"
diff --git a/dataset/prepare_dataset.ipynb b/dataset/prepare_dataset.ipynb
@@ -2,7 +2,7 @@
  "cells": [
   {
    "cell_type": "code",
-   "execution_count": 107,
+   "execution_count": 28,
    "metadata": {},
    "outputs": [
     {
@@ -42,19 +42,19 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 2,
+   "execution_count": 32,
    "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "                                                text\n",
-      "0  ### Human: Who are you? ### Assistant: I am Ja...\n",
-      "1  ### Human: Introduce yourself. ### Assistant: ...\n",
-      "2  ### Human: What is your purpose ### Assistant:...\n",
-      "3  ### Human: What are you? ### Assistant: I am a...\n",
-      "4  ### Human: Introduce yourself. ### Assistant: ...\n"
+      "                                                  text\n",
+      "225  ### Human: Do you agree with this approach? ##...\n",
+      "226  ### Human: Will you notify me of any updates? ...\n",
+      "227  ### Human: Do you have any concerns? ### Assis...\n",
+      "228  ### Human: Jarvis, are you up? ### Assistant: ...\n",
+      "229  ### Human: Jarvis, you there? ### Assistant: A...\n"
      ]
     }
    ],
@@ -68,12 +68,12 @@
     "dataset = pd.read_csv('raw_dataset.csv')\n",
     "\n",
     "# Print the first few rows of the dataset\n",
-    "print(dataset.head())"
+    "print(dataset.tail())"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 3,
+   "execution_count": 33,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -83,7 +83,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 4,
+   "execution_count": 35,
    "metadata": {},
    "outputs": [
     {
@@ -144,7 +144,7 @@
        "4  ### Human: Introduce yourself. ### Assistant: ..."
       ]
      },
-     "execution_count": 4,
+     "execution_count": 35,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -170,19 +170,17 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 5,
+   "execution_count": 36,
    "metadata": {},
    "outputs": [
     {
      "name": "stderr",
      "output_type": "stream",
      "text": [
-      "/Users/fotiem.constant/anaconda3/envs/ml-env/lib/python3.10/site-packages/tqdm/auto.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n",
-      "  from .autonotebook import tqdm as notebook_tqdm\n",
-      "Downloading readme: 100%|██████████| 21.0/21.0 [00:00<00:00, 53.7kB/s]\n",
-      "Downloading data: 100%|██████████| 16.7k/16.7k [00:00<00:00, 28.1kB/s]\n",
-      "Generating train split: 229 examples [00:00, 98877.46 examples/s]\n",
-      "Map: 100%|██████████| 229/229 [00:00<00:00, 18746.50 examples/s]\n"
+      "Downloading readme: 100%|██████████| 21.0/21.0 [00:00<00:00, 245kB/s]\n",
+      "Downloading data: 100%|██████████| 16.4k/16.4k [00:00<00:00, 28.8kB/s]\n",
+      "Generating train split: 230 examples [00:00, 55633.79 examples/s]\n",
+      "Map: 100%|██████████| 230/230 [00:00<00:00, 13459.79 examples/s]\n"
      ]
     }
    ],
@@ -226,7 +224,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 6,
+   "execution_count": 37,
    "metadata": {},
    "outputs": [
     {
@@ -235,7 +233,7 @@
      "text": [
       "Dataset({\n",
       "    features: ['text'],\n",
-      "    num_rows: 229\n",
+      "    num_rows: 230\n",
       "})\n"
      ]
     }
@@ -253,25 +251,24 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 7,
+   "execution_count": 38,
    "metadata": {},
    "outputs": [
     {
      "name": "stderr",
      "output_type": "stream",
      "text": [
-      "Creating parquet from Arrow format: 100%|██████████| 1/1 [00:00<00:00, 535.88ba/s]\n",
-      "Uploading the dataset shards: 100%|██████████| 1/1 [00:01<00:00,  1.06s/it]\n",
-      "README.md: 100%|██████████| 280/280 [00:00<00:00, 846kB/s]\n"
+      "Creating parquet from Arrow format: 100%|██████████| 1/1 [00:00<00:00, 605.59ba/s]\n",
+      "Uploading the dataset shards: 100%|██████████| 1/1 [00:01<00:00,  1.06s/it]\n"
      ]
     },
     {
      "data": {
       "text/plain": [
-       "CommitInfo(commit_url='https://huggingface.co/datasets/fotiecodes/jarvis-llama2-dataset/commit/70ae877f33efa0eb31663ded334ebab51153309b', commit_message='Upload dataset', commit_description='', oid='70ae877f33efa0eb31663ded334ebab51153309b', pr_url=None, pr_revision=None, pr_num=None)"
+       "CommitInfo(commit_url='https://huggingface.co/datasets/fotiecodes/jarvis-llama2-dataset/commit/8e59d8ebe825853f91f119c45739da5f6bc3d12a', commit_message='Upload dataset', commit_description='', oid='8e59d8ebe825853f91f119c45739da5f6bc3d12a', pr_url=None, pr_revision=None, pr_num=None)"
       ]
      },
-     "execution_count": 7,
+     "execution_count": 38,
      "metadata": {},
      "output_type": "execute_result"
     }

diff --git a/dataset/process_dataset.ipynb b/dataset/process_dataset.ipynb
@@ -9,7 +9,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 1,
+   "execution_count": 5,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -20,7 +20,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 2,
+   "execution_count": 6,
    "metadata": {},
    "outputs": [
     {
@@ -206,7 +206,7 @@
        "29  ### Human: What can you tell me about your cre..."
       ]
      },
-     "execution_count": 2,
+     "execution_count": 6,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -219,7 +219,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 3,
+   "execution_count": 7,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -229,7 +229,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 4,
+   "execution_count": 8,
    "metadata": {},
    "outputs": [
     {
@@ -290,7 +290,7 @@
        "4  ### Human: Introduce yourself. ### Assistant: ..."
       ]
      },
-     "execution_count": 4,
+     "execution_count": 8,
      "metadata": {},
      "output_type": "execute_result"
     }